The challenge of HIV/AIDS: Where does agroforestry fit in? - World ...

CitationGarrity, D., A. Okono, M. Grayson and S. Parrott, eds. 2006. World Agroforestry into the Future.Nairobi: World Agroforesty Centre.ISBN 92 9059 184 6© 2006 World Agroforesty CentreArticles appearing in this publication may be quoted or reproduced without charge, provided the source is acknowledged.No use of this publication may be made for resale or other commercial purposes.All images remain the sole property of their source and may not be used for any purpose withoutwritten permission of the source.Views expressed in this publication are those of the authors and do not necessarily reflect the viewsof the World Agroforestry Centre.The geographic designation employed and the presentation of material in this publication do not imply the expressionof any opinion whatsoever on the part of the World Agroforestry Centre concerning the legal status of any country, territory,city or area or its authorities, or concerning the delimitation of its frontiers or boundaries.

ContentsPrefaceOpeningvviiAgroforestry and the FutureChapter 1. Science-based agroforestry and the achievement 3of the Millennium Development GoalsDennis GarrityTrees and MarketsChapter 2. Trees and markets for agroforestry tree products: Targeting 11poverty reduction and enhanced livelihoodsRoger Leakey, Zac Tchoundjeu, Kate Schreckenberg, Tony Simons,Sheona Shackleton, Myles Mander, Rachel Wynberg, CharlieShackleton and Caroline SullivanChapter 3. The future of perennial tree crops: What role for agroforestry? 23Hubert Omont, Dominique Nicolas and Diane RussellChapter 4. Trees and Markets: Working Group Report 37Land and PeopleChapter 5. Confronting land degradation in Africa: Challenges for 43the next decadeM.J. Swift, Ann Stroud, Keith Shepherd, Alain Albrecht,André Bationo, Paramu Mafongoya, Frank Place, Thomas P. Tomich,Bernard Vanlauwe, Louis V. Verchot and Markus WalshChapter 6. Agroforestry innovations for soil fertility management in 53sub-Saharan Africa: Prospects and challenges aheadBashir Jama, Freddie Kwesiga and Amadou NiangChapter 7. Scaling up the impact of agroforestry: Lessons from three sites 61in Africa and AsiaS. Franzel, G.L. Denning, J-P. Lillesø-Barnekow and A.R. Mercado JrChapter 8. Policies for improved land management in smallholder 71agriculture: The role for research in agroforestry and naturalresource managementFrank Place and Yves-Coffi PrudencioChapter 9. Land and People: Working Group Report 79

Enhancing Environmental ServicesChapter 10. Agroforestry and environmental governance 85Brent Swallow, Diane Russell and Chip FayChapter 11. The potential for agroforestry to contribute to the conservation 95and enhancement of landscape biodiversityBrent Swallow, Jean-Marc Boffa and Sara J. ScherrChapter 12. Watershed functions in productive agricultural landscapes 103with treesMeine van Noordwijk, Farida, Pornwilai Saipothong, Fahmudin Agus,Kurniatun Hairiah, Didik Suprayogo and Bruno VerbistChapter 13. Opportunities for linking climate change adaptation 113and mitigation through agroforestry systemsSerigne T. Kandji, Louis V. Verchot, Jens Mackensen, Anja Boye,Meine van Noordwijk, Thomas P. Tomich, Chin Ong, Alain Albrechtand Cheryl PalmChapter 14. Environmental Services: Working Group Report 123Strengthening InstitutionsChapter 15. Sustainable agriculture and rural development – a response 129to the realities of rural AfricaTony DolanChapter 16. Capacity building in agroforestry in Africa and Southeast Asia 135O.A. ChivingeChapter 17. Institutional collaboration in agroforestry: Networking 141and knowledge managementPer Rudebjer, Nguyen Van So and John R.S. KaboggozaChapter 18. Building capacity for research in agroforestry 147Jemimah Njuki, Issiaka Zoungrana, August Temu and Janet AwimboChapter 19. Can e-learning support agricultural development 155in developing countries?Albert Dean Atkinson, Jan Beniest and Sheila RaoChapter 20. Strengthening Institutions: Working Group Report 163Agroforestry and Major Cross-cutting IssuesChapter 21. Trees outside forests: Facing smallholder challenges 169Syaka Sadio and Patricia Negreros-CastilloChapter 22. Women, land and trees 173Ruth Meinzen-DickChapter 23. The challenge of HIV/AIDS: Where does agroforestry fit in? 181Marcela Villarreal, Christine Holding Anyonge, Brent Swallowand Freddie KwesigaAuthor Contacts 193Acronyms and Abbreviations 195

PrefaceStanding where we are today, it is increasingly hard to believe that not so long ago agroforestrywas not a familiar word, let alone a recognized concept. It is a tribute to the vision ofmany thousands of professionals during the past quarter century that agroforestry has nowachieved such wide recognition as an integrative science and practice with enormous potentialto transform lives and landscapes in today’s and tomorrow’s world. Their hard work anddedication has meant that millions of people around the world now not only know aboutagroforestry but are also directly benefiting from it.Before the World Agroforestry Centre (ICRAF) was launched in 1978, agriculture and forestrywere commonly treated as mutually exclusive endeavours in research and in practice. Buttropical small-scale farmers eventually taught us that the integration of trees in agriculturallandscapes has enormous untapped potential to benefit people and the environment. Whatremained was to deploy the power of science to accelerate the knowledge generation andproductivity increases that would effectively exploit this potential.From the outset, ICRAF scientists collaborated with rural people to learn how agroforestrysystems are evolved, and to find ways to better adapt and scale up the most successfulscience-based agroforestry innovations. They were pioneers of an integrated research modelthat works to blend different disciplines in order to tackle complex land management challengesand opportunities.Upon reaching the Silver Anniversary mark, The World Agroforestry Centre convened aninternational conference to reflect on the accomplishments of agroforestry research, and toconduct a forward-looking assessment of the role of agroforestry science in addressing thekey global and regional challenges in the future. Distinguished speakers were invited to sharetheir analysis and views on a wide range of topical issues. These were presented in sessionsthat were organized around the four global themes of the Centre’s work: Trees and Markets,Land and People, Environmental Services, and Strengthening Institutions.This volume contains a selection of those presentations that were subsequently written up andpeer-reviewed. It represents a snapshot of current thinking on the science of agroforestry andexciting future opportunities for its application. We want to thank all those that participatedin and contributed to the conference and to this volume. And I want to offer particular thanksto the Canadian International Development Research Centre for their financial contributionto the completion of this publication, and to all of the donor institutions that have invested inthe Centre over the years.

viWorld Agroforestry into the FutureThis 25 th Anniversary Conference was a chance to mark the occasion of our birthday byexamining the impact that ICRAF and agroforestry have made and can make. While ourquarter-century may seem to be a watershed year, when we can look back on our achievementsand forward to our future, in fact we are doing this all the time. We have created aculture of science within the Centre and do not like to stand still or rest on our laurels. Mostrecently we restructured our focus so that we concentrate on the four global themes mentionedabove. This allows us to better appreciate the interconnectedness of the environmentand try to retain – or recapture – the natural balance. We recognize that, far from beingdestructive, change is important. If I can predict anything, it is that as different as the needsof 2003 were from 1978, those of 2028 will be different again.Dennis GarrityDirector GeneralWorld Agroforestry CentreMay 2006

OpeningOpening the conference on a personal note, the Honourable Kipruto arap Kirwa, Kenya’sMinister for Agriculture gave Kenya’s perspective on agroforestry. He noted that schools andyoung people should be more involved in tending to trees. This would nurture a responsibilityfor the global environment, land care and sustainable development in the next generation.“We look at forests and agroforestry as lessening the pressure on land and providing asource of cash… The more we invest in agroforestry, the better our future,” he observed. “Sothe challenge is how to inculcate our people so they understand that the more you invest inagroforestry, the better and brighter your life and that of your children.”In his welcoming remarks, Dr Eugene Terry, Chair of the Centre’s Board of Trustees, notedthat the Centre’s accomplishments have transformed lives and landscapes; not to mention theCentre itself. He added that the development of the Centre started from the inside out, as ittransformed itself into a truly global enterprise tackling global issues. That change continuestoday with the Centre’s recently launched theme-based structure that supersedes its previousfocus on individual programmes. “This conference will help us get valuable feedbackon these new ideas from our partners and stakeholders,” he added.Thanking all investors, particularly Kenya, which hosts the World Agroforestry Centre’s globalheadquarters, Dr Terry stated, “Only by working together in the future, as we have in the past,can we realize the potential of agroforestry and sustainable development. Each one of usshould rededicate our effort as we move into the next phase of this institution’s existence.”Managing the landscapeObserving that about a quarter of the world’s poor depend to some degree on forests, Mr IanJohnson, Chair of the Consultative Group on International Agricultural Research (CGIAR)and Vice-President of the World Bank, noted, “ICRAF has shown us that forests are morethan trees, and that farming is closely connected to landscape management. The Centre hasunderstood both the local and the global dimensions of this connection. The World Bank willincrease its commitment to forests, which are a survival base for the poor.”According to Mr Syaka Sadio of the Food and Agriculture Organization of the United Nations(FAO), land degradation is the most serious environmental issue in Africa. He added that theworld food situation has never been so worrying, especially in the face of rising populations,declining yields and the decreasing quality of land resources. “At the current rate, we willachieve our Millennium Development Goals 140 years later than planned,” he lamented.But there is a glimmer of hope in this bleak picture: “Agroforestry seems to have promisingpotential to reduce the decline in soil fertility and to increase income. Agroforestry should berespected as a tool for integrating many different sciences. Trees outside of forests, including

viiiWorld Agroforestry into the Futureall forms of forestry in urban areas and rural gardens, are a practical approach to increasingincomes and enhancing soil fertility. Over the last 25 years, ICRAF has accumulated a majorknowledge base that is essential to the advancement of these efforts.”Mr Bakary Kante, Director of the Division of Policy Development and Law at the UnitedNations Environment Programme (UNEP), reaffirmed the Centre’s important accomplishments:“Twenty-five years of passionate research in agroforestry is a major achievement, notjust for the World Agroforestry Centre, but also for the millions of farmers who benefit fromthe work of the Centre… Because of the wonderful work that is undertaken here, we aresuccessfully integrating agroforestry into the Clean Development Mechanism (CDM). Agroforestryintegrates environmental concerns and poverty eradication. In arid lands, trees likeshea and baobab are helping the poor to survive, and guaranteeing children three mealsa day.”Towards the future“Today’s needs are very different from those of 25 years ago. I am convinced that the Centre’scurrent staff, Board and management will carry the Centre confidently into the next 25 years,”predicted Ms Eva Ohlsson of the Swedish Agency for International Development Cooperation(Sida), who spent six years conducting research with ICRAF in Western Kenya in the1990s. Sida is a leading donor and supporter of the Centre’s early work on sesbania fallows inZambia. These tree fallows and other fertilizer tree systems are now widely practised acrosssouthern Africa.Other prestigious partners added their voice. “We continue to derive encouragement from ouraffiliation with the World Agroforestry Centre,” affirmed Ms Maureen O’Neil, President of theInternational Development Research Centre (IDRC). She explained that “the Centre that IDRChelped establish is a continuing source of pride for us. Integrated interdisciplinary participatoryresearch is one of the founding principles of both IDRC and the Centre. Only throughpartnerships can all the resources needed to solve a problem be harnessed.”Dr Bjorn Lundgren, the Centre’s Director General through the 1980s, remarked that networkingand partnership were the Centre’s way of doing business long before they became popularelsewhere. “We were one of the first interdisciplinary institutions. It was also in the 1980s thatwe laid the foundation of agroforestry as a science,” he recalled.Lundgern’s observations were echoed by Dr Pedro Sanchez, the Centre’s Director Generalduring the 1990s and winner of the World Food Prize in 2002. He commented that theCentre’s impact at its foundation extended further than networking, “We began to talk scienceand not just hope.” He then took a look into the future, and declared that. “ICRAF is at theforefront in the global science of watersheds, and in using light to predict soil properties.Using technologies such as these, we can simultaneously address the needs of the poor andconduct cutting-edge research.”Further reflections on the Centre’s past came from Dr M.S. Swaminathan, who also articulateda vision for the Centre’s future. As the first Vice-Chair of ICRAF’s Board of Trustees,

OpeningixDr Swaminathan has been intimately connected with the World Agroforestry Centre from itsvery earliest moments, having been part of a small group of 10-or-so visionary minds that metat IDRC’s headquarters in Canada in late 1977. This group helped to develop agroforestryas a concept, and successfully played ‘midwife’ to the birth of the Centre. That meeting wasconvened by David Hopper, the first President of IDRC. Inspiration was provided by the lateJohn Bene’s well-known paper that first proposed the imperative of an international centrededicated to championing the science and practice of agroforestry 1 . Bene (1910–1986) servedas ICRAF’s first Board Chair (1977–1979), and was succeeded by Swaminathan. 2In a video-taped speech, Dr Swaminathan urged agroforesters to broaden their reach and paygreater attention to the vast coastal ecosystems. Agroforestry can open new pathways to utilizingcoastal areas at a time when freshwater supplies are dwindling worldwide. He declaredthat the gender dimension of agroforestry and its impact on women is crucial and must not beoverlooked. He noted that agroforestry systems should be designed to ensure that they reallymeet the needs and problems of the people, while also seeking to understand the rationalebehind local land management practices.Charting a course for the future, Swaminathan encouraged the Centre to pursue a new paradigmof partnerships that emphasize public goods over patents, especially public goods thatbenefit resource-poor farmers and farm families throughout the tropical world. In conclusion,he recounted, “Agroforestry has always inspired me. I am glad that ICRAF has grown into itspresent stature and taken an appropriate name – the World Agroforestry Centre. Livelihoodsecurity and ecological security are two sides of the same coin. By choosing an agroforestryspecialist for the World Food Prize for the first time, the connection between food, forestryand agroforestry has been clearly recognized.”The 25 th Anniversary Conference of the World Agroforestry Centre (ICRAF) was officiallyopened by the Honourable Kipruto arap Kirwa, Kenya’s Minister for Agriculture. Notabledelegates, most of whom have played an important part in the history of the centre, spokeat the opening session. Also in attendance were distinguished guests, including: the UgandanHigh Commissioner to Kenya, His Excellency Matayo Kyaligonza; the Permanent Secretaryin Kenya’s Ministry of Agriculture, Joseph Kinyua; David Kaimowitz, Director General of theCenter for International Forestry Research (CIFOR); Carlos Sere Director General of the InternationalLivestock Research Institute (ILRI); Roger Leakey, a former Director of Research at theCentre; and directors and scientists from many national research institutions in Africa, Asiaand Latin America.1Bene, J.G., H.W. Beall and A. Côté 1977. Trees, food and people: land management in the tropics.International Development Research Centre (IDRC), Ottawa, Canada.2For a fuller version of ICRAF’s early history, refer to the companion booklet for this volume: ICRAF 2003.The World Agroforestry Centre: Looking back at the first quarter century and ahead to the next. ICRAF,Nairobi, Kenya.

Agroforestry and the Future

“Beyond question, agroforestrycan greatly improve life forpeople in the developing world,and do so within a reasonablyshort time.”Bene et al. 1977

Keywords:Millennium Development Goals, fertilizer tree systems,expanded tree cropping, improved tree product processingand marketing, agroforestry transformationChapter 1Science-based agroforestry and the achievementof the Millennium Development GoalsDennis Garrity, World Agroforestry CentreAbstractThe Millennium Development Goals (MDGs) of the United Nations (UN) are at the heart of the globaldevelopment agenda. This chapter examines the role of agroforestry research for development in lightof the MDGs. It reviews how agroforestry is materially assisting to achieve the goals. And it discusseshow the agroforestry science agenda can be realigned to further increase its effectiveness in helpingdeveloping countries to meet their MDG targets. Promising agroforestry pathways to increase on-farmfood production and income contribute to the first MDG, which aims to cut the number of hungry anddesperately poor by at least half by 2015. Such pathways include fertilizer tree systems for smallholders,and expanded tree cropping and improved tree product processing and marketing. These advancescan also help address lack of enterprise opportunities on small-scale farms and child malnutrition.The rate of return to investment in research on tree crops has been shown to be quite high (88%).But enterprise development and enhancement of tree product marketing have been badly neglected.Tree domestication, and the commercial processing and marketing of tree products and services is anew frontier for agroforestry research for development. A major role is also emerging in the domain ofenvironmental services, particularly the development of mechanisms to reward the rural poor for thewatershed protection, biodiversity conservation, and carbon sequestration that they provide to society.Agroforestry research for development is contributing to virtually all of the MDGs. But recognition forthat role must be won by ensuring that more developing countries have national agroforestry strategies,and that agroforestry is a recognized part of their development agenda.IntroductionAchieving the Millennium Development Goals (MDGs)is at the heart of the global agenda. The goals embodythe world’s aspirations to eliminate desperate hungerand poverty, ensure decent health, enable universaleducation and elevate the status of women, whileconserving and regenerating the global environment.Attaining these goals is the greatest challenge of ourgeneration. Their accomplishment will bring benefitsto everyone, including greater economic abundance,peace, and security to all people on the globe. Successin achieving the MDGs requires overcoming hungerand poverty in ways that are more thorough, comprehensiveand holistic than ever before. We must attackthese problems at their roots, through developmentthat permeates the heart of rural poverty in the developingworld.A clear vision is evolving that articulates how agroforestryresearch and development can contribute

4World Agroforestry into the Futurematerially to achieving these goals andaspirations. The vision drives a strategythat harnesses the best of global science.This chapter lays out that vision andstrategy, providing a framework for thesections and chapters that make up theremainder of this volume.Where agroforestry fits inTrees play a crucial role in almost all terrestrialecosystems. They provide a widerange of products and services to ruraland urban people. As natural vegetation iscleared for agriculture, trees are integratedinto productive landscapes – the practiceknown as agroforestry.Agroforestry is practised by millions offarmers, and has been a feature of agriculturefor millennia. It encompasses a widerange of working trees that are grown onfarms and in rural landscapes, and includesthe generation of science-based tree enterpriseopportunities that can be importantin the future. Among these are: fertilizertrees for land regeneration, soil healthand food security; fruit trees for nutritionand income; fodder trees that improvesmallholder livestock production; timberand fuelwood trees for shelter and energy;medicinal trees to combat disease, particularlywhere there is no pharmacy; and treesthat produce gums, resins or latex products(Garrity 2004). Many of these trees havemultiple uses, each providing a range ofbenefits.An estimated 1.2 billion rural people currentlypractise agroforestry on their farmsand in their communities, and dependupon its products (World Bank 2004).Their tree-based enterprises help ensurefood and nutritional security, increase theirincome and assets, and help solve theirland management problems. Trees play aparticularly pivotal role wherever peopledepend on fragile ecosystems for survivaland sustenance.During the past 30 years, agroforestryhas progressed from being a traditionalpractice with great potential to the pointwhere development experts agree thatit provides an important science-basedpathway for achieving important objectivesin natural resource management andpoverty alleviation. Despite its ubiquitoususe by smallholder farm families, there isinadequate awareness about the potentialof agroforestry to benefit millions ofhouseholds trapped in poverty. We need aglobal ‘agroforestry transformation’ to mobilizescience and resources to remove thesocio-economic, ecological and politicalconstraints to widespread application ofagroforestry innovations, and thereby helpattain the MDGs.Building on three decades of work withsmallholder farmers in Africa, Asia andLatin America, coupled with strategic allianceswith advanced laboratories, nationalresearch institutions, universities and nongovernmentalorganizations (NGOs) acrossthe globe, the World Agroforestry Centreand its partners are poised to foster such anagroforestry transformation.Aiming for an ‘agroforestrytransformation’The World Agroforestry Centre’s mission isto advance both the science and practiceof agroforestry to help realize this transformation.The target is a future in whichmillions of poor farming households haveaccess to a wide variety of adapted andproductive tree enterprises that improvelivelihoods in a holistic way (World AgroforestryCentre 2005). Underpinning this iscrucial scientific research that will ensurea stream of necessary technical, policy andinstitutional innovations. The Centre hasidentified seven major global challengesrelated to the MDGs to which we aim tocontribute. These challenges are:1. To help eradicate hunger throughpro-poor food production systemsin disadvantaged areas based onagroforestry methods of soil fertilityreplenishment and land regeneration.2. Reduce rural poverty through marketdriven,locally led tree cultivationsystems that generate income andbuild assets.3. Advance the health and nutrition of therural poor through agroforestry systems.4. Conserve biodiversity through integratedconservation-development solutionsbased on agroforestry technologies, innovativeinstitutions, and better policies.5. Protect watershed services throughagroforestry-based solutions that enablethe poor to be rewarded for their provisionof these services.6. Enable the rural poor to adapt to climatechange, and to benefit from emergingcarbon markets, through tree cultivation.7. Build human and institutional capacity inagroforestry research and development.Mission goalsTo address these seven global challenges,the Centre is pursuing four mission goals:Goal 1Enhance access by smallholders to highqualitytree germplasm and expandedmarket opportunities for smallholder treeproducts.Goal 2Advance understanding of the role oftrees in practical and productive land andfarm management and to foster integratedfarming systems based on appropriate

Chapter 1: Science-based agroforestry and the achievement of the Millennium Development Goals 5agroforestry systems for key agroecologicaldomains.Goal 3Increase recognition and deployment ofpro-poor agroforestry strategies that generatelocal benefits while providing globalenvironmental services.Goal 4Greatly improve the capacity for effectiveresearch, development and education inagroforestry by a wide range of individualsand institutions in the developing world.The Centre’s collaborative advantage inaddressing these challenges lies in its rolein being able to synthesize and integratethe science and practice of agroforestry atmultiple levels. Scientific teams deploy thenecessary experience in research, developmentand education to produce agroforestryinnovations in accordance with localneeds and priorities. Collaboration withlocal development partners helps integratethese innovations into their work with therural poor. Our research in more than 30developing countries allows for learningand synthesis across a wide range of social,economic, ecological and institutionalcontexts.Four integrated research anddevelopment-support themesThe mission goals are addressed throughfour global themes, each related to a correspondinggoal.Theme 1 – Trees and MarketsThis theme focuses on the key smallholdertree commodities, their cultivation, valueaddedprocessing and the market environmentfor tree-based products. Poor farmersin less-favoured environments often cannotcompete advantageously in the productionof basic food commodities. They need tomeet their basic needs while diversifyingtowards higher-value products. This requiresa new approach. New tree speciesneed to be domesticated. New strategiesand methods for tree product developmentand the diversification of cultivation systemsneed to be pursued to better meet theneeds of farmers and markets. The themefocuses on four aspects of this: tree domesticationwith intensification of tree cultivationsystems; sustainable seed systems andmanagement of genetic resources of agroforestrytrees; enterprise development andenhancement of tree product marketing;and farmer-led development, testing andexpansion of tree-based options.Theme 2 – Land and PeopleThis theme focuses on the household farmsystem and the integration of trees intoproductive enterprise portfolios that meetfamily needs. It seeks to understand thebasis for the role of trees in sound landmanagement and quantifies the long-termconsequences of agroforestry practices onsmall-scale agriculture in order to derivelocally relevant land management options.Smallholders need integrated portfoliosof tree enterprises that address their basicneeds and provide cash income. Theseare being assembled and disseminated ina range of agroecological domains, alongwith the best-bet agroforestry managementoptions. The components of these systemsinclude trees for improving soil fertility,fruits and vegetable trees for nutrition,fodder for livestock, live fencing, timber,fuelwood, and services such as microclimateregulation. Smallholder tree cashcrops (coffee, cocoa, rubber) are a majorcontributor to rural incomes in many tropicalcountries. Work on diversifying thesesystems through integrating other valuableagroforestry trees into them enables smallholdersto buffer their incomes in the faceof volatile and declining cash crop prices.The poor are often enmeshed in highlycomplex poverty traps. Thus, we are developingand fostering pro-poor participatorytechnology development approachesand the enabling policies to address thesecomplex constraints.Theme 3 – Environmental ServicesThis theme aims to develop pro-poor agroforestrystrategies for both local benefitsand global conservation. The work focuseson watershed protection, biodiversityconservation and climate change mitigationand adaptation. The goal is to identifyagroforestry systems and landscape mosaicsthat meet farmer needs for food andincome while enhancing these services.Centre scientists are refining the principlesand practices to enable communities tofarm sustainably while protecting watershedservices. Our recent breakthroughs in costeffectivemethods for rapidly assessing landquality have greatly enhanced this work.These are combined with the methods ofintegrated natural resource management(INRM) that the Centre has helped pioneerin its research around the world.The work in this theme is also advancingthe understanding of, and capacity tomanage biodiversity in human-dominatedlandscape mosaics in the tropics. Thiswork is helping to develop the scientificbasis for ‘ecoagriculture’ – an approach toincreasing agricultural productivity whileprotecting natural biodiversity. Agroforestryis also playing a critical role in developingintegrated biodiversity conservationapproaches for the most critical globalhotspots while enhancing the livelihoodsof the rural poor in adjoining areas. Centreteams are on the ground in many of theseareas, working to develop a global networkof successful cases, in collaborationwith the Center for International ForestryResearch (CIFOR) and other partners.

6World Agroforestry into the FutureWe are also clarifying the processes forsmallholder adaptation to climate changeby developing more resilient tree-basedfarming systems. ICRAF is also working tosuccessfully demonstrate how smallholderfarmers can benefit by adopting agroforestrysystems that sequester carbon and contributeto climate change mitigation. Andwe are providing science-based evidenceon the tradeoffs and complementarities betweenland use for environmental servicesand for livelihoods of smallholder farmers.Theme 4 – Strengthening InstitutionsAgroforestry is a relatively new field of scienceand development. The Centre recognizedearly on that there is a major need tobuild the capacity of regional, national andlocal institutions to effectively participatein generating and applying innovations inagroforestry and INRM. The work in thistheme thus supports the development ofhigh-quality and relevant education programmes,knowledge sharing networks,and mechanisms to link with farmersfor effective sharing and management ofknowledge for development. It adds valueto the efforts of national agricultural, forestryand natural resource institutions byenhancing their ability to develop vibrantagroforestry research programmes and linkthat research to human development.The Centre has fostered the developmentof two major networks that support educationalinstitutions throughout Africa andAsia. They incorporate multidisciplinaryapproaches to land management into curricula,and develop and improve teachingand learning resources and techniques inINRM. Recognizing that agroforestry is anexcellent means through which to conveyenvironmental concepts in elementaryschools, the Centre and the Food and AgricultureOrganization of the United Nations(FAO) have launched a global programmecalled Farmers of the Future to foster thisapproach in the primary and secondaryschool systems of developing countries.Strategic links have also been establishedwith farmers’ groups and developmentorientedorganizations to bolster awarenessamong policy makers and development institutionsand improve access to knowledgeproducts on agroforestry and INRM. Thiswork is fostering innovative communitybasedapproaches to sustainable land management.These include Landcare, a farmers’movement that exhibits great promisein Asia and Africa.Cross-thematic issuesAgroforestry and the advancement ofwomenIn the developing world 60–80 percentof farmers are women. Rural women indeveloping countries grow and harvestmost of the staple crops that feed theirfamilies. This is especially true in Africa,where women account for 75 percent ofhousehold food production (UNDP 1999,as cited in Bread for the World Institute2003). Food security throughout the developingworld depends primarily on women.Yet they own only a small fraction of theworld’s farmland and receive only a fraction(less than 10 percent) of agriculturalextension services.Agroforestry offers many entry points toimprove the status, income and health ofwomen and children. Rainwater harvestingand tree-growing on farms reduces thedrudgery of fetching water and fuel fromdistant areas. Research on gender andagroforestry is examining these issues, andexploiting important entry points throughwhich women’s property rights can beenhanced, and how household agroforestrysystems can specifically address their nutritional,health and economic needs.Agroforestry linkages with better healthand nutritionAdvances in agroforestry can improve thehealth and nutrition of the rural poor. Theexpansion of fruit tree cultivation on farmscan greatly increase the quality of children’snutrition. This is particularly importantbecause indigenous fruit tree resourcesin local forests are often overexploited.Work with national partners to domesticatea range of nutritious wild indigenousfruits seeks to save these species from overexploitationand develop them for local andregional markets. These efforts will contributeto MDG 4 on reducing child mortality.There are many complex linkages betweenagroforestry and the fight against HIV/AIDS.Forty million people currently live withHIV. There is potential for agroforestry togenerate much-needed income, improvenutrition, reduce labour demands andstabilize the environment in AIDS-affectedcommunities. The range of threats and thevarious opportunities have yet to be thoroughlyexplored, and incorporated into theresearch and development agenda.Regional programmesThe Centre focuses primarily on seven regionswhere the problems of poverty, foodinsecurity and environmental degradationare most acute. High population density,extreme poverty and land degradationoverlap to create strategic entry points forthe establishment of an agroforestry transformation.At the regional level, we concentrate onresearch, education and developmentpriorities for different agro-ecological

Chapter 1: Science-based agroforestry and the achievement of the Millennium Development Goals 7zones including dryland areas, humidforest zones, and tropical highlands. TheCentre also assists its partners in developingnational agroforestry plans and incorporatingagroforestry into poverty reductionstrategies, and food security and environmentalpolicies. Research networks andpolicy initiatives provide leverage to promoteagroforestry and increased impact.AfricaForty-eight percent of the African populationis desperately poor – the highestproportion in any region of the world. Percapita food production is declining, andmalnourishment and poverty continueto increase. But the global community ismobilizing its resources to address this intolerablesituation. An analysis of the farmingsystems in Africa by FAO (Dixon et al.2001), and recent comprehensive studiesby the InterAcademy Council of ScientificSocieties (2004) and the UN MillenniumProject (2005) have provided a thoroughpicture of the constraints and possibilitiesin attempting to alleviate hunger and ruralpoverty in Africa. These studies identifyhotspots where focused efforts can enhancefarm productivity, increase ruralincomes and transform agriculture to becomea more dynamic driver of economicgrowth. In their recommendations on howto address the constraints, these studieshighlight agroforestry as a crucial pathwaytoward greater prosperity.The World Agroforestry Centre currentlyinvests three-quarters of its income in itsfour African regional programmes. Weforesee a series of ‘evergreen revolutions’in Africa based on the intensification anddiversification of farming systems that havedemonstrated potential for major productivityincreases. The Centre has teams onthe ground investigating key opportunitiesfor science-based agroforestry to helpovercome poverty in nearly all of the keyhunger spots. This will involve bringingtogether the best technology for trees,crops, and livestock into integrated farmingsystems suited to the diverse ecologies,backed by better markets, more vibrantrural institutions, and a more conducivepolicy framework for development.AsiaOn this vast continent the Centre has tworegional programmes: Southeast Asia andSouth Asia. In Southeast Asia we focus onimproved land-use practices that integrateproductive trees into agroforest landscapesthat provide important environmental services.The incidence of desperate povertyis decreasing in this region, but povertyremains concentrated in the less-favouredupland environments. These areas areparticularly well suited to agroforestry.A unique network; Rewarding the UplandPoor for Environmental Services (RUPES)is investigating the nature of these servicesand developing the basis to recognizeproperty rights and transfer benefits. Theseadvances are based on the deployment ofnegotiation support systems, a methodologythat provides a science-based approachto managing the trade-offs among competinginterests in managing critical environments.In South Asia we focus on four mainecosystems with large populations of ruralpoor. We work in collaboration with astrong, well-established research, developmentand education system, and are linkingSouth Asian agroforestry science withAfrican regions that have similar ecologies.Latin AmericaHere we participate with the AmazonInitiative. This is a consortium that bringstogether a number of international andnational research institutions to focus onreversing natural resource degradation,while improving the livelihoods of the ruralpoor. In Latin America, Southeast Asia,and the African Humid Tropics regions,we are building on our long-term involvementwith the system-wide programme ofthe Consultative Group on InternationalAgricultural Research (CGIAR) known asAlternatives to Slash-and-Burn (ASB). ASBfocuses on the landscape mosaics (comprisingboth forests and agriculture) whereglobal environmental problems and povertycoincide at the margins of the remainingtropical forests.Connecting to the global policyenvironmentAgroforestry is one of the few productiveland uses that contribute directlyand synergistically to the objectives of allthe key international environmental andsustainable development conventions. Inparticular, it contributes to the goals of theConvention on Biological Diversity (CBD),the United Nations Framework Conventionon Climate Change (UNFCCC) and itsClean Development Mechanism (CDM),the United Nations Convention to CombatDesertification (UNCCD), and the UnitedNations Forum on Forests (UNFF). TheCentre also strives to ensure that the scienceand practice of agroforestry reachesthe mainstream within such internationalpolicy fora and initiatives as the New Partnershipfor Africa’s Development (NEPAD).Similarly, the MDGs provide importantmilestones that have helped to mobilizeand channel the interests of developingcountry governments, development banksand donor agencies.Forging strategic partnerships andalliancesThe science and practice of agroforestry areinherently complex, integrating a range ofdisciplines, communities and institutions.Thus, partnerships and alliances are crucialto fostering an agroforestry transformation.

8World Agroforestry into the FutureOur partners include farmers and farmingcommunities, national and internationalresearch organizations, government agencies,development organizations, NGOs,advanced research laboratories and otherCGIAR centres. Successful managementof agroforestry research, developmentand educational programmes hinges onthe balance and coordination of availableexpertise to achieve synergy.In forging partnerships with such institutions,we are able to bring together andfocus a critical mass of relevant disciplinesand resources to design effective agroforestrystrategies, programmes and activities.We need to span the continuum from analysisof research needs through technologydevelopment, testing, adoption and implementationof agroforestry innovations. And,in doing so, tap into opportunities providedby institutions and organizations that haveknowledge, experience, mandates andresources that complement those of theCentre. We must continue to promote localparticipation in advancing agroforestryscience and practice, thereby incorporatingindigenous knowledge and expertise intoour work, and vigorously ensure the scalingup and long-term sustainability of agroforestryas a science and a practice.In conclusionIn the study by John Bene and his colleagues(1977) – upon which the creation of aglobal centre dedicated to agroforestry wasbased – it is written:“A new front should be opened on thewar against hunger, inadequate shelterand environmental degradation. Thiswar can be fought with weapons thathave been in the arsenal of rural peoplesince time immemorial, and noradical change in their lifestyle will berequired… Beyond question, agroforestrycan greatly improve life for peoplein the developing world, and do sowithin a reasonably short time.”A quarter-century ago, when this Centrewas created, the world was a very differentplace in many respects. Then, as now,about a billion people lived in desperatepoverty and there was food insecurity forhundreds of millions in the developingworld. But one thing that has changedsince then is that the global communityis many times wealthier than it was, andhas the benefit of the intervening decadesof development experience. Perhaps mostimportant of all it has united behind a setof definitive goals, with explicit measurabletargets to hold us accountable for solvingeach problem. They enable us to focustogether on what is really important in ourworld today: to eliminate hunger and desperatepoverty, the most complex, demandingproblems that the human communityfaces. Clearly, there is a tremendous responsibilitythat scientists and developmentprofessionals must bear. We have considerablefreedom in choosing the problems onwhich we work, including the very complexintellectual challenges that tax ourintelligence and creativity. This is a specialprivilege. The issue is what we do with thatprivilege, with that special opportunity tomake a difference. I only hope that in another25 years our efforts will be judged tohave been adequate to that task.ReferencesBene, J.G., H.W. Beall and A. Côté 1977. Trees,food and people: land management in thetropics. International Development ResearchCentre (IDRC), Ottawa, Canada.Dixon, J., A. Gulliver, D. Gibbon and M. Hall(eds) 2001. Farming Systems and Poverty:improving farmers’ livelihoods in a changingworld. Food and Agriculture Organizationof the United Nations, Rome, Italy.http://www.fao.org/documents/show_cdr.asp?url_file=//docrep/003/y1860e/y1860e00.htmBread for the World Institute 2003. Agriculturein the Global Economy. Bread for theWorld Institute, Washington, DC, USA.164 pp.Garrity, D.P. 2004. Agroforestry and the achievementof the Millennium DevelopmentGoals. Agroforestry Systems 61: 5–17.InterAcademy Council of Scientific Societies2004. Realizing the Promise and Potentialof African Agriculture. http://www.interacademycouncil.net/CMS/Reports/AfricanAgriculture/PDFs.aspx?returnID=6959UN Millennium Project 2005. Halving Hunger:It Can be Done. Task Force on HungerReport. http://www.unmillenniumproject.org/reports/tf_hunger.htmWorld Bank 2004. Sustaining Forest: A DevelopmentStrategy. World Bank, Washington,DC. Appendix 2 p A-3 http://www.powells.com/cgi-bin/biblio?show=Softcvr%20W%2FComp%20Media:Sale:0821357557:20.00World Agroforestry Centre 2005. Trees ofChange: A Vision for an AgroforestryTransformation in the Developing World.ICRAF, Nairobi, Kenya.

Trees and Markets

“The scaling up of participatorydomestication to tens of millionsof new households across thedeveloping world is probably thebiggest challenge for agroforestry,both in terms of the logistics oftraining and supervision, and inthe adaptation to new species,environments and markets.”Leakey et al.

Keywords:Dacryodes edulis, Irvingia gabonensis, Sclerocarya birrea,eco-agriculture, Green Revolution, domestication,agroforestry tree products (AFTPs)Chapter 2Trees and markets for agroforestry tree products:Targeting poverty reduction and enhanced livelihoodsRoger Leakey, James Cook University, Australia; Zac Tchoundjeu, World Agroforestry Centre, Cameroon;Kate Schreckenberg, Overseas Development Institute, UK; Tony Simons, World Agroforestry Centre, Kenya;Sheona Shackleton, Rhodes University, South Africa; Myles Mander, Institute of Natural Resources, SouthAfrica; Rachel Wynberg, University of Cape Town, South Africa; Charlie Shackleton, Rhodes University, SouthAfrica and Caroline Sullivan, Centre for Ecology and Hydrology, UKAbstractAgroforestry tree domestication as a farmer-driven, market-led process emerged as an internationalinitiative in the early 1990s, although a few studies pre-date this. A participatory approach now supplementsthe more traditional aspects of tree improvement, and is seen as an important strategy for meetingthe Millennium Development Goals of eradicating poverty and hunger and promoting social equity.Considerable progress towards the domestication of indigenous fruits and nuts has been achieved inmany villages in Cameroon and Nigeria that focuses on ‘ideotypes’, based on an understanding of thetree-to-tree variation in many commercially important traits. Vegetatively propagated cultivars are beingdeveloped by farmers for integration into their polycultural farming systems, especially cocoa agroforests.However, if agroforestry is to be adopted on a scale that has meaningful economic, social andenvironmental impacts, it is crucial that markets for agroforestry tree products (AFTPs) are expanded.Detailed studies of the commercialization of AFTPs, especially in southern Africa, provide support forthe wider acceptance of the role of indigenous tree domestication in the enhancement of livelihoodsfor poor farmers in the tropics. Consequently, policy guidelines are presented in support of this newapproach to sustainable rural development – an alternative to the biotechnology approaches beingpromoted by some development agencies.IntroductionIn the context of reducing poverty and enhancing thelivelihoods of poor smallholder farmers, this chapterpresents the evolution over the last 10–15 years of treedomestication strategies, approaches and techniquesaimed at promoting the cultivation of trees and thedevelopment of markets for agroforestry tree products(AFTPs). It relates the domestication of agroforestrytrees to the commercialization of their products andexamines the important role that markets play in theadoption of agroforestry and in the achievement ofsome of the Millennium Development Goals (MDGs).Finally, it suggests that the domestication and commercializationof AFTPs represents a rural developmentparadigm that is appropriate for wider implementationin developing countries.

12World Agroforestry into the FutureTreesThe origins of tropical treedomesticationThe domestication of many species for foodand other products has been carried outfor thousands of years in almost every partof the world, often arising from extractiveuses by indigenous people (Homma 1994).The concept of domesticating trees was firstpresented by Libby (1973), but at this timeit was focused on timber trees and was virtuallysynonymous with tree improvement,including the emerging clonal approaches.In 1992, a conference was held in Edinburgh,entitled ‘Domestication of TropicalTrees: The Rebuilding of Forest Resources’(Leakey and Newton 1994a; 1994b), whichembraced tree cultivation within the conceptof domestication. The recent interestin domestication is not restricted to treespecies; a range of new herbaceous cropsare also being studied (Smartt and Haq1997). Many indigenous vegetables arecandidates for domestication (Schippers2000) and can be components of multistratasystems, where there is a need fornew shade-tolerant crops.These days, the World Agroforestry Centre’stree domestication activities fall withina Trees and Markets research theme thatstresses the commercialization of AFTPs inthe overall poverty alleviation strategy ofthe Centre. While the focus of this chapteris on the development of marketable productsfrom agroforestry trees, interest in treedomestication encompasses trees for otherpurposes such as soil amelioration, fodder,fuelwood, timber, boundary demarcation,and so on.The aim of tree domesticationThe definition of tree domestication, establishedat the 1992 Edinburgh conference,encompasses the socioeconomic and biophysicalprocesses involved in the identificationand characterization of germplasmresources; the capture, selection and managementof genetic resources; and the regenerationand sustainable cultivation of thespecies in managed ecosystems (Leakey andNewton 1994a; 1994b). This concept hassubsequently been refined and expandedwith emphasis on it being a farmer-drivenand market-led process (Leakey and Simons1998; Simons 1996; Simons and Leakey2004) that takes a participatory approachto involve local communities (Leakey et al.2003; Tchoundjeu et al. 1998). Furthermore,not just species but also whole landscapescan be domesticated as a result of changingplant exploitation practices (Wiersum 1996).Since the mid-1990s, a growing number ofdonors have recognized the potential of treedomestication to achieving ICRAF’s visionand mandate for agroforestry to contributeto both poverty alleviation and the provisionof environmental services (see Figure 1). Therationale is that domestication and commercializationof indigenous trees throughagroforestry will provide an incentive forsubsistence farmers to plant trees in waysthat will reduce poverty and enhance foodand nutritional security, human health andenvironmental sustainability. In this way,agroforestry tree domestication is seenas an important component of strategiesto achieve the Millennium DevelopmentGoals (Garrity 2004; also see Chapter 1 thisvolume, especially goals relating to environmentalsustainability and food security(Goals 1 and 3–8: www.un.org/millenniumgoals).It is important to note, however, thatto bring about effective outcomes from thisresearch the messages have to be clearlyand widely disseminated to farmers, forestersand many relevant institutions.The genetic improvement of trees has usuallybeen the prerogative of national andinternational research institutes, but sincethe start of the Centre’s Tree DomesticationProgramme (a forerunner to its Trees andMarkets theme), the approach pursued hasbeen a participatory one.The Humid Lowlands of West and CentralAfrica (HULWA) was the first of theCentre’s regions to develop participatoryapproaches that went beyond simply collectinggermplasm. This started with thedevelopment of guidelines for speciespriority setting, derived by a partnership ofinternational and national scientists withfarmers and both non-governmental andcommunity-based organizations (Franzel etal. 1996). This project has since evolved inseveral ways:1. It now includes a range of differentspecies.2. It has used, disseminated and refined asimple low-technology system for thevegetative propagation of tropical trees,appropriate for use in small, low-costvillage nurseries (Leakey et al. 1990;Mbile et al. 2004; Shiembo et al. 1997).3. It has been quantitatively examining thetree-to-tree variation in a range of fruitand nut traits to determine the potentialfor highly productive and qualitativelysuperior cultivars (e.g. Anegbeh et al.2005; Atangana et al. 2002; Leakeyet al. 2005c; Ngo Mpeck et al. 2003;Waruhiu et al. 2004).4. Perhaps most importantly, it has beensuccessfully scaled-up to regional level(Tchoundjeu et al. 1998) and nowencompasses 40 villages in southernCameroon (about 2500 farmers) 11 villagesin Nigeria (2000 farmers), 3 villagesin Gabon (800 farmers) and 2 villages inEquatorial Guinea (500 farmers).Together these developments result in amodel participatory domestication strategy.This strategy is aligned with the United

Chapter 2: Trees and markets for agroforestry tree products13Nations Environment Programme’s Conventionon Biological Diversity (Leakeyet al. 2003; Simons and Leakey 2004;UnsustainablecroppingLoss of incomefrom wildlifeand plantsLoss of incomefrom cropsExternal factors:war, environmentaldisasters, etc.HIV/AIDSDeforestationTchoundjeu et al. 1998), by recognizing therights of local people to their indigenousknowledge and traditional use of nativePovertyEcosystem degradationand soil erosionLoss of biodiversityBreakdown of ecosystem functionLoss of crop yieldsMalnutritionDeclining livelihoodsOvergrazingBreakdown ofnutrient cyclingand loss of soilfertility/structureIncreasedtransport ofnitrates towater tableIncreasedhealth risksFigure 1. The cycle of biophysical and socioeconomic processes causing ecosystemdegradation, biodiversity loss, and the breakdown of ecosystem function, in agriculturalland in many tropical countries.plant species, and to benefit from commercialdevelopment of this knowledge.The extraction of fruits and medicinalproducts from indigenous trees by huntergatherersis a traditional practice in mosttropical regions (Sullivan 1999). It has alsobeen noted that farmers frequently maintainindigenous fruit and nut trees withintheir farming systems and sell the productslocally. Yet despite these observations, andthe findings of the species prioritizationprocess, international donors were initiallysceptical about investing in new crop species.This scepticism perhaps stemmedfrom a deep commitment to the GreenRevolution, coupled with a top-down approachto rural development in the tropics– still evident in the on-going pursuit ofnew biotechnological solutions (Lipton1999; McCalla and Brown 1999).As proof, a study of the frequency distributionpatterns of traditional species foundthat subsistence households are indeedcommitted to their traditional food species(Leakey et al. 2004). This conclusion hasbeen corroborated by the quantificationof the numbers of indigenous fruit treesin farmers’ fields in Cameroon, especiallyon small farms (Degrande et al. in press;Schreckenberg et al. 2002). In Benin, relativedensities of widely used species aretypically higher in farmers’ fields than inthe natural savanna vegetation because ofpreferential retention by farmers (Schreckenberg1999). Interestingly, evidence fromSouth Africa indicates that the yield ofmarula (Sclerocarya birrea), a traditionallyimportant indigenous fruit tree, is increasedby 5- to 15-fold through cultivation inhomestead plots and fields (Shackleton etal. 2003a). Mean fruit size is also greaterfrom trees in these plots, again with someevidence for domestication by farmers(Leakey 2005; Leakey et al 2005a; 2005b).

14World Agroforestry into the FutureIdentification, capture, retention andprotection of genetic diversityDomestication has been defined as humaninducedchange in the genetics of a speciesto conform to human desires and agroecosystems(Harlan 1975). It is not surprisingtherefore, that much of the work to domesticateagroforestry trees has focused on boththe identification of intraspecific genetic variabilityof the priority species and the vegetativepropagation techniques to capturethese superior combinations. However, onedesirable trait is not necessarily correlatedwith another: thus large fruits are not necessarilysweet fruits, and do not necessarilycontain large nuts or kernels. This multitraitvariation, coupled with the variability ofeach individual trait, results in a considerableopportunity for selection of trees withgood combinations of traits, but also makesit more unlikely that an ideal tree will befound. Thus, large numbers of trees have tobe screened to find the rare combinations oftraits. This rapidly becomes impractical andvery expensive. Consequently, the practicalapproach is to search for trees that have particularmarket-oriented trait combinations(or ideotypes) – such as big, sweet fruits forthe fresh fruit market (a fruit ideotype) orbig, easily extracted kernels for the kernelmarket (kernel ideotype), etc.Trees can also be selected for production traitssuch as yield, seasonality and regularity of production,reproductive biology, and reductionof susceptibility to pests and diseases (Kengueet al. 2002). High yield is obviously a desirabletrait in any cultivar, but, within reason, may notbe as important in the early stages of domesticationas the quality attributes. Fruiting seasontime/length, ripening period and seedlessnessare other important variables that could be selectedfor (Anegbeh et al. 2005).Such great intraspecific genetic diversityneeds to be preserved. Domestication isgenerally considered to reduce geneticdiversity, a situation that may occur wherethe domesticated plant replaces or dominatesthe wild origin, but is probably notthe case at the current level of domesticationof agroforestry trees. For example, therange of fruit sizes in on-farm populationsof Dacryodes edulis and Irvingia gabonensishas been increased by the early stagesof domestication (Leakey et al. 2004).Nevertheless, the maintenance of geneticdiversity is essential. Modern moleculartechniques can identify the ‘hot-spots’ ofintraspecific diversity (Lowe et al. 2000),which should, if possible, be protectedfor in situ genetic conservation, or be thesource of germplasm collections if ex situconservation is required. In addition, whendeveloping cultivars, they should originatefrom unrelated populations with very differentgenetic structures.Having identified the superior trees withthe desired traits, the capture of tree-to-treevariation using techniques of vegetativepropagation is relatively simple and wellunderstood (Leakey 2004b; Leakey et al.1996; Mudge and Brennan 1999). Cuttingsfrom mature trees have a low rate of propagativesuccess, and the number of peoplewith the appropriate skills to carry it outmay be a constraint to its widespread applicationin the future (Simons and Leakey2004). However, propagation by juvenileleafy cuttings is very easy for almost all treespecies and is currently the preferred optionfor participatory domestication in villagenurseries (Mbile et al. 2004; Mialoundamaet al. 2002; Shiembo et al. 1996;Tchoundjeu et al. 2002b).Cultivation and the growth of cultivarsThe final stage of the domestication processis the optimal integration of selected plantsinto the farming system (Leakey and Newton1994a; 1994b). In African farmland, awide range of densities and configurationsare grown (Kindt 2002). In Cameroon, forexample, cocoa agroforests have beenreported to contain around 500 cocoabushes growing with 15 other types of treesand shrubs (Gockowski and Dury 1999).Agroforestry is expected to provide positiveenvironmental benefits on climate changeand biodiversity (Millennium DevelopmentGoal 7). However, research is needed todetermine the impacts of such diversity onagroecosystem function (Gliessman 1998;Leakey 1999b; Mbile et al. 2003); carbonsequestration (Gockowski et al. 2001) andtrace gas fluxes; and on the sustainabilityof production and household livelihoods.MarketsThe term agroforestry tree products (AFTPs)is of very recent origin (Simons and Leakey2004) and refers to timber and non-timberforest products (NTFPs) that are sourcedfrom trees cultivated outside of forests, todistinguish them from NTFPs extractedfrom natural systems. However, someproducts will be marketed as both NTFPsand AFTPs during the period of transitionfrom wild resources to newly domesticatedcrops. Consequently, both terms are usedin the following sections.Economic and social benefits fromtrading AFTPsTo be effective, there must be a link betweentree domestication and productcommercialization, which requires theinvolvement of food, pharmaceutical andother industries in the identification of thecharacteristics that will determine marketacceptability (Leakey 1999a). In West andCentral Africa, a number of indigenousfruits and nuts, mostly gathered from farmtrees, contribute to regional trade (Ndoyeet al. 1997). In Cameroon, the annual tradein products from five key species has been

Chapter 2: Trees and markets for agroforestry tree products15valued at US$7.5 million, including exportsworth US$2.5 million (Awono et al.2002). Women are often the beneficiariesof this trade; they have especially indicatedtheir interest in marketing D. edulis fruitsbecause the fruiting season coincideswith the time to pay school fees and tobuy school uniforms (Schreckenberg etal. 2002). It is also the women who arethe main retailers of NTFPs (Awono et al.2002). Marula (Scleocarya birrea) is anotherfruit with a harvesting season thatcoincides with the start of the school year,and therefore the greater involvement ofwomen.These tangible market benefits are supplementedby additional benefits such as theavailability of products for domestic consumption,the use of household labour forharvesting/processing free of charge, andease of access to informal markets, etc. Becausethe production and trading of AFTPsare based on traditional lifestyles, it is relativelyeasy for new producers to enter withminimal skills, little capital and with fewneeds for external inputs. Together thesemake this approach to intensifying productionand enhancing household livelihoodsvery easy, and adoptable by poor people.The linkages between domesticationand commercialization of AFTPsAs already indicated, domestication thatis market-orientated has the greatest likelihoodof being adopted on a scale thathas impact on the economic, social andenvironmental problems afflicting manytropical countries. This requires that agroforesterswork closely with the companiesprocessing and marketing the products(Leakey 1999a). However, in doing this itis important to remember that smallholderfarmers are the clients of the research anddevelopment (R&D) work and that thereneeds to be a functional production–to–consumption chain; principles that wereapparently forgotten during recent domesticationof peach palm (Bactris gasipaes) inAmazonia, resulting in the underperformanceof the market (Clement et al. 2004).In many cases, the successful commercializationof AFTPs relies on domestication toensure that supply can keep up with thegrowing demand of a developing market.Through cultivar development, domesticationcan also help to overcome anotherconstraint to commercialization: variabilityof quality (taste, size and purity). Domesticationcan also lead to an extended seasonof production, as is being done in WestAfrica with D. edulis, making it easier tosupply industries throughout the year. Kiwifruit (Actinidia chinensis) and macadamianuts (Macadamia integrifolia) are good examplesof co-ordinated domestication andcommercialization.The important question here is whetheragroforestry can prevent the negative impactsthat result from domesticating cropsin a monoculture system, which can causeenvironmental degradation through deforestation,soil erosion, nutrient mining andloss of biodiversity. These systems can alsoresult in social inequity and the ‘povertytrap’ for small-scale producers who are unableto compete in international trade withlarge or multinational companies. In theory,agroforestry is beneficial to the environmentand beneficial to the poor farmer.However, if the domestication of AFTPs isso successful that the market demand forone of them reaches the point where monocultureplantations, either in the countryof origin or in some overseas location, areviable, this could undermine the wholepurpose of developing new crops. Withoutmarkets there will not be the opportunityfor subsistence households to increase theirstandard of living, while expanded marketopportunities could lead to their exploitationby unscrupulous entrepreneurs. Havingsaid that, recognizing the traditionalrole of NTFPs/AFTPs in food security,health and income generation, it is clearthat the potential benefits from domesticationoutweigh the risks, and that commercializationis both necessary and potentiallyharmful to small-scale farmers practisingagroforestry (Leakey and Izac 1996).Important areas for further study includethe complex issues surrounding commercializationof genetic resources and benefitsharing (ten Kate and Laird 1999) and traditionalknowledge (Laird 2002) and waysin which smallholder farmers can securetheir intellectual property rights on farmerderivedinnovations.One strategy that reduces risk is to domesticatea wide range of AFTP tree species,especially those with local and regionalmarket potential. In this way, coupled withstrong indigenous rights, it is very unlikelythat the market demand will attract majorcompanies and, even if products of a fewspecies do become international commodities,there will be others that remain.Not all interest from international companiesin agroforestry is unwelcome. Forexample, Daimler-Benz has taken a smallholder,multistrata agroforestry approachto producing raw materials for its C-classMercedes-Benz cars in Brazil, and inpartnership with the International FinanceCorporation has been developing this asa new paradigm for public–private sectorpartnerships (Mitschein and Miranda 1998;Panik 1998). Smallholder cocoa farmers inAfrica and Asia are supported by chocolatierMasterfoods (formerly M&M Mars) asthey diversify their cocoa farms into cocoaagroforests, integrating fruit trees (oftenindigenous species) into the cocoa farm

16World Agroforestry into the Futureso that the shade trees are also companioncrops (Leakey and Tchoundjeu 2001). Thishas been done as a risk-aversion strategyto provide new sources of income, in responseto fluctuating market prices. Interestingly,cocoa is not the only former plantationcash crop to now be an importantagroforestry species. Rubber is perhaps thebest example, especially in Southeast Asia(Tomich et al. 2001), while tea and coffeeare moving in the same direction.A somewhat different but interesting exampleof AFTP commercialization is thecase of marula, a tree of dry Africa, whichis starting to be marketed by subsistencefarmers for traditional beer and for industrialprocessing as an internationallymarketed liqueur, ‘Amarula’, by DistellCorporation. Marula kernel oil (‘Maruline’)is also breaking into international cosmeticsmarkets. This species thus provides anopportunity to examine the impact of differentcommercialization strategies on thelivelihoods of the producers, the sustainabilityof the resource and the economicand social institutions. In other words, whoor what are the winners and losers arisingfrom the commercialization of indigenousfruits and nuts?Winners and losers: impacts onlivelihoodsThe Centre for Ecology and Hydrology,UK,in collaboration with a wide range of institutions,conducted a large, multidisciplinary,multi-institutional study to determinethe ‘winners and losers’ of the various commercializationstrategies for a number ofdifferent NTFP products from two tree species(S. birrea and Carapa guianensis) in differentenvironments and in structurally andethnically different communities (Shackletonet al. 2003a; Sullivan and O’Regan2003). The study specifically examined theeffects of commercialization on the fiveforms of livelihood capital (human, social,financial, natural and physical). In brief,the authors concluded that to improve thelivelihood benefits from commercializingNTFPs it is important to improve:• The quality and yield of the productsthrough: domestication and the disseminationof germplasm; and enhancing theefficiency of post-harvest technology (extraction,processing, storage, and so on).• The marketing and commercializationprocesses by: diversifying markets forexisting and new products; investing inmarketing initiatives and campaigns;and promoting the equitable distributionof benefits.The following lessons were learnt forNTFP commercialization from the study ofS. birrea (abridged from Shackleton et al.2003b), that apply equally to AFTPs:• NTFPs are most important for poor andmarginalized people.• NTFPs make up income shortfalls butdo not significantly alleviate poverty.How domestication may change this stillneeds to be determined.• Engagement in NTFP commercializationand the extent of benefits is variableeven among the poorest households.• Benefits of NTFP commercializationmust be weighed against the negativesocial and cultural costs of commercialization.• Land and usufruct rights must be clear,government intervention must bepitched at the appropriate level, andpolitical support for the NTFP industrymust be secured.• NTFP commercialization can lead to improvedmanagement and conservationof the resource in certain circumstances.• NTFP cultivation needs to be community-ownedand driven.• Benefits can be accrued at the local level.• Intellectual property right (IPR) systemsthat promote poverty alleviation, foodsecurity and sustainable agriculture areurgently needed.• Models of commercialization based onpartnerships between producer communities,non-governmental organizations(NGOs) and the private sector are mostlikely to succeed.• The diversification of species used,products produced, markets traded, andplayers involved, is an extremely importantstrategy to minimize the risks ofNTFP commercialization for rural communities.• Scaling up and introducing new technologiescan shift benefits away fromwomen and the most marginalized producers.• NTFPs form only part of a far broaderecological, economic, social and politicallandscape. For example, continuedland clearance, the need for biomass energy,and wood for woodcarvings can bea greater threat than the commercializationof a fruit product.• NTFP trade and industries are dynamicin space and time. There are seldom permanentwinners and losers.The conclusion from this study was thatNTFP commercialization can create bothwinners and losers, but positive outcomescan be maximized if external players promotecommunity involvement, and if thecommunities themselves work together anduse their own strengths to manage and usetheir resources effectively. This is supportedby the findings of a study investigating therole of tree domestication in poverty alleviation(Poulton and Poole 2001). Nevertheless,to ensure that those engaged inparticipatory domestication are winners,the current difficulties facing farmers

Chapter 2: Trees and markets for agroforestry tree products17wishing to protect their rights to their cultivarsneed to be resolved.Policy guidelinesInevitably, in a new research area such asthis, many questions remain unanswered;indeed they cannot be answered until thetechniques and strategies outlined abovehave been in use for longer periods and onlarger scales. Nevertheless, there seems tobe growing confidence on the part of institutionslike ICRAF, and their donors, thatthis approach to agroforestry and the alleviationof poverty has merit. This is emphasizedby suggestions that these conceptshave a role to play in the achievement ofseveral of the Millennium DevelopmentGoals (Garrity 2004).One clear policy message is that it is importantto recognize the ‘chicken and egg’relationship between domestication andcommercialization (Leakey and Izac 1996)– and the folly of doing one without theother. However, it is clear that the relationshipbetween domestication and commercializationis delicately balanced. Both thelack of a market and the excessive growthof a market pose a threat. Sound policyinterventions will probably be needed toensure that smallholder subsistence farmersare the beneficiaries of the domesticationof AFTPs. Policy makers tend not to thinkmuch about the differences between a monoculturalapproach to growing a new cropversus an agroforestry approach. However,in the extreme 20 million trees can eitherbe grown by four farmers planting 5 millionapiece, or by 1 million farmers eachgrowing only 20 trees. Each scenario willlikely have very different social and economicoutcomes.Desirable policy interventions (fromTchoundjeu et al. 2004; Ndoye et al. 2004;Wynberg et al. 2003) may be to:• Promote the participatory domesticationof tree species fitting a variety of onfarmniches.• Focus domestication activities on thecapture and use of intraspecific variationexisting in wild/semi-domesticatedpopulations and utilize the relativelyquick economic and social returns fromparticipatory domestication.• Promote local-level processing andmarketing of indigenous fruits, nuts andother tree products in parallel with domestication.• Recognize the considerable training andextension needs of rural communitiesthat are required to achieve the scalingup necessary to meet the MillenniumDevelopment Goals.• Clarify land and usufruct rights to facilitatethe successful and effective commercialdevelopment of AFTPs, recognizingthat Western approaches may notbe appropriate for indigenous resourcetenure systems.• Develop and implement systems to protectcommunity-based cultivars (throughparticipatory domestication) as partof legislative reforms for biodiversitymanagement, indigenous knowledgeprotection, and plant genetic resourceconservation and use.• Ensure the continued use of a widerange of NTFPs to support rural livelihoods.• Establish basic management, financialand institutional capacities to ensurethat local people capture a greater shareof the benefits from commercialisation.Features of this approach torural developmentAlthough this chapter has focused on thereduction of poverty and the enhancementof smallholder livelihoods, the problemsof poverty, land degradation, loss of biodiversity,social deprivation, malnutrition,hunger, poor health and declining livelihoodsare all inextricably linked and cyclical(Figure 1). Consequently any attemptsto alleviate the problems have to target anumber of different points within the cycle.Agroforestry is advocated as one of manyFigure 2. The relationship between the two functions of agroforestry trees and theirpotential to mitigate global problems arising from unsustainable land use.Source: Leakey and Tomich (1999).Tree servicesGlobal challenges:• Improved human welfare• Alleviation of poverty• Reduced environmentaldegradationEnvironmental resilienceAgroforestryTree products(AFTPs)Increased income forpoverty alleviationFood andnutritional security

18World Agroforestry into the Futuremeans of meeting these globalchallenges (Figure 2).Domestication for improvements in quality and yieldThe potential of this approachof course comes with somerisks (Figure 3). Furthermore,the domestication of AFTPsmay reduce the market-shareof wild-collected NTFPs,thereby disadvantaging landlessrural people. However,the number of people benefitingfrom this domesticationprobably greatly outweighsthose who are disadvantaged.A number of studies implythat the income from AFTPscan contribute to meeting theMillennium DevelopmentGoal of halving the numberof people living on less thanUS$1 per day. For example, inCameroon, studies of farmersgrowing indigenous fruits havefound that the net presentvalue per hectare of cocoa isabout US$500 greater whengrown with indigenous fruitsthan when grown without(Gockowski and Dury 1999).To these benefits can alsobe added the AFTP productsused in domestic consumption,which represent a savingon expenditure, and the cashearned from selling AFTPs thatmay be reinvested in the farmsin the form of new and betterinputs. It is clear therefore,that it is difficult to evaluatethe total benefits obtainedfrom marketable AFTPs.Thus the challenge posed bythe Millennium DevelopmentProtection offarmers’ IPR’sIncreased incomeTarget: povertyImproved well-being:• Increased social benefits• Improved nutrition and healthTarget: livelihoodsSupporting environmentalbenefits (= agroforestry)Supporting tradition,culture and social needsDiversified and increased biological resourceTarget: biodiversityMarket expansionIncreased environmental services and ecosystem functionTarget: ‘environmental degradation’Wise use of genetic diversityIncreased market value/appealIncentive to plant treesIncreased scaleof tree plantingSUSTAINABLE RURAL DEVELOPMENTFigure 3. Potential impacts on sustainability of domesticating agroforestry trees.Loss of geneticdiversityDANGERDetrimental to tradition and cultureDANGERDANGERPromotion of largescalemonoculturesDANGER

Chapter 2: Trees and markets for agroforestry tree products19Goals is not so much how to find a way toachieve them at the household level, butmuch more how to scale up AFTP productionbetween now and 2015 to reach themillions of poor rural families (60 millionin HULWA alone) for whom AFTPs mightprovide a step out of poverty. The AFTPapproach could be thought of as a ‘reallygreen revolution’ (Leakey 2001).Development issues for thefutureThe scaling up of participatory domesticationto tens of millions of new householdsacross the developing world is probablythe biggest challenge for agroforestry,both in terms of the logistics of trainingand supervision, and in the adaptation tonew species, environments and markets.Techniques including vegetative propagation,and the acquisition and protection of‘community plant breeders rights’ on thecultivars created by communities, are alsoareas where urgent action is needed. Failingto achieve this will discourage villagersfrom investing their time, effort and limitedresources in a venture that could be takenaway from them. Policy makers should realizethat participatory domestication thatenables community rights to be protectedand realized represents a new and acceptableapproach to biodiscovery – the antithesisof biopiracy.As demand grows, markets will start tobe more interested in quality rather thanquantity. This will require refinements inthe ideotypes for each particular market;necessitating, in turn, better marketinformation than is currently available.Therefore, to avoid the potential pitfalls ofdomestication (Figure 3), strategies such asdeliberate retention of intraspecific variationfor pest and disease resistance, etc.will be important (Leakey 1991). In addition,as commercial interests increase, itwill be important to maintain a focus ondiversified agroforestry production thatshould promote integrated pest management(Leakey 1999b).Around the world, agricultural R&D institutionsmust be helped to develop new skillsin the domestication of indigenous species,the processing/storage of their products,market analysis and in developing marketlinkages (Garrity 2004). This level of expansionwill also require high-level policysupport to ensure a coordinated and coherentapproach to the domestication andcommercialization of AFTPs.ConclusionsIn the 9 years since agroforestry treedomestication was institutionalized at theCentre, great progress has been made.This review has focused on progress in thehumid zone of West and Central Africaand in Southern Africa, but similar programmesare in progress in the Sahel, EastAfrica, Amazonia and Southeast Asia, aswell as outside the Centre. Hopefully, theexperiences reported here for agroforestrybased on locally relevant tree species andmarkets will be of great benefit to otherareas of the world embarking on similarpeople-centred concepts for rural development.We suggest that this approach offersa viable alternative to biotechnology-basedadvances in agricultural science for developingcountries.ReferencesAnegbeh, P.O., V. Ukafor, C. Usoro, Z. Tchoundjeu,R.R.B. Leakey and K. Schreckenberg2005. Domestication of Dacryodes edulis:1. Phenotypic variation of fruit traits from100 trees in southeast Nigeria. New Forests29: 149–160.Atangana, A.R., V. Ukafor, P.O. Anegbeh,E. Asaah, Z. Tchoundjeu, C. Usoro,J-M. Fondoun, M. Ndoumbe and R.R.B.Leakey 2002. Domestication of Irvingiagabonensis: 2. The selection of multipletraits for potential cultivars from Cameroonand Nigeria. Agroforestry Systems55: 221–229.Awono, A., O. Ndoye, K. Schreckenberg, H.Tabuna, F. Isseri and L. Temple 2002. Productionand marketing of Safou (Dacryodesedulis) in Cameroon and internationally:Market development issues. Forest, Treesand Livelihoods 12: 125–147.Clement, C.R., J.C. Weber, J. van Leeuwen,C.A. Domian, D.M. Cole, L.A.A. Lopezand H. Argüello 2004. Why extensive researchand development did not promoteuse of peach palm fruit in Latin America.Agroforestry Systems 61: 195–206.Degrande, A., K. Schreckenberg, C. Mbosso, P.O.Anegbeh, J. Okafor, J. Kanmegne and M.Trivedi (In press). Driving forces behindlevels of fruit tree planting and retentionson farms in the humid forest zone of Cameroonand Nigeria. Agroforestry SystemsFranzel, S., H. Jaenicke and W. Janssen 1996.Choosing the right trees: setting prioritiesfor multipurpose tree improvement. ISNARResearch Report 8.Garrity, D. 2004. World agroforestry and theachievement of the Millennium DevelopmentGoals. Agroforestry Systems 61: 5–17.Gliessman, S.R. 1998. Agroecology: EcologicalProcesses in Sustainable Agriculture. AnnArbor Press, Chelsea, USA.Gockowski, J.J. and S. Dury 1999. The economicsof cocoa-fruit agroforests in southernCameroon. Pp. 239–241, in: F. Jiménezand J. Beer (eds), Multi-strata AgroforestrySystems with Perennial Crops. Centro

20World Agroforestry into the FutureAgronómico Tropical de Investigación yEnseñanza, Turrialba, Costa Rica.Gockowski, J., G. Blaise Nkamleu and J. Wendt2001. Implications of resource-use intensificationfor the environment and sustainabletechnology in the Central Africanrainforest. Pp. 197–219, in: D.R. Lee andC.B. Barrett (eds) Trade-offs or Synergies?Agricultural Intensification, EconomicDevelopment and the Environment, CABInternational, Wallingford, UK.Harlan, J.R. 1975. Crops and Man, AmericanSociety of Agronomy/Crop Science Societyof America, Madison, USA.Homma, A.K.O. 1994. Plant extractivism in theAmazon: Limitations and possibilities.Pp. 34–57, in: M. Clusener-Godt and I.Sachs (eds) Extractivism in the BrazilianAmazon: Perspectives on Regional Development,MAB Digest 18, Man and theBiosphere, United Nations Educational,Scientific and Cultural Organization,Paris, France.Kengue, J., F.N. Tchuenguem Fohouo and H.G.Adewusi 2002. Towards the improvementof Safou (Dacryodes edulis): Populationvariation and reproductive biology. Forests,Trees and Livelihoods 11: 73–84.Kindt, R. 2002. Methodology for tree speciesdiversification planning for African ecosystems.PhD thesis, University of Ghent,Belgium.Laird, S.A. 2002. Biodiversity and TraditionalKnowledge: Equitable Partnerships inPractice. People and Plants ConservationSeries, Earthscan, London, UK.Leakey, R.R.B. 1991. Towards a strategy forclonal forestry: Some guidelines basedon experience with tropical trees. Pp.27–42, in: J.E. Jackson (ed) Tree Breedingand Improvement. Royal Forestry Societyof England, Wales and Northern Ireland,Tring, UK.Leakey, R.R.B. 1999a. Potential for novel foodproducts from agroforestry trees, FoodChemistry 64: 1–14.Leakey, R.R.B. 1999b. Agroforestry for biodiversityin farming systems. Pp. 127–145, in:W.W. Collins and C.O. Qualset (eds) Biodiversityin Agroecosystems. CRC Press,New York, USA.Leakey, R.R.B. 2001. Win:Win landuse strategiesfor Africa: 2. Capturing economic andenvironmental benefits with multistrataagroforests, International Forestry Review3: 11–18.Leakey, R.R.B. 2005. Domestication potential ofMarula (Sclerocarya birrea subsp caffra) inSouth Africa and Namibia: 3. Multi-traitselection. Agroforestry Systems 64: 51–59.Leakey, R.R.B. and A-M. Izac 1996. Linkagesbetween domestication and commercialisationof non-timber forest products:Implications for agroforestry. Pp. 1–7, in:R.R.B. Leakey, A.B. Temu and M. Melnyk(eds) Domestication and Commercialisationof Non-timber Forest Products, Non-Wood Forest Products No. 9, Food andAgriculture Organization of the UnitedNations, Rome, Italy.Leakey, R.R.B. and A.C. Newton 1994a. Domesticationof ‘Cinderella’ species as thestart of a woody-plant revolution. Pp. 3–4,in: R.R.B. Leakey and A.C. Newton (eds)Tropical Trees: The Potential for Domesticationand the Rebuilding of Forest Resources,Her Majesty’s Stationery Office,London, UK.Leakey, R.R.B. and A.C. Newton 1994b. Domesticationof Tropical Trees for Timber andNon-Timber Forest Products. MAB Digest17, Man and the Biosphere, United NationsEducational, Scientific and CulturalOrganization, Paris, France.Leakey, R.R.B. and A.J. Simons 1998. Thedomestication and commercialisation ofindigenous trees in agroforestry for thealleviation of poverty. Agroforestry Systems38: 165–176.Leakey, R.R.B. and T.P. Tomich 1999. Domesticationof tropical trees: from biology toeconomics and policy. Pp. 319–338, in:L.E. Buck, J.P. Lassoie and E.C.M. Fernandes(eds) Agroforestry in SustainableEcosystems, CRC Press/Lewis Publishers,New York, USA.Leakey, R.R.B. and Z. Tchoundjeu 2001. Diversificationof tree crops: Domestication ofcompanion crops for poverty reductionand environmental services. ExperimentalAgriculture 37: 279–296.Leakey, R.R.B., J.F. Mesén, Z. Tchoundjeu, K.A.Longman, J. McP. Dick, A.C. Newton,A. Matin, J. Grace, R.C. Munro and P.N.Muthoka 1990. Low-technology techniquesfor the vegetative propagation oftropical tress. Commonwealth ForestryReview 69: 247–257.Leakey, R.R.B., A.B. Temu, M. Melnyk andP. Vantomme (eds) 1996. Domesticationand Commercialization of Non-TimberForest Products for Agroforestry. Non-Wood Forest Products No 9. Food andAgriculture Organization of the UnitedNations, Rome, Italy.Leakey, R.R.B., A.R. Atangana, E. Kengni, A.N.Waruhiu, C. Usuro, P.O. Anegbeh andZ. Tchoundjeu 2002. Domestication ofDacryodes edulis in West and Central Africa:Characterisation of genetic variation.Forests, Trees and Livelihoods 12: 57–72.Leakey, R.R.B., K. Schreckenberg and Z.Tchoundjeu 2003. The participatory domesticationof West African indigenousfruits. International Forestry Review 5:338–347.Leakey, R.R.B., Z. Tchoundjeu, R.I. Smith, R.C.Munro, J-M. Fondoun, J. Kengue, P.O.Anegbeh, A.R. Atangana, A.N. Waruhiu,E. Asaah, C. Usoro and V. Ukafor 2004.Evidence that subsistence farmers havedomesticated indigenous fruits (Dacryodesedulis and Irvingia gabonensis) inCameroon and Nigeria. Agroforestry Systems60: 101–111.Leakey, R.R.B., S. Shackleton and P. du Plessis2005a. Domestication potential ofMarula (Sclerocarya birrea subsp caffra) in

Chapter 2: Trees and markets for agroforestry tree products21South Africa and Namibia: 1. Phenotypicvariation in fruit traits. AgroforestrySystems 64: 25–35.Leakey, R.R.B., K. Pate and C. Lombard 2005b.Domestication potential of Marula(Sclerocarya birrea subsp caffra) in SouthAfrica and Namibia: 2. Phenotypic variationin nut and kernel traits. AgroforestrySystems 64: 37–49.Leakey, R.R.B., P. Greenwell, M.N. Hall, A.R.Atangana, C. Usoro, P.O. Anegbeh,J-M. Fondoun and Z. Tchoundjeu 2005c.Domestication of Irvingia gabonensis: 4.Tree-to-tree variation in food-thickeningproperties and in fat and protein contentsof Dika Nut. Food Chemistry 90:365–378.Libby, W.J. 1973. Domestication strategies forforest trees. Canadian Journal of ForestResearch 3: 265–276.Lipton, M. 1999. Reviving global poverty reduction:What role for genetically modifiedplants? Sir John Crawford Memorial Lectureat CGIAR Centers Week, Washington,DC (October 28 th 1999), ConsultativeGroup for International AgriculturalResearch Secretariat, The World Bank,Washington DC, USA.Lowe, A.J., A.C.M. Gillies, J. Wilson and I.K.Dawson 2000. Conservation genetics ofbush mango from central/west Africa:implications from random amplifiedpolymorphic DNA analysis. MolecularEcology 9: 831–841.Mbile, P., Z. Tchoundjeu, A. Degrande, E.Asaah and R. Nkuinkeu 2003. Mappingthe biodiversity of ‘Cinderella’ trees inCameroon. Biodiversity 4: 17–21.Mbile P., Z. Tchoundjeu, A. Degrande, M-L.Avana and C. Tsobeng 2004. Non-mistvegetative propagation by resourcepoor,rural farmers of the forest zone ofCameroon: Some technology adaptationsto enhance practice. Forest, Trees andLivelihoods 14: 43–52Mbofung, C.M.F., T. Silou and I. Mouragadja2002. Chemical characterisation of Safou(Dacryodes edulis) and evaluation of itspotential as an ingredient in nutritiousbiscuits. Forests, Trees and Livelihoods 12:105–118.McCalla, A.F. and L.R. Brown 1999. Feedingthe developing world in the next Millennium:A question of science? Proceedingsof Conference on Ensuring Food Security,Protecting the Environment, ReducingPoverty in Developing Countries. Can BiotechnologyHelp? 21–22 October 1999,World Bank, Washington DC, USA.Mialoundama, F., M-L Avana, E. Youmbi, P.C.Mampouya, Z. Tchoundjeu, M. Mbeuyo,G.R. Galamo, J.M. Bell, F. Kopguep, A.C.Tsobeng and J. Abega 2002. Vegetativepropagation of Dacryodes edulis (G. Don)H.J. Lam by marcots, cuttings and micropropagation.Forests, Trees and Livelihoods12: 85–96.Mitschein, T.A. and P.S. Miranda 1998. POEMA:a proposal for sustainable development inAmazonia. Pp. 329–366, in: D.E. Leihnerand T.A. Mitschein (eds) A Third Millenniumfor Humanity? The Search for Pathsof Sustainable Development. Peter Lang,Frankfurt am Main, Germany.Mudge, K.W. and E.B. Brennan 1999. Clonalpropagation of multipurpose and fruittrees used in agroforestry. Pp. 157–190,in: L.E. Buck, J.P. Lassoie and E.C.M. Fernandes(eds) Agroforestry in SustainableEcosystems. CRC Press/Lewis Publishers,New York, USA.Ndoye, O., M. Ruiz-Perez and A. Ayebe 1997.The markets of non-timber forest productsin the humid forest zone of Cameroon.Rural Development Forestry Network,Network Paper 22c, Overseas DevelopmentInstitute, London, UK.Ndoye, O., A. Awono, K. Schreckenberg andR.R.B. Leakey 2004. Commercialisingindigenous fruit for poverty alleviation:A policy briefing note for governments inthe African humid tropics region. OverseasDevelopment Institute, London, UK.Ngo Mpeck, M.L., E. Asaah, Z. Tchoundjeu andA.R. Atangana 2003. Strategies for the domesticationof Ricinodendron heudelotii:Evaluation of variability in natural populationsfrom Cameroon. Agriculture andEnvironment 1: 257–262.Panik, F. 1998. The use of biodiversity and implicationsfor industrial production. Pp.59–73, in: D.E. Leihner and T.A. Mitschein(eds) A Third Millennium for Humanity?The Search for Paths of Sustainable Development.Peter Lang, Frankfurt am Main,Germany.Poulton, C. and N. Poole 2001. Poverty andFruit Tree Research: Issues and OptionsPaper, www.nrinternational.co.uk/forms2/frpzf0141b.pdf, Department for InternationalDevelopment, Forestry Research Programme,Wye College, Ashford, UK.Schippers, R.R. 2000. African Indigenous Vegetables:An Overview of the CultivatedSpecies, Natural Resources Institute/AfricaCaribbean and Pacific – European UnionTechnical Centre for Agricultural and RuralCooperation, Greenwich, London, UK.Schreckenberg, K. 1999. Products of a managedlandscape: Non-timber forest products inthe parklands of the Bassila Region, Benin.Global Ecology and Biogeography 8:279–289.Schreckenberg, K, A. Degrande, C. Mbosso,Z. Boli Baboulé, C. Boyd, L. Enyong, J.Kanmegne and C. Ngong 2002. The socialand economic importance of Dacryodesedulis (G.Don) H.J. Lam in southernCameroon. Forests, Trees and Livelihoods12: 15–40.Shackleton C.M., J. Botha and P.L. Emanuel2003a. Productivity and abundance ofSclerocarya birrea subsp. caffra in andaround rural settlements and protectedareas of the Bushbuckridge lowveld, South

22World Agroforestry into the FutureAfrica. Forests, Trees and Livelihoods 13:217–232.Shackleton, S.E., R.P. Wynberg, C.A. Sullivan,C.M. Shackleton, R.R.B. Leakey, M.Mander, T. McHardy, S. den Adel, A. Botelle,P. du Plessis, C. Lombard, S.A. Laird,A.B. Cunningham, A. Combrinck and D.P.O’Regan 2003b. Marula commercialisationfor sustainable and equitable livelihoods:Synthesis of a southern African casestudy. In: Winners and Losers in ForestProduct Commercialization, Final TechnicalReport to Department for InternationalDevelopment, Forestry Research ProgrammeR7795, Appendix 3.5 Centrefor Ecology and Hydrology, Wallingford,England, UK.Shiembo, P.N., A.C. Newton and R.R.B. Leakey1996. Vegetative propagation of Irvingiagabonensis Baill., a West African fruittree. Forest Ecology and Management 87:185–192.Shiembo, P.N., A.C. Newton and R.R.B. Leakey1997. Vegetative propagation of Ricinodendronheulelotii (Baill) Pierre ex Pax, aWest African fruit tree. Journal of TropicalForest Science 9: 514–525.Simons, A.J. 1996. ICRAF’s strategy for domesticationof indigenous tree species. Pp.8–22, in: R.R.B. Leakey, A.B. Temu, M.Melnyk and P. Vantomme (eds) Domesticationand Commercialization of Non-TimberForest Products in Agroforestry Systems,Non-Wood Forest Products 9. Food andAgriculture Organization of the United Nations,Rome, Italy.Simons, A.J. and R.R.B. Leakey 2004. Tree domesticationin tropical agroforestry. AgroforestrySystems 61: 167–181.Smartt, J. and N. Haq 1997. Domestication,Production and Utilization of New Crops.International Centre for UnderutilizedCrops, Southampton, UK.Sullivan, C.A. 1999. Linking the past with thefuture: Maintaining livelihood strategiesfor indigenous forest dwellers in Guyana.In: Redclift, M. (ed) Sustainability? LifeChances and Livelihoods. Routledge,London, UK.Sullivan, C.A. and D.P. O’Regan 2003. Winnersand Losers in Forest Product Commercialization.Final Report to DFID ForestryResearch Programme (R7795), Centre forEcology and Hydrology, Wallingford, UK,www.ceh-wallingford. ac.uk/research/winners/literature.htmlTchoundjeu, Z., B. Duguma, J-M. Fondoun andJ. Kengue 1998. Strategy for the domesticationof indigenous fruit trees of West Africa:case of Irvingia gabonensis in southernCameroon. Cameroon Journal of Biologyand Biochemical Sciences, 4: 21–28.Tchoundjeu, Z., M.L. Avana, R.R.B. Leakey, A.J.Simons, E. Asaah, B. Duguma and J.M. Bell2002b. Vegetative propagation of Prunusafricana: effects of rooting medium, auxinconcentrations and leaf area. AgroforestrySystems 54: 183–192.Tchoundjeu, Z., A. Degrande, R.R.B. Leakeyand K. Schreckenberg 2004. Participatorydomestication of indigenous trees forimproved livelihoods and a better environment:A policy briefing note forgovernments in the African humid tropicsregion. Overseas Development Institute,London, UK.Ten Kate, K. and S. Laird 1999. The CommercialUse of Biodiversity: Access to geneticresources and benefit-sharing. Earthscan,London, UK.Tomich, T.P., M. van Noordwijk, S. Budidarsono,A. Gillison, T. Kusumanto, F. Stolle andA.M. Fagi 2001. Agricultural intensification,deforestation and the environment:assessing tradeoffs in Sumatra, Indonesia.Pp. 221–244, in: D.R. Lee and C.B. Barrett(eds) Tradeoffs or Synergies? AgriculturalIntensification, Economic Developmentand the Environment. CAB International,Wallingford, UK.Waruhiu, A.N., J. Kengue, A.R. Atangana, Z.Tchoundjeu and R.R.B. Leakey 2004.Domestication of Dacryodes edulis:2. Phenotypic variation of fruit traits in200 trees from four populations in thehumid lowlands of Cameroon. Food, Agricultureand Environment 2: 340–346.Wiersum, K.F. 1996. Domestication of valuabletree species in agroforestry systems: evolutionarystages from gathering to breeding.Pp. 147–158, in: R.R.B. Leakey, A.B. Temu,M. Melnyk and P. Vantomme (eds) Domesticationand Commercialization of NontimberForest Products, Non-Wood ForestProducts 9. Food and Agriculture Organizationof the United Nations, Rome, Italy.Wynberg, R.P., S.A. Laird, S. Shackleton, M.Mander, C. Shackleton, P. du Plessis, S.den Adel, R.R.B. Leakey, A. Botelle, C.Lombard, C. Sullivan, A.B. Cunninghamand D. O’Regan 2003. Marula policy brief.Marula commercialisation for sustainableand equitable livelihoods. Forests, Treesand Livelihoods 13: 203–215.

Keywords:Perennial tree crops, plantations, exports, price variability,smallholders, farmer organizations, diversification, certification,sustainable development, quality controlChapter 3The future of perennial tree crops:What role for agroforestry?Hubert Omont and Dominique Nicolas, Centre de coopération internationale en recherche agronomique pourle développement, France and Diane Russell, World Agroforestry Centre, KenyaAbstractPerennial tree crops play a fundamental role in the economies of many of the least developed countriesin the tropics where they occupy millions of hectares and few alternative agricultural enterprises exist.They can be a major factor in local poverty alleviation, and global demand for such tree products aschocolate, coffee and rubber continues to grow. Today smallholders produce 80–95% of tree products.As perennial landscape features tree crops can impact land tenure and provide many of the biodiversityand soil protection functions of natural forests. They rarely compete with food crops and by incorporatingvaluable intercropped annual or perennial species they can provide the basis for productive agroforestrysystems.Perennial tree crops currently face many problems, including product price instability,increased industrial concentration within certain commodity chains, withdrawal of state support andweak research efforts. Neither past attempts to regulate trade nor more recent market liberalization hasstabilized prices for their products. Producers now need to focus more on diversifying and increasingproduct quality to ensure more stable incomes. Few poor countries currently have the resources to helpproducers diversify, and increasingly it will be up to farmer organizations to attract investments fromthe private sector by improving their product quality. Even the most complex and diverse tree crop/agroforestry systems are generally not very intensive and are often low-yielding. There is potential toincrease the range and quality of their products and services. Since most perennial tree crops are producedon smallholdings, social concerns must be taken as seriously as economic and technical aspects.Research is needed on diversifying plantations and the agronomic effects of tree shading, as well as onbuilding the capacity of farmer groups and focusing on commodity quality, certification and value addition.Land tenure regulations, taxes and funding must be addressed. It is also important to promoteexamples of existing industries that have successfully diversified their outputs while maintaining thequality and supply of the main commodity.IntroductionOver recent decades, there has been an explosion inthe number and variety of perennial tree crops plantedin the humid tropics. We define these crops as plantspecies with a woody support system that periodicallyproduce a valuable crop (for food, income or environmentalbenefit) other than, or in addition to, timber.A sustained, worldwide rise in the demand for goodssuch as chocolate, coffee and natural rubber has ledto plantations being established in every continent inthe region so that they now occupy tens of millions ofhectares. For instance, world cocoa consumption hasrisen from barely 100,000 tonnes at the beginning ofthe 20 th century to 3.2 million tonnes in 2003–2004

24World Agroforestry into the Future(ICCO 2005). There is a similar trend foroil crops such as coconut (oil is extractedfrom the dried kernel – the copra) and,more recently, oil palm. Natural rubber,which was a gathered crop in Amazonia atthe end of the 19 th century, is now harvestedfrom plantations covering more than 10million hectares worldwide (InternationalRubber Study Group 2004).These crops play a fundamental role in theeconomy of many developing countries,particularly in those that are the least developedand most heavily indebted. Mostof the crops are bound for the export marketand are an important pillar of overallgrowth and rural development. “More than50 developing countries depend on threeor fewer commodities for more than halfof their export earnings. All heavily indebtedpoor countries (HIPCs) depended onprimary commodities for more thanhalf of their merchandise export earningsin 1997,” (World Bank 1999). Except forpalm oil, where estates still represent asignificant amount of the planted area,most of these perennial crops (80–95%)are grown on small to very small farms(Gilbert and Ter Wengel 2000; World Bank2002; International Rubber Study Group2004).Some of these crops (e.g. oil palm and coconut)contribute to food security at bothlocal and regional levels, while most ofthem generate income at the householdlevel. They also play a major role in generatingforeign exchange at the national level.In addition, such crops can contributeto the sustainability of agricultural systemsand, in some cases, play an important rolein the preservation of species and ecosystemdiversity (Ruf and Zadi 1998; Gockowski2001; Schroth et al. 2004)With market liberalization and globalizationraising concerns about the sustainablemanagement of land and other naturalresources, many stakeholders in treecrop commodity chains are concernedabout the future of their crops. In orderto address these issues there have been anumber of major conferences on the futureof tree crops, some of which looked at arange of products (e.g. the perennial cropsconference in Yamoussoukro, Côte d’Ivoirein 2001), while others concentrated onone commodity (e.g. the United NationsConference on Trade and Development(UNCTAD) cocoa conference in 2001,the International Coffee Organizationconference in Bangalore in 2004, and theRoundtable for Sustainable Palm Oil thathosted its third roundtable in November2005 [www.sustainable-palmoil.org]).Other meetings have focused on possiblesolutions to land management, whichinclude agroforestry and forms of certificationand quality zoning approaches (e.g.presentations at the World AgroforestryCongress in Orlando, Florida and theWorld’s Wildest Coffee event in Nairobi,Kenya, both in 2004).This chapter provides a brief overview ofthe significance of tree crops to sustainabledevelopment and the relationship betweentree crops and agroforestry systems. It thenreviews the major problems – in somecases crises – in the main tree crop sectors.Finally it suggests ways to create moresustainable tree crop systems, focusing onhow these solutions tie into an agroforestryresearch and development (R&D) agenda.It shows how agroforestry R&D may playa role either directly (e.g. in improving theproductivity of farming systems) or indirectly(e.g. through capacity building of farmerorganizations) in tree crop systems.Tree crops and sustainabledevelopmentEconomic importancePerennial tree crops have become criticalcomponents of many of the nationaleconomies in the humid tropics. Millionsof hectares are planted to cocoa, coffee,coconut, rubber and oil palm plantations,mostly in developing countries (Table 1).Production is calculated in millions oftonnes; markets are growing and generallyabsorb any surges in supply, albeit with‘booms and busts’ that can and have destabilisedlocal economies. Geographicaldistribution varies depending on the crop:almost 70% of cocoa is produced in sub-Saharan Africa; more than half of all coffeecomes from Latin America; and more than90% of natural rubber, palm oil and coconutoil is produced in Asia (Table 2).Tree crops account for a significant percentageof total agricultural exports inmany countries: in Côte d’Ivoire theycomprise 35%; in Ethiopia 26%; in Ghana25%; and in Kenya 23%, while in Ugandathey account for a massive 53% of all agriculturalexports. In Uganda, increasedearnings from coffee exports were responsiblefor half of the drop in the percentageof people living under the poverty threshold,which fell from 54% in 1992 to 35%in 2000 (World Bank 2002).The total value of tree crop exports forAfrica amounted to almost US$5 billionin 2000 (World Bank 2002), includingaround US$1.5 billion for Côte d’Ivoireand US$640 million each for Ghanaand Kenya. The value of global trade inproducts such as coffee is considerable.For instance, the United States importedUS$1.7 billion worth of green coffee in

Chapter 3: The future of perennial tree crops25Table 1. Areas planted with perennial tree crops in the humid tropics, 2003.Cocoa62 countries6 980 000 haCoconut88 countries10 707 000 haCoffee82 countries10 025 000 haOil Palm42 countries11 678 000 haRubber28 countries8 250 000 haCountries % Countries % Countries % Countries % Countries %Côte d’Ivoire 24 the Philippines 29 Brazil 24 Malaysia 30 Indonesia 32Ghana 21 Indonesia 25 Indonesia 11 Nigeria 28 Thailand 23Nigeria 16 India 18 Mexico 7 Indonesia 26 Malaysia 15Brazil 8 Sri Lanka 4 Colombia 6 Guinea 3 Vietnam 5Indonesia 7 Thailand 3 Vietnam 5 Thailand 2 India 5Source: FAOSTAT 2005.Table 2. Production of major perennial tree crops, 2003.Cocoa 13 102 000 tCoconut (copra) 25 281 000 tCoffee 36 204 000 tSources: 1. ICCO 2005; 2. Oil World 2005; 3. ICO 2005; 4. International Rubber Study Group 2004.Oil Palm 227 920 000 tRubber 47 980 000 tCountries % Countries % Countries % Countries % Countries %Côte d’Ivoire 43 the Philippines 45 Brazil 28 Malaysia 48 Thailand 36Ghana 16 Indonesia 24 Vietnam 14 Indonesia 37 Indonesia 22Indonesia 14 India 13 Colombia 11 Nigeria 3 Malaysia 12Nigeria 5 Papua New Guinea 3 Indonesia 6 Thailand 2 India 9Brazil 5 Mexico 2 Mexico 4 Colombia 2 China 62003 alone (FAOSTAT 2005). However,some products such as palm oil are usedalmost entirely by the domestic market. Themajority of tree crop products are exportedin a raw, unprocessed state. In order toreduce their dependence on internationalmarkets, many producing countries arenow attempting to carry out more processinglocally to add value to their products.Social importanceHistorically, growing tree crops was thepreserve of large private or public investors.Nowadays, tens of millions of smallholdersparticipate and tree crops provide employmentand income for hundreds of millionsof people in the rural zones of countries inthe humid tropics. This shift has taken placepartly because of the high cost of maintaininglarge plantations – particularly given thesignificant market fluctuations – and partlybecause the end of colonialism has led toa loss of plantation rights. However, largescaleplantations still exist, for examplesun-grown coffee is produced over widetracts of Brazil, large-scale oil palm plantationsare found in Malaysia and Indonesia,and tea plantations are common in southAsia and eastern Africa.Smallholdings, which usually employ familylabour, have more flexibility in the face ofsocial and economic factors. Consequently,smallholder farming systems now largelydominate copra, cocoa and natural rubberproduction: there are 700,000 cocoa producersin Côte d’Ivoire and 1.6 million inGhana. The same is also true for other commodities:Vietnam’s remarkable surge in coffeeproduction, making it the world’s secondlargest producer (it went from an annualaverage of 100,000 tonnes in the 1980s tomore than 800,000 tonnes in 2001–2002),comes from small production units.Tree crops bring in cash for agricultural environmentsthat often have few alternatives.They also help to integrate local economies

26World Agroforestry into the Futurewith regional and international markets.Wealth created in this way generally hasa multiplier effect that improves the entirelocal economy and can also benefit the nationalcommunity.Moreover, several studies have shown thatthere is rarely any real competition betweenfood crops and perennial tree crops. Foodsecurity is generally better in perennial treecrop zones than elsewhere (Krueger andBerg 2002). However, there are still risksin farming tree crops for a livelihood, particularlywhen there is centralized controlof tree crop markets that also extends tomarket and extension systems for food cropsgrown by the same farmers. If the marketfor tree products declines then the wholerural infrastructure may decline, as was thecase during the cocoa crisis in Cameroon inthe early 1990s and the more recent coffeecrisis in Kenya (Rice 2003; Ruf 1995) Treecrops have a considerable impact on landtenure. In societies where land ownershipis communal or customary, the perennialnature of the crops introduces profoundchanges in land distribution, hence in socialrelations because plantations tend to beowned by individuals or families rather thancollectively. Clearing land for plantations isan important avenue for securing tenure inareas with little formal land registration.The fact that tree crops are at the heart ofchanges in social and territorial structuresis only just beginning to be realized insome countries. “Tree crops have gonehand-in-hand with a territorializationprocess,” commented Charlery de laMasselière at Yamoussoukro (Charleryde la Masselière 2001).Many different kinds of social or tradingnetworks may be established or transformedbecause of tree crops. The qualityand nature of such networks is a decisiveelement in how well tree crops can contributeto sustainable community development.In many countries in the past, farmerscould not realize the benefits of theirwork owing to state and parastatal controlof market chains. Policies such as thosecontrolling taxation, extension, ‘forcedcooperation’ and farmers’ organizationsweakened the ability of farmers to improvetheir market position and hence alleviatepoverty. Those who benefited the mostwere the early adopters and landowners,including those who rented out land toothers and made a living from the profits(Berry 1975). Owing to this history, deregulationhas not helped because farmers arenot well organized to deal with large-scale,private-sector actors.Environmental importanceFor a long time, tree crop plantations wereextended by clearing forest on pioneerfronts, i.e. areas with low land occupationpressure that are therefore cheaper. Thiscontributed to deforestation, but at a timewhen environmental concerns were not apriority.Since the Rio Summit in 1992, environmentalconcerns have been acknowledgedas integral components of sustainabledevelopment. Faced with the threat ofdepletion of primary forests, land that hasalready been cultivated needs to be replanted.Perennial tree crops are forest-typecultivated ecosystems that can constitutesustainable systems:• They help to protect existing forests bysupplying wood for industry and energy.For example, in Sri Lanka “perennialcrop-based farming systems supply over50% of national timber and 80% of thefuelwood needs”, (Pushpakumara 2001).• They make a substantial contributiontowards carbon sequestration. For example,rubber trees can sequester morethan 100 tonnes of carbon per hectareover 33 years (Hamel and Eschbach2001).• The permanent cover and fairly systematicuse of cover crops effectively protectssoil from erosion. “A well-managedtea plantation results in an annual [soil]loss of only 0.24 t ha –1 , compared with25–100 t ha –1 for vegetables, potatoesand tobacco, and 0.3 t ha –1 for denseforest,” (Kaosa-ard and Rerkasem 1999).• Soil cover also reduces the risk of leaching,and legumes – often associated withtree crops – help to improve the nitrogenbalance.• Use of pesticides is limited: either theyare not necessary or farmers cannotafford them.• It is possible to maintain a degree ofdiversity in multi-storey plots such asthose for ‘jungle rubber’ in Indonesia,cocoa agroforests in Cameroon andintercropped coffee in Ethiopia, Indiaand Indonesia• Most perennial crops are less sensitive tofertility levels than food crops, and someof them can help to stabilize the agroecologyof marginal or degraded lands.In Central America, Arabica coffee treeshelp to fix many fragile mountain soils.While these are positive aspects of tree cropsystems, conservation biologists warn thateven complex tree crop/agroforestry systemsare not equivalent to natural forests.Furthermore, farm inputs such as pesticides,particularly the copper-based fungicidesused on cocoa, can harm other species. Treecrop plantations can encroach significantlyon protected areas (e.g. the cocoa beingplanted within Lore Lindu Park in Sulawesi,Indonesia; Schroth et al. 2004).

Chapter 3: The future of perennial tree crops27When managed in smallholder plantations,individual agroforestry patches oftenbecome fragmented making it harder toretain a working ecosystem. Recent thinkingin conservation science calls for waysto connect forest patches and smallholdertree crop/agroforestry systems with largerprotected forests to boost the conservationbenefits to the landscape scale. There arenumerous challenges to these proposalsincluding the spread of diseases, pesticideuse and the effects of animals on crops (seecase studies in Schroth et al. 2004).Tree crops within agroforestrysystemsAccording to ICRAF, “Our vision is anagroforestry transformation in the developingworld, resulting in a massive increasein the use of working trees on workinglandscapes by smallholder rural householdsthat helps ensure security of food,nutrition, income, health, shelter and energyand a regenerated environment” (WorldAgroforestry Centre 2005). This all-embracingdefinition means that, in many cases,perennial tree crops can be consideredto be agroforestry systems; they providemany of the services identified by ICRAFas being relevant, including: income generation,soil fertility enhancement, carbonsequestration, weed control, microclimateimprovement and reclamation of degradedlands. Even the most complex and diversetree crop/agroforestry systems, includingthe jungle rubber systems in Indonesia, aregenerally not very intensive and are oftenlow-yielding, usually because of a lackof high-yielding varieties or appropriatecultural techniques. While some agroforestrysystems, such as shade coffee, canbe extremely efficient, knowledge of themechanisms that govern these systems andthe ideal crop combinations remain highlyspecific to each region, limiting the possibilitiesof sharing results between regions.If these systems are to be improved sustainably,their assets – such as the environmentaland biodiversity benefits, risk-sharingand labour use – need to be promotedand improved. Similarly, their weaknesses,which are primarily of a technical or economicnature relating to low productivityand quality of the products, need to be addressed.Challenges to tree cropsystemsDespite the undeniable advantages offeredby perennial tree crops, there are concernsabout the role they should play in sustainabledevelopment. Price instability, marketinefficiencies, difficulties in diversifying,gaps in commodity chain organization,problems in renewing the means of production,and quality demands are a fewof the problems that need to be dealt withwhen considering their future.Variability in prices and marketinefficienciesVariability in commodity prices is oneof the most serious risk factors for thefuture of certain commodity chains (Ruf1995). Perennial tree crops underwent theirgreatest crisis in the 1990s when tropicalFigure 1. Commodity prices trends 1960–2004.2000150010005000Source: World Bank 2005.agricultural prices declined (Figure 1). Overthe present decade, countries also shiftedemphasis between the crops they weregrowing (Table 3). At the end of this period,copra was trading at around US$200 pertonne and palm oil was between US$250and US$300 per tonne, i.e. around thesame level as in the 1960s. Given that theAmerican price index had risen five-foldin that time, in real terms copra and palmoil prices were a fifth of their 1960s value(World Bank 2005)While market prices have improved forproducts such as cocoa, rubber and oil,the gap between prices paid by end usersand those paid to producers is widening(Oxfam 2002). In the case of coffee, forwhich prices have remained dramaticallylow, price weakness is combined with anextremely volatile market, with fluctuationsdepending on numerous uncontrollablefactors. The mere hint of a frost forecast inBrazil – the world’s leading producer – cansend coffee prices soaring; a denial a fewweeks later can cause a slump. Similarly,political uncertainties in Côte d’Ivoire arehaving a visible impact on cocoa prices.In order to cope with such variability,stakeholders have attempted to establishCoconutPalm oilCocoaRubberCoffee RobustaCoffee Arabica1980 1985 1970 1975 1980 1985 1990 1995 2000 2004

28World Agroforestry into the FutureTable 3. Area planted with each of five main tree crops ('000 000 ha).Cashew Cocoa Coffee Oil Palm Rubber1995 2003 1995 2003 1995 2003 1995 2003 1995 2003Brazil 700 673 739 589 1870 2396 33 52 50 103Côte d’Ivoire 70 125 1900 1700 920 400 134 141 46 70Ghana 2 13 1000 1500 10 8 100 115 15 18Kenya 0.85 2 – – 160 170 – – – –Vietnam 189 258 – – 155 500 – – 278 437Source: FAOSTAT 2005.trade regulation mechanisms: internationalagreements with or without buffer stocks,stabilization funds or cartels of producingcountries. None of these arrangements haveso far withstood the test of time and theydo not harmonize with the current movestowards liberalization and globalization.International funding organizations suchas the International Monetary Fund (IMF)and the World Bank have in fact pushed forthe dismantling of such state control andregulatory mechanisms through structuraladjustment programmes (SAPs).As well as affecting international trade, pricedeclines and fluctuations increase the riskto producers, who no longer have any guaranteeof remuneration and may end up withdebts that require them to sell or mortgagetheir land. These situations are exacerbatedfor perennial tree crops, which, by definition,are a long-term investment. Farmerswho obtain a poor price for their groundnutscan change their crop and production methodfor the next cycle; changing a coconutbasedfarming system is an entirely differentmatter. The upside to agroforestry systems is,however, that farmers can be more flexibleabout whether they choose to harvest a treecrop or not, although in many cases this isscarcely enough to compensate.So far, market liberalization has not producedmechanisms or institutions that canadjust supply to a somewhat inflexibledemand and therefore play a regulatoryrole to satisfy all concerned. In part thiscan be attributed to the weakness of farmerorganizations, particularly those in Africa,which is the result of having previously hadState-run enterprises and cooperatives. Insome countries, farmer organizations arehighly politicized, leading to in-fightingand rigid organizational hierarchies. Theprivate sector is unlikely to create appropriateregulatory institutions by itself becausefirms compete with each other and thedominant firms prefer private market arrangements.The weakening of the State byliberalization and the SAPs was justified onthe grounds of State mismanagement, butthe vacuum it has left has not been filled.Disinvestment has in fact considerablyweakened national research and extensioncapabilities that at least used to providesome support to farmers.One must also question whether liberalizationis actually being implemented. Controlson trade still exist in many countries,particularly in Africa. For example, pricesfor food crops may be kept low by controlsat the national level, or there may be capson the price of tree crops (timber and nontimberforest products) that make themunprofitable.Difficulties in diversifyingMany tree crop experts consider diversificationto be vital to protect product pricesagain entering a heavy downward trend.But such diversification, which needs totake place on both farm and producingcountry scales, comes up against numerousobstacles.In many countries, impoverished farmerslack both the information or training onplantation diversification and the technicalor financial resources required to diversify.This is all the more true for farmersof perennial tree crops, which experiencelong immature periods and wide price variations.They are also more likely to faceuncertainties surrounding land tenure andtree use regulations, particularly aroundprotected and indigenous species, whichcan seriously curb diversification efforts.On a national level, few poor countriescurrently have the human, technical andfinancial resources needed to help producersdiversify and increase local processing.For example, in most countries in Africa it

Chapter 3: The future of perennial tree crops29is difficult for smallholders to obtain qualitytree germplasm that might make suchprocessing feasible (Ræbild et al. 2005).Perhaps the most serious problem is theweak internal markets and market chains inmany countries. For tree products in highdemand such as charcoal, there are numerousregulations that inhibit markets (Russelland Franzel 2004). While smallholder systemsof timber intercropped with coffee oranother tree crop are being developed inCentral America, southern Asia and SouthAfrica, they are still in their infancy in mostof sub-Saharan Africa. There are great differencesin the markets for fruit and fruitproducts throughout Africa: they are weakin eastern and southern Africa but strong inWest Africa. Furthermore, export has nottaken off to any great extent and there is littlevalue addition; countries in eastern andsouthern Africa find it hard to compete forexport markets with the horticultural expertiseof South Africa (Petit and Barghouti1992; Maizels 1999).Not enough made of technicalinnovations?The development of most perennial treecrops has taken place outside the naturalrange of the plants, and has been basedon pioneer-type models exploiting existing,untouched forest resources and usingcheap labour. First cropping cycles thereforebenefit from soils that are fertile fromtheir forest past and where specific parasitesare absent.Major research efforts have led to significantprogress in improving plantation yieldand productivity by improving varieties,cultural practices and parasite control.Such efforts have increased yields of somecrops five-fold over the last 50 years; harvestsof natural rubber, for example, havegone from 500 kg ha –1 to 2,500 kg ha –1 .In general, their pioneering mindset leadsfarmers to increase their yields by expandingarea cultivated rather than by increasingproductivity per unit area. Cocoa is aclear example: Spanish reports of CentralAmerican native plantations dating fromthe colonial period indicate annual yieldsto be around 700 kg ha –1 which is higherthan current average world production ofless than 500 kg ha –1 per annum.Today, some crops have aged, as have thepeople who planted them and the soils thatbear them; this type of extensive farming isreaching its territorial limits and nearing itsend. Although high-yielding varieties mayexist locally, they are rarely easily accessibleand affordable for smallholders. New systemsand technical innovations are neededto permit replanting and intensification (Rufand Konan 2001; Ruf and Schroth 2004)The integrated pest management approachis in its infancy in most areas and few producerscan afford to buy the recommendedphytosanitary products. And while multiplespecies cropping systems are widely practisedin plantations, little is known abouthow they function. Post-harvest processingoperations are often insufficiently masteredto market a product that meets increasinglystrict quality requirements.Many countries experience problems inraising funds for research into perennialtree crops, particularly when the prices formost commodities are enduringly low. Donorsupport for these research projects hasgenerally been limited, as the emphasis formost donors is on food crops that directlycontribute to food security rather than oncommercial agriculture as a whole. Manydonor agencies consider that research intotree crops must be funded by the industrialsectors that make use of their products.They are reluctant to invest public moneyin what they consider produces privategoods (although this may be changing: theWorld Bank for example is developing atree crop strategy for Africa). On the otherhand, the private sector only contributesfunding for specific projects that are closelylinked to the competitiveness of its industry,especially in the downstream/post-harvestpart of the chain, and many producers’concerns remain neglected or even ignoredby the industry.Quality requirementsProduct quality requirements are both aconstraint and an opportunity for producingcountries. Draconian technical andsanitary standards, often associated withthe importing country’s health concerns,have to be respected, but can be difficult toachieve for developing countries. However,they can be of benefit if meeting them providesthe producer with a distinct qualityadvantage that can be promoted.Unfortunately, many low-income countriescannot rapidly adopt the new technologiesrequired to meet exacting exportstandards, and lose out to better-preparedmedium-income countries. The situationis exacerbated by the previously discussedState withdrawal.One way to promote quality that is availableto developing as well as developedcountries is to introduce the terroir concept,whereby a specific crop (in France itis wine) is only produced by a particularregion. Similar products from areas outsidethe region cannot be given the same name.A study in Honduras laid the foundationfor possible coffee terroirs (Avelino 2002).Similar studies have been undertaken inother countries such as the DominicanRepublic, Indonesia, Kenya and Rwanda.It has also been shown that local consumerscan appreciate different qualities, as demonstratedwith red palm oil in Côte d’Ivoire

30World Agroforestry into the Future(Cheyns 2001). And there is growing worldwidedemand for fair trade and sustainableproduction, which bring new requirementsto producers. A good example is the recentcreation of the RSPO – the Roundtable forSustainable Palm Oil (http://www.sustainable-palmoil.org)– that gathered togethermany stakeholders from the sector. Somecommodities are obviously more advancedthan others: coffee quality, for example, hasmore consumer recognition while few knowabout different types of rubber.However, promotion of the terroir conceptand other certification processes often facesopposition from downstream industrial sectorswhere the feeling is that end-productquality depends exclusively on their knowhow.They do not accept the idea that thelocation, crop variety and technical skills ofthe farmer can play a role in the quality offeredto consumers. At the moment farmersare not organized, end-consumers know virtuallynothing about the commodity chainsinvolved (except perhaps for coffee), andit remains to be seen if the costs of variouscertification processes can be justified byprice improvements.The way forwardsAlthough perennial tree crops are only partof the economic tapestry of national andlocal development, they possess uniquecharacteristics and it is worth paying attentionto how they can better contribute.This section suggests policy options andfinancial and technological actions thatwill ensure that tree crop systems play anincreasing role in sustainable development.Political and financial mechanismsfor sustainabilityThere are three essential aspects to improvingthe tree crop sector: land tenure regulations,taxes and the funding of activities.Firstly, it is important for the sustainabilityof tree crop-based farming systems that therelationship between land tenure and perennialtree crop development be considered.This relationship is not straightforward: localconditions are diverse and complex andsimple models cannot be extrapolated. Butpublic authorities need to be aware of howformal and informal tenure arrangementsplay a role in the tree-crop sector.Secondly, as with many goods, the tax rateapplied to tree crops destined for export isa decisive element in the sustainability ofthe production chain. A satisfactory balancemust be found between a low rate thatencourages commodity chain vitality and ahigher rate to meet the State’s requirements.As perennial crops are often one of its mostvisible and lucrative resources, the State hastended to over-tax in the past.Finally, farmers’ access to financial resourcesrelies on two things: production earningsand access to credit. As discussed earlier,wide variation in product prices creates incomeinstability; itself a factor in productioninsecurity. Several options have been developedto help including one by the WorldBank, which has taken the initiative to setup an international task force to examinethe feasibility of commodity price risk management.This involves offering producersrisk-management instruments, implementedby local transmission mechanisms, cooperatives,agricultural banks or exporters. Thefirst step has to be to strengthen the financialinstitutions in many countries otherwisethese funds will be wasted.Access to credit is often difficult for poorfarmers. In order to facilitate such access,land tenure regulations could be changed togive perennial tree crops a guarantee valueto help secure a loan. However, this mightalso lead to a rise in land speculation andland being appropriated by the rich, increasingrural landlessness. Hence a holistic andforward-thinking strategy that includes attractingmedium-scale industrialization oftree crop industries has to be envisaged.Overall what is needed is a clear demonstrationof how tree crops contributeto food security and poverty reduction atboth household and national levels. Theproblem is that there persists a simplisticnotion that food security is mainly to dowith household food production; in fact itdepends upon the ensemble of strategiesthat people use to get food and allocatetheir land and labour (for food and also foreducation). There is evidence that tree cropincome, coming regularly throughout theyear, contributes greatly to the formationand maintenance of social capital, goingtowards school fees, house building andsocial stability (e.g. marriage payments andhealth costs; Russell and Tchamou 2001).Social capital creates a safety net that bufferscommunities against risk.Tree crops and other plants, such as thosewith medicinal value, intercropped inplantations are a source of income and nutrition.These crops may provide jobs for arural labour force that is desperately underemployed.The addition of small enterprisessuch as tree nurseries and primary processingof tree products adds stability to fragilerural economies. In some cases the localvalue of intercrops can exceed the value ofthe traditional tree crop (Gockowski andDrury 1999).Organization of producersWith small farms and limited financialresources, isolated smallholders hold nomore sway over product prices or commoditychain balances than they do overthe agricultural policies of their countries,especially in a context of globalization.

Chapter 3: The future of perennial tree crops31Producer organizations enable farmers toexert a true negotiating influence and totake part in drawing up agricultural policies,i.e. to become fully fledged stakeholders,not only in the commodity chainbut also in the sustainable development oftheir country.There are also research and development(R&D) implications: farmer groups canbecome involved with capacity building,increasing the emphasis on quality ratherthan quantity. They can push for the creationof quality zones and other certificationmechanisms, and provide the demand fordevelopment of diversification optionswith high and sustained economic value.National and international support forthese organizations should come in theform of training and information, to givethe farmers the resources they need to exertnegotiating power within a commoditychain. Social sciences research can helpby analysing local situations and proposingapproaches that have been tried and testedelsewhere in similar circumstances.There are many types of farmer organization.These include shareholding schemes,where producers buy a share of a downstreambusiness (e.g. bulking or primaryprocessing); cooperatives, which do notinvolve financial transactions, instead thefocus is on sharing information, trainingand advocacy; or bulking centres that arelinked to farmers and farmer groups, whichis being proposed, for example, in the Kenyandairy industry.Whatever form the group takes, it is importantthat the impetus comes from thefarmers themselves. Groups created bynon-governmental organizations (NGOs)or governments are often doomed to failurebecause the farmers joining are notnecessarily the best representatives of theircommunities, and their motives for joiningmay be coloured by the expectation ofhandouts. Perhaps the best way for NGOsand governments to help is by networkingand building the capacity of existinggroups that have stood the test of time(Tanui 2003).Integrating tree crop and agroforestry extensioncould enhance benefits to farmergroups and reduce risks by introducing arange of options. Agroforestry groups couldprovide an additional platform of actionthat is perhaps less politicized than treecrop commodity chains. Better land managementcan also be integrated. The challengeis to maintain or even increase thequality of the tree commodity.Cooperation within commoditychainsIn the short term, the interests of the differentstakeholders in a commodity chain maydiverge and take the form of a power struggle,but in the long term it is the prosperityof the entire chain that is everyone’s mainconcern. If certain categories of stakeholderssystematically lose out, they will leavethat field of activity, which can be fatal tothe commodity chain (this happened tocastor oil producers a few decades ago).Most commodity chains have either formalor informal structures in which stakeholdersmeet, exchange information and sometimesimplement joint projects. Rampingup such cooperation could guarantee moreequitable relations between stakeholdersand promote services that ensure thesmooth running of commodity chains.The Global Forum on Agricultural Research(GFAR) has an initiative to promote globalresearch programmes for perennial treecrops, in a move to strengthen cooperationamong commodity chain stakeholders. Inthe context of financial restrictions that areseriously affecting research structures indeveloping countries, better cooperationwithin commodity chains (and sometimeseven between commodity chains) couldhelp in putting all these assets to more effectiveuse, so as to concentrate resourceson what needs true scientific and technicalinnovations. The GFAR Global Programmes(GPs) are a step in that direction. GPs consistof coordinated sets of activities, carriedout by a wide range of programme participantsor partners, and directed towardsspecific problems or sets of problemsidentified at a global level. ProMusa, forbanana and plantain, and Procord for coconutare the two existing GPs. The cocoasector is discussing the feasibility of a sustainablecocoa production GP, and otherapproaches are being developed for differentcommodities (Frison et al. 2000; GFAR2002; Omont and Frison 2002).These stakeholder meetings and processesare important steps, but a great deal remainsto be done to harmonize researchacross a large number of countries, manyof which have weak research and extensionsystems. The most crucial next step isto design a range of options at the interfaceof environmental and market imperativesthat will reduce the cost of certificationschemes to smallholders and poor countries,and to find cost-effective ways to supportbetter land management. These couldinclude environmental service paymentsor special arrangements between localitiesproducing high-value tree crops that borderfragile or protected areas. Stakeholdergroups need to address the proliferationof certification and quality standardsand also build internal markets for somecrops where certification schemes such asEurepGap and others present a formidablebarrier.

32World Agroforestry into the FutureScientific and technical innovationsWhen markets were booming, technicaland scientific innovation was primarilygeared towards achieving higher yields andgreater productivity. Today we have a morecomplex situation. The success of scientificand technical innovation can no longer beassured without the involvement of all thestakeholders and a clear understanding ofsocial and political contexts. The complexbehaviour of those stakeholders and theshifting policy environments at many levelsmean that social and policy sciences haveto be much better integrated into technologyresearch.The fact that most perennial tree crops areproduced on smallholdings means thatsocial concerns must be taken as seriouslyas the economic and technical aspects.It is essential to know more about thestakeholders in commodity chains: theirmotivation; the reasons for their choices;the scope of their decisions; and how newtechnologies can be adapted to their requirementsand effectively transferredto them.In terms of raising production and productivity,numerous technical possibilitiesalready exist: new varieties, disease controlmethods, crop management sequencesand post-harvest technologies for instance;some of which need to be adapted to newor different socioeconomic contexts. Butsuch possibilities are not always availablefor the areas where they would be mostuseful, and those who need them do notalways have the relevant information. Forsome commodity chains that are overproducing,such as coffee, it is no longernecessary to raise production, though it remainsrelevant for other commodity chainsthat expect a shortage such as cocoaand natural rubber. However, translatingimproved productivity into increased producerincomes remains a priority for manycommodity chains. Quality, be it in termsof industrial technical standards or healthstandards, is becoming the main concern.On top of all these concerns, environmentalissues must also be integrated into theR&D process. This will require a change inthe prevailing attitude and paradigm, notto mention specific scientific and technicalinnovations. Perennial tree crops can havea positive impact on the environment, butassumptions about how to improve landscapeconnectivity and hence biodiversitythrough agroforestry and growing treecrops need to be rigorously tested. Thereare also many serious questions about animalhabitats, choice of tree species, flow ofpests and diseases, and tree genetic diversityon a landscape scale that need answering(Schroth et al. 2004).Agroforestry approachesFor producers, the aim is to share risk bydiversifying crops, either in separate patchesor intercropped within the trees. It is amatter of making better use of land, financialresources and farm labour as well asattempting to maintain (or even improve)the fertility of cultivated ecosystems.If there are no severe land occupation constraints,diversification can take the formof a succession of monocultures withoutintercrops. However, in many countries,farming systems already use several plantsor crops, often in complex combinations.Growing food crops in rows within ayoung tree crop plantation is a commonpractice (known as the taungya system).Several research centres are working onrationalizing these approaches with systemsof biomass transfer and improvedfallow with coppicing and/or quick growthshrubs. More recently, long-term intercroppingof perennial crops has been tested inseveral countries with apparent successfrom a technical aspect (Erhabor et al.2002; Herath and Takeya 2003; Maheswarappaand Nanjappa 2000; Ndeayo et al.2001; Ollivier et al. 2001; Osei-Bonsu etal. 2002; Rajasekharan and Veeraputhran2002; Suja et al. 2003). Indeed, in intercroppingterms, the only limit appears tobe the imagination of the farmers.To assist diversification efforts, researchinstitutions need to make examples of successfuldiversification available (e.g. coffeediversification in Ethiopia and India ,timber and coffee in Costa Rica, fruit treesand cocoa in Cameroon, etc.), provideguidelines, focus more on meeting demandin local and regional markets, and linkdiversification to certification. The CASCAproject (sustainability of coffee agroforestrysystems in Central America [http://www.casca-project.com]) is a good exampleof what R&D could bring to the issue.Another example comes from the rubbersector – jungle rubber – where a multistakeholders’project is being developedin Indonesia, with the aim of improvingthe yield of rubber trees while maintainingthe good environmental achievements ofthe jungle rubber agroforests (Herath andTakeya 2003).The private sector may be encouraged toinvest more in R&D for smallholders is if ithelps maintain or increase quality. Whenquality is declining dramatically, high-endproducers lose out. This is the case forcoffee in Kenya, considered to be amongthe most valuable in the world. Somecompanies are also concerned about theirreputations, for example in the wake of thecoffee crisis that devastated smallholdersbut created record profits for manufacturingcompanies such as Nestlé. The use of childlabour in tree crops is another concern thathas reached world attention. M&M Mars,

Chapter 3: The future of perennial tree crops33the chocolate manufacturers – a familyownedcompany, has a goal of substantiallyincreasing farmer revenues within the next5 years through quality improvements anddiversification (Shapiro and Rosenquist2004). They are embarking on diversificationefforts into medicinal commodities thatcan be grown in smallholder cocoa plantations.Reducing child labour and increasingthe use of integrated pest management (andtherefore reducing chemical use) was themotivation behind the creation of the SustainableTree Crops Programme (STCP) inWest Africa, which is backed by US chocolatemanufacturers. Chocolate producers arealso interested in such commodities as sheabutter, which is used in chocolate manufactureand thus may sponsor research toimprove these tree products.Local consumers could stand to benefitgreatly if important local trees and plantsare integrated into tree crop plantations insuch a way that both supply and quality increasewithout adversely affecting the qualityand supply of the main commodity. Thisis the tree crop/agroforestry challenge.ConclusionPerennial tree crops have numerous assetsthat need to be highlighted to ensure sustainabledevelopment.• They play a fundamental role in theeconomy of numerous producing countries;the additionalmonetary incomescontribute towards poverty alleviationand the maintenance of social balances.• They play a role in structuring societies,e.g. by bringing in capital at differentlevels that enable investments, allocationof land and organization of commoditychains.• They play a decisive role in maintaininga forest-type cultivated ecosystem, bothin terms of fixing carbon and preventingsoil erosion.But the future of these crops is threatenedby several negative factors:• The volatility of product prices leads tochronic insecurity within commoditychains, which can be especially bad forsmallholder farmers in the absence ofregulatory mechanisms.• Increased industrial concentrationwithin certain commodity chains isworsening the imbalance of power, thatparticularly affects producers who arepoorly, or not at all organized.• The total withdrawal of State supportfrom some commodity chains is just asdetrimental as its over-involvement withsome of them in the past.• The weakness of public research efforts,in both developed and developingcountries is not being compensated forby private research efforts.The political, economic and social contextof tree crop farming is undergoing profoundchanges, with the emergence of newissues centring on the sustainable managementof territories and natural resources,quality, food safety and ethical issues (e.g.,child labour).The development of environmentally sustainableand competitive production systemsfor these crops, including agroforestry,and the availability of long-term incentivesfor smallholder producers will require newapproaches and a concerted effort on thepart of the international agricultural community,but it will also involve all commoditystakeholders.Agroforestry research for its part has tostrongly embrace the tree crop R&D agenda.Agroforestry systems that comprise thesehigh-value crops but also provide a range ofproducts and services can be true engines offuture sustainable development. The commonagenda that is emerging focuses on:• Studying technical, economic, and socialissues in diversification of tree cropplantations.• Focusing on commodity quality and agronomicquestions related to shading oftree crops.• Building capacity of farmer groups inintegrated plantation-agroforestry.• Strengthening farmer groups and associationstechnically and organizationally.• Testing approaches to certification andlocal value addition, including terroirthat strengthen farmers’ capacity, improvetheir negotiating power and maintainmarket value over time.• Determining local demand and need foragroforestry and tree products that canbe integrated into plantations.• Networking information on sustainablemanagement, diversification and valueaddition across the continents.ReferencesAvelino, J. 2002. Identifying terroir coffees inHonduras. Research and Coffee Growing,Centre de coopération internationale enrecherche agronomique pour le développement(CIRAD), Montpellier, France.Berry, S. 1975. Cocoa, Custom and SocioeconomicChange in Rural Western Nigeria,Clarendon Press, Oxford, UK.Charlery de la Masselière, B. 2001. Petits planteurset construction territoriale en Afriquenoire. Quels enjeux pour les culturespérennes? International Conference onthe Future of Perennial Tree Crops, 5–9November 2001. Yamoussoukro, Côted’Ivoire. CIRAD, Montpellier, France.

34World Agroforestry into the FutureCheyns, E. 2001. La consommation urbaine del’huile de palme rouge en Côte d’Ivoire:Quels marchés? International Conferenceon the Future of Perennial Tree Crops, 5–9November, 2001. Yamoussoukro, Côted’Ivoire. CIRAD, Montpellier, France.FAOSTAT 2005. Website coffee statistics for2003, Food and Agriculture Organizationof the United Nations, Rome, Italy.Erhabor, J.O., A.E. Aghimien and G.C. Filson2002. The root distribution pattern ofyoung oil palm (Elaeis guineensis jacq)grown in association with seasoned cropsin southwestern Nigeria. Journal of SustainableAgriculture 19 (3): 97–110.Frison, E., H. Omont and S. Padulosi 2000.GFAR and international cooperation oncommodity chains. Synthesis paper preparedfor the Global Forum on AgriculturalResearch, 2000 Conference. Dresden,Germany. Deutsche Gesellschaft für TechnischeZusammenarbeit (GTZ) GmbH,Eschborn, Germany.GFAR (Global Forum on Agricultural Research)2002. Final report of the Workshop onGlobal Programmes for CommodityChains – Results and Prospects. 6–7 June2002. Montpellier, France. http://www.egfar.org/tools/april2002/aprilexsec.htmGilbert, C.L. and J. Ter Wengel 2000. Commodityproduction and marketing in a competitiveworld. CFC/UNCTAD panel discussion.United Nations Conference on Tradeand Development X, Bangkok, Thailand.Gockowski, J. and S. Dury 1999. The economicsof cocoa–fruit agroforests in southernCameroon. In F. Jiménez and J. Beer, eds.Multi-strata Agroforestry Systems withPerennial Crops, p. 239–241. CATIE,Turrialba, Costa Rica.Gockowski, J. 2001. Smallholder Cocoa Systemsin the Moist Tropics of West and CentralAfrica: A Pathway to Sustainable Development.International Institute of TropicalAgriculture, Ibadan, Nigeria.Hamel, O. and J.M. Eschbach 2001. Potentialimpact of the clean development mechanismin the future of perennial crops.International Conference on the Futureof Perennial Tree Crops, 5–9 November2001. Yamoussoukro, Côte d’Ivoire.CIRAD, Montpellier, France.Herath, P.H.M.U. and H. Takeya 2003. Factorsdetermining intercropping by rubbersmallholders in Sri Lanka: A logit analysis.Agricultural Economics 29 (2): 159–168.ICCO (International Cocoa Organization) 2005.Annual Report 2003–2004. InternationalCocoa Organization, London, UK.ICO (International Coffee Organization) 2005.International Coffee Organization: Statistics.http://www.ico.org/. 20 March 2005International Rubber Study Group 2004. RubberStatistical Bulletin, 58 (8).Kasoa-ard, M. and B. Rerkasem 1999. TheGrowth and Sustainability of Agriculturein Asia. Asian Development Bank, Manilla,the Philippines and Oxford UniversityPress, Oxford, UK.Krueger, A. and A. Berg 2002. Trade, growth andpoverty: A selective survey. Annual BankConference on Development Economics.World Bank, Washington, DC, USA.Maheswarappa, H.P. and H.V. Nanjappa 2000.Dry matter production and accumulationin arrowroot (Maranta arundinacea L.)when intercropped with coconut. TropicalAgriculture 77 (2):76–82.Maizels, A. 1999. Economic dependence oncommodities. Paper prepared for theHigh-Level Roundtable on Trade andDevelopment: Directions for the TwentyfirstCentury. UNCTAD X. February2000, Bangkok, Thailand. http://scholar.google.com/scholar?hl=en&lr=&q=cache:NzUTcUVNyvwJ:r0.unctad.org/en/docs/ux_tdxrt1d6.en.pdf+Maizels+1999+economic+dependance+commodities.Ndaeyo, N.U., G.S. Umoh and E.O. Ekpe 2001.Farming systems in southeastern Nigeria:Implications for sustainable agriculturalproduction. Journal of Sustainable Agriculture17 (4): 75–89.Oil World 2005 Statistics. http://www.oilworld.biz/app.php, 20 March.Ollivier, J., W. Akus, L. Beaudoin-Ollivier, X.Bonneau and T. Kakul 2001. Replanting/underplanting strategy for old coconutplantations in Papua New Guinea. Impacteconomique de differents scenarios dereplantation de vieille cocoteraie. OCL-Oleagineux Corps Gras Lipides 8 (6):659–665.Omont, H. and E. Frison 2002. Activity reportof the GFAR Montpellier Facilitation Unit— 1999–2003. Workshop on Global Programmesfor Commodity Chains – Resultsand Prospects. 6–7 June 2002. Montpellier,France. GFAR, Rome, Italy.Osei-Bonsu, K., K. Opoku-Ameyaw, F.M. Amoahand F.K. Oppong 2002. Cacao–coconutintercropping in Ghana: Agronomic andeconomic perspectives. AgroforestrySystems 55 (1): 1–8.Oxfam 2002. Mugged: Poverty in Your CoffeeCup, Oxfam, Oxford, UK.Petit, M. and S. Barghouti 1992. Diversification:Challenges and Opportunities, WorldBank, Washington, DC, USA.Pushpakumara, D.K.N.G. 2001. Environmentalbenefits of perennial crop farmingsystems. International Conference onthe Future of Perennial Tree Crops,5–9 November 2001. Yamoussoukro,Côte d’Ivoire. CIRAD, Montpellier, France.Ræbild, A, B. Bassirou, J-P. Barnekow Lillesø,E-L. Yago and P. Damas 2005. Farmers’planting practices in Burkina Faso. Surveycarried out by the project ImprovedSeed Supply for Agroforestry in AfricanCountries (ISSAAC). Forest and LandscapeWorking Papers No.5. Forest and Landscape,Denmark and World AgroforestryCentre, Kenya.

Chapter 3: The future of perennial tree crops35Rajasekharan, P. and S. Veeraputhran 2002.Adoption of intercropping in rubber smallholdingsin Kerala, India: A tobit analysis.Agroforestry Systems 56 (1): 1–11.Rice, R. 2003. Coffee production in a time ofcrisis: social and environmental connections,SAIS Review vol. XXIII no. 1(Winter–Spring).Ruf, F. 1995. Booms et Crises du Cacao: Les Vertigesde l’Or Brun, Karthala, Paris, France.Ruf, F. and A. Konan 2001. Les difficultés de lareplantation. Quel avenir pour le cacaoen Côte d’Ivoire ? International Conferenceon the Future of Perennial Tree Crops,5–9 November, 2001. Yamoussoukro,Côte d’Ivoire. CIRAD, Montpellier, France.Ruf, F. and G. Schroth 2004. Chocolate forestsand monocultures: a historical review ofcocoa growing and its conflicting role intropical deforestation and forest conservation.Pp 107–134 in: Schroth, G., A. Gustavo,B. Fonseca, C.A. Harvey, C. Gascon,H.L. Vasconcelos and A-M.N. Izac (eds),2004. Agroforestry and Biodiversity Conservationin Tropical Landscapes, IslandPress, Washington, DC, USA.Ruf, F. and H. Zadi 1998. Cocoa: From deforestationto reforestation. Paper contributedto the First International Workshop onSustainable Cocoa Growing, 29 March –3 April 1998. Panama, Republic ofPanama.http://natzoo.si.edu/smbc/Research/Cacao/ruf.htm.Russell, D. and S. Franzel 2004. Trees of prosperity:Agroforestry, markets and theAfrican smallholder, Agroforestry Systems,61–62 (1–3): 345–355.Russell, D. and N. Tchamou 2001. Soil fertilityand the generation gap: the Bënë ofsouthern Cameroon In: Carol J. PierceColfer and Yvonne Byron (eds) PeopleManaging Forests: the Links between HumanWell-being and Sustainability, RFFPress – Resources for the Future, Washington,DC, USA.Schroth, G., A. Gustavo, B. DaFonseca, C.A.Harvey, C. Gascon, H.L. Vasconcelosand A-M.N. Izac 2004. Agroforestry andBiodiversity Conservation in TropicalLandscapes, Island Press, Washington,DC, USA.Shapiro, H.Y. and E.M. Rosenquist 2004.Public–private partnerships in agroforestry:the example of working togetherto improve cocoa sustainability, AgroforestrySystems, 61–62 (1–3): 453–462.Suja, G., V.M. Nair, P. Saraswathy andR. Pushpakumari 2003. Plant populationand sett size effects on white yam(Dioscorea rotundata Poir.) intercroppedin coconut gardens. Tropical Agriculture80 (3): 157–162.Tanui, J. 2003. What about the land user? AnAfrican Grassroots Innovation for Livelihoodand Environment (AGILE) Approach.African Highlands Initiative, AHI Brief No.D1. November 2003. Nairobi, Kenya.World Agroforestry Centre 2005. Trees ofChange – A Vision for an AgroforestryTransformation in the Developing World.ICRAF, Nairobi, Kenya.World Bank 1999. Discussion paper for theroundtable on commodity risk management,World Bank, Washington, DC, USA.World Bank 2002. Reinvesting in African SmallholderAgriculture: The Role of Tree Cropsin Sustainable Farming Systems. WorldBank, Washington, DC, USA.World Bank 2005 Data and Research – Commodityand Price data (pink sheets). http://web.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTDECPROSPECTS 0,contentMDK:20268484~menuPK:556802~pagePK:64165401~piPK:64165026~theSitePK:476883,00.html, 20 March 2006.

Chapter 4Trees and Markets:Working Group ReportIntroductionICRAF’s Trees and Markets theme created workinggroups to discuss the enhancement and future of thefour focal areas that comprised the theme in 2003(a fifth was added in 2004). These were: TM1 – marketanalysis and support for tree product enterprises; TM2– developing sustainable seed and seedling systemsand sound management of agroforestry genetic resources;TM3 – tree domestication with intensificationand diversification of tree cultivation systems; andTM4 – farmer-led development, testing and scaling upof tree-based options.Each group was given five questions:• What is your 2015 vision for the focal area?• What are main problems, opportunities and assumptionsin the focal area?• What new partners and partnerships do you suggest?• Can you suggest promising new project ideas?• What types of deliverables or outputs are needed toaccomplish the results we seek?These are the conclusions that the groups reached.TM1 – market analysis and support for treeproduct enterprisesAlthough ICRAF has organized many discrete marketorientedprojects in the past that have focused onthis subject, TM1 is still a relatively new direction. In2003, there were only a few staff working in this focalarea and outputs were limited. Despite this, it becameclear that people had quite different visions for thisfield, and the first question yielded a wide range ofanswers. One member of the group stated: “ICRAFand partners should, by 2015, have identified a ‘resilientset’ of diversified tree crops (by region) withmarkets with characteristics that are capable of bufferingeach other in the face of market fluctuations.”Another mentioned that it would be important to seestrong federated farmer groups in diverse tree cropsectors. Others put the emphasis on buffering risks (bykeeping prices stable, dealing with diseases and pests,and retaining or creating diversity), building cottageindustries, the importance of tree domestication, andon finding new products to diversify traditional treeplantations.In terms of the main problems and opportunities thatneed to be tackled in the focal area there was moreconsensus. The working group identified:• A lack of credit, capital and private sector involvement• A lack of clarity about which market level should betargeted• The existence of regulatory and trade barriers• Unequal access to opportunities• A lack of ‘cottage industry’ know-how• A need for more institutional capacity, and• A lack of focus on which products to develop.The working group identified new partnership opportunitiesthat would help to advance the focal area,including with the Global Forum on Agricultural Research(GFAR), the Underutilized Crops Unit of theInternational Plant Genetic Resources Institute (IPGRI),developed-country consumer groups, national andinternational private enterprise, fair trade groups, theWorld Business Council for Sustainable Development(WBCSD), and appropriate technology institutes.

38World Agroforestry into the FutureProject suggestions within TM1 spannedthe entire range from helping farmersto start producing, to developing nichemarkets for new products. For those farmerswho do not have enough capital tostart production, the group suggested thatit would be a good idea to start a creditscheme. Further en route to market, it isthe post-harvest processing of producethat is crucial – and something that caneasily be overlooked. However, to achievereal success in the global market, thereshould be product differentiation andways of ensuring that goods are uniqueand of a high quality. One project thegroup suggested would test the appellationconcept for agroforestry products.This is a scheme employed in Europe toidentify regions of excellence for certainproducts. Another way to create a nichemarket is to promote indigenous or localproducts; for example essential oils orcraftwork. All these developments needto be supported by the right agroforestrypolicies, and therefore starting a discussionand maybe even lobbying on policyimprovement for agroforestry productswould be a step in the right direction.To reach these visions, ICRAF wouldneed a product development strategy thatincluded production of enterprise developmentmanuals, market developmentguides, consumer surveys for at least sixproducts (preferably worldwide), periodicmarket bulletins and databases of products,prices and other market information.Some of this could be accomplished byincluding market information on existingwebsites. Scientific publications, policybriefs and reports on lessons learned arealso important outputs. Some membersof the working group felt that a big mediapromotion campaign would help to getproducts and marketing processes rolling.TM2 – developing sustainable seedand seedling systems and soundmanagement of agroforestry geneticresourcesThe group discussing TM2 felt that it washard to predict the desirable situation in2015. However, they agreed that flexibilityis the key as demand for seeds and seedlingsis always changing, particularly interms of the quality, quantity, variety anddiversity of agroforestry products needed.The main problems the working groupidentified included:• Intermittent demand that is unpredictableand often disappears totally• Lack of marketplace intelligence• Inadequate information and training• Inadequate or expensive seed supply invillages• The existence of free handouts that constrainprivate sector development• Farmers not knowing their options, and• Inadequate policies and institutional arrangements.Partnerships (both new and old) with variousinstitutions were suggested by the group asways of countering these problems. Thosementioned included communities, nationalseed centres, non-governmental organizations(NGOs), research centres, local governments,private entrepreneurs (nurseryowners, seed growers), and ministerial-levelgovernment officials. To raise awareness andmoney it was suggested that grammar/primaryschools could sell tree seed and thatICRAF could publish a seed calendar. ICRAFshould make people aware of the availabilityof germplasm outside their area.To build up knowledge and expertise inthe TM2 area, a lot of research needs to bedone, and the working group suggested thatone useful project would be to look at thetree seed sector as a whole to gain soundbackground knowledge. This could lead tothe development of small-scale strategieswith regional benefits, for example wherelocally grown seed was sold to local farmersand not necessarily grown for a large,national programme. Tying in with the questionon partnerships, the working group feltthat developing strategies under the umbrellaof national agroforestry networks, whocould then take the ideas further with theirown task forces and would also enhanceour knowledge of the field.In order to ensure that these ideas arebrought to fruition, the working groupthought that what is needed the most ismore information, which can be taught byICRAF or shared between farmers. ICRAFshould also play a major role in the sharingof tree germplasm – either by organizingexchanges itself or by helping to developgermplasm supply strategies at the nationallevel.TM3 – tree domestication withintensification and diversification oftree cultivation systemsThe working group noted that farmers whogrow many different types of crops/treesreduced their environmental and economicrisk and this contributed to poverty alleviation.However, sustainability – both oflivelihoods and of traditional ways – is theimportant factor. Therefore, the the group’svision for TM3 included using domesticationstrategies to capture indigenous knowledgefor future generations. In a similar vein, itwas suggested that by 2015 we will havelearnt to appreciate natural products (e.g.natural dyes) and to recognize the untappedcommercial potential in intact ecosystems.On-farm, the group predicted that therewould be more cultivation of particular

Chapter 4: Trees and markets39tree species and intensification of foddersystems, especially for those trees that canprovide foods for famine times in droughtproneareas. Furthermore, improvements tobe made to tree domestication would comefrom using a combination of participatoryand biotechnology approaches rather thandepending on one or the other.Because there are lots of opportunitieswithin agroforestry to improve farmers’incomes and choices while reducing theirexposure to risk, it is vital that the full impactof farming decisions are known andconsidered up front. There is a serious riskthat relying on only one, or a handful ofagroforestry species could reduce choiceand push the less popular trees towards extinction.Some indigenous trees are alreadyin decline, particularly those with recalcitrantseeds that are hard to grow ex situ,which will reduce natural biodiversity aswell as degrade the indigenous knowledgethat goes with them. It is therefore importantto determine the best trees that can begrown in any situation, whether for generalor niche consumption, and what the economicreturns of each are. Furthermore, thegroup highlighted the fact that fruit treesshould be well managed (appropriatelyspaced, well fed, etc.) as well as geneticallyimproved to get the best results, andthese two processes should feed off eachother. Other important problems, opportunitiesand assumptions identified include:• Raising community awareness of agroforestrypractices and benefits• Priority setting for species selection• The need for landscape planning, and• Establishing subsidised or free localnurseries so that fruit trees can interactbeneficially with crops, inputs (especiallychemicals) are kept to a minimum,and diversity (ethnic, landscape andspecies) can be managed.There is a lot of opportunity for new partnershipswithin TM3, and the workinggroup recognized that the retention ofold partnerships (e.g. with Ministries ofAgriculture, Forestry, Health, Science andTechnology) is just as important as forgingnew ones. Any new partnerships withintree domestication should involve heightenedaccess to market information and alsoto specialist knowledge, for example inmedicinal trees.Bearing these factors in mind, the workinggroup suggested some areas where newprojects would come in useful. Some of themost important areas concern the removalof social or economic constraints, such asgender inequality or lack of micro-creditsor subsidies for small-scale farmers. Thereare quite a lot of untapped opportunitiesthat could be investigated, such as growingwood for small dimension, high qualitycabinet making, as well as the extensionof existing tree growing and craft-making.Overall there is a need to keep communitiesinvolved and to focus on local,small–medium-sized enterprises. Othersuggestions include:• More collaboration between tree usersand molecular biologists/other scientists• Identification of new technologies thatcan help with creating demand for treeproducts (e.g. in areas such as milling orfermentation), and• Establishment of an ‘endowment tree’ ineach environment, consisting of a treeor set of trees that will mature in around15 years and add value each year at anabove-interest rate.The working group perceived that one ofthe requirements to meet this future includedeveloping a participatory toolkit sothat communities can undertake participatorydomestication activities. This wouldinclude information on species characteristics,cultivation methods, products anduses, quality control and market data.TM4 – farmer-led development,testing and scaling up of tree-basedoptionsEmpowering farmers is one of ICRAF’s majoraims; consequently the working grouphad many grand visions for TM4. One ofthe recurring themes from the panel wasthat of inclusion, that everyone in a community– women, HIV/AIDS sufferers,special-interest groups and those living inmarginal areas – can participate in agroforestryadvances. And the numbers beingconsidered are quite substantial as well:one working group member predicted 10million farmers would adopt agroforestrytechnologies by 2015.Such a wide-scale and widespread adoptionshould become internally sustainableas empowered farmers start to demandmore services. Scaling up would becomea natural course of action as farmers participatemore fully in ‘bottom up’ researchand development, even establishing andrunning their own nurseries. This in turnwould create a range of new technologiesthat would be reflected in a more diverselandscape. Farmers will be able to choosewhich type of agroforestry system was bestfor them, and adapt it as necessary. By2015, the panel envisioned that agroforestrywould have advanced so far that it couldreach 20% of farmers in countries whereICRAF operates.To develop in this way, there needs to bea supportive policy environment that isregularly reviewed to identify any constraints.To further its promotion, agroforestryshould be brought into the syllabusand institutionalized. The end product of

40World Agroforestry into the Futureall this home-grown technology would,the working group predicted, be improvedlivelihoods – better health, more security,better education and more money.To reach a future with such a high potential,the working group suggested manynew project ideas. A lot of them are longterm,multi-strata projects that can take ideasfurther and deeper than before. To reallymake progress in this field, it is importantthat the fundamentals are understood.This means a lot of basic work is requiredto create inventories of useful plants (e.g.medicinal plants), document them (i.e.ethnobotany), study what works and whatdoes not, and examine participatory communicationtools used when scaling-up.There need to be adequate resources forscaling human, physical and financial capital.There should also be adequate suppliesof seed and planting material, and thatthe germplasm from threatened species isidentified and protected. At the institutionallevel it is important that there is the rightdegree of protection for intellectual property,tenure rights and partnerships and thatthe best policies exist for boosting membershipof the agroforestry community.Carrying on the principle of inclusion, it isvital that local and indigenous knowledgeand conservation issues are integrated intoscientific knowledge. Projects that focuson gender issues are also a step in the rightdirection. Furthermore, ICRAF should focuson projects that build on and foster localinnovation and which promote participatorycommunication – as well as investigatingother ways to disseminate ideas.To really achieve the widespread adoptionof agroforestry, the very way in which R&Dis carried out needs to be examined. Thisincludes looking at the nature of partnershipswithin agroforestry R&D, conductingsurveys of innovations in integrated naturalresource management and assessingfarmer-trainer systems – who carries themout, who they are aimed at and how theirdevelopment can be facilitated. Aboveall, the working group decided, there is aneed for people to understand the usefulnessof linking agroforestry products to theappropriate markets – locally, regionally,nationally and internationally.

Land and People

“Trees influence landscape scaledynamics more than any otherorganisms (although, of course,humans have now appropriatedthis claim). Investigation of thiskeystone role must remain notjust a major part of the Centre’sresearch agenda, but at the veryheart of it, because of the hugenumber of secondary interactionsthat flow from the incorporationof trees within any land usesystem.”Swift et al.

Keywords:Integrated natural resource management,scales, holism, institutionsChapter 5Confronting land degradation in Africa:Challenges for the next decadeM.J.Swift, Tropical Soil Biology and Fertility Institute of Centro Internacional de Agricultura Tropical,Kenya and Institut de recherche pour le développement, France; Ann Stroud and Keith Shepherd,World Agroforestry Centre; Alain Albrecht, World Agroforestry Centre and Institut de recherche pour ledéveloppement; Andre Bationo, Tropical Soil Biology and Fertility Institute of Centro Internacional deAgricultura Tropical; Paramu Mafongoya, Frank Place and Thomas P. Tomich, World Agroforestry Centre;Bernard Vanlauwe, Tropical Soil Biology and Fertility Institute of Centro Internacional de AgriculturaTropical; Louis V. Verchot and Markus Walsh, World Agroforestry CentreAbstractThis chapter presents five challenges for integrated natural resource management research at the WorldAgroforestry Centre within the theme of Land and People:1. What are the determinants of the adaptive and adoptive advantages of the available technologicaloptions for sustainable soil fertility management?2. How can the functions of the soil community be optimized with respect to different ecosystemservices?3. What are the trade-offs between the storage of organic matter in the soil (to counter climate changeeffects of gaseous emissions) and its use to drive nutrient cycling, crop production and other ecosystemservices?4. What are the key questions arising from interactions in the chain linking resource management–system intensification–market access–policy?5. What are the rules governing cross-scale transitions in natural resource management?These issues will require both holistic and reductionist interdisciplinary methodologies working acrossa range of scales, but many of the necessary tools have been put in place during previous work.Introduction: land degradation in AfricaThe purpose of this chapter is to identify some of the majorchallenges facing the research community concernedwith combating land degradation in Africa and its effectson human welfare. This purpose can be rephrased in theform of a question: What type of science, to do where,and for whom? There is no intent to review in detail theproblem of land degradation and its multiple causes, orthe scientific response in terms of methods and results.These have been documented in exhaustive detail else-where and reference is made to the key sources whereappropriate. Five major challenges for future researchhave been identified, but the most central issue for theWorld Agroforestry Centre remains that of investigatingthe keystone roles of trees as regulators of landscape dynamicsand providers of goods and services for humanwelfare.Land degradation has many characteristics, includingsoil erosion and nutrient depletion, decreasing quality

44World Agroforestry into the Futureand quantity of available water, and loss ofvegetative cover and biological diversity.These have knock-on effects on the prevalenceof disease – of plants, animals andhumans – and, most importantly, on humanwelfare and well-being by the disruptionof food production and other ecosystemservices. The causes of land degradationare multiple and interactive. This complexchain of cause and effect has been analysedand documented in detail; most recently inthe proposal by the Forum for AgriculturalResearch in Africa (FARA) for a ChallengeProgramme for sub-Saharan Africa and itsmany supporting documents (FARA 2003).The bottom line is that problems of thiscomplexity require holistic solutions.At different scales in space the associatedinterests of different sectors of society becomedominant and the issues of importanceto these stakeholders also change(Figure 1). It is at the plot and farm scalesthat the natural resources of soil, waterand biota are often most intensively managedand their dynamics altered by theinterventions of humans. For the farmers,therefore, the availability, quantity andquality of resources at this scale, and thefactors influencing their capacity to convertthese resources into food and marketableSCALENRMPERSPECTIVEproducts, is their major (but not their only)concern. At higher-level scales, movingthrough the hierarchy of catchments acrossthe landscape to the aggregate of riverbasins, additional issues become the concern,not only of farmers and other directland users, but also of urban society. Theseinclude ecosystem goods and services beyondfood production, such as the impactof land management on water availabilityand quality. At a global level, the effects ofland use on climate and biodiversity havebecome issues of significance. ResourceSECTORS OFSOCIETYThe challenge of scales andemergent propertiesIn proposing a holistic response to theproblems of land degradation, one of themajor structuring features must be a multiscaleapproach, embracing both space andtime. Learning to work across scales (plot,farm, land use type, landscape) with the associatedhuman perspectives (farmer, farmfamily, community, district planner, forestrymanager, etc.) is already one of the majorconcerns of the Consultative Group on InternationalAgricultural Research (CGIAR).The World Agroforestry Centre has mademajor contributions in this respect andChapters 7 and 10–13 illustrate many ofthe innovative and successful advancesthat have been made as well as addressingmany of the most important methodologicalissues (see for instance the scalar approachesin the work of the Alternatives toSlash and Burn (ASB) Programme as describedin Palm et al. 2000). Nonetheless,our facility in moving between scales, andin translating results learned on one scaleinto possible implications on scales aboveand below, remains limited.GlobePoliticalzoneCatchmentlandscapeFarmand plotBiologicaldomainsResourceconservationResourcerightsResourceallocationResourceuseResourcedynamicsGlobalconventionsNationalpoliticsCommunityrulesFarmers’decisionsScientificadviceFigure 1. Perspectives on natural resource management at different scales. The left-handcolumn provides a convenient classification of scales. The middle column indicates someof the major issues in resource management at each of these scales (although of coursethey overlap). The third column designates stakeholders with the dominant role in tropicalland management at each of the given scales.

Chapter 5: Confronting land degradation in Africa45protection becomes a feature of decisionmaking at these higher scales, sometimesconflicting with the aspirations of the farmlevelland users with respect to rights ofallocation and use. Likewise, forests can beharvested by private enterprise or government,and the resultant land cover changescan influence the availability of water tolocal people.In the following discussion we deal successivelywith challenges at the scales ofplot and farm (the most common level ofconcurrence of agricultural practice andscientific exploration); the scales below theplot where biological dynamics in the soilinfluence agricultural productivity and otherecosystem services; and finally with the‘landscape’, an aggregation of scales abovethe farm. It is perhaps necessary to note thatthis chapter is predominantly concernedwith biological aspects of the managementof natural resources. Those to do with theeconomic, social, cultural and institutionalaspects are covered in Chapters 7: ‘Scalingup the impact of agroforestry’ by Franzelet al. and 8: ‘Policies for improved landmanagement in smallholder agriculture’ byPlace et al. in this volume. This is simplya matter of convenience and, hopefully, itdoes not need to be said that the biologicalproblems associated with ecosystem serviceprovision and those of the sociology of needand acceptability are inseparable and mustbe tackled holistically.Plot and farmThe degradation of soil fertility, specificallythe capacity of the soil to support agriculturalproduction, has been identified as oneof the main causes of Africa’s agriculturalfailure (Buresh et al. 1997). It has beenrecognized that the problem of soil fertilitydegradation is a microcosm of that of landdegradation as a whole. TSBF/ICRAF (2002)states: ‘The soil fertility problem remainsintractable largely because of the failure todeal with the issue in a sufficiently holisticway. Soil fertility decline is not a simpleproblem. In ecological parlance it is a slowvariable, which interacts pervasively overtime with a wide range of other biologicaland socioeconomic constraints to sustainableagroecosystem management. It is notjust a problem of nutrient deficiency butalso of inappropriate germplasm and croppingsystem design, of interactions withpests and diseases, of the linkage betweenpoverty and land degradation, of oftenperverse national and global policies withrespect to incentives, and of market and institutionalfailures such as lack of extensionservices, inputs or credit opportunities.’Tackling soil fertility issues thus requiresa long-term perspective and a holistic approachthat integrates biological and socialelements (e.g. Swift and Palm 2000). As expressedin the African Highlands Initiative,integrated natural resource managementembodies the following (Stroud and Khandelwal2001:• principles for improving livelihoods;• inclusion of the perceptions, needs,opportunities and positions of multiplestakeholders;• formulation and adoption of strategiesto better balance the differing goals ofthose primarily concerned with environment,economic growth, equity or governance;• facilitation of institutional arrangementsand linkages within organizations andbetween various actors so as to achievebetter coordination;• fostering of synergies and informationexchange between stakeholders to promotesustainable development;• promotion of institutional and technologicalinnovations and policies thatcontribute to local ownership and stewardship;and• building upon local assets (financial,physical, knowledge and skills) topromote self-determinism and limitdependency.For more than two decades, the Centreand its partners have focused on developingtechnological options for sustainablesoil management that are biologically effective,economically viable and sociallyadoptable (Raintree 1987). This work hasproduced a substantial database of empiricalknowledge, including a series ofbooks, e.g. Young (1997), Buresh et al.(1997), Bergstrom and Kirchmann (1998),Tian et al. (2001), Vanlauwe et al. (2002),Gichuru et al. (2003), Schroth and Sinclair(2003) and Bationo (2004). From this work,a number of general lessons have emerged,with significant success in adoption andimpact (see Jama et al. Chapter 6, this volume).Most of this success has been centredround the recognition that combined use ofinorganic and organic sources of nutrientshas greater benefit than either alone (Figure2). As described in these publications, themenu of available technologies is broad andthe potential for identifying the appropriateone under a given set of conditions is nowhigh. The cropping designs that promoteintegrated nutrient management include integrationof legumes as grain or cover crops,rotations and intercrops, improved fallows,integration with livestock (i.e. use of manure)and conservation tillage. Agroforestryhas contributed successful options to manyof these generic systems. A key featureof the success in soil fertility research inAfrica over the last decade has been theintegration of ecological and participatorysocial science research.The focus on integrated nutrient managementas the basis of soil fertility managementis, of course, a relearning of an oldlesson, but one that has been accompanied

46World Agroforestry into the Futurea. 9000ControlCrop residuesb.8000 FertilizerCrop residues + fertilizer7000Total millet dry matter (kg ha —1 )6000500040003000Dry matter/N applied (kg ha —1 )120100806040200020100001982 1984 1986 1988 1990 1992 1994 1996Year0Without cropresiduesWith cropresiduesFigure 2. The effects of crop residues and nitrogen fertilizer on (a) total millet dry matter yield and (b) nitrogen use efficiency at Sadoré,Niger (modified from Bationo et al. 1999).by significant advances in scientificunderstanding and practice. These includehigh-precision fertilizer management (e.g.Baidu-Forson and Bationo 1992) and thedevelopment of knowledge-based organicmatter management practices (e.g. Palmet al. 2001 and 2002; Vanlauwe et al.2003). The target of the these practices is,of course, not only to replenish soil nutrientsand improve crop yields but also to(re)build soil fertility for long-term sustainablesoil management. This has always beena target of soil fertility management but hasgained additional impetus with realizationof the wide range of ecosystem goods andservices that stem from the maintenance ofhigh levels of organic matter in the soil. Inparticular, the sequestration of carbon insoil as a means of alleviating the climaticimpacts of excessive greenhouse gas emissionshas become a global objective (Felleret al. 2001; Albrecht and Kandji 2003).Significant questions remain, however. Itis still unclear how to predict the optimumquantitative mix of fertilizers and the variousqualities of organic inputs. Such modelsmust take into consideration the substantialsite variability commonly found onfarmers’ fields, as well as the characteristicheterogeneity of tropical climatic conditions.Furthermore, while good formulaeexist for the provision of nitrogen andphosphorus, the interaction with other nutrients,particularly micronutrients, remainslargely undefined. A significant opportunityfor development exists in this areaof research now that very high priority isbeing given internationally to plant breedingfor micronutrient provision. Most significantly,all the successes tend to remaindisappointingly local in scale. The majorchallenge remains that of determining thenecessary actions to multiply the local successesto achieve impact at a continentalscale (FARA 2003). This requires betterinsight into the relationship between thetechnology performance and the prevailingconditions of the biophysical, institutionaland socioeconomic environments. Thefirst challenge therefore addresses both theadaptive range of the technological menuand the social and economic circumstancesthat influence farmers’ willingnessto adopt.Challenge 1: What are thedeterminants of the adaptive andadoptive advantages of the availabletechnological options for sustainablesoil fertility management?

Chapter 5: Confronting land degradation in Africa47Below the farm: harnessingthe biosphereThe community of organisms in the soilperform many essential processes. They actas the primary driving agents of nutrientcycling; regulate the dynamics of soil organicmatter, soil carbon sequestration andgreenhouse gas emission; modify soil physicalstructure and water regimes; enhancethe amount and efficiency of nutrient acquisitionby the vegetation through mycorrhizaand nitrogen-fixing bacteria; and influenceplant health through the interactionof pathogens and pests with their naturalpredators and parasites. These processesalso provide a range of services to humans,including maintaining the availability andquality of water resources, erosion control,biological control of pests, climate regulationand, of course, food production.The status and activity of the soil communityhas only rarely been taken into accountin modern approaches to soil management.Any form of integrated nutrient managementnonetheless relies on the capacity ofthe decomposer community to process theorganic inputs. Similarly, the benefits ofconservation tillage rely on regulation ofthe soil’s physical condition by earthwormsand other biological ploughs. There is thusincreased awareness of the need to understandthe functioning of the living soilcommunity, as expressed in the ‘secondparadigm’ for soil fertility managementproposed by Sanchez (1994) (see also Swift1998), i.e. for improved crop growth, ‘relyon biological processes by adapting germplasmto adverse soil conditions, enhancingsoil biological activity and optimizingnutrient cycling to minimize external inputsand maximize the efficiency of theiruse’ (our emphasis). The second challengedirectly addresses the central, but largelyunrealized, clause in this paradigm.Challenge 2: How can the functionsof the soil community be optimizedwith respect to different ecosystemservices?One of the major reasons for the slowprogress in understanding the functionalbiology of the soil biota has been the lackof sensitive methods for investigating soilmicro-organisms. This problem has comecloser to solution since the advent of molecularmethods for identifying and trackingspecific soil organisms and for assessingchanges in overall biodiversity (Ammanand Kuhl 1998; Amman and Ludwig 2000).The Centre has already embarked on innovativestudies of this kind (e.g. Bossio etal. 2005). Combining such work with continuingstudies of the keystone role of treesin agroecosystems (see following section)offers great promise as a means of linkingthe driving functions of ecosystem servicesacross scales from below the plot to thelandscape.Beyond the farm: landscapesand institutionsThe biggest challenges for the managementof natural resources probably lie at thebroadest scales and can be placed underthe inexact term of ‘the landscape’. Atthese scales, predictions of effects derivenot so much from specific biological processes,but from their aggregate and interactiveeffects. The main driver of these effectsis the nature and location of different landusesystems on the landscape, includingtheir history and management. Trees influencelandscape scale dynamics more thanany other organisms (although, of course,humans have now appropriated this claim).Investigation of this keystone role mustremain not just a major part of the Centre’sresearch agenda, but at the very heart of it,because of the huge number of secondaryinteractions that flow from the incorporationof trees within any land use system.One specific feature that is strongly influencedby the presence of trees is the abundanceand quality of soil organic matter.This, in its turn, influences both soil fertilityand all the other ecosystem services derivedfrom soil. It is therefore both a majorresource and an indicator of soil status. Thework of Shepherd and Walsh in developingmethods for remote sensing and mappingof soil carbon status and its linkages toother soil properties at different scales hasprovided extremely powerful tools for bothassessing and predicting the environmentaleffects of land use change (Shepherd andWalsh 2002).Soil carbon is now of global interest becauseof the opportunity to utilize sequestrationas a mechanism for correcting theimbalances in emissions of carbonic gasesthat are believed to be driving climatechange at an unacceptable rate. However,organic matter is not an inert componentof soil. It is the substrate for many soilorganisms and thence contributes energyfor many of the essential biological processesthat support plant production andsoil structure maintenance. There is thusa major challenge to assess (for differenttypes of land use) how to optimize the useof soil carbon. This is fundamentally an issueof balancing the needs of the farmer (atthe scale of the plot) to exploit soil organicmatter energy for crop production with theneeds of society in general (at the scale ofthe landscape) to conserve carbon (Tomichet al. 2005). It is possible to hypothesizethat these two goals may be mutually incompatibleor have significant quantitative

48World Agroforestry into the Futurelimits under some conditions. Izac andSwift (1994) argued, for instance, that asustainable agricultural landscape mightnecessitate a balance between areas of exploitationof resources and areas in whichthey are permitted to accrue. This relates tothe third challenge.Challenge 3: What are the tradeoffsbetween the storage of organicmatter in the soil (to counter theclimate change effects of increasedgaseous emissions) and its use todrive nutrient cycling, crop productionand other ecosystem services?Socio-biophysical interactions are apparentat all scales, but primarily at the landscapescale where the impacts of decisionsmade by different stakeholders across arange of scales interact (Figure 1). The twoSystemwide Programmes managed by theCentre (the African Highlands Initiativeand the Alternatives to Slash and BurnProgramme) have been at the forefront indeveloping approaches and methods forassessing the interactions between environmental,economic, social and politicalfactors in natural resource management(Stroud 2001; Stroud and Khandelwal2003; Palm et al. 2005). The proposal forthe Challenge Programme for sub-SaharanAfrica drew upon these lessons by picturingan interactive chain of cause and effectin land degradation and unsustainable agriculture.This chain links the degradation ofnatural resources to failures in market accessand performance, thence to inappropriatepathways of system intensification,and finally to inadequate policies (FARA2003). The analysis provided in the ChallengeProgramme documentation serves inparticular to direct attention to the ‘interactions’between these sectors of the researchenterprise as well as to the issues withineach of them (the fourth challenge).Challenge 4: What are the keyquestions arising from interactions inthe chain linking resource management–systemintensification–marketaccess–policy?Integrating across scalesThe rules governing resource management,and the institutions making them, changeas scales change. For example, rules (orthe lack of them) made from the nationalperspective can strongly influence localbehaviour and may result in significantfeedback effects (Figure 1). This complexityis compounded by changes in dominanceof the factors determining natural resourcedynamics at different scales (e.g. van Noordwijket al. 2004 with respect to hydrologicalflows and Swift et al. 2004 in relation tothe significance of biodiversity). The type ofmanagement (communal or individual, governmentor private) apparent within the landuse types and the gender and wealth dimensionis also important. Thus, a range ofsocial parameters enters the equation. Theseissues have been analysed by the CGIARTaskforce on Integrated Natural ResourceManagement and the reports and papersemanating from that group (e.g. Campbelland Sayer 2003; Sayer and Campbell 2001)together with the framework developed byIzac and Sanchez (2001) offer some of thebest analyses of the methods, approaches,successes, opportunities and challenges thatface a scientific community committed toissues of ‘Land and People’.In a recent study of watershed managementissues that cut across scales and social perspectivesin the East African Highlands (reportedby German 2003 and Stroud 2003),five main categories were identified:1. Issues involving the management ofcommon property resources which compromiseeither the quantity or quality ofthese resources.2. Issues involving limited access andinequitable distribution of resources(absolute and relative shortages).3. Trans-boundary effects between neighbouringfarms and villages.4. Areas in which collective action couldsignificantly enhance farm productivity,either through increased access toproductive resources (natural resources,labour, capital) or through cooperationto conserve resources that are underthreat (biodiversity, local knowledge).5. Areas in which collective action iscurrently needed to enhance incomeor livelihood more broadly (publicworks, governance of existing resources,marketing).Of equal interest to variation across spatialscale is the influence of change over differentscales in time. Crowley and Carter(2000) provided a detailed and perceptiveanalysis of the historical factors that haveinfluenced the current state of natural resourcesand agricultural practice in westernKenya. Such analysis influences an importantdebate in natural resource managementresearch. It is often asserted that the characteristicsof the natural resource base andits management are highly site-specific, anobservation largely derived from the hugebiophysical variation that is commonly seenbetween neighbouring fields with respectto soil fertility status and other biologicalproperties. These observable differencesmay derive, in some cases, from variationsin underlying materials, but are more frequentlya product of the history of humanmanagement of the natural resources ofthe plots, farms and regions concerned in

Chapter 5: Confronting land degradation in Africa49response to risks posed by weather, market,food and feed needs, energy and land usepolicies, etc. This paints a potentially chaoticpicture resulting from dynamic evolutionover time. The evolutionary biologist,Stephen Jay Gould, has made an eloquentplea for scientists to appreciate the importanceof historical analysis as an integraltool in the biological sciences (Gould2000). In particular, he points to the importanceof understanding the degree to whichpresent conditions are contingent on eventsthat occurred in the past. Whilst his argumentsare largely concerned with the processesof biological evolution, they surelyalso apply to the development of ecologicalsystems over time, particularly those influencedby agriculture.The recognition of site specificity at the plotlevel has led to the pessimistic assertionthat there is therefore no opportunity forgeneric scientific or technological solutionsto natural resource management problems.This is a confusion of principle and practice.It is certainly now generally accepted thatmonolithic zonal technology recommendations(e.g. for fertilizer dosage) are ineffective.They have been largely replaced bymenus of multiple options, and the choiceof option is determined by local conditions.However, the origins of the menu optionsare no less based on scientific principlesthan are (for example) those created by producingdifferent crop genotypes. Indeed, itcould be argued that failures in realizing thepotential of genotypes have often resultedfrom failure to recognize the environmentalvariations that are taken for granted in naturalresource management research.This brief analysis of multiscale issues inspace and time serves to emphasize fourcross-cutting issues:1. Recognition of the hierarchical linkagesacross scales and their interactions in termsof problems and potential solutions (Swift1999).2. The value of understanding the historicalbasis of present conditions.3. The importance of merging biological,social and institutional analyses in order tounderstand the dynamics of influence bothacross and within scales.4. The value of identifying ‘entry’ points forresearch and intervention, i.e. simplifyingaccess to the complexity of interactive effectswithin any natural resource managementproblem by tackling them throughaccessible and influential components.These issues can be summarized as a fifthchallenge.Challenge 5: What are the rulesgoverning cross-scale transitions innatural resource management?Conclusion: what type ofscience, to do where, and forwhom?The five specific challenges presented aboveprovide a response, but by no means a completeanswer, to the question posed in theopening paragraph. The greatest challengefor any institution whose role is science fordevelopment is that of choice: choice ofone scientific topic versus another; choiceof criteria for research locations; choice ofwhat type of scientific approach to use; andchoice of which of the myriad stakeholdersto work directly with, in what manner andacross what scales in space and time.The scientist in any development-relatedtopic will always be faced by decisions asto where to place her or his activities in theresearch-to-adoption spectrum. Whether toconcentrate on relatively basic research, removedfrom the ultimate client but generatingknowledge that may open up areas ofprogress hitherto inaccessible; or to focuson actions to disseminate knowledge andtechnology that interact directly with, andprovide identifiable benefits for, a selectedgroup of such clients. This dilemma is farfrom peculiar to international agriculturalresearch. The Nobel Prize laureate, immunologistand incisive writer on the philosophyof science, Sir Peter Medawar, picturedtwo ‘Conceptions of Science’ that exist inthe popular imagination (Medawar 1982).He described the ‘Romantic Conception’with the words of the English poet and essayist,Samuel Taylor Coleridge: “The firstman of science was he who looked into athing, not to learn whether it could furnishhim with food, or shelter, or weapons, ortools, or play-withs, but who sought toknow it for the gratification of knowing.”Medawar contrasted this concept withwhat we may term a ‘Pragmatic Conception’:“Science above all else [is] a criticaland analytical activity; …scientific researchis intended to enlarge human understanding,and its usefulness is the only objectivemeasure of the degree to which it does so.”Medawar acknowledged that these two descriptionswere caricatures and concludedthat: “Anyone who has actually done orreflected deeply upon scientific researchknows that there is in fact a great deal oftruth in both [conceptions].” Internationalagricultural research must indeed serveboth these pursuits. By choosing to workfor the poorest of the poor we have alreadychosen to follow the Pragmatic Conception.But our contribution is likely to begreater by applying what we are most suitedto do – to exercise our curiosity and imaginationto empower people with the bestthat the scientific adventure can provide.The discussion in the preceding sectionsargues in almost every part for a holisticapproach to agricultural research for

50World Agroforestry into the Futuredevelopment – and this is perhaps the mostexacting scientific challenge of all. Theobvious danger in any broad approach,however, is that of working on absolutelyeverything and losing all useful focus. Ecosystemecology is helpful in this regard, forit has become clear that, whatever the levelof complexity, there are always some featuresof the system that are more influentialin its regulation than others. This existenceof keystone organisms and processes isparticularly good news for agroforesters,because asking the question ‘what is therole of the tree(s)?’ in any given ecosystem isunlikely to be distant from asking ‘how doesthis system work?’ Likewise, the disciplinesof anthropology and sociology draw uponknowledge of repeated patterns in socialbehaviour and actions that can be linked tobio-political processes. The key for the resourcemanagement scientist in agriculturalresearch is thus to identify key features ofthe system during the diagnostic phase ofa research programme, and to derive appropriateentry points to the research thatwill enable manipulation of these keystoneorganisms, processes or properties.In seeking to wade in the ‘bathwater’ ofholistic science we should not, however,cast out the ‘baby’ of reductionism. Asmentioned above, Stephen Jay Gouldargued forcefully throughout his career fora wider and less mechanistic concept ofwhat constitutes the scientific method thanmere reductionist experimentation. Yet, inso doing, he also said: “Only a fool or anenemy of science could possibly deny theextraordinary power and achievements ofreductionism…” (Gould 2004). It is thussurely not a question of either reductionismor holism, but the challenge of when andin what manner to employ the power ofthe former within a framework of the latter.As outlined above, the tools to assist suchchoices are already part of the World AgroforestryCentre’s armoury – what remains ishow we choose to use them.AcknowledgementsOur grateful thanks go to Antonia Okonofor her assistance beyond the call of dutyin the preparation of this chapter.ReferencesAlbrecht, A. and S.T. Kandji 2003. Carbonsequestration in tropical agroforestrysystems. Agriculture, Ecosystems andEnvironment 99: 15–27.Amman, R. and M. Kuhl 1998. In situ methodsfor assessment of microorganisms andtheir activities. Current Opinion in Microbiology1: 352–358.Amman, R. and W. Ludwig 2000. RibosomalRNA-targeted nucleic acid probes forstudies in microbial ecology. FEMS MicrobiologyReviews 24: 555–565.Bationo, A. (ed) 2004. Managing Nutrient Cyclesto Sustain Soil Fertility in Sub-SaharanAfrica. African Academy of Science Publishers,Nairobi, Kenya.Bationo, A., S.P. Wani, C.L. Bielders, P.L.G. Vlekand A.U. Mokwunye 1999. Crop residueand fertilizer management to improvesoil organic carbon content, soil qualityand productivity in the desert margins ofWest Africa. In: Lal, R., J.M. Kimble andB.A.Stewart (eds) Global Climate Changeand Tropical Ecosystems. Advances in SoilScience. CRC Press, Florida, USA.Baidu-Forson, J. and A. Bationo 1992. Aneconomic evaluation of a long-termexperiment on phosphorus and manureamendments to sandy Sahelian soils usinga stochastic dominance model. FertilizerResearch 33: 193–202.Bergstrom, L. and H. Kirchman (eds) 1998.Carbon and Nutrient Dynamics in Naturaland Agricultural Tropical Ecosystems. CABInternational, Wallingford, UK.Bossio, D., M.S. Girvan, L. Verchot, J. Bullimore,T. Borelli, A. Albrecht, K.M. Scow, A.S.Ball, J.N. Pretty and A.M. Osborn 2005.Soil microbial community responseto land use change in an agriculturallandscape of western Kenya. MicrobialEcology 49: 50–62Buresh, R.J., P.A. Sanchez and F. Calhoun 1997.Replenishing Soil Fertility in Africa. SoilScience Society of America, Madison, WI,USA.Campbell, B.M. and J.A. Sayer (eds) 2003. IntegratedNatural Resource Management:Linking Productivity, the Environment andDevelopment. CAB International, Wallingford,UK.Crowley, E.L. and S.E. Carter 2000. Agrarianchange and the changing relationshipsbetween toil and soil in Maragoli, westernKenya (1900–1994). Human Ecology 28:383–414.FARA 2003. Securing the Future for Africa’sChildren: Building sustainable livelihoodsthrough research for integrated agriculturalresearch for development. Forum for AgriculturalResearch in Africa, Accra, Ghana.Feller, C., A. Albrecht, E. Balnchart, Y.M. Cabidoche,T. Chevallier, C. Hartmann, V.Eschenbrenner, M.C. Larre-Larrouy andJ.F. Ndandou 2001. Soil organic carbonsequestration in tropical areas. Generalconsiderations and analysis of someedaphic determinants for Lesser Antillessoils. Nutrient Cycling in Agroecosystems61: 19–31.German, L. 2003. Beyond the Farm: A new lookat livelihood constraints in the highlandsof eastern Africa. African Highlands InitiativeBrief A2. World Agroforestry Centre,Nairobi, Kenya.Gichuru, M.P., A. Bationo, M.A. Bekunda, H.C.Goma, P.L. Mafongoya, D.N. Mugendi,H.K. Murwira, S.M. Nandwa, P. Nyathiand M.J. Swift (eds) 2003. Soil Fertility

Chapter 5: Confronting land degradation in Africa51Management in Africa: a Regional Perspective.African Academy of Sciences,Nairobi, Kenya.Gould, S.J. 2000. Wonderful Life: The BurgessShale and the Nature of History. Vintage,London, UK.Gould, S.J. 2004. The Hedgehog, the Fox and theMagister’s Pox: Mending and Minding theGap Between Science and the Humanities.Vintage, London, UK.Izac, A-M.N. and P.A. Sanchez 2001. Towards anatural resource management paradigm forinternational agriculture: the example ofagroforestry research. Agricultural Systems69: 5–25.Izac, A-M.N. and M.J. Swift 1994. On agriculturalsustainability and its measurement insmall scale farming systems in sub-SaharanAfrica. Ecological Economics 11: 105–125.Medawar, P. 1982. Pluto’s Republic. Oxford UniversityPress, Oxford, UK.Palm, C.A., Izac, A-M.N. and S.A. Vosti 2000.Procedural Guidelines for Characterization.Technical Report. World AgroforestryCentre: Nairobi, Kenya. http://www.asb.cgiar.org/Palm, C.A., K.E. Giller, P.L. Mafongoya and M.J.Swift 2001. Management of organic matterin the tropics: translating theory into practice.Nutrient Cycling in Agroecosystems61: 63–75.Palm, C.A., C.N. Gachengo, R.J. Delve, G.Cadisch and K.E. Giller 2002. Organicinputs for soil fertility management intropical agroecosystems: application of anorganic database. Agriculture, Ecosystemsand Environment 83: 27–42.Palm, C.A., S.A. Vosti, P.A. Sanchez and P.J. Ericksen(eds) 2005. Slash and Burn: The Searchfor Alternatives. Columbia University Press,New York, USA.Raintree, J.B. 1987. D & D User’s manual: AnIntroduction to Agroforestry Diagnosis andDesign. World Agroforestry Centre, Nairobi,Kenya.Sanchez, P.A. 1994. Tropical soil fertility research:towards the second paradigm. Transactionsof the 15th World Congress of Soil Science,10–16 July, Acapulco, Mexico.Sayer, J. and B.M. Campbell 2001. Researchto integrate productivity enhancement,environmental protection, and humandevelopment. Conservation Ecology 15(2)available online at: http://www.consecol.org/vol5/iss2/art1Shepherd, K.D. and M.G. Walsh 2002. Developmentof reflectance spectral libraries for charcterisationof soil properties. Soil ScienceSociety of America Journal 66: 988–998.Schroth, G. and F.L. Sinclair 2003. Trees, Cropsand Soil Fertility: Concepts and ResearchMethods. CAB International, Wallingford,UK.Stroud, A. 2001. End of Phase I (1995–1997). AfricanHighlands Initiative Technical Report.World Agroforestry Centre, Nairobi, Kenya.Stroud, A. and R. Khandelwal 2003. In Searchof Substance: an analytical review of conceptsand approaches in natural resourcemanagement in sub-Saharan Africa. AHIWorking Paper No 20. African HighlandsInitiative, Nairobi, Kenya.Swift, M.J., A-M.N. Izac, and M.N. van Noordwijk2005. Biodiversity and ecosystemservices in agricultural landscapes – are weasking the right questions? Agriculture, Ecosystemsand Environment 104: 113–134.Swift, M.J. 1998. Towards the second paradigm:integrated biological management of soil.In: Siqueira, J.O., F.M.S. Moreira, A.S.Lopes, L.G.R. Guilherme, V. Faquin, A.E.Furtini Neto and J.G. Carvalho (eds) SoilFertility, Soil Biology and Plant NutritionInterrelationships. Sociedade Brasileira deCiencia do Solo/Universidade Federal deLavras/Departamento de Ciencia do Solo(SBCS/UFLA/DCS), Lavras, Brazil.Swift, M.J. 1999. Integrating soils, systems andsociety. Nature and Resources 35: 12–20.Swift, M.J. and C.A. Palm 2000. Soil fertility as anecosystem concept: a paradigm lost andregained? Proceedings of the InternationalUnion for Soil Sciences Symposium, April2000, Bangkok, Thailand.Tian, G., F. Ishida and D. Keatinge 2001. SustainingSoil Fertility in West Africa. Soil ScienceSociety of America, Madison, USA.Tomich, T.P., A. Cattaneo, S. Chater, J.H. Geist, J.Gockowski D. Kamowitz, E.F. Lambin, J.Lewis, O. Ndoye, C.A. Palm, F. Stolle, W.D.Sunderlin, J.F. Valemtin, M. van Noordwijkand S.A. Vosti 2005. In: Palm, C.A., S.A.Vosti, P.A. Sanchez and P.J. Ericksen (eds)Slash and Burn: the Search for Alternatives.Columbia University Press, New York,USA.TSBF/ICRAF 2002. Soil Fertility Degradationin sub-Saharan Agriculture: Leveraginglasting solutions to a long-term problem.Conclusions from a workshop held at theRockefeller Foundation Bellagio Study andConference Centre, 4–8 March, TropicalSoil Biology and Fertility Programme,Nairobi, Kenya.Vanlauwe, B., J. Diels, N. Sanginga and R.Merckx 2002. Integrated Plant NutrientManagement in Sub-Saharan Africa: fromConcept to Practice. CAB International,Wallingford, UK.Vanlauwe, B., C.A. Palm, H.K. Murwira and R.Merckx 2003. Organic resource managementin sub-Saharan Africa: validation ofa residue quality-driven decision supportsystem. Agronomie 22: 839–846.van Noordwijk, M., J. Poulsen and P.J. Ericksen2004. Quantifying off-site effects of landuse change: filters, flows, and fallacies.Agriculture Ecosystems and Environment104: 19–34.Young, A. 1997. Agroforestry for Soil Management(2nd Edition). CAB International,Wallingford, UK and World AgroforestryCentre, Nairobi, Kenya.

Keywords:Agroforestry, improved fallows, trees,biomass transfer, nitrogen, phosphorusChapter 6Agroforestry innovations for soil fertilitymanagement in sub-Saharan Africa:Prospects and challenges aheadBashir Jama, World Agroforestry Centre, Kenya; Freddie Kwesiga, World Agroforestry Centre, Zimbabwe andAmadou Niang, World Agroforestry Centre, MaliAbstractImproving soil fertility is a key entry point for achieving food security, reducing poverty and preservingthe environment for smallholder farms in sub-Saharan Africa. Given the high cost of inorganic fertilizers,an integrated approach that combines promising agroforestry technologies – particularly improvedfallows and biomass transfer – with locally available and reactive phosphate rock – such as the Minjinguof northern Tanzania – is described in this chapter. Leguminous tree fallows of several speciescan accumulate significant amounts of nitrogen in their leaves in the short duration (from 6 months to2 years). Incorporating these leaves into the soil before planting can increase crop yields several-fold.Improved fallows can also contribute to the control of weeds (including Striga hermontheca) and providewood for energy and for staking climbing crops. Some of the species also have fodder value thatcan improve manure quantity and quality. For biomass transfer, use of Tithonia diversifolia is the mostpromising because of its high nutrient content and rapid rates of decomposition. This plant is now beingused more widely for high-value crops such as vegetables. To facilitate the scaling up of these fertilityoptions, future research and development needs to address recommended application rates, impacts atboth farm and landscape levels, and the method by which high-value trees, crops and livestock can beintensively farmed to provide a natural progression out of poverty.BackgroundLand degradation and declining soil fertility are increasinglybeing viewed as critical problems affectingagricultural productivity and human welfare in tropicalAfrica. It is estimated that an average of 660 kg ofnitrogen (N) ha –1 , 75 kg of phosphorus (P) ha –1 and 450kg of potassium (K) ha –1 have been lost during the last30 years from around 200 million ha of cultivated landin 37 countries in sub-Saharan Africa (SSA) (Stroorvogelet al. 1993). The estimated value of such losses averagesabout US$4 billion per year (Drechsel and Gyiele1999). This figure is probably higher than the annualofficial development assistance given to the agriculturalsector in Africa during the last three decades.The underlying socioeconomic causes of nutrient depletion,their consequences and the various strategiesfor tackling this constraint are fairly well known (Bureshet al. 1997; Smaling 1993). It is important, however,to underscore that the spatial distribution of nutrientdepletion in Africa is not uniform at regional or farmscales. Regions that have not been subject to intensive,

54World Agroforestry into the Futurecontinuous cultivation, or which have awidespread history of fertilizer use, do notexhibit this problem (Scoones and Toulmin1999). Localized differences in farmerwealth ranking, field-use history and theuse of organic additions (typically in fieldsclose to the homestead) generally produce‘islands’ of high soil fertility (Shepherd andSoule 1998.Given the acute poverty and limited accessto mineral fertilizers, a promising approachis one that integrates organic and inorganicfertilizers. Organic fertilizers include theuse of improved fallows of leguminoustrees, shrubs, herbaceous legumes andbiomass transfer. The improved fallows systemis the product of more than 10 years’of agroforestry research and developmentefforts by the World Agroforestry Centre(ICRAF) and its many partners in SSA. Bothresearch and development dimensions arediscussed in this chapter. We do this bydrawing particular reference to the Centre’scollaborative work in three regions ofAfrica – East and Central Africa, southernAfrica and the Sahel. Declining soil fertilityis a major concern faced by smallholderfarmers in all these regions (Franzel 1999;Sanchez and Jama 2002).Improved fallowsThe concept and practiceAlthough neither the idea nor the researchon improved fallows is new (Nye andGreenland 1960), critical examination ofthe practice, and the wide-scale evaluationof suitable species, is relatively recent(Sanchez 1995). Planted fallows of leguminoustrees or shrubs can biologically fixconsiderable amounts of N – for example,between 60–80 kg ha –1 – in above-groundbiomass (Gathumbi 2000). The rest of therecycled N in such leguminous trees orshrubs is accessed from sub-soil N – Oxisolsand Oxic Alfisols – which is unavailableto crops (Mekonnen et al. 1997). Underconditions such as those in western Kenya,where the soils possess substantial anionexchange capacity, net N mineralizationexceeds N uptake by crops and high rainfallcarries nutrients to the sub-soil, resultingin a build-up of sub-soil N that rangesfrom 70 to 315 kg ha –1 (Hartemink et al.1996). Nitrogen that accumulates in theabove-ground biomass of planted tree fallowsis returned to the soil upon clearing;the fallow biomass is incorporated into thesoil for subsequent cropping. Additionally,fallows increase the amount of labile fractionsof organic soil matter, which supplynutrients to crops following fallows (Barrioset al. 1997). They can also contribute toimproving soil structure, build up of soilorganic matter and its carbon (C) stocks,thus contributing to C sequestration.The choice of which species to plant inthe fallow period is influenced by bothbiophysical and socioeconomic conditions.The ideal tree species is typicallyfast-growing, N-fixing and efficient atnutrient capture and cycling. Examples ofpromising species include Crotalaria grahamiana,Tephrosia vogelii, Cajanus cajan(pigeonpea) and Sesbania sesban (sesbania).Coppicing species can also be used,and Gliricidia sepium (gliricidia) and Calliandracalothyrsus (calliandra) are becomingincreasingly popular with farmers inKenya, Malawi and Zambia because theyare perennial and, unlike the non-coppicingspecies, there are no costs involved inreplanting them once they are cut back.Agronomic and economic benefitsSeveral agronomic studies demonstratethat improved fallows of 1–3 seasons (8–21months) can increase soil fertility and improveyields considerably. For instance,Kwesiga and Coe (1994) in premier fieldstudies demonstrated that 2- and 3-yearsesbania fallows can increase maize yield,compared to unfertilized maize monoculture,for at least three cropping seasons afterharvest of the fallows on an N-deficientsoil in Zambia. This was confirmed laterin multilocational trials in eastern Zambia(Kwesiga et al. 2003). In western Kenya,similar observations have also been madewith use of several species and fallowdurations (Jama et al. 1998a; Niang et al.1996a; Rao et al. 1998). Recent trials in theSahel that were conducted within the subhumidregion of Mali also demonstrate theability of several species to improve soilfertility and crop yields considerably (Figure1). These studies have led to the generalconclusion that total farm productioncan be greater with improved fallow–croprotations than with continuous cropping,even though crop production is skipped forone or more seasons with improved fallows(Sanchez et al. 1997).In areas such as southern Malawi with lowrainfall and sandy soils, gliricidia fallowsthat coppice when cut back perform betterthan those of sesbania, which does notcoppice well. This has been demonstratedthrough long-term trials that also show thatthe highest yields are obtained when improvedfallows are used in conjunction withrepeated application of the recommendedrates of inorganic fertilizers (Figure 2).In soils that are severely depleted of nutrients,the addition of inorganic fertilizersincreases the productivity of improved fallows.In western Kenya, for instance, thereis increasing evidence that 1–2 season-longfallows do not overcome N deficiency inhighly degraded soils, especially when deficienciesof other nutrients are overcomeand when high-yielding crop varieties areused. Fertilizer use is, however, limited by

Chapter 6: Agroforestry innovations for soil fertility management in sub-Saharan Africa55Sorghum grain yield (t ha —1 )2.52.01.51.00.50.01.4Figure 1. Effect of improved fallows using different species on the grain yield of sorghumat Bamako, Mali.Source: Niang (unpublished data).2.31.1T.candida S.sesban C.siberia C.cajan NaturalfallowCropping system1.21.30.6Continuouscropits high cost. In SSA, fertilizers cost around3–4 times the international price largelybecause of poor roads and the associatedhigh transport costs in many countries.However, fertilizers are needed for theintegrated nutrient management approachproposed for replenishing soil fertility inAfrica, and hence should be made affordableto farmers.Economic analysis indicates that improvedfallows are generally attractive (Franzel etal. 1999; Swinkels et al. 1997). Accordingto sensitivity studies conducted by Placeet al. (2000) in eastern Zambia, which isprone to droughts, this is the case even underdrought conditions. In western Kenya,however, economic benefits are marginal.Even though the soils in this region are P-deficient and require application of P-richfertilizers, that are prohibitively expensive(costing more than US$500 t –1 ).Maize grain yield (t ha —1 )8765432= SedOther benefitsControl of Striga hermontheca, a parasiticweed of many cereal crops, is an addedbenefit of the repeated use of improved fallows(Barrios et al. 1998; Gacheru and Rao2001). Striga causes large yield losses inthe Lake Victoria area of the East and CentralAfrica basin. Although the processesare not well understood, it is suspected thatthe fallow species excrete substances thatcause suicidal early germination of Striga.101996 1997 1998 1999 2000 2001 2002 2003Figure 2. Maize grain yield for eight seasons using Sesbania sesban and Gliricidia sepiumfallows. Sed = Standard error of difference of means. M+F = Maize that was fertilized with200 kg ha –1 of a compound fertilizer (N = 100 g kg –1 , P = 90 g kg –1 and K = 80 g kg –1 )at sowing, and 92 g N ha –1 as urea 4 weeks after emergence; M-F is maize not fertilized.Source: Kwesiga (unpublished data).Years/seasonGliricidia M+F M–F SesbaniaThe provision of fuelwood is another benefitof improved fallows. Depending on thespecies and fallow duration, considerableamounts of wood can be obtained fromimproved fallows. For instance, in westernKenya, calliandra, which produces woodwith good fuelwood properties, can generatemore than 10 t ha –1 of wood from asearly as the third year of establishment. Thisis enough to meet the fuelwood needs of atypical rural household with 6–7 members

56World Agroforestry into the Futurefor 6–8 months. The wood can also be usedto support climbing beans and other climbingcrops.LimitationsTwo important factors that need to be consideredwith short-term improved fallowsystems are: how much leaf biomass thefallow species produces and the quantityof nutrients recycled with it. Several factorsinfluence biomass production. In degradedsites (nutrient-depleted and eroded), mostfallow species grow poorly and producelittle biomass. This is also the case in dryareas and those with Vertisols (heavy clays)that drain poorly during the wet season.Such conditions prevail around the LakeVictoria basin where leaf biomass yieldsare typically less than 1 t ha –1 . This willgive less than the 80–100 kg N ha –1 requiredto produce a 2 t ha –1 maize grainyield (Palm 1995). There are options thatcan be explored to increase biomass yieldwithout necessarily increasing the fallowperiod. These include the use of coppicingspecies, and under-sowing the treefallow with herbaceous green manurelegumes such as mucuna (Mucuna puriens)and macroptilium (Macroptilium atropurpureum).In P-depleted soils, trees respondto P application and can benefit from havingP applied to crops planted within them(Jama et al. 1998b).The incidence of pests and diseases is anotherimportant limitation and there aretwo aspects to this problem. Firstly, thereare pests and diseases that affect the treesthemselves and limit their productivity. Forexample, sesbania is damaged, sometimesseverely, by the defoliating beetle Mesoplatysochroptera. Crotalaria grahamiana,until now a promising species for improvedfallows in western Kenya, is attackedand defoliated severely by lepidopterousAmphicallia pactolicus caterpillars. Controllingthese pests is vital to ensure thatthe productivity of species used and promotedfor improved fallows is maintained.Secondly, there is need to understand andcontrol the effects of these pests on thecrops that succeed the fallows. A case inpoint are the root-knot nematodes associatedwith sesbania that also affect beansand tomatoes (Desaeger and Rao 2000).There is also the potential for some of thespecies used in fallows to become invasiveweeds – although no such occurrence hasbeen reported so far. Prolific seeders likecrotalaria and leucaena species are examplesof the types of fallow plants most likely tobecome problematic. Other species may startto seed prolifically when taken out of theirecological range. Thus control mechanisms,including prevention, early detection andrapid response, need to be developed. Thisrequires cross-regional collaborative efforts.Biomass (green manure)transferApart from improved fallows, existinghedges on farm borders are another sourceof organic nutrients for biomass transfer.More than 10 species with potential for thispurpose have been screened in westernKenya (Niang et al. 1996b), and the mostpromising of all is Tithonia diversifolia ofthe family Asteraceae (tithonia). Althoughit is not a legume, the fresh leaf biomassof tithonia has levels of N as high as thosefound in many N-fixing legumes. This commonshrub is also rich in P and K: the freshleaves contain 3.5% N, 0.3% P and 3.8%K. The leaf biomass decomposes rapidlywith a half-life of about one week especiallyduring the rainy season (Gachengo1996).Many field studies report that the applicationof tithonia biomass results in highercrop yields than application of inorganicfertilizers, and it has longer residual effects(Gachengo 1996; Jama et al. 2000). Part ofthe yield benefits associated with tithoniacould be due to increased availability ofnutrients. Phosphorus release from tithoniafresh-leaf biomass is rapid, and the supplyof plant-available P from tithonia can be atleast as effective as an equivalent amountof soluble fertilizer. Nziguheba et al. (1998)reported that incorporation of green tithoniabiomass equivalent to 5 t dry matter ha –1to an acid soil in western Kenya increasedP in soil microbial biomass and reducedP sorption by soil (Table 1). In this study,the plots were kept free of weeds and notcropped in order to eliminate plant uptakeof P as a factor affecting soil P fractions andprocesses. Increased P in soil microbial biomass2 weeks after tithonia incorporationpresumably indicates enhanced biologicalcycling and turnover of P in labile poolsof soil P. Enhanced microbial biomass Pfollowing integration of tithonia with triplesuperphosphate, and not with sole applicationof triple superphosphate, supports thehypothesis that tithonia increases soil labileP. Soil microbial P before maize plantinghas been shown by Buresh and Tian (1997)to be directly correlated to maize yield ona P-deficient soil in western Kenya.Availability of sufficient quantities of tithoniabiomass and the labour required toharvest and transport it to cropped fieldsare likely to be two major constraints tothe wide-scale adoption of this technologyby farmers. Recognizing these limitations,most farmers in western Kenya are usingtithonia on small parcels of land and onhigh-value crops such as tomato and kale(Brassica oleraceae var acephala; ICRAF1997). They are also experimenting withtithonia in maize–bean (Phaseolus vulgaris)intercrops, where it could be more financiallyattractive than in sole maize because

Chapter 6: Agroforestry innovations for soil fertility management in sub-Saharan Africa57Table 1.Effects of 15 kg P ha –1 as either green tithonia biomass or triple superphosphate (TSP) on increase in microbial biomass P anddecrease in sorbed P at 0.2 mg P L –1 solution.Weeks afterapplicationIncrease in microbial biomass P (mg P kg –1 ) Reduction in sorbed P (mg P kg –1 )Tithonia TSP Tithonia + TSP Tithonia TSP Tithonia + TSP2 4.3 ** 1.8 7.8 ** 49 ** 41 ** 30 *16 1.6 0 3.7 ** 27 * 10 * 20* indicates significance at P = 0.05 and ** at P = 0.01. All values are relative to a control with no added TSP or tithonia.Source: Nziguheba et al. (1998)beans are of higher value than maize. InZambia, farmers are doing the same in the‘dambos’ (wetlands) during the dry season(Kuntashula et al. 2004). In Mali, it is increasinglybeing used for vegetable farmingin urban and peri-urban agriculture. Indeed,economic analyses indicate positivereturns from the use of tithonia on highvaluevegetables but not for low-pricedmaize (ICRAF 1997).Based on feedback from farmers, researchon tithonia is now focused on several issuesof practical importance: i) identifyingways of increasing its production on-farmby growing it in small niches such asaround farm boundaries and in soil conservationstructures; ii) integrating it withinorganic fertilizers to reduce the requiredquantities of each material; iii) using it tocomplement low-quality organic materialssuch as crop residues and farmyard manurethat are used as fertilizers; iv) identifyingthe minimum acceptable quantities oftithonia for application to vegetables andcereals; and v) optimizing its use efficiencythrough timely application and appropriateplacement.Livestock manureFor smallholder farms, farmyard manureis a major source of nutrients. However,quality is poor and quantities available areoften low, especially in densely populatedregions like western Kenya where farmerskeep few animals (Kihanda and Gichuru1999). Quality can be improved throughbetter management, including feedingnutrient-rich tree fodder to cattle. Manurefrom livestock fed with calliandra foddercan be especially high in P, for example, asdemonstrated through studies in westernKenya (Jama et al. 1997). Application of thismanure at rates typically used by farmers inthe area more than doubled maize yieldsin P-deficient soils, and effects were evengreater when it was spot applied (placedin the planting hole) instead of broadcast.However, much more assessment is neededon improvements in tree fodder and manurequality, including a better understandingof their interaction with inorganic fertilizersand how they affect overall householdeconomic conditions.Need for phosphorus inputsPhosphorus deficiency is widespread inSSA. This is particularly pronounced in westernKenya where, for instance, more than 80percent of the farms are severely deficientin P, with less than three parts per millionof available P when analysed by the Olsenprocedures. As a consequence, crop yieldsremain low. Under these conditions, P inputis a must if crop yields are to be improved.Although trees can add some P to the soil,this is mostly by recycling what is alreadythere and not through new additions. Theexception is biomass transfer. Even then, theamounts that can be added through the biomassof trees are often low.Options for P inputs are phosphorous fertilizersand phosphate rock (PR), dependingon which is cost-effective. There are severalPR deposits in Africa that could be of agronomicuse (van Straaten 2000), for examplethe Tilemsi in Mali and the Minjingu innorthern Tanzania. The agronomic effectivenessof Minjingu rock phosphate wasexamined in long-term (5-year) field trialsin western Kenya. Two different strategies ofphosphorus application were compared –a large one-time application (250 kg P ha –1 )that is expected to provide a strong residualeffect for at least 5 years, and annual applicationsof 50 kg P ha –1 applied to the rainyseasonmaize crop. Over the 5 years ofthe study, cumulative maize yield was significantlyincreased by P fertilization, andthe cumulative grain yields were almostthe same, regardless of which P source orapplication method was employed (Figure3). This clearly demonstrates the utility ofMinjingu and other reactive PRs in soil fertilitymanagement approaches in SSA.

58World Agroforestry into the FutureCumulative maize grain yield (t ha —1 )252015105baaaa0No P MPR MPR TSP TSP0 50 x 5 250 x 1 50 x 5 250 x 1P source and rateFigure 3. Cumulative maize yields over 5 years (five crops, ‘long rains’ cropping season only) in western Kenya. Nitrogen (N) andpotassium (K) were supplied. Minjingu phosphate rock (MPR) or triple superphosphate (TSP) fertilizer were added, either once at 250 kgP ha –1 , or at a rate of 50 kg ha –1 each year for 5 years. There was also a control plot with no added P.Source: Sanchez and Jama (2002).Conclusion and way forwardImproved fallows of leguminous speciesand biomass transfer are both promisingagroforestry techniques that can contributeto integrated soil fertility managementpractices in smallholder farms. They canalso provide other benefits such as controlof pests and diseases, and in the case ofimprove fallows, provide fuelwood that isin short supply in many rural settings. Toenhance the impact of these technologies,there are a few remaining challenges thatneed to be addressed. These include:1. Determining the recommendation domains(geographic areas and householdtypes where the technologies are feasibleand profitable), something that is necessarygiven the large biophysical and socioeconomicvariability that exists within andbetween farms. 2. Developing strategies tomake fertilizers affordable, especially thosecontaining P that organic produce cannotsupply adequately. 3. Promoting widelysynergistic technologies such as biologicalsoil and water conservation measures.4. Promoting the keeping of livestock toproduce manure, and developing bestmanagement practices for its use. 5. Developingstrategies for wide-scale disseminationof the options available, particularlythose that deal with overcoming the prevailingconstraints of germplasm supplyand information on their use. 6. Assessingecological benefits of fallow plant specieswhile mitigating potential problems ofthem becoming invasive weeds. 7. Determiningways in which high-value trees,crops and livestock can be more intensivelyfarmed, providing a natural progressionout of poverty.ReferencesBarrios, E., F. Kwesiga, R.J. Buresh, and J.I.Sprent 1997. Light fraction soil organicmatter and available nitrogen followingtrees and maize. Soil Science Society ofAmerica Journal 61: 826−831.Barrios, E., F. Kwesiga, R.J. Buresh, J.I. Sprentand R. Coe 1998. Relating preseason soilnitrogen to maize yield in tree legumemaizerotations. Soil Science Society ofAmerica Journal 62: 1604–1609.Buresh, R.J., P.C. Smithson and D.T. Hellums1997. Building soil capital in Africa. Pp.111–149 in: Replenishing Soil Fertility inAfrica. American Society of Agronomyand Soil Science Society of America. SSSASpecial Publication No. 51.Buresh, R.J. and G. Tian 1997. Soil improvementby trees in sub-Saharan Africa. AgroforestrySystems 38: 51–76.

Chapter 6: Agroforestry innovations for soil fertility management in sub-Saharan Africa59Desaeger, J. and M.R. Rao 2000. Parasitic nematodepopulations in natural fallows andimproved cover crops and their effects onsubsequent crops in Kenya. Field CropsResearch 65: 41–56.Drechsel, P. and L.A. Gyiele 1999. The EconomicAssessment of Soil Nutrient Depletion:Analytical Issues for FrameworkDevelopment. International Board for SoilResearch and Management (IBSRAM),Bangkok, Thailand.Franzel, S. 1999. Socioeconomic factors affectingthe adoption of potential improved tree fallows.Agroforestry Systems 47: 305–321.Gachengo, C.K. 1996. Phosphorus release andavailability on addition of organic materialsto phosphorus fixing soils. MSc thesis,Forestry Department, Moi University, Eldoret,Kenya.Gacheru, E. and M.R. Rao 2001. Managingstriga infestation in maize using organicand inorganic nutrient sources in westernKenya. International Journal of Pest Management47: 233–239.Gathumbi, S. 2000. Nitrogen sourcing by fastgrowinglegumes in pure and mixed speciesfallows in western Kenya. PhD Thesis,Wye College, Department of BiologicalSciences, University of London, UK.Hartemink, A.E., R.J. Buresh, B. Jama and B.H.Janssen 1996. Soil nitrate and waterdynamics in sesbania fallows, weed fallowsand maize. Soil Science Society ofAmerica Journal 6: 568−574.ICRAF 1997. Annual report 1996. InternationalCentre for Research in Agroforestry, Nairobi,Kenya.Jama, B.A., R.J. Buresh, J.K. Ndufa and K.D.Shepherd 1998a. Vertical distribution ofroots and soil nitrate: Tree species andphosphorus effects. Soil Science Societyof America Journal 62: 280−286.Jama, B.A., R.J. Buresh and F.M. Place 1998b.Sesbania tree fallows on phosphorus-deficient sites: Maize yield and financialbenefit. Agronomy Journal 90: 717–726Jama, B.A., R.A. Swinkels and R.J. Buresh 1997.Agronomic and economic evaluation oforganic and inorganic sources of phosphorusin western Kenya. Agronomy Journal89: 597−604.Jama, B.A., C.A. Palm, R.J. Buresh, A.I. Niang, C.Gachengo, G. Nziguheba and B. Amadalo2000. Tithonia diversifolia as a green manurefor soil fertility improvement in westernKenya: A review. Agroforestry Systems49: 201–221.Kihanda, F.M. and M. Gichuru 1999. ManureManagement for Soil Fertility Improvement.Tropical Soil Biology and FertilityInstitute, Nairobi, Kenya.Kuntashula, E., P.L. Mafongoya, G. Sileshi and S.Lungu 2004. Potential of biomass transferin sustaining vegetable production in thedambos of eastern Zambia. ExperimentalAgriculture 40: 37–51.Kwesiga, F. and R. Coe 1994. The effect of shortrotation Sesbania sesban fallows on maizeyield. Forest Ecology and Management64: 199–208.Kwesiga, F., F.K. Akinnifesi, P.L. Mafongoya,M.H. McDermott and A. Agumya 2003.Africa research and development in thesouthern Africa region during the 1990s:Review and challenges ahead. AgroforestrySystems 59: 173–186.Mekonnen, K., R.J. Buresh and B.A. Jama 1997.Root and inorganic nitrogen distributionsin sesbania fallow, natural fallowand maize fields. Plant and Soil 188:319–327.Niang, A., S. Gathumbi and B. Amadalo 1996a.The potential of short-duration improvedfallow for crop productivity enhancementin the highlands of western Kenya. Pp.218–230 in: J.O. Mugah (ed) People andInstitutional Participation in Agroforestryfor Sustainable Development. Proceedingsof the First Kenya Agroforestry Conference.Kenya Forestry Research Institute,Nairobi, Kenya.Niang, A., S. Gathumbi, B. Amadalo and C.Obonyo 1996b. Maize yield response togreen manure application from selectedshrubs and tree species in western Kenya:A preliminary assessment. Pp. 350–358 in:J.O. Mugah (ed) People and InstitutionalParticipation in Agroforestry for SustainableDevelopment. Proceedings of the First KenyaAgroforestry Conference. Kenya ForestryResearch Institute, Nairobi, Kenya.Nye P.H. and D.J. Greenland 1960. The soil undershifting cultivation. Technical CommunicationNo. 51. Commonwealth Bureauof Soils, Harpenden, UK.Nziguheba, G., C.A. Palm, R.J. Buresh and P.C.Smithson 1998. Soil phosphorus fractionsand adsorption as affected by organic andinorganic sources. Plant and Soil 198:159–168.Palm, C.A. 1995. Contribution of agroforestrytrees to nutrient requirements of plants.Agroforestry Systems 30: 105–124.Place, F. S. Franzel, J. de Wolf, R. Rommelse,F.R. Kwesiga, A.I. Niang and B.A. Jama2000. Agroforestry for soil fertilityreplenishment: Evidence on adoptionprocesses in Kenya and Zambia. Paperpresented at workshop: UnderstandingAdoption Processes for Natural ResourceManagement Practices in sub-SaharanAfrica, 3–5 July 2000, at the World AgroforestryCentre, Nairobi, Kenya.Rao, M.R., A. Niang, A. Kwesiga, B. Duguma,S. Franzel, B. Jama and R.J. Buresh 1998.Soil fertility replenishment in sub-SaharanAfrica: New techniques and the spreadof their use on farms. Agroforestry Today10(2): 3−8.Sanchez, P.A. 1995. Science in agroforestry.Agroforestry Systems 30: 5–55.

60World Agroforestry into the FutureSanchez, P.A., K.D. Shepherd, M.J. Soule, F.M.Place, R.J. Buresh, A.-M. Izac, A.U. Mokwunye,F.R. Kwesiga, C.G. Ndiritu andP.L. Woomer 1997. Soil fertility replenishmentin Africa: An investment in naturalresource capital. Pp. 1–46 in: Buresh, R.J.,P.A. Sanchez and F. Calhoun (eds) ReplenishingSoil Fertility in Africa. SSSA SpecialPublication No. 51, Soil Science Societyof America, Madison, WI, USA.Sanchez, P.A. and B.A. Jama 2002. Soil fertilityreplenishment takes off in East and SouthernAfrica. Pp. 23–45 in: Vanlauwe, B., J.Diels, N. Sanginga and R. Merckx (eds)Integrated Plant Nutrient Management inSub-Saharan Africa. CAB International,Wallingford, UK.Scoones, I. and C. Toulmin 1999. Soil nutrientbudgets and balances: What use forpolicy? Managing African Soils No. 6.IIED Drylands Programme, InternationalInstitute for Environment and Development,Edinburgh, UK.Shepherd K.D. and M. J. Soule 1998. Soil fertilitymanagement in west Kenya: dynamicsimulation of productivity, profitabilityand sustainability at different resourceendowment levels. Agriculture, Ecosystemsand Environment 71: 131–145.Smaling, E.M.A. 1993. An agroecological frameworkfor integrated nutrient managementwith special reference to Kenya. PhD thesis,Agricultural University, Wageningen,the Netherlands.Stroorvogel, J.J., E.M.A. Smaling and B.H.Janssen 1993. Calculating soil nutrientbalances in Africa. I. Supranational scale.Fertilizer Research 35: 227–235.Swinkels, R.A., S. Franzel, K.D. Shepherd, E.Ohlsson and J.K. Ndufa 1997. The economicsof short rotation improved fallows:Evidence from areas of high populationdensity in western Kenya. AgriculturalSystems 55: 99–121.van Straaten, P. 2000. Rocks for crops: Agromineralsfor sub-Saharan Africa. WorldAgroforestry Centre (ICRAF), Nairobi,Kenya.

Keywords:Extension, farmer-centred research, fodder shrubs,landcare, natural vegetative strips, tree fallowsChapter 7Scaling up the impact of agroforestry:Lessons from three sites in Africa and Asia 1S. Franzel, G.L. Denning, J-P. Lillesø-Barnekow, World Agroforestry Centre, Kenya and A.R. Mercado Jr, WorldAgroforestry Centre, the PhilippinesAbstractThis chapter assesses recent lessons learned from attempts to scale up agroforestry improvements,drawing on three case studies: fodder shrubs in Kenya, improved tree fallows in Zambia and naturalvegetative strips coupled with the Landcare movement in the Philippines. Currently, more than 15 000farmers use each of these innovations. Based on an examination of the main factors facilitating theirspread, 10 key elements of scaling up are presented. These include: taking a farmer-centred researchand extension approach; providing a range of technical options; building local institutional capacity;sharing knowledge and information; learning from successes and failures; and strategic partnershipsand facilitation. Three other elements are important for scaling up: marketing, germplasm productionand distribution systems, and policy options, although the three case study projects had only a marginalreliance on these. As different as the strategies for scaling up are, they face similar challenges. Facilitatorsneed to develop exit strategies, find ways to maintain bottom-up approaches as innovations spread,assess whether and how successful strategies can be adapted to different sites and countries, examineunder which circumstances they should scale up innovations and under which circumstances theyshould scale up processes, and determine how the costs of scaling up may be reduced.IntroductionDuring the past two decades, researchers have workedwith farmers throughout the tropics to identify anddevelop improved agroforestry practices that buildon local indigenous knowledge and offer substantialbenefits to households and the environment (Cooper etal. 1996; Franzel and Scherr 2002; Place et al. 2002;Sanchez 1995). Research and development projectshave demonstrated in many instances that agroforestryincreases household incomes, generates environmentalbenefits, and is particularly well suited to poor andfemale farmers. But in most cases these success storieshave been confined to localized sites, often with unusuallyconcentrated institutional support from researchand development organizations.As a consequence, considerable attention has beendevoted in recent years to ‘scaling up’ the benefits ofresearch, that is, ‘bringing more quality benefits tomore people over a wider geographical area, morequickly, more equitably, and more lastingly’ (IIRR2000). The issue of scaling up is particularly important1This chapter is a shortened version of a longer published paper: Franzel, S., G.L. Denning, J.P.B. Lillesø and A.R. Mercado 2004.Scaling up the impact of agroforestry: Lessons from three sites in Africa and Asia. Agroforestry Systems 61–62(1–3): 329–344.

62World Agroforestry into the Futureto agroforestry and natural resources managementinnovations, because they arerelatively ‘knowledge intensive’ and, unlikeGreen Revolution technologies, may notspread easily on their own. Drawing on arange of expertise, Cooper and Denning(2000) identified 10 essential elementsfor scaling up agroforestry innovations:farmer-centred research and extension approaches,technology options, buildinglocal capacity, germplasm, market options,policy options, knowledge and informationsharing, learning from successes and failures,strategic partnerships, and facilitation(Figure 1).The objective of this chapter is to assessrecent lessons learned in scaling upagroforestry benefits, drawing on threeStrategicpartnershipsLearning fromsuccess andfailurescase studies in Kenya, Zambia, and thePhilippines. Two of these, from Kenya andthe Philippines, were reported in Franzelet al. (2001a), but this chapter will showimportant developments since then. Firstly,concepts and definitions of scaling up arereviewed. Secondly, the case studies arepresented, followed by a discussion oftheir use of the 10 fundamental elements.Finally, conclusions are drawn andresearch challenges are discussed.Scaling up: Definitions andconceptsThere is a proliferation of terms to describescaling up (Gündel et al. 2001; Uvinand Miller 1996). For instance, Uvin andMiller’s typology involves 17 differentAvailablegermplasmskinds of scaling up, focusing variously onstructure, when a programme expandsits size; strategy or degree of politicalinvolvement; and resource base, referringto organizational strength.In this chapter we follow Gündel et al.(2001), who adopt the IIRR (2000) definitionof scaling up, which notes that the‘scaled-up state’ can either occur spontaneouslyor because of the deliberate, plannedefforts of governments, non-governmentalorganizations (NGOs) or other changeagents. Much can be learned from studyinghow spontaneous dissemination ofinnovations takes place, and in particularthe role of farmer-to-farmer dissemination.Scaling up is a communication process,and change agents have to understand howfarmers receive, analyse, and disseminateinformation in order to facilitate it. There isemerging literature on agricultural knowledgeand information systems, exploringhow those involved in the creation of agriculturalknowledge acquire, transmit andexchange information (Garforth 2001).MarketoptionsTechnologyoptionsCase studies from Kenya,Zambia, and the PhilippinesPolicyoptionsFacilitationFigure 1. Essential elements for scaling up agroforestry innovations.Source: Cooper and Denning (2000).SCALING UPAGROFORESTRYINNOVATIONSLocalinstitutionalcapacityKnowledge/informationsharingFarmer-centredresearch andextensionapproaches1. Fodder shrubs, KenyaThe low quality and quantity of feedresources is a major constraint to dairyfarming in central Kenya. Most farmersalso grow Napier grass as fodder, but it isinsufficient in protein and the daily yieldof cows fed on it is only around 8 litres.Commercial dairy meal is available butfarmers consider it expensive and mostdo not use it (Franzel et al. 2003).Development of the innovationIn the early 1990s, researchers (from theKenya Agricultural Research Institute –KARI, the Kenya Forestry Research Institute

Chapter 7: Scaling up the impact of agroforestry63– KEFRI, and the World Agroforestry Centre– ICRAF) and farmers around Embu, Kenya,tested several fodder shrubs. Most of thetrials were farmer-designed and managed.Calliandra calothyrsus emerged as the bestperforming and most preferred by farmers.It was found to grow in a range of‘neglected niches’ on their farms, includingin hedges along internal and externalboundaries, around the homestead, alongthe contour for controlling soil erosion,or intercropped with Napier grass. Whenpruned to a height of 1 m, the shrubs didnot compete with adjacent crops. Growing500 shrubs increased farmers’ incomes byaround US$98–124 per year. By the late1990s, two other shrub species, Morus alba(mulberry) and Leucaena trichandra, wereintroduced to farmers following successfulon-farm testing (Franzel et al. 2003).Scaling upBy 1999, 8 years after the introduction offodder shrubs, about 1000 farmers aroundthe research sites had planted them. However,there was limited scope for reachingall the 625 000 dairy farmers in Kenya;seed was scarce, and the farmers, extensionstaff and NGOs away from the onfarmtrials were not aware of the work.During 1999–2001, KARI, ICRAF and theInternational Livestock Research Institute(ILRI) collaborated in a project to scale upthe use of fodder shrubs in central Kenya.An extension facilitator, working with arange of government and NGO partners,assisted 180 farmer-development groups(comprising 3 200 farmers across sevendistricts) to establish nurseries and plantfodder shrubs. This approach proved to bevery effective: by 2002, each farmer hadan average of 340 shrubs and had giveninformation and planting material to an averageof six other farmers. Sixty percent ofparticipating farmers were women.Beginning in 2002, a project financed bythe Forestry Research Programme of DFID(the UK’s Department for InternationalDevelopment) and implemented by theOxford Forestry Institute and ICRAF helpeda range of partner organizations to increasethe adoption of fodder shrubs. By early2003, there were about 22 000 farmersplanting fodder shrubs in Kenya and severalthousand in four other countries. Facilitatorsare helping to train the extensionstaff of a range of different organizations,including government, NGOs, churches,community-based organizations, farmergroups and private-sector firms. The projectis also helping to facilitate the emergenceof private seed producers and dealers, andto help link them to buyers in areas whereseed demand is high (Franzel et al. 2003;Wambugu et al. 2001).2. Improved tree fallows, ZambiaThe plateau area of eastern Zambia ischaracterized by a flat to gently rollinglandscape, with annual rainfall around1000 mm. Approximately half the farmerspractice ox cultivation, the others cultivateby hand hoe. Maize is the most importantcrop, and sunflower, groundnuts, cottonand tobacco are also grown.Surveys identified soil fertility as the farmers’main problem; fertilizer use had beencommon during the 1980s but was indecline as farmers now lacked the cash topurchase it (Franzel et al. 2002; Howard etal. 1997; Kwesiga et al. 1999).Development of the innovationIn 1987, a Zambia–ICRAF agroforestryresearch project began research on improvedfallows, using Sesbania sesban.By 1995, several hundred farmers wereinvolved. In researcher-led trials, farmerschose among three different species andtwo different management options – intercroppingwith maize versus growing thetrees in pure stands. In farmer-led trials,farmers planted and managed the improvedfallows as they wished. Most farmersopted for a 2-year fallow and plantedtheir main food crop, maize, for twoseasons following the fallow. Tephrosiavogelii, Cajanus cajan and Gliricidia sepiumwere the main fallow species used.Maize yields following improved fallowsaveraged 3.6 t ha –1 , almost as high as forcontinuously cropped maize with fertilizer(4.4 t ha –1 ) and much higher than maizeplanted without fertilizer (1.0 t ha –1 ).Scaling upExtension activities began in earnest in1996 when an extension specialist in theZambia–ICRAF project set up demonstrations,facilitated farmer-to-farmer visits,and trained staff from the Ministry of Agriculture,several NGOs and developmentprojects in Eastern Province. The projecthelped launch an adaptive research anddissemination network, consisting of representativesfrom several organizations,farmers’ associations and projects (Katangaet al. 2002). The extension effort receiveda big boost with the start of a United StatesAgency for International Development(USAID)-financed agroforestry project in1999, covering five districts. The Centrealso facilitated the visits of farmers fromMalawi, thus helping to launch the practicethere (Böhringer et al. 1998). Scaling upobjectives included sensitization, buildinggrassroots capacities, developing effectivepartnerships, promoting policies more conduciveto adoption, monitoring and evaluation,and conducting research on the scalingup process (Böhringer et al. 2003). By2001, more than 20 000 farmers in easternZambia had planted improved fallows(Kwesiga et al. 2003).

64World Agroforestry into the Future3. Natural vegetative strips (NVSs)and Landcare, the PhilippinesThe upland municipality of Claveria islocated in northern Mindañao, the Philippines.Annual rainfall of 2 200 mm allowsa farming system of two maize crops peryear. However, with this high rainfall,coupled with cultivation of sloping fieldsand use of animal tillage, soil loss througherosion had degraded lands and led todeclining maize yields.Applied research began in 1985 oncontour hedgerow systems using nitrogenfixingtrees to minimize erosion, restoresoil fertility and improve crop productivity.But adoption of this system was slow, andmany hedgerows were abandoned owingto the high labour requirement to maintainthem, poor adaptation of leguminous treesto acid soils, and competition between thetrees and the maize crop.Development of the innovationThrough participatory on-farm experiments,ICRAF researchers concluded that theconcept of contour hedgerows remainedpopular and that farmers were concernedabout soil erosion and loss of productivity.Researchers observed that farmers oftenploughed along contour lines, leavingcrop residues and/or natural vegetation instrips between ploughed fields. The latterinnovation evolved into natural vegetativestrips (NVSs) and emerged as a crucial entrypoint for reversing land degradation onsloping fields.Over several years, the NVS technology,coupled with contour ploughing, spreadspontaneously among farmers. This innovativefarmer-based system and itscomponents were the subjects of intensiveon-farm research. Farmer innovations suchas the ‘cow’s back method’ (using the viewof the ox’s backbone when ploughing tomaintain a reasonable trajectory for layingout contour lines) were identified as acceptablealternatives to the more technical‘A-frame’ technique (ICRAF 1997). For thestrips, some farmers demonstrated interestin such cash crops as fruit, timber andcoffee; others preferred improved foddergrasses and legumes. In all cases, theseinnovations built on and enriched thefoundation of the NVSs.Scaling upWith the spontaneous visible spread ofNVSs in and around ICRAF’s applied researchsites, considerable interest emergedfrom communities, local and provincialgovernment agencies, and NGOs to learnmore about this innovation. In 1996, theCentre responded to communities’ requestsfor technical support and training by introducingand testing the appropriateness ofLandcare, a participatory, community-basedapproach from Australia involving the developmentof groups in partnership withlocal government to promote conservationfarmingpractices (Campbell and Siepen1994; Catacutan et al. 2001; Mercado et al.2001). Farmers’ interest led to the formationof the Claveria Landcare Association, whichhas emerged as the platform for widespreaddissemination of conservation farmingbased on NVSs. In 1999, Landcare wasextended to another ICRAF research sitein nearby Lantapan municipality, and by2002 there were an estimated 500 Landcaregroups, involving more than 15 000farmers in the Philippines.Comparing the key elementsof scaling upFarmer-centred research andextensionParticipatory research, in which farmersplay a critical role in the design, implementation,and evaluation of research, hasbeen shown to improve the effectiveness ofresearch and to reduce the time betweeninitial testing and uptake (CGIAR/PRGA1999).Farmers were involved in the early stagesof the development of technologies at allthree sites. In both Kenya and Zambia,researcher-led and farmer-led trials wereconducted simultaneously: the formerprimarily to assess biophysical response,the latter for socioeconomic assessment(Franzel et al. 2001b). Encouraging farmersto experiment with the new practicesas they wished led to new innovations andgreatly improved the practices at both sites– reducing costs, promoting adoption andmaking scaling up more rapid.In the Philippines, it was a farmer innovation– leaving crop residues along the contour,where they revegetate forming NVSs– that proved very popular. Researcherslater proved that these strips were effectivein controlling soil erosion and required littlemaintenance. The use of NVSs spreadrapidly and farmers continued to innovate(Mercado et al. 2001). Also, establishing along-term field presence in Claveria enabledresearchers to identify and validatefarmers’ innovations, such as the cow’sback method, and to help farmers adjustthe NVS system to better reflect their interests,in particular by introducing such cashgeneratingenterprises as timber and fruit.There was some variation in extensionstrategies among the three case studies.In Kenya, extension facilitators providedtraining to government extension and NGOstaff and representatives of village-basedfarmer development groups, resulting ina significant amount of farmer-to-farmerextension. A similar strategy was implementedin eastern Zambia, except that

Chapter 7: Scaling up the impact of agroforestry65facilitators established a network of farmertrainers.In the Philippines, partnershipwith farmers in on-farm research paved theway for active farmer participation in thescaling up of both the technical innovationsand the Landcare approach.Technical optionsOffering a range of options to farmersrather than a specific recommendation isimportant for several reasons (Franzel et al.2001c):• Diversification minimizes productionand market risks, and allows for differentpreferences.• Farmers’ resources vary and differentoptions often have different resourcerequirements.• Different options allow for a variableenvironment.In all three sites, researchers and farmersquickly developed a range of options forthe technologies in question. In Kenya,farmers have the choice of three foddershrubs and a herbaceous legume, whichcan be planted in a range of differentniches and arrangements on their farms.Moreover, they can feed the leaves to theiranimals fresh or dry, or store them.In Zambia, farmers choose from four differentspecies and a range of management optionsfor their improved fallows, dependingon their preferences and available labour.They can plant the crops in pure stands orintercropped.In the Philippines in the early 1990s,farmers and researchers began with a singleinnovation – the NVS. But by the endof the decade, farmers had introduced 31different perennials, on their own initiativeor with advice from facilitators. Thesedifferent options included fruits, timbertrees, fodder grasses and legumes. Manyplanted with the intention of earning cash(Mercado et al. 2001).Local institutional capacityEmpowering local communities to plantheir own development and mobilize resourcesis fundamental to any successfuldevelopment strategy (Binswanger 2000).The three case studies used different approachesto building local institutionalcapacity. In central Kenya and easternZambia, extension facilitators providedtraining to village-based groups on thetechnologies they were promoting, butthere were few direct efforts to otherwisebuild the capacities of these groups.In Eastern Province, Zambia, in the mid-1990s, ICRAF assisted partner organizationsto form an adaptive research anddissemination network to plan, implementand evaluate on-farm research, training anddissemination activities. The network facilitatesthe involvement of local groups inthe plans and activities of research and developmentorganizations, which enhancestheir capacity and feelings of ownership ofthe network and practices (Katanga et al.2002).In the Philippines, Landcare has gonefurther in building local institutional capacity.It has enabled communities to shareknowledge and experience, influence theagenda of researchers and local policymakers, and mobilize financial resources.Mercado et al. (2001) noted that the “greatestsuccess of Landcare” was the changein the attitudes of farmers, policy makers,local government and landowners withrespect to land use and environmentalmanagement.GermplasmThe lack of planting material is repeatedlyidentified as one of the most importantconstraints to the wider adoption of agroforestryinnovations (Simons 1997). NationalTree Seed Centres have been unable to deliverseed to large numbers of smallholdersand, as with crop seed, “the seed demand–supply relationship in a large proportionof Africa’s smallholder farming systemsappears to represent a situation of marketfailure” (DeVries and Toenniessen 2001).Successful production and distribution ofquality tree seed to resource-poor farmersdepends on a number of factors, someof which are biophysical, for example,identifying adapted provenances and seedsources or ensuring sufficient genetic variation,while other factors are economic,organizational and institutional, such asthe protection of and ownership of seedsources, and cost-efficient production anddistribution networks.The Kenyan calliandra case study shows atypical dilemma: farmers unfamiliar withthe new practice cannot be expected tobuy seed, yet provision of free seed discouragesthem from harvesting it and underminesthe emergence of private-sectormarketing systems. ICRAF and KARI aretrying to improve the situation in four ways:• Helping link dealers in western Kenyato buyers in other parts of Kenya.• Assisting fodder shrub growers andprivate nurseries in central Kenya toproduce high-quality seed and seedlingsand become seed dealers.• Working with an NGO, Farm Input ProvisionServices, to help private dealers topackage and sell seed through stockists.• Encouraging private firms to producefodder shrub seed or to buy seed fromseed dealers.The situation of improved fallows inZambia has many similarities. One solutiontried here through a USAID-financed

66World Agroforestry into the Futureproject is to loan seed to farmers in returnfor a promise to give back double theamount they took. The sustainability of thissystem is uncertain; no private seed dealershave yet emerged, despite the wide-scaleadoption of improved fallows.In the expansion of Landcare, hundredsof communal and private individual treenurseries have been established to provideseedlings for fruit and timber species. InLantapan, farmers organized themselvesto create the Agroforestry Tree Seed Associationof Lantapan (ATSAL), a farmeroperatedseed collection, production,processing and marketing association. Theorganization has trained more than a thousandfarmers in both exotic and indigenoustree species and has extended its operationsto other areas of the country.MarketingLinking farmers to markets and adding valueto raw products have great potential forimproving the incomes of smallholders andfacilitating the scaling up process (Deweesand Scherr 1996). All three of the mainpractices promoted in the case studiesproduce inputs: fodder for increased milkproduction, and soil erosion control andsoil fertility for crop production. However,only one of them, fodder, can be sold, explainingthe relatively low emphasis givento marketing and product transformation inthe case studies. Nevertheless, the uptakeof the new practices depends on the availabilityof markets for the final products.As mentioned above, efforts are needed inall three cases to promote the marketing ofseed and seedlings. Moreover, there are alsooptions for increasing the marketing of fodderfrom shrubs, which could be promotedas a cash crop for farmers who do not ownlivestock. In Kenya, there is also great potentialfor selling leaf meal as a protein sourceto millers producing dairy concentrates,who currently import protein in the form offish meal, soybeans and cottonseed cake.For thousands of low-income farmers in thePhilippines, the NVS system has evolvedas a means to graduate from subsistencemaize farming to cash cropping. Claveriais well connected by road to the large portcity of Cagayan de Oro, opening up potentialmarkets for a range of agroforestryproducts. NVS adopters in Claveria arenow observed to be growing a wide rangeof timber and fruit trees and are increasinglyexpressing interest in backyard livestockenterprises to diversify and stabilize theirincomes. Market access has been criticalfor the intensification and diversification ofthe NVS system.Policy optionsPolicy affects scaling up operations in severaldifferent ways: policy constraints maylimit adoption of new practices, policyincentives help promote adoption, andpolicy makers themselves may be engagedto promote or even finance scaling upactivities – a relatively untapped resource(Raussen et al. 2001).In Zambia, local leaders played importantroles in promoting improved fallows intwo ways. Firstly they helped sensitize andmobilize their constituents to plant improvedfallows. Secondly, they passed, andin some cases, promoted the enforcementof bylaws to remove two of the main constraintsto agroforestry adoption: the settingof uncontrolled fires and free grazing oflivestock (Ajayi et al. 2002).The Landcare movement has benefitedfrom and, in turn, reinforced the Philippinegovernment policy of decentralizationand devolution of responsibilities to localgovernment. The local government units(LGUs) are now seen as important partnersin local natural resource managementinitiatives, providing policy support forinstitutionalizing Landcare and conservationfarming practices, training staff, andfinancing Landcare activities (Catacutanand Duque 2002; Catacutan et al. 2001).Knowledge, information sharing andlearning from successes and failuresThe dissemination of knowledge andinformation about scaling up among stakeholdersis necessary for making effectivedecisions. Monitoring and evaluation systems,both formal and informal, ensure thegeneration of such information at a rangeof different scales and from the perspectivesof different stakeholders (Cooper andDenning 2000).In Kenya and Zambia, monitoring andevaluation have been conducted in severaldifferent ways. Village workshops enabledresearchers to gain an up-front understandingof farmers’ assessments and expectationsof the technologies they are using. Inboth Zambia and Kenya, Centre staff andpartners engage in collaborative monitoringand evaluation. These studies includeeconomic analyses, impact assessmentsand assessing factors affecting adoption.The system in both countries is not withoutproblems, not all organizations involved inscaling up participate and some are unableto collect even the minimum data required.But the collaborative mechanism givespartners a greater sense of ownership andbuy-in as well as access to more informationand feedback (Nanok 2003).Knowledge sharing and learning are prioritiesat all three sites. As highlightedearlier, Landcare groups have proved to bean effective vehicle for knowledge sharingin areas of conservation farming andlivelihood improvement. This institutional

Chapter 7: Scaling up the impact of agroforestry67platform for knowledge sharing is especiallyvaluable in heterogeneous biophysicalenvironments and dynamic local economieswhere farming systems are constantlyevolving.Strategic partnerships andfacilitationWhen scaling up, partnerships offer highpotential benefits. Organizations with complementarystrengths, resources and ‘reach’can improve the efficiency and effectivenessof scaling up efforts (Cooper andDenning 2000). However, these need tobe weighted against such potential risks ashigh costs in terms of time and resources,compromised impact, and loss of identity(Jacquet de Haveskercke et al. 2003).Landcare in the Philippines is based on astrategic partnership of farmers, LGUs andtechnical facilitators (such as ICRAF andgovernment line agencies). These alliancesare an important aspect of the Landcaremovement as they strengthen the influenceof farming communities on policy formulationand resource allocation decisions.One concern is that ICRAF continues toplay a seemingly indispensable role in facilitation;more effort is needed to developlocal skills, preferably in the community orwith NGOs that have a long-term commitmentto the local communities.Conclusions: ResearchchallengesThe review of the three case studieshighlights the fact that there is no singlerecipe for scaling up (Table 1). Differentapproaches can all be successful, dependingon the innovation, the environment andthe resources at hand. The key elementscontributing to improved impact in the casestudies were: taking a farmer-centred researchand extension approach; providing arange of technical options; building local institutionalcapacity; sharing knowledge andinformation; learning from successes andfailures; and creating strategic partnershipsand facilitation. Three other elements areimportant for scaling up: marketing, systemsof germplasm production and distribution,and policy options, but the performance ofthe cases on these was, at best, mixed.Although the facilitators in the three casestudies used many of the same elements,their scaling up strategies were very different.The case of fodder shrubs in Kenyaoffers the simplest approach, in which afocus on on-farm research and facilitatingextension services, NGOs and farmergroups reaped surprisingly high benefits.Certainly the limited scope of the innovation– it being one among several importantmanagement practices in the dairyenterprise – prevents certain scaling upapproaches from being used.Table 1.Agroforestry practices, extension strategy, policy options and institutional innovations promoted in the three scaling upcase studies.TechnologyExtensionstrategy Policy options Institutional innovationsTypeOriginFacilitatingextensionservices,NGOs,farmergroupsEngaginglocalgovernmentinfacilitativeroleObtaininglocalgovernmentfinancingLobbyingfor localpolicychangesLobbying fornational policychangesCapacitybuildingof farmergroupsFacilitatingthecreation offederationsof farmergroupsPromotingconsortiaof partnersFodder shrubs,KenyaResearcherandfarmer-ledtrialsXXImprovedfallows,ZambiaNaturalVegetative Strips/Landcare,the PhilippinesResearcherandfarmer-ledtrialsFarmerinnovationwith researchinputsX X X X XX X X X X X X X

68World Agroforestry into the FutureImproved fallows in Zambia is an intermediatecase, in which a more complex managementpractice relevant to several enterprisesis being scaled up. Facilitators are using severalstrategies in addition to those used byfodder shrub facilitators, including engaginglocal government in a facilitative role, lobbyingfor policy changes, and promotinga network of partners. These have greatlyadded to the success of the innovation andto its spread across eastern Zambia.The case of NVS/Landcare in the Philippinespresents the most extensive set ofinnovations, a technical one accompaniedby an institutional one. The technical innovationis simple, yet serves as a platformfor a multiplicity of other technical innovationsand, indeed, a transformation of thefarming system. The institutional innovation,Landcare, has had far-reaching ramifications,as federations of farmer groups canwield not only increased economic powerbut political power as well. In addition tothe strategies used by those promoting foddershrubs and improved fallows, facilitatorsin the Philippines have obtained localgovernment financing and have facilitatedthe establishment of federations of groups.Moreover, they have persuaded policymakers to incorporate the Landcare approachinto local and national policy.But as different as the case studies are, theyface five similar challenges:• Articulation of a clear exit strategy, toleave farmers on their own to continueto implement and disseminate the innovations,with limited local backstopping.• Maintaining the bottom-up, participatorynature of the scaling up process,which contrasts with the top-downapproaches of many government servicesand NGOs.• Adapting the scaling up innovations andprocesses from one site or country foruse at another site or country.• Deciding under what circumstancesfacilitators should seek to scale up technologies,and under what circumstancesto scale up the process by which adoptionand adaptation have taken place.In other words, is a scaling up strategyapplicable only for a particular technology,or can it be used for several innovations,for any type of agricultural innovationor for agriculture in general?• Making sure that the benefits of scaling upout-weigh the costs. This includes promotingor formalizing farmer-to-farmer informationsystems, and encouraging farmerorganizations such as Landcare to take onsome of the functions of these systemsAll of the above issues are at least to someextent researchable. For example, carefulassessments of the costs and benefits, andadvantages and disadvantages of differentstrategies can be made. Simple plannedcomparisons of different scaling up mechanismscan be undertaken. Just as learningand knowledge sharing are critical functionsin the scaling up of innovations, theyare critical for identifying effective andefficient scaling up strategies. Investmentin understanding scaling up processeswill reap important rewards leading to improvedlivelihoods of beneficiaries.AcknowledgementsThe comments and suggestions of ChrisGarforth and Robert Tripp on earlier draftsare gratefully acknowledged.ReferencesAjayi, O.C., R. Katanga, E. Kuntashula and E.T.Ayuk 2002. Effectiveness of local policiesin enhancing adoption of agroforestry technologies:The case of by-laws on grazingand fire in eastern Zambia. Pp. 93–99 in:Kwesiga, F., E. Ayuk, and A. Agumya (eds)Proceedings of the 14 th Southern AfricaRegional Review and Planning Workshop,3–7 September 2001, Harare. SouthernAfrica Development Community–WorldAgroforestry Centre (SADC–ICRAF) RegionalProgramme, Harare, Zimbabwe.Binswanger, H.P. 2000. Scaling up HIV/AIDSprograms to national coverage. Science288: 2173–2176.Böhringer, A., E.T. Ayuk, R. Katanga and S. Ruvuga2003. Farmers’ nurseries as a catalystfor developing sustainable land usesystems in southern Africa. Part A: Nurseryproductivity and organization. AgriculturalSystems 77: 187–201.Böhringer, A., N. Moyo and R. Katanga 1998.Farmers as impact accelerators: Sharingagroforestry knowledge across borders.Agroforestry Today April–June 1998,World Agroforestry Centre (ICRAF),Nairobi, Kenya.Campbell, A. and G. Siepen 1994. Landcare:Communities Shaping the Land and the Future.Allen and Unwin, Sydney, Australia.Catacutan, D.C. and C.E. Duque 2002. LocallyledNatural Resource Management. Proceedingsof a regional workshop, 8–9November 2001, Valencia City, Bukidnon,the Philippines.Catacutan, D.C., A.R. Mercado and M. Patindol2001. Scaling up the Landcare andNRM planning process in Mindanão, thePhilippines. LEISA (Low External Inputand Sustainable Agriculture) Magazine17(3): 31–34.

Chapter 7: Scaling up the impact of agroforestry69CGIAR/PRGA (Consultative Group on InternationalAgricultural Research/ParticipatoryResearch and Gender Analysis) 1999.Crossing perspectives: Farmers and scientistsin participatory plant breeding. Cali,Colombia.Cooper, P.J.M. and G.L. Denning 2000. ScalingUp the Impact of Agroforestry Research,World Agroforestry Centre (ICRAF),Nairobi, Kenya.Cooper, P.J., R.R.B. Leakey, M.R. Rao and L.Reynolds 1996. Agroforestry and the mitigationof land degradation in the humidand sub-humid tropics of Africa. ExperimentalAgriculture 32: 235–290.DeVries, J. and G. Toenniessen 2001. Securingthe harvest: Biotechnology, breeding, andseed systems for Africa’s crops. CAB International,Wallingford, UK.Dewees, P.A. and S.J. Scherr 1996. Policies andmarkets for non-timber tree products.Environment and Production TechnologyDivision Discussion Paper No. 16. InternationalFood Policy Research Institute,Washington, DC, USA.Franzel S. and S.J. Scherr (eds) 2002. Trees onthe Farm: Assessing the Adoption Potentialof Agroforestry Practices in Africa. CABInternational, Wallingford, UK.Franzel, S., P. Cooper and G.L. Denning (eds)2001c. Scaling up the benefits of agroforestryresearch: Lessons learned and researchchallenges. Development in Practice(Special volume) 11(4): 524–534.Franzel S., C. Wambugu and P. Tuwei 2003.The adoption and dissemination of foddershrubs in central Kenya. Agricultural Researchand Network (AGREN) Series PaperNo. 131. Overseas Development Institute,London, UK.Franzel S., D. Phiri and F.R. Kwesiga 2002.Assessing the adoption potential of improvedtree fallows in eastern Zambia. Pp.37–64 in: Franzel, S. and S.J. Scherr (eds)Trees on the Farm: Assessing the AdoptionPotential of Agroforestry Practices in Africa.CAB International, Wallingford, UK.Franzel, S., P. Cooper, G.L. Denning and D.Eade (eds) 2001a. Development and Agroforestry:Scaling up the Impacts of Research.Development in Practice (Specialvolume) 11(4): 405–534.Franzel, S., R. Coe, P. Cooper, F. Place and S.J.Scherr 2001b. Assessing the adoptionpotential of agroforestry practices in sub-Saharan Africa. Agricultural Systems 69:37–62.Garforth, C. 2001. Agricultural knowledge andinformation systems in Eritrea: A study insub zoba Hagaz. The University of Reading,UK.Garrity, D. 2000. The farmer-driven Landcaremovement: An institutional innovationwith implications for extension and research.Pp. 7–9 in: Cooper, P.J.M. andG.L. Denning (eds) Scaling Up the Impactof Agroforestry Research. World AgroforestryCentre (ICRAF), Nairobi, Kenya.Gündel S., J. Hancock and S. Anderson 2001.Scaling up strategies for research in naturalresources management: A comparativereview. Natural Resources Institute,Chatham, UK.Howard, J.A., L. Rubey and E.W. Crawford1997. Getting technology and the technologyenvironment right: Lessons frommaize development in southern Africa.Paper presented at the Meeting of theInternational Association of AgriculturalEconomics. 10–16 August 1997, Sacramento,USA.ICRAF (World Agroforestry Centre) 1997.Annual Report 1996. ICRAF, Nairobi,Kenya.IIRR (International Institute of Rural Reconstruction)2000. Going to Scale: Can We BringMore Benefits to More People More Quickly?IIRR, Silang, Cavite, the Philippines.Jacquet de Haveskercke, C., A. Böhringer, R.Katanga, L. Matarirano, S. Ruvuga andD. Russell 2003. Scaling out agroforestrythrough partnerships and networking: Theexperience of the World Agroforestry Centrein southern Africa. Southern Africa AgroforestryDevelopment Series No. 2. WorldAgroforestry Centre, Zomba, Malawi.Katanga, R., D. Phiri, A. Böhringer and P. Mafongoya2002. The adaptive research anddissemination network for agroforestry:A synthesis of the adaptive workshops.Pp 93–99 in: Kwesiga, F., E. Ayuk andA. Agumya (eds) Proceedings of the 14 thSouthern Africa Regional Review andPlanning Workshop, 3–7 September2001, Harare. Southern Africa DevelopmentCommunity–International Centre forResearch in Agroforestry (SADC–ICRAF)Regional Programme, Harare, Zimbabwe.Kwesiga, F.R., S. Franzel, F. Place, D. Phiri andC.P. Simwanza 1999. Sesbania sesbanimproved fallows in eastern Zambia: Theirinception, development, and farmer enthusiasm.Agroforestry Systems 47: 49–66.Kwesiga, F., S. Franzel, P. Mafongoya, O. Ajayi,D. Phiri, R. Katanga, E. Kuntashula, F.Place and T. Chirwa 2003 Improved fallowsin eastern Zambia: History, farmerpractice and impacts. Conference PaperNo. 12. Internationale Weiterbildung undEntwicklung/International Food PolicyResearch Institute/New Partnership forAfrica’s Development/Technical Centrefor Agricultural and Rural Cooperation(InWEnt/IFPRI/NePAD/CTA) Conference'Successes in African Agriculture', 1–3December 2003, Pretoria, South Africa.Mercado, A.R., P. Marcelino and D.P. Garrity2001. The Landcare experience in thePhilippines: Technical and institutionalinnovations for conservation farming. Developmentin Practice 11 (4): 495–508.Nanok, T. 2003 Assessment of the adoption anddissemination of improved fallows andbiomass transfer technologies in westernKenya. MSc. thesis. Department of Agri-

70World Agroforestry into the Futurecultural Economics. Egerton University,Njoro, Kenya.Place, F., S. Franzel, J. DeWolf, R. Rommelse,F. Kwesiga, A. Niang and B. Jama 2002a.Agroforestry for soil fertility replenishment:Evidence on adoption processesin Kenya and Zambia. Pp. 155–168 in:Barrett, C.B., F. Place and A.A. Aboud(eds) Natural Resources Management inAfrican Agriculture: Understanding andImproving Current Practices. CAB International,Wallingford, UK.Raussen, T., G. Ebong and J. Musiime 2001. Moreeffective natural resources managementthrough democratically elected, decentralisedgovernment structures in Uganda. Developmentin Practice 11(4): 460–470.Sanchez, P.A. 1995. Science in agroforestry.Agroforestry Systems 9: 259–274.Simons, A.J. 1997. Delivery of improvement foragroforestry trees. In: Dieters, M.J. and C.A.Matheson (eds) Tree Improvement for SustainableTropical Forestry. Queensland ForestryResearch Institute, Gympie, Australia.Uvin, P. and D. Miller. 1996. Paths to scaling-up:Alternative strategies for localnongovernmental organizations. HumanOrganization 55(3): 344–354.Wambugu, C., S. Franzel, P. Tuwei andG. Karanja 2001. Scaling up the useof fodder shrubs in central Kenya. Developmentin Practice 11(4): 487–494.

Keywords:Policy, land management, agriculture,agroforestry, researchChapter 8Policies for improved land management in smallholderagriculture: The role for research in agroforestry andnatural resource managementFrank Place, World Agroforestry Centre and Yves-Coffi Prudencio, World Bank 1AbstractThis main purpose of this chapter is to identify significant research areas for the World AgroforestryCentre (and other natural resource management research centres) that will contribute to improvedpolicies for land management. To do this, the chapter first describes the nature and extent of land degradationand argues why this is a major policy issue. It then applies a conceptual framework to identifythe proximate and underlying causes of land degradation. Special attention is paid to the effect of policies,especially those related to land tenure, soil management, and research and extension. It concludesby reviewing the Centre’s experiences in policy research and advocacy and proposing three key areasfor intervention in the future: identification of land management problems and opportunities, design ofinvestment/response priorities, and monitoring the impact of policy reforms.IntroductionIt is well documented that smallholder farming areasthroughout the developing world are beset by highpoverty rates, suffer from low (and sometimes declining)productivity, and have increased occurrences ofland degradation. For example, about 50 percent of ruralhouseholds in India, and about 60 percent of thosein sub-Saharan Africa (SSA) are living below the povertyline (World Bank 2000). In fact, in SSA, about 85percent of the all the poor reside in rural areas (Mwabuand Thorbecke 2001). Some South American and EastAsian countries have succeeded in significantly reducingpoverty through urban growth, but this is decadesaway from being feasible in many South Asian countriesand almost all African countries (United Nations2003). Thus, coinciding with the recent heightenedattention on poverty and the setting of the MillenniumDevelopment Goals (MDGs) (United Nations 2000),there has been renewed debate on how policy reformscan spur investment in smallholder agriculture to increasefood security and income. These debates payparticular attention to the impact of such policies onsustainable land management, which is increasinglybeing recognized as a necessary step along the pathtowards meeting the MDGs in rural areas.The major aim of this chapter is to suggest a set ofpriorities for research into improved land managementin smallholder agricultural landscapes, which willpotentially inform policy decisions. The section thatfollows will show why smallholder land management isan issue meriting public policy attention: examining the1The opinions expressed in this chapter are those of the author and not of the World Bank.

72World Agroforestry into the Futureextent of land management problems, theirproximate and root causes, and their consequencesfor society. The third section willthen examine how past and current policieshave contributed to such land managementproblems. The fourth section of the chapterdiscusses positive policy reforms that havetaken place along with those that requirefurther attention. Finally, the contributionof research to the process of policy reformis discussed, and priorities for natural resourcemanagement (NRM) and agroforestryresearch centres are proposed.Land degradation as a publicpolicy issueExtent of land degradation in theworldMany researchers, for example Oldemanet al. (1991), estimate that severe land degradationis already pervasive. Significantdamage has been observed in the Balkans,eastern Europe, much of Southeast Asia,the farming areas bordering the SaharaDesert, and the central United States, whilethere is substantial but patchy degradationin areas of South Asia, eastern andsouthern Africa, and the eastern Amazon.In their paper, Oldeman et al. estimatedthat around 25 percent of land in developingcountries is degraded, rising to 65percent when focusing solely on Africanagricultural land. However, the severity ofsuch large-scale land degradation is contentious.A recent study by Kaiser (2004)for example, supported by Wiebe (2003),has tempered some of the earlier alarmingnational and global rates of soil loss.What can be said is that land degradationis found throughout the world, but it isa more complex problem in developingcountries where it tends to reinforce and bereinforced by poverty.Why land degradation is a publicpolicy issueThe vast extent of degradation indicatesthat it is a major problem, but does not byitself imply that it is a public policy issue.However, there are complementary reasonsthat clearly make it a policy issue ofimportance at local, national, regional, andglobal scales. Firstly, there are significantsocial costs and benefits (externalities) associatedwith bad and good land management(for example, more dust or carbonsequestration respectively). Externalitiescan have local/regional effects, such asthe sedimentation of Lake Victoria fromdegraded upland agricultural communities,or global impacts, such as dust cloudsmoving from the desertified Sahel to NorthAmerica. Secondly, poverty reduction andfood security are the major goals of mostdeveloping-country governments, and thepoor are overwhelmingly rural. Helping thepoor, which is a public policy issue, thereforemust involve some attention to theland resource. This is important for overalleconomic growth as well as for equity considerations.Finally, future generations haverights to a viable soil resource and governments,as ‘custodians’ of the land (and,indeed, as owners of much of it), have anobligation to ensure that this happens.Reasons for land degradation insmallholder farming systemsThere are some causes of land degradationthat are either not possible to prevent orare too costly to prevent at farm or landscapescales. These include climatic catastrophessuch as major storms that bringforth massive rain and wind erosion, environmentallyinduced droughts, fires, majorpests or diseases that destroy vegetativecover, and human-induced or perpetuateddegradation such as massive health calamitiesor wars.But many other types of degradation canbe prevented or at least mitigated throughland management. Under similar physicaland climatic conditions, some communitiesare found to manage their land resourcesmuch better than their neighbours(see Pretty 1995 for a number of examples).Likewise, within the same villagelocation, some farms will have intact soilswhile others exhibit considerable types oferosion. So what explains the differences inland degradation or, more specifically, whyare large areas of lands not being investedin and managed sustainably?In their 1994 paper, Scherr and Hazell offeredsix main reasons why smallholderfarmers may not invest in NRM or adoptNRM technologies. These reasons are listedin a similar order to a decision-makingprocess, from first consideration to last:1. Lack of recognition of a natural resourceproblem2. Lack of importance of the natural resourceor its problem3. Lack of willingness to invest in the resource4. Lack of capacity to invest in the resource5. Lack of economic and other incentivesto invest in the resource6. Lack of information and support servicesthat are necessary to implement investments.We will now briefly discuss these reasonsspecifically in terms of land managementby smallholders (although these reasons arealso quite valid when considering collectivemanagement of landscapes).Lack of recognition of a naturalresource problem. Resource degradationprocesses, such as nutrient leaching ormore subtle sheet erosion of topsoil,are not always easy to detect. Even if

Chapter 8: Policies for improved land management in smallholder agriculture73productivity shows signs of weakening,a farmer may not attribute this to erodedresources. In other cases, a farmer mayrealize that the resource base is declining inquality, but it does not yet have a measurableimpact on productivity. So, for example,cumulative sheet erosion may be takingplace, but not be posing a ‘problem’ ondeeper soils.Lack of importance of the naturalresource or its problem. There are severalreasons why smallholders who can observea natural resource problem may feel that itis unimportant. Firstly, it may be occurringin a small or insignificant area, such as aremote corner of the farm. Secondly, eventhough the magnitude may be significant,a rural household may place more attentionon livelihoods that do not depend on thedegrading resource, for example, if thehousehold derives more of its income fromservices, processing, or wage employment,its priorities for investment will not be onfarming natural resources. Thirdly, lack ofappreciation of the natural resource orproblem may be due to a lack of awarenessor education, or due to historical or culturalfactors. For example, immigrants may notattach a high value to tree resources becausethey lack awareness of their potentialvalues.Lack of willingness to invest in theresource. Willingness does not referto capacity, but rather one’s interest ininvesting. This hinges significantly onsmallholders’ property rights and timehorizons. For example, if a farmer cannotgain long-term rights to land, there is littleincentive to make investments in that land,irrespective of the potential effectiveness ofthe investment. Furthermore, householdsthat rent land, or female farmers who donot control the benefits from investmentsthey make, are also unlikely to be willingto invest time and effort on improvements.In terms of time, some smallholders mayhave very short time horizons because ofhigh risk aversion or extreme poverty, and insuch cases will not even consider a numberof types of investments. In particular, thediscount rates of the poor are very high dueto immediate survival needs; they thereforesteeply discount benefits that may accruefrom long-term investment.Lack of capacity to invest in theresource. Some households mayrecognize a problem and have thepotential to benefit from investments tosolve it, but simply cannot assemble theresources required to make the investment.Setting aside land for permanent cover orlong fallows is not attractive to householdswith very small farms. Often, labourshortages prevent significant soil andwater conservation investments from beingundertaken, especially in areas ravagedby malaria or HIV/AIDS. But perhaps themost serious constraint is lack of financialcapital, since in theory households withcash can obtain labour or other resourcesneeded for investment.Lack of incentives to invest in theresource. In this situation, farmers mayhave access to the resources required tomake investments in natural resources,but the payoff from doing so appears tobe unattractive. This can be a result ofthe (in)effectiveness of the investment.For instance, few profitable investmentsmay be available in the more arid orsandy environments. But incentives arealso highly related to prices, access toinputs and markets for outputs. Whereoutput markets are lacking, farmers arediscouraged from purchasing inputs. Evenwhere markets exist, in many areas of ruralAfrica the ratio of farmgate input costs tooutput prices is so high as to discourage allbut the minimum of investment.Lack of information and supportservices to implement investments.While some types of investments arestraightforward, others may require adegree of technical knowledge. Withoutbasic training, farmers may not fullyunderstand how and why certain typesof investments will work. Likewise, sometypes of investment require materials (e.g.shovels or seedling pouches) that maynot normally be available locally. Moresignificant public investment to supportimproved land management is clearly vitalfor impoverished rural areas so that thistype of vicious circle of poverty and landdegradation can be broken.Many of these circumstances may be feltat the household level, but are actually affectedby larger-scale cultural, economicand political factors. Some of these are:• lack of crop/agriculture insurance toprotect against climatic risk;• almost complete absence of rural creditsystems;• weak healthcare systems, including lowavailability of medicines;• weak extension systems in terms of personnel,transportation and motivation;• poor infrastructure and markets leadingto unattractive prices for outputs andinputs; and• poor or non-existent rights to land forwomen, migrants, settler communitieson state land and in a variety of othersituations.Some of these large-scale factors stem frommarket or even social failures, but all ofthem are driven by policy and are thereforealso policy failures.

74World Agroforestry into the FuturePolicy factors contributingto poor smallholder landmanagementIn this section, we explore in more detailsome of the policy problems related directlyto land management and neededreforms. This list is not exhaustive, butdoes reflect the priorities noted by severalother recent authors (Dorward et al. 2004;Hazell and Johnson 2002; Lee et al. 2001).We also recognize that in fact there havebeen positive reforms undertaken in manysectors that support smallholder agriculturein general and land management in particular,for example, policies with a focuson poverty, decentralization and new landlegislation.Enabling incentives for theagricultural sectorIt goes without saying that incentives providedto rural communities and agricultureare critical in promoting improved landmanagement. Policy incentives can coversuch areas as market, pricing, infrastructureand credit. In some cases, even withgrowing pressure on resources, gettingthese policies right can lead to improvedland management (this has been thecase in Machakos, Kenya, for instance).However, in many cases, merely puttingwell-intentioned broad policies in placeis not sufficient to ensure sustainable landmanagement, as witnessed by low nutrientinputs throughout much of Africa. Whileacknowledging the importance of theseenabling policies, we do not discuss themin detail, instead devoting more time tospecific land management policy issues.Readers may refer to the studies noted inthe introduction to this section for moreinformation.Land tenure policyThere is no disagreement that propertyrights are important in shaping incentivesfor investment, management, mortgagingand transacting in land (for investment insoil management, see Prudencio 1993, andManyong and Houndekon 2000 for examples).There is no property rights systemthat is ideal for all circumstances, but thereare many recognizable faults with someexisting systems (see Deininger 2003 for agood summary). Some policy problems arean effect of unresolved struggles for authoritybetween formal and informal tenureinstitutions (e.g. state versus chiefdoms inareas of western and southern Africa), orthey may be caused by excessive state controlover land and other resources (e.g. inEthiopia, Indonesia and the Philippines), oroverlapping or conflicting tenure administrationacross ministries (of lands, agriculture,natural resources, and so on). Thesesituations contribute to poor formulationand implementation of tenure policies thatcan particularly affect the property rightsof women and migrants (or in some caseindigenous peoples). Each of these factorsmay have considerable impact on NRM.By increasing the uncertainty over anyindividual or group’s claims to the benefitsgenerated by the application of resources,these tenure problems reduce incentivesfor investment and management of naturalresources.While in many areas, indigenous tenuresystems offer sufficient incentives for investmentin land management, there remainsome stubborn problems to resolve.As a general principle, the clearer the policiesare (i.e. no overlapping inconsistenciesor gaps), the better, all else being equal.Furthermore, there needs to be functionaland fair property rights enforcement mechanismsin place to back up those policies.This calls for local services and for localand national administration to be linked.For the specific tenure problems that exist,transparent resolution processes need tobe developed, substantiated by scientificinformation about the desirability of anyspecific new tenure arrangement.One lesson that has been learned fromrecent reforms is that national policy alonewill not necessarily lead to tenure changeon the ground (for example, with respectto women’s rights; Razawi 2003). Suchchange also requires interventions on a localscale, which may include policies butwill also need such institutional reforms asa fairer judicial system and technologicalinterventions such as practices that womenwill use. Also because tenure systems andtenure security respond to driving forces,indirect policies (e.g. on market liberalization)may lead to desired tenure changes intenure institutions.Land/soil management policyLand management policy is often the amalgamationof dozens of laws and hundredsof regulations, established by several differentministries and possibly two or moreadministration levels. This may be unavoidableto some extent, yet the result is usuallyconflicting policies resulting in a lack ofclarity at government level and certainly atthe level of land manager. Most countriesare attempting to consolidate policy makingby forming Environmental ManagementAuthorities, but they have not been givenample resources to tackle this colossalmandate. Indeed, the natural resources sectoris often underfunded and must dependlargely on externally funded projects.It is unfortunate that policies concerningdevelopment and welfare goals are oftendetached from those focusing on conservationand the environment. This is evidentin the significant lack of attention that ispaid to land and soil management in thepoverty reduction strategy papers (PRSPs)of African nations, such as Burkina Faso,

Chapter 8: Policies for improved land management in smallholder agriculture75the Democratic Republic of Congo, Kenya,Mozambique, Niger, Tanzania and Zambia.To see how this lack of clarity affects landmanagement, we can take the example ofagroforestry. Land and tree property rights,which affect agroforestry investment incentives,are influenced to some degree byministries of forestry, agriculture, lands,energy and water. In addition, in manycountries, there is insufficient distinctionbetween trees on farms and those in forests.Rules and laws are thus made to applyto all trees, and logging bans and protectedspecies regulations have unintended disincentivesfor farmers to plant and managethe trees on their farms. Agricultural extensionstaff are increasingly active in promotingagroforestry enterprises for farmers, buttree seed is still mainly in the domain offorest departments. At the same time, forestdepartments do not have enough staff to interactsufficiently with farmers. So developmentprogrammes continue to struggle tobuild sustainable agroforestry systems.Thus, the profile of land managementneeds to be raised, and new or existinginstitutions should become focal pointsor coordination units for land and soil.They would be responsible for assemblingdatabases on soils, monitoring changes inland resources and championing improvedland management practices to practitionersand policy units. There are many areas thatrequire attention; here we discuss in moredetail some policy reforms that could contributeto improved soil nutrient managementby smallholder farmers.Most countries’ policies on sources andmanagement of nutrients require attention.The starting point is generally fertilizer prices;in a country where markets and infrastructurefunction well, the ratio of input tooutput prices at the farm is sufficiently lowto attract high demand and use of fertilizer.However, in much of the developing worldthese conditions are not met. To encouragethe use of fertilizer, governments can try toalter the ratio (e.g. to entice farmers to cultivatehigher-value crops) or try to directlyinfluence prices (e.g. by subsidizing fertilizers).Fertilizer subsidization has been triedin the majority of countries, and remains inplace in a few. In countries such as Chinaand India fertilizer application rates arevery high – even greater than those in somedeveloped countries (FAO 2004), leadingto environmental and human healthconcerns. Several governments that discontinuedsubsidies are considering otheroptions to make fertilizers more attractiveto farmers. Some, such as the governmentsof Malawi and Zambia, have embarked onlimited subsidy programmes that are designedto be small starter kits or targeted tothe poor. In other countries, such as Kenya,fertilizer import taxes have been loweredand competition encouraged so that thereare now many large importers, which reducesprofit margins at the importer level(Jayne et al. 2003). None of these policiesby themselves will lead to desirable levelsof fertilizer use because there are still highpoverty rates coupled with poor rural creditavailability. Thus, the exploration of othernutrient sources becomes important.Organic sources of nutrients are gainingrecognition as not only feasible andappropriate, but necessary in certainsituations (Palm et al. 1997; Place et al.2003). Crop–livestock farms are commonthroughout the smallholder sectors ofdeveloping countries, but little of the manureproduced is used as an input to cropproduction because of lack of knowledge,lack of labour, use of manure for energy,and grazing systems that do not favour theconcentration of manure near the farms.More attention needs to be paid to manuremanagement and application as well asto the development of manure markets. Interms of plants, a number of herbaceousand woody legumes have been found toproduce large amounts of organic matterand certain elements such as nitrogen.Moreover, they are cheap to establish andtherefore attractive to the poor. However,policy makers continue to be unawareof these systems and thus have not madethem part of mainstream development programmes.In places, such as most of sub-Saharan Africa, where multiple constraintsto land investment exist at community andhousehold levels it is likely that over thenext 10–20 years the use of nutrients fromall sources – mineral fertilizers, animalmanure and green biomass – will becomeeven more important.Research and extension policySound agricultural and natural resourceresearch systems have emerged in Brazil,China, and India and in some of theother emerging nations of Asia and SouthAmerica, but remain weak elsewhere. Thereasons for weak systems are manifold andinclude under-appreciation and neglect bygovernment, insufficient external funding,poor and rigid management, and low staffmotivation. Another criticism of researchsystems, especially in Africa, is that theydo not integrate well with extension systems.But sometimes it is the extensionsystems themselves that come under intensivescrutiny (World Bank 2003). Disseminatingnew information or fosteringinnovative processes through technicalsupport are major challenges in rural areasof developing countries and are exacerbatedby poor communication infrastructure,multiplicity of languages and high levels ofilliteracy. The poorer countries have beenunable to sustain investment in extensionsystems leading to failures and calls forchange.

76World Agroforestry into the FutureUntil recently, national research programmestended to be organized aroundcommodity sectors, which hampered integratedresearch into land managementissues. Now national research organizationsneed to attract funds to fulfil theirnew mandates in natural resources andland management research. The currentorthodoxy is that for such research to besuccessful, it should embrace the followingguidelines: be demand-driven throughparticipatory research, be interdisciplinaryin approach, be market-oriented, and helpintegrate technologies, institutions andpolicies. In doing so it will be able to offerfarming communities a range of landmanagement options that can overcomeconstraints and generate sustainable productivityincreases. Governments, includingthose in developing countries, need tomake long-term commitments to researchfunding, and must ensure that researchresults are better integrated into nationalpolicies and programmes. The experiencesof Brazil, China and India are a testamentto the extraordinary achievements that canresult from making strong commitments toresearch, and these impacts are being feltaround the world.Extension systems must become more flexiblein their approaches; farmers are increasinglydemanding information on marketopportunities and processing techniques inaddition to their typical production questions.Agents or facilitators must have muchlarger information networks and must haveadequate resources to access the informationand then transmit it to clients. Farmersalso demand different levels of service provision,from simple message transmissionto more sustained technical support. To improveland management, it is necessary topromote knowledge-intensive practices suchas integrated nutrient management as wellas more simple transmission of informationon output prices in different markets. Somesigns of this are occurring, for example,participatory rural appraisal is becomingmainstream, but financial resources are oftenthe binding constraints. There have alsobeen attempts to reduce national governments’share of costs and to move towards afee-for-service system, but the feasibility ofthis for poor smallholder farmers is a majorconcern. How to implement these conceptsin practice has remained elusive for themost part.The role of research inshaping land managementpolicy reformsContributions by ICRAF and itspartners in policy reform processesAlthough the Centre has not emphasizeda policy research programme, it has hadmodest success in contributing to positivepolicy change. One of the key policy areasin which the Centre has undertaken a significantamount of research is land and treetenure in Africa and Asia (Fay et al. 1998;Otsuka and Place 2001; Place 1995; Tomichet al. 1997; Traore and van Dorp 2004).This has helped to generate knowledge onthe effects of different tenure arrangementson NRM, and has led to several local tenurechanges in Africa, the development ofa pilot tenure reform in Indonesia (Tomichet al. 1998), and the provision of technicaladvice to global tenure reform processes(e.g. presentations at World Bank and UN-Economic Commission for Africa workshopson tenure issues for Africa).ICRAF and its partners (national agriculturalresearch institutes and universities)have undertaken a large number of studiesrelated to understanding farmers’ NRMdecision-making processes and outcomes.This has also taken place throughout Africaand Asia and has helped to identify majorconstraints and opportunities on whichnew policies and programmes can bebased (Barrett et al. 2002). These have beenpublished in several different media, includingpolicy briefs, and have found theirway into some of the major developmentinitiatives from the United Nations MillenniumProject and the World Bank.Apart from policy research per se, the Centrehas invested significant effort into seeingthat its broader research results reachpolicy makers. The Centre has helpedto bring the concept of agroforestry intomainstream policy documents, researchprogrammes, extension services and developmentprojects. Examples include:the formation of agroforestry research programmesin many African national researchinstitutions, where previously they did notexist; the creation of agroforestry curriculain more than a hundred universities andcolleges throughout the developing world;generation of national agroforestry strategiesand networks in some countries; theexplicit mention of agroforestry in increasingnumbers of global and national policydocuments; and contribution to specificmajor policy and programmatic initiativessuch as the New Partnership for Africa’sDevelopment (NEPAD) agricultural strategy(NEPAD 2003), the World Bank’s SoilFertility Initiative and the UN’s Hunger TaskForce strategy to reduce hunger in Africa.Supporting future policy reformsThere are three major areas where researchfrom an international organization such asthe Centre can best support policy debatesand reforms and improve smallholder landmanagement.1. Identification of land managementproblems and opportunities. Thisinvolves the understanding of how bio-

Chapter 8: Policies for improved land management in smallholder agriculture77physical problems of land degradationvary and how they interact with poverty,policy and other drivers and incentives.Research work must therefore analysenot only land management problems butalso the underlying policy and incentivestructures that shape land management.In essence, this research will identify priorityareas for intervention based on thesignificance of problems and opportunitiesfor rehabilitation or enrichment. The Centreis well placed to do this since it can drawupon research from a range of settings.2. Design of investment/responsepriorities. Research should identify theinvestments – in technological, institutionaland policy reforms and programmes –that will lead to the greatest impact insustainable land management for povertyreduction. Some of the crucial areas ofresearch will be: synthesizing lessonslearned from around the world; analysingtrade-offs and synergies of differentinterventions or sets of interventions;identifying entry points or sequencingpatterns of investments; and analysingalternative implementation strategies.It will be particularly important to assessinterventions comprehensively in terms oftheir social, economic and environmentalimplications as well as implications atdifferent spatial and temporal scales.3. Monitoring the impact of policyreforms. This will involve collaborationwith policy makers in the pilot testing ofland management policy and programmereforms and the monitoring and assessmentof these interventions. Governments ofdeveloping countries are generally weakin the area of monitoring, while an internationalcentre is in a good position todevelop comparative studies of reformsand other interventions across countries.Such studies will not only evaluate specificinterventions in certain countries butshould also be able to understand thefactors associated with how successful anintervention is, and to identify beneficialmodifications to reforms or interventions.These are very broad research areas; theCentre will need to focus on areas connectedto agroforestry or, more broadly, tonatural resources in an agro-ecosystem.Building on strengths, the Centre shouldcontinue to give emphasis to issues such as:germplasm supply; property rights (includingintellectual property rights); communityand watershed management and regulation;the interactions between poor communitiesand farmers on the one hand and landmanagement and degradation on the other,especially among the poor; research and extension;markets; commercialization; NRM;and the potential for promoting environmentalservices and improved land managementin smallholder communities.In order to have impact in these areas, theCentre will need to be active in two ways.Firstly, it should directly participate inresearch and advocacy activities in highpriorityareas where it may generate globalpublic goods. It will be most effectivewhen attempting to learn or disseminatelessons across regions and countries, givenits global position. Secondly, the Centreshould be a champion for promoting policyresearch and dissemination on issues thatare vital for improving land management.This will involve public awareness, such asthe dissemination of technological successstories and presentations at key nationaland international fora. It will also requirecapacity building of research and advocacygroups so that they may integrate agroforestryissues into their own agendas. Clearly,the latter approach is going to be morecritical in having policy impact in a largenumber of countries so that more and moresmallholders might be able to benefit fromimproved land management.ReferencesBarrett, C.B., F. Place and A. Abdillahi 2002.Natural Resources Management in AfricanAgriculture: Understanding and ImprovingCurrent Practices. CAB International,Wallingford, UK.Deininger, K. 2003. Land Policies for Growthand Poverty Reduction. World BankResearch Report. The World Bank,Washington, DC, USA and OxfordUniversity Press, Oxford, UK.Dorward, A., J. Kydd, J. Morrison and I. Urey2004. A policy agenda for pro-poor agriculturalgrowth. World Development32(1): 73–89.FAO (Food and Agriculture Organization of theUnited Nations) 2004. FAOSTAT. (http://www.faostat.fao.org)Fay, C., H. de Foresta, M. Sirat and T.P. Tomich1998. A policy breakthrough for Indonesianfarmers in the Krui Damar agroforests.Agroforestry Today 10: 25–26.Foster, M. and P. Mijumbi 2002. How, Whenand Why does Poverty get Budget Priority:Poverty Reduction Strategy and PublicExpenditure in Uganda. Overseas DevelopmentInstitute Working Paper 163,London, UK.Hazell, P. and M. Johnson 2002. Ending Hungerin Africa: Only the Small Farmer Can DoIt. International Food Policy Research Institute(IFPRI) Brief. www.IFPRI.org.Jayne, T., J. Govereh, M. Wanzala and M. Demeke2003. Fertilizer market development:a comparative analysis of Ethiopia,Kenya, and Zambia. Food Policy 28:293–316.

78World Agroforestry into the FutureKaiser, J. 2004. Wounding Earth’s fragile skin.Science 304: 1616–1618.Lee, D., C.B. Barrett, P. Hazell and D. Southgate2001. Assessing Tradeoffs and Synergiesamong Agricultural Intensification, EconomicDevelopment and EnvironmentalGoals: Conclusions and Implications forPolicy. In: Lee, D. and C.B. Barrett (eds)Tradeoffs or Synergies? Agricultural Intensification,Economic Development andthe Environment. CAB International, Wallingford,UK.Manyong, V. and V. Houndekon 2000. LandTenurial Systems and the Adoptions of aMucuna Planted Fallow in the DerivedSavannas of West Africa. Collective Actionand Property Rights Working Paper No. 4,International Food Policy Research Institute,Washington, DC, USA.Mwabu, G. and E. Thorbecke 2001. Rural Development,Agricultural Development,and Poverty Reduction in sub-SaharanAfrica. Paper presented at the AfricanEconomic Research Consortium BiannualResearch Workshop, December 1–6 2001,Nairobi, Kenya.NEPAD (New Partnership for Africa’s Development)2003. Comprehensive African AgricultureDevelopment Programme. (www.nepad.org).Oldeman, L.R., R. Hakkeling and W.G. Sombroek1991. World Map of the Status ofHuman-induced Soil Degradation: An ExplanatoryNote (2 nd edition). InternationalSoil Reference and Information Centre,Wageningen, the Netherlands.Otsuka, K. and F. Place 2001. Land Tenure andNatural Resource Management: A ComparativeStudy of Agrarian Communities inAsia and Africa. International Food PolicyResearch Institute, Washington, DC andJohns Hopkins University, Baltimore, USA.Palm, C., R. Myers and S. Nandwa 1997.Combined use of organic and inorganicnutrient sources for soil fertility maintenanceand replenishment. Pp. 193–218in: Buresh, R., P. Sanchez and F. Calhoun(eds) Replenishing Soil Fertility in Africa.Soil Science Society of America SpecialPublication No. 51, Soil Science Societyof America, Madison, USA.Place, F. 1995. The Role of Land and Tree Tenureon the Adoption of Agroforestry Technologiesin Uganda, Burundi, Zambia andMalawi: A Summary and Synthesis. LandTenure Center, University of Wisconsin,Madison, USA.Place, F., C. Barrett, H. Freeman, J. Ramischand B. Vanlauwe 2003. Prospects for integratedsoil fertility management using organicand inorganic inputs: evidence fromsmallholder African agricultural systems.Food Policy 28: 365–378.Pretty, J. 1995. Regenerating Agriculture: Policiesand Practice for Sustainability andSelf-Reliance. Earthscan Publications,London, UK.Prudencio, C.Y. 1993. Ring management of soilsand crops in the West African semi-aridtropics: The case of the Mossi farmingsystem in Burkina Faso. Agriculture, Ecosystem,and Environment 47: 237–264.Razawi, S. 2003. Agrarian Change, Gender, andLand Rights. Blackwell Publishers, Oxford,UK.Scherr, S. and P. Hazell. 1994. Sustainable AgriculturalDevelopment Strategies in FragileLands. Environmental and ProductionTechnology Development Division DiscussionPaper No. 1, International Food PolicyResearch Institute, Washington, DC, USA.Tomich T.P., J. Kuusipalo, K. Menz and N. Byron1997. Imperata economics and policy.Agroforestry Systems 36: 233–261.Tomich, T., M. van Noordwijk, S. Budidarsono,A. Gillison, T. Kusumanto, D. Murdyyarso,F. Stolle and A. Fagi 1998. Alternatives toSlash and Burn in Indonesia: SummaryReport and Synthesis of Phase II. ASB-Indonesia Report No. 8, ASB-Indonesiaand International Centre for Research inAgroforestry, Southeast Asia Programme,Bogor, Indonesia.Traore, C. and M. van Dorp 2004. Atelier surl’impact de la reglementation forestieresur les systemes agroforestiers. Proceedingsof a workshop held in Segou, Mali,8–10 December 2003.United Nations 2000. United Nations MillenniumDeclaration Resolution Adopted bythe General Assembly of the United Nations,55/2. 18 September 2000. UnitedNations, New York, USA.United Nations 2003. World Urbanization Prospects:The 2003 Revision. United NationsPopulation Division, Rome, Italy.Wiebe, K. 2003. Linking Land Quality, AgriculturalProductivity and Food Security.Agricultural Economic Report No. 823,Economic Research Service, United StatesDepartment of Agriculture, Washington,DC, USA.World Bank 2000. World Development Indicators.World Bank, Washington DC, USA.World Bank 2003. Extension and Rural Development– Converging Views on InstitutionalApproaches? Summary of workshopheld on November 12–15 2002, at theWorld Bank, Agriculture and Rural DevelopmentDepartment, Washington, DC,USA.

Chapter 9Land and People:Working Group ReportAfter some discussion, four working groups wereformed to brainstorm over priorities for the Land andPeople research and development agenda in the comingyears. The four groups corresponded to the topicsof the presentations: land and soil management, landmanagement policy, scaling up of agroforestry interventions,and the scientific challenges for the Centrepresented in Chapter 5.Working group on land and soil managementThis group identified two primary topics under land andsoil management where it felt ICRAF should significantlyincrease its role. The first is to develop relevant and usefulland-quality indicators. Within this, it was recognizedthat much research had already been undertaken thatdemonstrated the critical role of organic carbon (C) insoil fertility processes. However, more work should bedone to understand the links between organic C and soilqualityindicators. Furthermore, soil biology databasesand the links between different soil biota and soil fertilityrequire strengthening. This work needs to be conductedwithin a strategic framework with the landscape as a keyunit of research, which could have the additional resultof describing land capability classes that can guide decisionsabout future land-cover dynamics.The second area on which the Centre should concentrateis the greater application of geographic informationsystems (GIS) and remote-sensing tools. Thesecan guide priority setting for research on agroforestrytechnologies for improved land management, assist theCentre in moving innovations to scale and allow theorganization to identify strategic sites for testing thatexhibit different soil problems. GIS can also be use-ful in informing current development efforts. In termsof geographical emphasis, the working group felt thatthere was a need for ICRAF to pay greater attention tothe semi-arid and arid lands, where degradation andpoverty hotspots occur.Achievement in these two areas should produce ‘internationalpublic goods’, in particular publicly availableinformation that can be efficiently stored in databaseson soils and used in decision-support guides on technologicaloptions for different problem areas.The working group identified three key roles for ICRAFin this research:• contribute to better diagnosis of problems, andidentify entry points for agroforestry;• build capacity and develop human resources(including in civil society and for smallholderfarmers); and• produce research and development outputs inaccordance with the Centre’s comparative advantage.Working group on policyThis working group spent considerable time takinginto account the internal processes under which policyresearch is identified and implemented at the Centre.The group felt that more discussion is needed to answerthe following: How are policy research topics selected?What is the priority-setting process? Is this processbased on existing global or national strategies or policydocuments?The working group concluded that the manner inwhich policy research themes and directions are

80World Agroforestry into the Futureselected at ICRAF needs to be more transparentand rigorous.With regards to the output of policy research,there was considerable discussionwithin the group about whether it waspreferable to focus on developing a processor a product. Some experiences showthat leading with process work can havemore of an effect and creates its own demandsfor specific research products. Thisapproach should be considered throughoutthe institution. The next question waswhether the Centre should therefore beproactive in shaping policy processes andagendas. The group felt that building upfarmer representative groups as a voice fordefining and lobbying for priority policyagendas would enable the organization tomake better decisions.The group also recommended that ICRAFgives due attention to the administrativelevel and position of the policy makers withwhom it interacts. Although the organizationis in the process of learning about appropriatelevels of engagement, from localto international, nevertheless it is importantto be aware that specific policy problemsoften require specific types of decision makers.The Centre can help to bring policymakers from different sectors and scalestogether with researchers to debate policyissues. Even here, there are many optionsand pathways to choose from. Therefore, itis important to spend more effort on identifyingthe policy makers with whom ICRAFwishes to influence and interact, what thesepolicy-makers need, and how to communicateeffectively with them.Working group on scaling up,adoption and impact – reaching thevulnerable groupsThis working group discussed lessons fromits experiences both at the Centre and inother organizations. The group recognizedthat scaling up should not take a broadbrushapproach. The critical, vulnerablegroups should be identified and targeted,and different approaches are needed foreach one. Reaching the most-vulnerablepeople can be difficult; some studies havefound that they are less likely to belong toeasily identifiable groups in villages. So isit possible to reach such solitary farmers?GIS tools and maps that have advancedour ability to target technologies to specificagroclimatic conditions may be useful inidentifying these socioeconomic variablesand targeting vulnerable groups.Not only should scaling up efforts be targetedto particular groups, different situationsrequire different approaches. And notall situations require elaborate scaling upmechanisms: the group noted that sometechnologies ‘sell themselves’ and undergorapid spontaneous diffusion. Therefore, onelesson to learn is that we need to ensurethat all new technologies work well in theeyes of the farmers.The group agreed that ICRAF should investmore resources in learning from experiencesof scaling up, especially those thatinvolved different situations and technologies.Certain questions need to beanswered: What are the widely applicablemethods and lessons for disseminationprocesses? What is known about successfulextension materials? How can cadresof local people be persuaded to share theirexperiences? Institutionalization of lessonsis also important, to ‘cement them in’ andenable them to be built upon. Therefore,what are the best ways to take such lessonsinto mainstream extension, education andpolicy?The diffusion of new innovations/modificationsis a complicated process that facesmany challenges. However, it can bestrengthened by better horizontal and verticalfeedback systems that spread informationabout innovations. This will also helpto prevent negative consequences, such aswhere new information becomes dilutedduring the diffusion processes and leads toundesirable management practices.For adoption and impact of new technologies,researchers need to look at a rangeof indicators that relate to poverty. Thisis because some indicators move slowly,while others may not show a significanteffect from only one agroforestry improvement.This problem is exacerbated by alack of understanding of the key drivers ofadoption. More attention should be paidto studies of mature technologies. Oneparticular research gap the group identifiedwas how adoption of various practices isaffected by human health.An even weightier question the workinggroup discussed was the balance betweenreach and impact. Is it better for ICRAF totry and reach a large number of farmerswith little impact or fewer farmers but withlarger impacts? For example, should it aimfor broad productivity impacts or should ittry to invest more in targeting the poor andvulnerable? The group agreed that the Centrehas started on the right track by formingpartnerships, but that it does not yet haveall the right ones (i.e. improvements couldbe made in partnerships for research, educationand development) and there is stilltoo much fragmentation.Finally the group concluded with a brainstormingsession of what it considers to bepossible international public goods outputsin the area of scaling up:• methods that analyse scaling up experiencesand which could be applied bypartners;

Chapter 9: Land and people81• scaling up approaches that could beapplied to many situations over widerscales;• strategies for testing of unfinished technologieswith farmers to fine tune theirapplication;• use and integration of high-level scientificapproaches such as GIS and biotechnology;• building capacity among national researchand development organizations toenable ICRAF to effect exit strategies; and• identification of the Centre’s comparativeadvantage in dynamic research anddevelopment environment.Working group on Mike Swift’schallengesThe working group addressed many of thechallenges presented in the opening sessionon Land and People which was ledby Mike Swift, eminent soil scientist andformer Director of the Tropical Soil Biologyand Fertility Programme. Here is a list of thechallenges followed by the responses of theworking group that were more of a reflectivebrainstorm about the challenges.What are the key determinants of the adaptiveand adoptive advantages of the availabletechnological options? What works,where do they work, when do they work,and for whom do they work?• There is a need to be able to better predictthe performances of different technicaloptions under different conditionsExplore whether the creation of targetzonedomains for various options basedon biophysical and socio-economicconditions is practical and beneficial.• Test options across varying conditions sothat limits of the options are known.• Conduct risk analysis for various options.What are the main questions arising fromthe interactions in the chain that goes fromresource management to system intensificationto market access to policy?• Our analysis of natural resource management(NRM) programmes shouldstart with determining what markets existfor goods and services in an area andthis will drive our research agenda.• Our approaches should take a holisticyet incremental approach (covering markets,policies, institutions and so on).• Based on one aspect of a very complexproblem, try to look at an integratedmodel of the situation and work in amultidisciplinary team.What are the rules governing transitionsacross spatial scales (for example, hownutrient cycling processes or biodiversityfunctionlinkages differ at plot, farm andlandscape scales)?• Be aware of the spatial scale of the workin both biophysical and socioeconomicresearch and be able to better link researchthat takes place at different scales.• Some of the work within the Centredoes deal with issues at various scales– this is very innovative science. Theseinteractions and transitions are complex,which makes ICRAF’s work in thisarea complex compared to research oncrops. However this work has practicalpayoffs and should be continued, despitethe difficulties.What are the trade-offs between the storageof organic matter in the soil and its use todrive nutrient cycling, crop production andother ecosystem services?• Much is known about nitrogen (N) andits role in crop production.• We need to go beyond N and studyother nutrients, issues such as soil acidity,and environmental services, and linkthis with price, policies, farmers’ needsand the resilience of agroecosystems.This is an area that is largely unexplored.How can the functions of the soil biologicalcommunity be optimized with respectto different ecosystem services?• Functional biodiversity (as opposed tospecies richness) is the key issue in relationto various ecosystem services aswell as such issues as soil health, soilmacrofauna, soil microbes, and soilphysical properties.• These components/systems need to bestudied in relation to agroecosystemstability, productivity, and resilience ofsystems.• This type of research should be pursuedvigorously over the next 5 years.

Enhancing Environmental Services

“Forest conservation is no longerseen as the only appropriatemeans to achieve environmentalconservation, nor is afforestationseen as the only way to reverseenvironmental damage.”Swallow et al.

Keywords:Property rights, land tenure, buffer zones,forest classification, payments for environmental services,Africa, AsiaChapter 10Agroforestry and environmentalgovernanceBrent Swallow, Diane Russell and Chip Fay, World Agroforestry CentreAbstractEnvironmental governance is in a state of change throughout the developing world. Power and authorityare shifting from national offices to global and regional fora and to local user groups. Regulatoryapproaches to environmental management are gradually being augmented by incentive- and marketbasedapproaches. Private organizations and firms are becoming more involved in the provision of suchenvironmental goods as water, energy and timber, and environmental services like conservation andwatershed protection. Forest conservation is no longer seen as the only appropriate means to achieveenvironmental conservationm, nor is afforestation seen as the only way to reverse environmental damage.Integrated approaches to ecosystem and landscape management, which include local residents asimportant partners, are being given more emphasis. These trends are creating new opportunities andconstraints for agroforestry. While there are very few pieces of legislation or rural institutions that focussolely on agroforestry, there are many laws and rural institutions that shape farmers’ incentives to plantand manage trees in their agricultural landscapes. This chapter reviews the five policy issues that havegreatest impact on agroforestry: land and tree tenure, forest classification, biodiversity and forest conservation,environmental service reward mechanisms, and global environmental governance. Targetedapplied research and engagement in local policy processes increases the beneficial impacts of agroforestrydevelopment within local policy terrains and contributes to policy reform at the national andglobal levels.IntroductionEnvironmental governance – including all policies andinstitutions affecting the state of the environment – isin a state of change throughout the developing world,with significant consequences for on-farm tree plantingand management. Changes are occurring in severaldimensions. Formal authority is shifting from nationalforestry agencies to decentralized multistakeholdercommittees and local user groups. Rules and prohibitionsare gradually being augmented by incentive- andmarket-based approaches to environmental management.Private firms are becoming more involved in theprovision of such environmental goods as water, energyand timber and environmental services like biodiversityconservation and watershed protection. Integrated approachesto ecosystem and landscape management,which include local residents as important partners, arebeing given more emphasis in international agreementsand the programmes of influential international organizations(Tomich et al. 2004).

86World Agroforestry into the FutureWhile few developing countries havespecific laws or policies on agroforestry,a range of environment and developmentpolicies and structures of administrationand governance affect the practice. Herewe follow the Leakey (1996) definition ofagroforestry as a “dynamic, ecologicallybased,natural resource management systemthat, through the integration of trees onfarms and in the landscape, diversifies andsustains production for increased social,economic and ecological benefits”. Thisdefinition has three direct implications forgovernance. First, agroforestry involves thedeliberate management of trees, includingtree planting and various intensities offarmer management of trees in multiplefunction landscapes. Second, it depictsagroforestry as a natural resource managementsystem that includes land use practicesand the institutions (including rules,regulations and norms) that shape thoseland use practices. Third, the definitionexplicitly recognizes agroforestry as a landusesystem practiced at the farm and landscapescales. Institutions and policies thatgovern land use and environmental managementat those scales will affect farmers’incentives to plant and manage trees.At the farm scale, the most important institutionalarrangement affecting agroforestryis property rights. Property rights to landand trees on farms shape farmers’ expectationsof whether and how they will beable to appropriate long-term benefits frominvesting in tree management and planting.Property rights are also important atthe landscape scale since property rightsregimes (state, common or open access)governing tree resources outside of individualfarms affect the use of those resourcesand the incentives for farmers to plant treeson farm. One of the key determinants ofproperty rights to trees outside of privatefarms is the system of forest classificationand governance. State systems of forestgovernance generally reflect a combinationof state control of valuable forest resourcesand concern for the public interest in theenvironmental services that they provide.As property rights and forest governancesystems have evolved over the last two decades,other governance arrangements havebecome important. The focus of biodiversityconservation has widened from lookingsolely at protected areas to including theirboundaries and the surrounding landscape.Agroforestry is recognized as having unfulfilledpotential to contribute to biodiversityconservation at the landscape scale. Environmentalservice reward mechanisms are beingexplored in some locations, with agroforestryoften seen as a desirable land use from theperspective of biodiversity conservation,carbon sequestration, renewable energy productionand reversal of land degradation. Thegrowing importance of global environmentalagreements is increasing motivation for someof these environmental service mechanisms.The widespread implementation and national‘domestication’ of global environmentalagreements provides a mix of opportunitiesfor and constraints to agroforestry.This chapter reviews evidence concerninglinks between agroforestry and the fivecomponents of environmental governancedescribed in the previous two paragraphs:i) property rights to land and trees; ii) landclassification; iii) biodiversity and forestconservation; iv) environmental servicereward mechanisms; and v) global environmentalgovernance. Table 1 summarizesinformation on the links between each ofthese components and agroforestry.The World Agroforestry Centre (ICRAF)works with a range of partners to implementa three-pronged approach to addressthese policy challenges. Firstly, we seekto enhance understanding of the links betweenagroforestry, forestry, protected areamanagement and social objectives relatedto the environment. In some cases, the resultschallenge conventional wisdom andconventional approaches to environmentalmanagement. Widespread disseminationof key principles and empirical findings isachieved through scientific publicationsand engagement in local, national and internationalpolicy fora. Secondly, we seekto broaden understanding of how policiesand institutions affect the incentives offarmers to manage and plant trees in highprioritysituations. Commonalities and contraststend to emerge across research sites,implying that there are no universal policysolutions. Thirdly, in high-priority situationswe work with policy makers and policyshapers to promote reform or effective implementationof policies and regulationsthat have high impact on the effectivenessof agroforestry. The following sections summarizelinks to agroforestry, relevant researchfindings and policy impacts for thefive components described in Table 1.Property rights to land andtreesA large body of literature on the relationshipsbetween property rights and treemanagement has grown up during the25 years since the Centre was founded.While economic theory indicates straightforwardrelationships between tree plantingand land tenure security, the evidenceindicates complex interrelationships betweenmanagement of natural vegetation,tree planting, perceptions of land and treetenure security, gender relations and theoperations of customary and formal tenurearrangements. Uncovering the more complexrelations requires research approachesthat draw upon institutional economics,social and economic theories of innovation

Chapter 10: Agroforestry and environmental governance87Table 1.Characterization of the links between five components of environmental governance and agroforestry.Relation to farmer incentiveto deliberately manage treesLocation of policy makingrelative to farmerTrends in policy and governancecontextProperty rights to landand treesFarmer assurance of future benefitsfrom current investments; farmerincentive to obtain tree productson own farm or elsewhere in thelandscapeLocal norms; decentralized governmentagencies; national policiesGradual individualization; decentralizedstate agencies generally becomingmore importantForest classificationand governanceFarmer incentive to obtain treeproducts from forest areas; farmerincentive to manage and protectnearby forestsDecentralized forest agencies;national forest agenciesDecentralization of state agencies;some movement away from commandand control approachBuffer zone and landscapeapproachesto conservationIncentives/disincentives to managetrees near protected areas; typesof trees allowed and encouraged indifferent parts of the landscapeDecentralized conservation/forestagencies; national forest agencies;international conservation pressuresMixed success with integratedconservation and developmentprojects; landscape approachesstill largely experimentalEnvironmental servicemechanismsMost environmental service mechanismsinvolve tree and vegetationmanagement by individual farmers,groups and/or local governmentsRegional dialogue for watershedservices; national policies and internationalmechanisms for biodiversityand carbonBecoming part of government approachesin many countries in LatinAmerica; small experiments in otherregionsGlobal environmentalgovernanceUNFCCC 1 , UNCBD 2 , UNCCD 3 andGEF 4 all have significant forestrycomponents, with inadequate provisionfor smallholder agroforestry; nospecific forestry convention sincethe Rio conference in 1992National ratification and domesticationof global agreements and fundingopportunitiesUNFCCC has progressed furthestin explicitly considering potential ofagroforestry and smallholdersSource: Authors’ summary from this chapter and literature review.1United Nations Framework Convention on Climate Change2United Nations Convention on Biological Diversity3United Nations Convention on Combating Desertification4Global Environment Facilityand collective action, and a variety of quantitativeand qualitative research tools. Muchof this research has been conducted in associationwith the Collective Action andProperty Rights Initiative of the ConsultativeGroup on International Agricultural Research(CGIAR) (Meinzen-Dick et al. 2002).In the 1980s and early 1990s, much of theevidence on the links between agroforestryand property rights in Africa emerged fromjoint efforts by ICRAF and the Land TenureCentre at the University of Wisconsin–Madison(Bruce 1989; Fortmann 1985; Fortmannand Bruce 1988; Place 1995 and Raintree1987). Bruce (1989) summarizes the resultsof these studies by noting that agroforestryprojects may be associated with severalproblems of land and tree tenure. Firstly, aproject may disturb or destroy rights to otherimportant uses of the land or trees. Secondly,customary tenure systems that providemultiple uses of land and tree resourcesmay make it difficult for individual farmersto protect tree seedlings. Thirdly, some categoriesof intended clients may be unable toparticipate in a project because they do nothave the right to plant or own trees. This includeslandless people and women in somesocieties. Fourthly, farmers may undertaketree planting as much to establish rights toland as for the direct products of the trees.In the mid-1990s, ICRAF, the InternationalFood Policy Research Institute (IFPRI) andTokyo Metropolitan University engaged

88World Agroforestry into the Futurenational partners from across Asia and Africain studies of the effects of land tenureon tree management at the farm and communityscales (Otsuka and Place 2001;Suyanto et al. 2001). Several importantresults have emerged from these and othersimilar recent studies. Firstly, most customaryland rights systems provide sufficienttenure security to encourage deliberatetree management in at least some land useniches (although some state-imposed tenuresystems have actively discouraged treemanagement or created de facto open accesssituations that encourage overuse andunder-investment in long-term tree management).Secondly, land tenure security,tree planting and management of naturalvegetation are interdependent in manycustomary societies in Asia and Africa.Because both clearance of natural vegetationand tree planting are markers of landimprovement, it is possible to observe bothreduced and increased tree cover as landrights become more individualized (Placeand Otsuka 2002; Suyanto et al. 2001).Unruh (2002), for example, describes theimportance of cashew trees for markingland claims in post-conflict Mozambique.Thirdly, many African and Asian societiesand national governments have propertyrights systems biased against womenplanting and managing long-term agroforestryinvestments such as timber trees andwoodlots (e.g. Fortmann 1998). Even insuch systems, however, women are oftenable to benefit directly from short-termagroforestry investments such as improvedfallows (Gladwin et al. 2002), processingand marketing interventions that add valueto agroforestry products and intra-familyallocation mechanisms that distribute thebenefits and costs of longer-term investments.Agroforestry interventions targetedto particular niches of land controlled bywomen, such as home gardens, may beadopted in the short term, but in the longterm may encourage men to try to wrestcontrol of such lands (Schroeder 1999).Property rights to land and trees have alarge impact on farm tree managementin the Lampung province of Sumatra, Indonesia.Suyanto et al. (2005) found thatareas designated as protected forests hadmore frequent and devastating fires duringthe Suharto era than in the years since1998, when the regime fell, implying thatthere now is less deliberate use of fire asa weapon to claim property rights. Fayand Michon (2003) describe part of thepolitical motives and ideology underlyinggovernment expropriation of large areas ofland previously used by individual familiesand governed by indigenous ethnicgroups. Fortunately, the greater politicalfreedom (‘reformasi’) and decentralizationthat have developed since the fall of theSuharto regime have created opportunitiesfor enhancing farmers’ tenure security andrevitalizing customary institutions. ICRAF/AMAN/FPP (2003) describe the outcomesof these processes for the disenfranchisementand loss of land rights for millions ofindigenous Indonesian people. Scientistsfrom the Centre have actively contributedto the restoration of indigenous rights in Indonesia(see Box 1). In northern Thailand,Centre scientists and collaborators (e.g.Care–Thailand) are increasing recognitionof the property rights of indigenous communitiesby helping upland tribal groupsto develop accurate maps of their villagedomains. This active engagement in policyprocesses as a facilitator, supplier of empiricalevidence and supporter of indigenouspractitioners of agroforestry is a hallmark ofthe Centre’s approach to agroforestry policy(van Noordwijk et al. 2001).Box 1. Promoting indigenous rightsin IndonesiaScientists from the World AgroforestryCentre have actively contributed tothe restoration of indigenous rights inIndonesia. Our research and engagementwith policy processes in the mid/late-1990s contributed to the first communityforest law in Indonesia: a historicdecree by the Indonesian Minister ofForestry that recognized and protectedthe rights of the Krui community to collectiverights to damar agroforests. Wehave also promoted the subsequent recognitionof the right of the Governmentto designate a forest for such specialpurposes and facilitated dialogue amongindigenous groups. The direct effect wasthat 15 000 households in the Krui areawere granted more secure rights to over35 000 hectares of agroforest land; theindirect effect was that it became easierfor indigenous communities throughoutIndonesia to register rights to agroforestlands.Source: ASB 2001Forest classification andgovernanceClassification of forestland is a relatedaspect of environmental governance.Throughout the developing world, largetracts of land have been declared as stateforests. There are two key components ofthis designation: state and forest. Both haveimplications for farmer incentives to plantand manage trees. Centre research in thisarea is motivated by the general question:how can forest policies be reformed to

Chapter 10: Agroforestry and environmental governance89harmonize the stewardship of trees in thelandscape and on individual farms?Colonial-era forestry was designed aroundthe needs of the colonial state, i.e. promotionof exports and large industry andcontrol over local communities. Policiesof extraction and central control had theirroots in the feudal systems of medievalEurope. Policies designed to protect theenvironmental values of forests by excludingpeople began in Europe only in theearly 1900s and spread to colonial areasin the later stages of the colonial era (Fayand Michon 2003). Anthropologists suchas Cronon (1996) have shown that manyforested landscapes now considered to bepristine primary forests have, in fact, hadlong histories of human impact. Forceddisplacement of people from conservationareas has led to the impoverishment oftens of thousands of former forest dwellersin the Congo Basin (Cernea and Schmidt-Soltau 2003) and to long-term conflictbetween rural people and governments inmuch of Southeast Asia (Fay and Michon2003; Tomich et al. 1998).A large body of research and experiencehas accumulated on central forest managementand the advantages and disadvantagesof devolution to municipal governmentsand local community groups (Agrawal andOstrom 2001; WRI 2003). Three majorconcerns have been expressed about stateforest management: i) some state agenciesand agents have used their positions as forestregulators as platforms for extractionof resources through concessions to commercialcompanies and/or bribes; ii) stateagencies are accused of not understandingor respecting the ways that local communitiesand indigenous people use and manageforests; and iii) many state forests arehighly degraded. As a consequence, therehas been undue suffering for many forestdwellers. Although state forest authoritiescontinue to operate in many countries,some countries have experienced realdevolution of authority away from centralauthorities toward communal managementand co-management regimes. Differentvariants have developed in different places:joint forest management in India, extractivereserves in Brazil, community forest usergroups in Nepal, and community forests inCameroon (see also Box 1).In the last few years, Centre scientists inIndonesia have given more attention on the‘forest’ part of state forest management. Fayand Michon (2003) argue that the forestryregulatory framework has been inappropriatelyapplied to large parts of Indonesia andother countries in Southeast Asia. In manyinstances this classification has, in fact, beenan a posteriori justification that suits thedominant political and economic interestsand disenfranchises smallholders, whilefavouring large-scale plantations and forestconcessions. Farmers practicing agroforestryhave suffered as a result. Fay and Michon(2003) argue that the forestry regulatoryframework should instead be reserved forareas that clearly protect ‘public environmentalservices’ such as watershed protectionand biodiversity conservation. Areasthat don’t generate such environmental servicesshould be reclassified under the less-restrictiveagrarian regulatory framework.The Forest Codes of francophone West Africaare renowned for the disincentives theyprovide for participatory forest managementand agroforestry. Some of them havechanged since the 1990s, with the paceof change varying from country to country(Russell et al. 2001). Cameroon introducedcommunity forestry in the late 1990s andseveral lessons have been learned from itsexperience. Niger passed its new ForestryAct in April 2004. A new forestry code forthe Democratic Republic of Congo is stillunder consideration with much discussionon the role of communities in a situationwhere all land and resources continue tobe legally the property of the state. Ashleyet al. (2005) shows that continued uncertaintyabout forest classification is creatingdisincentives for agroforestry and forestmanagement in Cameroon and Mali.Buffer zone and landscapeapproaches to conservationThere is now general agreement that conservationof valuable natural resourcesand biodiversity requires the designationof protected areas and better managementof the land surrounding them. Agroforestrycontributes to landscape approaches toconservation by enhancing the diversity ofvegetation in farming areas, increasing thehabitat value of land-use mosaics aroundprotected areas, and reducing pressureon protected areas. Schroth et al. (2004)conclude their review of the potential foragroforestry to contribute to biodiversityconservation with the statement that “…theeffective integration of agroforestry intoconservation strategies is, however, a majorpolicy and institutional challenge”.Attempts to address that institutional challengehave been undertaken in severalcountries, including Nepal and the Philippines(see Box 2). In 2003/4, Centre researchersconducted studies of the policyterrain affecting agroforestry in severalprotected areas (national parks or classifiedforests) in Cameroon, Mali and Uganda.Key conclusions from the studies were asfollows (Ashley et al. 2005):• policy and institutional support to agricultureand agroforestry in buffer zonestends to be very minimal;• extension and development agenciesthat support agroforestry in buffer zones

90World Agroforestry into the Futuretend to focus on a small number of exotictrees, putting little emphasis on theindigenous trees that would be bettersuited from an ecological perspective;• reserved species laws, originally designedto conserve indigenous treespecies, tend to provide disincentivesfor agroforestry; and• the overall policy and regulatory terraintends to have many inconsistenciesbetween forestry, environment and landpolicies.Centre scientists are following up these studieswith targeted research and developmentprojects around protected areas in severalcountries, including Cameroon, Indonesia,Kenya, the Philippines, Thailand and Uganda.The fundamental question still beingasked is: where and how do the integrationand segregation options for human–environmentinteraction have greatest potential tomeet conservation and rural developmentobjectives? (van Noordwijk et al. 1997.)Box 2. Buffer zone approaches in Nepal and the PhilippinesIn Nepal, the Worldwide Fund for Nature and the King Mahendra Trust for NatureConservation created a rosewood plantation/agroforest around the Royal ChitwanNational Park, a valuable conservation area for native forest and wildlife, including theendangered tiger. As part of the Biodiversity Conservation Network, this approach wasmonitored for its effectiveness in both conservation (reducing pressure on park resources)and contribution to local livelihoods. An additional benefit was empowerment of localcommunities in park management (WWF 1997).Environmental servicemechanismsDuring the past decade, there has been increasedinterest in mechanisms linking supplyand demand of environmental services.The environmental services of greatest interestinclude carbon sequestration, watershedprotection and biodiversity conservation.The different environmental services havelargely different populations of demandersand suppliers. Carbon sequestration is aglobal environmental service being financedby emitters of greenhouse gases in thecontext of the United Nations FrameworkConvention on Climate Change (UNFCCC)(see next section). The global benefits ofcarbon sequestration are basically the sameno matter where the carbon is sequestered.This contrasts with environmental servicemechanisms for watershed protection. Inany particular watershed, there may or maynot be specific populations (e.g. urban waterusers) or individual actors (e.g. hydro-powercompanies) who demand watershed protec-In the Philippines, the World Agroforestry Centre was part of a group of organizations thatconducted research and development around the Mount Kitanglad National Park, one ofthe most important biodiversity areas in the country. The Landcare approach to land management,which links community groups, municipal governments and research organizations,was tested in the conditions prevailing around the park boundaries. Hundreds offarmers joined sub-village Landcare chapters around the edge of the Park. After severalyears, this approach has led to improved agricultural production, increased tree cover,and a substantial reduction in encroachment into the Park (Garrity et al. 2002).tion, and specific populations of land userswho can supply those services. Biodiversityconservation falls somewhere betweenthese two extremes; those who demand biodiversityconservation often demand conservationof species and ecosystems at bothglobal and local levels.Several factors account for increased interestin environmental service reward mechanisms.Firstly, many organizations arelooking for new ways to finance conservation.Secondly, changes in the regulatoryenvironment and liberalization of marketsare resulting in increased private-sectorparticipation in conservation, domestic watersupply and carbon offsets. Private firmsappear to be more interested in marketapproaches to protect the integrity of theirresource base. Thirdly, international environmentalagreements are creating spacefor more market-oriented approaches.The Clean Development Mechanism(CDM) of the UNFCCC creates new opportunitiesfor developing-country farmers tobenefit from their contributions to carbonsequestration and renewable energy. Interestin agroforestry has increased since areport by the Inter-Centre Panel on ClimateChange (IPCC 2001) indicated that changesin land use from annual crops to agroforestryis one of the most promising approachesfor sequestering carbon throughCDM-approved afforestation. Although thecarbon sequestration value of agroforestryhas received greater attention to date, thereis also evidence that agroforestry has goodpotential to generate renewable energyin the form of biomass and biodiesel thatcould qualify for the CDM if it can beshown to replace non-renewable sources(Venema and Cisse 2004).Simple calculations show that the monetaryvalue of the carbon sequestration

Chapter 10: Agroforestry and environmental governance91benefits of most tree production systemsare small in relation to the value of thetimber produced. However, Chaco etal. (2002) and Tomich et al. (2002) haveused data from the Alternatives to Slashand Burn (ASB) programme in Indonesiato predict how carbon sequestration paymentswould change the relative returns toalternative land use systems. Their resultsindicate that carbon payments could besufficient to increase returns to smallholderagroforestry systems to levels comparableto those generated by oil palm plantations.This makes agroforestry attractive toCDM since projects must be shown to addvalue to the existing situation. Pilot carbonsequestration schemes with smallholderfarmers are currently in progress in severaldeveloping countries, with the most experienceaccumulated in Latin America. TheCentre is currently involved in pilot carbonsequestration schemes in Kenya, the Philippinesand Uganda.Experience to date shows that institutionaland governance factors determine thefeasibility, performance and impacts ofenvironmental service mechanisms. Formalinstitutions are often designed in ways thatrequire market participants to incur transactioncosts that cannot be feasibly metby individual smallholders (Landell-Millsand Porras 2002; Krey 2004; Chaco et al.2002). Moreover, where land rights are unclear,environmental service mechanismsmight compel powerful people to usurpotherwise marginal lands and evict poorland users (Grieg-Gran and Bann 2003).The Rewarding Upland Poor for EnvironmentalServices (RUPES) project wasestablished in 2001 to address possibilitiesfor environmental service mechanisms inAsia, with particular emphasis on potentialfor the upland poor to benefit from themechanisms. The project conducts actionresearch at pilot intervention sites acrossAsia to examine the provision of environmentalservices, decide who benefits andwho pays, and determine the institutionaland policy environment to enable fair andequitable distribution. An inclusive view istaken on payment, including rewards thatprovide upland farmers with enhancedland tenure security in exchange for followingland use agreements (RUPES 2004).Global environmentalgovernanceThe Rio Convention of 1992 marked asharp increase in the importance of globalenvironmental governance, includingseveral conventions and mechanisms thathave direct and indirect relevance for agroforestry.The United Nations Convention onBiological Diversity (UNCBD), the UNFC-CC and the United Nations Convention onCombating Desertification (UNCCD) arethe most important for agroforestry.The UNCCD has a Thematic ProgramNetwork (TPN) in Asia and Africa on agroforestryand soil conservation. The WorldAgroforestry Centre has provided technicalinput on agroforestry to the TPN forAfrica and is increasing its links with theTPN for Asia. The TPNs can also benefitfrom greater consideration of the links withenvironmental governance. In other words,while tree-based solutions have great technicalpotential for the problems of landdegradation, harnessing that potentialrequires institutional arrangements that appropriatelyshare benefits and costs, fosterlocal collective action in tree managementand provide individual farmers and farmcommunities with appropriate incentives.Comparative studies on agroforestry inthe drylands of South Asia and Africa canprovide valuable information. One successstory that may be replicated is the ‘Ngitili’system for farmer-managed natural regeneration(Barrow and Mlenge 2003).The UNCBD has adopted an expandedprogramme of work on forestry that hasmany connections with agroforestry, includingraising awareness of the problemsof invasive alien species. Recent Centreresearch in the Baringo area of Kenya isexploring how policies and institutionscan shape the benefits and costs associatedwith the alien invasive tree speciesProsopis juliflora. One approach to moreeffective management of P. juliflora wouldbe to organize collective harvesting andprocessing of charcoal made from itswood.The Centre has been engaged in the UN-FCCC for over 5 years. In 2001, the IPCCissued its third assessment report on climatechange, with a strong endorsement ofthe potential for agroforestry to contributeto increased carbon stocks in agriculturallands, while contributing to the welfareof smallholder farmers: “Agroforestry canboth sequester carbon and produce a rangeof economic, environmental and socioeconomicbenefits. For example, trees inagroforestry farms improve soil fertilitythrough control of erosion, maintenance ofsoil organic matter and physical properties,increased nitrogen, extraction of nutrientsfrom deep soil horizons and promotion ofmore closed nutrient cycling” (IPCC 2001).The Centre influences CDM policy processesin several ways. Firstly, we seek to providescientific data and information on therelations between agroforestry systems andgreenhouse gases, including carbon andnitrogen compounds. Secondly, we seekto understand the potential for agroforestryto buffer farmers against climate risks.Thirdly, we seek to evaluate how smallholderfarmers could be involved in carbon

92World Agroforestry into the Futuresequestration projects, and the implicationsof alternative mechanisms for exploitingthis potential. Finally, we provide relevantinformation to a variety of stakeholders atinternational, national and local levels.Conclusions and implicationsfor future research anddevelopmentEnvironmental governance shapes thecontext in which farmers make decisionsabout where and when to invest time andresources in planting and managing trees.Farmers are encouraged to protect existingvegetation and invest in new agroforestrysystems when they have secure rights tothe products generated by the trees, whenthere are certain markets for those products,and when they capture value from thepositive environmental services that theirtrees generate. Land and tree tenure, forestclassification, conservation policies, environmentalservice mechanisms and globalenvironmental agreements are componentsof environmental governance that affectthose incentives through various pathways.They are also policy levers that are used bygovernments to advance forest conservation,environmental protection, economicgrowth and other national objectives.Most developing countries have had regimesof environmental governance thatstressed forest conservation by centralagencies without due regard for the valueof the environmental services produced bythose forests, the performance of the regulatoryagencies or the negative impacts offorestry laws on farmers’ incentives to practiceagroforestry. Changes in environmentalgovernance are unfolding in many developingcountries, with some decentralizationof governance institutions and moreemphasis on the environmental effects ofland use outside of forests. In many cases,the result is a very uncertain and unevenpolicy terrain, particularly regarding therelatively new discipline of agroforestry.The review presented in this paper suggeststhat additional research is needed on thefollowing:• The landscape and watershed level effectsof different types of property rightsin farm areas and different configurationsof property rights in non-farmareas. Suyanto et al. (2005) has takenthis approach to fire management inSumatra; Swallow et al. (2001) outlinea similar approach for watershed management;and Ashley et al. (2005) dothe same thing for protected area landscapes.• Appropriate negotiation platforms formulti-functional landscapes. van Noodwijket al. (2001) have made majorcontributions to this with their work onnegotiation support systems.• The potential for environmental servicemechanisms that enhance the supply ofenvironmental services and the welfareof smallholder agroforesters in multifunctionallandscapes. The Centre is graduallyexpanding work on environmental servicemechanisms from specific locationsin Southeast Asia to key locations in LatinAmerica and South Asia.• The ways that global environmentalagreements can be modified or implementedto maximize the potential foragroforestry to synergize the objectivesof the agreements with that of reducingpoverty.AcknowledgementsThe authors wish to thank Regina Birnerand Tom Tomich for constructive commentson a previous version of this chapter.ReferencesAgrawal, A. and E. Ostrom 2001. Collectiveaction, property rights and devolution offorest and protected area management.In: Meinzen-Dick, R., A. Knox and M. diGregoria (eds), Collective Action, PropertyRights and Devolution of NaturalResource Management, Proceedings ofan International Conference 21–25 June1999, Puerto Azul, the Philippines. DeutscheStiftung für Internationale Entwicklung,Feldafing, Germany.Ashley, R., D. Russell and B. Swallow 2005.The policy terrain in protected area landscapes:challenges for agroforestry inintegrated landscape conservation. Biodiversityand Conservation (in press).ASB (2001). ASB Policy Brief 2: Putting Community-BasedForest Management on theMap. Alternatives to Slash and Burn Program,World Agroforestry Centre, Nairobi,Kenya.Barrow, E. and W. Mlenge 2003. Trees as keyto pastoral risk management in semi-aridlandscapes in Shinyanga, Tanzania andTurkana, Kenya. Paper presented at theInternational Conference on Rural Livelihoods,Forests and Biodiversity 19–23May, Bonn, Germany. Centre for InternationalForestry Research, Bogor, Indonesia.Bruce, J.W. 1989. Rapid Appraisal of Tree andLand Tenure. Community Forestry Note5. Food and Agriculture Organization ofthe United Nations. Available online athttp://www.fao.org/documents/show_cdr.asp?url_file=/DOCREP/006/T7540E/T7540E01.htmCernea, M.M. and K. Schmidt-Soltau 2003. Biodiversityconservation versus populationresettlement: risks to nature and risks topeople. Paper presented at the InternationalConference on Rural Livelihoods,Forests and Biodiversity 19–23 May,Bonn, Germany.

Chapter 10: Agroforestry and environmental governance93Chaco, O.J., G.R. Marshall and M. Milne 2002.Smallholder Agroforestry Projects: Potentialfor Carbon Sequestration and Poverty Alleviation.Food and Agriculture Organizationof the United Nations: Rome, Italy.Cronon, W. 1996. Uncommon Ground:Rethinking the Human Place in Nature.Norton, New York, USA.Fay, C. and G. Michon 2003. Redressing forestryhegemony: when a forestry regulatoryframework is best replaced by an agrarianone. Paper presented at the InternationalConference on Rural Livelihoods, Forestsand Biodiversity 19–23 May, Bonn,Germany.Fortmann, L. 1985. The tree tenure factor inagroforestry with particular reference toAfrica. Agroforestry Systems 2: 240–243.Fortmann, L. 1998. Why women’s propertyrights matter. Paper presented at the InternationalConference on Land Tenurein the Developing World 27–29 January,University of Capetown, South Africa.Fortmann, L. and J.W. Bruce (eds) 1988. WhoseTrees? Proprietary Dimensions of Forestry.Westview Press, Boulder, USA.Garrity, D., V.B. Amoroso, S. Koffa, D. Catacutan,G. Buenavista, P. Fay and W. Dar2002. Landcare on the poverty-protectioninterface in an Asian watershed. ConservationEcology 6(1): 12. Available onlineat www.ecologyandsociety.org/vol6/iss1/art12.Gladwin, C.H., J.S. Peterson, D. Phiri andR. Uttaro 2002. Agroforestry adoptiondecisions, structural adjustment and genderin Africa. In: Barrett, C.B., F. Placeand A.A. Aboud (eds) Natural ResourcesManagement in African Agriculture: Understandingand Improving Current Practices.CAB International, Wallingford, UK,pp. 115–129.Grieg-Gran, M. and C. Bann 2003. A closerlook at payments and markets for environmentalservices. In: Gutnam, P. (ed) FromGoodwill to Payments for EnvironmentalServices: A Survey of Financing Optionsfor Sustainable Natural Resource Managementin Developing Countries. WorldwideFund for Nature, Gland, Switzerland.IPCC 2001. Third Assessment Report on Mitigation.Inter-Governmental Panel on ClimateChange. United Nations EnvironmentalProgramme and World MeteorologicalOrganization, Geneva, Switzerland. Availableonline at http://www.grida.no/climate/ipcc_tar/ICRAF/AMAN/FPP 2003. In Search of Recognition.World Agroforestry Centre (ICRAF),Alliance of Indigenous Peoples of theArchipelago, (AMAN) and Forest People’sProgramme (FPP), Bogor, Indonesia.Krey, M. 2004. Transaction Costs of CDMProjects in India – An Empirical Survey.Report 238. Hamburg Institute of InternationalEconomics, Hamburg, Germany.Landell-Mills, N. and I.T. Porras 2002. SilverBullet or Fool’s Gold? A Global Review ofMarkets for Forest Environmental Servicesand their Impact on the Poor. InternationalInstitute for Environment and Development,London, UK.Leakey, R.R.B. 1996. Definition of agroforestryrevisited. Agroforestry Today 8(1): 5–7.Meinzen-Dick, R., A. Knox, F. Place and B.Swallow (eds) 2002. Innovation in NaturalResource Management: The Role ofProperty Rights and Collective Action inDeveloping Countries. Johns HopkinsUniversity Press, Baltimore, USA.Otsuka, K. and F. Place 2001. Land Tenure andNatural Resource Management: A ComparativeStudy of Agrarian Communities inAsia and Africa. Johns Hopkins UniversityPress:, Baltimore, USA.Place, F. 1995. The Role of Land Tenure in theAdoption of Agroforestry in Burundi,Uganda, Zambia and Malawi: A Summaryof Synthesis. Land Tenure Centre: Madison,USA.Place, F. and K. Otsuka 2002. Tenure and treemanagement in Uganda and Malawi. In:Meinzen-Dick, R., A. Knox, F. Place andB. Swallow (eds) Innovation in Natural ResourceManagement: The Role of PropertyRights and Collective Action in DevelopingCountries. Johns Hopkins UniversityPress, Baltimore, USA.Raintree, J. (ed) 1987. Land, Trees and Tenure.Land Tenure Centre, Madison, USA andInternational Centre for Research in Agroforestry,Nairobi, Kenya.RUPES 2004. Rewarding Upland Poor for EnvironmentalServices. Home page at: www.worldagroforestrycentre.org/sea/Networks/RUPESRussell, D., J. Ribot and F. Swartzendruber 2001.Central Africa and forest governance:counter-balancing the powers of publicand private sectors. CARPE Issue Brief No.11. Central African Regional Program forthe Environment, Washington, DC, USA.Schroeder, R.A. 1999. Shady Practices: Agroforestryand Gender Politics in The Gambia.University of California Press, Berkeley,USA.Schroth, G., C.A. Harvey and G. Vincent 2004.Complex agroforests: their structure,diversity, and potential role in landscapeconservation. In: Schroth, G., G.A.B. daFonseca, C.A. Harvey, C. Gascon, H.L.Vasconcelos and A.M.N. Izac (eds) Agroforestryand Biodiversity Conservation inTropical Landscapes. Island Press, Washington,DC, USA,Suyanto S., T.P. Tomich and K. Otsuka 2001.Agroforestry management in Sumatra. In:Otsuka, K. and F. Place (eds) Land Tenureand Natural Resource Management: AComparative Study of Agrarian Communitiesin Asia and Africa. Johns HopkinsUniversity Press, Baltimore, USA andLondon, UK.Suyanto, S., R.P. Permana, N. Khususiyah andL. Joshi 2005. Land tenure, agroforestry

94World Agroforestry into the Futureadoption, and reduction of fire hazard ina forest zone: a case study from Lampung,Sumatra, Indonesia. Agroforestry Systems65(1): 1–11.Swallow, B.M., D.P. Garrity and M. van Noordwijk2001. The effects of scales, flows andfilters on property rights and collectiveaction in catchment management. WaterPolicy 3(6): 449–455.Tomich, T.P., A.M. Fagi, H. de Foresta, G. Michon,D. Murdiyarso, F. Stolle and M. vanNoordwijk 1998. Indonesia’s fires: smokeas a problem, smoke as a symptom. AgroforestryToday 10(1): 4–7.Tomich, T.P., H. de Foresta, R. Dennis, Q.M. Ketterings,D. Murdiyarso, C. Palm, F. Stolle,S. Suyanto and M. van Noordwijk 2002.Carbon offsets for conservation and developmentin Indonesia? American Journal ofAlternative Agriculture 17(2): 125–137.Tomich, T.P., D. Thomas and M. Van Noordwijk2004. Environmental services and land usechange in Southeast Asia: from recognitionto regulation or reward. Agriculture, Ecosystemsand Environment 104(1): 229–244.Unruh, R.D. 2002. Land dispute resolution inMozambique: evidence and institutions ofagroforestry adoption. In: Meinzen-Dick,R., A. Knox, F. Place and B. Swallow (eds)Innovation in Natural Resource Management:The Role of Property Rights andCollective Action in Developing Countries.Johns Hopkins University Press,Baltimore, USA, pp. 166–185.van Noordwijk, M., T.P. Tomich, H. de Forestaand G. Michon 1997. To segregate orto integrate? The question of balancebetween production and biodiversity conservationin complex agroforestry systems.Agroforestry Today 9 (1): 6–9.van Noordwijk, M., T.P. Tomich and B. Verbist2001. Negotiation support models forintegrated natural resource managementin tropical forest margins. ConservationEcology 5(2): available online at www.ecologyandsociety.org/vol5/iss2/21Venema, D.H. and M. Cisse 2004. Seeing theLight: Adapting to Climate Change withDecentralized Renewable Energy in DevelopingCountries. International Institutefor Sustainable Development (IISD) andClimate Change Knowledge Network,Winnipeg, Canada.WRI 2003. World Resources 2002–2004: Decisionsfor the Earth: Balance, Voice, andPower. United Nations DevelopmentProgramme, United Nations EnvironmentProgramme, World Bank, World ResourcesInstitute, Washington, DC, USA.WWF 1997. Ecotourism in the Forests/Grasslandsof Royal Chitwan National Park,Nepal. Worldwide Fund for Nature: availableonline at http://www.worldwildlife.org/bsp/bcn/projects/chitwan97.htm

Keywords:Homegardens, buffer zone, conservation planning,tropics, Africa, Asia, alien invasive speciesChapter 11The potential for agroforestry to contributeto the conservation and enhancement oflandscape biodiversityBrent Swallow and Jean-Marc Boffa, World Agroforestry Centre and Sara J. Scherr, Forest TrendsAbstractAgroforestry is increasingly being acknowledged as an integrated land use that can directly enhanceagrobiodiversity and contribute to the conservation of landscape biodiversity, while at the same timeincrease, diversify and sustain rural incomes. There are valid concerns, however, that the biodiversitybenefits of agroforestry may be misunderstood and the risks to biodiversity understated. This chaptertherefore reviews some of the growing literature on agroforestry and biodiversity in order to clarify keyrelationships, including factors and processes that amplify or limit the contributions of agroforestry tobiodiversity conservation. Four propositions are presented, with reference to evidence for the propositionsand caveats to them. We conclude that agroforestry generally produces higher biodiversitybenefits than both annual and perennial monoculture crop production, and that agroforestry is of thegreatest benefit to biodiversity when it is a component of an integrated approach to land use. Importantknowledge gaps remain, however, regarding the ways in which tree domestication and agroforestrypromotion can be designed to stimulate new agroforestry systems that have greater positive impactson wild biodiversity.IntroductionAgroforestry is increasingly being identified as an integratedland use that can directly enhance plant diversitywhile reducing habitat loss and fragmentation (Nobleand Dirzo 1997). There are major concerns, however,that the deforestation benefits of agroforestry have beenoverstated (Angelsen and Kaimowitz 2004) and thatthe risks associated with agroforestry have not beenadequately acknowledged. It is therefore more importantthan ever that both the scientific and developmentcommunities develop a more accurate and subtle understandingof the multiple links between biodiversity andagroforestry. This chapter reviews evidence that linksagroforestry with biodiversity in an attempt to clarifykey relationships. It also examines the factors and processesthat may amplify and limit the contributions ofagroforestry to biodiversity conservation. It is organizedas follows: the first substantive section presents importantorganizing concepts; the second section reviewsthe available evidence for and against four propositionsabout the relationship between biodiversity and agroforestry;and the final section discusses a number ofissues for follow-up research.Organizing conceptsThe United Nations Convention on Biological Diversity(UNCBD) defines ‘biodiversity’ as “…the variability

96World Agroforestry into the Futureamong living organisms from all sources,including, inter alia, terrestrial, marine andother aquatic ecosystems, and the ecologicalcomplexes of which they are part; thisincludes diversity within species, betweenspecies, and of ecosystems.” Spatial andecological scales are therefore fundamentalconcepts in biodiversity studies. TheUNCBD further defines agrobiodiversity asbiodiversity that is important for agriculturalproduction, including crop and livestockgenetic diversity, wild biodiversity closelyassociated with domesticated species, andother wild biodiversity sharing the resources.‘Wild biodiversity’ is biodiversity that hasnot been domesticated, while ‘domestication’is the dynamic process of how humansselect, improve, manage, propagate andintegrate trees or other plants into land usesystems. While ICRAF and its partners haveconducted a great deal of work on belowgroundbiodiversity (e.g. van Noordwijk etal. 2004), we concentrate here on abovegroundbiodiversity at the landscape scale,explicitly focusing on the links between theplanting and management of trees by farmersand biodiversity in the landscape.Several definitions of the term ‘agroforestry’are used in science and practice. Leakey’s(1996) definition is used most frequently:“a dynamic, ecologically based, naturalresource management system that, throughthe integration of trees on farms and in thelandscape, diversifies and sustains productionfor increased social, economic andecological benefits.” Three aspects of thisdefinition are important for the biodiversityvalue of agroforestry. Firstly, agroforestryinvolves the deliberate integration of treeswith farms and landscapes, which mayhave direct and indirect effects on farm andlandscape biodiversity. Secondly, there aretrade-offs and complementarities betweenthe social, economic, ecological and biodiversitybenefits of agroforestry comparedto other land use systems; indeed, thequantification of trade-offs has been at theheart of the research agenda of the Alternativesto Slash and Burn (ASB) Programmecoordinated by ICRAF (Tomich et al. 2001).Thirdly, while some agroforestry practicesin certain circumstances contribute greatlyto diversification and sustainability, thereare other circumstances where it contributesvery little.Propositions aboutrelationships betweenagroforestry and biodiversityA number of recently completed reviewpapers suggest ways in which agroforestrycontributes to the conservation and protectionof biodiversity, including that of bothwild species and species more directlyrelated to agricultural production (Boffa1999; Buck et al. 2004; Cunningham et al.2002; McNeely 2004; McNeely and Scherr2003; Schroth et al. 2004a; van Noordwijket al. 1997). These and other studies suggestfour key relationships between agroforestryand landscape biodiversity.1. Agroforestry farmers and systemsas promoters of plant diversityProposition: While modifying naturalvegetation for their productive use, farmersdevelop and maintain agroforestrysystems that make substantial contributionsto biodiversity in multi-functionallandscapes.The proposition that agroforestry will resultin ‘substantial contributions’ to biodiversityis supported by a good deal of evidenceregarding the diversity of tree and vascularplant species across a variety of landscapes,including those containing agroforestrysystems. There are important caveatsto the proposition, however: i) there arelarge differences in the biodiversity valueof different agroforestry systems; ii) some ofthe more diverse agroforestry systems maybecome less diverse under high levels ofpopulation pressure; and iii) the commoditiesthat underpin many of the most diverseagroforestry systems are subject to fluctuationsand declines in profitability whenadopted on a large scale.The ASB programme has evaluated the biodiversityassociated with a range of typicalland use types, including agroforestry, thatare found at the frontiers of tropical forestsin Southeast Asia, the Congo Basin, and theAmazon Basin. Methods used and resultsgenerated by this comprehensive set ofstudies are available on the ASB web page(www.asb.cgiar.org). Summary results arealso presented in Tomich et al. (1998) andTomich et al. (2001). In general the resultsshow that multistrata agroforestry systemscontain an intermediate level of plant biodiversitythat lies between primary forestsand monocrop perennials or field crops.For example, Murdiyarso et al. (2002) comparedthe number of plant species found indifferent types of land use in the Jambi areaof central Sumatra. They found that continuouslycultivated cassava had 15 speciesper 1.5-hectare plot, oil palm plantationshad 25 species per plot, rubber agroforestshad 90 species per plot, while primary forestshad 120 species per plot.Gillison et al. (2004) found that complexagroforestry systems and shade-growncoffee both had much higher levels ofbiodiversity than simple sun-grown coffee,although all coffee systems had lower biodiversitythan primary or secondary forests.

Chapter 11: The potential for agroforestry97Similarly high levels of tree diversity arealso reported for complex cocoa systemsfound in West Africa and Central America(Schroth et al. 2004b) and the intensehomegarden systems found in many partsof Africa and Asia (Khan and Arunachalam2003; Michon and de Foresta 1995).However, recent data from the Chaggahomegardens in Tanzania indicate that treepopulations in established gardens may becomeless dense and more fragmented overtime if population pressures rise to veryhigh levels (Misana et al. 2003; Soini 2005).An alternative approach to agroforestry/forest management that has proved particularlyeffective in parts of East Africaand the West African Sahel, is describedby one analyst as farmer-managed naturalregeneration (Chris Reij, seminar at ICRAF,Nairobi, September 2004). The agroforestryparklands of the Sahel are one example– for several generations, farmers acrossthe Sahel have deliberately selected andprotected valuable indigenous trees locatedin their agricultural fields (Boffa 1999).The ‘Ngitili’ system practised in westernTanzania is another example. The Sukumapeople – as individuals and groups – havetraditionally set aside parcels of land andmanaged them as biodiversity reserves andfall-back resources. After years of neglect,this system has been revived in large areasof western Tanzania. Additional value hasbeen added to many ‘Ngitili’ exclosuresthrough the planting and management ofvaluable timber and fruit trees (Barrow andMlenge 2003).Assessing the value of farmer tree managementto biodiversity is challenging.On-farm surveys in Cameroon, Kenya andUganda show that the diversity of tree specieshosted on African farms is greater thanoriginally thought (Kindt 2002). However,in these agroforestry systems, althoughthere may be high species richness it isoften accompanied by the infrequent occurrenceof many species. For example,while 47 percent of species recorded inUganda’s Mabira Forest were found onsurrounding farms, more than half of theidentified species numbered 10 individualsor less (Boffa, unpublished data), whichmay not be sufficient to sustain genetic diversityin the long term. Equally importantis the extent to which agroforestry systemsspecifically contribute to the conservationof rare or threatened forest species. Datafrom these three countries indicate that fewvulnerable or threatened species have actuallybeen observed in agroforestry systems.More research is needed on the importantfunctions and roles that tree diversity playsin landscapes in terms of conserving lesserknown aspects of biodiversity, providingother environmental services and benefitinglivelihoods.A crucial question is how agroforestrysystems with increased biodiversity valuecan be stimulated or enhanced in newenvironments. Agroforestry systems havethe potential to evolve through successiontoward mature, productive systems to forma mosaic of patches on a landscape whileproducing marketable tree products for improvedlivelihoods. Leakey (2004) proposesthat domestication of valuable indigenoustrees is a key starting point. The propositionis that farmers who recognize and are ableto capitalize on the value of indigenoustrees will be impelled to plant and protecttrees of various types. ICRAF’s agenda ofresearch on domestication, seed productionand marketing of indigenous fruit andmedicinal trees is largely based on thisproposition. One of the challenges is to integratedomestication work with a broaderconservation framework.2. Agroforestry and pressures onforests and protected conservationareasProposition: The increased uptake ofagroforestry in multi-functional landscapescan reduce pressure on forestsand protected conservation areas.This proposition is not supported by a largebase of empirical evidence, but nonethelesshas become the basis for includingagroforestry in many integrated conservationand development projects. For example,one of the global champions forprimate research and conservation, JaneGoodall, now supports agroforestry developmentas a way of protecting the remainingchimpanzee populations in the Gombenational park in Tanzania (http://www.janegoodall.ca/inst/inst_tacare_hist.html).Thereare four main caveats to this proposition:i) agroforestry will only result in reducedpressures on a protected area if the mainpressure on that area is farmers’ collectionof tree products; ii) agroforestry has the potentialto increase pressure on forests andconservation areas if it results in increasedclearance of primary forest for agroforestry;iii) the potential impact of agroforestry onprotected areas depends upon the policyand institutional context affecting treemanagement and protected area use; andiv) it is difficult to separate the effects ofagroforestry from other elements of bufferzone management that successfully reducepressures on protected areas.The proposition that agroforestry canreduce pressure on conservation areasis mainly based on evidence of the productivityof agroforestry systems comparedto more extensive systems of land

98World Agroforestry into the Futuremanagement. For example, Ramadhani etal. (2002) found that 5-year-old woodlotsof Acacia crassicarpa in the Tabora districtof Tanzania produced five times as muchwood as mature ‘miombo’ woodlands.Simple calculations show that if all thewood needed for tobacco drying camefrom woodlots instead of the ‘miombo’,then 8 675 hectares of woodland would beconserved each year in the Tabora district.Govere (2002) attempted to test this substitutionhypothesis for the example of improvedfallows in eastern Zambia. His resultsare mixed: in one village the adoptersof improved fallows gathered less woodthan non-adopters; in another village adoptersand non-adopters gathered roughly thesame amount of wood.A study by Garrity et al. (2002) around theMount Kitanglad Range National Park inMindañao, the Philippines, provides supportfor a link between agroforestry and reducedpressure on protected areas. Farmersaround this area of high biodiversity wereeducated about the use of natural vegetativestrips to stabilize hillside farming areas,and improved germplasm and nurserytechniques to enhance on-farm productionof fruit and timber. The key institutionalinnovation was Landcare – farmer-ledknowledge-sharing organizations inspiredby the Landcare movement in Australia.After a number of years, this combinationof technical and institutional interventionsproduced positive impacts in terms of increasedmaize yields, greater density offruit and timber trees, reduced runoff anderosion, enhanced environmental awareness,reduced encroachments into the park,and restored stream corridor vegetation.By 2002 there were more than 800 householdsin Mindañao that belonged to villageLandcare chapters around the parkboundary.Another study of the buffer zone of theKerinci Seblat National Park, Indonesiahighlights the relationship between farmdiversification and reliance on adjacent nationalpark resources (Murniati et al. 2001).Comparing a sample of rice-only farms,mixed garden farms and a combination ofboth, the authors found that farms practisingboth rice growing and mixed gardeninghad 80 percent lower dependency on parkresources. Factors associated with a higherpropensity to extract from protected forestresources were low farm income and lowsupply of on-farm tree-based products,suggesting that agroforestry systems wereparticularly relevant in the buffer zones.ICRAF research around the Mabira ForestReserve in Uganda suggests that largerscale economic forces and forest policycan have greater impact on protected areasthan agroforestry and other developmentinterventions undertaken around them.While resource extraction by adjacentcommunities increased with proximity tothe forest, agroforesry in the buffer zonecould not have any significant impact onthe quantitatively far more significant pressuresoriginating from outside the bufferzone, particularly from fuelwood marketsfor sugar and tea processing, and for brickand charcoal making (Mrema et al. 2001a;2001b; 2001c; 2001d).Angelsen and Kaimowitz (2004) argue thatthe conservation benefits of agroforestryhave often been overstated, particularly inplaces where the forest frontier is still opento settlement and harvesting. Angelsen andKaimowitz (2004) and Tomich et al. (2001)point out there are likely to be trade-offsassociated with profitable agroforestry: onone hand, there will be pressure to convertprimary forest to profitable alternative landuses; on the other hand, degradation ofagroforestry systems may lead to conversionto less desirable land uses. A classiccase of these trade-offs is cocoa. Conversionof primary forest to cocoa productionhas been a major source of biodiversityloss in many parts of the humid tropics.However, compared to sun-grown cocoaor competing annual crops, shade-growncocoa agroforests retain much higher levelsof biodiversity (Donald 2004).3. Agroforestry and habitat for wildspeciesProposition: Agroforestry can createhabitat for wild species in landscapematrices surrounding forest conservationareas.The integration of trees into multiple-uselandscape matrices can contribute to wildbiodiversity through the maintenance oflandscape connectivity, heterogeneity andcomplexity of vegetation structure, integrityof aquatic systems, and cleaner water. Treescan contribute nesting sites, protectivecover against predators, access to breedingterritory, access to food sources in allseasons, and encourage beneficial speciessuch as pollinators. Evidence of the natureof these relationships has been generatedthrough a fairly large number of field studies,most of which have focused on birds.One caveat is that there have been limitedstudies to date on how the spatial configurationof trees on farms and in landscapesaffects the conservation of different types ofbiodiversity.Buck et al. (2004) reviewed 12 studies thatfound agroforestry systems to provide habitatfor diverse populations of birds, withthe greatest amount of evidence pointingtowards the habitat value of shade-grown

Chapter 11: The potential for agroforestry99coffee and cocoa systems in Southeast Asiaand Central America. However, there arealso contrasting results: Soini (2004) foundlow levels of bird diversity in the multistrataChagga homegardens of Kilimanjaro,Tanzania. Soini postulates that the veryhigh levels of human population in thoseareas have created an inhospitable habitatfor most bird species.Naidoo (2004) presents a novel analysis ofthe relationship between forest types andbird types in and around the Mabira forestin Uganda. He analysed the diversity ofsongbirds along transects across differenttypes of landscapes, from intact primaryforest, to regenerating secondary forests andagricultural fields. Songbirds were classifiedas forest specialists, forest generalists, forestvisitors and open habitat species. He foundroughly similar numbers of total songbirdspecies in each of the three land-use types,but marked differences in the percentagesof different species groups. Forest specialistswere not found in the agricultural area;open habitat species were not found in theintact forest. Statistical models of the habitat–speciesrelationship showed that treedensity and distance to intact forest had thegreatest impacts on number of forest species.Model results indicate that greater treedensity in agricultural fields could result ina sizeable expansion in the habitat of forestspecialists within the forest and forest generalistsin the forest margin.Agroforestry can enhance connectivity andlandscape heterogeneity in multi-functionalconservation landscapes. Zomer etal. (2001) found that an agroforestry systeminvolving Alnus nepalensis and cardamomcontributed to the integrity of riparian corridorsfor wildlife conservation around theMakalu Barun National Park and ConservationArea of eastern Nepal.Griffith (2000) suggests a different ecologicalmechanism by which agroforestry cancontribute to biodiversity – by providinga low risk refuge in the case of fire. Heassessed bird biodiversity in two agroforestryfarms in the buffer zone of the MayaBiosphere Reserve in Guatemala in orderto determine whether those farms hadserved as biodiversity refuges during thefires of 1998 that burned eight percent ofthe reserve. He found high numbers of birdspecies, including forest specialists and forestgeneralists – birds that are not usuallyfound in agroforestry areas.4. Agroforestry and the threats ofinvasive alien speciesProposition: Agroforestry developmentcan be implemented in a way thatreduces the risk of alien invasive speciesto acceptable levels, if adequateprecautions are taken.In the introduction to this chapter wenoted that there are major concerns in theconservation community about the potentialthreat that farmer planting of treesmay pose to biodiversity. For example, theUNCBD Thematic Programme of Workstates: “Tree plantations and agroforestry areimportant sources of biological invasions…Of species used for agroforestry aroundseven percent are said to be weeds undersome conditions, but around one percentare weedy in more than 50 percent of theirrecorded occurrences.”Evidence from across the world indicatesthat agroforestry projects have contributedto the ecological problems associated withalien invasive species. News of impending‘fuelwood crises’ a generation ago ledto the creation of a large number of newagroforestry projects across the developingworld in the late 1970s and early 1980s.While many of these projects undoubtedlycontributed to increased energy supplies,they have also had negative consequencesfor welfare, biodiversity and water availability.Better design of the current generationof agroforestry projects should help tominimize negative impacts in the future. Forexample, ICRAF has adopted a policy thatfocuses on reducing the risk of introducinginvasive alien species as part of newagroforestry research and developmentprogrammes. We are also conducting researchon effective management of selectedinvasive alien species. For example, ongoingresearch on Prosopis juliflora inthe Baringo area of Kenya indicates thepotential benefits and limitations of effectivemanagement through sustained use.Challenges for the futureThe overall conclusion that emerges fromthis review is that agroforestry generallyproduces biodiversity benefits that are intermediatebetween monocrop agricultureand primary forests. The overall contributionof agroforestry to biodiversity conservationdepends, therefore, on the type of land usethat it replaces and on the attributes of thespecific agroforestry system. The effectivenessof agroforestry in biodiversity conservationdepends on the design of the system and thenature of the biodiversity to be conserved.Agroforestry is not a stand-alone approachto conservation. Rather, it needs to be seenas an element of conservation strategies,which also include policy and institutionalchanges, and spatial configurations that emphasizemaintenance of natural habitats.Additional research, including appropriatemeasurement, modelling and

100World Agroforestry into the Futureexperimentation, is needed, containedwithin the following recommendations:• Broaden the agroecological focus ofagroforestry and biodiversity studies toinclude more drylands and annual cropbasedsystems.• Identify the key features of agroforestrysystems – species composition, configuration,management, landscape position– that are most critical to supportingbiodiversity in the landscape and inmultiuse areas around protected areas.• Evaluate the conditions under whichmarket-led domestication and on-farmhusbandry of valuable indigenous treescan stimulate a sequence of increasedtree planting, more intensive land use,and less pressure on forest and land resources.• Assess the landscape-level effects ofnew agroforestry systems, such as theimproved fallows and rotational woodlotspromoted in southern Africa.• Give higher priority to the challengesof alien invasive species, with specialemphasis on the development ofmanagement plans for species that havebeen associated with agroforestry.• Expand the use of agroforestry systemsin degraded lands to help restore theproductivity and biodiversity of marginallands.• Fully explore the refuge value of agroforestrysystems, such as those studiedby Griffith (2000).• Conduct more research into theimportant functions and roles thattree diversity plays in landscapes forconserving lesser-known aspects ofbiodiversity, providing other environmentalservices, and benefitinglivelihoods.AcknowledgementThe authors gratefully acknowledge commentsfrom Louise Buck, MohammedBakarr and Tom Tomich on a previousdraft of this chapter.ReferencesAngelsen, A. and D. Kaimowitz 2004. Is agroforestrylikely to reduce deforestation? Chapter5, pp. 87–106, in: Schroth, G., G.A.B.da Fonseca, C.A. Harvey, C. Gascon, H.L.Vasconcelos and A.M.N. Izac (eds) Agroforestryand Biodiversity Conservation inTropical Landscapes. Island Press, Washington,DC, USA.Barrow, E., and W. Mlenge 2003. Trees as key topastoralist risk management in semi-aridlandscapes in Shinyanga, Tanzania andTurkana, Kenya. Paper presented at theInternational Conference on Rural Livelihoods,Forests and Biodiversity 19–23May 2003, Bonn, Germany.Boffa, J-M. 1999. Agroforestry parklands in sub-Saharan Africa. FAO Conservation Guide34. Food and Agriculture Organization ofthe United Nations, Rome, Italy. (http://www.fao.org/DOCREP/005/X3940E/X3940E00.htm)Buck, L.E., T.A. Gavin, D.R. Lee, N.T. Uphoff, D.C.Behr, L.E. Drinkwater, W.D. Hively and F.R.Werner 2004. Ecoagriculture: A Review andAssessment of its Scientific Foundations.Cornell University, Ithaca, USA.Cunningham, A.B., S.J. Scherr and J.A. Mc-Neely 2002. Matrix Matters: BiodiversityResearch for Rural Landscape Mosaics.Center for International Forestry Researchand World Agroforestry Centre, Bogor,Indonesia.Donald, P.F. 2004. Biodiversity impacts of someagricultural commodity production systems.Conservation Biology 18(1): 17–37.Garrity, D., V.B. Amoroso, S. Koffa, D. Catacutan,G. Buenavista, P. Fay and W. Dar2002. Landcare on the poverty–protectioninterface in an Asian watershed. ConservationEcology 6(1): 12.Gillison, A.N., N. Liswanti, S. Budidarsono, M.van Noordwijk, and T.P. Tomich 2004.Impact of cropping methods on biodiversityin coffee agroecosystems in Sumatra,Indonesia. Ecology and Society 9(2): 7.Govere, I. 2002. Improved tree fallow andnatural miombo woodland use in EasternZambia: The potential of agroforestry inthe conservation of indigenous forests.MSc Thesis, Department of AgriculturalEconomics and Extension, University ofZimbabwe, Harare, Zimbabwe.Griffith, D.M. 2000. Agroforestry: A refuge fortropical biodiversity after fire. ConservationBiology 14(1): 325–326.Khan, M.L. and A. Arunachalam 2003. Traditionalagroforestry as a viable choice toconserve agro-biodiversity in the northeastIndia. Pp. 95–105 in Pathak, P.S. and R.Newaj (eds) Agroforestry: Potentials andOpportunities. Agrobios (India) and IndianSociety of Agroforestry, Jodhpur, India.Kindt, R. 2002. Methodology for tree speciesdiversification planning for African agroecosystems.PhD thesis, University ofGhent, Ghent, Belgium.Leakey R.R.B. 1996. Definition of agroforestryrevisited. Agroforestry Today 8(1): 5–7.Leakey, R. 2004. Domestication and marketingof novel crops for ecoagriculture. Paperpresented at the International EcoagricultureConference and Practitioner’s Fair 27September–1 October 2004, World AgroforestryCentre, Nairobi, Kenya.

Chapter 11: The potential for agroforestry101McNeely, J.A. 2004. Nature vs. nurture: Managingrelationships between forests, agroforestryand wild biodiversity. AgroforestrySystems 61: 155–165.McNeely, J.A. and S.J. Scherr 2003. Ecoagriculture:Strategies to Feed the Worldand Save Wild Biodiversity. Island Press,Washington, DC, USA.Michon, G. and H. de Foresta 1995. The Indonesianagro-forest model. Forest resourcemanagement and biodiversity conservation.Pp. 90–106 in: Halladay, P. and D.A.Gilmour (eds) Conserving BiodiversityOutside Protected Areas. The Role OfTraditional Agro-ecosystems. WorldConservation Union, Forest ConservationProgramme, Andalucía, Spain.Misana, S.B., A.E. Majule and H.V. Lyaruu 2003.Linkages between changes in land use,biodiversity and land degradation on theslopes of Mount Kilimanjaro, Tanzania.LUCID Working Paper Number 38. InternationalLivestock Research Institute,Nairobi, Kenya.Mrema, M., D. Wafula and J. Sekatuba 2001a.Livelihood strategies and the use of treeand forest products in the Mabira Forestbuffer zone. World Agroforestry Centre(ICRAF), Kampala, Uganda.Mrema, M., D. Wafula and J. Sekatuba. 2001b.The use of tree and forest products by localindustries: The case of sugar and teaestates in the Mabira buffer zone of Mukono.World Agroforestry Centre (ICRAF),Kampala, Uganda.Mrema, M., D. Wafula and J. Sekatuba. 2001c.The use of tree and forest products bylocal institutions and households in theurban centres surrounding the Mabirabuffer zone. World Agroforestry Centre(ICRAF), Kampala, Uganda.Mrema, M., D. Wafula and J. Sekatuba. 2001d.Marketing of tree and forest products inthe buffer zone of Mabira Forest and surroundingtowns. World Agroforestry Centre(ICRAF), Kampala, Uganda.Murdiyarso, D., M. van Noordwijk, U.R. Wasrin,T.P. Tomich and A.N. Gillison 2002.Environmental benefits and sustainableland-use options in the Jambi transect,Sumatra, Indonesia. Journal of VegetationScience 13: 429–438.Murniati, D.P. Garrity and A.N. Gintings 2001.The contribution of agroforestry systems toreducing farmers’ dependence on the resourcesof adjacent national parks: A casestudy from Sumatra. Agroforestry Systems52: 171–184.Naidoo, R. 2004. Species richness and communitycomposition of songbirds in a tropicalforest-agriculture landscape. Animal Conservation7: 93–105.Noble, I.R. and R. Dirzo 1997. Forests as human-dominatedecosystems. Science 277:522–525.Ramadhani, T., R. Otsyina and S. Franzel 2002.Improving household incomes and reducingdeforestation using rotational woodlotsin Tabora district, Tanzania. Agriculture,Ecosystems and Environment 89: 229–239.Schroth, G., G.A.B. da Fonseca, C.A. Harvey, C.Gascon, H.L. Vasconcelos and A.M.N. Izac(eds) 2004a. Agroforestry and BiodiversityConservation in Tropical Landscapes.Island Press, Washington, DC, USA.Schroth, G., C.A. Harvey and G. Vincent 2004b.Complex agroforests: Their structure,diversity and potential role in landscapeconservation. Chapter 10, Pp. 227–260in: Schroth, G., G.A.B. da Fonseca, C.A.Harvey, C. Gascon, H.L. Vasconcelos andA.M.N. Izac (eds) Agroforestry and BiodiversityConservation in Tropical Landscapes.Island Press, Washington, DC, USA.Soini, E. 2004. Birds in three habitats on thesouthern slopes of Mt. Kilimanjaro, Tanzania.Unpublished manuscript.Soini, E. 2005. Dynamics of livelihood strategiesin the changing landscapes of the southernslopes of Mt. Kilimanjaro, Tanzania.Agricultural Systems 85(3): 306–323.Tomich, T.P., M. van Noordwijk, S.A. Vosti andJ. Witcover 1998. Agricultural Economics19 (1998): 159–174.Tomich, T.P., M. van Noordwijk, S. Budidarsono,A. Gillison, T. Kusumanto, M. Murdiyarso,F. Stolle and A.M. Fagi 2001. Agriculturalintensification, deforestation, and the environment:Assessing tradeoffs in Sumatra,Indonesia. In: Lee, D.R. and C.B. Barrett(eds) Tradeoffs or Synergies? AgriculturalIntensification, Economic Developmentand the Environment. CAB International,Wallingford, UK.van Noordwijk, M., G. Cadish and C.K. Ong2004. Below-ground Interactions in TropicalAgroecosystems. CAB International,Wallingford, UK.van Noordwijk, M., T.P. Tomich, H. de Forestaand G. Michon 1997. To segregate orto integrate? The question of balancebetween production and biodiversity conservationin complex agroforestry systems.Agroforestry Today 9(1): 6–9.Zomer, R.J., S.L. Ustin and C.C. Carpenter 2001.Land cover change along tropical andsub-tropical riparian corridors within theMakalu Barun National Park and ConservationArea, Nepal. Mountain Researchand Development 21(2): 175–183.

Keywords:Agroforestry, buffering water flow, coffee, criteria and indicators,litter layer, macroporosity, protective garden, soil bulk density,soil structure, watershed functionsChapter 12Watershed functions in productive agriculturallandscapes with treesMeine van Noordwijk and Farida, World Agroforestry Centre, Indonesia; Pornwilai Saipothong, World AgroforestryCentre, Thailand; Fahmudin Agus, Centre for Soil and Agroclimate Research, Indonesia; KurniatunHairiah and Didik Suprayogo, Brawijaya University, Indonesia and Bruno Verbist, World Agroforestry Centre,Indonesia and Katholieke Universiteit Leuven, BelgiumAbstractWatershed functions are often discussed in terms of deforestation and reforestation, but require a morecareful diagnosis of problems and solutions. Criteria and indicators that are based on the quantity, timingand quality of river flows are influenced by a combination of effects, including the green and browncover provided by plant canopies and surface litter layers, the soil surface properties and soil structure,and the landscape-level drainage network. Opportunities for agroforestry and other forms of conservationfarming to maintain and restore watershed functions are dependent on the relatively rapid optionsfor restoring green and brown cover, the asymmetric (rapid degradation, slow recovery) dynamics ofsoil structure and on modification of landscape-level drainage. Data for the watersheds of Mae Chaemin northern Thailand and Way Besai in Lampung, Indonesia, indicate that land-cover change has arelatively small effect on low river flow. We focus here on the changes in soil structure as the ‘slow variable’that tends to dominate the long-term opportunities for keeping watersheds productive as well assuppliers of quality water at the desired time.IntroductionWatershed functions are nearly everybody’s concern.Clearing natural forests to grow crops or build roadscan reduce the amount of water that enters the soil andincrease overland mudstream flows. Human habitationand industry can lead to streams becoming pollutedwhile increasing the demand for clean water. Buildingon floodplains and wetlands can reduce water storageand buffer capacity and put the new developments atrisk of flooding. New fast-growing crops and plantedtrees can use more water than existing vegetation.And governments can claim control of waterways andimpose national solutions on them that do not takeaccount of the local effects.The end result of all these changes is that there are‘problems with watershed functions’ that affect peopleone way or another. These problems will generally beattributed to deforestation, and reforestation is the defaultsolution in public debate. The standard approachto ‘rehabilitation of watersheds’ is to plant trees, usuallyunder the control of foresters, in the hope of recreatingthe benign conditions of a natural forest. Naturalor planted forests, however, provide livelihood optionsonly at low population densities, so reforestation cannotreally solve current pressures on the land. Furthermore,tree planting in relatively dry areas may actually increasethe problem: fast-growing trees with high water use willreduce dry-season flows of streams and rivers.

104World Agroforestry into the FutureAgroforestry can make solid contributionsto resolving the apparent trade-off betweenmaintenance of watershed functions andproductive agriculture, if it addresses theissues in a way that links patch, field, farmand landscape scales.In this brief description of current approachesto agroforestry solutions to watershedproblems, we will consider the followingfour basic steps, and discuss the conceptsand tools required for each:1. Diagnosis of problems at watershedscale.2. Comparing land-use options on the basisof buffering functions.3. Modelling physical degradation andrehabilitation processes in the analysisof trade-offs between profitability andwatershed functions.4. Negotiations between stakeholders ofsolutions on the basis of trade-offs.Diagnosis of problems atwatershed scaleBecause there are many potential solutionsto problems with watershed function, weneed to be clear and specific about whatthe problem is, and this requires a commonperception (criteria and indicators; Figure 1).A list of criteria for the contribution of watershedsto water quantity (the capacity totransmit water, buffer peak flows and releasewater gradually), water quality (reduce sedimentloads and other pollutants and maintainaquatic biodiversity) and integrity of theland surface (control landslides and reduceloss of fertile topsoil through erosion), needsto be combined with criteria that relate tobiodiversity conservation and to the socialand economic welfare of the people livingin watershed areas.The relationship between full (as providedby a forest) and partial (agroforestry) treecover and hydrological functions in termsof the five watershed functions listed inFigure 1 involves different time scales andtrade-offs between total water yield and thedegree of buffering of peak river flows relativeto peak rainfall events. The role of landuse can be analysed in terms of changes inevapotranspiration, linked to the presenceof trees; infiltration, linked to conditions ofthe soil; and the rate of drainage linked tothe drain network in the landscape.van Noordwijk et al. (2003) completed adetailed analysis of both the 4000 km 2 MaeChaem catchment in northern Thailand(mean annual rainfall 1500 mm, populationdensity 20 km –2 ; mean annual river flow20–30 m 3 s –1 ) and the 500 km 2 Way Besaicatchment in Lampung, Indonesia (meanannual rainfall 2500 mm, population density160 km –2 ; mean annual river flow 15–20m 3 s –1 ). Daily rainfall and river flows forthese two watersheds are shown in Figure 2.The two rivers have very different patterns:the largely forested Mae Chaem shows avery strong seasonal pattern, falling nearlydry for a few months of the year; the WayBesai (only 15 percent forest) has approximatelycontinual flow. These differences,Function/criteriaMain indicator1. Transmit water • Q/P=1–(E/P)2. Buffer peak rain events • ? Q abAvg /? P abAvg3. Release gradually • Q slow /P = (P inf –E S+V )/P4. Maintain quality • Qual out /Qual in5. Reduce mass wasting • ? riskof course, primarily relate to the rainfallpattern. They show, however, that commonlyused indicators such as the ratio ofmaximum and minimum flow of the river,Q max/Q mincannot be used to analyse thecondition of watersheds, without regards torainfall.The indicators of Figure 1 are all expressedin dimensionless form, relating river flow(discharge) to rainfall. For the analysis ofthe Mae Chaem and Way Besai situations,a new ‘buffering indicator’ was developed(van Noordwijk et al. 2003) that relates thefrequency distribution of daily river flowto the frequency distribution of point-levelrainfall. It can be used to test perceptionsof increased flooding and peak flows. TheWay Besai data relate to a 23-year periodwhere forest cover was reduced from almost30% to less than 10% in 2002. Themain effect of this land-cover change wasto increase the total water yield as a fractionof total rainfall. The total discharge inthe month with the lowest flow, expressedas a fraction of annual rainfall, showedconsiderable variation between years butdid not change along with total water yield.The buffering indicator was negatively correlatedwith the total water yield, but forScaledependentFigure 1. Indicators for the five criteria. The quantitative properties of river discharge changealong the river course, and lead to scale dependence of three out of the five criteria. Q =river flow; P = precipitation; E S+V= total evapotranspiration minus evaporation of canopyintercepted water; abAvg = sum of all above average values; inf = infiltration.

Chapter 12: Watershed functions in productive agricultural landscapes105Way BesaiMae ChaemDaily rainfall (station level) mm day 1DaysADaily rainfall (station level) mm day 1DaysBCDRiver flow m 3 sec –1River flow mm 3 sec –1DaysDaysFigure 2. The records of rainfall (A and B) and river discharge (C and D) of the Way Besai for the 1975–1998 period (A and C), and MaeChaem for the 1988–2000 period (B and D). The thin dark lines trace the maxima and minima of daily values for the observation periods,the solid lighter line indicates the mean daily values (van Noordwijk et al. 2003).the other indicators the trade-off explainedonly a small part of the total variation. Thissuggests that in this catchment area the potentialdownstream benefits of more waterare not associated with negative changes inriver flow during the driest month or withless buffering of peak events.A spatially distributed water balance modeldetermined the current land cover situationas being between natural vegetation offorest on a porous soil, and degraded landwith grassland on a compacted soil. Withoutfine-tuning of the model, an acceptableagreement between the model and actualmeasurements was obtained for the WayBesai (Table 1) and Mae Chaem (Table 2)catchment areas for each of the variousindicators, within the range of forest to degradedlands.This analysis, presented in very condensedform here, suggests that changes in forestcover can modify a number of quantitativecharacteristics of river flow, but that rainfall(and any change in rainfall characteristicsbetween measurement periods) dominatesthe outflows. As rainfall tends to have complexpatterns of variation over time, it is noteasy to tease out a land-use change signalfrom the noisy background. Much of theattribution of change in river flow to landcoverchange in public debate may notsurvive close scrutiny.By comparing the results from the modelwith the measurements taken in the fieldwe can conclude that models that link thespace–time characteristics of rainfall via

106World Agroforestry into the FutureTable 1.Indicators of watershed functions for Way Besai, comparing actual data (averaged over 20 years) with simulations of differentenvironments: the current LU (land use) mix; an ‘all forest’ approximation of natural vegetation; and a ‘degraded lands’scenario with grass cover on a compacted soil. GenRiver simulations use rainfall data, soil information, land-cover type andsub-catchment structure of the watershed area (van Noordwijk et al. 2003).IndicatorsActual dataGenRiverCurrent LU Current LU Natural vegetation Degraded landTotal discharge fraction 0.61 0.53 0.44 0.62Buffering indicator 0.79 0.82 0.80 0.68Relative buffering indicator 0.66 0.66 0.55 0.49Buffering peak events 0.86 0.81 0.76 0.78Highest monthly discharge relativeto mean rainfallLowest monthly discharge relativeto mean rainfall1.92 2.19 1.65 1.580.39 0.54 0.50 0.46Overland flow fraction * 0.11 0.00 0.36Soil quick-flow fraction * 0.10 0.02 0.00Slow flow fraction * 0.30 0.29 0.25* indicates data not available.Table 2.Indicators of watershed functions for Mae Chaem, comparing actual data (averaged over 20 years) with simulations of differentenvironments: the current LU (land use) mix; an ‘all forest’ approximation of natural vegetation; and a ‘degraded lands’scenario with grass cover on a compacted soil. GenRiver simulations use rainfall data, soil information, land-cover type andsub-catchment structure of the watershed area (van Noordwijk et al. 2003).IndicatorsActual dataGenRiverCurrent LU Current LU Natural vegetation Degraded landTotal discharge fraction 0.21 0.19 0.13 0.32Buffering indicator 0.89 0.90 0.93 0.81Relative buffering indicator 0.49 0.45 0.54 0.40Buffering peak events 0.91 0.88 0.91 0.79Highest monthly discharge relativeto mean rainfallLowest monthly discharge relativeto mean rainfall3.16 3.67 3.01 3.370.20 0.22 0.27 0.24Overland flow fraction * 0.00 0.00 0.00Soil quick-flow fraction * 0.08 0.03 0.17Slow flow fraction * 0.14 0.08 0.12* indicates data not available.

Chapter 12: Watershed functions in productive agricultural landscapes107the dynamics of macropores in the soil tothe dynamics of river flow can fairly wellreproduce the time series of data from intensivelystudied (sub)catchments.Comparing land-use optionson the basis of bufferingfunctionsLong time-series with consistent data ofland-cover change are scarce and muchof the existing variation in land cover inagriculturally used landscape mosaics isnot represented in empirical data. Furtherinference on land-use options has to relyon analysis of the various contributing factorsand on synthetic models. Essentially,watershed functions that relate to the quantity,timing and quality of water flows canbe understood by considering steps in thepathway of water through the landscape(Ranieri et al. 2004). The main factors are:• Green cover – leaves intercept raindropsand modify the drip size (and thereforethe splash power they have when theyreach the ground), keeping a relativelysmall amount of water as water film onwet surfaces for rapid evaporation.• Brown cover – the litter layer on the soilsurface protects the soil from splash erosion,feeds soil biota that enhance soilstructure, and acts as a filter for overlandflow, reducing the sediment load.• Soil structure – at the surface and in thesoil determines the speed at which watercan infiltrate and hence the amount ofexcess rainfall that travels over the soilsurface as overland flow. Depending onslope and connectivity of the horizontalflow pathways (pipes) a substantialamount of water can be passed on tostreams as interflow in a matter of hoursafter a rainstorm.• Soil water deficit – water uptake byvegetation between rain events createsspace in the soil pores to absorb water;if the soil structure allows this water toinfiltrate fast enough, water use can thusreduce overland flow.• The drainage network – the network offurrows, gullies, drains, roads, soil profileintersections along roads, temporarystorage sites in ponds and wetlands,streamlets and streams determines howrapid overland flows and subsurface(inter)flows can reach rivers. Whereland-use change affects the timing offlow at a minutes-to-hours scale, the significanceof changes in pathways loosesimportance with increased spatial scale(say for distances more than 10 km), asthe travel time in the river itself (and itsinfluence by the degree of channelling,propensity for use of flood plains andriparian wetlands) starts to dominate.• Properties of the riverbed – if the riverbedconsists of stones and the riverbanks are stable, it can transport cleanwater at high velocity. Where the riverflows through (or meanders in) a landscapewith alluvial material, the rivercan pick up sediment along its way duringpeak flows and carry high sedimentloads regardless of the degree of soilprotection in the uplands. Landslides(linked, for example, to earthquakes,road construction or decrease in soil anchoringby decay of deep tree roots) andvolcanic ash deposits can provide soilmaterial for transport, over and beyondwhat comes from the hillsides.• Point sources of organic and chemicalpollutants – direct use of surface waterfor drinking and other domestic use isnot generally safe downstream of humanhabitation. Water quality for otherpurposes, as well as for maintenanceor restoration of the aquatic ecosystem,its biodiversity and use values, can benegatively affected by point sources oforganic and chemical pollutants. Use ofpesticides, imbalances between fertilizerinputs, uptake by plants (Cadisch et al.2004) and deposition of harvested productsor manure into streams by domesticlivestock (or domesticated elephants inecotourism areas in northern Thailand)can all make other efforts to maintainwatershed functions useless from a userperspective.Modelling physical degradationand rehabilitation processesin the analysis of trade-offsbetween profitability and watershedfunctionsA range of tools and models (e.g. Matthewset al. 2004; Ranieri et al. 2004) exist to relatethe overall performance of a landscapeto (subsets of) this list of influences, as wellas to the ‘natural capital’ (including rainfallregimen, slope, intrinsic soil conditionsand nature of the vegetation replaced byhuman land use).For the specific analysis of agroforestrymosaics in Southeast Asia we use the Wa-NuLCAS (Water, Nutrient and Light Capturein Agroforestry Systems) model at plot level(Khasanah et al. 2004; van Noordwijk et al.2004c), GenRiver and SpatRain for dailytime steps at watershed scale (Farida andvan Noordwijk 2004) and FALLOW (Forest,Agriculture, Low-value Lands Or Waste;Suyamto et al. 2004) to analyse longer-termtrends in land-use change linked to internaldrivers of change. In the remaining part ofthis chapter we will focus on the changes insoil conditions – as this may be the easiestpart to manage for practitioners of agroforestryand other forms of eco-agriculture.Using the soils under old-growth forestas a reference or baseline, soil degradationinvolves the loss of organic matter, adecline in soil nutrient reserves, a changein soil biota and below-ground food-webs,

108World Agroforestry into the Futuresoil compaction and a change in waterretention. The latter includes the capacityof soil to absorb water during rainfallevents; release water during the first day(s)after a rainfall to groundwater and streamsto reach field capacity; and retain water attensions that are appropriate for plants totake up water (Figure 3).The effects of compaction on these propertiesvary with soil type, but can be approximatedby relating the actual bulk density(mass per unit volume) to a reference valuethat can be estimated from the soil texture(and which depends on sand, silt, clay andorganic matter content) on the basis oflarge datasets for agricultural soils (Wöstenet al. 1998). As a first estimate, we mayexpect topsoils under natural forest to havea bulk density (BD) of about 70 percent ofthis reference value, while severely compactedsoils may reach 1.3 times the referencevalue (BDref).Averaged over the 10 main soil groupsrepresented in the database of Suprayogoet al. (2003), the decrease in water-holdingcapacity from a natural forest to along-term agriculturally used soil will be0.136 cm 3 cm –3 , equivalent to the abilityto temporarily store up to about 25 mm ofrainfall in 20 cm of topsoil. This is storagecapacity that can be re-used in a rain eventon the next day, as the water will by thenhave found its way to streams and rivers(or deep groundwater stores, if these arenot yet saturated). Upon further degradationfrom agricultural to degraded lands,a further 0.081 cm 3 cm –3 (or the ability toabsorb 15 mm of rainfall) can be lost. Thisloss of storage capacity is likely to induceoverland flow conditions that can lead toflash floods and erosion.The loss of plant-available water owingto soil compaction is small relative to thepF = log(—matric potential)7.06.05.04.0PermanentwiltingpointPlantavailableSoilQ3.0flowField capacity2.0(drainage dependent)1.0Saturation0.00.0 0.1 0.2 0.3 0.4 0.5 0.6Volumetric water content cm 3 cm —3Figure 3. The main properties of the soil–water retention curve are the total water contentat saturation, the amount retained one day after heavy rain (field capacity), and thepermanent wilting point. Soil compaction primarily affects the soil close to saturation; thecapacity for soil quick-flow (SoilQflow) or interflow depends on the difference betweenfield capacity and saturated soil water content.loss of temporary storage capacity. Theconsequences of soil compaction for thepathways of excess water flows (overland,subsurface lateral flow or deep groundwaterpathways) are thus likely to be morepronounced than those for plant-wateravailability on site.Compaction can, however, negatively affectthe aeration of plant root systems, anda value of air-filled porosity at field capacity(numerically equal to the soil quickflowcapacity) of 0.1 is often interpreted asa critical threshold for sensitive crops.A relatively simple method to visualizeand analyse changes in soil macroporositylinked to land cover makes use of the infiltrationof a dye (Figure 4). The infiltrationpatterns can be interpreted on the basis ofthe general macroporosity of the soil andspecific impacts of cracks, old root channelsand activity of earthworms or othersoil biota.Soil compaction can be rapid; bulldozers,cars, animal hooves and people can allapply sufficient pressure to compact a soil,especially when the latter is wet. In the absenceof soil cover, detachment of fine soilparticles and a process called ‘slumping’also has the same effect. The reverse process,creation of macroporosity, is slow; itprimarily depends on the activities of earthwormsand similar ‘engineers’ and the turnoverof woody roots. Once a soil is severelycompacted, the recovery process maytake decades or up to a century. Soil tillageis a poor substitute for biological structureformation: its effects are short-lived and bydestroying biological structures it in factcreates an addictive effect – once tillagestops, the soil structure generally degradesrapidly. Strategic tillage-like interventions,such as planting holes or crust breakingcan, however, set a long-term biologicalsoil recovery process in motion.Physical soil degradation can also have itsprimary effect via the reduction of the potentialsurface infiltration rate, through theformation of crusts on the soil surface. Inrelatively dry climates this may even be theprimary effect that leads to overland flow

Chapter 12: Watershed functions in productive agricultural landscapes109Depth (cm)Depth (cm)ForestMultistrata coffeemain processes are captured in simulationmodels that have reached considerablepredictive ability, the dynamics of soilstructure in terms of decay and recoveryare still largely a black box, constrainingfurther precision of models of water balancefor example. The WaNuLCAS model(van Noordwijk et al. 2004c) uses theempirical reference value for bulk density,BDref , as a ‘fall-back’ value to which soilstructure decay reverts in the absence ofspecific macropore creation activities,which create macropores directly (vanNoordwijk et al. 2004d). This model descriptionsuggests that the most importantparts of a tree for land rehabilitation are thedead leaves that it sheds and the fine andcoarse root turnover it induces.Simple shade coffeein conditions where the soil remains farfrom saturated. Where surface phenomenasuch as crusting rather than soil compactiondominate in the soil physical degradationprocess, recovery may be faster: anytype of mulch that protects the soil fromthe direct impact of rain and sunshine andthat stimulates soil biological activity maylead to recovery in a timeframe of months.Sun coffeeFigure 4. Infiltration patterns for a dye that leaves a dark trace in all macropores it passesthrough. This simulates what may happen during heavy rainfall on four types of land usein the Sumberjaya benchmark area in West Lampung, Indonesia; see Hairiah et al. (2004)and Widianto et al. (2004) for details on the methods and sites.It is thus important to correctly diagnosewhat type of degradation dominates in agiven location, as this will influence thetimeframe for potential recovery. Avoidingcompaction at sites that are still in a naturalforest condition is probably more effectivethan trying to rehabilitate degraded sites.Where surface processes dominate, however,rapid gains by mulch-based restorationactivities can be expected.Standard soil physical textbooks and handbookof methods specify how BD can bemeasured – but not how the data can be interpreted.Bulk density is strongly related tosoil texture and soil organic matter content(which in itself depends on texture), so fora valid interpretation in the context of compaction,we need to derive a reference valuefor a soil with the same texture. A simplescheme is available in spreadsheet form onwww.ICRAF.org/sea as part of the ecologicalmodels that can be freely downloaded.While the water, nutrient and carbon balanceof soils are well understood, and theA further complication arises when werealize that surface litter, depending onits size and weight, is prone to be carriedaway by wind or overland flow of water,leading to a differentiation of the land intomutually enhancing zones of high infiltrationwith deposition of surface mulch,and zones of crusted soil with high runoff.Classification of litter sources by their propensityto transport is only just starting.A macro version of the transport–depositioneffect is known as the ‘tiger bush’striped pattern in semi-arid lands – wherethe degraded zones act as water harvestingsource areas for the vegetated parts.Land rehabilitation can aim at strategicallymodifying the scale of this pattern, but notat a fully homogeneous state.For a full understanding of the tradeoffsbetween productivity (or profitability) ofland use and the implication for watershedfunctions we thus have a reasonably wellequippedtool kit. There are complications,however, such as the differences in time

110World Agroforestry into the Futurecourse of profitability and the substantialvariation in soil properties, often at shortspatial range, with substantial differencesbetween soils in susceptibility tocompaction.Negotiations betweenstakeholders of solutions onthe basis of trade-offsThe basics of watershed functions arewell understood in most local ecologicalknowledge systems that have so far beenexplored (Joshi et al. 2004), as well as informal ecohydrological science. Their representationin general public debate andpolicy circles, however, leaves much scopefor improvement.Indonesia is rich in examples of landscapeswhere farmers have combined the use oftrees and other elements of the natural forestthat provide environmental serviceswith areas that are used for intensive foodcrop production. These agroforestry mosaiclandscapes can be seen as ‘kebun lindung’(protective gardens) that offer great opportunityfor combining development andenvironment targets (Pasya et al. 2004; vanNoordwijk et al. 2004a). Yet, there are obstaclesto the recognition of these systems,as they may not meet the legal definitions offorest or be in harmony with existing landuseregulation systems and policies – eventhough they could pass the test when functionalcriteria and indicators would be used.In negotiating solutions to local problems,the following aspects may require specificattention:1. Creation of local infiltration sites is oftenthe first step required to break out froma soil degradation–surface runoff erosioncycle. Such sites will both reduce negativeimpacts on downhill neighbouringzones and allow for a positive feedbackloop of vegetation that stimulates formationof soil structure, increasing infiltrationand acting as a further stimulus toplant growth. Triggers of such a positivefeedback can be remarkably simple:stone lines (as used in the Sahel), plantingholes made for trees (that may be thebest part, initially, of reforestation effortsand is often not considered as such) orsmall strips left to natural vegetationsuccession in between ploughed fields(‘natural vegetative strips’, see Chapter 7this volume) as used in the Philippinesand Indonesia.2. Taking natural forest soil as a baseline,soil compaction will initially have astronger effect on the lateral flows thataffect watershed functions than on theon-site productivity of the soil. Whereprotection of forest soils is feasible byreduction of the drivers of degradation,it is likely to be much more effectivethan efforts to rehabilitate degraded locations.Unfortunately, environmentalgovernance and reward systems tendto be reactive, and have difficulties indealing with avoidance of degradation,while rehabilitation is considered worthyof public investment.3. Enhancing soil organic matter levels haslittle direct influence on plant-availablewater, but a strong indirect effect via soilstructure, depending on the texture ofthe soil and the rainfall regime. Susiloet al. (2004) discuss the relationshipbetween total organic input in the agroecosystemand the various levels of thebelow-ground food-web.4. The most important part of a forest froma perspective of soil and water flows islikely to be in the litter and root turnovereffects, and that in turn supports soil biotato maintain soil structure. Half-open(agroforestry) land-use systems with treescan approach the same functionalitywhile providing better livelihood opportunitiesand income (see van Noordwijket al. 2004b, for discussion of trade-offbetween relative ecological and relativeagronomic functions, or REF and RAF).5. For assessment and monitoring purposes,new methods and models thatprovide internal controls in the form ofreference values for soil carbon and BDcan be used to deal with the inherentvariation in soil properties and the relationshipsbetween lateral flow processacross spatial scales.The discussion so far has highlighted theecological/technical side of soil structureand function. If agroforestry is to achieve itsaims, understanding of and actions targetingthese technical aspects at farm-managementscale will have to be embedded in a structureof rules and incentives that relate boththe downstream users of landscapes and thestakeholders in maintenance of watershedfunction to the decisions made on-farm. Thepast focus of watershed managers on forestcover per se may now give way to a moresubtle view in which land uses such as the‘kebun lindung’ in Indonesia get the recognitionthat they are due (Pasya et al. 2004;van Noordwijk et al. 2004a).AcknowledgementsOur research on these topics in Indonesia inthe context of the Alternatives to Slash andBurn consortium is supported by the AustralianCentre for International AgriculturalResearch (ACIAR) and the UK’s Departmentfor International Development (DFID), butthe views expressed remain the authors’responsibility.

Chapter 12: Watershed functions in productive agricultural landscapes111ReferencesCadisch, G., P. de Willigen, D. Suprayogo, D.C.Mobbs, M. van Noordwijk and E.C. Rowe2004. Catching and competing for mobilenutrients in soils. Pp. 171–191, in: vanNoordwijk, M., G. Cadisch and C.K. Ong(eds) Below-ground Interactions in TropicalAgroecosystems. CAB International,Wallingford, UK.Joshi, L., W. Schalenbourg, L. Johansson, N.Khasanah, E. Stefanus, M.H. Fagerströmand M. van Noordwijk 2004. Soil and watermovement: combining local ecologicalknowledge with that of modellers whenscaling up from plot to landscape level.Pp. 349–364, in: van Noordwijk, M., G.Cadisch and C.K. Ong (eds) BelowgroundInteractions in Tropical Agroecosystems.CAB International, Wallingford, UK.Farida and M. van Noordwijk 2004. Analisisdebit sungai akibat alih guna lahan danaplikasi model GenRiver pada DAS WayBesai, Sumberjaya. [Analysis of changes inriver flow in response to land use change,and application of the GenRiver model tothe Way Besai watershed in Sumberjaya].Agrivita 26: 39–47.Hairiah, K., D. Suprayogo, Widianto, Berlian, E.Suhara, A. Mardiastuning, C. Prayogo andS. Rahayu 2004. Alih Guna Lahan HutanMenjadi Lahan Agroforestri Berbasis Kopi:ketebalan seresah, populasi cacing tanahdan makroporositas tanah. [Thickness ofthe litter layer, earthworm populationsand soil macroporosity as influenced byland use in coffee-based agroforestry].Agrivita 26: 68–80.Khasanah, N., B. Lusiana, Farida and M. vanNoordwijk 2004. Simulasi LimpasanPermukaan dan Kehilangan Tanah padaBerbagai Umur Kebun Kopi: Studi Kasusdi Sumberjaya, Lampung Barat. [Simulationof surface runoff and soil loss inrelation to age of coffee gardens: casestudy in Sumberjaya, West Lampung].Agrivita 26: 81–89.Matthews, R., M. van Noordwijk, A.J. Gijsmanand G. Cadisch 2004. Models of belowgroundinteractions: their validity, applicabilityand beneficiaries. Pp. 41–60, in: vanNoordwijk, M., G. Cadisch and C.K. Ong(eds) Belowground Interactions in TropicalAgroecosystems, CAB International,Wallingford, UK.Pasya, G., C. Fay and M. van Noordwijk 2004.Sistem pendukung negosiasi multi tatarandalam pengelolaan sumberdaya alamsecara terpadu: dari konsep hingga praktek.[Negotiation support systems forsustainable natural resource management:from concept to application]. Agrivita 26:8–19.Ranieri, S.B.L., R. Stirzaker, D. Suprayogo, E.Purwanto, P. de Willigen and M. vanNoordwijk 2004. Managing movementsof water, solutes and soil: from plot tolandscape scale. Pp. 329–347, in: vanNoordwijk, M., G. Cadisch and C.K. Ong(eds) Belowground Interactions in TropicalAgroecosystems, CAB International, Wallingford,UK.Suprayogo, D., Widianto, G. Cadish and M. vanNoordwijk 2003. A Pedotransfer resourcedatabase (PTFRDB) for tropical soils: testwith the water balance of WaNuLCAS. In:D. Post (ed) MODSIM proceedings, July2003, Townsville, Australia.Susilo, F.X., A.M. Neutel M. van Noordwijk, K.Hairiah, G. Brown and M.J. Swift 2004.Soil biodiversity and food webs. Pp. 285–307, in: van Noordwijk, M., G. Cadischand C.K. Ong (eds) Belowground Interactionsin Tropical Agroecosystems. CABInternational, Wallingford, UK.Suyamto, D.A., M. van Noordwijk and B.Lusiana 2004. Respon petani kopi terhadapgejolak pasar dan konsekuensinyaterhadap fungsi tata air: suatu pendekatanpemodelan. [Coffee farmers’ responseto market shocks and consequencesfor watershed functions: a modellingapproach]. Agrivita 26: 118–131.van Noordwijk, M., J. Richey and D. Thomas2003. Landscape and (sub)CatchmentScale Modeling of Effects of ForestConversion on Watershed Functionsand Biodiversity in Southeast Asia. Alternativesto Slash and Burn, BNPP Activity 2Technical Report. [http://www.worldagro-forestrycentre.org/sea/upload/meine/FVO-BAct2_all_2.pdf]van Noordwijk, M., F. Agus, D. Suprayogo,K. Hairiah, G. Pasya, B. Verbist andFarida 2004a. Peranan agroforestry dalammempertahankan fungsi hidrologi daerahaliran sungai (DAS). [Role of agroforestryin maintenance of hydrological functionsin water catchment areas]. Agrivita 26:1–8.van Noordwijk, M., G. Cadisch and C.K. Ong2004b. Challenges for the next decade ofresearch on below-ground interactions intropical agroecosystems: client-driven solutionsat landscape scale. Pp. 365–379,in: van Noordwijk, M., G. Cadisch andC.K. Ong (eds) Belowground Interactionsin Tropical Agroecosystems. CAB International,Wallingford, UK.van Noordwijk, M., B. Lusiana and N. Khasanah2004c. WaNuLCAS 3.01, Backgroundon a model of water nutrient and lightcapture in agroforestry systems. InternationalCentre for Research in Agroforestry(ICRAF), Bogor, Indonesia.van Noordwijk, M., S. Rahayu, S.E. Williams,K. Hairiah, N. Khasanah and G. Schroth2004d. Crop and tree root-system dynamics.Pp. 83–107, in: van Noordwijk, M., G.Cadisch and C.K. Ong (eds) BelowgroundInteractions in Tropical Agroecosystems.CAB International, Wallingford, UK.

112World Agroforestry into the FutureWidianto, D. Suprayogo, H. Noveras, R.H.Widodo, P. Purnomosidhi and M. vanNoordwijk 2004. Alih guna lahan hutanmenjadi lahan pertanian: apakah fungsihidrologis hutan dapat digantikan sistemkopi monokultur? [Conversion from forestto agricultural land: is the hydrologicalfunction affected by conversion from forestto coffee monoculture?] Agrivita 26:47–52.Wösten, J.H.M., A. Lilly, A. Nemesand C. andLe Bas 1998. Using existing soil data toderive hydraulic parameters for simulationmodels in environmental studies andin land use planning. Report No. 156,Staring Centre (SC-DLO), Wageningen, theNetherlands.

Keywords:Agroforestry, vulnerability, agricultural research,tropical agriculture, rural developmentChapter 13Opportunities for linking climate change adaptationand mitigation through agroforestry systemsSerigne T. Kandji and Louis V. Verchot, World Agroforestry Centre; Jens Mackensen, United NationsEnvironment Programme; Anja Boye, Meine van Noordwijk, Thomas P. Tomich, Chin Ong, World AgroforestryCentre; Alain Albrecht, World Agroforestry Centre and Institut de recherche pour le développement andCheryl Palm, The Earth Institute at Columbia University, USAAbstractAgroforestry systems not only provide a great opportunity for sequestering carbon, and hence helpingto mitigate climate change, but they also enhance the adaptive capacity of agricultural systems intropical and subtropical regions. Agricultural research over the last few decades has been driven by thequest to increase the productivity and resilience of agricultural systems. While increasing productivityrelates directly to the ability of a system to accumulate and retain carbon, improving the resilience ofagricultural systems is largely the result of enhancing the capacity of such systems to cope with adverseclimatic changes. This chapter presents data that examine the mitigation and adaptation potential ofdifferent agroforestry systems as well as their significance for income generation for rural populations.New areas of research are proposed and a better use of existing agricultural management knowledgeis called for.IntroductionClimate change will affect developing countries moreseverely because of their low capacity for adaptation(IPCC 2001). Within these countries, the agriculturalsector is particularly vulnerable, putting rural populationsat risk. Furthermore, climate change is an additionalthreat that might affect a country’s ability tomeet urgent rural development demands including theimprovement of food security, poverty reduction, andprovision of an adequate standard of living for growingpopulations. There is a real risk of losing the gains ofthe Green Revolution, which has largely eliminated thedanger of famines such as those seen in the 1950s and1960s. Several modelling studies carried out in SouthAsia to assess the impact of climate change (Aggarwaland Mall 2002; Aggarwal and Sinha 1993; Berge et al.1997; Kropff et al. 1996; Rao and Sinha 1994; Saseendranet al. 2000) have shown that increases in temperaturelead to a decrease in the length of the growing seasonand the yield of most crops. Maize production inthe tropics is predicted to decline by 10 percent (Jonesand Thornton 2003), with regions such as the Sahel andsouthern Africa suffering disproportionately.Within the United Nations Framework Convention onClimate Change (UNFCCC) negotiation process, mitigationand adaptation activities have been largely dealtwith as separate matters. Carbon sequestration throughland use, land-use change and forestry (LULUCF) as ameasure for mitigating climate change has been a very

114World Agroforestry into the Futurecontentious issue during recent negotiations.However, agreements have beenmade on the modalities and procedures forLULUCF projects, which offer, inter alia,opportunities for agroforestry activitiesunder the Clean Development Mechanism(CDM). Adaptation, on the other hand,was only recently recognized as an importantand separate topic as expressed, forexample, in the Delhi Declaration of theUNFCCC eighth session of the Conferenceof the Parties (COP 8) in 2002.The discussion on the potential synergiesbetween adaptation and mitigationmeasures is just starting and is all toooften reduced to a discussion of the costsof global adaptation vs. global mitigation.A practical understanding of the linkbetween adaptation and mitigation measures,particularly with respect to land useand land management, does not yet exist.Yet agricultural research in the last fewdecades has been addressing the need tocope with adverse and irregular climaticconditions including rainfall variability orshifting weather patterns. Similarly, therehas been a major emphasis on improvingthe productivity of agricultural systems,leading to the understanding that increasingsoil carbon stocks in degraded lands isessential for enhanced productivity. Agroforestryprovides a unique opportunity toreconcile the objectives of mitigation of,and adaptation to, climate change.Agroforestry and climatechange mitigationA wide range of studies (Albrecht andKandji 2003; IPCC 2000; Palm et al. 2005)have substantiated the fact that agroforestrysystems, even if they are not primarily designedfor carbon sequestration, presenta unique opportunity to increase carbonstocks in the terrestrial biosphere (Table 1).Worldwide it is estimated that 630 x 10 6ha are suitable for agroforestry. Carbon isparticularly useful in agricultural systems(Figure 1), making agroforestry a quantitativelyimportant carbon sink.Agroforestry systems in the humid tropicsare part of a continuum of landscapesranging from primary forests and managedforests to row crops or grasslands. They aremostly perennial systems such as homegardensand agroforests in which the tree componentcan stay in the field for more than20 years. Agroforestry trees play importantroles including shading tree crops such ascocoa, nutrient cycling and improving themicroclimate. Since trees and crops growat the same time, these systems are referredto as simultaneous systems. Figure 2 showsthat converting primary tropical forests toagriculture or grassland results in a massiveloss of carbon storage capacity. While agroforestrysystems contain less carbon thanprimary or managed forests, the fact thatthey contain significantly higher carbonTable 1.Potential carbon (C) storage 1 for agroforestry systems in different ecoregions ofthe world.Ecoregion System Mg C ha –1Africa humid tropical high agrosilvicutural 29–53South AmericaSoutheast Asiahumid tropical lowdry lowlandshumid tropicaldry lowlandsagrosilvicuturalagrosilvicutural39–10239–19512–22868–81Australia humid tropical low silvopastoral 28–51North Americahumid tropical highhumid tropical lowdry lowlandsstocks than row crops or pastures suggeststhat the introduction and proper managementof trees in crop lands has a great potentialfor carbon sequestration, in additionto rehabilitating degraded land.Unlike simultaneous systems, improved fallowsare tree–crop rotation systems wherebyfast-growing, often leguminous, treesare cultivated for a period of 8 monthsto 3 years to enhance nutrient-depletedsoils and degraded lands in the sub-humidtropics. Even in drier areas such as theSudan–Sahel zone of West Africa, recentfield experiments have shown that thistechnology could significantly contributeto curbing land degradation and improvingfarm productivity. Typically, improvedfallows are short-term rotation systems andas such sequester much less carbon aboveground than perennial systems. However,several studies on soil carbon dynamicshave indicated that soil organic matterincreases after a few seasons of tree plantingon degraded soils. On-farm trials in thesilvopastoralsilvopastoralsilvopastoral133–154104–19890–175Northern Asia humid tropical low silvopastoral 15–181Carbon storage values were standardized to a 50-year rotation.Sources: Dixon et al. 1993; Krankina and Dixon 1994; Schroeder 1993; Winjum et al.1992).

Chapter 13: Opportunities for linking climate change adaptation and mitigation115sub-humid tropics of Togo and Kenya haveshown various degrees of success dependingon location (rainfall and soil type),fallow species, duration of the fallow phaseWetlandrestorationRestoration ofdegraded landsAgroforestryForestmanagementGrazingmanagementRicemanagementCroplandmanagementFigure 1. Carbon (C) sequestration potential (in millions of tonnes per year) of differentland use and management options.Source: IPCC (2000).C stock (Mg ha —1 )3002502001501005000 100 200 300 400 500 600 700Primary forestPotential C sequestration by 2040 (Mt C y —1 )Managedforestand sampling depth; soil organic carbonaccretions through employing improvedfallow were estimated to be between 1.69and 12.46 Mg ha –1 (Table 2).Agroforestry systemsEcosystemsCrops, pasturesand grasslandsFigure 2. Summary of carbon (C) stocks in different ecosystems of the humid tropics. Dataare from the benchmark sites of the Alternatives to Slash and Burn (ASB) Programme of theConsultative Group on International Agricultural Research (CGIAR).Although carbon fluxes in agroforestry systemsare well documented, we have a muchpoorer understanding of the effects of thesepractices on non-carbon dioxide (CO 2)greenhouse gases. In the case of nitrous oxide(N 2O) emissions, much depends on thepresence or absence of legumes in the system.In general, agroforestry systems, whichpromote the use of legumes as fertilizer orshade trees, may increase N 2O emissionscompared to unfertilized systems. Similarly,tree-based systems that encourage the introductionand development of livestockfarming may contribute to increasing methane(CH 4) emissions. While efforts shouldbe made to minimize the emission of thesetrace gases, what ultimately matters in termsof climate change mitigation is how theseemissions compare to the amount of carbonsequestered in agroforestry systems. Forexample, in an improved fallow–maize rotationsystem in Zimbabwe, N 2O emissionswere found to be almost 10 times those ofcontinuous unfertilized maize (Chikowoet al. 2003), but these levels were still extremelylow when compared to the increasein the amount of carbon stored.Enhancing farmer adaptivecapacity through agroforestryAs adaptation emerges as a science, therole of agroforestry in reducing the vulnerabilityof agricultural systems (and the ruralcommunities that depend on them for theirlivelihood) to climate change or climatevariability needs to be assessed more effectively.Rainfall variability is a majorconstraint in the semi-arid regions and tothe upland farms in Southeast Asia that donot have access to irrigation. However,the effects of variable rainfall are oftenexacerbated by local environmental degradation.Therefore, curbing land degradationcan play an important role in mitigatingthe negative impacts of climate change and

116World Agroforestry into the FutureTable 2.Soil organic carbon (SOC) increase over the duration of the fallow phase in a few tropical soils with different tree species in thesub-humid tropics.SOC increaseCountryFallow duration(years)Soil typeFallow speciesSampling depth(cm)Total(Mg ha –1 )Annual(Mg ha –1 yr –1 )Togo 5 Ferric Acrisol (sandy) Acacia auriculiformisAlbizzia lebbeckAzadirachta indicaCassia siamea0–10 3.41–12.46 0.68–2.49Kenya 1.5 Arenosol (sandy) Crotalaria grahamianaC. paulina0–20 1.69–2.15 1.13–1.43Kenya 1.5 Ferralsol (clayey) Crotalaria grahamianaC. paulinaTephrosia vogelii0–20 2.58–3.74 1.72–2.49Source: Albrecht and Kandji (2003).variability, and that is where agroforestrycan be a relevant practice.Successful and well-managed integration oftrees on farms and in agricultural landscapesoften results in diversified and sustainablecrop production, in addition to providing awide range of environmental benefits suchas erosion control and watershed services.In western Kenya, the World AgroforestryCentre, together with various partners, hastested the potential of improved fallowsystems for controlling soil erosion, usingfast-growing shrubs such as Crotalaria spp.and Tephrosia spp. These species showedgreat promise in reducing soil losses (Boyeand Albrecht 2005). At the same time a significantimprovement in soil water storagehas been observed in the improved fallowsystems (Figure 3). We now understand thatclimate change may translate into reducedtotal rainfall or increased occurrence of dryspells during rainy seasons in many semiaridregions. Therefore, optimizing the useof increasingly scarce rainwater throughagroforestry practices such as improved fallowcould be one way of effectively improvingthe capacity of farmers to adapt to drierand more variable conditions.Under many of the different farmer practicesin Africa, crops will still fail completelyor yield very little in drought years. Resultsfrom improved fallow trials were used tomodel these various systems. The modelsuggested that it would be possible to producean acceptable amount of food in lowrainfall years if practices such as improvedfallows were pursued (Table 3). As expected,maize production was higher afterimproved fallow than in a continuous croppingsystem in good rainfall years (typically962–1017 mm of rain). A similar trend wasobserved in low rainfall years (< 600 mm).Most interestingly, the model predictedthat maize yield in a low rainfall year aftera Sesbania spp. fallow period was evenhigher than maize yield in the continuouscropping system in a good rainfall year.If we define rainfall use efficiency (RUE) asthe amount of maize (in kg) produced witheach mm of rainwater, then, apparently,the maize crop after improved fallow madebetter use of the available water than thecontinuous crop, especially when rainfallwas low (Table 3). In low-rainfall years, wateravailability to crops is paramount andseems to be the dividing factor betweenabsolute crop failure and reasonable foodproduction. Buffering agricultural cropsagainst water deficiencies is, therefore, animportant function agroforestry would haveto play in the adaptation battle.There are other mechanisms such as improvedmicroclimate and reduced evapotranspirationthrough which agroforestrypractices may improve the adaptive capacityof farmers. In the African drylands,where climate variability is commonplace,farmers have learned to appreciate the roleof trees in buffering against production

Chapter 13: Opportunities for linking climate change adaptation and mitigation117T. candidaC. grahamianaC. paulinaNatural fallowContinuous maizerisk (Ong and Leakey 1999). The parklandfarming system, in which trees are encouragedto grow in a scattered distributionon agricultural land, is one example. Oneof the most valued (and probably mostintriguing) trees in the Sahel is Faidherbiaalbida. Thanks to its reversed phenology(the tree sheds its leaves during the rainyseason), F. albida significantly contributesto maintaining crop yield through biologicalnitrogen fixation and provision of afavourable microclimate while minimizingtree–crop competition. A study on an0 10 20 30 40Overall water increase (mm)Figure 3. Change in soil water stocks (0–60 cm depth) in a western Kenyan soil undercontinuous maize, natural fallow and improved fallow systems using either Tephrosiacandida, Crotalaria grahamiana or Crotalaria paulina.Source: Orindi (2002).F. albida–millet parkland system in Nigerdemonstrated that shade-induced reductionof soil temperatures, particularly at thetime of crop establishment, is critical forgood millet growth (Vandenbeldt and Williams1992).This type of reversed phenology is not observedin other parkland trees such as theshea butter tree (Vitellaria paradoxa) andnéré (Parkia biglobosa), which have a negativeshading effect that may reduce milletyield under the tree by 50 to 80 percent insome cases (Kater et al. 1992). Farmers arewell aware of this loss in yield, but do notmind it since the economic benefits fromharvesting marketable tree products largelycompensate for the loss of crop yield. However,in extremely hot conditions (which wemay have to face in the future), the shadingeffect of these evergreen trees could compensatefor the yield losses due to excessheat in the open areas of the field. Such ahypothesis has been validated by the workof Jonsson et al. (1999), who measuredvariables including temperature, photosyntheticallyactive radiation (PAR is thelight in the 400–700 nm waveband of theelectromagnetic spectrum that is useful forphotosynthesis) and millet biomass underand away from tree canopies in a parklandsystem (Table 4). The results showed thatdespite the heavy shading, similar amountsof millet biomass were obtained from theareas under these trees and in the open.This absence of yield penalty under treeswas, to a great extent, explained by the factthat millet seedlings under tree canopiesexperienced only 1–9 hours per week ofsupra-optimal temperatures (> 40°C) comparedwith 27 hours per week in the open.In other words, the shorter exposure toextreme temperatures compensated for themillet biomass loss that would otherwisehave occurred as a result of shading. Thisunderscores the important role trees couldTable 3.Grain yield (kg ha –1 ) and rainfall use efficiency (RUE, kg mm –1 ) of maize in continuous maize and improved fallow (IF;Sesbania sesban) systems across five seasons in Makoka, Zambia.Season 1(rainfall = 1001 mm)Season 2(1017 mm)Season 3(551 mm)Season 4(962 mm)Season 5(522 mm)Maize IF Maize IF Maize IF Maize IF Maize IFGrain yield 990 1100 1300 2400 600 1850 1100 2300 500 1180RUE 0.99 1.10 1.28 2.36 1.09 3.36 1.14 2.39 0.96 2.26

118World Agroforestry into the FutureTable 4.TreatmentMean temperature (T), duration when temperature exceeds 40°C (H40),photosynthetically active radiation (PAR) and millet biomass harvested underand away from the tree canopies. (Standard errors in parentheses).T(°C)V. paradoxa (large) –V. paradoxa (small)P. biglobosa (large)P. biglobosa (small)Control 129.10 (0.3)28.30 (0.5)27.00 (0.3)29.98 (0.4)H40(h week –1 )–19527PAR(µE m –2 s –1 )429 (57)541 (64)451 (57)660 (45)2158 (40)Millet biomass(g dry weight plant –1 )46.2 (16.5)43.3 (17.5)56.2 (14.6)36.8 (14.3)39.8 (15.2)Crops Research Institute for the Semi-AridTropics (ICRISAT) in Niger during 2002.The estimated income from a 1-ha farmwas US$600, some 12 times the value of atypical millet crop (Table 5). The estimatedcosts of establishing the SEF are not high;the plant material costs about US$60 perha, and the one-time application of fertilizerabout US$10. The labour requirementsfor land preparation and tree planting aremet by farmers and their families.1Control plots away from tree canopies.Source: Jonsson et al. (1999).play in mitigating the negative effects ofextreme temperatures on crops, especiallyin semi-arid regions.Pests, diseases and weeds already standas major obstacles to crop production inmany tropical agroecosystems and there arestrong reasons to believe that their prevalenceand deleterious effects on crops mayincrease with a warmer climate (Beresfordand Fullerton 1989; Hill and Dymock1989; Rosenzweig et al. 2000). It is stronglybelieved (Altieri and Letourneau 1982;Speight 1983), yet not sufficiently tested,that enhancing plant biodiversity and mixingtree and herbaceous species in agriculturallandscapes can produce positiveinteractions that could contribute towardscontrolling pest and disease outbreaks. Thepotential of agroforestry to control both ordinaryweeds (Gallagher et al. 1999; Impala2001) and parasitic weeds such as Strigahermonthica (Rao and Gacheru 1998) hasalso been demonstrated.Income generation throughtree productsBesides the biophysical resilience, whichallows the various components of the agroforestrysystems to withstand shocks relatedto climate variability, the presence of treesin agricultural croplands can provide farmerswith alternative or additional sourcesof income, so strengthening the socioeconomicresilience of rural populations. Treeproducts (including timber, fodder, resinsand fruits) are normally of higher valuethan maize or hard grains such as milletand sorghum, and can buffer against incomerisks in cases of crop failure.The Sahelian Eco-Farm (SEF) providesan eloquent example of how an agroforestry-basedintegrated natural resourcemanagement regime can help improve thelivelihood of the rural poor in vulnerableregions such as the Sahel (Pasternak et al.2005). The SEF is an integrated land-usesystem that incorporates high-value multipurposetrees/shrubs with soil and waterconservation structures. The value producedis in the form of food, fuelwood and forage(which can all be converted into cash),plant nutrients, biomass for mulch (whichcontributes to increased infiltration of rainfall,and addition of organic matter to thesoil), and protection from wind erosion. Thefirst on-station test of the SEF took placeat the Sahelian Center of the InternationalIn the semi-arid zone of Kenya, the parklandsystem is showing similar success.The fast-growing indigenous species Meliavolkensii is highly compatible with cropsand can provide high-value timber in 5–10years (Stewart and Blomley 1994). A studyby Ong et al. (2002) in the Kitui district ofKenya showed that in an 11-year rotationperiod, the accumulated income from treeproducts exceeds the accumulated valueof crop yield lost through competition. Thisincome difference is worth US$10 or 42percent during average years, and US$22or 180 percent if a 50 percent rate of cropfailure owing to drought (reasonable forKitui) is assumed. In such a hostile environment,where crops normally fail every otheryear, good and secure financial returnsfrom M. volkensii even in drought yearscan provide significant relief for farmers.This will be all the more necessary as extremeclimate events (droughts and floods)are likely to increase in frequency and inmagnitude in the near future.ConclusionsThe impact of climate change will be felton several levels in the agricultural sector.Most of the effects will hit the rural poorin developing countries, who are the mostvulnerable because of their poor ability toadapt. The adaptive capacity of farmers indeveloping countries is severely restricted

Chapter 13: Opportunities for linking climate change adaptation and mitigation119Table 5.Value of Sahelian Eco-Farm (SEF) products from SEF–ICRISAT (International Crops Research Institute for the Semi-Arid Tropics)Sadoré station during 2002.Species Quantity per unit area Yield per unitUnit value(US$)Total revenue(US$ ha –1 )Acacia colei 320 trees ha –1 2 kg seeds tree 0.14 kg –1 90Zizyphus mauritiana 63 trees ha –1 30 kg fresh fruit tree –1 0.12 kg –1 225Andropogon gayanus 567 metres ha –1 1 bundle 10m 0.8 bundle –1 45Millet 1/3 ha 1500 kg ha –1 0.1 kg –1 50Cowpea 1/3 ha 1260 kg ha –1 0.2 kg –1 84Roselle 1/3 ha 400 kg ha –1 0.8 kg –1 106Total 1 ha 600Source: Pasternak et al. (2005).by their heavy reliance on natural factorsand a lack of complementary inputs andinstitutional support systems.The concepts of resilience and sustainableproductivity are well established inagriculture and can be linked directly tothe discussions about adaptation to andmitigation of climate change. Thus, policymakers can draw upon a substantial bodyof knowledge in this respect. However,the adaptation and mitigation synergies ofagroforestry management systems warrantfurther investigation.Within international fora, there is muchtalk about bringing adaptation into themainstream of planning processes. Wehave shown above, through the specificcase of agroforestry, that some mitigationmeasures simultaneously provide opportunitiesto increase the resilience of agriculturalsystems. It is suggested that such synergiesought to be promoted more intensivelythrough the channels of the UNFC-CC such as the CDM. However, if agroforestryis to be used in carbon sequestrationschemes including the CDM, several areasneed to improve, for example, we needbetter methods of assessing carbon stocksand non-CO 2emissions. Furthermore, thedebate on durable wood products is ongoing,but what is known is that farmers willneed provisions to allow them to marketwood products from their agroforestrysystems, and we should develop methodsto account for the lifetime of the carbonsequestered in agroforestry products. Assmall-scale farmers are enrolled in carbonoffsetprojects, we will need to develop abetter understanding of the implications ofthese for carbon sequestration by agroforestryand what it means to livelihoods. Finally,the CDM has very stringent rules forparticipation that may be beyond the reachof small-scale farmers to understand or toprovide evidence of compliance. There is aneed for institutional support by national,regional and international centres of excellenceto facilitate effective participation ofsmall-scale farmers in the CDM.In their attempts to develop adaptationstrategies for the agricultural sector, scientistsand policy makers must consider thecomplex interactions of constraints createdby changing climates in the light of otherstress factors. Government and internationalsupport in terms of research, education,and extension will be required to helpfarmers in developing countries cope withthe additional stresses created by climatechange and increased climate variability.Agroforestry can very likely contribute toincreasing the resilience of tropical farmingsystems. However, our understanding ofthe potential of agroforestry to contribute toadaptation to climate change is rudimentaryat best. Better information is requiredon the role of agroforestry in bufferingagainst floods and droughts from both thebiophysical (e.g. hydraulic lift or soil fertility)and financial (e.g. diversification andincome risk) points of view.Agroforestry promises to create synergiesbetween efforts to mitigate climate changeand efforts to help vulnerable populationsadapt to the negative consequences ofclimate change. The research agenda in thisarea is fairly well defined; much is alreadyknown and putting these ideas into practiceon the ground with small-scale farmers willallow us to learn important lessons.

120World Agroforestry into the FutureReferencesAggarwal, P.K. and R.K. Mall 2002. Climatechange and rice yields in diverse agroenvironmentsof India. II. Effect of uncertaintiesin scenarios and crop models onimpact assessment. Climate Change 52:331–343.Aggarwal, P.K. and S.K. Sinha 1993. Effect ofprobable increase in carbon dioxide andtemperature on productivity of wheat inIndia. Journal of Agricultural Meteorology48: 811–814.Albrecht, A. and S.T. Kandji 2003. Carbonsequestration in tropical agroforestry systems.Agriculture, Ecosytems and Environment99: 15–27.Altieri, M.A. and D.K. Letourneau 1982. Vegetationmanagement and biological controlin agroecosystems. Crop Protection 1:405–430.Beresford, R.M. and R.A. Fullerton 1989. Effectsof climate change on plant diseases. DSIRPlant Division Submission to ClimateChange Impacts Working Group, Departmentof Industrial and Scientific Research,Wellington, New Zealand.Berge, H.F.M., P.K. Aggarwal and M.J. Kropff(eds) 1997. Applications of Rice Modelling.Elsevier Publishers, Amsterdam, theNetherlands.Boye, A. and A. Albrecht 2005. Soil erodibilitycontrol and soil carbon losses under shortterm tree fallows in western Kenya. Advancesin Soil Science (in press).Chikowo, R., P. Mapfumo, P. Nyamugafata andK.E. Giller 2003. Mineral N dynamics,leaching and nitrous oxide losses undermaize following two-year improved fallowson a sandy loam soil in Zimbabwe.Plant and Soil 259: 315–330.Dixon, R.K., K.J. Andrasko, F.A. Sussman, M.A.Lavinson, M.C. Trexler and T.S. Vinson1993. Tropical forests: Their past, presentand potential future role in the terrestrialcarbon budget. Water, Air and Soil Pollution70: 71–94.Gallagher, R.S., E.C.M Fernandes and E.L. Mc-Callie 1999. Weed management throughshort-term improved fallows in tropicalagroecosystems. Agroforestry Systems 47:197–221.Hill, M.G. and J.J. Dymock 1989. Impact ofClimate Change: Agricultural/HorticulturalSystems. DSIR Entomology DivisionSubmission to New Zealand ClimateProgramme. Department of Scientific andIndustrial Research, Wellington, NewZealand.Impala 2001. First annual report of ImpalaProject, covering the period October2000–December 2001. ICRAF, Nairobi,Kenya.IPCC 2000. Land-Use, Land-Use Change andForestry. Special Report of the IntergovernmentalPanel on Climate Change. CambridgeUniversity Press, Cambridge, UK.IPCC 2001. Climate Change 2001. Third AssessmentReport of the Intergovermental Panelon Climate Change (http://www.grida.no/climate/ipcc_tar).Jones, P.G. and P.K. Thornton 2003. The potentialimpacts of climate change on maizeproduction in Africa and Latin America in2055. Global Environmental Change 13:51–59.Jonsson, K., C.K. Ong and J.C.W. Odongo 1999.Influence of scattered néré and karité treeson microclimate, soil fertility and milletyield in Burkina Faso. Experimental Agriculture35: 39–53.Krankina, O.N. and R.K. Dixon 1994. Forestmanagement options to conserve andsequester terrestrial carbon in the RussianFederation. World Resources Review 6:88–101.Kater, L.J.M., S. Kante and A. Budelman 1992.Karité (Vitellaria paradoxa) and néré(Parkia biglobosa) associated with cropsin South Mali. Agroforestry Systems 18:89–105.Kropff, M.J., P.S. Teng, P.K. Aggarwal, B. Bouman,J. Bouma and H.H. van Laar 1996.Applications of Systems Approaches at theField Level. Volume 2. Kluwer AcademicPublishers, The Hague, the Netherlands.Ong, C.K. and R.R.B. Leakey 1999. Why treecrop interactions in agroforestry appear atodds with tree-grass interactions in tropicalsavannahs. Agroforestry Systems 45:109–129.Ong, C.K., J. Wilson, J.D. Deans, J. Mulatya,T. Raussen, N. Wajja-Musukwe 2002.Tree-crop interactions: manipulation ofwater use and root function. AgriculturalWater Management 53: 171–186.Orindi, V.A. 2002. Soil and water conservationunder some improved fallow plant speciesin Vihiga district, Kenya. Master’s degreethesis, Kenyatta University, Nairobi,Kenya, 93pp.Palm, C.A., M. van Noordwijk, P.L. Woomer, J.Alegre, L. Arevalo, C. Castilla, D.G. Cordeiro,K. Hairiah, J. Kotto-Same, A. Moukam,W.J. Parton, A. Rice, V. Rodriguesand S.M. Sitompul 2005. In: Cheryl A.Palm, Stephen A. Vosti, Pedro A. Sanchez,and Polly J. Ericksen (eds). Carbon lossesand sequestration following land usechange in the humid tropics. Alternativesto Slash and Burn: The Search for Alternatives.Columbia University Press, NewYork, USA.Pasternak, D., A. Nikiema, D. Fatondji,J. Ndjeunga, S. Koala, A. Dan Gomaand T. Abasse 2005. The Sahelian Eco-Farm. Pp 286–296 in: Omanya, G. andD. Pasternak (eds). Sustainable AgricultureSystems for the Drylands. InternationalCrops Research Institute for the Semi-AridTropics (ICRISAT), Patancheru, India (inpress).Rao, M.R. and E. Gacheru 1998. Prospects ofagroforestry for striga management. AgroforestryForum 9(2): 22–27.Rao, G.D. and S.K. Sinha 1994. Impact ofclimatic change on simulated wheatproduction in India. Pp. 1–10 in:Rosenzweig, C. and I. Iglesias (eds)

Chapter 13: Opportunities for linking climate change adaptation and mitigation121Implications of Climate Change for InternationalAgriculture: Crop ModellingStudy, Environmental Protection Agency,Washington, DC, USA.Rosenzweig, C., A. Iglesias, X.B. Yang, P.R.Epstein and E. Chivian 2000. ClimateChange and US Agriculture: The Impactsof Warming and Extreme Weather Eventson Productivity, Plant Diseases and Pests.Center for Health and the Global Environment,Harvard Medical School, Boston,USA.Saseendran, S.A., K.K. Singh, L.S. Rathore, S.V.Singh and S.K. Sinha 2000. Effects ofclimate change on rice production in thetropical humid climate of Kerala, India.Climate Change 44: 495–514.Schroeder, P. 1993. Agroforestry systems: Integratedland use to store and conservecarbon. Climate Research 3: 53-60.Speight, M.R. 1983. The potential of ecosystemmanagement for pest control. Agriculture,Ecosystems and Environment 10: 183–199.Stewart, M. and T. Blomley 1994. Use of Meliavolkensii in a semi-arid agroforestrysystem in Kenya. Commonwealth ForestryReview 73: 128–131.Vandenbeldt, R.J. and J.H. Williams 1992.The effect of soil temperature on thegrowth of millet in relation to the effectof Faidherbia albida trees. Agriculturaland Forest Meteorology 60: 93–100.Winjum, J.K., R.K. Dixon and P.E. Schroeder1992. Estimating the global potential offorest and agroforest management practicesto sequester carbon. Water, Air andSoil Pollution 64: 213–228.

Chapter 14Environmental Services:Working Group ReportIntroductionWorking groups were designated around the four focalareas of the environmental services theme: watershedmanagement, biodiversity, climate change and environmentalpolicy. The groups considered three tasks:• Identify problems and challenges for environmentalservices and pro-poor agroforestry.• Identify significant roles that the Centre can undertaketo address these challenges.• Identify targets for the Centre to achieve over thenext 10 years.The following conclusions were reached.Watershed management: pro-poorstrategies to enhance watershedfunctionsProblems and challenges of watershedmanagement and pro-poor agroforestryThe last 20 years have seen a wide variety of approachesto watershed management across the developingworld. There has been some synthesis of this experienceduring the last few years and four major conclusionsstand out. First, watershed management programmesneed to be redesigned to better involve multiplestakeholders and link upstream resource use withdownstream impacts. Second, there are vital gaps inlocal and global knowledge about land–water and landuse interactions that inhibit appropriate action for improvedwatershed management, particularly over largeland areas. Third, there is a need for better networkingamong people involved in research, development andpolicy for watershed management. And fourth, there isincreased realization that the main output of watershedmanagement is water.Forest management has traditionally been tightlyassociated with good watershed management. However,with increased pressure on available water andstate forest resources, there is a need to reassess therelationships between trees, water and soils in watersheds.Improved understanding of the effects of treeson different watershed functions – water quality, floodrisk, landslide risk, dry-season flow – can help to targetforest conservation, afforestation and agroforestry interventions.The United Nations Convention to Combat Desertification(UNCCD), the UN Convention on BiologicalDiversity (UNCBD) and the UN Framework Conventionon Climate Change (UNFCCC) are creating newsituations in which public and private agencies areplanting trees for the sake of improved environmentalmanagement. Without due care, this might put unacceptablenew pressures on available water resources.There is a need to assess these risks and tradeoffs.Roles for ICRAF in pro-poor watershedmanagementThree main questions motivate the Centre’s work onpro-poor watershed management. First, what is thepotential for agroforestry systems to conserve or restorewatershed functions in multiple-function watersheds?Second, how can agroforestry systems be designed andadapted to have the greatest beneficial impact on waterquality, flood risk, landslide risk and dry-season flowin the wide range of environments found in watersheds

124World Agroforestry into the Futureacross the less-developed countries of theworld? Third, how can poor farmers benefitfrom the watershed benefits produced bytheir agroforestry systems?During the past 5–10 years, the Centre hasengaged in watershed management researchprogrammes in a number of sites inSoutheast Asia and East Africa. It is essentialthat the Centre continue to engage withlocal stakeholders in a range of conditions.Ground-level work is particularly importantfor evaluating tree–water–land interactionsand for assessing negotiation support andenvironmental service mechanisms thatcan link different stakeholders. Outputsshould be synthesized across sites andimplications for larger areas can be drawn.There is an increasing role for ICRAF in policydevelopment and advocacy, connectingand linking tree farmers (small- and largescale),non-governmental organizations(NGOs) and government agencies to developimproved schemes for reforestation.The Centre can also help in forging linksbetween different line agencies.Furthermore, ICRAF can play a part in networkingat the regional and global levels.It has begun to do this through co-hostingthe African regional workshop for thenext generation of watershed managementprogrammes.ICRAF priority outputs in pro-poor watershedmanagement for next 10 years:• Develop guidelines and manuals foragroforestry and negotiation support inwatersheds. This should include informationon how to deal with new speciesand potentially invasive species.• Help to develop reward mechanismsand joint systems for watershed management.ICRAF will need to developdifferent models for different water usesand water management objectives (totalwater yield, water quality, reduced sedimentation,and so on).• Develop and test alternative financialand reward mechanisms for linking upstreamand downstream watershed management(e.g. the polluter [and wateruser] pays principle).• Consider where and how protection andnatural regeneration of vegetation maybe preferable to new tree planting.Biodiversity: use andconservation of biologicaldiversity in workinglandscapesProblems and challenges ofbiodiversity in working landscapesScientists are increasingly recognizing thatmuch of the world’s valuable biodiversityexists outside of protected areas, and that itis strongly influenced and shaped by humanactivity. Recent reviews show that changesin agroforestry practices can affect wildbiodiversity in both positive and negativeways (see Chapter 11, this volume). A positivecontribution results from the provisionof habitat areas and corridors for wildlifeand plant geneflow in multi-use landscapemosaics. Generally, the impact of alieninvasive species on indigenous biodiversityis potentially negative. In the conservationcommunity, there is increasing awareness ofthe need for ecosystem and landscape approachesto conservation planning, includingagroforestry and community forestry. Forexample, the UNCBD recognized the valueof the ecosystem approach in 2000, whilethe World Conservation Union (IUCN) hasnow embraced ecosystem management asone of its major themes (www.iucn.themes/cem). There is currently a need to identifymodels that should be scaled up, and toidentify policies that shape land use in conservationlandscapes.For landscape approaches to biodiversityconservation to be convincing to local andnational policy makers, however, goodquantitative analysis of the contributionsof biodiversity to rural livelihood improvementis needed. There are both directcontributions – through agroforestry andnon-timber forest products – and indirectcontributions, through the ecosystem servicesaffected by biodiversity (e.g. nutrientcycling and water balances). It is importantto go beyond simply measuring the compositionof biodiversity in order to gainan improved understanding of its impacton ecosystem structure and function. Forexample, there are possible trade-offsbetween the genetic diversity within onespecies and across many species. We alsoneed to know more about the contributionof tree domestication to the livelihoods oflocal people.Information on the trade-offs associatedwith different landscape configurationsneeds to be developed and shared with therange of stakeholders who have interestsin conservation areas and the surroundinglandscapes. Negotiated agreements on theconservation, use and shared benefits ofbiodiversity require those various stakeholdersto have a shared understanding.Furthermore, there is a need to creategreater awareness of the ways in whichagroforestry can contribute to achievingthe social goals related to biodiversity.The messages need to be simple enoughto shape policy and programme design,but should not oversimplify complexprocesses.The main roles for ICRAF inbiodiversityICRAF has a comparative advantage in researchon agro-ecosystem biodiversity relativeto most other organizations concernedwith biodiversity. That is, agroforestry links

Chapter 14: Environmental Services125plot to landscape scales, agrobiodiversityto wild biodiversity, biodiversity to ecosystemfunction and livelihoods, and linksnational systems, development agenciesand international conservation organizations.Specifically, research should identify:i) the role of biodiversity conservation inrisk management, system resilience andlivelihoods; and ii) native tree species thatare valuable but under threat, as prioritiesfor domestication. Location-specificresearch should be done in a way thatcontributes to the identification and negotiationof local solutions. The researchprogramme should cut across and integratethe four themes of ICRAF:• Land and People – concerned with below-groundbiodiversity and the ecosystemservices of biodiversity;• Trees and Markets – concerned withon-farm biodiversity and tree domestication;• Strengthening Institutions – concernedwith building capacity in biodiversityassessment and domestication; and• Environmental Services – concernedwith landscape-scale biodiversity.ICRAF should also be an information clearinghouse,providing storage, organizationand management for enhancing user access.The Centre should promote and convenenon-conventional partnerships:• be an active participant in research aswell as a convenor and catalyst of priorityresearch;• catalyse a Working Group on Biodiversitythat draws together diverse scientificperspectives (agriculture, conservation,development and social science) withinCGIAR and the international conventions;and• continue to be an active member inEcoagriculture Partners.ICRAF should play a vital role in raisingawareness and education, by:• catalysing development of educationalmaterials;• raising community-level awareness tostrengthen local processes; and• participating in sensitizing donors.Major ICRAF outputs for the next10 years in biodiversity:• Conduct a review and develop an actionplan for a strategic research programmethat moves the agricultural biodiversitydebate beyond landraces to ecosystemstructure and function at the landscapelevel.• Support the development of nationalpolicies that take account of the potentialcontributions of agroforestry to biodiversity.• Support landscape initiatives in highprioritylocations with respect to bothbiodiversity and agroforestry.Climate change: climatechange mitigation andadaptation for ruraldevelopmentProblems and challenges relatedto climate changeClimate change will have consequences forfarmers all over the developing world. Theseeffects will play out over long periods oftime, while farmers and policy makers tendto have relatively short planning horizons.There is increasing interest in the idea thatagroforestry can contribute to the mitigationof net greenhouse gas emissions. Thisapproach uses mechanisms that offset continuedcarbon emission in more-developedcountries with carbon sequestration indeveloping countries. The Clean DevelopmentMechanism (CDM) of the KyotoProtocol on Climate Change is one suchagreement that most countries around theworld have already signed. In addition,there has been a mushrooming of similarvoluntary mechanisms outside the CDM.An initial assessment of the potential foragroforestry to contribute to greenhouse gasmitigation was included in a crucial reportof the Intergovernmental Panel on ClimateChange. However, that assessment wasbased on a very thin set of biophysical data.Thus the effects of agroforestry on the emissionof all relevant greenhouse gases needsto be investigated thoroughly, including thegathering of multiyear, multisite evidence.To date, the various discussions about thepotential of agroforestry to contribute tocarbon sequestration have been kept separatefrom discussions of the environmentalconsequences of tree planting and its effecton the livelihoods of the poor.Roles for ICRAF in climate change:• Create a regional project to networknational agriculture research systems,environmental research institutes andcentres of the CGIAR to assess the strategiesthat farmers and communities useto cope with multiple stresses.• Use hotspots of agroforestry adoption aslearning sites for environmental serviceassessment.• Link crop models to climate-changemodels.ICRAF outputs for the next 10 yearsin climate change:• Act as a broker, provider and facilitatoron information on climate change toenable debates at various levels (for example,by supporting farmers to attendinternational global forums).

126World Agroforestry into the Future• Conduct measurements in different agroecologicalzones and establish the role ofagroforestry in mitigation and adaptation.Environmental governance:policies to harmonizeenvironmental stewardshipand rural developmentProblems and challenges inenvironmental governanceAgroforestry is a cross-cutting and integrativeland use. This means that it is affectedby a wide variety of sectoral policies, especiallythose regarding forestry, land, water,agriculture and environment. It also meansthat it can be challenging to link agroforestrywith processes for planning andfinancing of rural development and povertyalleviation.Better analytical tools are needed for environmentalgovernance. This includes toolsthat assist with understanding and strengtheningcollective action and property systemsat multiple scales. It also includestools that clarify the links between environmentalservices, governance and servicereward themes.It is important to recognize that environmentalgovernance incorporates formalgovernment structures as well as locallybased and community-based natural resourcemanagement systems.Major roles for ICRAF in environmentalgovernanceICRAF must play a leading role in showinghow agroforestry is affected by sectoralpolicies and regulatory frameworks. TheCentre can take the lead in promotingintegrative approaches, for example inthe way that we engage with investors inagroforestry. Furthermore, it can help governmentsto cut through the complexityof environmental governance by fosteringprinciple-based and evidence-based environmentalmanagement at multiple scales.ICRAF should continue to catalyse importantpartnerships that create links acrossagriculture, the environment, forestry andgovernance [for example, by linking withthe environmental service and governanceprogrammes of the Center for InternationalForestry Research (CIFOR)].Priority outputs for ICRAF for thenext 10 years in environmentalgovernance:• Gather a critical mass of national andregional expertise in effective methodsfor multistakeholder assessment andnegotiation support.• Empower vulnerable groups of indigenouspeople who rely on agroforestry bygiving them more secure property rights.• Develop, test and disseminate policyand institutional options for promotingthe role of agroforestry in enhancing thelivelihoods of smallholder farmers, whileat the same time meeting environmentalobjectives.• Encourage governments in prioritycountries to enact changes in environmentalgovernance that recognize themultifunctional nature of landscapes.• Contribute to the modification of internationaland regional conventions,agreements and action plans to helpsmallholder farmers practise agroforestry.

Strengthening Institutions

“Agroforestry can and does playa major role in halting and evenreversing the decline in agriculturalproduction. It can alsohelp to increase food security,reduce poverty and malnutrition,improve health and reduce landdegradation. Agroforestry educationtherefore represents a directpathway to reducing Africa’spoverty and increasing fertilityand agricultural production.”Chivinge

Keywords:Sustainability, core values, integrated,commitment, youth, practicalChapter 15Sustainable agriculture and rural development– a response to the realities of rural AfricaTony Dolan, Baraka Agricultural College, KenyaAbstractOne of the main challenges for African agricultural education, training and extension is to devise systemsthat will give smallholders and pastoralists the hope, self-confidence, knowledge and skills theyneed to play a central role in the development process. In most African countries, agriculture is themajor source of livelihoods, and current standards of agricultural training must be improved if we are toachieve the goal of sustainable development. Many systems have been tried, from the ‘bwana shamba’(government extension agent) in each village, to the World Bank’s ‘training and visit’ system. All havemet with only limited success. Baraka Agricultural College is a small private institution, owned by theCatholic Diocese of Nakuru. Our mission is to promote sustainable agriculture and rural development(SARD) in eastern Africa through dialogue, participatory education, training, research and extension.Agroforestry is one of the non-negotiable programmes for sustainable agriculture and Baraka is anardent promoter of the practice. Bottom-up development is the approach, and the aim is to promotedevelopment through participatory action research. Collaboration and partnership are the two fundamentalvalues underpinning the work of the College. Baraka’s experience has shown that the SARDstrategy works. We believe that it should be adopted as a tool for rural development throughout Africa.IntroductionBaraka Agricultural College is a small private institutionsituated in the highlands of Kenya, 7 km from Molotown. Owned by the Catholic Diocese of Nakuru, theCollege is managed by Franciscan Brothers. The collegewas established in 1974 to respond to the educationand training needs of the newly settled farmers of RiftValley Province. Initially, these needs were met byproviding short courses in close collaboration with theMinistries of Agriculture and Livestock. However, theproblems of running an institution on short coursesalone soon became apparent. It proved difficult tocreate a learning atmosphere and maintain staff motivationbecause of a lack of continuity in attendanceof course participants. To solve this problem, we introduceda 1-year course in general agriculture aimedat school leavers entering farming. The curriculum forthe new course was derived mainly from Europeanagricultural colleges, with the emphasis being on highexternal input/high output agriculture.By the mid-1980s, it had become clear that the collegecurriculum was not supporting the needs of increasingnumbers of smallholder farmers. An international workshopon sustainable agriculture was held at the collegein 1986, and this proved to be a turning point in theBaraka approach. After this, we focused on refiningthe concept of sustainable agriculture and develop-

130World Agroforestry into the Futureing curricula and technologies to promoteit. The concept of rural development wasadded in the early 1990s, when it becameapparent that efficient farming alone wouldnot solve the economic and social problemsof rural eastern Africa.Conscious of the need to ensure that thefocus of Baraka remained on primaryproducers in rural communities, a strategicplanning process was started in early2003. This involved intensive work by collegemanagement, staff and many partnerorganizations, including Kenya’s Ministriesof Agriculture and Education, Science andTechnology. The resulting strategic plancharts a clear path for the future. BakaraCollege will continue its role of strivingto empower rural communities throughpromoting the concept of sustainable agricultureand rural development (SARD).This chapter is based on the practical experienceof Baraka College over the past 29years, hence we do not refer to the literature.Box 1. Baraka Agricultural CollegeOur Vision: ‘Rural communities wherepeople live in dignity and harmony withtheir environment and God’.Our Mission: ‘To promote sustainableagriculture and rural development ineastern Africa through dialogue, participatoryeducation, training, research andextension’.Dependency on agricultureA cursory social analysis of the countriesof eastern Africa gives a clear picture ofthe social, economic, environmental andpolitical realities of the region. Seventy percentof the population lives in rural areasand 60 percent of rural inhabitants live inabsolute poverty; the physical and socialinfrastructure is very poor; education levelsare low; government development servicesare inadequate and not likely to improve;government decision making is influencedby a small percentage of the population;cultural values are strong, but cultural practiceshave not adapted to changed realities;the informal education system has beenneglected; and there is over-emphasis onacademic university education and a correspondingneglect of the middle-level educationinstitutions. The hasty dismantling ofState Marketing Boards and other essentialrural development services has left smallholderfarmers and pastoralists with littlepractical State support. The nurturing of dependencyand the neglect of the informaleducation system has contributed to a lackof organization by rural people and theygenerally lack the confidence to contributeto or influence the national and internationalpolicies that affect them.Yet according to government planning, thesmallholder–pastoralist sector will becomemore important in providing employmentand creating wealth in the future. Table 1clearly shows the scale of the challenge.By 2008, the small-scale agriculture sectoris expected to directly absorb 10.09 millionof an estimated 18.05 million labour force.By that time, the formal sector of the economywill employ a mere 1.83 million people.Baraka Agricultural College is convincedthat SARD is the most appropriate responseto the economic, social, environmental andpolitical realities of eastern Africa.Sustainable agriculture andrural developmentSARD is a development concept that promotesefficient, environmentally friendlyfarming and integrated rural development.Sustainable agriculture is concerned witheducating and training farmers, studentsand development workers to look onfarming as an environmentally friendlybusiness. The emphasis is on using locallyavailable resources and technologies.Criteria in all decision-making processesinclude the economic, environmental,social and cultural impacts of such decisions.The concept includes activities outsidethe farmgate, including the need forcooperation with neighbours, for example,self-help groups, marketing associations,cooperative societies or other farmer-representativestructures. From the farmer’sperspective, for a farm to be sustainable,it must satisfy the food security, cash andenergy needs of the family depending on it.Therefore, home management, social andcultural issues also need to be addressed.Sustainable agriculture aims to promote efficientfarming, but no matter how efficient,farming alone will not eradicate poverty.Rural development encourages the integrateddevelopment of rural areas throughadding value to farm produce, improvingthe efficiency of the informal industrial sectorand attracting local and foreign capitaland expertise to invest in rural areas, sobringing jobs to the people. Rural developmentis also concerned with improving educationand medical services, security andrecreation facilities, and generally makingrural areas more attractive to live in.Baraka College takes community developmentas a prerequisite for the success ofSARD. Youth development, gender, theenvironment and HIV/AIDS are some of thecrosscutting issues that have to be addressedwithin the SARD concept. We emphasizethe value of informal education and trainingsystems and promote integral developmentof the individual and resources.

Chapter 15: Sustainable agriculture and rural development131Table 1.Projected employment in Kenya.Numbers employed (millions)Box 2. The ANAFE approach – anillustrationSectorFormal agricultureSmall-scale agricultureRural informalUrban informalFormal non-agriculture2000 2008 (projected)0.348.411.582.851.450.3510.092.073.711.83Total labour force 14.63 18.05Source: Republic of Kenya National Development Plan 2002–2008.Principles of SARDThe fundamental principles underpinningour work are those of collaboration andpartnership. As a private institution, Bakarabelieves that it should be innovative andprepared to takes risks in experimentingwith new ideas, approaches and technologies.It is appreciated that the private sectorhas the responsibility to be at the cuttingedge in order to be competitive. Partnershipwith the Kenya Ministry of Agriculturehas been a major factor in the success ofBaraka. We also collaborate closely with theKenya Agricultural Research Institute (KARI),Egerton University, the African Network forAgroforestry Education (ANAFE) and manynon-governmental organizations (NGOs)and churches. Partnership with externalagencies is vital, as it brings a shared experienceaccumulated by partner agenciesfrom other parts of the world (and crucialfinancial support). Baraka has worked inclose partnership with agencies such as theGerman Catholic Bishop’s Organization forDevelopment Cooperation (MISEREOR),the Overseas Development Agency of theCatholic Church in Ireland (Trócaire), theFreedom from Hunger Council of Ireland(GORTA), Self Help Development Internationaland the Government of Ireland’sAgency for Personal Service Overseas(APSO). Advice and support from theseagencies has made an important contributionto Bakara's success.However, partnership and collaborationwith local and external agencies is only ofvalue when the rural community is put atthe centre of the equation. As Baraka seesit, the challenge is to create a learning community.That is, to create ‘mass action’, sothat development momentum is generatedand the community – as a corporate entity– gains the capacity to continue developingwhen the project or programme is over.No single methodology will achieve this. Ingeneral, the Baraka approach is one of ‘participatoryaction research’. This approachputs the needs of the community at thecentre, and all development agencies workingwith the community are encouraged torespond to these needs in a spirit of collaboration.Rather than giving directives on whatis right or wrong, communities use theirown experience to develop solutions, withadvice from the professionals. These are theprinciples followed by ANAFE, of whichBaraka is a member. The establishment ofANAFE is having a very positive impact onthe promotion of sustainable developmentin Africa through supporting third-level educationand training institutions.When a Baraka extension worker wasdiscussing the value of recycling organicmatter on the farm with a group of farmers,one of the participants suggested that,from his experience, he got a much bettercrop of maize after burning the stalks ofthe previous year. Instead of saying ‘youmust be wrong!’ the extension workerasked the individual and the group to reflectand encouraged them to researchwhy this might be so. She used the opportunityto develop a better understanding ofthe many aspects of soil fertility.The role and approach of the trainer, extensionworker or researcher is vital.There is no room for the so-called ‘expert’coming in with pre-packaged solutions.Technology alone will not solve developmentproblems. There is a need for aholistic approach and direct involvementof the community – and the developmentof community leadership and selfconfidenceis central to the process. In ourwork we must realise that development isa slow process and that no two communitiesare the same.Baraka programmesBaraka College runs six programmes all ofwhich aim to improve the development capacityof rural communities.• Certificate in sustainable agriculture andrural development (CSARD)• Diploma in sustainable agriculture andrural development (DSARD)• Short courses• Beekeeping development programme• Area-based development programme• Day-release courses.

132World Agroforestry into the FutureThe CSARD course is a 16-month programmeand the college has the capacityto take 80 students each year. Participantscome from Kenya, Sudan, Tanzania andUganda. The curriculum has been developedover the years to meet the needs ofparticipants and their communities. Thecourse aims to empower communities fromwithin, by encouraging them to nominateintelligent women and men who havedemonstrated a commitment to communitydevelopment. A scholarship scheme, sponsoredby GORTA, allows students fromsome of the poorest regions to participate.Starting in 2006 Baraka College will offera diploma course in sustainable agricultureand rural development (DSARD). This willbe a full-time residential eighteen-monthprogramme offered at Baraka. The courseis aimed at certificate in agriculture/ruraldevelopment graduates who a have a minimumof two years experience working withrural communities.Participants know that there will be very fewopportunities for employment with NGOs,government or church organizations. However,Bakara realizes that graduates of thiscourse will have problems satisfying theiraspirations for a better quality of life inrural communities. Consequently, we havea support structure for past students, theBaraka Agricultural College Old StudentsAssociation (BACOSA), which gives practicalsupport and includes a saving and creditscheme. In addition, the College has designedspecial short courses in ‘starting yourown business’ and ‘participatory projectplanning’ for past students of the CSARDcourse. We believe this is a cost-effectiveand efficient way of improving the developmentcapacity of rural communities.On average, 500 farmers and developmentworkers complete one-week coursesat Baraka every year. The content of thesecourses ranges from organic farming to e-business. The emphasis in all short coursesis on the participant acquiring practicalskills that she/he can apply on their farm orin their community or small-scale business.The beekeeping development unit providestraining and extension services in all aspectsof beekeeping, an enterprise that has goodpotential to create sustainable livelihoods,especially in arid and semi-arid areas.The area-based development programmeis concentrated in the Kamara and Laredivisions of Nakuru district. This programmeapplies the principles and methodsof SARD used in all other Bakara programmes.It also demonstrates that theconcept works and provides a practicaltraining ground for course participants.Day-release courses are aimed at teachers,school and college students and membersof the general public. The objective is topromote the SARD concept among thegeneral public.Sustainability of BarakaCollegeIncome generating units (IGUs) are animportant part of our strategy to develop asustainable college. Baraka has four suchunits: the farm, workshop, highlands honeyand shop. Until 2002, these units were rundirectly under college management andwere an overall financial liability to thecollege. In 2002, they were separated fromdirect college management and set up asan independent department. Efficiency hasimproved and the units are now contributingapproximately 5 percent of the collegebudget. The aim is to increase this to 20percent by 2007. It is clear that managementof IGUs demands a very differentmanagement style from that required forsocial service provision!However, no matter how successful localfinance generation is, it is evident thatfinance from external sources will have tobe provided if institutions such as Barakaare to fulfil their missions. In developingcountries, in the absence of direct Statefinancial support, this funding will haveto come from external sources; hencethe need for loyal long-term externalpartners.ConclusionThere are several lessons learned fromBaraka’s experience that have broaderapplication in the challenging work ofbringing smallholders and pastoralists intothe heart of the development process. Froma global perspective, theorists and policymakers must respond to realities – and doso quickly. The symptoms are all there –international terrorism, ecological disasters,uncertainty over energy supplies, disease,urban violence, unhappiness in the developedworld, especially among youth – andthe paradigm that is responsible for thesesymptoms needs to be challenged andalternatives proposed.In Baraka’s view, SARD is a conceptthat has the potential to achieve thesustainable development of the Africancontinent. Sustainable agriculture, properlyunderstood and applied by Statepolicy, will promote economically efficient,environmentally friendly farming.Rural development requires strongnational and international policies andsupport and, in rural communities, workis needed to create the preconditions fortake-off. However, with current developmentsin information and communicationtechnology, there is no reason why

Chapter 15: Sustainable agriculture and rural development133environmentally friendly rural industrializationcannot be promoted.The uptake of SARD in rural communitiesrequires close collaboration betweeneducation, extension, training and researchservice providers. The objective of creatinga learning community through participatoryaction research also requires the collaborationof line ministries, such as Health,Water, Community Services, etc. Too manygovernment development programmesare donor-dictated and short-term. Thereis a temporal infatuation with the latestparticipatory approach and time is seldomallowed for any of these systems to make alasting impact. Rural communities requirecommitted, dedicated and available developmentworkers who have vision and whosee their profession as a vocation rather thanjust a way to make a living.The experience of Baraka AgriculturalCollege shows that, even in times oflimited resources, there are innovativeways of providing rural communities withthe education, training and extensionservices necessary to achieve sustainabledevelopment.

Keywords:Educational impact, sustainability, rural development,farmers of the future, environmental policy, responsiveentrepreneurship, integrated approachesChapter 16Capacity building in agroforestry in Africa andSoutheast AsiaO.A. Chivinge, University of ZimbabweAbstractSince its formation in 1978, the International Centre for Research in Agroforestry (now the WorldAgroforestry Centre) has generated a lot of general knowledge on agroforestry. The need to advancethis knowledge for specific geographical locations led to the formation of the African Network forAgroforestry Education (ANAFE) in 1993 and the Southeast Asian Network for Agroforestry Education(SEANAFE) in 1999. Since then these networks have built condsiderable human capacity and developedregion-specific material. Many lecturers in agriculture, natural resource management and forestryhave been trained in aspects of agroforestry through short-term training programmes and workshopsorganized by ANAFE and SEANAFE. Capacity building has taken the form of training of trainers, curriculadevelopment, writing teaching materials, postgraduate scholarships for thesis research, staff andstudent exchange programmes, student attachment, establishment of demonstration plots, farmers ofthe future (FoF) programme (aimed at young school children), short courses and workshops for policymakers, farmers and education–research and extension links. As a result of these efforts, many universitiesand colleges are now teaching agroforestry either as part of a course or as a full course at diploma,undergraduate and postgraduate levels. However, there are still many challenges to capacity buildingand dissemination of technologies that need to be met in order to have adequate human resources tohelp reduce poverty and food insecurity, improve people’s livelihoods and reduce malnutrition. Theseinclude more capacity building, continuous review of curricula, establishing agroforestry courses in allinstitutions offering agriculture and natural resource management, securing more financial resources,and linking education with research and development to continuously build capacity in agroforestry.IntroductionAgroforestry can and does play a major role in haltingand even reversing the decline in agricultural production.It can also help to increase food security, reducepoverty and malnutrition, improve health and reduceland degradation. Agroforestry education thereforerepresents a direct pathway to reducing Africa’s povertyand increasing fertility and agricultural production(Kung’u and Temu 2004). In their 2002 paper, Chivingeet al. emphasized the need and role of agroforestry capacitybuilding at the regional level, while the situationat continental level is given by Temu et al. (2003).Since the formation of ICRAF in 1987, much useful andrelevant knowledge and information on agroforestryhas been developed. However, there were limitationsto how this information could reach relevant stakeholdersaround the world. Hence in April 1993 the

136World Agroforestry into the FutureAfrican Network for Agroforestry Education(ANAFE) was established by African collegesand universities teaching agricultureand natural resource sciences, and in 1999Southeast Asian colleges and institutionsfollowed suit by establishing the SoutheastAsian Network for Agroforestry Education(SEANAFE; Temu 2003). At that time, theseinstitutions were not offering courses orprogrammes in agroforestry, and agroforestrywas not part of the curriculum. However,they realized that agroforestry has a criticalrole to play in agriculture and natural resourcemanagement (NRM), and their leaders– including vice-chancellors, deans offaculties and chairpersons of departments– decided to revise curricula to include it.ANAFE and SEANAFE’s objectives includestrengthening agroforestry and NRM educationat African and Southeast Asian universitiesand colleges; linking education withresearch and extension; improving availabilityand access to multidisciplinary land-useeducation; creating awareness among policymakers of the role of agroforestry and NRM;enhancing the availability and quality ofteaching, which includes human resources,teaching material and facilities; and linkinginstitutional teaching and land-use educationthrough networking. These organizationstarget higher educational institutionsincluding universities, colleges and researchinstitutions as well as secondary and primaryschools, local farmers and farmingcommunities, and policy makers.Since their foundation, ANAFE and SEAN-AFE have become increasingly effectiveas agents for advancing integrated naturalresource management (INRM) educationin Africa and Southeast Asia, and are alsostrategic vehicles for sharing agroforestryand NRM knowledge. As one of its firsttasks, ANAFE undertook a study of thecurrent educational technologies offered byits members and came up with a strategyto improve teaching and learning, includingapplication of modern informationand communication technologies (ICTs).As a result, both ANAFE and SEANAFEhave strengthened training of trainers,development of teaching materials andtraining facilities. Now, the academicweight of agroforestry ranges from merelyfeaturing as a topic in a course to being afull programme at diploma, undergraduateand postgraduate levels. The sheernumber of workshops and meetings heldon the subject have led to policy issues onagroforestry being implemented in manyinstitutions. All universities offering postgraduateprogrammes in agroforestry arebeing monitored, and the networks assist inprogramme and content design to ensurequality capacity building.Why educate in agroforestry?Agricultural production in some areas ofAfrica and Asia is in decline, leading tofood insecurity, increased poverty, malnutritionand poor health. The Green Revolutionemphasized monocropping and reliedon the use of machinery and syntheticchemicals; completely neglecting the conceptof smallholder farming. In many partsof the world, this led to natural resourcesbeing destroyed and land degraded, resultingin environmental problems anda decrease, rather than the intended increase,in crop yields and other agriculturalactivities. Soil fertility still declines inmany countries therefore agroforestry isan important tool for rural farmers to usein successful agricultural production andland conservation. Prior to the formationof ANAFE and SEANAFE, it was clear thatfarmers were not maximizing the opportunitiesfor agroforestry. INRM is criticalfor successful agricultural production. Furthermore,agroforestry products need to bemarketed, and hence product processingand adding value are also critical steps inthe process. It was therefore essential to includeagroforestry education at certificate,diploma, undergraduate, masters and Ph.D.levels, and for farmers, extension/developmentworkers and policy makers to beaware of it.AchievementsBoth ANAFE and SEANAFE are active inagroforestry capacity building in Africanand Southeast Asian educational institutions,and in raising awareness among policymakers, extension staff and farmers, andat community levels. The number of institutionsteaching agroforestry at various levelsis increasing. Currently 123 universitiesand colleges in 35 countries under ANAFE,and 76 institutions in five countries underSEANAFE are offering agroforestry courses(ICRAF 2003). ANAFE and SEANAFE havehad many notable achievements, includingtraining of trainers on how to develop extensionmaterials, curricula development,preparation of agroforestry teaching materialsfor universities and colleges, establishmentof demonstration plots at teachinginstitutions and on farms, agroforestrytheses at the B.Sc. and postgraduate levels,and provision to the institutions of agroforestrypublications. ANAFE and SEANAFEhave both produced newsletters to shareinformation on agroforestry education andnetworking.Training of trainersVarious training-of-trainer courses andworkshops have been conducted byANAFE and SEANAFE in different countriesand institutions, resulting in about 300

Chapter 16: Capacity building in agroforestry in Africa and Southeast Asia137lecturers being trained or retooled in suchaspects of agroforestry as curricula developmentand review, and writing agroforestrymaterials (Table 1).Through fellowships provided by ANAFE20 agroforestry monographs have beenwritten by various teams, for the benefit ofthe region. To date 100 lecturers have attendedcurricula-development workshopsunder SEANAFE, where teachers weretrained in agroforestry theory and practicesand pedagogic skills and took courses ontranslating material into local languages.Furthermore 25 faculties from universitiesand colleges have attended trainingin participatory on-farm experimentation.Short-term courses have been conducted toenable lecturers to update the agroforestrycurricula in both Africa and Southeast Asia.SEANAFE has held participatory curriculadevelopment courses with 20 lecturers ineach of the countries it covers. Through scientificwriting workshops, lecturers are beinggiven the opportunity to develop theirown writing skills. Some short-term coursesfor farmers/extension staff are being heldin collaboration with ICRAF. Under ANAFE152 farmers have received short-termtraining on various aspects of agroforestryin universities, colleges and vocationalinstitutions.Curricula developmentUnder SEANAFE and ANAFE, many universitiesand colleges have developed and updatedagroforestry curricula at certificate,diploma, graduate and postgraduate levels.ANAFE has conducted many workshopsfor curricula reviews incorporating agroforestryand INRM, reaching a total of 154people between 1993 and 2002 (Table 2).Currently, 31 colleges and universities inthe Philippines offer B.Sc. degrees in agro-Table 1. The number of people who attended ANAFE curricula review 1993–2002.Discipline/Level Certificate Diploma First degree Postgraduate TotalAgriculture 2 4 15 2 23Forestry 7 8 6 2 23Other (e.g. rural development)New agroforestryprogrammes1 2 3 0 60 4 5 6 15Total 10 18 29 10 67Table 2. Number of trainers/lecturers trained between 1993 and 2002.Level/Discipline Agriculture Forestry Other TotalUniversities 47 17 33 97Colleges 26 18 10 54Vocational institutes 2 0 0 2Extension organizations 43 5 8 56Total 118 40 51 209forestry. In Africa, more than 30 universitiesoffer agroforestry as a thesis research area,and 20 universities offer postgraduate programmes.One of the best new products onoffer is the new agroforestry diploma fromNyabyeya Forestry College in Uganda,which many other institutions are nowadopting in part or in full. In terms of M.Sc.availability, a regional agroforestry curriculumguide was produced and distributedto all member institutions under SEANAFE,which should lead to M.Sc. curricula beingdeveloped. SEANAFE, like ANAFE, hasmade relevant teaching material availableto institutions. It has also produced teachingmaterial and manuals in local languages,which are now available to mostinstitutions. Both networks have producedteaching and field-based materials andmanuals at various levels.In both Africa and Southeast Asia, agroforestryhas been incorporated in educationand training programmes at unprecedentedrates; many undergraduate and postgraduateprogrammes in agriculture, forestry andNRM now offer agroforestry as an area ofspecialization. In collaboration with ICRAF,many institutions are developing andimplementing programmes on outreachactivities, and there is also a focus on interdisciplinaryapproaches in agriculture and

138World Agroforestry into the FutureNRM. New topics are being added to theagroforestry curricula that place emphasison indigenous trees and crops, value addition,processing and marketing that arecritical areas for rural communities.Teaching materialdevelopmentTo date, ANAFE has distributed many publicationsto its member institutions, including23 book titles, various course and workshophandouts, and other materials, mostlyused by ICRAF in training courses. Eightfocal institutions have received multiplecopies of publications, books and equipmentto support training. Together with theRegional Land Management Unit (RELMA),the Centre for Agricultural and BiosciencesInternational (CABI) and others, ANAFE hassupplied books, manuals, CD-ROMs andother teaching material in many areas ofagroforestry. In Asia, 46 publications fromSEANAFE, ICRAF and the Food and AgricultureOrganization of the United Nations(FAO) have been made available to allSEANAFE member institutions. SEANAFEhas also developed an activities handbook,agroforestry teaching materials for differentlevels, and short-term teaching materialsand brochures, all of which have been distributedto relevant member institutions.Demonstration plotsIn order to capture farmers’ indigenouspractices for part of the curriculum, ANAFEand SEANAFE have funded demonstrationplots at institutions and on farms aroundthe region, providing various technologiesto strengthen teaching and income generation.ANAFE has also held agroforestryfarmers’ field days at colleges, universitiesand on farms. Sixteen institutions inAfrica and Southeast Asia have developedagroforestry demonstration plots, related totheir curricula and specific to local problems.These plots have been set up specificallyto help students, extension staff andfarmers.On-farm demonstration plots can be usedfor both teaching and capacity building forextension services, while Kenyan on-farmagroforestry trials in fertility have beenestablished to benefit farmers (Ngumi etal. 2004). Through such field days, farmershave been exposed to and learned muchabout available agroforestry technologiesthat can result in improved and sustainablecrop production. More farmers are nowemploying the technologies they find appropriatefor their communities and socioeconomiccircumstances. Seventy four percentof the farmers who visited field daysare now using the technologies. This hasresulted in maize grain yields of 1.7–4.8 tha –1 , compared to 1.5 t ha –1 where no agroforestrytechnologies are applied (Ngumi etal. 2004).Institutions use demonstration plots forpractical aspects of their courses and to fulfiltheir mission – to build human capacity inresearch for sustainable economic development,poverty reduction and improved foodsecurity. Some also use them for in-servicetraining and research; conservation and domesticationof high-value trees and shrubs;and for growing herbs as medicinal plants.Postgraduate fellowshipsAs part of capacity building, many postgraduatefellowships have been offered tovarious institutions in member countries.Since the formation of ANAFE 160 postgraduatefellowships covering educators,researchers and development workers havebeen awarded: 136 provided by ANAFE(103 males and 33 females). In SoutheastAsia, in many activities, there is equalrepresentation of men and women. Whilethere are teaching staff at various levelsin agroforestry, there are not yet sufficientnumbers to teach the large number of students,although scientists and developmentspecialists are now available.Agroforestry graduates work with variousorganizations including educational institutions,governments, international organizationsand developmental organizationspromoting agroforestry. Their theses havecontributed significantly to agroforestryscience, knowledge and capacity buildingat various levels, and have contributed toAfrican and Southeast Asian farmers’ improvedagricultural production.Student and staff exchangeANAFE has funded staff and student exchangeprogrammes with many institutionsand ICRAF research institutes. Throughthese human resource exchange programmes,the network has been able tomake capacity available that can be deployedwhere the greatest needs are found(Zoungrana et al. 2004). As a result of theincrease in human resource capacity, manyinstitutions have developed a critical masscapable of delivering agroforestry courses.SEANAFE and ANAFE sponsorship has enabledmany students to do M.Sc. or B.Sc.theses in agroforestry. Staff exchanges alsoallow the networks to share knowledge andskills in curriculum development, trainingof teachers and development of agroforestryteaching materials.Education research extensionlinksThrough the Regional Agroforestry TrainingTeam (RAFTT) meetings, workshops andfield days for farmers at universities, collegesand farms, considerable capacity has been

Chapter 16: Capacity building in agroforestry in Africa and Southeast Asia139built at farmer and extension levels. Someinstitutions have developed brochures foruse by extension staff and farmers. In bothANAFE and SEANAFE, some training materialshave been developed in local languages,thereby helping many farmers. Programmesdeveloped by colleges and universities haveresulted in farmers and extension staff beingtrained in various aspects of agroforestry aswell as the development of training materialsand guidance on methodology in adultlearning (andragogy) as short-term courses.Extension staff have been taught to teach incolleges.Farmers of the futureThe FoF programme, established byANAFE, aims to teach those school pupilswho will become farmers without a universityor college education. The aim of theFoF programme is to build capacity withpolicy makers and educators to implementagroforestry in primary and secondaryschools. Teachers are being taught how todevelop teaching materials. This initiativeis very important as it connects the agroforestryeducation continuum from primaryschool to university level.ChallengesDespite the successful capacity buildingto date, there are still many challenges relatedto agroforestry education. Changes ineducational programmes take a long timeto produce visible results, and it is importantto continue to address critical NRMand environmental issues through capacitybuilding in agroforestry.Although agroforestry is now accepted asdiscipline in many institutions of higherlearning, transferring technology to farmersand stakeholders is still challenging. Somelecturers lack the skills to develop teachingmaterial in a pedagogically acceptablemanner. Others lack the ability to mergetheory and practices – an essential elementin agroforestry. Capacity at the localand institutional levels needs to be built inorder that more sustainable and profitablelivelihood options are created. High-qualityresearch, education and extension advisoryservices are therefore key elements inbuilding up capacities of farmers and communitiesin adopting agroforestry technologies.Capacity building in agroforestry is aviable pathway to integrate rural developmentand hence should be given priority.One of the current limitations in agroforestryresearch and development is thesmall number of scientists and professionalswho are planning to undertake researchin multidisciplinary land-use programmes.The challenge facing African institutionsof higher learning is to build agroforestryprogrammes that are adequately integrative.Furthermore, many of the scientifictheories and practices generated throughstudent theses need to be transformed intoappropriate educational materials. There isa shortage of curricula and teaching materialin institutions in some regions, suchas West Africa, yet many lecturers lack theskills to write appropriate educational material.Existing materials in English shouldbe translated into Portuguese and French.All of these activities require funds for capacitybuilding, especially for student agroforestryresearch theses.The futureGood agroforestry education must involvepeople from environmental science,forestry, agriculture, rural development,social sciences, and veterinary science.It is important to ensure that all relevantstakeholders have access to agroforestrytechnologies, and that agroforestry materialdevelopment takes account of the differencesin each region. Adequate trainingmodules for extension staff/developmentalworkers should be made available, andthere should be incentives to keep the capacitydeveloped within each region. It isessential that appropriate agroforestry technologiesbe fully utilized by the relevantstakeholders.Training on later stage processes, such asfood processing and marketing of agroforestryproducts should be implemented inthe curricula. The role of agroforestry inalleviating the effects of HIV/AIDS shouldbe emphasized, and ways should be foundto reduce the human capacity lost throughthese diseases.In education, institutions not teaching agroforestryshould be encouraged to do so.There is also a need to build capacity at local/communitylevels on technical aspectsof agroforestry scaling up. For the young,the FoF programme is going to be increasinglyimportant, as children will be theones to promote agroforestry techniquesand knowledge in rural communities in thefuture.AcknowledgementWe are thankful to the Swedish Agency forInternational Development Cooperation(Sida) for ANAFE and SEANAFE activities.

140World Agroforestry into the FutureReferencesChivinge, O.A., J. Skagerfalt and A. Temu 2002.Ten years of capacity building in agroforestryin the southern Africa region. Paperpresented at the ICRAF Southern AfricaWorkshop 22–27 July 2002. Waterbath,South Africa.ICRAF (International Centre for Research in Agroforestry)2003. 1999–2002 Final ProjectReports to Sida. World Agroforestry Centre,Nairobi, Kenya.Kung’u, J. and A.B. Temu 2004. The role of agroforestryeducation in sustaining society andenvironment. Agroforestry has the potentialto solve social and environmental problems.in: Temu, A.B., S. Chakeredza, K.Mogotsi, D. Munthali and R. Mulinge (eds)Rebuilding Africa’s capacity for agriculturaldevelopment: the role of tertiary education.ANAFE Symposium on Tertiary Education14–16 April 2003, World AgroforestryCentre, Nairobi, Kenya. World AgroforestryCentre, Nairobi, Kenya.Ngumi, N.G.N., H. Boga, H. Njenga, P. Maina,J. Muthanga and M. Kaibui 2004. AgroforestryDemonstration sites in tertiaryeducation institutes as a tool for capacitybuilding in agriculture and naturalresource management. in: Temu, A.B., S.Chakeredza, K. Mogotsi, D. Munthali andR. Mulinge (eds) Rebuilding Africa’s capacityfor agricultural development: the roleof tertiary education. ANAFE Symposiumon Tertiary Education 14–16 April 2003,World Agroforestry Centre, Nairobi, Kenya.World Agroforestry Centre, Nairobi, Kenya.Temu, A.B. 2004. Towards better integration ofland use disciplines in education programmes.in: Temu, A.B., S. Chakeredza, K.Mogotsi, D. Munthali and R. Mulinge (eds)Rebuilding Africa’s capacity for agriculturaldevelopment: the role of tertiary education.ANAFE Symposium on Tertiary Education14–16 April 2003, World AgroforestryCentre, Nairobi, Kenya. World AgroforestryCentre, Nairobi, Kenya.Temu, A.B., I. Mwanje and K. Mogotsi 2003. ImprovingAgricultural and Natural ResourcesEducation in Africa: A Stitch in Time. WorldAgroforestry Centre, Nairobi, Kenya.Zoungrana, I., K.K. Mogotsi, A.B. Temu and P.Rudebjer 2004. Promoting INRM trainingand education through networks. inTemu, A.B., S. Chakeredza, K. Mogotsi, D.Munthali and R. Mulinge (eds) RebuildingAfrica’s capacity for agricultural development:the role of tertiary education. ANAFESymposium on Tertiary Education 14–16April 2003, World Agroforestry Centre,Nairobi, Kenya. World Agroforestry Centre,Nairobi, Kenya.

Keywords:Networking, research–extension–education links,agroforestry landscapes, training needsassessment, educational changeChapter 17Institutional collaboration in agroforestry:Networking and knowledge managementPer G. Rudebjer, World Agroforestry Centre, Thailand; Nguyen Van So, University of Agriculture and Forestry,Vietnam and John R.S. Kaboggoza, Makerere University, UgandaAbstractAs the focus of rural development has shifted from sector-based to integrated approaches, agroforestryhas emerged as a key element of integrated natural resources management (INRM). Integratedapproaches require effective links across the research, education and extension continuum. Thischapter discusses weaknesses in this continuum that lead to the unsatisfactory impact of investmentin agroforestry development, with emphasis on the education system. We describe the change fromopportunistic agroforestry teaching – often initiated by alumni of the World Agroforestry Centre’s trainingcourses – towards regional agroforestry education networks. The African Network for AgroforestryEducation (ANAFE) was formed in 1993 and the Southeast Asian Network for Agroforestry Education(SEANAFE) in 1999 to improve the access to and quality of agroforestry education. Activities includecurriculum development and review using participatory approaches, training of trainers, teachingmaterials support, graduate thesis research, information and policy advocacy.IntroductionA widespread traditional practice among farmers, agroforestryhas emerged as a science only during the past25 years. The education system has taken note andagroforestry courses are now offered widely in universitiesand technical colleges in Africa and Southeast Asia,as well as in Latin America and South Asia. Many universitiesin developed countries also teach the subject.Agroforestry science initially focused on classificationof agroforestry systems, intercropping research andthe development of agroforestry technologies – approachesthat were embraced by extension and educationsystems. But efforts to disseminate agroforestryas a technical package to increase food security andincome and to protect the environment showed mixedresults. Meanwhile, many traditional agroforestry practiceswere overlooked. Today, agroforestry science hasbroadened its scope and now includes multidisciplinaryresearch on landscape functions and the livelihoodsof people. One obstacle to developing agroforestryin an integrated, participatory and innovativemanner has been a lack of adequately trained agroforestryresearchers, extension specialists and teachers.Integrated natural resource management (INRM) isa new approach to agricultural research and developmentthat has emerged to address these complexinteractions. The INRM paradigm differs notably fromthe traditional crop improvement paradigm that wassuccessful in bringing about the Green Revolution(Izac and Sanchez 2001). INRM reflects the broad

142World Agroforestry into the Futureinteractions required to simultaneouslyreduce poverty, increase food securityand achieve environmental protection.It also recognizes ecological, social andeconomic interactions at different scalesin time and space (CGIAR 2000). Thesetrends influence development strategiesin Africa and Southeast Asia too, as theyevolve from sector-oriented towards integratedrural development. Agroforestrypractices play an important role in suchintegrated approaches to natural resourcemanagement.Complex problems require new organizationalforms for their solution. Interorganizationalnetworks among public, privateand grassroots organizations have emergedto meet this need (Boje and Wolfe 1989).In the 1990s, regional networks wereformed in Africa and Southeast Asia to improvethe access to and quality of highereducation in agroforestry. The two networksof universities and technical collegescontribute to educational change and tobuilding the capacity of present and futureagroforestry and natural resource managementprofessionals. This chapter showshow institutional collaboration, in the formof networking, can be a powerful tool formanaging knowledge about agroforestry,thus underpinning the complex processesof rural development.development efforts have often had weakor missing links in this continuum, whichled to unsatisfactory or sub-optimal impactof investments in agroforestry development.The situation that is yet to be fullycorrected, was characterized by:• poor adoption and slow scaling up ofagroforestry innovations;• technology oriented research and extensionand local knowledge not sufficientlyrecognized by research and educationinstitutions;• research results not effectively reachingor entering education programmes;• poor capacity among graduates to useparticipatory approaches in developingrural areas; and• government research, education andextension departments being locatedin separate units, resulting in poor linksand a fragmented approach.Underlying causesTo understand the reasons for the missinglinks, we need to look at each of the threecomponents in our model: education, researchand extension.Education was often not geared towardsdevelopment – this was supposed to betaken care of by the extension systemalone. Similarly, universities in Africa andSoutheast Asia often have weak researchprogrammes (due to, for example, low salariesand poor facilities). Thus, educationprogrammes often have limited researchor extension content (although there areexceptions). Theoretical bias is common:programmes are too academic and do nothave practical learning methods, makingit difficult for graduates to face realitywhen they enter the job market. At thesame time, lecturers tend to lack field-levelskills. Too many curricula have been developedin a top-down, programme-basedmanner, as opposed to participatory andneeds-based approaches. Finally, there isa lack of integration between universityand ministry structures (hampering interdisciplinaryeducation), and between researchersand educators.The research system generally showsweak links with education and extension.This is because: i) research and extensionin most countries are handled by separateinstitutions; ii) the research agenda is toonarrow (i.e. not systems-oriented) andfarmers do not participate sufficiently inidentification of research topics, the conductof research or feedback on results;Missing links in theresearch–education–extension continuumThe starting point of our discussionis a model of the linksbetween research, educationand extension – a continuumthat depends on, and interactswith, a range of otherstakeholders and the policyframework (Figure 1). RuralAppropriate policy(decision makers,administrators)Extension(extensionists, farmers)Education(lecturers, graduates)Figure 1. The research–education–extension continuum.Other stakeholders’influences (markets, traders,processing companies, etc.)Research(scientists)

Chapter 17: Institutional collaboration in agroforestry143and iii) there is a lack of interdisciplinaryteam work, leading to sub-optimal use ofexisting human resources and a lack ofsynergy. Research results are therefore notdisseminated effectively.The extension system also has severalbottlenecks that hamper the free flow ofinformation with research and educationalorganizations. These include: i) a hierarchicalextension approach, using one-waycommunication for spreading nationalpolicies and tending to overlook localknowledge and practices; ii) a focus ontechnologies that do not consider socioeconomicor cultural aspects; iii) weak institutionalsupport systems (e.g. resources,facilities, human resources, knowledge andskills) hamper the acquiring and sharingof knowledge; and iv) limited experienceand lack of capacity in using participatorymethods. In addition, cross-cutting policyand institutional factors influence the linksbetween research, extension and education.Policy makers are often not sufficientlyinvolved in the development process ata local level. Institutional structures do nothelp either; there are often several differentministries involved, which may or may notcollaborate.As a result of these bottlenecks and missinglinks, educational institutions face difficultiesin teaching subjects that requireinterdisciplinary skills and a good grasp ofcurrent research and extension paradigms.Examples of such complex areas are thelivelihoods in the ethnically diverse uplandsof Southeast Asia, the links betweenlocal land use and environmental services,or farmers’ postharvest processing andmarketing.Given the increasing interest in agroforestryeducation at tertiary level, how can individualinstitutions tackle the kind of issuesdiscussed above? They cannot change sucha complex situation alone. To have a strongvoice, they need to unite. Networkingamong universities and colleges has beenfound to be an effective tool.Development of regionalnetworksSeveral universities and colleges in Africaand Southeast Asia began to take an interestin agroforestry education during themid-1970s. This was triggered by populationincrease, rapid changes in land use(including extensive deforestation) andissues raised by the global society aboutsustainable development and the environment,such as widespread soil erosion andland degradation. Educational institutionswere also influenced by external factors,such as advances in international agroforestryresearch and development. Theprocess of strengthening capacity for agroforestryeducation and training evolvedthrough the following steps:1. International training courses, such as theWorld Agroforestry Centre’s ‘Introductionto agroforestry research and development’in the 1970s and 1980s, exposededucators and researchers to agroforestryprinciples and practices.2. Alumni tried to introduce agroforestrycourses into their home institutions inAfrica, Asia and Latin America, with varyingdegrees of success. This was done opportunistically.3. A broader interest emerged in universitiesand colleges to incorporate agroforestryinto education programmes, particularlyin faculties of forestry.4. Some institutions developed degreeprogrammes (B.Sc. and M.Sc.) in agroforestry.5. In time, many institutions began teachingagroforestry courses or programmes attechnical, B.Sc. and M.Sc. levels, but therewere few mechanisms for institutionalcollaboration nationally or regionally(compared to forestry and agriculture).6. The need for joint curriculum standardsand sharing of resources was recognized.7. Regional workshops were held, culminatingin decisions by educational institutionsto establish regional networks. Thisprocess was jointly facilitated by the Centre’sAfrican and Southeast Asian officesand key universities in the two regions.Status and needs assessments and institutionalvisits to universities and colleges(conducted in the early 1990s in Africa andin 1997/8 in Southeast Asia) revealed a seriesof constraints to agroforestry education(Hansson 1992; Temu and Zulberti 1994;Rudebjer and del Castillo 1999):• agroforestry was not recognized as specializationor discipline;• agroforestry curricula were inadequate:they were often incomplete and lackeda common standard;• training materials were in short supply:they were to few, too specialized, or inthe wrong language, and even whenmaterials were available, the librariescould not afford them;• there was limited research capacityamong staff and graduates;• lecturers needed training in all aspectsof agroforestry because agroforestry sciencehad developed so fast that therewere few trained teachers; and• there were inadequate links with fieldpractices.At the same time, opportunities fornetworking were appearing. Agroforestryprogrammes were offered within many educationalinstitutions. Stronger institutionswanted to take the lead while the ‘weaker’ones wanted to learn from others. Therewas recognition of the challenges relatedto land sub-division and intensification of

144World Agroforestry into the Futureland use. Agroforestry practices seemedto provide viable solutions. Governments,non-governmental organizations (NGOs)and multilateral organizations also tookan increasing interest in agroforestrydevelopment. Support from the policylevel increased, but human capacity wasneeded to implement such programmes.Partnerships with like-minded projects andorganizations emerged, including linkswith social and community forestry effortsand sustainable agriculture.Agroforestry network development tooka major step forward when the SwedishAgency for International DevelopmentCooperation (Sida) offered its support and,in 1993, the African Network for AgroforestryEducation (ANAFE) was established.This network now has123 member collegesand universities in 34 African countriesand is organized into four regional subnetworks,some of which include nationalsub-networks. The Southeast Asian Networkfor Agroforestry Education (SEAN-AFE) was established in 1999. There are76 member colleges and universities infive national sub-networks (Indonesia, LaoPeople’s Democratic Republic [PDR], thePhilippines, Thailand and Vietnam).Latin American countries are also followingsuit and a planning workshop for a LatinAmerican agroforestry network was held in2002.Membership, managementand activitiesThe networks link institutions rather thanindividuals. Membership is free. In Africa,any relevant institution may apply formembership. In Southeast Asia, given thepotentially very large number of institutionsin some countries, membership is basedon invitation as advised by the NationalAgroforestry Education Committee in eachcountry.Ownership by the members is securedthrough elected network leadership atregional, sub-regional (Africa) and national(Southeast Asia) levels. The members showcommitment through sharing the costs ofnetwork administration and meetings.The medium- to long-term direction andstrategies are discussed at regular meetings.The network coordination/facilitationunits are located at the Word AgroforestryCentre offices in Nairobi (Kenya) and Bogor(Indonesia), where they benefit from thelatest advances in agroforestry research.The networks conduct similar types ofactivities, with variations depending onnational and sub-regional needs. The keyactivities are:• curriculum review (using participatoryapproaches) and publication of guidelinesfor such reviews (e.g. Temu et al.1995; Rudebjer et al. 2001);• training-of-trainers in agroforestry theoryand teaching methods;• preparing, developing, translating andadapting teaching materials;• supporting graduate thesis research inagroforestry;• linking the networks to the regionalagroforestry research agenda;• pooling of resources and exchange ofstaff and experiences between institutionsin the networks (where ‘stronger’institutions assist ‘weaker’ ones);• providing information on network outputsand activities through publications,newsletters and websites; and• inviting policy makers to key events.AchievementsSEANAFE and ANAFE have already becomepowerful mechanisms for managing knowledgeand communicating and sharing experiencesin agroforestry among educationalinstitutions through their publications, newsletters,websites and databases. They arethe largest working networks of educatorsin Africa and Southeast Asia and are recognizedinternationally – being regional hubsfor the International Partnership on ForestryEducation (IPFE), an initiative launched atthe World Forestry Congress in 2003. IPFE,with initial support from the World Bank,aims to strengthen university-level educationabout forests and forestry worldwide,by facilitating and supporting collaboration.The efficiency and relevance of the networksis enhanced by their regional and nationalsub-networks, which encourage localsolutions to local problems. The networksform a platform for multidisciplinary dialogueamong educators, researchers and developmentworkers – effectively encouraginggreater integration and synergy. Collegesand universities are realizing that, to beeffective, they need to work more closelywith farmers and to capture their experiencesinto teaching programmes. New andrevised educational programmes emergeevery year, all addressing various aspects ofagroforestry and INRM. Due to changes ineducation policies, agroforestry and INRMare being accepted as important componentsof college and university education;institutions now consider agroforestry asa suitable platform for launching broadernatural resource management programmes,such as watershed management or environmentalconservation. At the national level,the networks have developed agroforestryteaching manuals for B.Sc. courses in locallanguages and institutions have establishedpractical field sites for training, research andoutreach activities. In addition, the Centresupports thesis research, staff exchangeand attachments at their own research sitesthroughout the regions.

Chapter 17: Institutional collaboration in agroforestry145In conclusion, the institutional networkshave proven valuable in terms of:• changing attitudes among educators:there is now greater understanding ofthe need for local context in rural development;• providing leadership for, and analysis of,agroforestry education within countriesand regions;• sharing experiences and enhancingprogrammes;• bringing integrated approaches tonatural resource management intoeducation systems; and• facilitating interaction betweenacademics, researchers, policy makers,extension workers and farmers.ChallengesAlthough member institutions of ANAFE andSEANAFE have responded strongly to theneed for curricula reform, there is still muchto do, since the science of agroforestry isdeveloping rapidly. For example, the emergingbroad landscape view of agroforestry isstill not widely covered. There is still a needto incorporate such knowledge into educationprogrammes and develop methods forfield-based learning with farmers.ANAFE and SEANAFE also need to addressthe challenge of growing demandfor participation and membership in thenetworks. The establishment of new, moredecentralized sub-networks brings issues ofsustaining leadership and communicationat the regional level. Regional meetingsare expensive to organize and mobilizingresources to support the growing networkswill not be easy. It is easier to find resourcesfor specific activities, such as trainingand teaching material development, thanfor network management. Furthermore,as demand grows it will be difficult todevelop and distribute sufficient academicmaterials to meet the growing need, particularlywhere countries use local languagesof instruction (e.g. Lao PDR and Vietnam).Finally, there is still a lack of policy-levelrecognition of agroforestry as a field of studyand a career path. It is generally felt thatthere is need for agroforestry competencebut a shortage of specific agroforestry jobs.Box 1. On-campus field laboratory in the PhilippinesFuture opportunitiesThe two networks are very well placed toaddress weakness in the education systemfor natural resources managementand to capture opportunities for educationalchange. The networks have broughtindividuals and institutions into long-termpartnerships. They have come to knoweach other, which also opens opportunitiesfor partnerships beyond agroforestry.The Misamis Oriental State College of Agriculture and Technology (MOSCAT) in thePhilippines began offering agroforestry education in 1995. Two programmes were offered:a diploma and a bachelor degree. Both required practical experience, but the lackof a convenient field site proved problematic. So, in 1998 25 hectares on campus weredesignated as an agroforestry field laboratory. The college itself had extremely limitedfinancial resources, so the development of the field laboratory was based on formingpartnerships with local agencies, international research centres, NGOs and the privatesector. The bulk of the financial support was provided by SEANAFE.Initially a simple banana and coffee plantation, the site now has a woodlot, windbreaks,a multistorey system with free-range chickens, silvipasture with free-range goats andsheep, alley cropping with improved natural vegetative strips, a nursery and a fishpond.Farm income (from production of maize, lanzones, rambutan, sweet potato, jackfruit,‘marang’, chayote, cassava and chickens) increased from US$117 in 2000 to US$425 in2003. The centre also has goats, sheep, and cattle. Networking with other stakeholdershas promoted a multisectoral approach to agroforestry development at the local level.Future plans include:• domesticating indigenous tree species;• producing seedlings through macro-propagation;• collecting non-timber forest products;• enhancing existing agroforestry systems for improved production;• developing an agro/eco-tourism village; and• strengthening links with national government agencies, NGOs and people’s organizationsthrough collaborative research and development.

146World Agroforestry into the FutureThe network approach can be applied ina broader context. For example, ANAFEis not strictly about agroforestry only. It isabout natural resources management, integratedbeyond forestry, beyond agriculture.There is a change of attitude among educatorstowards putting education into contextin rural development. Experiential learningmethods and tools are emerging andthe teaching and learning environment ischanging to include farmers’ participation.Agroforestry programmes can thus serve asvehicles for broader rural development.The networks also reach out to regionsand countries outside their core area (e.g.Latin America and South Asia) to influencechange. This role can be further strengthened,for example through IPFE.Finally, while the networks in some casesare being institutionalized (i.e. their programmesare becoming part and parcel ofthe institutions’ regular work), the implementationof many good ideas will dependon resource mobilization activities thatattract donors. Better funding will ensurethe networks can play an important role inreducing poverty and conserving the environmentin Africa and Southeast Asia.ReferencesBoje, D.M. and T.J. Wolfe 1989. TransorganizationalDevelopment: Contributions totheory and practice. In: Leavitt, H.J., L.R.Pondy and D.M. Boije (eds) Readings inManagerial Psychology. The University ofChicago Press, Chicago, USA.CGIAR 2000. Integrated Natural ResourceManagement Research in the CGIAR.A brief report on the Consultative Groupon International Agricultural Research(CGIAR) Integrated Natural ResourcesManagement (INRM) workshop held inPenang, Malaysia 21–25 August 2000.Consultative Group on International AgriculturalResearch, Washington, DC, USA.Hansson, B. 1992. Inventory of agroforestryteaching at universities and technicalcolleges in Africa. Training and EducationReport No. 20. World Agroforestry Centre,Nairobi, Kenya.Izac, A.M.N. and P.A. Sanchez 2001. Towardsa natural resource management paradigmfor international agriculture: the exampleof agroforestry research. AgriculturalSystems 69: 5–15.Rudebjer, P. and R.A. Del Castillo (eds) 1999.How Agroforestry is taught in SoutheastAsia – a status and needs assessment inIndonesia, Lao PDR, the Philippines, Thailandand Vietnam. Training and EducationReport No. 48. World Agroforestry Centre,Bogor, Indonesia.Rudebjer, P.G., P. Taylor and R.A. Del Castillo(eds) 2001. A Guide to Learning Agroforestry.A framework for developingagroforestry curricula in Southeast Asia.Training and Education Report No. 51.World Agroforestry Centre, Bogor,Indonesia.Temu, A.B., W. Kasolo and P. Rudebjer 1995.Approaches to agroforestry curriculumdevelopment. Training and EducationReport No. 32. World Agroforestry Centre,Nairobi, Kenya.Temu, A.B. and E. Zulberti (eds) 1994. AgroforestryEducation in Africa. Proceedings ofthe first workshop of the African Networkfor Agroforestry Education (ANAFE) 19–21April 1993, Nairobi, Kenya. World AgroforestryCentre, Nairobi, Kenya.

Keywords:Capacity building, agroforestry, postgraduate research,gender, partnershipsChapter 18Building capacity for research in agroforestryJemimah Njuki, International Center for Tropical Agriculture; Issiaka Zoungrana, International Plant GeneticResources Institute; August Temu and Janet Awimbo, World Agroforestry CentreAbstractThis chapter discusses the key components of capacity building and the mechanisms that have beenused to develop research capacity in agroforestry. These include building new expertise, developinginstitutions, forming networks, involving stakeholders and strengthening the links between institutionsthat have an interest in agroforestry research (including those in the North and the South, and thosewith national and international mandates). We discuss the efforts of the World Agroforestry Centre tobuild research capacity, including training at postgraduate level, postgraduate research, review of curriculato include agroforestry and networking between institutions. We also cover the recommendationsof a 1982 conference on professional education in agroforestry and the extent to which these havebeen implemented, and present the results of the Centre’s graduate training programme. Finally, we outlinethe challenges to building capacity for agroforestry research, along with some suggested strategiesand new opportunities.IntroductionCapacity building: definitions and componentsCapacity building is the structured process by whichindividuals, groups, organizations, institutions and societiesincrease their abilities to perform core functions,solve problems and define and achieve objectives, inorder to understand and deal with their developmentneeds in a broad context and in a sustainable manner.Temu and Garrity (2003) suggest that institutions withthe right capacity have a good policy environment,good strategies, control capital and financial resourcesand have the needed expertise to successfully mobilizetheir capacity.Szaro et al. (1997) define capacity building as enablingthe indigenous peoples of developing countries to carryout development processes successfully by empower-ing them through strengthening of domestic institutions,provision of domestic markets and improvementof local government efforts to sustain infrastructures,social institutions and commercial institutions.An effective research capacity building strategy aimsto build scientific, technological and managerialabilities and capacities at the individual, institutionaland organizational levels. To be successful, it shouldemphasize mechanisms that bring about a qualitativeimprovement in the way research is planned and implemented,and in the way results are disseminated.In order to facilitate the transfer and adoption ofknowledge generated, it is essential that capacitybuilding efforts involve all stakeholders in the researchprocess (including identification of issues, prioritisation,definition of research themes, activity monitoring

148World Agroforestry into the Futureand evaluation) and integrate traditionalknowledge (Olsson 1996; Thulstrup 1996).This contributes to building capacity for theapplication of research findings.Capacity building has three major components:• strengthening institutions (legal andpolicy framework, support mechanisms,conducive environment);• creating individual competence andstrengthening a critical mass of humanresources that is capable of planningand implementing agroforestry researchprojects to address the national agenda;and• developing infrastructure (researchequipment, buildings, facilities).Temu (2003 unpublished) gives a rich illustrationof the key components of capacitybuilding. These include enabling policies,credible strategies and programmes, physicalinfrastructure, human resource capacity,financial resources, institutional power andvoice, and networks or links with peers,clients and stakeholders (Figure 1).Mechanisms for capacity buildingThe pooling of resources through networkingis an essential element of anyresearch capacity building project (Owino1994). Networking can reduce the feelingof isolation and build critical massesof scientists within specialty groups in avirtual environment. Networking can takeCrediblestrategyandprogrammesEnablingpoliciesPhysicalresourcesFinancialresourcesHuman resources capacityLinks with clients,peers, stakeholdersInstitutional power/voiceFigure 1. Key components of institutional capacity.Source: Temu (2003 unpublished).

Chapter 18: Building capacity for research in agroforestry149many forms, including electronic, printor personal contact. All forms of networkingshould be encouraged, from local andnational to regional and international. Networkingensures that neighbouring countries,and even institutions within the samecountry, share lessons and do not duplicatesimilar work. It must be pointed out, however,that networking has transaction costssuch as communication and systems establishment.Well-developed scientific programmesare of little benefit if there is noparticipation from stakeholders and policymakers. Local input is essential to ensurethe relevance of the work, and to createa sense of ownership once the sciencemoves towards implementation in resourcemanagement and policy. Stakeholder engagementis a prerequisite for an effectivecapacity building strategy (Owino 1994).The effectiveness of capacity building alsodepends on a step-by-step approach, beginningwith existing capacity and activities.The goal is to make capacity buildinga unified process, within which particularactivities can be organized and deliveredin a logical order. Whatever the specificobjectives of commitments or projects,capacity building is, above all, a long-termprocess that must emphasize the developmentof local structures and organizations.In recent years, there has been increasingdemand for research-trained manpower inkey fields in many developing countries.There are several reasons for this, includingglobal technology change towardsmore efficient research-based methods inboth industry and agriculture, and increasingenvironmental concerns. The needfor research-trained experts is often veryspecific for a given country and locality,and requires continuous production ofresearch-trained human resources in thefield (Thulstrup 1993).One primary focus of research capacitybuilding should be to engender and encouragerelationships between institutions. Theseshould focus on the links between developedand developing-country institutions,and between institutions within developingcountries. Institutions may ask themselvesseveral questions before deciding withwhich institutions to collaborate; for example,which organizational structures are tobe involved? What institutional relationshipsneed to be established between them? Howare these relationships to be established?Which are the most crucial institutions withwhich to collaborate: those with similar interests,those with complimentary interestsor those with completely different interests?(Szaro et al. 1997).Another focus for capacity building is thestrengthening of institutions, not only interms of human capacity but also in termsof research strategies, policy and legalframeworks, communication capacity andresearch management capacity.The World AgroforestryCentre’s efforts to buildresearch capacitySince its inception in 1978, the Centre hasbuilt capacity in agroforestry in variousways. In the early days, capacity buildingtook the form of preparing training materials,guiding in-house trainees, givingagroforestry lectures and contributing toseminars and world literature on agroforestryand its place in education. In 1982,the Centre held a major conference to discussprofessional education in agroforestry(Zulberti 1987). The main recommendationswere to: a) develop agroforestry intoan experimental science that can be taughtwithin the context of existing professionallinks; b) integrate agroforestry into existingcourses, such as land use; c) incorporateagroforestry programmes at postgraduatelevel where appropriate; d) encourageshort courses in agroforestry within institutions;e) develop agroforestry training materialsthat will be regularly updated; and f)encourage South–South and North–Southcollaboration between institutions. TheCentre was seen as a major player in termsof supporting institutions to incorporateagroforestry and in providing material foragroforestry training.Building expertise in agroforestrythrough graduate fellowsOne of the main recommendations fromthe 1982 conference was to support postgraduatetraining in agroforestry throughfield research. This is producing new scientificknowledge about agroforestry systemsin the context of agriculture, forestry andintegrated management of natural resources.The knowledge is being absorbedand disseminated by teaching, learningand extension systems. It is also producingpolicy and technological innovationsthat encourage farming communities toadopt agroforestry. Collaboration betweenresearchers, educators, development workersand farmers, and building on existingfarmer knowledge/traditions is crucial inthis process.The World Agroforestry Centre has beenhelping to build the competence of individualsand the capacity of institutionsthrough giving both financial and technicalsupport. The technical support includesattaching students to Centre scientists andcarrying out thesis research in Centre fieldsites. In the last 10 years, the Centre hassupported 276 graduates at M.Sc. andPh.D. level (the majority from Africa) tocarry out field research in agroforestry(Figure 2). The highest percentage of traineescame from East and Central Africa(ECA). This is because the Centre has been

150World Agroforestry into the FutureHULWA(14.5%)South Africa(10.2%)Australia(0.7%)Europe(17%)North America(3.9%)Socialsciences(15.00%)Researchmanagement(3%)Other(3%)Soilsand water(31%)Asia(9.1%)Sahel(13.4%)ECA(31.2%)ECA = East and Central Africa;HULWA = Humid lowlands of West AfricaAgroforestrytechnologies(24%)Treedomestication(24%)Figure 2. Geographical distribution of World AgroforestryCentre graduate trainees (1993–2003).Figure 3. Thematic areas pursued by WorldAgroforestry Centre degree fellows (1993–2003).very active in ECA and has many scientistsbased there. There are also more institutionsteaching agroforestry at postgraduatelevel in ECA than in other regions.The Centre has also trained many studentsfrom Europe, Asia and Latin America. Mostof these are sent by major donors and arefunded by their respective projects. Out ofthe 276 fellows 214 have pursued Mastersprogrammes and 62 have pursued Ph.D.programmes. It must be noted, however,that the Centre has provided financialsupport (especially through the regionalnetworks) to more than the 276 fellowsreported here.The thematic areas studied in postgraduateresearch are very varied and includesoils and water, agroforestry technologies,tree domestication, social sciences andresearch management (Figure 3). Researchunder the social sciences has focused onissues such as marketing, extension, therole of agroforestry in poverty alleviation,adoption studies and studies on indigenousagroforestry knowledge. At the 1982 workshop,it was noted that agroforestry knowledgewithin the social sciences was verypoor. However, Figure 3 shows that wehave made positive developments in thisarea, with 15 percent of the fellows carryingout research in the social sciences between1993 and 2003. Research management,although important, has only beenstudied by nine of the 276 fellows andhas focused on geographical informationsystems (GIS), computer modelling and theeffectiveness of agroforestry research withrespect to funding levels.Gender representation, although not equal,has been very encouraging (Figure 4).Around 32 percent of M.Sc. students and37 percent of Ph.D. students were female.This is high compared to the general percentageof women working in agriculturalsciences (12–15 percent).One of the major achievements of thegraduate programme is that it has builta critical mass of agroforestry experts in80706050403020100MastersPh.D.MaleFemaleFigure 4. Gender balance of WorldAgroforestry Centre degree fellows (1993–2003).national agricultural and forestry services,especially in Africa. Although we have notformally followed up on individual trainees,we believe that many of them are workingin forestry and agricultural departments ofresearch institutions or universities.

Chapter 18: Building capacity for research in agroforestry151Postgraduate programmes ininstitutions of higher learningAt the time of the 1982 agroforestry conference,there were hardly any postgraduatecourses in agroforestry being offered byhigher learning institutions, although a fewprovided courses on land use. By 2003,Africa alone had 31 institutions offeringagroforestry as a postgraduate course(Table 1). Many others are now teachingTable 1.RegionNumber ofuniversitiesAfrica 31Australasia 4Central and SouthAmericaPostgraduate education inagroforestry.Eastern Europe andChinaNorth America 20Western Europe 41Southeast Asia 153 1? 1South Asia ? 11Information not completeSource: Temu (2003, unpublished).important component of the Centre’s capacitybuilding strategy. The networks haveexperimental sites in different countriesin Africa, which provide excellent experientiallearning for scientists from nationalresearch institutions and from the WorldAgroforestry Centre itself. Most of the researchby the Centre’s graduate fellows (asdescribed above) is carried out at researchsites managed by the networks.There are now three main agroforestry educationnetworks: the African Network forAgroforestry Education (ANAFE), the SoutheastAsian Network for Agroforestry Education(SEANAFE) and the Latin AmericanAgroforestry Network (LANAFE). These aredescribed by Rudebjer et al. in Chapter 17of this volume. In addition to these regionalnetworks, many countries have also formednational agroforestry networks.Issues and challengesin building capacity foragroforestry researchAgroforestry has come a long way sinceit was originally recognized as a simpleextension of forestry. It was categorized inthis way because the idea of planting treesappeared to fit best with the forestry profession,and foresters saw agroforestry asa way of helping farmers to produce theirown tree products and reduce their dependencyon natural forests. By and large,the early proponents of agroforestry wereforesters (Temu, unpublished).Agroforestry is now seen as a science thatis of increasing interest to a wide varietyof disciplines. Perspectives are changingand many new agroforestry programmesare being developed within agriculture,forestry, environmental education andother land-use programmes. Agroforestry iscurrently considered as an important entrypoint for holistic natural resources managementstudies within educational institutions.Temu and Garrity (2003) observe thatagroforestry also provides an entry pointfor biodiversity education. Despite viewingagroforestry as a broad-based disciplinetouching on various sectors, most nationalagricultural institutes and university facultiesstill remain very sector-based withseparate institutions for agriculture, forestry,wildlife, livestock, etc. Integrating theagroforestry agenda therefore represents asignificant challenge, especially in caseswhere there is no institutional collabora-agroforestry as a subject within other disciplines,such as forestry, agriculture and environmentalsciences. In addition, severalinstitutions have modified their curricula toinclude agroforestry (Table 2). The reviewand incorporation of agroforesty within thecurricula of national universities ensuresthe sustainability of these programmes andfacilitates interdisciplinary interactions.Agroforestry research and educationnetworksAgroforestry research networks were startedin Africa in the late 1980s and form anTable 2. Number of curricula reviewed to incorporate agroforestry in Africa (1993–2002).Discipline/level Certificate Diploma First degree Postgraduate TotalAgriculture 2 4 15 2 23Forestry 7 8 6 2 23Other (e.g. ruraldevelopment,horticulture)New agroforestryprogrammes1 2 3 0 60 4 5 6 15Total 10 18 29 10 67Source: Temu (2003, unpublished).

152World Agroforestry into the Futuretion between sector-based institutions,ministries or faculties.The sectoral nature of institutions, ministriesand faculties brings new challenges withinnational programmes in terms of coordinatingagroforestry activities. Despite recommendationsat the conference in 1982 formore integration, this has not happenedin most countries. In institutions of higherlearning that do not offer postgraduatecourses in agroforestry, students wishing tocarry out their research in agroforestry haveto find their own points of integration.Funding constraints often prevent nationalagricultural research organizations(NAROs) and public universities retainingwell-trained agroforestry staff. Once theybecome skilled, staff may move to theprivate sector, international research organizationsor non-governmental organizations(NGOs), or go out of the region on teachingor research assignments. Although theymay contribute in one way or another toresearch efforts in their home countries orregions, the high turnover presents a challengefor the national programmes. Thisproblem is not specific to agroforestry andis linked to the more general ‘brain drain’of expertise from the South to the Northand, within countries, from the public tothe private sector. For those that remain inthe national programmes, regular salariestend to be insufficient for sustaining an acceptablestandard of living and a secondincome is often a necessity. Only rarely aresecond jobs conducive to research activities;usually they will take time away fromresearch and (graduate) education. Lack oftime and incentive for research thereforepresents a serious problem for many developing-countryresearch programmes.Mainstreaming agroforestry into nationalprogrammes has been successful to acertain extent. In institutions of higherlearning, agroforestry has become an areaof study, even at graduate level, implyingmajor policy changes in the recognition ofagroforestry by universities. However, mostof the agroforestry activities conducted inNAROs are donor funded and agroforestryhas not become part and parcel of theircore programmes and priorities. This is duemainly to persisting sectoral orientation.Effective communication and sharing ofinformation across sectoral barriers remainsa daunting challenge at both localand national level. However, these problemsare only part of the broader globalcapacity building challenge to promote,legitimize and institutionalize effectivesharing of ideas and information acrosssectoral, national, cultural, linguistic andsocioeconomic barriers. The increasingimportance of electronic communicationhas often made matters worse for developing-countryresearchers, since the requiredfacilities may not be available. In the longterm, however, electronic communicationis likely to become a valuable tool, helpingdeveloping-country researchers to overcomethe long distances between them andother researchers in their fields.Measuring the impact of capacity buildingefforts presents another challenge. Thereare various ways to do it, but each methodhas its own shortcomings. As with any impactassessment, issues of cause and effectwill arise. A ‘with and without’ analysissuffers from bias, since the very effort thatwent into identifying target institutionswould make any comparison suspect. A‘before and after’ comparison assumes thatover that same time period, there were noactivities or initiatives affecting the institutionother than the particular researchcapacity building efforts for which impactis being measured. Individual institutionstherefore need to develop a combinationof measures that will give the best resultsbased on the objectives of their capacitybuilding efforts. Examples of aspects thatcan be measured include:• The number of people that have beentrained (critical mass of expertise withinorganizations). This is really a measureof building research competence ratherthan research capacity.• The number of publications producedby trainees. This approach has distinctlimitations in that not all the good researchis published, and there may begood reasons why some people do notor cannot publish their research results.Some work may be more effectively disseminatedby means other than publishingin peer-reviewed journals. In manycases, there is a considerable time lagbetween the research results and publication.• The funding attracted by researcherswho have benefited from the capacitybuilding effort. This has limitations,however, as comparisons may not bevalid. Different institutions, especially indifferent countries, may attract differentlevels of funding based on their relationswith donors. And this only measures researchers’ability to attract funding, notthe impact of their research on the enduser, i.e. farmers and the poor.• The networks built by the institution.This measures the extent to which increasingthe capacity of an institutionincreases its ability to establish partnershipswith other international, regionaland national organizations. While thisis a good measure of the institutionalgrowth as a result of the built capacity,it does not address issues of impact onthe other institutions nor on the end usersof the institutions’ research products.• Inclusion of agroforestry in nationalprogrammes is another measure of

Chapter 18: Building capacity for research in agroforestry153impact that evaluates the extent of institutionalizationof agroforestry within aninstitution’s or a country’s programmes.This can be measured in various waysincluding the evaluation of policydocuments to assess whether and howagroforestry is articulated in these policydocuments, the perceptions of key policymakers on the role of agroforestry,changes in practice, etc.This kind of impact assessment and evaluationmay be internal or external or acombination of both. There are advantagesand disadvantages in internal and externalevaluations and the two should be combinedin order to reap maximum benefitfrom their advantages. While an internalevaluation allows and encourages learning,corrective measures and improvements,it may be subjective, since actors may failto highlight weaknesses and concentrateon positive aspects. Building capacity inevaluation for learning and change can,however, reduce this risk. External evaluationsmay provide a more objective view ofthe strengths and weakness of an institutionand the recommendations should be usedto make improvements.learning that do not offer agroforestry as afull postgraduate course, it is still left to studentsof agriculture and forestry to integratethe two if they want to pursue researchin agroforestry. Measuring the impact ofcapacity building efforts also presents achallenge.New opportunities for capacity building inagroforestry are presented by forming linkswith sub-regional and regional organizations,such as the Association for StrengtheningAgricultural Research in East andCentral Africa (ASARECA) and the Forumfor Agricultural Research in Africa (FARA).The World Agroforestry Centre forms acentral point for linking educational institutions,NAROs, NGOs, donors andregional organizations (Figure 5). In thepast, links between educational institutionsand research institutions have been weakand have failed to address issues of strategy,such as joint research and educationprogrammes involving different types ofDonorsinstitutions. While links between researchand extension have improved, there is stilla gap between research and education. Inmost cases, the only points of contact areuniversity students, who attend researchinstitutions for experiential learning and tocarry out their thesis research, and researchstaff, who attend university courses. Theselinks need to be strengthened to promotesharing of information and development ofcoordinated agroforestry programmes.A good model for improving the coordinationof agroforestry education is the formationof a national agroforestry coordinationcommittee. This would include representativesfrom different ministries or departments,universities, national agricultural andforestry research institutions, NGOs, theprivate sector and international agriculturalresearch centres that are engaged in or havean interest in agroforestry. The World AgroforestryCentre is well placed to support thedevelopment of such committees.Conclusions andrecommendationsCapacity building in agroforestry has comea long way since the 1982 agroforestryeducation conference. Most of the majorrecommendations from that conferencehave been implemented and the WorldAgroforestry Centre has played a leadingrole. However, significant challengesremain. The sectoral nature of research,education and administration has affectedthe institutionalization of agroforestry. Lackof funding and loss of staff to better-payingjobs also limits the effectiveness of capacitybuilding efforts. In the institutions of higherUniversities(North and South)WorldAgroforestryCentreFigure 5. Institutional links in agroforestry.FarmersNAROs(research and extension)and NGOsRegional andsub-regionalorganizations

154World Agroforestry into the FutureThe Centre has already been successful increating national and regional partnershipsand networks in developing countries. Thenext step will be to strengthen the linksbetween them and agroforestry institutionsin the North. The issues to consider whenforming new types of partnerships are: a)the organizational structures that need tobe involved; b) the institutional relationshipsthat need to be established; c) thecomparative advantage that each institutionbrings; and d) how to establish the relationships,including the sharing of credit/authorshipsand issues of intellectual propertyrights.Relationships with institutions in the Northwould provide several benefits, includingopportunities for overseas studies, jointreview of agroforestry curricula, scientistexchange programmes, information sharingand improved communication technology,joint research and publications and access/training in new research methods andtools.ReferencesOlsson, B. 1996. Developing a CreativeResearch Environment in Least DevelopedCountries. In: Thulstrup, E.W. andH.D. Thulstrup (eds) Research Training forDevelopment. Roskilde University Press,Frederiksberg, Denmark.Owino, F. 1994. The Capacity Building inForestry Research Project (CBFR) of TheAfrican Academy of Sciences. Paper presentedat Supporting Capacity Buildingin Forestry Research in Africa. Proceedingsof the First International Symposiumheld in Nairobi, Kenya. 28 June–1 July,International Foundation for Science (IFS),Stockholm, Sweden.Szaro, R.C, E.W. Thulstrup, W.W. Bowers, O. Souvannavongand I. Kone 1997. Mechanismsfor Forestry Research Capacity Building.Paper presented at the International Consultationon Research and Information Systems.7–10 September, Gmunden, Austria.Temu, A. 2003. Global Education in Agroforestry.World Agroforestry Centre, Nairobi, Kenya(unpublished).Temu, A. and D. Garrity 2003. Building InstitutionalCapacity to Enhance the Adoptionof Innovations: Experiences of The WorldAgroforestry Centre. Paper Presented at aUnited Nations Conference on TechnologyTransfer and Capacity Building, 23–27June, Trondheim, Norway.Thulstrup, E.W. 1993. Scientific research fordevelopment education and social policy.Discussion Paper. The World Bank, Washington,DC, USA.Thulstrup, E.W. 1996. Strategies for ResearchCapacity Building through Research Training.In: Thulstrup, E.W. and H.D. Thulstrup(eds) Research Training for Development.Roskilde University Press, Frederiksberg,Denmark.Zulberti, E. 1987. Professional Education in Agroforestry:Proceedings of an InternationalWorkshop organized by the World AgroforestryCentre, 7–11 September 1982, ICRAF,Nairobi, Kenya, International Centre forResearch in Agroforestry, Nairobi, Kenya.

Keywords:E-learning, agricultural education,Africa, Southeast AsiaChapter 19Can e-learning support agricultural development indeveloping countries?Albert Dean Atkinson, International Rice Research Institute; Jan Beniest, World Agroforestry Centre and SheilaRao, World Agroforestry Centre and The Commonwealth of LearningAbstractAdvances in information and communication technology are opening up new opportunities for distancelearning. Various centres within the Consultative Group for International Agricultural Research(CGIAR) have been developing electronic learning materials for use on the Internet, such as on-lineshort courses, CD-ROMs, on-line communities and collaborative sites. The centres have also developedvirtual university courses for longer-term training. The CGIAR centres collaborate with nationaland regional partners in developing and using these e-learning materials. The centres are also creatinga joint Learning Resources Centre. This will gather and coordinate learning materials based on ongoingresearch from contributing CGIAR centres, especially material with cross-disciplinary applicability,such as statistical analyses or scientific writing. These materials will be adapted for a broad rangeof applications, and will be presented in different media and levels of interaction, depending on localcircumstances and the context of learning goals.IntroductionAgricultural research and development (R&D) institutionsin developing countries often have problemskeeping up-to-date with advances in internationalscience and technology. Members of the ConsultativeGroup on International Agricultural Research (CGIAR)are working with many national institutions to helpthem improve in this area. For countries with few scientistsand weak institutions, the CGIAR may help directlyby organizing training programmes or providingmaterials and resources; while in countries with manycompetent scientists and well-functioning institutions,they adopt a collaborative and partnership approachthrough networking and sharing knowledge.The training and education materials developed by theCGIAR are in great demand and are reaching audiencesfar beyond the participants of in-house trainingevents. The materials can be translated and adapted tomeet the needs of specific audiences in regional settings,but they need to be produced in partnership withnational institutions to ensure ownership.Face-to-face-based learning tends to be costly, particularlybecause of travel and accommodation expenses.Advances in information and communication technology(ICT) mean widely scattered audiences can now findout about recent developments in agricultural R&D andnatural resource management. Although the innovations

156World Agroforestry into the Futureare proving successful in developed countries,problems with communications infrastructureand access to computer facilitieshave made this approach more difficult indeveloping countries. This chapter highlightssome of the difficulties and takes a criticallook at the new CGIAR approaches towardsbuilding capacity and strengthening institutionsusing electronic-based ICT.ICT in training and educationWhen considering ICT, experienced educatorsknow that the characteristics of theaudience for whom the training is targetedand the content of the training, (for instance,will it close the knowledge gap?)are more important than the technologythat delivers the training. For over 50 years,training professionals have relied on theinstruction systems process as a guidingmethod for designing training (see Figure 3).organizations resembles that of corporationsin 2000/01. John Chambers, Presidentof Cisco Systems Inc. claimed in 1999that “the next big killer application for theInternet is going to be education. Educationover the Internet is going to be so bigit is going to make e-mail usage look like a‘rounding error’ in terms of the Internet capacityit will consume” (Chambers 1999).So what happened? Figure 1 highlights thecorporate e-learning experience from 1996to 2003.When the ‘dot com’ bust occurred in 2001,corporations began to realize that their corporateuniversity investments were not providinga good return and they started to dissolvethe mechanisms. Within six months,Electronic Data Systems, one of the world’slargest ICT service providers, cut their corporateuniversity from 245 staff to 25.Kruse (2003) states: “Then 2001 broughtthe harsh, steep slope of unfulfilled promises.Several high-profile providers shuttheir doors while many more announcedlarge-scale layoffs in the face of missedrevenue targets and crashing stock prices.E-learning advocates retreated to the moredefensible ground of ‘blended learning.’”The World Bank’s 1999 World DevelopmentReport suggested that advances in electronicand computer-based ICT would allow ‘unprecedentedopportunities’ for the world’spoor. “Advances in communications havetransformed society before: movable type,photography and telegraphy, the telephone,television, and the fax machine have allpushed outward the limits of our ability tostore and transmit knowledge. Now theconverging of computing and telecommunicationsappears ready to shatter those limits,Information and communication technologyis not new, nor is it limited to transferringlarge amounts of information at the speedof light across the Internet. Essentially, ICTis any technology that transmits information– whether it is cave paintings, smokesignals, kites, books, CD-ROMs, television,radio, or computers and the Internet. “Itis an expanding assembly of technologiesthat can be used to collect, store and shareinformation between people using multipledevices and multiple media” (Chapman andSlaymaker 2002). Despite the broad definitionof ICT, we focus on advances in electronicand computer-based ICT here.Development organizations working to improveaccess to information and educationseem to be intrigued by the possibilities ofICT and there are good levels of fundingfor ICT-driven capacity-building projects.It is interesting to note that the pattern ofinterest in e-learning from developmentVISIBILITYE-LEARNING HYPE CYCLE1996 1997 1998 1999 2000 2001 2002 2003TechnologytriggerSCORM 1.0 releasedLearning portal fad beginsAsymetrix becomes Click2Learn.comFocus changes from tools toenterprise wide management systemsMacromediaFlash 1.0NYU Online foundedSaba foundedFirst article on“Intranet-based Training”in Training MagazineOnly one “IBT”session at ASTDCBT Systems becomes SmartforceMasie launchesTechlearn 97Figure 1. E-learning hype cycle.Peak of inflatedexpectationsDigitalThink stock at US$89.44John Chambers declares“E-Learning is the next killer application”Headlight.com closesPensare closesCaliber closesPeer3 closesFocus on Blended learningLCMSs emergeNYU Online closesSCORM 1.2 releasedSmartForce and Skillsoftannounce mergerTrough ofdisillusionmentDigitalThink stock at US$1.34TIMEEnlightenment and productivityNote: Figure illustrates the wave of interest and investment in e-learning. The acronyms are not relevant inthis context.Source: www.e-learningguru.com/articles/hype1_1.htm

Chapter 19: Can e-learning support agricultural development in developing countries?157making it possible to send vast amounts ofinformation anywhere in the world in seconds– at an ever-decreasing cost. This newtechnology greatly facilitates the acquisitionand absorption of knowledge, offeringdeveloping countries unprecedented opportunitiesto enhance educational systems,improve policy formation and execution,and widen the range of opportunities forbusiness and the poor.” (World Bank 1999).While encouraging, the World Bank statementplaces a tremendous responsibility oneducators and trainers. But while the technologywill greatly facilitate informationacquisition, it will do nothing for the ‘absorptionof knowledge’ if the content is notstructured and organized in a meaningfulway. In other words, technology alone doesnot ‘create’ knowledge; it is the quality anddesign of the content that matters. Poorlydesigned and structured material deliveredby distance learning is just as ineffective aswhen it is delivered face-to-face.Computer and Internet availability, policiesand training vary widely throughout Africa.For example, in 2001 in Kenya 1.6 per centof the total population were Internet users.In Zambia, the figure was less than onepercent (World Bank 2002). These figuresare increasing rapidly as people realizethe benefits of ICT. Professor Abdulrazak,Deputy Vice Chancellor of Research andExtension at Egerton University in Kenyabelieves that “investing in computer-baseddistance learning allows greater opportunitiesfor future generations of students andfaculty to access and to fully take part inthe global information society” (Abdulrazak2004). However, Internet access islimited, at present, to well-resourced, wellfundedinstitutions and regions.Table 1 shows that most Internet users arein North America, Europe and, to a lesserextent, Asia and the Pacific Rim. Africahas the lowest user numbers, next to LatinAmerica. It also illustrates the large gap betweenusage, indicating the difficulties associatedwith developing e-learning. However,the number of users is rising everywhere.Limited Internet connectivity and bandwidthavailability is a major obstacle toincreasing Internet use and e-learning inAfrica. Figure 2 shows the connectivity inAfrica through bits per capita (BPC). Thistakes into consideration the wide range ofInternet applications available, from personaluse to cybercafes and business transactions.The map shows the Internationallevel (bpc) bandwidth available in eachAfrican country with the darker shadedcountries having the most accessablity andthe lighter shaded countries with the least.The high percentage of outsourcing Internetand satellite coverage outside of Africamakes bandwidth availability expensivebecause of the telecommunication and administrativeupkeep of such a system, andlimits the amount of Internet use in educationalsettings. The average cost of using alocal dial-up Internet account for 20 hoursa month in Africa is about US$60 (Etta andParvyn Wamahiu 2003).Table 1.Source: Dennis (2000).Despite the technical barriers to computerbasedICT development in Africa (and to alesser extent in other regions), e-learningis developing in response to the growingneed for access to education. Even wherereliable Internet connection is not yet possible,e-learning is being developed, basedon the assumption that ICT infrastructurewill continue to grow in the coming years.Therefore, establishing the theoretical andpedagogical background in e-learning willensure valuable benefits in the future.Figure 2. Internet connectivity in Africa.Source: IDRC 2003. For the full colour versionof this map see http://web.idrc.ca/en/ev-6568-201-1-DO_TOPIC.html.Active adult Internet users aged 14 and over, worldwide (in millions).2000 2001 2002 2003 2004North America 97.6 114.4 130.8 147.7 160.6Europe 70.1 107.8 152.7 206.5 254.9Asia and the Pacific Rim 48.7 63.8 85.4 118.8 173.0Latin America 9.9 15.3 22.1 31.0 40.8Africa and the Middle East 3.5 5.3 7.2 9.0 10.9Worldwide total 229.8 306.6 398.2 513.0 640.2

158World Agroforestry into the FutureLearning in e-learningIt is important to recognize that humanlearning is a complex psychological process.It can be structured or unstructured, formalor informal, supervized or unsupervizedand carried out with or without the assistanceof a teacher. ‘Distance’ or ‘e-learning’is also referred to as correspondence education,distance education or training, distributedlearning, on-line learning, open learningand web-based learning. For the sake ofclarity, we consider the term ‘e-learning’ tobe a catch-all phrase that suggests learningat a different time and in a different place(asynchronous learning), aided by a computerconnected to the Internet.A literature review of e-learning finds recurringreferences to its advantages overface-to-face learning. But, just as e-learningis neither inherently cost-efficient norcost-effective, its advantages cannot alwaysbe assumed. For example, e-learning instructorsneed the same amount of time tofacilitate a course as a face-to-face instructor.The student–instructor interaction stillexists, but through the telephone or e-mail.E-learning allows students to maintain flexiblehours and reduces travel and relocationcosts. However, this is based on theassumption that the learner has continuousaccess to a computer and/or the Internet atany given time. In some regions, e-learningtakes place in cyber cafes where accessibilityis limited.with a desired state, then examining thegap between the two to see how it can bebridged. The point is to design content in away that brings users closer to the desiredstate. The content is then developed anddelivered. An evaluation closes the loop tocheck whether the design of the instructionwas effective. The ISD model ensures thelearning materials are of good quality andthat the learning process, the instructionaldesign of the materials and the learningoutcome is not compromised as a result ofthe technological tools used.Training and education andcomputer-based ICT“ICTs, as any tools, must be consideredas such and be used and adapted to serveeducational goals. It is educational needsand goals, not materials or technology thatmust drive educational change. Many ethicaland legal issues concerning widespreaduse of ICTs in education are yet to besolved” (UNESCO 2004).Terry Hilberg, Chief Executive Officer ofNextEd suggests, “In Asia, people like to sitin classrooms with a professor at the frontof the class, handing out pearls of wisdomto a silent student body” (Raths 2000). Thislearning style is the opposite of learning ata distance, where the ‘pearls of wisdom’are the product of student interaction anddiscussion, facilitated by an instructor.Therefore, culture – not access – may bethe main barrier to overcome before theCGIAR partners and target audiences beginto appreciate e-learning. As a result, coursecompilers are moving away from structurede-learning courses and towards packagedreference guides, fact sheets and standalonelearning objects that can be translatedand customized by national partnersto provide training in the local context.Table 3 summarizes the opportunities forusing computer-based ICT, the issues thatarise in training and education, and therole CGIAR centres can play in this area.In March 2003, the World AgroforestryCentre conducted an e-survey of 45 academicagricultural institutions located invarious countries in Africa and Asia (Tossell2003). The survey explored the degree ofinterest of partner institutions in e-learningactivities, their level of Internet usage andtheir Internet connection reliability in fourcategories: news updates, on-line databases,on-line courses and on-line communities.The intent was to gauge the degree ofinterest in using selected e-learning activities.A second part of the survey looked atInternet use and connection availability toestablish possible links between personalinterest, use and availability.The results of the survey showed that therewas interest in expanding the use of ICTMany of the often-claimed advantages of e-learning will be realized only if the instructionis properly designed. So how can thisbe assured? The answer lies in the mechanicsof the instructional systems design (ISD)process model (Figure 3).AnalysisDesignEvaluation Delivery DevelopmentThe ISD process begins by analysing anaudience’s current state, comparing itFigure 3. The instructional systems design process model.

Chapter 19: Can e-learning support agricultural development in developing countries?159Box 1. The Rice Knowledge BankThe Rice Knowledge Bank at the International Rice Research Institute (IRRI) is a repositoryof training and extension resources. It receives an average of 634 visitors a day,with a total of more than 2.5 million web hits since its launch in April 2002. Its contentis screened before publication with three criteria in mind: i) is it credible? ii) is it valueadded,i.e. has it been organized in a format and sequenced in a way that promoteseasy understanding? and iii) is it demand driven? For the demand-driven criterion, theKnowledge Bank uses statistical web traffic-monitoring software. This provides:• A live view of the formats users prefer, measured by the frequency of downloads forspecific formats.• Keyword query analysis that indicates the subject areas users are most interested in,as indicated by the frequency of certain keywords that are typed into the knowledgebank’s search field and those that are used in the Internet’s leading search engines tofind content on the knowledge bank.• Demographic analysis of where the users are located and the types of systems andsoftware they are using to access the Knowledge Bank.Source: Albert Dean Atkinsen www.knowledgebank.irri.org.Table 3.Summary of e-learning tools.Type of e-learning sourceWeb-based on-line coursesFunctionAdministered through the Internet, have flexible timeschedules, alleviate geographical barriersfor disseminating information and education,despite a lack of reliable Internetconnection (Table 4). When asked aboutpersonal Internet use, 53 per cent of respondentssaid they used it frequently, butonly 31 per cent thought their connectionwas reliable. The survey also showed thatmost institutions viewed on-line databasesand news reports as the most useful formsof e-learning.The relatively low interest in on-linecourses and communities suggests a correlationbetween the level of interest andusage in computer-based technologies andlack of reliability in Internet connectionsas well as limited relevance of current Internetcourses. Those with little access orslow connections and minimal technicalsupport may not see on-line communitiesas appropriate or feasible. The low figuresalso suggest that exposure to and understandingof on-line communities is limited,and thus they are not perceived as useful.Similarly, on-line courses are perceived toneed a more efficient Internet connectionthan is currently available in some areas.Using on-line communication as part of anexisting course or lesson plan may be morereadily accepted.On-line databasesModerated listservesOn-line communitiesCD-ROMsVirtual universitiesNetwork host provides a central repository of informationaccessible from anywhere in the worldTechnologically easy to set up, e-mail lists are well adaptedto low bandwidth situationsRequires consistency in moderating list activitiesBuilds international links based on research, elaborates onexisting education and training coursesLess common than e-mail, more affected by bandwidthrestrictionsConverts hard-copy documents into electronic versionsWidespread usabilityProvides on-line courses to those without access to universityfacilitiesRequires high-powered networks and computersUsing a combination of various types ofe-learning or ‘blended learning’ such ason-line databases, recent agroforestry newsupdates, on-line communities and on-linecourses may be appropriate to match thebenefits of ICT in education and training touser needs and capabilities. For example,a particular web site or repository maycontain links to on-line databases, provideupdated research information and offershort-term courses or sections of trainingcourses. Alternatively, an existing coursemay contain on-line components or opportunitiesfor discussions pertaining to faceto-facelearning and course content.

160World Agroforestry into the FutureTable 4. Interest in different methods of e-learning.Method of ICT delivery Most useful method (%)News and current events about agroforestry research 32.5On-line databases of agroforestry information, including research papers and training materials 29.4On-line courses covering news and current events about agroforestry research 19.7On-line communities, using group e-mail and special web sites to share information with agroforestrypractitioners worldwide18.2Source: Tossel (2003).On-line communitiesVarious types of e-learning methods can becategorized as on-line communities. Theyrange in complexity from very simple listservesto more complex interactive conferencingor live chats. Recent workshops andcourses hosted by the Centre have usedon-line communities to give internationalcourse participants access to informationbefore the course (for example, backgroundinformation on course topics, preliminaryreading and contact informationof other participants). After the course, theon-line community can act as a venue forfollow-up discussion and resource sharing.The concept of the on-line community wastested after two Centre training courses andthe investigators found that most membersparticipated at least once, but problemswith Internet access limited the consistencyof participation.Developments in ICT allow CGIAR centresto work together when developing learningresources. This helps avoid overlapand duplication of effort in subject matter,such as experimental design, participatoryresearch, scientific writing, data analysisand training methods. Using software suchas MicroSoft SharePoint Team Services (seewww.microsoft.com/windowsserver2003/techinfo/sharepoint/wss.mspx) or Internetservices such as D Groups (see www.dgroups.org) gives research and resourceteams the ability to create and manageweb sites where they can communicate,share documents and work together ondeveloping training materials of mutualinterest, regardless of the physical locationof team members. Issues surroundingaccess to the server from different regionsprevents interregional collaboration due tobandwidth restrictions and varied Internetinfrastructure. D Groups was designed withlow bandwidth regions and limited Internetaccess in mind. It provides a service forinternational collaboration, provided theappropriate type of community is implemented,depending on where communitymembers are located. Communities can bebased on particular topics relevant to developingcountries. For example, a globalcommunity was formed to discuss the useof scenarios as a decision-making tool forcommunities located at the tropical forestmargins.In rural areas, on-line communities can beas simple as a listserve accessible throughan internet cafe or telecentre. Recent Centreactivities for teachers from eastern andsouthern Africa used listserves to follow upthe course and to share and exchange information.Most participating teachers werebased in rural areas with limited access tothe Internet. They overcame this by havingInternet-based e-mail addresses and usingInternet cafes in nearby cities. Thus, the onlinecommunity, in this context, suited theaudience and information could be shared.Virtual universitiesAdvances in ICT have allowed open anddistance learning universities to change theway they provide formal education. Correspondencecourses can now be deliveredvia radio, television, fax, telephone and theInternet. Virtual universities are an alternativeto open universities and administertheir courses over the Internet as interactiveweb sites, on-line seminars, e-mail discussions,CD-ROMs and on-line video conferencing.Many of these concepts are used inuniversities in developed countries and themodel is being adapted in an attempt toincrease access to education in developingcountries.The CGIAR centres have generated largeamounts of scientific knowledge, have theresource persons to impart such knowledgeand have contributed to the education of avast number of students. The CGIAR GlobalOpen Agriculture and Food University(GO–AFU) aims to extend collaboration

Chapter 19: Can e-learning support agricultural development in developing countries?161with national institutions and provides opportunitiesto enhance teaching and learningin developing countries. This initiativewill provide a link between technical andtheoretical expertise from well-establisheduniversity institutions in developed countriesand educational institutions in developingcountries. Opinions on whetherthis is needed vary amongst participatingcentres. Some feel that it goes against thegrain of strengthening national institutionsand risks competing with and potentiallyweakening them. Others are concernedabout the cost-effectiveness. CGIAR scientistscontribute to short-term training andeducation activities as resource persons,but that is less of a commitment thanteaching or supervizing students in an openuniversity.A similar initiative headed by the Norwayoffice of the United Nations DevelopmentProgramme (UNDP) will soon begin to offerglobal courses on environmental anddevelopment issues under the Global VirtualUniversity (GVU). “Courses offered byGVU will be primarily directed at studentsfrom developing countries” (UNDP 2003).However, the kinds of technologies andsoftware being used are congruent withdeveloped-country standards of e-learningrather than existing electronic technologyavailable in developing countries.The African Virtual University (AVU) offerson-line courses that are administeredin Africa but originate from Europe andAmerica. This World Bank-funded projectwas established in 1995 to serve studentsfrom sub-Saharan Africa. The initial phasetook place in Kenya, where the virtualuniversity “rightly focused on science, engineering,business and the medical fields,as a technology-based distance educationalternative” (Amutabi and Oketch 2003).Universities in several African countriesparticipate in these courses and act asfocal points. Sourcing the project externally,however, increases tuition fees anddependency on global support instead ofbuilding local infrastructure that wouldalso build local knowledge. The bulk of thecourse materials come from Colorado StateUniversity, the University of Massachusettsand the New Jersey Institute of Technology(in the USA) and University College Galway(in Ireland), which increases the costof administration and delivery. In addition,since the course content originates fromoutside Africa, it is not oriented specificallyto African situations. Therefore, the longtermimpact on African students dependson the students’ employment options and,more importantly, whether they can affordthe course.E-learning resourcesUntil recently, the CGIAR centres producedmaterials for their training events in traditionalformats (print on paper, slide series,videos and audiotapes). Advances in ICTallow them to develop and produce materialson floppy disks and CD-ROMs or asdownloadable files from web sites. One ofthe main advantages of electronic formatsis that they can be updated easily to reflectchanges and advances in subject matter,and they can be adapted to suit the needsof different audiences.E-learning resources can be viewed on-lineby anyone interested in training and educationin agriculture and natural resourcemanagement. In 2003, a number of CGIARcentres decided to collaborate as a trainingcommunity and proposed a project entitledthe CGIAR Learning Resources Centre.The idea is to package learning materialsproduced by different CGIAR centres (suchas documents, models, images, video clipsand presentations) into ‘value-added’ learningmaterials (i.e. in addition to instructionaldirection they will provide additionalreferences and course materials, contactpersons, etc.). The content is based on individualcentre research focus (e.g. naturalresource management, genetic resources,livestock, roots and tubers, rice, wheat andmaize). Materials are offered in various formats,such as an electronic reference guide,an e-learning course, a fact sheet or a slidepresentation. All are designed in a way thatallows further formatting depending on theusers’ needs, while maintaining a commonlook and feel. It is hoped that the materialswill be used by the CGIAR training community,national partners and the internationalagricultural community in general.ConclusionThe Internet and computer technology offermany opportunities for supporting trainingin international agricultural research anddevelopment. However, technologies mustbe practicable for the intended audiences.Content developers must understand howcomputer-based systems can manage,facilitate and support – but not replace– conventional training methods. Internetaccess also opens the way for many newand long-distance partnerships. Materialsoffered by e-learning must be adapted tothe local context, usually by consultingwith national education institutions. Thereare various approaches to e-learning, eachof which is appropriate in a different situation.Computer-based learning has thepotential to increase accessibility to educationand training although, in some parts ofthe world, lack of Internet access currentlypresents a barrier. A ‘blended approach’applies the most suitable ICT to a regionand particular audience. Often face-to-faceteaching combined with computer-basedtechnology provides added value to theeducational process.

162World Agroforestry into the FutureThe potential for a project like the CGIAR’son-line learning resources can only berealized with sound planning and technicalappropriation. Using inexpensive, lowbandwidth programmes allows greateraccess by those directly in need of agriculturaland natural resource management informationand those with little experienceof the Internet or computer technology.Although e-learning no longer carries thehype from previous years, it has the capacityto enhance learning and expand accessto education and training in agriculture andnatural resource management at the global,regional and local levels.ReferencesAbdulrazak, S.A. 2004. Presentation at a conferenceon On-line Learning Resources heldat International Crops Research Institutefor the Semi-Arid Tropics, Patancheru,India,14–18 June 2004. World AgroforestryCentre, Nairobi, Kenya.Amutabi, M.N. and M.O. Oketch 2003. Experimentingin distance education: the AfricanVirtual University (AVU) and the paradoxof the World Bank in Kenya. InternationalJournal of Educational Development23: 57–73.Chambers, J. 1999. Quoted in Friedmans, T.L.Next, it’s E-ducation. The New York Times,November 17th 1999.Chapman, R. and T. Slaymaker 2002. ICTs andrural development: review of the literature,current interventions and opportunities foraction. Working Paper 192. Overseas DevelopmentInstitute, London, UK.Dennis, S. 2000. Just five percent of World online.www.computeruser.com/news/00/12/11/news15.html retrieved June 2001.Etta, E.F. and S. Parvyn Wamahiu 2003. Informationand communication technologiesin Africa. The experience with communitytelecentres. Volume 2. InternationalDevelopment Research Centre Ottawa,Canada.IDRC 2003. Acacia Project. InternationalDevelopment Research Centre, Ottawa,Canada. www.idrc.ca/acadiaKruse, K. 2003. The State of e-Learning: Lookingat History with the Technology HypeCycle. www.e-learningguru.com/articles/hype1_1.htmRaths, D. 2000. Looking to the East. www.timclark.net/media/looking.to.east.htmlTossell, I. 2003. E-learning Needs Assessmentin East Africa. Unpublished report. WorldAgroforestry Centre, Nairobi, Kenya.UNDP 2003. Global Virtual University (GVU)Brochure. United Nations DevelopmentProgramme, Arendal, Norway.UNESCO 2004. United Nations Educational,Scientific and Cultural Organization Informationand Communication Technologiesand Education Programme. http://www.unescobkk.org/education/ict/v2/info.asp?id=10954World Bank 1999. World Development Report1998/99: Knowledge For Development.World Bank, Washington DC, USA. http://www.worldbank.org/wdr/wdr98/contents.htmWorld Bank 2002. ICT at a Glance: Zambia.http://www.worldbank.org/cgi-bin/sendoff.cgi?page=%2Fdata%2Fcountrydata%2Fict%2Fzmb_ict.pdf accessed March 2004.

Chapter 20Strengthening Institutions:Working Group ReportStrengthening capacity for scaling upagroforestry technologiesGoalTo create a high level of awareness of the benefits ofagroforestry among stakeholders. This will encourageagroforestry development projects and contribute towider rural development.Strategic issuesThe World Agroforestry Centre needs to define its visionand strategy for scaling up and its role in relationto those of its partners. Successful scaling up means improvingthe quality and quantity of learning resourcesavailable to extension staff. Training in the areas of policy,management, markets, biotechnology, land degradationand indigenous fruits should also be enhanced.Developing support networks and a mechanism toupdate knowledge on natural resource managementshould enhance the links between research, educationand extension. It is therefore important to conduct aneeds assessment for training at this level, and to establishbaselines for impact. Another crucial issue is toidentify partners who can contribute to scaling up.New approaches and opportunitiesNew approaches to capacity building should make use of:• people-centred and site-specific approaches to sustainablerural development;• national agroforestry centres of excellence, includingnational agroforestry training teams;• information and communication technology (ICT)for natural resources management programmes;• national agroforestry dissemination centres;• farmer extension skills;• networking between institutions and communitybasedorganizations; and• seminars and demonstration plots.There is a trend for many developing countries to decentralizegovernment administration and create grassrootsownership of the development process through basingresearch on local needs. Furthermore, natural resourcesand agricultural policies are being reviewed to includeagroforestry. These developments provide opportunitiesfor strengthening the role of local communities in policyand strategy formulation, which will help in the scalingup and out of agroforestry innovations. The Centreshould tap into the potential of existing regional andnational education networks and forge links with suchregional bodies as the Forum for Agricultural Research inAfrica (FARA) and its sub-regional organizations.ConstraintsIn many national institutions, the infrastructure to supportagroforestry is not well established. Agroforestryis often handled within forestry or classified under‘agriculture’ and, in many cases, it appears in bothdisciplines. Whatever the institutional arrangementsadopted by different countries, the key issue is theneed for better coordination of agroforestry research,education and development. The sectoral orientationof national agricultural development programmes hasnot helped matters. Thus, in building agroforestry capacity,the Centre should pay attention to aspects ofinstitutionalization. Agroforestry training materials are

164World Agroforestry into the Futureavailable in few languages other than English.For local communities to really benefitfrom them, they need to be translated andadapted to fit local social, cultural and environmentalconditions. In some countries,there are no financial incentives to planttrees. It is important to build capacity atpolicy level to ensure opinion leaders understandand support agroforestry and treeplantingactivities and the links betweenagroforestry and livelihood improvement.Strategic partnershipsThe Centre’s partners include training andextension institutions, non-governmentalorganizations (NGOs) and governmentagencies working with farmers, nationalagricultural research institutions, farmersand farmer organizations, private sectoragro-food processors and private investors.The Centre should identify the differentroles played by these partners andtheir comparative advantages for variousactivities.Further strengthening ofeducational programmesGoalsAgroforestry should be accepted as a broaddiscipline and institutionalized by educationalsystems. This would result in strongercurricula, programmes and delivery ofresources, as well as a team approach toteaching across disciplines. Stronger linksbetween research, education and extensionare essential. Moreover, postgraduateresearch and theses contribute significantlyto research outputs.Strategic issuesAgroforestry should be marketed as amultidisciplinary field of study. An analysisof education policies, institutions andenvironment could identify strengths andweaknesses in current agroforestry teachingand learning. Experiential learning for bothstaff and students is one way forward. Currentefforts to improve curricula should beintensified and, possibly, linked to efforts tobuild programmes in eco-agriculture, biodiversityand environmental management.The greatest challenge is to train more educators.Many institutions still lack adequatehuman resource capacity in agroforestry.Collaboration and sharing (networking)among institutions would help to temporarilyalleviate the shortage, but long-terminvestment in training educators is an imperative.Institutions of learning also needto forge better links with farming communities.For a long time, educational institutionshave taken learning as an objective initself. There is a need for them to link learningto social and economic development oflocal communities.New approaches and opportunitiesThe Centre should form strategic allianceswith funders of graduate studies to increaseagroforestry thesis research. The emphasisshould be on integrated watershed management(multidisciplinary approaches)and participatory development of methodsand tools in agroforestry education.Greater use of information and communicationtechnology (ICT) to distributeteaching materials should be explored andprivate-sector participation should be encouraged.African leaders have expresseda commitment to enhance agriculture andthis is an opportunity to raise awarenessof the benefits of agroforestry, which willlead to its wider inclusion in educationalprogrammes.ConstraintsThe resources and facilities to support agroforestryeducation are generally weak. Insome countries, unfavourable policies andlack of recognition deter people from enteringcareers in agroforestry. National governmentsand institutions should thereforebe encouraged to commit more resourcesto support agroforestry education.Strategic partnershipsThere is a need to work with a broad rangeof partners – farmers, NGOs, industry,policy makers and international agencies,especially those working on tree products.There should be greater private-sectorinvolvement, for example through agroindustryand equipment manufacturers.Links should be cultivated with microfinanceinstitutions and entrepreneurs ineducation and educational technologies, aswell as twinning arrangements with developed-countryinstitutions.Improving national researchcapacityGoalsTo mainstream agroforestry and integratednatural resource management initiativesinto national agricultural research systems(NARS). With good research–education–developmentlinks and sound postgraduateprogrammes at local universities, agroforestryresearch outputs from local institutionswould have greater impact on agriculturaldevelopment.Strategic issuesCurrent research support strategy is to facilitatethe development of better agroforestryresearch programmes by national agriculturalresearch organizations (NAROs).This can happen if agroforestry is acceptedas a priority in solving real developmentproblems. However, agroforestry researchcan only be successful if the various sectors

Chapter 20: Strengthening institutions165involved collaborate. The Centre shouldhelp build capacity for joint resourcemobilization and encourage institutionallinks, especially among different disciplinesand across countries and regions. Bringingtogether researchers in such disciplines asagriculture, forestry and animal sciencesis the key to forming links among differentdisciplines.Another aspect is the need to raise the humanresource capacity, especially throughpostgraduate fellowships. This is a priorityfor many institutions, irrespective of disciplinarybackground. In this context, thesisresearch should be aligned to national prioritiesand, preferably, implemented as partof a national agenda. Weak areas, suchas biometrics and participatory research,require special attention. Follow-up andsupport of alumni are necessary to enhancefuture research collaboration and networking,especially across disciplines.New approaches and opportunitiesIt is vital to establish mechanisms to mentorinstitutions, as well as institutionallinks, partnerships and networks throughNorth–South and South–South collaborativeprogrammes. One way to do this isto establish peer review mechanisms.Colleges, universities, research institutesand extension organizations generallyoperate as independent, de-linked institutionsand would benefit from mechanismsto facilitate better coordination. IntegratedNARS that include research, education andextension, should be promoted.Strategic partnershipsStrategic partners for national research institutionsinclude universities, internationalagricultural research centres, investors,regional economic bodies, advanced researchinstitutions and funders of researchand postgraduate education.Improving knowledgemanagementGoalsTo make stakeholders in agroforestry moreconscious and more active in advancingknowledge management as well as increasingthe application of knowledge todevelopment. This will lead to a better flowof knowledge on agroforestry and naturalresources management through knowledgesystems at all levels.Strategic issuesKnowledge is increasingly playing a centralrole in development. Knowledge managementis therefore very crucial for everysociety and institution. Advances in electronictools for sharing information areenhancing our capacity to gather, analyse,systematize/organize, store, retrieve andapply knowledge. Indeed, the World AgroforestryCentre is lending support to theConsultative Group on International AgriculturalResearch (CGIAR) Learning ResourceCentre (see Chapter 19, in this volume).All materials posted here must be peerreviewedand partners must participate inthe development and management of webbasedtools. However, in some countries theelectronic infrastructure is still weak. Thereis therefore a need to also work with moretraditional tools to make sure stakeholderswith poor access to electronic resources arenot excluded. On-farm practical learningfacilities enhance the participation of localcommunities and guarantee capture and useof indigenous knowledge.Research scientists should adopt knowledgemanagement principles and practices.To achieve this, it is important to mentorresearchers in communication and publicawareness. The Centre should also developa content taxonomy/classificationfor agroforestry knowledge and shape it toaddress the needs of different stakeholders(research and academic institutions anddevelopment partners).New approaches and opportunitiesThere is a need to strengthen the use ofnetworks. Available knowledge is largelyin print format and in English. It shouldbe produced in multimedia formats andtranslated into other languages to facilitateaccess by the majority of stakeholders.ConstraintsIn knowledge sharing, there is a real risk ofbeing technology-driven, rather than demand-driven,in delivery strategies. Manypartner institutions have inadequate informationand communication infrastructureand the Centre should be sensitive to this.Knowledge should be managed in closecollaboration with national institutions, sothey increasingly take leadership. This isnecessary for the long-term sustainability ofthe effort. Maintenance costs for databasesand web sites should be mainstreamed intonational and private systems.ConclusionsThe World Agroforestry Centre continues toplay a leading role in agroforestry capacitybuilding. In the future, national institutionsand regional networks are expected to playincreasing roles. They will only succeed ifstrategies and resources are made availableto facilitate the mentoring of national institutionsand networks. The Centre will needto work closely with its partners, especiallyin developing strategic programmes andmobilizing resources. In this context, farmers’organizations should be recognized asinstitutions and be supported, especiallyin their approaches to dissemination ofknowledge. The Centre’s new ‘StrengtheningInstitutions’ theme is a timely initiativeand will help address the new challenges.

Agroforestry and MajorCross-cutting Issues

168“The biological characteristicsand wide range of resourcetypes provided by trees onfarms makes them invaluable inefforts to promote sustainablerural development and restoredamaged ecosystems.”Sadio and Negreros-Castillo

Keywords:Trees outside forests (TOF), agroforestry, livelihoods,sustainable rural development, improving the globalenvironment, landscape restorationChapter 21Trees outside forests:Facing smallholder challengesSyaka Sadio, Food and Agriculture Organization of the United Nations and Patricia Negreros-Castillo,Iowa State University, USAAbstractDuring the past three decades, rapid population growth has been accompanied by increased demandfor agricultural land and forest products. As a result, widespread degradation of natural resources, desertificationand loss of biodiversity has occurred, together with food insecurity and extreme povertyin the most vulnerable areas. Trees offer great potential to restore degraded ecosystems, to enhancelivelihoods based on production of food and medicines and to provide environmental and economicbenefits. It is now agreed that a bold approach is needed to promote tree planting and improve themanagement of trees outside forests (TOF). The sustainable management of TOF resources will play anincreasing role in efforts to reduce land degradation and poverty and improve food security, if appropriatepolicy measures are taken and farmers are placed at the centre of development efforts.IntroductionDuring the past three decades, rapid populationgrowth and expansion of agriculture has led towidespread degradation of natural resources. Thecumulative effects of this expansion have producedserious global environmental, social and economicproblems, including loss of biodiversity and extremepoverty among people living in the more vulnerableareas.Trees have the potential to rehabilitate degraded landsand ecosystems, restructure the landscape, provide arange of benefits and products (wood and non-woodproducts for food and medicines), and render environmentaland socioeconomic services. The importanceof trees in addressing these issues has been wellunderstood by farmers through the centuries and hasbeen clearly demonstrated in traditional tree-basedagricultural farming and land-use systems, such asshifting cultivation in the humid tropics and grazing inthe semi-arid savanna areas.The concept of trees outside forestsAlthough they play an important role in environmentalprotection, landscape restructuring andlivelihoods, trees on non-forest land received littleattention from scientists, practitioners, planners, decisionmakers and policy makers until the early 1990s.At the Kotka meeting in 1993 (FAO 2000), expertsrecognized the uniqueness and importance of differentresource types and defined a new concept of‘trees outside forests’ (TOF), which considers trees atall levels and in all aspects, emphasizing their social

170World Agroforestry into the Futureand economic role and promoting theirconsideration at the policy level and inthe assessment of the world’s forest resources.‘Trees outside forests’ are defined as ‘treesfound on non-forest and non-wood lands’,such as: agricultural lands, urban and settlementareas, roadsides, homegardens,hedgerows, pasture/rangelands and scatteredin the landscape (FAO 2000). Theconcept recognizes the biological characteristicsof trees and their ability to provideenvironmental, social, cultural and economicbenefits.Since 1999, the Food and AgricultureOrganization of the United Nations (FAO)has been working to raise awareness of TOFand to dispel the idea that tree resourcesare only important for small-scale farmersor those that make a limited contribution tosustainable forest resource management.Multiple roles and benefits oftrees outside forestsFood and other essential goodsTOF have been called ‘trees that nourish’,particularly for many poor and landlesspeople who obtain essential productsfrom them (Manu and Halavatau 1995).Many tree species found in African andAsian agroforestry systems (e.g., Borassusaethiopum, Balanites aegyptiaca, Ziziphusmauritiana) are planted for their ability toproduce large quantities of food and othernon-wood forest products (see Box 1).Wood and fuelwood productsIn developing countries, TOF providesa large proportion of the fuelwood fordomestic energy (cooking and heat). Forexample, 50 percent of fuelwood used inThailand, 75–85 percent in Indonesia, Java,Pakistan, the Philippines, Sri Lanka andVietnam and 83 percent in Keralla, India(FAO 2001a; Jensen 1995) is harvestedfrom farmland and other non-forest land.In many countries, farmers and smallholdersplant trees (especially valuable timberspecies) as a means of saving for the future(FAO 2001a; Negreros-Castillo and Mize2002).Fodder supplyMany pastoralists use TOF as a source offodder for livestock including cattle, camels,sheep and goats (see Box 2). Recentcase studies from Latin America have highlightedthe importance of trees in livestockproduction, since they provide shade, shelterand supplementary fodder, particularlyin the dry rangelands and low forest coverareas (Sanchez et al. 1998). In the aridparts of sub-Saharan Africa, three-quartersof the 10 000 woody species that grow insilvipastoral systems are thought to be usedas fodder, supplying up to 50 percent oflivestock feed (FAO 2001a), particularly inthe dry season when grass and crop residuesare not available.Sustainability of agriculturalproductionThe most universally recognized roleplayed by trees in agricultural systems isthat of soil conservation and the replenishmentof soil fertility (Chivaura-Mususa et al.2000; Sanchez et al. 1998). In many partsof the world, improved tree-based systems,e.g., shelterbelts, windbreaks, alley cropping,hedgerows and tree cover crops, e.g.,coffee, cacao, coconut, olive and citrushave been integrated within croplands(Huaxin 2001; IRDC 2000; Jensen 1995).The use of such nitrogen-fixing woodyspecies in agroforestry parklands in WestAfrica as Acacia albida, Vitallaria paradoxaBox 1. TOF for foodBorassus aethiopum (Fan palm)The kernel and mesocarp of the fruitcan be eaten raw or roasted and providesnourishing carbohydrate, proteinand minerals. The jelly-like, immaturefruit is delicious, and a decoction of theroots makes a cooling drink for infants.Sprouts grown from the nuts are commonlyeaten as a vegetable, the sap isused to make wine and the leaves aresuitable for weaving roofs and fences.Vitellaria paradoxa (Shea nut tree)The pulp is eaten raw and suppliescarbohydrate, minerals and vitamins.The kernel supplies oil used in cooking,cosmetics, candles, and even forwaterproofing the walls of farmers’homes. In Burkina Faso, annual yields of48–65 kg ha –1 of fresh nuts are common.Some 40 000–75 000 t are exported toEurope and 10 000–15 000 t to Japan,where they are used in cosmetics, pharmaceuticalsand baking.Source: FAO (2001a)Box 2. TOF for fodderA 1989–1990 survey in Bamako, Malifound that home-reared sheep were fed1.8 kg of Pterocarpus erinaceus andKhaya senegalensis leaves each day,and over 1400 t of fresh Pterocarpuserinaceus leaves were sold in Bamakodaily as feed for small ruminants. In SriLanka, leaves of Gliricidia sepium area popular fodder for goats and sheep.Fodder products from trees providecarbohydrates, nitrogen, magnesium,potassium and oligo-elements.

Chapter 21: Trees outside forests 171and Acacia senegal is a good illustration ofa traditional tree-based farming system thathelps to restore soil fertility and structure.Environmental services and socioculturalvaluesTOF improve air quality and the microclimate,particularly in urban areas. Treesare a valuable carbon ‘sink’ (World Bank2002) and help to reduce soil erosion bychecking wind velocity and water runoff.They are highly valued by many communitiesliving in hot climates as providers ofshade and have significant symbolic, social,religious and cultural status. They alsoprovide many other environmental benefitsand services, such as minimizing the lossof mineral elements through leaching andimproving soil structure.The way forwardIt is generally agreed that increasing agriculturalproductivity is central to growthand poverty alleviation in rural areas. It isalso well understood that increasing productionoften results in the destruction offorest cover, depletion of fertile soil andsevere land degradation.At national and international levels, theimportance of TOF as a resource is oftenoverlooked. Only limited initiatives existedprior to the Kotka meeting in 1993. Sincethen, many case studies conducted at nationaland regional levels have identifiedthe serious livelihood challenges faced bysmallholder farmers. To be effective, treeissues need to be addressed in a holistic,people-centred vision, focusing on themultiple functions of trees.Population pressure on land andforest resourcesIncreasing population pressure on limitedagricultural land has led to the breakdownof many traditional tree-based systems.The result has been diminishing vegetationcover, soil erosion and reduced agriculturalproduction. One of the greatest challengesis that of preventing further erosion of forestcover. Promoting TOF systems withinthe context of land conservation and restorationis one of the most effective ways toimprove productivity of the existing agriculturalland area, thereby limiting pressureon remaining forest resources.Appropriate policy to meet localdevelopment needsMillions of vulnerable people living in ruraland peri-urban areas rely heavily on treeresources for their livelihoods, but they lackan effective voice in decision making. Anurgent challenge is that of extending smallholders’rights in order to give TOF moreprominence as a route towards more sustainablelivelihoods. Farmers will plant moretrees when they are given policy and marketincentives. Formal acknowledgement of userrights over trees growing on farmland wouldprovide a major incentive for the conservationof such trees. On-farm tree managementgenerally makes good economic sensewhen benefits are taken into account in a‘whole farm’ evaluation approach.The best solution would be to integrate TOFinto national agriculture and forestry developmentplans, by focusing on the needs oflocal people to create their own woodlots,protect their environment and improve theirlivelihoods. In order to convince the governmentaldecision-makers and planners toinclude TOF in their national policies andtree planting programmes, however, it is essentialto demonstrate the benefits of trees inthe national economy.Farmers will need to be supported with appropriatelegal measures, market incentivesand the removal of barriers to land accessand tree tenure. National forestry lawsseldom favour small-scale on-farm treeplanting and private investment in forestryis often limited by rigid land tenure systemsand restrictions (FAO 1993). Legal changesto land and tree tenure are critical for TOFpromotion because they secure benefits forthe stakeholders.We believe that a bold approach, that givesstakeholders a voice and the right to decidewhat is necessary, should be the focus ofany policy targeting sustainable development.This requires simultaneous action,however, such as multiple-scale activitiesand careful consideration of how socialand political changes influence the successof different interventions and managementpractices.Building knowledge and capacitiesThe past two decades have witnessedthe development of techniques to designlandscape mosaics based on tree-cropintegration and environmental protection.However, results of many recent case studies,meetings and workshops (FAO 2001a;FAO 2002) highlight the need for furthertechnologies and research to enhance theproper use of trees in landscape restorationand for increasing environmentaland economic viability. In addition, as apriority action, there is a need to evaluatetraditional knowledge and practices for treemanagement. A combination of naturalresource management approaches is likelyto be required.Gender-differentiated managementWomen are the first to be concerned withthe selection and harvesting of non-woodforest products (leaves, roots, fruits, etc.)and have good overall knowledge of theiruse, conservation and processing. A studyin Java showed that 60 percent of a ruralfamily’s food typically comes from home-

172World Agroforestry into the Futuregardens in which trees are prominent, andthat these gardens are mostly managed bywomen (FAO 2001b). This important skillshould be taken into consideration andseen as reason enough to enhance andstrengthen women’s roles in TOF resourcemanagement.ConclusionsThe biological characteristics and widerange of resource types provided by TOFmakes them invaluable in efforts to promotesustainable rural development andrestore damaged ecosystems. They can alsohelp improve the environment of urbanareas and raise yields and profits on lowproductivityagricultural land. Finally, if thenew vision of TOF is embraced by manynations and expressed in the form of policieslinking trees to all land-use systems,the accumulated effect will be reflectedin a significant improvement in the globalenvironment and progress towards sustainablerural development.ReferencesChivaura-Mususa, C., B. Campbell and W.Kenyon 2000. The value of mature treesin arable fields in the smallholder sector,Zimbabwe. Special section: land useoptions in dry tropical woodland ecosystemsin Zimbabwe. Ecological Economics33(3): 395–400.FAO 1993. Forest Policies in the Near EastRegion: Analysis and Synthesis. Food andAgriculture Organization of the UnitedNations, Rome, Italy.FAO 2000. Trees Outside Forests. Conceptpaper. Forest Conservation Working PaperSeries. Food and Agriculture Organizationof the United Nations, Rome, Italy.FAO 2001a. Trees Outside Forests. ConservationGuide No. 35. Forest Conservation Service,Rome, Italy.FAO 2001b. State of the World’s Forests. Foodand Agriculture Organization of theUnited Nations, Rome, Italy.Huaxin, Z. 2001. Agroforestry in arid and semiaridareas of China: A country report. In:Proceedings of First Regional Workshopon TPN2. Regional Programme underUNCD/ICRISAT. 18–21 December,Patancheru, India.Jensen, M. 1995. Woodfuel productivity ofagroforestry systems in Asia. FAO FieldDocument No. 45. Food and AgricultureOrganization of the United Nations,Bangkok, Thailand.Manu, V.T. and S. Halavatau 1995. Agroforestryin the food production systems in theSouth Pacific. Proceedings of AustralianCentre for International AgriculturalResearch 56: 63–68.Negreros-Castillo, P. and C.W. Mize 2002.Enrichment planting and the sustainableharvest of mahogany (Swietenia macrophyllaKing) in Quintana Roo, Mexico.In: Lugo, A., J. Figueroa-Cólon and M.Alayón (eds) Big-Leaf Mahogany: Genetics,Ecology and Management. SpringerVerlag, New York, USA.IRDC 2000. Emerging products and services fromtrees in lower rainfall areas. ResearchUpdated Series 2 No. 00/171. RuralIndustries Research and DevelopmentCorporation, Kingston, Australia.Sanchez, P.A., R.J. Buresh and R.R.B. Leakey1998. Trees, Soils, and Food Security. LandResources on the Edge of the MalthusianPrecipice? CAB International in associationwith the Royal Society, New York,USA and Wallingford, UK.van der Linde, J. 1962. Trees Outside theForest. Forestry and Forest Products Studies15: 139–208. Food and AgricultureOrganization Conservation guide 35,Forest Conservation Service, Rome, Italy.World Bank 2002. Launch of $100 million biocarbonfund provides new opportunitiesfor rural poor. News Release 2003/133/S:Tokyo, Japan.

Keywords:Gender, property rights,tenure, agroforestryChapter 22Women, land and treesRuth Meinzen-Dick, International Food Policy Research InstituteAbstractThe distribution of rights to land and trees between men and women has important implications foragricultural productivity, women’s empowerment and household welfare. Rights to land and trees tendto shape women’s incentives and authority to adopt agroforestry technologies more than other cropvarieties because of the relatively long time horizon between investment and returns. Some of the complexitiesinvolved in understanding women’s control over resources are discussed, not only in the mainagricultural fields, but also in important ‘interstitial spaces’ that are often overlooked. There are manyfactors affecting distribution of rights and, often, simply changing legislation may not change rights inpractice. However, this heterogeneity offers considerable potential for women and outside agencies towork together to strengthen women’s rights over land and trees.IntroductionThe World Agroforestry Centre has conducted pioneeringwork on ‘women, land and trees’, so this is an appropriatesubject for a 25 th Anniversary publication. Itis also appropriate to look ahead, because these issuescontinue to affect many of the research themes relatedto agroforestry. Among its many accomplishments, theCentre and its partners have been successful in increasingunderstanding of people’s rights to trees as separablefrom (but linked to) rights to land, and how rights toboth land and trees are affected by gender (Fortmannand Bruce 1988; Fortmann and Rocheleau 1985).The recognition of rights to trees has important implicationsfor agroforestry adoption, particularly wherewomen are restricted from planting or have low incentivesto plant certain types of trees. But, rather thanusing these obstacles as a reason to focus on men (whoare often easier to reach) the Centre, to its credit, recognizeswomen as major clients who are likely to playan increasingly important role in agriculture. Research-ers have therefore redoubled their efforts to understandwomen’s use of agroforestry and to ensure that newtechnologies and dissemination programmes reachthem. For example, Rocheleau’s research (Rocheleau1988; Rocheleau and van den Hoek 1984) showed theimportance of spaces that are often overlooked, such aspatio gardens and the ‘interstitial spaces’ of hedgerows,roadsides, and the space between trees on ‘men’s’land. Programmes such as biomass transfer have builton this information and developed technologies thatuse the resources that can be accessed and controlledby women. At the same time, there has been carefulattention to men’s incentives to plant and maintaintrees. For example, the Centre’s collaborative researchin Malawi and Uganda identified links between treerights, tree density and marriage and inheritance patterns,where patterns of matrilineal inheritance (landpasses to nephews rather than sons) means that menmove to their wives’ villages, where they have less incentiveto plant trees, because the trees would not passto their own children (Otsuka and Place 2001). Thus,

174World Agroforestry into the Futurecareful attention to gender relations, mobilityof residence and control over resourcesin different locations is needed to maximizethe impact of agroforestry programmes.Why are women’s rightsimportant?Property rights for women are important foragricultural productivity, women’s empowermentand household welfare. Rights toland and trees tend to shape women’s incentivesand authority to adopt agroforestrytechnologies more than other crop varietiesbecause of the relatively long time horizonbetween investment and returns. Studies inCameroon, Kenya, Mali, Uganda and Zambiahave found that tenants without longtermland rights are restricted in their rightsto plant or harvest from trees because ofinsecurity of tenure (Place 1994). In communalareas of Zimbabwe, Fortmann et al.(1997) found that the potential loss of landand trees following widowhood or divorcetended to make women feel insecure andlimited the amount of trees they planted onhousehold land. At the same time, womenand men were equally likely to plant treeson community woodlots because rightsover those trees derived from communitymembership and investment, not maritalstatus, and hence there were fewer genderdifferences in tenure security. However, differencesin socioeconomic status of householdswere a bigger factor than gender differencesalone in explaining tree plantingbehaviour overall: the poorest householdshad least ability to plant, and tended tofocus on trees for subsistence needs, ratherthan commercialization.Because credit, extension, and other servicesare generally directed preferentially tolandowners, the agricultural productivityof women without land rights is furtherrestricted by a lack of complementaryinputs. This is particularly a constraint foragroforestry, where access to informationand credit are important for adoption.Where women obtain land through theirhusbands (for example, in the dual-farmingsystems of Africa), they are often requiredto work on their husbands’ fields in orderto obtain their own plots for growing food,which restricts labour availability on women’sfields. When men are absent (due tomigration, divorce or death), the problemsof labour shortage are even more serious(Davison 1988; Lastarria-Cornhiel 1997).The high labour requirements for someagroforestry systems pose a particularproblem for women, who not only faceless control of others’ labour compared tomen, but also have high labour requirementsfor domestic tasks. Studies indicatethat reducing the gap between men’s andwomen’s control over capital and inputscould increase agricultural productivity insub-Saharan Africa (SSA), for example, by10–20 percent (Quisumbing 2003).There is a growing body of empirical evidenceindicating that property rights raisewomen’s status in the household as well asin the community, and this translates intogreater bargaining power (Agarwal 1994;Quisumbing 2003). When women controlresources, such as land and trees, they aremore likely to be managed in a way that isconsistent with women’s priorities. Longtermrights to land and trees also providesecurity for women in the event of widowhood,divorce or family crisis. Whetherbecause of this improved fallback positionor because of higher status, studies in variouscontexts have found that women withcontrol over land have more influence overdecisions at home, and may be subjectedless to domestic violence.The higher income and stronger bargainingpower of women with control overresources has implications for the distributionof welfare within the household,since women and men spend incomeunder their control in systematically differentways (Quisumbing 2003). Womenare observed to spend a higher proportionof their income on food and health care ofthe children (particularly of girls), whichhas important implications for overallfamily welfare and long-term poverty reduction.Thus, improving women’s statusand resources improves child health andnutrition (Smith et al. 2003). Trees can playa role in this, not only as an asset that mayincrease women’s bargaining power, butalso as a source of food, medicines, andfuelwood, which all tend to be importantto women.The complexity surroundingwomen’s rightsIt is one thing to recognize the importanceof women’s rights, but quite another tostrengthen them, particularly concerningland and trees. It is important to beginwith an understanding of existing rightssystems, which often involve complexrelationships between different uses andusers of the resources. Rather than simple‘ownership’ of resources, we often findseparate bundles of rights; for example,one person may have the right to plant atree and use its fruits, another to grow anannual crop on the land around the trees,and a third to graze their flocks on theland in the dry season. In other situations,one person has the right to use the land,but another holds the controlling or decision-makingrights. The different rights maybe held by different households (landlordand tenant), or even by different memberswithin a household (husband, wife andchildren). The duration of rights also varies,from a growing season (or less) to thelong term.

Chapter 22: Women, land and trees175It is also essential to consider the robustnessof rights, that is, their ability to withstandchallenges from others. For example,in Mozambique, Vijfhuizen et al. (2003)found that women do not usually planttrees when they live in their husband’sfamily homestead because they do notfeel they have secure tenure there. However,they do plant trees when they moveto an independent homestead with theirhusbands, or when they are allocatedtheir own land by the chief or in-laws. Inthe case of women’s rights over land, weoften find (particularly in SSA), that womenacquire the right to cultivate land fromtheir husbands, fathers or sons, but theircontrolling or decision-making rights arerestricted. This is especially problematicfor agroforestry, because planting trees is amanagement (decision-making) right that isoften restricted to landowners, particularlymale ones. Moreover, if women acquirerights through a man, their rights are highlydependent on their relationship with thatman and, if they become widowed or divorced,they may lose those rights. Evenwhen married women have land use rights,they often have to get permission from theirhusband to plant a tree. As a result, femaleheadedhouseholds with land may be morelikely than women in male-headed householdsto adopt agroforestry, because theyhave more autonomy (Gladwin et al. 2002;Hansen et al. 2005).Many analyses of land rights focus on themain agricultural plots or on residentialand commercial property. To understandwomen’s property rights, and particularlytheir rights over trees, it is essential to lookbeyond these to consider tenure and treeswithin the whole landscape. In many cases,common lands (which may be officially designatedas community or state property) areimportant sources of trees and tree products,particularly for women. For example, in thePhilippines, Flora (2001) found that womendepended on the commons more heavilythan men for domestic and market-orientedproduction, but men’s interests tended toprevail where markets had developed, bothin crop production and in land tenure, andwomen often lost access when land wasprivatized.Rocheleau and Edmunds (1997) drawattention to the importance of such ‘interstitialspaces’ as homesteads and patiogardens (trees planted between the houseand fields), hedgerows (trees or bushes betweenfields and roads or other fields), andinter-cultivation of annual crops betweenplanted trees in fields. These areas producevaluable products, such as wood, fodder,vegetables, medicines and wild foods; butthis type of production may be ignored bygovernment statistics or even by researchers.Interstitial spaces are particularly importantfor people who have little controlover the main farmland. In addition to theirproductive and livelihood values, they playimportant ecological roles, particularly inwatersheds, where stream banks, hedgerowsand wetlands act as filters and sinksfor reducing water pollution and controllingsoil erosion (Swallow et al. 2001).Rights over these interstitial spaces areoften not clearly defined, partly becausethey are boundary areas and attempts todefine them may generate conflict. Onthe one hand, unclear rights give accessto the landless, but on the other, it meansthat responsibilities are also not clearly defined.Thus, it is important to consider howmen and women relate to these interstitialresources, both in terms of their use of theproducts and their involvement in maintainingthe resource.Another important complication in rightsover land and trees arises from the multiplesources of claims for property rights. Statetitle is only one (albeit important) source ofproperty rights and it is one to which manyAfricans have no access. However, claimsmay also be based on a range of customaryor religious laws or even local norms. Forexample, a country may have laws specifyingthat all children are entitled to inheritan equal share of assets from their parents,while Islamic law specifies that daughtersreceive half the share of sons, and localnorms may prescribe that women shouldnot cultivate their land, but give it to theirbrothers. Even where there is agreementon rights, it may be difficult for people tophysically use their claims, particularly inthe face of social pressures. On the otherhand, men and women use their socialconnections to access land and the labourto farm it, particularly for agroforestry.Thus, both state law and local norms, particularlythe interplay of gender and powerrelations, play a crucial role in shapingwomen’s rights to land and trees.Figure 1 illustrates how state law and genderrelations are linked to women’s rightsover land and trees, and how these, in turn,are linked to access to inputs, agroforestryadoption, agricultural productivity andhousehold welfare.Sources of change in women’srights over land and treesIn spite of their complexity, property rightsare highly dynamic; they change over time,as a result of external and women’s ownactions. Most of the linkages are two-wayrelationships. For example, women’s landrights may influence agroforestry adoption,but agroforestry adoption can also affectland tenure. Similarly, changes in householdincome or welfare can have feedbackeffects on women’s rights to resources.Figure 2 illustrates how each of the links towomen’s rights over resources can become

176World Agroforestry into the FutureState lawAgroforestryadoptionCustoms,gendered powerrelationsWomen’s rightsto land, treesAccess to inputsinformation, etc.Figure 1. Links to women’s rights over land and trees.HouseholdwelfareProductivity,incomeState lawCustoms,gendered powerrelationsAgroforestryadoptionAccess to inputsinformation, etc.Productivity,incomeFigure 2. Ways of influencing women’s property rights.a point of leverage to increase genderequality. We now consider each of thesein turn.Legal reformsLegal reforms have received considerableattention from women’s movements andothers advocating gender equality. Certainlychanging state law is an important meansof strengthening women’s property rights. Insome cases this involves repeal of discriminatorylaws, such as those of Lesotho andSwaziland that declare women to be legalminors, and therefore unable to hold property.Other legal mechanisms to strengthenwomen’s property rights include reformof inheritance laws or provision for jointtitling of land in the names of husband andwife that can provide women with strongerclaims on the land while their husbands areliving, and greater security in the case ofwidowhood or divorce. While much of theattention on legal reforms has focused onprivate lands, strengthening recognition ofrights over public or common lands – andin many cases, preventing privatizationof the common resources that many poorwomen depend upon – can be an equallyimportant mechanism for securing women’saccess to resources.However, legal changes do not automaticallytranslate into changes on the ground.For new property rights laws to becomeeffective, both the implementers of thelaw and the public need to be aware of

Chapter 22: Women, land and trees177the changes. Legal literacy campaigns aretherefore needed for both audiences, andwomen need to have access to administrativeand judicial channels. Access can berestricted physically, by women’s morerestricted mobility due to childcare orsafety concerns. The greater the distancepoor rural women have to travel to registera claim or fight for it in court, the lesslikely they will be to use formal channelsto claim their rights. Social distance alsomatters, for when women have less educationor social standing than those to whomthey must go to claim their rights, they areless likely to seek redress. For this reason,in Tanzania and Uganda, land managementand adjudication bodies at all levelsare required to have female representation.However, these bodies require resourcesto be effective, and governments may nothave the personnel, funds or political willto provide for them at the local level. Evenwith reforms in legislation and implementingbodies, women’s rights may not changedramatically if there is major differencebetween statutory and customary law, butstate law does provide a basis upon whichthey can appeal. Eventually, with enoughappeals to such laws, even social normsand power relations can change.Community programmesCommunity programmes offer the potentialto change gendered power relations directly.Many programmes run by governmentor non-governmental organizations (NGOs)use group-based approaches for a rangeof social development or natural resourcemanagement (NRM) goals. In many cases,the programmes introduce new discourseabout gender relations and resource rightsthat affect local customs. If women areincluded in such groups, either through allwomen’sgroups (as in many microfinanceor education programmes), or throughactive participation of women in mixedgendergroups (as in some, though not all,forestry groups), it can have an empoweringeffect on them. The interaction betweenmen and women in such groups can alsohelp to change gender-related norms andeven power relations. However, in othercases, it can cause a backlash by men towardmore ‘traditional’ practices.Collective action programmesCollective action programmes can havea direct influence on women’s propertyrights, as when women group together toacquire land and then cultivate it individuallyor collectively (Agarwal 1994; Rocheleauand Edmunds 1997; Schroeder 1993).Women’s participation in NRM groups hasa direct bearing on their property rights.Under many devolution programmes, usergroups are charged with managing resourceslike forests, irrigation, or watersheds thathad been under (nominal) state control.Although many government programmesare reluctant to give away explicit rightsover resources (preferring to focus on theresponsibilities that user groups should undertake),when user groups take over, theyBox 1. Collective action in Nepaleffectively acquire management and oftenexclusion rights over the resource. Thus,effective participation in the user groups isimportant to ensure that the resources areused in a way that meets women’s needs.Without women’s participation, for example,watershed management groups in Indiaoften close off areas for grazing or fuelwoodcollection in such a way that poorwomen lose access and bear the greatestcosts, while the benefits go to the richerlandholders (often men) downstream. Kerr(2002) found that where NGOs had givenspecial attention to social organization(especially among women and marginalizedgroups) before starting to addresswatershed management, those groups werebetter able to articulate their interests, andthe programmes were more equitable andmore sustainable. However, collective actionprogrammes can be highly complex(see Box 1).Agroforestry interventionsAgroforestry interventions have considerablepotential to influence property rightsbecause planting or clearing trees is aThe complexities involved in collective action programmes are illustrated by programmesin Nepal, which provide 40-year leases of degraded forest to groups of poor people. Thepilot project included 25 percent women members, but even when women were not themembers, both husbands and wives were involved. Gender training for all project staffand hiring women as group promoters were seen as instrumental to project success. Inaddition to significant regeneration of vegetation, the focus on developing fodder andfuelwood sources that were of particular importance to women reduced household timespent on gathering them by an average of 2.5 hours per day (Brett et al. 2004). Womenreported that they became more empowered, in part because they learned to have a sayin group decision making. However, granting more secure tenure to a subset of the poorestpeople challenged local power structures and excluded others who had customaryuse rights over that land (Baral and Thapa 2003). Without strong local organizations toenforce rules, there were conflicts that decreased tenure security.

178World Agroforestry into the Futurecommon means of establishing claims, notonly on the trees, but also on the underlyingland, particularly in customary Africanland tenure systems. Indeed, so potent istree planting for establishing tenure claimsthat in some societies, women are prohibitedfrom planting trees, or may be restrictedin the species that they can plant. In suchcases, agroforestry programmes that introduceshrubs or species that women cancontrol are more likely to strengthen women’srights to both land and tree resources.Place (1994) suggests that agroforestrycan even be instrumental in changing socialnorms about women’s tree planting,by blurring the distinction between treesand perennial crops. Types of trees thatmeet women’s needs and fit within theirresource constraints are also more likely tobe adopted. For example, trees that replenishsoil fertility have been welcomed bywomen who lack the cash to buy fertilizers(Gladwin et al. 2002). The World AgroforestryCentre’s recent work on fruit andmedicinal trees is particularly promising,because these are important to women.When gender relations regarding trees areignored, programmes can even weakenwomen’s rights, as in The Gambia, whena tree planting scheme introduced for ‘environmentalrehabilitation’ and targetedto (male) landowners pushed out highlyproductive women’s gardens that were cultivatedon that land (Schroeder 1993).Credit, information and inputsCredit, information and inputs helpwomen to acquire land and invest in trees.Strengthening women’s land rights aloneis not enough; other constraints need to beaddressed if women are to be able to usethe land in a productive and sustainableway. In many cases, women’s yields areless than those of men because they haveless access to seeds, fertilizer and labour(Quisumbing 2003). The land that womenacquire is often less productive because oflow soil fertility or lack of water. Thus, extensionand other programmes that explicitlyseek to redress gender imbalances canhelp women to use their land productively,rather than mortgaging or renting it out.Community nurseries (often managed bywomen’s groups) can provide planting materialsand the knowledge needed to growthem effectively. Microfinance programmeshave targeted poor women in many countries,but, because of the small size ofavailable savings or loans, often cannothelp women to purchase land. However,microfinance can help them to buy trees(which can become a form of savings account)or other necessary inputs. Many extensionsystems bypass women, particularlywhen they are not the landowners, andhence they may not acquire informationabout improved practices, including agroforestry.Recognizing this constraint, theCentre’s biomass transfer and improved fallowsprogrammes in Zambia and westernKenya have used group-based approachesand simple dissemination materials to ensurethat women are included (Gladwin etal. 2002; Place et al. 2003). The use of amulch from cut branches of Tithonia shrubsfrom hedgerows (interstitial spaces, whichare under women’s customary usage) allowsthose with limited land to adopt thisapproach to soil fertility enhancement.In practice, these types of change do notexist in isolation, but interact. Quisumbingand Otsuka’s (2001) study of the evolutionof land tenure in western Ghana providesan apt illustration (see Box 2).Challenges for research andactionRecognizing the importance of women’srights over land and trees for productivity,equity and household welfare is only thebeginning. Much remains to be done inboth research and practice to achieve genderequity in property rights. The first stepis to make sure that our understanding ofthe issues is adapted to the local context.The processes involved are complex, andwill vary from place to place and accordingto a host of intersecting identities, includingreligion, ethnicity and culture. Asthe example from Ghana indicates, broadgeneralizations can be misleading. Thus,it is essential to identify the key gender/tenure interactions at each site, and howthese relate to the use of resources and distributionof welfare in society (and withinhouseholds). Although the specific answerswill differ from place to place, Gladwinet al. (2002) show that there are manycommon factors facilitating or limiting theplanting of trees, indicating a core set ofquestions that researchers can ask about.Cost-effective diagnostic tools are nowavailable for assessing customary as well asstatutory rights (Freudenberger 1994), so attentionto these issues is no longer restrictedto the research community that can undertakedetailed study. Focus group meetingsand key informant interviews may start withmapping the local resources, then discussingwho uses each resource and the rulesgoverning that use. Time and trend linescan indicate how access to these rights haschanged over time. However, it is essentialto discuss these issues with women andmen, younger and older generations, andto look beyond the private farmlands to thecommons and even the in-between spacesand resources. With such an approach, appliedprojects can and should develop anunderstanding of the distribution of rights toland and trees in each site.Although our understanding of propertyrights is growing, there is a need for furtherresearch on the complex interactions

Chapter 22: Women, land and trees179Box 2. Evolution of land tenure in western GhanaRapid population growth in western Ghana has put unsustainable pressure on customarysystems of acquiring land by clearing forests. As a result, agroforestry (particularlycocoa production) became more profitable than shifting cultivation, which created localpressure to individualize land tenure. While individualization of tenure frequently led towomen losing their customary access to land (Lastarria-Cornhiel 1997), in this case theintroduction of cocoa increased demands for women’s labour. Men needed to provide incentivesfor their wives to work in the cocoa fields. Although land was customarily heldonly by men, women acquired use rights through their relationships with men, and traditional‘gifting’ ceremonies, witnessed by the community, were adapted so husbandscould transfer individual land rights to their wives in exchange for labour on the cocoafields. Thus, customary practices were used to adapt the land tenure and give womenrelatively secure rights to land and trees. (While this represents a significant advance inwomen’s rights to land, it does not represent full equality. Women had to plant 40–50percent of the land to cocoa before receiving rights to it, whereas men only had to plant20–25 percent of the land before receiving the rights.) At the same time, the statutory lawwas changed by the 1985 Intestate Succession Law, which provides for a wife and childrenif a man dies without a will. According to the new law, the distribution of assets wasto be 3/16 to the spouse, 9/16 to the children, 1/8 to the parents and 1/8 to the matriclan(mother’s extended family). The common interpretation of this distribution, however,was 1/3 each to surviving spouse, children and matrilineal family. Thus, the local law waseven more favourable towards women than the formal statute, and legal reforms cameafter changes in local practice.Source: Quisumbing and Otsuka (2001).between land, trees and water, and on howproperty rights influence the managementof these interacting resources. The impactsof these interactions are seen on individualand household welfare and on the landscape.This is particularly important inwatershed management. Here, the Centre’swork (Swallow et al. 2001) brings togetherbiophysical and social scientists to addressthese complex relationships. The work indicatesthat it is not only private land thatmatters; in many cases, collective rights (orlack of them) to critical landscape features,especially water supply points, wetlands orriver banks, has implications for women.For example, work in the Nyando basin ofKenya indicates how privatization of most ofthe land, including that adjoining rivers, hasrestricted access to water supplies. Whenthe owners of land around springs are persuadedto set aside the land and plant nativetree species for spring protection, this canhave important benefits for women’s timeand resource use, along with environmentalbenefits in terms of improved water qualityand reduced soil erosion.Moving from research to practice, identifyingeffective ways to strengthen women’srights to resources remains a key challenge.The framework presented in this chaptersuggests several different intervention pointsthat can enhance (or weaken) women’s accessto and use of land and trees. However,much remains to be done to assess the effectivenessof alternative intervention strategies,and to understand the ways in whichdifferent interventions interact with strategiesundertaken by women. External policies andlegislative reforms can influence change inlocal norms, but changes in local norms canalso influence the implementation of policies.Technologies (including agroforestry)that increase the returns to women’s labourcan strengthen their bargaining power withinthe household. Supporting programmesto disseminate technologies and complementarycredit and inputs can enhance thisprocess, enabling women to use their landand trees more effectively.The relationships among gender, tenure,technologies and household welfare arecomplex. Rather than shying away fromthis complexity, development researchersand practitioners need to understandthe relationships and how they are likelyto affect outcomes of projects in any particularcontext. This web of interactionsmeans that any single intervention, suchas legal reform, is not likely to achieve anobjective by itself, but it also offers multiplepoints through which women’s rightsover resources can be strengthened, withimportant implications for the adoption ofagroforestry technologies and agriculturalproductivity, and also for women’s empowermentand overall household welfare.AcknowledgementsI am indebted to Agnes Quisumbing andFrank Place for generously sharing their researchand insights with me, and to DianeRussell, Carin Vijfhuizen, Paul Hebinck,Marcus van Maanen and Marleen Nooij forhelpful reviewers’ comments. Responsibilityfor any errors is mine.

180World Agroforestry into the FutureReferencesAgarwal, B. 1994. Gender and command overproperty: A critical gap in economicanalysis and policy in South Asia. WorldDevelopment 22: 1455–1478.Baral, J.C. and Y.B. Thapa 2003. Nepal’s LeaseholdForestry for the Poor: Looking at theunintended consequences. MountainForum http://www.mtnforum.org/resources/library/barax03b.htmBrett, N., F. Ohler and J.K. Tamrakar 2004.Providing the Poor with Secure Access toLand in the Hills of Nepal. Case Study forScaling Up Poverty Reduction: A GlobalLearning Process and Conference, May25–27, 2004, Shanghai, China.Davison, J. (ed) 1988. Agriculture, Women, andLand: The African Experience. WestviewPress, Boulder, USA.Flora, C.B. 2001. Access and control of resources:Lessons from the SANREM CRSP. Agricultureand Human Values 18(1): 41–48.Fortmann, L., C. Antinori and N. Nabane 1997.Fruits of their labors: Gender, propertyrights, and tree planting in two Zimbabwevillages. Rural Sociology 62(3): 295–314.Fortmann, L. and J.W. Bruce 1988. Whose Trees?Proprietary Dimensions of Forestry. WestviewPress, Boulder, USA.Fortmann, L. and D. Rocheleau 1985. Womenand agroforestry: Four myths and threecase studies. Agroforestry Systems 2:253–272.Freudenberger, K.S. 1994. Tree and land tenure:Rapid appraisal tools. FAO Forests, Treesand People Programme and Network.Food and Agriculture Organization of theUnited Nations, Rome, Italy.Gladwin, C., J. Peterson, D. Phiri and R. Uttaro2002. Agroforestry adoption decisions,structural adjustment and gender in Africa.In: Barrett, C.B., F. Place and A.A. Aboud(eds) Natural Resources Management inAfrican Agriculture: Understanding andImproving Current Practices. CABI Publishing,Nairobi, Kenya.Hansen, J.D., M.K. Luckert, S. Minae and F.Place 2005. Tree planting under customarytenure systems in Malawi: An investigationinto the importance of marriageand inheritance patters. AgriculturalSystems 84(1): 99–118.Kerr, J.M. 2002. Watershed Development Projectsin India: An Evaluation. Research Report127. International Food Policy ResearchInstitute, Washington, DC, USA.Lastarria-Cornhiel, S. 1997. Impact of privatizationon gender and property rights inAfrica. World Development 25: 1317–1333.Otsuka, K. and F. Place 2001. Land Tenure andNatural Resource Management: A ComparativeStudy of Agrarian Communities inAsia and Africa. Johns Hopkins UniversityPress, Baltimore, USA and London, UK.Place, F. 1994. The Role of Land and Tree Tenureon the Adoption of Agroforestry Technologies:A summary and synthesis. LandTenure Center, University of Wisconsin,Madison, USA.Place, F., M. Adato, P. Hebinck and M. Omosa2003. The impact of agroforestry-basedsoil fertility replenishment practices onthe poor in Western Kenya. FCND DiscussionPaper 160. International Food PolicyResearch Institute, Washington, DC, USA.Quisumbing, A.R. 2003. Household Decisions,Gender, and Development: A Synthesis ofRecent Research. International Food PolicyResearch Institute, Washington, DC, USA.Quisumbing, A.R. and K. Otsuka 2001. Land,trees, and women: Evolution of LandTenure Institutions in Western Ghanaand Sumatra. Research Report 121. InternationalFood Policy Research Institute,Washington, DC, USA.Rocheleau, D. 1988. Gender, resource managementand the rural landscape: Implicationsfor agroforestry and farming systemsresearch. In: Poats, S.V., M. Schmink andA. Spring (eds) Gender Issues in FarmingSystems Research and Extension. WestviewPress, Boulder, USA.Rocheleau, D. and D. Edmunds 1997. Women,men and trees: Gender, power and propertyin forest and agrarian landscapes.World Development 25: 1351–1371.Rocheleau, D. and A. van den Hoek 1984. TheApplication of Ecosystems and LandscapeAnalysis in Agroforestry Diagnosisand Design: A case study from Kathamasub-location, Machakos district, Kenya.International Centre for Research on Agroforestry,Nairobi, Kenya.Schroeder, R. 1993. Shady practice: Gender andthe political ecology of resource stabilizationin Gambian garden/orchards. EconomicGeography 69: 349–365.Smith, L.C., U. Ramakrishnan, A. Ndaiye, L. Haddadand R. Martorelle 2003. The Importanceof Women’s Status for Child Nutrition inDeveloping Countries. Research Report.131. International Food Policy ResearchInstitute, Washington, DC, USA.Swallow, B.M., D.P. Garrity and M. van Noordwijk2001. The effects of scales, flows andfilters on property rights and collectiveaction in watershed management. WaterPolicy 3: 457–474.Vijfhuizen, C., C. Braga, L. Artur and N. Kanji2003. Gender, Markets and Livelihoodsin the Context of Globalisation. A study ofthe cashew sector in Mozambique. Mainreport: Phase 2: The South. www.iied.org/sarl/research/projects/t3proj01.html.

Keywords:Forestry, health, Africa, herbal medicine,nutrition, agricultureChapter 23The challenge of HIV/AIDS: Where doesagroforestry fit in?Marcela Villarreal and Christine Holding Anyonge, Food and Agriculture Organization of the United Nations;Brent Swallow and Freddie Kwesiga, World Agroforestry CentreAbstractIn its early stages, the global HIV/AIDS epidemic was predominantly an urban problem. It affectedmore men than women, and those with relatively higher incomes. The epidemic has moved rapidlyinto rural areas and now, the majority of people living with and dying from HIV/AIDS are the ruralpoor. Among them, women comprise a disproportionately high number. Although up to 80 percentof the people in the most affected countries depend on agriculture for their livelihoods, there havebeen limited responses from governmental and non-governmental actors in the agriculture and naturalresource sectors. This chapter discusses the impact of HIV/AIDS on rural livelihoods and the ways inwhich agroforestry could help mitigate those impacts. The chapter concludes that agroforestry interventionscan improve communities’ long-term resilience against HIV/AIDS and other external shocks inways that agricultural interventions alone cannot. Agroforestry technology can be better tuned to respondto the cash, labour, food and asset shortages faced by AIDS-affected communities. By providingoptions for producing nutritious food, managing labour, generating income and enhancing soil fertility,agroforestry technologies can help reduce hunger and promote food security. The authors recommendthat current and future agroforestry programmes and forest policies should be reviewed to assess theireffects on key determinants of HIV vulnerability. They also recommend some responses that can bemade by agroforestry research and development organizations.Introduction“Over the past few years, there has been a major revolutionin the world’s thinking about HIV. The epidemichas been understood, not just as a health issue that willalways remain, but as a major threat to developmentand to human security.” 11Peter Piot, Executive Director, Joint United Nations Programmeon HIV/AIDS (UNAIDS), Keynote Address to the UN Symposiumon Nutrition and HIV/AIDS 2 April 2001, Nairobi, Kenya.In its earlier stages, the HIV/AIDS epidemic was predominantlyan urban problem, affecting more men thanwomen, and those with relatively high incomes. Theepidemic is now moving rapidly into rural areas, hittingthose who are least equipped to deal with its consequences.Today, 95 percent of people living with – andan even higher proportion of those dying of – HIV/AIDSlive in developing countries. The overwhelming majorityare the rural poor and, among them, women comprisea disproportionately high number. The epidemic isresponsible for undoing decades of economic and social

182World Agroforestry into the Futuredevelopment and causing rural disintegration.For example, in sub-Saharan Africa,HIV/AIDS is depleting the region of its foodproducers and farmers and decimating theagricultural labour force for generations tocome (FAO 2004a).Although up to 80 percent of the people inthe most affected countries depend on agriculturefor their livelihoods, the greatest responseto the epidemic has come from thehealth sector. The agricultural sector cannotcontinue ‘business as usual’ in communitieswhere large numbers of adults havedied, leaving the elderly and children toproduce food. Agriculture and natural resourceresponses can play essential rolesin controlling the epidemic, so researcherswill have to revise the content and deliveryof services and the process of transferringagricultural knowledge.This chapter illustrates the specific impactsof the HIV/AIDS pandemic on agroforestryand proposes relevant strategies thatcould mitigate them. Since agroforestryis the science and practice of integratingtrees into farming systems and agriculturallandscapes, and there are many types andproducts of trees, there are many ways inwhich agroforestry can contribute.The impact of HIV/AIDS onrural livelihoodsThe impacts of HIV/AIDS are many andintertwined. While health and demographicimpacts have been studied most,the effects on agriculture and food securityhave become clearer over the last fewyears (FAO/UNAIDS 2003; Mushati et al.2003; Yamano and Jayne 2004). The impactscan be felt most dramatically in thereduction of the labour force, impoverishmentand the loss of knowledge that istransferred from one generation to another.All exacerbate food insecurity and poverty.Moreover, the consequences of HIV/AIDScontribute to making the rural poor morevulnerable to HIV/AIDS infection. Thisdevastating cycle must be broken, andagroforestry has a critical role to play.The impacts of HIV/AIDS that have a longlastingeffect on forestry, agroforestry andrural livelihoods stem largely from: a) reductionof the productive age groups andagricultural labour force; b) acute impoverishmentof households; and c) loss ofknowledge.Regarding loss of productive age groups,the Food and Agriculture Organizationof the United Nations (FAO) has estimatedthat some countries could lose up to26 percent of their agricultural labour by2020 (FAO 2004a). A lack of availablelabour was found to be associated with anincrease in forest fires in Malawi, wherecommunities resorted to clearing the landby burning the forest rather than selectivecutting/processing. Fires destroy largerareas of woodlands than selective cuttingand remove all the products and servicesthat woodlands provide to communities(Mike Jurvelius, FAO, personal communication2002). Other effects of labourshortage are shown as a sharp reduction inthe area of land cultivated and a shift fromcash to food crops (FAO 2002).When a member of a household becomesinfected with HIV/AIDS there is a need topay for medical expenses and/or a funeraland productive assets are often sold tomeet these expenses. The consequent lossin purchasing power has led to less moneybeing spent on food. For example, a study inEthiopia calculated the cost of treatment ofone AIDS patient was more than the entirefarm’s average annual income (Demeke1993).The loss of a generation to HIV/AIDS isinterfering with the transfer of agriculturalknowledge, practices and skills that arenormally passed from one generationto the next. This knowledge is critical toboth sustainable agricultural productionand cultural identity. The epidemic is alsoresponsible for a significant loss in institutionalknowledge, since staff of such agriculturalservice institutions as the extensionservices are also affected.The overall impact of the epidemic dependson the actual stage of evolution ofinfection in the population. As shown inFigure 1, the rate of infection of the HIVvirus follows distinct phases. In Phase I,which can take several decades, prevalencerates remain low and increase slowly.Phase II starts when somewhere around5 percent of the population becomesinfected and exponential growth rates ininfection occur. Phase III represents a levellingoff of infection rates, and finally, adecrease occurs in Phase IV. The preciseshape of the curve and the duration ofeach phase will vary from area to area anddepend on a range of factors, includingpolicy and action taken to prevent andmitigate the epidemic. Early and decisiveaction may keep rates low, without everreaching Phase II, as in the case of mostdeveloped countriesPrevalence ratePhase I II III IVTimeFigure 1. Phases in the HIV prevalence curve.Source: Villarreal (2003).

Chapter 23: The challenge of HIV/AIDS183HIV/AIDS is a slow epidemic with slowimpacts. Impacts only start to be obviousat the population level years after a sharpincrease in prevalence rates (Villarreal2003). This time lag varies in duration,but large numbers of deaths can be expected5–10 years after the onset of PhaseII, depending on the level of nutrition ofthe population and other factors. Death isone of the main impacts, but before deathoccurs, the person suffers weakness anddecreased ability to work for about twoyears (again, depending on nutritional,medical and other factors). Impoverishmentis another effect, starting more or lessat the time of decreased work ability. Impactsat the population level that lag someyears behind the death curve include theproblems faced by children who have lostfirst one then both parents. These ‘doubleorphans’ face tremendous problems in securingboth short- and long-term learningand livelihoods. Hypothetical ‘prevalence’and ‘death’ curves are shown in Figure 2.The time lag between the peaks of prevalenceand impact means there may be littleHIV prevalence/deathsFigure 2. Prevalence and death curves: Window of opportunity.Source: Villarreal (2003).Window ofopportunityHigh HIV rates/low impactobservable impact on a population evenwhen prevalence rates are very high. Inaddition, low prevalence rates may occurat the same time as high levels of impactduring Phase IV. Policies, programmes andother initiatives designed to prevent thespread of HIV/AIDS and mitigate its effectsneed to take account of the stage of theepidemic (Topouzis 2001).Very importantly for agroforestry initiatives,the lag between rising prevalence andimpact creates a ‘window of opportunity’for specific interventions. For example, asdiscussed below, some of the proposedlabour-saving initiatives are initially labour-intensive, and only later will producelabour-saving benefits. If they aresuccessful, the people could begin to reapthe advantages of such techniques whentheir need is greatest. Analysis of the progressionof AIDS at the household levelalso shows a time lag between the timeof infection of the first adult and the developmentof full-blown AIDS. Householdsgo through similar stages of prevalenceand impact and the death of differentPhase 1 Phase II Phase III Phase IVTimeDeathsPrevalencefamily members has different effects on thehousehold.Recent studies have shown that the preciseeffects of HIV/AIDS depend upon: a)the previous demographic structure of thehousehold; b) who and how many peoplein the household are chronically ill or die;c) the length of time that the household hashad to cope with the effects of the epidemic;and d) the resources the household hadat its disposal for dealing with increaseddemands (Mushati et al. 2003; Yamano andJayne 2004).Figure 3 depicts the different compositionof households affected by AIDS over time.There are 13 different types of householdsthat may exist in an area of high HIV/AIDSprevalence. In many communities, all13 types may exist at the same time. Eachtype of household has distinct resources,challenges and needs.Agroforestry possibilities forthe mitigation of HIV/AIDSimpactsIn this chapter we will be using a landscapeperspective on agroforestry: “Agroforestryrefers to a dynamic, ecologicallybased natural resources managementsystem that, through integration of treesin farms and in the landscape, diversifiesand sustains production for increasedsocial, economic and environmentalbenefits of land use at all levels,” (Leakey1996).Agroforestry can play an extremely importantrole in ensuring rural livelihoods survivean epidemic of HIV/AIDS because it:• enhances food security through improvingsoil fertility;• produces nutritious foodstuffs (fruits,berries, leaves) that can boost the

184World Agroforestry into the FutureMale-headed,Two adult residents,prime working age,own childrenOlder female-headedor two older adults,Non-resident childrenproviding support, possiblycaring for grandchildrenMale-headed,Male adultchronically illMale-headed,Female adultchronically illTwo unaffected adults,chronically ill childrenOther adultscaring formore childrenand ill adultsMale deathFemale-headedMale deathFemale-headedFemale deathMale-headedOther adultscaring formore childrenOrphan-headedSurvivingchildren mergedinto anotherhouseholdSurvivingchildren moveto urban areaTime and severity since beginning of epidemic in communityFigure 3. The evolution and different composition of households affected by AIDS.immune system and help protect againstopportunistic disease;• includes medicinal trees and other productsthat can help treat opportunisticinfections;• provides income generation opportunitiesthat are not labour-intensive;• offers a safety net of subsistence andincome (e.g. firewood for consumptionand for sale, animal fodder, potentiallyhigh-value tree products, building andthatching materials)• marks ownership of land; and• offers short-term and long-term labourmanagement possibilities.Improving soil fertilityOne of the greatest challenges for agriculturalhouseholds affected by HIV/AIDS isto maintain food production while copingwith reduced disposable income forpurchasing agricultural inputs. The WorldAgroforestry Centre and its partners in easternand southern Africa have developedseveral agroforestry methods for enhancingsoil fertility and maintaining soil quality. Inthe highly populated bimodal rainfall areasof western Kenya, the emphasis has beenon short-duration improved fallows andbiomass transfer. In the relatively sparselypopulated, unimodal rainfall areas of easternZambia, the emphasis has been on2–3-year fallows. Some farmers in easternZambia are using biomass transfer as aninput into the production of garlic, a plantknown for its anti-oxidant properties. In thedensely populated areas of Malawi, leguminoustrees are being intercropped withmaize. All these systems have met withapproval, with a total of some 250,000farmers now testing or adopting one/someof the practices by 2004. In areas with suitableproduction characteristics, improvedfallows allow farmers to produce maizeyields roughly similar to those obtainedusing recommended levels of inorganicnitrogen fertilizer, and two to five timeshigher than yields obtained under continuousmaize production without fertilizer.Economic returns to land and labour tendto be 20–100 percent greater using improvedfallows than with continuous maizeproduction without fertilizer, although thefigures do not reach those achieved whenmaize is grown with fertilizer (Rommelse2001; Franzel et al. 2002). Table 1 illustratesthe success of improved fallows ineastern Zambia.Studies on these systems generally showthat adoption levels are relatively similarfor male- and female-headed householdsand for households with different levels ofeducation. In addition to addressing issuesof soil fertility and quality, improved fallowsalso provide households with nearbysources of fodder and fuelwood, thus contributingto reduced overall labour requirementsfor the family (Ajayi et al. 2003).Agroforestry foodsPeople living with HIV are trapped in avicious cycle in which repeated episodesof illness weaken the body and acceleratethe onset of AIDS. HIV weakens the immunesystem and people become ill morefrequently. Repeated illness reduces appetiteand, at the same time, nutrients are lost

Chapter 23: The challenge of HIV/AIDS185Table 1.Labour requirements, maize production and returns to land and labour of Sesbania sesban improved fallows and continuouslycropped maize over a 5-year period, using an average farm budget.OptionWork daysha –1Maizet ha –1Returns to land:net present value US$ ha –1Returns to labour:net returns US$ work day –11996 prices 1998 prices 1996 prices 1998 pricesContinuous maize, no fertilizer 499 4.8 6 6 0.47 0.79Improved 2-year,sesbania fallow441 8.5 170 215 1.11 1.64Continuous maize with fertilizer 645 21.9 229 544 1.04 2.18Source: Franzel et al. (2002). Note that the economic analysis was conducted under two scenarios, 1996 and 1998 prices. Prices in 1996 were low,following a bumper harvest, while prices in 1998 were high, following a poor harvest.from the body through vomiting and diarrhoea.Some medicines also cause nutrientloss, while infections interfere with thebody’s ability to absorb and use the nutrientsin food. This has serious consequencesfor the poor, who are more likely to bemalnourished even before they become infected.Malnutrition may also be associatedwith an increased risk of HIV transmissionfrom mothers to children (FAO 2004b).Epidemiological evidence shows how vitamindeficiency, protein deficiency and lowimmunity make people much more susceptibleto the disease (Stillwagon 2002).Forest resources can help to provide thenutritional requirements of people who areHIV-positive. For example, leaves from thebaobab (Adansonia digitata) are a source ofcalcium, Vitamin A and Vitamin C (Boukariet al. 2001) and protein (Nordeide et al.1996). In Mali, the World Agroforestry Centrehas been working with several women’sgroups who are managing baobab plantsto produce leaves in a similar way to tea.Moringa oleifera is a multipurpose tree thatoriginated in the eastern Himalaya and hasbeen introduced to many tropical countries.In recent years, it has been promoted bynon-governmental organizations (NGOs)and faith-based organizations to meet thenutritional needs of communities affectedby HIV/AIDS. Moringa oleifera seeds arevery effective in clarifying and treating water;fresh moringa leaves contain very highlevels of micro- and macronutrients (protein,carotene, calcium, iron, Vitamins A, B andC); and moringa pods and dried leaf powderare used as nutritional supplements (McBurneyet al. 2004).Throughout sub-Saharan Africa, peopleharvest indigenous fruits to supplementtheir diets and incomes. In southern Africa,the Centre and its partners have initiated aresearch and development programme topromote the on-farm planting and managementof selected varieties of indigenous fruittrees. Greatest progress has been made withUapaca kirkiana. Scientists are selectingelite germplasm and looking at propagationand preservation, processing and marketingof fruit products. However, the domesticationand dissemination of indigenous fruitspecies is a challenging process. Mithoeferet al. (2004) estimated that the returns toplanting non-improved U. kirkiana wereonly 10–25 percent as high as the returns togathering the fruit from the wild because ofthe high costs associated with tree plantingand maintenance. Where wild fruit is stillavailable, farmers would find it profitable toplant U. kirkiana only if the domesticatedtrees yield fruit 2–4 years after planting, ifthe production per tree was increased bya factor of 8, and/or the price per kg of thedomesticated fruit was twice that of thewild fruit.Medicinal plantsPlant products have been used to curedisease since ancient times. In Africa, it isestimated that 80 percent of the populationuse natural products to treat various ailments.In the Shinyanga region of Tanzaniaalone, over 300 plants have been identifiedfor their medicinal values (Dery et al.1999). Although there is no hard evidenceto show that traditional medicines cantreat HIV and cure AIDS (FAO 2002), it isknown that certain tree products can beused to treat opportunistic infections associatedwith HIV/AIDS and/or to relievesome of their symptoms.In many villages, basic pharmaceuticaldrugs are not available, and householdsrely entirely on wild medicinal plants fortreatment of opportunistic infections associatedwith AIDS (Kolberg and HoldingAnyonge 2002). Results of two recentFAO country studies show that in someareas, such plants are becoming scarcer

186World Agroforestry into the Future(Sitoe 2004; Kayambazinthu et al. 2005).Encouraging village forest committees andextension officers to incorporate medicinalspecies into their management plans couldcontribute to sustainable management ofwild species. There is also potential for thedomestication of medicinal plants.New sources of medicines from plantscontinue to be discovered. For example,researchers at the University of Lausannehave found that the African tree Bobgunnuamadascarienis contains an anti-fungal substancethat combats Candida albicans, thebacteria responsible for fungal skin problems,and mycosis, a condition that commonlyaffects the eyes of AIDS patients. Itis also said to fight Aspergillis, a fungus thatcan cause fatal lung disease (SAF 2004).Income generationTrees provide many products (food, fuel,fibre, timber, poles and fodder) that householdscan use or sell. Livelihoods can alsobe supported by sale of woodland productssuch as honey and mushrooms. An NGOin Malawi has encouraged young womento make and sell charcoal briquettes insteadof engaging in commercial sex fortheir income (Ngwira et al. 2001).Safety net resourcesForest foods have traditionally complementedagriculture and often sustainpeople during severe food shortages(Shackleton et al. 2001). The ‘miombo’woodlands of southern Africa provide sucha traditional ‘safety net’ and they occurthroughout the area most affected by HIV/AIDS. Despite the mediocre fertility of thesoils, the vegetation provides a wide rangeof products including foods, medicinalplants, firewood and timber. Studies havefound that indigenous fruit can be a significantsource of food and cash income,especially for poorer households, womenBox 1. Rotational woodlots providefood and cashIn the tobacco-growing Tabora regionof Tanzania, farmers have traditionallyharvested wood from natural ‘miombo’woodlands to make poles for drying thetobacco leaves. To protect these woodlandsand create an alternative source ofincome, the Centre and its partners havedeveloped a rotational woodlot system,where farmers plant fast-growingacacia (primarily Acacia crassicarpa) in1-ha plots. They continue to grow maizein the same plots for the first two years,then wait until the fifth year to harvestthe wood. Compared to the customarymaize/fallow system, rotational woodlotsrequired about 2.5 times as muchlabour, mostly needed to harvest thewood in the fifth year. Despite the highlabour cost and longer payback period,the net current value of rotational woodlotswas over six times that of maizealone and the return to labour from rotationalwoodlots was more than twicethat of maize. Many farmers are nowadopting the system (Ramadhani et al.2002). While loss of family membersdue to HIV/AIDS may create long-termlabour shortages, it appears that extralabour can usually be hired for harvestingthe wood.and children. In the communal areas ofZimbabwe, for example, Cavendish (2000)found that the poorest 20 percent of householdsgenerated 7–9 percent of their totalhousehold income from selling collectedwild foods. At two sites in rural Zimbabwe,Mithhoefer and Waibel (2003) foundthat virtually all households consumedsome indigenous fruits (Uapaca kirkiana,Strychnos sp., Parinari curatellifolia) andthat 7–20 percent of households sold someUapaca kirkiana. When food is plentiful,it is mostly the children who eat the fruits.However, in times of food scarcity, Uapacakirkiana, Strychnos sp. and Parinari curatellifoliabecame the main food for over 70percent of households at one of the studysites. In total, the three indigenous fruitcontributed 5–7 percent of total householdincome. Returns to family labour investedin gathering, processing and selling indigenousfruit in both villages were found tobe higher than returns to crops, livestock,horticulture, exotic fruit trees and casuallabour.Marking ownershipIn many sub-Saharan African social systems,when a man dies, his relatives take overall productive assets (and sometimes otherproperty) from the widow. In some cultures,the widow, land, property and children are‘inherited’ by a brother of the deceased.With the spread of AIDS, and worries thatthey may be ‘inheriting’ infected people,families of deceased heads of householdsmay refuse to care for the widow andchildren, yet still claim the land that theirbrother had farmed. Widows are then leftwith no productive assets (Drimie 2002).Access to and ownership of land can havean important influence on the viability ofHIV/AIDS-affected households. Trees havelong been an indicator of tenure in Africa.There is some prospect that planting trees inabandoned fields can preserve the land forthe family and, at the same time, rehabilitatewasted soils and provide fuelwood, fodderand fruits. On the other hand, investing intrees could also encourage more powerfulfamily members to take over that portion ofland. The interpretation of customary practicewith regard to land and tree tenure willvary between adjacent communities sharingthe same cultural heritage. More evidence

Chapter 23: The challenge of HIV/AIDS187needs to be compiled on how aspects ofland and tree ownership can assist peoplesuffering from the effects of HIV/AIDS.Labour managementMost African agriculture depends on manuallabour and there are peaks in labourdemand, for example, for land preparation,planting and harvesting. When sickness,death and funerals occur during thesecritical periods, crop productivity will begreatly affected. Caring for the sick alsodemands time and energy and reducesavailability of labour, especially women’s,for agricultural tasks. In Ethiopia, a studyfound that AIDS-affected households spent50–66 percent less time on agriculture thanhouseholds that were not affected (Baryoh2000). In Tanzania, researchers found thatwomen spent 60 percent less time on agriculturalactivities when their husbandswere ill (Tibaijuka 1997).Different improved fallows agroforestry systemsadvocated by the Centre have differentimplications for total and seasonal labourdemand. Short-duration improved fallows,as developed in western Kenya, require lesstotal labour than the typical two-seasonpattern of maize production, but there isa greater seasonal labour demand duringthe land preparation phase of the long rainproduction season. If labour hiring is not aviable option, then shortage of labour mayimpede adoption (Rommelse 2001). The 2–3-year fallows developed in eastern Zambiatypically entail less labour per hectare andper unit output than the continuous maizesystems (no fertilizer) that they replace. Duringthe first year of tree establishment, thefallows do require extra labour, but thereis quite a wide variation between differentfallow systems. Steve Franzel, ICRAF (personalcommunication) used farm data fromeastern Zambia to calculate the extra labourtime required to establish 0.27 ha plots ofSesbania sesban and Tephrosia vogelii fallows.Since the average farm in this regionhas 1.08 ha of cultivated land, planting 0.27ha to improved fallow each year for 4 yearswould allow most farmers to convert theirfarms to the agroforestry system within 4years. The calculations showed that purestands of S. sesban and T. vogelii require anaverage of 36 and 22 extra labour days duringthe establishment year. If, however, thefallows are intercropped with maize duringthe first year, then only an extra 16 daysfor S. sesban and 3 days for T.vogelii arerequired. Different systems are therefore appropriatefor households at different stagesof AIDS impact. Households that have alreadysuffered significant labour losses maynot be well advised to plant pure standsof S. sesban, but they could still managethe extra labour required to intercrop theirmaize with T. vogelii.Reducing labour peaksA study in eastern Zambia showed that landpreparation, weeding and harvesting accountfor 70 percent of the labour demandassociated with the production of maize(29%)(18%)under improved fallows (Ajayi 2003: seeFigure 4). Land preparation and weeding arethe most demanding, since several essentialactivities have to be carried out over arelatively short time. Any interventions thatreduce labour requirements during thesephases will therefore be attractive.Further analysis showed that in agroforestryfields, farmers spent 27 percent of total labouron land preparation compared to 19percent in non-agroforestry fields (Table 2).However, in the non-agroforestry fields,farmers spent 34 percent of their labourtime on weeding activities compared to26 percent in agroforestry fields. Weedingis time-consuming and must be completedwithin a short time to prevent a poor harvest.Thus, by reducing the proportion oftime allocated to weeding from 34 to 26percent, improved fallows help labourconstrainedhouseholds to have a betterchance of a good yield. However, fallowsentail more labour for land preparation.The trade-off is in favour of fallows becausethe time ‘window’ for land preparation isless critical than that for weeding.Figure 4. Distribution of total labour inputs by type of field operation (% of absolute inputs).Source: Ajayi (2003).(2%) (2%)(3%)(8%)(3%)(23%)(12%)TransportationNursery operationsLand preparationRe-ridgingAgroforestry plantingMaize plantingFertilizer applicationWeedingHarvesting

188World Agroforestry into the FutureLong-term labour managementCommunities differ in their capacity torecover from external shocks. The timeof recovery after a drought, for example,depends on external factors, such as thetype of agro-ecological zone, and variablefactors, such as labour availability, knowledge,skills and food stocks, including storagefacilities and postharvest management.HIV/AIDS affects all the variable factorsand creates an increased dependency ratioin households, since it kills mostly productiveage adults.Traditional agroforestry systems vary indifferent agro-ecological zones and aresocially and culturally specific. Appropriateagroforestry strategies can help increasethe resilience of communities to externalshocks. Combinations of trees and cropscan be developed jointly with local communitiesin line with their short- and longtermneeds and according to the currentphase of HIV/AIDS. Techniques that havehigh labour requirements to get themstarted should be promoted during the lowimpact phase. Trees planted in the earlystages, when labour is still available, willprovide a source of food several years later,when labour supplies in the household andcommunity may have dwindled.So, although some agroforestry technologiesappear to be labour intensive, especially inthe early stages, they offer a number of benefitstowards the mitigation of HIV/AIDS.1. Increased yields might allow farmers toplant a smaller area of land with maize,reducing labour demand for land preparation,weeding and harvesting. The savingof labour and land would allow thefarmer to grow something else, such asvegetables and/or fruits.2. Most of the agroforestry technologies forimproving soil fertility also produce fuelwoodon farms thus saving the labourTable 2.Field operationWith agroforestry(improved fallows)(%)and energy normally spent gatheringwood (especially relevant for women).3. In some societies, planting trees enhancessecurity of land ownership.4. Improved fallows suppress weeds, therebycutting down the amount of labourneeded for weeding.5. Agroforestry technologies improve soilfertility and produce fodder, thereby reducingthe need for expensive inorganicfertilizer or livestock feed. However,some agroforestry technologies tradelow cost with higher demand for labour.6. Growing medicinal plants on farm preventsover-harvesting of wild varieties.Increasing the relevance ofagroforestry to HIV/AIDSmitigationOne of the defining attributes of agroforestryis its complexity. Across the Africancontinent, farmers have adopted agroforestrysystems that vary greatly in termsof types of goods and services produced,length of production period, market andenvironmental risks, ecological complexity,land, labour and managerial intensity, anddependence on input and output markets.While this complexity may make it difficultto recommend a standard set of agroforestryinterventions, it also means that thevariety of options are relevant to a wideWithout agroforestry(%)Land preparation 27.1 19.2Weeding 26.2 33.7Harvesting 16.9 18.5Source: Ajayi (2003).Time devoted to the three most time-consuming field activities during thecropping season.range of circumstances encountered byhouseholds and communities affected byHIV/AIDS. For example, households thathave absorbed non-family members maybe in a position to establish new agroforestrysystems on crop fields, while householdsthat have lost family members mayadapt trees into their home gardens.Another defining attribute of agroforestryis its tight connection with forestry. Farmersaccess and use trees throughout their landscapes,whether on their farms, on communallands, or on the margins of forests.Access to forests and communal lands affectsfarmers’ decisions on what trees theypreserve, what trees they plant, and howthey manage trees on their farms. Agroforestryinterventions therefore need to bebased on a good understanding of the treesand tree products that are already availableand the potential of agroforestry and otherforest management options for enhancingthe supply of consumable and marketableproducts (such as fodder and fruits) and forsoil fertility and conservation.Preserving and adapting knowledgeAgroforestry systems depend upon localagricultural and biodiversity knowledge tomaintain production. When a productivegeneration is lost, they can no longer passon their livelihood skills and agroforestry

Chapter 23: The challenge of HIV/AIDS189knowledge. The consequence is a youngpopulation who are ill equipped to managethe impacts of the epidemic and tomaintain successful production. Communityknowledge of the environment andlocal genetic diversity are fundamental fornurturing and preserving cultural identity.Indigenous knowledge and, often, technology-relatedknowledge are typicallygendered, with some aspects passed on bymen and some by women. Gaps in knowledgewill therefore occur when a parentdies. Effective initiatives in facilitating genderedlinks in indigenous and communityknowledge in single-parent householdsneed to be designed, implemented andmonitored, and agroforestry educationneeds to be targeted to the rural youth.It is also important to strengthen formalinstitutions, since human resources arebeing lost from ministries of forestry andagriculture, thereby hampering the developmentand implementation of agroforestrystrategies. In general, the loss of all typesof government cadres is creating seriousgovernance problems in the most-affectedcountries. However, instead of developingstrategies to replace lost human resources,which is already difficult for the worst hitcountries, it is necessary to rethink governmentfunctions and streamline them toadapt to the situation. Extension workers,for example, need to be trained to addressthe emerging clientele (widows, orphans,etc.) with specific information and knowledgeto match their needs. Vocationaltraining institutions may also be requiredto review their staffing, length and priorityfoci of courses, in light of the impact of thepandemic and changing human resourcerequirements. Specific staff policies needto be developed in the relevant ministries,including awareness building, behaviouralchange, communication, stigma and discrimination,voluntary counselling and testing,modification of working conditions ofemployees exposed to high-risk situations,improved access to medicine, etc.should be reviewed to assess their effects onkey determinants of HIV vulnerability. Theseinclude social inequalities, exclusion, creationof cash economies/disposable incomes,displacement and migrant labour. A reviewprocess would assist project programmersand policy makers to identify where andfor whom prevention and mitigation effortsshould be targeted and concentrated.ConclusionsAgroforestry interventions can play aunique role in the mitigation of the impactsof HIV/AIDS. They can improve communities’long-term resilience against this andother external shocks, in a way that agriculturalinterventions on their own cannot.Strengthening institutionsStrengthening local institutions is an essentialcomponent of the sustainabilityof any agroforestry intervention. Such anapproach also marries well with the currenttrend in extension towards supportingcollective action and empowering localcommunities to design and manage theirown development initiatives. Village forestrycan be the main cash generator in acommunity, and tree resources (customarywoodlands, village plantations and trees onfarms) have seen communities through periodsof severe hardship in the past.Forestry policyEffective forestry policy needs to take theeffects of HIV/AIDS into account. For example,policy makers need to be aware oflabour availability and the labour implicationsof interventions. Extension servicesneed to adapt to a new clientele, with veryspecific knowledge and service needs.Barany et al. (2005) recommend that currentand future forest policies and programmesBox 2. Strengthening local communitiesAgroforestry technology can be carefullytuned to respond to the AIDS-affected communities’lack of labour and cash, bothin the short term and in the long term. Byproviding labour management possibilities,agroforestry technologies can reducehunger and promote food security.The capacity to generate alternative lowinputincome-generating activities andA project in Katunga, Malawi, aimed to enhance the production of woody vegetationand strengthen the capacity of local communities to manage the resource by establishingeucalypts (Eucalyptus saligna) on hillsides surrounding the village. After 13 years, thetrees were handed over to the community’s natural resource management committee.The area is relatively fertile, with a wide range of crops and trees. Trees, not agriculture,now provide the major source of cash income and the villagers have built a new schoolclassroom. HIV/AIDS is a serious problem in the area and although native medicinalplants are not readily available, funds from the trees support 20 orphans under 5 yearsold, a basic pharmacy, transport to hospital, and contribute to funeral costs.Source: Kolberg and Holding Anyonge (2002).

190World Agroforestry into the Futureprovide essential nutrients means thatagroforestry interventions can help breakthe vicious cycle of impoverishment–malnutrition–AIDS. Medicinal plants andtrees frequently provide the only source ofsymptomatic relief available to the poor.The specific needs of a new clientele createdby the epidemic, with high dependencyratio households and unique compositions,must be taken into account when designingagroforestry interventions. Efforts shouldbe made to ensure that basic agriculturalskills are passed on to the younger generation,and that local knowledge, includingbiodiversity and gender-specific skills, arepreserved. If a strategy can be developedthat can respond effectively to these needs,a significant contribution will be madeto preventing and mitigating the consequencesof HIV/AIDS within agroforestrycommunities.AcknowledgementsWe wish to acknowledge many contributionsfrom colleagues at FAO and theWorld Agroforestry Centre who have supportedour individual and joint effortsto address the links between HIV/AIDS,forestry and agroforestry. At FAO, wethank Tage Michaelsen for his belief in,and his strong support of, our endeavours.The continued support of El Hadji Sene,Manuel Paveri and Hosny El Lakany, ForestryDepartment, FAO and other colleaguesin Rome, Harare and Accra should also beacknowledged. At the World AgroforestryCentre, we thank Paul Thangata, AggreyAgumya and Joyce Mitti for their contributions,and Olu Ajayi for the labour data.We appreciate the many helpful commentsprovided by Scott Drimie.ReferencesAjayi, O. 2003. Labour and financial analysisof agroforestry and conventional maizeproduction systems. Paper presented tothe Department of Agricultural Economics,Faculty of Agriculture, University ofZimbabwe, Harare, Zimbabwe.Ayaji, O.C., S. Franzel, E. Kunashula and F.Kwesiga 2003. Adoption of improvedfallow technology for soil fertility managementin Zambia: empirical studies andemerging issues. Agroforestry Systems59: 317–326.Barany, M., D. Kayambazinthu, A. Sitoe andC. Holding Anyonge 2005. HIV/AIDSand the miombo woodlands of Mozambiqueand Malawi: an exploratory study.Synthesis report. Forestry Working Paper,Food and Agriculture Organization of theUnited Nations, Rome, Italy.Baryoh, A. 2000. Socio-economic impact ofHIV/AIDS on women and children inEthiopia. United Nations DevelopmentProgramme: Addis Ababa, Ethiopia. Unpublished.Boukari, I., N.W. Shier, X.E. Fernandez, R.J.Frisch, B.A. Watkins, L. Pawloski and A.D.Fly 2001. Calcium analysis of selectedwestern African foods. Journal of FoodComposition and Analysis 14(1): 37–42.Cavendish, W. 2000. Empirical regularitiesin the poverty–environment relationshipof rural households: evidence fromZimbabwe. World Development 28(11):1979–2003.Demeke, M. 1993. The potential impact ofHIV/AIDS on the rural sector of Ethiopia.Unpublished manuscript, Addis AbabaUniversity, Addis Ababa, Ethiopia.Dery, B.B., R. Otsyina and C. Ng’atigwa 1999.Indigenous knowledge of medicinal treesand setting priorities for their domesticationin Shinyanga Region, Tanzania. WorldAgroforestry Centre, Nairobi, Kenya.Drimie, S. 2002. The Impact of HIV/AIDS onLand: Case studies from Kenya, Lesothoand South Africa. Synthesis report preparedfor the Food and Agriculture of theUnited Nations Southern African RegionalOffice. Human Sciences Research Council,Pretoria, South Africa.FAO 2002. Living well with HIV/AIDS. A manualon nutritional care and support forpeople living with HIV/AIDS. Food andAgricultural Organization of the UnitedNations, Rome, Italy.FAO 2004a. AIDS: a threat to rural Africa. Foodand Agriculture Organization of the UnitedNations. http://www.fao.org/FOCUS/E/aids/aids1-e.htmFAO 2004b. HIV/AIDS and food security. Foodand Agriculture Organization of theUnited Nations. http://www.fao.org/hivaids/impacts/nutrition_en.htmFAO/UNAIDS 2003. Addressing the impact ofHIV/AIDS on ministries of agriculture:Focus on eastern and southern Africa.Food and Agriculture Organization of theUnited Nations, Rome, Italy.Franzel, S., D. Phiri and F.R. Kwesiga 2002.Assessing the adoption potential of improvedtree fallows in eastern Zambia. In:Franzel, S. and S.J. Scherr (eds) Trees onthe Farm: Assessing the Adoption Potentialof Agroforestry Practices in Africa. CABInternational, Wallingford, UK.Kayambazinthu, D, M. Barany, R. Mumba andC. Holding Anyonge 2005. Miombowoodlands and HIV/AIDS interactions:Malawi country report. Forestry WorkingPaper, Food and Agriculture Organizationof the United Nations, Rome, Italy.Kolberg, S. and C. Holding Anyonge 2002. Joint agricultureand forestry sector mission report toMalawi. Food and Agriculture Organizationof the United Nations, Rome, Italy.Leakey, R.R.B. 1996. Definition of Agroforestryrevisited. Agroforestry Today 8(1): 5–7.

Chapter 23: The challenge of HIV/AIDS191McBurney, R.P.H., C. Griffin, A.A. Paul andD.C. Greenberg 2004. The nutritionalcomposition of African wild food plants:from compilation to utilization. Journal ofFood Composition and Analysis 17 (3–4):277–289.Mithoefer, D. and H. Waibel 2003. Income andlabour productivity of collection and useof indigenous tree products in Zimbabwe.Agroforestry Systems 59: 295–305.Mithoefer, D., J. Wesseler and H. Waibel 2004.Private investment in biodiversity conservation:a real option approach. Paper presentedat the 8th International Conferenceon Real Options: Theory Meets Practice,June 17–19, Montreal, Canada. http://www.realoptions.org/papers2004/WesselerMithoeferWaibel.pdfMushati, P., S. Gregson, M. Mlilo, C. Zvidzaiand C. Nyamukapa 2003. Adult mortalityand erosion of household viability inAIDS-afflicted towns, estates and villagesin eastern Zimbabwe. Paper presentedat the Scientific Meeting on EmpiricalEvidence for the Demographic and Socio-Economic Impact of AIDS. 26–28 March,Durban, South Africa.Ngwira, N., S. Bota and M. Loevinsohn 2001.HIV/AIDS, agriculture and food security inMalawi: Background to action. RENEWALWorking Paper 1. Regional Network onHIV/AIDS, Rural Livelihoods and FoodSecurity: Lilongwe, Malawi and TheHague, the Netherlands.Nordeide, M.B., A. Hatloy, M. Folling, E. Liedand A. Oshaug 1996. Nutrient compositionand nutritional importance of greenleaves and wild food resources in anagricultural district, Koutiala, in southernMali. International Journal of FoodScience and Nutrition 47(6): 455–468.Ramadhani, T., R. Otsyina and S. Franzel 2002.Improving household incomes and reducingdeforestation using rotational woodlotsin Tabora, Tanzania. Agriculture, Ecosystemsand Environment 89/3: 227–237.Rommelse, R. 2001. Economic assessment ofbiomass transfer and improved fallows trialsin western Kenya. World AgroforestryCentre Working Paper 2001–3. WorldAgroforestry Centre, Nairobi, Kenya.SAF 2004. Substances found in two trees mayoffer potential AIDS treatment. Society ofAmerican Foresters. http://www.safnet.org/archive/aids301.htm (also published inThe Forestry Source March 2001).Shackleton, C.M., S.E. Shackleton and B. Cousins2001. The role of land-based strategiesin rural livelihoods: the contribution ofarable production, animal husbandry andnatural resource harvesting. DevelopmentSouthern Africa 18: 581–604.Sitoe, A. 2004 Miombo woodlands and HIV/AIDS interactions: Mozambique countryreport. Forestry Working Paper. Food andAgriculture Organization of the UnitedNations, Rome, Italy.Stillwagon, E. 2002. HIV/AIDS in Africa: fertileterrain. Journal of Development Studies38(6): 1–22.Tibaijuka, A.K. 1997. AIDS and economicwelfare in peasant agriculture: Casestudies from Kagabiro village, Kageraregion, Tanzania. World Development25(6): 963–75.Topouzis, D. 2001. Strategy Paper on HIV/AIDS,eastern and southern Africa. InternationalFund for Agricultural Development, Easternand Southern Africa Division ProgrammeManagement Department, Rome, Italy.Villarreal, M. 2003. HIV/AIDS impacts and responses:Context and opportunity. Foodand Agriculture Organization of the UnitedNations, Rome, Italy.Yamano, T. and T.S. Jayne 2004. Measuring theimpacts of working-age adult mortalityon small-scale farm households in Kenya.World Development 32(1): 91–119.

Author ContactsFahmudin Agusisri@indo.netChip Fayc.fay@cgiar.orgParamu Mafongoyap.mafongoya@cgiar.orgAlain Albrechtalain.albrecht@mpl.ird.frS. Franzels.franzel@cgiar.orgMyles Mandermanderm@iwr.unp.ac.zaAlbert Dean Atkinsona.atkinson@cgiar.orgJanet Awimbojanetawimbo@yahoo.comAndre Bationoa.bationo@cgiar.orgJan Beniestj.beniest@cgiar.orgJean-Marc Boffaj.boffa@cgiar.orgAnja Boyeanja@swiftkisumu.comO.A. Chivingechivinge@admin.uz.ac.zwG.L. Denningglenn.denning@unmillenniumproject.orgTony Dolanbaraka@africaonline.co.keFaridaaifarida@yahoo.comKurniatun Hairiahsafods.unibraw@telkom.net ork.hairiah@cgiar.orgChristine Holding Anyongechristine.holdinganyonge@fao.orgBashir Jamaashir.jama@unmillenniumproject.orgSerigne T. Kandjis.kandji@cgiar.orgJohn R.S. Kaboggozakaboggoza@forest.mak.ac.ugFreddie Kwesigaf.kwesiga@cgiar.org orfkwesiga@fara-africa.orgRoger Leakeyroger.leakey@jcu.edu.auJ-P. Lillesø-Barnekowjpbl@cgiar.orgJens Mackensenjens.mackensen@unep.orgA.R. Mercado Jra.mercado@cgiar.orgRuth Meinzen-Dickr.meinzen-dick@cgiar.orgPatricia Negreros-Castillopnc@iastate.eduAmadou Niangamadou.niang@unmillenniumproject.orgDominique Nicolasdominique.nicolas@cirad.frJemimah Njukij.njuki@cabi.orgChin Ongc.ong@cgiar.orgHubert Omonthubert.omont@cirad.frCheryl Palmcpalm@iri.columbia.eduFrank Placef.place@cgiar.org

194World Agroforestry into the FutureYves-Coffi Prudencioyprudencio@worldbank.orgKate Schreckenbergk.schreckenberg@odi.org.ukThomas P. Tomicht.tomic@cgiar.orgSheila Raosherao@gmail.comTony Simonst.simons@cgiar.orgBernard Vanlauweb.vanlauwe@cgiar.orgPer G. Rudebjerp.rudebjer@cgiar.orgDiane Russelldiandeva@aol.comSyaka Sadiosyaka.sadio@fao.orgPornwilai Saipothongp.saipothong@cgiar.orgSara J. Scherrsscherr@ecoagriculturepartners.orgCharlie Shackletonc.shackleton@ru.ac.zaNguyen Van So (deceased)Ann Strouda.stroud@cgiar.orgCaroline Sullivancsu@ceh.ac.ukDidik Suprayogosafods.unibraw@telkom.netBrent Swallowb.swallow@cgiar.orgM.J. Swiftswift@mpl.ird.fr orswiftmj2003@yahoo.co.ukMeine van Noordwijkm.vannoordwijk@cgiar.orgBruno Verbistb.verbist@cgiar.orgLouis V. Verchotl.verchot@cgiar.orgMarcela Villarrealmarcela.villarreal@fao.orgMarkus Walshm.walsh@cgiar.orgRachel Wynbergrwynberg@science.uct.ac.zaSheona Shackletons.shackleton@ru.ac.zaZac Tchoundjeuz.tchoundjeu@cgiar.orgIssiaka Zoungranai.zougrana@afdb.orgKeith Shepherdk.shepherd@cgiar.orgAugust Temua.temu@cgiar.org

Acronyms and AbbreviationsACIARAFTPAHIAMANANAFEAPSOASARECAASBATSALAVUBACOSABDBPCCCABICARPECASCACBDCDMCGIARCIATCIFORCIRADCOP 8CSARDDFIDDSIRECAEMBRAPAFAOFALLOWFARAAustralian Centre for International AgriculturalDevelopmentagroforestry tree productsAfrican Highlands InitiativeAlliance of Indigenous Peoples of the ArchipelagoAfrican Network for Agroforestry EducationAgency for Personal Service Overseas (Ireland)Association for Strengthening Agricultural Research inEast and Central AfricaAlternatives to Slash and BurnAgroforestry Tree Seed Association of LantapanAfrican Virtual UniversityBaraka Agricultural College Old Students’ Association(Kenya)bulk densitybits per capitacarbonCentre for Agricultural and Biosciences InternationalCentral African Regional Program for the EnvironmentSistemas Agroforestales de Café en America CentralConvention on Biological DiversityClean Development MechanismConsultative Group on International AgriculturalResearchCentro Internacional de Agricultura TropicalCenter for International Forestry ResearchCentre de coopération internationale en rechercheagronomique pour le développement (France)Eighth Session of the Conference of the Parties (UNFCCC)Certificate in Sustainable Agriculture and RuralDevelopmentDepartment for International Development (UK)Department of Industrial and Scientific Research(New Zealand)East and Central AfricaEmpresa Brasileira de Pesquisa AgropecuariaFood and Agriculture Organization of the United NationsForest, Agriculture, Low-value Lands or WasteForum for Agricultural Research in AfricaFoFFPPGEFGFARGISGO–AFUGORTAGPGVUHIPCHIV/AIDSHULWAIACSSIBSRAMICOICCOICRAFICRISATICTIDRCIFPRIIFSIGUIIEDIIRRIISDILRIIMFINRMIPCCIPFEIPGRIIPRIRRIISDISNARIUCNfarmers of the futureForest Peoples Programme (Indonesia)Global Environment FacilityGlobal Forum on Agricultural Developmentgeographic information systemsGlobal Open Agriculture and Food UniversityFreedom from Hunger Council of Irelandglobal programme (GFAR)Global Virtual Universityheavily indebted poor countryHuman immunodeficiency virus/acquired immunodeficiencysyndromeHumid Lowlands of West and Central AfricaInterAcademy Council of Scientific StudiesInternational Board for Soil Research and ManagementInternational Coffee OrganizationInternational Cocoa OrganizationWorld Agroforestry CentreInternational Crops Research Institute for the Semi-AridTropicsinformation and communication technologyInternational Development Research Centre (Canada)International Food Policy Research InstituteInternational Foundation for Scienceincome generating unitsInternational Institute for Environment and DevelopmentInternational Institute of Rural ReconstructionInternational Institute for Sustainable DevelopmentInternational Livestock Research InstituteInternational Monetary Fundintegrated natural resource managementInter-Centre Panel on Climate ChangeInternational Partnership on Forestry EducationInternational Plant Genetics Research Instituteintellectual property rightsInternational Rice Research Instituteinstructional systems designInternational Support to National Agricultural ResearchWorld Conservation Union

196World Agroforestry into the FutureKKARIKEFRILANAFELGULUCIDLULUCFMDGMISEREORMOSCATMPRNNARONARSNEPADNGONRINRMNTFPNVSODIPPARPRPRGAPRSPRAFRAFTTREFRELMAR&DRIRDCRSPORUEpotassiumKenya Agricultural Research InstituteKenya Forestry Research InstituteLatin American Agroforestry Networklocal government unitsLand Use Change, Impacts and Dynamics (ILRI)land use, land-use change and forestryMillennium Development GoalsThe German Catholic Bishops’ Organisation forDevelopment CooperationMisamis Oriental State College of Agriculture andTechnologyMinjingu phosphate rocknitrogennational agricultural research organizationnational agricultural research systemsNew Partnership for Africa’s Developmentnon-governmental organizationNatural Resources Institute (UK)natural resource managementnon-timber forest productsnatural vegetative stripsOverseas Development Institute (UK)phosphorusphotosynthetically active radiationphosphate rockParticipatory Research and Gender Analysis (CGIAR)poverty reduction strategy papersrelative agronomic functionRegional Agroforestry Training Teamrelative ecological functionRegional Land Management Unitresearch and developmentRural Industries Research and DevelopmentCorporation (Australia)Round Table on Sustainable Palm Oilrainfall use efficiencyRUPESSADCSAFSAPSARDSARECSEANAFESEFSidaSSASSSASTCPSWCTOFTMTPNTrocaireTSBFTSPUNAIDSUNCBDUNCCDUNCTADUNDPUNEPUNESCOUNFCCCUNFFUSAIDWaNuLCASWBCSDWWFWRIRewarding Upland Poor for Environmental ServicesSouthern Africa Development CommunitySociety of American Forestersstructural adjustment programmessustainable agriculture and rural developmentSouth Asia Regional Energy CoalitionSoutheast Asian Network for Agroforestry EducationSahelian Eco-Farm (ICRISAT)Swedish Agency for International DevelopmentCooperationsub-Saharan AfricaSoil Science Society of AmericaSustainable Tree Crops Programmesoil and water conservationtrees outside forestsTrees and Markets (ICRAF)Thematic Program Network (UNCCD)Overseas Development Agency of the Catholic Churchin IrelandTropical Soils Biology and Fertility (CIAT)triple superphosphateUnited Nations Programme on HIV/AIDSUnited Nations Convention on Biological DiversityUnited Nations Convention on Combating DesertificationUnited Nations Conference on Trade and DevelopmentUnited Nations Development ProgrammeUnited Nations Environment ProgrammeUnited Nations Educational, Scientific and CulturalOrganizationUnited Nations Framework Convention on ClimateChangeUnited Nations Forum on ForestsUnited States Agency for International DevelopmentWater, Nutrient and Light Capture in Agroforestry SystemsWorld Business Council for Sustainable DevelopmentWorldwide Fund for NatureWorld Resources Institute

CreditsFront cover photo: Karen Robinson/Panos PicturesEditing: Michelle Grayson and Sue Parrott/Green Ink (www.greenink.co.uk)Design and Layout: Christel Blank/Green Ink (www.greenink.co.uk)Proofreading: Sue Hainsworth/Green Ink (www.greenink.co.uk)Technical Coordination: Warwick Easdown/ICRAF (www.worldagroforestry.org)Printing: Pragati Offset Pvt Ltd, India (www.pragati.com)

World Agroforestry into the FutureThis book has been compiled from contributions by world experts in agroforestry who met in Kenyain November 2003 to chart the future on the occasion of the 25 th anniversary of the World Agroforestry Centre.The World Agroforestry Centre is recognized as an international leader in agroforestry research, education and developmentsupport.It was established in 1978 as the International Council for Research in Agroforestry (ICRAF), to promote agroforestryresearch in developing countries, in response to a visionary study by Canada’s International Development Research Centre(IDRC). During the 1980s, the Centre operated as an information council focusing primarily on Africa. ICRAF joined theConsultative Group on International Agricultural Research (CGIAR) in 1991 to conduct global research on agroforestry,transforming itself into the International Centre for Research in Agroforestry. ICRAF is one of 15 international centres of theFuture Harvest Alliance supported by the CGIAR. Its first strategic plan as a CGIAR member addressed poverty, food securityand environmental degradation in smallholder farms of sub-humid and semi-arid Africa. The Centre then expanded to LatinAmerica, South Asia and Southeast Asia, while strengthening its activities in four regions of Africa. In 2002, it adopted thebrand name, the World Agroforestry Centre, but retained the International Centre for Research in Agroforestry (ICRAF) asits legal name.The mission of the CGIAR is to achieve sustainable food security and reduce poverty in developing countries throughscientific research and research-related activities in agriculture, livestock, agroforestry, fisheries, policy and natural resourcesmanagement (NRM).United Nations Avenue, GigiriPO Box 30677-00100Nairobi, KenyaTel: +254 20 722 4000 or via USA +1 650 833 6645Fax: +254 20 722 4001 or via USA +1 650 833 6646www.worldagroforestry.org

World Agroforestry into the Future