The proposed Parc Regional de Belomotse - Frontier-publications ...
The proposed Parc Regional de Belomotse - Frontier-publications ...
The proposed Parc Regional de Belomotse - Frontier-publications ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
<strong>Frontier</strong> Madagascar Environmental Research<br />
REPORT 6<br />
<strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong><br />
<strong>Belomotse</strong><br />
Biodiversity survey and conservation evaluation<br />
<strong>Frontier</strong>-Madagascar<br />
2003
<strong>Frontier</strong> Madagascar Environmental Research<br />
Report 6<br />
<strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong><br />
<strong>Belomotse</strong><br />
Biodiversity survey and conservation evaluation<br />
Emmett, D. A., Fanning, E. & Olsson, A. (eds)<br />
<strong>Frontier</strong>-Madagascar<br />
University of Toliara<br />
<strong>The</strong> Marine Sciences Institute<br />
Madagascar<br />
<strong>The</strong> Society for Environmental Exploration<br />
UK<br />
Toliara<br />
2003
Suggested Technical Paper citation:<br />
<strong>Frontier</strong>-Madagascar (2003) <strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>: biodiversity survey and<br />
conservation evaluation. <strong>Frontier</strong>-Madagascar Environmental Research Report 6. Society for<br />
Environmental Exploration, UK and the Institute of Marine Sciences, University of Toliara,<br />
Madagascar.<br />
Suggested Section citations:<br />
Emmett, D. A. & Rasoambolaniaina, Y. (2003) Flora: In <strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>:<br />
biodiversity survey and conservation evaluation. pp. 13 – 18. <strong>Frontier</strong> Madagascar Environmental<br />
Research Report 6. Society for Environmental Exploration, UK and the Institute of Marine Sciences,<br />
University of Toliara, Madagascar.<br />
Emmett, D.A., Olsson, A, Bloxam, C., Henson, D. & Thomas, H. (2003) Fauna: In <strong>The</strong> <strong>proposed</strong> <strong>Parc</strong><br />
<strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>: biodiversity survey and conservation evaluation. pp. 19 – 62. <strong>Frontier</strong><br />
Madagascar Environmental Research Report 6. Society for Environmental Exploration, UK and the<br />
Institute of Marine Sciences, University of Toliara, Madagascar.<br />
Olsson, A., Henson, D. & Emmett, D.A. (2003) Resource Use and Socio-Economics : In <strong>The</strong> <strong>proposed</strong><br />
<strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>: biodiversity survey and conservation evaluation. pp. 62– 64. <strong>Frontier</strong><br />
Madagascar Environmental Research Report 6. Society for Environmental Exploration, UK and the<br />
Institute of Marine Sciences, University of Toliara, Madagascar.<br />
Bloxam, C., Henson, D., Olsson, A. & Emmett, D.A. (2003) Tourism: In <strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong><br />
<strong>de</strong> <strong>Belomotse</strong>: biodiversity survey and conservation evaluation. pp. 64 – 71 <strong>Frontier</strong> Madagascar<br />
Environmental Research Report 6. Society for Environmental Exploration, UK and the Institute of<br />
Marine Sciences, University of Toliara, Madagascar.<br />
Emmett, D. A. & Olsson, A. (2003) Conclusions and Recommendations: In <strong>The</strong> <strong>proposed</strong> <strong>Parc</strong><br />
<strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>: biodiversity survey and conservation evaluation. pp. 71 – 76. <strong>Frontier</strong><br />
Madagascar Environmental Research Report 6. Society for Environmental Exploration, UK and the<br />
Institute of Marine Sciences, University of Toliara, Madagascar.<br />
This report series was created in 2005 and incorporated previous reports published by <strong>Frontier</strong>-<br />
Madagascar. <strong>The</strong> previous citation for this report was:<br />
<strong>Frontier</strong> Madagascar (2003) Emmett, D. A., Fanning, E., & Olsson, A. (eds) <strong>The</strong> <strong>proposed</strong> <strong>Parc</strong><br />
<strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>: biodiversity survey and conservation evaluation. <strong>Frontier</strong> Madagascar<br />
Environmental Research Report 6. ISSN 1479-120X Society for Environmental Exploration, UK and<br />
the University of Toliara & Institute of Marine Sciences, Madagascar.<br />
<strong>The</strong> <strong>Frontier</strong> -Madagascar Environmental Research Report Series is published by:<br />
<strong>The</strong> Society for Environmental Exploration<br />
50-52 Rivington Street,<br />
London, EC2A 3QP<br />
United Kingdom<br />
Tel: +44 (0)20 7613 3061<br />
Fax: +44 (0)20 7613 2992<br />
E-mail: research@frontier.ac.uk<br />
Web Page: www.frontier.ac.uk<br />
ISSN 1479-120X (Print)<br />
ISSN 1748-3719 (Online)<br />
ISSN 1748-5126 (CD-ROM)<br />
© <strong>Frontier</strong>-Madagascar 2003, 2005
<strong>The</strong> Society for Environmental Exploration (SEE)<br />
<strong>The</strong> society is a non-profit making company limited by guarantee and formed in 1989. <strong>The</strong> society’s<br />
objectives are to advance field research into environmental issues and implement practical projects<br />
contributing to the conservation of natural resources. Projects organised by <strong>The</strong> Society are joint<br />
initiatives <strong>de</strong>veloped in collaboration with national research agencies in co-operating countries.<br />
<strong>The</strong> University of Toliara / IHSM<br />
<strong>The</strong> University of Toliara was established in 1976 as a centre for learning and research in the biological<br />
sciences. <strong>The</strong> Institut Halieutique et <strong>de</strong>s Sciences Marines (IH.SM) is the marine <strong>de</strong>partment of the<br />
University of Toliara.<br />
<strong>The</strong> University is studying the flora and fauna of Madagascar and is conducting research into the<br />
maintenance and improvement of the environment and the sustainable use of the natural resources in<br />
the region.<br />
<strong>The</strong> WWF Madagascar Dry Forest Programme<br />
<strong>The</strong> WWF Madagascar Dry Forest Programme was initiated in 1998. It is a 50-year programme with<br />
the aim of conserving key habitats throughout southwest Madagascar. This eco-region is suffering<br />
more rapid habitat loss than any other region in Madagascar. <strong>The</strong> project aims to implement<br />
management plans for 15-25% of the region, incorporating all major habitat types and including areas<br />
of high biodiversity.<br />
<strong>The</strong> <strong>Frontier</strong>-Madagascar Forest Research Programme<br />
<strong>The</strong> Society for Environmental Research, the University of Toliara and the WWF Madagascar Dry<br />
Forest Programme have been conducting collaborative research into environmental conservation issues<br />
since July 2001 un<strong>de</strong>r the title of the <strong>Frontier</strong>-Madagascar Forest Research Programme. From July<br />
2001 until July 2002, the Programme has been working in the forests of the Sept Lacs region. This<br />
activity was aimed at i<strong>de</strong>ntifying core areas of biodiversity and establishing baseline biodiversity and<br />
resource-use information in those areas. In addition, tourism feasibility surveys were un<strong>de</strong>rtaken.<br />
FOR MORE INFORMATION<br />
<strong>Frontier</strong> -Madagascar<br />
BP41, Antsiranana, 201<br />
MADAGASCAR<br />
Tel/Fax: +261 (0) 20 82 23117<br />
E-mail: frontier@wanadoo.mg<br />
L’Insitute Halieautique et <strong>de</strong>s Sciences Marine<br />
(IHSM)<br />
Zone Portuaire, BP 141,<br />
Tulear 601<br />
MADAGASCAR<br />
Tel: +261 (0) 20 94 43552<br />
Fax: +261 (0) 20 94 43434<br />
E-mail: ihsm@wanadoo.mg<br />
Society for Environmental Exploration<br />
50-52 Rivington Street,<br />
London, EC2A 3QP. U.K.<br />
Tel: +44 (0) 20 7613 3061<br />
Fax: +44 (0) 20 7613 2992<br />
E-mail: research@frontier.ac.uk<br />
Internet: www.frontier.ac.uk
Contents<br />
List of Figures<br />
List of Tables<br />
List of Appendices<br />
List of Abbreviations<br />
Executive Summary<br />
Acknowledgements<br />
Page Number<br />
i<br />
i<br />
ii<br />
ii<br />
iii<br />
v<br />
1. INTRODUCTION 1<br />
1.1 Madagascar 1<br />
1.2 <strong>The</strong> history of conservation in Madagascar 2<br />
1.3 Threats to conservation 3<br />
1.4 Rationale 4<br />
2. DESCRIPTION OF THE SURVEY AREA 5<br />
2.1 <strong>The</strong> region 5<br />
2.2 Location of the <strong>proposed</strong> <strong>Regional</strong> Park 5<br />
2.3 Local communities 6<br />
2.4 Site history 6<br />
2.5 Topography, geology, and climate 7<br />
2.6 Vegetation 8<br />
2.7 Previous research 8<br />
3. THE STUDY SITES 9<br />
3.1 Data collection methods 9<br />
3.2 Zoological trapsites 10<br />
3.3 Trapsite procedure 12<br />
4. PLANTS 13<br />
4.1 Introduction 13<br />
4.2 Methods 13<br />
4.3 Results 14<br />
4.3.1 Gallery forest 15<br />
4.3.2 Spiny forest 16<br />
4.4 Discussion 16<br />
4.4.1 Dominant species and forest structure 16<br />
4.4.2 Plant-resource use 17<br />
4.4.3 Threats to habitats 18<br />
5. MAMMALS 19<br />
5.1 Introduction 19<br />
5.2 Methods 19<br />
5.2.1 Pitfall trapping 19<br />
5.2.2 Shernam and mesh trapping 20<br />
5.2.3 Observations, tracks and calls 20<br />
5.2.4 Mist-nets and harp traps 20<br />
5.2.5 Owl pellets 20<br />
5.2.6 Specimen handling 21<br />
5.3 Results 21
5.3.1 Trapping success 21<br />
5.3.2 Rattus rattus population composition 22<br />
5.3.3 Observations, tracks and calls 23<br />
5.3.4 Owl pellet analysis 23<br />
5.4 Discussion 23<br />
5.4.1 Ecological distribution and species richness 23<br />
5.4.2 Introduced species 25<br />
5.4.3 Seasonal variation 26<br />
5.4.4 Rare species and range extensions 27<br />
5.4.5 Threats to mammal conservation 28<br />
6. TRIAENOPS POPULATION STUDY 29<br />
6.1 Introduction 29<br />
6.2 Methods 29<br />
6.3 Results 30<br />
6.4 Discussion 31<br />
6.4.1 Temporal variation 31<br />
6.4.2 Seasonal variation 31<br />
6.4.3 Breeding status 32<br />
6.4.4 Population <strong>de</strong>nsity and conservation status 32<br />
7. DIURNAL LEMUR POPULATION STUDY 32<br />
7.1 Introduction 32<br />
7.2 Methods 33<br />
7.3 Results 34<br />
7.3.1 Line transects 34<br />
7.3.2 Direct observation 34<br />
7.4 Discussion 35<br />
7.4.1 Comparison of methods 35<br />
7.4.2 Lemur <strong>de</strong>nsity and distribution 35<br />
7.4.3 Conservation status 36<br />
8. BIRDS 36<br />
8.1 Introduction 36<br />
8.2 Methods 37<br />
8.3 Results 37<br />
8.4 Discussion 38<br />
8.4.1 Rare species and en<strong>de</strong>mism 38<br />
8.4.2 Ecological variation 38<br />
8.4.3 Threats to bird conservation 39<br />
9. REPTILES 40<br />
9.1 Introduction 40<br />
9.2 Methods 40<br />
9.3 Results 41<br />
9.4 Discussion 41<br />
9.4.1 Biogeography and species richness 41<br />
9.4.2 Seasonal variation 42<br />
9.4.3 Ecological distribution 42<br />
9.4.4 Rare species 44<br />
9.4.5 Range extensions and locality records 45<br />
9.4.6 Threats to reptile conservation 45<br />
10. AMPHIBIANS 46
10.1 Introduction 46<br />
10.2 Methods 47<br />
10.3 Results 47<br />
10.4 Discussion 48<br />
10.4.1 Biogeography and species richness 48<br />
10.4.2 Seasonal variation 48<br />
10.4.3 Ecological distribution 48<br />
10.4.4 Threats to amphibian conservation 48<br />
11. INVERTEBRATES 49<br />
11.1 Butterflies 49<br />
11.1.1 Introduction 49<br />
11.1.2 Methods 49<br />
11.1.3 Results 50<br />
11.1.4 Discussion 50<br />
11.1.4.1 Ecological distributions 50<br />
11.1.4.2 New species, rare species and range extensions 51<br />
11.2 Malaise trapping 53<br />
11.2.1 Introduction 53<br />
11.2.2 Methods 53<br />
11.2.3 Results 53<br />
11.3 Leaf litter extraction 54<br />
11.3.1 Introduction 54<br />
11.3.2 Methods 54<br />
11.3.3 Results 54<br />
11.3.4 Discussion 54<br />
11.4 Dragonflies 56<br />
11.4.1 Introduction 56<br />
11.4.2 Methods 56<br />
11.4.3 Results 56<br />
11.5 Molluscs and millipe<strong>de</strong>s 57<br />
11.5.1 Introduction 57<br />
11.5.2 Methods 57<br />
11.5.3 Results 58<br />
11.6 Spi<strong>de</strong>rs and scorpions 58<br />
11.6.1 Introduction 58<br />
11.6.2 Methods 58<br />
11.6.3 Results 58<br />
12. RESOURCE USE AND SOCIO ECONOMICS 59<br />
12.1 Introduction 59<br />
12.2 Methods 59<br />
12.2.1 Ground-truthing and GIS data layers 60<br />
12.2.2 Biological diversity and habitats maps 60<br />
12.2.3 Human settlements, land, and resource use 60<br />
12.3 Results 61<br />
12.3.1 Biological diversity and distribution 61<br />
12.3.2 Human populations and local management 61<br />
12.3.3 Resource use 62<br />
12.3.3.1 Gallery and spiny forests 62<br />
12.3.3.2 Wetlands 62<br />
12.3.4 Agriculture 63<br />
12.4 Discussion 64
13. TOURISM 64<br />
13.1 Introduction 64<br />
13.2 Methods 65<br />
13.3 Results and discussion 65<br />
13.3.1 Priority areas for tourism 65<br />
13.3.2 Transport 66<br />
13.3.3 Campsite location 67<br />
13.3.4 Campsite logistics 68<br />
13.3.5 Attractions and activities 68<br />
13.3.5.1 Gui<strong>de</strong>d walks 68<br />
13.3.5.2 Gui<strong>de</strong>s 69<br />
13.3.5.3 Boat trips 69<br />
13.3.5.4 Bird hi<strong>de</strong>s 69<br />
13.3.6 Amenities within the area 70<br />
13.3.7 Potential problems 70<br />
13.3.8 Monitoring 70<br />
14. CONCLUSIONS AND RECOMMENDATIONS 71<br />
REFERENCES 73
List of Figures<br />
Figure 1:<br />
Figure 2:<br />
Figure 3:<br />
Figure 4:<br />
Figure 5:<br />
Figure 6:<br />
Figure 7:<br />
Figure 8:<br />
Figure 9:<br />
Figure 10:<br />
Map of the Toliara region, showing the Onilahy River. (Reproduced from<br />
Lonely Planet, Madagascar 4 th Edition, May 2001.)<br />
Map of the <strong>proposed</strong> <strong>Regional</strong> Park (from WWF, Toliara)<br />
Map of the survey area from Antafoky Lake to the Sept Lacs region.<br />
Annual trapping of small mammals<br />
Annual composition of R. rattus composition<br />
Triaenops nocturnal activity pattern<br />
Diurnal lemur line-transect locations in the survey area<br />
Proportions of crops grown in the survey area<br />
Map of core area of <strong>Regional</strong> Park of interest to tourists<br />
Position of <strong>proposed</strong> tourist walks and birdhi<strong>de</strong>s at Antafoky Lake<br />
List of Tables<br />
Table 1:<br />
Table 2:<br />
Table 3:<br />
Table 4:<br />
Table 5:<br />
Table 6:<br />
Table 7:<br />
Table 8:<br />
Table 9:<br />
Table 10:<br />
Table 11:<br />
Summary of biodiversity and en<strong>de</strong>mism of taxa surveyed<br />
Trapsite <strong>de</strong>scription<br />
Number and en<strong>de</strong>mism of tree and scrub families, genera and species in<br />
different habitats<br />
Most cost common tree species in different habitats.<br />
Number and <strong>de</strong>nsity of diurnal lemurs within the survey area.<br />
Distribution of reptile species within the survey area<br />
Amphibian species list from the survey area<br />
Summary of butterfly species from the survey area<br />
Ecological distribution of butterflies within the survey area<br />
Ant species list from the survey area<br />
Proportion of crops grown and sold by local communities within the survey<br />
area<br />
i
List of Appendices<br />
Appendix 1:<br />
Appendix 2:<br />
Appendix 3:<br />
Appendix 4:<br />
Appendix 5:<br />
Appendix 6:<br />
Appendix 7:<br />
Appendix 8:<br />
Appendix 9:<br />
Appendix 10:<br />
Appendix 11:<br />
Appendix 12:<br />
Appendix 13:<br />
Trapsite habitat <strong>de</strong>scription<br />
List of plant species found in riparian forest<br />
List of plant species found in spiny forest<br />
Plant numbers and specific uses – Sept Lacs (west)<br />
Plant numbers and specific uses – Antafoky<br />
List of mammal species found in the <strong>proposed</strong> <strong>Regional</strong> Park<br />
List of bird species positively i<strong>de</strong>ntified in the <strong>proposed</strong> <strong>Regional</strong> Park<br />
List of bird species seen but not positively i<strong>de</strong>ntified in the <strong>proposed</strong><br />
<strong>Regional</strong> Park<br />
List of reptile species found in the <strong>proposed</strong> <strong>Regional</strong> Park<br />
Butterfly species collected in the <strong>proposed</strong> <strong>Regional</strong> Park, with associated<br />
habitat data.<br />
Description of <strong>proposed</strong> tourist walks<br />
Description of <strong>proposed</strong> bird hi<strong>de</strong>s<br />
List of collaborators for specimen i<strong>de</strong>ntification and <strong>de</strong>position<br />
List of Abbreviations<br />
ANGAP: Association Nationale pour le Gestion <strong>de</strong>s Aires Protégées<br />
GELOSE: Gestion Locale Securisee<br />
GIS: Geographical Information System<br />
GPS: Global Positioning System<br />
IH.SM: Institut Halieutique et <strong>de</strong>s Sciences Marines<br />
IUCN: <strong>The</strong> World Conservation Union<br />
MGF: <strong>Frontier</strong> Madagascar Forest research programme<br />
ONE: Office Nationale pour l’Environnement<br />
SAGE: Service d’Appui a la Gestion <strong>de</strong> L’Environment<br />
SEE: Society for Environmental Exploration<br />
WWF: World Wildlife Fund<br />
ii
EXECUTIVE SUMMARY<br />
This report <strong>de</strong>scribes the biodiversity survey of the recently <strong>de</strong>marcated <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong><br />
<strong>Belomotse</strong>, Toliara Province, Madagascar, conducted as part of the <strong>Frontier</strong>-Madagascar Forest<br />
Research Programme, between July 2001 and September 2002.<br />
<strong>The</strong> <strong>Frontier</strong>-Madagascar surveys involved a study of the extent and condition of the different<br />
vegetation types; vegetation species composition; and the collection of baseline biodiversity and<br />
ecological data for mammals, birds, reptiles, amphibians, butterflies and other selected<br />
invertebrates. Additional studies investigated resource use in the area, and the potential for the<br />
<strong>de</strong>velopment of ecotourism.<br />
<strong>The</strong> surveyed area was, until recently, ungazetted and did not have a formally <strong>de</strong>marcated<br />
boundary. It has now been <strong>de</strong>signated as a <strong>proposed</strong> <strong>Regional</strong> Park. Within this area is a core<br />
zone, managed through the locally enforced Gestion Locale Securisée (GELOSE). This core<br />
area, <strong>de</strong>scribed for the purposes of this report as the ‘Sept Lacs region’, covers over 3,000<br />
hectares between 40 and 250m asl, approximately 500 hectares of which are riparian forest and<br />
wetland habitats. <strong>The</strong> area surveyed by <strong>Frontier</strong>-Madagascar extends beyond the locally managed<br />
(Sept Lacs) core zone, and also inclu<strong>de</strong>s a lakesi<strong>de</strong> habitat, Antafoky Lake, and associated areas<br />
of riparian forest covering approximately 150 hectares. Thus, the report <strong>de</strong>tails surveys of the<br />
Sept Lacs region and the Antafoky Lake region, although both sites are contained within the<br />
<strong>proposed</strong> <strong>Regional</strong> Park.<br />
<strong>The</strong> surveyed area contains approximately 800 inhabitants, distributed through five villages.<br />
<strong>The</strong>se inhabitants belong to the Mahafaly, Tandroy and Masikoro tribes.<br />
<strong>The</strong> surveyed area contains many rare species. Four species of mammal, two species of bird and<br />
four species of reptile recor<strong>de</strong>d during this survey are classified as vulnerable (IUCN, 2002).<br />
Many of the species collected are too poorly studied to be <strong>de</strong>signated IUCN status at this time.<br />
Furthermore, several un<strong>de</strong>scribed species were discovered during this survey, including a Blind<br />
snake (Typhlops sp.), a pipistrelle bat (Pipistrellus sp.) a long-fingered bat (Miniopterus sp), and<br />
one butterfly species (Strabena sp.). It is expected that the results of the entomological<br />
collections will yield many new species.<br />
<strong>The</strong> species richness and en<strong>de</strong>mism of the surveyed taxa are summarised in Table 1.<br />
Table 1. Summary of biodiversity and en<strong>de</strong>mism of taxa surveyed.<br />
Taxon: Total no. No. of Percentage en<strong>de</strong>mism<br />
of species en<strong>de</strong>mics<br />
Trees and shrubs 142 16* -<br />
Mammals 26 21 81%<br />
Birds 79 33 42%<br />
Reptiles 55 50 91%<br />
Amphibians 6 5 83%<br />
Butterflies 61 - -<br />
* En<strong>de</strong>mic genera (species en<strong>de</strong>mism unknown)<br />
<strong>The</strong> riparian forest and wetlands within the Toliara region are threatened with<br />
<strong>de</strong>struction due to clearance or drainage for agricultural purposes. Charcoal production is<br />
leading to the loss of many trees of all sizes, and larger trees are logged for the<br />
production of pirogues (dugout canoes) or for local burial purposes. Grazing by domestic<br />
animals leads to further damage of the disturbed areas. Due to the dry climate,<br />
iii
egeneration is slow. Some species will become locally extinct unless the current trend in<br />
environmental <strong>de</strong>gradation is reversed.<br />
<strong>The</strong> major un<strong>de</strong>rlying cause of environmental <strong>de</strong>gradation is the low level of agricultural<br />
productivity, obliging local people to exploit natural resources. Supplementary factors are<br />
rapid population growth, low educational levels, harmful agricultural methods due to poor<br />
soils and techniques, and specific aspects of the Malagasy culture. Ultimately, in or<strong>de</strong>r to<br />
reduce exploitation of natural resources, living standards of local people must be raised.<br />
However, this will take a long time to achieve, and in the short term it is necessary to reduce<br />
the rate of environmental <strong>de</strong>gradation by promoting sustainable agriculture and natural<br />
resource use, raising awareness about environmental problems, and initiating appropriate<br />
forest management schemes.<br />
<strong>The</strong> <strong>proposed</strong> <strong>Regional</strong> Park contains some of the most important areas of riparian forest and<br />
wetland in southwest Madagascar. However, the remaining areas are becoming fragmented.<br />
Habitat <strong>de</strong>struction within the GELOSE-managed Sept Lacs core zone is occurring at a<br />
comparatively low rate, and may be sustainable in the short term. However, this area contains<br />
only a small number of wetland habitats, and many of the riparian forests are already <strong>de</strong>gra<strong>de</strong>d<br />
and fragmented.<br />
<strong>The</strong> riparian forest around Antafoky Lake is the largest unfragmented area of riparian forest in<br />
the surveyed area. However, this area is suffering from <strong>de</strong>forestation at an alarming rate and,<br />
unless checked, it will be mostly <strong>de</strong>stroyed within three or four years. Also, this lake and the<br />
associated forests contain the highest level of biological diversity found within the <strong>proposed</strong><br />
<strong>Regional</strong> Park. All vulnerable species found during the surveys were recor<strong>de</strong>d at this site, as were<br />
all but one of the un<strong>de</strong>scribed mammal, reptile and butterfly species. <strong>The</strong>refore, it is essential that<br />
immediate support be given to this site in or<strong>de</strong>r for it to remain a viable and integral part of the<br />
<strong>Regional</strong> Park.<br />
iv
ACKNOWLEDGEMENTS<br />
This summary report has been written based on research conducted during five <strong>Frontier</strong>-<br />
Madagascar phases, running consecutively from July 2001 to September 2002, fun<strong>de</strong>d by the<br />
Society for Environmental Exploration.<br />
<strong>The</strong> research team comprised <strong>Frontier</strong>-Madagascar field-staff, scientists from WWF-<br />
Madagascar, researchers from the University of Toliara, and <strong>Frontier</strong>-Madagascar volunteer<br />
research assistants. <strong>Frontier</strong>-Madagascar is a collaboration between the Society for<br />
Environmental Exploration, UK and the Institute of Marine Sciences, Madagascar.<br />
This report is the culmination of the advice, co-operation, hard work and expertise of many<br />
people. In particular, acknowledgments are due to the following:<br />
L’INSTITUT HALIEUTIQUE ET DES SCIENCES MARINE (IHSM)<br />
F-M Co-ordinators:<br />
Dr. Man Wai Rabenevanana<br />
Dr. Mara Edouard Remanevy<br />
SOCIETY FOR ENVIRONMENTAL EXPLORATION (SEE)<br />
Managing Director Eibleis Fanning<br />
Development Programme Manager Elizabeth Humphreys<br />
Programme Manager (Research) Nicola Beharrell<br />
Programme Manager (Operations) Matthew Willson<br />
WWF-MADAGASCAR<br />
<strong>Regional</strong> Director<br />
Research & Training Officer<br />
Biodiversity Officer/Herpetologist<br />
Sub-regional co-ordinator<br />
Mark Fenn<br />
Pierre Jules Rakotomalaza<br />
Achille Raselimanana<br />
Voninjatovo Andriamiarisoa<br />
FRONTIER-MADAGASCAR<br />
Project Co-ordinator: David Emmett<br />
Research Co-ordinator:<br />
Annette Olsson<br />
Assistant Research Co-ordinators: Craig McCoy, Catherine Bloxam, Esther Ball, and<br />
David Henson<br />
Logistics Manager:<br />
Hannah Thomas<br />
Field Assistants:<br />
Vomash, Ennanetse Alfred<br />
ASSOCIATION NATIONALE POUR LE GESTION DES AIRES PROTEGEES<br />
(ANGAP)<br />
<strong>Regional</strong> Director<br />
Jocelyn Rakotomalala<br />
CHICAGO FIELD MUSEUM<br />
Curator of Mammology<br />
Steven Goodman<br />
AMERICAN MUSEUM OF NATURAL HISTORY (AMNH)<br />
Curator of Herpetology<br />
Christopher Raxworthy<br />
CALIFORNIAN ACADEMY OF SCIENCES (CAS)<br />
Curator of Entomology Brian Fisher<br />
NATURAL HISTORY MUSEUM (NHM), LONDON<br />
v
Research Associate (Entomologist)<br />
UNIVERSITY OF LEEDS<br />
Postgraduate Entomologist<br />
David Lees<br />
Alison Cameron<br />
UNIVERSITY OF ANTANANARIVO<br />
DEA Mammology stu<strong>de</strong>nt<br />
Research Assistants during the five phases of fieldwork were:<br />
Juliana Vonimanitra Rahantavololona Rasoma<br />
July-Sept 2001 Oct-Dec 2001 Jan-March 2002<br />
Ross Adamson Rose Adams Geoff Allen<br />
Holly Baldwin Leo Barasi Kathryn Baskerville<br />
Marcus Barker Aaron Alexandra-Bloch Christopher Chapman<br />
Andrew Britton Noa Dychtwald Katie Foster<br />
Alastair Campbell Amanda Greenfield Steven Grzywacz<br />
Helen Dacre Patrick Griffiths Chris Hodgson<br />
Emily Fox Samantha Hanvey Rebecca Hughes<br />
Tony Kavanagh GeorgeLow Lara Hunt<br />
Sharon Ketley Ben Rayner Akiko Kumagae<br />
Jamil Khan Cat Simmons Charlotte Lawlor<br />
Emily Kilborn Lizzy Tennant Susan Loose<br />
Alwyn Mortimer James Thorburn Edward Morrison<br />
Tara O’Kelly<br />
Kate Wilkins<br />
Victoria Peaple<br />
Richard Wiltshire<br />
Anna Scaife<br />
April-June 2002 July-Sept 2002<br />
Philip Bowles<br />
Janette Buckley<br />
Thomas Carr<br />
Emma Butterworth<br />
Charlotte Corke<br />
Karen Dawson<br />
Katie Francis<br />
Colin Higgins<br />
Deborah Hendson<br />
Charlotte Jones<br />
Katherine Howard<br />
Curstain Keltie<br />
Melvin Jarmin<br />
Katy Littler<br />
Sarah Taylor<br />
Michael San<strong>de</strong>rson<br />
Helen Seyler<br />
vi
1. Introduction<br />
1.1 Madagascar<br />
Madagascar is the fourth largest island in the world, with a surface area of approximately<br />
587,000 sq km. <strong>The</strong> country stretches from 12º S, 50º E in the north to 26º S, 44º E in the<br />
south. A series of mountain ranges runs down the centre of the island from north to<br />
south, which creates a central highland plateau. To the east steep slopes fall towards the<br />
coast, whereas to the west the slopes are longer and gentler.<br />
Tra<strong>de</strong> winds approaching the island from the Indian Ocean to the east bring rain that falls<br />
over the escarpment. <strong>The</strong> west coast lies in a rain shadow and receives very little rain.<br />
Monsoon rains approach the island from the north and their strength <strong>de</strong>creases towards<br />
the south. <strong>The</strong> north-eastern part of the island is therefore the wettest, receiving up to<br />
4,000 mm rain per year, while the south-western part is the driest, with some areas<br />
receiving less than 300 mm per year. <strong>The</strong> highland plateau ranges between 750 and<br />
1,350 metres above sea level (m a.s.l.) and the Tsaratanana massif in the north rises to<br />
2,876 metres. Numerous rivers run from their sources in these highlands towards the<br />
coasts (Garbutt, 1999; Rogers, 2001).<br />
Due to the differences in altitu<strong>de</strong> and latitu<strong>de</strong> throughout the island, the climate varies<br />
both from area to area and throughout the year. In the central highlands the night<br />
temperature drops to near freezing point during the dry season, and in the north-east the<br />
day temperature often exceeds 40 º C.<br />
<strong>The</strong> variation in geography and climate creates a wi<strong>de</strong> variety of habitats. <strong>The</strong> humid<br />
areas to the east are dominated by lowland and sub-montane evergreen rainforests.<br />
Grasslands stretch over the highland plateau and <strong>de</strong>ciduous and dry forests dominate<br />
the western areas. Spiny forest and scrub thrive in the arid south. Along the coastal<br />
areas, marshlands, mangroves and coastal forests are common. <strong>The</strong>se different<br />
habitats are home to a great diversity of plants and animals.<br />
Madagascar is classified amongst the top eight mega-diversity countries on Earth. It<br />
ranks amongst the top four of the world’s most threatened biodiversity hotspots, and<br />
contains exceptionally high levels of en<strong>de</strong>mic species, many of which are classified as<br />
vulnerable or endangered (Myers et al, 2000).<br />
<strong>The</strong> flora of Madagascar shows very high levels of en<strong>de</strong>mism. It has been estimated<br />
that over 96% of all trees and shrubs are en<strong>de</strong>mic solely to Madagascar. This inclu<strong>de</strong>s<br />
approximately 161 en<strong>de</strong>mic genera (Schatz, 2001). Only one other country, Brazil,<br />
surpasses these figures. Over 95% of all plant species found in the dry forests of south<br />
and southwest Madagascar are regionally en<strong>de</strong>mic; that is, they are not only en<strong>de</strong>mic<br />
to Madagascar, they are also en<strong>de</strong>mic to the dry southern eco-region.<br />
<strong>The</strong> high levels of en<strong>de</strong>mism in Madagascar are not restricted to the flora. Over 45%<br />
of all bird species are en<strong>de</strong>mic, including 5 en<strong>de</strong>mic families and 35 en<strong>de</strong>mic genera<br />
(Morris & Hawkins, 1998). <strong>The</strong> figure for reptiles is even higher, with 91% species<br />
en<strong>de</strong>mism and 56% genus en<strong>de</strong>mism. At present, nearly 350 reptile species are<br />
known. At least 230 different amphibian species have been i<strong>de</strong>ntified, and over 99%<br />
of the species <strong>de</strong>scribed are en<strong>de</strong>mic to Madagascar and its offshore islands. This<br />
1
epresents one of the highest numbers of en<strong>de</strong>mic amphibians from any country in the<br />
world. However, it is with the primates that Madagascar shows the highest <strong>de</strong>gree of<br />
en<strong>de</strong>mism; 100% species en<strong>de</strong>mism, encompassing 14 genera and 5 families (Garbutt,<br />
1999). Madagascar contains over one third of the world’s primate families.<br />
1.2 <strong>The</strong> History of Conservation in Madagascar<br />
Madagascar has suffered extremely high levels of extinction since man arrived on the<br />
island. <strong>The</strong> date (and countries of origin) of man’s first arrival in Madagascar has long<br />
been the subject of studies. Archaeological evi<strong>de</strong>nce seems to indicate that man arrived<br />
approximately 2,000 years ago (Burney et al, 1997). Since this remarkably late date,<br />
man has precipitated a landsli<strong>de</strong> of extinctions. Again, archaeological evi<strong>de</strong>nce validates<br />
the ‘Blitzkrieg hypothesis’ put forward by Martin (1984), whereby as the human<br />
population on Madagascar began to increase, man began to over-hunt the animals and<br />
burn the land for cultivation, leading to a population crash amongst the en<strong>de</strong>mic<br />
megafauna. This in term led to the extinction of many species, including up to 15 lemur<br />
species, incorporating 8 genera. Other losses inclu<strong>de</strong>d two genera of elephant bird and at<br />
least two species of pygmy hippo. <strong>The</strong> extinction of these megafauna was remarkably<br />
rapid, and relatively recent; the pygmy hippos and elephant bird are thought to have gone<br />
extinct less than 1,000 years ago.<br />
Madagascar has had a long history of attempted conservation efforts. After French<br />
colonisation in 1895, prohibitive regulations were implemented (Kull, 1996). <strong>The</strong>se<br />
covered such aspects as banning the killing of lemurs, the initiation of replanting<br />
schemes using eucalyptus and pine and, in 1927, the initiation of ten nature reserves.<br />
However, France, through intensive logging activities combined with cash cropping of<br />
coffee, precipitated further, rapid <strong>de</strong>struction of the remaining rainforests. This forced<br />
many Malagasy to move into the highlands to find land for subsistence cultivation. In<br />
turn, the French government banned tavy (slash and burn agriculture). This led the<br />
Malagasy people to revolt, which in turn led to in<strong>de</strong>pen<strong>de</strong>nce.<br />
After in<strong>de</strong>pen<strong>de</strong>nce, the lack of funding to Madagascar proved disastrous for<br />
conservation efforts. In addition, in 1975 the presi<strong>de</strong>nt of Madagascar, Didier Ratsiraka,<br />
established a socialist regime and expelled all foreigners from the country; effectively<br />
halting all outsi<strong>de</strong> support and severely hin<strong>de</strong>ring progress (Richard and O’Connor,<br />
1997). However, in the face of mounting external pressure, in 1985 an International<br />
Conference on Conservation for Development was held in Antananarivo.<br />
This led to a dramatic increase in external support for Malagasy conservation<br />
<strong>de</strong>velopment by organisations such as WWF and World Bank. It also led to an influx of<br />
foreign scientists. In turn, a new Malagasy non-governmental organisation was set up;<br />
the National Association for the Management of Protected Areas (ANGAP), which led<br />
to the introduction of implemented conservation and <strong>de</strong>velopment strategies in specific<br />
protected areas, or National Parks. Today, ANGAP control over forty National Parks and<br />
protected areas.<br />
General elections in 2002 led to a new presi<strong>de</strong>nt, Marc Ravalomanana. It is to be hoped<br />
that this will lead to further improvements in conservation <strong>de</strong>velopment and support in<br />
Madagascar.<br />
2
1.3 Threats to Conservation<br />
<strong>The</strong> high diversity of habitats and species in Madagascar is un<strong>de</strong>r severe threat. <strong>The</strong><br />
human population in Madagascar is increasing by about 3% per year and has now<br />
reached approximately fifteen million (Bille, 2002; Durbin, 1994). With a growing<br />
population, the need for food and other resources also increases, putting further pressure<br />
on natural resources throughout the country.<br />
Over the last few <strong>de</strong>ca<strong>de</strong>s, the Malagasy culture has shifted towards a more materialistic<br />
society. This has led to a <strong>de</strong>mand for non-basic goods, leading to the need for a financial<br />
income. In turn, this has led to excess production of food-crops and cattle for export and<br />
sale.<br />
Natural vegetation is being cleared by the traditional slash and burn method, tavy, for<br />
maize cultivation. Forest and scrub areas in valleys and on hillsi<strong>de</strong>s are partly cut down,<br />
and then the area is burnt. <strong>The</strong> cleared land is used to grow maize. After two or three<br />
seasons the soil is usually too poor in nutrients for crops to grow, so the area is<br />
abandoned and new sites are found. This method is highly <strong>de</strong>structive and makes large<br />
areas of land barren for a long period after use. This is especially true in the dry forests<br />
of the south and west of Madagascar, where regeneration is exceptionally slow.<br />
Most of the crops are consumed locally and nationally, but there is now an increasing<br />
trend for maize to be exported to mainland Africa, Mauritius, Réunion and the<br />
Seychelles (Seddon et al., 2000). This exported maize is mainly used for animal food.<br />
Over 80% of Malagasy households are <strong>de</strong>pen<strong>de</strong>nt on wood or charcoal for cooking.<br />
Certain tree species are more useful for charcoal production than others and areas are<br />
either selectively logged or are cleared of all trees. <strong>The</strong> latter technique is becoming<br />
more common as the <strong>de</strong>mand for charcoal is rising. Often, only a small proportion of the<br />
charcoal from a production area is used locally. <strong>The</strong> majority of the charcoal is taken for<br />
sale to urban areas and regions where charcoal production is not possible. Charcoal is<br />
mainly produced in areas with a fairly good infrastructure, as it is difficult to transport<br />
the large heavy sacks without a vehicle.<br />
Although alternative materials such as bricks and fibreglass are used in some areas for<br />
construction of buildings and boats, wood is still the most commonly used material and<br />
logging of specific tree species is carried out to provi<strong>de</strong> planks and poles. Also, when<br />
bricks are ma<strong>de</strong>, the process uses a large amount of firewood.<br />
Zebu cattle and other livestock play an important role in Malagasy life and culture. Zebu<br />
are seen as a sign of wealth and status by the Malagasy people (M. Fenn, pers. comm.,<br />
2002). Large herds of zebu, goats and sheep are frequently found grazing in the forests<br />
near rural villages. Intensive grazing prevents plant regeneration. Saplings, shrubs and<br />
grasses are eaten, dramatically changing the structure of the forest plant communities.<br />
Also, during periods of very dry weather the branches from large trees may be cut down<br />
to provi<strong>de</strong> additional food for the livestock.<br />
<strong>The</strong>se pressures on natural resources have led to the <strong>de</strong>gradation of many habitats, and<br />
remaining areas are often fragmented. At present, more than 90% of the original natural<br />
vegetation has been lost in Madagascar (Myers et al., 2000).<br />
3
Many species of mammals, reptiles and birds are often hunted for food (Ferguson, 2002).<br />
Specific reptile species are collected for their skins, or for export and sale in the pet<br />
tra<strong>de</strong>, mostly to Europe and America. Excessive hunting or mass-collections of species<br />
can cause severe <strong>de</strong>clines in local populations, which can cause extinctions at the local<br />
and national level.<br />
Exotic species that have been <strong>de</strong>liberately or acci<strong>de</strong>ntally introduced by man to<br />
Madagascar can cause severe threats to populations of native species. Many introduced<br />
species often adapt well to new surroundings and can be very successful competitors.<br />
Furthermore, introduced species often carry diseases and parasites to which the native<br />
species have no resistance, causing serious <strong>de</strong>clines in affected populations.<br />
1.4 Rationale<br />
In 1996, WWF i<strong>de</strong>ntified 234 ecological regions (eco-regions) around the world on<br />
which to concentrate their present and future efforts. Five eco-regions were i<strong>de</strong>ntified in<br />
Madagascar. In 1997, two eco-regions were selected as priority regions for WWF-<br />
Madagascar. <strong>The</strong>se were the dry <strong>de</strong>ciduous forests of the west, and the spiny forests of<br />
the south and southwest. <strong>The</strong> spiny forest eco-region comprises over 6,612,850 hectares<br />
of land, less than 3,000,000 hectares of which is forested. Only 3.2% of the entire ecoregion<br />
is un<strong>de</strong>r protective management. Conversely, over 20% of the rainforests of<br />
Madagascar are un<strong>de</strong>r protective management (Lowry et al, 1997).<br />
Since 1998, the WWF-Madagascar Dry Forest Programme has been un<strong>de</strong>rtaking<br />
reconnaissance and research throughout the region, the objectives being to i<strong>de</strong>ntify sites<br />
of biological interest and to collect biological and socio-economic information. <strong>The</strong> aim<br />
has been to provi<strong>de</strong> long-term protective management for 15-25% of the eco-region. To<br />
this end, research has focussed on the i<strong>de</strong>ntification of biodiversity hotspots and unique<br />
or species-rich habitats. In or<strong>de</strong>r for long-term management to be successful, community<br />
participation and integration is vital. WWF has also focussed on i<strong>de</strong>ntifying the<br />
un<strong>de</strong>rlying social issues and aims to address these issues at the ‘grass-roots’ level, with<br />
involvement and support from national organisations.<br />
<strong>The</strong> primary aim of the <strong>Frontier</strong>-Madagascar Forest Research Programme is to promote<br />
conservation of the natural resources within selected study areas. <strong>The</strong> objectives of the<br />
research programme aim to provi<strong>de</strong> information on the biological value and<br />
importance of surveyed areas within the spiny forest eco-region in or<strong>de</strong>r to i<strong>de</strong>ntify<br />
sites of biological conservation importance. <strong>The</strong> project will also survey resource use<br />
and report on the status of, and threats to, the habitats in the surveyed areas.<br />
<strong>The</strong> research programme will also help to <strong>de</strong>velop GIS maps of the surveyed areas<br />
showing priority habitats (e.g. riparian forest), as well as areas of high biodiversity<br />
and sites of conservation importance. In addition, the programme will <strong>de</strong>velop a<br />
system of methods for monitoring aspects of biodiversity and resource-use, and assist<br />
in the training of Malagasy personnel for long-term monitoring programmes in the<br />
region.<br />
<strong>The</strong> biodiversity and resource use surveys conducted by the <strong>Frontier</strong>-Madagascar<br />
Forest Research Programme provi<strong>de</strong> information that can be combined with data<br />
4
collected by the WWF-Madagascar Dry Forest Programme and other conservation<br />
organisations. This can be used in the i<strong>de</strong>ntification of high priority sites of<br />
conservation interest, and provi<strong>de</strong> support for the implementation of sustainable forest<br />
management plans and activities. <strong>The</strong>se activities can then be monitored over time<br />
based on biodiversity and ground-truthed resource use data collected during the<br />
surveys, and can utilise personnel trained in biodiversity surveying and monitoring.<br />
2. Description of the survey area<br />
2.1 <strong>The</strong> Region<br />
Southwest Madagascar is the most arid region on the island, with some areas receiving<br />
an annual rainfall of less than 300 mm (Rogers, 2001). <strong>The</strong> area is dominated by spiny<br />
forest composed mainly of xerophytic plants. <strong>The</strong>se plants are well adapted to the dry<br />
climate. Typical adaptations inclu<strong>de</strong> succulent leaves, low fruit production and reduced<br />
branching and foliage (MacPhee, 1986). <strong>The</strong> area is referred to as the spiny forest ecoregion.<br />
This region stretches from the Mangoky River on the west coast to Fort Dauphin<br />
in the southeast (WWF, 2002; Seddon, 2000).<br />
<strong>The</strong> coastal town of Toliara has a population of approximately 140,000 inhabitants<br />
(Billé, 2002), and is the largest town in southwest Madagascar. <strong>The</strong> region around<br />
Toliara is one of the poorest in the country. <strong>The</strong> majority of the rural communities breed<br />
cattle and poultry and cultivate crops that are adapted to the dry environment, mainly<br />
manioc and maize.<br />
<strong>The</strong> spiny forest eco-region is the least studied region in Madagascar, but the few<br />
surveys that have been carried out suggest a very high diversity of animals (Nicoll &<br />
Langrand, 1989; Raxworthy, 1995). Deforestation and habitat fragmentation are<br />
especially prominent in the dry forests of the west and south, which are un<strong>de</strong>rgoing the<br />
most rapid <strong>de</strong>forestation of any region in Madagascar (Sussman & Rakotozafy, 1994;<br />
Dupuy & Moat, 1996; Smith, 1997). In fact, tropical dry forests are now consi<strong>de</strong>red to<br />
be the most endangered of all major tropical ecosystems (Janzen, 1988). This is due to<br />
extremely slow regeneration of the vegetation, caused by low rainfall, poor soils and the<br />
rapid onset of erosion following land clearance for cultivation or grazing.<br />
2.2 Location of the <strong>proposed</strong> <strong>Regional</strong> Park<br />
<strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong> is located along the Onilahy River, east of<br />
Toliara, see figure 1 for regional map. <strong>The</strong> area is dominated by spiny forest on the<br />
calcareous Mahafaly plateau. However, it has a core area containing lakes, marshes and<br />
areas of riparian forest. This is called the Sept Lacs region, see figure 2. <strong>The</strong> Sept Lacs<br />
region (S 23º 28' - S 23º 31', E 44º 04' – E 44º 10') covers over 3,000 ha between 40 and<br />
250m asl, and is located along the north bank of the Onilahy River between the villages<br />
of Antafoky and Ifanato. To the west of this core area is Antafoky Lake, with a large<br />
area of associated riparian forests (S 23º 28', E 044º 03'). <strong>The</strong> lake covers approximately<br />
150 ha, the gallery forest covers a further 150 ha.<br />
5
Both sites are contained within the <strong>proposed</strong> boundaries of the <strong>Regional</strong> Park. <strong>The</strong> areas<br />
surveyed by the <strong>Frontier</strong>-Madagascar Forest Research Programme inclu<strong>de</strong> the core Sept<br />
Lacs region, the Antafoky Lake area and the corridor of gallery forest along the Onilahy<br />
River.<br />
2.3 Local Communities<br />
<strong>The</strong>re are a number of villages within the park, representing several tribes. <strong>The</strong> small<br />
village of Antafoky, located on the western bor<strong>de</strong>r of the core area, contains villagers<br />
that are mainly from the Masikoro tribe. Mahaleotse, in the central part of the park,<br />
consists of three small settlements; Bevolavo, Man<strong>de</strong>rano and Befas. <strong>The</strong> village of<br />
Ifanato is located near the eastern bor<strong>de</strong>r of the core area. Mahaleotse and Ifanato mainly<br />
consist of people from the Mahafaly tribe.<br />
Figure 1. Map of the Toliara region, showing the Onilahy River. (Reproduced from Lonely Planet, Madagascar<br />
4 th Edition, May 2001.)<br />
<strong>The</strong> villagers cultivate the surrounding land, the most common crops being rice, maize,<br />
beans and manioc. In addition they herd livestock, and collect fish from the Onilahy<br />
River and associated ponds. <strong>The</strong> villagers also hunt wild ducks and gather natural<br />
products such as wood, reeds, medicinal plants and honey.<br />
2.4 Site history<br />
<strong>The</strong> site of the <strong>proposed</strong> <strong>Regional</strong> Park inclu<strong>de</strong>s the last remaining fragments of<br />
riparian forest along the northern bank of the Onilahy River (Raxworthy, 1995). In<br />
or<strong>de</strong>r to preserve these habitats the Department of Eaux et Forets implemented a<br />
system of local management, called Gestion Locale Securisee (GELOSE), in this area<br />
in March 2000.<br />
<strong>The</strong> principal aim of GELOSE was to provi<strong>de</strong> the local communities with the authority<br />
to manage and direct the sustainable use of the remaining forested areas. <strong>The</strong> regional<br />
division of the National Environment Office (ONE) provi<strong>de</strong>s technical support in this<br />
6
initiative. This regional <strong>de</strong>partment, called Service d’Appui a la Gestion <strong>de</strong><br />
l’Environment (SAGE), is working in conjunction with socio-economic advisors from<br />
the WWF-Madagascar Dry Forest Programme.<br />
Antafoky Lake and the associated riparian forests are not inclu<strong>de</strong>d within the GELOSE<br />
management plan, although they contain large areas of riparian forest and a unique<br />
wetland habitat.<br />
2.5 Topography, Geology and Climate<br />
<strong>The</strong> <strong>proposed</strong> <strong>Regional</strong> Park<br />
lies at the western edge of the<br />
Mahafaly Plateau, a limestone<br />
formation that extends south<br />
from Toliara to Cap Ste<br />
Marie. This area bor<strong>de</strong>rs the<br />
north bank of the Onilahy<br />
River. <strong>The</strong> lowest point has<br />
an altitu<strong>de</strong> of approximately<br />
30m, rising to over 250m on<br />
the plateau. <strong>The</strong> flanks of the<br />
plateau are fairly gentle,<br />
although there are many large<br />
cliffs along the edge of the<br />
river. Narrow, shallow valleys<br />
extend into the hills.<br />
<strong>The</strong> lake at Antafoky, the<br />
Sept Lacs river system and<br />
numerous other ponds in the<br />
area are fed by a number of<br />
un<strong>de</strong>rwater springs, with<br />
water flowing continuously<br />
throughout the year.<br />
Figure 2. Map of the <strong>proposed</strong> <strong>Regional</strong> Park, outlined in red.<br />
(Reproduced courtesy of WWF Toliara)<br />
<strong>The</strong> region has acquired its shape due to the dominant rock type in this region,<br />
carboniferous limestone. This has created large areas of comparatively high-standing<br />
hills and plateau, interspersed with channels and gullies cut into the limestone and the<br />
less erosion-resistant sandstone. A large limestone pavement can be found on the<br />
plateau. Exposed subterraneous caverns can also be found in the region, along with<br />
the remnants of stalagmites and stalactites, which are often found in un<strong>de</strong>rground cave<br />
systems.<br />
<strong>The</strong> local geology consists of well-bed<strong>de</strong>d white sandstones, sub-roun<strong>de</strong>d<br />
conglomerates and bed<strong>de</strong>d limestone. <strong>The</strong> sandstones contain a large proportion of<br />
7
marine shell fragments (mostly coral stems, bivalves, and gastropods). <strong>The</strong> limestone<br />
is less well bed<strong>de</strong>d than the sandstone unit and is generally thicker (up to 5m). <strong>The</strong><br />
limestone contains an abundance of fossils, mainly gastropods, bivalves and<br />
fragments of corals.<br />
<strong>The</strong> climate is semi-arid, with an annual average rainfall of around 400 mm. Over<br />
85% of this annual rainfall occurs during the rainy season, from November to March.<br />
Average maximum temperatures vary from 26ºC in the dry season to 36ºC in the wet<br />
season, occasionally reaching as high as 42ºC with a high humidity (Du Puy and<br />
Moat, 1996).<br />
2.6 Vegetation<br />
<strong>The</strong> <strong>proposed</strong> site of the <strong>Regional</strong> Park comprises a vegetation mosaic of unique<br />
habitats, including riparian and spiny forest. <strong>The</strong> riparian forest is limited to the<br />
valleys, whereas the spiny forest dominates the hillsi<strong>de</strong>s and plateau. <strong>The</strong> riparian<br />
forest, which occurs along the river system and around areas of wetland, has been the<br />
most heavily disturbed by the activities of man.<br />
In addition to the forest habitats, the <strong>proposed</strong> <strong>Regional</strong> Park contains large areas of<br />
agricultural land and scrub. Agriculture is concentrated on the flat land along the river<br />
edges and hilltops. Scrub areas are found where forest has been cleared around the<br />
edge of cultivated areas, and on the plateau in areas that have been subjected to<br />
burning.<br />
2.7 Previous research<br />
Southwest Madagascar has very high levels of floral and faunal diversity and<br />
en<strong>de</strong>mism. However, comparatively few biodiversity surveys have been un<strong>de</strong>rtaken in<br />
the region, and very little reliable or comprehensive data have been collected<br />
(Sussman & Rakotozafy, 1994).<br />
Previous research in southwest Madagascar has focussed on the Beza Mahafaly<br />
Special Reserve, the Kirindy Forest Reserve and the Mikea Forest region. <strong>The</strong>re are<br />
no published studies available for the <strong>proposed</strong> site of the <strong>Regional</strong> Park. <strong>The</strong> area has<br />
only been formally surveyed on one occasion, during a short visit to the area by an<br />
Earthwatch team (Raxworthy, 1995). Occasionally, stu<strong>de</strong>nts from the University of<br />
Toliara have visited the area, but data are often inaccessible and unreliable.<br />
8
3. <strong>The</strong> Study sites<br />
Work was carried out in the survey areas over a period of 15 consecutive months from<br />
July 2001 to September 2002. <strong>The</strong> surveys were divi<strong>de</strong>d into five periods of<br />
fieldwork, the dates of which are given below:<br />
Period 1 1 July to 6 September 2001<br />
Period 2 4 October to 15 December 2001<br />
Period 3 7 January to 20 March 2002<br />
Period 4 4 April to 1 July 2002<br />
Period 5 1 July to 6 September 2002<br />
<strong>The</strong> surveys focussed on the Antafoky Lake area for fieldwork periods 1, 2 and 3.<br />
During these three fieldwork periods, a significant proportion of the time was also<br />
spent in the Sept Lacs region. During fieldwork period 4 the surveys were based on<br />
the western si<strong>de</strong> of the Sept Lacs region, around Man<strong>de</strong>rano, see figure 3. Fieldwork<br />
period 5 focussed on the central and eastern areas of the Sept Lacs region.<br />
Figure 3. Map of the core region from Antafoky Lake to the Sept Lacs (produced based on GPS data collected by<br />
the <strong>Frontier</strong>-Madagascar Forest research Programme).<br />
3.1 Data Collection Methods<br />
<strong>The</strong>re are many methods for assessing zoological biodiversity. <strong>The</strong> methods used by<br />
the <strong>Frontier</strong>-Madagascar Forest Research Programme are a combination of suitable<br />
methodologies from comparative surveys in Madagascar, based on similar studies by<br />
<strong>Frontier</strong>-Tanzania using over twelve years of biodiversity experience in coastal forests<br />
and rainforest.<br />
<strong>The</strong> methodologies and associated data-recording sheets were i<strong>de</strong>ntified and compiled<br />
through the following process:<br />
9
A review of existing methodologies through a library search of references relating to past<br />
biodiversity, ecological and behavioural surveys in the country and region.<br />
<strong>The</strong> i<strong>de</strong>ntification of methodologies that could be used, consi<strong>de</strong>ring the environmental,<br />
logistical and personnel constraints of the project and region.<br />
<strong>The</strong> i<strong>de</strong>ntification, through field trial phases, of methodologies that would be suitable.<br />
<strong>The</strong> i<strong>de</strong>ntification of an efficient system for data collection, storage and use.<br />
<strong>The</strong> production of a series of accurate, simple data sheets and databases that enable<br />
efficient data collection and utilisation.<br />
<strong>The</strong> herpetological methodologies follow those of Raxworthy and Nussbaum (1994)<br />
and the small mammal surveys follow those of Goodman et al. (1999). <strong>The</strong> butterfly<br />
collection methodologies follow those of Kremen (1994) and the general invertebrate<br />
collection methods follow those of Fisher (1999). <strong>The</strong> botanical surveys follow those<br />
of Seddon et al (2000) and WWF methodologies (2002). <strong>The</strong> resource use surveys<br />
follow the methods of Kremen et al (1999).<br />
Behavioural and ecological studies were un<strong>de</strong>rtaken when the opportunity arose to<br />
survey a particular, usually threatened, species. <strong>The</strong>se surveys inclu<strong>de</strong>d a study of the<br />
<strong>de</strong>nsity and distribution of diurnal lemurs, following the methods used by Whitesi<strong>de</strong>s et<br />
al. (1988) and Rabinowitz (1997), and the estimation of populations of two species of<br />
bats, following the method used by Wilson et al. (1996).<br />
Trapping could not be fully relied upon to provi<strong>de</strong> a representative sample of species<br />
from an area as not all species were targeted with the standardised methods. So,<br />
opportunistic day and night searches were carried out to ensure the capture or<br />
observation of the majority of species present.<br />
See appendix 13 for a list of collaborators for specimen i<strong>de</strong>ntification and <strong>de</strong>position.<br />
See Methodology Training Manual (<strong>Frontier</strong> Madagascar, 2003) for <strong>de</strong>tailed<br />
<strong>de</strong>scriptions of the methods used.<br />
3.2 Zoological trap sites<br />
<strong>The</strong> biodiversity of the study area was assessed by applying a set of different methods<br />
for different target animal groups. <strong>The</strong>se inclu<strong>de</strong>d mammals, birds, reptiles, amphibians<br />
and invertebrates. <strong>The</strong> set-up of a combination of the chosen methods in a certain place<br />
for a set time-period was called a ‘trapsite’. <strong>The</strong> trapsites were typically established for<br />
eight-day periods and consisted of a combination of methods including pitfall trapping,<br />
Sherman and mesh trapping, bat collection using mist nets and harp traps, butterfly<br />
collection using sweep-nets and canopy traps, invertebrate collection, and casual<br />
collections and observations.<br />
10
<strong>The</strong> trapsites were positioned such that all habitats in the study area were represented.<br />
<strong>The</strong>y were set up in suitable localities for specimen collection and observation. Ten of<br />
the trap sites were un<strong>de</strong>rtaken during the dry season, one at the onset of the rains and six<br />
during the rainy season. This optimised opportunities for collecting and observing all of<br />
the chosen taxa.<br />
<strong>The</strong> surveys focussed on nine trap-sites around Antafoky Lake and a further eight<br />
trap-sites within the Sept Lacs region. See Appendix 1 and table 2 for trapsite location<br />
<strong>de</strong>scriptions.<br />
Table 2. Description of the zoological trap sites.<br />
Trap site Date and<br />
(phase No.) duration<br />
1 (013 1) 13-20 July<br />
8 days<br />
2 (013 2) 24-31 July<br />
8 days<br />
3 (013 3) 8-11 Aug.<br />
4 days<br />
4 (013 4) 17-24 Aug.<br />
8 days<br />
5 (014 1) 13-20 Oct.<br />
8 days<br />
6 (014 2) 23-30 Oct.<br />
8 days<br />
7 (014 3) 12-21 Nov.<br />
10 days<br />
8 (014 4) 25 Nov.-2<br />
Dec. 8 days<br />
9 (021 1) 21-28 Jan.<br />
8 days<br />
10 (021 2) 3-10 Feb.<br />
8 days<br />
Season Position Altitu<strong>de</strong> Habitat <strong>de</strong>scription * Trap effort<br />
Dry S 23º 28' 44.7"<br />
E 44º 03' 58.0"<br />
Dry S 23º 29' 16.0"<br />
E 44º 04' 39.1"<br />
Dry S 23º 28' 23.8"<br />
E 44º 04' 22.8"<br />
Dry S 23º 31' 41.9"<br />
E 44º 09' 20.0"<br />
Dry/wet S 23º 28' 44.7"<br />
E 44º 03' 58.0"<br />
Wet S 23º 29' 16.0"<br />
E 44º 04' 39.1"<br />
Wet S 23º 31' 29.2"<br />
E 44º 09' 33.2"<br />
Wet S 23º 31' 05.4"<br />
E 44º 08' 18.0"<br />
Wet S 23º 28' 44.7"<br />
E 44º 03' 58.0"<br />
Wet S 23º 29' 22.9"<br />
E 44º 04' 43.1"<br />
50 masl. Gallery forest, distance<br />
to lake 1-60 m.<br />
50-80 Gallery forest, distance<br />
masl. to lake 1-70 m.<br />
80-90 Thorny scrub, no water<br />
masl. association<br />
80-90 Gallery forest, distance<br />
masl. to river 5-10 m.<br />
50 masl. Gallery forest, distance<br />
to water 10-100 m.<br />
50 masl. Gallery forest, distance<br />
to lake 2-70 m.<br />
100-130 Transitional forest and<br />
masl. thorny scrub<br />
50 masl. Gallery forest, distance<br />
to marsh 5-80 m.<br />
50-80 Gallery and transitional<br />
masl. forest.<br />
50-250 Gallery forest,<br />
masl. transitional scrub and<br />
forest.<br />
120 bucket nights<br />
48 trap nights<br />
120 bucket nights<br />
48 trap nights<br />
60 bucket nights<br />
24 trap nights<br />
120 bucket nights<br />
48 trap nights<br />
264 bucket nights<br />
80 trap nights<br />
264 bucket nights<br />
96 trap nights<br />
330 bucket nights<br />
120 trap nights<br />
264 bucket nights<br />
96 trap nights<br />
264 bucket nights<br />
96 trap nights<br />
264 bucket nights<br />
96 trap nights<br />
11
11 (021 3) 1-10 Mar.<br />
10 days<br />
Wet S 23º 31' 29.2"<br />
E 44º 09' 33.2"<br />
120-150<br />
masl.<br />
Gallery and transitional<br />
forest<br />
330 bucket nights<br />
110 trap nights<br />
12 (022 1) 4-11 May<br />
8 days<br />
Dry S 23º 31' 26.9"<br />
E 44º 05' 34.3"<br />
70-80<br />
masl.<br />
Gallery and transitional<br />
forest.<br />
264 bucket nights<br />
48 trap nights<br />
13 (022 2) 11-18 May<br />
8 days<br />
Dry S 23º 30' 36.4"<br />
E 44º 04' 57.5"<br />
60-70<br />
masl.<br />
Gallery forest<br />
176 bucket nights<br />
800 trap nights<br />
14 (022 3) 22-29 May<br />
8 days<br />
Dry S 23º 28' 25.1"<br />
E 44º 04' 24.7"<br />
50-90<br />
masl.<br />
Gallery and transitional<br />
forest<br />
176 bucket nights<br />
800 trap nights<br />
15 (022 4) 3-10 June<br />
8 days<br />
Dry S 23º 29' 15.2"<br />
E 44º 04' 43.2"<br />
70 masl. Gallery forest 176 bucket nights<br />
800 trap nights<br />
16 (022 5)<br />
16-23 June Dry S 23º 31' 26.9" 70-90 Gallery and transitional 176 bucket nights<br />
8 days<br />
E 44º 05' 34.3" masl. forest<br />
800 trap nights<br />
17 (023 1)<br />
4-11 Aug. Dry S 23º 31' 15" 150-170 Gallery and transitional 264 bucket nights<br />
8 days<br />
E 44º 08' 35" masl. forest<br />
48 trap nights<br />
Total 136 days<br />
3632 bucket nights<br />
4158 trap nights<br />
* See appendix 1 for <strong>de</strong>tailed habitat <strong>de</strong>scriptions of the trapsites.<br />
3.3 Trapsite procedure<br />
When the trapsites were assembled, and during the days of checking the traps, care was<br />
taken not to disturb the area. This was to ensure that target species were not scared away<br />
from the area. <strong>The</strong> sites were only visited when checking the traps.<br />
<strong>The</strong> day of trapsite assembly was regar<strong>de</strong>d as ‘day 0’. <strong>The</strong> following days were regar<strong>de</strong>d<br />
as ‘days 1 - 8’. A ‘trap night’ or ‘bucket night’ refers to the period of 24 hours during<br />
which the traps and pitfall buckets were in use, from late afternoon after checking the<br />
traps to the following afternoon before checking the traps. For example, eleven pitfall<br />
buckets used for eight days would be <strong>de</strong>scribed as 88 bucket nights.<br />
<strong>The</strong> pitfall and box traps were checked and emptied in the morning between 06.00 and<br />
08.00 hours, and the butterfly canopy traps were baited. Between 16.00 and 18.00 hours<br />
in the afternoon the pitfall and butterfly traps were re-checked and the mammal traps<br />
were opened and baited. When a trapsite was located in open exposed areas, the pitfall<br />
traps were additionally checked at midday as the high temperatures could kill captured<br />
specimens.<br />
<strong>The</strong> target mammal species were mostly nocturnal, so the mammal box traps were open<br />
throughout the night and closed during the day. This also prevented trapped animals<br />
dying from heat exhaustion during the day.<br />
<strong>The</strong> Sherman and mesh traps were baited with fish, fruit (mostly banana) or shrimps to<br />
attract all possible mammal species. <strong>The</strong> butterfly canopy traps were baited with<br />
fermented banana. Pitfall traps were not baited.<br />
All captured individuals were taken to the base-camp for i<strong>de</strong>ntification. Individuals that<br />
were not preserved were returned immediately after i<strong>de</strong>ntification and released near the<br />
location of capture. Sherman traps were washed and returned to their previous position<br />
after the capture of an animal.<br />
12
4. Plants<br />
4.1 Introduction<br />
<strong>The</strong> Malagasy flora encompasses at least 107 families and 490 native genera of trees<br />
and large shrubs, of which 161 genera are en<strong>de</strong>mic to Madagascar and the Comoro<br />
Islands (Schatz, 2000). <strong>The</strong> en<strong>de</strong>mic tree genera are represented by 940 species. <strong>The</strong><br />
329 non-en<strong>de</strong>mic tree genera are represented by 3,280 species, of which 95% are<br />
en<strong>de</strong>mic. In total, over 96% of the estimated 4,220 species of Malagasy trees and<br />
large shrubs are en<strong>de</strong>mic (Schatz, 2001).<br />
Three families – Euphorbiaceae, Fabaceae and Rubiaceae – account for almost one<br />
third of the genera of Malagasy trees and large shrubs. Of the remaining families, 52<br />
are represented by only a single genus of trees in Madagascar, of which 14 are<br />
en<strong>de</strong>mic. <strong>The</strong>re are seven tree families en<strong>de</strong>mic to Madagascar – Asteropeiaceae,<br />
Didymelaceae, Kaliphoraceae, Melanophyllaceae, Physenaceae, Sarcolaenaceae and<br />
Sphaerosepalaceae. With over 700 species, the tree family Rubiaceae is the second<br />
largest plant family in Madagascar, after the family Orchidaceae.<br />
Southwest Madagascar has the highest percentage plant en<strong>de</strong>mism of all domains in<br />
Madagascar. However, in the arid regions of southern Madagascar, no <strong>de</strong>tailed study<br />
of floristic composition exists (Sussman & Rakotozafy, 1994). As a prerequisite to<br />
conservation and monitoring, baseline data on forest composition are vital.<br />
4.2 Methods<br />
Sixteen quadrats were laid out randomly along a series of 500 m-long transects that<br />
were placed through either gallery or spiny forest. <strong>The</strong> locations of these transects<br />
were selected because two distinct plant communities were recognised and the study<br />
was attempting to i<strong>de</strong>ntify indicative species from each habitat and measure the<br />
differences between them. Within the studied areas, the transects were positioned to<br />
pass through representative forest. Representative gallery and spiny forest habitats<br />
were i<strong>de</strong>ntified based on a number of factors. For example, it was assumed that<br />
comparatively undisturbed forests near water, with high canopy and Tamarindus<br />
indica as the dominant species in a complex plant community, could be <strong>de</strong>fined as<br />
gallery forest. <strong>The</strong>se forests also typically showed a strong collaboration with <strong>de</strong>ep,<br />
moist soils on fairly shallow gradients. By comparison, low canopy forests on the dry<br />
calcareous hills and plateau could be <strong>de</strong>fined as spiny forest. <strong>The</strong>se forests typically<br />
contained xerophytic, spiny plant communities on dry soils.<br />
Two quadrats were assigned to each transect. Quadrats were placed at random along<br />
each transect by generating random numbers and using them to correspond to the<br />
appropriate distance along the transect. <strong>The</strong> quadrat areas were calculated using<br />
minimum area curves. This entailed establishing a 1m x 1m quadrat and i<strong>de</strong>ntifying<br />
the number of different species that were present. <strong>The</strong> quadrat area was then doubled<br />
and the number of new species recor<strong>de</strong>d. This was continued to produce a species<br />
accumulation curve. When the curve levelled off, the suitable quadrat size was<br />
achieved. For gallery forest, this quadrat size was 16 m x 16 m. For spiny forest, the<br />
quadrat size was 32 m x 32 m.<br />
13
All trees, shrubs and stem herbs that were rooted within each quadrat were measured<br />
and the Diameter at Breast Height (DBH) was recor<strong>de</strong>d. Each plant was assigned to a<br />
DBH size class (< 2.5 cm, 2.5 - 5 cm, 5 - 10 cm, 10 - 25 cm, >25 cm) and plants<br />
within each size class were combined for analysis. <strong>The</strong> DBH was taken for each plant<br />
to help assess the current state of the habitat and to predict previous forest use. <strong>The</strong><br />
DBH measurements were also used to assess the effect of species-specific plant use<br />
on chosen trees and shrubs, compared to other tree species present within the same<br />
quadrat. Thus, the DBH data could be used to help show previous resource use within<br />
the area, and could indicate whether villagers selectively used certain plants or sizeclasses<br />
of plants.<br />
Plant i<strong>de</strong>ntifications were ma<strong>de</strong> in the field. Problematic individuals that could not be<br />
i<strong>de</strong>ntified had their vernacular name taken for future i<strong>de</strong>ntification. Fanaterany<br />
Manantiry, a local village el<strong>de</strong>r with extensive knowledge of the regional flora,<br />
provi<strong>de</strong>d the vernacular names and also gave information regarding local plant uses.<br />
Further information on plant use, both by man and by local fauna, was obtained from<br />
Schatz (2001).<br />
<strong>The</strong> plants within the quadrats were not marked or tagged. However, the co-ordinates<br />
of the individual quadrats were recor<strong>de</strong>d using GPS. <strong>The</strong> survey was inten<strong>de</strong>d as a<br />
species inventory, as plant data for the region are scarce. At present, regional<br />
emphasis is on the i<strong>de</strong>ntification of species and habitats. However, using the GPS<br />
points it will be possible to return to each quadrat in the future to monitor the state of<br />
the plant communities using the species-specific size data and <strong>de</strong>signated DBH sizeclasses.<br />
4.3 Results<br />
A total of 3,880 plants were sampled, from which 142 different tree and large shrub<br />
species were i<strong>de</strong>ntified. Family and genera names follow the classifications given by<br />
Schatz (2001). <strong>The</strong> vernacular names of all specimens were also recor<strong>de</strong>d (see<br />
appendices 2 and 3). Local uses of plants are also listed (appendices 4 and 5).<br />
4.3.1 Gallery forest<br />
Thirty families of plants were represented over eight 16 x 16 m quadrats, of which<br />
plants from twenty-five families were classified as trees or large shrubs (see table 3).<br />
<strong>The</strong> family Euphorbiaceae had the most species (7), followed by Fabaceae (6) and<br />
Rubiaceae (5). Two of the families present, Physenaceae and Sphaerosepalaceae, are<br />
en<strong>de</strong>mic to Madagascar.<br />
A total of forty-one tree and large shrub genera were i<strong>de</strong>ntified in the gallery forests,<br />
seven (17%) of which are en<strong>de</strong>mic. <strong>The</strong> total number of species that were collected<br />
was sixty-one, with forty-six being i<strong>de</strong>ntified and named to the species level. <strong>The</strong><br />
flora of Madagascar is too poorly known to accurately assess the level of species<br />
en<strong>de</strong>mism at this time.<br />
14
Table 3. <strong>The</strong> number and en<strong>de</strong>mism of tree and large shrub families, genera and species per habitat.<br />
Gallery<br />
forest<br />
Spiny<br />
forest<br />
Family Genus Species Uni<strong>de</strong>ntified<br />
Number En<strong>de</strong>mic Number En<strong>de</strong>mic Number En<strong>de</strong>mic Number<br />
25 2 41 7 46 - 15<br />
31 2 60 12 86 - 32<br />
Throughout all surveyed areas of gallery forest, the dominant large (DBH > 25 cm)<br />
tree species was Tamarindus indica (see table 4). This species formed a closed canopy<br />
15-25 m in height in undisturbed areas. <strong>The</strong> tree species that was dominant through<br />
the lower size ranges was Celtis bifida. Other common gallery forest species in the<br />
surveyed areas inclu<strong>de</strong>d Diospyros sakalavarum and Olax lanceolata.<br />
Table 4. Number of individuals of the two most common tree species in each size class and habitat<br />
Riparian Forest quadrats<br />
Spiny Forest quadrats<br />
Species 25cm 25cm<br />
ANACARDIACEAE<br />
Operculicarya <strong>de</strong>caryi 16 6 11 26<br />
BURSERACEAE<br />
Commiphora arafy 2 5 10 8 3<br />
Commiphora sp1 1 2 27 21 14 16 2<br />
Commiphora sp2 43 26 21 3<br />
CELTIDACEAE<br />
Celtis bifida 99 47 15 9 3<br />
EBENACEAE<br />
Diospyros latispatula 100 44 17<br />
Diospyros sakalavarum 68 23 12 2 13 5<br />
EUPHORBIACEAE<br />
Croton sp.3 178 2<br />
Securinega seyrigii 1 45 32 14 5<br />
FABACEAE<br />
Alantsilo<strong>de</strong>ndron sp. 49 53 37 14<br />
Tamarindus indica 5 2 20<br />
MALVACEAE<br />
Grewia grevei 1 140 19 3<br />
OLACACEAE<br />
Olax lanceolata 37 26 30 6 1 1 1<br />
4.3.2 Spiny forest<br />
Thirty-nine families of plants were represented over eight 32 x 32 m quadrats, of<br />
which plants from thirty-one families were classified as trees or large shrubs. <strong>The</strong><br />
15
families Euphorbiaceae and Fabaceae had the most species (14 each), followed by<br />
Malvaceae (7) and Burseraceae (6). Two of the families present, Physenaceae and<br />
Sphaerosepalaceae, are en<strong>de</strong>mic to Madagascar.<br />
A total of sixty tree and large shrub genera were i<strong>de</strong>ntified in the spiny forests, twelve<br />
(20%) of which are en<strong>de</strong>mic (see table 3). <strong>The</strong> total number of species that were<br />
collected was 118, with 86 being i<strong>de</strong>ntified and named to the species level. As with<br />
the gallery forest species, en<strong>de</strong>mism at the species level is not inclu<strong>de</strong>d.<br />
Throughout all surveyed areas of spiny forest, there were few large (DBH > 25 cm)<br />
tree species. <strong>The</strong> upper size ranges were dominated by Antsilo<strong>de</strong>ndron sp.,<br />
Commiphora spp. and Operculicarya <strong>de</strong>caryi. <strong>The</strong>se species formed an open, low<br />
canopy 3-10 m in height in undisturbed areas. <strong>The</strong> canopy height <strong>de</strong>pen<strong>de</strong>d on the<br />
gradient and substrate of the area. <strong>The</strong> tree species that were dominant through the<br />
lower size ranges were Croton sp. 3, Grewia grevia and Diospyros latispatula.<br />
4.4 Discussion<br />
<strong>The</strong> most significant achievement of the botanical survey was the production of a<br />
preliminary species list for the two main natural habitats within the <strong>proposed</strong> regional<br />
park. Few inventory studies have been un<strong>de</strong>rtaken in the region, so this species list<br />
can be used for comparative studies and monitoring when further studies have been<br />
completed in other areas. At present, the analyses of survey data are awaiting the<br />
completion of comparative studies, in or<strong>de</strong>r to allow comparisons to be ma<strong>de</strong> and<br />
specific regional habitats to be i<strong>de</strong>ntified. <strong>The</strong>refore, this discussion will focus on<br />
forest structure, dominant species, resource use and threats to the spiny and gallery<br />
forests.<br />
4.4.1 Dominant species and forest structure<br />
<strong>The</strong> two most dominant tree and shrub species within each size class and habitat are<br />
shown in Table 4. <strong>The</strong> data show clearly that there are very few dominant habitatgeneralist<br />
species. By that, it is meant that the dominant species within the spiny<br />
forests are almost entirely absent in the gallery forests, and vice versa. This seems to<br />
indicate that, within the region, habitats that may be only metres apart are clearly very<br />
different in terms of plant dominance and forest structure. It also indicates the habitatspecific<br />
diversity within the spiny and gallery forests of southwest Madagascar.<br />
<strong>The</strong> results in Table 4 also show that in gallery forest, a small number of species are<br />
dominant, but they are dominant through more than one size class. For example,<br />
Celtis bifida was one of the two most dominant plant species within every size class.<br />
This leads to the ground layer, shrub layer and canopy layer of gallery forest being<br />
dominated by a small number of species.<br />
In contrast, the spiny forests show a much greater diversity of dominant species<br />
between the different size classes. None of the plant species that are dominant in the<br />
lower size classes are even found in the upper size classes. This reflects the diversity<br />
of species that are found in the ground layer, shrub layer and canopy layer, and shows<br />
the level of niche-specificity that spiny forest plant species exhibit.<br />
16
<strong>The</strong>se results seem to show that the spiny forests contain a high diversity of plant<br />
species within a complex community. <strong>The</strong> gallery forests seem to show a lesser<br />
<strong>de</strong>gree of plant diversity, with a more simple plant community dominated by<br />
relatively fewer species.<br />
4.4.2 Plant-resource use<br />
Appendices 4 and 5 show some of the uses of specific tree and shrub species, both by<br />
man and by specific fauna. As can be seen, trees from gallery forest that are typically<br />
used for construction purposes by man were found in low numbers and only in the<br />
lower size classes. For example, Bivinia jalberti and Colvillea racemosa, both of<br />
which are used for the construction of houses and pirogues, were found in low<br />
numbers and only as saplings. Conversely, many of the trees from the spiny forest that<br />
could be used for construction purposes were found in fairly high numbers, often as<br />
large trees. It therefore appears that specific tree species of a particular size are being<br />
targetted and removed from areas of gallery forest for the purposes of construction,<br />
leading to the loss of large trees from this habitat. However, the effect of speciesspecific<br />
tree removal from areas of spiny forest appears to be comparatively less. This<br />
may be due to the relative difficulties of access to areas of spiny forest, or it may<br />
reflect a preference for the larger trees found within gallery forest. It may also be that<br />
the surveyed areas of spiny forest were not representative.<br />
Many of the plants listed in appendices 4 and 5 are used by man for purposes other<br />
than construction. For most of these plant species, such purposes either require large<br />
trees, such as for the collection of edible fruits, or, the plant is used in a non<strong>de</strong>structive<br />
way, such as the collection of leaves for medicinal purposes. <strong>The</strong> results<br />
appear to show comparatively large numbers of tree and shrub species of varying size<br />
classes that are utilised for resources other than wood. Without comparative data, it is<br />
not possible to state whether these species have been adversely affected by local use.<br />
However, there appear to be significant numbers of these plants, in all size ranges,<br />
both in gallery forest and spiny forest. For example, Tamarindus indica produces<br />
fruits that are collected and eaten or sold by local villagers. <strong>The</strong>se trees are dominant<br />
in the upper size ranges in all surveyed areas. It therefore appears that, based on<br />
current data, the non-<strong>de</strong>structive use of certain plants has had a negligable negative<br />
effect on the numbers of these plants within their habitat. It may even be that, in<br />
certain circumstances, specific plants may benefit from their ‘usefulness’ to local<br />
villages.<br />
<strong>The</strong> fruits of some tree species are edible to lemurs. In particular, Tamarindus indica,<br />
Ficus spp. and Diospyros spp. are important food sources for lemurs within the<br />
region. In<strong>de</strong>ed, Ficus spp. and Tamarindus indica could be classified as keystone<br />
species, such is their importance to the regional fauna. <strong>The</strong>se species are found in<br />
comparatively large numbers, often as adult trees. This is partially due to the use of<br />
their fruits by man, and partly due to cultural beliefs. Tamarindus indica and Ficus<br />
spp. have important spiritual significance in southwest Madagascar; in many areas<br />
large trees of these species are sacred. This may partially explain the dominance of<br />
these tree species in the upper size classes within the areas of gallery forest in the<br />
region, along with the apparent low inci<strong>de</strong>nce of clearance of adult trees of these<br />
17
species for timber. It is fortunate, or perhaps well judged, that the two keystone<br />
species within gallery forest are provi<strong>de</strong>d with such protection.<br />
4.4.3 Threats to habitats<br />
<strong>The</strong> areas of spiny forest and gallery forest are very different in terms of plant<br />
composition and structure, but the threats to both of these habitats are similar. Large<br />
areas of gallery and spiny forest have been cleared cultivation, predominantly maize<br />
and cassava. Many trees have been felled for timber, and areas have been cleared for<br />
the production of charcoal. Grazing of cleared areas has prevented regeneration.<br />
<strong>The</strong>se <strong>de</strong>structive uses of plant resources have led to the clearance of almost all<br />
significant areas of gallery forest along the south bank of the Onilahy River, mainly<br />
for cultivation purposes. <strong>The</strong> gallery forests contained within the surveyed areas<br />
represent the last significant forested areas along the Onilahy River. On the hillsi<strong>de</strong>s,<br />
large areas of spiny forest have been cleared for cultivation and charcoal production.<br />
Within all surveyed gallery forest quadrats, there was clear evi<strong>de</strong>nce of previous<br />
forest clearance for the purposes of cultivation or charcoal production. <strong>The</strong> gallery<br />
forest quadrats around Lake Antafoky contained evi<strong>de</strong>nce of fairly recent fire damage<br />
and slash-and-burn activities. In fact, the area was observed in the process of<br />
clearance in September 2001. <strong>The</strong> need for agricultural land is great around Lake<br />
Antafoky, so the threat to the areas of gallery forest in this area is severe. <strong>The</strong> gallery<br />
forest quadrats to the west of the Sept Lacs region (around the village of Mahaleotse)<br />
showed evi<strong>de</strong>nce of past use for small-scale charcoal production. However, the areas<br />
that were cleared for charcoal production appeared to be fairly small and old, and<br />
regeneration was evi<strong>de</strong>nt. Unfortunately, goats were regularly seen in the gallery<br />
forests throughout the surveyed area, which will have affected the forest structure and<br />
reduced the rate of regeneration consi<strong>de</strong>rably.<br />
<strong>The</strong> areas of spiny forest around Lake Antafoky showed comparatively low levels of<br />
disturbance within the quadrats and on the hillsi<strong>de</strong>s, although large areas of cleared<br />
forest were visible on the hilltops. <strong>The</strong>se areas had been cleared several years<br />
previously for the cultivation of maize. <strong>The</strong> spiny forest within the Sept Lacs region<br />
was fairly intact within the surveyed quadrats and on the hillsi<strong>de</strong>s, but an area of<br />
cleared forest was located on the plateau. This area was extremely large, extending for<br />
several kilometres across the plateau, and had been cleared primarily for maize<br />
cultivation.<br />
Taking into account the relative areas both of the spiny forest and gallery forest, it is<br />
evi<strong>de</strong>nt that the habitat of primary concern is gallery forest, particularly around Lake<br />
Antafoky. <strong>The</strong> remaining areas of gallery forest within the region are fairly small and<br />
are gradually becoming <strong>de</strong>gra<strong>de</strong>d through selective logging and localised clearance.<br />
<strong>The</strong> structure and quality of these forests are therefore affected, and they are<br />
becoming fragmented. In comparison, the spiny forests within the surveyed region<br />
show large areas of relatively intact forest on the hillsi<strong>de</strong>s, although clearance for<br />
cultivation has severely impacted the spiny forest in large areas across the plateau.<br />
18
5. Mammals<br />
5.1 Introduction<br />
Mammals occur in a wi<strong>de</strong> variety of terrestrial and aquatic habitats throughout the world,<br />
and constitute an essential aspect of the fauna in Madagascar. Because of the long period<br />
of isolation of Madagascar from other landmasses, the mammal fauna has evolved<br />
separately and created species and communities en<strong>de</strong>mic to Madagascar (Hutcheon,<br />
1994; Cook et al, 1991). As with many islands, Madagascar lacks many familiar higher<br />
taxa such as the wild cats, dogs, monkeys, apes and the large herbivores that dominate<br />
the plains of mainland Africa (Cook et al, 1991; Garbutt, 1999). Only six or<strong>de</strong>rs are<br />
present on Madagascar - Chiroptera, Insectivora, Ro<strong>de</strong>ntia, Carnivora, Primates and<br />
Artiodactyla. Of all the native species, of which more than 90 % are en<strong>de</strong>mic to the<br />
island, the best known are the primate or<strong>de</strong>r of lemurs and insectivore or<strong>de</strong>r of tenrecs.<br />
Asi<strong>de</strong> from the charismatic lemurs, most mammal species are poorly studied, in<strong>de</strong>ed<br />
the subor<strong>de</strong>r of microchiroptera are one of the most diverse or<strong>de</strong>rs and yet the least<br />
studied (Stephenson, 1994; Goodman, 1996). <strong>The</strong>re have been no comprehensive<br />
small mammal studies in the region of the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>. A<br />
short survey was conducted in this area (Raxworthy, 1995), but no mammals were<br />
captured. No other reported mammal studies have been un<strong>de</strong>rtaken in the region prior<br />
to this survey, so the mammal fauna is comparatively un-studied and unknown.<br />
Small mammals have important roles in their ecosystems because they are central to the<br />
food chain. <strong>The</strong>y feed on vegetation and invertebrates and are preyed upon by<br />
carnivorous mammals, reptiles and birds. Furthermore, mammals often function as seed<br />
dispersers and are therefore important for plant species distribution and regeneration.<br />
Small mammals are comparatively easy to catch, handle, and i<strong>de</strong>ntify. <strong>The</strong>y have<br />
short life spans so changes in diversity, <strong>de</strong>nsity, and community composition over<br />
time are relatively easy to <strong>de</strong>tect and monitor. Small mammals are a useful ecological<br />
indicator group. Data from repeated surveys can be used to monitor the level of<br />
disturbance to an area and thereby contribute to the management of these areas.<br />
As with most other groups of animals in Madagascar, the mammals are threatened by<br />
habitat <strong>de</strong>struction, excessive hunting and competition by introduced species.<br />
5.2 Methods<br />
As mammals have adapted to many different habitats, several methods are necessary to<br />
reveal the species occurring within an area.<br />
5.2.1 Pitfall trapping<br />
Small insectivores and ro<strong>de</strong>nts were captured using pitfall traps. Pitfall traps are a cheap<br />
and effective method, used to sample ground dwelling small mammal, reptile and<br />
amphibian species. <strong>The</strong> method consists of buckets sunk into the ground along a<br />
plastic sheeting drift fence, that channels the animals into the buckets (commonly<br />
called bucket lines). <strong>The</strong> buckets were 275 mm <strong>de</strong>ep, with a top diameter of 290 mm.<br />
Three 100-metre long bucket lines, each with 11 buckets, were used at every trapsite.<br />
19
5.2.2 Sherman and mesh trapping<br />
Ro<strong>de</strong>nts, large insectivores and mouse lemurs were captured using either Sherman or<br />
mesh traps. <strong>The</strong> Sherman traps were of the standard size, 225 mm x 87 mm x 75 mm.<br />
<strong>The</strong> mesh traps measured 400 mm x 150 mm x 150 mm. <strong>The</strong> traps were placed in<br />
suitable habitats at each trapsite, typically with two of each type of trap near each pitfall<br />
line. A total of six Sherman traps and six mesh traps were used at each trapsite apart<br />
from trapsites 13-16, when 80 Sherman traps and 20 mesh traps were used. Suitable<br />
locations for placing the traps were in areas with a <strong>de</strong>nse un<strong>de</strong>rstorey, along fallen and<br />
rotting logs, or near obvious food sources such as fruiting trees. Traps were situated as<br />
far away from the pitfall lines as possible, with a minimal distance of 10 metres. <strong>The</strong>y<br />
were typically positioned at least 50 m apart. For the capture of arboreal species, traps<br />
were fastened with tape to a suitable branch or vine.<br />
5.2.3 Observations, tracks and calls<br />
Diurnal and nocturnal observations, tracks, dung and vocalisations were used to i<strong>de</strong>ntify<br />
mammal species that could not be trapped. Specimens were photographed wherever<br />
possible to provi<strong>de</strong> confirmation of species i<strong>de</strong>ntification. I<strong>de</strong>ntifications of species<br />
through photographs, tracks and calls were ma<strong>de</strong> using the book ‘Mammals of<br />
Madagascar’ (Garbutt, 1999).<br />
5.2.4 Mist nets and harp traps<br />
Microchiroptera were captured using mist nets and harp traps. <strong>The</strong> traps were set in<br />
suitable locations along flight-paths and around roosts or feeding sites. I<strong>de</strong>al places were<br />
across forest clearings, forest edges, paths, roads, streams and ponds. Mist nets were<br />
placed in enclosed areas with overhanging and surrounding vegetation whenever<br />
possible, increasing the capture rate by channelling the bats into the nets. <strong>The</strong> mist net<br />
composition <strong>de</strong>pen<strong>de</strong>d on the location and habitat. <strong>The</strong> number of mist nets used varied<br />
<strong>de</strong>pending on the habitat, but typically consisted of three or four nets. Short nets were<br />
suitable in <strong>de</strong>nse forest and on narrow paths and roads, whereas long nets were suitable<br />
for open spaces and over water. Harp traps were i<strong>de</strong>al for trapping around roosts where<br />
there were a large number of bats, as it was much easier and more efficient to collect<br />
large numbers of individuals in a harp trap than to untangle them from a mist net. Harp<br />
traps used in conjunction with mist nets ensured a high probability of capturing the<br />
majority of species that were present. Some bats learn to avoid the mist nets and harp<br />
traps, so trapping was un<strong>de</strong>rtaken for no more than 2-3 days at each location.<br />
5.2.5 Owl pellets<br />
Owl pellets were collected. Most owls feed partly on small mammals. <strong>The</strong>y usually<br />
swallow their prey whole, and after digesting the flesh, the indigestible fur and bones are<br />
regurgitated as compact pellets. Owl pellets can be broken apart and analysed for prey<br />
species. Small mammals can be readily i<strong>de</strong>ntified from their skulls, which are fairly<br />
intact in the pellets. <strong>The</strong> analysis of owl pellets can contribute to i<strong>de</strong>ntifying the small<br />
mammal species present within a locality. Owl pellet analysis is particularly useful for<br />
<strong>de</strong>tecting the presence of species that are rare or not readily collected with the used<br />
methods. Useful owl species inclu<strong>de</strong> the Madagascar long-eared owl, Asio<br />
20
madagascariensis and the Barn Owl, Tyto alba. Both species feed mainly on mammals<br />
and return after foraging to their roosting sites, where they digest and regurgitate pellets.<br />
Owl pellets were collected frequently and regularly during the survey period, so the<br />
data could be used to i<strong>de</strong>ntify seasonal variation in small mammal populations.<br />
5.2.6 Specimen handling<br />
Captured specimens were measured, weighed and marked for individual recognition.<br />
<strong>The</strong> sex and breeding status of the captured individuals was also recor<strong>de</strong>d, based on<br />
visible sexual characteristics. Specimens were <strong>de</strong>fined as juvenile, sub-adult or adult,<br />
<strong>de</strong>pending on their size, weight and level of sexual maturity.<br />
For a <strong>de</strong>tailed study of Rattus rattus, captured specimens were allocated a specific age<br />
class. A juvenile was <strong>de</strong>fined as having a head-body length of less than 100 mm and a<br />
weight of less than 50 g. A sub-adult was <strong>de</strong>fined as having a head-body length between<br />
100 mm and 150 mm and a weight of less than 100 g, whereas an adult was <strong>de</strong>fined as<br />
having a head-body length of over 150 mm and a weight of over 100 g.<br />
See <strong>Frontier</strong>-Madagascar Forest Research Programme Methodology Training Manual<br />
(<strong>Frontier</strong> Madagascar, 2003) for <strong>de</strong>tailed <strong>de</strong>scriptions of all methodologies used.<br />
5.3 Results<br />
Seventeen mammal species, representing 4 or<strong>de</strong>rs and 6 families, were captured<br />
during a total of 3,632 pitfall nights, 4,158 trap nights, and 1,829 mist net hours. A<br />
further 9 species, representing a further 3 or<strong>de</strong>rs and 5 families, were i<strong>de</strong>ntified either<br />
from direct observation, or from tracks and dung. In total, 26 species representing 7<br />
or<strong>de</strong>rs and 11 families were found in the surveyed areas during the study.<br />
Classifications follow Garbutt (1999).<br />
<strong>The</strong> or<strong>de</strong>rs represented inclu<strong>de</strong> Macrochiroptera (1 species), Microchiroptera (7<br />
species), Insectivora (6 species), Ro<strong>de</strong>ntia (3 species), Carnivora (2 species),<br />
Artiodactyla (1 species) and Primates (6 species). See Appendix 6 for species list.<br />
5.3.1 Trapping success<br />
Sherman and mesh traps traps proved most successful in the dry season, during the<br />
months of July and August, with capture rates of 12.5% and 11% respectively.<br />
Conversely, the capture rate of mammals in pitfall traps was highest in the wet season,<br />
during the months of January and February, with capture rates of 18% and 10%<br />
respectively. Mist nets and harp traps were successful throughout the year, with<br />
gallery forest habitats proving most successful. <strong>The</strong> spiny forest habitats were<br />
extremely damaging to the mist nets and the bats were dispersed over a wi<strong>de</strong> area,<br />
whereas the gallery forest canopy provi<strong>de</strong>d a more suitable environment for trapping.<br />
Traps placed across water proved to be successful, as did trapping near to roost sites<br />
(see Section 6).<br />
<strong>The</strong> number of ro<strong>de</strong>nts captured was highest in July (24 captures), whereas the number<br />
of insectivores captured was highest in January (50 captures). Ro<strong>de</strong>nts were most<br />
frequently caught in the dry season (June to August) where they ma<strong>de</strong> up 89-100% of<br />
21
the total live-trap captures (see Figure 4). Conversely, insectivores were most<br />
frequently caught in the wet season (January and February) when they ma<strong>de</strong> up 96-<br />
100% of the total captures.<br />
<strong>The</strong> most frequently captured mammal species throughout the survey was Rattus<br />
rattus, with thirty-three caught individuals from an accumulated total of fifty captures.<br />
Composition and annual patterns in the surveyed populations could therefore be<br />
<strong>de</strong>tected. Of the 19 adult individuals captured, 10 were males and 9 were females,<br />
indicating an even sexual distribution in the population.<br />
Annual mammal trapping<br />
No. of captures<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
July<br />
August<br />
October<br />
November<br />
January<br />
February<br />
Month<br />
March<br />
May<br />
June<br />
Figure 4. Annual trapping of ro<strong>de</strong>nts and insectivores<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
0<br />
% of all captures<br />
ro<strong>de</strong>nt<br />
captures<br />
Insectivore<br />
captures<br />
% ro<strong>de</strong>nts of<br />
all captures<br />
%<br />
insectivores<br />
of all<br />
captures<br />
5.3.2 Rattus rattus population composition<br />
In October, most captured individuals of R. rattus were breeding adults, whereas in<br />
May most captured individuals were juveniles (see Figure 5). This indicated that the<br />
majority of the mature population mated around the onset of the rainy season. As food<br />
is most abundant during the wet season, this would provi<strong>de</strong> the necessary resources<br />
for <strong>de</strong>veloping juveniles. A small number of breeding adults and sub-adults were<br />
captured throughout the survey. This indicates that, although there is a clear ‘peak<br />
season’, this species does continue to breed throughout the year.<br />
Annual R. rattus population composition<br />
12<br />
10<br />
No. of<br />
individuals<br />
8<br />
6<br />
4<br />
2<br />
Adult<br />
breeding<br />
Adult<br />
Sub-adult<br />
0<br />
July/August October November May June<br />
Month<br />
Juvenile<br />
Figure 5. Annual composition of the Rattus rattus population.<br />
22
5.3.3 Observations, tracks and calls<br />
Observations, tracks and vocalisations accounted for the i<strong>de</strong>ntification of nine species<br />
that were not captured using standard trapping methods (see Appendix 6 for <strong>de</strong>tails).<br />
This inclu<strong>de</strong>d five species of lemur. Primates are protected species, and specific<br />
permission is necessary in or<strong>de</strong>r for specimens to be collected. <strong>The</strong>refore,<br />
i<strong>de</strong>ntifications based on observations are most suitable.<br />
<strong>The</strong> bush pig, Potamochoerus larvatus; the Indian civet, Vivvericula indica; and the<br />
fosa, Cryptoprocta ferox, are large, shy nocturnal animals who are rarely trapped.<br />
Tracks were seen, as were droppings. <strong>The</strong> Indian civet was sighted on a small number<br />
of occasions.<br />
5.3.4 Owl pellet analysis<br />
<strong>The</strong> analysis of owl pellets is a time-consuming process and at the time of writing this<br />
report the analyses have not been completed. <strong>The</strong> results will be ad<strong>de</strong>d in a later<br />
version of this report.<br />
5.4 Discussion<br />
5.4.1 Ecological distribution and species richness<br />
Many of the plants and animals of Madagascar display an exceptionally high level of<br />
en<strong>de</strong>mism, and the mammals are no exception. Of the twenty-six species found in the<br />
surveyed areas, twenty-one species (81%) were en<strong>de</strong>mic. <strong>The</strong> infraor<strong>de</strong>r<br />
Lemuriformes, of the or<strong>de</strong>r Primate, are en<strong>de</strong>mic to Madagascar. Six lemur species<br />
were i<strong>de</strong>ntified during the surveys. Five species that were present in the surveyed<br />
areas are not en<strong>de</strong>mic to the country, at least three of which are thought to have been<br />
introduced by human colonists. Although this is a fairly low number, these species<br />
can damage the native ecosystems and are seen as pests in Madagascar.<br />
Compared to other tropical and sub tropical regions throughout the world, the<br />
mammals of Madagascar show low diversity (Lundahl & Olsson, 2000). Also, the<br />
humid areas of the east coast show a higher diversity of mammals than the arid areas<br />
of the west. However, the surveyed areas within the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong><br />
<strong>Belomotse</strong> showed high diversity compared to the findings of other surveys in similar<br />
habitats (see Raxworthy, 1995). <strong>The</strong> high number of mammal species found in the<br />
survey area may be due to the diversity of habitats in the surveyed areas. It may also<br />
reflect the fact that the areas were surveyed over a long period of time, covering both<br />
the wet and the dry season, so the probability of catching and observing many<br />
different species was high.<br />
<strong>The</strong> or<strong>de</strong>r of chiroptera is the most diverse of Madagascar’s small mammal fauna. <strong>The</strong><br />
taxonomy for many genera and species is still being revised and new species and<br />
range extensions are frequently being found. During the survey, seven species of<br />
microchiroptera were i<strong>de</strong>ntified, two of which are un<strong>de</strong>scribed species. Three species<br />
of megachiroptera occur in Madagascar. <strong>The</strong> two largest species, Pteropus rufus and<br />
Eidolon dupreanum, have previously been found in the southwest. During this survey,<br />
fruit bats were observed roosting in caves or overhangs on cliff faces, and feeding in<br />
23
fruiting trees at night. <strong>The</strong> specimens appeared to be Eidolon dupreanum. As no<br />
specimens were captured, this i<strong>de</strong>ntification is based on the fact that E. dupreanum<br />
typically roosts in caves, whereas Pteropus rufus typically roosts in trees. Roosting<br />
colonies seemed to vary in size <strong>de</strong>pending on the availability and capacity of suitable<br />
caves. Colonies with as few as ten individuals were observed. <strong>The</strong> roosting animals<br />
were often semi-active and noisy throughout the day, emerging at dusk.<br />
All species of the larger tenrecs; Tenrec ecaudatus, Setifer setosus and Echinops<br />
telfairi, are known from southwest Madagascar. All of these species were collected<br />
from the surveyed areas. T. ecaudatus was heard foraging in the un<strong>de</strong>rgrowth at night<br />
and young individuals were caught in pitfall traps. This species was found both in<br />
gallery forest near to water and in the arid spiny forest on the hills.<br />
Large-eared tenrecs and shrew tenrecs, belonging to the genera Geogale and<br />
Microgale consecutively, are known from southwest Madagascar. Geogale aurita is<br />
restricted to the southwest region where it is typically found in dry <strong>de</strong>ciduous and<br />
spiny forests. <strong>The</strong> genus Microgale contains approximately twenty species that are<br />
found in many different habitats throughout the island. During the surveys, G. aurita<br />
was captured in low numbers in the dry spiny forest on the hillsi<strong>de</strong>s and plateau. One<br />
species of Microgale was found during the survey. See section 5.4.4 for a discussion<br />
of this species. M. pusilla has previously been collected from the Toliara region<br />
(Garbutt 1999), though it was not i<strong>de</strong>ntified in the surveyed areas. This species is<br />
typically found in comparatively high numbers in suitable habitats, so it is suggested<br />
that this species does not occur in the surveyed area.<br />
Four species from the family Cheirogaleidae (Mouse and Dwarf Lemurs) were<br />
i<strong>de</strong>ntified within the surveyed areas. <strong>The</strong> genus of mouse lemurs, Microcebus, was<br />
previously assumed to consist of two species, with one of them occurring in the west,<br />
but recent studies suggest that at least seven species are found in western Madagascar<br />
(Rasoloarison et al. 2000). Two species, Microcebus murinus and M. griseorufus,<br />
were i<strong>de</strong>ntified during the survey based on observations, DNA samples and the<br />
preliminary analysis of owl pellets (Goodman, pers. comm). M. murinus is distributed<br />
along the west coast of Madagascar, where it appears to be common and<br />
comparatively abundant. During the survey, individuals were typically observed in the<br />
dry transitional scrub and spiny forest. M. griseorufus was observed in similar<br />
habitats.<br />
<strong>The</strong> fat-tailed dwarf lemur, Cheirogaleus medius, is distributed along the west coast<br />
of Madagascar in dry <strong>de</strong>ciduous forest and gallery forest. During the survey, it was<br />
observed frequently between the months of October and March. This species was<br />
fairly inactive during the dry season from May to September, only occasionally<br />
emerging. C. medius appears to be comparatively abundant, although it was only<br />
observed in gallery forest habitats. Coquerel’s dwarf lemur, Mirza coquereli, is<br />
known from the north-west of Madagascar in the Sambirano region, and also from<br />
several locations in the west. Previous unconfirmed sightings suggested the existence<br />
of populations south of this range, along the north bank of the Onilahy River. <strong>The</strong><br />
species was positively i<strong>de</strong>ntified in this area during the surveys. <strong>The</strong> presence of M.<br />
coquereli in this area probably <strong>de</strong>fines the most southern limit to the range of this<br />
species, as no sightings have been ma<strong>de</strong> south of the Onilahy River. In comparison to<br />
24
Cheirogaleus medius, the Coquerel’s dwarf lemur was not seen to reduce its activity<br />
levels during the dry season. It was observed all year round, though only in gallery<br />
forest habitats.<br />
Verreaux’s sifaka, Propithecus verreauxi verreauxi, and the ring-tailed lemur, Lemur<br />
catta, were found in large numbers throughout the surveyed areas. Both species have<br />
been recor<strong>de</strong>d in the Toliara region and were known to be present in the Sept Lacs<br />
region. <strong>The</strong>y were typically found in gallery forest and spiny forest. However, all<br />
observed groups of both species appeared to have the core area of their home ranges<br />
in the galley forest. Both species spent much more time in the gallery forest during the<br />
dry season than during the wet season, when they often foraged in the spiny forest on<br />
the hillsi<strong>de</strong>s during the day. <strong>The</strong> two species co-existed in the same areas, often<br />
occupying the same trees.<br />
<strong>The</strong> elusive fosa, Cryptoprocta ferox, is distributed throughout much of the island in<br />
many different habitats. It is very shy and difficult to observe. No confirmed direct<br />
observations were ma<strong>de</strong> of the fosa during the survey. However, on several occasions,<br />
unconfirmed sightings were ma<strong>de</strong> of a large animal by field staff and volunteers.<br />
Also, dung was found containing the fur and spines of Tenrec ecaudatus. This species<br />
of tenrec could only have been eaten by either C. ferox or a large dog (S. Goodman,<br />
pers. comm), although domestic dogs were not seen in the remote location where the<br />
dung was found. Furthermore, local villagers have pointed the species out from<br />
photos, stating that the species has been seen around their village at night. It is<br />
therefore tentatively assumed that the fosa is present in the surveyed area.<br />
5.4.2 Introduced Species<br />
<strong>The</strong> impact of introduced species on the native wildlife of Madagascar is not yet fully<br />
un<strong>de</strong>rstood, and although relatively few mammal species have been introduced to<br />
Madagascar, further studies are nee<strong>de</strong>d to i<strong>de</strong>ntify the direct and indirect effects that<br />
these species are having on populations of native species.<br />
Two introduced species from the sub-family Murinae were found during the survey,<br />
Rattus rattus and Mus musculus. Both were found in high <strong>de</strong>nsities. Both species were<br />
probably introduced to Madagascar within the last 2,000 years, since man arrived on<br />
Madagascar. Being generalist and opportunistic, both species appear to have adapted<br />
well to the new surroundings. Both R. rattus and M. musculus breed rapidly and<br />
populations quickly reach high <strong>de</strong>nsities. <strong>The</strong>se <strong>de</strong>nsity levels cause extensive<br />
resource competition with those native species occupying the same guilds. Without<br />
exception, R. rattus now inhabits all habitats on the island, even penetrating into the<br />
most <strong>de</strong>nse forested mountains far from human habitation (Goodman, 1995). Previous<br />
surveys revealed low numbers of R. rattus in the Toliara region (Raxworthy, 1995).<br />
This led to the conclusion that R. rattus may not be as abundant in native habitats in<br />
this region as elsewhere in Madagascar. However, during this study R. rattus was<br />
caught frequently in all habitats surveyed, showing that the species is very abundant<br />
and wi<strong>de</strong>spread. Furthermore, it was frequently observed around the campsites near<br />
tents and food sources. It is well known by the local people, who consi<strong>de</strong>r it to be a<br />
pest.<br />
25
Rattus rattus has been shown to be a harmful pest, bringing several diseases to<br />
Madagascar. One of these diseases, plague, which is carried by the fleas of R. rattus,<br />
only occurred on the island after the introduction of rats. Today, the disease is still a<br />
problem, affecting approximately 800 humans every year (S. Goodman, pers. comm.,<br />
2002). Furthermore, other parasites and viruses carried by this animal have caused<br />
severe problems for native ro<strong>de</strong>nt species. With little immunity to these exotic<br />
diseases, populations have <strong>de</strong>creased and several species may have been driven close<br />
to extinction as a result. <strong>The</strong> en<strong>de</strong>mic ro<strong>de</strong>nt Eliurus myoxinus was found in low<br />
numbers in the surveyed sites compared to the introduced ro<strong>de</strong>nts R. rattus and Mus<br />
musculus. Comparatively little is known about the populations of E. myoxinus in the<br />
region, and further surveys need to clarify whether the low <strong>de</strong>nsity observed is normal<br />
for the species, or whether it is suffering from competition with the introduced<br />
species.<br />
<strong>The</strong> pygmy musk shrew, Suncus madagascariensis, from the family Soricidae, has<br />
been found during the survey. This species is the world’s smallest terrestrial mammal,<br />
with a head-body length of 50- 55 mm and a weight of 3-4 g (Garbutt, 1999).<br />
However, the individuals that were captured during this survey were all consi<strong>de</strong>rably<br />
smaller, with a head-body length of 35-40 mm and a weight of approximately 2 g.<br />
<strong>The</strong> en<strong>de</strong>mic status of this species is uncertain, but it is currently assumed that the<br />
populations on Madagascar are introduced (S. Goodman, pers comm). This species<br />
was commonly found both in gallery and spiny forest habitats, but the impact of the<br />
species on native species of ro<strong>de</strong>nts and tenrecs is not known. Being insectivorous, it<br />
may compete with other small insectivores for food.<br />
<strong>The</strong> introduced ungulate Potamochoerus larvatus larvatus occurs along the west coast<br />
of Madagascar. Spoors were found both in marshy and in forested areas within the<br />
surveyed areas. This species poses a threat to some native species in these habitats. It<br />
forages vigorously in the un<strong>de</strong>rgrowth, digging up the soil as it feeds on roots, fruits,<br />
invertebrates, vertebrates and bird eggs. It is also known to feed on tortoise eggs.<br />
<strong>The</strong> introduced civet, Viverricula indica, is potentially harmful as it preys upon a wi<strong>de</strong><br />
variety of native animal species, particularly small vertebrates. V. indica was observed<br />
in several locations throughout the surveyed areas, often near to villages.<br />
5.4.3 Seasonal variation<br />
Due to the arid climate in the region, some animal groups ten<strong>de</strong>d to lower their<br />
activity during the dry season, when much of the vegetation dried out and food<br />
sources became scarce. However, ro<strong>de</strong>nts appeared to stay active all year round and<br />
were observed regularly throughout the year in all surveyed areas. <strong>The</strong>y were trapped<br />
in higher numbers during the dry season than during the wet season, see figure 4.<br />
Being omnivorous and opportunistic in their feeding behaviour, ro<strong>de</strong>nts can survive<br />
on a wi<strong>de</strong> variety of food items, e.g. fruits, seeds and invertebrates, which will allow<br />
them to be active at times where other groups are inactive. Stephenson (1994) found<br />
the same levels of activity in a year-round trapping survey in central Madagascar. He<br />
suggested that due to the abundance of ‘natural’ food sources in the wet season, the<br />
trap bait seem less attractive to the animals, so trapping success for ro<strong>de</strong>nts is reduced<br />
at this time. This hypothesis explains the high trap success during the dry season in<br />
the present survey, with the accompanying dip in trapping during the wet season.<br />
26
However, only two year-round trapping surveys have been carried out in Madagascar<br />
and more long-term studies are nee<strong>de</strong>d.<br />
Insectivores were mainly caught during the wet season, with almost total absence<br />
during the dry season (July and August). During these dry periods, insectivores lower<br />
their metabolism and enter a state of torpor, hiding in burrows, caves, fallen trees and<br />
other such sites. Insectivores are highly <strong>de</strong>pen<strong>de</strong>nt on the availability of insects and<br />
other invertebrates as a food source. During the dry season the temperature can be as<br />
low as 25º C during the day and below 10º C at night. Combined with the lack of<br />
precipitation, this leads to a seasonal climate that is unfavourable for invertebrates.<br />
<strong>The</strong> numbers of invertebrates during the dry months are therefore low and may not be<br />
sufficient for the insectivorous species.<br />
Although little is known about the ecology and behaviour of the musk shrews Suncus<br />
murinus and S. madagascariensis, they are thought to be active and breed throughout<br />
the year (Garbutt, 1999; S. Goodman, pers. comm., 2002). During this survey,<br />
trapsites were repeated in similar locations during both the dry and wet seasons. S.<br />
madagascariensis was the only insectivorous species collected during the months of<br />
June and July, though only in small numbers. This species was only captured in July<br />
or August on one occasion (after light showers), whereas between October and March<br />
it was captured in comparatively high numbers. This indicates that the species has a<br />
longer activity period than the other collected insectivore species, and also that it does<br />
not hibernate for a long period, if at all.<br />
5.4.4 Rare species and range extensions<br />
Two species of Triaenops bats, Triaenops furculus and T. rufus, were found in the<br />
surveyed area. Both have previously been found in lowland areas of northern and<br />
western Madagascar (Garbutt, 1999). <strong>The</strong> i<strong>de</strong>ntification of both of these species<br />
within the region of the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>, combined with nonconfirmed<br />
sightings further south, suggest that the ranges of these species are greater<br />
than previously assumed. T. furculus is regar<strong>de</strong>d as vulnerable, with a <strong>de</strong>creasing<br />
population (IUCN, 2002), so the presence of a large population of this species within<br />
the survey area is of high importance to the conservation of the species, see section 6.<br />
A species of insectivorous bat from the genus Miniopterus was captured from a<br />
number of locations throughout the surveyed areas. This species has characteristics<br />
unlike any others from this genus and is thought to be an un<strong>de</strong>scribed species (S.<br />
Goodman, pers. comm., 2002). Further specimen collections, with associated tissue<br />
samples, are nee<strong>de</strong>d to confirm the i<strong>de</strong>ntity of this species.<br />
<strong>The</strong> taxonomy of the bat genus Pipistrellus is unclear in Madagascar. One species<br />
from this genus was collected during the surveys. This appears to be an un<strong>de</strong>scribed<br />
species (S. Goodman, pers. comm., 2002).<br />
Microgale brevicaudata is distributed throughout a number of habitats on the<br />
northeast and west coast of Madagascar. It is also found south of Toliara. An<br />
unknown Microgale species was captured within riparian and spiny forest habitats<br />
throughout the region of the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>. <strong>The</strong> species has<br />
27
proved difficult to i<strong>de</strong>ntify, as the measurements and external characteristics do not<br />
resemble those of M. brevicaudata. This may be an un<strong>de</strong>scribed species of Microgale,<br />
or a previously un<strong>de</strong>scribed regional variant of M. brevicaudata. Specimens and DNA<br />
samples have been sent to Chicago Field Museum for analysis. Results are not yet<br />
available. Until formal i<strong>de</strong>ntification has been completed, the species has been<br />
tentatively assigned as M. brevicaudata.<br />
<strong>The</strong> Coquerel’s dwarf lemur, Mirza coquereli, is currently listed as vulnerable with a<br />
<strong>de</strong>creasing population (IUCN, 2002). During this survey this species was frequently<br />
observed, foraging high in the canopy in gallery forest. It was not observed in<br />
<strong>de</strong>gra<strong>de</strong>d gallery forest with low or fragmented canopies, nor was it observed in spiny<br />
forest. This indicates that the species is <strong>de</strong>pen<strong>de</strong>nt on undisturbed riparian forest.<br />
Furthermore, the species is solitary with large home range sizes (Sussman et al.,<br />
1985), so large areas of continuous forest are necessary in or<strong>de</strong>r to sustain viable<br />
populations. Comparatively little gallery forest is left in Madagascar, and the<br />
remaining areas are highly fragmented. <strong>The</strong>se factors are part of the reason for the<br />
classification of M. coquereli as vulnerable. <strong>The</strong> unexpected existence of<br />
comparatively large populations in the surveyed area is a positive sign and an<br />
important find for the conservation of this species.<br />
5.4.5 Threats to mammal conservation<br />
A number of factors affect mammal species <strong>de</strong>nsity and distribution throughout<br />
Madagascar. Several of these factors impact severely on specific mammal species,<br />
causing <strong>de</strong>clines in their overall populations and threatening their survival.<br />
Throughout Madagascar, large areas of land are being cleared for agriculture. In<br />
addition, forested areas are cleared for the production of charcoal and the collection of<br />
firewood and building materials. Some en<strong>de</strong>mic mammal species, such as Microcebus<br />
murinus, are distributed throughout a wi<strong>de</strong> variety of habitats and adapt well to<br />
changing environments. However, many en<strong>de</strong>mic mammal species are sensitive to<br />
change, so populations are often adversely affected by habitat <strong>de</strong>gradation. Many<br />
mammal species are limited to specific habitats and comparatively small areas in<br />
Madagascar. Clearance and fragmentation of the habitats and areas can be reflected by<br />
a <strong>de</strong>crease in the populations of these mammals, which can lead to species becoming<br />
vulnerable or endangered with extinction.<br />
<strong>The</strong> remaining areas of gallery forest within the region of the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong><br />
<strong>de</strong> <strong>Belomotse</strong> are un<strong>de</strong>r severe threat through clearance for cultivation and charcoal<br />
production. <strong>The</strong> corridor of gallery forest between Antafoky Lake and the Sept Lacs<br />
region is very fragmented and it is doubtful whether it can be classified as a corridor<br />
for certain mammal species. <strong>The</strong> small populations of many species that have become<br />
isolated by fragmentation of habitats may be in danger of local extinction. An<br />
example of this would be the dwarf lemur, Mirza coquereli, which is sensitive to<br />
habitat fragmentation. This species was only found in gallery forest and now appears<br />
to form three fragmented populations; one around Lake Antafoky, a second in an area<br />
of forest near to Mahaleotse and the last in the upper reaches of the Sept Lacs region.<br />
Further <strong>de</strong>gradation of these sites will limit the range of this species still further.<br />
28
In many areas, species are hunted for food or collected for the pet tra<strong>de</strong>. Mammal<br />
species that are impacted by these activities inclu<strong>de</strong> lemurs and tenrecs. <strong>The</strong> large<br />
tenrec species are seen as a staple food source in many areas. For example, Tenrec<br />
ecaudatus is frequently hunted throughout much of its range. However, the species<br />
reproduces often and has large litters, so <strong>de</strong>clines in distribution and population levels<br />
have not been reported, although this may be the result of a lack of research and<br />
monitoring. Relatively high numbers of T. ecaudatus were caught during the survey<br />
and population <strong>de</strong>nsities seemed to be high. However, as the human population (both<br />
regionally and nationally) increases, the long-term impact of hunting and collection<br />
on this species needs to be monitored throughout Madagascar. <strong>The</strong> fruit bats Eidolon<br />
dupreanum and Pteropus rufus are hunted for food, potentially causing local<br />
extinctions and threatening the future of these species in certain areas (Hutcheon,<br />
1994). E. dupreanum was found throughout the surveyed areas. However, because of<br />
the inaccessible locations of many of the roosting sites, the impact of hunting on this<br />
species does not seem severe.<br />
Ring-tailed lemurs, Lemur catta, are collected from the <strong>proposed</strong> park area for the pet<br />
tra<strong>de</strong>. Juveniles of this species are sold to hotels and restaurants in Toliara and nearby<br />
tourist resorts. Infant mortality in the wild is high for this species, so such collections<br />
may not severely impact upon the populations. However, if the mothers are killed to<br />
facilitate the collection of juveniles, or if collection rates increase, this could impact<br />
upon ring-tailed lemur populations in the area.<br />
6. Triaenops population study<br />
6.1 Introduction<br />
Preliminary surveys within the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong> revealed three<br />
species of microchiroptera roosting within a cave system in the Sept Lacs region. <strong>The</strong><br />
i<strong>de</strong>ntified species were Hipposi<strong>de</strong>ros commersoni, Triaenops rufus and T. furculus.<br />
Very little is known about the ecology, distribution or behaviour of these species in<br />
Madagascar. In particular, little is known about the en<strong>de</strong>mic species T. furculus,<br />
which is classified as vulnerable with a <strong>de</strong>creasing population, or T. rufus, which is<br />
classified as data <strong>de</strong>ficient (IUCN, 2002).<br />
A population study was un<strong>de</strong>rtaken in or<strong>de</strong>r to ascertain the approximate number of<br />
individuals within the cave system. Associated behavioural data were also recor<strong>de</strong>d.<br />
<strong>The</strong> study aimed to contribute to the basic knowledge and conservation status of these<br />
species.<br />
6.2 Methods<br />
In August 2002, an eight-day mark-recapture survey was un<strong>de</strong>rtaken, following Wilson<br />
et al. (1996), to estimate the populations of the species present within the cave system.<br />
<strong>The</strong> cave consisted of a large subterraneous cavern in limestone that opened to the<br />
outsi<strong>de</strong> through a narrow sinkhole. This hole measured approximately 2m x 3m.<br />
Previous explorations of the cave system by qualified cavers had i<strong>de</strong>ntified a flask-<br />
29
shaped limestone cave with an extremely <strong>de</strong>ep, branching central tunnel. <strong>The</strong> small,<br />
single exit to the cave provi<strong>de</strong>d an i<strong>de</strong>al opportunity to comprehensively survey the<br />
entire bat community that lived within the cave system.<br />
Two 1 m x 1 m harp traps were used in conjunction with two mist nets (9 m and 12 m),<br />
which were placed in suitable locations around the cave entrance. Trapping was<br />
maintained throughout the night from dusk until dawn. <strong>The</strong> time of each capture was<br />
recor<strong>de</strong>d, then captured individuals were i<strong>de</strong>ntified to species, sexed, measured, marked<br />
(by cutting the fur between the shoul<strong>de</strong>rs) and released. Individuals were marked over a<br />
total of 6 nights, encompassing approximately 144 harp trap hours and 1,260 mist net<br />
hours. During the following 2 nights, trapping was continued but no specimens were<br />
marked. Instead, the time of capture and numbers of marked and unmarked individuals<br />
were recor<strong>de</strong>d. This two-night trapping period encompassed approximately 42 harp trap<br />
hours and 480 mist net hours.<br />
6.3 Results<br />
<strong>The</strong> survey revealed large populations of Triaenops rufus and T. furculus. No<br />
individuals of the species Hipposi<strong>de</strong>ros commersoni were captured or observed, even<br />
though previous trapping had revealed a population of this species within the cave<br />
system.<br />
Comparing the ratio of marked to unmarked individuals captured during the final twonight<br />
collection period, the total population of each Triaenops species was estimated.<br />
Using the Lincoln in<strong>de</strong>x, the population of T. rufus was estimated to be approximately<br />
32,000 individuals, with a standard error of 3,060. <strong>The</strong> population of T. furculus was<br />
estimated to be approximately 9,000 individuals, with a standard error of 1,900. <strong>The</strong><br />
combined population totals of both Triaenops species equate to approximately 41,000<br />
individuals within the cave system.<br />
Approximately 2,700 individuals in total were captured, marked and released over 6<br />
nights, with consistently high numbers of individuals captured every night.<br />
<strong>The</strong> sex ratio of captured individuals was almost equal for both species, although T.<br />
rufus had a slightly higher number of males and T. furculus had a slightly higher<br />
number of females.<br />
30
6.4 Discussion<br />
6.4.1 Temporal variation<br />
At dusk and during the first hour of darkness (between 6.00 and 7.00 p.m.) many<br />
thousands of bats emerged from the sinkhole. Between 6.00 and 6.30 p.m. the<br />
majority of captured individuals were T. rufus. Large numbers of T. furculus emerged<br />
from the sinkhole slightly later, between 6.30 and 7.30 p.m. Between approximately<br />
7.00 until 9.00 p.m., few bats were seen or captured. After this time, a small number<br />
of bats began to return. Between approximately 9.00 p.m. and midnight the bats<br />
showed another peak in activity, with large numbers of both species returning to the<br />
cave. After this time, activity was low until approximately 4.30 a.m. when the<br />
remaining bats returned to the cave prior to sunrise. See figure 6.<br />
Nightly activity pattern<br />
Number of individuals<br />
1000<br />
900<br />
800<br />
700<br />
600<br />
500<br />
400<br />
300<br />
200<br />
100<br />
0<br />
rufus total<br />
Furculus total<br />
18:00-19:00<br />
20:00-21:00<br />
22:00-23:00<br />
Time<br />
00:00-1:00<br />
2:00-3:00<br />
4:00-5:00<br />
Figure 6. Triaenops rufus and T. furculus nocturnal activity pattern<br />
6.4.2 Seasonal variation<br />
During November 2001, a preliminary <strong>Frontier</strong> study of the species that inhabited the<br />
cave system had i<strong>de</strong>ntified a population of Hipposi<strong>de</strong>ros commersoni. A number of<br />
individuals were captured and released at that time. However, no individuals of this<br />
species were observed or collected during the mark-recapture study in August 2002.<br />
It may be that the population no longer roosts within the cave, or it may be that the<br />
species shows a reduction in activity during the dry season. During discussions with<br />
villagers and local gui<strong>de</strong>s within the area, many stated that the ‘large bat’ hibernates<br />
<strong>de</strong>ep within the cave during the dry season. H. commersoni is one of the largest<br />
insectivorous bats in Madagascar, with a weight of 40-80g (Garbutt, 1999). This large<br />
body weight may enable the species to retain the necessary reserves to remain inactive<br />
during the dry season. Alternatively, it may be that the cave is only used by H.<br />
commersoni as a communal nursery colony during the breeding season.<br />
31
6.4.3 Breeding status<br />
Many of the captured female T. furculus had swollen genitalia, indicating that the<br />
species was in the breeding season. During a previous survey, in November 2001,<br />
many of the captured females of both Triaenops species were pregnant. Most of the<br />
other mammal species that occur in the surveyed areas also had a breeding season<br />
towards the end of the dry season. This would mean that the births would coinci<strong>de</strong><br />
with the onset of the wet season, which would ensure the provision of food for the<br />
offspring.<br />
6.4.4 Population <strong>de</strong>nsity and conservation status<br />
<strong>The</strong> survey showed exceptionally high <strong>de</strong>nsities of Triaenops rufus and T. furculus<br />
within the cave system. Previous studies of these species have i<strong>de</strong>ntified populations<br />
roosting within caves or equivalent sites, often in mixed-species colonies with<br />
Miniopterus, Myotis and Otomops species. <strong>The</strong>se populations typically contain several<br />
hundred individuals (Garbutt, 1999).<br />
T. furculus is classified as vulnerable, with a <strong>de</strong>creasing population. T. rufus is data<br />
<strong>de</strong>ficient (IUCN, 2002). Consi<strong>de</strong>ring the given status of these species, this survey has<br />
i<strong>de</strong>ntified important populations of both Triaenops species. <strong>The</strong> presence of<br />
approximately 9,000 individuals of a vulnerable mammal species within a single site<br />
warrants specific conservation attention. In conjunction with the exceptional numbers<br />
of T. rufus and the suspected presence of Hipposi<strong>de</strong>ros commersoni, the site should be<br />
classified as one of the most important known microchiroptera roost sites in southwest<br />
Madagascar.<br />
7. Diurnal lemur population study<br />
7.1 Introduction<br />
Preliminary biodiversity studies within the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong><br />
i<strong>de</strong>ntified the presence of two species of diurnal lemur; Verreaux’s sifaka, Propithecus<br />
verreauxi verreauxi, and the ring-tailed lemur, Lemur catta. Both species are classified<br />
as vulnerable (IUCN, 2002). While there have been extensive studies of both species<br />
across their ranges in west and south-west Madagascar, no previous studies had been<br />
un<strong>de</strong>rtaken in the <strong>proposed</strong> park area. Whilst lemurs are not necessarily reliable<br />
indicators of forest biodiversity (Ganzhorn 1999), their en<strong>de</strong>mism and status as<br />
Madagascar’s only primate species makes them an extremely important family to<br />
monitor both for biodiversity and conservation management. In addition, they are<br />
charismatic flagship species of importance for tourism, raising public awareness and<br />
providing support to fund-raising.<br />
A population distribution and <strong>de</strong>nsity study was un<strong>de</strong>rtaken in or<strong>de</strong>r to ascertain the<br />
range and approximate number of individuals of both species of diurnal lemur within<br />
the surveyed areas. <strong>The</strong> study aimed to contribute both to conservation and tourism<br />
<strong>de</strong>velopment within the region, and to provi<strong>de</strong> a basis for future monitoring.<br />
32
7.2 Methods<br />
Between May and August 2002, <strong>de</strong>nsities of diurnal species were calculated and<br />
compared in surveys of three separate areas within the <strong>proposed</strong> regional park. Due to<br />
the topographical and logistical constraints of surveying steep spiny forest, the survey<br />
habitats were stratified and lemur <strong>de</strong>nsities were calculated for gallery forest alone.<br />
Given the <strong>de</strong>pen<strong>de</strong>nce of both diurnal lemur species on gallery forest during the dry<br />
season and taking into account seasonal changes in home ranges, a study of lemur<br />
<strong>de</strong>nsity within gallery forest during the dry season should still reflect the general<br />
population distributions and group dynamics within the area.<br />
Line-transect methods outlined by Whitesi<strong>de</strong>s et al. (1988), Chapman et al. (1988)<br />
and Rabinowitz (1997) were used to establish a survey <strong>de</strong>sign. Twenty 500-metre<br />
transects were walked twice daily for a minimum of five days each. All sightings of<br />
Verreaux’s sifaka and ring-tailed lemur groups were recor<strong>de</strong>d. Date, time, bearing of<br />
the transect, bearing of the spotted group or individuals, distance to the group, lemur<br />
species, group height and numbers of individuals were all noted down for each<br />
sighting.<br />
Three areas within the <strong>proposed</strong> park area were surveyed. <strong>The</strong> first area, around<br />
Antafoky Lake, comprised a fairly substantial band of gallery forest. <strong>The</strong> second<br />
survey area investigated the thin strip of gallery forest along the north bank of the<br />
Onilahy River and the forested areas within small tributary valleys. This area formed<br />
a narrow, fragmented corridor between the first area and the third area in the Sept<br />
Lacs region. <strong>The</strong> third area surveyed the sections of gallery forest located along the<br />
A<br />
B<br />
Antafoky Lake<br />
C<br />
D<br />
A<br />
B<br />
Antafoky<br />
track road<br />
source<br />
agriculture A<br />
B<br />
Andranomite<br />
A<br />
(sacred spring)<br />
A<br />
MAHALEOTSE<br />
Bevoalavo<br />
D<br />
ONILAHY RIVER<br />
C<br />
Camp 1<br />
Man<strong>de</strong>ranoC D B<br />
sand<br />
C<br />
Camp 2<br />
B<br />
D<br />
SEPT<br />
Ifanato<br />
Ifanato River and lake system, within the Sept Lacs region (see figure 7).<br />
Figure 7. Positions of the twenty lemur transects. <strong>The</strong> red and green transects were within the Antafoky<br />
Lake area. <strong>The</strong> Purple transects were within the Sept Lacs area and the blue and orange transects were<br />
located within the forest corridor.<br />
33
Line transect methods have been <strong>de</strong>veloped for estimating the <strong>de</strong>nsity of primates in<br />
continuous forest and have not been adapted for fragmented habitats such as those found<br />
within the survey area. However, the survey team spent long periods of time in each<br />
area, allowing for repeated observations of the lemur groups. <strong>The</strong>refore, total counts of<br />
both species were also estimated from these direct observations.<br />
7.3 Results<br />
<strong>The</strong> survey revealed populations of Propithecus verreauxi verreauxi and Lemur catta<br />
in all of the three surveyed areas (see table 5).<br />
7.3.1 Line transects<br />
Assessments of lemur population numbers using line transects i<strong>de</strong>ntified high<br />
numbers of ring-tailed lemurs around Antafoky Lake and along the corridor of forest<br />
between the lake and the Sept Lacs region, with a comparably lower number within<br />
the Sept Lacs region. <strong>The</strong> <strong>de</strong>nsity of ring-tailed lemurs appeared to be highest around<br />
Antafoky Lake.<br />
A comparable pattern was observed for Verreaux’s sifaka, with high numbers around<br />
Antafoky Lake and along the forest corridor, and comparatively lower numbers within<br />
the Sept Lacs region. <strong>The</strong> <strong>de</strong>nsity of sifaka appeared to be consi<strong>de</strong>rably greater around<br />
Antafoky Lake than at the other two sites.<br />
7.3.2 Direct observation<br />
Assessments of ring-tailed lemur populations by direct observation i<strong>de</strong>ntified high<br />
population numbers around Antafoky Lake and along the forest corridor, with lower<br />
numbers within the Sept Lacs region. <strong>The</strong> <strong>de</strong>nsity of ring-tailed lemurs appeared to be<br />
slightly higher around Antafoky Lake.<br />
A similar pattern was observed with Verreaux’s sifaka, with high numbers observed<br />
around Antafoky Lake. <strong>The</strong> <strong>de</strong>nsity of sifaka appeared to be greater around Antafoky<br />
Lake than at the other two sites.<br />
Table 5. Minimum numbers and <strong>de</strong>nsity of diurnal lemurs at three locations within the survey area,<br />
using two different methods.<br />
Lemur catta<br />
Area of<br />
gallery forest<br />
Min. No. by<br />
direct count<br />
Density (No. / ha.)<br />
by direct count<br />
Min. No. by line<br />
transect<br />
Density (No. / ha.)<br />
by line transect<br />
Antafoky Lake 140 ha. 138 1 299 2.1<br />
Sept Lacs 110 ha. 75 0.7 158 1.4<br />
Corridor 220 ha. 150 0.7 260 1.2<br />
P. v. verreauxi<br />
Antafoky Lake 140 ha. 42 0.3 110 0.8<br />
Sept Lacs 110 ha. 25 0.2 17 0.2<br />
Corridor 220 ha. 25 0.1 52 0.2<br />
34
7.4 Discussion<br />
7.4.1 Comparison of methods<br />
<strong>The</strong> two lemur <strong>de</strong>nsity assessment methods produced very different results. <strong>The</strong> direct<br />
observation method obtained results through repeated sightings and countings, and<br />
gave a minimum figure for the number of individuals present within the surveyed<br />
areas. However, it is probable that not all groups and individuals were observed and<br />
counted. Conversely, the line transect method produced much higher figures for lemur<br />
numbers and <strong>de</strong>nsities than the direct observations, in some cases the estimates were<br />
over twice as high.<br />
Based on the relative accuracy of repeated counts based on direct observations within<br />
fairly small areas of forest, the estimates produced by the line transects appear to be<br />
too high. <strong>The</strong> areas of gallery forest surveyed were fairly narrow and a large field<br />
team spent a long period of time in each survey area. <strong>The</strong> presence of such large<br />
numbers of diurnal lemurs would have been observed during direct observations.<br />
<strong>The</strong> actual number of individuals for each of the diurnal lemur species is probably<br />
closest to the figures given by direct observation. Previous surveys within similar<br />
areas and habitats have shown that the line transect method is difficult to apply, and<br />
that counts based on direct observations have been found to be more accurate (J.<br />
Durbin, pers. comm., 2002).<br />
7.4.2 Lemur <strong>de</strong>nsity and distribution<br />
<strong>The</strong> areas of gallery forest differ consi<strong>de</strong>rably in size between the three surveyed areas<br />
(see Table 5). This appears to have been reflected in the numbers and <strong>de</strong>nsities of<br />
lemurs that were observed in these habitats. High <strong>de</strong>nsities of both diurnal lemur<br />
species were estimated for the area of forest around Antafoky Lake where, <strong>de</strong>spite<br />
recent forest clearance for crop cultivation, the forest is both large and fairly<br />
undisturbed. With an estimated area of approximately 140 ha, this is the single largest<br />
fragment of gallery forest within the <strong>proposed</strong> park area.<br />
<strong>The</strong> corridor of gallery forest contained fairly high numbers of P. verreauxi verrauxi<br />
and very high numbers of L. catta. However, the <strong>de</strong>nsity of lemurs was low. <strong>The</strong><br />
combined areas of gal1ery forest along this corridor were estimated to be<br />
approximately 220 ha in size, although the quality of the forest varied consi<strong>de</strong>rably.<br />
<strong>The</strong> small tributary valleys contained some areas of relatively undisturbed forest,<br />
whereas the gallery forest that bor<strong>de</strong>red the river was narrow, fragmented and<br />
extremely <strong>de</strong>gra<strong>de</strong>d in some areas. It appears that this fragmentation to the gallery<br />
forest along the corridor, combined with the fairly high levels of <strong>de</strong>gradation in some<br />
areas, may have led to the comparatively low lemur <strong>de</strong>nsities that were observed.<br />
Low numbers of both diurnal lemur species were recor<strong>de</strong>d within the Sept Lacs<br />
region. However, the <strong>de</strong>nsities of both species were comparatively high. <strong>The</strong> forest<br />
was less disturbed than the corridor area but only covered approximately 110 ha in a<br />
narrow band along the Ifanato River. This may have adversely affected lemur <strong>de</strong>nsity.<br />
However, several other factors should be taken into account, such as plant species<br />
composition and <strong>de</strong>tection bias (lemurs could not be heard over the sound of the<br />
35
waterfalls). Also, the Sept Lacs region is visited by a small number of tourists, which<br />
may have affected the behaviour of the diurnal lemurs. <strong>The</strong>refore, further studies may<br />
be nee<strong>de</strong>d to clarify these findings.<br />
7.4.3 Conservation status<br />
<strong>The</strong> ring-tailed lemur and Verreaux’s sifaka are both classified as vulnerable (IUCN,<br />
2002). Both species are distributed throughout southwest Madagascar and are found<br />
within several protected areas. However, both species rely on restricted habitats that<br />
are known to be diminishing rapidly due to clearance for agriculture, fuel wood and<br />
building materials. In addition, both species are hunted for food throughout much of<br />
their range and are collected for the pet tra<strong>de</strong>.<br />
Large populations of both species were recor<strong>de</strong>d within the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong><br />
<strong>de</strong> <strong>Belomotse</strong>. <strong>The</strong> survey data provi<strong>de</strong> a good baseline estimate of diurnal lemur<br />
<strong>de</strong>nsity and can go some way to reflecting the abundance of lemurs in this relatively<br />
unstudied area. However, consi<strong>de</strong>ring the fact that some areas of gallery forest are<br />
un<strong>de</strong>r severe threat within the <strong>proposed</strong> park area, more work needs to be done to<br />
maintain and improve the remaining forested areas. Further conservation efforts<br />
should also focus on the maintenance and improvement of the forest around Antafoky<br />
Lake and the corridor along the Onilahy River. Studies should also focus on the<br />
i<strong>de</strong>ntification of the combination of factors that affect the <strong>de</strong>nsities of both lemur<br />
species throughout the area.<br />
8. Birds<br />
8.1 Introduction<br />
<strong>The</strong> long isolation of Madagascar from mainland Africa has not been such a<br />
restriction to colonisation for early bird species as for the mammals, reptiles and<br />
amphibians. As a consequence, birds show a lesser <strong>de</strong>gree of en<strong>de</strong>mism than these<br />
other taxonomic groups, but one that is still remarkably impressive by global<br />
standards. Madagascar boasts at least four en<strong>de</strong>mic families – the mesites<br />
(Mesitornithidae), vangas (Vangidae), ground-rollers (Brachypteraciidae), and<br />
cuckoo-rollers (Lepstosomatidae), as well as the subfamily of asities (Eurylaimidae).<br />
A total of thirty-seven genera are en<strong>de</strong>mic to Madagascar and the Comoros islands.<br />
Several of these genera are monotypic, including the Madagascar Harrier Hawk and<br />
the Helmet Vanga. At the species level, 59% of breeding birds are en<strong>de</strong>mic.<br />
This significant level of en<strong>de</strong>mism unfortunately correlates with the level of<br />
threatened bird species in Madagascar. <strong>The</strong> island’s extraordinary biogeography<br />
means that habitat loss has a huge impact on bird species that have become<br />
evolutionarily specialised, particularly wetland and rainforest species. Thirty-three of<br />
the thirty-seven en<strong>de</strong>mic genera are only found in forested areas. Several are un<strong>de</strong>r<br />
threat from forest clearance. In total, twenty-seven en<strong>de</strong>mic bird species are currently<br />
threatened (IUCN, 2002), of which five are ‘critically endangered’. Four of these are<br />
wetland species. Six species are ‘endangered’ and the remaining sixteen are<br />
consi<strong>de</strong>red ‘vulnerable’. A further fourteen bird species are classified as near<br />
36
threatened. Several species, including the recently discovered Red-shoul<strong>de</strong>red Vanga<br />
(Calicalicus rufocarpalis), are classified as ‘data <strong>de</strong>ficient’ and further studies are<br />
nee<strong>de</strong>d to ascertain their IUCN status.<br />
Protected areas form a crucial element to bird conservation since nearly all threatened<br />
forest species are found in these areas. However, wetland species receive virtually no<br />
protection and only two of the critically endangered waterbirds have populations or<br />
pairs in protected areas. Permanent wetland areas in the south and west of Madagascar<br />
are scarce. <strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong> contains forest and wetland<br />
habitats that are suitable for a wi<strong>de</strong> variety of bird species. <strong>The</strong> <strong>proposed</strong> park area<br />
also contains some of the last remaining areas of riparian forest along the Onilahy<br />
River. <strong>The</strong> Onilahy River appears to be a biogeographical limit for many bird species,<br />
which makes it an interesting area to study.<br />
<strong>The</strong> primary aim for studying the bird communities in the <strong>proposed</strong> park area was to<br />
create a species inventory. Furthermore, the study i<strong>de</strong>ntified and compared the bird<br />
species that occupied different habitats, to provi<strong>de</strong> baseline information that could<br />
support habitat-specific conservation recommendations.<br />
8.2 Methods<br />
Bird observations were recor<strong>de</strong>d and a list of species was produced, with<br />
accompanying ecological data. All positive i<strong>de</strong>ntifications were confirmed from<br />
captures and sightings, as field-staff were not experienced enough to i<strong>de</strong>ntify species<br />
solely from vocalisations.<br />
Bird observations were mostly opportunistic. All members of the group ma<strong>de</strong> these<br />
inci<strong>de</strong>ntal bird records, although one staff member confirmed the majority of positive<br />
i<strong>de</strong>ntifications. In or<strong>de</strong>r to minimise the margin of error caused by the use of<br />
volunteers when making field observations, a strict protocol was observed to confirm<br />
a positive i<strong>de</strong>ntification. Information was recor<strong>de</strong>d at the time of the sighting. If the<br />
recor<strong>de</strong>d distinguishing features of the bird matched the distinguishing features of one<br />
specific species in the field-gui<strong>de</strong>s, a positive i<strong>de</strong>ntification could be ma<strong>de</strong> and the<br />
species was recor<strong>de</strong>d as a confirmed sighting. If a <strong>de</strong>scription was ma<strong>de</strong> that did not<br />
match any species in the book, the <strong>de</strong>scription was recor<strong>de</strong>d in a logbook and further<br />
observations were ma<strong>de</strong> in the area.<br />
Species were i<strong>de</strong>ntified using the following field gui<strong>de</strong>s: ‘Birds of Madagascar - A<br />
Photographic Gui<strong>de</strong>’ by Pete Morris and Frank Hawkins (1998), and ‘Birds of the<br />
Indian Ocean Islands’ by Ian Sinclair and Oliver Langrand (1998).<br />
See <strong>Frontier</strong>-Madagascar Methodology Training Manual (<strong>Frontier</strong> Madagascar,<br />
2003) for <strong>de</strong>tailed <strong>de</strong>scription of methods.<br />
8.3 Results<br />
Over 300 man-hours of field observations revealed 1,971 bird sightings, the vast<br />
majority of which were confirmed to species level.<br />
37
Seventy-nine species were positively i<strong>de</strong>ntified in the area. See Appendix 7 for the<br />
complete species list. A further eleven birds were not i<strong>de</strong>ntified to species, but were<br />
distinct from the positively i<strong>de</strong>ntified species, see Appendix 8. Classifications follow<br />
those used by Morris & Hawkins (1998) and Sinclair & Langrand (1998).<br />
66% of the positively i<strong>de</strong>ntified birds were found both around Antafoky Lake and<br />
within the Sept Lacs region, most probably because they contain similar habitats. A<br />
total of 26.5% of bird species were found only around Antafoky Lake, although this<br />
may reflect the higher birding effort in this area rather than a habitat-specific<br />
difference in species distributions. Many of the birds that were i<strong>de</strong>ntified from only<br />
one location were positively i<strong>de</strong>ntified on only one or two occasions.<br />
8.4 Discussion<br />
8.4.1 Rare species and en<strong>de</strong>mism<br />
Of the positive i<strong>de</strong>ntifications ma<strong>de</strong>, two were classified as vulnerable (IUCN, 2002);<br />
the Humblot’s Heron (Ar<strong>de</strong>a humbloti) and the Madagascar Pond Heron (Ar<strong>de</strong>ola<br />
idae). One near threatened species was i<strong>de</strong>ntified, the Madagascar Sparrowhawk<br />
(Accipiter madagascariensis).<br />
<strong>The</strong> Peregrine falcon (Falco peregrinus) was observed on several occasions. This<br />
species is noted as uncommon throughout Madagascar (Morris & Hawkins, 1998;<br />
Sinclair & Langrand, 1998). A number of other observed bird species are noted as<br />
uncommon in southwest Madagascar, including the Madagascar wagtail (Motacilla<br />
flaviventris), the Madagascar long-eared owl (Asio madagascariensis) and the<br />
Madagascar white-eye (Zosterops ma<strong>de</strong>raspatana). Many water birds, including the<br />
Black-Crowned Night Heron (Nycticorax nycticorax), Great egret (Egretta alba) and<br />
Hammercop (Scopus umbretta) are also stated as uncommon in the dry south and<br />
western regions.<br />
A total of thirty-three species (41% of all species found) were en<strong>de</strong>mic or en<strong>de</strong>mic<br />
bree<strong>de</strong>rs to Madagascar. A total of twenty-one species (27%) were regionally<br />
en<strong>de</strong>mic.<br />
8.4.2 Ecological variation<br />
Sixty bird species (76%) were observed in gallery forest, many of which are reliant on<br />
the presence of this habitat. For example, the cuckoo-roller (Leptosomus discolor) was<br />
observed in gallery forest and is said to be restricted to this habitat in the south<br />
(Morris & Hawkins, 1998). Five bird species were observed solely in gallery forest, of<br />
which four were observed only in the gallery forest around Lake Antafoky (see<br />
Appendix 7). This may reflect the quality and size of this forested area. Many bird<br />
species were observed in disturbed areas of gallery forest, as well as comparatively<br />
undisturbed forest. It may be that this is indicative of the habitat restrictions imposed<br />
on bird species by the fragmented areas of gallery forest, rather than typical habitat<br />
preferences of these species.<br />
38
Nine species were observed solely in spiny forest throughout the surveyed areas. All<br />
other bird species that were observed in this habitat were also seen in disturbed<br />
habitats.<br />
A total of fifteen species of water bird were observed within the surveyed areas, many<br />
of which are recor<strong>de</strong>d as rare in southwest Madagascar Madagascar (Morris &<br />
Hawkins, 1998; Sinclair & Langrand, 1998). <strong>The</strong> Humblot’s Heron and the<br />
Madagascar Pond Heron, both classified as vulnerable with a <strong>de</strong>creasing population,<br />
were regularly seen at Antafoky Lake and the marshes along the Onilahy River. <strong>The</strong><br />
variety of water birds observed reflects the large areas of wetland that occur in the<br />
area. <strong>The</strong> <strong>proposed</strong> park area contains a variety of different wetland habitats including<br />
lakes, ponds, marshes and stream edges. <strong>The</strong>se habitats are distributed as a series of<br />
core areas linked by the Onilahy River. <strong>The</strong>se are the only permanent wetlands in<br />
approximately 2,500 sq km, providing a unique and vital habitat for water birds in the<br />
region.<br />
8.4.3 Threats to bird conservation<br />
Many bird species are threatened through habitat <strong>de</strong>struction due to forest clearance<br />
for agriculture, charcoal production and other resource uses. <strong>The</strong> gallery forests within<br />
the surveyed areas are un<strong>de</strong>rgoing variable rates of <strong>de</strong>gradation. <strong>The</strong> area of forest<br />
around Antafoky Lake is un<strong>de</strong>rgoing a rapid rate of <strong>de</strong>forestation, which will reduce<br />
the area of viable habitat for many species.<br />
<strong>The</strong> wetlands are un<strong>de</strong>r threat from drainage for rice cultivation. In addition, the reeds<br />
are collected for use as building materials. Two vulnerable bird species, Ar<strong>de</strong>a<br />
humbloti and Ar<strong>de</strong>ola idae, rely on these wetlands. <strong>The</strong> <strong>de</strong>struction of these habitats<br />
could lead to a further <strong>de</strong>cline in the populations of these species.<br />
Many bird species are threatened through hunting exploitation. In particular, the<br />
White-faced Whistling duck and Comb duck are hunted around the wetlands within<br />
the surveyed areas, often by visitors from Toliara. Villagers within the region also<br />
hunt birds and have been seen hunting for buttonquail, ground couas and guinea fowl.<br />
This can cause dramatic reductions in populations and can cause noise disturbance<br />
leading to migration of other species from the sites.<br />
Introduced fauna can adversely affect bird communities within the area. <strong>The</strong> bush-pig,<br />
Potamochoerus larvatus, was present within the surveyed areas. It was particularly<br />
common in the upper reaches of the Ifanato River, within the core Sept Lacs area.<br />
This area contains many marshes. P. larvatus plun<strong>de</strong>rs the nests of ground-laying<br />
birds and disturbs water-birds and their nests, affecting populations within the area.<br />
<strong>The</strong> small indian civet, Viverricula indica, has been seen throughout the surveyed<br />
area. This introduced animal preys on small vertebrates and will take birds and eggs.<br />
<strong>The</strong> introduced rat, Rattus rattus, can have an impact on breeding populations of<br />
birds, <strong>de</strong>stroying eggs of ground-laying and arboreal-nesting bird species.<br />
39
9. Reptiles<br />
9.1 Introduction<br />
<strong>The</strong> Field-Gui<strong>de</strong> to the Reptiles and Amphibians of Madagascar (Glaw & Vences,<br />
1994) lists 290 species of reptiles for Madagascar, although this figure has now risen<br />
to 346 known species (Raselimanana, pers. comm). This figure is likely to increase<br />
further, as un<strong>de</strong>scribed species are frequently being found and many of the genera are<br />
in need of revision. <strong>The</strong> Or<strong>de</strong>r reptilia exhibits extremely high levels of diversity and<br />
en<strong>de</strong>mism in Madagascar. Of the 346 known species, 314 are en<strong>de</strong>mic to Madagascar.<br />
A recent global survey of biodiversity hotspots found just two other regions<br />
(Caribbean and Meso-America) that had a greater diversity of en<strong>de</strong>mic reptile species<br />
(Myers et al., 2000).<br />
Southwest Madagascar has become one of the worlds leading herpetological hotspots,<br />
with an extremely high proportion of en<strong>de</strong>mic species. This is mainly a result of the<br />
region’s long isolation, its variable and often extreme conditions, and the relative lack of<br />
competition from other animal groups (such as mammals). However, due to the<br />
inaccessibility and difficult working conditions, surprisingly little work has been carried<br />
out on the herpifauna of the region. Many species are known from only one or very few<br />
specimens so there is extremely little information on their physical dimensions,<br />
population <strong>de</strong>nsities, ecology, behaviour and distribution. Because of the lack of<br />
information, comparatively few reptile species have IUCN status.<br />
Reptiles are worthy of study not only because so little is known about them in<br />
Madagascar, but also because they can be good indicators of habitat types and levels<br />
of disturbance. During the study, a species list was compiled and ecological,<br />
biogeographical and physiological data were collected.<br />
9.2 Methods<br />
Reptiles were observed and collected using systematic and opportunistic methods at<br />
all selected trapsites. Pitfall traps were set up in representative habitats to target<br />
ground-dwelling and fossorial species. In addition, fossorial species were collected by<br />
opportunistic digging and leaf-litter searches in suitable habitats. Crevice searches<br />
were also un<strong>de</strong>rtaken. This inclu<strong>de</strong>d looking un<strong>de</strong>r bark, fallen logs, rocks, and<br />
looking insi<strong>de</strong> vegetation. Large diurnal snakes and lizards were captured by hand or<br />
with the use of sweepnets. At night, representative habitats were targeted for nightwalks,<br />
using head-torches to locate specimens. In addition, ongoing casual collections<br />
were ma<strong>de</strong> during other survey activities.<br />
Representative specimens were collected and ecological and geographical data<br />
recor<strong>de</strong>d for each capture site. Specimen colouration was recor<strong>de</strong>d in life, prior to<br />
killing the animals with an injection of saturated methane sulphate (MS222) solution.<br />
Once killed, the specimens were fixed in a 5-7 % formalin solution. DNA samples<br />
were taken from all species collected.<br />
40
9.3 Results<br />
A total of fifty-five different reptile species, belonging to thirty-four genera and ten<br />
families, were recor<strong>de</strong>d during the study period. <strong>The</strong>se are listed in Appendix 9 with<br />
associated ecological data. IUCN status is also listed (IUCN, 2002). <strong>The</strong> classification<br />
and en<strong>de</strong>mic status is based on Glaw & Vences (1994) and Raselimanana (pers.<br />
comm., 2002). Appendix 9 also shows the habitat localities for observed and collected<br />
specimens during this survey. However, this may not necessarily reflect the general<br />
habitat distribution of these species.<br />
<strong>The</strong> asymptote for reptile species accumulation curves were reached both for the<br />
Antafoky Lake area and the Sept Lacs region (see figure 8). It is therefore assumed<br />
that the majority of species present were collected using the methods <strong>de</strong>scribed above.<br />
60<br />
Antafoky Lake area<br />
60<br />
Sept Lacs region<br />
50<br />
50<br />
No. of Species<br />
40<br />
30<br />
20<br />
No. of Species<br />
40<br />
30<br />
20<br />
10<br />
10<br />
0<br />
0<br />
0 50 100 150 200 250 300 350 400 450<br />
0 50 100 150 200 250 300 350 400<br />
Effort<br />
Effort<br />
Figure 8. Reptile species accumulation curves for the surveyed areas.<br />
9.4 Discussion<br />
9.4.1 Biogeography and species richness<br />
<strong>The</strong> biodiversity survey produced an extremely comprehensive reptile species list for<br />
the site of the <strong>proposed</strong> <strong>Regional</strong> Park. Prior to this survey, only one short<br />
herpetological study had been un<strong>de</strong>rtaken in this area (Raxworthy, 1995). During that<br />
survey, twenty-nine species were recor<strong>de</strong>d, all of which were also recor<strong>de</strong>d during<br />
this study.<br />
Of the fifty-five species collected, fifty (91%) are en<strong>de</strong>mic, of which nineteen (35%)<br />
are limited to the dry forests of southwest Madagascar, see Appendix 9. <strong>The</strong> nonen<strong>de</strong>mic<br />
species that were collected or observed inclu<strong>de</strong>d the Nile crocodile<br />
(Crocodylus niloticus), two species of freshwater turtle (Pelusios castanoises and<br />
Pelomedusa subrufer), and two gecko species (Hemidactylus mercatorius and H.<br />
mabouia). All of these species are also found on the African mainland. <strong>The</strong> means by<br />
which these non-en<strong>de</strong>mic species arrived on Madagascar is not clear, although it is<br />
probable that man introduced the house gecko, Hemidactylus mabouia.<br />
<strong>The</strong> high level of species richness in the surveyed areas can be attributed to the<br />
diversity of habitats in this area. <strong>The</strong>se habitats inclu<strong>de</strong> riparian and spiny forest,<br />
lakes, temporary marshlands, open sandy grassland, cultivated areas and limestone<br />
41
cliffs. In addition, many areas contain habitats with a negligible level of human<br />
disturbance.<br />
9.4.2 Seasonal variation<br />
<strong>The</strong> region has an extremely variable climate. This provi<strong>de</strong>s seasonal variation of<br />
conditions and habitats, to which many of the reptile species have adapted. Many<br />
reptiles in southwest Madagascar exhibit behavioural adaptations during the dry<br />
season (May-October).<br />
Certain species were dormant during the dry season, but were collected, often in large<br />
numbers, following the onset of the rains. <strong>The</strong>se species inclu<strong>de</strong>d two species of<br />
freshwater turtle, Pelomedusa subrufer and Pelusios castanoi<strong>de</strong>s; a water-<strong>de</strong>pendant<br />
snake, Liopholidophis lateralis; and a burrowing snake, Typhlops <strong>de</strong>corsei. All other<br />
species were collected throughout the year, although the <strong>de</strong>nsity of active specimens<br />
of many of these species varied consi<strong>de</strong>rably during the dry season.<br />
During the months of May – October, the temperature in the region is comparatively<br />
cool, reaching as low as 10ºC during the night. <strong>The</strong>re is little vegetation on many of<br />
the plants, particularly those found in the dry habitats on the hillsi<strong>de</strong>s and plateau.<br />
Also, the number of active invertebrate species is comparatively low. <strong>The</strong>se factors<br />
have a significant effect on the levels of activity of the reptile populations in the<br />
region at this time. Many reptile species appeared to reduce their levels of activity<br />
during the dry season. In particular, nocturnal reptiles became less active. This<br />
reduction in activity was particularly noticeable for nocturnal species from the<br />
families Boidae and Colubridae. Also, chameleons from the genus Furcifer were less<br />
frequently observed during the dry season. During this time, several individuals of<br />
Furcifer verrucosus were observed in a dormant state, buried in shallow soil. This<br />
behaviour has been noted before in this region (C. Raxworthy, pers. comm., 2002). It<br />
appears that many reptile species reduce their activity during the dry season as an<br />
adaptation to one or more factors. <strong>The</strong>se factors inclu<strong>de</strong> a reduction in temperature, a<br />
reduction in available food, and an increase in levels of predation.<br />
Not all reptile species reduced their activity during the dry season. In particular, many<br />
species and specimens from the genera Mabuya and Paroedura were collected at this<br />
time. This did not appear to represent an increase in the number of active individuals,<br />
though, as comparable numbers were also collected during the rainy season. However,<br />
an apparent increase in the levels of activity of some small colubrids was noted. All<br />
but one specimen of the colubrid species Liophidium trilineatum, L. torquatum and<br />
Pseudoxyrhopus quinquilineatus were collected in pitfall traps during the dry season.<br />
<strong>The</strong> increase in numbers of specimens collected at this time may reflect an increase in<br />
leaf-litter activity due to a reduction in available niches during the dry season, or it<br />
may reflect an increase in the level of feeding activity necessary with the associated<br />
reduction in numbers of invertebrates.<br />
9.4.3 Ecological distribution<br />
Reptiles were found in all of the major habitats in the region. <strong>The</strong> majority of species<br />
collected from the surveyed areas were found both in riparian and spiny forest. Out of<br />
a total of 22 species (40%) that were found in both habitats throughout the surveyed<br />
42
areas, all of them were found in the Sept Lacs region and all but one were found in the<br />
Antafoky Lake area, see Table 6. This inclu<strong>de</strong>d species from eight different families.<br />
This shows that many species have adapted to a variety of habitats, and that these<br />
species are comparatively wi<strong>de</strong>spread throughout the surveyed area.<br />
Table 6. Ecological distribution of reptile species in the <strong>proposed</strong> <strong>Regional</strong> Park<br />
(Percentages in perentheses).<br />
No. of Species<br />
Location<br />
Riparian<br />
forest<br />
Riparian and<br />
spiny forest<br />
Spiny forest<br />
Water<br />
Antafoky<br />
Lake area<br />
Sept Lacs<br />
region<br />
17 (35%) 21 (44%) 8 (17%) 2 (4%)<br />
18 (37%) 22 (46%) 6 (13%) 2 (4%)<br />
Both 21 (38%) 22 (40%) 9 (16%) 3 (6%)<br />
A total of twenty-one reptile species (38%) were found only in riparian forest in the<br />
surveyed areas. Of these, fourteen were snake species, many of which typically<br />
<strong>de</strong>pend on trees (e.g. Ithycyphus oursi, Langaha madagascariensis, and the<br />
Madagascar tree boa, Sanzinia madagascariensis), or live in the leaf-litter (e.g.<br />
Liophidium trilineatum, L. torquatum and Pseudoxyrhopus quinquilineatus). Other<br />
species of note inclu<strong>de</strong>d Paroedura vahiny, Androngo trivittatus and Mabuya dumasi,<br />
all of which are regionally en<strong>de</strong>mic. <strong>The</strong> pygmy chameleon, Brookesia brygooi, was<br />
found only in riparian forest, as were Geckolepis maculata and G. petiti. <strong>The</strong>se<br />
species are <strong>de</strong>pendant on riparian forest and rainforest throughout their range (Glaw &<br />
Vences, 1994).<br />
A total of 9 reptile species (16%) were found only in spiny forest habitats. All of these<br />
species have adapted to the dry environment, and inclu<strong>de</strong> the radiated tortoise,<br />
Geochelone radiata, and Standingi’s day gecko, Phelsuma standingi, both of which<br />
are classified as vulnerable (IUCN, 2002). Other species inclu<strong>de</strong> the burrowing snake,<br />
Typhlops <strong>de</strong>corsei, of which very little is known, and the iguanas Chalarodon<br />
madagascariensis and Oplurus cyclurus. Also collected in spiny forest were the skink<br />
Mabuya gravenhorstii, the geckos Paroedura picta and Blaesodactylus sakalava, and<br />
the gerrhosaurid Tracheloptychus madagascariensis. All of these species are<br />
<strong>de</strong>pendant on dry forests throughout their range (Glaw & Vences, 1994).<br />
A total of 3 aquatic reptile species were observed or collected. <strong>The</strong> Nile crocodile,<br />
Crocodylus niloticus, was observed in Antafoky Lake and in the Onilahy River<br />
around Man<strong>de</strong>rano. <strong>The</strong> freshwater turtle Pelusios castanoi<strong>de</strong>s was collected at<br />
Antafoky Lake, where it was recor<strong>de</strong>d in large numbers during the wet season. This<br />
species is found in ponds and lakes throughout Madagascar. <strong>The</strong> turtle Pelomedusa<br />
43
subrufer was collected from floo<strong>de</strong>d marshes and temporary water bodies in the Sept<br />
Lacs region. This species is recor<strong>de</strong>d as inhabiting temporary areas of water<br />
throughout Madagascar.<br />
9.4.4 Rare species<br />
<strong>The</strong> radiated tortoise, Geochelone radiata, was observed in areas of spiny forest on<br />
the hillsi<strong>de</strong>s around Lake Antafoky. It was not observed in the Sept Lacs Region,<br />
although extensive searches were carried out in suitable habitats. This en<strong>de</strong>mic<br />
species is classified as Vulnerable (IUCN, 2002) and is promoted as a flagship species<br />
throughout south and southwest Madagascar. Very few specimens were seen, and the<br />
population seems to consist of isolated, low <strong>de</strong>nsity populations.<br />
Standingi’s day gecko, Phelsuma standingi, was observed and collected from the<br />
spiny forest on the hilltops around Lake Antafoky. It was not observed in the Sept<br />
Lacs Region. This en<strong>de</strong>mic species is classified as Vulnerable by IUCN (2002). Only<br />
a single specimen was collected, as the species is apparently very uncommon in the<br />
surveyed areas.<br />
<strong>The</strong> en<strong>de</strong>mic skink, Mabuya dumasi, has a localised distribution. It has only been<br />
collected from areas of riparian forest along the Onilahy River. Specimens were<br />
collected from the forests around Lake Antafoky and from the Sept Lacs region. In<br />
both localities, it was commonly found.<br />
<strong>The</strong> nocturnal snake, Madagascarophis ocellatus, is a regional en<strong>de</strong>mic. It is<br />
restricted to the dry forests of southwest Madagascar. Specimens were observed and<br />
collected from the forests around Lake Antafoky and from the Sept Lacs region. In<br />
both localities, it was commonly observed.<br />
<strong>The</strong> nocturnal, en<strong>de</strong>mic snake, Lycodryas gaimardi, is rarely collected in the Toliara<br />
Region of southwest Madagascar. One specimen was collected in riparian forest<br />
around Lake Antafoky, and a further two specimens were collected from the Sept<br />
Lacs region.<br />
<strong>The</strong> en<strong>de</strong>mic burrowing snake, Typhlops <strong>de</strong>corsei, is rarely collected. It is limited to<br />
the dry regions of southwest Madagascar. Only few specimens have been collected,<br />
and the ecology of the species is poorly un<strong>de</strong>rstood. Specimens were collected from<br />
both surveyed sites, though only from spiny forest areas. During and soon after<br />
rainfall, the species was comparatively easy to collect and appeared to be fairly<br />
common.<br />
A second en<strong>de</strong>mic burrowing snake, Typhlops sp., was collected from both surveyed<br />
sites. This snake is similar in appearance to Typhlops arenarius, but differs in midbody<br />
scale count and pigmentation. Specimens of this potentially un<strong>de</strong>scribed species<br />
were commonly collected in areas of riparian forest around Antafoky Lake and<br />
throughout the Sept Lacs region. It was the most commonly collected colubrid snake<br />
in these areas.<br />
44
9.4.5 Range extensions and locality records<br />
Range extensions were recor<strong>de</strong>d for several species. In addition, new locality records<br />
were ma<strong>de</strong> for many species at the surveyed sites.<br />
<strong>The</strong> en<strong>de</strong>mic snake, Liophidium trilineatum, has only previously been recor<strong>de</strong>d south<br />
of the Onilahy River. Specimens of this species were observed and collected in<br />
riparian forest around Lake Antafoky, on the north bank of the Onilahy River. This<br />
species was not collected from the Sept Lacs region, and therefore appears to have a<br />
localised distribution in the area.<br />
<strong>The</strong> en<strong>de</strong>mic pygmy chameleon, Brookesia brygooi, has previously been recor<strong>de</strong>d as<br />
far south as Sakaraha. A single specimen collected from riparian forest in the Sept<br />
Lacs region represents a range extension of approximately 160km. This cryptic<br />
species is difficult to observe. However, extensive searches only revealed one<br />
specimen, so it can be assumed to be uncommon in the area.<br />
<strong>The</strong> surveys conducted by <strong>Frontier</strong> i<strong>de</strong>ntified twenty-six species that had not been<br />
previously collected in the surveyed regions. <strong>The</strong>se represent new locality records for<br />
the area.<br />
9.4.6 Threats to reptile conservation<br />
<strong>The</strong> most significant threat to reptile species diversity in the surveyed region is habitat<br />
loss. In particular, the riparian forests in the region are rapidly becoming fragmented.<br />
Twenty-one species of reptile were collected only from areas of riparian forest, and<br />
are vulnerable to the loss of this habitat. In particular, species of gecko from the genus<br />
Geckolepis were only collected from areas of undisturbed riparian forest, showing<br />
sensitivity to disturbance. <strong>The</strong> fragmented distribution of several riparian forest<br />
<strong>de</strong>pendant species (eg Liophidium trilineatum and Brookesia brygooi) may indicate<br />
that localised extinctions of forest <strong>de</strong>pen<strong>de</strong>nt species are already occurring.<br />
Nine species of reptile showed <strong>de</strong>pen<strong>de</strong>nce on the dry spiny forests, large areas of<br />
which are being cleared for cultivation. Forested areas are cleared by burning, which<br />
severely impacts the reptile populations, particularly the radiated tortoise, Geochelone<br />
radiata. Also, large trees from the spiny forests are selectively felled for use as<br />
building materials. <strong>The</strong>se trees are the preferred habitat of the rare Standingi’s day<br />
gecko, Phelsuma standingi.<br />
<strong>The</strong> pet tra<strong>de</strong> has affected the populations of many reptiles in Madagascar,<br />
particularly chameleon, tortoise and day gecko species. <strong>The</strong>re is not a recent history of<br />
reptile collection in the surveyed areas, as regional collections have focussed on the<br />
dry forests inland from Ifaty, where Phelsuma standingi have been collected in large<br />
numbers in the past. However, if specimen collection for the pet tra<strong>de</strong> were to occur in<br />
the area, it would severely affect the populations of Geochelone radiata and Phelsuma<br />
standingi, which have been found in such low <strong>de</strong>nsities that their populations already<br />
appear to be critically threatened.<br />
<strong>The</strong> area supports populations of two boa species, Sanzinia madagascariensis and<br />
Boa dumerili. <strong>The</strong>se species are both collected for the pet tra<strong>de</strong>. <strong>The</strong>se species are also<br />
45
collected for their skins, which are sold in tourist markets throughout the country.<br />
<strong>The</strong>se species were common throughout the surveyed areas, but both species have<br />
slow reproductive rates and the populations of these snakes could be threatened if<br />
collecting was to occur.<br />
<strong>The</strong> Nile crocodile, Crocodylus niloticus, is also present in the surveyed areas, and is<br />
un<strong>de</strong>r threat throughout Madagascar due to hunting. Crocodiles are killed for their<br />
skins and for food. However, the crocodiles in Lake Antafoky are locally consi<strong>de</strong>red<br />
to be fady and are not hunted.<br />
Local communities around Antafoky Lake have been observed trapping and collecting<br />
the freshwater turtle, Pelusios castanoi<strong>de</strong>s, for food. Also, shell fragments of the<br />
radiated tortoise, Geochelone radiata, have been observed around the village of<br />
Antafoky, including the shell of an individual that had been marked and released by<br />
<strong>Frontier</strong> earlier in the year. This specimen had been released over two kilometres from<br />
the village, showing that tortoise collection occurs over a fairly large area. Only five<br />
radiated tortoises were found in the surveyed areas over a period of fifteen months, so<br />
it appears that this species is in danger of becoming locally extinct.<br />
10. Amphibians<br />
10.1 Introduction<br />
<strong>The</strong> Field-Gui<strong>de</strong> to the Reptiles and Amphibians of Madagascar (Glaw & Vences,<br />
1994) lists 187 <strong>de</strong>scribed species of amphibian for Madagascar, although as with the<br />
reptiles, the actual figure is proving to be much higher. At least 230 different species<br />
have been i<strong>de</strong>ntified, and the status of more than 45 additional forms remain to be<br />
clarified (Raselimanana, pers. comms). <strong>The</strong> amphibians show high levels of en<strong>de</strong>mism,<br />
with over 99% of the species en<strong>de</strong>mic to Madagascar and its offshore islands. This<br />
represents one of the highest percentages of amphibian species en<strong>de</strong>mism in the world.<br />
Only a few other countries, such as Brazil, Columbia and Mexico, have more en<strong>de</strong>mic<br />
amphibian species (Raselimanana, pers. Comm., 2002).<br />
Southwest Madagascar contains a low number of amphibian species, compared to other<br />
regions of Madagascar. This is mainly due to the arid climate and short rainy season.<br />
However, several of those species present are en<strong>de</strong>mic to the region and show high<br />
levels of adaptation. For example, the genus Scaphiophryne has adapted well to the dry<br />
habitats of south and southwest Madagascar and several species can be found.<br />
<strong>The</strong> amphibian fauna of the <strong>proposed</strong> site of the <strong>Regional</strong> Park was studied as part of<br />
the <strong>Frontier</strong>-Madagascar biodiversity survey. A species list was compiled, and<br />
ecological, biogeographical and physiological data were collected. As with the<br />
reptiles, amphibians can be good indicators of general habitat types and levels of<br />
disturbance.<br />
46
10.2 Methods<br />
Amphibians were observed and collected using both systematic and opportunistic<br />
methods at all selected trapsites. Pitfall traps were set up in representative habitats to<br />
target ground-dwelling species. After rain, typical amphibian habitats were targeted<br />
for sampling. In addition, ongoing casual collections were ma<strong>de</strong> during other survey<br />
activities.<br />
Specimens were collected whenever possible, and ecological and geographical data<br />
recor<strong>de</strong>d for each capture site as outlined in the Methodology Training Manual<br />
(<strong>Frontier</strong> Madagascar, 2003). Specimen colouration was recor<strong>de</strong>d in life, prior to<br />
killing the animals in methane sulphate (MS222) solution. Once killed, the specimens<br />
were fixed in a 5 % formalin solution. DNA samples were taken from all species<br />
collected.<br />
10.3 Results<br />
A total of 6 different amphibian species, belonging to 5 genera and 3 families, were<br />
recor<strong>de</strong>d during the study period. <strong>The</strong>se are listed in Table 7 with associated<br />
ecological data. <strong>The</strong> classification and en<strong>de</strong>mic status is based on Glaw & Vences<br />
(1994), Glaw & Vences (2002), and Raselimanana (pers. Comm., 2002.). Table 7 also<br />
shows the habitat localities for observed and collected specimens during this survey,<br />
although this may not necessarily reflect the general habitat distribution of these<br />
species.<br />
Many specimens were captured in pitfall traps. However, nearly all Boophis doulioti<br />
and all Mantidactylus curtus specimens were recor<strong>de</strong>d from opportunistic collections.<br />
Table 7. List of amphibian species found in the Sept Lacs region.<br />
No. SPECIES<br />
LOCATION HABITAT<br />
ENDEMIC<br />
Antafoky Sept Lacs Riparian Spiny<br />
Lake Region<br />
STATUS<br />
forest scrub<br />
Family: Ranidae<br />
1 Ptycha<strong>de</strong>na mascareniensis N<br />
Family: Mantellidae<br />
2 Mantidactylus curtus E<br />
3 Boophis doulioti RE<br />
4 Laliostoma labrosum E<br />
Family: Microhylidae<br />
5 Scaphiophryne brevis RE<br />
6 Scaphiophryne calcarata RE<br />
Total 4 6<br />
KEY TO ABBREVIATIONS.<br />
En<strong>de</strong>mic status:<br />
• RE - <strong>Regional</strong> En<strong>de</strong>mic: En<strong>de</strong>mic species with limited distribution, occurring only in southwest Madagascar.<br />
• E - En<strong>de</strong>mic: Occurring only in Madagascar.<br />
• N – Non-en<strong>de</strong>mic.<br />
47
10.4 Discussion<br />
10.4.1 Biogeography and species richness<br />
<strong>The</strong> most significant achievement of the amphibian survey was the production of a<br />
comprehensive species list for the <strong>proposed</strong> site of the <strong>Regional</strong> Park. Prior to this<br />
survey, the only inventory of this region was produced following a short study by<br />
Raxworthy (1995). A total of three species were recor<strong>de</strong>d, all of which were also<br />
recor<strong>de</strong>d in this study.<br />
Five en<strong>de</strong>mic species were recor<strong>de</strong>d, three of which are en<strong>de</strong>mic to the dry coastal<br />
forests of south and southwest Madagascar (see Table 7). <strong>The</strong> non-en<strong>de</strong>mic species,<br />
Ptycha<strong>de</strong>na mascareniensis, originated from mainland Africa and has spread<br />
throughout Madagascar. It can be found in almost every large body of fresh water<br />
throughout the country.<br />
<strong>The</strong> low level of species richness in the Sept Lacs region can be attributed to the dry<br />
climate and low number of suitable habitats for amphibians. <strong>The</strong> species collected<br />
during this survey are typical for the dry habitats of southwest Madagascar.<br />
10.4.2 Seasonal variation<br />
<strong>The</strong> seasonal distribution for several of the amphibian species within the survey area<br />
was characterised by a single peak in species abundance during the rainy season. <strong>The</strong><br />
species showing this pattern were Scaphiophryne brevis, S. calcarata and Laliostoma<br />
labrosum. <strong>The</strong>se species burrow into the ground at the end of the rainy season and<br />
remain dormant during the dry season, emerging again after the first rains. <strong>The</strong> other<br />
species collected during the survey showed a high affinity to water, and the<br />
abundance of these species appeared to be fairly constant throughout the year.<br />
10.4.3 Ecological distribution<br />
Amphibians were found in all of the natural habitats within the surveyed area,<br />
although the majority of species were located only in close proximity to permanent<br />
bodies of water. During the dry season, no specimens were collected in area of spiny<br />
forest. However, following the onset of the rains, Laliostoma labrosum,<br />
Scaphiophryne brevis and S. calcarata were collected from areas of spiny forest that<br />
were fairly close to water. <strong>The</strong> other species were collected around permanent water<br />
bodies. <strong>The</strong> most frequently recor<strong>de</strong>d amphibian species was Ptycha<strong>de</strong>na<br />
mascareniensis, which was collected from every area of permanent water within the<br />
surveyed areas, as was Boophis doulioti. However, the distribution of Mantidactylus<br />
curtus was limited to the fast-moving stream within the core area of the Sept Lacs<br />
region.<br />
10.4.4 Threats to amphibian conservation<br />
<strong>The</strong> most significant threat to amphibian species diversity in the location of the<br />
<strong>proposed</strong> <strong>Regional</strong> Park is habitat loss. All species show affinity to water bodies and<br />
48
the associated riparian forest, the <strong>de</strong>terioration or loss of which will threaten their<br />
continued survival.<br />
However, all species are present within the Sept Lacs region, managed by GELOSE.<br />
Also, all species are present within the core area of the Sept Lacs region, which<br />
currently has the highest <strong>de</strong>gree of protective management. In particular,<br />
Mantidactylus curtus is only found within this area. It is hoped the <strong>de</strong>signation of the<br />
area with reserve status will allow the continued existence of this isolated population.<br />
11. Invertebrates<br />
11.1 Butterflies<br />
11.1.1 Introduction<br />
Butterflies are often consi<strong>de</strong>red to be important indicators of overall biodiversity in an<br />
area (Ehrlich, 1988). However, at present little work has been done on the butterflies<br />
of Madagascar. <strong>The</strong>re are no field gui<strong>de</strong>s specific to the country, and species lists are<br />
incomplete. Ecological, biogeographical and seasonal data and lacking for many<br />
species. As yet, the presence or absence of specific species cannot be used as an<br />
indication of habitat quality until further studies have i<strong>de</strong>ntified these links.<br />
<strong>The</strong> first comprehensive list of the butterflies of Madagascar is in the process of<br />
completion, the author of which has provi<strong>de</strong>d the taxonomic i<strong>de</strong>ntifications here. <strong>The</strong><br />
specimens and ecological and seasonal data collected by the <strong>Frontier</strong>-Madagascar<br />
Forest Research Programme will be incorporated into this work.<br />
<strong>The</strong> primary aims of this survey were to produce a species inventory for the Sept Lacs<br />
region, and provi<strong>de</strong> data on distribution with respect to habitat type and season.<br />
11.1.2 Methods<br />
Canopy traps and sweep-nets were used for the collection of butterflies. Canopy traps<br />
were an integral part of the biodiversity trapsites. <strong>The</strong> traps were placed at different<br />
heights and in different habitats (e.g. near the ground in a clearing, or high up in thick<br />
vegetation). <strong>The</strong> bait used in canopy traps was fermented banana. <strong>The</strong> banana bait was<br />
ma<strong>de</strong> in advance and left in a suitable container in the sun for at least one week to allow<br />
it to ferment. Traps were baited in the morning and checked at noon and at dusk.<br />
Sweep-netting focussed on the peak activity time for most butterfly species, around<br />
midday. However, in or<strong>de</strong>r to maximise the chances of collecting all species present,<br />
opportunistic sweep-netting was un<strong>de</strong>rtaken throughout the day from dawn to dusk.<br />
Representative specimens were collected, and ecological and geographical data<br />
recor<strong>de</strong>d for each capture site as outlined in the Methodology Training Manual<br />
(<strong>Frontier</strong> Madagascar, 2003). Specimens were killed either by pinching the thorax or<br />
by placing the butterfly in a container with ethyl acetate vapours. <strong>The</strong> specimens were<br />
49
stored in butterfly envelopes and kept dry. DNA samples were taken from many of the<br />
species collected.<br />
11.1.3 Results<br />
During the survey, a total of 340 specimens were retained for taxonomic purposes.<br />
<strong>The</strong>se represented 61 species from 37 genera and 5 families; Hesperiidae, Lycaenidae,<br />
Nymphalidae, Papilionidae and Pieridae (see Table 8). See Appendix 10 for the full<br />
species list. <strong>The</strong> classifications follow those of Lees (Pers. Comm. 2002).<br />
11.1.4 Discussion<br />
11.1.4.1 Ecological distribution<br />
A <strong>de</strong>finitive field-gui<strong>de</strong> to the butterflies of Madagascar has yet to be produced, and<br />
as such, the current ecological distributions of many of the butterfly species of<br />
Madagascar are unknown. <strong>The</strong>refore, it is not always possible to compare trends in<br />
ecological distribution that have been i<strong>de</strong>ntified during this survey with results from<br />
other regions of Madagascar. What is apparent, however, is that there are many<br />
butterfly species in the Toliara region that appear to have restricted ecological<br />
distributions.<br />
Table 8. Summary of butterfly specimens collected.<br />
Family No. of genera No. of species<br />
Hesperiidae 2 4<br />
Lycaenidae 7 7<br />
Nymphalidae 16 31<br />
Papilionidae 3 5<br />
Pieridae 9 14<br />
Total 37 61<br />
<strong>The</strong> surveyed region contained two major habitats: riparian forest and spiny forest. A<br />
total of 45 species were collected solely from areas of riparian forest. Four species<br />
were collected solely from areas of spiny forest, and 12 species were collected from<br />
both habitats. See Table 9 for a summary of the ecological distribution of butterfly<br />
species.<br />
As a general rule, riparian forest is associated with water, whereas spiny scrub is dry.<br />
This may lead to incorrect assumptions about the number of riparian forest <strong>de</strong>pendant<br />
species, as there are very real grounds for suggesting that many of the species are<br />
associated with the water source, not the riparian forest plants, and that these species<br />
can survive in any moist environment. In<strong>de</strong>ed, it may be that several of the species<br />
categorised as gallery forest <strong>de</strong>pendant in this survey will also be found in areas of<br />
open or cultivated land near to water sources in the region. However, several of the<br />
species collected show distinct patterns of distribution and, based on current<br />
50
knowledge, are linked solely to riparian and rainforest habitats. <strong>The</strong>se species inclu<strong>de</strong><br />
Aterica rabena, Leptosia alcesta, Mylothris phileris and Tagia<strong>de</strong>s insularis. Other<br />
species that were collected solely in riparian forest, for which little data are known to<br />
suggest whether there is actual habitat <strong>de</strong>pen<strong>de</strong>ncy, are Charaxes andara, C. cacuthis,<br />
Apaturopsis paulianii, Libythea tsiandava, Melanitis leda, Neptis kiki<strong>de</strong>li, and<br />
Pseudacraea imerina.<br />
As shown in Table 10, the families Hesperiidae and Papilionidae appear to show a<br />
high <strong>de</strong>gree of habitat <strong>de</strong>pen<strong>de</strong>ncy, with all species from these families being found<br />
only in riparian forest. However, it should be noted that many of the species from<br />
these two families are attracted to damp soil, particularly species from the genera<br />
Papilio and Coelia<strong>de</strong>s. Further study is nee<strong>de</strong>d to i<strong>de</strong>ntify whether these and other<br />
species collected show <strong>de</strong>pen<strong>de</strong>ncy on the gallery forest and associated plant species,<br />
rather than <strong>de</strong>pen<strong>de</strong>ncy on water and sha<strong>de</strong>.<br />
Table 9. Ecological distribution of butterfly species in the <strong>proposed</strong> <strong>Regional</strong><br />
Park.<br />
Habitat<br />
Family Riparian forest Spiny forest Both<br />
Hesperiidae 4 0 0<br />
Lycaenidae 3 1 3<br />
Nymphalidae 23 2 6<br />
Papilionidae 5 0 0<br />
Pieridae 10 1 3<br />
TOTAL 45 4 12<br />
Several species were only collected from the hillsi<strong>de</strong>s and plateau, in spiny forest.<br />
<strong>The</strong>se species appear to show <strong>de</strong>pen<strong>de</strong>ncy on this habitat. However, further studies on<br />
the food-plants of the larvae are necessary before they can be <strong>de</strong>fined as spiny forest<strong>de</strong>pendant.<br />
<strong>The</strong> species found only in spiny forest were Acraea encedon, A. zitja,<br />
Colotis amata and Hemiolaus cobaltina.<br />
A further twelve species were collected both in riparian and spiny forest, many of<br />
which are wi<strong>de</strong>spread throughout Madagascar and mainland Africa. <strong>The</strong>se inclu<strong>de</strong>d<br />
Leptotes pirithous, Eurema floricola, Danaus chrysippus aegyptius and Zizula hylax.<br />
It appears that these species are able to live in a variety of habitats.<br />
It is to be hoped that ecological data collected from this and other studies in<br />
Madagascar will be used to i<strong>de</strong>ntify indicator species for different habitats. This will<br />
be of use for rapid assessments of survey sites, allowing target species to be i<strong>de</strong>ntified<br />
and collected, and assumptions and <strong>de</strong>cisions ma<strong>de</strong> based on the presence or absence<br />
of these species.<br />
11.1.4.2 New species, rare species and range extensions<br />
A total of sixty-one butterfly species were collected and i<strong>de</strong>ntified during this study.<br />
<strong>The</strong>se inclu<strong>de</strong>d 1 un<strong>de</strong>scribed species, 1 possible new sub-species and 7 rare species<br />
51
(D. Lees, pers. comm, 2002.). In addition, data and DNA collected are to be used in<br />
the revision of species complexes.<br />
A specimen was collected at the Antafoky Lake site that was tentatively i<strong>de</strong>ntified as<br />
Strabena goudoti. Further examination of this species by Dr Lees (NHM) showed the<br />
specimen to be an un<strong>de</strong>scribed species. An individual of this species was also<br />
discovered from a site further north at around the same time. <strong>The</strong> specimen has been<br />
labelled as morphospecies 77 at this time, and is awaiting formal i<strong>de</strong>ntification.<br />
Several of the species that were collected are extremely rare, or rarely collected. <strong>The</strong>y<br />
shall be discussed here separately.<br />
Two specimens of the species Acraea obeira were collected from the Antafoky Lake<br />
region, a male and a female. This species is wi<strong>de</strong>spread throughout Madagascar, but<br />
rare throughout its range. Few specimens have been collected, and the Natural History<br />
Museum has only two specimens.<br />
During the dry season, two specimens of the species Apaturopsis paulianii were<br />
collected from the Sept Lacs region, a male and a female, with a DNA sample. Only<br />
one specimen of this species had ever been collected from southwest Madagascar. It is<br />
a very interesting find of an extremely rarely collected butterfly (D. Lees, pers.<br />
comm., 2002).<br />
Three specimens of the species Byblia anvatara anvatata were collected, with DNA<br />
samples. <strong>The</strong> specimens were collected both at the Antafoky Lake site and the Sept<br />
Lacs region. <strong>The</strong> specimens conform to the form seriata collected in ‘Morondawa’ by<br />
Rothschild and Jordan in 1903. Only that initial collection of 10 specimens was<br />
known, until they were collected during this study. This represents a range extension<br />
of nearly 300km, and a rediscovery of the species.<br />
Two specimens of Charaxes cacuthis were collected from riparian forest around<br />
Antafoky Lake, a male and a female. This species is rare in the west of Madagascar,<br />
and may even be a new subspecies (D. Lees, pers. comm., 2002). <strong>The</strong>se specimens,<br />
and the DNA sample collected, will be used to help ascertain subspecies status.<br />
One specimen of the species Neptis kiki<strong>de</strong>li was collected from the Sept Lacs region.<br />
This species is very rarely collected, and nothing was known of its habitat preference.<br />
It was found in riparian forest.<br />
One specimen of the species Pseudacraea imerina was collected from riparian forest<br />
in the Antafoky Lake area. This is a very rare species (D. Lees, pers. comm., 2002). It<br />
is of great interest for DNA studies, so the specimen will be extremely useful.<br />
One specimen of the species Tagia<strong>de</strong>s insularis was collected from the Sept Lacs<br />
region. This also is a very rare species, with few specimens having been collected in<br />
southwest Madagascar.<br />
52
11.2 Malaise Trapping<br />
11.2.1 Introduction<br />
Invertebrate taxonomy is complex. Very few studies have been un<strong>de</strong>rtaken on the<br />
invertebrate fauna of Madagascar, and even fewer surveys collect flying invertebrates.<br />
Many studies that have been un<strong>de</strong>rtaken have focussed mainly on large and easily<br />
captured species, such as butterflies and dragonflies.<br />
Malaise trapping in the <strong>proposed</strong> location of the <strong>Regional</strong> Park was <strong>de</strong>signed to<br />
survey flying invertebrates. Malaise traps typically collect large numbers of<br />
specimens from a wi<strong>de</strong> range of or<strong>de</strong>rs, including Lepidoptera (butterflies and moths)<br />
and Odonata (dragonflies). <strong>The</strong>y also collect many Hymenoptera (particularly wasps,<br />
flying ants and parasitoids of other insect groups). Also, vast numbers of Diptera<br />
(flies) are collected, particularly when the malaise trap is set up close to water, when<br />
large numbers of Nematocera (craneflies, mosquitoes and gnats) are captured. Malaise<br />
trapping also produces large numbers of Coleoptera (beetles).<br />
11.2.2 Methods<br />
Malaise traps are large, tent-like stationary mesh nets that are placed in suitable habitats.<br />
<strong>The</strong> insects meet a vertical barrier un<strong>de</strong>rneath a pitched canopy and instinctively fly<br />
up towards the sky where they are directed towards the collecting bottle. Once in the<br />
bottle they are unlikely to escape and soon fall into the collecting fluid - usually 70-<br />
95% alcohol. <strong>The</strong>se traps were located in natural corridors for flying invertebrates, such<br />
as the edges of <strong>de</strong>nse vegetation, gaps through vegetation, paths, or the edges of streams.<br />
<strong>The</strong> traps were set up according to the methods outlined in the Methodology Training<br />
Manual (<strong>Frontier</strong> Madagascar, 2003). <strong>The</strong>y were placed in representative habitats over<br />
all seasons, in or<strong>de</strong>r to collect a wi<strong>de</strong> variety of invertebrate species. Traps were placed<br />
in suitable habitats for a period of 2-4 days in each location. After that time, specimens<br />
were removed and the trap was taken down.<br />
Specimens were stored in at least 75 % ethanol. All specimens from the same malaise<br />
trapping-location were stored together. Ecological and geographical data were<br />
recor<strong>de</strong>d for each collection, as was the date and duration of the trapping period.<br />
11.2.3 Results<br />
<strong>The</strong> malaise trapping proved a successful way of collecting flying invertebrates. Large<br />
numbers of specimens from the focal or<strong>de</strong>rs were collected. However, while awaiting<br />
the i<strong>de</strong>ntification of the specimens, little can be ascertained regarding the invertebrate<br />
fauna of the region.<br />
53
11.3 Leaf litter Extraction<br />
11.3.1 Introduction<br />
<strong>The</strong> biota of the leaf-litter and soil are a major component of the number of<br />
extinctions caused by tropical <strong>de</strong>forestation (Wilson, 1992). <strong>The</strong> number of leaf-litter<br />
and soil dwelling species that have not been discovered or <strong>de</strong>scribed has been<br />
estimated in the millions, and inclu<strong>de</strong>s insects, spi<strong>de</strong>rs, molluscs and millipe<strong>de</strong>s.<br />
<strong>The</strong>se organisms play a very limited role in conservation <strong>de</strong>cisions due to the lack of<br />
information on the majority of such species (Kremen, 1994).<br />
Leaf-litter extraction surveys in the <strong>proposed</strong> location of the <strong>Regional</strong> Park were<br />
<strong>de</strong>signed to survey ground-dwelling invertebrate Or<strong>de</strong>rs. This would inclu<strong>de</strong><br />
Hymenoptera (ants and solitary bees), Coleoptera (beetles), Diptera (fly larvae),<br />
Lepidoptera (moth and butterfly larvae), Diplopoda (millipe<strong>de</strong>s), Chilopoda<br />
(centipe<strong>de</strong>s), Orthoptera (crickets), and primitive wingless groups such as Collembola<br />
(springtails).<br />
11.3.2 Methods<br />
Leaf-litter extractions sampled the forest floor invertebrate community. Leaf-litter<br />
samples were systematically collected along 100 m transects in representative<br />
microhabitats throughout the survey area. <strong>The</strong> samples were filtered, and the<br />
invertebrates extracted. See Methodology Training Manual (<strong>Frontier</strong> Madagascar, 2003)<br />
or Fisher (1999) for full <strong>de</strong>tails of the extraction techniques.<br />
Specimens were stored in at least 75 % ethanol. All specimens from the same leaflitter<br />
transect were stored together. Ecological and geographical data were recor<strong>de</strong>d<br />
for each collection, as was the date of the trapping period.<br />
11.3.3 Results<br />
<strong>The</strong> leaf-litter extractions proved a successful way of collecting ground-dwelling<br />
invertebrates. Large numbers of specimens from all focal or<strong>de</strong>rs were collected.<br />
While awaiting the i<strong>de</strong>ntification of the specimens, little can be ascertained regarding<br />
the invertebrate fauna of the region. However, ant specimens that were collected have<br />
been i<strong>de</strong>ntified to the genus level, see Table 10. Specimens from nineteen genera of<br />
ants were collected, of which one genus is <strong>de</strong>pendant on the dry forests of western<br />
Madagascar and three genera are <strong>de</strong>scribed as rarely found in southwest Madagascar (B.<br />
Fisher, pers. comm, 2002.). <strong>The</strong>se genera were Aphaenogaster, Cerapachys, Platythyrea<br />
and Pyramica respectively. Work is un<strong>de</strong>rway to i<strong>de</strong>ntify the specimens to species level.<br />
11.3.4 Discussion<br />
Specimens from the genus Aphaenogaster were collected from the Sept Lacs region.<br />
This genus has only been collected from the dry forests of western Madagascar. <strong>The</strong><br />
specimens were collected both in riparian forest, closely associated with water, and on<br />
the hillsi<strong>de</strong>s in dry spiny scrub.<br />
54
<strong>The</strong> genus Cerapachys is rarely found. Species from this genus are extremely<br />
specialised, feeding only on the eggs of other ant species. A single collection of<br />
specimens from this genus was ma<strong>de</strong> from an area of disturbed riparian forest in the<br />
Antafoky Lake area, see table 10.<br />
Table 10. Ecological distribution of ant genera from sampled areas in the <strong>proposed</strong><br />
<strong>Regional</strong> Park.<br />
Antafoky Lake area Sept Lacs region<br />
Disturbed<br />
Riparian<br />
Riparian<br />
Riparian<br />
Forest<br />
Forest<br />
Genus<br />
Forest<br />
Spiny forest<br />
Anochetus 7<br />
Aphaenogaster 3 3<br />
Camponotus 1<br />
Cardiocondyla 1<br />
Cerapachys 1<br />
Hypoponera 2 7 8 15<br />
Leptogenys 1<br />
Monomorium 18 18 19 20<br />
Oligomyrmex 3 6 7 2<br />
Pachycondyla 6 5<br />
Paratrechina 4 9<br />
Pheidole 3 5 18 19<br />
Plagiolepis 2 2 12 14<br />
Platythyrea 1 1<br />
Pyramica 1 6 8 2<br />
Simopone 1<br />
Strumigenys 2 6 1<br />
Tapinoma 1<br />
Technomyrmex 2 2<br />
Tetramorium 13 17 19 19<br />
Tetraponera 1 1 8<br />
TOTAL 11 11 14 15<br />
Specimens from the genus Platythyrea were collected from the Sept Lacs region. Species<br />
from this genus are rare in southwest Madagascar. However, specimens were found both<br />
in riparian forest and on the hillsi<strong>de</strong>s in spiny scrub.<br />
<strong>The</strong> genus Pyramica is also rarely found in southwest Madagascar. Specimens from this<br />
genus were collected from the Antafoky Lake area and the Sept Lacs region. In both<br />
sites, they were found in riparian forest. <strong>The</strong>y were also collected from spiny scrub in the<br />
Sept Lacs region.<br />
From the results shown here, the Sept Lacs area appears to have a greater diversity of ant<br />
genera than the Antafoky Lake area. However, further leaf-litter transects have been<br />
un<strong>de</strong>rtaken in these sites, and the results of these collections must be known before<br />
conclusions can be drawn.<br />
55
11.4 Dragonflies<br />
11.4.1 Introduction<br />
Dragonflies, from the or<strong>de</strong>r Odonata, show extremely high levels of diversity in<br />
Madagascar. However, few studies of dragonflies have previously been un<strong>de</strong>rtaken in<br />
many regions. <strong>The</strong> arid southwest of Madagascar has been very poorly studied. <strong>The</strong>re<br />
are no taxonomic keys or field-gui<strong>de</strong>s to the dragonfly fauna of Madagascar, and<br />
i<strong>de</strong>ntification is extremely difficult.<br />
<strong>The</strong> study aimed to sample the dragonfly communities within the survey areas, and<br />
i<strong>de</strong>ntify the species present.<br />
11.4.2 Methods<br />
Dragonfly specimens were collected with the use of a sweep-net. Collections were ma<strong>de</strong><br />
along watercourses and in areas of wetland. Specimens were collected in the early<br />
morning or late evening, when the dragonflies were flying slowly or resting on<br />
vegetation near the water.<br />
Specimens were killed using the vapours of ethyl acetate. <strong>The</strong>y were stored in<br />
butterfly envelopes and kept dry. Ecological and geographical data were recor<strong>de</strong>d for<br />
each collection location, as <strong>de</strong>scribed in the Methodology Training Manual (<strong>Frontier</strong><br />
Madagascar, 2003).<br />
11.4.3 Results<br />
A total of at least twenty-five species of dragonfly were collected. While awaiting the<br />
i<strong>de</strong>ntification of the specimens, little can be ascertained regarding the dragonfly fauna<br />
of the region.<br />
56
11.5 Molluscs and Millipe<strong>de</strong>s<br />
11.5.1 Introduction<br />
Madagascar has a land snail fauna currently consi<strong>de</strong>red to consist of about 1,200<br />
species, although estimates suggest a figure as high as 4,000 species (Emberton,<br />
2001). It is consi<strong>de</strong>red one of the most interesting snail faunas in the world. Twentythree<br />
families and seventy-one genera are known. <strong>The</strong> dominant families are entirely<br />
different from those found in Africa, and consist of the families Cyclophoridae,<br />
Pomatiasidae, Acavidae and Ariophantidae.<br />
In general, the humid calcareous areas of the north, west and south of Madagascar<br />
have the richest mollusc fauna. Most species are found on or not far from the coast.<br />
Few species survive in the arid regions. Southwest Madagascar is inhospitable to<br />
terrestrial molluscs due to the intense heat and low rainfall. However, those species<br />
that are found in the dry environments of this region show a high <strong>de</strong>gree of adaptation<br />
to this harsh environment. <strong>The</strong> most common adaptation is that of dormancy during<br />
the dry season. Many species bury <strong>de</strong>ep into the soil after the rains have en<strong>de</strong>d, to<br />
emerge following the onset of the next rains. This is especially true of species from<br />
the genus Clavator, which are found in the sandy regions of southwest Madagascar.<br />
Land snails can be useful indicators for conservation assessment, due to their<br />
restricted distributions and ecological fragilities (Emberton, 2001).<br />
<strong>The</strong> millipe<strong>de</strong>s of Madagascar are very poorly studied. Collections are nee<strong>de</strong>d to<br />
clarify the level of diversity and en<strong>de</strong>mism at the regional and national level.<br />
11.5.2 Methods<br />
Mollusc and millipe<strong>de</strong> surveys were un<strong>de</strong>rtaken prior to and during the onset of the<br />
rains. Molluscs were sampled in three 1m x 1m quadrats per trapsite, over eight<br />
trapsites. Millipe<strong>de</strong>s were sampled in three 3m x 3m quadrats per trapsite, also over<br />
eight trapsites. Each quadrat was established near a pitfall trap-line. <strong>The</strong>y were<br />
located in or<strong>de</strong>r to encompass a range of microhabitats.<br />
Two man-hours were spent searching for molluscs in the leaf litter and top 15cm of<br />
soil within each quadrat. All molluscs encountered were collected. Four man-hours<br />
were spent searching for millipe<strong>de</strong>s in the leaf litter and top 15cm of soil within each<br />
quadrat. All millipe<strong>de</strong>s encountered were collected. In addition, specimens were<br />
collected opportunistically during other survey activities.<br />
Live molluscs were placed in tobacco-water to kill them and to draw their bodies partway<br />
out of the shell for ease of preservation and i<strong>de</strong>ntification. Molluscs and<br />
millipe<strong>de</strong>s were stored in 70-95 % ethanol.<br />
Ecological and geographical data were recor<strong>de</strong>d for each quadrat, as <strong>de</strong>scribed in the<br />
Methodology Training Manual (<strong>Frontier</strong> Madagascar, 2003). Specimens were sent for<br />
formal i<strong>de</strong>ntification to Dr Brian Fisher of the Californian Aca<strong>de</strong>my of Sciences.<br />
57
11.5.3 Results<br />
A total of at least nine species of mollusc were collected. Field i<strong>de</strong>ntifications showed<br />
the presence of the giant African snail, Acatina fulica. This species has been<br />
introduced from mainland Africa. Also present were uni<strong>de</strong>ntified species from the<br />
genera Clavator and Tropidophora. While awaiting formal i<strong>de</strong>ntification of the<br />
mollusc specimens, little can be ascertained regarding the mollusc fauna of the region,<br />
or their usefulness as indicators of habitats.<br />
A total of at least sixteen species of millipe<strong>de</strong> were collected, including two species of<br />
pill millipe<strong>de</strong>. While awaiting the i<strong>de</strong>ntification of the specimens, little can be<br />
ascertained regarding the millipe<strong>de</strong> fauna of the region.<br />
11.6 Spi<strong>de</strong>rs and Scorpions<br />
11.6.1 Introduction<br />
<strong>The</strong> spi<strong>de</strong>rs and scorpions of Madagascar are extremely diverse, yet comparatively<br />
poorly studied. Little is known about the distribution of these species, many of which<br />
are still un<strong>de</strong>scribed.<br />
<strong>The</strong> study aimed to i<strong>de</strong>ntify spi<strong>de</strong>r and scorpion species present within the <strong>proposed</strong><br />
<strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>, in an attempt to increase the knowledge of spi<strong>de</strong>r and<br />
scorpion diversity within the region.<br />
11.6.2 Methods<br />
Spi<strong>de</strong>r and scorpion specimens were collected through a variety of different methods.<br />
Pitfall trapping provi<strong>de</strong>d the i<strong>de</strong>al framework to produce a comprehensive collection<br />
with minimal effort, and specimens were collected during each trapsite. Crevice searches<br />
also yiel<strong>de</strong>d specimens, as did daytime and night-time walks (the eyes of a spi<strong>de</strong>r reflect<br />
blue or orange in torchlight).<br />
Specimens were placed in at least 75 % ethanol to kill and preserve them, and stored<br />
according to the locality from which they were collected. Ecological and geographical<br />
data were recor<strong>de</strong>d for each collection, as <strong>de</strong>scribed in the Methodology Training<br />
Manual (<strong>Frontier</strong> Madagascar, 2003).<br />
Specimens were sent for formal i<strong>de</strong>ntification to Dr. Brian Fisher of the Californian<br />
Aca<strong>de</strong>my of Sciences.<br />
11.6.3 Results<br />
A total of at least four species of scorpion and thirty-two species of spi<strong>de</strong>r were<br />
collected. While awaiting the i<strong>de</strong>ntification of the specimens, little can be ascertained<br />
regarding the scorpion and spi<strong>de</strong>r fauna of the region.<br />
58
12. Resource use and Socio economics<br />
12.1 Introduction<br />
Madagascar’s economy is based mainly on agriculture, which employs about 80% of<br />
the workforce, provi<strong>de</strong>s 70% of the countries export earnings and generates about one<br />
third of the GDP. Madagascar produces a wi<strong>de</strong> variety of crops both for consumption<br />
and for export, predominantly rice, maize, cassava and sugar cane. Until 1971,<br />
Madagascar produced enough rice to feed itself and have surplus for export. But in<br />
1976, with the return of the presi<strong>de</strong>nt Didier Ratsiraka to power, rice fields were<br />
nationalised. Although this lasted only 8 years, the production of rice dropped<br />
dramatically. Combined with an increase in soil erosion due to <strong>de</strong>forestation, this has<br />
meant that yields and production are so low that Madagascar now imports rice.<br />
Madagascar was once the biggest exporter of vanilla in the world, which<br />
predominantly went to the U.S for ice cream production. However, the world <strong>de</strong>mand<br />
for vanilla has <strong>de</strong>creased. This, combined with the Comoros un<strong>de</strong>rcutting the market,<br />
has left coffee and cotton as Madagascar's main agricultural exports.<br />
Madagascar has sold its large territorial fishing waters to Japan in return for money<br />
and supplies of vehicles. However, prawns are still sufficiently plentiful to provi<strong>de</strong> a<br />
good source of export revenue.<br />
Traditional communities throughout the island use resources from the surrounding<br />
areas. <strong>The</strong> arid south-western areas of Madagascar are the least fertile, with only spiny<br />
scrub growing naturally on the limestone plateau. However, along the few lakes and<br />
rivers in the area water is constant and plentiful, creating a good basis for agriculture.<br />
<strong>The</strong> Onilahy River is wi<strong>de</strong> with broad strips of flat land lined with marshes and<br />
riparian forest. Many of these areas have been cleared or drained, and cultivated by<br />
the local communities. In addition, areas of spiny forest have been burned and cleared<br />
for the cultivation of maize, cassava and peanuts.<br />
<strong>The</strong> forested areas of southwest Madagascar are also un<strong>de</strong>r threat from clearance for<br />
charcoal production and logging. In addition to these high-impact uses of resources,<br />
hunting is also common throughout the region. Species of hunting interest inclu<strong>de</strong><br />
birds, lemurs and tenrecs. Fishing is also a common practise in lakes, ponds and rivers<br />
throughout the region, as well as intensive saltwater fishing. Medicinal plants and<br />
natural food products, such as honey and fruits, are collected.<br />
12.2 Methods<br />
Areas of visible resource use were mapped, as were representative vegetation habitats<br />
within the surveyed areas. In addition, non-visible uses of forest resources were<br />
i<strong>de</strong>ntified through semi-structured interviews with suitable individuals or groups from<br />
local villages. Also, interviews were held to discuss and i<strong>de</strong>ntify the agricultural<br />
practices used within the region<br />
59
12.2.1 Ground-truthing and GIS data layers<br />
Ground-truthing and mapping data were collected for incorporation into geographic<br />
information systems (GIS) as data layers. Specific layers of information were<br />
collected.<br />
Elements of human land use were mapped including village territories, settlements,<br />
and access routes (roads, tracks, paths, and rivers). Wetlands and forested areas were<br />
mapped. This inclu<strong>de</strong>d gallery forest, spiny forest, rivers, lakes, ponds, and marshes.<br />
Areas of high biodiversity were mapped, including home ranges for diurnal lemur<br />
species. Areas of land use were mapped including agricultural land on river flood<br />
plains, areas of forest cleared for cultivation, and areas of forest impacted through<br />
charcoal production and logging.<br />
Surveyed areas were digitised from global positioning systems (GPS), with data<br />
collected using Garmin 12’s.<br />
Ground-truthing of forest edges and specific zones of resource use was conducted on<br />
foot within the core area of the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>, between the<br />
villages of Antafoky and Ifanato. GPS points were collected at approximately 0.05 to<br />
0.2 km intervals. <strong>The</strong> points were plotted onto FTM 1/100,000 Carte <strong>de</strong> Madagascar<br />
regional maps. Forest boundaries and zones of resource use were traced. <strong>The</strong> map was<br />
used to i<strong>de</strong>ntify sites for further work and to provi<strong>de</strong> maps for other survey and<br />
reporting activities.<br />
12.2.2 Biological diversity and habitat maps<br />
Species inventory studies were conducted in different habitats and environments<br />
likely to influence species distributions. <strong>The</strong>se were gallery forests, spiny forests,<br />
wetland areas and disturbed habitats. Altitudinal ranges were consi<strong>de</strong>red too small to<br />
influence species distributions.<br />
Trapsites were established at seventeen different localities within these habitats across<br />
the surveyed areas. <strong>The</strong>se sites were mapped and the faunal data were recor<strong>de</strong>d for the<br />
specific habitat, including mammal, reptile, amphibian, bird and invertebrate data.<br />
This information was recor<strong>de</strong>d for each of the trapsites, which were mapped, which<br />
will allow distribution maps of faunal data to be produced throughout the region.<br />
12.2.3 Human settlements, land, and resource use<br />
Human settlements were mapped, with GPS points taken at the centre of each village<br />
and around the periphery of the village. Surrounding agricultural or cleared land was<br />
also mapped. Information was recor<strong>de</strong>d for each village, including the number of<br />
households, village population, and village status (permanent or temporary).<br />
Socioeconomic and agricultural data were obtained from semi-structured interviews<br />
with village groups and individuals from five villages. Data collected inclu<strong>de</strong>d aspects<br />
of resource use, agricultural practices and restrictions on land use, including locally<br />
managed areas, or areas protected by fady.<br />
60
12.3 Results<br />
12.3.1 Biological diversity and distribution<br />
High levels of habitat specificity were observed from the biological inventory studies.<br />
Species distribution ranges were i<strong>de</strong>ntified for indicator taxa throughout the <strong>proposed</strong><br />
park area in different habitats. Faunal zonation was observed between spiny forest and<br />
gallery forest habitats, with each habitat having a complement of restricted species.<br />
<strong>The</strong> faunal species distribution data were examined to assess and i<strong>de</strong>ntify the habitats<br />
and areas of conservation management importance, if diversity and representivity is to<br />
be maintained within the <strong>proposed</strong> regional park. In particular, species of special<br />
concern were i<strong>de</strong>ntified (locally en<strong>de</strong>mic species or those currently classified as<br />
vulnerable by IUCN) and their distributions were assessed and compared.<br />
One locality (Antafoky Lake) was i<strong>de</strong>ntified that contained all vulnerable mammal,<br />
reptile and butterfly species. This locality also contained the highest diversity of bird<br />
species. A second locality (the Sept Lacs region) was i<strong>de</strong>ntified that contained all the<br />
vulnerable mammal species. In total, three localities within the <strong>proposed</strong> regional park<br />
contained all of the locally en<strong>de</strong>mic, rare or vulnerable mammal, bird, reptile,<br />
amphibian and butterfly species. <strong>The</strong>se localities encompassed all vegetation zones,<br />
geographical features and elevations.<br />
This information is structured to allow incorporation into the GIS map, providing a<br />
data layer for habitat-specific diversity and distribution of key faunal taxa.<br />
12.3.2 Human populations and local management<br />
A total of 5 permanent villages were observed within the core area of the <strong>proposed</strong><br />
regional park, all of which were situated along the Onilahy River. <strong>The</strong> average<br />
number of inhabitants of these villages was approximately 120, with the largest<br />
village, Ifanato, containing approximately 280 inhabitants. <strong>The</strong> total number of<br />
inhabitants for these villages is approximately 800 individuals.<br />
Village territories typically inclu<strong>de</strong>d areas of flat land on the Onilahy flood plain and<br />
within forested valleys. <strong>The</strong> extent of the village territory was based on closeness to<br />
other villages, and traditional and cultural factors such as the age of the village and<br />
the location of fady areas and tombs. Four of the five village territories inclu<strong>de</strong>d<br />
significant areas of agricultural land. <strong>The</strong> fifth village, Antafoky, has suffered from<br />
high levels of erosion by the Onilahy River due to the location of the village on an<br />
outer bend on the river. This has led to the recent loss of large areas of agricultural<br />
land during floods.<br />
<strong>The</strong> villages and associated areas of agricultural land were mapped and this<br />
information recor<strong>de</strong>d to allow incorporation into the GIS map.<br />
Four of the five villages managed the surrounding forests through a programme of<br />
local management, Gestion Local Securisee (GELOSE). <strong>The</strong> fifth village, Antafoky,<br />
was outsi<strong>de</strong> of the <strong>de</strong>signated GELOSE area.<br />
61
12.3.3 Resource use<br />
12.3.3.1 Spiny and gallery forests<br />
Large areas of spiny forest on the plateau had been cleared for the cultivation of maize<br />
and cassava. Access routes were found between villages and the agricultural areas on<br />
the plateau. <strong>The</strong>se are used to transport drinking water to villagers working in the<br />
area, and to transport harvested crops down from the plateau.<br />
Areas of gallery forest had been cleared for cultivation, mostly around Lake<br />
Antafoky. All areas of gallery forest east of the village of Ifanato have been cleared<br />
for at least 12 km along the river, leaving only small patches of trees around the edges<br />
of agricultural land. <strong>The</strong> same is true of the gallery forest between the villages of<br />
Ambohimahavelona and Antafoky. This means that the areas of gallery forest around<br />
Lake Antafoky and along the river between Antafoky and the Sept Lacs region<br />
represent the last remaining significant areas of gallery forest on the north bank of the<br />
Onilahy River.<br />
Many areas of gallery forest had been selectively logged for pirogues, building<br />
material or charcoal production. Cut poles and charcoal pits were frequently observed.<br />
Between the villages of Antafoky and Ifanato, within the GELOSE managed area,<br />
almost all charcoal pits were old. However, recent or unburnt charcoal mounds were<br />
regularly seen in the gallery forest and spiny forest around Antafoky Lake.<br />
Recently cleared areas of forest were often further impacted by grazing from cattle<br />
and goats, which were observed in dry and riparian forest in most of the surveyed<br />
areas.<br />
Natural forest products such as tamarind fruits and tubers were collected to<br />
supplement the local diet and for sale. Plants were collected for medicinal use from<br />
the spiny and gallery forests. However, due to the low population in the area, the<br />
amount collected is low, so impacts are minimal. Honey was collected throughout the<br />
year and was <strong>de</strong>pen<strong>de</strong>nt on gallery forest and spiny forest at different times of the<br />
year. Honey produced in the dry forests is regar<strong>de</strong>d as very good quality and it is<br />
collected for sale in Toliara.<br />
Several species of animals are collected or hunted, either for food consumption locally<br />
or for sale as food or pets. Hunting was observed in all surveyed areas. Specific<br />
targets inclu<strong>de</strong>d birds, tenrecs and freshwater turtles. Young ring-tailed lemurs were<br />
occasionally collected, often by children. <strong>The</strong>se were kept in the villages or taken to<br />
Toliara for sale.<br />
12.3.3.2 Wetlands<br />
All permanent villages within the surveyed areas are located near the Onilahy River.<br />
<strong>The</strong> inhabitants of these villages fish in the river. During the rainy season, prawns are<br />
collected in large numbers from the river.<br />
Wetlands along the Onilahy River are threatened with <strong>de</strong>struction by drainage so that<br />
the land can be used to cultivate rice. However, many permanent marshes and ponds<br />
were found to be in good condition within the GELOSE managed area.<br />
62
Reeds are collected from the shallow marshy areas. <strong>The</strong>y are dried and used for<br />
roofing. However, this activity seems to be sustainable at current levels. Selective<br />
harvesting was observed from the expansive marshes around the village of<br />
Man<strong>de</strong>rano.<br />
Lake Antafoky is <strong>de</strong>ep and has steep-sloping banks around most of the lake,<br />
preventing the cultivation of rice in these areas. <strong>The</strong> villagers from Antafoky catch<br />
several species of fish from the lake, as well as freshwater turtles. Fish are sold to<br />
nearby villages. <strong>The</strong>re are several fady about the lake, often involving the small<br />
population of resi<strong>de</strong>nt Nile crocodiles. This has prevented excessive exploitation of<br />
the resources from the lake.<br />
All areas of gallery forest and wetlands were mapped and this information was<br />
structured in or<strong>de</strong>r to allow incorporation into the GIS map as a habitat data layer.<br />
12.3.4 Agriculture<br />
<strong>The</strong> land along the Onilahy River is fertilised annually by alluvial <strong>de</strong>posits from the<br />
river. Much of this land is used for agricultural purposes. <strong>The</strong> wet season floods<br />
provi<strong>de</strong> these areas with water at a known time. However, water can be a limiting<br />
factor. If the rains are insufficient or excessively plentiful this will have a major<br />
impact on yields. Many different crops are grown in the fertile floodplains, usually<br />
giving a good yield to the local villages.<br />
Table 11. Proportion of different crops grown, and the percentage that is sold.<br />
Crop<br />
Percentage of all Percentage sold<br />
grown<br />
Maize 33 70<br />
Manioc 16 0<br />
White beans 10 75<br />
Sweet potatoes 10 30<br />
Rice 8 0<br />
Butter beans 5 70<br />
Sugar cane 5 80<br />
Tomatoes 4 90<br />
Peanuts 4 55<br />
Kidney beans 3 70<br />
Pumpkin 1 17<br />
Bananas 1 70<br />
Maize is the most important crop grown in the area. It is produced in high quantities<br />
and the surplus is sold. Manioc is also grown in large quantities and is the staple diet<br />
of the villagers. Both of these crops are predominantly grown in cleared areas of spiny<br />
forest on the plateau. Manioc is also grown in dry valleys.<br />
Beans and other vegetables are grown on the Onilahy River flood plain. <strong>The</strong>y are<br />
grown in comparatively small quantities (see Table 11 and Figure 8). A proportion of<br />
the harvested produce is consumed locally and the surplus is sold.<br />
63
Proportional Crop Agriculture<br />
MANIOC<br />
RICE<br />
MAIZE<br />
WHITEBEANS<br />
TOMATOES<br />
SWEET POTATO<br />
BUTTERBEANS<br />
KIDNEY BEANS<br />
SUGAR CANE<br />
PUMPKIN<br />
BANANAS<br />
PEANUTS<br />
Figure 8. Proportions of crops grown by the local villagers within the central Sept Lacs area of the<br />
<strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>.<br />
All areas of agricultural land along the river between the villages of Antafoky and<br />
Ifanato have been mapped. Areas of spiny and gallery forest that have been cleared<br />
have also been mapped. This information was structured in or<strong>de</strong>r to allow<br />
incorporation into the GIS map as a resource use data layer.<br />
12.4 Discussion<br />
<strong>The</strong> natural resources within the <strong>proposed</strong> regional park are used in a variety of<br />
different ways. <strong>The</strong> sustainability of many of these resource use activities needs to be<br />
assessed and the effects monitored. In particular, clearance of gallery forest needs to<br />
be restricted and the remaining areas must be carefully monitored. <strong>The</strong> same is true of<br />
the spiny forest on the plateau, although the loss of gallery forest will have longerterm<br />
effects on the biodiversity of the region.<br />
<strong>The</strong> areas of gallery forest around Antafoky Lake are suffering rapid rates of<br />
clearance for agriculture, due to the lack of agricultural land along the Onilahy River.<br />
This area has exceptional levels of biodiversity, both faunal and floral. <strong>The</strong> area is<br />
outsi<strong>de</strong> of the GELOSE managed zone. With the current rate of <strong>de</strong>forestation, the<br />
gallery forests around Lake Antafoky will rapidly become too small and fragmented<br />
to harbour viable populations of many species. <strong>The</strong> use of forest resources around the<br />
village of Antafoky must therefore be monitored, and alternative, sustainable<br />
activities i<strong>de</strong>ntified.<br />
13. Tourism<br />
13.1 Introduction<br />
<strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong> has great potential for tourism due to its<br />
spectacular scenery, interesting geological features, varied terrestrial and wetland<br />
habitats, abundant wildlife and fascinating local community and culture. <strong>The</strong> site,<br />
which is close to the city of Toliara, has a relatively good infrastructure. Within the<br />
<strong>proposed</strong> regional park is an existing small-scale tourism attraction, the ‘Sept Lacs<br />
64
egion’. <strong>The</strong> <strong>de</strong>velopment of the regional park would focus tourism on a larger area<br />
with more features and involvement from more local communities.<br />
Tourism has the potential to provi<strong>de</strong> a workable alternative to non-sustainable<br />
resource use activities and could contribute to sustainable conservation in the region.<br />
However, in or<strong>de</strong>r for this to work it is vital that the operation is set up in a way that is<br />
going to have minimal negative impacts on the environment and culture. In turn, the<br />
<strong>proposed</strong> tourism operation must directly benefit the community that primarily relies<br />
on the land for survival, and must be sufficiently beneficial to prevent the people<br />
wanting to use the land in a more profitable way.<br />
Despite the presence of unique and fascinating flora, fauna and scenery, the Toliara<br />
region receives comparatively few tourists. Toliara is easily reachable both by road<br />
and by air, and has a wi<strong>de</strong> range of tourist accommodation both in Toliara and in<br />
tourist resorts along the coast. <strong>The</strong> region contains many beautiful and interesting<br />
natural sites that have great potential to become excellent tourist attractions if given<br />
sufficient support, <strong>de</strong>velopment, advertising and management. Increasing the number<br />
of visitors to the area would contribute to the general income for the region and could<br />
benefit the local communities.<br />
13.2 Methods<br />
During a series of field studies, the potential for tourism within the core area of the<br />
<strong>proposed</strong> park region was assessed. This core area encompassed gallery forest, lakes<br />
and rivers between the villages of Antafoky and Ifanato. All existing footpaths were<br />
walked and notes were taken regarding the duration, level of difficulty, scenery and<br />
wildlife. As a result of the systematic bird observations un<strong>de</strong>rtaken during the survey,<br />
information regarding suitable locations for bird hi<strong>de</strong>s was easily extracted. Options<br />
for possible campsites were assessed as the project moved around during the course of<br />
the field studies and established several long-term base camps and short-term satellite<br />
camps. Conducting field studies over a period of approximately 15 months provi<strong>de</strong>d<br />
the opportunity to assess the infrastructure and logistical support available upon which<br />
a tourist operation could be based.<br />
<strong>The</strong> volunteer research assistants working on the project during the field studies<br />
originated from a variety of backgrounds and countries. Many had never travelled<br />
outsi<strong>de</strong> of Europe or the United States, which placed them in an excellent position to<br />
i<strong>de</strong>ntify features and attractions that would appeal to potential tourists, and to i<strong>de</strong>ntify<br />
those features that nee<strong>de</strong>d <strong>de</strong>velopment or would not be suitable. <strong>The</strong>se ‘eco-tourists’<br />
contributed with objective opinions relating to potential tourism issues.<br />
13.3 Results and Discussion<br />
13.3.1 Priority areas for tourism<br />
Within the region of the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>, several areas were<br />
located with potential for tourism. In particular, two areas were i<strong>de</strong>ntified that would<br />
be of interest to the vast majority of visitors. <strong>The</strong>se were Antafoky Lake and the Sept<br />
Lacs region (see Figure 9).<br />
65
<strong>The</strong> Antafoky Lake area contains comparatively large, intact areas of gallery forest.<br />
Many wildlife species that are of interest to tourists were found here – sightings of<br />
Verreaux’s sifaka and ring-tailed lemurs could be virtually guaranteed. Dry spiny<br />
forest and vast sandstone cliffs surround the lake and gallery forest, presenting a<br />
dramatic biotic and abiotic contrast to the lake and surrounding forest. <strong>The</strong> spiny<br />
forest contains many interesting bird and reptile species and geological formations.<br />
<strong>The</strong> lake itself is a picturesque attraction, with opportunities for boat trips to observe<br />
water birds, crocodiles and lemurs.<br />
Camp 1<br />
Antafoky<br />
Ankotrofoty<br />
and Toliara<br />
Antafoky<br />
track road<br />
Befasy<br />
Bevoalavo<br />
MAHALEOTSE<br />
Man<strong>de</strong>rano<br />
Andranomite<br />
(fresh water<br />
spring)<br />
ONILAHY RIVER<br />
Camp 2<br />
Bat Colony<br />
Walk 6<br />
Walk 5<br />
Antsoha<br />
Sept Lacs<br />
region<br />
Ifanato<br />
Figure 9. <strong>The</strong> core area of the park, of major interest to tourists<br />
<strong>The</strong> Sept Lacs region is an existing tourist attraction that is currently visited by a low<br />
but regular number of visitors throughout the year. <strong>The</strong> forest and lakes are of great<br />
aesthetic appeal and swimming is permitted. A visit could easily be combined with<br />
walks or boat-trips to one or more of the large colonies of bats that roost nearby.<br />
<strong>The</strong> track that links Antafoky Lake with the Sept Lacs region is easy to walk,<br />
scenically beautiful, and sha<strong>de</strong>d for the most part. It also presents opportunities to see<br />
local villages, lemurs and water birds en-route.<br />
Given the spatial location of the main attractions of the park, locations for campsites<br />
have been suggested both for Antafoky Lake and the Sept Lacs region, with a possible<br />
improvement of pre-existing amenities en-route to make the walk between sites more<br />
enjoyable for visitors.<br />
13.3.2 Transport<br />
A rough road runs through the <strong>proposed</strong> park from east to west, following the north<br />
bank of the Onilahy River. This road is passable in a 4 x 4 vehicle, quad bike or<br />
motorbike all the way to Sept Lacs, although it is in fairly poor condition in some<br />
places. For those on a smaller budget, there is public transport from the centre of<br />
Toliara to the village of Ankotrofoty, which is approximately 3 hours walk from<br />
Antafoky Lake. Once in Antafoky, the transfer to the campsite could take place by<br />
66
oat or on foot. From the lake, the walk to the Sept Lacs region takes a further 4<br />
hours. Alternatively, it is possible to travel by zebu cart. <strong>The</strong> <strong>de</strong>velopment of “luxury<br />
zebu carts” would be encouraged, with the addition of padding and a sunsha<strong>de</strong>. This<br />
would enhance the travelling experience, while providing traditional transport to<br />
tourists. <strong>The</strong>se zebu chariots could collect tourists from the village of Ankotrofoty and<br />
provi<strong>de</strong> a system of transport throughout the core tourist area.<br />
<strong>The</strong> walk from Antafoky Lake to the Sept Lacs region is fairly easy, with a low level<br />
of physical fitness required. <strong>The</strong> route follows a flat road through gallery forest along<br />
the Onilahy River. <strong>The</strong>re are good opportunities to see wildlife along the route,<br />
particularly ring-tailed lemurs, Verreaux’s sifaka, and a variety of reptiles and water<br />
birds. Spectacular views across the river can be seen along this route. <strong>The</strong>re are<br />
several resting places for walkers, including a natural, spring-fed swimming pool just<br />
before the village of Man<strong>de</strong>rano (approximately 1 hour after Antafoky Lake) and a<br />
freshwater spring at a shrine approximately 1 hour from the Sept Lacs region.<br />
Luggage could be sent on ahead by zebu cart and the walk could be un<strong>de</strong>rtaken<br />
ungui<strong>de</strong>d. <strong>The</strong> campsite suggested in the Sept Lacs region is adjacent to the road,<br />
easing logistical difficulties for arriving tourists.<br />
13.3.3 Campsite location<br />
Providing an environmental experience should overri<strong>de</strong> the need for provision of<br />
luxurious accommodation or sophisticated facilities. Many facilities could be kept<br />
simple, to minimise local disturbance and long term impact.<br />
When consi<strong>de</strong>ring a campsite location, key criteria were consi<strong>de</strong>red. <strong>The</strong>se inclu<strong>de</strong>d:<br />
• Accessibility - the ease of getting to the campsite from the road<br />
• Proximity to biodiversity – the level of biodiversity near the campsite and the ease<br />
of observation<br />
• Surroundings - the beauty of the scenery and immediate surroundings<br />
• Level of disturbance – the amount of work nee<strong>de</strong>d to <strong>de</strong>velop the campsite and the<br />
short and long-term effects of this disturbance on local flora and fauna<br />
Campsite 1 at Antafoky Lake, see Figure 9, is not easily accessible on foot, but<br />
visitors and their luggage can easily be transported from the road to the campsite by<br />
boat. It is possible to reach the <strong>proposed</strong> campsite on foot, although this would<br />
involve wading through water on a number of occasions. Building walkways or<br />
raising existing footpaths could alleviate this. <strong>The</strong> <strong>proposed</strong> campsite is within an area<br />
of gallery forest that is rich in biodiversity. <strong>The</strong> area contains five lemur species and<br />
many birds, and the surroundings provi<strong>de</strong> natural beauty and sha<strong>de</strong>. <strong>The</strong> site is in a<br />
pre-existing clearing in gallery forest, but currently lacks any facilities.<br />
Campsite 2, within the Sept Lacs region, see figure 9, is located near to the road on<br />
the si<strong>de</strong> of a small hill, near to the Ifanato River and associated lakes. This site has<br />
been recently <strong>de</strong>veloped by WWF and has tent space for approximately twenty tents,<br />
with simple shower and toilet areas. <strong>The</strong>re is a small restaurant in the village of<br />
67
Ifanato, 2 km from the campsite. <strong>The</strong> village also contains a small guesthouse<br />
(hotely), providing an alternative for those who do not wish to camp.<br />
13.3.4 Campsite <strong>de</strong>sign<br />
To minimise disturbance to the fragile ecosystems within the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong><br />
<strong>de</strong> <strong>Belomotse</strong>, campsites should be kept small and fairly simple. I<strong>de</strong>ally, the campsites<br />
would contain a series of areas for tents (open or covered), a covered cooking area<br />
and table, washing facilities and a simple toilet. A covered pit latrine or basic toilet<br />
would be suitable. A simple covered shower area would also be nee<strong>de</strong>d. Water could<br />
be pumped in, or collected from the nearby watercourse in buckets.<br />
Tourists could bring bottled drinking water from Toliara, or lake or river water could<br />
be treated prior to drinking. All litter should be collected and taken out of the park<br />
area for disposal, preferably by the tourists or tour gui<strong>de</strong>s.<br />
<strong>The</strong>re would not be a necessity for electrical power. A simple level of lighting, either<br />
from candles or kerosene lamps, could provi<strong>de</strong> illumination for the communal eating<br />
and cooking areas.<br />
13.3.5 Attractions and activities<br />
13.3.5.1 Gui<strong>de</strong>d Walks<br />
A variety of walks and wildlife-watching excursions are possible from both <strong>proposed</strong><br />
campsites, see Figures 9 and 10. In or<strong>de</strong>r to keep the disturbance on the environment<br />
to a minimum, pre-existing paths can be used.<br />
<strong>The</strong> <strong>proposed</strong> walks were i<strong>de</strong>ntified based on several criteria. <strong>The</strong>se inclu<strong>de</strong>d the<br />
potential for viewing wildlife and other points of interest, the level of<br />
ease/accessibility, the length and duration, and the predicted level of disturbance and<br />
environmental impact. See Appendix 11 for a summary of the <strong>proposed</strong> walks.<br />
Viewpoints were i<strong>de</strong>ntified and photographed. See <strong>Frontier</strong>-Madagascar Tourism<br />
Feasibility Study Report (in prep) for further <strong>de</strong>tails of walks, viewpoints and<br />
viewable wildlife.<br />
Due to the difficult terrain, fragile ecosystems and potentially dangerous wildlife<br />
(there are crocodiles in Lake Antafoky), it is strongly recommen<strong>de</strong>d that gui<strong>de</strong>s are<br />
compulsory for tourist groups. Also, the forests contain several areas of cultural<br />
significance, including sacred tombs. It is imperative that the cultural beliefs and<br />
traditions of the local communities are respected, so tourists should avoid these areas.<br />
Gui<strong>de</strong>s would be necessary to i<strong>de</strong>ntify these sacred areas and highlight the<br />
significance of the sites. This would ensure that the villagers’ traditions are respected,<br />
as well as providing cultural information for tourists.<br />
68
Figure 10. <strong>The</strong> approximate position of walks and bird hi<strong>de</strong>s at<br />
Antafoky Lake.<br />
13.3.5.2 Gui<strong>de</strong>s<br />
Two local gui<strong>de</strong>s, Enanetse Alfred and Mre Silvestre, have been i<strong>de</strong>ntified and trained<br />
in simple English and French. In addition, they have had wildlife observation training<br />
and have had basic tourism etiquette explained. Both gui<strong>de</strong>s know the areas of forest<br />
within the Sept Lacs area and the corridor of forest between Antafoky and the Sept<br />
Lacs.<br />
13.3.5.3 Boat Trips<br />
<strong>The</strong>re are several options for boat trips within the <strong>proposed</strong> park area. At Antafoky<br />
Lake, boat-trips could be co-ordinated with several of the walks listed above, or they<br />
could form a separate excursion. Many bird and lemur species can be easily be<br />
observed and approached from the water, providing a different perspective to the<br />
walks. A second option would be a boat-trip from the village of Ifanato, within the<br />
Sept Lacs region, to a large colony of fruit bats within a limestone cave, in an area of<br />
gallery forest on the south bank of the Onilahy. <strong>The</strong> cave can only be approached<br />
from the water and the bats are easily visible. A third option would be to use pirogues<br />
as a return route to Toliara. Tourists could sail down-river in a large pirogue to St.<br />
Augustin, where they could take a taxi-brousse back to Toliara. Alternatively, from St<br />
Augustin tourists could extend their tour by travelling south to the coastal tourist<br />
resort of Anakao or Lac Tsimanampetsotse.<br />
13.3.5.4 Bird Hi<strong>de</strong>s<br />
Initially, the choice of bird hi<strong>de</strong> location was to be quantified by comparing the time<br />
spent at a particular site with the number and variety of bird species observed.<br />
69
However, birding experience suggested that species in forest habitats were most easily<br />
observed on foot, enabling birds to be followed. Also, the high <strong>de</strong>nsity of vegetation<br />
in the hillsi<strong>de</strong> scrub provi<strong>de</strong>d poor visibility for bird watching. <strong>The</strong>refore, all of the<br />
suggested bird hi<strong>de</strong> locations are around Antafoky Lake.<br />
See Appendix 12 for a summary <strong>de</strong>scription of the <strong>proposed</strong> bird hi<strong>de</strong>s.<br />
13.3.6 Amenities within the area<br />
<strong>The</strong>re are a series of pre-existing amenities within the <strong>proposed</strong> park area that could<br />
be adapted slightly to effectively provi<strong>de</strong> tourist support. This would also provi<strong>de</strong> a<br />
mo<strong>de</strong> by which local inhabitants could financially benefit from tourism.<br />
Small shops and stalls are located in the villages of Antafoky, Mahaleotse and Ifanato,<br />
selling basic items such as soap, cigarettes and sweets. At present, these shops are not<br />
well indicated and visiting tourists may well not notice them. A simple sign could<br />
solve this problem. <strong>The</strong> same applies to the basic “hotelys” which exist in Antafoky,<br />
Mahaleotse and Ifanato, selling coffee and basic food. <strong>The</strong>se could provi<strong>de</strong> good<br />
resting points for tourists travelling between Antafoky Lake and the Sept Lacs region,<br />
or during a gui<strong>de</strong>d walk. A simple guesthouse and restaurant already exists in Ifanato.<br />
<strong>The</strong>se pre-existing facilities could be highlighted with park publicity and could<br />
<strong>de</strong>velop as the number of visiting tourists increase.<br />
13.3.7 Potential Problems<br />
One of the few current logistical problems facing the <strong>de</strong>velopment of tourism within<br />
the area is the location of the village toilet in Antafoky. At present, this is besi<strong>de</strong> the<br />
lake, along a path that tourists would use. It is assumed that tourists would be<br />
extremely unhappy about walking through an open toilet, thus making a large<br />
proportion of the lake inaccessible to visitors on foot. A possible relocation of the<br />
toilet could be discussed with the villagers. A second toilet area is located along the<br />
road, away from the <strong>proposed</strong> tourist circuits, so this could be <strong>de</strong>veloped further. This<br />
would also probably enhance the health of the village, as the current toilet location is<br />
polluting the drinking water of the village.<br />
<strong>The</strong> access road passes through a corridor of gallery forest between Lake Antafoky<br />
and the Sept Lacs region. This road is in poor condition and is in need of<br />
improvement. However, the <strong>de</strong>velopment of the road will cause further disturbance to<br />
the gallery forest, which is already becoming <strong>de</strong>gra<strong>de</strong>d due to erosion caused by the<br />
existing road. <strong>The</strong>re is no other location for the road, so improvements need to be<br />
un<strong>de</strong>rtaken with care. Also, improvements to the road could lead to an increase in the<br />
rate of charcoal production and maize cultivation in the area, due to the improved<br />
transportation route. <strong>The</strong>se activities need to be monitored.<br />
13.3.8 Monitoring<br />
In or<strong>de</strong>r for the tourism operation to run in a sustainable manner with minimal<br />
negative environmental and cultural impacts and maximum community benefits,<br />
monitoring must take place. Without a monitoring system, negative impacts caused by<br />
70
the <strong>de</strong>velopment of tourism and associated infratructures may go unnoticed and<br />
potential positive impacts may go unrecognised and un<strong>de</strong>veloped.<br />
With a good monitoring system in place, key aspects of the tourism operation can be<br />
observed. As a result of this system, an assessment could be ma<strong>de</strong> as to whether the<br />
current management was working, and steps could be taken to mitigate any negative<br />
impacts and alter aspects of operation management that are not effective.<br />
See <strong>Frontier</strong>-Madagascar Forest Research Programme Tourism Feasibility Report<br />
(in prep.) for further <strong>de</strong>tails.<br />
14. Conclusions and recommendations<br />
<strong>The</strong> <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong> is an important site for conservation<br />
management, supporting the maintenance of biodiversity and representivity within the<br />
region and the conservation of biodiversity in Madagascar. <strong>The</strong> regional park, if<br />
implemented, would protect approximately 470 ha of gallery forest. This fragile habitat<br />
has been highlighted for conservation priority in the government Environmental Action<br />
Plan for the region. Nationally, large areas of gallery forest are represented in very few<br />
protected areas, specifically Beza Mahafaly Special Reserve and Berenty Private<br />
Reserve. <strong>The</strong> <strong>proposed</strong> regional park would represent one of the largest areas of<br />
protected gallery forest in Madagascar. In, addition, the <strong>proposed</strong> park area contains<br />
significant areas of wetland, a habitat type that is uncommon in the dry forests of<br />
southwest Madagascar.<br />
<strong>The</strong> <strong>proposed</strong> regional park contains many rare, locally en<strong>de</strong>mic or vulnerable species,<br />
including flagship species such as Verreaux’s sifaka, Propithecus verreauxi verreauxi;<br />
the ring-tailed lemur, Lemur catta; and the radiated tortoise, Geochelone radiata.<br />
<strong>The</strong> surveys i<strong>de</strong>ntified four principal direct threats to biodiversity in the core section<br />
of the <strong>proposed</strong> regional park. <strong>The</strong>se were slash-and-burn agriculture, both in the<br />
gallery forests and spiny forests; drainage of wetlands for agricultural purposes; overharvesting<br />
of specific tree species for pirogues, building material or charcoal<br />
production; and over-harvesting of non-timber forest products (animals and plants).<br />
Of these, slash-and-burn agriculture has the most long-term and severe effects,<br />
causing <strong>de</strong>struction of spiny forest that will not regenerate. Large areas of grassland<br />
on the plateau show that the progression of the habitat following spiny forest<br />
clearance is conversion to grassland. In addition, the loss of gallery forest severely<br />
threatens the survival of the many faunal species that are <strong>de</strong>pendant on this habitat.<br />
<strong>The</strong> gallery forests within the <strong>proposed</strong> regional park are the last fragments on the<br />
north bank of the Onilahy River, so many <strong>de</strong>pendant species would not be able to<br />
migrate to other areas if these areas were cleared and fragmented. <strong>The</strong> loss of these<br />
gallery forests would therefore cause the extinction of significant populations of<br />
vulnerable and rare species.<br />
71
<strong>The</strong> forests on the hillsi<strong>de</strong>s and plateau have an important role in watershed protection,<br />
both for local people and the Onilahy River system. Erosion caused by <strong>de</strong>forestation<br />
on the plateau will cause run-off that will have long-term affects on the ecosystem. In<br />
addition, run-off will cause the sedimentation of the reefs that are found near the river<br />
estuary.<br />
Fortunately, there are a comparatively small number of villages and inhabitants within<br />
the core region of the <strong>proposed</strong> regional park. In addition, much of this area is un<strong>de</strong>r<br />
local GELOSE management, which appears to be maintaining a sustainable level of<br />
resource use in the gallery forest and wetland areas. However, even in these areas the<br />
spiny forest on the plateau is rapidly becoming <strong>de</strong>gra<strong>de</strong>d. In addition, the single largest<br />
area of gallery forest, around Antafoky Lake, is not currently un<strong>de</strong>r GELOSE<br />
management. This area contains the highest levels of biodiversity within the entire<br />
<strong>proposed</strong> park area. If this area could be inclu<strong>de</strong>d within the GELOSE management<br />
scheme, the gallery forest could be managed and monitored to prevent or reduce further<br />
fragmentation.<br />
<strong>The</strong> core region of the <strong>proposed</strong> regional park has great potential as a <strong>de</strong>stination for<br />
domestic and foreign tourists. This could create an alternative source of sustainable<br />
revenue for local communities, while promoting environmental awareness within the<br />
regional park. However, if tourism is to help support the conservation of the area,<br />
sensitive management is nee<strong>de</strong>d incorporating the interests and concerns of all the<br />
relevant stakehol<strong>de</strong>rs.<br />
Many areas within the <strong>proposed</strong> regional park have suffered from human disturbance<br />
in the past, although many of the disturbed areas still have the potential to return to a<br />
natural condition. Replanting activities in specific areas could merge fragmented<br />
habitats. In particular, specific disturbed areas of gallery forest around Lake Antafoky<br />
and along the corridor between Antafoky and the Sept Lacs region could be the focus<br />
of replanting activities, allowing the corridor to function effectively and linking<br />
isolated forested areas.<br />
<strong>The</strong>re are certain issues, such as the inclusion of Antafoky within GELOSE<br />
management and the sustainable management of the remaining areas of gallery forest,<br />
which must be addressed rapidly. However, providing that this takes place, the<br />
<strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong>e <strong>de</strong> <strong>Belomotse</strong> has a promising future.<br />
72
References<br />
Billé, R. & Mermet, L. 2002. Integrated coastal management at the regional level:<br />
lessons from Toliary, Madagascar. Ocean & Coastal Management 45, 41-58.<br />
Burney, D.A., James, H.F., Rafamantanantsoa, J., Ramilisonina, Wright, H.T., &<br />
Cowart, J.B. 1997. “Environmental change, Extinction, and Human Activity:<br />
Evi<strong>de</strong>nce from Caves in NW Madagascar”. Journal of Biogeography, Vol. 24.<br />
Chapman, C., Fedigan, L. M. & Fedigan, L. 1988. A Comparison of Transect<br />
Methods of Estimating Population Densities of Costan Rican Primates. Brenesia 30:<br />
67-80.<br />
Cook, J. M., Trevelyan, R., Wall, s. s., Hatcher, M & Rakotondraparany, F. 1991. <strong>The</strong><br />
ecology of Hypogeomys antimena, an en<strong>de</strong>mic Madagascan ro<strong>de</strong>nt. Journal of<br />
Zoology, London 224, 191-200.<br />
Durbin, J. C., & Ralambo, J. A. 1994. <strong>The</strong> Role of Local People in the successful<br />
Maintenance of Protected Areas in Madagascar. Environmental Conservation, vol. 21,<br />
Nr. 2.<br />
Dupuy, D. J. & Moat, J. 1996. A refines classification of the primary vegetation of<br />
Madagascar based on the un<strong>de</strong>rlying geography: using GIS to map its distribution and<br />
to assess its conservation status. In: Biogéographie <strong>de</strong> Madagascar. (ed. Laureço, W.<br />
R.) Paris.<br />
Ehrlich, P.R. 1998. Tomorrow’s world: why saving biodiversity is today’s priority.<br />
World Birdwatch 10 (2): 6-7.<br />
Emberton, K. C., (2001). Exploratory phylogenetic and biogeographic analyses within<br />
three land snail families in southeastern-most Madagascar. Biological Journal of the<br />
Linnean Society, 72: 567-584.<br />
Fergusson, B. 2002. Opportunities for Community Based Conservation in the Androy<br />
Region of Madagascar’s Southern Spiny Forests. Non published paper.<br />
Fisher, B.L., 1999. Improving inventory efficiency: a case study of leaf litter ant<br />
diversity in Madagascar. Ecological Applications, 9: 714-731.<br />
<strong>Frontier</strong> Madagascar 2002. Methodology Training Manual. Edited by Emmet, D.,<br />
Olsson, A. and Bloxam, C. Non published document.<br />
Ganzhorn, J. U., Fietz, J., Rakotovao, E., Schwab, D. & Zinner, D. 1999. Lemurs and<br />
the Regeneration of Dry Deciduous Forest in Madagascar. Conservation Biology, vol.<br />
13, No. 4, pp. 794-804.<br />
Garbutt, N. 1999. Mammals of Madagascar. Pica Press.<br />
73
Glaw, F., & Vences, M. 1994. A Field Gui<strong>de</strong> to the Reptiles and Amphibians of<br />
Madagascar. Pica Press.<br />
Glaw, F., & Vences, M. 2002. Molecular phylogeny of Boophis tephraeomystax: a<br />
case test for east-west vicariance in Malagasy anurans (Amphibia, Anura,<br />
Mantellidae). Spixiana, Vol 25 (1): 79-84.<br />
Goodman, S. M. 1996. Results of a Bat Survey of the Eastern Slopes of the Réserve<br />
Naturelle Intégrale d’Andringitra, Madagascar. In: Goodman, S. M., ed., A floral and<br />
faunal inventory of the eastern slopes of the Réserve Naturelle Intégrale<br />
d’Andringitra, Madagacsar: With reference to elevational variation. Fieldiana:<br />
Zoology, n.s., 85: 1-319.<br />
Goodman, S.M. 1995. Rattus on Madagascar and the Dilemma of Protecting the<br />
En<strong>de</strong>mic Ro<strong>de</strong>nt Fauna. Conservation Biology Volume 9, No. 2, April 1995.<br />
Goodman, S.M., M.D. Carleton & M. Pidgeon, 1999. Ro<strong>de</strong>nts of the Reserve<br />
Naturelle Integrale d’Andohahela, Madagascar. Pp. 217-249, In Goodman, S.M.<br />
(Ed.), A floral and faunal inventory of the Reserve Naturelle Integrale d’Andohahela,<br />
Madagascar: with reference to elevational variation. Fieldiana: Zoology, new series,<br />
94.<br />
Hawkins, F. & Morris, P. 1998. Birds of Madagascar, A Photographic Gui<strong>de</strong>. Pica<br />
Press.<br />
Hutcheon, J. M. 1994. <strong>The</strong> Great Red Island: A Future for its Bats BATS, Vol. 12,<br />
No. 2.<br />
Kremen, C., 1994. Biological inventory using target taxa: a case study of the<br />
butterflies of Madagascar. Ecological Applications, 4(3), pp. 407-422.<br />
Kremen , C., Razafimahatratra, V., Guillery, R. P., Rakotomalala, J., Weiss, A. and<br />
Ratsisompatrarivo, J. 1999. Designing the Masoala National Park in Madagascar<br />
Based on Biological and Socioeconomic Data. Conservation Biology Vol. 13, No. 5.<br />
Kull, Christian A. 1996. “<strong>The</strong> Evolution of Conservation efforts in Madagascar.”<br />
International Environmental Affairs, Vol. 8, No. 1.<br />
IUCN 2002. Red List of Threatened Species. IUCN, Gland, Switzerland.<br />
Janzen, D.H. 1988. In E. O. Wilson & F. M. Peter (eds), Biodiversity, pp. 130-137.<br />
National Aca<strong>de</strong>my Press, Washington, D.C.<br />
Lowry, P.P., Schatz G.E., & P.B. Phillipson. 1997. <strong>The</strong> Classification of Natural and<br />
Anthropogenic Vegetation in Madagascar. pp. 93-123. In: Goodman, S.M. and B.D.<br />
Patterson, eds., Natural Change and Human Impact in Madagascar. Smithsonian<br />
Institution Press, Washington, D.C.<br />
74
Lundahl, S. L. L. & Olsson, A. 2000. Small Mammal Communities in Lowland<br />
Rainforest of Krau Wildlife Reserve, Peninsular Malaysia- with reference to human<br />
disturbance. MsC. <strong>The</strong>sis, unpublished. University of Copenhagen.<br />
Martin, P. (eds.) 1984. Prehistoric Overkill: <strong>The</strong> global Mo<strong>de</strong>l. In Quaternary<br />
Extinctions: A prehistoric Revolution. Tucson: University of Arizona Press.<br />
MacPhee, R. D. E. 1986. Environment, Extinction, and Holocene Vertebrate<br />
Localities in Southern Madagascar. National Geographic Research 2 (4):441-455.<br />
Morris, P. & Hawkins, F. 1998. Birds of Madagascar. A Photographic Gui<strong>de</strong>. Pica<br />
Press.<br />
Myers, N., Mittermeier, R.A., Mittermeier, C.G., <strong>de</strong> Fonseca, G.A.B., Kent, J. 2000.<br />
Biodiversity Hotspots for Conservation Priorities. Nature, 403, 853-858.<br />
Nicoll, M. E. & Langrand, O. 1989. Madagascar: Revue <strong>de</strong> la Conservation et <strong>de</strong>s<br />
Aires Protégées. WWF, Gland, Switzerland.<br />
Rabinowitz, A. 1997. Wildlife Field Research and Conservation Training Manual.<br />
Wildlife Conservation Society<br />
Rasoloarison, R. M., Goodman, S. M. & Ganzhorn, J. U. 2000. Taxonomic Revision<br />
of Mouse Lemurs (Microcebus) in the Western Portions of Madagascar. International<br />
Journal of Primatology, Vol. 21, No. 6.<br />
Raxworthy, C.J., Nussbaum, R. A. 1994. A rainforest survey of Amphibians, Reptiles<br />
and Small mammals at Montagne D’Ambre, Madagascar. Biological Conservation<br />
69, pp 65-73.<br />
Raxworthy, C. J. 1995. Field survey of Amphibians, Reptiles and Small Mammals<br />
from the Toliara Region, Madagascar. Unpublished field Report.<br />
Richard, Alison F., and Shiela O’Connor. 1997. “Degradation, Transformation, and<br />
Conservation: <strong>The</strong> Past, Present, and Possible Future of Madagascar’s Environment.”<br />
In Natural Change and Human Impact in Madagascar, edited by Steven M. Goodman<br />
and Bruce D. Patterson. Washington, D.C.: Smithsonian Institution Press.<br />
Rogers, H. (Eds) 2001. Madagascar. Lonely Planet Publications Pty Ltd.<br />
Seddon, N., Tobias, J., Yont, J. W., Ramanampamonjy, J. R., Butchart, S. &<br />
Randrianizahana, H. 2000. Conservation issues and priorities in the Mikea Forest of<br />
south-west Madagascar. Oryx Vol 34 No. 4.<br />
Schatz, G.E. 2000. En<strong>de</strong>mism in the Malagasy tree flora. Pp. 1-9, In: W. R. Lourenco<br />
& S. M. Goodman (eds), Diversite et En<strong>de</strong>misme a Madagascar, Memoires <strong>de</strong> la<br />
Societe <strong>de</strong> Biogeographie, Paris.<br />
75
Schatz, G. E. 2001. Generic Tree Flora of Madagascar. Royal Botanical Gar<strong>de</strong>ns,<br />
Kew & Missouri Botanical Gar<strong>de</strong>n. Cromwell Press.<br />
Sinclair, I. & Langrand, O. 1998. Birds of the Indian Ocean Islands. Struik Pblishers<br />
(Pty) Ltd. Cape Town.<br />
Smith, A. P. 1997. Deforestation, fragmentation and reserve <strong>de</strong>sign in western<br />
Madagascar. In: Tropical Forest Remnants, Ecology, Management and Conservation<br />
of fragmented Communities (eds. Lawrence, W. & Bierregaard, O. W.). University of<br />
Chicago press, Chicago.<br />
Stephenson, P. J. 1994. Seasonality effect on small mammal trap success in<br />
Madagascar. Journal of Tropical Ecology 10:439-444.<br />
Sussman, R. W. & Rakotozafy, A. 1994. Plant Diversity and Structural Analysis of a<br />
Tropical Dry Forest in Southwestern Madagascar. Biotropica 26 (3): 241-254.<br />
Sussman, R. W., Richard, A. F. & Ravelojaona, G. 1985. Madagascar: Current<br />
Projects And Problems In Conservation. Primate Conservation, No. 5.<br />
Whitesi<strong>de</strong>s, G. H., Oates, J. F., Green, S. M. & Klubendanz, R. P. 1988. Estimating<br />
Primate Densities from Transects in a West African Rain Forest: A Comparison of<br />
Techniques. Journal of Animal Ecology, 57, 345-367.<br />
Wilson, E. O. (Ed), 1992. <strong>The</strong> Diversity of Life. Harvard University Press,<br />
Cambridge.<br />
Wilson, D. E., Cole, F. R., Nichols, J. D., Rudran, R. & Foster, M. S. 1996.<br />
Measuring and Monitoring Biological Diversity. Standard Methods for Mammals.<br />
Smithsonian Instistution.<br />
WWF 2002. Action Plan 2002-2005. WWF-Madagascar Dry Forest Programme,<br />
Toliara.<br />
76
Appendix 1.<br />
Trapsite (TS) Habitat Descriptions<br />
TS 1:<br />
Antafoky (Andamalamy). Secondary riparian forest. Some old selective logging.<br />
Dominated by tamarind (Tamarindus indica) and fig (Ficus) trees. Some large (>30<br />
cm DBH) trees. Good regeneration with many saplings.<br />
Ground layer < 10 %, scrub layer < 10 %, canopy layer > 50 %. Canopy height 15-20<br />
metres. Distance to water (lake) 1-60 metres.<br />
TS 2:<br />
BL 1+2:<br />
BL 3:<br />
Antafoky (Tongomboa).<br />
Secondary riparian forest. Some old selective logging. Dominated by tamarind<br />
(Tamarindus indica) and fig (Ficus) trees. Some large (>30 cm DBH) trees. Good<br />
regeneration with many sapling.<br />
Ground layer < 10 %, scrub layer < 10 %, canopy layer > 50 %. Canopy height 15-20<br />
metres. Distance to water (lake) 1-70 metres.<br />
Cultivated area. Slash and burn, two to three years old maize field. No live trees, soil<br />
pow<strong>de</strong>ry and mostly ash. Ground, scrub, and canopy layer 50 %. Canopy height 15-20 metres. Distance to water<br />
(stream/river) 0-50 metres.<br />
TS 5:<br />
Antafoky (Andamalamy). Secondary riparian forest. Some old selective logging.<br />
Dominated by tamarind (Tamarindus indica) and fig (Ficus) trees. Some large (>30<br />
cm DBH) trees. Good regeneration with many sapling.<br />
Ground layer < 10 %, scrub layer < 10 %, canopy layer > 50 %. Canopy height 15-20<br />
metres. Distance to water (lake) 10-100 metres.<br />
TS 6:<br />
BL 1+2:<br />
BL 3:<br />
Antafoky (Tongomboa).<br />
Secondary riparian forest. Some old selective logging. Dominated by tamarind<br />
(Tamarindus indica) and fig (Ficus) trees. Some large (>30 cm DBH) trees. Good<br />
regeneration with many saplings.<br />
Ground layer < 10 %, scrub layer < 10 %, canopy layer > 50 %. Canopy height 15-20<br />
metres. Distance to water (lake) 2-60 metres.<br />
Riparian forest. Highly disturbed from recent slash and burn. Some live trees.<br />
Ground layer < 10 %, scrub layer < 10 %, canopy layer 10-50 %. Distance to water ~<br />
70 metres.<br />
TS 7: Sept Lacs.<br />
BL 1+2: Transitional forest, some selective logging. Ground layer < 10 %, scrub layer 10-50<br />
%, canopy layer > 50 %. Canopy height 5-10 metres. Distance to water (stream/river)<br />
2-100 metres.<br />
BL 3: Thorny scrub, hilltop, some selective logging. Ground layer < 10%, scrub layer ><br />
50%, canopy layer < 50%. Canopy height 3-4 metres. No water association.<br />
77
TS 8:<br />
Sept Lacs. Tall riparian forest, some selective logging. Valley along road. Some large<br />
(DBH > 30 cm.) trees, good regeneration. Ground layer < 10%, scrub layer < 10 %,<br />
canopy layer > 50 %. Canopy height 15-20 metres. Distance to marsh 5-80 metres.<br />
TS 9:<br />
BL 1+3:<br />
BL 2:<br />
Antafoky (Andamalamy).<br />
Secondary riparian forest. Some old selective logging. Dominated by tamarind<br />
(Tamarindus indica) and fig (Ficus) trees. Some large (>30 cm DBH) trees. Good<br />
regeneration with many saplings.<br />
Ground layer < 10 %, scrub layer < 10 %, canopy layer > 50 %. Canopy height 15-20<br />
metres. Distance to water (lake) 3-50 metres.<br />
Transitional scrub/forest. Some selective logging. Rocky and dry.<br />
Ground layer < 10 %, scrub layer 10-50 %, canopy layer > 50 %. Canopy height ~ 5<br />
metres. No water association.<br />
TS 10:<br />
BL 1:<br />
BL 2:<br />
BL 3:<br />
Antafoky (Tongomboa)<br />
Secondary riparian forest. Some old selective logging. Dominated by tamarind<br />
(Tamarindus indica) and fig (Ficus) trees. Some large (>30 cm DBH) trees. Good<br />
regeneration with many sapling.<br />
Ground layer < 10 %, scrub layer < 10 %, canopy layer > 50 %. Canopy height 15-20<br />
metres. Distance to water (lake) 3-100 metres.<br />
Transitional scrub. Hillsi<strong>de</strong>. Some old charcoal production. Sign of Zebu grazing.<br />
Ground layer < 10 %, scrub layer > 50 %, canopy layer ~ 50 %. Canopy height ~ 3-4<br />
metres. No water association.<br />
Transitional forest. Hilltop. Some old selective logging. Ground layer < 10 %, scrub<br />
layer 10-50 %, canopy layer > 50 %. Canopy height 5- 7 metres. No water<br />
association.<br />
TS 11:<br />
Sept Lacs. Transitional forest. Rocky area, some succulents. No tamarind/fig. Mostly<br />
small, new trees (DBH < 10), but some large trees (DBH > 10). Ground layer < 10%,<br />
scrub layer 10-50 %, canopy layer > 50%. Emergents 10-15 metres. Distance to<br />
stream/river 3-200 metres.<br />
TS 12: Man<strong>de</strong>rano.<br />
BL 1+2: Tall riparian forest, some selective logging and some grazing. Large tamarind and<br />
fig trees, some saplings and regeneration. Ground layer < 10%, scrub layer < 10%,<br />
canopy layer > 50%. Canopy height 15-20 metres. Distance to marsh ~ 20-40 metres.<br />
BL 3: Transitional forest, some selective logging. Ground layer < 10%, scrub layer < 10%,<br />
canopy layer 10-50%. Canopy height 5-10 metres. No water association.<br />
TS 13: Man<strong>de</strong>rano, valleys along road towards Antafoky.<br />
BL 1: Riparian forest. Some selective logging. Ground layer < 10%, scrub layer < 10%,<br />
canopy layer > 50%. Canopy height 10-15 metres. Distance to marsh ~ 50 metres.<br />
BL 2: Riparian forest. Disturbance from logging and charcoal burning. Ground layer < 10%,<br />
scrub layer < 10%, canopy layer > 50%. Canopy height 10-15 metres. Distance to<br />
marsh ~ 50 metres.<br />
TS 14:<br />
BL 1:<br />
BL 2:<br />
Antafoky (Andamalamy)<br />
Secondary riparian forest. Some old selective logging. Dominated by tamarind<br />
(Tamarindus indica) and fig (Ficus) trees. Some large (>30 cm DBH) trees. Good<br />
regeneration with many saplings. Ground layer < 10 %, scrub layer < 10 %, canopy<br />
layer > 50 %. Canopy height 15-20 metres. Distance to water (lake) 1-60 metres.<br />
Transitional scrub/forest. Some selective logging. Rocky, dry.<br />
78
Ground layer < 10 %, scrub layer 10-50 %, canopy layer 50 %. Canopy height ~ 5<br />
metres. No water association.<br />
TS 15:<br />
BL 1:<br />
BL 2:<br />
Antafoky (Tongomboa).<br />
Secondary riparian forest. Some old selective logging. Partly maize cultivation in<br />
between riparian forest trees. Ground layer < 10 %, scrub layer < 10 %, canopy layer<br />
10-50 %. Canopy height 10-15 metres. Distance to water (lake) 1-70 metres.<br />
Old cultivated area. Slash and burn, two to three years old maize field. Some live<br />
trees, soil pow<strong>de</strong>ry and mostly ash. Ground, scrub, and canopy layer 10-50 %. No<br />
water association.<br />
TS 16: Man<strong>de</strong>rano.<br />
BL 1: Tall riparian forest, some selective logging and some grazing. Large tamarind and<br />
fig trees, some saplings and regeneration. Ground layer < 10%, scrub layer < 10%,<br />
canopy layer > 50%. Canopy height 15-20 metres. Distance to marsh ~ 20-40 metres.<br />
BL 2: Transitional forest, some selective logging. Rocky. Ground layer < 10%, scrub layer<br />
< 10%, canopy layer 10-50%. Canopy height 5-10 metres. No water association.<br />
TS 17 :<br />
BL 1:<br />
Sept Lacs.<br />
Transitional forest. Some selective logging. Rocky. Ground layer
Appendix 2.<br />
Numbers of individuals of plants encountered in eight 16 x 16m quadrats in<br />
gallery forest. (Nomenclature based on Schatz, 2001.)<br />
Tree DBH<br />
Vernacular name Family and species<br />
ANACARDIACEAE<br />
Sohihy Abrahamia grandidieri 1<br />
ASCLEPIADACEAE<br />
Bokabe Mar<strong>de</strong>nia cordifolia 1<br />
BIGNONIACEAE<br />
25cm<br />
Somontsoy Fernandoa madagascariensis 5 2 1<br />
BORAGINACEAE<br />
Varo Cordia varo 3 2<br />
BRASSICACEAE<br />
Paky Boscia madagascariensis 1<br />
Kalaogna Crateva excelsa 1 2 2<br />
BURSERACEAE<br />
Daro Commiphora sp1 1 2<br />
CELASTRACEAE<br />
Reampy Buxiella sp 3 1<br />
Tsingilofilo Maytenus linearis 4<br />
CELTIDACEAE<br />
Tsiambanilaza Celtis bifida 99 47 15 9 3<br />
COMBRETACEAE<br />
Vahifoty Combretum sp. 1<br />
Talinala Terminalia rhopalophora 1<br />
EBENACEAE<br />
Maintifototse Diospyros sakalavarum 68 23 12 2<br />
ERYTHROXYLACEAE<br />
Montsoy Erythroxylum sp. 17 1<br />
EUPHORBIACEAE<br />
Hazombalala Croton mahafaliens 7 1 3<br />
Pisopiso Croton sp1 5<br />
Hazondimohy Drypetes sp. 16 16 8 2<br />
Laro Euphorbia laro 1 1 3 2<br />
Voafogna Flueggea virosa virosa 1 1<br />
Hazomena Securinega seyrigii 1<br />
Fitrekantsifake 19 12 4 3 1<br />
80
Appendix 2. Continued.<br />
FABACEAE<br />
Robontsy Acacia morondavensis 2<br />
Mendoravy Albizia greveana 1 1<br />
Bagnake Bauhinia hil<strong>de</strong>bradtii 6 9 1<br />
Sarongaza Colvillea racemosa 4<br />
Magnary Dalbergia greveana 1<br />
Kily Tamarindus indica 5 2 20<br />
HYPPOCRETACEAE<br />
Vahimpindy Hyppocratea angustipetala 4 2<br />
LAMIACEAE<br />
Tsilaitsora Capitanopsis angustifolia 39 2<br />
LOGANIACEAE<br />
Relefo Strychnos <strong>de</strong>cussata 5 1<br />
Bakoa Strychnos madagascariensis 8 4 1<br />
MALVACEAE<br />
Malimatsy Grewia bororum 1<br />
Katepoky Grewia grevei 1<br />
MELIACEAE<br />
Hompy Quivisianthe papinae 1<br />
Mampisaraky Turraea sericea 1 3 1<br />
MORACEAE<br />
Mandresy Ficus mormorata 1 1 1<br />
Fihamy Ficus pyrifolia 1<br />
OLACACEAE<br />
Tsilaitsy Olax lanceolata 37 26 30 6<br />
Tanjaky Olax pervilleana 12 5 7<br />
PANDANACEAE<br />
Fandra Pandanus sp. 1<br />
PAPILIONACEAE<br />
Fangitsy Dolichos fangitsy 1<br />
PHYSENACEAE<br />
Fandriandambo Physena sessiliflora 13 4<br />
RHAMNACEAE<br />
Roihavitse Scutia sp. 3<br />
Tapisaky Xericisyos dangui 1<br />
81
Appendix 2. Continued.<br />
RUBIACEAE<br />
Beholitse Hymenodictyon <strong>de</strong>caryum 1<br />
Voafotake Rothmannia' sp. 2 2<br />
Mantsake Tarenna pruinosum 7 1 4<br />
Lamotimboay Catunaregam spinosa 1<br />
Litsake 2 1<br />
RUTACEAE<br />
Monongo Zanthoxylum tsihanimposa 1<br />
SALICACEAE<br />
Lalipito Bivinia jalberti 3 3 3<br />
SAPINDACEAE<br />
Matitihena Stadmania oppositifolia 18 3 2<br />
SPHAEROSEPALACEAE<br />
Talafoty Rhopalocarpus lucidus 1 3 1<br />
VITACEAE<br />
Laza Cyphostema laza 1 1<br />
UNIDENTIFIED<br />
Bika 1<br />
Diamainty 1<br />
Filankibo 14 4<br />
Hazomposa 1<br />
Malaignarety 1<br />
Roimena 1<br />
Sandrisandry 11 10<br />
Sarevaro 1 1<br />
Tagnatagna 1<br />
Tagnatagnanala 2<br />
Tambakobe 1<br />
Tamenaky 4<br />
Tandrokaondrilahy 1<br />
Timbatse 1<br />
Tsimatindrahavelo 1<br />
82
Appendix 3.<br />
Numbers of individuals of plants encountered in eight 32 x 32m quadrats in<br />
Spiny forest. (Nomenclature based on Schatz, 2001.)<br />
Tree DBH<br />
Vernacular<br />
name<br />
Family and species 25cm<br />
ANACARDIACEAE<br />
Jabihy Operculicarya <strong>de</strong>caryi 16 6 11 26<br />
Sakoakomoke Poupartia birrea caffra 2 8<br />
ASCLEPIADACEAE<br />
Ranga Cynancum perieri 4<br />
Try Cynancum sp. 1<br />
Taritariky Lepta<strong>de</strong>nia madagascariensis 1<br />
Bokabe Mar<strong>de</strong>nia cordifolia 1<br />
ASTERACEAE<br />
Malaignevotsy Psiadia altissima 1 1<br />
BIGNONIACEAE<br />
Somontsoy Fernandoa madagascariensis 4 2 1<br />
Mangarahara Stereospermum euphoroi<strong>de</strong>s 3<br />
BORAGINACEAE<br />
Lambotaho Erhetia ambovombes 17 6 3<br />
BRASSICACEAE<br />
Paky Boscia madagascariensis 14 8 1 3<br />
Maharoake Boscia sp. 8 3 1 2<br />
Mahafangalitse Thilachium monophyllum 5 1<br />
BURSERACEAE<br />
Mamy Canarium madagascariensis 1<br />
Arofy Commiphora arafy 2 5 10 8 3<br />
Sengatsy Commiphora simplicifolia 8 1 3 7<br />
Daro Commiphora sp1 27 21 14 16 2<br />
Tariby Commiphora sp2 43 26 21 3<br />
Daromena Commiphora sp3 1<br />
CELASTRACEAE<br />
Reampy Buxiella sp 1<br />
COMBRETACEAE<br />
Kapikala Combretum grandidieri 1 2<br />
Talinala Terminalia rhopalophora 1 4 1 2<br />
Taly Terminalia tricristata 3 3<br />
CONNARACEAE<br />
Vahimainty Agelaea peutagyna 2<br />
CONVALLARIACEAE<br />
Tavevola Dracaena sp 31 2<br />
83
Appendix 3. Continued.<br />
CYCADACEAE<br />
Fatra Cycas thouarsii 27 18 3 1<br />
EBENACEAE<br />
Hazonta Diospyros humbertiana 9<br />
Kobaintsihotse Diospyros latispatula 100 44 17<br />
Maintifototse Diospyros sakalavarum 13 5<br />
ERYTHROXYLACEAE<br />
Montsoy Erythroxylum sp. 2 3<br />
EUPHORBIACEAE<br />
Hazombalala Croton mahafaliens 14 17 9<br />
Andriambolafotsy Croton sp.1 70 1 1<br />
Somoro Croton sp.2 43 6<br />
Pisopiso Croton sp.3 178 2<br />
Karimbola Croton sp.4 32 5 1<br />
Teloravy Dalechampia elematidifolia 1 1<br />
Antso Euphorbia antso 23 9 9 7<br />
Pira Euphorbia intisy 37 24 15 1<br />
Laro Euphorbia laro 3 1 1<br />
Betondro Euphorbia oncoclada 11 6 4<br />
Famata Euphorbia sp 3 1<br />
Avoha Jatropha curcus 37 5 9 1<br />
Katratra Jatropha mahafaliensis 28 13 4 2<br />
Hazomena Securinega seyrigii 45 32 14 5<br />
FABACEAE<br />
Fatipatike Acacia angustifolia 9 3 2<br />
Ambilazo Alantsilo<strong>de</strong>ndron sp. 49 53 37 14<br />
Mendoravy Albizia greveana 6 2 2 1 1<br />
Hafobakondry Bauhinia grandidieri 101 11<br />
Bagnake Bauhinia hil<strong>de</strong>bradtii 6 2<br />
Kitomba Caesalpinia sp 1 1<br />
Sarongaza Colvillea racemosa 1 2<br />
Fengoke Delonix sp 2 4 1 1<br />
Lovanafy Dicraeopetalum mahafaliense 35 21 2 1<br />
Roy Mimosa sp. 15 7 1<br />
Talamena Neoapaloxylon sp 5 2 3 2<br />
Sofasofa Ormocarpopsis sp 10 2<br />
Vaovy Tetrapterocarpon geayi 30 7 3 1<br />
Tsikidrakitsy Voganhia mahafaliensis 6 6<br />
Kapaipoty<br />
HERNANDIACEAE<br />
Gyrocarpus americanus<br />
capuronianus<br />
1 3 5 1<br />
HYPPOCRETACEAE<br />
Vahimpindy Hyppocratea angustipetala 6<br />
LILIACEAE<br />
Vaho Aloe divaricata 1<br />
Vahontsoy Aloe sp. 2 1<br />
84
Appendix 3. Continued.<br />
LOGANIACEAE<br />
Relefo Strychnos <strong>de</strong>cussata 2 1<br />
LYTHRACEAE<br />
Lelaosy Woodfordia fructicosa 2 1<br />
MALVACEAE<br />
Tainkafotsy Dombeya ambongoensis 12 8 4<br />
Satro Dombeya sp 16 5 5 1<br />
Malimatsy Grewia bororum 13 4 4 1<br />
Katepoky Grewia grevei 140 19 3<br />
Selibohoky Grewia sp. 1 2<br />
Sely Grewia sp1 2 1<br />
Ampelambatotse Helmiopsiella madagascariensis 1<br />
MELIACEAE<br />
Handimbohitse Neobeguea mahafaliensis 3<br />
MORACEAE<br />
Fihamy Ficus pyrifolia 1<br />
Nonokamboa Ficus sp. 1<br />
OLACACEAE<br />
Tsilaitsy Olax lanceolata 1 1 1<br />
Tanjaky Olax pervilleana 11 3 1<br />
PANDANACEAE<br />
Fandra Pandanus sp. 1<br />
PAPILIONACEAE<br />
Fangitsy Dolichos fangitsy 2<br />
Malaikifafa Indigofera tinctoria 1<br />
Kifafa Plubango aphylla 6<br />
PASSIFLORACEAE<br />
Hola A<strong>de</strong>nia holaboainaris 1 1<br />
PEDALIACEAE<br />
Farehitse Uncarina sp. 8 2<br />
PHYSENACEAE<br />
Fandriandambo Physena sessiliflora 3<br />
PTAEROXYLACEAE<br />
Katrafaifilo Cadrelopsis grevei 2 1<br />
RHAMNACEAE<br />
Roihavitse Scutia sp. 1<br />
Tapisaky Xericisyos dangui 3 2<br />
RUBIACEAE<br />
Lamotimboay Catunaregam spinosa spinosa 1<br />
Voligejy Gar<strong>de</strong>nia rutenbergiana 4<br />
85
Appendix 3. Continued.<br />
RUBIACEAE (cont.)<br />
Kaposera Poe<strong>de</strong>rum grevei 50 22<br />
Voafotake Rothmannia' sp. 1 4 1<br />
Lavaravy Tarenna sp. 1<br />
RUTACEAE<br />
Katrafay Cedrelopsis gracilis 121 25 7 3<br />
Katrafaintsifake Cedrelopsis simplifolia 1<br />
Monongo Zanthoxylum tsihanimposa 3 1<br />
SALICACEAE<br />
Lalipito Bivinia jalberti 2<br />
SAPINDACEAE<br />
Matitihena Stadmania oppositifolia 3 1 1<br />
SPHAEROSEPALACEAE<br />
Talafoty Rhopalocarpus lucidus 4 1 3<br />
Lombiry Rhopalocarpus sp 2<br />
UNIDENTIFIED<br />
Andriambelo 4 5 1<br />
Atolimpoly 28 4<br />
Berotsy 7 3 1<br />
Boramena 2 1<br />
Eringitse 12<br />
Filankibo 2<br />
Hatambogna 5 3<br />
Hazofio 33 12 5 1<br />
Hazofoty 51 16 2 1<br />
Hazomamy 7<br />
Hazondragnaty 2 7 1<br />
Hazonkoaky 1<br />
Hazovola 5 1<br />
Katrafay 43 8 3<br />
Kopitse 3<br />
Lamendra 1<br />
Malaitambio 102 9 1<br />
Roimena 1<br />
Sandrisandry 89<br />
Saresakoa 10 4 4 2<br />
Sariboy 1<br />
Sp1 22<br />
Tainkoaky 5<br />
Tambarikosy 3 1<br />
Tandrokaondrilahy 1<br />
Timbatse 1<br />
Tsianagnampo 1<br />
Tsimarefe 1 2<br />
Tsimatindrahavelo 1<br />
Vahimena 2 2<br />
Valae 1<br />
Volomboto 6 13 6 1<br />
86
Appendix 4.<br />
Plant numbers and specific uses – Sept Lacs (west)<br />
Family Genus and species Notes<br />
< 2.5cm 2.5 - 5cm 5 - 10cm 10 - 25cm >25cm<br />
Gallery forest<br />
SALICACEAE Bivinia jalberti Wood - construction 1<br />
BRASSICACEAE Boscia madagascariensis Edible fruit, leaves - rheumatism 1<br />
FABACEAE Colvillea racemosa Wood - construction, pirogues 4<br />
BURSERACEAE Commiphora sp1 Wood - construction 1 2<br />
FABACEAE Dalbergia greveana Wood - construction & furniture 1<br />
EBENACEAE Diospyros sakalavarum Fruit - edible for lemurs 63 20 11 2<br />
MORACEAE Ficus mormorata Fruit - edible for lemurs 1 1 2<br />
LOGANIACEAE Strychnos madagascariensis Fruit - edible for lemurs and man 8 4 1<br />
FABACEAE Tamarindus indica Fruit - edible for man and lemurs 1 3 17<br />
Spiny forest<br />
BRASSICACEAE Boscia madagascariensis Edible fruit, leaves - rheumatism 9 4 1 2<br />
FABACEAE Colvillea racemosa Wood - construction, pirogues 1 2<br />
BURSERACEAE Commiphora arafy Wood - construction 2 5 10 7 3<br />
BURSERACEAE Commiphora simplicifolia Wood - construction 6 1 3 7<br />
BURSERACEAE Commiphora sp1 Wood - construction 3 6 2 4 1<br />
BURSERACEAE Commiphora sp2 Wood - construction 1<br />
BURSERACEAE Commiphora sp3 Wood - construction 2 6 6 1<br />
FABACEAE Delonix sp Resin - glue, wood - pirogues 1 3 1 1<br />
EBENACEAE Diospyros humbertiana Fruit - edible for lemurs 3<br />
EBENACEAE Diospyros latispatula Fruit - edible for lemurs 66 31 7<br />
EBENACEAE Diospyros sakalavarum Fruit - edible for lemurs 4<br />
EUPHORBIACEAE Jatropha mahafaliensis Whole plant - living fence 3 3 2<br />
FABACEAE Neoapaloxylon sp Root - edible, bark - rope 5 2 3 2<br />
FABACEAE Ormocarpopsis sp Wood - construction & tool handles 4 2<br />
ANACARDIACEAE Poupartia birrea caffra Fruit - edible 4 8<br />
PEDALIACEAE Uncarina sp. Roots - medicinal 8 2<br />
87
Appendix 5.<br />
Plant numbers and specific uses – Antafoky<br />
Family Genus and species Notes<br />
< 2.5cm 2.5 - 5cm 5 - 10cm 10 - 25cm >25cm<br />
Gallery forest<br />
SALICACEAE Bivinia jalberti Wood - construction 2 3 3<br />
EBENACEAE Diospyros sakalavarum Fruits - eaten by lemurs 5 3 1<br />
FABACEAE Tamarindus indica Fruits - eaten by lemurs and man 4 3<br />
Spiny forest<br />
SALICACEAE Bivinia jalberti Wood - construction 2<br />
BRASSICACEAE Boscia madagascariensis Edible fruit, leaves - rheumatism 5 4 1<br />
BRASSICACEAE Boscia sp. Edible fruit, leaves - rheumatism 8 3 1 2<br />
BURSERACEAE Canarium madagascariensis Wood - pirogues, fruit - eaten by lemurs 1<br />
BURSERACEAE Commiphora arafy Wood - construction 1<br />
BURSERACEAE Commiphora simplicifolia Wood - construction 2<br />
BURSERACEAE Commiphora sp.3 Wood - construction 34 19 15 2<br />
BURSERACEAE Commiphora sp2 Wood - construction 7 1<br />
BURSERACEAE Commmiphora sp 1 Wood - construction 24 15 12 12 1<br />
FABACEAE Delonix sp Resin - glue, wood - pirogues 1 1<br />
EBENACEAE Diospyros humbertiana Fruits - eaten by lemurs 6<br />
EBENACEAE Diospyros latispatula Fruits - eaten by lemurs 34 13 10 1<br />
EBENACEAE Diospyros sakalavarum Fruits - eaten by lemurs 9 5<br />
MORACEAE Ficus pyrifolia Fruits - eaten by lemurs 1<br />
MORACEAE Ficus sp. Fruits - eaten by lemurs 1<br />
EUPHORBIACEAE Jatropha mahafaliensis Whole plant - living fence 15 10 4<br />
FABACEAE Ormocarpopsis sp Wood - construction and tool handles 6<br />
COMBRETACEAE Terminalia rhopalophora Wood - construction 1 4 1 2<br />
88
Appendix 6.<br />
List of mammal species found in the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>.<br />
HABITAT<br />
No. SPECIES Method Gallery Transitional En<strong>de</strong>mic<br />
forest scrub / forest Status<br />
Family: Pteropopidae<br />
1 Eidolon dupreanum O E V<br />
Family: Hipposi<strong>de</strong>ridae<br />
2 Hipposi<strong>de</strong>ros commersoni C, O E<br />
3 Triaenops furculus C E V<br />
4 Triaenops rufus C E<br />
Family: Vespertilionidae<br />
5 Myotis goudoti C E<br />
6 Miniopterus sp C E<br />
7 Pipistrellus sp C E<br />
Family: Molossidae<br />
8 Tadarida leucogaster* C E<br />
Family: Tenrecidae<br />
9 Tenrec ecaudatus C, O E<br />
10 Setifer setosus C E<br />
11 Echinops telfairi C E<br />
12 Geogale aurita C E<br />
13 Microgale sf brevicaudata C E<br />
Family: Soricidae<br />
14 Suncus madagascariensis C N<br />
Family: Muridae<br />
15 Eliurus myoxinus C, O E<br />
16 Rattus rattus C, O N<br />
17 Mus musculus C, O N<br />
Family: Viverridae<br />
18 Cryptoprocta ferox T E En<br />
19 Viverricula indica O N<br />
Family: Cheirogaleidae<br />
20 Microcebus murinus** C, O E<br />
21 Microcebus griseorufus O E<br />
22 Cheirogaleus medius O E V<br />
23 Mirza coquereli O, T E V<br />
Family: Lemuridae<br />
24 Lemur catta O, T E V<br />
Family: Indriidae<br />
25 Propithecus verreauxi<br />
verreauxi<br />
Family: Suidae<br />
26 Potamochoerus larvatus<br />
larvatus<br />
IUCN<br />
Status<br />
O, T E V<br />
T N<br />
KEY<br />
Methods: C= captured, O= observed, T= tracks, calls, dung.<br />
Status: E= en<strong>de</strong>mic to Madagascar, N = Non-en<strong>de</strong>mic. V = vulnerable, En = endangered.<br />
* Colony roosting in building in village. ** I<strong>de</strong>ntified from the colouration of observed individuals.<br />
89
No.<br />
Appendix 7<br />
List of bird species found in the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong><br />
SPECIES<br />
LOCATION<br />
Antafoky<br />
Lake<br />
Sept Lacs<br />
region<br />
Gallery<br />
forest<br />
HABITAT<br />
Spiny<br />
forest<br />
Disturbed<br />
ENDEMIC<br />
STATUS<br />
IUCN<br />
STATUS<br />
1 Nycticorax nycticorax<br />
N<br />
2 Ar<strong>de</strong>ola idae RE V<br />
3 Bubulcus ibis N<br />
4 Butori<strong>de</strong>s striatus<br />
N<br />
5 Egretta dimorpha<br />
N<br />
6 Egretta alba<br />
N<br />
7 Ar<strong>de</strong>a purpurea<br />
N<br />
8 Ar<strong>de</strong>a cinerea N<br />
9 Ar<strong>de</strong>a humbloti E V<br />
10 Scopus umbretta<br />
N<br />
11 Dendrocygna viduata<br />
N<br />
12 Sarkidiornis melanotos<br />
N<br />
13 Aviceda Madagascariensis<br />
E<br />
14 Milvus aegyptius N<br />
15 Polyboroi<strong>de</strong>s radiatus E<br />
16 Accipiter madagascariensis<br />
E NT<br />
17 Accipiter francesii<br />
RE<br />
18 Buteo brachypterus<br />
E<br />
19 Falco newtoni<br />
RE<br />
20 Falco concolor N<br />
21 Falco peregrinus N<br />
E<br />
22 Margaroperdix<br />
madagascariensis<br />
23 Numida meleagris<br />
N<br />
24 Turnix nigricollis E<br />
25 Dryolimmas cuvieri RE<br />
26 Gallinula chloropus<br />
N<br />
27 Actitis hypoleucos<br />
N<br />
28 Sterna caspia<br />
N<br />
29 Pterocles personatus<br />
E<br />
30 Sreptopelia picturata RE<br />
31 Oena capensis N<br />
32 Treron australis<br />
RE<br />
33 Coracopsis vasa<br />
RE<br />
34 Coracopsis nigra<br />
RE<br />
35 Agapornis canus<br />
E<br />
36 Cuculus rochii EB<br />
37 Coua gigas E<br />
38 Coua coquereli<br />
E<br />
39 Coua cursor<br />
E<br />
40 Coua ruficeps<br />
E<br />
41 Coua cristata<br />
E<br />
90
Appendix 7. Continued.<br />
42 Centropus toulou RE<br />
43 Tyto alba N<br />
44 Otus rutilus<br />
RE<br />
45 Ninox superciliaris<br />
E<br />
46 Asio madagascariensis<br />
E<br />
RE<br />
47 Caprimulgus<br />
madagascariensis<br />
48 Cypsiurus parvus<br />
N<br />
49 Alcedo vintsioi<strong>de</strong>s<br />
RE<br />
50 Merops supersciliosus<br />
N<br />
51 Eurystomus glaucurus<br />
N<br />
52 Leptosomus discolor N<br />
53 Upupa marginata E<br />
54 Phedina borbonica<br />
RE<br />
55 Motacilla flaviventris<br />
E<br />
56 Coracina cinerea<br />
RE<br />
57 Hypsipetes madagascarensis<br />
RE<br />
58 Vanga curvirostris E<br />
59 Xenopirostris xenopirostris E<br />
60 Falculea palliata<br />
E<br />
61 Artamella viridis<br />
E<br />
62 Leptopterus chabert<br />
E<br />
63 Saxicola torquata<br />
N<br />
64 Copsychus albospecularis E<br />
65 Thammornis chloropetoi<strong>de</strong>s E<br />
66 Newtonia brunneicauda<br />
E<br />
67 Newtonia archboldi<br />
LE<br />
68 Cisticola cherina<br />
RE<br />
69 Neomixis tenella<br />
E<br />
70 Neomixis striatigula E<br />
71 Terpsiphone mutata RE<br />
72 Nectarinia souimanga<br />
RE<br />
73 Zosterops ma<strong>de</strong>raspatana<br />
RE<br />
74 Lonchura nana<br />
E<br />
75 Ploceus sakalava<br />
E<br />
76 Foudia madagascariensis E<br />
77 Acridotheres tristis N<br />
78 Dicrurus forficatus<br />
RE<br />
79 Corvus albus<br />
N<br />
73 58 60 50 58<br />
KEY:<br />
Status: E= en<strong>de</strong>mic to Madagascar, EB – En<strong>de</strong>mic Bree<strong>de</strong>r, RE – <strong>Regional</strong> En<strong>de</strong>mic, LE – Locally en<strong>de</strong>mic, N = Non-en<strong>de</strong>mic,<br />
V = vulnerable, NT = Near Threatened.<br />
91
Appendix 8.<br />
Bird species that are different from all other positive i<strong>de</strong>ntifications but that could not<br />
be confirmed to species level.<br />
LOCATION<br />
HABITAT<br />
Common<br />
No. SPECIES<br />
Antafoky Sept Lacs Gallery Spiny Disturbed name<br />
region forest forest<br />
1 Phoenicopterus sp Flamingo<br />
2 Anas sp Teal<br />
3 Charadrius sp Plover<br />
4 Sterna Tern<br />
5 Anous stolidus Brown<br />
noddy<br />
6 Zoonavena grandidieri Madagascar<br />
spinetail<br />
7 Apus barbatus African<br />
black swift<br />
8 Monticola sp Rock thrush<br />
9 Tylas eduardi Tylas vanga<br />
10 Nestillas Warbler<br />
11 Hartlaubius auratus Madagascar<br />
starling<br />
92
Appendix 9.<br />
List of reptile species found in the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>.<br />
No. SPECIES<br />
Antafoky<br />
Lake area<br />
Family: Testudinidae<br />
LOCATION<br />
Sept Lacs<br />
region<br />
Riparian<br />
forest<br />
HABITAT<br />
Spiny forest<br />
ENDEMIC<br />
STATUS<br />
IUCN &<br />
CITES<br />
STATUS<br />
1<br />
Geochelone radiata<br />
RE V,<br />
CITES I<br />
2 Pelusios castanoi<strong>de</strong>s N<br />
3 Pelomedusa subrufa N CITES III<br />
Family: Boidae<br />
4<br />
Acrantophis dumerili<br />
E V,<br />
CITES I<br />
5<br />
Sanzinia madagascariensis<br />
E V,<br />
CITES I<br />
Family: Typhlopidae<br />
6 Typhlops sp 1 RE<br />
7 Typhlops <strong>de</strong>corsei RE<br />
Family: Colubridae<br />
8 Dromicodryas bernieri E<br />
9 Heteroliodon occipitalis RE<br />
10 Ithycyphus oursi E<br />
11 Langaha madagascariensis E<br />
12 Leioheterodon geayi RE<br />
13 Leioheterodon madagascariensis E<br />
14 Leioheterodon mo<strong>de</strong>stus E<br />
15 Liophidium torquatum E<br />
16 Liophidium trilineatum RE<br />
17 Liophidium vaillanti E<br />
18 Liopholidophis lateralis E<br />
19 Lycodryas gaimardi E<br />
20 Madagascarophis colubrinus E<br />
21 Madagascarophis meridionalis RE<br />
22 Madagascarophis ocellatus RE<br />
23 Mimophis mahfalensis E<br />
24 Pseudoxyrhopus quinquelineatus E<br />
Family: Crocodilia<br />
25 Crocodylus niloticus N CITES I<br />
Family: Gekkonidae<br />
26 Blaesodactylus sakalava RE<br />
27 Geckolepis maculata E<br />
28 Geckolepis petiti E<br />
29 Hemidactylus mabouia N<br />
30 Hemidactylus mercatorius N<br />
31 Lygodactylus tolampyae E<br />
32 Lygodactylus tuberosus E<br />
93
LOCATION<br />
No. SPECIES Antafoky GELOSE<br />
protected<br />
area<br />
Riparian<br />
forest<br />
HABITAT<br />
Spiny forest<br />
ENDEMIC<br />
STATUS<br />
IUCN &<br />
CITES<br />
STATUS<br />
33 Paroedura bastardi E<br />
34 Paroedura pictus RE<br />
35 Paroedura vahiny RE<br />
36 Phelsuma mutabilis E CITES II<br />
37<br />
Phelsuma standingi<br />
RE V,<br />
CITES II<br />
Family: Gerrhosauridae<br />
38 Tracheloptychus madagascariensis E<br />
39 Zonosaurus karsteni RE<br />
40 Zonosaurus laticaudatus E<br />
Family: Scincidae<br />
41 Amphiglossus igneocaudatus E<br />
42 Amphiglossus ornaticeps E<br />
43 Androngo trivittatus RE<br />
44 Mabuya dumasi RE<br />
45 Mabuya elegans E<br />
46 Mabuya gravenhorstii E<br />
47 Mabuya vato RE<br />
48 Voeltzkowia fierinensis RE<br />
Family: Iguanidae<br />
49 Chalaradon madagascariensis E<br />
Family: Opluridae<br />
50 Oplurus cyclurus RE<br />
51 Oplurus quadrimaculatus E<br />
Family: Chameleonidae<br />
52 Brookesia brygooi E CITES II<br />
53 Furcifer lateralis E CITES II<br />
54 Furcifer oustaleti E CITES II<br />
55 Furcifer verrucosus E CITES II<br />
TOTAL 48 48<br />
KEY TO ABBREVIATIONS<br />
En<strong>de</strong>mic status:<br />
• RE - <strong>Regional</strong> En<strong>de</strong>mic: En<strong>de</strong>mic species with limited distribution, occurring only in southwest Madagascar.<br />
• E - En<strong>de</strong>mic: Occurring only in Madagascar and associated islands (Comoros, Reunion, Mauritius).<br />
• N – Non-en<strong>de</strong>mic.<br />
IUCN status:<br />
• EN - Endangered<br />
• V - Vulnerable<br />
CITES status<br />
• I – Export and Import authorisations are necessary<br />
• II – Export authorisations are necessary<br />
• III – Exportation must inclu<strong>de</strong> documentation stating the country of origin of the specimen<br />
94
Appendix 10.<br />
List of butterfly species found in the <strong>proposed</strong> <strong>Parc</strong> <strong>Regional</strong> <strong>de</strong> <strong>Belomotse</strong>.<br />
No. SPECIES Riparian forest Spiny forest<br />
Family: Hesperiidae<br />
1 Coelia<strong>de</strong>s ernesti <br />
2 Coelia<strong>de</strong>s rama <br />
3 Coelia<strong>de</strong>s ramanatek ramanatek <br />
4 Tagia<strong>de</strong>s insularis insularis <br />
Family: Lycaenidae<br />
5 Chila<strong>de</strong>s miniscula <br />
6 Cupidopsis jobates jobates <br />
7 Deudorix antalus <br />
8 Hemiolaus cobaltina <br />
9 Leptomyrina phidias <br />
10 Leptotes pirithous <br />
11 Zizula hylax <br />
Family: Nymphalidae<br />
12 Acraea dammii <br />
13 Acraea encedon encedon <br />
14 Acraea igati <br />
15 Acraea lia <br />
16 Acraea obeira obeira <br />
17 Acraea ranavalona <br />
18 Acraea turna <br />
19 Acraea zitja <br />
20 Apaturopsis paulianii <br />
21 Aterica rabena <br />
22 Byblia anvatara anvatara <br />
23 Charaxes andara <br />
24 Charaxes antamboulou <br />
25 Charaxes cacuthis <br />
26 Charaxes zoolina betsimisaraka <br />
27 Danaus chrysippus aegyptius <br />
28 Eurytela dryope lineata <br />
29 Heteropsis ankaratra <br />
30 Heteropsis narcissus fraterna <br />
31 Junonia hierta paris <br />
32 Junonia rhadama <br />
33 Libythea tsiandava <br />
34 Melanitis leda <br />
35 Melanitis leda helena <br />
36 Neptidopsis fulgurata fulgurata <br />
37 Neptis kiki<strong>de</strong>li <br />
38 Neptis saclava saclava <br />
95
No. SPECIES Riparian forest Spiny forest<br />
39 Pseudacraea imerina imerina <br />
40 Sallya amazoula <br />
41 Sallya madagascariensis <br />
42 Morphospecies 77 <br />
Family: Papilionidae<br />
43 Graphium evombar <br />
44 Papilio <strong>de</strong>modocus <strong>de</strong>modocus <br />
45 Papilio epiphorbas epiphorbas <br />
46 Papioio erithonioi<strong>de</strong>s <br />
47 Pharmacophagus antenor <br />
Family: Pieridae<br />
48 Belenois creona prorsus <br />
49 Belenois grandidieri <br />
50 Catopsilia thauruma <br />
51 Colotis amata crowleyi <br />
52 Colotis evanthe <br />
53 Colotis mananhari <br />
54 Colotis zoe <br />
55 Dixea charina narena <br />
56 Eurema floricola <br />
57 Eurema floricola floricola <br />
58 Gi<strong>de</strong>ona lucasi <br />
59 Leptosia alcesta sylvicola <br />
60 Mylothris phileris <br />
61 Pinacopteryx eriphia mabillei <br />
96
Appendix 11.<br />
Descriptions of Suggested Walks<br />
Walk 1<br />
Begins at campsite 1 and follows the edge of the lake to Antafoky village. <strong>The</strong> time<br />
nee<strong>de</strong>d is approximately 1 hour. A medium level of fitness would be required as there<br />
is sometimes a steep gradient and footing may be difficult on the narrow pathway.<br />
Visitors can expect to see a variety of wildlife - in particular the bird life is very<br />
prevalent and obvious, especially water birds (cattle and great egrets, Madagascan<br />
pond, purple, green-backed and night crowned herons), which are abundant. <strong>The</strong> walk<br />
provi<strong>de</strong>s a won<strong>de</strong>rful view of the spectacular limestone cliffs that frame the lake. At<br />
Antafoky village it is possible to rest and have simple refreshments un<strong>de</strong>r the mango<br />
trees overlooking the Onilahy River.<br />
Return is via the same route or boat.<br />
Possible hazards encountered in this walk are narrow paths and loose stones that may<br />
require careful footing. <strong>The</strong> path also passes through several shallows around the lake<br />
and tourists would need to wa<strong>de</strong> up to waist <strong>de</strong>ep in water.<br />
Walk 2<br />
Begins in Antafoky, goes up the steep hill behind the village and then <strong>de</strong>scends down<br />
into the valley to join the road back to Antafoky. <strong>The</strong> time nee<strong>de</strong>d is approximately<br />
five hours. This walk is fairly arduous and requires a high level of fitness due to the<br />
steepness.<br />
Reptiles are abundant and radiated tortoises have been seen here. This walk also<br />
encompasses a wi<strong>de</strong> range of vegetation types - gallery forest, transitional forest,<br />
thorny scrub and some cultivated areas. A spectacular dried limestone waterfall, dried<br />
up riverbed, caves and huge cliffs can be seen. <strong>The</strong> walk inclu<strong>de</strong>s the village of<br />
Antafoky and a tomb on the hillsi<strong>de</strong> which visitors may find interesting.<br />
A pirogue can be taken back to the camp or Walk 1 will take walkers back.<br />
<strong>The</strong> hike is fairly arduous and loose rock and thorny scrub may cause minor injuries<br />
such as scratches. Plenty of water, good walking boots, long trousers and hat and<br />
suncream are recommen<strong>de</strong>d. <strong>The</strong>re are plenty of picnic sites along the route - the best<br />
being the limestone waterfall due to the spectacular viewpoint it provi<strong>de</strong>s. Certain<br />
areas of the path need to be cleared.<br />
Walk 3<br />
This walk starts at campsite 1 and travels up the valley before going up a small hill<br />
and then curving around and returning to camp. <strong>The</strong> estimated required time is two<br />
hours. <strong>The</strong> level of fitness required is fairly low to medium.<br />
97
Many species of birds can be seen, as well as lemurs on the cliff face and many reptile<br />
species. <strong>The</strong> landscape features inclu<strong>de</strong> a fossilised coral reef, sandstone cliffs and a<br />
limestone plateau.<br />
<strong>The</strong> path leads through a cultivated area where there is a lack of sha<strong>de</strong> and the path<br />
needs to be cleared of rocks and vegetation. This area could be the focus of a<br />
replanting scheme.<br />
Walk 4<br />
This is a stunning walk but fairly strenuous. <strong>The</strong> route begins at campsite 1 and goes<br />
around the si<strong>de</strong> of some impressive sandstone cliffs where a spectacular view across<br />
the lake can be seen. From here, the path ascends up a fairly steep path that levels out<br />
on the plateau to give a magnificent view of the surrounding valleys. <strong>The</strong> sandstone<br />
cliffs, a dried-up waterfall and rock amphitheatre are just some of the geological<br />
points of interest to note as you pass along the riverbed towards the lake. This walk<br />
takes 8 hours and requires a mo<strong>de</strong>rate to high level of fitness.<br />
Several areas of the path need to be cleared, but this would only need minor work, and<br />
possibly benches could be placed at strategic places around the walk for bird watching<br />
or just resting and admiring the viewpoints.<br />
Walk 5<br />
This walk begins at campsite 2 at the Sept Lacs and goes up the valley taking in all<br />
seven lakes. At the final pool, a track leads to a zebu-cart path that can be followed<br />
back down to the main Antafoky – Ifanato road, and hence back to camp.<br />
This walk would take around an hour and a half, but most people would want to stop<br />
and swim in at least one of the pools. <strong>The</strong>re are numerous places to do this and there<br />
are many i<strong>de</strong>al picnic spots.<br />
<strong>The</strong> walk could be done by anyone with a mo<strong>de</strong>rate level of fitness. <strong>The</strong> path passes<br />
in and out of the river and may need some careful clearance and management.<br />
Walk 6<br />
This walk begins at campsite 2 and can either begin by following walk 5 and turning<br />
right when the path meets up with the zebu track, or it can follow the zebu track<br />
directly from the road.<br />
<strong>The</strong> walk winds up through the dry forest, affording spectacular views over the<br />
Onilahy River valley, and then passes down through the Sept Lacs River until the<br />
valley opens out again into a patch of gallery forest. A track then continues up to the<br />
left out of the gallery forest into spiny forest, passing a sinkhole that contains a<br />
sizeable colony of bats.<br />
This walk would take around 4 hours. A medium level of fitness would be necessary<br />
for the majority of the way, and it passes through the river for a short distance.<br />
98
Ring-tailed lemurs and Verreaux’s sifaka are abundant, as are many bird and reptile<br />
species. <strong>The</strong> bat colony in the sinkhole comprises approximately 40,000 individuals of<br />
three different species. <strong>The</strong> bats begin to leave their roosts around dusk (about 6 p.m).<br />
<strong>The</strong> first hour is the most spectacular and would be of great interest to many tourists.<br />
Walk 7<br />
This walk starts at campsite 2 and follows walk 6 until the path reaches the area of<br />
gallery forest. From here, follow a small path that runs parallel to the river, taking in a<br />
further three lakes, including an extremely <strong>de</strong>ep pool that is perfect for swimming and<br />
high diving. <strong>The</strong> path and river continue for over five kilometres.<br />
<strong>The</strong> time nee<strong>de</strong>d is approximately 5 hours. A medium level of fitness is required,<br />
although the path is fairly level.<br />
Birds are extremely abundant in the upper sections of the river, particularly birs of<br />
prey. Ring-tailed lemurs and Verreaux’s sifaka can be seen along the route.<br />
<strong>The</strong> path needs to be cleared and clearly signposted in places.<br />
99
Appendix 12.<br />
Descriptions of <strong>proposed</strong> bird hi<strong>de</strong>s<br />
Hi<strong>de</strong> 1<br />
Surroundings – Located at Antafoky Lake in water hyacinth raft area with views<br />
across lake to opposite hillsi<strong>de</strong>s.<br />
Features – Good potential for crocodile watching.<br />
Disturbance – General lake activity (fishing and transportation by pirogue).<br />
Visibility – Excellent unobstructed views across lake and water hyacinth rafts.<br />
Special hi<strong>de</strong> requirements – approach to hi<strong>de</strong> to be constructed of floating woo<strong>de</strong>n<br />
boards seclu<strong>de</strong>d by existing aquatic vegetation growth. Hi<strong>de</strong> could be floating to<br />
manage changes in lake levels.<br />
Suggested combined activities - Close proximity to walk 4 to valley with natural<br />
amphitheatre, plunge pool and cliff nesting sites for raptors and other bird species.<br />
Birds – Potentially high variety on lake - gallery forest transition. Grey Heron and<br />
Great Egret regularly seen on water hyacinth rafts.<br />
Means of Access – By foot using walk 1 from suggested camp via Antafoky village.<br />
Hi<strong>de</strong> to be accessed via floating boar<strong>de</strong>d walk from lake si<strong>de</strong> path.<br />
Time from <strong>proposed</strong> camp – 1hr30min by foot.<br />
Hi<strong>de</strong> 2<br />
Surroundings – Located at Antafoky Lake – gallery forest transition with view across<br />
water hyacinth raft area and surrounding hillsi<strong>de</strong>s.<br />
Features – Good potential for crocodile watching.<br />
Disturbance – General lake activity (fishing and transportation by pirogue) and lake<br />
path disturbance.<br />
Visibility – Excellent unobstructed views across lake and water hyacinth rafts.<br />
Suggested combined activities - Close proximity to walk 4 to valley with natural<br />
amphitheatre, plunge pool and cliff nesting sites for raptors and other bird species.<br />
Birds – Potentially high variety on lake-gallery forest transition. Grey Heron and<br />
Great Egret regularly seen on water hyacinth rafts.<br />
Means of Access – By foot using walk 1 from suggested camp via Antafoky village.<br />
Time from <strong>proposed</strong> camp – 1hr30min by foot<br />
100
Hi<strong>de</strong> 3<br />
Surroundings – Located in gallery forest commanding views of forest canopy and<br />
lakesi<strong>de</strong> near <strong>proposed</strong> campsite 1.<br />
Features – Good potential for lemur watching.<br />
Special bird features – Canopy views afford good views of Vasa Parrots and other<br />
canopy bird species.<br />
Disturbance – General lake activity (fishing and transportation by pirogue).<br />
Visibility – Good view of canopy and lake.<br />
Suggested combined activities - Walk 4 to sandstone cliff and further walk to valley<br />
with natural amphitheatre, plunge pool and cliff nesting sites for raptors and other bird<br />
species.<br />
Birds – Potentially high variety on lake-gallery forest transition. Cliff location –<br />
potential of seeing Rock Thrushes.<br />
Means of Access – By foot using walk 1 from suggested camp via Antafoky village.<br />
Time from <strong>proposed</strong> camp – 1hr30min by foot.<br />
Hi<strong>de</strong> 4<br />
Surroundings – Located at Antafoky Lake – gallery forest transition at lake bend with<br />
view of opposite scrub hillsi<strong>de</strong> and raft of lake water hyacinths.<br />
Features – Good potential for watching crocodiles and lemurs in forest opposite.<br />
Disturbance – General lake activity (fishing and transportation by pirogue). Path to<br />
hi<strong>de</strong> currently not in regular use as a main village footpath.<br />
Visibility – Some overhanging vegetation and small bushes at the lakesi<strong>de</strong> obscure the<br />
view from the hi<strong>de</strong>.<br />
Suggested combined activities - Close proximity to walk 4 to valley with natural<br />
amphitheatre, plunge pool and cliff nesting sites for raptors and other bird species.<br />
Birds – Potentially high variety on lake-gallery forest transition. Grey Heron and<br />
Great Egret regularly seen on water hyacinth rafts.<br />
Means of Access – By foot using walk 1 from suggested camp via Antafoky village.<br />
Time from <strong>proposed</strong> camp – 1hr50min by foot.<br />
Difficulty – Path to bird hi<strong>de</strong> fairly difficult due to boul<strong>de</strong>rs and narrow path.<br />
101
Hi<strong>de</strong> 5<br />
Surroundings – Located at Antafoky Lake – gallery forest transition at lake bend with<br />
view of opposite scrub hillsi<strong>de</strong> and raft of lake water hyacinths.<br />
Features – Good potential for watching crocodiles and lemurs in forest opposite.<br />
Disturbance – General lake activity (fishing and transportation by pirogue). Path to<br />
hi<strong>de</strong> currently not in regular use as a main village footpath.<br />
Visibility – Some overhanging vegetation and small bushes at the lakesi<strong>de</strong> obscure the<br />
view from the hi<strong>de</strong>.<br />
Suggested combined activities - Close proximity to walk 4 to valley with natural<br />
amphitheatre, plunge pool and cliff nesting sites for raptors and other bird species.<br />
Birds – Potentially high variety on lake-gallery forest transition. Grey Heron and<br />
Great Egret regularly seen on water hyacinth rafts.<br />
Means of Access – By foot using walk 1 from suggested camp via Antafoky village.<br />
Time from <strong>proposed</strong> camp – 1hr50min by foot.<br />
Difficulty – Path to bird hi<strong>de</strong> fairly difficult due to boul<strong>de</strong>rs and narrow path.<br />
Accommodation in hi<strong>de</strong>s:<br />
It is envisaged that small groups (5 people maximum) with gui<strong>de</strong> will visit hi<strong>de</strong>s at<br />
any one time. Each hi<strong>de</strong> should a<strong>de</strong>quately accommodate at least 2-3 persons at any<br />
given time. Seating (e.g. woo<strong>de</strong>n bench) is required in addition to a shelf below<br />
viewing slits for standing telescopes and other equipment. Hi<strong>de</strong>s should be<br />
waterproof.<br />
Hi<strong>de</strong> materials, construction and <strong>de</strong>sign:<br />
Hi<strong>de</strong>s should be constructed with minimal impact to the local environment. It is<br />
suggested that all hi<strong>de</strong>s are built from wood incorporating simple <strong>de</strong>sign features<br />
including a concealed entrance door and viewing slits on three hi<strong>de</strong> si<strong>de</strong>s. Use of local<br />
labour and skills in both construction and maintenance is <strong>de</strong>sirable.<br />
Birdwatching information :<br />
Tourists will require specific information on what bird species can be seen in the area<br />
and may require general information such as bird field gui<strong>de</strong>s to Madagascar. It is<br />
suggested that bird species-lists and a map of local bird features are ma<strong>de</strong> available<br />
for tourists. Local knowledge of the gui<strong>de</strong> will complement these sources of<br />
information.<br />
Birdwatching equipment :<br />
Visitors will have to bring their own equipment, such as binoculars, gui<strong>de</strong>-books,<br />
torches (for night walks), tripods and telescopes.<br />
102
Appendix 13.<br />
Specimen i<strong>de</strong>ntification and <strong>de</strong>position<br />
For i<strong>de</strong>ntification purposes, selected specimens are sent to the following taxonomists<br />
and institutions:<br />
Botany:<br />
Pierre Jules Rataromalaza WWF Toliara WWF Toliara, B.P. 220<br />
Toliara (1), Madagascar<br />
Zoology:<br />
Small mammals:<br />
Steven Goodman Field Museum of Natural History Field Museum of Natural<br />
History, 1400 Roosevelt Road,<br />
Chicago, Illinois 60605, USA<br />
Daniel Rakotondravony Dept of Animal Biology University of Antananarivo,<br />
B.P. 906, Antananarivo,<br />
Madagascar<br />
Reptiles and Amphibians:<br />
Chris J. Raxworthy US Museum of Natural History American Museum of Natural<br />
History, New York, USA<br />
Achille Raselimanana WWF Antananarivo Biodiversity Officer, WWF<br />
Antananarivo<br />
Butterflies:<br />
David Lees and Natural History Museum Dept of Emtomology, <strong>The</strong><br />
Alison Cameron<br />
Natural History Museum,<br />
Cromwell Road, London,<br />
SW7 5BD<br />
Ants and other invertebrates:<br />
Brian Fisher Dept of Entomology California Aca<strong>de</strong>my of<br />
Sciences, Gol<strong>de</strong>n gate<br />
Park, San Francisco,<br />
California 94118, USA<br />
103