FORESTRY - Institute of Foresters of Australia

The IFA thanks the following sponsors: 

Gold Sponsors 

Silver Sponsors 

Conference Dinner Sponsor Field Trip Sponsors Exhibitors 

• Queensland Commodity Exports • Healthcare Insurance Limited 

• Griffi th University - Centre for • Narromine Transplants 

Forestry and Horticultural Research • Private Forestry Southern Queensland 

• National Forestry Masters Program 

Advertising Sponsor • Aerial Acquisitions 

• SciWest Consulting • LREye Imagery 

Poster Prize Student Support 

• Gary Bacon • Alan Irving 

Supported through DERM Forest Products, a 

commercial business group of the Department 

of Environment and Resource Management 

PROCEEDINGS OF THE INSTITUTE OF FORESTERS OF AUSTRALIA 2009 CONFERENCE 

IFA 2009 

FORESTRY: A CLIMATE OF CHANGE 

THE INSTITUTE OF FORESTERS OF AUSTRALIA 2009 CONFERENCE 

Proceedings 

6 - 10 September 2009 Caloundra, Queensland, Australia

Forestry: a climate of change 

Pre-Conference Proceedings of the 

Biennial Conference of the 

Institute of Foresters of Australia 

Caloundra, Queensland, Australia 

6 – 10 September 2009 

Edited by: 

Robert Thistlethwaite, David Lamb, Russell Haines 

August 2009

Published by: 

The University of Queensland 

With the assistance of: 

Dr John Herbohn and Annerine Bosch, School of Integrative Systems, The University of 

Queensland, Brisbane, Qld 

Edited by: 

Robert Thistlethwaite, David Lamb, Russell Haines 

National Library of Australia Cataloguing-in-Publication entry: 

Forestry: a climate of change. Proceedings of the Biennial Conference of the Institute of 

Foresters of Australia, Caloundra, Queensland, Australia, 6 – 10 September 2009 

ISBN: 978-0-646-52024-7 

1. Trees-Forest Governance-Congresses. 

2. Sustainable Forestry-Climate Change-Congresses. 

I Thistlethwaite, R. II. Institute of Foresters of Australia 

© Institute of Foresters of Australia 2009 

This publication and its individual papers should be cited as: 

1. General: 

Thistlethwaite, R., Lamb, D and Haines, R. (eds) 2009. Forestry: a climate of change. Proc. 

IFA Conference, Caloundra, Queensland, Australia, 6 – 10 September 2009. 

2. Specific: 

Feary, S., Kanowski, P., Baker, R. and Altman, J. 2009. Managing forest country: Aboriginal 

Australians and forests. In: Forestry: a climate of change, Thistlethwaite, R., Lamb, D. and 

Haines, R. (eds). pp. 422–432. Proc. IFA Conf., Caloundra, Queensland, Australia, 6 – 10 

September 2009. 

Cover design: 

Cameron Coward 

Electronic layout: 

Annerine Bosch, Alison Williams, Robert Thistlethwaite

Forestry: a climate of change 

PRE‐CONFERENCE PROCEEDINGS OF THE BIENNIAL CONFERENCE OF THE 

INSTITUTE OF FORESTERS OF AUSTRALIA 

CALOUNDRA, QUEENSLAND, 6–10 SEPTEMBER 2009 

TABLE OF CONTENTS 

KEYNOTE ADDRESS 

John Innes The role of sustainable forestry in climate change 1‐8 

David Brand The changing role of forests: state of 

development of markets for ecosystem services 

9‐18 

Kathryn Adams Capturing the value of innovation: R&D for 

forest and wood products industries 

19‐27 

SESSION 1 Climate change panel 

SESSION 2 Mitigation opportunities and risks 

Rodney Keenan Native forest management options for climate 

mitigation in Australia and PNG 

28‐35 

Jody Bruce, Michael Battaglia, Australia’s plantation vulnerability to climate 

36 

Libby Pinkard 

change 

Paul Dargusch, Steve Harrison, Environmental markets and Australian forestry: 37‐43 

John Herbohn 

the emergence of a new period of industry 

development 

Scott Arnold Lessons learned from building estate‐scale 

carbon accounting systems 

44‐53 

SESSION 3A Delivery of environmental and ecosystem services 

John Tredinnick and Jodie Carbon stores and bioenergy supplies: 

54‐63 

Mason 

opportunities to develop sustainable plantation 

industries in low rainfall agricultural landscapes 

Himlal Baral, Sabine Kasel, GIS‐based classification, mapping and valuation 64‐71 

Rodney Keenan, Julian Fox, of ecosystem services in production landscapes: 

Nigel Stork 

a case study of the Green Triangle region of 

south‐eastern Australia 

David Lee, John Huth, David Selecting species for carbon sequestration under 72‐82 

Osborne, Bruce Hogg 

climate change scenarios in sub‐tropical 

Queensland 

Frank Batini Responses to a drying climate in the northern 

Jarrah forest 

83‐90 

SESSION 3B Forestry Education 

Lyndall Bull, Peter Kanowski The National Forestry Masters Program: a new 

era of collaboration in professional forestry 

education 

91‐96 

Antoinette Hewitt: Education and training pathways 97 

Chris Weston and Katherine Forestry is a handy word but not the right one for 

Whittaker 

attracting students 

Matthew Doig Training operations staff in the field: challenges 

and opportunities 

98‐102 

103‐110

SESSION 4 Forestry education panel 

SESSION 5A Resource assessment 

Phil West: Practical use of 3P sampling in forest inventory 111‐119 

Andrew Haywood and Michael Estimating forest characteristics in young 

120‐130 

Sutton 

Victorian ash regrowth forests using field plots 

and airborne laser scanning data 

Andrew Haywood and Christine Object‐based analysis of forest stand delineation 131‐141 

Stone 

on high spatial resolution imagery using open 

source software 

Adam Gerrand, Erik Lindquist, The Global Forest Resource Assessment Remote 142‐150 

Mette Wilkie and Rod Keenan Sensing Survey 

Hugh Stone and David Wood Sustainable pole supply project 151‐160 

SESSION 5B Innovation in utilisation 

Russell Washusen, Andrew Processing performance and sawn product 

161‐169 

Morrow, Dung Ngo, Graeme recovery from thinned native forest regrowth logs 

Siemon, Tim Wardlaw, Mike 

Ryan, Martin Linehan and Daniel 

Tuan 

from southern Australia 

Mark Brown Diesel‐electric hybrid drive technology for reduced 

fuel consumption and carbon emissions in forest 

operations 

170‐174 

Russell Wa+shusen Modelling mill door sawlog prices of Australian 

plantation‐grown eucalypts with CSIROMILL 

175‐184 

Mark Brown Reduced fuel use in forestry transportation 

through the use of higher productivity vehicles 

(HPV) 

185‐190 

Martin Strandgard Onboard systems for Australian forest operations 191‐197 

SESSION 5C Wood 

Gregory Nolan Timber from native forest and plantation 

eucalypts ‐ Users will quickly find that they are not 

the same thing 

198‐207 

Kevin Harding, Terry Copley, Wood property variation in mature Queensland 208‐216 

Marks Nester, Kerrie Catchpoole, 

Anton Zbonak 

exotic pine silviculture experiments 

Robert Evans, Josh Bowden, Modelling the influence of climate change on 217‐224 

Kevin Harding, Terry Copley, 

Kerrie Catchpoole, Marks Nester 

plantation wood properties 

Sasha Alexander Competition and collaboration in the Australian 

timber furniture industry ‐ a value chain approach 

toward higher value creation for local and global 

markets and pathways for sustainable forest 

resource use 

225‐233 

Kevin Harding, Terry Copley, Caribbean Pine graded sawn recovery variation in 234‐243 

Marks Nester, Kerrie Catchpoole, a mature spacing by thinning experiment in 

Anton Zbonak 

Queensland 

SESSION 5D Biological performance 

Peter St. Clair Rehabilitation of forests in decline: Mt. Lindsay 

State Forest 

244‐254 

Simon Lawson Using a process‐based model, ParopSys, to predict 

the impact of climate change on the eucalypt 

plantation pest Paropsis atomaria 

255‐262 

Tim Smith and Geoff Pegg Impact of tree boron status on tree growth and 

susceptibility to Quambalaria shoot blight 

(Quambalaria pitereka) in Corymbia sp. 

263‐270

Vivien de Remy de Courcelles, 

Bhupinderpal‐Singh, Mark 

The influence of climate change on soil respiration 

with Eucalyptus saligna 

271‐279 

Adams 

SESSION 5E Fire protection 

Roger Underwood Bushfire governance in Australia in 2009: a note to 

future historians 

280‐284 

Miguel Cruz and Jim Gould National Fire Behaviour Prediction System 285‐291 

Jennifer Hollis and Lachlan Woody fuel consumption and carbon in the 

292‐302 

McCaw 

changing climate of Australia 

Mal Tonkin Victorian Bushfires 2009 ‐ A plantation company’s 

experience 

303‐308 

SESSION 5F Silvicultural aspects 

David Doley Silviculture for a climate of change 309‐320 

Paul Warburton, Paul Macdonell, Improving grey gums to sequester carbon on 321‐329 

Jeremy Brawner, John Huth, 

David Lee 

marginal sites in sub‐tropical Australia 

Jeremy Brawner, David Bush, The utilisation of red mahogany for high value 330‐339 

Paul Macdonell, David Boden, 

Simon Potter, Paul Warburton, 

Paul Clegg 

plantation forestry in the tropics ‐ 

Don Willis Growing teak for the roots – the jiffy solution 340 

SESSION 6A Reduced emission from deforestation and degradation 

Zoe Harkin Lessons learned from implementation of REDD – a 

collection of analogies, metaphors and clichés 

341‐346 

Grahame Applegate Developing a REDD scheme for post 2012: the 

Kalimantan forests and climate partnerships 

347‐354 

Majella Clarke Reducing emissions from deforestation and 

degradation (REDD) in Lao Peoples’ Democratic 

Republic 

355‐364 

Andrew Morton and Blair 

Freeman 

REDD in Asia Pacific: Challenges and implications 365‐373 

Majella Clarke A characterisation of the current market for 

reducing emissions from deforestation and 

degradation (REDD) and forest carbon offsets 

374‐384 

SESSION 6B Changing needs in forest governance and management 

Aidan Flanagan Tasmania’s New Forest Industry Plan: Reshaping 

the forestry agenda 

385‐394 

Sean Ryan Reshaping the forestry agenda in Queensland 395‐402 

David Williams Improving business performance in softwood 

plantations 

403‐412 

Dick McCarthy and Gabriel Samol Addressing climate change impacts on the 

business of PNG forestry: Implications for 

Australia and PNG 

413‐421 

SESSION 6C Indigenous interests and aspirations 

Sue Feary, Peter Kanowski, Managing forest country: engaging indigenous 422‐432 

Richard Baker and Jon Altman communities and the forests sector 

Mark Annandale Community consultation and indigenous forestry 

opportunities in Cape York Peninsula 

433‐440 

Peter Shepherd A Forester’s Guide to working with Indigenous 

Australians: An introduction 

441‐446 

Braden Jenkin Michael Blyth, An integrated approach to species regime 

447‐457 

Peter Kanowski and Don Yakuma selection in PNG: beginning with local people's 

needs

Proceedings of the Biennial Conference of the 

Institute of Foresters of Australia, Caloundra, 2009 1 

THE ROLE OF SUSTAINABLE FORESTRY IN CLIMATE CHANGE 

John Innes 1 

ABSTRACT 

Forests will only help mitigate climate change if they can survive the changes in 

climate. Consequently, any attempts to consider mitigation must also consider 

adaptation. A recent report from the Collaborative Partnership on Forests has examined 

adaptation in detail. It is very difficult, if not impossible, to take a prescriptive approach 

to forest management when considering climate change. Instead, there are a wide range 

of possible actions that could be taken, and the precise combination will depend on the 

local conditions, the anticipated changes in climate, and the objectives for the forest 

(which may themselves have to be changed because of climate change). A key issue is 

that forestry regulations and codes of practice must be sufficiently flexible to allow 

foresters to try different strategies appropriate to their particular situations. 

INTRODUCTION 

For those living at higher latitudes, particularly in the northern hemisphere, there is a great deal of 

evidence that the climate is changing. Meteorological records indicate that temperatures have 

increased significantly, with the most obvious change being a reduction in the coldest extremes during 

winter. Precipitation records are more variable and no generalization is possible. The changes of 

temperature have been accompanied by changes in both the physical and biological phenomena 

present in the region. Glaciers are melting in most parts of the world, permafrost is melting, slope and 

stability has decreased. Tree-lines are moving upwards and northwards (there appears to be no 

documented evidence of southward extensions of tree-lines in the southern hemisphere), and 

aboriginal peoples are reporting that species never previously seen are now occurring. In some cases, 

these are invasive species that are spreading, such as the European starling (Sturnus vulgaris), in other 

cases a northward shift in the distributions of some species is clearly occurring. 

It is natural to assume that these changes are all linked to the phenomenon known as ‘global warming’, 

and the recent reports of the Intergovernmental Panel on Climate Change frequently cite such evidence 

(Intergovernmental Panel on Climate Change 2007), but some argue that they could equally well be 

natural changes following on from the ‘Little Ice Age’ that peaked in the 17 th and 18 th centuries. 

During this period, the Thames River in the UK froze over regularly, something that would be 

considered impossible today. Many glaciers reached their maximum Holocene extent, and many 

marginal agricultural areas had to be abandoned (Grove, 1988). In this paper, I do not wish to address 

the relative merits of the arguments surrounding the causes of today’s changing climate. That is dealt 

with by others at this conference. Instead, I focus on the implications for today’s foresters of a 

changing climate, regardless of cause. 

SOME IMPACTS 

We often speak about climate change as if it is a new phenomenon that will somehow affect us only in 

the future. It is not new: climate change has always occurred, and the implications of the possibility of 

accelerated warming have been discussed in forestry for over 20 years (e.g., Booth and McMurtrie 

1988). However, in the northwest of the North American continent, there is already evidence of 

substantial changes. These include: 

• A major outbreak of the Mountain Pine Beetle (Dendroctonus ponderosae) in central British 

Columbia that has resulted in the death of ca. 14.5 million ha of lodgepole pine (Pinus 

contorta), and the sudden availability of over 620 million m 3 of logs 

1 Professor John L.Innes, Faculty of Forestry, University of British Columbia, 2045-2424 Main Mall, Vancouver, 

Canada V6T 1Z4. Tel: (+1) 604 822 6761. Email: john.innes@ubc.ca.



• Major outbreaks of the spruce beetle (Dendroctonus rufipennis) that has devastated large areas 

of white spruce (Picea glauca) in the Kenai Peninsula of Alaska (600,000 ha) and southwest 

Yukon (390,000 ha) 

• A series of droughts in coastal British Columbia that appears to be causing increased mortality 

in western red cedar (Thuja plicata) 

• Reductions in the length of the winter operations season, when snow and frozen ground makes 

forestry operations feasible and roads are passable. 

The bark beetle outbreaks remain controversial – they can be related to both past management 

activities (or lack thereof) that resulted in the widespread occurrence of mature and old stands of pine 

and spruce. However, it is likely that changes in climatic conditions enabled the epidemics to become 

as severe as they have (Berg et al. 2006). In the case of the mortality, which is evident in a number of 

species in western North America, the data are quite limited, and the precise role of climate change is 

difficult to determine (van Mantgem et al. 2009). 

FUTURE CHANGES 

If model predictions are correct, it is very likely that many parts of the Earth’s surface will experience 

significant changes in climate over the next 100 years. It is currently not possible to specify what these 

changes will be, as they will depend to a large extent on the manner in which the world develops and 

responds to climate change over the same period. As a result, modellers have developed a number of 

scenarios based on different development and mitigation options. However, there is evidence that 

already carbon dioxide emissions are exceeding even the most pessimistic of these options. The choice 

of scenario is therefore a cause of considerable uncertainty. 

The predicted emissions of carbon dioxide and other greenhouse gases are fed into Global Change 

Models. There are a number of these, and the results they produce differ. Early models were very 

crude, but they have gradually become more and more sophisticated and better at reproducing the 

climate change that has occurred over the past 100 years. The predicted future climate for a given 

scenario does however vary depending on the choice of model. 

Global Change Models, by their very nature, work at coarse global scales, and cannot take into 

account fine-scale topography. Consequently, they are likely to be very poor in predicting climate 

change in mountainous areas. Also, if there are any local effects, such as land and sea breezes or 

anabatic and katabatic winds, these are unlikely to be accounted for. Consequently, downscaling is 

necessary, whereby the output from the global models is adjusted for local conditions. This is a further 

source of uncertainty. 

The net effect of these three major sources of uncertainty is that it is not possible to predict with any 

degree of confidence the future climate at a specific place. Yet such predictions are precisely what the 

forester, thinking of the next 100 years or so, needs. Much the same problem is faced by the engineer: 

structures being designed today need to be able to cope with climate change while remaining 

economically feasible (Hallegatte 2009). 

MITIGATION VS. ADAPTATION 

Given such uncertainty, what should a forester do? There are two major strategies that have been 

discussed in detail: mitigation and adaptation. Mitigation presupposes a causal link between 

greenhouse gas emissions and climate change and involves efforts by foresters to reduce the amount of 

greenhouse gases, particularly carbon dioxide, in the atmosphere through selected forestry activities. It 

includes increased sequestration, through an increase in global forest cover and the increased use of 

wood products in long-lived products such as houses and furniture, reduced emissions from forests 

through reductions in deforestation and forest degradation and reductions in the carbon dioxide 

released by forest fires, and substitution of substances associated with greenhouse gas emissions, such 

as coal, concrete and steel, by wood and wood composites.



The other major strategy is adaptation. This presumes that climate change is inevitable (regardless of 

cause), and involves the adaptation of forest practices to enable forests and forestry to cope with these 

changes. Clearly, this is potentially problematic, as we do not know the nature or extent of future 

changes in climate. Consequently, we need to identify changes in management strategy that will be 

beneficial for forests and forestry whatever the future changes in climate are likely to be, or even in 

the unlikely event that there is no change. 

Is there such a “win-win” strategy? Fortunately there is, in the form of sustainable forest management. 

Many of the criteria associated with sustainable forest management, together with their associated 

indicators, could reduce the vulnerability of forests to climate change. There are a number of different 

definitions of sustainable forest management and also a number of criteria and indicator schemes. 

Here, I focus on the Montreal Process, as this is the scheme used in Australia, and is also used for 

more than 85% of the temperate and boreal forest area. 

THE MONTREAL PROCESS CRITERIA 

The Montreal Process contains seven criteria, all of which are relevant to climate change. However, 

climate change is creating some questions about the continued relevance of current interpretations of 

the criteria. For example, ‘Maintain biodiversity’ is an admirable concept in a static environment, but 

may not be so applicable in the context of climate change. Within the context of climate change, this 

criterion needs to be interpreted as maintaining biodiversity appropriate for the site conditions. The 

emphasis then is on enabling the forest to adapt as climate changes, rather than preventing any 

changes that may represent a part of that adaptation. This will require some flexibility on the part of 

managers, as highly desirable tree species may be replaced by less desirable but better adapted tree 

species. 

Having the flexibility to make such decisions is critical. Forestry institutions, broadly defined, include 

the governance structures, formal and informal education systems, professional bodies and other actors 

involved in forests. Many of these are traditionally very conservative, with the result that adaptation to 

the rapid changes in forests induced by climate may be difficult. For example, in some jurisdictions 

(particularly in developed countries), governance systems are highly prescriptive, with detailed 

regulations on exactly how all aspects of forestry are to be undertaken. These may be amongst the 

most inflexible to adaptation. Conversely, where a results-based approach to forestry has been 

adopted, the focus is on particular management objectives, with the management options being taken 

to achieve these objectives being largely left to the manager to decide. Such systems will be much 

more flexible. 

Table 1 (appended) reveals some of the issues that may be involved, including both direct and indirect 

effects. For example, the inadequate communication between forest scientists and policy makers is 

partly a result of the rapidity of the changes in knowledge, but also a function of the lack of 

appropriate institutions and the early failures of the forestry community to be actively involved with 

policy makers in this area. In addition, it should be noted that in many cases the institutional 

framework in other sectors may have a greater impact on the ability of the forest sector to adapt to 

climate change than that of the forest sector itself. Promotion of biofuel production from agricultural 

products, for example, already has increased forest conversion into oil-palm plantations in Asia (Koh 

and Wilcove 2007, 2008). Free trade agreements may also affect directly or indirectly the local 

capacity to deal with climate change, in some cases with positive effects, in others not. 

STRATEGIES FOR FORESTERS 

Every forest manager (or forest policy maker) faces a unique set of problems, before climate change is 

even considered. As a result, the actions taken to adapt to climate change will vary from place to 

place, and no generalizations are possible. In fact, prescriptive approaches to climate change would be 

highly misleading, and most likely erroneous. While it is tempting to provide broad recommendations 

applicable to, for example, boreal forests, this would be disingenuous. The managers of forests in 

Alaska, Canada, Fennoscandia and Russia have very different objectives. Even within these units, 

there may be major differences, such as between the manager of a forest biological reserve and that of 

a forest biomass plantation. Managers must be given the freedom to choose which management



options are the most appropriate for their particular piece of land. Policy makers must ensure that 

appropriate mechanisms exist to enable this flexibility. This suggests strongly that forest policy must 

move away from prescriptive approaches to results-based approaches, allowing managers the freedom 

to adopt the best way to achieve those results, given local circumstances. 

The majority of tropical production forests are not being managed (sustainably), almost 20 years after 

ITTO defined the needs of sustainable forest management (e.g., International Tropical Timber 

Organization 1992, 1993). From the perspective of mitigation through reductions in deforestation and 

forest degradation, this represents a major challenge for the global forest policy community. 

Despite years of negotiation, the poor uptake of sustainable forest management has not been 

satisfactorily addressed by groups such as the United Nations Forum on Forests (Persson 2005, 

Humphreys 2006, Capistrano et al. 2007). Similarly, illegal logging remains rife, contributing to 

emissions from deforestation and forest degradation (Tacconi 2007). There is a need for effective 

policy initiatives to address these issues. 

While recommendations for strategies such as improved fire-fighting capabilities or the ability to 

combat insect infestations are possible, such actions would be appropriate in some circumstances, but 

not others. Some areas might indeed face such problems, but others will not. Consequently, general 

prescriptions run the risk of encouraging scarce resources to be allocated inappropriately. 

The key message is that although every situation is different, there are both procedures and 

management options that will help forest managers adapt to climate change. To date, few actions have 

been taken, and there is a major need to educate foresters of both the need to adapt to climate change 

and the options that are available to them. They need to be given the freedom to choose the options 

that are most applicable to their particular situations. However, there is a need to recognize that the 

definition of sustainable forest management is continuously evolving, and managers must be prepared 

to address these “shifting goal posts”. 

There is an urgent need to ensure that forests utilized for timber exploitation are managed sustainably. 

This is a task for the international forest policy community, but their actions to date suggest that other 

mechanisms may need to be involved. To a certain extent, this is already occurring. The current 

international interest in illegal logging and REDD occurred despite, rather than because of, the forest 

policy community. 

The Intergovernmental Panel on Climate Change was largely independent of the forest policy 

community, as was the Stern report (Stern 2006), but their reports contain many references to forests 

and their adaptation to climate change. There is a real danger of the forest policy community 

becoming increasingly distanced from the emerging mainstream forest policy issues, but climate 

change has provided an impetus that is now driving the forestry agenda. 

The feasibility and/or effectiveness of an adaptation practice will depend on the characteristics of the 

forested system, such as its soil, existing biodiversity, pre-existing natural and anthropogenic stresses, 

the status of the local ecosystem (e.g., early to late successional), and on the characteristics of the 

social, economic, and political system interacting with, or dependent on, the ecosystem. 

Increasing the adaptive capacity of stakeholders may involve improving access to appropriate 

information and technologies to address the impacts and opportunities of the changing climate, 

strengthening social networks and traditional forest management and governance institutions, 

increasing food security, alleviating poverty, and policy changes to expand or enhance current 

management opportunities. For example, existing management practices may need to be applied in 

new and different situations (such as prescribed fire outside of the ‘traditional’ fire management 

seasons). 

Understanding the opportunities and barriers that affect successful implementation of management 

practices to enhance adaptive capacity is a critical need. Enhancements or barriers to adaptation may 

include biological, physical, economic, social, cultural, institutional, or technological conditions. As 

such, it will be necessary to adopt cross-sectoral approaches to the issues.



CONCLUSIONS 

Climate change poses another set of uncertainties for foresters trying to plan over the medium- to 

long-term. These uncertainties should not be taken an as excuse for taking no action. Instead, foresters 

should choose those actions that are most likely to benefit sustainable forest management while at the 

same time maintaining the ability of the forest to adapt to climate change. No specific prescriptions are 

possible, partly because of the uniqueness of every situation, and partly because the future changes in 

climate remain unknown. 

Forestry is increasingly regulated, especially in a country such as Australia. It is important that these 

regulations do not hinder foresters in managing their forests to adapt to climate change. If the 

regulations are too prescriptive, then the ability of foresters to allow their forests and forestry practices 

to adapt to climate change may be compromised. At the same time, many groups see highly 

prescriptive regulations as the only way to control forestry. An alternative however is a results-based 

approach, where there is general agreement on the endpoint being sought and foresters are allowed the 

flexibility to try different paths to that point. This, however, requires a well-educated forestry 

workforce that is able to recognize the potential implications not only of their practices, but of the 

changes in the environment (specifically the climate) that are occurring around them. 

REFERENCES 

Berg, E.E, Henry, J.D., Christopher, L., Fastie, C.L, Volder, A.D. and Matsuoka, S.M. 2006. Spruce beetle 

outbreaks on the Kenai Peninsula, Alaska, and Kluane National Park and Reserve, Yukon Territory: 

Relationship to summer temperatures and regional differences in disturbance regimes. Forest Ecology 

and Management 227(3) 219–232 

Booth, T.H. and McMurtrie, R.E. 1988. Climate change and Pinus radiata plantations in Australia. In: Pearman, 

G.I. (ed.) Greenhouse: Planning for climate change. CSIRO, Melbourne, pp. 534-545. 

British Columbia Ministry of Forests and Range (BCMOF) 2006. Preparing for climate change: Adapting to 

impacts on British Columbia’s forest and range resources. Province British Columbia, Victoria, British 

Columbia, Canada. 80pp. 

Capistrano, D., Kanninen, M., Guariguata, M.R., Barr, C., Sunderland, T.C.H., Raitzer, D.A. 2007. Revitalizing 

the United Nations Forum on Forests: critical issues and ways forward. Centre for International Forestry 

Research, Bogor, Indonesia. 

Chapin, F.S. III, Peterson, G., Berkes, F., Callaghan, T.V., Angelstam, P., Apps, M., Beier, C., Bergeron, Y., 

Crépin, A.-S., Danell, K., Elmqvist, T., Folke, C., Forbes, B., Fresco, N., Juday, G., Niemelä, J., 

Shvidenko, A. & Whiteman, G. 2004. Resilience and vulnerability of northern regions to social and 

environmental change. Ambio 33(6): 344–349. 

FAO 2008. Conference Report for “Adaptation of forests and forest management to changing climate with 

emphasis on forest health: A review of science, policies and practices. Umeå, Sweden, 25–28 August 

2008. UN Food and Agriculture Organization, Rome, Italy. 

Grove, J.M. 1988. The Little Ice Age. Routledge. London and New York, 498 p. 

Hallegatte, S. 2009. Strategies to adapt to an uncertain climate change. Global Environmental Change – Human 

and Policy Dimensions 19 (2): 240 – 247. 

Humphreys, D. 2006. Logjam. Deforestation and the crisis in global governance. Earthscan Publications, 

London, UK. 

Intergovernmental Panel on Climate Change. 2007. Impacts, adaptation and vulnerability: Contribution of 

Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. 

Parry, M.L., Canziani, O.F., Palutikof, J.P., van der Linden, P.J. and Hanson, C.E. (eds.). Cambridge 

University Press, Cambridge, UK. p. 976. 

International Tropical Timber Organization 1992. ITTO guidelines for the sustainable management of natural 

tropical forests. International Tropical Timber Organization, Yokohama, Japan. p. 18. 

International Tropical Timber Organization 1993. ITTO guidelines for the establishment and sustainable 

management of planted tropical forests. International Tropical Timber Organization, Yokohama, Japan. 

p. 39. 

International Tropical Timber Organization 2006. Status of tropical forest management 2005. ITTO Technical 

Series No. 24. International Tropical Timber Organization, Yokohama, Japan. p. 302.



Johnston, M., Williamson, T., Price, D., Wellstead, A., Gray, P., Scott, D., Askew, S. & Webber, S. 2006. 

Adapting forest management to the impacts of climate change in Canada. BIOCAP Research 

Integration Program Synthesis Paper. 96 pp. 

Kellomaki, S., Peltola, H., Bauwens, B., Dekker, M., Mohreh, F., Badeck, F.W., Gracia, C., Sanchez, A., Pla, E., 

Sabate, S., Lindner, M. & Pussinen, A. 2005. European Mitigation and Adaptation Potentials: 

Conclusions and Recommendations. In: Kellomaki, S. & Leinonen, S. (eds). SilviStrat: Management of 

European Forests under changing climatic conditions. Tiedonantoja Research Notes No. 163. Faculty of 

Forestry, University of Joensuu, Joensuu, Finland. 427 pp. 

Koh, L.P., Wilcove, D.S. 2007. Cashing in palm oil for conservation. Nature 448, 993-994. 

Koh, L.P., Wilcove, D.S. 2008. Is oil palm agriculture really destroying tropical biodiversity? Conservation 

Letters 1 (2), 60-64. 

Lemmen, D. & Warren, F. (eds.). 2004. Climate change impacts and adaptation: A Canadian perspective. 

Climate Change Impacts and Adaptation Directorate, Natural Resources Canada. Ottawa, Ontario. 

Ogden, A.E. & Innes, J.L. 2007. Incorporating climate change adaptation considerations into forest management 

planning in the boreal forest. International Forestry Review 9: 713–733. 

Persson, R. 2005. Where is the United Nations Forum on Forests going? International Forestry review 7 (4), 348- 

357. 

Spittlehouse, D.L. 2005. Integrating climate change adaptation into forest management. Forestry Chronicle 

81(5): 691–695. 

Spittlehouse, D.L. & Stewart, R.B. 2003. Adaptation to climate change in forest management. BC Journal of 

Ecosystems and Management 4(1): 1–11. 

Stern, N. 2006. The economics of climate change. The Stern Review. Cambridge University Press, Cambridge. 

p. 692. 

Tacconi, L. (ed.) 2007. Illegal logging. Law enforcement, livelihoods and the timber trade. Eathscan 

Publications, London, UK. 

van Mantgem, P.J., Stephenson, N.L., Byrne, J.C., Daniels, L.D., Franklin, J.F., Fule, P.Z., Harmon, M.E., 

Larson, A.J., Smith, J.M., Taylor, A. H. and Veblen, T.T. 2009. Widespread increase of tree mortality 

rates in the western United States. Science 323 (5913): 521-524.



Table 1. Strategic (S) and operational (O) climate change adaptation options that may be considered 

to achieve the management objective of ensuring the appropriate legal, institutional and economic 

framework is in place for forest conservation and sustainable management. Table adapted from Ogden 

and Innes (2007). Sources: BCMOF (2006), Chapin et al. (2004), FAO (2008), Johnston et al. (2006), 

Kellomaki et al. (2005), Lemmen and Warren (2004), Spittlehouse (2005), Spittlehouse and Stewart 

(2003). 

Impact S/O Adaptation Options 

Forest 

S Development of flexible forest management plans and policies that 

management plans incorporate uncertainty and are capable of responding to climate 

and policies lack 

change. 

the flexibility that Maintain an overall management process where targets and results 

is required to 

and the effectiveness of chosen options are revisited and assessed 

respond to climate 

change 

periodically against emerging issues and changing environmental 

conditions 

Provide long term tenures to encourage long term considerations 

within short term decisions 

Relax rules governing the movement of seed stocks from one area to 

another 

Measure, monitor and report on indicators of climate change and 

sustainable forest management to determine the state of the forest 

and identify when critical thresholds are reached 

Evaluate the adequacy of existing environmental and biological 

monitoring networks for tracking the impacts of climate change on 

forest ecosystems 

Practice adaptive management. 

O Conduct a priori vulnerability assessments 

Conduct advanced planning of salvage logging procedures 

Train forest managers to deal with uncertainty and to manage risk 

effectively 

Implement practices that reduce vulnerability and increase resilience 

even at the expense of growth 

Forest 

S Development of forest management plans that reduce vulnerability 

management plans of forest and forest dependent communities to climate change 

and policies 

enhance the 

vulnerability of 

forests and forest 

dependent 

communities to 

Assess vulnerability of forests, forest dependent communities and 

society in general 

Support research on climate change, climate impacts, and climate 

change adaptations and increase resources for basis climate change 

impacts and adaptation science 

climate change 

Support knowledge exchange, technology transfer, capacity building 

and information sharing on climate change; maintain or improve 

capacity for communications and networking 

Incorporate new knowledge about the future climate and forest 

vulnerability into forest management plans and policies 

Involve the public in an assessment of forest management adaptation 

options 

O Diversify forest-based and non-forest based income sources 

Increase local governance of local forest resources 

Increase general capacity for the detection and management of 

climate change impacts



Forest 

management 

policies and 

incentives do not 

encourage 

adaptation to 


Forest institutions 

are unable to cope 

with the needs of 


Inadequate 

communication 

between forest 

scientists and 

S 

S 

Remove barriers and develop incentives to adapt to climate change. 

Provide incentives and remove barriers to enhancing carbon sinks 

and reducing greenhouse gas emissions 

Provide opportunities for forest management activities to be 

included in carbon trading systems (as outlined in Article 3.4 of the 

Kyoto Protocol) 

Increase awareness of the implications of climate change to 

encourage behavioural change that will facilitate planned adaptation 

Encouragement local, regional and national forest management 

policies and plans to consider their role globally (think global, act 

local) 

Establish new or enhance existing institutions and infrastructure for 

proactive or proactive adaptation 

Develop multiple partnerships and break down institutional barriers 

to better enable the recovery from climate-mediated crises 

Develop mechanisms to encourage multilateral solutions to the 

problems brought about by climate change 

S Develop mechanisms to enable research results to be shared and 

communicated to decision- and policy-makers 

Elevate the importance of the social sciences in the adaptation of 

communities, land-uses and forest management to climate change 

policy makers Raise the awareness of the research community to the type of 

information needed by decision- and policy makers



THE CHANGING ROLE OF FORESTS: STATE OF DEVELOPMENT 

OF MARKETS FOR ECOSYSTEM SERVICES 

David Brand 1 

INTRODUCTION 

For over twenty years there have been concerted efforts to reduce the loss and degradation of forests. 

While some success has been seen in countries like Australia where regulation of land clearing has 

been effective in reducing conversion rates (WWF, 2009), in other regions, particularly tropical 

forests, reduction in forest cover has continued (FAO, 2005). Much of the systematic degradation of 

ecosystems, including forests, has been driven by the expansion of agriculture, and to a lesser extent 

by human settlement, mining, oil and gas and infrastructure (Millennium Ecosystem Assessment, 

2005). Ultimately the process of ecosystem degradation is driven by the size and expansion of the 

global economy, which is driven by human population and per capita economic growth. 

We now have a population of 6.7 billion, likely to rise to 10.5 billion before peaking (UN DESA, 

2009). The global economy is approximately $62.2 trillion (CIA, 2009), and growing on average 

between 2% and 3% per annum. This means that the global economy is likely to double every 24 to 

36 years. Many forms of consumption have been outstripped economic growth. For example meat 

production, energy consumption, and paper usage have all grown at or faster than economic growth 

(FAOSTAT, 2009). Growing consumption impacts ecosystems both directly and indirectly. 

Conversion of Brazilian rainforest to cattle grazing, or of Indonesian rainforest to palm oil, has direct 

and long-lasting consequences. Excessive pollution and climate change have more chronic impacts on 

ecosystems via potential changes to disturbance factors like wildfire and insect epidemic, and shifts in 

species distributions. 

In fact since the Earth Summit in 1992 we have been able to foresee that the global system would be 

systematically changed by consumption and that the three great challenges would be related to soil 

conservation and access to freshwater, biodiversity conservation and global climate change. Forest 

conservation fundamentally intersects these three challenges—forests are one of the key elements of 

the global carbon cycle (Malhi, Y et al, 2002), regulate soil conservation and water quality, and 

provide the basis for something like 50% of the diversity of life on earth (Millennium Ecosystem 

Assessment, 2005). As our population grows and the economy becomes larger, there is a conundrum. 

On the one hand we know that these ecosystems provide services such as water purification, carbon 

storage and soil conservation, but on the other we continue to convert these ecosystems to production 

systems to support ever-increasing demand for food, fiber, human settlement and infrastructure. 

No one would realistically expect that we might end up with a planet that has no natural ecosystems 

left. But how do we diverge from the pattern of economic growth linked to conversion of ecosystems 

from conservation functions to production functions? The problem is that production systems 

generate profit and attract capital for development, while conservation functions do not. Effectively, 

nature is unpriced, and therefore is used wastefully as a low cost input to production. In our present 

economic system it is more cost effective to convert more land to production than it is to intensify 

production on the existing landbase. It is easy to argue that this represents a market failure. 

There have been many attempts to use planning, policies, aid programs and philanthropy to halt the 

degradation and loss of forests (e.g. the Forest Principles developed at the 1992 Earth Summit and the 

FAO Tropical Forest Action Plan which was developed in 1985). They have largely failed, because 

they don’t change the underlying price signals and economic drivers. As a result there is now a 

rapidly escalating effort to create price signals for ecosystems and for ecosystem services as a way of 

actually making conservation functions valuable (Ecosystem Marketplace, 2009). Once something 

1 

Dr David Brand, Managing Director, New Forests Pty Limited, The Zenith Centre, Level 19 Suite 1905, 821 Pacific Hwy, 

Chatswood NSW 2067, Australia.



becomes valuable, it can attract capital and will become more costly to use as an input to production. 

This paper aims to discuss how these markets for ecosystem services are emerging, and how they 

might contribute to a kind of global end-game where production and conservation functions are 

stabilized, and human society effectively makes an accommodation with nature. 

THE CONCEPT OF MARKETS FOR ECOSYSTEM SERVICES 

The question is how to introduce these price signals into an economic system that has benefited from 

nature being unpriced. In some ways it is not as difficult as it may seem. Markets have long 

understood the concept of auctioning fishing quotas, timber rights, grazing rights and irrigation water. 

It is only a step more to commodify carbon storage, watershed or catchment management and species 

conservation. The challenge in any emerging market, however, is to create both supply and demand. 

The demand side of ecosystem services has been lacking. 

There are a growing range of examples of markets for ecosystem services, and while these remain 

small they are expanding rapidly. In these markets demand is generally created by either voluntary 

action or by government regulation. Voluntary action examples might include the sale of carbon 

credits from Australian re-vegetation projects to automobile owners by Greenfleet (Greenfleet, 2009), 

payment to landowners by Coca-Cola for clean water (Clinton Global Initiative, 2008), and 

sponsorship of duck habitat conservation by duck hunters (Ducks Unlimited, 2009). 

Yet voluntary funding is invariably a marginal act, and does not transform price signals. That only 

occurs through regulation. During the 1970’s the US Government passed the Clean Water Act (US 

Environmental Protection Agency, 2008) and the Endangered Species Act (United States Fish and 

Wildlife Service, 2009). These two acts provide an interesting case study of the establishment of price 

signals for ecosystems. In each case the Act created a no net loss philosophy for wetlands and 

endangered species respectively. Over time regulatory instruments emerged to allow for wetlands 

mitigation banking and endangered species banking. These were market-based instruments. For 

example if I am a property developer wanting to convert habitat for an endangered species, I will need 

to pay someone else to enhance the conservation or expand the habitat for the same species elsewhere. 

These markets often worked by creating large areas of functional habitat, or by rehabilitating degraded 

areas into productive wetlands. In many cases the ‘banks’ created to sell wetland or endangered 

species conservation became highly valuable. Many developers found that their properties were more 

valuable if managed for conservation rather than property development. 

The mitigation banking industry in the US is now a well-regulated sector with hundreds of millions of 

dollars in sales annually (speciesbanking.com, 2008). The sector has small and medium sized 

businesses, increasingly attracts private equity funding, and there is a growing sophistication in the 

industry. In fact it could be seen as a model for how an ecosystem market can unfold. 

The other evolving market that came from the US was the Clean Air Act (US Environmental 

Protection Agency, 2009) and its creation of a market for reductions in sulfur dioxide emissions 

causing acid rain. The regulations set a cap on emissions and allowed polluters to trade the allowances 

necessary to comply with the cap. This meant that a company who could reduce their emissions more 

inexpensively would reduce more than required and sell the excess reductions to companies with a 

higher cost of compliance. This market based approach proved highly successful and the cost of 

meeting the targeted reductions was far lower than had been predicted by economic models (Gutt, E. 

et.al, 2000). 

The US government also promoted the use of markets to address greenhouse gas reductions in the 

1990’s. While the US ultimately pulled out of the Kyoto Protocol, the use of flexibility mechanisms 

like carbon trading was a legacy of US negotiating strategies. The US and other countries also 

promoted the incorporation of forest conservation and reforestation into the global carbon market as an 

offset (or carbon credit against emissions). While this ultimately was curtailed in the Kyoto Protocol 

and rejected for incorporation to the European Union Emissions Trading Scheme (European Union



Emissions Trading Scheme, 2009) forestry has now begun to see a resurgence of interest in non- 

European carbon markets such as the Australian Carbon Pollution Reduction Scheme (Department of 

Climate Change, 2009) , the NZ Emissions Trading Scheme (Ministry for the Environment, 2009), the 

California Forestry Protocols (Climate Action Reserve, 2009) and most recently and significant the US 

Federal Waxman-Markey Bill (Committee on Energy and Commerce, 2009). Significant innovation is 

occurring related to new concepts like Reductions in Emissions from Deforestation and Forest 

Degradation (“REDD”) (VCS Association, 2008 and Avoided Deforestation Partners, 2008) that could 

create whole new markets for forest conservation. 

So there are functional examples of regulatory systems across carbon, water and biodiversity 

attributes. The challenge is to make these instruments ubiquitous in the global economy, and subject 

to a comprehensive no net loss policy at national and international levels. This will not be easy, but 

the ultimate solution may well be found in the problem itself—effectively embedding ecosystem 

services into the supply chains to consumers. 

CONSERVATION AND PRODUCTION – LANDSCAPES AND HUMAN SOCIETY 

Several recent studies confirm the degree to which human society and its demand for goods and 

services now dominates nature (see for example the Millennium Ecosystem Assessment or the WWF 

Living Planet Report). Economic growth is inexorable, but ecosystems are finite. Therefore as the 

economy grows we have seen a steady erosion of natural systems resulting in a series of negative 

consequences. Loss of ecosystem function is linked with soil erosion, loss of productivity, the spread 

of weeds, feral animals and disease, increased flooding, nutrient leaching, hypoxic zones, and a host of 

other outcomes. It is as if the demand for goods is now seriously impeding the capacity to maintain 

ecosystem services, and some form of balance must be sought. The lack of pricing for ecosystem 

services means that there is no optimal outcome where the marginal benefit of increased production is 

balanced against the marginal cost of lost ecosystem services. The market based pricing of goods will 

dominate the unpriced public services of ecosystems even if there is a huge negative net cost or 

‘externality’ to society. So the answer has to be to price ecosystems and their services. 

The approach that has the most merit is to regulate those industries that either impact or benefit from 

water quality or biodiversity, or which have greenhouse gas emissions, such that they have to avoid, 

reduce and ultimately mitigate their environmental impacts. In parallel we need to commoditize the 

ecosystem services so that they can be conserved and enhanced by private capital flows the same way 

that oil palm or beef is produced. The production functions then will need to acquire ecosystem 

services credits in order to continue to operate or expand operations. This embeds the conservation 

functions in the production functions and consumption ultimately must pay for the costs of 

maintaining ecosystem services. As the global economy grows and demand rises, ecosystems should 

become steadily more valuable. This will lead to greater emphasis on resource use efficiency, 

production efficiency, recycling, etc. Effectively there should be some kind of economic balance 

where the cost of further impacts to ecosystems becomes too expensive to contemplate. 

The supply side instruments are starting to become clear. The first is the emerging forest carbon 

market. While the Kyoto protocol has provided limited innovation around forest carbon products, 

there has been substantial new thinking coming from the United States and the voluntary carbon 

market (see AD Partners website cited above). 

Standards for creating carbon conservation value related to reforestation, improved forest management 

and forest conservation are all emerging into a workable framework. The US Waxman Markey bill 

was a key step forward and comprehensively integrated both domestic and international forest carbon 

management into a legislative framework. The work undertaken by US legislators is likely to support 

a strong US position on the incorporation of forest carbon instruments into the negotiations at COP15 

on a successor international agreement to the Kyoto Protocol in 2013.



The second key instrument is a biodiversity product. Unlike carbon, biodiversity is not easily 

measurable or commodifiable. Biodiversity is also a surrogate itself for a suite of ecosystem services 

related to purification of water, pollination, disease control, and general maintenance of productivity 

and natural function. It is hard to comprehend a completely homogenized world with complex natural 

systems largely gone, and human managed exotic production systems completely dominant. It is 

probably hard to think through what such a world might look like, but there are certainly regions of the 

earth already well down the path of complete homogenization. 

The mitigation banking industry in the US is probably the most likely model for a global biodiversity 

conservation product. The approach is based on institutionalizing a philosophy of no net loss and 

then, based on assessments, allowing development activities to mitigate their impact by buying 

‘credits’ from mitigation bankers. The ultimate expression of this approach could be large scale 

protected areas funded by mitigation banking fees. New Forests has recently established the Malua 

Biobank (MWHCB Inc, 2009) in Sabah, Malaysia, that is trying to experiment with exporting the 

mitigation banking approach to conserving tropical rainforests with a globally significant suite of plant 

and animal species. In this case serialized Biodiversity Conservation Certificates are created and 

listed on the TZ1 exchange (TZ1 Market, 2009). Each certificate represents the rehabilitation and 

conservation management of 100 square metres of dipterocarp forest. Oil palm plantations can buy 

these certificates and attach them to their exports of crude palm oil. In fact the product works very 

well with the palm oil supply chain. Each hectare of palm oil plantation produces about 100 tonnes of 

crude palm oil during its 25 year life. If a biodiversity certificate is attached to each tonne produced, 

the hectare of palm oil plantation effectively sponsors the rehabilitation and conservation management 

of one hectare of the biobank. This makes the palm oil producer the sponsor of biodiversity 

conservation rather than the cause of its depletion. 

The third key instrument is a water quality related product. The impacts of production activities on 

water quality include soil erosion and resultant turbidity, increases in dissolved nutrients, particularly 

nitrates and phosphates, and runoff of pesticides. These impacts affect water quality directly for 

downstream human use, but also affect aquatic system health via impacts on benthic fauna, fish, bioaccumulation 

of pesticides, algal blooms, hypoxia, and coral bleaching. There are also specific 

regional impacts on water quality from temperature changes, salinisation, acid sulphate soils and shifts 

in pH. 

Like the biodiversity related instruments, water quality problems are often localized and need to be 

regulated at the level of catchments or watersheds. In the US, again, there have been a number of 

experiments and pilot programs related to water quality trading (Environmental Trading Network, 

2009). In general the nutrient trading regimes have been designed so that point source polluters like 

sewage treatment plants could buy offsets after achieving a certain level of in-house pollution control. 

This reflects the reality that the last 2 or 3 percent reductions may have extremely high costs, and 

paying farmers to fence riparian zones, or use precision fertilization practices may deliver additional 

reductions in nutrient loss at a far lower cost. 

A major water quality initiative is currently being undertaken in the Australian state of Queensland, 

where catchments draining into the Great Barrier Reef have been put under a stringent continuous 

improvement and monitoring system. All land management practices have been benchmarked for 

their water quality impacts, and payment schemes have been instituted to support landowners in 

making a transition to lower impact land management practices. The catchments have been modeled 

and the level of nutrients and other pollutants can be forecast under different levels of program takeup. 

The specific goal of the Australian Government is to reduce nutrient loads by 50% in the Great 

Barrier Reef Lagoon. The Government is paying approximately $AU200 million over four years to 

landowners under the Reef Rescue Program (Caring for our Country, 2009). The question is what will 

happen as Government funding comes to an end. The Government funding was a by-product of 

selling the national telephone company. Further initiatives would need to come from tax revenues, 

which is unlikely given the current budget deficit.



New Forests has put forward a concept for a water quality bank for the Great Barrier Reef Catchments 

that will perpetuate the price signals for sustainable land management via a Nutrient Bank. The 

concept is based on setting a baseline of current loads of nutrients, sediment and chemicals and 

creating a kind of water pollution unit. Monthly water quality monitoring at key points can be used to 

build an annual water quality rating. If the water quality rating is below the baseline, then there are 

credits created. These credits would be listed and saleable. The revenue gained from selling these 

credits would be used for three purposes—to pay landowners an upfront price for improved land 

management, to generate a return to private investors in the bank and to pay dividends to landowners 

who maintain their properties under higher levels of land management practice. Downstream 

beneficiaries of the improved water quality like tourism operators could be buyers or the credits could 

be embedded in the agricultural commodities, principally sugar. The government could also be a 

buyer, at least initially to act on behalf of the general public good. 

WHAT IS THE END-GAME? 

There have been continuing attempts to set targets and policy mechanisms for conservation of 

ecosystems. The global goal of 10% protected areas in each bioregion is a major example, supported 

by the United Nations Convention on Biological Diversity, the IUCN and others. Conservation 

International has embraced the concept of protecting key ‘hotspots’ which are areas of unique 

endemism and biodiversity. The concept of protected areas works reasonably well in developed 

countries with substantial resources, but even there the protected areas tend to be over-represented in 

non-productive ecosystems (alpine, mountainous terrain, arctic regions, deserts, etc.) and underrepresented 

in productive systems used for agriculture, grazing and human settlement. 

Given that most productive areas are under private ownership or management, it proves expensive and 

controversial for governments to use tax-payers dollars to acquire these areas for rehabilitation or 

conservation. Some NGO’s have sought to augment formally protected areas in Government 

ownership with informal conservation areas often established by the purchase of conservation 

easements or development rights from property owners. These philanthropic conservation funds have 

made substantial contribution to protected areas networks, but the general view is that they are not 

capable of out-competing private investment in development activities. A recent study by the Union 

of Concerned Scientists and the UK government suggested that funding of $US20 to $US33 billion 

per annum would be needed to reduce deforestation by 50% over the next ten to twenty years. That is 

250% to 400% of the current combined funding of the World Bank, Overseas Development Assistance 

and Philanthropy. If we wish to completely stop all further rainforest conversion we would potentially 

be looking at funding of $50 billion per annum or more. 

Another way to look at the problem of forest conservation finance is to explore the goals for emissions 

reduction in 2030 and work backwards. It has been suggested that global agreements should aim for 

80% reductions in greenhouse gas emissions by 2050 and that this will require approximately 30 

billion tonnes per annum of greenhouse gas reductions relative to business as usual by 2030. Analysis 

of that target suggests that at least 20% of the net emission reduction will need to come from rainforest 

conservation by 2030. That is 5 billion tonnes of carbon dioxide equivalent reductions per annum. At 

a global carbon price of $US20 that represents a revenue stream of $US100 billion per annum. If we 

securitized such a cash flow at a 10% real cash yield, it would represent an asset value of the world’s 

rainforests of $US1 trillion. That is approximately equal to 5% of the institutional financial assets in 

the United States, or approximately equal to the entire assets of the Australian Superannuation 

industry. 

On the other hand if we compare this with the value of agricultural commodities it is less daunting. 

For example, Malaysia exports around 15 million tonnes of crude palm oil for $US8 billion per 

annum. Globally oil from oils seeds is approximately a $US50 billion per annum industry (FAO). 

Meat, sugar, timber, and other commodities are even greater in value. The Union of Concerned 

Scientists recently suggested that global timber imports are $US160 billion per annum – a 5% tax



would generate $US8 billion per annum for REDD or other forest conservation initiatives (Union of 

Concerned Scientists USA, 2009). 

A tax is generally a blunt instrument as it affects both sustainable and unsustainable producers equally. 

A more effective approach would be to create certification or accreditation processes that allow 

sustainably produced goods to either trade at a premium, or to have lower costs associated with their 

lower environmental impacts. There are a number of these certification processes now in operations. 

The first and longest running is the Forest Stewardship Council (Forest Stewardship Council, 2009) 

which certifies the sustainable management of forests and provides for chain of custody certification 

so that end users can differentiate between wood products coming from Forest Stewardship Council 

Certified forests and those that are not. There are now a series of commodity Round-Tables (eg Soy, 

Palm Oil, Sugar) that are proposing certification standards and mechanisms to differentiate sustainable 

production from unsustainable production (RSPO, 2009). 

This is a critical first step, but not quite enough to achieve a no net loss end-game. The sustainable 

production systems are accredited to best practice in terms of not clearing high conservation value 

forests, minimizing pesticide use, implementing fair employment practices, etc. They do not however 

require zero environmental impact. This means that as the global economy doubles and redoubles in 

size, the consumption continuously chips away at natural systems, albeit less rapidly than under 

unsustainable production systems. That is why the idea of actually embedding the ecosystem services 

products described above can more realistically move us towards an ultimate solution. 

The way this might work is that each bio-region would have an inventory of its ecosystems and state 

of the ecosystem services. In some cases the goal might be to stabilize the system, in other cases it 

could be to rehabilitate and recover some ecosystem services to a higher level (eg in our Great Barrier 

Reef example). In some rare cases it may even be acceptable to draw down further on the ecosystems 

before determining a stabilization point. Whatever that point is, then the ecosystems providing those 

services would be commodified into REDD projects, bio-banks and water quality banks. The credits 

could then be attached to the commodities produced in those bio-regions to create a systematic 

sponsorship of the conservation functions by the production functions. As an example of this, the 

Malua bio-bank can generate a commercially acceptable return from selling its Biodiversity 

Conservation Certificates (BCC) at $US10 per 100 square metres of forest rehabilitation and 

conservation management. Therefore, attaching one BCC to one tonne of crude palm oil only adds 

1.5% to the current price (Figure 1). 

The bio-regions may be supporting not only local commodity agribusiness, and point source industries 

like mining, oil and gas, but could be servicing global industries like electricity generation via REDD 

markets. In this case the bio-banks could be established to sell multiple ecosystem commodities 

including biodiversity certificates to palm oil companies, REDD credits to overseas energy sector 

firms, and water quality credits to downstream water users. Over time as the global economy grows 

the eco-commodities could become hugely valuable in line with their growing importance in 

supporting an ever larger global economy. The result would be landscapes that integrate production 

and conservation functions on a commercial basis (Figure 2).



Figure 1. An example of how biodiversity conservation credits can be integrated into the crude 

palm oil supply chain.



Figure 2. Landscapes with more functional ecosystems can be made more valuable 

CONCLUSION 

We appear to be on the verge of some breakthrough ideas that could shift the balance in the economics 

of deforestation and forest degradation. Not only is there a renewed emphasis on incorporating forests 

into the global carbon market, but the global agri-business industry is under pressure to introduce third 

party certification of the sustainability of production systems. These trends need to be linked with 

mechanisms to standardize the ecosystem services products like REDD, bio-banks and water quality 

or watershed conservation banks. We can already see the potential instruments emerging in voluntary 

markets and in some national level regulatory experience. The challenge now is to move ahead with 

implementation fast enough that the tide can be turned before there is little left to save.



REFERENCES 

Avoided Deforestation Partners (2008). AD Partners, Washington D.C., viewed 13/07/2009, 

www.adpartners.org. 

Caring for our Country (2009). Caring for our Country, Canberra, viewed 13/07/2009, www.nrm.gov.au. 

Central Intelligence Agency (CIA) (2009). CIA, Washington D.C., viewed 13/07/2009, 

www.cia.gov/library/publications/the-world-factbook/geos/XX.html 

Climate Action Reserve (2009). Climate Action Reserve, Los Angeles, viewed 13/07/2009, 

www.climateactionreserve.org. 

Clinton Global Initiative (2008). Clinton Global Initiative, New York, viewed 13/07/2009, 

http://www.clintonglobalinitiative.org//Page.aspx?pid=2646&q=270340&n=x 

Committee on Energy and Commerce (2009). Committee on Energy and Commerce, Washington D.C., viewed 

13/07/2009, http://energycommerce.house.gov/index.php?option=com_content&task=view&id=1560. 

Department of Climate Change (2009). Carbon Pollution Reduction Scheme, Canberra, viewed 13/07/2009, 

www.climatechange.gov.au/emissionstrading/index.html. 

Ducks Unlimited (2009). Ducks Unlimited, Memphis, viewed 13/07/2009, www.ducksunlimited.org. 

Ecosystem Marketplace (2009). Ecosystem Marketplace, Washington D.C., viewed 13/07/2009, 

www.ecosystemmarketplace.com. 

Environmental Trading Network (2009). Environmental Trading Network, Michigan, viewed 13/07/09, 

http://www.envtn.org. 

European Union Emissions Trading Scheme (2009), European Commission, Brussels, viewed 13/07/2009, 

http://ec.europa.eu/environment/index_en.htm. 

FAO Statistics Division (2009). Data available online at http://faostat.fao.org 

Food and Agriculture Organization of the United Nations (FAO) (2005). Global Forest Resources Assessment 

2005. Rome: Food and Agriculture Organization of the United Nations. Available online at 

www.fao.org/forestry/fra2005/en/ 

Food and Agriculture Organization of the United Nations (FAO) (1985). Tropical forestry action plan. Rome: 

Food and Agriculture Organization of the United Nations. 

Forest Stewardship Council (2009). FSC, Bonn, viewed 13/07/2009, www.fsc.org. 

Greenfleet (2009) Greenfleet, Melbourne, viewed 13/07/2009, www.greenfleet.com.au. 

Gutt, E., Patton, V. and Spencer, N. (2000), Building on 30 Years of Clean Air Act Success: The Case for 

Reducing NOx Air Pollution, viewed 13/07/2009, www.edf.org/documents/398_CAAReport.PDF. 

Malhi, Y., Meir, P. & Brown, S. (2002) Forests, carbon and global climate. Philosophical Transactions of the 

Royal Society. A 360, 1567–1591. 

Millennium Ecosystem Assessment (2005). Ecosystems and human well-being : current state and trends : 

findings of the Condition and Trends Working Group / edited by Rashid Hassan, Robert Scholes, Neville 

Ash. Island Press, Washington, DC 

Millennium Ecosystem Assessment (2005). Ecosystems and Human Well-being: Synthesis. Island Press, 

Washington, DC 

Ministry for the Environment, (2009). The New Zealand Emissions Trading Scheme, Wellington, viewed 

13/07/2009, www.climatechange.govt.nz/emissions-trading-scheme/index.html. 

MWHCB Inc (2009). Malua BioBank, viewed 13/07/2009, www.maluabiobank.com. 

RSPO (2009). RSPO, Kuala Lumpur, viewed 13/07/2009, www.rspo.org. 

SPECIESBANKING.COM (2008). The Katoomba Group, Washington D.C., viewed 13/07/09, 

www.speciesbanking.com. 

TZ1 Market (2009). TZ1 Market, viewed 13/07/2009, www.tz1market.com. 

Union of Concerned Scientists USA (2009). Union of Concerned Scientists, Cambridge, MA, viewed 

13/07/2009, www.ucsusa.org. 

United Nations (2000). Forest Principles, viewed 13/07/2009, 

http://www.un.org/documents/ga/conf151/aconf15126-3annex3.htm. 

United Nations, Department of Economic and Social Affairs, Population Division (2009). World Population 

Prospects: The 2008 Revision, Highlights, Working Paper No. ESA/P/WP.210



United States Fish and Wildlife Service, (2009). USFWS, Washington D.C., viewed 13/07/2009, 

www.fws.gov/laws/lawsdigest/ESACT.html. 

US Environmental Protection Agency (2008). Clean Water Act, Washington D.C., viewed 13/07/2009, 

www.epa.gov/watertrain/cwa. 

US Environmental Protection Agency (2009). Clean Air Act, Washington D.C., viewed 13/07/2009, 

www.epa.gov/air/caa. 

VCS Association (2008). VCS, Washington D.C., viewed 13/07/2009, www.v-c-s.org. 

WWF-Australia (2009) WWF-Australia, Australia, viewed 10/07/2009, www.wwf.org.au/articles/feature08 

ABOUT NEW FORESTS 

New Forests (see www.newforests.com.au) is an investment management and advisory services firm 

specialising in forestry and land-based environmental markets, such as timber, carbon, biodiversity 

and water. The company’s investment philosophy seeks to deliver traditional timber returns as well as 

returns from eco products, such as certified timber, renewable energy, carbon credits, biodiversity 

benefits and water-quality improvements. The company is headquartered in Sydney, Australia, with 

offices in Washington, D.C., San Francisco and Kota Kinabalu, Malaysia. New Forests holds an 

Australian Financial Services Licence. 

New Forests manages over $250 million in sustainable forestry and eco product (carbon, biodiversity, 

and water) assets in the United States, Australia, New Zealand, Southeast Asia and the Pacific Islands. 

This includes the full value chain of services, from the development of investment theses and portfolio 

management to operational execution and asset management. New Forests Advisory Services 

business provides policy and market analysis, regulatory advice, technical modelling capabilities and 

investment strategy to external clients and internal investment management teams. The business line is 

focused on developing commercial solutions at the intersection of land-based investment, conservation 

and climate change mitigation by identifying and quantifying environmental asset opportunities. 

New Forests' staff includes experienced professionals across timberland investment, operational 

forestry, environmental management and finance, and the company provides its clients with a unique 

combination of forestry, carbon and financial management skills. Staff members have been involved 

in forest carbon transactions since the 1990s and have contributed to the development of forestry 

offset rules in Australia, New Zealand, California, the United States, Canada and through the 

UNFCCC, as well as the drafting of the Agriculture, Forestry and Other Land Use guidelines for the 

Voluntary Carbon Standard.



CAPTURING THE VALUE OF INNOVATION – 

R&D FOR FOREST AND WOOD PRODUCTS INDUSTRIES 

Kathryn Adams 1 

ABSTRACT 

Australia has invested significantly in the development of a well managed forest and 

wood products industry, backed by internationally recognised certification for sustainable 

forest management and chain of custody. To maintain and improve its competitiveness 

the industry should not lose sight of the need to continue to invest in its innovation capital 

through focussed scientific research. 

In Australia the government provides mechanisms through Forest and Wood Products 

Australia (FWPA) and the Cooperative Research Centre (CRC) program. However, this 

joint investment can dilute direct engagement of commercial end users and they may miss 

some of the value that this opportunity provides unless they actively participate. 

This paper provides a reminder of the need for a clear industry strategic direction on 

which to base research investment priorities, followed by answers to the questions ‘why 

are we doing it, who will use it, how will they use it, are the answers there already, how 

should we manage the intellectual property to properly capture the value’? If this is done 

well, project design will focus the research on impact and outcomes rather than around 

convenient inputs. 

ADDRESS 

Ladies and Gentlemen, 

It is a privilege for me to be here today as a Keynote Speaker for this session on ‘Promoting 

Innovation in Forest Management and Processing’. But before I start I would like to emphasise that 

the views that I express here are my own and are not those of any organisation with which I might be 

associated, including Forest and Wood Products Australia Ltd and Australian Forestry Standards Ltd. 

This theme of ‘promoting innovation’ is critical for the future of the forestry and wood products sector 

if it is to realise its full potential in Australia and overseas. Australia has invested significantly in the 

industry to ensure that this renewable resource is adding value to the environment, carbon reduction, 

carbon storage and the economy. This is supported by internationally recognised certification schemes 

for sustainable forest management and chain of custody. To maintain the benefit from this investment, 

the sector needs to be able to continue to substantiate its position through ongoing investment in 

science based research and development (R&D). It needs to take care that this investment keeps pace 

with market needs and that it don't use up its current store of innovation assets faster than it is creating 

new ones. 

The Macquarie Dictionary defines ‘innovate as: 

‘to bring in something new; make changes to anything established’. 

This can happen at a very simple level by changing the way information is filed or answering the 

phone in a more consistent and informative way. At a higher level it can include introducing new 

technology or production systems. More complex innovation is generally based on scientific R&D 

from within a company, commissioned by a company or arising from joint investments such as 

through FWPA or a CRC. If a company undertakes or commissions the R&D it is more likely to 

adopt the outcomes as it is intimately involved in the project and has commercial reasons for 

1 

Director, Forests & Wood Products Australia, Level 4, 10-16 Queen Street, Melbourne VIC 3000; and Director, Australian 

Forestry Standard Limited.



investing. Where the R&D comes from an external source, even though companies might be 

contributing, it is more difficult to ensure that the R&D is meeting a specific need. I intend to focus 

on this ‘shared investment’ R&D and to emphasise the need for industry engagement at the company 

level if the full value of the investment is to be captured. 

I don't propose to tell you anything new but rather to repeat some of the key attributes that the industry 

could address to help maximise the return through the opportunities provided in Australia for joint 

investment in R&D. I will look at the issues under four headings: 

1. The forest and wood products sector in Australia today 

2. Some key principles for capturing the value from R&D 

3. Promoting and marketing the benefits 

4. Challenges for industry 

THE FOREST AND WOOD PRODUCTS SECTOR IN AUSTRALIA TODAY 

Industry Overview 

The Forest and Wood Products Australia (FWPA) Strategic Plan 2009-2013 (2008, p. 7), has a table 

showing the contribution of the forest and wood products sector to the Australian Economy: 

Table 1: Some industry statistics 

Industry turnover $19 billion/year 

Direct employment 83,000 people 

Proportion of Gross Domestic Product 1% 

Proportion of manufacturing industry 7% 

The sector (including pulp) contributes about 1% of GDP. In relation to the non-pulp sector, figure 1 

shows that the estimated gross value of log production in Australia has increased (not adjusted for 

inflation) from $1,337 m in 2001-02 to $1,872 m in 2007-8. 

(from numbers in ABARE, 2009, p.13). 

In 2007-08 the apparent consumption of sawn wood increased by almost 9 per cent, mainly from 

increased use of softwood both from Australia and overseas (ABARE, 2009 p2). Although growth 

indicators for the current year are not as positive due to the economic downturn, the sector contributes



significantly to the Australian economy and the gross value of log production has been increasing (in 

unadjusted terms) at least since June 2002. 

Investment in R&D 

For the sector to continue to contribute in this way to the Australian economy it will need to maintain 

its competitiveness and credibility both in Australia and overseas. Much depends on the industry’s 

ability to innovate and substantiate its practices through scientific analysis. There is extensive 

literature on the role of scientific R&D in assisting with industry growth and commercial viability. Of 

itself R&D is not enough and needs to be accompanied by innovative management, marketing and 

sales – it is an integral part of a well managed business. 

Historically, State governments in Australia have produced the wood, while processing has been with 

private sector companies, many of them small sawmillers. Research relating to production of timber 

was undertaken by government research organisations and some universities whereas processing 

research was undertaken predominantly by CSIRO, universities and individual companies. As 

plantation timber has increased, more companies and government corporations have become involved 

in growing timber. 

To try and increase the level of private sector investment in R&D, the Forest and Wood Products 

Research and Development Corporation (FWPRDC) was established in 1994 as a Commonwealth 

Statutory Authority to invest in R&D for the industry. It was based on the model of other Primary 

Industry R&D Corporations (RDCs), with a levy on production of logs for processing, which was 

matched by the Australian Government. In addition there was also a levy on imported logs to 

minimise any perceived advantage or ‘free rider’ effect that might accrue to importers of logs. The 

pulp and paper sector opted not to contribute (and take advantage of the Commonwealth contribution) 

by setting its levy at zero. Although producers of wood (predominantly the States) did not contribute 

to the levy (for constitutional reasons), they benefited through support for their research programs 

where there was clear value to the processing sector. 

In recent years a number of RDCs have been replaced by an industry owned company limited by 

guarantee to integrate R&D with marketing and industry promotion. In September 2007 Forest and 

Wood Products Australia Ltd (FWPA) was established in this way to replace FWPRDC and expand its 

role into other industry services, including generic promotion and marketing. As part of the agreement 

with the Australian Government, the investment in R&D was not to be lower in dollar terms than the 

average of the last 3 years of FWPRDC ie $6.73 m per annum. Expenditure on R&D is matched by the 

Australian Government up to 0.5% of industry GVP (the levy is currently only about 0.2% GVP and 

the pulp sector still has a zero levy). Expenditure on promotions and marketing for the industry is not 

matched. 

In addition, for the first time the forest growers are contributing to FWPA (the States on a voluntary 

levy basis) so the whole industry, from ground to wood product, has an interest in the value add and 

increased return on investment that can be obtained through FWPA's investment. 

The other avenue for joint investment in R&D in Australia is through the CRC Program, where 

research providers (including universities), companies and other R&D investors come together to 

invest in a specific, focussed R&D program. The Commonwealth runs a competitive bid process and 

contributes to the successful CRCs for 7 years initially, provided that the CRC meets its agreed 

research objectives. There have been at least 3 CRCs focussed on the forest and wood products sector. 

In a recent draft report for FWPA, Turner and Lambert (2009 – unpublished p 24) found that R&D 

investment by the sector in Australia has increased in unadjusted dollars, since 1982, from about $44.5 

million per annum to $104.6 million per annum in 2007-8 (6.2% of value of logs harvested). However, 

in real terms this is a decrease in investment to about $37.5 million in 1982 terms (figure 2). In 

addition, the R&D effort is fragmented, with over 50 organisations and 500 scientists and technicians, 

all with different strategic agendas. In 2007-8, 52% of research expenditure was provided directly or 

indirectly by the Commonwealth, 28% by State Governments and 20% by private companies (Turner 

and Lambert 2009 unpublished p14).



Total expenditure ($ million) 

120 

100 

80 

60 

40 

20 

0 

1982 1986 1990 1995 2002 2008 

Assessment Year 

Total Actual 

Total 1982 Dollars 

Figure 2. Total research costs estimated over the study periods from 1981/82 to 2007/08 as both 

actual dollars and adjusted dollars (1982 base). (from Turner and Lambert 2009 

unpublished, p15). 

CAPTURING THE VALUE 

Capturing the value from investment in R&D is not always easy. There is a general rule of thumb that 

says only about 1 in 100 projects will yield results that get to commercial pilot stage and that maybe 1 

in 1000 will actually get to market and make money. These figures differ depending on the proximity 

to market, the complexity of the innovation and the level of change that is required in an operating 

business to implement the innovation. For example, if a whole new processing line is required and the 

existing one still has 20 years of life, then the value add of the new line will need to be considerable. 

However if the innovation can be readily integrated into the existing system, then the likelihood of 

more immediate change is much greater. 

In 2007 Agtrans completed a cost benefit study for FWPA which showed that for the 36 completed 

projects in the sample there were a range of quantitative and qualitative benefits. The weighted 

average benefit at 5% discount rate, using 2005-6 dollars, was 11:1. Similarly the RDCs have just 

completed a joint cost-benefit analysis of a range of programs. They too found a return of about 11:1 

from 32 randomly selected projects rated across their combined portfolio. In addition a cost benefit 

analysis was done on 36 projects rated as ‘highly successful’ across the 15 RDCs. For a total $465 

million investment ($265 million from RDCs plus $200 m from partners), the value added was $10.5 

billion (22:1) being $5.5 billion in direct industry benefits and $5.0 billion in other benefits (eg 

environmental and social benefits). The project was undertaken by seven independent consultant 

groups and the methodology was reviewed by a number of key economic development agencies 

including: 

Treasury 

Department of Finance and Deregulation 

Department of Agriculture Fisheries and Forestry 

Productivity Commission 

Australian Bureau of Agricultural and Resource Economics.



In the project Summary Report (Rural R&D Corporations (2008), p3) it was noted that: 

The returns attributable to the RDCs’ $265 million investment – $5.9 billion – will more than pay 

for the entire $4.5 billion invested by RDCs across 600 projects over the past 10 years'. 

These types of analyses are undertaken by most rural RDCs, CRCs and others who invest in R&D and 

innovation both in the private and public sector. For both FWPA and the RDC model in general, this 

is an excellent result and these analyses are a valuable tool in assessing the potential return on R&D 

investment. 

However, for the forest and wood products sector, where Turner and Lambert have noted the 

fragmented nature of the R&D effort, the question has to asked as to whether there may be ways of 

improving this return even further by placing more emphasis on simple principles that are included in 

the current investment processes but where companies, investors and researchers could place more 

emphasis. 

R&D investment principles 

I would like to look at whether there are some more basic steps that can be taken to improve the 

usability and value add of R&D outputs. I will focus particularly on research programs that are for the 

benefit of an industry and the Australian community and are not company specific; for example where 

a group of interested partners are investing jointly eg through FWPA, CRCs, Joint Ventures. I would 

also like to concentrate on the industry benefits in particular, while recognising that joint funding with 

government also requires community benefits. My purpose here is to highlight the need for more 

commercial engagement, on the basis that the government is engaging to ensure that the community 

benefits are being achieved in a more direct way than the company investors. 

Where a company is investing directly in R&D for its own benefit, it will take a close interest in the 

direction and progress of the project. Where there is more than one investor the specific benefit to each 

one is a little further removed and responsibility for the success of the investment can become diluted 

the greater the number of investors. It is then left to the researcher, an industry association or the 

coordinating organisation to do their best to interpret what they think the end users want. 

To help bring the focus back to end user driven R&D, I have built a simple framework based on eight 

fundamental questions to be asked by those identifying investment priorities and preparing and 

assessing investment proposals: 

• What R&D should we be investing in, or from another perspective, why are we investing in 

this research? 

• What outcomes/impact do we want to achieve as a result? 

• Who will use the outputs to achieve these outcomes? 

• How will the outputs be used? 

• How is industry ‘usability’ monitored during the project? 

• How is intellectual property identified and managed to maximise benefits to industry? 

• Are we making the best use of the information already available in this area? 

• How to we design a project to achieve the outcomes and address each of the above issues? 

The effectiveness of this approach depends on how seriously those assessing the proposed investment 

are in requiring complete and substantiated answers. Focus on outcomes rather than inputs can result 

in much more directed investment and more usable, value-adding outputs for the end users. 

The approach I am taking can be applied to all levels of research, whether it be short-term problem 

solving or longer term ‘good ideas’ research. It also applies whether the outcomes are industry or 

community specific. The answers may not be as clear cut for the longer term issues and may change 

over time, but the important thing for an industry is to have a focus for R&D as an integrated part of 

an industry strategic plan. It can then be monitored and modified as circumstances change. 

Experience with a number of R&D investing organisations suggests that there is a general acceptance 

that this approach is already being taken, but while at the Board and top management levels this may 

be the case, time and other pressures mean that there can be considerable deviation on the ground. For 

example, most application forms from RDCs or CRCs have some questions around commercialisation



and path to market, but very few of the project milestones actually reflect an integrated approach to 

market need, focussed research and industry usability. Similarly, The project summary is written 

before the questions about commercialisation and adoption so it is clear that the project is designed 

and then paths to market and industry use are made to ‘fit’ the project. 

What research should we invest in and why are we investing in this research? 

These are fundamental questions but are often glossed over with greater focus being given to inputs 

rather than addressing real needs. How an investor decides where the priority investment areas are and 

then which particular projects to invest in should be identified by 'market need'. ‘Who wants this 

research and how will the outputs value-add’ are key questions? To help this process it is important 

for industry to be able to identify its strategic directions and then the market drivers. 

• Industry Strategic Directions 

The first questions to ask is ‘does the forest and wood products sector in Australia have an 

industry strategic plan to identify where it wants to be in 5-50 years and what R&D, 

promotion and marketing does it need to get there? There have been some sector based 

approaches to this, including the ‘Plantations 2020 Vision’ developed in 2002, and the 2008 

Forest and Forest Products Committee ‘Forest Research Strategic Directions 2008-2011’. 

Neither of these really provides an industry driven, integrated approach to a strategic future 

which would provide a firmer base for decision making. If the Australian forest and wood 

products sector could develop this sort of strategic framework, it could be used for a range of 

purposes, but would be particularly useful for organisations like FWPA when developing 

investment priorities. 

To be successful, such a strategy needs the CEOs of the companies to engage and drive the 

strategic direction – it cannot be driven by researchers or industry association staff; they may 

have the ideas for innovation but the companies, at a senior level, must provide the strategic 

commercial direction and market drivers for R&D focus. This is easier said than done as 

market forces change rapidly and response time, particularly in the tree growing area, is long. 

CEOs often do not have the luxury of time to focus on industry wide strategies and where they 

do they need to be cognisant of Trade Practices Act restrictions. 

• Is there a market? 

Once a strategic direction has been developed and the broad areas for research investment 

have been outlined, researchers and investors can take a market based approach to project 

development by asking 'what research is needed to meet industry market expectations'. 

What outcomes/impact do we want to achieve as a result? 

The next question is to clearly articulate the benefits (quantitative and qualitative) that the investment 

is meant to achieve. This may not always be directly translated into dollars but may include practice 

change within a community or contribution to the science behind policy development. The main thing 

is that the desired impact can be identified and measured so that investment in R&D has clear goals. 

Who is going to use the outputs from the research? 

While at the higher strategic level, an area of research might be identified as being valuable, it is also 

important to identify the potential end users at the project level and ask them if they would use the 

results from this research in their business? If not, why not? Could the outputs be changed in any way 

to make them more usable? If they would use it, how long before they would be in a position to do 

so? What further work/investment would they need in-house? 

When asking the question it is important to understand the types of end-users. Those that are looking 

for new ideas are more likely to give positive answers than those that wait until something is fully 

proven and has been in the market a long time. This is important as the first group may be the ones to 

focus on at this stage in an investment program. 

It is also important to make sure that the target user is who you think it is. For example, who is the 

end user of research into new species with greater stiffness characteristics? Is it the grower, the 

sawmiller, the timber specifier/architect/engineer? Is it all of them and if so do they each have 

different ‘usability’ criteria?



How are they going to use it? 

Once it has been ascertained that the end users are likely to use the outputs from the research the next 

question is how will they use it? This is critical for the usability of the outputs. With many research 

projects the last milestone is the production of a final report and/or a research paper. But is this usable 

by the end users? If the output is computer software, is it user friendly? Will the end users need to 

undertake further adaptation work? Do they have the in-house skills to do that or should the research 

deliver a 'user' package as well? 

Will usability be monitored during the project? 

The usability of project outputs can often be improved during the project if users are kept engaged or 

asked for feedback at regular intervals. Build this into the project milestones, with ‘go-no-go’ 

decision points. Waiting until the end can often mean that changes to project design that could have 

easily been made at an earlier stage, become major. 

Make sure that the users are giving real feedback by asking how will you use this in your business? 

Can we make it easier for you? How? Also make sure they are the ‘real’ users and not go betweens. 

There is the story of the industry project steering group who met every six months to discuss progress 

and were very supportive until at the end they were asked 'will you use this in your business' and most 

said 'no'! The concept was good but the cost-benefit of implementation was not there. Often when 

people are on a committee, they take a less accountable perspective than they would if they were 

assessing an investment using their own company’s money. 

How will intellectual property be identified and managed to optimise industry benefits? 

This is an issue that is not always handled well where there are multiple beneficiaries and where 

governments have contributed. The consequences of not handling it well and taking the easy 'public 

domain' route can mean that the attractiveness of the R&D outputs from a commercial perspective, are 

diminished and therefore the level of uptake and industry use may be lower than anticipated. 

R&D is the business of producing intellectual property (IP) and its management should be second 

nature to researchers, but our training does not do a good job in this area. All R&D produces IP. It 

takes many forms: some of it is directly protectable (eg PBR, patents, trade marks, copyright); some 

may need packaging into a 'system' before it can be protected; some can be protected through common 

law (eg trade secrets and confidential information) and other can be protected through contract. The 

important thing is to be clear why the research is being done, who for and how it will be used. Then a 

case by case 'structured' decision making process on IP management can be applied to each project, in 

the context of the portfolio of projects. Early disclosure may shut off commercial avenues not only for 

one project but for others in the portfolio as well. 

If it is clear who the beneficiaries are then IP management/industry use plans can be developed to 

maximise benefits. Integrated protection can include more than one form eg trademark to protect an 

overall brand and send a message to the market about the product or service, together with copyright 

in manuals, brochures and publicity materials, Plant Breeder's Rights in new varieties and patents in 

new products or processes. Decisions in relation to who will have access to the IP and under what 

conditions are critical for maximising return on investment. For example, licensing the IP to particular 

groups, such as those who have contributed to the project, can provide incentive to invest/use, whereas 

putting the IP into the public domain gives access to everyone, which may reduce the commercial 

value. 

Are we making the best use of information already available? 

One aspect of industry focussed R&D that always troubles me is whether we are duplicating what has 

happened in the past and whether we are making the best use of the information already available. 

Good practice for any research project dictates that there be a literature search before a new project is 

started to ensure that the wheel is not being reinvented or impinging on someone else's intellectual 

property. What concerns me is that I don't know that this is done particularly well. 

The other aspect that bothers me is whether all the information in final reports for projects undertaken 

over the last 10 years is being used as well as it could be. When you see the number of reports 

produced and the little time that is given to reviewing them, particularly by industry, I sometimes ask



should we stop where we are for 6-12 months and review/analyse these existing reports to see what we 

could use and what would be needed to make the outputs more usable – have we missed something? 

Can I design a project that will address the questions above and optimise the commercial usability? 

Traditionally projects are designed to meet broad research priorities and provide technical outputs. 

These technical solutions are then provided to the end user, only to discover that the fit with 

commercial reality is not always good. If the questions about why, who and how are answered before 

the project is designed, the usability of the outputs may be increased without jeopardising the scientific 

integrity of the project. 

Inclusion of an ‘industry use’ plan as part of the project with milestones related to ‘usability’ and 

‘market’ testing throughout the project rather than at the end can keep researchers and industry 

focussed on the work that is being done and the return that might be achieved. 

All of this may seem ‘motherhood’ to many, but experience has shown that a lot of our jointly funded 

research is based on available inputs rather than a thorough analysis of the why, who and how with 

design emphasis on usability and value add. 

PROMOTION AND MARKETING THE BENEFITS 

As FWPA is aware, as it seeks to move the industry forward through its 'Wood Naturally Better' 

campaign, you cannot promote a product unless you have sound scientific backing to substantiate the 

benefits. If that backing is in place, then a campaign is less likely to stall due to adverse response from 

competitors or the community. In addition the market research that monitors and evaluates and guides 

the direction of the campaign also delivers an insight into future market drivers and therefore R&D 

needs. You can't have one without the other and the integration of these two formerly separate 

activities can only have synergistic effects where 1+1=3 + . 

Along similar lines, the internationally endorsed (through the international Program for Endorsement 

of Forestry Certification Schemes - PEFC) Australian Forestry Standards for Sustainable Forest 

Management and Chain of Custody can only strengthen the viability of forest and wood products in 

Australia and overseas, but again there cannot be a standard without solid scientific backing and 

ongoing research to keep our industry at the leading edge globally. 

This integration of R&D and generic marketing and promotion for the industry is a major step forward 

but its main benefits will occur when industry, through company engagement, uses the market 

intelligence to drive its strategic planning and from that focus and coordinate its investment in R&D 

for innovation. 

From an FWPA perspective, this is beginning, with a number of CEOs being appointed to the Board in 

late 2008. The role and focus of industry advisory groups is also being reviewed to obtain relevant 

industry engagement. The challenge is to engage the right people at the right level into the future. 

THE CHALLENGES 

To maximise the return on their joint investment in R&D, I see the challenges for our industry as: 

developing an industry strategic plan to identify where it sees itself being positioned in the 

next 5,10, 20 and 50 years to help focus joint R&D investment for the future 

identifying key market drivers where R&D could value add and assist with delivering the 

strategic outcomes 

taking a stronger role in the avenues available for industry feedback on R&D eg FWPA 

industry Advisory Groups, steering committees etc. If you can't answer 'yes' to the question 

'would I use that in my business' (assuming it is in the relevant target user group), then the 

investment should be questioned; similarly industry advisory groups should have a majority of 

members from companies rather than industry associations – FWPA is your company and you 

should help it achieve your goals for you. 

encouraging young industry people to be engaged



asking the questions set out above, or pushing others to do so for every new R&D investment 

with the project design as the last task, not the first; and not glossing over the questions – 

focussing on outcomes rather than inputs 

asking itself: if the return on investment from FWPRDC (now FWPA) is of the order of 11:1, 

as the Agtrans study indicates, supported by a similar result from all users of the RDC model, 

why is the industry investment (via levies) only about 0.2% of GVP when the Australian 

Government will match up to 0.5% of GVP?; similarly why is the pulp and paper sector not 

taking advantage of this opportunity; and lastly 

could the return be greater if there was a less fragmented approach than that outlined by 

Turner and Lambert and is FWPA the glue that could help bring it together? 

I will end with the example provided at the recent CRCA Conference in Canberra earlier this year 

from the Roger Campbell, CEO of the Pork CRC (Campbell, 2009). In basic terms, the key outcome 

that the CRC is seeking from its investment in R&D is more $/kg of carcass weight. All projects are 

assessed against their ability to contribute to that outcome and they are regularly monitored to ensure 

they are continuing to meet it. If they are not performing against that target, they are either refocussed 

or discontinued. Not all organisations may want or be able to have such a single measure but it is an 

excellent example of the need to question what we are doing and why – is this contributing to our 

targeted outcomes and are we getting the best return from our investment? 

REFERENCES 

ABARE 2009, Australian Forest and Wood Products Statistics, September and December quarters, 2008, p.13. 

http://www.abare.gov.au/publications_html/afwps/afwps_09/afwps_may09.pdf 

Campbell, R 2009, ‘Measuring Economic Impact’, presentation at Pathfinders: the Innovators Conference, 

Cooperative Research Centres Association Annual Conference, Canberra, 28 May 2009. 

http://www.crca.asn.au/conference/pdf/presentation/03_thursday/03_Murray_Room/1300_- 

_1600/Roger_Campbel_%20-_Economic_Impact.pdf 

Forest and Forest Products Committee 2008, Forest Research strategic Directions 2008-2011. 

http://www.forestscience.unimelb.edu.au/informing_government/rpcc/RPCC_Forest_Research_Priorities_ 

Short.pdf 

FWPA 2008, Strategic Plan 2009-2013, p. 7. 

http://www.fwpa.com.au/Resources/About/5year/FWPA_Strategic_Plan_2009-2013.pdf 

Plantations for Australia: the 2020 Vision 2002, An industry/government initiative for plantation forestry in 

Australia. http://www.plantations2020.com.au/assets/acrobat/2020vision.pdf 

Rural R&D Corporations 2008, Measuring economic, environmental and social returns from Rural Research 

and Development Corporations’ investment, Report Summary. 

http://www.ruralrdc.com.au/WMS/Upload/Resources/Evaluation/Rural%20RDC%20Eval%20Summary% 

20low%20res.pdf 

The Macquarie Dictionary, 1981, Macquarie Library 

Turner, J and Lambert, M 2009, ‘Expenditure on forestry research and forest products research in Australia 

2007-2008’, unpublished review for Forest and Wood Products Australia, March 2009. (final version 

should be on www.fwpa.com.au by September 2009.)



ABSTRACT 

NATIVE FOREST MANAGEMENT OPTIONS FOR 

CLIMATE MITIGATION IN AUSTRALIA AND PNG 

Rodney Keenan 1 

Under the Kyoto Protocol, the use of forests to mitigate greenhouse gas emissions has 

been largely restricted to increasing forest carbon stocks through expanding the area of 

forests compared with the area in 1990, and reducing emissions due to land clearing. 

Options to manage existing forests to reduce greenhouse gas emissions or increase 

carbon sequestration have been discussed for some time and were eventually included 

in the Kyoto Protocol on a voluntary basis. With the development of the national 

Carbon Pollution Reduction Scheme and a new international climate mitigation regime 

for the post-2012 period scheduled for agreement by December 2009, it is timely to 

review the options to manage existing native forests for climate change mitigation. For 

developed countries, forest management is likely to remain voluntary and broadly 

defined, including both emissions due to harvesting and sequestration in regrowth from 

all types of activities on managed forests. In developing countries, the potential 

inclusion of degradation under Reduced Emissions from Deforestation and Degradation 

(REDD) may mean that accounting is unbalanced, with emissions associated timber 

harvesting included but not regrowth sequestration. This paper analyses the situation for 

Australia and Papua New Guinea (PNG). Both countries currently have limited capacity 

to comprehensively report on carbon dynamics in managed native forests. In Australia, 

emissions from harvesting are more than offset by managed regrowth, with emissions 

associated with wildfire likely to be the major factor impacting on carbon stocks. In 

PNG, harvesting is resulting in a net emission of carbon dioxide, even when regrowth 

sequestration is included. In PNG fire has limited impacts on carbon stocks. Options for 

reducing emissions are discussed. Establishing an agreed baseline against which to 

compare changes in emissions reductions remains a continuing policy and technical 

challenge. 

INTRODUCTION 

Vegetation and soils are major components of the global carbon cycle. The importance of effective 

management of these carbon sinks is recognised in the Framework Convention on Climate Change, 

which committed signatories to ‘promote sustainable management, and promote and cooperate in the 

conservation and enhancement, as appropriate, of sinks and reservoirs of greenhouse gases … 

including biomass, forests, oceans, and other terrestrial, coastal and marine ecosystems’ and other 

international agreements such as the Montreal Process on criteria and indicators of sustainable forest 

management (MIG 2008). 

In Australia, much of the focus has been on reducing greenhouse gas emissions from land clearing 

and increases in carbon storage in newly established forests. Comprehensive accounting of all carbon 

emissions and sinks has been proposed as a desirable long-term goal but there are still considerable 

risks and uncertainties associated with this approach (Australian Government 2008). 

Globally, the amount of carbon stored in terrestrial ecosystems, and fluxes between ecosystems and 

the atmosphere, can change over time due to variation in climate and disturbances such as fires, 

storms, pests and diseases or human activities. Greenhouse gas emissions from deforestation are a 

significant proportion (18-20%) of total human emissions but there is also a substantial uptake in 

vegetation recovering from past clearing or disturbance, which varies considerably from year to year 

(Canadell et al. 2007). 

1 

Professor Rodney J. Keenan, Department of Forest and Ecosystem Science, Melbourne School of Land and Environment, 

The University of Melbourne. Email: rkeenan@unimelb.edu.au.



Forests are part of the global carbon cycle through the natural processes of growth, disturbance, death 

and decay and human activities such as forest conversion, afforestation and utilisation of wood for 

energy, timber and other uses of wood and non-wood forest products. It is challenging to integrate all 

these components into a comprehensive assessment and policy framework for climate change 

mitigation. 

FORESTS AND CLIMATE CHANGE: THE POLITICS 

Inclusion of changes in carbon stocks in forests in Kyoto Protocol targets was controversial (Noble 

and Scholes, 2001, Keenan 2002). After considerable negotiation, the Protocol provided for Annex I 

countries to include change in carbon stocks due to afforestation, reforestation and deforestation and 

inclusion of eligible activities under Article 3.4 in broadly defined categories (forest management, 

cropland management, grazing land management and revegetation) at the discretion of the country. 

During the first commitment period, a Party that selected any or all of these activities needs to 

demonstrate that such activities have occurred since 1990 and are human-induced. Voluntary election 

of article 3.4 allowed Parties to leave out activities where there were methodological problems that 

meant there were high uncertainties related to net emissions or where the risk of emissions due to 

natural disturbances was potentially high. Australia chose not to include activities under Article 3.4, 

primarily due to risks of emissions associated with wildfire and drought. 

The 2007 Bali Action Plan put in place two negotiating processes, one addressing arrangements for 

Parties with commitments under the Kyoto Protocol beyond 2012 and another seeking global 

agreement for further co-operative action. Decisions under both processes are expected in 

Copenhagen in December 2009. Forests feature in both, with considerable debate over inclusion of 

activities under Article 3.4 and for developing countries incorporating Reduced Emissions from 

Deforestation and Degradation (REDD). Measuring and accounting for changes in forest carbon 

stocks is a continuing challenge, as is the establishment of appropriate baselines. Adaptation of 

societies to climate change, and achieving increased resilience and livelihood improvements, is a 

growing concern and aligning climate mitigation, poverty alleviation and biodiversity conservation 

objectives is a major policy challenge. 

In this paper, case studies from the Oceania region are considered to illustrate different issues 

associated with accounting for greenhouse gas emissions and removals due to forest management. 

Australia and PNG are near neighbours with quite different ecologies, indigenous cultures, European 

settlement histories, current land uses and stages of economic development. It has been proposed that 

the two countries become more deliberately linked in greenhouse gas emission reduction objectives, 

with reduced land-based emissions in PNG offsetting emissions in other sectors in Australia (Garnaut 

2008). In international negotiations, the Australian Government has suggested the inclusion of 

changes in forest carbon stocks in the context of comprehensive accounting under ‘forest 

management’ but highlighted the need to factor out the unbalanced, episodic effects of events such as 

wildfire or drought. PNG has been an active promoter of the inclusion of REDD in future global 

arrangements for climate change mitigation. 

CARBON ACCOUNTING FOR FOREST MANAGEMENT 

In previous negotiations, accounting for greenhouse gas emissions and removals due to forest 

management in Article 3.4 considered defining activities as ‘broad’ or ‘narrow’. A broad definition 

(eg. forest management) required accounting for the effects of all practices on an area of land subject 

to the activity. A narrow definition would assess the effects of individually-defined practices (eg. 

fertilisation or species change). A broad approach was agreed, supported by land-based accounting. 

Activity-based approaches may re-enter the carbon accounting scene. For example, the recent 

Waxman-Markey Bill that recently passed the US House of Representatives provides for the 

inclusion of a long list of narrowly and broadly defined land management activities to meet proposed 

US emission reduction targets. 

Carbon pools specified for forest-based greenhouse gas accounting are: above and below ground 

living biomass, dead wood, litter and soil organic matter (IPCC 2006). Ideally, a framework for



assessing GHG emissions associated with forest management involves (IPCC 2006, Penman et al 

2003): 

• Assessment of the forest area subject to the management activity. 

• Determining the extent of change in carbon stock associated with the activity. 

• Comprehensive accounting to include lands subject to past or present management. 

• Balanced accounting including all changes in carbon stocks. 

The extent of reduction in carbon stock can be estimated directly or by using emissions factors 

associated with indirect estimates of different activities. For example, reductions in carbon stock due 

to timber harvesting are often assessed using data on timber removals, because these statistics are 

more readily available than in-forest measures. 

Assessing the effect of change in practices or measures to reduce greenhouse gas emissions involves 

comparing the carbon stock change in one period (the baseline) with the change in the target period. 

1990 is the baseline year for the first commitment period of the Kyoto Protocol,. For afforestation or 

reforestation under Article 3.3 it was effectively assumed that there were no net emissions on cleared 

land in 1990. For deforestation, the 1990 emissions included the immediate loss of carbon associated 

land clearing and emissions due to decay of wood or soil carbon changes associated with clearing 

prior to 1990. For countries that included Forest Management under Article 3.4 accounting was on a 

‘gross-net’ basis with no 1990 baseline estimate because countries felt they could be penalised for 

imbalances in the age-class of their forest estate associated with past land use changes. Other 

measures were put in place to limit the extent to which countries could use this activity to meet 

emissions targets. 

Future requirements for accounting for greenhouse gas emissions in native forests are likely to be 

similar for forest degradation under new global arrangements and forest management under the 

Kyoto Protocol. 

Beyond 2012, a number of countries are arguing that a voluntary approach to Article 3.4 activities 

continues. Others have suggested that countries should at least demonstrate that carbon stocks in 

managed forests are being maintained (European Union 2008). 

Three alternative baseline options are currently under consideration: (i) maintaining a base year of 

1990; (ii) using net emissions over a ‘base period’ covering a number of years to produce a reference 

carbon stock change; or (iii) developing a ‘forward-looking’ baseline based on business-as-usual. 

Assessing carbon dynamics over a longer period under option (ii) would reduce the effect of 

imbalances in age-classes due to past human or natural disturbance, although that would require this 

period to cover a full cycle of plantation management (10-40 years) or even longer for cycles of 

wildfire and regrowth. The challenges in the forward-looking baseline are determining the forward 

period, establishing a credible and verifiable baseline for that period and considering the potential 

impact of a carbon price. 

NATIVE FOREST HARVESTING IN AUSTRALIA 

Forests cover 149 million hectares, 147 million hectares of native forests and nearly 2 million 

hectares of forest plantations. The area of public forests zoned for multiple-use and available for 

harvesting declined from 11.4 M ha in 2000 to 9.4 M ha in 2006, and the area of public nature 

conservation reserves increased from 21.5 M ha to about 23 M ha over the same period, as a result of 

processes to develop the conservation reserve system (MIG 2008). 

The carbon dynamics in intact native forests are complex. Forests vary widely in condition, with 

significant areas of old growth, with large trees and high carbon stocks (Raison et al. 2003, Keith et 

al. 2009). Others are in a regrowth condition or mixed-age class following past disturbance. There 

has been no systematic, repeated field-based inventory of Australian native forests and forest carbon 

dynamics are generally poorly understood (Keenan 2002). Estimates of native forest timber removals 

are currently used as the basis for emissions reporting in native forest management, together with 

estimates of growth in regrowing forests following harvesting or other disturbance. Harvest rates are 

a function of the area of forest available for harvesting, estimated growth and market conditions.



Native forest harvest rates have declined by about 10% since the 1990s, to an average of about 9.2 M 

m 3 per year (ABARE 2009). 

The current national greenhouse gas inventory accounting approach uses national harvest figures to 

determine the quantity of carbon removed in timber harvest, combined with an emission factor 

associated with slash (currently about 0.9 times the timber removed, DCC 2009). Using these factors, 

a crude estimate of harvesting emissions is 5.68 MtC (20.8 Mt of CO2) or 2.3 t CO2 per cubic metre 

of timber removed. Some emissions from harvesting slash actually occur over time through decay, 

and a proportion of the timber removed is added to the wood products pool rather than the 

atmosphere. Emissions from burning firewood are estimated to be about 10 Mt of CO2 (MIG 2008). 

Thus, total emissions associated with timber harvesting and firewood removal are about 31 Mt CO2 

per year. 

Under a comprehensive assessment of forest management, these emissions are offset by carbon 

sequestered in regrowth. Regrowth native forests were estimated to take up about 43.5 Mt CO2 per 

year in 2005 (MIG 2008). This suggests that carbon stocks in managed native forests are actually 

increasing by about 12.5 Mt of CO2 per year. The rate of increase is probably higher compared to a 

potential baseline period because of the reduced rate of timber harvesting, the shift from harvesting in 

old growth or mature forests to regrowth forests (with reduced slash emission factors). Also, 

estimated carbon uptake in regrowth does not include sequestration in a substantial area of regrowth 

forests that has been included in new conservation reserves and is not considered part of the 

‘managed’ forest estate. Emissions associated with firewood replace emissions that might otherwise 

occur due to heating with fossil fuels. 

In some local areas, long-term carbon stock reductions are occurring where old growth or mature 

forest is converted to regrowth through clearfelling or regrowth managed more intensively, with 

thinning and shorter rotations. Past timber harvesting or other disturbance may mean that carbon 

stocks in native forests are below their potential carbon carrying capacity (Roxburgh et al 2006) and 

it has been suggested that reducing native forest timber harvesting could result in long-term increases 

in native forest carbon stocks (Mackey et al. 2008). This is open to debate. A recent modelling study 

suggested that light selection harvesting could result in higher total forest carbon stock than in 

unharvested areas (Ranatunga et al. 2008). 

Additional reductions in emissions from native forest harvesting could be achieved by reducing the 

‘emissions factor’ associated with timber harvesting or increasing carbon stocks through regeneration 

and restoration of understocked areas. Reducing the emissions factor could be achieved by reducing 

the extent of clearfelling and conversion of old growth forest to even-aged regrowth forest, reducing 

the impact on retained stems during selection harvesting operations, retaining live standing trees in 

harvested areas, or maintaining a greater level of biomass in woody debris on sites after harvesting, 

perhaps through a reduction in intensity of slash burning and site preparation. 

Reducing removals from native forests may result in increased carbon stocks. However, this requires 

further analysis of sequestration capacity at different forest growth stages. Given that harvesting is 

now mostly of regrowth stands, harvesting at peak MAI may be the optimal short term carbon 

sequestration option, particularly if accounting includes storage in wood products. The overall effect 

of reducing native forest harvesting also requires assessment of ‘leakage’ associated with increased 

timber imports or more intensive utilisation of plantations and analysis of the cost of this emission 

mitigation measure compared to other options. Achieving maximum carbon potential will depend on 

effective management to maintain continued sequestration rates and consideration of the impacts of 

wildfire. This is discussed in the next section. 

FIRE IN AUSTRALIAN NATIVE FORESTS 

Fire is a widespread feature of the Australian landscape and much of the vegetation is adapted to or 

even dependent on fire and fire-related disturbances for regeneration and survival (Bradstock et al. 

2002). Wildfires have significant impacts on life, health, property, infrastructure and primary 

production systems (Whelan et al. 2006) and fire affects nutrient cycling and availability, forest 

productivity, vegetation composition, wildlife habitat, soil biota and hydrological functioning. Large-



scale wildfires can occur as a result of deliberate (arson) or unintentional (eg. powerlines) human 

activities, or naturally through lightning strikes. Prescribed fire is widely used as a management tool, 

although this practice is not without controversy and the rate of burning in eastern Australia has 

declined in recent years after peaking in the early 1980s (Adams and Attiwill 2008). 

The incidence and severity of wildfire varies considerably from year to year and decade to decade, 

depending on climatic conditions, vegetation type, fuel loads and human actions, and there may be 

prolonged periods (75-150 years) between ‘stand replacing’ fires (McCarthy et al. 1999). The area of 

forest impacted by fire is not well-monitored and fires vary greatly in extent, intensity and severity of 

their impacts. Accounting for greenhouse gas emissions from fire is also problematical. ‘Emissions 

factors’ associated with different fire intensities have not been developed. Fire results in some 

emissions directly as CO2 but smoke is a complex mix of organic compounds, particulates and 

carbon monoxide. After a fire, a considerable amount of wood is left in the forest. In many forest 

types impacted trees can recover quickly. Dead standing trees decompose over time and some of the 

remaining biomass is converted to charcoal that may have long-term stability in the soil. 

In south eastern Australia, where much of the native forest timber harvest is now focused, there has 

been a relatively high incidence of large scale wildfires since 2000, with major events in the summers 

of 2002-03, 2006-07 and in February 2009, partly as a result of a prolonged period of below average 

rainfall. Over 2.6 M ha have been burnt in these events with significant consequences for forest 

carbon stocks. Estimates of carbon dioxide emissions associated with these fires range from 40 Mt 

for the first (MIG 2008) to 550 Mt for both (Attiwill and Adams 2008). The most recent events on 7 

February 2009 which resulted in the deaths of 173 people and the loss of over 2,000 homes may have 

resulted in carbon emissions of over 50-70 Mt CO2. This is about three times the emissions that have 

occurred due to timber harvesting over the past 10 years (excluding firewood emissions). 

If the fire interval is sufficiently infrequent, this emitted carbon will be sequestered over time in 

regrowth following the fires. However, successive fires in the same area may have long-term impacts 

on forest carbon. Higher fire frequency can lead to shifts in vegetation composition and reduced 

forest carbon stocks. Regular prescribed fire can reduce the intensity of wildfires, so the potential 

interactions are complex. Under future climate change scenarios, the frequency of severe fire weather 

days is projected to increase over the next 20-40 years (Hennessy 2007) with significant implications 

for forest composition, structure and carbon stocks. 

Research is required to support improved accounting of forest carbon dynamics (particularly soil 

carbon) associated with different types of fire to provide a better basis for incorporation of forest fire 

management in climate change policy (Attiwill and Adams 2008). The extent to which wildfires are 

considered human-induced and assessing the effect of management practices (such as prescribed 

burning) that aim to mitigate fire impacts will be key policy design issues. 

FOREST HARVESTING IN PAPUA NEW GUINEA 

Papua New Guinea has a wide variety of environments and forests are characterised by high species 

diversity. Human societies in PNG are also highly diverse, with over 700 different language groups 

and a large number of different cultural and ethnic groups in the population of about 6 million 

people, and many people are dependent on forests for their livelihoods. 

There are five primary drivers of forest change in PNG: agricultural conversion, fire, mining, 

subsistence agriculture and timber harvesting (Filer et al. 2009). Conversion to intensive agriculture 

has been relatively limited in PNG. About 120,000 ha of oil palm plantation has been established 

over 30 years of development and not all of this has resulted in forest conversion. Fire has been 

shaping PNG’s vegetation patterns through thousands of years. If the interval is not too frequent, 

forests generally recover from fire and the structure of forests has in some parts been determined by 

previous fires (Haberle et al. 2001). Mining has locally significant impacts on forest cover, particular 

in Western Province, where deposition of mine tailings from Ok Tedi and overbank flooding is 

continuing to cause the death of significant areas of riparian forest. Subsistence agriculture appears to 

take place largely in cycles of cultivation and regrowth fallow and does not appear to be impacting 

extensively on primary forest (Filer et al. 2009). Timber harvesting is therefore the activity with the



greatest potential impact on forest carbon stocks. It has been suggested that harvesting is resulting in 

long-term degradation of forest carbon stocks and emissions of 75 – 85 Mt/year of CO2 and that there 

is little recovery of carbon stocks after harvesting (Shearman et al. 2008). 

Almost all land in PNG is under customary or tribal ownership. Forests have been seen as the basis 

for economic development for some time. Under the National Forest Plan, 11.9 million ha of land is 

identified as the production forest estate. Most of the forest subject to timber harvesting is allocated 

to processors and exporters through timber rights purchase and management agreements negotiated 

between the government and landowner groups, and approved by the National Forest Board. Most of 

the timber removed is exported, and log exports rates have varied considerably, ranging between 1 to 

2 M m 3 per year over the last 10 years. 

Harvesting occurs largely in primary forests. It is selective, with the intensity of felling and 

associated damage varying widely with the density of merchantable species and the market 

requirements, and the skill of the forest operator. Some areas are permanently converted due to 

roading or inadequate regeneration. Average timber removals are 10-20 m 3 /ha (Keenan et al. 2005). 

Using a figure of 15 m 3 /ha combined with the reported annual log export volume results in an 

estimated total of about 3.2 M ha impacted by selective harvesting up to 2007. 

Analysis of 125 permanent sample plots across the country (Fox et al 2009) indicated that timber 

harvesting reduces the average carbon density in above-ground live biomass (AGLB) by about 33%, 

or about 40 tC/ha. Applying this stock reduction to the harvested area associated with timber 

removals gives average CO2 emissions associated with timber harvesting of about 26 Mt per year 

over the 10 years from 1999 to 2008 (including an allowance for total loss of carbon stock on areas 

subject to roading and conversion to gardens), or about 17 t CO2 per cubic metre of timber removed 

(7.5 times the Australian value). 

The long-term impact on carbon stock in these forests will depend on their recovery capacity. Eighty 

nine of the above plots had measurement periods longer than about 5 years. About 76% showed an 

increase in basal area over the period (Yosi pers. comm.). Applying this percentage and carbon 

uptake figure of 2 tC/ha/year (Fox pers. comm.) to the cumulative area harvested results in an 

estimated rate of sequestration 12.8 Mt CO2/year. 

Thus, timber harvesting in primary forests in PNG is resulting in a net reduction in forest carbon 

stocks of about 13 Mt CO2/year when regrowth is included. Whether this stock reduction is long-term 

will depend on future timber harvesting or other human disturbances. The extent of future harvesting 

will depend on the residual stocking and rate of growth of merchantable species, the extent and 

location of the forest, future market conditions and, possibly, the value to the forest owners of any 

future avoided greenhouse emissions. Some accessible areas of are already being subjected to further 

cutting, either by larger companies or small-scale sawmilling. 

Reduced emissions could be achieved by reducing the rate of harvesting, reducing the impacts of 

harvesting (Putz et al. 2008) or both. Incorporating these reductions into a payment mechanism such 

as REDD will require establishment of a baseline. This will vary considerably depending on the 

period chosen because of the large fluctuation in historical harvest levels (Bank of PNG 2008). 

Ongoing monitoring will be required to assess the whether stock reductions are long-term and 

degradation is occurring. 

DISCUSSION 

The state of knowledge of carbon dynamics across the Australian and PNG landscapes is still highly 

uncertain. Research can improve capacity to quantify carbon fluxes across the landscape, but 

uncertainty is likely to remain high until comprehensive forest monitoring systems are established in 

both countries. In Australia, national accounting of greenhouse gas emissions due to deforestation is 

based on well-established system of remote sensing coupled with models of wood decay, soil carbon 

dynamics and regrowth sequestration (DCC 2009). Accurate accounting for carbon stock changes 

associated with forest management is less amenable to such approaches and requires a combination 

of remote sensing and ground-based measurements to verify carbon stock changes. The quantum of 

emissions and removals associated with different types of activities are not insignificant and justify



investment in improved monitoring, particular when it can be integrated with monitored timber 

stocks, vegetation composition and other values. 

A particular challenge is development of historical baselines. Establishment of a sound baseline 

using historical remote sensing data or aerial photographs requires considerable local knowledge to 

properly interpret past forest condition. To effectively incorporate changes in carbon stocks due to 

timber harvesting, fire or other forest management actions is likely to require the establishment of a 

retrospective baseline. Separating human-induced impacts (including indirect effects on forest 

growth such CO2-fertilisation) and legacy effects of pre-1990 activities (that have resulted in current 

forest age structures) from natural processes such as fire and inter-annual climatic variability will be 

a continuing scientific and policy challenge. 

A comprehensive accounting approach, that includes both losses due to disturbances and 

sequestration in regrowth, is important to provide balanced accounting and to provide incentives for 

effective establishment and management of regrowth, to restore degraded or failed areas and to 

maintain fully stocked stands. 

The interaction of timber harvesting and fire impacts is likely to become an increasingly important 

issue under future forest management and climate change scenarios. It has been argued that 

harvesting exacerbates the risk of fire in tropical forests but some types of harvesting activities may 

mitigate the effects of wildfire in temperate regions (Hurteau et al. 2008). 

Other forest values are also important. Maintaining variability in forest structure and understorey 

species can maintain carbon stocks and provide wildlife habitat benefits. Analysis of forest 

management effects on carbon stocks also should not stop at the forest boundary. The recent Fourth 

Assessment Report of the IPCC noted that carbon storage in wood products, replacing energy 

intensive building materials and fossil fuel emissions using biofuels from wood can increase forest 

carbon benefits. We need to utilise extracted timber as efficiently as possible to further increase 

carbon stocks in wood products and replace emissions from fossil fuels. The optimal solution to 

forest carbon management needs to consider the whole carbon lifecycle (Lindner, M. et al. 2008). 

ACKNOWLEDGEMENTS 

Colin Filer, Bryant Allen, John MacAlpine provided valuable background on PNG. Figures on 

carbon stocks and dynamics in PNG have been the result of long-term collaboration with the Papua 

New Guinea Forest Research Institute funded by the Australian Centre for International Agricultural 

Research (ACIAR). Julian Fox and Cossey Yosi provided data and other input from this work. 

REFERENCES 

ABARE 2009. Quarterly Forest Products Statistics. Australian Bureau of Agriculture and Resource Economics, 

Canberra. 

Attiwill PM, Adams MA (2008) Harnessing forest ecological sciences in the service of stewardship and 

sustainability: a perspective from ‘down-under’. Forest Ecology and Management 256, 1636-1645. 

Australian Government 2008. Land Use, Land-Use Change and Forestry (LULUCF) Sector Submission to the 

AWG-KP and AWG-LCA, UN Framework Convention on Climate Change. 

Bradstock, RA, Williams, JE and Gill, AM, (Eds) 2002, Flammable Australia: the Fire Regimes and 

Biodiversity of a Continent, Cambridge University Press, Cambridge. 

Canadell JG, Corinne Le Quéré, Michael R. Raupach, Christopher B. Field, Erik T. Buitehuis, Philippe Ciais, 

Thomas J. Conway, RA. Houghton, Gregg Marland (2007) Contributions to accelerating atmospheric 

CO2 growth from economic activity, carbon intensity, and efficiency of natural sinks. Proceedings of the 

National Academy of Science, 0702737104. 

DCC 2008. Agenda paper: International accounting rules relating to emissions from landuse: implications for 

scheme design’, Australian Government Department of Climate Change, Canberra. 

http://www.climatechange.gov.au/emissionstrading/consultation/pubs/international-accounting-rules- 

19may2008.pdf (accessed 9 August 2009). 

DCC 2009. National Inventory Report 2007. Appendix 7.B. Harvested Native Forests. Australian Government 

Department of Climate Change, Canberra.



Filer, C. Keenan, RJ, Allen, BJ and MacAlpine J. 2009. Deforestation and forest degradation in Papua New 

Guinea. Annals of Forest Science (in press) 

European Union 2008. Submission by France on behalf of The European Community and its member states: 

Land use, land-use change and forestry (AWG-KP). Submission to UN Framework Commission on 

Climate Change. 

Fox, JC, Yosi, CK, Oavika, F., Nimiago, P., Lavong, K. and Keenan, RJ 2009a. Assessing forest carbon in 

Papua New Guinea. Biotropica (in review). 

Garnaut, R, 2008. The Garnaut Climate Change Review: Final report. Cambridge University Press, Melbourne. 

Haberle, SG, Hope, GS and van der Kaars, S, 2001. Biomass burning in Indonesia and Papua New Guinea: 

Natural and human induced fire events in the fossil record. Palaeogeogr. Palaeoecol. 171(3/4): 259-268. 

Hennessy, K, Lucas, C, Nicholls, N, Bathols, J, Suppiah R, and Ricketts, J. 2005. Climate change impacts on 

fire-weather in south-east Australia. CSIRO, Melbourne. 

Hurteau M.D., Koch G.W., Hungate B.A. 2008. Carbon protection and fire risk reduction: towards a full 

accounting of forest carbon offsets. Frontiers in Ecology and Environment 6, 493-498. 

IPCC 2006. Good practice guidance for land use, land-use change and forestry (edited by J. Penman, M. 

Gytarsky, T. Hiraishi, T. Krug, D. Kruger, R. Pipatti, L. Buendia, K. Miwa, T. Ngara, K. Tanabe and F. 

Wagner). Institute for Global Environmental Strategies (IPCC National Greenhouse Gas Inventories 

Programme), Kanagawa. 

Keenan, R.J. 2002. Historical vegetation dynamics and the carbon cycle: current requirements and future 

challenges for quantifying carbon fluxes in Australian terrestrial ecosystems. Aust. J. Bot. 50:533-544. 

Keenan, RJ, Ambia, V, Brack, C, et al. 2005. Improved timber inventory and strategic forest planning in Papua 

New Guinea. Bureau of Rural Sciences, Canberra, and Forest Research Institute, Lae. 

Lindner, M., Green, T, Woodall, TW, Perry, CH, Nabuurs, GJ and Sanz, MJ 2008. Impacts of forest ecosystem 

management on greenhouse gas budgets. Forest Ecology and Management, 256: 191-193. 

McCarthy, MA, Gill, AM, and Lindenmayer, DB 1999. Fire regimes in mountain ash forests: evidence from 

forest age structure, extinction models and wildlife habitat. Forest Ecology and Management 124: 193- 

203. 

MIG 2008. Australia’s State of the Forest Report 2008. Montreal Implementation Group and Bureau of Rural 

Sciences, Canberra. 

Noble, I. and Scholes, R. 2001. Sinks and the Kyoto Protocol. Climate Policy, 1:5-25. 

Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L. Miwa, K. Ngara, T. 

Tanabe, K. and Wagner, F. (eds). Definitions and Methodological Options to Inventory Emissions from 

Direct Human-induced Degradation of Forests and Devegetation of Other Vegetation Types. 

Intergovernmental Panel on Climate Change and Institute for Global Environmental Strategies, 

Kanagawa. 

Putz, F.E., Zuidema, P.A., Pinard, M., Boot, R.G.A., Sayer, J.A., Sheil, D, Sist, P. Vanclay, J.K. 2008. 

Improved Tropical Forest Management for Carbon Retention. PLoS Biology 6:1368-1369. 

Raison, J, Keith, H, Barrett, D, Burrows B, and Grierson, P 2003. Spatial estimates of mature biomass. Tech. 

Rep. 44, Australian Greenhouse Office, Canberra. 

Keith, H, Mackey, BG and Lindenmayer D. 2009. Re-evaluation of forest biomass carbon stocks and lessons 

from the world’s most carbon-dense forests. Proceedings of the National Academy of Science, 

0901970106. 

Ranatunga, K, Keenan, RJ Wullschleger, SD Post, WM and Tharp, ML 2008. Effects of harvest management 

practices on forest biomass and soil carbon in eucalypt forests in New South Wales, Australia: 

Simulations with the forest succession model LINKAGES. Forest Ecology and Management, 255:2407– 

2415. 

Roxburgh, SH, Wood, SW, Mackey, BG, Woldendorp, G and Gibbons, P 2006, Assessing the carbon 

sequestration potential of managed forests: a case study from temperate Australia, Journal of Applied 

Ecology, 43:1149–59. 

Shearman, PL, Bryan, JE, Ash, J, Hunnam, P, Mackey, B and Lokes, B, 2008. The state of the forests of Papua 

New Guinea: Mapping the extent and condition of forest cover and measuring the drivers of forest change 

in the period 1972-2002. University of Papua New Guinea, Port Moresby. 

Whelan, R, Kanowski, P. Gill, M. and Andersen, A. 2006. Living in a land of fire. prepared for the 2006 

Australian State of the Environment Committee. 

http://www.environment.gov.au/soe/2006/publications/integrative/fire/effects-of-fire.html (accessed 10 

August 2009).



ABSTRACT 

AUSTRALIA’S PLANTATION VULNERABILITY 

TO CLIMATE CHANGE 

Jody 1 Bruce, Michael Battaglia, Libby Pinkard 

Predicting the likely impact of climate change on plantation productivity and 

security is fraught with uncertainty. While there is now some confidence as 

to the climate to 2030, models and scenarios beyond that period diverge. In 

addition we know little about how our plantation species will respond to 

elevated C02. Finally, the evidence we have to date suggests that climate 

change impacts will depend on the site conditions that present at a particular 

plantation locale. Changes in pest/pathogen distributions and the ability of 

forests to recover from damage will depend on the inherent fertility of 

individual sites and the future climate. Other disturbances such as fire 

introduce further uncertainty. We developed a series of surfaces across 

Australia’s plantation estate for Eucalyptus globulus, E. nitens, radiata pine 

and hybrid pine using a range of photosynthetic responses to elevated C02 in 

combination with varying fertility and future climates to investigate the 

changes in productivity and risk of drought death. 

This study presents an analysis that attempts to understand the spatial 

distribution of this uncertainty across Australia, identify on which parts of 

the landscape do the assumptions we make matter, and establish broad trends 

of change and risk under different combinations of assumptions. This 

allowed us to determine where climate change is likely to be damaging, 

where neutral or beneficial and where we are uncertain without more 

information and detailed site by site analysis. Some adaptation strategies to 

minimize risks are also investigated. 

1 

CSIRO Sustainable Ecosystems & CRC for Forestry, Private Bag 12, Hobart, Tas, 7001. Ph- +61 3 6237 5616. 

Email: Jody.Bruce@csiro.au



ENVIRONMENTAL MARKETS AND AUSTRALIAN FORESTRY: THE 

EMERGENCE OF A NEW PERIOD OF INDUSTRY DEVELOPMENT 

Paul Dargusch 1 , Steve Harrison 1 and John Herbohn 1 

ABSTRACT 

This paper investigates how carbon markets can influence change at an industry level 

by studying the case of the Australian forest industries. The influence of carbon 

markets on the current structure of the Australian forest industries is mostly evident in 

the emergence of 24 new firms that are primarily engaged in the business of 

afforestation-based offsets. Whilst these firms currently account for only a small 

amount of business (in terms of financial turnover and area of tree plantings) compared 

with the traditional constituents of the Australian forest industries, their emergence is 

indicative of a changing strategic outlook for the industry in which carbon markets are 

likely to play a much more influential role. Fervent debate in Australia during the last 

two years over a proposed mandatory emissions trading scheme (which allows the use 

of offsets generated through afforestation projects undertaken within Australia) has 

made the traditional constituents of the Australian forest industries give special 

consideration to how carbon markets will affect their businesses. The case study 

highlights the important role that entrepreneurial first-movers play in evoking change 

in industries. The authors argue that regulated carbon market policy should therefore be 

designed to support the capacity of entrepreneurial firms to engage in carbon markets, 

particularly in the early stages of the introduction of a mandatory emissions trading 

scheme and particularly in terms of how policy influences the entrepreneurial firms’ 

capacity to secure finance. 

INTRODUCTION 

Carbon markets are defined here as those markets associated with the trade of greenhouse gas 

emissions offsets and regulator-issued permits (mandatory and voluntary) to emit greenhouse gases. 

Carbon markets are particularly interesting in the Australian context because the current Australian 

government has proposed legislation to implement a mandatory greenhouse gas emissions trading 

scheme (Carbon Pollution Reduction Scheme (CPRS)) commencing in 2011. The CPRS has received 

substantial attention in the Australian mainstream media over the past two years and firms have begun 

to strategize to deal with the scheme’s pending introduction. The case of forest-related industries in 

Australia is also particularly interesting because the proposed CPRS allows offsets generated from 

afforestation-based activities carried out within Australia to be used by organizations to meet their 

CPRS compliance obligations. 

The Australian forest industries have had a turbulent history, characterized by a handful of notable 

political, market and investment-related events. It is convenient to consider the history of the 

Australian forest industries within five broad periods (Figure 1). The way in which carbon markets 

have influenced the changing structure and strategic outlook of the Australian forest industry – period 

5 - is discussed in the following sections of this paper. Evidence is presented in the form of an analysis 

of new types of businesses entering the industry. The paper concludes by discussing what this change 

might mean for the traditional constituents of the Australian forest industries. 

1 School of Integrative Systems, St Lucia Campus, University of Queensland Email: p.dargusch@uq.edu.au



Industry and Native Forests 

Small-scale 

harvesting for 

local markets 

1 st Period 

(up to 1960s) 

Industrial-scale 

native forest 

harvesting, 

mostly for 

export markets 

Industry and Plantations 

Plantation 

development 

mostly for 

research 

purposes 

Focus on 

avoiding 

overexploitation 

of native 

forests 

Substantial expansion in 

the softwood plantation 

estate by state governments 

Industry and Government Forest Policy 

Focus on policies to 

support wood product 

supply security and 

national selfsufficiency 

2 nd Period 

(1960s-1990s) 

Principles of ecological sustainable development 

exercised through the National Forest Policy 

Statement and Regional Forest Agreements guide 

native forest management. 

3 rd Period 

(1990s) 

Substantial expansion of 

the hardwood plantation 

estate by Managed 

Investment Scheme firm 

Focus on fostering 

hardwood plantation 

development using private 

sector funds 

4 th Period 

(2000s) 

Plantations for 

the primary 

purpose of 

carbon 

sequestration 

Focus on 


mitigation and 

adaptation 

5 th Period 

(emerging) 

Figure 1. History of the Australian forest industries conceptualized as five periods of industry 

development. The three sub-headings pertain to the three categories of themes, namely 

(1) how Australian forest industry has interacted with native forests, (2) how 

Australian forest industry has interacted with plantations, and (3) the main focus of 

Australian government forest policy as it relates to industry development. 

CASE ANALYSIS OF CARBON MARKETS AND AUSTRALIAN FOREST INDUSTRIES 

There now appears to be an emerging fifth period in the evolution of the Australian forest industries, 

characterised by how the industry engages in and is influenced by carbon markets. The proposed 

CPRS is a critical economic consideration for Australia. When implemented, it will constitute 

arguably the most far-reaching economic reform undertaken in the country for over 30 years. The 

proposed scheme has been intensely debated in media, political, industry and research circles in 

Australia for over two years, and as of July 1 st 2009, was at the stage of proposed legislation being 

deliberated by the lower and upper houses of the Australian Parliament. Whether the legislation is 

passed into law is uncertain but its high profile in mainstream debate has meant that many businesses 

and related stakeholders have given the scheme special consideration, and have begun strategizing and 

become increasingly engaged in actions in preparation for the scheme’s introduction.



Many of the 700 or so firms mandated to participate in the CPRS will hold substantial emissions 

liabilities 2 . These firms could manage the cost of their compliance obligations in various ways, 

including by buying permits via government-operated auctions, buying permits through the secondary 

market (including using various financial instruments such as derivates), increasing their ratio of 

revenue to greenhouse gas emissions by implementing cleaner forms of production, restructuring their 

businesses so that less of their business falls under the coverage of the CPRS, or by acquiring offsets. 

Given these options, it is likely that firms covered by the CPRS will be increasingly interested in 

buying or investing in greenhouse gas emissions offsets because offsets can offer reduced investment 

risk and greater cost certainty 

Importantly, afforestation is recognized under the CPRS as the only legitimate option that firms can 

use to offset their greenhouse gas emissions outside of their facility or corporate operational 

boundaries within Australia 3 . Permits created via afforestation projects registered under the CPRS will 

be bankable, tradable and extinguishable in the same way as government-issued permits. 

Consequently, there is likely to be substantial and growing interest over the next five years (coinciding 

with the proposed first commitment period of the CPRS) in afforestation projects developed within 

Australia that are registered under the CPRS. For example, a recent analysis by the Australian Bureau 

of Agricultural and Resource Economics estimated that, given a carbon price under an enacted CPRS 

of AU$28/tCO2e, an additional 21.8 million ha of agricultural land in Australia would be become 

economically viable for tree planting (in the context of existing competing land uses) (Lawson et al. 

2008). If these potential areas were planted, they would equate to almost a 10 fold increase in the 

Australian forest plantation estate. 

To be eligible for registration under the CPRS, Australian-based afforestation projects must meet 

stringent eligibility criteria including Kyoto Protocol compliance relating to forest type and land-use 

history. However, not all existing afforestation projects within Australia that comply with these Kyoto 

Protocol provisions will be eligible for registration under the CPRS. Rather, only those currently 

registered under the NSW Greenhouse Gas Abatement Scheme (NSW GGAS) or the Australian 

Commonwealth Government’s Greenhouse Friendly Program, will be eligible to be registered under 

the CPRS. To the best of the authors’ knowledge, this means that none of the existing plantings of the 

larger Australian MIS firms will be eligible to be registered under the CPRS. In other words, the 

industrial-scale traditional forest growing constituents of the Australian forest industries are unable to 

use their existing forest plantation estates to supply the demand for offsets that is likely to occur if the 

CPRS is introduced. The CPRS-driven demand for offsets would therefore need to be met by (1) 

existing constituents of the Australian forest industries changing the focus of their business to include 

the production of offsets, and (2) new entrants to the industry which choose to become directly 

involved in the business of afforestation-based offsets to meet their CPRS compliance obligations or 

seek profit from the business of offset production and trade. 

A number of Australian companies have recently been established that are primarily involved in the 

business of afforestation-based emissions offsets. Table 1 provides a summary of the characteristics of 

these firms. Part A presents a summary of the 24 Australian companies 4 that the authors identified 

were primarily in the business of afforestation-based offsets as of February 25 th 2009 while Part B 

presents a summary of 11 of these 24 companies that were primarily in the business of managing 

2 Whilst most types of industries are included in the coverage of the CPRS (e.g. electricity generators, transport, 

manufacturing, mining, waste) firms that operate businesses in the ‘forestry industry’ can choose to engage or 

not to engage in the scheme and firms involved in agriculture are excluded. 

3 Firms covered by the Australian CPRS will also be allowed to acquire offsets from Clean Development 

Mechanism and Joint Implementation projects based in other countries. 

4 The trading names of these 24 companies are Carbon Balance Consulting, Carbon Planet, Carbon Reduction 

Institute, Carbonza, Green Pass, Ark Climate, Australian Carbon Traders, Brokers Carbon, Canopy, Carbon 

Pool, Cool Planet, Emit Environmental Brokers, Green Pig, Auscarbon International, Carbon Conscious, Carbon 

Neutral, CarbonSmart (Landcare), CO2Australia, Elementree, Greenfleet, Greenhouse Balanced, Greening 

Australia, Offset Emissions and TreeSmart. The first 11 companies listed are those described in Part B of the 

Table as companies that were primarily in the business of managing afforestation-based offset generation 

projects within Australia.



afforestation-based offset generation projects within Australia. Note also that there are a number of 

Australian companies in addition to the 24 reviewed in Table 1 that are either primarily involved in the 

business of non-afforestation-based emissions offsets or are also involved in the business of 

afforestation-based offsets but for whom the business of afforestation-based offsets is secondary to 

other types of business activities. The nature and motivations of these additional firms is the subject of 

another study currently being undertaken by the authors. 

Most of the 24 companies listed in Table 1 are young (22 of the 24 are less than 10 years old, with 15 

less than five years old), small (17 of the 24 companies had an annual turnover of less than AU$1 

million), privately-owned (19) and ‘for-profit’ (19). There is an approximately equal mix of firms that 

operate as traders, managers of afforestation-based emissions offset generation projects, and advisors. 

All 13 firms that operate as either traders or advisors had an annual turnover of less than $AU1 million 

in 2007/2008 and were for-profit privately-owned companies with less than five employees. 

Interestingly, four of the five not-for-profit companies listed in Part A of Table 1 were in the business 

of managing afforestation-based offset projects and these are some of the largest companies by 

turnover of the 24 companies reviewed (four of the not-for-profit companies turned over more than 

AU$1 million per annum). These data suggest that not-for-profit companies, although relatively small 

in number, currently play a disproportionately influential (and arguably market-leading) role in 

Australian carbon markets, particularly amongst firms in the business of managing afforestation-based 

emissions offsets projects. 

Table 1 Summary of characteristics of Australian firms primarily engaged in the business of 

afforestation-based emissions offsets. 

Part A Summary characteristics of 24 Australian companies primarily engaged in the business of 

afforestation-based offsets 

Age of 

0 to 5 years 5 to 10 years 10 to 20 years >20 years 

company 15 (63%) 7 (29%) 1 (4%) 1 (4%) 

Commercial 

For-profit Not-for-profit 

purpose 19 (79%) 5 (21%) 

Commercial Trader or broker Project manager Advisor 

role 8 (33%) 11 (46%) 5 (21%) 

Ownership 

Private 

19 (79%) 

ASX listed 

1 (4%) 

Member based 

4 (17%) 

Clientele Governments Large companies Sees Individuals 

Primary 2 (8%) 13 (54%) 5 (21%) 4 (17%) 

Secondary 7 (29%) 0 4 (17%) 13 (54%) 

Are the traded 

Certified 

Not certified 

offsets 

(NSWGGAS or ‘Climate Friendly) (NSWGGAS or ‘Climate Friendly) 

certified? 12 (50%) 12 (50%) 

Estimated 

turnover 

AU$5 million 

3 (12%) 

Part B Summary characteristics of 11 Australian companies primarily engaged in the business of 

managing afforestation-based emissions offsets generation projects 

Age of 

0 to 5 years 5 to 10 years 10 to 20 years >20 years 

company 4 (36%) 5 (46%) 1 (9%) 1 (9%) 

Commercial 

For-profit Not-for-profit 

purpose 7 (63%) 4 (37%) 

Clientele Governments Large companies Sees Individuals 

Primary 1 (9%) 7 (63%) 0 3 (28%) 

Secondary 4 (36%) 0 2 (18%) 5 (46%) 

Land 

Share farming Lease Purchase 

acquisition 2 (18%) 7 (64%) 2 (18%) 

Target land 

750 mm pa 

rainfall 11 (100%) 0 

Distance to port 

200 km 

11 (100%)



Tree species 

planted 

Species traditionally used for wood 

production 

0 

Species not traditionally used for wood 

production 

11 (100%) 

Are the traded 

Certified 

Not certified 

offsets 

(NSWGGAS or ‘Climate Friendly) (NSWGGAS or ‘climate friendly) 

certified? 7 (64%) 4 (36%) 

Estimated 

turnover 

AU$5 million 

3 (28%) 

Source: Websites and other company reports as at February 25 th 2009. 

About half of the primary clients of the 24 companies in Table 1 are large companies (13 of 24 

companies had large companies as their main clients) and close to the same proportion of the existing 

secondary clients were individuals (13 of the 24 companies had individuals as their second most 

common clients). Furthermore, only half of the 24 companies stipulated that the afforestation-based 

offsets that they deal in must be certified under either the NSWGGAS or the Commonwealth 

Government’s Greenhouse Friendly program. These figures suggest that whilst many of the companies 

target large companies that will have a CPRS liability as clientele, there are a similar number that 

target either companies that do not have a CPRS liability or individuals that are perhaps more driven 

by the subjective personal values of offsets as opposed to their CPRS compliance value. 

Part B of Table 1 reveals a number of interesting features of how the 11 companies that manage 

afforestation-based offset projects go about the business of afforestation. All of the firms developed 

afforestation projects more than 200 km away from port facilities in semi-arid regions with annual 

rainfall typically less than 750 mm and all afforestation projects involved species not traditionally used 

for wood production but rather most used species endemic to the planting area. These differ 

substantially from traditional Australian plantations. 

Future growth of this new ‘carbon market’ sector of the Australian forest industries is likely to be 

fueled by demand for offsets from firms wanting to meet their CPRS obligations. Demand would also 

arise from firms which, whilst not having a compliance obligation under the CPRS or NGER, are 

motivated to minimize their ‘carbon footprint’ for corporate legitimacy, ethical and marketing reasons. 

This market growth has implications for conventional Australian forest enterprises, many of which 

may need or decide to adapt their existing capabilities to take advantage of new opportunities 

associated with carbon markets. Moreover, it is likely that government policies and protocols for the 

measurement, verification and commercialization of offsets from afforestation projects will be refined 

and made more ‘industry friendly’. As well, issues such as how to best include and manage soil carbon 

in carbon markets and how to best integrate trees and agricultural crops in agroforestry and 

silvopastural systems, are being included in the policy debate. 

DISCUSSION AND CONCLUDING REMARKS 

The entrance of the 24 new firms primarily in the business of afforestation-based offsets into the 

Australian forest industries is arguably the most notable indication of how carbon markets are 

influencing change in the industry. These new firms are typically entrepreneurial; their business 

activities are closely aligned with principles of ecologically sustainable development, and most are 

young and small-scale. Such features are characteristic of so-called eco-preneurial firms (a concept 

coined by Blue, 1990). Some research (e.g. Isaak, 2002; Schaper, 2002; Dixon and Clifford, 2007) has 

indicated that eco-preneurial firms can invigorate industry-wide change towards more sustainable 

business practices by introducing ‘innovation, adaptation and new ideas’ (Schaper, 2002, p. 27). It 

seems likely that the 24 new eco-preneurial entrants into the Australian forest industries are positioned 

to invigorate similar change in their industry. However the proposed CPRS and related climate change 

mitigation policies in Australia contain relatively few provisions that are specifically targeted towards 

supporting the development of these eco-preneurial new entrants. The authors believe that the 

effectiveness of regulated carbon market policy in achieving climate change mitigation objectives 

could be greatly enhanced by including more effective provisions that more specifically support the 

business development of eco-preneurial firms (particularly in relation to the capacities of those firms 

to secure finance).



Some of the 24 new entrants (including Greening Australia and Landcare, which are amongst the 

largest of the 24 firms by turnover) have been in the business of planting trees for more than 10 years 

but have only recently started to engage in carbon markets as a means of revenue generation. For most 

of their existence, Greening Australia and Landcare have utilized funding from corporate-based 

philanthropic sources or from the Australian Commonwealth government, to undertake extensive tree 

planting for ecosystem rehabilitation. With much of this public funding no longer available, and with 

increased interest in carbon markets, many of these organisations are now making the transition to 

business models with greater reliance on non-government funding. How successful these firms are in 

making this transition will be an important factor in how the Australian forest industry changes in 

response to carbon markets over the next 10 years. 

Three of the four largest new entrants into the Australian forest industries that are in the business of 

afforestation-based offsets are not-for-profit companies. Whether these companies would be able to 

raise enough capital or attract sufficient appropriately skilled staff to enable a rapid expansion of their 

business operations in the event demand for offsets increased dramatically is questionable. The authors 

are concurrently undertaking a more detailed survey of these not-for-profit firms to understand better 

their capacity constraints and expansion intentions. But regardless of these capacity constraints, the 

question needs to be asked whether not-for-profit companies are actually better suited to the 

production of offsets in carbon markets than profit-driven alternatives? 

Carbon markets are embedded in broader notions of sustainability and firms which have a compliance 

obligation under a regulated carbon market such as the CPRS are generally expected by stakeholders 

to not only seek the lowest cost means of compliance, but also to comply in a manner that contributes 

broader community benefits and enhances the sustainability of their business practices. 

It follows that perhaps not-for-profit companies are better placed than their profit-driven 

contemporaries to provide offset products that, whilst perhaps produced on a smaller and more costly 

scale, provide preferred sustainability benefits, such as the ability to engage the community in tree 

planting projects, smaller-scale plantings that more effectively enhance complementary ecological 

services, and better socio-economic outcomes at a regional scale. In this sense, government-owned 

organizations may also be well suited to afforestation-based offset production. Furthermore, it may be 

preferable for some firms which have a compliance obligation under the CPRS to engage in the 

business of producing offsets themselves (i.e. growing the trees), particularly in cases where those 

firms have some knowledge of land use and forestry, or have access to land suitable for afforestation 

projects. The authors are currently involved in research investigating the nature and motivations of a 

number of firms which are engaged in these types of activities. 

While this paper has focussed on the CPRS, voluntary carbon markets can also play an important role 

in industry change. Indeed, multiple factors are likely to influence change at an industry level, 

including the emergence of both regulated and voluntary carbon markets, dynamics of ecopreneurship, 

dynamics of industry economics and competition, and processes of organisational change 

in larger, traditional industry constituents. But the case of the Australian forest industries certainly 

does indicate that carbon markets are a prominent factor in industry change. 

The case study has highlighted how regulated carbon market policy provisions relating to offsets from 

forests can invigorate eco-preneurial business activity, and there is reason to believe that ecopreneurial 

activity can act as a catalyst for wider industry change that supports better climate change 

mitigation outcomes. 

This finding has relevance to the design of other regulated carbon markets, particularly those that 

might include provisions for offsets from forests, as is the case with proposed regulated carbon market 

policy included in the US Clean Energy and Security Act 2009. The effectiveness of carbon market 

policy should be enhanced if it included provisions to address the specific business development needs 

of eco-preneurial firms. Furthermore, it seems likely that not-for-profit and government organisations 

may be better suited to the production of forest-related offsets, given their ability to better support 

complementary social and ecological services.



The overall benefits of regulated carbon market policy could be increased if policy included provisions 

supporting the development of these types of organisations as well. Whilst the authors believe that the 

most pending types of assistance that eco-preneurial firms need are measures that support the firm’s 

ability to secure finance, the authors also acknowledge that a complex mix of factors influence ecopreneurship. 

The effectiveness of carbon market policy would be increased if policy decision-making 

was supported by a better understanding of the dynamics of eco-preneurship. Therefore, future 

research on carbon markets should investigate what Beveridge and Guy (2005, p. 672) described as 

the ‘processes and practices of emergence, negotiation and innovation’ of eco-preneurship. 


We would like to thank Neil Byron and Ian Ferguson for their helpful comments during the 

development of this paper. 

REFERENCES 

ABARE 2008. Australian Forest and Wood Product Statistics, March and June quarters 2008. Australian 

Bureau of Agriculture and Resource Economics, Canberra. 

Ajani J. 2008. Australia's transition from native forests to plantations: the implications for woodchips, pulpmills, 

tax breaks and climate change. Agenda: A Journal of Policy Analysis and Reform, 15(3): 3-10. 

Beveridge, R. and Guy, S. 2005. The rise of the eco-preneur and the messy world of environmental innovation. 

Local Environment, 10(6): 665-676. 

Blue, R. 1990. Ecopreneuring: Managing for Results. Scott Foresman, London. 

Dargavel, J. 1995. Fashioning Australia’s Forests. Oxford University Press, Oxford 

Dargusch, P 2008 ‘Understandings of sustainable corporate governance by Australian managed investment 

schemes and some implications for small-scale forestry in Australia’, Small-scale Forestry, 7(1): 67-75. 

Dixon, S. and Clifford A. 2007. Ecopreneurship - a new approach to managing the triple bottom line. Journal of 

Organizational Change Management 20(3): 326-345. 

Herbohn, J. and Harrison, S. 2004. The evolving nature of small-scale forestry in Australia. Journal of Forestry 

102(1): 42–47 

Isaak R. 2002. The making of the ecopreneur. Greener Management International 38(2): 81-91. 

Lawson, K, Burns, K, Low, K, Heyhoe, E and Ahammad, H 2008, Analysing the economic potential of forestry 

for carbon sequestration under alternative carbon price paths, ABARE, Canberra. 

NAFI 2008. Submission on CPRS Green Paper. National Association of Forest Industries, Canberra. 

Routley, R. and Routley, V. 1975. The Fight for the Forests – the Takeover of Australian Forests for Pines, 

Woodchips and Intensive Forestry. RSBS, ANU, Canberra. 

Schaper, M. 2002. The essence of ecopreneurship. Greener Management International 38(2): 26-30. 

Slee, B. 2001. Resolving production-environment conflicts: the case of the Regional Forest Agreement Process 

in Australia. Forest Policy and Economics 3(1-2_:17–30.



ABSTRACT 

LESSONS LEARNED FROM BUILDING ESTATE-SCALE 

CARBON ACCOUNTING SYSTEMS 

Scott Arnold 1 

In 2008 the Victorian Department of Sustainability and Environment embarked on a 

project to construct a set of carbon accounts for approximately 7.5 million hectares of 

public land in Victoria. The land comprised mostly National Park and State Forest 

tenures but also included a range of other crown land tenures. The area was stratified on 

the basis of historical event data and geographic region. The large volumes of historical 

data were simplified and used to determine the stock trajectories for each stratum. The 

initial data sets resided in a GIS and a database system was developed to pass data to and 

from FullCAM, where carbon stocks were calculated. The carbon stocks were most 

heavily influenced by wildfire events, principally as a result of the very large areas that 

are sometimes burnt. Harvesting events had a greater impact on a per-hectare basis but 

the relatively small areas involved limited their influence. 

INTRODUCTION 

The Department of Sustainability and Environment (DSE) manages approximately 7.5 million 

hectares of public land across Victoria under a range of tenures. Parks and Reserves (including 

National Parks, State Parks and Regional Parks) and State forests constitute the vast majority of that 

land, being 3.6 million hectares and 3.3 million hectares respectively. Public land represents one third 

of the total area of Victoria. 

Interest in the carbon stock held on that public land has increased in recent years and that interest has 

accelerated with the Commonwealth Government’s proposed Carbon Pollution Reduction Scheme 

(CPRS). 

Land Carbon Project 

In December 2008 DSE commenced a project, known as the Land Carbon Project, aimed at 

quantifying the amount of carbon stored on publicly managed land. The stated aims of the Land 

Carbon Project were to: 

• Provide an indicative set of carbon accounts for Victoria’s publicly managed land, 

• Increase the Department’s understanding of fluxes in carbon stocks, and 

• Inform the policy development process in areas relating to managing carbon as an asset. 

A Project Team of three staff was assembled in order to deliver a product by the end of April 2009. 

DELIVERABLES 

The Land Carbon Project Team were faced with a situation where there was very limited 

understanding of exactly what was expected in terms of deliverable products. Aside from the clear 

requirement to produce a ‘map of the carbon stocks’ the balance of the deliverable products were not 

enunciated in any detail. 

Given the ‘open’ nature of the project, the Land Carbon Project Team sought to clarify project 

expectations through a series of formal and informal meetings. Two key themes emerged from the 

meetings: 

1. The project must use data that, where possible, was of Victorian origin, and 

1 Department of Sustainability and Environment. Level 3, 8 Nicholson Street, PO Box 500, East Melbourne Vic 3002, 

Ph: (03) 9637 8520 Email: scott.arnold@dse.vic.gov.au



2. The impact of management events such as harvesting and burning must be identifiable. 

These themes played a significant role in shaping the project. The Land Carbon Project Team were 

now charged with using DSE data to construct a set of carbon accounts that described both total stocks 

and stock changes as a result of harvesting and burning on publicly managed land in Victoria. 

LAND CARBON STRUCTURE 

With a more refined project scope in place, the structure of the Land Carbon Project could be defined. 

Figure 1 illustrates the main components of the Land Carbon Project. 

Figure 1: Land Carbon Project structure 

GIS 

• Tree Cover 

• Harvesting 

• Fire 

• IBRA 

Map 

Products 

Database 

• Link to FullCAM 

• Format results 

Project 

Report 

FullCAM 

• Templates 

• Plot Files 

• Carbon models 

Analyses 

• Total stock 

• Impact of events 

Geographic Information System 

Carbon accounts are data-oriented products that, depending on their level of sophistication, can require 

large amounts of input data. Almost all of the data used in the Land Carbon Project was spatial data 

stored in the DSE geographic information system (GIS) library. 

Each of the GIS data inputs could be categorised as being either biophysical or administrative. The 

biophysical data described the events or characteristics that influence the carbon stocks on site and 

included harvesting extent, fire history, vegetation type and geographic region. The administrative 

data served two purposes; one was to define the extent of publicly managed land in Victoria and the 

other was to provide a basis for determining how the total carbon stock was distributed among the 

various State government land management agencies (eg: Catchment Management Authorities 

(CMAs), Local Government Areas). 

Database System 

After the input data had been prepared in the GIS, additional formatting work was required before the 

data could be passed on to the carbon model. A Microsoft Access 2003 database was built to create 

files in a format that allowed data to be interpreted by the carbon model. 

This same database was also used to receive all the outputs of the carbon model and to present that 

data in a format that allowed the results to be joined back to the GIS for map production and further 

analysis. 

FullCAM 

FullCAM is a stock balance carbon model for forest and agricultural land uses developed by the 

Commonwealth Department of Climate Change (DCC), formerly known as the Australian Greenhouse 

Office (AGO). FullCAM was chosen as the carbon model for the Land Carbon Project. This decision 

was made on the basis of allowing the project to leverage the credibility that FullCAM has while also 

maximising the ‘compatibility’ of project estimates with estimates contained in the national accounts. 

The results of FullCAM simulations were imported into the database for further analyses and linkage 

back to the GIS.



Map Products 

In order to generate map products, a series of tables from the database were linked to the GIS and from 

there maps showing the distribution of any variable, at any point in time, could be made. Figure 2 

shows an early draft map of the distribution of total onsite carbon as at 2005, with the impact of the 

2003 alpine fire event clearly evident. 

Figure 2: Sample map product. 

Analyses 

In order to minimise the number of calculations in FullCAM, the decision was made not to include any 

administrative data in the actual carbon model and only the biophysical data was used to describe the 

publicly managed land. Once the results of FullCAM were linked to the GIS it was then possible to 

intersect the carbon data with any administrative data layer to produce both maps and tabular data 

breaking down the FullCAM results. 

As an example, the layer of total onsite carbon stocks could be unioned with a layer of all the CMAs 

to calculate stocks for each CMA. 

Project Report 

The Project Team maintained a very strong commitment to ‘document as you go’, generating a series 

of procedural documents that served to record all the decisions made and actions taken for each 

discrete task. The task of preparing the Project Report was greatly simplified by the availability of the 

procedural documents. 

POPULATING THE MODEL 

The first stage in producing the carbon accounts was to assemble all the input data required to 

populate a model of the target estate. This task involved using the GIS to prepare layers that described 

harvesting history, fire history, forest cover, Interim Biogeographical Regions of Australia (IBRA) and 

the extent of Victoria’s publicly managed land. 

While almost all of the GIS processes were relatively unaffected by the number of records being 

processed, the Project Team was acutely aware that once data was passed to FullCAM then the 

number of records would have a huge impact as there is no capacity within FullCAM to ‘batch’



operations. If a particular process needed to be applied to each record (or Plot as they are called in 

FullCAM) then that process would have to occur manually – one at a time. For this reason there was a 

very strong emphasis on simplifying the data when working with the GIS layers. 

Harvesting History 

The DSE GIS library contains several layers that contain data on the extent and timing of harvest 

events dating back to the early 1930’s. Four key decisions were made in order to reduce the sheer 

volume of harvesting data. 

Firstly, it was decided that only the most recent harvesting event would be included in the model. 

Secondly, it was decided to aggregate harvesting events into decades. Thirdly, it was decided that 

only one type of harvesting event would be modelled. The alternative was to model thinning, group 

selection and clearfall events and to do so would triple the size of the overall model. 

Finally, it was decided that all records less than 10ha would be discarded. Again, this decision was 

taken to reduce the number of records in the harvesting layer and the figure of 10ha was chosen after 

examining the distribution of areas (see Figure 3). Removing small areas reduced the number of 

records to less than 9,800 records while removing only 2% of the area harvested. 

Figure 3: Distribution of harvesting records by area. 

No attempt was made to include planned future harvesting events in the model. 

Fire History 

One of the management events that was clearly identified as essential when determining the Land 

Carbon Project deliverables was the impact of planned burning on carbon stocks. The DSE GIS 

library contained several layers that differentiated between planned burns and wildfires and contained 

data going back to the 1930’s. 

As with harvesting data, a process was undertaken to simplify the fire history data 

Forest Cover 

Victoria’s publicly managed land comprises both forested and non-forested areas. The presence or 

absence of forest was assumed to be the single most significant influence on carbon stocks. 

The DSE GIS library contained a layer that represented the distribution of tree cover as at 1995. Areas 

described as non-forested in 1995 were assumed to have been non-forest throughout the model.



Interim Biogeographical Regions of Australia 

The IBRA layer (Natural Resource Management Ministerial Council, 2004) was used to assist in 

differentiating between areas with similar event histories but occurred in locations with distinctly 

different climatic, soil and vegetation characteristics. 

The IBRA layer was sourced from a Commonwealth Department of the Environment, Water, Heritage 

& the Arts website (http://www.environment.gov.au/metadataexplorer/explorer.jsp). 

Publicly Managed Land Extent 

The final GIS layer used to populate the model was one that would define the extent of the area to be 

included in the model. The relevant layer of the DSE GIS, containing data on all publicly managed 

land in Victoria was used. This layer also required simplification prior to being included in the spatial 

data model. 

Combining Layers 

Once the five GIS layers were in place they were unioned to create the Input Layer. Critically, the 

Input Layer was created as a multipart layer, meaning that all polygons with the same attribute set 

were stored in the attribute table as a single record. By using a multipart layer and the processes to 

simplify the harvesting, fire and public land data, the total number of records was kept to 822. 

Each record in the Input Layer then had its area, in hectares, calculated as well as the latitude and 

longitude, in decimal degrees, of the polygon’s centroid. These three values along with the record 

identifier would be passed to FullCAM. 

USING FULLCAM 

FullCAM is a highly complicated collection of carbon models and this paper considers FullCAM only 

from the perspective of a software user. FullCAM uses a proprietary file format as the basis for using 

the model. 

There is no capacity to ‘batch’ processes in order to treat a number of Plots in the same fashion. This 

is unfortunate for users wishing to generate Plot level data, as was the case with the Land Carbon 

Project where we required data for each Plot to be joined to the GIS in order to compile map products 

and perform spatial analyses (such as dividing carbon stocks by CMAs). 

Of equal significance is the inability to directly access the back-end database in instances where 

changes need to be made to large numbers of plots. Even a change that took only 30 seconds for each 

Plot represented almost a full day’s work when applied to all 822 Plots. 

Using Templates 

The Research edition of FullCAM supports the use of templates to set up Plots that hold settings and 

values that users wish to apply to a number of Plots. For example, template Plots could hold settings 

for a series of events and then all new Plots created from that template would all have the same event 

series. The Land Carbon Project used this capacity to ensure that all Plots with the same treatment 

histories were the same. Figure 4 illustrates the concept of templates. 

Configuring Templates 

The configuration of the template Plots was the most challenging aspect of the entire Land Carbon 

Project. All estimates of carbon stocks and their fluxes were driven by the way both the vegetation 

and the events applied to the sites were modelled. 

Plots in FullCAM, quite literally, have hundreds of variables that the user can adjust. Wherever 

practical, the Land Carbon Project Team accepted the default values of FullCAM and made changes 

only where deemed necessary. 

Each harvesting event is followed by a post-harvest burn event. The post-harvest burn event is 

assumed to impact on 80% of the harvested area. The settings used to describe combustion rates and 

transfers for the trees and debris are shown in Figures 5a and 5b respectively.



Figure 4: Using templates in FullCAM. 

Text File 

Template 

Plot Files 

Plot Area Lat Long Harvesting Fire 

Plot1 

Plot2 

Plot 

Plot2 

146. 

914. 

Area 

146. 

914. 

-36.22 

-35.92 

Figure 5a and 5b: Post-harvest burn impact on trees and debris. 

Lat 

-35.92 

Wildfire events were assumed to impact 80% of the nominated burnt area. The settings used to 

describe combustion rates and transfers for the trees and debris are shown in Figures 6a and 6b 

respectively. These settings are designed to represent a ‘typical’ wildfire in terms of severity. 

Figure 6a and 6b: Wildfire impact on trees & debris. 

70 

Plot 

Plot1 

Area 

Post-harvest Burn Impact on Trees 

85 

10 

Wildfire Impact on Trees 

12 

5 

18 

Lat 

-36.22 

Combusted 

To Debris 

Alive 

Combusted 

To Debris 

Alive 

+ 

= 

155.729 

154.836 

Long 

Long 

155.729 

154.836 

39 

58.5 

Harvesting 

1965 

Harvesting 

1965 

1965 

Fire 

1985 

Fire 

1985 

1985 

Post-harvest Burn Impact on Debris 

1 

Wildfire Impact on Debris 

1.5 

60 

40 

Combusted 

To Soil 

Unburnt 

Combusted 

To Soil 

Unburnt



Planned burn events were modelled to be more mild fires with lower combustion and transfer rates for 

both trees and debris. Planned burns are assumed to impact 50% of the nominated burnt area. 

During the development of the burning events (both wildfire and planned burns), a lack of stand 

response after the event was detected. This issue was raised with the DCC who advised that fire 

events in FullCAM did not contain any capacity for the stand to respond to the disturbance and that 

stand response was only modelled for harvesting events. 

To overcome this lack of stand response to burning, each burn event was immediately followed by a 

harvest event. The harvest event removed no volume but it’s presence ‘alerted’ the model to the 

changed stand circumstances that were present after the burn event and FullCAM subsequently 

modelled a stand response (even though it was, in theory, responding to the wrong event). 

Templates for each of the 247 different combinations of events were constructed. Managing the 

timing of events in FullCAM was important, as the chronological sequence of events impacted on the 

relative size of each pool at the time an event occurs. 

Event Timing 

FullCAM offers two options for managing the timing of events; events can be scheduled by either 

absolute dates or dates relative to the start of simulation. Both options have their pros and cons and 

the Land Carbon Project chose to use relative dates for all events. 

The Plots were configured to start simulation at ‘year 0’ (corresponding to the calendar year 1900) and 

run for 150 years. For example, a Plot with harvesting in the 1940’s and a wildfire in the 1960’s 

would have the harvesting scheduled for year 45 and the wildfire scheduled for year 65. 

After examining the model output it became clear that the model needed a longer period of time to 

reach stable levels of carbon in the soil pools. The starting time for all plots was then brought back 

400 years to allow the model to fully stabilise before any events were scheduled. Figure 7 shows how 

long soil carbon takes to stabilise. 

Figure 7: Soil carbon stocks vs. tree carbon stocks. 

Estate Disturbance 

Figure 7 also shows the impact of an event scheduled to occur in 1860. This event was included to 

provide disturbance of the estate prior to events occurring. Omitting this ‘early’ event meant that 

FullCAM was assuming all stands were untouched, fully mature forests, when in fact almost all of the 

forests in the model have been burnt or otherwise disturbed in the past. 

All reporting was confined to the period 1930 to 2050 because that includes the period for which there 

is event data (1935 to 2005) as well as a period over which stock trends could be seen. All stock 

estimates for earlier periods were discarded.



Model Validation 

One of the FullCAM model outputs is total stem volume. The model estimates of stem volume were 

compared with stem volume estimates derived from field measurements as a means of verifying the 

performance of FullCAM. 

A set of 18 Plots were selected for validation against State Forest Resource Inventory (SFRI) data. 

The SFRI data was collected between 1999 and 2002 and the data was updated in 2006. The 

validation was performed against the FullCAM estimate of stem volume for the year 2006. 

Figure 8 shows the initial plot of FullCAM and SFRI data, indicating that FullCAM was under 

estimating stem volume. 

Figure 8: Initial plot of FullCAM and SFRI data. 

NCAT modelled VolStems m3 ha-1 

350 

300 

250 

200 

150 

100 

50 

y = 0.3038x + 81.646 

R 2 = 0.1575 

0 

0 100 200 300 400 500 600 

SFRI VolStems m3 ha-1 

Although the regression statistics suggested that the FullCAM estimates should be tripled, this was not 

pursued. Instead, the FullCAM estimates were modified by introducing a Type 2 treatment (Snowdon, 

2002) immediately after the first disturbance event. The coefficient of change for the Type 2 

treatment was 1.2. This value was chosen because it allowed stands to return to their pre-disturbance 

volumes without exceeding them. Introducing the Type 2 treatment reduces the intercept from 81.6 to 

34.8 and increases the R2 from 0.15 to 0.21. 

The Project Team attempted to improve the regression statistics by adjusting the forest type according 

to the corresponding IBRA region (up to this stage all plots had been modelled as Medium Eucalypt 

forest). This did not lead to any improvement and the original forest typing was retained. 

Selecting Model Outputs 

FullCAM offers approximately 800 model outputs. The Land Carbon Project Team had to decide 

which model outputs to keep and which to ignore. To minimise the likelihood of making changes to 

the outputs (thereby necessitating significant volumes of work), the Land Carbon Project took the 

approach of including any value that was thought ‘might’ be of interest. As a result, 86 outputs were 

selected and this set of outputs was propagated to all Plots. 

Simulating Plots 

Each plot was simulated for the standard simulation period and the standard list of outputs selected. 

The model results were then saved as a Microsoft Excel file and the files were named after the Plot 

they represented.



UNCERTAINTY 

When providing estimates of modelled forest resource data it is common to provide an estimate of the 

level of uncertainty associated with those estimates. The normal practice is to calculate the standard 

deviation or coefficient of variation for a set of estimates of a single parameter such as stand volume. 

The Land Carbon Project is rather more complex, making use of data from various sources in a model 

system that is essentially a collection of other models (FullCAM makes use of elements of RothC, 

3PG and CAMFor). There is also the factor of scale, the reliability of the estimates of carbon stocks 

may be acceptable at the State level but unacceptable at the regional or stand level. 

Estimating Uncertainty 

Two methods of estimating uncertainty were applied. The first was to make use of the Sensitivity 

Analysis function of FullCAM and the second was for the Project Team to make an informed estimate 

of uncertainty (at the State level). 

In order to perform Sensitivity Analysis, a licensed copy of Palisade’s @Risk Developer’s Kit is 

required. During Sensitivity Analysis of four plots with different event histories, the values for key 

model drivers, such as proportion of canopy removed in harvesting, were systematically varied over 

thousands of simulations and the results were presented as a distribution of possible carbon stocks.The 

results of Sensitivity Analysis for each of the four selected Plots are shown in Table 2. 

Table 2: Sensitivity Analysis in FullCAM 

Plot Type PLE (90% confidence) Total area of plots with the same 

event history 

Harvesting and fire 24 467,317 

Harvesting and no fire 17 213,576 

No harvesting and fire 18 4,874,142 

No harvesting and no fire 18 1,696,618 

The Sensitivity Analysis did not include uncertainty associated with the model parameters and the list 

of elements included in the analysis was not exhaustive. The quantity of uncertainty associated with 

the model parameters within FullCAM was not pursued. 

By comparison, the Land Carbon Project Team’s estimate of uncertainty, which was settled prior to 

performing the FullCAM Sensitivity Analysis, was a PLE of 40% at 90% confidence. 

PROCESSING RESULTS 

After all analysis in FullCAM was complete, there were 822 Microsoft Excel files that contained 

results in a particular format. That format could not be imported directly into the Land Carbon 

Database and so some intermediate formatting steps were required. 

Once the results of FullCAM simulation were stored in the Land Carbon Database a series of queries 

were run to create a set of tables that would be joined to the GIS. Each of those tables contained data 

for a single FullCAM output as a time series with one record in the table for each Plot. Structuring the 

tables in this manner allowed these tables to be joined to the Input Layer of the GIS. 

All FullCAM output was expressed in total tonnes of carbon whereas tonnes of carbon per hectare 

were considered more suitable for map production. During the process of creating the new tables, all 

total carbon values were converted to per hectare values. 

Some additional processing was required to separate emissions attributable to planned burning or postharvest 

burning from those emissions from wildfire. The objective was to be able to report on 

anthropogenic and non-anthropogenic emissions. 

Joining to the GIS 

Once the set of GIS join tables had been created they were added to the ArcMap project file so that 

they were available to be joined to the Input Layer. Once a result table had been joined to the Input 

Layer it was possible for the user to select data from any year and use that as the basis for displaying



the result in a map. By saving maps from successive years as Portable Document Format (PDF) files, 

it was possible to create a ‘movie sequence’ showing changes in carbon stocks over time. 

Estate Analysis 

Information about the whole of Victoria’s publicly managed land was analysed in the Land Carbon 

Database where sets of queries were used to create tabular summaries for inclusion in the Project 

Report. Tabular summaries of total carbon stocks, stocks for individual pools, emissions by source 

and removals from the atmosphere were all created. 

Individual sets of accounts for ‘areas of interest’, such as State forest tenure were created using the 

GIS to ‘clip’ out those areas. Maps and tabular summaries for those areas were generated using the 

same methods as applied to the entire publicly managed land estate. 

RESULTS 

The results of the Land Carbon Project were not publicly available at the time of preparing this paper. 

CONCLUSIONS 

The provision of a set of carbon accounts led to an immediate request for further information and in 

that regard these initial accounts have served to sharpen focus on the management of carbon stocks on 

Victoria’s publicly managed land. 

There are clear policy implications for fire management and the heavy influence of fire events on total 

stocks generally support the notion that Australia would be unwise to sign up to Article 3.4 of the 

Kyoto Protocol. 

The Land Carbon Project Team acknowledges the limitations of these accounts but remain confident 

that they have met the objectives for which they were created. 

REFERENCES 

Natural Resource Management Ministerial Council (2004), Directions for the National Reserve System – 

A Partnership Approach, Australian Government, Department of the Environment and Heritage, 

Canberra, ACT 

Snowdon, P. 2002. Modelling Type 1 and Type 2 growth responses in plantations after application of 

fertiliser or other silvicultural treatments. Forest Ecology and management 163 (2002) 229-244



CARBON STORES AND BIOENERGY SUPPLIES: 

OPPORTUNITIES TO DEVELOP SUSTAINABLE PLANTATION 

INDUSTRIES IN LOW RAINFALL AGRICULTURAL LANDSCAPES 

John Tredinnick 1 and Jodie Mason 1 

ABSTRACT 

There is now a global focus on climate change that provides opportunities to develop 

financially viable industries on Australian agricultural lands. The opportunities will be 

assisted by new policies and regulation favouring low carbon fuels and sustainable land 

management practices. Planting of mallee crops on cleared land in central and southern 

Australia is ideally structured to meet these objectives. The mallee system focuses on the 

planting of belts of trees across the landscape, fully integrated with cropping and livestock 

production, while providing an alternative income source and providing public and private 

environmental benefits. Mallees therefore create a unique opportunity for the development 

of sustainable products, including wood products and bioenergy, as a dedicated carbon sink, 

or even providing all products from the same area of land. Current forest carbon policy 

frameworks support the creation of carbon forests, however these policies don’t allow 

accounting for the use of the harvested wood products, nor is there a direct benefit to the 

grower from the displacement of fossil fuels when biomass is harvested for energy creation. 

This paper suggests that the greenhouse and regional socio-economic benefits of a 

harvesting regime that leads to the substitution of fossil fuels or the retention of carbon in 

wood products can be greater than those from dedicated carbon sinks, and the greenhouse 

benefits more permanent. 

INTRODUCTION 

Sustainable and secure fuel and energy supplies will need to be developed in the future. There are 

very promising opportunities for meeting this demand from energy crops such as mallee eucalypts 

planted in the lower rainfall regions of Australia. Although there is some concern that energy crops 

will compete for land for the production of food, the mallee system can be structured to balance the 

demands for both energy and food production. 

For the industry to develop, the trees will need to provide at least an equivalent financial return to 

existing farm enterprises. Strategies for industry development also must seek to overcome three key 

impediments to the development of new industries: 

• The technologies and markets associated with the processing of mallees into industrial 

products on a large scale are not yet developed to the stage where any single product, or 

combination of products, provides a clear basis for economically viable development; 

• If clear opportunities were currently available, current mallee plantings are not at a 

sufficient scale for a processing industry to be developed; and 

• The demand for mallees as part of a dedicated carbon sink is far greater than that for 

industrial production. Under current policy settings this could result in suitable land 

being “locked up” on a non-harvest regime for more than 70 years. 

The first two impediments result in a “chicken and egg” scenario where the commitment from 

processors towards developing an industry is tempered by the lack of resource in the short term, and 

potential growers are looking for clear evidence of future profitability before they develop tree crops. 

1 URS Forestry, Level 3, 20 Terrace Road, East Perth, WA 6004, Australia. Email: john_tredinnick@urscorp.com



This paper draws on work undertaken as part of the development of the Oil Mallee Industry 

Development Plan for Western Australia (URS 2008) to summarise the silvicultural regimes that have 

potential to maximise productivity. It also describes the potential for mallee eucalypts to produce a 

number of commercial products and the potential economic returns from these products. It then 

illustrates how the impediments to industry development might be overcome by capitalising on the 

opportunities that can be provided by carbon trading. Such policy developments have the potential to 

provide a way forward for the industry that includes both carbon sequestration and the production of 

energy. 

RESOURCE DEVELOPMENT 

In NSW and Victoria a eucalyptus oil industry based on blue mallee (Eucalytpus polybractea) in 

native woodlands has existed for nearly 100 years. These traditional industries have also commenced 

their own planting to secure and expand the resource. 

Mallee plantings in the WA wheat belt region began in the early 1990s, based on the concept of 

developing a commercially viable tree crop to reduce agricultural salinity on the necessary scale 

(Bartle and Shea 2002). Woody crops like mallee, with their industrial products potential, were also 

seen as an option to diversify farm business and regional economies in the face of the long term 

decline in the terms of trade for agricultural products (Bartle 2006). 

The WA Department of Conservation and Land Management (CALM) 2 coordinated the first broad 

scale investment and establishment of oil mallee plantings in the period 1994-96 during which a total 

of six million mallee trees were planted (Bartle 2006). By 1995 there was a substantial interest in 

furthering industry development and a body of growers formed the Oil Mallee Association (OMA) 

(Bartle and Shea, 2002, OMA 2007). In 1997 the Oil Mallee Company (OMC) was established by the 

OMA to form a commercial arm for the oil mallee industry in WA (OMA 2007). Significant 

investment into mallee plantings for carbon sequestration in WA began in 2003, when the OMC 

established 1,000 ha of plantings across 24 farms for voluntary offsets purchased by a Japanese power 

producer, General Environmental Technos Co Ltd (previously Kansai). 

Both the OMA and OMC are still in operation today, furthering the expansion and diversification of 

the industry. Other companies that are focussed on the economic benefits of carbon trading have also 

driven an expansion of the resource base over the previous 12 months. This is likely to continue as a 

result of the current Federal Government’s commitment to a Carbon Pollution Reduction Scheme 

(CPRS) by 2011 has provided further impetus for the mallee industry in NSW and this is likely to 

follow in other states. 

Integration with Agriculture 

Mallees are readily combined with large scale annual cropping (Bartle et al 2007). However, to make 

a real contribution to improving agricultural productivity mallees have to be complementary to, not 

just compatible with, annual plant agriculture. Although water tends to be the most limiting resource 

in southern Australian, current farm management systems allow a proportion of rainfall to run-off or 

infiltrate below root zones and to be lost from production, leading to rising water tables and dryland 

salinity. A real complementary role for mallee is achieved when some of this water is captured. 

Carefully designed narrow belts of deep rooted perennials like mallee are being used for this purpose. 

Robertson et al (2006) and Sudmeyer and Goodreid (2007) have shown that belts of Eucalyptus 

species can develop a broad (up to 20 m either side) and deep (beyond 10 m) zone of dewatered soil 

that can theoretically be used as a sink for lateral flows and leakage from the adjacent annual plant 

agriculture (Figure 1). 

In the evaluation of mallee performance the interactions between mallee belts and adjacent annuals 

must also be taken into account. The strong lateral root systems of mallee in belts can impose 

competition on adjacent crop or pasture. 

2 Now the Department of Environment and Conservation (DEC)



Source: Huxtable and Bartle (2007) 

Figure 1: Water capture by mallees 

Optimising Yields 

The rate of growth of mallee plantings is affected by a number of variables. Apart from species and 

provenance, almost all of the factors are linked to water availability. These variables include climate 

(rainfall, temperature and evaporation); soil (depth, texture and structure); and the quality of site 

preparation and management – particularly weed and pest control. 

The most common belt configuration has been four row belts with a width of 2 metres (m) between 

rows. The effective belt width to the drip line of the trees under this configuration is 10 m 3 . The 

distance between mallee belts allows integration with agricultural practices and typical farm plans 

have the belts 40-100 m apart, depending upon farm machinery requirements. 

Current analysis by DEC (Bartle pers comm. 2008) shows that the effectiveness of mallee in water use 

means that economic factors will push down planting density and row number. It appears likely that a 

two row or even a one row belt may be the economic optimum from the biomass yield perspective, but 

other considerations like operational efficiency, cost, aesthetics or shelter benefit may work against 

this. Mallee water use effectiveness will also push design of belt layout into configurations that will 

facilitate better interception of lateral water flows. 

Huxtable and Bartle (2007) report that actual yields of above ground biomass range from 5-10 

bdmt 4 /ha/annum based on measurements at ages 7 to 11 years. Using these growth increments to 

adjust outcomes from Cooper et al (2005), Huxtable and Bartle (2007) suggest that productivity will 

lie in the range 5-15 bdmt/ha/annum over the rainfall range 350-550 mm pa. It should be noted that 

these yields have been standardised such that they are based on two row belts (i.e. where four row 

belts had been established only the two outer rows have been included in the productivity estimates). 

If a given area is to be established using four row belts the overall productivity is expected to be 

around 25% lower. Block plantings will not enable the same benefits to be concentrated in the area 

planted and may only achieve 30-50% of the productivity of two row belts over the area planted. This 

is shown diagrammatically in Figure 2. 

3 

This width meets Australia’s current definition of Kyoto compliant forests under Article 3.3 of the Kyoto Protocol 

4 Bone dry metric tonnes



Two row belts 

Assuming 7 bdmt/ha/annum of growth, yield at age 5 is expected 

to be: 

• 35 bdmt per net hectare planted 

• 1.75 bdmt per gross (paddock) hectare, assuming 50% of 

soils suitable and 10% of suitable soils planted. 

Source: URS (2008) 

Figure 2: Examples of mallee planting configurations and yields 

Four row belts 

Assuming 5.25 bdmt/ha/annum of growth (75% of two row 

productivity), yield at age 5 is expected to be: 

• 26.25 bdmt per net hectare planted 

• 1.3 bdmt per gross (paddock) hectare, assuming 50% of 

soils suitable and 10% of suitable soils planted. 

Block planting 

Assuming 2.8 bdmt/ha/annum of growth (40% of two row 

productivity), yield at age 5 is expected to be: 

• 14 bdmt per net hectare planted 

• 7 bdmt per gross (paddock) hectare, assuming 50% of soils 

suitable and all suitable soils planted.



USES AND POTENTIAL PRODUCTS 

A range of potential processing options and markets exists for mallees. The commercial viability of 

these options is dependent on a number of factors, including: 

• Costs of production from planting to processing; 

• Demand for products; 

• Prices that can be achieved in markets; and 

• Achieving necessary resource requirements (scale) for market development. 

Historically, there has been a significant focus on processing mallees to produce multiple products. 

Integrated wood processing (IWP) was considered as a production option in WA for its potential to 

efficiently direct specific biomass fractions to higher value products while directing residues to lower 

value uses such as bioenergy and eucalyptus oil, and achieve a commercially viable industry. Western 

Power (now Verve Energy) completed construction of a trial IWP plant in 2006 to undertake 

operational scale testing. During the trial phase the plant produced activated carbon from the wood 

fraction of the mallees and Eucalyptus oil from the leaves, while all residues, including spent leaf and 

waste heat, were used for bioelectricity (the production of electricity). The trial indicated that 

production of activated carbon, eucalyptus oil and energy from mallee biomass is feasible. 

Verve Energy is now investigating the potential to attract investment in the construction and operation 

of a larger plant and to replicate the plant in other areas of the WA wheat belt and other parts of 

Australia. The prospects for integrated processing are well developed and appear promising. The 

potential scale of the mallee resource is such that a range of additional markets should be considered 

as either stand alone industries or as part of an integrated production process. Table 1 provides a 

summary these products and markets. 

Table 1: Summary of market opportunities 

Product Opportunities and constraints 

Bioelectricity is the production 

of renewable electricity from 

solid biomass through direct 

combustion of wood, co-firing 

with other fuel sources, or via 

production of gaseous and liquid 

fuels. 

Carbon The carbon dioxide 

removed from the atmosphere 

and stored in biomass is 

recognised as a carbon sink 

which offsets greenhouse gas 

emissions and thereby helps to 

mitigate global climate change. 

Wood pellets are produced by 

grinding wood material into 

small particles, then 

compressing the material to bind 

the wood together. Energy is 

produced upon combustion. 

• Strong market opportunities exist in Australia for bioelectricity production. Small scale co-firing is 

being implemented around Australia and there are currently several proposals to develop dedicated 

bioenergy plants. 

• Opportunities for renewable electricity in Australia are expected to grow as a result of government 

policies, particularly the Commonwealth Government’s Mandatory Renewable Energy Target 

(MRET) scheme. The increased renewable energy target of 20% by 2020 being introduced by the 

Federal Government is likely to be a catalyst for significant investment in renewable energy 

infrastructure 

• The placement and size of bioenergy plants can be linked closely to the electricity grid in an area 

with excess network transmission capacity, and must have sufficient long term feedstock supplies 

within viable transport distances. 

• Distributed energy supply can also be provided to isolated projects. 

• The Federal Government established the Department of Climate Change and water (DCC) on 3 

December 2007. Shortly thereafter, the Australian Government officially ratified the Kyoto 

Protocol, and in doing so committed to limiting its national greenhouse gas emissions to 8% above 

1990 levels during the first Kyoto period of 2008 to 2012. The Government also committed to 

implement the CPRS in 2011. 

• Principal markets for wood pellets are in Europe, Canada, Japan and the United States. There is 

increasing interest from Australian investors to export wood pellets. Tests have shown that pellets 

manufactured from mallee will meet the standards required for the export market. 

• Export represents the clearest opportunity for Australian wood pellets in the short term, 

notwithstanding the significant transport distances and increasing domestic production in Europe. 

• Prices for renewable electricity in Australia may increase and potentially support a domestic market. 

• Domestic heating could also provide a market for pellets as a replacement for firewood as the pellets 

have proven energy efficiency and environmental advantages.



Product Opportunities and constraints 

Activated carbon is a form of 

charcoal that is valued for its 

ability to purify liquids and 

gases. It can be produced from 

wood through carbonisation and 

activation processes. 

Composite wood products such 

as Medium Density Fibreboard 

(MDF), particleboard and 

Engineered Strand Lumber 

(ESL) are used extensively in a 

range of applications in the 

construction, manufacturing and 

furniture sectors. 

Eucalyptus oil is produced from 

the leaves of trees and used as an 

ingredient in a range of products. 

Charcoal presents two 

opportunities: industrial carbon 

which is used in the steel 

industry as a substitute for 

coking coal or in the non-ferrous 

industry as a specialty reductant, 

and biochar which is used as a 

soil additive. 

Lignocellulosic ethanol is a 

liquid biofuel produced from 

wood by breaking down biomass 

to isolate the component sugars, 

which are then processed to 

produce ethanol. Its most likely 

immediate application is in the 

blended fuel product – E10. 

Biomass to Liquid fuels are 

produced when biomass 

undergoes a gasification process 

and gases are then synthesised 

into a range of liquid fuels such 

as methanol and synthetic diesel. 

Bio-oil is produced through a 

process of pyrolysis where 

woody biomass is heated in the 

absence of air. It is a 

combustible product and has 

potential as a substitute for fossil 

fuels in a range of applications. 

Technology has recently been 

commercialised overseas. 

ECONOMIC MODELLING 

• Opportunities exist to replace imports of activated carbon into Australia, particularly for use in the 

gold industry and water filtration systems where it is used in the recovery process. 

• Feasibility will depend on production costs, but the quality of the product that can be produced from 

mallees has proven to be suitable. 

• Australia is a significant exporter of MDF and consumption of composite wood products is likely to 

continue to increase in Australia. Tests have shown that mallee is a suitable input for the production 

of MDF. 

• Global markets for engineered wood products such as ESL are expanding and there are strong 

opportunities to supply the ESL market based on growing international demand for this relatively 

new product. 

• Most of Australia’s eucalyptus oil consumption is based on imports from China and there appears to 

be an attractive opportunity for domestically produced oil to replace imports. 

• There are strong precedents to the use of naturally occurring terpenes (such as those found in 

Eucalyptus oil) as industrial solvents and cleaners, albeit at significantly lower prices than current 

uses of Eucalyptus oil. 

• Eucalyptus oil is currently a niche product and large scale sales have the potential to significantly 

lower prices and reduce the viability of the industry. 

• It will be very difficult for industrial carbon produced from biomass to compete with the current 

coking coal industry, where supply is well established and low prices for coal are highly competitive. 

However, specialised activities requiring wood charcoal for reduction could provide a market for 

certain fractions of mallee. 

• Development of the biochar product is being undertaken, including trials to demonstrate the benefits 

of biochar as a soil enhancer and partial replacement for fertiliser. Other financial incentives, 

including official recognition of biochar as a tradeable carbon sink, will increase the potential of the 

market. 

• Lignocellulosic ethanol offers several advantages over traditional ethanol production from 

agricultural feedstocks (such as sugar cane and grain) in that it utilises a potentially cheaper 

feedstock and has a better energy balance as a result of lower greenhouse gas emissions during 

production. Lignocellulosic ethanol is also less likely to compete with human and agricultural food 

markets. 

• The lignocellulosic process is yet to be commercialised but is currently undergoing intensive research 

and development overseas. Several commercial prototypes of the technology are planned over the 

next 3-5 years. 

• Biomass to Liquid fuels present an opportunity to supply Australia’s diesel market. Growing 

demand and increasing prices for diesel will increase the viability of synthetic fuels. 

• Individual technologies are well understood, but integrated production requires further development 

and is a current constraint to the industry. 

• Markets for bio-oil are limited by the fact that it cannot currently be used as a fuel in existing forms 

of transportation. However, it can be used as a boiler fuel or in gas turbines for power generation. 

• Research and development is being undertaken to allow blending with hydrocarbon fuels which is 

hoped to broaden the overall market for the product. 

This section describes a model that was developed to allow the assessment of oil mallee production 

alternatives. The model was developed using principles of benefit-cost analysis to provide a 

discounted cash flow analysis of biomass production. The model allows comparison between



locations and areas with different growth potential, and between biomass production options at 

different scales. A 30 year timeframe has been used and a real, pre-tax discount rate of 7% has been 

assumed. Further details of the assumptions are outlined in URS (2008). 

Case study examples have been developed in the south west of Western Australia to provide a 

comparison between rainfall zones and areas adjacent to a port, rail or major electrical transmission 

line. The characteristics of the four examples are described in Table 2. 

Table 2: Characteristics of case study examples 

Region 

Rainfall 

(mm) 

Characteristics Potential industries 

On 

SWIS 1 

Near a 

port 

Near other 

biomass 

residue 

sources 

Pellet 

export 

Small bioenergy 

plant 

(5MW) 

Medium 

bio-energy 

plant 

(25MW) 

Engineered 

wood 

products 

Eucalyptus 

oil product 

Activated 

carbon 

products 

1 500-600 Yes Yes Yes Co-product Co-product 

2 400-500 No Yes Yes Co-product Co-product 

3 300-400 Yes No No Co-product Co-product 

4 300-400 No No No Co-product Co-product 

Very good potential Good potential Some potential 

Note 1. South West Interconnected System. The SWIS consists of nearly 88,000 kilometres of powerlines stretching from Kalbarri in the 

north to Kalgoorlie in the east and south to Albany. 

Table 3 shows the value of returns from biomass over a range of factory gate prices for mallee 

biomass and the value of carbon sequestered. 

Table 3: Annualised on-farm returns from biomass ($/ha pa) 

Factory gate value ($/gmt) 

Region 1 -$34.02 $35 $40 $45 $50 $55 

$0 -$3 $21 $44 $67 $90 

$20 $39 $62 $85 $108 $131 

Carbon value ($/t CO2e) $40 $80 $103 $126 $150 $173 

$60 $121 $145 $168 $191 $214 

$80 $163 $186 $209 $232 $255 

Region 2 $10.99 

$0 -$8 $10 $28 $46 $64 

$20 $24 $42 $60 $78 $96 


$60 $88 $106 $124 $142 $160 

$80 $120 $138 $156 $174 $192 

Region 3 -$53.93 

$0 -$41 -$28 -$15 -$2 $11 

$20 -$18 -$5 $8 $21 $34 


$60 $28 $41 $54 $67 $80 

$80 $51 $64 $77 $90 $103 

Region 4 -$64.11 

$0 -$36 -$23 -$10 $3 $16 

$20 -$13 $0 $13 $26 $39 


$60 $33 $46 $59 $72 $84 

$80 $56 $69 $82 $95 $107



The estimates of returns are presented as annualised values which are based on the net present value of 

returns from biomass plantings harvested regularly over a thirty year period. These returns should be 

compared against returns from agriculture or other land uses that might be achieved in each rainfall 

zone. 

In the analysis, the value of CO2e impacts directly on the value that the resource has as a carbon offset. 

In the case of wood fibre being supplied to a domestic energy producer, the value of CO2e will also 

have an impact on the factory gate value, as will the value of RECs and the value of any capacity 

credits 5 . The potential interaction between markets for CO2e, RECs and capacity credits is complex 

(see Garnaut 2008). As a general guide it can be assumed that as the value of CO2e increases, so will 

the factory gate price that can be paid by an energy producer for the mallee biomass. 

In making comparisons between agricultural returns and the returns from tree crops under various 

values of CO2e, it needs to be noted that there are also likely to be changes in agricultural production 

values that result from carbon trading if agriculture becomes a covered sector under the CPRS. These 

impacts are likely to be in the form of liabilities incurred by agricultural production and could make 

any advantages of tree crops more favourable. 

PROMOTING CARBON SEQUESTRATION AND BIOMASS PRODUCTION 

To a large extent, policy initiatives that support development of new products based on the mallee 

resource are already in train, particularly the federal Government’s proposed increase in renewable 

energy targets and the development of a national CPRS. Investors, processors and energy producers 

are expected to respond to these initiatives. Hence any proposed changes to policy need to be 

focussed on the development of a mallee resource that is available to be utilised by project proponents 

as technologies develop over the next five years. 

To do this, it is necessary to break the “chicken and egg” scenario of industry development versus 

resource development that is described in the introduction to this paper. Carbon trading provides the 

circuit breaker, but current carbon trading settings do not encourage optimal establishment of mallees 

for biomass production and carbon sequestration from the same area of land. This can be changed 

through initiatives that recognise that the objective of minimising greenhouse gas emissions can be 

achieved by both harvesting trees for the production of bioenergy or wood products and by creating a 

carbon offset credit from these products as well as the standing tree biomass. 

Under proposed guidelines for the CPRS, forest owners who “opt-in” will be offered units for the 

carbon sequestered on their land that are equivalent to the average amount of carbon in standing 

timber over the sequestration period after allowing for periodic harvesting. In the same way that there 

is increasing support for the inclusion of carbon stored in harvested wood products as offset credits, 

there must also be potential for the production of woody biomass for uses where it displaces fossil 

fuels, or the use of biochar as a soil additive, to generate an offset credit for the grower. 

Figure 3 describes the process of sequestration via fossil fuel displacement. The figure shows that 

there are ongoing greenhouse benefits from the biomass production process, in the same way that 

carbon can be sequestered in a dedicated sink. Furthermore, the greenhouse benefits of a harvesting 

regime that leads to the substitution of fossil fuels can be greater (per hectare planted), and more 

permanent, than the development of tree crops as dedicated carbon sinks. 

In proposing this benefit to mallee growers it is acknowledged that the use of woody biomass as a 

substitute for fossil fuels in energy production is likely to be recognised as a carbon neutral source of 

energy that reduces the potential liability of energy producers. It could be expected that these benefits 

will be recognised financially through market prices that are influenced by the value of RECs, the 

penalties to energy producers for carbon emissions under a future CPRS, or as a combination of the 

impacts of both factors. However, values under this implicit market will be dependent on the 

interaction of a number of supply/demand factors and energy producers will not necessarily pass on 

5 

A capacity credit is a measure of an electric power generator’s expected or actual contribution to meeting system reliability 

goals.



the full market value of carbon to the grower. 

Figure 3: Cumulative sequestration via fossil fuel displacement versus non-harvest sinks 

tC/ha 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

0 10 20 30 40 50 

Time (years) 

Recognition for fossil fuel displacement or wood products 

Above ground biomass (Harvest) 

Litter / debris 

Below ground biomass (Harvest) 

Change in soil carbon 

--- Non- harvest regime 

Source: URS (2008) 

Note: Figure and scale is indicative only and not intended to represent actual rates of sequestration 

Furthermore, without the direct recognition of fossil fuel displacement as an offset credit, large 

emitters of carbon dioxide (such as oil and gas producers) that establish mallees for offset credits will 

need to forego their accumulated credits at the time of harvest. They will then need to buy another 

offset on the market with the proceeds of sales from the biomass to meet their obligations under any 

future emissions cap. This exposes the organisation to a price risk associated with future carbon 

trading that they were seeking to avoid through the early action of planting mallees. 

CONCLUSIONS 

The development of an Australian CPRS should encourage parties wishing to establish offsets via tree 

crops to also participate in a future industry based on biomass harvesting. Under such a framework 

the resource owner would retain carbon credits at the time of harvest that are equivalent to the net 

greenhouse benefit from replacing fuels that release more CO2e in the production of energy or 

commodities. 

The advantages of such an initiative are significant and numerous: 

• Investors in carbon sinks would have an incentive to keep their options open for future 

harvesting; 

• A mallee resource would be created as a result of the demand for carbon offsets that also provides 

the basis for a future processing industry; 

• Periodic harvesting controls the height and spread of the mallee belt, keeps them in the more 

physiologically active young coppice stage of growth and moderates the competition imposed on 

adjacent crops and pastures; 

• In addition to environmental benefits of mallees, there is also the potential for new jobs and 

industries and associated socio-economic benefits in regional communities as a result of 

processing industries that would not arise from dedicated sinks; and 

• An increase in offsets/ sequestration per hectare planted.



ACKNOWLEDGMENTS 

URS Forestry would like to thank the following individuals and groups for their assistance, comments 

and contributions towards the development of the Oil Mallee Industry Development Plan for Western 

Australia: 

• Simon Dawkins (Oil Mallee Association) 

• Tim Emmott (Oil Mallee Association) 

• Paul Brennan (Forest Products Commission) 

• John Bartle (Department of Environment and Conservation) 

• Col Stucley (Enecon) 

• Jim Bland (Enecon) 

This Oil Mallee Industry Development Plan for Western Australia was funded by the Australian 

Government and the Government of Western Australia through the National Action Plan for Salinity 

and Water Quality 

REFERENCES 

Bartle, J.R. (2006). New Non-Food Crops and Industries for Australian Dryland Agriculture. Paper presented at 

the Green Processing Conference, Newcastle, NSW, 5-6 June 2006. 

Bartle J.R. and Shea S. (2002). Development of mallee as a large-scale crop for the wheat belt of WA. In 

'Australian Forest Growers 2002 National Conference: Private Forestry - Sustainable accountable and 

profitable'. Albany, WA, Australia pp. 243-250. 

Bartle J.R., Olsen G., Cooper D. and Hobbs T. (2007). Scale of biomass production from new woody crops for 

salinity control in dryland agriculture in Australia. International Journal of Global Energy, Issues 27 2: 

115-137. 

Cooper, D., Olsen, G. and Bartle, J.R. (2005). Capture of agricultural surplus water determines the productivity 

and scale of new low-rainfall woody crop industries. Australian Journal of Experimental Agriculture 

45:1369-1388. 

Garnaut R. (2008). Garnaut climate change review. Draft report June 2008. Report commissioned by the 

Commonwealth of Australia. 

Huxtable, D. and Bartle, J. (2007). .Predicting mallee biomass yield in the WA wheat belt. An interim report on 

investigations, Department of Environment and Conservation, February 2007. 

OMA (2007). The WA mallee story. A situation analysis prepared for the national mallee workshop, Oil Mallee 

Association, Canberra, 2007. 

Robinson N, Harper RJ, Smettem KRJ (2006). Soil water depletion by Eucalyptus spp integrated into dryland 

agricultural systems. Plant and Soil. 286:141-151 

Sudmeyer RA and Goodreid A (2007). Short rotation woody crops: a prospective method for phytoremediation 

of agricultural land at risk of salinisation in southern Australia. Ecological Engineering. 29: 350-361 

URS (2008). Oil Mallee Industry Development Plan for Western Australia. Report prepared for Forest Products 

Commission, Western Australia and the Oil Mallee Association of WA.



GIS-BASED CLASSIFICATION, MAPPING AND VALUATION 

OF ECOSYSTEM SERVICES IN PRODUCTION LANDSCAPES: 

A CASE STUDY OF THE GREEN TRIANGLE REGION 

OF SOUTH-EASTERN AUSTRALIA 

Himlal Baral 1* , Sabine Kasel 1 , Rodney Keenan 2 , Julian Fox 1 , Nigel Stork 3 

ABSTRACT 

This paper presents a GIS-based approach for classification, mapping and valuation of 

selected ecosystem services using market and non-market valuation techniques. First, we 

identified and compiled a variety of spatial and non-spatial data and developed a land 

cover typology of the study area into a GIS environment. Second, we estimate the annual 

flow of economic value of each service using various economic valuation techniques. We 

found that the economic value of market ecosystem services such as timber and carbon 

was relatively straightforward. However the quantification and valuation of non-market 

services such as biodiversity was complicated. Finally, we produced an annual flow of 

total economic value of sub-catchment G8 of the Lower Glenelg Basin, south-eastern 

Australia using the spatial economic valuation technique. We expect that this work will 

highlight research avenues to advance the ecosystem services framework as an 

operational basis in plantation dominant production landscapes in Australia and 

elsewhere. 

INTRODUCTION 

Forested ecosystems provide a variety of benefits for human beings – such as food, fibre, flood 

protection, clean water, and clean air. The benefits human populations derive, directly or indirectly, 

from ecosystem functions are labelled as ecosystem services (Costanza et al. 1998). According to this 

definition the goods and services are derived from the ‘functions’ and are utilised by ‘people’. The 

goods and services produced by the ecosystem are not always complementary, because the production 

of certain goods and services often results in the depletion or degradation of other goods and services. 

For example, timber extraction can affect visual quality, water quality and recreation. However, with 

proper management, there is a potential to integrate a variety of goods and services in production 

landscapes. The nature and typology of ecosystem services has been extensively elaborated elsewhere 

(de Groot et al. 2002; Boyd & Banzhaf 2007; Fisher & Kerry 2008; Wallace 2008; Fisher et al. 2009). 

The process of identifying, classifying, quantifying and mapping ecosystem services at the landscape 

level is increasingly recognised as an essential prerequisite for the efficient allocation of natural 

resources (Heal et al. 2005). Estimating the economic value of ecosystem services can play an 

important role in conservation planning and ecosystem-based management (Plummer 2009; Stenger et 

al. 2009). Conversely, a lack of economic valuation can potentially conceal the importance of such 

resources. However, economic valuation of ecosystem services requires up-to-date and reliable 

information and considerably better understanding of the landscapes that provide such services (Troy 

& Wilson 2006). 

We present a framework for classifying and mapping ecosystem services for a production landscape 

and assess the value of both market and non-market goods and services in a Geographic Information 

1 

Department of Forest and Ecosystem Science, University of Melbourne, 500 Yarra Boulevard, Richmond, Victoria, 3121, 

Australia. Email: h.baral@pgrad.unimelb.edu.au 

2 

Department of Forest and Ecosystem Science, University of Melbourne, Water Street, Creswick, Victoria, 3363, Australia. 

3 

Department of Resource Management and Geography, University of Melbourne, 500 Yarra Boulevard, Richmond, Victoria, 

3121, Australia.



System (GIS). This provides a baseline estimate of the ecosystem services value (ESV) of selected 

ecosystem services, such as timber, carbon, water regulation and biodiversity. 

CLASSIFICATION, MAPPING AND ECONOMIC VALUATION 

GIS as an ecosystem services mapping tool 

Ecosystem services are not homogenous across the landscape and their supply changes through time 

(Fisher et al. 2009). For this reason, ecosystem services are best expressed and most easily studied at 

particular spatial and temporal scales (MEA 2003). GIS provides land managers with a tool for 

quantifying and mapping the values of multiple ecosystem services across landscapes for improved 

resource planning and decision planning. Moreover, the valuation of ecosystem services, and 

understanding of how these resources interact, also provides an indication of trade offs and synergies. 

This paper will demonstrate how GIS can be used to analyse disparate data to generate spatially 

explicit results within a production landscape. 

Economic valuation of ecosystem services 

The knowledge and recognition of the importance of ecosystem services in providing benefits to 

society, and as the basis for the sustainable functioning of natural systems, have grown in recent years. 

Despite their importance, ecosystem services are yet to be incorporated into decision making (Chan et 

al. 2006). Many of these goods and services are not valued on markets and there is a gap between 

market valuation and the economic value of many ecosystem services (Stenger et al. 2009). 

Ecosystem goods and services can be divided into two broad categories: (i) the provision of direct 

market goods or services such as timber, pulp and carbon; and (ii) the provision of non-market goods 

or services, which include biodiversity and habitat for plant and animal life (Wilson & Hoehn 2006; 

Mertz et al. 2007). Economic valuation of the former is straightforward while the latter is more 

complicated and controversial (Wilson & Hoehn 2006; Stenger et al. 2009). However, resource 

managers and policy analysts involved in protecting and managing natural resources must make 

decisions which involve multiple trade-offs in allocating resources. These are mainly economic 

decisions and based either explicitly or implicitly on the value society places on services. To this end, 

economic valuation provides a useful tool for justifying set priorities and programs, policies, or actions 

that protect or restore ecosystem and associated services. 

Ecosystem valuation techniques 

Market price method: This method estimates the economic value of ecosystem goods or services that 

are bought and sold in commercial markets. This method can be used to value changes in either the 

quantity or quality of a good or service (Wilson & Hoehn 2006). In this study, this method was used to 

estimate the economic values of timber/pulp and carbon sequestration. 

Non-market valuation methods: Values for many ecosystem goods and services are not readily 

captured in market transactions, and thus require non-market valuation methods, such as travel cost 

method, hedonic approach, and contingent valuation (Wilson & Hoehn 2006; Stenger et al. 2009). 

Environmental value transfer: Value transfer is an accepted economic methodology which obtains 

an estimate for the economic value of non-market goods or services through work conducted at 

another site or group of sites (Troy & Wilson 2006). The ‘transfer’ itself refers to the application of 

economic values and other information from the original ‘study site’ to a ‘policy site’. This technique 

was used in this study to estimate the economic value of biodiversity and associated services. 

METHODOLOGY 

Study area 

The study location lies within the Green Triangle region of southern Australia, some 300 km west of 

Melbourne, Victoria, and covers approximately 305 km 2 (Figure 1). The area is known as subcatchment 

G8 of the Glenelg Hopkins Catchment. This area was selected because of its concentration 

of commercially valuable hardwood and softwood plantations where the focus is on production 

forestry with the provision of a variety of environmental services. 

The area has a variable annual rainfall of approximately 800 mm and mean annual temperature of 8 ºC 

(min) to 19 ºC (max). The altitudinal gradient ranges from 8–198 m above sea level. The major land



use in G8 is production forestry including both hardwood (Eucalyptus globulus) and softwood (Pinus 

radiata) plantations. (Refer .Figure 3).There are also some large areas of uncleared crown land 

(reserved forest) and limited grazing of sheep and cattle. Few social networks and services are 

available in sub-catchment G8 with Digby the only township, with a population of about 50 people. 

Community based groups include Rifle Downs Landcare, Miakite-Grassdale Tree Group, and the 

Smokey River Land Management Group. 

Fig. 1 Location of the study area – sub-catchment G8, Lower Glenelg Basin within The Green 

Triangle region of south-eastern Australia. Shading represents areas of high to low relief 

according to a digital terrain model. 

Methods 

The approach is based on a combination of market-based and non-market valuation techniques and has 

six key steps: (i) define the study area, identify the data requirement and compile data; (ii) develop 

land cover typology; (iii) select ecosystem services for economic valuation; (iv) map land cover and 

associated ecosystem services; (v) quantify goods and services and calculate economic value; and (vi) 

assign economic value in GIS and calculate total ecosystem service value (Figure 2). 

Fig. 2 Conceptual framework for the method of classification, mapping and economic 

valuation of ecosystem services used in this study. 

The first step was to identify both spatial and attribute data requirements, and then compile the data 

from a variety of sources. Datasets included: (i) plantation GIS data, which includes location, plant 

year, species, previous land use; (ii) ecological vegetation class (EVC); (iii) location of threatened



flora and fauna; (iv) a digital terrain model; and (v) topographic data such as contours, roads and 

watercourses. EVCs are systems of classifying native vegetation types based on differences in broad 

landscape features and environmental regimes (DNRE, 1997). EVC and plantation statistics provide 

an important overview of land cover types of the study area as vegetation is a highly visual component 

of the landscape and is an important part of timber resources as well as flora and fauna habitats. This 

study used the EVC dataset developed by the Victorian Department of Sustainability and 

Environment, and plantation GIS data available from the forestry companies represented in the study 

area (ITC, Timbercorp, Hancock Victorian Plantations, Wollybutt, Midway Afforestation, Great 

Southern Plantations, Green Triangle Plantation Forests, and Auspine). 

The second step entails development of land cover typology for the study area and starts with a 

preliminary analysis of available GIS data and EVCs. Land cover types from EVC and plantation data 

present in the study area are listed, a range of ecosystem services produced from each land cover type 

are identified and possible means of quantification and economic valuation techniques are reviewed. 

Ecosystem services identified in step two are screened in the third step in order to determine which are 

most relevant to the planning and decision making, and to set priorities for further assessment and 

valuation. Given the limited time and resources, it was not possible to assess in detail all the 

ecosystem services and assign economic value. 

In the fourth step, a land cover map is created using analysis tools in GIS which combine input layers 

from the diverse data sources to produce the final land cover map. Although significant variation 

within each land cover type may exist due to various factors, such as slope, aspect, and soil type, this 

level of detail was not measured in this preliminary study. Once each land cover type is finalised, then 

goods and services provided by each land cover type is quantified and economic value is assigned 

(Step 5). The economic value of ecosystem services was summed and cross-tabulated by each good 

and service and land cover type. Finally, a total spatial economic value map was produced (Step 6) by 

transferring ecosystem services value calculated in Step 5 using an ArcGIS 9.2 (from ESRI Inc.). The 

equation below illustrates the calculation of total ecosystem services value (ESV) in GIS environment 

(Troy and Wilson 2006). 

n 

V ( ES ) = ∑ A( LU )· V ( ES 

i i 

k = 1 

) 

ki ……(i) 

Where A(LUi) = area of land use/cover type (i), and V(ESki) = annual value per unit area of ecosystem 

service type (k) generated by each land cover type (i) 

Ecosystem services identified and valuation for this study 

Although there are large numbers of ecosystem services produced by the sub-catchment G8 (timber or 

fibre, carbon sequestration and climate regulation, water regulation, biodiversity and associated 

services, erosion control, salinity mitigation, tourism, recreation and cultural value), this study only 

focused on four goods and services – timber, carbon, water, and biodiversity and associated services. 

Timber: Plantation GIS data available from various forestry companies were used to identify and map 

the areas for timber production. Timber yield per hectare was calculated using a generic growth model 

(Fig. 4) available from Farm Forestry Toolbox (FFT) (Version 5.0, Private Forests Tasmania). Current 

timber value per unit area was estimated by multiplying the current stumpage price for hardwood and 

softwood logs and for pulpwood. Current market prices (2009 as a base year) were reviewed from 

Product Disclosure Statements (PDS) issued by various Managed Forestry Investment Scheme (MIS) 

forestry projects. The accuracy of growth and yield calculated from FFT was not validated with actual 

growth and yield data due to high level of confidentiality of inventory data. 

Carbon: A generic Carbon Sequestration Predictor (Version 3.1, New South Wales Department of 

Primary Industries) was used to estimate average carbon sequestration per hectare per year. Plantation 

GIS data were used to classify sites and determine the area and age of plantation. The average market 

value of current CO2 trading in Australian and international markets was used to calculate carbon 

value.



Biodiversity: Classifying, measuring and valuing biodiversity was a challenging issue for a number of 

reasons – (i) it was complicated by the wide spectrum of biotic scales at which biodiversity operates, 

ranging from the molecular specific to the ecosystem level; (ii) even for a given level of diversity, e.g., 

presence of certain threatened flora or fauna, there is no well established and agreed means for 

defining, measuring and valuing biodiversity; and (iii) a number of different indicators have been 

proposed which neither provide consistent nor comparable results on which to base general 

interpretations (Bene & Doyen 2008). 

As detailed mapping and valuation of biodiversity was not within the scope of this study, we were left 

with two alternatives – (i) transfer biodiversity values from other studies; or (ii) use an approximate 

dollar value employed by the Australian and various State government initiatives to conserve native 

vegetation on private land, such as ‘bush tender’ and ‘habitat hectare’ in Victoria, ‘biodiversity 

banking’ in New South Wales, ‘NatureAssist’ in Queensland, and ‘levy/biodiversity offset package’ in 

South Australia. The second option was used due to lack of comparable studies to transfer appropriate 

values for this site. The estimated value was transferred to each land cover type of native remnant 

vegetation. 

Water: Estimating accurate water use by plantations and various vegetation types is also a difficult 

issue. It is generally agreed that forest plantations use such a large quantity of water that downstream 

flow may be affected. Following harvest, water quantity increases but the quality may decrease where 

sediment discharge in the overland flow reaches stream channels. 3-PG, a simple, process-based forest 

growth model, developed by Landsberg and Waring (1997), was used estimate annual plantation water 

use per unit area. The average water harvesting charge per megalitre available from the Independent 

Pricing and Regulatory Tribunal (IPART) NSW was used to estimate the cost of water utilised by 

forest plantations. 

ESV transfer 

The ecosystem service value of selected ecosystem services, estimated by market and non-market 

valuation techniques,, was transferred to the GIS and the total economic value of selected ecosystem 

services was calculated using equation (i) above. 

RESULTS 

The landcover map of sub-catchment G8 is dominated by ‘heathy woodlands’ (27.4 percent) followed 

by ‘plain woodlands’ (19.3 percent). Hardwood and softwood plantations cover approximately 18 

percent of the sub-catchment (Table 1 and Figure 3). 

Table 1. Land cover typologies (and proportion of area occupied) for sub-catchment G8, Lower 

Glenelg basin 

• Heathy woodlands (27.4%) 

• Plain woodlands (19.3%) 

• Agriculture/pasture (18%) 

• E. globulus plantation (10.2%) 

• Herb-rich woodlands (8.8%) 

• P. radiata plantation (7.3%) 

• Lowland forests (3.9%) 

• Riparian or swampy scrubs (2.4%) 

• Heathlands (1.9%) 

• Riverine grassy woodlands (0.5%) 

• Wetlands (0.3%) 

The values of the selected ecosystem services – timber, carbon, biodiversity and water – have 

distinctly different spatial distributions, although some areas are of high value for multiple services 

while others are low value (Fig. 5). For example, the preliminary assessment and economic valuation 

of these ecosystem services indicate plantations represent the highest economic value per unit area. 

This is mainly due to the relatively detailed valuation of consumptive services available from 

commercial plantation areas. The areas with higher timber yields also produce more carbon which 

further increases the economic value per unit area.



Fig. 3 Land cover map of sub-catchment G8, Lower Glenelg Basin. 

Fig. 4. Site productivity (top) and mean annual increment (bottom) for Pinus radiata (4a, b) and 

Eucalyptus globulus (4c, d) used to estimate timber yield ha -1 ,(where, MDH = mean 

dominant height, MDDob = mean dominant diameter over bark and MAI = mean 

annual increment.)



Fig. 5 Annual flow of selected Ecosystem services of sub-catchment G8, Lower Glenelg Basin, 

the large light grey area indicates ESV less then $1350 ha -1 yr -1 or no data available at 

this stage. 

The value of native vegetation is relatively low due to our use of the conservation value based on 

various government initiatives to conserve native vegetation in private land (Section). This does not 

reflect the true value of all native vegetation and the value of a number of other services such as 

erosion control and water regulation are not assessed in this study. 

The land cover map (Fig 3) was based on the most recent, updated plantation and native vegetation 

datasets. However, improved estimates for growth and yield, together with improved conservation 

values, will permit further refinement of our valuation of ecosystem services for sub-catchment G8. 

DISCUSSION 

The quantification and valuation of timber and carbon were relatively straightforward due to readily 

available datasets. Similarly, the accuracy level is relatively high because prices are well-defined, and 

real data and accepted economic techniques were used. However, some issues, such as the true 

economic value of goods or services, may not be fully reflected in market transactions which are 

subject to market imperfections, seasonal variations and unforeseen impacts on markets. 

Other services which are not readily bought and sold in the market place, such as biodiversity and 

water quantity and quality, are difficult to quantify and value. The complexity and uncertainty 

underlying the functioning of biodiversity contribute to the difficulty in assessing such services (Bene 

& Doyen 2008). Although it is difficult to put an accurate price tag on such services, it is important to 

estimate the value of these ecosystem services because of the growing environmental market. 

While the current global financial downturn threatens markets for timber resources, eco-products are 

gaining in market popularity due to perceived environmental challenges, such as drought, declining 

supplies of drinking water, and loss of biodiversity (Metz et al. 2007). The economic valuation of 

ecosystem services provides better understanding of these forms of natural capital, and this is critical 

to our ability to plan and manage it over long term. 

The critical underlying assumption of the value transfer approach is that the economic value of 

ecosystem goods or services at the study site can be inferred with sufficient accuracy from the analysis 

of existing valuation studies. Clearly, as the level of information increases within the source literature, 

the accuracy of the value transfer improves. For this reason we are currently compiling recent forest 

ecosystem valuation studies in Australia as most existing datasets are outdated.



We are also in the process of accessing CABALA (CArbon BALAnce), a more advanced model for 

predicting forest growth and carbon sequestration in plantations and managed forests developed by 

CSIRO, and anticipate that tree growth and carbon estimation will improve as a result. 

CONCLUSIONS 

This study presented a simple GIS-based process of classification, mapping and valuation of 

ecosystem goods and services at the landscape scale. Rather than a single methodological approach, 

this study used a number of tools to estimate ESV using market and non-market based valuation 

approaches in a spatially explicit manner. Although the study is still in its early stages and does not 

account for all non-market ecosystem services, it demonstrates an important contribution of such 

services at a landscape level. The approach and methodology can be used for more precise future 

mapping for all ecosystem goods and services, including agriculture, pasture and other forest goods 

and services. 


The research was a part of PhD study (H. Baral) funded by The University of Melbourne and The 

CRC for Forestry. The GIS data were provided mostly by the Victorian Department of Sustainability 

and Environment, while plantation GIS data were supplied by forestry companies through the South 

East Resource Information Centre. Additional geospatial data were available from the Bureau of 

Meteorology, Geosciences Australia, and the Glenelg Hopkins Catchment Management Authority. 

REFERENCES 

Bene, C & Doyen, L 2008, 'Contribution values of biodiversity to ecosystem performances: A viability 

perspective', Ecological Economics, 68 (1-2), 14-23. 

Boyd, J & Banzhaf, S 2007, 'What are ecosystem services? The need for standardized environmental accounting 

units', Ecological Economics, 63 (2-3), 616-626. 

Chan, KMA, Shaw, M, Cameron, DR, Underwood, E & Daily, G 2006, 'Conservation Planning for Ecosystem 

Services', Plos Biology, 4 (11), e379. 

Costanza, R, d'Arge, R, de Groot, R, Farber, S, Grasso, M, Hannon, B, Limburg, K, Naeem, S, O'Neill, RV, 

Paruelo, J, Raskin, RG, Sutton, P & van den Belt, M 1998, 'The value of ecosystem services: putting the 

issues in perspective', Ecological Economics, 25 (1), 67-72. 

de Groot, RS, Wilson, MA & Boumans, RMJ 2002, 'A typology for the classification, description and valuation 

of ecosystem functions, goods and services', Ecological Economics, 41 (3), 393-408. 

Fisher, B & Kerry, T 2008, 'Ecosystem services: Classification for valuation', Biological Conservation, 141 (5), 

1167-1169. 

Fisher, B, Turner, RK & Morling, P 2009, 'Defining and classifying ecosystem services for decision making', 

Ecological Economics, 68 (3), 643-653. 

Heal, GM, Barbier, EB & Boyle, KJ 2005, Valuing Ecosystem Services: Toward Better Environmental Decision- 

Making, Washington, DC. 

MEA 2003, Ecosystems and Human Well-being: A framework for Assessment, Island Press, Washington DC. 

Mertz, O, Ravnborg, HM, Lovei, G, Nielsen, I & Konijendijk, CC 2007, 'Ecosystem services and biodiversity in 

developing countries', Biodiversity and Conservation, 16 (10), 2729-2737. 

Metz, B, Davidson, O, Bosch, P, Dave, R, and Meyer, L 2007, Climate Change 2007: Mitigation of Climate 

Change, Cambridge University Press, London. 

Plummer, ML 2009, 'Assessing benefit transfer for the valuation of ecosystem services', Frontiers in Ecology 

and the Environment, 7 (1), 38-45. 

Stenger, A, Harou, P & Navrud, S 2009, 'Valuing environmental goods and services derived from the forests', 

Journal of Forest Economics, 15 (1), 1-14. 

The former DNRE,1997, Victoria’s Biodiversity: Directions in Management, Department of Natural Resources 

and Environment, East Melbourne. 

Troy, A & Wilson, MA 2006, 'Mapping ecosystem services: Practical challenges and opportunities in linking 

GIS and value transfer', Ecological Economics, 60 (2), 435-449. 

Wallace, K 2008, 'Ecosystem services: Multiple classifications or confusion?', Biological Conservation, 141 (2), 

353-354. 

Wilson, MA & Hoehn, JP 2006, 'Valuing environmental goods and services using benefit transfer: The state-ofthe 

art and science', Ecological Economics, 60 (2), 335-342.



SELECTING SPECIES FOR CARBON SEQUESTRATION UNDER 

CLIMATE CHANGE SCENARIOS IN SUBTROPICAL QUEENSLAND 

David Lee 1, 2 , John Huth 1 , David Osborne 1 , Bruce Hogg 1 

ABSTRACT 

Optimal matching of species to sites is required for a sustainable hardwood plantation 

industry in the subtropics. This paper reports on the performance and adaptation of 60 

taxa (species, provenances and hybrids) across two rainfall zones and a range of soil types 

in southern Queensland. Specifically, taxa performance is compared across five 

replicated taxa-site matching trials at age six years. Three trials are in a 1000 mm rainfall 

zone of the Wide Bay near Miriam Vale and two in a drier (c. 750 mm) rainfall zone near 

Kingaroy in the South Burnett. 

In the higher rainfall zone, the taxa with the fastest growth in the three trials at age six 

years are Corymbia citriodora subsp. variegata Woondum provenance ranked 1 st 6 th and 

5 th , E. longirostrata Coominglah provenance ranked 3 rd 2 nd and 3 rd respectively; and two 

sources of E. grandis Copperlode provenance (ranked 4 th and 1 st ) and SAPPI seed orchard 

(ranked 6 th and 4 th ) planted in only two of the three trials. Similarly in the lower rainfall 

zone, E. grandis and its hybrids appear promising from the growth data; however, this 

excellent early growth has not continued in either rainfall zones, with these taxa now 

(over eight years old) showing signs of stress and deaths. Based on trial results in these 

two rainfall zones, the taxa that appear the most promising for sustainable plantation 

development with high average annual volume indexes and low incidence of borer attack 

is Corymbia citriodora subspecies variegata (6.7 m³/ha). E. grandis and E. longirostrata 

both have better average annual volume indexes (8.2 m³/ha and 7.4 m³/ha respectively) 

but are highly susceptible to borer attack. The current and long term productivity and 

sustainability of plantation forestry in these rainfall zones is discussed. Further, the 

implications of predicted climate change (particularly reduced rainfall) on growing trees 

for fibre production and carbon sequestration are explored. 

INTRODUCTION 

The expansion of commercial hardwood plantations in Queensland has been dramatic over the past 

decade from 400 ha in 1997 to 49400 ha in 2007 (ABARE, 2008). This has come at a time when there 

has been an escalation in the price of land in the higher rainfall zones (>1000 mm) and therefore a 

reduction of the amount of land available for economically viable plantation establishment. This has 

resulted in hardwood plantation growers expanding into lower rainfall zones which have traditionally 

been regarded as marginal for production forestry. The problem faced by the plantation growers is 

that, while there is potentially a large number of species and inter-specific hybrids that may grow in 

these regions, there is little long-term growth, carbon sequestration, wood quality and pest tolerance 

data available on which to base species selection and estimate potential plantation productivity and 

profitability. To address these problems 150 taxa (species, provenance and hybrid) trials have been 

established over the past 12 years on a range of site types across Queensland and northern New South 

Wales. 

This paper presents the results from five large trials (11952 trees) in south-east Queensland at age six 

years. Three trials are in a moderately high rainfall zone (1000–1200 mm) in the Wide Bay region 

near Miriam Vale, and two are in the lower rainfall zone (750–800 mm) in the South Burnett region, 

near Kingaroy (Figure 1). 

1 

Horticulture and Forestry Science, Queensland Primary Industries & Fisheries, LB 16 Fraser Road, Gympie Qld 4570, 

Australia. 

2 

Faculty of Science, Health and Education, University of the Sunshine Coast, Maroochydore DC Qld 4558, Australia. Email: 

dlee@usc.edu.au.



MATERIAL AND METHODS 

Study sites 

The five trials were established over two years. Two trials in the Wide Bay were planted in 2000 

(trials 506a and 506b) and a third (trial 506c) was planted in 2001. Both trials in the South Burnett 

(523c and 523d) were planted in 2000 (Table 1, Figure 1). All trials were planted on sites that were 

being used by commercial hardwood plantation growers. Site details are presented in Table 1. Before 

planting each trial site was deep ripped, mounded and sprayed with pre-plant herbicide. After 

planting, weed free conditions were maintained in the planting row until the end of the following wet 

season. 

Table 1. Site descriptions for the five taxa trials. 

Trial Australian Soil 

Classification 1 

Wide Bay trials (higher rainfall) 

506a Red Chromosol to Yellow 

Dermosol 

506b Red Chromosol to Yellow 

Dermosol 

506c Yellow Chromosol and 

Black Vertosol 

Latitude 

(°S) 

Longitude 

(°E) 

Mean annual 

rainfall (mm) 

Elevation 

(m asl) 

Average annual 

rainfall during 

trial (mm) 2 

24.60 151.64 1033 65 877 low 

Frost 

risk 

24.60 151.64 1033 55 877 medium 

24 38 151 52 1138 90 868 medium 

South Burnett trials (lower rainfall) 

523c Red Ferrosol 26.36 151.72 786 440 598 high 

523.d Red Dermosol 26.55 151.73 760 480 594 high 

1 Isbell (1996) 

2 Interpolated daily observations from the SILO database see: http://www.longpaddock.qld.gov.au/silo/ (Jeffrey et al., 2001) 

Fertiliser was applied to all trials shortly after planting. Trials in the Wide Bay were fertilised with a 

low dose of fertiliser (2.1 kg/ha of nitrogen, 3.7 kg/ha of phosphorus, 2 kg/ha of potassium and 3.8 

kg/ha of sulphur, plus trace elements). The very low levels of elements were used by the plantation 

owner as a base dressing as these trials were established on ex pasture sites that had a high number of 

fertiliser applications over past years (exact details are not known). The trials in the South Burnett 

were given a higher dose of fertiliser (72.6 kg/ha of nitrogen, 89.5 kg/ha of phosphorus, 58.6 kg/ha of 

potassium and 8.6 kg/ha of sulphur plus trace elements. 

The rainfall over the period since planting has been much lower than the long-term average with the 

Wide Bay trials receiving 84 per cent and the South Burnett trials 76 per cent respectively (Table 1). 

Experimental design 

Each trial was established as a randomised incomplete block design with three replications. In each 

trial, one or more provenances of best-bet species and or hybrids (Eucalyptus, Corymbia) as well as 

traditional soft wood plantation species of south-east Queensland (Araucaria and Pinus) were tested. 

Allocation of taxa to trials was based on expert opinion of the potential of the taxa as well as plant 

availability. Details of the 60 taxa planted in the trials and their origins are given in Table 2; this table 

also details the abbreviations used to describe the taxa throughout this paper (e.g. acmen -neer). Six 

taxa: ccc -glad, ccv -brooy, ccv –woon, g × c -dend, g × u -csir and longi -coom were planted across 

all trial sites. Plot sizes varied with site due to seedling and land availability. All measure plots had 

isolations of the same taxa planted around the outside of the plot. Trials 506a and 506b were planted 

approximately 800 m apart, on the same property with 32-tree measure plots (4 rows × 8 trees). Trial 

506c was established with 40-tree measure plots (4 rows × 10 trees) and trials 523c and 523d with 20tree 

measure plots (4 rows × 5 trees). Trials in the 506 series were planted at 794 stems per hectare 

whereas the two trials in the 523 series were planted at 1111 stems per hectare. Trials were un-thinned 

at the time of the latest measure.



Measurement and assessments 

Survival was assessed and total height and diameter at breast height over bark were measured for all 

surviving trees in each trial at age six years. As no specific volume equations were available for these 

species a volume index (VI) was calculated using the following formula: 

VI (m³) = [1/3 tree height (m) × basal area at 1.3 m (m 2 )] 

In trials 506a, 506b and 506c the incidence of borer attack was also assessed at age six years (presence 

or absence of longicorn beetles Phoracantha spp. or giant wood moths Endoxyla cinerea). 

Figure 1. Location of trial sites in the Wide Bay (series 

506) and Burnett (series 523) regions of subtropical 

Queensland referred to in the paper. 

Statistical analysis 

Analysis of variance of the data, for each 

trial, was carried out using the Genstat 

Version 9.2 software package with a 

randomised complete block design 

model: 

Yijk = μ + Ti + Rj + TRij+ Eijk 

where, Yij = the observation on the i th 

taxa in the j th replicate; μ = overall mean; 

Ti = the effect of the i th taxa; Rj = the 

effect if the j th replicate, TRij is the 

interaction between the i th taxa and j th 

replicate and Eijk the random error 

associated of the ijk th observation. This 

model was considered appropriate as 

most taxa (species and hybrids) did not 

have nested taxonomic groupings 

(multiple provenances or seedlots). 

Arcsine transformation of binary data 

(survival and borer incidence) was 

performed, but did not alter the result of 

the analysis. Post hoc Fischer’s 

Protected least significant difference test 

was applied if significant differences 

were detected by ANOVA. Means are 

reported, and treatment differences or 

interactions were regarded as significant 

at P



significant attack (16 to 100 per cent; Table 3). E. grandis and its hybrids were heavily attacked by 

borer with incidence ranging from 44 to 100 per cent. E. urophylla and its hybrids were also highly 

susceptible to borer attack with incidence ranging from 89 to 100 per cent. Mean survival of the six 

taxa represented in all trials was similar, ranging from 79 to 84 per cent. Across the trials survival of 

the taxa ranged from 6 to 98 per cent. 

The average VI of the six taxa planted across all the trials (ccc –glad, ccv –brooy, ccv –woon, g × c – 

dend, g × u –csir and longi –coom) was calculated to provide an indication of site potential. Based on 

this, trial 506a had the lowest site potential at age six (24.5 m 3 /ha) followed by 523d (26.0 m 3 /ha), 

506b (28.5 m 3 /ha), 506c (48.4 m 3 /ha) with trial 523c exhibiting the best site potential at age six (55.3 

m 3 /ha; Table 3). 

Performance of taxa in trials in the high rainfall region 

Volume index in trial 506a at age six years ranged from 2.8 m³/ha for uro -mt.egon to 28.5 m³/ha for 

ccv -woon, which along with seven other taxa formed a group that had significantly better growth than 

the other 21 taxa in the trial (Table 3). Survival ranged from 19 per cent for uro -mt egon to 92 per 

cent for g × c -dend. Incidence of borer attack ranged from 0 per cent for ccv -woon, ccv -brooy, cloez 

-home, cloez -mungy and sphaero -bla to 100 per cent for uro -mt ergon. Incidence of borers in E. 

grandis and E. grandis hybrids ranged from 50 to 97 per cent while that of E. longirostrata ranged 16 

to 41 per cent across the three trials. 

Trial 506b was on the same property as trial 506a but planted in a lower position in the landscape 

(creek flats) 800 m away. In this trial at age six years VI ranged from 0.1 m³/ha for sphaero -bla to 

45.6 m³/ha for grandis -cop, which along with longi -coom and g × r -csir had significantly larger 

volumes than the rest of the taxa in the trial. Survival ranged from 6 per cent for sphaero -bla to 91 

per cent for the g × r -csir seedlot. Incidence of borer attack ranged from 0 per cent for ccv -brooy to 

100 per cent for the g × r -csir seedlot. Borer attack in E. grandis and its hybrids ranged from 44 to 

100 per cent. 

Six year VI in trial 506c was generally higher than the other two trials in the higher rainfall zone (as 

indicated above in the section on site potential). In this trial age six years VI ranged from 9.0 m³/ha 

for sphaero -bla to over 90 m³/ha for grandis -koo and grandis -dpi, which had a significantly larger VI 

than the rest of the taxa in the trial. Survival ranged from 28 per cent for maid -bondi to 98 per cent 

for pch. Borer incidence in trial 506c was lower than the other two trials in the higher rainfall zone 

ranging from 0 per cent for 12 taxa including the six Corymbia taxa to just over 90 per cent for grandis 

-koo and g × u -csir. 

Performance of taxa in trials in the low rainfall region 

Thirty four taxa were planted in trial 523c. Age six VI ranged from 2.6 m³/ha for hoop to 84.3 m³/ha 

for saligna -clo. At age six years this latter clone, along with grandis-wed, grandis-dpi and g × r-csir 

seedlots (71.3, 72.1 and 74.4 m³/ha respectively) had significantly higher VI than the other taxa in the 

trial. The VI for E. grandis and its hybrids at age six years was in the range 41.2 to 74.4 m³/ha. The 

VI of the Corymbia species and hybrids in this trial ranged from 38.7 to 57.6 m³/ha. Survival ranged 

from 32 per cent for u hybrid -mix to 98 per cent for siderox -ural. Borer attack was not recorded at 

this site or the other site in the South Burnett but anecdotal observations indicated that borers are an 

issue for plantations in these regions. 

Trial 523d had 28 taxa planted. Six year VI ranged from 0.2 m³/ha for hoop to 33.0 m³/ha for maid - 

bolaro which did not have a significantly better VI than 12 other taxa in the trial. The VI of the 

Corymbia species ranged from 23.2 to 31.7 m³/ha and for E. grandis and its hybrids the range was 

17.5 to 32.0 m³/ha. Survival ranged from 43 per cent for hoop to 98 per cent for the g × c -dend 

clones. 

Trends across the five trials are given in the discussion.



Table 2. Details of seedlots and clones planted in the trials. If the seedlot number is followed by an ‘a’ it came from 

the Australian Tree Seed Centre, otherwise it came from the Forestry Tree Seed Centre, Queensland. The 

details in the taxa column are used in Table 3 to describe the taxa. This table is sorted by the taxon 

abbreviation. 

Species / hybrid Provenance Taxon abbreviation 

Seedlot / 

Clones identity 

No. seed 

parents 

Latitude 

(°S) 

Longitude 

(°E) 

Eucalyptus acmenoides Neerdie acmen -neer 10804 / 15874 >7 25.98 152.77 70 

E. acmenoides Woolgoolga acmen -woolg 15534a 8 30.12 153.17 80 

E. argophloia Ballon argo -ballon 5520 18 26.50 150.92 350 

Corymbia hybrids QPIF 1 CP 2 families C hybrid – 12 – – – 

C. citriodora subsp. citriodora Gladstone ccc -glad 20016a 8 – – – 

C. citriodora subsp. citriodora Yeppoon ccc -yepp 4971 / 11246 >7 23.10 150.73 30 

C. citriodora subsp. variegata Brooyar ccv -brooy 10248 12 26.17 152.50 90 

C. citriodora subsp. variegata Woondum ccv -woon 5567 21 26.25 152.82 40 

C. henryi Lockyer ch -lock 10250 10 27.47 152.28 150 

C. henryi Nerang ch -ner 10257 11 27.98 153.32 100 

E. cloeziana Home cloez -home 4363 8 26.05 152.70 220 

E. cloeziana Mungy cloez -mungy 10823 20 25.28 151.35 

E. cloeziana South African CSO 3 cloez -saso – – – – – 

E. dunnii SAPPI 4 dunnii -sapp 10513 unknown – – – 

E. grandis Wedding Bells grandis -wed 10542 18 30.17 153.12 100 

E. grandis Copperlode grandis -cop 5970 28 16.97 145.67 425 

E. grandis Mixed selects grandis -dpi – 5 – – – 

E. grandis Koorlong grandis -koo 20261r unknown – – – 

E. grandis SAFCOL 5 grandis -saf 10520 >10 – – – 

E. grandis SAPPI grandis -sap 10519 unknown – – – 

E. grandis × E. camaldulensis CSIR 6 g × c -csir 11307 unknown – – – 

E. grandis ×E. camaldulensis Dendros clones g × c -dend – 9 clones – – – 

E. grandis ×E. pellita QPIF CP families g × p -mix 2000 g × p 15 – – – 

E. grandis ×E. resinifera CSIR g × r -csir 5991 / 11304 10 – – – 

E. grandis ×E. tereticornis CSIR g × t -csir 10502 / 11306 unknown – – – 

E. grandis ×E. tereticornis CSIR and DPI&F CP g × t -mix – 10 – – – 

E. grandis ×E. urophylla Dendros clones g × u -clone – 6 clones – – – 

E. grandis ×E. urophylla CSIR g × u -csir 11302 / 11301 unknown – – – 

E. grandis hybrid mix g × t, g × u and g × p hybrid -mix – unknown – – – 

Araucaria cunninghamii FPQ 7 CSO seed hoop – unknown – – – 

E. longirostrata Coominglah longi -coom 19312a / 11153 5 24.92 151.00 400 

E. macarthurii SAPPI mac -sappi 10854 unknown – – – 

E. globulus subsp. maidenii Bolaro maid -bolaro 18728a 78 35.67 150.03 380 

E. globulus subsp. maidenii Bondi maid -bondi 19454a 24 37.18 149.46 480 

E. moluccana Ravenshoe mol -ravens 5696 10 17.60 145.48 913 

Pinus caribaea var. hondurensis FPQ Kennedy CSO pch pch unknown – – – 

E. pellita Kuranda pell -kur 5081 15 16.75 145.57 440 

E. pellita Ellerbeck SSO 8 PNG-IJ pell -png 5952 26 – – – 

E. pilularis Beerburrum pil -beer 248 9 26.93 152.88 159 

E. pilularis Deongwar pil -deong 5504 15 27.31 152.25 550 

Pinus elliottii var. elliottii × 

P. caribaea var. hondurensis FPQ best-F2 hybrid pine hybrid 

CSO families 

bulk unknown – – – 

E. reducta Ravenshoe reduct -rave 272 >10 17.55 145.45 980 

E. resinifera Beerburrum resin -beer 13981a 10 26.95 152.83 100 

E. resinifera Dorrigo resin -dorr 13976a 11 – – – 

E. resinifera Ravenshoe resin -ravens 314 unknown 17.55 145.45 980 

E. saligna subsp. saligna Dendrotech – clone 56 saligna -clo – 1 – – – 

E. siderophloia Mt Mee siderop -mtmee 17388a 4 27.02 152.70 350 

E. sideroxylon Uralla siderox -ural 20394a 9 30.63 151.50 895 

E. sphaerocarpa Blackdown sphaero -bla 391_392 – 23.83 149.08 760 

E. tereticornis Rundle teret -rund 10456 7 23.63 151.00 30 

E. tereticornis Zimbabwe SSO teret -zimb so 10872 unknown – – – 

E. urophylla hybrids u × p, u × d and u × g × c u hybrid -mix – unknown – – – 

E. urophylla Dongmen, China uro -dongmen 5979 10 – – – 

E. urophylla Sumatra provenances 9 uro -mix 1121 20 – – – 

E. urophylla Mt Egon Flores uro -mt egon 14531a 10 08.63 122.45 515 

E. urophylla Sumatra uro -sum 11152 unknown – – – 

E. urophylla Uhak Wetar uro -uhak 17836i 10 7.57 126.50 

E. urophylla × E. camaldulensis CSIR u × c 10508 – – – – 

E. urophylla × E. pellita SAPPI u × p -sappi 10551-1–7 5 – – – 

E. urophylla × E. tereticornis CSIR u × t -csir 10505 6 – – – 

Altitude 

(m asl) 

1 QPIF = Queensland Primary Industries and Fisheries, 2 CP = control pollination, 3 CSO = clonal seed orchard, 4 SAPPI = Sappi Forest 

Products, 5 SAFCOL = South African Forestry Company Ltd, 6 CSIR = Council for Scientific and Industrial Research, South Africa, 7 FPQ = 

Forestry Plantations Queensland, 8 SSO = seedling seed orchard, 9 Provenances are: Tele, Aek Nauli, Samosir and Harbinsaran.



DISCUSSION 

The large scale establishment of commercial hardwood plantations in Queensland’s subtropics only 

commenced in 1997. The selection of taxa to grow in these plantations was initially based on overseas 

experience and on utilisation of the species traditionally harvested from the native stands (Lee et al., 

1999). For example, the establishment of E. grandis and E. grandis × E. urophylla and E. grandis × E. 

camaldulensis hybrids was based on data from South Africa and Brazil (e.g. Eldridge 1993; Verryn 

2000) and the planting of E. dunnii in China (e.g. Wang et al., 1999; Arnold and Luo, 2002). As a 

result some of these hardwood plantations have failed. The trial series discussed here were established 

to provide a better basis for taxa selection. 

The taxa planted in these trials showed significant differences in VI, survival and borer incidence at 

age six years. The average rainfalls during this growth period for the three Wide Bay trials were very 

similar ranging from 868 to 877 mm (which is approximately 15 per cent below the long-term 

average). Nevertheless, soil moisture conditions and frost risk would have varied with trial position in 

the landscapes and soil types thus providing variable conditions over which to assess taxa potential. 

Moreover, the droughty conditions provided test environments simulating those being predicted for 

Queensland’s subtropics in the future as a consequence of climate change. These conditions would 

have favoured taxa better able to tolerate drought conditions. In these three trials (506a, 506b, 506c) 

the taxa showing consistently high VI at age six were ccv -woon (ranked 1 st , 6 th and 5 th ) and longi - 

coom (ranked 3 rd , 2 nd and 3 rd ; Table 3). In 506a and 506b, grandis -cop ranked 4 th and 1 st and grandissap 

ranked 6 th and 4 th respectively. In these trials a bulk of the nine commercial g × c -dend clones 

ranked 8 th , 11 th and 7 th . The other commercial taxa planted in the region, dunnii -sap, ranked 5 th and 

10 th respectively in trials 506a and 506b. Other taxa that performed well on at least one site in this 

region are grandis -koo and grandis-dpi ranked 1 st and 2 nd in 506c, g × r -csir ranked 3 rd in 506b and 

ccv -brooy which ranked 2 nd and 8 th in 506a and 506c respectively. When borer attack is taken into 

account, the stand-out taxa across the three trials is ccv -woon. This result confirms other results (Lee 

et al., 2001; 2005) on this taxa which is generally only surpassed by Corymbia hybrids (Lee 2007, Lee 

et al., 2009) in terms of growth across a range of sites in Queensland’s subtropics. 

In the South Burnett the average annual rainfall during the period reported in this paper was between 

590 and 600 mm, 23 per cent below the long-term average. In these two trials (523c and 523d) the 

taxa with a consistent high volume index at age six years were: grandis -dpi (ranked 3 rd and 4 th ), 

grandis -wed (ranked 4 th and 7 th ), g × c -dend (ranked 8 th and 2 nd ), maid -bolaro (ranked 7 th and 1 st ), 

and ccv -woon (ranked 11 th and 5 th ). In 523c taxa that had high VI in only in this trial were saligna - 

clo (ranked 1 st ) g × r -csir (ranked 2 nd ), maid -bondi (ranked 5 th ) and grandis -koo (ranked 6 th ). 

Similarly, in 523d the taxa with high VI were: ch -lock (ranked 3 rd ), ch -ner (ranked 6 th ) and ccv - 

brooy (ranked 8 th ). 

Since our measurement in 2006 and 2007, E. grandis and its hybrids, E. dunnii, E. globulus subsp. 

maidenii and E. pellita have suffered considerable mortality (up to 25 per cent), possibly due to water 

stress and borers. It is expected that at age 10 the performance of these taxa will show marked 

reductions in volume yield on a per hectare basis. This projection is similar to findings for E. globulus 

in West Australia where early-age performance did not foreshadow later-age performance (Walsh et 

al., 2008). White et al. (2003) warns the susceptibility to drought is an important consideration 

particularly in low rainfall regions where rainfall is highly variable. For many parts of Queensland the 

long-term average rainfall might appear to be relatively high (e.g. 1000–1200 mm in the Wide Bay 

region) but it is highly variable and species such as E. grandis, E. globulus subsp. maidenii, E. saligna 

and E. dunnii which have water use strategies similar to E. globulus may be at risk. 

Indications of taxa most suitable for various products across all trials 

A sustainable hardwood plantation industry in south-east Queensland requires optimal choice of 

species in relation to sites (landscape positions, soils) and changing climate conditions. Regarding the 

latter, a five per cent reduction in rainfall is predicted by 2030 in Queensland’s subtropics followed by 

at least a 10 per cent reduction by 2070 (Anon. Office of Climate Change, QLD 2008). Along with



this, there are predicted increases in temperature, more intense severe weather events and higher 

evapotranspiration rates. The trials have been subjected to a period of lower than average rainfall. 

Therefore taxa performance reported here should be a good indicator of their potential when grown 

under the more extreme climatic conditions as a result of climate change. 

While it is important to match species to sites it is probably even more important to understand which 

species with suitable wood properties can be grown across a wide range of sites and climatic zones. 

The stability of growth of ccv -woon and longi -coom across all trials confirms the early promising 

potential of these species, identified in Queensland by Lee et al. (2001) and Gardner et al., (2001; 

2007) in South Africa. Although ccv -woon has low annual VI increments at age six (



Impact of site on taxa performance 

The growth variation within the three sites in the higher rainfall zone is thought to be related to soil 

types and topographic position. Trials 506a and 506b were established within the same plantation; 

with 506b sited on a lower slope position. While the soil types are fairly similar, the lower slope 

position was expected to have a higher moisture status, which may explain the small growth 

advantage. Trial 506c is on a different property and is located on a broad alluvial area with deeper 

soils. This site has a 70 per cent VI advantage (using the mean of the six common taxa) over 506b. 

The better growth on this site can probably be explained by deeper and possibly more fertile soils and 

access to more water than that available in 506a and 506b. 

The two trials in the lower rainfall zone are on similar soils; however, there is overall a 113 per cent 

advantage in VI for trees at site 523c site over those in experiment 523d (using the mean of the six 

common taxa). While soil and site differences may account for some of this variation, anecdotal 

evidence indicates that there are shallow water tables throughout the area and it is possible that there is 

underground water available to trees in 523c. 

CONCLUSION 

The taxa with the best VI at age six years in these trials were: ccv –woon (28.5 m³/ha in 506a), for 

grandis -cop (45.6 m³/ha in 506b), grandis -koo (97.6 m³/ha in 506c), for saligna -clo (84.3 m³/ha in 

523c) and for maid -bolaro (33.0 m³/ha in 523d). This equates to annual VI increments 4.7, 7.6, 16.3, 

14.1, and 5.5 m³ /ha /annum respectively. If the better annual VI increments could be achieved on a 

broad-scale, then the success of the hardwood plantations in this region would be assured. 

Unfortunately, selecting sites where higher growth rates are achieved is a hit and miss affair the southeast 

Queensland. The Symphyomyrtus species (E. dunnii, E. grandis and its hybrids, E. longirostrata, 

E. globulus subsp. maidenii and E. saligna) are all highly susceptible to borer attack so they are 

unlikely to supply high-value solid wood products. Further these species have water-use strategies 

similar to that of E. globulus subsp. globulus so the variable rainfall pattern in south-east Queensland 

could put these species at risk. If the product is solid wood then the taxa that meet this criteria in 

Queensland’s subtropics are select provenances Corymbia citriodora subsp. variegata (e.g. ccv – 

woon), the Corymbia hybrids and possible E. cloeziana. If carbon sequestration is the objective then 

select provenances of species with good growth and high densities such as E. longirostrata and C. 

citriodora subsp. variegata should be considered. 


We thank many staff at Department of Employment, Economic Development and Innovation and its 

predecessor departments: Cliff Raddatz, Murray Johnson, and Alan Ward helped in the establishment 

of the trials and Tony Burridge validated the measure and assessment data and prepared Figure 1. 

Thanks are given to Integrated Tree Cropping and Forest Enterprises Australia for provision of land 

and with help in the establishment and measurement of the trials. We thank Garth Nikles and Mike 

Shaw for a helpful prepublication review of the paper. 

REFERENCES 

ABARE. (2008). Australian forest and wood products statistics. March and June quarters 2008 

(Canberra: Australian Bureau of Agricultural and Resource Economics), pp. 82. 

Anon. (2008). Climate change in Queensland what the science is telling us (Brisbane: The State of 

Queensland. Environmental Protection Agency Office of Climate Change), pp. 36. 

Arnold, R., and Luo, J. (2002). Cold Tolerant Plantation Eucalypts for South Central China. In 

International Symposium on Eucalyptus Plantations - Guangdon, P.R. China (China), pp. 1- 

13. 

Atyeo, W.J., Hopewell, G.P., and McGavin, R.L. (2008). Evaluation of wood characteristics of 

tropical post-mid rotation plantation Eucalyptus cloeziana and E. pellita: Part (d) Veneer and 

plywood potential. (Melbourne: Forest and Wood Products Australia, Limited), pp. 47.



Bailleres, H., Hopewell, G.P., and McGavin, R.L. (2008). Evaluation of wood characteristics of 

tropical post-mid rotation plantation Eucalyptus cloeziana and E. pellita: Part (c) Wood 

quality and structural properties (Melbourne: Forest and Wood Products Australia, Limited), 

pp. 65. 

Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall, N., Hyland, B.P.M., Johnson, R.D., Kleinig, 

D.A., McDonald, M.W., and Turner, J.D. (2006). Forest Tree of Australia. (Melbourne: 

CSIRO Publishing). 

Eldridge, K., Davidson, J., Harwood, C., and van Wyk, G. (1993). Eucalypt domestication and 

breeding. (Oxford: Clarendon Press). 

Gardner, R.A.W. (2001). Alternative eucalypt species for Zululand : Seven year results of site : 

species interaction trials in the region. Southern African Forestry Journal 190, 79-88. 

Gardner, R.A.W., Little, K.M., and Arbuthnot, A. (2007). Wood and fibre productivity of promising 

new eucalypt species for coastal Zululand, South Africa. Australian Forestry 70, 37-47. 

Isbell, R.F. (1996). The Australian soil classification. (Melbourne, Victoria: CSIRO Publishing). 

Jeffrey, S.J., Carter, J.O., Moodie, K.M., and Beswick, A.R. (2001). Using spatial interpolation to 

construct a comprehensive archive of Australian climate data. Environmental Modelling and 

Software 16, 309-330. 

Lawson, S. (2003a). Susceptibility of eucalypt species to attack by the giant wood moth (Endoxyla 

cinerea) in Queensland. Hardwoods Queensland report No 7. Queensland Forestry Research 

Institute, Agency for Food and Fibre Sciences, Queensland Department of Primary Industries, 

pp. 7. 

Lawson, S. (2003b). Susceptibility of eucalypt species to attack by longicorn beetles (Phoracantha 

spp.) in Queensland. Hardwoods Queensland report No 10. Queensland Forestry Research 

Institute, Agency for Food and Fibre Sciences, Queensland Department of Primary Industries, 

pp. 11. 

Lee, D.J. (2007). Achievements in forest tree improvement in Australia and New Zealand 2: 

Development of Corymbia species and hybrids for plantations in eastern Australia. Australian 

Forestry 70, 11-16. 

Lee, D.J., Nikles, D.G., and Walker, S.M. (1999). Strategy and progress on the genetic improvement 

of hardwood timber species in Queensland. In Proceedings of the 11th Australian Plant 

Breeding Conference (Adelaide, Australia: CRC for Molecular Plant Breeding), pp. 38-40. 

Lee, D.J., Nikles, D.G., and Dickinson, G.R. (2001). Prospects of eucalypt species, including 

interspecific hybrids from South Africa, for hardwood plantations in marginal subtropical 

environments in Queensland, Australia. Southern African Forestry Journal 190, 89-94. 

Lee, D.J., Huth, J.R., Brawner, J., and Dickinson, G.R. (2009). Comparative Performance of Corymbia 

Hybrids and Parental Species in Subtropical Queensland and Implications for Breeding and 

Deployment. Silvae Genetica, (in press). 

Lee, D.J., Debuse, V.J., Pomroy, P.C., Robson, K.J., and Nikles, D.G. (2005). Developing genetically 

adapted tree varieties for marginal areas of northern Australia (Barton, Canberra: Rural 

Industries Research and Development Corporation), pp. 58. 

Listyanto, T. Glencross, K., Nichols, J.D., Schoer, L. and Harwood, C. (submitted) Performance of 

eight eucalypt species and interspecific hybrid combinations at three sites in northern New 

South Wales, Australia. Australian Forestry (submitted). 

Verryn, S.D. (2000). Eucalyptus hybrid breeding in South Africa. In Hybrid breeding and genetics of 

forest trees. Proceedings of QFRI/CRCSPF Symposium, D.G. Nikles, ed (Noosa, 

Queensland, Australia: Department of Primary Industries, Brisbane), pp. 191-199. 

Walsh, P.G., Barton, C.V.M., and Haywood, A. (2008). Growth and carbon sequestration rates at age 

ten years of some eucalypt species in the low- to medium-rainfall areas of New South Wales, 

Australia. Australian Forestry 71, 70-77. 

Wang, G., Arnold, R.J., Gardiner, C.A., Zhang, J., and Wu, Z. (1999). Seed source variation for 

growth in Eucalyptus dunnii; results from trials in south central China. Australian Forestry 

62, 120-127. 

White, D.A., Battaglia, M., Macfarlane, C., Mummery, D., McGrath, J., F., and Beadle, C.L. (2003). 

Selecting species for recharge management in Mediterranean southwestern Australia: some 

ecophysiological considerations. Plant and Soil 257, 283-293.



Table 3 Performance means and differences among 60 taxa in five trials for six year volume index, survival and borer incidence 1 . 

Definitions of taxa are provided in Table 2. 

Volume index at age six years (m³/ha) Survival at age six years (%) Borer incidence at age six years (%) 

Taxa 506a 506b 506c 523c 523d 506a 506b 506c 523c 523d 506a 506b 506c 

acmen -neer 9.1 abc 7.1 ab 82.3 hijkl 64.6 bcd 87.5 efgh 3.8 a 3.8 ab 0.8 a 

acmen -woolg 13.9 ab 85.8 efgh 0.0 a 

argo -ballon 36.7 cde 16.7 cdef 85.0 ghijkl 82.8 efg 

Corymbia hybrid 38.7 cdef 41.1 abc 

ccc –glad 2 15.4 cdef 15.3 bcde 28.5 abcdef 45.0 efghi 23.2 efghijkl 84.4 jkl 62.5 bcd 91.7 fgh 85.0 ghijkl 80.0 defg 1.0 a 6.0 ab 0.0 a 

ccc -yepp 10.5 abc 39.9 cdefg 46.5 efghij 28.3 hijklm 71.9 ghi 83.3 efgh 85.0 ghijkl 70.0 bcde 1.4 a 0.0 a 

ccv -brooy 27.6 gh 26.0 ghij 47.0 efgh 53.1 fghijk 28.8 hijklm 88.5 kl 86.5 gh 94.2 gh 71.7 defgh 76.7 cdef 0.0 a 0.0 a 0.0 a 

ccv -woon 28.5 h 31.9 ijkl 57.6 ghi 57.6 hijklmn 29.7 jklm 86.5 jkl 86.5 gh 88.3 efgh 95.0 jkl 95.0 fg 0.0 a 1.0 a 0.0 a 

ch -lock 35.6 bcdefg 45.0 efghi 31.7 lm 74.2 cdefg 76.7 efghij 83.3 efg 0.0 a 

ch -ner 39.3 cdefg 55.6 ghijkl 29.3 ijklm 77.5 defgh 86.7 ghijkl 83.3 efg 0.0 a 

cloez -home 22.4 fgh 15.5 bcdef 47.1 efghij 26.2 ghijklm 86.5 jkl 72.9 cdefg 85.0 ghijkl 80.0 defg 1.1 a 1.4 a 

cloez -mungy 22.5 fgh 50.1 efghijk 19.9 defgh 86.5 jkl 82.8 ghijkl 58.3 abc 0.0 a 

cloez -saso 23.3 fgh 87.5 kl 0.0 a 

dunnii -sapp 24.9 gh 25.3 fghij 86.5 jkl 79.2 efgh 39.6 cd 42.7 bc 

grandis -wed 71.3 mnop 28.9 hijklm 86.7 ghijkl 70.0 bcde 

grandis -cop 25.6 gh 45.6 m 70.8 gh 80.2 efgh 97.5 fg 93.0 d 

grandis -dpi 94.5 j 72.1 nop 30.4 klm 92.5 fgh 96.7 kl 85.0 efg 64.4 ef 

grandis -koo 97.6 j 62.0 klmno 26.0 ghijklm 86.7 efgh 83.3 ghijkl 70.0 bcde 90.6 g 

grandis -saf 57.4 hijklm 23.2 efghijkl 93.3 ijkl 70.0 bcde 

grandis -sap 24.8 gh 33.8 jkl 69.8 fg 62.5 bcd 96.9 fg 95.1 d 

g × c -csir 11.5 abcd 58.3 ef 81.9 ef 

g × c -dend 23.6 fgh 22.8 efghi 50.8 fghi 59.5 ijklmn 32.0 lm 91.7 l 82.3 efgh 89.2 efgh 95.0 jkl 98.3 g 50.0 d 44.1 bc 29.7 bc 

g × p -mix 62.9 hi 82.5 efgh 53.9 de 

g × r -csir 21.4 fgh 39.7 klm 74.4 op 21.9 defghijk 77.1 ghijk 90.6 h 88.3 hijkl 81.1 defg 85.8 efg 63.9 cd 

g × t -csir 20.6 efgh 21.6 efgh 41.2 defg 83.3 ijkl 75.0 defg 58.3 cde 75.4 e 89.9 d 

g × t -mix 21.1 fgh 28.6 hij 45.8 cd 64.6 bcd 96.5 fg 95.2 d 

g × u -clone 57.6 ghi 54.9 ghijkl 86.7 efgh 86.7 ghijkl 86.1 fg 

g × u -csir 24.4 gh 33.1 jkl 35.8 bcdefg 57.5 hijklm 17.5 cdefg 57.3 de 70.8 cdef 56.7 bcd 60.0 cdef 56.7 abc 95.1 fg 100.0 d 90.7 g 

hoop 2.6 NA 0.2 a 83.3 ghijkl 42.8 a 

hybrid -mix 28.9 abcdef 55.8 bcd 63.1 de 

longi -coom 27.4 gh 41.6 lm 71.0 i 59.0 ijklmn 24.6 fghijklm 88.5 kl 85.4 gh 83.6 efgh 83.3 ghijkl 66.7 bcde 41.0 cd 35.1% abc 16.1 ab 

mac -sappi 50.1 efghijk 71.7 defgh



Table 3 cont. 

Volume index at age six years (m³/ha) Survival at age six years (%) Borer incidence at age six years (%) 

Taxa 506a 506b 506c 523c 523d 506a 506b 506c 523c 523d 506a 506b 506c 

maid -bolaro 

26.1 abcde 61.1 jklmno 33.0 m 53.3 bc 83.3 ghijkl 85.0 

efg 15.3 

ab 

maid -bondi 11.8 a 68.7 lmno 25.9 fghijklm 28.3 a 73.3 defgh 61.7 abcd 42.0 cd 

mol -ravens 36.2 bcde 10.6 bc 90.0 hijkl 80.0 defg 

pch 38.7 cdefg 98.3 h 0.0% a 

pell -kur 21.6 fgh 28.8 hij 83.3 ijkl 84.4 fgh 27.4 bc 36.8 abc 

pell -png 19.4 defg 17.6 cdefg 82.3 hijkl 68.8 cde 82.6 efg 60.4 cd 

pil -beer 22.2 abcd 46.0 efghi 17.2 cdefg 77.5 defgh 86.7 ghijkl 55.0% ab 0.0 a 

pil -deong 12.6 a 39.9 def 12.8 bcd 50.0 ab 53.9 bcd 45.0 a 0.0 a 

pine hybrid 25.8 abcde 94.2 gh 0.0 a 

reduct -rave 8.3 abc 17.4 abc 52.1 b 70.8 bcdef 3.5 ab 1.1 a 

resin -beer 38.1 cdefg 20.9 defghij 80.8 efgh 75.0 bcdef 4.2 a 

resin -dorr 42.5 defgh 43.8 efgh 20.2 defghi 79.2 efgh 75.0 efghi 81.7 defg 1.1 a 

resin -ravens 20.4 efgh 30.9 hijk 87.5 kl 89.6 h 21.9 b 34.7% abc 

saligna -clo 84.3 p 93.3 ijkl 

siderop -mtmee 38.0 cde 22.4 efghijk 98.3 l 93.3 fg 

siderox -ural 24.3 abc 15.0 cde 68.3 defg 86.7 efg 

sphaero -bla 11.9 bcde 0.1 a 9.0 a 21.9 ab 4.8 ab 82.3 hijkl 6.3 NA 68.3 bcde 78.3 fghijk 43.3 a 0.0 a 5.5 ab 0.0 a 

teret -rund 18.7 defg 86.5 gh 46.5 c 

teret -zimb so 13.1 bcde 84.4 fgh 40.3 abc 

u hybrid -mix 20.3 a 32.2% a 

uro -dongmen 6.7 abc 30.2 ab 88.5 efg 

uro -mix 5.0 ab 8.1 abc 30.2 ab 28.1 a 99.8 g 99.8 d 

uro -mt egon 2.8 a 18.8 a 99.7 g 

uro -sum 6.3 ab 9.0 abcd 34.4 bc 28.1 a 99.9 g 95.6 d 

uro -uhak 6.7 abc 50.0 de 89.7 efg 

u × c 58.3 hijklmn 71.1% defgh 

u × p -sappi 19.4 defg 75.0 ghij 88.5 efg 

u × t -csir 16.5 bcdefg 60.4 bc 89.6 d 

Mean six 

common taxa 2 24.5 28.5 48.4 55.3 26.0 82.8 79.0 83.9 81.7 78.9 31.2 31.0 22.8 

1 

Taxa differences were significant at the P



ABSTRACT 

RESPONSES TO A DRYING CLIMATE 

IN THE NORTHERN JARRAH FOREST 

Frank Batini 1 

I am deeply concerned at the health of the jarrah forest ecosystem. Some of the symptoms 

measured over the past three decades include: a change to rainfall patterns, lower average 

rainfall, a fall of several metres in regional water tables, a reduction in streamflows by seventy 

five percent, a summer drying of streams that were once considered as perennial, consequent 

changes to stream biota and the death of old, large trees on shallower soils. 

These changes are due, in part, to climate change, but also to other hydrologic factors. In the 

past ten years it now takes fifty percent more rain for some streams to commence to flow, and, 

for an equivalent rainfall, the stream now yields only one-third of the previously recorded flow. 

Why is this so? 

Forest management has also changed. Large areas are now protected as reserves and are not 

available for silvicultural operations. There has been a reduction in the frequency of prescribed 

burning. Bauxite mining has altered the hydrology of the rehabilitated mine-pits. All these 

changes compound the climatic effects. 

However the situation is retrievable, if we have the will as a community to implement alternate 

policies and management practices. Research trials have shown that thinning, the utilisation of 

biofuels and regular prescribed burning can increase water tables and streamflow. I estimate that 

sound forest management on 100,000 hectares of jarrah forest in the high rainfall zone would 

benefit forest health, increase water tables, improve stream flow and aquatic biodiversity as well 

as producing 80-100MW of power annually from biofuels and increasing yield into the surface 

dams by between 30 and 50 GL each year, at a fraction of the cost of desalination. 

INTRODUCTION 

As a forester and scientist with a lifetime of experience in the jarrah forests north of Collie, I am 

deeply concerned for the future for this ecosystem. Let me share my concerns with you. 

• In 1975 the streamflow into dams that are part of the Integrated Water Supply System 

averaged 350 billion litres. Since that year it has fallen to one-quarter and in 2006 to 

one-sixth of that amount. 

• In Wungong Brook the Vardi road gauging station has been operating for 21 years. 

Five of the six lowest flows recorded there have occurred in the last eight years. 

• Until recently, this Brook was considered to be perennial. In 2006, a very dry year, the 

stream ceased to flow for about 12 weeks and even the larger pools were dry. The 

stream also ceased to flow in the following summer, even though the 2007 rainfall was 

above average. 

• Data published by Alcoa scientists (Croton and Reed 2007) show that while water 

yields from experimental catchments in the northern jarrah forest subjected to bauxite 

mining initially increased (during mining), they are then significantly reduced 

following rehabilitation. The decline between 2001 and 2005 in four experimental 

catchments ranged between 31 and 67 mmpa, or between 29 and 58 percent of the 

average streamflow. In another near-by catchment where most of the rehabilitation is 

younger, the reduction in yield is currently much smaller, 4mmpa and 5 percent of 

flow. 

1 Consultant in the Management of Natural Resources. Email: fbatini@bigpond.net.au



• These changes in flow volumes and flow patterns of forested streams in bauxite 

mining areas are replicated across the jarrah forest, and have already had measurable 

effects on stream fauna. Data collected by University of Western Australia Dr Storey 

show significant shifts in faunal assemblages, differences in species richness and 

abundance (Storey 2007) 

• Regional water-tables have fallen by as much as 12 metres in the past 30 years (Batini 

2004). If the specific coefficient of the soil is taken as 8 percent, this equates to a drop 

of one metre of water, or 10 000 kilolitres per hectare, or a staggering 4500 billion 

litres from the forested catchments that feed into the Integrated Water Supply Scheme. 

• The key hydrologic process operating is evapotranspiration by trees and understorey. 

In the jarrah forest this can account for 90–95 percent of the annual rainfall ( Doley 

1967, Greenwood et al 1985). 

• The forest is now using more water than is available through rainfall. It is essentially 

‘mining’ the reserves that are stored in the soil profile and water table. 

• Data provided to me by the Department of Environment and Conservation (DEC) on 

‘Soil dryness index’ show that the upper layers of soil, the litter, heavy fuels and living 

vegetation are becoming progressively drier. 

• Jarrah trees have died from drought stress on shallow soils close to exposed rock 

surfaces. This is not unusual but what is most concerning now is that some of these 

trees are very large, probably 120–150 years old. There is a similar situation in the 

wandoo forest, where large old trees are dying back from the crowns ― an obvious 

symptom of drought. Having survived previous droughts, why are they dying now? 

• Hydrologists and water supply engineers agree that the reductions in water yields 

appear to follow a step-like process, which is then difficult to reverse. For example, at 

the Cobiac gauging station, a rainfall of 967 mm in 2006 now yields only one-third the 

flow of a similar rainfall 10 years ago. This stream now needs about 300mm of rain to 

begin to flow, whereas 10 years ago it only needed 200mm. In that time the vegetation 

has not changed, but water tables have fallen by about 3–5 m on upland sites and 1–2 

m on lower slopes (Wungong Whispers 2009). 

• Most water movement in the jarrah forest is by ‘interflow’ that occurs within a metre 

or so of the soil surface. These flows raise the shallow watertable, saturating the 

swamps and near-stream zones that then act as ‘roaded’ catchments. Any rain that hits 

these wetted areas cannot penetrate into the soil and will then flow into the stream. A 

fall in the shallow watertable means that the ‘sponge’ that needs to be filled before any 

streamflow can occur is much larger. It is this ‘disjunction’ that may explain these 

anomalous results. 

You may consider that all of these effects are solely due to ‘climate change’ and thus outside of our 

direct control. This is not so. While lower rainfall is a key driver of change, other factors must also be 

considered. Let us examine how the forest has changed since 1829. 

The jarrah forest in 2008 is very different to that of 1829 when settlement at Perth began. Early 

photographs show that the original forest was dominated by old, large, mature to overmature trees. It 

had a sparser understorey and was probably burnt on a very regular cycle by the Noongar people, and 

by lightning-caused fires (Burrows and Abbot 2003). There was no logging, dieback, bauxite mining 

or roads. Access by rail and road began in the 1880s and by the 1920s all of the Wungong catchment 

and much of the western jarrah forest had been logged. Most then burnt under wildfire conditions, but 

the forest is very resilient and regenerated successfully, from coppice, lignotubers and seed. 

The logging and burning changed the structure and composition of this forest. A smaller number of 

large, old trees were replaced by large numbers of smaller, younger trees. Many areas were again 

logged, some as many as four times since 1880. The current forest is now dominated by regrowth, the 

oldest up to 120 years of age. The density of this regrowth (as measured by basal area) is not much 

different from that of the virgin forest but there are now at least twice as many, smaller trees on each



hectare (Inventory data supplied by DEC). For a given density, smaller trees have a much greater 

‘sapwood area’ and it appears that they are able to maintain higher transpiration rates than larger trees 

(C McFarlane CSIRO). While most areas regenerated well, the proportion of sheoak (Allocasuarina 

fraseriana) also increased on some sites. There are now thickets of sheoak on sites where only large 

jarrah stumps remain. 

Dieback disease was introduced into the western jarrah forest in about the 1920s and was well 

established in the Wungong catchment by the 1940s, when the first aerial photos became available ( 

Batini 1973). The disease spread rapidly during the wetter period between 1950 and 1970 and is now 

found over hundreds of thousands of hectares. Bauxite mining began in the forest in 1964 and now, 

some 40 years later, over 13000 hectares have been mined and rehabilitated. The rehabilitated minepits 

have been deep ripped to encourage infiltration, are planted densely with tree and understorey 

species and are engineered to retain water. 

Fire regimes have also changed, from the very frequent burning by the Noongar, through a period of 

attempted fire exclusion in the 1920–40s, to prescribed burning on 5–7 year cycles during the 1960– 

1990s to a cycle now averaging 10–12 years over the forest area (DEC pers. comm.). 

From this discussion about these landscape-wide changes, the following five factors now play a more 

significant role in affecting streamflow: 

1. Tree density. If the forest has too many trees, it will transpire more and cause a drop in 

water-tables. The forest has always been fully stocked (i.e. stocked to the limit of its water supply) 

but, since the rainfall has decreased, it is now overstocked with younger regrowth resulting from 

past commercial operations. It now contains a high proportion of smaller trees that not only 

transpire more than larger trees, but are only suitable for use as re-constituted wood, as bio-fuels 

or bio-energy, and are currently not able to be widely utilised. Data show that the canopy cover of 

the original stand can be reached or exceeded by regrowth at a very early age, between 10 and 20 

years of age (Stoneman 1988, Grigg and Grant 2009). 

2. Understorey density. The understorey can account for up to 30–40 percent of the total 

evapotranspiration (Greenwood et al 1985, Marshall et al 1994). If the understorey is denser, it 

will transpire more and cause a drop in water-tables. As a result of changes to prescribed burning 

rotations, lengthening these from about 5–7 years (Forests Department) to about 10–12 years 

(DEC, pers. comm.), I believe that the understorey is now denser than it has been in the past. In 

addition, some stream areas are left unburnt for “habitat”. 

3. Bauxite mining. If bauxite mining has altered the sub-surface, interflow patterns and the 

rehabilitation is too dense, transpiration will increase and runoff will reduce. By removing the 

bauxitic layer, deep ripping, ensuring that any surface runoff is captured within pits, and by 

seeding densely with native shrub and tree species, mining has severely reduced streamflow in 

some catchments (Croton and Reed 2007). This is not a criticism of Alcoa World Alumina. These 

practices were mutually-agreed to between the mining company and State Agencies as providing 

appropriate environmental outcomes, especially the protection of water quality, but are probably 

more suited to rainfall conditions of 30–40 years ago. 

4. Larger reserves. Many areas of the forest that are now overstocked have been allocated either a 

‘formal’ or ‘informal’ reserve status. Some stream reserves are now 400 metres wide, yet these 

are the key “source areas” for streamflow. In addition, large Fauna Habitat Zones are now 

required at regular spacing within State forest (Conservation Commission 2004) An example is 

the Wungong catchment, where one-third of the total area is not available for any form of 

silviculture and prescribed burning is only done at intervals of as much as 10–12 years — 

possibly longer.



5. Dieback disease. Dieback disease is the exception. By killing many species in both the 

overstorey and understorey, dieback disease has in fact increased the water yield from affected 

areas. (Batini et al 1980) Also, many of the dieback areas are close to streamlines and contribute 

to enlarging the ‘source area’ for flow. However, this contribution has decreased in recent years as 

many areas have now been actively replanted or seeded and others are regenerating naturally. 

These factors have major negative implications on water yield and management will need to be 

changed on the 100,000 hectares of higher rainfall, forested catchment that I believe should be 

designated with a priority of ‘water supply’ and managed for this purpose. I believe that the concepts 

of Ecologically Sustainable Forest Management and of Multiple Use allow for the identification of 

some areas, such as Nature Reserves where conservation of biodiversity is a priority, and other areas 

where water production should be a priority, leading to different management regimes. 

Is it all doom and gloom? No, I don’t think so. The situation is retrievable, at least in part, if we are 

willing to work together as true partners in the management of these forests rather than as ideological 

competitors for ‘this piece of turf’ or as protectors of the ‘status quo’. Land management policies that 

were developed at a time when the hydrologic regime was quite different need to be re-examined and 

many will need to be changed. 

We also need to give considerable thought to re-defining our land use priorities in the jarrah forest. 

Just as Nature Reserves rightly have a priority for conservation, and National Parks a priority for 

conservation and visitor use, the higher rainfall areas of State forest within catchment areas (whether 

surface or groundwater) should have a priority for the sustained production of good quality water, 

while recognising other values such as biodiversity ( Forests Department 1978, 1980). The State’s 

current investment in our dams, pipelines and infrastructure in these areas now exceeds a billion 

dollars. In addition, each hectare of forest on surface water catchments in the higher rainfall zone can 

yield an average of 100 mm per annum of high quality drinking water valued at between $1000 and 

$1500. I believe that this yield can be increased by 30- 50 percent by good forest management. 

The priorities for drinking water catchments should be set by the Department of Water (DoW), as I 

believe that they are not sufficiently well articulated in the current Forest Management Plan 2004- 

2013 nor given adequate recognition by either the Conservation Commission (responsible for forest 

planning) or DEC (responsible for the management of these forests). This is understandable, given 

that that Agency’s prime functions are environmental protection and the conservation of biodiversity. 

I see that there is a clear need for a very strong leadership role here from the Department of Water. 

The key ecosystem drivers in the jarrah forest are energy (from the sun), nutrients and water. We have 

plenty of the first but the other two are both limiting factors. In my view there is now too much 

emphasis on research into individual species (eg the quokka, threatened flora, Carnaby cockatoo) and 

too little thought given to the impact that landscape-wide processes have on nutrient and water 

availability and on forest health. Let me now discuss some of these. 

Landscape wildfire vs regular prescribed burning 

There is minimal direct surface runoff from forested catchments in WA, except after wildfire. 

Following the 27,700 hectare-2005 Hills wildfire, the soil was bare and hydrophobic. The pattern of 

flow observed at the gauging station and field checks showed severe surface runoff and massive 

amounts of erosion and nutrient loss during the 2005 winter (Batini and Barrett 2007). This pattern is 

not observed after a cooler prescribed burn: run-off is increased, but this is not accompanied by soil 

damage and erosion. 

We also know that although water yield increases temporarily after a wildfire, by as much as 200 

percent after the “Hills fire of 2005”, it then decreases for many years due to massive regeneration. In 

contrast, a statistical analysis for the Water Corporation of two large catchments showed that regular 

prescribed burning increased water yield by at least 20 percent for 12–18 months after each burn (C 

Terry pers com). The more frequent the burn, the greater the increase. However, the DEC has some



data from two small experimental catchments that do not show an increase in yield after prescribed 

burning ( Stoneman pers com). 

It is estimated that the 2005 wildfire killed between 1.5 and 2.3 million trees. Many were large, habitat 

trees. Regular prescribed burning protects the jarrah forest by allowing these wildfires to be contained 

sooner. 

Silviculture 

A forest ecosystem that is managed with appropriate silviculture will not only yield sustainable 

products for our use, but will also increase the water available for use by the remaining trees, the 

understorey plants, the soil organisms, the stream flora and stream fauna, possibly by as much as 80– 

100 mmpa. Eventually more water will also flow into dams and be available for use by human beings. 

There are considerable research data from the jarrah forest thinning trials during the 1980’s-1990’s to 

support these statements. (Bari and Ruprecht 2003, Marshall and Chester 1992, Ruprecht et al 1991, 

Ruprecht and Stoneman 1993, Schofield et al 1989 and Stoneman 1993). 

Data from the 1980’s show that in regrowth forests there is competition for water and nutrients at 

basal area densities underbark of about 15 m 2 /ha, say about 20m 2 /ha overbark (Stoneman et al 1996). 

With the reductions in rainfall, this value could now be between 10-12 m 2 /ha. Some recent modelling 

by J Croton (pers com) shows that, with current rainfall regimes, water tables and streamflows cannot 

be maintained if basal areas exceed 10m 2 /ha. However, in the high rainfall zone, the mean basal area 

of 35 m 2 /ha is well in excess of these values. Thinning will substantially increase the growth rate on 

the remaining trees (Stoneman et al 1996), and will not greatly alter the composition of the 

understorey (Mattiske pers com). To maintain the increases in water and timber yields, coppice and 

seedling regrowth will need to be controlled. In a time of declining rainfall, the silviculture and 

utilisation of regrowth forests on water catchments needs to be increased, not reduced, as is being 

advocated by some interest groups, including the Conservation Commission. 

Utilisation 

DEC records show that most of the jarrah forest on drinking water catchments has been cut over at 

least twice, some three or four times. We now have a forest structure which is dominated by smaller 

trees that are not large enough to produce sawlogs and can only be utilised on the scale available for 

producing re-constituted wood products, bio-energy or second generation bio-fuels. Preliminary 

calculations show that the resource of bio-energy in the 100,000 hectares of State forest within the 

western, high-rainfall zone of surface water catchments could produce 80-100MW of electricity 

annually for at least 20-25 years (J Clarke Forest Products Commission, pers com). It is likely that any 

move by Government in this direction will be strongly opposed by the environmental movement. 

However there are really only two other choices: to carry out non-commercial thinning, at a 

considerable cost ($400-1000/ha), and leave the dead trees in the forest; or do nothing and let the 

forest sort itself out in the longer term, by dying from drought. This option, preferred by some in the 

environmental movement, may take decades to have full effect and in the meantime the stream 

environments and the catchment water supply system will continue to decline. 

Forest health 

The Forest Management Plan prepared by the Conservation Commission is based on the concept of 

Ecologically Sustainable Forest Management. A forest ecosystem where water and nutrients are not 

limiting will be a healthier ecosystem. There are data from the 1980s, when rainfall was considerably 

higher, that show that the growth rates of jarrah are severely reduced at densities in excess of 20 m 2 in 

basal area, due to competition for these scarce resources. Over 85 percent of the forest now exceeds 

this density. 

Significant ill-health and extensive drought deaths are a real possibility in the future. Trees such as 

jarrah and marri are very long-lived organisms (150 to 300+ years) and have therefore developed 

multiple strategies to cope with threatening processes such as unusual weather, fire, climatic change



and attack by insects and fungi. These processes are dynamic; tree crowns may partially recover, and 

then be affected once again. 

However, because of their great age, these older trees may not necessarily be in balance with the 

current environment, especially at a time of rapid change. Visiting foresters often remark at the 

current sorry state of jarrah crowns. 

Drought strategies all use less water and therefore mean that a lesser amount of photosynthetic 

material is produced. Where a forest is subjected to chronic drought or nutrient stress it is also much 

less able to resist attack by insect pests or fungal disease. Active response to infection or insect attack 

requires mobilisation of defence mechanisms that need additional photosynthetic products. These new 

leaves are often much more palatable to insect grazers leading to an increase in decline and further 

stress on the tree. 

Where the negative impacts on the trees are sustained over long periods (some years or decades) by a 

chronic climatic change, drought or continuous insect or fungal attack, the availability of food 

resources becomes limiting, defence mechanisms are severely stretched and may then fail. External 

influences that would have been dealt with by the tree in a normal situation may now lead to severe 

debility and tree death. 

. 

While many elements of the jarrah forest will survive, there will be considerable ecological change 

with a shift to the hardier, more drought tolerant plants as well as the suites of animals and insects that 

prefer these plants. While jarrah also grows in lower rainfall areas it is more open, lower in height, 

smaller in girth and has a lower basal area. Significant shifts in plant populations can occur over a 30– 

40 year period. The biota that are likely to be most negatively affected in the early phases are found 

within the stream ecosystems. These negative effects can already be measured (Storey 2007). 

In the past 30 months, the Water Corporation of Western Australia has been funding fieldwork for a 

large- scale experiment in the Wungong water-supply catchment that incorporates commercial logging 

by the Forest Products Commission, as well as non-commercial thinning and widespread prescribed 

burning by the Department of Environment and Conservation (DEC) as contractors. Twenty 

independent research/monitoring programs are also funded. The trial is anticipated to last for 12 years 

and cost $20 million (Water Corporation 2005, www.watercorporation.com.au/wungong). 

OTHER STATES 

Despite sustained drought and water restrictions in other mainland capital cities, I am not aware of any 

other Water Authority in Australia that is actively managing its forested catchments over large areas 

with prescribed fire and silviculture in order to maintain streamflow and biodiversity. In fact, quite the 

opposite it appears. 

There is limited prescribed burning, often leading to recent, widespread wildfire in several States 

(Victoria, NSW, ACT), including major water-supply catchments. The consequences of these largescale 

wildfires on reductions in yield as the forest recovers will be considerable and are still to be felt. 

Most of the major water-supply catchments are still “closed” to logging and there is controversy even 

over the logging of small areas of State forest such as those in the Thompson dam catchment in 

Victoria ( M Poynter pers com). 

CONCLUSION 

I believe we do have a choice. I can accept that Nature Reserves and National Parks in the jarrah 

forest be managed conservatively, without silviculture and with longer fire intervals, but the 

consequences of this management option must be closely monitored. These areas will then provide 

our baseline, or reference areas, in terms of the impacts of this method of management on forest 

health, biodiversity, stream health and water yield. However, this “experiment ”is being conducted by



the Conservation Commission over one million hectares of former State forest, with minimal 

monitoring of the consequences. 

I do not accept that similar management practices, focussing on the priority of conservation of 

biodiversity, should also apply to all of the remaining State forest, particularly State forest in the 

higher rainfall area of water catchments. In some of these areas, for example the 13000 ha Wungong 

catchment, I propose the need to be able to trial (and monitor) alternative management systems that 

favour streams, water yield, timber use, biodiversity and forest health without the imposition of 

unreasonable environmental constraints by the Conservation Commission, the EPA, the DEC or the 

environmental movement. What are some of these “ unreasonable” constraints? 

• The Conservation Commission refuses to permit thinning operations below the approved level 

of 15 m 2 /ha, even though current data suggest that this level is too high to maintain water tables 

and streams. 

• The Conservation Commission and DEC refuse to reduce the interval between prescribed burns 

to between 4-6 years required for control of understorey, preferring an 8-10 year cycle, citing 

vague “biodiversity concerns” 

• The Conservation Commission refuses to reduce the width of stream buffers, some of which are 

now 400 m wide and therefore negate the effects of any thinning upslope. A suitable width of 

buffer could well be 30m either side of the streamside vegetation. 

• The Conservation Commission will not allow any non-commercial thinning to occur within the 

existing wide buffers. This operation could be carried out on foot, targeting only the smaller 

trees. 

• The DEC has refused a proposal for a trial logging to be conducted under “moist soil” 

conditions within areas previously mined and rehabilitated. The reason given is “soil damage”, 

but without a trial, this assumption cannot be tested. 

• The DEC has refused a proposal to keep some of the rehabilitated areas that could be clearfelled 

open for between 4 and 6 years to allow time for water tables to build up. The reason 

given is that the current guidelines require that rehabilitation must commence within 30 months. 

• The Government’s Policy is to allow the use of biofuels, but only from plantation residue and 

not from native forest. This is due to pressure from the conservation lobby. However this leads 

to excessive costs, waste of material and visual impact. 

If these constraints can be overcome, we will then have at least two, possibly more, very different 

management systems available for comparison. Given what we know now, which is likely to be the 

more robust in a time of climate change? 


The Water Corporation of Western Australia. 

This paper was first presented at a public seminar organised by the Department of Water, October 15 th 

2008. The Abstract was published by DoW in the seminar’s proceedings. The paper was expanded 

subsequently. I have benefited from discussions with colleagues, especially K Barrett, A Reed, J 

Bradshaw and R Underwood. 

BIBLIOGRAPHY 

Bari M A and J K Ruprecht ( 2003) Water yield response to land use change in south-west Western 

Australia Department of Environment Report No SLUI 31 

Batini F E (1973) Jarrah Dieback- a disease of the jarrah forest of Western Australia,. Bulletin Forests 

Department Western Australia No 84 pp45. 

Batini FE, Black R E, Byrne J and Clifford P J ( 1980) An examination of the effects of land use 

changes in catchment condition on water yield in the Wungong catchment Western Australia 

Aust For Res10: 29-3



Batini F E and K Barrett ( 2007)-Monitoring the effects of wildfire on water, vegetation and 

biodiversity. The Forester 50, 3, p21. 

Batini F E ( 2004) Health of wandoo ( Eucalyptus wandoo) in the Helena catchment in relation to 

depth and reductions in water tables. Report to the Wandoo Recovery Group. 

Burrows, N. and Abbott I. (2003) Fire in south-west Western Australia: synthesis of current 

knowledge, management implications and new research directions. Fire in ecosystems of 

south-west Western Australia; impacts and management. CALM, 2003, Perth, WA. 

Conservation Commission of Western Australia (2004) Forest Management Plan 2004-2013 

Croton J T and Amanda J Reed (2007) Hydrology and bauxite mining on the Darling Plateau. 

Restoration Ecology 15,4,40-47. 

Doley, D (1967) Water relations of Eucalyptus marginata Sm under natural conditions. J.Ecol.55:597- 

614. 

Forests Department of Western Australia (1978). A Perspective for Multiple Use Planning in the 

Northern Jarrah Forest. 

Forests Department of Western Australia (1980). Land Use Management Plan Northern Jarrah Forest 

Management Priority Areas. 

Greenwood, E.A. N., Klein, L., Beresford, J. D., Watson, G. D. and Wright, K. D. (1985) Evaporation 

from the understorey in the jarrah (Eucalyptus marginata Don ex Sm) forest, south-western 

Australia. Journal of Hydrology 80: 337-349. 

Grigg A H and C D Grant ( 2009)- Overstorey growth response to thinning, burning and fertiliser in 

10-13 –year-old rehabilitated jarrah ( Eucalyptus marginata ) forest after bauxite mining in 

south-western Australia. Aust For 72, no 2, pp 80-86. 

Marshall, J. K., Chester, G. W. and Colquhoun, I. J. (1994) Water use by rehabilitation vegetation. 

Alcoa of Australia, Report No 94/28, 12p. 

Marshall J K and G W Chester (1992) Effect of forest thinning on jarrah ( Eucalyptus marginata) 

water uptake. Land and Water Research News no 14.24-27 

Ruprecht, J.,Schofield, N., Crombie, D.,Vertessy, R. and Stoneman, G. (1991). Early hydrological 

response to intense forest thinning in south-western Australia. Journal of Hydrology 127: 261- 

277. 

Ruprecht, J. and Stoneman, G. (1993) Water yield issues in the jarrah forest of south-western 

Australia. Journal of Hydrology 150: 369-391. 

Schofield, N. J,. Stoneman, G. L. and Loh, I. C. (1989) Hydrology of the jarrah forest. In B. Dell, J. 

Havel and N. Malaczjuk (eds). The jarrah forest-A complex mediterranean ecosystem, pp 179- 

201. Kluwer Academic Publishers, Dordreck. 

Stoneman G L, P W Rose and H Borg (1988) –Recovery of Forest Density After Intensive Logging in 

the Southern Forest of Western Australia. Technical report no 19. Department of Conservation 

and Land Management. 

Stoneman G L (1993) Hydrological response to thinning a small jarrah ( Eucalyptus marginata) 

catchment . Journal of Hydrology 150; 393-407. 

Stoneman G L, D S Crombie, K Whitford, F J Hingston, R Giles, C C Portlock, J H Galbraith, and G 

M Dimmock- (1996) Growth and water relations of Eucalyptus marginata (jarrah) stands in 

response to thinning and fertilisation. Tree Physiology 16, 267-274 

Storey A ( 2007) Wungong catchment trial project. Aquatic fauna biodiversity assessment. Report to 

the Water Corporation 

Water Corporation (2005)- Wungong Catchment Environment and Water Management project. 

Water Corporation( 2009)-Wungong Whispers vol 8, p2-3.



ABSTRACT 

THE NATIONAL FORESTRY MASTERS PROGRAM: 

A NEW ERA OF COLLABORATION 

IN PROFESSIONAL FORESTRY EDUCATION 

Lyndall Bull 1 , Peter Kanowski 1 

Both competition and cooperation have characterised Australian forestry education over the 

past century. The recent establishment of the National Forestry Masters Program, a 

collaborative framework for graduate professional forestry education, is the most significant 

cooperative endeavour in Australian forestry education since the establishment of the 

Australian Forestry School in 1926. We argue that the NFMP is a logical means for sustaining 

and advancing professional forestry education in the larger context of the Australian higher 

education sector, and that a strategy which identifies and enables the roles and responsibilities 

of all key actors with significant interests in forestry education is necessary to ensure that 

professional forestry education in Australia survives and prospers. 

INTRODUCTION 

Recognition of the need for professional forestry education in Australia dates back to 1887, when the 

Indian Forest Service Conservator, F.D’A Vincent, recommended establishment of a forestry school in 

Victoria (Roche and Dargarvel, 2008). It is now nearly 100 years since the Victorian School of 

Forestry was founded in 1910. Negotiation between the states about the establishment of a national 

forestry school continued until its establishment in 1926 (Carron 1985), and exemplified the mix of 

competition and collaboration between the state agencies and universities that continued to 

characterise Australian forestry education. The recent development of the National Forestry Masters 

Program (NFMP – see http://www.forestry.org.au/masters) is the most explicit example of 

collaboration in forestry education since the establishment of the Australian Forestry School. This time 

round, at least, the initiative includes Victoria. The current shortage of forestry graduates (Figure 1) in 

Australia is now recognised within the sector as a pressing problem. 

80 

70 

60 

50 

40 

30 

20 

10 

0 

1996 

1997 

1998 

1999 

Source: NAFI/A3P 2006, Figure 9.9 

2000 

2001 

2002 

2003 

2004 

2005 

Australian National University 

University of Melbourne 

Southern Cross University 

University of Queensland 

1 Fenner School of Environment & Society, College of Medicine, Biology and Environment, The Australian National 

University, Canberra ACT 0200 Australia and National Forestry Masters Program, Australia. 

Total



Figure 1: University forestry graduates 1996-2005 

The 2006 forestry sector skills report (NAFI/A3P 2006) documented the strong demand for 

professional forestry skills and graduates, with around 70% of forest growing and management 

organisations reporting a shortage of foresters 2 . The lack of suitable candidates to fill vacancies in the 

Australian forest sector has forced many organisations to recruit internationally to fill professional 

vacancies. Whilst the globalisation of forestry employment has many benefits for employees, 

employers and the profession, it also has direct and indirect costs, and the recent reliance of the 

Australian forestry sector on international recruitment suggests substantial local market failure in the 

extent of forester demand – supply imbalance. These trends in undergraduate enrolment are also 

evident in North America and many European countries (Kanowski 2008). 

From the start of Australian forestry education until the 1980s, most students studying forestry in 

Australia were supported by state forestry agencies or other scholarships. Nor were there many 

alternative environment-focused degrees until the late 1980s. As a result, the number of forestry 

students in Australia grew to a peak in the 1980s, and has declined subsequently. There are various 

reasons to which this decline can be attributed; some reflect broader societal changes, such as the 

urban shift in Australia’s population, and an attendant decline in interest in professions associated with 

rural and regional Australia. Australian agricultural science education faces a similar challenge in this 

respect (Pratley and Leigh 2008). Other reasons may also be in common with agriculture, such as 

community perceptions of a dumb, sunset industry rather than a smart, innovative one. Unpleasant and 

misinformed though they may be, such perceptions are not necessarily entirely without foundation – 

for example, the proportion of the agriculture, forestry and fishing sector with a degree qualification, 

at around 7% in 2004, remains well below that for competitor sectors of the economy - eg 17% for 

mining and 24% for services (Productivity Commission 2005, Chapter 5). 

These changes are taking place in the context of diminished national investment in education, at least 

in relative terms. Recent reviews of Australian innovation and education (Commonwealth of Australia 

2008, Cutler 2008) recognised that while higher education lies at the heart of Australia’s research and 

innovation system, Australia is falling behind other countries in investment in its higher education 

system. 

One of the consequences of the diminished investment in higher education, and of the university 

funding models that have prevailed over the past decade, has been the increasing inability of 

universities to sustain programs with low student numbers. This is particularly problematic for 

professional degrees such as forestry, which require breadth across a range of topics and problem 

solving skills (Brown 2003). 

In response to these factors, five Australian universities offering forestry education - Australian 

National University, Southern Cross University, University of Melbourne, University of Queensland 

and University of Tasmania - cooperated to initiate the Australian National Forestry Masters Program 

(http://www.forestry.org.au/masters/), with seed funding over 3 years of $1.56 million from the 

Australian Government. The NFMP was modelled in part on the relatively new European masters 

programs, including two in forestry (SUFONAMA and SUTROFOR), which are predicated on student 

mobility. The seed funding has been used principally to employ program convenors at most 

participating universities, and to provide c. 70 student mobility scholarships. Political support from 

forest sector peak bodies was instrumental in securing the seed funding, and support from the Institute 

of Foresters, the CRCs for Forestry and Bushfire, Greening Australia, and other forest sector 

organisations has been fundamental in delivering the program. 

2 The range of skills shortages reported by the sector incorporates a number of skills sets, both professional (e.g 

Roberts 2007) and vocational. Given the focus of the NFMP, this paper’s focus is limited to tertiary education 

for professional foresters.



THE NATIONAL FORESTRY MASTERS PROGRAM 

The National Forestry Masters Program is a coordinating framework linking graduate coursework 

degrees already offered at the five participating universities. Students enrol at one of the participating 

universities, and follow its degree rules, but can access courses offered by the other partners. The 

program thus offers students access to the best available teaching, field experience, industry and 

research opportunities across Australia. It also encourages the development of professional networks, 

and links students to forestry in the Asia-Pacific region, by requiring students to participate in two 

joint courses, one of which is conducted abroad. 

The NFMP’s collaboration at the postgraduate coursework level, rather than the undergraduate level, 

reflects a number of factors: new institutional initiatives, such as the “Melbourne Model”, which focus 

professional education at the graduate level; the continuing challenges of recruiting students into 

undergraduate forestry programs; and opportunities for encouraging a more diverse group of students, 

including those with first degrees in entirely different topic areas, and those with professional 

experience both within and outside the forest sector. Nor is the approach new for Australian forestry 

education – it emulates, in a contemporary context, that which was used at the Australian Forestry 

School. 

The NFMP therefore plays an important complementary role to the undergraduate forestry programs 

still offered at the Australian National University and Southern Cross University. In addition to 

attracting a different cohort of students, the NFMP is helping to maintain courses – such as forest 

operations – which undergraduate numbers alone cannot sustain. While it may be the case that a twoyear 

graduate program cannot deliver experience and learning identical to that of a traditional fouryear 

undergraduate degree, it is also the case that professionally-oriented masters are becoming a 

common means – both internationally and within Australia - of delivering professional education from 

the basis of a more generalist undergraduate degree, or one in a different topic area. Nor is it prudent 

for the Australian forestry sector to seek to continue to rely on the graduates of undergraduate forestry 

degrees, unless either the numbers of students attracted to undergraduate forestry programs increases 

substantially, or educational policy changes to recognise the need for specific, greater investment to 

sustain specialist undergraduate programs such as forestry. 

The NFMP has been successful in its goals of both attracting new candidates to the profession and 

assisting the further development of individuals already working within the sector. To date, more than 

50 students are participating in the program. Current NFMP students: 

• include individuals from a wide range of backgrounds - including information 

technology, landscape architecture, natural resource management, 

telecommunications and physics; 

• range in age from the mid 20s to the mid 50s, and; 

• are based in five states and territories. 

Most NFMP courses are offered as two-week blocks, to facilitate student mobility and enable 

participation of professionals already in employment. There are both pedagogical advantages and 

disadvantages to this model, as with any other, but it is increasingly common at all levels of tertiary 

education. The learning challenges presented by compression of the course into a concentrated period 

can largely be addressed by pre- and post-contact activities, and by thoughtful structuring of the 

contact time; an associated challenge for many students is finding sufficient time pre- and post-contact 

to prepare adequately, and to complete assessment requirements. 

Other NFMP activities 

The NFMP has also served as a vehicle for collaboration between universities to deliver other 

educational programs. In 2008/9, the NFMP consortium was funded under the Australian 

Government’s Asia Pacific Forestry Skills and Capacity Building Program (APFSCBP), administered 

by the Department of Agriculture, Fisheries and Forestry, to deliver 3 forestry training programs in the 

Asia – Pacific region in 2008/09. These were:



• Certification training – delivered to c. 40 participants from across the region at Deramakot 

Forest, Sabah; 

• Participation in NFMP coursework and a work placement in Australia - for students from 

Malaysia, Laos and the Philippines; and 

• Leadership training for c.10 emerging forestry leaders in the Asia-Pacific region, conducted in 

Japan in association with the British Council’s Climate Cool program. 

The offshore components of the APFSCBP have also been open to small numbers of NFMP students, 

providing them with additional learning and networking opportunities. 

Next steps for the NFMP 

The Australian Government seed funding for the NFMP will cease at the end of 2009, although some 

funds will carry over through 2010. The current NFMP model is predicated on student mobility, and it 

is hard to envisage any collaborative model that does not include some level of student mobility. Staff 

mobility, while possible in principle and already a small part of the NFMP, is less attractive because it 

does not deliver efficiency gains in terms of either class size or lecturer workload. Consequently, the 

central challenges for the NFMP in the near and medium terms are to secure sufficient funding to 

sustain a minimum level of provision of mobility scholarships, and to continue recruiting activities for 

both graduate and undergraduate forestry programs. The Institute of Foresters of Australia, in 

conjunction with the NFMP partner universities, has taken the lead in the former, through the 

establishment of a forestry education trust fund. Supporting that trust fund, and developing and 

sustaining effective recruiting strategies, requires an effective partnership across the whole Australian 

forest sector. 

One way of thinking about this partnership is to identify the respective primary and shared roles and 

responsibilities of the key actors concerned with forestry education. The principal of these are 

summarised in Table 1. 

The structure represented in Table 1 is intended to emphasize that all interested parties have 

complementary and important roles in supporting forestry education. As discussed at the IFA Forestry 

Education Summit in May 2008, and at subsequent meetings, these roles need to be coordinated within 

an overall strategy for fostering forestry education. This strategy would recognise the 

interdependencies between both activities supporting forestry education and the roles of the different 

actors. 

It is imperative that those parties committed to sustaining and developing forestry education – the IFA, 

the universities, and the forest sector businesses who employ foresters and agencies which have 

responsibility for advancing the sector’s development – agree and give effect to a forestry education 

strategy. We have at most until mid-2010 to do this if the NFMP is to be sustained. 

The consequences of not agreeing and implementing such a strategy are likely to be that the 

cooperative mechanism represented by the NFMP will collapse – not for lack of goodwill, but for lack 

of resources to enable it to continue. Should the NFMP not continue, forestry education will revert to 

being the responsibility of individual universities. Unless the policy settings and funding for higher 

education change dramatically to favour forestry, the ultimate consequence is likely to be progressive 

loss of capacity for forestry education at both undergraduate and graduate levels, an enhanced risk that 

forestry education in Australia will end, and further deterioration in the availability of forestry 

professionals with skills relevant to Australian forestry. 

Conversely, sustaining a collaborative model should allow participating universities to evolve their 

contributions to the NFMP, and forestry education more generally, to reflect both their strengths and 

the strategic directions of their institutions, and minimise the risk of loss of forestry education capacity 

nationally. It will also provide the vehicle for continuing collaborative engagement with forestry 

education and training in and for the Asia-Pacific region, and more widely, which will itself further 

support Australian capacity in forestry education.



Table 1. Principal roles and responsibilities of actors associated with forestry education 

Activity 

Partner 

Student recruitment 

Scholarship funding 

Professional development 

and networking 

Ensure program meets 

needs of sector 

Professional body Universities 

Engage membership in 

effective recruiting 

Continue to take lead on 

behalf of the sector 

Facilitate student 

membership, networking 

activities, and mentoring 

Provide fora for connecting 

students and employers (eg 

advertising within 

newsletters, web portal) 

Active role in communicating 

members’ views 

Promote forestry degrees 

within overall recruiting 

strategy 

Pursue funding for forestry 

scholarships within the 

scholarships and 

endowment portfolio 

Actively engage sector in 

course delivery and extracurricula 

events 

Engage with industry 

events 

Establish mechanisms to 

enable external partners to 

contribute to curriculum 

review and development 

Forest sector businesses, 

agencies and bodies 

Promote forestry as a 

profession as part of the 

overall business and 

communication strategy; 

encourage and allow relevant 

staff to pursue degrees 

Recognise the need for 

scholarship funding from the 

sector, and contribute to 

scholarship pool 

Provide vacation work 

placements and internships. 

Be responsive to requests for 

engagement at universities 

Support employees to 

undertake professional 

development opportunities 

within NFMP 

Contribute constructively to 

curriculum review and 

development, cognisant of 

constraints within universities 

Government (DEEWR & 

other relevant departments) 

Recognise and support forestry 

as a sector of national 

importance and critical skills 

shortage 

Recognise the need for 

scholarship funding in the 

sector; contribute to scholarship 

pool; ensure tax regulations 

encourage corporate and 

individual support 

Support for international 

exchange programs and 

placements 

Support program and 

curriculum development to 

meet sector needs. 

95



CONCLUSION 

The NFMP has proved in its initial phase to be a feasible and effective model for delivery of 

forest education in the broader context of the contemporary Australian higher education 

system. It was established and has succeeded to date because of the commitment of many 

partners in the forest sector; it needs their continuing commitment and support to be 

sustained. The cooperative model represented by the NFMP offers the Australian forestry 

sector the best strategy for the continued delivery and further development of specialist 

forestry education, at both graduate and undergraduate levels, as the basis for meeting the 

sector’s needs for professional foresters. The development and implementation of a strategy 

based on the respective roles and responsibilities of key actors with interests in forestry 

education is the next critical step in sustaining forestry education; we have only a little time to 

complete this task. 

REFERENCES 

Brown, N. 2003. “A Critical Review of Forestry Education” Bioscience Education E-journal, 1-4. 

http://www.bioscience.heacademy.ac.uk/journal/vol1/beej-1-4.aspx 

Carron, LT. 1985. A history of forestry in Australia. ANU Press. 355 p. 

Commonwealth of Australia. 2008. “Review of Australian Higher Education” Department of 

Education, Employment and Workplace Relations, Canberra, Australia. 

http://www.deewr.gov.au/he_review_finalreport. 

Cutler, T. 2008. “Report on the Review of the National Innovation System” Department of Innovation, 

Industry, Science and Research, Canberra, Australia. 

http://www.innovation.gov.au/innovationreview/Pages/home.aspx 

Kanowski, P. 2008. “Centennial challenges: professional forestry education in Canada and the USA in 

2007, and learnings for Australia; Forest and Wood Products Australia 2007 Dennis Cullity 

Fellowship Report. Forest and Wood Products Australia, Melbourne Australia 

National Association of Forest Industries and Australian Plantation Products and Paper Industry 

Council. 2006. “Wood and Paper Products Industry Skills Shortage Audit” 

Pratley, J and Leigh, R. 2008. “Agriculture in decline at Australian Universities” Global Issues 

Paddock Action. Proceedings of the 14th Australian Agronomy Conference. September 2008, 

Adelaide South Australia. 

Productivity Commission. 2005. “Trends in Australian Agriculture,” Research Paper, Canberra. 

Roberts, R. 2007. “The role of graduates and their education for the Australian wood processing sector 

– results from a forest industry survey” Forest and Wood Products Research and Development 

Corporation, Melbourne, Australia. 

Roche, M and Dargavel, J. 2008. “Imperial Ethos, Dominions Reality: Forestry Education in New 

Zealand and Australia, 1910–1965” Environment and History. 14: 523-43.



Summary of Presentation 

EDUCATION AND TRAINING PATHWAYS 

Antoinette Hewitt 1 

ForestWorks’ core business is facilitating skills development for the forest, wood, paper and timber 

industry workforce. ForestWorks has Industry Skills Council (ISC) status giving it responsibility for 

the development of nationally consistent competency standards and qualifications. 

Skills Australia’s recent review of the governance of the Vocational Education and Training (VET) 

sector, and the Council of Australian Governments’ (COAG) push to integrate workforce 

development, employment and education priorities, have seen a move towards focusing on unity 

within the tertiary education and training area consisting of the two sectors: VET and Higher 

Education. 

Government is driving a future in which pathways between the two sectors will be increasingly 

seamless with multiple entry and exit points. 

In the push for economic and social sustainability and the pivotal role that forests will play, it is 

increasingly important that forests are managed by a skilled workforce. Alarmingly, trends in 

Australia show that the numbers of professional foresters are declining to the extent that a skills 

shortage exists and, with the current small number of university enrolments, this is unlikely to ease in 

the foreseeable future. 

Over the past year, ForestWorks has been engaging with universities and industry on this issue in 

order to see if ForestWorks can in some way contribute to the efforts to overcome a shortage of people 

skilled in Forest management. This has involved participation in the national Foresters’ Summit in 

Canberra, and more recently with discussions with various universities and industry. In the future 

these discussions must also involve the IFA. 

ForestWorks is now looking at developing a strategy to expand the traditional pathways into Forester 

activities by including existing workers who may be or have been studying in VET programs. Such 

inclusion would boost the potential number of graduates eligible for entry to the Forestry Masters 

Program. Part of this strategy will be to explore whether VET outcomes can be delivered via the 

university-based Masters program allowing additional, funded, VET students to become part of the 

program and thereby increasing the likelihood of maintaining a critical mass of students at the 

universities with forestry courses. 

1 Manager - Industry Skills Council Project, ForestWorks, 559A Queensberry Street 

North Melbourne,Vic 3051. Ph: +61 9321 3512 Email: ahewitt@forestworks.com.au



FORESTRY IS A HANDY WORD BUT NOT THE RIGHT ONE 

FOR ATTRACTING STUDENTS. 

Chris Weston and Katherine Whittaker 1 

ABSTRACT 

Building interest in bachelor qualifications in forest science has proven difficult over the 

last decade, with graduate forester numbers recently falling, and now well short of forest 

sector demand. This paper relates our experience over the last two years in attracting 

students to the Master of Forest Ecosystem Science at The University of Melbourne and 

degree-level forest science in general. Although job and career prospects, teaching mode, 

course location and financial support are all important drivers of student choice of degree 

– an overriding factor is the need to appeal to prospective students’ sense of idealism and 

desire to contribute to big picture issues that are globally significant. In this paper we 

explore this theme and present our vision for the steps to a future well populated with 

forest science graduates. 

INTRODUCTION 

It's hard to believe that a nation of over 20 million people with more than150 million hectares of 

forest, including 40 million hectares of productive open forest formations and almost 2 million 

hectares of plantation forests, would struggle to maintain a forest-specific degree in tertiary education. 

Yet it's true, and the problem has exercised the minds of our best exponents of forest disciplines in 

universities and in the forest sector more broadly for over a decade (Kanowski, 2006). 

Analysing the problem of low student numbers in forestry and the causes is difficult, as there are few 

sources of hard data to draw on other than student numbers, forest sector job numbers and the nature 

of existing tertiary forestry courses (NAFI and A3P, 2006). 

Here we consider the key elements of the problem by first making a brief overview of the 

development of university-level forest education nationally, followed by discussing our experience in 

attracting students to forest science over the last two years. Our focus here is at degree level – 

bachelor and master qualifications – rather than diploma qualifications. 

We propose that awareness of forest science, the role of foresters, and the importance and role of the 

forest sector is poorly developed in the general community and that forestry as a catch all for forest 

disciplines is a poor choice of term for attracting students to bachelor and coursework master degrees. 

It seems the term forestry neither inspires interest nor conveys the broader relevance of forest 

disciplines to society and thus to a new audience of potential students. 

FORESTRY DISCIPLINES - have a long heritage in Australian universities 

Turning briefly to the history of forestry education we see that forestry has been a part of university 

curricula in Australia since the University of Adelaide established a Forestry Department in 1911. The 

forestry disciplines developed at Adelaide were eventually transferred to the Australian Forestry 

School in 1927 under the control of the Commonwealth Forestry Bureau, and reformed as the 

Forestry Department of the ANU in 1965. 

In Victoria, the Forests Act of 1907 prescribed an examination system for the training of professional 

foresters leading to the first intake of forestry diploma students in 1910 at Creswick (Ferguson and 

Youl, 1998). Select graduates from Creswick entered a Bachelor of Science at The University of 

Melbourne (UM) where a Bachelor of Science (Forestry) was eventually formalized in 1943. 

Two aspects of this history of forest education are relevant here. The first is that forest science 

degrees emerged in the university system as a continuation of diploma level forestry qualifications. 

1 School of Forest and Ecosystem Science, University of Melbourne. Tel: 53214103 Email: Weston@unimelb.edu.au



We think this legacy highlights the need to adopt a different name for tertiary or degree level 

qualifications to distinguish them from the more applied and technical diploma forestry qualifications. 

A general lack of clarity around the distinction between diploma and degree qualified foresters is 

likely a contributing factor to declining interest in forestry degrees at university. The second aspect is 

that both State and Federal government funds external to the university system have been crucial to 

the development of forest science disciplines and the training of foresters. Unless student patronage of 

forest courses at universities increases very soon, these external funds will be required again to shore 

up forest disciplines. 

Today forest science disciplines persist across half a dozen or so universities and each of them has 

challenges in maintaining sufficient capacity to offer comprehensive forest science programs. The 

National Forestry Masters Program (NFMP) has begun the difficult task of seeking to establish a 

practical and viable way to offer comprehensive forest science education where the expertise is drawn 

from across the participating universities. 

The whole forest sector has a stake in the success of the NFMP and its drive to support and sustain 

both undergraduate and postgraduate coursework in forestry – as set out by Lyndall Bull, Peter 

Kanowski and the NFMP partners elsewhere in these proceedings. It is crucial that the forest sector 

and the university partners in the NFMP agree on how to move this initiative from seed funding by 

government to an ongoing basis. These aspects are discussed in the NFMP paper. 

ENROLMENTS IN FOREST SCIENCE DEGREES – the Melbourne experience 

Here we briefly review the experience at The University of Melbourne to understand how enrolments 

in forest science have changed over the last few decades. 

Graduates from Melbourne's Bachelor of Science (Forestry), created as a degree in 1943, have 

numbered between about 2 and 20 students from the 1950’s until the early 1980’s. Throughout the 

1980's the Bachelor of Forest Science (BForSci) attracted between 15 and 25 students each year; first 

year enrolments did not exceed 30 students until the early 1990’s. Enrolments increased in 1994 with 

the introduction of combined degrees (five-year with Science or Commerce), reaching a peak of about 

60 students into the 1997 class. The peak years were 1993 to 1997, when between 50 and 62 students 

commenced a degree in forest science. 

Because first year completions were about 10 to 15 less than commencing students during these high 

enrolment years, the maximum number of students entering second year at Creswick has always been 

less than 48 and has only exceeded 40 in two years (1995 and 1997). 

BForSci enrolments, including combined degree students, declined to about 22 in 2002 and to around 

10 to 12 annually over 2003 to 2007, when the degree was discontinued at Melbourne in a university 

level overhaul of undergraduate course offerings. 

To summarize, annually the number of students continuing to 2nd year of the BForSci rose from 

around 10 to 20 in the 1980s to about 40 in the mid 1990s and then fell back to 8 to 10 in the years 

from 2005 to 2007. 

We can see from this history of enrolment in forest science that student numbers have never been 

large – Melbourne has graduated about 1000 BForSc (or equivalent) students over the last 65 years. 

The peak interest in forest science at the University of Melbourne coincided with a relatively fleeting 

popularity of double degrees in the mid 1990s. The 5-year forest science/science or commerce double 

degrees likely lost patronage due to the rising costs of university education for students, combined 

with the increasing tendency to work part-time to support tertiary studies. 

Undoubtedly a rise in the offering of natural resource management and environmental science degrees 

throughout the 1990s had a significant impact on forest science enrolments. These degrees were 

generally of 3 years and more accessible than forest degrees which were 4 years duration and for most 

students incurred travel and living away from home costs.



FROM BACHELOR TO MASTER – the forest discipline at the University of Melbourne 

from 2008 and the National Forestry Masters Program 

In shifting from discipline-specific degrees of 4 or 5 years to a broader 3-year general undergraduate 

degree followed by 2-year graduate programs (The Melbourne Model), the University of Melbourne 

moved the forest science discipline to graduate school. 

The last intake to the Bachelor of Forest Science degree in 2007 has been accompanied by the 

introduction of a two-year graduate coursework program in 2008 – the Master of Forest Ecosystem 

Science (MFES). 

Melbourne's shift of forest science to graduate coursework has provided a strong impetus to the 

National Forestry Masters Program (NFMP) as shown by the enrolment of more than 80% of NFMP 

students in the Master of Forest Ecosystem Science. 

The new coursework masters has been named Master of Forest Ecosystem Science to emphasize the 

central understanding of forests as systems in developing students’ understanding of forest disciplines 

and to appeal more broadly to potential students. Many of the 45 MFES students who have 

commenced in 2008 and 2009 have commented that they were initially attracted to the ecosystem 

component of the degree. Our experience in establishing the MFES has shown that gaining the interest 

of potential students is the first crucial step to increasing enrolments. 

Once engaged in investigating our course, website, or in conversation, the key aspects of a forest 

ecosystem science degree can be explained. Once the students have studied a foundation subject that 

outlines the importance of forests to human welfare both globally and in Australia they come to 

appreciate the importance of more traditional forestry disciplines and to go on to choose these subjects 

in their course. 

The structure and content of the MFES, especially the first few subjects, is very important to drawing 

in students who are new to the discipline area, and showing them the relevance of forests and forest 

disciplines to big picture problems facing society. 

In becoming wise to the relevance of forest sciences the students can then choose to study any or each 

of technical, social and policy or economic aspects of forestry. So far we have found that students new 

to forestry relish the opportunity to study across a broad range of forest subjects, the niche 

specializations offered, and to take advantage of national and international study options. 

STIMULATING STUDENT INTEREST IN FOREST SCIENCE 

Searle and Bryant (2009) surveyed students in the Bachelor of Science (Forestry) at ANU in the late 

1990s and from responses concluded that general lack of public awareness of forestry and the 

relevance of forest disciplines was central to declining student numbers. Their paper describes well the 

declining interest in courses labelled forestry and their conclusions accord with our experience at 

Melbourne. 

We think it is unfruitful to persist with the term forestry in the naming of university degrees and in 

promoting professional forester qualifications. Forestry in a degree title or in the promotional text and 

dialogue conveys a narrow industrial focus that does not reflect broader course content and the full 

range of career possibilities known to arise from a degree in forest science. 

The forestry moniker thus fails to capitalise on interest in new and emerging employment 

opportunities for which forest science professionals are well qualified. Attracting student interest in 

forest science qualifications is our biggest challenge. 

To illustrate our key points we summarized in Table 1 our strategy in promoting the Master of Forest 

Ecosystem Science.



Table 1. A summary of our key messages and actions in both promoting the Master of Forest 

Ecosystem Science and maintaining students in the course. 

1. We have emphasized the role of forests in key environmental issues facing society 

Climate change – forests central to carbon cycling, C sequestration and biofuels 

Water – forested catchment management and protection crucial for water security 

Fire – knowledge of forests and fire central to community response to rising fire risk 

Landscape restoration – is a big part of future farming landscapes and land management 

2. We have made explicit the distinctive features of studying in the Master of Forest Ecosystem 

Science relative to more general environmental science degrees 

Study based on a core of forest subjects with a wide choice of electives 

Two week intensive teaching allows for study during short breaks from work 

Strong emphasis on field experience and related practical exercises 

The course retains a good mix of traditional forest science subjects such as silviculture, wood 

science, forest operations and forest inventory – these subjects are rarely available elsewhere 

and are a distinctive feature of the course 

Part-time study is possible and students can start in the program at any stage throughout the year 

The course structure and learning and teaching mode results in strong cohort and collaborative 

learning experience 

Strong forest sector employer connections linked to internship study opportunities and research 

project work. 

Good opportunities for developing a professional network including international study and 

employment opportunities 

3. We emphasize the financial incentives 

NFMP scholarships are attractive to students and important in differentiating forest masters 

from more general environmental masters courses 

The availability of commonwealth supported places and eligibility for Centrelink payments 

Students eligible for financial assistance from the University of Melbourne's Graduate Access 

program 

A professional coursework degree in contrast to many more general environmental management 

degrees 

THE UNIVERSITY ENVIRONMENT – challenges in maintaining forestry academics and their 

disciplines 

A key driver of the NFMP – as discussed in Lyndall Bull and Peter Kanowski's contribution to this 

conference – is the decline in forestry disciplines in universities. Broadly, academics are selected and 

maintained in universities based on potential teaching and research demand in their area of expertise, 

and on their record and perceived ability to maintain a strong original research output as demonstrated 

by publication in international journals.



Academics who score highly on ability to attract teaching income, research higher degree students and 

external research grants, survive and prosper in the system. With dwindling student teaching loads, 

forest academics are challenged to maintain relevance in competitive university environments. 

In Australia we have already reached the point at which comprehensive forestry programs are difficult 

to mount with the existing academic staff. Universities have dealt with this by maintaining a small 

core of forestry academic staff and have contracted external expertise to cover teaching in essential 

forestry disciplines. The NFMP offers a better solution to this problem as it has the potential to 

strengthen the position of existing academic appointments. 

CONCLUSIONS – a changing context for tertiary forest education in Australia 

Degree qualified foresters carry, develop and contribute knowledge and skills essential for the social, 

economic and environmental wellbeing of our society. The problem is that you have to be one or know 

one to recognize the critical role of foresters' skills in society. All foresters have a key role in 

educating the community as a basis for attracting prospective students into tertiary forest education. 

Foresters can take advantage of the rapid emergence of climate change in the public consciousness, 

and the many challenges it poses that link forests with community safety and wellbeing – such as 

management of fire risk, water security in forest catchments, and growth and protection of carbon 

sinks for greenhouse gas reduction. This provides a good opportunity to promote forester skills. 

Knowledge and expertise in these areas can be linked to the more traditional understanding of foresters 

roles — e.g. in the stewardship of forests for biodiversity conservation, wood for high value timber 

and paper products. This strategy will help the community to understand the crucial role of foresters 

and lead to interest in forest science in education. 

Our key point, repeated throughout this paper, is that forestry is not the right term for the forest 

profession and forest educators to adopt in the ongoing campaign to attract new students – as it 

conveys a narrow industrial concept of foresters' skills. It is time to universally embrace new language 

such as forest sector instead of forest industry, forest ecosystem services or benefits instead of forest 

products, and forest science instead of forestry. 

As professional foresters we should redouble our efforts to speak with the broader community in 

language that leads community attitudes and shows the immediate relevance of foresters to the 

pressing issues of the times from practical through to policy level, as well as reinforce the broad range 

of employment available. In doing so we will be appealing to the idealism of potential students and 

attracting their patronage. 

Finally, in reaching out to a broader audience we must look ahead and speak in positives rather than 

look back and search among the dark perceptions of forestry – there is too much at stake for Australia 

as a nation if we loose our forest disciplines in tertiary education. 

REFERENCES 

Searle S, Bryant C (2009) Why students choose to study for a forestry degree and implications for the forestry 

profession. Australian Forestry 72: 71-79. 

NAFI and A3P (2006) Wood and paper products industry skills shortage audit. NAFI and A3P 249 pp. 

http://www.nafi.com.au/skills/ (accessed July 2009). 

Kanowski PJ (2006) Forestry education – where are the students and what should we do? Australian Forestry 4: 

241-241. 

Ferguson IS, Youl R (1998) Forestry at Creswick and the University, 1910. 

http://www.landfood.unimelb.edu.au/dean/book2/index.html (accessed July 2009)



ABSTRACT 

TRAINING OPERATIONS STAFF IN THE FIELD - 

CHALLENGES AND OPPORTUNITIES 

Matthew Doig 1 

The last decade has seen major changes in the role of Forest Operations staff. 

They are now expected to undertake and supervise tasks with a greater degree 

of complexity and responsibility than in the past. Foresters now face major 

challenges in training these people in the various aspects of forest management. 

A project to develop a series of operationally based Capability Statements and 

associated Assessment Tools covering key areas of forest operations is 

examined. This examination shows how these Capabilities and Tools ensure a 

consistency of approach and clear directions for trainers. 

The paper highlights the following: 

• The greater attention needed to meet Operations Staff training 

needs. 

• Industry based Capability Standards can help meet that need. 

• Training can be better targeted, and validated, through such 

capabilities, together with Assessment Tools. 

• Capabilities and Tools can complement and strengthen existing 

training activities. 

INTRODUCTION 

Forestry, professionally and operationally, has undergone significant changes over the last 

twenty years. This has been reflected in such things as the technology employed and in the 

way Forestry work itself is undertaken. A 1949 promotional brochure for the School of 

Forestry Creswick lists what foresters will learn and apply in the course of their career. Some 

of the tasks mentioned on this brochure are now very much in the province of operations staff, 

many of whom do not hold any qualifications in Forestry and who formerly operated at the 

sub-professional level under direction from foresters. As this situation becomes more 

common the need for better training for this emerging group assumes greater urgency. 

Providing training for operations staff continues to be a significant challenge for Forestry 

organisations, particularly due to the limited options such people have in terms of external 

programs focused on them. This training response can be better achieved via a process of 

identifying industry or even company specific capabilities and developing assessment tools 

with which to validate learning. This can be illustrated with the example of VicForests which 

has responsibility for the development of many operations staff and which has sought to 

ensure a consistent level of training via such an approach. 

VicForests is a leading native forest timber harvesting and sales business in Victoria, 

Australia. According to its mission statement it seeks to ensure “a key role in ensuring the 

economic prosperity of Victoria’s timber industry while ensuring operations are sustainable in 

the long term.” VicForests prides itself on “following strict sustainable forest management 

systems and standards that are globally recognized” (VicForests 2008). Like all major forestry 

organisations, such a statement has implications for clearly defining the required capabilities 

of staff, addressing them through appropriate training and ensuring that skill and knowledge 

sets which support certification of its products and operations are clearly validated. 

1 Melliodora Solutions, 13 Hamel Street, Hampton, Vic. Ph: 0412081199 Email: mattdoig@bigpond.net.au.



VicForests identified the need to develop and implement its own practical, operationally 

based Capability Statements and supporting Assessment Tools, which would support its 

Certification process and ongoing adherence to Australian Forestry standards. The 

compliance-based nature of much of the work VicForests undertakes further highlighted the 

need for staff capabilities to be comprehensively validated. 

Existing training packages at the VET level were not considered comprehensive enough to 

cover Forestry Operations in a stand alone sense and did not provide the depth to cover the 

needs of operations staff. A decision was made to draft the Capabilities in a similar format to 

Competency Standards used within the VET system, but with an emphasis on being practical, 

widely used and operationally based. Complementary assessment tools were also required to 

be developed along these lines. Technical input was provided by employees within VicForests 

who had current responsibility for training operations staff in the various areas covered by the 

capability. 

How then are operations staff classified? Such staff can include employees who are subprofessional, 

i.e. who do not possess qualifications in forestry, but who are nonetheless 

increasingly doing work formerly completed by qualified foresters. Anecdotal comments 

suggest that the vision of an operations staff member who had a patch in the field and walked 

this as part of his work is no longer the case and in many instances they are now actually 

office bound and carrying out supervision of contractors or other junior staff. 

So how has the role of Operations staff changed over time? Whilst Foresters still retain 

control over key areas, there is now more significant delegation of duties to the subprofessional 

groups. Experience with students who undertook the old Diploma in Forestry at 

The School of Forestry, Creswick, provided insights into the expanded range of responsibility 

being delegated to Operations staff. One observation was that everything in forestry had 

moved up a rung, with contractors carrying out a lot of work formerly performed by 

operations staff, and the operations staff doing a lot of the work formerly done by foresters. 

The title of “Forest Supervisor”, which many organisations now use, serves to remind us of 

the expanded role of operations people generally. 

Whilst there remains a strong need for industry and company focused training for this group, 

their actual opportunities for training beyond their organisations are somewhat limited. The 

former Diploma in Forestry provided an opportunity for these people to acquire some 

professional recognition, and the theoretical and practical foundations important to the sort of 

work they were doing. The course provided opportunities for people who wished to pursue 

careers as technical supervisors in the forestry and forest industries and as such was more 

vocationally oriented. 

One feature of the course was the large number of students who were drawn from Forestry 

Operations within companies across Australia and who were sponsored by their employers. 

There was an acknowledgement that this sort of training was unique and also complementary 

to the sort of work they would be expected to do. The block release nature of the course also 

assisted in them getting time off work to attend, particularly for those from interstate. 

With the demise of this program, there are fewer options for this group to receive training 

outside of their own organisations. This makes it more likely that training of this group is 

going to have to be provided “in house” by existing staff. As part of certification, forestry 

organisations must be able to show that appropriate training of staff has been carried out and 

ensure that this is done properly against correct performance measures. 

This necessity to train means that a framework must be used to ensure that such training 

covers all necessary areas of skill and knowledge. This raises a clear need for such training to 

be conducted against standards and measures. Whilst these exist in the form of various 

technical notes, publications and guidelines, they are not always user friendly. Hence, 

VicForests desire to ensure that their training of operations staff be conducted against a 

standardised set of capabilities reflecting current practice. VicForests quest is to ensure



standards are adhered to via reference to Capability Statements. In the VET sector there are 

many thousands of nationally endorsed competencies which are grouped together into 

qualifications. Whilst many are generic in nature and rely on further interpretation down to a 

certain level, there are often specific things missing which are felt to be essential to a 

company being able to meet its compliance requirements. For this reason VicForests sought 

its own specific capabilities with which they can confidently undertake staff development. 

The Capability Statements developed for VicForests have been designed to broadly reflect the 

design of Competency standards in the VET system. The key structure should equate with 

overall VET standards but should also reflect detail based on the level of operations with 

which the company is involved. The capability statements themselves need to closely match 

the suite of tasks required to be carried out by operations staff in support of the overall 

company mission. For this reason they must highlight quite clearly the areas of skill and 

knowledge required and how this is to be applied in a range of contexts. 

In the case of the VicForests capabilities, the foundation for this was the Native Forest 

Silvicultural Guidelines produced by the then Department of Conservation and Natural 

Resources. These guidelines already contained excellent information on the knowledge and 

application of skills required within these areas of silvicutural activity. Whilst these served as 

a useful guide, the process involved in using them to deliver training was not well defined. 

The information they contain is sequentially based, and there is a lot of detail, but they were 

more of a reference and key points were not easily extracted for training purposes. For this 

reason the guides were used as the basis for the capability statements which were themselves 

based on the key result areas described in the guidelines. 

A Capability Statement and the first element relating to Browsing Management is provided 

below. 

BROWSING MANAGEMENT CAPABILITY STATEMENT 

Description 

This statement specifies the capability required to successfully manage the browsing 

of regeneration and reforestation sites and undertake appropriate risk assessment 

and monitoring activities necessary to mitigate the effects of browsing on company 

operations. 

Application of Capability 

Foresters within Vic Forests need to understand the impact of browsing on 

regeneration and reforestation operations and the necessary steps required to protect 

these operations from predation from a range of animals. This will include skills 

involved in the identification of animal species and trees involved, assessing types of 

damage and planning, and implementing and assessing various control measures. 

This capability is required across selected coupe sites experiencing browsing 

pressure. 

Element 1: Assess Browsing activities 

Performance Measures 

1.1 Use knowledge of pertinent forest types and coupe locations to determine 

areas likely to be affected by browsing 

1.2 Investigate history of sites to determine browsing impact and likely affect on 

future operations 

1.3 Be able to identify the key browsing animals and the species of tree most 

susceptible to browsing predation



Evidence Guidelines for Supervisor/assessor 

The following list provides guidance to the Supervisor/Assessor of the key evidence 

categories required to be matched in order to determine that the capability has been 

achieved. To demonstrate achievement against this capability, employees must be 

able to provide evidence that they have an appropriate knowledge of underlying 

principles relating to Browsing Management as well as an ability to apply this 

knowledge via necessary procedures and techniques. The Assessment Tool located 

at section 2 provides the Supervisor/Assessor with the necessary means to obtain 

this evidence via questioning, observed activities, projects etc. 

Knowledge Requirements: 

Forests types; Browsing animal types; Susceptible tree species; Browsing control 

methods 

Relevant codes of practice and instructions 

Skills and attributes: 

Ability to interpret key principles, codes of practice, instructions and written 

procedures 

Ability to apply the above in a practical context or in support of other operations 

High level communication skills to relate knowledge to others 

Analytical skills to apply techniques and procedures in a range of contexts 

Research/Investigation, planning and self-organisation skills 

Field identification of plants and animals 

Application of browsing risk assessment methods 

Practical examples of Evidence: 

Presentation of relevant sections of guides, instructions and procedures to support 

actions 

Demonstration of plant animal identification techniques in a field setting 

Record of application of browsing risk assessment method 

Examples of completed and verified remedial actions relating to Browsing 

Management 

Oral and written explanations of browsing control methods applied in different 

situations 

Resources Required for Evidence Collection 

Natural Forest Silviculture Guideline No 7 

Code of Forest Practices 

Copy of browsing risk assessment sheets 

Coupe plans 

Field location for practical demonstration of skills associated with Browsing 

Management 

Equipment used in browsing control 

In the above example, the Capability has been interpreted from the Natural Forest Silviculture 

guidelines and the following key sections developed. There is a general description which 

provides an overview of the capability and what the knowledge and skills within contribute to 

in terms of completing work place tasks. This is designed to enable a manager to get a 

snapshot of what the capability enables staff to do and how this fits in a particular location or 

context. There is also a brief guide on how this capability might be applied and a description 

of its application in terms of a particular location or context. The capability is divided into 

elements which enable its key parts to be described. This is important both in terms of being 

able to show what the capability actually this is required. 

Each element has key Performance Measures which give a clear indication of what a staff 

member has to be able to do, know and think in order to demonstrate the appropriate skills, 

knowledge and attitude to carry this out in a work setting. The emphasis on skills, knowledge 

and attitude is particularly important in specifying what the people carrying out the operations 

to an acceptable standard must be able to demonstrate. These are provided in a sequential and



easily understood format so that a manager can look quickly at the required measures for each 

element. 

An Evidence Guide is included to assist the manager in the process of meeting each of the 

performance measures through submission of appropriate evidence. This gives listings under 

the headings of Knowledge Requirements, Skills and Attributes and Practical examples of 

Evidence, and can be used as part of an assessment process as a guide for both the 

training/assessing manager and the staff member. 

To assist the manager’s use of these lists there is also a guide to the resources required for the 

evidence collection. This includes various internal and external documents such as the Native 

Forest Silviculture Guideline and legislation, as well as pertinent equipment. This is to ensure 

that the capabilities can be tested in the appropriate setting and the manager has recourse to 

these resources whilst planning and conducting assessment. 

Research undertaken as part of the Capability project has shown that whilst the overall 

standard of field training is good, it has tended to be more dependent on the style and whims 

of individual trainers rather than adhering to a set formula in terms of what each person needs 

to know and be able to do. Training in the operational setting does need to suit local 

conditions and requirements but still must be able to be delivered in such a way that it is seen 

to be adhering to major forestry and management principles, which in themselves are a key 

part of appropriate compliance and certification. Staff will only be well trained if they are 

trained correctly and this training can be tested to the satisfaction of management against 

objective results. 

This testing process has traditionally been left to managers who would mark off someone’s 

development against internal performance management plans. The capabilities developed 

specifically for VicForests are able to be used across a range of operations in different 

locations and contexts, and provide better confirmation of the skills and knowledge being 

attained for the process of operational effectiveness. Testing of these capabilities had to be 

easily achievable by managers who, whilst well versed in their areas of expertise, were not 

necessarily experienced in the training function. VicForests also addressed this by arranging 

training for selected staff in units drawn from the TAA40104 Certificate IV in Training and 

Assessment. This was done prior to the Capability project to ensure that more staff 

responsible for training in the field have the necessary skills in training and assessment, which 

makes their background and ability in Forestry more easily deployable to the task of the 

ongoing development of staff. 

Whilst the training of trainers has been an important step, there will always be, due to 

transfers, changing roles and attrition, a group of people responsible for the training of 

operations staff who have not had the benefit of this kind of development. In this instance, 

there is a great need for both the Capability Statement and associated Assessment Tool to be 

as complementary and self explanatory as possible. For this reason the language on the 

Capability statement was written to reflect the “jargon” of Forest Operations whilst at the 

same time being relatively easy to follow by people newer to the industry who were both 

training and being trained. 

In order to facilitate this process, a complementary and easy to use Assessment Tool was 

developed concurrently with the Capability statement. This is designed to be used by anyone 

with responsibility for training (see below). The tool commences with instructions to assist 

the Manager who is responsible for the assessment of the particular capability. This includes 

not only advice on the process of assessment but also measures to ensure that staff with 

existing skills and knowledge pertinent to the capability could produce this in lieu of having 

to undertake some or all of the Assessment tasks. This will also speed up the process, where 

the need to test against the capability remains but potentially avoids time consuming 

questioning, demonstrations etc.



The assessor is taken through the three recommended forums for the Capability (in this case 

Browsing Management) which are used to collect evidence of the staff member’s capacity to 

carry out all aspects. 

The interview, questioning and activities are all carried out under each element of the 

capability so that the assessor can be confident that evidence is collected against the whole 

capability and the staff member can also follow this process sequentially. Of critical 

importance to the process is the need to carefully record the results of the Assessment against 

the Capabilities Elements and Performance Measures. The Assessment Tools final section is a 

recording grid which mirrors the Capability Statement but also includes the methods of 

assessment used, evidence submitted by the staff member and a confirmation that the 

capability was demonstrated against every element. In this way nothing can be left out and the 

trainer can feel confident in signing of the staff member. 

BROWSING MANAGEMENT ASSESSMENT TOOL 

Instructions for Assessor 

Depending on the nature of the duties required from the staff member you may need to 

assess against some or all of the elements of the Capability Statement. If you normally work 

with the applicant you will be able to complete the record of evidence progressively over an 

extended period. To undertake assessment you will need to see the candidate in a one on 

one capacity to enable questioning to be undertaken, skills to be demonstrated and any 

records of past work experience to be examined against the Capability. In some cases the 

staff member may be able to demonstrate that they already have the required capability or 

parts of it and this will need to be verified via past work records and supervisor verification. 

You should also make arrangements to observe the candidate undertaking tasks and 

applying knowledge in a practical setting. 

Evidence to be gathered 

The Assessor should aim to collect three types of evidence as part of the verification process 

against the Capability. One or more types are required against each performance measure 

and should be written beside the measure in the record of evidence column. Evidence will be 

collected via: 

Interview 

The candidate should undertake an initial interview with the assessor to determine their 

current duties experience. The process of assessment and evidence collection should 

also be explained and any previous work experience verified and recorded. 

Questioning: 

Questions are used to collect evidence about the candidate’s knowledge. These must be 

answered without help although the assessor may assist by clarifying questions. 

Activities: 

Candidates will be observed undertaking activities specified under each element of the 

assessment tool and demonstrating various skills in the workplace. The assessor will 

record all observations and demonstrations. 


Interview: 

1. What kinds of experience have you had in assessing the impact of browsing animals 

on regenerating Eucalypt forests? 

2. Can you produce verifiable records of past activities/achievements in this element?



Questions: 

1. How does the forest type and location of a coupe influence the extent to which it is 

susceptible to browsing damage? 

2. In what ways can the history of a site be useful in determining future browsing 

potential? 

3. Describe some of the key browsing species within your region and the impact they 

have on various trees. 

Observation of Activities/Demonstration of Skills: 

1. Produce a copy of the relevant guidelines and instructions relating to Browsing 

Management. Show the assessor the pertinent sections relating to the management 

of biodiversity. 

2. In a field location, identify evidence of predation on production tree species. 

3. Using evidence such as tracks and scat, determine which browsing species are 

present within a coupe. ASSESSMENT RECORD 


Date of Assessment: Methods of 

Assessment 

1.1 Use knowledge of pertinent forest types and coupe 

locations to determine areas likely to be affected by 

browsing 

1.2 Investigate history of sites to determine browsing 

impact and likely affect on future operations 

1.3 Be able to identify the key browsing animals and 

the species of tree most susceptible to browsing 

predation 

Interview 

Questioning 

Observation 

Demonstration 

Evidence 

supplied 

Capability 

demonstrated 

Yes/No 

Assessor Comments: 

________________________________________________________________ 

________________________________________________________________ 

________________________________________________________________ 

________________________________________________________________ 

________________________________________________________________ 

________________________________________________________________ 

________________________________________________________________ 

Assessor Signature: ________________________________________ 

Name: ___________________________________________________ 

Workplace Coach/Supervisor Signature: ________________________ 

Name: ___________________________________________________ 

Date: Recommendation 

C / NYC



The Capability Statements and Assessment Tools currently being trialed by VicForests are 

designed to support rather than replace conventional training. The time honoured process of 

learning on the job that both graduates and other field workers must go through can be 

expected to continue, but it will now be conducted within a more identifiable framework 

provided by the Capabilities. The major change required will relate to the use of the 

Capability Statements to ensure that the training undertaken covers all required areas of 

performance and that trainees are in a position at the end of this demonstrate achievement of 

the Capability against all the required tasks on the Assessment Tool. Because the Capabilities 

were derived in large part from the DSE instructions already used by the company, they 

should not represent a large departure from current and accepted practices and if anything, 

should make the process of drawing necessary training approaches simpler as the Capabilities 

still sub head the key skill and knowledge areas. 

The trialing process at VicForests is seeking comment from trainers in the field on how well 

they have been able to apply training from the Capability Statements and how this has fitted 

in with the Assessment Tools. The project was careful not to be too proscriptive in terms of 

the way training should be conducted, as this will obviously vary somewhat due to locations, 

personnel and work contexts, but having a clear structure and starting point as well as 

definitive final results will help the process greatly. 

The extent to which the Capabilities and Assessment Tools are easily used will be revealed by 

field trials currently being carried out by VicForests. This involves a number of managers 

using the newly developed format to carry out development activities with their own 

Operations Staff. 

Whilst it could not be expected to be a flawless process, it is anticipated that such an approach 

will boost an area of Forestry training in major need of attention. There is also a continuing 

challenge of how the sub-professional groups can achieve recognition beyond validation 

against company Capabilities, but being able to undertake a formal process like this is a step 

in the right direction. 

REFERENCES 

VicForests Australia 2009, viewed 26 th February 2009, 

http://www.vicforests.com.au/ 

Mark Poynter, Peter Fagg, 2005. Native Forest Silviculture Guideline No 7, Department of 

Sustainability and Environment,



PRACTICAL USE OF 3P SAMPLING IN FOREST INVENTORY 

Phil West 1 

ABSTRACT 

Before starting an inventory of a forest area, broad-scale information is often available 

about the forest area. Commonly, this is obtained through remote sensing, using aerial 

photography, airborne laser scanning or satellite imagery. In combination with ground 

measurement of a sample of stands, this information is used to determine the mean and 

confidence limit for whatever is being assessed in the inventory (stand sequestered carbon 

say); sampling with probability proportional to size, model-based sampling and stratified 

random sampling are techniques used in such cases. When this sort of prior information is 

not available, sampling with probability proportional to prediction (3P sampling) may 

serve a similar purpose; the alternative is simple random sampling, which is generally a 

far less efficient sampling technique. Previously, 3P sampling has required that each and 

every individual sampling unit in the entire population be visited during the sampling 

process; this has restricted its use in inventory to small populations only. Recent 

developments have removed this requirement. This paper describes those developments 

and applies 3P sampling to example forest populations. 

INTRODUCTION 

Forest inventory generally involves the selection of a sample of stands from throughout the total forest 

area being included in the inventory (that is, from throughout the forest ‘population’). In each of those 

sample stands, measurements are made on the ground of whatever it is it is desired to assess in the 

inventory, perhaps wood volume or the amount of carbon sequestered in the tree biomass of the forest. 

These sample data are then used to estimate the stand mean for the population, of whatever is being 

assessed, together with a confidence limit about that mean. When the mean and its confidence limit are 

multiplied by the area of the entire forest, an estimate is obtained of the amount available across the 

entire forest of whatever is being assessed. 

In forest inventory today, additional information about the forest is frequently obtained through remote 

sensing, using techniques such as aerial photography, airborne laser scanning or satellite imagery. This 

can provide certain measures of the forest at any point across its entire area, at least to the resolution 

possible with the remote sensing device; such measures might be the average height of the trees, their 

crown cover or some measure of the productive capacity of the forest at any point. These measures 

may be correlated, at least partially, with whatever it is that is being assessed in the inventory. For 

example, it would be expected that the amount of carbon sequestered at any point in the forest would 

be larger if the trees were taller at that point. 

Variables which are correlated with whatever is being assessed in inventory are known as ‘covariate’ 

variables. When values of covariates are available for points right across the entire forest area, those 

values may be used to improve the ‘efficiency’ of the inventory. That is, they may allow use of 

inventory methods which produce a lower confidence limit of the estimate of the mean with the same 

sample size. Given the substantial expense involved in forest inventory, a method which provides a 

more efficient result is clearly desirable. Inventory techniques such as ‘sampling with probability 

proportional to size’, ‘model-based sampling’ or ‘stratified random sampling’ all make use of such 

covariate information to provide a more efficient inventory. Descriptions of these techniques are 

available in texts on forest inventory (eg Schreuder et al. 1993, Shiver and Borders 1996, West 2009). 

1 SciWest Consulting, 16 Windsor Court, Goonellabah, NSW 2480 and School of Environmental Science and Management, 

Southern Cross University, Lismore, NSW 2480. Email: pwest@nor.com.au. Web: http://www.nor.com.au/users/pwest



However, suppose that such covariate information is not available, whether from remote sensing or 

from any other source. Remote sensing is expensive (involving things like aircraft hire or purchase of 

satellite imagery) and usually requires that specialists be employed to extract the required covariate 

information and make it available in a readily accessible form. Smaller forest owners may not be able 

to afford the expense, or, perhaps it is desired to undertake a preliminary inventory of a forest area, on 

a budget which precludes such expense. 

Under these circumstances, the most basic method by which a sample is selected in inventory, ‘simple 

random sampling’ may be used; in that type of sampling, any point in the entire forest area has the 

same chance of being included as part of the sample. An option is to use ‘sampling with probability 

proportional to prediction’ (often abbreviated as ‘3P sampling’). This does not require that covariate 

information be available from the entire inventory area. However, it has the potential to provide an 

estimate from the inventory which is just as efficient as that which might be obtained when the 

covariate information is available. 

Whilst 3P sampling has been used to some extent in forest inventory in the past, some restrictions in 

the way it has been practised has limited its usefulness generally. Recent developments are removing 

those restrictions. This paper describes those developments and considers some practical applications 

of the method. 

3P SAMPLING 

The 3P sampling method was invented in the 1960s by the American forest biometrician L.R. 

Grosenbaugh. Shiver and Borders (1996) and Iles (2003) give some of the background of how the 

method developed. In America particularly, it seems to have had some use in forest inventory. 

The method requires that a suitable covariate variable be identified, a variable which can be quickly 

and easily measured at any point in the forest. However, there is no need to have available values for 

that covariate variable measured right across the entire forest population. Instead, 3P sampling 

requires that the covariate be measured only for those sampling units (that is, points in the population 

which might constitute one of the inventory sample points) which are actually being considered for 

inclusion in the sample and only at the time the sample is actually being selected and measured in the 

field. 

Traditionally, the covariate variable used in 3P sampling has been a visual estimate in the field, by the 

person doing the sampling, of whatever is being assessed in the inventory (stand sequestered carbon, 

say); as long as the sampler has some reasonable experience of the forest concerned, such estimates 

should correlate quite well with the variable being assessed and so be suitable as covariate values. 

However, the present author has recognised that any other quickly and easily measured variable, 

which correlates to a reasonable extent with the variable being assessed, can be used as the covariate. 

Sample selection 

Before starting the selection of the sampling units to be included in a 3P sample, the sampler must 

obtain an estimate of the maximum and minimum values of the covariate which will be encountered 

anywhere in the population. These would usually be determined during a preliminary survey of the 

population. A decision has to be made also as to what sample size is desired for the inventory. 

Given these decisions, the sampler then sets out to select the 3P sample. In the field, the sampler visits 

a sampling unit. Whilst the choice of which sampling units are visited should be made objectively, the 

order in which they are visited is not particularly important. However, by the time the sample selection 

has been completed, the sampler should have visited sampling units spread widely throughout the 

entire population. 

As each sampling unit is visited, the sampler obtains the value of the covariate variable for it. This 

may be done by the ‘estimation by eye’ process, as discussed above, or by any other measurement 

method. The important thing is that the covariate value should be obtained quickly and easily, to 

minimise the time involved in doing so. 

Once the covariate value has been obtained for a sampling unit, the sampler selects a value chosen at 

random from within the range of the minimum and maximum values assumed for the covariate. To



obtain random values in the field, the sampler might carry a calculator or laptop computer or might 

simply refer to a list of such values, printed before sampling started. 

If the covariate value is greater than or equal to the random value chosen for that sampling unit, the 

sampling unit is then included in the sample. The measurement crew accompanying the sampler would 

then measure the actual value, for that sampling unit, of the variable that is being assessed ultimately 

in the inventory. 

If the covariate value is less than the random value, that sampling unit is simply ignored and the 

sampler moves on to the next sampling unit. This process continues until the required number of 

sampling units has been selected to make up the sample. 

The disadvantage of 3P sampling is that many sampling units may be visited and rejected from the 

sample. If an inventory is being carried out over a large forest area, a lot of time is usually spent by the 

sampler and the measuring crew moving around the forest to visit sampling units. In the case of 3P 

sampling, a lot of time and effort can be spent getting to a sampling unit, only to have it rejected 

immediately from the sample. 

However, it is not this disadvantage which has rendered 3P sampling to only have limited use in forest 

inventory to date. Over the years since it was introduced, a more or less ‘standard’ protocol (in the 

terminology of West 2005) has been developed by American foresters as to how 3P sampling should 

be done (e.g. Shiver and Borders 1996, Avery and Burkhart 2002, Iles 2003). In particular, this 

protocol has required that each and every sampling unit in the entire population be visited in the field 

and a covariate value be obtained for each. Of course this is quite impractical if the population is very 

large, with many thousands of sampling units. As well, this protocol has considered that the only way 

of obtaining the covariate value for any sampling unit is ‘estimation by eye’ of the variable being 

assessed ultimately by the inventory. 

It has now been recognised (West 2005, 2009) that there is an alternative way of viewing 3P sampling. 

This no longer requires that a covariate value be obtained for each and every sampling unit. Instead, it 

is necessary to obtain a covariate value only for those sampling units which are visited during the 

sample selection process. It recognises also that any variable is suitable as a covariate variable, not just 

an ‘estimated by eye’ variable, as long as the variable is quick and easy to measure and has some 

reasonable degree of correlation with whatever variable is being assessed ultimately by the inventory. 

The removal of both these constraints from the 3P sampling process renders the method generally 

more practical for use in forest inventory. 

The only disadvantage of West’s approach to 3P sampling is that the whole process is rendered invalid 

if, during the selection of the 3P sample, a sampling unit is encountered with a covariate value outside 

the range within which it was assumed they would lie. This problem is allowed for in the standard 

protocol. 

Calculation of results 

The key to West’s approach to 3P sampling is recognition that it is in fact a form of ‘sampling with 

variable probability of selection’. This means that different sampling units within the population are 

assigned different probabilities (that is, different chances) of being selected as part of the sample. 

Provided the probabilities are assigned appropriately, this can lead to a more efficient form of 

sampling than simple random sampling. 

The properties and mathematical approach to sampling with varying probability of selection are well 

known (eg Schreuder et al. 1993). Earlier work on 3P sampling did not use this approach and unique 

mathematical treatments were developed for it (see Shivers and Borders 1996, pp. 303-305 and Avery 

and Burkhart 2002, pp. 262-263, as reiterated by West 2005). 

Suppose a forest population contains a total of N sampling units, any of which might be included in a 

sample of size n drawn from it. Suppose that cx and cm are the maximum and minimum, respectively, 

values, which are assumed to occur anywhere in the population, of a covariate variable. When 3P 

sampling is done, as described above, to select from a population a sample of size n, West (2005,



2009) recognised that the i th (i=1…n) of those sampling units had a probability pi that it was included 

in the sample, where 

pi = (nv/N)(ci−cm)/(cx−cm) (1) 

where ci is the covariate value measured on that sampling unit and nv is the total number of sampling 

units which were visited in the 3P sampling process before the final n were selected in the sample. 

Suppose that a value yi of the variable that is being assessed in the inventory was then measured on the 

i th selected sampling unit. Following the theory for sampling with variable probability of selection, the 

estimate from the sample data of the population mean (YM) is then (as adapted from Eqs. 3.7 and 3.9 

of Schreuder et al. 1993) 

YM = [Σi=1...n (yi/pi)]/N (2) 

and of its variance (VM) is 

VM = {1/(2N 2 )}Σi,j=1...n, i≠j{[(pipj−pij)/pij][yi/pi−yj/pj] 2 } (3) 

where pij (j=1…n) is the probability that both sampling units i and j were included in the sample, 

where 

pij = pipjN(n−1)/[n(N−1)] (4) 

If it is now assumed that the population size is much greater than the sample size, that is, N»n, then 

with some algebraic manipulation of Eqs. (1−4), it can be shown that YM and VM can be approximated 

very closely as 

and 

YM ≈ [(cx−cm)/nv] Σi=1...n[yi/(ci−cm)] (5) 

VM ≈ {[cx−cm] 2 /[2(n−1)nv 2 ]} Σi,j=1...n, i≠j{[yi/(ci−cm)]−[yj/(cj−cm)]} 2 (6) 

In practice for most forest inventories, it is perfectly reasonable to assume that N»n, that is, to assume 

that only a relatively small sample is taken from a large forest population. Eqs. (5) and (6) are then of 

particular interest, because neither involves the population size, N. This means that it is possible to 

apply 3P sampling without knowing the total population size, a matter overlooked by West (2005). 

However, it is clear from Eqs. (5) and (6) that the estimates of both the mean and its variance are 

functions of cx and cm, the maximum and minimum values of the covariate which were assumed to 

occur in the population. Thus, it will be necessary to consider carefully what values are to be used for 

those before setting out to take a 3P sample; that issue is considered below. 

The estimate of the confidence limit of the population mean (CM) may be determined as 

CM = t√VM 

(7) 

where t is Student’s t, for whatever probability level is desired and with (n−1) degrees of freedom. 

PRACTICAL EXAMPLES OF 3P SAMPLING 

The application of 3P sampling in practical forest inventory will be considered using data collected 

from two plantation forests. The first was a small (1 ha in area) planting of a mixture of rainforest 

species, established in 1997 on a farm a few kilometres from Alstonville (28°50’S, 153°26’W) in 

subtropical, north-eastern NSW. At the time of measurement the plantation was unthinned and had a 

stocking density of about 880 stems/ha. 

The second was a compartment (19 ha in area) of the 1996 Pinus radiata plantations of Forestry SA, 

established near Wandilo (37°43’S, 140°44’W) in south-eastern SA. The plantation had been thinned 

at 9 years of age to a residual stocking density of about 790 stems/ha. Both plantations were measured 

when they were 11 years old. 

A simple random sample of plots was selected from each plantation. Plots averaged about 0.01 ha in 

size in the mixed-rainforest plantation and about 0.02 ha in the P. radiata plantation. In each plot, the



diameters at breast height over bark of all the trees in the plot were measured. The data were used to 

determine both the stand basal area over bark and the amount of carbon sequestered in the stand total 

tree biomass (above- and below-ground) of each plot. For the mixed-rainforest plantation, tree 

biomasses were estimated using the individual tree biomass function, for plantations of this nature, 

developed by Specht and West (2003); this function predicts total oven-dry biomass of a tree (bT, 

tonne) as 

bT = 0.000364d 2.00 (8) 

where d (cm) is stem diameter at breast height over-bark (cm). For the P. radiata plantation, tree 

above-ground biomasses were estimated using the individual tree biomass function for P. radiata in 

Australia (Snowdon et al. 2000, Table 1.4); this function predicts above-ground oven-dry biomass of 

a tree (bA, tonne) as 

bA = 0.000118d 2.25 (9) 

Once this function had been used to estimate stand above-ground biomass (BA, tonne/ha), an 

additional function (Snowdon et al. 2000, Table 3.6) was then used to convert the above-ground 

biomass to a stand total oven-dry biomass (BT, tonne/ha) as 

BT = BA + 0.677BA 0.712 .(10) 

In both plantations, it was then assumed that 50% of oven-dry biomass was carbon (West 2009). A 

summary of the inventory data for both plantations is given in Table 1. 

These data were used as follows, to simulate results which might be obtained if 3P sampling was used 

instead of simple random sampling in these plantations. Firstly, the data sets were simply repeated 

many times to create, in effect, data collected from a very much larger simple random sample of the 

plantations. Stand basal area was used as the covariate variable, for selection of a 3P sample from this 

larger data set. Values were chosen for cx and cm, the maximum and minimum covariate values which 

were assumed to occur in the population. Then, a 3P sample of size n=20 was selected from these data 

using the procedures described above. If this was a real inventory, it might be envisaged that a quick 

and easy way to measure the covariate variable would be to take a point sample, as any point in the 

forest was being considered for inclusion in the sample. The number of plots which had to be 

considered (that is, ‘visited’), before this sample size was reached, was counted to give a value for nv. 

The estimates of the mean, variance and confidence limit were then determined from this 3P sample 

using Equations. (5-7). 

Table 2 shows an example of how this 3P sample selection was done, using the random sample data 

from the mixed-rainforest plantation. In this instance, values of cx=26.2 m 2 /ha and cm=2.2 m 2 /ha were 

assumed. As each plot was considered, a random value was selected within the range 2.2−26.2 m 2 /ha; 

the values selected are shown in the third column of the table. If the stand basal area of the plot 

exceeded the random value, the plot was included in the 3P sample; selected plots are indicated with a 

‘y’ in the fourth column of the table. In this case, nv=38 plots were ‘visited’ to obtain a sample size of 

n=20; it is data for only those 38 plots which are shown in the table. The stand total carbon of the 

selected plots is shown in the last column of the table. 

Table 1. Summary of inventory data from two plantation forest blocks, using simple random 

sampling. Shown are sample sizes and the minimum−mean−maximum of stand basal 

area and estimates of total amount of carbon sequestered in the above- and below- 

ground biomass of the trees. 

Mixed-rainforest Pinus radiata 

Sample size 20 24 

Stand basal area (m 2 /ha) 3−14−20 18−24−30 

Stand total carbon (tonne/ha) 7−33−47 35−45−59



Table 2. An example illustrating the selection of a 3P sample from a simple random sample 

from the mixed-rainforest plantation. In this case, it was assumed cx=26.2 m 2 /ha and 

cm=2.2 m 2 /ha. Plots included finally in the 3P sample are marked with a ‘y’ in the 

fourth column. 

Plot 

Stand basal 

area (m 2 /ha) 

Random 

value 

Included in 3P 

sample (y) 

Stand total carbon 

(tonne/ha) 

1 3.2 14.7 

2 18.0 19.2 

3 16.3 14.8 y 37.8 

4 13.3 7.1 y 30.9 

5 19.3 19.7 

6 3.2 11.5 

7 14.8 10.0 y 34.2 

8 18.6 10.2 y 43.0 

9 19.8 19.1 y 46.0 

10 10.1 4.8 y 23.4 

11 12.0 7.8 y 27.8 

12 12.4 22.3 

13 18.6 15.1 y 43.0 

14 9.9 18.2 

15 17.5 2.4 y 40.5 

16 12.4 19.1 

17 13.8 23.3 

18 14.6 23.8 

19 14.8 13.0 y 34.2 

20 3.2 16.9 

21 8.3 8.6 

22 13.3 10.6 y 30.9 

23 8.3 8.4 

24 10.1 12.6 

25 11.8 12.8 

26 15.6 2.8 y 36.0 

27 16.3 11.9 y 37.8 

28 15.6 17.1 

29 19.3 20.6 

30 18.0 3.1 y 41.8 

31 11.8 10.9 y 27.4 

32 10.1 7.2 y 23.4 

33 19.8 4.4 y 46.0 

34 20.1 13.4 y 46.7 

35 15.6 25.7 

36 10.1 25.2 

37 17.5 5.5 y 40.5 

38 18.6 12.9 y 43.0 

The 20 plots selected in this process were then used to determine the inventory results from the 3P 

sample, using Equations (5-7). The results are shown in the first column of values in Table 3. The 

corresponding results obtained from a simple random sample with n=20 are shown in the second 

column. Clearly, 3P sampling led to a much smaller 95% confidence limit than simple random 

sampling (0.7 tonne/ha of carbon as opposed to 2.7 tonne/ha). Similar results are shown for the P.



radiata plantation in the last two columns of Table 3. Again, 3P sampling was far more efficient than 

simple random sampling. 

For both the mixed-rainforest and P. radiata plantation data, similar results for 3P sampling were then 

obtained many times, but using different values of cx and cm each time to obtain different estimates of 

the means and variances. It follows from Equations (5) and (6) that these will vary, depending on the 

values used for cx and cm. The effect of this on estimates of the confidence limit from the 3P samples is 

shown in Figure 1. 

Table 3. Estimates from inventory of the mean stand carbon amount, and its 95% confidence 

limit, for two plantation forest blocks using either 3P sampling or simple random 

sampling. 

(a) 

95% Confidence limit 

3 

2 

1 

0 

Mixed rainforest Pinus radiata 

3P 

sample 

Simple 

random 

sample 

3P 

sample 

Simple 

random 

sample 

cx (m 2 /ha) 26.2 - 32.6 - 

cm (m 2 /ha) 2.2 - 0 - 

n 20 20 20 20 

nv 38 - 26 - 

Mean carbon (tonne/ha) 34.3 32.3 47.0 46.7 

95% confidence limit (tonne/ha) 0.7 2.7 0.2 3.2 

3.2 

2.2 

1.0 

10 30 50 70 

cx-cm 

(b) 

7 

6 

5 

4 

3 

2 

1 

0 

12.6 

9.0 

0 20 40 60 80 100 

Figure 1. Change in size of 95% confidence limit (tonne/ha) from 3P sampling as the range 

(cx−cm) changes for data from the (a) mixed-rainforest and (b) P. radiata plantations. 

The three solid lines ( ______ ) shown on each graph are for three different values of cm, 

which are marked adjacent to the line. The confidence limit from a simple random 

sample is shown also (- - -). In all cases, the sample size was n=20. 

95% Confidence limit 

cx-cm 

0



Firstly, these results suggest that there is generally a slight tendency for the estimate of the confidence 

limit from 3P sampling to increase as the range within which the covariate values are assumed to lie 

(cx−cm) increases. Far greater than this is the effect on the number of sampling units which must be 

visited to obtain the requisite sample size of n=20; the number visited increased approximately in 

direct proportion to the width of the range, rising to well over 100 in both plantation types, at the 

largest values of (cx−cm) used here (data not shown). That is to say, the narrower the range chosen for 

(cx−cm), the fewer sampling units will have to be visited to obtain the 3P sample. However, the 

narrower the range, the greater the likelihood of encountering a sampling unit of which the covariate 

value lies outside the chosen range, a circumstance which, as discussed above, invalidates the present 

3P sampling process. 

What is particularly evident in the results of Fig. 1 is that the value chosen for cm has a very large 

effect on the estimate of the confidence limit from 3P sampling. In these two example cases, the closer 

the value chosen for cm approached the actual minimum stand basal area in the data set (3.2 m 2 /ha in 

the mixed-rainforest plantation and 18.0 m 2 /ha in the P. radiata plantation), the larger was the estimate 

of the confidence limit. In the P. radiata plantation, once a value of cm greater than about 10 m 2 /ha 

was used, the estimate of the confidence limit from the 3P sample was greater than that from simple 

random sampling. 

It will require further study of the theory developed here to understand why the value chosen for cm 

has such large effects on the estimate of the confidence limit from the 3P sample. Ultimately, it will be 

desirable to develop some system for choosing cm and cx, through some preliminary survey of the 

population before 3P sampling starts, so that the 3P sampling leads to the most efficient estimate 

possible of the desired population characteristic. 

CONCLUSIONS 

The results here suggest that 3P sampling does have potential as a sampling method for forest 

inventory, when suitable covariate information from the entire forest population is not available. 

Recognition of 3P sampling as a case of sampling with varying probability of selection has allowed it 

to be applied without the need to obtain, in the field, covariate information for each and every 

sampling unit in the population. In earlier consideration of the theory of 3P sampling, this limitation 

rendered the technique practical for use only in very small forest populations. 

However, the efficiency of 3P sampling as applied here was highly dependent on the value assumed to 

be the minimum covariate value which occurred in the population. Further theoretical studies may 

show exactly how and why this is so and may offer ways to select the value to render the application 

of 3P sampling at its most efficient. Until such studies are undertaken, it may require preliminary 

simple random sampling of the population under consideration and use of simulation studies, similar 

to those done here, to determine empirically, for any population under consideration, the most 

appropriate minimum covariate value to use. 


I am indebted to Messrs John Gough Snr and Jnr for access to their mixed-rainforest plantation and to 

Forestry SA for access to their P. radiata plantation. The data used here were collected by students in 

the 2007 and 2008 ‘Measuring Trees and Forests’ undergraduate classes of Southern Cross University. 

REFERENCES 

Avery, TE and Burkhart, HE 2002, Forest measurements, 5th edn, McGraw-Hill, New York. 

Gifford, RM 2000, Carbon contents of above-ground tissues of forest and woodland trees, National Carbon 

Accounting System Technical Report No 22, Australian Greenhouse Office, Canberra. 

Iles, K 2003, A sampler of inventory topics, Kim Iles & Associates Ltd, Nanaimoa, British Columbia. 

Schreuder, HT, Gregoire, TG & Wood GB 1993, Sampling methods for multiresource forest inventor, Wiley, 

New York. 

Shiver, BD & Borders, BE 1996, Sampling techniques for forest resource inventory, Wiley, New York.



Snowdon, P, Eamus, D, Gibbons, P, Khanna, PK, Keith, H, Raison, RJ & Kirschbaum, MUF 2000, Synthesis of 

allometrics, review of root biomass and design of future woody biomass sampling strategies, National 

Carbon Accounting System Technical Report No 17, Australian Greenhouse Office, Canberra. 

Specht, A and West, PW 2003, ‘Estimation of biomass and sequestered carbon on farm forest plantations in 

northern New South Wales, Australia’, Biomass & Bioenergy, vol. 25, pp. 363–79. 

West, PW 2005, ‘An alternative approach to selecting a 3P sample’, Paper to a meeting, Western Forest 

Mensurationists, Hilo Hawaii, (Available at http://www.growthmodel.org/wmens/m2005/ west(a).pdf). 

West, PW 2009, Tree and forest measurement, 2nd edn, Springer, Berlin.



ESTIMATING FOREST CHARACTERISTICS IN 

YOUNG VICTORIAN ASH REGROWTH FORESTS 

USING FIELD PLOTS AND AIRBORNE LASER SCANNING DATA 

Andrew Haywood 1,2 and Michael Sutton 1 

ABSTRACT 

Airborne laser scanning data have the ability to measure the vertical and horizontal 

structure of forest vegetation. The aim of this study was to investigate how metrics 

derived from laser scanning data could be used in simple regression models for 

estimation of eucalypt top height, basal area and stems per hectare on 20m by 20m field 

plots. The study was located in the Central Forest Management Area in Victoria. The 

target population was ash dominated regrowth forests aged 20 to 60 years old. A linear 

regression function was able to provide precise estimates of eucalypt top height (R 2 = 

0.87; RMS error = 3.8 m) using a single height percentile variable. On the strength of 

these results it should be possible to predict top height with a precision that is close to 

traditional field measurement methods. Regression estimates of eucalypt basal area 

were less precise (R 2 = 0.56; RMS error = 14.69 m 2 ) than the top height model. The 

model included both a height percentile and intensity variable. Regression modelling 

was unable to provide a reasonable estimate (R 2 = 0.41) of eucalypt stems per hectare 

using height percentiles, intensity and canopy structure variables. This model should be 

used with caution. Basal area and height models were applied across the target regrowth 

stands and used to estimate future timber volume from these areas. 

INTRODUCTION 

There is a need for accurate and up-to-date information on resource information for sustainable forest 

management of Victorian native forests. The Statewide Forest Resource Inventory (SFRI) program has 

delivered comprehensive mapping and sawlog volume information for mature and older regrowth 

forests (DNRE, 2000). However, the stand mapping associated with this process is based on aerial 

photography now in excess of 15 years old. In addition, the SFRI field sampling focused on stands of 

1939 and older age classes. The younger regrowth stands were not sampled and are currently assumed 

to have the same productivity as the sampled 1939 (and older) stands. Following the Victorian fires of 

2003 and 2006/07 there is an increased need for resource information to aid sustainable forest 

management. In particular, improved resource information on the younger (un-sampled) regrowth 

stands is required. 

Previous research has shown that laser scanning data can be used to provide accurate estimates of 

forest structure variables that are currently measured using manual field collection and aerial photo 

interpretation methods (Naesset 1997, Hyyppä et al. 2000, Persson et al. 2002, Holmgren 2003). 

Forest estimates (of height, basal area and stems per hectare), using laser data, are often based on 

statistical measures derived from the distribution of laser point data. Magnussen and Boudewyn (1998) 

showed that for a given plot size and canopy structure, a certain percentile in the laser height 

distribution exists that corresponds to the canopy height of interest (i.e. top height). Other researchers 

have also noted that the inclusion of measures of canopy characteristics derived from the laser height 

distribution, in combination with selected laser height percentiles, have proven useful for estimating 

tree density (Naesset 2002), timber volume (Nelson et al. 1984; Means et al. 1999; Naesset & Økland 

2002) and identifying species (Donoghue et al., 2007). 

1 

Victorian Department of Sustainability and Environment, PO Box 500., East Melbourne, Victoria, 3002, Australia 

2 

Corresponding author: E-mail: Andrew.Haywood@dse.vic.gov.au



Much of this work has been focused on estimating forest attributes at a local scale. The development 

of new airborne laser scanning instruments allows forest information to be measured in three 

dimensions with precision over moderately large areas at low unit cost. The data derived from such 

laser scanning surveys are increasingly being used operationally over large areas, both internationally 

(Nilsson, 1996; Næsset, 2002; Watt, 2005; Holmgren & Wallerman, 2006) and domestically in 

Australia (Turner, 2007; Turner 2008). 

This project investigates the potential of airborne scanning and field data to provide improved younger 

ash regrowth resource information for the Central Forest Management Area (FMA) in Victoria, 


MATERIALS AND METHODS 

Study area 

The study area is located in the Central Forest Management Area (FMA) in Victoria (see Figure 1). 

The Central FMA covers some 290 000 hectares of public land located north of the Great Dividing 

Range, south of the Goulburn River and between the Mt. Disappointment forest near Broadford on the 

Hume Highway and the upper reaches of the Big River south of Lake Eildon. Of the forested land, one 

third is set aside for National or State Park and the remaining is designated State Forest. The area 

covers a variety of forest types, ranging from Ash species (mountain ash and alpine ash) in the higher 

elevations. As the elevation decreases, these forest structures are replaced by Messmate, and in drier 

areas, Peppermints, Boxes and Candlebark. 

Figure 1. The location of the Central Forest Management Area, Victoria Australia. The grid 

area represents the distribution of laser scanning data collected, the red circular points 

represents the field plot data collected. 

Field plot data 

The field data collected comprise forest plot measurements for young ash-regrowth forests in the 

Central FMA. A total of 88 square 0.04 hectare sample plots were measured. The sampling design was 

based on a stratified random sampling methodology. The stratification was based on existing SFRI 

data to include stands which regenerated between 1950 to 1989, species dominated by one or more of 

mountain ash (Eucalyptus regnans), alpine ash (Eucalyptus delegatensis) or shining gum (Eucalyptus 

nitens) and currently zoned for commercial timber harvesting. The location of each field plot was 

acquired using differential GPS. In each plot eucalypt top height, basal area and stems per hectare 

were measured. Table 1 summarises the plot data.



Table. 1. Summary of 88 field plot measurements. 

Minimum Maximum Mean Standard Deviation 

Top Height (m) 14.2 63.9 39.2 10.4 

Basal Area (m2/ha) 1.2 93.1 42.5 22.1 

Stems per hectare 25.6 1472.4 408.4 295.7 

Laser scanning data acquisition 

High-density laser scanning data were acquired over the study area with an Optech ALTM 3100 EA 

system during December 2007 and January 2008. The system settings and flight parameters are shown 

in Table 2. The vendor provided raw laser scanning data consisting of XYZ coordinates, off-nadir 

angle, and intensity for all laser returns within the area. In addition, the vendor provided a filtered 

ground data set consisting of points presumed to be measurements of the terrain surface, identified via 

a proprietary filtering algorithm. These filtered ground returns were used to generate a 1 m digital 

terrain model (DTM). The airborne laser scanning system collected up to four returns from each laser 

pulse, and all returns were used in this paper. 

Table 2. Laser scanning data specifications. 

Flying height 1 300 m Flight Direction NE/SW 

Swath width 945 m Laser pulse density 0.96 pulses/m 2 

Swath overlap 25% Laser footprint size 0.26 m 

Other optical data available included digital 1:40 000 scale aerial photography flown in January and 

March 2007. The aerial photography was used to assist in validation of the laser scanning processing. 

METHODS 

The analysis of the data was divided into a four-stage process, with the main objective being to 

evaluate their potential to provide improved resource information for the Central FMA. 

• Calculation of laser scanning plot-level variables, such as height percentiles and coefficient 

of variation of above ground pulse responses. Laser scanning data were also used to 

determine ground height within the plots. 

• Exploratory analysis of the two datasets (field measurements and laser scanning data) to 

investigate their underlying structure. 

• Bivariate regression methods to generate relationships at plot level between field 

measurements (of top height, basal area and stems per hectare) and height percentile laser 

estimates. 

• Multiple regression analysis to assess whether using more than one height percentile laser 

variable and/or intensity and canopy structure variables improved the relationship. 

Laser scanning data extraction 

The following variables were calculated from the laser scanning data and extracted over co-located 

field plots for quantitative analysis: laser height percentiles; mean intensity percentiles; standard 

deviation of laser dispersals; percentage of ground returns; coefficient of variation; skewness and 

kurtosis. 

Laser height percentiles provide information on the structure of the forest canopy at different canopy 

height levels. Using the laser scanning data, the pulses above 0.5 m were divided into quantiles 

corresponding to every 10th percentile from the 10th to the 100th, as well as the 5th, 95th and 99th 

percentiles. The 0.5 m cut-off was used as a threshold to account for undulations in terrain. Other 

researchers have used higher height thresholds (up to 3 m) to eliminate laser returns from the 

understorey vegetation layer (Naesset 2002; Riaño et al. 2004). Lower thresholds are appropriate in 

Victorian ash forests, because of the generally high numbers of trees regenerated which leads to early



forest canopy closure and limited development of the understorey. This provided 13 variables of an 

average laser canopy height by percentile. 

Like most discrete return systems, the Optech system records intensity for each pulse in the near infrared 

(1064 nm) region at a very narrow spectral width of 10 nm. The intensity of each return pulse, 

sometimes referred to as laser amplitude, represents the reflected energy from a highly culminated 

beam of light. It provides a concentrated measurement of the object's reflectivity unaffected by 

shadows or occlusions. This reflectance may vary based on the reflectance properties and porosity of 

the targeted material, path length and incidence angle of the pulse. Accordingly, for this study the data 

are regarded as uncalibrated, and are only used as a relative measure of intensity. The mean intensity 

for each of the 13 height percentiles was calculated for each plot. 

The percentage of vegetation returns provides a measure of canopy density, and is calculated by 

dividing the sum of all vegetation returns with height values above 0.5 m by the total number of 

returns. All returns above this threshold are considered to be canopy hits. Areas with low numbers of 

vegetation returns will be those with sparser, more open canopies. 

The standard deviation of laser dispersals provides a simple measurement of the variation or dispersal 

within the laser height distribution of each field measurement plot. 

The coefficient of variation (CV) summarises the relative variation, or dispersion, of the laser height 

distribution within each sample plot. It is the ratio of standard deviation and mean, and is expressed as 

a percentage. As a measure of crown density, higher CV values indicate sparse, open canopies and low 

CV values dense, closed canopies (e.g.



Multiple regression analysis was conducted to determine if further variation in the models could be 

explained by the inclusion of laser scanning-derived measures of intensity and canopy 

structure/density. 

RESULTS 

Exploratory data analysis 

Summary statistics for the field measurement data have been given in Table 1. The histograms in 

Figure 2 illustrate their distribution. 

Figure 2 shows the general distribution of the field measurement data. Top height and basal area are 

noticeable for their relatively normal distributions. Stems per hectare demonstrate a positive skew. A 

single plot with stems per hectare greater than 1400 trees/ha has a major influence on the distribution. 

Table 3 shows the correlations between field data variables and a selection of laser scanning derived 

variables. The data show that basal area is positively correlated with stems per hectare. Top height is 

negatively correlated with basal area due to biological growth. Top height is found to be negatively 

correlated with stems per hectare; this is basically due to mortality increasing as the plots get older 

(taller). 

The laser scanning data can be broken into three categories: height percentiles; intensity 

measurements; and canopy density/structure measurements. The height percentiles are generally 

positively correlated with each other. As expected the 50th height percentile is positively correlated 

with basal area and top height and negatively correlated with stems per hectare. The laser scanning 

canopy density/structure metrics have a stronger correlation with stems per hectare than the height 

percentiles. 

Table 3. Summary of correlations between field measured and LiDAR-derived variables. 

Stems per hectare 

(SPH) 

-0.30 SPH 

Basal Area (m2/ha) 0.54 0.20 BA 

Standard Deviation 0.90 -0.40 0.51 sd 

50 th Height percentile 0.73 -0.09 0.70 0.63 p50 

Mean intensity -0.50 -0.08 -0.17 -0.42 -0.31 p50i 

Coefficient Variation 0.26 -0.37 -0.14 0.52 -0.25 -0.25 CV 

Percentage of 

Ground Returns 

0.08 -0.20 -0.24 0.11 -0.19 -0.19 0.24 pcveg 

Skewness -0.14 -0.25 -0.42 0.06 -0.61 -0.20 0.73 0.46 skew 

Eucalypt top height prediction using laser height percentiles 

The percentile height with highest R 2 and lowest RMS error was selected as the predictor for 

estimation of top height. In this case, laser height values corresponding to the 95th percentile were 

used (Table 4). When the height percentile is greater than 70 % the differences in R 2 values are small, 

which implies that if separate models for each percentile (> 70th) were generated then each model 

would explain a similar amount of variance. However, if RMS error of each top height model is 

considered then differences between the percentiles become apparent (Table 4). 

Figure 3 illustrates there is a strong linear relationship (R 2 = 0.87, RMSE error = 3.89 m) between the 

95th height percentile and field-measured top height. With an R 2 of 0.87 using a single variable it is 

clear that a simple model that uses a single height percentile is the most effective approach to a 

predication of top height. A small number of plots had much higher laser percentile heights than field 

measured heights. After examination of the digital aerial photography this was explained by large over 

hanging canopies from trees outside the field measurement plots, and as a result these plots were not 

removed as outliers. Canopy density variables did not add any additional value to the predictive model 

in terms of explaining the remaining variation.



Table. 4. Summary of laser height percentiles used to estimate eucalypt top height. 

Height 

percentile 

Frequency 

Frequency 

Basal Area (m2/ha) 

0 5 10 15 20 

0 5 10 15 20 25 

R2 RMSE 

(m) 

Height 

percentile 

R2 RMSE 

(m) 

Height 

percentile 

R2 RMSE 

(m) 

1 0.0047 10.61 40 0.2794 9.06 90 0.8725 3.81 

5 0.0139 10.59 50 0.5464 7.19 95 0.8736 3.79 

10 0.0065 10.63 60 0.7989 4.78 99 0.8444 4.21 

20 0.0134 10.59 70 0.8488 4.15 Max height 0.8399 4.27 

30 0.0985 10.12 80 0.8667 3.90 

10 20 30 40 50 60 

Top height (m ) 

0 500 1000 1500 

Stocking (stems/ha) 

Frequency 

Frequency 

0 5 10 15 

0 5 10 15 20 

0 20 40 60 80 100 


1955 1965 1975 

Ageclass 

1985 

Top height (m) 

20 30 40 50 60 

20 30 40 50 60 

Lidar 95th Percentile Height (m) 

Figure 2. Histograms of field plot data. Figure 3. Eucalypt top height against laser 

scanning 95 th height percentile. 

80 

60 

40 

20 

0 

10 

20 

30 40 50 

50th Lidar Percentile Height(m) 


80 

60 

40 

20 

0 

20 30 40 50 

Mean intensity for 99th laser Percentile Height(m) 

Figure 4. Eucalypt basal area against 50 th laser height percentile and mean intensity from 95 th 

height percentile.



Eucalypt basal area prediction using laser scanning data 

Multiple regression showed that the 50th height percentile and the mean intensity value of the 95th 

percentile were important variables. The model containing these variables had an R 2 of 0.56 with an 

RMS error of 14.69 m 2 . Figure 4 shows the relationship between basal area and the two significant 

variables, both are positively correlated with basal area. The laser height percentile can be interpreted 

as representing stand development. The biological interpretation of the mean intensity value of the 

95 th height percentile was aided by plotting it against the proportion of eucalypt stems to total stems 

per hectare (Figure 5). It can be seen that the intensity variable is positively correlated with the 

proportion of eucalypts per hectare. 

Figure 6 shows that there are no strong outliers in the dataset causing undue influence on the 

regression. There are no major patterns or structure in the residuals, which indicates that the model 

predicts basal area reasonably well at both high and low basal area. 

Eucalypt stems per hectare prediction using laser scanning data 

Analysis of the field measurement data indicates that the stems per hectare (sph) range from 26 to 

1472 sph and that the distribution is skewed to the left. As a consequence a square root transformation 

was made to stabilise the residuals and variance. 

Multiple regression analysis was used to investigate whether height percentiles, intensity and canopy 

density measures were useful in predicting sqrt(stems per hectare). Beginning with an empty model, 

all variables were considered but a variable was only added if the p-value fell below 5%. (i.e. where 

inclusion of a variable was warranted because it significantly improved the model). In all four 

variables were added to the model including two laser height percentile (5th and 100th), one measure 

of intensity (60th height) and one canopy structural variable (skewness). The model containing these 

variables had an R 2 of 0.41 with an RMS error of 5.581 (sqrt(sph)). 

Given the strong relationships between the different height percentiles and top height observed in the 

scatterplots in Figure 3, collinearity between variables in the model may be a problem. The presence 

of collinearity was assessed using variance inflation factor. According to Rabe-Hesketh and Everitt 

(2000) multicollinearity exists if VIF values are larger than 10 and if the mean VIF is larger than the 

VIF for individual variables. This was not the case for the variables in the model. 

It can be seen in Figure 7 there a are no strong outliers in the dataset causing undue influence on the 

regression. However, the RMS is high (5.581 sqrt(trees/ha)) limiting its practical use for providing 

precise estimates of tree density. 

Mean Intensity for 95th Height Percentile 

50 

40 

30 

20 

0.2 0.4 0.6 0.8 1.0 

Proportion of Eucalypt stems/ha to total stems/ha 

Figure 5. Proportion of eucalypt stems total 

stems versus intensity of 95 th height 

percentile. 

35 

30 

sqrt(Stems per Hectare (sph/ha)) 

25 

20 

15 

10 

5 

10 15 20 25 30 

Predicted sqrt(Stems per Hectare (sph) 

Figure. 6. Predicted versus actual sqrt 

(eucalypt stems per hectare).




80 

60 

40 

20 

0 

20 30 40 50 60 70 80 

Predicted Basal Area (m2/ha) 

Figure 7. Predicted versus actual eucalypt basal area. 

DISCUSSION 

This project sought to evaluate the potential benefit of using field and airborne laser scanning data to 

improve forest resource estimates of young regrowth stands. It is anticipated that some of the methods 

described in this paper will be operationalised within Victoria and that laser scanning data will be used 

to provide stand-based estimates of eucalypt top height and basal area, which are key growth and yield 

model parameters. At this stage, for the forest plots studied, it appears that laser scanning data is able 

to provide reasonable estimates of eucalypt mean top height (R 2 =0.87), basal area (R 2 =0.56). The 

following discussion compares this study against results in the literature and discusses some 

operational considerations. 

Comparisons with related research 

Eucalypt top height 

Since laser scanning data are capable of providing a very accurate measure of distance, it is well suited 

to providing estimates of canopy height. Over forests, previous research has shown that for a given 

plot size and canopy structure, a certain percentile in the height distribution exists that corresponds to 

the canopy height of interest (Magnussen & Boudewyn 1998; Næsset & Økland 2002; Næsset & 

Bjerknes 2001). 

In this project a strong relationship (R 2 = 0.87; RMS error = 3.89 m) between eucalypt top height and 

laser derived height was observed above the 70th percentile; within this range little difference was 

observed between height percentiles. This is likely due to factors such as the density and homogeneity 

of the forest canopies studied (Yu et al., 2005). The results from this study suggest that the linear 

model is relatively insensitive to laser height distribution percentiles above 70%. Combined, these 

results agree with the best available science which shows that the accuracy of laser-derived height, 

after calibration to field height measurements, is comparable to that of manual field survey methods 

(Donoghue & Watt, 2006; Holmgren, 2003; Hyyppä et al., 2000; Næsset, 1997b; Persson et al., 2002; 

Watt, 2005; Lim et al., 2003; Lim & Treitz, 2004; Nelson et al., 2004; Popescu et al., 2003, 2004). 

It should be noted that in order to achieve a good level of accuracy the density of laser returns must be 

sufficient to (a) define the underlying terrain and (b) capture variations in terms of stand height and 

spatial arrangement. Generally a survey that records at least 4 returns/m 2 at a scan angle of



height measure can be interpreted as being a measure of the development phase of the plot which is 

related directly to changes in basal area. The intensity variable is most likely reflecting the eucalypt 

crown cover. The estimates of basal area observed in this study are similar to other relevant research. 

In European conifer-dominated forest Næsset (2002) reported R 2 values of 0.86 for basal area in 

southern Norway. Lim et al. (2003) reported basal area estimates of R 2 = 0.86 in a Canadian hardwood 

forest and Lucas et al. (2006) reported R 2 = 0.61 from the Australian forests. 

Eucalypt stems per hectare 

Eucalypt stems per hectare were not as reliably predicted using laser scanning measurements. A 

model, based on height percentiles, intensity and canopy structure variables provided the best model 

(R 2 of 0.41). However, the RMS error is relatively high (5.581 sqrt(sph)) limiting its practical use for 

providing accurate estimates of eucalypt tree density. 

This result is in contrast to other studies where the inclusion of measures of canopy characteristics 

derived from the laser height distribution, in combination with selected laser height percentiles, have 

proven useful for estimating tree density (Hudak et al. 2006; Næsset 2002; 2004a). One explanation is 

that in young regrowth often contain non-eucalypt species which influences the laser distribution at 

certain points in time. This makes it difficult to model eucalypt stems per hectare in young regrowth 

stands because while the distribution of the returns may change the eucalypt stems per hectare may 

not. 

Operational considerations 

It is likely that in the near future that laser scanning data will be used operationally to provide resource 

information to aid the sustainable management of Victoria’s commercial native forests. An appropriate 

approach would be to fly laser scanning instruments over GPS-located field plots measured in 

commercial forest areas. A number of findings identified as part of this study are considered 

applicable to an operational roll out: 

Estimate volume directly from the laser scanning data: It may be prudent to do this as the analysis 

shows that there are errors associated with each of the forest variables predicted. If these are used as 

inputs to another model then the errors may become additive. 

Refinement and validation of relationships: Previous research has indicated that if the physical 

structure of the forest changes then relationships may change. Therefore, it is expected that more than 

one estimation model (per variable) may be needed to adequately capture variation between 

silviculture regimes and forest areas. Validation of model predictions should also be frequently 

reviewed to ensure estimates are reliable. 

For future surveys, use laser scanning data to map the variation prior to establishing field plots: Since 

laser scanning data are sensitive to changes in forest structure and canopy cover, it would be possible 

to use these data to map forest variation. This information could be used to stratify the area prior to the 

forest inventory. From a modelling perspective this is important, and especially if estimates are to be 

extrapolated to areas where no field plots are available. 

Stems per hectare estimates: The methods tested in this study have shown that stems per hectare 

estimates are generally difficult to obtain from laser scanning data. An alternative approach would be 

to calculate stems per hectare by running an automated tree detection routine over the laser scanning 

dataset. Although single tree detection algorithm generally work best when detecting the number of 

stems on sparse stands, the error is likely to be high in dense eucalypt forests, limiting its practical use. 

It would probably be difficult to use this technique on very dense stands (greater than 1,000 sph) due 

to the coalescing of crowns. Increasing the laser scanning point density is one option to improve the 

accuracy of the detection, but this may not be economically viable. 

CONCLUSION 

This study has demonstrated that airborne scanning data provide an alternative operational approach to 

estimate, at the plot-level, resource information for Victorian ash forest. Eucalypt top height and basal 

area can be determined with acceptable accuracy.



REFERENCES 

DNRE, 2000. State-wide Forest Resource Inventory: Forest Stand Classification. Forest Services Technical 

Reports 00-1. Department of Natural Resources and Environment Victoria. 

Donoghue, D.N.M., Watt, P.J., Cox, N.J., Wilson, J., 2007. Remote sensing of species mixtures in conifer 

plantations using LiDAR height and intensity data. Remote Sensing of Environment 110, 509-522. 

Donoghue, D.N.M. and Watt, P.J., 2006. Using LiDAR to compare forest height estimates from IKONOS and 

Landsat ETM+ data in Sitka spruce plantations. International Journal of Remote Sensing. 27, 2161-217. 

Holmgren, J., Nilsson, M., Olsson, H., 2003. Estimation of tree height and stem volume on plots using airborne 

laser scanning. Forest Science 49, 419-428. 

Holmgren, J. and Wallerman, J., 2006. Estimation of tree size distribution by combining vertical and horizontal 

distribution of LIDAR measurements with extraction of individual trees. Proceedings from the EARSeL 

ISPRS Workshop, 3D Remote Sensing in Forestry, Vienna, 14-15 February 2006. 

Hyyppä, J., Hyyppä, H, Ruppert, G., 2000. Automatic Derivation of Features Related to Forest Stand Attributes 

Using Laser Scanner. IAPRS 2000 33, 421-428. 

Lim, K., Treitz, P., Wulder, M., St-Onge , B., Flood, M., 2003. LiDAR remote sensing of forest structure, 

Progress in Physical Geography 27 88–106. 

Lim, K., and P. Treitz, 2004. Estimation of aboveground forest biomass from airborne discrete return laser 

scanner data using canopy-based quantile estimators. Scandinavian Journal of Forest Research 19, 558- 

570. 

Lucas, R.M., Cronin, N., Lee, A., Moghaddam, M., Witte, C., and Tickle, P. 2006. Empirical relationships 

between AIRSAR backscatter and LiDAR-derived forest biomass, Queensland, Australia. Remote Sensing 

of Environment 100, 407–425. 

Næsset, E., 1997. Determination of mean tree height of forest stands using airborne laser scanner data. ISPRS 

Journal of Photogrammetry and Remote Sensing 52, 49-56. 

Næsset, E., 2002. Predicting forest stand characteristics with airborne scanning laser using a practical two-stage 

procedure and field data. Remote Sensing of Environment 80, 88-99. 

Næsset, E., 2004. Estimation of Above- and Below-Ground Biomass in Boreal Forest Ecosystems. In: 

Proceedings of the ISPRS working group VIII/2, Laser-Scanners for Forest and Landscape, Freiburg, 

Germany 03-06 October 2004, 145-148. 

Næsset, E. amd Bjerknes, K.O. 2001. Estimating tree heights and number of stems in young forest stands using 

airborne laser scanner data. - Remote Sensing of Environment 78, 328-340. 

Næsset, E. and Okland, T. 2002. Estimating tree height and tree crown properties using airborne scanning laser 

in a boreal nature reserve. Remote Sensing of Environment 79, 105–115. 

Nelson, R., Krabill, W., Maclean, G. 1984. Determining forest canopy characteristics using airborne laser data. 

Remote Sensing of Environment 15, 201-212. 

Nelson, R., Oderwald, R., Gregoire, T. G. 1997. Separating the ground and airborne laser sampling phases to 

estimate tropical forest basal area, volume, and biomass. Remote Sensing of Environment, 60, 311-326. 

Nilsson, M. 1996. Estimation of tree heights and stand volume using an airborne lidar system. Remote Sensing 

of Environment 56, 1-7. 

Persson, Å., Holmgren, J., Soderman, U., 2002. Detecting and measuring individual trees using an airborne laser 

scanner, Photogrammetric Engineering and Remote Sensing 68, 925-932. 

Popescu, S.C., Wynne, R.H., Nelson, R.F., .2002. Estimating plot-level tree heights with lidar: local filtering 

with a canopy-height based variable window size, Computers and Electronics in Agriculture 31, 71-95. 

Popescu, S.C., Wynne, R.H., Nelson, R.F., 2003. Measuring individual tree crown diameter with LiDAR and 

assessing its influence on estimating forest volume and biomass, Canadian Journal of Remote Sensing. 

Vol. 29(5): 564 - 577 

Riaño, D., Meier, E., Allgöwer, B., Chuvieco, E.,Ustin, S.L., 2003. Modelling airborne laser scanning data for 

the spatial generation of critical forest parameters in fire behaviour modelling. Remote Sensing 

Environment 86, 177-186. 

Stephens, P.R., Watt, P.J., Loubser, D., Kimberley, M.O. 2007. Estimation of carbon stocks in New Zealand 

planted forests using airborne scanning LiDAR. In proceedings of IAPRS Volume XXXVI, Part 3 / W52. 

Turner, R. 2007. An overview of Airborne LiDAR applications in New South Wales state forests. In proceedings 

of ANZIF 2007 Growing Forest Values conference, Coffs Harbour.



Turner, R. 2008 Applications for remote sensing tools in forestry operations. In proceedings of ForestTECH 

2008 Conference – Albury, NSW, Australia – April 2008. 

Watt, P.J., 2005. An evaluation of LiDAR and optical satellite data for the measurement of structural attributes 

in British upland conifer plantation forestry. Doctoral thesis. Department of Geography, University of 

Durham, England. 

Watt, P.J., and Haywood, A. 2006. Evaluation of airborne scanning LiDAR generated data as input into 

biomass/carbon models. Pöyry Forest Industry, contract report 38A08068 to Ministry for the Environment, 

Wellington, New Zealand. 

Yu, X., Hyyppä, J., Kaartinen, H., Maltamo, M., 2004. Automatic Detection of Harvested Trees and 

Determination of Forest Growth Using Airborne Laser Scanning. Remote Sensing of Environment 90, 451- 

-462.



OBJECT-BASED ANALYSIS OF FOREST STAND DELINEATION 

ON HIGH SPATIAL RESOLUTION IMAGERY 

USING OPEN SOURCE SOFTWARE 

ABSTRACT 

Andrew Haywood 1 and Christine Stone 2 

For decades, aerial photo interpretation has been, and to a good extent is still, the 

method of choice for producing fine-scale native forest stand mapping. Recent 

computer techniques have eased the task of the interpreter, who is now able to delineate 

polygons through on-screen digitising in a Geographical Information System (GIS) 

environment. Even with these advances, a great deal of skill is required in the polygon 

delineation. In an effort to contribute to the automation of this process, we introduce an 

open-source object based solution to the mapping of forest stand boundaries using 

attributes derived from digital aerial photography and laser scanning data (lidar) 

acquired over a study area in the Victorian Central Highlands. This methodology 

transforms remotely sensed imagery (single or multi-channel) and laser scanning 

derived canopy raster layers into polygon vector layers. It is intended that the resultant 

polygon layer should resemble the product derived by an aerial interpreter, without any 

prior knowledge of the scene. The derived product aims to produce a layer comprising 

of relatively homogeneous polygons all exceeding a minimum size. This layer can be 

used as a template by the interpreter. The interpreter can then aggregate (and sometimes 

correct) pre-delineated regions by simple drag-and-click operations. The derived 

product is only meant to be an intermediate aid for the work of the interpreter. The 

relationship between spectral, texture and laser scanning derived features for forest 

stand boundary delineation, and human interpreted boundaries is not straight forward. 

However, this approach is relatively cheap and flexible, being a workable compromise 

between (expensive) software automation and interpreter supervision for classifying 

native forest landscapes. Preliminary results are encouraging; both in regard to 

automating the process and the delivery of robust delineation of stand boundaries. 

Future research will focus on appropriate input resolution to reduce computation 

requirements and improved data fusion methods to obtain more accurate forest stand 

delineation. 

INTRODUCTION 

Approximately a third (8.3 million hectares) of Victoria’s land mass is covered by forest, of which 3.4 

million hectares is classified as State Forest and 3.7 million hectares as national parks and other 

reserves. Privately owned forest accounts for 1.2 million hectares, of largely native forest and 360 000 

hectares of plantations (predominantly Pinus radiata and Eucalyptus globulus) (DSE 2005). 

The State Forests are managed for wood production and the provision of non-wood production values 

including recreation, biological and landscape diversity. Since these forests provide many functions 

that are important to society, there is a necessity to monitor their sustainable management and to 

understand the causes of change. At the time of publication, the Victorian Department of 

Sustainability and Environment (DSE) was responsible for the sustainable management of public land 

in Victoria, including the public forest estate. As a consequence, DSE engages in a number of 

processes to monitor the sustainability of Victoria’s forests. These include Victoria’s State of the 

Forests Report (produced every five years), Sustainability Charter for Victoria’s State Forests (DSE 

2006) and associated Criteria and Indicators for Sustainable Forest Management in Victorian Forests 

(DSE 2007). These mechanisms are designed to enable Victoria to critically assess and evaluate 

progress towards achieving its sustainable forest management objectives and targets. The State of the 

1 

Victorian Department of Sustainability and Environment, PO Box 500, East Melbourne, Victoria, 3002, Australia. Email: 

Andrew.Haywood@dse.vic.gov.au. 

2 

Forest Science Centre, NSW Department of Industry and Investment, PO Box 100, Beecroft, NSW, 2119, Australia.



Forests Report seeks to include all forest land tenures in Victoria. However, a focus on Victoria’s State 

Forests is evident due to the limited data available on other land tenures. This is an issue that needs to 

be addressed. 

Forest stand delineation is one of the key building blocks for both forest management and monitoring 

and reporting. The process of delineating forest stands should allow the 1) production of forest maps 

precise enough to derive wood volume estimates, 2) characterisation of ecosystem structures, 3) 

production of forest structure maps useful for habitat and biodiversity evaluation. Within Victoria’s 

State Forests, stands are currently mapped using aerial photography interpretation (API) in 

combination with field surveys and are manually mapped in accordance with the State-wide forest 

resource inventory (SFRI) Forest Stand Classification (DNRE 2000). This approach acknowledges 

three primary stand components: eucalypt species, eucalypt structure and disturbance. Eucalypt 

structure comprises of crown cover, cover form, stand height and crown size. All of these attributes 

are visually assessed and assimilated into a single code for each forest stand by the photo interpreters. 

The resulting map classes are called forest stand classes. Manual delineation and interpretation of air 

photos is a consuming process and accuracies are dependant of the skills and experience of the photo 

interpreters. The majority of this mapping has occurred within State Forests and is increasingly 

becoming outdated. In addition, there are major mapping gaps across other land tenures. 

Internationally it has been recognised that current forest mapping procedures are falling short of forest 

agency expectations for three reasons: 1) the decreasing availability of trained API analysts; 2) the 

laborious, time consuming nature of manual feature identification and 3) the high labour costs 

involved (e.g. Leckie et al. 2003; Opitz and Blundell 2008; Wulder et al. 2008a and b). As a 

consequence, cost effective, semi-automated stand classification mapping techniques to supplement 

traditional methods need to be developed (e.g. Leckie et al. 1998; 2003; Hay et al. 2005; Wulder et al. 

2008a and b). 

Over the last two decades here has been considerable research into semi-automated forest mapping 

based on remotely sensed data, with varying success (e.g. Kilpeläinen and Tokala 1999; Almeida- 

Filho and Shimabukuro 2002; Flanders et al. 2003; Leckie et al. 2003; Tiede et al. 2004; Zhang et al. 

2004; Hay et al. 2005; Chubey et al. 2006; Antonarakis et al. 2008; Castilla et al. 2008; Pascual et al. 

2008; Wulder et al. 2008a and b). For the purposes of this study it is hypothesised that spectral, 

textural and laser scanning attributes can be utilised in a practical open-source solution to semiautomate 

forest stand delineation in Victoria’s natural forests. 

SPECTRAL INDICES IN FOREST STAND DELINEATION 

Spectral reflectance patterns of forest vegetation in the visible spectrum is generally controlled by the 

absorption features related to chlorophyll content. In contrast, the spectrum in the near infrared region 

is generally influenced by water content and the contribution of other organic materials. Different 

vegetation types have characteristic reflectance spectra. Simple image transformations have been 

useful in classifying imagery into discrete categories. For instance, band ratios have been used to 

detect and map different vegetation types extracted from remotely acquired spectral imagery (e.g. 

Gamon et al. 1995). The Normalised Difference Vegetation Index (NDVI) is a commonly used index 

that provides a measure of the amount and vigour of vegetation at the surface (e.g. Gamon et al. 1995). 

Unfortunately in complex eucalypt forests the water content and contribution of chlorophyll are 

spatially and temporally dynamic even within the same species and this can cause considerable 

variation in reflectance values (Goodwin et al. 2005). In addition, the reflectance of two different 

species can be very similar (Coops et al. 2004). As a consequence it is likely that spectral indices 

alone will not be sufficient to successfully delineate forest stands and that additional information such 

as texture indices and Laser scanning derived indices will be required. 

TEXTURE INDICES IN FOREST STAND DELINEATION 

The structurally complex and dynamic nature of native forests limits the ability of traditional statistical 

classification procedures to create stand level polygons based solely on the spectral response pattern at 

individual pixel locations. Spectral indices are a cumulative measure of all components within a pixel 

that influence the spectral values. In addition to chlorophyll and water content, factors such as



difference in crown closure, shadows, or stand density may result in markedly different stand 

structures, yet may still be represented by the same spectral value. The increased availability of higher 

spatial resolution sensors has provided imagery commonly having a pixel size smaller than the object 

of interest. This can result in visually discernable objects being populated with pixels having a high 

degree of spectral variation producing a ‘salt and pepper’ effect. Per-pixel image classifications are 

commonly applied at low or moderate spatial resolution but sometimes do not operate well at high 

spatial resolution. Classification accuracies, however, have increased through incorporating the ideas 

of image texture, shape and context. Texture, in the context of digital processing, is the spatial 

variability of image tones which describes the relationship between elements of surface cover (Wulder 

et al. 1998). The above-ground organisation of forest elements or forest structure is accordingly 

represented in texture. There are several approaches to texture processing. Commonly texture 

measures are derived from moving a fixed-size, odd-numbered window through the image and 

calculating a variety of different statistics such standard deviation (first-order) or second-order 

statistics derived from image spatial co-occurrence (Haralick et al. 1973) or spatial autocorrelation 

functions (e.g. semivariance). Numerous studies have added texture transformations with the 

spectrally derived indices and improved forest stand classification (e.g. Cohen et al. 1990; Gong et al. 

1992; Franklin and McDermid 1993; Ryherd and Woodcock 1996; Wulder et al. 1998; Franklin et al. 

2001; Coburn & Roberts 2004; Zhang et al. 2004). 

LASER SCANNING INDICES IN FOREST STAND DELINEATION 

Discrete-return lidar sensors operate by rapidly emitting a laser pulse toward a target (e.g. a forest 

stand) and recording the time, location and quantity of the reflected energy. Most attention has focused 

upon using lidar as a tool for characteristing vertical forest structure – primarily the estimation of tree 

and stand heights with volume and biomass also of interest (e.g. Rooker et al. 2006; Pascual et al. 

2008). A number studies have investigated fusing remotely sensed spectral imagery and laser scanning 

data and then statistically combining metrics derived from both sources of data for stand delineation 

(e.g. Hill and Thomson 2005; Hyde et al. 2006). There are two different sources of laser scanning 

information that can be used to stratify forest areas by canopy structure and/or composition.. Both can 

be applied at the tree or plot/stand level. The first, and most common, uses statitical measures based on 

the distribution of pulse laser returns from the forest canopy (Evans et al. 2006; Donoghue et al. 2007; 

Pascual et al. 2008). This assumes that the distribution of pulses will be different for forest class and / 

or density due to canopy structural characteristics such as canopy porosity, branching and form. The 

second alternative uses the near infrared intensity of the return from the forest canopy to help 

differentiate forest types. For discrete Laser scanning systems the intensity value (sometimes referred 

to as the amplitude) of each return pulse provides a measure of the amount of energy reflected from a 

target. Intensity values are influenced not only by sensor specifications and atmospheric conditions but 

also the reflectivity of the target in the near infrared range of wavelengths (Donoghue et al. 2007). It 

is assumed therefore that lidar intensity data contain information relating to forest type and condition 

(Kim et al. 2009). 

IMAGE SEGMENTATION IN FOREST STAND DELINEATION 

After spectral, texture and laser scanning pre-processing are conducted, further processing is required 

to achieve stand delineation, a task for which a range of different image segmentation methods can be 

employed (e.g. Wulder et al. 2008a). Image segmentation is the partitioning of a digital image into a 

set of jointly exhaustive and mutually disjointed regions than are more uniform within themselves than 

when compared to adjacent regions (Wulder et al. 2008a and b). In addition to the limitations already 

discussed with pixel-based methods applied to high-resolution imagery, classical edge-detection 

algorithms have had limited success, partly due to the fact that forest stand boundaries are often 

diffuse and difficult to define, even for an experienced aerial photo-interpreter (Leckie et al. 2003; 

Castilla et al. 2008; ). Attention recently, however, has been on object-orientated approaches, and in 

particular, automated segmentation that exploits the spatial information inherent in remotely sensed 

imagery in addition to the spectral information (Wulder et al. 2008a and b). Several of these methods 

have been applied to forest stand delineation, with the most common approach being based on region 

merging (e.g. Baatz and Schape 2000; Pekkarinen 2002; Hay et al. 2005; Castilla et al. 2008; Wulder 

et al. 2008a and b). This typically involves the sequential aggregation of adjacent regions according to



their similarity until a stop criterion is reached, which can be base on either a similarity or a size 

threshold. Once the digital scene has been segmented into ‘objects’ or polygons, clustering methods 

are then applied to classify the objects based on features (i.e. spectral and spatial properties) and 

meaningful information extracted from the objects (Chubey et al. 2006). 

Although several region-based image analysis techniques have been used successfully for forest 

information extraction purposes, adoption of these techniques by Australian forest agencies has been 

limited. This has been partly due the computational complexity of some of the procedures and the 

empirical nature of the customised requirements such as the trial and error basis for determining the 

optimal parameterization for the segmentation (e.g. Flanders et al. 2003; Hay et al. 2005). 

Nevertheless, with increasing availability of new commercial region-based image analysis software 

such as Definiens Developer 7 ® (Definiens® , Germany, formerly eCognition) and Feature Analyst® 

(Visual Learning Systems Inc., Missoula, Montana), coupled with diminishing availability of skilled 

photo interpreters there is new interest in region-based image analysis of remotely sensed data to 

support manual delineation and interpretation (Hay et al., 2005; Chubey et al. 2006; Wulder et al. 

2008a and b) We advocate that robust open-source protocols for implementing semi-automated image 

segmentation for individual forestry applications are a cost-effective alternative to the current 

commercial software packages. 

PRACTICAL AND OPEN-SOURCE APPROACH IN FOREST STAND DELINEATION 

Since it is likely that significant research remains until fully automated forest stand delineation can be 

achieved, the general approach should be as practical as possible. Consequently the interim goal rather 

than trying to replace photo interpreters, should be to support them in generating more timely, 

consistent and accurate products (Leckie et al., 1998; Castilla et al. 2008; Wulder et al. 2008a and b). 

Therefore, new and or better tools are required that produce incremental improvements in these areas. 

These tools need not provide final solutions or 100% correct results, they simply need to be tools that 

are useful and that can be easily corrected when things go awry. Specifically, they must be simple to 

apply, not require expensive equipment, not substantially alter the mapping workflow, nor involve 

inordinate fine-tuning by the interpreter (Leckie et al. 1998; Wulder 2008a and b). In order to facilitate 

these requirements, a semi-automated forest stand delineation methodology has been developed for 

application to moist native forests in south eastern Australia. 

Although there are a number of existing segmentation algorithms available for delineation of forests 

stands, they have all been developed overseas for non-eucalypt forests and significant effort is 

required to become familiar with these algorithms and associated software modules (e.g. Definiens 

Developer 7; Castilla et al. 2008) . As a consequence, an alternative approach that developes an 

integrated workflow process utilising open-sourced software was taken in this study. The 

Geographical Resources Analysis Support System platform (GRASS www.grass-itc.it; Neteler and 

Mitasoua 2004) was chosen due to its popularity within the open-source community and that it fully 

integrates with the open-source statistical software R (www.r-project.org; R Development Core Team 

2005), along with the python scripting language (van Rossum and Drake 2001). 

In this paper we present a methodology that combines spectral and texture attributes, Laser scanning 

derived Canopy Height Model layers and a simple region growing algorithm to delineate and 

characterise moist sclerophyll forest stands. The study area is described in the next section followed by 

a description of the three datasets used in the analysis. A detailed presentation of the algorithms 

utilised is made, followed by an examination of the results. The paper is concluded with a discussion 

of the method and its extension. 

METHODS 

Study area and datasets 

The study area is located 70 km north east of Melbourne, in Victorian Central Highlands, southeastern 

Australia. A high proportion of this mountainous area supports Wet Forests, dominated by 

Eucalyptus regnans (Mountain Ash). The area was selected to represent a variety of ash-forest types 

and forest conditions.



The area experiences a cool temperate climate, with mild summers and cool winters. Average annual 

rainfall exceeds 1200 mm over most of the area. Soils tend to be free draining, friable, brown 

gradational, have high water holding capacities, and have developed on a variety of volcanic parent 

rock materials (DNRE 1998). Fire is the major natural disturbance associated with the study area. 

Several fires have occurred within the study area over the past 150 years, the most extensive being in 

1926; 1939 (McHugh 1991; Jeremiah and Roob, 1992) and the recent extreme fire event of January 

2009. 

The study area is subject to intensive hardwood timber harvesting. Large-scale timber cutting, 

generally selective harvesting and sawmilling occurred in these forests in the latter part of the 

nineteenth and early twentieth centuries. Massive salvage operations followed major wildfires, 

particularly the extensive 1939 fires. Since the 1960s, clearfellling has been the major silvicultural 

system practised (Squire et al. 1991). The study area is 4 km by 4km, for a total area of 1 600 ha. 

Stand delineation from aerial photographic interpretation 

In Victoria, API formed the basis of a State wide Forest Resource Inventory project, whereby the 

derived polygons describing the forest attributes were considered based upon homogenous attributes 

such as eucalypt species, eucalypt structure and disturbance. An interpreter delineated the inventory 

polygons using visual interpretation of stereo images. The forest inventory polygons are assumed to be 

positionally accurate to ± 20 m, following State guidelines (Black 1996). In this study, we utilised 

stand boundaries interpreted from aerial photography acquired in the year 1999 at a scale of 1:20 000 

(DSE 2007). These boundaries have been updated to 2008 using logging and fire history records. 

Digital aerial photography 

Digital aerial photography data at a scale of 1:40 000 were acquired over the study area during January 

and March 2007 using a ZI Digital Mapping Camera (AAM Hatch Ltd). The Land Victoria 20 m 

contours plus photogrammetric breaklines were used to orthorectify the image. Colour balancing, 

mosaicing and resampling to a pixel size of 0.5 m followed the orthorectification process. The images 

were saved as multiband Geotiff files having visible red (R), green (G) and blue (B) channels. 

According to the image supplier the average location error of the orthorectified images is less than 1 

m. Only the central parts of the photographs were used in the mosaic to decrease the effects of 

bidirectional reflectance, exposure falloff, and relief displacement on the image analysis. This photo 

mosaic was used as the dataset for the spectral and texture analysis. 

Laser scanning data 

High-density laser scanning data were acquired over the study area with an Optech ALTM 3100 EA 

system (AAM Hatch Ltd) during December 2007 and January 2008. The system settings and flight 

parameters are shown in Table 1. The vendor provided raw laser scanning data consisting of XYZ 

coordinates, off-nadir angle, and intensity for all laser returns within the area in LAS file format. In 

addition, the vendor provided a filtered ground dataset consisting of points presumed to be 

measurements of the terrain surface, identified via a proprietary filtering algorithm. These filtered 

ground returns were used to generate a 1 m digital terrain model (DTM). 

Table 1: Laser scanning data specifications. 

Flying height 1 300 m Flight Direction NE/SW 

Swath width 945 m Laser pulse density 0.96 pulses/m 2 

Swath overlap 25% Laser footprint size 0.26 m 

Workflow structure 

The workflow presented here was set up on a LINUX platform using GRASS GIS as the environment 

for the implementation of new modules. GRASS GIS is designed for both interactive use with a 

Graphical User Interface (GUI) and command line use. The built in commands can be batched 

together with LINUX programs using shell scripting as an interface, which gives broad possibilities on 

adaptations of the existing GIS functionalities (Neteler and Mitsova 2004). Furthermore new modules 

are added as python scripts into the workflow (Python Software Foundation 2006).



The workflow included the following steps: Data processing, Imagery segmentation, Feature 

extraction, Unsupervised region classification, Merging and mapping of unsupervised classes to 

general forest classifications and Comparison between SFRI mapping and semi-automated delineation 

Processing the Laser scanning data; DCHM and colour aerial photography 

The elevations of the laser canning measurement were converted to a digital canopy height model 

(DCHM) by subtracting off the elevation of the underlying terrain. The canopy height model was 

calculated at a resolution of 0.5 m and then re-sampled to 1.0, 2.0 and 5.0 m resolutions. 

Unfortunately the visible colour RGB aerial photography used in this study does not allow the 

calculation of the widely used Normalised Difference vegetation Index (NDVI) as reflectance in the 

NIR spectral region was not captured. Instead, a vegetation index (VI) (Equation 1) based on the green 

(G) and red (R) bands was used. The ratio of these two bands has been successfully used in Canada to 

map and monitor forest disturbance caused through stand mortality from mountain pine beetle attack 

(Wulder et al. (2009). 

Vegetation Index = R/G Equation 1 

The image attributes were obtained using a square shaped windows (size 20 by 20 m). Holopainen and 

Wang (1998) have stated this size appears generally to be the near-optimal window size for extracting 

aerial photograph attributes for forest classification. The image attributes were extracted using the 

original 0.5 m resolutions as well as images re-sampled to 1.0, 2.0 and 5.0 m resolutions. 

The extracted neighbourhood attributes derived using the R/G VI were average, standard deviation, 

number of unique pixel values and range of pixel values. Additionally, four texture features based on 

the image gray-level co-occurrence matrices (Haralick et al., 1973; Haralick 1979) were computed: 

angular second moment (ASM), contrast (CON), correlation (COR) and entropy (ENT). 

These texture features were calculated from the pixel window using horizontal (0), vertical and 

diagonal (45 and 135) directions resulting in four values per feature. Prior to conducting the texture 

analysis the band data were requantified (Equations 2, 3, 4) to a 6 bit image (64 levels). This process 

does not change the actual grey levels distributions but greatly reduces the computation time. 

The two ‘best” neighbourhood and texture features were selected for input into the segmentation 

algorithm. In this study ‘best’ was defined through a iterative process of re-running all pair 

combinations to visually identify the best segmentation results. 

Imagery Segmentation 

Within GRASS there are a number of segmentation modules, including pixel-based (maximum 

likelihood classifier- i.maxlik and the sequential maximum a posteriori algorithim-i.smap), edgedetection 

(v.lidar.edge detection) and region growing modules (v.lidar.growing). 

Studies presented in the literature and personal experience suggest that region-based segmentation 

methods are generally more appropriate for semi-automatic stand delineation. Unfortunately, the 

current region growing modules implemented in GRASS are tailored for pure laser scanning object 

identification and do not lend themselves to the inclusion of spectral and textural information. As a 

consequence part of this study was to develop a more generic region growing module for GRASS. 

Traditional region growing methodology can be viewed as an iterative process by which regions are 

merged starting from individual pixels (or any initial segmentation), and growing iteratively until 

every pixel is processed. In broad terms, it can be described by the following steps: 

1. Segment the entire image into pattern cells (1 or more pixels); 

2. each pattern cell is compared with its neighbouring cells to determine if they are similar, using 

a similarity measure. If they are similar, merge the cells to form a fragment and update the 

property used in the comparison; 

3. continue growing the fragment by examining all of its neighbours until no joinable regions 

remain. Label the fragment as a completed region; and 

4. move to the next uncompleted cell, and repeat these steps until all cells are labelled.



The drawback of this traditional scheme is that, at each iteration, one or several merges occur so that 

the resulting segmentation is dependent on the order of the merges (lack of global convergence). In 

order to solve this problem, a number of different approaches have been suggested; including iterative 

parallel (Tilton and Cox, 1983), Hierarchical Step-Wise (Beaulieu and Goldberg, 1989) and size 

constrained region merging (Castilla and Hay et al. 2008). The segmentation algorithm proposed here 

is simpler than these approaches but is has been providing satisfactory results in the segmentation of 

some forest and agricultural images. 

New region growing segmentation algorithm 

The approach implemented was based on the traditional region growing technique, with some 

modifications which partially solve the problem of the dependence on the order of the merges. The 

algorithm can be applied to any number of layers but in this study was limited to the DCHM and the 

two best texture and neighbourhood metrics 

Feature extraction 

The aim of the feature extraction phase is to determine the statistical attributes of each candidate forest 

stand outlined by the polygons resulting from the segmentation process. The result of the feature 

extraction phase is a file that contains a list of candidate stands sorted in descending order according to 

their area. For each candidate stand on the list, the file contains several statistics from the images 

selected by the user. These statistical attributes are then used to conduct a supervised or unsupervised 

classification; in both cases, to determine the similarity measure between two candidate stands 

outlined by polygons. 

Unsupervised classification 

In this study an unsupervised classification based on a clustering algorithm was applied to the list of 

candidate stands, characterized by their statistical attributes, defined in the feature extraction file. The 

clustering algorithm uses the covariance matrix and mean vector of the candidate stands to estimate 

the centres of the classes. 

Merging of classes 

This is an interactive phase where all resulting classes were reclassified into broad forest classes. The 

user defines the reclassification rules by comparing the aerial photography and the canopy height 

model with the classification results. 

Comparison of automated segmentation versus aerial photographic interpretation delineation 

Originally, two comparison assessment techniques were envisaged. The first accuracy assessment test 

was to use existing irregularly spaced Statewide Resource Forest Inventory plots. However, this 

dataset was found to be unsuitable. There were too few points and too large a gap between points, 

relative to the mapping scale and the resulting number of polygons delineated. The second accuracy 

assessment was a qualitative comparison with the manually mapped SFRI classes. A qualified photo 

interpreter assessed the semi-automated delineation and where possible attributed to the automatically 

delineated polygons. The comparison of the two delineations was largely qualitative, as per Leckie et 

al (2003) and Wulder et al (2008). It must be recognised that there are a number of limitations 

associated with this comparison approach: 

1. The original SFRI delineation does not necessarily represent the “truth”. This process is 

afflicted with some unknowns (subjectivity, abstraction, and different semantics). 

2. The two products are based on different imagery sources in terms of date of acquisition and 

resolution– the SFRI mapping based on benchmarked 1980’s imagery versus the semiautomated 

delineation based on 2007 imagery. 

3. Unfortunately, there is no universally accepted procedure for comparing thematic maps. In 

addition, there is increasing evidence that traditional accuracy assessment tools based on 

cross-tabulation spreadsheets referring to points or pixels may not be suitable for object-based 

classification (de Kok, 2001; Blaschke 2003, Radoux and Defourney, 2007). Because objects 

include information about shape, area, area-shape relationships, topology etc., this additional 

information may have to be evaluated in another manner.



RESULTS 

The automated polygon segmentation was generated using the following three layers: 

1. Digital Canopy Height Model (DCHM) 

2. Average Vi with a 5m × 5m moving window (VI average) 

3. Contrast Vi with a 5m × 5m moving window (VI contrast) 

The results of the semi-automated segmentation versus manual delineation applied to the study area 

are shown in Figure 1. Using a simple visual examination, several differences in the polygons are 

obvious. Specifically, the semi-automated segmented polygons are more detailed than those resulting 

from the manual delineation. While human interpretation intrinsically includes generalisation – which 

is often important and necessary for many applications – automated image segmentation can also 

result in very complex outlines following shapes of small stands of trees. These differences are 

reflected in the mean perimeter of the polygons generated from the two processes: the semi-automated 

output had a mean polygon perimeter that was approximately 100% greater than the perimeter of the 

SFRI-manually delineated polygons, while the mean polygon size for the semi-automated polygons is 

only 30% greater than the manual SFRI polygons (Table 2). The manual SFRI polygons have a much 

greater range in polygon sizes, with a standard deviation that is more than double the automated 

delineated polygons. 

Figure 1: Screenshots of the resulting maps benchmarked manual delineation (left) and 

semi-automated polygon delineation (right). 

In addition, the benchmarked SFRI manually interpreted polygons were provided using ortho aerial 

images and corporate fire and logging history GIS layers. The result is a combination of visual 

interpretation (proofed by field surveys) updated with fire and logging history post photo date. 

However, Figure 1 shows the benchmarked SFRI data contain errors. 

The red arrows mark two points, where the line management is not comprehensible. These errors are a 

result of the logging history data not being up to date. 

A qualified photo interpreter assessed the semi-automated delineation and attributed each polygon 

with a simplified SFRI code. The photo interpreter noted several attributes of the automated delineated 

polygons: 

1. As previously mentioned, the boundaries delineated by the automated process are not as 

straight as those produced manually by a human interpreter; 

2. Approximately 60% of the area contained sensible polygons that were easily attributed; 

3. The remaining 40% polygons would require minor modifications to be meaningful, or were 

not appropriate for reasons discussed below.



Table 2: Properties of delineated polygons. 

Attribute Manual Semi-automated 

Total # of polygons 261 189 

Mean polygon size (ha) 6.13 8.46 

Minimum polygon size (ha) 0.01 1.51 

Maximum polygon size (ha) 50.48 112.15 

Standard deviation polygon size (ha) 7.95 14.35 

Mean polygon perimeter (m) 1291.75 2458.87 

Minimum polygon perimeter (m) 9.09 701.45 

Maximum polygon perimeter (m) 9021.90 25225.73 

Standard deviation polygon perimeter (m) 1268.29 2829.54 

An example of the first issue of more ‘precise’ delineations compared to manual mapping is shown in 

Figure 2. This is caused by the fact that the automatic delineation depends on the boundaries of the 

individual tree crowns. This means the older the trees within a forest stand the more frayed 

(fragmented) the boundary will be. 

Figure 2:Geometric differences for object delineation of forest species classes (left: manually 

mapped by human interpreter; right: semi-automatically constructed through image 

segmentation). 

Examples of the second situation, where the semi-automated delineation produced polygons that were 

sensible and easy to type, are provided in Figure 3. 

Figure 3: Examples of where semi-automated delineation performed well. 

These polygons represent areas with homogenous texture and tone, which have relatively uniform 

species composition, age, and height –demonstrating the ability of semi-automated segmentation 

routines to perform well in uniform stand conditions.



In contrast, Figure6 illustrates situations where the semi-automated delineation did not perform well. 

In Figure 4 (left), some mature vegetation has been included within a larger re-generation polygon. In 

Figure 4 (right), the highlighted polygon seems no different than the area to the left. The main reason 

for these errors is how the algorithm deals with the weighting between the three input layers. In both 

cases shown below the digital canopy height model layer is driving the incorrect delineation. This 

coupled with the minimum mapping unit size tends to be the main issues. 

Figure 4: Examples of where semi-automated delineation performs poorly. 

The quantitative comparisons for the five major species groups identified within the study area results 

in an overall accuracy of 65% (Kappa Index: 44% see Table 3). A closer look at the classes shows 

partly very good results (for the mixed Eucalypt classes), but very bad results (user accuracy for the 

‘unknown’ class). One reason is the inexact and/or subjective reference dataset; another reason 

originates in the resemblances of the classes. Confusion occurred most between the classes ReM - 

DeM and ReP - DeP, while the most stable subdivision occurred between mixed (Mixed E. regnans 

and E. delegatensis) and the pure stands. If only the four stable classes (ReM, ReP, DeM, DeP) are 

taken into account, overall accuracy rises up to 80% (Kappa index: 62%). 

Table 3:Accuracy assessment for the five major forest species classes in the study area. 

Development phases 

Mixed 

E. regnans 

(ReM) 

Pure 

E. regnans 

(ReP) 

Mixed 

E.delegatensis 

(DeM) 

Pure 

E.delegatensis 

(DeP) 

Unknown 

User Accuracy* in % 80 42 69 54 12 

Producer Accuracy in % 54 48 89 49 30 

Overall Accuracy: 65% Kappa Index: 44% 

However, it should be noted that the reference benchmarked SFRI data are subjective (and in some 

places obviously incorrect). The SFRI class descriptions can be fuzzy and it is sometimes difficult to 

distinguish forest classes in the field, if they are proximate in time. 

CONCLUSION 

It is generally accepted that manual aerial photo interpretation provides better classification accuracy 

than automated delineation, because the interpreter can use additional clues, such as area, shape and 

contextual information, including topographic information, results of previous classification, etc. In 

this paper, a region-growing segmentation technique, in which the algorithm uses some contextual 

attributes besides spectral values of the pixels, is an attempt to improve digital classification. Even this 

alternative can result in an inadequate classification, so the procedure proposed here, similar to other 

techniques, will require editing after the digital analysis steps to eliminate misclassified set of pixels 

(polygons).



Overall, the segmentation accuracies are moderate from the semi-automated approach but it must be 

recognised that often these forest classes are spectrally similar and the forest texture and Laser 

scanning height data vary within each class. Notwithstanding, the results generally encourage further 

investments in data acquisition and methodological development. However, some critical points 

remain. In particular the following tasks need to be solved for supporting or partially replacing 

traditional aerial photography interpretation techniques: 

• Quantify to what extent the results simplify API using aerial photography combined with 

stand ground surveys (time, exactness, costs) 

• Supporting the API work with interim analysis results (e.g. stand height maps, differentiation 

vegetation, gap distribution etc) to get more objective and faster results in case of unclear 

classification. 

• Integrating statistical results in the classification which allow statements about how certain a 

forest vegetation classification is, or how uncertain the distinction from a particular other 

class is. 

For the foreseeable future, aerial photo interpretation coupled with ground based surveys will be 

necessary (in particular in the production forest) especially for planning purposes. But the proposed 

method in this paper is believed to potentially reduce the efforts required. API can then focus on 

uncertain situations and on the most important forest stands concerning forest planning and ecological 

questions. 


This study was supported by the Victorian Government through the Resource Outlook project. We 

would like to thank Lee Miezis, Mike Sutton, Fiona Hamilton, Fred Cummings, Kristen Thrum, 

Andrew Mellor, Kristen Thrum and Kate Nolan for their support (Department of Sustainability and 

Environment). 

REFERENCES 

(Due to space limitations three pages of references are available upon request from the 

main author)



THE GLOBAL FOREST RESOURCE ASSESSMENT 

REMOTE SENSING SURVEY 

Adam Gerrand 1 , Erik Lindquist 1 , Mette Wilkie 1 and Rod Keenan 2 

ABSTRACT 

Global concern is growing over deforestation because of its climate impacts as well as 

loss of biodiversity and other forest services. FAO has been reporting on the world's 

forests at 5 to 10 year intervals since 1946 through the Global Forest Resources 

Assessments (FRA). As part of FRA 2010, FAO, its member countries and partner 

organizations will undertake a new global remote sensing survey of forests between 2008 

and 2011. The survey will substantially improve knowledge on land use change including 

deforestation, reforestation and natural expansion of forests. The assessment will sample 

the whole land surface of the Earth with over 13 000 Landsat Global Land Survey image 

sections covering 10km by 10km at each of the intersections of the latitude and longitude 

degree lines. 

The main outcomes will be improved information at the global and ecozone level on 

changes in forest cover and land use including trends in the rate of deforestation, 

afforestation and natural expansion of forests from 1990 to 2005. The work also has 

potential as a global monitoring framework for a range of other purposes. A significant 

effort will also be made to provide training to build on the capacity of many countries to 

improve their forest monitoring and reporting. The paper will describe the FRA2010 

Remote Sensing Survey and encourage involvement of others who can provide useful 

additional data or analyses for validation. 

INTRODUCTION 

The world’s forests provide vital economic, social and environmental benefits. They mitigate climate 

change by storing carbon, provide wood and non-wood forest products, support human livelihoods, supply 

clean water and provide habitat for half the species on the planet (Wilson, 1988). 

The Food and Agriculture Organization (FAO) of the United Nations provides detailed information on 

global forest cover and forest land use at 5 to 10 year intervals from 1946 to 2005 in a series of reports 

called the Global Forest Resources Assessments (FRA). Each is based on data that countries provide 

to FAO in response to a questionnaire. FAO compiles and analyses the information and presents the 

current status of the world’s forest resources and their changes over time. Historically, FRA reporting 

has evolved to reflect the major issues of concern at the time. Early reports focused on timber stocks in 

response to post-war (WWII) needs for building materials while more recent emphasis has shifted to 

deforestation and conservation issues. The breadth and quality has also improved as individual 

countries gain reporting capacity and information availability increases. The most recent FRA report, 

in 2005, was the most comprehensive in scope ever and aimed at assessing progress towards 

sustainable forest management (FAO, 2006a). 

The need for reliable information on forests 

During the 2008 G-8 Summit, world leaders “encouraged actions for Reducing Emissions from 

Deforestation and Forest Degradation in Developing Countries (REDD) including the development of 

an international forest monitoring network building on existing initiatives” (G8 Summit 2008). 

As part of the FRA 2010, FAO, its member countries and partners are undertaking a systematic remote 

sensing survey which will form the basis for a consistent long-term global forest monitoring system and 

1 UN Food and Agriculture Organisation, FAO Forestry Department, Viale delle Terme di Caracalla, Rome 00100, Italy 

Rome. Tel: +3906 5705 3063, Fax: +3906 5705 5137. adam.gerrand@fao.org 

2 School of Forest and Ecosystem Science, University of Melbourne, Water Street Creswick, Victoria Australia 3363.



help meet a number of international forest information needs, including the G-8 recommendations (FAO, 

2008). Forest loss has effects on local, regional, and global scales on a wide range of factors including 

impacts on climate, biodiversity and ecosystem services (Lambin and Geist, 2006). With respect to climate 

alone, the latest estimates state that “forestry” (including deforestation, forest biomass decomposition after 

logging, and release of CO2 from peat soils) is responsible for about 17 % of human-produced greenhouse 

gas emissions (IPCC, 2007). The Clean Development Mechanism under the UNFCCC, and other 

initiatives such as Reducing Emissions from Deforestation and Degradation (REDD) are being developed 

to help reduce the negative effects of forest loss on climate but issues of accounting for and monitoring 

forest resources remain difficult to resolve. 

According to the latest FRA assessment, approximately 13 million hectares of forest cover is cleared annually 

worldwide (FAO, 2006a). Few countries can adequately report on changes in forest area over time so there is 

a need to improve the quality of reporting and harmonisation of definitions of forests (FAO, 2006, Hansen et 

al., 2008b). FAO is working to improve the results through strengthening national capacity as part of FRA 

2010 and improved reporting guidelines (FAO, 2007b). 

Satellite remote sensing offers the advantage of broad area coverage, repeatable and systematic 

observations of the Earth’s surface. Though remote sensing does not replace the need for field-collected 

data, it offers distinct benefits when conducting large-area surveys for broad vegetation-type categories. 

One major benefit is the ability to provide maps rather than tabular numerical results summarised by 

country or region that better illustrate where changes in forest cover are occurring. Forest loss is variable 

across forest types and geographical locations, a fact increasingly well documented in regional remote 

sensing studies from the tropics to the boreal (Potapov et al., 2008; Hansen et al., 2008b, Achard et al., 

2002). 

The FRA 2010 Remote Sensing Survey 

The FRA 2010 Remote Sensing Survey (RSS) brings together the comprehensive Landsat Global 

Land Survey satellite imagery database (Tucker et al., 2004) and is a partnership of many of the 

world’s leading land cover remote sensing scientists to analyse satellite data and engage with country 

experts in over 150 countries (see Acknowledgements). 

The goals of the RSS are to obtain systematic information on the distribution and changes in forest cover 

and forest land use from 1990 to 2000 to 2005 at regional, ecozone and global levels. The RSS also 

provides a consistent framework upon which future global assessments and more detailed regional 

assessments can be based. As presented in this paper, the RSS is a work in progress; the methods, 

discussions and conclusions are preliminary and may be amended as the study develops. Notably, the RSS 

is intended to provide information complementary to and in some cases strengthen existing national 

reporting systems - but not replace them. 

METHODS 

Sampling framework 

The FRA 2010 RSS global sampling grid consists of 13 689 sites and covers the globe between 75 

degrees North and South in latitude. A systematic sampling design based on each longitude and 

latitude intersection has been implemented, with a reduced intensity above 60 degrees North/South 

latitude due to the curvature of the Earth (every second intersection sampled in between 60 and 75 

degrees North/South). 

Each sample tile will cover a 10 by 10 kilometre box at every one-degree latitude and longitude 

junction (approximately 100km apart). This grid of sample plots is the same basic layout but a lower 

intensity (wider spacing) than the national forest assessments supported by FAO (FAO, 2008). 

Antarctica was excluded from the land mask. Sample locations within deserts and areas with 

permanent ice are also excluded from analysis leaving a total number of 9 078 tiles (Figure 1).



Figure 1: The 13 689 sample locations of the FRA RSS (light black dots) and the 9 078 sample sites 

where at least a portion of the 10x10 km sample tile has tree cover > 10% as derived from 

the circa 2000 MODIS VCF (dark black dots). 

Tasmania 10 sample tiles 

across different forest 

types 

Landsat samples available on-line: Draft viewing, analysis and labelling software: 

Figure 2: The FRA RSS sampling grid with 700 samples across Australia and the 10 samples in 

Tasmania with one sample tile shown in detail with Landsat images from the FRA website 

(lower left) and draft analysis (lower right).



Remotely sensed inputs 

The United States Geological Survey’s Landsat Global Land Survey dataset (GLS) provided the 

imagery data for interpretation and classification. The GLS is a co-registered, multi-date dataset 

composed of the single best Landsat image acquisition covering most of the Earth’s land surface and 

centered on the years 1975, 1990, 2000, and 2005 (Gutman et al., 2008; Tucker et. al., 2004). The 

FRA 2010 RSS will focus initially on the GLS1990, GLS2000 and GLS2005. If time permits and 

techniques are developed to handle earlier Landsat MSS data, then we may include data from 1975 to 

make a 30 year time series. 

For each survey tile, Landsat bands 1-5 and 7 (also 8 in the case of ETM+) of the GLS acquisitions were 

compiled. These were clipped to a 20km by 20km box centered on each one-degree latitude and 

longitude intersection to create imagery ‘chips’. This produced 56 219 individual imagery chips for the 

three time periods. The central 10km by 10km box of each sampling tile will be used for area 

calculations and statistical analysis. This was adjusted to preserve an integer number of Landsat data 

pixels (30m spatial resolution). For 10km by 10km blocks, 334 by 334 pixels tiles were created for a tile 

extent of 10 020 by 10 020m. 

The images will be processed through a series of steps described in more detail in Gerrand et al. (2009); 

Achard et al (2009) and Bodart et al (2009). The aim is to develop polygons of a similar spectral signal 

and label these with a simple legend with 6 main land cover classes (Table 1). 

Labeling objects (polygons) in each segmented survey tile is a two-step process. The first step is a 

draft pre-labeling exercise followed by label correction and validation by in-country experts. Polygons 

are pre-labeled in one of two ways: (i) automatically using spectral training signatures and (ii) by 

visual interpretation. Automated labeling by the Joint Research Centre (JRC) will be applied to survey 

tiles in the humid tropics and Russian boreal forests (Bodart et al, 2009). For all other regions, FAO 

specialists will use visual interpretation and a FAO-developed software program called GEOVIS 

Mapping Device – Change Analysis Tool (MADCAT) to label polygons. 

The FAO-developed Land Cover Classification System (LCCS) (FAO, 2005) has been adapted for 

labeling polygons by land cover. The legend has been simplified to include only 5 land cover classes 

(Table 1). Survey tiles processed by the JRC use a slightly different initial legend and will be re-coded to 

fit the FRA cover classes. A simple system of 9 Land-use codes has also been developed for use in the 

RSS based on FRA definitions (FAO, 2007b). 

Table 1: Land cover and land-use classes to be used in the FRA 2010 RSS. 

Land Cover Class Land-Use Class 

Tree Cover Forest 

Other wooded land 

Other land with tree cover 

Shrub Cover 

Herbaceous Grass and herbaceous cover 

Agricultural crops 

Other Land Built up habitation 

Bare land 

Water Water 

No data No data 

Validation and changes in forest cover and use 

Pre-labeled polygons in ESRI shapefile format and the remotely sensed imagery will be provided to all 

countries with sample tiles for in-country, expert validation. Polygon labels will be checked for 

accuracy against each time period of imagery. Ancillary, country-specific land cover data sets (such as 

forest inventory and vegetation type maps where available) and qualitative information obtained from 

the Degree Confluence Project (www.confluence.org) and Google Earth (www.earth.google.com) can 

also be used for validation. Important data is also held in many other organisation such as State and



Territory and local organisations and these are encouraged to submit data to the on-line data portal being 

set up under the FRA website: http://www.fao.org/forestry/fra2010-remotesensing/en/. 

Forest cover changes both positive (afforestation or natural expansion) and negative (deforestation) 

between time periods will be calculated by summing polygon area for those polygons that changed 

labels from time one to time two or three. All polygons with tree cover and those identified as forest 

cover change will also be assigned a cover and land use code for each time period. Changes in land 

use will be tallied in the same manner as forest cover. Global ecological zones (FAO, 2001), and 

global political and administrative boundaries (FAO, 2006b) will be used as summary units. 

EXPECTED RESULTS 

Summaries of forest/tree cover and land use will be produced at regional, eco-zone and global levels 

for 1990, 2000, and 2005. Change in forest/tree cover and land use between periods will also be 

summarized. Country-level statistics will not be reported in the RSS as the global sampling design 

precludes statistically valid estimates at the national level in most cases due to the small number of 

samples. The MODIS VCF product based on an updated version of Hansen (2003) will use the RSS 

sample tiles to validate the latest version (circa 2005) at 250-meter spatial resolution. This analysis will be 

used to generate a new global map of the world’s tree cover planned for release in 2011. 

A database of all imagery chips used in the RSS is available for free download via the internet 

(http://www.fao.org/forestry/fra2010-remotesensing/en/). This repository of global, multi-temporal 

imagery and ancillary information has the potential to become an important data source for researchers, 

land managers and policy makers and can serve as both a base for future studies and an archive of the 

RSS. 

A pilot study analysis of around 250 samples worldwide is currently underway at the time of writing this 

paper in July 2009. Australia is participating in the Pilot Study and has 13 sample tiles to process. 

DISCUSSION 

Statistical sampling versus wall-to-wall mapping 

The RSS is employing a systematic sampling approach covering most of the land surface of the Earth. 

Though global coverage remotely sensed datasets do exist, they exhibit a relatively coarse spatial resolution 

(250+ meters) and many forest cover changes take place at spatial scales smaller than can adequately be 

measured with ‘large’ pixels, especially in the tropics (Hansen, 2008b). 

Landsat data was chosen as the preferred data source for the FRA RSS because it has a suitable pixel size 

(30m) to detect small patches of forest change and because it has the best historical archive of global data 

(Williams, 2006). The commendable decision in 2008 by the USGS to open the entire Landsat archive for 

use overcomes one of the major historical limitations to use of these data. However, the large data volumes 

and difficulties of automating the processing to produce successful results across a wide variety of 

ecosystems, currently limit the ability to develop a global, wall-to-wall map of the worlds forests at 

Landsat-scale spatial resolution. These constraints are being addressed and promising results are 

emerging on regional and continental scales in North America (Masek et al., 2006), Australia (Caccetta 

et al., 2007), central Africa (Hansen et al., 2008a), and Europe (Pekkarinen et al., 2009). 

A sampling approach using 30m spatial resolution imagery was chosen as a solution that achieves 

manageable data volumes, comprehensive coverage of sample plot locations, and returns statistically valid 

results at regional and global spatial scales. A thorough review of statistical sampling compared with wall-towall 

mapping methods can be found in Stehman et al., 2003. A detailed justification for the FRA RSS 

sampling strategy can be found in FAO, 2007a. 

Land cover and land use information are both needed 

The RSS will produce results on both forest cover and forest land use. Land cover is what can be detected 

from a remotely sensed instrument, like the Landsat satellite, and refers to the biophysical attributes of the 

Earth’s surface. Land use implies a human dimension or purpose characterizing a location. It can sometimes 

be inferred from remotely sensed data, however, is typically only verified with local, expert knowledge or



data collected in situ for a particular location. Accurate information on land use is critical to understand the 

drivers of forest cover change (Lambin and Geist, 2005) and help develop effective policies and strategies to 

reverse forest loss that is one of the global objectives for forests (UNFF, 2007). 

Country involvement and capacity building 

FAO and its partners will work closely with national governments and institutions and with a wide range 

of other international and non-governmental organizations. Countries will be involved in the analysis to 

include national data and local knowledge to help ensure the results are validated and as accurate as 

possible. 

The FAO developed computer software for viewing the imagery and labelling the changes will be made 

freely available to participating countries. A series of training workshops will be held around the world 

in regional centres to improve the technical capacity of many staff in analysing remote sensing imagery. 

The access to free remote sensing data and software will particularly benefit developing countries with 

limited forest monitoring data or capacity. The long-term aim of this is to strengthen the abilities of 

countries to regularly monitor and manage their forests and enable informed decisions as well as meet 

the reporting requirements of many national and international processes. 

Australia’s involvement 

Australia’s forests and ecosystems are significant natural assets and are of global importance. They are highly 

valued, have many uses and provide a wide range of valuable products and benefits for society. Australia has 

over 149 million hectares of forests (NFI, 2008), which equates to around 4% the world’s forests (FAO, 2006). 

There are increasing requirements to measure and monitor the extent and condition of Australia’s forests 

for management purposes and for domestic and international reporting requirements. The traditional focus 

on wood production values and estimates of sustainable yield has been long recognised as being too 

narrow; however, data collection systems for monitoring the broader range of forests and forest ecosystems 

are limited in both area and length of study. Australia has set up a sophisticated National Carbon 

Accounting System (Richards and Brack, 2004) to monitor the stock and change in carbon stocks. This 

system has been designed to meet the requirements for monitoring and reporting change in tree cover for 

the purposes of reporting to UNFCC and the Kyoto Protocol. There are opportunities for using the NCAS 

data to assist the validation of the FRA samples by being able to spatially compare the presence of trees. 

However, the NCAS on its own does not fulfil the data needs for FRA because methods and definitions are 

different (e.g. the NCAS is a pixel based system with one class of woody vegetation cover with a minimum 

mapping unit (MMU) of 1ha and the FRA RSS uses an object polygon based approach with several 

vegetation classes (trees, shrubs, herbaceous, and has a MMU of 5 ha. The Australian National Forest 

Inventory data is based on State and Territory forest and vegetation mapping and includes forest type which 

is needed to translate into the global FRA classes. Neither NCAS nor the NFI datasets meet the FAO 

requirements alone as FRA forest definition includes a land-use component (FAO, 2006). Thus additional 

information such as that collected by the Australian Collaborative Land Use Mapping Programme 

(ACLUMP Leslie et al. 2006) will be required to obtain satisfactory final results. 

There is still no adequate national forest monitoring system for a range of other values such as biodiversity 

and other values (Lindermayer et al. 2007, Brack, 2007). Gerrand and Clancy (2007) argued that Australia 

should be setting up a national framework for a monitoring system because current approaches using 

periodic map compilations are not adequate to report on change in forest extent or condition. The 

proposed Continental Forest Monitoring Framework (Wood et al. 2007) developed by the National 

Forest Inventory and supported by States and Territory agencies as a concept would help address many 

of these concerns. The sampling system proposed under the CFMF was for a 20km grid spacing to 

provide many more samples to get better statistical results at a regional level. 

The FRA samples are approximately 100km apart and thus do not have enough samples to provide 

reliable results at a local scale. For example, in Tasmania there are only 10 samples - including one each 

on King and Flinders Islands (Figure 2). This is clearly not enough to capture the diversity or accurately 

estimate the area of forest or change in Tasmania and other methods are needed to report on forest extent 

and change at the local scale.



The CFMF has not yet been implemented in Australia due to a lack of funding. With the FRA sampling 

there now exists a starting platform of data that can help build the data inputs to get a CFMF style system 

up and running at reduced cost. The biggest costs of course are the field work but even the start with the 

image samples and processing is a useful step forward. Australia has not formally confirmed it will 

complete the analysis of the FRA RSS sample sites at the time of writing this paper (July 2009) and one 

aim of this paper is to encourage awareness of the work to build support and commitment to the work. 

CONCLUSIONS 

The FRA 2010 RSS will be a systematic, comprehensive, global study of tree cover and forest land-use 

changes from 1990 to 2000 to 2005. It presents a consistent methodology for monitoring forest change 

at a global level that can be expanded for more detailed studies. It is expected that the RSS will 

improve understanding of total forest area changed, the patterns resulting from this change, and the 

processes driving forest cover change globally. This is information that governments, land managers, 

researchers and civil society groups can use to make better-informed decisions regarding the world’s 

forest resources. 

FAO outreach and training activities will help build technical capacity to monitor forest resources in 

many countries. FAO will provide access to remote sensing imagery either through the internet or hard 

media. The image processing software can be used for other studies and monitoring purposes. 

Additionally, a global network of FRA RSS specialists will be built representing a powerful human 

resource for improved technical capacity and proficiency in many countries. 

The FRA 2010 and the RSS combined will provide a basis for reporting on progress towards sections 

of: (i) the United Nations Convention on Biological Diversity’s target of reversing biodiversity loss by 

2010, (ii) the Millennium Development Goals, (iii) the Global Objectives of the UN Forum on Forests, 

(iv) the International Tropical Timber Organization’s Objective 2000. If countries choose and have the 

resources to do so, the methods have the potential to form a platform for developing more detailed 

reporting capabilities at a national level such as those required for the land use and land use change 

results for the UN Framework Convention on Climate Change and the Kyoto Protocol and the 

Reducing Emissions from Deforestation and Forest Degradation in Developing Countries (REDD). 

Australia is encouraged to participate in the FRA RSS and to complete the analysis of the samples as 

part of its contribution to this important global study and also as an important step to developing a 

better national forest monitoring system. 


The FRA RSS is a partnership between FAO, the Joint Research Centre of the European Commission, US 

Geological Survey and NASA, South Dakota State University, and Friedrich-Schiller University and ,many 

national experts. We would like to thank: Ralph Ridder, Frederic Achard and colleagues at JRC (see Bodart et. 

al.), Thomas Loveland, John Latham, Antonio Di Gregorio, Antonio Martucci, Ilaria Rosati. Many others 

have helped — please accept our thanks. Funding was supported by initial grants from NASA, the 

Governments of Finland, Australia, and funding for 2009-11 from the European Commission. For more 

information please visit: http://www.fao.org/forestry/fra2010-remotesensing/en/ 

LITERATURE CITED 

Achard, F., Eva, H.D, Stibig, H.J., Mayeaux, P., Gallego, J., Richards, T., Malingreau, J.P., 2002. Determination of 

deforestation rates of the world’s humid tropical forests. Science, 297, pp. 999-1002. 

ACLUMP, 2006. Guidelines for land use mapping in Australia: principles, procedures and definitions. A technical 

handbook supporting the Australian Collaborative Land Use Mapping Programme Edition 3. Eds R. Leslie et al. 

Bureau of Rural Sciences, Canberra, 83pp. 

Bodart, C., Beuchle, R., Simonetti, D., Eva, H., Raši, R., Carboni, S., Brink, A., Stibig, H. Achard, F., Mayaux, P.,2009. 

Global monitoring of tropical forest cover changes by means of a sample approach and object-based classification 

of multi-scene Landsat imagery: pre-processing and first results. (ISRSE – this meeting proceedings). 

Brack, C.L., 2007. National Forest Inventories and Biodiversity Monitoring in Australia. Plant Biosystem 141: 

104-112. 

Caccetta, P.A., Furby, S.L., O'Connell, J., Wallace, J.F. and Wu, X. 2007. Continental Monitoring: 34 Years of 

Land Cover Change Using Landsat Imagery, 32nd International Symposium on Remote Sensing of 

Environment, June 25-29, 2007, San José, Costa Rica.



Desclée, B, Bogaert, P., Defourny, P., 2006. Forest change detection by statistical object-based method. Remote 

Sensing of Environment, 102, pp. 1-11. 

Duveiller, G., Defourny, P., Desclée, B., Mayeaux, P., 2008. Deforestation in Central Africa: estimates at regional, 

national and landscape levels by advanced processing of systematically-distributed Landsat extracts. Remote 

Sensing of Environment, 112, pp. 1969-1981. 

FAO, 2001. Global Ecological Zoning for the Global Forest Resources Assessment, 2000. Forest Resources 

Assessment - Working Paper 56. Rome, Italy. http://www.fao.org/docrep/006/ad652e/ad652e00.htm. 

FAO, 2005. Land Cover Classification System (LCCS), Version 2: Classification Concepts and User Manual. Ed. 

Di Gregorio, A. http://www.fao.org/docrep/008/y7220e/y7220e00.htm 

FAO, 2006a. Global Forest Resources Assessment 2005, FAO Forestry Paper 147. 

ftp://ftp.fao.org/docrep/fao/008/a0400e/ 

FAO, 2006b. The Global Administrative Unit Layers (GAUL). 

http://www.foodsec.org/docs/GAUL_Doc01_Ver15.pdf 

FAO 2007a. Global Forest Resources Assessment 2010, Options and recommendations for a global remote 

sensing survey of forests. Forest Resources Assessment programme Working Paper 141. 

ftp://ftp.fao.org/docrep/fao/010/ai074e/ai074e00.pdf. 

FAO, 2007b. Specification of National Reporting Tables for FRA 2010. Forest Resources Assessment 

Programme Working Paper 135. http://www.fao.org/forestry/media/14119/1/0/ 

FAO, 2008. National Forest Monitoring and Assessment Manual for integrated field data collection. NFMA Working 

Paper 37/E. 

Gerrand, A. and Clancy, T. Why we need a national forest and ecosystem monitoring system: Combining the 

Continental Forest Monitoring Framework and Long Term Ecological Research network into a broader 

Terrestrial Ecosystem Research Network. Proc. Inst. of Foresters of Australia Conference, Coffs Harbour, 2007. 

Gutman, G., Byrnes, R., Masek, J., Covington, S., Justice, C., Franks, S. and Kurtz, R., 2008. Towards monitoring 

land-cover and land-use changes at a global scale: The Global Land Survey 2005. Photogrammetric 

Engineering and Remote Sensing, 74, pp. 6–10. 

Hansen, M.C., Dimiceli, C., Sohlberg, R.A., 2003. Global percent tree cover at a spatial resolution of 500 

meters: First results of the MODIS vegetation continuous fields algorithm, Earth Interactions, 7, paper 

no. 10, 15pp. 

Hansen, M., Roy, D.P., Lindquist, E., Justice, C., Altstatt, A., 2008a. A method for integrating MODIS and 

Landsat data for systematic monitoring of forest cover and change in Central Africa. Remote Sensing 

of Environment, 112, pp. 2495–2513. 

Hansen, M.C., Stehman, S.V., Potapov, P.V., Loveland, T.R., Townshend, J.R.G., DeFries, R.S., Pittman, K.W., 

Arunarwati, B., Stolle, F., Steininger, M.K., Carroll, M., Dimiceli, C., 2008b. Humid tropical forest clearing from 

2000 to 2005 quantified by using multitemporal and multi resolution remotely sensed data. Proceedings of the 

National Academy of Science: 105, pp. 9439-9444. 

Intergovernmental Panel on Climate Change, 2007. Climate change 2007—The Physical Science Basis: 

Contribution of Working Group I to the Fourth Assessment Report of the IPCC (Cambridge University 

Press, Cambridge, UK). 

Lambin, E., and Geist, H., (eds), 2005. Land Use and Land Cover Change: Local Processes, Global Impacts (Springer, 

New York). 

Lindenmayer, D. et al. 2007. On borrowed time: Australia's environmental crisis and what we must do about it. 

Penguin books, Australia. 138 pp. 

Masek, J.G., Vermote, E.F., Saleous, N.E., Wolfe, R., Hall, F.G., Huemmrich, K.F., Gao, F., Kutler, J., Lim, T., 2006. 

A Landsat surface reflectance dataset for North America, 2990-2000. IEEE Geoscience and Remote Sensing 

Letters, 3, pp. 68-72. 

Pekkarinen, A., Reithmaier, L. Strobl, P., 2009. Pan-European forest/non-forest mapping with Landsat ETM+ 

and Corine Land Cover 2000 data. ISPRS Journal of Photogrammetry and Remote Sensing 64, 171-183. 

Potapov, P., Hansen, M.C., Stehman, S.V., Loveland, T.R., Pittman, K., 2008. Combining MODIS and Landsat 

imagery to estimate and map boreal forest cover loss. Remote Sensing of Environment, 112, pp. 3708- 

3719. 

Richards, G. P. and C. Brack, 2004. "A continental biomass stock and stock change estimation approach for 

Australia." Australian Forestry 67(4): 284-288. 

Stehman, S.V., 2003. Statistical sampling to characterize land cover change. Remote Sensing of Environment, 86, pp. 

517-529. 

The Group of Eight Summit, 2008. Toyako G8 Summit Leaders Declaration Hokkaido Toyako, 8 July 2008. 

http://www.g8summit.go.jp/eng/doc/doc080714__en.html 

Tucker, C.J., Grant, D.M., Dykstra, J.D., 2004. NASA’s global orthorectified Landsat data set. Photogrammetric 


UNFF, 2007. Non-legally Binding Instrument on all Types of Forests. UN General Assembly A/C.2/62/L.5.



Williams, D.L., Goward, S. and Arvidson, T., 2006. Landsat: yesterday, today, and tomorrow. Photogrammetric 


Wilson, Edward O. (ed.), 1988. Biodiversity. National Academy Press, Washington, DC, 521pp.



ABSTRACT 

SUSTAINABLE POLE SUPPLY PROJECT 

Hugh Stone 1 and David Wood 2 

Ergon Energy is facing a critical shortage in the future supply of hardwood power poles 

with demand expected to increase from 13,000 to 25,000 p.a. over the next 15 years. 

This is unlikely to be met by traditional suppliers. With over 900,000 poles in the 

network, of which approximately 94% are hardwood, Ergon Energy is facing a 

significant shortfall, commencing as early as 2009. Hardwood is the preferred pole type 

due to its economic advantages and its superior electrical insulation properties. It is safe 

and reliable. 

The Sustainable Pole Supply Project has been initiated to ensure that Ergon Energy has a 

reliable and economic future supply of poles. To achieve this, the corporation seeks to 

acquire native forests and to establish pole plantations. These forests will be located in 

diverse geographic areas to minimise risk and will be selectively harvested to provide a 

renewable supply of hardwood power poles while preserving ecosystems and 

maintaining habitats. Additionally, the benefits of carbon capture and biodiversity 

conservation can offset other activities of the corporation in meeting its sustainability 

objectives. 

This paper discusses the risks and benefits to Ergon Energy and the community of 

undertaking the project. 

INTRODUCTION 

Ergon Energy manages $5.6 billion worth of electricity infrastructure assets over one million square 

kilometres of regional Queensland. Ergon Energy’s service area effectively covers 97% of the state - 

equivalent to most of the eastern seaboard of the United States. This represents one of the largest and 

most diverse infrastructure networks in the western world (EECL 2007). A large majority of this 

network is held up, literally, by approximately 900 000 poles of which approximately 94% are 

hardwood. The map in Appendix 1 shows the area serviced by Ergon Energy’s network in Queensland 

and the area covered by the distribution network. 

DEMAND FOR POLES 

The demand for timber poles arises from network maintenance, network expansion to service 

consumer demand, and network upgrading to ensure feeders are not over loaded and enough capacity 

exists to meet peak demand. 

Maintenance of the Existing Network 

The maintenance of poles and other assets associated with the Ergon Energy network is driven by 

policies to meet customer expectations of reliability, employee, public and environmental safety; the 

present and future needs of the assets; legislative requirements; and internal and external benchmarks 

The vast geographical spread of Ergon Energy’s service area has a number of impacts on the 

performance of the distribution system. The geographic and related environmental features which 

reinforce the need for a robust maintenance program include: 

• High exposure to cyclone risk; 

• High storm and lightning activity; 

• Significant summer-winter and day-night temperature variations; 

• High rainfall areas; 

• Other weather impacts (e.g. the Channel Country flooded for hundreds of kilometres); 

• Active termite populations affecting power pole integrity; 

1 

Ergon Energy, Toowoomba, Qld. Email: Hugh.Stone@ergon.com.au. 

2 

Ergon Energy, Townsville, Qld.



• Unstable soil types (e.g. Darling Downs black cracking soils). and 

• Bushfire hazards. 

All electrical utilities have in place strategies and plans to inspect assets and ensure they are 

maintained and replaced on a cyclic basis. This usually entails physically inspecting each structure or 

piece of equipment, with cycles determined by legislation or engineering best practice (Sanders & 

Noonan 2008). Ergon Energy has a routine inspection and maintenance program for poles and 

associated equipment covering the entire network on a 4 yearly cycle, known internally as the Asset 

Inspection and Defect Management Program. 

This inspection program commenced in April 2003 (initially on a 3 year cycle) and from that time 

approximately 1.5 million pole inspections have been carried out. The first cycle of inspections 

resulted in 10,595 poles being identified as defective and consequently replaced, with a further 19,959 

poles treated with pole base reinforcement (installation of a large metal stake around the pole base as 

support, to extend the life of poles which have ground line or below ground defects) (EECL 1997). 

Network Expansion 

The growth in the population of Queensland has seen a commensurate growth in domestic demand for 

electricity supply, and has been accompanied by a significant industrial demand for electricity supply 

especially in the mining and allied sectors. 

For the last 5 years customer numbers have grown around 2% annually (EECL 2007). In 2007/08 

customer numbers increased by 2.6%. This increase has been driven largely by the expansion and 

development of the regional coastal centres of Hervey Bay, Mackay, Townsville and Cairns, and 

growth in the Toowoomba area. 

Network Augmentation 

As population increases and new industry is established there is an increase in demand from existing 

assets, with a need to expand the capacity of lines and substation assets to cope with increased loads.. 

This can result in reconfiguration of distribution lines to transmission standards with larger poles and 

associated equipment to accommodate actual and projected increases in demand. 

Number of Poles 

26000 

24000 

22000 

20000 

18000 

16000 

14000 

12000 

10000 

8000 

6000 

4000 

2000 

0 

2004 

2006 

2008 

Future Hardwood Poles Requirement 

2010 

2012 

2014 

2016 

2018 

Year 

2020 

2022 

2024 

Defective Replacement New Extension Augmentation 

Figure 1 - Future hardwood pole requirement 

Combining the pole demand requirements from maintenance of the existing network, network 

expansion to meet demand from new customers and network augmentation, Ergon Energy’s current 

2026 

2028 

2030



demand for hardwood poles is around 13,000 annually. Research conducted internally suggests that 

this demand is set to increase steadily to around 25,000 poles annually by 2031 (EECL, nda). Figure 1 

presents this projected pole demand. 

This extrapolation assumes the age profile of existing poles created from historical data, will continue 

to apply. Analysis indicated a pole life of 51 years ± 11 years. The projected curve of future pole 

requirements flattens off from about 2022 due largely to an expectation that a higher proportion of 

new development will move cable underground. 

POLE SUPPLY 

The South East Queensland Forest Agreement (SEQFA) signed in 1999 was an agreement between the 

Queensland Government, the timber industry and conservation groups that sought to secure a viable 

timber industry while securing an adequate and representative conservation estate. The agreement 

involves the transition of publicly owned commercial native forests (State Forests) into the 

conservation estate over a 25 year period with essentially a final harvest of all merchantable wood 

products as each area is completed. 

In the area covered by the SEQFA the supply of poles from State land is governed by a policy 

stipulating that the current supply, which extends until 2014, will be maintained at historic levels 

existing prior to commencement of the SEQFA. Supply arrangements will be reviewed at the 

completion of this period and there is no in-principle reason why new contracts could not be entered 

into for pole supply up until the end of the SEQFA period in 2024. Present Government policy 

stipulates that there will be no net increase in the supply of pole products, and that substitute policies 

will promote a transition from products supplied from native forest to plantation products (Stephen 

Walker DNRW, pers. comm. 16/2/09) under the SEQFA. 

The phasing out of native forest harvesting is accompanied by a transition plan to move to plantation 

timber (both hardwood and softwood) for all forest products. Despite comprehensive strategies to 

allow industry to make this transition, the use of plantation grown poles is largely untested and has not 

provided a degree of confidence sufficient for Ergon Energy to rely solely upon plantations to meet a 

shortfall of pole supplies as native forest logging is wound up. 

Traditionally, hardwood poles have been obtained 35% from public resources and 65% from private 

property. (Chris Bragg, DNRW pers comm. 20/02/09). However, while public land accounts for only 

35%, they are typically among the larger pole sizes (12.5m and 14m), while private resources provide 

poles of a lower length and strength range (EECL, n.d.). 

Given then that Ergon Energy faces a sharp increase in demand for hardwood poles, and that current 

suppliers are unable to increase their supply, and particularly for the common 12.5m and 14m pole 

lengths, either a new approach to sourcing hardwood timber poles is required, or a suitable alternative 

has to be found. 

ALTERNATIVES TO HARDWOOD POLES 

Plantation Softwood 

An in-service trial of slash pine poles commenced in the mid eighties. After 15 years, results were 

promising, suggesting physical deterioration of CCA treated slash pine poles is only marginally higher 

than that of CCA treated hardwoods (Powell , nda). However, the strength of softwood poles has been 

questioned and consequently they are not currently considered suitable for all Ergon Energy 

applications. 

While the substitution of plantation grown softwood poles is feasible, this option would require a 

relatively long planning horizon to realise, on the assumption that a) existing maturing plantations are 

likely to be committed already under wood supply agreements; and b) a softwood pole clearly needs to 

be larger in diameter to be equivalent in strength to a hardwood pole, and would therefore need to be 

sourced from end of rotation stands. Thus while softwood poles may provide a longer term option, 

they cannot be considered a realistic substitute for hardwood poles for at least the next 25 years, until 

young unallocated softwood plantations reach the necessary pole size.



Concrete Poles 

Concrete poles can be readily supplied in specific sizes and they offer an option that matches or 

exceeds the strength properties of hardwood poles. Additionally, they have greater structural 

reliability and reduced maintenance costs. As such they are preferred for use in transmission and sub 

transmission (66kV and above). Their use in the distribution network is not common due to the initial 

high capital cost of concrete poles (refer Table 1), higher field installation costs, considerations of 

current leakage and the lack of versatility in accommodating additional future fittings (eg crossarms). 

While their use is warranted within the distribution network as transformer poles and on higher 

voltage sub transmission feeders, they are not as versatile as hardwood poles and their higher capital 

costs make it difficult to justify them for universal use. 

The environmental impact and sustainability of using concrete poles also needs to be considered. 

Manufacture of these poles requires use of non-renewable raw materials and large amounts of energy 

producing a relatively large carbon footprint in comparison to the use of timber which is both a 

renewable resource and a carbon sink. 

Ergon Energy continues to research alternatives to solid hardwood timber poles with long term field 

trials. Other alternatives include steel poles, fibre composites, laminated timber and hybrids using a 

combination of materials. Hardwood timber maintains its status as the preferred option for use in the 

distribution network mainly because of cost. It has desirable electrical properties, it is safe, strong and 

reliable and importantly, in a time when climate change is a real concern and must be considered in 

long term business planning, it is a renewable resource that sequestrates carbon. 

Table 1 A comparison of pole costs for timber and concrete 

Pole Size 

length/strength 

Average Pole cost ex 

depot ($ per pole) 

Con- 

Wood 

crete 

Typical use 

9.5/5 $140 $400 Urban Service pole 

11/5 $180 $490 SWER intermediate pole or Urban LV reticulation pole 

12.5/5 $220 $690 Urban HV/LV reticulation pole 

12.5/12 $490 $950 Urban Transformer pole 

14/8 $410 $800 Rural backbone feeder intermediate pole 

18.5/12 $870 $1750 66kV transmission intermediate pole 

20/12 $1360 $2000 66kV transmission intermediate pole 

SUSTAINABLE POLE SUPPLY PROJECT (SPSP) 

With no immediate substitute for hardwood timber poles, an immediate forecast shortage of poles over 

the next 20 years, a likely continuing demand for poles well beyond the 20 year projections demand 

cited above, it was logical step for Ergon Energy to embark on a project which will provide a 

sustainable supply of hardwood timber poles – the SPSP. 

The SPSP will ultimately provide a sustainable source of poles capable of satisfying 50% of the 

projected demand. The project scope allows for the purchase or lease of up to 10 000 hectares of land 

suitable for growing timber. The majority of the areas covered by the project will be native forests, 

where enrichment planting would be done in understocked or degraded forest. There is also provision 

for the conversion of previously cleared farm land to hardwood plantations. Exercising these options 

would optimise production of poles from areas with a mosaic of past land use, and add some benefits 

by means of more efficient land management and enhanced environmental values. 

Internally, the project is directed to the search for and acquisition of land considered suitable for the 

organisation’s needs, while an external forest manager is engaged to provide advice on the suitability 

of potential properties and to create and implement a forest management plan for each acquired 

property. It is intended that, once the SPSP is established, minimal internal resources will be required 

to manage the project with a heavy reliance on external resources to provide ongoing forest 

management expertise.



Ideally the properties will be 1,000 to 2,000 hectares in size to allow geographical separation of each 

property in an effort to minimise risks associated with pathogens, fire and climatic influences. The 

first property was purchased in 2008 under the project and selection of additional properties is well 

advanced. 

The SPSP contains other options than outright purchase or lease.. Agreement has been made with 

Wide Bay Water to purchase poles from plantations established to use sewage effluent in the Hervey 

Bay area. Discussions have also been held with representatives from Australian Forest Growers on 

long term sales agreements for suitable species. No specific model of management and procurement is 

stipulated at this stage; rather, a range of options will be pursued to bring more flexibility into the 

supply chain. 

Property Selection 

To assist in the location of potential properties for purchase or lease, a combination is employed of 

desktop GIS analysis, field investigation and inspection, followed by contact with stock and station 

agents. The first step is to define areas more likely to contain appropriate vegetation types with 

desirable pole species, soils and climate conditions. A GIS consultant was engaged to identify suitable 

land parcels consistent with a pre determined scope and set of criteria. 

Figure 2. Example of map output with Regional Ecosystems containing favoured species shaded 

in different colours. Orange–Corymbia citriodora, Light Blue-Eucalyptus acmenoides, 

Dark Blue-Eucalyptus sideroxylon. (Fitzpatrick 2009) 

The data sets used in the GIS analysis are: 

• Queensland Department of Natural Resources and Water (DNRW) FPC data for 

entire Queensland provided in MGA zones 54, 55, 56 (QLD DNRW 2007); 

• Queensland Department of Natural Resources and Water (DNRW) LANDSAT scene 

data for entire Queensland (QLD DNRW 2007); 

• Queensland Local Government Boundaries for entire Queensland (DNRW). 

• Queensland Environment Protection Agency’s Version 5.0 Regional Ecosystems and 

Remnant Vegetation (RE) data for entire Queensland (QLD EPA 2007); and



• Cadastre with tenure attributes over the Stage One project area (QLD Lands 

Department). 

• Ergon list of acceptable and preferred pole species 

The criteria in use to identify potentially suitable properties that warrant further on-ground 

investigation are: 

• Searches are restricted to freehold land only. 

• A minimum land parcel size of 100 hectares will be used as a filter. 

• DNR&W Foliage Projection Cover (FPC) based on 12% FPC which equates to an 

approximate canopy cover of 20%.(to capture regrowth areas). 

• A refined list of Regional Ecosystems (RE’s) most likely to contain suitable forest 

structure (predominantly woodland and tall forest) and species with Durability Class 1 

& 2 species. 

• Only RE’s with a Vegetation Management Act status of ‘Not of Concern’ are 

desirable. It was considered filtering to RE’s listed as ‘Of Concern’ or ‘Endangered’ 

carried risks for future availability for harvest. This list was created manually based on 

published descriptions of each Regional Ecosystem and local knowledge. 

This process creates a map output that can be used to target geographic locations with mapped suitable 

vegetation. It should be noted that non-remnant forested areas suitable for the project are also under 

consideration, and in some ways are more desirable in terms of offset trading. These are difficult to 

refine through GIS desktop analysis as the vegetation type is not mapped; however, assuming the FPC 

criteria are met, and mapped ‘Not of Concern’ remnant vegetation is adjacent or in the vicinity, they 

are included in the final output. 

A comprehensive due diligence process has been developed to ensure that the growing stock and site 

quality of purchased properties will provide a reasonable return on investment. Legal due diligence 

has also included assessment of environmental risks. As a result, several properties have been rejected 

because their growth potential has not been able to match possible liabilities presented by major weed 

infestations. 

Investigation of suitable properties is ongoing and properties have been purchased at Ravensbourne 

(‘Eagle Rock’), North East of Toowoomba and at Tiaro, North of Gympie. A forest management plan 

for Eagle Rock has been created and its implementation commenced with trials of both commercial 

and non commercial thinning in progress, with a view to improving stands to focus site resources on to 

future pole crop trees. Some enrichment planting of old log dumps, rehabilitation of access tracks and 

weed control has also been carried out. 

Timber stands will be selectively harvested according to the Queensland Code of Practice 

administered by DNRW. Management will adopt the principles of Ecological Sustainable Forest 

Management (ESFM) with the intent of maximising all forest values including biodiversity and 

conservation values. 

BENEFITS ASSOCIATED WITH THE PROJECT 

The main benefit of the SPSP is a sustainable supply of hardwood poles to help meet internal current 

and projected demand for hardwood poles. While a proportion of Ergon’s forecast pole demand will 

still be sourced on the open market, having its own sustainable supply of poles offers a degree of 

protection from market forces and security for business planning. 

Other benefits associated with the project include carbon sequestration; biodiversity offset credits, and 

conservation and ecosystem health.



Carbon Sequestration 

The introduction of an Emissions Trading Scheme (ETS) by the Australian Government is unlikely to 

provide Ergon Energy with carbon credits available for trade, as sequestration of carbon from native 

forests is not included as a source of carbon credits within the current proposed ETS framework. 

While plantations established on previously cleared land acquired under the project could be certified 

for carbon trading benefits, the benefits associated with carbon sequestration in terms of corporate and 

community responsibility are desirable even without an economic gain in the form of emission 

credits.. The impacts on climate change from carbon emitted into the atmosphere through human 

activity are generally accepted as a threat to the health of the global environment, and individuals and 

corporations are encouraged to reduce their carbon footprint as best they can. The sustainable pole 

project works to capture and store atmospheric carbon and effectively reduce Ergon Energy’s 

footprint. 

Biodiversity Offset Credits 

The Queensland Government introduced the Queensland Government Environmental Offsets Policy 

(QGEOP) on the 1st July 2008 across the whole of government in an effort to replace environmental 

values lost through development. The policy is supported by offset policies that provide detailed 

direction for specific environmental issues (vegetation, koala habitat, marine fish habitat). These 

Activity-Specific Offset Policies are developed and administered by different government agencies 

according to their environmental portfolio. 

The QGEOP defines an offset as “an action taken to counterbalance unavoidable, negative 

environmental impacts that result from an activity or a development. An offset may be located within 

or outside the geographic site of the impact. Offsets can be either direct or indirect. A direct offset is 

aimed at on-ground maintenance and improvement of habitat or landscape values, such as long term 

protection of existing habitat, or legally securing regrowth vegetation for inclusion as remnant 

vegetation upon maturity. An indirect offset aims to improve knowledge, understanding and 

management leading to improved conservation outcomes. Examples of indirect offset include funding 

species-specific research, sponsoring implementation of species recovery plans, or providing 

infrastructure, such as nest boxes or fauna crossings, to minimise impacts on fauna. Generally 

speaking a direct offset is preferable to an indirect offset, because the impact is easier to measure. 

According to the QGEOP, where Ergon’s activities affect mapped remnant vegetation, essential 

habitat (under the Vegetation Management Act) or listed Rare, Vulnerable or Endangered species 

(under the Nature Conservation Regulations 2008) an offset must be provided in line with the relevant 

Activity-Specific Offset Policy, potentially at a ratio greater than 1:1. 

While the implementation of the QGEOP with its associated activity-specific offset policies is in its 

infancy, and remains largely untested, it currently requires any adverse impact on native vegetation to 

be offset to ensure there is no overall loss of environmental value as a result of the activity. 

A general interpretation of the Activity-Specific Offset Policy for vegetation would be the securing of 

like vegetation not currently mapped as “remnant” 3 . Preferably this vegetation would be located 

geographically close to the area being disturbed. For example, if 5 hectares of “Of Concern” remnant 

vegetation is cleared, the administering authority may issue a conditional permit requiring the holder 

to legally secure a floristically and ecologically “like” area of non-remnant vegetation for future 

inclusion as mapped remnant vegetation – essentially replace what will be removed from the mapped 

remnant vegetation map as a result of their development. Our experience to date has seen disturbed 

areas mandated to be offset at ratios greater than 1:1 depending on the conservation status of the 

ecosystem disturbed. 

A general interpretation of the policy under an activity-specific offset policy for biodiversity may 

require a net increase of individuals from pre disturbance levels for each individual species impacted. 

Given this, the impact to Ergon Energy is largely dependent on interpretation. As the NCA defines all 

native vegetation as protected, a strict interpretation of the policy would define any disturbance to a 

3 

“Remnant” status is technically achieved when the upper stratum of a stand reaches 70% of site height and crown cover of 

the upper stratum is => 50%.



native species, regardless of its individual or ecosystem conservation status, as a conservation loss 

which therefore must be offset. By definition of the policy, the offset must result in a net increase in 

individual numbers for each species impacted. 

For land owners the introduction of this policy brings financial opportunity through the sale of offset 

credits. In the vegetation example above, the permit holder could secure an offset from a land owner 

by purchasing the vegetation community values of their property (or part thereof) by way of a caveat 

over the property. Essentially this would result in the permit holder paying a nominated value per unit 

area to the land owner, who in return is legally bound to retain the native vegetation which in time 

would reach “remnant” status, be classified, and subject to legislative controls. 

Regardless of the interpretation of the policy, with Ergon’s continually expanding distribution 

network, the corporation will be regularly required to secure offsets. Being a landowner will 

potentially make the process of sourcing and securing offsets much more efficient. Furthermore, land 

acquired under the project will provide opportunities for sale of offset credits to other entities seeking 

offsets for their impacts. The sale of offset credits would provide a positive financial injection back 

into the project, effectively reducing the price paid for the property and providing funds to manage 

biodiversity threats, such as fire, pests and weeds, to ensure remnant status is achieved. 

Current experience suggests the sale of offset credits will provide a per hectare payment up-front to 

compensate the landowner for retaining the vegetation, and a further per hectare sum for management 

costs associated with restoring the vegetation to remnant status. Management costs are assigned for the 

management of fire, pests, weeds and any other long term threat to the health of the vegetation. An 

offset agreement is drawn up detailing responsibilities of all parties in ensuring the vegetation reaches 

remnant status. Once the regulator assesses the area as remnant, the offset area is included on the 

official remnant vegetation mapping and the offset agreement is deemed to be satisfied. 

In summary, the Sustainable Pole Supply Project will provide a residual benefit through biodiversity 

offset credits which can be used internally to satisfy offset obligations stemming from other arms of 

the business, and/or sold externally to provide income to the project and funds for land management. 

Conservation and Ecosystem Health 

Timber poles can be produced without compromising biodiversity or conservation values. Through 

the application of the principles of Ecologically Sustainable Forest Management (ESFM) all forest 

values can be maintained. Sensitive areas, such as riparian zones, will be identified and then excluded 

from the harvest area. In the case of Eagle Rock, past land use has led to a degraded forest structure 

and health. Ergon Energy is confident that biodiversity and conservation values will increase over 

time as the benefits from sustainable silvicultural treatment and management are realised. 

Initial approaches have been made to universities for the conduct of surveys for threatened species and 

to monitor biodiversity programs. This will produce information on biodiversity changes over time, 

and provide an education resource for the universities. 

RISKS ASSOCIATED WITH THE PROJECT 

In managing the projected increase in demand for hardwood poles against a predicted reduction in 

market supply, the greatest risk to Ergon Energy would stem from doing nothing. The acquisition of 

land provides Ergon with an asset which can be liquidated quickly, while the identified benefits justify 

the SPSP proceeding. Against these benefits, the risk associated with the project is considered 

acceptable. Other risks include: 

a) The objectives of the project may not be realised, or the percentage of poles produced may be less 

than anticipated. However if a tree does not achieve the form and quality standards to be utilised as a 

pole, it would be saleable as a quota sawlog or salvage grade round timber, at a lower price. In all 

stands there will be a proportion of trees that do not meet pole specifications. 

b) This brings with it an associated risk in the need to focus on non-core business, that is, managing 

timber sales. However the revenue returned from timber sales should at least neutralise the associated 

costs.



c) The major changes in government laws and regulations than occur over the lifetime of even a shortrotation 

pole plantation are recognised as a significant risk, and one that Ergon Energy cannot avoid 

through planning and good management. The greatest risk in this regard would be a change in 

government policy on the legality of harvesting privately owned native forests. Another area of 

uncertainty are the permissible activities in mapped remnant areas. At the time of writing, the media 

has speculated on the impacts of a review of laws associated with native forest regrowth and its 

associated mapping. The conversion of mapped “regrowth” areas to a “remnant” vegetation category 

is one example of how a legislative change could dramatically change the pole supply resource. 

Should this happen, the concept of biodiversity offset trading under the Vegetation Management Act 

could be compromised. 

d) Another significant risk, which has been evident from the commencement of the project, has been 

competition from other land uses, such as agriculture and horticulture for the better quality land sought 

for forestry. 

e) Another recognised risk is the false perception among some agricultural sectors that the demand for 

land for forest establishment will threaten their industry. This has been seen, for example, with some 

members of the sugar industry, and has been accompanied by public and political lobbying. However 

the Sustainable Supply Pole Project is not expected to meet this type of resistance because of its focus 

mainly on existing regrowth, the dispersed nature of property purchases, and the low key approach 

adopted in seeking suitable properties. 

f) The supply of poles from Ergon owned and managed forests could adversely impact private 

suppliers. This is an unlikely risk as pole delivery can be managed so that no suppliers are in fact 

adversely impacted. This could simply be effected by ensuring Ergon poles are distributed across 

Ergon regions and depots. As other electricity utilities face the same dilemma of future shortages of 

hardwood poles, the formation of joint ventures with private growers, or other corporate structures, to 

ensure security of supply is anticipated. 

CONCLUSION 

The hardwood pole is a reliable, safe, renewable asset on which Ergon Energy relies for the 

distribution of electricity across Queensland. The likely unavailability of hardwood poles from current 

suppliers has been the catalyst for Ergon Energy’s proactive initiation of the Sustainable Pole Supply 

Project with the acquisition of native forests to help meet Ergon’s future pole needs, and the 

management of those forests sustainably for a wide range of forest values. 


The authors wish to acknowledge the assistance of David Carleton, Manager of the Sustainable Pole 

Supply Project, and Bernard Fitzpatrick, CTG Consulting, who provided some of the background 

information presented in the paper. 

REFERENCES 

Ergon Energy Corporation Limited. n.d. Hardwood Pole Shortage Report. Internal Report prepared by J. Brooks 

and P. Le. Network Maintenance and Standards, Ergon Energy Corporation Ltd, Queensland. 

Ergon Energy Corporation Limited. 2008. Network Management Plan Part A - Electricity Supply for Regional 

Queensland 2008/09 to 2012/13. Ergon Energy Corporation Ltd, Queensland. 

Sanders, A & Noonan, K. 2008 Wood Pole Management , Overhead Lines Seminar 4 - 5 March 2008 Sydney. 

CIGRE Australian Panel B2- Overhead Lines. 

Powell, M. n.d. Inspection and Assessment of Slash Pine Utility Poles after approximately 15 years in-service 

exposure in several areas of Queensland. Queensland Forestry Research Institute, Queensland. 

Wood Pole versus Concrete Pole Comparison. Internal Report, Network Maintenance and Standards, Ergon 

Energy Corporation Ltd, Queensland. 

Fitzpatrick, B. Identification of potential land parcels for poles, South East Queensland Biogeographic Region. 

Internal report produced for Ergon Energy by CTG Consulting February 2009. 

Queensland Government Environmental Offsets Policy Queensland Government Environmental Protection 

Agency, 2008, //www.epa.qld.gov.au/publications/p02501aa.pdf/



APPENDIX 1 – ERGON ENERGY’S SERVICE AREA



PROCESSING PERFORMANCE AND SAWN PRODUCT RECOVERY 

FROM THINNED NATIVE FOREST REGROWTH LOGS 

FROM SOUTHERN AUSTRALIA 

Russell Washusen 1 , Andrew Morrow 1 , Dung Ngo 1 , Graeme Siemon 2 , 

Tim Wardlaw 3 , Mike Ryan 4 , Martin Linehan 5 and Daniel Tuan 5 

ABSTRACT 

Logs of conventional sawlog quality, diameter and length were obtained from thinned 

E. fastigata (New South Wales), E. sieberi (Victoria), E. regnans (Victoria), E. regnans 

(Tasmania) and E. diversicolor (Western Australia). They were sawn and dried using 

normal industry processing methods for the respective species. The wood processing 

performance and product quality were compared with samples of logs of similar grade, 

diameter and length, from normal log supplies from much older unthinned regrowth. 

Losses due to board end-splitting, board deflection and drying performance of the 

wood from the thinned regrowth was comparable to, and in some cases, better than the 

older regrowth; and the general wood quality tended to be better in the younger thinned 

logs due to less frequent kino and insect damage. The results give support to the 

thinning programs applied in even-aged forests by forest management agencies across 

southern Australia. 

INTRODUCTION 

In eucalypt species that tend to produce even-aged stands, thinning has been conducted experimentally 

by forest management agencies across southern Australia since about the 1950’s. This early work led 

to the first serious attempts to develop commercially viable and silviculturally beneficial mid-rotation 

thinning operations during the “Young Eucalypt Program” in the late 1980’s (Kerruish and Rawlins 

1991). Commercial thinning of small areas of eucalypt production forests is now a common practice. 

The majority of these thinned forests are yet to contribute to a significant supply of sawlogs. However, 

some of the earlier experiments have produced trees of a harvestable size and the thinned stands are 

generally approaching the time when they will produce a significant supply of logs for conventional 

processing. 

Thinning treatments, where they have been monitored, have often produced growth responses in the 

retained trees (Webb 1966, Goodwin 1990, West 1991, O’Shaughnessy and Jayasuriya 1993, Fagg 

2006). The most important effect is accelerated diameter growth rate and as a consequence increased 

sawlog yields. 

Despite the work of Wardlaw et al (2004) who examined differences between large thinned regrowth 

logs and smaller unthinned regrowth, and Innes et al. (2005) who examined differences in wood 

properties of trees of different ages and sizes, there is little information available about the comparable 

performance of logs from thinned regrowth and older conventional sources. This is because direct 

comparisons have not been made between logs of comparable diameter and grade obtained from older 

conventional sawlog supplies and thinned forests. In the latter case, the logs will tend to produce a 

sheath of wood after thinning (Figure 1) that has different wood properties than much slower grown 

unthinned regrowth. 

Some processors are concerned that this growth response may alter wood properties and consequently 

processing performance. The magnitude of longitudinal tensile growth stresses may be different, 

1 

CSIRO Materials Science and Engineering and CRC for Forestry, Contact author. Email: Russell.Washusen@csiro.au 

2 

Forest Products Commission 

3 

Forestry Tasmania 

4 

VicForests 

5 

Forests NSW



impacting on board end-split severity, board deflection during sawing and product undersizing and, as 

a consequence, impact on product recovery and value. During drying the wood may produce different 

drying performance and influence the commercially important defects of surface and internal 

checking. 

Figure 1. Radial strips from the base of the stem from trees of similar diameter from 1939 

regrowth E. regnans (bottom) and 1972 regrowth E. regnans thinned in 1990 (top) 

showing differences in growth ring width in the outer heartwood and sapwood. 

This project was initiated at the request of the processing industry in Australia which wanted to 

quantify differences in processing performance and general wood quality in thinned regrowth forests 

where there had been a growth response. To do this, log samples were selected from thinned and 

unthinned forests so that log diameter, quality and height in the stem were similar. 

The species examined were suggested by industry because of their importance to processors and their 

availability from thinning experiments. They were E. fastigata from New South Wales, E. sieberi from 

Victoria, E. regnans from Tasmania and Victoria, and E. diversicolor from southwest Western 

Australia. It was the primary objective of this project to quantify differences in wood processing 

performance and product quality in logs of a size and quality suitable for processing in existing 

sawmills. 

This work produced five reports for Forest and Wood Products Australia (FWPA), one report for each 

species by region (Washusen et al. 2007 a,b,c,d,e). This paper summarises these sub-projects and 

presents the main findings. 

METHODS 

The evaluation of each species by region was conducted in sawmills located close to the forest of 

interest, and currently processing the respective species from unthinned regrowth sources. This meant 

that each evaluation had unique attributes, partly because of differences in processing methods, and 

partly due to differences in target markets and consequential customer product requirements. While 

these differences were inevitable, each sub-project aimed to understand the effect of increased growth 

rates following thinning on key indicators of wood quality and value. In order to do this each subproject 

adhered as close as possible to the structure that follows. 

Initially, trees from thinned stands were selected and harvested to produce either a single butt log or 

two logs from the lower stem. These were then matched with logs from the same height so that the 

diameter and log grade (for the respective forest management agencies) were as similar as possible.



Coupe selection was also structured to minimise variation in elevation, slope, rainfall, soil types and 

genetic origins. Examples of coupe maps and thinned forests are shown in Figures 2 and 3. These are 

coupe maps for the E. regnans from the Styx Forest Block in Tasmania (Fig. 2) and thinned E. 

regnans from the Central Highlands of Victoria (Figure 3). 

Figure 2. Coupe maps for unthinned E. regnans (left) and thinned E. regnans from the Styx 

Forest Block, Tasmania. 

Figure 3. Thinned E. regnans from the Central Highlands, Victoria. 

Where possible, harvesting of the respective thinned and unthinned coupes for each species by region, 

was conducted at about the same time and by the same methods. Logs were transported to the mills 

either in sawlog length for a single log per tree, or bushlog length where two logs were produced. At 

the mill the logs were measured and graded according to the requirements of the respective State forest 

management agencies. The logs were colour coded so each board produced could be tracked 

throughout processing and individual log statistics produced.



The randomization of the logs for sawing ensured that the following conditions were imposed on the 

processing: 

• Identity of the logs was unknown throughout processing except by supervising research 

staff. 

• Logs were processed by identical strategies in each sawmill. 

• Boards from each sample were randomised within a single drying batch and so were dried 

with the same schedule, and the position within the kiln was reduced as much as possible 

as a variable in drying performance. 

• During the grading by industry graders (Figure 4), resource identity was unknown. 

• During processing staff were employed on one task, and normal staff rotations avoided to 

ensure consistency in processing strategies. 

Many of the conditions imposed on the research restricted the size of the project because of the 

expense in handling and tracking material. Sample size was restricted to a maximum of 30 m 3 of 

sawlogs from each thinned and unthinned forest. This was considered imperative in order to eliminate 

as many as possible of the variables that may impact on processing performance. These variables 

included sawing techniques, resulting in differences in product orientation and sizing, differences in 

drying schedules and drying conditions, and differences in interpretation of board grade and final 

product size. 

A general summary of the log sample locations, age at establishment and thinning, sawing and drying 

methods and final products produced are given in Tables 1 and 2. The year of thinning for most 

species ranged from 1990 to 1994 so that at harvest at least 12 to 16 years had elapsed since thinning. 

One sample of E. regnans from the Styx Forest Block in southern Tasmania was thinned in 1970. 

The coupes selected from the unthinned forests were scheduled for normal harvesting operations. They 

ranged from mature stands or stands dominated by mature trees of unknown age (E. fastigata, E. 

diversicolor and E. sieberi) with 1939 and 1934 regrowth E. regnans from Victoria and Tasmania 

respectively. One stand of unthinned 1974 regrowth E. diversicolor was also selected in Western 


Processing methods (sawing and drying) and final products for the respective species were normal for 

the sawmills where the processing was conducted. Back-sawing strategies and quarter-sawing 

strategies to produce sized boards for drying were applied for E. diversicolor and E. sieberi and E. 

regnans (from Tasmania) and E. fastigata respectively. Quarter-sawing strategies to produce slabs 

dimensioned only in thickness were applied to E. regnans in Victoria. The slabs were ripped to final 

board sizes after drying. With exception of E. regnans from Victoria, the production was conventional 

appearance grade products suitable for flooring or furniture, and tile battens and pallet grade boards. A 

dual strategy where both F17 structural products and appearance grade products was employed for the 

Victorian E. regnans. 

Figure 4. Industry graders grading dried and skip dressed boards from E. sieberi.



Table 1. Log sample coupe locations, year of establishment and thinning and log diameter, length and grade 

Species Location Year 

Established 

Year 

Thinned 

Log length 

(m) 

Log diameter 

(cm) 

Log grade 

E. fastigata Nalbaugh SF, NSW 1956 1994 4.9 52.2 SFNSW Quota grade 

E. fastigata Cathcart SF, NSW Unknown Unthinned 4.9 54.0 SFNSW Quota grade 

E. diversicolor Pemberton, WA 1972 1994 6.0 40.4 WA FPC Grade 1 

E. diversicolor Pemberton, WA 1974 Unthinned 6.0 36.9 WA FPC Grade 1 

E. diversicolor Pemberton, WA Unknown Unthinned 6.0 35.7 WA FPC Grade 1 

E. regnans Central Highlands, Vic 1971 1990 5.4 52.3 VicForests B, B/C and C 

E. regnans Central Highlands, Vic 1939 Unthinned 5.4 50.0 VicForests B, B/C and C 

E. regnans Styx Forest Block, Tas 1946 1970 5.4 / 4.9 45.7 FT Category 3 

E. regnans Styx Forest Block, Tas 1934 Unthinned 5.4 / 4.9 49.3 FT Category 3 

E. sieberi Jirrah Block, Orbost, Vic 1965 1991 4.5 49.1 VicForests B and C 

E. sieberi Jirrah Block, Orbost, Vic Mature & 1957 Unthinned 4.5 59.7 VicForests B and C 

Table 2. Log sample locations, year of thinning, processing methods and final products 

Species Location Established Thinned Sawing Drying Final product thickness /grade 

E. fastigata Nalbaugh SF, NSW 1956 1994 Quarter-sawn Predrier / kiln 19 mm / flooring and tile battens 

E. fastigata Cathcart SF, NSW Unknown Unthinned Quarter-sawn Predrier / kiln 19 mm / flooring and tile battens 

E. diversicolor Pemberton, WA 1972 1994 Back-sawn Air dry / kiln 21 and 25 mm / appearance and pallet 

E. diversicolor Pemberton, WA 1974 Unthinned Back-sawn Air dry / kiln 21 and 25 mm / appearance and pallet 

E. diversicolor Pemberton, WA Unknown Unthinned Back-sawn Air dry / kiln 21 and 25 mm / appearance and pallet 

E. regnans Central Highlands, Vic 1971 1990 Quarter-saw Predrier / kiln 38 mm / appearance and F17 

E. regnans Central Highlands, Vic 1939 Unthinned Quarter-saw Predrier / kiln 38 mm / appearance and F17 

E. regnans Styx Forest Block, Tas 1946 1970 Quarter-sawn Air dry / kiln 24 mm / appearance grades 

E. regnans Styx Forest Block, Tas 1934 Unthinned Quarter-sawn Air dry / kiln 24 mm / appearance grades 

E. sieberi Jirrah Block, Orbost, Vic 1965 1991 Back-sawn Predrier / kiln 19 mm / flooring and tile battens 

E. sieberi Jirrah Block, Orbost, Vic Mature & 1957 Unthinned Back-sawn Predrier / kiln 19 mm / flooring and tile battens



From data collected during processing a number of wood quality indicators were calculated. These 

included the recovery of all boards after docking end-splits, the grade recovery, the percentage of the 

volume of standard grade and better boards



Table 4. Log sample means for the calculated processing, feature and recovery and product value variates (bold = significant differences between 

thinned and unthinned samples at p≤0.05). 

Species 

Thinned 

Surface check (mm m -2 ) 

Processing characteristics Feature characteristics Recovery and product value 

# internal checks : # boards 

Spring (mm m -1 ) 

End split docked: board volume 

Knot (m m -1 ) 

E. fastigata 1994 82 0.078 0.58 0.064 0.14 0.63 0.10 21.4* 8.8* 15.3 268 

E. fastigata Unthinned 73 0.021 0.63 0.053 0.17 0.80 0.30 20.8* 3.8* 16.7 208 

E. diversicolor 1994 15 0.000 3.34 0.056 0.02 0.35 0.35 27.6** 15.5** 23.8 170 

E. diversicolor Unthinned 43 0.000 2.86 0.038 0.05 0.40 0.44 27.6** 11.2** 23.4 141 

E. diversicolor Unthinned 51 0.000 2.19 0.025 0.03 0.57 0.59 28.2** 7.9** 32.0 126 

E. regnans 1990 16 0.431 0.17 0.050 0.32 0.79 0.04 27.7** 8.8** 7.1** 0.9 279 

E. regnans Unthinned 41 0.844 0.18 0.080 0.36 0.87 0.05 25.8** 3.7** 11.4** 2.3 234 

E. regnans 1970 12 0.000 0.37 0.027 0.24 0.46 0.00 28.7** 13.9** 4.2 220 

E. regnans Unthinned 13 0.000 0.19 0.023 0.21 0.32 0.00 30.8** 19.3** 4.0 263 

E. sieberi 1991 347 0.151 0.77 0.059 0.54 0.87 0.86 29.1** 1.0** 9.4 278 

E. sieberi Unthinned 738 0.113 0.72 0.089 0.53 0.99 0.95 31.1** 0.0** 5.1 291 

* Recovery based on final product size 

** Recovery based on nominal dry dimensions after skip dressing 

Kino (m m -1 ) 

Insect damage (m m -1 ) 

Recovery all grades (% log 

volume) 

Recovery of select grade (% log 

vol) 

Recovery of F17 (% log vol) 

Standard grade +



Spring 

There were significant differences in spring for E. regnans from Tasmania and E. diversicolor. In both 

cases it was worse in the thinned regrowth. However, spring was generally minor and not value 

limiting. For example, in the case of the thinned E. regnans from Tasmania the calculated mean spring 

of 0.37 mm m -1 equates to approximately 1.6 mm of spring for a 5.0 m long board. Spring of this 

magnitude was not grade limiting. Spring was much worse in E. diversicolor than any other species, 

because the long log length, and relatively small diameter of the logs, accentuated the effect of growth 

stress release. In this case spring was reduced to acceptable levels for strip flooring during processing. 

End-splits 

The only significant difference in the volume docked to eliminate end-splits : original board volume 

was for E. diversicolor. The thinned sample had significantly more end-splits. The ratio can be 

expressed as a percentage of log volume by multiplying by 100 and then by the recovery of all grades 

(Table 4). In this case the ratio of 0.056:1 indicates that less than 1.6% of log volume was lost due to 

docked end-splits indicating that it was a minor value limiting defect. 

Knots, kino and insect damage 

There were no significant differences for knots. There were significant differences in kino in all 

species by region and there were significant differences in insect damage for E. fastigata and E. 

diversicolor. Generally kino and insect damage were less common in the thinned samples. The only 

case where this did not occur was for kino from the thinned E. regnans from Tasmania. 

Kino and insect damage were the major value limiting defects observed. These defects limited board 

grade to standard grade, utility grade, pallet or F17, depending on the grading strategy applied. As a 

consequence while overall recoveries tended to be similar for each species by region the product value 

per cubic metre of log input was generally higher for the thinned samples. 

The only case where this did not occur was with the E. regnans from Tasmania, where many boards 

were limited to standard grade because of the presence of kino. However, it is important to note that of 

the 11 log samples processed in this series of trials, the incidence of kino in the thinned Tasmanian E. 

regnans was the 4 th lowest recorded, and in comparison to both the thinned and unthinned E. regnans 

samples from the comparable Victorian trials, was a much scarcer defect (0.460 m m -1 versus 0.792 m 

m -1 and 0.875 m m -1 for the thinned and unthinned samples respectively). This indicates that kino vein 

occurrence may have been low in the unthinned Tasmanian sample by the overall E. regnans forest 

standards. 

Despite the lower product values from the thinned Tasmanian E. regnans logs, there appears to be a 

considerable advantage from thinning in terms of sawn timber value MAI ($ ha -1 year -1 ) and pulpwood 

yields that more than compensates the lower sawn product values and extra harvesting costs. Estimates 

from harvest records in the two coupes selected for this work found that the sawn timber value MAI 

was 191 ($ ha -1 year -1 ) and 880 ($ ha -1 year -1 ) for the unthinned coupes and thinned coupes 

respectively. At the same time MAI of pulpwood production was 5.0 (m 3 ha -1 year -1 ) and 6.1 (m 3 ha - 

1 -1 

year ) from the unthinned and thinned coupes respectively. 

Differences in tree dominance and kino and insect damage 

By ensuring that the trees were selected so that the logs were as close as possible to the same diameter 

for the thinned and unthinned regrowth, there remains a question about the differences in tree 

dominance and its possible association with defect occurrence. This may account for some of the 

differences in kino and insect damage. However, for this evaluation, which was interested only in the 

impact of differences in growth rates on processing performance and product quality in conventional 

sawmills, this was always a compromise that had to be accepted. It would therefore be useful to obtain 

further information on the occurrence of these defects in both thinned and unthinned regrowth. 

CONCLUSION 

While more work is needed to understand differences in kino and insect damage, this series of trials 

gave a clear indication that the existing processing industry would not be disadvantaged by processing 

logs from thinned regrowth of conventional sawlog size and quality.



Faster growth after thinning did not contribute to any serious decline in wood processing performance 

and there is some evidence of improvements in wood drying performance, possibly due to more 

uniform wood properties. 

Given increases in sawlog yields that may be expected from thinning in even age stands these results 

support these thinning strategies in production forests. 


This research was funded by the FWPA, CSIRO, VicForests, Forests NSW, Forestry Tasmania and the 

WA Forest Products Commission as part of the FWPA PN06.3015. Processing was conducted by Blue 

Ridge Hardwoods, Eden, NSW; McCormack Demby Timbers, Morwell, Victoria; ITC Southwood and 

Launceston, Tasmania and Auswest Timbers, Pemberton, Western Australia. 

REFERENCES 

Fagg, P. (2006). Native Forest Silviculture Guideline No. 13. Thinning of Ash Eucalypt Regrowth. Department 

of Sustainability and Environment Victoria. 

Goodwin, A. (1990) Thinning response in eucalypt regrowth. Tasforests, 2(1): 27-35. 

Innes, T., Armstrong, M. and Siemon, G. (2005). The impact of harvesting age/tree size on sawing, drying and 

solid wood properties of key regrowth eucalypt species. Project No. PN03.1316, Forest and Wood 

Products Research and Development Corporation, Melbourne. pp 99. 

Kerruish, C.M. and Rawlins, W.H.M. (Eds.) (1991) The Young Eucalypt Report: Some Management Options for 

Australia’s Regrowth Forests. CSIRO, East Melbourne. 

O’Shaughnessy, P. J. and Jayasuriya, M. D. A. (1993). (Eds) Third progress report, North Maroondah, 

Melbourne Water, Report Number MMBW-0017. 

Wardlaw T.J., Plumpton B.S., Walsh A.M. and Hickey J.E. (2004) Tasforests Volume 15 June 2004. 

Washusen, R., Morrow, A., Linehan, M,. Bojadzic, M. Ngo Dung and Tuan, D. (2007a). The effect of thinning 

on wood quality and solid wood product recovery of regrowth forests: 1. E. fastigata from southern 

New South Wales. Ensis Client report no. 1770 for Forests and Wood Products Australia PN06.3015. 

May 2007. 35 pp. 

Washusen, R., Morrow, A., Siemon, G. and Ngo Dung (2007b). The effect of thinning on wood quality and solid 

wood product recovery of regrowth forests: 2. E. diversicolor from southwest Western Australia. Ensis 

Client report no. 1821 for Forests and Wood Products Australia PN06.3015. 

Washusen, R., Morrow, A., Ryan, M. and D. Ngo (2007c). The effect of thinning on wood quality and solid 

wood product recovery of regrowth forests: 3. E. regnans from Central Highlands Victoria. Ensis Client 

report no. 1828 for Forests and Wood Products Australia PN06.3015. 

Washusen, R., Morrow, A., Wardlaw, T. and D. Ngo (2007d). The effect of thinning on wood quality and solid 

wood product recovery of regrowth forests: 4. E. regnans from Southern Tasmania. Ensis Client report 

no. 1827 for Forests and Wood Products Australia PN06.3015. 

Washusen, R., Morrow, A., Ryan, M. and D. Ngo (2007e). The effect of thinning on wood quality and solid 

wood product recovery of regrowth forests: 5. E. sieberi from Eastern Victoria. Ensis Client report no. 

1826 for Forests and Wood Products Australia PN06.3015. 

Webb, A. W. F. (1966). The effect of thinning on the growth of Mountain Ash in Victoria. MSc thesis, Forest 

Commission, Victoria. 

West, P.W. (1991). Thinning response and growth modelling. In Kerruish C. M. and Rawlins W. H. M.editors 

(1991) The young eucalypt report - some management options for Australia's regrowth forests. CSIRO 

Australia, pp 28-49.



DIESEL-ELECTRIC HYBRID DRIVE TECHNOLOGY FOR 

REDUCED FUEL CONSUMPTION AND CARBON EMISSIONS 

IN FOREST OPERATIONS 

Mark Brown 1 

ABSTRACT 

The forest industry is well placed to play a positive role in climate change through the 

management of forests as carbon sinks. However, forest operations generate significant 

carbon emissions. Therefore forest companies have the opportunity to play a positive role 

in climate change by reducing their carbon emissions by implementing environmentally 

friendly technology and work methods. 

The CRC for Forestry has partnered with a developer of retrofit hybrid-drive technology, 

Innovative Transport Solutions Pty Ltd (ITS Electric), to develop and test diesel-electric 

hybrid conversions for forestry applications. The overall aim is a 10-25% reduction in 

fuel consumption for a reduction in operating costs and carbon emissions. The first trial 

vehicle, a 6X6 truck used to move chip trailers around forest harvesting sites, will be 

tested under field conditions in Western Australia during the last half of 2009. 

This paper discusses why hybrid technology is believed to have significant potential for 

reducing fuel use and carbon emissions for some areas of forest operations; provides an 

overview of the unique conversion approach to create hybrid drive vehicles; and 

describes the field trial that will be conducted over the next two years. 

INTRODUCTION 

Forest operations are energy intensive with harvest and haulage methods being responsible for the 

bulk of the energy consumed. Commonly used forest harvesting equipment in Australia consumes up 

to 40 L of fuel per hour and the typical haulage trucks use up to 80 L/100 km. The use of such a large 

amount of fuel in forest operations is of concern with its increasing cost, but it also has important 

carbon emission implications. In Sweden, where harvesting systems are among the most fuel efficient 

in the world with fuel consumption between 10 and 25 L/hr, studies have shown CO2 emissions of 

about 5.9 t per m 3 of wood harvested and over 6.7 t per m 3 for haulage (Berg & Lindholm, 2005). 

In an effort to further advance the positive role of forestry in the climate change issue the industry is 

actively exploring options to improve the fuel efficiency of their harvesting and haulage operations. 

This interest is highlighted by handbooks of best practice, such as ‘In Forest Operations Fuel 

Economy Counts’ from Canada, which focus entirely on what can be done both by equipment 

selection and by adjusting work techniques to reduce the use of fuel. While these handbooks are 

popular and successful in achieving outcomes in Canada, it is a fact that up to 80% of the fuel use is 

influenced by machine and engine design. This highlights the need for a continued drive for more 

efficient machine and engine designs. 

HYBRID TECHNOLOGY 

Hybrid technology is the use of more than one power source in a vehicle’s drive or power system that 

allows for the optimisation of the two power sources over the work cycle for minimum energy use. 

Many examples of hybrid technologies have the added advantage that one or more of the power source 

technologies have the capability of recovering and storing energy when not providing power, either 

through the normal function of the operations (moving down hill, braking, etc.) or by recovering 

excess energy output. Without the energy recovery aspect of hybrid technology, optimisation between 

power sources can only offer very limited efficiency gains and in many cases none at all. This is due 

1 University of Melbourne, CRC Forestry, Harvesting and Operations. Email: mwbrown@unimelb.edu.au



to the difficulty of effectively optimising the power sources over complex work cycles (Lin et.al., 

2003). As a result the application of hybrid technologies which offer the greatest efficiency gains are 

those used in work cycles that have frequent and large variations in power load. In the transport 

industry these are operations with frequent starts and stops, as highlighted by Langer (2004) where 

medium duty trucks are projected to be able to increase fuel economy by 71% due to the start and stop 

nature of the work. 

The two most common hybrid technology combinations are an internal combustion engine (ICE) with 

a battery powered electric motor, and an ICE with an accumulator powered by a hydraulic motor. The 

characteristics of the hydraulic hybrid system are a limited capacity to store energy, and best operation 

with a quick release of the stored energy. Due to these characteristics, the hydraulic hybrid system is 

only practical in a limited number of applications where the power input from the hydraulic power 

system is relatively short and regeneration of the hydraulic power is sufficient between power input 

demands. Electric hybrid technology tends to be more versatile, with an increased capacity to respond 

to more frequent and/or sustained power demands depending on the size and type of battery used. In 

part, this increased flexibility has led to this project’s exploration of electric hybrid technology. 

The development of electric hybrid technology has largely focussed on integrated drive systems where 

the electric drive and regenerative system is integrated with the transmission system. This integrated 

approach offers the highest level of control to return the highest possible gains in efficiency from the 

hybridisation. However, this integrated approach also means that retrofitting the technology to existing 

ICE is either impossible due to a lack of space and/or cost prohibitive. Therefore the introduction of 

the electric hybrid technology would be limited to purchasing new equipment and this would involve a 

relatively high cost. With forestry vehicles having a potential working life up to 15 years and 

companies trying to minimise purchase costs, the realisation of the potential efficiency gains would 

therefore be slow. 

The Australian company, Innovative Transport Solutions Pty. Ltd. (ITS Electric), has developed an 

electric hybrid approach which can be applied to many ICE applications as a conversion. Like other 

electric hybrid technology, the ITS approach adds an electric motor to the ICE to maximise the 

efficiency of both and uses the electrical system and battery to recover energy for further efficiency 

gains. The main difference in this case is that the electric motor is introduced directly at the axle 

without interfering with the traditional drive system and the hybridization is controlled through an 

independent external controller. 

This approach means that while there are limitations in the level of integration which can be achieved 

with the ICE drive system, it also means most ICE drive systems can be retrofitted in a cost effective 

way, with projected payment periods for many truck applications of less than three years. 

POTENTIAL APPLICATION OF THE HYBRID CONVERSION TECHNOLOGY 

IN FORESTRY OPERATIONS 

As stated earlier, hybrid technology is most efficient in operations with frequent and large variations in 

power demand. In transport operations, some of the best results with hybrid technology have occurred 

where there are a large number of starts and stops in a normal work cycle, such as urban garbage 

collection, local delivery trucks and city busses. 

Blohm and Anderson 2004 looked at the potential of hybridisation in garbage trucks in the United 

Kingdom. Through GPS tracking they identified that, in an average day’s work, the trucks stopped 459 

times and spent over 40% of the time idling. They identified fuel economy benefits of 30-40% 

depending on the type of hybridization used. A US company, United Parcel Service (UPS), which 

operates a large fleet of delivery vehicles, announced in a 2007 press release that they would continue 

their the expansion of hybrid vehicles in their fleet with an additional 50 hybrid trucks, and expected 

to save about 166,500 litres of fuel per year, based on their experience since 2001 with hybrid 

vehicles. 

Much of the equipment used in forest operations has frequent and large variations in power demand, 

including skidders, forwarders, shunt trucks, chippers, haul trucks, feller-bunchers and harvesters. All 

of this equipment may benefit from an adapted hybrid system.



When the work cycle of forestry primary transport equipment is examined most tend to cover short 

distances, frequently start and stop, and often have large amounts of idle time or low power demands. . 

A series of short operational trials carried out by the CRC for Forestry in Western Australia examined 

the basic productivity of harvesting equipment. Skidders working in plantation clearfelling operations 

have a minimum of 100 to 140 start/stops in a ten hour shift and spend only about 40% of the time 

under heavy power demand, travelling loaded. Similarly, forwarders have a minimum of 120 to 200 

start/stops in a ten hour shift and spend less than 25% of the time under heavy power demand 

travelling loaded. 

A desktop analysis of the work cycle of shunt trucks as it relates to chipper operations indicates a 

minimum of 90 to 140 start/stops in a ten hour shift, over 50% of the time idling, and only about 20% 

of the time pulling a full load that has a large power demand. 

Fuel savings of 10-25% are achievable for these three primary transport functions. For a skidder this 

represents up to 15,000 litres of fuel saved per year, and about half as much for the forwarder and for 

the shunt truck, since a forwarder is considerably more efficient than a skidder, while a shunt truck 

typically has a much smaller work load. 

In the longer term, or with more aggressive hybrid conversions, even greater savings are possible by 

using a smaller ICE in combination with the electric motor, as the electric system could support the 

limited amount of heavy power demand. 

Beyond these primary transport applications, a more detailed analysis of work cycles is required to 

explore the potential for other forestry machines and such equipment is not addressed in this report. 

With such applications, while there would be a limited start/stop type function for energy recovery, 

there would be an opportunity to recover energy at times of low demand in order to power an electric 

motor to support times of high demand. 

THE CRC FORESTRY INDUSTRY TRIAL 

After extensive discussion with ITS Electric and the forestry industry, an initial trial has been planned 

with shunt trucks. Significant effort had already been put into designs to fit class 8 trucks with hybrid 

technology that would directly apply to shunt trucks, since shunt trucks are a version of a class 8 truck. 

Being able to directly apply this existing knowledge and design to the shunt truck application will 

minimize risk and cost. 

From the forest industry perspective, while shunt trucks have limited application, the 6X6 trucks 

required are difficult to get affordably on the used truck market, where they are typically sourced, due 

to competition from the mining industry. As a result the approach of purchasing a more readily 

available used class 8-truck and converting it to 6X6 with the electric hybrid conversion technology 

can be done for about the same cost as the traditionally used 6X6 trucks, thereby reducing any 

associated risks and costs. 

From an operational trial perspective, the use of the shunt truck as the “proof of concept” for this fuel 

saving technology is very attractive since the work cycle of the shunt truck is highly repetitive, the 

operating conditions are relatively easy to control and there is frequent operator change over. 

Considering: 

o the high level of repetition, 

o the opportunity to control certain operating conditions, and 

o the ability to eliminate or at least identify operator influences, 

the project will be able to evaluate the performance of the shunt truck effectively and isolate, with 

reasonable confidence, the influence of the hybrid technology without the high cost of highly 

controlled bench tests that are often required to measure fuel impacts.



As shunt trucks have been identified as a good approach for all three collaborators, the trial design has 

four main objectives: 

1. To determine if the hybrid conversion concept is practical and functional. 

2. To test the functional capabilities of the converted hybrid 6X6 shunt truck against 

currently acceptable 6X6 diesel trucks to ensure it can perform the task. 

3. To quantify the fuel efficiency gains and emission reduction potential of the technology in 

a 6X6 shunt truck application. 

4. To evaluate the short to medium term reliability and costs of the hybrid conversion 

technology. 

The evaluation of the hybrid conversion technology will be done in four distinct phases over a 24 to 30 

month period of time. 

o Phase one will look at the costs of the vehicle and its setup as the basis for economic 

comparisons. 

o Phase two will include some specific comparative functional tests on turning and pulling 

capabilities to ensure the hybrid truck is able to complete the required operational tasks 

o Phase three will be a detailed evaluation of operating cost, performance and reliability on a 

short to medium term to complete the economic evaluation as well as environmental impacts. 

o Phase four will be a medium to long term evaluation of operating costs and reliability to 

better understand maintenance costs. 

To evaluate the comparative initial costs of the hybrid conversion against the traditional 6X6 truck in 

Phase one, all components and their cost for the conversion to hybrid will be recorded, including 

labour and customization by the forest company operator for safety and ease of operation. 

Considering that prototypes tend to cost as much as three times that of production models and, if the 

technology is successful, on a large scale there will be cost benefits to large volume production of 

certain components, estimates of future costs with sensitivity analysis will be done. All these initial 

and preparation costs will be compared against the cost of the last three to six 6x6 trucks acquired, 

based on availability of information from the cooperating forest company, including all preparation 

costs. This comparison will look at both simple purchase/setup costs and annual distribution of these 

costs based on projected lives for the two different trucks. 

With Phase two, functional tests to ensure the truck will be able to perform the required operational 

tasks without significant risk of failure or decreased safety will be carried out. The turning radius will 

be checked with controlled test methods in good (flat, dry asphalt), moderate (flat, loose gravel) and 

poor (sloped, very loose gravel) conditions. Pulling strength as compared to the traditional 6X6 truck 

will be examined with a controlled acceleration test where the distance and time to attain a defined 

operating speed from a dead stop will be compared under good (flat, dry asphalt), moderate (flat, loose 

gravel) and poor (sloped, very loose gravel) conditions with a loaded trailer. For both sets of 

controlled tests in phase two, each test condition will be repeated to ensure three results are obtained 

that are within +/- 5% of each other (to ensure the results are true and repeatable). 

During Phase three a detailed, side-by-side operational tracking of the hybrid truck as compared to the 

traditional 6X6 truck will be performed for six months. With the support of onboard tracking 

technology, shift level manual data records and detailed observation studies data will be collected on: 

• equipment availability, 

• utilization and effective time 

• ICE use and performance (idle time RPM profile, travel speeds, fuel consumption, etc.) 

• hybrid technology use and performance (use of electric motor, battery 

levels/charge time, etc.) 

• work or duty cycle of the truck



• short and long operational delays and root cause of the delays 

• fuel use 

• scheduled and unscheduled repair and maintenance 

• harvesting system results that influenced shunt truck use/performance 

• general operating conditions and operational productivity 

The data will be analysed to quantify the performance of the hybrid shunt truck relative to the 

traditional 6X6 trucks and clearly identify any impacts on operating cost, operating performance and 

fuel use. Particular emphasis will be given to identifying conditions where the hybrid demonstrates 

significant strengths or weaknesses. 

Phase four of the project will see the continued collection of similar data as for phase three, but at a 

lower data collection intensity for an additional 18 months, and will focus on longer term issues with 

performance and operating costs, such as maintenance and delays. 

CONCLUSION 

While the forestry industry has a positive role to play in climate change through carbon sequestration 

in the growth and sustainable management of forests, forest operations represent an opportunity for 

further improvement by reducing fuel use and emissions through the appropriate application of new 

technology. 

One such opportunity is the use of hybrid drive technology, as a large amount of the equipment used 

in forest operations have work cycles that match the conditions where hybrid technology has been 

most successful — a large amount of start and stop activity, with large periods of work time at idle or 

low power demand levels. 

With the more common integrated hybrid technology, its introduction would be limited to new 

equipment and even then, within the price sensitive forest industry, this would be slow, based on the 

current mechanical life of forest equipment and replacement cycles. As a result, the new “Made in 

Australia” conversion approach is an attractive alternative which would allow the introduction of the 

technology without equipment replacement and with relatively short payback periods. 

The CRC for Forestry will work with the technology developer and the Australian forest industry to 

adapt the hybrid conversion technology for shunt trucks in the forest industry. The project will seek to 

verify the base performance and any impact the technology has on the costs, efficiency and 

performance of the shunt trucks. If successful, the hybrid shunt truck will provide a cost effective, 

environmentally friendly technology option for the industry. In addition, the hybrid shunt truck will 

provide a good base for investment into hybrid conversion technology, and pave the way for its 

introduction to other forestry-related machinery, or the conversion of existing equipment. 

REFERENCES 

Berg S. and Lindholm E-L., 2005, ‘Energy use and environmental impacts of forest operations in Sweden’, 

Journal of Cleaner Production vol. 13, pp. 33-42. 

Blohm T. and Anderson S., 2004, ‘Hybrid Refuse Truck Study’ paper presented to Product Development 

Conference, Huntington Beach, California. 

Forest Innovation Partnership, In Forest Operations Fuel Economy Counts, Sainte-Foy, Quebec, Canada. 

Langer T., 2004, ‘Energy Savings Through Increased Fuel Economy For Heavy-Duty Trucks’, prepared for the 

National Commission on Energy Policy. 

Lin C-C., Peng H., Grizzle J.W. & Kang J-M., 2003, ‘Power Management Strategy for a Parallel Hybrid Electric 

Truck’, IEEE Transactions on Control Systems Technology, vol. 11 no. 6, pp. 839-849.



MODELLING MILL DOOR LOG PRICES FOR 

PLANTATION EUCALYPTS WITH CSIROMILL 

Russell Washusen 1 

ABSTRACT 

CSIROMILL incorporates financial models of sawmills, developed from processing 

experiments with the assistance of industry, to assist establishment of new plantation 

eucalypt resources. During a recent international project supported by the Australian 

Centre for International Agricultural Research (ACIAR), one module of CSIROMILL 

was used to model mill door prices of unpruned 9-year-old E. pilularis, processed with a 

HewSaw R200 PLUS linear sawmill. The results indicated that it would not be viable to 

grow the unpruned E. pilularis, despite the efficiencies of the linear flow sawmill. Further 

work was conducted with two additional modules of CSIROMILL using pruned 17-22 

year-old Corymbia spp. (spotted gum), processed with a HewSaw SL250 Trio linear 

sawmill; and, pruned 17-22 year-old spotted gum, processed with a conventional 

reciprocating flow eucalypt sawmill. The mill door prices generated demonstrate the 

importance of producing high quality logs. The modelling also gave an advantage in mill 

door log values of $90-120 per cubic metre with the linear system compared to the 

reciprocating system. The results demonstrate how CSIROMILL can be applied, and the 

importance of both log quality and sawmill efficiency in developing viable plantations for 

the production of sawlogs. 

INTRODUCTION 

Recent experiments with plantation-grown E. nitens, E. pilularis and Corymbia citriodora subsp. 

variegata indicated that the HewSaw R200 and R250 sawmills can be applied to saw small-diameter 

eucalypt logs (Washusen et al. 2007, 2008, 2009). The sawing technology incorporated into these 

sawmills is usually associated with the softwood industry where, unlike in eucalypts, longitudinal 

peripheral growth stresses pose few problems during processing. 

In each of the experiments it was found that, with the application of appropriate cutting patterns, the 

technology appears to be capable of processing small diameter eucalypts with a large range of growth 

stress levels. The main characteristic associated with growth stress release that may hinder processing 

efficiency was bow in boards cut from near the log periphery. However, the magnitude of bow was 

considered small enough to allow efficient transport of boards in the green mill if minor modifications 

were made to board conveyors. 

In comparison to conventional eucalypt processing, the HewSaw incorporates technology that 

produces fundamental differences in processing performance for eucalypt saw logs. These differences 

are: 

(i) Chippers are applied with small diameter circular saws during sawing in such a way that 

wood near the log periphery (assumed to be the most highly stressed wood) is removed 

and boards are profiled before or simultaneously with sawing. The removal of chip from 

near the log periphery eliminates the need to handle distorted ‘roundbacks’ in the mill, it 

tends to reduce board deflection (bow), and board and log end-splitting appears to be 

controlled if sufficient material is chipped from the log periphery (Washusen et al 2007, 

2008, 2009). 

(ii) End dogging of logs is eliminated which can reduce log end-splitting induced during the 

sawing process (Washusen and Innes 2007). 

1 Dr Russell Washusen, Principal Research Scientist, CSIRO Materials Science and Engineering, Project Leader 2.3, CRC 

for Forestry, Bayview Av, Clayton South, email: russell.washusen@csiro.au



(iii) The adoption of multi-saw (and chipper) technology allows linear flow of material. In 

contrast, conventional eucalypt sawmills, even those where twin saw systems are 

employed, require the saws to be moved repeatedly through the logs and flitches, or the 

logs and flitches to be passed back and forwards through a stationary saw (reciprocating 

flow). Either of these methods produces a relatively slow throughput of logs and flitches, 

and bottle necks in the processing line often occur. 

The potential for improved material flow, while maintaining or improving board behaviour, allows 

high log throughput rates to be achieved. For example, the most modern dedicated conventional 

reciprocating flow sawmills built in Australia have log throughput rates of around 25,000-45,000 m 3 

of log annum -1 in a single shift (pers. comm. Geoff Bertolini, Whittakers Timber Products, Western 

Australia; Ian Whiteroad, Integrated Tree Cropping, Tasmania). 

In comparison, single shift annual log throughput rates for sawing softwood with the HewSaw R250 

and HewSaw SL250 Trio PLUS are around 130,000 m 3 of log annum -1 and 200,000 m 3 of log annum -1 

respectively (pers. comm. John Marshall, Carter Holt Harvey, Victoria; Kennett Westermark, Veisto 

Oy, Finland). In comparison to conventional reciprocating systems, this improved material flow has 

the potential to reduce processing costs for eucalypts, with flow-on effects of improved returns to both 

processors and growers. 

In order to quantify the impact of improved material flow a HewSaw R200 PLUS system was 

modelled to predict mill door prices for 9-year-old unpruned E. pilularis with the CSIROMILL H 200 

PLUS 12000 AD BB module (Washusen et al. 2008). CSIROMILL is a modelling tool developed by 

CSIRO in partnership with industry to assist planning of new plantation-grown eucalypt resources. 

The primary function of CSIROMILL is to simultaneously model sawmill profitability and mill door 

prices using information collected from ‘real life’ experimental processing trials. 

The modelling of the unpruned E. pilularis indicated that processing could be profitable, although mill 

door prices were too low for profitable growing because of poor log quality. As a result of this work 

further modelling was undertaken with high quality sawlogs to further assess the potential mill door 

prices for eucalypts processed with linear sawing systems. For the purpose of comparing outcomes 

with conventional processing, comparisons were made with a modern reciprocating flow sawing 

system currently used to process small diameter native forest eucalypts. 

For this work, two additional modules of CSIROMILL were employed. CSIROMILL also has a data 

base of information obtained from numerous trials in sawmills across Australia. The data base has 

recorded log quality information for pruned eucalypts that can be readily produced by growers, and 

measured product recoveries and values matched to logs of a specified quality. This information was 

used to model mill door prices for pruned logs so that comparisons could be made between the 

unpruned E. pilularis and linear and reciprocating sawing systems processing larger diameter, pruned 

eucalypts. 

LOG QUALITY, RECOVERIES AND PRODUCT VALUES USED FOR MODELLING 

For modelling the following information was used. The pruned Corymbia spp (spotted gum) log Grade 

2 adopted by Washusen (2006) was used. The log specifications are given in Table 1. 

Table 2 gives the recoveries of wood chip (percent of log volume) and sawn product based on nominal 

dry dimensions (percent of log volume) for the pruned log Grade 2 for spotted gum. 

These values were obtained from trials conducted at the Auswest Timbers Pemberton sawmill 

(Washusen 2006). Grade recoveries met the requirements of the Australian Standard AS 2796.1 

(Timber-Hardwood-Sawn and milled products, Part 1: Product specification). The synonym for 

standard grade is medium feature grade in AS 2796.1. The logs were sourced from Boyup Brook and 

Vasse in southwest Western Australia and ranged in age from 17-22 years.



Table 1. Specifications for pruned spotted gum log Grade 2. 

Log characteristic Specifications 

Diameter of defect core < 15.0 cm 

Log mean diameter 30.1-35.0 cm 

Sweep in 2.4 m length*



Figure 1. Example sawing pattern for a 35 cm small end diameter log showing the sequence of 

cuts on the twin saw and resaw cuts. 

All wood prices used were AUD mill-door-wholesale prices free of GST. 

Prices for sawn boards were adapted from Washusen (2006) using the values of the actual board sizes 

recovered. Victorian ash prices were used to value sawn wood with a 50% discount for select and 

standard grade boards shorter than 1.8 m and a 10% discount for all other boards



THE CSIROMILL METHOD AND MODULES USED 

The CSIROMILL system and the methodology used for modelling are described by Washusen et al. 

(2008) and readers are directed to that report for background information. 

Two CSIROMILL modules were used for modelling (Table 4). One was a WH module and the other a 

HSL 250 module. These are described below. 

Table 4. CSIROMILL Modules and the processing systems they represent. 

Module System Annual log intake range 

(m 3 ) 

WH 30000 AD SG Reciprocating system with twin and 

multi-saws 

40,000-60,000 

SL 250 150000 AD SG Linear system with three 

chipping/sawing units in line 

250,000-350,000 

The WH Modules of CSIROMILL 

The WH modules of CSIROMILL are reciprocating flow systems representing modern twin-saw 

technology coupled with a downstream multisaw resaw and trim-saw systems. The reciprocation of a 

35 cm SED log during part of the sawing process through the actual twin saw represented in the 

module is shown in Figure 2. 

The WH modules are based on the Whittakers Timber Products small log line in Western Australia 

with some modifications. This is the newest hardwood mill in Australia having commenced 

production in 2006. The modules reflect actual construction costs of the mill (for most components). 

The operating costs have been sourced from a number of hardwood mills located in Western Australia, 

Victoria and Tasmania. 

Capital, operating costs, staffing levels and summary sheets of the CSIROMILL WH 30000 AD SG 

module are given in the output files in Washusen (2008). 

For the WH 30000 module, logs are cut to length either in the forests or the log yard. They are 

debarked on the sawmill in-feed, sawn with the twin bandsaw and multisaw resaw coupled with a 

trim-saw system. The twin saw is controlled by a computer. The logs are scanned and the computer 

determines the best sawing pattern and the sawing is then automatic. The overhead end-dogging 

system is equipped with a ‘turn-down’ device eliminating the need to release the log. The debarking 

and sawing system has capacity to process logs from 20 cm SED to 40 cm SED. 

Figure 2. Pruned eucalypt log being processed on the twin saw at the Whitakers Timber Product 

sawmill in Western Australia, showing the reciprocation of the log and log rotation to 

produce the log break-down pattern shown in Figure 1.



The log throughput range is 20,000-30,000 m 3 annum -1 for a single shift. Based on trials conducted in 

the Whittakers Timber Products mill, the actual capacity is approximately 25,000 m 3 annum -1 in a 

single shift (Washusen unpublished data). The sawing and material handling systems appear to have 

greater capacity than this, but log throughput is highly dependant on staff capability. The throughput 

rates tested in the analysis are 45,000-55,000 m 3 annum -1 for a two shift operation. 

Sawn residues are chipped and transported by conveyors to a stock pile or for loading directly into 

trucks. Bark and 50% of the sawdust is transported to bins and sold to local markets. The remaining 

sawdust is used green as fuel to heat kilns. 

Immediately after sawing, boards containing sapwood are segregated and treated with a standard 

boron diffusion treatment. 

Drying incorporates air-drying in sheds with drop down sides to control air flow. Following air-drying 

the wood is reconditioned in steam then kiln dried to a final moisture content of 10-12%. Heating for 

kilns is via pressurised hot water, heated from a furnace with sawdust and dry shavings being used as 

fuel. 

The analysis is applicable to the production of oversized skip dressed boards that are sold for further 

manufacturing into flooring. Each analysis assumes that the entire intake is of the logs specified. In 

reality there would be a range of log sizes and possibly more than one species. 

The SL 250 Modules of CSIROMILL 

The SL 250 modules are based on the HewSaw SL250 Trio sawing line. Figure 3 shows the linear 

configuration of the separate machines that make up the sawing line. Figure 4 shows an example 

sawing pattern produced during the sawing process. The sawing strategies are different to those 

produced by the HewSaw R250 and R200 in that the sawing is not completed within the one machine. 

However, the rip saw produces a similar cutting pattern to the R250, and the chipping prior to 

application of these saws produces the same stress release. 

The separation of the machines in the HewSaw SL250 sawing line has several advantages over the 

HewSaw R250 and R200. It allows the cant to be turned so that the sawing produces back-sawn 

boards. The cant saw recovers either 1+1 or 2+2 side boards that are not recovered by the HewSaw 

R250 and R200. This results in higher recovery in larger diameter logs and in pruned logs these boards 

would be produced from the clearwood zone. The minimum log length that can be sawn on the SL250 

Trio is also 2.4 m compared to 3.0 m for the HewSaw R250. This gives greater capacity to reduce 

board deflection by reducing log length, and to reduce the impact of log taper on recovery. Because of 

these features, in comparison to the HewSaw R250 and R200, the SL 250 Trio is better suited to 

processing pruned eucalypts. 

The SL 250 modules are based on a number of mills in Finland, as no mills of this type exist in 

Australia. Two examples of Finnish mills are shown in Figure 5. One of the features of the modules is 

the incorporation of tray sorters on the outfeed from the saw for the four or eight outer boards (Figure 

5) and bins for the centre boards. Tray sorters can be set up so they pick up boards from specific parts 

of the sawing pattern. As these boards are the same dimension for logs of similar diameter, sorting of 

boards is not required. Transport distances can also be reduced and some elevators eliminated. All of 

these features can be potentially beneficial to transporting and sorting hardwood boards that may be 

unstable to transport on conventional softwood conveying systems (pers. comm. Kennett Westermark, 

Viesto Oy, Finland). 

As with the WH modules, actual capital and construction costs applicable to Australia have been used 

using a conversion rate of € 1 = AUD 1.69, where capital costs were supplied to CSIRO in Euros. The 

operating costs have been sourced from a number of mills located in Victoria and Tasmania, with the 

maximum cost used for calculations.



Figure 3. The configuration of the HewSaw SL250 Trio line (adapted from www.hewsaw.com). 

Capital, operating costs, staffing levels and summary sheets of the CSIROMILL SL 250 150000 AD 

SG module used for modelling are given in the output files in Washusen (2008). 

The module represents a mill where logs are scanned and merchandised in the log yard and debarked 

on the sawmill in-feed. The sawing system incorporates a HewSaw SL 250 PLUS trio sawing line 

with scan and set capability and a ring reducer immediately after the debarker. The debarking and 

sawing system can process logs from 2.4 to 6.0 m in length and from 10 cm SED to 40 cm SED and 

50 cm LED. Line speed is 60 to 150 m min -1 . Side boards are handled with a tray sorter and stacker, 

and there is a separate stacker for centre boards. 

Figure 4. Pattern produced by the HewSaw SL250 Trio line (adapted from www.hewsaw.com).



Figure 5. The HewSaw SL250 sawing line at United Sawmills Ltd., Pori, Finland (left) and a 

tray sorter on the outfeed of the HewSaw R250 at Havesa Timber, Hamina, Finland 

(right). 

In normal processing of softwood in Finland (Norway spruce and Scots pine), the HewSaw SL 250 

Trio has a maximum log throughput rate of 220,000 m 3 annum -1 for a single shift. In the CSIROMILL 

module the capacity has been reduced for spotted gum to a range of 125,000 to 175,000 m 3 annum -1 

for a single shift operation. This is to take into account log volume, higher wood density, increased 

board deflection and shorter logs (shorter boards), all of which will slow down material flow rates. 

The throughput rates modelled in the analysis ranged from 250,000 to 350,000 m 3 annum -1 for a two 

shift operation (approximately 60-80% of the maximum log throughput rate for softwood). 

Sawn residues are chipped and transported by conveyors to a stock pile or for loading directly into 

trucks. Bark and 50% of the sawdust is transported to bins and sold to local markets. The remaining 

sawdust is used green as fuel to heat kilns. 

Immediately after sawing, boards containing sapwood are treated with a standard boron diffusion 

treatment. 

Drying incorporates air-drying in sheds with drop down sides to control air flow. Following air drying 

the wood is reconditioned in steam then kiln dried to a final moisture content of 10-12%. Heating for 

kilns is via pressurised hot water, heated from a furnace with sawdust and dry shavings used as fuel. 

The modelling is applicable to the production of oversized skip dressed boards that are sold for further 

manufacturing into flooring. Each model assumes that only the entire intake is of the logs specified. In 

reality there would be a range of log sizes and possibly more than one species. 

RESULTS 

Mill door prices for three annual log intake rates for both mills and IRR’s for the mills of 10, 15 and 

20% are given in Table 5 and Figure 6. Taxation rate on profit was 30%. Detailed output files for an 

example analysis for the SL 250 150000 and WH 40000 modules are given in Washusen (2008). Table 

5 also gives the Mill door prices for E. pilularis from the H200 PLUS 120000 module from Washusen 

et al. (2008). 

For 50,000 and 300,000 m 3 throughput, for the WH 30000 and SL 250 150000 modules respectively, 

the mill door prices were approximately AUD 90-110 per cubic metre higher for the SL 250 150000. 

The difference in log prices was due to reduced costs in all sectors of the mill, with greatest reductions 

in the sawmill. Costs attributed to sawing were AUD 14 and AUD 52 per cubic metre of log input for 

the linear and reciprocating system respectively. 

The results indicate that, on a comparable basis, there is an improvement in mill door prices with 

improved log quality and with the application of linear sawing systems for a species that can produce 

back-sawn boards that can be dried readily. Therefore, for such species there is considerable potential 

to improve returns to growers with the adoption of pruning to produce high quality sawlogs, and 

through the application of linear sawing systems. For this potential to be realised, resources of the



appropriate size need to be developed within an economically viable supply zone. For plantations 

producing 10 m 3 ha -1 annum -1 of sawlogs this would require a total plantation area of 25,000 – 30,000 

ha and 4,500 – 5,500 ha for the linear and reciprocating system respectively. For the linear system, it 

may be difficult to achieve this plantation area in a single species. However, the mill design has 

capacity to process softwood and eucalypt logs in batches. This was the approach taken by FEA, 

processing Pinus radiata and E. nitens with the HewSaw R200. This would require a modest 

plantation area that could be expanded given good economic returns to the mill and growers. In 

southern Australia, there are already in existence a number of softwood mills processing below their 

capacity that could take advantage of appropriate resources. However, the plantation area available is 

very small and it may be several years before there are serious attempts to process large volumes of 

high quality pruned eucalypts in linear sawing systems. 

Table 5. Mill door prices (in bold AUD per cubic metre of log intake) for pruned spotted gum 

estimated with CSIROMILL modules WH 30000 AD SG and SL 150000 AD SG and 

unpruned E. pilularis estimated with CSIROMILL module H200 PLUS 120000 AD 

BB. 

WH 30000 log intake (m 3 annum -1 ) 

Logs processed IRR (%) 55,000 50,000 45,000 

Pruned spotted gum 10 123 111 95 



SL 250 150000 log intake (m 3 annum -1 ) 





H 200 PLUS 120000 log intake (m 3 annum -1 ) 


Unpruned E. pilularis 10 42 35 26 



CONCLUSIONS 

This paper estimates mill door prices for pruned 30-35 cm small end diameter (SED) plantation-grown 

spotted gum logs modelled using the WH 30000 ADSG and HSL250 150000 ADSG modules of 

CSIROMILL. These modules represent the state of the art reciprocating and linear sawing 

technologies that are capable of processing eucalypt logs within the SED range of 20-40 cm. 

Variables tested were a 30% taxation rate, IRR’s of 10, 15 and 20%, product values of $295 per cubic 

metre of log input and log input ranges of 45,000-55,000 m 3 annum -1 and 250,000-350,000 m 3 annum -1 

for the WH 30000 ADSG and SL 250 150000 AD SG modules respectively. 

Mill door prices were approximately AUD 90-100 higher for the SL 250 150000 ADSG module 

(linear system) for the most likely log throughput rates for both systems. This difference in log prices 

was due to reduced processing costs in all sections of the mill, with the largest reductions attributed to 

the sawmill (AUD 52 to AUD 14 per cubic metre of log intake). 

The Mill Door prices were also much higher from both systems than those reported for unpruned and 

unthinned E. pilularis mill door prices modelled with the H 200 PLUS 150000 module of 

CSIROMILL in earlier work.



The results indicate considerable potential to improve returns to growers with the adoption of both 

pruning to produce high quality sawlogs, and the application of linear sawing systems. For this to be 

realised resources of the appropriate size would need to be developed within an economically viable 

supply zone. 

REFERENCES 

Washusen, R. (2006). Evaluation of product value and sawing and drying efficiencies for low rainfall hardwood 

thinnings. A report for the RIRDC/Land and Water Australia/FWPRDC/MDBC Joint Venture 

Agroforestry Program. Project No CSF-65A. pp 52. 

Washusen, R. and Innes, T. C. (2007). Processing plantation eucalypts for high value timber. In proceedings of 

the Plantation eucalypts for high value timber conference, 9-12 October 2007, Moorabin, Melbourne. 

pp 91-108. 

Washusen, R., Morrow, A., Ngo, D,. Bojadzic, M., Henson, M., Porada, H., Northway, R., Boynton, S., 

Chen Shaoxiong, Peng Yan, Nguyen Quang Trung and Bui Chi Kien. (2007). Genetic variation in 

growth stress related wood behaviour of small diameter Eucalyptus nitens logs processed with a 

Hewsaw R250 sawmill. Report for ACIAR project FST/2001/021: Improving the value chain for 

plantation-grown eucalypts in China, Vietnam and Australia: sawing and drying. Client Report No 

1799. pp 39. 

Washusen, R., Morrow, A., Dung Ngo, Northway, R., Boynton, S. and Henson, M. (2008). Processing air and 

kiln dried sawn wood from unthinned and unpruned 9-year old Eucalyptus pilularis from northern New 

South Wales with a HewSaw R200. Report for ACIAR project FST/2001/021: Improving the value 

chain for plantation-grown eucalypts in China, Vietnam and Australia: sawing and drying. CSIRO 

Material Science & Engineering Client Report No CMSE(C)-2008-300. pp 50. 

Washusen, R. (2008). A comparison of mill door log prices for pruned Corymbia spp processed with 

conventional reciprocating and linear flow sawing systems. Report for ACIAR project FST 2001/021. 

Improving the value chain for plantation grown Eucalypts in China, Vietnam and Australia: sawing and 

drying. CSIRO Materials Science & Engineering Client Report No: CMSE (C)-2008-301. pp 42. 

Washusen, R., Morrow, A., Dung Ngo, Harding, K. and Innes, T. (2009). Processing air and kiln dried sawn 

wood from unthinned and unpruned 7.5 year-old Corymbia citriodora subsp. variegata from 

Queensland with a HewSaw R200. Report for ACIAR Project FST 2001/021. Improving the value 

chain for plantation grown Eucalypts in China, Vietnam and Australia: sawing and drying. CSIRO 

Materials Science & Engineering Client Report No: CMSE-2009-046. pp 43.



REDUCED FUEL USE IN FORESTRY TRANSPORTATION 

THROUGH THE USE OF HIGHER PRODUCTIVITY VEHICLES (HPV) 

Mark Brown 1 

ABSTRACT 

In transportation, payload is critical to vehicle efficiency and economics. The focus of 

improvements in transportation has mainly been on the economics of improved payload, 

but fuel usage for a given transport task is equally important. 

In the load-constrained context of transport on public road networks, potential gains in 

payload must be achieved by reduced tare weight or the introduction of new 

configurations that have higher load limits. With the introduction of performance based 

standards (PBS) in Australia, the forest industry is exploring the opportunity to introduce 

new high productivity vehicles (HPV). 

From an analysis of weighbridge data from six forestry companies, a potential 13% 

reduction in fuel use has been identified through lower tare weight specifications. A 

further 71% reduction may be possible through new HPV designs. 

INTRODUCTION 

Transportation in forestry has the draw back that a significant cost is involved in the delivery of wood 

and it users a large amount of energy. Studies in Sweden have shown that transportation represents 53 

to 56% of the total energy used in the production and delivery of timber (Berg and Lindholm, 2005). 

Similarly, in western Canada the Forest Engineering Research Institute of Canada (FERIC) found 

haulage accounts for 51% of the total energy used in forest operations (Sambo, 2002). While 

operations in Australia certainly differ from those in Sweden and Canada, the fact that similar results 

are reported across a broad geographical area, and that the equipment used is similar to that used in 

Australia, suggests that transportation is likely to be a significant proportion of total energy use in 

Australian forest operations. 

As fuel use in truck transport is of interest outside of forest transportation, there is no shortage of 

technology and training available to help reduce fuel use. While many solutions offer significant 

opportunity for reducing fuel use, they often come at a high cost and/or are difficult to implement 

effectively. Solutions typically involve improving technology so that the same level of work is 

achieved with lower energy use, or by adapting work methods to reduce “wasted energy”. 

An alternative approach is for the technology to do more work with the same or a nominal increase in 

energy usage in order to achieve an overall reduction of energy to complete a freight task. An 

example of this approach was demonstrated by FERIC when examining the potential to reduce fuel 

consumption by using large off-highway trucks. In the study FERIC found that an off-highway truck 

that consumed almost twice as much fuel as the normal haulage truck on a l/100km basis actually used 

36% less fuel on a l/tonne transported basis to complete the same freight task (Michaelsen, 2007). 

While the introduction of very large off-highway trucks is not feasible across many forest operations 

in Australia, the concept of doing more work with nominally the same or even more energy is 

applicable here. As most wood in Australia is transported via public roads for part of the journey from 

forest to mill, the total load or gross vehicle weight (GVW) for a given vehicle configuration is limited 

by public road restrictions. In some cases the actual vehicle configuration that is accepted on the route 

will also be limited. 

Until recently these vehicle limits have been very restrictive with most Australian states allowing only 

two or three configurations which fall within the GVW restriction of 42 to 62.5 tonnes. Consequently, 

1 University of Melbourne, Program Leader, CRC Forestry, Harvesting and Operations



to increase the vehicle work output under these restrictions, the empty or tare weight needs to be 

reduced, to allow an increase in payload without exceeding the GVW. In Case Study 1 (below) data 

from six Australian forest companies are examined to see how well tare weights are managed, and 

what opportunities exist for improvement. 

There has recently been a change in the Australian road rules for vehicle configurations, with a shift 

from prescriptive standards (predefined configurations like B-doubles) to performance based standards 

(PBS). As a result, industry is no longer confined to a limited set of predefined configurations. 

Industry is now able to design a configuration ideally suited to a given application, provided it can 

demonstrate that it meets the performance standards which ensure it is safe and infrastructure-friendly 

(Prem, Ramsay, and McLean, 1999). Having a greater variety of vehicle configurations is of 

particular interest to the forest industry, as the energy efficiency of its vehicles can be increased, and 

the transportation costs and associated pollution can be reduced. 

Through discussion with the forest industry and trailer manufacturers, the CRC for Forestry has 

identified two new vehicle configurations which offer attractive potential gains in efficiency. These 

gains will be evaluated in operational trials. 

Case Study 2 examines the potential impact of a PBS design which uses off-the-shelf technology to 

produce a quad axle B-double. Case Study 3 looks at the application of new stearable wheel 

technology to produce a high productivity semi-trailer. 

CASE STUDY 1: Improving the current configuration with a focus on tare weight 

Before exploring the use of new technology and trailer designs, the first step is to see what potential 

exists for improvements in vehicle efficiency through improvements with currently used 

configurations. This approach is attractive, as potential improvements might be made at relatively low 

cost and low risk simply by promoting the application of current best practice without venturing into 

the unknown. 

This approach is based on the theory that each configuration is limited by the GVW it can transport — 

a combination of the vehicle tare weight and the payload. As forest products are transported as a bulk 

commodity, load management is highly important and care must be taken to ensure each load is as 

close as possible to the maximum load, without exceeding load limits. 

When this is done, every kilogram of vehicle tare weight represents a kilogram of lost payload each 

trip. This is based on the core assumption that those vehicles with a lower tare weight will be able to 

transport more wood per trip at the same GVW than those vehicles with higher tare weights, but still 

use only the same amount of fuel. 

To understand what the potential savings might be, weighbridge data (by load delivered) providing 

details of truck identification, date, time, GVW and tare weight were collected from six forest 

companies from Western Australia, South Australia and Victoria over a period from 6 to 12 months. 

After eliminating corrupted data points, partial loads, and trucks making only few deliveries (less than 

10), the combined fleet for analysis comprised 427 trucks of four configurations which delivered a 

total of 52,306 loads within the analysis period. 

Table 1 presents a summary of the data, including the average tare weight for the lightest 20%, the 

average tare, and the average tare weight for the heaviest 20% of the vehicles in a configuration class. 

The table indicates there is a large variation in vehicle tare weights within the same configuration 

class, with a difference of up to 26% between the lightest 20% and heaviest 20% in a group. 

In the case of the 82t road trains, the tare weight variation within the configuration is greater than the 

gain achieved by moving up to the 82t road train configuration from the 79t road train.



Table 1. Summary of weigh bridge data 

Trucks Loads 

Allowable 

GVW 

Lightest 

20 % 

Average tare 

Whole 

group 

Heaviest 

20 % 

# # Kg Kg Kg Kg 

Semi-Trailers 97 9962 42500 16627 18275 20056 

B-doubles 87 17212 62500 21107 22651 24679 

79t Road trains 209 20071 79000 26008 28585 32867 

82t Road trains 34 5061 82500 27194 29450 31924 

For simplicity, the analysis in Table 2 assumes that vehicle fuel consumption is unchanged with a 

change in tare weight. Therefore if more wood can be moved per trip there is an opportunity to 

decrease fuel use for a given freight task with no change in vehicle configuration. Since vehicle fuel 

consumption increases with weight (USEPA,2004) further savings are likely as the time traveling 

empty would experience lower fuel consumption under the same operating conditions. Further 

research however is required to quantify fuel savings when the vehicle is empty. 

Table 2. Fuel savings relative to the heaviest 20% through increased payload 

Potential payload 

Fuel savings as compared to 

heaviest 20 % in class 

Lightest 

20 % 

Average 

Heaviest 

20 % 

Lightest 

20 % 

Average 

Kg Kg Kg 

Semi-Trailers 25873 24225 22444 13 % 7 % 

B-doubles 41393 39849 37821 9 % 5 % 

79t Road trains 52992 50415 46133 13 % 8 % 

82t Road trains 55306 53050 50576 9 % 5 % 

To put this into perspective, many forest companies have indicated they would invest in upgrading 

from an average B-double to an average 79t road train were that configuration allowed elsewhere apart 

from Western Australia. Such an investment would yield just over 20% improvement in efficiency, but 

requires significantly more effort to introduce and a higher initial investment. 

CASE STUDY 2: HPV Quad-B-double using current self-steering axle technology 

With the opportunity to develop alternative high productivity configurations through PBS, many 

transportation industries have explored options to increase payload to improve efficiency. One concept 

which has shown promise for several bulk transport applications, including wood chip transport, is the 

Quad-B-double configuration shown in Figure 1. 

Figure 1. Quad-B-double



By including two additional self-steering axles over a traditional B-double, the GVW is increased by 

14.5 tonnes. This approach is favoured by the transport industry as it stays with design and technology 

with which they are familiar, and the rig is only 1 metre longer than a B-double. Furthermore, the 

industry believes public acceptance for this configuration will be possible with a reasonable level of 

effort. 

Initial evaluation of the new Quad-B-double configuration for wood chip transport has led to the 

conclusion that the configuration will have a sufficient storage volume to reach full loads based on 

normal wood chip density. As the Quad-B-double is not currently hauling woodchips, determining the 

tare weights for evaluation of the impact on efficiency requires the use of certain assumptions. Firstly, 

as the Quad-B-double design is the same as the standard B-double, it is assumed the tare weights 

would be about the same with an additional increase in weight for the two additional axles and the 

extra one metre of length. Each axle weighs between 800 kg and 1200 kg depending on the axle 

design and suspension, with self-steering axles being at the heavier end of the range. The weight of the 

extra metre in length would vary between 200 Kg and 400 Kg depending on the frame and wall 

materials used. Based on these ranges, it is assumed that the Quad-B-double would weigh, on 

average, about 2400kg more than a standard B-double. Finally, since the heavier trucks in the fleet 

tend to be older, it is assumed the new Quad-B-doubles would cover a range of tare weights 

comparable to the lower end of the range reported in Case Study 1, giving a probable tare weight 

range of 23 500 Kg to 25 100 Kg, and a potential payload range of 51 900 Kg to 53 500 Kg. 

Table 3 shows the impact on fuel use with the shift to the Quad-B-double from the standard B-double 

for a given freight task. With this shift in configuration there are potential fuel savings of over 40%. 

Table 3. Difference in fuel use between Quad-B-double and Standard B-double. 

B-double 

Difference in fuel use Lightest 

Heaviest 

Average 

20 % 

20 % 

Quad-B- Light 29 % 34 % 41 % 

double Heavy 25 % 30 % 37 % 

CASE STUDY 3: HPV semi-trailers using new steerable wheel technology 

In addition to new and creative uses of existing technology, the introduction of PBS has provided an 

opportunity to incorporate new technology into the design of completely new HPV configurations. 

One example of this is the steerable wheel concept developed by Steerable Wheel Systems Pty Ltd 

(SWS). SWS has developed an innovative modular steerable wheel system that comprises 

electronically controlled twin wheels with integrated suspension (Prem, May, & Davey, 2008). The 

advantage of this new technology, of particular interest for woodchip transport, is the lack of an axle 

allowing for greater load volumes with a lower centre of gravity. In addition, the actively steered 

wheels allow for longer single trailer units which can be unloaded more quickly than trains. 

Furthermore, the wheel spacing options allow for a higher GVW within current semi-trailer and Bdouble 

overall dimensions. 

To examine the potential of this new concept for the forest industry, two steerable wheel 

configurations are being evaluated with the industry, including a 19 metre configuration (Figure 2) 

which would be an alternative to the current semi-trailer configuration, and a 26 metre configuration 

(Figure 3) which would be an alternative for the current B-double configuration. 

Figure 2. A 19 m steerable wheel 

configuration — 55.7 t GVW and 41.3 t 

payload



Figure 3. A 26 m steerable wheel configuration — 68 t GVW and 51 t payload 

Since steerable wheel trailers are a new concept based on a new technology, with new trailer design 

tailored to the technology, there are no examples on the road from which to draw actual weight 

information. Therefore tare weights and payload need to be derived from theoretical weights provided 

by design models. In the case of the 19 m configuration, the allowable GVW based on PBS will be 

55.7 t and the current design predicts a tare weight of 14.4 t for a payload of just over 41 t. 

For the 26 metre configuration, the PBS design would allow for a 68 t GVW and the design model 

predicts a tare weight of 17 t for a payload of 51 t. Assuming the predicted tare weights may be 

optimistic, the analysis in Table 4 summarises the impact on fuel use of moving from the current semitrailer 

configuration to the 19 m steerable wheel configuration: 

o at the predicted tare weight, and 

o at a tare weight that is 10% higher than predicted. 

Table 4. Difference in fuel use between semi-trailers and 19 m steerable wheel configuration 

Difference in fuel use 

Semi-trailer 

Lightest 20 % Average Heaviest 20 % 

19 m Predicted tare 60 % 70 % 84 % 

Steerable +10 % 54 % 65 % 78 % 

Table 5 summarises the impact of moving from the current B-double configurations to the 26 m 

steerable wheel configuration on the same basis. 

Table 5. Difference in fuel use between B-doubles and 26 m steerable wheel configuration 

Difference in fuel use 

Lightest 20 % 

B-double 

Average Heaviest 20 % 

26 m Predicted Tare 23 % 28 % 35 % 

Steerable +10 % 19 % 24 % 30 % 

Even at a tare weight 10% higher than predicted for the steerable wheel configurations, fuel use 

savings of: 

o up to 78 % are possible for the semi-trailer alternative, and 

o up to 30 % for the B-double alternative. 

These results are quite comparable to the savings achieved with the Quad-B-double in Case Study 2, 

but in this case the steerable wheel configuration has fewer wheels, a lower center of gravity and can 

be easily unloaded as a single unit. 

CONCLUSION 

Fuel consumption is an important factor in both the cost and environmental impact of transportation 

operations. One major focus of the transportation industry is the control these impacts by reducing 

fuel use while doing the same work. 

The three case studies presented, however, show more work can be achieved by using the same 

amount of fuel and, similarly to the FERIC study in Canada, by consuming a more nominal amount of 

fuel.



Each of the reported case studies examines progressively higher risk approaches to improving payload 

and thus the efficiency of forest transportation on public roads in Australia. In Case Study 1, fuel 

savings of up to 13% can be achieved by specifying vehicles that are as light as the current best 20% 

in Australia. As this involves no change in vehicle configuration nor the application of new 

technology, the effort to implement the change is relatively small, and returns potentially high. 

In Case Study 2, by taking advantage of new PBS rules and using off-the-shelf technology, even 

greater improvements are possible, with potential fuel savings of up to 41%. This approach requires 

the introduction of a new configuration built from relatively proven technology. Therefore the 

technical risk is low, but effort may be high to win public acceptance and ultimately route access for 

the use this new configuration. But the reward is high. 

Finally, Case Study 3 examines the use of a new configuration based on new design and technology to 

achieve savings as high as 84%. This approach is accompanied by a significant cost and risk in 

dealing with new technology and, as with the Quad-B-double, will require significant effort to gain 

route access. 

However this vehicle configuration offers significant potential fuel savings, and offers some design 

features that will benefit other aspects of the operations. 

This paper has focused more on the fuel savings that can be achieved. However, changes in vehicle 

configuration may also reduce green house gas emissions. Based on the fact that each litre of diesel, 

when burned, produces 2.7 kg of C02, green house gas emissions would be reduced by the same 

percentages, with increased payloads. Thus the industry would demonstrate its social responsibility in 

protecting the environment. The outcome of the case studies is important, and will become 

increasingly so with the introduction of carbon emission trading or carbon tax schemes in Australia. 

REFERENCES 

Berg S. and Lindholm E-L., 2005. ‘Energy use and environmental impacts of forest operations in 

Sweden’, Journal of Cleaner Production vol. 13, pp. 33-42. 

Michaelsen J., 2007. Reducing fuel consumption, emissions of GHG and transportation costs by the 

implementation of partial off-highway hauls in forest operations, Contract Report- CR-0293-1, 

Forest Engineering Research Institute of Canada, Montreal, Quebec. 

Prem H., Mai L., and Davey G., 2008. ‘A new steerable wheel system for road transport applications’, 

Proceedings of the International Conference on Heavy Vehicle, Paris 2008, paper 31. 

Prem H., Ramsay E., and McLean J., 1999. Performance based standards for heavy vehicles: 

Assembly of case studies, Report to National Transport Commission Australia. 

Sambo S.M., 2002. Fuel consumption for ground-based harvesting systems in western Canada, 

Advantage Report v.3 no.29, Forest Engineering Research Institute of Canada, Montreal, 

Quebec. 

Smartway Transport Partnership, 2004. A glance at clean freight strategies weight reduction, U.S. 

Environmental Protection Agency, viewed 8 April 2009, 



ONBOARD SYSTEMS FOR 

AUSTRALIAN FOREST OPERATIONS 

Martin Strandgard 1 

ABSTRACT 

Management of harvesting operations using onboard technology has been highly 

developed overseas, particularly in Scandinavia, but is relatively still in its infancy in 

Australia. Overseas experience has identified productivity gains of over 30% in some 

areas and expectations of further significant gains. On the basis of this overseas 

experience, the CRC for Forestry Program 3 has commenced to examine the range of 

onboard technology available, and to identify and trial equipment suitable for Australian 

forest harvesting conditions. The paper presents the preliminary findings of the project 

and the expected costs and gains from the adoption of the technology. 

INTRODUCTION 

Harvesting and transport account for over 50% of the costs for a rotation. Mechanical harvesting 

operations in Australia typically cost between $20 - $40 per tonne (averaging about $25 per tonne) and 

are currently used on about 80% of the 25 million tonnes harvested annually, a percentage that is 

increasing. International experience has shown that effective use of onboard computing technology 

can increase machine utilisation and productivity by up to 30% through identifying operational 

bottlenecks, delays and improving operator performance. Targeted operator training and damage 

prevention techniques have produced a reduction of up to 40% in particular maintenance costs. 

Limited detailed time-studies in Australia have found machine utilisation rates are typically between 

70 and 85%, with some under 60% (Acuna et al., 2009). 

OBJECTIVES 

This paper is based on a project funded by Forests and Wood Products Australia (FWPA) and three 

industry partners: Timbercorp Ltd, ForestrySA and VicForests. As the industry partners cover a broad 

cross-section of the Australian forest industry it is expected that the results will be widely applicable to 

Australian forestry. 

The project objectives are: 

• Improvement in utilisation and/or productivity of forestry equipment of at least 10% 

• Improvement of machine availability of at least 5% 

• Measurable improvements in fuel efficiency of the operations 

• Measurable improvement in operation safety and operator job satisfaction 

The project will consider only Australian harvesting operations and only those activities associated 

with harvesting, processing, transport to the coupe edge and onsite chipping. 

Utilisation is the proportion of scheduled time that a machine actually works. For most machines it 

can be further broken down into travel and work times. Utilisation is improved mainly through 

identifying and reducing delays. Reducing travel time can also be important. 

Productivity is the amount of wood cut or transported in a given time (typically m 3 or tonnes/hr). It 

can be expressed in terms of either productive (actual work hours) or scheduled machine hours. 

Productivity is strongly related to piece size (Hartsough and Cooper, 1998) and utilisation and, for 

forwarding, load capacity and hauling distance (Jiroušek et al. 2007). Operator skill is also a factor 

(Ovaskainen et al., 2004). Both utilisation and productivity can be affected by machine interactions in 

1 Department of Forest and Ecosystem Science, University of Melbourne. Email: mnstra@unimelb.edu.au.



a “hot deck” operation (Spinelli and Visser, 2008), for instance if a forwarder has to wait for 

harvesters to produce wood or a mobile chipper has to wait for wood to be delivered. 

In sawlog cut to length operations, value recovery is of more interest to forest owners than 

productivity. For example in ForestrySA’s operations, cross-cutting decisions can result in large 

differences in value recovery without changing volume output. Value recovery is being investigated 

in another project but potential synergies between the projects will be identified. 

Machine availability is the proportion of scheduled time that the machine is mechanically available for 

work. As such, it reflects machine characteristics (e.g. make/model, age, work and maintenance 

history), site factors (e.g. rockiness, slope) and also, to an extent, operator skills and attitude. 

Fuel efficiency is generally expressed in terms of fuel consumed per unit of wood cut or transported. 

Fuel use per unit of wood extracted in Sweden has been reduced from 5.4 l/m 3 to 3.7 l/m 3 over the 20 

years to 2006, mainly through the introduction of lighter, more fuel efficient machines and operator 

training. For a given site and machine, fuel efficiency is strongly related to utilisation and productivity. 

Operation safety is measured in terms of number of lost time injuries and amount of lost time per unit 

time (often expressed as Lost Time Injury Frequency Rate (LTIFR) which is the lost time per million 

hours worked). Job satisfaction is difficult to measure accurately and is dependent on both work 

related and external factors. 

DELIVERABLES 

The primary project deliverable is a selection guide to allow users to match their circumstances (e.g. 

plantation, native forest, grower, contractor, machine types, hot/cold decking, etc) and the issues they 

would like to resolve or understand (e.g. poor utilisation, productivity, etc) with an appropriate 

onboard system (or systems). This will be accompanied by an implementation guide and basic 

software data analysis tools. 

It is expected that the primary delivery mechanism will be the web, due to the ease of implementation, 

distribution, and management of the selection guide and other material via this medium. 

AVAILABLE ONBOARD COMPUTING EQUIPMENT 

A review of available equipment identified three major categories of onboard computing technology 

which have the potential to meet the project requirements: 

1. Purpose – built forestry systems 

Purpose-built systems have the advantages of associated software to manage, analyse and 

display forestry specific data and the ability to obtain user inputs, particularly delay causes. 

• Manufacturer supplied systems. All recently manufactured harvesting equipment 

has some level of onboard computing technology. Often a number of computers are 

available with different capabilities (and prices). Data are transferred by USB 

memory device, or over the mobile phone network, or in some cases by floppy disk. 

The data complies with the StanForD standard (Skogforsk, 2007), allowing data from 

different manufacturers to be processed using any manufacturer’s software. 

• Multidats. Multidats were developed by FPInnovations (formerly Feric) in Canada to 

collect long-term utilisation data from forestry machines. The basic configuration 

records machine activity using a vibration sensor. There are 4 channels for 

connection of analogue sensors to switches, current/pressure sensors, etc and an 

optional GPS input. Operators can also enter delay causes in categories. Operator 

feedback can be provided with the optional Multipad display. Data are transferred 

with a PDA or laptop connected to the Multidat by cable.



2. Telematics 

Telematics is the remote monitoring of machine function and location. It originated with the 

trucking industry and has recently been applied to off-road vehicles. Several major 

manufacturers of forestry equipment have produced proprietary telematic products for their 

construction equipment (e.g. Product Link from Caterpillar and Komtrax from Komatsu). 

There are also a number of generic telematic solutions e.g. Mix Telematics and Topcon Tierra. 

Telematic equipment will soon be standard on new construction equipment and can be 

retrofitted to older machines or to machines from other manufacturers (with limited 

functionality). 

The key characteristic of telematic equipment is that data collected by the onboard technology 

are sent periodically to a central computer via satellite or mobile phone, and accessed via a 

web site. The service is provided by a one-off payment or ongoing fee-for-service basis for 

each machine. 

The basic function of telematics is tracking of vehicle movements via GPS. Most telematic 

equipment can also collect data from the machine’s Canbus (see below). Some can collect 

additional data using digital or analogue sensors. There are currently no standards as to what 

data are collected and how they are presented to end-users, though this is being addressed by 

the Association of Equipment Management Professionals. 

Telematics is being largely driven by the requirements of the trucking and construction 

industries. That said, there is a great deal of potential overlap in requirements, such as vehicle 

location, hours worked and fuel consumption. However, some degree of customisation will be 

required to meet the requirements of the forest industry. 

There are a small number of telematic systems that encompass both heavy vehicles and 

forestry applications, such as RouteHawk (Strong Engineering), FCGIS-H (Bracke Systems) 

and TAGASIS (Aldata). 

3. Generic Dataloggers 

A wide range of generic dataloggers is available which could be used to collect data from 

forest harvesting machines (e.g. Datataker, BusDAQ and ASLH309). Some have been used 

for this purpose, mainly in research trials. Typically they have multiple channels that can 

collect digital, analogue, or Canbus (see below) data, or a combination of these. GPS input is 

an option on some dataloggers, and some can be interfaced with displays to provide feedback 

to, or input from, machine operators. 

Depending on the make and model, generic dataloggers can transfer data by a variety of 

means including USB memory devices, wifi (Bluetooth or wlan), radio modems and mobile 

phone modems. 

Equipment installation 

With the exception of manufacturer supplied equipment, data collection hardware needs to be wired 

into each machine to collect the required data using either stand-alone sensors, such as a vibration 

sensor, connections to existing switches or installation of new switches to detect on/off actions (e.g. 

pedal movements), or electric current or hydraulic pressure changes. The alternative to this approach 

is to monitor the Canbus. Canbus is a two wire communications bus originally developed by Bosch to 

simplify wiring looms in luxury cars. It has been widely adopted in many areas, including heavy 

vehicles such as trucks and forest machines. Canbus promises to deliver a great deal of useful 

information about machine performance and activities with a single connection (eg. total operating 

time, work time, travel time, fuel use, engine rpm/load). Forest harvesting machines control the 

engine and transmission using a Canbus compliant with the heavy equipment standard J1939 and the 

remainder of the machine functions using proprietary manufacturer codes which differ between 

manufacturers and machine models. The data available from the J1939 compliant component is



unknown at this stage as there is no compulsion for a manufacturer to fully implement the standard or 

to publicly document their implementation. 

The Canbus is accessed either via an existing diagnostic port or by wiring an additional port into the 

system (if the system has this capability) which then must be enabled programatically. 

As the use of Canbus for data collection on forestry machines is a relatively unknown factor, 

equipment selected with Canbus capability will need to be able to collect data using analogue and/or 

digital channels to minimise the risk of not being able to collect useful data via the Canbus. 

INDUSTRY PARTNER REQUIREMENTS 

Each industry partner will be installing equipment on 3 to 4 forest machines which work together on 

an operation. To test a range of onboard computing equipment and to avoid data integration problems, 

it is anticipated that each partner will test a different type of equipment and will use one brand of 

equipment (where possible) within their organisation. 

ForestrySA manage radiata pine plantations primarily for sawlog production. As mentioned above, 

ForestrySA have a particular interest in value recovery while their contractors are focussed on 

utilisation and productivity. Manufacturer supplied equipment is the only means to achieve both these 

requirements. Equipment for this study has to be capable of optimising cross-cutting decisions and 

allowing operators to enter delay causes. Contractors are unlikely to allow ForestrySA staff to view 

data related to utilisation and productivity, so segregation of data needs to be factored into the data 

collection and transfer process. To meet these requirements Dasa computers will be installed. Dasa 

equipment is installed by manufacturers and as aftermarket equipment which makes it ideally suited 

for this project. 

Timbercorp are a major blue gum MIS company producing woodchips for export that owns and 

operates their own harvesting equipment. Their typical harvesting operation is to fell and debark 

stems at the stump, and use forwarders to transport whole stems to a roadside chipper. This type of 

operation is primarily concerned with efficient throughput of timber. The fundamental drivers are the 

number of stems each harvester processes, the number of forwarder loads and the tonnes of wood 

processed by the chipper. These are in turn dependent on the utilisation percentage. Machine 

interactions can also affect individual machine performance. 

Timbercorp are interested in all aspects of utilisation, productivity, fuel efficiency, etc. Timbercorp 

currently have Multidats installed on a range of harvesting equipment as part of another trial. 

RouteHawks will be installed on the Timbercorp equipment to test the telematics approach including 

data collection from the Canbus and potentially remote data collection via mobile or satellite phones. 

Although Timbercorp are currently under administration, it is anticipated that this will be resolved in 

time for the project to continue. If this proves not to be the case, the project will proceed with the 

other two industry partners. 

VicForests extract timber from Victoria’s native forest estate using a mixture of hand-felling and 

mechanised operations depending on tree size. In clearfell operations, stems are generally cut using a 

felling head on an excavator and transported to a landing by a grapple skidder. Excavator-mounted 

grapples are then used to debark, cut to length and load logs onto trucks for transport to the mill. 

Although price premiums apply to high quality log products, it is generally not possible to identify log 

quality due to internal defect until the stem has been cut. Regrowth thinning operations generally use 

harvesters and/or feller-bunchers and forwarders to harvest and process trees. 

As with ForestrySA, VicForests’ interests differ from those of their contractors. The recommended 

solution is to install Multidats on a range of equipment, to obtain basic productivity and utilisation 

information, and then examine the breaking down of utilisation to travel and work times, and the 

breaking down of causes of delay into categories, such as mechanical, personal and operational delays.



Each mobile machine will also have a Garmin GPS/GIS to display coupe boundaries, site hazards and 

the machine’s location. 

MACHINE DATA COLLECTION 

This section describes one or more alternative means of collecting the data required to achieve the 

project objectives. 

Utilisation and Productivity 

• Harvester utilisation can be measured using a vibration sensor or engine rpm/load to 

determine when the machine is active. Travel and work components of utilisation can be 

distinguished in a number of ways, including sensors to detect when the boom or harvester 

head is active, the travel alarm, speed (GPS), or connections to the travel pedal or brake 

circuit,. The only practical source of detailed information on log dimensions is the 

manufacturer supplied system. Simpler production measures include stem counts based on the 

harvester head tilt up switch and log counts based on number of main saw cuts or time spent 

by the harvester head travelling along the stem as an indication of stem length. 

• Forwarder productivity is measured by the amount of wood transported in a given time. To 

a large extent this is determined by the load capacity of the forwarder, piece and stack size and 

travel distance. Other factors include the operator’s loading and unloading efficiency and 

delays. Productivity can be estimated using a GPS to estimate travel distance and number of 

loads or a load cell to estimate number of loads and total weight transported. The operator can 

also manually enter into the onboard computer when the machine is being loaded or unloaded. 

Forwarder utilisation can be determined with a vibration sensor or engine load. Utilisation 

can be broken down to work and travel, using a GPS or by monitoring engine load. 

• Skidder productivity (as with the forwarder) is measured by the amount of wood transported 

in a given time. However, it is more difficult to estimate the amount of wood a skidder 

transports because a skidder drags logs rather than carrying them. Estimates of productivity 

can be made using a GPS to estimate travel distance and number of loads or a strain gauge to 

estimate load weight. Utilisation can be determined with a vibration sensor or engine load. 

Utilisation can be broken down to work and travel with a GPS or by monitoring engine load. 

• Mobile chipper productivity is directly related to its utilisation, all else being equal. Spinelli 

and Visser (2009) reported that organisational delays, such as waiting for wood delivery and 

chip trucks, were the most significant factors affecting chipper utilisation. Operator and 

mechanical delays were also important. Travel time and operator skill are relatively minor 

issues for chipper utilisation and productivity. 

Utilisation can be determined using a vibration sensor or Canbus monitoring of engine load. 

Given the significance of delays to chipper utilisation and productivity, recording of delay 

causes is an essential requirement for a chipper. Entry of delay causes is currently only 

possible using manufacturer supplied equipment or a Multidat or RouteHawk. 

Fuel efficiency 

Fuel use can be measured using a flow meter installed in machine fuel lines or by recording refill 

quantities. The Canbus also reports fuel consumption but it is potentially inaccurate as the value is 

calculated rather than being measured directly. 

Fuel efficiency is strongly related to machine utilisation and productivity hence, unless a specific need 

is identified, it will be monitored in the project by recording machine refill quantities rather than by 

fitting sensors to machines. 

Machine availability 

Machine availability is monitored by recording machine utilisation and delay causes, in order to 

identify delays that are mechanical in nature and hence result in the machine being unavailable for



work. Currently only manufacturer supplied equipment, Multidats and RouteHawks are capable of 

recording delay causes to calculate and manage machine availability. 

Operator health, safety and job satisfaction 

Lost time injuries in forestry where harvesting is mainly mechanical are typically around 10-20 per 

million work hours. At this frequency it is unlikely that safety improvements could be monitored with 

the small sample in the trial. Therefore the trial will focus on means to reduce identified injury causes. 

Most injuries occur outside the cabin. Machine maintenance was identified by Nieuwenhuis and 

Lyons (2002) as a major source of injuries in Irish forest operations, so steps to identify and reduce 

delays resulting from machine breakdowns will have the additional benefit of reducing time lost 

through injury. 

Operating machines has created another range of health and safety issues. Stress and fatigue have 

been identified as important issues for machine operators, which can be alleviated by frequent short 

breaks (Kirk et al., 1997) that can be monitored by onboard equipment. Anecdotal evidence suggests 

that fitting onboard optimising computers has reduced stress in harvester operators as the computers 

now make most of the decisions as to what products to cut from each stem. Onboard GPS/GIS 

systems can improve safety by alerting operators to potential site hazards and by keeping machines 

within site boundaries. 

A major means of improving operator job satisfaction is to provide feedback on operator performance 

and rewards for improved performance. Several onboard computing options can provide operator 

feedback (manufacturer systems and Multidats with the Multipad option and RouteHawks). It may 

also be possible to achieve by adding screens to generic dataloggers. 

PRELIMINARY COSTS AND RETURNS 

Indicative purchase costs (Australian dollars) for each unit used in the study are as follows: 

• DASA computer. A price was not available at the time of writing 

• Multidat. A new price list is being compiled with reduced prices. Each unit is expected to 

cost under $5,000 including optional GPS. The Garmin GIS display units are ~$900 each. 

• RouteHawk. Basic unit cost is $5,000. Use of the central database to store and distribute data 

will incur a small ongoing cost per machine. Mobile or satellite phone modem will add costs 

(up to ~$2,000 for a satellite phone) as well as call costs which will depend on the amount and 

frequency of data transferred. 

As all the units are imported, exchange rate movements will impact these prices. 

Installation and operating costs will depend on the options installed and the machine type. The most 

basic installation is to wire the unit to the machine’s power supply. Other options will include 

installation of GPS aerials, and switches or connections to existing switches. Operating costs will 

include data transfer and analysis and will depend on the degree to which this is automated. Manual 

data transfer may require purchase of a suitable PDA. 

At the time of writing, the equipment for the project had not been installed. However, the CRC for 

Forestry Program 3 has several other projects using Multidat onboard computers. In one project, 

~10% of potential productivity gains were identified through the use of the onboard technology. This 

represents an estimated annual saving ($/tonne) of over $40,000 per harvester. 

Details of the changes required to achieve the productivity gains will be made available initially to 

CRC for Forestry Program 3 members.



CONCLUSION 

Onboard computing technology has been shown in overseas studies to be able to significantly increase 

forest harvesting machine utilisation and productivity, and reduce maintenance and fuel costs. Studies 

performed in Australian forest harvesting operations have indicated that there is the potential for these 

gains to be achieved in Australia. 

The initial stage of the project has been to identify available onboard computing technology and match 

it to industry partner requirements. 

• ForestrySA manage radiata pine plantations primarily for sawlog production and would be 

best suited with Dasa onboard computers as they can both meet ForestrySA’s requirement to 

increase value recovery as well as improving the harvesting contractor’s performance. 

• Timbercorp own their harvesting equipment and so are interested in the full gamut of project 

objectives. As they have a Multidat installation on a related project, their needs will best be 

met by trialling the RouteHawk telematics solution potentially including testing of Canbus 

connections and remote data collection techniques. 

• VicForests use contractors to harvest timber from Victoria’s native forests. The nature of 

their harvesting operations is best suited to installing Multidats and Garmin GPS/GIS units on 

a range of equipment to obtain basic utilisation and productivity data and to examine delay 

causes and improve operator safety. 

REFERENCES 

Acuna, M., Wiedemann, J. and Strandgard, M. (2009). Evaluation of an in-field chipping operation in 

Western Australia. CRC for Forestry Program 3, Bulletin 4. 

Hartsough, B.R. and Cooper D.J. (1998) Cut-To-Length Harvesting of Short Rotation Eucalyptus at 

Simpson Tehama Fiber Farm. Proceedings of the Short-Rotation Woody Crops Operations 

Working Group, Second Conference, 25-27 August 1998, Vancouver, Washington, USA 

Jiroušek, R. Klvač, R. and Skoupý, A. (2007) Productivity and costs of the mechanised cut-to-length 

wood harvesting system in clear-felling operations. J. For. Sci. 53(10): 476-482 

Kirk, P.M., Byers, J.S., Parker, R.J. and Sullman, M.J. (1997) Mechanisation Developments Within 

the New Zealand Forest Industry: The Human Factors. Int. J. For Eng. Vol. 8(1) 

Nieuwenhuis, M. and Lyons, M. (2002). Health and Safety Issues and Perceptions of Forest 

Harvesting Contractors in Ireland Int. J. For Eng.Vol. 13(2) 

Ovaskainen, H., Uusitalo, J. and Väätäinen, K. (2004) Characteristics and Significance of a Harvester 

Operators' Working Technique in Thinnings Int. J. For Eng. Vol. 15(2) 

Ovaskainen, H. (2009). Timber harvester operators’ working technique in first thinning and the 

importance of cognitive abilities on work productivity. Dissertationes Forestales 79 University 

of Joensuu, Faculty of Forest Sciences 

Skogforsk (2007). Standard for Forest Data and Communications. Skogforsk. 

Spinelli, R and Visser R.J.M. (2008). Analyzing and estimating delays in harvesting operations. Int. J. 

For. Eng. 19(1) 

Spinelli, R and Visser R.J.M. (2009). Analyzing and estimating delays in wood chipping operations. 

Biomass and Bioenergy 33(3): 429-433



TIMBER FROM NATIVE FOREST AND PLANTATION EUCALYPTS - 

USERS WILL QUICKLY FIND THAT THEY ARE 

NOT THE SAME THING 

Gregory Nolan 1 

ABSTRACT 

Solid eucalypt hardwood timber has largely shifted in Australia from a general 

construction material to primarily an architectural selection in building. Yet, this section 

of Australia’s construction industry has established preferences for relatively clear, stable, 

larger section timber recovered from older native forest eucalypt logs, traditionally the 

base resource for the Tasmanian solid hardwood production industry. Over recent 

decades, the transition from this resource to younger regrowth forced significant industry 

changes, but these are comparatively minor to the changes that are likely to occur as the 

industry’s base resource becomes smaller eucalypt plantation logs selected for their 

growth and not necessarily wood quality. This paper reports on recent work on the 

comparative properties of the mature, regrowth and plantation resource available for the 

Tasmanian solid hardwood production industry and discusses the likely implications their 

different properties for profitable processing, value adding and use. 

INTRODUCTION 

Forests provide society with a range of environment and economic services and products, and an array 

of human experiences and associations. This paper deals with just a few, focusing on the solid wood 

products drawn from native and plantation eucalypt forests, the associations these products have when 

used in artefacts such as buildings, and the potential for plantation eucalypts to maintain those 

associations and support (or maintain) a viable processing industry. 

Wood products can be as varied as the forests from which they are drawn. A large part of this variety 

results from the quality of the log and the log’s transformation during different production processes 

into increasingly reduced primary wood elements: first, solid wood products, then short grained 

elements such as chips or flakes; fiberal elements, and finally chemical elements (Marra 1972, p.202). 

These primary wood elements can be viewed as products ready for artefact-making or as the 

ingredients for second generation products such as plywood, laminated veneer lumber (LVL) and fibre 

boards, again widely used in artefact-making. Another large part of the variety of wood products 

results from the temporal, genetic, and site characteristics of the forests that yield the logs for 

processing. If Tasmania’s eucalypt log supply is taken as an example, logs for sawn boards can 

originate from one of a dozen tree species from 80 to hundreds of years old drawn from native forests 

or from 15 to 25 year old monocultural native plantation forests of Southern blue gum, E. globulus, 

and Shining gum, E. nitens. Each forest provides logs, and these logs can be converted into at least 

some part of the range of potential wood products. The best (largest, straightest, and cleanest) logs 

provide a resource for the widest range of products, including high quality solid timber boards. The 

worst (smallest, youngest or most crooked) logs provide a resource for a much more limited product 

range, usually only chip or fibre production. 

The vast bulk of wood products generated through this matrix of resource and process and used in 

artefact construction are general commodity products: structural and industrial grade timbers and sheet 

materials used where considerations of economy, strength or durability dominate their use. The unseen 

structural elements of buildings are an example of this. As pine products are consistently cheaper to 

produce than hardwood ones, they have come to dominate this sector over recent decades, largely 

replacing hardwood products. However, there is a group of applications where the importance of the 

1 Centre for Sustainable Architecture with Wood, School of Architecture and Design, University of Tasmania, Locked Bag 

1-324, Launceston, Tasmania 7250 +61 3 6324 4478, Email: Gregory.Nolan@utas.edu.au.



aesthetic character of the timber and wood’s special association with nature and human culture are 

relatively high and more expensive differentiated products are used. These are primarily appearance 

applications in buildings and furniture. 

Unable to compete against cheaper pine products, solid eucalypt hardwood timber has largely shifted 

from being a general construction material to become primarily an architectural selection in building. 

In Australia, most appearance wood products are hardwoods and those produced locally are drawn 

almost exclusively from native forests. In Tasmania in particular, the most valuable products are 

drawn from mature native forest logs. Over the last two decades, industry’s transition from processing 

this resource to younger native forest regrowth, coupled with the shift in market from structural to 

appearance products has forced significant industry changes but these are likely to be relatively minor 

compared to the changes that are and will occur as native forest logs become relatively rare and the 

base resource for industry moves from regrowth to smaller eucalypt plantation logs of species selected 

their growth characteristics and not necessarily wood quality. 

There are several factors to explore in this transition: the nature of the desire for particular 

characteristics in wood and the demand this generates; the characteristics found in the timber from 

native forest logs necessary for the appearance applications that will support a differentiate product; 

the combination of age, species, and growing conditions necessary to produce trees with those 

characteristics; the particular characteristics of the plantation resources due to come on stream in 

Tasmania from about 2020, and the effects this difference will have on the timber processing industry 

and designers. 

DESIRE AND DEMAND 

Wood products possess a culturally-embedded attraction or appeal far beyond their simple form or 

functionality and designers draw on this to enhance the artifacts they design or make. This attraction is 

not just based on an aesthetic reaction to the wood’s colour, grain and touch but includes wood’s 

important associations with nature, its approachability of form and assembly, and senses of continuing 

culture. 

Designers recognise wood as a medium for creating more friendly and approachable environments for 

living (Kock 1972, p.15). For renowned Finish architect, Alvar Aalto, wood was a primary means of 

expressing nature’s impact upon the built environment and its capacity to maintain and even enrich 

life. Wood’s ‘biological characteristics, its limited heat conductivity, its kinship with man and living 

nature, the pleasant sensation to the touch it gives (Aalto 1956, p.142)’ made it a suitable material 

through which to design a sympathetic world. Menin and Flora, (2003, p.79) believe Aalto’s use of 

wood ‘became a trope for the individuality of his buildings and indeed his mind, as well as for their 

yearning for something natural amidst the new forms and material of the modern epoch’. Earle (1972, 

p. 223) reflected on this kinship between man, nature and wood, similarly contrasting it to the 

characteristics of man-made materials. 

The infinite shape plasticity of modern technology's production, made of materials and 

forms that deny any self identification in their total commitment to their function, is not 

easily related to man's experience with living. Instead the knotty, splintery, irregularly 

dimensioned, warping, shrinking, swelling, directionally grained, uneven strengthened 

piece of wood expresses better man's grasp on the imperfect realities of life. 

Wood’s democracy of use (Nolan 1994) is also part of making more livable environments. Kock (1972 

p.16) believed that wood’s workability provides us with an opportunity to create for ourselves the 

objects of beauty and usefulness that we desire. That is, many people can make (and enjoy making) 

useful wooden things for themselves. By inference, those living in wood-rich environments are likely 

to have a better intuitive understanding of the origins and character of their surrounds and be more 

comfortable in them. 

Banhan (1972, p. 8) noted a more ingrained cultural association as ‘our hands become immemorially 

conditioned to the grasp of wood and it has acquired for us an ancient familiarity… The grasp of wood 

is probably as basic to human culture as the making of fire and the writing of language.’



Figure 1. Unique high feature woods (Left to Right): figured burl, wavy grain, gum cluster, 

and blackheart figure 

Designers of buildings and other artefacts intuitively recognise these desirable associations and 

attempt to borrow and blend them with the more obvious aesthetic appeal of grain and texture in the 

items that they design and make. For this, they regularly exploit high-quality appearance woods, often 

with limited amounts of natural feature, or unique woods, where the extent of the natural feature 

establishes the material’s appeal. Shown in Figure 1, these unique woods can contribute to the most 

special applications. 

Wood included in any visible surface or structural element carries these associations but it is strongest 

in elements close to the eye, the hand and the touch. These close-at-hand objects, such as the pews 

shown in Figure 2, also place the most demanding requirements on the appeal of the wood. Elements 

in view but further away, such as architectural structures, have lesser demands on quality. 

Figure 2. Tasmanian oak cathedral pews Figure 3. Cathedral nave and wood panelling



The key wood properties for appearance applications are colour, grain, feature, workability and critical 

aspects of functionality (such as board size, stability, hardness and durability). Ranking the importance 

of each of these properties for particular applications is generally subjective (due to the number of 

variables included) but examining price in the market place can provide a reasonable objective ranking 

of the desirability of grade, dimension and colour for more general building design. Tables 1 to 3 

(from Nolan et al. 2005, p.7) show key relativities. Material graded Select under AS 2796 (Standards 

Australia 1999) was worth about 75% more per m 3 than structural material of the same size (Nolan et 

al. 2005, p.8). Select appearance material was worth significantly more than Standard grade and about 

twice as much as high feature material. A premium is paid for wide material. The additional price paid 

for increased thickness can probably be attributed to both increased desirability and production costs. 

Table 1. Relative prices of flooring milled form nominal 100 x 25 mm material 

Grade Select Standard High Feature 

Relative price /m 3 119% 100% 61% 

Table 2. Relative prices for Select boards nominally 25 mm thick 

Width (nom) 75 100 125 150 

Relative price /m 3 93% 100% 100% 117% 

Table 3. Relative prices for Select boards nominally 100 mm wide 

Thickness 25 38 50 

Relative price /m 3 80% 100% 114% 

In summary, there is a general pattern that appearance products are valued significantly more than 

structural ones, and wide, select (low feature) boards are valued far more than other appearance 

products. Designers wish to use these products more and their clients are willing to pay more for them. 

Like all general patterns however, there are notable exceptions. When a wood is particularly unique or 

the level of feature in the wood reaches a certain level of consistency and intensity (Nolan 1998), their 

value and appeal can also increase. They become unique woods, used in special applications. 

The Australian, and especially the Tasmanian, hardwood production industry naturally recognized 

these realities. Largely unable to compete with cheaper sawn and engineered pine product in the 

structural markets, it generally focuses on maximizing recovery of appearance product, and of select 

material in particular. Logs from mature and large regrowth native forest tree have shown that they 

can supply these products in an economically sustainable way. They can also provide a supply of 

special timbers. 

AGE AND SPECIES EFFECTS 

To successfully substitute for older material, a replacement resource will need to provide logs of a 

suitable wood quality and of a size that can be economically processed. In this, age and species are 

important. The age of the resources generally influences log size, the uniformity of the wood and the 

generation of special features. Species characteristics also significantly influence wood quality. 

As shown in Figure 4, Washusen and Clark (reported in Nolan et al. 2005) established that the 

recovery of select grade board is strongly related to log size. This is due to a range of factors including 

geometric characteristics, the percentage of lyctus-susceptible sapwood in the log and growth stress. 

However, as shown in Figure 5, the plantation logs to be provided to the Tasmanian industry are 

forecast to be considerably smaller on average than those from native forest. As a result, the 

proportion of Select material recovered is likely to decrease.



Recovery of select grade 

(% of log volume) 

60 

50 

40 

30 

20 

10 

r = 0.68 

0 

15 20 25 30 35 40 45 50 55 60 65 

Log small end diameter (cm) 

Figure 4. Plot of log diameter and recovery for 

pruned eucalypts milled with std. industry 

practice (Washusen and Clark, in prep) 

Figure 5. Log size distribution of “high 

quality” sawlogs from hardwood plantation 

and native forest (reproduced from FT 2007). 

Age also affects the uniformity of the wood. A tree grows by depositing a new layer of wood, under 

the bark, on the outside of the previous year's deposited wood. Freshly deposited wood differs from 

the wood deposited in the previous year’s growth, up until the tree reaches an age nominally between 

20 and 40 years old, after which the characteristics of the deposited wood are relatively consistent. 

Figure 6 is a general diagram of the relationship of wood density to age. As a result, as shown in 

Figure 7, a quarter sawn board cut from a 25- year-old plantation-grown log, cut from the zone of 

wood property change (in yellow), is likely to have much greater variability in wood properties across 

the board than a board cut from the zone of unchanging wood properties (brown) from a mature log 

from a natural stand. The material currently supplied to Tasmanian hardwood producers probably has 

an average age of 90 years, while the plantation material to be supplied has largely been planted since 

1995 and will be about 25 years old when harvested. 

Figure 6. Generalised pith-to-bark variation in 

average basic density (Nolan et al. 2005, p.72) 

Figure 7. Boards cut position from 110 year 

and 25 year logs (Nolan et al. 2005, p.72) 

Age affects generate many of the most desirable characteristics of special timbers. Fiddle back and 

wavy grain form as the mass of the standing tree compresses and buckles the grain at the tree’s base.. 

Burl is formed through the repeated healing of a wound. Blackheart is from long fungal infestations. 

Few if any of these occur in younger regrowth trees and it is unlikely any will occur in plantations. 

Species characteristics affect wood quality. In southern Australia, plantation eucalypt are generally 

one of two species: Southern (or Tasmanian) blue gum, E. globulus, planted in areas not susceptible to 

frost and Shining gum, E. nitens, planted in areas that are frost susceptible. In 2004, about 60% of 

Australia’s plantation resource was E. globulus. As most of Tasmania’s plantations are occasionally 

susceptible to frost, E. nitens makes up the majority of material planted in that state to supply an 

ongoing sawn hardwood industry with logs. However, these species were generally selected for their 

growth characteristics and general pulping properties, not their properties as solid wood products.



TASMANIA’S YOUNG EUCALYPT AND PLANTATION RESOURCE 

By 2020, about 50% of Tasmania’s high quality sawlog resource will be drawn from a plantation 

eucalypt resource. Young regrowth will make up some proportion of the remainder. However, recent 

studies indicate that recovery of appearance grade products from young regrowth and plantation 

material in Tasmania is likely to be significantly less than from mature and older regrowth material. 

Reduced log size is only one reason for this. Increased drying degrade such as distortion and internal 

check is likely to lead to loss of grade and profitability and for the main Tasmanian plantation sawlog 

species, thinned and pruned E. nitens, potentially even complete loss of economic viability. 

After milling regrowth messmate, E. obliqua, from 1901, 1934, 1949 and 1967, Innes et al. (2005) 

reported a range of age and site effects, including: 

• Younger messmate logs had a lower heartwood proportion than older logs. 

• The youngest messmate (1967 regrowth) underwent significant internal checking and its 

properties were generally more variable. 

• Boards cut from younger material shrank less than older, but yield of select grade was 

substantially lower due to gum vein. 

• It appears that the oldest logs had the highest growth stress, as those logs had the most end 

splitting and boards cut from them the most spring. 

• There was no trend with age of basic density, initial moisture content, drying rate, strength or 

hardness. 

Recovery of boards graded Select to AS 1684 on both sides and of boards subject to internal check is 

included in Table 4. 

Table 4. E. obliqua - % of boards Select both sides and % of boards internally checked 

Year % Select both sides % Internally checked 

1901 92 0 

1934 78 4 

1949 44 0 

1967 20 15 

In 2005, a major study sought to assess the economics of processing the major southern Australian 

species of plantation eucalypts of known origin using current industry-standard equipment and 

procedures (Innes et al. 2008). It also sought to identify the factors most directly affecting the value of 

dry output given current market conditions. 

Table 5. Log classes 

species E.globulus E.globulus E nitens E. nitens E.globulus 

age 47 13 26 19 

silviculture fibre wide spaced, pruned thinned, pruned fibre thinned, pruned 

location Gippsland Gippsland Ridgley Ulverstone 

forest owner GRP Frank Hirst Gunns Robyn May 

log description run-of-bush pruned butt pruned 

butt 

unpruned 

upper 

run-ofbush 

pruned butt 

As detailed in Table 5, two species, E. nitens and E. globulus, harvested from four sites in two States 

were milled across three operational hardwood mills, ITC Timber’s small regrowth log mill at 

Newood in Tasmania and two conventional mills, ITC Timber Heyfield in Victoria and Gunns 

Lindsay Street in Tasmania. A total of 12,778 sawn boards were produced. This material was dried in 

a combination of conventional drying yards (ITC Heyfield and ITC Mowbray) or recently 

commissioned predryers (Gunns Lindsay Street). Some boards were milled into final products and all



boards were graded and tallied. Value reducing defects were assessed. Market values were estimated 

and applied to the recovered material. 

Of the logs milled in Tasmania, sawn recovery was dependent on the milling equipment available 

while total high-grade recovery was low. Total dry recovery from the newer ITC Newood mill was 

significantly higher from the same resource than the conventional mills. ITC Newood achieved a dry 

recovery of 40% from 26-year-old thinned-pruned E. nitens butt logs. Gunns Lindsay Street recovered 

34%. For E. globulus butt logs, recoveries were 37% and 29% respectively. 

dried board recovery (nominal m3 

board/m3 log) 

45% 

40% 

35% 

30% 

25% 

20% 

15% 

10% 

5% 

0% 

13% 10% 11% 4% 

4% 

4% 

5% 

21% 22% 20% 

4% 

25% 

0% 0% 0% 

0% 

1 2 3 4 5 6 

log grade 

HF/merch standard/MF select 

Figure 8. 26-year-old E. nitens thinned-pruned butt recovery- ITC 

dried board recovery (nominal m3 

board/m3 log) 

45% 

40% 

35% 

30% 

25% 

20% 

15% 

10% 

5% 

0% 

6% 

3% 

6% 

18% 17% 

0% 

2% 

3% 4% 

26% 

3% 

17% 17% 

3% 

2% 5% 

1 2 3 4 5 6 

log grade 

HF/merch standard/MF select 

Figure 9. 26-year-old E. nitens thinned-pruned butt recovery – Gunns 

Assessment of recovered boards milled into final products showed that: 

• Boards from E. nitens butt logs (both thinned-pruned and fibre managed) showed higher levels 

of surface check as the primary reason for downgrade (25-30%). 

• There was an unexpected high loss in recovery when skim-dressed boards were milled into 

final products, especially of E. globulus. 12% of sawn volume from the thinned-pruned E. 

nitens butt logs and 30% of sawn volume from the thinned-pruned E. globulus butt logs was 

rejected. This was mainly due to distortion related effects. 

• Boards cut from E. globulus thinned-pruned butt logs show higher level of “gum features” as 

the primary reason for downgrade. 

0% 

13%



100% 

90% 

80% 

70% 

60% 

50% 

40% 

30% 

20% 

10% 

0% 

Gunns ENIT TP butt - 100mm - all grades 

Gunns ENIT UTUP - 125mm - all grades 

Gunns ENIT TUP tops - 125mm - all grades 

Gunns EGLOB TP butt - 125mm - all grades 

NST EGLOB TP butt - all sizes, all grades 

NST ENIT TP butt - all sizes, all grades 

no check minimal check moderate check heavy check 

Figure 10. Levels of internal check by log type. 

A high level of internal check was also found in the Tasmanian E. nitens butt logs harvested from both 

the fibre-managed and thinned and pruned stands. As shown in Figure 10, 25-31% displayed minimal 

check while between 6-17% showed moderate/heavy check. The check classed as minimal was still 

sufficient to be a serious defect and be grade limiting. The rate of checking was consistent for air-dried 

and pre-dried material, the two major hardwood drying technologies. Internal check in the E. nitens 

top logs and the E. globulus was significantly less. 

The high level of internal check observed was (and remains) of serious concern as the studied logs 

were from an expensive silvicultural regime, similar to the regime adopted by major growers 

managing E. nitens plantations for clear wood log production. The results indicate that between 15 and 

40% of boards from similar thinned and pruned full-rotation E. nitens can be expected to contain 

significant levels of internal check if dried with current industry technology. The situation could be 

even worse in full-scale production if 5.4 m logs were milled into long length boards. As stated above, 

there was an unexpected high loss in recovery when skim-dressed boards were milled into final 

products with dimensional affects making the boards unacceptable at any grade under the relevant 

Australian standard. Further, quartersawn boards of both species cut at normal full length (5.4 – 6.0 m) 

can be expected to incur significant downgrade from spring distortion. These combined results cast 

doubt on assumptions of board and value recovery from other studies based on the assessment of 

skim-dressed material only. 

product value (/m3 log) 

$400 

$350 

$300 

$250 

$200 

$150 

$100 

$50 

$0 

log grade 1 log grade 2 log grade 3 log grade 4 log grade 5 log grade 6 

sawlog grade 

E. nitens 26yo - thinned, pruned 

butt log - ITC Newood 

E. nitens 26yo - thinned, pruned 

upper log (unpruned) - ITC 

Newood 

E. nitens 26yo - fibre - ITC 

Newood 

,E. globulus 19yo - thinned 

pruned butt log - ITC Newood 

Figure 11. Total product value per cubic metre of sawlog at ITC Newood



The economics analysis for the study found that the value of solid products recovered from the thinned 

and pruned E. globulus logs and probably logs salvaged from the fibre-managed E. nitens stands were 

likely to be sufficient for a suitable and sustainable return. See Figure 11. However, growing and 

processing the thinned and pruned E. nitens was marginal or uneconomic, mainly due to the loss of 

value from internal checking, especially in timber from the pruned butt log. 

In a further study, 21-year-old trees, pruned at age six years to a height of 6.4 m, were selected from a 

silvicultural trial in north-east Tasmania which tested five stockings (100, 200, 300, 400 stems per 

hectare (SPH) and un-thinned control with ~700 SPH (Washusen et al. 2007). The logs were milled 

and the recovered boards were dried in a research predryer before final drying in a commercial kiln. 

On assessment, internal checking was found to be a common defect in the boards and was 

significantly more prevalent in boards from the lower logs than from the upper log. Frequency of 

occurrence did not differ significantly between the two sawing methods (see Table 6). However, the 

mean numbers of internal checks per board was significantly greater for back-sawn boards than 

quarter-sawn boards, as well as there being significantly more checks in the boards from the lower 

logs than the upper logs for both sawing categories. 

Table 6. Percentage of boards displaying internal checking 

Lower log Upper log 

Back-sawn 73.3 34.3 

Quarter-sawn 73.1 23.1 

CONCLUSIONS 

This paper focuses on the solid wood products drawn from native and plantation eucalypt forests, the 

associations these products have when used in artefacts such as buildings, and the potential for 

plantation eucalypts to maintain those associations and support (or maintain) a viable processing 

industry. 

It links together the supply chain for these solid hardwood products in Australia: from the user seeking 

to enjoy an artefact; the producer suppling the material for the artefact, and the grower providing the 

raw material. Because of its visual qualities and its inherent inability to compete in more competitive 

markets, solid hardwood is used primarily as an appearance product in this country. It demands a high 

price in the market because of its innate attractions in building and especially in applications close to 

the eye, the hand and the touch. The key wood properties for these applications, established during 

centuries of practice, are colour, grain, feature, workability and critical aspects of functionality. 

The industry milling and drying of these products depends on the premium they demand over other 

wood products, as the cost of producing hardwood is systemically more expensive than milling 

softwood. To mill these products, the industry requires a resource, and in Tasmania, 50% of the high 

quality log supply to be provided to industry from state forests from 2020 is predominantly plantation 

E. nitens. 

Unfortunately, given current technology, this resource is not likely to be suitable for the applications 

on which the current hardwood production industry depends. The boards either distort too much to be 

acceptable for high quality appearance products, or suffer from an unacceptable and unrecoverable 

drying degrade — that of internal check. These large hurdles overwhelm consideration of more subtle 

price and recovery influences, such as colour, stability, hardness and log size. 

This mismatch of resource and market is likely to have several effects. Australia’s designers (or more 

correctly, suppliers acting on their behalf) will import hardwood from overseas to overcome the lack 

of a suitable local product. The supply of unique wood from Tasmania’s forest, already significantly 

constrained, will decrease further. Lacking a resource, a proportion of hardwood producers will cease 

to trade. Others will continue to operate with the remaining native resource and some proportion of 

recovered plantation wood. Some will attempt to compete in the structural market but they will need to 

pursue very selective production strategies to overcome the added complications and associated costs



of hardwood production. The grower will be left to try to recover some value from the standing 

resource. However, the economics of this will be interesting, given the expensive silvicultural regimes 

pursued to produce the logs in the first instance. 

There is potentially a technological solution to processing the resource currently in the ground. 

However, current indications are that this solution will require a radical rather than incremental change 

to current production techniques. Recent studies have generally failed to reveal any statistically 

relevant differences to exploit in results from variations on current approaches. Unfortunately, it 

appears that E. nitens wants to check, especially in timber from the butt log, and E. globulus wants to 

distort. 

REFERENCES 

Aalto, A 1956, ‘Wood as a building material’, Arkkitehti-Arkkitekten in Schildt, Sketches. 

Banhan, R 1972, ‘Is there a substitute for wood grain plastic’ in Anderson EA & and Earle GF (ed.), Design and 

aesthetics in wood, State University of New York Press, Albany. 

Earle, GF 1972, ‘Is there a substitute for wood grain plastic’ in Anderson EA & and Earle GF (ed.), Design and 


Forestry Tasmania 2007, Sustainable High Quality Eucalypt Sawlog Supply from Tasmanian State Forest: 

Review No. 3, Forestry Tasmania Planning Branch, August. 

Innes, T and Greaves, BL and Nolan, G & Washusen, R 2008, Determining the economics of processing 

plantation eucalypts for solid timber products, Forest & Wood Products Research & Development 

Corporation, PN04.3007 

Innes, T Armstrong, M & Siemon G 2005, The impact of harvesting age on sawing, drying and solid wood 

properties of key regrowth eucalypt species, Forest & Wood Products Research & Development 

Corporation, PN03.1316 

Kock 1972, ‘Design and the Product in the Future Economy’ in Anderson EA & and Earle GF (ed.), Design and 


Marra 1972, ‘Wood Products in the Future – A Technological Extrapolation’ in Anderson EA & and Earle GF 

(ed.), Design and aesthetics in wood, State University of New York Press, Albany. 

Menin, S., Flora, S 2003, Nature and Space: Aalto and Corbusier, Routledge 

Nolan, G Greaves, BL Washusen, R & Parson M 2005, Plantations for Solid Wood Products in Australia – A 

Review, Forest & Wood Products Research & Development Corporation, PN04.3002 

Nolan, GB 1994, The Culture of using Timber as a Building Material in Australia, Proceedings, the Pacific 

Timber Engineering Conference, Surfer's Paradise, July 

Nolan, GB 1998, ‘Specifying natural feature timber’, 1st International Furniture Technology Conference 

Proceedings, 1998, Sydney, pp. 9. 

Standards Australia 1999, Australian Standard AS2796: Timber – Hardwood – Sawn and Milled Products, 

Standards Australia, Sydney. 

Washusen, R Harwood C, Morrow A, Valencia JC, Volker P, Wood M, Innes T, Ngo D, Northway R & 

Bojadzic M 2007 Technical report 168: Gould’s Country Eucalyptus nitens thinning trial: solid wood 

quality and processing performance using conventional processing strategies, CRC Forestry, Hobart.



WOOD PROPERTY VARIATION IN MATURE QUEENSLAND 

EXOTIC PINE SILVICULTURE EXPERIMENTS 

Kevin Harding 1 , Terry Copley 1 , Marks Nester 2 , Kerrie Catchpoole 2 & Anton Zbonak 1 

ABSTRACT 

Three mature silvicultural experiments that produced large differences in growth rate 

responses were sampled to investigate the impact of these treatment responses on wood 

properties among and within trees. These experiments were a Caribbean pine 

cultivation and weed control experiment (26-years-old), an exotic pine mounding by 

taxa experiment (21-years-old) and a Caribbean pine spacing by thinning experiment 

(28-years-old). Results for wood density varied in consistency but overall displayed a 

small negative trend with growth rate. Microfibril angle tended to increase with faster 

growth rate but across the trials increases were small and produced inconsistent trends. 

Overall the responses in these wood properties were small compared to the large 

DBHOB differences evident in the experiments. Taxa differences studied in one 

experiment displayed a negative relationship between average DBHOB and average 

wood density ranking from highest to lowest: slash pine seedlings, F2 hybrid seedlings, 

F1 hybrid cuttings and Caribbean pine seedlings. 

INTRODUCTION 

In 2005, Forestry Plantations Queensland (FPQ) and the Wood Quality Improvement group within the 

Product Quality section of Horticulture and Forestry Science (H&FS), DPI&F, commenced a long 

term project to obtain samples from three mature silviculture/genetics trials to collect data on tree 

growth and wood quality. The main goal was to collect robust data from the three trials that would 

underpin wood property and growth modeling as part of an overall decision support system being 

developed by FPQ (Catchpoole et al. 2007). By modeling growth rate and wood property interactions 

produced by the silvicultural treatments present in these trials it should be possible to identify an 

optimised silvicultural regime to deliver the best structural grade recovery and value returns able to be 

produced from exotic pine plantings in Queensland. 

The data collection focused on within and between-tree variation to consider the impacts of different 

silvicultural factors on wood density, microfibril angle and juvenile wood proportion, both at breast 

height, and at multiple heights within the merchantable stem. The sampled experiments (132BOW, 

195MBR and 532NC) provided opportunities to examine the effect of site preparation and weed 

control, mounding and taxon , and initial spacing and on wood density, microfibril angle and juvenile 

wood proportion. The experiments were targeted due to their maturity (20 to 28 years old), their 

design and their observed large differences in growth rates in response to the treatments applied. This 

variation provided excellent scope to evaluate and model the relationship between large productivity 

responses and wood traits. 

The first experiment, Expt 132BOW, is a Caribbean pine cultivation and weed control trial planted in 

1980 and wood sampled at age 26 years. Expt 195MBR is an exotic pine mounding by taxa trial, 

comparing slash pine, Caribbean pine, F1 hybrid cuttings and F2 hybrid seedlings, planted in 1985 and 

wood sampled at age 21 years. Expt 532NC is a Caribbean pine spacing and thinning trial planted in 

1977, wood sampled at age 28-29 years, and used in a sawing study at age 30 years. 

The use of these designed and controlled experiments to study silviculture and growth rate effects on 

wood quality is much more cost effective than attempting to model these responses in routine 

1 

Horticulture and Forestry Science, Department of Primary Industries and Fisheries, Queensland 

2 

Forestry Plantations Queensland



plantation stands. The confounding influence of site (both within and between site differences) age 

and spacing/thinning/fertiliser history means much larger samples of trees on many more sites would 

need to be studied to attempt to define the significance of responses. The latter approach very quickly 

becomes prohibitively costly and lacks the scientific rigour needed to produce repeatable results. 

MATERIAL 

Silviculture experiments 

Table 1. Relevant experiment details for the selected treatments sampled in three experiments 

grown in coastal southern and central Queensland. 

Experiment 132BOW 195MBR 

PEE seedlings, PEE × 

532NC 

Species† 

and stock type 

PCH seedlings 

PCH F1 cuttings, PCH 

seedlings and PEE × PCH 

F2 seedlings 

PCH seedlings. Kennedy 

seed orchard batch T251 

Location 

Byfield 

22˚ 48’ S, 150˚ 39’ E 

Tuan 

25˚ 35’ S, 152˚ 45’ E 

Toorbul 

27°05'S 152°58'E 

Planted 1/1980 at 3.0 × 3.0 m 6/1985 at 5.0m × 2.1m 2/1977 

Date sampled - 

wood properties 

5/2006 3/2006 6/2005 

Spacing/stocking; 

Treatments Ploughing; Weed Control Mounding 

Thinning (at age 10 

years) 

Ploughing … Nil; 

Ploughed 

Weed control … Nil, 

Treatment levels woody weeds on inter- Nil, 0.5 m 

rows only, all woody 

weeds, and total weed 

control 

2 

Spacing (stocking) 

…1.8m 2 ( 3088spha), 

2.4m 2 (1737spha), 3.0m 2 

(1111spha) and 3.6m 2 

(772 spha) 

Thin … Nil, 200, 400 

and 600 spha 

† PCH = Pinus caribaea var. hondurensis; PEE = Pinus elliottii var. elliottii; PEE × PCH = interspecific hybrid cross 

METHODS 

Standing tree assessments 

In each experiment sample trees were selected to represent the diameter at breast height over bark 

(DBHOB) distribution in the treatment combinations studied. Each tree was assessed for standing tree 

acoustic velocity, height, and DBHOB and a 12mm core was extracted. Two acoustic velocity 

measurements (ST300) were obtained on opposite sides of each tree in approximately the same 

alignment as the increment core sample was taken; i.e. along the shortest diameter to minimise 

inclusion of compression wood. The wood cores were removed from the closest mid-whorl below 

breast height. All wood cores were resin extracted in methanol and hot water in a soxhlet apparatus 

using a boiling cycle of 8 hours in methanol, followed by 10 hours in hot water. 

In Expt 532NC approximately 105 trees per thinning treatment were assessed providing 403 trees in 

total due to incomplete survival or unacceptable tree form, size or damage. Numbers sampled varied 

across initial spacing × thinning treatment combinations as surviving tree numbers were small in 

widely spaced and heavily thinned plots and were large in unthinned plots. In Expt 195MBR 20 trees 

were sampled for each taxon × mounding combination (160 trees in total). Fifteen trees per weed 

control treatment plus the nil plough × weed control treatment used as a control were sampled in two 

replications of Expt 132BOW (150 trees in total). 

SilviScan 

Based on the DBHOB and gravimetric density results, 80 trees (5 trees per treatment) from 532NC 

and 40 trees (10 trees per taxon) from 195NC were systematically selected for felling and analysis of 

wood properties with height within the stem. For this study only results for the breast height samples 

have been reported. The best defect-free radius of each core was used for SilviScan analysis. X-ray



scanning was performed to capture density readings at every 50 microns from pith to cambium along 

each radial sample. The positioning of growth ring boundaries were assigned by comparing SilviScan 

pith to bark density patterns to standard profiles developed using a selection of samples with 

prominent and readily discernable growth ring patterns, as well as reference to actual samples under 

magnification. This combined approach was used to minimise the erroneous inclusion of false growth 

rings in the individual growth ring data. Radial increments to each growth ring boundary were used to 

estimate the basal area of each growth ring and to provide basal area weighted averages of density 

results. The controlled environment used for x-ray scanning equates to approximately 8% moisture 

content in the scanned samples so SilviScan density refers to density assessed at this moisture content. 

Analysis 

Analysis of variance was undertaken using GenStat (2002). 

RESULTS 

Expt 132BOW 

Ploughing was found by Debuse and Nester (2004) to have had a significant effect on the impact of 

weed control on mean adjusted DBHOB at age 18.1 years (weed control x ploughing: F 3,20 = 3.2, P = 

0.045). At age 18.1 years, complete weed control in combination with ploughing yielded significantly 

greater adjusted DBHOB measures than all other treatments, with the exception of ‘nil plough + 

complete weed control’, and ‘plough + complete woody weed control’ treatments. These DBHOB 

results for the full set of experiment treatments indicated large differences in growth rate between the 

control treatment (nil plough and nil weed control) and the four ploughed weed control treatments that 

were sampled for this wood study. 

Analysis of variance revealed no significant difference among sampled ploughing × weed control 

treatments for area-weighted SilviScan means of wood density (F4,145=1.05, P= 0.385) and microfibril 

angle (F4,145=1.57,P=0.186) estimated on breast height radial core samples. Wood density displayed a 

consistent negative trend with more intensive weed control treatments which reflect growth rate 

improvements. Microfibril angle increased with better weed control from around 13 degrees in the 

control to about 14.6 degrees in the treatments with more intensive weed control. Trends for these 

traits with relative distance from the pith are plotted in Figure 1. 

Wood density (kg m -3 ) 

750 

700 

650 

600 

550 

500 

450 

400 

350 

Nil plough-Nil weed Plough-1 nil weed Plough-2 Partial woody weed 

Plough-3 Complete woody weed Plough-4 Complete weed control 

MfA (deg) 

35 

30 

25 

20 

15 

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 

Wood density MfA 

Relative distance from pith (0%) to bark (100%) 

Figure 1. Wood density and microfibril angle variation among Expt 132BOW plough × weed 

control treatments relative to proportional distance from pith. Weed control 

treatments: 1=nil; 2=woody inter-row weeds; 3= all woody weeds and 4 = complete.



Treatment (Number) Nil ploughing Ploughing 

Nil weeding 

(1) 

Partial woody weed 

removal and inter-row 

cultivation 

(2) 

Complete woody weed 

removal, inter-row 

cultivation, and basal 

spray 

(3) 

Complete weed control - 

grass and woody weeds 

(4) 

Wood density zones: 

400 – 500 kg/m 3 

500 – 600 kg/m 3 600 – 700 kg/m 3 

Figure 2. Graphic illustration of the average relative proportion of wood density zones within 

Expt 132BOW treatments. Tree cross-sections have been scaled to indicate relative 

mean size of trees in the treatments. 

Using the actual growth ring increment values from SilviScan results, Figure 2 was constructed to 

relative scale to illustrate the differences between treatments in both average tree size and proportions 

of wood density zones within each.



Although arbitrarily assigned, the three wood density zones defined approximate juvenile, 

intermediate and mature wood values. Although the incremental increase in tree size is very small 

from weed control treatment 2 to treatment 3, the proportion of wood in the lowest density zone varies 

markedly. Treatment 3 has resulted in a lower density inner core than other treatments, which is likely 

to have negative implications for the grade recovery of sawn framing from such a regime. 

As significant amounts of the higher density mature wood zone are lost in squaring up the log during 

sawing, the strength and stiffness properties of timber framing sawn from trees displaying these 

differences are likely to vary considerably. Ongoing analysis and data modelling of these relationships 

will be undertaken for the decision support system development that these studies support. The latter 

will need to consider the potential to gain economic returns from optimising silviculture regimes to 

balance average plantation stem volume production and the quality and profile of the wood produced. 

The density profiles observed within this experiment suggest that there may be considerable scope for 

optimisation. 

Expt 195MBR 

This experiment compared three mound sizes (0.5m 2 , 1.0m 2 and 1.5m 2 ), which Debuse and Nester 

(2002) found had a significant effect on DBHOB growth for all taxa combined at age 16 years (F 3,5 = 

9.2, P = 0.018). Trees planted on all sizes of mound showed significantly greater DBHOB than those 

planted on no mounds. However, this mounding effect did not vary significantly among taxa (F 12,31 = 

0.5, P = 0.898) and therefore only the 0.5m 2 mounds were sampled in the wood study as these were 

closest in size to the current commercial practice. DBHOB measures differed significantly among taxa 

for all mound sizes combined at age 16 years (F 4,31 = 86.2, P < 0.001). PCH exhibited significantly 

larger DBHOB measures than all other taxa, while PEE had significantly smaller diameters than the 

other taxa (Debuse and Nester, 2002). 

Analysis of variance revealed a significant difference among taxa for breast height area-weighted 

SilviScan mean wood density (F3,152=45.70,P=0.001) but no significant difference due to mounding 

(F1,152=1.21,P=0.274) or for the interaction between these treatments (F3,152=0.56,P=0.645). There was 

also no significant difference in breast height area-weighted mean microfibril angle due to taxa 

(F3,152,P= 0.541), mounds (F1,152=2.85,P=0.094) or treatment interaction (F3,152=1.27,P=0.287). The 

area-weighted averages obtained from SilviScan analysis are presented in Figures 3 and 4. 


750 

700 

650 

600 

550 

500 

F1 CUT F2 PCH PEE 

Taxon 

Nil mound 

0.5m^2 mound 

Figure 3. Comparison of average area weighted wood density for taxa planted on nil mounds 

(treatment 1) and 0.5m 2 mounds (treatment 2) in Expt 195MBR.



Microfibril angle (deg) 

18 

17 

16 

15 

14 

F1 CUT F2 PCH PEE 

Taxon 

Nil mound 

0.5m^2 mound 

Figure 4. Comparison of average area weighted microfibril angle for taxon planted on nil 

mounds (Treatment 1) and 0.5m 2 mounds (treatment 2) in Expt 195MBR. 

The wood density ranking among taxa is in agreement with the findings of Kain (2003) and Harding 

and Copley (2000). Although it is not statistically significant, the parental taxa appear to show a 

response to mounding for microfibril angle that is much less pronounced in the hybrid taxa. The latter 

is likely to reflect better growth rate achieved on mounds. 

Wood density trends for the taxa are plotted in Figure 5. Although the relative trends among the taxa 

are similar both with mounds and without there are larger differences among the taxa on the 

unmounded site, and these differences are more accentuated in the outer wood. 


900 

800 

700 

600 

500 

400 

300 

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 

Nil mound 0.5m^2 mound 

Relative distance from pith to bark 

F1 CUT 

F2 

PCH 

PEE 

Figure 5. Air dry density variation relative to proportional distance from pith for taxa in Expt 

195MBR planted with and without mounds. 

Expt 532NC 

The DBHOB trends with age plotted in Figure 6 indicate the large differences in growth rate observed 

among treatments that were studied. The magnitude of these DBHOB differences rendered this



experiment an ideal candidate for wood property sampling to provide the opportunity to investigate 

wood property response to very large differences in growth rate. 

DBHOB (cm) 

50 

45 

40 

35 

30 

25 

20 

15 

10 

5 

0 

0 5 10 15 20 25 30 

Age (years) 

S 772 - UT 

S 1111 - UT 

S 1736 - UT 

S 3086 - UT 

S 772 - Thin to 600 

S 1111 - Thin to 600 

S 1736 - Thin to 600 

S 3086 - Thin to 600 

S 772 - Thin to 200 

S 1111 - Thin to 200 

S 1736 - Thin to 200 

S 3086 - Thin to 200 

S 772 - Thin to 400 

S 1111 - Thin to 400 

S 1736 - Thin to 400 

S 3086 - Thin to 400 

Figure 6. DBHOB (cm) trends with age for unthinned initial spacing × thinning treatments 

sampled for wood studies in Expt 532NC. 

Wood density responses to the pre- and post-thinning stocking combinations illustrated in Figure 7 are 

not consistent across thinning treatments within initial spacing treatments. There is a tendency for 

area-weighted density to be higher in the more heavily thinned treatments. However, analysis of 

variance produced no significant effects for initial stocking (F3,49=1.49, P=0.229), for postthinning 

stocking (F3,49=1.76, P=0.168) or for the interaction of these treatments (F9,49=0.49, 

P=0.871). 


650 

620 

590 

560 

530 

500 

772 spha 1111 spha 1737 spha 3088 spha 

Initial stocking 

200 spha 

400 spha 

600 spha 

Unthinned 

Figure 7. SilviScan area-weighted wood density variation in Expt 532NC with pre- and postthinning 

stocking levels.



The underlying trend with age plotted in Figure 8 is strong and clearly illustrates juvenile and mature 

wood density zones while also indicating the variability among the treatments. 

Basic density (kg m 3 ) 

600 

550 

500 

450 

400 

350 

300 

Pith-5 Rings 06-10 Rings 11-15 Rings 16-20 Rings 21-25 Rings 26-bark 

Growth ring segment 

772 - 200spha 

772 - 400spha 

772 - 600spha 

772 - Unthinned 

1111 - 200spha 

1111 - 400spha 

1111 - 600spha 


1737 - 200spha 

1737 - 400spha 

1737 - 600spha 


3088 - 200spha 

3088 - 400spha 

3088 - 600spha 


Figure 8. Basic density variation from pith to bark for initial stocking × thinning treatment 

combinations in Expt 532NC 

There is an indication that microfibril angle is influenced by growth rate and analysis of variance 

found a significant response to initial stocking (F3,49=9.15, P=0.001) which is reflected in the negative 

trend in Figure 9 for area-weighted microfibril angle with initial stocking level. No significant 

difference was found for post-thinning stocking level (F3,49=0.39, P=0.760) or for the initial stocking × 

post-thinning stocking interaction (F9,49=1.12, P=0.366). 

Microfibril angle (deg) 

20 

18 

16 

14 

12 

10 

772 spha 1111 spha 1737 spha 3088 spha 

Initial stocking 

200 spha 

400 spha 

600 spha 

Unthinned 

Figure 9. Area-weighted microfibril angle variation in Expt 532NC with pre- and post-thinning 

stocking levels. 

The results of a sawing study on 56 stems from this experiment have been reported separately 

(Harding et al. 2009) and draw on wood property results from this study to consider relationships 

between standing tree wood properties and sawn product quality and value.



CONCLUSION 

Wood property variation in the three mature silviculture experiments studied has revealed some 

significant variation with growth rate. However, given the size of the growth rate responses observed 

in these experiments (particularly between control treatments and the silviculture and/or taxon 

treatments that have provided the largest differences in DBHOB), the average changes in wood 

density and microfibril angles observed are relatively small. The data collected provide a very detailed 

data set for ongoing wood property modelling, which will underpin decision support systems for 

Queensland exotic pine plantations. The practical significance of the results observed requires detailed 

analysis and conversion modelling to refine predicted impacts on product recovery, grade and value. 

However, it is clear that there is considerable potential to gain economic returns from optimising 

silvicultural regimes to balance average plantation productivity and wood quality. The latter 

opportunities require the pith to bark wood density and wood properties profiles of the trees to be 

monitored and managed. Links are needed to key environmental drivers and their interaction with 

silvicultural regimes to optimise the quality and value returns from these plantations. 

REFERENCES 

Catchpoole, K.J., Nester, M.R. and Harding , K.J. (2007) Predicting wood value in Queensland Caribbean pine 

plantations using a decision support system. Aust. J. Forestry 70(2): 120-124. 

Debuse, V.J. and Nester, M.R. (2002) The effect of mound size on the growth of Pinus elliottii var elliottii, 

Pinus caribaea var hondurensis and their hybrids on poorly drained sites.– Experiment 195MBR. 

Unpublished confidential QFRI project report for DPI Forestry, Queensland. 27pp. 

Debuse, V.J. and Nester, M.R. (2004) The effect of weed competition and cultivation on growth and early 

survival of Pinus caribaea var. hondurensis – Experiment 132BOW. Unpublished confidential QFRI 

project report for DPI Forestry, Queensland. 22pp. 

GenStat (2002) GenStat for Windows, version 6.1 Lawes Agricultural Trust. 

Harding, K.J. and Copley, T.R. (2000) Wood property variation in Queensland-grown slash ×Caribbean pine 

hybrids. In “Hybrid Breeding and Genetics of Forest Trees” Proceedings of QFRI/CRC-SPF Symposium, 

9-14 April 2000, Noosa, Queensland, Australia. (Compiled by Dungey, H. S., Dieters, M. J. and Nikles, D. 

G.), Department of Primary Industries, Brisbane. 

Harding, K.J., Copley, T.R., Nester, M.J., Catchpoole, K.J. and Zbonak, A.(2009) Caribbean pine graded sawn 

recovery variation in a mature spacing by thinning experiment grown in Queensland. Paper submitted for 

presentation to the Institute of Foresters of Australia National Conference, 6-10 September 2009, 

Caloundra, QLD 

Kain, D.P. (2003) Genetic parameters and improvement strategies for the Pinus elliottii × Pinus caribaea var. 

hondurensis hybrid in Queensland Australia. PhD Thesis, Australian National University. 456pp.



ABSTRACT 

MODELLING THE INFLUENCE OF CLIMATE CHANGE 

ON PLANTATION WOOD PROPERTIES 

Robert Evans 1 , Josh Bowden 1 , Kevin Harding 2 , Terry Copley 2 , 

Kerrie Catchpoole 3 , and Marks Nester 3 

As world-wide climate change is accelerating, the need for understanding forest 

responses to these changes is of increasing importance for commercial and environmental 

risk management. Development of models to predict wood quality changes for a range of 

species over a range of predicted climatic conditions is a step towards optimisation of the 

value of future forest resources. The modelling is based on statistical analysis of growth 

and wood properties in many trees over a range of known climatic conditions. Wood 

quality data were obtained from a range of Pinus caribaea var. hondurensis samples 

using the SilviScan system. The samples were from a Forestry Plantations Queensland 

exotic pine thinning and spacing trial. This study is a work in progress and begins by 

showing the qualitative connections between environmental variation and wood property 

variation (density, tracheid diameter, microfibril angle, wall thickness, coarseness and 

modulus of elasticity). The plantation growth model 3-PG has been used to give 

predictions of annual growth for use in mapping wood properties onto a time axis. All 

properties are shown to be strongly affected by short-term variations in weather, 

especially rainfall. False rings in earlywood, produced in dry conditions, result in 

increased density and modulus of elasticity, and decreased microfibril angle and tracheid 

diameter. 

INTRODUCTION 

1.2 million cubic metres of softwood sawlog are produced commercially in Queensland every year 

from trees that have experienced variations in climate over decades. Once the seeds (or cuttings) and 

sites have been chosen, the quantity and quality of the wood are controlled by weather and 

management practices such as thinning and application of fertiliser. Commercial forests are valued on 

the quantity and quality of the wood produced, therefore climate variations have a direct impact on the 

value, and sometimes the viability, of these resources (Broadmeadow et al., 2005). 

By combining SilviScan analyses with climate model predictions (for example, the Australian 

Community Climate and Earth-System Simulator (ACCESS) (Kowalczyk et al. 2006)) we aim to 

estimate the effects of climate change on properties such as timber stiffness, strength, pulp yield and 

paper density and strength. The premise is that the current (and recent) variation in climate and 

corresponding wood properties over the range of commercial wood resources is sufficient to generate 

relationships that will allow the prediction of the effects of climate change using existing climate 

models. 

In the last decade, very large numbers of wood samples from across Australia and around the world 

have been measured by SilviScan. Encoded in these results are climate signals that could be used to 

find climate/wood property relationships. In this context, ‘climate’ initially involves temperature and 

rainfall, which are two of the most important extrinsic variables affecting tree growth and wood 

structure. The goal is to find climate signals in the noise arising from the many uncontrolled and 

unmeasured variables (e.g. variations in site, microclimate and genetics), and eventually to generate 

1 

CSIRO Materials Science and Engineering. Bayview Ave Clayton, Victoria. 

2 

Queensland Primary Industries and Fisheries, Department of Employment, Economic Development and Innovation, 

Townsville and Indooroopilly, Qld. 

3 

Forestry Plantations Queensland, Brisbane and Gympie, Qld.



predictive models that can be applied to different species over a range of sites in Australia and around 

the world. 

We will here describe the data, methods and initial observations for a set of trees collected from a 

South East Queensland Pinus caribaea var. hondurensis trial. 

MATERIAL AND METHODS 

A set of 30-year-old Pinus caribaea var. hondurensis samples from Beerburrum, in south-east 

Queensland (about 60km north of Brisbane), analysed in a previous project, was found to have a 

complex ‘false ring’ structure resulting from environmental variation over many years. These rings are 

regions of high wood density which occur within the normal annual ring growth of a tree, either before 

or during the normal rise in density at the end of a season, and is thought to occur at the cessation of 

shoot and needle elongation (Chudnoff and Geary, 1973). Twelve radial samples from an original 520 

(65 trees, 8 heights per tree) were selected for high resolution analysis on SilviScan-3 at the 4m 

sampling level in 12 trees. The samples were in the form of 2mm wide (tangential dimension) and 

7mm high (longitudinal dimension) pith-to-bark (radial) strips. 

They were scanned in three stages. First, automated microscopy and image analysis of the polished top 

(radial-tangential) plane gave growth-ring boundary orientation, ray orientation, radial tracheid 

diameter and tangential tracheid diameter at 25 micron intervals. Second, automated x-ray scanning 

densitometry gave density variation (10 micron intervals), normalised to the actual average sample 

density obtained by independent gravimetric analysis. Third, automated scanning x-ray diffractometry, 

using 100 micron step size, gave microfibril angle (MFA) and together with the density information 

was used to estimate longitudinal modulus of elasticity (MOE). Tracheid wall thickness and 

coarseness were calculated from tracheid diameter and density on the assumption that dry cell wall 

density is constant at 1500 kg/m 3 . All properties were placed on a 25 micron step scale. Annual ring 

boundary identification was performed interactively to avoid the false rings. An annual ring boundary 

was generally recognisable by the large, rapid drop in density going from latewood formed at the end 

of one growing season to the earlywood of the next growing season.. 

Weather data for the Beerburrum site (27.05S, 153.00E) were interpolated using 0.05 degree steps 

from nearby weather station information by the Australian Bureau of Meteorology Silo Data Drill, and 

consisted of maximum temperature, minimum temperature, rainfall, evaporation, radiation, vapour 

pressure, maximum relative humidity and minimum relative humidity. 

The Keetch-Byram Drought Index (KBDI, http://www.kruckysweather.org/kbdi.htm) is an indicator 

of water deficiency commonly used in bush fire research. KBDI integrates weather conditions such as 

temperature and rainfall. During periods of high temperature and inadequate rainfall, this index 

increases rapidly as the soil dries out. The index returns to a low value after sufficient rain. Such 

intermittent conditions are responsible for the false rings seen in these trees. 

The wood property profiles were then remapped on to an annual scale by linear interpolation between 

ring boundaries. Intra-annual alignment was carried out using the false rings to ensure that subsequent 

averaging of the sample profiles did not eliminate the false ring structure. Prominent false ring 

positions were marked using the density profiles, identifying rising and falling edges as well as peak 

positions. The medians of all wood property profiles were then produced. As the ring boundaries are 

associated with the beginning of spring (July-August, or nominally 60% of the way through the 

calendar year – Shepherd et al., 2003) the time axes are annotated accordingly in the following Figures 

below. The range of years displayed was chosen to avoid the relatively featureless juvenile wood 

closest to the pith, the thinning operation in 1987, and crown closure. 

RESULTS AND DISCUSSION 

Figure 1 shows the radial profiles for MOE (Evans, 2006), which combines information from 

diffractometry and densitometry. Density can also be considered a composite property of wood, 

combining cell wall thickness, cell diameter and cell wall density. The profiles in Figure 1, which also 

show the annual rings and false ring structures, may therefore be considered to represent a summary of



many of the properties measured on SilviScan. The lack of structural variation in the very juvenile 

wood is evident in these profiles. 

Under ideal conditions, Southern pines such as P. caribaea grow rapidly, displaying a distinctive 

“square-wave” pattern of alternating low density early wood and high density latewood. In these 

samples, the local weather patterns have caused intermittent wood growth to produce very complicated 

radial profiles of wood properties, and a relatively wide range of average growth rates. 

longitudinal modulus of elasticity 

0 50 100 

distance from pith / mm 

150 200 

Figure 1. Longitudinal modulus of elasticity profiles for the 12 selected radial samples. 

Radius / mm 

250 

200 

150 

100 

50 

0 

thinned 

1987 

1980 1985 1990 1995 2000 2005 

Year 

Figure 2. Individual radial profiles for the 12 samples. Thinned to 200 stems/ha (squares), 400 

stems/ha (diamonds), 600 stems/ha (triangles). Unthinned - 1100 stems/ha (solid 

circles).



The ring boundary positions were used to generate radial growth curves for the samples, as shown in 

Figure 2. The plots from which the samples were taken had been thinned to 200, 400 and 600 stems/ha 

(plus unthinned control) in 1987 at age 10. The change in rate of growth after thinning increased with 

thinning intensity. It should be noted that the false ring patterns were not affected by the level of 

thinning. As ring boundary positions were difficult to pinpoint within the juvenile wood (Fig 1), only 

data after 4 years is shown in Figure 2. 

An initial attempt to connect the annual growth of the samples with the weather was made using the 

modelling software 3-PG (Landsberg et al., 2001, Coops and Waring 2001, Coops et al., 2005) and 

results are shown in Figure 3. Manual parameterisation of 3-PG was carried out by modifying the 

parameters for P. radiata as shown in Table 1. Note that Tmax, Tmin and Topt are dramatically 

different from those of the temperate climate pine. Figure 3, which also leaves out juvenile growth 

prior to 1982, shows a reasonable prediction of the 3-PG model of annual yearly increment, although 

some features, such as the spike in growth during 1999-2000, are inadequately modelled. The adjusted 

parameters reasonably predict yearly increments, but the prediction of monthly growth is far from 

optimal, indicating further work is needed before 3-PG can be used to map wood properties onto a 

time scale. Optimisation of the parameters Tmin, Topt and Tmax for this species over a range of sites 

with varying climate along the Queensland coast would improve the 3-PG model and is now in the 

planning stages. 

ring width / mm 

1.2 

1 

0.8 

0.6 

0.4 

0.2 

0 

1980 1985 1990 1995 2000 2005 2010 

year of growth 

Figure 3. Average ring widths for the 12 samples predicted by 3-PG (diamonds) using the 

available weather conditions and 3-PG parameters modified from radiata pine as 

seen in Table 1. Actual ring widths are also shown from SilviScan data (circles). 

Table 1. 3-PG parameters for P. radiata and modified for P. caribaea. 

3-PG parameters P. radiata P. caribaea 

(this study) 

Foliage: stem partitioning ratio @ D=20 cm (pFS20) 0.4 0.5 

Constant in the stem mass v. diam. relationship (aS) 0.0243 0.11 

Power in the stem mass v. diam. relationship (nS) 2.72 2.7 

Maximum fraction of NPP to roots (pRx) 0.6 0.95 

Minimum fraction of NPP to roots (pRn) 0.25 0.45 

Minimum temperature for growth (Tmin) / °C 0 8 

Optimum temperature for growth (Topt) / °C 20 32 

Maximum temperature for growth (Tmax) / °C 32 38 

Relative age to give fAge = 0.5 0.5 0.55



Minimum basic density - for young trees (rhoMin) / t/m 3 0.480 0.500 

Maximum basic density - for older trees (rhoMax) / t/m 3 0.480 0.600 

Age at canopy cover 0 24 

3-PG Initialisation data: 

Fertility rating 0.6 

Initial stocking / stems/Ha 1100 

Soil class Sandy loam 

Max avail. soil water/ mm 250 

Initial avail. soil water 200 

Silvicultural event Age = 10 Stocking = 600 

Although many properties were measured by SilviScan, interpretation of their variation is greatly 

complicated by strong local correlation between properties. An example is shown in Figure 4. MFA 

and density are inversely correlated in general. The quantitative nature of the correlation varies over 

time and between samples, so that one property cannot be estimated from the other. Such correlations 

affect the interpretation of variation in all wood properties. 

density / kg/m3 

1200 

700 

200 

-300 

wood density (upper line), MFA (lower line) 

-800 

0 20 40 60 80 100 120 140 

distance from pith / mm 

Figure 4. Microfibril angle and density radial profiles for a typical radial sample. 

Mapping on to a time scale 

For comparison with weather / climate information, the wood property data should be presented on a 

time scale. No detailed intra-annual growth information was available for these trees, therefore 

converting the distance-based property profiles into time-based profiles could only be done using 

annual ring boundaries. 

density kg/m3 

1200 

700 

200 

-300 

-800 

5 

wood density (upper line), MFA (lower line) 

1989.6 1990.6 1991.6 1992.6 1993.6 1994.6 1995.6 1996.6 1997.6 1998.6 1999.6 

Figure 5. Median profiles of wood density and MFA showing their complementary local 

variation, including that caused by environmental influences. The profiles have been 

linearly remapped on to annual intervals. 

year 

After the mapping as described in the Methods section, the intra-annual position of wood property 

features is not strictly time-based because radial growth rate is not constant within a year. Faster 

30 

25 

20 

15 

10 

60 

50 

40 

30 

20 

10 

0 

MFA / deg 

MFA / deg.



spring growth, for example shifts the features to the right between the annual ring boundaries. This 

should be noted for all Figures from Figure 5 onwards. 

density / kg/m3 

1200 

700 

200 

-300 

-800 

wood density (upper line), tracheid population density (lower line) 

1989.6 1990.6 1991.6 1992.6 1993.6 1994.6 1995.6 1996.6 1997.6 1998.6 1999.6 

year 

Figure 6. Environmental influences on median wood density and the median population density 

of tracheids. The profiles have been linearly remapped on to annual intervals. 

Figure 5 shows approximate time-based profiles for density and MFA. The inverse relationship 

between these properties is very clear. In Figure 6, wood density variation is compared with that of 

tracheid population (tracheids per mm 2 ). Note the similarity of the patterns of intra-annual features in 

Figures 5 and 6. The three properties (density, tracheid population and MFA) are measured 

independently. Environmental conditions have had a very large effect on the wood properties. Peaks in 

false ring density and tracheid population are often double the levels in the surrounding wood. MFA 

fell by up to a third in the same regions. 

KBDI is shown in Figure 7, together with wood density for a sequence of 9 years. There are many 

spikes in KBDI that could be associated with the density spikes (false rings), bearing in mind that the 

intra-annual positions of the false rings and the strictly time-based KBDI peaks are not expected to 

coincide, but lag by varying amounts. Wood radial growth is much more rapid in the first half of the 

year than in the second half, therefore it is expected that climate features in spring (true time scale) 

should occur before those in the property profiles. 

density kg/m3 

1200 

700 

200 

-300 

-800 

wood density (upper line), KBDI (lower line) 

1989.6 1990.6 1991.6 1992.6 1993.6 1994.6 1995.6 1996.6 1997.6 1998.6 1999.6 

Figure 7. Median wood density variation and drought index. The density profile has been 

remapped on to annual intervals for closer comparison with the time-based drought 

index. 

year 

Rainfall appears to be the limiting factor in the earlywood growth of fast-growing tropical pines (Slee 

1972). Figures 8 and 9 show rainfall accumulated over 30 days (moving mean) with profiles of density 

and radial tracheid diameter. Note that over any 30 day period, the accumulated rainfall rarely exceeds 

200mm, and never in early spring. The peaks in rainfall are not associated with peaks in density and 

decreases in tracheid diameter, but with the restoration of ‘normal’ wood properties. Rain is expected 

to bring an end to the false rings and increase tracheid diameter. 

Therefore, the peaks in accumulated rainfall should be associated with rapid decreases in density and 

rapid increases in tracheid diameter. As with KBDI, many of the features seen in the rainfall patterns 

are reflected in the false ring structures. 

2200 

2000 

1800 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

80 

70 

60 

50 

40 

30 

20 

10 

0 

KB drought index 

tracheids /mm2



density kg/m3 

1200 

700 

200 

-300 

-800 

wood density (upper line), rainfall (lower line) 

1989.6 1990.6 1991.6 1992.6 1993.6 1994.6 1995.6 1996.6 1997.6 1998.6 1999.6 

Figure 8. Median wood density profile remapped on to annual intervals, and rainfall 

accumulated over the previous 30 days. 

year 

Current efforts are directed towards the quantification of the effects of environmental variation on the 

wood properties at different scales. False ring structures that can be reliably associated with weather 

events will be used to estimate the magnitude of the local excursions in wood properties. Average 

wood properties within rings (and perhaps within earlywood and latewood) will be associated with 

climate, as well as climate variability, to generate relationships more suitable for the prediction of 

future macroscopic wood properties. 

tracheid radial 

diameter /um 

60 

50 

40 

30 

20 

10 

0 

tracheid radial diameter (upper line), rainfall (lower line) 

1989.6 1990.6 1991.6 1992.6 1993.6 1994.6 1995.6 1996.6 1997.6 1998.6 1999.6 

Figure 9. Median tracheid radial diameter remapped on to annual intervals, and rainfall 

accumulated over the previous 30 days (moving mean). 

CONCLUSIONS 

year 

• Growth modelling software 3-PG gave a qualitatively similar ring width pattern to that 

measured. Further efforts on 3-PG parameterisation to predict growth at higher 

(monthly) time resolution could improve the mapping of the distance scale on to the 

time scale. Accurate time scale mapping will improve statistical relationships between 

climate and wood properties, allowing more accurate predictions of future wood 

properties. 

• False ring structures associated with environmental variation are evident in all 

measured wood properties. 

• Initial indications are that the false rings result from an alternation between insufficient 

and sufficient rainfall for efficient growth. 

• Reduced rainfall during earlywood formation would be expected to increase density 

and MOE, decrease MFA and tracheid diameter. 

700 

600 

500 

400 

300 

200 

100 

0 

0 

30 day rainfall 

/mm 

700 

600 

500 

400 

300 

200 

100 

30 day rainfall /mm



REFERENCES 

Broadmeadow, M. S. J., D. Ray and. Samuel C.J.A. 2005, Climate change and the future for broadleaved tree 

species in Britain. Forestry, 78 (2), 145-161. 

Chudnoff, M. and Geary, T.F. 1973, Terminal shoot elongation and cambial growth rhythms in Pinus caribaea. 

Commonwealth Forestry Review, 52(4), 317-324. 

Coops, N.C. and Waring, R.H. 2001, Assessing forest growth across South-Western Oregon under a range of 

current and future global change scenarios using a process model, 3-PG. Global Change Biology, 7(1), 15- 

29. 

Coops, N.C., Waring, R.H. and Law, B.E. 2005, Assessing the past and future distribution and productivity of 

ponderosa pine in the Pacific Northwest using a process model, 3-PG, Ecological. Modelling, 183, 107– 

124. 

Evans, R. 2006, Wood stiffness by x-ray diffractometry. In: Characterisation of the cellulosic cell wall, Chapter 

11. Proceedings of the workshop 25-27 August 2003, Grand Lake, Colorado, USA. Southern Research 

Station, University of Iowa and the Society of Wood Science and Technology. D. Stokke and L. Groom, 

eds. Blackwell Publishing. 

Kowalczyk, E.A., Wang, Y. P., Law, R. M., Davies, H. L., McGregor, J.L. and Abramowitz, G. 2006, The 

CSIRO Atmosphere Biosphere Land Exchange (CABLE) model for use in climate models and as an offline 

model. CSIRO Marine and Atmospheric Research paper 013. 

Landsberg, J. J., Johnsen, K. H., Albaugh, T. J., Allen, H. L. and McKeand, S.E. 2001, Applying 3-PG, a Simple 

Process-Based Model Designed to Produce Practical Results' to Data from Loblolly Pine Experiments. 

Forest Science, 47(1), 43-51. 

Shepherd, M., Cross M., Dieters M. J., Harding K., Kain D. and Henry R. 2003, Genetics of physical wood 

properties and early growth in a tropical pine hybrid. Canadian Journal of Forest Research, 33(10), 1923– 

1932. 

Slee, M.U. 1972, Growth patterns of slash and Caribbean pine and their hybrids in Queensland. Euphytica, 

21(1), 129-142.



COMPETITION AND COLLABORATION 

IN THE AUSTRALIAN TIMBER FURNITURE INDUSTRY – 

A VALUE CHAIN APPROACH TOWARD HIGHER VALUE 

CREATION FOR LOCAL AND GLOBAL MARKETS AND PATHWAYS 

FOR SUSTAINABLE FOREST RESOURCE USE 

ABSTRACT 

Sasha Alexander 1 

This paper explores the development of an innovative value creation methodology which 

extend value chain analysis techniques that highlight opportunities for performance, 

profitability and inter-firm relationship improvements to a complete timber value chain 

for a furniture product, from forest to end-consumer. The research design employs a 

mixed method approach comprising qualitative industry survey interviews, and 

quantitative and qualitative industry and end-consumer surveys, guided by a value chain 

analysis approach. 

The findings support more direct engagement with end-consumer needs, providing value 

chain insights toward those timber furniture attributes which require the greatest 

attention in the creation of value in the timber value chain, including timber material 

sourcing and production phases. 

A key theme to emerge is the positive correlation of organisational commitment toward 

creating further value for timber end-consumers through environmental stewardship, and 

capitalising on timber’s unique end-consumer-valued attributes, with the end-consumer 

as central to the development of the timber value proposition. 

INTRODUCTION 

The need to increase the competitiveness of Australian firms participating in the timber furniture 

industry has become more critical in recent years with the marked increase in the volume of imported 

furniture products representing about AUD2 billion of a AUD5 billion furnishings industry (FIAA 

2002). The global furniture marketplace is considerable at about AUD80 billion of which Australian 

production contributes a small fraction, especially to global export and import replacement (United 

Nations, 2005). Australia is grouped within a series of countries which possess high costs in 

production and therefore is seen at a competitive disadvantage in comparison to countries representing 

low cost production. This poses a considerable direct threat to competitiveness of so-called high cost 

countries (HCC). 

Coupled with a perception of reduction in access to particular Australian forest resources, the need to 

focus on higher value-added products has become an imperative in planning for sustainable forest 

resource use and achieving returns on investment in the Australian furniture sector (AEGIS, 1999). 

Australian furniture products have surged in domestic and export popularity but the industry faces a 

strong threat from imports. The United Nations International Trade Database has valued furniture 

imports into Australia in excess of $470m in 1998 (United Nations 1999), $583m in 2001, moving 

progressively to $1.5bn in 2005 (United Nations 2005). Australia’s furniture sector exports in a trade 

balance comparison were $55.3m in 1998 (United Nations 1999), $76m in 2001, then climbing to 

$100m in 2005. 

The nature of international competition has become more complex during this period and Australian 

industries have experienced reduced import tariff protection in a bid to encourage competitiveness, a 

1 

University of Western Sydney, Locked Bag 1797, Penrith South DC, NSW 1797. Ph: 61 2 9852 5222. 

Email: S.Alexander@uws.edu.au.



rise in labour costs in comparison to countries with relatively lower labour structures, cost increases in 

local materials and other commodities, and consolidation of providers in several global trading sectors. 

This trade gap imbalance creates an increasing concern for the competitiveness of the Australian 

timber furniture industry. If sustained, this trend could render traditional timber furniture 

manufacturing based employment unsustainable, and erode domestic growth opportunities. 

Furthermore, a state of reduced competitiveness in adding value to processed materials could equate to 

further losses in wealth generation capability, potentially relegating Australia to a raw materials 

supplier to the global timber furniture products sector and a reduced economic value generator. 

The negative impact on the competitiveness of high cost countries (HCC) such as Australia, the USA, 

and the members of the European community and others, from low cost countries (LCC) of 

manufacture, namely China, has transformed global competition and challenged competitors to adapt 

to new consumer demands (Navarro et al 2008). 

A firm’s competitive ability to succeed in meeting new consumer demands within an industry sector 

and the marketplace is based on maintaining consumer interest in the products and services it offers 

(Woodruff 1997). The creation of unique value for the end-consumer is pivotal in building competitive 

advantage over rival firms (Barney 2002). 

THEORETICAL BACKGROUND: A VALUE CHAIN APPROACH 

In the analysis and building of competitive advantage, value chain methods have gained popularity as 

the efficiency of each process step is assessed in detail from raw materials to the eventual user, 

whereas the preceding supply chain models focused on activities that take raw materials and subassemblies 

into a manufacturing operation smoothly and economically (Porter 1985, Barney 2002). 

The importance an end-consumer places on the purchase of a forest product not only underwrites the 

viability of the timber furniture retailer but the suppliers of each preceding process. Firms can be seen 

to merely move money around by paying each other for services but it is the end consumer that is the 

ultimate source placing money back into the entire system or chain of activities known as the value 

chain. 

The value chain perspective highlights interdependence between firms aligned along a common value 

chain embracing raw material supply, logistics, product development, manufacturing and 

merchandising instead of primarily being about purchasing and selling. In recent years increases in the 

dominance of some value chains global competition has tended to be between value chains and 

precluded individual firms from re-entering the marketplace. This methodology is of value to a wide 

range of forest product industries and the timber furniture industry serves as an example of the 

potential. 

Morris (2000) suggests that a micro-level understanding of value creation of each firm within an 

interdependent and inter-firm context can be effective in focussing attention toward process efficiency 

and increases in productivity and economic rewards. This level of understanding if incorporated in 

cooperative intra-firm, inter-firm, regional and an industry-wide basis suggests a move to greater 

understanding of higher added value performance based on collaboration between firms and a 

heightened awareness of the cost in creating value for all suppliers and value for money for endconsumers. 

The increase in global competition has driven many organisations to search for new ways to achieve 

and retain competitive advantage by pursuing new value creation. 

In describing the attributes in new value creation, Bitici (2004) states that the more successful 

companies have a very clear and unambiguous value proposition, with the value proposition based on 

focussed and refined developments of key competencies, enabling value proposition delivery to 

consumers and suppliers. The key competencies include product leadership, operational excellence, 

and customer intimacy. Woodruff (1997) contends that the next major source of competitiveness will 

come from more outward orientation toward consumers and call on organisations to compete on 

superior consumer value delivery.



The extant literature does not directly link the concept of end-consumer value to the overall efficiency 

of the value chain, though there is discussion that value is reliant on value attributed by the endconsumer 

at the timber product retail level or retail showroom, as is the case with domestic or 

commercial timber furniture. 

Traditionally, each firm level decision is relatively independent of the end-consumer as value chain 

members are mostly isolated from having an overall viewpoint in how value creation is prioritised, 

information shared across value chain levels, nor the effectiveness and efficiency of activities and 

profitability in their value chain. The focus of value chain research to date has been on production 

efficiency and inter-firm communication and collaboration, rather than an extended model of 

collaboration integrating end-consumer requirements with overall value chain ambitions which are 

often fragmented, at best. 

The sought-after linkage between value chain effectiveness and end-consumer value may be 

addressed, in part, by the concept of the co-creation of value (Prahalad and Ramaswamy 2004, p.8). 

The co-creation of value between firms and consumers provides the linkage between the needs of the 

consumer and the needs of the firm, without isolating the firm or the consumer from the context of use 

and the creation of the consumer experience. 

A co-created forum suggests a market where there is an uninterrupted flow of information from 

individual consumer experience to the value chain and back “rather than passive pockets of demand 

for the firm’s offerings” (Prahalad and Ramaswamy 2004, p.16). The relationships are summarised in 

an adapted industry map placing the furniture consumer as the central focus of value creation activities 

and that all firm-level activities have a co-dependent effect on the success of the whole (Figure 1). 

Source: Adapted from Prahalad and Ramaswamy 2004, p.9 

Figure 1. Co-creation experience model 

The challenge for the future growth of value creation opportunities in the forest products sector or 

furniture sector may be reliant on the success of moving from an informal understanding of a value 

chain member’s role to a more formally recognised interdependent value creation model. Value chain 

members often conduct isolated efforts toward value chain outcomes communicating as a supplier to 

only one buyer. Firms may be unaware that their efforts are contained within a chain in which all 

inter-firm actions are interdependent. 

An alternate industry model where each value chain member is acknowledged by all value chain 

members in the value contributed to the value proposition delivered to the end-consumer has merit in 

the mounting of strategies to compete against rival chains with similar motivations. The prospect of 

receiving and providing assistance to enhance the efficiency and profitability for collaborators within 

one’s own value chain also adds impetus to building competitive advantage and creating further value 

for end-consumers as a common objective.



METHODOLOGY 

This paper identifies specific areas for improving the competitiveness of the Australian timber 

furniture industry, thereby influencing the forest products sector. The outcomes may have resonance in 

forest products industries and other sectors as the concept of competitiveness and a user-centred 

approach is not exclusive to the forest products sector alone. The most promising mechanisms for 

building value through industry and end-consumer collaboration using a value chain approach are 

explored. A mixed method research design was implemented in three phases and included the 

participation of several levels of the value chain from the forest source to the end-consumer. 

The confluence of decision making in the value chain was observed during interviews at multiple 

levels of the value chain. The levels investigated included the forest harvest, timber mill, timber 

merchant, the furniture manufacturer, furniture sub-contractor, and furniture retail levels of the value 

chain. As the research was focused primarily on the Australian timber furniture industry, the timber 

furniture manufacturing level was selected as the starting point in the identification of further players 

in the timber furniture value chain. 

The analysis of data from surveys gathered in the production areas of the value chain and with endconsumers 

in the furniture retail environment was interpreted and conclusions were drawn to ascertain 

if the intentions and ambitions of the value chain were complementary to the needs of the endconsumer. 

Participation from the industry levels of the value chain was a key component to the study. This 

included firms associated with forest supply, timber milling and distribution, furniture manufacture, 

transport logistics and furniture retail. Respondents from timber furniture manufacturing level of the 

value chain overwhelmingly cited the threat posed by low cost entrants into the Australian timber 

furniture market as a continuing competitive barrier to gaining market share. This emphasis and preoccupation 

in gaining local market share may be adversely affecting the psychology of firms in the 

creating of export opportunities at the furniture manufacturer level. Furniture manufacturer level 

resources are primarily focused on survival rather than growth as most firms in the furniture industry 

possess less than ten employees and retain few resources to generate innovations through new product 

development. 

The overall research strategy focused on the timber furniture value chain and the value creation 

process from forest plantation to furniture retail environment and the end-consumer. Barnes and 

Morris (1999) suggest that a focus on an individual product rather than only firm specific issues 

provides the ability to assess a production pipeline and therefore its value chain in its capacity to meet 

particular market demands (Womack and Jones 1996 cited in Barnes and Morris 1999). The 

production pipeline signifies the actions of value chain member firms in the production of a product. 

The selection of furniture type in this study was based on the suitability for timber value chain study 

within available time and resources. The selection suitability required the timber product to be 

relatively simple in construction including range of furniture construction materials used, as very 

complex products often constitute complex and extended value chain study, and resources. Access to a 

reliable study sample through industry referrals and the researcher’s familiarity with single seat 

furniture design promoted timber dining chairs as the favoured timber product to be traced through the 

timber value chain. 

A mixed method approach employing a triangulated research design (Creswell and Plano Clark 2003 

cited in Creswell and Plano Clark 2007 p.62) was implemented with the view to interpret the results of 

both industry and end-consumer surveys in establishing the extent of the value chain and factors 

affecting value chain competitiveness (Figure 2). 

The research design was formulated in two industry data collection phases: Phase 1 - exploratory and 

primarily qualitative; and Phase 2, a more targeted survey informed by the industry interviews of 

Phase 1, with both Phases containing a mixed approach seeking quantitative and qualitative data. A 

third, and final value chain level survey phase seeking end-consumer data provides the emphasis 

placed on successful timber furniture attributes and end-consumers’ perception of value (Phase 3).



Source: Adapted from Creswell & Plano Clark (2007 p.63). 

Figure 2. Mixed Method – Triangulation Design: Convergence Model 

The gathering of separate industry and consumer data was proposed as the most effective data 

gathering starting point until data from industry and consumer survey could be analysed for overlap 

and actors of significance (AEGIS 1999). This is unique to this study and not yet identified in the 

extant literature in the Australian furniture sector. 

OPERATIONALISATION: THE VALUE CHAIN SAMPLING TECHNIQUE 

Initially nine value chains were administered. This later was reduced to three value chains based on 

the timeframe, level of cooperation, and geographic differentiation and complexity of the value chain. 

Limitations experienced included the securing and timing of interviews in several Australian states; 

intellectual property issues; or issues undisclosed. Using this sample frame, over twenty-five 

respondents were engaged equating up to eight respondents per value chain on average. 

The measurement technique used in data gathering Phases 1, 2, and 3 of the study was a selfadministered 

survey technique containing a semi-structured interview, industry and end-consumer 

survey questionnaires respectively. The semi-structured interview technique is particularly effective in 

generating descriptive, explanatory, and exploratory perspectives. This method was used to test a 

proposed model based on the constructs of trust and collaboration, communication, and value creation. 

The standardisation of the survey for very different industry levels gives the study the potential to 

provide a balanced approach well suited to a value chain approach. A potential drawback is that the 

researcher may miss what is most important to the responder in their core competence area. Questions 

for this reason were as general and as open-ended as possible. 

In addition, the consideration of data from a wide sample of industry respondents also enhances the 

potential generality of findings. The semi-structured interview represented eight constructs of interest 

providing a number of variables of consequence to value chain performance and contained over thirtytwo 

open-ended questions within eight categories. The categories included firm level strategy; interfirm 

relations; resources and core competencies; local and global competitive threats; competitiveness 

from quality, policy and market data perspective, firm level performance; end-consumer value; and 

profitability trends. 

The research design Phase 2 incorporated a structured industry survey questionnaire targeted to the 

identified primary decision-making level of the value chain representing leadership in the particular 

forest product area: the timber furniture manufacturer. The structured industry survey instrument of 

Phase 2 contained an extended and refined version of the semi-structured interview process



represented in Phases 1 and more targeted survey questions. These questions included new product 

development (NPD), innovation processes, intellectual property including registered designs or 

patents, quality assurance, performance measurement, market intelligence, competitive advantage 

strategies, brand recognition, collaboration with suppliers and buyers, communication with suppliers 

and buyers, and timber certification. A series of questions were also directed to the level of 

understanding of the end-consumer needs, industry perceptions of consumer value, and the successful 

timber attributes industry believed the end-consumer held including environmental issues and the 

influence on consumer purchasing decisions. 

The final survey Phase 3 was represented by an end-consumer survey questionnaire administered at 

the timber furniture retail showroom environment. The questions were directed at members of the 

general public whose intention was to purchase timber furniture. The end-consumer questionnaire 

sought to establish an end-consumer’s opinion of the success of timber as a furniture material and the 

timber attributes that contributed to this consumer perspective. The survey contained five sections 

including the level of consumer approval of timber as a furniture material, successful end-consumer 

timber attributes, end-consumer buyer preferences, brand character responses, and end-consumer 

perceptions of timber certification and related environmental issues affecting timber furniture 

ownership. 

RESULTS: INDUSTRY SURVEY 

From the industry survey results the following points can be summarised as aspects concerning the 

timber furniture industry by firms contributing to the furniture value chain (FVC). 

A) Strategy: Respondents were confident that their firms more likely held sound new product 

development strategies; that they possessed unique product or services over competitors; that 

competitors were assessed for competitiveness periodically through mostly informal methods; and 

that internal performance measurement applying benchmarking techniques were prevalent. 

B) Inter-firm relations: Respondents were more likely to seek better relationships with suppliers, 

buyers, and industry bodies and less likely to seek further collaboration with government in their 

keen need to derive more reliable market intelligence. Higher collaboration was linked to better 

quality outcomes with suppliers and buyers, and that firms, mostly manufacturers, were more 

likely to be the initiators of communication with suppliers and buyers. Business strategies of 

interacting firms were not widely shared, nor were profit margins. 

C) Resources / core competencies: Respondents were positive in high and appropriate technology 

access and utilisation; reliable access to skilled human resources, reliable materials inventory and 

handling competencies, strong order processing efficiency and procurement effectiveness. The 

marketing of services was seen as a concern but effective selling was not. 

D) Competitive threats: Respondents were more likely to encounter difficulties in maintaining 

reliable suppliers and customers, and that international trade policy would adversely affect profits 

and with competition based on cost alone. 

E) Competitiveness: An increase in wider industry compliance to quality standards was linked to 

competitiveness. The design of imported products was not seen as a key competitive determinant, 

nor did timber certification give the imported products the competitive edge. 

F) Assessing efficiency – firm level performance: Strong support for understanding efficiency in 

firm’s and with collaborative firms, and preferences to meeting suppliers and buyers face-to-face. 

A very high level of satisfaction was present with current internal performance systems. 

G) Consumer value: The needs of end-consumers were held in very high regard by respondent firms 

and the furniture buyers according to the survey of firms, but the suppliers were thought to have 

less interest. Respondents perceived that end-consumers were more likely to have an interest in 

environmental issues and the belief that timber held superior material qualities over comparable 

materials in the end-consumer’s mind and that origin of timber both location and type had some 

influence on end-consumer choice. Respondents did link higher consumer knowledge of the 

process to higher profits by better informing the end-consumer’s purchase decision.



RESULTS: END-CONSUMER SURVEY 

From the end-consumer survey results the following points can be summarised as aspects concerning 

end-consumer attitudes to timber. 

A) Timber as a furniture material: End-consumers supported timber as a recyclable material but did 

not convincingly support timber as being environmentally safe. Timber was seen to be attractive to 

almost all respondents, and furniture construction material that was practical with inherent 

timeless qualities. Timber was seen as highly durable and possessing inheritable qualities with 

intergenerational appeal. There was some disconnection in terms of timber as fashionable with 

timber seen as youthful and a material of the future and also as an old-fashioned material with 

positive and negative associations possibly related to timber furniture design. 

B) Successful end-consumer timber attributes: A positive end-consumer attitude was expressed in 

timber possessing superior qualities as a furniture material in comparison to alternate materials. 

Emphasis was placed on attributes in five key areas including ecology, craftsmanship, touch 

characteristics, timber in use, and visual response characteristics. Successful visual characteristics 

account for the most favoured timber attributes with warmth, ageless, natural and good looking. 

The single most referred to timber attribute was the durability of timber. 

C) End-consumer buyer preferences: Respondents were highly to value the timber name and variety 

though the forest source location was not of considerable interest to end-consumers. Timber 

characteristics associated with timber quality were highly regarded by end-consumers as were 

levels of quality detail in timber exposing the qualities of the natural timber material with a high 

proportion in strong agreement. 

D) Brand character: Tradition linked to timber craftsmanship in production was respected by endconsumers 

though not overwhelmingly. Having knowledgeable retail sales staff was in demand 

and post-purchase furniture timber care was also sought-after. More than one visit to the retail 

showroom was essential for end-consumer prior to a furniture buying decision with cost and 

timber quality factoring highly in end-consumer decision-making. 

E) Timber certification: Selection of timber from approved forest sources was supported by half of 

respondents with a quarter strongly agreeing that timber certification would influence their buying 

decision. Over seventy percent of end-consumers inaccurately described the time required for a 

hardwood timber to reach maturity prior to harvest, and were only moderately more successful in 

estimating softwoods. 

RELATIONSHIP BETWEEN THE INDUSTRY AND END-CONSUMER SURVEY RESULTS 

There was encouragingly both interest from the industry level of the value chain toward end-consumer 

needs, and interest form the end-consumer level of the value chain towards the forestry and furniture 

manufacturing levels of the value chain. 

The end-consumers appear to possess a genuine interest in the variety of and qualities in timber but are 

less drawn to environmental issues than the industry level responses to the survey had expected. The 

industry has appeared to over-estimate the level of interest of the end-consumer in timber certification. 

In the industry position defence associated with timber certification, end-consumer responses in the 

survey also appear ambiguous toward timber certification where many respondents were undecided if 

timber certification would or would not affect their purchasing decisions. This may suggest that endconsumers 

may be experiencing a dilemma of environmental conscience where the end-consumer may 

need furniture for example but may be reluctant make a decision that may ‘adversely’ impact on a 

forest resource that appears to have a direct connection and resonance to overall planetary 

environmental health. This may be more a result of interpretations of general media than strategic 

long-range forestry management. 

The industry survey responses also linked more highly informed end-consumer purchasing decisions 

including forest source, timber variety, and qualities in craftsmanship with higher industry 

profitability. This was supported comprehensively in the end-consumer survey in the strong interest 

from end-consumers to understand the forest product intended for purchase from a timber variety and 

timber quality perspective. This is very encouraging and signifies precedence toward a level of



cooperation on the key determinants in the process of development of new products. It also supports 

the value chain technique advancing as a legitimate evolutionary step of a well-planned and connected 

industry focussed on end-consumer value, and industry sustainability. 

DISCUSSION 

It is apparent that there is a strong emotional and empathy-to-nature-based connection to timber in 

general, shared at all levels of the value chain including the end-consumer. This is a positive 

implication for the future of targeted timber use for particular industries, as all parties in the value 

chain converge on the same level of agreement. 

This level of consensus may be represented by the follows points: 

A) Timber holds unique qualities over alternate materials in particular applications, including timber 

furniture products; and 

B) The use of timber in high profile applications in close visual and textural contact with endconsumers 

holds long-term, endearing qualities which impact on longer term social, 

environmental and economic aspects that few alternate consumer products can provide over time. 

The barriers to maintaining industry competitiveness focused on the heightened future impairment of 

finding buyers, but not suppliers. Further obstacles included a consistent, formal or informal, internal 

performance measurement system, despite cited confidence in efficient internal procedures which was 

inconsistent in the analysis of survey responses. Furthermore, a lack of access to market intelligence 

hindered firm-level planning and targeted investment. 

End-consumer survey phase returns in the furniture retail environment established consumer attitudes 

to timber as a furniture material. Both the industry survey and consumer surveys concurred that timber 

was seen a material that held superior qualities in comparison to other materials. Future timber 

furniture purchasing was also positively affected by timber certification but not comprehensively. 

The industry survey returned the view that consumers may be interested in the process of production 

from forest to retail, and timber related environmental issues. This was consistent with end-consumers 

though also focused on timber quality, craftsmanship, and how to care for timber. Most consumers had 

little idea how long a tree may take to grow to maturity to be of use in timber furniture manufacture, 

especially hardwood species. Only about half of end-consumer respondents preferred timber from an 

approved forest source leaving room for industry discussion. 

CONCLUSION 

This paper explored value creation in the Australian timber furniture value chain and examined the 

level of synchronicity between the needs of end-consumers and value chain suppliers in what 

constitutes overall value for the end-consumer. The study establishes that in a fragmented industry 

despite high levels of trust and communication there is a search for leadership. A method by which the 

collective knowledge and expertise of the value chain is brought together to address competitive 

shortfalls is sought. It is not evident that the players in the timber furniture industry understand that 

they reside and participate within a value chain. As a consequence the industry may be unaware of the 

benefits that may be within arm’s reach in building competitive advantage. 

The gap in knowledge of the industry players namely the timber furniture manufacturers appears 

substantial. Their consistent request for access to more reliable buyers and access through buyers to 

the needs of the end-consumers is urgently required. Access to this vital information through focus on 

the entire value chain from forest to consumer may produce positive outcomes for all stakeholders. 

A mixed method research design approach to value chain thinking suggests that multiple level analysis 

of the industry and its consumer base is essential in understanding the attributes of the industry. This 

may assist in further understanding competition in local and global markets and how best to prepare to 

compete in those markets. By focusing on new value creation the paper indicates how the industry 

may synchronise the value needs of the industry and the end-consumer within the timber value chain. 

The targeting and prioritising the efficient use of forest resources toward the highest value-added 

yields per unit of material may support economic and environmental imperatives.



The focus on a high value added timber commodity such as timber furniture provides an opportunity 

to explore the application of value chain methodology. This methodology is much attributed in the 

building of competitive advantage in sectors outside wood and forest products industry and Australia 

in general and can act as a point of industry reflection and change. Within the understanding of the 

consequence of all inputs and outputs in the creation of value throughout production and the endconsumer-led-decision-making 

process lie the answers to building sustainable competitive advantage 

and challenges for forest products industries. 

LIMITATIONS 

This study is exploratory and does not claim any larger population or industry generalisations other 

than supporting a theoretical model for end-consumer centred value creation within the timber 

furniture value chain. The study did not seek the participation of entire value chain based on time and 

resources precluding a wider sampling of all value chain level inputs and therefore a sample 

representing the timber furniture manufacturing level has been emphasised in its leadership role in 

decision making in the selected timber value chain. 

REFERENCES 

Barnes,J. and Morris,M., 2000. Value Chains, Production Pipelines and Systemic Efficiency. Authors’ 

manuscript, South Africa. 

Barney,J.B., 2002. Gaining and Sustaining competitive advantage. New Jersey,USA: Prentice Hall. 

Bititci,U., et al 2004. Creating and managing value in collaborative networks. International Journal of 

Physical Distribution and Logistics Management; 2004, 34,3/4; 251-268. 

Commonwealth of Australia and AEGIS, 1999. The Australian Furnishings Industry: An Analysis. 

Commonwealth of Australia: Canberra. 

Creswell,J.W., and Plano Clark,V.L., 2007. Designing and conducting mixed methods research . 

Thousand Oaks, California. Sage Publications. 

FIAA, 2002. Furniture Trend 2002. Furnishing Industry Association of Australia. Sydney Australia 

Morris, M., and Barnes,J. 1999. Using Production Pipelines as a research methodology for 

understanding competitiveness : a case study of an automotive plastics component. School of 

Development Studies (incorporating CSDS) University of Natal, South Africa. 

Navarro,J., Hayward,P., and Voros,J. 2008. How to solve a wicked problem? Furniture foresight case 

study. Foresight, 2008; 10, 2; 11-29. 

Pakarinen, T.,1999 Success factors as wood as a furniture material. Forest Products Journal: 1999, 49, 

9; 79-85. 

Porter, M., (1985). Competitive advantage: creating and sustaining superior performance. New York 

:Free Press. 

Prahalad,C.K. and Ramaswamy.V. 2004. The future of competition: co-creating unique value with 

customers. Boston:Harvard University Press. 

United Nations, 2005. United Nations International Trade Database. http://comtrade.un.org/ 

Woodruff,R.B. (1997). Customer Value: The next source for competitive advantage. Journal Academy 

of Marketing Science: Spring, Vol. 25, Issue 2;p139,15pgs.



CARIBBEAN PINE GRADED SAWN RECOVERY VARIATION 

IN A MATURE SPACING BY THINNING EXPERIMENT 

GROWN IN QUEENSLAND 

Kevin Harding 1 , Terry Copley 1 , Marks Nester 2 , 

Kerrie Catchpoole 2 , Anton Zbonak 1 

ABSTRACT 

A mature Caribbean pine silviculture experiment provided initial square spacing 

treatments of 1.8 m 2 , 2.4 m 2 , 3.0 m 2 and 3.6 m 2 (equal to 3088, 1737, 1111 and 772 

stems/ha) that were thinned at age 10 years to 600, 400 and 200 stems/ha and an 

unthinned control. The trial was assessed for standing tree wood properties at 28 years 

and some of these trees were felled for a sawing study at 30 years. 

Dried dressed recovery (DDR) was strongly related to tree size and the volume ‘lost’ to 

defects tended to be lower in unthinned treatments. Ingrade recovery decreased with 

higher post-thinning stocking level for plots originally planted at 1111 and 772 spha 

and value per tree was highest in the lowest stocking treatment (772 spha). Log value 

decreased across all treatments consistently from butt to top log. The value per hectare 

was highest in unthinned plots due to high stem numbers per hectare. A complete 

economic analysis considering all cost structures is required to investigate the optimal 

silviculture required to maximise economic returns. 

INTRODUCTION 

In 2005, Forestry Plantations Queensland (FPQ) and the Wood Quality Improvement group within the 

Product Quality section of Horticulture and Forestry Science (H&FS), DPI&F, commenced a project 

to obtain samples from three mature silviculture/genetics trials to collect data on tree growth and wood 

quality. The main goal was to collect robust data from the three trials that would underpin wood 

property and growth modeling as part of an overall decision support system being developed by FPQ 

(Catchpoole et al. 2007). The data collection focused on within and between-tree variation to consider 

the impacts of different silvicultural factors on wood density, microfibril angle and juvenile wood 

proportion, both at breast height and at multiple heights within the merchantable stem. The 

experiments were targeted due to their maturity (20 to 28 years old), their design and their observed 

large differences in growth rates in response to the treatments applied. This variation provided 

excellent scope to evaluate and model the relationship between large productivity responses and wood 

traits. 

This paper focuses on the sawn recovery variables assessed in Pinus caribaea var. hondurensis 

experiment 532NC which was the only trial for which sawing was undertaken. This 28-year-old 

clearfall age experiment was intensively studied in a staged sampling program that provided: 

1. Standing tree non-destructive evaluations of wood properties in breast height increment cores and 

using standing tree acoustic assessment. 

2. Destructive sampling to examine up the stem variation with height including individual growth 

ring patterns of variation from pith to bark captured for key heights by SilviScan x-ray 

densitometric scans complemented by average 5-ring segment data at other tree heights. 

3. Sawn board graded recovery for 70 × 35mm structural product. 

Each of these assessment stages was based on fewer trees as the resources required and the expense of 

the assessments increased. However, the selection of each subsample was informed by the results of 

previous characterisations so as to select trees representative of the diameter distribution and acoustic 

properties observed in the experimental treatments. 

1 

Horticulture and Forestry Science, Department of Primary Industries and Fisheries, Queensland 

2 

Forestry Plantations Queensland



MATERIAL 

Experiment 532NC was planted in 2/1977 at Toorbul (27°05'S 152°58'E) and was 28-years-old when 

wood sampled in 6/2005 and 30-years-old when sampled for sawing in 08/2007. The experiment was 

planted at six initial stockings: 772, 1111, 1310, 1370, 1737 and 3088 spha resulting from five square 

spacings (1.8 m 2 , 2.4 m 2 , 2.7 m 2 , 3.0 m 2 and 3.6 m 2 ) plus one rectangular (3.5 m x 2.1 m). It was 

thinned at age 10 years to 600 spha, 400 spha and 200 spha with a control treatment left unthinned. 

For the purposes of this study four initial stocking levels (772, 1111, 1737 and 3088 spha from 1.8 m 2 , 

2.4 m 2 , 3.0 m 2 and 3.6 m 2 square spacings) were sampled as they provided extreme treatment levels 

and spacing comparisons with regular incremental increases across the treatments. 

For the sawing study, sampling was restricted to ten of the 16 spacing × thinning treatment 

combinations to meet logistical and resourcing requirements. The selection strategy used is detailed 

below. 

METHODS 

In each thinning treatment, approximately 105 trees were assessed for standing tree acoustic velocity, 

height, diameter at breast height and 12mm cores extracted giving a sample of 403 trees due to 

incomplete survival or unacceptable tree form, size (



Sawing strategy 

All stems were cut to 4.8m log lengths to allow direct comparison of recovery with height across 

treatments. Top logs were also recovered and sawn to a small end diameter merchantable limit of 

120mm DUB in a commercial length of 2.4, 3.0, 3.6 or 4.2m when the top log was shorter than 4.8m. 

Logs were sawn to maximise recovery of 70 × 35 mm framing dimension boards from centre cant/s 

and wings. All logs were sawn to only 70 × 35 mm framing dimension boards so non-structural 

recovery was not captured. The aim was to maximise the number of structural dimension boards 

recovered from each log and stem to provide a basis for comparison among and between treatments 

that would not be biased by product size differences among trees and logs. Standardising log length 

and product size should provide a sensitive basis for comparisons among trees and treatments. 

Analysis 

Correlation analysis at both individual tree level and treatment mean level was undertaken using the 

data analysis tools of GenStat (2002). 

RESULTS 

At the tree level, dried dressed recovery (DDR) ranged from 16% to 39%, with an average recovery of 

33%. DDR for individual treatment levels are presented in Figure 1. Results indicate that the highest 

recovery is achieved from the treatment with the lowest initial stocking level (772 stems/ha). It is also 

clear that recovery increases with lower numbers of stems retained after thinning. 

Dried dressed recovery (%) 

38 

36 

34 

32 

30 

28 

26 

Unthin 400spha 600spha Unthin 400spha 600spha Unthin 200spha 400spha Unthin 

1 (3088) 2 (1737) 4 (1111) 5 (772) 

Initial and thinning stocking (stems/ha) 

Figure 1. Average dried dressed recovery by initial stocking and thinning treatment in 

Experiment 532NC 

Tree size/volume significantly affects DDR. In general terms the more intense thinning regimes tend 

to produce larger trees and is reflected in higher DDR values. The differences in DDR were more 

pronounced for the smaller trees (< 1m 3 volume) covering a range of nearly 20% DDR with lower 

differences observed in DDR among the larger trees (>1m 3 volume) with a range of around 10%. 

Lost volume due to defects 

During visual grading wood defects such as knots, splits, resin pockets, resin shakes, etc. were 

recorded. If the measured value of the defects exceeded the visual grading threshold values defined in 

the machine stress grading standard (AS/NZS 1748, 2006), the stick was docked to remove these 

defects. The initial low stocking treatment (772 spha) exhibited the highest percentage of lost volume,



which is most likely associated with heavier branching and larger knot size in these trees. With the 

exception of treatment 772-unthinned, the unthinned plots showed lower values of PLV. 

Graded recovery 

All structural size sticks (70 × 35mm) were assessed for acoustic velocity using the CIRAD Forêt 

BING measurement system (Brancheriau and Baillères, 2003). Using the acoustic velocity properties 

with the stick average air-dry density and the maximum knot area ratio, an estimate of average 

modulus of elasticity (MoE) was calculated for each stick. Using these MoE estimates assignment of 

sticks to MGP (Machine graded pine) classes was undertaken. Grade assignment starts with the 

highest values of MoE and assigning these sticks to an MGP grade population until its average reaches 

the desired threshold value for each grade (MGP15 = 15200MPa, MGP12 = 12700MPa, MGP10 = 

10000MPa, Utility < 10000MPa). The upper and lower limits of each grade were recorded and used to 

assign the grade for each stick in that population. 

Figure 2 shows an overall distribution of grades across all investigated treatments. The level of out-ofgrade 

(reject and utility) is approximately 47% when results for all treatments are combined. MGP 15 

grade makes up only 2.4% of the total recovery of all sticks. In reality, it is not practical to have a 

single grade with this low proportion so the recovery of MGP15 was also included in a separate 

analysis as part of the MGP12 population to examine impacts on total product value. This latter 

process also impacted the distribution of MGP12, MGP 10 and utility grades (Figure 2). All 

subsequent analysis undertaken is reported only for Option B (i.e. without an MGP 15 population). 

Proportion of grades (%) 

50 

40 

30 

20 

10 

0 

Incl MGP15 Excl MGP 15 

MGP15 MGP12 MGP10 Utility Reject 

Grade 

Figure 2. Proportion of machine graded pine classes for all stocking and thinning treatments 

combined with options of inclusion and exclusion of MGP15 grade. 

The overall distributions of product grades in Figure 3 differ considerably across the treatments. In 

general, the grade proportions suggest that initial high stocking (3088 and 1737 stems/ha) produced a 

higher quantity of MGP 12 grade. Additionally, the proportion of MGP12 grade recovery generally 

increases with stronger intensity of thinning as might be expected due to improved recovery of mature 

wood from the larger trees produced. However, although grade proportions are an important indicator 

of quality they must be considered with the total volume recovered to quantify the total economic 

potential of a treatment response.



Actual value per log (AUD) 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 


1 (3088) 2 (1737) 4 (1111) 5 (772) 


Figure 3. Proportion of grade classes across different stocking and thinning treatments in 

Experiment 532NC 

Economic value recovery 

To consider the economic importance of the grade recovery distribution across treatments the value of 

recovery by product was estimated. Product value was assigned to each stick based on its grade using 

the wholesale price estimates per m 3 provided in Table 2. 

Table 2. Approximate December 2008 wholesale softwood product prices (AUD/m 3 ) for 70mm 

structural product grades and length 

MGP15 MGP12 MGP10 

Grade Length Length Length Length Length Length Utility 

>4.6m 2.4-4.6m >4.6m 2.4-4.6m >4.6m 2.4-4.6m 

Price 

AUD/m 3 $729 $663 $605 $582 $501 $456 $247 

Value recovery at log level 

Figure 4 presents the distribution of actual value of log along the longitudinal tree axis and across 

treatments. In all treatments the butt log is the most valuable part of the tree and total log value 

declines up the stem. The gradient of value decline is not uniform across treatments. Most notable is 

the steep difference observed in the fastest grown treatment, 772-200 stems/ha. On the other hand the 

unthinned compartment with the highest initial stocking level (3088 stems/ha) exhibited a moderate 

change in log value with log position in tree. It can be expected that these trends in part reflect branch 

size and therefore knot size impacts on grade. 

Value recovery at tree level 

The full distribution of DBHOB classes within each individual plot was used to calculate a weighted 

average value recovery at tree level across different spacing and thinning treatments. Using this 

approach the effect of influential individuals with low representation within the sample distribution is 

reduced. Figure 5 shows differences in value recovery for simple average and weighted average 

values. In general, unthinned plots with a higher initial stocking level provide less valuable trees than 

compartments with less initial stocking and more intense thinning regime. This is to be expected with 

average tree size increasing with more intensive thinning and providing improved recovery of better 

quality mature wood timber from the larger average log size. 

butt 

2 

3 

4 

5



Proportion of grades 

100% 

80% 

60% 

40% 

20% 

0% 

Unthin 

1 

(3088) 

400spha 

600spha 

Unthin 

400spha 

600spha 

Unthin 

200spha 

400spha 

2 (1737) 4 (1111) 5 (772) 


Unthin 

Reject 

Utility 

MGP10 

MGP12 

Figure 4. Value recovery at log level across different stocking and thinning treatments in 

Experiment 532NC. 

Value recovery/tree (AUD) 

250 

200 

150 

100 

50 

0 

Averaged value/tree Weighted average value/tree 


1 (3088) 2 (1737) 4 (1111) 5 (772) 


Figure 5. Comparison between average value recovery and weighted average value recovery at 

tree level across different stocking and thinning treatments in Experiment 532NC. 

Value recovery prediction for per hectare comparisons 

The experiment results were used to extrapolate to per hectare value comparisons by multiplying 

weighted average tree value with utilizable number of trees per hectare taking into account survival 

and tree size. Only trees meeting a minimum DBHOB of ≥ 22cm were considered merchantable. 

These comparisons are presented in Figure 6. Unthinned plots, despite lower average tree value, 

provide higher actual value on a per hectare basis than plots with high thinning intensity. The latter is 

unsurprising as it reflects numbers of trees and biomass on a site and is divorced from the commercial 

realities of establishment costs including costs per tree planted, and market requirements for both



piece size in log and sawn form that impact harvesting, haulage and processing cost structures and 

product size requirements. The 600 spha post-thinning treatments in the higher initial stocking 

treatments (1737 and 1111 spha) and the 400 spha treatment in the low initial stocking (772 spha) 

treatment display intermediate returns. These results suggest that economic modelling of returns 

taking into account all establishment, harvesting and processing costs should provide important 

insights in to the optimal establishment and final stocking regimes needed to maximise value returns. 

Value recovery per hectare (AUD) 

100000 

80000 

60000 

40000 

20000 

0 


1 (3088) 2 (1737) 4 (1111) 5 (772) 


Figure 6. Value recovery per hectare comparison across stocking and thinning treatments in 

Experiment 532NC for stems ≥ 22 cm DBHOB. 

Correlation analysis 

The matrices in Tables 3 and 4 (Appended) show the directions and strength of the correlations among 

investigated properties at individual tree level (all 56 trees sawn, ignoring treatment) and at treatment 

mean level combining 5-6 trees from each thinning treatment into single average values (10 data 

points). The results revealed that the tree growth properties such as DBHOB, tree height and tree 

volume were significantly correlated with almost all sawing properties. The most notable correlation 

(r=0.98) was found between tree volume and actual dollar value of graded products per tree. This is 

not surprising since increased dried dressed recovery would be expected as tree size increases. Larger 

trees should also produce a higher proportion of good quality grades from improved recovery of 

mature wood as a proportion of overall recovery which is then reflected in overall higher mean value 

of structural graded products. 

There were no significant correlations between tree growth characteristics and basic density at 

individual tree level but they emerged at treatment level showing that density is positively related with 

DBHOB and tree volume. This reflects the balance of juvenile wood versus mature wood increment 

with lower proportions of higher density mature wood in the unthinned plots with slower growth rates 

compared to heavily thinned stands which produce bigger trees from better growth rates throughout 

the life of the stand. 

Acoustic velocity as assessed using an ST300 device was poorly correlated with all other investigated 

properties. The only significant correlation found for acoustic velocity was with MOE measured on 

framing sticks (r=0.49) and with the proportion of MGP12 and MGP10 grades combined (r=0.42) at 

individual tree level. However, it should be noted that ST300 velocity is used to predict wood stiffness 

in the outer wood of the tree which is then mostly lost during sawing in waney-edge boards.



Basic density of core samples was weakly correlated with dried dressed recovery (r=0.32 at individual 

tree level; r=0.80 at treatment level), and the proportion of MGP12 and MGP10 grades recovered. 

There was a strong relationship between average core density and average framing stick density 

(r=0.77) as might be expected. 

CONCLUSIONS 

This sawing study has produced very encouraging results which suggest that returns may be optimised 

by silvicultural regimes that are within the spectrum of current commercial practice. Initial planting at 

772 to1111 spha with thinning to between 400 and 600 spha produces relatively good dried dressed 

recovery percentages, structural grade distribution, log value and predicted product value per hectare.. 

Although the results suggest that higher stocking regimes can produce higher average log and tree 

values and returns per hectare, these are unlikely to be economically viable once all establishment, 

harvesting and processing costs are considered, as well as the product requirements (piece size and 

grade distribution) of processors. 


This work was funded by Forestry Plantations Queensland and represents the planning and 

collaborative efforts of a large team of colleagues from both FPQ and DPI&F Innovative Forest 

Products including: Ian Last and Paul Keay (FPQ) and Megan Prance, Martin Davies, Bill Atyeo, 

Adam Redman, Rapheal Tschupp, Dr Henri Bailleres, Eric Littee and Rob McGavin (Innovative 

Forest Products group, DPI&F). 

REFERENCES 

Brancheriau, L. and H. Baillères (2003) Vibration Spectra as Grading Technique for Structural 

Timber. Holzforschung 57: 644–652. 

Catchpoole, K.J., Nester, M.R. and Harding , K.J. (2007) Predicting wood value in Queensland 

Caribbean pine plantations using a decision support system. Aust. J. Forestry 70(2): 120-124. 

GenStat (2002) GenStat for Windows, version 6.1 Lawes Agricultural Trust. 

Standards Association of Australia and Standards Association of New Zealand. (2006) Timber - 

Mechanically stress-graded for structural purposes. AS/NZS 1748, Standards Australia/Standards 

New Zealand, Sydney and Wellington 12 pp.



Table 3. Correlation coefficients with p-level value at individual tree level. Trees from all sampled treatments combined (only statistically 

significant coefficients with p-level



Table 4. Correlation coefficients with p-level value at treatment mean level (only statistically significant coefficients with p-level



REHABILITATION OF FORESTS IN DECLINE: 

MT. LINDESAY STATE FOREST 

Peter St.Clair 1 

ABSTRACT 

Forests suffering decline were rehabilitated using a combination of harvesting, burning, 

planting and herbicide treatments at Mt. Lindesay State Forest in Northern NSW. The 

severity of the decline ranged from slight through to severe, where, in the latter, there 

were few forest trees alive, little eucalypt regrowth, a dense understorey of lantana, and 

abundant colonies of bell miners (Manorina melanophrys, Meliphagidae). Yellow 

bellied gliders and koalas were the only arboreal mammals found at the site. Two years 

after treatment the forest contained healthy regrowth, much reduced lantana infestation, 

reduced bell miner populations and no measurable change to yellow bellied glider and 

koala populations. Tree crown health improved significantly and bell miner populations 

decreased. 

INTRODUCTION 

Eucalypt decline and shrub encroachment in the absence of fire is a widespread and increasing 

problem in Australia (Anon. 2007) and it usually involves a range of secondary pests, pathogens and 

parasites of eucalypts (Jurskis 2005). Near Urbenville in north-eastern NSW, the area of declining 

forest increased from less than 1,000 ha in 1992 to at least 20,000 ha in 2004 (Jurskis 2005). 

Declining forests in this region typically have a dense shrubby understorey often dominated by 

lantana (Lantana camara), and dying overmature, mature and pole stage eucalypts infested with sapsucking 

insects called psyllids and high numbers of bell miners. This form of decline has been 

termed Bell Miner Associated Dieback (BMAD). There are basically two divergent models of the 

decline process (e.g. Stone 1996, 1999, 2005, Jurskis and Turner 2002, Jurskis 2005, Turner and 

Lambert 2005, Stone et al. 2008, Turner et al. 2008). Jurskis and others propose that exclusion of 

fire from open forests causes changes in soils which adversely affect the roots of established 

eucalypts and promote plants and animals that attack or compete with the trees. Stone and others 

alternatively consider that disturbance of forest structure allows shrub encroachment which provides 

nesting habitat for bell miners. Bell miners exclude other psyllid-eating birds resulting in elevated 

psyllid populations which eventually kill the trees. 

The BMAD specific model proposed by Stone and others cannot explain why decline involving 

psyllids in some forest types can occur in the presence or absence of disturbance and/or bell miners, 

and it cannot explain why a similar process affects a wide range of undisturbed forests in most of 

Australia where there are no bell miners. Research at Richmond Range and at Jilliby by Stone and 

others confirmed many elements of the model proposed by Jurskis and others. Canopy decline was 

associated with infrequent fire, dense shrubbery, low and wet positions, elevated soil nitrogen, and 

areas relatively undisturbed by recent logging, whereas young regrowth stands were healthy. This 

and other research also showed that psyllid plagues and decline of susceptible eucalypts are not 

constantly associated with the presence of bell miners (e.g. Moore 1961, Stone et al. 1995, Stone 

and Simpson 2006, Dare et al. 2007, Jurskis 2008). It is clear from long term observations of 

eucalypt decline involving bell miners and from the high stockings of dead stags in many of the 

declining stands, that complete canopies in undisturbed, fully stocked stands are gradually reduced 

as crowns become thin and die, and under-storey density increases due to increased sunlight and 

water availability. Elevated soil N was also correlated with decline in eucalypt plantations near 

Kyogle (Mews 2008). To improve tree health, the challenge is to control the understorey density and 

soil N content. 

1 Forests NSW, NSW Department of Primary Industries, PO Box 688, Casino NSW 2470. Email: peters@sf.nsw.gov.au



Culling of bell miners and removal of their watering points have been trialled to ameliorate forest 

decline. On the South Coast of NSW 1500 birds were removed resulting in rapid return of other bird 

species, and some improvement in tree health, but subsequent re-colonisation by bell miners and 

deteriorating tree health. Removal of a dam at Toonumbar had no measureable effect on bell miner 

numbers (Sommerville pers. com.). Bell miners can accelerate the decline process by excluding 

other birds that would have a damping effect on psyllid populations, but they are now considered by 

the BMAD Working Group to be a contributor rather than a cause of decline. However bell miner 

populations are easy to monitor, and may be used as a surrogate for psyllid numbers and an indicator 

of the effects of rehabilitation treatments on the forest decline process. 

Combinations of harvesting, cultural treatments and fire regimes were trialed to improve and 

maintain the health of deteriorating stands and to rehabilitate severely declining stands. Objectives 

of the project were: 

1. Lantana cover reduced to less than 15% 

2. Increased health of retained trees 

3. Decrease in abundance of bell miners (An indication of reduced habitat or food) 

4. Maintenance of grassy understoreys 

5. Restoration of severely degraded stands with natural regeneration, supplementary seeding 

and enrichment planting of native over-storey species 

6. Integration of harvesting and rehabilitation 

METHODS 

Site description 

The trial was conducted in Mt. Lindesay SF (Cpts 276, 279, 280) in the head of the Richmond River 

catchment. The trial area of 536 ha was considered sufficiently large to minimise edge effects, 

accessible, and substantially affected by decline. The Chocolate soils contain about 45% clay and are 

derived from Tertiary Lamington Volcanics weathered over the Walloon Coal Measures (sandstone, 

siltstone and mudstone) with a small area of Marburg sandstone in the north of Compartment 279. 

Forest decline is prevalent on the Walloon Coal Measures (Jurskis 2004). Annual rainfall is about 

1400mm with most falling January to April. Mean maximum temperatures range from 17 0 C in July 

to 29 0 C in January and mean monthly minima range from 3 0 C to 16 0 C. Prior to 1950’s the forests 

were extensively grazed and burnt to maintain grass with only a small number of high quality logs 

removed. During 1974-77 large high quality logs (>40cm dub) were selectively removed. The area 

was logged again during 1983-84 for low quality logs and durable poles and girders, yielding 

1100m 3 (3 m 3 ha -1 ). Parts of Cpt 280 were harvested during 1996. The roadside of Cpt 276 was 

harvested in the mid 80’s for the Mt. Lindesay Highway realignment. 

The stands ranged from low to high quality, with variable structure including patches of regrowth of 

various age and mature stands. All forest types except Brush Box (Lophostemon confertus) were 

declining, though the severity varied widely, and regeneration in most recently disturbed areas was 

stifled by lantana. The drier forest types originally had a grassy understorey but had over recent 

decades developed an impenetrable lantana understorey (Weaver 2006 pers. com.). There was only 

one record of fire (in 1994) during the preceding 20 years, burning 40 ha on Hildebrands Rd. Nine 

forest types are mapped within the compartments. Areas of Narrowleaved White Mahogany-Red 

Mahogany-Grey Ironbark occur along ridge tops and upper slopes along with Grey Box, Grey Gum- 

Grey Ironbark-White Mahogany, Sydney Blue Gum, Tallowood-Sydney Blue Gum and Rock. Mid 

to lower slopes are dominated by Brush Box or Flooded Gum or rainforest, and Forest Red Gum 

dominates low flats. 

Treatments 

Variable forest health and structure as well as a diversity of forest types necessitated a variety of 

silvicultural treatments. A Silvicultural Decision Tree (SDT) was developed to allocate treatments 

across the trial area (Figure 1). Treatments included a combination of harvesting, selective shrub 

removal (mainly Lantana and Cissus), ground preparation, burning, planting, weed control, and 

reintroduction of low intensity fire in appropriate areas. Due to dense lantana understorey it was not



physically possible to determine in advance the treatment to be applied at a fine scale. The location 

of all treatments was mapped using GPS and detailed records were maintained for each discrete unit. 

Harvestable 

trees 

YES 

No 

Harvestable 

trees 

IS THERE ESTABLISHED REGENERATION? 

(>150spha>4m high) 

Harvest as 

Forest Types Forest Types 

per IFOA 46,47,48 60,62,65 

Top disposal 

LIB = Low Intensity burning 

Push, rip, 

spray, burn, 

plant 750- 

1500spha. 

LIB 

Follow up 

Lantana 

control 

Machine 

disturbance 

to assist LIB 

Enrichment 

plant 

Figure 1. Silvicultural decision tree used for rehabilitation area at Mt. Lindesay SF in 2007 

Monitoring and Methodology 

Sixty plots were established in treated and control areas with stratification based on broad forest 

types (Dry and Wet as defined Forest Commission of NSW Research Note 17 and classified as per 

Table 1). 

Table 1. Allocation of monitoring plots in Mt. Lindesay SF Rehabilitation Trial in 2007 

Location Treatment Area Control Area 

Dry Forest Types(FT 60, 62 and 65) 20 Plots 10 Plots 

Moist Forest Types(FT 46, 47 and 48) 20 Plots 10 Plots 

Plots were located on a grid randomly oriented across the area. Plots were not established within 

harvest exclusion areas or equivalent areas in controls. Control plots were established in 

Compartment 280 immediately east of the treatment area to minimise edge effects, facilitate fire 

exclusion from the control area, maximise efficiencies for planning, harvesting and rehabilitation 

treatments and to allow for the trial to continue for 15 years. 

No fuel management or prescribed burning will be undertaken in the control area for the 15 years of 

the trial. Monitoring methodology was according to agreed minimum standards (BMAD WG 2007). 

Lantana cover was estimated for the whole plot, individual tree crown score by Stone et al. (2008), a 

3 minute bird score on a 30m radius (Craig and Roberts 2002). Fire and harvesting intensities were 

estimated from the amount of disturbance or residual ground cover vegetation left in the plot. 

Mammal surveys were conducted as per Threatened Species Licence protocol. 

NO 

Seed Trees 

8pha 

Eucalypts 

top disposal 

Harvest 

additional as 

per IFOA 

Brushbox 

top disposal 

as required



RESULTS 

Application of Treatments 

Harvesting was conducted over the period May to September 2007 and 9260 m 3 were removed from 

the net harvest area of 425 ha at an average of 22 m 3 ha -1 . This yield included 2650 m 3 of salvage 

logs and 900m 3 of logs destined for hardboard manufacture. Gross royalty was $392 000 at an 

average of $42 m -3 or $922 ha -1 with product royalties ranging from $218 m 3 to $2.40 m 3 . 

Herbicide was sprayed along roadsides, snig tracks and drainage depressions to generate drier fuel 

for ignition. Herbicide was also used after harvesting and burning to control accessible outbreaks or 

residual clumps of lantana. Burning was conducted between 16/10/2007 and 25/10/2007. Weather 

indices are shown in Table 2. 

Table 2. Fire weather conditions at Mt. Lindesay SF during burning operations in 10/ 2007 

Date 

Oct 

2007 

Max 

Temp 

( o C) 

Min 

RH 

(%) 

Wind 

Direct n 

Wind 

(km/hr) 

16 30 24 W 5 80 9 14 

17 24 38 SE 10 82 9 9 

22 28 35 NE 5 85 9 10 

25 27 40 NE 5 86 9 8 

BKDI DF FDI Comment 

Western Area 

115 ha, 15 plots 

Southern area 


Eastern area 


Nth Hwy 

25 ha, 1 plot 

Rain commenced on the evening of 25 th October 2007 and continued incessantly for 6 months 

eliminating further burning opportunities. Of the 40 plots planned for burning, 11 were not burnt for 

various reasons, 21 experienced intense fire and the balance lower fire intensities. Of the 40 plots 8 

were not harvested, and 6 had minor harvesting disturbance. These were mostly not burnt. The most 

intensively harvested plots experienced higher fire intensities. Rainfall was favourable for the 

establishment of planted seedlings and natural regeneration of vegetation on disturbed sites. In the 

period Oct 2007-Feb 2008, 75000 eucalypt or brush box seedlings were established in about 34 ha of 

clearings and about 62 ha of sparse stands requiring enrichment planting. 

Integration of harvesting and rehabilitation 

Harvesting disturbance facilitated rehabilitation of lantana infested forest by opening up tracks and 

creating dry fuel for burning. After harvesting according to the SDT it was possible to identify sites 

for selective clearing prior to burning. These areas were burnt at the same time as general logging 

slash. Attempts to burn green lantana on lower slopes near drainage lines where harvesting was not 

conducted were unsuccessful because the fuels were green and access was difficult. 

Lantana cover 

Lantana cover was lower in 2008 than in 2007 in all treatments and recovered slightly in 2009 in all 

treatments except the controls. The recovery was significant in the moderate and intense fire 

treatments and in all treatments combined (Figure 2). There was no significant difference in the 

lantana cover in unburnt and minor fire treatments before and after treatment. The more intense fire 

treatment had the greatest pre-treatment lantana cover (80%) and following fire had the lowest (5%). 

The moderate and intense fire treatments, and all fire treatments combined showed some recovery of 

lantana cover in 2009 compared with 2008 (Figure 2). In 2009 the most intense harvesting treatment 

had significantly lower lantana cover than less intense harvesting or treatments that were not 

harvested (data not presented). The weighted average lantana cover for all plots was 14.7% in 2009 

compared with 66% in 2007 before treatments were applied (Table 3).



Within three months after more intense fire and harvesting treatments there was a proliferation of 

‘fire weeds’ such as inkweed (Phytolacca octandra), yellow weed (Sigesbeckia orientalis), 

nightshade (Solanum nigrum), fire fern (Doodia aspera), native grasses, poison (aka native) peach 

(Trema aspera), native hibiscus (Hibiscus heterophyllus), bleeding heart (Homalanthus 

populifolius), tobacco bush (Solanum mauritianum), wattles, and some eucalypt regeneration. The 

exotic weeds scotch thistle (Circium vulgare), ragweed (Ambrosia artemisifoila), noogoora burr 

(Xanthium occidentale) and purple-top verbena (Verbena bonariensis) also established. The rapid 

growth and density of these fire succession species was almost sufficient to eliminate lantana and 

cissus from intensely burnt plots. 

Average % lantana cover 

(incl std error) 

90 

80 

70 

60 

50 

40 

30 

20 

10 

0 

Average % lantana cover according to treatment, 

Mt Lindesay State Forest Project Area 2007-2009 

Control 

Harvested/unburnt 

Minor fire 

Moderate fire 

Treatment type 

Intense fire 

All fire treatments 

2007 

Census 

2008 

Census 

2009 

Census 

Figure 2. The effect of fire 

treatments on the average lantana 

cover following harvesting and fire 

treatments imposed after the 2007 

census. (n=60) 

Table 3. Weighted average of lantana cover of plots in the Mt. Lindesay project area in March 2007 

and 2009 following fire treatments in Oct 2007. 

Fire Treatment Nil 1- Minor 2-Moderate 3-Intense Total 

#Plots 11 3 4 22 40 

Plot % 27.5 7.5 10 55 100 

2007 Lantana % 52 46 43 80 

2007 weighted % 14.3 3.4 4.3 44 66 

2009 Lantana % 28 21 21.5 6 

2009 weighted % 7.7 1.6 2.1 3.3 14.7 

Health of Retained Trees 

Tree canopy health improved in both controls and treated plots from 2007 through to 2009. The 

health of trees in treated plots was lower than trees in control plots at each measure (Figure 3). 

Average canopy condition score (incl std 

error) 

25 

20 

15 

10 

(un)Treated 

2007 

Average canopy condition score for all plots, 

Mt Lindesay State Forest Project Area 2007-2009 

Treated 2008 Treated 2009 Control 2007 Control 2008 Control 2009 

Census date and treatment type 

Figure 3. Crown canopy condition score of 

treated and untreated plots as measured in 

March 2007 prior to harvesting (April- 

September 2007) and burning treatment 

(Oct 2007) and measured again in March 

2008 and March 2009. Treatments n=40, 

Control n=20. Crown score system is that 

published by Stone et al. (2008) and was 

calculated for all trees retained as at 

March 2009 compared with their 

corresponding crown score condition in 

March 2007 prior to treatments being 

applied.



Bell miner numbers 

Bell miner numbers were lower than in controls immediately after treatment but had recovered to 

similar levels to controls in 2009 (Figure 4). Bell miner populations in the controls were much lower 

in 2008 and 2009 compared with 2007. The bell miner score was strongly correlated with the lantana 

cover (Figure 5). 

Average bell miner bird count 

class (incl std error) 

Average bell miner bird count class for control and treated 

plots, Mt Lindesay State Forest Project Area 2007-2009 

2.5 

2.0 

1.5 

1.0 

0.5 

0.0 

2007 census 2008 census 2009 census 

Census year 

control 

treated 

Figure 4. Average score of bell miners in 

control and treated plots treated with 

harvesting and burning after the 2007 

census. Bell miner score 0= nil birds. 1 

=1-5 birds, 2 =5-20 birds. 

Figure 5. Relationship between 

lantana cover and bell miner numbers. 

Data averaged for each of 6 censuses 

where n=60. 

Use of low intensity burning 

There was only about 2 ha of grassy forest remaining in the study area. Trees in this area were 

healthy, part of a Highway visual protection zone and not of a size and quality for harvesting. This 

area was burnt and maintains its grassy appearance. Some other nearby forests that still exhibit 

extensive grassy areas were burnt during 2007 and 2008. A single low intensity fire did not kill 

lantana. 

Eucalypt and Brush Box Regeneration 

Regeneration was stimulated by fire and averaged 800-1000 stems ha -1 and was marginally increased 

at the higher level of harvesting disturbance (Figure 6). 

Number seedlings per hectare (incl 

std error) 

Average regeneration (stems per hectare) by fire 

treatm ent, Mt Lindesay State Forest Project Area 2009 

2000 

1800 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

0 

0 1 2 3 

Fire Intensity 

Number seedlings per hectare 

(incl std error) 

Average regeneration (stems per hectare) by harvest 

treatment, Mt Lindesay State Forest Project Area 2009 

2500 

2000 

1500 

1000 

500 

0 

0 1 2 3 

Harvest intensity 

Figure 6. Effect of fire and harvesting treatments in 2007 on counted regeneration of eucalypt 

seedlings in plots in March 2009. 1=Minor, 2=Moderate, 3=Intense 

Regeneration was however variable as indicated by the standard error bars. Highest retained 

stocking coincided with lowest regeneration (Table 5). Brush Box regeneration was two orders of 

magnitude greater than the eucalypts (Figure 7). Planted seedlings were found in 6 plots. All 

showed good survival and early growth. Planted seedlings in the largely cleared areas ranged in



stocking from 150 to 1400 stems ha -1 whilst in enrichment planting areas they were difficult to 

locate and distinguish from natural regeneration as they were not planted in a grid pattern. 

Table 5. New regeneration in plots and retained stocking and basal area of trees with 

diameter at breast height over bark greater than 10cm for corresponding plots at 

Mt. Lindesay SF 

Stocking stems ha -1 No 

regen 

0-100 101-250 251-500 501-1000 >1000 Total 

#Plots 13 8 5 2 3 9 40 

% plots 32.5 20 12.5 5 7.5 22.5 100 

Retained stocking 

stems ha -1 

202 194 175 88 108 139 

Retained BA m 2 ha -1 20.1 12.7 9.7 5.1 22.7 11.2 

Number seedlings (log scale) 

Total natural regeneration by species following fire, Mt Lindesay State Forest Project Area 

2007-2009 

1000 

100 

10 

1 

Brushbox 

Sydney blue gum 

White mahogany 

Grey gum 

Forest redgum 

Tallowood 

Species (common names) 

Red mahogany 

Bloodwood 

Grey ironbark 

Figure 7. Total natural 

regeneration by species 

in 40 plots of aggregate 

area of 1.6 ha following 

treatments in 2007 as 

measured in plots in 

March 2009. Note log 

scale. 

Rehabilitation costs 

Costs include tractor time, herbicide, plants and planting, fertiliser, fire trail construction, burning, 

planning and supervision. In moderate and severely declining stands site preparation is the main 

fixed cost and plants/planting and fertiliser cost were a variable cost of $1 per seedling. Estimated 

total cost for the treatments in this project were $189000, most expense going into the moderately 

and severely impacted areas (Table 6). 

Table 6. Estimated total rehabilitation cost (AUD) at Mt. Lindesay SF, 2007 

Decline Rating Area 

(ha) 

0 Healthy 

Full regeneration 

1/2 Slight. Minor Lantana 

EnrichmentPlanting 

3 Moderate. Lantana. 

Enrichment planting 

4 Severe. Heavy lantana. 

No regeneration 

Planted 

seedlings/ha 

Total 

seedlings 

Rehab n cost 

$/ha 

Total 

$ cost 

229 0 0 100 22900 

100 100 10000 200 20000 

62 500 31000 1000 62000 

34 1000 34000 2500 85000 

TOTAL 425 75000 189000 

Impacts on mammals 

Following harvesting and burning in April to October 2007 numbers of arboreal mammals were 

similar to numbers recorded in 1999, 2000 and 2007 using the same methods in the same sites 

(Table 7).



Table 7. Occurrence of yellow bellied gliders and koalas as recorded over a 3 night period in 

Mt. Lindesay SF Project Area harvest area. *Koalas sighted whilst measuring plots. 

Date recorded July November March March- 

1999 2000 2007 April 2008 

Yellow bellied glider 14 9 23 16 

Koala 0 0 3 2* 

DISCUSSION 

Lantana cover 

Harvesting and herbicide spraying created dry fuels relative to the surrounding untreated areas. This 

made lantana combustible at relatively low Fire Danger Indices. In the absence of harvesting 

disturbance fire intensity was insufficient to consistently kill lantana. After intense fire, dense 

lantana cover was replaced with dense ‘fireweeds’, grasses and eucalypt regeneration, however there 

was significant recovery of lantana in the moderate and intense fire treatments where cover at least 

doubled in the year following treatment. Nevertheless, the objective of lantana cover being reduced 

to less than 15% was met. Growth of shrubs including lantana is likely to cause problems with 

reintroduction of low intensity fire regimes in areas without a complete and even regenerating 

eucalypt canopy. The need for costly rehabilitation operations can be avoided by reintroducing low 

intensity fire regimes into open forests before understorey development makes this too difficult. 

Although a single low intensity fire may not kill lantana, repeated fires will gradually reduce or 

eliminate it (e.g. Kington et al 2009) When shrub encroachment is well advanced, integrated 

harvesting and rehabilitation operations are likely to be the only feasible solution to forest decline on 

a broad scale. 

Overstorey recovery 

Canopy health in forests in north eastern New South Wales generally improved after 2007 with 

ongoing rainfall events. In this project, tree crown health improved in both control and treated areas 

but remained lower in treated areas than in controls. There is no evidence that the treatments 

improved the health of the retained trees at this stage. High intensity fire scorched the crowns and 

promoted epicormic foliage which is palatable and nutritious to psyllids. In stands with sparse 

regeneration, immediate reintroduction of low intensity fire will be necessary to improve the health 

of retained eucalypts. 

Bell Miner abundance 

Bell miners can contribute to forest decline and accelerate the process by excluding other birds 

which could more effectively damp irruptions of psyllids. In a previous project at Donaldson SF it 

was found that the most severe treatment (mechanical disturbance and hot fire) produced a greater 

reduction in bell miner numbers and increase in other bird species than did either burning alone or 

herbicide spraying alone (Mews 2006). In the current study there was a proportionately greater 

reduction of bell miners in treated areas than controls in 2008, but bellbird numbers had recovered 

by 2009 and were not significantly different to controls. It appears that harvesting and burning 

resulted in loss of nests and fledgings, and loss of psyllids (i.e. food resources) in scorched canopies. 

Stone et al. (2008) found a strong correlation between bell miner abundance, understorey density 

and tree crown health. The current study indicated that reduced bell miner populations in controls 

were a result of better crown health following ongoing rainfall and consequently reduced psyllid 

populations. Reduced lantana cover in the controls was apparently due to increased shading provided 

by healthier crowns. Although there was a much greater reduction of lantana in treated areas than in 

controls (Figure 2), canopy health and bell miner numbers in 2009 were not affected by the 

rehabilitation treatments (Figures 3, 4) indicating that canopy health controls psyllid and bell miner 

populations, and understorey density. If this is the case, then removal of bell miners and poisoning 

or burning of lantana per se will not improve tree health. The phenomenon of linked lantana, psyllid 

and bell miner invasions is a consequence of poor tree health caused by deteriorating root function 

under changing soil conditions in the absence of fire as proposed by Jurskis (2005). Frequent low 

intensity fire regimes maintain stable soil conditions and forest health in naturally grassy forests



whilst exclusion of fire causes tree decline, shrub invasion and pest outbreaks (Jurskis 2005, Turner 

et al. 2008). 

Forest rehabilitation and regeneration 

A previous fire and mechanical clearing trial in a declining stand of Forest Red Gum in Donaldson 

SF virtually eliminated lantana and produced a proliferation of fire succession species (mainly native 

peach and wattle) similar to the more intense fire treatment in this study. There was insufficient 

eucalypt regeneration to restore a forest canopy (Shipman 2007). Shipman (2007) reported that 32 % 

of 60 plots had no eucalypt regeneration, 36 % had a stocking of 250 stems ha -1 and only 32 % had 

a stocking of 500 stems ha -1 or more. In this project about half the plots had regeneration stocking 

of less than 100 stems ha -1 and about a quarter of the plots had more than 1000 stems ha -1 . The low 

levels of regeneration can be partially attributed to the high stockings and retained basal areas (Table 

5) and the absence of regeneration where fire was not used (Figure 6). 

High intensity fire such as used in parts of Donaldson and Mt.Lindesay State Forests resulted in 

dense understorey development but achieved large reductions in lantana understoreys, compared 

with less intense or no fire. There is need to be cautious with fire treatments as high intensity fire can 

exacerbate forest decline by promoting N cycling in the dense nitrophilous understoreys and shading 

the ground (e.g. Jurskis et al. 2003). This is true of native and other exotic understoreys as well as 

lantana. Furthermore, lantana showed significant recovery in moderate and high intensity fire 

treatments within a year after treatment. Equally, high intensity fire may scorch crowns and promote 

epicormic foliage which is palatable and nutritious to psyllids. 

The variability of regeneration success in plots with lower residual stockings and basal areas can 

primarily be explained by whether or not there was brush box regeneration (Figure 7). The poor 

regenerative capacity of the eucalypts can be explained by the paucity of seed in their debilitated 

crowns. Several years of normal shoot development are necessary for eucalypts to initiate buds, 

flower, and produce mature seed. This cannot occur in trees suffering chronic decline. Where brush 

box seed trees were present there was prolific seed fall and establishment, particularly after fire 

treatments, irrespective of intensity, in harvested areas. From a forest regeneration viewpoint, 

retention of well spaced brush box seed trees is desirable, however planting of eucalypt seedlings is 

vital to maintain a natural species composition in mixed stands. 

Cost of project 

Whilst the cost of the project was significant, the opportunity cost of doing nothing is greater. The 

cost of rehabilitation was less than the likely loss of production if the forest continued to decline and 

die (Table8). 

Table 8. Estimate of the economics of BMAD rehabilitation assuming growth rates impacted 

by BMAD and actual average royalty rates and rehabilitation costs at Mt. Lindesay. 

Decline Net growth Value 

Rating rate $/ha @ 

(m3/ha/yr) $40/m 3 

Opportun y 

Rehab cost Rehab Net gain 

cost (OC) Over 40 yrs cost (RC) production 

($/ha/yr) ($) ($/ha/yr) (OC-RC) 

($/ha/yr) 

0 None 1 40 100 4 0 

1/2 Slight 0 0 40 200 25 15 

3 Moderate -1 -40 80 1000 50 30 

4 Severe -2 -80 120 2500 75 45 

IMPLICATIONS OF THE PROJECT 

Priority 1. Use of practices to limit understorey invasion of grassy forests so that costly 

rehabilitation is not required. Reintroduction of low intensity fire and/or return of grazing to forest 

areas. Whilst it maybe argued that cattle are not a natural component of forest ecosystems, neither is 

the exclusion of fire or the presence of lantana. Cattle grazing provides the opportunity for closer 

understorey management through occupier weed control, more fire management capacity and 

browsing of regenerating shrubs and lantana by cattle.



Priority 2. Use of harvesting, burning and planting treatments to restart degraded stands to ensure 

rapid growth and dominance of the overstorey, thus controlling the understorey. The regenerating 

stands of drier forest types may begin to decline as they approach pole stage and sapwood basal area 

declines (e.g. Jurskis 2005) unless low intensity fire regimes are introduced when forest structure 

and fuel permits. 

Priority 3. Whilst harvesting is currently not an option for assisting rehabilitation of lantana 

infestations in National Parks or in State Forests areas reserved from harvesting, wildfires may 

present an opportunity to restart patches of forest provided seeding or planting of overstorey species 

can be done before the explosion of fire succession species and/or subsequent re-invasion of lantana. 


The author acknowledges the Bell Miner Associated Dieback Working Group and members of the 

Scientific Reference Group (Harry Recher, Ross Florence, Ross Goldingay) for assistance with 

design and support of this project, partial funding by the National Heritage Trust, Forests NSW staff 

who planned, supervised operations and measured plots, especially Casino staff Flavio Bugno, Jayne 

Fitzpatrick and Craig Pavy, Lynden Pavy, Jamie Churchill, Ken Wheatley, Dan Hutton, Kevin 

Harvey, Paul Meek and Steve Rayson. Vic Jurskis and Doland Nichols for their reviews and 

comments, Sarah Nadler and Craig Wall from Department of Environment and Climate Change for 

database creation and data management, and Adam Kirby of Southern Cross University for data 

analysis and presentation and Border Ranges Contracting and the Githabul people for planting 

operations. The plots will be remeasured in March 2011. Visit www.bmad.com.au for more info. 

REFERENCES 

Anonymous (2007). Tree decline in the absence of fire. Fire Note 13. Bushfire Cooperative Research Centre, 

East Melbourne. 

BMAD Working Group (2007). Minimum standards for monitoring. DECC Coffs Harbour. 

Craig, M.D. and Roberts, J.D. (2001). Evaluation of the impact of time of day, weather, vegetation density and 

bird movements on outcomes of area searches for birds in eucalypt forests of south-western Australia. 

Wildlife Research 28, 33-39. 

Dare, A.J., McDonald. P.G. and Clarke. M.F. (2007) The ecological context and consequences of colonization 

of a site by bell miners (Manorina melanophrys). Wildlife Research 34, 616-623. 

Forestry Commission of NSW (1989) Research Note 17 Forest Types of NSW. ISBN: 0 73055649 2 

Integrated Forestry Operations Approval 

Jurskis, V., (2004). Overview of forest decline in coastal New South Wales. In: White, T.C.R., Jurskis, V. 

(Eds.), Fundamental Causes of Eucalypt Forest Decline and Possible Management Solutions. 

Proceedings of a Colloquium at Batemans Bay, 18 and 19 November 2003, State Forests of NSW, 

Sydney, pp. 4–7. 

Jurskis, V. (2005) Eucalypt decline in Australia, and a general concept of tree decline and dieback. Forest 

Ecology and Management, 215, 1–20 

Jurskis, V., 2008. Drought as a factor in tree declines and diebacks. In: Sanchez, J.M. (Ed.) Droughts: Causes, 

effects and predictions. Nova Science Publishers Inc. New York. pp 331-341. 

Jurskis, V. and Turner, J. (2002) Eucalypt dieback in eastern Australia: a simple model. Australian Forestry 

65 (2): 87-98 

Kington, W., Kington, D. and Burnham, M. (2009) Guide for planned burning. Open forests and Woodlands 

V2. South-east Queensland Parks and Wildlife. Qld Government CD 2009 

Mews, J. (2006) Effects of understorey modifications trial on bird populations in Bell Miner Associated 

Dieback Affected Forest. Undergraduate Project. Southern Cross University, Lismore. 

Mews, J. (2008) The role of interactions between understorey, soil properties and foliar nutrient status in the 

development of Bell Miner Associated Dieback (BMAD). Honours Thesis. Southern Cross University. 

Moore, K.M., (196)1. Observations on some Australian forest insects. 8. The biology and occurrence of 

Glycaspis baileyi Moore in New South Wales. Proc. Linn. Soc. N.S.W. 86 (2), 185–200. 

Shipman, R (2007) Regeneration/treatment of native forest bell miner affected areas. Undergraduate Project. 

Southern Cross University, Lismore.



Stone, C., (1996), The role of psyllids (Hemiptera: Psyllidae) and bell miners (Manorina melanophrys) in 

canopy dieback in Sydney Blue Gum (Eucalyptus saligna Sm). Australian Journal of Ecology 21. 

Stone, C., (1999), Assessment and monitoring of decline and dieback of forest eucalypts in relation to 

ecological sustainable forest management: a review with case study. Australian Forestry 62 (1) 

Stone, C., (2005), BMAD at the tree crown scale: A multitrophic process. Australian Forestry 68 (4) 

Stone, C., Kathuria, A., Carney, C. and Hunter, J. (2008), Forest canopy health and stand structure associated 

with bell miners (Manorina melanophrys) on the central coast of New South Wales. Australian 

Forestry 71 (4) 

Stone, C. and Simpson, J.A. (2006) Leaf, tree and soil properties in a Eucalyptus saligna forest exhibiting 

canopy decline. Cunninghamia 9, 507-520. 

Stone, C., Spolc, D.and Urquhart, C. A., (1995), Survey of Crown Dieback in Moist Hardwood Forests in the 

Central and Northern Regions of NSW State Forests (Psyllids/Bell miner Research Programe). 

Research paper No. 28. Research Division, State Forests of NSW. Sydney. 

Threatened Species Licence, under the Threatened Species Conservation Act (1995) Upper North East Region 

Turner, J. and Lambert, M. (2005), Soil and nutrient processes related to eucalypt forest dieback. Australian 

Journal 68 (4) 

Turner, J., Lambert, M., Jurskis, V., Bi, H., (2008). Long term accumulation of nitrogen in soils of dry mixed 

eucalypt forest in the absence of fire. For. Ecol. Manage. 256, 1133-1142.



USING A PROCESS-BASED MODEL, PAROPSYS, 

TO PREDICT THE IMPACT OF CLIMATE CHANGE ON THE 

EUCALYPT PLANTATION PEST Paropsis atomaria. 

Simon Lawson 1 

ABSTRACT 

The eucalypt leaf beetle, Paropsis atomaria Olivier, is an increasingly important pest of 

eucalypt plantations in subtropical eastern Australia. A process-based model, ParopSys, 

was developed using DYMEX TM and was found to accurately predict the beetle 

populations. Climate change scenarios within the latest Australian climate model forecast 

range were run in ParopSys at three locations to predict changes in beetle performance. 

Relative population peaks of early generations did not change but shifted to earlier in the 

season. Temperature increases of 1.0 to 1.5 ºC or greater predicted an extra generation of 

adults at Gympie and Canberra, but not for Lowmead, where increased populations of 

late season adults were observed under all scenarios. Furthermore, an additional 

generation of late-larval stages was predicted at temperature increases of greater than 1.0 

ºC at Lowmead. Management strategies to address these changes are discussed, as are 

requirements to improve the predictive capacity of the model. 

INTRODUCTION 

Chrysomelid leaf beetles are important insect defoliators of young eucalypt plantations in Australia in 

most growing regions. Damage by these beetles can lower growth rates significantly, with losses of 

up to 30 percent in productivity observed over a ten year period in unprotected trees in Tasmania (Elek, 

1997). There, Paropsisterna bimaculata and Pa. agricola are key pests of E. nitens and E. globulus 

plantations. Pa. bimaculata populations are managed using a sophisticated integrated pest 

management system that is based on more than 20 years of research data on the biology, ecology and 

impact of this beetle on tree growth rates (Elliott et al., 2002, Elek, 1997, Elliott et al., 1992). 

In the subtropics of northern New South Wales and southeast and central Queensland, significant 

expansion of eucalypt plantations has only occurred since the late 1990’s, with about 60,000 ha now 

established in Queensland and 80,000 ha in New South Wales (Gavran and Parsons, 2009). The 

chrysomelid beetle, Paropsis atomaria Olivier, has emerged as a key pest of several plantation taxa in 

the region during this time, including Eucalyptus cloeziana, E. dunnii, E. grandis and its hybrids and 

Corymbia citriodora ssp. variegata (Nahrung et al., 2008b, Nahrung et al., 2008a, Nahrung, 2006). 

Severe outbreaks have been commonly reported by plantation growers and applications of insecticides 

have been used to manage beetle populations on an ad hoc basis, not as part of an intensive IPM 

system (Lawson et al., 2008). 

While the biology of P. atomaria had been well-studied in south-eastern Australia (Carne, 1966a, 

Carne, 1966b, Ohmart, 1991, Ohmart et al., 1987), intensive field censuses of the beetle and its natural 

enemies, and laboratory studies undertaken since 2004 in Queensland have begun to address 

deficiencies in knowledge of the biology and ecology of this species in subtropical plantations. Low 

temperature thresholds, temperature-induced mortality rates, and day-degree requirements for 

immature stages were obtained in the laboratory using beetles sourced from temperate and subtropical 

zones and combined with field mortality estimates of these stages to produce a process-based 

population model, ParopSys, using DYMEX TM (Maywald et al., 2007). This model has been 

successfully validated for sites in central and southeastern Queensland, and for Canberra in temperate 

southeast Australia (Nahrung et al., 2008b). The model has since been used to predict population 

trends in plantations and to optimize timing of insecticide applications. 

1 Queensland Primary Industries and Fisheries, Department of Employment, Economic Development and Innovation, 80 

Meiers Rd, Indooroopilly, Qld 4068. Ph: +61 7 38969613 Email: Simon.Lawson@deedi.qld.gov.au



A feature of the DYMEX TM modeling system is the ability to run climate change scenarios and test the 

impact of these on the population dynamics of the organism or system being modeled. Since insects 

are poikilothermic, changes in temperature directly affect developmental rates and hence population 

dynamics. This module was used to test the effects of climate change on the population dynamics of P. 

atomaria using a series of scenarios derived from the most recent projections for Australia (Climate 

Change in Australia Technical Report 2007) for three locations in Queensland and the Australian 

Capital Territory that were used in the initial model validation. 


ParopSys 

Details of the structure and values of variables used in ParopSys are detailed by (Nahrung et al., 

2008b). In brief, the model was run using the same climatic data used to validate the model and the 

simulations run over the period 28 September 2005 to 10 May 2006, the period when adult beetles are 

known to be active in the field. P. atomaria predominantly overwinters in the adult stage and so the 

model was initialised with 0.5 pre-reproductive adults on day one, with the same number added on 

each of the following nine days, giving a total of five adults. Pre-reproductive adults are used to 

initialise the model because, as they emerge from overwintering, adults must feed for a period of time 

before they are able to begin reproducing. 

Temperature variation in the climate change module can use a number of different input factors, 

depending on the location of the climate data being used. For this study, inputs which raised the 

maximum and minimum temperatures for ‘winter’ (defined in the model as March 2 to September 30) 

and ‘summer’ (defined as October 1 to March 1) by a defined number of degrees were used (see 

below). 

Regions 

The three locations originally used to validate the model and which represent a large part of the 

distributional range of this beetle were chosen to test the climate change scenarios. These were: (1) a 

plantation near Gympie, southeast Queensland (26º 06'S; 152º 45'E); (2) Lowmead, central 

Queensland (24º 30'S; 151º 42'E); and (3) Canberra, A.C.T (35º 27’S; 149º 12’E). 

Meteorological data used in ParopSys were obtained from the Silo Data drill website 

(http://www.nrw.qld.gov.au/silo/datadrill/). 

Climate Change Scenarios 

Variation in temperature was used as the only predictor in these climate change scenarios, since 

rainfall, evapotranspiration and other potential climate change variables are not currently used as 

inputs to drive ParopSys. The model therefore assumes that beetle access to suitable food resources 

(young, expanding foliage for larvae and young or adult foliage for adults) is not limiting. It is 

anticipated that future versions of ParopSys will incorporate explicit modelling of resources such as 

presence of flush foliage and immigration and emigration processes. 

Climate change projections for the regions being modelled were obtained from the Climate Change in 

Australia Technical Report 2007 website (http://www.climatechangeinaustralia.gov.au). A variety of 

projections are available from this modelling, using high, medium and low carbon/ CO2 emission 

levels and 10, 50 and 90 percentile estimates of likelihood, with projections to 2030, 2050 and 2070. 

For this modelling exercise, we chose what would appear to be the most likely scenarios, i.e., medium 

level emissions, 50 percentile likelihood (considered to be the best estimate) for projection to 2030, 

2050 and 2070. Across all locations, the 2030 scenario gives summer and winter temperature 

increases of 0.6 to 1.0 ºC (Fig. 1), for 2050 1.0 to 1.5 ºC for southeast and central Queensland and 1.5 

to 2.0 ºC for Canberra, and for 2070 1.5 to 2.0 ºC for southeast and central Queensland and 2.0 to 2.5 

ºC for Canberra. 

We ran seven climate change scenarios in ParopSys to encompass this forecast range in temperature 

variation. Data for the 2005-2006 season for each site were used as the control, with increases of 0.25,



0.5, 1.0, 1.5, 2.0 and 2.5 ºC in mean daily temperature. From these model runs we obtained outputs of 

numbers of all Paropsis lifestages over the length of the model run (28 September to 10 May). Given 

that most damage by this beetle is caused by late stage larvae (third and fourth instars) and adults we 

focussed on the effect of the climate change scenarios on these stages. 

RESULTS 

Outputs of ParopSys for the seven climate change scenarios simulations are shown in Figure 1 (adult 

beetles) and Figure 2 (third and fourth instar larvae). 

Numbers of Adults Numbers of Adults 

Numbers of Adults 

180 

160 

140 

120 

100 

80 

60 

40 

20 

0 

300 

250 

200 

150 

100 

50 

0 

70 

60 

50 

40 

30 

20 

10 

0 

28/09/2005 

18/10/2005 

No Change 

0.25 

0.5 

1.0 

1.5 

2.0 

2.5 

7/11/2005 

27/11/2005 

(a) Gympie 

(b) Lowmead 

(c) Canberra 

17/12/2005 

6/01/2006 

26/01/2006 

15/02/2006 

Simulation Date 

Figure 1. Comparison of adult P. atomaria population responses to seven climate change 

scenarios at three sites using the simulation model ParopSys: (a) near Gympie, 

southeast Queensland, (b) Lowmead, central Queensland and (c) Canberra, temperate 

south-eastern Australia. 

7/03/2006 

27/03/2006 

16/04/2006 

6/05/2006



Numbers of L3 + L4 Larvae 



1000 

800 

600 

400 

200 

0 

3500 

3000 

2500 

2000 

1500 

1000 

500 

0 

600 

500 

400 

300 

200 

100 

0 

28/09/2005 

18/10/2005 

No change 

0.25 

0.5 

1 

1.5 

2 

2.5 

7/11/2005 

27/11/2005 

(a) Gympie 

(b) Lowmead 

(c) Canberra 

17/12/2005 

6/01/2006 

26/01/2006 

15/02/2006 

Simulation Date 

Figure 2. Comparison of late stage (3 rd and 4 th instar) larval P. atomaria population responses to 

seven climate change scenarios at three sites using the simulation model ParopSys: (a) 

near Gympie, southeast Queensland, (b) Lowmead, central Queensland and (c) 

Canberra, temperate south-eastern Australia. 

Without climate change, ParopSys predicts two beetle generations per year for Canberra, three for 

Gympie, and four for Lowmead. (Nahrung et al., 2008b) validated this prediction for Gympie and the 

outputs were also shown to fit field observations for Lowmead (Lawson pers. obs.) and for Canberra 

(Carne, 1966a). 

When temperature was varied according to the six climate change scenarios, adult beetle populations 

at each site differed in their response to increases in temperature. The Gympie and Canberra 

populations were similar in that an additional generation began to appear in the upper range of climate 

change scenarios. For the Gympie population, a fourth generation of beetles began to emerge with 

temperature increases of 1.5 ºC and above (Fig. 1a), while the Canberra population added a third 

generation at the same elevations in temperature (Fig. 1c). For temperature increases below 1.5 ºC, 

7/03/2006 

27/03/2006 

16/04/2006 

6/05/2006



population peaks were predicted to occur earlier in the season but peak heights did not change. All 

increases in temperature resulted in an increased fourth generation peak population for the Lowmead 

site, with a more than two-fold increase in peak height being observed at temperatures from 1.5 to 2.5 

ºC (Fig. 1b). 

Changes in peaks and population levels of late stage larvae with increased temperature showed similar 

patterns for the Gympie and Canberra sites, where additional late stage larval generations were only 

predicted for the top two temperature increases of 2.0 and 2.5 ºC (Fig. 2a,c). For the Gympie site, 

neither of these additional larval population peaks exceeded the previous third generation peak. At the 

Canberra site, the peak of the third generation late stage larvae was only predicted to exceed that of the 

second generation for a 2.5 ºC increase. For Lowmead, additional larval population peaks that almost 

equalled, or exceeded, the third generation peak occurred at all temperature increases of 1.0 ºC and 

above. 

From Figures 1 and 2 it appears that there is a threshold of temperature change below which 

population dynamics appear relatively stable and above which there are significant changes in number 

of generations and size of population peaks. This mainly seems to be the case for the southernmost 

sites of Gympie and Canberra. To test this observation, temperature change was plotted against 

population numbers of the last peak in the season for both adults and larvae (Fig. 3). 

Numbers of late season peak adults 

Numbers of late season peak L3+L4 

300 

250 

200 

150 

100 

50 

0 

3500 

3000 

2500 

2000 

1500 

1000 

500 

Gympie 

Low mead 

Canberra 

Gympie 

Low mead 

Canberra 

0 

0.0 0.5 1.0 1.5 2.0 2.5 

Temperature change ºC 

Figure 3. Relationship between increase in average temperature and predicted numbers of 

adults (top) and late stage larvae (bottom) in late-season population peaks for three 

sites: near Gympie, southeast Queensland, Lowmead, central Queensland and 

Canberra, temperate south-eastern Australia.



For both the Gympie and Canberra sites, a threshold temperature of between 1.0 and 1.5 ºC is 

predicted to produce an extra generation of adults. A similar trend was observed for these two sites 

for larvae, but the threshold at which an extra generation of these stages occurred was at a higher 

temperature of between 2.0 and 2.5 ºC. By comparison, no such distinct thresholds were observed for 

the Lowmead site, where there the change is restricted to an increase in numbers of the already 

existing fourth generation. 

DISCUSSION 

The effects of climate change on herbivorous insect population dynamics are complex and involve 

interactions between the direct effects of temperature on the developmental and mortality rates of the 

insects themselves, the effects of environmental variables on their host plants, and the interactions 

between these factors. Effects on host plants are largely mediated through the effects of elevated 

temperature and enhanced atmospheric levels of CO2 on plant growth rates and changes in plant water 

relations as a function of differing rainfall and evapotranspiration patterns. 

Temperature was used as the only climate change variable in this study since the ParopSys model does 

not currently use other climate inputs to drive beetle populations. This is an obvious limitation to the 

applicability of these results given that populations of these beetles are also driven by a range of other 

variables, in particular the availability of abundant, expanding juvenile foliage, which with eucalypt 

species is closely associated with soil moisture reserves and rainfall during the active growing period. 

However, even with this limitation the model has been shown to accurately predict the number, timing 

and relative population peak heights for P. atomaria in southern Queensland, and, at a minimum, the 

number of generations at sites in central Queensland and temperate south-eastern Australia (Nahrung 

et al., 2008b). Therefore, the trends that emerged from this modelling are potentially useful to forest 

managers in predicting potential changes in the population dynamics of this beetle over its 

distributional range as climate warms. 

General outcomes of the modelling can be summarised as follows. No changes in relative population 

peaks of the early generations of either beetle stage were observed at any site or with any of the 

temperature change scenarios, but these generations became time-shifted to earlier in the season (Figs 

1&2). These time shifts become more marked as the season progresses, with shifts of up to three 

weeks earlier for the third generation at Gympie and Lowmead and the second generation at Canberra 

being observed. Beetle damage can thus be expected to occur earlier in the season at all sites as 

temperature increases. 

For the two more southern locations of Gympie and Canberra, temperature elevations of 1.0 to 1.5 ºC 

or greater (equivalent to the temperature changes predicted post 2030 for these sites) predict the 

development of an extra generation of adults but not additional large numbers of late stage larvae. 

High populations of late season adults can cause severe defoliation of mature foliage which impacts on 

the ability of trees to recover in the following spring through reduced photosynthetic capacity. For 

these two sites, the model therefore predicts that, up to 2030, damage is largely restricted to that 

caused by larval stages (i.e. upper canopy defoliation) and that this damage is progressively shifted 

earlier in the season with increased temperature, but post-2030 greater late stage defoliation of mature 

foliage by adult beetles can be expected. 

For the northernmost site of Lowmead, the model predicts no extra adult generation even at the 

highest temperature increase, but does predict progressive increases in population levels of this stage 

at all scenarios modelled (a greater than twofold increase in population at the 2.5 ºC elevation). With 

late-larval stages an additional generation begins appearing at temperature increases of 1.0 ºC and 

greater, equivalent to post 2050 scenarios for this site. At Lowmead, the model therefore predicts 

increased levels of late season adult damage to mature foliage for all temperature increases through to 

2070, with an extra larval generation causing additional late season upper canopy defoliation from 

2050. From 2050, this effectively means that greater levels of whole-tree defoliation may be expected 

at this site. 

In terms of practical management outcomes, the modelling presented here suggests that monitoring of 

beetle populations should commence earlier in the season than currently is practised as temperature



increases. The development of additional generations of adults (Gympie, Canberra), and higher 

numbers of late season adults and an additional larval generation (Lowmead) would indicate that 

strategies to reduce beetle populations earlier in the season are going to be increasingly important to 

lower the risk of highly damaging late-season populations developing. At some sites, multiple control 

applications may be required. 

As stated above, there are a number of limitations to the ParopSys model, including unknown effects 

of temperature on the survival of overwintering stages and high temperature mortality responses. 

Elevated winter temperatures could potentially increase survival of overwintering adult beetles, 

resulting in higher beetle populations earlier in the season. Experimental data is required so that this 

potential response to climate change can be incorporated into the model. Whilst P. atomaria is 

distributed as far north as tropical Queensland, damaging populations have not been recorded in 

plantations there and the beetle appears to be quite rare. High temperatures may therefore be a 

limiting factor in development of populations of this beetle. We have temperature associated mortality 

data for the immature stages of P. atomaria and this will be included in a future version of ParopSys to 

improve its predictive power. 

Furthermore, the model assumes constant favourability of its tree host for survival and reproduction. 

Addition of a ‘flushing’ module, including tree response to defoliation, would therefore be an 

important refinement since the presence of flush foliage is required for egg-laying and subsequent 

development of beetle immature stages. 

Other predicted effects of climate change on plants are not currently included in the model. Elevated 

CO2 levels are predicted to have various effects on plant chemistry and physiology including: (a) 

higher plant growth rates, often leading to higher rates of feeding by insects due to higher biomass but 

lower foliar nitrogen levels and (b) increased carbon to nitrogen ratio, enabling plants to potentially 

produce higher levels of carbon-based defensive chemicals, such as phenolics and tannins (important 

components in the leaves of eucalypts), and lowering their ability to produce nitrogen based defensive 

compounds (such as alkaloids and cyanogenic glycosides; (Trumble and Butler, 2009). More complex 

responses arise when temperature and CO2 are elevated simultaneously (Zvereva and Kozlov, 2006). 

Few specific data exist on the effects of climate change on eucalypt leaf chemistry, with the exception 

of (Gleadow et al., 1998) who examined the effects of elevated CO2 levels on nitrogen chemistry 

(particulary cyanogenic compounds) in the leaves of Eucalyptus cladocalyx. Under normal conditions 

leaf nitrogen has not been shown to be a significant limiting factor for development of P. atomaria 

(Ohmart et al., 1987, Ohmart, 1991) but further specific studies on the performance of P. atomaria on 

hosts grown under elevated temperature and CO2 are essential to further develop the climate change 

predictive ability of the ParopSys model. 


Sincere thanks to Dr Valerie Debuse and Dr Manon Griffiths of the Department of Employment, 

Economic Development and Innovation for their help in improving earlier versions of this manuscript. 

Development of the ParopSys model was partly carried out under Australian Research Council 

Linkage Projects Program (LP0454856) and in conjunction with Forestry Plantations Queensland 

(formerly DPI-Forestry). I gratefully acknowledge all organisations for their support, and especially 

Dr Helen Nahrung, Dr Mark Schutze, Dr Anthony Clarke, Michael Duffy, and Elizabeth Dunlop who 

were instrumental in the development of the ParopSys model. 

REFERENCES 

Carne, P. B. (1966a) Ecological Characteristics Of The Eucalypt-Defoliating Chrysomelid Paropsis Atomaria. 

Aust. J. Zool., 14, 647-72. 

Carne, P. B. (1966b) Growth And Food Consumption During The Larval Stages Of Paropsis Atomaria 

(Coleoptera : Chrysomelidae). Entomologia Exp. Appl., 9, 105-112 Pp. 

Elek, J. A. (1997) Assessing The Impact Of Leaf Beetles In Eucalypt Plantations And Exploring Options For 

Their Management. Tasforests, 9, 139-154.



Elliott, H. J., Bashford, R., Greener, A. & Candy, S. G. (1992) Integrated Pest Management Of The Tasmanian 

Eucalyptus Leaf Beetle, Chrysophtharta Bimaculata (Olivier) (Coleoptera: Chrysomelidae). Forest 

Ecology And Management, 53, 29-35. 

Elliott, H. J., Elek, J. A. & Bashford, R. (2002) Developing Pest Management Systems For Eucalypt Plantations. 

Forspa Publication, 139-146. 

Gavran, M. & Parsons, M. (2009) Australia’s Plantations 2009 Inventory Update, National Forest Inventory, 

Bureau Of Rural Sciences, Canberra. 

Gleadow, R. M., Foley, W. J. & Woodrow, I. E. (1998) Enhanced Co2 Alters The Relationship Between 

Photosynthesis And Defence In Cyanogenic Eucalyptus Cladocalyx F. Muell. Plant, Cell & 

Environment, 21, 12-22. 

Lawson, S. A., Mcdonald, J. M. & Pegg, G. S. (2008) Forest Health Surveillance Methodology In Hardwood 

Plantations In Queensland, Australia. Australian Forestry, 71, 177-181. 

Maywald, G.F., Kriticos, D.J., Sutherst, R.W. & Bottomley, W. (2007) Dymex, Version 3.0. Hearne Scientific 

Publishing, Melbourne 

Nahrung, H. F. (2006) Paropsine Beetles (Coleoptera: Chrysomelidae) In South-Eastern Queensland Hardwood 

Plantations: Identifying Potential Pest Species. Australian Forestry, 69, 270-274. 

Nahrung, H. F., Duffy, M. P., Lawson, S. A. & Clarke, A. R. (2008a) Natural Enemies Of Paropsis Atomaria 

Olivier (Coleoptera: Chrysomelidae) In South-Eastern Queensland Eucalypt Plantations. Australian 

Journal Of Entomology, 47, 188-194. 

Nahrung, H. F., Schutze, M. K., Clarke, A. R., Duffy, M. P., Dunlop, E. A. & Lawson, S. A. (2008b) Thermal 

Requirements, Field Mortality And Population Phenology Modelling Of Paropsis Atomaria Olivier, An 

Emergent Pest In Subtropical Hardwood Plantations. Forest Ecology And Management, 255, 3515-3523. 

Ohmart, C. P. (1991) Role Of Food Quality In The Population Dynamics Of Chrysomelid Beetles Feeding On 

Eucalyptus. Forest Ecology And Management, 39, 35-46. 

Ohmart, C. P., Thomas, J. R. & Stewart, L. G. (1987) Nitrogen, Leaf Toughness And The Population Dynamics 

Of Paropsis Atomaria Olivier (Coleoptera: Chrysomelidae) - A Hypothesis. Journal Of The Australian 

Entomological Society, 26, 203-207. 

Trumble, J. T. & Butler, C. D. (2009) Climate Change Will Exacerbate California's Insect Pest Problems. 

California Agriculture, 63, 73-78. 

Zvereva, E. L. & Kozlov, M. V. (2006) Consequences Of Simultaneous Elevation Of Carbon Dioxide And 

Temperature For Plant-Herbivore Interactions: A Metaanalysis. Global Change Biology, 12, 27-41.



IMPACT OF TREE BORON STATUS ON TREE GROWTH 

AND SUSCEPTIBILITY TO QUAMBALARIA SHOOT BLIGHT 

(QUAMBALARIA PITEREKA) IN CORYMBIA SPECIES 

Timothy Smith 1 and Geoff Pegg 1 

ABSTRACT 

An experiment was conducted to assess the impact of tree boron (B) status on tree growth 

and susceptibility to Quambalaria shoot blight caused by the fungal pathogen 

Quambalaria pitereka. Increasing tree B status from deficient to sufficient was associated 

with increased tree growth and reduced susceptibility to Quambalaria shoot blight. 

Corymbia citriodora sb.sp. variegata Woondum provenance seedlings, Corymbia 

torelliana seedlings and C. torelliana x C. citriodora sb.sp. variegata hybrid (CTVA 1) 

were grown in an aeroponics system under glasshouse conditions and subjected to 6 B 

treatments administered as regular root sprays over a period of 3.5 months. Increasing B 

to roots significantly increased the youngest fully expanded leaf (YFEL) B and provided a 

range in tree B status from deficient to sufficient. A foliar inoculation of Quambalaria 

spores was applied to trees after 3 months and disease infection incidence and severity 

assessments were conducted after 14 days. As tree B status increased from deficient (9-12 

µg g -1 YFEL B) to sufficient B status (22-24 µg g -1 YFEL B) there were increases in tree 

height (107%, P



Lesion development on immature leaves and stems starts on average 5 days after infection by Q. 

pitereka and is visible as discrete chlorotic spots with necrotic centres (Self et al., 2002; Pegg et al., 

2005). These may develop into large, sporulating lesions 10 to 14 days after infection under 

favourable conditions. Conidial germination is triggered when Relative Humidity levels exceeded 90 

% and commencing within two hours in the presence of free water (Pegg et al., 2009). Conidial 

germination and hyphal growth occurs on the upper and lower leaf surfaces with penetration occurring 

via the stomata or wounds on the leaf surface or juvenile stems. Following penetration through the 

stomata, Q. pitereka hyphae grow only intercellularly without the formation of haustoria or interaction 

apparatus. Instead an interaction zone is evident between host and pathogen cells (Pegg et al., 2009). 

Several soils used to support Eucalypt plantations in Eastern Australia are limited in their supply of 

boron (B). In Ferrosol soils (as classified using Isbell, 1986), B was found to be the third most 

limiting element for growth of E. nitens after nitrogen and phosphorus (Smith, 2007). Ferrosols are 

characterised by >5% free iron content (Isbell, 1986). The iron and aluminium oxides in these soils 

have a high affinity for adsorption of B. On this soil type, B deficiency was found to cause poor form 

of E. nitens (Smith, 2007) and Ccv. (Smith, unpublished data). Similarly, Lambert et al. (1997) 

reported that B was the most common micronutrient deficiency limiting growth of Radiata pine (Pinus 

radiata) in extensive areas of Australia, New Zealand and Chile. In China, the Philippines and 

Australia, B deficiency in eucalypts was induced through applications of macronutrients, which 

accelerated tree growth, but B uptake was inadequate to meet demand (Dell et al., 2008). 

The reported roles of B in plants include: bonding of the primary cell wall, specifically the apiose side 

chains of rhamnogalacturonan II fractions of the pectin network (Matoh, 1997; O’Neill et al., 2004); 

stability of membranes (Goldbach et al., 2001), possibly through the formation of complexes with 

membrane glycoproteins (Goldbach and Wimmer, 2007); and anti-oxidative mechanisms associated 

with defence (Cakmak and Romheld, 1997). Interestingly B was found to play a role in quorum 

sensing of the bacterium V. harveyi whereby bonding with B was required to form a biologically 

active signal molecule (Chen et al., 2002). The role of B in plant defence is rapidly being uncovered 

and B has been shown to reduce the severity of a variety of diseases (Dordas, 2008), but the influence 

of plant B concentrations on the development and virulence of fungi is largely unknown. 

This experiment was specifically designed to test the influence of root applied B, at several 

concentrations, on tree growth and susceptibility of treated trees to attack by Q. pitereka. 


Mist culture system 

A mist culture B trial was conducted under glasshouse conditions using a randomised complete block 

design with 6 B treatments, 3 species, 2 solution tank blocks and 3 replicates (108 trees in total). The 

mist culture system used in the experiment is described in detail in Smith (2004). The system was a 

modified version of earlier mist culture designs used by Carter (1942), Martin and Hendrix (1966) and 

Hendrix and Lloyd (1968). The basics of the system consists of 12 nutrient solution tanks (200 l 

double lined food grade polypropylene with screw on lids) which each supply 9 atomiser chambers 

(white 25 l high density polyethylene mist chamber buckets) each. The atomiser chambers were 

randomly allocated to 3 benches in a temperature controlled glasshouse, with 3 chambers per bench. 

The water used in the experiment was purified by passing through a carbon filter, a water softener, a 

reverse osmosis unit (IBC 4000R model), then through an IBC® twin column unit with anion and 

cation resin beds to supply bulk purified water with an electrical conductivity of less than 0.1 µS cm -1 . 

The water was stored in a 3000 l polypropylene tank. It was then passed through an inline UV filter 

and a borate specific resin bed (Amberlite 743 resin), before being used in the experiment. 

The nutrient solutions were made up as concentrates (x 2000) in 20 l containers. Two concentrate 

solutions were used for the basal nutrient solutions, and were termed the ‘red solution’ (which 

contained iron and calcium) and the ‘blue solution’. Basal concentrate solutions were made using



analytical reagent grade salts and quantities are detailed in Table 1. The desired concentrations in 

solution were attained by adding 100 ml of red solution concentrate, 100 ml of blue solution 

concentrate and 100 ml of the relevant boron treatment solution concentrate per 200 l of purified water 

(Table 2). Boron treatments were randomly allocated to tanks and associated mist chambers were 

randomised in the glasshouse. 

Table 1. Essential element concentrations (excluding B) in nutrient solutions atomised on to 

tree roots during the 4-month trial period. 

Element Elemental 

concentration in 

nutrient 

solutions 

Salt used Red solution 

concentrate* 

g salt l -1 

Blue solution 

concentrate* 

g salt l -1 

N 

μM 

357.5 

1414.9 

1400 

NH4NO3, 

KNO3 

Ca(NO3)2.4H2O 

28.6 

2890.8 a 

13224 b 

2830.8 c 

P 129.4 KH2PO4 704.2 d 

K 1544.2 KNO3 & KH2PO4 Inc. above a Ca 1400 Ca(NO3)2.4H2O Inc. above 

Inc. above b 

Mg 411.9 MgSO4.7H2O 4059.2 e 

S 411.9 MgSO4.7H2O Inc. above e 

Fe 14.3 C10H12N2O8FeNa 210.6 

Mn 

Zn 

1.46 

1.23 

MnSO4.H2O 

C10H12N2O8Zn.4H2O 

9.86 

23.1 

Cu 0.315 C10H12CuN2O8.2Na 5.0 

Mo 0.011 Na2MoO4.2H20 0.1 

*Concentrate solutions are 2000x the concentration of that in final nutrient solutions 

Table 2. Boron treatments applied in nutrient solutions atomised on to tree roots during the 4month 

trial period. 

Boron 

treatment 

Elemental concentration in 

nutrient solutions 

μM 

c & d 

Salt used Boron concentrate* 

g salt l -1 

1 0 H3BO3 0 

2 0.078 H3BO3 0.00966 

3 0.3125 H3BO3 0.03864 

4 1.25 H3BO3 0.1546 

5 5.0 H3BO3 0.6183 

6 20.0 H3BO3 2.473 

*Concentrate solutions are 2000x the concentration of that in final nutrient solutions 

The nutrient solutions were pumped using Onga® 442 pressure pumps (240V, 300W, water pressure 

100-200kPa) to the atomiser chambers via 25 mm polypropylene pipe (main lines) with 4 mm 

polypropylene risers connecting to plastic Netfin® mist nozzles fitted through the side of the buckets. 

Control boxes fitted with Hager EG400 programmable timers were used to turn on and off pumps (240 

V) and in-line solenoid valves (12 V) at programmed intervals. 

The main lines circulated surplus solution back into the top of tanks, agitating the nutrient solution in 

the process. Polypropylene bungs with 2 mm holes drilled in the centre were placed on the end of 

each return line into the nutrient solution tanks. This restricted the flow rate to 5.7 ml s -1 back into the 

tank and increased pressure in the line to ensure adequate atomisation of nutrient solutions inside the 

mist chamber buckets. 

Excess nutrient solutions from inside the mist chambers were run to waste to maintain a set nutrient 

concentration contacting the roots. A 13 mm black polypropylene drainpipe was inserted at a height of



20mm from the base of the buckets for draining excess solution to waste. The nutrient solution in the 

base of each bucket maintained humidity in the atomiser chamber between sprays. 

Tree growth 

Two Corymbia species Corymbia torelliana, (open pollinated) (Ct.), C. citriodora subsp. variegata 

(Woondum Bulk) (Ccv.) and one hybrid of these species Ct. x Ccv. (CTVA 1 = CT2-002 x CV2-018) 

were used in this experiment. The trees were propagated as cuttings on the 7 th December 2006 and 

grown under glasshouse conditions at an ambient temperature of 30 o C for 1 month, prior to being 

moved into shade bays to grow to approximately 25cm in height. 

On the 1 st February 2007, a selection of uniform cuttings was removed from 90 mm plastic tray inserts 

(hyco) inserts and potting mix was washed from the plant’s roots using tap water. Plants were placed 

in a container of deionised water once all of the potting mix had been removed. When all 36 plants for 

that species had been prepared they were taken to the glasshouse. Individual plants were placed in 

foam squares with a slit cut halfway into one side. Plants were then placed in a polypropylene concrete 

chair and suspended through holes cut into the lids of the atomiser chamber buckets. Where required, 

plants were pruned back to a single dominant stem. Medium size white garbage bags were placed over 

each plant to produce a humid microclimate for 1 week whilst the trees became established in the 

system. 

The nutrient solutions were atomised onto the roots for a period of 10 seconds at intervals of every 30 

min between 4-6 am, every 20 min between 6–12 am, every 10 min between 12 am-3 pm, every 20 

min between 3–6 pm, every 30 min between 6-8 pm and then every 60 min between 8 pm-4 am 

(totalling 55 times a day). Each atomisation event sprayed approximately 95 ml of nutrient solution 

onto roots in each mist chamber. Therefore, approximately 47 l of nutrient solution per day was used 

to supply 9 trees. Glasshouse temperatures were set at 25 o C days and 20 o C nights. 

The trees were grown in their respective treatments in mist culture and inoculated in the glasshouse 

with a single isolate of Q. pitereka on the 23 rd April 2007. 

Inoculation 

Quambalaria pitereka isolate (BRIP 48385) was obtained from single lesions and grown on potato 

dextrose agar (PDA) for 2 to 3 weeks in the dark at 25 o C. A spore suspension (1x10 6 spores ml -1 ) was 

obtained by washing plates with sterile distilled water (SDW) to which two drops of Tween 20 had 

been added prior to inoculation. 

Foliage of trees was inoculated using a fine mist spray (2.9 kPa pressure) generated by a compressor 

driven spray gun (Iwata Studio series 1/6 hp; Gravity spray gun RG3, Portland, USA), to the upper 

and lower leaf surfaces until runoff was achieved. All trees were covered with plastic bags 

immediately after inoculation to maintain high humidity levels and to increase the period of leaf 

wetness. Bags were removed after 48 hours. Sub-samples of the spore suspension applied to the trees 

were placed onto PDA and incubated at 25 o C for 48 hours to ensure that the spores were viable. 

Disease incidence (I) and severity (S) was assessed for all trees 14 days after inoculation, from which 

a Quambalaria shoot blight (QSB) score was calculated (I xS/100). 


An analysis of variance (ANOVA) was performed on all data using GenStat® version 9 (Lawes 

Agricultural Trust, Rothamsted Experimental Station). For each ANOVA, the mean separation was 

determined by calculating the least significant difference (Fisher’s LSD) at P



RESULTS 

As tree B status increased from deficient (9-12 µg g -1 YFEL B) to sufficient B status (22-24 µg g -1 

YFEL B) there were increases in tree height (107%), stem diameter (62%), total tree weight (194%) 

and root weight (144%)(Tables 3), whilst Quambalaria shoot blight incidence, severity, stem incidence 

and disease score, and resultant shoot death decreased (Table 4). 

Table 3. Effect of root applied B treatment sprays to Corymbia species on foliar B and tree 

growth (averaged across Corymbia species) 

Treatment YFEL B Tree height Stem diameter Total tree Root weight 

(at base) 

weight 

(μM) (mg kg -1 ) (cm) (mm) (g) (g) 

0 11.8 94 1.02 285 95 

0.078 8.9 124 1.21 439 126 

0.3125 9.6 162 1.47 670 194 

1.25 17.4 165 1.21 504 122 

5.0 22.2 191 1.53 805 201 

20.0 23.5 195 1.65 839 233 

P



Table 5. Effect root applied B treatments sprays to Corymbia species on tree growth. 

YFEL B Tree height Total tree weight 

(mg kg -1 Treatment 

(μM) 

) (cm) (g) 

Ct. Ct. x Ccv. Ct. Ct. x Ccv. Ct. Ct. x Ccv. 

Ccv. 

Ccv. 

Ccv. 

0 16.5 8.5 10.3 93 118 70 361 398 96 

0.078 10.5 7.5 8.9 136 143 94 546 560 210 

0.3125 11.7 8.8 8.2 186 173 128 949 764 297 

1.25 19.9 17.2 15.1 159 176 161 670 459 382 

5.0 24.0 18.8 23.8 212 194 167 1182 719 514 

20.0 24.7 20.4 25.5 201 200 184 1226 718 572 

P 0.012



loss of apical dominance leading to bushy crowns and shrub like growth characteristics, and in severe 

cases, tree death (Pegg et al., 2009). Symptom development is rapid and occurs after 7 to 10 days 

under conditions of high humidity. While the identification of host resistance is a key to the further 

development of spotted gum as a plantation species, risk minimisation strategies are also required to 

reduce potential losses. The finding that a balanced tree nutrition approach and specifically B can 

reduce the susceptibility to Q. pitereka infections is encouraging and emphasis needs to be placed on 

the role of nutrition in plant defence as well as tree growth in deciding appropriate nutrient 

management systems in plantation forestry. 

The Ct. x Ccv. hybrid had reductions in Q. pitereka infection rates due to increasing tree B status from 

deficient to sufficient and appreciably lower susceptibility to Q. pitereka infections than Ccv., 

demonstrating a successful outcome of the breeding program (Lee, 2007). A breeding strategy that 

encompasses selection for disease tolerance and nutrient efficiency as well as tree growth and wood 

quality factors will be more suitable for widespread deployment of clones across a range of 

environments. However, even with improved progeny, underlying nutrient deficiencies or imbalances 

need to be addressed to maintain health and vigour in new and existing plantations. 

The function that B plays in plant defence against fungal attack is unclear and further research is 

required to determine the mechanism for reduced susceptibility to Q. pitereka in trees supplied 

adequate B verses B deficient trees. The treatment levels were insufficient to have a direct fungicidal 

effect on the Q. pitereka, however we are uncertain about whether B is playing a direct or indirect role 

in plant defence. Pegg et al. (2009) demonstrated that infection and colonisation of Corymbia spp. by 

Q. pitereka occurs through intercellular growth of hyphae following penetration and remains 

intercellular until host cell death. A localised interaction zone is apparent at points of interaction 

between hyphae and host cell walls. It was assumed that Q. pitereka obtains nutrients from the host 

cells through these interaction zones. While the influence of B on disease development requires more 

in-depth studies, B is known to have a direct function in cell wall structure (Matoh, 1997; O’Neill et 

al., 2004; Brown et al., 2002). The reduction in cell wall permeability to fungal products produced by 

Q. pitereka during the infection process may influence the development of disease. Similarly more 

research is required to investigate the plant defence implications of other reported roles of B in the 

stability of membranes (Goldbach et al., 2001) and anti-oxidative mechanisms (Cakmak and Romheld, 

1997). 

These studies have shown that B is required for early tree growth of Corymbia species and hybrids, 

and that increasing tree B to sufficient reduced the susceptibility to Q. pitereka infections in the 

establishment phase. Critical foliar B concentrations appear to be in the order of 17.2-20.4 mg B kg -1 

for the Ct. x Ccv. hybrid and >25 mg B kg -1 for Ccv.. However, more research is required to develop a 

better understanding of pathogen biology, such as variability in isolate aggressiveness and the effect of 

repeated infection on overcoming host resistance to help formulate a disease management strategy and 

to determine the effectiveness of increasing tree B status to sufficient on disease rates of trees in 

plantations that are already infected or have older foliage that may have different host disease 

interactions. It must also be stressed that due care needs to be taken with B application rates which 

vary depending on soil type and soil B adsorption characteristics (Smith, 2004), to avoid B toxicity 

scenarios, particularly on course textured soils. Aiming towards a balanced nutrient management 

system in plantation forestry is desirable for plant health and applications of B alone are unlikely to 

achieve that goal if other essential elements are in limited supply. 


The authors would like to acknowledge the valuable contributions of Donna Richardson for 

maintaining the trial and assistance with measures and other operations.



REFERENCES 

Atkinson D and McKinlay RG 1997 Crop protection and its integration within sustainable farming systems. 

Agriculture, Ecosystems and Environment 64, 87-93 

Brown PH, Bellaloui N, Wimmer MA, Bassil ES, Ruiz J, Hu H, Pfeffer H, Dannel F, Romheld V 2002, Boron in 

plant biology. Plant Biology 4,205-223 

Burgess TI and Wingfield MJ 2002, Impact of fungi in natural forest ecosystems; A focus on Eucalyptus. In 

`Microorganisms in Plant Conservation and Biodiversity’ (Eds. K Sivasithamparam, KW Dixon, RL 

Barrett). Kluwer Academic Publishers, Dordrecht, pp. 285-306 

Cakmak I and Romheld V 1997, Boron deficiency-induced impairments of cellular functions in plants. Plant and 

Soil 193, 71-83 

Carnegie AJ 2007, Forest health condition in New South Wales, Australia, 1996-2005. II. Fungal damage 

recorded from eucalypt plantations during forest health surveys and their management. Australasian 

Plant Pathology 36, 225-39 

Carter W 1942, A method of growing plants in water vapour to facilitate examination of roots. Phytopathology. 

19, 407-412 

Chen X, Schauder S, Potier N, Van Dorsselaer A, Pelczer I, Bassler BL, Hughson FM 2002, Structural 

identification of a bacterial quorum-sensing signal containing boron. Nature 415 (6871), 545-549 

Dell B, Hardy G, and Burgess T 2008, Health and nutrition of plantation eucalypts in Asia. Southern Forests 

70(2), 131-138 

Dordas C 2008, Role of nutrients in controlling plant diseases in sustainable agriculture. A review. Agronomy 

and Sustainable Development 28, 33-46 

Drenth A 2004, Fungal epidemics – does spatial structure matter? New Phytologist 163, 4-7. 

Goldbach HE, Yu Q, Wingender R, Schulz M, Wimmer M, Findeklee P, and Baluska F 2001, Rapid response 

reactions of roots to boron deprivation. Journal of Plant Nutrition and Soil Science 164, 173-181 

Goldbach HE and Wimmer M 2007, Boron in plants and animals: Is there a role beyond cell-wall structure? 

Journal of Plant Nutrition and Soil Science 170, 39-48 

Hendrix JW and Lloyd EH 1968, An improved Washington State University mist chamber for root disease and 

root research. Washington Agricultural Experimental Station Bulletin 700, 5 

Isbell RF 1986, `The Australian Soil Classification’. CSIRO Publishing, Collingwood, Victoria. pp. 41-44 

Lambert MJ, Turner J, and Knott J 1997, Boron nutrition of radiate pine plantations in Australia. In `Boron in 

Soils and Plants’ (Eds. R.W Bell and B. Rerkasem). Kluwer Academic Publishers, Netherlands. pp. 83-88 

Lee D 2007, Development of Corymbia species and hybrids for plantations in eastern Australia. Australian 

Forestry 70, 11-16 

Martin NE and Hendrix JW 1966, An apparatus for the mist culture of wheat. Plant Disease Reporter. 50(6), 

369-371 

Matoh T 1997, Boron in plant cell walls. Plant and Soil 193, 59-70 

Old KM 1990, Diseases caused by fungi. In `Trees for Rural Australia’ (Ed. KW Cremer). Inkata Press, 

Melbourne, Australia. 

O’Neill MA, Ishii T, Albersheim P, and Darvill AG 2004, Rhamnogalacturonan II: Structure and function of a 

borate cross-linked cell wall pectic polysaccharide. Annual Review of Plant Biology 55, 109-139 

Pegg GS, Drenth A, Wingfield MJ 2005, Quambalaria pitereka on spotted gum plantations in Queensland and 

northern New South Wales, Australia. International Forestry Review 7, 337. 

Pegg GS, O’Dwyer C, Carnegie AJ, Wingfield MJ, Drenth A 2008, Quambalaria species associated with 

plantation and native eucalypts in Australia. Plant Pathology 57, 702-14 

Pegg GS, Webb RI, Carnegie AJ, Wingfield MJ, Drenth A 2009, Infection and disease development of 

Quambalaria spp. on Corymbia and Eucalyptus species. Plant Pathology Doi:10.1111/j.1365- 

3059.2009.02087.x 

Smith TE 2004, Boron nutrition of Hass avocado (Persia americana Mill.). PhD Thesis, the University of 

Queensland, awarded 19 th July 2004. 196 p. 

Smith TE 2007, Poor tree form in Eucalyptus nitens linked to boron deficiency. In `Advances in Plant and 

Animal Boron Nutrition’ (Eds. F Xu, H.E. Goldbach, PH Brown, RW Bell, T Fujiwara, CD Hunt, S 

Goldberg, and L Shei). Springer. pp 163-169 

Self NM, Aitken, EAB, Dale MD 2002, Susceptibility of provenances of spotted gums to Ramularia shoot blight. 

New Zealand Plant Protection 55, 68-72. 

Walker J, Bertus AL 1971, Shoot blight of Eucalyptus spp. caused by an undescribed species of Ramularia. 

Proceedings of the Linnean Society of New South Wales. 96, 108-15 

Wingfield MJ, Slippers B, Roux J, Wingfield BD 2001, Worldwide movement of exotic forest fungi especially 

in the tropics and Southern Hemisphere. Bioscience 51, 134-40



ABSTRACT 

THE INFLUENCE OF CLIMATE CHANGE ON 

SOIL RESPIRATION WITH EUCALYPTUS SALIGNA 

Vivien de Rémy de Courcelles 1 , Bhupinderpal-Singh, Mark Adams 

Belowground respiration takes a major part in the cycle of carbon between the 

atmosphere and the soil; the soil possesses the biggest pool of the terrestrial carbon. It 

might be influenced by the current rapid increase in atmospheric [CO2]. Within the 

Hawkesbury Forest Experiment (HFE) at the University of Western Sydney (Richmond), 

twelve Whole Tree Chambers (WTC) were installed. Each WTC enclosed a Eucalyptus 

saligna tree that was grown from a seedling stage to a maximum of 10 m height for up to 

2 years. These trees in WTC were subjected to a factorial combination of ambient or 

elevated atmospheric [CO2] and irrigation or drought treatments. We measured the soil 

CO2 efflux from March 2008 to February 2009. First results showed strong inter-seasonal 

variations in soil respiration with more respiration occurring during the warmer months. 

There was no significant difference between the two CO2 treatments. On the other hand 

the irrigation treatment resulted in a significant higher rate of soil respiration than the 

drought treatment. 

INTRODUCTION 

The atmospheric concentration of carbon dioxide has increased by 36% since the mid 18 th century, 

from a pre-industrial value of about 280 ppm to 379 ppm in 2005, and continues to increase by an 

average of 19 ppm/year (Forster et al., 2007). Soil CO2 efflux (SCE), commonly referred to as 

belowground or soil respiration, is a major flux of carbon dioxide toward the atmosphere. Averaged 

over 18 European forests, SCE equaled 69% of the total terrestrial ecosystem respiration (Janssens et 

al., 2001). In other words, an estimated 10% of the carbon dioxide of the atmosphere cycles through 

soils each year (Raich and Potter, 1995), which represents about 10 times the amount of CO2 produced 

by burning fossil fuels. 

Despite its importance in the carbon cycle, belowground respiration is not as well known as its aboveground 

counterpart. In part this is due to the difficulty of isolating and quantifying its components. 

Belowground soil respiration must be divided between autotrophic and heterotrophic sources. 

Autotrophic respiration can be clearly identified as root respiration (Kuzyakov, 2006). On the other 

hand, heterotrophic respiration includes respiration from free-living microorganisms in soil (Cisneros- 

Dozal et al., 2006). Because of the methodological constraints, respiration from rhizosphere 

microorganisms, including mycorrhizal fungi, is usually coupled with the autotrophic respiratory 

component and the argument for this is that these microorganisms are highly dependent on plants for 

their supply of recently-fixed carbohydrates via photosynthesis. In the strict sense, soil respiration 

from these sources should however be classified as heterotrophic respiration. 

The increase in atmospheric [CO2] and the resulting changes in climate can impact on part or all 

components of soil respiration. Equally, the response in soil CO2 efflux will depend on the land use, 

species composition, age class of the vegetation as well as disturbances other than climatic (Zak et al., 

2000). 

Because of the size of trees and that of their systems, field based experimental designs for studying the 

effect of climate change on trees and forest require heavy and costly equipment. Some of those 

devices include open top chambers (OTC) (Leadley and Drake, 1993, Mandl et al., 1973) and closed 

top or whole tree chambers (WTC) (Kellomaki et al., 2000, Medhurst et al., 2006) at the tree level, 

whereas Free Air CO2 Enrichment (FACE) (Hendrey et al., 1999, Lewin et al., 1994) was used for 

studies at the plot level. 

1 University of Sydney. Email: v.decourcelles@usyd.edu.au.



Here we present results gathered over a year of an experiment where Eucalyptus saligna are growing 

in WTC under the climate change conditions that are predicted to happen within the next 100 years in 

eastern Australia. 

MATERIALS AND METHOD 

Site description 

The Hawkesbury Forest Experiment (HFE) is located at the University of Western Sydney campus in 

Richmond, North-West of Sydney, Australia (Lat. 33°36’40” S, Lon. 150°44’26.5” E). It sits on the 

alluvial floodplain of the Hawkesbury River at elevation 25 m a.s.l. The soil is a chromosol, sandy in 

texture, with a pH around 5.5 in the first 40 cm, and low in C (



Two environmental variables were employed inside the chambers: moisture and CO2 concentration. 

At the start of the experiment all WTC and outside control plots were irrigated with 10 mm of water 

every third day to ensure a good establishment of the trees. A first drought cycle started on 16 

February 2008 when all irrigation ceased for half of the experimental plots. The drought was broken 

after more than 200 days on 18 September 2008 in order to rewet the upper soil level. A second 

drought cycle started on 26 October 2008 and lasted until the end of the experiment. 

The [CO2] was either ambient or elevated and the treatment started at the day of planting. The ambient 

CO2 varied between 380 ppm and 500 ppm depending on day or night-time, and elevated CO2 was 

maintained 240 ppm above the ambient level. 

There were 3 replicates of each combination of water and CO2 treatment, i.e. ambient [CO2]-dry, 

ambient [CO2]-irrigated, elevated [CO2]-dry, elevated [CO2]-irrigated. 

Six controls including one tree each were also set up to evaluate the WTC effect on the measured 

parameters: a frame bearing a plastic sheet was installed 300 to 200 mm over the soil and around the 

stem preventing rainfall from reaching the soil. Sprinklers controlled the moisture delivered to the soil 

around the trees in accord with the same irrigated and drought treatments applied inside the WTC. 

Belowground respiration measurements 

Soil respiration measurements were first done using static chambers 204 mm tall and 250 mm outside 

diameter for 236 mm internal diameter. Chambers were inserted in the soil in February 2008 within 

each WTC as well as in the control plots to a depth of about 50mm: Two chambers were placed at 

about 500 mm from the stem of the trees and 2 more at about 1 metre distance. Measurements were 

taken on a monthly basis from March to May, and again in September and October 2008. A 25 ml gas 

sample was taken when the lid was closed and immediately transferred to an evacuated 12ml 

Exetainer tube. A second sample was taken 30 minutes later. Samples were analysed on a CO2 gas 

analyser at the University of Western Sydney (UWS). The rate of soil respiration was extrapolated 

from the slope of the concentration line joining the two measurement points. In a separate campaign, 

we collected the CO2 gas samples at three intervals (0, 30, 60 min) over a 60 min period in order to 

check for linearity in the gas accumulation rate inside the chamber. We always found a strong linearity 

(R2 > 0.95) in the gas accumulation rate. 

In September 2008, we installed 2 sets of cores into the soil of each WTC and control plots based on 

the design by Heinemeyer et al (Heinemeyer et al., 2007). Each set consists of one 50 mm tall x 200 

mm diameter ring pressed against the soil. Two 250 mm tall x 200 mm diameter cores were inserted to 

a depth of 200 mm into the soil. They were equipped with four 50 x 50 mm windows on their side 

fitted with a 38-μm nylon mesh or a 1-μm nylon mesh that allows or prevents the in-growth of 

mycorrhizal hyphae inside the cores, respectively (see Heinemeyer et al., 2007 for details). A deeper 

700 mm long and 100 mm diameter core without windows was also inserted in each WTC or control 

plot. It excludes any roots, including the deeper growing ones. The timeframe for this study did not 

allow enough time for roots to decompose inside the mesh-collars, and only the results from the 

shallow collars are considered here. 

We measured soil respiration out of these mesh-collars with a Vaisala carbocap GM 343 mounted on a 

5.25 litter lid. The lid-Vaisala-collar association works as a static chamber equipped with its own CO2 

analyser. The lid was left on the collars for 10 minutes out of which only the last eight were 

considered in order to calculate the rate of evolution of CO2. The recording during the first 2 minutes 

was discarded in order to allow the [CO2] to stabilise. 

Respiration measurements were carried out in the first week of November and December and then 

every 3 weeks until 9 February. 

In order to compare the respiration results obtained using static chambers-gas sampling and Vaisala 

direct measurements, both techniques were used in October. Out of ten plots, seven had a soil 

respiration measured with the Vaisala within ±20% of the respiration measured using the static 

chambers; one was about 21% higher and the remaining two were more than 50% lower than those 

measured using the static chambers.




Belowground respiration between [CO2] treatments, between ambient WTC and outside controls, and 

between irrigated and drought treatments were analysed using two-way ANOVA. 

RESULTS 

Respiration rate over the year 

Respiration rate was very variable throughout the year for WTC and control plots. It varied between 

158−181 mg CO2/m 2 /h in September or May 2008 to between 416−502 mg CO2/m 2 /h in March 2008 

or February 2009, indicating that soil respiration was higher in the warmer than colder months. 

Response to the [CO2] treatment 

Soil respiration followed a similar annual pattern in the elevated and ambient [CO2] WTC (Figure 1). 

Soil CO2 efflux for the ambient plots ranged from a low 181 mg CO2/m 2 /h in September 2008 to a 

high 502 mg CO2/m 2 /h in December 2008. The elevated [CO2] plots released a minimum of 158 mg 

CO2/m 2 /h in September 2008 and a maximum of 416 mg CO2/m 2 /h in December 2008 (Figure 2). 

The belowground efflux from soil of the ambient WTC was greater than the elevated ones except in 

March and April 2008. However, there was no significant difference (P>0.05) in belowground 

respiration between ambient or elevated plots for any of the measurement days, apart from January 

2009. 

SCE (mgCO2/m 2 /h) 

600 

500 

400 

300 

200 

100 

0 

1/3/08 1/5/08 1/7/08 1/9/08 1/11/08 1/1/09 1/3/09 

Figure 2. Time series of Soil CO2 efflux (SCE) at HFE averaged over the 6 control plots (…○…), 

the 6 elevated [CO2] WTC (–■–) and the 6 ambient [CO2] (–□–) displayed with 

negative standard error bars 

Response to irrigation treatment 

Across all treatments, the soil of the irrigated plots produced significantly more carbon dioxide than 

the soil of the dry plots, except on three occasions in March, April and October 2008 (Figure 3). In 

April, the soil respiration values were not significantly different across the WTC regardless of the 

treatment, but there was a significant difference between wet and dry treatment in the control plots. 

The irrigated control plots showed a significantly (P< 0.05) higher rate of soil respiration compared to 

the droughted ones in April and from December to January inclusive, with a less marked difference in 

October (P=0.1) and November (P=0.057). Inside the ambient [CO2] WTC, during the summer 

months, the rate of soil respiration was significantly (P



SCE (mgCO2/m 2 /h) 

600 

500 

400 

300 

200 

100 

0 

1/3/08 1/5/08 1/7/08 1/9/08 1/11/08 1/1/09 1/3/09 

Figure 3. Time series of Soil CO2 efflux (SCE) measured at HFE between March 2008 and 

February 2008 plotted against the watering treatment. Mean of 3 control plots, 3 

ambient WTCs and 3 elevated WTCs subjected to drought treatment (receiving no 

water) (–■–) or water treatment (irrigated) ( … ○ … ) 

Interaction between treatments 

No interaction between [CO2] and watering treatment was recorded for any measurement day or across 

all measurements. 

Chamber effect 

There was a significant difference in belowground soil respiration rates between the open air control 

plots and the ambient WTC plots in March, September and from November 2008 to February 2009. 

However in March and September 2008, the soil in the control plots released more carbon dioxide 

than the soil inside the ambient WTC whereas the trend was reversed from November 2008 (see 

Figure 1). 

DISCUSSION 

Elevated [CO2] effect 

For most of the measurements, more carbon dioxide was emitted from soils in the ambient than the 

elevated [CO2] WTC, but there was no significant difference except for January 2009. Most studies 

noted an increase in carbon dioxide efflux from soils under vegetation exposed to an atmosphere with 

elevated [CO2] (Bernhardt et al., 2006, Hamilton et al., 2002, Lin et al., 2001, Pajari, 1995, Wan et al., 

2007, Zak et al., 2000). Others have not recorded any effect of elevated atmospheric CO2 enrichment 

(Oberbauer et al., 1986, Tingey et al., 2006). At time of writing, we are yet to link specific soil 

respiration with the growth of the tree for each plot. The results will be presented during the 

conference. 

While root respiration has been evaluated as representing 10 to 90 % of total soil respiration in 

different studies (Hanson et al., 2000), the development of new techniques has allowed to refine these 

values. Root and mycorrhizal respiration accounted for up to 56% of total soil respiration for the year 

following a tree girdling experiment (Hogberg et al., 2001); this difference was increased up to 65% in 

the second year i.e. when contributions from stored carbohydrates in roots and from decaying 

mycorhizal roots were reduced to a very low value two years after the girdling treatment 

(Bhupinderpal et al., 2003). Heinemeyer et al. (2007) found that heterotrophic respiration accounted 

for 60% of total soil CO2 efflux, whereas ectomycorrhizas and roots contributed about 25% and 15%, 

respectively. In our experiment, we didn’t achieve that same level of separation of the components of 

soil respiration using similar mesh-collar chambers as used by Heinemeyer et al. (2007) in a boreal 

forest soil system. However, our 700 mm long collar must have excluded most roots and their



associated mycorrhiza and therefore the respiration recorded on these collars must be of ‘true 

heterotrophic’ origin only. Therefore, that portion of soil respiration due only to contribution from 

roots was, on average, 63% in November and 66% for the two measurements made in December 2008. 

There was no difference between CO2 treatments. It should be noted that the floor of the WTCs 

prevented litter from falling on the ground. We have therefore missed the input of microbial 

decomposers feeding on the surface litter that would be considered in a natural environment, and thus 

these numbers are probably an overestimation of the contribution of root respiration to the total 

belowground CO2 efflux. 

Comparison across studies can be deceptive since the intensity of stimulation of soil respiration varies 

with the age or the composition of the stands (King et al., 2004, Pregitzer et al., 2006) and the length 

of exposition to elevated [CO2] (King et al., 2004). Kasurinen et al. (2004) observed a positive effect 

of elevated [CO2] on soil carbon dioxide efflux under one clone of silver birch (Betula pendula) 

whereas the effect was negative on the soil respiration under another clone with the difference 

increasing after three years. Therefore conclusions should not be drawn before considering the 

particular conditions of any experimental design. 

Irrigation treatment 

Our results of greater soil CO2 efflux from the watered WTCs than from the droughted WTCs are 

consistent with the results of many other studies. In Brazil, irrigated clones of Eucalyptus grandis x 

urophylla produced a higher rate of soil respiration than the rainfed counterparts over 2 years of 

experiment (Stape et al., 2008). The drought and rainy periods of the year were matched by 

respectively low and high rates of SCE in eucalypts forest (Stape et al., 2008) as well as in conifers, 

broadleaf mixed and evergreen broadleaf forests (Tang et al., 2006). 

Jassal et al. (2008) have found that water stress in a Douglas fir forest significantly reduced carbon 

dioxide emissions from soils. Furthermore, under a threshold of 11 m 3 .m -3 of soil water at 4 cm depth, 

belowground respiration was not affected by soil temperature, but was linearly dependent on soil 

moisture content. The drought treatment in our experiment achieved a similar reduction in soil CO2 

efflux most of the time. The only exceptions occurred right at the start of both drought cycles. The 

most notable one was the October measurement following irrigation of the dry plots. They showed a 

greater, though not significant, belowground respiration rate than did their irrigated counterparts. 

The increase in soil respiration rate occurring after the rewatering of soils previously submitted to a 

lengthy dry episode, is called “Birch effect” named after H.F. Birch, one of the first to describe this 

phenomenon (Jarvis et al., 2007). Four reasons are commonly listed to explain the Birch effect: 

1. Successive drying and wetting of soils make organic matter available for 

decomposition by breaking soil aggregates. 

2. The dry spells kill soil micro-organisms which are subsequently decomposed when 

wetting occurs, hence releasing nutrients. 

3. Wetting triggers a sudden growth of microbial biomass and fungal hyphae. 

4. Microbes respond to the osmotic shock created by the rewetting by releasing carbon 

rich solutes that accumulated in their cytoplasm during the dry periods (Jarvis et al., 

2007). 

Recent studies have shown both an increase in yearly forest soil CO2 emissions (Jarvis et al., 2007) 

and no extra release of CO2 (Muhr et al., 2008) following the wetting of previously dry soils. In our 

case, the increase of SCE in the dry plots following watering, led to a higher rate of respiration than 

the irrigated plots but not significantly so. Moreover, the effect of the rewatering was cancelled by the 

following month measurement. These results would tend to indicate that over a year no significant 

amount of extra CO2 would be emitted from our soils because of the Birch effect. 

The Australian climate of the 21 st century is predicted to bring longer dry spells broken by heavier 

rainfall episodes (CSIRO and BOM, 2007). These irregular wetting-drying events make prediction of 

the pattern of future belowground respiration a very hard task.



Chamber effect 

Similar chamber effects (i.e. greater soil respiration in the control plots than in the WTC) as the one 

we recorded in March and September 2008 have been observed by Niinisto et al (2004) in their WTC 

experiment. However, over the 4 years of their experiment, there was no reversal of the trend of 

greater soil respiration in the control plots than in the ambient WTC to lower respiration in the control 

plots than in the WTCs. 

Belowground respiration is linked positively to the temperature measured at the soil surface (Niinisto 

et al., 2004, Rustad et al., 2001, Bronson et al., 2008, Pajari, 1995) as well as the soil moisture (Keith 

et al., 1997, Tang et al., 2006, Orchard and Cook, 1983). Our measurements of soil temperatures 

(result not presented) show that for both the September and October measurements, soils in the outside 

control plots were on average 1.4ºC warmer than in the ambient WTC, a significant difference, 

whereas the difference ranged from a 0.1 ºC cooler to a 0.9 ºC warmer for the other months. Except in 

May, soil moisture was always lower in the control than the WTC plots. Both these factors play a role 

in the soil CO2 efflux and their respective importance might change according to the seasons: soil 

moisture can become a limiting factor in summer whereas temperature would be a limiting factor in 

winter. Their interaction is a determinant for soil respiration and would explain most variations. The 

WTC experimental design at HFE has an effect on soil conditions that varies between seasons; this 

could explain the observed switch from greater to lower soil respiration in the control plots compared 

to ambient WTC. 

CONCLUSIONS 

Increase in the atmospheric [CO2] from an ambient level to an elevated level did not induce significant 

changes in the soil CO2 efflux from a Eucalyptus saligna plantation. However, compared to the 

irrigated treatment, the drought treatment led to a significant decrease in soil CO2 efflux. . 

Considering the predictions of increasing dry spells in future climate change scenarios in Australia, 

soil respiration could be predicted to decrease in the next few years; however, irregular rainfall events 

may lead to greater soil respiration than in areas subjected to long dry spells. Because of the chamber 

effect that we observed and the relatively short duration of the present study, extrapolation of these 

results to field and ecosystem scales should be undertaken with caution. 

Long-term, studies are needed to better understand the future effect of climate change on soil 

respiration. 

REFERENCES 

BERNHARDT, E. S., BARBER, J. J., PIPPEN, J. S., TANEVA, L., ANDREWS, J. A. & SCHLESINGER, W. 

H. (2006) Long-term effects of free air CO2 enrichment (FACE) on soil respiration. Biogeochemistry, 77, 

91-116. 

BHUPINDERPAL, S., NORDGREN, A., LOFVENIUS, M. O., HOGBERG, M. N., MELLANDER, P. E. & 

HOGBERG, P. (2003) Tree root and soil heterotrophic respiration as revealed by girdling of boreal Scots 

pine forest: extending observations beyond the first year. Plant Cell and Environment, 26, 1287-1296. 

BRONSON, D. R., GOWER, S. T., TANNER, M., LINDER, S. & VAN HERK, I. (2008) Response of soil 

surface CO2 flux in a boreal forest to ecosystem warming. Global Change Biology, 14, 856-867. 

CISNEROS-DOZAL, L. M., TRUMBORE, S. & HANSON, P. J. (2006) Partitioning sources of soil-respired 

CO2 and their seasonal variation using a unique radiocarbon tracer. Global Change Biology, 12, 194-204. 

CSIRO & BOM (2007) Climate change in Australia-Technical report 2007. Australian Bureau of Meteorology 

(BOM) 

FORSTER, P., RAMASWAMY, V., ARTAXO, P., BERNTSEN, T., BETTS, R., FAHEY, D. W., HAYWOOD, 

J., LEAN, J., LOWE, D. C., MYHRE, G., NGANGA, J., PRINN, R., RAGA, G., SCHULZ, M. & 

DORLAND, A. R. V. (2007) Changes in Atmospheric Constituents and in Radiative Forcing. IN 

SOLOMON, S., QIN, D., MANNING, M., CHEN, Z., MARQUIS, M., AVERYT, K. B., TIGNOR, M. & 

MILLER, H. L. (Eds.) Climate Change 2007: The Physical Science Basis. Contribution of Working 

Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, 

United Kingdom and New York, NY, USA, Cambridge University Press. 

HAMILTON, J. G., DELUCIA, E. H., GEORGE, K., NAIDU, S. L., FINZI, A. C. & SCHLESINGER, W. H. 

(2002) Forest carbon balance under elevated CO2. Oecologia, 131, 250-260.



HANSON, P. J., EDWARDS, N. T., GARTEN, C. T. & ANDREWS, J. A. (2000) Separating root and soil 

microbial contributions to soil respiration: A review of methods and observations. Biogeochemistry, 48, 

115-146. 

HEINEMEYER, A., HARTLEY, I. P., EVANS, S. P., CARREIRA DE LA FUENTE, J. A. & INESON, P. 

(2007) Forest soil CO2 flux: uncovering the contribution and environmental responses of 

ectomycorrhizas. Global Change Biology, 13, 1786-1797. 

HENDREY, G. R., ELLSWORTH, D. S., LEWIN, K. F. & NAGY, J. (1999) A free-air enrichment system for 

exposing tall forest vegetation to elevated atmospheric CO2. Global Change Biology, 5, 293-309. 

HOGBERG, P., NORDGREN, A., BUCHMANN, N., TAYLOR, A. F. S., EKBLAD, A., HOGBERG, M. N., 

NYBERG, G., OTTOSSON-LOFVENIUS, M. & READ, D. J. (2001) Large-scale forest girdling shows 

that current photosynthesis drives soil respiration. Nature, 411, 789-792. 

JANSSENS, I. A., LANKREIJER, H., MATTEUCCI, G., KOWALSKI, A. S., BUCHMANN, N., EPRON, D., 

PILEGAARD, K., KUTSCH, W., LONGDOZ, B., GRUNWALD, T., MONTAGNANI, L., DORE, S., 

REBMANN, C., MOORS, E. J., GRELLE, A., RANNIK, U., MORGENSTERN, K., OLTCHEV, S., 

CLEMENT, R., GUDMUNDSSON, J., MINERBI, S., BERBIGIER, P., IBROM, A., MONCRIEFF, J., 

AUBINET, M., BERNHOFER, C., JENSEN, N. O., VESALA, T., GRANIER, A., SCHULZE, E. D., 

LINDROTH, A., DOLMAN, A. J., JARVIS, P. G., CEULEMANS, R. & VALENTINI, R. (2001) 

Productivity overshadows temperature in determining soil and ecosystem respiration across European 

forests. Global Change Biology, 7, 269-278. 

JARVIS, P., REY, A., PETSIKOS, C., WINGATE, L., RAYMENT, M., PEREIRA, J., BANZA, J., DAVID, J., 

MIGLIETTA, F., BORGHETTI, M., MANCA, G. & VALENTINI, R. (2007) Drying and wetting of 

Mediterranean soils stimulates decomposition and carbon dioxide emission: the "Birch effect". Heron 

Publishing. 

JASSAL, R. S., BLACK, T. A., NOVAK, M. D., GAUMONT-GUAY, D. & NESIC, Z. (2008) Effect of soil 

water stress on soil respiration and its temperature sensitivity in an 18-year-old temperate Douglas-fir 

stand. Global Change Biology, 14, 1305-1318. 

KASURINEN, A., KOKKO-GONZALES, P., RIIKONEN, J., VAPAAVUORI, E. & HOLOPAINEN, T. (2004) 

Soil CO2 efflux of two silver birch clones exposed to elevated CO2 and O-3 levels during three growing 

seasons. Global Change Biology, 10, 1654-1665. 

KEITH, H., JACOBSEN, K. L. & RAISON, R. J. (1997) Effects of soil phosphorus availability, temperature and 

moisture on soil respiration in Eucalyptus pauciflora forest. Plant and Soil, 190, 127-141. 

KELLOMAKI, S., WANG, K. Y. & LEMETTINEN, M. (2000) Controlled environment chambers for 

investigating tree response to elevated CO2 and temperature under boreal conditions. Photosynthetica, 38, 

69-81. 

KING, J. S., HANSON, P. J., BERNHARDT, E., DEANGELIS, P., NORBY, R. J. & PREGITZER, K. S. 

(2004) A multiyear synthesis of soil respiration responses to elevated atmospheric CO2 from four forest 

FACE experiments. Global Change Biology, 10, 1027-1042. 

KUZYAKOV, Y. (2006) Sources of CO2 efflux from soil and review of partitioning methods. Soil Biology & 

Biochemistry, 38, 425-448. 

LEADLEY, P. W. & DRAKE, B. G. (1993) Open top chamber for exposing plant canopy to elevated CO2 

concentration and for measuring net gas-exchange Vegetatio, 104, 3-15. 

LEWIN, K. F., HENDREY, G. R., NAGY, J. & LAMORTE, R. L. (1994) Design and application of a free air 

carbon-dioxide enrichment facility. Agricultural and Forest Meteorology, 70, 15-29. 

LIN, G., RYGIEWICZ, P. T., EHLERINGER, J. R., JOHNSON, M. G. & TINGEY, D. T. (2001) Timedependent 

responses of soil CO2 efflux components to elevated atmospheric [CO2] and temperature in 

experimental forest mesocosms. Plant and Soil, 229, 259-270. 

MANDL, R. H., WEINSTEIN, L. H., MCCUNE, D. C. & KEVENY, M. (1973) A Cylindrical, Open-Top 

Chamber for the Exposure of Plants to Air Pollutants in the Field. J Environ Qual, 2, 371-376. 

MEDHURST, J., PARSBY, J., LINDER, S., WALLIN, G., CESCHIA, E. & SLANEY, M. (2006) A whole-tree 

chamber system for examining tree-level physiological responses of field-grown trees to environmental 

variation and climate change. Plant Cell and Environment, 29, 1853-1869. 

MUHR, J., GOLDBERG, S. D., BORKEN, W. & GEBAUER, G. (2008) Repeated drying-rewetting cycles and 

their effects on the emission of CO2, N2O, NO, and CH4 in a forest soil. Wiley-V C H Verlag Gmbh. 

NIINISTO, S. M., SILVOLA, J. & KELLOMAKI, S. (2004) Soil CO2 efflux in a boreal pine forest under 

atmospheric CO2 enrichment and air warming. Global Change Biology, 10, 1363-1376. 

OBERBAUER, S. F., OECHEL, W. C. & RIECHERS, G. H. (1986) Soil respiration of Alaskan toundra at 

elevated atmospheric concentration. Plant and Soil, 96, 145-148. 

ORCHARD, V. A. & COOK, F. J. (1983) Relationship between soil respiration and soil moisture. Soil Biology 

& Biochemistry, 15, 447-453.



PAJARI, B. (1995) Soil respiration in a poor upland site of Scots pine stand subjected to elevated temperatures 

and atmospheric carbon concentration. Plant and Soil, 168-169, 563-570. 

PREGITZER, K., LOYA, W., KUBISKE, M. & ZAK, D. (2006) Soil respiration in northern forests exposed to 

elevated atmospheric carbon dioxide and ozone. Oecologia, 148, 503-516. 

RAICH, J. W. & POTTER, C. S. (1995) Global patterns of carbon-dioxide emissions from soils. Global 

Biogeochemical Cycles, 9, 23-36. 

RUSTAD, RUSTAD, L., CAMPBELL, CAMPBELL, J., MARION, MARION, G., NORBY, NORBY, R., 

MITCHELL, MITCHELL, M., HARTLEY, HARTLEY, A., CORNELISSEN, CORNELISSEN, J., 

GUREVITCH, GUREVITCH, J., GCTE, N. & GCTE, N. (2001) A meta-analysis of the response of soil 

respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem 

warming. Oecologia, 126, 543-562. 

STAPE, J. L., BINKLEY, D. & RYAN, M. G. (2008) Production and carbon allocation in a clonal Eucalyptus 

plantation with water and nutrient manipulations. Forest Ecology and Management, 255, 920-930. 

TANG, X. L., ZHOU, G. Y., LIU, S. G., ZHANG, D. Q., LIU, S. Z., LI, J. & ZHOU, C. Y. (2006) Dependence 

of soil respiration on soil temperature and soil moisture in successional forests in southern China. Journal 

of Integrative Plant Biology, 48, 654-663. 

TINGEY, D. T., LEE, E. H., WASCHMANN, R., JOHNSON, M. G. & RYGIEWICZ, P. T. (2006) Does soil 

CO2 efflux acclimatize to elevated temperature and CO2 during long-term treatment of Douglas-fir 

seedlings? New Phytologist, 170, 107-118. 

WAN, S., NORBY, R. J., LEDFORD, J. & WELTZIN, J. F. (2007) Responses of soil respiration to elevated 

CO2, air warming, and changing soil water availability in a model old-field grassland. Global Change 

Biology, 13, 2411-2424. 

ZAK, D. R., PREGITZER, K. S., KING, J. S. & HOLMES, W. E. (2000) Elevated atmospheric CO2, fine roots 

and the response of soil microorganisms: a review and hypothesis. New Phytologist, 147, 201-222.



BUSHFIRE GOVERNANCE IN AUSTRALIA IN 2009: 

A NOTE TO A FUTURE HISTORIAN 

Roger Underwood 1 

ABSTRACT 

Historians at the end of the 21st century will look back with bewilderment to the 

repeated bushfire disasters in Australia during the period 1983-2009. They will be 

perplexed as to why an educated, civilised, forest-loving, prosperous and science-based 

community adopted a forest management system that made effective bushfire 

management impossible. This paper will provide one perspective. I focus on the 

deficiencies in public policy, the lack of leadership and accountability, the capacity for 

non-accountable pressure groups to influence political decision-making, and the 

transfer of bushfire responsibilities from land management to emergency services 

agencies; in short the failure of governance at all levels. 

INTRODUCTION 

This is a note to a future bushfire historian. I am writing it because I feel sure he or she will look back 

on the late 20 th and early 21st century with bewilderment. How did it happen, the question will be 

asked, that the Australians of those times developed an excellent bushfire management system..... and 

then discarded it, thus needlessly exposing themselves to death and destruction by fire. 

The answer? Read on. 

A POTTED HISTORY 

Serious bushfires in Australian forests are nothing new. Their occurrence goes back 200 years or so to 

the early 19 th century, when European settlers rejected the bushfire management approach of 

Aboriginal Australians and replaced it with an imported European approach. 

There were a few decades of transition, and there were even a few European Australians who could 

see that the Aboriginal approach made sense. But this didn’t last. The insertion of fire vulnerable 

townships, homesteads, crops, fencing and stock into bushland designed by nature to burn regularly, 

and then the failure to burn this bush regularly, did not seem at the time to be a particularly silly idea. 

Even when it produced the inevitable result: bushfire disasters. They became milestones in our rural 

history, with successive disasters named after days of the week. Black Friday, Ash Wednesday, 

Incinerated Tuesday and so on. 

The Empire fought back. Harking back to their British heritage, the new Australians remembered the 

concept of the fire brigade, and introduced it to their new land. Bushfire brigades, progressively better 

organised and resourced over the years, at first voluntary but increasingly paid, came into being. Vast 

paramilitary organisations were developed. Their efforts were underpinned by European ecophilosophy: 

bushfires were not inevitable, but were an anomaly. The default position for the 

Australian bush was one in which no fires occurred. It was the bushfire equivalent of Terra Nullius. 

Those espousing this philosophy believed in an ideal world, forests without fire. They convinced 

themselves that if a sufficiently well-trained and well-equipped force of fire-fighters could be set up, 

1 Roger Underwood is Chairman of the Bushfire Front Inc. He is a former district and regional forester and senior land 

management administrator in Western Australia and a Fellow of the Institute of Foresters. He was the recipient of the 

Institute's NW Jolly Medal in 2008. Email: yorkgum@westnet.com.au



the ideal could be achieved. All fires would be pounced upon the moment they started, and 

extinguished. Serious fires would become a thing of the past. The war on fire was declared. 

Early Australian foresters were firm devotees of this philosophy, indeed they provided the scientific 

underpinning, based on their knowledge of northern hemisphere temperate hardwood forests. They 

saw fire as an ecological disaster. In this view they were encouraged to observe that the US Forest 

Service had adopted the very same philosophy. Wildfire was also to be expunged from North 

American forests, providing an example to the rest of the world. 

There was only one problem with this philosophy: it didn’t work. It certainly didn’t work in the USA. 

It especially didn’t work in Australia where fire fighting resources were so limited and our eucalypt 

forests were so fire-loving. Over time, keen observers started to spot the trend: the longer fires were 

excluded from the bush, the more fire fuels accumulated. The heavier the fuels became, the more 

fierce the inevitable fire. And the fiercer the fire, the less able were fire-fighters to suppress it and the 

more damage was done. 

No-one much remembers the 1950s these days, but they were a decade of appalling forest fires all 

over the country, especially in my home state of Western Australia, where they brought to an end 

three decades of attempted fire exclusion in the jarrah forest. And just as happened in Victoria a halfcentury 

later, it all culminated in disaster. Towns were burnt, hundreds of thousands of hectares of 

forest and farmland were severely burned, and there was a Royal Commission. 

THE PENNY DROPS 

The decades of attempted fire exclusion and the subsequent fire disasters in the 1950s and 1960s 

signified a new era in Australian fire management. The penny finally dropped. Australia is not 

Europe. It is not even the USA. A new generation of forest ecologists who studied the bush rather 

than English and French text books realised that eucalypt forests actually thrive under a regime of 

frequent mild fire. They were pleasantly surprised when anthropologists pointed out that the 

Aborigines had understood this for thousands of years. For the first time, foresters in the districts were 

encouraged to start putting into practice the practical wisdom they had been accumulating during the 

bad years. They well understood that in forests which are periodically burned under mild conditions, 

fuel levels remain low, and fire suppression is easier and safer. With a prescribed burning program, 

fires could not be eliminated, but the number, and the damage caused by large high intensity forest 

fires declined significantly. What was better, the frequently burned forest was healthier and more 

beautiful and better able to sustain the whole range of forest values. 

Later, practical bush wisdom was supported by practical fire research. These put numbers to fire 

behaviour, and enabled rates of spread and intensity to be predicted, which in turn allowed the 

development of prescribed burning guides. The innovation of aerial burning revolutionised prescribed 

burning operations, as maximum use could be made of optimum conditions. 

For the first time since the Aborigines lost control of the bush, Australians had a home-grown fire 

management system that was ecologically sound and highly effective in minimising the occurrence of 

landscape-level high-intensity fires. The new philosophy drew on medical rather than military 

science. “Prevention is better than cure” became the catchcry; fuel reduction by prescribed burning 

was regarded as immunising the forest against a wildfire disease epidemic. People came from all over 

the world to see what we were doing, and were filled with admiration....even the Americans. 

THE PENNY IS PICKED UP AGAIN 

The trouble with successful bushfire management is that it is its own worst enemy. Success breeds 

apathy and apathy breeds mismanagement. Mismanagement leads to system degrade. And in the case 

of bushfire management, system degrade leads to serious bushfires. This is exactly the situation 

witnessed right across forested Australia over the last 15 years.



There is a theory to explain all this, called Chaos Theory. Systems always come apart once they lose 

energy. The admirable fire management systems developed for Australian forests in the 1960s and 

1970s, and evolving successfully through the 1970s and 1980s, suddenly had the energy sucked from 

them, and in the 1990s, they came to pieces. 

There were several reasons for this, but the fundamental problem was incompetent government. The 

1990s was the time in Australian history when urban pressure groups took over policy-making in land 

management. It was the time when academics who were still wedded to the idea that Australia was 

Europe, began to wield serious influence over political processes and bureaucrats. It was the time 

when people who had not started at the bottom took over at the top, when for the first time our land 

management agencies were being led by people with nil bushfire experience. In the blink of an eye 

the system flip-flopped; the Australian philosophy was thrown out the window and the discredited 

European philosophy and its American counterpart were reinstated. 

The result was that all over forested Australia, land managers began to turn away from prescribed 

burning and to look to the fire brigades as their bushfire panacea. The tired old idea that “all fires are 

bad” and must be immediately extinguished again took hold. The intelligentsia emerged from their 

leafy campuses to preach a message about the irreparable ecological damage caused by periodic mild 

fire. Prescribed burning, it was said, was leading to fauna extinctions, to dieback, to soil infertility, to 

weed infestations, to killer smoke hazes, even (horror of horrors) to tainted wine. It was the late 19 th 

and early 20 th century all over again, the only difference being that now we had jazzier fire fighting 

equipment and our fire chiefs had sexier uniforms. 

But, lo and behold! The fire brigade approach still did not work! Not only did it fail to eliminate large 

intense bushfires, their number increased. Luckily there was an excuse: unstoppable bushfires were 

the result of global warming, of Mother Nature counter-attacking! 

I think not. The decline in bushfire management in Australia in the last 15 years has a much less 

palatable explanation: the incompetence of our governance, and the failure of our leadership. This is 

readily demonstrated. 

ANALYSING GOVERNANCE – A CASE STUDY 

During 2003, my colleagues and I in The Bushfire Front developed a model for Best Practice in 

Bushfire Governance for a jurisdiction like WA. The model set out ten basic requirements which a 

government could adopt either as goals during a period of system development, or as performance 

measures against which progress and achievements could later be gauged. 

The ten Best Practice requirements are: 

1. Western Australia has a State Bushfire Policy providing leadership, guidance and overarching 

direction to all government agencies, land managers and local government. 

2. There is clear accountability for bushfire management and for fire outcomes at Ministerial 

level, and within the land management and emergency services bureaucracies. 

3. There is a unified and consistent approach to prevention, preparedness, damage mitigation, 

suppression, fire recovery and community education across government and local 

government. This approach is spelled out in writing and signed off by the government, and 

assigned to a single Minister for implementation and review. 

4. Land management, bushfire and emergency services legislation is up-to-date and assigns 

accountability, responsibility, priorities, standards and an efficient hierarchy of 

administration, leadership and control to Ministers and to agencies.



5. Bushfire management on crown lands is the responsibility of a competent land management 

agency, not the emergency service. 

6. The land management agency has a published Bushfire Strategy that focuses on bushfire 

prevention and damage mitigation as well as suppression, incorporates an effective program 

of fuel reduction burning, and is supported by ongoing appropriate research. 

7. Performance outcomes, standards and targets for bushfire management on crown lands 

(directed at minimising the number and size of high intensity wildfires) have been developed 

and published. 

8. There is a public report at the end of every fire season based on independent monitoring of 

bushfire outcomes compared with performance standards and targets, and listing towns and 

settlements at the urban/bushland interface which are most at risk from bushfires in the 

coming summer due to proximity of heavy long-unburnt fuels. 

9. The government has developed and is implementing a professional communications strategy 

to educate the community (including schoolchildren) about fire science, fire preparedness, 

damage mitigation and action in the event of a fire. 

10. The government has empowered rural and semi-rural communities to practice effective 

bushfire management, and has funded the placement of professional fire management 

specialists within local government authorities. 

We found that the Western Australian government failed utterly (in fact scored zero) on points 1, 2, 3, 

4, 7, 8, 9 and 10. The most critical failures were: 

• Policy. WA has no overarching Bushfire Policy. Instead individual agencies and local 

governments each have their own policies, developed independently and usually in conflict 

with each other; 

• Accountability. There is no single Minister in the government with responsibility for fire and 

who can be held accountable for bushfire outcomes. Instead we found six Ministers with a 

finger in the bushfire pie, and accountability was so diffuse that no one Minister could ever 

be held accountable for any outcome. 

• Agency mismanagement. We found that within the Department of Environment and 

Conservation (DEC), the agency responsible for management of national parks, regional 

parks, State forests, nature reserves and Unallocated Crown Land in WA, fire management 

did not have the status of “core function”. 

Our protocol was rejected out of hand by the government. Apparently, bushfires were not a problem. 

We were told to mind our own business, that bushfire management was in good hands, and that they 

were very satisfied. Every subsequent attempt we made to get the government to focus on system 

deficiencies was thwarted by grossly over-confident and poorly informed Ministers, no doubt guided 

by their political advisers and public servants. 

The change in this attitude after the Victorian disaster in February this year was dramatic. All of a 

sudden we detected an awful realisation in the corridors of power that if a similar disaster occurred in 

WA they could not say they had not been warned. 

WHAT ARE THE PROSPECTS FOR CHANGE? 

The ghastly Victorian disaster has had one positive outcome. Almost everyone now acknowledges 

that the current approach to bushfire management in Australia is a failure. Putting all the eggs in the 

suppression basket only works for small fires under mild conditions, i.e., the sort of fires that do little 

damage anyway.



Realising that there is a problem is the first step towards solving it. I am encouraged to think that most 

sensible people have now moved past the point of denial. I detect that many politicians now accept 

that they have been seriously misled over the years – misled by their political advisers, misled by the 

environmentalists and the whining academics, even misled by their fire chiefs. If this process of 

facing up to reality survives subsequent elections, it will mean that the lives and assets lost in the 

Victorian fires will not have been completely in vain. 

However, I can see no prospect for change unless Australian governments start to govern. This will 

require four things: 

1. Governments must reject and reverse the concept of placing responsibility for bushfire 

management with the emergency services rather than the land managers. Fire management 

must become holistic and focus on prevention and damage mitigation, not just suppression. 

2. Each jurisdiction must design and face up to an appropriate Best Practice protocol, and get on 

with the job of building an effective fire management system, starting with policy and ending 

with monitoring and reporting of outcomes. 

3. There can be no more gutless hiding behind the pathetic excuse of “global warming”. If the 

temperature does go up a degree or two, then let’s be ready for it, not use it to excuse 

incompetent leadership and outdated philosophies. 

4. Significant cultural change must be induced within the senior ranks of our land management 

agencies. The people in charge of our national parks and State forests must be made to realise 

that unless they manage fire, they cannot manage for anything else, and that attempts to 

manage fire by the fire brigade approach will always fail. They must summon up the courage 

to reject the false prophets from academia. 

None of this will happen under the current system of diffuse Ministerial responsibility and 

accountability. Each state needs a Bushfire Supremo, one person clearly running the show, whose job 

depends on performance — the degree to which effective fire management systems are designed, 

implemented and energised. This person must sit outside and above both the land management 

agencies and the emergency services, and report directly to Parliament. 

So, dear historian of the future, this is what happened back in the years 1983-2009. Basically we 

stuffed things up. But not so badly that some of us did survive, and we still have an opportunity to put 

things right. We do not see ourselves as “powerless in the face of unstoppable fires caused by global 

warming” but as intelligent people who love our forests, value our community assets and recognise 

our duty of care to our fellow-humans. What’s more, we actually know already what has to be done! 

If the technology of 2109 allows it, get back to me with your review, will you? I’m keen to know how 

we go. 

September 2009



ABSTRACT 

NATIONAL FIRE BEHAVIOUR PREDICTION SYSTEM 

Miguel Cruz 1,2 , Jim Gould 1,2 

The estimation of fire behaviour is an important component of any fire management 

approach, allowing the determination of the impacts of fire on ecosystem components and 

supporting forest fire management decision-making. Fire behaviour prediction combines 

quantitative and qualitative information based on experience and scientific principles of 

describing the combustion and behaviour of fire influence by topography, weather and 

fuel. Predictions are based on mathematical models that integrate important factors in a 

consistent way. The National Fire Behaviour Prediction (NFBP) system will consist of 

four primary components (fuel models, fuel moisture models, wind models, and fire 

behaviour models) to predict fire characteristics (e.g., rate of spread, flame height, fireline 

intensity, onset of crowning spotting potential, etc). This paper will focus on the fire 

behaviour component of the NFBP system. This component integrates a suite of models 

covering the main fuel types of Australia, eucalyptus forests, exotic pine plantations, 

grasslands, shrublands and Mallee-heath. 

The desired accomplishments of the proposed National Fire Behaviour Prediction 

Systems is to provide fire managers with better operating models to implement prescribed 

burning programs, suppression resources, risk and biodiversity management programs. 

The fuel type specificity of the fire models, its greater accuracy and updated calculation 

methods allow also for more accurate simulations of the impact of hypothetical climate 

change scenarios on fire potential and risk in Australia. 

INTRODUCTION 

Recent extreme bushfire seasons have increased the focus on how rural fire and land management 

activities determine landscape level fuel dynamics, fire regimes, bushfire risk and bushland urban 

planning. There is need to improve ecosystems health and to reduce the likelihood of catastrophic 

fires. Recommendations from the Council of Australian Governments (COAG) National Inquiry on 

Bushfire Mitigation and Management (Ellis et al. 2004) that the provision of additional resources 

jointly by the Australian Government and the state and territory governments to accelerate the research 

necessary for the characterisation of fuel loads and dynamics for Australian ecosystems (both natural 

and exotic), the characterisation of fire behaviour and ecological responses, the development of 

‘burning guides’ from this information, and the compilation of this information and knowledge in 

nationally accessible databases. 

Current weather patterns, society perception of fire, and land management agencies policies and 

constrains have increased the complexity and magnitude of the wildland fire problem. Information is 

the key in sound decision making to mitigate detrimental effects of wildland fire. The ability to predict 

and comprehend fire behaviour in relation to its drivers is fundamental to a safe, effective and 

ecosystem enhancing fire management decision-making. Fire behaviour, “the manner in which fuel 

ignites, flame develops, and fire spread and exhibits other related phenomena as determined by the 

interaction of fuels, weather and topography” (Merrill and Alexander 1987) is a vital component on 

the decision making process related to prescribed fire use planning and execution, dispatching, firefighting 

safety, definition of bushfire suppression strategies and tactics, evaluating fuel treatments 

effectiveness and recurrence. 

1 Bushfire Dynamics and Applications, CSIRO Sustainable Ecosystems and CSIRO Climate Adaptation Flagship, 

GPO Box 284, Canberra, ACT 2601, Australia. 

2 Bushfire Cooperative Research Centre, East Melbourne, VIC, 3002, Australia.



A review of the pertinent fire behaviour research carried out in Australia in the past 30 years (e.g., 

Cheney et al 1992, Burrows 1994, Marsden-Smedley and Catchpole 1995, McCaw 1997, Gould et al. 

2007, Cheney and Sullivan 2008, Burrows et al. 2009) and the tools currently used by fire managers 

reveals a large disconnection between the knowledge available and the knowledge used. 

This is partially due to the lack of effective technology-transfer programs and a considerable 

technology gap regarding the format decision support tools are available to users. A lack of 

sophistication is manifest in the broad reliance on slide rules introduced by McArthur (1966, 1967) 

and tables (Sneeuwjagt and Peet 1985) to support fire management decision-making. It is paradoxical 

that although the cost of fire fighting, our understanding of fire phenomena and the complexity of the 

fire manager job increased significantly through time, the tools that are used to predict fire behaviour 

and support decision making at any temporal and spatial scale are based in technology introduced in 

the 1950’s. 

The use of slide rules, tables and other field guides have a role in producing first approximations of 

fire behaviour characteristics and intended mainly to be used as a field reference guides. Nonetheless, 

precise fire behaviour predictions, combining spatially explicit fuel and topography information with 

detailed weather forecast, are best made using computing-based representation of fire behaviour 

models. The integration of new technologies, such as, Geographical Information Systems, satellite 

information, large computing power, internet and handheld devices, to harness available datasets and 

conduct fire behaviour simulations, either at the local or landscape level, is episodic, and unfortunately 

not the norm. 

Furthermore, the overly reliance on the old methods to predict fire behaviour results in a lack of 

integration of recent research on fuel moisture (e.g., Matthews et al. 2007), fuel dynamics (Gould et al 

2007) and wind speed (Sullivan and Knight 2001), decreasing the certainty of fire predictions. 

The situation in Australia contrasts with the evolution of fire management decision support systems in 

the US and Canada. It is well accepted that in the 1960’s these three countries had the most advanced 

fire behaviour research programs in the world. At that time the technologies used by fire practitioners 

were comparable, although the development of the grassland and forest fire danger meters (McArthur 

1966, 1967) gave Australian fire managers a tool still inexistent to others in North America, the 

capacity to predict rate of spread and intensity from information on weather and fuel characteristics. 

The advent of reliable fire behaviour models in both the USA and Canada, in the mid-seventies and 

late eighties respectively, was accompanied by the development of computer software of various 

levels of complexity that allow fire managers to simulate fire propagation at a range of spatial and 

temporal scales, e.g., BEHAVE, FARSITE, PROMETHEUS, FSPro (Andrews and Finney 2007, 

Tymstra et al. 2006). These systems not only form the backbone of effective fire management 

programs in these countries but are also extensively used in research applications. 

In Australia there have been attempts to provide users with computer-based simulation systems 

(Colman and Sullivan 1996) although they seemed to not have been adopted by the fire management 

community. 

CONCEPT 

This paper presents a proposal for a National Fire Behaviour Prediction (NFBP) system aimed at 

addressing the needs of fire managers in regards to fire behaviour information, namely quantifying 

fuel hazard, assessing fire danger and predicting site specific fire behaviour. The purpose of the 

system is to integrate all available and peer-reviewed fire behaviour knowledge into a set of userfriendly 

tools that can be used by fire managers to deal with fire management issues over a range of 

spatial and temporal scales, and complexity levels. The availability of such a system will improve fuel 

management programs, lead to more effective and safe firefighting, enhance protection of rural 

communities and reduce detrimental effects to natural resources.



Figure 1. Flow diagram of the National Fire Behaviour Prediction System (NFBPS) illustrating 

the links between fire drivers, the knowledge base (models and data) and output 

systems. 

The heart of the model system is composed by two distinct components (Figure 1), the core fire 

models module and a fire behaviour knowledge base module. The core fire models module is the 

engine of the system, providing simulations of fire behaviour aimed at respond to a large array of fire 

management questions. Main physical fire behaviour quantities incorporated in the model are: 

• Sustainability of fire spread 

• Rate of surface fire spread and intensity; 

• Flame dimensions (height, depth and angle) and residence time 

• Spotting potential 

• Onset of crowning and crown fire behaviour 

• Coarse woody fuel consumption 

• Initial fire development potential (area and perimeter assuming point source ignition)



Along with the above described fire behaviour models, an assortment of sub-models will provide 

accessory information related to first order fire effects: 

• Burn patchiness; 

• Scorch height; 

• Emissions; 

Figure 2. Prediction of Cheney et al (1998) natural pasture fire-spread model as a function of 

wind speed against experimental data. Data has been normalized for the effect of fuel 

moisture. The 68 and 95% prediction intervals are shown. Figure from Cheney et al. 

(1998). 

The fire behaviour knowledge base module provides a setting where fire behaviour simulations can be 

compared with real world data. This module will incorporate all available fire behaviour data in 

Australian ecosystems and will allow users to visualize their simulations within the context of data, 

illustrating the uncertainty in the predictions and the limitations of the datasets. A grassfire rate of 

spread model (Cheney et al 1998) output as a function of wind speed (data corrected for fuel moisture 

effect) provides an example of such concept (Figure 2). The figure not only provides information on 

how rate of spread increases approximately linearly with wind speed, but also indicates the variability 

in observed fire spread. 

For the area where the bulk of the data exists (wind speeds between 15 and 25 km/h) the rates of 

spread seem to spread +/-33% of the predicted value, with a few cases showing extreme variability. 

The graph also provides the bounds of the dataset used to build models, leading to caution by users 

when their simulations are required for conditions not covered by the existent datasets. The system 

considers that users can input their own datasets (collected in prescribed burns or wildfires) to verify 

simulations and strength the predictive capacity of the system. This analysis will provide users with 

an understanding of the predictive capacity of models, something that is only subconsciously acquired 

after a long career at predicting fire behaviour.



MODELLING UNCERTAINTY 

The operational prediction of fire spread to support fire management operations relies on a 

deterministic approach where a single “best-guess” forecast is produced from the best estimate of the 

environmental conditions driving the fire. Although fire can be considered a phenomenon of low 

predictability and the estimation of input conditions for fire behaviour models are fraught with 

uncertainty, no error component is associated with these predictions. The NFBPS will incorporate 

modelling techniques that allow describing the inherent uncertainty in fire behaviour into the 

simulation process. Ensemble methods that consider the variability of model inputs and Monte Carlo 

sampling (Cruz, in press) will be integrated in the core fire modelling component. This will allow to 

describe prediction statistics and estimate the likelihood that extreme phenomena, e.g., onset of 

crowning, will occur. These outputs allow users to ascertain with some confidence what will and what 

will not happen. The probabilistic outputs from the ensemble method extend the range of questions 

that can be answered by fire behaviour models, enabling linkages between fire behaviour models and 

quantitative risk analysis. 

In situations where users provide observed fire data to the system, data assimilation methods will be 

used to improve the prediction potential of the model system. 

FUELS 

The NFBPS will cater for a large variety of fuel types existent in Australia, incorporating research 

carried out since the 1980s. 

• Grasslands 

o Natural grasslands (Cheney et al 1998) 

o Grazed grasslands (Cheney et al 1998) 

o Eaten out grasslands (Cheney et al 1998) 

o Grassland/woodland (Cheney and Andrews 2008) 

o Spinifex grasslands (Burrows et al. 2009) 

• Shrubland 

o Coastal heaths (Catchpole et al, in preparation) 

o Buttongrass Moorlands (Marsden-Smedley and Catchpole 1995) 

o Mallee-heaths (McCaw 1997, Cruz et al, in preparation) 

o Mallee-spinifex 

• Forest 

o Exotic pine plantations (Sneeuwjagt and Peet 1985, Cruz et al 2008a); 

o Dry sclerophyll eucalypt forest (Sneeuwjagt and Peet 1985, Burrows 1994, Cheney et 

al. 1992, Ellis 2000, Gould et al 2007); 

• Logging slash 

OUTPUT COMPONENTS 

The NFBPS is comprised of several output modules that aim respond to the requirements of distinct 

applications of fire behaviour information (Figure 1). The four module systems are: 

• Fuel Hazard Assessment System 

• Fire Danger Rating System 

• Fire Behaviour Prediction System 

• Bushfire Forecasting System 

Fuel Hazard Assessment System (FHA) 

Fuel hazard assessment aims to describe qualitatively the contribution of fuel complex structure to 

potential fire behaviour (McCarthy et al 1998, Gould et al. 2007). It allows integrating information 

from the various fuel strata into an overall hazard. Within the NFBPS fuel hazard assessment is linked



to the quantitative component of fire behaviour modelling and is used to support landscape level fuel 

management, namely evaluating the effectiveness of fuel modification treatments, prescribed burn 

planning and evaluate WUI (Wildland Urban Interface) specific fuel hazards and propose mitigation 

actions. 

Fire Danger Rating System (FDR) 

Fire Danger is a process systematically evaluating and integrating the individual and combined factors 

influencing fire potential (e.g., ignition, rate of spread, difficulty of control, fire impact) at a large 

spatial scale (Merrill and Alexander 1987). Fire danger is calculated to optimize pre-suppression 

planning, resource allocation, and initial attack fire fighting tactics. It is also a vital component of 

public awareness to potential fire hazard drives public policy related to certain fire issues. In NFBPS 

fire danger is intimately linked to fire behaviour. By integrating spatial information this adjusts fire 

potential to the specificity of a region vegetation cover and topography. This approach makes use of 

available datasets to provide detailed information on the local conditions, necessary for some 

applications, while still allowing the scale up of the results to regional of state wide fire danger 

analysis. 

Fire Behaviour Prediction System (FBP) 

The fire behaviour prediction system aims to provide the site specific fire behaviour information that 

is necessary to plan and conduct prescribed burns, support wildfire suppression strategies and tactics, 

advise firefighters of safety concerns (e.g., red flag warnings) and gauging fuel management 

effectiveness. The temporal and spatial scope of the simulations depend on the information available 

(point data vs. GIS database) and the intended use of the simulation output. 

Bushfire Forecasting System (BF) 

The Bushfire Forecasting System is a module that aims to extend the temporal and spatial scale of the 

FBP system to look at long range fire potential (5 to 20 days). This system will combine spatial 

information on vegetation and topography with long-term weather forecasts and climate data and 

produce probabilistic outputs of fire impact zones and impact severity. 

FORMATS 

The system will be available in different formats. The core system is software that has the capacity to 

be used at multiple spatial and temporal scales. The scale used for a particular simulation will depend 

solely upon the availability of the necessary inputs and the scope of the simulation. A user can carry 

out simple simulations with only a few basic input parameters (weather, fuel type, slope), extend such 

predictions for a 12- or 24-h period if forecasted weather is available, or simulate fire propagation 

across the landscape over complex topography under variable fuels and evolving weather conditions. 

This later simulation will require spatially explicit information of fuels, topography and weather. 

Aside from stand alone computers running the software, it is envisioned that the system should also be 

able to run in a centralized server (localized in a Fire Behaviour Service Center) with the results being 

able to be visualised in mobile devises (laptops, smart phones, etc) through the internet. This will 

allow processor intensive simulations to be carried out quickly and the results could be available to a 

suit of users that would not have the time or training to run the simulations. 

The system will be also available in different formats suitable for quick verification in field conditions 

namely slide rules, tables and nomograms. In these formats simplified version of the core models will 

be used to produce first approximations of fire behaviour characteristics and intended mainly to be 

used as field reference guides. 

CONCLUDING REMARKS 

Fire is a multifaceted phenomena, and knowledge of its behaviour is necessary to comprehend its 

likely impact on human live, property and natural resources, develop proactive fire management plans 

and establish effective fire suppression strategies and tactics. The increase in complexity of the land 

and fire manager job in face of new climate scenarios, weather extremes and conflicting land 

management objectives has not been accompanied by the development of decision support tools that 

allow fire prediction using the most up to date fire behaviour science and technologies.



We propose a decision support system to predict fire behaviour and impacts based on the integration 

of the latest fire behaviour and peer-reviewed science within state of the art IT tools. The success of 

fire behaviour prediction tools in supporting fire management decision-making in North America 

resulted in part from the extensive training received by users to understand how to use the model 

systems, and most importantly, the limitations associated with then and how they should be taken into 

consideration. There are still large areas of uncertainty in fire behaviour prediction and it is of utmost 

importance that users understand those. Failure to do so can lead to erroneous interpretations of output 

results, which at best can lead to a decrease in the trust users put on the models, and at worst can put 

lives and human property at risk. The development of the NFBPS needs to be accompanied by a well 

supported technology transfer program that ensures users are aware of system limitations and 

application bounds. 

REFERENCES 

Andrews, P. and Finney, F. (2007). Predicting wildfires. Scientific American, August issue. 47-55. 

Burrows, N.D. (1994). Experimental development of a fire management model for jarrah (Eucalyptus marginata 

Donn ex Sm.) forest. Aust. Natl. Univ., Canberra, Australian Capital Territory. Ph.D. thesis. 293 p. 

Burrows, N.D., Ward, B., Robinson, A. (2009). Fuel dynamics and fire spread in spinifex grasslands of the 

western desert. Pages 69-76 in (Tran C. Ed.) The Proceedings of the Royal Society of Queensland – 

Bushfire 2006 conference. Brisbane 6-9 June 2006. 

Cheney, N.P., Gould, J.S., Catchpole, W.R. (1998). Prediction of fire spread in grasslands. International Journal 

of Wildland Fire 8(1) 1-13. 

Cheney, P., Sullivan, A. L. (2008). Grassfires: fuel, weather and fire behaviour. 2 nd Ed, CSIRO Publishing, 

Melbourne, Australia. 150p. 

Cheney, N.P., Gould, J.S., Knight, I. (1992). A prescribed burning guide for young regrowth forests in silvertop 

ash. Technical report. Forestry Commission of New South Wales., Sydney. 

Coleman, J., Sullivan, A. L. (1996). A real-time computer application for the prediction of fire spread across 

Australian landscape. Simulation 67,230-240. 

Cruz, M.G., Alexander, M.E., Fernandes, P.A.M. (2008a). Development of a model system to predict wildfire 

behaviour in pine plantations. Australian Forestry 71, 113-121. 

Cruz, M.G. xxxx. Application of a Monte Carlo based ensemble method to predict the rate of spread of grassland 

fires. Int. J. Wildland Fire, in press. 

Ellis, S, Kanowski, P, Whelan, R. (2004). National Inquiry on Bushfire Mitigation and Management. 

Commonwealth of Australia, Canberra. 

Ellis, P.F. (2000). The aerodynamic and combustion characteristics of eucalyptus bark – a firebrand study. Ph.D. 

Thesis, Australian National University, Canberra, Australia. 187 p. 

Gould, J.S., McCaw, W.L., Cheney, N.P., Ellis, P.F., Knight, I.K., Sullivan, A.L. (2007). Project Vesta : fire in 

dry eucalypt forest : fuel structure, fuel dynamics and fire behaviour. (Ensis-CSIRO, Canberra ACT, and 

Department of Environment and Conservation, Perth WA) 

Matthews S., McCaw, W.L., Neal J.E., Smith R.H. (2007) Testing a process-based fine fuel moisture models in 

tow forest types. Can J. For. Res. 37:23-35. 

Marsden-Smedley, J.B., Catchpole, W.R. (1995). Fire behaviour modelling in Tasmanian buttongrass moorlands 

II. Fire behaviour. Int. J. Wildland Fire 5:215-228. 

McArthur, A.G. (1966). Weather and grassland fire behaviour. Forestry and Timber Bureau of Australia Leaflet 

1000. 

McArthur, A.G. (1967). Fire behaviour in dry eucalypt forest. Forestry and Timber Bureau of Australia Leaflet 

107. 

McCaw, W.L. (1997). Predicting fire spread in Western Australian mallee heath shrubland. PhD thesis. 

Canberra, Australia: University of New South Wales, University College, School of Mathematics and 

Statistics. 235 p. 

Merrill, D.F., Alexander, M.E. (1987). Glossary of fire management terms. Canadian Committee on Forest Fire 

Management, NRCC Nº26516, Ottawa, Canada. 91 p. 

Sneeuwjagt, R.J., Peet, G.B. (1985). Forest fire behaviour tables for Western Australia. Department of 

Conservation and Land Management, Perth, Western Australia 59p. 

Sullivan, A.L., Knight, I.K. (2001). Estimating error in wind speed measurements for experimental fires. Can. J. 

For. Res. 31:401-409. 

Tymstra, C,, Bryce, R.W., Wotton, B.M., Armitage, B. (2006). Prometheus – the Canadian Wildland Fire 

Growth Model: Model Development and Validation. Inf. Rep. In preparation. Alberta Sustainable 

Resource Development/Nat. Resour. Can., Can. For. Serv., North. For. Cent.



WOODY FUEL CONSUMPTION AND CARBON 

IN THE CHANGING CLIMATE OF AUSTRALIA 

Jennifer Hollis 1,3 and Lachlan McCaw 2,3 

ABSTRACT 

Woody fuel consumption was assessed at prescribed burns in four different forest types 

across southern Australian eucalypt forests using variations of the line intersect 

technique. Research into the effect of climate change in fire potential in Australian 

ecosystems indicated that future climate scenarios will result in an increase in the severity 

of fire seasons and fire intensity. The study aimed to quantify the link between fireline 

intensity and the fraction of coarse woody fuel consumed and carbon released in 

bushfires. Results from 39 prescribed burns ranging in intensity between 53 and 5000 

kW/m were analysed for the relationship between fireline intensity and proportion of 

woody fuel consumed. Without the ability to isolate the effect of fireline intensity from 

other variables such as fuel moisture, it was difficult to assess its direct relationship with 

woody fuel consumption. While fireline intensity appears to effect the consumption of 

fine fuels and to a lesser extent the small woody fuels (



New South Wales. This included detailed assessment of woody fuel moisture, density and wood 

decay. The authors found a strong relationship between woody fuel consumption and fire intensity and 

noted that the greater the degree of decay, the greater the proportion of consumption. This suggests 

that CWD age may also be an important variable affecting fuel consumption. 

From 184 studies available in literature (Mackensen et al. 2003) found that in 57% of all cases, the 

calculated lifetime of CWD (time to when 95% of mass is lost) is longer than 40 years (the median of 

the distribution was 49 years and the mean 92 years). In fact, coarse woody debris in some species 

such as jarrah (Eucalyptus marginata) were found to have a lifetimes of up to 120 years. 

CWD is significantly affected by fire disturbances making accurate accounting for its’ contribution to 

the carbon stock of a forest ecosystem particularly complex. In Australian forests where CWD 

contributes approximately 18% of the total forest above-ground biomass and carbon stock in 

Australian dry sclerophyll forests and 16% in wet sclerophyll forests (Woldendorp et al. 2002), fire 

can significantly modify CWD volume resulting in changes to greenhouse gas emissions and carbon 

stocks. This will vary from forest to forest and due to differences in the conditions under which they 

are burnt. One of the studies by (Hingston et al. 1980) in the jarrah forest (Eucalyptus marginata) of 

southwest Western Australia, found coarse woody debris proportion of above ground biomass was 

32%, significantly higher than the average found in the culmination of studies by (Woldendorp et al. 

2002). 

In the event of a fire, prescribed or otherwise, consumed CWD will directly add to carbon emissions 

(Apps et al. 2006), however, the event may not necessarily result in the equivalent reduction in the 

coarse woody debris fuel load. As a result of the fire, newly fallen trees and branches will be 

transferred from overstorey and understory trees to a new coarse woody debris load (Waterworth and 

Richards 2008) which could even be as high as the pre-fire load depending on variables such as 

species, stand structure, fire behaviour (e.g., fire intensity, residence time) and termite damage. 

As the mass of carbon in coarse woody debris is a function of carbon concentration (%) and density, 

uncertainties exist in determining the contribution of coarse woody debris to carbon stocks. This can 

be attributed to the scarcity of information on coarse woody debris volume and decay rates (Brown et 

al 1996), and the impact of decay on wood density in Australian forests (Grierson et al. 1992; 

Mackensen et al. 2003). As a result, reliable calculations of biomass, carbon stock values and CO2 

emissions are currently fraught with uncertainty until further data is collected through long-term 

studies of coarse woody debris in different environments (Mackensen et al. 2003). Current estimates 

of carbon in coarse woody debris mass range from 45 to 50 percent (Woodwell et al. 1978); Tilman et 

al. 2000; Mackensen and Bauhus 1999). For the purposes of this report, the conversion factor 

currently used by the Australian Greenhouse Office (Mackensen and Bauhus 1999) will be used where 

the average carbon content is 50% of the coarse woody debris biomass. 

Woody fuels and climate change 

Fires have long been part of the Australian bush and have played a major role in shaping vegetation 

structure and species composition (Bradstock et al. 2002; Burrows 2008; Gill et al. 1981). Australian 

fire behaviour and regimes are closely related to meteorological conditions including temperature, 

humidity, temporal and seasonal rainfall patterns (and their affect on fuel moisture content) and wind 

speed (Catchpole, 2002; Cheney 1981; Luke and McArthur 1977; McAurthur 1973; McCaw et al. 

2003) as well as opportunities for ignition from lightning or man-made sources. Climate change 

projections for Australia indicate that each of these variables are likely to be impacted by climate 

change including a rise in annual mean temperatures as well as local variations in rainfall patterns that 

include precipitation regimes that have longer dry spells broken by heavier rainfall events (Lucas et al. 

2007). 

Climate change has the potential to alter a number of components of the fire regimes including fire 

frequency, size, intensity and length of the fire season including the period suitable for prescribed 

burning (Lucas et al. 2007). These changes have the potential to result in an increase in the area burnt 

by wildfires (Amiro et al. 2001; Gould and Cheney 2007) and as a result, increase instantaneous 

releases of greenhouse gases linked to biomass (including coarse woody debris). Understanding the 

effects that these changes could have on the amount and quality of coarse woody debris and improving



our ability to predict fuel consumption and the resulting implications for carbon stocks and cycles 

within a forest ecosystem is becoming increasingly important. 

In this paper we assess the particular impact of fire intensity on woody fuel consumption and carbon 

release at prescribed burns in 4 different forest types across Australia. This includes: 

1. jarrah (Eucalyptus marginata) forest in the south-west of Western Australia 

2. blue gum (Eucalyptus globulus)/ manna gum (Eucalyptus viminalis) forest in northern-central 

Victoria 

3. mountain gum (Eucalyptus dalrympleana) / narrow-leaf peppermint (Eucalyptis radiata) 

forest in south-eastern New South Wales 

4. stringybark (Eucalyptus obliqua) forest Tasmania 

We test the hypothesis that fireline intensity is a significant variable determining woody fuel 

comsumption. This hypothesis will be tested over the spectrum of woody fuel size classes. 

METHODOLOGY 

Woody Fuel Consumption Project (WFCP): Behind the Flaming Zone 

Woody fuel consumption was assessed during three prescribed burns in southwest Western Australia 

including Wilga, Quilben and Hester blocks (Figure 1) in 2007 and 2008. Woody fuel consumption 

was also assessed in the Tallarook State Forest located in the Victorian northern central district (Figure 

1). Multiple plots were burnt under the same weather, season and fuel conditions at the Hester site in 

order to determine the specific effect of varying fireline intensity. This was achieved by varying fire 

direction in each plot, i.e. head-fire, backing fire and flank fires. 

Woody fuel loads (>0.6cm) were determined pre and post fire using Van Wagner’s line intersect 

method (Van Wagner 1968) with four fixed 100m transects (400m total transect length) placed within 

the burning plots. Five size classes were adopted including; Size 1: 0.6-2.5cm, Size 2: 2.5-7.5cm, Size 

3: 7.5-22.5cm, Size 4: 22.5-50.0cm, Size 5: >50.0cm. Fuels crossing each transect were counted and 

measured for diameter prior to the fire and wires were tied around the circumference of the fuels 

greater than 2.5cm to determine change in volume during post-fire assessment. Each fuel item was 

scored on a scale of 1-5 for decay, suspension, arrangement, charing and bark hazard. When possible, 

the species was also recorded. Pre and post-fire woody fuel loads were calculated using Brown’s 

Woody Material formula (Brown 1974). 

Rates of spread were calculated using thermologgers and the time intervals at which the fire passed 

known grid points at 320°C within the plots. Profile, near surface and surface fine fuel moisture 

(



Figure 1. Location of woody fuel consumption research sites across Australia including Wilga, 

Quilben, Hester and McCorkhill (Project Aquarius) in Western Australia, Tallarook 

in Victoria, Tumbarumba in New South Wales and the Warra LTER site in Tasmania 

Project Aquarius 

Woody fuel consumption was assessed at 32 experimental fires ignited under dry summer conditions 

at McCorkhill block in the southwest of Western Australia in 1983 (Figure 1). 

Woody fuel loads (>1.0cm) were determined pre and post fire using Van Wagner’s line intersect 

method (Van Wagner 1968) with 20m transects placed at regular intervals along established grid lines 

placed 100m apart within each fire plot. The total transect length for each plot varied according to plot 

size. Fuels in eight size classes were counted including; 10-25mm, 25-50mm, 50-75mm, 75-100mm, 

100-150mm, 150-200mm, 200-300mm and >300mm where the fuels horizontal diameter was also 

measured. These were subsequently converted to those used in the Woody Fuel Consumption Project 

to enable comparison across projects. Pre and post-fire woody fuels loads were calculated using 

Brown’s Downed Woody Material formula (see above Equation 1 and 2 where the fuel diameter is 

known). 

Patterns of fire development within the plots was measured from periodic mapping with an infra red 

line (IR) scanner. The digital data from the scanner was corrected for scale, scene compression, jitter 

by aircraft movements and geometrically rectified onto a known planar map to enable fire spread rates 

to be calculated.



Profile and surface fine fuel moisture (7.0cm. Brown’s equation (1974) was adopted to calculate fuel load which incorporates an angle 

correction factor as well as a slope correction factor from Brown and Roussopoulus (1974). 

The quadratic mean diameter (QMD) for size classes 2.5-5.0cm and 5.0-7.0cm was determined during 

field sampling by recording diameters within each size class and using Van Wagner’s equation to 

calculated the QMD (Van Wagner 1982). 

Species densities were calculated using the water displacement method (TAPPI, 1994) and were 

means of at least 20 branches in each diameter class for each species. 

The ignition method used, centre fire ignition (convection), did not allow detailed measurements of 

fire behaviour properties. Flame heights were estimated to be 4 m. In the absence of fire behaviour 

information fireline intensity was estimated based on the Gould et al (2007) relationship between 

mean flame height and mean fire intensity. 

Tumbarumba 

Woody fuel consumption research was undertaken in February 2004 within the Maragle State Forest, 

Tumbarumba, located in south-eastern New South Wales. 

Detailed background, methodology and fire behaviour from this research was given by Tollhurst et al. 

(2006). Pre and post fire woody load (>2.5cm diameter) was determined using the Van Wagner (1968) 

line intersect method with 3 x 30m transects in each of the plots (i.e. 90m transect lengths). The 

diameter of fuels intersecting the transect was recorded along with a rating for decay and suspension 

class. Each fuel circumference was tied with 2mm wire to determine volume consumed. Four size 

classes were adopted including 2.6-7.5cm, 7.6-22.5cm, 22.6-50cm and >50cm diameter. 

Woody fuel moisture was assessed across experimental blocks. Three measures of fuel moisture were 

obtained for each wood sample including two by oven determination; from ‘inner’ and ‘outer’ 

locations of the sample, and one by electronic moisture meter (T-H Fine Fuel Moisture Meter (Chatto 

and Tolhurst 1997)) from the saw dust generated during the cutting of the sample . 

Woody density was measured in a range of diameter and decay classes by determining the volume of 

wood samples from the weight increase after submersion in water. 

Plot ignition was from a 100m continuous line and on the up-wind, down-slope position of the block 

with a final burn area approximately 4ha. Fire intensity was calculated using Byram’s (1959) fireline 

intensity equation as described previously.



Table 1. Site and Burn Characteristics 

Burn ID Av. Annual 

Rainfall 

(mm) 

Dominant Species Burn Type Ignition Technique 

WCFP - Wilga 830 Eucalyptus marginata Silvicultural Long Line 

WCFP - Quilben 1012 Eucalyptus marginata Ecological / Fuel Reduction Long Line 

WCFP - Hester 830 Eucalyptus marginata Ecological / Fuel Reduction Long Line 

WCFP - 

595 Eucalyptus globulus Ecological / Fuel Reduction Long Line 

Tallarook 

Eucalyptus viminalis 

Project Aquarius 1140 Eucalyptus marginata Ecological / Fuel Reduction Long Line/ Multiple 

- McCorkhill 

Ignition Point 

Warra LTER 883 Eucalyptus obliqua Silvicultural Central Ignition 

Tumbarumba 975 Eucalyptus dalrympleana 

Eucalyptus radiate 

Ecological / Fuel Reduction Long Line 

RESULTS AND DISCUSSION 

Pre-fire fuel load distribution 

The pre-fire fuel load distribution by size class is similar across sites with the exception of the Warra 

LTER sites which appear to have significantly higher fuel loads in each of the size classes (Figure 2). 

The Quilben site in Western Australia also has a higher than average fuel load for fuels greater than 

50cm (size class 5). At each site, the larger size classes, particularly sizes 4 and 5, form much of the 

overall woody fuel load (on average 35 and 30% respectively). 

This highlights the importance of the larger fuels to overall woody fuel consumption. This is an 

important characteristic of the problem under analysis, namely when we attempt to investigate and 

understand the effect of different environmental variables and fire behaviour characteristics, such as 

fireline intensity, on fuel consumption. 

Fuel Load (t/ha) 

250 

200 

150 

100 

50 

0 

Wilga 

Quilben 

Hester 

Tallarook 

Aquarius 

Warra LTER 

Tumbarumba 

Pre-fire Fuel Load Distrubution by Size Class 

Fine Fuel (0-0.6cm) Size 1 (0.6-2.5cm) Size 2 (2.5-7.5cm) Size 3 (7.5-22.5cm) Size 4 (22.5-50cm) Size 5 (>50cm) 

Size Class 

Figure 2. Pre-fire fuel load distribution by size class. 

Fire behaviour and fuel condition 

A wide range of weather and seasonal influences are represented in the dataset including burns 

conducted under typical spring and autumn prescribed burning conditions as well as those 

characteristic of dry summer wildfires (see Table 2 where burning conditions have been averaged for 

each site). The broad range of burning conditions is also reflected by fireline intensity which ranged 

from 53 to 5000 kW/m and in the fine (profile) and woody fuel (log average) moisture contents, 

ranging from 8-72% and 33-56% respectively (Table 3).



Table 2. Summary of average fire behaviour characteristics across burn sites. Range across 

multiple burns (minimum to maximum) in italics. 

Site/Mean 

Characteristics 

No. 

of 

fires 

RH % 

Temp 

( o C) 

10m Open 

Wind 

Speed 

(kph) 

KBDI SDI 

ROS 

(m/hr) 

Residence 

Time 

(s) 

Fireline 

Intensity 

(kW/m) 

WFCP - 

Wilga 

1 27.5 24.9 7.8 42.9 97.7 93.7 299.0 

WFCP - 

Quilben 

1 68.5 21.1 5.5 84.9 52.2 26.9 209.9 

WFCP - 

Hester 

4 56.4 

(48-63) 

24.8 

(23-27) 

13.7 

(11-17.5) 

147.5 

(148- 

148) 

105.0 

(15-217) 

21.6 

(10-40) 

348.7 

(53-678) 

WFCP - 

Tallarook 

2 50.1 

(34-66) 

16.5 

(13-20) 

8.6 

(8.2-8.9) 

36.8 

(14-60) 

139.5 

(136- 

143) 

52.3 

(20-85) 

77.5 

(28-127) 

234.0 

(76-393) 

Project 

18 45.7 24.5 5.2 139.1 372.8 

not 

measured 

1680.5 

Aquarius 

(20-61) (18-33) (2.5-24) 

(129- 

163) 

(153- 

774) 

(585-3304) 

Warra 

LTER 

11 67.4 

(52-90) 

18.1 

(17-19) 

not 

measured 

51.0 

(51-51) 

not 

measure 

d 

not 

measured 

5000.0 

(5000- 

5000) 

Tumba- 

2 32.5 27.0 7.3 122.0 369.9 

not 

measured 

2430.6 

rumba 

(20-45) (26-28) (6.5-8) 

(122- 

122) 

(122- 

618) 

(955-3906) 

Table 3. Summary of fuel moisture conditions and fuel consumption outcomes across burn 

sites. Range across multiple burns (minimum to maximum) in italics. 

Site/ 

Mean 

Characteristics 

Profile 

Fuel 

Moisture 

Content 

(%) 

Log 

Moisture 

Content 

>0.6cm 

(%) 

Total Prefire 

Fine 

Fuel Load 

0.6cm 

(t/ha) 

Woody 

Fuel 

Consum 

ption 

(%) 

Carbon 

Release 

(t/ha) 

WFCP - 

Wilga 

11.6 38.7 5.9 0.2 42.3 22.1 47.6 10.1 

WFCP - 

Quilben 

24.6 37.3 7.8 3.0 175.0 121.7 30.5 26.7 

WFCP - 16.6 32.9 6.6 0.6 93.8 47.3 49.4 23.3 

Hester (16.6-16.6) (33-33) (6.0-7.0) (0.1-1.3) (76-106) (40-62) (42-57) (17-30) 

WFCP - 51.9 45.0 8.1 1.4 52.2 31.6 39.7 10.3 

Tallarook (32.3-71.5) (35-56) (7.3-8.9) (0.3-2.6) (49-55) (28-36) (36-43) (10-11) 

Project 

10.4 not measured 8.8 0.0 60.5 27.1 54.9 16.7 

Aquarius (8.3-13.2) (6.3-12.0) (0-0) (33-107) (5-54) (33-90) (8-36) 

Warra not measured not measured 44.1 9.0 599.5 336.8 46.4 131.4 

LTER (30-53) (6.7-10.6) (226-1322) (71-795) (9-69) (31-263) 

Tumba- 

11.0 47.2 12.8 0.0 86.3 52.3 44.2 17.0 

rumba (11.0-11.0) (47-47) (12-13) (0-0) (49-123) (22-83) (33-56) (14-20) 

Woody fuel consumption and fireline intensity 

At a plot/fire specific level woody fuel (>0.6cm) consumption ranged from 9 to 90%. However the 

average for each site ranged from 31 – 55% (Table 3). Using the amount of fuel consumed at each plot 

this equates to a range of carbon release between 8 and 263t/ha, for an average of 50t/ha. At a plot/fire 

level woody fuel consumption appears to increase with fireline intensity up to approximately 700 

kW/m (Figure 3) after which no definable relationship appears to exist. This is illustrated in the weak



regression relationship for site fuel consumption with fireline intensity (sizes 1-5 combined) (R 2 = 

0.01) in Table 4 below. 

By following the diameter reduction relationships of individual fuel items at each of the Woody Fuel 

Consumption Project (WFCP) burns, it has been possible to assess woody fuel consumption by size 

class and intensity. Figure 4 illustrates the scatter of the consumption (by intensity) across the WFCP 

sites. There appears to be a weak relationship between fireline intensity and wood fuel consumption of 

the fine (R 2 = 0.48) and size class 5 (R 2 = 0.41). The regression explanatory power for the other classes 

was weaker, with an R 2 of 0.23 obtained for size class 1, and R 2 below 0.1 for sizes 2, 3 and 4. 

This supports the hypothesis that fireline intensity is an influencing variable in the consumption of fine 

fuels and possibly the small woody fuels (i.e. size class 1 (0.6-2.5cm). It also appears that fireline 

intensity may influence the consumption of fuels greater than 50cm (size 5). This may suggest that the 

consumption of larger proportions of the fine and small woody associated with higher fireline 

intensities, is required to ignite and consume the fuels greater than 50cm. 

Woody Fuel (>0.6cm) Consumption (%) 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

Woody Fuel Consumption & Fireline Intensity 

0 

0 1000 2000 3000 4000 5000 

Fireline Intensity (kW/m) 

Figure 3. Scatterplot of consumption and fireline intensity across all prescribed burns. 

Woody Fuel (>0.6cm) Consumption (%) 

100 

80 

60 

40 

20 

Fine Fuel Consumption & Fireline Intensity 

0 

0 100 200 300 400 500 600 700 

Fireline Intensity (kW/m) 

Fine Fuel (50cm) 

Fine Fuel 

Size 1 

Size 2 

Size 3 

Size 4 

Size 5 

Project Aquarius 

T umbarumba 

WFCP WA 

WFCP Tallarook 

Warra LTER 

Figure 4. Scatterplot and regression (y=a ln(x)) analysis of consumption and fireline intensity by 

size class at the Woody Fuel Consumption Project (WFCP) sites.



Table 4. Prediction equations for woody fuel consumption by size class using fireline intensity 

at the Woody Fuel Consumption Project (WFCP) sites. 

Fuel size class Regression Equation (linear) R² 

Fine fuel (50cm) 13.6 + 0.0731 Fireline Intensity 0.41 

Site (class 1-5 combined) * 

* Includes data across entire dataset 

51.7 - 0.00081 Fireline Intensity 0.01 

It is noted that the variety of conditions under which each plot and site have been burnt will also 

contribute significantly to the varied consumption outcomes. Some of the variables are also likely to 

be highly correlated, for example woody fuel moisture – log decay, and fireline intensity - Soil 

Dryness Index. Given this, the best example of the effect of fireline intensity on woody fuel 

consumption was found at the Hester site burns which were burnt concurrently (same day, same time). 

Through this method we were able to burn the various plots under the same fuel moisture conditions, 

and variable fireline intensity. This was achieved by burning distinct areas of the plot by head, back 

and flank fires. For this prescribed burn woody fuel consumption ranged from 42.2 – 56.9% with the 

highest consumption occurring within the plot with the highest fireline intensity (678 kW/m)(Table 5). 

Table 5. Woody fuel consumption by size class at the Hester site burns. 

Fireline 

Intensity 

(kW/m) 

Fine 

Fuels 

Fuel Consumption by Size Class (%) 

1 2 3 4 5 

Woody Fuel 

>0.6cm 

(sizes 1-5 

combined) 

Hester 1 678 99.0 76.7 42.6 52.4 62.5 56.5 56.9 

Hester 2 380 94.5 57.8 48.3 33.5 46.4 67.3 55.2 

Hester 3 284 92.4 55.8 48.0 50.8 35.0 47.4 42.2 

Hester 4 53 81.3 68.2 45.0 46.2 56.5 16.7 43.5 

Within each of the size classes, fine fuel consumption increased with increasing intensity however the 

same relationship was not clear in any of the woody fuel size classes greater than 0.6cm. It appears 

that the low intensity (53 kW/m) of the Hester 4 burn may have had an effect on the consumption of 

the woody fuels greater than 50cm (size 5) which was minimal (16.7%). This could be the result of the 

small energy quantity being released by the surface fire not meeting the energy requirements to ignite 

the large fuels. 

Climate change implications associated with fireline intensity 

Climate change has the potential to affect fire regimes by modifying fire intensity. While the 

relationship between fine fuel consumption and fireline intensity appears to support theories that fire 

intensity influences the burn patchiness and proportion of fine fuels consumed (e.g., Moreno & 

Oechel, 1989), the effect on woody fuel consumption and carbon release is unclear. This could be 

because there are so other variables, some of them correlated, that affect woody fuel consumption, 

making it difficult to isolate the relationship between fireline intensity and woody fuel consumption. 

It may also be that the relationship between woody fuel consumption and fireline intensity is very 

weak, possibly less important than other variables such seasonal dryness and associated large fuel 

moisture content. It is reasonable to expect that under the future altered fire climate, with longer and 

more severe (drier) fire seasons, the consumption of woody fuels will increase as a greater proportion 

of areas are burned under drier, more intense (high fire danger indices) wildfire conditions. This is an 

important aspect to take into consideration when planning and conducting prescribed burns. 

It is possible that fireline intensity plays a role in influencing the consumption of large size 5 fuels 

(>50cm), and further research is required to look at this relationship as they form a large proportion of



the CWD fuel load (on average 30%). Their ignition and consumption are important processes 

responsible for the release of large quantities of stored carbon. 

CONCLUSIONS 

The relationship between fireline intensity and the consumption of woody fuels (>0.6cm) and 

associated carbon release in southern Australian eucalypt forests is unclear making it difficult to assess 

the affect of potential climate change scenarios if only fireline intensity is considered.. While fireline 

intensity appears to effect the consumption of fine fuels and to a lesser extent the small woody fuels 

(



Mackensen J, Bauhus J, Webber E (2003) Decomposition rates of coarse woody debris-A review with particular 

emphasis on Australian tree species. Australian Journal of Botany 51, 27-37. 

Marsden-Smedley JB, Slijepcevic A (2001) Fuel characteristics and low intensity buring in Eucalyptus obliqua 

wet forest at the Warra LTER site. Tasforests 13, 261-280. 

Moreno, J.M, Oechel, W.E. 1989. A simple method for estimating fire intensity after a burn in California 

chaparral. Acta OEcologica 10(1):57-68. 

Nelson RM 2003 Reaction times and burning rates for wind tunnel headfires*. International Journal of Wildland 

Fire 12, 195-211 

Ottmar RD (1987) Prescribed Fire and Fuel Consumption in Uncured Slash--Preliminary Results. In 

'Proceedings of the Ninth Conference on Fire and Forest Meteorology'. San Diego, CA 

Ottmar RD, Burns MF, Hall JN, Hanson AD (1993) 'Consume Users Guide.' USDA Forest Service Pacific 

Northwest Research Station. 

Ottmar RD, Prichard SJ, Nihanek RE, Sandberg DV, Bluhn A (2006) 'Modification and Validation of Fuel 

Consumption Models for Shrub and Forested lands in Southwest, Pacific Northwest, Rockies, Midwest, 

Southeast and Alaska.' Fire and Environmental Applications Team, Pacific Northwest Research Station, 

Pacific Wildland Fire Sciences Laboratory, Seattle, Washington. 

Sandberg DV, Ottmar RD (1983) Slash burning and fuel consumption in the Douglas-fire subregion. In 'Seventh 

Conference on Fire and Forest Meteorology Symposium Proceedings'. Fort Collins, CO pp. 90-93. 

(American Meteorological Society) 

Slijepcevic A (2001) Loss of carbon during controlled regeneration burns in Eucalyptus obliqua forest. 

Tasforests 13. 

Sullivan AL, Knight IK, Cheney NP (2002) Predicting the radiant heat flux from burning logs in a forest 

following a fire. Australian. Forestry 65, 59–67. 

TAPPI 1994. Basic density and moisture content of pulpwood. Test Method T268 om-94. TAPPI, Atalanta, 

Georgia, USA. 5p. 

Tilman D, Reich P, Phillips H, Menton M, Patel A, Vos E, Peterson D, Knops J (2000) Fire Suppression and 

Ecosystem Carbon Storage. Ecology 81. 

Tollhurst KG, Anderson WR, Gould J (2006) Woody Fuel Consumption experiments in an undisturbed forest. In 

'Conference Proceedings of the V International Conference on Forest Fire Research'. (Ed. DX Viegas) 

Van Wagner CE (1968) The line intersect method in forest fuel sampling. Forest Science 14, 20-26. 

Van Wagner CE (1982) Notes: Graphical Estimation of Quadratic Mean Diameters in the Line Intersect Method. 

Forest Science 28, 852-855. 

Woodwell GM, Whittaker RH, Reiners WA, Likens GE, Delwiche CC, Botkin DB (1978) The Biota and the 

World Carbon Budget. Science 199, 141-146.



ABSTRACT 

VICTORIAN BUSHFIRES 2009 - 

A PLANTATION COMPANY’S EXPERIENCE 

Malcolm Tonkin 1 

February 7 th 2009 was a day of tragedy in Victoria on which many lives were lost, along 

with houses and other community assets. The fire weather conditions reached and 

exceeded predicted abnormal extremes and HVP Plantations lost around 16,500 ha of its 

plantation estate, while 7,800 ha of its native forest was also burnt. 

As the drought continues beyond 12 years, the level of Company fire preparedness 

increases each year, the fire seasons get longer and their intensity increases; betweenseason 

relief for staff gets shorter and the potential for asset loss is more sustained. This 

is a reflection of the broader Victorian community experience. Although the drought 

conditions and fuels were similar to the much larger 2006 Alpine fires, the severity of the 

weather and the geographic context of the fires on February 7 th provided very different 

outcomes for the community and the company. 

INTRODUCTION 

HVP Plantations (HVP) is Australia’s largest, private, timber plantation company, with assets of over 

$800 million. The company is owned jointly by Australian and US superannuation and investment 

funds and manages 245,000 hectares of land in Victoria including 170,000 hectares of plantations 

These plantations supply over 3 million tonnes of wood to rural and regionally based processing 

industries, including sawmills, paper mills, newsprint mills and panel plants, and generating 

employment for over 3,000 people. 

In the current continuing drought, fire is a significant risk to the plantations and to the industries and 

communities dependent on the timber resources they provide. To manage this risk, HVP is a 

substantial contributor to the prevention and suppression of fire in rural Victoria. HVP lost over 2000 

ha of plantation in each of the large 2003 and 2006 Alpine and associated fires and 16,500 ha in the 

recent January /February 2009 fires. These events have significant Company and community impacts. 

The most recent year in which there was widespread, above-average rain in Victoria was 1996. At 

some locations, such as Melbourne, there have been 12 consecutive years of below-average rainfall. 

In this area, twelve-year rainfall totals have been around 20% below the 1961-90 average, and 10-13% 

below the lowest on record for any twelve-year period prior to 1996. 

Overlayed on the seasonal preconditions for fire generated by the drought, there was a specific build 

up of warm to hot weather conditions in the 11 days leading to February 7 th 2009. 

The weather conditions for this day were accurately forecast well in advance of the day itself, and 

clearly indicated a day of abnormally high risk with Forest Fire Danger Indices over most of the state 

exceeding 100. 

HVP ROLE IN FIRE PREVENTION AND SUPPRESSION 

Wildfire in the rural landscape is a community issue and as a rural landowner HVP plays a role both 

within and around its estate to protect life and assets; in particular its plantation assets. 

HVP maintains seven Forest Industry Brigades (FIBs) under the legislative framework of the Country 

Fire Authority Act (1958). These HVP FIBs are located, managed, equipped, maintained, structured 

1 rd 

General Manager, Stewardship and Risk, HVP Plantations,3 Floor, 517 Flinders Lane, Melbourne, Victoria 3000. 

Ph: 03 9289 1400 Email: MTonkin@hvp.com.au



and financed by the company but, like other Country Fire Authority (CFA) brigades, they operate 

legally in accordance with the Standing Orders of the Chief Officer of the CFA. 

HVP exceeds its statutory obligations in relation the provision of trained brigade members and 

equipment, providing crews to man 20 large (3,000-4,000 litre) forest fire tankers and 38 light (400 

litre) first-attack slip-on units. The company also places two first-attack helicopters on standby during 

key periods in the summer. 

In response to the fire emergency of early 2009, HVP provided more than 200 trained and experienced 

fire-fighters, including more than 50 drawn from other operations interstate and overseas. 

The company enjoys strong relationships with the Department of Sustainability and Environment 

(DSE) and the CFA at both the local and corporate level, as a brigade within the CFA and as a major 

neighbour of DSE, with whom it shares about 70% of its property boundaries in Victoria. 

HVP radio communications are compatible with DSE and CFA and all of the company’s fire vehicles 

have real time GPS radio tracking for additional safety during fires. 

HVP crews attend fires within what it considers to be its area of interest as plantation losses 

commonly occur from fire entering a plantation from outside, rather than from fire igniting within the 

company’s plantations. As FIBs within the CFA, HVP operates at fires within the control plan for that 

incident as determined by the responsible fire control agencies (DSE and CFA). HVP chooses not to 

provide staff for roles in the Incident Control Teams managing fires, but prefers to provide a liaison 

person who is able to emphasise the areas of strategic importance to HVP assets and ensure that our 

crews and adequate equipment are deployed largely under our control to those areas. The role of 

liaison officers has proven to be very effective in sharing information and co-ordinating operations. 

In large fires the company benefits from managing its own logistics. Even though HVP works within 

the Incident Action Plan, Company fire-fighters in debriefing sessions reinforce the need to maintain 

our own logistic functions where possible. This is to ensure efficient service rather than be subject to 

logistics provided by the large emergency organisational arrangements which can become less 

efficient due to their scale. HVP has at all major fires successfully managed its own logistics, 

including equipment, meals, accommodation, fuel distribution mapping services, together some 

additional planning functions. 

HVP Regions each have a fire plan and an annual works program which schedules fire prevention 

maintenance works and additional summer fire crews and equipment availability. Contractors are 

required to have specified fire equipment and HVP has a graduated forest closure schedule to 

minimise internal fire ignition risk which ultimately shuts down all operations in the plantation on bad 

days when the FFDI is predicted to be 45 or above. 

As part of the wider forest industry, HVP has worked co-operatively with other plantation companies 

in fire prevention and suppression for many years. The aggregated resources and skill levels of the 

plantation industry in Victoria are a significant private contribution to the state’s fire suppression 

resources and are the prime example of a public-private partnership in the provision of emergency 

services. 

FEBRUARY 7 TH 2009 

On Saturday February 7 th , after several weeks of temperatures in the mid to high 30’s, the forecast for 

Ballarat (as an example) was for 41 o C, NNW winds of 65 km/hour gusting to 85 km/hour, with a dry 

wind change to the SW in the afternoon at 60 km/hr. The Forest Fire Danger Index (FFDI) was 

forecast to be 185. While it is not clear what FFDI’s of this magnitude really mean, the day was 

clearly forecast in advance to be abnormally severe. 

As a precursor to February 7 th , the first Gippsland fires occurred on 29 th January. The Delburn fire 

(6,400ha) was caused by multiple deliberate lights (a suspect has been charged by police) on the 

afternoon of 29 th January 2009. It burnt around 2,700 ha of HVP plantation (eucalypt and pine) and 

other assets including 29 houses in the Boolara area. The fact that this 6,000 ha fire did not run again 

a week later, under the most extreme conditions parallel to the Churchill fire into the Strzelecki



Ranges, is a great credit to those CFA, DSE and HVP crews who worked on it both before and on that 

day. 

Fires which were controlled but not yet safe at Delburn in Gippsland, and several lightning strikes in 

the North East, were priority tasks identified by HVP in the lead up to February 7 th , as potential fires 

to threaten plantation assets. 

On Saturday February 7 th four major fires burnt into HVP plantations. All four fires were managed by 

joint incident control between the DSE and the CFA. As a Forest Industry Brigade within the CFA, 

HVP resources were tasked by the Incident Control Team for each fire and HVP liaison officers 

placed in key Incident Control Centres. 

The Churchill fire (24,500 ha) in Gippsland started around 1.30 pm. It was allegedly deliberately lit 

and a suspect has been charged by police. It reportedly travelled 6km in 10 minutes in the HVP pine, 

eucalypt plantations and native forest of the Strzelecki Ranges. Substantial loss of life, community 

property and 8,000ha of HVP plantations occurred under the NW wind and then after the SW wind 

change. 

The Murrindindi fire of unknown cause (suspected to be deliberate) started at around 3pm and 

travelled quickly through 15km of native forest to spot into HVP plantations near Buxton within an 

hour or so. Substantial loss of life, community property and 3000 ha of HVP plantations occurred 

under the NW wind and also after the SW wind change. This included the devastation of the town of 

Marysville. The mill of a long-time HVP customer at Narbethong, was also lost, as was a harvesting 

contractor’s equipment. 

The Kilmore East fire is alleged to have been started by fallen power lines around 11.30pm and 

travelled quickly through 25km of the Mt Disappointment forest and burnt HVP’s 980 ha Kinglake 

West plantation. Substantial loss of life and community property occurred both under the NW wind 

and also after the SW wind change including the town of Kinglake, some 35km from the origin. A 

cable harvesting tower and other harvesting equipment was lost in the fire. The Kilmore East and 

Murrindindi fires eventually merged with a total area of over 250,000 ha. 

The Beechworth fire (31,000ha) is alleged to have been started by fallen power lines at 6.20pm, 

burning a number of HVP plantations totalling 1,900 ha under the NW wind and subsequently on the 

SW wind change the next day. 

There was little effective work which could be done to save plantations on the day of February 7 th 

Many HVP staff were involved in saving houses and other assets. Fortunately, all except one of our 

staff could return to a home, although many staff homes were threatened. There were no serious 

injuries and in the subsequent weeks no mills ran out of logs to compound the community distress. 

Sadly there are many in our communities who were not so fortunate. 

Support from the timber industry was critical in enabling HVP staff to get some rest, while still 

ensuring protection of plantation assets. A crew of 9 from Hancock Forest Management in NZ 

assisted on the Delburn fire and were held over in anticipation of the bad weather forecast for Saturday 

7 th February and thereafter a second crew of 8 also came to assist. 

On subsequent days there were many other offers of assistance from the timber industry and crews 

came from plantation companies from both within and outside the state. 

COMPARISON BETWEEN THE 2006 ALPINE FIRES AND FEBRUARY 7 TH 2009 

The 2006 Alpine fires and the 2009 fires both became major fire campaigns within the state. The 

comparison is of interest because while they were both destructive and disruptive events for the state 

and for HVP business, the circumstances were quite different and both scenarios need to be 

incorporated in future plantation management, risk and response strategies. 

Following serious winter/spring rainfall deficits, the fires in 2006 commenced with multiple (80+) 

lightning strikes in remote alpine locations in north-eastern Victoria. These fires had great potential to 

run and coalesce, but it was not until 5 days later that the fire weather caused significant expansion. 

These fires then continued for about 2 months, burning over 1 million hectares (5% of Victoria).



These enlarged fires exceeded the capacity of fire-fighting resources to contain them. No lives and 

few houses were lost. In 2003, 1.2 million ha was burnt similarly in the Alpine regions. 

While again following serious winter/spring rainfall deficits, the fire scenario in 2009 was quite 

different. The fires were limited in number but attributed to arson with their ignition occurring under 

extreme fire weather conditions. They were large fires and again exceeded the capacity of the firefighting 

resources to contain them; however, almost all the destruction happened in the first 12 hours, 

Some 173 persons died and about 2,000 houses destroyed. 

HVP Plantations lost around 2000ha of plantations in 2006. However, due to the nature of the fires, 

the company could use its fire-fighting resources strategically and was thus able to avoid greater loss. 

On the day of February 7 th 2009, when HVP lost about 14,000 ha of plantations, the company could do 

little to avoid loss under horrendous conditions. On subsequent days, only relatively minor losses were 

incurred due to the success of suppression action. In addition to the plantation area, some 7,800 ha of 

managed native forest was also burnt. 

Many HVP plantations are widely dispersed around the northern and southern foothills of the Alpine 

Region. The sheer size of the 2006 fires resulted in different Company plantations coming under 

threat over an extended period and this placed great pressure on staff and resources. With what seems 

to be an increasing regime of large fires in Victoria, the risks to Company resources and the strain on 

available staff are a matter of great Company concern. 

COMPANY RECOVERY 

These fire events have created massive trauma to affected Victorian communities. They also caused a 

massive disruption for HVP. For a three week period almost all staff were diverted for some or all of 

their time on fire related tasks. With tremendous support from contractors and customers, no 

processing plants ran out of logs and this continuity was critical factor in ensuring rural communities 

were not subject to further trauma through loss of income. 

The four main components of the company’s recovery phase are considered to be: salvage, reestablishment, 

staff recovery, and legal processes. 

Salvage 

Approximately 13,000ha of the burnt plantation was too young, or unviable, for salvage operations. 

This land will be cleared for replanting in subsequent years. The 3,500ha remainder were stands older 

than 20 years of age (2,500 ha softwood and 1000 ha hardwood) which might, depending on the fire 

damage, produce logs of merchantable quality. 

These older stands were classified as either: 

• Conventional salvage i.e. logs with burnt bark but no burnt fibre and thus could be supplied to 

existing customers; burnt butt logs go to export); 

• Black salvage (logs containing burnt fibre, which could not be delivered to existing customers 

but are deemed economic to harvest for sale “out of specification” or to export markets); 

• Uneconomic salvage (not scheduled for harvest due to terrain, stand quality or market 

constraints). Where applicable some will be harvested to reduce silviculture costs for reestablishment. 

Softwood (Pinus radiata) salvage was planned to be mostly completed by June 2009 due to the 

anticipated progressive reduction in log quality caused by blue stain, however some salvage operations 

will continue until there are no viable markets remaining. The ability to salvage many stands is 

reduced due to depressed market conditions and a current market surplus of chips and roundwood. 

Opportunities for bio-fuels remain open and are being investigated. 

Hardwood salvage is likely to continue for at least 2 years. Priorities to harvest Eucalyptus globulus 

and E.regnans stands will be primarily set by seeking to return the most profitable sites to production 

quickly. 

Indications are that around 600,000 tonnes of softwood and 400,000 tones of hardwood may be 

salvaged.



Re-establishment 

The HVP Board stated after the fires that “HVP is committed to the Victorian timber industry and its 

people” and “We will … be replanting as quickly as possible.” A strategy has been developed to 

replant the fire affected areas over 5 years. This represents a 40% increase in the establishment 

program and requires a build-up of production in Company nurseries to around 10 million plants per 

year. The strategy for replanting is primarily based on such considerations as spreading the workload 

across the regions, weed management, profitability of sites, and capacity to access appropriate planting 

stock. 

The Company’s research program will need to be reviewed as 330 ha of research plots were lost. 

Staff 

Many HVP employees had their personal homes at risk at some stage during the fires and left the fire 

line in order to protect their own assets and families. 

A significant number of employees have been personally touched by the fires, through the loss of life 

or loss of house of a friend or family member, or through the tragic scenes they witnessed. There is 

likely to be an ongoing impact on the mental and emotional welfare of staff. 

HVP employed a counsellor to tour the regions and provide group and one-on-one counselling where 

appropriate. Reflecting the broader affected community, there were a number of issues of concern 

such as insomnia, exhaustion and fatigue; survivor guilt; anxiety about future fires; and anxiety about 

the future of their employment given the significant loss. It is expected that some of these issues will 

linger and ongoing counselling service is offered for employees and their families. 

Legal 

A number of legal processes arise from such major events. Because of the extensive loss of resource, 

force majeure clauses in a number of contracts had to be reviewed, due to a possible inability to meet 

contract commitments. The operation of these clauses can be quite complex in long term contracts 

where the impacts may not be felt for a number of years. This is new ground for HVP. 

An arrest has been made and a prosecution is pending for the deliberate lighting of the Churchill Fire 

which started on HVP land, and the Victorian Police have sought company information to assist their 

investigative process. Evidence tendered to coronial enquiries will inevitably involve HVP. 

Documentation, maps, ground and aerial photos have been collated in anticipation of enquiries which 

may continue for a number of years. 

The Victorian State Government established a Royal Commission into the Bushfires on 16 February 

with very broad terms of reference, and instructions to deliver an interim report by 17 August 2009, 

and a final report by 31 July 2010. HVP lodged a submission and will follow the deliberations closely 

in anticipation of changes to fire prevention and suppression measures in Victoria which may have 

implications for the company. 

Two fires which caused plantation loss to HVP, the Kilmore East and the Beechworth are alleged to 

have been ignited by powerlines in the electricity distribution network. There is likely to be a class 

action and other damages claims in response to these fires. 

HVP must manage these legal processes while ensuring they do not become a distraction from the 

company business. 

LESSONS 

The extended drought in southern Australia continues to require review and re-assessment of the risk 

of fire to the plantation industry and the management of that risk. There have been a number of 

lessons: 

1. The longer and more intense fire seasons provide greater stress and less time for fire 

suppression personnel to recover. This season has shown that fire fighting forces can still 

perform at a high level despite signs of fire season fatigue, and they continue to provide a 

strong commitment and focus on fire suppression. However given the increasing burden both



physically and emotionally, innovative schemes must be created to reduce the pressure and 

intrusion on personal lives over such a long summer in order to retain skills and experience. 

2. Fire-fighting has inherent dangers, but with few fire-related lost time injuries this past 

summer, HVP is confident that well trained and experienced staff and crews can minimise the 

safety risks while maintaining an aggressive and effective approach to fire control. 

3. The extended drought has raised the cost of fire risk management. Suppression costs to save 

forest plantations have increased due to larger fires burning for longer periods, and smaller 

fires being more threatening. In the drier conditions, the focus for fire suppression extends 

further afield from the Company’s forest boundary, with more fires being attended as a result. 

Fire prevention costs increase as additional on-ground maintenance works are programmed 

and additional measures are put in place, such as the two first-attack helicopters on standby for 

key summer periods. 

4. Fuel reduction burning has been shown again to have great value for managing wildfire. 

There are good examples of fuel reduction burning carried out in recent years by the 

Department of Sustainability and Environment which were instrumental in containing key 

sectors of the Beechworth fire and in turn saving plantation assets. 

5. Good strategy development very early in the fire event is a key factor in containing a fire and 

thereby protecting all assets. In recognising this, HVP places a liaison person in the Incident 

Control Centre to ensure strategy relating to HVP assets and the deployment of HVP firefighting 

resources are aimed at containment of the sections of fire relevant to the minimisation 

of plantation loss. Fire suppression resources at these early stages of large fires can often be 

diverted to the immediate protection of houses and buildings while insufficient focus is placed 

on medium term strategy development for fire containment. 

6. Working within the fire control structures and the Incident Action Plan of the responsible fire 

authorities is critical for ensuring Company fire operations are both legally protected, safe and 

effectively co-ordinated. To work independently in a potentially dangerous environment 

would be an unacceptable risk, and would not attract the mutual support required to achieve 

the plantation protection objective. 

CONCLUSION 

Extended drought creates a higher, more sustained, level of fire risk which must be managed as it 

places a strain on fire-fighting personnel beyond previous norms. A high level of co-operation is 

required with the responsible fire authorities together with an aggressive strategic approach to fire 

prevention and control by the forest owner to ensure that timber resources remain available to 

processors. Growers and processors are significant employers in the rural and regional communities 

in which they are based and days such as February 7 th 2009 are destructive and distracting for both the 

industry and their communities. Effective recovery plans are crucial to return community and industry 

confidence.



SILVICULTURE FOR A CLIMATE OF CHANGE 

David Doley 1 

ABSTRACT 

Climatic changes within the life span of a tree may lead to species extinctions from their 

present environments. Withdrawal of silviculture from native forests follows from the 

assumption that natural processes will achieve the desired conditions. Benign neglect 

will not ensure success in a changed environment. The nature and extent of physical 

changes and the biological attributes of species must be understood in order to apply 

adaptive measures. Most plant species will grow satisfactorily beyond their natural 

distribution ranges provided various growth stages are facilitated and physical factors 

and competition are managed appropriately. Silviculture can and should be used to 

maintain species in desired locations and to introduce them to suitable new locations. 

Native forest silviculture is complex and our understanding of most species is limited, 

but we must attempt to overcome the effects of past disturbances and realise the many 

benefits that forests can provide. 

INTRODUCTION 

Climate may be interpreted broadly as the total environment or domain within which something exists. 

Changes in the forest domain include the meteorological climate, the physical situation of forests in 

the landscape, economics and societal expectations. All these factors impact on the ways in which 

forests function, how they could be managed and how they will actually be managed. The problem is 

complex but it can be addressed by examining the threats and identifying the conditions that are 

needed to secure the stability of the forest. These are the objectives of silviculture in a climate of 

change. 

This review will show that the maintenance of healthy and productive forests in Australia is a complex 

task, often undertaken in the absence of detailed information about the species involved, and 

sometimes without a clear resolution of the objectives that may be held for a particular area of forest. 

If forests have experienced no human influence and they exist in a stable physical environment, then 

over the lifespan of a tree (say 100 to 1,000 years), it can be expected that the required combination of 

conditions will result in sufficient regeneration events to replace the species and that growing season 

conditions will enable the trees to develop to their genetic potential. 

Most Australian forests have been impacted to some degree by humans and are not necessarily in a 

state of ecological equilibrium. One view of forest land management is that human intervention should 

be minimised, whereupon natural processes will restore the forest structures and functions that existed 

before the arrival of European settlers, the ‘benign neglect’ described by Brown (1996). 

Another view is that disturbances to the forest have caused changes that may not be rectified by 

natural processes. These conditions may apply where humans have altered the forest or the 

surrounding environments, where exotic or pest species have been introduced, or when rapid and 

systematic climate change is occurring. Sadly, almost all of the native forests of Australia have 

suffered these disturbances to some extent, as well as the changes that have followed commercial 

timber removals. If natural processes cannot ensure the recovery of forests to the pre-European 

settlement condition, continued intervention may be essential to ensure that future changes in the 

forest are in the directions that will result in the pre-European or some other desired forest condition 

(Doley 1991). 

It does not take long to realise that both views of the responses of forests to disturbance may be 

correct. Small human-induced changes within a large area of forest may have less impact than natural 

disturbance events such as fire and cyclones. Extensive or radical human-induced changes are likely to 

1 The University of Queensland, Centre for Mined Land Rehabilitation, St Lucia Qld 4072, Australia



have greater effects than natural disturbance, especially if the characteristics of the surrounding lands 

have been altered. 

Here is the dilemma. How can forests be maintained in a constant condition while the surrounding 

environment is changing? If constant forest conditions are desired while the forests and their 

environments are not at equilibrium, then we can not turn our backs on continuing management inputs. 

Silvicultural techniques must be accepted by the community if they are to be applied for the long term. 

In the past, forest authorities were left to their own devices in state forests, but increasing demands for 

public participation in land use determinations resulted in the questioning of forest practices and the 

transfer of substantial areas of former commercial forest to reserves (Dovers 2003). At about the same 

time, administrators adopted a market approach to resource management, so that the new objectives no 

longer accommodated some of the silvicultural practices of the earlier times (Dovers 2003). Despite 

the commercial pressures to simplify procedures and disperse skills, there has been progress towards 

the development of silvicultural systems, ranging in Tasmania from ‘benign neglect’ of large areas 

reserved for conservation to intensive management of forests in locations more suited to production 

forestry (Hickey and Brown 2003). 

It is important to focus on the aspects of forest functioning that may require or may benefit from 

intervention, especially in a changing environment. To do that, we must appreciate what is changing, 

how the changes might affect forests, whether the changes can be mitigated and what might be done to 

enhance adaptation to changes that cannot be avoided. 

THE NATURE OF ENVIRONMENTAL CHANGE 

The present climate and physical situation of the forest can be described with some certainty and these 

parameters are widely used to predict the growth rates of growth of trees and the rates of production of 

dry matter or wood in forests. Because of the recorded and projected changes in climate in recent 

times (IPCC 2007), it is appropriate to consider whether the changes in the production functions for 

forest goods and services call for changes in the silvicultural practices in Australian forests. 

Projected climatic changes include increasing atmospheric concentrations of carbon dioxide and other 

greenhouse gases, increasing mean temperature and, depending on the locality, decreasing or 

increasing mean precipitation, increasing seasonality of precipitation, more extreme weather events 

(IPCC 2007). Throughout the world, the concentration of carbon dioxide in the air is increasing by 

about 1.9 ppm yr -1 and the recent increases have been greater than those of earlier decades (IPCC 

2007). For eastern Australia, the rate of increase in mean surface temperature is predicted to be about 

0.25 o C per decade, resulting in mean summer temperatures rising by between 0.6 and 1.5 o C across 

Australia by 2030 (CSIRO 2009) (Figure 1a). At the same time there has been a general poleward shift 

in the high pressure systems, resulting in longer periods for which dry air flows over the Australian 

continent, leading to reduced winter rainfall and hotter summer weather with higher extreme 

temperatures. Slight decreases in annual precipitation occurred over eastern Australia between 1979 

and 2005, but there were slight increases over much of western Australia (IPCC 2007). In contrast, 

CSIRO (2009) predictions to 2030 indicate less rainfall in the west of the continent, with little change 

on parts of the east coast and in the north (Figure 1b).Mean annual relative humidity is predicted to 

decrease in the west and south of Australia (Figure 1c), with an associated increase of up to 4% in 

predicted potential evapotranspiration rates by 2030 (CSIRO 2009) (Figure 1d). 

These and other physical changes impact on biological processes that affect forest growth. For 

example, a change in mean annual temperature of 1 o C is equivalent to a latitudinal shift towards the 

equator of approximately 2 o (Charles-Edwards et al. 1986). This rate of change means that, for the 

condition shown in Figure 1a, a given temperature regime will be found up to 3 o latitude 

(approximately 330 km) farther from the equator after 40 years. At the same time, an increase in 

temperature of 1 o C is equivalent to an increase in altitude associated with a given environment of 

approximately 100 m. This would suggest that altitudinal temperature boundaries in Australia could 

rise by up to 150 m between 1990 and 2030. What might this mean for the functioning of trees or 

other types of plants in the forest?



(a) (b) 

(c) (d) 

Figure 1. Predictions of mean changes in (a) summer temperature ( o C), (b) annual rainfall (%), 

(c) annual relative humidity (%) and (d) potential evapotranspiration (%) for 

Australia between 1990 and 2030, assuming a high greenhouse gas emission scenario. 

From CSIRO (2009). 

RESPONSES TO ENVIRONMENTAL CHANGE 

Biological responses 

For plant species, the responses of many functions to changing temperature can be described by a 

curve with a broad optimum and limiting upper and lower temperatures at which the function ceases or 

the plant dies. For example, Battaglia et al. (1996) showed that in summer, Eucalyptus nitens leaves 

function almost uniformly between about 10 and 30 o C, but in winter the optimum temperature is 

between 10 and 15 o C (Figure 2). In contrast, E. globulus showed an optimum temperature close to 20 

o o 

C in spring and summer, and between 10 and 15 C during winter. It would appear that Eucalyptus 

nitens might be more suited to higher environmental temperatures than E. globulus. This characteristic 

has not prevented selected genotypes of E. globulus from being planted far beyond its original range. 

Slatyer (1977) showed that when snow gum (Eucalyptus pauciflora) plants were grown at different 

temperatures, their optimum temperatures for photosynthesis varied in a predictable manner. Over 

temperature regimes from about 6 to 30 o C there was substantial but not complete adjustment of the 

optimum temperature (Figure 3). It is interesting that the summer sunny day maximum temperature for 

each site studied by Slatyer and Morrow (1977) was lower than the optimum temperature for the 

provenance. Does this mean that snow gum would grow faster at lower altitude than in its present



range? The prediction from these photosynthesis measurements alone would be ‘yes’, but there is 

more to growth than photosynthesis. 

Net photosynthesis (μmol/m 2 /s) 

25 

20 

15 

10 

5 

0 

0 10 20 30 40 

Temeprature ( o C) 

E. globulus spring 

E. globulus summer 

E. globulus winter 

E. nitens summer 

E. nitens winter 

Figure 2. Variation in rates of photosynthesis with temperature of plantation-grown Eucalyptus 

globulus and E. nitens leaves during different seasons in Tasmania. Redrawn from 

Battaglia et al. (1996). 

Optimum temperature ( o C) 

40 

30 

20 

10 

0 

Waste Point 915 m 

Daners Gap 1645 m 

T opt = 0.27T gr + 20.64 

T opt = 0.37T gr + 15.02 

0 10 20 30 40 

Mean growth temperature ( o C) 

Figure 3. Relationship between the optimum temperature for photosynthesis in two provenances 

of snow gum (Eucalyptus pauciflora) from different altitudes and mean growth regime 

temperature in a phytotron. The intersection of the regression line for a provenance 

and the 1:1 line for optimum temperature vs. growth temperature is defined as the 

preferred temperature. The arrows above and below the regression lines indicate the 

sunny summer day maximum air temperatures for the two locations. Redrawn from 

Slatyer (1977).



Temperature response functions of tree growth are difficult to organise and measure at the field scale. 

However, growth can be predicted with reasonable precision by models based on the responses of 

physiological processes such as photosynthesis and transpiration. Several of these models have been 

developed in Australia, by Landsberg and Waring (1987), McMurtrie et al. (1992), Kirschbaum (1999) 

and Battaglia et al. (2004). 

Considering all the factors associated with likely changes in temperature and CO2 concentrations, 

Kirschbaum (2000) concluded that an increase in temperature of 2 o C would cause small changes in 

wood production, but larger and more variable changes would accompany an increase in CO2 

concentration from 350 to 700 ppm. By comparison with seasonal variations, the temperature 

differences of 1 or 2 o C being considered in connection with climate change are small and there may 

be very effective adaptation to the new regimes (Kirschbaum 2000). However, increasing mean 

temperature is associated with an increased tendency for water loss (Figure 1), leading to stomatal 

closure and a reduction in photosynthesis. 

At the local scale, comprehensive climate modelling has predicted that the areas available for many 

tropical rainforest plant communities will contract because they are associated with certain altitudinal 

bands within the wet tropics of Queensland (Hilbert et al. 2001). The changes in forest environment 

are predicted to impact on the habitats for threatened fauna such as the golden bower bird (Prionodura 

newtonia) to an extent that may cause the extinction of the species (Hilbert et al. 2004). 

These habitat contractions are associated chiefly with changes in precipitation, including the frequency 

of cloud or mist at higher altitudes, which has been identified as one factor influencing the distribution 

of mammal species (Kanowski et al. 2001). An increase in air temperature of 1 o C will raise the 

altitude of the cloud base by close to 100 m. This lifting of the cloud base could have much more 

dramatic consequences for species survival than the associated change in temperature. On the Main 

Range of south-east Queensland, throughfall resulting from the interception of cloud by a large tree 

crown increased the annual precipitation beneath that tree by more than one-quarter over the rainfall in 

the open (Hutley et al. 1997). 

The distribution of this cloud interception through the dry season could be especially critical to the 

survival of drought-sensitive species in marginal rainforest environments. Climate-change induced 

movement of distribution limits of European beech (Fagus sylvatica) has been described for the upper 

temperature and dry limits (Jump et al. 2006). Both small saplings and large trees from Central Europe 

(where droughts are uncommon) have been shown to be sensitive to water availability and relative 

humidity (Jump et al. 2006, Lendzion and Euschner 2008) while in Greece populations exposed to 

more frequent droughts were not so affected (Fotelli et al. 2009). This finding emphasises the 

conclusions from Australian studies that ecotypic variation is important in determining the adaptability 

of eucalypt species to environmental change (Ladiges 1974, Li et al. 2000). 

Defining the management objectives 

This catalogue of threats might indicate that catastrophic changes to forest communities are imminent. 

Much of the concern relating to environmental change, whether natural or anthropogenic, is associated 

with its likely impact on biodiversity. While predictable changes could result in many adverse effects, 

including species extinctions, it does not follow that the threats cannot be mitigated by carefully 

planned human intervention (Jones et al. 2007). For forests, the intervention could be in the form of 

silvicultural practices that assist in the maintenance of species in desired areas, the introduction of 

species to areas where they may be expected to survive in future and in the maintenance of forest 

structures that may be crucial to the survival of dependent species. 

The maintenance of biodiversity is commonly the principal function of native forest reserves, and it is 

an important function of native forests managed for commercial timber production. Even if it were 

possible to resolve the constantly varying objectives for native forest management, simple and cheap 

procedures for securing and maintaining regeneration of particular species may no longer be adequate. 

Ferguson (1996) observed that most forest management activities are based on perceptions of market 

value. The monetary valuation of conservation is extremely difficult and a satisfactory valuation 

mechanism has eluded managers up to this time. Despite these difficulties, we are confronted with 

changing conditions that call for planning and action at the physical level so that management options



are still available in 50 or 100 years’ time. One of the critical decisions in relation to forest 

management for both production and conservation is a determination of where species may be able to 

exist and under what conditions, that is, their ecological niche. 

The ecological niche is the environmental space occupied by a species (Austin and Smith 1989) and it 

has two forms. The fundamental niche is described as the range of physical conditions within which 

the species can survive. The realised niche is the actual space occupied by the species. These two 

niches may be identical for species in extreme environments containing very few individuals, but for 

most species they are not congruent and the optimum environment for the realised niche may not even 

be close to the optimum for the fundamental niche (Austin and Smith 1989). This difference can be 

seen in a species such as Pinus radiata, which has a restricted natural range (realised niche) but has 

been planted successfully across a much wider environmental range (fundamental niche). 

Figure 4 presents a hypothetical example of how two species with the same optimum environmental 

condition but different ranges and maximum values for some attribute might be distributed if that 

attribute was critical for their survival and performance. It shows that natural distributions do not 

necessarily indicate whether a species could survive in an environment that differs slightly from its 

existing environment or if management makes the new environment available to that species. 

Relative performance 

4 

3 

2 

1 

0 

0 5 10 15 20 

Environmental value 

Figure 4. A hypothetical example of differences between fundamental ecological niches (1F and 

2F) and realised ecological niches (1R and 2R) for two species with identical optimum 

conditions for an environmental value but different ranges and maximum values for 

some attribute (e.g. photosynthesis or growth). Relative performance is the amount of 

an attribute or the rate of a process for a species. 

Realising the biological potential for change 

From the point of view of determining whether a species will adapt to a new environment, growth 

models can predict the rates of dry matter accumulation under conditions that are not represented by 

the present geographical range of a species. This capacity is most important for indicating the likely 

fate of species with very restricted distributions. For example, Hughes et al. (1996) found that 210 

Eucalyptus species had narrow ecological niches as indicated by a range of mean annual temperature 

of less than 1 o C. Although a narrow range and a small population might indicate limited capacity for 

evolutionary change (Skelly et al. 2007), this does not necessarily mean that the physiological 

tolerance of a species is limited to exactly the location of their present occurrence because the 

distribution may be limited by edaphic requirements or by competition from other species. 

Kirschbaum (2000) showed that conservative assumptions about the extent of adaptation of 

photosynthesis to changing temperature and the response of a species to water deficits can be applied 

to growth models in order to predict the potential range of a species. 

Nevertheless, restricted species distributions may mean that the natural rates of dispersion are slower 

than the rates of environmental translation through climate change (Lindenmayer and Franklin 2002; 

1F 

2F 

1R 

2R



IPCC 2007). As a result, the preferred thermal environment for the species may not occur at a 

contiguous location with suitable nutrient or water availabilities and the future existence of the species 

may be threatened if it is not able to adapt rapidly to environmental change or cannot be relocated to a 

suitable new environment. 

One aspect of the ecological niche may be illustrated by the requirements for seed germination. Weed 

species can often germinate speedily and completely over a wide temperature range, whereas species 

from a forest understorey may have slower germination over a much more limited temperature range. 

However, once germinated and established, the plant may be able to grow successfully and reproduce 

under a different or a wider range of conditions. Therefore, the realised niche for a species may be 

manipulated to ensure the survival of the species or to enhance the yield of a desired product. This is 

the basis of artificial regeneration practices and they should be considered for conservation as well as 

production objectives. 

The physical requirements for growth, based on the environmental constraints on photosynthesis, 

reflect the fundamental niche of the species (Booth et al. 1988). Typically, plants have a broad 

tolerance of temperature and light regimes although some tropical species are sensitive to temperatures 

below about 10 o C and species from humid and shaded environments are subjected to photoinhibition 

and heat stress when exposed to high light and dry atmospheres. Compared with these extreme 

changes, climate change is predicted to cause small changes in average temperature. Therefore, it is 

the likely occurrence of extreme conditions of high temperature and low humidity that are most likely 

to result in damage to plants. 

One aspect of plant growth that may be sensitive to relatively small environmental changes is 

flowering and seed production. In several important forest species, seed set is not uniform from year to 

year and the triggers for mast seeding are not fully understood. One factor that is important is the 

accumulation of sufficient reserves, but there is good evidence in hoop pine (Araucaria cunninghamii) 

that rainfall and seasonal variation in temperature are also important in cone initiation. Seed 

production of hoop pine at coastal or humid locations in Queensland (e.g. Imbil) is sporadic and 

generally limited, but in the more continental environment near the inland limit of the species 

distribution (Yarraman) seed set is more regular and prolific. As a result, the more productive seed 

orchards are located at Yarraman rather than at Imbil (Nikles 1996). Despite these limitations in seed 

production, hoop pine grows naturally over a wide latitudinal and climatic range. Therefore, it is 

possible to establish the species successfully despite some barriers to seed production. The critical 

issue is that natural regeneration is not relied upon in plantations and the same approach could be 

applied to the regeneration of hoop pine and other desired species in natural forests. 

Abundant natural regeneration of a species may occur rarely. If the life span of a tree is 1000 years, 

then theoretically an individual needs to be replaced only once in 1000 years. However, the probability 

of any tree surviving for 1000 years is very small and it might be necessary for millions of seeds to 

produce hundreds of thousands of seedlings and hundreds of young trees in order to overcome the 

risks of individual deaths. 

If seedlings are placed deliberately and then managed to minimise the risk of loss, the number of 

individuals that need to be introduced to a site can be reduced dramatically, possibly to ten instead of 

thousands. The management of the seedlings – their placement and tending – then becomes the key to 

successful regeneration. 

Enrichment planting in native forest is not undertaken very often, principally because of the cost and 

because the established vegetation may prevent the newly planted seedling from obtaining the required 

resources for growth. Shading and water shortages are the principal recognised causes of death in 

newly planted seedlings but in rainforests, burial under litter or total removal by foraging birds and 

animals (Song 2007) may be the most frequent causes of death. For high-value plants, the use of tree 

guards may be necessary to secure seedling establishment. 

SILVICULTURE FOR CHANGE MANAGEMENT 

With some notable exceptions, the recovery plans for many endangered species do not consider 

extending their geographic ranges and could be described more accurately as triage documents. If 

small populations are considered to be expanding from a few individuals and are therefore new, then



the environmental range of sites at which they might be expected to survive may be greater than that 

of the sites of their present occurrence. If dispersal of a species was removed as a restriction to its 

spread, introductions of species to potentially suitable habitats would be possible. The resulting 

increase in the number of habitats should reduce the risk of catastrophic loss in one habitat. 

Introductions of species to national parks may not be desired, but there should be less concern over the 

introduction of species into managed forests. In this way, production forests could greatly enhance the 

opportunities for biological conservation. If the small populations of plants are regarded as old and 

contracting (e.g. Wollemia nobilis and Idiospermum australiense), then it is even more appropriate to 

consider ex situ conservation. 

The effort to understand the biology of Wollemi pine (Wollemia nobilis) (NSW Department of 

Environment and Conservation 2006) and to capitalise on its rarity is an example of commercial 

conservation that follows in the footsteps of the plant hunters of the eighteenth century who carried 

unusual specimens back to Europe for cultivation in garden environments far different from the 

original habitats of the species. Should Wollemia be re-introduced into the wild? Why not? It has 

already been introduced deliberately to areas far beyond its present natural range. 

Idiospermum australiense (ribbonwood), one of the more primitive angiosperms, is a rainforest 

canopy tree with a large and poisonous seed that is dispersed only by gravity. It occurs naturally along 

a few coastal creeks in the wet tropics of Queensland. If left to its own devices, this species is bound 

for extinction as dispersal occurs only seawards and sea levels are predicted to rise. Should this 

species be allowed to become extinct, or should it be relocated to higher positions in the landscape? 

Humankind has a poor record of just leaving things alone and Idiospermum australiense has already 

been introduced to a number of locations outside its natural range. One such planting, by then- 

Governor-General Bill Hayden, commemorated the initiation of the Cooperative Research Centre for 

Tropical Rainforest Ecology and Management. This action may be taken as giving vice-regal 

endorsement of species translocation in the interests of conservation. It is a small step to consider 

management of established plants for the same end. Despite this, there is no recovery plan in place or 

in preparation for Idiospermum australiense. 

Clearly, for each species there needs to be a comprehensive understanding of all of the conditions 

associated with its success in different environments. If a new environment conforms to the central 

portion of the fundamental niche or even if it resembles the former realised niche, then manipulation 

of other stresses (competition) in the new environment could enable the species to survive and 

possibly flourish. Silviculture is the manipulation of a forest in order to secure the optimal 

development of desired individuals of desired species. The application of silviculture to a long list of 

species is much more complex than plantation silviculture of a single species, but it is not impossible. 

Rainforest silviculture must consider individual trees and their immediate environment and decisions 

must be made about favouring one individual over another. While this is requires skill and is timeconsuming 

and therefore expensive, it can be done. If endangered species are accorded a high value, 

then it follows that the resources necessary for their survival in environments appropriate to their 

fundamental niches must be available. 

Most plant species will grow satisfactorily beyond their natural distribution ranges, which may be 

constrained more by the complex interactions of climate and soil on the processes of reproduction and 

natural establishment, rather than on vegetative growth. The ability to cultivate species is already 

highly developed and this knowledge can be applied to the maintenance of species in desired 

locations, or the introduction of species to new locations in order to ensure their continuation on earth. 

We are often told that the speed of environmental change in the last 50 years has been equal to that 

normally experienced over thousands of years (Alley et al. 2003). If that is so, then the processes of 

translocation and establishment of species and their cultivation, even in natural areas, need to be 

accelerated accordingly. This is the time for further and resolute intervention, to restore stability to our 

forest environments. 

Hickey and Brown (2003) described a range of silvicultural techniques that had been tested in the 

mountain ash forests of Tasmania. Identification of the appropriate mixture of techniques and scapes 

of application is critical for the optimisation of the different and often conflicting outputs that are 

expected from these forests. There has been a move from large-scale radical and uniform treatments of



felling and regeneration areas to a variety of opening size, proportion and characteristics of retained 

trees and other vegetation types. One important factor in these treatments is the need for skilled design 

and execution of the working plans, which represents a need for professionals who understand the 

needs of the tree species as well as those of forest-dwelling fauna. 

The need for intervention in Tasmanian national parks was indicated by Marsden-Smedley and 

Kirkpatrick (2000) so that fire regimes more closely resembled those of pre-European times. Hickey 

and Brown (2003) considered it essential to attack bushfires threatening the very fire sensitive King 

Billy pine (Athrotaxis selaginoides) in Tasmanian reserves. Following such fires, further intervention 

in the form of artificial planting may also be warranted in order to secure regeneration of the species 

when natural processes have failed. 

In Victoria, the maintenance of habitats for Leadbeater’s possum (Gymnobelideus leaadbeateri) has 

been a critical issue in the montane ash forest type. Squire (1993), Attiwill (1994a, b, 1995) and 

Lindenmayer and Franklin (2002) have outlined in some detail practices that might contribute to the 

maintenance of commercially productive forests and robust populations of Leadbeater’s possum. 

Suggested measures include the retention of certain densities of trees that are suitable as nesting sites 

now or some time in the future, the provision of varied age structures in the forest and the avoidance 

of salvage operations following disasters. Unanimity regarding the desired or required measures is yet 

to be reached but the existence of varied silvicultural approaches in the montane ash forests indicates 

that the essential experimentation is occurring. 

As the size of a forest management unit decreases, the intensity of management is likely to increase 

and the appropriateness of the use of heavy machinery diminishes. For many of silvicultural 

operations, hand working may be the most effective method. Unfortunately, cost structures in 

Australia almost prohibit manual tending of the forest, so there may be nothing between the extremes 

of extensive disruption by machinery and neglect. 

FORESTS FOR CARBON STORAGE 

Climate change effects and responses in forests cannot be divorced from the issue of carbon storage. 

Forests are viewed as providing the simplest land-based mechanism for ameliorating the effects of 

carbon dioxide released from fossil fuel burning (Garnaut 2008). Increased carbon accumulation in 

forests is expressed mainly in the form of an increase in the quantity of wood that is retained in the 

forest. An important assumption in the modelling of the contribution of forests is the price of carbon, 

which is predicted to increase from about $20 per tonne CO2 equivalent (CO2-e) (in 2005 Australian 

dollars) at present to nearly $600 per tonne by 2100. 

For the first half of the century, however, the price of carbon is not expected to exceed $100 per tonne 

so we should concentrate on the near rather than the very distant view. Polglaise et al. (2008) estimate 

that the wooded lands of Australia could remove up to 143 Mt CO2-e per year and Mackey et al. 

(2008) estimated that the south-eastern eucalypt forests alone could remove 136 Mt CO2-e per year at 

an average rate of approximately 9 t CO2-e ha -1 yr -1 . At a price of $20 per tonne or $180 ha -1 yr -1 , it 

could be more profitable than some present activities. In lower rainfall environments, the potential 

returns from carbon accumulation will be more modest and the risk of loss may be much greater. 

Mackey et al. (2008) argued that if the eucalypt forests of south-eastern Australia were left 

undisturbed they would accumulate much more carbon in ‘green’ or non-tree biomass than has been 

considered in the National Carbon Accounting System (Kesteven et al. 2004). For the wettest eucalypt 

forests that are likely to burn only rarely, it may be appropriate to use these much greater estimates of 

above-ground carbon storage capacity (up to 640 tonne ha -1 ). 

However, in the aftermath of the Victorian fires of 2009, the durability of non-tree biomass in much of 

the south-eastern Australian eucalypt forests must be questioned, especially as these forests are 

expected to become drier in future (IPCC2007, CSIRO 2009). It may be more prudent to aspire to 

lesser goals and to accept that forest management for the accumulation of wood is a lower risk 

alternative to the maximisation of non-wood carbon stocks in forests.



CONCLUSIONS 

Society accepts readily the need to intervene in the management of species that are on the brink of 

extinction. This is demonstrated by the number of recovery plans that have been approved throughout 

the country (examples). For species that occur in forested landscapes, it would seem to be more 

effective to develop comprehensive landscape (forest) management plans that take into account the 

needs of all the species. Foresters have been pilloried for concentrating in the past only on the 

commercial timber species in their forests, but modern management plans are already far more 

cognisant of the needs for conservation of other species, even though almost nothing may be known 

about the requirements for their regeneration and growth. These considerations persuade me that 

continuing, but appropriate, management is required for the whole forest estate. The specific 

objectives of management may vary from place to place and from time to time but the overall 

objective must be to maintain the landscape in the most stable and healthy condition possible, subject 

to the ability to pay for the work. 

National parks may be left alone so that the effects of climate change can be observed with minimal 

interference from humans. However, there is an important role for forests outside these reserves to act 

as repositories for threatened species as well as sites where resources with commercial value can be 

produced sustainably. Hickey and Brown (2003) indicated several directions in which more complex 

management systems were being applied to the forests of Tasmania. This complexity increases the 

price of any product, but if forest management is to be tested against market based criteria, then all 

products and services should be accorded appropriate values. Unfortunately, the timber market is 

global, and Australian products compete against those from countries that do not have the same 

environmental goals. If an open market for wood products is to be retained, then it may not be possible 

for the costs of matrix silviculture to be met from the sale of produce and they must be made up from 

other sources such as general taxation. 

Despite the extent of disturbance caused by timber harvesting, commercial forestry does not appear to 

have been associated with the loss of plant species. Therefore, it can be argued that past silvicultural 

practices have not been totally detrimental to biodiversity in forest areas and the considered 

application of silviculture should provide many opportunities for the cultivation of species that may be 

threatened with extinction due to climate change. 

Silviculture has a crucial role to play in the maintenance of forested landscapes in a changing 

environment. Far from withdrawing from the forest, there should now be a redoubling of effort to 

overcome the effects of past disturbances and to optimise the many benefits that forests can provide. 

REFERENCES 

Alley, R.B., Marotzke, J., Nordhaus, W.D., Overpeck, J.T., Peteet, D.M., Pielke, R.A., Pierrehumbert, R.T., 

Rhines, P.B., Stocker, T.F., Talley, L.D. and Wallace, J.M.(2003) Abrupt climate change. Science 299, 

2005-2010. 

Attiwill, P.W. (1994a) The disturbance of forest ecosystems: the ecological basis for conservative management. 

Forest Ecology and Management 63, 247-300. 

Attiwill, P.W. (1994b) Ecological disturbance and the conservative management of eucalypt forests in Australia. 


Attiwill, P.W. (1995) Managing Leadbeater’s possum in the mountain ash forests of Victoria, Australia - Reply. 


Austin,M.P. and Smith, T.M. (1989) A new model for the continuum concept. Vegetatio 83, 35-47. 

Battaglia, M.C., Beadle, C. and Loughhead, S. (1996) Photosynthetic temperature response of Eucalyptus 

globulus and Eucalyptus nitens. Tree Physiology 16, 81-89. 

Battaglia, M., Sands, P., White, D. and Mummery, D. (2004) CABALA: a linked carbon, water and nitrogen 

model of forest growth for silvicultural decision support. Forest Ecology and Management 193, 251-282. 

Booth, T.H., Nix, H.A., Hutchinson, M.F. and Jovanovic, T. (1988) Niche analysis and tree species introduction 


Brown, M.J. (1996) Benign neglect and active management in Tasmania’s forests: a dynamic balance of 

ecological collapse? Forest Ecology and Management 85, 279-289. 

Charles-Edwards, C.A., Doley, D. and Rimmington, G.M. (1986) Modelling Plant Growth and Development. 

Academic Press, Sydney. 

CSIRO (2009) Climate Change Australia at http://www.climatechangeinaustralia.gov.au/nattemp1.php.



Doley, D. (1991) The C.T. White Lecture for 1990: The future forest: management or curation. The Queensland 

Naturalist 30, 100-109. 

Dovers, S. (2003) Are forests different as a policy challenge? In D.B. Lindenmayer and J.F. Franklin (eds.) 

Towards Forest Sustainability, pp 15-29. CSIRO Publishing, Melbourne. 

Ferguson, I.S. (1996) Sustainable Forest Management. Oxford University Press, Melbourne. 

Fotelli, M.N., Nahm, M., Radoglou, K.m Rennenberg, H., Halvyopoulos, G. and Matzarakis, A., (2009). 

Seasonal and interannual ecophysiological responses of beech (Fagus sylvatica) at its south-eastern 

distribution limit in Europe. Forest Ecology and Management 257, 1157–1164. 

Garnaut, R. (2008) The Garnaut Climate Change Review: Final Report. Cambridge University Press, Port 

Melbourne. 

Hickey, J.E. and Brown, M.J. (2003) Towards ecological forestry in Tasmania. In D.B. Lindenmayer and J.F. 

Franklin (eds.) Towards Forest Sustainability, pp 31-46. CSIRO Publishing, Melbourne. 

Hilbert, D.W., Bradford, M., Parker, T. and Westcott, D.A. (2004) Golden bowerbird (Prionodura newtonia) 

habitat in past, present and future climates: predicted extinction of a vertebrate in tropical highlands due 

to global warming. Biological Conservation 116, 367-377. 

Hilbert, D.W., Osterndorf, B. and Hopkins, M.S. (2001) Sensitivity of tropical forests to climate change in the 

humid tropics of north Queensland. Austral Ecology 26, 590-603. 

Hughes, L., Cawsey, E.M. and Westoby, M. (1996) Climatic range sizes of Eucalyptus species in relation to 

future climate change. Global Ecology and Biogeography Letters 5, 23-29. 

Hutley, L.B., Doley, D., Yates D. J. and Boonsaner A.. 1997. Water balance of an Australian subtropical 

rainforest at altitude: the ecological and physiological significance of intercepted cloud and fog. 

Australian Journal of Botany 45, 311-329. 

IPCC 2007 Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth 

Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. 

Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)] Cambridge University 

Press, Cambridge, UK, 996 pp. 

Jones, R.N., Dettmann, P., Park, G., Rogers, M. and White, T. (2007) The relationship between adaptation and 

mitigation in managing climate change risks: a regional response from North Central Victoria, Australia. 

Mitigation and Adaptation Strategies for Global Change 12, 685-712. 

Jump, A., Hunt, J.M. and Penuelas, J. (2006) Rapid climate change-related growth decline at the southern range 

edge of Fagus sylvatica. Climate Change Biology 12, 2163–2174. 

Kanowski, J., Hopkins, M.S., Marsh, H. and Winter, J. (2001) Ecological correlates of folivore abundance in 

north Queensland rainforests. Wildlife Research 28, 1-8. 

Kesteven, J., Landsberg, J. and URS Australia (2004) Developing a National Forest Productivity Model. 

National Carbon Accounting System - Technical Report No. 23. Australian Greenhouse Office, Canberra. 

Kirschbaum, M.U.F. (1999) CenW, a forest growth model with linked carbon, energy, nutrient and water cycles. 

Ecological Modelling 181, 17-59. 

Kirschbaum, M.U.F. (2000) Forest growth and species distribution in a changing climate. Tree Physiology 20, 

309-322. 

Ladiges, P.Y. (1974) Variation in drought tolerance in Eucalyptus viminalis Labill. Australian Journal of Botany 

22, 489-500. 

Landsberg, J.J. and Waring, R.H. (1997) A generalised model of forest productivity using simplified concepts of 

radiation-use efficiency, carbon balance and partitioning. Forest Ecology and Management 95, 209-228. 

Lendzion, J. and Leuschner, C. (2008). Growth of European beech (Fagus sylvatica L.) saplings is limited by 

elevated atmospheric vapour pressure deficits. Forest Ecology and Management 256, 648–655. 

Li, C., Berninger, F., Koskela, J. and Sonninen, E. (2000) Drought responses of Eucalyptus microtheca 

provenances depend on seasonality of rainfall in their place of origin. Australian Journal of Plant 

Physiology 27, 231–238. 

Lindenmayer, D.B. and Franklin, J.F. (2002) Conserving Forest Biodiversity: A Comprehensive Multiscale 

Approach. Island Press, Washington. 

Mackey, B.G., Keith, H., Berry, S.L. and Lindenmayer, D.B. (2008) Green Carbon: The role of natural forests in 

carbon storage. Part 1. A green carbon account of Australia’s south-eastern eucalypt forests, and policy 

implications. Australian National University, Canberra. 

Marsden-Smedley, J.B. and Kirkpatrick, J.B. (2000) Fire management in Tasmania’s Wilderness World Heritage 

Area: ecosystem restoration using indigenous-style fire regimes? Ecological Management and Restoration 

1, 195-203. 

McMutrtrie, R.E., Comins, H.N., Kirschbaum, M.U.F.and Wang, Y.-P. (1992) Modifying existing forest growth 

models to take account of effects of elevated CO2. Australian Journal of Botany 40, 657-677. 

Nikles, D.G. (1996) The first 50 years of the evolution of forest tree improvement in Queensland. In Dieters, 

M.J., Matheson, A.C., Nikles, D.G., Harwood, C.E. and Walker, S.M. (eds.) Tree Improvement for



Sustainable Tropical Forestry. Proceedings of QFRI-IUFRO Conference, Caloundra, Queensland, 

Australia. 27 October-1 November 1996. 

NSW Department of Environment and Conservation (2007) Wollemi Pine (Wollemia nobilis) Recovery Plan. 

Department of Environment and Conservation, Hurstville, New South Wales. 

Polglaise, P., Paul, K., Hawkins, C., Siggins, A., Turner, J., Booth, T., Crawford, D., Jovanovic, T., Hobbs, T., 

Opie, K., Almeida, A. and Carter, J. (2008) Regional Opportunities for Agroforestry Systems in 

Australia: A Report for the RIRDC/L&WA/FWPA/MDBC Joint Venture Agroforestry Program, RIRDC, 

Barton, ACT. 

Skelly, D.K., Joseph, L.J., Possingham, H.P., Freidenburg, L.K., Farrugia, T.J., Kinnison, M.T. and Hendry, A.P. 

(2007) Evolutionary responses to climate change. Conservation Biology 21, 1353-1355. 

Slatyer, R.O. (1977) Altitudinal variation in the photosynthetic characteristics of snow gum, Eucalyptus 

pauciflora Sieb. ex Spreng. VI. Comparison of field and phytotron responses to growth temperature. 

Australian Journal of Plant Physiology 4, 901-916. 

Slatyer, R.O. and Morrow, P.A. (1977) Altitudinal variation in the photosynthetic characteristics of snow gum, 

Eucalyptus pauciflora Sieb. ex Spreng. I. Seasonal changes under field conditions in the Snowy 

Mountains area of south-eastern Australia. Australian Journal of Botany 25, 1-20. 

Song, G-Z.M. (2007) Patterns of Regeneration of Tree Species in a Subtropical Rainforest. PhD Thesis, The 

University of Queensland. 

Squire, R.O. (1993) The professional challenge of balancing sustained wood production and ecosystem 

conservation in the native forests of south-eastern Australia. Australian Forestry 56, 2237-248.



IMPROVING GREY GUMS TO SEQUESTER CARBON ON 

MARGINAL SITES IN SUBTROPICAL AUSTRALIA 

Paul Warburton 12 , Paul Macdonell 2 , 

Jeremy Brawner 2 , John Huth 3 , David Lee 3,4 

ABSTRACT 

Expected decreasing rainfall in subtropical Australia, driven by elevated atmospheric 

carbon dioxide levels, will present a range of challenges for the forest plantation 

industry. One challenge for forestry research providers is identifying species with 

adaptive plasticity for use over a wide range of sites. The grey gums—Eucalyptus 

longirostrata, E.biturbinata (syn. punctata), E.propinqua, E.major, E.caniculata and 

E.grisea—have high density wood suitable for poles, sleepers, flooring and other uses 

that require strong, durable timber (Boland et al 2006). The performance of some of 

these species in taxa and provenance trials indicates that they have potential as 

hardwood plantation species. As these are dense-wooded species (up to 1070 kg·m –3 ) 

they will exhibit higher levels of carbon sequestration for the same piece size as lower 

density trees and may also have an important role as carbon sinks to mitigate the 

effects of elevated CO2 levels in a changing environment. Results from three year old 

trials, established across south-east and central Queensland and northern New South 

Wales by CSIRO and Queensland Primary Industries and Fisheries (QPIF), indicate 

there are significant differences between species and provide early indications of their 

potential in marginal environments (below 800 mm mean annual rainfall). Results 

indicate that E. longirostrata (14 provenances, 165 families) and E. biturbinata (five 

provenances, 67 families) are the most productiive of the grey gum species and that 

E. longirostrata has the greatest potential across most sites. We also report on the 

establishment of a new E. longirostrata breeding population as part of a collaborative 

improvement program to develop improved germplasm for timber production and for 

offsetting carbon. 

INTRODUCTION 

Identifying species with adaptive plasticity for use over a wide range of sites, particularly those of 

lower rainfall and poorer soils, is an immediate challenge for forestry research providers. It is 

predicted that elevated atmospheric carbon dioxide levels will result in increases in temperature 

between 1.0°C and 6.0°C and changes in annual rainfall from –35% to +10% in most regions of 

Australia by 2070, which will result in marked changes in evapo-transpiration rates and soil moisture 

balances (CSIRO 2001, Henry et al. 2002). 

The grey gums—E.longirostrata, E.biturbinata (syn. punctata), E.propinqua, E.major, E.caniculata 

and E.grisea— are members of the sub-genus Symphyomyrtus. They have a natural distribution that 

primarily covers coastal regions from central New South Wales to south-east Queensland although 

populations of E.longirostrata and E.major can be found further west in Queensland in the Carnarvon 

Ranges and Blackdown Tablelands. These species occur on a variety of sites including hills, ridges 

and slopes and on moderately fertile to poor soils (Boland et al. 2006). 

It was reported that in 2004, commercial forestry in Australia accounted for 43.7 million tonnes of 

carbon dioxide removed from the atmosphere (FWPRDC, date unknown). There are therefore 

opportunities for hardwood plantations to play a significant role in greenhouse abatement. Growth in 

trials established in Australia and South Africa indicates that some grey gum have excellent potential 

1 

Corresponding Author 

2 

CSIRO Plant Industry, Forest Biosciences, Queensland Biosciences Precinct, Brisbane, Australia 

3 

Queensland Primary Industries and Fisheries, Horticulture and Forestry Science, Fraser Road, Gympie 

4 

University of Sunshine Coast, Sippy Downs. Queensland



as hardwood plantations species. With high density (up to 1070 kg·m -3 in natural stands), hard and 

durable wood suitable for poles, sleepers, flooring and other uses requiring heavy grade timber, grey 

gums should play an important role as carbon sinks to help mitigate the effects of elevated carbon 

dioxide levels in a changing environment. 

This paper reports the overall performance of grey gum species in three year old trial established by 

CSIRO and QPI&F with closer investigation of growth data of E. longirostrata and E. biturbinata. 


Figure 1 (A-F) illustrates the geographical locations of the original seed sources established in all three 

trials, the trial locations and the natural distribution of the species included. 

Figure 1. Location of CSIRO and QPIF trials, species’ natural distribution and geographical 

locations of the original seed sources of A. E.longirostrata B. E.biturbinata/E.punctata, 

C. E.propinqua, D. E.major, E. E.moluccana, F. E.grisea. 

CSIRO trials 

Two grey gum species – provenance trials were established in May and September 2006 (Table 1). 

The first trial planted in May at Mt Alma (29 km west by south of Calliope) in central Queensland, 

contained 261 individual families from the five grey gum species. Gum-topped box (E.moluccana) 

and Dunn’s white gum (E.dunnii) were planted as controls in this trial. The trial was established in a 

randomised complete block design with four tree line plots replicated 5 times and a spacing of 4 m × 

2.8 m to give a nominal stocking density of 990 stems per hectare. Fertiliser was applied to the site at 

the following rates; 54.3 kg/ha Amsul, 11.2 kg/ha muriate of potash, 34.4 kg/ha MAP, 5.1 kg/ha DAP, 

1.77 kg/ha zinc oxysulphate, 6.88 kg/ha boronate, 4.86 kg/ha copper oxysulphate. Trees were 

supplemented with 500ml saturated water-retaining gel in the root zone at time of planting. The Mt 

Alma trial site is a free-draining alluvial flat that had been planted with improved pasture prior to the 

trial establishment. 

A subset of one hundred and seventy-five of the seedlots from the Mt Alma trial were established in a 

second trial in September 2006 at Pagan’s Flat (47 km west-south-west of Casino) in northern New



South Wales (Figure 1). This trial was established in a similar design as the Mt Alma trial with five 

tree line plots replicated 5 times and a spacing of 4.5 m × 2 m to give a stocking density of 1111 stems 

per hectare. Fertiliser was applied in the rip line at a depth of 40cm during site preparation at a rate of 

100kg/ha of N:P:K:S (13:14:15:1). No additional treatment was applied. Pagan’s Flat is characterised 

by rolling low hills with a large alluvial plain associated with the Clarence River. The soil at this trial 

site is a very dark, strongly-structured clay loam over medium clay to a depth of 150–200 cm. 

Table 1. Treatment details of CSIRO grey gum trials, established at Mt Alma, Queensland and 

Pagan’s Flat, New South Wales, 2006. 

No. entries 

Species Provenance Latitude Longitude Mt Alma Pagan’s Flat 

Ballon (20899) 26° 27´ 150° 49´ 18 12 

Barakula (20404, 20898) 26° 19´ 150° 41´ 22 12 

Blackdown Tablelands (20008) 23° 46´ 149° 04´ 8 6 

Chinchilla (16008) 26° 22´ 150° 27´ 5 0 

Coominglah (a) (19312) 24° 48´ 150° 57´ 4 2 

Coominglah (b) (20464) 24° 49´ 150° 58´ 7 4 

Goodger (20943) 26° 38´ 151° 49´ 6 5 

Kroombit Tops (20954) 24° 20´ 150° 56´ 3 2 

Nanango (20942) 26° 38´ 152° 00´ 5 4 

E. longirostrata Starkvale Creek (20007) 25° 20´ 149° 15´ 8 5 

Chaelundi (19812) 29° 57´ 152° 22´ 6 4 

Girad (19809) 28° 58´ 152° 15´ 6 4 

Jimna (20405) 26° 37´ 152° 24´ 5 3 

Maryland (20406) 28° 28´ 152° 11´ 5 3 

E. biturbinata Mt Colliery (20407) 28° 17´ 152° 19´ 5 3 

Blackdown Tablelands (20009, 20491) 23° 45´ 149° 04´ 12 10 

Boxvale (20959) 25° 21´ 148° 38´ 3 2 

E. major Gympie (15603) 26° 18´ 152° 49´ 5 4 

Brooweena (20967) 33° 49´ 150° 23´ 5 2 

Blackdown Tablelands (20492) 23° 47´ 149° 04´ 5 4 

Coffs Harbour (15145) 30° 26´ 152° 58´ 7 6 

Goomeri (20949, 20950) 26° 13´ 152° 00´ 10 6 

Kin Kin (20969) 26° 14´ 152° 54´ 3 3 

Nanango (20961) 34° 45' 150° 11´ 4 3 

Taylors Arm (18674) 30° 47´ 152° 39´ 6 5 

E. propinqua Unumgar (20499) 28° 25´ 152° 42´ 6 5 

Curryall (20149) 32° 07´ 149° 53´ 6 3 

Kedumba Valley (19280) 33° 49´ 150° 23´ 7 4 

Nullo S.F. (19797) 32° 45´ 150° 12´ 6 3 

E. punctata Wingello (19352) 34° 45´ 150° 11´ 6 4 

Ballon (15877; n=20) 26° 27´ 150° 49´ 1 1 

Biloela (20951) 24° 15´ 150° 32´ 5 4 

Calliope (20484) 23° 57´ 151° 06´ 5 4 

Coominglah (20771) 24° 49´ 150° 58´ 6 4 

Cooyar (15221; n=10) 26° 52´ 152° 00´ 1 1 

Crediton S.F. (20010) 21° 15´ 148° 29´ 6 5 

Gunnawarra (20960) 18° 10´ 145° 17´ 3 2 

Long Mile Range Creek (20408) 29° 59´ 152° 35´ 5 3 

Monto (15225; n=8) 24° 50´ 150° 57´ 1 1 

Mt Garnett (20957) 17° 43´ 145° 02´ 5 4 

Running Creek (20955) 25° 54´ 152° 19´ 5 3 

Tairo (20965) 25° 45´ 152° 35´ 4 3 

Tomoulin (20958) 17° 32´ 145° 26´ 5 4 

E. moluccana Wondai S.F. (17752; n=3) 26° 19´ 151° 54´ 1 1 

South African SSO n.a n.a 1 0 

E. dunnii Yabbra S.F. (20850; n=30) 28° 35´ 152° 30´ 1 1 

Unknown 2 1



QPIF trial 

Trial 519b grey gum species – provenance trial (Table 2) was established at Pechey (31 km north-east 

Toowoomba), Queensland (see Figure 1) by QPIF in February 2002 and contains 13 bulked 

provenance seedlots from three species. Five provenances of E. longirostrata (Barakula, Coominglah, 

Starkvale Creek, Chinchilla and Blackdown Tablelands) and five provenances of E.biturbinata 

(Chaleundi, Girad, Jimna, Maryland and Mt Colliery) had seedlots in-common with the CSIRO trials. 

The QPIF trial was established as an incomplete block design of 12 tree line-plots with three 

replications at a spacing of 4 m × 2 m, giving a planting density of 1250 stems per hectare. The trial 

was thinned to 300 stems per hectare at age four years. The trial was established in a plantation 

previously planted with Pinus patula. The soil is described as a red ferrosol with a pH of 5.5 in the A 

horizon. The site has an elevation of 720 m a.s.l. 

Table 2. Treatment details of QPIF grey gum trial 519b, established at Pechey, Qld, 2001. 

Species Provenance Latitude Longitude No. Parent trees 

E. biturbinata Chaelundi (19812) 29° 57´ 152° 22´ 10 

Girad (19809) 28° 58´ 152° 15´ 15 

Jimna (20405) 26° 37´ 152° 24´ 5 

Maryland (20406) 28° 28´ 152° 11´ 5 

Mt Colliery (20407) 28° 17´ 152° 19´ 5 

E. grisea Mt Moffat (19702) 25° 05´ 148° 07´ 10 

E .longirostrata Barakula (20404) 26° 19´ 150° 41´ 8 

Blackdown Tablelands (20008) 23° 46´ 149° 04´ 8 

Chinchilla (16008) 26° 22´ 150° 27´ 5 

Coominglah (a) (19312)) 24° 48´ 150° 57´ 5 

Coominglah (b) (20464) 24° 49´ 150° 58´ 7 

Monto (a) (15227) 24° 52´ 150° 58´ 27 

Monto (b) (9662) 24° 52´ 150° 57´ 6 

Starkvale Creek (20007) 25° 20´ 149° 15´ 9 

Climate 

Interpolated climate data for all trial sites, including temperature, rainfall and accumulated monthly 

vapour pressure deficit (VPD) are presented in Table 3. 

Table 3. Temperature and rainfall data for Alarm Creek, Pagan’s Flat and Pechey trial sites 

(source: Datadrill©, Queensland Dept of Natural Resources and Management, 2009) 

Trial 

Location Rainfall Temperature VP deficit 

Latitude Longitude 

M.A.R 

(long 

term) 

M.A.R 

(since 

planting) 

Mean 

maximum 

(hottest month) 

Average accrued 

monthly VPD 

(since planting) 

Mt Alma 24° 01´ 150° 56´ 865 675 31.7°C (Jan) 14.62 hPa 

Pagan’s Flat 28° 56´ 152° 34´ 780 1004 30.6°C (Jan) 11.75 hPa 

Pechey 27° 18´ 152° 04´ 897 746 27.7°C (Jan) 10.94 hPa 

Assessment 

The Mt Alma, Pagan’s Flat and Pechey 519b trials were assessed for survival, height and stem 

diameter at breast height over bark (DBHOB) at 32 months, 27 months and 36 months respectively. 

Due to different assessment ages across the trials, it was necessary to set an arbitrary age for 

comparison purposes. Therefore the assessments of Mt Alma, Pagan’s Flat and Pechey will be treated 

as one age class (nominally age three years). Trees with dead tops, or tops blown out and trees blown 

over were omitted from the growth calculations but were included for survival. Tree height and 

DBHOB measurements were used to calculate volume indexes (VI). As no volume equation is 

available for these species, the following formula was used: 

VI (m 3 ) = 1/3 × Height (m) × Basal Area (m 2 ) 

The Giant Wood Moth (Endoxyla cinerea) was present at Mt Alma and the incidence of attack on 

trees in the trial was recorded. Trees were scored for the presence or absence of an entry or exit hole.



Analysis 

The data from the three grey gum trials were analysed separately to evaluate the performance of 

species and provenances from south-east Queensland and northern New South Wales. Analysis of 

variance of the data for each trial was undertaken using Genstat version 11.1. The statistical model 

used was; 

yijk = + Ri + Sj + Pjk + Eijkl 

where, μ= overall mean; Ri is the effect of the i th replicate; Sj is the effect of the j th species; Pjk is the 

effect of the k th provenance within the j th species and Eijkl is the random error associated with the ijkl th 

plot mean. Transformation (Log, Sqrt and arcsin) used on survival data but did not alter results of the 

analysis and therefore results of the analysis of non-transformed data are presented here. Post hoc 

Fischer’s Protected Least Significant Difference tests were applied to plot means if significant 

differences were found by the ANOVA. 

RESULTS 

Mean survival was good across all species and sites, ranging from 82.5% to 96% (Table 4). Only Mt 

Alma showed a significant difference in survival (P



Based on their performance at the species level, provenance performance data of E.longirostrata and 

E.biturbinata was selected for separate analysis (Table 5). There was little difference in survival 

between provenances at all three sites and only Pechey exhibited any significant differences. The 

Chaelundi provenance of E. biturbinata had significantly lower survival (P



DISCUSSION 

In Australia, the National Agriculture and Climate Change Action Plan (2006–2009) identifies forests 

as carbon sinks (Natural Resource Management Ministerial Council 2006). Schemes such as the 

Greenhouse Gas Reduction Scheme (GGAS) allow abatement certificates for sequestering carbon to 

be generated by plantation growers. Provision in the GGAS Carbon Sequestration Rule recognises 

permanent carbon storage whilst rotating harvests from plantations in the sequestration pool. 

Assuming that net carbon dioxide flux is proportional to the biomass stocks in a given area, there are 

clear opportunities for greenhouse gas abatement via plantations and examples of plantings for carbon 

sequestration can be found in many countries (Christie and Scholes 1995, Singh et al. 2000). The 

main hardwood species currently dominating solid wood plantations in subtropical Australia are 

Corymbia citriodora ssp. variegata, E.grandis x E.camaldulensis hybrids, E.cloeziana, E.dunnii and 

E.argophloia. Current expansion of plantations into new areas, expected changes in our climate and 

the lack of a species that would be a panacea for Australian wood production drive the need for 

genetic diversity within the northern plantation estate. Numerous species trials in South Africa have 

shown that E.longirostrata has potential on wet sites in Zululand, ranking alongside E. grandis for 

growth and interest is now turning to its adaptability to drier sites (Gardner et al. 2007, Gardner and 

Little 2007). Lee et al. (2001; 2009) identified that this species has potential in Queensland and 

Henson et al. (2008) suggested E. longirostrata had potential for marginal sites in New South Wales. 

In this study, E.longirostrata and E.biturbinata consistently ranked well for growth, quantified as 

volume index (VI), against the other species established in these trials. E. biturbinata had a greater VI 

than E.longirostrata at Pechey although the difference was not significant. The best VI across all trials 

was shown by E.longirostrata at Mt Alma. At age three, the mean annual increment (MAI) for 

E.longirostrata was 7.7 m 3 /ha at Mt Alma and 4.3 m 3 /ha at Pagan’s Flat which was slightly lower than 

the top performer, E.dunnii (5.1 m 3 /ha). It should be noted however, that while E.dunnii performed 

better at Pagan’s Flat than the grey gums as a species, some E. longirostrata provenances (e.g. Ballon 

and Coominglah) had a similar VI at the provenance level. E.biturbinata achieved a MAI of 6.0 m 3 /ha 

at Pechey. 

At the time of planting, E.biturbinata and E.punctata were distinct species. Recent taxonomic 

changes have placed E. biturbinata in synonomy with E. punctata (CPBR 2006). We have retained 

the distinction for the purposes of this study but the affiliation between E.biturbinata and E.punctata is 

reflected in the fact that there was no significant difference in VI between the species at either Mt 

Alma or Pagan’s Flat. The slightly lower mean values for E.punctata may be due to the more southern 

origins of the seedlots with respect to the climatic differences with the trial sites. With the lowest 

VPD and mean hottest temperature and highest long-term MAR, the Pechey trial site appears to 

provide the most conducive conditions to good tree growth however, three-year growth was generally 

greatest at Mt Alma, followed by Pechey and Pagan’s Flat. This is despite Mt Alma receiving the 

lowest rainfall and reaching the highest mean maximum temperatures during the trial period. Gardner 

et al. (2007) reported the potential of E.longirostrata on productive wet sites and these results 

demonstrate that E.longirostrata also has potential on sites receiving



for volume along with E. longirostrata Monto (a) and Coominglah (P



are susceptible to damage by Giant Wood Moth. Early indications of differences in susceptibility 

between species and provenances within species require thorough investigation at more advanced 

ages. A large progeny trial has been established by CSIRO and QPIF to further improve the growth 

characteristics and wood properties of E. longirostrata. The objectives for this trial is to make 

selections of best individuals for breeding, the production of improved seed and selection of clonal 

germplasm for distribution to the commercial hardwood plantation industry. 


We thank many staff from CSIRO (David Boden, Tim Vercoe and Darren Morrow) and at the 

Department of Employment, Economic Development and Innovation and its predecessor departments 

(Murray Johnson, Alan Ward, Tony Burridge and John Oostenbrink) for their assistance in 

establishing and measuring the trials and collating data. Thanks are also due to Integrated Tree 

Cropping and Great Southern Limited for their support and commitment in hosting the CSIRO trials. 

REFERENCES 

Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall, N., Hyland, B.P.M., Johnston, R.D., Kleinig, D.A., 

McDonald, M.W and Turner, J.D. (2006) Forest Trees of Australia. 5 th ed. CSIRO Publishing, 

Collingwood, Victoria, Australia. 736 p. 

Christie, S.I. and Scholes, R.J. (1995) Carbon storage in Eucalyptus and pine plantations in South Africa. 

Environmental Monitoring and Assessment 38:231–241 

CPBR 2006. Euclid – eucalypts of Australia, 3 rd edition. Centre for Plant Biodiversity and Research, Canberra. 

DVD. 

CSIRO (2001) www.dar.csiro.au/publications/projections2001.pdf 

Darrow, W.M. 1997 Eucalypt site-species trials in Zululand. ICFR Bulletin Series 3 – 97. i–26 

FWPRDC (Forest and Wood Products Research and Development Corporation) pub date unknown. Forests, 

Wood and Australia’s Carbon Balance. 

Gardner, R.A.W. (2001) Alternative Eucalypt species for Zululand: Seven year results of site : species 

interaction trials in the region. Southern African Forestry Journal. 190: 79-88 

Gardner, R.A.W. (2006) Early performance of promising cold-tolerant and sub-tropical eucalypt species in the 

warm temperate climate zone of KwaZulu-Natal. ICFR Bulletin Series 13/2006 i+ 21pp. 

Gardner, R.A.W., and Little, K.M. (2007). Investigating the commercial potential of alternative Eucalyptus and 

Corymbia species for northern, coastal Zululand. In IUFRO WG 2.08.03 Improvment and culture of 

Eucalypts (Durban, Republic of South Africa: Institute for Commercial Forestry Research), pp. 1-14. 

Gardner, R.A.W., Little, K.M., Arbuthnot, A. (2007) Wood and fibre productivity potential of promising new 

eucalypt species for coastal Zululand, South Africa. Australian Forestry. 70:37–47. 

Henry, B.K., Danaher, T. McKeon G.M. and Burrows, W.H. (2002) A review of the potential role of 

greenhouse abatement in native vegetation management in Queensland’s rangelands. Rangelands Journal. 

24(1): 112–132. 

Henson, M., Smith, H.J. and Boyton, S. (2008) Eucalyptus longirostrata: a potential species for Australia’s 

tougher sites? New Zealand Journal of Forestry Science. 38(1): 227–238. 

Lawson, S. (2003) Susceptibility of eucalypt species to attack by longicorn beetles (Phoracantha spp.) in 

Queensland. Hardwoods Queensland Report No. 10. Queensland Forestry Research Institute, Agency for 

Food and Fibre Sciences, DPI. 11??pp 

Lee, D.J., Nikles, D.G., and Dickinson, G.R. (2001). Prospects of eucalypt species, including interspecific 

hybrids from South Africa, for hardwood plantations in marginal subtropical environments in Queensland, 

Australia. Southern African Forestry Journal 190, 89-94. 

Lee, D.J., Huth, J.R., Osborne, D.O., and Hogg, B.W. (2009). Selecting Hardwood Varieties for Fibre 

Production in Queensland's Subtropics. In Australasian Forest Genetics Conference (Fremantle, West 

Australia: Forest Products Commission, Western Australia). 

Natural Resource Management Ministerial Council (2006) National Agriculture and Climate Change Action 

Plan, 2006–2009. Department of Agriculture, Fisheries and Forestry. 14pp 

Singh, T.P., Varalakshmi, V. and Ahluwalia, S.K. (2000) Carbon sequestration through farm forestry: case 

from India. Indian Forester. 126(12): 1257–1264.



THE UTILISATION OF RED MAHOGANY FOR HIGH VALUE 

PLANTATION FORESTRY IN THE TROPICS 

Jeremy Brawner 1 2 , David Bush 3 , Paul Macdonell 2 , 

David Boden 2 , Simon Potter 4 Paul Warburton 2 , and Paul Clegg 5 

ABSTRACT 

The expansion of red mahogany (Eucalyptus pellita) plantations in the north of 

Australia has increased interest in its domestication. A review of an overseas breeding 

population was undertaken to enhance our current understanding of genetic parameters, 

which are essential for the development of advanced generation breeding programs. 

The large genetic differences between and within provenances, as well as the moderate 

heritability estimates for growth and form traits, imply that extensive tree breeding can 

create improved breeds and reduce variability in plantation forests. Our analysis 

suggests that breeding value predictions of parents tested in first generation 

provenance/progeny trials are less indicative of genetic merit than those from second 

generation trials. The low heritability estimate in the first generation relative to the 

second generation, and the reduced inter-generational correlations will influence the 

level of genetic gain that can be realised. The implications of these findings for the 

management of E. pellita breeding populations and the production of improved seed for 

plantation forestry in the tropics are discussed. 

INTRODUCTION 

Eucalyptus pellita is a forest tree of the humid and subhumid tropics, which occurs naturally in 

Southern New Guinea and Northern Queensland (Harwood 1998). Red mahogany is sympatric with 

Acacia mangium, A. crassicarpa, A. aulacocarpa, Xanthostemon sp., Stenocarpus sp., Syzygium sp. 

and Eucalyptus brassiana between rainforest and savannah woodland (Vercoe and McDonald 1991). 

The species has been grown in plantations to a limited extent in Australia with a recent expansion of 

the plantation estate as Managed Investment Schemes move into the tropical north. 

A tree improvement program using a wide range of New Guinea and Queensland provenances 

established in replicated progeny trials was initiated by the Commonwealth Scientific and Industrial 

Research Organisation and the Queensland Department of Primary Industries and Fisheries in 1991 

and 1992. This program has led to the establishment of many second generation seedling seed 

orchards and the recent establishment of a third generation progeny trial in 2007. 

In these second generation trials, one year after planting the mean height of the second generation 

orchard families was 10 percent greater than the mean of natural-provenance controls (Harwood et al. 

1997). The species has also been established in collaborative trials in the Northern Territory of 

Australia as well as across 22 sites in Brazil, in Dongmen, China, South Kalimantan, Indonesia and 

Sabah, Malaysia (Harwood 1998). 

1 Corresponding Author 

2 CSIRO Plant Industry, Forest Biosciences, Queensland Biosciences Precinct, Brisbane, Qld, Australia 

3 CSIRO Plant Industry, Forest Biosciences, Yarralumla, ACT, Australia 

4 CSIRO Plant Industry, Forest Biosciences, Clayton, Vic, Australia 

5 Toba Pulp Lestari, Medan, Sumatra, Indonesia



Figure 1. – Native range of Eucalyptus pellita 

Toba Pulp Lestari (TPL) is one of the smaller companies that form part of the APRIL group, which 

has a pulp and paper production focus. The company has several mills and produces pulp and paper 

for both the domestic (Indonesian) and export markets. TPL’s plantation forest estate is located 

around Lake Toba, Sumatra, Indonesia at elevations from 900 to nearly 2,100 metres above sea level. 

The soils are mostly volcanic ash from the Toba eruption which occurred most recently 50,000 years 

ago. The soil is classed locally as ‘Toba Tuff’, which is a typic Andisol of volcanic origin with 

characteristics of very high permeability, high Al content with resulting low N and P status. An 

annual rainfall of about 2,000 to 3,000 mm occurs throughout the year, with lower rainfall from May 

to August. Temperatures range from about 17-22 o C. The environment of the TPL plantation estate is 

dramatically different from that encountered in the lowlands of Sumatra, where APRIL has established 

extensive Acacia mangium plantations. 

The E. pellita breeding population is a major part of TPL’s eucalyptus tree improvement program, 

which has focused on increasing per hectare pulp yield. The company has invested significant 

resources on the genetic improvement of eucalypts for use in the high-elevation plantation estate and 

has advanced generation breeding populations for other eucalypt species (E. grandis and E. urophylla) 

at various stages of domestication. Although E. pellita may not be used extensively for reforestation 

as a pure species in the highlands of Sumatra, it has considerable value as: 1) a hybrid parent, 2) an 

alternative for Acacia mangium on the Sumatran non-peat soils of the lowlands and 3) a base 

population for TPL’s sister companies operating in other regions.



For a variety of reasons, interest in alternative species for the lowland Acacia mangium estate has 

increased recently. This has prompted the company’s tree improvement program to establish an E. 

pellita progeny trial and a range of other genetic material in the lowland estate of its sister company 

Riau Andalan Pulp and Paper (RAPP), which is located around the town of Pekanbaru in Sumatra. 

These tropical lowland environments are typically not suitable for plantations of eucalypt species due 

to very high pest and disease pressure (Harwood 1998). If trial and pilot scale plantations can cope 

with the disease pressure common to the wet lowlands and yield is satisfactory, the creation of an E. 

pellita based breed for the lowlands could greatly increase the utility of the species for operational 

reforestation across the tropics. 


The foundation of TPL’s E. pellita breeding program is a base population contained within a 

first generation progeny trial established in 1995. This trial was assessed to three years of age and 

each eight-tree row-plot was phenotypicaly thinned to leave the best tree and create a seedling seed 

orchard (Eldridge et al 1993). The resulting open-pollinated seedlots were collected from the best 

individuals in each family for second generation trials. One second generation progeny trial was 

established close to the first generation trial, and although these trials were planted several years apart, 

they have experienced very similar growing environments. Within the upland estate, the rainfall and 

temperature patterns are generally constant within and across years. The two upland field trials sites 

were selected to be relatively flat with homogenous, very deep and well-drained ‘Toba tuff’ soils. The 

first and second generation trials were established at a different spacing to reflect the changing 

operational practices of TPL with 1333 (2.5 X 3 metres) and 2222 (1.5 X 3 metres) trees per hectare 

planted respectively. Supplementary genetic material from the native range (19 families) was infused 

into the E. pellita base population at the establishment of the second generation upland and lowland 

trials. 

Additionally, a smaller second generation progeny trial was established in the lowlands of 

Sumatra. This trial was established to evaluate the productivity of E. pellita in the lowlands and 

provide an indication of the stability of family performance across the contrasting lowland and upland 

environments. Based on previous experience and the findings of other research (Harwood et al 1997, 

Pegg and Wang 1994, Werren 1991) showing Queensland sources were less productive than New 

Guinea sources when established in the wet tropics. It has a very pronounced dry season)., the openpollinated 

families established in the lowland trial were exclusively from New Guinea. A seedlot that 

was a bulk of selected Queensland families identified in the first generation progeny trial was included 

to assess the merit of expanding the representation of Queensland germplasm in future trials. 

Table 1. Description of Eucalyptus pellita field trials established in Sumatra, Indonesia by 

Toba Pulp Lestari 

Trial Location 

EB1 EB15A EB15B 

Near Lake Toba Near Lake Toba Near Pekanbaru 

Highland 

Highland 

Lowland 

Generation First Second Second 

Date Established July 1995 February 2003 March 2003 

Soil Toba Tuff Toba Tuff Lowland clay 

Spacing 2.5 X 3 m 1.5 X 3 m 1.5 X 3 m 

Replications 8 8 8 

Design 8-tree row-plots 10-tree row-plots 10-tree row-plots 

Treatments 66 73 29 

DBH Assessed last 36 months 30 months 30 months 

Genetic material assessed in field trials 

The majority of the germplasm established in the three progeny trials of E. pellita were 

established by the TPL breeding program from seeds collected within the species native range for the 

first generation and from seeds collected within the thinned first-generation progeny trial for the 

second generation. Additionally, controls of locally produced clones and bulk seedlots were planted 

within these trials. Descriptive statistics were generated for each trial individually, using data sets that



included all controls and native range E. pellita families. For these descriptive statistics, an additional 

country code labeled ‘TPL’ was included to congregate the various controls that were used in these 

trials. A description of the genetic material established and the data analyzed for each trial is included 

in Table 2. 

Table 2. Description of provenances and number of trees within each of the Eucalyptus pellita 

field trials established by Toba Pulp Lestari 

Source EB1 EB15A EB15B 

Australia 

Abergowrie 

Bloomfield 

56 

16 

80 

80 

Cardwell 8 80 

Daintree 24 80 

Helenvale 64 240 

Kirrama Range 64 80 

Kuranda 1120 3240 

Tinaroo Creek 16 

Wonga – Daintree 64 80 

Indonesia 

Papua New Guinea 

Toba Pulp Lestari 

Bupul Muting 64 80 60 

Papua (Province of Indonesia) 312 860 670 

Keru to Kumbalusi 64 80 60 

Keru to Mata 56 60 

Serisa 56 

South of Kiriwo 16 80 60 

Tokwa 16 270 290 

Control Clone 1 80 80 

Control Clone 2 80 80 

Control E.grandis 24 60 60 

Control E.pellita Lowland Seed 70 180 

Control E.pellita QLD selects 80 60 

Control E.urophylla 8 60 60 

Total trees 2048 5760 1720 

Statistical Analyses 

Data were analyzed in two stages; descriptive statistics were generated to provide general information 

on the performance of the different material included in the trials prior to undertaking across site 

analyses to produce genetic parameters. Data cleaning, standardisation, pedigree file production and 

the generation of descriptive statistics within each trial were completed using SAS (SAS 1991). To 

generate descriptive statistics for each site, a mixed model including the following factors was used to 

produce least square means for normally distributed traits: overall trial mean, replication, plot within 

replication, country of origin, locality within country, country by replication, locality by replication 

interaction and residual error. For these single site analyses, all factors were fixed, with the exception 

of the plot within replication (family by replication) and the error terms. 

Estimates of genetic parameters for diameter growth at latest assessment 

A single pooled-site analysis was used to generate breeding value predictions and variance component 

estimates using the latest-age, pre-thinning diameter at breast height (DBH) data available for each 

trial. Data was standardized to have a mean of zero and a variance of one within each trial to remove 

scale effects that contribute to genotype by environment interactions and facilitate convergence of the 

Restricted Maximum Likelihood process. All controls, treatments that were not open pollinated 

seedlots of E. pellita, were removed to ease variance component estimation and clarify their genetic 

interpretation. An individual-tree model was implemented in ASREML with standard errors for 

genetic parameters calculated using the Taylor series expansion method within ASREML (Gilmour et 

al. 2005). For the pooled-site analysis, the following tri-variate mixed-effects model was used to



obtain estimates of variance and covariance components for diameter growth at each of the three trial 

sites: 

y = Xb + Z p p + Zii 

+ e , 

where y i is the vector of observations of the latest DBH assessment undertaken by TPL, 

b is the vector of fixed effects including overall mean, and replicate within trials and X is an 

incidence matrix relating the observations in y to the fixed effects in b , p is the vector of random 

2 

plots nested within replicate where a separate plot variance ( P) was estimated for each site and Zp is 

a known incidence matrix relating the observations in y to the plot effects, i is the vector of random 

individual tree effects ~MVN ( 0, G ⊗ A , where, 

) 

2 ⎡σ 

1 

⎢ 

G = ⎢σ 

12 

⎢ 

⎣σ 

13 

σ 12 

2 

σ 2 

σ 23 

σ ⎤ 13 

⎥ 

σ 23 ⎥ 

2 

σ ⎥ 

3 ⎦ 

A = the numerator relationship matrix and Zi is a known incidence matrix 

2 2 2 

relating the observations in y to effects in i, 1, 2 and 3 are the genetic 

variances for trial 1, 2, and 3 respectively with covariances between the trials 

on the off-diagonal, e is random vector of residuals where a separate error 

term, 

2 

E, was estimated for each trial to account for heterogeneous residual errors between trials. It 

was assumed that there was no correlation of residuals or plot effects between trial sites. 

The mixed model equations were modified by including genetic groups to implicitly fit fixed effects 

for ‘provenances’ or ‘localities’ (Westell et al. 1988, Quass, 1988, Gilmour et al. 2005). The primary 

reason for using genetic groups in this analysis was to differentiate between the common environment 

(or pollen source) of the seed collected from the native range and seed collected from within the first 

generation trial. Each the 16 native range provenances included in the progeny trials were assigned to 

a different genetic group. Using volatile leaf oil composition, Doran et al (1995) concluded there was 

no clear evidence to support the subdivision of E. pellita from northern Queensland, Cape York and 

southern New Guinea into sub-populations. In a separate study using isozymes House and Bell (1996) 

determined there was no support for the separation of New Guinea and Cape York populations as a 

distinct grouping from the more southern populations of E. pellita. It was therefore decided to create 

‘finer’ genetic groups based on provenances or localities of seed collection as reported by TPL rather 

than group the populations arbitrarily by region or country. An additional genetic group was created to 

represent the common pollen parent of all seedlots collected from within the first generation for 

establishment in the second generation trials. For comparative purposes, a pedigree file assigning 

groups back to the provenance of origin was also used to determine the effect of this new genetic 

group on genetic parameter estimates. 

Individual-tree narrow-sense heritability and inter-site genetic correlation estimates for diameter 

growth at the latest age of assessment were taken from the single pooled-site analyses following 

Falconer and Mackay (1996): 

V 

2 

2 A ˆ 

σ G 

h = ≈ 

is the estimated individual-tree narrow-sense heritability from the 

2 2 2 

VP 

σ G + σ P + σ E 

pooled-site analysis, where VA is and estimate of the additive genetic variance, VP is and estimate of 

2 2 2 2 2 

the phenotypic variance, G is the sum of the genetic variance estimates ( 1, 2, 3), P is the 

sum of the within plot variance estimates and 

2 E is the sum of the estimated error variances. As the 

genetic analysis included seedlots from both natural stands and the first generation progeny trial, the 

coefficient of relationship used to estimate of additive genetic variance was not altered from 0.25 (Luo 

et al 2006 and House and Bell 1996). 

V σ 

= is the estimated individual tree narrow-sense heritability from each site 

hˆ 2 

i 

Ai 

VPi 

≈ 2 

σ P i 

2 

G i 

2 2 

+ σ P + σ i E i 

2 

Gi is the genetic variance from site i, 

i, where 

variance for site i. 

2 Pi is the plot error for site i, and 

2 Ei is the residual



The genetic correlation between trial sites for growth was calculated as: rˆ 

ADD = 

σ g 

2 2 

σ gσ 

g 

, with 

covariance ( g) between sites (ie. 1, 2 and 3) and genetic variances defined as above (Burdon 1977, 

Gilmour et al. 2005). 

RESULTS 

There were significant differences between the various origins of material established in these progeny 

trials at all ages for the assessed growth traits. Survival was similar at each site with 71, 88 and 83 

percent of the trees remaining in EB1, EB15A and EB15B respectively at the final assessment. The 

locally produced TPL material was consistently superior across all ages of assessment in each trial, 

with the exception of the 36 month DBH assessment at the first generation trial (EB1). Within this 

first generation trial, the material from Australia performed particularly well at 36 months of age and 

was significantly larger than the PNG (p



Table 4. Descriptive statistics of provenances, localities and controls included in three Toba Pulp Lestari E. pellita field trials 


DBH 

Origin HT 12 HT 24 DBH 24 DBH 36 HT 18 DBH 18 DBH 30 HT 18 18 DBH 30 

AUS Abergowrie 3.1 9.2 8.9 12.6 8.5 7.2 9.6 

AUS Bloomfield 2.6 7.9 7.2 9.7 6.7 5.8 7.9 

AUS Cardwell 3.0 8.8 8.1 10.7 6.8 5.8 7.8 

AUS Daintree 3.1 9.2 8.6 12.9 6.5 5.9 8.1 

AUS Helenvale 3.0 8.7 8.5 12.4 7.1 6.1 8.6 

AUS Kirrama Range 2.9 9.0 8.6 13.0 9.2 7.8 10.5 

AUS Kuranda 3.5 9.7 9.5 13.3 8.1 6.8 9.1 

AUS Tinaroo Creek 3.3 10.9 10.1 13.5 

AUS Wonga – Daintree 3.3 9.1 8.2 11.6 9.0 6.8 9.5 

IND Bupul Muting 3.2 8.4 8.2 11.7 6.4 5.7 7.8 9.3 6.9 8.6 

IND Irian Jaya 3.2 8.7 8.1 11.1 8.0 6.8 9.6 10.9 7.6 9.6 

PNG Keru to Kumbalusi 2.7 8.0 7.5 10.6 9.9 8.3 11.7 11.9 8.7 11.1 

PNG Keru to Mata 2.3 6.7 6.4 8.7 9.0 6.4 7.8 

PNG South Kiriwo 3.1 8.6 8.6 11.6 7.1 6.3 8.9 9.9 6.9 8.6 

PNG Serisa 2.5 7.7 7.2 10.2 

PNG Tokwa 3.0 7.9 8.3 11.0 6.9 5.8 7.5 9.7 6.6 8.1 

TPL E.pellita Lowland Seed 8.1 6.8 9.3 11.5 7.9 9.9 

TPL E.pellita QLD selects 7.7 6.8 9.1 10.3 7.1 9.0 

TPL E.grandis 4.0 11.4 9.8 12.2 9.9 7.8 11.1 9.6 6.2 8.7 

TPL E.urophylla 3.5 10.6 8.1 11.2 8.3 7.2 10.2 10.9 7.4 8.8 

TPL Clone 1 High 11.0 8.4 11.3 

TPL Clone 2 High 10.2 7.9 11.5 

TPL Clone 1 Low 11.1 8.1 9.5 

TPL Clone 2 Low 13.1 9.3 12.4 

SED 0.49 0.80 0.92 1.25 0.50 0.41 0.62 0.53 0.46 0.64 

Pr > F 0.0378 0.0012 0.0006 0.0012



The germplasm originating from the TPL breeding program that was used as controls generally 

performed well in these trials with substantial increases in relative performance from the first 

generation to the second. The drop in the growth of the E. grandis and E. urophylla controls relative 

to the other controls when established in the second generation lowland trial is likely due to fungal leaf 

pathogens reported to affect these species when established in the wet tropics (Dionese et al. 1984, 

Harwood et al. 1997, Werren 1991). While the survival of the E. grandis and E. urophylla controls in 

the second generation highland trial was 82 and 85 percent respectively, survival was just 13 and 50 

percent in the lowland trial. This compares to an overall second generation trial survival in the 

highland trial of 88 percent and an overall trial survival in the lowland trial of 83 percent. The 

differential impact of disease was less apparent in the E. pellita controls from the lowland and 

highland sources; material selected in a lowland orchard was significantly larger than a bulk of 

selected Queensland families (9.9 cm vs 9.1 cm) and survival was 79 and 68 percent respectively 

when planted in the lowland trial. Within both second generation trials the clonal controls grew very 

well, showing significant superiority to most of the families originating from the native range. 

When comparing the first generation and second generation trials in the upland region, improvements 

in form and stem defects are visually apparent. In the first generation trial top-breakage was a 

problem with a trial average of 4.9 percent of the trees being broken. This varied greatly across 

families in the trial, ranging from 0 percent to as high as 44 percent for some families. Although 

confounded with differences between the first and second generation trial sites, the effect of selecting 

against top-breakage by removing individuals with broken tops prior to seed collection for the second 

generation appeared to be very effective. The incidence of top breakage was nearly halved in the 

second generation trial EB15A when compared to EB1 with a reduction from nearly 5 to 2.5 percent. 

The offspring of the individuals selected from within a highly damaged first generation family, which 

had 44 percent of the tops broken out, reduced to 13.3 percent top-breakage in the second generation 

trial. 

Table 5. Genetic parameters and standard errors of estimates from the latest diameter 

assessment taken in the three trials 


Heritability 0.15 (0.06) 0.33 (0.07) 0.24 (0.09) 

Genetic Correlation 0.54 (0.31) 0.49 (0.59) 

Genetic Correlation 0.79 (0.16) 

The across site heritability estimate from the three trials was 0.24 (standard error of 0.05), which was 

very similar to the average of the individual site heritability estimates detailed in table 4. The 

heritability estimates increased markedly between the first and second generation trials, indicating a 

substantial rise in the ability of differentiate between families in the second generation. The principal 

reason the across-site analysis was undertaken was to estimate genetic correlations between the trials, 

specifically between the first and second generation trials. Family rankings from the first generation 

trial, from which seed was collected for the second generation trials, did not correlate well with 

rankings in either second generation trial. The standard errors of these correlation estimates were 

large and indicated little precision in their estimation. This contrasted with the relatively high 

correlation between the two second generation trials, which were established in strikingly different 

environments. 

The assumption that an additional genetic group was required to account for the common pollen parent 

was checked by allocating second generation parents to the original provenance from which the seed 

was collected. This alternative approach generated very similar heritability estimates for both across 

and individual site analyses. However, correlations between the first and second generation trials were 

considerably lower (0.45 (±0.32) for EB1-EB15A, 0.42 (±0.61) for EB1-EB15B) while the correlation 

between the two second generation trials was similar (0.79 (±0.16)).



DISCUSSION 

Establishing a range of Eucalyptus plantations in the highlands of Sumatra has been successful for a 

range of species. Comparisons of locally produced seedlots and clones indicate that significant 

genetic improvement can be achieved relative to the E. pellita families in these trials. The utility of 

the species has increased recently following the demonstration of its resistance to a range of leaf fungi, 

which typically cause eucalypt plantations in the wet tropics to fail (Werren 1991, Harwood 1998), as 

well as the species ability to produce productive hybrids (Luo et al. 2006). While E. pellita seedlings 

may not be the primary source of material for deployment within the Sumatran highlands estate, the 

worldwide E. pellita plantation estate has expanded in recent years and derived hybrids have been 

identified by various tree improvement programs in the tropics. 

As previously reported, the most productive source in the first generation trial was from Kuranda (Luo 

et al. 2006 and Harwood 1998) with nearby Tinaroo Creek also performing well. Kuranda has been 

identified as having a particularly high outcrossing rate when compared to other natural populations 

from Cape York and Irian Jaya (House and Bell 1996). Although these trials were rather 

comprehensive in the number of sources included, many of the provenances that were compared to 

Kuranda were represented with few families and their true genetic potential may not have been 

indicated. The relative superiority of Kuranda diminished in the second generation trial, possibly 

because the growth advantage resulting from naturally high outcrossing rates was not present in the 

second generation trial where other families had equal chances of outcrossing. Low outcrossing rates 

of eucalypt species have been shown to increase within-family variation, because many selfed or 

inbred individuals perform poorly as a consequence of inbreeding depression (Eldridge et al 1993). 

Increased within-family variation would tend to bias genetic parameter estimates downwards if 

differential selfing rates are prevalent in the families included in these trials. On the other hand, related 

matings resulting in a larger number of full-sib offspring within families would tend to bias heritability 

estimates upwards (Eldridge et al 1993). 

The main focus of the genetic analysis was on the estimation of genetic parameters within the first and 

second generation trials for comparative purposes. The heritability estimate of the first generation trial 

was low and standard error was high when compared to both second generation trials. The impact of 

differential outcrossing rates on heritability estimates has been discussed at length by Eldridge et al 

(1993). The distribution of E. pellita, as reviewed by Harwood (1998), has been described as ‘heavily 

dissected’ in the Fly-Diogel shelf of New Guinea by Paijmans (1976), ‘small scattered populations’ in 

Southern Papua, ‘extensive tall open forests’ from Bupul to Muting in Irian Jaya, ‘discontinuous and 

scattered’ in ‘narrow bands’ in Northern Queensland with a ‘major disjunction’ between populations. 

Due to the dispersed occurrence of individuals within parts of the native range, it is expected that 

selfing would be prevalent in some of the families included in these trials. While an increase in 

related mating would tend to bias heritability upwards due to an increase in the true coefficient of 

relationship, a downward bias in the heritability estimate of the first generation progeny trial could be 

caused by differential levels of selfing inflating the within family variance. In addition to the 

differential selfing rates, other reasons the heritability estimate was found to be lower in the first 

generation trial than in the second generation trials could be: fewer trees per plot, reduced silvicultural 

intensity or any cause of increased experimental error. The reduced heritability in the first generation 

implies the rankings of families are not well predicted and there is therefore an expectation that 

rankings would differ in subsequent generations. 

Genetic correlations have provided an indication of the relative importance of between-generation and 

between-site differences on the ranking of families of E. pellita. Low correlation estimates between 

the first and second generation trials indicate that substantial changes in ranking occurred following 

the thinning that was used to convert the progeny trial into a seedling seed orchard. Although the 

genetic correlation between the first and second generation highland trials is effected by both the 

climatic and management differences as well as outcrossing, it was expected that the correlation 

between the highland and lowland trials would be lower than the correlation between the two highland 

trials. Nevertheless, the correlation between the two second generation trials was found to be much 

greater than the correspondence of family rankings in the first and both of the second generation trials. 

The selection of improved families for tropical highlands and deployment of these same families in the



lowlands appears to be a possible given the results of these trials. Although the statistics and genetic 

theory imply a single population would generate more genetic gain by focusing resources, the 

differences in climate and soils in the lowland and highland estate are so dramatic that intuition does 

not lead prudent tree breeders into forming a single population for deployment into these two 

environments. 

CONCLUSIONS 

The results of this study will provide some guidance for other organizations initiating tree 

improvement programs for Eucalyptus species. First generation trials of eucalypts provide a first 

indication of the relative merit of populations and families, but results from these trials should be 

treated with caution (Eldridge et al. 2003, White et al. 2007). Culling families at the early stages of 

breeding population development could eliminate individuals that would have been valuable once 

given the opportunity to outcross. Relatively simple and extensive breeding strategies that quickly 

move through the evaluation of native-range populations and do not purge poorly performing families 

may be an effective path to creating diverse populations for advanced generation breeding. 

REFERENCES 

Burdon, R.D. 1977. Genetic correlation as a concept for studying genotype-environment interaction in forest 

tree breeding. Silvae Genet. 26: 168-175. 

Dionese, J.C., Haridasan, M. and Moraes, T.S. de A. 1984 Tolerance to “mal do Rio Doce”, a major disease of 

Eucalyptus in Brazil, Tropical Pest Management 30(3), 247-252 

Doran, J.C. Williams, E.R. and Brophy, J.J. 1995. Patterns of variation in the leaf oils of Eucalyptus urophylla, 

E.pellita and E.scias. Australian Journal of Botany 43, 327-336 

Eldridge, K., Davidson, J., Harwood, C. and van Wyk, G. (1993): Eucalypt Domestication and Breeding. Oxford 

University Press. Oxford, UK. 

Falconer, D. S. and Mackay, T. F. C. (1996): Introduction to Quantitative Genetics. 4th Edition. Addison Wesley 

Longman Ltd. Essex, UK. 

Gilmour, A.R., Gogel, B.J., Cullis, B.R., and Thompson, R. 2005. ASReml User Guide Release 2.0. VSN 

International Ltd., Hemel Hempstead, HP1 1ES, UK. 267 pp. 

Harwood, C.E., 1998. Eucalyptus pellita – An annotated Bibliography. CSIOR Forestry and Forest Products, 

Canberra. 70 pp. 

Harwood, C.E., Nikles, D.G., Pomroy, P.C. and Robson, K.J. 1997. Genteic improvement of Eucalyptus pellita 

in north Queensland, Australia. Pp 219-226, Vol1 in IUFRO Converence on silviculture and improvement 

of Eucalypt, 1997, Slavador. EMBRAPA Centro Nacional de Pesquia de Florestais 

House, A.P.N. and Bell, J.C. 1996. Genetic diversity and systematic relationships in two red mahoganies, 

Eucalyptus pellita and Eucalyptus scias. Australian Journal of Botany 44, 157-174 

Luo, J. Arnold, R.J. and Aken, K. 2006. Genetic variation in growth and typhoon resistance in Eucalyptus pellita 

in south-western China. Australian Forestry. Vol 69, No 1:38-47 

Pegg, R.E., and Wang, G.X. 1994. Results of Eucalyptus pellita trials at Dongmen, China. In Brown, A.G. (ed.) 

Australian tree species research in China: Proceedings of and international workshop held at Zhangzhou, 

Fujian Province, PRC, 2_5 November 1992. ACIAR Proceedings no. 48. ACIAR Canberra, 226p 

Paijmans, K. 1976) New Guinea Vegetation. Australian National University Press, Canberra, 213 pp. 

Quass, R.L. 1988. Additive genetic model with groups and relationships. J. Dairy Sci. 71:1338-1345 

SAS Institute, Inc. 1991 SAS Language, SAS Institute Inc., Cary, NC. USA. 

Vercoe, T.K. and McDonald, M.W. 1991. Eucalyptus pellita F.Muell and Acacia seed collections in New 

Guinea, September – October 1990. Forest Genetic Resources Information 19, 38-42 

Werren, M. 1991 Eucalyptus plantation development in Indonesia. Pp 1160-1166 in AGP Schonau, (ed.) 

Symposium on Intensive Forestry: The role of Eucalypts. IUFRO and Souther African Institute of 

Forestry, Pretoria, South Africa 

Westell, R.A. Quass, R.L., and van Vleck, L.D.. 1988. Genetic groups in animal models. J. Dairy Sci, 71:1310- 

1318 

White TL, Adams WT, Neale DB (2007) Forest genetics. CABI, Wallingford, p 682 

Yamada, Y. 1962. Genotype by environment interaction and genetic correlation of the same trait under 

different environments. Jap. J. Genet. 37: 498-509.



GROWING TEAK FOR THE ROOTS – THE JIFFY SOLUTION 

ABSTRACT 

Don Willis 1 

Have you ever managed your plantation for what you do not see? What if you were 

able to produce Teak seedlings in the nursery in a shorter time period and plant out a 

younger root system that is ground-ready? 

Soft-walled pellets mimic nature offering unrestricted root growth symmetry and 

superior lateral root growth. Through air pruning and proper water management in the 

nursery, Jiffy pellets allow roots to maximize stability in Teak plantations, resulting in 

improved wood quality. Maximizing Teak seedling crop management includes: 100% 

tray stocking at anytime, easy grading for quality at anytime, shipping at anytime, and 

water conservation while simultaneously producing a long-term viable root system for 

the life of the plantation. 

The Jiffy Pellet System is a value added, quality seedling container and media in one 

solution that simplifies nursery management. In an era of environmental economics, 

businesses that lead us into the future choose superior products that lead the way 

towards zero waste. The Jiffy Pellet System is the next generation. 

1 Don Willis RPF, Global Forestry Product Manager, Jiffy Products, 125 Industrial Park, Shippagan, NB E8S 3H1, Canada, 

Ph (+1) 506 – 336 – 2284 Web: www.jiffypot.com.



LESSONS LEARNED FROM IMPLEMENTATION OF REDD – 

A COLLECTION OF ANALOGIES, METAPHORS AND CLICHÉS 

Zoe Harkin 1 

ABSTRACT 

Fauna & Flora International, together with the Macquarie Group, have partnered to 

form a ‘Carbon Forests Taskforce’ (CFT), with the aim of implementing at least six 

projects for Reducing Emissions from Deforestation and Degradation (REDD), located 

in threatened landscapes spread across south east Asia, South America and Africa. It is 

planned for avoided emissions generated from the projects to be verified and sold on 

voluntary carbon markets (i.e. markets driven by corporate social responsibility and 

speculators). It is also hoped that the resultant REDD credits will be accepted by 

international compliance-based (i.e. formal) markets for REDD, if and when they 

become operational. 

The paper provides a summary of observations and lessons learned by the CFT Forest 

Carbon Specialist, following one year’s experience in the implementation of REDD. 

To facilitate a common understanding of the sometimes complex concepts associated 

with REDD, the paper is presented as a series of analogies, metaphors and clichés. 

REDD: Calm before the storm... but clearly the next gold rush 

In his 2006 ‘Review of the Economics on Climate Change’, the head of the UK Economic Service, 

Lord Nicholas Stern, proclaimed that “curbing deforestation is a highly cost-effective way of reducing 

greenhouse gas emissions.” Stern’s economic models suggested that significant quantities of REDD 

credits could be developed for less than USD 5 per tonne of CO2 (compared with marginal abatement 

costs for Carbon, Capture and Storage (CCS) of up to USD 270 per tonne CO2). 

Stern’s conclusions grabbed the attention of the developed nations of the world - hungry for cheap, 

proven ‘technologies’ that could be mobilised rapidly to mitigate climate change. REDD had the 

added appeal of aligning with the objectives of other international treaties such as the UN Human 

Development goals, Convention to Combat Desertification, and the Convention on Biological 

Diversity. Likewise, forested developing nations, long unable to reign in their rampant deforestation 

rates, saw REDD as a mechanism to attract compensation for efforts to reduce deforestation, and for 

what they perceived to be foregone development opportunities. 

In the international policy arena, REDD came into the international spotlight at the 13 th Conference of 

the Parties (COP) to the United Nations Framework Convention on Climate Change (UNFCCC), held 

in Bali in December 2007. Here, the ‘Bali Roadmap’ paved the way for REDD to be included within 

a post-Kyoto international climate change agreement, to be implemented from 2013 onwards. By all 

appearances, REDD had the unanimous support of all signatory countries to the UNFCCC, including 

particularly strong support from Australia. 

This international in-principle endorsement of REDD spawned a flurry of activity in the budding 

voluntary carbon market. Environmental groups rushed to stake out their claim on areas of forest 

considered at risk of conversion to other land uses. Investors looked to partner with the 

environmental groups, in the hope of securing cheap credits ahead of the rush, and perhaps hoping to 

polish their corporate images at the same time. The Governments of forested developing countries 

looked on, sensing a significant opportunity while at the same time wary of being ripped off, yet 

again, in a market they did not fully understand. The World Bank, the United Nations Collaborative 

1 Dr Zoe Harkin, Fauna & Flora International. Ph: 03 9416 5220 Email: zoe.harkin@fauna-flora.org.



Programme on Reducing Emissions from Deforestation and Forest Degradation in Developing 

Countries (UN-REDD) and other donor countries formed massive funds to help fund REDD 

‘readiness’ activities. And Indigenous Peoples human rights activists grew increasingly weary of 

waiting for the aforementioned REDD proponents to consult with them. And so the ‘REDD gold 

rush’ was born.... 

At the time of writing, this ‘REDD gold rush’ is in a state of calm before the storm, awaiting formal 

endorsement of REDD at COP 15 in Copenhagen in December this year. If the draft negotiating text 

for the meeting is any indication, it appears almost certain that REDD will be included within the 

post-2012 UNFCCC agreement. If and when this occurs, carbon market analysts suggest that 

demand, and subsequently prices for REDD credits, will increase rapidly. 

REDD: the ultimate horse/cart, chicken/egg problem 

In our numerous presentations on REDD, we have used linear or circular process diagrams, such as 

the one depicted below, in order to explain the ‘REDD implementation process’. 

Figure 1 Schematic depicting the REDD implementation process 

Unfortunately this schematic is an over-simplification. Implementation of REDD is not quite the 

linear process we had envisioned. For example: in commencing consultation with local stakeholders 

and indigenous groups, one of the first things they want to know is an indication of the benefits they 

might receive from the proposed REDD project. In order to develop a reasonable estimate of the 

benefits of the REDD project, it is necessary to conduct a forest carbon inventory. However investors 

may be reluctant to provide a significant investment in a carbon inventory, unless they have the inprinciple 

support of the local communities and Government, which can only be acquired following 

extensive consultation. Furthermore, under existing standards in the voluntary carbon market, the 

REDD project cannot be verified unless the project developer has secure tenure over the site. 

Securing tenure can be an expensive business, particularly if this involves buying out existing 

commercial stakeholders or paying hefty application fees. Therefore an investor might reasonably 

want assurance that the project is verifiable before going to the considerable effort and expense of 

securing tenure... and so goes the REDD chicken-egg-horse-cart merry-go-round. 

Ultimately, it appears that REDD is an integrated assessment in the truest sense of the term. It 

requires pursuit of all facets of the project – carbon inventory, community consultation, government



endorsement and tenure security – at the same time. The project will fail if just one of these aspects is 

not achieved, and the inter-relationship between all facets largely precludes a sequential, linear 

approach. 

Forecasting baselines: smoke and mirrors, or clear as day? 

In order to quantify avoided emissions due to implementation of REDD, current accounting rules in 

the voluntary carbon market typically require the project developer to forecast the forest carbon 

profile under the project (or REDD intervention) scenario. This is then compared to a forecast of the 

forest carbon profile under the baseline (or ‘business-as-usual’) scenario (Figure 2). 

Figure 2 Forecast of baseline and project scenarios to estimate avoided emissions 

This rule defines eligible avoided emissions are those that occur in excess of ‘business as usual’ 

activity, which are said to be ‘additional’. The intention of this accounting rule is to “incentivise” 

genuine behavioural changes, rather than provide credits for activities that would have occurred in the 

absence of carbon markets anyway. 

During the very early stages of REDD market development, forecasting of the baseline carbon profile 

could be described as little more than crystal ball gazing. Under this ‘smoke and mirrors’ approach, a 

few references to the scientific literature, supplemented with some commentary on the deforestation 

drivers affecting the site, appeared to suffice. As the voluntary carbon market has matured, however, 

the technically prescriptive standards such as the Voluntary Carbon Standard (VCS) are requiring 

baseline forecasts to be developed using more advanced methodologies. These methodologies are 

tailored to different baseline scenarios, depending on whether the baseline deforestation/degradation 

is planned (i.e. sanctioned by the government, such as oil palm conversion licences); or unplanned 

(i.e. unsanctioned by the government, such as expansion of slash and burn agriculture, clearing for 

grazing, or illegal logging). 

In cases of planned deforestation, the REDD project developer is now required to collate an extensive 

audit trail including documentation such as regional development plans, zoning maps, management 

plans and licences. This documentation needs to convince the auditors of the exact time and location 

that the deforestation or degradation event was planned to have occurred. In the case of unplanned 

deforestation, the methodologies required to forecast the rate and location of deforestation are far 

more complex. These typically require the use of GIS based software to calculate the historical rate of 

land use conversion in a suitable reference (or analogous) area, and then correlate this conversion rate 

with specific indicator variables such as distance to roads, population centres, sawmills, nearest 

distance to clearing, etc. 

Regardless of which methodology is used to forecast baselines, auditors are demanding more detailed 

analysis, documentation and support from the scientific literature in order to approve baseline



forecasts. It is also important to keep in mind that it may eventuate that baselines are not required in 

future, if and when REDD tends towards national level accounting. This is because avoided 

emissions would be sold relative to the collective national reference level, regardless of additionality 

at the project level. 

Additionality is a double-edged sword 

As described above, additionality refers to the requirement that eligible avoided emissions are those 

that would have occurred in excess of business-as-usual activity. This additionality requirement is 

currently included in almost all forest carbon standards on the voluntary carbon market today 2 . 

In order to demonstrate additionality, the project developer is required to show proof that the forest 

was under imminent threat of deforestation or degradation. Since forward crediting is not allowed 

under the VCS, then REDD credits can only be claimed at the time at which the deforestation or 

degradation would have occurred. The combined effect of these rules is that forests considered most 

at risk of deforestation or degradation are most attractive as a REDD project. 

This approach has obvious advantages. Primarily, the additionality rule attracts REDD investment to 

the forests that are most at threat. On the flipside, forests that are most at threat are often the hardest 

(and arguably the most expensive) to save. Intervening just prior to deforestation is likely to require 

compensation of a whole suite of stakeholders, all of whom have invested significant time and energy 

in pursuit of the business-as-usual activity. The more advanced the stage of this development, the 

more costs that are likely to have been incurred, and the more committed these parties are to the 

activity. Averting this course of action ‘at the last minute’ can be extremely tricky, and potentially 

quite expensive. 

REDD project developers clearly need to weigh up the pros and cons of this ‘double edged sword’ 

phenomenon – assessing the risk of investing in projects with strong additionality, against the 

attractiveness of the more immediate revenue stream that can be delivered from such projects. 

Unravelling tenure issues is like peeling the layers of an onion 

Lack of clarity on forest tenure is often cited as one of the fundamental drivers of deforestation or 

degradation. This is because uncertainty around who is eligible to do what in the forest creates a 

‘tragedy of the commons’ scenario. Each stakeholder ultimately feels that if they don’t utilise the 

forest to their maximum utility, then someone else will. The negative impacts of the 

deforestation/degradation event will be shared to some extent by all, and particularly those living 

closest to the forest. Therefore the marginal utility of deforestation for an individual is generally 

greater than the negative impacts they might experience, even if the overall negative impacts shared 

by society exceed this benefit. 

Lack of clarity around forest tenure in many developing countries can be due to an ‘onion-like’ 

layering of land rights and use claims. The uppermost layer might typically involve some sort of legal 

permit to clear the land. However multiple levels of government, and multiple ministries within the 

same government, can lead to a situation where the same piece of forest land has been licensed to 

multiple parties for overlapping, and often conflicting, land uses. A common example is approval of a 

mine on an area also approved for commercial forestry. In part, the ‘onion-like’ land tenure situation 

arises due to corruption, whereby government officials receive financial benefits from each successive 

licence approval. The government official may have little interest in which party ultimately benefits 

from their licence. If the multiple licensees are aware of each other’s presence, this can lead to a ‘race 

against time’ to undertake their licensed activities before the other licensees. 

Beneath these government-endorsed tenure arrangements often lie customary or historical land use 

claims. These may or may not be formally recognised by the government. And the customary land 

use claims can also overlap, for example where different indigenous groups or local communities 

claim ownership or customary use of the same area of land, or when the ’bundle’ of forest use rights 

is divided between or shared amongst multiple forest users. This can be particularly complex where 

indigenous groups have been displaced, and migrants may have moved in to the area. Such events 

2 With the exception of the Chicago Climate Exchange.



may have occurred many years in the past, however the indigenous group may still have historical 

links to the area, while the ‘newer’ migrants feel a sense of ownership of the forest because of their 

daily interaction with and dependence on the forest. 

This ‘onion-like’ layering of tenure issues often leads to a hierarchy of drivers of deforestation. If the 

primary driver of deforestation is addressed, then the secondary drivers of deforestation may become 

dominant. This tenure situation is similar to Liebig’s ‘barrel analogy’ of forest nutrition (Justus von 

Liebig (1803-1873), whereby forest growth is theoretically constrained by the most limiting factor. 

Likewise, deforestation is determined by the most dominant driver. 

What this means is that implementation of REDD requires a holistic approach, where each successive 

driver of deforestation is identified, and ultimately addressed. 

The ‘baby out with the bathwater’ problem 

The commercial viability of REDD is highly sensitive to assumptions around the price of carbon. 

During these early stages of REDD market development, the assumed carbon price is quite low. As a 

result, marginal increases in project development costs can quickly make a potential REDD project 

unviable, or considered too risky. This creates a ‘baby out with bathwater’ problem – if the cost of 

developing a REDD project becomes too high, then many good prospective REDD projects will be 

abandoned before they even start. 

Alongside this cost sensitivity, the proliferation of technologies to measure and monitor forest carbon 

are growing in sophistication (and cost). It is currently unclear as to which of these methodologies and 

technologies will be required for implementation of REDD, either in the voluntary or compliance 

market. In the face of this uncertainty, a prudent REDD project developer might seek to adopt ‘best 

practices’ carbon inventory and baseline development practices, in order to reduce the risk that the 

project is considered non-compliant when REDD methodologies are clarified. This might involve 

utilisation of advanced remote sensing technologies, complex GIS-based land use change forecasting 

algorithms, combined with a high density of field sample plots. There is a danger that the cost of such 

‘best practices’ methodologies may become prohibitive, if they become a requirement for all REDD 

projects. A more sensible approach might be for project developers to commit to ‘continuous 

improvement’; developing procedures for retrospective back-calculation of avoided emissions, thus 

allowing the project developer to apply the best practice technologies consistently across the time 

series once they can be afforded. 

This ‘baby out with bathwater’ problem has been recognised by a number of stakeholders involved in 

REDD methodology development. For example, a US-based group, Avoided Deforestation Partners, 

adopted a ‘de minimus’ approach in development of their methodologies for submission to the VCS. 

This is based on the principle that REDD methodologies should be based on the minimum effort 

required in order to achieve best practices. 

Robbing Peter to pay Paul: the leakage problem 

One of the major criticisms of REDD has been its potential to cause ‘leakage’ issues. This relates to 

the concern that attempts to reduce deforestation or degradation in one area, may simply displace 

these activities to another area, resulting in little or no net benefit to the atmosphere. In effect, poorly 

designed REDD projects could effectively be ‘robbing Peter to pay Paul’. 

There are two main solutions to the leakage problem. The first is to conduct carbon accounting on a 

national basis, thereby capturing all leakage effects within a single national reporting system. The 

second is to address the fundamental drivers of deforestation, rather than simply addressing the 

symptoms. The example most relevant to foresters is timber supply. Timber is extracted for two 

reasons: 1) as a source of income generation for those extracting it; and 2) as a supply of wood 

products for paper, construction, furniture and other uses. If a REDD project cuts off timber supply 

without providing alternative income sources for the local timber industry, or without implementing 

demand-side abatement measures, then the project will likely result in leakage. 

Within reason, there is a danger that demand-side abatement measures could quickly result in perverse 

greenhouse outcomes – if low greenhouse intensity wood products were displaced by more



greenhouse intensive substitutes such as steel, aluminium or concrete, then the REDD project could 

ultimately result in more greenhouse gas emissions than it avoided. 

Attempts to find alternative income sources for the local timber industry are not always readily 

identifiable; do not necessarily guarantee a reduction in pressure on the forest; and local communities 

may resent any project that severely restricts their access to the forest. As a result, in many cases it 

appears that implementation of some form of sustainable forestry regime may be one of the best ways 

to address leakage issues. It allows a continued supply of timber from the forest, provides a continued 

stream of income for local people, and adoption of reduced impact logging techniques can avoid 

significant greenhouse gas emissions when compared with conventional logging. In this way, REDD 

is likely to become an important tool in achieving sustainable forest management objectives, 

particularly if coupled with certification and measures to reduce illegal logging. 

Light at the end of the tunnel... 

From the stream of issues and problems presented above, one might reasonably surmise that the 

obstacles involved in implementation of REDD are insurmountable. However, a quick ‘back of 

envelope’ calculation on avoided forest emissions reveals why many investors are interested in 

REDD. The numbers on avoided emissions multiply very quickly, creating a ‘light at the end of the 

tunnel’ that is sure to reward project developers and investors with the persistence and tenacity to 

overcome the numerous problems that REDD presents. Plus the level of international enthusiasm for 

implementation of REDD is astounding. Very rarely does a proposal get such unanimous agreement 

from all countries in the world. 

Finally, the ability for REDD to assist in addressing a suite of other global issues that have long 

eluded policy-makers, such as global poverty alleviation, biodiversity conservation, sustainable timber 

production and agriculture, causes one to resort to a final cliché: REDD is a true a win-win (win, win, 

win....) situation.



DEVELOPING A REDD SCHEME FOR POST 2012: THE 

KALIMANTAN FORESTS AND CLIMATE PARTNERSHIP 

Grahame Applegate 1 

ABSTRACT 

The Kalimantan Forests and Climate Partnership (KFCP) is a demonstration for reducing 

greenhouse gas (GHG) emissions from avoided deforestation and forest degradation 

(REDD), with a focus on peat swamp forests in Indonesia. Incentive-based approaches 

will be adopted that can help pave the way for large scale financing of avoided 

deforestation and degradation with methodologies provided for the UNFCCC 

negotiations for a post 2012 climate change agenda. KFCP is identifying the enabling 

conditions for supporting environmental governance and institutional requirements, 

incentives, payment mechanisms and processes which link national, provincial, district 

and village level payments to performance based outcomes. KFCP is also developing and 

implementing appropriate systems for peat swamp forest and GHG measurements and 

accounting at the project and national level, taking account of additionality, leakage and 

permanence. KFCP is linked to the national carbon accounting system being developed 

by the Government of Indonesia. 

INTRODUCTION 

Peatlands cover 3 per cent of earth’s land area and store a large fraction of the world’s terrestrial 

carbon; up to 528 000 million tonnes, equivalent to 70 times the current annual global emissions from 

burning of fossil fuel (Hooijer et al. 2006). Peat consists of plant debris which has accumulated over 

thousands of years in water-logged, acidic conditions that prevent it from decomposing and releasing 

the stored carbon into the atmosphere. Peatlands emit around 800 million t CO2 annually, three 

quarters of which are from South-East Asia (Parish et al. 2007). 

Indonesia has around 22.5 million ha of peatlands, (Page and Banks 2007) which store 55± 10 Gt of 

carbon (Jaenicke et al. 2008). The carbon stored in below ground biomass (peat) is 18.6 times higher 

than intact Indonesian peat swamp forests which have an above ground carbon content of 140.5 t C /ha 

(Uryu et al.2008). The peat swamp forests are the habitat of a large range of plants and animals, 

including endangered species such as orangutans. Over 26% of peatland in Indonesia is found on the 

island of Borneo, where 3.1 million ha, much of it originally covered in peat swamp forests, is found 

in Central Kalimantan (Hooijer et al.2006). 

Tropical peatlands are increasingly being cleared and drained for logging, farming and plantations of 

oil palm and fast growing tree plantations, with the result that almost half of Indonesia’s peat swamp 

forests have been converted for other uses (Hooijer et al. 2006). Clearing and draining peatlands 

speeds up the decomposition of the peat and exposes it to fire, both of which result in large amounts of 

carbon dioxide emissions. Hooijer et al. (2006) estimated that annual emissions from deforestation and 

burning of tropical peatlands are 2 billion tonnes CO2, with over 90% coming from Indonesia, the 

country’s single largest source of greenhouse gas emissions. Fires in peatlands were a major cause of 

CO2 emissions resulting from the burning of 2.1 million ha of peatland during the large fires in 1997- 

98 that blanketed large parts of South-East Asia in haze (Page et al. 2002). 

Peat Swamp Forests and Factors Causing Deforestation and Degradation 

As shown in Figure 1, rivers often demarcate the boundaries of large continuous areas of peat. 

Furthermore, each continuous area of peat has its own hydrological unit, and can form a ‘dome’ up to 

fifty km in diameter (see Figure 2). 

1 

Consultant, AusAID, Australian Embassy, Jl. H.R. Rasuna Said, Kav. C15-16, Jakarta, 12940, Indonesia. Email 

grahame.applegate@iafcp.or.id



The Ex Mega Rice Project Master Plan (Euroconsult et al. 2008) provides a strong argument for 

managing the peatlands at a landscape level and taking a “whole-of-dome” approach based on 

hydrological boundaries. This approach and understanding is fundamental to developing solutions for 

managing peat to reduce GHG emissions from avoided deforestation and degradation. It is important 

to understand that affecting a change in water level in one area of the dome will impact over time on 

the remainder of the dome. Draining areas of shallow peat containing forest on the edge of the dome, 

for example for agricultural purposes, may lower the water table in the central part of the dome, which 

may result in drying-out areas of intact forest (making them susceptible to fire) and possibly dieback 

due to a lack of water. Although further research is required to quantify these relationships, there is 

evidence that the construction of canals to drain peat swamp forest areas in Indonesia has adversely 

impacted on them and continues to impact a much larger area in which the canals are located 

(Euroconsult et al. 2008). 

Source: Bappenas- National Strategy and Action Plan for Sustainable Management of Peatlands, 2006 

Figure 1. Cross section of a peat dome showing Peat Swamp Forest 

While additional research is required on these phenomena, the risk of not taking a whole-of-dome 

approach to avoiding deforestation or degradation is too high, and could jeopardise the success and 

longevity of any interventions designed to maintain forest cover. 

Restoring and Protecting Peat Swamp Forests 

Most of the peat swamp forests in Indonesia and elsewhere in S.E Asia were cleared by first 

constructing large canals which intersected across the ‘peat dome’ and drained the water to adjacent 

streams. This then allowed access to the area for logging and subsequent clearing and burning. Many 

of the peat swamp forests which were logged but not cleared, also had canals constructed for the 

extraction of the logs. These canals were usually small and excavated by chain saw and were 

particularly prevalent for illegal logging activities in logged-over or expired timber production 

concessions. See Figure 2. 

One of the best known examples of peatland degradation was the ‘Mega Rice Project’ in Central 

Kalimantan, which cleared 1 million ha of peat swamp forest for rice production in the mid 1990s. The 

peatlands were found to be unsuitable for growing rice and the project was abandoned resulting in 

large areas of degraded and exposed peatland which continue to emit large quantities of GHGs from 

decomposition and annual fires. The Indonesian Government is now committed to restoring these 

deforested and degraded peat lands (Presidential Instruction 2007). 

In order to maintain existing peat swamp forests requires blocking the canals to further restrict the 

drainage and the oxidation and to protect the forests from further fire and exploitation. The degraded 

peat swamp forests can be hydrologically restored by raising the water table back to original levels (by 

blocking the canals previously used to drain them) and reintroducing vegetation to hasten the 

rehabilitation and hydrological restoration process.



CO2 

Canals 

Source: Delft Hydraulics 

Figure 2. Drainage of the peat dome 

River 

In order to maintain existing peat swamp forests requires blocking the canals to further restrict the 

drainage and the oxidation and to protect the forests from further fire and exploitation. The degraded 

peat swamp forests can be hydrologically restored by raising the water table back to original levels (by 

blocking the canals previously used to drain them) and reintroducing vegetation to hasten the 

rehabilitation and hydrological restoration process. 

INTRODUCTION TO KALIMANTAN FORESTS AND CLIMATE PARTNERSHIP 

The International Forest Carbon Initiative (IFCI) is Australia’ contribution to the global effort on 

reduced emissions from avoided deforestation and degradation (REDD). The IFCI aims to 

demonstrate that REDD can be part of equitable and effective global climate change activities, with a 

central element of this work developing REDD demonstration activities with Indonesia. The 

Kalimantan Forests and Climate Partnership (KFCP) is the first demonstration activity being 

implemented under IFCI through the Indonesia- Australia Forest Carbon Partnership (IAFCP), part of 

the bilateral cooperation between Australia and Indonesia. 

Goal 

The goal of the KFCP is: 

• to demonstrate a credible, equitable, and effective approach to reducing greenhouse gas 

emissions from deforestation and forest degradation, including from the degradation of peat 

swamp forest lands, 

• inform a future international climate change framework, and 

• enable Indonesia’s meaningful participation in future international carbon markets. 

Location and Site Description 

The REDD demonstration site represents a complete dome and covers an area of approximately 

120,000 hectares in the northern part of the Ex Mega Rice Project Area (Euroconsult et al. 2008) area 

in Central Kalimantan (centre is approximately 2 o South and 115 o East). See Figure 3. The site is 

bordered by the Kapuas River to the west and south-west, and the Mantangai River to the east and 

south east. The demonstration site lies completely within Kapuas District. Kapuas District is 

administered from Kuala Kapuas, some 100 km by river to the south. 

Demography 

The area has a low population density relative to much of Indonesia, with most villages located along 

the Kapuas River. Around a dozen villages are likely to have some connection (either currently or in 

the past) with the demonstration site, primarily for the use of natural resources from the forest or river 

systems, harvesting rattan or use of agricultural land. The villages primarily contain members of the 

local populations of Dayak. There are no plans for families to be settled in the area under Indonesia’s 

transmigration scheme.



Land use 

Other than community land, which consists of a strip of land (3-5 km) along the river adjacent to each 

village, the vast majority of the site is government owned land, under control of the Ministry of 

Forestry, but administered by the Province. 

The northern half of the demonstration site is heavily forested which was logged using elevated 

railways and the ‘kuda kuda’ system 2 . While there are some areas of remaining forest in the southern 

half of the site, the majority of this area was drained using an intensive network of canals, and has 

largely been deforested and burnt. 

The land adjacent to the Kapuas River is used by communities for agricultural activities, including 

food crops and commercial rubber. Fishing is a major activity for both income and subsistence 

purposes, (including a range of native fish species and freshwater prawns). Forest areas are similarly 

important for a range of commercial products, including jelutung and gemor, and a range of 

subsistence products used traditionally for buildings, food (both plants and animals), medicines and 

handicrafts. 

Illegal logging occurs across the site, although the extent and severity has reduced over recent years. 

Illegal logging is likely to be undertaken by people from both the local communities and others from 

further a field (most likely further downstream, where there is an abundance of small to medium scale 

sawmilling operations). 

Social characteristics 

Although the Kapuas River is used as the main transport route in the area, the impacted villages are 

relatively remote and have limited public infrastructure, including power supply. The level of access to 

and quality of both health and education services is relatively poor. The isolation also limits the range 

of economic opportunities available, and the deforestation of large areas of land since 1996 has greatly 

reduced these opportunities. Although specific information is not available on the incidence of poverty 

in the impacted villages, there is a high incidence of poverty in peat areas in general within Central 

Kalimantan. 

Ex-Mega Rice Project Central Kalimantan 

Figure 3. Location of the Kalimantan Forests and Climate Partnership demonstration area 

in Central Kalimantan 

2 This is a logging system used in peat swamp forests in SE Asia which involves logs being manually hauled in sledges on 

wooden skids constructed on the forest floor. The logs are usually manually skidded to small rail heads or river landings 

from where they are transported to the processing mills.



KFCP IMPLEMENTATION STRATEGY AND COMPONENTS 

KFCP is designed to implement four components to achieve the goal. These are: emissions reduced 

from avoiding forest /peatland deforestation and degradation includingcapturing and communicating 

knowledge from this REDD demonstration; forest and GHG baseline measurement and monitoring; 

developing and implementing effective incentive based payment mechanisms; and REDD 

management, including capacity building and readiness at the provincial and district and village levels. 

Component 1: Emissions Reduced from Avoided Deforestation and Degradation 

This is a key component of KFCP activities and provides a framework in which the other components 

are integrated. The community based activities implemented at the village level are designed to be 

coordinated with the information dissemination of REDD at the village level, changes in behaviour, 

peatswamp forest restoration and hydrological restoration; GHG monitoring and incentive based 

payment mechanisms. As the REDD is a new concept with many new ideas being introduced and 

tested, it will bring with it a high level of risk from the social and political and technical perspective. 

However it does provide an opportunity to significantly improve tropical forest management 

(Applegate and Smith 2000). 

Community engagement: Essential to the success of REDD, is the involvement of the local villages, 

as gaining support from these communities is a precondition for emissions reduction from the peat 

swamp forests. It will be essential that the local people have the ability to replace lost income from the 

limitations placed on them on the destructive use of forest resources under some form of tenure and 

land use right on village lands. Hence, the village engagement is designed to be flexible, participatory 

and provide for informed consent and adaptive in nature. It will ensure that improved livelihoods are 

compatible with REDD and offer real income opportunities which do not exacerbate gender and social 

differences. The activities will be planned within the Government of Indonesia village level planning 

process, which enable these plans to be integrated into the higher levels of government spatial 

planning. 

Peat swamp forest rehabilitation and hydrologic restoration: The hydrologic restoration process is 

based on the whole-of-dome approach as the protection strategy. An intact patch of forest will be 

severely impacted in both the short and long term and become degraded and eventually destroyed if 

water levels are not maintained by blocking any canals located downstream and rehabilitating the 

degraded forest areas. The dams them selves will not raise levels sufficiently in the short term to 

inundate the higher parts of the peat dome, but this first step is essential to stop further drying of peat 

in the vicinity of the canals, and therefore reducing the rate of GHG emissions. Dams along the canals 

will assist in reducing access to areas by people which also has the added benefit of reducing the risk 

of fire and by increasing moisture levels of the peat close to the canals. Rehabilitation of degraded 

peat swamp forests down stream from intact forests, either by promoting natural regeneration or 

replanting with appropriate species, will ‘kick-start’ the ecological processes essential for keeping the 

peat surface moist and reducing water runoff from the upper sections of the dome. Reforestation 

(natural or artificial) in this context is not carried out for the sake of establishing more trees in the 

traditional sense of reforestation, but is essential for controlling water flows and starting the ecological 

process essential for protecting the remaining forests upstream. The same rationale exists for 

improving livelihoods of local communities as part of a fire prevention strategy designed to protect the 

forests and reduce GHG emissions. Fire is one of the main causes of deforestation and degradation, so 

preventing fires from starting in degraded forests has a major influence on determining the integrity of 

the remaining peat swamp forest. 

The rehabilitation of the degraded peat swamp forest will involve planting on a demonstration basis to 

test a new range of suitable tree species . This will be scaled up to a size which allows techniques and 

procedures for measuring and monitoring the impacts to be credible. Rehabilitation will demonstrate 

various approaches (e.g., full planting, partial planting, natural regeneration) for reforesting near the 

central part of the dome in areas where land tenure is not in dispute. Orangutan food species and those 

appropriate for forest protection and conservation will be planted and promoted near the centre of the 

dome and species with non-timber values to be utilised by the communities towards the edge closer to 

their villages. Community groups or individuals will be encouraged to develop nurseries with 

technical assistance, and sell seedlings for the rehabilitation activity. This commerce will provide an



opportunity for women to benefit financially and enhance skills in the early stages of the 

demonstration. Local people will be hired (as part of the incentive mechanism) to plant trees, either 

directly, or by arrangement with implementing partners or private contractors. Trees must be planted 

to coincide with dam construction to enable access for planting material and people. A rehabilitation 

and strategy will guide the planning process for the rehabilitation which will focus on promoting 

natural regeneration as well as replanting where required. The trees may have a direct economics value 

as non-timber forest products on designated community land, food source for orang-utans in the 

Protection Forests, as well as forming ‘dams’ on the peat to reduce run-off and thus keep the peat from 

drying out and oxidising. 

Construction of appropriate dams on the canals could reduce emissions quickly by blocking access and 

reducing the water draining from the site and stabilizing the water table. The basic approach is to start 

dam construction at the centre of the dome and work outwards, spacing dams relatively closely (20 cm 

-50 cm vertical elevation) to avoid putting too much hydraulic pressure on each dam (this has caused 

dam failure in the past). Options for developing dams at different vertical intervals and cost 

implications will be part of the rehabilitation and restoration strategy. Methods for constructing more 

cost-efficient dams will be explored including ways of achieving economies of scale through bulk 

purchase and transport of materials and use of contractors in agreed areas. 

Livelihood interventions and incentives: One of the key aspects of a successful REDD mechanism 

will include the development of appropriate incentives to adopt sustainable land-use practices. These 

incentives and how the mechanisms will function are a key part of the overall strategy to mitigate 

against the further deforestation and degradation, as farming and land use preparation methods depend 

on fire for land clearing and preparation. Incentives to encourage sustainable practices will 

include: 1) input-based payments, which comprise payment or other direct benefits for adopting and 

undertaking interventions, such as dam construction, tree planting, provision of dam-building and treeplanting 

materials, or fire suppression; 2) performance-based payments for maintaining the 

interventions so as to achieve the desired results, such as maintaining dams in order to keep water 

levels high, protecting forest from encroachment, or reducing the incidence and extent of fire; 3) 

outcome-based payments commensurate with the level of GHG emissions reductions. These will 

initially be a proxy for a future forest carbon market which will later be based on tradeable credits in a 

real market. 

Measurement and monitoring: In order to effectively monitor the impact of the interventions on the 

reducing emissions and reducing the loss of peat swamp forests and their degradation, initial 

measurements will be undertaken on a number of parameters prior to any interventions in order for 

monitoring and evaluation to have a sound basis. These measurements will be also used to determine 

the various baselines and current emission levels and Reference Emission Level (REL). It is important 

to recognise what parameters need to be measured prior to the commencement of the interventions. 

For example, social and village baselines and GHGs emissions for determining the REL from the 

current area of the peat swamp forests in the KFCP area is well underway and needs to be carried out 

before interventions are implemented. In terms of the REL, KFCP will determine the carbon content 

of the various components of above and below ground component of the peat swamp forests and the 

methodologies by which the changes in avoided emissions and changes in forest quality and area will 

be estimated. In terms of GHGs, the KFCP will undertake a scanning Light Detection and Ranging 

(LIDAR)study prior to the interventions to determine the elevation of the peat in the peat swamp 

forests, but will then need to undertake research into carbon content of the below (peat) and above 

ground forest biomass (no data at present), carbon content of peat at different depths and position in 

the dome (determined by the forest type when peat was developed), bulk density which varies with 

depth and position in the dome, size of the project area and peat depth, in order to determine estimate 

avoided emissions from the interventions and the impact of the interventions. 

KFCP as a demonstration activity is designed to capture the knowledge and lessons learned and 

providing this information to a range of stakeholders and policy makers. The audience for this 

information are local communities, the Government of Indonesia and Australia and the international 

community such as the UNFCCC. These groups which will be kept informed by implementation of 

both a communications strategy and a knowledge capture strategy. The work will include



identification of the KFCP audiences, their information needs and the best media for communicating 

with them. The strategy to capture knowledge will be developed to collect, and store the information 

coming from multiple sources within and outside the KFCP and then disseminate the “lessons learned” 

and other knowledge-based results to target audiences. 

Component 2: Emissions Estimation and Monitoring 

The current international agreements concerned with forest carbon emissions do not cover reductions 

of emissions from deforested and degraded peat swamp forest, so KFCP will endeavour to provide 

information on the methodologies to estimate and monitor emissions from peat and peat swamp forests 

to enable them to be included in a future climate change agreement, and whether it would be 

incorporated into future action on REDD. Hence this work will be focused improving and informing 

the international debate and providing lessons learned through contribution to UNFCCC discussions 

through 2009, in areas such as: research required to develop the methodologies required for estimating 

changes in GHG emissions from land use changes and measurement and monitoring of peatland 

characteristics, GHG emissions. To facilitate this process, the KFCP has convened a group of 

specialists for establishing a Project (KFCP) specific emissions baseline (REL) and determine 

scientific methodologies and implementation guidelines for the measurement and monitoring changes 

in peat swamp forests. 

To facilitate the implementation of this component, the work is divided into two distinct, but 

interrelated tasks: 

• developing, testing, and validating a GHG emissions baseline and monitoring system that 

estimates emissions changes against interventions and will be accepted as scientifically 

valid in climate change negotiations (i.e. methods to be between Tier 2 and Tier 3); and 

• operationalise GHG estimating and monitoring through remote sensing and direct ground 

measurement in ways that will meet the requirements of a future REDD carbon market 

and can be integrated into the Indonesian National Carbon Accounting System (NCASI). 

Component 3: Practical and Effective REDD GHG Payment Mechanisms 

One of the crucial aspects of a successful compliance based REDD scheme is the development of an 

equitable and effective payment mechanism for the potential REDD credits. The KFCP provides the 

opportunity and flexibility to test different approaches, guided by experience with payment 

mechanisms for other environmental services globally and other services within Indonesia. Testing 

payment systems will therefore benefit from having an overarching mechanism to pay for emission 

reduction incentives and have to be closely linked to monitoring of GHG emissions and socioeconomic 

impact in order to verify reductions, create a system that is credible, and inform the 

development of NCASI. 

The KFCP will demonstrate payment mechanisms, and in doing so, will inform the 

development of a national REDD system by; testing accountable, transparent and equitable 

payment mechanisms that create positive incentives for achieving emissions reductions; 

calculates the cost involved in implementation, impact mitigation, monitoring and regulating 

(including nominal costs per tonne of CO2), and compares whether KFCP implementation costs are 

lower than opportunity costs, the potential market value of the emission reductions it has realised 

(viability) and develops arrangements for revenue allocation between all stakeholders that takes 

sufficient account of equity considerations as well as capacity to address causes of deforestation. The 

mechanism will also involve keeping a balance between cost recovery and an attractive investment 

climate for REDD activities with effective incentives to ensure that communities are motivated to 

implement and a change in behavior that supports REDD. 

Component 4: REDD Management/Technical Capacity and Readiness Developed 

KFCP will integrate REDD into governance arrangements at the province and district levels by 

assisting to develop management institutions, a legal framework, and provide a technical capacity to 

support demonstration activities which will allow local integration into a REDD carbon market. The 

current levels of support at the province level for a REDD Working Group will be further supported at



the district level through consultation and awareness raising. The Ministry of Forestry and the EMRP 

Master Plan support the development of management units based on hydrological characteristics. The 

Forest Management Unit concept which would encompass the KFCP site provides a promising way to 

improve the institutional and technical capacity within the district to manage REDD and improve land 

management. 

CHALLENGES AND NEXT STEPS 

The multidimensional partnerships involved in a REDD demonstration activity such as KFCP is a 

management challenge. This becomes even more interesting by the fact that KFCP is essentially a 

community development project superimposed by a science and learning, socialization, policy 

development, and multiple audience communications project. In addition, there are the ever present 

surrounding international issues; additionality where REDD will be required to reduce deforestation 

below a “business as usual” baseline; leakage which involves avoiding increased deforestation 

elsewhere in the region as a result of REDD, and permanence, which involves maintaining GHG 

reductions for many years. 

KFCP is also aware that as REDD is new to many, we are on a steep learning curve, so there is a need 

to share with others what is learned. The REDD process does not work in isolation and needs to 

consider a holistic approach, with the host (community, national and sub-national governments and 

other stakeholders) leading the process. Finally the international community may be able to assist in 

the process as the outcome is designed to inform further UNFCCC processes leading to a post 2012 

climate change agenda. 

ACKNOWLEDGEMENT 

This paper is based on the Design for the Kalimantan Forests and Climate Partnership in Indonesia funded 

through the International Forest and Carbon Initiative. Contributions to the Design were provided by consultants 

to AusAID, and the Government of Australia and Indonesia. The views expressed in this paper are solely those 

of the author. 

REFERENCES 

Applegate, G. and Smith, J. (2000), Could Trade in Forest Carbon Contribute to Improved Tropical Forest 

Management. Centre for International Forestry Research (CIFOR). 

Bappenas. (2006), National Strategy and Action Plan for Sustainable Management of Peatlands, 

Euroconsult, Mott MacDonald and Deltares (2008), Master Plan for the Rehabilitation and Revitalisation of the 

Ex Mega Rice project area in Central Kalimantan, Indonesia, October 2008. 

Hooijer, A., Silvius, M., Wösten, H. and Page, S. (2006), Peat CO2: Assessment of Co2 emissions from drained 

peatlands in SE Asia. Delft Hydraulics and Wetlands International. 

Jaenicke, J., Rieley, J.O., Mott, C., Kimman, P. and Siegert, F. (2008), Determination of the amount of carbon 

stored in Indonesia’s peatlands. Geoderma 147, 151-159. 

Page, S. E. and Banks, C. (2007), Tropical peatlands: distribution, extent and carbon storage- uncertainties and 

knowledge gaps. Peatlands International 2007, (2), 6-27. 

Page, S.E., Siegert, F., Rieley, J.O., Boehm, H.-D.V. and Jaya, A. (2002), The amount of carbon released from 

peat and forest fires in Indonesia in 1997. Nature 420:61-65. 

Parish, F., Sinn, A., Charman, D., Joosten, H., Minayeva, T. and Silvius, M. (Eds), (2007), Assessment of 

Peatlands, Biodiversity and climate change: Executive Summary. Global Environment Centre, Kuala 

Lumpur and Wetlands International, Wageningen. 

Presidential Instruction 2007 (No 2), Acceleration of the Rehabilitation and Revitalisation of the Ex-Mega Rice 

Project Area. Government of Indonesia, 2008. 

Uryu, Y., Mott, C., Foead, N., Yulianto, K., Budiman, A., Takakai, F., Nursamsu, Sunarto, Purastuti, E., 

Fadhli, N., Hutajulu, C.M.B., Jaenicke, J., Hatono, R., Siegert, F., and Stüwe, M. (2008), Deforestation, 

Forest Degradation, Biodiversity Loss and CO2 Emissions in Riau, Sumatra, Indonesia. WWF Indonesia, 

Technical Report Jakarta, Indonesia, 74pp.



REDUCING EMISSIONS FROM 

DEFORESTATION AND DEGRADATION (REDD) 

IN LAO PEOPLE’S DEMOCRATIC REPUBLIC 

Majella Clarke 1 

ABSTRACT 

This paper sets out to synthesize the current activities and strategies for Reducing 

Emissions from Deforestation and Degradation (REDD) in Lao PDR, and discusses the 

potential implications of REDD strategies for rural livelihoods and sustainable 

development. The drivers of deforestation and degradation within the country are 

discussed, and the correlation between canopy density thresholds and deforestation drivers 

is reviewed. It is argued that canopy density threshold has an important impact on the 

potential allocation of REDD emission reduction credits, and plays an important role in the 

REDD strategy formulation process. Previous Lao forest strategies for avoiding 

deforestation in Lao PDR are discussed, noting the improvements made and challenges 

ahead. 

INTRODUCTION 

The objective of this paper is to present a synthesis of the challenges and opportunities for capacity 

development and strategy formulation for reducing emissions from deforestation and degradation 

(REDD) in Lao PDR. The paper will introduce the present forest landscape in Lao PDR and discuss the 

potential and recognized drivers of deforestation and degradation within the country. The correlation 

between canopy density thresholds and drivers of deforestation will also be reviewed in this respect. The 

paper will argue that the canopy density threshold parameter has an important impact on the potential 

REDD emission reduction credits allocated and plays an important role in the REDD strategy formulation 

process. Finally, a presentation of REDD strategies to overcome the main drivers of deforestation and 

degradation in Lao PDR will be discussed. 

Lao PDR has one of the most intact forest areas in South East Asian region with over 40% of its land area 

classified as forest. Some 80% of the population live in rural areas and are dependent on forest functions 

and resources for their livelihoods. Non-Timber Forest Products (NTFPs) are crucial for meeting 

subsistence needs and perform an important role in providing food security to the majority of rural Laos. 

While there have been few studies conducted, Foppes and Ketpanh (1997) found that NTFPs provide on 

average 55% of family cash income compared with 15% of farm income and 30% of household incomes 

are derived from livestock, while income generation from NTFPs has slightly declined (more recent 

estimates state 50%, Ingles (2006)), it is still representative of their important role in food security and 

income generation in rural livelihoods in Laos. 

As part of efforts to conserve its forest resources and biodiversity, Lao PDR established a National 

Protected Area (NPA) system designed to preserve natural resources, and protect nature and preserve the 

natural landscape. These NPAs are now National Biodiversity Conservation Areas (NBCAs) and cover 

22% of the country’s land area (that is, 5.3 million ha) and contains 20 NBCAs in addition to two 

corridors. 

Production forests play an important role in timber export revenues contributing to some 15% of GDP, 

and to meet ambitious forest sector targets, the government intended to establish 400,000 ha of 

plantations from 1993. However, only some 57,000 ha had so far been planted as of 2005 and many of the 

plantations were replanted several times because of management failure and fire, and most performed 

poorly in strict economic terms (World Bank (2005). This has however, recently improved with large 

foreign investments in jatropa and teak plantations, as well as rubber, coffee and banana plantations 

within the agriculture sector. 

1 Dr Majella Clarke, Department of Forestry, Vientiane, Lao PDR. Email: majella.clarke@indufor.fi.



However, expanding shifting cultivation areas with reduced fallow periods, hydroelectric, energy and 

mining projects, unsustainable timber extraction techniques and forest land conversion are some of the 

potential and recognized driving causes of deforestation and forest degradation in Lao PDR. In addition, 

little research and capacity development has addressed drivers of deforestation in the past in Lao PDR 

and therefore monitoring such drivers, and establishing historical trends is already creating a great 

challenge within the REDD R-PLAN preparation. 

FOREST ASSESSMENT IN LAO PDR 

A brief review of the history of the forest inventory illustrates why Lao PDR has accumulated a variety of 

methods through a number of institutional arrangements over the past century for monitoring its forests. 

The first forest assessments were conducted between 1909 and 1943 by the French colonists but it was 

almost 30 years before another forest assessment would be attempted by the Mekong River Commission 

in 1970. In 1974, Cooperation between the Governments of Lao PDR and Canada provided an assessment 

of forest cover but only in some provinces. It was not until 1982 that the first national reconnaissance 

survey was conducted on forests with the assistance of the USSR. Between 1989 and 1990 a national 

level forest inventory at the provincial level was completed. In 1990, the Mekong River Commission did 

a second assessment of forest cover. This was followed shortly by the second national reconnaissance 

survey with the assistance of the Swedish International Development Agency (SIDA). During 1995- 

2000, the project FORMACOP did a number of provincial inventories in production forests supported by 

the Finnish Government and the World Bank. In 2002, the first digital forest assessment was produced 

with the assistance of SIDA and the Japanese International Cooperation Agency (JICA), in which an 

assessment of forest cover and land use for 1982-1992-2002 was produced and published in 2005. 

Finally, the Sustainable Forestry and Rural Development Project (SUFORD) supported by the Finnish 

Government and the World Bank, produced a number of inventories for production forests between 2004 

and 2008, with ongoing support for an expansion of further inventories planned for 2009-2012 

(Chanhsomone, 2008). 

Such efforts have led to a fragmentation in the collection, storage and publication of forest inventory data. 

The most reliable source of forestry data and which the majority of governmental decisions are based 

upon is the 1982-1992-2002 Assessment of Forest Cover and Land Use published by the Department of 

Forestry (DoF) in 2005. The preliminary reference scenarios within this paper are also based on the 

assessment, as it is the most recent national level assessment of forest cover and land use and has also 

been digitalized. 

The current forest definition in Lao PDR is in line with the definition parameters used to report to FAO 

Global Forest Resource Assessment, and meets the UNFCCC COP decision no. 19/CP9 (A/R CDM). 

Based on the national forest resource assessment carried out between 1992 and 2002 the tables below 

illustrate the forest types and their respective areas in Lao PDR. 

Table 1. Forest Vegetation types and Distribution in Lao PDR in 2002 

“Current Forest” Vegetation Type Area (1000 ha) % of land area 

Dry Dipterocarp 1,317.2 5.5 

Lower Dry Evergreen 65.0 0.2 

Upper Dry Evergreen 1,387.9 5.9 

Lower Mixed Deciduous 881.0 3.7 

Upper Mixed Deciduous 5,499.5 23.2 

Gallery Forest 28.2 0.1 

Coniferous 89.1 0.4 

Mixed coniferous and broadleaved 525.8 2.2 

Wood Plantation 40.0 0.2 

Total 9,824.7 41.5 

Source: DoF (2005) 

“Current Forests” include natural forests and forest plantations and is applicable to land with tree canopy 

density of more than 20%, on an area of more than 0.5 ha, in which the trees should be able to reach a 

minimum height of 5 metres. A change in the canopy density parameter to 10% has implications and will



be reviewed later in this article. Table 1 presents the forest vegetation types and respective areas included 

in the definition of current forests. 

Potential forests refer to previous forest areas where the canopy density has been reduced below 20% for 

some reason (logging or shifting cultivation), and also include bamboo, old shifting cultivation areas 

(young secondary forests) and temporary unstocked areas. Table 2 presents the forest vegetation types 

and respective areas included in the definition of potential forests. 

Table 2. Potential forest vegetation types and distribution in Lao PDR 2002 

Potential Forest Vegetation Type Area (1000 ha) % of land area 

Bamboo 539.0 2.3 

Unstocked 10,096.3 42.6 

Old shifting cultivation areas 516.9 2.2 

Total 11,152.2 47.1 

Source: DoF (2005) 

Other Wooded Areas are areas with trees where site conditions are so poor that the canopy density can 

never be expected to exceed 20%. This includes savannah forests and heath, and stunted or scrub forests. 

(Table 3) 

Table 3. “Other Wooded Areas” and distribution in Lao PDR 2002 

Other Wooded Area Vegetation Types Area (1000 ha) % of land area 

Savanah/Open Wood Lands 94.4 0.4 

Heath, Scrub Forests 192.1 0.8 

Total 286.5 1.2 

While current and potential forest areas may look optimistic at first glance compared with other countries 

within the South-East Asian region, deforestation and forest degradation have reduced the forest cover 

from about 16-17 million ha (70% of land area) in 1940 , to 9.8 million ha by the completion of the latest 

forest assessment in 2002. One study put forest clearing in 2001 at a rate of 300,000 ha per annum 

(AusAID 1996). The government attributed the deforestation rate of about 200,000 ha per annum to 

shifting cultivation and a further 100,000 ha per annum to activities other than shifting cultivation (STEA 

2000). The following section attempts to give a brief overview of the potential and recognised drivers of 

deforestation and forest degradation in Lao PDR. 

POTENTIAL AND RECOGNISED DRIVERS OF DEFORESTATION IN LAO PDR 

Hydropower damn construction 

Hydropower is the most abundant and cost effective energy source in the Greater Mekong River Basin 

with theoretical hydroelectric potential of about 18,000 MW in Lao PDR (AusAID, 1996). However, as 

of 2001, less than 5 % (624 MW) of the country’s potential for hydroelectric power had been developed 

(DANIDA, 1998). As of 2007, Lao PDR had 10 hydropower plants operational, with an additional 12 

sites planned to be ready to export energy to Thailand by 2015. According to the latest forest assessment, 

31 forest areas for hydropower dam construction have been identified as potential sites in line with the 

government policy to respond to regional demand, amounting to 140,635 ha altogether, (DOF, 2005). The 

hydropower sites are drivers of deforestation as areas designated for development are financially 

attractive logging sites since no regeneration of forests is required after logging, and no other area to 

replant and offset what is lost is required to be regenerated. 

.Mining and Mineral Exploration 

To date, there are no studies which analyse the relationship between mining and deforestation in Lao 

PDR. In 2000, although Lao PDR has good mining potential, mining activities only accounted for about 

1% of GDP in 2000 (STEA,(2000) although recent figures now suggest that the mining sector is a major 

contributor to GPD. So far, Laos has approved 181 mining projects by a total of 118 companies, 74 of 

which are foreign owned. The Lao government reportedly plans to approve more than ten projects 

between 2008 and 2010. The majority of those projects will be bauxite and aluminium exploration and 

mining in Champasak and Attapeu provinces (Bird, 2008). Given the rich unexploited mineral resources



in Lao PDR, including tin, coal, iron, copper, gold, gypsum, zinc, sapphire and other minerals, and the 

role mining activities have played in deforestation elsewhere, mining activities, both legal and illegal, 

could be a potential driver of deforestation in Lao PDR, and which should also be further researched. 

Slash and Burn for Land Conversion 

It is generally agreed that slash and burn systems can be sustainable with long fallow periods when the 

population densities are low (Roder, 2001). For Lao PDR, slash and burn agriculture is of particular 

importance as it is a major land use practice which involve more than 150,000 households, or 25% of the 

rural population (Lao PDR, 1999). Productivity of land used in slash and burn agriculture is extremely 

low, and even lower if it takes into consideration fallow land. Even with its low productivity, slash and 

burn is still practiced widely in Northern Laos today. With many areas still not easily accessible by roads 

and with no communication infrastructure in most rural areas, providing alternatives for income 

generation and food security remains a challenge. Moreover, the practice of slash and burn could be 

difficult to eradicate because of its significance to the traditions and culture of the Northern upland 

people, even though it is part of the environmental strategy 2020 to phase out all slash and burn practices 

by 2010. 

However, due to the potential emissions which can be reduced from phasing out slash and burn, it has 

recently become the focus of a number of REDD pilots and projects. To understand the scale of slash and 

burn, whether for shifting cultivation or illegal land conversion, a map produced by the University of 

Maryland, in Thomas (2008) using MODIS satellite fire data for Laos 2007 is shown below, illustrating 

some 200,000 fire detections in 2007. The darker the dot, the higher the fire intensity and number of 

burning days detected. 

Source: University of Maryland MODIS Fire Data in Thomas (2008) 

There are other drivers of deforestation in Lao PDR, including conversion to rubber plantations, 

resettlement, legal and illegal mining, unsustainable timber harvesting, illegal logging and illegal land 

conversion. The lack of studies on the impact of, these drivers means that accurate up to date statistics 

are hard to find, and such drivers are given only a general mention in most of the relevant government 

documents (DOF,2005, STEA, 2000, World Bank, 2005).



DRIVERS OF FOREST DEGRADATION 

Shifting Cultivation with Reduced Fallow Cycles 

Shifting cultivation practices and forest fires are still the main cause of forest degradation, particularly in 

the north (DOF 2005). There, the lack of flat land for permanent and stable agriculture contributes to the 

traditional practice of shifting cultivation. Generally, yields tend to be low, increasing pressure on 

cultivation areas for rice self-sufficiency. While the State of the Environment Report 2001 identified 

600,000 ha of area as being under shifting cultivation, the government attributed deforestation and 

degradation due to shifting cultivation at about 200,000 ha per annum in 2001 (STEA 2000). Of more 

concern is the declining trend in the percentage of “current forest” in the North of Lao over the past 20 

years, while there has been an increasing trend in the “potential forest” areas (which incorporate 

unstocked forests caused by shifting cultivation (Figure1). 

% 

Figure 1. Forest and land distribution in Northern Lao PDR 

Source: Department of Forestry (2005) 

A recent concern for shifting cultivation practices in Lao PDR, and particularly in the northern region, is 

that traditionally a 12-16 year cultivation cycle is required for the recovery of soil fertility. However, in 

recent years the fallow period has been shortening and slash and burn agricultural systems with fallow 

periods of only 1-3 years are now common in Luang Prabang province, where upland rice farming is the 

main livelihood. (Kiyono et al. 2007). Some of the reasons for the shortening fallow and increased land 

pressure stem from Government policies for forest management which mandated the conversion of some 

fallow slash and burn areas to conservation forests, which in turn reduced the area under the slash and 

burn system. Teak plantations established in the 1990s also decreased the amount of arable land in the 

northern region (Roder et al. 1995). 

Unsustainable exploitation of timber and Non-Timber Forest Products is another cause of forest 

degradation. For example, approximately 80% of domestic energy consumption for cooking is based on 

fuel wood. The estimated volume of annual fuel wood consumed by local communities is 4-5 million 

m 3 /yr leading to excessive gathering of fuel wood, tree felling, and further pressure on remaining forests 

(Phanovong 1997). Excessive timber harvesting occurs most seriously in the central and southern regions 

of Lao PDR with many forest concession areas being exploited at a rate greater than 15 m 3 /ha, which has 

been set as the sustainable rate of timber extraction (STEA 2000). This has led to forest fragmentation 

and increased area classified as temporary unstocked forests. 

CURRENT REDD ACTIVITIES IN LAO PDR 

In 2007, the Government of Lao PDR submitted an expression of interest to the World Bank’s Forest 

Carbon Partnership Facility (FCPF) to receive support for developing the Readiness Project Idea Note (R- 

PIN) and subsequently Readiness Plan (R-PLAN) for REDD preparation. Activities supported under the 

R-PLAN include stakeholder consultations, developing a national REDD strategy, designing a REDD



implementation framework, developing the required reference scenarios, and establishing a monitoring, 

reporting and verifications system at the national level for changes in forest cover and carbon stocks. 

With respect to any REDD project implementation, forest carbon monitoring can be the most critical 

exercise in assessing the amount of carbon and biomass that forests store. Forest carbon monitoring 

requires remote sensing capacity, and the activities used to monitor general forest parameters are fairly 

new to Lao PDR, with the first digital forest assessment accomplished in 2002. In addition to the need for 

a high level of expertise, remote sensing activities for forest carbon assessments require expensive inputs, 

such as satellite data, field verification, aerial surveys and sometimes the establishment of permanent 

sample plots, altogether. 

THE IMPORTANCE OF THE CANOPY DENSITY THRESHOLD 

AS A PARAMETER IN THE FOREST DEFINITION 

The canopy density threshold is the most important parameter in the forest definition when accounting for 

carbon stocks, as it defines the boundaries of forest cover. Changes in this parameter are currently under 

review in Laos. The proposed change is from a 20% canopy density threshold, which now defines 

“current forests”, to 10%, which includes “current forests” and “potential forests”. This would be a 

significant change for Lao PDR.. On one hand, a change from 20% to 10% would greatly enlarge the area 

eligible for REDD emission reduction credits from about 9 million ha to 15.5 million ha, taking into 

account an average linear deforestation trend from 2002 forest assessment estimates. This change would 

also minimize the risk of degradation emissions; for example, if a forest with 40% canopy cover is 

degraded to 15% canopy cover, under a lower threshold, it is not clear that any penalties would apply. 

However, under the 20% threshold penalties would apply, as it would be classified as deforestation. A 

definition for degradation accepted within the international climate change framework has still to be 

agreed upon. 

Figure 2 illustrates the difference in forest cover area between a 10% canopy density threshold and a 20% 

canopy density threshold, and gives an idea of what future reference scenarios required under REDD may 

look like. The deforestation rates are also different, and under the current REDD and reference scenario 

models used in the voluntary market projections with high potential deforestation rates against a baseline 

generate more emission reduction credits. 

Million ha 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

1982 

1984 

Figure 2. Forest cover loss average linear trend 

1986 

1988 

1990 

1992 

1994 

1996 

1998 

2000 

2002 

2004 

2006 

2008 

2010 

2012 

2014 

2016 

2018 

2020 

2022 

2024 

Forest Cover Loss Average Linear Trend 

Av 0,003%/yr 

Av 1,26%/yr 

2026 

2028 

2030 

2032 

2034 

2036 

CD 20% 

CD 10% 

2038 

2040 

2042 

This is illustrated in Figure 3 which shows that under the assumption that deforestation for the relevant 

canopy threshold can be stabilized from 2012, the threshold which has the highest deforestation rate also 

receives more emission reduction credits. This has been one of the scrutinizing factors within the REDD



framework, as it is perceived as a mechanism which rewards the “bad guys” who have historically high 

rates of deforestation, and gives little reward to nations which have had generally low historical 

deforestation rates. The following two diagrams are based only on the data from the Forest Cover and 

Land Use Assessment between 1982-2002, DoF (2005). 

So while a decrease in the canopy density threshold from 20% to 10% will greatly enlarge the area 

available for REDD credits, it could potentially reduce the emission reduction credits allocated based on 

the current reference scenarios because the 10% canopy density threshold has a much lower historical 

deforestation rate than the 20% canopy density threshold, and this is illustrated by the area under the 

baseline, above the projected forest loss between 2012-2042. Note that the baseline assumes that 

deforestation rate goes to 0% in 2012 and forest cover is maintained throughout the reference period. This 

is particularly optimistic baseline was designed to show the maximum possible benefits of REDD. More 

realistic and up to date reference scenarios and baselines will be developed within the R-PLAN process 

and will most certainly have a deforestation rate for canopy density of 20%, above 0% deforestation, 

while the canopy density threshold of 10% would most likely be very close to 0% deforestation. 

Million ha 

20 

18 

16 

14 

12 

10 

8 

6 

4 

2 

0 

1982 

1984 

1986 

1988 

1990 

1992 

1994 

1996 

1998 

2000 

2002 

2004 

2006 

2008 

2010 

CD 20% CD 10% 

2012 

2014 

2016 

2018 

2020 

2022 

2024 

2026 

2028 

2030 

2032 

2034 

2036 

2038 

2040 

2042 

Figure 3. Comparison of emission reduction credits available under assumed baseline 

The canopy density thresholds will also have an influence on the types of REDD strategies formulated, as 

we can see from the graphs and optimistic baseline scenarios, that the canopy density threshold of 20% is 

a more risk adverse option compared with the canopy density threshold of 10%. However, one of the 

most important areas of REDD is yet to be clarified – that is the inclusion of emissions from forest 

degradation. Most of the forest cover area between canopy density threshold 20% and 10% could 

potentially represent degraded forests. 

According to the latest forest cover assessment of 2002, about 10 million ha, or some 42% of the land 

area in Laos, is classified as unstocked forest, which is one of the three categories within the potential 

forest definition (more than 10%, less than 20% canopy threshold). The national forest inventory defines 

unstocked forest areas as: 

“Unstocked forest areas are previously forest areas in which crown density has been reduced to less 

than 20% because of logging, shifting cultivation or other heavy disturbance. If the area is left to grow 

again undisturbed, it can become a forest again. Abandoned ray (or shifting cultivation areas) and 

disturbed stands with a crown density of less than 20% should also be classified as unstocked forest 

areas.” (DOF, 2005)



Given that 42% of the total land area of Lao PDR is classified as “unstocked forest area”, emissions from 

degradation could be quite significant and will be one of the focal areas in the REDD strategy, 

Nonetheless, it highlights the importance of reaching an internationally acceptable definition for forest 

degradation, so that such emissions can be accounted for. 

STRATEGIES FOR AVOIDING DEFORESTATION 

Deforestation and forest degradation are not new phenomena in Lao PDR and forest strategies have been 

drafted in the past to address such problems. This section will review the past forest strategies of Lao 

PDR noting the improvements made and challenges they face. Overall, the Government of Lao 

recognized the rapidly deteriorating forest resource situation and has set development targets for 2005, 

2010 and 2020. The targets include stabilizing shifting cultivation by 2005 and phasing it out completely 

by 2010. Tree plantations are also strongly promoted, and the classification, delineation and special 

management for protection forests, production forests and conservation forests has commenced (MAF 

2005). 

Establishing National Biodiversity Conservation Areas (NBCAs) 

Through Prime Ministerial Decree in 1993, 18 NBCAs were set aside specifically for conservation. 

Logging, collecting ,NTFPs, excavation or mining, expansion of shifting cultivation, exploitation of 

cultural or historical assets, activities that degrade the environment such as the use of explosives, 

chemicals, poisons, and burning are prohibited on these areas. Since then, several more areas have been 

added so that there are currently 20 NBCAs and 2 green corridors, covering 5.3 million ha or about 22% 

of the total land area, under some degree of protection. However, management of the NBCAs is still in 

the initial stage, with many still lacking clear delineation of boundaries, and specific management plans 

for managing high conservation values (e.g. watershed and soil conservation values), and resource 

depletion continues with illegal harvesting and trade in wildlife and non-timber forest products occurring 

in such areas (MAF 2005). 

Reduction in Annual Harvest Quota of Logs by the Government 

There are now several Prime Ministerial Orders and Decrees controlling the harvesting and sales of forest 

products. Consequently, government annual log harvest quotas have fallen from 734,000m 3 in 1999 to 

150,000 m 3 in 2004/5. However, the operational capacity of sawmills is still far above the harvesting 

levels set by the Government and puts pressure on natural forests. Over the next few years the 

Government is implementing pilots for certification projects to examine the costs and benefits under 

existing policy and conditions; however, it is still unclear as to how cost effective certification is for 

sustainably managed production forests in Lao PDR (MAF 2005). 

Stabilizing Shifting Cultivation 

In 1996, a land and forest allocation program which aimed to encourage cash crop production, stabilize 

shifting cultivation and promote forest conservation was adopted. More than half the rural population 

(some 4 million) was allocated land for farming and tree planting, and forests for management by 2005. 

Consequently, the area used for shifting cultivation was reduced between 1996 and 2005. However, the 

program was implemented very quickly without a thorough stakeholder consultation with villagers, and 

other sectors, and village forest officers received little training. Thus the results of the total program were 

limited, with poor communication and lack of support being cited as reasons for not realizing the full 

program objectives. The forest strategy for 2020 will reintroduce the land and forest allocation system, 

working more closely with villages in the land use planning process. The strategy also states that a 

national land use policy will be formulated over the coming decade (MAF 2005). 

Tree Planting 

Tree planting and improving the existing forest areas have been identified as one of the major sector 

targets, which must be achieved to contribute to poverty reduction. The targets for 2020 are ambitious 

with natural regeneration occurring on up to 6 million ha and planting trees on up to 500,000 ha in 

temporary unstocked forests. In the past, tree planting increased by 1700 ha/yr in the early 1990s, to 

17000ha/yr in 2000s. Teak in particular, has started to bring benefits to farmers from this program, 

however, past challenges included an inadequate maintenance of young stands and a lack of attention to 

the thinning (rarely done) important for quality wood production. The surveys for planted trees prior to 

2005 were not well done with respect to the selection of sites and species. There is therefore a strong need



to improve tree plantation profitability, technology and market research during the implementation of the 

2020 strategy. 

Improving Governance and the Regulatory Framework 

The legislative and regulatory framework are still being formulated and developed, and it is not 

uncommon to implement annual Prime Ministerial Degrees to cover gaps in the forest law when they are 

noticed (DOF 2005). Problems with forest law enforcement and governance relate mostly to harvesting 

and use of timber, and to NTFPs (MAF 2005). In response, a forest surveillance unit was established 

under the Ministry of Agriculture and Forestry in 2008, which aims to improve forest governance and 

decrease the rate of illegal logging. 

Overall, there have been some improvements in livelihoods due to improved forest health, and while 

progress appears to be slow in some areas, the REDD process provides an opportunity to renew such 

commitments. Some projects funded by donors are beginning to produce results which suggest that the 

participatory sustainable management of forests can lead to poverty reduction and increased earning 

opportunities. 

CONCLUSION 

REDD policies and mechanisms which support REDD strategies can impact rural forest livelihoods. 

However, policies and mechanisms from other sectors, such as energy, hydropower, mining and 

agriculture can have a greater impact on rural livelihoods, and potentially even on REDD activities, 

particularly in a country like Lao PDR where all land and forests are state owned. 

However, carefully planned community based REDD activities could, on the other hand, provide an 

additional source of income and work to rural communities in Lao PDR. Programs that focus on 

sustainable forest management and conservation, eco-tourism and protected areas, providing alternatives 

to slash and burn, afforestation and land rehabilitation, as well as community based forest surveillance 

can all have positive impacts on rural livelihoods, in addition to providing income flows. This has already 

been found with a few existing avoided deforestation/REDD projects in Africa and Latin America. 

In the past, forest strategies in Lao PDR to avoid deforestation and degradation and increase forest cover 

have met with mixed success and can provide important lessons for future REDD strategies. Expanding 

shifting cultivation areas with reduced fallow periods, hydroelectric, energy and mining projects, slash 

and burn, unsustainable timber extraction techniques and forest land conversion are some of the drivers of 

deforestation and forest degradation in Lao PDR. Of particular concern, are the potential emissions from 

forest degradation caused by shifting cultivation, usually preceded by slash and burn activities. One of the 

key challenges in meeting the information requirements for strategy formulation associated with REDD 

activities in Lao PDR is the lack of available information on the direct and indirect drivers of 

deforestation and forest degradation, and the lack of necessary data infrastructure to support forest carbon 

monitoring requirements. 

REFERENCES 

AusAID (1996) Natural Resource Management in the Mekong River Basin: Perspective for Australian Development 

Cooperation, Final Overview Report, Australian Agency for International Development, Online Document: 

http://usyd.edu.au/su/geography/hirsch/5/5/htm 

Bird, J. (2008) Intergrated Water Resources Management in the Rapidly Growing Private Sector Development 

Context: the Mekong Basin. Presentation at the World Water Week, 19 th August 2008, Stockholm, Sweden. 

Chanhsomone, P., Somchai, S., and Chittana, P. (2008). Current Status of Forest Cover in Lao PDR. Presentation to 

the 2 nd event of the Symposium on Preparing for Mitigation of Climate Change in the Mekong Region and the 

Workshop for preparing the REDD programme. Hanoi, 3-5 November, 2008, Vietnam. 

DANIDA (1998). Environmental Problems of the Energy Sector, Presentation to Danida Natural Resource and 

Environment Programme, DANIDA, Vientiane. 

Department of Forestry (DOF) (2005). Report on The Assessment of Forest Cover and Land Use During 1992-2002. 

Ministry of Agriculture and Forestry, Vientiane, Lao PDR. 

Foppes, J. and S. Ketphanh, (1997). “The use of Non Timber Forest Products in Lao PDR, paper presented at the 

International Workshop on Sustainable Management of Non-Wood Forest Products, UPM, Serdang, Selangor, 

Malaysia, 14-17 October 1997.



Ingles, A. (2006). Non-timber Forest Products and Rural Livelihoods in Lao PDR: reducing poverty through forest 

development and conservation interventions. Greater Mekong Sub-region BCI Symposium April 27 th 2006. 

Kiyono, Y., Ochiai, Y., Chiba, Y., Asia, H., Saito, K., Shiraiwa, T., Horie, T., Songnoukhai, V., Navongxaui, V. 

And Inoue, Y. (2007). Predicting Chronosequential Changes in Carbon Stocks of Pachymarph Bamboo 

Communities in Slash and Burn Agricultural Fallow, Northern Lao People’s Democratic Republic. Journal of 

Forest Research 12: 371-383. 

Lao PDR (1999). The Government’s Strategic Vision for The Agriculture Sector, Ministry of Agriculture and 

Forestry, Vientiane, Lao PDR. 

MAF (2005). Forest Strategy to the Year 2020 of the Lao PDR. Ministry of Agriculture and Forestry, Department of 

Forestry, Vientiane, Lao PDR. 

Phanovong (1997). Energy Demand and Supply in Lao PDR in World Bank (2005) Lao PDR Environment Monitor. 

World Bank, USA. 

Robinson, D,M. & McKean, S.T. (1992). Shifting Cultivation and Alternatives: An Annotated Biography 1972- 

1989. Wallingford UK, Centro Internacial de Agriculura Tropical and CAB International. 

Roder,W. (2001). Slash and Burn Rice Systems in the Hills of Northern Lao PDR: Description, challenges and 

opportunities. International Rice Research Institute, Manila, Philippines. 

Roder, W., Keoboualapha, B., Manivanh, V. (1995). Teak (Tectona grandis), fruit trees and other perennials used by 

hill farmers of Northern Laos. Agro forestry Systems 29:47-60. 

Roder, W., Manivong, V, Soukaphone, H., and Lealock, W. (1992). Fanning Systems Research In the Uplands of 

Laos: In Proceedings of the upland Rice-based farming systems research planning meeting, Chiang Mai, 

Thailand, p.39-54. 

STEA (2000). National Environmental Action Plan 2000, Science Technology and Environment Agency, Vientiane, 

Lao PDR. 

Thomas, I. (2008). How to Map Land Cover and Forest Change in Lao PDR. Presentation to the Lao REDD Task 

Force, Department of Forestry, November, 2008, Vientiane, Lao PDR. 

World Bank (2005). Lao PDR Environment Monitor. World Bank, USA.



REDD IN ASIA PACIFIC: 

CHALLENGES AND IMPLICATIONS 

Andrew Morton 1 and Blair Freeman 1 

ABSTRACT 

The scope for Reducing Emissions from Deforestation and Forest Degradation (REDD) to 

address climate change is significant. This paper outlines some of the challenges that need 

to be addressed for effective REDD projects to be put in place. Key design challenges 

include the further development of national governance and registry systems; cost–effective 

baseline assessment methodologies; and practical systems for equitable and efficient 

distributions of carbon payments to stakeholders. Emerging carbon standards will play an 

important role in supporting ongoing investment in REDD projects. Recent cost 

assessments indicate the full costs of REDD projects will tend to be higher than opportunity 

costs based on regional assessments. Furthermore, project proponents will seek returns 

commensurate with the high level of risks around REDD during this formative stage of 

development. Notwithstanding these challenges, REDD does offer enormous potential to 

address climate change and can support the broader objectives of sustainable forest 

management. If REDD is to succeed in the longer term, integrated approaches to forest 

management will be required to tackle the underlying drivers for deforestation and 

degradation, some of which comes from beyond the traditional forest sector. 

THE EMERGENCE OF REDD 

There is a relatively new forest-based solution now attracting significant attention worldwide: the 

scope for Reducing Emissions from Deforestation and Forest Degradation (REDD) is widely 

considered to have great potential to assist global efforts to curb climate change. 

Avoided deforestation was initially excluded by the United Nations Convention on Climate Change 

(UNFCC) from eligible activities within the land use, land-use change and forestry sector. However 

REDD could be incorporated in the post-2012 framework for the successor to the Kyoto Protocol. 

The premise of REDD is simple: Deforestation and forest degradation in tropical forests account for a 

substantial proportion of global carbon emissions. Emissions from tropical deforestation in the 1990s 

were estimated to be approximately 1.6 billion tonnes of carbon per year, equating to approximately 

20% of global carbon emissions (IPCC 2007). 

REDD projects offer the opportunity to utilise funding from developed countries to reduce 

deforestation and forest degradation in developing countries. REDD projects incorporate interventions 

to stopping or reducing the release of CO2 from forest lands. For example: 

• A forest that would have been cleared for oil palm is protected and therefore the CO2 is not 

released; 

• A forest that would have been logged conventionally is logged using Reduced Impact Logging 

(RIL) systems, thereby reducing damage and therefore CO2 emissions; and 

• An area that is excessively burnt is protected, thereby reducing forest degradation and carbon 

emissions. 

REDD is a compelling mechanism as it represents avoided emissions, rather than post-emission 

sequestration and offsets. It also offers scope for substantial reductions in global greenhouse gas 

1 URS Forestry, Level 6, 1 Southbank Boulevard, Southbank, VIC 3006, Australia. Email: blair_freeman@urscorp.com



emissions, potentially at relatively low cost (as much of the project creation will be undertaken in 

developing countries) and with additional environmental and social benefits. 

In 2005, Papua New Guinea and Costa Rica, supported by eight other Parties to the United Nations 

Framework Convention on Climate Change (UNFCC), proposed a mechanism for Reducing 

Emissions from Deforestation in Developing Countries. The proposal received wide support from 

Parties and the convention established a contact group and thereafter began a two year process to 

explore options for REDD. 

The scope for REDD to be included in a post-2012 UNFCCC framework was discussed further in Bali 

in 2007. Under the Bali Action Plan (or Roadmap), the Conference of the Parties agreed that if REDD 

is to be incorporated beyond 2012, a decision about what a REDD mechanism will look like and what 

it will include needs to be agreed by next meeting of the Parties in Copenhagen in December 2009. 

A REGIONAL CONTEXT 

REDD is proposed as part of a global solution to climate change, and may be developed for projects 

based in tropical countries including Central and South America, Africa and Southeast Asia. 

From an Australian perspective, there is particular interest in opportunities in Southeast Asia, 

specifically Indonesia and also Malaysia and Papua New Guinea. According to FAO reports, 

deforestation in Indonesia totaled approximately 1.8 million hectares per year through the 1990s, and 

the rate has increased from 1.7% to 2% per year since then (FAO 2006). 

A key feature of Indonesia’s capacity for REDD is its extensive peat swamp forests, which hold vast 

stores of carbon. Peat can extend to 20 metres in depth and can contain upwards of 20,000 tonnes of 

CO2 per hectare. Recent estimates indicate that in the order of 42 - 55 gigatonnes of carbon are stored 

in Indonesia's peatlands (Jaenicke et al, 2008; Hooijer et al, 2009). 

However, deforestation, draining and burning of peatland is a major contributor to Indonesia’s carbon 

emissions. According to Hooijer et al (2009), global CO2 emission caused by decomposition of 

drained peatlands was between 355 and 855 million tonnes per year in 2006, of which 82% came from 

Indonesia, largely Sumatra and Kalimantan. Based on these estimates, CO2 emission from peatland 

drainage in Southeast Asia is contributing the equivalent of 1.3 to 3.1% of current global CO2 

emissions from the combustion of fossil fuel. 

The opportunity to reduce this conversion through REDD projects is demanding attention due to the 

scale of emission reductions that could be achieved in targeted areas. Malaysia and Papua New 

Guinea have a range of adjacent forested lands and similar pressures on forests as in Indonesia. 

DESIGN CHALLENGES 

While the development of REDD has been rapid, it is still emerging as a policy option and there are 

numerous design issues and challenges that confront its proponents across a range of regions. 

Several key design challenges are outlined below, providing some perspectives on the complexity of 

REDD project development and implementation. These challenges include: 

a) Development of national governance and registry systems; 

b) Baseline determinations, incorporating additionality, leakage and permanence; and 

c) Equitable and effective distribution of payments to stakeholders. 

Governance and registry systems 

REDD has been prioritised in international negotiations about the climate change policy framework 

beyond 2012 and incorporated in consideration of the capacity of individual countries to meet 

emission reduction targets. REDD credits would accrue nationally or sub-nationally, rather than the 

smaller scale project-based approach fostered under the Clean Development Mechanism. 

National sovereignty will play a major role in the shaping of REDD projects in different countries. 

Possible government roles would appear to be as: a seller; a buyer from a sub-national devolved 

payment system; or regulator and/or broker. Mayers et al 2008 note that high levels of central



coordination will be important – strong and fair rules and institutions, macroeconomic and agricultural 

policies aligned with forest policies, effective monitoring – and issues of tenure at local level will be 

critical. 

National and sub-national governments need to resolve the role/s that they will fulfil and the processes 

and systems in which they will conduct these roles. In some respects these design issues are well 

developed, while in other respects they remain unclear. 

In Indonesia for example, there has been substantial pilot project activity at the provincial and 

kabupaten level, on the basis of direct support and facilitation from provincial governments and local 

stakeholders. However, the Central Government has more recently moved to establish its ultimate 

authority on REDD implementation. In May 2009, the Government enacted a Regulation of the 

Minister of Forestry on Procedures for Reducing Emission from Deforestation and Forest 

Degradation (P.30/Menhut-II/2009), which: 

• Recognises the establishment of a REDD Commission to manage REDD implementation; 

• Requires REDD proponents to submit a proposal to the Minister, incorporating requirements 

outlined in the Decree that include: (i) clear designation of the land tenure for the project area; (ii) 

recommendation for REDD implementation from the local government; (iii) detailed 

specifications in respect of location and baseline criteria and indicators for REDD implementation; 

and (iv) a REDD implementation plan. 

• Recognises the Minister will request the REDD Commission to assess the REDD proposal, and 

within 14 days after receiving the assessment of the REDD Commission, the Minister can decline 

the proposal or approve the proposal by issuing a licence; 

• Recognises the REDD Commission will request an Independent Appraiser Institution to conduct 

verification of monitoring reports, with costs borne by the proponent; In the event that all 

requirements are fulfilled, the Commission will issue a Carbon Emission Reduction Certificate, 

which can be traded by the proponent. 

These government guidelines have been established in advance of the decision by the UNFCC on the 

mechanism of REDD implementation at the international level, for the purpose of REDD 

demonstration activity capacity building, and transfer of technology and voluntary carbon trading. In 

this climate of change, further government regulations may follow shortly. 

In addition to the development of these guidelines, a World Bank workshop on developing REDD in 

Indonesia (2008) noted that key government challenges are to effectively implement the Forest 

Resource Information System and the National Carbon Accounting System, which will support central 

registry systems in which REDD projects can be reconciled against actual and project emission levels. 

Baseline determinations 

REDD project baseline determinations incorporate three key elements. These are: 

1. Current carbon stocks, which is effectively the carbon stored within the REDD project area at 

project inception, based on measurements and determinations of the carbon pools comprising 

above ground and potentially below ground biomass; 

2. Future carbon stocks without the proposed REDD intervention, taking into account deforestation 

or forest degradation that is expected to occur; and 

3. Future carbon stocks with the proposed REDD intervention, taking into account the impacts of the 

intervention within the REDD project area, the buffer area and any leakage impacts. 

Figure 1 illustrates the nature of a carbon baseline, and how it can change over time. For example, in 

this schematic, the REDD intervention may result in a short term reduction in emissions that is 

substantially larger than the longer term annual emission reductions. 

Current carbon stocks can to a large extent be readily determined with established forest inventory 

assessments, which include remote sensing technologies that provide scope for rapid appraisals across 

large forested areas. This is particularly the case for above ground biomass.



In the case of below ground biomass and soil carbon however, cost effective methodologies for 

measuring carbon stocks are not as well developed (Gibbs et al 2007). For forests on peat soils in 

particular, the absence of standardised, cost effective and rigorous methodologies is a significant 

technical challenge for the development of reliable baseline estimates to underpin REDD projects. 

CO2 stock 

(t CO2/ha) 

Figure 1. - Establishing a carbon baseline for REDD projects 

A 

Source: URS Forestry 

B 

Time 

Project 

eg: forest conservation, reduced 

impact logging and improved 

peat management 

Baseline 

eg: conversion to agriculture and 

degradation 

In respect to developing projections of future carbon stocks, there is a critical need for a sound 

understanding of the underlying drivers for deforestation and degradation. This will generally require 

technical forestry expertise and social research expertise to assess forestry utilisation under different 

baseline scenarios, within the REDD project area and the broader region. 

The success of the planned intervention will depend to a large extent on the nature of the underlying 

drivers, local stakeholders’ attitudes towards land rights and forest stewardship, and the way in which 

the planned intervention will address these drivers. For example, deforestation and degradation may 

be caused by (i) commercial interests in conversion to agricultural crops, (ii) markets for high value 

timbers within the project area, or (iii) land right disputes. Unless the underlying cause of 

deforestation and degradation is addressed, there is a threat to the permanence of the intervention or a 

threat of leakage to surrounding areas. 

The assessment of forestry utilisation and the risk of deforestation or degradation is also central to the 

claims that a project can make on ‘additionality’. Additionality refers to the basis on which a project 

can show that the carbon benefit (from avoided emissions or emission offsets) would not have 

happened but for the intervention. Furthermore, if the risk of deforestation or degradation is imminent, 

this can generally provide greater assurance of additionality. This can lead to some perverse outcomes 

in terms of process or in on-ground activity. For example, it could result in efforts to bring forward or 

otherwise enhance the threat of deforestation / degradation to substantiate additionality claims. 

Countries with different forest covers and historic deforestation rates will hold different interests in the 

way the baseline reference levels are constructed, and involving countries with high forest covers and 

low historic deforestation rates will be necessary to reduce perverse incentives. 

Equitable and effective distribution of payments 

The concept of REDD is based on funding from developed countries to reduce deforestation and forest 

degradation in developing countries. How this funding is distributed (financial flows) among 

stakeholders is likely to be critical to the long term success of REDD projects. How can the benefits 

from REDD be distributed to forest communities in a just, equitable way that minimises rent seeking 

of the benefits by other stakeholders or local elites? 

REDD projects can potentially comprise many key stakeholders that would likely seek entitlement to a 

share of REDD benefits. These include:



• project proponents, which may the project developers or carbon brokers, or both; 

• tiers of government – including national, provincial and district/local government authorities; 

• concession license holders; 

• village leaders and clan leaders; 

• land owners and farmers; and 

• forest based community people dependent on the use of land as a source of economic activity, 

permitted or not. 

In addition, there are project support services, including project development assessments (comprising 

technical forestry and social services as well as legal and administrative services) and ongoing 

monitoring and verification assessment services. 

The distribution of payments from the proponent to local communities in the REDD project area will 

generally be net of service fees and distributed through the tiers of government. How the government 

manages this will relate to the respective positions on REDD governance and registry systems. 

Clearly there is a threat that a cascading series of payments may substantially diminish the benefits 

payable to whose livelihoods may depend on the forests subject to REDD interventions. This is a key 

design challenge for project proponents and regional governments. In their assessment of tenure 

considerations for REDD, Cotula and Mayers (2009) highlighted the need to balance efficiency and 

fairness, noting “REDD simply will not work unless it is locally credible: it will be undermined and 

overthrown”. 

EMERGING STANDARDS 

Concurrent with REDD policy developments, carbon standards are emerging to support REDD 

projects through project inception and ongoing monitoring and validation. There are a range of 

standards and investment guidelines that are relevant to development of REDD. 

For example, the Voluntary Carbon Standard (VCS) has emerged as one of the leading carbon 

standards worldwide and can be regarded as the most prominent of carbon standards considered for 

REDD forestry projects in the Asia Pacific region currently. The VCS program has launched a global 

registry system (comprising multiple registries) which will provide the chain of custody platform for 

trading of Voluntary Carbon Units. 

The Climate, Community & Biodiversity Alliance (CCB) Standard is also prominent in existing forest 

carbon projects in Asia Pacific region. In some cases, forest carbon projects have chosen to pursue 

certification under both the VCS and the CCB standards to either progressively attain certification to 

credible standards or to enhance the credibility of the project within target markets. 

The development of the VCS and CCB standards may converge with carbon finance support programs 

and associated standards that REDD proponents are considering as relevant and provide additional 

assurances to potential purchasers. Relevant examples of related programs include: 

• The International Finance Corporation (IFC) project selection criteria established by the Carbon 

Finance Unit. This IFC unit is set up to assist project sponsors in emerging markets to access the 

market for carbon credits. Eligibility for assistance under this program is determined against 

criteria that include the location (with the primary focus on emerging countries ratified to the 

Kyoto Protocol), the amount of credits, environmental and social impact assessments and 

independent verification; and 

• Third party sustainable forest management certification, such as certification under the Forest 

Stewardship Council (FSC) or Programme for Endorsement of Forest Certification (PEFC). Such 

schemes may be aligned with or complementary to the outcomes and risks for forest carbon 

purchaser requirements.



The further development of internationally recognised carbon standards will be important to 

supporting ongoing investment in forest carbon projects, particularly during the formative phases of 

voluntary and compliance markets. 

MARKET DYNAMICS 

In the absence of regulatory demand for avoided deforestation and deforestation, voluntary carbon 

markets are leading the establishment of the market dynamics and a carbon price. 

Forest carbon transactions and prices 

In their State of the Voluntary Carbon Markets 2009, Hamilton et al (2009) noted that between 2007 

and 2009, stakeholders seeking a means of halting deforestation had begun to aggressively influence 

policy and markets to create incentives for REDD. However, despite the positive sentiments towards 

REDD, the review found that REDD-based credits declined from 1.4MtCO2e in 2007 to 0.7MtCO2e in 

2008. This trend was attributed to the difficulties in developing projects due to a range of factors, 

including layers of complexity in relation to working with communities, mid-levels of government, 

national governments, the policies and regulations around carbon ownership, technical issues around 

measuring carbon. 

Comparing over the counter transaction prices between 2007 and 2008, Hamilton et al (2009) found an 

overall increase in credit prices for most project types, including avoided deforestation – up from a 

weighted average price of US$4.80/tCO2e in 2007 to US$6.30/tCO2e in 2008. However, transaction 

data for avoided deforestation credits is limited at present. This comparison of prices for avoided 

deforestation is based on 10 and 11 transactions in 2007 and 2008 respectively, principally in Africa 

and Latin America (none in Asia). Figure 2 shows the range of prices in 2008 varied between US$4.80 

– US$28/tCO2e. 

Figure 2. - Transaction prices in voluntary markets, 2008 

USD/tCO2 

$50 

$40 

$30 

$20 

$10 

$0 

Afforestation 

plantation 

Afforestation 

reforestation 

Avoided 

deforestation 

Source: Hamilton et al 2009, Ecosystem Marketplace, New Carbon Finance. 

Costs of REDD 

A range of recent studies have estimated the cost and potential of reducing emissions for deforestation. 

Kanounnikof (2008) noted that the economic concept of the supply curve suggests there is no single 

cost-value for REDD, but alternative levels of REDD supply are associated with different costs. It is 

important to recognise there are separate cost components of REDD. These include: 

o Opportunity costs – the foregone profits from alternative land uses such as cash or food crops and 

timber sales, for which the landowner or concessionaire will seek to be compensated;



o Transaction costs – the costs associated with establishing the project or scheme, including the 

search for project areas and partners, project design, consultation with stakeholders, verification 

with standards and establishing contractual arrangements for all aspects of the project; and 

o Implementation costs – the operational costs associated with managing the project or scheme and 

the compliance costs associated with monitoring and evaluation for the life of REDD credits 

generated by the project or scheme. 

To date, opportunity costs have generally been recognised as the largest portion of REDD costs. 

Table 1 sets out a range of estimated opportunity costs of REDD in 2020, presented by the Union of 

Concerned Scientists based on recent research covering a variety of regional and global studies. 

Table 1. - Estimated opportunity costs of REDD in 2020* (US$/CO2e) 

Approach Average Range 

Regional $3.51 $1.84 – 5.18 

Stern review $6.52 $3.76 – 9.28 

Global models $12.26 $7.77 - $18.86 

Source: Boucher, 2008; * Cost (in 2005 dollars) to reduce 1tCO 2e if overall there is a 46% 

reduction in global deforestation. 

Recent CIFOR research has also found global models have yielded far higher REDD prices than 

empirical models, including the Stern estimate (Kanounnikoff, 2009). One explanation is that global 

simulation models not only consider the opportunity costs, but also the costs arising from 

interrelations with other sectors and from the fact that for practical reasons land users are likely to be 

paid a uniform price, not differentiated according to their opportunity costs (Eliasch 2008). 

Other related work includes research on the opportunity costs associated with conversion of natural 

forests in Kalimantan to oil palm plantations. Based on a financial analysis of conversion scenarios, 

the University of Queensland estimated that halting oil palm conversion in Kalimantan would cost 

between US$10 and US$33 per tonne of CO2 (Venter et al 2008). These opportunity costs are higher 

than current average prices in voluntary markets (Figure 4), but lower than prices in the order of 

US$30 per tonne in Kyoto compliance markets. 

Venter et al (ibid) also noted that opportunity costs can be considerably lower on peat soils, due to the 

substantially higher carbon stores in peat forests than in mineral soils. By targeting only peat areas for 

REDD, the opportunity cost was estimated to drop to a range of US$1.63 – $4.66 per tonne of CO2. 

This is of particular significance to the further development of REDD across Indonesia. 

However, transaction costs do need to be taken into account, particularly for the early movers, who 

face considerable risk at this early stage of REDD development. The full costs of establishing a project 

will incorporate substantial investment in project design documentation, a range of technical and 

social assessments, consultation with a broad range of stakeholders, and obtaining all necessary 

approvals at various tiers of government. Transaction costs could potentially be as high or exceed 

opportunity costs in some settings. 

In addition, there is the return on risk, or profit motive, that project developers will require to meet 

investment requirements. The risks associated with investing in REDD projects are high at present, 

and it is expected that investors will require a relatively high return on their investment, above the full 

range of costs associated with developing and implementing their projects. 

Global investment 

The success of REDD will ultimately be determined to a considerable extent by the market value of 

REDD credits relative to the opportunity cost of other carbon emissions and alternative land uses.



Based on its review of costs of REDD, the Union of Concerned Scientists estimated that for 

US$5 billion a year, REDD could be applied to nearly 20% of the tropical forests in danger of 

deforestation, and US$20 billion a year could cover about half of the tropical forests at risk (Boucher, 

2008). This study concluded that REDD can greatly reduce tropical deforestation and greenhouse gas 

emission with modest funding. 

REDD AND SUSTAINABLE FOREST MANAGEMENT 

From a forestry perspective, REDD must be considered in the context of sustainable forest 

management objectives and frameworks. Broadly, the objective of REDD is to reduce to greenhouse 

gas emissions by reducing deforestation and degradation in tropical countries. To some extent these 

are two distinct objectives, and proponents may be motivated by one or other, or both. 

The objectives of sustainable forest management, as set out under the Montreal Process, are broader. 

These objectives include maintenance of forest contribution to global carbon cycles, in addition, 

maintenance of the productive capacity of forest ecosystems and the maintenance and enhancement of 

long-term multiple socio-economic benefits to meet the needs of society. 

If REDD is focussed primarily on excluding activity within a forested region - potentially a 

preservation regime - there is a risk that it foregoes sustainable forestry and livelihoods. Sustainable 

forest management is based on comprehensive forest planning to identify the range of values within 

forests and the means of managing for multiple long-term benefits. Contula and Mayers (2009) noted 

the ‘old worry’ among foresters was that the forest sector is so complex that it will not figure in 

climate change regimes; the new worry is almost the opposite: that forest carbon finance is coming 

forward so quickly that it will not support sustainable forestry and livelihoods. 

If REDD is to succeed in the longer term it will need to be nested within more integrated approaches 

to forest management. These approaches need to tackle the underlying drivers of deforestation and 

degradation, some of which comes from beyond the traditional forest sector. Sustainable forest 

management also needs to incorporate all of the various aspects of forest-based mitigation – such as 

forest restoration and reforestation - as well as adaptation to climate change. 

LOSING SIGHT OF THE FOREST FOR THE CARBON? 

Clearly the international interest in the scope for REDD is growing rapidly. The World Bank’s 

Learning Workshop in Indonesia noted that the COP13 Bali Action Plan led to burgeoning interest in 

implementing REDD pilot and demonstration projects, and created Indonesia’s “carbon rush” (not 

unlike the California gold rush in the mid 19 th century), with its touted high financial returns. Further, 

the potential earnings from avoided deforestation carbon credit sales in Indonesia have been estimated 

in a range from US$500 million to $2 billion per annum in today’s voluntary market (World Bank 

Indonesia, 2009). 

Amid this carbon rush, there is an important if not vital role for foresters to play in assessing and 

reviewing the key elements of the project. 

For example, recent technical project experience has found some projects that have been developed to 

a considerable extent, at considerable expense, without any field reconnaissance of the proposed 

project area by experienced forestry personnel – which would have identified some key issues or 

material impacts on carbon stock assessments. 

Similarly, deforestation is commonly the last act in a forest following a series of forest-degrading 

activities over time. Successful REDD initiatives will recognise and support sustainable forest 

management as a key and critical intervention. 

This experience highlights the important role that foresters can play in the development of REDD as 

an effective mechanism for reducing greenhouse gas emissions, utilising core forestry disciplines in 

the context of broader sustainable forest management objectives.




In this climate of change, REDD could play a major role in shaping the future management of tropical 

forests within Asia Pacific. This paper highlights the compelling nature of REDD, and the challenges 

that need to be addressed for effective REDD projects to be put in place. 

Design challenges include the further development of national governance and registry systems; cost– 

effective baseline assessment methodologies; and practical systems for equitable and efficient 

distributions of carbon payments to stakeholders, including local communities. 

Emerging carbon standards will play an important role in supporting ongoing investment in REDD 

projects. 

Recent cost assessments indicate that the full costs of REDD projects will tend to be higher than 

opportunity costs based on regional assessments. Furthermore, project proponents will seek returns 

commensurate with the relatively high level of risks around REDD during this formative stage of 

development. 

Notwithstanding these challenges, REDD does offer enormous potential to assist in curbing climate 

change and can support the broader objectives of sustainable forest management. If REDD is to 

succeed in the longer term, integrated approaches to forest management will be required to tackle the 

underlying drivers for deforestation and degradation, some of which comes from beyond the 

traditional forest sector. 

REFERENCES 

Boucher, D. 2008. 2008. Estimating the cost and potential of reducing emissions from deforestation. 

Union of Concerned Scientists. Available online at: www.ucsusa.org/REDD.html. 

Cotula, L. and Mayers, J. 2009. Tenure in REDD: Start-point or afterthought? Natural Resource 

Issues No. 15, IIED, London, UK. 

Eliasch, J. 2008. The Eliasch Review – Climate Change: Financing Global Forests. UK Office of 

Climate Change. Available online at: http://www.occ.gov.uk/activities/eliasch.htm. 

FAO, (2006), Global forest resource assessment 2005: progress towards sustainable forest 

management. Forestry Paper 147. 

Gibbs, H., Brown, S., O Niles, J. and Foley, J. 2007. Monitoring and estimating tropical forest carbon 

stocks: making REDD a reality. Environmental Research Letters, 2 (2007) 045023. 

Hamilton K, Sjardin, M., Shapiro, A. and Marcello T. 2009. Fortifying the Foundation: State of the 

Voluntary Carbon Market 2009. Ecosystem Marketplace, New Carbon Finance: USA. 

Hooijer, A., S. Page, J. G. Canadell, M. Silvius, J. Kwadijk, H. Wösten, and J. Jauhiainen, 2009. 

Current and future CO2 emissions from drained peatlands in Southeast Asia, Biogeosciences 

Discuss., 6, 7207–7230, 2009, www.biogeosciences-discuss.net/6/7207/2009/. 

Intergovernmental Panel on Climate Change (IPCC) 2007 Climate Change 2007: The Physical 

Science Basis: Summary for Policymakers. 

Jaenicke, J et al. 2008. Determination of the amount of carbon stored in Indonesian peatlands. 

Geoderma. Vol. 147. pp. 151-158. 

Mayers, J., Bass, S. Bigg, T., Bond, I., Bradstock, A. and Grieg-Gran, M. 2008. Forest-based climate 

strategies: making REDD and other initiatives work. IIED, London, UK. 

Parker, C, Mitchell, A, Trivedi, M. and Mardas, N, 2008. The Little REDD Book, Global Canopy 

Programme, Oxford, UK. 

Richards, M. 2008. REDD, the last chance for tropical forests? Policy brief, FRR, Bristol, UK. 

World Bank, 2009. Convenient solutions to an Inconvenient Truth: Ecosystem-based Approaches to 

Climate Change. Environment Department, World Bank.



ABSTRACT 

A CHARACTERISATION OF THE CURRENT MARKET FOR 

REDUCING EMISSIONS FROM DEFORESTATION AND 

DEGRADATION (REDD) AND FOREST CARBON OFFSETS 

Majella Clarke 1 

Forest carbon offsets from REDD and forestry projects are providing increasing 

opportunities in the voluntary carbon market. Currently, forest carbon offsets from 

REDD and most forestry projects are only able to be traded in the voluntary carbon 

market on a project by project basis, and on a deal by deal basis through the Over-The- 

Counter market. The paper argues that because of the global nature and flexibility of the 

voluntary market, project based crediting mechanisms have been set up under a range of 

different standards for accounting, monitoring, verifying and distributing carbon credits 

from forest carbon projects, leading to growing product complexity, difficult 

marketability and little overall transparency. The paper concludes that market and 

liquidity fragmentation have implications on achieving a common price signal and that 

inefficient price formulation is, at this stage, an unavoidable characteristic of the 

voluntary market for forest based carbon credits. 

INTRODUCTION 

The objective of the paper is to present a characterisation of the current market for forest carbon 

offsets with a particular focus on Reducing Emissions from Deforestation and Degradation (REDD). It 

builds upon the State of the Voluntary Market Report 2008 (Hamilton et al. 2008) but diverges to 

focus specifically on carbon offsets from forest projects and REDD to give a market characterisation. 

It distinguishes itself by reviewing a number of REDD projects generating offsets on the voluntary 

market through a desk review of Project Design Documents (PDD), project websites, verification 

reports, newsletters, communications with project implementers and specific project presentations 

were reviewed for 12 REDD projects currently selling emission reductions as a result of the project 

activities on the voluntary carbon market. Also, carbon offset prices were collected from 79 vendors 

selling carbon offsets online with forestry projects in their portfolio. 

The first section of this paper will summarise the general trends in the trade of forest based carbon 

offsets and briefly reviews current REDD projects, project types, their mechanisms and the spatial 

distribution. The second section looks into the voluntary market structure for forest carbon based 

offsets. It argues that because of the global nature and flexibility of the voluntary market, project based 

crediting mechanisms have been set up under a range of different standards for accounting, 

monitoring, verifying and distributing carbon credits from forest carbon projects, leading to growing 

product complexity, difficult marketability and little overall transparency. The third section of this 

paper argues that market and liquidity fragmentation have implications on achieving a common price 

signal and that inefficient price formulation is, at this stage, an unavoidable characteristic of the 

voluntary market for forest based carbon credits. 

General Trends 

The market structure for REDD is, at this stage, to a large degree undefined as it is still not clear what 

“product functions” will be included within the international framework. Will deforestation and 

degradation be considered under the same mechanism? Or will it just be deforestation? Will there be 

multiple mechanisms? The most obvious and important market structure determinants for REDD will 

be what is offered and how it is regulated, as these determinants will affect the supply and demand of 

emission reduction offsets directly. However, it may also follow the market development of other 

forest carbon offsets, in which the Over-The-Counter (OTC) voluntary market has shown potential for 

filling compliance market gaps. 

1 Dr Majella Clarke, Department of Forestry, Vientiane, Lao PDR. Email: majella.clarke@indufor.fi



According to Hamilton et al. (2008), OTC transactions for avoided deforestation credits grew from 3% 

to 5% of total OTC offset credits in 2006 and 2007 respectively. Because of the nature of avoided 

deforestation offset credits being traded solely OTC, prices for such credits had large variations 

ranging from about USD 2.50 - 30 tCO2e in 2007, with 11 data points used in calculating the volume 

weighted average of USD 4.80 /tCO2e for avoided deforestation offsets. One of the key reasons for 

the divergence in price formulation is because avoided deforestation offsets are sold on a project by 

project basis, and each project is unique. 

Types of Projects and Mechanisms to Reduce Emissions from Deforestation and Degradation 

Of the 12 REDD projects reviewed, at total of 2,572,670 ha of forest area was being used to 

implement REDD projects. one project was reviewed from Australia, two projects were reviewed from 

the USA, six projects from Latin America, two from Africa, both in Madagascar and finally, one 

project from Asia in Indonesia. Project areas ranged from 756 ha in the USA’s Fred Van Eck Forest 

Foundation registered on the California Climate Action Registry, to 750,000 ha of the Indonesian Ulu 

Masen REDD project, in which VERs have been sold to private investors in cooperation with donors 

and governments. So far, Latin America has the largest area and the most REDD projects within the 

sample reviewed. Within these REDD projects, a total of 362,315,886 tCO2e are expected to be offset 

over the course of the project. Project based emission reduction durations ranged from 20 years to 100 

years. So far, the Latin American region has contributed to generating the most emission reductions. 

All 12 projects reviewed had the central objective of avoiding deforestation and reducing emissions. 

However, their emission reduction strategies included objectives like reducing emissions through 

avoided slash and burn of forests, avoided land conversion, avoided forest fragmentation, avoided 

logging concessions, etc. Figure 1 presents the different objectives in projects in addition to avoided 

deforestation. 

N. of Projects 

14 

12 

10 

8 

6 

4 

2 

0 

Avoided 

Deforestation 

Avoided slash 

and burn 

Avoided land 

conversion 

Avoided illegal 

activities that 

lead to 

deforestation 

Avoided forest 

fragmentation 

and 

encroachment 

Avoided 

logging 

conessions 

Avoided land 

degradation 

Focus 12 5 5 4 2 4 1 

Figure 1. A Comparison of different project objectives to Reducing Emissions from 

Deforestation and Degradation (REDD) 

Avoided slash and burn, and avoided land conversion featured prominently throughout most REDD 

projects currently in implementation. One possible explanation for this is that avoiding slash and burn 

and avoiding land conversion (which often involves slash and burn in developing countries) can 

generate more emission reduction credits tCO2e/ha, than avoiding logging concessions without 

burning. A prime example of this is the Noel Kempff project in Bolivia, in which there are two 

components:



Component 1 Stopping Industrial Timber Harvesting on an area of 524,000 ha generates 791,444 

tCO2 of certified offsets, or between 1997-2005, an average of 1.51 tCO2/ha 

Component 2 Avoided Slash and Burn Agriculture on an area of 756 ha generates 371,650 tCO2 

certified offsets, or between 1997-2005, an average of 491.6 tCO2/ha (Seifert- 

Granzin, 2007) 

This project illustrates that different REDD objectives yield different magnitudes of emission 

reduction credits, and may shed light on why an increasing number of REDD projects, pilots and 

initiatives are including additional components addressing slash and burn agriculture and land 

conversion. In response to the different emission reduction strategies presented above, a number of 

standard project mechanisms for REDD are presented in Figure 2. 

No. of REDD Projects 

9 

8 

7 

6 

5 

4 

3 

2 

1 

0 

Figure 2. Project mechanisms for the implementation of REDD 

In general, two types of project mechanisms featured prominently within the 12 REDD projects 

reviewed. 

• Protected Areas and Corridors were the focus of REDD projects in Latin America and Africa 

and include: 

o Establishing Protected Areas 

o Enlarging Protected Areas 

o Creating/protecting Green Corridors 

o Protected Area Management 

• Community Development, Agreements and Land Use Planning were the focus of most REDD 

projects in developing countries. This includes activities like finding alternative incomes and 

employment for activities which deforest, land use planning and zoning at community and 

village levels so that livelihoods are improved, and agreements with surrounding communities 

on land and forest use. 

Other project mechanisms common in REDD projects include agroforestry, Forest Law Enforcement 

and Governance initiatives, sustainable forest management with conservation, indigenous rights and 

ecotourism. Within the brief overview of on-going REDD activities which have received or are 

applying to obtain verified offsets within the voluntary carbon market, there are several general 

conclusions which can be drawn from this so far:



• There are different methods for reducing emissions from deforestation and degradation. 

• There are many different standards through which an REDD project can be verified and 

implemented upon. 

• Some activities are much more efficient (tCO2/ha) in yielding emission reduction credits than 

others (e.g. the Noel Kempff example). 

• There are a number of mechanisms which can be used to implement an REDD project, which 

require completely different financing requirements. 

• REDD activities and their emission reduction credits are not limited only to developing 

countries, like the CDM A/R, but are also able to be implemented in developed countries. 

VOLUNTARY MARKET FOR REDD AND FOREST CARBON OFFSETS 

Market Structure Determinants and the OTC Market 

Following Willet (1931) and Senn (2002), Over-The-Counter (OTC) markets can be characterised by 

some of the following features: 

1. Small capitalization. 

2. Limited distribution. 

3. Lack of speculative interest. 

4. High price. 

5. Trading is not centralised 

6. Dealers are normally market makers 

7. Desirability for portfolios of institutional investors (banks, insurance companies, etc.) who 

wish to negotiate the purchase or sale of a large block at one price. 

These characteristics are highly reflective of the REDD and forest carbon offsets market. Often offsets 

are purchased by individuals or businesses, on a deal by deal basis to offset household emissions, 

flights or events. There is no specific or central location for the sale and purchase and there is an 

increasing number of internet vendors (for example, see www.carboncatalogue.com and 

www.carbonoffsetguide.com.au). A single purchase of offsets is usually very small compared with the 

offering (small capitalisation). Because the individual purchase of forest carbon offsets have the sole 

purpose of offsetting emissions, it is void of speculation. 

In general, there are several market structure determinants that are common and applicable to buying 

and selling forest carbon offsets on the OTC market: 

1. Product complexity and marketability 

2. Data and pricing transparency 

3. Liquidity 

Product Complexity and Marketability 

Product complexity refers to the increasing number of product options an entity has to offer, and can 

affect the marketability and adoption of the product (AMD (003, UGS PLM Solutions 2004). The 

rationality behind this is that the more simple a product is, the easier it is to sell. This is a particularly 

relevant market structure determinant to the selling of forest carbon offsets due to the methodological 

complexities in measuring and monitoring forest carbon stocks and changes, multiple project options 

and strategies which can be used to implement an REDD or forest project, and the different forest 

values which can be offered in addition to forest carbon credits, e.g. biodiversity conservation, 

community development etc. In all, it can be difficult for consumers of forest carbon offsets to know 

what they are buying, and therefore, an evolving list of standards have emerged to verify a certain 

quality and standard are met by forest carbon offset projects. 

According to the recent study by Hamilton et al. (2008), over the past 2 years, the number of 

independent verification standards has risen in line with market trends which require the independent 

verification of projects which reduce GHG emissions. There are now more than a dozen standards for 

verification in existence to prove project activities reduce emissions. Some have been developed



specifically for REDD and avoided deforestation while others have been developed for general 

emission reductions verification. In the State of the Voluntary Market Report for 2008, the emergence 

of carbon standards and registries was one of the most noticeable trends in 2007, in response to 

proving the legitimacy and quality of carbon credits. Table 1 illustrates the growing number of 

standards and the different types of projects for forest based offsets in the voluntary market. 

Table 1. Standards on the voluntary market for forest offsets 

Standard Project types 

Climate, Community & Biodiversity Climate, Community & Biodiversity Alliance (CCBA) verifies all 

(CCB) Standard 

land based project types 

Climate, Community & Biodiversity 

(CCB) Standard, 2 nd In addition, it takes into account indigenous rights and high 

Edition biodiversity values 

Voluntary Carbon Standard (VCS) Afforestation, reforestation and revegetation, agricultural land 

Agriculture, Forestry and Other Land management, improved forest management, REDD 

Uses (AFOLU) 

Carbon Fix Standard Conversion of non-forest land to forest land, conservation forests, 

planted sustainably managed forests, protected areas leading to land 

use change of non forest to forest. Not for REDD. 

Carbon Fix Standard 2.1 As above, also includes agro-forestry projects 

Plan Vivo Afforestation, reforestation, agro-forestry, restoration, conservation, 

improved forest management and REDD. 

The Gold Standard Only REDD through fuel efficiency. 

VER+ Land Use, Land Use Change & Forestry (LULUCF) projects, 

including REDD, are accepted if implemented with a buffer approach 

to address the risk of potential non-permanence 

Green e Agriculture, Forestry and Other Land Uses (AFOLU) projects under 

Voluntary Carbon Standard (VCS) 2007, are eligible as long as the 

seller provides proof that the native species requirements under the 

Green-e Climate Standard are met. 

ISO 14064 Special importance for the emerging voluntary approaches to 

greenhouse gas declarations by companies. Does not specify or 

exclude any project types. 

Chicago Climate Exchange (CCX) Protocol for sustainably managed forests, standard for afforestation 

and tree planting. 

Greenhouse Friendly Special guidelines for forest sink abatement projects. 

The Carbon Neutral Protocol Forestry and land use, manage forests to ensure their continued health 

and carbon storage capacity, make effective use of forest and farm 

residues. 

California Climate Action Registry Includes Forest Sector Protocols. 

Protocols 

Social Carbon The methodology is based on the Sustainable Livelihood Approach, 

and considers six basic resources: Social, Human, Financial, Natural, 

Biodiversity and Carbon. 

Voluntary Offset Standard (VOS) Only accepts Verified Emission Reductions (VERs) from projects 

implemented using CDM methodologies and Gold Standard offsets. 

United National Framework for Only Afforestation and Reforestation, does not include REDD yet, 

Climate Change Convention though there is a proposal for the inclusion of REDD under the CDM 

(UNFCCC) Clean Development to be negotiated within the United Nations framework for climate 

Mechanisms (CDM) Afforestation change. 

/Reforestation (A/R) 

Sources: See websites for standards in references, in addition to Merger (2008), Kollmuss (2008) Ecosystem Marketplace, New Carbon 

Finance.



Data and Pricing Transparency 

Data and pricing transparency is an important aspect of pricing and valuation. This is backed up by 

Redmond (1989), Leach and Madhaven (1993) and García-Alonso et al. (1997). Price transparency is 

often limited by the private nature of the trades in the OTC market, and there are studies to conclude 

that prices on an exchange can better reflect information than prices in the over the counter market 

(Bessimbinder 1999). In order to deal with the lack of transparency of the OTC market, in addition to 

the standards, a number of carbon credit accounting registries have evolved, and are like public 

databases which provide transparency over ownership claims concerning emission reductions. 

Table 2. Registries and transparency 

Registry Host Transparency Forest 

Offsets 

American Carbon Registry USA Projects and Account Information Public No 

Blue Registry Germany Projects and Account Information Public No 

Chicago Climate Exchange USA CCX verified and registered offset projects, list of Yes 

(CCX) Registry 

offset providers and projects, list of members, but 

linkages between members, providers and projects is 

undisclosed. 

Asia Carbon Registry 

Singapore List of projects, place, Certified Emissions 

Reductions (CERs) and Verified Emission 

Reductions (VERs). Account info not public 

Yes 

Australia Climate Exchange Australia Linked with Greenhouse Friendly Abatement Yes 

Registry 

Register. Account info undisclosed 

Australian Carbon Traders 

Registry 

Australia Project and account info public Yes 

New South Wales Greenhouse 

Gas Reduction Scheme (GGAS) 

Australia Project and account info public Yes 

Greenhouse Gas (GHG) Clean 

Projects Registry 

Canada Account info and projects are public Yes 

Canadian Greenhouse Gas (GHG) Canada Project info is public, but transaction info is not Yes 

Reductions Registry 

recorded on this registry 

California Climate Action USA Tracks and registers voluntary projects that reduce Yes 

Registry (CCAR) 

emissions of GHGs. Publically available. 

Triodos Climate Clearing House Netherlands Undisclosed Yes 

Globe Carbon Registry Canada The GLOBE Carbon Registry will ensure market 

integrity by issuing a unique serial number to each 

posted offset, avoiding double counting and allowing 

chain of custody control from creation to retirement. 

Unclear 

Gold Standard Registry for Switzerland The registry’s infrastructure prevents double- Possible 

Verified Emission Reductions 

counting of VERs and provides numerous public 

(VERs) 

reports and a full audit of all transactions to ensure 

the integrity of the Gold Standard VERs. 

TZ1 Registry New 

Zealand 

Full account and project info available to public. Yes 

Carbon Offset Project Registry UK Project and credit info available, account info 

undisclosed to public 

Yes 

APX Voluntary Carbon Standard USA The Registry will offer publicly accessible reports Yes 

(VCS) Registry 

showing the following: a directory of organizations Monitored, 

using the Registry, a list of Voluntary carbon Units measured, 

(VCU) registered projects, issued and retired VCUs. & verified 

only. 

While the registries aid in transparency over the transactions of credits and trace transactions and 

ownership, pricing transparency is still a problem as the nature of the OTC is a deal-by-deal basis. It is 

up to the buyer to look around and find the right price offering. There is some independent online 

information which lists carbon offset providers (www.carboncatelog.org), their projects and their 

prices, and based on a review of carbon price data collected from 79 vendors selling forest carbon 

offsets within their portfolio (note that several of the vendors reviewed included forestry offsets mixed 

with other non-forest emission reduction projects), Figure 3 illustrates the variance in the price for 

forest carbon offsets (t/CO2).



USD/CO2 

80 

70 

60 

50 

40 

30 

20 

10 

0 

27.43 

13.15 

17.36 

Europe (8) USA (14) UK (16) Australia (34) Canada (6) New Zealand (1) 

13.76 

14.72 

16.54 

Series2 

Series1 

Source: Project reviews and price data were collected from www.carboncatalogue.com, March 19th 2009, 

www.carbonoffsetguide.com.au, May 22 nd 2009 . Numbers in parentheses indicate the number of carbon offset sellers. 

Figure 3. Average price of forest carbon offsets Q1, Q2 2009 

The range of price in which forest carbon credits are traded over the counter (i.e. not in auctions or on 

exchanges) in different regions of the developed world in the first quarter of 2009 shows that Europe 

(excluding the UK) had by far the largest carbon price range and the highest average price of carbon 

per ton for forest carbon offsets, while the other regions had the average price of carbon between 

USD10-20. With such a variance in price for tCO2, it is clear that there is a strong need to improve the 

price discovery process for carbon offsets traded on the over the counter market. 

Market and Liquidity Fragmentation in the Price Formulation Process 

Markets have two broad functions – liquidity and price discovery. The Stern Review, Stern (2006), 

states that where markets function, two conditions must hold to reduce GHG emissions efficiently. 

Firstly, the marginal social cost of carbon must equal the cost of abatement, and the second, more 

relevant to this analysis, is that to deliver emission reductions at least cost, a common price signal 

is required across countries and different sectors at a given point of time. The deeper and liquid a 

market, the harder it is for an individual trade to affect the overall price and the less volatile the market 

will be. Liquidity fragmentation, however, inhibits the efficient exchange interaction which would 

normally occur in an open integrated market, as each pool draws its own trades, market data becomes 

more difficult to collect and analyse, which can lead to imperfect information, volatility and inefficient 

price signalling and price formulation processes. This can already be seen from the lack of available 

pricing information available, different standards and registries being used across the board to verify 

and validate offsets, and large variations in price ranges of carbon offsets across countries (see Figure 

3). One of the key concerns of market fragmentation for REDD and forest carbon offsets is that it 

could affect the efficient price formulation process via liquidity fragmentation. There is abundant 

literature which supports the link between market fragmentation and liquidity fragmentation 2 . 

2 Bennett and Wei (2003) showed that compared with a fragmented market structure, an exchange which requires all orders 

to interact and compete, produces higher quality price formulation, lower volatility and lower execution costs. Mendelson 

(1987) showed that a fragmented markets creates lower liquidity, and higher price volatility, while Madhaven (1995) showed 

that market fragmentation increases both the price variations and leads to price inefficiency. A number of papers have also 

shown that market fragmentation can reduce liquidity. Cohen, Maie, Schwartz and Whitcomb (1982) point out that off-



Markets fragment when instruments embodying the same risk trade in multiple locations or forms. 

Fragmentation may occur across multiple markets that trade identical instruments or across multiple 

instruments whose values depend on the same underlying risks (Harris and Mayhew 2005). This is 

already prevalent in the OTC dealings for forest carbon offsets with different projects offering carbon 

emission reduction offsets using a range of implementation methods (sustainable forest management 

with conservation, enlargement of protected forest area, strategies to reduce emissions from slash and 

burn etc), different standards (ISO 16404, Gold Standard etc) and having the information and 

transaction information stored in an array of registries. There are different levels and strata of linkages 

between the standards, registries and exchanges. Some registries have linkages with other registries 

that have linkages with exchanges, other registries are linked through standards. For example, the 

Chicago Climate Exchange (CCX) also hosts the CCX registry and has a board that administer CCX 

standards and protocols. Another example would be Triodos Climate Clearing House, which functions 

as an independent registry and trading platform in the transaction of carbon credits, yet standards 

accepted are based on discretion. 

One apparent observation from this exercise is that there are a number of voluntary market carbon 

instruments, which often carry the same type of risks (e.g. permanence) trading across multiple 

markets, and in different locations. Information on these trades in the OTC market, is generally 

registered to increase transparency, however, not only are the markets clearly fragmented, the 

information based in the registries has also become fragmented with the increasing number of 

evolving standards and registries. As stated before, markets function by providing liquidity and a price 

discovery process, however, as illustrated, registries, standards and exchanges have linkages. Given 

the relatively small capital for voluntary market forest offsets - 0.000527% of total global carbon 

market value (in 2006, a total of USD 21.06 million worth of forest offsets was traded on the voluntary 

market, compared with a total global carbon market worth USD 40 billion, Hamilton et al. (2008)) 

disbursed over a growing number of market instruments, liquidity fragmentation has become 

inevitable. 

What this means for REDD and forest offsets is that transaction costs for each project are higher than 

exchange traded offsets because it requires a broker to introduce new offset buyers to new offset 

sellers due to product and market complexity. When liquidity becomes highly fragmented for forest 

carbon offsets, large projects may become difficult to finance and will also take time to finance as the 

broker/contributor looks for different sources to leverage funding. 

To counter this anomaly, some REDD projects which sell offsets rely to some degree on donor funds 

to cover capacity building and monitoring costs. In addition, there are an increasing number of funds 

being established, and pledges made to tackle deforestation in tropical countries. Each fund is unique 

in terms of contributors, its aim and the amount it commits, as well as regional focus. The amounts of 

money are actually quite significant – Norway pledged EUR 1.8 billion to REDD over 5 years, and 

Germany pledged EUR 500 million, with many other nations making large contributions(see Bellassen 

et al. (2008). At this stage it is unclear whether some of these funds will be directly linked to the 

carbon market to generate carbon offsets, however, a number of proposals have suggested that there 

should be future efforts to link funds with a carbon market, see The Little REDD Book, for an outline 

of the 12 international proposals on approaches to market REDD credits (Parker et al. 2008). 

CONCLUSION 

Forest carbon offsets from REDD and forestry projects are providing increasing opportunities in the 

voluntary carbon market. There are a number of options on how to reduce emissions from 

deforestation and forest degradation including avoiding slash and burn of forests, avoiding illegal and 

legal land conversion, avoiding illegal logging and encroachment, and the purchase of logging 

concession sites before harvest. There are a number of different instruments which can be used to 

exchange executions may benefit brokers but harm the market as a whole. Cohen, Conroy and Maier (1985) show that a 

fragmented market may result in a wider bid-ask spread because of decreased opportunity for order interaction. Mendelson 

(1987) finds that the fragmented market has less liquidity and increases price variances faced by investors. Madhavan (1995) 

demonstrates that fragmentation results in higher price volatility and violations of price efficiency. Empirically, Amihud, 

Lauterbach and Mendelson (2002) provide evidence that trading consolidation improves liquidity and adds value to investors.



achieve reduced emissions forest deforestation and forest degradation. Protected area enlargement, 

establishment and/or management, and community development and community level land use 

planning are the two most popular types of instruments for REDD projects, based on the review of 12 

REDD projects. Other instruments include sustainable forest management with conservation, 

improving forest law enforcement and governance, agroforestry, ecotourism and working with 

indigenous rights. 

Currently, forest carbon offsets from REDD and most forestry projects are only able to be traded in the 

voluntary carbon market on a project by project basis, and on a deal by deal basis through the Over- 

The Counter market. The market for REDD and forest based carbon credits can be characterised by 

the following trends: 

Growth in the number of projects offering forest based carbon offsets OTC 

Increasing number of standards for forest based carbon offsets 

Increasing number of registries to improve transparency 

Increasing number of funds available for REDD activities 

Uncertainty of future linkages with the carbon market 

The market for REDD and forest based carbon credits can be characterised by the following market 

characteristics: 

Only available on the Voluntary Market OTC 

Market Fragmentation due to the nature of the OTC market 

Large degree of price variation USD/tCO2, particularly within the EU 

Lack of a common price signal across markets and sectors 

REDD and forest carbon offsets are difficult to market 

Small capitalisation distributed over large geographical scope and across many markets has 

led to liquidity fragmentation 

High transaction costs 

Due to the nature of the OTC market, market fragmentation is unavoidable due to the product 

complexity, lack of data and price transparency, and divergent regulations to which different project 

regions are subjected. The lack of data and price transparency and product complexity means that 

forest carbon credits are traded on a deal by deal basis, with prices ranging from USD 2.50-68 per ton 

of CO2 in the first quarter 2009, based on price data collected from 79 vendors. As mentioned earlier, 

to reduce GHG emissions efficiently, one of the key market criteria which must hold is that a common 

price signal is required across countries and different sectors at a given point of time (The Stern 

Review). As this paper shows, this is far from the case and that inefficient price formulation is caused 

by market and liquidity fragmentation inherit in the over-the-counter market with small capitalisation. 

Given that reducing deforestation rates substantially will require significant levels of finance (USD 

17-30 billion per year to halve the emissions from the forest sector by 2030), there is an urgent need to 

consolidate standards, harmonise pricing and transparency mechanisms, to develop efficient and 

decisive methods to consolidate the required capital to reduce emissions from deforestation and 

degradation on a significant scale. 

REFERENCES 

American Carbon Registry - http://www.americancarbonregistry.org/carbon-registry [last viewed 19th March 

2009] 

AMD Global Consumer Advisory Board (2003) Technology Terminology and complexity Study. Online: 

http://www.amd.com/us-en/Weblets/0,,7832_8551,00.html 

APX VCS Registry http://vcsregistry.apx.com/ [last viewed 19th March 2009] 

Asia Carbon Global Registry Services - http://www.asiacarbon.com/Carbon_Registry_Services.html [last viewed 

19th March 2009] 

Australian Climate Exchange Registry http://www.climateexchange.com.au/Content/rOperation.aspx [last 

viewed 19th March 2009]



Australian Carbon Traders Registry http://www.australiancarbontraders.com/carbonmart/public/register [last 

viewed 19th March 2009] 

Bellassen, V., Crassous, R., Dietzsch, L. & Schwartzman, S. (2008) Reducing Emissions from Deforestation and 

Degradation: What Contribution from Carbon Markets? Mission Climat of Caisse des Dépôts . Climate 

Report Issue 14, September, 2008, France. 

Bennett, P. & Wei, L. (2003) Market Structure, Fragmentation and Market Quality. Evidence from Recent 

Listing Switches. Draft Paper http://media.terry.uga.edu/documents/finance/bennett.pdf 

Bhattacharya, Mita and Bloch, Harry and Chen, Jong-rong. 2008. Determinants of domestic market structure: A 

comparative study between Japanese and Taiwanese manufacturing. Asia Pacific Journal of Economics and 

Business 11 (2): 25-38. 

Blue Registry - https://www.netinform.de/BlueRegistry/Projects.aspx [last viewed 19th March 2009] 

Bissembinder, H. (1999) “Trade Execution Costs On NASDAQ And The NYSE: A Post- 

Reform Comparison,” Journal of Financial and Quantitative Analysis, v34[3,Sep], 387-407. 

California Climate Action Registry - http://www.climateregistry.org [last viewed 19th March 2009] 

California Climate Action Registry Protocols - http://www.climateregistry.org/tools/protocols/projectprotocols.html 

[last viewed 19th March 2009] 

Cameron, P. (1995) Gas Regulation in Europe. From Monopoly to Competition. Fuel and Energy Abstracts, 

Volume 36, Number 4, July 1995, pp255. 

Canadian GHG Reductions Registry - http://www.ghgregistries.ca/reductions/index_e.cfm [last viewed 19th 

March 2009] 

Carbon Catalogue - http://www.carboncatalog.org/ [last viewed 19th March 2009] 

Carbon Fix Standard - http://www.carbonfix.info/ [last viewed 19th March 2009] 

Carbon Offset Guide – http://www.carbonoffsetguide.com.au/ [last viewed 10 th July 2009] 

Chicago Climate Exchange, Forestry Carbon Emission Offsets - 

http://www.chicagoclimatex.com/content.jsf?id=242 [last viewed 19th March 2009] 

Chicago Climate Exchange Registry - http://www.chicagoclimatex.com/content.jsf?id=582 [last viewed 19th 

March 2009] 

Climate, Community and Biodiversity Standard - http://www.climate-standards.org/ [last viewed 19th March 

2009] 

Crampton, P. & Kerr, S. (2002) Tradable Carbon Permit Auctions. How and Why to Auction Not Grandfather. 

Energy Policy 30, 333-345. 

EU_Commission (2008). Proposal for a Directive of the European Parliament and of the Council amending 

Directive 2003/87/EC so as to improve and extend the greenhouse gas emission allowance trading system 

of the Community (presented by the Commission). 2008/0013 (COD) COM(2008) 16 final. 

Federer, J.P. The rise of the Over the Counter Market. Department of Economics, Malacester College, 

USA.pp39. 

García-Alonso, M.D.C. & Levine, P. (1997) Strategic Procurement and Market Structure. International Journal 

of Industrial Organization, Volume 26, Issue 5, September 2008, Pages 1180-1190. 

GHG Clean Projects Registry http://www.ghgregistries.ca/cleanprojects/index_e.cfm last viewed 19th March 

2009] 

Green e Standard - http://www.green-e.org/getcert_ghg_standard.shtml [last viewed 19th March 2009] 

Greenhouse Friendly Guidelines - http://www.climatechange.gov.au/greenhousefriendly/publications/gfguidelines.html 


Hamilton, K., Sjardin, M., Marcello, T., and Xu, G. (2008) Forging a Frontier: State of the Voluntary Carbon 

Markets 2008. A Report by Ecosystem Marketplace and New Carbon Finance. May 2008, USA. 

Harris, L.E. & MayhewS s. (2005) Market Fragmentation Metrics. Independent Paper. California, USA. 

IPCC (2007) Fourth Assessment Review. http://www1.ipcc.ch/ [last viewed 19th March 2009] 

ISO 14064 - http://www.climatechange.sgs.com/iso_14064_climatechange [last viewed 19th March 2009] 

Kollmuss, A., Zink, H., & Polycarp, C. (2008) Making Sense of the Voluntary Carbon Market. A Comparison of 

Carbon Offset Standards. Stockholm Environment Institute. Tricorona, Sweden. 

Leach, J.C. & Madhaven, A.N. (1993) Price Experimentation and Security Market Structure. Review of 

Financial Studies, Volume 6, Number 2, pp 375-404.



Madhaven A (1995) Consolidation, Fragmentation and the Disclosure of Trading Information. Review of 

Financial Studies, 8, 579-603. 

Madhaven, A. (2000) Market Microstructure: A Survey, Journal of Financial Markets, 3, 205-258. 

Merger, E. (2008) Forestry Carbon Standards 2008 - A comparison of the leading standards in the voluntary 

carbon market and the state of climate forestation projects. Carbon Positive. 

Mendolson, H. (1987) Consolidation, Fragmentation and Market Performance. Journal of Financial and 

Quantitative Analysis, 22, 189-208. 

Parker, C., Mitchell, A., Trivedi, M., & Mardas, N. (2008) The Little REDD Book. Global Canopy Programs, 

Oxford, UK. pp 113. 

Nelson, J.P. Market Structure and Incomplete Information: Price Formation in a Real-World Repeated English 

Auction. Journal of Economic Behvaiour and Organisation , Volume 27 Number 3, August 1995, pp 

421.237. 

NERA Economic Consulting. (2007) Market Mechanism for Reducing GHG Emissions from Agriculture, 

Forestry and Land Management. Department for Environment, Food and Rural Affairs. United Kingdom. 

NSW GGAS Registry - http://www.ggas-registry.nsw.gov.au/ [last viewed 19th March 2009] 

Plan Vivo Standard - http://www.planvivo.org/fx.planvivo/scheme/news.aspx [last viewed 19th March 2009] 

Ravindranath, N.H., Ostwald, M. (2008) Carbon Inventory Methods: Handbook for Greenhouse Gas Inventory, 

Carbon Mitigation and Roundwood Production Projects, Springer, 304 pages. 

Redmond, W.H. (1989) The Effect of New Product Pricing on the Evolution of Market Structure. Journal of 

Product Innovation Management. Volume 6, Number 2, June 1989, pp 99-108. 

Regi - http://www.regi.co.nz/regPublic/registry.mt_public.home [last viewed 19th March 2009] 

Senn, M. (2002) Supermontage in the American Securities Market Context. Journal of International Financial 

Markets. Volume 4. Issue 6. 2002. 

Seifert-Granzin, J. (2007) The Noel Kempff Climate Action Project: Challenges for Project Base Approaches in 

future REDD Schemes. Presentation to the Conservation Finance Workshop, Vlim, 29th July-Aug 3 rd , 

2007. 

Social Carbon Standard - http://www.socialcarbon.org [last viewed 19th March 2009] 

Social Carbon registry - http://www.socialcarbon.org/Registry/ [last viewed 19th March 2009] 

Stern, N. (2006) The Economics of Climate Change. The Stern Review. Cabinet Office, HM Treasury, 

Cambridge Press, UK. 

The Carbon Neutral Protocol - http://www.carbonneutral.com/uploadedfiles/TCNC%20Protocol%202008.pdf 


The Carbon Neutral Offset Project Registry - http://www.carbonneutral.com/pages/Offsetprojectregistry.asp [last 

viewed 19th March 2009] 

The Globe Carbon Registry - http://www.globecarbonregistry.com/ [last viewed 19th March 2009] 

The Gold Standard - http://www.cdmgoldstandard.org/ [last viewed 19th March 2009] 

Triodos climate Clearing House - http://www.triodos.com/com/climate/ [last viewed 19th March 2009] 

UGS PLM Solutions (2004) Reducing Complexity. NX Digital Product Development White Paper. 

UNFCCC (2007). COP 13. Revised draft decision -/CP.13 Ad Hoc Working Group on Long-term Cooperative 

Action under the Convention. FCCC/CP/2007/L.7/Rev.1 14 December 2007. Bali. 

VER+ Standard - http://www.tuev-sued.de/uploads/images/1179142340972697520616/Standard_VER_e.pdf 


Voluntary Carbon Standard - http://www.v-c-s.org/afl.html [last viewed 19th March 2009] 

Voluntary Offset Standard - http://www.carboninvestors.org/documents [last viewed 19th March 2009] 

Willett, Meyer. “Over-the-Counter Possibilities.” Wall Street Journal (December 28, 1931): p11.



TASMANIA’S NEW FOREST INDUSTRY PLAN: 

RESHAPING THE FORESTRY AGENDA 

Aidan Flanagan 

ABSTRACT 

Tasmania is now past the half-way mark in relation to the Regional Forest Agreement. 

Over the last 10 years significant changes have occurred within the sector, reflecting 

community, environmental, commercial and political pressure. To ensure that 

Tasmania’s future incorporates a strong forest industry, the Forests and Forest Industry 

Council of Tasmania has facilitated the development of a New Forest Industry Plan 

(NFIP) which will fully utilise over 10 million tonnes of wood fibre annually by 2020. 

The NFIP contains commitments, strategies and actions which build on Tasmania’s 

competitive strengths and deliver investments which focus on: a) creating wealth by 

maximising the value of forest products and delivering positive economic benefits to 

regions; b) delivering rewarding careers for our children; c) creating stronger community 

engagement initiatives and responding to community concerns; d) generating healthier 

environments and protecting our natural values; and e) adapting to climate change and 

reducing the State’s carbon emissions. 

OVERVIEW 

Tasmania’s native forests have ancient origins that influence their unique characteristics, how they can 

be used and why they are highly valued by Tasmanians. Forests are an integral part of the landscape. 

These forests also produce some of the strongest and most beautiful and valuable timbers in the world 

which are renowned for their texture and character. In addition to the versatile Eucalypts (dominated 

by E. delegatensis, E. regnans and E. obliqua), other important species include Blackwood, Myrtle, 

Sassafras, and Celery Top, Huon and King Billy Pine. 

Figure 1. Tasmania’s land use 

In Tasmania, the forests and forest industry has a long history of effective innovation. It is an engine 

for advances in improved technology and management systems, and a driver of regional economies 

and establishing social bonds that make it a special feature of the Tasmanian way of life. 

Forests and Forest Industry Council of Tasmania, Level 5-2 Kirksway Place, Battery Point Tas 7004. 

Ph +613 62338221. Email aflanagan@ffic.com.au



Forest operations around the world have gone through dramatic changes over the last 20 years and 

Tasmania is no exception. Gone are the days when a forest operation included a crew of men on 

chainsaws working in front of a relatively low cost tractor. Today’s forest industry is a high tech 

industry. 

Today, only 40% of Tasmania’s forests are potentially available for wood production (as indicated in 

Figure 1). However, the areas available for harvesting are likely to be less than the 40% identified as 

areas potentially available for timber production are often discounted. In Tasmania, these discounts 

include the Forest Practices Code and other conservation provisions, accessibility, economics, safety 

and operational constraints apply. 

The continuing success of the industry has been achieved by focusing on maximising value, improving 

productivity and profitability, committing to sustainable management practices, investing in new 

technologies and skilled career development, and responding to market demands. In summary, 

maximising value has been about adopting world’s best practice in a competitive market, and then 

adapting these to the Tasmanian forest environment. 

Its natural resources make Tasmania a State of opportunity which can build on its competitive 

strengths and deliver investments which focus on: 

creating wealth by maximising the value of forest products and delivering positive economic 

benefits to regions; 

delivering rewarding careers for our children; 

creating stronger community engagement initiatives and responding to community concerns; 

generating healthier environments and protecting out natural vales; and 

adapting to climate change and reducing the State’s carbon emissions. 

Tasmania’s forests and forest industry’s international competitive position is supported by the security 

of access to native forest resources provided under the Regional Forest Agreement (RFA); clear 

regulatory processes delivered through the Forest Practices Act; a growing hardwood pulpwood and 

sawlog plantation resource supported by Plantations Australia: the 2020 Vision; a diverse private 

contracting and processing sector; and highly skilled employees. 

In 1985, Tasmania became the first state in Australia to introduce a comprehensive planning system 

for forestry, encompassing forest practices legislation, policies and processes. 

The Forest Practices Act 1985 provides the world class framework under which balanced triple 

bottom-line principles are monitored and regulated “to achieve sustainable management of Crown and 

private forests with due care for the environment”. Tasmania remains a leader in forest regulatory and 

management practices. This system was independently assessed by researchers from Yale University 

who concluding in their 2008 report that: 

“Tasmania is unique among case study Organization for Economic Co-operation and 

Development jurisdictions in applying the same forest practice policies to both public and 

private lands in regards to riparian buffers, clearcut size limits, reforestation and road 

building 1 ”. 

Over the last decade, the Tasmanian RFA has supported a culture of continuous improvement and 

adaptive management which are embraced and driven by the forest and forest industry, and accredited 

under the Australian Forestry Standard. Security provided under the RFA has underpinned 

investments in the industry which has diversified agricultural economies by injecting substantial 

capital into the State. Since 1997, the industry has generated over $14 billion in turnover, $2.5 billion 

in wages, and $2.3 billion in new forests, new technologies and new processing capacity. The RFA 

provides a balanced approach for managing natural and heritage conservation values in Tasmania’s 

forests. This was confirmed in 2008, when the joint UNESCO World Heritage 

1. McDermott C.L., Cashmore B. and Kanowski P, ‘A Global Comparison of Forest Practice Policies Using Tasmania as a 

Constant Case’, viewed at



Centre/IUCN/ICOMOS mission report on the conservation of the Tasmanian Wilderness World 

Heritage Area and concluded that: 

“The area managed under the TWWHA management plan provides a good representation 

of well-managed tall Eucalyptus forest and there is similar forest outside the property 

which is also well-managed, but for both conservation and development objectives.” 

And 

“The threats to these forests from production forestry activities are well managed and 

there is no need for the boundary of the property to be changed to deal with such 

threats.” 

Ongoing support for the RFA and balanced regulatory policies will support an expansion in 

investment confidence by reducing sovereign risk and providing certainty for financial institutes to 

release the capital necessary for businesses to operate. It will also provide resource security and allow 

the industry to respond and contribute positively to the environmental challenges posed by climate 

change, by adopting new practices, investing in and developing new products and markets, and 

delivering community focused outcomes. 

Tasmania’s forests have also been a focus for active and robust community debate. This debate has 

reflected the variety of views and values held by people about the use and management of Tasmania's 

forests. These differences in views relate mainly to the value that individuals and groups place on 

forests and the wood products that flow from them. The espoused values are shaped by many factors, 

including economic dependency on forests. That is, the views held by tourist operators, contractors, 

apiarists, sawmillers, and craft wood users, differ from those with a non-commercial dependence, 

such as recreationists and tourists. 

Irrespective of these differences, it is clear that Tasmania has now achieved a balance between the 

needs of forest-based industries, the community, existing agricultural activities, and the environment. 

Research undertaken by FFIC (unpublished) indicates that the majority of Tasmanians support 

appropriate ‘triple bottom line’ outcomes from sustainable forestry activities. This research has shown 

that employment opportunities, retention of young people within the State, and wealth creation are 

recognised and valued by many within the wider community, along with forest-based tourism, 

indigenous participation, and environmental and cultural values. 

Ultimately, the success of the Tasmanian forests and forest industry will be determined by investment 

decisions, made outside of government, which are driven by innovation, financed by private 

investment, and responsive to market and community signals. 

It is now appropriate that the future forest industry focuses on new approaches and new opportunities. 

In order to support a future within which the Tasmanian forests and wood products industry has a lead 

role, the New Forest Industry Plan (NFIP) was developed to shape positive future government and 

industry programs and policy initiatives, and provide a sound basis for future planning and investment 

decisions which promote growth and innovation across Tasmania’s forest and forest industry supply 

logistics and value chains. 

The NFIP has been developed by applying three principles. 

1 There is only one forest industry in Tasmania and it is about highly skilled people and 

communities. Over 10,000 people and their families are directly dependent on the industry for 

employment and a sense of ‘place’. 

2 While the industry is made up of many businesses and many people, all are interdependent 

and reliant on the survival and prosperity of the supply, logistic and value chains as a whole. 

3 Good land management in Tasmania is no longer restricted by boundaries and there is no 

longer a need to extend reservations or further restrict access to forest resources. 

The NFIP builds on the success of the 1991 Secure Futures for Forests and People: the Forests and 

Forest Industry Strategy, and the 1997 Tasmanian RFA. It is also consistent with the National Forest



Policy Statement and Plantations for Australia: The 2020 Vision. The NFIP is an industry plan, and 

was developed from input received through a series of industry workshops, and a number of broader 

symposia and strategic meetings held in early 2009. The NFIP provides strong commitments. These 

are supported by strategies, priority actions and general actions which promote investment and growth 

opportunities identified under five Elements: 1) Wealth Creation; 2) Industry of Choice; 3) 

Community Engagement; 4) Healthier Environments; 5) Climate Change. 

The NFIP provides the framework for industry, governments and the wider community to work 

constructively together to drive innovation and stimulate economic development essential to sustain 

services that support our aging population. It provides opportunities to create rewarding careers for our 

children, and create stronger links within the community so that Tasmanians can be confident that our 

forests are sustainably managed and our forest products meet international certification standards. It 

will create healthier environments by adopting new approaches which provide solutions to 

environmental challenges facing Tasmania, including climate change. 

The NFID also recognises that there is only one future for Tasmania’s forest industry. It is a future 

that incorporates native and plantation forests, and private and public forests. It includes saw and pulp 

logs; paper and specialty manufacturing; growing trees and processing wood; large companies and 

small businesses. It also consists of contractors and their employees; honey production and tourism; 

furniture and craft wood products; recreation and cultural heritage; and bioenergy and carbon storage. 

The NFIP recognises that the growth in the plantation estate, increased productivity from native 

forests, higher value processing and market development are creating opportunities for growth. It also 

sets forth commitments and strategies which will be progressed by supporting regulatory systems 

which promote open, competitive and non-discriminating markets. 

Tasmania’s extensive plantation and native forest resources are expected to provide over 10 million 

tonnes of wood products annually by 2020. This expanding resource creates opportunities for existing 

businesses to expand, and for new industries to develop. Over $2.1 billion of investment in new 

processing opportunities has been identified. These investments would benefit Tasmania by directly 

creating up to an additional 855 long term, highly technical career opportunities and directly generate 

an extra $1.17 billion annually in wealth to the state’s economy. Table 1 summarises income and 

employment impacts of new processing investments opportunities within Tasmania 2 . 

Table 1. Income and employment impacts of new processing investments 

Facility 

Minimum 

capital 

investment 

($ million) 

Annual direct 

income 

($ million) 

Long term 

employment in 

new facilities 

(No. of jobs) 

Input volume of 

wood required 

(million m 3 or tpa) 

Hardwood pulp mill 1,450 660 290 4 Mt 

ESL plant 225 290 150 0.55 Mt 

Hardwood sawmilling 

- 3 reciprocated mills 

60 50 200 0.24 Mm 3 

Hardwood sawmilling 

65 50 65 0.25 Mm 3 

- 1 linear mill 

Hardwood plywood 15 20 50 0.25 Mm 3 

Softwood sawmilling 10 30 30 0.165 Mm 3 

Bioenergy – (based 

on 3 bio-electricity 

plants and 1 export 

wood pellet plant) 

370 120 165 1.15 Mm 3 

Total (reciprocated 

hardwood sawmills 2,130 1,170 885 

option) 

Total (linear hw 

sawmill option) 

2,135 1,170 750 

2 URS Forestry report, 2009. “Economic Impacts of Potential Forest Industry Developments in Tasmania’.



There are also opportunities to increase supply chain efficiencies, enhance forests management, reduce 

the threat of wildfire and increase the use of wood products. These opportunities can, in themselves, 

significantly improve competitiveness by increasing value recovery across the supply and logistic 

chains. Figure 2 displays the value recovery opportunities within the harvesting and haulage sector 3 . 

Sub optimal log 

making 

52% 

Felling Damage 

7% 

Thinning 

2% 

Sub optimal choice 

grades or stands 

30% 

Extraction breakage 

2% 

High stumps and 

butt damage 

5% 

Log Making 

Damage 

2% 

Figure 2. Value recovery opportunities within the harvesting and haulage sector 

These new opportunities, when added to the servicing of existing processing and manufacturing 

businesses, will provide secure opportunities for forest contractors. These opportunities will require 

progressive investments over 10 years in new harvesting machinery of around $230 million and 

transport investments of around $135 million. Investments in harvesting and haulage contracting 

capacity will directly support over 650 jobs. This level of investment will generate revenues of $240 

million annually, while delivering efficiencies and reducing delivery costs. 

However, the transition from native to plantation forest resources is reliant on the resolution of a 

number of technical and economical challenges, including loss in the value of dry output associated 

with creating high value appearance products from plantation grown Eucalyptus nitens. Currently E. 

nitens is unsuitable as a replacement for traditional native forest products, particularly within the solidwood 

‘appearance’ grade market. Further research is required to evaluate the increased suitability and 

utilisation of E. nitens through improved breeding and silviculture, harvest and transport handling 

systems, and sawing and drying methods. 

Maximising the value of Tasmania’s forests and forest products also will benefit the wider Tasmanian 

community. Applying standard forestry employment and income multipliers, additional support and 

service investments would generate at least an additional 1000 jobs and around $1 billion in direct and 

indirect income within the State. Service and support sectors which are likely to benefit from 

increased forest processing and manufacturing capacity include: 

• housing construction and accommodation – increased regional employment will stimulate 

demand 

• accounting, financial and administration services – there is likely to be demand for capital 

financing support, accounting and administration services 

• community services – increased regional employment will require community services, such 

as schooling, health services, and social and sporting facilities 

• communication services – efficient communication networks will be required to support sales 

and marketing systems 

• engineering, mechanical and technical services – these include computer maintenance, 

electrical, mechanical, civil and structural skills required during the design and construction 

phase, as well ongoing plant and equipment maintenance requirements 

• environmental services – construction and operations will require ongoing environmental 

assessment and monitoring services 

3 CRC for Forestry, ‘Maximizing Value Recovery along the Forest-to-Mill Supply Chain’, June 2009 workshop.



• labour services – ongoing career and skill development services will be required, and will 

drive demand for education and training support. 

• transport – increased movement of people (employees and their families) and materials will 

generate demand for efficient transport services within regions, and between regions and ports 

CREATING WEALTH 

Tasmanian forests produce around six million cubic metres of logs annually. These are used to make 

sawn timber and panels for homes and commercial buildings; industrial material for construction; 

unique and high quality flooring, joinery, designs in furniture and craft products; composite products 

such as Timbercrete (a brick which incorporates plantation timber waste with cement, sand and 

binders); and a range of high value products such as writing, packaging, tissue and other paper 

products. Forest and mill residues may also be used to produce landscaping material, firewood and 

bioenergy. Most of these products are sold within Australian markets, although exports are an 

important part of the industry. 

This diversity in processing and manufacturing has helped to cushion our forest-reliant businesses, 

employees, their families and many rural communities from prolonged and deepening impacts of 

market fluctuations experienced by many other sectors of Tasmania’s economy. 

Figure 3. Composition of the forests and forest industry employment in Tasmania 

Forestry and timber processing continues to contribute significantly to Tasmania’s economic 

prosperity and social wellbeing. The forests and forest industry: 

• is the second highest manufacturing contributor to Tasmania’s gross state product and 

contributes up to $1.6 billion to the State economy; 

• is an integral part of the Tasmanian community, and directly employs over 6000 people 

(Figure 2) and in excess of 10,000 in businesses providing support and services; 

• encompasses 1600 private forest growers (the majority being farm based); 500 businesses 

which includes large forest growers, contracting and transport providers, and processors; 

• is unique in that it is a net exporter of forest products; and 

• is a nationally important contributor to reducing our reliance on imported wood products.



If the forests and forest industry is not investing, developing and applying innovative solutions and 

approaches, and improving competitiveness, its ability to adapt and continue to provide ongoing 

benefits to the economy and community of Tasmania will be diminished. 

People within the forest and wood products industry have demonstrated a commitment and capacity to 

adapt to change by embracing technology, enhancing their skills and thus broadening their career 

opportunities. This commitment has resulted in a rationalisation of opportunities within certain sectors 

in the short to medium term, while providing new career opportunities in others. 

The relatively recent change in resource reliance from old-growth to smaller diameter regrowth and 

plantation sawlogs has resulted in significant adaptation along the supply and logistics chains, with 

innovation playing a major role in improving timber utilisation by focusing on maximising recovery. 

The growing and processing of smaller diameter logs is driving the adoption of new approaches to 

forest management, harvesting and transport, mill processing and drying, business management and 

market development. 

The increasing importance of carbon storage will require innovative solutions to achieve more 

accurate systems for measuring, accounting and reporting carbon sequestration within plantations and 

native forests, as well as assessing new products to reduce carbon emissions, such as bioenergy and 

biofuels. 

The NFIP provides a range of options for what the industry could look like in 10+ years and highlights 

areas of innovation to: 

develop and adopt new processing and re-tooling technologies to deal with changing resource 

qualities and allow for more efficient utilisation of resources; 

respond to emerging markets for new products that may arise from global initiatives to combat 

climate change; 

drive efficiency and add value across the supply and logistics chains; 

adopt new designs and marketing strategies which strengthen consumer linkages; 

achieve the necessary scale required to remain internationally competitive; 

invest in people and create skilled careers; 

adopt strategic management techniques to target high value market opportunities with 

products that feature Tasmanian timber; and 

invest in sophisticated sales and distribution systems which respond to information and 

products sought by architects and specifiers. 

The NFIP identifies value adding opportunities for a range of new approaches in forest management, 

in supply and logistics, and in processing. These opportunities include the need to expand our existing 

pulp and paper production and to restructure existing hardwood sawmilling capacity to include greater 

utilisation of plantation resources. It also includes the establishment of engineered wood product 

production facilities, an ongoing need to focus on high value appearance type products for domestic 

and export markets, and the establishment of sustainable bioenergy production. These opportunities 

will require ongoing innovation, modernisation and development by industry partners, and should be 

viewed as evolutionary, reflecting a continuum in adding value rather than radical change. 

Tasmania’s extensive plantation and native forest resources are expected to provide up to 10.5 million 

tonnes of wood products annually (22.3% from native forests). This expanding resource creates 

processing investment opportunities totalling over $2.2 billion. These will benefit Tasmania by 

directly creating an additional 850+ long term, highly technical career opportunities and generate an 

extra $1.1 billion annually in wealth to the State’s economy. 

In addition, these new opportunities, will provide secure opportunities for forest harvesting contractors 

and transport providers,, and a wide range of service providers. 

Discriminatory or inefficient policies will increase sovereign risk and undermine investor confidence, 

jeopardise critical developments and place at risk Tasmania’s future economic and job creation 

prosperity. The NFIP recognises the need for governments to review polices to reduce sovereign risk



and support economic growth. Strong, consistent and non-discriminatory regulatory systems will 

continue to support investments in Tasmania only where such a balance is achieved. The NFIP also 

identifies further scope to refine existing frameworks to ensure they deliver enhanced security and 

thereby contribute to attracting new investment. 

The extensive involvement of governments in promoting sustainable regional development means that 

they have an important role in developing a strong economy through a consistent approach to regional 

planning, and investing in integrated infrastructure networks which support efficiencies across supply 

chain networks and promote international competitiveness and market development. 

INDUSTRY OF CHOICE 

The NFIP recognises that people are the State’s most valuable natural resource. Tasmania’s success in 

today’s competitive, global economy depends on the professionalism and skills of the people 

employed in businesses, and the daily interactions in workplaces. In effect, the success of the 

Tasmanian forests and forest industry relies on it being seen as ‘an industry of choice’ with attractive 

careers that offer interesting work, safe and healthy work conditions, ready access to training and skill 

development, and remuneration of an acceptable level. 

The strength of this State’s economy relies on local businesses to provide career opportunities while 

maximising productivity and maintaining viable regional communities. However, Tasmania is 

entering ‘the age of the ageing’, and this demographic change is impacting on the industry’s ability to 

attract and retain skilled employees at all levels of the labour force. This is constraining industry 

development and investment by increasing costs and uncertainty for investors. 

Tasmania’s forest resources continue to change and the skills required in the areas of technology and 

environmental management will continue to evolve. Due to these changes, employees will require a 

broader and higher level of skills, and career development will require the support of industry decision 

makers and governments to meet these evolving requirements. 

Government policies must support stable employment, create opportunities for new careers, and 

ensure Tasmania’s economy remains strong during difficult times. The NFIP identifies the lead role 

that governments have in facilitating stronger links between industry, education and training 

organisations. We must also build on the strengths and synergies provided by vocational training and 

higher education structures while fostering the development of systems that support career 

development, not just job placement. 

The NFIP identifies future initiatives which will need to deliver targeted, relevant and flexible training 

systems, which attract and retain the people the forest industry requires if it is to become an industry 

of choice. 

Tasmania’s forest industry is committed to the continuing provision of career opportunities for 

indigenous people, and is proud to have the highest proportion of indigenous employees in the State. 

The industry supports initiatives, such as the National Indigenous Forestry Strategy. This provides a 

framework for industry to work with indigenous communities to build partnerships which focus on the 

creation of career opportunities across all sectors while protecting cultural values, and encouraging 

greater participation in land management. 

The NFIP identifies further employment and career development opportunities across many different 

disciplines and supports strong and stable communities. The industry is committed to working with 

governments and the community to develop a stronger and more resilient economy, through a 

consistent approach to regional and career project planning. 

COMMUNITY ENGAGEMENT 

Community-focused engagement is a central platform of the NFIP. The industry is committed to 

addressing community concerns by developing innovative solutions which are supported by 

scientifically based evidence, better practices and supportive policies. 

In Tasmania, rural populations continue to decline as families leave the regions, and farms are 

amalgamated into larger holdings. This has long been a global trend, but one which has been



accelerated in Tasmania by the adoption of mechanisation which requires fewer workers, depressed or 

volatile prices for products, encouragement of youth to pursue alternative city-based careers, and 

family hardship associated with ongoing drought. 

During these times of change, improved management of private forests and expanded plantation 

development have helped to maintain rural populations. Thus, the expansion of plantations is one 

aspect of the continuum of change to rural landscapes, and helps maintain a viable traditional 

agricultural communities. 

It is therefore important that government and industry policy fosters innovative new investments in 

regional communities and alternative land uses as a means of creating employment, retaining young 

people, and maintaining regional economic, community and environmental values. The Tasmanian 

forests and forest industry brand will help to build and sustain community trust by promoting 

initiatives, such as the ‘Forestry Good Neighbour Charter’, and the uptake of forest certification 

systems. The industry is also committed to developing clear, and consistent approaches for 

communicating achievements in sustainability and regulatory compliance; and promoting public 

understanding of the environmental and carbon-friendly credentials of sustainable forestry. 

HEALTHIER ENVIRONMENTS 

Tasmania still has a higher percentage of native forest than any other State in Australia, and some 65% 

of this native forest is not available for wood production. 

Tasmania’s forests and forest industry is committed to improving the health of the urban and natural 

environments. It will achieve this by: 

promoting independent certification schemes; 

reducing forest impacts on the community’s health and lifestyle ; 

enhancing biodiversity values; 

conserving cultural heritage sites; 

reducing the threat of wildfire; 

improving the quality of water; and 

reducing our reliance on chemicals. 

Scientific research supports the view that conservation and biodiversity values are enhanced through 

the management practices being implemented by Tasmania’s forests and forest industry. Managed 

forests can redress many of the environmental and social problems facing rural and regional 

communities if they are strategically planned and located. 

Commercial forestry activities fund environmental and heritage initiatives which support publicly 

funded programs with over 23,500 hectares of formal reserves having been established under Forest 

Practices Plans lodged with the Forest Practices Authority, which administers the Forest Practices Act 

1985. 

Private forest owners manage more land for non-commercial returns than any other sector of the 

economy. In total, over 68,000 hectares of private plantation (19%) land are managed for noncommercial 

returns such as conservation, water quality enhancement and fire protection — all without 

public funding. In addition, over 600,000 (>50% of total) of State multi use forests are also managed 

for non-commercial returns. 

These commitments and achievements are being supported by the adoption of independent, 

international recognised forest management and sustainability standards. Over 1.83 million hectares 

of managed forests in Tasmania are accredited as being sustainably managed under the Australian 

Forestry Standard (AFS), which is endorsed internationally by the Programme for the Endorsement of 

Forest Certification — the world's largest forest certification body. 

The NFIP recognises the potential for organisations and governments to undermine the significant 

environmental and community benefits of the AFS certification processes through the adoption of 

restrictive trade and discriminatory or punitive approaches to certification standards.



Consequently, the NFIP identifies the need for governments to adopt appropriate procurement 

policies, and build standards and associated regulations which recognise, , all internationally 

accredited, independent national forest certification schemes. Governments should act against the use 

of public funds to support organisations and businesses that restrict trade by discriminating against 

Tasmanian timber which meets third party certification. 

CLIMATE CHANGE 

The forestry and forest products industry helps to reduce the adverse impacts of climate change 

through reducing Tasmania’s reliance on imported coal-based electricity generation. From a national 

perspective, the total annual increase in carbon stored in managed native and plantation forests is 2.4 

times that lost through decay of forest wastes produced during harvesting. The industry’s ability to 

capture carbon within its forests and store it in wood products has reduced the State’s total carbon 

emissions by nearly 30% as indicated in Figure 4. 

million tonnes CO2-e 

3.0 

2.0 

1.0 

0.0 

-1.0 

-2.0 

-3.0 

Agriculture 

Figure 4. Tasmania's GHG emissions by sector 4 

Transport 

Energy: 

Manufacturing and 

construction 

Industrial 

processes 

Energy: Electricity 

.The NFIP has identified strategies and actions which will help the industry to reduce the State’s 

carbon emissions and increase carbon storage capacity by: 

capitalizing on carbon stored in forests through enhanced management practices; 

maximizing the use of wood products in residential and commercial constructions; 

expanding the capital value of farms by integrating forestry to offset on-farm carbon 

emissions; and promoting renewable bioenergy opportunities. 

To achieve the benefits provided through these initiatives, the NFIP advocates a government review of 

policies and the removal of impediments which distort markets by restricting the management and 

expansion of forests, and discriminate against the use of wood products in construction and the 

generation of power and heat. 

CONCLUSION 

Tasmania’s future economic and social growth relies on a strong forest industry embracing change by 

focusing on maximising value across the supply, logistics and value chains. The career development 

opportunities and benefits offered by an innovative and profitable industry are significant. The NFIP 

will be achieved by developing partnerships between the industry and the community, and focusing on 

building trust and delivering on commitments. Tasmania’s forest industry can help in establishing a 

healthier Tasmania by providing solutions to environmental challenges, including climate change. 

The industry has a long history of innovation, and the NFIP represents a commitment by individuals 

and businesses to continue to adapt to change, and embrace solutions and technologies which meet the 

social, environmental and economic needs of the State. 

4 DCC (2008b). State and Territory Greenhouse Gas Inventories. Department of Climate Change. Available at: http://www.climatechange.gov.au/inventory/2006/index.html 

generation 

Energy: Other 

industries 

Waste 

Land use change 

Afforestation & 

Reforestation



ABSTRACT 

RESHAPING THE FORESTRY AGENDA IN QUEENSLAND 

Sean Ryan 1 

Private Native Forestry will need to fill the coming void in hardwood production in 

Queensland. In 15 years with the closure of State Native forests and the production lag 

in the fledgling State-based nanoscale hardwood plantation program, there needs to be 

a seismic shift in forestry foresight and planning. It needs to be recognised that 

plantations with their significant initial carbon footprint, genetic pollution and cost 

factors are not the only solution to future production scenarios. In SE Queensland there 

are over 1.2 million hectares of remnant classified private native forests, mostly in a 

very unproductive condition due to lack of management. However, more importantly, 

from a future productivity point of view, there is in excess of 500 000ha of regrowth, 

non-remnant classified forest. This young regrowth has generally not had the negative 

impacts of high-grading and suppression evident in our mature forests, and is still in or 

near ‘free-growth’ mode. Thinning trials established by Private Forestry Southern 

Queensland in 10 different locations and forest types across south east Queensland, and 

the commencement of a thinning harvest of 1000ha of regrowth from a 20 year old 

clearfall operation, has demonstrated a cost effective, environmentally sound system to 

rapidly fill the hardwood production shortfall. 

The paper outlines the initial condition of the regrowth forests, the process involved 

including costs and growth data gathered to date and the improvements in forest health, 

ground cover and forest productivity achieved. It will also outline the potential extent 

of the resource, forecast product range and potential growth rates. 

INTRODUCTION 

Australia has supported a 2 billion dollar national forest product deficit for many years, which includes 

a sawn wood deficit of 450 000 m³. Over the last ten years the trade deficit (the extent to which 

imports exceed exports) has been running at a fairly constant level between $1.7 and $2.2 billion per 

year with an average of $1.9 billion (Stanton 2005). 

Queensland is a net importer of overseas forest products, with an overall trading deficit of about $447 

million in 2002/03, an increase of about $166 million, compared to a decade earlier. In 1991/92 

approximately 2 per cent of the total turnover of the forest industry was sold overseas. By the end of 

the decade, in 1999/2000, this ratio had risen to 5.5 per cent. 

The import penetration of the Queensland market for forest products (that is, the share of imports in 

total estimated sales to the domestic market) rose from 12.6 per cent in 1991/92 to 17.5 per cent in 

1999/2000. An estimated 860 000 cubic metres of sawn timber products came out of the Queensland 

industry in 2002/03, of which 88 per cent was sold locally. Queensland consumes 1.1 million cubic 

metres of sawn timber product each year. The current hardwood sawn timber deficit of around 100 

000m³is expected to increase to around 200 000 m³ by 2015 (Fegely & Follas, 2006). 

About two-thirds of Queensland’s overseas imports of forest products are sourced from producers in 

four countries, New Zealand, Indonesia, Malaysia and China, with about 33% ($216 million) sourced 

from New Zealand (Marohasy 2005). 

1 

Executive Officer, Private Forestry Southern Queensland, 224 Mary Street Gympie 4570. Ph. 07 54836535. 

Email -pfsq@bigpond.com.



CURRENT PRODUCTION 

Forestry Plantations Queensland (FPQ) states in its 07/08 Annual Report that its business objectives 

are to operate as a responsible commercial plantation forest manager and maximise the market 

value of its assets consistent with: 

- best practice principles for sustainable plantation forest management; 

- the directions of its responsible Minister; repairs 

- the operations and financial parameters of its operational plan; 

- the Queensland government’s commitments for hardwood plantations. 

Although FPQ does not have as a specific business objective to increase the supply of timber to meet 

Queensland’s Timber deficit, it does produce 1.26 million m³ of exotic pine sawlogs and .42m m³ of 

Hoop Pine sawlogs and purchased an additional 4,942ha of land suitable for exotic pine plantation 

development in the 07/08 financial year. (07/08 Annual report). 

POLICY PLANNING VOID 

Queensland does not have a development plan to address the trade deficit. It has been 15 years since 

the Federal and State Governments signed the National Forest Policy Statement (1992). This 

Statement outlined agreed objectives and policies for the future of Australia's public and private 

forests. 

The strategy and its policy initiatives were to lay the foundation for forest management in Australia 

into the next century. Some of the Policy Initiatives specific to Private Native Forests in the document 

include: 

• Sustainable management of private native forests will be encouraged through a combination of 

measures that may include dissemination of information about, and technical support for, forest 

management, education programs, conservation incentives, land-clearing controls, harvesting 

controls, and codes of forest practice. 

• Encouraging private landowners to manage forests for long-term economic use by removing any 

unnecessary impediments or disincentives. The Governments will develop a range of incentives 

and programs to promote sustainable management of native forests on private land. These 

incentives and programs will be designed to ensure active management of private native forests 

for both ecologically sustainable wood production and nature conservation, so that the private 

native forest estate will remain a permanent resource. 

In his recent Statement to Federal Parliament ‘Preparing our Forest Industries for the Future’ Tony 

Burke, Minister for Agriculture, Fisheries and Forestry, stated that ‘The Government committed $20 

million to assist industry and support jobs, particularly in regional communities, through measures 

that: 

• invest in value-adding – through the Forest Industries Development Fund; 

• address long-term skills and training shortages – through the new Forest and Forest Products 

Industry Skills Council and database development; 

• deal with climate change – by addressing major knowledge gaps; and 

• work with our Asia-Pacific neighbours and industry to tackle illegal logging – by investing in 

capacity building, certification, improving governance and in developing a Regulatory Impact 

Statement’. 

In Queensland’s hardwood industry the problem is not the processing industry; it has always been the 

resource, or lack of it, that has resulted in seven hardwood mills closing down in SE Queensland in the 

last 12 months. Dr Gary Bacon in a recent interview for the in-wood magazine states ‘I am saddened 

by this because native forests management is a decentralised regionally-based activity. On one hand 

you hear this litany from government jobs, jobs, jobs, and yet they take away from these communities a 

sustainable production from native forests’.



The failed RFA process committed many millions of dollars to the mills with state hardwood 

allocation for retooling and value adding, but nothing to improving the productivity of the resource, 

where the problem clearly lay. 

Minister Burke’s statement went on to say ‘The Rudd government remains fully committed to RFAs as 

the primary mechanism to sustain jobs and support industry, to ensure high conservation values, and 

for the protection of biodiversity and threatened species.’ But of course Queensland does not have an 

RFA and there is no State-based management plan for the future of the hardwood industry other than a 

commitment to a total of 20,000ha of hardwood plantation to replace the 3 million hectares of Native 

hardwood resource. 

THE REGROWTH RESOURCE 

PFSQ’s mapping shows there is over 1.2 million 

ha of remnant classified private native forest in SE 

Queensland. Remnant classification was given to 

any vegetation considered to meet three criteria, 

namely: 

1. 50% of the original canopy cover 

2. 70% of the original canopy height and 

3. supporting the original species mix 

Surrounding and interspersed with the remnant 

forests is a similar area of non-remnant classified, 

regrowth forest. It is this resource that holds the 

key to a highly productive native forest resource. 

The majority of remnant forests have been subject 

to 100 years of opportunistic harvesting with little 

or no follow up management. Stocking rates of 

800-1000 stems/ha over 10cm dbh is not 

Map 1. (Above) Aerial Photo showing 

remnant overlay and non-remnant 

d 

Map 2. (Left) Showing Plot location for: 

Spotted Gum remnant 

Non-remnant 

uncommon, and without a radical silvicultural 

reset these forests tend to demonstrate a very 

low productivity with diameter growth of 0.1- 

0.3cm per year. PFSQ’s research has 

demonstrated that even after a radical thin, 

existing stems take up to five years to regain a 

reasonable growth rate. Even then is at best 50% of that expected from a tree never subjected to 

excessive competition or suppression. It is not until the regeneration from the thin comes through that 

the productivity will reach acceptable levels.



PLOTS AND DEMONSTRATION SITES 

PFSQ has established 11 demonstration sites across SE Queensland over the last ten years (Map 2.), 

five in remnant forests and six in non-remnant forests. The plots are generally within the 900-1200mm 

rainfall zone dominated by Spotted Gum (7 of the 11 sites). Most sites have 3 replicated thinning 

regimes (nominally 200, 120, and 80 stems/ha) and a control. (Full site description and procedures 

have been written up as case studies for each site and appear at www.pfsq.net) 

The demonstration sites were implemented to ascertain the growth rates currently being achieved from 

our unmanaged private native forest resource and what management procedures need to be 

implemented to improve that productivity. There was little or no information on the dynamics of this 

significant resource. 

When early growth data was analysed in subsequent years it was clear that there had been a very slow 

response to the management intervention. It was then that we turned to the regrowth resource to 

investigate its productivity potential. 

REMNANT FORESTS 

Three locations were initially 

chosen for the Remnant forest 

demonstration sites, one in 

Nanango, Iron Pot and Miva. The 

pre-thinning stand data for the 

three remnant forests demonstrate 

an overstocked forest all of which 

had been subject to a medium 

harvest in the previous ten years. 

The plots were established in 

2000 (Miva NW Gympie), 2002 

(Iron Pot NW Kingaroy) and 2006 

Nanango. Trees selected for 

retention were generally in a 

dominant or co-dominant position 

with a Grimes Crown score in 

excess of 20 points. Any trees not 

marked for retention with a 

Photo 1. Miva plot 7 - thinned to 80 stems/ha 

merchantable product were 

harvested. Any trees left after the harvest that were unmarked were tree injected with 4:1 Tordon®. 

Table 1 represents the original stand stocking rates for each diameter class. 

Table 1. Original plot stocking rates for each diameter class 

DBH Classes 0 - 10 10 - 20 20 - 30 30 - 40 40 + Total 

Av Stems/ha - Iron Pot 149 213 111 14 10 497 

Av Stems/ha - Miva 130 91 94 70 - 385 

Av Stems/ha – Nanango 320 241 85 31 17 693 

Av Stems/ha - Rathdowney 63 118 14 6 8 209 

Av Stems/ha – Gin Gin 866 170 10 5 8 1059 

Eight years after the thinning operation annual diameter growth increments are still low to very low, 

ranging from .15 to .37 cm/yr dbh (see Table 2). 

NON-REMNANT FORESTS 

After completing a value adding case study in 2001 at a property owned by the Thompson’s at 

Gundiah, it became clear the management systems that they had introduced over the last 60 years, 

managing regeneration encroaching on to their open paddocks, were achieving growth rates well



above the traditional State-based systems. It became evident that thinning Spotted Gum when it was 

young, and retaining only trees that never came under suppression or excessive competition, would 

triple productivity. 

Three locations were chosen as regrowth demonstration sites. The first was obviously at Thompson’s 

and were in good condition with appropriate stocking levels (130stems/ha). The other two sites Gin 

Gin, (200 km NW of Gympie) and Rathdowney (130 km south of Brisbane) were clearly regrowth. 

The Gin Gin stand had a significant regeneration response (866 stems



Clearly the forest requires thinning from 1100 stems/ha to around 200, this can be achieved by 

A. Chemically treating the unwanted stems - at over 900 stems/ha to be injected it would be an 

expensive operation ($500-600/ha) and leave a very high volume of dead standing timber, a 

future work place and fire hazard, and leave little opportunity for Blackbutt to regenerate in the 

areas where woodsiana is dominating. 

B. Fall to waste, leaving up to 100 tonnes/ha fuel load 

C. Harvest the material as Masonite billets to supply the factory outside Ipswich (100km haul). 

This was considered the only viable option even though the mill gate price for the billets was 

only $42/tonne and it would cost $20/tonne haulage. 

To date 3 thinning trials have been undertaken, namely: 

1. Minor (1ha) hand falling operation with the product snug out with a 55HP 4wd tractor mounted 

with a hydraulic grapple. 

2. 10 ha mechanical harvest using a ‘Valmet 941’ (270 hp, 6 wheel drive with 11.5m reach). 

3. 7 ha mechanical harvest using a ‘Timbco 415’ (215hp, tracked drive with 6.5m reach). 

The trial’s objective was to test the viability of harvesting the non-productive/non-commercial portion 

of the stand to reduce the stand from around 1160 stems/ha, retaining only trees of a commercial 

species with a good quality bole and healthy crown. In simple terms the majority of the stems removed 

would be Angophora woodsiana a non-commercial species with only pulp values. The retained 

stocking rate varied with the species mix; in areas with a high proportion of woodsiana, all stems are 

removed creating a gap that is suitable for enrichment planting. In areas dominated by pilularis or 

microcorys the retained stocking levels would be around 200 stems/ha, if that quantity of straight 

healthy stems were available. 

The advantage of harvesting the material was to remove the bulk of the non-productive sector of the 

stand from the site instead of leaving it standing dead if chemically injected. It was considered that this 

may be able to be undertaken for the same price or cheaper than chemically treating the stand. 

The trial areas were paint marked for retention ahead of the harvest to ensure the best quality stems 

were retained and for ease of operation as the harvest operator did not have the skill set or the visual 

perspective from his machine to make a reasonable tree selection decision. 

MANUAL HARVEST OPERATION 

The hand falling operation was undertaken by two cutters using ‘660 Stihl’ chainsaws cutting on a 

face across the hectare. The 4wd tractor then moved through the stand, snigging the logs to a dump at 

either end of the stand. The cutters cut @ 3.1 tonnes/hr, and the tractor snug @ 6.8 tonnes/hr removing 

approximately 55 tonnes of product from the 

hectare. Net return = $ -280/ha. 

‘Valmet 941’ x 40 hr Trial 

The machine opened its own track between the 

trees and removed unmarked trees up to 11 

metres either side of the opened track. The 

harvester directional felled each tree, cut it to 

length and then stacked it for removal by the 

forwarder. 

The harvester proved very effective, harvesting 

stems with little or no damage to the residual 

stand. The large rubber tyres also rolled over the 

ground cover without exposing soil to possible 

erosion. The machine also broke the harvest 

residues into small pieces and left them flat on 

Photo 3. 55 hp 4wd tractor and grapple 

the ground reducing the potential fire risk and 

leaving a good layer of mulch.



The forwarder also reduced site impacts by lifting all harvest material and transporting it off site by 

trailer in place of snigging the material along the ground, exposing soil in the process. Average rate of 

production achieved was approximately 11.87 tonnes per hour completing 475 tonnes in the 40 hours. 

Net = return $-594/ha. 

Photo 4. Valmet 942 wheeled harvester - 

$150/hr 

Photo 5. Forwarder unloading logs at the dump 

$140/hr 

‘Timbco 415’ Trial 

The ‘Timbco 415’ (considerably older than the Valmet) proved to be very slow and resulted in a 

higher level of soil impact and damage to the residual stand. Harvesting at only 5.9 tonnes/hr. Net 

return = $ -1080/ha 

DISCUSSION 

The costs of the operation are significant and as such must be considered as a cost/return calculation 

per hectare. This varies considerably according to the density of the timber and its size class, ease of 

access, type of harvesting machine etc and as such may vary considerably according to the site. The 

results of these small trials demonstrate similar costs per hectare to chemical treatment per hectare 

($450/ha), however the results are very different. The advantages of harvesting can be considered from 

two positions, namely: 

1. The stand is not left with a significant layer of dead trees (fuel load, visual and 

WH&S benefits). 

2. The stand is immediately opened up to allow for crown development in the retained 

trees and a new regeneration layer. 

3. The areas with large gaps are available for immediate enrichment planting. 

4. The stand is left with a significant mulch layer, immediately reducing soil erosion and 

increasing soil carbon. 

5. Product is utilised and the carbon locked into long term building products. 

6. Further differentiation of the product i.e. sorting fencing materials, minor saw logs 

should improve the returns in areas where durability one species and saw logs occur. 

Disadvantages of harvesting include: 

1. With variation in site quality a significant drop in returns may make the process 

unviable (difficult to ascertain from a small trial). 

2. As a result of analysing the cost of the operation, the owners of the machinery indicate 

the cost of the hourly rate of the machine for ongoing work would be considerably 

higher. However there are other available machines for a similar cost but performance 

will need to be tested.



3. Many of the stumps will coppice and there is also a layer of Angophora woodsiana 

regeneration that will need further treatment or it will again dominate the stand. This 

section of the trial will trial three chemicals, ‘Glyphosate’, ‘Grazon’, and ‘Garlon’ to 

establish which chemical is the most effective at killing the stump coppice as well as 

the cheapest. The harvested area will be broken into three units and all stump coppice 

that is not required for regeneration, as well as unwanted regeneration (mostly 

Angophora woodsiana) will be sprayed using a double reeled ‘Quickspray’ unit with 

remote automatic hose retraction capability. 

RECOMMENDATION AND WORK PROGRAM 

There are significant advantages in the thinning harvest scenario in areas that are viable for this type of 

operation. However, the scope needs to be extended to an integrated harvest that includes removal of 

fencing material, trees in decline where sawlog material is available and the removal of trees for pulp 

(bent or non-commercial) that are not required for habitat or species distribution. 

REFERENCES 

Bacon G (2009) ‘Time for Fresh’, Interviewed by Neilson T in In-Wood Magazine. http://www.inwoodmag.com 

Burke T (2009) ‘Preparing our Forest Industries for the Future’, Department of Agriculture, Fisheries and 

Forestry Ministerial Statement. Dated: Wednesday, 24 th June 2009. 

De Fegely R and Follas C (2006) ‘The Options for Increasing Harvest Security and Investment in Private 

Natural (Native) Forests in Queensland’, Report by Poyry Forest Industry for AgForests Queensland. 

Marohasy J (2005) ‘Terrorist or Freedom Fighter?’ The Politics and Environement Blog July 21, 2005,viewed 

18 June 2009, http://www.jennifermarohasy.com/blog/archives/000748.html 

Stanton R (2005) ‘Tracking the trade deficit is becoming harder’, Australian Forest Grower. 28 (1): p. 10



IMPROVING BUSINESS PERFORMANCE 

IN SOFTWOOD PLANTATIONS 

David Williams 1 

ABSTRACT 

Australia’s plantation estate is now approaching 2 million hectares and is composed of 

similar areas of softwood and hardwood species. There has been limited expansion of 

the softwood estate over the last 2 decades. As a consequence the current tight softwood 

log supply will become critical as demand continues to grow while supply remains 

static. Investment in new plantings has been limited because of poor returns. Best case 

new plantation investment returns that could be expected currently would be around 

4%. This paper considers some initiatives to improve financial outcomes from existing 

businesses which are a pre-requisite for any investment in new plantings. The initiatives 

include improved log prices, win-win opportunities through growerprocessor 

partnerships and improved value recovery from implementing optimising technology 

into harvesting. These initiatives have potential to add a further 3-4% return on 

investment. 

INTRODUCTION 

The national plantation estate has been expanding at a healthy rate for the last four decades. The 

current national plantation estate of 1.97 million ha represents a nine fold expansion over the preexpansion 

program estate of the early 1960’s. The estate is now almost equal proportions of 

softwood and hardwood species. These plantations were established for different purposes with 

different end markets. Changing government policies, the reduction in native forest harvesting and 

Vision 2020 have created drivers for an expansion of hardwood over softwood in the past 15 years. 

Traditionally softwood plantations have been managed under longer rotations (~30 years) to 

maximise sawlog production for processing into structural timber. The more recently established 

hardwood plantations are managed under shorter rotations (10-15 years) primarily for woodchip 

production destined for export to Japanese (or North Asian) paper companies. 

Despite increasing demand for softwood sawlogs, the softwood plantation estate has remained almost 

static for two decades. Net new plantings have averaged less than 1,000 ha per annum since 1992. 

Given the time lag for maturing of long rotation plantations, the lack of new plantation investment is 

now becoming apparent. Ironically this gloomy outlook follows the golden decade of increased log 

supply and processing investment. 

This paper focuses on Australia’s softwood plantations to show that whilst the aggregate picture of 

the national plantation estate reveals a positive outlook, the future for the softwood plantation sector 

is dire. A large expansion in new softwood plantations is desperately required. Softwood plantation 

owners have various opportunities to improve the commercial outcomes from their existing 

businesses a pre-requisite for investing in new plantings. These initiatives are actions available to 

softwood plantation owners and have been implemented by HVP. 

THE NATIONAL SOFTWOOD PLANTATION ESTATE 

There are three recognisable phases covering the national softwood plantation estate, as presented in 

Figure 1 below. 

Phase 1 – Prolonged early period of modest plantings 

Plantings commenced in the 1870’s and it took 90 years to establish the first 200,000 ha. The driver 

for these plantations was the shortage of local sources of softwood species. 

1 rd 

General Manager, Business Development, HVP Plantations, 3 Floor, 517 Flinders Lane, Melbourne, Victoria 3000. 

Ph: 03 9289 1400 Ph: 03 9289 1430. Email: dwilliams@hvp.com.au.



Phase 2 – The estate rapidly expands 

The national estate increased five-fold to almost 1 million ha over the period from the early 1960’s to 

the late 1980’s. The plantings were undertaken by State governments, with Federal government 

financial assistance. Government investment was aimed at rapidly expanding the estate consistent 

with post World War II government policy aimed at self sufficiency in strategic industries. It should 

be noted that many of these plantations were established on Crown land which was cleared of native 

vegetation; an action now banned across Australia. 

Phase 3 – Plantation expansion hits the wall 

Government investment ceased by the late 1980’s as Australia moved into recession. New plantings 

have almost stalled since government investment ceased. Community pressure to halt the clearing of 

native vegetation resulted in many agencies purchasing and planting cleared agricultural land. A 

subsequent backlash and difficult economic climate saw these agencies halt plantation expansion. 

The net softwood estate has expanded by less than 1,000 hectares per annum since the one millionth 

hectare was planted in 1992, following a period when expansion averaged more than 30,000 hectares 

per annum. Planting of first rotation areas by Willmott Forests and other softwood MIS companies is 

off-setting the area lost due to conversion to other species or land uses. 

'000 ha 

1,200 

1,000 

Figure 1 : Average rate of softwood plantation establishment in 

Australia 

800 

600 

400 

200 

0 

1870 

early modest 

plantings 

rapid expansion 

1965 

1992 

expansion 

hits wall 

THE SECTOR’S GOLDEN DECADE 

Log production increases as the estate matures 

Log production increased over the last decade as the national softwood estate matured and the 

plantings from the 1970’s and 1980’s were harvested. These plantations are routinely re-planted. The 

current softwood sawlog production of 14.4 million m 3 represents a 39% increase over the decade 

(BRS, 2009). 

Processors invest in expanding production capacity as log supply increases 

During the past decade, sawn timber production has increased 85% to 3.9 million m 3 (BRS, 2009). 

Progressively, the sawn timber processing sector invested in larger, world scale capacity mills using 

improved technology to process the high quality logs at lower unit costs. 

The other processing sectors of panels and paper products have also expanded production over the 

last decade by 27% and 32% respectively on the back of increased log availability. 

2008



Disappointing prices during decade of expansion 

Despite increases in supply and efficiencies in processing and subsequent import replacement, timber 

prices have declined substantially in real terms over this period of unprecedented expansion in log 

supply and processing capacity. Sawlog and pulp log prices have declined in real terms by 22% and 

9% respectively (KPMG, 2009). Figure 2 below illustrates the decline in sawlog prices in real terms 

over the period from 2000-2008. Sawlog prices shown are the weighted average of the four log 

classes presented in the Australian Pine Log Price Index (APLPI)(KPMG, 2009). 

Sawlog price $/m cubic metre 

$51.00 

$49.00 

$47.00 

$45.00 

$43.00 

$41.00 

$39.00 

$37.00 

$35.00 

Figure 2 : Average softwood sawlog prices 2000-2008 

2000 2001 2002 2003 2004 2005 2006 2007 2008 

APLPI price - flat real $ APLPI price - actual real $ 

STAGNANT LOG SUPPLY FOR DECADES AHEAD 

No net increase in plantation areas over the last 2 decades limit opportunities 

Softwood sawlog supply is currently fully committed, predominantly under long term contracts to a 

sector with processing capacity which exceeds the log supply. Log supply will remain static for the 

next 30 years even if major new plantings commenced now. While demand exceeds supply, the 

domestic demand results in a national trade imbalance in timber and paper products of $2 billion per 

annum. Advances in productivity will require increased supply which is not available in the short 

term, or rationalisation through mergers to maintain world scale efficiencies. 

For more than 50 years, the demand for timber products has been correlated with population growth. 

On average, Australians consume a little over one cubic metre of log equivalent volume of timber 

products per person per year (BRS 2009). Domestic consumption of timber products will therefore 

continue to grow over future decades in line with population growth. Future population growth is 

influenced by fertility and morbidity rates as well as government immigration and family policies. It 

is observed that population growth increased by an average of 1.5% per annum for the 5 years until 

June 2008 (www.census.abs.gov.au) This is a comparatively high growth rate historically. However, 

one can expect future population growth to remain positive at a rate of around 1% per annum. 

Assuming a 1% future growth rate, timber demand will increase by more than 200,000 cubic metres 

of log equivalent per year. By 2019 the supply-demand gap for softwood log will have increased by a 

further 8 million cubic metres per annum. This will cause further deterioration in the trade imbalance 

in timber products.



Investment to maintain competitive mills will be challenging 

Processors have invested in new mills by securing increased log volumes from growers over past 

decades. This has resulted in more efficient mills which were world scale when initially constructed. 

Maintaining competitiveness requires ongoing investment in new milling technology and this will be 

more challenging for the processing sector in the future. Because of supply constraints, further 

upgrades will be difficult without rationalisation through amalgamations or merging of operations. 

This is a more expensive option than the past approach of securing increased volume directly from 

growers. Upgrades do not increase overall log supply or production capacity. 

Business expansion is limited for growers 

Competition and increasing prices for land limits the ability for the softwood plantation businesses to 

expand their existing estates into the future. Supply expansion will depend on making the most of the 

current estate through improving operational efficiencies, genetics, nutrition and other plantation 

management activities. Government-supported past research provided a basis for substantial 

productivity gains in softwood plantations. Cessation of this source of support will require other 

parties, including individual growers, to increases their investment in plantation research if ongoing 

productivity gains are to continue. 

There will also be opportunities for softwood plantation owners to benefit through new 

environmental values, including carbon sequestration, bio-fuels, and other environmental values. 

THE CHALLENGE OF ATTRACTING NEW INVESTMENT 

Past players and future investors 

Governments and public companies largely financed the existing softwood plantation estate. These 

traditional bodies are unlikely to be major investors in further new plantings. 

There has been a changing of the guard in the softwood plantation sector. Timberland Investment 

Management Organisations (TIMO’s) and Managed Investment Schemes (MIS) companies are more 

recent entrants into the Australian forestry sector and may continue as investors if comparable 

mainstream returns can be achieved from investing in new softwood plantations. MIS companies are 

finding it more difficult to justify and secure new land for planting, opting in some cases to rent 

second rotation land from established plantation growers. 

The nature, structure and motivations of Australian plantation investors varies considerably. There 

are a number of different investor categories that have contributed to the national plantation sector. 

In 1950-51 more than 90% of Australia’s total plantation estate was publicly owned. Public 

ownership has declined substantially since and new players, including TIMOs and MIS companies, 

now own almost half of the plantation estate (BRS, 2009) as shown in Figure 3. 

Figure 3 : Australian plantation ownership 2008 

Other private 

growers 

Timber companies 

Super funds 

MIS 

Governments



Governments – unlikely future investors 

Governments were the major drivers behind the establishment of the existing national estate but these 

sources of investment fell away sharply 2 decades ago. The previous investment motivation relied on 

government industry policies to create regional employment and infrastructure and to reduce 

dependence on imported wood and paper products. Once established, governments withdrew from 

their activity to be replaced by private enterprise and so the former policies are no longer relevant. 

Further, government ownership of plantation businesses has been progressively reduced through 

either the creation of commercial Crown corporations or the complete privatisation of plantations. 

Having withdrawn from the sector, Governments are unlikely to again be direct investors in new 

plantings. Those plantations operated by Crown corporations may well invest in new plantings but 

only based on business models which show reasonable returns on investment using commercial 

investment criteria similar to private investors. 

Public companies – unlikely future investors 

Vertically integrated plantation and processing companies made significant past contributions to the 

existing national estate. The trend over recent decades has been for integrated companies to divest 

their plantations and focus on their processing businesses. This move was precipitated by perceived 

poor returns from their timberlands businesses based on poor commercial models and internal pricing 

between enterprises. The original motivation for their plantation investment was resource security 

for their downstream processing businesses. Revision of business models has suggested that it is 

more cost efficient for corporations to concentrate on their processing activities, sell the plantations 

and retain long term supply contracts for the future volume from these plantations. Thus the existing 

public companies involved in processing are unlikely to be major direct investors in new plantings in 

future. 

Private companies – likely future investors 

Private companies are those with a limited number of shareholders and are unlisted on the Australian 

Stock Exchange. These are often TIMOs which are active investors in private timberlands. They 

have been formed on the back of opportunities provided by public integrated companies and 

governments divesting their timberlands. TIMOs bring together interested investors often with 

considerable access to funds. They apply streamlined and efficient management with long term 

philosophy to what were poorly performing businesses. 

The first TIMO to enter the Australian plantation sector was Hancock Natural Resources Group 

when it acquired the previously owned Victorian government plantation business, Victorian 

Plantation Corporation. The parent created Hancock Victorian Plantations (now renamed HVP 

Plantations) in 1998 to manage the Victorian plantations. Since this time a number of other TIMOs 

have become involved in plantation ownership and or management. 

MIS is a relatively new investment model for establishing new plantations. There were 10 major MIS 

companies offering a variety of investment options for individuals seeking investment in plantations 

and other agricultural businesses. The sector has expanded at a rapid rate since the mid 1990’s. The 

overwhelming share of total investment has been in short rotation hardwood plantations and this has 

been the driver in the rapid expansion of the nation’s hardwood plantation estate – expansion of more 

than 800,000 ha since 1996 (BRS, 2009). The investor attraction to these hardwood plantations has 

been the relatively short investment period. 

There has also been some smaller scale investment by local and foreign investors in short rotation 

hardwood plantations, often by entities directly or indirectly associated with north Asian processors 

or trading houses. 

One MIS company, Willmott Forests, is an exception to this general picture. Willmott Forests invests 

in softwood plantations on behalf of individual investors. Over the last 20 years Willmott has 

established an estate of more than 40,000 ha of softwood plantations through MIS. 

The current outlook for MIS investment is unclear with the two largest MIS companies going into 

receivership. Time will reveal whether the apparent shortcomings of these companies will limit 

future investment via MIS companies. With longer term investment horizons it should possible for 

MIS to be a significant driver for new softwood plantations if the shortcomings of the current failing



companies have not soured an investment model for new plantings based on reasonable returns and 

established processing. 

Both TIMOs and MIS companies are potential investors in new softwood plantations provided 

realistic investment returns can be achieved. 

WHAT’S NEEDED FOR PROFITABLE INVESTMENT 

Mainstream investment in new plantations will require normal returns on investment over the life of 

the business. Expected returns on investment vary with the nature and risk of the business. A 

definitive acceptable rate of return on new plantation investment is debatable. However, let us 

assume, for the purposes of this discussion, double-digit annual return of at least 10% is a reasonable 

expectation for a mainstream investment in new plantations. The general term investment return or 

return on investment as used in this paper refers to economic internal rate of return (IRR). 

It is estimated that the investment returns for many existing softwood plantation businesses range 

from 2-10% per annum (Industry Edge, 2008). Businesses in the upper quartile of returns, which 

includes HVP, are likely to be providing satisfactory returns to their investors. Returns for businesses 

in the lower quartile would not satisfy normal investors and are therefore probably disappointing 

their current owners. 

Why wouldn’t existing successful plantation businesses invest in new plantings? 

Returns from existing plantations are substantially greater than new plantations. The difference is the 

cash flow profiles. Having the ability to generate cash from older plantations is important to 

investment returns. Existing plantations usually have a range of age classes that provides an 

immediate and ongoing cash flow whereas cash flow from new plantation investment is delayed for 

many years. 

In addition, the valuation approaches used to assess investment returns for most existing plantation 

businesses either do not consider land value or apply conservative accounting treatments to land. 

Thus the cost of purchasing or leasing land for a full rotation or in perpetuity is the defining 

difference between existing plantations businesses and new plantings. Carrying the land cost at 

prevailing discount rate for around 30 years until the plantation is finally harvested is a burden for 

new plantation investment. Much of the existing national estate was established by governments 

clearing native forests on Crown land, so there were no land costs for these plantations. 

Analysis based on costs and future cash flows incorporating land value at a prevailing discount rate is 

an appropriate model for considering new plantings. A hypothetical investment scenario illustrates 

the substantially superior investment outcomes for existing plantations compared with new plantings. 

Consider a new plantation investment assuming medium growth rate (assume m.a.i. = 22 m 3 /ha/a), 

prevailing log prices (assume APLPI price – KPMG, 2009), average plantation and log production 

costs, and 2 thinnings with final harvest at 28 years. If land is available at no cost, an IRR of 8-9% 

would be a reasonable expectation. If land purchase cost $5,000/ha, or land lease at commercial rent 

of 5% per annum, then the expected IRR might be in the order of 4%. 

There is a myriad of variables that affect costs, growth rates, plantation regimes and sales returns, 

therefore determining the hypothetical optimal new planting option is problematic. An IRR of close 

to 4% is likely to represent the best outcome achievable for investing in new plantings assuming best 

case plantation variables and including full land cost. Such a best case investment option falls short 

of minimum expected market returns by around 6% per annum. 

INITIATIVES TO IMPROVE BUSINESS PERFORMNACE 

This section considers three initiatives that have the potential to improve returns from existing 

plantation businesses and therefore provides opportunities to contribute to bridging the gap for new 

plantation investor returns. 

Initiative 1. Log prices 

Log prices have been declining in real terms over the period when log supply has been progressively 

tightening until the current time where log supply is in deficit.



Normally when supply exceeds demand for any commodity in competitive markets, increasing prices 

reflecting the normal principles of supply and demand can be expected. This has not been observed 

in the Australian softwood log markets for a number of reasons including the fact that log sales are 

generally not exposed to competitive market pressures. 

A competitive market requires a number of willing sellers and buyers. Australian softwood log 

markets do not reflect normal competitive markets because in most cases there is only one or two 

major log sellers and buyers in each regional market. In some cases there is only one seller and one 

buyer in a market. Increasing log supply by importing logs from New Zealand, North America or 

Chile is not viable and has not occurred. Thus the lack of normal competitive markets excludes 

supply and demand pricing in the large majority of sales. 

HVP has experienced increased prices in all cases when it has had the opportunity to offer logs 

through a competitive sale process, illustrating the potential of competitive markets. Anecdotal 

information indicates other plantation owners also achieve higher log prices whenever they are able 

to offer logs for sale through competitive market processes. 

With expanding demand and tightening log supply, processors seek greater supply security which has 

additional value to the processor. Once the aggregate log supply is fully committed, long term supply 

contracts reduce opportunities for changes in the log volume shares between processors. Thus market 

share of processed timber is also locked in by log supply contracts. A processor cannot challenge 

another competitor's market without in fact withdrawing from some part of its own existing market. 

Fixed market shares provide a unique pricing environment for processors. There is less pressure for 

price discounting when market shares cannot be challenged. Thus long term supply contracts present 

unique advantages to processors providing opportunities for real price increases for timber. 

Real price increases for domestic sawn timber establishes the circumstance for real log price 

increases. The challenge is for processors to maintain their profitability through timber price 

increases and reducing processing costs during the future period of market-driven increasing log 

prices. 

Whilst the increasing deficit of domestic timber supports price increases, domestic timber also faces 

import competition which can constrain domestic prices. Other materials used in house building 

have achieved better price outcomes than domestic timber, notwithstanding tough competition within 

their own sectors. 

annual % increase 

8.0% 

7.0% 

6.0% 

5.0% 

4.0% 

3.0% 

2.0% 

1.0% 

0.0% 

Figure 4 : Price changes for some building materials 

used in house building 

All groups Timber, board, 

joinery 

Concrete, cement, 

& sand 

Steel products 

av increase - last 5 years av increase - last 40 years



Figure 4 illustrates that the average annual price for concrete, cement and sand, and for steel products 

increased by 6.7% and 6.8% respectively over the last 5 years. Over the same period, timber, board 

and joinery managed an average annual price increase of only 1.8%. However, the trend over the last 

40 years has been comparatively flat around 6% per annum 2 . It would seem, therefore, that, during a 

period of major increase in log availability and expansion in timber production, timber prices have 

actually declined when compared with other house building materials. 

Scarcity provides opportunity for substantial real price increases for both logs and timber. A real 

price increase in log prices of 10-15% is realistic for the medium term, and this could add 1% to the 

IRR for new plantation investment. 

This thus constitutes another opportunity for growers seeking profitable new plantation investment. 

Initiative 2. Grower Processor partnerships 

Partnerships are defined here as grower and processor working together where business aspects 

overlap or interface. The benefits of the partnership are allocated between the parties as profit shares. 

Traditional growerprocessor relationships involve the grower supplying logs to the processor with 

very little business cross-over. Growers maintain their focus on growing, harvesting and delivering 

logs, while processors receive the logs and process them into timber products. The businesses remain 

separate with few additional benefits beyond those formal arrangements detailed at the time supply 

agreements are executed. 

This relationship can inhibit opportunities for mutual benefits that could flow from cooperative 

actions aimed at improved overall commercial outcomes. The past has seen opportunities go begging 

because of the inability to agree on how the costs and benefits of enhanced net outcomes can to be 

shared. 

For example, a change to an aspect of supply or log specifications could improve the value of the 

parcel of logs but this may incur some additional cost to the grower. The inability to agree on how 

the additional cost is to be offset and how the net upside is to be shared may prevent the beneficial 

changes being adopted. Such situations represent wasted opportunities. 

The most prospective areas for partnerships will be in business activities that are more likely to be 

successful if carried out jointly. For example, where inputs are complementary, such as expertise, 

funds, materials, other resources and markets. The possibilities could include value adding 

opportunities where action is required by both parties to achieve the improved outcome. One 

example is the production and sale of chips. This is the basis of the successful partnership of 

Softwood Plantation Exporters (SPE) (discussed further below). Other potential areas for 

collaboration or for joint involvement include: 

• Bio-fuel production including co-generation with fibre coming from the mill and the 

plantations. 

• Changes to enhance the value of a log parcel beyond expectations in the original supply 

agreement. Changing the parcel to add value for the processor could incur cost to the grower 

but the net result could still represent increased value overall. 

• Additional log volume provided preferentially to the processor. 

• Enhancing the value of supply contracts by amending conditions such as extending term. This 

is valuable to the processor as it locks in market share by excluding access by competitors to 

a fully committed national log market. There is a particular opportunity to realise this value at 

the time a processing business is sold. Amending supply contracts prior to sale is an example 

of where grower and processor could act together to add and share value which is 

immediately realisable in monetary terms. 

The formal arrangements for such collaboration can come in many forms. There is a full range of 

options covering the format of growerprocessor partnerships, including formal partnerships, joint 

ventures (JV) and marketing arrangements. These partnerships do not require one party intruding into 

2 http://www.abs.gov.au/AUSSTATS/abs@nsf/Lookup/6427.0Main+Features1Junpercent202004?Open



the core business of the other party. The assumption is that parties are expert and successful at their 

core businesses and will continue as separate businesses. 

Implementing successful partnerships is challenging as they need to exhibit some or all of the 

attributes of a high level of confidence and trust between the parties, a strong base of common 

objectives, complementary inputs, and shared business style and culture. 

Softwood Plantation Exporters (SPE) – an example of a successful partnership 

HVP and AKD Softwoods, one of HVP’s major sawlog customers, formed a JV partnership more 

than 10 years ago to export softwood chips to Japanese paper customers. 

At the time, HVP’s plantation value was constrained by limited market outlets for pulp logs from 

thinning operations. This prevented full implementation of optimal thinning regimes. AKD was also 

adversely affected by limited volume outlet for its residual chips. The parties came together through 

necessity after each concluded there were limited opportunities for their stand alone pulp log and 

chip volumes. The larger total chip volume allowed consideration of new domestic processing or 

export options. Exporting chips to the Japanese paper market was the preferred option. The parties 

formed a JV and the export project was developed as a greenfield business with the parties sharing 

equally all aspects of the JV and selling their chips under joint arrangements. 

The business has been successful, providing improved financial returns for both parties as well as 

building a platform for expansion of the business under the SPE banner and providing an opportunity 

to approach aspects of the traditional log supply as a partnership. AKD and HVP both receive good 

prices for sawmill chips and plantation pulp logs as well as an export margin. The export chip 

business is a significant addition to the respective partners with an aggregate turn over of almost 

$300 million since 1997. 

Thus such partnerships are cost effective and provide potential to improve the financial performance 

for both parties. A plantation owner could reasonably aspire to increased returns sufficient to 

improve IRR by 1% or more. 

Initiative 3. Value Recovery from harvesting 

Value recovery (VR) refers to maximising the value of trees at harvest. A tree stem can be cut into 

various combinations of log quality (veneer, sawlog, pulp log), diameter and length. Maximising the 

aggregate value of the mix of log products provides improved financial returns from harvesting 

operations. 

Growers’ primary benefit derives from value recovery. Conventional mechanised harvesters have 

limited capacity to select the optimal mix of log products and therefore cannot maximise recovered 

value from harvesting operations. Optimisation technology is one method for determining the best 

log combinations. 

Invariably processors have a log product range they are seeking. Using machine-mounted computer 

systems and competent operators, the grower and the processor can select various log product 

combinations to maximise value. An effective system incorporates reconciliation processes to 

measure VR performance against pre-harvest inventory and forest model benchmarks, and timely 

auditing of operations to provide positive feedback to harvesting operators. A VR program needs to 

be supported by comprehensive procedures. 

An optimisation program with these attributes is being implemented by HVP across its estate. 

Procedures have been developed, supervisors and operators trained and optimising technology 

progressively fitted to harvesters. Two thirds of harvesters in HVP plantations are now fitted with 

optimising technology and most supervisors and machine operators have been trained in VR 

procedures. The majority of HVP’s logs are now produced by optimising harvesters with very 

positive results. Recovered value is consistently better than non-optimised operations and value 

recovery exceeds forest modelling forecasts. Three per cent overall value improvement from 

harvesting is a conservative achievable target if an effective VR program is in place. There will also 

be opportunities for further improvement as effective VR programs are developed beyond the initial 

phase. 

A 3% value recovery improvement could correspond to a 0.5% increase in IRR.



CONCLUSION 

Australia has a substantial trade imbalance in timber products including imported products processed 

from softwood logs. Currently softwood log demand exceeds a supply which will remain static for 

several decades as a result of lack of investment in new plantations. Historically timber demand has 

grown in line with population growth, so the current gap between supply and demand will further 

widen. 

New plantation investment is urgently required to provide opportunities for Australian softwood 

plantation growers and processors to redress the increasing national timber trade imbalance. The lack 

of investment is attributable to perceived low returns from new plantation investment and the 

diversion of potential funds to short rotation hardwood plantations. 

Improving returns from existing plantation businesses is a first step in creating an environment that 

may be attractive to private investors. There are a number of potential initiatives that existing 

plantation businesses could adopt. Some of these include increased log prices resulting from the 

increasing scarcity of logs and domestic timber, shared improved outcomes from win-win 

partnerships between growers and processors, and improved value recovery from harvesting by 

utilising optimising technology. These initiatives have been successfully incorporated into HVP’s 

business and all have added to overall business performance. 

These initiatives, and others, have the potential to move the currently low returns from investment in 

new plantation areas closer to the IRR expected by mainstream investors. Achieving these outcomes 

could make new plantation investment attractive and be a catalyst for a reinvigorated expansion of 

softwood plantations. 

REFERENCES 

BRS, Department of Agriculture Fisheries and Forestry, Australian Government (2009). Australia’s forests at a 

glance 2009 Data to 2007-08. 

KPMG (2009). Australian Pine Log Price Index Updated to December 2008. 

Industry Edge (2008). A Critical Analysis of the Softwood Industry in Eastern Australia.



ADDRESSING CLIMATE CHANGE IMPACTS ON THE BUSINESS OF 

PNG FORESTRY: IMPLICATIONS FOR AUSTRALIA AND PNG 

Dick McCarthy 1 and Gabriel Samol 2 

ABSTRACT 

This paper addresses the climate change impacts and adaptation challenges for forestry 

on the forests of PNG (Papua New Guinea) which are botanically related to the isolated 

pockets of surviving Gondwanan flora in Australia such as the wet tropical rainforests 

of northeastern Queensland. 

PNG, through large scale plantation development, has the potential and the capacity 

(land, soils, water, air and adaptable plantation species) to capture and store carbon in 

its biomass, soils and forest products in perpetuity. If, through the proper introduction of 

an emissions trading scheme, increased value and returns to the forest owners and 

developers can be realised, PNG has the capacity to increase: 

• Five fold its commercial timber tree plantation estate to some 300,000 hectares. 

• 20 % increase in forest industry employment and flow on effects to other 

employment sectors 

• an increase in value of forest products production of more than 100 % 

INTRODUCTION 

The origins of Australia’s forests can be traced to the beginning of the Cretaceous period, when the 

super continent Gondwana began to fragment into Africa, South America, India, Australia/Antarctica 

and many smaller islands. Although this happened at least 135 million years ago, similarities can still 

be found in the flora and fauna of these now widely separated lands. 

Over the intervening millennia, Australia has evolved a new biota. About 38 million years ago, 

Australia broke away from Antarctica and shifted northwards, colliding with Asia about 13 million 

years ago. During Australia’s northward journey, the climate became progressively warmer and drier, 

and the vegetation adapted accordingly. Cool and warm rainforests were replaced by sclerophyllous 

genera such as Eucalyptus and Acacia. 

PNG is a land subject to continual catastrophe. The mountains are young and continuing to uplift as 

the Australian plate subducts below the Pacific plate so earthquakes with associated landslips are 

frequent on the young steep slopes. There are numerous active volcanoes which create lava flows and 

mudflows and thick ash deposits. Strong destructive winds occasionally occur. In exceptionally dry 

years these forests, always slightly seasonal, become unusually dry and may catch fire. The big rivers 

which run on the coastal plains have unstable courses. Shifting cultivation and associated regrowth 

forest is also extensive. 

In reviewing PNG forest dynamics, it is no surprise that, in a listing of timber tree species for a tract of 

lowland rainforest, there is usually a considerable proportion of pioneers, such as species of Albizia; 

Paraserianthes and Serianthes, and Eucalyptus deglupta (New Britain), and strongly light-demanding 

climax species, for example, Campnosperma; Pometia and Terminalia. 

As PNG has been Australia’s wood shed for over 100 years and politically continues to be a hardwood 

producing nation, this paper will cover, especially from a PNG perspective: 

• Climate change scenarios 

1 McCarthy & Associates (Forestry) Pty Ltd, formerly Executive Director, PNG Forest Industries Association 

2 Executive Officer PNG Forest Industries Association, Inaugural President Association of Foresters of Papua New Guinea



• Effects on PNG communities and on the full range of forest values 

• Management of risks and identification of opportunities 

The implications for Australia and PNG of political, economic and security development in terms of 

PNG forestry is that there is a long standing and enormous wood trade between Australia and PNG 

encompassed within a sphere of similar forest types and forest products bound by many existing long 

term areas of mutual co-operation within the respective government and private sectors. Climate 

change scenarios should be no different 

PNG’s Vegetation 

PNG’s constant high temperature is characteristic of the tropical climate and hence creates tropical 

moist rainforest vegetation over the whole island. 

This tropical moist rainforest vegetation of PNG has considerable variation dependent on varying 

responses to rainfall, soil water; soils; elevation, and soil fertility which give rise to the following 

distinct forest formations: 

• Rainforests (where every month is wet (100 mm+ of rainfall per month)) and graded by 

elevation into: Lowland rainforest; Lower montane forest; Montane forest; Subalpine forest; 

Heath 

• Savannah forests (where there are several dry months of 60 mm or less of rainfall) 

• Soil water creating forest formations as mangrove forests; beach vegetation; peat swamp 

forests and freshwater periodic swamp forests. 

Nature of PNG Wood Business - Current Situation 

Currently PNG’s annual wood flow approaches 8 million m 3 (5 million m 3 firewood (FAO estimate); 

1.8 million m 3 log export; processed export 0.65 million m 3 , domestic industrial 0.2 million m 3 ) 

• Forestry and forest products are PNG’s second largest industry. After 60 years, in terms of 

revenue generation and forest management, only 3.7 million hectares are under active forest 

management regimes and there are only 65,000 hectares of plantations 

• Even so, if allowed to operate efficiently, this small forest base is capable of producing a 

sustainable cut of 3.9 million m 3 with a contribution of US$270 million to PNG’s GDP 

annually; with US$85 million in export taxes/levies, while landowners receive some US$20 

million in direct payments 

• The total value of manufactured forest products in 2002 was US$33 million, with log exports 

totalling US$100 million. 

• The forestry industry has various sectors: harvesting, sawn timber, plywood manufacture, 

veneer production, furniture making and forest plantation activities. 

• The wood business trade is based on price and individual consumer demand 

PNG Market Changes for Solid Wood Products – Export Markets 

• In recent years, PNG has seen a rapidly declining market for its forest produce in Japan, and 

now its China market is diminishing. 

• Yet China is becoming the top exporter of secondary wood processed products in the 

developing world and is expected soon to overtake Germany as the third largest exporter 

globally. 

• The PNG solid wood export products market is being taken over by panel products, wood 

composites and now even wood/plastic composites. 

• The Australian sawn timber and panel product is now becoming PNG’s fastest growing 

market for PNG sawn timber and PNG plywood. 

• The Australian market is the closest export market for PNG 

PNG Market Changes for Solid Wood – Domestic Markets 

• PNG has lost much of the domestic market for timber products to other non-wood materials, 

due mainly to failure of wood in use because of the lack of wood preservative treatments.



Property owners and developers, together with architects, engineers and builders have sought 

other materials because of the inherent dangers in using untreated wood. 

• This loss of domestic market share correlates directly with the cessation of proper monitoring 

with compliance checks of correct timber treatments and verification by the responsible 

government authorities; i.e. government agencies are not doing the job for which they are 

mandated.. 

Future Development of the PNG Solid Wood Market 

• The existing annual sustainable wood harvest is 3.9 million m 3 per annum. Current production 

is averaging 2 million m 3 . Of this harvest approximately 30% is processed. 

• The 10 proposed new projects cleared for further development by the Moratorium/review of 

the former government would add 1.3 million m 3 to the annual cut. This is well within the 

annual sustainable (allowable) cut. 

• A phased introduction of those 10 projects would provide a net gain in direct export tax of 

K80 million per year, and an increase in cash and benefit payments to landowners of K24.5 

million per year. 

CLIMATE CHANGE SCENARIOS 

The Politics of “Ecology” or Climate Change 

In the 1960’s, “ecology” was the buzz word. Mass media bestsellers assured us that by the 1980’s we 

would all be starving or suffocated by our own wastes. What went wrong? Or what went right? 

Did the world’s problems go away in the intervening period? Of course not — but the difference is 

that we know a lot more now than we did in the 1960’s.The ozone layer and the oceans appear to be 

rather more resistant to pollution than we once thought. We are more conscious of the crises being 

caused by acid rain; deforestation; desertification and carbon dioxide build-up and population buildup. 

PNG is no different to the rest of the world with regard to the politics of population; the politics of 

food; or the politics of forests and the politics of forest protesters. The politics of climate change is 

treated no differently. 

The politics of climate change. – “The developed world wants a carbon trading scheme where PNG’s 

clean environment, climate cleaning mechanisms and carbon storage facilities are utilised without 

having to pay for them”. Anonymous 

Effects of Change on Rainforest Communities 

Tim Whitmore, one of the world’s foremost rainforest experts, studied the world’s tropical rainforests 

for many years. His observations typify PNG’s forest development. These observations include: 

• In lands where tropical rainforests occur, man has lived in closest dependence on them since 

time immemorial. 

• Europeans became aware of them over two millennia ago. 

• Increased knowledge since the Renaissance with the voyages of discovery and then the 

colonial era revealed that there are in fact many different kinds of tropical rainforest. Plants 

exist in luxuriance and a diversity of bizarre forms undreamed-of in temperate latitudes. 

Animal life is also rich and diverse. 

• Tropical forests have waxed and waned since geological time, and the present patterns of 

species distributions are a result of these historical events. 

• The former idea that these great forests have survived immutable “since the dawn of time” is a 

romantic fallacy, as investigations over the last three decades have shown. 

• The forests are continually changing at the other end of the time scale, the life span of an 

individual tree. 

• The elucidation of forest dynamics has been the other major breakthrough of recent years. 

Now we know a great deal about the ecology of individual tree species and the particular



requirements for growth of their seedlings in canopy gaps formed by the deaths of big trees. 

Silviculture, the manipulation of forests by Man to favour tree species of his choice is 

dependent on understanding these innate characteristics. 

• Tropical rainforest nutrient cycles are now reasonably understood. The old idea of a closed 

cycle and with nearly all the nutrients in the plants has not survived. 

• Shifting agriculture in now well known to be a sustainable form of farming, suitable for 

infertile soils. 

• Sustainable human utilisation of forest lands for crops or trees depends, as with silviculture, 

on working within the natural limits of the nutrient cycle. 

Whitmore identified further major impacts on tropical societies as: 

• The colonial era, which had a profound impact on tropical societies. Now largely disappeared 

it left tell-tale traces in forest structure or species composition. The spice trade shaped the 

history of the world. Useful plants were moved between continents and introduced to new 

regions via Botanic gardens. Today, staple foods and many fruits are pan-tropical as a result. 

• The industrial revolution increased demand for many forest products such as rubber and 

resins. Now trade is greater in tropical hardwood timbers. 

• Today, tropical woody vegetation is being altered at a rapid rate in one of two different ways.. 

Some areas are converted to other land uses (agricultural pursuits). Other areas are logged but 

left to regenerate, but may be destroyed later by shifting agriculture as farmers gain readier 

access to land being opened up by logging roads. 

Climate Change Scenarios and their Impact on PNG’s Vegetation 

Patterns of distribution of plants, animals and vegetation have long fascinated biologists. Over the past 

few decades two major causes have been discovered for many of the present day patterns seen 

amongst rain forest plants and animals. These are continental drift and past fluctuations of climate. 

Similarities in flora between the three regions of tropical rainforest occur because all are part of old 

Gondwanaland, The major evolution of the flowering plants had occurred before Gondwanaland 

began to break-up and has continued on the different fragments. 

Climate is the principal factor which controls the growth of plants and constitutes the conditions which 

render a country suitable for the abode of man and animals. Of course other factors such as 

topography, altitude and soil type also influence plant growth. 

These changes in forest cover can thus induce feedback effects on the climate by modifying surface 

temperatures and by influencing carbon dioxide concentrations. Forests have a lower reflectivity than 

other ecosystems and through their extensive root systems have more access to soil water than other 

types of vegetation. In consequence, they absorb more solar energy, which can lead to heating and loss 

of more water through evaporation, which can lead to cooling. In tropical zones, evaporative processes 

tend to dominate and the net effect of forests is to cool and moisten the environment. 

Forests which have been managed primarily for timber production could also be managed for climate 

mitigation and other environmental values. 

The greatest amount of carbon is captured and stored when tree growth is most rapid which is in their 

infant stages. This paper explores how PNG can best exploit that attribute. 

MANAGEMENT OF RISK AND IDENTIFICATION OF OPPORTUNITIES 

The question of how to make the best use of PNG’s forest resources is a complex one and is affected 

by the laws of the market; supply, demand and competition, as well as by environmental concerns and 

even by history.



For PNG producers and exporters to be successful, they must be able to optimize their logistics chain 

to make their integrated timber harvesting/wood production processes an efficient and economical 

material flow and data flow. Without this, PNG producers will not be able to meet the increasing trend 

of globalization in creating a buyer’s market, which is forcing suppliers to produce customer tailored 

goods more cost effectively and more swiftly. Importers need to know volumes of product on offer, 

schedules of delivery and, above all, consistency, quality, quantity and reliability of products being 

exported. 

The problem facing any successful PNG producer/exporter is to find markets for PNG’s residual logs 

and lower quality forest products. 

Risks Facing Forest Developers Include: 

1. Historical Risk 

The natural forests of PNG are not public forests – they are privately owned and unless they become 

part of a forest management ethic and practice on the part of resource owners, it can be questioned 

whether there will be any effective base for the long term management goals of forest policy. 

2. Erosion of Forest Investment Capital 

Unless PNG can halt the decline of the existing forest industrial base through the erosion of 

investment capital — due to impediments such as unsustainable and discriminatory taxation burdens, 

lack of rural infrastructure development and maintenance, and a lack of reinvestment by PNG back 

into the forest sector — there will be enormous poverty in the rural areas. 

3. Impact of PNG Government Decisions 

The PNG forest industry is on the “user end” of government decisions and actions with regard to a 

stable investment climate, resource security, and consistent administration of rules and regulations, 

especially those relating to forest revenue systems in the context of fiscal stability. Without such 

government action, the sector will falter and PNG will fare poorly in terms of the sustainable 

development of its forest resources. 

4. PNG Forest Revenue System 

PNG needs to harness the potential of the forest industry to become partners in development by 

providing realistic operating and fiscal conditions, which would bring back confidence to the sector. 

The forest revenue system, including export tax, needs to be a tool for achieving sustainable forest 

management, not just a means of raising government revenue. 

5. Consistency of Long Term Policies 

With regard to policies and practices which would encourage industry development, it is not just a 

question of policies being conducive to attracting the initial investment. Rather it is the whole policy 

matrix, and how practices are formulated within that matrix, that will influence the investment 

outcome. To achieve an internationally competitive industry, policies must focus on international best 

practice to be successful. Accordingly, policies, which guarantee a secure resource base, in large 

enough volumes to support the investment, must be the starting point. 

6. PNG Forest Industry Association members are restricted by regulation from undertaking many 

development activities. For example: 

• Reforestation levies which are paid to government to plant more trees. 

• The government is responsible to maintain infrastructure in concession areas. 

• Roads in concession areas are constructed only to allow harvesting to be undertaken. 

7. Poor Land Use Planning Controls allowing the emergence of alternative economic 

development in designated forest development areas, which permanently remove forest cover, such as 

oil palm plantations, other cash cropping, and mining.



HOW CAN GOVERNMENT AND INDUSTRY COLLABORATE TO REMOVE OBSTACLES 

TO GROWTH IN THE FOREST SECTOR? 

• Together, industry and government must identify and then resolve issues which impede the 

development of a viable downstream processing industry in PNG. 

• The role of government is to act as a catalyst for change by providing support framework to 

help implement change and assist with specific actions. 

In the forest sector of PNG there is significant potential for new employment creation together with 

wealth creation and rural development for resource owners and the nation. This is of particular 

relevance when it is remembered that the forest sector of the economy operates in rural areas where 

job opportunities and development have been in decline over recent years. 

However, if PNG wishes to alleviate poverty by generating wealth and creating new employment it is 

trade and new trade opportunities that will be the catalyst, not foreign aid, . 

Success will be achieve if strong and competent operators are maintained, and attracted to invest more 

in PNG, under regulations which give a fair deal to all parties – resource owners, resource investors 

and GOPNG, both for current and future harvesting cycles from production forest. 

OPPORTUNITIES FOR PNG FOREST INDUSTRY 

The Capacity of PNG Species for Plantation Development and Carbon Capture 

Much work is required to bring rainforest under scientific management because, in a broad sense, 

rainforest varies in physiognomy and floristic composition with varying environmental factors which 

shape all plant communities. Any management techniques must ultimately rest upon a sound 

understanding of forest ecology. 

If PNG were to use the natural forests as carbon sinks, the management of these natural forests would 

pose difficulties since it requires such detailed ecological knowledge of a vast number of species. In 

the natural forest there is considerable variation in the growth rates of the valuable species besides an 

abundance of secondary species which have no solid wood value and may in fact be poor carbon fixers 

due to their poor growth rates. 

This is a gross oversimplification, as economic factors in particular frequently outweigh the ecological 

factors in significance. The basic premise remains, however, that economic conditions in this 

technological age can change with alarming rapidity, but the ecological behaviour of rainforest 

communities and of the species making up those communities is relatively immutable and can be 

disregarded only at great risk to the final success of management operations. 

However, for PNG, it has the notable feature that many of its native species behave well in plantation 

format and, with their notable feature of fast growth rates, demonstrate the capability to produce, in 

half the time or less of the northern and southern hemispheres, wood with acceptable properties 

suitable for all forms of value-adding processes, besides capturing and storing large amounts of 

carbon. 

Through silvicultural plantation techniques, further refinement of plantation growth (and hence carbon 

capture) can be undertaken by: 

• Matching species to site conditions for optimum growth 

• Improving establishment, tending and maintenance regimes 

• Manipulating the environment and nutrients 

Potential PNG Plantation Species 

Once the end users have been clearly defined, the choice of species, whether indigenous or exotic, is 

dependent on the end use or the environmental role of the plantation, coupled with the matching of 

specific species to specific sites. For PNG proven plantation species include:



INDIGENOUS EXOTIC 

A. cunninghamii; Araucaria hunsteinii; Acacia 

mangium; other acacias; Eucalyptus deglupta; 

E.pellita; Terminalia sp; Pometia pinnata; 

Casuarina spp; Pterocartpus indicus; 

Octomeles spp 

Tectona grandis 

Pinus spp 

Ochroma lagopus 

Benefits of Plantation Development in PNG 

• It represents a major opportunity for growth in PNG’s long term wood supply as well as emerging 

emission trading scheme markets. 

• It would be a significant contributor to regional economic development by value adding locally to 

primary production and generating downstream processing jobs. 

• It provides a long term agricultural crop which can be managed to produce large volumes of wood 

per unit area. 

Action Plan for a properly constructed Emission Trading Scheme using PNG Forest Resources 

Such an action plan should include not just government policy measures and the supportive roles of 

government agencies, but more importantly also indicate how forest plantation activities can be made 

an attractive business proposition to investors through various financing options. 

For the partnership in emissions trading to be sustainable and commercially successful, an action plan 

needs to ensure: 

• The optimum species are used for both wood quality and carbon capture 

• Verification of the mechanism for carbon capture 

• Development of core competencies in-country to accomplish the two dot points above 

• Establishment of a reputable carbon market and a sound regulatory framework 

• Development and adoption of a national financial incentive mechanism to fight deforestation 

• The proper use of forest products such that carbon is stored and not released to the atmosphere 

• Resolution of the conflict between the use of forest products as a source of biomass energy 

and its consequent release of CO2 

• Mobilization of the existing hardwood industries to undertake plantation development 

• Development of access to investment capital (for establishment and processing) 

• Encouragement of confidence in the economics of plantations 

• Mechanisms are in place to develop and implement a communication strategy to deal with: 

o The lack of silvicultural knowledge 

o Support for the implementation of plantation establishment 

o The need to influence landholder, industry and community attitudes 

• Identification of the land base and product priorities through planning initiatives and 

management over 5, 10, and 20 year horizons, in conjunction with regional development 

organizations under the auspices of the National Forest Plan 

The Need for Strategic Industry Status for PNG Industrial Forest Plantations 

PNG is at the crossroads of forestry development, with the painful experience (economically and 

socially) of developing its natural forest resources further, whilst at the same time greatly expanding 

its forest plantation resources. Of course, the key elements of structural change in the PNG forest 

industry, and especially plantation development, are related to a variety of issues concerning demand, 

supply and international trade in forest products. Above all, what products are going to be made and 

marketed from plantation grown resource? 

From industry experience, the challenge in the longer term if the industry wishes to survive is to 

establish its own plantations and to grow wood for a range of markets rather than a single product, 

removing the uncertainties of the single-product markets. A bigger challenge for the industry is that, if 

it is to create new forest resources to compete in the world markets, it must be competitive.



The importance of forest plantation development in PNG to augment the supplies of logs from the 

natural forest needs to be recognized by both the government and the private sector. Plantation forests 

are seen as complementary to natural forests, never able to replace all the values associated with the 

natural forests but, where appropriately developed, helping to divert some of the pressures away from 

them. 

The government needs to accord “strategic industry” status to the forest plantation sector. The 

response to date has not been encouraging (only some 60,000 hectares since 1975) and much more 

needs to be done to study the issues surrounding forest plantation development. 

There have been no programs to encourage small and large investors for many years. Yet, under 

previous programs of some 35 years ago, forest extension programs alone saw over one million trees 

being planted each year by small holders for fuel wood, shelter wood and to augment future sawlog 

supplies. 

The issues surrounding forest plantation development have been discussed at numerous meetings, 

seminars and conferences. Reasons for poor performance such as inadequate or ineffective incentives, 

poor support from financing institutions, resource security in terms of land base and social issues, 

markets and shortage of sizeable land areas have been cited. As forest plantation development is 

becoming increasingly critical for the sustainability of the PNG wood based industry, it is imperative 

that concrete steps be taken to address the issues regarding forest plantation development and a 

practical action plan that could be implemented as soon as possible be formulated. 

May be this action plan could be included in the next revision of the PNG National Forest Plan? 


It is hoped that this policy paper will create opportunities for the PNG forest sector to diversify into 

new markets, assist to set the scene; ensure proper market outlooks, allow for growth in the sector 

though privatisation and competition by attracting investment and strengthening PNG’s research and 

development base. 

PNG, through large scale plantation development, has the potential and the capacity (land, soils, water, 

air and adaptable plantation species) to capture and store carbon in its biomass, soils and forest 

products in perpetuity. If, through the proper introduction of an emissions trading scheme, increased 

value and returns to the forest owners and developers can be realised, PNG has the capacity to 

increase: 

• Five fold its commercial timber tree plantation estate to some 300,000 hectares. 

• 20 % increase in forest industry employment and flow on effects to other employment sectors 

• an increase in value of forest products production of more than 100 %. 

REFERENCES 

Appanah S & Weinland G 1993. Planting Quality Timber trees in Peninsular Malaysia –a review Malayan 

Forest Record # 38 Forest Research Institute Malaysia 1993. 

Baur George 1961/62. The ecological basis of rainforest management. FAO 

Hillary, Sir Edmund 1984. Ecology 2000. – The changing face of earth. Editor. Michael Joseph London. 

McCarthy R B 2003. The Business of PNG Forestry – Implications for Australia and PNG. PNG Forest 

Industries Presentation 20 th Australian PNG Business Forum Cairns. 

Papua New Guinea Forest Industries Association, 2008 Submission to the Garnaut Review on the Impact of an 

ETS on commercial forestry and development outcomes in PNG. PNG Forest Industries Association. 

Silver, Jerry 2008. Global warming and climate change demystified. McGraw Hill. 

Tennesen Michael . The complete idiot’s guide to global warming. 

Whitmore T. C. 1990 An Introduction to Topical Rainforest. Oxford University Press.



2009 - 2059 PNGFIA PLANTATION VISION 

“Five fold increase in commercial timber plantation area to 300,000 hectares.” 

APPENDIX 1 

1. PNGFIA goal is to build an internationally competitive, market orientated industry driven by 

private sector investment 

2. PNG has the necessary land resources, rainfall, commercial tree growth productivities 

compared to rest of world 

3. However, PNG needs the correct investment climate and resource security in order to 

undertake large scale plantation investment 

4. Benefits of large scale plantation development in PNG is that there is a creation of separate & 

alternative income streams for rural dwellers; enhanced regional development opportunities; 

increased employment in rural areas and environmental benefits e.g. reduced soil erosion, 

improved water quality, carbon sinks under a properly constructed emissions trading scheme 

5. The achievement of 2059 vision is the responsibility of the plantation growing industry, i.e. 

industry will be planting the trees, not the government 

6. Plantation wood cannot replace all timber uses. PNG will require continuing use of native 

forests for wood production with plantations supplying an economically viable, reliable and 

high quality wood resource which complements wood supplied from native forests 

7. PNG one of only a few countries capable of increasing its plantation base in next 50 years 

8. PNG plantation species (native and exotic) have great potential particularly if properly dried 

and presented. However, any development must be internationally competitive. 

9. PNG needs to look at the use of effluent for fuel wood plantations within urban areas 

10. There needs to be development of PNG commercial tree growers/ use of SRPM model for 

commercial woodlot owners 

11. PNG needs to develop cabinet wood plantations 

12. PNG needs to develop a code of practice for afforestation companies



MANAGING FOREST COUNTRY: 

ABORIGINAL AUSTRALIANS AND FORESTS 

Sue Feary 1 , Peter Kanowski, Richard Baker and Jon Altman 

ABSTRACT 

Aboriginal Australians have a diversity of interests in their forests, encompassing 

cultural, economic, environmental and social values. Historically, the agencies and 

industries comprising the forests sector have engaged with only some of these interests, 

and have typically done so in a segmented fashion. Our research with Aboriginal 

communities around Australia suggests a myriad of opportunities for a broadly defined 

forests sector, but this requires improved relationships between Aboriginal people and 

the dominant culture and much deeper understanding of Aboriginal aspirations at the 

local level. The National Indigenous Forestry Strategy promotes these aspirations, but 

requires a much stronger commitment from governments if it is to deliver them. 

INTRODUCTION 

Aboriginal people’s views and aspirations about the forest sector have not always been well 

understood by forest managers, policy makers, or the community at large. Forest debates of the 1970s 

in Australia and subsequent adoption of the principles of sustainable forest management (SFM) have 

enabled Aboriginal voices to be heard through several consultative processes. These have consistently 

demonstrated three important goals for Aboriginal people: employment and economic development in 

the forest sector; recognition of customary rights over forests; and protection of the spiritual and 

cultural values of forests. 

Relevant national policy initiatives, such as the National Indigenous Forestry Strategy, aim to achieve 

these multiple goals, but this can be difficult where they are seemingly at some odds with each other. 

One of the barriers is the sectoral approach that state and federal governments have adopted for the 

management of forests, with a dichotomy established between conservation of forest lands on the one 

hand and their commercial exploitation for timber on the other. Thus, for example, forestry operations 

provide employment, but logging and road construction can threaten significant cultural places. 

Alternatively, reservation of forests in conservation areas ensures protection of cultural values of 

forests but tends to offer limited opportunities for economic development and may restrict certain 

cultural activities. This creates something of a dilemma for Aboriginal people for whom traditional 

processes of ‘caring for country’ involved both resource exploitation and protection. 

At the local level, Aboriginal people engage with the forest sector in a variety of ways, reflecting both 

the diversity of the forest sector and the social heterogeneity of Aboriginal culture. This paper 

presents the results of recent doctoral research throughout Australia, demonstrating that a ‘cultural 

match’ between the type of forest-related activity and the type of Aboriginal social unit can result in 

many different forms of productive partnerships within the broad paradigm of ‘Aboriginal forestry’. 

This research demonstrates that, in some instances, the forestry agenda has already been reshaped in 

response to local community situations. The challenge of the National Indigenous Forestry Strategy is 

to scale up local successes for wider application. 

CREATING A SPACE FOR ENGAGEMENT 

Native forests are part of the traditional landscape of Australia’s Indigenous people who have 

managed them for their utilitarian and spiritual values for thousands of years. It has been argued that 

colonial history was a dual war – a war against nature and a war against the natives (Rose, 2002:2). 

What was once a holistic landscape imbued with Indigenous meaning has become fragmented and its 

Aboriginal owners have, for the most part, become detached from it. Little attention was paid to 

1 The Australian National University, Canberra. Contact author: suefeary@hotkey.net.au, Phone 02 44286312



Aboriginal people’s views on forest management until after the 1970s, when timber production forests 

became the centre of debates about whether or not the forestry practices of the time were sustainable 

(Dargavel, 1995). The subsequent recognition of the multiple values of forests provided the first 

formal opportunities for Aboriginal people to articulate their views and interests at a national level. 

The principles of sustainable forest management have been important in creating a space for 

engagement between Aboriginal people and the forest sector in three main areas. Firstly, a greater 

recognition of environmental and socio-cultural values reflects a move from a forest management 

approach towards a forest ecosystem approach (Holling et. al., 1998). This is important from an 

Indigenous perspective because recognising trees as part of a functioning landscape of complex 

ecosystems and natural cycles resonates with indigenous worldviews of holistic, integrated ecosystems 

that also include humans (Rose, 1996). The second concerns the major objective of sustainability -— 

finding balances between social, cultural, environmental and economic factors. This opens the way for 

using participatory mechanisms to include Aboriginal people in negotiations, and recognises that they 

have interests in each of these arenas. Finally, both the principle and the practice of SFM specifically 

acknowledge the rights of indigenous peoples to have their concerns heard and to participate in 

decision-making over forests (Higman et al., 2005). 

Nationwide consultation with Aboriginal people about forests and forest management has occurred 

three times over the last three decades. The first was through the Resource Assessment Commission’s 

(RAC) Inquiry into the Forest and Timber Industry in the late 1980s, when anthropologist Dr Scott 

Cane conducted a review of Aboriginal attitudes to forests through a series or workshops and 

interviews (Cane 1990). Cane’s report highlighted a multiplicity of Aboriginal interests in the forest 

sector, including land ownership, site protection, consultation, land management and access for 

cultural purposes, such as hunting. 

The second consultative process was associated with the Regional Forest Agreements (RFA), 

emerging from the 1992 National Forest Policy Statement, which coincidentally was agreed in the 

same year as recognition of native title. A review of social assessments carried out during the 

Comprehensive Regional Assessments between 1996 and 2000 (Black et al., 2003; Brooks et al., 

2001), demonstrated that Aboriginal people had reiterated their previous concerns. Indigenous 

people’s interests in the forest sector, arising from the RAC and CRA processes, are summarised 

below: 

¤ Consultation and participation in forest management decisions 

¤ Land management –soil and water quality 

¤ Access for cultural activities 

¤ Site management and protection 

¤ Cross-cultural awareness training for non-Aboriginal agency staff. 

¤ Land claims/native title./prior ownership recognition 

¤ Business development 

¤ Employment and training opportunities 

¤ Compensation/royalties 

¤ Formation of agreements and partnerships 

¤ Control of cultural information/ mechanisms for ensuring confidentiality for cultural 

information 

The most recent forest-related consultation with Aboriginal people was during development of the 

National Indigenous Forestry Strategy (NIFS) in 2003 (Australian Government 2005). Aboriginal 

people’s views were canvassed at workshops on perceived opportunities and barriers to forest-based 

development in their local areas (BDO Consulting (SA) Pty Ltd, 2004). The focus of these 

consultations was on economic development, rather than on the broader matters of land management, 

rights and social justice, and drew attention to Indigenous interests in plantations, something which 

had not been documented previously.



Despite the intensely localised nature of the demands by Aboriginal people, three major themes, 

emerged from these consultations: economic development, recognition of rights, and protection of 

cultural values. The fact that these aspirations have been articulated by Aboriginal people since the 

RAC inquiry in 1989 demonstrated that overcoming social and economic disadvantage is not just 

about employment and wealth generation, but must be situated in broader social justice agendas. 

It is important to acknowledge that the benefits of creating a space for engagement between 

Aboriginal people and the forest sector are not just one-sided. Aboriginal Australians now own 16% of 

Australia’s land area and 14% of Australia’s forests (Montreal Process Implementation Group 2008). 

The National Indigenous Forestry Strategy noted that the forest sector could benefit from access to 

Aboriginal land, both for growing trees and for utilising Aboriginal owned forests and woodlands 

(Australian Government 2005). 

Additionally, since 1993, all forests on crown land are potentially subject to native title. Managers of 

both wood production and protected forests need to have dialogue with traditional owners to address 

uncertainty in regard to the impact of native title rights and interests on their activities. In most 

instances, these negotiations are opting for agreements and settlements outside of the lengthy court 

hearings, for example, Indigenous Land Use Agreements or joint management arrangements. 

Agreements of this kind go some way to addressing the three main aspirations of Aboriginal people 

while also providing some level of certainty to forest management authorities. The negotiation and 

engagement pathways of timber production and conservation agencies are discussed later in this paper. 

CONCEPTUALISING ENGAGEMENT 

The forest sector and Indigenous society are separate complex domains comprising multiple parts. 

Dovers (2003:26) points out that policy processes have not acknowledged that there is more than one 

forestry industry. Meanings ascribed to ‘forestry’ have broadened as a result of societal change and 

political processes over the last thirty years. Thus, ‘forestry’ is no longer just about growing tall, 

straight trees to cater for the needs of the timber industry, but about forest management catering for a 

range of forest based interests and values. Globally, the economic values of forest products have 

expanded to include non-timber forest products and, the commercial use of forest’s intrinsic values 

through ecotourism. Most recently, debate on climate change has focused attention on the 

environmental services of forests, and the role that indigenous people can play in environmental 

stewardship (Wunder, 2005). 

Turning to the Indigenous side of the equation, there is a prevalence in Australian society to construct 

all Aboriginal people as united and part of a same-thinking, pan-Indigenous society, regardless of 

whether its subjects live in inner-city Redfern or in the desert. Many non-Aboriginal people see the 

diversity of views and opinions expressed within an Aboriginal community group as signs of 

dysfunction and barriers to resolving an issue. Consequently, rather than developing systems for 

responding to diversity, effort focuses on trying to get consensus (Sullivan, 2004). This is not confined 

to Aboriginal Australians; any minority group tends to be regarded by the dominant culture as 

homogeneous (Dyck, 1985). 

Conversely, recognising diversity and heterogeneity within both the forest and Indigenous domains 

creates a useful space for exploring connections between the fine-scaled elements of each within the 

respective bigger pictures. The conceptual tool used to bring these two complex systems together is a 

matrix, comprising categories of the forest sector along one axis and Indigenous social units along the 

other, as shown in Figure 2. Historical legacies and geographical location comprise third and fourth 

dimensions, in recognition of their influences on contemporary Aboriginal culture. 

The depiction of Indigenous social structures shown in Figure 2 does not purport to reflect any 

particular theoretical position in anthropology. Different kinds of Indigenous organisations and 

structures are formed in response to different legislative requirements or other external influences: 

some may be strongly influenced by customary factors, others less so. Rather, this model is schematic 

and illustrative of Aboriginal representation. Similarly, the forest sector axis is indicative rather than 

exhaustive.



Figure 2: A conceptual diversity matrix linking Aboriginal people and the forest 

sector. 

The matrix is a heuristic device for exploring the diversity of connections between Aboriginal society 

and a broadly defined forest sector, premised on finding the right ‘match’ between forestry sector 

category, and ‘type’ of Indigenous social structure. Examples of appropriate matches could be: a 

natural resource management project in a remote area involving a group of traditional owners on their 

country; or an apprenticeship in a timber yard for a young Aboriginal man living in a regional 

township. Examples of other potential matches are shown in Figure 2. An appropriate ‘match’ between 

the forest sector unit and the Indigenous unit may be a significant factor in creating an enabling 

opportunity for addressing disadvantage. The Harvard project in America is having considerable 

success in using a similar model for developing effective local governance structures in Indigenous 

American communities (Cornell and Begay, 2003), and is gaining popularity in Australia (Behrendt, 

2003). Application of this model in forms appropriate to Australia offers good prospects for enabling 

mutually beneficial cooperation between Indigenous Australians and the forest sector. 

TESTING THE MODEL 

Case study research was used to conduct in-depth analysis of the relationships between the forest 

sector and Aboriginal people, and to test the matrix model described above. Four main themes guided 

case study selection – remoteness, management of Crown native forests, involvement with the 

plantation industry and forest-based economic development. Research was conducted in four forested 

case study locations – western Cape York in far north Queensland, southwest Western Australia, the 

Riverina where it straddles the NSW/Victorian border, and southeastern New South Wales (Feary, 

2007). 

Remoteness 

Forest based relationships in remote settings were explored through partnerships between the 

Queensland government and the small Aboriginal communities of Injinoo at the tip of Cape York and 

Napranum near Comalco’s mining town of Weipa. Over the last decade, government agencies have 

invested effort in establishing forestry enterprises to assist social and economic development in a



region where opportunities for employment and business development are extremely limited and the 

majority of the Aboriginal population receives welfare support (Annandale and Feary, 2009). 

The Injinoo sawmilling project was a partnership between the Injinoo community and the Queensland 

Government, aimed at building a team of skilled local people capable of using their forest resource to 

construct houses and visitor facilities (Annandale et. al., 2002). From 2001, local Aboriginal men were 

trained and accredited in harvesting, transportation, milling and grading of timber (Annandale et. al., 

2002). On returning to the area to evaluate progress in 2005, government staff were met with marked 

trees still unfelled and an idle sawmill (Feary, 2007). Although still enthusiastic, the community 

expressed little commitment to keeping the project running. Reasons included tensions between 

different community organisations and the need for ongoing mentoring by the government to build 

confidence and expertise among the trained men. 

The situation was different at Napranum in 2005, where bauxite mining, which displaced Aboriginal 

people from their traditional lands in the 1950s, is now assisting in a viable community-run sawmilling 

operation. Until recently, Darwin stringybark forests in the mining leases were cleared and burnt by 

Comalco to allow access to the bauxite layer beneath (Annandale and Taylor, 2007). Not only was this 

a waste of usable timber, burning of extensive areas of forest was contributing to greenhouse gases. In 

the early 1990s, a new Indigenous owned business called Nanum Tawap Ltd was established, to 

harvest useable timber from the lease area prior to mining. The business is run by members of the five 

clan groups whose traditional country is covered by Comalco’s mining lease. The sawmill sells to both 

a local and export market and currently has a non-Aboriginal manager and employs five local 

Aboriginal men on average. 

Superficially it could be a small sawmill operation anywhere in Australia, but conversations with 

sawmill workers gave insight into the connections between cultural behaviours, economic activity, and 

the capacity for adaptation. Both Aboriginal and non-Aboriginal people recognised barriers to 

mainstream employment arising from a history of lack of workforce participation, limited formal 

education and social problems characteristic of poor, marginalised peoples. The sawmill operation has 

some innovative and pragmatic adaptations for ensuring its economic viability. Firstly, in recognition 

that the workforce will vary in its numbers and skill levels from day to day, the make of sawmill, 

although relatively more expensive, can be operated by one man if necessary, i.e. if nobody turns up to 

work. Secondly, the issue of vouchers to buy sought after white goods at the local electrical goods 

store was an unorthodox productivity incentive. 

The Nanum Tawap sawmill is a mainstream forestry operation situated in a cultural setting. There are 

arguments both for and against separating culture from business in Aboriginal communities (Altman, 

2001) but the approach taken at Napranum has been to separate the business from the operations and 

activities of the rest of the community. The workplace is separated from the wider Napranum 

community by physical, administrative, and cultural boundaries. But, cultural traditions are recognised 

at higher levels; in the administrative and management structures of the business, where the Board has 

real power in decision-making and where traditional clan territories are recognised and respected. 

Evaluation of the timber resource by the Queensland government indicated that the sawmill’s capacity 

would need to be substantially increased to maximise economic returns from timber salvage 

operations (Annandale and Taylor, 2007). The Nanum Tawap Steering Committee considered these 

options and supported a business model that would maximize local employment by having three small 

sawmills, one in each community, rather than a single large mill requiring fewer people but needing 

higher levels of technical skills. This decision is significant for demonstrating that building social 

capital in local communities can take precedence over maximizing economic returns. 

Management of Crown native forests 

Naturally vegetated lands, including forests, are significant to many Aboriginal people. They are a 

reminder of a once holistic Indigenous-owned landscape from which they were alienated during 

colonisation, especially in ‘settled’ Australia. Since most native forests are owned by the crown 

(Dargavel 1995), Aboriginal connections with native forests occur mainly in the context of 

government instrumentalities. All tiers of government recognise, to a greater or lesser extent, social 

justice and indigenous rights, together with state land rights and native title legislation, in their



policies, vision statements, codes of practice and the like. Most include mechanisms for consulting 

with Aboriginal people with varying degrees of effectiveness. 

Mechanisms for involving Aboriginal people in forests on Crown land, both wood/fibre production 

and protected area forests, were explored through case study research in Western Australia, the 

Riverina and southeastern NSW (Feary, 2007). Interviews with Aboriginal people suggested that 

consultation processes were very important in defining their relationship with the resource extraction 

sector. In fact, Aboriginal people frequently judged forestry agencies as much on their consultative 

and participatory processes as on their skills and abilities as forest managers. 

Interviews with non-Aboriginal people revealed many of the complexities of consulting with 

Aboriginal people, especially in relation to representation and identifying those with the ‘right to 

speak’. The increasing numbers of Aboriginal community organisations with interests in land 

management, together with intra- and inter-community contestations over traditional connections with 

the land, often result in unsatisfactory outcomes or prolonged periods of negotiation. Despite the 

existence of agency policies for consultation, both sets of actors feel that the other side could do better. 

Case study research suggested that timber production agencies have applied the principles of 

sustainable forest management quite narrowly; limiting it to consultation, protection of tangible 

cultural heritage, some limited employment and access for local communities to conduct cultural 

activities. At a global scale, this is something of a contrast to other settler societies (NZ and the 

Americas) and developing countries, where the broader rights and interests of indigenous people have 

been incorporated into SFM in these countries. The narrow focus is problematical because there is a 

danger of consultation being an end in itself, representing the beginning and end of Aboriginal 

involvement in a commercial forestry operation. 

Increasingly there are exceptions, including the recent Githabul Indigenous Land Use Agreement over 

forests in northern NSW and southern Queensland, and agreements for managing the river red gum 

forests in Victoria, over which the Yorta Yorta people lost their native title claim (Morgan et al 2006). 

On the NSW far south coast, an intensely negotiated Eden RFA delivered arrangements between the 

Eden Local Aboriginal Land Council and the local forestry office for contract employment and the 

transfer of ownership of some Crown forests to the Land Council. 

Consultation processes were not the main focus of Aboriginal interviewees’ interests in natural 

resource management and conservation. Instead, conversations moved beyond consultation to focus 

more on participatory processes and partnerships that integrated scientifically based land management 

and research with customary Aboriginal knowledge. The intrinsic synergies between nature 

conservation and Aboriginal worldviews about caring for the environment have resulted in relatively 

greater expectations for conservation agencies to play a major role in addressing Indigenous rights and 

social justice. 

However, the views of Aboriginal people reflected ambivalence towards protected area management. 

The challenge for both Aboriginal and non-Aboriginal people, is understanding the similarities and 

differences between nature conservation as practiced by western cultures and Aboriginal worldviews 

about land management, often labelled ‘caring for country’. There is a commonly held belief among 

non-Indigenous Australians that conservation of nature through protected area management is a 

modern equivalent of traditional practices of ‘caring for country’. It is true that many overlaps do exist 

and joint management of national parks across the country attests to the synergies. But, at the same 

time these cooperative arrangements mask the inevitable tensions between Aboriginal owners and 

national park agencies. From an Aboriginal perspective, ‘caring for country’ includes the exploitation 

of natural resources through hunting and collecting, which is at odds with the objectives of 

contemporary nature conservation goals (Young 2001). Understandably, many Aboriginal people 

consider that conservation objectives of the western world are too restrictive and serve only to further 

separate humans from nature. At the same time, they also know that a national park (as opposed to 

development) offers the best chance to protect the natural landscape and its associated cultural values.



Involvement in the plantation industry 

Plantation forestry offers an alternative to logging native forests. Although the focus is on commercial 

wood and wood fibre production, the functions of plantations in reversing land degradation are 

becoming recognised (Varmola, 2005). Aboriginal people’s engagement with plantation forestry may 

occur as employees or as landowners, and both were examined through case study research in Western 

Australia and Cape York. 

Plantation forestry is a major industry in southern WA with thousands of hectares planted to 

Tasmanian blue gums and other species. An Esperance based company, Integrated Tree Cropping 

(ITC), contracted the Esperance Aboriginal Corporation (EAC) to undertake various projects at their 

plantations. ITC has organised training courses with a view to making Aboriginal workers ‘job ready’ 

and is encouraging its Aboriginal workers to become proficient in harvesting techniques, to take 

advantage of new job opportunities when the current plantations are ready to harvest. 

Case study research in WA suggested that local Aboriginal people did not have a very good reputation 

in relation to work ethics, which, together with lack of training and education, has created barriers to 

Aboriginal employment in plantation forestry. However, ITC were making a considerable effort to 

engage Aboriginal people, while acknowledging that it required extra effort in providing training and 

may result in less efficiency and be less productive. The company gave Forest Stewardship Council 

(FSC) certification as a basis for their commitment to providing opportunities for Aboriginal 

employment. Recognising the rights and interests of indigenous peoples is an FSC criterion (Forest 

Stewardship Council, 2007). 

EAC members enjoyed the work because it was outside, enabled people to work together and had 

flexible working hours and to this end, demonstrated a strong work ethic. Other Indigenous people 

interviewed in WA supported sharefarm arrangements for plantations on their land because of long 

term and short economic returns, providing useful employment as well as helping to restore degraded 

agricultural land. 

Similar sentiments were expressed by an Indigenous man involved with sandalwood plantations on 

Cape York : 

‘The land needs to be in good condition when handed back after mining. That’s why we 

are growing sandalwood, to make the land healthy…‘We’ll just keep planting 

sandalwood so when Comalco does decide to shift out, traditional owners can still get 

money from the land by harvesting the sandalwood… Planting sandalwood will 

hopefully provide jobs and put money back into the community, reduce handouts from 

the government. That’s what I’d like to see’ (Feary, 2007). 

The need for land to grow trees, combined with positive responses by Aboriginal people to 

plantations, situates plantation forestry favourably for contributing to Indigenous people’s social and 

economic wellbeing. However, plantation forestry is highly commoditised, with engagement occurring 

in the mainstream market economy in a largely privatised industry. Aboriginal people are competing 

in a labour market for which they may not be ready or want to enter. There is little room for 

‘indigeneity’ in an investment driven industry and, without targeted education and training, neither 

individuals nor organisations will be competitive. The plantation industry is unlikely to enter into 

contractual arrangements if there is a real or perceived risk to business profits. Historically, the private 

sector has also been reluctant to reduce this risk by investing in training and development, tending to 

leave it to the public sector, although this is changing (Timber 2020). The Integrated Tree Cropping 

example indicates that forest certification may play an increasingly important role in building the 

capacity of Aboriginal people to be employed in plantation forestry. 

Forest-based economic development 

Arts and crafts and forest-based ecotourism: Outside of main stream employment, native forests offer 

a broad range of potential economic opportunities at the local level. Small-scale enterprises such as 

arts and crafts and cultural tourism are often attractive to Aboriginal communities because they can be 

managed locally, do not require high levels of technology and provide local jobs. Most importantly, 

they are a mechanism for using customary knowledge and law for deriving an income in the market 

economy. In some remote areas it is one of the few ways for Indigenous people to gain some income



from activity on their lands while maintaining social and cultural practices (Hall, 2007). However 

pressures of the market can undermine customary systems, as research into the use of the tree Bombax 

for making wooden artefacts has demonstrated (Koenig, et al 2005). 

Agroforestry for the commercial production of bushfoods is another enterprise with potential for local 

Aboriginal communities and there has been recent research into the economic viability of small-scale 

plant harvest of a range of food species (Whitehead et. al., 2006). Non-wood forest products from 

Australian forests, particularly bush foods and medicines are gaining popularity in Australia and 

overseas, and a considerable amount of research and development is directed towards the genetic 

development, domestication and finding sustainable levels of wild harvest of a wide range of native 

species (see www.rirdc.gov.au for examples). However, issues still surround the intellectual property 

rights of Aboriginal Australians (Janke, 1998) and although community based bush food enterprises 

seem an ideal, currently the industry is dominated by non-Indigenous companies. 

Cultural heritage management – managing and protecting sites – is an area where traditional 

knowledge forms the basis of viable economic ventures in Crown forests. All states have legislation 

for protecting Aboriginal heritage, together with policies requiring Aboriginal consultation. These 

have provided a springboard for employment and have greatly enhanced input by Aboriginal people 

into management of their cultural heritage. For example, Forest NSW engages the local Aboriginal 

Land Council to undertake heritage surveys prior to logging and the work has become a significant 

source of income for the Land Council. 

Western Arnhem Land Fire Abatement project: Research in west Arnhem Land has shown that 

strategic fire management in savannah landscapes can reduce greenhouse gas emissions from 

savannah fires. This research has led to a greenhouse gas offset agreement between 

ConocoPhillips, the NT Government, and local Aboriginal groups. Indigenous fire managers 

will be paid around $1 million a year for 17 years to provide this fire management service. 

The project seeks to increase the proportion of controlled early dry season fires to create fire 

breaks and patchy mosaics of burnt and unburnt country to minimise the occurrence of 

destructive late dry season wildfires, thereby reducing greenhouse gas emissions (NAILSMA 

2006). 

UNDERSTANDING ENGAGEMENT 

Analysis of case study data has demonstrated that engagement between the forest sector and 

Indigenous people took the form of individual and community based employment, various forms of 

partnerships and agreements in economic and non-economic capacities, and a wide range of 

consultative and participatory processes. The type of engagement was shown to be affected by the 

element of the forest sector under consideration, although this was tempered by physical setting and its 

associated historical legacies. 

Indigenous cultural factors were always present at the forest sector-Indigenous interface, but took 

many forms. Sometimes culture was expressed by the tangible and intangible values of the forest 

itself. Sometimes it was about the means for Aboriginal elders to pass on customary knowledge and 

for the next generation to practice and renew culture. In ‘settled’ Australia, many Aboriginal people 

wish to both participate in mainstream society and maintain cultural traditions. The Aboriginal owned 

land base in settled Australia is small, and naturally vegetated lands are mostly owned by the Crown. 

Uncertainty over native title and other processes affording Indigenous rights and interests over Crown 

native forests is a pragmatic reason for both state and federal governments to establish effective 

mechanisms for negotiating outcomes. 

Characterizing Aboriginal engagement with the forest sector prompts reflection about whether the 

stamp of ‘indigeneity’ can be disentangled from the influences of social and economic disadvantage. 

In his analysis of the relationship between identity and culture in the pastoral industry, Richard Davis 

notes that Aboriginal stockmen sometimes were distinctively Aboriginal while at other times their 

work was indistinguishable from that done by white stock workers and managers. He suggests further 

that Aboriginal stockmen also fashioned their own identity by positioning themselves in both worlds 

(Davis 2005).



An examination of Aboriginal engagement with the market components of the forest sector has 

demonstrated some parallels with the model developed by Davis for Aboriginal participation in the 

pastoral industry. Thus certain components of the forest sector, such as working in sawmills makes an 

Aboriginal person indistinguishable from a non-Aboriginal person, notwithstanding barriers arising 

from persistence of negative stereotypes and poor levels of education and training. Voices from case 

study research suggest that in these situations, ‘work is work’. That is, no special dispensation was 

given or sought for indigeneity and paid work was separated from social and cultural activities. Thus it 

is the similarities between cultures that take precedence, rather than the differences. For many parts of 

Australia it is possibly little different from when Aboriginal people worked in local sawmills fifty 

years ago. 

In other examples, Aboriginal people undertake activities in Aboriginal ways. The forest heritage 

business, ecotourism enterprises and arts and crafts all rely on customary knowledge, and their 

expressions are distinctively Aboriginal. However, all operate within a modern market economy 

where Aboriginal people can be vulnerable to market pressure, leading to ecologically unsustainable 

practices, unequal partnerships and a possible concomitant reduction in the all important ‘Aboriginal 

flavour’. 

For remote areas such as Cape York, commerce and culture are intertwined in ways not easily 

explained by extant economic models. Nanum Tawap’s sawmill operation is firmly embedded in a 

cultural landscape but is separated from it physically and administratively. There are special 

adaptations such as the type of saw and use of white goods as work incentives, but essentially it 

operates as a mainstream workplace. Application of cultural practices occurs through traditional 

owners who are empowered through their positions on the Board of the company, where the important 

decisions are made. Culture is still evident in the management of sawmilling operations which is based 

on customary social boundaries and a desire to build social capital rather than maximise profit. 

Case studies dealing with commercial native forestry showed that sustainable forest management was 

focused largely on consultation processes and protection of tangible cultural heritage. Case studies of 

conservation demonstrated that in addition to these aspects, a broader rights and social justice agenda 

operated, leading to active participation in land management and ownership of land, although with 

legislative constraints on customary activities such as hunting and gathering. Plantations and 

agroforestry have the potential to offer employment, business and revenue in the mainstream as was 

shown in the WA case study. But they can also contribute to ‘caring for country’ by repairing previous 

land damage or by providing food and raw materials for Aboriginal communities such as on Cape 

York. 

CONCLUSION 

This paper has presented the results of research that has demonstrated that ‘Aboriginal forestry’ is a 

spectrum of activities and values, ranging from purely economic to purely social. These reflect local 

responses by Indigenous people for balancing desires for maintaining cultural values and traditions 

with aspirations for social and economic development. Mainstream forest based employment and 

business development with no special recognition of indigeneity was found to be a preferred approach 

in several of the case studies. This is relevant for plantation forestry, which could capitalise on 

Aboriginal people’s support for plantations by developing policy to underpin effective and equal 

partnerships for growing trees that delivered economic, environmental and cultural benefits on 

Aboriginal land. In other cases, associations with forests were embedded in cultural responsibilities to 

care for the ancestral beings. 

As with all natural resource use by a nation state, the management and use of Australia’s forests is 

embedded in broader agenda of social justice for indigenous peoples. Experiences from mining have 

shown that royalties in return for resource exploitation do not overcome disadvantage on their own. 

Overcoming Indigenous social and economic disadvantage requires linkages between programmes for 

employment and business development and the broader social goals of health, education, 

empowerment and self-determination. Government policy must be pluralistic by having policies that 

allow Aboriginal people choices about the relative importance they want to place on the economic, 

cultural and environmental values of forests.



Ultimately, forests can be landscapes of reconciliation, through recognising and respecting cultural 

‘difference’: the sentient values of the forests to Aboriginal people, the cultural places they contain 

and their importance for maintaining cultural identity. Forests can also be landscapes for recognising 

‘sameness’ and overcoming discrimination between Aboriginal and non-Aboriginal workers in the 

mainstream market economy. 

REFERENCES 

Altman, J., 2001. Indigenous communities and business:Three perspectives, 1998-2000, Working Paper No. 

9/2001, Centre for Aboriginal Economic Policy Research, ANU. 

Annandale, M., McGavin, R. and Venn, T., 2002. Injinoo Sawmilling Project Phase 1: Small-scale Timber 

Processing., Queensland Forestry Research Institute. 

Annandale, M. and Feary, S., 2009. Consulting with Aboriginal and Torres Strait Islander communities: 

Agroforestry in North Queensland, Canberra, Rural Industries Research and Development Corporation 

Publication 08-168. 

Annandale, M. and Taylor, D., 2007. Forest Futures - Indigenous Timber and Forestry Enterprises on Cape 

York., In Forestry for Indigenous Peoples. Proceedings of Technical Session 130, XXII IUFRO World 

Congress, 8-13 August 2005, Brisbane, Australia., (Ed.), Feary, S., ANU Fenner School of Environment 

and Society, Canberra, pp. 49-59. 

Australian Government, 2005. The National Indigenous Forestry Strategy, Canberra, Commonwealth of 


BDO Consulting (SA) Pty Ltd, 2004. Opportunities and Barriers for Greater Indigenous Involvement in 

Australia's Forestry Industry, Commonwealth of Australia, Canberra. 

Behrendt, L., 2003. Achieving Social Justice, The Federation Press, Sydney. 

Black, A., Baines, P. and Bekle, H., 2003. Towards a Framework for Consultation and for Input of Indigenous 

Knowledge Pertinent to Ecologically Sustainable Forest Management at a Regional Level., Victoria, 

Forest and Wood Products Research and Development Corporation. 

Brooks, K., Kelson, S. and Tottenham, R., 2001. Summary of Australian Regional Forest Agreement Social 

Assessments and a Recommended Assessment Methodology, Canberra, Forest Industry Branch, AFFA. 

Cane, S., 1990. Trees on the Land, Canberra, Report to the Resource Assessment Commission Forest and 

Timber Inquiry. 

Cornell, S. and Begay, M., 2003. What is Cultural Match and Why is it so Important? Lessons from 14 years of 

the Harvard Project, Building Effective Indigenous Governance, Jabiru, NT. Available at: 

http://www.nt.gov.au/cdsa/indigenous_conference/web/html/papers.html (accessed 18/4/2005). 

Dargavel, J., 1995. Fashioning Australia's Forests, Oxford University Press Australia, Melbourne. 

Davis, R., 2005. Identity and Economy in Aboriginal pastoralism, In The Power of Knowledge. The Resonance 

of Tradition, (Eds.), Taylor, L., Ward, G., Henderson, G., Davis, R. and Wallis, A., Aboriginal Studies 

Press, Canberra, pp. 49-60. 

Dovers, S., 2003. Are forests different as a policy challenge?, In Towards Forest Sustainability, (Eds.), 

Lindenmayer, D. and Franklin, J., CSIRO publishing, Collingwood, pp. 15-30. 

Dyck, N., 1985. Aboriginal Peoples and Nation States: An Introduction to the Analytical Issues, In Indigenous 

Peoples and the Nation State: Fourth World Politics in Canada, Australia and Norway., (Ed.), Dyck, N., 

Canada, pp. 1-26. 

Feary, S., 2007. Chainsaw Dreaming: Indigenous Australians and the Forest Sector. PhD, School of Resources, 

Environment and Society, The Australian National University: Canberra. 

Forest Stewardship Council, 2007. FSC Australia Update, Available at: www.fscaustralia.org/whats-new/fscaustralia-update/ 

(accessed 7/10/2007). 

Hall, N., 2007. Telling the story: performance and narrative as reflective frameworks for Indigenous tourism, 

Sydney, CAUTHE Conference2007: Past Achievements, Future Challenges, 11-14 February, 2007, 

Higman, S., Mayers, J., Bass, S., Judd, N. and Nussbaum, R., 2005. The Sustainable Forestry Handbook. A 

practical guide for tropical forest managers on implementing new standards, Earthscan, London. 

Holling, C., Berkes, F. and Folke, C., 1998. Science, sustainability and resource management, In Linking social 

and ecological systems for resilience and stability, (Eds.), Berkes, F. and Folke, C., Cambridge University 

Press, Cambridge, pp. 342-362. 

Janke, T., 1998. Our culture our future: report on Australian indigenous cultural and intellectual property 

rights, Canberra, Australian Institute of Aboriginal and Torres Strait Islander Studies.



Koenig, J., Altman, J. and Griffiths, A. D., 2005. 'Too Many Trees' : Aboriginal Woodcarvers in Australia, In 

Carving out a Future, (Eds.), Cunningham, A., Campbell, B. and Belcher, B., Earthscan, pp. 136-146. 

Montreal Process Implememtation Group, 2008. Australia’s State of the Forests Report 2008. Bureau of Rural 

Sceinces: Canberra. 

Morgan, M., Strelein, L. and Weir, J., 2006. Authority, Knowledge and Values: Indigenous Nations Engagement 

in the Management of Natural Resources in the Murray-Darling Basin, In Settling with Indigenous 

People, (Eds.), Tehan, M., Palmer, L., Langton, M. and Mazel, O., The Federation Press, Sydney, pp. 

135-158. 

North Australia Indigenous Land and Sea Management Alliance (NAILSMA), 2006. 

http://www.nailsma.org.au/projects/indigenous_carbon_abatement.html (accessed 23/7/2009). 

Rose, D., 1996. Nourishing Terrains: Australian Aboriginal Views of Landscape and Wilderness, Australian 

Heritage Commission, Canberra, ACT. 

Rose, D., 2002. Love and Reconciliation in the Forest: A Study in Decolonisation., Magill, SA, Hawke Institute 

Working Paper Series No. 19, Hawke Institute, University of South Australia. 

Sullivan, P., 2004. Indigenous Fragmentation, Minority Rights Scepticism and the Need for Both Re-integration 

and Co-integration, Canberra, Paper given at the AIATSIS Conference, Canberra, 22-25 November 2004. 

Timber2020 2009 Indigenous Forestry Great Southern Region, Western Australia. Strategy and Action Plan, 

Govt of Western Australia and Timber 2020. 

Whitehead, P., Gorman, J., Griffiths, A., Wightman, G., Massarella, H. and Altman, J., 2006. Feasibility of 

Small Scale Commercial Native Plant Harvests by Indigenous Communities., Canberra, RIRDC 

Publication No. 04/149, Joint Venture Agroforestry Program, RIRDC. 

Wunder, S., 2005. Payments for environmental services: some nuts and bolts, Jakarta, CIFOR Occasional Paper 

No. 42, Center for International Forestry Research. 

Varmola, M., 2005. Diversifying Functions of Planted Forests, In Forests in the Global Balance - Changing 

paradigms,(Eds.), Mery, G., Alfaro, R., Kanninen, M. and Lobovikov, M., IUFRO World Series Volume 

17, pp. 117-136. 

Young, E., 2001. Looking after country is men's and women's business: institutional support for Indigenous land 

management, Dialogue, 20(2): 28-32.



COMMUNITY CONSULTATION AND INDIGENOUS FORESTRY 

OPPORTUNITIES IN CAPE YORK PENINSULA 

ABSTRACT 

Mark Annandale 1 

This paper explores community consultation with Indigenous people through the 

development of agroforestry projects in north Queensland. Indigenous people have been 

practicing various forms of agroforestry in Australia for thousands of years, as custodians 

of the natural environment. Traditional land management, often called ‘caring for 

country’, was practiced through strategic use of fire to encourage growth of certain 

species. Some species were planted and others were selectively cultivated to favour 

specific species or faunal habitats. Elements of these traditional systems of land 

management are contained in the modern practice of agroforestry, yet there are few 

examples of commercial agroforestry enterprises involving Indigenous Australians today. 

The case study resulted from an extensive consultation process that developed and tested a 

methodology for effective community engagement. This paper provides a review of 

Queensland sandalwood and discusses options for growing sandalwood species to provide 

sustainable development opportunities in Cape York Peninsula (CYP) in far north 

Queensland. Strategies for further development of the sandalwood industry in CYP are 

discussed. 

INTRODUCTION 

The Sandalwood species that occurs naturally in Cape York Peninsula CYP is Santalum lanceolatum 

(R.Br), commonly known in northern Australia as Queensland sandalwood. Taxonomists have 

recently separated this species from the sandalwood which occurs south of 20 degrees latitude with 

suggested common names of northern and southern sandalwood respectively (Harbaugh, 2007). The 

first recorded sighting and collection of the plant by Joseph Banks and Daniel Solander occurred in 

1770 at the Endeavour River in south eastern CYP (Wharton, 2005). From around 1896 to the early 

1930s sandalwood cutters relied on Aboriginal knowledge of country and for labour to exploit the 

species in north Queensland (Pike, 1983; Wharton, 1985; Statham, 1990). 

Sandalwood is one of the most valuable timbers in the world (Baruah, 1999; Bragg et al., 2004). The 

high price commanded by sandalwood is a result of the increased commercial demand and 

international trade in sandalwood timber and its associated products, compounded by a decline in 

supply. New sandalwood plantations are required to address the overexploitation of the native forest 

resources experienced in most countries and to meet demand for sandalwood products. In the CYP 

minoforestry, the integration of forestry production systems into mine rehabilitated landscapes, 

provides an opportunity for Indigenous business, employment and a range of environmental benefits. 

The CYP is a diverse and ecologically important region of tropical Australia. It covers approximately 

13.72 M ha. The vegetation is diverse and includes extensive areas of dry eucalypt woodland 

dominated by Darwin Stringybark (Eucalyptus tetradonta), Melville Island Bloodwood (Corymbia 

nesophylla) and to a lesser extent Cooktown Ironwood (Erythrophleum chlorostachys), occurring 

mostly on deeply weathered plateaus (Neldner & Clarkson 1996). Areas of rainforest occur along the 

eastern coast, mostly in national parks, with extensive tropical grasslands, heathlands and mangroves 

are also present throughout the region. 

The current population of CYP is approximately 18,000 people, over half of whom are of Aboriginal 

and Torres Strait Islander origin and live in a number of Indigenous communities. Most are relatively 

small and isolated settlements which are self-governing, run by an elected council made up of local 

residents and funded by government, with the largest Indigenous community comprising about 1200 

1 Director. Environment Land Heritage Pty Ltd and PhD candidate at School of Natural and Rural Systems Management. 

University of Queensland St Lucia Qld, 4067 mark.annandale@landheritage.com.au



people. The two regional centres include Cooktown, which serves as an administrative centre, and 

Weipa which is a large mining community, each comprising about 3000 people, including many 

Indigenous people from the region. There are well-recognised opportunities for Indigenous 

communities in CYP to develop, manage and operate viable forestry business enterprises based on 

sustainable management of native forests and plantations (Annandale, 2000; Annandale and Taylor, 

2000; Hopewell, 2001; Annandale et al., 2002; Annandale, 2003; Annandale and Bragg, 2004; Venn, 

2004). Several studies recognise the potential for commercial forest activities associated with pre- and 

post- mining land uses around the western CYP communities, including plantation-grown sandalwood 

and enterprises that can process timber cleared prior to mining activity (Annandale, 2003; Annandale 

and Bragg, 2004). 

Forestry-based business can play a lead role in regional development because of its potential to 

provide resources for value-adding and employment and encourage business. It will help foster 

regional economic development, including forestry-based businesses through development of 

infrastructure, introduction of new technology and employment supported by training initiatives. 

Agroforestry is the practice of combining trees, shrubs and forests with agricultural systems. 

Indigenous people in north Queensland have expressed a desire to develop sustainable commercial 

agroforestry enterprises to provide local employment, enable people to work on country and to 

improve health and wellbeing. 

There is increasing recognition that the social justice agenda for Indigenous Australians incorporates 

the right to derive economic benefit from land-based enterprises and the expanding Indigenous-owned 

land base offers opportunities for various agroforestry enterprises. 

COMMUNITY CONSULTATION 

All non-Aboriginal people intending to work with Indigenous communities should undertake 

comprehensive induction and cultural awareness programs prior to entering communities. They must 

develop the interpersonal skills to freely exchange information and transfer skills that Indigenous 

people require to give a community greater autonomy. Additionally, skills and knowledge cannot be 

passed on without a basic level of understanding of the culture and society of community members. 

When working with Indigenous communities, one must first make steps towards understanding some 

of the background issues which have impacted on the communities and individuals in both recent and 

past times. Be aware that attitudes will be influenced by consultation processes undertaken in previous 

years, some of which may not have been adequate to meet the needs and aspirations of the community. 

The building of meaningful relationships is critical for the successful delivery of any project. This can 

often take a considerable amount of time, especially when many Indigenous people are involved. Lack 

of time to build relationships and a mutual understanding is the main reason that cross-cultural 

projects fail, so it is important to factor sufficient time into work programs. Once established, projects 

and enterprises evolve and mature. There is a vast amount of existing knowledge and experience in 

Indigenous communities, with people who have insights into what types of projects work. For 

successful project development there needs to be an exchange of information and experiences, which 

occurs over a long period of time. 

When discussing project ideas or opportunities both parties must be committed, have confidence in the 

process, be open, honest and provide reliable information. The project must be discussed at length and 

clearly outlined so that expectations remain realistic. Indigenous people are then in a better position to 

decide whether or not they wish to participate in projects. 

A holistic approach is required providing the communities with the tools to freely determine their own 

economic and social development, to empower people to control their destiny. The criteria for 

measuring the success of any business need to be identified and agreed to by all those involved. It 

should not be determined by government or any other external stakeholders. This will keep all 

expectations at a realistic level. 

Education and training needs to be adaptive and responsive to community needs. People need the 

opportunity to have inputs into curricula, including incorporation of cultural knowledge.



Community consultation does not have to be complicated, but it does need to be inclusive and follow 

some basic protocols. Protocol means following the customs or lores of the people with whom you are 

working. 

Effective consultation needs to be designed to meet the unique aspects of the situation and to identify 

and clearly define the issues. Enough time must be allocated for the consultation process and detailed, 

balanced and accurate information must be provided to enable informed discussion and negotiation. 

Once a proposed activity requiring involvement of Indigenous people has been identified, a 

consultation protocol should be followed to ensure that the appropriate Indigenous people are 

consulted in mutually agreed ways. The consultation process should begin as early as possible and be 

flexible. 

CASE STUDY 

Indigenous people in far north Queensland are interested in pursuing socially, culturally, 

environmentally and economically sustainable development on their own land. They have identified 

the need to become more self-sufficient and to engage with mainstream society on an equal footing. 

Indigenous people identify a common desire to return to country. They want to get back to country 

with a purpose, which includes educating younger people in traditional ways, looking after country 

and/or being able to make a living compatible with the lifestyle of the people involved. Therefore 

projects that involve people getting back onto country are attractive and often provide an opportunity 

for younger people to get to know their country in greater detail. Elders in several Cape York 

communities have talked about the time when many people lived on country, in outstations, as a time 

when many of the social problems, and dependency on welfare did not exist. There is a desire to find a 

balance between living in towns or communities and living on country. Agroforestry business 

development may provide this opportunity. 

Currently, sustainable land use options available to Indigenous communities are limited. New 

industries involving some form of agroforestry could offer opportunities for ecologically sustainable 

development and provide long-term business and employment. 

People intending to work with Indigenous people should be adequately prepared. This may involve 

background research on the local Indigenous history and culture, and developing an awareness of local 

Indigenous political and governance structures. 

Consultation or other participatory processes are an essential component of any project. Inappropriate 

or inadequate consultation can cause a project to fail and result in ill feeling which jeopardises future 

projects. 

Sustainability is an essential part of any community development, achieved through a combination of 

scientific and traditional knowledge. However, pressures of the market economy need to consider 

customary laws concerning resource use. To ensure that agroforestry projects are addressing cultural 

and environmental sustainability, appropriate monitoring and evaluation mechanisms must be in place. 

These should include feedback from Traditional Owners on the impacts of the project on the natural 

environment and on cultural integrity. 

All issues associated with Indigenous Intellectual and Cultural Property Rights need to be 

acknowledged and respected prior to project implementation. To ensure appropriate and agreed use of 

Indigenous peoples’ cultural knowledge and traditions. 

THE SANDALWOOD OPPORTUNITY 

The harvesting of native sandalwood in Australia has long been an important industry, with Australia 

being one of the world’s two largest producers of native sandalwood (with India)(Statham, 1990; 

Radomiljac et al., 1999; Vernes and Robson, 2002; Bragg et al., 2004), supplying a global market 

estimated at 3000 tones per annum. Queensland and Western Australian sandalwood, is exported to 

South-East Asia, primarily Taiwan, where it is powdered and mixed with various resins and other 

aromatics to make incense sticks (Jones, 2001). In 1988 Australia supplied 94% of the global incense 

stick sandalwood market (Statham, 1990).



Of more than 30 sandalwood species globally, about seven are economically important. Of the four 

Santalum species that occur in Australia, S. lanceolatum is the most widespread, occurring across 

northern Australia (Applegate et al., 1990b; Harbaugh, 2007), but Santalum album (L), Indian 

sandalwood, and Santalum spicatum (R.Br), Western Australian sandalwood, are currently the two 

most commercially valuable species (Bragg et al., 2004). 

A recent study of natural sandalwood populations in CYP has identified several with outstanding oil 

qualities, and have high potential for domestication. 

In part because of its high-value, the exploitation of sandalwoods around the world has led to some 

uncertainty about the sustainability of the natural resources of most Santalum species. An attractive 

alternative or complement to a sandalwood industry based on Australian native forest is the 

establishment of plantations which can be managed to provide a sustained yield (Applegate et al., 

1990a; Baruah, 1999; Vernes and Robson, 2002). 

Sandalwoods have specific ecological and silvicultural requirements. They are all hemi-obligate root 

parasites, with inefficient root systems that extract a range of nutrients from the root system of their 

hosts. The oil content in the heartwood of sandalwood varies across their distribution (Baruah, 1999). 

Sandalwood products and markets 

The scented heartwood of sandalwood is a valuable commodity used historically by Indigenous people 

all over the world. In traditional medicine, sandalwood oil is used for a wide variety of conditions 

ranging from an antiseptic and astringent to the treatment of headache, stomach-ache, and urogenital 

disorders. The oil, with its distinctive and woody aroma, is a highly sought after fragrance.. It is also 

valuable for its binding qualities with other fragrances to create a characteristic bouquet for a branded 

perfume. In volume terms, its use in the perfumery industry far outweighs its medicinal usage. 

Sandalwood oil is produced commercially by steam distillation. Variation in oil content is attributable 

to tree age, site and position of heartwood within the tree. 

The heartwood is close-grained, finely and evenly textured, hard durable and renowned as a carving 

material. The timber seasons well when dried slowly. The wood can be worked to a smooth finish and 

takes on a satin-like polish. The heartwood was used for many centuries for carvings, prayer poles and 

other religious artifacts, valuable handicrafts, fuel for funeral pyres, coffins and incense sticks 

(Statham, 1990; Baruah, 1999). 

Due to the high value of sandalwood species, illegal harvesting has become a management problem in 

some countries, to the extent that in some areas natural stands of sandalwood are threatened. 

Queensland sandalwood background 

The native forest sandalwood industry commenced in Queensland in the 1890s when trees were 

harvested from the northern part of CYP, and later near Normanton and the Mitchell River in the 

earlier parts of last century (Applegate et al., 1990a; Statham, 1990; Bristow et al., 2000). Reports 

dating back as early as 1902 detail sandalwood being exported to China from Somerset, at the top of 

CYP (Statham, 1990). There is information referring to tonnes of sandalwood being cut and exported 

from CYP, adjacent to several Aboriginal missions in the late 19 th century (Wharton, 1985, 2005). 

Sandalwood cutters relied on Aboriginal labour and knowledge of country to exploit the species. The 

Aboriginal labourers located the trees, cut off the bark and sapwood and helped transport the billets of 

heartwood to the coast for export (Wharton, 1985). 

The sandalwood cutters first harvested sandalwood from the east coast of CYP, using the services of 

up to two hundred Aboriginal people, including women and children, to dig up the valuable stumps 

and roots. When that resource became scarce, the sandalwood cutters spread to other areas of CYP. In 

the Aurukun area over 100 aborigines were employed to cut sandalwood and paid in tobacco and basic 

food rations, exporting 300 tonnes up to 1920 (Pike, 1983). The sandalwood industry in Queensland 

declined in the 1930s and ceased following the outbreak of civil war in China in 1946, and the 

collapse of the sandalwood market (Kealley, 1989). Some cutting resumed from 1983 (Applegate et 

al., 1990b; Statham, 1990).



Northern sandalwood is commonly harvested by pulling the entire tree and stump from the ground, 

because a large amount of the oil (in some species most of the oil) and a higher proportion of fragrant 

heartwood are found in the root ball (Doran et al., 2005). 

The natural regeneration of sandalwood species occurs through seed dispersal and vegetatively, 

including root suckering. In Australia, the critical factors for sandalwood establishment, growth and 

production appear to be the regularity of rainfall, the length of the growing season, grazing of the 

palatable leaves by animals, fire, and the regularity of frosts (Bristow et al., 2000; Bragg et al., 2004). 

Community development opportunities in the Cape York Peninsula: A case study 

Historically, the rate of commercial development in Cape York Peninsula has been slow, with most 

development activity occurring after World War II. Extensive cattle grazing and mining, together with 

some forestry operations, have been the main enterprises. A steady improvement in infrastructure 

(roads, transport, communication, towns) has accompanied the expansion of the mining, cattle and 

tourism sectors. Commercial forestry activities have been small-scale and low intensity harvesting of 

native forest (Annandale and Crevatin, 2002). 

Commercial sandalwood plantings present an opportunity for the Indigenous communities of CYP to 

participate in sustainable business development. The unique ecology of the sandalwood, coupled with 

the high value of sandalwood products, allow for versatile silvicultural systems which in CYP could 

focus on sandalwood products and associated value-adding to create long-term business opportunities 

and alternative employment opportunities. In contrast, the plantation sandalwood industry in the Ord 

River region of WA, has adopted a highly mechanised, large-scale industrial forestry approach,. 

Specialised sandalwood plantations which target niche and high-value markets have the potential to 

meet both community expectations and high-value market opportunities. Sandalwood plantings in 

CYP could be designed to suit the ecological and socioeconomic conditions of the area through an 

adaptive management approach, including enrichment and minoforestry planting systems, utilising 

flexible labour systems that could include flexi-time, part-time, or seasonal work.. 

Historically people in remote regions, are subject to decisions made by people often far removed from 

where the decisions have an impact, which are generally influenced by broader commercial directives 

and affected by government policies. These decisions may not suit the biophysical environments of 

CYP; they don’t acknowledge the level of traditional forest knowledge resident in the communities 

and are not necessarily aligned with everyday life in this remote region. Industrial-scale forestry in 

Australia commonly relies on external investment together with mainstream economic systems and 

timetables that influence plantation management decisions. Also, plantation forestry companies 

already growing sandalwood typically require substantial financial and technological investments 

currently not available in CYP. 

Small-scale sandalwood planting systems provide an opportunity to maximise returns through 

integration of multiple goals to achieve both commercial and non-commercial benefits, tailored to suit 

other land uses, various land tenures, including freehold, community owned land and bauxite mine 

leases of the western CYP. Commercial benefits could be complemented by environmental and socioeconomic 

benefits. Such developments could be small in scale both spatially (e.g. planting, tending 

and harvesting small areas each year) and in terms of labour demands (e.g. providing jobs for small 

teams on a seasonal basis). In this way, sandalwood plantations could offer opportunities for the 

Indigenous communities and individuals to determine scale, management regimes and other factors 

required for sustainable development in the socio-cultural environment of CYP. 

Strategies for further development of the sandalwood industry in CYP 

In Western CYP around 600,000 hectares of land is covered by bauxite mine leases near the mining 

town of Weipa, and around the Indigenous communities of the Mapoon, Napranum and Aurukun. 

Mining in this region has taken place for 50 years, and it is conservatively estimated that the bauxite 

resource will last for at least another 40 years (Rio Tinto Alcan newsletter 1, 2008). Bauxite mining 

involves clearing of vegetation, removal of the upper 0.3-1 m of soil, digging up of the bauxite 

through open-cut mining and progressive mine site rehabilitation. Mine rehabilitation can be 

developed to include plantation sandalwood, complemented by other commercial plantings, species



for conservation outcomes and value-adding enterprises (Annandale 2003; Annandale and Bragg 

2004). On the completion of mining the land will be rehabilitated and eventually returned to the 

Traditional Owners. 

There is questionable value in regenerating native vegetation for marginal environmental benefit, as 

this provides no economic, cultural or recreational services to the community (Annandale, 2003) nor 

provides a long-term and sustainable business opportunity when the bauxite resource has been mined. 

A range of minoforestry options have developed (Bragg et al., 2004), including sandalwood 

establishment and management in either plantations or enrichment planting designs. Early enrichment 

planting trials in Weipa have established sandalwood seedlings in either native or mine rehabilitated 

forests to enrich the value of the forest. Presently both designs are being trialed on a small scale, but 

opportunities exist for broadacre systems. This integrated management approach, if developed to 

appropriate scale and commercialized, will increase the long-term sustainability of communities. 

Minoforestry — an integrated approach 

Minoforestry, the integration of forestry production systems into rehabilitated mine landscapes, 

provides an opportunity to include a commercial return on rehabilitation investments (miners have to 

pay for rehabilitation as part of the licence to operate), such as Indigenous business employment and a 

range of environmental benefits. There are many design options for minoforestry, some of which have 

demonstrated potential for mine rehabilitation and some commercial returns. These include: 

integration of seed of high-value species into standard rehabilitation seed mixes; enrichment planting 

into established mine rehabilitation areas; block plantings of selected commercial species; plantations 

as windbreaks and shelter for stock in pasture areas; and other inclusion of native species for food, 

fibre and medicinal purposes. In addition to economic and environmental benefits, some species may 

provide community health benefits i.e., traditional foods. 

DISCUSSION 

Consultation with Indigenous people invariably involves consultation, often involving large numbers 

of people from widely dispersed communities. The way consultation is undertaken will determine the 

success or otherwise of the process and ultimately the project itself. A large number of matters must 

be taken into consideration, some of which apply to all consultations and some that are unique to 

circumstances where the Indigenous people live. 

Sandalwood growing, harvesting, and research have a long history in north Queensland. Studies 

undertaken by Traditional Owners, James Cook University researchers and collaborators such as 

Harbaugh (2007) have provided evidence that sandalwood in CYP is able to meet international 

standards for sandalwood oil and compete with Indian sandalwood on the world market. The outcomes 

of this work provide support for further development of sandalwood plantations in the region through 

the selection of superior trees for the development of cultivars for plantation development. 

Experimental forestry plantations established over the last 40 years at Weipa have identified forest 

species suitable for mined land. Plantation on mined land can be developed to establish a viable 

forestry industry and associated value-adding business opportunities to create long-term and 

alternative business and employment opportunities for the Indigenous communities of the Western 

CYP. Specialised sandalwood plantations that target niche and high-value markets have the potential 

to meet both community expectations and high value market opportunity. 

In the CYP Indigenous community involvement in forestry is evolving. Community aspirations and 

expectations for mined land acceptable for relinquishment back to Traditional Owners include the 

need for industry diversification to support Indigenous business and employment opportunities when 

mining operations have ceased. 

Opportunities for growing sandalwood on Cape York Peninsula include plantation and forest 

management, minoforestry, or mixed agroforestry systems on community lands. Conservation 

outcomes are further supported through growing sandalwood now only found in isolated and small 

populations, including preservation of in-situ trees.




This paper draws on a variety of work with Alan Bragg and Mila Bristow when they and the author 

worked with the Queensland Forestry Research Institute in north Queensland. and on work with Sue 

Feary of the Australian National University on Indigenous community consultation, The inputs of 

Alan, Mila and Sue were instrumental in developing some of the concepts discussed in this paper and 

are gratefully acknowledged. 

REFERENCES 

Annandale, M., 2000. Management and viability of forestry plantations at Weipa: Weipa Forestry Plantation 

Assessment. In. Queensland Forestry Research Institute, Department of Primary Industries, Atherton. 

Annandale, M., 2003. Economic development opportunities for indigenous communities: provided by 

regeneration of mined lands in the Western Cape York Peninsula. In, Land relinquishment criteria 

workshop, Weipa, Queensland. 

Annandale, M., Bragg, A., 2004. Minoforestry development at Weipa. In. Department of State Development and 

Innovation, Cairns, p. 31. 

Annandale, M., McGavin, R.L., Venn, T.J., 2002. Injinoo sawmill project phase 1:small-scale timber processing. 

In. Queensland Forestry Research Institute, Department of Primary Industries, Atherton. 

Annandale, M., Taylor, D., 2000. Developing a sustainable forest management system for the forests of Cape 

York Peninsula. In. Queensland Forestry Research Institute, Department of Primary Industries, Atherton. 

Applegate, G.B., Chamberlain, J., Daruhi, G., Feigelson, J.L., Hamilton, L., McKinnell, F.H., Neil, P.E., Rai, 

S.N., Rodehn, B., Statham, P.C., Stemmermann, L., 1990a. Sandalwood in the Pacific: A State-of- 

Knowledge Synthesis and Summary form the April 1990 Symposium. In: Hamilton, L., Conrad, C.E. 

(Eds.), Symposium on Sandalwood in the Pacific. USDA Forest Service, Honolulu, Hawaii, pp. 1-11. 

Baruah, A.D., 1999. The economics of sandal-oil. Indian Forester 125, 640-643. 

Bootle, K.R., 1983. Wood in Australia: types, properties and uses. McGraw-Hill Book Company, Sydney. 

Bragg, A., Annandale, M., Wapau, J., 2004. Minoforestry - Past, present and future commercial tree plantings at 

Weipa. In: Bevege, D.I., Bristow, M., Nikles, D.G., Skelton, D. (Eds.), Prospect for high-value hardwood 

timber plantations in the 'dry' tropics of northern Australia. Published as a CD ROM by Private Forestry 

North Queensland, Kairi, Queensland. 

Brand, J.E., 1993. Preliminary observations on ecotypic variation in Santalum spicatum 2. Genotypic variation. 

Mulga Research Centre Journal, Curtin University 11, 13-19. 

Brand, J.E., 1999. Conserving sandalwood (Santalum spicatum) in the rangelands, Western Australia. In. 

Department of Conservation and Land Management, Western Australia, pp. 1-4. 

Bulai, P., Nataniel, V., 2002. Research, development and extension of sandalwood in Fiji: A new beginning. In, 

Regional Workshop on Sandalwood Research, Development and Extension in the Pacific Islands and Asia. 

In Press, Noumea, New Caledonia. 

Bule, L., Daruhi, G., 1990. Status of Sandalwood Resources in Vanuatu. In: Hamilton, L., Conrad, C.E. (Eds.), 

Symposium on Sandalwood in the Pacific. USDA Forest Service, Honolulu, Hawaii, pp. 79-84. 

DPI Forestry, 1995. DPI Forestry yearbook 1994-95. In, Brisbane.Fox, J.E.D., Harbaugh, D.T., 2007. A 

taxonomic revision of Australian northern sandalwood (Santalum lanceolatum, Santalaceae). Australian 

Systematic Botany 20, 409-416. Harbaugh, D.T., Baldwin, B.G., 2007. Phylogeny and biogeography of the 

sandalwoods (Santalum, Santalaceae): Repeated dispersals throughout the Pacific. American Journal of 

Botany 94, 1028-1040. 

Hopewell, G., 2001. Characteristics, utilisation and potential markets for Cape York Peninsula timbers. In. 

Forest Products program, Queensland Forestry Research Institute, Department of Primary Industries, 

Brisbane. 

Kealley, I.G., 1991. The management of sandalwood. In. Department of Conservation and Land Management, 

Perth. 

Pike, G., 1983. The Last Frontier. Pinevale Publications, Mareeba. 

Radomiljac, A.M., 1998. The influence of pot host species, seedling age and supplementary nursery nutrition on 

Santalum album Linn. (Indian sandalwood) plantation establishment within the Ord River irrigation Area, 

Western Australia. Forest Ecology and Management 102, 193-201. 

Rio Tinto Alcan newsletter 1, 2008. South of the Embley Project. Rio Tinto Alcan Brisbane Australia, June 

2008).



Statham, P.C., 1990. The sandalwood industry in Australia: a history. In: Hamilton, L., Conrad, C.E. (Eds.), 

Symposium on Sandalwood in the Pacific. USDA Forest Service, Honolulu, Hawaii, pp. 26-38.Venn, T.J., 

2004. Visions and realities for a Wik forestry industry on Cape York Peninsula, Australia. Small-scale 

Forest Economics, Management and Policy 3, 431-451. 

Vernes, T., Robson, K., 2002. Indian sandalwood industry in Australia. In, Sandalwood Research Newsletter, pp. 

1-4. 

Wharton, G., 1985. Antiquarians and sandalwood-getters: the establishment of the Cape York collection at 

Weipa. In, North Australian Mine Rehabilitation Workshop No. 9, Weipa, Queensland. 

Wharton, G., 2005. Northern Sandalwood (Santalum lanceolatum) on Cape York Peninsula. A report on 

historical sources for the Queensland Department of State Development and Innovation. Prepared by Geoff 

Wharton, Infotracker Historical and Information Research Services October 2005. Brisbane. 

Wilson, G. 2005. Traditional Owner Kanju elder Personal Communication.



A FORESTER’S GUIDE TO WORKING WITH 

INDIGENOUS AUSTRALIANS: 

AN INTRODUCTION 

Peter Shepherd 1 

ABSTRACT 

Indigenous issues have gained in importance in recent years. Forest managers are now 

required to work more closely with traditional land owners with the risk of possible 

litigation, long delays and high legal expenses if proposals are not supported by the 

local Indigenous community. Foresters will benefit from the broad introduction to 

Indigenous culture presented in this paper. Topics discussed include approach to land 

ownership, society structure, indigenous language groups and significance of flora to 

Indigenous people. The risk of treating the Indigenous community as a homogeneous 

society is recognised. Hence this paper is presented as a general guide and introduction 

to alert foresters about some topics they may need to consider when managing forests 

on land that is significant to Indigenous communities. 

INTRODUCTION 

Many foresters must work with Indigenous people who have an interest in the land; be they owners, 

traditional owners, neighbours or interested people. Over the last few decades, the need to work with 

Indigenous Australians has increased as their stake holding in land has increased. 

As Indigenous Australians are now expecting greater influence over decisions for land in which they 

are stakeholders, foresters need to be able to form positive relationships with Indigenous individuals 

and communities. This can only happen with respect, understanding and sensitivity. This paper will 

introduce some key information about Indigenous culture at a very basic level in order to help increase 

the general understanding of how to work constructively with Indigenous people. 

However, I am writing this paper with some trepidation since it is wrong to suggest Indigenous 

Australians are a homogenous population. Hence generalisations are dangerous. Further, I am not 

claiming to be an expert in Aboriginal communication and culture. This paper is presented as a very 

general and basic guide to inform and help those who are unfamiliar with working with Indigenous 

people to avoid mistakes and misunderstandings. 

I apologise in advance to any Indigenous person who feels I have made sweeping generalisations that 

do not apply to their culture. 

INDIGENOUS NATION STRUCTURE 

The Indigenous Nation is in effect a conglomerate of smaller cultural groups, like most countries. 

Language, culture and geography define these groups. It was assumed originally that there were over 

700 language groups, of which 250 have been recorded and over 200 different languages have been 

mapped (see http://www.abc.net.au/indigenous/map/ for an interactive map and Diagram 1). Extensive 

and strong relationships between language groups exist with respect for each “country” where the 

traditional owners are based. Respect is shown by acknowledging the owners of the land, the Elders 

and their ancestors, early in an interaction, event or entry to the territory. 

Each group has distinct cultural aspects which distinguish it from all others. 

1 Dr. Peter Shepherd, Manager, Business Development, Batchelor Institute of Indigenous Tertiary Education, 

ph 0419309281 (m) (08) 89397273; email: peter.shepherd@batchelor.edu.au



Diagram 1: Map of Indigenous Language Groups 

(source: Tindale – revised 1974) 

LAND OWNERSHIP 

Indigenous people do not view land ownership in the same way that western cultures do. The concepts 

of freehold and land title are alien. In remote communities in north and central Australia, housing is 

provided by government (much of it well below standard). Many Indigenous people in their traditional 

setting do not understand how any person can own land or hold “title” to a house; hence there is no 

concept of mortgage over land or property. Urban Indigenous people in larger towns and cities do 

have a better understanding of home “ownership” but in many cases live in government-provided 

housing. 

Indigenous people have a very different relationship with land. Many believe they are culturally 

responsible to their “country”, which is part of their Dreaming, and they must act as custodians for that 

land. This is a spiritual obligation to their forefathers and traditional leaders. 

KINSHIP AND FAMILY 

Indigenous family and kinship relations are complex and difficult for untutored westerners to 

understand. I will start with family and work to the community. Sometimes a few related new mothers 

combine to raise a group of youngsters. Hence it is possible for an indigenous person to claim 2, 3 or 

more “mothers” all treated with similar respect.



An Aboriginal person will belong to an inherited skin group which is descended from a particular 

ancestral creator spirit. The skin group can be inherited from a person’s mother or father depending on 

the clan. There are very strict rules based on skin groups. The skin group has a strong ruling on a 

person’s selection of a marriage partner, land they will inherit and access to resources. 

Marriage between people of the same skin group is strictly forbidden. The women elders of a clan 

carefully considered all marriage options and arranged marriages to ensure correct skin group lineage. 

Each skin group has at least one opposing group with whom they cannot make contact,; especially 

people of the opposite sex. This includes every day contact such as talking, eye contact and even 

passing items to a forbidden person. It is understood these rules are in place to avoid relationships that 

are genetically close, for example, to forbid first, second or third cousins marrying. 

Because of the complex spiritual beliefs and the intertwining of clans through marriage, the estate 

boundaries would not always have been clearly definable as boundary lines on a map (such as 

Diagram 1), but none-the-less are clearly perceived and understood by individuals knowing 

geographic markers on their traditional land and passing this on to younger clan members as part of 

Indigenous knowledge.. Through intermarriage and other alliances people were able to access land and 

resources far beyond their own estates. Access to land and resources were negotiated through 

discussion, marriage, ceremony and adherence to Aboriginal lore. 

Clans were united in alliances based on a shared language, spirituality, seasonal hunting and gathering 

events and lore. These alliances are called language-culture groups. 

It needs to be remembered that a young indigenous child will be exposed to a number of languages. 

The child’s mother will speak in many cases a few of the common dialects in the area. The child’s 

father will speak the same languages and several others needed for cultural interaction purposes. Some 

speak 7 languages or more to a high level of proficiency. English may well be the 4 th (or more) 

language spoken in their house, only at school, and rarely in the community. 

DREAMING AND TIME CONCEPT 

Western people often equate dream time to the creation, or some time in the past. This is true in one 

sense, but a mistake in that it is not limited to the past. Dreaming is also today. Dreaming is infinite 

and, for Aboriginal people, links the past and the present to the future. Dreaming can explain how 

landscapes and environments were shaped. But an Indigenous person will also have a “Dreaming”; a 

special inheritance and link to the land. This may be represented in his/her totem. Dreaming stories are 

the truth from the past which becomes the code of law for today. Often the story is only a little part of 

a much greater story which provides insight into behaviour and future being. Some dreaming stories 

are very sacred when told in their country and can only be told in special places under specific 

circumstances. 

In most stories of the Dreaming, the Ancestor Spirits came to the earth in human form and as they 

moved through the land, they created the animals, plants, rocks and other forms of the land. They also 

created the relationships between groups and individuals to the land, the animals and other people. 

Once the ancestor spirits had made the world, they did not disappear but changed into trees, the stars, 

rocks, watering holes or other objects. These are the sacred places and organisms of Aboriginal culture 

and have special properties. Because the ancestors did not disappear at the end of the Dreaming, but 

remained in these sacred sites, the Dreaming is never-ending, linking the past and the present, the 

people and the land. Hence an Indigenous person has his/her “Dreaming” which is their relationship 

to the land and ancestors and a special creation story often through a plant species, landscape 

formation or animal. 

I have a very special boomerang made by an aboriginal elder who made a special effort to put his 

Dreaming on to it. It is a special item of great significance to me. It only takes one look to feel and 

know the difference between this and mass produced tourist items.



The protocols for social behavior and consequences, including punishments and disciplines learnt, are 

evident in Dreaming stories. Virtue in Aboriginal religion lies in the obligation to follow ancestral 

precedent, which involves keeping the Dreaming stories alive. This takes the forms of painting, song, 

dancing or ceremony - all of which are therefore necessarily inextricably linked. This is part of a living 

tradition based on ritual practices. Traditions and practices also merge with economic and ecological 

responsibilities for 'looking after country'. Looking after country means to continue to express these 

ritual forms of the Dreaming. Clan groups have the right to use the land regarded as their 'territory, 

estate or country, and any of its products, based on their duties to tend the land through the 

performance of ceremonies. 

Smith (2005, pp.4-5) outlines a mistake made by some Park Managers who were explaining the 

origins of fruit bats in North Queensland to an Aboriginal community when they suggested they had 

migrated from a colony further away. The local Indigenous people were amused because local 

knowledge pertaining to flying foxes was that they came from the mouth of the Rainbow Serpent, a 

powerful creator that was still present and active in the region’s landscape. Smith’s account is valuable 

because it discusses how difficult it was for park management professionals to accommodate 

Indigenous knowledge into their thinking without being patronising. The result is a reluctance for 

Aboriginals to share and demonstrate a public dissent with non-aboriginal perspectives. Smith 

continues to outline how constructive relationships with local Aboriginals around land management 

can be created and is recommended reading for those considering working in this area. 

It only takes a moment to ponder the vast differences between the spiritual and ancestral values and 

the Western system of valuing land and forest products. 

TOTEMS 

A totem is an object or thing in nature that is adopted as a family or clan emblem. Different clans are 

assigned different totems and in some cases individuals are given personal totems at birth. In some 

areas personal pendants are worn and these pendants are mostly carved out of wood, turtle shell or 

shells and often represent the person's totem. There are well established rules as to when the pendants 

can be worn, often only allowed during ceremonies or rituals. 

Some Aboriginal people can be identified by their totems, which can be birds, reptiles (like turtles), 

plants, trees, sharks, crocodiles and fish. They are an important part of their cultural identity and are 

especially significant in song, dance and music as names and on cultural implements. Some clans 

forbid their individuals from eating the animal that is their totem, while other tribes make exceptions 

for special occasions such as ceremonies. People can collect seeds, fruits and other forest products 

from trees and plants as they have been provided by the Ancestor to the Clan. But the special Totem 

Trees cannot be harvested as they are the Dreaming of ancestors.



SORRY TIME AND FESTIVALS 

“Sorry Time” is the mourning period after a person has died. This can last up to a month and it is 

mandatory for clan members to be present. Employees will not be able to attend work for this period. 

Further, few meetings will be attended or decisions made. People will travel huge distances to attend a 

funeral. The dead person’s name cannot be mentioned in the community. 

There is usually an annual festival in the clan’s social calendar. This is generally to celebrate the end 

of a period of difficult weather, so it is usually held at the end of the Wet Season in the north or in 

spring in the south. The celebrations can last for days or weeks. Most Indigenous people will want to 

attend part or all of their community’s festival to meet clan and family obligations. 

LEARNING SYSTEMS 

As with many Pacific Island cultures, Australian Aborigines are a “talking” culture. The need to take 

time to talk with visitors, to get to know them and establish a relationship before they will have 

meaningful talks is paramount. 

Indigenous learning is obtained from narrative leading to understanding and insight. Learning and 

understanding come from stories, songs and skills acquired in the field under the supervision of a more 

senior clan member. Learning only takes place within the context of a narrative, preferably by doing or 

observing a related fact or skill from the real world, Dreaming and Elder. Smith (2005) gives a good 

insight into how Indigenous knowledge is passed on. 

I cringe when western thinking managers and even teachers believe that they can further their 

objectives by preparing a PowerPoint based on our western experience and learning paradigm. Classic 

PowerPoint presentations are sure to quickly loose the interest of an Indigenous audience, fail to 

achieve the outcomes and do a disservice to the reputation of the person who delivers it in the eyes of 

the clan. 

LEADERSHIP 

Finally a word about leadership. Defining leadership is a vexed task in both the business literature and 

reality of professional life. I will leave discussions about definitions of leadership to others. The focus 

here is on the different approaches to leadership between Western and Indigenous cultures so that 

foresters can be alert to some traps that await them when working with Aboriginal people. 

Firstly, an Elder is an old and respected member of a clan. This position has been obtained by the fact 

of being a senior member. Their opinions are important. A leader is usually an Elder but does not 

necessarily have to be. Usually a Leader has a significant leadership role requested by the senior 

members of the clan and has a significant relationship with the clan such as being a traditional owner. 

Indigenous leadership is not about having a high profile and being seen at the front. Working with an 

Elders’ Council may be seen as a slow and cumbersome process lacking in focus by a person 

expecting a western decision making experience. 

The leader may not be the person greeting you or chairing the meeting. The most influential people 

may be the Elder women sitting under a tree at the back of the meeting listening quietly and observing 

you carefully. 

Some country is Women’s Country. If this is the case, women will have a different position in 

discussions than usual. It is an advantage to have a female as part of the delegation who can meet with 

the women. However, consult the Elders first to ensure you are not making a mistake. 

Indigenous people are shy, charming, witty, charismatic and delightful to work with but they take time 

to develop relationships and respect for an outsider. Attempt a quick, in/out meeting with great caution 

as it would be better to spend more time, looking, listening while becoming familiar with the people 

and context.



Smith (2005, pp: 9-11) describes the work of David Claudie who compared Indigenous thinking (he 

called it “round thinking”) to western thinking (“square thinking”). Claudie describes bureaucrats who 

can’t listen and perceive themselves as “flash” and better than Aboriginal people. This may not be the 

case but is how they are perceived. A number of causes of this misunderstanding of people are offered 

including a misalignment between Western and Indigenous people’s “way of knowing”. 

Noel Pearson (2009, pp: 251-253) speaks of a leadership continuum made up with realism and 

idealism on opposite ends of a two dimensional plane. Indigenous leadership models are at the 

idealism end while Western leadership is more pragmatic. Pearson suggests those who harbour ideals 

need to work with those who have real power to form a compromise. 

“It takes insight, skill and creativity, careful calculation as well as bold judgement, prudence 

as well as risk, perseverance as well as preparedness to alter course, belief as well as humility 

and great competence as well as ability to make good from mistakes to bring ideals closer to 

reality” (Pearson 2009, p. 252). 

Clearly the instruction to foresters here is to prepare well, gain insight, think carefully, listen and be 

flexible when dealing with Indigenous leaders and communities. 

CONCLUSION 

Working with other cultures is not rocket science. To be successful just takes a little time and thought. 

The web has vast resources available. A search for “working with Aboriginal Communities” will 

result in a vast collection of resource and guide books as well as bibliographies. 

It will take much more than this brief overview of some parts of Indigenous culture to ensure 

successful interactions for people who are not familiar with working with other cultures. There are a 

couple of key behaviours that need to be adopted that will help. Firstly listen and do more listening 

than talking. It is so important to listen to individuals, communities and read the press. If possible, 

attend community visits with someone who is experienced, known and respected by the people. 

Observe carefully how this person behaves. Develop expectations that are realistic for the Indigenous 

decision making model and not a western expectation. And finally think about what you have read and 

observed to create a deep behaviour model. Understanding will come with work and effort. 

Failure to do the work will result in problems. 


Thank you to Patrick Anderson from Batchelor Institute for proof reading this paper and providing to 

me many deep and insightful explanations of the mysteries of Indigenous culture. 

REFERENCES 

Pearson, N. (2009). “Up from the Mission”. 400 pp. Black Inc. 

Smith, B. R. (2005). “We got our own management”: local knowledge, government and development in Cape 

York Peninsula. Australian Aboriginal Studies 2005/2 pp:4-15.



AN INTEGRATED APPROACH TO SPECIES REGIME SELECTION 

IN PNG: BEGINNING WITH LOCAL PEOPLE’S NEEDS 

ABSTRACT 

Braden Jenkin 1 , Michael Blyth 2 , Peter Kanowski 3 and Don Yakuma 4 

Selection and development of a tree species regime to deliver benefits to small-scale treegrowers 

must be driven by the needs of the individuals, families and local communities 

involved. A species regime must consider all elements of the production system from the 

tree species requirements through each stage of the supply chain to the point of sale. As 

part of an Australian Centre for International Agricultural Research project (ACIAR 

FST/2004/050), a systematic approach to planted woodlot and individual tree species 

regime choice is being developed to assist local communities in three pilot locations 

(Madang area, Madang Province; Ramu Valley, Morobe Province; the Fly River corridor, 

Western Province) in Papua New Guinea. A parallel component of the project seeks to 

better understand the decision making process of why people would need or wish to plant 

trees. 

The key variables of a species regime selection framework are the physical and financial 

input requirements, the timing of returns, and inherent and exogenous risk and uncertainty 

factors. Individuals’ preferences for the timing of returns may be linked to significant 

household events. In some cases, a species regime may provide intergenerational returns 

with trees treated as a bank account. Other regimes may provide continuous cash flow 

after a startup period, with a final lump sum (e.g. rubber trees). A species regime that 

provides additional non-financial benefits may be of interest (e.g. nut production as part of 

a food security). 

The ACIAR project has developed protocols to consider market and processing needs, to 

assist focus on least-risk options that are feasible under current conditions. While other 

options may become attractive in the future, a risk minimisation approach suggest such 

options should be regarded as the potential icing on a likely cake. 

INTRODUCTION 

There is an increasing need in Papua New Guinea (PNG) to investigate, develop and implement tree 

growing regimes which assist rural communities to generate income from planted trees and forests, in 

part to relieve pressures on natural forests and in part to enhance landowners’ livelihoods. Everywhere 

in PNG, tree growing and management of trees are incorporated into both traditional and modern 

farming systems. However, because there has been little incentive to focus on commercially-valuable 

forestry species, such species have seldom been adopted. Where a critical mass of resource can be 

established, commercial tree-growing appears a good prospect for landowners with limited incomegeneration 

alternatives (ACIAR, 2009). Smallholders’ (in the PNG context, usually described as 

‘landowners’) resource allocation decisions to bridge these two issues require support and guidance in 

a number of critical areas, in particular species choice. 

Kanowski et al., (2007) conducted a pilot assessment (the Scoping Study for ACIAR Project 

C2005/050) of the potential to value add to current PNG agroforestry systems and facilitate 

smallholder uptake of high-value tree species. This project identified suitable candidate regions and 

1 

Managing Director, Sylva Systems Pty Ltd (contact point), PO Box 1175, Warragul, Vic, 3820 braden@latrobe.net.au. 

2 

Director, Four Scenes Pty Ltd., PO Box 50, Kippax, ACT 2615. 

3 

Professor, The Fenner School of Environment & Society ANU Forestry Building 48, Linnaeus Way, The Australian 

National University, Canberra ACT 0200, Australia. 

4 

Forestry Program Coordinator, Ok Tedi Development Foundation, Ok Tedi Mining Limited, P.O. Box 1, Tabubil, WP, 

Papua New Guinea.



partners, together with preliminary tree species and production systems (ACIAR, 2009), and resulted 

in a subsequent project (ACIAR project FST/2004/050). A multidisciplinary team was formed to 

research the adoption of commercial-scale high-value tree growing in PNG, developed through a 

relationship fostered between landowners and selected business partners (ACIAR, 2009), and to apply 

the results to benefit smallholders (see Kanowski et al., 2008). 

A systematic approach to tree species selection has evolved to assist local communities in three pilot 

locations (Madang area, Madang Province; Ramu Valley, Morobe Province; the Fly River corridor, 

Western Province). A parallel component of the project is to develop an understanding of smallholder 

motivations and decision making and is the subject of ACIAR-sponsored PhD research (see Mulung, 

2007, p.26). 

This paper describes the development of a decision-making framework to assist individual 

landholders, communities and their advisors through the critical step towards successful tree species 

selection for commercial production. 

KEY CONCEPTS 

Smallholder needs: An appropriate species regime must satisfy the needs of those growing the trees, 

and hence this project includes a foundation in social research. Chambers (2007, p.38-41) considered 

that the dimensions often omitted in social research are: 

• Tropical seasonality: disease, ground hardness, lack of food and the impact on ability to 

work; 

• Places of the poor: people can be isolated and lack access to infrastructure; 

• Poverty of time and energy: an imbalance between the desire to work and lack of work 

options; 

• The body: often the only resource available, which may be weakened by poor health and 

nutrition. 

The analysis will focus on the needs of the local people while considering these points of caution. 

High-value: High-value is a philosophical adjective, unless qualified by a statement of who is the 

beneficiary. In some cases, what is high-value to one party may be at the expense of another party; or 

there may be little connection between value and return to growers: e.g. does an increase in high-value 

exports contributing to gross domestic product necessarily result in higher profits for producers? 

Non-standard economics: Non-standard economics looks beyond simply financial costs and returns. 

Economic analyses relevant to PNG landowners’ decisions should include measures such as utility and 

returns on labour (both of which are standard economic parameters) to fully assess livelihood assets 

deployment options and outcomes. 

SMALLHOLDER AGROFORESTRY 

‘In most countries in Southeast Asia, government driven reforestation (under both traditional and socalled 

community-based reforestation approaches) have had limited success’ (Bertomeu et al., 2008, 

p.27). By contrast, agroforestry systems implemented by smallholder farmers can re-establish tree 

cover, provide environmental services and enhance local livelihoods (Bertomeu et al., 2008, p.27). 

Agroforestry is a dynamic, ecologically based natural resource management system that, through the 

integration of trees on farms and in the agricultural landscape, diversifies and sustains production for 

increased social, economic, and environmental benefits for land use at all levels (Leakey, 1998). 

Smallholder agroforestry systems typically follow less intensive management regimes than industrialscale 

plantations. After a site is prepared and trees are planted, ongoing management of the tree crop 

such as application of fertilizer, thinning and pruning trees and weeding during the establishment 

phase are less likely to occur on a strict schedule than in an industrial plantation. 

However, where silvicultural activities directly benefit the agricultural crops inter-planted with the 

trees, they are practised more commonly. While small-scale agroforestry systems can achieve



profitability levels comparable to large-scale plantations, such performance levels may negatively 

affect the productivity of associated agricultural crops (see Tomich et al., 2001; Holding and 

Roshetko, 2003). Within an agroforestry system, trade-offs between the productivity of timber 

production and the productivity of the agricultural food crops have to be made, to varying degrees. 

Food production or food security is, necessarily, often the priority of smallholders. 

Given the situation of smallholders, their decisions to invest in tree planting cannot be expected to 

parallel those of large-scale industrial plantation investors. As Byron (2001, p.212) suggested ‘there 

may be fundamental differences in objectives, which inevitably lead to the choice of quite different 

processes and techniques’. He explained that where an industrial plantation operator may aim to 

maximise yield per hectare, a smallholder may choose to maximise the cash return per unit of labour. 

Understanding the motivation or purpose of smallholders to integrate trees into their agricultural 

production system is a critical first step in determining successful small-scale agroforestry investment 

strategies. 

Successful decision-making for agroforestry depends firstly on the purpose for which a landholder 

intends to plant trees. If the purpose is achievable, then success is possible. Smallholders plant trees 

for a variety of purposes, including: timber for various end uses, fuelwood, fruits, fodder, medicines, 

resins, shade for livestock, wind protection and for conservation of soil and water resources (Holding 

and Roshetko, 2003). Timber may be a secondary product, harvested after trees have served their 

primary purpose. In some cases, farmers plant trees as a form of savings for future family needs such 

as education expenses or as an intergenerational investment to help children and grandchildren. Where 

the objective is to produce roundwood for a third party on a commercial basis, other issues come into 

play, and the process of species selection becomes a marketing issue, rather than being based purely 

on utility. 

PNG SMALLHOLDERS 

In PNG, trees have been and continue to be grown and managed as an integral part of customary 

farming and land use systems. The diversity and dynamism of these systems reflect high levels of 

innovation and adaptation in agriculture by land users (Kanowski et al., 2007, p.4). Many land users 

have experiences growing coffee, cocoa, coconut, oil palm and a range of fruit trees. In some districts, 

such as in the Western Province smallholders have many years experience growing rubber. There are 

four broad customary agroforestry systems described, based primarily on differences in geographic 

factors (Kanowski et al., 2007, p.8&9): 

• Highlands agroforestry systems – characterised by sweet potato and casuarina (Casuarina 

oligodon); 

• Lowlands to mid-montane agroforestry systems – characterised by combinations of annual and 

perennial food plants and trees including breadfruit (Artocarpus altilis), pandanus, betel nut 

palm, sago palm and bananas; 

• Coastal agroforestry systems – characterised by fruit and nut tree species including breadfruit 

and Canarium species in combination with cassava, sweet potato and bananas; 

• Islands agroforestry systems – characterised by fruit and nut tree species including Canarium, 

Barringtonia, breadfruit and coconut, in association with numerous species of annual and 

perennial food crops. 

DECISION MAKING 

The time lag between planting a tree and yielding a benefit must be considered in the planning: what 

by when? This involves a process beginning with the end in mind whether physical, financial or other 

outcomes. Formal analytical approaches are available: e.g. Mendoza et al., (1986) presented an 

application of multiple objective programming as applied to agroforestry as a mechanism to consider 

multiple attributes inherent is such systems. 

Analysis should include an assessment of associated risk and uncertainty. Risk is defined as ‘a state in 

which the number of possible future events exceeds the number of events that will actually occur, and 

some measure of probability can be attached to them’ (Bannock et al., 1992, p. 374). Uncertainty is 

defined as ‘the state in which the number of possible outcomes exceeds the number outcomes and



when no probabilities can be attached to each possible outcome’ (Bannock et al., 1992, p. 431). 

Bigsby (2001) presents a series of papers considering plantation investment risk, primarily in terms of 

impacts on financial outcomes. Kirby et al., (1993) presented an appraisal of agroforestry investment 

under uncertainty by applying a range of values to cash flow parameters resulting in net present values 

(NPV) at set discount rates. This level of sophistication is appropriate if the parameters and the likely 

range is understood. For some individuals and households in PNG, levels of risk and uncertainty may 

lead to catastrophic outcomes, if food security strategies are compromised or people cannot buy high 

protein / high fat foods deficient in local diets (Bourke and Allen, 2008). 

A DECISION MAKING FRAMEWORK 

Development of an understanding of smallholder motivations and decision making related to land use 

and tree growing is a primary objective of the ACIAR project (ACIAR FST/2004/050; ACIAR, 2009) 

under which the work reported in this paper is being undertaken. A species regime must fit within an 

individual’s or household’s livelihood framework and with the strategies developed to target 

individuals and household goals and aspirations. ‘A livelihood comprises the capabilities, assets 

(stores, resources, claims and access) and activities required for a means of living: a livelihood is 

sustainable which can cope with and recover from stresses and shocks, maintain or enhance its 

capabilities and assets, and provide sustainable livelihood opportunities for the next generation; and 

which contributes net benefits to other livelihoods at the local and global levels and in the short and 

long term’ (Chamber and Conway, 1991, p.6). A long-standing understanding is that Melanesians 

living in resource rich landscapes are often in a state of ‘subsistence affluence’. Hence when offered 

commercial cropping opportunities to increase individual income and to boost national development, 

uptake is minimal or only to a level to meet their vital needs (Kennedy and Clarke, 2004, p.31). If a 

minimalist approach is taken, it is important to understand how such targets evolve with individuals’ 

or families’ life stages, as they influence cash flow needs. 

A range of guides which assess livelihood strategies may assist in species regime selection (e.g. 

Messer and Townsley, 2003). Ashley and Carney (1999, p.47) presented an often used sustainable 

livelihood framework (SLF). It is a dynamic tool that links the deployment of social group livelihood 

assets to invest in strategies to achieve livelihood outcomes. Livelihood assets are classified as human, 

natural, financial, social and physical capital (see Messer and Townsley, 2003, p.8&9). The SLF 

defines the resources available and the formation of strategies to deploy such scarce resources. 

Although dynamic, the framework lacks prioritisation of scarce resource allocation and target 

outcomes. Individuals’ and families’ livelihood priorities are unlikely to be linear, and could be 

described as a series of buckets. The first priority is to fill bucket A to 25% capacity to reach a point 

considered to be sufficient, after which resources can be directed to buckets B and C, filling each up to 

a pre-determined level. Once this has been achieved, resources can be directed back to bucket A and it 

is filled to say 50%. The process will continue until the available resources are exploited. The order of 

priority and levels of allocation will be an individual decision. 

In PNG, 87% of the population live outside of urban areas. About one million people live in severe 

poverty (Bourke and Allen, 2008), which takes a variety of forms in different locations (Hanson et al., 

2001). For many, capital assets are limited. The average life expectancy for men in rural PNG is 53 

years (Bourke and Allen, 2008). Therefore, tree species regimes should complement the need for food 

security within the allocation of those livelihood assets necessary to achieve priority outcomes over 

one or more generations. 

A FRAMEWORK TO GUIDE DECISION MAKING FOR SMALLHOLDER 

AGROFORESTRY 

Several elements need to align or be aligned before successful agroforestry can occur within a given 

area. For example, while a particular tree species may grow successfully on a smallholder’s land, 

sustained commercial viability within an agroforestry system is complex, involving many factors. 

Before planting it is necessary to investigate and analyse market prospects for the products from 

particular species and assess the economic, institutional and technical performance of the supply chain 

including harvesting, transport, primary and secondary processing. The main elements of the 

agroforestry decision support framework for smallholders, rural communities and their advisors in



PNG are perceptions, attitudes and motivation of smallholders, institutional and legal factors, financial 

or economic factors, social and community factors as well as physical and biological factors. 

These have been discussed by Kanowski et al., (2007, p.24&25) and others (e.g. see: Mercer, 2004; 

Scherr, 2004; Pannell et al., 2006). In addition to these factors uncertainty has been identified as a 

particularly important factor in adoption of farming practices such as agroforestry which are 

associated with land conservation. Consideration of uncertainty is critical in agroforestry investment 

assessments. An example of institutional factors which may constrain commercial tree growing is the 

current export ban on roundwood of teak (Tectona grandis) (Teak); balsa (Ochroma pyramidale); 

Blackbean (Castanospermum australe); Cordia (Cordia dicotoma); Black ebony (Diospyros ferrea); 

rosewood (Pterocarpus indicus), and all coniferous species logs (Customs Tariff Act 1990: Schedule 2, 

Export Item 44.03). 

If, after initial assessment, it appears that each of the framework elements is satisfactory, or potentially 

so, then a prospective grower can move to the next step on their agroforestry investment pathway. It is 

likely that some factors will provide more substantial constraints than others; these are the areas that 

require attention before an agroforestry investment is likely to be successful. Some of these may be 

within the control or influence of an individual or community investor, while others may not. For 

example, improving inadequate or unreliable infrastructure is beyond the financial capacity of an 

individual or a community. Prospective growers will make judgments based on the best information 

available to them, reflecting their risk aversion and the quality of their analysis. 

Questions that persons advising smallholders should consider when deciding what tree species to 

recommend include (based on Holding and Roshetko, 2003): 

• Is there a domestic and/or export market demand for resources and products from planted 

forests? What product characteristics do markets or buyers require (e.g. for timber the market 

may specify characteristics such as form, length, volume and/or specific wood properties such 

as density; certification may be a requirement for some markets)? 

• Can harvested forest resources that meet market requirements be transported to processors and 

traders from the proposed production site efficiently and reliably? Can processing be 

conducted effectively – i.e. able to consistently meet market requirements? 

• Will sufficient volume of resources be produced in the area to attract and sustain 

establishment of processing facilities or are there existing processing facilities within 

economic reach of the growing area that can accommodate increased throughput? 

• Can quality germplasm of the right species or mix of species be sourced for the proposed site 

(where species are selected in accordance with growth potential for the site and the potential 

to deliver wood properties required by markets)? 

• Will the selected mix of species allow growers to balance market requirements, household 

needs, agricultural crop productivity and sustainability of the tree resource? 

• Do growers have access to technologies and practices to ensure optimal site productivity and 

resource quality? 

FACTORS INFLUENCING THE DECISIONS OF PNG LANDOWNERS 

Factors influencing adoption of agroforestry by smallholders can be grouped under the following 

headings. 

Growing trees within garden systems: Application of the SLF outlined above highlights the broader 

issue of the impact of tree growing on livelihood outcomes. As the term agroforestry implies, forestry 

is one component of a smallholder’s farming activities. This must be taken into account when 

measuring the benefits of planting trees. Assessing the returns from the trees is necessary but not 

sufficient to determine the benefits of an agroforestry system. The net returns from the total 

agroforestry operation must be assessed and compared with other land uses to establish whether or not 

agroforestry is a better land use option. Assessment is further complicated where the decision to plant 

trees is not a purely economic or financial decision but includes social (family and community) and 

environmental considerations, as is often the case.



Financial focus: While an acceptable net present value (NPV) or internal rate of return (IRR) 

indicates an attractive investment option, a smallholder’s decision may consider other factors, such as 

the length of time before a positive cash flow occurs or the need to achieve a target income level by a 

particular year which aligns with anticipated education or other significant family expenses. Studies of 

the motivations for tree planting by small-scale forest growers in Australia found that financial returns 

per se were a minor factor (Venn, 2005). Non-commercial benefits such as shade, shelter, land 

rehabilitation, water table management, wildlife habitat conservation and stock feed were more 

important factors. The influence of social and environmental factors on landholders’ decisions to adopt 

farm forestry suggests that it is a more than a purely financial decision for many landowners (Schirmer 

et al., 2000; Schirmer, 2004). For smallholder agroforestry, conventional performance measures of 

NPV, IRR, land expectation value (LEV), annual equivalent value (AEV) and return to labour will 

need to be supplemented with specific measures that reflect the particular needs and preferences of 

smallholder investors. This involves understanding the processes that smallholders apply when making 

land use decisions. 

Attitudes to risk and uncertainty: A challenge for allocation of livelihood assets to forestry is the 

length of time before returns can be realised. Rotation lengths can range from four to six years for fastgrown 

pulpwood species (e.g. mangium: Acacia mangium), to 20 years for exotic sawnwood species 

(e.g. teak: Tectona grandis), and substantially longer for native hardwood species (e.g. kwila: Instia 

bijuga). Growers’ advisors and growers have to make judgments on how institutional, financial, social 

and physical conditions may change from establishment to harvest. If changes are likely, growers must 

consider how they might affect the fate of production. For some factors there may be a high degree of 

certainty about likely outcomes (e.g., a government decision may have been made to develop roads or 

provide power supply that will benefit growers and downstream processors and enhance delivery to 

markets in the future; for other factors, there may be a high degree of uncertainty about their future 

state or level (e.g., will market requirements for timber resource characteristics be more or less 

specific in 20 to 30 years time?). Dealing with the uncertainty of future conditions is a major issue in 

forestry investment for small-scale and large-scale growers alike. 

Innovation, experience and adoption: There are many factors to consider before a tree is planted with 

confidence that it will be part of a successful agroforestry production system. As a land use system, 

agroforestry is more knowledge intensive than agricultural innovations such as a new crop variety or 

fertilizer regime. Consequently the uptake of agroforestry is not a simple exercise in enterprise or 

input substitution but involves modification to the whole farming operation. Effective implementation 

requires substantial investment in education and training (Mercer, 2004) complemented by a good 

understanding of the capacity and motivation of prospective growers. 

There is added uncertainty for smallholders where agroforestry requires establishing a new mix of 

trees and annual food and fodder crops often with a new land management technique such as contour 

hedgerows, alley cropping or enriched fallows (Mercer, 2004). Successful outcomes depend on access 

to good knowledge and technology inputs and the capacity and patience of growers to apply and adapt 

that knowledge to their situation. This inherent uncertainty of agroforestry production systems will 

require a slow initial adoption rates suggesting ‘that agroforestry projects will require longer time 

periods before becoming self-sustaining and self diffusing than the earlier Green revolution 

innovations’ (Mercer, 2004, p.312). In the initial stages of implementation of agroforestry systems, it 

will not be known what impacts tree crops will have on the productivity and/or quality of annual crops 

until various combinations have been tried and demonstrated, and estimates of the impacts of annual 

cropping regimes on the productivity and quality of the tree crops need to be made over a number of 

years. Smallholder agroforestry growers and their advisors will benefit from tools that can guide them 

to identify the best species to plant and how to manage inherent uncertainties and risks to best meet 

their objectives. 

ASSESSING THE RISKS – SPECIES SELECTION 

Where sufficient data and information exist, the decision framework should be sufficient to provide an 

indication of the feasibility of agroforestry regimes in a given area. Where there are several gaps,



alternative approaches may be needed to support decisions, allowing prospective investors to explore 

possible futures in a highly uncertainty environment. 

Markets: Markets exist for a range of natural forest and plantation species grown in PNG. The 

existence of a market is important to collect price information and roundwood specifications to help 

inform the grower’s allocation of livelihood assets. This can be combined with species experience to 

better understand likely uncertainty. Table 1 presents a matrix based on these two factors for a sample 

group of species. The least uncertain option is where a planted species has been grown and sold into a 

market, provided that the market continues. Thus, teak is a least uncertain species; in contrast, 

although a market exists for natural forest Intsia bijuga, it is not clear whether plantation grown 

material will be accepted, or if that market will still exist if there is a substantial gap between reduced 

supply from natural forests and wood from plantations coming on stream in perhaps 50 years time. 

Market 

Existing New 

New species Intsia bijuga (natural forest) Intsia bijuga (plantation) 

Existing species Acacia mangium, Tectona grandis, 

Eucalyptus deglupta, Araucaria hunsteinii 

Table 1. An example of an assessment matrix based on experience with the species in 

plantations and its present market status. 

Species experience: Where there is experience with the silviculture of a species, the projected 

outcomes of management should be more certain. In PNG there are few reliable data on indigenous 

species as planted forests on which to base commercial management decisions (Karmar et al., 2004, 

p.16). The level of domestication and breeding of a species is an important factor in reducing 

uncertainty. Domestication and breeding is a series of sequenced repeated selections and mating to 

change gene frequencies (Eldridge et al., 1997, p.1). Leakey (2006, p.4) presents a species 

commercialisation continuum for non-timber forest products, indicating that the process has a series of 

stages. The development of trees for wood production would require a similar development cycle. 

Species that have already been grown to full operational rotations would represent the least level of 

uncertainty, whereas wild seed of an untested species would have the greatest uncertainty. A 

smallholder formulating strategic plans for future cashflows, e.g. for education fees, is likely to prefer 

planting a species with more certain outcomes. The other dimension to risk and uncertainty 

management is rotation length as a proportion of a person’s life. The higher the proportion, the lower 

the chance of undertaking another crop and the greater the likelihood that the plantation will be an 

intergenerational crop. 

ASSESSING THE BENEFITS OF SMALLHOLDER AGROFORESTRY 

Fit with food security. The SLF proposes that livelihood assets are first allocated to ensure food 

security as a primary concern. Where subsistence agriculture is practiced, development of a species 

options must consider labour requirements for different components of the land use system. Figure 1 

presents the labour inputs to establish, maintain, and harvest to roadside on-truck mangium pulpwood 

in the Madang area. Mangium is of interest in part because its production interval complements that of 

a garden rotation in the Madang region, which includes a 4 to 5 year fallow period. The initial crop 

establishment (including year 0 maintenance = 21.0% of labour) and the final harvest (39.5% of 

labour) consume the greatest labour. Assuming that total labour available per year is 260 days (net of 

community and church commitments), tree growing would be a significant proportion of labour, 

potentially at the expense of food production. However, site visits indicate that the local farmers were 

incorporating mangium into their garden system, suggesting that farmers were managing the time 

demands for both gardening and mangium growing.



Figure 1. The labour inputs to establish, manage and harvest a mangium pulpwood crop 

established at 1,100 stems/ha in the Madang area. The site was assumed to be 7 

years of regrowth forest on an ex-garden site. 

Returns on labour: agroforestry can generate community benefits, provide employment opportunities 

for local communities and contribute to social stability within a community or region, especially 

where there is excess labour capacity. A smallholder must find a balance between allocation of their 

own and contract labour, which will depend on their broader community focus compared to their own 

interests. Taking caution from Chambers (2007, p.38-41), labour productivity will be impacted on by a 

range of factors. Returns on labour, (with cash crops) can be used to compare crop options by 

smallholders or the option to work for wages (K8/day is the base wage). Figure 2 presents returns on 

labour and total labour input for PNG cash crops (taken from Bourke and Harwood, in press) and a 

series of mangium management scenarios (using Jenkin, 2002). The aim was to test for drivers of 

returns, while keeping the returns constant. The impact on garden output has been excluded as, at this 

stage of the project, this information is yet to be collected. 

Figure 2. Returns on labour and labour demands for PNG cash crops (Bourke and Harwood 

in press) and a series of mangium management scenarios. The tree crop was assumed 

to grow at a mean annual increment of 13.3 m 3 /ha/y at age 6 years. The percentage 

indicates the area covered e.g. 25% would indicate 1 in 4 rows are planted. ‘G’ indicates 

gardening applied for the number of years indicated, replacing maintenance inputs. ‘FG’ 

indicates gardening throughout the rotation. ‘FM’ indicates full labour maintenance 

applied. ‘CH’ indicates that contract harvesting is used. Seedlings are assumed to be 

provided free to the smallholders. 

Smallholder intercropping results in frequent and intensive tending operations for the crops, improving 

planted tree survival and growth rates (Bertomeu et al., 2008, p.28). Increasing replacement of 

maintenance labour by gardening improves returns on the marginal increase in labour for the planted 

tree crop e.g. from K8.88 / day when providing all labour in a woodlot to K217.33 / day with planting



labour only and the maintenance by gardening throughout the rotation. The practicalities of each 

scenario will be determined in the field. The use of contract labour at K8.00 / day to harvest to 

roadside and load trucks is a key driver of smallholder returns. 

Further scenario analysis, assuming different levels and forms of labour inputs, of a mangium 

pulpwood regime was conducted (Figure 3). The analysis indicates the impact of increasing reliance 

on gardening to achieve tree inputs, and of the choice to contract harvest to roadside. The use of 

contract harvest labour was observed in the field, reflecting local landowners’ assessment of the most 

efficient use of their time. The analysis indicates that, under the assumptions made, integration of tree 

crops as a marginal activity within a garden system provides an important cash returns on the marginal 

smallholder labour deployed. 

Figure 3. Returns on labour and labour demands for a series of mangium management 

scenarios. The tree crop was assumed to grow at a mean annual increment of 13.3 

m 3 /ha/y at age 6 years. The labour inputs are: ‘Full labour’ = 100% smallholder labour; 

‘M1’ = 3 rounds of 6days/ha/round of maintenance per year; ‘M2’ = 2 rounds of 

6days/ha/round of maintenance per year; ‘M3’ = 1 rounds of 6days/ha/round of 

maintenance per year; ‘G’ = gardening to achieve maintenance in the years indicated; 

‘P’ = smallholders planted the tree crop; ‘CH’ indicates that contract harvesting is used. 

Seedlings are assumed to be provided free to the smallholders. 

CONCLUDING COMMENTS 

In situations where risk and uncertainty can have catastrophic results, such as farming systems 

decisions in many parts of PNG, it is necessary ensure that any recommendations for changes to these 

systems helps minimise risk and uncertainty in decision making. This paper has reviewed a conceptual 

framework for appropriate decision making, and applied that approach to a sample of species and tree 

growing systems used in PNG. The next stage of the project will integrate this approach with the 

outcomes of research investigating landowners’ attitudes to tree growing, and more precise 

information about the basis for their decision making. 


The project is funded by the Australian Centre for International Agricultural Research (ACIAR). We 

thank ACIAR and its Forestry Program Manager, Dr Russell Haines, for their support. The project 

team has been supported by many PNG institutions – including the PNG Forest Authority, PNG 

University of Technology, and Ok Tedi Development Foundation. We have been hosted by many 

people in PNG during field visits; their assistance is invaluable. The knowledge of the other team 

members has helped frame the concepts presented; in particular, we thank Dr Mike Bourke, whose 

PNG experience and passion has been a key ingredient, and acknowledge the contributions of Hartmut 

Holzknecht, Lastus Kuniata, Andrew McGregor, Kulala Mulung and Ruth Turia.



REFERENCES 

Ashley, C. & Carney, D., 1999. Sustainable livelihoods: lessons from early experience. London: Department for 

International Development. 

Australian Centre for International Agricultural Research, 2009. Value-adding to PNG agroforestry systems. 

Available at: http://www.aciar.gov.au/project/FST/2004/050 [accessed 04/03/09]. 

Bank of Papua New Guinea, 2009. Table 8.7. Non-agricultural exports – quantities exported of major 

commodities. Available at http://www.bankpng.gov.pg [accessed 05/03/09]. 

Bannock, G., Baxter, R.E & Davis, E., 1992. The Penguin dictionary of economics. 5 th ed. England: Penguin 

Press. 

Bertomeu, M., Roshetko, J.M. & Fay, C. 2008. Domesticating landscapes through farmer-led tree cultivation: an 

agroforestry approach to reforestation in Southeast Asia, In: IUFRO International Symposium, Smallscale 

rural forest use and management: global policies versus local knowledge. Gerardner, France, 23- 

27 June 2008. Nancy: France. 

Bigsby, H.R. ed., 2001. Assessment and management of forest investment risk. Canterbury, New Zealand, 

Lincoln University Press. 

Bourke, M. & Allen, B., 2008. Poverty in rural Papua New Guinea. [Presentation at Divine World University, 

Madang] May 2008. Canberra: Land Management Group, The Australian National University. 

Bourke, R.M. & Harwood, T. eds, (in press). Food and Agriculture in Papua New Guinea. Canberra: The 

Australian National University. 

Byron, N., 2001. Keys to smallholder forestry in developing countries in the tropics. In: Harrision, S. R. and 

Herbohn, J. L. eds. 2001. Sustainable Farm Forestry in the Tropics: Social and Economic Analysis and 

Policy. UK: Edward Elgar Publishing. Part. IV 

Chambers, R., 2007. Poverty research: Methodologies, mindsets and multidimensionality. IDS Working Paper 

293. Brighton (UK): Institute of Development Studies, The University of Sussex. 

Chambers, R. & Conway, G.R., 1991. Sustainable rural livelihoods: practical concepts for the 21 st century. IDS 

Discussion Paper 296. Brighton (UK): Institute of Development Studies, The University of Sussex. 

Customs Tariff Act 1990 (Schedule 2: Rates of Export Duties). Port Moresby: Government of Papua New 

Guinea. Available at http://www.customs.gov.pg [accessed 05/03/09]. 

Eldridge, K., Davidson, J., Harwood, C. & van Wyk, G., 1997. Eucalypt domestication and breeding. Oxford: 

Clarendon Press. 

Hanson, L.W., Allen, B.J., Bourke, R.M. & McCarthy, T.J., 2001. Papua New Guinea rural development 

handbook. Canberra: The Australian National University. 

Holding, C. & Roshetko, J.M., 2003. Farm-level timber production: orienting farmers towards the market. 

Unasylva 212. (54), pp.48-56. 

Jenkin, B.M., 2002. The search for transparency in determining the share of residual woodchip stumpage returns 

to growers. The Australian Forest Grower, 25(4), Special Liftout No. 62. 

Karmar, M., Spencer, D. & Aken, K., 2004. Report on the assessment of the Papua New Guinea Forest 

Authority’s (Forest Research Institute) planted trials of indigenous species. [Draft report] March 2004. 

Part of ACIAR Project FTS/1998/115 Domestication of Papua New Guinea’s indigenous forest species. 

Canberra: CSIRO Forestry and Forest Products. 

Kanowski, P., Holzknecht, H. & McGregor, A., 2007, Value adding to Papua New Guinea agroforestry systems 

through the incorporation of high-value tree species. [Scoping Project Report] November 2007. 

ACIAR Project FST/2005/050. Canberra: The Australian National University. 

Kanowski, P, Holzknecht, H, Mulung, Blyth, M, Jenkin, B & McGregor, A. 2008, Fostering commercial tree 

growing by small-scale landowners in Papua New Guinea from a base in traditional agroforestry 

systems, In: IUFRO International Symposium, Small-scale rural forest use and management: global 

policies versus local knowledge. Gerardner, France, 23-27 June 2008. Nancy: France. 

Kennedy, J. & Clarke, W., 2004. Cultivated landscapes of the Southwest Pacific. Resource Management in Asia- 

Pacific Working Paper No.50. Canberra: Resource Management in Asia-Pacific, The School of Pacific 

and Asian Studies, The Australian National University 

Kirby, C.M., Sinden, J.A. & Kaine, G.W., 1993. Appraisal of agroforestry investment under uncertainty: A 

South Australian case study. Australian Forestry, 56 (2), pp.109-119. 

Leakey, R. 1998. Editorial. Agroforestry Today. September.



Leakey, R.R.B., 2006. Domestication and commercialisation of agroforestry trees for AFTPs. In: 

Agroforestry.net, Trees for improving profitability, sustainability, and resource conservation on farms 

and ranches 2006: Pacific Island Agroforestry Workshops and Field Visits 2006. Keauhou, Kona, 

Hawaii, 16 th to 19 th May, 2006. Available at: http://www.agroforestry.net [accessed 13/03/09]. 

Mendoza, G.A., Campbell, G.E. & Rolf, G.L., 1986. Multiple objective programming: An approach to planning 

and evaluation of agroforestry systems – part 1: Model description and development. Agricultural 

Systems, 22, pp.243-253. 

Mercer, D.E., 2004. Adoption of agroforestry innovations in the tropics: A review. Agroforestry Systems, 61(1), 

pp.311-328. 

Messer, N. & Townsley, P., 2003. Local institutions and livelihoods: Guidelines for analysis. Rome: Rural 

Development Division, Food and Agricultural Organisation of the United Nations. 

Mulung, K, 2007, Land owner decision making processes in relation to commercial tree growing in Papua New 

Guinea (PNG). [Draft research proposal] October 2007. ACIAR Project FST/2004/050. Canberra: 

Australian National University. 

Pannell, D. J., Marshall, G. R., Barr, N., Curtis, A., Vanclay, F. & Wilkinson, R., 2006. Understanding and 

promoting adoption of conservation practices by rural landholders. Australian Journal of Experimental 

Agriculture. 46 (11), pp.1407-24. 

Scherr, S. J., 2004. Building opportunities for small-farm agroforestry to supply domestic wood markets in 

developing countries. Agroforestry systems. 61(1), pp.357-370. 

Schirmer, J., Kanowski, P. & Race, D., 2000. Factors affecting adoption of plantation forestry on farms: 

implications for farm forestry development in Australia. Australian Forestry, 63 (1), pp.44-51. 

Schirmer, J., 2004. Review of social perceptions of plantation expansion in Australia. In: Woodhead, A., 

Hyndman, D. & Dunn, T., eds 2004. Pratt Water: Timber Plantation Project. Social Assessment. 

Canberra: Bureau of Rural Resources, Department of Agriculture, Fisheries and Forestry. 

Tomich, T. P., Van Noordwijk, M., Budiddarseno, S., Gillison, A., Kusumanto, T., Murdiyarso, D., Stolle, F. 

and Fagi, A.M., 2001. Agricultural intensification, deforestation, and the environment: assessing 

tradeoffs in Sumatera, Indonesia. In: D.R Lee and C.B. Barrett, eds. 2001. Tradeoffs or synergies? 

Agricultural intensification, economic development and the environment. Wallingford UK: CAB- 

International, pp.221-24. 

Venn, T. J., 2005. Financial and economic performance of long-rotation hardwood plantation investments in 

Queensland, Australia. Forest Policy and Economics. 7(3), pp.437-454.

FORESTRY - Institute of Foresters of Australia

Create successful ePaper yourself

Delete template?

Save as template?