Landscape-Based Conservation Approaches in the Developing World: the Case for Managed Elephant Ranges in Sri Lanka

Landscape-based conservation approaches in the

developing world:

The case for Managed Elephant Ranges in Sri Lanka

Sanjay Kalpage

A thesis in fulfilment of the requirements for the degree of

Doctor of Philosophy

Institute of Environmental Studies School of Arts & Social Sciences

October 2013

THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Kalpage

First name: Sanjay Other name/s: Rajnish Perera

Abbreviation for degree as given in the University calendar: PhD

School: Arts & Social Sciences Faculty: Institute of Environmental Studies

Title: Landscape-based conservation approaches in the developing world: the case for Managed Elephant Ranges in Sri Lanka

Abstract 350 words maximum: (PLEASE TYPE)

While establishing protected areas (PAs) has helped conserve the Earth’s diminishing biodiversity, the ‘traditional’ view of conservation underlying this approach—perceiving nature as static, and considering people apart from nature—has compromised PA effectiveness. Moreover, competing land uses often hamper new PA establishment. Consequently, landscape-based approaches could better achieve conservation objectives while catering to human needs. Unfortunately, these approaches have often proved difficult to implement, especially in developing countries. This thesis explores how a landscape-based approach could address Sri Lanka’s most pressing conservation issue: human-elephant conflict resulting in the deaths of about 75 people and 200 elephants annually, and causing substantial property damage. Government policy makers are under tremendous pressure to resolve this conflict—as the endangered Asian elephant is a national icon—but their efforts have largely proved ineffective. Some conservation biologists have proposed a unique conflict-mitigation approach based on research showing that traditional ‘shifting agriculture’ creates optimal elephant habitat, and that differences in timing of land-use by farmers and elephants enables resource-sharing with minimal conflict (irrigated agriculture has been the conflict’s main driver). They recommend establishing ‘Managed Elephant Ranges’ (MERs) outside, but mostly adjacent to, PAs. My thesis focuses on implementing this approach within a particular landscape in southeast Sri Lanka. Given the problem’s complexity, I adopted an interdisciplinary approach—incorporating GIS-based spatial analysis, stakeholder analysis and economic modelling—to demarcate potential MERs, identify stakeholder needs (e.g. of farmers, government officials, NGOs, business-people), and estimate sources of economic value within the landscape (e.g. from nature tourism, carbon credits). Combining my results with lessons from other countries, I have developed a sustainable, economically viable MER implementation plan that meets key stakeholder needs. My approach constitutes a blueprint for implementing MERs in Sri Lanka’s human-elephant conflict areas, while my overall methodology has wider applications in areas of human-wildlife conflict worldwide. Ultimately, my research provides practical applications of natural capital valuation and sustainable natural resource use in a developing country context.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

…………………………………………………………… ……………………………………..……………… ……….……………………...…….… Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

i ORIGINALITY STATEMENT

I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.

Signed ……………………………………………...... …………......

Date ……………………………………………...... …………......

I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.

Signed ……………………………………………...... …………......

Date ……………………………………………...... …………......

AUTHENTICITY STATEMENT

I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.

Signed ……………………………………………...... …………......

Date ……………………………………………...... …………......

This page has intentionally been left blank

iii ABSTRACT

While establishing protected areas (PAs) has helped conserve the Earth’s diminishing biodiversity, the ‘traditional’ view of conservation underlying this approach— perceiving nature as static, and considering people apart from nature—has compromised PA effectiveness. Moreover, competing land uses often hamper new PA establishment. Consequently, landscape-based approaches could better achieve conservation objectives while catering to human needs. Unfortunately, these approaches have often proved difficult to implement, especially in developing countries. This thesis explores how a landscape-based approach could address Sri Lanka’s most pressing conservation issue: human-elephant conflict resulting in the deaths of about 75 people and 200 elephants annually, and causing substantial property damage. Government policy makers are under tremendous pressure to resolve this conflict—as the endangered Asian elephant is a national icon—but their efforts have largely proved ineffective. Some conservation biologists have proposed a unique conflict-mitigation approach based on research showing that traditional ‘shifting agriculture’ creates optimal elephant habitat, and that differences in timing of land-use by farmers and elephants enables resource-sharing with minimal conflict (irrigated agriculture has been the conflict’s main driver). They recommend establishing ‘Managed Elephant Ranges’ (MERs) outside, but mostly adjacent to, PAs. My thesis focuses on implementing this approach within a particular landscape in southeast Sri Lanka. Given the problem’s complexity, I adopted an interdisciplinary approach—incorporating GIS-based spatial analysis, stakeholder analysis and economic modelling—to demarcate potential MERs, identify stakeholder needs (e.g. of farmers, government officials, NGOs, business-people), and estimate sources of economic value within the landscape (e.g. from nature tourism, carbon credits). Combining my results with lessons from other countries, I have developed a sustainable, economically viable MER implementation plan that meets key stakeholder needs. My approach constitutes a blueprint for implementing MERs in Sri Lanka’s human-elephant conflict areas, while my overall methodology has wider applications in areas of human-wildlife conflict worldwide. Ultimately, my research provides practical applications of natural capital valuation and sustainable natural resource use in a developing country context.

This page has intentionally been left blank

v TABLE OF CONTENTS

Acknowledgements ix

List of publications and presentations arising from the writing of thesis xi

Abbreviations and symbols xiii

List of Figures xv

List of Tables xix

Chapter 1: Problem definition and research overview 1

1.1 The global biodiversity crisis and the importance of protected areas 1

1.2 Issues with relying on protected areas to conserve biodiversity and 6 tropical forests

1.3 The nature of the required solution 7

1.4 An alternative approach: landscape-based conservation 15

1.5 Implementing the landscape approach in the developing world: issues 23 and areas for further research

1.6 Focus of research 27

1.7 Frameworks, approaches and methodology 31

1.8 Thesis structure 41

Chapter 2: Overview of Sri Lanka’s biodiversity resources, land-use and 43 conservation-related institutions

2.1 Biophysical context 44

2.2 Socio-economic context 48

2.3 Land-use 51

2.4 Historical context 55

2.5 Conservation-related institutions 71

vi 2.6 Legislative and policy framework 76

2.7 Key conservation-related issues in Sri Lanka 78

2.8 Overview of key points in the chapter 85

Chapter 3: Sri Lanka’s imperilled national icon 87

3.1 The Asian elephant: an introduction 87

3.2 Human-elephant conflict 94

3.3 The Asian elephant in history and culture 95

3.4 Sri Lanka’s Asian elephant subspecies 98

3.5 The elephant within Sri Lanka’s history and culture 101

3.6 The current status of the elephant in Sri Lanka 107

3.7 Sri Lanka’s human-elephant conflict 111

Chapter 4: An approach to mitigate Sri Lanka’s human-elephant conflict 113

4.1 The importance of resolving human-elephant conflict 113

4.2 Measures taken to resolve the conflict and their effectiveness 118

4.3 The problem with current approaches used to resolve human-elephant 124 conflict

4.4 The nature of the required solution 129

4.5 An innovative approach based on recent research on elephant 131 behaviour

4.6 Overview of key points in Chapters 3 and 4 137

Chapter 5: Introduction to case study 139

5.1 Sri Lanka’s southeast landscape 139

5.2 Land use within the study area 151

vii 5.3 Livelihoods of people in the study area 165

5.4 The elephant population and human-elephant conflict within the study 176 area

5.5 Potential Managed Elephant Ranges (MERs) within the study area 179

Chapter 6: Stakeholder analysis 181

6.1 Selecting a method to assess and meet stakeholder needs 181

6.2 Stakeholder analysis theory and methods 182

6.3 Identifying stakeholders within Sri Lanka’s southeast landscape and 187 understanding their role in human-elephant conflict

6.4 Differentiating among and categorising stakeholders 223

6.5 Investigating relationships between stakeholders 225

6.6 Implications of the stakeholder analysis 228

Chapter 7: The economic viability of Managed Elephant Ranges 231

7.1 The economic explanation for deforestation and the need for valuation 231

7.2 The opposition to valuation 242

7.3 The concept of total economic value (TEV) 246

7.4 The components of TEV 257

7.5 Estimating the TEV of natural resources within Sri Lanka’s southeast 266 landscape

7.6 Summary and implications of the economic analysis 288

Chapter 8: Implementing Managed Elephant Ranges 291

8.1 Required characteristics of Managed Elephant Ranges 291

8.2 Potential challenges to implementing Managed Elephant Ranges 323

viii 8.3 Implementing Managed Elephant Ranges 334

8.4 Summary of key points in chapter 342

Chapter 9: Conclusions and implications of research 345

9.1 What is (and has been) Sri Lanka’s approach to conservation, and what 347 are key issues associated with this approach?

9.2 What have been the impacts of Sri Lanka’s ‘traditional’ approach on 348 elephant conservation and human-elephant conflict?

9.3 Could a landscape-based approach help conserve elephants while 349 addressing stakeholder needs in southeast Sri Lanka?

9.4 What lessons could be drawn from this research in the context of Sri 352 Lanka and other developing countries?

9.5 Final thoughts 361

Appendices 363

Appendix 1: Problem definition and research overview 363

Appendix 2: Overview of Sri Lanka’s biodiversity resources, land-use 375 and conservation-related institutions

Appendix 3: Sri Lanka’s imperilled national icon 393

Appendix 4: An approach to mitigate Sri Lanka’s human-elephant 396 conflict

Appendix 5: Introduction to case study 398

Appendix 6: Stakeholder analysis 417

Appendix 7: The economic viability of Managed Elephant Ranges 423

Appendix 8: Implementing Managed Elephant Ranges 445

References 455

ix ACKNOWLEDGEMENTS

A wise person (or possibly even a PhD student) once remarked that: “the trouble with good ideas is that they quickly degenerate into hard work”. My experiences in the past three and a half years have certainly illustrated to me the wisdom contained in these words. I have been extremely fortunate, however, that many people helped ‘lighten my burden’. In particular, I would like to thank:

• My supervisors, John Merson and Daniel Robinson, whose timely, insightful comments on my work, and unstinting encouragement throughout the entire process more than made up for the ‘good ideas’ they gave me from time to time!

• All those who went out of their way to help me with my research in Sri Lanka: Manori, Sumith, Eric, Bandara and Prasanga introduced me to interviewees, helped me obtain required information, and provided valuable perspectives on a range of conservation-related issues in Sri Lanka;

• All those, both in Sri Lanka and elsewhere, who agreed to be interviewed; my special thanks to Rob, Les and Ian in South Africa;

• Chinthaka, who encouraged me to apply for a PhD at UNSW in the first place, and provided me with keen insights and sound advice throughout the past few years;

• My colleagues at IES, who made the entire process so enjoyable, especially Sarah, Alex, Tani, Crelis and Nahid;

• Nancy, who has had the (mis)fortune of reading (almost) everything I have ever written, including every word of this thesis; her incomparable editorial skills and insightful comments made this document much more readable;

• My parents, for their encouragement and inspiration in all my endeavours, and their unwavering belief that I could do anything I put my mind to; and,

• My brother Pravin, my friend Chanaka, and, above all, my wife Thimali, with whom I have spent many unforgettable moments in the wilderness, inspiring me to pursue my interest in conservation and undertake this research; I will always treasure our shared experiences, wide-ranging travels, and thought-provoking discussions (often in the unlikeliest of places and situations).

This page has intentionally been left blank

xi LIST OF PUBLICATIONS AND PRESENTATIONS ARISING FROM THIS THESIS

Kalpage, S., J Merson, and D. Robinson 2012, ‘Landscape-based conservation and sustainable resource-use in the developing world: a case study from Sri Lanka’, in J Merson, R. Cooney and P. Brown (eds), Conservation in a Crowded World: Case Studies from the Asia Pacific, University of New South Wales Press, Sydney, pp. 16-33.

Kalpage S. 2012, ‘Socioeconomic aspects of a landscape-based approach to mitigate Sri Lanka’s human-elephant conflict’, paper presented at the 4th International Wildlife Management Congress, Durban, South Africa, 9 – 12th July.

Kalpage S. 2011, ‘The role of protected areas in climate change mitigation and adaptation in developing countries such as Sri Lanka’, paper presented at the International Conference on Climate Change & Social Issues, Colombo, Sri Lanka 13 – 15th December.

Kalpage S. 2011, ‘Mitigating human-elephant conflict in Sri Lanka: lessons from Southern Africa’, paper presented at the 7th International Wildlife Ranching Symposium, Kimberly, South Africa, 10 – 13th October.

Kalpage S. 2010, ‘Landscape-based conservation and sustainable resource-use: a case study from Sri Lanka’, paper presented at the Conservation through Sustainable Use Workshop, Phuket, Thailand, 10 – 12th July.

xii

This page has intentionally been left blank

xiii ABBREVIATIONS AND SYMBOLS

Abbreviation Explanation

AWF African Wildlife Foundation BBC British Broadcasting Corporation CAMPFIRE Communal Areas Management Programme for Indigenous Resources (of Zimbabwe) CBD Convention on Biological Diversity CBNRM Community Based Natural Resource Management CCR Centre for Conservation Research CDM Clean Development Mechanism CEA Central Environmental Authority (of Sri Lanka) CITES Convention on International Trade in Endangered Species of Wild Fauna and Flora

CO2 Carbon dioxide COP Conference of Parties DWC Department of Wildlife Conservation (of Sri Lanka) EFL Environmental Foundation Limited (Sri Lankan NGO) EIA Environmental Impact Assessment FCPF Forest Carbon Partnership Facility (of the World Bank) FD Forest Department (of Sri Lanka) FFPO Fauna and Flora Protection Ordinance (of Sri Lanka) GDP Gross Domestic Product GIS Geographic Information Systems GNP Gross National Product ha Hectares HEC Human-elephant conflict IPCC Intergovernmental Panel on Climate Change IUCN International Union for the Conservation of Nature JICA Japan International Cooperation Agency kg Kilograms km Kilometres LTTE Liberation Tigers of Tamil Eelam m Metres

xiv Abbreviation Explanation

MERs Managed Elephant Ranges mm Millimetres MOP Muriate of Potash (fertiliser) MW Mega Watts NGO Non-Governmental Organisation NTFP Non-Timber Forest Products NPV Net Present Value OECD Organisation for Economic Cooperation and Development PA Protected Area PREM Poverty Reduction and Economic Management Network (NGO) REDD Reduced Emissions from Deforestation and Degradation Rs. Sri Lankan Rupees SANParks South African National Parks SLAITO Sri Lanka Association of Inbound Tour Operators SLFP Sri Lanka Freedom Party (Sri Lankan political party) TEV Total Economic Value TSP Triple Super Phosphate (fertiliser) TULF Tamil United Liberation Front (Sri Lankan political party) UK United Kingdom UN United Nations UNESCO United Nations Educational, Scientific and Cultural Organization UNFCCC United Nations Framework Convention on Climate Change UNP United National Party (Sri Lankan political party) US United States US$ United States Dollars UV Ultra Violet VHF Very High Frequency WCS Wildlife Conservation Society WDPA World Database on Protected Areas WNPS Wildlife and Nature Protection Society (of Sri Lanka) WTP Willingness to Pay WWF World Wide Fund for Nature

xv LIST OF FIGURES

Figure 1.1 Growth in nationally designated protected areas: 1911 – 2011 8 Figure 1.2 Objective and overall approach 33 Figure 1.3 Oral qualitative methods in human geography 39 Figure 2.1 Major climatic zones and rainfall isohyets in Sri Lanka 45 Figure 2.2 Protected areas in Sri Lanka administered by the Forest 54 Department and the Department of Wildlife Figure 2.3 Growth of Sri Lanka’s protected area system in the forestry and 64 wildlife subsectors Figure 2.4 Distribution of protected areas by size 83 Figure 3.1 The former and present distributions of Asian elephants 91 Figure 3.2 Area of remaining and protected wildlands inside the range of the 93 Asian elephant Figure 3.3 The approximate range of Asian elephants in Sri Lanka 100 Figure 4.1 Elephants’ use of chenas throughout the year 135 Figure 5.1 Sri Lanka’s 24 administrative districts, and the three districts that 141 are the focus of the case study Figure 5.2 District Divisions included in landscape considered in the case 143 study (map) Figure 5.3 Sri Lanka’s three ancient kingdoms in about 300 B.C. (map) 146 Figure 5.4 Sithulpahuwa temple, within Yala National Park (photograph) 148 Figure 5.5 Map of landscape types within my study area in southeast Sri 152 Lanka Figure 5.6 Protected areas within Sri Lanka’s southeast landscape (map) 155 Figure 5.7 Timing of rainfall patterns and rice cultivation in Sri Lanka’s dry 158 zone Figure 5.8 Ripening paddy, Moneragala (photograph) 159 Figure 5.9 Threshing paddy, Hambantota (photograph) 159 Figure 5.10 Timing of rainfall patterns and chena cultivation in Sri Lanka’s 161 dry zone Figure 5.11 Chena plot in the buffer zone of Yala National Park (photograph) 162 Figure 5.12 The area around Yodakandiya town, located close to the Yala 164 National Park (map) Figure 5.13 Percentage of farmers engaged in chena by month 169 Figure 5.14 Harvesting tomatoes in a chena near the Yala National Park 172 (photograph)

xvi Figure 5.15 Livestock herding, Hambantota district (photograph) 175 Figure 5.16 Fishing boats, Ampara district (photograph) 175 Figure 5.17 Potential MER bordering the Yala National Park (map) 179 Figure 5.18 Potential MER bordering the Uda Walawe and Lunugamvehera 180 National Parks (map) Figure 6.1 A schematic representation of stakeholder analysis steps, and 186 associated methods Figure 6.2 Elephants in paddy field, after paddy harvest (photograph) 190 Figure 6.3 Gate to a chena damaged by elephants (photograph) 190 Figure 6.4 The house of Farmer 29, which was damaged by an elephant 191 (photograph) Figure 6.5 Elephant attacks by month in southeast Sri Lanka 192 Figure 6.6 A rudimentary electric fence around a chena (photograph) 194 Figure 6.7 A yaya (group of paddy fields) surrounded by an electric fence 195 (photograph) Figure 6.8 Solar panel and room containing the electric circuit powering an 195 electric fence around a yaya (group of paddy fields) (photograph) Figure 6.9 Banana plantation damaged by elephants in the Thammanna 198 village (photograph) Figure 6.10 Concrete posts holding electric wires damaged by elephants in 198 the Thammanna village (photograph) Figure 6.11 Evidence of the rapid development in Sri Lanka’s southeast 207 landscape (photograph) Figure 6.12 Villagers protest the killing of a woman by an elephant at the 208 entrance to Bundala National Park (photograph) Figure 6.13 Key stakeholder groups, and select sub-groups, in my case study 223 located on the influence-importance matrix Figure 6.14 Actor-linkage matrix for key stakeholders in Sri Lanka’s 226 southeast landscape Figure 7.1 The underlying drivers of deforestation 232 Figure 7.2 Beneficiaries of forest services 234 Figure 7.3 The logic of market-based mechanisms for forest services 236 Figure 7.4 Valuing marginal changes 250 Figure 7.5 The different categories of value of tropical forests (and 252 biodiversity) Figure 7.6 Process for estimating the TEV of land-use for Sri Lanka’s 266 southeast landscape Figure 8.1 Key findings from analyses in Chapters 5, 6 and 7 292 Figure 8.2 Male elephants at the electric fence bordering Uda Walawe 296

xvii National Park, southeast Sri Lanka (photograph) Figure 8.3 Sketch of electric fences around Ketanwewa village in the 300 records of the Farmers’ Association responsible for maintaining them (photograph) Figure 8.4 The fence at the Thammana village that was breached after 302 several years, mainly because villagers allowed elephants opportunities to challenge it (photograph) Figure 8.5 Chaos at a leopard sighting at Yala National Park (photograph) 309 Figure 8.6 Proposed process for implementing MERs 334 Figure 8.7 Proposed organisation for implementing MERs 337

Figures in Appendices

Figure A1 Estimated extinction rates—distant past, recent past and future 363 Figure A2 Sri Lanka’s major vegetation types 375 Figure A3 Results of 2011 census of wild elephants in Sri Lanka 395 Figure A4 Four key stakeholder categories arising from an influence- 418 importance matrix Figure A5 The components of Total Economic Value (TEV) 435

xviii

This page has intentionally been left blank

xix LIST OF TABLES

Table 1.1 IUCN protected area management categories 5 Table 1.2 Characteristics of a more effective approach to conservation 14 Table 1.3 Contrasting approaches to conservation: the traditional approach 21 versus the landscape approach Table 1.4 Multi-disciplinary analyses used in my thesis 33 Table 2.1 Sri Lanka’s population and growth rates 49 Table 2.2 Land ownership in Sri Lanka as of 1991 51 Table 2.3 Area under principal crops and food crops, 2009 52 Table 2.4 Extent of protected areas managed by the Forest Department and 53 the Department of Wildlife Table 2.5 Sri Lanka's closed canopy forest cover 81 Table 3.1 Population estimates of Asian elephants by country 89 Table 3.2 Area of remaining wildlands inside Asian elephant ranges, km2 92 Table 3.3 Examples of the use of the elephant in contemporary Sri Lankan 108 society Table 4.1 Elephant and human deaths in Sri Lanka, 2006 – 2010 114 Table 4.2 Causes of elephant deaths in Sri Lanka, 1998 – 2007 116 Table 4.3 Numbers of elephants within Sri Lanka’s protected area network 126 Table 4.4 Analysis of how the proposed approach corresponds to a 137 landscape approach Table 5.1 Details on District Secretariat Divisions covered in the case study 142 Table 5.2 Land area, population and population densities of Hambantota, 144 Ampara and Moneragala districts, 2010 Table 5.3 Median per capita incomes of people living Ampara, Hambantota 144 and Moneragala districts, 2009/2010 Table 5.4 Poverty indicators for Ampara, Hambantota and Moneragala 145 districts, 2009/2010 Table 5.5 Land-use within my study area in southeast Sri Lanka 151 Table 5.6 Protected areas within the study area 154 Table 5.7 Land-use within study area in southeast Sri Lanka, excluding 156 protected areas Table 5.8 Employed population (as a percentage of working population) by 157 major industry for Moneragala, Hambantota and Ampara districts, 2010 Table 5.9 Select dry zone crops: proportion produced in Hambantota, 160 Ampara and Moneragala for the 2009 season Table 5.10 Annual income estimate for chena farmer households 166

xx Table 5.11 Annual income estimate of households engaged in paddy 166 farming Table 5.12 Incomes of farmers surveyed compared to official poverty lines 168 Table 5.13 Chemical fertiliser use by chena farmers in sample 170 Table 5.14 Pesticide use by chena farmers in sample 170 Table 5.15 Number of years cultivated by chena farmers in sample 171 Table 5.16 Types of small and micro-enterprises interviewees engaged in 173 Table 5.17 Annual income estimate for households that did not engage in 173 chena or rice cultivation Table 6.1 Methods that could be used to conduct each step of the 184 stakeholder analysis Table 6.2 Sources of information by stakeholder group 187 Table 6.3 Farmers’ views on methods—both currently used and those that 197 may be used in the future—to mitigate their elephant-related issues Table 6.4 Livelihood impacts related to the environment, apart from 202 human-elephant conflict Table 6.5 Land ownership of paddy farmers 205 Table 6.6 Farmers’ views of success of the Department of Wildlife in 209 addressing their issues with elephants Table 7.1 Components of Total Economic Value (TEV) 247 Table 7.2 Economic valuation methods 253 Table 7.3 Sources of bias in contingency valuation studies 255 Table 7.4 Land-use within my study area in southeast Sri Lanka 267 Table 7.5 Fertiliser subsidies provided by the Sri Lankan government, 2009 271 Table 7.6 Summary of valuation methods used in economic model 276 Table 7.7 Summary of estimated annual benefits, costs and economic 277 profits from land-use within Sri Lanka’s southeast landscape Table 7.8 Land-use in Scenario C 281 Table 7.9 Estimated annual benefits and costs in three land-use scenarios in 283 Sri Lanka’s southeast landscape Table 7.10 Estimated annual economic profits and required investment in 284 three land-use scenarios Table 8.1 Required characteristics of Managed Elephant Ranges 293 Table 8.2 Responses from foreign tourists who visited Bundala, Uda 310 Walawe and Yala National Parks Table 8.3 Summary of bad and good responses from foreign tourists who 310 visited Bundala, Uda Walawe and Yala National Parks Table 8.4 The direct and indirect economic contribution of tourism in 324

xxi leading nature tourism destinations compared to Sri Lanka in 2010 Table 8.5 ‘Best practices’ associated with the management planning 338 process

Tables in Appendices

Table A1 Numbers of gap species in the current PA network 365 Table A2 IUCN management categories with detailed descriptions 369 Table A3 Acceptability of trade-offs at different spatial scales 372 Table A4 Forest cover in Sri Lanka over time 376 Table A5 Species richness of fauna in Sri Lanka 376 Table A6 Population, surface area and population density in South Asian 377 countries in 2007 Table A7 Percentage of urban population by country, 1950 – 2005, and 378 projections to 2030 Table A8 Life expectancy in South Asian countries, 1950 – 2000 (years) 378 Table A9 Adult literacy rate by sex at age 15 and over in the countries of 379 South Asia Table A10 Development indicators in the countries of South Asia 379 Table A11 Land-use in Sri Lanka, 2008 380 Table A12 Recorded conservation activities by Sri Lankan Kings 380 Table A13 Land alienation by the state, 1953 – 1985 383 Table A14 Extent of land acquired through the 1972 and 1975 land reforms 384 Table A15 Principal recipients of land under the Land Reform Programme, 385 up to October 1976 Table A16 Vision, mission, responsibilities and organisation structure of the 386 Forest Department and the Department of Wildlife Table A17 Key conservation-related NGOs in Sri Lanka 387 Table A18 Selected environmental laws in Sri Lanka and the authorities in 390 charge of supervision, regulation and/or enforcement Table A19 Key international agreements related to the conservation of 391 biodiversity which have been ratified by Sri Lanka Table A20 Examples of human-elephant conflict reported from Asian 393 elephant range states Table A21 Distribution of wild elephants in Sri Lanka according to the 2011 394 elephant census Table A22 Details of elephant drives conducted by the Department of 396 Wildlife by year, 2007 – 2011

xxii Table A23 Elephant translocations conducted by the Department of Wildlife 397 by year, 2007 – 2011 Table A24 Annual cost and number of ‘elephant thunders’ (thunder flashes) 397 distributed by the Department of Wildlife to people living in areas where there is human-elephant conflict Table A25 Annual compensation paid by the Department of Wildlife, 2005 397 – 2010 Table A26 Examples of on-going and planned infrastructure development in 398 southeast Sri Lanka Table A27 Visitor numbers to Sri Lankan parks, 2010 400 Table A28 Visitor revenue from Sri Lankan parks, 2010 400 Table A29 Rice production by district for the 2009/2010 Maha season and 401 the 2010 the Yala season Table A30 Numbers of interviews conducted by stakeholder group 402 Table A31 Details of interviews conducted with farmers 404 Table A32 Details of interviews conducted with government officials 407 Table A33 Details of interviews conducted with tourism companies 408 Table A34 Details of interviews conducted with other locals 409 Table A35 Details of interviews conducted with NGOs 411 Table A36 Details of sample allocation and coverage of World Bank survey 411 Table A37 Composition of households in World Bank survey 412 Table A38 Overall household statistics in World Bank survey 412 Table A39 Comparison between chena and non-chena households in World 413 Bank survey Table A40 Chena cultivation by village in World Bank survey 413 Table A41 Crop types cultivated in chenas in World Bank survey 414 Table A42 Estimates of wild elephant populations contained within several 415 important landscapes Table A43 Distribution of wild elephants in Sri Lanka according to the 2011 416 elephant census Table A44 Urban respondents’ ranking of those who should be major 417 contributors to farmers’ compensation Table A45 Land-use options governments consider when making forest 423 policies Table A46 The value of non-timber forest products 424 Table A47 Tourism/ recreational values in tropical forests 427 Table A48 Indirect use values of forests 428 Table A49 Non-use values of forests 430

xxiii Table A50 The three Kyoto Mechanisms 432 Table A51 Structure of the model corresponding to TEV components 437 Table A52 Specifications of well-constructed fences in the Sri Lankan 445 context Table A53 Potential approaches to achieve sustainable agriculture 446 Table A54 Distribution of tourists, by type and location, in the World Bank 450 tourism survey Table A55 Wildlife species diversity – results for Bundala, Uda Walawe and 451 Yala National Parks, numbers of responses Table A56 Congestion – results for Bundala, Uda Walawe and Yala 451 National Parks, numbers of responses Table A57 Knowledge of guide – results for Bundala, Uda Walawe and 451 Yala National Parks, numbers of responses Table A58 Accommodation – results for Bundala, Uda Walawe and Yala 452 National Parks, numbers of responses Table A59 Diversity of activities – results for Bundala, Uda Walawe and 452 Yala National Parks, numbers of responses Table A60 Friendliness of staff – results for Bundala, Uda Walawe and Yala 452 National Parks, numbers of responses Table A61 Transport experience – results for Bundala, Uda Walawe and 453 Yala National Parks, numbers of responses Table A62 Value for money – results for Bundala, Uda Walawe and Yala 453 National Parks, numbers of responses Table A63 Overall experience – results for Bundala, Uda Walawe and Yala 453 National Parks, numbers of responses Table A64 Preferences of eco tourists as revealed by survey commissioned 454 by the Sri Lanka Tourist Board

xxiv Chapter 1: Problem definition and research overview

1.1. The global biodiversity crisis and the importance of protected areas

Life on Earth is facing a crisis, as biological diversity (or ‘biodiversity’)1 is being threatened with a major pulse of extinction. According to Dirzo and Raven (2003) extinction rates for the past hundred years have been at least several hundred times historical values. While extinction is the ultimate fate of all species—there have been five significant extinction events in the past—what distinguishes the current event is its severity and underlying causes (Dirzo and Raven 2003). The number of species lost in previous extinctions was probably much lower2, and the underlying causes were environmental (Dirzo and Raven 2003). In contrast, the current spasm of extinction is largely driven by anthropogenic factors, which are “implicated in a 100 to 10,000 fold increase in the species extinction rate expected from gradual environmental change, newly established competitive interactions, and occasional change catastrophes” (Bradshaw et al. 2009; p.80). An alarming portion of remaining species is threatened with extinction including some 12% of bird species, 23% of mammals and about a third of amphibians (Stuart et al. 2004; Millennium Ecosystem Assessment 2005). Consequently, the projected future extinction rate is ten times higher than the current rate (Millennium Ecosystem Assessment 2005; p.4)3.

Meffe and Carroll (1997) identified five broad drivers of biodiversity loss: habitat destruction, introduced species, overexploitation, pollution and synergistic influences. Global climate change is an example of the last factor, as a few degrees of temperature change or slight changes in rainfall patterns could be especially harmful for species already stressed through other human activities (Rustad 2008). Unfortunately, these effects are likely to be increasingly significant, given predictions of rising temperatures, and more numerous and severe extreme weather events (Rustad 2008).

1 Plants, animals, fungi and microorganisms, their genetic variation, and the ecosystems of which they are a part. 2 Partly due to a lower number of total species in the past (Sri Lanka Tourist Board 2003; p.8-11). 3 More details are provided in Appendix 1.

1 Global biodiversity is not distributed evenly. According to Ninan (2007), half of all vertebrates, 60% of known plant species and possibly 90% of the world’s total terrestrial species are found in tropical forests. Although estimates of the number and proportion of contained species vary, there is scientific consensus that tropical forests, which cover about 14% of the Earth’s surface, contain an exceptional wealth of biodiversity (Ninan 2007; Sodhi 2008; Bradshaw et al. 2009). Furthermore, tropical forests contain high levels of endemic species, and constitute the majority of ‘biodiversity hotspots’4 (Dirzo and Raven 2003; Schmitt et al. 2009).

Unfortunately, tropical forests are being lost rapidly. Since 1980, global forest cover is estimated to have declined by 225 million ha per year due to human action; an estimated 15 million ha is being lost annually, an area the size of Bangladesh (Eliasch 2008; Bradshaw et al. 2009). There is a significant risk of extinction due to tropical deforestation given the high concentration of threatened species they contain. Accordingly, an analysis of the IUCN5 Red list of threatened species determined that a large majority of threatened species are in tropical countries (Dirzo and Raven 2003). For example, Indonesia tops the list of countries with high numbers of threatened mammal species, followed by India, Brazil, China, and Mexico (Dirzo and Raven 2003).

The consequences of biodiversity and tropical forest loss

This rapid loss of biodiversity and tropical forests has been termed a ‘crisis’ given their direct contribution to human well-being through provisioning (e.g. food, freshwater, wood, fuel), regulating (e.g. climate regulation, disease regulation, water purification) and cultural (e.g. aesthetic, spiritual, educational, recreational) means (Millennium Ecosystem Assessment 2005). Forests also provide indirect supporting benefits such as nutrient recycling, soil formation and primary production. These services—which are mostly ignored in conventional economic measures—have been valued at about US$ 33 trillion per year, or twice the global Gross National Product6 (Constanza et al. 1997; p.253).

4 Areas that have high levels of species endemism. 5 The International Union for the Conservation of Nature. 6 (Dirzo and Raven 2003) estimate that the value of the world’s ecosystem services is about US$ 33 trillion per year, compared to world GNP (in 1997) of US$ 18 trillion per year.

2 The continuing loss of biodiversity and tropical forests would have severe negative consequences for the Earth’s ecosystems in general, and humans in particular. For example, the loss of topsoil due to deforestation could reduce rice output by 1.5 million tons per year, an amount that would feed up to 15 million people per year (Bradshaw et al. 2009). Tropical forest loss is currently responsible for roughly one-fifth of all human carbon emissions (which contribute to climate change), more than the entire global transportation sector (Laurance 2008). Habitat loss could also facilitate the spread of some diseases—deforestation seems to result in an increase in the distribution of mosquitoes with concomitant increases in malaria transmission (Sodhi 2008). Poor people in developing countries are often severely affected by biodiversity and forest decline, given their close dependence on these resources which often represent an ‘insurance policy’ during times of stress such as natural or human-induced disasters (Ninan 2007).

Given the significant benefits of biodiversity and tropical forests, there is an urgent need to stem their decline. This urgency was reflected during the target setting process for the Millennium Development Goals—Target 7B for the goal ‘ensure environmental sustainability’ is to “reduce biodiversity loss, achieving, by 2010, a significant reduction in the rate of loss”, and a key indicator for monitoring this progress is ‘the proportion of land area covered by forest’ (UN Statistics Division 2008).

The significance of protected areas in reducing biodiversity and tropical forest loss

Since the late 19th century, protected areas (PAs) such as nature reserves and national parks have been a key component of biodiversity conservation efforts (Bengtsson et al. 2003). According to Hole et al. (2009; p.420) “networks of PAs remain our single most valuable resource for conserving global biodiversity”.

The 130,709 PAs listed in the World Database on Protected Areas encompass 24.2 million square kilometres (UNEP-WCMC 2012)7. Of these, terrestrial PAs cover 16.2

7 The WDPA is only a record of PAs created by governments and hence diverse forms of community- based conservation are not included. Recent estimates suggest that forests under community conservation

3 million square kilometres, while the rest are marine PAs (UNEP-WCMC 2012). Globally, 12 per cent of national territories, comprising both terrestrial and marine areas, are covered by nationally designated PAs (UNEP-WCMC 2012). Terrestrial PA coverage reaches 12.7 per cent, but the marine environment has received much less attention—PAs cover 7.2 per cent of the world’s territorial seas and less than 1 per cent of high seas (UNEP-WCMC 2012).

The global PA network has grown rapidly over past decades. From 1900 to 1950 about six hundred official PAs were created worldwide, but by 1960, there were almost a thousand; as pointed out above, there are now over 130,000 PAs worldwide (Dowie 2009; p.20). Some countries have made significant additions to their PA networks in recent years. For example, Gabon established 13 national parks comprising 11 per cent of the country in 2002, while the Democratic Republic of Congo established a vast rainforest reserve in 2007, which, at 30,570 km2, is approximately the size of Belgium (Brockington et al. 2008; UNEP-WCMC 2008).

Protected area categorisation

The International Union for the Conservation of Nature (IUCN) plays a global leadership role in defining different types of PAs, and influencing how PA systems are developed and managed (Locke and Dearden 2005). It is a confederation of about 1,300 government and non-governmental organizations, including the World Wide Fund for Nature, the World Bank, the United Nations, and the Organization for Economic Cooperation and Development. The IUCN defines a PA as a “clearly defined geographical space, recognised, dedicated and managed, through legal or other effective means, to achieve the long-term conservation of nature with associated ecosystem services and cultural values” (Dudley 2008; p.8). As illustrated in Table 1.1, this definition encompasses different types of PAs with varying objectives.

on Africa, Asia and the Americas cover 3.7 million km2 and are at least as extensive as forest PAs Constanza et al. (1997; p.253).

4 Table 1.1: IUCN protected area management categories (from UNEP-WCMC 2008; p.5)

Category Title Managed for

Ia Strict Nature Reserve Science

Ib Wilderness Area Wilderness protection

II National Park Ecosystem protection and recreation

III Natural Monument Conservation of specific natural features

IV Habitat/Species Management Area Conservation through management intervention

V Protected Landscape/ Seascape Landscape/seascape conservation and recreation

VI Managed Resource Protected Area Sustainable use of natural ecosystems

These definitions have evolved over past decades. At its inception in 1948, the IUPN8 advocated the protection of pristine landscapes for scientific, aesthetic, recreational and educational purposes; it listed three PA categories in 1966: national parks, scientific reserves and national monuments (Kalamandeen and Gillson 2007; Brockington et al. 2008). However, expanded categories were proposed in the early 1970s, recognising people’s roles in landscapes, reflected in categories V and VI in Table 19 (Weeks and Mehta 2004).

Having explored the important role of PAs in global efforts to address the biodiversity crisis, I now consider the extent to which they have accomplished their objectives.

8 The International Union for the Protection of Nature, which became the IUCN in 1956. 9 These categories were finalized in 1992 (West and Brockington 2006).

5 1.2. Issues with relying on protected areas to conserve biodiversity and tropical forests

According to Margules and Pressey (2000), the extent to which PAs fulfil their role of protecting biodiversity depends on their success in meeting two objectives: (1) ‘representativeness’, or “the need for reserves to represent, or sample, the full variety of biodiversity, ideally at all levels of organisation”; and (2) ‘persistence’, the ability of PAs, once established, “to promote the long-term survival of species and other elements of biodiversity they contain by maintaining natural processes and viable populations by excluding threats” (Margules and Pressey 2000; p.243).

Several studies have revealed that existing PA networks contain significant gaps in their coverage—in other words, they lack representativeness. For example, Rodrigues et al. (2004) concluded from a study of terrestrial vertebrates that at least 12% of species assessed were not represented in any PA10. Other research has revealed significant gaps in coverage of biodiversity-rich habitats, such as ‘biomes’ defined by the WWF11. According to a study by the United Nations Environmental Programme, a significant portion of WWF biomes, especially those in tropical realms, have levels of protection below 10 per cent, the minimum target recommended by the Convention on Biological Diversity12 (UNEP-WCMC 2008; p.21). The main reason for these gaps is the general lack of systematic conservation planning and the bias toward establishing PAs in areas lacking commercial potential (Margules and Pressey 2000; Pressey et al. 2002). Moreover, environmental policy-makers sometimes focus on protecting remote, rugged, ‘wilderness’ areas, without considering their significance from a biodiversity viewpoint (Bengtsson et al. 2003; p.389).

Areas surrounding PAs are key to the ‘persistence’ value of these PAs (Margules and Pressey 2000). According to Margules and Pressey (2000; p.248) “if reserves become remnants of natural habitat surrounded by alien habitat such as cropland or pasture, changes brought about by isolation and exposure have implications for the persistence

10 This study is considered in more detail in Appendix 1. 11 The WWF has divided the terrestrial world into biogeographic realms which are “geographical regions where distinctive assemblages of plants and animals occur” and 14 biomes within these realms according to major plant communities found in them (Weeks and Mehta 2004). 12 The CBD target is that at least 10 per cent of each of the world’s ecological regions [should be] effectively conserved’ and ‘areas of particular importance to biodiversity [should be] protected’ (UNEP- WCMC 2008; p.20).

6 of species within them”. Laurance et al. (2012) conclude that over half the tropical PAs included in their comprehensive study “are experiencing an erosion of biodiversity that is often alarmingly widespread taxonomically and functionally”, i.e. their ability to safeguard the ‘persistence’ of species has been severely eroded (Laurance et al. 2012; p.290). The authors believe that most threats to PA persistence originated from areas immediately outside reserves since “tropical PAs are often intimately linked ecologically to their surrounding habitats, and a failure to stem broad-scale loss and degradation of such habitats could sharply increase the likelihood of serious biodiversity declines” (Laurance et al. 2012; p.290).

Global climate change would further reduce PA effectiveness in terms of both representativeness and biodiversity persistence. Hannah et al. (2007) predict that warmer temperatures would drive species from reserves, reducing PA representativeness. Climate change would also impact PA persistence by causing a range of stresses—invasive and pest species better adapted to the new climate may move in, competing for resources and causing disease outbreaks; fire and floods may also become more frequent (IPCC 2007; Mansourian et al. 2009).

1.3. The nature of the required solution

Having identified deficiencies within the global PA network, I now consider how these issues could be addressed. First, I examine solutions proposed by the ‘traditional’ approach to conservation that was instrumental in creating this network. However, as the ‘traditional’ approach itself may be problematic, I look at the assumptions behind it, to identify the parameters of an approach that would achieve better conservation outcomes.

The solution according to the ‘traditional’ conservation perspective

According to the ‘traditional’ conservation perspective, the solution to deficiencies in the current PA system—in terms of lack of ‘representativeness’ and threats to ‘persistence’—would be to expand the PA network. An example is the Convention on Biological Diversity’s Programme of Work on Protected Areas, which, in 2004, urged

7 “the greater expansion of the PA network across the globe to secure long-term representativeness of ecosystems and help species adapt to climate change” (Mansourian et al. 2009; p.67). This type of thinking has contributed to rapid growth in the area under protection through nationally designated PAs, especially in the past few decades (Figure 1.1).

25,000,000

Cumulative Terrestrial Area 22,500,000 Cumulative Marine Area 20,000,000 Cumulative Total Area ) 2 17,500,000

15,000,000

12,500,000

10,000,000

7,500,000

5,000,000 Total area protected (km

2,500,000

0 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001 2011 1916 1926 1936 1946 1956 1966 1976 1986 1996 2006 Year

Figure 1.1: Growth in nationally designated protected areas: 1911 – 2011 (from UNEP- WCMC 2012)

PAs now constitute one of the most significant forms of human land use on the planet (Chape et al. 2005). The land under conservation is greater than the entire landmass of Africa, and equal to half the planet’s endowment of cultivated land (Dowie 2009; p.20). There has also been significant expenditure on PAs, especially by international donor agencies. For example, from 1991 to 2006, the Global Environment Facility and its funding partners alone distributed about US$ 2 billion to support PAs in Africa and Latin America (Wittemyer et al. 2008). Overall, it has been estimated that biodiversity-

8 related aid comprises almost 3% of ‘official development assistance’13, representing about US$ 3 billion per year (Biodiversity Indicators Partnership 2010; p.2).

Issues with the ‘traditional’ conservation perspective

It is becoming increasingly difficult for governments—especially those in developing countries—to justify drastically increased expenditures on PAs, given numerous other demands on land (e.g. for agriculture, landless rural citizens) and resources (e.g. poverty alleviation programmes, improving health care) (Margules and Pressey 2000). More fundamentally, deficiencies within the traditional approach itself imply that simply expanding the current PA network would not effectively protect biodiversity. Critics have raised potential problems with two fundamental concepts that underlie the traditional approach: (i) the idea of nature as static; and (ii) the perception that people are apart from nature. I now examine each of these concepts, exploring their impact on PA effectiveness.

The idea of nature as static

An underlying assumption of traditional conservation approaches is the idea that nature exists at or near a static equilibrium condition, and in order to conserve nature, it must be maintained in (or returned to) an unchanging ‘pristine’ or ‘original' condition (Gunderson and Holling 2002; Bengtsson et al. 2003; Fisher et al. 2008). For instance, Shultis and Way (2006) characterise conservation efforts in early national parks as “focused on the preservation of specific species in what were conceived as primordial, untouched wilderness areas, landscapes characterised by relatively static, steady-state environments through linear, ordered succession” (Shultis and Way 2006; p.224). Recent research has revealed significant, naturally occurring, perturbations within most ecosystems; moreover, because ecological responses to these disturbances are difficult, if not impossible, to predict, it becomes impossible to define a ‘natural’ or ‘native’

13 This indicator measures measures aid contributions via the Development Assistance Committee (DAC), a principal body of the OECD which deals with issues related to co-operation with developing countries. The DAC is an international forum of 24 members: 23 donor governments and the European Commission (UNEP-WCMC 2008).

9 ecosystem (Shultis and Way 2006). Furthermore, although traditional conservation techniques attempt to prevent disturbances such as fire, it is now evident that these disturbances often comprise intrinsic ecosystem processes, rendering such efforts counterproductive, as described by Bengtsson et al. (2003; p.390):

The classical view of a single equilibrium in nature has until recently dominated conservation and management efforts, especially in government agencies, but also in natural resource management research. It has been reflected in, for example, rules such as fixed sustainable yields, and popular notions such as the balance of nature. It has led to what has been referred to as the pathology of natural resource management. Such management tends to apply strategies that aim at preventing or excluding disturbances. Since disturbances usually are intrinsic parts of ecosystems, it will rarely be possible to exclude them. Some, such as fire and flooding, may accumulate to large temporal and spatial scales and thereby cause large-scale impacts. Others happen anyway, for example, windstorms and pathogens. Recent ecological understanding of ecosystem dynamics across scales, and how to relate to such dynamics, is critical for successful management of nature reserves and their biological diversity. The view of nature as static disregarded important interactions between ‘wilderness areas’ and the landscape surrounding them (Bengtsson et al. 2003). As a result, landscape management in general, and conservation in particular, has historically been approached from the scientific tradition of reductionist thinking, where each landscape component was considered in isolation (Fisher et al. 2008). Reductionism is a way of understanding complex things (e.g. objects, phenomena, theories) by breaking them up into a set of small, easy-to-understand elements. This approach has been applied widely to land use planning (Maginnis et al. 2004; Fisher et al. 2008). Reductionist land-use planning involves discretely and rigidly compartmentalising land according to its functionality or use. For example, in certain areas food is grown using intensive agricultural techniques, while in other areas timber is cultivated in large plantations, and there is no overlap between the two (Maginnis et al. 2004; Tscharntke et al. 2005). Conservation has been approached in a similar manner. Wildlife has been corralled into ‘protected areas’ chosen for a variety of reasons— sometimes on the basis of biological importance, but often because the land had previously been protected as a hunting reserve, or was simply of no interest to anyone with power or influence (Maginnis et al. 2004). A crude, but not inaccurate, characterisation of this model of land-use compartmentalisation is that “wildlife is removed from agricultural areas and (poor) people removed from PAs” (Maginnis et al. 2004; p.322).

10 Reductionist approaches to land management have resulted in many PAs becoming “islands in intensively managed landscapes” (Siegfried et al. 1998; Bengtsson et al. 2003; p.389). Land-use practices within these landscapes may negatively impact biodiversity ‘persistence’ within PAs. For example, agricultural pesticides may enter reserves through natural processes such as the water cycle, or invasive species could spread into PAs from surrounding areas (Harvey et al. 2008; Chazdon et al. 2009). Chazdon et al. (2009; p.142-143) elaborate further:

The fates of biodiversity in PAs and surrounding landscapes are inextricably linked. Most PAs in tropical regions are embedded within a matrix of heterogeneous land uses and are often directly or indirectly affected by forest fragmentation, road construction, agrochemicals, hunting, cattle grazing, agricultural incursions, fire, invasive species, over- harvest of non-timber forest products, logging, and mining. These human activities often threaten species in PAs. On the other hand, certain types of agriculture, agroforestry, fallow vegetation, and forest patches surrounding PAs can support significant levels of biodiversity, while also providing valuable ecosystem services, such as carbon sequestration and hydrological protection. In summary, the view that nature is static has failed to incorporate important ecosystem processes into natural resource management. It has led to reductionist approaches to policy-making, which do not incorporate landscape-level dynamics within land-use decisions, especially those related to PAs and areas surrounding them. These two factors, in turn, have negatively impacted biodiversity and ecosystem conservation.

The perception that people are apart from nature.

Supporters of the traditional approach to conservation have typically assumed that humans should be excluded from PAs because they have a negative influence on the landscape. Their conservation ethic stems from their desire to preserve wild places for their intrinsic and spiritual value, and natural beauty (Kalamandeen and Gillson 2007; p.167). Brockington et al. (2008; p.20) believe that “the desire to protect wilderness from encroachment and human interference has been one of the most powerful forces behind PA establishment”.

The motivation to separate people from nature has sometimes led to the physical displacement of people, which, according to some recent estimates have been in the

11 range of 10 – 20 million people globally to date (Agarwal and Redford 2009; Dowie 2009). However, data on actual numbers of evictions are scarce, because governments and other agencies have historically not collected such information, possibly because it is likely to be extremely sensitive (Brockington et al. 2008; p.74). Consequently, “nearly all of what we have to say about the overall magnitude of conservation-induced displacement depends on heroic extrapolation” (Agarwal and Redford 2009; p.4). Nevertheless, it is widely accepted that PAs have sometimes caused the displacement of local communities—often it is the magnitude that is disputed, given the highly political nature of the issue (Agarwal and Redford 2009).

PAs could also adversely affect local people by denying them ‘customary’ access to resources (Fisher et al. 2008; p.9). This raises issues of equity, since as Adams et al. (2004; p.1146) point out “the eviction of former occupiers or right holders in land and resources, can cause the exacerbation of poverty as well as the contravention of legal or human rights”. For example the Karruyu—a nomadic pastoral group in Ethiopia—lost access to their traditional grazing grounds through the creation of Awash National Park; this disrupted their way life and has meant that they are now caught in a “permanent food crisis” (Borrini-Feyerabend et al. 2004; p.5).

In addition to raising equity and human rights concerns, marginalisation of local communities may be short-sighted from a conservation perspective, because locals who are denied customary access to natural resources often end up contributing to their destruction (Brockington et al. 2008). In Uganda, for example, villagers “set about killing as much wildlife as they could to try and avert the re-creation of the Mburo National Park” (West and Brockington 2006; p.614). In India, denial of traditional access to forest products to villagers living near a tiger reserve led to increased poaching and other illegal activities such as timber extraction within the park (Nagendra et al. 2006). Given criticisms of the traditional approach to conservation, the IUCN created two new PA categories in 1992—V and VI—that captured elements of the interaction between humans and nature14 (Phillips 2002; Dudley 2008).

The problematic assumptions behind the traditional approach to conservation are evident in Yellowstone, the USA’s first national park. Yellowstone was the model for many subsequent parks in the late 19th and early 20th centuries in places like Australia,

14 These two categories are described in more detail in Appendix 1.

12 New Zealand and sub-Saharan Africa, and concepts embodied in this model influence PA establishment even to this day (Greiber et al. 2009; p.89). According to its founders, Yellowstone would enable the American public to enjoy its ‘unspoilt wilderness’, free from human interference (Kalamandeen and Gillson 2007; p.167). When the park was first established, it was a square block of 810,000 ha, an area deemed sufficient to preserve the ‘wilderness’ it contained ‘forever’ (Brockington et al. 2008). However, despite the park’s relatively large size, it quickly became apparent that it did not cover ranges of large predators such as grizzly bears; consequently, its boundaries had to be extended several times, but this has often been controversial, given rapid population growth in areas surrounding the park (Maginnis et al. 2004). Furthermore, the park’s founders blatantly ignored evidence that Native Americans had used the area for thousands of years; they relocated several tribes to nearby reservations, believing that Native Americans, hunting ‘illegally’ in the park, would scare away tourists (Kalamandeen and Gillson 2007; Dowie 2009)15.

Therefore, the entity claiming to be the ‘world’s first national park’16 is a stark reminder of the problematic assumptions behind the ‘traditional’ conservation approach. The problem isolation approach requiring the separation of ‘pristine nature’ from human influence failed to recognise that park ecosystems are not self-contained, static entities, but instead interact dynamically with surrounding areas. Given their desire to maintain the fiction of ‘unspoilt nature’, the park’s founders felt compelled to ignore that people had been an integral part of this purported ‘wilderness’ for thousands of years. Given these issues with the ‘Yellowstone model’, a new approach is needed to conserve biodiversity more effectively (Table 1.2).

15 The ‘Yellowstone model’ is considered in more detail in Appendix 1. 16 As pointed out in Appendix 1, this claim is false.

13 Table 1.2: Characteristics of a more effective approach to conservation

Assumptions behind Consequences/ issues Characteristics of an approach the ‘traditional’ that would address these issues approach

Nature is static (ideal • Failure to incorporate • Understand dynamic ecosystem of ‘pristine important ecosystem processes both within PAs and wilderness’) processes into natural across the larger landscape resource management (especially between PAs and

• Reductionist approach to surrounding areas) policy-making, which does • Take these processes into not incorporate important account when making land-use landscape-level dynamics decisions into land-use decisions

People are apart from • Physical displacement of • Clearly acknowledge that nature people people are part of the • Locals deprived of landscape, and understand their

‘customary access’ to role within it resources • Take steps to encourage locals • The above two factors result (and other stakeholders) to be in hostility among locals partners in conservation and to towards PAs, ultimately sustainably use components of compromising conservation biodiversity (plants, animals objectives and their parts)

The third column in Table 1.2 outlines the type of conservation approach needed based on recent scientific research, and socio-economic realities in developing countries17. I now present an alternative approach that meets these conditions: the landscape-based approach to conservation.

17 That contain most of the world’s biodiversity.

14 1.4. An alternative approach: landscape-based conservation

Some conservationists—along with experts from related fields such as natural resources management and land-management—propose a landscape-based approach to address deficiencies in the traditional approach to conservation (Sanderson et al. 2002; Morrison et al. 2009; Velázquez et al. 2009; Martín-López et al. 2011). This holistic approach not only situates PAs within the larger landscape, but also explores how areas outside PAs could be utilized to meet both conservation goals and human needs18.

Explanation of the terms ‘landscape approach’ and ‘landscape’

Sanderson et al. (2002; p.41) characterise a landscape-based approach as follows: Parks and reserves can effectively protect some elements of biodiversity and contribute to the conservation of nature, but often, strict protection is not possible over sufficient areas. As a result virtually all PAs are embedded in a landscape in which natural resource exploitation of multiple types occurs. Effective biodiversity conservation must therefore integrate use and protection across the landscape. For landscape scale conservation to be socially as well as ecologically sustainable, these strategies must succeed in a mosaic of different land uses that not only conserves biodiversity, but also allows people to make a living. Such a conservation strategy must integrate land uses as diverse as parks, large forestry concessions, indigenous management areas, extractive reserves, agricultural zones, and urban areas.

In this context, it is important to define the term ‘landscape’, since, as Phillips (2002; p.5) points out: “Landscape is a difficult word. It has many meanings and is interpreted differently by different people”. While physical geographers use the term quite precisely to describe a particular area based on enduring features, ecologists have developed a broader approach to ecosystem analysis based on ‘landscape ecology’ (Maginnis et al. 2004). Yet, Maginnis et al. (2004) point out that common usage depends more on cultural attributes, often reflecting the history of the land. Therefore, I will use the following definition of landscape proposed by Sayer et al. (2007; p.2679): ‘Landscape’ should be used to describe a geographical construct that includes not only the biophysical components of an area but also social, political, psychological and other components of that system. A landscape

18 It is important to keep in mind that trade-offs would have to be made between these objectives; a key issue to be addressed is how these trade-offs should be made.

15 is not therefore the same as an ecosystem—the latter is a predominantly biophysical construct. Consequently, landscapes reflect “self-definitions of people within a particular cultural context” and “this applies as much to a western conservationist as it does to a Kalahari San hunter” (Maginnis et al. 2004; p.331). This ambiguity however, does not negate the concept’s usefulness. As Maginnis et al. (2004; p.331) point out:

The real value of the landscape concept is that it allows one group of stakeholders to perceive how another group of stakeholders regards the space within which they live and work (their particular landscape), and this provides a context for understanding the landscape functions that they are most concerned about. In other words, it provides a framework for helping groups of stakeholders agree on how to balance the trade-offs inherent in land use. Landscapes defined by conservationists often overlap with several livelihood/cultural landscapes conforming to the realities of other key stakeholders. Such landscapes could be very limited in size, being composed of farms and woodlands within a valley bottom, or very extensive, such as the annual range used by nomadic pastoralists; they could even be defined on the basis of spirituality, such as Central Australia’s Uluru region (Maginnis et al. 2004). Livelihood/cultural landscapes are often, but not always, smaller than conservation landscapes (Maginnis et al. 2004).

Key characteristics of the landscape approach

The landscape approach appreciates the inability of PA networks alone to protect a significant portion of global biodiversity, and attempts to address major deficiencies of the traditional approach to conservation. In particular, instead of operating from the premise that nature is static, and rigidly compartmentalising types of land-use, it considers the relationships among land-uses across the entire landscape. Instead of viewing people as apart from nature and necessarily detrimental to it, the approach explicitly considers interactions between people and nature. I now explore these facets in further detail.

16 Considering the overall ecosystem

Ecosystems are complex, non-linear and dynamic—qualities that have eluded traditional scientific solutions, especially given their goal of finding the ‘right answer’ (Waltner-Toews et al. 2008). Complexity defies linear logic given characteristics such as self-organisation19 and feedback loops (Bengtsson et al. 2003). An example is the way some systems respond to disturbances. Given the buffering effect of feedback loops, the system can ‘absorb’ disturbances up to a threshold or tipping point with little apparent change. However, a slight change beyond this threshold would result in dramatic, often unpredictable, changes to the system (Waltner-Toews et al. 2008). Waltner-Toews et al. (2008) illustrate this phenomenon by describing effects of acid rain on a lake. Although acidic precipitation increased gradually over decades, the pH20 of lake water did not change substantially given a series of feedback loops. However, once the volume of precipitation exceeded the compensatory capacity of these loops, the system flipped to a different organisational state and became a ‘dead’ acidified lake.

Bengtsson et al. (2003) suggest that ‘ecological memory’ is an important component of a system’s ecological resilience, which consists of the network of species in the system, the dynamic interactions between each other and the environment, and the combination of structures that make reorganisation after disturbance possible (Bengtsson et al. 2003; p.389). Natural disturbance regimes could play an important part in system resilience. For example, some forests are adapted to fire, and ‘normal’ fires release nutrients, and stimulate dormant seedlings. Suppressing forest fires—as has been often done by environmental managers who depend on linear, causal models—hampers the rejuvenation process, allows fuel to accumulate, and sets the stage for huge, destructive conflagrations (Waltner-Toews et al. 2008; p.5).

Therefore, the entire ecosystem should be considered at the landscape-level, unlike in the traditional approach where, often, parts of ecosystems are considered in isolation. As described previously, traditional reductionist approaches to conservation have often resulted in PAs being isolated within landscapes that are used in a manner detrimental

19 Self-organisation is a phenomenon that gives us the sense that a system has an ‘identity’ of its own. An example is a school of fish or a herd of wildebeest—the entire school or herd seems to move of its own accord. Understanding this requires us to understand the relationships between individuals, rather than the behaviour of each individual itself. Self-organisation is about how coherent patterns of relationships are internally structured and develop over time (Biodiversity Indicators Partnership 2010; p.1). 20 A measure of acidity.

17 to biodiversity within these PAs. For effective conservation, therefore, it is important to gain an understanding of the overall ecosystem of which the PA is a part, and of resource-use by species within the ecosystem (Bengtsson et al. 2003). Sanderson et al. (2002; p.42) elaborate further:

Because nature is not homogenous, most organisms, at one scale or another, depend on heterogeneity for their survival. Accordingly, neither people, nor animals, use the landscape homogenously, but rather cluster their activities differentially among landscape elements… Some organisms use multiple habitats, depend on temporary concentrations of resources, and move around the landscape in non-random ways and according to spatially and temporally distributed requirements. Protecting organisms means protecting their ability to meet their requirements, inside and outside parks, among different human uses, and in spite of human economic and cultural requirements being met in the same times and places. In short, conservation planning requires a landscape sensibility. This type of landscape-level view is especially important in the case of wide-ranging species such as elephants, tigers and bears, which often use natural resources both within and outside PAs (Didier et al. 2009).

The landscape approach also asserts that land-use-related decisions should be made at the proper scale. As pointed out previously, developing-country governments face numerous demands on land, stemming from the needs of their growing populations. Therefore, government policy-makers have to make trade-offs among land-use types; the issue is how these trade-offs should be made, as explained by Maginnis et al. (2004; p.329):

Attempting to balance trade-offs at a relatively small scale—for example, to expect one small area of forest to fulfil all possible functions ranging from conservation of the most sensitive species to timber production— precludes the benefits of specialization and often results in suboptimal conservation and human development results. Conversely, allowing specialization of land uses over large areas can produce uniform, sterile landscapes. For example a 2,000-hectare forest plantation certified by the Forest Stewardship Council standards of good management is likely to be an asset to the landscape; however, this does not mean that the landscape will be improved further by establishing block upon block of similarly certified plantations on adjacent sites. The extent to which trade-offs could be made at different scales without undermining environmental functions is considered in more detail in Appendix 1.

18 Considering the needs of key stakeholders

As pointed out above, the traditional approach to conservation could have highly negative effects on local communities and ultimately the environment. In contrast, the landscape approach acknowledges that local people are an integral part of the landscape and explicitly considers how they use the landscape’s natural resources, exploring ways to make this use sustainable (Brown 2004). Moreover, it facilitates processes that help locals derive benefits from biodiversity resources, thereby incentivising local conservation initiatives (Sanderson et al. 2002; McNeely and Mainka 2009; Velázquez et al. 2009). This is in accordance with principles contained in the IUCN Policy Statement on Sustainable Use of Wild Living Resources, which states that the “use of wild living resources, if sustainable, is an important conservation tool because the social and economic benefits derived from such use provide incentives for people to conserve them” (IUCN 2000; point 7a).

The landscape approach also takes steps to involve relevant stakeholders in decision- making processes. Stakeholders are both people with power to control the use and management of resources, as well as those with no power or influence but whose livelihoods are affected by changing use or management (Brown 2004; p.238). Brown (2004) describes how the first step is to identify stakeholders and understand their relationships to the resources:

Stakeholder analysis is a system for collecting information about groups or individuals who are affected by decisions, for categorizing that information, and for explaining conflicts that may exist between important groups and identifying areas where trade-offs may be possible. It can be undertaken simply to identify stakeholders, to manage stakeholders and their interests, or to explore opportunities for getting groups or individuals to work together (Brown 2004; p.238). The author points out, however, that identifying stakeholders is only the first step. Developing trust between stakeholders—in the decision-making processes, and in the institutions and individuals making the decisions—is a necessary, but often very difficult, second step (Brown 2004). The final step involves designing appropriate institutions for integrating conservation and development that can be supported by deliberative inclusionary approaches and social learning (Brown 2004).

19 These types of inclusionary approaches are especially important given the risk that landscape approaches—with their wide scope—could result in centralised planning and actually harm the livelihoods of the local poor. Fisher et al. (2008; p.98-99) describe how this could occur, advocating the concept and practice of “negotiated landscapes” — through the types of participatory approaches described above—instead of “planned landscapes”:

The landscape concept is an entirely different way of looking at land use. Balancing land-use objectives over a wide scale is more useful than attempting to balance them at a site level. There are risks that the landscape concept can be used as a justification for centralised planning and an attempt to control the ways in which objectives are balanced. We are arguing for negotiated landscapes, not planned landscapes. Particular views about how landscapes should be conserved (or if they should be conserved) should be negotiated with other stakeholders. Whenever attempts are made to implement decentralised planning there is a tendency to revert to expert-driven land-use planning, perhaps with a little bit of community-level consultation thrown in for good measure. A number of well-intentioned conservation landscape (or ecosystem) approaches have inadvertently ended up reinforcing inequitable tenure and use rights and power relations, and exacerbating the insecurity of vulnerable rural livelihoods. In summary, the landscape approach explicitly considers the needs of important stakeholders, when making land-use trade-offs that do not affect important ecosystem functions within the landscape. Leading international conservation organisations such as the World Wildlife Fund, the Wildlife Conservation Society, and the African Wildlife Foundation are increasingly using landscape-based approaches in their efforts to conserve global biodiversity21.

Summary of differences between the traditional and landscape approaches to conservation

Before concluding this section, I thought it would be useful to summarise the two contrasting approaches to conservation we have considered thus far. The differences stem from the basic nature-related and socioeconomic paradigms of each, which, in turn, translate to policy goals, policy instruments, management approaches (especially

21 More details are presented in Appendix 1.

20 regarding land management), and even the disciplines involved. These differences between the two approaches are summarised in Table 1.3.

Table 1.3: Contrasting approaches to conservation—the traditional approach versus the landscape approach (adapted from Miller et al. 2011; p.951)

The traditional approach The landscape approach Nature-related • Nature as mechanistic/ linear • Nature (and ecosystems) as paradigms • ‘Balance of nature’/ complex, non-linear and ‘unchanging wilderness’ dynamic, i.e. the ecosystem approach • ‘Flux of nature’ Socioeconomic • Human presence is a threat to • Humans and biodiversity can paradigms biodiversity conservation co-exist productively • The goals of conservation and • Human livelihoods can improve development are often, if not with effective conservation inherently, incompatible, i.e. efforts (e.g., ecosystem ‘conservation versus services), i.e. ‘conservation for development’ development’ Policy goals • Biodiversity protection • Balancing biodiversity • Preservation of ‘wilderness’ protection with economic/social goals such as sustainable development, improving local livelihoods, etc. Policy • ‘Fortress conservation’, i.e., • Conservation both inside and instruments PAs with strict controls of outside PAs, i.e. across human activities within these landscape areas • Sustainable use of natural resources Management • Top-down approaches/ • Participatory approaches/ approach ‘planned’ landscapes ‘negotiated’ landscapes • Deterministic/ expert-based • Adaptive management management • Systems approach to land • Reductionist approach to land management, i.e., uses across management, i.e., land-uses in entire landscape considered at adjacent land management units the proper scale

21 The traditional approach The landscape approach considered independently (often managed by different authorities) Primary • Conservation biology • Conservation biology disciplines • Sociology/anthropology • Political ecology • Development economics

Table 1.3 encapsulates the discussion in sections 1.2 through 1.4, illustrating how the linear, mechanistic paradigm of nature within the traditional approach, along with the view that people must be excluded from this ‘pristine wilderness’ in order to preserve it, leads to the narrow policy goal of biodiversity conservation. This, in turn, drives the use of strictly controlled PAs as policy instruments and a top-down, deterministic approach to management in general, and a reductionist approach to land management in particular. Conservation biology is the predominant discipline within the traditional approach. In contrast, the landscape approach perceives nature to be complex, dynamic and non- linear, and consider humans and biodiversity as dependent on each other. This view results in a broad set of policy goals where both conservation and socioeconomic objectives are balanced across the landscape through policy instruments such as the sustainable use of natural resources, and more participatory, adaptive management approaches (i.e. ‘negotiated’ landscapes instead of ‘planned’ landscapes). Consequently the landscape approach utilises a broad suite of disciplines.

22 1.5. Implementing the landscape approach in the developing world: issues and areas for further research

Although landscape-based approaches are gradually gaining recognition, they are still being fine-tuned. I now consider some of the main challenges facing conservationists, policy makers, and others trying to implement landscape approaches in developing countries—this will help identify areas where further research is needed and illustrate the importance of applying the approach in practical situations.

Implementing the landscape approach in the local context

Conservation planning strategies—especially those used in landscape approaches—have become increasingly sophisticated, partly because of the tremendous complexity involved. However, the increasing complexity of mathematical-based modelling has led to what Knight et al. (2006; p.410) describe as an ‘implementation crisis’:

Most conservation planning studies published in peer-reviewed journals consist of scientific analyses in pursuit of ever-more precise information on, and efficient techniques for, prioritizing elements of nature (i.e., area selection algorithms). Far less scientific attention has been dedicated to planning the activities required to ensure the persistence of species, landscapes, and the processes that support them at selected priority areas. Ehrlich (1997) laments, “...journal[s] [are] still packed with papers describing more and more sophisticated analyses applied to more and more trivial problems”. Despite numerous references to the implementation crisis in academic literature, “few academic conservation planners regularly climb down from their ivory towers to get their shoes muddy in the messy, political trenches, where conservation actually takes place” (Knight et al. 2006; p.410).

Implementation issues are especially significant in developing countries where there is often a dearth of talent and resources within government agencies, making it difficult for them to carry out their narrow mandates such as policing national parks, let alone implementing more complex landscape approaches. Given these deficiencies in local institutions, large development organisations often come in and implement such approaches using a top down, project-based approach. This could result in projects

23 being designed by ‘experts’ who have little knowledge of the local context, leading to the abandonment of projects once funding has ceased. What Sayer and Wells (2004; p.38-39) term the ‘pathology of projects’ is all too common in the developing world:

Projects are intrinsically limited in space, time, and numbers of beneficiaries, while the main cause of biodiversity decline is the loss, fragmentation, and degradation of habitat over large areas due to many kinds of human activities. Halting or mitigating biodiversity loss requires changing the behaviour of large numbers of people dispersed over large areas for long periods of time. Projects are inherently unsuited for this... Within projects there tends to be a disproportionate emphasis on planning at the expense of implementation. Expatriate experts who will have no subsequent involvement in implementation often prepare detailed project plans. As pointed out previously, however, donor funding and expertise are often required to implement landscape approaches in developing countries given the lack of local expertise and resources; therefore these interventions must be designed in a way that mitigates the possible pitfalls. Consequently, there is an imperative to apply the landscape approach in specific situations within developing countries in a manner that takes into account the local context, builds local expertise and simplifies the implementation process without compromising the effectiveness of the approach.

Clearly identifying and explaining the benefits of different land-use options

There is a critical need to evaluate the opportunity costs of conservation within and across landscapes, especially since trade-offs are required among different land-use options. Chazdon et al. (2009) believe that this input would enable policy-makers to develop schemes to provide adequate financial benefits, equity, rights, and choices for rural people whose livelihoods depend upon sustainable agricultural production or resource extraction. The authors also point out that further research is needed to quantify the value of ecosystem services such as carbon sequestration “in a wide range of habitat types within human-modified landscapes, including silvopastoral systems, swidden agriculture, and agroforestry systems” (Chazdon et al. 2009; p.146). Valuing ecosystem services is particularly relevant in developing countries where government policymakers make land-use and resource allocation decisions based on multiple, often

24 competing, priorities such as poverty alleviation, rural development and biodiversity conservation (World Bank 2012).

Identifying and quantifying benefits of natural resources in specific landscapes would also help fill a gap in the literature related to providing case examples of the concept of sustainable use22. For example, according to Cooney (2007; p.37):

One of the failures of sustainable use advocates has been in clearly documenting and analysing examples of sustainable use yielding conservation benefits compared to strict protection approaches. These examples exist and many are known anecdotally, in grey literature, or through professional networks. But with few exceptions, few clear case examples are available for those who seek to justify IUCN policy and convince others of the potential of sustainable use as a conservation approach.

Incorporating local knowledge and practices

Land-use management approaches in developing countries, especially those concerning PAs, are usually grounded in western technical and legal precepts, often as part of a colonial legacy. However, since local people have often managed resources for centuries, it is important to incorporate their traditional knowledge as pointed out by Frost et al. (2006; p.8):

The needs and perceptions of rural communities often remain hidden to outsiders unless special efforts are made to uncover them. Because people’s decisions about managing natural resources are based on their knowledge and comprehension of those resources and their dynamics, understanding the nature of that knowledge is crucial to successful intervention, if only to identify what is needed to broaden the basis of decision-making. Kaimowitz and Sheil (2007) point out that considerable research is needed on how local people value natural resources and the techniques they use to manage them.

22 Introduced in section 1.4.

25 Effectively engaging stakeholders and incorporating their needs

Throughout the world, there has been a tendency for a small elite group of powerful people to control land-use decisions (Maginnis et al. 2004). The crudest manifestation of this phenomenon is when the strongest groups grab land, resources and rights; however, this could also happen in a subtler manner when small cadres of specialists, who act at the behest of political and economic elites, have a disproportionate influence on land-use policy (Maginnis et al. 2004). Even when there is a commitment to participatory decision-making, central, regional or local governments often pay most attention to those who are best placed to lobby them (Maginnis et al. 2004).

This ‘capture’ of decision making-processes by elites is evident, in various degrees and manifestations, in most developing countries. Sometimes, international conservation organisations contribute to the problem, as described by Maginnis et al. (2004; p.333): “conservationists have sometimes ignored the needs and interests of local communities by negotiating with regional or national authorities, in effect creating an intractable situation, which severely reduces options to negotiate trade-offs for the future”. Consequently, there is a need to understand how international organisations and government agencies that implement landscape approaches could build effective partnerships with local groups to share decision-making and risks, and achieve a mutually acceptable balance of rights and responsibilities (Kaimowitz and Sheil 2007).

Developing approaches that incorporate the effects of climate change

As pointed out before, climate change could drastically affect PA effectiveness both in terms of representativeness and persistence. This phenomenon would also affect the broader landscape as described by Chazdon et al. (2009; p.146): “climate change will have major impacts on agricultural production as well as on biodiversity within both human-modified landscapes and PAs throughout the tropics”. Consequently, the authors stress the need for research exploring how agricultural landscapes could be designed

26 and managed to maximize carbon sequestration and reduce emissions from deforestation23, thereby contributing to climate change mitigation.

Summary: issues and areas for further research

In summary, applying landscape approaches in developing countries could be challenging, due to issues with incorporating local contexts, knowledge and practices, identifying and estimating the approach’s benefits, engaging stakeholders and meeting their needs, and mitigating effects of climate change. There is a pressing need to apply the approach in practical situations, because although the specific manifestations of the issues may differ according to context, many themes are common across developing countries. As Kaimowitz and Sheil (2007; p.571) point out “as with any conservation investment, experiences should be reviewed, shared and learnt from”. Consequently, my thesis focuses on how a landscape-based approach could be implemented in Sri Lanka—a developing country—in a manner that addresses the challenges I identified above.

1.6. Focus of research

My thesis will apply a landscape approach to achieve better conservation outcomes in Sri Lanka, a developing country which has significant biodiversity resources24 that are under threat (Mittermeier et al. 2004; IUCN 2009). The country has historically adopted a ‘traditional’ approach to conservation, resulting in its PA network—though extensive—being inadequate to conserve its biodiversity, especially wide ranging species such as elephants and leopards (Green and Gunawardena 1997). The roots of this approach lie in the fact that British colonials established the country’s first ‘modern’ PAs25 using the ‘Yellowstone model’, which decreed that a ‘pristine wilderness’ must

23 Also termed REDD (Reduced Emissions from Deforestation and degradation); discussed in detail Chapter 7. 24 It has been designated as one of the 34 global biodiversity hotspots along with the Western Ghats of India. 25 As will be discussed in Chapter 2, Sri Lankan Kings established indigenous PAs well before the arrival of the British.

27 be protected from the destructive influence of human beings26 (Uragoda 1994). The British also put in place commensurate conservation-related legislation that promoted a reductionist approach to land management resulting in PAs being isolated within the landscape (which is still in force today albeit with several important amendments) (Jayasooriya et al. 2006).

While these colonial conservation activities resulted in the establishment of an extensive set of PAs, this traditional approach to conservation has proved inadequate to stem a loss of valuable biodiversity and ecosystems in recent times, especially in the context of the escalating demand for land in a developing country whose population has almost tripled since independence (Department of Census and Statistics 2011). Human-wildlife conflict is one of the most significant issues stemming from the Sri Lankan government’s continued use of the traditional approach to conservation, as it often involves charismatic species that capture the imagination the wider public (instead of only conservation biologists) while the people affected are mostly the rural poor, whose very livelihoods are threatened by the conflict. Consequently, human-wildlife conflict receives wide publicity and corresponding government attention.

The central hypothesis of the thesis

The central hypothesis of my thesis is that a landscape approach could achieve better conservation outcomes in Sri Lanka compared with the traditional approach that is currently used by the government. The landscape approach is especially suited to addressing what is perhaps the main issue in Sri Lankan conservation: how can the country more effectively conserve the Asian elephant, a flagship species that has played a special role in Sri Lanka’s history, culture, religion and folklore. As elephants range widely outside the country’s PA network, it results in human-elephant conflict which arises when the animals, faced with dwindling habitats, raid crops (Bandara and Tisdell 2002). The Sri Lankan government dedicates much effort and resources toward resolving this rapidly escalating conflict, given the cultural significance of elephants and pressure from local villagers whose property, and often even their lives, are threatened.

26 This approach (and its consequences) is described in more detail in Chapter 2.

28 Elephant conservation is also important from an ecological perspective because the Asian elephant is listed as ‘endangered’ under the IUCN Red List of Threatened Animals; moreover, given their large ranges, elephants are regarded as an ‘umbrella species’ because their conservation would also protect many other species occupying the same area (IUCN 2009). Given these ecological benefits and elephants’ importance from a cultural perspective, Sri Lanka must quickly develop a coherent policy for elephant conservation. This must cover the broader landscape, since more than two thirds of elephant habitat and ranges are outside PAs (Groom et al. 2006).

The Sri Lankan government’s main initiatives to address human-elephant conflict— such as translocating problem animals and conducting elephant drives 27 —are in accordance with its overall traditional approach to conservation. Its lack of an ecosystem-based approach, non-systematic approach to land-use planning, and inadequate management of designated buffer zones near PAs has resulted in land-use that exacerbates human-elephant conflict (discussed in detail in Chapter 4). The Department of Wildlife, which is responsible for elephant conservation, pays insufficient attention to the human dimension of the issue, leading to hostility between locals and department officials (as discussed in detail in Chapters 4 and 7). Consequently, my central hypothesis is that a landscape-based approach would prove far more successful, and, in Chapters 5 through 8, I explore in detail how this would work in a specific landscape in southeast Sri Lanka.

Research questions stemming from the central hypothesis

My research uses a landscape-based approach to develop and implement land-use options that permit resource sharing between elephants and human stakeholders such as farmers and government policymakers. In order to keep the analysis practical, I focus on a specific landscape in southeast Sri Lanka, which contains key elephant habitat but where rapid development is requiring policymakers to make difficult decisions regarding land-use.

My specific research questions are:

27 These intiatives are described in detail in Chapter 4.

29 1. What is (and has been) Sri Lanka’s approach to conservation, and what are key issues associated with this approach?

2. What have been the impacts of this approach on elephant conservation and human-elephant conflict?

3. Could a landscape-based approach help conserve elephants while addressing stakeholder needs in southeast Sri Lanka?

- How could resources—especially land—be shared among various stakeholders?

- How could government policymakers (usually the ultimate arbiters regarding land-use) and other relevant stakeholders be convinced to adopt this solution?

- How should the approach be implemented, and what enabling factors are required?

4. What lessons could be drawn from this research in the context of Sri Lanka and other developing countries?

The main focus of my research is the practical application of a landscape approach to address a pressing conservation issue in a developing country. The basic mechanism to share resources between elephants and local farmers (the key stakeholder group) was developed by a group of conservation biologists based on over a decade of research on elephant behaviour (Fernando et al. 2005; Fernando et al. 2008). Their proposal to establish mixed-use ‘Managed Elephant Ranges’ (MERs) adjacent to PAs was incorporated into national conservation policy in 2006, but has not yet been implemented (Government of Sri Lanka 2006; Bhattacharjya 2010). This type of situation is common, according to Meffe and Viederman (1995; p.327):

Many scientists in the field of conservation biology are only now beginning to understand how policy decisions are an important influence on natural systems. We can no longer simply do the science and hope that someone else uses the information to make good laws that protect species and their ecosystems. We now understand that much of what we do in conservation biology is essentially worthless if it is not translated into effective policy. All the theories, all the ecological and genetic models, and all the data amassed will have little effect if we do not influence policy

30 and human behaviour toward protection of biological diversity. We believe that the major advances in conservation action will take place not in scientific laboratories or field research sites, but in the political and economic arenas, because present limitations in conserving biodiversity do not typically occur through lack of knowledge, but rather poor implementation. Thus, the science of conservation biology is necessary, but not in itself sufficient, to stem species extinction and ecosystem degradation. The challenge is for science to inform policy to change individual and institutional behaviour. Given the difficulty of translating the work of conservation biologists with this particular situation in Sri Lanka—a problem that is common in other parts of the world—the focus of my research has been to develop a blueprint for implementing the approach in a specific landscape in Sri Lanka. To do this, I utilised a multi-disciplinary approach comprising techniques drawn from a variety of fields, as discussed next.

1.7. Frameworks, approaches and methodology

Overall approach

Brewer (1999) points out that the lack of an interdisciplinary approach has contributed to difficulties in translating scientific research into effective policy:

Much high-quality science illuminates environmental problems, but it is often poorly organized and incomplete. It often does not have an interdisciplinary integration and synthesis that permit problems to be seen in a larger context, especially in an ecologically sensitive and sensible one. It is often not geared to the scale needed to shed light on environmental problems of long-term importance to human well-being. In short, much essential knowledge is not capable of guiding the development of policy, heightening public awareness, or even informing and enlightening political debate (Brewer 1999; p.327). The author goes on to define interdisciplinary research, and how it is integral to the policy process (Brewer 1999; p.328 - 329):

Interdisciplinarity generally refers to the appropriate combination of knowledge from many different specialities especially as a means to shed new light on an actual problem. In notably effective efforts, the combination of disciplines adds value: the total is more interesting than the sum of the individual contributions or parts. Problems designate theory and methods, not the reverse, in sharp contrast to discipline-based and curiosity-driven inquiry. They also remind one of

31 the missing parts of an analysis and thus stimulate imagination to create unexpected policy alternatives for consideration. More importantly, raising and working to answer such questions help emphasize a problem orientation, contextuality, multiple methods, and an overriding concern for the human dimension and values. Taken altogether and elaborated over many years, these ideas form the core concepts of a distinctive field known as the policy sciences. Given the complex and multi-faceted nature of the problem I am addressing, I have adopted an approach incorporating perspectives from different fields—including conservation biology, geography, economics, public policy, sociology and history— comprising both quantitative and qualitative methods. In accordance with the view of Brewer (1999) quoted above, my aim was to combine several methods drawn from a number of fields into a whole that is “more interesting than the sum of the individual contributions or parts”.

Establishing context is another important dimension of public policy, as described by Clark (2002; p.29):

The contexts that generate and condition all public policy problems matter enormously in that all things are interconnected and that the meaning of anything depends on the context... The paramount importance of context demands that it be perceived by observers thoroughly and correctly. If the context is understood too narrowly, in a way that is too bounded, then significant details will be overlooked and misconstrued. Consequently, I have considered the specific case study of Sri Lanka’s southeast landscape, where human-elephant conflict is expected to increase, given the area’s rapid, often unplanned, development. My consideration of a specific case study provides a specific context within which an overall policy approach to elephant conservation within Sri Lanka can be tested. Given the immense problems human-elephant conflict is causing to a variety of stakeholders28, my findings would be of immediate relevance to government policymakers, and help realise some of the benefits identified by Clark (2002; p.142): “Case studies help people learn how to choose analytic approaches and design new ones, identify policy-relevant information, construct successful policy arguments, and communicate results and recommendations in a style appropriate to real decision-making contexts”.

My objective and overall approach is summarised in Figure 1.2.

28 Which I explore in detail in my stakeholder analysis in Chapter 6.

32 Landscape People

• Land use • Social indicators • History • Livelihoods • Development activities • Stakeholders related to establishing Managed Elephant Ranges

Environment Economics

• Key environmental • Key sources of issues economic value Objective • Elephant population • Changes, based on • Human-elephant Implement a landscape- land-use scenarios conflict based approach in • Non-economic factors southeast Sri Lanka

Figure 1.2: Objective and overall approach

Having identified the range of issues to be considered, I now describe how the relevant analyses were performed. As summarised in Table 1.3, I conducted my own analyses regarding some aspects of the problem, while using secondary sources of information for others.

Table 1.4: Multi-disciplinary analyses used in my thesis

Area Issue Discipline(s) Analysis and/or sources of information Landscape Land-use Geography Geographic Information Systems (GIS)-based analysis History History Literature review, press articles Development activities Economics, Literature review, press history, public articles policy People Social indicators Sociology Government publications/ data, literature review Livelihoods Sociology, Surveys, interviews, press

33 Area Issue Discipline(s) Analysis and/or sources of information economics articles, literature review Stakeholders related to Sociology Stakeholder analysis based establishing MERs on surveys and interviews, press articles, literature review Environment Key environmental issues Environmental Literature review, press studies articles, government reports Elephants Conservation Literature review biology Human-elephant conflict Conservation Surveys and interviews, biology, sociology literature review, press articles Economics Key sources of economic Economics Economic analysis/modelling value Changes in economic Economics Economic modelling value based on land-use scenarios Non-economic factors not Ecology, Surveys and interviews, captured in economic environmental literature review, press analysis (e.g. critical studies, sociology articles environmental thresholds, equity regarding income distribution)

As described in Table 1.3, I used a combination of primary sources (e.g. surveys, interviews) and analyses (economic modelling, GIS). These have been augmented by secondary information from a variety of sources. I reviewed academic literature to understand the historical and cultural context of issues being addressed in my thesis. For example, journal articles on Sri Lanka’s conservation history and land-use provided a context for current patterns of resource allocation within the landscape. I also used material from government agencies such as the Department of Wildlife and the Department of Census and Statistics, local and international conservation NGOs such as the Centre for Conservation Research and the World Bank, and articles from the Sri Lankan press.

Having described my broad approach, I now introduce the analyses I conducted—GIS, economic modelling and stakeholder analysis—providing brief descriptions of each

34 type of analysis; however, the relevant concepts and frameworks are presented in detail in the appropriate chapters.

Economic analysis

Conservation practitioners are recognising that traditional justifications for conservation, based solely on a combination of scientific and ethical arguments, would be insufficient to conserve rapidly declining ecosystems and biodiversity (Turner and Daily 2008; Perrings 2010). Economists point out that a key driver of the loss of these valuable natural resources, is that the valuable services they provide to human beings—so-called ‘ecosystem services’ such as flood control, watershed protection, crop pollination—are not valued in everyday economic decisions:

It has often been argued that a major reason for our failure to conserve natural ecosystems is that we do not realise how valuable they are. The farmers deciding whether to burn a hectare of forest to clear it for agriculture focus on the potential crop yields they may obtain, but pay little attention to the many ecological services that would go up in smoke. Likewise, national ministers of finance often base their budget decisions solely on the basis of indicators such as GDP29, foreign exchange balances, and tax receipts, in which ecosystems services either do not appear or are not recognised as such—indeed, perversely, GDP often identifies activities that destroy ecosystems as ‘benefits’. Not surprisingly, conservation budgets tend to get slighted (Pagiola et al. 2004; p.1). Consequently, I conducted an economic analysis to estimate the value of natural resources within Sri Lanka’s southeast landscape. I used the concept of total economic value, a holistic framework that captures ‘direct uses’ such as agriculture and tourism, ‘indirect uses’ such as watershed protection, ‘option values’ or deferring use of resources so that they could be used later, and ‘existence values’ or the satisfaction gained from knowing that these resources exist. I developed an economic model that captures the landscape’s major sources of economic value—especially those associated with natural assets—and created alternative scenarios to understand how these sources of value could be impacted by future land-use decisions.

My model utilises several important economic concepts. The first is the cost benefit framework that is widely utilised in public policy decision-making to compare

29 Gross Domestic Product.

35 alternative policy options on the basis of whether they lead to an efficient allocation of resources (Mullan and Kontoleon 2008; p.6). For a project or policy to qualify on cost- benefit grounds, its social benefits must exceed its social costs; ‘society’ is simply the sum of individuals (Pearce et al. 2006; p.16). In economic terms, an efficient policy change should meet the ‘Pareto’ condition that the maximum aggregate willingness to pay of the gainers (the social benefit) is greater than the aggregate minimum willingness to accept of the losers (the social cost) (Hanley and Barbier 2009; p.30). In freely functioning markets, these preferences are reflected in prices, but as implied above, in the case of environmental resources the social benefits are not reflected in market prices30; in economic terms, their ‘market prices’ are lower than ‘economic prices’, which in turn provides incentives for people to engage in economically inefficient behaviour such as deforestation (Hussen 2004; p.55 - 58; Farley 2008; p.1401). Consequently, the solution according to economics is to understand the ‘true’ benefits of natural resources (or the true costs to society as a whole when they are used in an unsustainable manner) through the process of valuation; policy choices are made so that these values are reflected in the prices of the relevant goods and services (Farley 2008; p.1401; Ha et al. 2008; Molinos-Senante et al. 2013).

There are several techniques to value environmental goods and services for which there are no properly functioning markets. These methods fall into two broad categories: revealed preference methods based on actual consumer behaviour, such as values implied through prices of related market goods; and, stated preference methods where peoples’ choices are assessed through surveys, or experiments where subjects are presented with hypothetical situations (Tietenberg and Lewis 2012; p.82). At the end of section 7.3 in Chapter 7, I describe in detail some of the main methods in both categories: simulated markets, travel cost, hedonic pricing methods, and avoidance expenditures (revealed preference methods); and, contingency valuation and attribute- based methods (stated preference methods). Later in the chapter (in Table 7.6), I link these theoretical concepts to the specific methods I used in my economic model.

Finally, as the costs and benefits from valuing the various TEV components occur at different points in time, my model utilises the concept of discounting, which recognises that “a sum of money available in the future is generally regarded as being worth less

30 This is due to presence of ‘externalities’, a concept that it described in detail in Chapter 7 (section 7.1).

36 than the same sum of money available now because if the future sum were available now it could be invested at the going rate of interest to ensure its return plus accumulated income from interest on it at the later date” (Tisdell 2005; p.239). The process of discounting enables the (discounted) costs and benefits resulting from a project or policy choice to be added together to arrive at a single figure—net present value—that enables comparison among these choices (Tisdell 2005; p.75). The specific discount rate used is extremely important, as it determines how much less the future is weighted relative to the present (Pearce et al. 2006; p.184). Some economists argue that interest rates used for many projects—especially those with environmental impacts— are too high from a social point of view, resulting in individuals investing in projects that are not socially optimal (Hussen 2004; p.182; Tisdell 2005; p.75). Consequently, at the end of section 7.1 (Chapter 7), I discuss my rationale for using a discount rate of 5% in my economic model.

In Chapter 7, I discuss in detail the theoretical concepts underlying my economic analysis, and how these concepts are used within my economic model. In sections 7.1 through 7.3, and at the beginning of Appendix 7, I present the theoretical basis for economic valuation, introduce the main valuation techniques used, and discuss the significant criticisms of this sometimes-controversial process (and how they are addressed in my analysis and interpretation of results). Next, in section 7.4, I point out how the specific components of TEV relate to Sri Lanka’s southeast landscape. Finally, in section 7.5 and in the corresponding section in Appendix 7, I describe my economic model (including model structure, assumptions and calculations), present my results and discuss their implications.

It is important to keep in mind that total economic value only captures benefits derived from an environmental resource in economic terms. In other words, it is an anthropocentric concept applicable only to goods or services over which individuals are willing to make trade-offs, and does not capture the value of services for which trade- offs are not possible, e.g. ‘intrinsic values’ such as cultural and religious values (Mullan and Kontoleon 2008). Moreover, there are inherent complexities, uncertainties and irreversible characteristics associated with ecosystems and biodiversity (Gilpin 2000; p.102). Therefore, when evaluating the scenarios derived from my economic model, I also considered potential impacts of key elements not captured in the model (section 7.5

37 in Chapter 7). Also, as described previously, I used economic valuation in conjunction with qualitative approaches such as stakeholder analysis, which I consider next.

Stakeholder analysis

As pointed out previously, considering the needs of stakeholders is an important characteristic of landscape-based approaches. Although methods such as participatory action research provide stakeholders with more opportunities to participate in the entire process (from research design to the analysis of results), they are difficult to incorporate within most research programmes given their significant time and resource requirements31 (Mackenzie et al. 2012). Consequently, I chose stakeholder analysis to understand the needs of key stakeholders in my study area, while specifying an institutional framework that would provide these groups with more opportunities to participate in the design and management of specific MERs (described in Chapter 8, section 8.3).

Mushove and Vogel (2005; p.185) define stakeholder analysis as “a range of tools or an approach for understanding a system by identifying the key actors or stakeholders on the basis of their attributes, interrelationships and assessing their respective interests related to the system, issue or resource”. While I have described my methodology in detail in Chapter 6, I now briefly look at interviews, my main source of primary information from key stakeholders.

Dunn (2005; p.79) defines an interview as a “face-to-face verbal interchange in which one person, the interviewer, attempts to elicit information or expressions of opinion or belief from another person or persons”. He points out that a main benefit of interviews is that they enable a researcher to collect a diversity of meaning, opinion and experiences. Winchester (2005) identifies a range of interviewing and related methods, from more qualitative and individualistic approaches to more structured methods; these are presented in Figure 1.3.

31 I describe my rationale for choosing stakeholder analysis in more detail at the beginning of Chapter 6 (section 6.1).

38 General Specific method Research questions methods

Biography Autobiography

Biography Individual Oral history

Interviews Unstructured

Semi-structured

Structured

Focus groups—open-ended

Surveys Focus groups—open-ended General/ Structured Questionnaires—structured

Figure 1.3: Oral qualitative methods in human geography (from Winchester 2005; p.7)

The main type of interviews I conducted were semi-structured interviews with key stakeholders. For example, interviews with farmers and other locals helped me understand their patterns of resource use and their views on MERs, while interviews with government policymakers provided insights on their decision-making processes, especially regarding human-elephant conflict. I also conducted some focus group interviews to understand the diversity of viewpoints among people within a single stakeholder group such as farmers32. In addition, I had access to data from a survey of 800 farmers within the landscape, conducted by the World Bank in 2009.

In Chapter 6 and Appendix 6, I discuss in detail stakeholder analysis theory and methods, identifying the particular set of tools I choose for my analysis (along with the rationale for my choices). In Chapters 5 and 6, and Appendices 5 and 6, I specify the processes related to my semi-structured interviews and the World Bank survey (e.g.

32 This method, which involves a small group of people discussing a topic or issues defined by a researcher, helps illustrate the interaction between members of the group (Waltner-Toews et al. 2008; p.4).

39 choice of interviewees, interview questions, data collection methods), and present my analyses of the resulting data (and their implications).

Geographic Information System techniques

Geographic information systems (GIS) are increasingly used to integrate and analyse the rapidly increasing amounts of conservation-related data (Salem 2003). Phua and Minowa (2005; p.209) describe new approaches that combine GIS with techniques to facilitate decision making among multiple stakeholders:

The ability of GIS to integrate with the multi-criteria decision-making approach has been shown in studies related to site determination for a nuclear waste facility and for a noxious waste facility. And the GIS-based multi-criteria decision-making approach has also extended to solving planning problems that involve conflicting multi-objectives such as land use allocation problems. Relatively few studies related to forest conservation have employed multi-criteria decision-making approach with GIS techniques but a GIS-based multi-objective evaluation was experimented with for management optimization of urban green parks in Italy.

I used GIS techniques to design potential MERs by mapping the landscape’s existing forested land, indicating areas important to various stakeholders. This process helped identify both areas that could be conserved with relatively little effort, as well as those likely to be claimed by multiple stakeholders requiring trade-offs and compromises. In Chapter 5 and Appendix 5, I describe the GIS data I obtained from the Survey Department of Sri Lanka, and present maps I created using these data (along with the associated analyses).

Having outlined my research focus and methodology, I conclude this chapter by outlining my thesis structure.

40 1.8. Thesis structure

In Chapter 1, I introduced my research context, focus and approach. First, I discussed the biodiversity crisis and how traditional approaches to conservation have not effectively conserved biodiversity and ecosystems. I then proposed the landscape approach as an alternative and identified the need to refine the approach through practical application in developing country contexts. Finally, I clarified the focus of my research, stating my research questions and outlining my methodology.

In Chapter 2, I provide an overview of Sri Lanka’s biodiversity resources and conservation landscape, establishing that while the country has valuable biodiversity resources, the approach it has historically taken to conserve these resources has been flawed. I discuss the country’s history and socio-economic profile, its traditional approach to conservation, conservation-related institutions, and the evolution of its legislative and policy framework. I end the chapter with a discussion on conservation- related issues in Sri Lanka.

In Chapter 3, I discuss the importance of the Asian elephant in Sri Lanka, describing how it is arguably the country’s most important species from cultural, historical and religious perspectives. I also point out that conserving Sri Lanka’s elephants is important from a global conservation perspective, since the Asian elephant is listed as endangered; moreover, since elephants range widely outside PAs, their conservation would also protect many other species within the landscape.

In Chapter 4, I illustrate why human-elephant conflict is perhaps the most important conservation-related issue in Sri Lanka, given the toll the conflict is having on humans and elephants alike. I describe how current approaches to address the issue have largely been unsuccessful, and present a landscape-based approach that could significantly enhance elephant conservation and mitigate human-elephant conflict.

Having introduced the problem and proposed a potential approach to address it, I then explore in detail how this approach could be applied to a specific landscape in southeast Sri Lanka. In Chapter 5, I introduce Sri Lanka’s southeast landscape, analysing the region’s land-use, socio-economic profile, and people’s livelihoods. I also describe the area’s elephant habitat and human-elephant conflict, and use GIS techniques to delineate potential MERs. In Chapter 6, I present my stakeholder analysis that identifies

41 needs and concerns of key stakeholders and helps determine the potential role each group could play in implementing MERs. In Chapter 7, I present the results of my economic analysis. First, I discuss the justification of economic valuation of natural resources, relevant theoretical concepts (e.g. prices and discounting), potential criticisms of valuation, valuation techniques, and the concept of total economic value. I then describe the economic model I developed using these concepts, discussing my findings in the context of several land-use scenarios. I also clarify important non- economic aspects associated with each scenario (that were not captured in the model). In Chapter 8, I bring my different analyses together to derive a set of essential MER characteristics, identify how potential challenges to the implementing these areas could be addressed, and recommend the organisation structure, strategic plan and approach that should be used to implement MERs in Sri Lanka’s southeast landscape.

In Chapter 9, I clarify and summarise my research findings, illustrating how they have addressed my research questions. I conclude with a discussion of the potential applications of my research in Sri Lanka and beyond.

42 Chapter 2: Overview of Sri Lanka’s biodiversity resources, land-use and conservation-related institutions

This chapter provides the context for many issues that will be discussed in the rest of the thesis. The first four sections provide an overview of Sri Lanka’s biophysical context, its unique biodiversity, its socio-economic context and land-use, showing that while Sri Lanka has valuable biodiversity resources, it is a densely populated developing country that must provide for its citizens’ needs. This is illustrated by its patterns of land-use: roughly a third each are devoted to agriculture and forests/wildlife. The government’s ownership of almost 80% of the country’s land makes it the most important entity involved in determining land-use in Sri Lanka.

The next section outlines the historical forces that have shaped the country’s socio- economic context and land-use patterns. It traces back through the major historical events that have shaped the country’s current approach to conservation and land-use, considering three main periods: the era of ancient kingdoms, colonial times and the six decades since independence. The section illustrates how colonial authorities’ approaches to both conservation and land-use were often very different to their ‘indigenous’ counterparts. The section ends with a description of conservation and land- use policies of post-independence governments.

The next section describes Sri Lanka’s conservation-related institutions, introducing major governmental and non-governmental players. These groups are important in the stakeholder analysis I present in Chapter 6. The penultimate section provides an overview of the country’s environmental legislative and policy framework, illustrating how many of Sri Lanka’s laws and policies have roots in the colonial era, through institutions such as the Forest Department founded by the British. These links have shaped the government’s current approach to conservation, which is characteristic of the ‘Yellowstone model’ described in Chapter 1 and Appendix 1. The chapter ends with a discussion of how this approach has contributed to several conservation-related issues in Sri Lanka, notably human-wildlife conflict.

43 2.1 Biophysical context

Sri Lanka is a continental island separated from southern India by the 20-metre-deep Palk Strait. The island—65,610 km² in size—has been repeatedly connected with India between successive interglacial periods, most recently until about 20,000 years ago by a land bridge up to about 140 km wide (Mittermeier et al. 2004). Sri Lanka is about 400 km in length and about 220 km at its widest point (Peebles 2006). The south-central mountainous region that covers 20% of its land area has 162 peaks ranging in elevation from 1000 – 2,500 m 33 (Chokkalingam and Vanniarachchy 2011; p.3). This is surrounded by an intermediate zone of upland ridges and valleys, which in turn is surrounded by an outer zone of lowlands (Jayasooriya et al. 2006). A coastal fringe consisting of sandbars and lagoons skirts the island.

Weather patterns, climatic zones and climate change

Given its location just north of the equator, Sri Lanka has a tropical climate, and is warm and moist the year round (Peebles 2006). Annual rainfall varies from 900 mm to 6,000 mm, the western slopes of the central hill country receiving more rain than the north-western and south-eastern lowlands (Jayasooriya et al. 2006). The island is influenced by two wind regimes: the southwest monsoon from May to September and the northeast monsoon from December to February.

Sri Lanka’s south-central mountainous region has a major influence on its climate. It intercepts the moisture-laden winds of the southwest monsoon, resulting in high rainfall in the southwest lowlands and windward slopes of the central highlands; the rest of the country remains dry. Most of the island receives rainfall from the northeast monsoon. Consequently, the northern, eastern and southern parts of the country average 1,000 – 2,000 mm per year as compared to 2,000 – 5,000 mm in the southwest (Jayasooriya et al. 2006). There is also some rain during the inter-monsoonal periods—March to April and October to November—which is mainly convectional (Jayasooriya et al. 2006). As indicated in Figure 2.1, the country is divided into wet (> 2500 mm annual rainfall),

33 The highest point, Mount Pidurutalagala, rises to 2,524 m above sea level (Cameron 2005).

44 intermediate (1750 – 2500 mm) and dry (< 1750 mm) zones on the basis of this climatic regime (Chokkalingam and Vanniarachchy 2011; p.3).

Major climatic zones Rainfall isohyets

Figure 2.1: Major climatic zones and rainfall isohyets in Sri Lanka (from Jayasooriya et al. 2006; p.19, 21)

There is little seasonal variation in temperatures. Lowland temperatures average 28ºC year round, with slightly higher temperatures in the north-central plains (Peebles 2006). The temperature in the highlands varies with elevation—in towns at higher elevations temperatures typically vary between of 10ºC and 20ºC (Peebles 2006).

Recent research indicates that Sri Lanka is experiencing climate change. De Costa (2008), who analysed mean annual temperatures and rainfall patterns over a 140 year period34, found that the entire distribution of temperatures had shifted upwards over

34 From 1869 to 2007.

45 time and mean annual rainfall had declined35. Worryingly, he also found that “in the majority of locations examined, the length of growing seasons for agricultural crops, which is primarily determined by the availability of water, is likely to become shorter” (de Costa 2008; p.86). According to a communication from the Sri Lankan government to the United Nations Framework Convention on Climate Change (UNFCCC), predicted impacts of climate change in the country include (Environmental Foundation 2009; p.61):

• Sea level rise: inundation, coastal erosion, salt water intrusion;

• Rising temperatures: bleaching of coral reefs and subsequent damage, increased evaporation from reservoirs impacting hydroelectricity generation, reduction in crop yields, outbreaks of crop diseases and pests, health impacts from heat;

• Increased droughts: reduction in water availability, reduction in power generation from hydroelectricity, impacts on irrigation for agriculture, spread of diseases due to lack of water; and,

• Increased rainfall: floods, soil erosion, impacts on agriculture, spread of disease.

Rivers and reservoirs

Sri Lanka’s network of 103 major rivers constitutes the main component of the country’s inland waters. Their total collective length is about 4,560 km and their basins cover an area of 59,245 km2—about 90% of the island’s land area (Jayasooriya et al. 2006). Most major rivers originate in the central highlands and flow across the lowland plains into the sea. The longest is the 355 km long Mahaweli Ganga which drains an area of 10,448 km2, carrying water from the wet zone to the dry zone (Jayasooriya et al. 2006). Other significant rivers are the Malwathu Oya, Kala Oya, Kelani Ganga, Yan Oya, Deduru Oya and Walawe Ganga, which range from 140 to 160 km. Many rivers show extreme seasonal variability—some have a propensity to flood, while others, that originate and flow mainly through the dry zone, may dry up considerably for several

35 The evidence for rainfall reductions is not as conclusive as that for increase of air temperature, primarily because rainfall at any given location shows much greater year-to-year variation than air temperature.

46 months during the year (Jayasooriya et al. 2006). The extensive marshes and swamps associated with Sri Lanka’s river system are rich in biodiversity (Ministry of Forestry and Environment 1999).

Although Sri Lanka does not have large natural lakes, many (artificial) tanks36 dot its dry zone (Ministry of Forestry and Environment 1999). In total, there are about 10,000 tanks varying in size from 6,500 ha to about 1 ha; they irrigate over 500,000 ha of land, and their combined surface area exceeds 170,000 ha (Ministry of Forestry and Environment 1999). In addition, several major reservoirs in the hill country have been built specifically for hydropower generation. These new and ancient reservoirs contribute to more than 80% of the island’s inland fishery production (Jayasooriya et al. 2006).

Habitat types and forest cover

Climate is the main factor determining the distribution of Sri Lanka’s natural vegetation. Lowland rainforest is the climax vegetation of the southwest wet zone, while montane rainforest and Sub-montane forest is present at higher altitudes. Moist mixed evergreen forest in the intermediate zone of the seasonally dry northern and eastern plains gives way to dry mixed evergreen forest in the dry zone. The arid zones of the island’s north- western and south-eastern extremities are covered with thorn scrub37.

Sri Lanka has about 1.74 million ha of forest covering about 27% of the country (Food and Agriculture Organisation 2010; p.10). However, of these, less than three quarter are likely to be dense forests, since, according a 1999 estimate, sparse forests and forest plantations accounted for 23% and 5% of Sri Lanka’s total forest cover (Bandaratillake and Fernando 2003; p.250). Further details on the country’s forest cover are presented in Appendix 2.

36 More details are provided in the section on Sri Lankan history. 37 A detailed description of Sri Lanka’s main vegetation types is provided in Appendix 2.

47 Biodiversity resources

Despite its relatively small size, Sri Lanka has a variety of ecosystems containing rich species diversity. The country’s exceptional biodiversity resources has been highlighted by its inclusion—along with the Western Ghats of India—among 34 global ‘biodiversity hotspots’ (Mittermeier et al. 2004)38. Sri Lanka’s wet zone harbours most of its endemic species, especially reptiles and amphibians (Jayasooriya et al. 2006). While the dry zone has comparatively lower levels of biodiversity, it provides key habitat for large, charismatic species such as leopards, elephants and sloth bears. Both wet and dry zone forests are extremely important habitats for a variety of species—for example 80% of Sri Lanka’s endemic birds are forest species which are usually unable to successfully colonize non-forested areas (Jayasooriya et al. 2006).

2.2 Socio-economic context

Population

Sri Lanka’s population is about 20.3 million (Department of Census and Statistics 2012; p.40). It is located in the extremely densely populated region of South Asia39, where nearly 24% of the world’s total population lives on less than 4% of the Earth’s land area (Veron 2008; p.9). At 292 people per km2, Sri Lanka is densely populated, although significantly less crowded than countries such as India (357) and Bangladesh (1,102) (Veron 2008; p.7). It is also the least urbanized country in the region—only 15% of Sri Lankans live in urban areas, much less than in Pakistan (35%) and India (29%), and significantly less than the world average of 50% (Veron 2008; p.59)40.

Sri Lanka’s population growth rate has slowed dramatically since the 1950s (Table 2.1). Kelegama (2006) describes how the country has gone through all three stages of the usual demographic transition in a comparatively short period of 70 – 100 years: The first phase, during the first half of the 20th century when there was slow population growth (high levels of mortality offsetting high rates of

38 Sri Lanka’s high degree of species diversity and endemism is described in Appendix 2. 39 South Asia is comprised of 8 countries: Afghanistan, Bangladesh, Bhutan, India, the Maldives, Nepal, Pakistan and Sri Lanka. 40 Details on how Sri Lanka’s population density and extent of urbanisation compare with other South Asian countries are presented in Appendix 2.

48 fertility); the second phase starting soon after independence and continuing through the 1970s when there was high population growth (sharp decline in mortality without an accompanying decrease in fertility); and finally, the third phase, picking up momentum in the 1980s and manifesting itself in the 1990s—once again slow population growth (where decline in mortality was accompanied by low fertility) (Kelegama 2006; p.21).

Table 2.1: Sri Lanka’s population and growth rates (from Veron 2008; p.71, 72; Department of Census and Statistics 2011)

Population Population growth Year Period (thousands) rate (annual, %) 1950 7,339 1950 - 1955 2.82 1970 12,342 1960 - 1965 2.46 1990 17,114 1970 - 1975 2.05 2000 18,714 1980 - 1985 1.39 2005 19,668 1990 - 1995 1.10 2008 20,217 2000 - 2005 0.43

Sri Lanka is a multi-ethnic, multi-religious country. Its main ethnic group is Sinhalese, but it has substantial populations of Tamils and Muslims. Buddhism is the major religion, but other faiths such as Hinduism, Islam and Christianity are practiced widely41.

41 According to the 2001 census, Sinhalese, Tamils and Muslims comprised 74%, 18% and 8% of Sri Lanka’s population, while Buddhists, Hindus, Muslims and Christians comprised 69%, 15%, 8% and 8% respectively (Peebles 2006).

49 Development indicators

Sri Lankans have a life expectancy of 70.8 years and an overall literacy rate of 91%— the highest in South Asia (World Bank 2010). These indicators compare favourably with India (62.9 years and 43%), Pakistan (63.6 years and 50%) and Bangladesh (62.0 years and 61%). Consequently, the country’s human development index42, at 0.755, is the highest in the region43. As will be discussed in section 2.4, these achievements are mainly due to the country’s substantial expenditure on welfare since independence.

System of government

Universal adult franchise was granted in 1931, when Sri Lanka was still a British colony44 (Holt 2011; p.230). At independence in 1948, the Soulbury Constitution, negotiated between the British and the locals, created a British-style parliamentary government (Peebles 2006). New constitutions were written in 1972 and 1978; the latter created a presidential system, with an elected president as head of state and government (de Silva 1981). The country’s unicameral Parliament has 225 seats, where members are elected by popular vote on the basis of a proportional representation system (Peebles 2006). Presidential and parliamentary elections must be held at least every six years. Although Sri Lanka initially had a largely centralised system of government, the 13th Amendment to the 1978 Constitution introduced a system of Provincial Councils in 1987 that resulted in some devolution of power (Peebles 2006; p.5).

42 Composite index measuring human development, including life expectancy, adult literacy rates, school enrolment ratios and per capita GDP. The closer it is to 1, the better the situation. 43 A comparison of Sri Lanka’s development indicators with other South Asian countries is presented in Appendix 2. 44 More details about Sri Lankan history will be provided in section 2.5.

50 2.3 Land-use

Sri Lanka is largely an agricultural country—a little over one third of the island’s total land area is devoted to agriculture, while another one third has been set aside for forest and wildlife (Survey Department of Sri Lanka 2008; p.122). About 80% is land under some form of state control (Table 2.2).

Table 2.2: Land ownership in Sri Lanka as of 1991 (Dent and Goonewardene 1993; p.16)

Ownership ha % of total Private land 17.7 Freehold 1,065,000 16.2 Land grants 101,000 1.5

State land 5,404,000 82.3 Alienated under various schemes 818 – 1,000 12.5 – 15.2 Vested in Land Reform Commissioner 406.000 6.2 Total land area 6,570,000 100

Agriculture

The agricultural sector is the mainstay of Sri Lanka’s rural population. While it only accounts for about 18% of GDP, 65% of Sri Lankans’ livelihoods is largely dependent on agriculture (Kelegama 2006; p.228). The country’s main crops are rice (paddy), tea, rubber and coconut45. Rice accounts for almost half the land under agriculture while coconut accounts for almost a quarter (Table 2.3).

45 The latter three are its main export crops.

51 Table 2.3: Area under principal crops and food crops, 2009 (Department of Census and Statistics 2011)

Crop Extent cultivated (ha, % of total land under % of country’s 2009) agriculture total land area

Rice (paddy) 719,078 44% 11% Tea 221,969 14% 4% Rubber 124,000 8% 2% Coconut 394,836 24% 6% Other 154,340 10% 2% Total 1,614,223 100% 25%

Forests and wildlife reserves

Sri Lanka’s forests and wildlife reserves are managed by two different agencies—the Forest Department (forestry) and the Department of Wildlife (wildlife)— largely due to historical reasons:

The Forest Department’s main functions, at one time, were equally divided between conserving forests and managing forests for supplying the nation’s timber needs…Since its establishment, the Department of Wildlife has focused its attention on establishing National Reserves and Sanctuaries to widen the range of habitats brought under protection and carrying out policing to prevent poaching (Ministry of Environment and Natural Resources 2003). Currently, almost 30% of Sri Lanka’s total land area is reserved and administered by either the Forest Department or Department of Wildlife, as illustrated in Table 2.4 and Figure 2.2.

52 Table 2.4: Extent of protected areas managed by the Forest Department and the Department of Wildlife (from Ministry of Environment Sri Lanka 2012; p.12)

% of total land Protected Area category Area in km2 area Areas managed by the Forest Department National Heritage Wilderness Area46 [N=1] 111 < 1% Conservation Forests [N=65] 962 2% Other Reserved Forests [N =366] 6,307 10% Forest Plantations 799 1% Total areas under the Forest Department 8,180 13%

Areas managed by the Department of Wildlife National Parks 5,262 8% Nature Reserves 571 1% Sanctuaries 3,491 6% Strict Natural Reserves 316 1% Jungle Corridors (Kaudulla- Minneriya) 88 < 1% Total areas under the Department of Wildlife 9,727 16% Total areas under both departments 17,907 29%

46 Also a World Heritage Area.

Forest Department Department of Wildlife

Figure 2.2: Protected areas in Sri Lanka administered by the Forest Department and the Department of Wildlife (from Jayasooriya et al. 2006; p.37-38)

Land under other types of use

In addition to agriculture, forestry and wildlife, land is also used for urban, industrial and cultural purposes (Survey Department of Sri Lanka 2008; p.122). Despite the country’s relatively low urban population growth47, the extent of urban land continues to grow, as evident in urban fringe areas where even prime agricultural land48 is being destroyed to make way for new human settlements.

Furthermore, as mentioned in section 2.1, Sri Lanka has large inland water bodies particularly in the dry and intermediate zones. After the completion of the main Mahaweli49 reservoir systems, the land area under internal water bodies rose to some 2,095 km2 accounting for 4.4% of the country’s total land area (Survey Department of

47 Discussed in section 2.3. 48 Such as paddy, rubber and coconut lands. 49 The Mahaweli irrigation and hydropower project, completed in the late 20th century, is discussed in more detail in the next section.

54 Sri Lanka 2008; p.122). Finally, some land is not utilised for ‘productive purposes’ due to issues such as salinity or rocky outcrops (Survey Department of Sri Lanka 2008).

Having examined Sri Lanka’s biophysical characteristics, socio-economic context and land-use patterns, I now consider its historical context to demonstrate how its use of land and other natural resources have been shaped by the country’s history.

2.4 Historical context

Sri Lanka has a long tradition of continuously recorded history. The two narratives Dipawamsa (4th century) and Mahawamsa50 (5th century) provide a chronology of the Sinhalese kingdom to about 300 A.D., and later texts continue the history to modern times (Peebles 2006). Although these texts are focused on the Sinhalese majority, and begin with mythical sections (e.g. magical connections of the Buddha to Sri Lanka), many later sections have been corroborated by modern research (Peebles 2006). The history of Sri Lanka can be divided into three main eras: (i) ancient kingdoms, from pre- history to about 1600, when the Portuguese established control over parts of Sri Lanka; (ii) the colonial era under the Portuguese, Dutch and the British; and (iii) the post- independence period from 1948 to modern times. The following sections describe key features and events of these eras.

Ancient kingdoms

The first clear evidence of human settlement in Sri Lanka has been dated to 28,000 B.C.—these are some of the earliest human remains found anywhere in South Asia (Peebles 2006). According to the Mahawamsa, the story of the Sinhalese peoples begins with the arrival of its founder Vijaya—a Prince from northern India51—and his 700 companions around the 5th century B.C. (de Silva 1981; p.6). According to de Silva (1981; p.6), there is a kernel of truth in this myth, as the island was probably colonised by immigrants from northwest India and the Indus region, and subsequently from east

50 Dipawamsa and Mahawamsa mean ‘Chronicle of the Island’ and ‘Great Chronicle’ respectively (Department of Census and Statistics 2011; p.8-9). 51 He had been banished by his father for misconduct (de Silva 1981).

55 India52. In addition, Dravidians from South India have come to Sri Lanka intermittently, crossing the narrow Palk Straight—they came from “very early times, although there is no firm evidence as to the date of their arrival” (de Silva 1981; p.12).

By the 3rd century B.C. the northern dry zone was undergoing rapid change as improved irrigation systems supported significant increases in rice production, and Anuradhapura emerged as South Asia’s largest city outside north India (Peebles 2006). A key event in Sri Lankan history was the establishment of Buddhism around 250 B.C., when the Indian Emperor Ashoka sent a mission to the country led by his son Mahinda, a Buddhist monk (Nissan and Stirrat 1990; p.20). The (Sinhalese) King made Buddhism the state religion and it spread swiftly among the island’s settlements, becoming a “potent factor of unification, primarily cultural no doubt, but one which strengthened the process of political unification within the island” (de Silva 1981; p.12). At the height of the Anuradhapura kingdom, an estimated 4 million to 7 million people—90% of the country’s inhabitants—lived in the northern dry zone (Peebles 2006; p.19). From the 7th century A.D. onward, Sri Lanka faced intermittent invasions from India, and South Indian rulers governed Anuradhapura from time to time, although they were subsequently repelled. In the early 11th century, the capital was moved to Polonnaruwa, located 100 km southeast (Holt 2011; p.11).

A key feature of these kingdoms was the massive irrigation system that enabled production of two rice crops each year instead of one (de Silva 2011; p.55). To reduce dependence on unreliable monsoon rains, Sri Lankans constructed canals, channels, water storage tanks and reservoirs “which developed into one of the wonders of the ancient world” (Peebles 2006; p.19). To this day, Sri Lanka’s dry zone is dotted by hundreds of tanks that are still extensively used for agriculture and fishing (Jayasooriya et al. 2006). Furthermore, “Sri Lanka’s traditional social, economic and cultural life has been closely knit with village tanks” (Jayasooriya et al. 2006; p.31).

Sinhalese rulers abandoned the northern dry zone in the 13th century after new invasions, malaria, and perhaps climate change53 brought about a precipitous decline in the population (Peebles 2006; p.27). Sinhalese political centres moved to the wet zone in southern Sri Lanka, and a Tamil kingdom was established in the northern peninsula.

52 The immigrants probably came from around Bengal and Orissa (de Silva 1981). 53 The ‘medieval warm period’ when world temperatures rose.

56 Sri Lankans became increasingly engaged in Indian Ocean trade, first with Arab traders and subsequently with Europeans (de Silva 1981). When Europeans threatened the coasts in the 16th century, there was a Sinhalese kingdom in Kotte (near modern-day Colombo) and another at Kandy in the central highlands (Peebles 2006; p.27).

The colonial era

From the 16th century onwards, Sri Lanka came under the sway of colonial powers that dominated the lucrative trade routes between the Indian Ocean and Europe. The first to arrive were the Portuguese in 1505, who displaced incumbent Muslim traders54 and forced the Kotte Kings to sell cinnamon, areca nuts and gems directly to them (Holt 2011; p.132). By the early 17th century, the Portuguese had conquered the coastal areas, including the Jaffna and Kotte kingdoms; however, the Kandy kingdom survived, mainly because its mountainous terrain and forests enabled Kandyans to launch successful guerrilla attacks on the invaders (Peebles 2006).

The Kandyan King Rajasingha II tried to use the Dutch—the emerging power in the region—to expel the Portuguese, promising to repay the costs incurred, and give the Dutch East India Company a monopoly of the spice trade (Peebles 2006). The Dutch duly obliged, capturing the fort of Colombo, the Jaffna kingdom and other areas held by the Portuguese (Holt 2011; p.189). However, the Dutch, who wanted nothing less than complete control of the spice-producing areas they had captured, refused to return these to the Sinhalese King55 (de Silva 1981). They took advantage of a rebellion against the King in 1665 to more than double their territory, to include the ports of Trincomalee and Batticaloa; like the Portuguese they never managed to conquer the Kandyan kingdom (Peebles 2006; p.44).

Taking advantage of the French occupation of the Netherlands during the Napoleonic Wars, the British sent an expeditionary force against Ceylon56; the Dutch capitulated in

54 Unlike the Persian and Turkic forces who invaded North India to establish empires, the Muslims who came to the coasts of South India and Sri Lanka were mainly Arab traders (de Silva 1981). These traders have been arriving on Sri Lankan shores since the 8th century A.D. (McGilvray and Raheem 2011; p.410). 55 They did this by pricing their expenses at a level above which the King could not afford to pay, and undervaluing the cinnamon, areca nuts, elephants and land revenue they obtained from the lands they controlled. 56 As Sri Lanka had been called since Portuguese times.

57 1796 (Peebles 2006; p.47). By 1798, the British made the entire island, except the kingdom of Kandy, a crown colony (Peebles 2006). In 1815, the British and the Kandyan chieftains deposed the King and signed the Kandyan Convention that “recognized British sovereignty in the island in return for preserving the laws of the country, including the King’s duty to protect and maintain Buddhism, the monastic orders and their temples” (Peebles 2006; p.50). The British ruthlessly crushed a rebellion that erupted two years later, unifying the entire island under a single government for the first time in its history (Peebles 2006; p.53).

Having gained full control, the British began transforming Ceylon into an export- oriented plantation economy. Coffee was their main choice initially—although they also invested in sugar, cotton, indigo and coconuts for export—and coffee plantations spread across the mountainous interior, flourishing from 1840 to 1883 (Peebles 2006). Under the Crown Land Encroachment Ordinance No. 12 of 1840, the government claimed all land to which there were no clear titles, including much of the suitable land for coffee in the hill country (Wenzlhuemer 2008; p.59). Since locals were not willing to put up with the miserable conditions on the plantations, workers were recruited from South India (de Silva 1981). When a leaf fungus decimated the coffee plantations, tea was introduced (de Silva 1981; p.366). The British used the revenue from the plantations to build roads and railroads, improve harbours and establish schools and churches.

The British rewarded their local allies with land grants, titles and administrative offices. Local elites “maintained order and were expected to assimilate to the rulers’ culture, which created an Anglicized elite that was increasingly distant from the people” (Peebles 2006; p.55). More than half of the two million acres of Crown land alienated by the government was purchased by Sri Lankans, some of whom acquired large expanses of land in order to live as a landed aristocracy (de Silva 1981). Some peasants also acquired titles to small plots of land, but many others lost land when the government refused to recognise title to chena57 land. Rice cultivation lagged because little was done to improve irrigation, as British planters preferred to import cheaper Indian rice to feed their labourers (Peebles 2006).

In the first few decades of the 20th century, the British gradually instituted constitutional reforms in response to local agitation. The Donoughmore Constitution of 1931 was the

57 The traditional system of ‘shifting cultivation’.

58 most important, giving “real power to a State Council elected on nearly universal franchise” (Peebles 2006; p.82). This body provided Sri Lankans legislative power over domestic issues 58, excluding areas reserved for the colonial government such as defence, foreign relations, law, justice and finance (Peebles 2006).

Post independence

Sri Lanka gained independence from Britain in 1948 (Holt 2011; p.3). “It was a peaceful transition from the colonial rulers to the Anglicized upper-class elite, who had already been substantially self-governing since 1931” (Peebles 2006; p.95). This elite smoothly adapted to a system of British-style parliamentary democracy instituted by the British before they left. Since independence, politics have been dominated by the United National Party (UNP), which has generally been more to the right of political spectrum, and the Sri Lanka Freedom Party (SLFP) which has identified more with the left (Peebles 2006). In addition there have been parties representing interests of minorities (e.g. the Tamil United Liberation Front or TULF), as well as other small Marxist parties (Spencer 1990; p.2).

Sri Lanka’s economic and social development record since independence has been mixed. On the one hand, as discussed previously, it has an exceptional human development record for a country with its level of income. However, a substantial portion of the population still lives poverty, and the country has just emerged from a ruinous civil war which flared up in 1983 (Wilson 1990; p.503). Kelegama (2006; p.20) describes the country’s post-independence development as “uneven, characterised by slow adjustment to internal and external shocks, missed opportunities and policy errors”. His explanation of this situation—based on three factors—is worth looking at, despite its (mainstream) economic focus.

The first factor is the country’s substantial expenditure on welfare, whose roots can be traced to the universal adult franchise Sri Lanka achieved in 1931 (Holt 2011; p.230). Although this approach certainly contributed to the improvement of social indicators, it was put under stress by rapid population growth in the first three decades after

58 It also provided them with valuable experience of democratic politics.

59 independence, as government expenditures were based more on “competitive political pressures rather than sustainable economic growth” (Kelegama 2006; p.20).

The second factor was Sri Lanka’s closed economy—characterized by ‘export pessimism’—which made it one of the most tightly controlled economies outside the socialist bloc in the early 1970s (Kelegama 2006; p.50). According to Kelegama (2006; p.21), “when economic liberalisation began in 1977, Sri Lanka was a late entrant into the export-led industrial world, and thus could not stage a significant breakthrough like the East Asian Tigers”.

Ethnic conflict—the third factor—further impeded Sri Lanka’s development, due to enormous losses of human life, and significant war expenditures and property losses. The roots of the conflict lie in the resentment among majority Sinhalese with what they saw as favouritism towards the Tamil minority under British administration. Consequently, nationalist Sinhalese governments enacted legislation to redress some of these imbalances. For example, the ‘Sinhala Only’ Act59 enacted by Prime Minister Bandaranaike in 1956, made Sinhala the official language in place of English (Manor 1990; p.602). After several ethnic riots, communal tensions erupted into a fully blown ethnic conflict in the 1980s, with Tamil militant groups pressing for a separate state they called Eelam. The Liberation Tigers of Tamil Eelam, popularly known as the LTTE, subsequently eliminated their more moderate rivals to become the main group fighting the government. Although most of the fighting took place in the north, the conflict also penetrated the heart of Sri Lankan society with Tamil Tiger rebels carrying out devastating suicide bombings in Colombo in the 1990s. After several military gains under its leader, Vellupillai Prabhakaran, the LTTE held swathes of territory in the north and east in the first decade of the 21st century (Peebles 2006; p.1). Hostilities ended in 2009 when government forces overran rebel positions and destroyed the entire LTTE leadership (Holt 2011; p.5). Overall, the conflict resulted in the deaths of more than 70,000 people and caused substantial economic damage (BBC 2012).

Since economic liberalisation in 1977, various governments have pursued policies encouraging export-led industrialization, trying to attract foreign investment in special ‘Export Processing Zones’, and offering incentives to domestic entrepreneurs. In the

59 The legislation was called the ‘Official Languages Act’.

60 1990s, under a ‘stabilization programme’ agreed with the International Monetary Fund, there was further export promotion, privatization of state enterprises and major poverty alleviation schemes (Kelegama 2006). There have also been numerous infrastructure development projects, mostly financed through foreign aid. An example is the Accelerated Mahaweli Development Programme, a “billion dollar land settlement-cum- hydroelectric project that was largely donor-funded, and involved such an extraordinary inflow of concessionary aid that accountability and efficiency became a low priority” (Kelegama 2006; p.110). Apart from its deleterious macroeconomic impacts, this programme had significant environmental consequences and displaced large numbers of people, whose resettlement process has been largely mishandled. These issues will be discussed in more detail in subsequent chapters.

Conservation history

I now trace Sri Lanka’s conservation history related to the above three periods, illustrating the country’s changing conservation paradigms, and drawing attention to key factors—such as the influence of Buddhism—that still permeate Sri Lanka’s conservation ethic.

The era of the ancient kingdoms

Sri Lanka had extensive forest cover during the era of the ancient kingdoms 60, despite the country’s relatively advanced hydraulic civilisation (Nanayakkara 1987). Having embraced Buddhist philosophy61, Sri Lankans generally tried to live in close co- existence with nature (Environmental Foundation 2009). In fact, there is evidence that Buddhism was instrumental in establishing the concept of the ‘public trust doctrine’, an important element of modern environmental legislation. According to this concept, “the government is deemed to be the trustee of the country’s environment and natural resources; these resources must be held or used by the government for the benefit of its people” (Environmental Foundation 2009; p.32). In Sri Lanka, this type of thinking has

60 Mountainous regions and the wet zone were covered in dense forests while agriculture was practiced in the plains, mainly in the dry zone. 61 In fact, the Buddha himself achieved enlightenment in a forest grove.

61 been associated with the establishment of Buddhism in the country, where, in a famous sermon, Arahat Mahinda62 told King Devanampiya Tissa—who was on a hunting trip— that:

O Great King, the birds of the air and the beasts have as equal a right to live and move about in any part of the land as thou. The land belongs to the people and all living beings; thou art only the guardian of it (Environmental Foundation 2009; p.34). Accordingly, chronicles such as the Manavamsa, and writings of British administrators and modern archaeologists record a range of Kings’ ‘conservation activities’ 63 . According to a report by British Forester J.R. Ainslie in 1920: “in the times of the Sinhala Kings and before the arrival of the Europeans, the King of Ceylon himself was generally considered the owner of the forest lands, but from time to time large grants of these were made to individuals, to various Buddhist communities” (Nanayakkara 1987; p.17). Almost every province of the Kandyan Kingdom had several ‘protected areas’ “where any kind of cultivation, felling of trees, hunting or fishing were strictly prohibited, and punishable by a heavy fine” (Ministry of Forestry and Environment 1999; p.6). The Kings appointed kele korales (forest officers) whose duties included protecting royal trees and preventing poaching (Jayasooriya et al. 2006).

The colonial era

During the colonial era Sri Lanka’s forest resources were exploited heavily, especially for timber and ‘hunting for sport’ (Uragoda 1994). However, there were also some positive developments, given that Sri Lanka’s current PA system and legislative framework took shape during British rule. The three colonial time periods are now explored briefly, although information on the Portuguese and Dutch periods is limited.

During the Portuguese period, coastal timber was utilised on a large scale, while elephant tusks and deer antlers and hides were exported (Nanayakkara 1987). (Nanayakkara 1987; Jayasooriya et al. 2006). The Dutch Governor Cornelius Simons has recorded in his memoirs how valuable timber such as ebony was exploited:

62 As stated previously, Arahat Mahinda brought Buddhism to Sri Lanka. 63 Some examples are presented in Appendix 2.

62 Slaves of the company as well as natives were employed for timber felling. The Master Attendant was responsible for supervision. Precious bushes and trees suitable for timber were marked separately and they may not be cut down and mixed up with firewood (Nanayakkara 1987; p.18). There was massive destruction of Sri Lanka’s forests and wildlife during the initial stage of the British colonial rule, as large tracts of forests in the central hills were cleared to establish coffee plantations64 (Peebles 2006; p.70). Consequently, while about 84% of the country had been forested in 1880, this proportion had dropped to 70% by 1900 (Kariyawasam 2007). Furthermore, many British ‘planters’65 and rural administrators took to hunting as a pastime, given the plentiful game within the extensive tracts of jungle still remaining. Uragoda (1994) describes how elephants—a popular target—were destroyed on a massive scale:

Men such as Major Rogers and Sir Samuel Baker have left behind a legendary trail of slaughter hard to beat in any country. The killings were counted in the thousands. When Major Rogers, the Assistant Government Agent of Badulla died in 184566, he was 41 years old. In his 12 years at Badulla, he had killed 1,200 elephants” (Uragoda 1994; p.100). Despite such actions the British also laid the foundations for wildlife and forest protection in the country. Joseph Hooker, the celebrated Head of Kew Gardens, wrote to the British government in 1873 after a visit to Sri Lanka, declaring that tea plantations would “inevitably lead to the destruction of the forests in those regions where it is of paramount importance that the forests be preserved” (Nanayakkara 1987; p.21). Although he was unable to prevent much of the destruction, he did succeed in reserving Crown land more than 5,000 feet above sea level to safeguard Montane Forest ecosystems (Sahajananthan 1987). In 1885, the colonial government passed a Forest Ordinance, and in 1887, appointed a Conservator of Forests, who was paid twice that of other Department Heads in order to procure the best available talent (Nanayakkara 1987). It also passed legislation, such as the 1891 Ordinance to “prevent wanton destruction of elephants, buffaloes and other game” that required licenses to shoot or capture specific animals, and the landmark 1937 Fauna and Flora Protection Ordinance that provided the basis for establishing National Reserves (Uragoda 1994). The latter is

64 As noted previously, these were subsequently replaced by tea plantations. 65 I.e., those who managed plantations. 66 Curiously he was struck by lightning, an occurrence which many Sri Lankans, especially those who are Buddhists, believe was divine retribution for his actions.

63 still in force today, albeit with several rounds of amendments. These, and other conservation-related legislation, will be discussed further in subsequent sections.

The post-independence era

The most significant conservation-related event immediately after independence was the establishment of a separate Department of Wildlife in 1949 to manage National Parks and other PAs (these had been managed by the Forest Department, under the Conservator of Forests)67. Since then, Sri Lanka has significantly expanded its PA system. For example, several new parks were established along with large-scale irrigation schemes that affected habitats of large species such as elephants. Figure 2.3 illustrates the growth of Sri Lanka’s PA network:

1400

1200 Department of Wildlife

Forest Department 1000

800 000 hectares) 000 ‘ 600

400 Total area ( area Total

200

0 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990

Year

Figure 2.3: Growth of Sri Lanka’s protected area system in the forestry and wildlife subsectors (from Green and Gunawardena 1997)

67 Sri Lanka became the first country in Asia to have such a division (Holt 2011; p.1).

64 As illustrated in Figure 2.3, the bulk of the forest reserves within the forestry sub-sector were established in the 1920s, although a number of smaller reserves were demarcated in subsequent decades. Within the wildlife sub-sector, the country’s PA network has grown progressively since the enactment of the Fauna and Flora Ordinance in 1937. Sri Lanka has also entered into a number of international agreements relating to biodiversity conservation, such as the Convention on Biological Diversity in 1992 (Jayasooriya et al. 2006). Sri Lanka’s PA system, the main governmental and non- governmental players in its conservation landscape, and its conservation legislation will be discussed in further detail in subsequent sections.

History of land use

The landscape approach stresses that conservation should be achieved across the wider landscape; consequently a country’s pattern of land-use significantly impacts its conservation efforts. This phenomenon is especially pertinent in the Sri Lankan context given the country’s different approaches to land-use during the various phases of its history. Moreover, current systems of land management have been shaped by these various influences, which, in turn continue have a significant impact on the country’s biodiversity and other natural resources. With this in mind, I now consider the history of land-use in Sri Lanka.

The era of the ancient kingdoms

The longstanding belief that the King was the sole ‘owner’ of land in ancient and medieval Sri Lanka has been challenged by de Silva (1981; p.36):

No doubt he had claims over most of the land in his kingdom, but this did not amount to anything approaching ‘fixed proprietary exclusiveness’. Implicit in the land grants of these centuries is the recognition of the ‘rights’ of individuals with regard to land. In none of these grants is there a mention of the king’s prior consent being a condition to alienation of land by individuals, while on the contrary some inscriptions provide evidence of kings actually buying ‘property’ for the purpose of subsequent donation.

65 Given the scarcity of evidence, it is difficult to get a clear picture of the land tenure system in Sri Lanka during the era of ancient kingdoms, but scholars have pieced together some of its characteristics based on available information:

• Abandoned and ownerless land belonged to the King. Where land was not cultivated or occupied, the King had prior rights to forests and timber, animals for hunting, and natural resources such as mines. This did not necessarily mean, however, that the people had to ‘buy’ land from the King to open up new cultivation (de Silva 1981; p.37).

• Inscriptions from around the 9th century A.D. refer to a form of tenure known as divel. These were property rights bestowed on an individual in return for services rendered to the grantor, and were terminable on the death of an employee or at the will of the granting authority. The recipient of a divel holding received revenue which the king had enjoyed earlier (de Silva 1981; p.38).

• Kings often granted land to chieftains for their loyalty and services, which was called Nindagam land. He also donated land to Buddhist temples (Viharas) and Hindu temples (Devalas); such land is known as Viharagams or Devalagams (Rathnayake 2007; p.54-62).

• Inscriptions, mainly after the 9th century A.D., refer to a type of tenure known as pamunu or paraveni, which conveyed the meaning of heritable right in perpetuity. Individuals could acquire pamunu property in at least three ways, namely royal grant, purchase and inheritance (de Silva 1981; p.37).

In Sri Lanka’s hydraulic civilisation, where irrigation was crucial for cultivating crops, especially in the dry season, water was a precious commodity that could be bought and sold as it passed through tanks, canals and fields. The ‘owners’ of tanks imposed a charge for water that passed through and in turn paid for water that came in—the King was the prime beneficiary of such levies because he controlled the largest tanks68 (de Silva 1981; p.36-37). In addition, there was a system of compulsory social service, or rajakariya69, where each adult was required to provide forty days of labour. This service

68 These were largely built and maintained by the state. 69 Literally, ‘service to the King’.

66 was typically used in capital works for the King, such as the maintenance of the irrigation system (Zubair 2005; p.168).

Several indigenous systems—such as Praveni, Thattumaru and Kattimaru—evolved to deal with land fragmentation and foster more equitable access to resources within villages, some of which survive to this day70. As most villagers only owned small plots of land (or did not have any land), however, chena cultivation became important. According to this practice, each farmer encroached on a few acres of state jungle to cultivate crops—such as grains, green gram, yams, bananas, onions and chillies— depending solely on rainwater from the northeast monsoon. Villagers generally respected the right to each other’s chenas; even if chena land was abandoned, other villagers would not cultivate this area without obtaining consent from the original cultivators (Rathnayake 2007; p.54-62).

Geiser (1995) describes indigenous land management as a decentralized system where the basic units, the villages (gamas), were interlinked through a huge irrigation infrastructure. As he explains further, there was a balanced combination of irrigated and rain-fed agriculture within villages, along with forestry and animal husbandry:

Each of the decentralized land management units (gama or village) included parts of lowland as well as highland and thus followed the soil catena: the valley bottoms were used for paddy cultivation; water was provided from a minor tank (a dam across the valley storing water). A wide range of paddy varieties was available to suit local soil conditions. The homesteads were located above the irrigation canal. The area on the adjacent highland was utilized in a multipurpose manner: rain-fed agriculture in the context of slash and burn (chena); pastures for cattle and buffaloes; timber for construction; firewood etc. Growing population was absorbed by land colonization, i.e. the establishment of new gamas. In the course of time, the basic concept has thus been replicated very often throughout the Dry Zone as is evident from topographic maps. In order to increase and ensure the supply of water to the minor tanks, canals were constructed to catch water from nearby streams. Such canals subsequently interlinked many village tanks and also supplied water to the emerging centres of power (e.g. the city of Anuradhapura).... Later, large dams were constructed across major rivers. These reservoirs however were generally not used for direct irrigation, but for supplementing village tanks; they functioned as intermediary storage basins, as a kind of "Central Bank" (Geiser 1995; p.188).

70 Some of these are described in Appendix 2.

67 Geiser (1995) emphasises that the system was not utopian—i.e., based on socially homogenous, egalitarian village communities—because the ruling class appropriated surplus to maintain the complicated central infrastructure and administrative system. Nonetheless, he argues that, as a result of the decentralized organization of the management of land resources, central authorities provided considerable leeway for land-related decisions to be made at local (i.e. village) level. Moreover, these balanced, decentralised systems of land-use, combined with the environmental ethic described previously, usually resulted in sustainable use of natural resources (Geiser 1995).

The colonial era

As described above, although the Portuguese and the Dutch ruled parts of Sri Lanka, it was the British who conquered the entire island, transforming it into an export-oriented plantation economy. As coffee plantations were established in the central hills, colonial planters and overseas investors71 became concerned about the security of land titles. This issue arose because the early coffee plantations were often established on former chena lands to save capital and labour otherwise required to clear the land (Wenzlhuemer 2008). As a result, there were frequent native claims to these areas, challenging the planters’ land titles. In order to address concerns of this important group, the colonial government enacted the Crown Lands Encroachment Ordinance of 1840 to define “a set of criteria for determining Crown and private rights to ‘waste lands’ and to create the machinery for ejection and/or punishment for encroachers” (Bandarage 1983; p.93-94). Clause 6 of the Ordinance, which specifically addresses chenas, is as follows:

And it is further enacted that all Forest, Waste, Unoccupied and Uncultivated Lands shall be presumed to be the property of the Crown; until the contrary thereof be proved; and all chenas and other lands which can only be cultivated after intervals of several years shall, if the same be situated within the Districts formerly comprised in the Kandyan provinces […] be deemed to belong to the Crown and not to be the property of any private person claiming to be the same against the Crown (Wenzlhuemer 2008; p.59). Under the Ordinance, the burden of proving land claims rested solely on claimants. As native claimants usually found it difficult to produce title deeds or other proof of

71 Overseas investors were mainly from Britain.

68 ownership that were acceptable to the colonial authorities, the Crown appropriated large extents of land which was sold to the planting community (Bandarage 1983). Wenzlhuemer (2008; p.59) points out that the law completely disregarded existing traditions of land tenure in the Kandyan regions and was designed solely to ensure planters’ security of land-titles. Although competition for land between planters and peasants subsided with the drop in coffee production following the coffee blight, demand for plantation lands intensified in 1890s with the subsequent expansion of tea estates. The British government responded with the “Waste Lands Ordinance of 1897”, which strengthened the provisions of 1840 Ordinance (Wenzlhuemer 2008; p.109).

The colonial government’s official policy was that chena was a ‘primitive form’ of cultivation that destroyed the soil, produced crops of ‘low nutritional value’, and made cultivators indolent because of its ‘land intensive nature’ (Bandarage 1983; p.101). Bandarage (1983; p.101) notes, however, that while chena was highly irrational from the viewpoint of planters and colonial authorities, it was the “safety valve to peasant subsistence”, since when irrigation and paddy harvests failed, or taxes on paddy fields became onerous, peasants often turned to shifting cultivation to stay alive. This was understood by the more enlightened British administrators, especially when agrarian poverty began to surface by the turn of the century as a result of colonial land policies (Jayawardene 1990). The vision of liberal minded British Governors like Sir Hugh Clifford led to the establishment of the First Land Commission of 1927, which in turn reoriented land policy towards the redistribution of Crown land among the locals (Jayawardene 1990). Accordingly, the Land Development Ordinance of 1935 gave legal recognition to this policy, institutionalising procedures for mapping out (i.e. assessing needs and surveying land), alienation (i.e. allocating state land to individuals and communities), and administration of land (Jayawardene 1990). It would also provide the foundation for later efforts of post-independence governments to improve the situation of peasant farmers.

Therefore, in colonial times, the clearing of land for plantations resulted in deforestation and land degradation in the hill country. Moreover, the British colonial authorities enacted legislation that dispossessed many Sri Lankans—especially chena farmers—of land they had cultivated for generations. Although the British ultimately enacted

69 important environmental legislation, on the whole, colonial land policies caused the destruction of valuable biodiversity and forest resources.

The post-independence era

After independence, successive governments have placed special emphasis on the development of rural areas and agriculture. In fact, scholars such as Bastian (2003; p.1) claim that “protecting the Sinhala peasantry as a class, and improving the lot of the small holder paddy growing farmer” has “became a central idea in the post-colonial state building project, one which has multiple objectives”. He notes that:

Protecting the peasantry and paddy production was not just another development activity. It encompassed a whole range of objectives. It was a means of redressing a past grievance and correcting a colonial injustice. It was expected to improve the living standards of the vast majority of the population. It was a means of protecting a way of life that truly reflects what Sri Lanka is. In other words, it was a principal foundation of the post-colonial state (Bastian 2003; p.5). The 1958 Paddy Lands Act was enacted as part of this process. The main objectives of the Act and its subsequent amendments were to regulate landlords’ authority over tenants, regulate rents paid by tenants and provide security of tenure of a permanent and heritable nature. It also attempted to prevent further fragmentation of paddy lands and encourage the consolidation of small-sized holdings. Finally, it regulated interest rates on loans made to paddy cultivators, fixed wages for agricultural labourers, and established cultivation committees to organise paddy cultivation (Bastian 2003; p.10).

Successive governments since independence have also taken steps to open up land for settlement, especially in the dry zone. These initiatives were essentially a continuation of policies instituted by the colonial government in decades prior to independence, which culminated in the Land Development Ordinance of 1935. D. S. Senanayake, the first Prime Minister of independent Ceylon, took a keen interest in developing of peasant agriculture, and his government launched land development and irrigation schemes, especially in the dry zone (Peebles 2006). An example is the Gal Oya scheme, the first major project since the days of the Polonnaruwa Kings (de Silva 1981; p.505). Later governments continued these efforts, which culminated in the Accelerated

70 Mahaweli Development Programme of the 1980s. These schemes involved the settlement of farmers from other parts of the country in newly irrigated lands.

The other set of important post-independence land reforms were those that established a ceiling for private ownership of agricultural land, enacted by the socialist-oriented coalition governments of the 1970s. While paddy land was virtually unaffected by these reforms, over half of tea land, about a third of rubber land and a tenth of coconut land was acquired by the state, encompassing 22% of the total area under agriculture in Sri Lanka (Peiris 1978; p.617). Most acquired lands were managed by the state, and their productivity declined as politicians began to use them as a means of doling out patronage to their supporters, especially in terms of employment opportunities (Bastian 2003; p.11). Consequently, more market-oriented governments of the 1990s began privatising these plantations, a process that has largely been completed (Bastian 2003; p.14).

Having examined Sri Lanka’s conservation and land-use history, I now look at its conservation-related institutions—both governmental and non-governmental—that have been deeply influenced by these historical events.

2.5 Conservation-related institutions

Governmental sector

The Ministry of Environment is the main government organisation responsible for managing Sri Lanka’s environmental resources. Its mission is to: “provide leadership to manage the environment and natural resources in order to ensure national commitment for sustainable development for the benefit of present and future generations” (Ministry of Environment Sri Lanka 2010; p.3). Key agencies under its purview are the Central Environmental Authority, the Forest Department, the Marine Pollution Prevention Authority, the State Timber Corporation, and the Geological Survey and Mines Bureau. The Department of Wildlife, which used to be under its purview, was moved to the Ministry of Agrarian Services and Wildlife in 2010 (Presidential Secretariat 2010).

71 The Central Environmental Authority

The Central Environmental Authority manages the overall legislative framework for the protection, conservation and management of the environment in Sri Lanka. It was established under the National Environmental Act of 1980 to make provisions for (a) the protection, management and enhancement of the environment; (b) the regulation, maintenance and control of the quality of the environment; and (c) the prevention, abatement and control of pollution. Its responsibilities include implementing procedures related to Environmental Protection Licenses and Environmental Impact Assessments, compliance monitoring for approved projects, initiating legal action against violators of the National Environmental Act, and maintaining environmental quality by developing standards, norms and criteria (Environmental Foundation 2009).

The Forest Department and the Department of Wildlife

As described previously, the Forest Department and the Department of Wildlife are the key agencies responsible for managing Sri Lanka’s forests and wildlife reserves. The Forest Department was established in 1887 by the colonial government, which was concerned about the rapid destruction of the country’s valuable timber reserves. R. Thompson, who had worked with the British Indian Forest Service, was appointed the first Conservator of Forests of Ceylon (Nanayakkara 1987; p.20). The government established the Department of Wildlife in 1949, recognising the need for a separate department to manage the increasing number of wildlife reserves in the country. The Conservator of Forests continued as the Acting Warden, until the new Wildlife Warden, C.W. Nicolas, was appointed in October 1950 (Department of Wildlife Conservation website, accessed January 2011). The department is now responsible for managing a PA network consisting of 20 National Parks, 4 Nature Reserves, 3 Strict Nature Reserves, 1 Jungle Corridor and 66 Sanctuaries72. The vision, mission, and organisation structure of the two departments are summarised in Appendix 2.

72 The vision, mission, and organisation structure of the Forest Department and the Department of Wildlife are summarised in Appendix 2.

72 Other government departments that impact on conservation

There are several other government departments and agencies whose activities impact conservation. The Land Commissioner General’s Department, which controls and administers lands belonging to the state, is responsible for conserving and developing these lands, issuing land to various individuals and institutions when required, and formulating land-use policies (Land Commissioner General’s Department website, accessed January 2011). These activities impact PA management, especially because they determine land-use in areas near PAs.

The Ministry of Agriculture—especially the Department of Agriculture under it—also impacts conservation because many Sri Lankan PAs are located within a matrix of agricultural land. Its responsibilities include increasing domestic agricultural productivity, adopting environmentally-friendly farming technologies and enhancing living standards of farming communities (Ministry of Agricultural Development and Agrarian Services 2007). Despite some efforts to minimise the impacts of farming on the environment, some Ministry policies have negatively impacted biodiversity. For example, its fertiliser subsidy often leads to the excessive use of agrochemicals, which in turn cause soil and water pollution.

Finally, the Mahaweli Authority of Sri Lanka, which manages the largest, multipurpose, rural development programme ever undertaken in Sri Lanka, tries to address certain environmental issues in areas under its purview. While its main objectives include irrigation, hydropower generation, settlement of landless poor and flood control, it acknowledges that “the success of this entire programme depends on the continued flow of adequate quantity and quality of water into the reservoirs and sustainable management of natural resources on the whole Mahaweli Basin” (Mahaweli Authority of Sri Lanka website, accessed January 2011). Consequently, the Authority institutes measures to conserve forests in watersheds and conducts environmental education programmes, targeting groups such as school children, local politicians and farmers.

73 Non-governmental sector

Sri Lanka’s biodiversity conservation action plan acknowledges that NGOs must play an important role in conserving the country’s biodiversity (Ministry of Forestry and Environment 1999; p.83). Several national and international organisations play an active role in Sri Lanka’s conservation landscape.

International organisations

International donor agencies have been providing assistance to Sri Lanka since it gained independence; over the years, several have been directly involved in the country’s conservation sector. For example, the Protected Area Management and Wildlife Conservation Project was implemented from 2001-2006 by the Asian Development Bank, supplemented by funds from the Global Environmental Facility and the Government of the Netherlands (Samaranayake 2007). The project had four main components: (a) enhancing the Department of Wildlife’s institutional capacity; (b) participatory management of pilot PAs; (c) collaborative conservation planning; and (d) sustainable financing for community partnership building (Samaranayake 2007; p.v).

The World Bank has also been engaged in environmental programmes since the 1990s. For example, the International Development Agency-financed Environmental Action 1 Project strengthened technical environmental management capacity with respect to urban and industrial issues. Building on this project, the World Bank’s Country Assistance Strategy (2009 – 2012) emphasises the need for continued capacity building in environment and natural resource management (World Bank 2010; p.4).

The IUCN—which established a local branch in Sri Lanka in 1988—has been involved in a number of projects related to conservation and management of critical habitats and biodiversity, environmental policy support, and environmental education (IUCN Sri Lanka website, accessed January 2011). Most of these have been undertaken in collaboration with various government agencies and local NGO partners.

74 Local organisations

The handbook on Environmental Law, developed by the Environmental Foundation and the United Nations Environmental Programme describes the role of local environmental NGOs as follows (Environmental Foundation 2009; p.96):

These organisations play a watchdog role guarding against violation of the law and regulations and providing an important monitoring role that supplements the government’s role in monitoring and supervision. A significant body of environmental litigation has developed in Sri Lanka due to the intervention of environmental organisations. The courts in Sri Lanka have recognised the locus standi of a person who can show genuine interest in the matter complained of and who has come before court as a public spirited person, concerned to see that the law is obeyed in the interests of all. Several NGOs focus on environmental and conservation related issues. The Environmental Foundation Limited (EFL) monitors implementation of environment- related legislation and litigates against abuses by the government or other parties. Organisations such as the Wildlife and Nature Protection Society (WNPS)—which was established during British times—focus specifically on wildlife-related issues. Others such as the Centre for Conservation Research and the Green Movement of Sri Lanka—a consortium of over a hundred community based organizations involved in natural resource management—have broader goals, such as developing solutions to reconcile conservation with local community development. Further information on some of Sri Lanka’s key conservation-related NGOs is provided in Appendix 2.

75 2.6 Legislative and policy framework

As described in Section 2.5, Sri Lanka has a rich tradition of environmental conservation based on Buddhist teachings and local traditions. The country’s constitution includes clauses relating to the Government’s and citizens’ conservation- related responsibilities (Jayasooriya et al. 2006; p.3):

The State shall protect, preserve and improve the environment for the benefit of the community [Article 27.14]. The exercise and enjoyment of rights and freedom is inseparable from the performance of duties and obligations, and accordingly it is the duty of every person in Sri Lanka to protect nature and conserve its riches [Article 28F]. Indeed, key components of Sri Lanka’s environmental legislation can be traced back to British times, as described previously. For example, the Timber Ordinance was introduced in 1848 to preserve forests for timber production, while the 1885 Forest Ordinance afforded some protection of forests, primarily for sustainable wood production but also provided limited protection of wildlife in forest reserves (Environmental Foundation 2009; p.24). This was developed further in 1907 with the Forest Ordinance No. 16, which had provisions for establishing reserved forests and village forests and for controlled exploitation of timber. In 1938 the clearing of forests above 5,000 feet was prohibited.

In 1937, the Fauna and Flora Protection Ordinance (FFPO) was passed to consolidate existing laws related to game protection; it also promoted the declaration of national reserves and sanctuaries (Jayasooriya et al. 2006; p.2). This ordinance—which is still in force after several amendments—is the basis of biodiversity protection in Sri Lanka (Saparamadu 2006; p.125). Amendments in 1964 and 1993 expanded PA categories, which now comprise Strict Natural Reserves, National Parks, Nature Reserves, Jungle Corridors, Marine Reserves and Buffer Zones (Pethiyagoda 2000; p.27). Sanctuaries established in the terms of the Fauna and Flora Protection Ordinance may be declared even over private lands (Environmental Foundation 2009; p.90).

The National Environmental Act of 198073—administered by the Central Environmental Authority—provides the overall legislative framework for the protection, conservation

73 Amended in 1988 and 2000.

76 and management of Sri Lanka’s environment. It adopts three primary approaches (Environmental Foundation 2009; p.77-83):

• Environmental protection: Regulating the discharge, deposit or emission of waste, especially by industrial and commercial undertakings;

• Environmental quality: Prohibiting any person from carrying out polluting activities such as: polluting inland waters, the atmosphere, soil or the surface of any land, making or emitting excessive noise and the disposal of litter; and,

• Environmental assessment and the approval of projects: Requiring Environmental Impact Assessment (EIA) approval for ‘prescribed projects’ listed in the regulations framed under the Act (Environmental Foundation 2009; p.32).

Further information on key environmental laws is presented in Appendix 2.

Sri Lanka has ratified several international agreements related to biodiversity conservation such as the Convention on Biological Diversity, the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), the UNESCO74 Man and the Biosphere Programme and the World Heritage Convention (Environmental Foundation 2009). These agreements influence state legislation and policy, as well as require the government to take certain actions to comply with the provisions contained in the agreements. For example, as required by the Convention on Biological Diversity, Sri Lanka prepared a Biodiversity Conservation Action Plan in 1998 (Ministry of Environment and Natural Resources 2003). Appendix 2 presents information on environmentally related international conventions, agreements, treaties and protocols that Sri Lanka has ratified.

74 United Nations Educational, Scientific and Cultural Organization.

77 2.7 Key conservation-related issues in Sri Lanka

Sri Lanka’s approach to conservation

Given the British colonial influence in establishing its current PA network, Sri Lanka continues to adopt a largely ‘traditional’ approach to conservation. This becomes evident when we assess its approach based on the characteristics of the traditional approach (described in Chapter 1).

The idea of nature as static

Although—as pointed out by Jayasooriya et al. (2006; p.12)—“Sri Lanka has been impacted by anthropogenic impacts for at least four thousand years, and few, if any, pristine environments exist”, the initial impetus in establishing nature reserves in the country during British times was motivated by a conviction that nature is static and should be preserved in its ‘pristine’ form. This view has endured within conservation- related government agencies—despite acknowledging the need to adopt an ecosystem- based approach, they have not adequately incorporated it into their decision-making processes. An example is the Department of Wildlife’s lack of concern with land-use near reserves. Although some parks have buffer zones, enforcement of land-use regulations within those zones is extremely lax, often catering to local political interests. Furthermore, there is no planning related to potential climate change impacts on the PA network. Jayasooriya et al. (2006; p.243) describe the department’s outlook as follows:

The terms such as PA network and network of forest reserves are often used superficially or out of context and the efforts to establish functioning networks of PAs to ensure long-term conservation of country’s biodiversity have rarely become apparent... In reality, what has taken place during the past several decades was the fragmentation of the previously existed network of forests, followed by peripheral degradation of the forest fragments resulting in increasing isolation of such areas. The view that nature is static has led to a reductionist approach to land management. For example, the government still assigns land-use categories such as ‘urban land’, ‘paddy land’, ‘forest land’ and ‘national parks’ which are managed by different departments and/or ministries that typically do not consult with each other when

78 developing land-use policies. Dalal-Clayton and Dent (2001; p.351-352) describe what they call an ‘imbroglio of institutions’ involved in land management:

A key issue is the fragmented and even jealous nature of the sectoral institutions… Twenty-six central government institutions are involved in land use planning and management and a new tier has since been created at provincial level. Thirteen different government departments and six research institutes are responsible for supplying primary data There are simply too many ministries, line departments, and statutory bodies responsible for the management of the land. Their authorities overlap. Decisions and policies are commonly contradictory. They do not cooperate. Many institutions lack the capacity to perform their assigned roles and, although several recent externally funded projects have included an institution-building component, there is a chronic shortage of effective professional staff. Agencies sometimes engage in ‘turf wars’. An example is the often-tense interaction between the Department of Wildlife and the Forest Department. The root of the problem lies in each department’s founding objectives: the Forest Department was set up to supply the nation’s timber needs, while the Department of Wildlife was established to protect sensitive habitats and prevent poaching (Ministry of Environment and Natural Resources 2003). The Department of Wildlife’s lack of a ‘profit motive’ and the Ministry of Environment’s budgetary practices have resulted in most national parks, even those with high visitation rates, being chronically short of funds75.

The perception that people are apart from nature

In the process of setting up their hunting reserves, which later became national parks, the British had to ‘resettle’ the people living in those areas. For example, the villages of Kumana and Patanangala were relocated when setting up the Yala Reserve, while some local villagers were ejected when the Wilpattu West Sanctuary and the Udawalawe park were established (Uragoda 1994; p.80).

In modern times, interactions between Department of Wildlife officials and local communities have often been adversarial, and largely related to the prevention of poaching (Kuruwita 2011). While some locals, such as park guides and jeep drivers,

75 Park fees and other forms of revenue have to remitted back to the Central Treasury, which then reallocates funds to each park according a pre-determined, woefully inadequate, budget allocation.

79 benefit from national parks with high visitation levels, these benefits are not widespread within the community. Although villagers engage in customary resource use such as chena cultivation and livestock grazing on Forest Department land, these activities are technically illegal (de Silva and de Silva 2007; p.217).

Conservation-related issues in Sri Lanka

As in other developing countries, Sri Lanka has many conservation-related issues, many which have arisen because of the country’s ‘traditional’ approach to conservation. I now consider some of the key issues.

Loss of habitat and consequent loss of biodiversity

At 261 persons per km2, the Western Ghats and Sri Lanka hotspot has the third highest population density among the 34 ‘biodiversity hotspots’ (Mittermeier et al. 2004). According to Cincotta et al. (2000), if population numbers are examined in isolation of other factors, the three hotspots with the most elevated risks are the Western Ghats and Sri Lanka, the Philippines and the Caribbean. Within Sri Lanka, the population density in the biodiversity-rich wet zone is 650 per km2, while that in the dry zone is only 170 per km2 (IUCN Sri Lanka 2007). Consequently, the main threats to the country’s remaining rainforests are anthropogenic in origin, and include habitat fragmentation, invasive species and edge effects. Although 27% of Sri Lanka is still forested, much of this forest is in the biodiversity-poor dry zone. Only 5% of the wet zone— encompassing an area of 800 km2—now contains closed-canopy rainforest, the remainder having been cleared for cinchona, coffee, tea and rubber cultivation.

Logging has led to much forest destruction in recent times, especially since chainsaws, skidders and loaders were used in mechanised logging conducted in the 1960s, instead of traditional methods using hand implements and elephants (Gunatilleke et al. 2008; p.43). This continued until 1989, when a moratorium was placed on all felling of trees in natural forests in an effort to conserve important watersheds and biodiversity (Jayasooriya et al. 2006; p.47). In addition, many forests managed by the Forest Department have now been designated as ‘conservation forests’ where no commercial

80 logging is allowed. Table 2.5 illustrates the rapid deterioration of Sri Lanka’s forests through British times to the present:

Table 2.5: Sri Lanka's closed canopy forest cover (from Environmental Foundation 2009; p.55; Food and Agriculture Organisation 2010; p.10)

Year % of forest 1881 cover80% 1956 44% 1983 27% 1992 24% 2010 21%

Sri Lanka’s expanding human population has resulted in significant encroachment onto state forests. For example, landless villagers living adjacent to wet zone forests encroach on these areas to plant crops such as tea, rubber and cinnamon, particularly when their market prices rise. On an average, the extent of encroachment by a household in the wet zone is about 0.5 ha (Jayasooriya et al. 2006; p.48). Increasing human populations near PAs also contribute to the spread of invasive species into parks. For example national parks such as Udawalawe and Bundala have been severely affected by invasive species such as Lantana camera (Gunatilleke et al. 2008; p.45).

The loss of habitat, combined with other anthropogenic factors, has adversely impacted Sri Lanka’s biodiversity, especially endemic species. For example, according the IUCN Red List of threatened fauna and flora of Sri Lanka, among Sri Lanka’s total inland indigenous vertebrate species, 223 (33%) were nationally threatened (IUCN Sri Lanka 2007). Sri Lanka’s traditional approach to conservation has contributed to habitat loss. The government’s reductionist approach of categorising land, managed by different departments and/or ministries that often have conflicting goals, has resulted in destructive land-use practices near PAs. For example, the Forest Department promoted mechanised logging until the late 1980s, even in biodiversity-rich pristine rainforests such as Sinharaja (Gunatilleke et al. 2008; p.43). Furthermore, as the country’s population has grown rapidly, departments responsible for increasing agricultural output have promoted practices that achieve increased crop yields in the short term, but are environmentally destructive over time. These practices, which have precipitated a drastic loss of ecosystem services, are discussed in more detail in the next section.

81 Loss of ecosystem services

In Sri Lanka, the side effects of human activities such as pollution and deforestation have resulted in the loss of ecosystem services76. For example, increased use of agrochemicals have polluted waterways and reservoirs, drastically affecting the environmental services they provide (Gunatilleke et al. 2008; p.46). The lack of an ecosystem-based approach to land management, due to the Sri Lankan government’s largely traditional approach to conservation, has contributed to many of these issues. As pointed out previously, even though the Department of Wildlife acknowledges the need for such an approach, it does not practice it, as evident in its lack of concern regarding land-use in areas near PAs, especially in designated buffer zones.

Inadequate coverage of the PA system

Most of Sri Lanka’s PAs are small and isolated, reflecting the fragmented nature of much of its remaining natural habitat (Green and Gunawardena 1997). As shown in Figure 2.4, 30% of PAs are less than 100 ha and 54% are less than 1,000 ha.

76 As described in Chapter 1, ecosystem services are the myriad benefits that people derive from nature. These include provisioning services such as food, water, and fibre; regulating services such as climate regulation and pollination; aesthetic services such as recreation and spiritual well-being; and supporting services such as nutrient cycling (Uragoda 1994).

82 30 DWC

FD 25

10 Number of reserves Number

0 < 100 101-999 1,000 - 9,999 10,000 - 99,999 > 100,000

Protected area size classes

Figure 2.4: Distribution of protected areas by size (from Green and Gunawardena 1997)

The single largest PA is Wilpattu National Park (133,571 ha) in the northwest arid zone, although the 144,513 ha complex of contiguous PAs in the southeast dry zone— comprising Ruhunu National Park, Yala East National Park and Yala Strict Nature Reserve—is slightly larger. PAs tend to be smaller in the intermediate and wet zones, where much more of the original forest having been converted to other forms of land use (Green and Gunawardena 1997). Therefore, despite the significant land devoted to conservation, a considerable portion of Sri Lanka’s biodiversity lies outside PAs, especially in the wet zone. This is especially true for endemics. Furthermore, several large charismatic species, such as elephants and leopards, range extensively outside reserves. Overall, as acknowledged in Sri Lanka’s Biodiversity Conservation Action Plan, “the present PA network does not adequately represent some floristic regions of the country, including areas that occur within the biodiversity rich wet zone… and about 15% of indigenous terrestrial species diversity is not found within the established PAs” (Ministry of Forestry and Environment 1999; p.22).

83 Several elements of Sri Lanka’s traditional approach to conservation have contributed to the inadequate coverage of the country’s PA system. Firstly, the lack of cooperation between the Department of Wildlife that is ‘responsible’ for species such as elephants and leopards, and the Forest Department that manages lands encompassing ranges of these species, has hampered efforts to ensure that these animals have adequate habitat for survival. Furthermore, the general lack of an ecosystem approach has led to types of land-use near PAs that have exacerbated issues arising from their isolation and size; the prevalence of large extents of chemically intensive agriculture adjacent to reserves is a prime example.

Escalating human-wildlife conflict

Human-wildlife conflict in Sri Lanka is a direct result of some of the factors described above. Firstly, a rapidly expanding human population reduces available habitat for animals, especially outside the country’s PA network. This, combined with the inability of the current PA network to provide habitat for wide-ranging species, results in frequent contact between wildlife and humans with often disastrous consequences for both groups. A prime example is the growing human-elephant conflict. Since over two thirds of elephant habitat in Sri Lanka is outside PAs, elephants inevitably come into contact with rapidly expanding human settlements and irrigated agriculture (Groom et al. 2006). Around 70 people and over 200 elephants die each year in the conflict that arises when elephants, faced with dwindling habitats, raid crops (Yatawara 2011).

A similar conflict occurs when leopards attack livestock in villages adjacent to national parks when their natural prey becomes scarce. Enraged villagers resort to poisoning half-eaten leopard kills, knowing that the cats usually return to finish their prey. The following account by Santiapillai and Jayewardene (2004; p.1063), outlining the challenges to leopard conservation, highlights several points that encapsulate Sri Lanka’s major conservation challenges:

The current strategy for conserving the leopard in Sri Lanka is based on the premise that people and predators do not mix. Conservation measures to date have largely depended on legislative protection of the species and setting aside of national parks and nature reserves to ensure that people and predators are kept apart. But such an approach may not be adequate to ensure the long-term survival of the leopard. If the leopard is to survive in

84 the wild, then healthy populations of the species must be maintained both within and outside PAs. But it is in the unprotected areas outside national parks and reserves that the leopard is coming under increasing threat from intensification of agriculture and an upsurge in human numbers and their livestock. Habitat modifications outside PAs may lead to the constriction of the life-support systems of the leopard. Agricultural expansion is encroaching on wildlife dispersal areas and forest corridors that are crucial for the integrity of the PA network. This has brought wildlife and people into increased contact and conflict with one another over diminishing unfenced land. Human-wildlife conflict would undoubtedly worsen due to some of the conflict drivers described above. For example, the loss of ecosystem services would reduce agricultural productivity, in turn intensifying competition for land between humans and animals. Climate change impacts would have similar effects as current land becomes unsuitable for cultivation, and animals are compelled to move out of PAs due to changes in habitat within reserves. Therefore, human-wildlife conflict has become one of Sri Lanka’s key conservation-related issues. Of these, human-elephant conflict is the most significant, as described in detail in Chapters 3 and 4.

2.8 Overview of key points in the chapter

This chapter has provided the context for much of the discussion in the rest of the thesis. It first described Sri Lanka’s biophysical context, including its weather patterns, hydrology and habitat types and highlighted potential impacts of climate change. It then described the unique biodiversity resources found in this ‘biodiversity hotspot’.

The discussion on Sri Lanka’s socioeconomic context illustrated that although population growth has slowed, the country is still densely populated (although less so than some of its neighbours). The impressive social indicators of this multi-ethnic, multi-religious country were discussed, and its democratic system of government was described briefly. The next section described land-use in Sri Lanka, demonstrating that a third of the country’s land area is devoted to agriculture, while another third is dedicated to forests and wildlife. It also showed that the government, which owns almost 80% of available land, is the main entity involved in land-use decisions.

The discussion then turned to country’s historical context. Firstly, ancient kingdoms, the colonial period and the post independence period were introduced as the three main eras

85 of Sri Lankan history. Secondly, the country’s conservation history was traced within the context of these three periods: the close co-existence with, and reverence for, nature during the time of ancient kingdoms; the largely exploitive use of resources during the colonial period, towards the end of which the foundations were laid for the country’s PA system and conservation legislation; the establishment of a separate Department of Wildlife; and the expansion of the PA network during the post-independence period. Finally, the history of land use was examined, describing in turn: the largely holistic and integrated traditional land management systems during the era of ancient kingdoms, which were managed in a relatively decentralised manner at the village level; the conversion of vast swathes of land to plantations during the colonial period, heavily impacting traditional forms of land-use such as chena; and the post-independence land reforms that emphasised rural development and agriculture.

Next, the responsibilities and goals of key conservation-related government institutions such as the Central Environmental Authority, the Forest Department and the Department of Wildlife as well as international and local NGOs were described. The country’s conservation-related legislative and policy framework was examined, highlighting important legislation such as the Fauna and Flora Protection Ordinance.

The final section focused on Sri Lanka’s major conservation-related issues. The country’s use of a traditional approach to conservation was illustrated using examples of the reductionist approach to land management (stemming from a view that nature is static), and the perception that people are apart from nature. The key issues faced by the country, namely, the loss of habitat and consequent loss of biodiversity, the loss of ecosystem services, inadequate coverage of the PA system, the threat of climate change, and escalating human-wildlife conflict were examined in the context of the country’s traditional approach to conservation. Therefore, section 2.7 directly addresses my first research question, stated in Chapter 1: ‘What is (and has been) Sri Lanka’s approach to conservation, and what are key issues associated with this approach?’

Chapters 3 and 4 will explore in more detail perhaps the most significant conservation- related issue that was highlighted in this chapter, i.e., human-elephant conflict.

86 Chapter 3: Sri Lanka’s imperilled national icon

Sri Lanka’s growing human-elephant conflict, which was highlighted as one of the country’s major conservation-related problems in Chapter 2, will be explored in more detail in Chapters 3 and 4. Chapter 3 provides an introduction to the Asian elephant and human-elephant conflict both across Asia and within Sri Lanka. It will illustrate that, given the importance of the elephant from historical, cultural, and ecological perspectives, conserving this flagship species and addressing human-elephant conflict is a priority for governments across Asia. This issue especially holds true in Sri Lanka where the elephant is a national icon. Chapter 4 shows how human-elephant conflict is perhaps the most important conservation-related issue in Sri Lanka, describes how measures used to address the conflict have been insufficient, and proposes a landscape- based approach to mitigate the conflict.

3.1 The Asian elephant: an introduction

The Asian elephant (Elephas maximus) is smaller than its African cousin (Loxodonta africana), but still may weigh over 4 tons and reach more than 3.5 m in height77 (Santiapillai and Jackson 1990; p.43). It is distinguished by its much smaller ears; its rounded back compared to the African elephant’s ‘saddle back’; its single ‘finger’ at the tip of the trunk compared to the African elephant’s two ‘fingers’; its double-domed forehead unlike the African elephant which has a single dome; and the fact that only males have tusks, while in African elephants both sexes have tusks (Bandara and Tisdell 2002; p.197).

77 The two species are the sole survivors of the family which includes various types of prehistoric elephants such as the mammoth and the mastodon with whom they share a large pig-like ancestor called Moeritherium, which lived about 25 million years ago (Layke 2009; p.3).

87 Characteristics and behaviour

Asian elephants inhabit a variety of habitats such as tropical evergreen forest, semi- evergreen forest, moist deciduous forest, dry deciduous forest and dry thorn forest, as well as cultivated and secondary forests and scrublands (IUCN 2009). They range over a wide spectrum of elevation levels, and can be found at sea level, as in southeast Sri Lanka, and at over 3,000 m above sea level in places such as the Eastern Himalayas in northeast India (IUCN 2009).

Elephants have a complex social system based on breeding groups of females and young (Santiapillai and Jackson 1990; p.vi). Herds form part of larger, related groupings called ‘clans’, which congregate periodically. Mature males live singly or in small groups and have no permanent ties with the females, but they may associate with them while feeding or mating (Santiapillai and Jackson 1990; p.vi). Asian elephants can live up to 70 years. Males reach sexual maturity at between 10 – 15 years, while females become sexually mature at 10 – 14 years, usually producing calves every 4 – 5 years (Sukumar 2006; p.3; IUCN 2009).

Asian elephants are mega-herbivores that may spend more than two-thirds of the day feeding (Sukumar 2006; p.1). On average, they consume about 150 kg of vegetation and drink about 140 litres of water a day (Bandara and Tisdell 2002; p.197). The proportions of the different plant types in their diet vary depending upon the habitat and season. For example, in southern India, about 70% of elephants’ dry-season diets were material obtained by browsing, while grasses made up about 55% of their wet season diets (IUCN 2009). Elephants’ ranges are dependent on availability of food and water— when resources are plentiful throughout the year, they tend to range across relatively small distances, but, if not, they may migrate hundreds of kilometres to find nourishment (Santiapillai and Jackson 1990). For example, home-ranges in excess of 600 km² have been recorded for females in south India, while much smaller home ranges of 30 – 230 km² have been recorded in Sri Lanka (Fernando et al. 2008; p.7). Moreover, a herd’s dry season range may be almost double that of their wet season range (Santiapillai and Jackson 1990; p.vi).

88 Numbers and distribution

Asian elephants are only found in 13 countries across south and southeast Asia: India, Sri Lanka, Nepal, Bhutan, Bangladesh, Myanmar, China, Thailand, Laos, Cambodia, Vietnam, Malaysia and Indonesia (Bandara 2004; p.285). The total population of wild elephants is between 40,000 and 52,000, while their total range is approximately 486,800 km² (IUCN 2009). Table 3.1 presents their population breakdown by country.

Table 3.1: Population estimates of Asian elephants by country (IUCN 2009)

Country Estimated number of elephants Bangladesh 150 – 250 Bhutan 250 – 500 Cambodia 250 – 600 China 200 – 250 India 26,390 – 30,770 Indonesia 2,400 – 3,400 Lao PDR 500 – 1,000 Malaysia 2,100 – 3,100 Myanmar 4,000 – 5,000 Nepal 100 – 125 Sri Lanka78 2,500 – 4,000 Thailand 2,500 – 3,200 Vietnam 70 – 150 Total 41,410 – 52,345

78 This is an underestimate according to the nationwide elephant census conducted in late 2011, which estimated that there are almost 6,000 elephants in Sri Lanka. More details are presented in section 3.4 and Appendix 3.

89 The ecological importance of Asian elephants

Asian elephant conservation is important from an ecological perspective. Given their requirements for large areas, elephants are regarded as an ‘umbrella species’ because conserving them also protects many other species occupying the same area (Riddle et al. 2010; p.653). Elephants are also regarded as a ‘keystone species’ because of their important ecological role; for example, the considerable amounts of dung they produce facilitate nutrient recycling and seed dispersal within the ecosystem (Sukumar 2006; p.4).

Conservation status and threats

The Asian elephant is one of the world’s most endangered large mammal species (Riddle et al. 2010). While over 600,000 African elephants range across 4.5 million km2 in 33 countries in Africa, even most optimistic figures indicate that there are less than 60,000 free ranging Asian elephants inhabiting less than 500,000 km2 in only 13 countries (Santiapillai 1997). Consequently, the Asian elephant is listed as ‘endangered’ under the IUCN Red List of Threatened Animals and appears in Appendix 1 of the Convention on International Trade of Endangered Species of Wild Fauna and Flora, or CITES79 (Sukumar 2006; p.2).

Prior to the large-scale modification of its habitat by humans, the range of the Asian elephant extended from the Euphrates-Tigris river systems in the west through southern Asia80 to Indochina and most of southern China in the east; an area covering over 9 million km² (Santiapillai and Jackson 1990; p.vii; IUCN 2009). Asian elephants are now extinct in West Asia, Java, and most of China81—the western-most populations of Asian elephants are found in northern and southern India, and only small populations persist in southern China and parts of Vietnam (Riddle et al. 2010; p.655-656). Even within their existing ranges, elephants have been in retreat for hundreds of years, and generally

79 Appendix I lists species that are the most endangered among CITES-listed animals and plants. They are threatened with extinction and CITES prohibits international trade in specimens of these species except when the purpose of the import is not commercial. 80 I.e. south of the Himalayas. 81 The western populations were probably extinct by 100 B.C., and the main Chinese populations disappeared sometime after the 14th century B.C. (Santiapillai and Jackson 1990; p.v).

90 survive only in highly fragmented populations (IUCN 2009). Their present and former distributions are indicated in Figure 3.1.

Figure 3.1: The former and present distributions of Asian elephants (from Santiapillai and Jackson 1990; p.vii)

Habitat loss has been largely responsible for declines in Asian elephant ranges (Riddle et al. 2010; p.656). The region’s rapid population growth and development activities in the past few decades have accelerated habitat loss, and most elephant range states have seen declines of 25% to 60% in forest cover during this period (Riddle et al. 2010; p.656). Growing rural populations, conversion of land to agricultural uses and deforestation have blocked traditional migration routes and pushed elephants into “ever- shrinking habitat islands” (Leimgruber et al. 2003; p.347). Unfortunately, habitat loss is expected to continue as 20% of the world’s population lives in and around areas inhabited by the Asian elephant, and at current growth rates, this population would double in 3 – 4 decades (Bandara 2004; p.286).

Issues associated with habitat loss have been exacerbated by poorly planned development activities that fail to integrate conservation requirements. The resulting

91 habitat fragmentation has created smaller, often poor-quality habitat patches, and genetically isolated elephant populations, reducing their prospects for survival in the long-term (Riddle et al. 2010; p.656). A detailed analysis by Leimgruber et al. (2003) on the fragmentation of Asia’s remaining wild-lands82 demonstrates that elephant habitat is shrinking and much of it lies outside PAs (Table 3.2 and Figure 3.2).

Table 3.2: Area of remaining wildlands inside Asian elephant ranges, km2 (from Leimgruber et al. 2003; p.350)

Country Geographic range of the Remaining Protected wildlands (% of Asian elephant wildlands remaining wildlands) Bangladesh 11,425 2,886 452 (16%) Bhutan 3,323 1,232 350 (28%) Cambodia 55,382 20,088 6,927 (34%) China 41,390 31,910 2,216 (7%) India 151,309 46,880 5,547 (12%) Indonesia 30,750 19,709 3,396 (17%) Laos 53,611 42,747 8,318 (19%) Malaysia 57,237 29,106 4,512 (16%) Myanmar 288,990 168,461 11,215 (7%) Nepal 2,195 88 0 (0%) Sri Lanka 27,816 4,102 1,664 (41%) Thailand 95,914 47,864 21,694 (45%) Vietnam 53,765 31,909 5,485 (17%)

82 Wildlands are defined as “large areas of natural vegetation unaffected by habitat fragmentation over time periods long enough that natural ecological processes (e.g. natural regeneration) determine ecosystem structure, composition and functions” (IUCN 2009).

Figure 3.2: Area of remaining and protected wildlands inside the range of the Asian elephant (from Leimgruber et al. 2003; p.350)

The authors summarise the disturbing results of their research (Leimgruber et al. 2003; p.355):

Our results indicate a dire situation for Asian elephants. We confirmed that only 51% of the geographic range consisted of unfragmented wildlands by the 1990s. Only 8% of the entire range was protected; little of that was unfragmented wildland. While 16% of the unfragmented wildland in the Asian elephant’s range is protected according to IUCN, many PAs may be paper parks, affording little protection for elephants. Our assessment was based on data from the 1980s and 1990s. Since then the situation has become worse. Annual deforestation rates in Southeast Asia are increasing and are close to 1.9% as compared to global rates of 0.2 – 0.3%. The disappearing forests may be prime elephant habitats. The resulting fragmentation may further threaten elephants by pushing them into shrinking habitat islands and increasing people–elephant conflicts. Apart from habitat loss and the inevitable conflict with humans, poaching is the other major threat to Asian elephants. For example, poaching for ivory has resulted in skewed sex ratios in elephant populations in southern India, while during the Vietnam war, elephants and other wildlife were actively hunted for commercial purposes in Cambodia, Laos, and Vietnam, resulting in drastic declines in local populations (Riddle

93 et al. 2010; p.657). Poaching continues to be a problem in northeast India, China and Myanmar (Riddle et al. 2010).

3.2 Human-elephant conflict

Human-elephant conflict is a direct consequence of the fragmentation and destruction of elephant habitat. Faced with dwindling sources of food and water within their ranges, elephants raid crops, often causing collateral damage to famers’ property, and sometimes even killing people trying to protect their crops. The often drastic measures taken by local communities to drive away marauding elephants result in escalating human-elephant conflicts.

A recent conservation review pointed out that “human-elephant conflict poses a serious threat to elephant conservation in Asia”, and that it “not only increases the local community’s hostility towards elephants, but also towards conservation in general” (Riddle et al. 2010; p.656). All elephant range states have reported instances of human- elephant conflict, while 10 of the 13 states reported related retaliatory killings of elephants83 (Riddle et al. 2010; p.656). Overall, growing human-elephant conflict represents one of the major challenges to Asian elephant conservation, as described by Fernando et al. (2005; p.2466):

The human-elephant conflict represents a widespread, complex, and intractable challenge to conservation and is the major threat to elephants across their range. Unprecedented human population growth in Asia has caused increasing conversion of natural habitat to human dominated landscapes, bringing elephants and humans into greater contact and conflict. Confronted with the escalating human-elephant conflict, the historical respect and reverence for elephants in Asian cultures and societies is rapidly eroding. People who have to contend with elephant depredation on a daily basis increasingly perceive them as agricultural pests, an unwelcome burden, and a threat to their survival and well-being. Resolving human-elephant conflict is a priority for governments across Asia. In Thailand, the King has initiated a programme to reduce conflict, which converts land previously under cultivation into elephant habitat and improves the incomes of local people (Perera 2009; p.46). In 2010, the Indian government established an Elephant

83 Some recent data related to the conflict is presented in Appendix 3.

94 Task Force to devise a strategy to conserve India’s Asian elephants and mitigate escalating human-elephant conflict (Rangarajan et al. 2010; p.72).

3.3 The Asian elephant in history and culture

The elephant—the largest terrestrial land creature—has captured the human imagination for millennia. In Hindu cosmology it is associated with the very creation of the universe:

When the cosmic egg was first cracked open by the Creator, its yolk became the oceans, the land, and the sky... From out of the eggshell in the Creator’s right hand emerged eight male tuskers, including the magnificent Airavata who would carry Indra, Lord of the Universe, on his back. From the eggshell in the Creator’s left hand emerged eight female elephants. This original herd of sixteen were the progenitors of the species and their offspring ‘ranged at will over the forests, rivers, and mountains of the whole world’ (Alter 2004; p.34). One of the most popular Indian Gods is the elephant-headed Ganesh, the God of Wisdom and ‘Remover of Obstacles’. He is evoked by Hindus at the beginning of any important undertaking (Santiapillai and Jackson 1990; p.vi). Hindu voyagers and settlers spread Ganesh worship throughout southeast Asia, and Ganesh temples are found in Thailand, Indonesia and Indochina (Santiapillai and Jackson 1990; p.vi).

Elephants have also been significant in Buddhist practices and rituals. For example, a white elephant is associated with the Buddha’s conception—his mother, Queen Maya, dreamt that such a creature, holding a white lotus, entered her right side (Cannon and Davis 1995; p.18). Wise men subsequently told her that this was a sign that she would soon give birth to a great man. The Buddha also used the elephant as a metaphor within his teachings, as in the following example on ‘overcoming moral difficulties’: “Just as the elephant, sunk in the mire, pulled himself from the bog and stood on firm ground... so you raise yourselves from the mud of moral defilements and put yourselves on the firm land of Nibbana” (Cannon and Davis 1995; p.22).

The ancient Greeks made many references to the elephant, although some contained fanciful claims, as in the following description in Pliny the Elder’s Natuaralis Historia, written in 77 A.D.:

95 The largest land animal is the elephant, and it is the nearest to man in intelligence: it understands the language of its country and obeys orders, remembers duties that it has been taught, is pleased by affection and marks of honour, nay more it possesses virtues rare even in man, honesty, wisdom, justice, also respect for the stars and reverence for the sun and moon (Alter 2004; p.52). Humans have used Asian elephants for various purposes for centuries. Some of the earliest historical evidence regarding their domestication comes from India, where they have been depicted on seals dating from the second millennium B.C. (Carrington 1958; p.164). By the time of the Greek conquest of India elephants were used extensively as beasts of burden and in war—a coin commemorating the Battle of Hydaspes in 326 B.C. shows Alexander the Great on horseback attempting to spear King Porus of Punjab, who is riding an elephant (Carrington 1958; p.164). Perhaps the most famous example of the use of elephants in western history is the crossing of the Alps by Hannibal with an army of 15,000 soldiers and 37 elephants, whose goal was to destroy the Roman armies in France and northern Italy before marching on Rome (Carrington 1958; p.187). Interestingly, although most of his elephants were African, his favourite appears to have been Asian, probably one which had been captured in Syria (Santiapillai and Jackson 1990; p.vi).

Elephants were also used in armies throughout Asia. For example, Thai historical records describe how the Kings of Sukhothai and Chod battled on elephant-back in the 14th century A.D. (Pimmanrojnagool and Wanghongsa 2001; p.35). The Moghul Emperors Jehangir and Akbar maintained huge corps of war elephants; the former was said to have owned 12,000 elephants, while there were 40,000 captive elephants in his empire (Carrington 1958; p.182; Riddle et al. 2010; p.656). Historical depictions of war elephants in Myanmar date back to the 1044 A.D. to the time of King Anawrahta of Bagan (Aung and Nyunt 2001; p.90).

Asian elephants have also been used extensively in the logging industry. For example, at the beginning of the 20th century there were around 100,000 domesticated elephants in Thailand, most of which were employed in the logging industry (Tipprasert 2001; p.157). These practices continue in modern times in some countries, albeit on a vastly reduced scale. For example, there were an estimated 4,075 elephants working in Myanmar’s timber industry in 2000 (Aung and Nyunt 2001; p.89).

96 The British used elephants extensively in their colonies, even in the early 20th century. Sir Sydney Cotton, the general who had subdued the mutinous sepoys at Peshawar noted the importance of elephants in warfare (Nongbri 2003; p.3190):

Strange as it may appear, there is no beast of burden in the world that can beat an elephant in traversing precipitous mountains, and the author never went without these mortars (2.5 kg/5.5 pound mortars conveyed on elephants). Of all animals of the creation there are none so intelligent and so useful in military operations in mountains as the elephants, and sufficient value and importance is not attached to them in the British (Indian) Services. One elephant can transport over long distances 6 British soldiers, with arms, ammunition and bedding, beside rations.

The Asian elephant as a ‘flagship species’

The preceding discussion illustrates that the Asian elephant is important not only from an ecological viewpoint, but also because it has played a significant role in the region’s history and culture. Consequently, it is often used as a ‘flagship species’ in conservation efforts (Venkataraman et al. 2002; p.1022). This is exemplified by the Indian government’s recent declaration84 of the elephant as a ‘National Heritage Animal’, with the view that “declaring it a National Heritage Animal will give it due place as emblem of ecological sensitivity. It will also mark recognition of its centrality in our plural cultures, traditions and oral lore” (Anonymous 2010; p.1).

The species likewise captures the imagination of people outside Asia. The USA’s Asian Elephant Conservation Act, for example, signed into law in 1997 by President Clinton after being passed by both the Senate and the House of Representatives, recognised that: “to reduce, remove, or otherwise effectively address threats to the long-term viability of populations of Asian elephants in the wild will require the joint commitment and effort of nations within the range of Asian elephants, the United States and other countries, and the private sector”, and established a grants programme85, to fulfil the purpose described by the Act (United States Government 1997).

The special role Asian elephants have played in Sri Lanka’s history and culture will be outlined in section 3.5. First, however, I briefly introduce the Sri Lankan subspecies.

84 In October 2010. 85 The Asian Elephant Conservation Fund.

97 3.4 Sri Lanka’s Asian elephant subspecies

There are three subspecies of Asian elephant. The Sri Lankan subspecies (Elephas maximus maximus) has the darkest skin, the Asian mainland subspecies (Elephas maximus indicus) has medium darkness of skin, and the Sumatran subspecies (Elephas maximus sumatranus) has the lightest skin colour and least depigmentation86 (Bandara and Tisdell 2005; p.4).

Characteristics of the Sri Lanka’s elephant subspecies

The Sri Lankan elephant has several unique characteristics in its genetic makeup, appearance and behaviour:

• Size: The Sri Lankan subspecies is the largest among the three subspecies (WWF International 2006; p.2).

• Tusks: Only about 7% of male elephants in Sri Lanka have tusks, compared to 45% of those in north India and almost 95% of those in south India (Fernando 1999; p.39).

• Genetic diversity: Although Sri Lankan elephants share several mitochondrial haplotypes with those in the mainland, Sri Lanka has the highest genetic diversity within its population of any of the 13 elephant range states (Santiapillai et al. 2006; p.92).

• Ranging behaviour87: Sri Lankan elephants have much smaller home ranges than their Indian counterparts. According to studies done by Fernando et al. (2008; p.7) using radio telemetry, their observed home ranges were between 30 and 230 km2, while similar studies conducted in India revealed those that were 3-5 times larger88. Male elephants in Sri Lanka dramatically increased their range during the musth period of about two months, covering an area almost four times their range for the rest of the year (Fernando 1999; p.40). Sri Lankan

86 I.e. patches of skin. 87 The implications of some of this behaviour will be explored in more detail in section 4.5. 88 The significant difference in home range size between the two populations is probably due to higher habitat productivity in Sri Lanka (Leimgruber et al. 2003; p.347).

98 elephants also displayed high fidelity to their home ranges, a pattern of behaviour that has also been observed among Indian elephants (Fernando et al. 2008; p.9). Finally, in contrast to previously held views, it has been found that Sri Lankan elephants do not undertake seasonal long-distance migrations (Fernando 1999; p.40)89.

• Density: Sri Lanka has one of the highest densities of elephants within the range of the species (Perera 2007; p.15; Fernando et al. 2008; p.9).

Overall, given its unique behaviour, high genetic diversity and high population density, the Sri Lankan elephant is extremely important in terms of overall Asian elephant conservation (Fernando et al. 2005; p.2466; Santiapillai et al. 2006; p.92).

Numbers and distribution

An island-wide elephant census conducted in August 2011 estimated that there are almost 6,000 elephants in Sri Lanka (Hettiarachchi 2011). As indicated in Figure 3.3, they are distributed throughout the lowlands in the country’s dry zone, apart from a small remnant population in the wet zone’s Peak Wilderness area (Santiapillai et al. 2006; p.93). The northwest, northeast and southeast regions contain the majority of elephants, accounting for 20%, 30% and 45% of the total population respectively (Hettiarachchi 2011)90.

89 Further details of findings from the latest elephant research in Sri Lanka are presented in section 4.5. 90 Details of the census are presented in Appendix 3.

Figure 3.3: The approximate range of Asian elephants in Sri Lanka91 (from Fernando et al. 2011; p.95)

DNA testing has revealed significant genetic differences among northern, mid-latitude and southern elephant populations in Sri Lanka (Fernando et al. 2000; p.369). This confirms some of the morphological differences that have been previously observed between elephants from different geographical locations in the country. For example the smaller stature southern elephants have been traditionally called ruhunu gatau while the

91 Areas of distribution are demarcated by a heavy red line. Elephants are absent in polygons demarcated by a thin red line. Green polygons denote PAs managed by the Department of Wildlife Conservation. Numbers and letters indicate areas referred to in the text: N=North, NW=Northwest, NC=North-central, S=South, 1=Udawalawe NP, 2=Lunugamvehera NP, 3=Yala NP, 4=Yala NP Block I, 5=Bundala NP, 6=Mattala, 7=Sinharaja, 8=Adam’s Peak, 9=Minneriya-Kaudulla NPs and Hurulu Reserve, 10=Kala Wewa, 11=Wilpattu NP.

100 exceptionally large and bulky elephants from the mid-latitude region have been called vil aliya (Fernando 1999; p.41). Therefore, Fernando (1999; p.41) suggest that the northern, mid-latitude and southern populations should be managed as distinct populations and that elephants should no longer be translocated between these regions.

3.5 The elephant within Sri Lanka’s history and culture

The elephant has enormous cultural and religious significance to both Buddhists and Hindus, Sri Lanka’s two main religious groups 92 (Santiapillai et al. 2006; p.92). Furthermore, elephants were a valuable commodity both during ancient times and the colonial period. We now examine the role of the elephant during the three main periods in Sri Lankan history: the era of the ancient kingdoms, the colonial era and the post- independence era.

The era of the ancient kingdoms

The Mahavamsa, which chronicles Sri Lankan history93, makes its first reference to domesticated elephants during the time of King Devanampiya Tissa who ruled around 300 B.C. (Bandara and Tisdell 2005; p.7). Contemporary Sinhala literature describes the state elephant on which the king rode—always a tusker, it was called Mangalahasti, and was housed in a special stable called hatthisala (Bandara and Tisdell 2005; p.7-8). A 12th century inscription on a stone seat at Polonnaruwa records that King Nissanka Malla sat upon it while watching elephant fights that were staged for the entertainment of nobles (Bandara and Tisdell 2005; p.8).

Perhaps the most famous elephant in Sri Lankan history is Kandula, the war elephant used by King Dutugemunu is his campaigns to regain control of the country from invaders from South India. The Mahawamsa records that Kandula was instrumental in attacking the fortresses built by the Tamil King Elara, and participated in 28 battles (Cannon and Davis 1995; p.24-25). Cannon and Davis (1995; p.25) describe how the outcome of the war was decided by a contest between the two kings:

92 This is because, as was pointed out in section 3.2, the animal is significant to both religions. 93 The significance of the Mahawamsa is described in section 2.5.

101 The two kings confronted each other on their respective elephants; King Elara on Mahapabbata, King Dutugemunu on Kandula. Fiercely they fought each other. As Kandula impaled Mahapabbata through the side with his tusk, King Dutgemunu cast his spear into King Elara. Both king and elephant fell to the ground and died. King Dutugemunu and his elephant Kandula paraded triumphantly, Dutugemunu claiming himself as the king of all of Lanka. Ancient kings captured and tamed elephants, which had a variety of uses. The first was in war—King Rajasinghe I used 2,200 war elephants against the Portuguese in the great siege of Colombo in 1588 (Saparamadu 2006; p.116). The second was for ceremonial occasions such as the annual Perahera, which still takes place in the city of Kandy. This ceremony, which dates back nearly 220 years, brings together over a hundred elephants that parade the streets at night on certain predetermined days in July and August each year (Bandara and Tisdell 2005; p.8). A specially chosen elephant gets the honour of carrying a casket containing sacred relics of the Buddha. Thirdly, kings used elephants as executioners, as described by the Englishman Robert Knox in 1848 (Saparamadu 2006; p.114). Finally they were used for entertainment—elephant fights were a popular form of sport which was called Gaja Keliya (Bandara and Tisdell 2005; p.8).

Given their importance within the cultural context of his kingdom, the King of Kandy maintained a special unit that dealt with all matters concerning elephants—including their capture, training, conservation and export—that was under a high-ranking officer known as the Gajanayake Nilame (Bandara and Tisdell 2005; p.8). All elephants were considered the property of the crown and the slaughter of elephants was “reckoned amongst the most heinous of offenses” (Saparamadu 2006; p.115).

Sri Lankan elephants were considered a valuable commodity in the ancient world. The export of elephants to India continued without interruption from the period of the First Punic War, as elephants from Sri Lanka were found to be more easily adapted for war and considered better than those from the mainland (Bandara and Tisdell 2005; p.8-9). Elephants were sent to Myanmar as tribute from Sri Lankan kings and exported to Egypt, probably for both war and ceremonial occasions (Bandara and Tisdell 2005; p.9). Sri Lankan elephants were well known to the Greeks even as far back as the 3rd century B.C., during the time of Alexander the Great, when an admiral of his fleet stated that the

102 elephants from Taprobane94 “are bigger, more fierce and furious for war service than those of India” (Bandara and Tisdell 2005; p.9).

The colonial era

The Portuguese continued the lucrative trade in Sri Lankan elephants after conquering the country’s maritime provinces in 1505 (Jayawardene 2001; p.43). It is recorded that they exported 37 elephants annually from two kraals95, and established a special unit called the ‘Elephant Hunt’— which employed over 3,000 people by 1655—for elephant capture and export (Jayawardene 2001; p.43; Bandara and Tisdell 2005; p.10).

The Dutch, who captured Sri Lanka’s maritime provinces from the Portuguese in 1658, expanded this trade in elephants. All elephants in Dutch-held territories were considered the property of the Dutch East Indies Company96 and their capture and training was organised through kraals held annually in the Matara and Wanni areas (Saparamadu 2006). The Dutch set targets for the collection of elephants, for example, requiring chiefs of the Vanni District to provide 80 elephants; similar tributes were extracted from the headmen in Batticaloa and Trincomalee (Saparamadu 2006). All elephants received by the Company were sold at very high prices in India resulting in an average annual profit of 200,000 guilders; the sale of elephants was the second largest source of revenue for the East Indies Company97 (Saparamadu 2006; p.113).

British colonial rule, which followed that of the Dutch, was probably the worst period for Sri Lankan elephants. Although, initially, the British continued the practice of capturing elephants, they soon found elephants and other wild animals to be a nuisance, especially after coffee and tea plantations were established in the central hills. Rewards were given for destruction of elephants, leopards and bears from 1831, and anyone who shot an elephant had only to produce the tail and trunk tip to claim the reward of a few shillings; the tusks, if any, were a bonus (Cannon and Davis 1995; Saparamadu 2006; p.119). The plight of Sri Lankan elephants worsened after British colonials began

94 Later Ceylon and then Sri Lanka. 95 A kraal was a timber enclosure used to capture, and subsequently tame and train wild elephants. 96 Known as the Vereenigde Oost-Indische Compagnie (VOC) in Dutch, this was a chartered company established in 1602. The States-General of the Netherlands granted it a 21-year monopoly to carry out colonial activities in Asia (Fernando et al. 2008; p.9). 97 Cinnamon was the first.

103 shooting elephants for sport. Nongbri (2003; p.3190) describes how the elephant became prime target of hunters:

Elephant hunting was considered a sport that was part of the sporting agenda of every ambitious big game hunter. Among British officials, several reckoned elephant shooting over and above that of the tiger as the most supreme and dangerous of all sports. Sir Samuel Baker, the greatest authority on elephants in the 19th century was one of the earliest to admit that “the wild elephant's attack (was) one of the noblest sights of the chase”. The scale of the destruction was staggering. British government records indicate that 5,194 elephants were destroyed between 1845 – 1859 and a further 3,253 were exported from 1853 – 1894 (Lorimer and Whatmore 2009; p.679). In addition to live animals, ivory was exported to India: between 1879 and 1883 about 1000 kg of raw ivory was exported annually from Sri Lanka (Santiapillai et al. 1999; p.177).

Elephant hunting became so intense that the government reward of ‘ten shillings per tail’ was reduced and even abolished in some districts (Cannon and Davis 1995; p.96). Major William Rogers, an Assistant Government Agent and District judge, killed over 1,400 elephants during the 12 years he spent in Sri Lanka98 (Uragoda 1994; p.100). Captain Galloway and Thomas Skinner both recorded over 700 kills each (Saparamadu 2006; p.119). These activities were extremely lucrative—in fact Rodgers is said to have purchased his commission in the Ceylon Royal Rifle Regiment from the sale of the tusks of the elephants he shot (Saparamadu 2006; p.119).

Despite the vast rewards associated with it, elephant hunting was a brutal activity as described by Lorimer and Whatmore (2009; p.678): As Harry Storey notes, writing at the start of the twentieth century: It would be a most unusual thing, if not an impossible occurrence, to find an elephant in Ceylon which has not been fired at, and more than once. In all the elephants I have killed any number of bullets, mostly round ball, often up to a dozen, have been picked up from head and body, after the flesh has disappeared. The jungles of Ceylon would have been full of maimed, blind and suffering animals.

98 Major Rogers was killed by lightning when he stepped outdoors during a thunderstorm; many Sri Lankans believed that this was divine retribution for the elephants he slaughtered (Saparamadu 2006).

104 The fact that elephant hunting was a “bloody, cruel and exploitative practice” was eventually recognised by the colonial government (Lorimer and Whatmore 2009; p.670). This was partly due to the growing concern in British society about animal welfare, “culminating in Queen Victoria sanctioning the addition of the prefix ‘Royal’ to the Society for the Prevention of Cruelty to Animals in 1840” (Lorimer and Whatmore 2009; p.680). This sentiment, as well as the realisation that game was declining rapidly due to indiscriminate slaughter, subsequently led to the promulgation of laws for the protection of wild animals, such as the ‘Ordinance to Prevent Wanton Destruction of Elephants, Buffaloes and Other Game’ of 1891, which required a license to capture or shoot an elephant (Uragoda 1994; p.4). Elephants were finally given full legal protection in the Fauna and Flora Protection Ordinance of 1937, but, by then, they had been effectively eliminated (in the wild) from Sri Lanka’s wet zone, despite having been distributed across much of the island even at the turn of the 20th century (Santiapillai et al. 2006; p.60, 93).

Bandara and Tisdell (2002; p.199) describe three phases of the decrease of Sri Lanka’s elephant population. The first is the use of the animal as an export commodity under the Portuguese, Dutch and eventually the British. As described above, elephants were regarded as a harvestable resource to be exploited and thousands of wild elephants were captured for this purpose. The second phase was the expansion of the plantation industry after the 1830s, when elephants were considered pests and shot throughout their range, leading to their elimination from the wet zone. The third phase covers the post-independence period, which will be described next.

105 The post-independence era

As noted in Chapter 2, one of the major preoccupations of post-independence governments was the development of rural areas and agriculture. Consequently, the largely undeveloped dry zone was targeted for development, especially through large- scale irrigation schemes. Unfortunately, this region, which had relatively high elephant numbers, but low human densities, contained most of the Sri Lanka’s elephants, leading to severe negative consequences for the species (Fernando et al. 2005; p.2467).

The scheme that had the most significant impact on the dry-zone landscape was the Accelerated Mahaweli Development Scheme initiated in 1978. Several dams were constructed across the Maheweli—Sri Lanka’s longest river—and the water was diverted to irrigate lands in the dry zone. Large areas of forests were cleared and brought under irrigated agriculture, cultivated by settlers who migrated from other parts of the country under a trans-migration and settlement program (Fernando et al. 2005; p.2467). Consequently, a significant portion of former elephant habitat was converted to agricultural land in under a decade, posing a significant threat to local elephant populations, as described by Santiapillai et al. (2006; p.95):

In the wake of the Accelerated Mahaweli Development Programme (AMDP), thousands of hectares of lowland forest in the dry zone were lost. Of the total area of 364,000 ha that were earmarked for irrigation, 260,000 ha were new land, which overlapped known elephant ranges. The elimination of natural habitat on such a large scale was one of the most serious problems facing elephant conservation in recent times. The elimination and fragmentation of vast areas of natural habitat in the lower reaches of the Mahaweli Ganga under the AMDP was the single most serious setback to Sri Lanka's elephants in recent times. In order to prevent conflict and to ‘save’ elephants that ranged in areas developed for agriculture, many were translocated into newly created national parks such as Maduru Oya and Flood Plains. However, translocation has failed to eliminate elephants from developed areas, and their continued presence has resulted in high levels of human- elephant conflict. This issue, perhaps the most significant in Sri Lankan conservation, is considered in more detail in section 3.7. Before that, however, I describe the current status of elephants in Sri Lanka.

106 3.6 The current status of the elephant in Sri Lanka

The elephant continues to be an integral part of Sri Lanka’s society and culture. Given this, as well as their historical importance, elephants have been afforded extensive legal protection within Sri Lanka, and several actions have been taken to look after their welfare. These will be explored in the rest of this section.

The elephant in Sri Lanka’s contemporary culture

The importance of elephants in Sri Lankan culture is summarised in the 2006 National Elephant Conservation Policy which states that “the elephant has been so closely associated with Sri Lanka’s history, culture, religions, mythology and even politics that it would be difficult to imagine the island without it” (Government of Sri Lanka 2006). As mentioned in section 3.5, religious events such as the Perahara, where the elephant plays an important role, continue to be held to this day. There are also many examples of the use of elephants in contemporary Sri Lankan society, ranging from their use as political symbols to corporate logos, or as the focus of corporate social responsibility activities of companies (Table 3.5).

107 Table 3.5: Examples of the use of the elephant in contemporary Sri Lankan society

Type of use Example Details

Political symbol Symbol of United National The elephant is the symbol of the Sri Party Lanka’s main opposition party. This is one of the most important political parties in the country, as it was founded in 1946, by D.S. Senanayake—Sri Lanka’s first Prime Minister—who was in the forefront in the

struggle for independence.

Corporate logo Logo of Elephant House Elephant House, which sells beverages and ice cream, is one of Sri Lanka’s most recognised brands. The company was established in 1866, during British colonial times.

Focus of Focus of CSR campaign of Elephant conservation is a key focus of the corporate social Dilmah tea corporate social responsibility activities of responsibility Dilmah, one of the Sri Lanka’s best-known activities tea brands. This globally recognised brand is sold in over 90 countries.

Legal status and government policy

The elephant is provided with a high degree of protection within the Sri Lankan legal system. The 1937 Fauna and Flora Protection Ordinance99—the country’s main piece of legislation which governs the conservation and use of its biodiversity—devotes an entire section to elephants, covering aspects such as (Pethiyagoda 2000):

• Making it illegal to “hunt, shoot, kill, injure or take any elephant” either inside or outside a national reserve or sanctuary. The exception to this is if elephants cause or are likely to cause “damage to any person, house, crop or plantation”, in which case the Director of Wildlife will issue a license to destroy the elephant.

99 Along with its various amendments.

108 However, persons who exercise the right under the license are required to report the killing of elephants to the police.

• Requiring a special permit to export an elephant, wild or tame, from Sri Lanka.

• Requiring domesticated elephants to be registered.

• Specifying fines for illegal activities such as trading in live elephants or elephant parts without first obtaining the required permits.

In 2006, the Sri Lankan cabinet approved a National Policy for the Conservation and Management of Wild Elephants in Sri Lanka. This is the first, and to date only, policy which is targeted to a specific species, demonstrating the importance of the elephant within Sri Lanka’s conservation landscape. The policy revolves around six policy objectives (Government of Sri Lanka 2006):

1. To ensure the long-term survival of the elephant in the wild in Sri Lanka,

2. To mitigate human-elephant conflict,

3. To derive socio-economic benefits from conserving the elephant,

4. To defray the costs imposed by human-elephant conflict on communities affected by it,

5. To adopt regulatory mechanisms for the removal of elephants from the wild for management reasons, and,

6. To promote scientific research as the basis for elephant conservation.

As many of the provisions are focused on the human-elephant conflict, the policy will be described in more detail in Chapter 4.

109 The Elephant Orphanage and the Elephant Transit Home

The government of Sri Lanka established an Elephant Orphanage in 1975 to care for wild elephants, especially calves, that were injured, orphaned, abandoned or separated from their herds, mainly due to human-elephant conflict (Tilakaratne and Santiapillai 2002; p.42-43). Situated on an 11 ha extent of coconut land bordering a river at Pinnawela, 88 km away from Colombo, the complex contained 82 elephants of all ages100, as of 2007 (Rajapaksa 2007; p.23). The animals are not fully tame, although some males are trained for basic types of work such as carrying bundles of fodder. The orphanage has been popular with both local and foreign tourists who visit in large numbers—about 50,000 a month—to watch the elephants being fed and bathed (Tilakaratne and Santiapillai 2002; p.43; Rajapaksa 2007; p.24). In 1982, the orphanage established an active breeding programme which resulted in the birth of 46 calves during the period from 1984 – 2007 (Rajapaksa 2007; p.25).

Given the criticism that the elephants at Pinnawela would remain in captivity for the rest of their lives, an Elephant Transit Home was established in 1995 at the Uda Walawe National Park, to re-integrate orphaned and abandoned juvenile elephants back into the wild (Miththapala 2009). Miththapala (2009; p.24) provides further details:

The Uda Walawe Elephant Transit Home was set up, where unlike at the Pinnawela Elephant Orphanage, the orphaned elephants were to have limited interaction with humans, so as to enable them to be reinstated back into the wild, at some point of time. The young are bottle fed, and then allowed to graze in a free ranging fenced area, abutting the western corner of the reservoir. In this manner, the elephants are allowed to feed and interact with each other without much human interaction for most of the day, except during feeding times and at night, when they are brought back into the central paddock. As of 2009, 66 elephants have been released into the Uda Walawe National Park. (Miththapala 2009; p.24).

100 This included 16 adult males and 30 adult females.

110 Domesticated elephants

As described in section 3.5, elephants have been used in Sri Lanka for various purposes such as working in the timber trade and participating in religious festivals such as the annual Perahera. However, with the advent of heavy machinery and the banning of large-scale timber extraction from natural forests, working elephants are now used mainly for religious and cultural purposes and in the tourist trade (Perera 2007; p.15). The number of domesticated elephants in Sri Lanka has declined gradually over the years—surveys in 1970, 1982 and 1997 revealed 532, 344 and 214 elephants, respectively (Jayawardene 2001; p.44). This is mostly due to a government ban on the capture of elephants from the wild that was instituted in the early 1970s (Lorimer 2010; p.499).

3.7 Sri Lanka’s human-elephant conflict

The conflict between humans and elephants in Sri Lanka has prevailed for centuries. Robert Knox, who was held prisoner by the King of Kandy from 1660 to 1697, described the damage caused by wild elephants to crops (Santiapillai et al. 2010; p.27). Writing in 1681, he noted that “they (i.e. elephants) do also great damage in their grounds, by night coming into their fields and eating up their corn and likewise their coker-nut trees...” (Santiapillai et al. 2010; p.27).

Unfortunately, the problem has worsened significantly in recent times, and many villagers who live close to or within elephant habitat deal have to contend with marauding elephants virtually on a day-to-day basis, especially when their fields are being cultivated. The Sri Lankan media is full of stories of challenges faced by these often desperately poor people; these excerpts from a recent newspaper article are a typical example (Perera 2010; p.1):

Government official Archchilage Weerasinghe says some 283 ha of land have been cultivated in the Konweva area for paddy. But the elephants have impeded further development. “We don’t plant in an area of about 350 acres (142 ha) because of elephants,” Weerasinghe says. The elephants cross the fields at will and trample the crops, villagers complain. Weerasinghe gave us a tour of some areas where the animals had roamed the week before. From a distance, it looked like the aftermath of a

111 meteorite storm. The elephants have also destroyed hundreds of coconut trees lining the village, he explains. It does not help that villagers are not entitled to compensation for damages to crops caused by the elephants if their fields are on government-owned land, which locals often use without permits. In short, residents here say, elephants are far from the adorable creatures seen on television. In July, a villager was trampled to death, and his wife injured, in an elephant attack. According to Weerasinghe, at least three villagers have been killed by elephants in the past year. One of the major points raised in the report is that, unlike many others in Sri Lanka, these farmers do not view elephants with affection, but instead as a dangerous threat to their livelihoods, and, on occasion, even their lives. Bandara and Tisdell (2002; p.492) describe how this negative attitude towards the species among people who live in or near elephant habitat is a threat to their conservation in the long term:

The animosity of the farmers and rural communities in the vicinity of the PAs and other nature reserves is an unfavourable portend for future elephant survival, particularly given the trend toward decentralised wildlife management throughout the Asian elephant’s range. Under current conditions, most local farmers would eliminate elephants from their environment if given the choice. Therefore, it is important that solutions to this problem take into account the welfare of the local people while meeting the conservation needs of elephants, as discussed in detail in Chapter 4.

112 Chapter 4: An approach to mitigate Sri Lanka’s human- elephant conflict

Although the Asian elephant is a flagship species and a cultural icon, people who share land and resources with elephants often view them as “dangerous, destructive agricultural pests” (Bandara and Tisdell 2002; p.491). In Chapter 3, I described human- elephant conflict across Asia and introduced it as a major conservation-related issue in Sri Lanka. In Chapter 4, I provide more details on the conflict within Sri Lanka, highlighting its profound socio-economic impact on locals and illustrating how it threatens elephant conservation in the country. I evaluate measures used by both the Department of Wildlife and farmers to protect people and crops from marauding elephants and discuss deficiencies in the country’s overall approach towards human- elephant conflict. Finally, I explore how recent research on elephant behaviour could be used to develop a landscape-based approach to mitigate the conflict.

4.1 The importance of resolving human-elephant conflict

As described previously, human-elephant conflict continues to have negative impacts on humans and elephants alike. I now explore these impacts in more detail, illustrating why the conflict is viewed as one of Sri Lanka’s major conservation-related challenges.

Overview of human-elephant conflict in Sri Lanka

Human-elephant conflict in Sri Lanka has worsened in recent times. From 1992 to 2001, 526 humans and 1,261 elephants died in the conflict at an annual rate of 53 humans and 126 elephants (Corea 2007; p.3). However, as presented in Table 4.1, the corresponding averages for the time period of 2006 – 2010 were 71 humans and 206 elephants.

113 Table 4.1: Elephant and human deaths in Sri Lanka, 2006 – 2010 (from Fernando et al. 2011; Yatawara 2011)

Year Elephant deaths Human deaths 2006 163 79 2007 189 72 2008 224 71 2009 228 50 2010 227 81 Total 1,031 352 Annual average 206 71

The impact of the conflict on people

Santiapillai et al. (2010; p.25) attempt to put in perspective the number of human deaths caused by elephants, by pointing out that, on average, about 1,500 people are killed annually by poisonous snakes in Sri Lanka. However, they acknowledge that “manslaughter by elephants receives greater publicity and evokes stronger emotions” (Santiapillai et al. 2010; p.25). Furthermore, the increasing numbers of human deaths due to elephants are given significant publicity in the local media.

Elephants also cause tremendous damage to crops and sometimes destroy people’s homes in attempts to access stored food. A recent study which recorded elephant attacks in southeast Sri Lanka describes the damage caused in the area (Campos-Arceiz et al. 2009; p.8):

From June 2004 to May 2005, we recorded 975 human-elephant conflict (HEC) incidents. Crop damage was the most common form of HEC, present in 92% of the incidents. Elephants damaged more than 30 different types of cultivated plants of which banana, paddy, and coconut were the most common... Seasonal crops were raided mainly when they were ripe. House damage accounted for 10.9% of the HEC incidents (118 houses were attacked). Elephants attacked houses to consume stored foods (mainly paddy), to drink from water deposits (e.g. home wells or water barrels), or—as claimed by some house owners—to consume salt from kitchens. Several houses were completely destroyed (22.0%) and many more suffered serious damages that could not be repaired (49.2%).

114 Bandara and Tisdell (2004; p.104) estimate crop damage by elephants at about Rs. 1,121 million (about US$ 10 million) annually, a figure has that probably increased significantly as their data are from the late 1990s. However, monetary estimates of damage do not adequately capture the conflict’s impact on human livelihoods, as a herd of elephants can destroy a farmer’s entire season’s crops in a single night. A recent newspaper article on elephant damage in northwest Sri Lanka provides some insights:

S. M. Sirisena, chairperson of a farmer association, said that herds of elephants have been destroying crops. “Farmers spend sleepless nights as the damage is high. An acre of paddy land costs us Rs. 15,000 for a single harvest,” Mr. Sirisena said... Calling the crop losses “unbearable”, another victim, a woman from the Puttalam district said that bank loans taken for farming could no longer be repaid. Almost all divisions in that district have been affected. More than 1,000 acres of coconut growing land was recently trampled on by elephants (Bhattacharjya 2010; p.1). An account by a former director of the Department of Wildlife also illustrates the impact of elephant damage on poor farmers’ livelihoods (Saparamadu 2006; p.136):

A herd of elephants can cause total havoc on a peasant’s chena and the loss of the crop means abject poverty for a whole year. I once saw a coconut plantation of about 5 acres at Buttala, which had been planted and looked after by a peasant for 10 years, totally destroyed by a herd of elephants in a single night. No amount of financial compensation can cover such a loss. Overall, human-elephant conflict has a profound socio-economic impact, since, as pointed out in Chapter 3, the people affected are among the poorest in the country. This is illustrated in a recent study done across areas of high human-elephant conflict101— almost 90% of the farmers interviewed were poor, earning less than Rs. 120,000 (US$ 1,000) an year; over 40% of them suffered damages exceeding a fifth of their income (Santiapillai et al. 2010; p.23). The authors describe the profound impact of the conflict on people they surveyed:

The plight of the small and marginal farmers is miserable, as evident from reports of their suicides. Many of them are in deep debt and are unable to move out of their poverty vortex. Their problems are compounded by wildlife, especially elephants, whose incursions into their farmlands can ruin their already perilous lives (Santiapillai et al. 2010; p.23).

101 The study included the North-Western, North-Central, Central, Uva and Eastern provinces.

115 The impact of the conflict on elephants

As marauding elephants destroy their crops and put their livelihoods at risk, farmers retaliate, often through cruel means. For example, they kill elephants using poison, concealed in palatable fruits such as pineapples (Santiapillai et al. 2010; p.27). Recently there has been increased use of a device called hakka patas—mixture of explosive matter, lead and iron made into a ball—that is inserted into a cucumber or a pumpkin (Bandara 2010). Unfortunate elephants that eat these fruits receive severe injuries to their mouths and trunks, ultimately suffering agonising deaths. The causes of elephant deaths from 1998 to 2007 are presented in Table 4.2.

Table 4.2: Causes of elephant deaths in Sri Lanka, 1998 – 2007 (from Yatawara 2010)

Cause of death Number of elephants % of total Gun shots 774 56% Electrocuted 135 10% Poisoned 22 2% Accidents 109 8% Natural deaths 63 5% Unidentified causes 285 21% Total 1,388 100%

Since 1950, at least 5,200 elephants have perished in Sri Lanka’s human-elephant conflict, a number equivalent to the country’s current elephant population 102 (Santiapillai et al. 2006; p.95). Comparing the average number of elephants killed in Sri Lanka (about 200 per annum) with the corresponding number in India (about 300) indicates that the problem, at least for elephants103, is much worse in Sri Lanka. India’s number of elephant deaths is only 50% higher than that in Sri Lanka although India’s total elephant population is 5 times greater. About 70% of the elephants killed in Sri Lanka’s conflict are adult bulls (Perera 2007; p.16). Consequently, researchers have

102 The authors have stated in 2006, using 2005 figures, that a minimum of 4,200 elephants have been killed in the conflict since 1950. An additional 1,031 elephants were killed in the period from 2006 – 2010, bringing the total to over 5,200. 103 In terms of the number of annual human deaths per elephant, the numbers from Sri Lanka and India are comparable at 0.01 for India (30,000 elephants and 300 deaths) and 0.011 for Sri Lanka (6,000 elephants and 70 deaths). However, when put in perspective of the total populations in the two countries it is evident why the human deaths are perceived as a much greater problem—India’s population is over 50 times that of Sri Lanka, while its number of deaths from human-elephant conflict is less than 5 times the corresponding number for Sri Lanka.

116 recorded skewed sex ratios among elephants in high-conflict areas, which put those populations in jeopardy:

In Sri Lanka, the average overall adult male to female sex ratio was found to be 1:3. But in certain areas in the northwest, as a result of human- elephant conflict, the adult male to female sex ratio is 1:7, indicating a greater loss of bull elephants. What is worrying is that in addition to the bulls, even cows and calves have also been killed in the conflict (Santiapillai et al. 2006; p.95).

Resolving the conflict—a national priority

The Department of Wildlife recognises human-elephant conflict as one of Sri Lanka’s most significant conservation-related issues. Moreover, the 2006 National Policy for the Conservation and Management of Wild Elephants states that:

The human-elephant conflict has been recognized as the most serious problem facing elephant conservation in Sri Lanka and has replaced poaching as the biggest threat to elephant conservation in the island. It is real and it is leading in just one direction; to the eventual destruction and elimination of elephants from non-conservation areas, and to the loss of crops, property and human lives (Government of Sri Lanka 2006; p.12). These views are echoed in statements by top officials such as the department’s Director General, Dr. Chandrawansa Pathiraja, who said that: “With the country heading towards rapid development, finding solutions to the human-elephant conflict is of paramount importance” (Yatawara 2011). Consequently the Department of Wildlife has spent more than 50 per cent of its working time and allocated budget to mitigate human-elephant conflict, numbers that are unlikely to reduce in the near future (Yatawara 2011). Another source estimated that Rs. 300 million (about US$ 2.7 million) is being spent each year to manage the conflict104 (Bhattacharjya 2010). Perhaps the clearest sign that the human-elephant conflict is an issue of national importance is that, when running for re-election in 2010, Sri Lanka’s President pledged to take further steps to mitigate the conflict. In his election manifesto he stated that:

Unplanned development and terrorism have destroyed our national parks and sanctuaries. Consequently elephants have invaded villages and have created an elephant human conflict. As a solution to this problem, I have implemented

104 The annual spend could easily exceed this figure in certain years. As noted later, a single elephant drive in 2006 cost Rs. 160 million (Uragoda 1994; p.100).

117 the “Gaja Mithuro” programme in 2009. As a part of this programme, ‘elephant control’ units and electric fences were established over 600 km during the last two years alone. I will complete these electric fences in all affected villages, including the north and the east, in the next three years (Rajapaksa 2010; p.69).

4.2 Measures taken to resolve the conflict and their effectiveness

The strategies used by particular groups to mitigate Sri Lanka’s human-elephant conflict are largely dependent on available resources (especially money). I now examine the key conflict mitigation measures taken by farmers and Department of Wildlife officials.

Traditional methods used by farmers

Farmers living in areas where crop raiding is frequent prefer traditional methods to chase away marauding elephants. Since crop raiding generally takes place at night, they keep watch from dusk till dawn, usually from huts built in trees (Perera 2009; p.45). When elephants arrive, people attempt to drive them away by shouting, wielding firebrands and lighting fire-cracker and/or thunder flashes called ali wedi105, which the Department of Wildlife sometimes distributes to farmers in high conflict areas (Fernando et al. 2008; p.6; Perera 2009; p.5). According a recent study, fire crackers are the most widely used method, mainly due to their “availability, low price and short-term effect”; although thunder flashes are more effective than fire-crackers, their use is limited due to low availability (Perera 2007; p.19).

In areas where people are used to living with elephants protection measures are often organised on a community basis. An example is the area near the Yala National Park in southeast Sri Lanka:

In Yala, crop protection was well organized in the single-crop permanent fields and individual farmers were assigned guard duty by farmer societies. Watch huts were constructed on the perimeter of tracts rather than on individual fields. Harvesting and guarding was co-ordinated so that the entire tract was guarded till harvesting was complete. Chena106 fields were guarded

105 Literally ‘elephant firecrackers’, these are large firecrackers approximately 25 cm in length and 2.5 cm diameter (Rodrigo 2011). 106 Traditional shifting agriculture, as described in Chapter 2.

118 on an individual basis, by keeping watch from a hut in the chena. When elephant presence was detected, neighbours co-operated to chase them from the vicinity, by shouting and lighting firecrackers (Fernando et al. 2005; p.2472). Sometimes farmers locate alarms on the periphery of their fields to detect elephants before they enter their fields (Fernando et al. 2008; p.4). Types of alarms used range from bottles and tin cans strung along trip wires to electric alarms made using car horns (Fernando et al. 2008; p.6). In areas of particularly intense conflict some farmers have resorted to measures that could injure elephants, such as shooting at them with shotguns, or throwing lighted coconut shells packed with ash soaked in kerosene oil onto them (Fernando et al. 2008; p.6; Perera 2009; p.45).

The effectiveness of these measures is mixed. In the past, when the conflict was less intense, and when elephants had abundant habitat and alternate food sources, traditional methods tended to prevent large-scale crop damage. Moreover, most methods used were non-lethal and did not cause significant harm to elephants (Fernando et al. 2008; p.5). Crop guarding tends to give the farmers a sense of empowerment, and if elephants can be successfully chased away most of the time, they are likely to tolerate occasional crop raiding (Fernando et al. 2008; p.5). However, in modern times, elephants facing dwindling habitats have begun to raid crops much more frequently and with a greater degree of desperation. Consequently they are becoming habituated to traditional measures and have even become more aggressive when confronted with these methods (Perera 2009; p.45).

Methods used by the Department of Wildlife

The main conflict mitigation methods used by the Department of Wildlife are translocating problem animals, conducting elephant drives and establishing electric fences. I will now address each of these methods in more detail.

119 Translocating problem animals

Although groups of elephants are sometimes captured and transferred to other areas, especially in Africa, in Sri Lanka, translocation is usually applied to adult bulls who habitually raid crops, destroy houses or kill people (Fernando et al. 2008; p.5). Between 2007 and 2011, 68 elephants (all male) have been translocated107 (Fernando et al. 2011; p.98). Identified individuals are darted with an anesthetic, captured, transported by truck and released, often into a PA (Fernando et al. 2011; p.99). If the correct animal is identified, it could help reduce the conflict in a particular area, since, in most cases, a few adult males are responsible for a large proportion of conflict (Fernando et al. 2008; p.24). However, there are several problems associated with translocation:

• Difficulties in identifying problem animals: As mentioned previously, most crop raiding takes place at night. This, and the usually considerable time lapse between the incidents and the arrival of the capture team, often “renders the positive identification of problem animals next to impossible” (Fernando et al. 2008; p.23). Therefore, there is a strong probability that the wrong elephant is translocated, rendering the entire exercise futile.

• High cost: Translocation is a costly exercise, especially in the context of Sri Lanka where the Department of Wildlife is chronically short of funds. It costs the department about US$ 2,500 to translocate a single elephant (Fernando et al. 2011; p.99). This includes “the services of two veterinarians, a capture team of about 15 persons, transport for personnel, a specialized truck to transport the elephant, as well as capture and restraining equipment, heavy machinery as anchor points and for ground clearing” (Fernando et al. 2008; p.24). Moreover, the average time of one week for a single capture/translocation is a significant drain on the department’s limited resources (Fernando et al. 2008; p.24).

• ‘Translocating the problem’: Research on translocated elephants in Sri Lanka suggests that adult males who are moved into unfamiliar surroundings are likely to be even more aggressive towards people (Fernando et al. 2008; p.24). Furthermore, their lack of knowledge of the area increases the risk of their coming into contact with people, making an accident even more likely.

107 Details are presented in Appendix 4.

120 Therefore, translocated elephants tend to do more damage in their new territories than in their previous ones (Fernando et al. 2011; p.99).

• Return of translocated elephants to their former ranges: In a well-known case, a translocated elephant travelled over 240 km in over a month to return to its previous range (Hettiarachchi 2009). Its movements could be tracked since it had been fitted with a radio collar prior to its release; the data revealed that the elephant had travelled through some heavily populated areas, putting humans as well as the elephant in danger (Hettiarachchi 2009). Therefore, elephants’ attachment to their home ranges could render their translocation ineffective.

Elephant drives

Elephant drives are used when large numbers of animals need to be driven from one place to another, usually from an area with increasing human habitation to a national park. Such drives have been conducted in Sri Lanka for many decades. Uragoda (1994; p.106-108) and Jayawardene (2010; p.31) have described 20 elephant drives which were conducted from the 1950s to the 1990s. Fernando et al. (2008; p.26) describe the process involved:

Elephant drives consist of a group of people walking through the forest creating a lot of disturbance and attempting to chase the elephants in a particular direction. A major drive involves up to a few hundred people and is conducted over a period of a few months. It is usually conducted at the height of the dry season, due to logistical difficulties imposed by the wet season and to make use of the scarcity of water to move the elephants in a particular direction. It involves people starting from one end of the drive area, lighting large firecrackers called 'thunder flashes', flares, banging on tin cans, shouting, shooting in the air etc. to drive elephants ahead of them. Water bodies in the drive area are guarded denying any access to the elephants so as to compel them to move to the next area... The cost of a drive is directly related to the extent of the drive area and the time it takes to complete the operation. Elephant drives are often well received by local communities as they are very visual exercises and the Department of Wildlife is perceived to be taking action to resolve human-elephant conflict in their area (Fernando et al. 2008; p.26). However, there have been several issues with this type of exercise:

121 • Failure to remove the real trouble-makers: Monitoring of drives in Sri Lanka has shown that elephant drives mostly remove herds from an area, but often fail to remove adult males who are responsible for most human-elephant conflict (Fernando et al. 2008; p.26).

• Endangering existing populations in PAs: As will be pointed out in section 4.5, most Sri Lanka’s PAs are already at carrying capacity in terms of elephant numbers. Therefore, translocating large numbers of elephants into these areas may exceed their carrying capacities, jeopardizing both translocated elephants and already existing populations. The practice would also adversely impact PA ecosystems.

• Stress on elephants: An elephant drive causes elephants high amounts of stress. Consequently those who backtrack may become more aggressive, often leading to increased levels of conflict (Fernando et al. 2008; p.27).

• High cost: An elephant drive conducted to clear elephants out of new agricultural land established due to the Walawe Left Bank irrigation scheme cost over US$ 1.6 million and took 1.5 years to complete (Fernando et al. 2011; p.98). Despite this large investment, the drive was a failure as described in the following news report (Rodrigo 2011; p.2):

Dr. Fernando estimates there are 300 – 400 elephants still in Hambantota. If the 2006 elephant drive which cost Rs. 160 million was successful, then how could there be such a large number of elephants remaining in Hambantota, questions Dr Fernando highlighting that the elephant drives are a failure. “Only the young and some herds mainly consisting of females could be moved by the elephant drive. The large male pachyderms who are really the trouble makers remained in the area,” pointed out the elephant biologist. “Like humans, elephants too are attached to their homes and those who have higher instincts also returned”.

122 Establishment of electric fences

The use of electric fences to deter elephants was first attempted in Sri Lanka in 1966 (Santiapillai et al. 2006; p.98). Since then, the Department of Wildlife has constructed many hundreds of kilometres of such fences in conflict areas108. Electric fences—which carry a high voltage current at low amperage—do not injure or disable elephants, but give them a sharp, memorable shock upon contact (Sukumar and Santiapillai 2006; p.12). The effectiveness and longevity of electric fences depends on their construction and maintenance, as described by Fernando et al. (2008; p.15):

Use of high quality components such as porcelain or UV109 stabilized insulators and other hardware, non corroding wires etc. make fences long lasting and can even last decades if constructed with quality materials coupled with good maintenance. Good maintenance requires daily inspection of fence for wire breaks, loosening, current leakage from objects or plants touching the wires, replacing damaged posts, inspection of batteries, powering energizer etc. Clearing vegetation from underneath the fence is essential to prevent current leakage from plants touching the wires especially where live wires run close to the ground. This can be done by manual clearing or use of chemicals. Although electric fences have been effective in some circumstances, there have been several issues associated with their use:

• Location of fence: Several fences erected on national park boundaries have elephants on both sides of them—due to the wide ranging behaviour of the animals—rendering them ineffective (Fernando et al. 2008; p.17).

• Issues with structure and maintenance: Some fences have major flaws in their design, structure and maintenance, severely reducing their effectiveness (Santiapillai et al. 2006; p.99).

• High cost: The current cost of the standard electric fencing used by the Department of Wildlife is around US$ 4,500 per km (Fernando et al. 2011; p.99).

• Elephants breaching fences: Elephants, are highly intelligent, and often figure out how to circumvent or destroy poorly constructed fences (Fernando et al. 2011; p.99). Their tactics include breaching fences using their tusks (which do

108 As described in the Sri Lankan President’s manifesto, referred to in section 4.1. 109 Ultra-violet.

123 not conduct electricity), pushing or kicking down fence posts110 and stepping over the fence using the thick soles of their feet to depress the wires (Fernando et al. 2008; p.14). Some animals even destroy fences by toppling trees onto them (Sukumar and Santiapillai 2006; p.13).

Despite these potential weaknesses, if used properly, an electric fence could be very effective in keeping elephants away from crops, especially by “making it visible and obvious, keeping it fully functional, and through judicious, strategic placement” (Fernando et al. 2008; p.14). For example, in Malaysia, where several thousand kilometres of fencing have been erected, it has been calculated that over a period of 5 years they may have saved crops valued as much as 70 times installation costs (Sukumar and Santiapillai 2006; p.12). As described later in section 4.5, electric fencing could help implement a landscape-based approach to mitigate human-elephant conflict. Before this, however, I identify the flaws in the current approach to the conflict and outline the nature of the required solution.

4.3 The problem with current approaches used to resolve human- elephant conflict

The current set of approaches used by farmers and the Department of Wildlife has not provided a lasting solution to Sri Lanka’s human-elephant conflict. The root-cause of this failure is the underlying traditional approach to conservation111. I now explore this problem in more detail, addressing each element of this approach, and outlining its consequences.

The idea of nature as static

As noted in Chapter 2, the initial impetus to establish nature reserves in Sri Lanka during British times was based on the belief that nature should be preserved in its ‘pristine’ form (i.e. a static view of nature). Consequently, although Sri Lanka has set aside a significant proportion of its land area as PAs, as Santiapillai et al. (2006; p.95)

110 That have not been electrified. 111 The traditional approach to conservation was discussed in detail in Chapter 1.

124 point out, “most of the reserves are established on marginal land to balance intensive land use in surrounding areas”. In the context of elephant conservation, they describe the situation as having “reversed from one in which human islands existed in a sea of elephants, to a sea of people with elephant islands" (Santiapillai et al. 2006; p.96).

The lack of an ecosystem-based approach and inadequate management of designated buffer zones near PAs has resulted in land-use that exacerbates human-elephant conflict. This phenomenon is exemplified by the conversion of large areas of forests bordering several key parks into sugarcane plantations:

In the Moneragala District, large areas of forest close to Yala (North), Gal Oya and Uda Walawe National Parks were converted to sugarcane plantations, despite the known appetite of elephants for sugarcane. Predictably the crops have been attacked by elephants resulting in massive economic losses to the sugarcane industry. The problem has been particularly severe in and around the Pelwatte Sugar Company plantation. Electrified fencing was established, but the elephant depredations go on, and both elephants and human beings are killed (Santiapillai and Jackson 1990; p.64). It was also pointed out in Chapter 2 that, despite the use of terms such as ‘PA network’ and ‘network of forest reserves’, there have been few efforts to achieve these networks, resulting in a small, isolated set of PAs. Of the 20 national parks in the country, only two are more than 1,000 km2 in size, while 16 are under 500 km2 (Survey Department of Sri Lanka 2008; p.86). Consequently, elephants are often compelled to range outside PAs for their resource requirements112, and as many of these areas are under permanent agriculture, this inevitably leads to escalating human-elephant conflict.

As pointed out in Chapter 1, the view that nature is static has often led to a reductionist approach to land management. This has happened in Sri Lanka, where land use planning and management lacks cohesiveness and a systematic approach113. Land is divided into categories such as ‘urban land’, ‘paddy land’, ‘forest land’ and ‘national parks’ that are managed by different departments and/or ministries that typically do not consult each other when developing land-use policies. Although the Department of Wildlife—whose jurisdiction regarding land management is limited to the PA system—is the only agency responsible for elephant conservation and conflict mitigation, a basic analysis of

112 Especially in the dry season. 113 As pointed out in Chapter 2.

125 elephant populations within reserves shows that the department alone cannot possibly address the issue by itself (Table 4.3).

Table 4.3: Numbers of elephants within Sri Lanka’s protected area network (from Perera 2007; p.15)

Number of elephants Name of PA Category Extent (ha) (minimum – maximum) Bundala National Park 6,216 80 – 100 Flood Plains National Park 17,350 50 – 100 Gal Oya National Park 62,936 150 – 200 Lahugala-Kitulana National Park 1,554 100 – 150 Maduaru Oya National Park 58,850 150 – 200 Madhu Road/ Mavillu Sanctuary 34,677 100 – 200 Minneriya-Giritale National Park 6,693 300 – 400 Peak Wilderness Sanctuary 22,380 50 – 60 Ruhuna National Park 126,782 300 – 350 Sinharaja National Heritage 8,864 0 – 10 Somawathiya National Park 37,762 50 – 100 Tirikonamadu Nature Reserve 25,019 200 – 250 Uda Walawe National Park 30,821 150 – 250 Victoria-Randenigala- Sanctuary 42,087 40 – 60 Rantambe Wasgomuwa National Park 37,063 150 – 250 Wilpattu National Park 131,693 100 – 150 Yala East National Park 18,149 30 – 40 Total 668,896 2,000 – 2,870

Comparing the figures in Table 4.3 with the country’s total number of elephants it is evident that a significant proportion of Sri Lanka’s elephants are found outside its reserve network114. Furthermore, even elephants found within PAs range outside them, especially in the dry season, and it is estimated that about 70% of elephant habitat in Sri Lanka lies outside its PA network (Santiapillai et al. 2010; p.21). Therefore, although the Department of Wildlife is responsible for managing elephants and mitigating the conflict, it only has jurisdiction over 30% of the entire range of elephants in the country!

114 When the numbers in Table 4.3 were complied the total population of elephants in Sri Lanka was believed to be about 4,000. However, as indicated in Chapter 3, according to the 2011 elephant census, there are close to 6,000 elephants in the country.

126 Land managed by the Forest Department consists of important elephant habitat, as these lands are often contiguous to the PA network. Unfortunately, as pointed out in Chapter 2, the Department of Wildlife and the Forest Department have often had a tense relationship, partly stemming from their differing departmental objectives. Therefore, they have been unable to work together effectively to maintain habitat suitable for elephants and take joint conflict mitigation measures. As pointed out in Chapter 3, the Mahaweli Authority also manages vast areas that were formerly prime elephant habitat. Recent research indicates that human-elephant conflict in these areas is perhaps the most intense in the country, but it has been difficult to work on a solution because land jurisdiction has been divided between the Mahaweli Authority and the Department of Wildlife; the departments do not adequately communicate with each other on land-use decisions (Fernando et al. 2005; p.2467). The problem is expected to worsen as the on- going irrigation and development projects would convert more elephant habitat to irrigated agriculture and settlements (Santiapillai et al. 2006; p.96). Furthermore, human-elephant conflict is likely to flare up in areas affected by ethnic conflict until 2009 (i.e. in Sri Lanka’s north and east) where paddy fields are being cultivated after a hiatus of several decades.

The perception that people are apart from nature

As described in Chapter 2, the basic premise behind Sri Lanka’s current conservation framework—that dates back to colonial times—is that people are apart from nature. This view has generally persisted among those who have drafted conservation-related legislation in the post-independence era—as pointed out by Santiapillai et al. (2006; p.95): “conservation measures to date have mainly depended on legislative protection of the species and reservation of habitat - essentially keeping people and elephants apart”.

However, unlike other pieces of conservation legislation, the 2006 National Policy for the Conservation and Management of Wild Elephants clearly acknowledges that the needs of people who live with elephants must to be taken into account (Government of Sri Lanka 2006):

For elephant conservation to succeed in Sri Lanka, conservation efforts should be tied to the welfare of the people who bear the brunt of elephant

127 depredations. Elephant conservation in Sri Lanka should not simply be a matter of protection of the charismatic species; it should also be about the well-being and survival of the people who share their land with elephants. Based on this, several compensation schemes have been launched by the government to address the livelihood impacts of damage caused by elephants. One such scheme that is administered by the Department of Wildlife compensates for injury and death, and another by the Social Services Department for crop and property damage115 (Fernando et al. 2008; p.35). Unfortunately, these initiatives have not been successfully implemented, as evident from the results of a recent study done across conflict areas116 (Santiapillai et al. 2010; p.23):

Of the families interviewed, 94% received absolutely no compensation at all from the government for their losses due to elephant depredations. Only 6% received some compensation but that too, according to them, was too little and too late. In one case, a woman from Aluth Oya in the North- Central province received a paltry sum of Rs. 250 (US$ 2.50) as compensation from the Department of Wildlife for damages incurred from elephant attack. In a village known as Kotavehera in the Eastern province, a total of 27 houses were destroyed by wild elephants in 2007, for which no compensation was given by the department. A related issue is the perception among poor farmers that elephants are “property belonging to the state” (Bandara and Tisdell 2002; p.493). Therefore, these people believe the Department of Wildlife is also responsible for elephant control, and blame the department for any damage caused by elephants as illustrated in the events described in the following news article (Manukulasooriya 2008):

Nearly 1,000 villagers from Gomadiyagala in the Polpithigama police area launched an agitation campaign against wildlife officers on Tuesday after a 25-year-old man was killed by a wild elephant near the Hakwatuna Oya reservoir in the Polpithigama Divisional Secretariat division...The villagers complained that although money had been allocated to erect an electrified fence as a remedial measure, no action has yet been taken by the wildlife officers... Meanwhile, an unidentified gang had set fire to the wildlife office at Dummeeya, Halmillawewa about 4.30 p.m. on Tuesday.

115 Some compensation has been paid, as illustrated in Appendix 3, but overall, the compensation process has not been widely (and effectively) implemented. 116 This is the same study referred to in section 4.1.

128 In summary, Sri Lanka’s lack of an ecosystems-based approach, its fragmented land management framework, and the Department of Wildlife’s insufficient attention to the human dimension of the issue, has hampered the introduction of an effective, holistic, landscape-based solution to human-elephant conflict.

4.4 The nature of the required solution

Given the failure of the traditional approach to address Sri Lanka’s human-elephant conflict, an approach that considers the overall landscape is required. As pointed out by Sukumar and Santiapillai (2006; p.9):

...[T]he most practical and appropriate strategy of conserving elephants in the long term is the landscape approach that incorporates PAs within a matrix of other forms of land use including logging areas, non-timber product extraction areas, grazing land for livestock, and even some agricultural activity. I now outline the elements of such an approach, based on the key characteristics described in Chapter 1.

Considering the overall ecosystem

As pointed out previously, a key driver of human-elephant conflict has been poor planning regarding land-use in areas adjacent to PAs. Consequently, a systems-based approach should be used to identify appropriate types of land-use, which take into account the complexity of the entire ecosystem that both elephants and human beings live in. This process would enhance ecosystem resilience, and enable the ecosystem to adapt to expected disturbances such as the effects of climate change117. As Bengtsson et al. (2003; p.389) point out, the goal is to ensure the survival of ecosystems encompassing both the current reserves as well as the land surrounding them:

In a world increasingly modified by human activities, the conservation of biodiversity is essential as insurance to maintain resilient ecosystems and ensure a sustainable flow of ecosystem goods and services to society. However, existing reserves and national parks are unlikely to incorporate the long-term and large-scale dynamics of ecosystems. Hence, conservation

117 Described in Chapter 2.

129 strategies have to actively incorporate the large areas of land that are managed for human use. Furthermore, as described previously, the Department of Wildlife cannot manage human-elephant conflict by itself because elephants range far beyond areas under its jurisdiction. Therefore, tackling the problem requires a holistic approach, implemented in cooperation with other agencies, that identifies important elephant habitat and uses this information to make land-use decisions which help maintain essential areas, while meeting the needs of Sri Lanka’s growing population, especially poor subsistence farmers. As pointed out in Chapter 1, it is crucial to make land-use trade-offs at appropriate spatial levels—trade-offs are only acceptable if instituted at the proper scale, since if attempted at too small a scale, as is often done, conflict and environmental degradation can occur. The required approach is summarised by Fernando et al. (2005; p.2478):

Land use planning for conservation landscapes, to take conservation beyond the boundaries of PAs is becoming the new paradigm in large mammal conservation. Thus, if Sri Lanka’s, and indeed Asia’s, elephants are to survive the next century, conservation planners should identify priority elephant conservation areas and work with policy-makers to agree on land use within these landscapes.

Considering the needs of key stakeholders

Human-elephant conflict has a significant socio-economic dimension. As discussed before, poor farmers often view elephants as destructive and dangerous pests and would “eliminate elephants from their environment if they could” (Bandara and Tisdell 2002; p.492; Santiapillai et al. 2010; p.28). Therefore it is critical that people who live among elephants perceive benefits from tolerating these animals, as this would provide them with incentives to conserve elephants, in accordance with the principle of conservation through sustainable use118. Potential measures include providing poor farmers adequate compensation for elephant-related damage, and sharing conservation benefits with them, such as proceeds from elephant-related nature tourism. As pointed out in Chapter 1, it is

118 As described in Chapter 1, the IUCN Policy Statement on Sustainable Use of Wild Living Resources states that the “use of wild living resources, if sustainable, is an important conservation tool because the social and economic benefits derived from such use provide incentives for people to conserve them” (Fernando et al. 2008; p.6).

130 first necessary to identify relevant stakeholders. In the case of the human-elephant conflict poor farmers and the government are obvious stakeholders, but there would be others depending on the specific context. For example, this could include groups such as conservationists and local political authorities.

Having outlined the nature of a landscape-based approach to address Sri Lanka’s human-elephant conflict, I now explore how recent research on elephant behaviour could be used to develop such an approach.

4.5 An innovative approach based on recent research on elephant behaviour

Scientists at the Centre for Conservation and Research have studied Sri Lankan elephant behaviour for over 15 years, using radio telemetry techniques119. For example, 10 adult elephants120 were fitted with VHF121 collars, and their ranging behaviour was tracked from 1996 to 2001122 (Fernando et al. 2008; p.6). Using this research, which has revealed some hitherto unknown details about elephant behaviour, they have proposed an innovative approach to the address the country’s human-elephant conflict.

I first summarise the key findings of the research and their implications, before outlining the proposed approach.

(i) Elephants range extensively outside reserves

The elephants in the study ranged extensively outside national parks and reserves, some of them having home ranges that were entirely outside the PA network (Fernando et al. 2005; p.2473; Santiapillai et al. 2006; p.98). They used their habitat fairly intensively, generally making maximum use of available food and water resources.

119 As of 2011, the organisation had radio tracked about 50 elephants (IUCN 2000; point 7a). 120 7 females and 3 males. 121 Very High Frequency. 122 Since then, the scientists have studied many other elephants in various parts of the country using similar approaches.

131 Implications: Conservation of elephants both within reserves and in their ranges outside them is important, and, consequently, the current practice of translocating animals into national parks can be extremely harmful since these PAs are probably already close to their carrying capacity. According to Fernando et al. (2008; p.11):

Non-conservation areas are a very important segment of Asian elephant range and removal of elephants from such areas will have a major impact on the population... Our study also suggests that elephants use habitat fairly intensively and that food is a limiting resource for them. Given the large energetic requirements of a mega-herbivore, translocating large numbers of elephants into PAs may exceed their carrying capacity, jeopardizing both translocated elephants and populations already there. Therefore, in situ management of Asian elephants outside PAs is critical for their conservation.

(ii) The ranges are small, but elephants show high fidelity to these ranges

Observed home range sizes were relatively small, i.e., 30 – 230 km2; in comparison, home ranges that were 3 – 5 times larger have been recorded in southern India (Fernando et al. 2008; p.7). Elephants showed high fidelity to their home ranges, possibly to ensure familiarity with resource availability, distribution and fluctuations (Fernando et al. 2008; p.9).

Implications: Familiarity with home ranges maybe an important fitness trait for elephants, so translocation may be detrimental to their survival (reinforcing the findings in point (i) above).

(iii) Lack of significant seasonal variation

Although dry season ranges were larger, there was extensive overlap between wet and dry season home ranges in all tracked elephants (Fernando et al. 2008; p.9). In other words Sri Lankan elephants do not embark on significant migrations. This is probably because perennial water sources and adequate fodder are available in the area all year round, although these become somewhat depleted in the dry season.

132 Implications: There is no need for long-distance elephant migration corridors linking PAs as previously assumed, and provided for, in the Fauna and Flora Ordinance (Pethiyagoda 2000). Santiapillai et al. (2006) make this point, emphasising the need for conservation of elephant habitat outside reserves:

The idea of a corridor to promote the movement of wild elephants from one area to another was first put forward in Sri Lanka in the 1950s based on the assumption that elephants migrated long distances. However, current research has shown that elephants in Sri Lanka have well- circumscribed small home ranges of 30 – 230 km2 and that they do not have seasonal long distance migrations. In addition, the existence of a corridor many kilometres in length and a few kilometres wide, through cultivation areas, is likely to exacerbate the conflict with humans as elephants use such areas as refuges from which to raid crops. What has been observed through radio tracking 18 elephants in Sri Lanka is that many groups as well as males, have home ranges entirely outside PAs. Therefore, rather than proposing narrow corridors linking PAs, a strategy of managing these outside areas to form contiguous elephant habitat between PAs, would be a more desirable approach (Santiapillai et al. 2006; p.98).

(iv) Elephants extensively utilise lands under chena cultivation, especially in the dry season

As discussed in Chapter 2, chena cultivation, or traditional shifting agriculture, remains an important form of land use, especially in dry zone areas. Although British colonial authorities actively discouraged this form of agriculture as a wasteful, ‘primitive’ form of land-use, it constitutes an important form of land utilisation, especially by poor villagers who use it to supplement their diets and incomes. Fernando et al. (2006; p.255) found that chena cultivation actually creates and maintains ideal elephant habitat:

Our research has indicated that traditional shifting agriculture, locally known as chena, is highly beneficial to elephants and is compatible with elephant conservation outside PAs. Under this regime, a patch of mature forest, usually about a hectare, is cleared, burned, and cultivated annually for about 3 – 4 years, then abandoned in favour of another patch. Because chena is rain-fed agriculture, the cultivation is limited to the rainy season. The abandoned patch is allowed to regenerate for about 10 – 12 years, before the farmer returns to

133 it123. In most areas where shifting cultivation is practiced, farmers also cultivate small rice paddies immediately down-stream of small reservoirs that capture rainwater during the wet season. The result is a landscape matrix of vegetation patches in various stages of succession ranging from newly regenerating vegetation in fallow fields to secondary forests, providing substantial food resources and refuges for elephants, and a network of reservoirs providing abundant water. Such landscape matrices represent optimal elephant habitats. The research also revealed that elephants largely used chena land during the dry season, when farmers’ fields were fallow. During the wet season, when the fields were cultivated, they were able to find food elsewhere, especially within PAs, as described by Fernando et al. (2006; p.256):

The few depredations which occur in the chena patches during the wet season are by solitary males, whereas herds rarely raid cropped chena. Although the frequency of both males and females entering chena patches increases during the dry season, the fields are fallow so there is no conflict. The researchers monitored selected chenas near Yala National Park for signs of elephants, to understand how the animals utilised these areas. The following figure from an article published in 2005 clearly shows the pattern described above.

123 In many areas, chena farmers now cultivate their plots more or less on a permanent basis due to the shortage of land. However, they still leave their fields fallow during the dry season, creating the optimal elephant habitat described by the authors.

134

Figure 4.1: Elephants’ use of chenas throughout the year (Fernando et al. 2005; p.2474)

Implications: There is an opportunity to utilise a traditional land-use method to maintain suitable elephant habitat. In fact, as Fernando et al. (2006; p.256) point out, farmers play a ‘keystone species’ role, creating and maintaining optimal conditions for elephants. Paradoxically, the Department of Wildlife’s ‘hands-off’ management approach in national parks has resulted in the expansion of mature forests, which are less optimal for elephants124 (Fernando et al. 2006; p.256). Even if active intervention were attempted within PAs, such efforts would fail given its prohibitive cost. The few depredations in the chena patches that occur in the wet season by solitary male elephants could be managed by constructing electric fences around cultivated chenas.

Santiapillai et al. (2006; p.98) summarise how the solution is a cost effective approach to create and maintain elephant habitat, one that benefits both people and elephants:

Considering the fact that a single elephant requires over 100 kg of food every single day, growing food for elephants in the PAs is not a practical option. While it would be possible to manage PAs for elephants by intensive habitat management through practices similar to slash-and-burn

124 However, mature forests are beneficial to a host of other species.

135 agriculture, such management would be detrimental to many other species that require undisturbed mature forest. In addition, habitat management within the PAs to support high elephant densities would need to be conducted indefinitely, and would require immense long-term allocations of funds. The use of traditional agricultural practices to manage outside areas for elephants has the advantage that there is no expenditure incurred by the government and that both elephants and people benefit from it.

The proposed approach

Land-use planning should be used to delineate three zones across the landscape: ‘human habitat’ from which elephants are excluded, such as towns, urban areas and irrigated/permanent agriculture; ‘Managed Elephant Ranges’ on land in elephant habitat outside reserves, where elephants co-exist with chena farmers; and ‘protected areas’ which conform to the current PA network. Managed Elephant Ranges (or MERs) and PAs would be contiguous, while electric fences would be constructed on the ecological boundaries between ‘human habitat’ and MERs (Groom et al. 2006). Temporary electric fences will be constructed around chenas within MERs in the wet season when they are under cultivation; these will be taken down once crops are harvested, to allow elephants access to regenerating scrub forest in these chena plots. This is different from the current approach of building permanent electric fences in a largely haphazard manner in a way that restricts elephants’ access to food and water resources, especially in their habitat which lies outside PAs125 (Fernando et al. 2006).

The approach would drastically reduce the need for translocation and elephant drives, although problem animals that stray into areas designated as ‘human habitat’ would have to be moved back into MERs. Furthermore, human-elephant conflict would be reduced, because elephants would have better access to resources in areas outside PAs, especially in the dry season126. Perhaps even more importantly, an escalation of the conflict would be averted in some areas. For example (as will be discussed in more detail in Chapter 5), while current levels of conflict are relatively low in the southeast landscape, this is expected to increase drastically if more elephant habitat is taken over

125 Many of these fences would be taken down if the suggested approach were to be adopted. 126 In the wet season, during which chenas are cultivated, food and water for elephants in more plentiful both within PAs, and in land within MERs that is not under chena cultivation.

136 by development activities and permanent irrigated agriculture. Table 4.4 illustrates how this proposed approach constitutes a landscape-based approach:

Table 4.4: Analysis of how the proposed approach corresponds to a landscape approach

Element of landscape Corresponding elements of suggested approach to mitigate approach human-elephant conflict

Considering the overall The approach is not confined to the current PA network alone, ecosystem but also considers land-use in surrounding areas, within the overall ecosystem.

The approach determines land-use trade-offs at a landscape scale, through the proposed zoning system where resources within MERs are shared temporally between elephants and chena farmers.

Consider the needs of key The approach enables resource-sharing between elephants and stakeholders poor farmers—the key local stakeholder group. If, as expected, human-elephant conflict is reduced, one of Sri Lanka’s major conservation-related issues would be addressed. Furthermore, the suggested approach provides the Department of Wildlife with a cost-effective method of creating and maintaining optimal elephant habitat outside PAs. These points will be discussed in more detail in Chapter 6.

4.6 Overview of key points in Chapters 3 and 4

Chapters 3 and 4 use the example of the Asian elephant to highlight an important conservation-related issue within the developing world: how could countries address habitat destruction and biodiversity loss in the face of escalating demands for resources—especially land—from their rapidly growing human populations?

Chapter 3 introduced the endangered Asian elephant as a flagship species for conservation efforts across the 13 countries that comprise its range, given its importance from both an ecological point of view as well as from historical, religious and cultural perspectives. Despite the almost universal admiration for this majestic animal, escalating human-elephant conflict is resulting in deaths of both people and elephants

137 and damage to crops and property. This issue—perhaps the most significant faced in Asian elephant conservation—generates intense hostility towards the species among communities that bear the brunt of their depredations. It is not easily resolved because the root of the problem is the fragmentation and loss of elephant habitat due to human actions, a phenomenon that is expected to increase in the future. Sri Lanka, which contains over 10% of the Asian elephant population, exemplifies these issues—the elephant is closely intertwined within the country’s history and culture, but escalating human-elephant conflict threatens the survival of the species outside the PA network. This is a major issue, since 70% of elephant habitat is outside reserves.

Chapter 4 illustrates that human-elephant conflict is perhaps the most important conservation-related issue in Sri Lanka given the toll the conflict is having on humans and elephants alike. It explores traditional measures—used mainly by farmers—to drive away crop-raiding elephants as well as conflict mitigation techniques used by the Department of Wildlife. Unfortunately, these measures have failed to make an appreciable impact given the country’s traditional approach to conservation that lacks an ecosystem approach and does not adequately incorporate the needs of affected locals. This section addresses the second research question stated in section 1.6: ‘What have been the impacts of the traditional approach on elephant conservation and human- elephant conflict’. Sections 4.4 and 4.5 begin to answer the third research question in section 1.6, namely: ‘Could a landscape-based approach help develop a solution to conserve elephants while addressing stakeholder needs?’ Section 4.4 outlines the nature of the required approach while section 4.5 proposes a landscape-based approach based on recent research on elephant behaviour that would permit resource-sharing between farmers and elephants on a temporal basis.

The question now is how this type of approach could be implemented within Sri Lanka’s current socio-economic and conservation landscape (described in Chapter 2). To keep the analysis practical, I now apply the approach to a specific landscape—Sri Lanka’s southeast region—that contains key elephant habitat but where rapid development is requiring policymakers to make difficult decisions regarding land-use. This region is introduced in Chapter 5.

138 Chapter 5: Introduction to case study

In Chapter 4, I proposed a landscape-based approach that could help mitigate Sri Lanka’s human-elephant conflict. I now apply this approach to a specific area—the country’s southeast landscape—in Chapters 5, 6 and 7. In Chapter 5, I introduce my study area, describing its boundaries and extent, providing an overview of its history and socio-economic context, and outlining development initiatives taking place within it. I then analyse landscape-use through GIS-based maps, and describe local livelihoods based on semi-structured interviews and surveys. Finally, I provide an overview of the area’s elephant population and human-elephant conflict, and map out potential MERs.

In Chapter 6, I use my survey and interview results to conduct a detailed stakeholder analysis, identifying the range of issues and viewpoints of various groups that could influence or be impacted by my suggested approach of establishing MERs. In Chapter 7, I address a major concern of several important stakeholder groups—the economic and financial impact of MERs. I discuss findings of an economic model I developed to evaluate economic benefits and costs of resource-use within the landscape. My analysis illustrates how different types of resource-use would impact major stakeholders, and identifies benefits and costs that are not captured by market forces (and therefore largely ignored in stakeholder decision-making processes).

5.1 Sri Lanka’s southeast landscape

My case study covers Sri Lanka’s southeast landscape, an area that contains important elephant habitat and a significant proportion of the country’s elephants127. Although the area’s human-elephant conflict has historically been lower than in other parts of the country—notably the northeast—it has escalated in recent years, largely as result of the rapid development occurring within the landscape. Moreover, this is one of Sri Lanka’s main centres of agricultural production, and the growth of agricultural activities due to a burgeoning human population is further encroaching on elephant habitat, thus exacerbating the conflict. Given these factors, Sri Lanka’s southeast landscape is a good place to explore the feasibility of implementing MERs to mitigate the growing conflict.

127 Details are provided in section 5.4.

139 In this sub-section I identify areas covered in my case study, provide an overview of the region’s historical and socio-economic context, and describe development activities taking place within the landscape.

Areas covered in the case study

Sri Lanka is divided into 9 provinces and 24 districts for administrative purposes; each district is further divided into District Secretariat Divisions. My case study focuses on Ampara, Moneragala and Hamabantota—three districts in southeast Sri Lanka—which together cover 12,226 km2, or about 19% of the country’s land area (Department of Census and Statistics 2011; Table 1.1). Figure 5.1 shows Sri Lanka’s provinces and districts (left), and delineates the District Secretariat Divisions contained within each district, highlighting the three covered in the case study (right).

140

Sri Lanka’s 24 administrative districts The three districts that are the focus of the case study (in dark colours), and the outlines of District Secretariat Divisions within each district (map created using GIS data obtained from the Survey Department of Sri Lanka)

Figure 5.1: Sri Lanka’s 24 administrative districts, and the three districts that are the focus of the case study (UN Humanitarian Information Center 2005)

To keep the analysis manageable I focus on specific District Secretariat Divisions within each of the three districts of Ampara, Moneragala and Hamabantota. I selected these based on the prevalence of human-elephant conflict, the presence of PAs and other important elephant habitat, the existence of chena cultivation and presence of areas targeted for rapid development. Table 5.1 provides details on the specific District Secretariat Divisions covered.

141 Table 5.1: Details on District Secretariat Divisions covered in the case study

District Divisional Secretariat Division Area (km2) Ampara Lahugala 830 Potuvil 214 Hambantota Agunakolapelessa 174 Ambalantota 206 Hambantota 338 Lunugamvehera 306 Suriyawewa 184 Tangalla 36 Tissamaharama 781 Moneragala Badalkumbura 228 Buttala 725 Kataragama 563 Moneragala 292 Sevanagala 187 Siyabalanduwa 867 Thanamalvila 480 Ratnapura Weligepola 44 Balangoda 12 Total 6,467

The landscape being studied is 6,467 km2 in extent, comprising about 10% of Sri Lanka’s land area128. A total of six District Divisions have been left out of Ampara, Moneragala and Hamabantota districts129, while two have been added from Ratnapura District because they contain part of Uda Walawe National Park. Figure 5.2 provides a detailed view of the District Divisions included in my study area.

128 As noted in Chapter 2, Sri Lanka’s total land area is 65,610 km2. 129 These are Damana (Ampara); Katuwana, Weeraketiya (Hambantota); and Madulla, Medagama, Wellawaya (Moneragala).

142

Figure 5.2: District Divisions included in landscape considered in the case study (map created using GIS data obtained from the Survey Department of Sri Lanka)

143 Socio-economic overview of study area

Hambantota, Ampara and Moneragala are sparsely populated districts, as evident in the date presented in Table 5.2.

Table 5.2: Land area, population and population densities of Hambantota, Ampara and Moneragala districts, 2010 (compiled using data from the Department of Census and Statistics 2011)

District Land area130 (km2) Population estimate Population density ( ‘000) (people per km2) Hambantota 2,496 571 229 Ampara 4,222 644 153 Moneragala 5,508 440 80 Sri Lanka 62,705 20,653 329

Table 5.2 shows that the population densities in each of the three districts are well below the national average. Together they contain 8% of Sri Lanka’s population in an area covering 19% of the country’s land area, resulting in a combined population density of only 135 people per km2 compared to 329 for the entire country. The per capita income of people living in Moneragala and Ampara is lower than national levels, while that for those living in Hambantota is higher (Table 5.3).

Table 5.3: Median per capita incomes of people living Ampara, Hambantota and Moneragala districts, 2009/2010 (compiled using data from Department of Census and Statistics 2011; p.11)

District Median monthly per capita income (Rs.) Hambantota 6,553 Moneragala 4,391 Ampara 4,120 Overall for country 5,803

Hambantota, Moneragala and Ampara were ranked 3, 19 and 21 among the 22 districts131 covered by the Household Expenditure and Income Survey for 2009/2010

130 Excludes inland water bodies.

144 conducted by the Department of Census and Statistics (Department of Census and Statistics 2011; p.11). Hambantota’s relatively high rank is due to recent development activities centred on that district (as described in the next sub-section). These income figures are consistent with poverty indicators for the three districts—Hambantota has a lower proportion of poor people and households compared to the national average, while the other two districts fare significantly worse (Table 5.4).

Table 5.4: Poverty indicators for Ampara, Hambantota and Moneragala, distrcits, 2009/2010 (compiled using data from Department of Census and Statistics 2011; p.11)

District Percentage of poor people Percentage of poor below the poverty line households in district Hambantota 6.9 5.4 Ampara 11.8 10.0 Moneragala 14.5 13.9 Overall for country 8.9 7.0

131 Three districts in the northern province—Vavuniya, Mullaitivu and Mannar—were not included in the survey due to difficulties in conducting the survey in those areas due to the ongoing ethnic conflict in those areas (when the survey was conducted).

145 History of Sri Lanka’s southeast

During the era of ancient kingdoms, Sri Lanka was divided into three kingdoms: ‘Pihiti’ or ‘Raja Rata’ with its capital in Anuradhapura, covering the north-central and northern areas; ‘Maya’ in the west with its capital at Kelaniya; and ‘Rohana’ or ‘Ruhunu’ in the south and southeast with its capital at Magama (de Silva and de Silva 2007; p.208). The boundaries of these kingdoms are shown in Figure 5.3.

Figure 5.3: Sri Lanka’s three ancient kingdoms in about 300 B.C. (de Silva 2009; p.3)

As indicated in Figure 5.3, my study area was part of the Ruhunu (or Rohana) kingdom. Although Anuradhapura was the main power centre in the country during this time132, the Ruhunu Kingdom became very important during the time of King Kavantissa in the 2nd century B.C. because Anuradhapura had fallen to Chola invaders from South India (de Silva 2009). Kavantissa’s son, Dutugemunu, was able to defeat the Chola King and

132 The King who ruled from Anuradhapura was generally recognised as the King of Sri Lanka, while the other two were mostly regional monarchs.

146 reunite all of Sri Lanka, securing his place in the country’s history as one of its greatest kings (Holt 2011; p.30).

Dutugemunu’s father’s capital, Magama, is believed to be modern-day Tissamaharama, one of the main towns in the landscape selected for the case study (de Silva 2009). Tissamaharama is just a few kilometres away from the famous Yala National Park also located in the landscape. In fact, Jazeel (2005; p.202) points out that Yala Park also features in Sri Lankan history and folklore:

The park is said to overlie the former Sinhalese kingdom, Ruhuna, which was the jungle refuge of the Sinhalese hero king Dutugemunu in the second century B.C. Legendary Sri Lankan history suggests that in this period the South Indian Tamil Elara conquered Anuradhapura, the capital of the northern Sinhalese kingdom. Dutugemunu, the deposed Sinhalese heir to Anuradhapura’s throne, took refuge in the wealthy and irrigated Ruhuna kingdom in the south. This inhospitable jungle, located at the very edge of the island, emerged as the home that nurtured and protected Dutugemunu, preparing him to wage a 15-year war with Elara that he eventually won to regain (Sinhalese) control in the north, at that time the heart of the island. These events are dramatized as central themes in the final chapter of the sacred Buddhist text the Mahavamsa, subtitled The great chronicle of Ceylon, its imagery feeding into contemporary narratives of national history, memory and geography. Consequently the entire area surrounding Tissamaharama town and the Yala National Park is scattered with important archaeological and religious sites, such as the Buddhist temples at Tissamaharama and Yatala, the Buddhist Monastery at Situlpahuwa and the Hindu temple at Kataragama.

147

Figure 5.4: Sithulpahuwa temple, within Yala National Park (photograph taken on 18/08/2010, by S. Kalpage)

As pointed out in Chapter 2, with the decline of the Kingdoms in Anuradhapura and Polonnaruwa, the dry zone was abandoned and Sri Lanka’s political centres shifted to the wet zone in the southwest. As result, much of the civilisation in the country’s southeast also went into decline and the area became covered by jungle. By the time of the British, it was one of the most under-developed areas in the country, as described by Leonard Woolf133 who served in the British Civil Service in Sri Lanka at the beginning of the 20th century:

I was nearly three years in Hambantota as the Assistant Government Agent. I grew to be extremely fond of the place and its peace. It was pure Sinhalese, no planters, no Europeans at all except a District Judge in Tangalla, two Irrigation Engineers and an Assistant Superintendent of Police. It was entirely rural agricultural in the west, and a vast stretch of jungle with some sanctuary in the east. There were no real towns, no lorry, hardly any roads (Woolf 2011; p.269).

133 Leonard Woolf is better known in the West as the husband who published the writings of his wife, Virginia Woolf (Fernando et al. 2011; p.98).

148 Development activities in the study area

Although there were some development activities in southeast Sri Lanka after independence, the area still lagged behind the rest of the country in developmental terms. This situation changed when Mahinda Rajapaksa was elected President in 2004. Unlike previous Presidents, who were mostly from western and central parts Sri Lanka, President Rajapaksa is from Weeraketiya in the Hambantota District. He, and several members of his government, made it a priority to develop southeast Sri Lanka. Their plans centred on the port-city of Hambantota, which was to be developed as a regional hub rivalling that of the capital, Colombo. A newspaper article describing the President’s aspirations included quotes from the President’s brother, the Minister for Ports, Aviation and Water Resources (Sirimane 2009):

Hambantota was once considered an area frequented by elephants and robbers, and under developed to such an extent that there wasn't even a decent tea kiosk to quench your thirst. There was no drinking water available, let alone sanitation. “Many people passed Hambantota on their way to Kataragama, but no one wanted to stop over at Hambantota as it did not have anything to offer,” said Ports, Aviation and Water Resources Minister Chamal Rajapaksa. This is the past. But, today a 'Development Tsunami' spearheaded by the government, has hit Hambantota and the South. “We have just started the development and in two years Hambantota would be one of the most developed districts in Sri Lanka,” he assured. “The other neighbouring districts, Matara, Galle and especially Moneragala will also reap the benefits by this development and would turn prosperous”, he said. In addition to the ruling party’s interest in the southeast, the Ampara district, which was formerly part of an area claimed as a separate state by the Tamil Tigers, has been a target for development activities under the ‘Negenahira Navodaya’ (‘Eastern Revival’) scheme since the end of hostilities (Krishnaswamy 2010). Consequently, there have been a number of state-led development efforts within the landscape over the last few years. These include a new international port, a new international airport, multipurpose irrigation schemes, rural electrification schemes, a road rehabilitation program and even the construction of a new international cricket stadium for the 2011 Cricket World Cup. Some of these projects are described in Appendix 5.

149 The irrigation and hydropower schemes around the Walawe Basin are examples of large-scale development projects within the landscape. There are 15 irrigation projects in the basin, with a total command area of 36,610 ha (Imbulana et al. 2009). The largest of these include Uda Walawe (23,700 ha), Ridiyagama (2,544 ha) and Liyangastota schemes (6,600 ha). These irrigation projects are complemented by several electricity generation schemes that consist of one large-scale hydropower station of 120 MW at Samanalawewa and several mini hydro plants (Imbulana et al. 2009).

Many environmental and social activists are concerned about the side effects of these large-scale development projects, and have pointed out their negative consequences “for the environment, for food security and most of all for poor and marginalised communities” (Unanthenna 2010; p.1). An example of these sentiments are the concerns expressed by Thilak Kariyawasam, Coordinator of Green Movement, regarding the Lunugamwehera project134, built with funding and technical expertise of the Asian Development Bank (Island newspaper 2006; p.1):

The people who planned the Lunugamwehera project have not considered the environmental destruction caused due to their erratic planning. Lagoons have been ruined, aquatic plants and aquatic living organisms have been exterminated and salinity of water bodies has increased considerably. Malnutrition among children is the highest in Lunugamwehera, education of children is adversely affected, family disputes and conflict situations are the few legacies that have been left behind by the ADB. Recent newspaper articles have focused on the negative impacts of projects such as the Mattala airport on locals (Kannangara 2012; p.1):

“Who benefits, from the development work of the Mahinda Chintanaya?” the villagers in Pahala Andara Wewa from the Hambantota District, and Kandhasurindugama, from Monaragala District, ask. Sadly enough, the people who brought the Rajapaksas to power have not benefitted, rather, the nearest and dearest of the First Family are enjoying all the fruits of State beneficence. The construction of the Mattala Airport Expressway has destroyed many of their homes and livelihoods. In order to make this possible, residents had to sacrifice their lands which included houses and paddy-fields and have received nothing in return, though six months have passed since the launch of the Expressway. Their only hope was Ms. Ganesha Amarasinghe, Divisional Secretary of Hambantota, who had promised to compensate

134 This was part of the irrigation and hydropower scheme in the Kirindi Oya basin.

150 them. However, she has now been transferred and a new Divisional Secretary has been appointed. The new Divisional Secretary, who is a government ally, has stated that no compensation will be paid as the villagers had ‘donated’ the land for a worthy cause… “All our attempts to obtain compensation for our lost livelihoods or our houses through the Grama Sevaka 135 , Road Development Board, Agriculture Development Board and the Divisional Secretary have fallen on deaf ears. Almost one season has gone, depriving us from our source of income which is paddy farming. How can we survive without an income?” queried the villagers. Since of late, herds of wild elephants have begun to attack the villagers and three persons have died as a result of these rampages. I conducted similar interviews with locals living in the landscape and will describe in detail some of the ill effects of development on their livelihoods in my stakeholder analysis in Chapter 6.

5.2 Land use within the study area

The study area consists mainly of agricultural areas, PAs and villages (i.e. homesteads). Table 5.5 shows the different types of land-use, while Figure 5 shows these on a map, which I developed using GIS data from the Survey Department of Sri Lanka.

Table 5.5: Land-use within my study area in southeast Sri Lanka (calculated from GIS data obtained from the Survey Department of Sri Lanka)

Land-use types Area (km2) % of total Paddy 491.14 8% Chena 542.88 8% Homestead/ gardens 725.93 11% Other cultivation, e.g. coconut, fruit trees 175.59 3% Forest 2,770.11 43% Scrub & grassland 1,339.77 21% Reservoirs and irrigation channels 213.96 3% Other water bodies (e.g. streams, water holes, lagoons) 93.52 1%

135 The government official assigned to the village.

151 Land-use types Area (km2) % of total Other natural (marsh, rocks, sand, salt pans) 105.33 2% Other human (e.g. built up areas, quarries, unclassified) 9.06 < 1% Total 6,467.29 100%

Figure 5.5: Map of landscape types within my study area in southeast Sri Lanka (map created using GIS data obtained from the Survey Department of Sri Lanka)

152 Table 5.5 and Figure 5.5 show that natural features such as forest, scrub, grasslands, rocks, sand and natural water bodies cover 67% of the landscape. The high proportion of natural features is largely because the area contains a number of PAs, but there is also considerable land covered in forest, scrub and grasslands outside PAs, mostly managed by the Forest Department. Agriculture is the other main type of land-use—paddy, chena, other crops and reservoirs and irrigation channels built to provide water to them, cover 22% of the land area. The remaining area (11% of the total) is devoted to human habitation, and many people cultivate fruit trees and vegetables in their home gardens. I now examine each of these land-use types in more detail.

Protected areas

The study area contains several important national parks and sanctuaries, including the famous Yala National Park. Although Yala was officially proclaimed as a national park in 1938136, it had been declared as a game sanctuary by the British colonial government as early as 1898, partly as a response to the rapid decline of wildlife in the country137 (Uragoda 1994; p.53). Several sections have been added to the park and some adjacent areas have been declared as sanctuaries, resulting in an interconnected PA complex138 covering over 1,700 km2 (Survey Department of Sri Lanka 2008; p.86). Yala is now perhaps Sri Lanka’s premier wildlife viewing area for locals and tourists alike, as evident from the description in the Lonely Planet Travel Book:

With trumpeting elephants, monkeys crashing through the trees, peacocks in their finest frocks and cunning leopards sliding like shadows through the undergrowth, Yala National Park (also known as Ruhunu) is The Jungle Book brought to glorious life. This vast region of dry woodland and open patches of grasslands is the big draw of this corner of Sri Lanka, and though it’s far from Kenya, a safari here is well worth all the time, effort and cost (Lonely Planet 2012; p.105). Uda Walawe National Park—another important PA within the landscape—is home to large herds of elephants; in fact it is one of the best places in Sri Lanka to view Asian

136 As per the Fauna and Flora Protection Ordinance of 1937, discussed in Chapter 2. 137 As discussed in Chapter 2, this was mainly due to the activities of professional game hunters. 138 The Yala Park Complex comprises Ruhunu National Park (Blocks I, II, III, IV, V), Yala Strict Nature Reserve, Yala East National Park (Block I, II), Katagamuwa Sanctuary, Kataragama Sanctuary and Kudumbigala Sanctuary.

153 elephants (de Silva and de Silva 2007; p.133). Its 308 km2 of habitat centres on the Uda Walawe reservoir, which covers over 10% of the park (de Silva and de Silva 2007; p.133). Sri Lanka’s southeast landscape also contains several important wetlands—of the five areas in Sri Lanka declared as ‘Wetlands of International Importance’ according to the Ramsar Convention on Wetlands, two are located within the landscape—Bundala National Park and Yala East National Park139 (Ramsar Convention 2012). The PAs within the chosen landscape (Table 5.6 and Figure 5.6) comprise almost 30% of Sri Lanka’s total PA network in terms of area.

Table 5.6: PAs within the study area (compiled using data from the Survey Department of Sri Lanka 2008; p.86, 88)

Year of Type of PA140 Name Extent (km2) declaration Ruhunu (Yala) 1938 979 Yala East (Kumana) 2006 357 National Park Lunugamwehera 1995 235 Bundala 1993 62 Uda Walawe 1972 308 Strict Nature Reserve Yala Strict Nature Reserve 1938 289 Bundala - Wilmanna 2006 33 Katagamuwa 1938 10 Kataragama 1938 8 Sanctuary Kudumbigala - Panama 2006 65 Nimalawa 1993 11 Wirawila-Tissa 1938 412 Total 2,769

139 Also known as Kumana. 140 National Parks can be visited by the public upon payment of a entry fee; Strict Nature Reserves are closed to the public and can only be visited by researchers with special permission from the Department of Wildlife; and Sanctuaries can be visited by the general public without a permit, but it is illegal to cause any harm to the wildlife that inhabits such a sanctuary (Holt 2011; p.268).

154

Figure 5.6: PAs within Sri Lanka’s southeast landscape (map created using GIS data obtained from the Survey Department of Sri Lanka)

These PAs are some of the most visited parks within the entire country; together, they account for 55% of total visitors and 65% of total revenue from Sri Lankan PAs (Department of Wildlife Conservation 2011). Details of visitor numbers and park revenues are provided in Appendix 5.

According to Table 5.7, even if PAs are excluded, about 50% of the landscape comprises a combination of forest, scrub and grassland. However, this number is almost certainly lower since the data are several years old and there has been much recent development activity in the area in the intervening period. Moreover, the region’s significant population increase has resulted in illegal encroachment onto government

155 lands. Nonetheless the landscape still has a significant area covered in forest and scrubland outside PAs, mostly managed by the Forest Department.

Table 5.7: Land-use within study area in southeast Sri Lanka, excluding PAs (calculated from GIS data obtained from Survey Department of Sri Lanka)

Land-use types Area (km2) % of total Paddy 472.12 11% Chena 499.58 12% Homestead/ gardens 711.61 17% Other cultivation, e.g. coconut, fruit trees 149.27 4% Forest 1,209.85 28% Scrub and grassland 997.44 23% Reservoirs and irrigation channels 111.65 3% Other water bodies (e.g. streams, water holes, lagoons) 61.33 1% Other natural (marsh, rocks, sand, salt pans) 39.47 1% Other human (e.g. built up areas, quarries, unclassified) 2.46 < 1% Total 4,255 100%

Agricultural areas

Agriculture is the most important source of employment within the districts of Moneragala, Hambantota and Ampara. Table 5.8 shows that the majority of the population in Moneragala are employed in agriculture, while the plurality of the populations in Hambantota and Ampara are employed in this sector, and the proportion of people employed in agriculture in each of the three districts is well above the national average.

156 Table 5.8: Employed population (as a percentage of working population) by major industry for Moneragala, Hambantota and Ampara districts, 2010 (Department of Census and Statistics 2011; p.37)

District Agriculture Industry Services Total Moneragala 55.0 11.6 33.4 100.0 Hambantota 43.6 23.3 33.2 100.0 Ampara 39.2 20.0 40.8 100.0 Overall for Sri Lanka 32.7 24.2 43.1 100.0

The main crop cultivated within the landscape is rice, which is grown during two seasons in Sri Lanka. The major season, known as Maha, is during the northeast monsoon that lasts from October to about mid April (de Silva and de Silva 2007; p.120). During this period, sufficient water for rice cultivation is usually available from rain as well as irrigation reservoirs, which often fill up rapidly. The minor season, or Yala season, lasts from May to September. The main source of rain during Yala is the southwest monsoon, which mostly brings rains to the wet zone in the southern and western parts of Sri Lanka. Therefore, cultivation during the Yala season in the dry zone usually depends mainly on irrigation water left over in reservoirs (de Silva and de Silva 2007; p.120). Figure 5.7 illustrates how the main stages of rice cultivation are coordinated with rainfall.

157 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sept

North east monsoon Inter South west monsoon (brings Rainfall (usually substantial in Inter monsoon monsoon little rain to the dry zone) the dry zone)

Rice cultivation

Land Irrigation & Harvesting Land Irrigation & Harvesting preparation growth preparation growth

Season Maha season Yala season

Source(s) Mainly rain water, if Need water from of water monsoon successful reservoirs

Figure 5.7: Timing of rainfall patterns and rice cultivation in Sri Lanka’s dry zone (adapted from de Silva and de Silva 2007; p.122)

Figure 5.7 illustrates how, during Maha, the substantial rainfall during the northeast monsoon is used during the growth phase of paddy, while the crop is harvested when it is relatively dry. Since the southwest monsoon brings little rain to the dry zone during Yala, farmers have to rely on water from reservoirs. During this time some rice fields are left uncultivated or planted with crops that are less water-intensive.

The study area is located mainly in the three districts of Hambantota, Ampara and Moneragala that are ranked number 1, 6 and 8, respectively, in terms of rice production, and together account for about a quarter of the country’s total production (Department of Census and Statistics 2011). Since the study area comprises significant portions of these three districts, it also includes prime rice producing areas. Figures 5.8 and 5.9 show paddy cultivation and harvest within the landscape.

158

Figure 5.8: Ripening paddy, Moneragala (photograph taken on 17/08/2010, by S. Kalpage)

Figure 5.9: Threshing paddy, Hambantota (photograph taken on 19/08/2010, by S. Kalpage)

159 In addition to rice, the three districts produce a significant portion of Sri Lanka’s other dry zone crops, as illustrated in Table 5.9.

Table 5.9: Select dry zone crops: proportion produced in Hambantota, Ampara and Moneragala for the 2009 season (compiled using data from Department of Census and Statistics 2011; Table 5.17)

Percentage of Sri Lanka’s total output for the 2009 Sinhala Crop growing season name Hambantota Ampara Moneragala Total (3 districts) Finger millet Kurakkan 10% 5% 15% 31% Maize Iringu 2% 9% 20% 31% Proso Millet Meneri 4% 0% 36% 39% Green Gram Mun Eta 14% 6% 20% 41% Cow Pea Cow Pea 4% 35% 16% 55% Ground nuts Rata Kadju 4% 7% 29% 40%

As per Table 5.9, the three districts produce over half of the country’s cowpea and over 40% of its green gram and groundnuts. Apart from the crops shown here, many other crops, such as chillies, manioc, and fruits such as watermelons are cultivated within the landscape. Many of these non-rice crops are grown in chenas141, an important form of traditional agriculture, which we now consider in detail.

Chena cultivation

Since many people living in the landscape do not have access to agricultural land, some practice a traditional form of agriculture called chena142. Some rice farmers also practice chena to supplement their income (de Silva and de Silva 2007; p.121). This form of cultivation, usually practiced in forested areas, is a major form of land-use in elephant habitat outside PAs (Campos-Arceiz et al. 2009; p.7).

In traditional chena, a patch of forest, usually about a hectare, is cleared and burnt before the onset of the northwest monsoon that prevails in the months of November to

141 The word chena refers to both the practice as well individual plots of land cultivated. So, farmers who practice traditional chena cultivation grow crops within their chenas. 142 This practice was described in Chapters 2 and 4.

160 February. Crops planted are those that can be harvested within 3 – 4 months, and include cereals such as maize and sorghum, pulses such as green gram and soy bean, spices such as chilli and mustard, vegetables such as tomato and onion, and fruits such as watermelon and banana (de Silva and de Silva 2007; p.121). After the harvest, which takes place around March, the land is left fallow for the next 5 – 6 months. Because no artificial fertilisers are used, the soil condition in the patch deteriorates over time and is abandoned after 3 – 4 years (Bandara 2007; p.7). Farmers return to formerly cultivated patches only after 10 – 12 years, by which time they are reforested (Wickramasuriya et al. 2009; p.2302). Figure 5.10 shows how the main stages of chena cultivation are coordinated with rainfall.

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sept

North east monsoon Inter South west monsoon (brings Rainfall (usually substantial in Inter monsoon monsoon little rain to the dry zone) the dry zone)

Chena cultivation

Land Crop growth Harvesting (may Chena preparation harvest some abandoned and planting shorter growth crops earlier, and more often)

Figure 5.10: Timing of rainfall patterns and chena cultivation in Sri Lanka’s dry zone (adapted from de Silva and de Silva 2007; p.122)

Present-day chena cultivation is slightly different from the traditional practice, as the increasing scarcity of forestland compels farmers to cultivate more or less permanent plots (Fernando et al. 2006; p.255). However, they still do so only during the rainy season, leaving the land fallow during the dry season comprising 5 – 6 months of the

161 year. They usually clear away only bushes, leaving taller trees intact, and generally use less fertiliser than other farmers (de Silva and de Silva 2007; p.217).

Most land on which chena is practiced is under the jurisdiction of the Forest Department. According to the law, chena farmers should obtain cultivation permits from the local government authorities. However, due to environmental concerns about the practice143 permits are rarely issued, especially within the study area (Bandara 2007; p.3). Consequently, much of the area’s chena cultivation is illegal. Despite lack of official sanction for the practice, however, social norms and customs dictate recognition of pre-existing family rights (Birner and Gunaweera 2001; p.7). Villagers generally respect the right to each other’s chenas; even if the chena land is abandoned, other villagers would not cultivate this land without the consent of the original cultivators (World Bank 2009; p.31).

Figure 5.11: Chena plot in the buffer zone of Yala National Park – note that many trees have been left intact (photograph taken on 25/04/2011, by S. Kalpage)

143 As pointed out in Chapter 2, there has been government opposition to chena since British times, but authorities have often turned a blind eye to the practice.

162 Other types of land-use

Human habitation is the third major type of land-use apart from PAs and agricultural areas (Table 5.5). While there are a few large towns in the area—such as the rapidly growing urban region surrounding Hambantota—many villages are scattered across the landscape. Figure 5.12 shows the area near the Yala National Park where there are several villages in the midst of rice fields (indicated in yellow). Most locals cultivate extensive home gardens (indicated in brown) where they grow fruit trees such as mango, coconut and papaya, and vegetables such as tomatoes, chillies and pumpkin. Produce from these home gardens is used for family consumption, while the excess is sold at village fairs or to local merchants.

163

Figure 5.12: The area around Yodakandiya town, located close to the Yala National Park (map created using GIS data obtained from the Survey Department of Sri Lanka)

164 5.3 Livelihoods of people in the study area

I conducted semi-structured interviews with locals living in the landscape to better understand their livelihoods. The majority of my interviews were with farmers—both rice and chena—since farming is the main occupation in the area, and because these people are among the most affected by human-elephant conflict. I also interviewed other locals who could potentially benefit from nature tourism, such as safari jeep drivers and small-business owners. In addition, I had access to data from a survey of 800 farmers living within the landscape, conducted by the World Bank in 2009144. Further details of my semi-structured interviews and the World Bank survey are presented in Appendix 5145. I present my analysis of data from these sources in the rest of this section, as well as in my stakeholder analysis in Chapter 6.

I now describe the livelihoods of farmers and other groups living within the landscape in light of this primary data. Where appropriate, I supplement my analysis with information from government agencies and press and journal articles.

Farmers

I estimated the annual income of people surveyed by the World Bank by analysing the relevant survey data. I found that farmers living in the landscape—especially in areas that comprise elephant habitat—are extremely poor. My results for two main groups of famers—those who cultivate chenas and other farmers—are presented in Tables 5.10 and 5.11. Rice is the main crop cultivated by the latter group, although they often also cultivate other crops when there is not enough water for rice (during the Yala season for example), or in their home gardens.

144 The surveys were conducted by the Institute of Policy Studies, a local research institute. 145 This includes dates and places the interviews were conducted; however, as per UNSW ethics requirements, interviewee names have been withheld.

165 Table 5.10: Annual income estimate for chena farmer households (based on analysis of data from the World Bank Survey)

Income Profits Source of income Costs (Rs.) % of total (Rs.) (Rs.) Agricultural income Chena 52,534 23,774 28,760 30% Paddy 13,084 5,760 7,324 8% Seasonal crops 924 305 620 1% Permanent crops 6,362 1,871 4,491 5% Livestock 423 0% Agricultural labour income 11,440 - 11,440 12%

Non-agricultural income Fisheries 2,644 1,903 741 1% Microenterprises 21,301 22% Formal employment 4,078 4,078 4% (government, private, NGO) Pension, disability, food stamps 261 261 0% Other cash income 399 399 0% Non-agricultural labour income 16,043 16,043 17% Total 108,191 33,613 95,456 100%

Table 5.11: Annual income estimate of households engaged in paddy farming (based on analysis of data from the World Bank Survey)

Income Profits Source of income Costs (Rs.) % of total (Rs.) (Rs.) Agricultural income Paddy 45,671 8,221 37,450 30% Seasonal crops 4,104 872 3,232 3% Permanent crops 6,483 1,611 4,872 4% Livestock 2,882 - 2,882 2% Agricultural labour income 7,924 - 7,924 6%

Non-agricultural income Fisheries 7,723 2,893 4,829 4%

166 Income Profits Source of income Costs (Rs.) % of total (Rs.) (Rs.) Microenterprises 42,061 - 42,061 33% Formal employment 4,082 - 4,082 3% (government, private, NGO) Pension, disability, food stamps 172 - 172 0% Other cash income 332 - 332 0% Non-agricultural labour income 17,892 - 17,892 14% Total 139,324 13,597 125,727 100%

As expected, both groups derive a major part of their income from agriculture. In addition, farmers (and other household members) appear to supplement their income working as labourers, in agriculture as well as in other enterprises such as construction. This is probably because chena cultivation is seasonal, and many paddy farmers cultivate rice for only part of the year due to insufficient water. Farming households also engage in microenterprises to supplement their income. According to the survey, the most popular activities were selling items at local fairs or even in front of homes, and small-scale manufacturing of items such as garments and mats. Some non-chena farmers engaged in fishing, mainly in inland water bodies.

Household income levels for both groups of famers are extremely low. The annual income of a typical chena farmer household is the equivalent of only about US$ 840, while the corresponding number for other farmer households—US$ 1,100—is only slightly better. These numbers are well below the government-declared poverty lines for the Hambantota District, where most of the interviews were conducted146 (Table 5.12).

146 Moreover, they are only a fraction of the mean annual household income for Sri Lanka of Rs. 437,412 (Campos-Arceiz et al. 2009; p.12).

167 Table 5.12: Incomes of farmers surveyed compared to official poverty lines (data from World Bank 2009; Department of Census and Statistics 2011)

Chena farmers Other farmers Annualised poverty line per person for 38,400 38,400 Hambantota district (Rs.) Average size of household (in survey) 4.8 3.9 Implied poverty line for household 184,320 149,760 (Rs.) Total household income of farmers 95,456 125,727 surveyed (Rs.)

It is important to bear in mind that figures for total household income (in Table 5.12) calculated from survey responses are probably underestimates. Firstly, people are likely to underestimate their income, either because they have forgotten some income sources and/or amounts, or because they want to qualify for government or other development programmes147. Other researchers, such as de Silva and de Silva (2007; p.212), who have conducted surveys among similar groups in Sri Lanka, found this to be the case. Secondly, farmers usually set aside some of their crops for household consumption, thus reducing their expenditure on food (resulting in some savings compared to urban households, for example)148. Despite the probable underestimation of income, the reported income of both groups of farmers is significantly below the poverty line; therefore it is likely that these farmers are still among the poorest Sri Lankans.

In additional to information about income of farmers, the survey revealed some insights into the specific nature of chena cultivation in the landscape. The first was on the timing of chena cultivation, presented in Figure 5.13.

147 Some respondents may have done this if they found out that a large international NGO such as the World Bank commissioned the survey. 148 According a recent household expenditure survey, Sri Lankans spend about 42% of their incomes on average, on food and drink; this is likely to be even higher for groups close to the poverty line, such these farmers (Department of Census and Statistics 2011; p.7).

168 100% % of farmers 90%

80%

70%

60%

50%

40%

30%

20%

10%

Figure 5.13: Percentage of farmers engaged in chena by month (n=252) (based on analysis of data from the World Bank Survey)

Figure 5.13 shows that the cultivation patterns of chena farmers in the landscape are similar to the expected general pattern illustrated in Figure 10. The few farmers who cultivate crops from March to July probably utilise small water bodies near their fields for irrigation.

169 The survey also provided insights on chena farmers’ fertiliser and pesticide use (Tables 5.13 and 5.14).

Table 5.13: Chemical fertiliser use by chena farmers in sample (n=252) (based on analysis of data from the World Bank Survey)

Interval: Fertiliser use (kg) % of farmers in interval None 76% <= 50 10% <= 100 7% <= 200 4% <= 300 2% > 300 1% Total 100%

Table 5.14: Pesticide use by chena farmers in sample (n=252) (based on analysis of data from the World Bank Survey)

Interval: Pesticide use (litres) % of farmers in interval None 73% <= 500 16% <= 1,000 3% <= 2,000 2% <= 5,000 4% <= 10,000 2% > 10,000 1% Total 100%

Tables 5.13 and 5.14 indicate that about three quarters of the chena farmers surveyed did not use chemical fertilisers or pesticides. Those using small amounts accounted for 10% and 16% respectively, bringing the total of those using little or no chemical inputs to almost 90%. This result, again, is as expected, although some researchers such as de Silva and de Silva (2007; p.217) found that the use of chemical fertilisers and pesticides is increasing among chena farmers. Moreover, some of the farmers I interviewed told me that they use pesticides and weedicides. An example was Farmer 7:

170 I think I must spend about a third of income on chemicals such as insecticides, weedicides and fertilizers! It is hard to grow crops without using these chemicals, because if I did that, my crops would be at risk (interview, Farmer 7, 06/04/11). The increased use of chemical inputs may be in part due to the growing use of foreign seeds—which are less adapted to local conditions compared to local varieties—a point some farmers talked about:

Many people cultivate new varieties of crops. For example, the popular variety of watermelon now grown in these parts is called ‘Malaysian’. As these cannot be grown using local seeds we have to buy them. Although this adds to our expenses, local varieties give us lower harvests and there is less demand for these crops. For example the hybrid version of Watakolu (Ribbed Gourd) is huge, unlike the traditional variety people used to cultivate around here. These new types of crops require more chemical inputs than traditional varieties (interview, Farmer 44, 06/09/11). However, de Silva and de Silva (2007; p.217) found that organic farming practices are re-emerging among farmers. This was confirmed by some of my interviewees, such as Farmer 45, who told me that he now predominantly cultivates traditional varieties of crops using natural fertilisers and traditional crop protection methods (interview, Farmer 45, 06/09/11).

Many farmers have been cultivating their chenas for long periods of time, as illustrated in Table 5.15.

Table 5.15: Number of years cultivated by chena farmers in sample (n=252) (based on analysis of data from the World Bank Survey)

Interval: Number of years cultivated % of farmers in the interval <= 10 35% <= 20 37% <= 30 18% > 30 9% Total 100%

As indicated in Table 5.15, about two thirds of the farmers had cultivated their chenas for over 10 years; the average is 17 years. The data also indicated that the average chena plot size is about two acres. Finally, the survey revealed information on types of crops

171 cultivated in chenas—the most popular vegetables were long beans, chillies, bitter gourd, green gram, okra, cowpea and tomatoes.

Figure 5.14: Harvesting tomatoes in a chena near the Yala National Park (photograph taken on 22/03/2010, by S. Kalpage)

More results from my analysis, especially those regarding threats to farmers’ livelihoods, are presented in the stakeholder analysis in Chapter 6.

Other livelihoods within the landscape

While farming is the main occupation within the landscape, there are several other types of livelihoods in the area. Some people fish in inland water bodies such as rivers and reservoirs, and in the sea, especially in coastal areas such as the harbour-town of Kirinda. Others work in small enterprises that are self-owned or run by local entrepreneurs. Many people, however, work at least part of the year as labourers in agriculture or in non-agricultural enterprises such as construction.

172 The World Bank survey revealed the various types of activities farmers and their families engaged in to increase household income. For example, chena farmers often worked as labourers during months when their chenas were abandoned. Other household members found different ways to generate income. Farmers’ wives, for instance, supplemented family income through home-based microenterprises such as sewing garments for sale. Even children earned money after school hours or during holidays: by selling produce from their parents’ fields or home gardens from roadside stands. Table 5.16 provides further information on the types of small and microenterprises people worked in; the data showed that almost half of household members engaged in some type of retail sales, or home-based manufacturing.

Table 5.16: Types of small and micro-enterprises interviewees engaged in (based on analysis of data from the World Bank Survey)

% of respondents engaged in activity Category (among those who answered the question) Selling (e.g. produce, handicrafts) 43% Handicrafts/ home manufacturing 27% Manufacturing services 19% Weaving/ garments 7% Other 4% Total 100%

I also analysed the income of households that did not engage in chena or paddy farming. The results are presented in Table 5.17:

Table 5.17: Annual income estimate for households that did not engage in chena or rice cultivation (based on analysis of data from the World Bank Survey)

Income Profits Source of income Costs (Rs.) % of total (Rs.) (Rs.) Agricultural income Chena 365 - 365 0% Paddy 2,788 1,525 1,264 1% Seasonal crops 4,158 1,790 2,367 2% Permanent crops 1,939 - 1,939 2% Livestock 6,744 - 6,744 7%

173 Income Profits Source of income Costs (Rs.) % of total (Rs.) (Rs.) Agricultural labour income 365 - 365 0%

Non-agricultural income Fisheries 77,186 48,737 28,448 28% Microenterprises 34,337 - 34,337 34% Formal employment 4,961 - 4,961 5% (government, private, NGO) Pension, disability, food stamps 479 - 479 0% Other cash income 1,140 - 1,140 1% Non-agricultural labour income 21,028 - 21,028 20% Total 155,123 52,052 103,071 100%

Table 5.17 shows the importance of micro-enterprises, which accounted for about a third of income. The fisheries industry was important, because some of the surveyed families earned a living from fishing, mainly in the ocean, and to a lesser extent in some of the area’s inland water bodies. A few people were engaged in formal employment, in the government, private or non-governmental sectors. Only 5% of income was derived from crops—probably from home gardens—while 7% was derived from livestock, mainly cattle, buffaloes and chickens (World Bank Survey).

174

Figure 5.15: Livestock herding, Hambantota district (photograph taken on 21/03/2010, by S. Kalpage)

Figure 5.16: Fishing boats, Ampara district (photograph taken on 17/08/2010, by S. Kalpage)

175 Having examined some of the main types of livelihoods of people living in the landscape, I now consider one of the major threats to many of those livelihoods—the area’s growing human-elephant conflict.

5.4 The elephant population and human-elephant conflict within the study area

The elephant population within the study area

Southeast Sri Lanka contains a significant proportion of the country’s elephants. The 2011 census of Sri Lanka’s wild elephant population conducted by the Department of Wildlife showed that the southern and eastern regions contained about 45% of the country’s elephants149. According to the survey, an area roughly corresponding to five districts—Galle, Matara, Hambantota, Ampara and Moneragala—contain 2,659 elephants, implying that 45% of the country’s elephants are contained within an area comprising about 25% of Sri Lanka’s land area (Department of Census and Statistics 2011; p.1).

A significant portion of these elephants live within my study area, since, as pointed out before, the landscape contains important elephant habitat (namely PAs such as the Yala Park Complex, Uda Walawe and Lunugamvehera, and surrounding forested areas). Although it is not possible to obtain the exact figures using the data released from the Department of Wildlife, we can arrive at an estimate by combining this information with previous figures published by de Silva and de Silva (2007), who estimated that about 30% of Sri Lanka’s elephants live in the chosen landscape. Combining this percentage with numbers from the 2011 elephant census, we can estimate that about 1,700 elephants live in the case study area. Since this proportion of elephants is contained within about 10% of the country’s land area, the elephant-density is high, as could be expected from the high quality of elephant habitat found within the landscape150.

149 Details of the census were presented in Appendix 3. 150 My detailed calculation is presented in Appendix 5.

176 Human-elephant conflict within the study area

Historically human-elephant conflict levels in southeast Sri Lanka have been much lower than those in the northwest and northeast (de Silva and de Silva 2007; p.210). Fernando et al. (2005; p.2467) attribute this to the area’s well-established complex of PAs and adjacent forested areas that provide elephants with adequate food and water resources. Moreover, local people have lived alongside wildlife for generations and adopt traditional practices such as chena, in forested ‘buffer zones’ between elephant habitat and villages, enabling both groups to share resources151.

However, rapid development in the region—described in section 5.1 above—has caused the conflict to escalate, as revealed in a detailed study conducted by Campos-Arceiz et al. (2009) in conflict areas in southeast Sri Lanka. Since the area studied by the researchers corresponds closely to my case study area, I consider their results in more detail. The authors collected data from six villages located close to PAs within the landscape. Over a span of one year—from June 2004 to May 2005—they collected details of 973 human-elephant conflict incidents (Campos-Arceiz et al. 2009; p.8). Incidents were defined as “any incident involving crop damage, house damage, and human threat or injury caused by a single elephant or group of them” (Campos-Arceiz et al. 2009; p.7). Crop damage was the most common form of conflict, occurring in 92% of cases. While elephants damaged more than 30 types of plants, they most preferred banana, rice and young coconut plants. Houses were damaged in about 11% of incidents (Campos-Arceiz et al. 2009; p.8). Of these about a fifth were completely destroyed, while another half sustained serious damages (Campos-Arceiz et al. 2009; p.8). The main motivation of elephants in these incidents was to get at food stored within the houses, mainly paddy.

The vast majority of incidents took place at night. Crop-raiding elephants came in small groups, and were typically males; over 50% of incidents were caused by single bull elephants (Campos-Arceiz et al. 2009; p.8). Based on this, and other similar studies, the authors conclude that the animals recognise that crop raiding is risky; consequently bulls seem more willing to take risks than herds comprised of cows and calves (Campos-Arceiz et al. 2009; p.10):

151 More details of this human-elephant co-existence were provided in Chapter 4.

177 These results indicate different behavioural strategies between adult males and female-led herds, with males more ready to assume the risks of crop raiding than females, demonstrating “high risk-high gain” behaviour... The fact that most conflict incidents happened at night also supports this, as elephants are normally active during daytime. Exclusive nocturnal raiding behaviour therefore indicates that elephants do recognize the risks associated with their activities. Sexual differences were even more apparent in attacks on houses and people, which involved higher risk than crop raiding. However, based on this and other studies, the authors conclude that as resources for elephants get even more scarce as their habitats are gradually destroyed, herds may also start to raid crops, resulting in a further escalation of the conflict (Campos-Arceiz et al. 2009; p.12):

Since female herds managed to obtain their resources without frequently feeding on crops, it can be said that crops are not a necessary component of the diet of elephants in the region. It has been suggested that females, initially less prone to assume the risks of conflicts with farmers, might become obligate crop raiders when living in highly fragmented habitats, where resources are more limited and scattered in the human-dominated landscape This means that the level of conflict in southeast Sri Lanka is likely to increase in coming years, as further developments and loss of habitats are on going. If such threshold exists, in which elephants become obligate rather than ‘optional’ raiders, conservation of elephants would become much harder once it is crossed, and therefore habitat management should be oriented to avoid these levels of habitat fragmentation and impoverishment. As the authors conclude, immediate steps must be taken to slow the destruction of elephant habitat to mitigate the conflict. Implementing MERs as discussed in Chapter 4 would be an effective way of accomplishing this. In the next section I discuss how this concept could be applied within my selected study area.

178 5.5 Potential Managed Elephant Ranges (MERs) within the study area

Given the availability of forested areas, scrubland and grasslands within the landscape, even outside PAs, several areas could be demarcated as MERs. Figure 5.17 shows a potential MER near Yala National Park. While most of the area comprises natural vegetation (shown in various shades in green), it also includes a few paddy fields (shown in yellow) and some home gardens (shown in brown). These areas should be protected by permanent, well-constructed electric fences. Moreover, temporary fences should be constructed around areas under chena cultivation (shown in light green with spots); these should be taken down after crops are harvested in March/April.

Figure 5.18, shows a potential MER near Uda Walawe and Lunugamvehera National Parks, which has been demarcated according to the same criteria used in Figure 5.17 (it should be managed in a similar manner).

Yala National Park

Potential Managed Elephant Range bordering park

Figure 5.17: Potential MER bordering the Yala National Park (map created using data obtained from the Survey Department of Sri Lanka)

179 Uda Walawe National Park

Lunugamvehera National Park

Potential Managed Elephant Range bordering parks

Figure 5.18: Potential MER bordering the Uda Walawe and Lunugamvehera National Parks (map created using data obtained from the Survey Department of Sri Lanka)

In this chapter, I explored some of the key characteristics of the chosen landscape—its socio-economic situation, history, land-use and local livelihoods. I also described the growing human-elephant conflict—one of the main threats to local livelihoods—and suggested potential areas where MERs could be established to mitigate the conflict. In the next chapter, I identify key stakeholders in the conflict, especially those who are affected and those who could be influential in implementing MERs.

180 Chapter 6: Stakeholder analysis

Having introduced my study area and identified potential MERs within the landscape, I now consider the various groups interested in mitigating the area’s human-elephant conflict, and therefore could influence MER establishment. In this chapter, I discuss my rationale for choosing stakeholder analysis to assess stakeholder needs, and introduce related theory and methods. I then use these tools to identify key stakeholders, analyse stakeholder needs, and draw lessons regarding MER implementation.

6.1 Selecting a method to assess and meet stakeholder needs

As noted in section 1.4, the landscape approach advocates ‘negotiated’ landscapes instead of ‘planned’ landscapes. Consequently, it is important to understand the needs of key stakeholders—especially of groups who may be marginalised—and ensure that their needs are met. The types of approaches that could be used for this purpose vary according to the degree of participation involved by the researcher (Neuman 2007; p.281). For example, some authors have advocated participatory action research— which closely involves stakeholders in the entire decision-making process from research design to the interpretation of results—as a way to ensure participation of all groups, especially regarding issues related to natural resource management (Castellanet and Jordan 2002; Mcintyre 2008; Mackenzie et al. 2012).

Unfortunately, participatory action research is extremely time and resource intensive and therefore difficult to implement, especially in developing countries, where government agencies are often very slow moving, lack the required resources, and take time to get comfortable with new ways of doing things (Mackenzie et al. 2012). Moreover, this type of approach does not fit the timelines advocated in most research projects, and is certainly beyond the scope of a PhD-relate project; for example, Castellanet and Jordan (2002; p.199) who evaluated a program in the Brazilian Amazon, found that “in actually solving management-resource problems, the project was only partially successful within the first five years”. Given that a participatory action research approach proved impractical, I decided to use stakeholder analysis to understand the needs of the key stakeholder groups in my research area. Although less collaborative than participatory action research, this process nonetheless yielded

181 important insights that helped inform MER design. Moreover, I have recommended a process that provides all major stakeholder groups with opportunities to participate in the planning and management of specific MERs; the suggested institutional mechanism for doing so is detailed at the end of Chapter 8 (section 8.3).

Having discussed my rationale for choosing stakeholder analysis, I now examine the related theory and methods in further detail.

6.2 Stakeholder analysis theory and methods

Policymakers are increasingly using stakeholder analysis in their decision-making processes, especially in instances where a range of individuals or organisations have varying interests in the outcome of a situation (Romanelli et al. 2011; p.463). These methods are often used to identify and resolve conflicts of interest in natural resource management, as these situations usually involve different groups who have a range of interests in the utilisation and/or conservation of natural resources (Grimble 1998; p.1).

Before discussing the methodology in detail, it is useful to clarify the use of the terms ‘stakeholder’ and ‘stakeholder analysis’. According to Grimble and Wellard (1997; p.175) stakeholders are “any group of people, organised or unorganised, who share a common interest or stake in a particular issue or system; they can be at any level or position in society, from global, national and regional concerns down to the level of household or intra-household, and be groups of any size or aggregation”. Consequently, stakeholder analysis is “a range of tools or an approach for understanding a system by identifying the key actors or stakeholders on the basis of their attributes, interrelationships and assessing their respective interests related to the system, issue or resource” (Mushove and Vogel 2005; p.185).

Stakeholder analysis is widely used in several fields. Having originated in management studies and business administration, it is now used in political science, development studies and environmental studies (Billgrena and Holmen 2008; p.552). Consequently, the scope of the approach can vary significantly depending on the context (Varvasovszky and Brugha 2000; p.338). Indeed, many authors believe that stakeholder analysis comprises a range of theories and methods (Billgrena and Holmen 2008; p.552).

182 Reed et al. (2009) describe how the approach is used for different purposes in business management and natural resource management:

In business management, the growing realisation that stakeholders could affect the success of a firm led naturally to the development of approaches to analyse stakeholders in order to understand their interests and influence, and how these could support or threaten the performance of the firm. As such, the business management community primarily used stakeholder analysis to mobilise, neutralise or defeat stakeholders, to meet the strategic objectives of firms. However within policy, development, and natural resource management, stakeholder analysis was increasingly seen as an approach that could empower marginal stakeholders to influence decision- making processes. Although this broadened the role of stakeholder analysis, enriching its theoretical basis and analytical methods, it also increased the complexity and difficulty of such research, since many additional conflicting and diverse agendas had now to be considered (Reed et al. 2009; p.1934). Given the specific situation in my case study—where I am considering how a landscape-based approach could mitigate human-elephant conflict—my focus is much closer to the natural resource management approach described above. My approach is similar to that adopted by de Lopez (2001) who developed a stakeholder management framework for Ream National Park, Cambodia, using stakeholder analysis to “understand and predict the behaviour and actions of stakeholders and devise strategies to ethically and effectively deal with them” (p.48). The questions I ask are similar:

a) Who are the stakeholders in the human-elephant conflict within Sri Lanka’s southeast landscape?

b) What is the nature of their interest in the issue?

c) What strategies should be adopted to achieve MER goals while meeting stakeholder needs and concerns?

I will revisit these questions at the end of this chapter, examining how my stakeholder analysis helps address them. I now consider the various available stakeholder methods, and select those best suited for my case study.

183 Stakeholder analysis methods

As discussed previously, a wide range of stakeholder analysis methods could be used, depending on the situation at hand. Reed et al. (2009; p.1936) categorise these methods in the context of three major steps: (i) identifying stakeholders, (ii) differentiating among and categorising stakeholders, and (iii) investigating relationships among stakeholders (Table 6.1).

Table 6.1: Methods that could be used to conduct each step of the stakeholder analysis152 (adapted from Reed et al. 2009; p.1937)

Analysis step Stakeholder method Description of method

Step 1: Identifying Focus groups A small group brainstorm stakeholders, their stakeholders interests, influence and other attributes

Semi-structured Interviews with a cross-section of stakeholders interviews (sometimes used to check/ supplement focus group data)

Snow-ball sampling Individuals from initial stakeholder categories are interviewed to identify new stakeholder categories

Step 2: Interest-influence Stakeholders are placed on a matrix according Differentiating matrices to their relative interest and influence among and Stakeholder-led Stakeholders themselves categorise categorising stakeholder stakeholders into categories which they have stakeholders categorisation created

Q methodology Stakeholders sort statements drawn from a concourse according to how much they agree with them, analysis allows social discourses to be identified

Radical transactiveness Snow-ball sampling to identify fringe stakeholders; development of strategies to address their concerns

152 In certain situations, a combination of methods could be used.

184 Analysis step Stakeholder method Description of method

Step 3: Actor-linkage matrices Stakeholders are tabulated in a two- Investigating dimensional matrix and their relationships relationships described among Social Network Analysis Used to identify the network of stakeholders stakeholders and measure relational ties between stakeholders through use of structured interviews/ questionnaires

Knowledge mapping Used in conjunction with Social Network Analysis; involves semi-structured interviews to identify interactions and knowledge

Having provided an overview of the aims, processes and methods associated with stakeholder analyses, I now select specific methods for my case study. I do this in the context of each of the three steps described in Table 6.1.

Methods used in case study

I used semi-structured interviews to identify stakeholders (i.e. step 1). One of the first groups I interviewed were conservationists involved in developing the MER approach, as they had done extensive fieldwork within my selected study area. I verified and added to their views through semi-structured interviews I subsequently conducted with other groups. Therefore, there was an element of snowball sampling in my approach to identifying stakeholders. Some of my initial interviews and discussions took place with small groups of conservationists, so there was an additional element of focus group methodology.

The semi-structured interviews I conducted provided valuable input to step 2, i.e. differentiating among and categorising stakeholders. This is consistent with Reed et al. (2009; p.1936) who agree that some of the methods may be used for more than one purpose. At the end of this step, I used an interest influence matrix to categorise the stakeholders.

185 In step 3, I used an actor linkage matrix to investigate relationships among stakeholders. Figure 6.1 presents a summary of the methods I used, according to the typology presented by Reed et al. (2009; p.1936); the method (or methods) used for each step are highlighted in yellow.

STEP 1 STEP 2 STEP 3

Analysis Identifying Differentiating among Investigating steps stakeholders and categorising relationships among stakeholders stakeholders

Analytical categorisation Actor linkage Focus groups (top down) matrices • Interest influence matrices • Radical transactiveness

Methods Semi structured Social network interviews analysis Reconstructive categorisation (bottom up) • Stakeholder-led stakeholder categorisation Snowball Knowledge • Q methodology sampling mapping

Figure 6.1: A schematic representation of stakeholder analysis steps, and associated methods; the methods used in the case study are highlighted in yellow (adapted from Reed et al. 2009; p.1936)

Having explained stakeholder analysis theory, and specified the methods used in my case study, I now describe my analysis, presenting my results for each step indicated in Figure 6.1, in section 6.2 (step 1), section 6.3 (step 2) and section 6.4 (step 3).

186 6.3 Identifying stakeholders within Sri Lanka’s southeast landscape and understanding their role in human-elephant conflict

As detailed above, I used semi-structured interviews to identify stakeholders within Sri Lanka’s southeast landscape. I interviewed 69 farmers—the main group affected by human-elephant conflict within this landscape. In addition, I obtained data from a survey of 800 farmers in those areas, conducted by the World Bank in 2009. Further details of my semi-structured interviews and my analysis of the World Bank survey data are presented in Appendix 5153. I also interviewed government officials—notably from the Department of Wildlife and the Forest Department—conservationists and NGOs whose work relates to the human-elephant conflict, tourism industry representatives, and other locals involved in or who benefit from nature tourism. I supplemented these interviews and surveys with information from press articles, government reports and journal articles to build up a comprehensive base of information—to the extent permitted by available time and resources—of the needs, issues and views of the various stakeholder groups. In Table 6.2, I describe my information sources for each of the identified stakeholder groups.

Table 6.2: Sources of information by stakeholder group

Stakeholder Sources of information group

Farmers • Semi-structured interviews with 69 farmers; these included both chena farmers growing crops on land under the jurisdiction of the Forest Department, mainly in PA buffer zones, as well as other farmers cultivating crops in areas outside parks but within elephant ranges

• World Bank survey of 800 farmers in villages experiencing significant human-elephant conflict

• Journal articles and press articles on human-elephant conflict

Government • Semi-structured interviews with officials from the Department of officials Wildlife and the Forest Department (10 officials)

• Statements to the press by officials from the two departments

153 This includes dates and places the interviews were conducted; however, as per UNSW ethics requirements, interviewee names have been withheld.

187 Stakeholder Sources of information group

• Government reports and press releases, including those by Ministries and Departments responsible for development and tourism

• Government websites

Tourism • Semi-structured interviews with 8 individuals involved in the tourism companies industry

• Press articles

Other locals • Semi-structured interviews with 13 locals who benefit from nature tourism, i.e. jeep drivers and local businesses

Sri Lankan • Published research on attitudes of rural and urban Sri Lankans towards public human-elephant conflict

• Press articles

Non- • Semi-structured interviews with 10 people working for conservation governmental and/or rural development-oriented NGOs organisations • Reports and press articles by these NGOs

• Websites of these NGOs

Having identified the main stakeholder groups, I now summarise my findings from interviews and other information sources154.

Farmers

As described in Chapter 5, there are two main groups of farmers in the landscape: those who cultivate rice and those who cultivate other crops such as maize, chillies and green gram according to traditional chena cultivation methods. Of the 69 farmers I interviewed, 35 cultivated chenas, 32 rice, and 2 fruits, mainly bananas155. The semi- structured interviews I conducted with each group revealed the major issues they faced,

154 Some findings also inform step 2 (differentiating among and categorising stakeholders) and step 3 (investigating relationships among stakeholders) as indicated in sections 6.3 and 6.4 respectively. 155 These numbers describe their primary cultivation types, as some cultivated both chena and rice.

188 of which human-elephant conflict was among the most significant. Farmers’ livelihoods were also threatened by natural disasters such as floods and droughts, for which climate change seemed at least partially responsible. A significant non-environmental issue chena farmers spoke about was their insecurity regarding land tenure. I now describe these issues in more detail.

Issues with elephants

All farmers I interviewed said that they had problems with elephants, but this was to be expected as my interviews were conducted in areas with considerable human-elephant conflict. Chena farmers cultivating crops in the buffer zone of Yala National Park said that during the growing season elephants came into the area almost daily, usually arriving at night in small, all-male groups. Elephants often tried to break through the fences that they had constructed around their plots of land. For example, Farmer 7156 told me:

We have a really difficult time protecting our chenas from elephants, who come every night. Recently, a group of 3 – 4 elephants have started visiting our area. During the last few nights they have broken 6 concrete posts in the electric fence (constructed by the Department of Wildlife) and have eaten crops in nearby chenas… During the day we work hard on our crops; at night, we have to stay awake to chase away elephants (interview, Farmer 7, 06/04/11).

156 Details on the interviews conducted—such as the date and interview location—are presented in Appendix 5.

189

Figure 6.2: Elephants in paddy field, after paddy harvest (photograph taken on 24/04/2011, by S. Kalpage)

Figure 6.3: Gate to a chena damaged by elephants (photograph taken on 06/04/2011, by S. Kalpage)

190 Elephants sometimes even break into farmers’ houses to get at paddy or other food stored within. Sometimes this happens even when people are inside, as illustrated by a recent incident Farmer 29 described as he showed me his damaged house:

An elephant broke into my house on November 18th, 2010. I was in hospital with a foot injury, my wife was visiting her family, but four relatives were staying at our house. A huge male elephant broke into the front room where paddy was stored. Luckily our relatives escaped through the back door. Elephants usually push on the walls with their heads making huge holes in the brickwork—see what happened! We are very lucky, because if this had been a wattle and daub house, the entire wall would have collapsed, and our relatives would have been killed. Our house is now structurally unsound; we will have to tear it down and re-build it (interview, Farmer 29, 04/09/11).

Figure 6.4: The house of Farmer 29, which was damaged by an elephant (photograph taken on 06/04/2011, by S. Kalpage)

Elephants often attack people who are trying to protect their crops and homes. Farmer 28 described some incidents in her home garden:

Elephants in our area are very aggressive. We try to chase them away using ali wedi (crackers), but are not always successful. My husband usually makes a fire in front of our house, staying up all night. Once, when my daughter and I were here alone, an elephant came right up to our front door. It came after us when we tried to drive it way, and chased us around the house. We were lucky to escape with our lives. On another occasion,

191 an elephant came after my father, who fell down trying to escape; he was lucky to survive (interview, Farmer 28, 04/09/11). Several farmers told me about family members who had been killed by elephants. For example, Farmer 4’s brother-in-law had been killed in 2006, and Farmer 26 told me about his father who died in 2001: “It was in the early hours of the morning; the elephant was probably trying to get back to the forest and he may have startled it” (interviews with Farmers 4 and 26, on 06/04/11 and 04/09/11).

Analysis of the World Bank survey data—presented graphically in Figure 6.5—shows the pattern of elephant attacks throughout the year.

All Chena Non chena

117

103

83 77 71 66 62 64 56 57 45 41 37 33 33 35 34 29 32 28 25 26 22 20 15 15 16 15 18 11 12 11 9 10 6 4

Figure 6.5: Elephant attacks by month in southeast Sri Lanka (analysis based on World Bank survey data)

Attacks are most intense during the rainy season when both rice and chena crops are being cultivated. This pattern is different from those found in studies conducted by Fernando et al. (2005) about 7 years ago when there was much less development in the area—as described in Chapter 4 they found that elephants utilised chena mainly in the

192 dry season. However, the area’s rapid development, and the numerous, haphazardly constructed electric fences found throughout the landscape has reduced elephants’ access to food and water even during the rainy season, making them more desperate. Some of the farmers I interviewed admitted that they had noticed that elephant attacks intensified when the animals were deprived of habitat in nearby areas:

The elephants have lost much habitat in recent years due to all the development in the area. Many roads have been constructed; large areas have been enclosed by electric fences for banana cultivation, irrigated by the Walawe left-bank scheme; and large extents of jungle have been cleared for various government schemes. So, naturally, elephants come to neighbouring fields and chenas looking for food (interview, Farmer 9, 07/04/11). Elephants have become a huge problem for us—during the last growing season, we had to harvest our paddy before it was fully ripe. This is mainly because short-sighted government schemes have reduce their habitat. The Walawe (irrigation) scheme has been established in prime elephant country, and a thousand acres—in an area where we used to see many elephants—was taken for Mattala airport (interview, Farmer 18, 25/04/11). When we first came into the area (i.e. in the 1960s and 1970s) we had no issues with elephants, and even used to store our harvest in our fields. People were also much more considerate of elephants’ needs. For example, some people avoided cultivating their fields in certain areas, at certain times of the year, to leave food for elephants’. However, now, with the rapidly growing population and so-called development, people are encroaching on elephants’ territory. For example, lots of ‘tsunami villages’157 were built right in the middle of elephant habitat, and the animals now come into our fields in search of food (interview, Farmer 27, 04/09/11). Many farmers I spoke to had constructed electric fences around their properties to protect their crops. Chena farmers typically had the simplest fences—usually a few strands of wire connected to a simple electric circuit powered by a solar panel—costing between Rs. 40,000 to 50,000 (US$ 350 to 440) for the battery, panel and circuit (interviews with Farmers 6 and 7, both on 06/04/11). Given their meagre incomes, they often get into debt to finance these types of infrastructure (more details on their precarious financial existence are provided shortly).

157 ‘Tsunami villages’ were villages that were constructed inland for people—previously living along the coast—whose homes were destroyed by the tsunami that devastated the area in December 2004.

193

Figure 6.6: A rudimentary electric fence around a chena. The solar panel used to power the fence is to the right of the hut (photograph taken on 25/04/2011, by S. Kalpage)

Many paddy farmers’ fields are within larger complexes of fields called a ‘yayas’. In areas in which I conducted my interviews, a yaya was between 100 and 200 acres in extent, comprising individual plots of around 2-5 acres. A yaya usually had an electric fence around its perimeter, powered by solar panels housed in small buildings set in a corner. Initial costs were often subsidized by NGOs such as the Centre for Conservation Research; the yaya’s farmers formed societies to share labour and other fence- maintenance costs. Farmers still watched over their crops at night— despite these fences—to chase away elephants before they tried to breach fences158. When elephants arrived, the farmers tried to chase them away by shouting and lighting crackers; the Department of Wildlife sometimes distributes powerful crackers called ‘ali wedi’ among farmers (e.g. interviews, Farmers 1, 2 and 3; all on 06/04/11).

158 Elephants are intelligent animals, and given enough time, often identify weak points in electric fences from where they can break in.

194

Figure 6.7: A yaya (group of paddy fields) surrounded by an electric fence. A watch hut can be seen at the perimeter of the fence (photograph taken on 04/09/2011, by S. Kalpage)

Figure 6.8: Solar panel and room containing the electric circuit powering an electric fence around a yaya (group of paddy fields) (photograph taken on 07/04/2011, by S. Kalpage)

195 Overall, electric fences were effective if constructed and maintained properly. For example, conservationists from the Centre for Conservation Research—who conducted the pioneering research on elephants’ use of chenas—have demonstrated the effectiveness of fences in a village called Thammanna. Their goal was to convince local chena farmers to use protective electric fences during the rainy season, but to take down fences soon after crops are harvested at the onset of the dry season, so that elephants could utilise the sprouting vegetation. A member of the team involved in the construction and maintenance of these fences described to me how the scheme worked:

The Centre’s first project was started in this village (i.e. Thammanna) in 2008. We first helped villagers construct a permanent fence around the perimeter of their village. We also taught chena farmers how to construct temporary fences around their fields and dismantle these fences after harvesting their crops; we provided them with material to construct fences, but they were responsible for fence maintenance. At the beginning, we inspected fences on a weekly basis, but since May 2010 have done so only monthly. A village society has been established to maintain the fence around the village (interview, NGO 1, 25/03/11). This system has appeared to work according to Farmer 19.

Before the electric fence was in operation, we had lots of issues with elephants. Since the fences were put up by CCR 159 we’ve had no problems—elephants have not destroyed posts here as they have done elsewhere. It’s pretty simple: if we don’t have the fence we can’t grow our crops! (interview, Farmer 19, 25/04/11). The farmers included in the World Bank survey agreed that electric fences are effective—over 50% believed that this was the case (Table 6.3). However, an equal number wanted elephants outside the fences chased away. This suggestion, however, is unlikely to be carried out, given the Department of Wildlife’s staffing and resource constraints described later in this sub-section160.

159 Centre for Conservation Research. 160 Even if the Department had the resources to do this, an obvious issue would be determining to which locale staff could chase these elephants. As described in Chapters 3 and 4, Sri Lanka’s PA network is not big enough to contain Sri Lanka’s elephant population.

196 Table 6.3: Farmers’ views on methods—both currently used and those that may be used in the future—to mitigate their elephant-related issues161

Method to reduce famers’ Percentage of Percentage of non- Percentage of all ‘elephant problems’ chena farmers chena farmers farmers answering answering ‘yes’ answering ‘yes’ ‘yes’ Properly maintain electric 58% 49% 52% fences Chase away elephants outside 63% 48% 53% fences Provide villagers with crackers 19% 18% 19% Other 7% 13% 11%

In spite of the relative success of electric fences, both the conservationists and villagers who led efforts to maintain fences stressed that farmers still needed to watch over their crops because a few elephants have become adept at breaking fences. According to Farmer 3: “Elephants are very clever and a few have figured how to deal with electric fences. At times, they roll logs onto fences and barge through” (interview, Farmer 3, 06/04/11). The problems arising from the lack of farmers’ vigilance was evident from the break-ins at Thammanna village in September 2011 (as discussed previously, this village was surrounded by an electric fence that had safeguarded farmers’ crops for several years). Farmer 39 told me how the lack of vigilance on the part of several farmers resulted in elephants breaching the fence and destroying banana cultivations located next to the fence (interview, Farmer 39, 05/09/11; Figures 6.9 and 6.10).

161 The numbers do not add up to 100%, because some farmers indicated more than one method.

197

Figure 6.9: Banana plantation damaged by elephants in the Thammanna village (photograph taken on 05/09/2011, by S. Kalpage)

Figure 6.10: Concrete posts holding electric wires damaged by elephants in the Thammanna village (photograph taken on 05/09/2011, by S. Kalpage)

198 Farmers, especially those who cultivate chenas, have a very precarious existence. Since they do not own land and have few other financial assets, they are not eligible for bank loans, and have to rely on moneylenders charging exorbitant interest rates of over 10% a month (interview: Farmer 2). Since they are heavily in debt until harvest-time, farmers are constantly worried that elephants would destroy crops they desperately need to feed their families and repay their loans. Farmer 6 explained his predicament:

The required infrastructure is very costly. For example we need to build electric fences around our properties, costing Rs. 42,000 for the battery, panel and circuit. We finance these investments by pawning our gold jewellery at banks, or borrowing from moneylenders at exorbitant rates. Our livelihoods are very risky, as we are totally dependent on a good harvest. It took me three seasons to repay Rs. 15,000 I had taken before I cultivated my crops! If people have a few bad harvests they are sunk, since the bank takes possession of their jewellery, and it is auctioned off. Overall, we live from hand-to-mouth, and if elephants destroy our crops we are ruined (interview, Farmer 6, 06/04/11). Many farmers had problems selling their crops at fair prices, and felt that middlemen and traders to whom they sold their produce were exploiting them. According to Farmer 44:

We sell generally our produce at Pannegamuwa or Tissamaharama. If we go through the middlemen, they take large cuts—if they buy from us at Rs. 25 a kg, they sell at Rs. 45. The government is supposed to buy our paddy at pre-determined prices, but rice mill-owners collude among each other to lower prices when they buy from us. They then sell this paddy to the government at higher prices, making huge profits. We are forced to sell to these mills as soon as we harvest our crops, because our creditors know when harvest time is and call our loans immediately. So, when we try to sell our produce to the mill-owners, prices are usually low. However, prices seem to rise soon afterwards, and they make a killing (interview, Farmer 44, 06/09/11). This situation has also been described by de Silva and de Silva (2007; p.207), and in the Sri Lankan press. For example, according to a newspaper article:

Chena farmers remain among the poorest people in the country. The vegetables they grow are purchased by mudalalis (i.e. merchants) from town at rock bottom prices. All chena vegetables flood the market at the same time-and ironically they are of the same kind, causing prices to plummet. So the net income for all that hard work is very meagre. Earning Rs. 15,000 to 30,000 out of a several-acre chena plot would be considered a good income. But this has to sustain the farmer family for the rest of the year (Dissanaike 2004; p.1).

199 Many farmers, who had lived in the area for many years, told me that elephant-related problems have worsened over time. When they first arrived in the area the animals were shy and avoided contact with people, but have now become more aggressive, probably driven to desperation due to dwindling water and food sources. The following account by Farmer 27 is typical:

I came to this area when I was still a child and have lived here for over 40 years. In those days, elephants and people kept their distance from one other. Elephants were scared of fire and would run away even if we flashed a torch. However, now, with our rapidly growing population and so-called development, we are encroaching on their territory. For example, lots of villages have been built in the middle of elephant habitat, so the animals have much less food to eat. These days some elephants—usually certain males—are much more aggressive, and have no hesitation in raiding paddy fields or home gardens. They have broken my fence twice. Instead of running away from torches, they actually attack people who shine torches at them. It is very difficult to chase them away (interview, Farmer 27, 04/09/11). Despite all these elephant-related problems, many farmers recognised that human beings were ultimately responsible for the recent escalation of the conflict. Several farmers told me that the area’s rapid development had deprived elephants of food and water; some even told me that farmers needed to recognise that they were encroaching on elephant habitat, as illustrated in the following quotes:

Our issues with elephants have worsened in recent years, because their habitat is being destroyed. The new airport at Mattala is being built on prime elephant habitat, and there are many—mostly illegal—rock quarries being operated in forested lands. People grow crops near local water sources, which are very tempting to hungry elephants (interview, Farmer 32, 04/09/11). In 1976, when I cultivated a chena, I used to leave for the chena (from the village) around 9 pm, but never had issues with elephants. Problems with them have become much worse since the 1990s, probably because roads have been constructed through the forest. Also, people have made ‘koratu’ (i.e. plots where cultivation takes place all year round) right by the main wewa (i.e. reservoir), restricting elephants’ access to the water. This is probably why they now come into the village (interview, Farmer 39, 05/09/11). The elephant is national treasure. We should always recognise that we are the ones who have taken over their habitat (interview, Farmer 34, 05/09/11).

200 In summary, while virtually all the farmers I interviewed struggled to protect their crops from elephants and faced significant threats to their livelihoods if their crops were destroyed, many recognised that elephants are a ‘national treasure’ worth protecting. However, many felt that they had been let down by the government. According to Farmer 7:

Elephants are a national treasure. We need to protect them for future generations, for our children who should be able to see elephants in the wild. However, people in this area are extremely poor and take drastic action when elephants destroy their crops, their main source of livelihood. No one thinks of us poor farmers. There are problems for us whether elephants live or die. If they live, our livelihoods are threatened. If they die, we are accused of killing them. There is no way we can win! (interview, Farmer 7, 06/04/11) Many farmers blamed the Department of Wildlife for failing to resolve human-elephant conflict. The following comments are typical:

Wildlife officials don’t care about our issues. They come by only if elephants have problems, not if we have problems (interview, Farmer 39, 05/09/11). Wildlife officials are useless. When we ask them for help, they say that they don’t have people or vehicles. They sometimes won’t come even if people are killed! (interview, Farmer 25, 04/09/11) The wildlife officials don’t maintain the electric fences around parks. If they did their job, we would have few problems with elephants (interview, Farmer 7, 06/04/11). Like these farmers, most locals believe that controlling elephants is the responsibility of the Department of Wildlife. Consequently, serious elephant attacks—especially those resulting in human deaths—often lead to confrontations between wildlife officials and local villagers. Several farmers told me about an incident that has led to much animosity and mistrust between the two groups.

We had serious issues with the department five years ago, when an elephant killed a villager. Some enraged villagers attacked local wildlife officials, and the magistrate instructed the police to catch 100 people (‘anyone’). The police came into our village and arrested many innocent people, as the culprits had already gone into hiding. Lots of innocent people received jail sentences, while the actual perpetrators got away (interview, Farmer 5, 06/04/11). It was completely unjust. The real culprits had already gone into hiding by the time the police came, so they caught innocent people. I was arrested

201 along with six chena farmers; I spent 6 months in jail and was fined Rs. 35,000 (interview: Farmer 7, 06/04/11).

Other environmental threats to farmers’ livelihoods

While crop raiding by elephants is a major threat to farmers’ livelihoods, several other environment-related issues emerged in both the survey and the semi-structured interviews. Table 6.4 presents some of the survey findings in this regard.

Table 6.4: Livelihood impacts related to the environment, apart from human-elephant conflict (n=800) (based on analysis of data from the World Bank Survey)

Percentage of total Type of threat Frequency162 respondents Forest fires 0 0% Drought 260 33% Dry spells 490 61% Floods 255 32% Other wild animal attacks 54 7%

Table 6.4 indicates that droughts and dry spells are the main threat to farmers’ livelihoods, as these impact their crops. This finding was confirmed by the semi- structured interviews. For example, according to Farmer 34:

I cultivate a paddy field in the village yaya. However, we have not been able to cultivate our fields for 3 years due to the lack of water, and have had to buy rice to feed our families (interview, Farmer 43, 05/09/11). Other farmers face the opposite problem, however, as indicated in Table 6.4, flooding, which has also been highlighted in recent news reports:

The recent unseasonal heavy rains have shattered the hopes and dreams for a prosperous new year for more than 60% chena cultivators in the Dambulla district. The unusually heavy rains have resulted not only in crop failure, but also in the poor quality of the crop (Ariyadasa 2012; p.1). Many farmers I interviewed told me they had noticed significant changes in weather patterns in recent times. In fact 38 of the 69 farmers I interviewed believed that climate

162 The numbers do not add up to 800 because some farmers cited more than one threat.

202 change was affecting their crops; those who had lived in the region for decades expressed their views emphatically:

Weather patterns have definitely changed over the years. I have noticed changes since the 1980s, when we faced prolonged droughts. Rains have been declining over the years, and are now much less predictable. The start of the rainy season is often delayed. These changes make it very difficult to cultivate crops (interview, Farmer 44, 06/09/11; he has been cultivating his chena since 1965). The weather has definitely changed over the last few years, and we have faced a series of severe droughts. When we first came here the rains arrived as expected, during the Yala and Maha seasons. Now it is hotter, and we often only get light showers. Earlier, we could cultivate paddy fields every 2 – 3 years; the wewa (i.e. the local reservoir) would fill up during the Maha rains, but now this rarely happens (interview, Farmer 17, 25/04/11; he first came to the area in 1963). Several farmers cited the rapid deforestation occurring in the area as a possible cause for these changes. For example:

Weather patterns have definitely changed over the years. Until about the 1980s, there was rain during both kannas (i.e. growing seasons). However, since about 1987, there is sufficient rain only during the Maha, and even that has become less predictable. This has probably happened because our local forests—especially in places such as Lunugamvehera and Uda Walawe—are being destroyed due to ‘development’ (interview, Farmer 39. 05/09/11). Those days, the rains came on time; it rains much less these days. Now, even for Maha, it rains only 4 – 5 times. Those days, we had forest all around us, which retained moisture. Now, the area looks like a desert, and rains have been affected (interview, Farmer 34, 05/09/11). Table 6.4 also showed that while elephants are the main threat to farmers’ crops, other animals also often pose problems. Farmers I interviewed told me that species such as cattle, wild pigs, monkeys, peacocks and other birds are a major nuisance, although, unlike elephants, they do not directly threaten human beings.

We have to protect our crops from monkeys during the daytime, and elephants and cattle at night (interview, Farmer 8, 06/04/11). If you consider threats to our crops, animals like peacocks and monkeys are sometimes worse than elephants. Monkeys raid our paddy fields and eat the ripening paddy (interview, Farmer 27, 04/09/11).

203 Issues with land tenure and the impact of local ‘development’ activities

Apart from human-elephant conflict and other environmental threats to their livelihoods, many of the farmers I interviewed—especially those who cultivated chenas—were worried about their continued rights to land they have been cultivating for many years. As discussed in Chapter 5, although chena cultivation is technically illegal, locals recognise traditional land rights, as Farmer 4 described:

All chena cultivation in this area is technically illegal, since it is on Forest Department land. However, people in this area recognise our traditional rights to this land. Any local will be able to show you “Hichchi Aiya’s chena” or “Somey Mama’s chena”. Our family has cultivated this chena for over 20 years (interview, Farmer 4, 06/04/11). Many farmers told me that land has become much more valuable in recent times as a result of the area’s rapid development. According to Farmer 17:

Because of the development in the area, there has been a huge increase in land value. In 1973, 3 acres cost Rs. 1,000. In 1995, 1.5 acres cost Rs. 50,000. Now 1.5 acres cost Rs. 5 to 6 lakhs163 (interview, Farmer 17, 25/04/11). Consequently, many chena farmers are worried about their continued rights to land they had cultivated for many years. Some of them talked about corrupt land deals in the area, and how land associated with state-sponsored irrigation schemes have been only allocated to government supporters:

We see a lot of ‘development’ taking place in the area. Many hotels are coming up. However, we worry that this will impact our chena land, as land gets more valuable. We urgently need security of land tenure—the lack of security is why we cultivate temporary crops. Unfortunately, it is easy to get illegal lands deeds by bribing the Grama Sevaka164. Even the priest at our local temple was recently involved in an illegal land deal, where he helped an hotelier obtain a 12 acre block of land near the temple, in exchange for a large donation. Now, people from Colombo have started to grab land in this area using money and influence to create false deeds (interview, Farmer 5, 06/04/11). Overall, there is a lot of corruption, and the ‘small man’ always gets squeezed. We have cultivated this land for over 20 years, but still don’t have deeds. However, people with money—especially from Colombo— have obtained land in the area by bribing officials to create false deeds.

163 A lakh is a 100,000. 164 The government official responsible for the village (or a cluster of small villages).

204 Many hotels have been constructed on land obtained in this manner (interview, Farmer 6, 06/04/11). We are currently cultivating our crops here illegally, but urgently need clear land titles. If each person from the village is given at least an acre of land with a clear title, we can take care of elephant-related issues ourselves. Tremendous injustice is being done, as people from outside the area are given land as part of the development scheme (i.e. the Walawe Left Bank Scheme), while locals (i.e. from Sooriyawewa or Muwan Pelessa) get nothing. Most of this land is distributed among the ruling party’s political supporters (interview, Farmer 12, 07/04/11). Although chena farmers had the most severe land-related issues, data from the World Bank survey showed that many rice farmers also did not own land they cultivated, as the majority are tenant farmers cultivating paddy according to a system called ‘ande govi’. This system involves a verbal agreement between the landowner and the tenant farmer to provide a portion of the harvest—usually between a quarter to a half—as rent for using the land to grow crops, usually rice165 (de Silva and de Silva 2007; p.217). In recent times, the transaction sometimes resembles a leasing agreement, as some farmers agree to pay a specified monetary amount upfront for the season. Table 6.5 shows land ownership patterns of paddy farmers included in the World Bank survey. As farmers in the categories ‘taken on Ande’, and ‘leased in’ do not own their land, over a third of the farmers in the sample are landless.

Table 6.5: Land ownership of paddy farmers (n=247) (based on analysis of data from the World Bank Survey)

Type of tenure Percentage of total Legally owned 49% Taken on Ande 22% Given on Ande 4% Leased in 12% Leased out 2% Other (e.g. shared ownership) 10% Total 100%

As a result of these issues with land tenure, many farmers (both paddy and chena) viewed the area’s rapid development with mixed feelings. For example, at the beginning

165 The landowner gives the farmer no assistance other than land.

205 of our interview, Farmer 31 acknowledged some of the benefits from local development activities:

I have now lived in this area for 25 years. When we came here it was very underdeveloped, heavily forested. We had no electricity or water facilities, and had to walk miles just to get drinking water. We got electricity 10 years ago and our water is now pumped from the Gonnoruwa tank. So, in these respects, development has improved local livelihoods (interview, Farmer 31, 04/09/11). These benefits not withstanding, however, this farmer soon started talking about some of the negative effects of development on local livelihoods, especially as it has coincided with an increase in corruption:

Cultivating chenas is how poor people previously eked out a living. However, chenas are no longer permitted near our village, and the Special Task Force166 (STF) conducts patrols to enforce this. The Grama Sevaka is involved as well, and cases are filed against people who violate the law. The issue is that these laws apply only to poor people. Rich people who are close to Ministers clear 50 to 100 acre plots using bulldozers. They destroy pristine forests, but no one stops them. In the end poor locals are left with nothing (interview, Farmer 31, 04/09/11). Several other farmers echoed this pessimistic view of the area’s rapid development, saying that while it had benefited some people, many of the poorer locals, especially farmers, have been negatively affected:

A six-lane highway is being constructed through our field. After we harvest our crops this season the fields will be bulldozed, and we will be left with nothing. 17 farmers will be losing their land. Government officials tell us: “We don’t have any land to give you in return, but we will compensate you monetarily”. Whether we get any of this compensation is another thing entirely; even if we do, it will be a pittance. Already, half of the property on which my house sat on has been appropriated, but they are actually asking for a further 20 feet, to grow shade trees along the highway! We have lived 20 – 30 years on this land, but our homes are now in peril. So, we have no benefits from development—even the little we have is being lost! (interview, Farmer 38, 05/09/11) There is a lot development in the area. A new port and airport are being built. Roads are being constructed. However, these do not have an impact on our livelihoods. There is a lot of corruption associated with the allocation of jobs in these ventures; people get jobs due to political connections. Everything has been politicised (interview, Farmer 25, 04/09/11).

166 An elite commando force created by the government during the civil war.

206 Although there is a lot development in the area we don’t benefit economically. In fact, the development is bad for us; only rich people benefit. Many restrictions have been imposed on farmers like us. Land has also become valuable and there is a lot of land grabbing in the area, mostly by government supporters (interview, Farmer 34, 05/09/11).

Figure 6.11: Evidence of the rapid development in Sri Lanka’s southeast landscape—the construction of a four lane highway through a rural village; the previous single-lane road is visible (photograph taken on 25/04/2011, by S. Kalpage)

Government officials

The two main government agencies that are stakeholders in the human-elephant conflict are the Department of Wildlife and the Forest Department—the former is responsible for elephant conservation, while the latter manages most elephant habitat outside PAs. However, given the national-level attention the conflict has received—especially in the popular press—several other government departments and ministries are also keen to resolve the conflict. I now consider each of these sub-groups.

207 The Department of Wildlife

As described in Chapter 2, the Department of Wildlife is responsible for the conservation and management of Sri Lanka’s animal and plant species—including elephants—as per the Fauna and Flora Protection Ordinance. As discussed in Chapter 4, resolving human-elephant conflict is the main focus of the department—the initiative to which it allocates most of its resources. The Department is anxious to find a solution to the conflict, especially because local villagers affected by elephant depredations often blame wildlife officials for not doing enough to reduce the conflict. As discussed previously, several farmers I interviewed told me that this issue has led to much tension between locals and wildlife officials. When elephants kill people in the area, villagers’ protests often turn violent; I witnessed such a protest on February 22, 2012 when villages burnt tires blocking the entrance to the Bundala National Park and threatened to assault local wildlife officers (luckily the officers managed to escape).

Figure 6.12: Villagers protest the killing of a woman by an elephant at the entrance to Bundala National Park (photograph taken on 06/04/2011, by S. Kalpage)

208 Such incidents have been highlighted in press articles; the following news report is an example (Rodrigo 2011; p.1):

Residents of the area blocked Palagala junction last week, demanding a solution for the human-elephant conflict. About 1,500 villagers gathered at this junction on July 20, protesting the deaths of 7 villagers killed by elephants in the last two months. Traffic from Kekirawa, Galewela and Mahawa was blocked, causing severe inconvenience to the public. The Department of Wildlife assured villagers that they would relocate the troublesome jumbos, and asked protesting villagers to disperse. The farmers included in the World Bank survey share these views (Table 6.6).

Table 6.6: Farmers’ views of success of the Department of Wildlife in addressing their issues with elephants (based on analysis of data from the World Bank Survey)

Perception of DWC Chena farmers Non-chena All farmers (%) intervention (%) (n=252) farmers (%) (n=800) (n=548) Not answered 6% 12% 10% Doing a good job 30% 25% 26% No significant intervention 46% 40% 42% No intervention at all 15% 17% 16% Negative perception 4% 7% 6%

According to Table 6.6, over 60% of each category of farmers perceives that the department’s efforts to address their problems are unsatisfactory while only about a quarter believe that they are doing a good job. The department’s acute lack of resources is a key contributor to its ineffectiveness. Like most government employees in Sri Lanka wildlife officials are under-paid, under-equipped, and over-worked. Although the department is responsible for managing a network of PAs covering about 14% of Sri Lanka’s land area, it only has about 800 employees (Kuruwita 2011). Moreover, since wildlife rangers often have to deal with armed poachers, their job is often dangerous. A recent article on a wildlife officers’ strike highlighted issues they are struggle with (Kuruwita 2011; p.1-2):

Sri Lankan park rangers and field officers began a work to rule campaign on Friday and say they will continue it till they win their demands. “With the work to rule campaign the rangers will not respond to any emergency which may occur after 4.45 pm. Usually we are on call 24

209 hours of a day because no one can say when elephants would raid villages, or when poachers will enter the PAs, or when we will receive news of illegal timber merchants operating within the parks,” said Premarathne Manawadu, secretary, All Island Wildlife Officers’ Association. They demand that the long-standing salary anomalies be rectified. The rangers who at times work a 24 hour shift are paid a basic salary between Rs.12,000 – Rs.13,000167 and an allowance of Rs. 500. Wildlife officers are responsible for almost 14% of the total territory of the country and they are also accountable for the flora and fauna of PAs. They patrol PAs, conduct raids and apprehend suspects, file cases, accept bookings, sell tickets, prepare accounts, rescue animals, chase elephants when necessary, and in this midst of this get regularly shot at by poachers – and are attacked by wild animals. In the last 25 years, 97 wildlife officers have died while carrying out their duties and hundreds have been injured or wounded. At present, there are only 800 officers and they find it extremely difficult to control the illegal activities that take place in PAs. In the last few years dozens of elephant calves have been snatched from the national parks, while the poaching of sambur, leopards and deer are at an all time high. Illegal felling of timber and sand mining are also on the rise and many attribute the deteriorating situation in the parks to the lack of manpower at the Wildlife Department. Therefore this work to rule might lead to a sudden spike in illegal activities. “This might also lead to an increase of casualties in the human- elephant conflict since wildlife officers will not be available when elephants enter villages as the villagers will be left on their own to defend themselves,” said Sajeewa Chamikara of the Environmental Conservation Trust. One of the farmers I interviewed was in agreement with the view of the conservationist cited in the article. He told me that when he and some other villagers had gone to speak to wildlife officials regarding some elephants that had broken into their village, the officials had refused to help because they had been on strike:

When we went to the department office to tell wildlife officers about the damage done by the elephants, they told me that they were on strike, and therefore could not do anything. Apparently there was a ‘work to rule’ action by the trade union, because department workers had not been given their promised increments by the government. In fact, they said they were happy that people needed them, since it would draw attention to their cause, and challenged us to report this to their superiors. They also told us: “we really want attention to this. Our monthly salaries are only Rs. 14,000, and our overtime and other bonus payments have not been approved” (interview, Farmer 40, 05/09/11).

167 US$ 105 – 115.

210 The conflict sometimes jeopardises careers of top wildlife officials as evident from the recent sacking of the Director of Wildlife who failed to adequately consider the political ramifications of his Department’s approach to the conflict (Nizam 2011; p.1):

Wildlife Director General, Dr. Chandrawansa Pathiraja, was removed from his post yesterday by Wildlife and Agrarian Services Minister S. M. Chandrasena. The Island learns that among the causes for his removal were elephant drives and relocation, particularly in the Wayamba area. Dr. Pathiraja took the initiative to relocate wild elephants which caused the human-elephant conflict in the Wayamba Province. This created problems to the people as the wild elephants caused more damages to lives and properties particularly farming lands. The initiative resulted in irate farmers and affected persons voting against the government, officials said. They said Dr. Pathiraja did not consider the political cost of his decision and therefore he was sacked. Department of Wildlife officials have realised that resolving the conflict would require attention to elephant habitat outside the PA network—land that is currently outside their purview. Therefore, they are likely to be supportive of MERs, as evident from a recent press interview by the Director General (Yatawara 2011; p.1):

“With the country heading towards rapid development, finding solutions to the human- elephant conflict is of paramount importance,” said Director General of the Department of Wildlife Dr. Chandrawansa Pathiraja. If we are to successfully conserve elephants and mitigate There is now an increasing awareness among consumers that traditional varieties are much better for their health There is now an increasing awareness among consumers that traditional varieties are much better for their health There is now an increasing awareness among consumers that traditional varieties are much better for their health human-elephant conflict it is recommended that the failed approach of restricting elephants only to PAs be curtailed. The new approach is based on creating Elephant Conservation Areas and Managed Elephant Ranges, which is advocated by the National Policy for the Conservation and Management of Wild Elephants formulated by the government in 2006 and ratified by the Cabinet of Ministers, Dr. Pathiraja said.

The Forest Department

Unlike the Department of Wildlife, the Forest Department has a commercial outlook, as described by de Silva and de Silva (2007; p.125-126):

From the time of their inception, the two departments had different outlooks. The emphasis of the Forest Department has been on the

211 controlled commercial exploitation of forests and reforestation of deforested areas with mono- and poly-culture using exotic species... The emphasis of the Department of Wildlife, on the other hand, has been on protection and conservation of wildlife and their habitats; harvesting and replacing of forests is not allowed in the reserves under the department. Most Forest Department land comprising elephant habitat is designated as ‘Other State Forests’ and is currently under illegal chena cultivation (de Silva and de Silva 2007; p.127). As the Department currently obtains little or no revenue from these areas, it is likely to be interested in exploring ways to generate revenue from these lands. An employee of a global development organisation told me in an interview, that the Forest Department had been receptive to the idea of MERs, when he once mentioned it to them:

In fact, I had a discussion on regulated chena with the Forest Department during the preparation of a project proposal; they were not opposed to the idea. So we may be able to work with Forest Department on this (interview, NGO 5, 08/03/12). He also told me, however, that bureaucratic competition between the Department of Wildlife and the Forest Department might be an obstacle to their working together to implement MERs:

A significant issue is the view among many wildlife officials that they have propriety rights on elephant viewing. During an ‘off the record’ discussion with a high ranking official, he told me that he wanted the media to raise the question as to why Forest Department is given rights to elephant viewing in Hurulu Eco Park and he wanted the media to put pressure on the Government to hand over Hurulu Eco Park to Department of Wildlife! So we need to first overcome this type of nonsensical thinking among wildlife officials who feel that only they should earn revenue from elephant viewing! (interview, NGO 5, 08/03/12) A recent press article describes how this inter-agency tension has resulted in the haphazard location of electric fences on administrative boundaries between the two departments, restricting elephants’ access to food and water resources (Ladduwahetty 2011; p.2):

One of the solutions that has been devised has been the deployment of electric fences, which can work and are practical when they are placed in the right location. Over 70% of electric fences are placed on boundaries of national parks and Forest Department land surrounding them. While there are administrative differences between the Wildlife Department and the Forest Department, the elephants do not realize that! When they come to

212 fences, they see food on either side. In fact, there is better food on the Forest Department side! Elephants are known to prefer degraded forests, and often break fences and go out. Elephants are intelligent, and to avoid getting electrocuted, they carry logs and place them on fences, breaking them. They cross over, and there is nothing to stop them from invading agricultural lands and villagers. Electric fences have a role, but in the right location. They have to be placed on ecological boundaries and not on administrative boundaries. The lack of cooperation between the two departments has probably been exacerbated because they report to two different ministries. The Department of Wildlife is now under the Ministry of Agrarian Services & Wildlife, while the Forest Department is under the Ministry of the Environment; prior to 2010, both were under the Ministry of Environment and Natural Resources (Presidential Secretariat 2010; p.52, 88).

Government departments involved in land-use planning and management

As pointed out in Chapter 2, apart from the Department of Wildlife and the Forest Department, the activities of several other government organisations impact conservation. Examples include agencies involved in land management such as the Land Commissioner General’s Department, those covering agriculture such as the Department of Agriculture, and those responsible for irrigation such as the Department of Irrigation and the Mahaweli Authority of Sri Lanka. These entities impact elephants, because, as was pointed out in Chapter 3, the animals range widely outside PAs, often coming into contact with expanding villages and irrigated agriculture. Therefore, when implementing MERs, it would be necessary to engage government authorities whose activities could impact, and be impacted by, the establishment of these areas. The specific bodies to be consulted should be determined on a case-by-case basis (depending on the specific area), and buy-in should be obtained from both central authorities and local government bodies such as District Secretariats and Divisional Secretariats.

A recent article advocating the need to consider elephant ecology and behaviour in land- use planning provides some insights to the numerous entities that should be consulted (Yatawara 2010; p.1):

213 An example of land-use planning taking into consideration elephant ecology, existing land use patterns, and needs of development is the zoning plan developed by the Urban Developmental Authority and the Central Environmental Authority for the greater Hambantota area. The Action Plan suggests the involvement of direct agencies involved in developing land use plans, such as the Urban Developmental Authority, the National Physical Planning Department, and the Land-Use Policy Planning Division of the Department of Agriculture, to take into account elephant presence when land-use planning is conducted.

High-ranking government officials

The rapid escalation of human-elephant conflict and the extensive press coverage regarding its impact on local villagers has brought the issue to the notice of high- ranking government officials. For example, the Minister of Agrarian Services and Wildlife, under whose purview the Department of Wildlife has now been placed, recently admitted that current efforts to resolve the conflict had been ineffective, but that he would resolve it “at any cost” (Edirisinghe 2011; p.1):

Minister of Agrarian Services and Wildlife S. M. Chandrasena said that he was well aware of this problem because he is from the Anuradhapura District. He said that no proper programme had so far been conducted to end human–elephant conflict. “I directed the ministry to start a programme to stop this,” Chandrasena said adding he will do it at any cost. The conflict has received attention at even higher levels within the government, given its social and political significance. As pointed out in Chapter 4, resolving human- elephant conflict was part of President Mahinda Rajapaksa’s 2010 election manifesto. Recent news reports confirm his continued concern (Weerasinghe 2011; p.1):

President Mahinda Rajapaksa has instructed Treasury Secretary P. B. Jayasundara to render his full cooperation to activities that concern resolving human-elephant conflict, the Daily News reliably learns. Meanwhile, steps are being contemplated at ministerial levels to pay a sum of Rs. 50,000 or Rs. 60,000 per acre of elephant crop damage to ensure the livelihoods of farmers living in villages near forests. Therefore, it is evident that mitigating human-elephant conflict is a priority not only for the Department of Wildlife, but also for high-ranking government officials and politicians, including even the President of Sri Lanka. Moreover, there are several other government agencies, including the Forest Department and various state institutions with jurisdiction over land, which would need to be consulted when implementing

214 MERs. Another government agency with a significant stake in conserving Sri Lanka’s elephants is the Sri Lanka Tourism Development Authority, which we consider next.

Tourism authorities

The Sri Lankan government has been trying to boost tourism in the country since the end of the country’s civil war in 2009. Showcasing the country’s biodiversity has been an integral part of the country’s tourism strategy; in fact, ‘Wildlife & Nature’ is one of the eight product categories included in the Sri Lanka Tourism Development Authority’s marketing campaign, as described by its Chairman, Nalaka Godahewa (Island newspaper 2011):

For the convenience of marketing we have summarized all attractions in Sri Lanka into 8 categories. The number 8 has been selected because there are 8 letters in the name Sri Lanka. The 8 different product categories that we want to promote are: 1. Beaches - Pristine 2. Sports & Adventure - Thrills 3. Heritage sites - Heritage 4. Mind and Body wellness - Bliss 5. Scenic beauty of the country - Scenic 6. Wildlife & Nature - Wild 7. People & Culture - Essence 8. Year-round Festivals - Festive Now just think of any tourism destination in the world where you can find all these in one place. Even if you do, where else can you cover all these within a few days? Sri Lanka is probably the only country which makes it possible. That’s why we can call Sri Lanka the ‘Wonder of Asia’.

Elephants are a key element in the ‘Wildlife & Nature’ component. Indeed, in 2011, the Lonely Planet travel guide declared the annual gathering of elephants at the Minneriya National Park in northwest Sri Lanka, the sixth greatest wildlife spectacle in the world (Weerasuriya 2011).

Given the preceding discussion, it is evident that a range of government departments should play a role in mitigating Sri Lanka’s human-elephant conflict. Unfortunately, coordinated action is difficult because these departments fall under the purview of

215 various ministries that may have differing or even conflicting priorities. For example, the Sri Lanka Tourism Development Authority falls under the Economic Development Ministry whose main purpose is fostering regional and local development and tourism while the Forest Department is under the Ministry of Environment whose aim is “preserving the environment for present and future generations” (Presidential Secretariat 2010; p.42, 52). Moreover, the Department of Wildlife is under the Ministry of Agrarian Services and Wildlife, which (among other tasks) seeks to “facilitate ecotourism in wildlife PAs with due consideration to the conservation of the ecosystems concerned” (Presidential Secretariat 2010; p.89), while the Department of Land Commissioner General is under the Ministry of Lands & Land Development, which is trying to allocate land among various competing uses (Presidential Secretariat 2010; p.63).

Tourism companies

Many Sri Lankan tourism companies—large conglomerates as well as smaller niche players—have demonstrated interest in the country’s growing nature tourism market. For example, a top official at Aitken Spence—a large tourism operator—explained his company’s strategy to the media:

“Sri Lanka is a small country, densely populated, with lean infrastructure,” said Malin Hapugoda, managing director of Aitken Spence Hotels, a big operator in Sri Lanka and the Maldives. “So we can't grow in numbers like Thailand. We need to be aware of our carrying capacity, and assess the potential impact of 2.5 million tourists on the country,” Hapugoda said Aitken Spence's strategy was to target niche-market, high-yielding tourists, given the country's small size and its biodiversity attractions. “We should use the country's biodiversity to get high-yielding, niche products,” he said. “Aitken Spence is going on that line to some extent. We're targeting the upper market segment while having a presence in the volume market. That would be the way to go without creating rooms and rooms, without a sustainable plan” (Lanka business online 2010; p.1). John Keels Group, a large conglomerate with significant interests in tourism, is also focusing on nature tourism—one of its recent initiatives is the extensive remodelling of its hotel located near Yala National Park in southeast Sri Lanka (Sunday Times newspaper 2011; p.2):

216 Chaaya Wild Yala, officially opened its doors on November 1, 2011 after undergoing a complete refurbishment. Situated bordering Sri Lanka’s most famed National Park, the resort’s offerings stem from responsible tourism practices aimed at conserving and being one with its precious environs. With over Rs. 400 million investment the 66-chalet world-class game lodge is placed in what is almost an extension of the Yala National Park with open boundaries to the surrounding wildlife. Sri Lanka’s Association of Inbound Tour Operators (SLAITO) organised several events and promotional activities in support of a government-declared wildlife month in September 2011 (Mendis 2011). Predictably, elephants were showcased: “announcing the line-up of events, SLAITO President Nilmin Nanayakkara stated that the presentations would touch on four key aspects of Sri Lankan wildlife, i.e. witnessing the ‘elephant gathering’ at Minneriya, bird viewing in Sinharaja, leopard sightings and blue whale sightings” (Mendis 2011; p.1).

The corporate sector’s interest in nature tourism was evident among representatives from some leading tourism companies I interviewed. For example:

Business has really picked up since the end of the war. We are now busy even during the off-season. About two-thirds of our clients are strictly nature tourists; the remaining third combine nature tourism with other types of activities (e.g. beaches, culture). Nature tourists, who are mostly from European countries such as the UK, Germany and the Netherlands, typically stay about two weeks. Our clients are sophisticated wildlife viewers, and expect a premium product (interview, Tourism operator 1, 13/09/12). The strictly nature-oriented tourists we serve visit Sri Lanka country solely for its wildlife. They stay here for 8 to 15 days. Group sizes range from 4 to about 20. Most are interested in viewing charismatic species such as leopards, elephants and whales. Most nature tourists visit Yala, combined with whale watching at Mirissa, while some dedicated birders visit places like Sinharaja and Bundala. If Sri Lanka remains peaceful, wildlife tourism is a huge opportunity for our company. Our industry should endorse wildlife tourism using international operators with big names and good reputations (interview, Tourism operator 2, 15/09/12). The head of one of Sri Lanka’s leading tourism operators told me that his company would welcome opportunities to establish operations in areas outside PAs, such as proposed MERs:

Private game viewing areas and reserves should be developed. There are thousands of acres of suitable land—especially in southern Sri Lanka— near national parks. Several companies should get together and purchase

217 land, or seek 99-year leases from the state. This would be the next level in nature tourism, and we need to make it happen in Sri Lanka. I have broached the idea with my colleagues; my company is definitely interested! (interview, Tourism operator 2, 15/09/12)

Other locals

Local groups such as jeep drivers and local businesses benefit from nature tourism in the southeast landscape, and are therefore potential stakeholders regarding MERs. I interviewed several jeep drivers who make a living by taking tourists to national parks. While their businesses had been badly affected by significantly reduced tourist arrivals during the civil war, they are now very busy due to the recent surge in park-oriented tourism. Not surprisingly, many are enthusiastic nature tourism supporters.

I have worked closely with the owner of Kulu Safaris for many years, and, in fact, helped him set up his first camp at the Yala National Park. He still contacts me when clients need transport from the airport to the park. I love nature, and enjoy working with Kulu Safaris, because it gives me an opportunity to be in the forest for a few days (interview, Other locals 4, 28/08/11; he had worked with Kulu Safaris, a small high-end nature tourism outfit that operates in Yala and Uda Walawe National Parks). I own three jeeps. I drive one and hire out the others. Of the Rs. 3,500 we charge per park visit, we spend about Rs. 2,000 on direct running costs; the rest is profit, but we need to set aside money for longer-term repairs and tires. During the high season—from November to early January, and during the summer months in northern hemisphere—we get at least one hire per day on average, but during the low season each jeep gets hired out only once every 2 – 3 days. Tourism is booming now, so we have a lot of business. However there were many periods, especially during the war years, when things were very tough (interview, Other locals 6, 06/09/11; he owns some jeeps which take tourists to the Yala National Park). Many local small businesses also benefit from nearby PAs. The owner of a small supermarket in Tissamaharama town, which is close to the Yala National Park, explained how the park helps his business:

In the Tissa area, only 5% of people have government jobs (or draw regular salaries); agriculture is the main occupation. Having the park nearby really helps locals, as the influx of park visitors—local and foreign—benefits local hotels, tour operators, jeep operators, restaurants and retail businesses such as mine. Things are very good now, but some years ago, the park was closed for about 1.5 years due to terrorist activity

218 within its borders. Local businesses, including mine, suffered; many went bankrupt (interview, Other locals 1, 21/03/11; he owns a supermarket).

The Sri Lankan public

As discussed in Chapter 3, the elephant is a national icon, given its important role in Sri Lanka’s history, culture and folklore. However, people who bear the brunt of elephant attacks are likely to have very different views of these animals compared to those of people living in cities. A study by Bandara and Tisdell (2003) provides some insights regarding the contrasting attitudes of these two groups. The authors surveyed 300 residents each in three areas in Colombo and six randomly selected villages in high conflict areas (Bandara and Tisdell 2003; p.331).

Each group was asked its views on a suggested policy that calls for reducing Sri Lanka’s elephant population by 50% to open up more land for agriculture and human settlements. While 94% of city dwellers rejected this proposition, about 94% of people living in rural areas accepted it (Bandara and Tisdell 2003; p.334). The authors point out that these findings illustrate “farmers’ animosity towards crop-raiding elephants” and “their dissatisfaction with the government’s commitment to resolve the country’s human-elephant conflict” (Bandara and Tisdell 2003; p.334).

The two groups also disagreed over how to deal with ‘problem animals’. When asked whether “local farmers in the vicinity of the nature reserves should be granted more freedom to control elephants which cause crop and property damage”, about 81% of urban respondents disagreed, saying that “if the local farmers were granted a free hand to control crop raiding elephants, they will use destructive methods such as guns, traps and poison…eventually leading to the elimination of the country’s entire elephant population” (Bandara and Tisdell 2003; p.334-335). In contrast, 79% of rural farmers agreed with the suggestion. Despite their opposition to allowing farmers to deal freely with crop raiding elephants, most urban dwellers recognise that farmers should be compensated for the damages to their crops and property; in fact 82% of urban respondents were in favour of this (Bandara and Tisdell 2003; p.340)168.

168 More details are provided in Appendix 6.

219 An interesting survey finding, however, was the degree of agreement between both groups as to whether elephants should be conserved—only 3% of urban residents and 8% of rural residents disagreed with the statement that “conservation of elephants in the wild is important for both economic and non-economic reasons” (Bandara and Tisdell 2003; p.335). Nonetheless, the authors point out that the underlying reasoning behind their decisions are somewhat different (Bandara and Tisdell 2003; p.336):

It appears that rural dwellers (farmers) give much more weight to historical, cultural and religious values than do the urban residents who relatively give more weight to the altruistic, bequest, and existence values of the elephant. Moreover, urban dwellers put more weight on the biodiversity and ecological value of the elephant than rural respondents. In summary, while rural Sri Lankans who face elephant attacks view the conflict differently from people who live in cities, both groups generally agree that elephants should be conserved, mainly for non-economic reasons.

Non-governmental organisations

An entire range of NGOs could be considered stakeholders in the human-elephant conflict. At one end of the spectrum are those focusing on biodiversity conservation in general, and elephant conservation in particular; some conservationists within these entities view local villagers as a threat to elephants. At the other extreme are organisations whose main aim is to alleviate rural poverty. Some people who work for these entities view elephants—especially those that range outside national parks—as a threat to the livelihoods of the people they are trying to help.

Some conservationists, however, believe that a lasting solution must consider needs of both elephants and locals. Professor Santiapillai is an example (Yatawara 2010; p.2):

“Wildlife is a sustainable natural resource. Some argue that all wildlife must be simply protected and not exploited. Others, including me, believe that it is only by allowing people who live in and around areas frequented by wildlife to derive economic benefit from it that they can be led to appreciate the value of wildlife and thus become committed to conserving it,” said Professor Santiapillai explaining the options left for Sri Lanka to mitigate human-elephant conflict.

220 Researchers at the Centre for Conservation Research, who proposed the establishment of MERs, also recognise the need for human-elephant coexistence. Press articles have highlighted the views of its chairman, Prithiviraj Fernando (Mendis 2009):

“Elephants are very intelligent, they are very attached to their original home ranges and most often always try to come back,” said Prithiviraj Fernando, who heads the Sri Lankan Centre for Conservation and Research. Fernando, who has been tracking elephants for nearly 15 years using radio collars, says Sri Lankans must learn to live alongside elephants (Dawn newspaper 2010; p.1). Centre for Conservation and Research Chairman Dr. Prithiviraj Fernando said the best elephant habitat is a mosaic of open areas and regenerating forests... “People have been generating favourable conditions for elephants for over 1000 years now. Chenas provide them food while the irrigation system provides them water. But the human-elephant conflict arises when human territory and habitats overlap.” Prithiviraj said that creating Managed Elephant Reserves, done in collaboration with the DWC as part of their National Policy, is a fool-proof strategy. “This involves a radical method of integrating human and elephant habitat to ensure co-existence of the two species by cultivating chena,” Prithiviraj said. “A chena provides a rich source of fodder for elephants. We are proposing that the existing chenas be regulated, while the PAs will continue to remain untouched by human activity” (Mendis 2009; p.1-2). The Centre for Conservation Research—which advocates a scientific, fact-based approach to mitigate the conflict—works closely with other stakeholders (CCR 2009). It collaborated with the Department of Wildlife to understand elephant ranging behaviour by fitting collars on several elephants169. Once researchers understood that elephants and locals could share resources in areas where chena cultivation is practiced, they worked with chena farmers to test an approach that aims to achieve this idea. The NGO has also helped locals and other farmers to erect fences around villages and rice fields to test the efficacy of properly located and maintained electric fences, and ultimately seek ways to implement MERs, a concept that the group has proposed.

Other NGOs have adopted a similarly balanced approach to elephant conservation. The Sri Lanka Wildlife and Conservation Society, for example, believes that “the only realistic path to sustainable conservation is to ensure a reasonable standard of living for all people - especially for the rural population of developing countries with some of the

169 This Centre is still collecting data from many of its collared elephants; it uploads data from one of its elephants onto Facebook.

221 world's richest biodiversity” (Sri Lanka Wildlife and Conservation Society 2011). The society has on-going projects in several parts of Sri Lanka that help farmers protect their crops from elephants. The Biodiversity and Elephant Conservation Trust is another organisation that believes that mitigating human-elephant conflict is essential to elephant conservation, and has established social programmes to help human victims; some of these programmes are described on the organisation’s website (BECT 2012).

Several NGOs, whose main objective is to improve the rural livelihoods in Sri Lanka— such as CARE international and the Poverty Reduction and Economic Management Network (PREM)170—have also worked with local farmers in southeast Sri Lanka to help install electric fences to protect their crops from elephants.

Overview of section 6.2

In this section I identified the main stakeholder groups, and used various information sources—especially semi-structured interviews and survey data—to explore the main issues, concerns and needs of each group. This analysis helps us understand how key stakeholders have been influenced by, or could influence, the escalating human- elephant conflict, and the role these groups could play in the establishment of MERs. It also provides key information required for the next step in my stakeholder analysis, i.e. differentiating among and categorising stakeholder groups, which I discuss next.

170 The Japanese International Cooperation Agency (JICA), an Overseas Development Agency, also has several projects in the area.

222 6.4 Differentiating among and categorising stakeholders

I categorised stakeholders using an ‘influence-importance matrix’, which has two axes: importance, or the extent of each stakeholder’s needs and interests in the issue at hand, and influence, or the power each stakeholder has over successfully implementing a solution171 (Grimble 1998; p.176; Reed et al. 2009). In the context of my analysis, importance is the extent of a stakeholder’s needs and interests regarding human- elephant conflict, while influence is the power each stakeholder has to successfully implement the proposed conflict mitigation approach—i.e. MERs. Figure 6.13 shows my attempt to locate the stakeholder groups on the matrix.

INTEREST

Subjects Key players

Chena farmers Dept. of high Wildlife

Other farmers NGOS

Sri Lankan Other locals public Forest Dept.

Tourism companies low Other key ministries & depts*

Crowd Context setters

low high INFLUENCE

* Those related to land development, agriculture, tourism

Figure 6.13: Key stakeholder groups, and select sub-groups in my case study located on the influence-importance matrix

Figure 6.13 illustrates how the main stakeholder groups fall into the categories proposed by Reed et al. (2009). The Department of Wildlife and NGOs—especially those

171 Further information about the ‘influence-importance matrix’ is presented in Appendix 6.

223 working on elephant conservation and/or improving rural livelihoods in conflict areas— are the ‘key players’. Both groups of famers are ‘subjects’. The general public and other locals who stand to benefit from nature tourism are also ‘subjects’, although less important ones. The Forest Department is the most important ‘context setter’—in fact it borders on being a ‘key player’, given the importance of some of the land under its management. However, since its interest in resolving the conflict is currently low, I classified it as a ‘context setter’. I also put government departments and tourism companies in this category, since, despite interest from some tourist companies in the MER concept it is still not widely accepted in this sector172.

It is important to realise that the positions of the various stakeholders in Figure 6.13 represent my best guess as to their current positions, as there is a certain degree of subjectivity associated with the entire exercise. In fact, there have been criticisms of categorising stakeholders in such a manner. Firstly, although the main purpose of stakeholder analysis is to highlight the interests of marginalized groups, differences in power among various stakeholder groups—made explicit through influence-interest matrices—could actually result in the opposite occurring, unless explicit participatory processes are put in place (Chevalier 2001). Secondly, the exercise may imply that stakeholder views are fixed, while in fact they may change over time and even according to the particular context, given “the multiple hats that stakeholders wear and the complex networks they belong to” (Chevalier 2001; p.7; Mayers 2005; p.13). I acknowledge these criticisms and stress that the main purpose of my stakeholder mapping is to understand current power differences among stakeholders so that these could be addressed in MER design173. In fact, in Chapter 8, I recommend measures to involve all the main stakeholder groups in the planning and implementation of these areas through participatory processes, and make specific suggestions as to how MERs could be designed to address most stakeholder needs (although some trade-offs would have to made).

To successfully implement MERs, actions should be taken to move several groups in the ‘right’ direction in the matrix, i.e. towards the top right hand quadrant. This is especially necessary in the case of important ‘context-setters’, and one of the main ways

172 Most Sri Lankan tourism companies would probably not even know what an MER is. 173 Appendix 6 presents further details on potential issues with my stakeholder analysis and how these issues are addressed.

224 to achieve this move is to establish MERs in a manner that ensures they have significant economic benefits, and clearly communicate these benefits to stakeholders. Perhaps the main stakeholder is the Forest Department, which has jurisdiction over for much of the land included in my proposed MERs. As it currently derives little revenue from these areas, it would be interested in proposals that make this land more profitable. Economic benefits are also crucial to important private sector stakeholders. As noted previously, large tourism operators I interviewed (e.g. Tourism operators 1, 2 and 8; interviewed on 13/09/11, 15/09/11 and 25/02/12) would welcome opportunities to expand their growing nature tourism business outside national parks; some even mentioned that they already had internal discussions regarding MERs. In addition to tourism companies, local entrepreneurs such as jeep drivers would be supportive if the proposed areas stimulate local nature tourism, giving them opportunities to expand their businesses. Therefore, presenting the economic rationale for these areas—as I will do in Chapter 7—would prove crucial in obtaining support from important stakeholder groups.

6.5 Investigating relationships between stakeholders

The final step in the stakeholder analysis (as described in Figure 6.1), entails investigating the relationships among stakeholders, which I have done using an actor- linkage matrix. Reed et al. (2009; p.1939) describe this method and its advantages:

A commonly used means of describing stakeholder interrelations is through actor-linkage matrices. These require stakeholders to be listed in the rows and columns of a table creating a grid so that the interrelations between them can be described, using key words. One popular method for example is to determine whether the relationships between each stakeholder are of conflict, complementary, or cooperation. The advantage of this approach is its simplicity of use and flexibility. I constructed an actor-linkage matrix using the results from my semi-structured interviews and information from other sources described in Table 6.2 (Figure 6.14).

225 Farmers Dept. of Wildlife Forest Dept. NGOs

Farmers

Dept. of Wildlife Conflict over destruction by elephants

Forest Dept. Conflict with chena Some bureaucratic farmers over land- competition use

NGOs Depends on NGO Some co-operation, Some co- goals, e.g. rural especially with operation development conservation NGOs especially with versus conservation conservation NGOs

Figure 6.14: Actor-linkage matrix for key stakeholders in Sri Lanka’s southeast landscape

For ease of analysis and presentation I only constructed the matrix visually for the four stakeholder groups that interact most with the others, i.e. farmers, the Department of Wildlife, the Forest Department, and NGOs. After describing these interactions in more detail, I will also briefly consider interactions of other stakeholder groups, i.e. other ministries and departments, tourism companies, other locals and the Sri Lankan public.

Both chena farmers and other farmers generally have adversarial relationships with the Department of Wildlife, since these groups generally believe that the department is not doing enough to stop elephant depredations, sometimes even leading to clashes between the two groups. There is tension between chena farmers and the Forest Department over land issues, as the former believe that they have traditional rights to land despite chena cultivation being technically illegal, and resent attempts by the latter to curtail their activities. Chena farmers also believe that Forest Department officials are corrupt and unfairly penalise them, while turning a blind-eye to illegal activities of people who are rich and/or politically connected.

226 The Department of Wildlife and the Forest Department generally have little interaction with each other, but, on occasion, engage in ‘turf warfare’. For example, as described previously, the Department of Wildlife believes that it should have proprietary rights to benefits from elephant viewing. Problems between the two departments are likely to have been exacerbated since they were placed under different ministries.

The relationship between NGOs and other stakeholders depends on each particular NGO’s focus. As described earlier, some conservation-oriented organisations like the Centre for Conservation Research, the Biodiversity and Elephant Conservation Trust and the Sri Lanka Wildlife Conservation Society, work closely with local people affected by human-elephant conflict. The Centre for Conservation Research has also worked closely with the Department of Wildlife to fit collars on elephants to better understand their ranging behaviour. Several other NGOs such as CARE International and the Poverty Reduction and Economic Management Network (PREM) help local farmers construct electric fences to protect their crops.

Other government ministries and departments have some interactions with various stakeholder groups. For example, those responsible for agriculture and irrigation may interact with local farmers, but usually not on matters related to elephants or human- elephant conflict. The interaction between these organisations and the Department of Wildlife and the Forest Department is typically minimal; as was pointed out in Chapter 2, weak inter-departmental communication channels make it difficult to address issues cutting across organisational jurisdictions. Although there is some dialogue between the Sri Lanka Tourism Development Authority and the Department of Wildlife regarding PA-related tourism, several interviewees told me that this is still insufficient (e.g., NGO 2, 3, and 4; interviewed on 20/09/11, 05/0312 and 06/03/12).

Tourism companies also interact with the Department of Wildlife and the Forest Department regarding PA-related nature tourism. After all, the companies such as Kulu Safari, which operate within national parks such as Yala and Uda Walawe, need to book campsites and pay the required fees. The companies sometimes meet with these departments and with the Sri Lanka Tourism Development Authority to suggest how processes could be streamlined to improve their clients’ experiences. Tourism companies also interact with some local villagers who derive income from nature tourism: hiring jeeps to take their guests into national parks and buying supplies from

227 local businesses. Tourism companies inevitably interact with members of the general public who are their clients.

Members of the Sri Lankan general public who do not live within the landscape generally have minimal interaction with other stakeholder groups. However, some people interested in conservation issues may support NGOs working in the area, or even seek membership in these organisations. Others who are interested in wildlife and photography visit parks and engage tourism companies to facilitate these visits. In the process, they would interact with park rangers and guides and may hire local jeep drivers and/or buy supplies from local retailers.

Having completed the three steps in the stakeholder analysis (described in Figure 6.1), I now consider the insights gained from this exercise.

6.6 Implications of the stakeholder analysis

In drawing implications from my stakeholder analysis, I consider how it has helped answer the questions I posed at the beginning of this chapter, namely:

a) Who are the stakeholders in the human-elephant conflict within the southeast landscape?

b) What is the nature of their interest in the issue?

c) What strategies should be adopted to achieve MERs goals while meeting stakeholder needs and concerns?

To answer questions (a) and (b) I will take in turn each of the stakeholder groups and summarise the nature of their stakes in the human-elephant conflict and the establishment of MERs.

The main interest of both groups of farmers is to protect their crops, homes and lives from marauding elephants that have become increasingly bold and desperate. As both groups are extremely poor, improving their financial condition is one of their major goals. In addition, chena farmers are extremely concerned about land tenure issues.

228 Among government entities, the Department of Wildlife is the most interested in mitigating the conflict, since this is one of its main responsibilities, and its lack of progress thus far has resulted in friction with locals, criticism in the press and the ire of its superiors. The Forest Department is not directly concerned about the conflict, but would like to increase the profitability of forested lands on which chena cultivation is currently taking place. Higher government officials—especially those who are politicians—seek a resolution to the conflict for political reasons.

The main goal of tourism companies and other entrepreneurs within the group ‘other locals’ is to maximise nature tourism-related revenues. However, since many people in these groups are conservation-minded, they would also like to contribute to elephant conservation.

The Sri Lankan public, especially those living in urban areas, are keen to conserve elephants for future generations. While their rural counterparts would also like to conserve elephants for historical and cultural reasons, those who live in areas of conflict are worried about the damage to life and property caused by elephants. The latter group overlaps significantly with the farmers described above.

NGOs would like to achieve the goals of their respective organisations. While those at extreme ends of the spectrum either focus exclusively on conserving elephants, or on improving rural livelihoods, several in the middle advocate approaches encouraging human-elephant coexistence. An NGO in the latter category suggested the idea of MERs; most other ‘moderate’ NGOs are likely to support the approach. As most NGOs are funded through individual donations, the private sector and/or development organisations, they would like to engage in activities that satisfy current donors and facilitate future fundraising initiatives.

Having answered questions (a) and (b), we can now use insights gained from the stakeholder analysis to answer question (c): ‘what strategies should be adopted to achieve MER goals while meeting stakeholder needs and concerns?’ Before I explore this in detail in Chapter 8, I will first address an important issue that concerns many stakeholder groups: namely what would be the economic benefits of MERs and how could these benefits be distributed among stakeholders? As pointed out in section 6.3 above, demonstrating that MERs could be economically viable would be crucial to

229 obtaining support from several important stakeholder groups. My analysis is presented in next chapter: the economic justification for establishing MERs.

230 Chapter 7: The economic viability of Managed Elephant Ranges

One of the key findings of my stakeholder analysis is that several important stakeholder groups would be interested in and/or concerned about potential economic benefits from MERs. Accordingly, in this chapter, I present the economic rationale for setting aside these areas. First, I explore the need for valuation according to economic theory, discuss some of the related theoretical concepts, and examine the main criticisms of this sometimes-controversial process to clarify how analysis results should be interpreted. I then present the total economic value framework on which my analysis is based, and apply it to Sri Lanka’s southeast landscape. Using an economic model I developed, I consider how different patterns of land-use could impact the total economic value of resources within the landscape. Finally, I consider potential land-use scenarios more holistically, bringing together my economic analysis with the social and equity aspects considered in Chapters 4 and 5.

7.1. The economic explanation for deforestation and the need for valuation

As pointed out in Chapter 1, tropical forests and the biodiversity they contain are being lost at a rapid rate; this is also happening in Sri Lanka (Chapter 2). Deforestation and loss of elephant habit significantly contribute to human-elephant conflict in Sri Lanka in general (Chapter 4), and in the southeast landscape in particular (Chapter 5). Therefore it is important to understand the key drivers of deforestation—especially from the perspective of economic theory—before exploring the economic viability of the proposed approach to mitigate the conflict, i.e. MERs.

Pearce et al. (1994) distinguish between proximate and fundamental causes of deforestation. Proximate causes—often part of popular explanations of forest loss— include logging, clearing forests for agriculture, and pollution. Fundamental causes, which lie behind these proximate causes, are rooted in economic, institutional and social factors (Pearce et al. 1994). For example, deforestation generally results from efforts to provide timber and agricultural products to meet global and local demand, usually

231 occurring when it is cheaper to meet this demand by supplying products from (previously) forested land than from other land (Eliasch 2008). Eliasch (2008; p.36) identifies several other important drivers of deforestation: (a) policy incentives (either direct, such as subsidies, or indirect, such as those for building roads into forests or encouraging migration to them); (b) land tenure systems; and (c) other governance- related incentives. Figure 7.1 summarises his view.

Population Rising Changing Policy Land tenure and growth incomes diets incentives governance

Growing demand for agricultural Extra incentives to products and timber deforest

Deforestation occurs when it is cheaper to meet demand from supplying products from converted forest land than from other land

Figure 7.1: The underlying drivers of deforestation (from Eliasch 2008; p.36)

The explanation provided by economic theory

Economic theory describes two kinds of market failure that contribute to issues such as deforestation—‘externalities’ and ‘intervention failure—which I now address in turn. Pearce et al. (1994; p.16) explain how externalities contribute to deforestation:

The main reason for the loss of forests is that there is an underlying disparity between the private and social costs and benefits of forest use and conservation. Private costs and benefits refer to those losses and gains as perceived by the immediate user of the environment: the farmer, the industrialist, the consumer. Social costs and benefits refer to the losses and

232 gains that accrue to society as a whole. Social and private interests often do not coincide: what is good for me as an individual may impose costs on the rest of society—so-called externalities. The theoretical explanation becomes clearer when we consider a concrete example: the case of farmers deciding whether to clear native forests for agriculture (Pagiola et al. 2002). In making this decision, farmers would weigh the benefits to them (e.g. harvest) against the costs (e.g. effort to clear the land, fertilizers and other inputs). However, they are unlikely to consider benefits that do not apply to them, such as watershed protection; although cutting down forests may increase downstream flooding and sedimentation, these will be borne by people living downstream. Moreover, the farmers are extremely unlikely to consider global benefits such as carbon sequestration. Consequently, from the perspective of farmers, the value of forests they plan to destroy appears to be far lower than it really is. In this example, forest benefits such as watershed protection and carbon sequestration—referred to as externalities in economic theory—are not captured by freely functioning markets based on narrow self-interest (Pearce et al. 1994). In economic terms, this situation is described as ‘market failure’ or distortions due to ‘missing markets’.

Figure 7.2 illustrates the above example in stylized form, each circle representing a service provided by the forest174. While there is some overlap among benefits, each stakeholder group perceives a different mix of benefits, determining each component’s importance based on their own priorities and preferences. The outer outline of the set of overlapping circles represents total benefits from the forest. The main insight from this example is that if local decision-makers are not compensated for providing benefits to other groups, they are unlikely to consider (and mitigate) impacts of their land-use decisions on these benefits.

174 Six are shown, although there could be more.

233 Non-timber products Carbon sequestration Biodiversity Fuel wood

Water Reduction of purification sedimentation

Benefits perceived by: Local forest managers Downstream water users Global community

Figure 7.2: Beneficiaries of forest services (adapted from Pagiola et al. 2002)

‘Intervention failure’ is the second category of ‘market failure’ identified by Pearce et al. (1994). This refers to situations where government interventions in markets— usually to serve social purposes such as helping low-income farmers—result in environmental degradation 175 . For example, the Brazilian government provides tax credits for ‘investment’ costs incurred to clear forests for cattle ranching; without this and other related subsidies176, it is estimated that Brazilian cattle ranching revenues would cover only about one third of the associated costs (Pearce 1991). In other words, without these subsidies, forest destruction for cattle ranching probably would not occur since it is highly unprofitable for ranchers.

175 The two boxes on the top right hand corner of Figure 7.1 are in this category. 176 Subsidised credit for livestock development, for example.

234 Possible approaches to address deforestation

There are several ways to respond to market failure associated with forests. For example, the government could assume direct responsibility for managing forest resources in the ‘public interest’. In the specific example above, it could declare the watershed as a ‘forest reserve’, prohibiting deforestation within its boundary. However, Pagiola et al. (2002) point out that governments often do not perform according to expectations, especially in developing countries where governments often lack information about various services177 provided by forests; even if this information is provided to them, they often lack funds to pay for the required conservation. Furthermore, they are not immune to political pressures, such as lobbying from businesses that would profit from deforestation (Pagiola et al. 2002).

Given these difficulties with government remedies, economists have proposed market- based mechanisms to address market failure. Figure 7.3 illustrates the basic logic of these mechanisms. Two hypothetical market-based mechanisms are shown: one that sells sedimentation reduction services to downstream water users, and another that sells carbon sequestration services to buyers needing credits to meet Kyoto emission reduction requirements178 (Pagiola et al. 2002; p.5-6). These payments essentially transfer some of the benefits received by these groups to the local forest managers. As a result, the total benefits of forest conservation, as perceived by forest managers, increase, perhaps substantially. Assuming that the benefits of converting forests to other uses remain unchanged, forest managers are more likely to decide to preserve forests.

It is important to note some key points:

• Even with these payments, the local farmers still perceive only a sub-set of all forest benefits. All that is needed is to provide them with sufficient incentives to preserve the forests (as compared to other possible uses).

• Some farmers may still cut down forests, e.g. if the payment is insufficient, if most of the payment is appropriated by a third party, or if a select group of

177 The ‘ecosystem services’ provided by forests are discussed in detail in section 7.4. 178 The Kyoto Protocol of the United Nations Framework Convention on Climate Change (UNFCCC) commits developed countries to legally binding targets to limit or reduce their greenhouse gas emissions. This is described in more detail in Section 7.4.

235 farmers get the majority of the benefits (i.e. due to a lack of equitable distribution).

• There may be practical issues with setting up the system of payments, especially if there are a large number of farmers.

Payment from carbon credit buyers to forest Non-timber managers products

Carbon sequestration Biodiversity

Fuel wood

Water Reduction of purification sedimentation

Payment from water users to forest managers

Benefits perceived by: Local forest managers Downstream water users Global community

Figure 7.3: The logic of market-based mechanisms for forest services (adapted from Pagiola et al. 2002)

The need for valuation

Regardless of whether the government approach or the market approach is chosen, assessing the value of the benefits provided by the forest would be important. In the case of the direct government regulation, it is crucial to identify which forest areas should be declared PAs, because there would be several alternate land-use options for the potential areas to be set aside. Governments would need to balance the differing needs of their citizens, for example farmers’ need for land to grow crops versus water use by people downstream. More generally, governments consider three options when deciding forest-related policies: (a) preservation, i.e. exclude human use; (b)

236 conservation, i.e. permit some human use; and (c) development, i.e. deforestation (Ninan 2007)179.

Pearce (1991) identifies two biases that often result when governments make these decisions. The first arises because development costs and benefits can be readily calculated since there are attendant cash flows and readily observable prices (e.g. for timber production), while conservation benefits are a mix of associated cash flows and ‘non-market benefits’ (e.g. watershed protection). Therefore, benefits from the development option appear more real (a phenomenon which he terms ‘misplaced concreteness’), and decisions are likely to be biased towards this option. The second bias occurs because non-market benefits do not usually accrue to those who clear forests; this ‘asymmetry of values’ also causes a bias toward the development option (Pearce 1991). Therefore, valuation of the ‘non-market benefits’ associated with the conservation option would help reduce the biases against it, and may persuade governments to choose this option180.

Valuation is also crucial if market-based mechanisms are chosen, to determine who should make the payments (and how much each individual or group should pay) and who should receive the payments (and how much each individual or group should receive).

Therefore, in summary, valuation helps society make informed decisions about trade- offs among alternative land-use options regardless of whether state-led or market-based approaches are adopted. Ninan (2007; p.6) cites three benefits arising from valuing forest resources:

1. It provides a way of trading off objectives;

2. It provides a way of arriving at a decision that maximizes well being181; and,

179 These three options are generalisations of points within a continuum of possibilities, ranging from preservation to deforestation; the actual policies of various governments could lie on any point in this continuum. Further details regarding the three options are presented in Appendix 7. 180 For example, when non-market benefits from mangroves such as their role in coastal protection, as breeding habitat for fish, and their ability to store carbon are taken into account, it is much more profitable to conserve them instead of destroy them for shrimp farms and coastal development as is currently being done throughout Asia (World Bank 2012; p.6). 181 As we will see in the next section, there are some aspects of human-well being that are not measured by economic valuation, so in many situations this approach should be used in combination with other approaches.

237 3. It is effective since it speaks in the economic language to which policy-makers listen.

The use of valuation within a cost benefit analysis framework

The person or people choosing among the various land-use options presented in the previous section182—regardless of whether they are from the government or the private sector—would weigh the costs versus the benefits of each option183. Economic theory attempts to capture this behaviour through its cost-benefit analysis framework. As described by Mullan and Kontoleon (2008; p.6), cost-benefit analysis “is a structured set of methods for comparing the benefits and costs associated with the provision of different levels of market and non-market goods and services, such as those derived from forest biodiversity”. Given that cost benefit analysis is central to my economic analysis of the natural resources within my chosen landscape, I now consider the method’s theoretical underpinnings as well two important concepts associated with it: prices and discounting.

The theoretical underpinnings of cost benefit analysis

Mullan and Kontoleon (2008; p.6) describe the theoretical basis for cost-benefit analysis:

Cost-benefit analysis is a tool-kit that operationalises the logic inherent in the main normative decision making criterion in modern economic science, namely the potential compensation Pareto criterion, which compares alternative policy options on the basis of whether they lead to an efficient allocation of resources. Put simply, a Pareto efficient policy change is that in which the net winners can potentially compensate the net losers from the change, and no other policy (or resource allocation) results in a greater overall level of utility184.

182 Preservation, conservation, and development. 183 It is important to note that all costs or benefits do not necessarily need to be financial or economic, as discussed in section 7.3. 184 It is also important to note that neither cost benefit analysis nor Pareto efficiency address equity issues; these are discussed further in section 7.5.

238 In their explanation above, the authors first indicate that cost benefit analysis is part of the ‘normative’ branch of economics. There are two main types of economic analysis that could be applied to increase our understanding of the relationship between the economic system and the environment: positive economics, which attempts to describe what is, what was, or what will be, and normative economics, which deals with what ought to be (Tietenberg and Lewis 2012; p.19). Consequently, disagreements within positive economics can usually be resolved by an appeal to the facts, but normative disagreements involve value judgments.

Another issue touched upon in the above explanation is how benefits and costs should be measured. For a project or policy to qualify on cost-benefit grounds, its social benefits must exceed its social costs; ‘society’ is simply the sum of individuals (Pearce et al. 2006; p.16). When measuring these costs and benefits, a fundamental requirement is that all relevant gains and losses are measured in the same units—otherwise they cannot be added together (aggregated), either across people or over time (Hanley and Barbier 2009; p.15). Conceptually, costs and benefits are calculated in units of utility, a term used by economists to represent those factors that make people happy, or that explain people’s choices (Tisdell 2005; p.2). While people can generally state their preferences—their preference for outcome A over outcome B for example—they usually find it more difficult to describe by how much. Therefore, economists generally use money metrics to measure underlying utility change, specifically “either the most that someone is willing to pay to acquire more of something desirable, or less of something undesirable; and the least that someone is willing to accept in compensation for giving up something desirable, or tolerating something undesirable” (Hanley and Barbier 2009; p.16).

Finally, as described by Mullan and Kontoleon (2008) above, an efficient policy change should meet the ‘Pareto’ condition that the maximum aggregate willingness to pay of the gainers (the social benefit) is greater than the aggregate minimum willingness to accept of the losers (the social cost) (Hanley and Barbier 2009; p.30). While compensation may not actually paid to losers, what is considered is whether the gainers could compensate the losers, and still be better off (Tisdell 2005; p.171).

239 The role of prices

Prices are crucial for the proper functioning of a market-based economy. Farley (2008; p.1400 - 1401) describe the role of prices, and their links to cost benefit analysis and to the concept of Pareto efficiency:

Prices serve a critical role in a competitive free-market economy, acting as a fulcrum that balances supply with demand, costs with benefits, and what is possible with what is desirable in three basic ways. First, prices signal scarcity. When market demand for a resource exceeds market supply, consumers bid up the price leading producers to supply more or create substitutes. When supply for a resource exceeds demand, suppliers lower their prices until the market again clears. Second, prices maximize net monetary value captured by all producers by allocating factors of production to whatever industry is willing to pay the most for them, which is the industry that generates the most profit from that factor. Third, prices maximize net monetary value captured by all consumers by rationing commodities to the individuals willing to pay the most for them. Market prices maximize the monetary value of both inputs and outputs. The outcome in theory is a competitive equilibrium in which it is impossible to improve the welfare of any one individual without making another individual worse off, a condition known as Pareto efficiency. This result is enshrined as the first fundamental theorem of welfare economics. Everything is driven by the decentralized, voluntary decisions of economic actors. However, as discussed at the beginning of the chapter, in the case of environmental goods such as ecosystem services, for which there are no properly functioning markets, private costs and social costs do not coincide. When negative externalities are present (as in the example of deforestation discussed above), social costs are greater than private costs (Hussen 2004; p.55). This mismatch is also reflected in the prices of environmental goods (the price of timber in this case). In economic terms, ‘market prices’ are lower than ‘economic prices’, which in turn provides incentives for people to engage in economically inefficient behaviour such as deforestation (Hussen 2004; p.55 - 58; Farley 2008; p.1401). Consequently, the solution according to economics is to understand the ‘true’ benefits of natural resources (or the costs incurred by using them in an unsustainable manner) through the process of valuation. This understanding enables the development of policy choices that ensure these values are reflected in the prices of the relevant goods and services (Farley 2008; p.1401; Ha et al. 2008; Molinos- Senante et al. 2013). I will discuss a range of various techniques at the end of section 7.3.

240 The discount rate

Costs and benefits of environmental goods and services are realised at various points in time. The process of discounting enables the (discounted) costs and benefits resulting from a project or policy choice to be added together to arrive at a single figure—net present value—that enables comparison among these choices (Hussen 2004; p.176). Tisdell (2005; p.75) describe these two concepts:

A sum of money available in the future is generally regarded as being worth less than the same sum of money available now because if the future sum were available now it could be invested at the going rate of interest to ensure its return plus accumulated income from interest on it at the later date. The sum of money which if invested now will accumulate to a future sum after addition of interest is the net present value or discounted present value of the future sum. Clearly the net present value of a future sum of money is lower the higher is the rate of interest and the more distant in time is the availability of the future sum. An important implication of the process of discounting is that the further into the future the benefit or cost occurs, the lower the weight attached to it. For example, even if we use a comparatively low discount rate of 4%, a gain or loss 50 years hence would be valued at only 14% of its value now (Pearce et al. 2006; p.184). Some economists argue that the interest rates used for many projects, especially those with significant environmental impacts, are too high from a social point of view, resulting in individuals investing in projects that are not socially optimal (Hussen 2004; p.182; Tisdell 2005; p.75). In other words, projects providing quick returns such as timber harvesting are excessively favoured in comparison to those with environmental benefits that are realised later (for example the immediate income from timber harvesting is valued more highly than the longer term benefits from forest-related environmental services) (Hussen 2004; p.182).

In spite of the problems with the process of discounting, using a discount rate of zero (i.e. not discounting at all) would unfairly penalise the current generation (Pearce et al. 2006; p.185). Gollier (2002; p.163) summarises the argument for a positive discount rate:

When public investment projects entail costs and benefits in the very long run, a question arises about the selection of the relevant discount rate to use for the cost–benefit analysis. Indeed, financial markets do not provide any guideline in this case. The main argument for using a positive discount rate

241 is the fact that the GNP185 per head is expected to grow over time. Therefore, projects whose costs today are as large as benefits in the future are clearly not desirable, since we do not see why current generations should sacrifice part of their consumption today for the benefit of future generations who will already be better off. The author suggests that variable rates should be used for different time horizons186: no larger than 5% for medium time horizons (i.e. up to 100 years) and about 1.5% for long time horizons (i.e. up to 200 years) (Gollier 2002; p.163). Consequently, I have used a discount rate of 5% in the economic model I developed to estimate the value of natural resources within Sri Lanka’s southeast landscape187 (described later in the chapter).

Having discussed the rationale for environmental valuation and its underlying theoretical framework, I now consider the opposition to valuation, and how some of these concerns could be addressed.

7.2. The opposition to valuation

Not surprisingly, there has been considerable debate on whether natural resources like biodiversity and tropical forests could be valued, or indeed should be valued188. There are two main categories of criticisms regarding such valuation. The first group challenges the basic premise of the practice, i.e., biodiversity valuation is inherently ‘anthropocentric’, only considering benefits to human beings. The second group challenges the use of valuation to determine biodiversity-related trade-offs. I now examine each of these views in more detail.

185 Gross National Product. 186 This is based on empirical evidence that suggests that the discount rate used by individuals to value the future is a decreasing function of the time horizon (Gollier 2002; p.157). 187 Gollier (2002; p.163) points out that most developed countries already use rates between 5% and 8%. 188 Most of the challenges focus on the valuation of biodiversity contained in forests, so the discussion in this section is centred on biodiversity. As pointed out before, most of the world’s biodiversity is contained in tropical forests, so biodiversity conservation requires forest conservation.

242 The challenge to anthropocentrism

As discussed above, a basic premise associated with valuation is that different groups of people perceive different benefits from biodiversity. Since valuation is used to determine trade-offs among possible uses of these resources, its underlying viewpoint undeniably revolves around humans, i.e., ‘anthropocentric’ (Kareiva et al. 2011; p.16).

This viewpoint has been criticised by the philosopher Arne Naess, who inspired the ‘deep ecology’ movement (Oksanen 1997). He believed that “the flourishing of human and non-human life on Earth has intrinsic value”, and that “richness and diversity of life forms are values in themselves and contribute to the flourishing of human and non- human life on Earth” (Oksanen 1997; p.544). These two ideas support the concept of bio-centric equality, or that “all organisms and entities in the ecosphere, as parts of the interrelated whole, are equal in intrinsic worth” (Oksanen 1997; p.544). However, practical difficulties arise when trying to implement the latter proposition. Firstly, it would be hard to prioritise species-conservation, since an elephant and a flea (and presumably even a human being) would have the same ‘intrinsic worth’. A greater difficulty arises with forms of biodiversity, such as some viruses, that are harmful to human beings—according to Naess’s proposition “we should undoubtedly allow malignant organisms to exist and, perhaps, to flourish; their human-caused extinction does not match with our fundamental moral obligation” (Oksanen 1997; p.544).

To address Naess’s criticisms, environmental philosophers distinguish between ‘strong’ and ‘weak’ anthropocentrism. ‘Strong anthropocentrism’ suggests that biodiversity exists only to fulfil human needs, and derives its value solely from its ability to meet specific human needs (Oksanen 1997). ‘Weak anthropocentrism’, on the other hand, acknowledges that, besides satisfying basic human needs, biodiversity contributes to the realization of human ideals (Oksanen 1997; p.542). In other words, while weak anthropocentrism maintains that the use of biodiversity to humans is an important consideration—especially since it is humans who make many decisions that are critical to their survival—it acknowledges that we have a moral obligation to preserve it.

In summary, Oksanen (1997) points out that although advocates of weak anthropocentrism and deep ecology have different viewpoints, they agree on the need to preserve biodiversity. Furthermore, when dealing with a complex issue like

243 biodiversity, no viewpoint or political strategy is universally applicable, and “real-life solutions require a great deal of flexibility and deliberation case by case” (Oksanen 1997; p.545).

The challenge to the use of valuation to determine trade-offs

O'Neill (1997) articulates why using economic valuation of biodiversity to determine trade-offs could be problematic, disputing each of the three benefits of valuation listed by Ninan (2007) in the previous section.

Firstly, O’Neill takes exception at the use of money as a single common measure to score alternative forms of land-use. His view is that “different values are incommensurable; there is no unit through which the different values to which appeal is made in managing a particular site can be placed upon a common scale” (O'Neill 1997; p.548). Consequently, he believes that there is “no substitute for good practical judgement that is informed by debate amongst practitioners and citizens” to resolve conflicts between landscape values, biodiversity, timber use, and cultural values (O'Neill 1997; p.548).

Secondly, O’Neill disputes that economic valuation could help us arrive at a decision that maximizes well-being since he believes that “environmental goods are matters of ethical principle” and “ethical preferences are not like non-ethical preferences” (O'Neill 1997; p.547). Consequently,

[o]ne should no more accept a price where issues of environmental value are involved than one should on issues of abortion, euthanasia, commercial surrogacy, hanging or any other issue of principle. To engage in monetary valuation in these arenas would be quite properly rejected as inappropriate, to ask willingness to pay questions an exercise in corruption. The proper mode of resolution is public debate in which utilitarians have to state their case with others, as they now do in issues of abortion, euthanasia, commercial surrogacy or hanging (O'Neill 1997; p.547). Finally, he believes that using economic valuation to put prices on goods that are not traded in markets—“theoretical prices constructed by economists”—results in values that are “unbelievable”, and using these values would show a lack of integrity on the part of environmentalists (O'Neill 1997; p.549).

244 Incorporating some of the (valid) criticisms of valuation to clarify its role in policy-making

O’Neill’s criticisms underline the importance using valuation as only one of several inputs into the policy-making process. He is right that the exercise of judgement is crucial in policy-making, and that the process should always include considerations such as cultural values and equity. He also correctly stresses the importance of stakeholder involvement and public debate, especially since valuation outcomes are often imprecise189. It is crucial that we always keep in mind the inherent complexities, uncertainties and irreversible characteristics of public goods such as biodiversity, using concepts such as the ‘precautionary principle’ and ‘safe minimum standards’ in decisions regarding how much we conserve190.

In summary, although it is impossible to calculate precise, completely defendable values of biodiversity or forests, valuation could make an important contribution to the policy- making process, especially because establishing PAs have a clear set of costs, both actual (e.g. the cost of maintaining national parks) and foregone opportunities or ‘opportunity costs’ (e.g. the value of timber which could have been harvested). In most cases—especially in developing countries—revenues from national parks would not justify their costs purely in conventional market-terms (e.g. through entry fees), making it increasingly difficult for governments to justify establishing such parks. As we shall see in subsequent analyses, however, the value of benefits such as ecosystem services could be orders of magnitude greater than those currently valued in the marketplace. Having even a rough idea of their value would significantly benefit policy-makers— especially those in developing countries—who face many competing priorities when making land-use decisions (e.g. increasing employment, reducing poverty, generating foreign exchange). Ninan (2007; p.8) summarises the case for valuation as follows:

Tropical countries like Malaysia and Indonesia, which rely on timber extraction for export earnings, would need appropriate economic incentives to forgo the development option and conserve their forest

189 This will be evident when we perform actual valuations in subsequent sections. 190 The Rio declaration on Environment and Development interprets the precautionary approach as meaning that “when there are threats of serious or irreversible damage to the environment, lack of full scientific certainty should not be used as a reason for postponing cost-effective measures to prevent environmental degradation” (O'Neill 1997). While critics of this approach are concerned about large commitments of resources to deal with vaguely defined problems, valuation, however imprecise, can help in this clarification.

245 resources. An idea of the development benefits foregone would thus be useful in designing conservation policies. The economic case for valuing biodiversity is, therefore, based on strong grounds. Therefore, valuation is extremely relevant to my case study, especially since government policymakers—perhaps the most influential stakeholder group—must agree to set aside land as MERs. Having established the case for valuation (and pointed out the caveats associated with using valuation), I now present the economic framework I have used to determine values of environmental resources in southeast Sri Lanka.

7.3. The concept of total economic value and how it could be measured

The concept of total economic value (TEV) captures the benefit or flow of services derived from an environmental resource in economic terms. In other words, it is an anthropocentric and relative concept that can only be applied to goods or services over which individuals are willing to make trade-offs (Mullan and Kontoleon 2008). Therefore, it does not capture the value of services for which trade-offs are not possible—the ‘intrinsic value’ for example. This caveat is further explored later in the section.

TEV asserts that environment resources have both ‘use value’ as well as ‘non-use value’. Use values arise from the use of the resource, including direct uses such as timber extraction and tourism, indirect uses such as watershed protection and nutrient recycling, and option value, or the deferral of direct and indirect use to make the resource available for use in the future. Non-use values are benefits people derive from knowing that biodiversity and forests continue to exist, as demonstrated by their contributions to environmental charities, for example. The concept of TEV becomes clearer when applied to a concrete example, such as to tropical forests.

246 Applying the TEV concept to tropical forests

Applying the concept of TEV to tropical forests, we come up with the following equation and components of value: A. Total economic value of tropical forests = use value + non-use value. Where: B. Use value = direct use value + indirect use value + option value C. Non-use value = existence value + bequest value. Table 7.1 explains each of the terms in (B) and (C).

Table 7.1: Components of TEV (adapted from Pearce 1991; Ninan 2007)

Term Explanation

Direct use value • This, in turn has three components: 1. Consumptive: Goods and services provided by forests, e.g. sustainable timber, non-timber products, medicinal plants, hunting, 2. Productive: plant breeding/ genetics, and 3. Non-consumptive: tourism, recreation, education.

Indirect use value • Ecosystem services such as watershed protection, nutrient and carbon cycling, carbon-capture (to mitigate global warming), biodiversity conservation, air pollution reduction.

Option value • The future use of both direct and indirect uses, e.g. the future value of drugs—this is like an insurance premium to ensure the supply of something, the availability of which would otherwise be uncertain (Pearce 1991; p.244).

Existence value • The value of forests to individuals who do not currently make use of forests but wish to see them preserved in their own right so that others can use them.

Bequest value • The preservation of forests for the use of future generations.

247 The correct use of the TEV concept

Since TEV attempts to value complex entities such as biodiversity and tropical forests, it is important to understand clearly how the concept should be used in the policy- making process. Therefore, I now address three concepts associated with TEV— aggregation, location dependence and monetisation—and identify what TEV does not take into account, i.e., distribution of value, critical thresholds and intrinsic values.

Aggregation

The components of TEV cannot simply be aggregated because there are trade-offs among different types of use values and between direct and indirect-use values (Pearce 1991). For example, if a forest is harvested extensively for its fauna and flora, this may negatively impact its value for tourism and degrade its ability to protect watersheds. Aggregation issues are especially important when we try to arrive at a number representing the total (measurable) value of tropical forests.

Location dependence

The nature and value of ecosystem services provided by a forest depend not only on the characteristics of the forest itself but also on the number and characteristics of beneficiaries of these services, as illustrated in the following example:

Two identical forests will provide very different water services if one is located in a watershed inhabited by many people and the other in an uninhabited area. Arguably, the services provided in the former case are worth more because they are important to a greater number of people. Differences in income can also affect the relative value of different forest services, to the extent that value reflects people’s ability to pay as well as their willingness to pay (Pagiola et al. 2002; p.18).

248 Monetisation

Monetisation is not necessarily an integral part of the concept of economic value191. Indeed, Rosales et al. (2005) describe how they used household rice sufficiency instead of money to get villagers in Laos to value products they collected from forests. Using prevailing market prices was not feasible since the villagers did not trade these products. However, expressing benefits in terms of monetary values is a useful convention as it allows us to express trade-offs using a single metric, namely money, which in turn facilitates the aggregation and comparison of costs and benefits within a cost-benefit analysis framework (Mullan and Kontoleon 2008).

What TEV does not address

Distribution of value (equity issues)

The concept of TEV, or the cost-benefit analysis framework in which it is used, does not address how the costs and benefits of ecosystem services provided by forests should be distributed. It is important that policy-makers keep this limitation in mind as it is important in state-sponsored conservation, but is even more crucial when designing market-based mechanisms.

Pagiola et al. (2002) point out that the extent to which poor producers are able to participate in market-based mechanisms is closely linked to the resulting market structure. In the case of carbon services and shade-grown coffee, for example, buyers have a wide range of potential suppliers to choose from, and the high transaction costs incurred in dealing with many small, dispersed producers place them at a competitive disadvantage in bidding to supply these services (Pagiola et al. 2002). Therefore, proactive interventions by outside actors are needed if the poor are to participate at all. These include securing land tenure for marginalised groups, supporting cooperative institutions for bundling and bargaining, facilitating access to training and start-up capital, and designing the market itself (Pagiola et al. 2002).

191 Based on how economic value was defined in the previous section.

249 Critical thresholds

TEV typically measures marginal (i.e. small-scale) changes in forest cover and the resulting flows of goods and services (Mullan and Kontoleon 2008). Therefore, while it may be feasible to estimate the welfare impacts of a partial reduction of particular forests, it is much less clear how to assess the welfare impacts of losing all forests in a country. Another reason for only estimating marginal changes, especially those related to biodiversity, is that there may be critical thresholds for the level of diversity, below which an ecosystem no longer functions. In such cases, a large reduction in biodiversity may have discontinuous and unpredictable impacts. Figure 7.4 illustrates how continuous, diminishing marginal values may exist for small changes in the flow of services from a natural resource (e.g. from Supply A to Supply B in the figure), until a critical threshold is reached, past which meaningful economic values cannot be obtained.

Figure 7.4: Valuing marginal changes (from Mullan and Kontoleon 2008; p.12)

250 Intrinsic values

As stated previously, the TEV framework only includes values of biodiversity that can be quantified in monetary terms192. However, these resources may also contain intrinsic values that some individuals—such as O’Neill (1997), described previously—place on biodiversity for cultural, historic or symbolic reasons, which they are unwilling to trade off against other factors. Similarly, other individuals may believe that permitting the loss of biodiversity is inherently wrong, or that humans have a duty to protect natural resources as stewards of the environment (Mullan and Kontoleon 2008). Such intrinsic or ‘moral’ values cannot be compared with marketed goods and services because they do not have a quantifiable effect on human welfare and therefore cannot be valued in monetary terms. Consequently, such non-economic values cannot be incorporated into cost-benefit analysis, but, as O’Neill (1997) pointed out, could potentially be incorporated in decision making processes via alternative means such as participatory and/or deliberative approaches.

Figure 7.5 presents a comprehensive view of the different categories of biodiversity including both economic (measured by TEV) as well as non-economic (not measurable by TEV) values.

192 Or other concrete measures of value such as rice.

251 Economic values Non Economic values Refers to absolute values Refers to relative values that can be assigned a monetary metric that cannot be assigned a monetary metric

Altruistic, Direct use Indirect use Option bequest and Intrinsic values values values values existence values

Output that is Ecological The value of Knowledge of Values for biodiversity enjoyed directly functions that retaining future continued existence that cannot be by consumers support and options, either or that others will expressed in terms of protect economic known of enjoy benefits of trade-offs or a monetary activity unknown forests & biodiversity metric

e.g. Wood, e.g. Flood e.g. Potential e.g. Contributions e.g. Cultural and recreation, control, storm bio-prospecting to environmental religious values NFTPs, etc. protection, values charities pollination, climate stabilisation, etc.

Figure 7.5: The different categories of value of tropical forests (and biodiversity)

Measuring TEV: valuation techniques

Calculating TEV is relatively straightforward for goods and services that are traded on markets, and therefore have prices, i.e. so-called market goods. However, as discussed in section 7.1, many environmental goods and services do not have properly functioning markets. Valuing such ‘non-market’ goods and services, therefore, requires the estimation of consumers’ willingness to pay either through examining their behaviour (by drawing inferences from their demand for related goods, for example), or through their responses to surveys (Tietenberg and Lewis 2012; p.82).

Valuation methods are generally classified into two broad categories: revealed preference methods and stated preference methods (Tietenberg and Lewis 2012; p.82). The former are based on observable consumer choices that allow resource values to be directly inferred from those choices, such as prices or values implied by consumer behaviour. The latter—which are used when preferences are not directly observable— are assessed through surveys, or experiments where subjects are presented with

252 hypothetical situations. Each of the two broad categories of methods includes both indirect and direct techniques (Prato 1998; p.314 - 315).

The available methods within each of the categories above are presented in Table 7.2.

Table 7.2: Economic valuation methods (adapted from Tietenberg and Lewis 2012; p.82)

Methods Revealed preference Stated preference

Direct Market price Contingency valuation Simulated markets

Indirect Travel cost Attribute-based methods, e.g. conjoint Hedonic pricing methods analysis Avoidance expenditures

I now briefly describe these methods as many are used in the economic model I developed to estimate the values of natural resources within Sri Lanka’s southeast landscape (which is described later in the chapter).

Among the revealed preference methods, those using market prices for traded goods are the most straightforward, e.g., the calculation of benefits from agricultural or timber production. Where markets are not fully developed, as in carbon credits for example, simulated market prices could be used. Indirect methods are used where prices are not available, but when human behaviour could be used to infer value. An example is the hedonic price method, which estimates the values of non-market goods by observing people’s behaviour in the markets for related goods. The starting point for this method is the observation that the price of a large number of market goods is a function of a bundle of characteristics (Tisdell 2005; p.167). For example, the price of a particular house is determined by factors such as its structure, the number and size of rooms, its location, the local environment, and nearby amenities (Pearce et al. 2006; p.93). So, by statistically analysing house prices across locations, and controlling for other characteristics, the value people place on factors such as ‘peace and quiet’ could be inferred (Hussen 2004; p.151). In short, the value of peace and quiet is then the premium that is paid for a quieter house compared with a noisier but otherwise identical one. The travel cost method—another indirect, revealed preference method—uses the costs which individuals incur in travelling to outdoor areas as surrogates for prices of

253 their visits. If visitors are drawn from a large catchment area, the cost per visit of those coming from further afield can be expected to be higher than those located closer to the natural area (Tietenberg and Lewis 2012; p.90). This cross-sectional data together with relative frequency of visits can be used to estimate a demand curve for visits to the natural area (Tisdell 2005; p.158). A final example of an indirect observable method involves examining ‘averting or defensive expenditures’, i.e., those designed to reduce the damage caused by pollution by taking some kind of averting or defensive action (Tietenberg and Lewis 2012; p.92). An example would households installing double- glazed windows to decrease exposure to road traffic noise. Essentially, double-glazing is a market good, which, in this example, acts as a substitute for a nonmarket good (peace and quiet in the sense of the absence of road traffic noise) (Pearce et al. 2006; p.99). Since people would not normally spend more to prevent a problem than would be caused by the problem itself, averting expenditures can provide a lower-bound estimate of the damage caused by pollution (Prato 1998; p.315).

Stated preference approaches are survey-based and aim to elicit people’s intended future behaviour in constructed markets. Among these, the primary direct valuation technique is contingency valuation. The simplest version of this approach involves asking respondents what value they would place on an environmental change (such as the loss of a wetlands or increased exposure to pollution) or on preserving the resource in its current state (Tietenberg and Lewis 2012; p.83). Alternative versions ask a ‘yes’ or ‘no’ question such as whether or not the respondent would pay $X to prevent the change or preserve the species. The answers reveal either an upper bound (in the case of a ‘no’ answer) or a lower bound (in the case of a ‘yes’ answer) (Prato 1998; p.322 - 323). This survey approach creates a hypothetical market and asks respondents to consider a willingness-to-pay question contingent on the existence of this market. The major concern with the use of the contingent valuation method has been the potential for survey respondents to give biased answers; the five main types of bias are presented in Table 7.3 below:

254 Table 7.3: Sources of bias in contingency valuation studies (adapted from Tisdell 2005; p.163 - 165; Tietenberg and Lewis 2012; p.83 - 84)

Potential source of bias Explanation/ details

Strategic bias Refers to the possibility that respondents will vary their answer in an attempt to influence the occurrence of their desired outcome. For example, if a respondent strongly desires the establishment of a national park in an area he/she may exaggerate the sum that he or she would be prepared to pay for its establishment. Information bias May arise from the way in which the alternatives are presented to respondents. For example, if individuals have not visited or seen a natural area that is being considered for protection, they may be largely dependent on the interviewer’s description of the area for their knowledge. If this is not accurate, bias may occur. Starting-point bias May arise in those survey instruments in which a respondent is asked to check off his or her answers from a predefined range of possibilities, i.e., how that range is defined by the designer of the survey may affect the resulting answers. A range of $0–$100 may produce a valuation by respondents different from, for example, a range of $10–$100, even if no bids are in the $0–$10 range. Hypothetical bias This type of bias could occur because the respondent is being confronted by a contrived, rather than an actual, set of choices. Since he or she will not actually have to pay the estimated value, the respondent may treat the survey casually, providing ill- considered answers. The observed discrepancy Respondents to contingent valuation surveys tend to report much between willingness to pay higher values when asked for their willingness to accept and willingness to accept compensation for a specified loss of some good or service than if asked for their willingness to pay for a specified increase of that same good or service. Economic theory suggests the two should be equal.

The potential for these types of biases to affect the results from contingency valuation must be considered when interpreting study results, as I have done in section 7.5, where

255 I discuss the output of my economic model of the Sri Lanka’s southeast landscape193. In spite of these potential issues with contingency valuation, this method has been extensively applied to the valuation of environmental impacts both in developed and developing countries (Pearce et al. 2006; p.106). Moreover, a panel of independent economic experts (including two Nobel Prize laureates) convened by the US National Oceanic and Atmospheric Administration to evaluate the use of contingent valuation methods in policy and judicial processes determined that “suitably designed surveys could eliminate or reduce these biases to acceptable levels” (Tietenberg and Lewis 2012; p.84 - 85).

While contingency valuation is a direct method, there are several indirect stated preference methods that include several attribute-based methods. Attribute-based methods such as conjoint models (or, equivalently, choice experiments) are useful when project options have multiple levels of different attributes (Tietenberg and Lewis 2012; p.87). Like contingent valuation, choice experiments are also survey based, but instead of asking respondents to state a willingness to pay, respondents are asked to choose among alternate bundles of goods. Each bundle has a set of attributes and the levels of each attribute vary across bundles. Since one of the attributes in each bundle is a price measure, willingness to pay can be identified (Tietenberg and Lewis 2012; p.87).

Given the time and cost of conducting original studies using both stated and revealed preference methods, there is the increasing use of ‘benefits transfer’ methods to adapt the results of original studies to other situations. Pearce et al. (2006; p.254) describe this process as “taking a unit value of a non-market good estimated in an original or primary study and using this estimate (perhaps after some adjustment) to value benefits that arise when a new policy is implemented”194.

I use many of the methods described above in my economic model in order to estimate values of the various TEV components related to the natural resources found within Sri Lanka’s southeast landscape (this information is summarised in Table 7.6 in section 7.5).

193 The results from a contingency valuation study are used in my model to estimate bequest and existence values for elephants. 194 I discuss the practical application of the benefit transfer method, in section 7.5, where I use it to estimate the value of non-timber forest products in my study area.

256 7.4. The components of TEV

Having outlined the concept of TEV, and discussed the various methods that could be used to estimate it, I now examine each of the components comprising the TEV of tropical forests195, focusing my discussion on elements relevant to Sri Lanka’s southeast landscape. This exercise provides a link between the theoretical concept of TEV to its practical application within my study area that I will describe I detail in section 7.5.

Direct use values

As described in Table 7.1, direct uses of forests include consumptive uses such as timber, non-timber forest products and hunting; productive uses such as plant breeding and genetics; and non-consumptive uses such as tourism.

I do not include timber in my analysis, as a direct use of forest biodiversity, because, while it is possible to harvest timber without destroying forest biodiversity, more often, a trade-off must be made between timber extraction and biodiversity conservation (Mullan and Kontoleon 2008). Moreover, timber extraction at the commercial scale is not legally permitted within the forests in southeast Sri Lanka.

I will exclude hunting for similar reasons. Hunting is generally considered destructive to forest biodiversity as illustrated by the ‘bush meat crisis’ in Africa and parts of Asia (Robinson et al. 1999). Although some conservationists—especially those in Southern Africa—recommend ‘trophy hunting’ as an integral value of conservation, this is not feasible in Sri Lanka, given local religious and cultural values196 (Bothma et al. 2009).

Finally, I will exclude bio-prospecting, given the probably low potential for this source of economic value within my study area—most studies concerning bio-prospecting conducted to date focus on Sri Lanka’s wet-zone, where there are a higher proportion of endemic plants. However, as the current lack of information does not definitely preclude future discoveries within the landscape, I have provided information on bio- prospecting in tropical forests in Appendix 7.

195 I consider in detail each of the elements described in Table 7.1. 196 This would also be the case in several countries in the region, notably India.

257 I now consider in detail two types of direct uses that are especially relevant to my case study context: non-timber forest products and tourism.

Non-timber forest products

Forests provide people a range of products other than timber, including food and forage for livestock; fibre for clothing and household objects; fuel for heating and cooking; and medicinal or cultural products (Mullan and Kontoleon 2008). These non-timber forest products (NTFPs) are often harvested on a small scale from the wild for direct subsistence use, although they may also be extracted on a larger scale or cultivated for commercial purposes (Mullan and Kontoleon 2008). Pearce (1991; p.247) describes how NTFPs could be an important source of revenue in Indonesia:

In Indonesia for example, exports of NTFPs rose from US$ 17 million in 1973 to US$ 154 million in 1985, comprising 12% of forest export earnings. They rose to US$ 238 million in 1987. Exports of rattan alone were US$ 80 million in 1985. Tropical forests also supply some essential oils such as camphor, cinnamon, clove and nutmeg, a trade worth some US$ 1 billion per annum, although most of the output of these products come from plantations. However, Pearce (1991) points out that the collection of these products could cause environmental damage. Extrapolating from a small area to a larger region (or to an entire country as in his example), furthermore, could be problematic since the markets for these products are unlikely to be very large, and as production expands, their prices are likely to fall (1991). Therefore, sustainable harvesting of NTFPs is crucial to maintaining their value in the long term.

NTFP values are usually calculated through surveys of households involved in their collection (Mullan and Kontoleon 2008). These surveys ask about quantities of one or more products harvested by the household over a given time period, for subsistence use and/or sale. Where products are marketed, a financial value can be obtained directly, although harvest costs should be estimated to derive net values. Where products are not marketed, or a local market does not exist, values may be imputed based on a close substitute, for example other fuels that could be used in place of fuel wood (Mullan and Kontoleon 2008). As was pointed out in section 7.3, household consumption of rice has been used for this purpose in a study in Laos.

258 Tourism

Ecotourism constitutes a potentially valuable non-extractive use of tropical rainforests (Pearce 2001). Ecotourism is a subset under nature-based tourism, the fastest growing segment of tourism today—it has been growing at annual rates of 10% – 12%, i.e., 3 times faster than the tourism industry as a whole (Coria and Calfucura 2012; p.47). Approximately 20 – 25% of leisure travel is classified as nature-based tourism (Parks et al. 2009).

However, ecotourism is a somewhat vague term, and as Isaacs (2000; p.61) points out, “it has been the proclaimed purpose of widely disparate enterprises, from low-impact activities like hiking wetland trails to energy-intensive activities as tourist submarines and underwater hotels”. Therefore, it is important to define clearly what we mean by ecotourism. Krüger (2005; p.580) proposes the following definition:

Travelling to relatively undisturbed or uncontaminated natural areas with the specific objective of studying, admiring, and enjoying the scenery and its wild plants and animals, as well as any existing cultural manifestations (both past and present) found in these areas. As ecotourism has grown in popularity, its definition has been expanded to incorporate ideas about ecotourism responsibility, environmentally-friendly destination management, and sustainable local development (Coria and Calfucura 2012; p.47). Stem et al. (2003; p.388) point out that, unlike many sustainable harvesting initiatives, ecotourism could consistently provide a return per hectare competitive with current land uses. Its numerous benefits include: long-term sustainable resource conservation, a more socially and environmentally acceptable approach to development, empowerment of local people, increased employment, the fostering of greater public and private cooperation, and opportunities for scientific and educational research (Parks et al. 2009).

Krüger (2005) analysed case studies of ecotourism projects to determine key success factors. An important finding was that ecotourism projects with no flagship species had an especially low proportion of sustainable case studies, while this proportion was especially high in projects with a charismatic bird (80%), charismatic mammal (81.3%) or a world-wide flagship species (72.2%).

259 Despite its potential benefits, ecotourism has been criticised as being good in theory, but often harmful in practice to nature or the cultural aspects it is supposed to sustain. For example Parks et al. (2009) complain that:

• Most countries lack standards for what can be called ‘eco’, and lack legislation to protect sites from overuse or degradation.

• The increased number of visitors often exceeds carrying capacities of ecotourism sites; tourists often damage or destroy natural areas.

• Equitable involvement of local people in planning, executing, and managing ecotourism projects is rarely accomplished.

In spite of these criticisms, Isaacs (2000; p.61) emphasises that the existence of negative externalities does not mean that ecotourism is necessarily detrimental. Indeed, ecotourism is less likely to damage the environment than alternative industries like agriculture, mining, timber, and urban development. Nevertheless, potential negative externalities must be recognized in assessing and designing ecotourism programs.

Indirect use values

Indirect uses of tropical forests include ecosystem services such as watershed protection, nutrient and carbon cycling, carbon-capture (to mitigate global warming), biodiversity conservation and air pollution reduction. Pagiola et al. (2002) describe three main types of services most often targeted by market-based mechanisms (and hence which are valued): watershed protection, biodiversity conservation and carbon sequestration. I now take each of these in turn.

Watershed protection

The loss or degradation of forest cover could negatively impact watershed functions, causing changes in water flow regulation, resulting in flooding or storm damage; increased soil erosion, leading to siltation and sedimentation of rivers and reservoirs; and loss of nutrients in agricultural soils (Mullan and Kontoleon 2008). Such impacts

260 could have economic consequences, such as damage to agricultural land or residential property due to flooding, reduced agricultural productivity, increased water treatment costs or loss of storage capacity in reservoirs, and damage to equipment used in hydroelectric facilities (Mullan and Kontoleon 2008). Pagiola et al. (2002; p.19) point out, however, that it is often difficult to establish a clear link between preserving forests and the resulting benefits, illustrating their point using the example of erosion:

In practice, establishing clear relationships between forest cover and erosion is extremely difficult. The clearest evidence concerns the role of forests in reducing sheet erosion. Research on the determinants of erosion suggests that forests are less important than other factors such as understory vegetation, soil composition, climate, raindrop size, terrain, and slope gradient. Although this issue could hamper the establishment of market-based mechanisms to conserve forests, Pagiola et al. (2002) present a number of examples where they have been successfully addressed, including Costa Rica’s system of ‘Payments for Environmental Services’ and the watershed management system in Sukhomajri, India.

Biodiversity conservation

As mentioned in Chapter 1, tropical forests contain significant biodiversity resources. Although biodiversity conservation typically involves providing suitable habitats for native plants and animals, it may also involve efforts to eradicate alien (exotic) species (Pagiola et al. 2002). Traditionally, this involved the state setting up and managing PAs or national parks, but, as pointed out previously, developing country governments often lack the required financial resources to do so. Consequently, there has been a proliferation of private reserves in recent years (Bothma et al. 2009). An example is Chile, where not only have individuals formed their own PAs, but commercial enterprises have entered the market to assist them (Pagiola et al. 2002). Schemes in Mexico and El Salvador to market shade-grown coffee are other examples of where the private sector has been willing to pay for biodiversity conservation (Pagiola et al. 2002). These schemes seek to capitalise on consumers’ willingness to pay for biodiversity conservation by inducing them to pay a premium for biodiversity-friendly cultivation practices.

261 Carbon sequestration

As pointed out before, forest ecosystems are a major carbon sink. It has been calculated that the world’s forest ecosystems—4 billion ha, covering 30% of the earth’s land area—hold more than double the amount of carbon contained in the atmosphere (Canadell and Raupach 2008). Typically, forests contain 20 to 100 times more carbon per unit of area than agricultural lands (Pearce 1991). As defined by the Intergovernmental Panel on Climate Change (IPCC), deforestation produces around 17% of global emissions, making it the third largest source of greenhouse gas emissions, larger than the entire global transport sector (Eliasch 2008; Dudley et al. 2010; p.31).

Overall, annual forest emissions are comparable to the total annual CO2 emissions of the US or China (Eliasch 2008). Forest-based carbon sequestration is based on two main approaches: active absorption in new vegetation, and avoided emissions from existing vegetation. The first approach includes any activity that involves planting new trees, such as afforestation, reforestation or agroforestry. The second approach involves preventing or reducing deforestation and land-use change, or reducing damage to existing forests (Pagiola et al. 2002).

In principle, it should be easier to develop markets for forest carbon sequestration services than for watershed protection or biodiversity conservation services because the value of this service is the same everywhere—a ton of carbon sequestered in one place makes much the same contribution to reducing climate change as a ton of carbon sequestered anywhere else (Pagiola et al. 2002). Moreover, measurement of carbon contained in plant biomass is simpler than linking land-use changes to hydrological functions or to biological diversity. Consequently, it is easier to account for increases or decreases in carbon storage, making it more convenient to monitor and trade emission reductions. Finally, estimates of carbon sequestration costs through forestry suggest that this method is much cheaper than most others used to tackle climate change, especially those aimed at reducing fossil fuel-based emissions. Given all these reasons, there have been significant efforts to establish a global carbon-trading framework.

The main instrument in current international climate change mitigation framework is the Kyoto Protocol of the United Nations Framework Convention on Climate Change (UNFCCC) that commits developed countries to legally binding targets to limit or reduce their greenhouse gas emissions, and provides mechanisms to allow emissions

262 trading (Eliasch 2008; p.101). This is done by allowing developed countries exceeding their emission targets to offset these excess emissions by investing in emission-reducing projects in developing countries. The key mechanisms of the international climate change mitigation framework are discussed in Appendix 7.

It is important to note that reduced emissions from deforestation and degradation (REDD) are currently not allowed under the Kyoto protocol—only afforestation and reforestation activities can generate carbon credits (Venter and Koh 2012; p.138). Consequently, there have been several rounds of painstaking, often contentious, negotiations to establish a global regime for implementing REDD, which is crucial to incentivising developing countries to reduce deforestation (details are provided in Appendix 7). Issues around additionality, permanence and leakage are major problems that need to be resolved:

• Additionality is the risk that payments for reduced emissions might be provided for reductions that would have occurred even without payments, i.e. the risk that REDD would not generate additional emissions reduction (Venter and Koh 2012; p.141). Therefore, a baseline emissions rate should be established for a given region or country, and adequate monitoring conducted to ensure that emissions reductions are motivated by the payments.

• Permanence is the risk that higher levels of deforestation and degradation in the future may offset current emissions reductions (Venter and Koh 2012; p.142). Therefore, terrestrial emissions should be continually monitored to ensure that payments are made only for permanent emissions reductions.

• Leakage is the risk that emissions reduced in one location might simply be shifted to another location through higher levels of deforestation in that other location (Venter and Koh 2012; p.142). Again, reliable global monitoring is necessary to ensure that emissions are not transferred spatially.

Some of these issues may be resolved through technology. For example, establishing a baseline rate requires data on current and past rates of forest loss, as well as field inventories to estimate how much carbon is stored in forests (Laurance 2008). Remote sensing is believed to be the most feasible way to gather these data, but developing nations would require financial and technical support to do this (Laurance 2008).

263 Although an international agreement on the broad contours of REDD+197 was reached in Cancun in 2010, a global unified financing mechanism seems unlikely given the overall lack of progress of climate-related talks and legislative setbacks in several major developed countries such as the US, Canada and Australia (Agrawal et al. 2011; p.374). The 2010 Cancun agreement describes three phases of REDD+ implementation. In the first phase, governments develop the institutional capacity to design REDD+ strategies through stakeholder dialogue, institutional strengthening, and demonstration activities; most developing nations participating in REDD+ are in phase 1 (Agrawal et al. 2011; p.380). In the second phase, developing country governments access REDD+ finance on the basis of results that are not necessarily tied to emissions reductions. Brazil, Guyana, Indonesia, and Tanzania are currently engaged in this type of funding, all from the Government of Norway (Agrawal et al. 2011; p.381). In the third phase, governments are ready to participate in performance-based funding, where REDD+ revenues are directly tied to verified reductions in greenhouse gas emissions from deforestation and forest degradation; no nation has reached phase 3 (Agrawal et al. 2011; p.381).

Despite the lack of a comprehensive, internationally agreed REDD+ financing mechanism, several international and national actors are already involved in helping developing nations implement REDD. The World Bank’s Forest Carbon Partnership Facility (FCPF) plans to use its US$ 180 million pledges by donor nations to support 37 countries in phase 1 activities, and aims to provide a further US$ 100 million to a smaller group of countries for phase 2 activities (Agrawal et al. 2011; p.381). The United Nations Collaborative Programme on Reducing Emissions from Deforestation and Forest Degradation in Developing Countries (UN-REDD)—that is providing support to 12 nations for phase 1 activities—is a partnership of the UN Environment Programme, the UN Development Programme, and the Food and Agriculture Organization (Agrawal et al. 2011; p.381). Finally, the ‘Paris-Oslo Partnership’ or the ‘Interim REDD+ Partnership’ is a multilateral process involving 58 donor nations and developing countries operating outside the UNFCCC ambit (Agrawal et al. 2011; p.381). The government of Norway is a global leader in REDD+ funding. In addition to committing funds to all three of the funding mechanisms discussed above, it has designed large-scale performance-based bilateral grants to slow deforestation. In 2008,

197 REDD+ contains some additional provisions around the sustainable management of forests. These are described in Appendix 7.

264 it made a US$ 1 billion commitment to slow deforestation in the Amazon, which will be disbursed through 2015 as Brazil demonstrates progress toward its target of reducing deforestation in the Brazilian Amazon by 70% by 2017 (Agrawal et al. 2011; p.381). Norway has also signed similar agreements with Indonesia and Guyana (Agrawal et al. 2011; p.381).

Non-use values

Individuals may place value on forests that they will never use, either because they value the knowledge that others elsewhere can use the resource, or because they gain satisfaction from the continuing existence of forest biodiversity in general, or of specific species (Mullan and Kontoleon 2008). This is generally known as ‘existence value’. Economists sometimes identify another non-use motivation where individuals place a value of conserving forests for future generations, terming this ‘bequest value’ (Ninan 2007). In practice however, it is extremely difficult to separate these two categories of value (Mullan and Kontoleon 2008).

Bandara and Tisdell (2003), who assessed the relative importance of the economic values of elephants in Sri Lanka, found that the majority of interviewees gave most weight to non-use values relating to elephants. Other researchers have found similar results regarding a range of other species. For example, Langford et al. (2001) claim that people mostly choose to pay for conservation of an endangered species to secure its existence primarily because it ensures a variety of subsidiary benefits for themselves and for future generations.

Having introduced the concept of TEV, and discussed its components in detail, I now use this concept to estimate economic values of various land-use types within Sri Lanka’s southeast landscape.

265 7.5. Estimating the TEV of natural resources within Sri Lanka’s southeast landscape

As discussed in section 7.1, valuation helps society make informed decisions on trade- offs among alternative land-use options. This approach is especially important for policy-makers, for whom the economic ramifications of a decision are extremely important, although they also consider social and political factors. Therefore, I now use a TEV framework to evaluate current land-use in Sri Lanka’s southeast landscape from an economic viewpoint, and understand how potential future land-use approaches— both environmentally friendly types incorporating MERs and those emphasising development with minimal environmental considerations—would impact economic value. I follow a three step process: first identifying TEV components that are relevant to the landscape, next building a model to estimate values of these components, and finally using the model to develop several land-use scenarios (see Figure 7.6).

STEP 1 STEP 2 STEP 3

Identify components Build model Develop scenarios

Estimates (USD million/ • Examine Component year) economic activities Costs Benefits Scenario Scenario Scenario and livelihoods A B C Direct values within landscape (x, y, z…)

• Determine TEV Indirect values (x, components that y, z…)

are relevant to Bequest & landscape existence values (x, y)

Total

Figure 7.6: Process for estimating the TEV of land-use for Sri Lanka’s southeast landscape

266 Step 1: Identifying relevant TEV components within landscape

In order to identify the TEV components relevant to Sri Lanka’s southeast landscape we need to revisit the types of land-use within the landscape discussed in Chapter 5. For convenience, I have reproduced the relevant table, renaming it Table 7.4.

Table 7.4: Land-use within my study area in southeast Sri Lanka (calculated from GIS data obtained from Survey Department of Sri Lanka)

Land-use types Area (km2) % of total Paddy 491 8% Chena 543 8% Homesteads/ gardens 726 11% Other cultivation, e.g. coconut, fruit trees 176 3% Forest 2,770 43% Scrub and grassland 1,340 21% Reservoirs and irrigation channels 214 3% Other water bodies (e.g. streams, water holes, lagoons) 94 1% Other natural (marsh, rocks, sand, salt pans) 105 2% Other human (e.g. built up areas, quarries, unclassified) 9 0% Total 6,467 100%

As discussed in Chapter 5, natural features such as forest, scrub, grasslands, rocks, sand and natural water bodies cover 67% of the landscape, given the number of PAs in the area, as well as forested land outside PAs198. The other major type of land-use is agriculture—paddy, chena, other crops and the reservoirs and irrigation channels used to provide water to them, cover 22% of the land area. The remaining area (11% of the total) is devoted to human habitation, although many people grow fruit trees and vegetables in their home gardens.

Keeping this pattern of land-use in mind, I now discuss relevant components of TEV within the context of the broad categories presented in Figure 7.5.

198 Managed by the Forest Department.

267 Direct use values

The most obvious direct use value is agriculture. As described in Chapter 5, paddy and chena farming are the main livelihoods of local people. Chena farmers produce most of the vegetables within the region, while both categories of farmers grow fruits and vegetables in their home gardens, selling the produce that is left over after domestic consumption.

In addition, people utilise the various natural resources within the landscape to obtain non-timber forest products. The livelihood survey described in Chapters 5 and 6 revealed that some people supplement their diets and incomes by fishing. Others gather materials for constructing their houses and for indigenous medicines. However, there is no significant timber industry in the area since there has been a ban on logging in natural forests in Sri Lanka since 1990 (Ministry of Environment Sri Lanka 2012).

Tourism is the other major component of direct-use value. As described in Chapter 5, the landscape contains some of Sri Lanka’s most visited parks, which together account for 65% of the country’s total PA revenue.

Indirect use values

Natural resources within the landscape—such as forests, grasslands, marshes and rivers—provide many valuable ecosystem services to local people including soil conservation, nutrient recycling, and watershed services such as flood attenuation, sediment control and water filtration199 (World Bank 2009; p.5). In addition to these largely local services, forested areas provide the global ecosystem service of carbon sequestration.

199 These are in addition to the more obvious services such as providing water for crops and opportunities to fish, which were covered under direct-use values.

268 Bequest and existence values

Many of the PAs in the landscape, and the flora and fauna they contain, are considered ‘national treasures’ (Ministry of Environment and Natural Resources 2003). Moreover, as noted in Chapter 5, about 30% of the country’s population of elephants—Sri Lanka’s national icon—live within the landscape. Therefore, much of the area’s natural resources would be of significant ‘bequest’ and ‘existence value’ to people across the country.

Having identified the TEV components relevant for the economic analysis, I now put these together in an economic model.

Step 2: Building an economic model of land-use

The overall structure of my economic model is based on the cost-benefit analysis framework described at the end of section 7.1. I estimated costs and benefits associated with each of the TEV components identified above, and added them up to arrive at an estimate of the total revenues and costs associated with the landscape’s current pattern of land-use. Economic profit is simply total revenues minus total costs. I developed my model in an excel workbook for flexibility and ease of use. This functionality was especially important for the next step in the process, i.e. constructing scenarios of land- use (as will be explained in more detail in the next section). Further details about the structure of the model are provided in Appendix 7.

I now describe how TEV components were calculated in my model, taking each by turn.

Agriculture

Revenue from rice production was calculated using data for the 2009/2010 growing season. Paddy yields for Ampara, Hambantota and Moneragala districts—obtained from the Department of Census and Statistics—were averaged, resulting in 4,988 and 4,734 kg per ha for the Yala and Maha seasons respectively. These were multiplied by the

269 extent of paddy-land in the case study area (from Table 7.2) to estimate paddy harvest200. Revenue was calculated using 2009 rice wholesale prices201, but since middlemen take a percentage, farmers do not receive the full amount; this percentage was assumed to be 15%, based on my interviews with farmers (described in Chapter 6).

‘Chena crops’ were selected based on the top six ‘seasonal’ crops202 grown in the districts of Hambantota, Moneragala and Ampara. These crops—maize, green gram, cowpea, ground nuts, manioc, ginger—accounted for 86% of the seasonal crops grown in the three districts, by extent of land cultivated in the 2009/2010 growing season. The proportions of each crop cultivated during that growing season (by extent) were distributed across the case study area’s chena land (as per Table 7.2). Crop yields, prices and revenues were calculated in a similar manner to that for rice.

Finally, I selected the crops grown in the land-use categories of ‘other cultivation’ and ‘homesteads/ gardens’ based on the top four ‘permanent and semi-permanent’ crops grown in the relevant District Secretariat Divisions203. These crops—coconut, bananas, lime and mangoes—accounted for 87% of the non-rice crops grown in the landscape, by extent of land cultivated, in the 2009/2010 growing season. Extents cultivated by crop were apportioned in a similar manner to that of chena crops204, and crop yields, prices and revenues were calculated by the same methods used for rice and chena crops.

Given difficulties in obtaining reliable cost data at the required level of detail205, and because costs can change drastically among regions and growing seasons due to changes in weather and pest outbreaks, costs were estimated using cost-income ratios of 80% for rice and 85% for other crops. The latter ratio is higher because chena farmers

200 This gives us the Maha harvest. Less rice is sown in the Yala season, so the ratio between extents sown in Yala and Maha for the 2009/2010 growing season were adjusted to account for this. 201 The wholesale price is the price the middlemen—such as rice-mill owners—sell the produce to retailers in cities and towns. 202 The Department of Census and Statistics classifies virtually all the crops grown in chenas—which were introduced in Chapter 5—within this category. Other crops such as coconut and fruits that are grown in home gardens are categorised under ‘permanent and semi-permanent crops’. Although some seasonal crops may also be grown in gardens as well, for simplicity, I put all of these under chenas, and the ‘permanent and semi-permanent crops’ under home gardens and ‘other crops’ (in terms of the land use categories in Table 7.2). 203 The way data were presented by the Department of Census and Statistics allowed a more precise calculation than for chena crops (where I had to use broader district-level data). 204 The available land for cultivation within homesteads/ gardens was estimated to be 50% of the total land in this category, since a part of the land would be used to build people’s homes (and not all people have extensive home gardens). 205 I.e. costs of cultivation per ha by crop.

270 are not eligible for government assistance such as fertiliser subsidies. Since rice can actually be loss-making in some seasons and marginally profitable in others (Samaratunga et al. 2001; p.8), these cost figures are probably an underestimate (and the resulting profit figures an overestimate). It is also important to take into account that the Sri Lankan government provides fertiliser to rice farmers at heavily subsidised rates, as indicated in Table 7.5.

Table 7.5: Fertiliser subsidies provided by the Sri Lankan government, 2009206 (adapted from Tibbotuwawa 2010; p.1)

Market price Subsidised price Subsidy per 50 Subsidy as % of Fertiliser per 50 kg per 50 kg (Rs.) kg (Rs.) market price Urea 5,300 350 4,950 93% Muriate of 7,200 350 6,850 95% Potash (MOP) Triple Super 6,300 350 5,950 94% Phosphate (TSP)

In 2009, these subsidies cost the Sri Lankan government US$ 234 million, comprising 2.24% of total government expenditures (Weerahewa et al. 2010; p.8). The fertiliser subsidy per ha in my economic model was estimated using figures from the Central Bank of Sri Lanka for the years 2008 – 2010 (Central Bank of Sri Lanka 2011; p.137). The subsidy per ha was calculated by dividing the subsidy allocated for paddy for each year, by the number of hectares of paddy cultivated that year. An average of the figures for the three years was used to estimate the total fertiliser subsidy for the 49,411 ha of paddy cultivated within my study area—over US$ 8 million.

The above fertiliser subsidies alone do not capture the full extent of government subsidies for paddy cultivation. Weerahewa et al. (2010; p.2) describe some of the other types of state assistance provided to farmers:

Governments have made longer-term investments with the objective of further improving the paddy sector. Some of the policies, programs, and strategies for enhancing paddy production have included large-scale irrigation projects coupled with land development and settlement schemes, free provision of irrigation water, provision of concessionary credit (and the writing off of previously obtained credit), extension services, seeds at

206 Urea, MOP and TSP are the most widely used fertilisers in rice cultivation.

271 concessionary rates, and guaranteed output and input prices. Of these policies, the fertilizer subsidy is the longest-lasting, the most expensive, and the most politically sensitive. I did not attempt to calculate the costs of all these programs due to difficulties in obtaining the required data. Therefore, my model underestimates agricultural production costs, especially those related to rice; consequently it overestimates agricultural profits.

Non-timber forest products (NTFPs) As I did not have the time and resources to gather primary data to conduct a detailed analysis of benefits from non-timber forest products, I used the benefit transfer method to estimate these benefits. Kareiva et al. (2011; p.36) describe this method in more detail:

Detailed non-market valuation studies appropriate to a particular service in a particular place tend to be time and resource intensive, limiting the applicability to broad-scale assessments. Benefit transfer provides a less time-intensive approach than production functions for generating broad- scale monetary estimates of ecosystem services. The most common application of the benefit transfer approach uses estimates of the value of services per unit area from a single or small number of locations and applies this value to other locations with the same ecosystem type. This approach relies on existing estimates and does not require any additional analysis, which is a distinct advantage if decisions are imminent and primary data collection is not feasible. One of main issues with this approach is whether the original studies from which values are ‘transferred’ are applicable to the situation at hand (Kareiva et al. 2011; p.36). To minimise this source of error, I used two studies from Sri Lanka—a study of the Muthurajawela Marsh by Emerton and Kekulandala (2003) and one on the value of mangroves in several parts of the country by Batagoda (2003). The values for specific services relevant to southeast Sri Lanka—such as fishing, gathering fuel-wood and obtaining basic construction materials—were used to estimate the per ha value of these products. These were then multiplied by the extent of forested areas accessible to locals within the landscape, i.e. those outside PAs, to arrive at an estimated benefit from non- timber forest products of approximately US$ 4 million an year.

272 Nature tourism

The number of tourists visiting the PAs in the study area was calculated based on actual 2010 numbers, obtained from the Department of Wildlife207. What tourists spent was estimated using data from a recent World Bank survey (World Bank 2009). This data shows that tourists who visit parks, but are not part of package tour, have a higher average daily expenditure than those who are part of package tours208 (World Bank 2009; p.12).

However, the World Bank study also pointed out that these expenditures have a ‘multiplier effect’ on the Sri Lankan economy (World Bank 2009; p.12):

When a tourist spends, say US$ 500, on hotels and restaurants, this generates a certain amount of wage income and net surplus that accrues to households, who then spend this income by consuming goods and services. In addition, when the hotel and restaurant sector produces US$ 500 worth of output, it must purchase inputs of food, beverages, water, electricity, communications, manufactures, and so on. These inputs to the sector are either imported or produced by other sectors in the Sri Lankan economy. When all transactions are added up it is possible to arrive at a total measure of the direct plus indirect production, value added, wages, tax revenues, and imports required to meet this demand. Estimated multipliers209 were used to calculate the total economic benefits, i.e. revenues, from foreign tourists within the landscape—US$ 7 million an year. In addition, there would be also be around US$ 2.4 million from existing high-end operators210, bringing the total economic impact from nature tourism to close to US$ 10 million an year. The revenue to the government was estimated by multiplying the number of tourists by an estimated average number of visits to the park of three game drives, and the entrance fee for foreigners of US$ 25 per visit. This number was added to the actual revenue from local park visitors for 2010211, bringing total government receipts to over US$ 2 million an year212.

207 The numbers of tourists that visited the relevant parks in 2010 were increased to reflect the increased numbers of tourists visiting Sri Lanka since the end of the civil war. 208 The daily expenditure of the two groups are US$ 73 and 41 respectively. 209 1.45 and 1.14 for package and non-package tourists respectively (World Bank 2009; p.14-15). 210 I will describe how I calculated the economic impact of high-end tourists in the next section, where I explain Scenairio B, the ‘environmentally-friendly land-use option’. 211 Obtained from the Department of Wildlife. 212 This number is a reasonable estimate since the Department of Wildlife obtained US$ 1.5 million from these parks in 2010.

273 To calculate costs associated with this revenue, amounts spent by the government to maintain parks, mitigate human-elephant conflict and pay compensation to people affected by the conflict, were added to crop damage caused by wild animals and amounts spent by farmers to protect their crops213. The resulting total cost was about US$ 5 million an year.

Carbon sequestration The first step in calculating the value of carbon sequestered within the landscape was to evaluate the area’s stock of carbon using estimated weights of above-ground and below- ground biomass214 for the vegetation types in Table 7.2, from Mattsson et al. (2012; p.33) and Chokkalingam and Vanniarachchy (2011; p.8). Multiplying these values by the extent in hectares of each vegetation type resulted in the total biomass of that vegetation type. These figures were then divided by 2.1267 to estimate the amount of carbon, and multiplied by 3.667 to obtain the equivalent amount of tons of CO2 (Anke et al. 2008; p.53; Chokkalingam and Vanniarachchy 2011; p.8). These calculations resulted in estimates of 45.1 and 43.4 million tons equivalent of CO2 for areas within and outside PAs respectively, resulting in a total of 88.5 million tons equivalent of CO2 for the entire landscape.

The next step was to estimate the annual release of CO2 from these stocks. An annual rate of deforestation of 1.5%, based on Mattsson et al. (2012; p.33), was applied to the 215 forested areas outside PAs , resulting in annual emissions of 630,000 tons of CO2. This is the amount that would be sequestered if those areas were converted to MERs.

The final step was to estimate the value of these avoided emissions. As pointed out in section 7.4 there is still no global regime that pays for carbon sequestration. Moreover, there have been issues with carbon pricing in the European Union’s Emissions Trading System; prices that reached a high of US$ 45 per ton of carbon sequestered, are now closer to US$ 15 (Economist 2012). Therefore, the value of carbon sequestered was

213 Data from a variety of reports and articles such as Fernando et al. (2011) and World Bank (2009) were used. 214 In tons per ha. 215 We are assuming zero deforestation within PAs—although there is some encroachment into certain PAs, e.g. Lunugamvehera, this can be assumed to be negligible from a landscape point of view, at least from the perspective of carbon sequestration.

274 calculated based on price points of US$ 15, 20 and 30, resulting in values of US$ 9, 12 and 18 million respectively. The middle value was used for the final analysis.

It is worth pointing out that the carbon stocks within parks were not included in the valuation because there is currently no process for rewarding countries for carbon sequestration within existing PAs, mainly because of issues of ‘additionality’ 216. However, during negotiations at the UNFCCC, the Sri Lankan government suggested that any future REDD+ mechanisms should include a one-time payment for ‘baseline stocks’ as an incentive to keep those important sinks intact (Mattsson et al. 2012; p.31).

The model also estimates some of the potential costs of obtaining payments for carbon sequestered. The main cost would be the research required to establish the ‘baseline’. Currently the UN-REDD program is supporting Sri Lanka with a grant of US$ 4 million for its ‘initial readiness activities’, so this amount is not included in the costs to the government (Mattsson et al. 2012; p.31). In addition, there would be monitoring costs to assure donors that Sri Lanka is abiding by its emission reduction commitments.

Watershed benefits

The watershed benefits from the rivers within the landscape were obtained from a World Bank study that valued environmental services within Sri Lanka’s southern province (World Bank 2009). The relevant estimates for the value of water for domestic and industrial purposes, and the provisioning, regulating and supporting services of the river basins within the selected landscape were used217. The value of water and the other services were US$ 5 million and US$ 24.5 million respectively, resulting in total annual watershed benefits of about US$ 29.5 million (World Bank 2009; p.34-35).

216 As pointed out in section 7.4, additionality is the risk that payments for reduced emissions might be provided for reductions that would have occurred even without payments, i.e. the REDD+ will not generate additional emissions reduction. 217 These were Karagan Oya, Malala Oya, Kirindi Oya, Menik Ganga and Bambawe Aru.

275 Bequest and existence values

A study by Bandara and Tisdell (2004) was used to estimate bequest and existence values within the landscape. The authors conducted a contingent valuation218 survey of willingness to pay for elephant conservation of a sample of urban residents living in three selected housing schemes in Colombo, and extrapolated the results to the rest of Colombo and to other urban areas in Sri Lanka to arrive at a monthly value of Rs. 735 million (p.103). This translates to an annual value of US$ 77 million. Since roughly 30% of Sri Lanka’s elephants live within my study area, the annual willingness to pay for this population would be about US$ 23 million. It is important to keep in mind, however, that these calculations relate only to the value placed on elephants by urban Sri Lankans. Therefore, presumably, the value this group would place on all the parks and wildlife within the landscape would be much higher than US$ 23 million219.

Table 7.6 summarises the various methods I used, in the context of the types of valuation methods I introduced at the end of section 7.3

Table 7.6: Summary of valuation methods used in economic model

TEV component Valuation methods used in economic model Agriculture Actual prices were used to calculate revenues (based on recent historical data). An estimated cost income ratio was used to estimate costs, i.e. simulated prices. Non-timber forest products The benefits transfer method was used, drawing from studies that estimated the values of non-timber forest products in similar locations in other parts of Sri Lanka. Nature tourism Revealed prices on tourist expenditures (based on World Bank survey) were used to calculate revenues. Actual government expenditures to maintain parks, mitigate human-elephant conflict and pay compensation to people affected by the conflict and crop damage caused by wild animals, and amounts spent by farmers to protect their crops were used to calculate costs. Carbon sequestration Simulated market prices for carbon credits were used (based on recent reports) Watershed benefits A variety of methods underlie the World Bank study from which I drew my estimate of watershed benefits within Sri

218 As described in section 7.3, Contingency Valuation is a survey-based method of assessing respondents’ Willingness to Pay (WTP) for a good or service (Muchapondwa et al. 2008; p.687). 219 Consequently, the number of US$ 23 million I have used in my economic model would be a significant underestimate of the bequest and existence values of the areas’ natural resources.

276 TEV component Valuation methods used in economic model Lanka’s southeast landscape. These include: • Averted expenditures, e.g. costs of added nutrients to maintain soil productivity and expenditure on erosion controls measures, • Actual costs of providing water, based on historical project costs, and, • Hedonic property value techniques. Bequest and existence values Contingency valuation was used in the study from which I inferred the amount attributable to the southeast landscape.

Table 7.7 summarises the estimated values of the TEV components for the southeast landscape.

Table 7.7: Summary of estimated annual benefits, costs and economic profits from land- use within Sri Lanka’s southeast landscape (US$ millions)

Estimate (US$ millions per year) TEV component Component Economic Benefits Costs profits Rice cultivation 102 90 12 Chena crops 56 47 9 Other crops 40 34 6 Direct values Non-timber forest 4 - 4 products Nature tourism 9 5 2 Carbon sequestration 12 1 11 Indirect values Watershed benefits 30 - 30 Bequest and 23 - 23 existence values Total 276 177 99

As expected, the majority of the revenues are derived from agriculture, which accounts for 72% of revenues. However, there are several other important revenues sources such as watershed benefits (11%), bequest and existence values (8%) and nature tourism (4%). In addition, there is potential to obtain US$ 12 million per year (4%) in carbon credits for forested areas outside PAs.

277 Although agriculture makes a significant contribution to revenues (72%) it accounts for a staggering 97% of all costs. Therefore, when considering economic profits, agriculture accounts for only 27%, while other components such as watershed benefits (30%), bequest and existence values (23%), carbon sequestration (11%) and nature tourism (6%) increase in significance.

Potential issues with the estimates

It is important to bear in mind that there is likely to be variation in the numbers presented in Table 7.7. In statistical terms, the numbers would have standard deviations due to unforeseen factors such as the weather and market forces, omission of revenue and cost sources, and methodological issues. I now consider how each of these factors could affect my calculations.

In any given year, various unforeseen natural and market forces could affect agriculture, the biggest revenue source within the landscape. For example, droughts or floods could ruin crops, as in early-2011, when floods inundated 200,000 acres of paddy lands in Sri Lanka’s five major paddy cultivation districts (Nizam 2011). Market forces could also cause price fluctuations, as happened in mid-2011, when vegetable prices dropped due to increased supply from northern Sri Lanka (Bandara 2011). Several farmers I interviewed said that they often have to sell their produce well below prevailing wholesale prices because there are local gluts due to farmers cultivating similar crops. Variation in global prices of imports such as petroleum and chemical fertilisers also sometimes result in unanticipated cost increases. Finally, as pointed out above, the price of carbon is extremely uncertain at present since the market is still at an early stage of development.

Omission of sources of revenues or costs could also be an issue. For example, as pointed out above, although the government incurs substantial costs to build and maintain irrigation facilities for paddy farmers, these are not included in my model due to difficulties in obtaining reliable data. Potential post-harvest losses, which could be as high as 15% of the rice yield, are also omitted (Sandika and Dushani 2009; p.74). Furthermore, there are inherent uncertainties in some model assumptions, such as

278 multipliers used to assess broader economic impacts of tourists, and the carbon content of various vegetation types within the landscape.

Finally, some of the calculation methods have inherent weaknesses. For example, Kareiva et al. (2011; p.37) point out that there are some issues with the benefits transfer method, because “assuming that every hectare of a given habitat is of equal value ignores well-demonstrated differences between sites in terms of scarcity, spatial configuration, size, quality of habitat, number of nearby people, or their social practices and preferences, all of which may be crucial in determining the value of ecosystem services”. Although I attempted to account for some of these weaknesses by using studies from Sri Lanka, other criticisms of the method are still valid. Similarly, there are criticisms of the contingent valuation method used in the study on which I based my bequest and existence value estimates. According to Kareiva et al. (2011; p.23):

In contingency valuation assessments of value, interviewees are asked what they would be willing to pay in order to prevent some real or hypothetical amenity. Many economists distrust results from survey approaches, claiming that individuals’ asserted preferences in the hypothetical circumstances posed by surveys bear no systematic relationship to their true preferences. Defenders of survey methods counter that, in many cases, surveys are the only method available. Despite these potential issues with the TEV calculations, it is important to bear in mind that the main objective of the exercise is to provide policy makers and other key stakeholders a rough idea of the potential value of the range of services provided by landscape’s natural resources. Although some estimates are admittedly imprecise, many ecosystem services are often completely ignored in current policy-making processes, implying that their (economic) value is negligible. Therefore, in that respect, these estimates provide some important insights, especially as we look at potential future scenarios of land-use, a topic I consider next.

279 Step 3: Potential future land use scenarios

Having built a model that estimates TEV components for current land-use, I varied model assumptions to construct two scenarios of potential future land-use. The first scenario, labelled Scenario A, is the current land-use within the study area (i.e. the ‘base case’). The second scenario, labelled Scenario B, assumes that the government further develops nature tourism assets within the landscape and implements MERs in forested areas outside PAs. The third scenario, labelled Scenario C, assumes that the government pursues a strategy of increasing agriculture and ‘traditional’ development activities, resulting in some environmental degradation.

These two scenarios were created to emphasise the dichotomy between ‘traditional development’ and ‘environmentally-friendly, nature tourism-led development’. Obviously, there could be various other intermediate scenarios, for example one that achieves some development while reducing environmental impacts (a ‘sustainable development’ scenario). However, as the government’s current development trajectory is much closer to Scenario C, this pattern of land-use is unfortunately quite realistic in the present context.

I now examine in further detail the assumptions contained in each of the two scenarios.

Scenario B

In this scenario, land-use patterns remain unchanged, but the government designates current forested areas adjacent to PAs—including chena plots within these areas—as MERs. Therefore, while existing cultivation would still continue within MERs220, further clearing of forests would not be permitted. Moreover, once chena farmers have harvested their crops in the dry season, they would be required to remove fences so that elephants could have access to regenerating vegetation within their plots.

The Forest Department—which has jurisdiction over much of the land contained within MERs—would issue licenses to high-end tourism operators to establish operations in these areas. I have assumed that 10 camps would be established, and that they would be

220 The Forest Department would probably have to map existing chena plots and issues permits to their cultivators.

280 similar to those of current operators at Yala National Park such as Kulu Safaris and Leopard Safaris221, i.e. tented mobile camps that could accommodate about 15 tourists at a time, each paying around US$ 250 a night. Tourists would stay 5 nights on average, and the occupancy rate is assumed to be 60%, to account for seasonality in the tourism business222. In addition, the Department of Wildlife would issue licenses for 5 other camps within PAs—the economics of these operations would be similar to those in MERs, except that they would be able to charge visitors US$ 300 a night given their premium locations within national parks.

Based on the economic multipliers referred to earlier in the section, the annual economic contribution from these high-end tourism operations would be over US$ 13 million, of which about half are from activities within MERs. The government would earn about US$ 1.6 million in additional park fees and campsite fees.

Scenario C

The main characteristic of Scenario C is that land-use has been changed to reflect the government’s current view of ‘development’. All forms of cultivation—paddy, chena, other crops, home gardens—and built-up areas have been increased at the expense of forest, scrub and grassland223, as per Table 7.8.

Table 7.8: Land-use in Scenario C (calculated from GIS data obtained from Survey Department of Sri Lanka)

Change from Resulting land- % of total of Land-use types current land use for Scenario resulting land use (km2) C (km2) use Paddy 800 1,291 20% Chena 100 643 30% Homesteads/ gardens 200 926 14% Other cultivation, e.g. coconut, fruit 300 476 7% trees

221 The websites of these operators are: http://www.kulusafaris.com/home.html and http://www.leopardsafaris.com/home.html. 222 July through January are considered the high season for tourism in Sri Lanka (Sri Lanka Tourism 2009). 223 The net change in zero.

281 Change from Resulting land- % of total of Land-use types current land use for Scenario resulting land use (km2) C (km2) use Forest -1,000 1,770 27% Scrub and grassland -500 840 13% Reservoirs and irrigation channels 0 214 3% Other water bodies (e.g. streams, 0 94 1% water holes, lagoons) Other natural (marsh, rocks, sand, 0 105 2% salt pans) Other human (e.g. built up areas, 100 109 2% quarries, unclassified) Total 0 6,467 100%

These land-use changes affect many TEV components. Cultivation of paddy and other crops increase by 72%, in terms of extent cultivated (compared to Scenario A), while forest, scrub and grassland decrease by 36%. The cost to income ratio is assumed to increase slightly for additional land brought under cultivation to reflect that, on average, it is likely to be of lower quality than existing cropland. Some investment would be required to prepare this new land for cultivation, which, at Rs. 20,000 per ha, comes to an additional US$ 23 million. The government would also spend an additional US$ 14 million annually on fertiliser subsidies.

The value of non-timber forest products, which are directly proportional to the available forested land outside PAs, go down by US$ 3 million, while there is a slight drop in the growth rate of foreign tourists visiting the PAs in the area, as compared to 2010224. Moreover, human-elephant conflict would worsen, requiring farmers to spend more to protect their crops; the government would need to allocate additional resources for conflict mitigation and compensation to affected farmers.

Indirect benefits are also affected. Carbon sequestration benefits, which are directly linked to the extent of forested land outside PAs, would drop by US$ 9 million each year, while watershed benefits would be impaired due to the reduced environmental safeguards and increased fertiliser use, resulting in an annual decrease of US$ 9 million.

224 The number would still grow compared to the base figures of 2010; however, the growth rate would be lower.

282 Table 7.9 compares Scenarios A, B and C.

Table 7.9: Estimated annual benefits and costs in three land-use scenarios in Sri Lanka’s southeast landscape (US$ millions)

TEV Scenario A: Scenario B: Scenario C: component Current Enhanced nature Development tourism Benefits Costs Benefits Costs Benefits Costs Rice cultivation 102 90 102 90 215 178 Chena crops 56 47 56 47 66 22 Other crops 40 34 40 34 94 56 Non-timber 1 79 4 - 4 - forest products Nature tourism 9 5 20 6 5 5 Carbon 3 12 1 12 1 1 sequestration Watershed 30 - 30 - 21 - benefits Bequest and 23 - 30 - 16 - existence values Total 276 177 294 178 420 341

Table 7.9 highlights the contrasting compositions of revenues in Scenarios B and C. In Scenario C, agriculture accounts for almost 90% of revenues, while in Scenario B, revenues are much better balanced among agriculture, nature tourism and indirect uses such as carbon sequestration and watershed services—the three areas accounting for 68%, 5% and 14% of revenues, respectively.

Overall revenues are highest in Scenario C; they are 52% and 43% higher than in Scenarios A and B, respectively. However, costs are also substantially higher in Scenario C—in fact they are almost double those of Scenario B. Consequently, as shown in Table 7.10 below, profitability is lowest in Scenario C and highest in Scenario B. This is primarily because agriculture, the main type of land-use in Scenario C, has very high input costs; this impact is exacerbated when the substantial government subsidies provided for paddy cultivation are taken into account. It is important to note that even if carbon sequestration benefits are disregarded—given that they cannot be accessed at present—Scenario B is still US$ 28 million more profitable than Scenario C.

283 Table 7.10: Estimated annual economic profits and required investment in three land-use scenarios (US$ millions)

Scenario A: Scenario B: Scenario C: TEV component Current Enhanced nature tourism Development Rice cultivation 12 12 16 Chena crops 8 8 9 Other crops 6 6 14 Non-timber forest 4 4 1 products Nature tourism 5 15 0 Carbon sequestration 11 11 2 Watershed benefits 30 30 21 Bequest and existence 23 30 16 values Total economic profits 99 116 79

Required one-time 3 7 25 investment

Table 7.10 indicates that Scenario B is significantly more profitable than the other two scenarios. However, it is important to note that two sets of benefits cannot be monetised at this time, i.e., bequest and existence values and carbon sequestration225. Despite this fact, these components are still important (as I argued at the beginning of the chapter); however it does mean that Scenario B does not look quite as outstanding at the present moment. Having said that, it is important to note that, even if we completely exclude these two components, Scenario B would still be more profitable than the other two Scenarios (with economic profits of US$ 75 million compared to US$ 65 million and $61 million for Scenarios A and C respectively); in fact Scenario B still would be

225 The bequest and existence value derived from the contingency valuation study conducted in Colombo cannot be monetised, but was included in the analysis to provide a indication of the potential satisfaction Sri Lankans derive from elephants. However, funds for elephant conservation are available globally, although it is difficult to obtain a reliable estimate of how much could be obtained by Sri Lanka on an annual basis. Likewise, REDD+ funding is not available at the moment because the global mechanims are still being worked out. However, pointed out in section 7.4, some bilateral schemes are already up and running.

284 almost 25% more profitable than Scenario C, the option that is currently being pursued by the Sri Lankan government.

As per Table 7.10, investment costs for Scenario C are several times higher than in the other two scenarios. Most of the incremental expenditure goes toward preparing additional agricultural land for cultivation. In reality even more government investment would probably be needed to construct irrigation facilities required for these new crop- growing areas. The investment required for Scenario B is mainly for establishing MERs—for example, to construct additional electric fences to safeguard villages and chenas during the growing season. The approximate cost per km of new electric fencing is approximately US$ 4,500, but existing fences that have been erected in the wrong locations—such as those on administrative boundaries—could be repurposed (Fernando et al. 2011; p.99). The investment required to obtain carbon credits is relatively low as international funding is available for countries to prepare for REDD+; as pointed out previously Sri Lanka has already secured US$ 4 million for this purpose (Mattsson et al. 2012; p.31)226. Start-up costs for high-end tourism operators in Scenario B have not been included, because they would be typically financed by debt and paid off during the course of operations227.

Non-economic factors related to the TEV calculation and the various Scenarios

As emphasised in sections 7.2 and 7.3 the concept of TEV does not consider several important factors such as critical thresholds, intrinsic values and equity-related issues. I now re-examine my scenarios—especially Scenarios B and C—in light of these factors.

Non-economic factors related to Scenario B

The assumption in Scenario B that agricultural land does not increase from current levels could impact paddy-farming families in which more than one child decides to go

226 Sri Lanka is likely to have access to further funding as it proceeds through the various stages of preparation (Mattsson et al. 2012). 227 In any case, they would be relatively low.

285 into farming228. As demonstrated by the analysis in Chapter 5, many of these farming families have incomes that are below the poverty line. A similar situation would arise among chena farmers, given restrictions on clearing land for additional chenas within MERs. This group—one of the poorest in the country—would potentially be affected if one of their main sources of livelihood were taken away. Many chena farmers I interviewed told me that even though they earn meagre profits, their crops help feed their families for at least part of the year.

Therefore, for equity reasons, if Scenario B is implemented, steps must be taken to address potential negative impacts on poor farming families. This could be done by redistributing a portion of the income from tourism in the MERs among this group of people. Moreover, concrete steps should be taken to ensure that they benefit from the increased economic activity arising from the influx of high-end tourists into the area, and the resulting multiplier effect on the local economy. I will discuss how this to could be done in Chapter 8.

Non-economic factors related to Scenario C

While increased agricultural and development activities in Scenario C would benefit some farming families, they could cause several other social issues, the most obvious being the rapid escalation of human-elephant conflict. As pointed out in Chapter 5, if the government continues on its current development trajectory, the conflict could worsen significantly, reaching levels currently being experienced in Sri Lanka’s northeast and northwest. If that happens, the livelihoods of poor paddy and chena farmers would worsen significantly, since, as revealed in my interviews, elephants are perhaps the main threat to their well-being. A larger proportion of their crops would be lost, more homes would be destroyed and a higher number of people would lose their lives. While some of these impacts are captured in the economic model, what it fails to account for adequately is the inequitable nature of the damages, as the people affected are extremely poor. Moreover, loss of lives are difficult to measure in economic terms, and therefore not considered in my analysis, apart from the compensation the

228 While one child could presumably cultivate his or her parents’ field, other children would have no land to cultivate. Alternatively, the lands would have to be divided. In either case there would be a shortage of paddy land.

286 government pays to families of people killed by elephants, or about US$ 1,000 per person killed.

Inevitably, more elephants would die, affecting one of Sri Lanka’s most important elephant populations. As pointed out in Chapter 3, people across the country would view this negatively as the elephant is a national icon. Moreover, the loss of elephants would be bad from a global conservation viewpoint, because Sri Lanka contains over 10% of the world’s Asian elephant population. All these issues would have political consequences for the government, especially since, as pointed out in Chapter 4, the current Sri Lankan President promised to mitigate the conflict as part of his 2010 re- election campaign, and many other high-ranking government officials have followed suit.

The increased environmental destruction resulting from Scenario C could have severe consequences if critical environmental thresholds associated with some natural resources are breached. As pointed out in section 7.4, if this happens, their ecosystem services may deteriorate rapidly, with unpredictable, potentially catastrophic, consequences. For example, pollution levels in some rivers may rise drastically, causing widespread health issues among local people. Unfortunately, there are precedents for this type of calamity in other areas of Sri Lanka, where excessive chemical fertiliser use has been linked to chronic renal disease by Bandara et al. (2011; p.439):

Chronic renal failure associated with elevated dietary cadmium among farming communities in the irrigated agricultural area under the Mahaweli River diversion scheme has increased significantly. Cadmium, derived from contaminated phosphate fertiliser, in irrigation water finds its way into reservoirs, and finally to food, causing chronic renal failure among consumers... Chronic renal failure associated with elevated dietary cadmium was first reported in North Central Province of Sri Lanka in 1993. Bandara et al. (2008) reported that the number of patients in the province was around 5,000. Chronic renal failure cases in the province steadily increased to 7,650 in 2009, as reported by the Health Ministry of Sri Lanka. The Government Information Department reported the number treated for chronic renal failure by October 2009 in Anuradhapura Hospital to be 9,000. These cases are mainly from North Central Province, Uva Province (Girandurukotte and Nikewewa), and North Western Province. The total number of deaths reported from chronic renal failure in Anuradhapura, North Central Province General Hospital is 1,082 to date, since 1993.

287 River pollution could also affect some of the area’s PAs because the Menik Ganga and Kumbukkan Oya which flow through Yala National Park are important water sources for animals, especially during the dry season, and if they become polluted or water levels drop significantly, the park’s biodiversity would be affected (Dissanayake and Smakhtin 2007; p.5). Apart from the obvious impact on tourism to Sri Lanka’s most visited park, the country would suffer significant ‘intangible’ or ‘intrinsic’ losses.

Finally, there could be cultural impacts if rivers used for religious festivals are affected due to the extensive development activities envisioned in Scenario C. Dissanayake and Smakhtin (2007; p.4) describe problems that are already occurring at Kataragama, an important religious site mentioned in Chapter 4:

Kataragama is a sacred city for Sri Lanka’s Buddhist and Hindu followers, in reverence to a god that believers of both religions worship. Each year, the Kataragama Festival attracts about 100,000 people per day over a 15- day period during July and August. The main event of the festival is the water cutting ceremony held in gratitude to God Kataragama. The water cutting ceremony ideally requires about 1.2 to 1.5 m of water in the river. The minimum water depth required for the festival around Kataragama temple has been estimated as 0.6 m with the corresponding discharge at Kataragama gauging station of 2.0 cubic meters per second (m3/s). However, over the last decade, the river did not carry this much flow at Kataragama during the months of the festival. Analysis of available flow records for a period of 1977-1998 shows that the long-term mean flow in a river during August, for example, is less than 0.6 m3/s. Another cultural practice at the festival is the offering of pure water to gods and bathing in ‘holy’ water of the Menik Ganga prior to visiting God Kataragama to worship. Due to increasing levels of water pollution arising from low flows, festival officials have deemed the river water unsuitable for bathing. The 12 tube wells in the shrine premises alone cannot satisfy this requirement.

7.6. Summary and implications of the economic analysis

In this chapter I looked at how economic analysis could help society make informed decisions about trade-offs among alternative land-use options. Given that a main driver of deforestation and other forms of environmental destruction is policy-makers’ failure to consider services provided by natural resources, I explored how this deficiency could be addressed by the concept of TEV, which considers economic benefits in a holistic manner.

288 I applied this framework to Sri Lanka’s southeast landscape to better understand economic revenues, costs and profits associated with current forms of land-use. As expected, the majority of revenues are derived from agriculture, but there are several other important benefits such as watershed services, bequest and existence values, nature tourism and carbon sequestration potential. When considering economic profits, other benefits increase in significance given the high costs associated with agriculture. Although there is likely to be variance in my estimates due to factors such as the weather and market forces, omission of some revenues and costs, and methodological issues, they nonetheless provide useful insights regarding the economic profitability of different land-use types.

I then considered two scenarios: Scenario B, where MERs are established and high-end nature tourism promoted, and Scenario C, where there agriculture and infrastructure-led development occurs without adequate environmental safeguards, resembling the government’s current approach. Overall revenues were significantly higher in Scenario C, but they were more balanced in Scenario B. However, costs and required investment were much higher in Scenario C, drastically reducing its economic profitability. Consequently Scenario B ends up being 47% more profitable than Scenario C, demonstrating that, in economic terms, establishing MERs is a sensible form of land- use.

Finally, I looked at the impacts of some non-economic factors that are not captured by the TEV framework. Scenario C has numerous issues when considered from this perspective. Firstly, human-elephant conflict would increase, impacting livelihoods of some of Sri Lanka’s poorest people, and resulting in more deaths of human beings and endangered, culturally significant elephants. Secondly, extensive pollution could push some natural resources to their limits of ‘normal’ function, with potentially catastrophic consequences. Finally, important cultural services may be affected. In summary, Scenario C is socially undesirable for both economic and non-economic reasons.

Since Scenario C proves to be less profitable than Scenario B, and has many non- economic problems, the latter is clearly the better alternative. However, there are some potential equity issues in Scenario B, as some poor farmers may be affected by restrictions on the expansion of paddy and chena land. These factors should be considered during MER implementation, which is the topic of my next chapter.

289

This page has intentionally been left blank

290 Chapter 8: Implementing Managed Elephant Ranges

Having introduced MERs as a possible approach to mitigate Sri Lanka’s human- elephant conflict (Chapter 4), I delineated potential ranges within the country’s southeast landscape, an area containing 30% of Sri Lanka’s elephants (Chapter 5). I then conducted a stakeholder analysis to understand needs and concerns of the landscape’s key stakeholders (Chapter 6), and an economic analysis to determine the economic viability of potential MERs (Chapter 7). I now focus on how MERs could be implemented, bringing together information and results from my previous chapters. I identify required characteristics of MERs based on my stakeholder and economic analyses, discuss potential challenges to implementing the concept, and suggest a three- step process to establish MERs in southeast Sri Lanka.

8.1 Required characteristics of Managed Elephant Ranges

I now revisit the main findings of analyses conducted in Chapters 5, 6 and 7 to identify key factors that should be considered when implementing MERs. These are summarised in Figure 8.1 below:

291 • Farmers: human elephant conflict, ways to improve livelihoods Stakeholder • Chena farmers: land tenure, human elephant conflict, ways to concerns improve livelihoods • Department of Wildlife: human elephant conflict • Forest Department: obtain revenue from land • Politicians: human elephant conflict, improve local livelihoods • Tourism companies & entrepreneurs: business opportunities • Sri Lankan public: elephant conservation • NGOs: elephant conservation, improve local livelihoods

Economic • Agriculture sources of value • Nature tourism (for Scenario B) • Carbon credits

Non economic • Equity issues: impacts on poor farmers issues (for Scenario B)

1! Figure 8.1: Key findings from analyses in Chapters 5, 6 and 7

Figure 8.1 summarises the needs of the main stakeholder groups that should be considered—paddy and chena farmers, government agencies such as the Department of Wildlife and the Forest Department, politicians and high-ranking government officials, tourism companies and local entrepreneurs, the Sri Lankan public, and NGOs working in the landscape. These stakeholder needs should be met while accessing key sources of economic value within the landscape—such as agriculture, nature tourism and carbon credits—in an environmentally friendly manner, ensuring that MERs are financially viable. Watershed benefits and bequest and existence values are not included here, as they cannot be monetised229; however, they will be considered during the management planning process in section 8.3. Finally, potential impacts on livelihoods of poor farmers—the non-economic issues associated with Scenario B—should be addressed for equity reasons.

229 This does not reduce their importance, however. As discussed in Chapter 7, if environmental services such as watersheds are affected, local livelihoods would be severely affected.

292 Taking into account all these factors, I propose several important characteristics of MERs in Table 8.1:

Table 8.1: Required characteristics of Managed Elephant Ranges

Stakeholder group(s) whose Economic and non- Characteristic needs/concerns are economic issues satisfied230 addressed

1. Reduced human-elephant conflict • Paddy and chena farmers • Address some

within landscape (both within and • Department of Wildlife equity issues outside MERs) through improving • Politicians and high- livelihoods of • Reduced impact of elephants ranking government poorest people in on local livelihoods, officials especially those of farmers landscape • Conservation NGOs • Adequate habitat, food and • NGOs focused on water sources for elephants improving local livelihoods

• Urban Sri Lankans

2. Improved chena cultivation • Chena farmers • Address some practices within the landscape • Forest Department equity issues through improving • Land tenure security for • Conservation NGOs chena farmers livelihoods of • NGOs focused on poorest people in • A set of rules—agreed with improving local landscape chena farmers—to improve livelihoods the productivity of chena • Reduce likelihood • Government departments cultivation within MERs, of environmental related to agriculture while making it more thresholds being environmentally and elephant reached friendly

230 Elephants are arguably stakeholders as well, but since they cannot act/behave as ‘conventional’ stakeholders, their interests are looked after by the Department of Wildlife and Conservation NGOs.

293 Stakeholder group(s) whose Economic and non- Characteristic needs/concerns are economic issues satisfied230 addressed

3. Opportunities for nature tourism • Tourism companies • Increase economic activities • Forest Department profits

• Tourism authorities

• Local entrepreneurs

4. Credits for carbon sequestration • Forest Department • Increase economic profits

5. Improved local livelihoods and • Local entrepreneurs • Address some local support for conservation • Poor locals, including equity issues • Income generating many chena farmers through improving opportunities for locals livelihoods of • NGOs focused on poorest people in • Economic benefits shared improving local landscape with poor locals livelihoods • Increase economic • Local politicians profits

These five actions outlined in Table 8.1 would address key issues and concerns of the main stakeholder groups identified in Chapter 6. They also help access some of the landscape’s main sources of economic value, and promote a more equitable distribution of income, the main non-economic issue identified in Chapter 7. I now take each of the actions in turn, using examples from outside Sri Lanka to provide practical insights on how to implement them effectively.

294 Reduced human-elephant conflict within the landscape

The entire purpose of establishing MERs would be to reduce human-elephant conflict within the landscape and provide elephants with suitable habitat. Perhaps the first step towards meeting this goal is to carefully select MERs, i.e. they should contain adequate food and water resources for the animals. In Chapter 5, I presented a preliminary attempt at this, mapping out potential MERs in my chosen landscape based on suitability of vegetation (forest and scrub) and location, i.e. being adjacent to existing PAs to ensure continuous elephant habitat within and outside parks. This exercise should be improved upon using more up-to-date mapping data231 and through land surveys in selected areas to verify the correspondence between survey data and the current situation on the ground.

Measures should also be taken to reduce human-elephant conflict, both within MERs and in the larger landscape. As pointed out in Chapter 4—where the concept was introduced—this should be done through the judicious location of electric fences232, a method that has been used to reduce the conflict in elephant ranges throughout Asia and Africa, as described by Kioko et al. (2008; p.52):

Electric fences are increasingly being used to reduce crop damage by elephants. The fences act as physical as well as psychological barriers to separate elephants from settled areas. In Kenya, more than 1,200 km of electric fencing has been installed to protect farmlands from elephants and an additional 1,300 km of fencing is planned.

231 While I used the most recently available data from the Survey Department of Sri Lanka, this information usually takes 1 – 2 years to produce, after initial surveys are undertaken. Given the rapid development within the southeast landscape, some of this information may already be dated. 232 As discussed in Chapter 4, many other measures to reduce human-elephant conflict—both traditional ones, mainly by farmers, and measures such as elephant drives and relocation of problem animals into parks by the Department of Wildlife—have been tried, but they have not been very successful.

295

Figure 8.2: Male elephants at the electric fence bordering Uda Walawe National Park, southeast Sri Lanka (photograph taken on 21/03/2010, by S. Kalpage)

As discussed in Chapter 4, factors such as fence location, construction and maintenance impact the efficacy of electric fences in reducing human-elephant conflict. I now examine these factors in more detail.

Location and design

The first step to maximise the effectiveness of an electric fence is to carefully choose its location. Nelson et al. (2003; p.13) point out that: “knowledge of the ecology, distribution and movement patterns of the local elephant population is vital”, and that “ignoring these might result in a fence in the wrong place, or an unacceptable maintenance demand”. Furthermore, de Silva and de Silva (2007; p.232), who have studied issues with the location of fences in Sri Lanka, recommend that particular care be taken to ensure that elephants are not trapped on one side of the fence. As pointed out in Chapter 4, the practice of constructing fences on administrative boundaries such as national parks, coupled with the fact that many elephants range widely outside PAs,

296 has led to situations where elephants are found on both sides of fences, rendering them completely ineffective! MERs were proposed to prevent this type of situation, by taking down fences between parks and elephant habitat outside parks so that elephants could range widely to obtain required food and water resources, especially in the dry season.

Nelson et al. (2003; p.13) describe three basic types of fence design:

i. Enclosing agricultural land and/or houses and people;

ii. Enclosing elephants and their range; and,

iii. A straight-line barrier between elephants and agriculture (e.g. along a park boundary).

Based on evidence—mainly from Africa—they conclude that the third design is largely ineffective. This seems to be the case in Sri Lanka as well, based on my interviews within the southeast landscape; authors such as de Silva and de Silva (2007) have come to similar conclusions. Nelson et al. (2003; p.13) conclude that the first design type seems to be the most effective since “problem elephants appear not to be easily deflected, rather it seems to make most sense to identify their target and then to keep them out”. This would also be the recommended option for MERs, i.e. permanent fences would be constructed around any villages within these areas, while temporary fences would be used to protect chena crops. The latter would be taken down after crops are harvested, so elephants could access regenerating vegetation within chena plots during the dry season.

Construction

The effectiveness and the longevity of electric fences depend on their construction and maintenance. Fernando et al. (2008; p.15) point out that: “use of high quality components such as porcelain or UV233 stabilized insulators and other hardware, such as non-corroding wires, make fences long lasting and can even last decades if constructed with quality materials coupled with good maintenance”. Gunaratne and Premarathne (2006) describe how some of the main components of fences—the energiser, the solar

233 Ultra-violet.

297 power system, fence wires and fence posts—should be designed, based on a detailed study of fences in several parts of Sri Lanka. Some of their main findings are summarised in Appendix 8.

Maintenance

Issues with maintenance are often the key cause for the failure of electric fences (Hoare 2003; p.2; Fernando et al. 2008; p.15). Hoare (2003; p.2) identify a number of potential maintenance issues:

The most serious maintenance problems with electric fences are nearly always associated with the power supply, especially if this involves the use of solar panels and batteries, as opposed to mains electric power that is seldom available in rural areas. Vandalism and theft of components (particularly solar panels, energizers and wire) is extremely common in community-managed fences. Not only does this inactivate the fence but frequently creates the knock-on effect of the maintenance demand outstripping its budget, leading to total collapse of the project… Keeping vigorous growth of vegetation clear of a fence line in the growing season is a perennial problem that characterizes the management of electric fences. Vegetation contact causes power leakages and overgrowth conceals the fence from being an obvious barrier to elephants. These problems occur widely in Sri Lanka, and were mentioned by several conservationists and farmers I interviewed. Community support and participation is essential for the maintenance of fences since government agencies such as the Department of Wildlife neither have the time nor resources to maintain the several hundred kilometres of electric fences that have been constructed across Sri Lanka. Gunaratne and Premarathne (2006; p.45)—who conducted a study of the effectiveness of fences in several parts of Sri Lanka—found this to be the case:

The study indicates that community support is key to the success of a fence. Fences are often cut by illegal timber fellers and illicit liquor producers. Such events can be policed by well-organised community- based organisations. Community support is also critical in several other ways. Community labour is required to establish a fence and, most importantly, for fence maintenance. Community organisations have a role to play in protecting a fence and keeping its route clear. I interviewed some farmers—in a few of the villages within the landscape—who led efforts to maintain the local electric fences, to understand the activities related to fence

298 maintenance and the associated issues. Farmers 36 and 37 provided me with some useful insights:

In 2008, CARE built an electric fence, but it was not successful because the current was too weak and elephants destroyed it. In 2010, JICA234 helped us construct a new fence. This initiative has been much more successful, especially since it was constructed with the cooperation and assistance of villagers. The Farmers’ Association maintains the fence. This body has 42 members, but some others also have a stake in the fence, so we have 62 people involved (although only 50 – 55 regularly attend meetings). People contribute Rs. 50235 (per household per month) for the upkeep of the fence (interview, Farmer 36, 05/09/11). I walk around the fence (with other people in the society) to identify any issues. The solar panel and the circuit are located on my property. The Grama Sevaka236 and the farmers’ association are responsible for ensuring the fence is in good working order. Several parties have signed an agreement regarding the fence, including: the Department of Wildlife, the District Secretariat, the Provincial Secretary, the Centre for Conservation Research, JICA, and the Farmers’ Association. Mr. Bandara from the Centre for Conservation Research comes by every month to check on our fence (interview, Farmer 37, 05/09/11).

234 The Japanese International Cooperation Agency. 235 About US$ 0.50. 236 The government official assigned to the village.

299

Figure 8.3: Sketch of electric fences around Ketanwewa village in the records of the Farmers’ Association responsible for maintaining them. The fences are indicated in red, the crops are in green, and the squares represent individual houses (photograph taken on 05/09/2011, by S. Kalpage)

It is important that all electric fences in an area are well-constructed and properly maintained, because elephants that succeed in breaching ineffective fences often learn to break well-maintained fences (Fernando et al. 2011; p.99). Fernando et al. (2008; p.14) describe some of the tactics of habitual ‘fence-breakers’:

Some elephants eventually adapt (to electric fences) with prolonged exposure. Elephants have been known to breach electric fences by using tusks, which do not conduct electricity, pushing or kicking down fence posts and stepping over the fence using the thick soles of their feet to depress the wires. Some elephants also learn that an electric shock does not harm them and simply barge through the wires. Once an elephant learns to breach an electric fence, the fence becomes useless against the particular individual. There have been similar experiences with African elephants; consequently, a robust system to ‘enforce the fence’ was put in place at the Ol Pejeta Conservancy in Kenya (Graham et al. 2009; p.7):

300 A mobile rapid response team was created to respond to reports of elephants attempting to challenge the perimeter fence. The team aimed to get to the site and scare away the elephant/s before the fence could be damaged. Elephants that did break the perimeter fence were identified by a trained elephant researcher. If these identified elephants challenged the fence more than three times, they were destroyed by or with the consent of the Kenya Wildlife Service. It is unlikely that the cultural norms in Sri Lanka would permit the destruction of elephants in this manner. However, the relevant lesson that could be learnt from the system at the Ol Pejeta Conservancy—which drastically improved the effectiveness of the fence—is that both farmers and wildlife officials should take steps to prevent elephants from challenging fences. The Centre for Conservation Research—an NGO working to reduce the conflict in Sri Lanka’s southeast landscape by helping locals erect electric fences—stresses to farmers the necessity of watching over their crops at night. Recent incidents at Thammanna village—which had a successful fence for many years—illustrates what could happen if farmers fail to do this. Farmer 39 described what happened, when I interviewed him:

The fence was constructed around the village in 2008. It has been very effective and most of our issues with elephants were resolved. However, in the past few days, elephants have started breaking the fence. This is entirely our own fault, because we failed to properly maintain the fence. We were told not to put in ‘fake’ wires (i.e. wires that were not electrified), but some villagers did that. People have also been lax in watching over their crops at night. Our negligence and laziness has backfired on us (interview, Farmer 39, 05/09/11).

301

Figure 8.4: The fence at the Thammana village that was breached after several years, mainly because villagers allowed elephants opportunities to challenge it (photograph taken on 05/09/2011, by S. Kalpage)

Financing fence construction and maintenance

Electric fence construction costs vary widely. Kioko et al. (2008; p.53) estimate that two fences near the Amboseli National park in Kenya cost US$ 9,000 per km, while fences installed along the border of the Indira Gandhi Wildlife Sanctuary in the Anamalai hills cost only around US$ 2,829 per km (Fernando et al. 2008; p.14). Obviously, costs depend significantly on the type of fence being constructed, and the specific materials used. Fernando et al. (2011; p.99) estimate that fences suitable for Sri Lanka cost around US$ 4,500 per km.

In Sri Lanka, fences on park boundaries are typically constructed by the Department of Wildlife, while those around villages are built by NGOs such as the Sri Lanka Wildlife Conservation Society, the Centre for Conservation Research, and Practical-Action (Fernando et al. 2011; p.99). Chena farmers sometimes build (and maintain) rudimentary fences around their plots. As pointed out previously, village fences are

302 maintained by village societies while those on administrative boundaries are maintained by the Department of Wildlife.

In MERs, the government ideally should fund the construction of fences, with NGO assistance if possible. Chena farmers and villagers should be responsible for basic maintenance of their respective fences and they should devise a system to watch over these fences, which deters elephants trying to breach them. Ideally, a centrally funded rapid response unit should be put in place to assist farmers in these activities—this could be funded by revenue generating activities within MERs such as nature tourism and carbon credits. However, villagers would still need to be involved to ensure that they feel a sense of ownership over the fence, as it would be far too costly to create an external unit responsible for maintaining the entire fence.

Improved chena cultivation practices within the landscape

As described in Chapter 2, although the practice of chena has been criticised widely as an environmentally harmful form of land-use, British colonial administrators propagated this view to dispossess Sri Lankan peasants of land deemed necessary for colonial plantations. However, findings of recent studies on environmental and social effects of swidden agriculture—the more generic name for chena—are mixed. Bruun et al. (2009; p.376), who evaluated the practice in Southeast Asia, provide some insights:

Swidden cultivation is the traditional land use system in the sloping uplands of Southeast Asia… Regardless of a lack of substantiating data, the swidden systems in the upland areas have frequently been deemed environmentally destructive causing deforestation, soil degradation in terms of erosion and negative nutrient balances and contributing to CO2 emissions. This perception is, however, increasingly challenged as numerous studies have shown that swidden cultivation in many situations can be a rational economic and environmental choice for resource poor farmers and that swidden cultivation besides being a production system provides a range of ecosystem services in terms of hydrology, biodiversity and carbon storage in soil and vegetation. When adapting chena cultivation to the context of MERs, however, it is important that the potential issues associated with the practice are taken into account, such as those described by Perera et al. (1983; p.187):

303 One of the main reasons why chena cultivation has persisted from ancient times to the present is that it requires very little material inputs for crop production. The importance of this factor cannot be overemphasized when the uncertainties posed by the environment are taken into consideration, i.e. erratic rainfall, the threat of wild animals to the crops, etc. Despite its obvious merits, the practice of chena cultivation results in the diminishing of forests, driving wild animals from their natural habitat, the silting of tanks and intensive degradation of soil. However, the most serious limitation on continued chena cultivation is the increasing population of the dry zone area during recent years and the consequent mounting of pressure on the use of land. In addition to the issues pointed out by Perera et al. (1983), the increasing use of pesticides and chemical fertilisers (de Silva and de Silva 2007; p.217) would be a cause for concern, as the excessive use of these substances could pollute water bodies within MERs237 (as described in Chapter 8).

Agalawatte and Abeygunawardena (1994; p.68) iterate that lack of property rights could contribute to chena farmers’ relative lack of concern regarding the impacts of their cultivation practices:

Most of the chena lands are owned by the government, where property rights are not imposed. This makes these lands very much similar to an open access resource. In a situation where the user rights are not properly defined, many have access to the resource and the present user would have every incentive to derive the maximum benefit of the resource. There are very little or no incentives for these farmers to develop an institutional mechanism to maintain the productivity of these lands for a longer period. Furthermore, as described in Chapter 6, many chena farmers told me that they are very worried about losing access to the plots they had cultivated for many years, because the practice is still technically illegal. The economist Hernando de Soto described the importance of providing property rights to poor people238, pointing out that, apart from enabling them to use their land as collateral, it gives them incentives to use these assets sustainably (de Soto 2001; p.17-18).

Consequently, the Forest Department should legalise existing chena cultivation within MERs. Before doing this, they would need to conduct a detailed survey of existing plots and their owners, and work out the exact legal status that could be afforded to these farmers—since it is technically government land, a transferrable permit system may

237 Although the World Bank data I analysed revealed that 76% of chena farmers did not use chemical inputs (as described in Chapter 5), several chena farmers I interviewed said that they did. 238 Based on research conducted in developing countries, particularly in Latin America.

304 work. Providing farmers with a formal title to the land they cultivate could help improve their livelihoods. For example, the dire financial circumstances of most chena farmers can be traced back to the fact that they do not own their land, preventing them from using their land as collateral and increasing their dependence on moneylenders charging exorbitant interest rates. Farmer 44 described his situation:

We live a hand-to-mouth existence and are eternally in debt. As we don’t have any form of title to our land, we cannot use it to obtain bank loans (like many paddy farmers). We have to pawn our valuables—such as gold and motorcycles—to get the money we need to prepare our fields. Money- lenders charge us exorbitant interest rates of over 20% per month—if we borrow Rs. 10,000, we have to repay Rs. 20,000. Everything we earn is finished once we pay off our debts. If something happens to our crops, we are in debt for many years (interview, Farmer 44, 06/09/11). In return for legalisation of their plots, the Forest Department should require chena farmers to use environmentally and elephant friendly cultivation techniques. This is in line with co-management approaches advocated by conservationists, where communities are granted land tenure in exchange for their commitment to observe certain rules and restrictions in the exercise of those rights. Barber et al. (2004; p.115) argue that:

Devolution (or restoration, in some cases) of tenurial rights over land and resources is a key element of many new approaches to PA governance. Apart from the strong equity and rights-based arguments for restoring indigenous lands to their indigenous owners, it is also often argued that tenurial security—not only for indigenous peoples but also for local communities who may not hold the long-standing claims of indigenous groups—is an important incentive for conservation at the local level. Those with a long-term property interest, it is argued, are more likely to be good stewards of resources, avoiding the “open access” free-for-all that characterises publicly-controlled but scantily-managed parks and other public lands, particularly in developing countries. Under many co- management schemes, conservation agencies explicitly trade greater community tenurial control for community commitments to conservation objectives such as coral reef protection or reforestation. Officials from the Forest Department and the Ministry of Agriculture should work with farmers and other agricultural experts239 to develop a set of chena cultivation practices that reduce the use of chemical inputs and encourage the use of traditional methods to improve soil fertility and combat pests. As pointed out in Chapter 7, the overuse of chemical inputs in Sri Lankan agriculture have caused many environmental and health

239 Those from universities and development agencies, for example.

305 issues. When interviewing chena farmers, I found out that while some use chemical inputs, others, such as Farmer 44, are aware of their drawbacks, and have taken steps to address these issues.

There is now an increasing awareness among consumers that traditional varieties are much better for their health. I still only cultivate old varieties, without pesticides; I use traditional methods to protect my crops. I also use less fertiliser, so my harvest is naturally lower than if I had used chemicals. If others get 15 bushels240 per acre, I get 12. However, because my inputs are less costly, I think that things even out in the end. More importantly, my wife and kids consume some of my produce, so I don’t want to poison them by using too much chemicals (interview: Farmer 44, 06/09/11)241. Older farmers such as Farmer 44 are acquainted with methods that chena cultivators traditionally used to reduce environmental impacts of the practice, such as judicious crop selection and piling up debris to prevent erosion. Bandara (2007; p.3-6) provides further details based on his research on traditional Sri Lankan farming methods:

Chena cultivators adopted an ecologically sound system based on rituals and traditions. The selection of land for the slash and burn process was done systematically by the elders based on the topography and existing vegetation. The slope of the land was the main criterion used to avoid erosion after the land was cleared. Once the chena was burnt, there would be a lot of unburnt debris, such as short logs, sticks, and partly burnt charcoal. Logs and sticks that could be used as firewood were stacked between the forks of standing trees. The rest was piled as ridges, diagonally across possible water flows. Left over debris was piled up around the lower-end of the boundary to arrest further erosion. Crops were chosen so that they matured at different times of the season so that harvesting was also staggered. The objective of using a wide diversity of crops in a chena was to maintain a more natural ecosystem for effective pest management and for regular ground cover to prevent erosion during the northeast monsoon. In addition, this helped meet family needs in terms of carbohydrates, proteins, beverages, spices and fruits. While much could be learnt from traditional methods, these alone would probably be insufficient to ensure sustainable chena cultivation within MERs. This is because, as pointed out in Chapter 4, it is no longer possible to leave land fallow for long periods of time to regenerate, due to the southeast landscape’s rapid population increase; consequently most chena plots are cultivated annually. Therefore, lessons should be

240 A local unit of measure. 241 I also included this quote in Chapter 5, section 5.3.

306 drawn from the extensive body of research on sustainable agriculture. Gomiero et al. (2011; p.15-18) describe a number of different approaches to agriculture management that have been proposed and implemented in this regard; Appendix 8 summarises those relevant to the southeast landscape.

Implementing sustainable agriculture methods would require coordination among the relevant government agencies, farmers, and probably the private sector (who purchase farmers’ produce). Apart from helping farmers adopt efficient, environmentally friendly cultivation techniques, the government should also take steps to resolve issues associated with marketing their produce. Farmers may also require assistance to plan their cultivation processes to avoid the current phenomenon of oversupply of particular crops, leading to dramatic price reductions at harvesting time. Farmers’ Societies could play an important role in representing farmers’ interests—something many of them currently fail to do, as pointed out by a chena farmer I interviewed:

It is very hard to make a living these days as a chena farmer. The prices of tomatoes and maa karal242 have plummeted. One of the main reasons is because our Farmers’ Society is inactive. The society was established in 1994, but it still has not been able to achieve basic things like getting the local wewa243 built. All they do is to write lots of letters—I have a whole stack back there… The Farmer’s Society should do a study before the growing season to determine how much of each crop variety could be reasonably sold without depressing local prices. They should communicate that to farmers at society meetings. This will prevent us from cultivating the same crop, causing prices to plummet. When this happens, traders take only the produce that they feel they can sell. They do not take the rest, so poor farmers are left with stacks of unsold vegetables, most of which do not keep for long, so they make huge losses. The poor farmers are usually in debt, so they cannot repay their debts (interview, Other locals 13244, 21/02/12). So, in summary, the rules and processes governing chena cultivation within the MERs should be worked out through discussion among the main stakeholder groups. The process of adaptive management—which will be discussed later—should be used to implement what is agreed.

242 Yard-long beans. 243 Reservoir. 244 He was a shopkeeper whose main customers were chena farmers; he also cultivated a chena.

307 Opportunities for nature tourism activities

As described in Chapter 7, there is significant potential for nature tourism within MERs. Marketing these areas to high-end tourists—the most profitable and environmentally friendly option245—would require providing them with the experiences they seek. A useful starting point to attract this type of clientele is to resolve some of the issues at Yala and the landscape’s other world-class PAs. As park visitation increased rapidly following the end of Sri Lanka’s civil war in 2009, Yala has become a victim of its own success, as described in a recent BBC246 report (Haviland 2012; p.1):

Yala is a place of magic, of rocky outcrops, big trees, ancient lakes and the rushing sound of the Indian Ocean never far off. It is a place of leopards, elephants, sloth bears, and a rich bird life of peacocks, hornbills and more. But conservationists in Sri Lanka are warning that anarchic behaviour in some national parks is endangering the wildlife and the ecology of wilderness areas. They say safari vehicles are flagrantly breaking speed limits and that marauding behaviour by drivers and tourists is grossly insensitive to fauna and flora. This is especially the case in Yala in the southeast, the most famous habitat for leopards—the only big cat found on the island. But there have been serious consequences for wildlife in the area. Four months ago a female leopard cub was killed in Yala by a hit-and-run driver. The BBC has been told of jeep drivers going up to 100 km per hour (the nominal limit is 40 per hour), bottles and dung being thrown into bushes to entice the animals out, and widespread littering. It is also not uncommon to see jeep ‘jams’ caused by the frenzied use of mobile phones to spread the word about a wildlife sighting. Bad behaviour by jeep drivers, especially at sightings of charismatic species such as leopards, inevitably diminishes the tourist experience, as evident from Figure 8.5.

245 As described in Chapter 7. 246 British Broadcasting Corporation.

308

Figure 8.5: Chaos at a leopard sighting at Yala National Park (photograph taken on 09/03/2011, by S. Kalpage)

A recent World Bank survey gathered the opinions of almost 2,000 tourists, both international and local. Interview questions included those on respondents’ satisfaction levels regarding parks they visited247. I obtained access to the raw data from the survey and analysed the sample of tourists that was most relevant to potential MERs, i.e. foreign tourists who visited PAs within the southeast landscape248. My results are summarised in Tables 8.2 and 8.3 (further details are presented in Appendix 8).

247 However, 80% of the sample, did not visit PAs. This is, in fact, slightly less than the percentage of tourists who visit PAs, so the sampling is reasonable. Appendix 8 presents more details on the survey. 248 The size of this segment was 119 people.

309 Table 8.2: Responses from foreign tourists who visited Bundala, Uda Walawe and Yala National Parks (analysis of data from World Bank 2009)

Number of responses (n=119) Factor 1 = very 3 = 2 = bad 4 = quite good 5 = very good bad neutral Wildlife species 25 26 21 41 5 diversity Congestion 41 26 30 19 3 Knowledge of guide 36 18 34 14 16 Accommodation 31 30 29 22 7 Diversity of 42 36 25 10 6 activities249 Friendliness of staff 30 22 22 39 6 Transport experience 32 25 22 18 22 Value for money 27 38 20 23 10 Overall experience 36 19 23 24 16

Table 8.3: Summary of bad and good responses from foreign tourists who visited Bundala, Uda Walawe and Yala National Parks (summary of results of Table 4)

Bad scores Good & neutral scores Factor Difference (i.e., 1, 2) (i.e., 3, 4, 5) Wildlife species diversity 51 67 16 Congestion 67 52 -15 Knowledge of guide 54 64 10 Accommodation 61 58 -3 Diversity of activities 78 41 -37 Friendliness of staff 52 67 15 Transport experience 57 62 5 Value for money 65 53 -12 Overall experience 55 63 8

249 This refers to the range of activities allowed in the PA. For example, in some parks in southern Africa, there are product offerings such as walking safaris, hot air ballooning, canoe safaris, etc. As evident from the responses, there are few of these currently offered in Sri Lankan PAs.

310 Table 8.2 summarises foreign tourists’ responses to a series of questions designed to assess their park visits. Interviewees were asked to assess the factor addressed in the question on a scale from 1 to 5—1 and 2 were negative responses, 3 was neutral and 4 and 5 were positive responses. In Table 8.3, the difference between the positive plus neutral response and the negative responses are calculated, to arrive at (an admittedly crude) measure to assess whether the overall response for that question is positive or negative overall. The results are striking. The diversity of wildlife, friendliness of staff and the overall experience were the top three positive categories, while the diversity of activities on offer, congestion in parks and the value for money are the three most negative. In addition, the overall quality of accommodation was also judged to be poor. This feedback provides valuable insights on what could be done to improve the quality of tourists’ experiences in Sri Lankan PAs, especially those in the country’s southeast.

Along with this feedback, it is important to understand the key characteristics of the target customer segment of high-end nature tourists (or eco-tourists). The IUCN stresses the importance of customer segmentation and the need to understand the particular needs of target segments, to sustainably maximise tourism potential (Eagles et al. 2002; p.21-22):

There is no such thing as the ‘average PA visitor’. In reality, markets comprise many segments, each of which has somewhat different characteristics, expectations, activity participation and spending patterns. Marketing exploits these visitor segments by comparing and matching them with the biophysical and cultural attributes of the park, and then sensitively promoting appropriate PA attributes to the targeted segment. This reduces adverse impacts on the PA, increases the economic benefits and makes it more likely that visitors are satisfied. The value of segmentation is that it can predict behaviour, and thus help managers to plan for this behaviour. Segmentation by perceived product benefit can be used to develop an understanding of what tourists really seek in a visit to a PA, and so establish an appropriate management response. In this way, visitors will gain greater satisfaction from the products and services offered. Although the study points out that sophisticated research capabilities are normally required to understand the market in this way (Eagles et al. 2002; p.21), we could draw useful insights from a study commissioned by the Sri Lanka Tourist Board that identified characteristics of high-end eco-tourists (Sri Lanka Tourist Board 2003; p.7-8):

• Income: higher than the average traveller;

311 • Occupation: professional, technical, managerial;

• Education: well-educated and interested in learning;

• Age: all ages are interested, with general/occasional eco tourists tending to be younger, and frequent/experienced eco tourists tending to be older;

• Party size: mainly couples, with a substantial number travelling alone; and,

• Season: most European and North American tourists prefer to travel in summer, but there is more winter season interest among eco tourists than there is among general travellers.

According to the report, the international ecotourism market is centred on North America and Europe, with most of the travellers coming from the US, UK, Germany, Canada, France, Australia, the Netherlands, Sweden, Austria, New Zealand, Norway and Denmark (Sri Lanka Tourist Board 2003; p.14). However, Japan, Southern Europe and the newly industrialised Asian countries are also generating increasing numbers of eco-tourists (Sri Lanka Tourist Board 2003; p.14). The report also includes results from a comprehensive survey conducted in important ecotourism markets from the Sri Lankan perspective such as US, UK and Australia (Sri Lanka Tourist Board 2003; p.8- 11). The preferences indicated in this survey—regarding motivating factors for the trip, and choices regarding activities, accommodation, and food—are summarised in Appendix 8.

The Sri Lanka Tourist Board study points out that high-end tourists have a global outlook and choose from a wide array of potential countries and operators when planning their holidays (Sri Lanka Tourist Board 2003; p.14). In particular, this means that Sri Lanka has to compete with well-established wildlife destinations such as Kenya, which already have sophisticated product offerings:

Kenya is one of the leading tourism destinations in sub-Saharan Africa, besides South Africa, Tanzania, and Uganda. Key visitor activities in Kenyan wildlife preserves include game viewing, photography, adventure travel, balloon safaris, walking, lake boating, guided tours, river rafting, horse riding, fishing, cycling, etc. Further, Kenyan parks and reserves provide specialized camping/lodging facilities in divergent locations satisfying a wide range of tourist interests,

312 therefore providing convenient forums for excursions and over-night stays (Odunga and Maingi 2011; p.28). Consequently, it is important that MERs offer a wide range of product offerings—far beyond those currently offered at Sri Lankan PAs—to compete successfully with nature tourism destinations such as Kenya.

While ecotourism could potentially bring economic benefits that could fund biodiversity and habitat conservation, there have been many instances where it has damaged the very resources it was intended to protect:

In Canada, tourists are alleged to harass polar bears by approaching too closely. Whales have been harassed and even killed in Quebec and the Canary Islands. Wildlife observers drive cheetahs off Kenyan preserves, exposing the cats to danger and the risk of inbreeding. Sea turtles are distracted by electric lights at shore-side tourist facilities (Isaacs 2000; p.64). Most of these cases occur when the concept of ecotourism has been applied incorrectly, and there has been inadequate planning and monitoring of ecotourism ventures. It is important, therefore, to be aware of potential detrimental impacts of the ecotourism, so steps could be taken to prevent or minimise them. Eagles et al. (2002; p.33) point out that the construction of accommodation, visitor centres, infrastructure, and other services often have direct impacts on the environment, through the removal of vegetation, impacts on drainage, and disturbance to animals and their habitats. Some tourism companies, however, such as the Sabi Sands game reserve, have managed to establish operations that help conservation and minimise environmental impacts (Eagles et al. 2002; p.67):

Sabi Sands is an 800 ha private game reserve, adjacent to Kruger National Park. It has three operating lodges. An artificial wetland is a key part of the sewage treatment system. Sewage is collected by gravity feed to a three-chambered 10,000-litre holding and separation tank. Sludge is pumped out and trucked away, as needed. Liquids are pumped to settling ponds, and electrified fences protect adjacent vegetation. Water drains gradually through plant roots to a small swamp area. The ponds and swamp support a range of water birds, including the saddle-billed stork, which feeds on frogs. In addition to these direct impacts of poorly conceived and managed tourism, there could be negative social and economic impacts on local communities. Brockington et al. (2008; p.138) argue that while “ecotourism can offer opportunities for economic

313 diversification, it can also exacerbate existing resource management conflicts that are rooted in the historical context of local power relations”.

Given the findings from the World Bank survey, the Sri Lanka Tourist Board study, and the general literature on eco tourism, several steps should be taken to enhance tourism potential within MERs and mitigate potential negative impacts:

• Develop a diverse and compelling product offering: Government policy makers—especially in the Forest Department which manages most of the land within MERs, and the Tourism Development Authority—should work with tourism operators to develop a diverse, compelling range of activities that would meet expectations of their well-travelled, sophisticated target customer segment. Apart from the game drives offered in regular PAs, activities such as walking safaris, night drives, balloon safaris, and canoe trips on local waterways should be considered.

• Enhance the customer experience: As noted in the Sri Lanka Tourist Board survey, most eco tourists seek learning opportunities. Therefore, guides should be provided with structured, comprehensive training that increases their knowledge regarding animal behaviour and ecology, local history and culture, and equips them with essential language skills to communicate effectively with their clients250. Ideally, as many locals as possible should be provided with opportunities to train as guides, since they would already have some local knowledge; this would also help build local support for MERs.

• Establish accommodation that meets customer requirements: Relevant government authorities should work with tourist operators to develop a range of world-class, environmentally friendly, accommodation within MERs, including well-designed lodges and camping operations.

• Take steps to safeguard natural assets: The Forest Department should work with the Department of Wildlife to develop a set of rules governing operator and visitor behaviour within MERs, and strictly enforce them to provide tourists

250 While English is the most important language given its almost universal use among high-end eco tourists, guides could be provided with opportunities to learn additional languages such as German, French or Japanese.

314 excellent opportunities to enjoy nature and view wildlife with minimal disturbance to animals and the environment. A portion of revenues from MERs should go towards activities that enhance the conservation value within the areas—the rehabilitation of degraded habitats, for example.

To successfully implement these measures, the government agencies taking the lead in implementing MERs—the Forest Department and the Department of Wildlife—should work closely with other important stakeholders. One group is ‘other government agencies’ such as the Tourism Development Authority to help design the overall tourism experience within MERs and include these areas within Sri Lanka’s tourism marketing plan; and the Economic Development Ministry to help tourism operators— especially those from other countries—establish local business operations, providing them with various incentives if they meet certain criteria. Another group are ‘local communities’, whose support is vital to the success of the MERs in general, and enhancing the tourism experience in particular. We will consider this group in more detail later. Finally, the implementing agencies would need to work closely with potential tour operators to develop compelling tourism products, including activities and accommodation.

Credits for carbon sequestration

As calculated in Chapter 7, there are significant potential benefits from carbon sequestration, which represents about US$ 11 million of economic profit annually— within Sri Lanka’s southeast landscape. Much of this potential would be accessible if MERs are established, since they include much of the land used in my calculation251. At present, however, Sri Lanka is some years away from being able to access these revenues. As discussed in Chapter 7, there are three phases of REDD implementation according to the international agreement reached in 2010 at Cancun:

1. Readiness: governments develop the institutional capacity to design REDD+ strategies through stakeholder dialogue, institutional strengthening, and demonstration activities;

251 As described in Chapter 7, I only considered forested land outside the current PA network. Most of this area would be included in potential MERs.

315 2. Implementation of a National REDD+ Strategy: the national government in the developing country accesses REDD+ finance on the basis of results that are not necessarily tied to emissions reductions; and,

3. Full scale REDD+ implementation: the national government is ready to participate in performance-based funding, in which flows of REDD+ revenues are directly tied to measured and verified changes in greenhouse gas emissions from deforestation, forest degradation, and forest carbon enhancement.

Like most developing nations, Sri Lanka is currently in Phase 1252, and has taken several steps to prepare for the implementation of REDD+. The Sri Lankan government ratified the UNFCCC in 1993 and the Kyoto Protocol in 2002 (Yamane 2003; p.5). As a party ratifying the UNFCCC, Sri Lanka was obliged to prepare a national communication report (Yamane 2003; p.5). In its ‘Initial National Communication’, submitted to UNFCCC in October 2000, the country identified nine sectors considered by experts to be most vulnerable to the impact of global climate change in Sri Lanka; one of them is the forestry sector253 (Ministry of Environment Sri Lanka 2000; p.xi). These were consolidated into five sectors 254 in the ‘National Climate Change Adaptation Strategy for Sri Lanka 2011 to 2016’, and sector vulnerability profiles were created for each of these sectors (Ministry of Environment Sri Lanka 2010; p.5). In January 2012, the Ministry of the Environment submitted a ‘Readiness Preparation Proposal’ to the UN-REDD Policy Board, which sets out a comprehensive plan to steer the country through the first of the three phases of REDD+, the Readiness Phase (Ministry of Environment Sri Lanka 2012; p.5).

Despite this progress, Sri Lanka has considerable work ahead of it before it is adequately prepared for REDD+. Mattsson et al. (2012; p.37) describe how information deficiencies constitute a major gap the country needs to address:

Sri Lanka currently has difficulties producing a robust Tier 2255 reference level for deforestation (even more so for degradation, carbon enhancement

252 Only Brazil, Guyana, Indonesia, and Tanzania are currently in phase 2, while no nations are yet in phase 3. 253 The others were: energy, industry, transport, agriculture, water resources, the coastal zone, health, human settlements and public utilities. 254 Agriculture and fisheries, water, health, urban development, human settlements & economic infrastructure, biodiversity and ecosystem services. 255 The different tiers is a set of guidelines for estimating greenhouse gas inventories at different levels of quality, from Tier 1 (simplest to use; globally available data), to Tier 2 (nationally derived data which

316 and reforestation) due to lack of consistent and repeated national inventories of the extent of forest cover… In order to produce a robust reference level for Sri Lanka it is thus imperative to carry out new analyses of the extent and changes in forest area, using recent and historic data on forest cover and forest cover change that is consistent. This calls for technical advice, training, and capacity- building to develop and use high-tech methods, such as satellite remote sensing in Sri Lanka, underpinned by financial support, something that have been stressed by Sri Lanka in the latest UNFCCC submission on the issue. Local expertise also needs to be used to account for national circumstances and validate the outcomes of methodological assessments. In October 2009, Sri Lanka was granted observer status on the Policy Board of the UN- REDD program, providing it with access to networking and knowledge sharing (Mattsson et al. 2012; p.31). Consequently the UN-REDD program is supporting Sri Lanka with US$ 4 million for initial readiness activities, and a REDD+ National Joint Program was set up in 2011 to help develop an effective REDD regime (Mattsson et al. 2012; p.31). This assistance would help the country gradually move toward full REDD+. Although the global regime will take time to evolve, there may be opportunities for Sri Lanka to sign bilateral agreements with developed countries such as Norway, which— as described in Chapter 7—has signed agreements with countries like Brazil and Indonesia.

Improved local livelihoods and local support for conservation

As pointed out in Chapters 4 and 5, many locals in the southeast landscape have extremely negative attitudes towards elephants because these animals pose significant threats to their property, livelihoods, and sometimes even their lives. As discussed in Chapter 1, there is ample evidence that the success of conservation efforts could be seriously threatened if not supported by local communities. Eagles et al. (2002; p.31-32) describe how PA-related tourism can negatively affect local livelihoods, eroding local support:

Increased numbers of tourists may disturb community activities, and compete for recreation places and other services. Poorly planned tourism development can lead to increased congestion, littering, vandalism and allow for more precise estimates where changes in carbon stock are calculated), and Tier 3 (high resolution methods specific for each country and repeated through time) (Barber et al. 2004; p.162-165).

317 crime. Governments may exacerbate these problems if they put short-term economic considerations before all else, for example by building inappropriate infrastructure or failing to establish the needs of local communities. When this happens, the local support for the PA may be put at risk. Where PA agencies develop visitor management regulations that also affect local residents, there may be negative socio-cultural impacts (e.g. prohibitions on traditional uses such as fuel wood gathering or on spiritual uses which require entry to the PA). Other negative impacts may occur where local traditions become commercialized, and lose their integrity or authenticity. An example would be dances, which had once had a vital social role but which are now put on only for the entertainment of visitors. Therefore, steps should be taken to enable local communities to benefit from MERs; at the very least, locals should not be negatively affected. Potential measures that could be taken in this regard include:

• Providing income-generating opportunities for locals: Tourism companies should be provided with incentives to hire and train locals as guides and to obtain food and other supplies locally—chena farmers who adhere to prescribed cultivation practices within MERs should be given preference. Tourism companies could even work with farmers to agree on which crops they should cultivate and undertake to buy a certain portion of their harvest, thus eliminating the middlemen who often exploit poor, indebted farmers. The Forest Department could also hire locals to help maintain MER-related infrastructure—to construct and maintain electric fences for example.

• Sharing some economic benefits with poor locals: The Forest Department should set aside a portion of MER revenues—from tourism concessions and carbon credits for example—to help benefit local communities. It could work out arrangements with village societies to use these funds to construct essential infrastructure such as schools and piped water, and provide additional insurance for victims of human-elephant conflict.

• Establishing community education programmes: The Forest Department and tourism companies should collaborate to educate locals about the importance of conservation of natural assets—especially those providing valuable ecosystem services.

318 Fisher et al. (2008; p.102) point out that these types of measures—especially if implemented with adequate consultation with the relevant groups 256 —have both practical as well as ethical benefits:

The rationale for including multiple stakeholders in resource management decisions has a pragmatic aspect (including people who carry out natural resource management in decision making increases the likelihood that they will modify their practices in conformity with stakeholder agreements) as well as an ethical one (including people who will be affected by forest management decisions will make it more likely that decisions will reflect their needs and interests). In establishing measures that benefit local communities, Sri Lankan policy-makers should look at experiences of their counterparts in other countries, especially because many of them have been grappling with similar issues. For example, Southern African conservationists have recognised the importance of local community support for conservation for many decades (Scheepers et al. 2011; p.1). This was particularly necessary in South Africa, where parks were perceived by locals as being managed solely for the benefit of a white minority, and therefore PAs were at risk of being converted to other uses by the African National Congress after it gained political power in 1994 (Child 2009; p.23). Therefore, local community ‘upliftment’ is now among the three fundamental objectives of South African national parks (or ‘SANParks’), which are: “firstly, to conserve the country’s biodiversity; secondly, to maintain a relationship of community ‘upliftment’ and capacity building (entrepreneurial, improving quality of life, etc.) among people living in the areas in and around the parks; and lastly, to provide a recreational outlet for people to experience and enjoy the wonders of the parks” (Saayman and Saayman 2010; p.1038).

SANParks has a division called ‘People in Conservation’ covering a range of functions, including indigenous knowledge management, community relationship building, local economic empowerment, environmental education, and interpretation and training (Swemmer and Taljaard 2011; p.2). This division has pioneered initiatives such as the ‘Extractive Resource Use Programme’ where local communities are allowed to harvest certain natural resources from national parks such as tissue from the pepper bark tree which are used in traditional medicines (Scheepers et al. 2011). An adaptive

256 As described in detail in section 8.3.

319 management approach257 is being used to establish sustainable harvest levels (Scheepers et al. 2011). Saayman and Saayman (2010), who studied four South African national parks, found that these programmes have had positive impacts: “All the national parks have a very positive impact on their communities. The parks also affect the communities’ personal quality of life positively, indicating the positive perception that the communities have of the relevant national parks” (Saayman and Saayman 2010; p.1048).

Southern African countries have taken community participation a step further through pioneering Community Based Natural Resource Management (CBNRM) programmes that have devolved wildlife management rights to local communities. Child (2009; p.188) describes the main objectives of these programmes, which have given locals a direct stake in the success of conservation initiatives:

The CBNRM ‘model’ as understood in southern Africa is a multifaceted approach that combines economic, political and institutional goals. Economically its objective is to provide institutions of collective action that improve the productivity of non-agricultural systems by allowing higher-valued land uses, such as wildlife, to be adopted. Zimbabwe’s CAMPFIRE258 programme, which was started in the late 1980s, has been used a model for subsequent initiatives in the region and elsewhere (Frost and Bond 2008; p.776). This programme involves the sale by rural communities of rights to access wildlife to entrepreneurs, who in turn organise safaris for hunters and eco- tourists (Frost and Bond 2008; p.777; Taylor 2009). Namibia has a growing number of similar initiatives—by the end of 2006 there were 13 formal joint venture tourism enterprises in community conservancies (Jones and Weaver 2009; p.229).

Despite the overall success of some of Southern Africa’s CBNRM programmes virtually all of them have struggled with significant issues, especially during initial stages. There have been some instances of mismanagement of funds in Zimbabwe’s CAMPFIRE programme (Balint and Mashinya 2008; p.792). In Botswana, community entities, with their collective decision-making structures, have had difficulties in running successful businesses (Rozemeijer 2009; p.254). In Namibia, the lack of secure land tenure has hampered the CBNRM process in some areas (Jones and Weaver 2009;

257 This approach will be discussed in more detail in section 8.3. 258 The acronym for ‘Communal Areas Management Programme for Indigenous Resources’.

320 p.240). Consequently, it is important to keep in mind that communities are complex entities, often comprising multiple actors with differing priorities and objectives:

The term community in community-based conservation is gloss for a complex phenomenon because social systems are multi-scale and the term community hides a great deal of complexity. Idealized images of coherent, long-standing, localized sources of authority tied to what are assumed to be intrinsically sustainable resource management regimes are just that— idealized. As many conservationists know, it is often difficult to find a cohesive social group to work with in the field. Communities are elusive and constantly changing. A community is not a static, isolated group of people. Rather, it is more useful to think of communities as multidimensional, cross-scale, social-political units or networks changing through time (Berkes 2004; p.623). Therefore, implementing community-based programmes requires patience and persistence from government organisations and other stakeholders. Furthermore, an adaptive management259 approach should be adopted since CBNRM is “a complex, long-term process that stretches beyond the time-frame and imagination of donor projects” that are often used to implement them (Child 2009). Although it would be difficult to establish CBNRM programmes in Sri Lanka at this point in time, the above examples could provide policymakers with insights on how MER governance could evolve in the future.

In terms of the more immediate need to allocate a portion of funds generated from MERs to local initiatives, however, Sri Lankan policy-makers could consider mechanisms that have been used for this purpose in other countries. One way is to encourage tourism operators within MERs to contribute to local development. For example, a private camp near South Africa’s Addo Elephant National Park “has specifically allocated a percentage of its turnover to support a local community trust fund, enhancing nature-based tourism opportunities and ‘upliftment’ in previously disadvantaged communities” (Castley et al. 2009; p.316). Alternatively, a government- administered levy, incorporated into the legal framework governing MERs, could be instituted, as in India260:

259 Roux and Foxcroft (2011; p.1) describe adaptive management as “learning-by-doing in a scientific way, adapting behaviour and overall direction as new information becomes available”. This approach is discussed in more detail in section 8.3. 260 This was included in the recently released set of ‘Guidelines for ecotourism in and around protected areas’ by India’s Ministry of Environment and Forests.

321 As part of the state-level ecotourism strategy, the state government should levy a ‘local conservation cess261’ as a percentage of turnover, on all privately-run tourist facilities within 5 km of the boundary of a PA. The rate of cess should be determined by the state government, and the monies thus collected should be earmarked to fund PA management, conservation and local livelihood development, and not go to the state exchequer. The rationale for a local conservation cess should be clearly explained to the public at large, including through clear signage at local tourist facilities. (Government of India 2011; p.6). Costa Rica—a renowned eco-tourism destination—extensively used environmental education to build public support for its sustainable tourism initiatives. This was necessary, because when the national park and wildlife reserve system started in the 1960s, there was a general lack of understanding among the public of the need to conserve the country’s natural heritage (Eagles et al. 2002; p.30). The government developed public understanding and appreciation of Costa Rica’s exceptional biodiversity through a number of initiatives such as (Eagles et al. 2002; p.30):

• Encouraging park visitation by residents (valuing through experience),

• Channelled entry to education facilities in prominent national parks (e.g. via construction of an interpretive visitor centre at the entrance to Volcan Poas National Park),

• Interpretation on site (e.g. through displays, materials and interpreters),

• A national school environmental education programme (ensuring future generations understood the country’s natural heritage), and,

• Encouragement of school visits (through ‘active schools programmes’ introduced by some parks).

Although Sri Lanka’s entire PA network would benefit from similar programmes, the Forest Department could develop education initiatives associated with MERs in the southeast landscape, which could serve as a pilot programme for future initiatives in other areas.

261 This is essentially a conservation tax. ‘Cess’ is part of the jargon used by the Indian bureaucracy to indicate any type of tax, e.g. ‘irrigation-cess’, ‘educational-cess’.

322 8.2 Potential challenges to implementing Managed Elephant Ranges

Despite the many potential benefits of MERs, described in Chapters 4, 5, 6 and 7, it may not be easy to implement them successfully in the Sri Lankan context. Potential challenges include: (a) difficulty in convincing government policy-makers of the economic viability of MERs; (b) deficiencies in key government agencies involved in conservation; and (c) inadequate provisions in Sri Lanka’s legislative framework to encourage conservation outside state PAs. I now explore these challenges in further detail and discuss potential approaches to address them, based on lessons from other countries.

(a) Difficulty in convincing government policy-makers of the economic viability of MERs

The Sri Lankan government has made rapid economic development its priority since the end of its long running civil war in 2009. In his 2010 election campaign, the President pledged to double per capita income over the following six years, maintaining a continuous growth rate of 8% per annum (Rajapaksa 2010; p.3-4). The government’s 2006 – 2016 development framework envisages significant infrastructure development, including rehabilitating existing irrigation schemes and constructing new irrigation schemes (Government of Sri Lanka 2005). These activities would put pressure on existing elephant habitat262, but since government policy makers face competing demands for land-use, they would need to be convinced that allocating land for MERs would be economically viable and socially beneficial.

Experiences of nature tourism destinations in Southern and Eastern Africa, Southern and Central America and some Asian countries could provide Sri Lankan government- policy makers with examples of economic and social benefits of conservation landscapes, both within and outside PAs. As discussed in Chapter 7, while increased nature tourism is the most obvious benefit from MERs, emerging sources of revenue such as carbon credits could also be tapped. Moreover, conservation landscapes provide vital ecosystem services that are seldom accounted for within conventional economic

262 As described in Chapter 4.

323 measures such as GDP. I now discuss briefly how these various types of benefits have made conservation an economically viable form of land-use in several countries.

Table 8.4 illustrates the valuable economic contribution of tourism in some well-known nature tourism destinations. Tourism’s contribution Sri Lanka’s economy is significantly less than in all the other countries, but on par with that of South Africa. However, South Africa has much larger, diversified economy than Sri Lanka, and, in any case, the magnitude of tourism’s economic impact is significantly higher263. Therefore, Sri Lanka has an opportunity to increase its economic contribution from tourism, especially given its significant cultural and natural assets.

Table 8.4: The direct and indirect economic contribution of tourism in leading nature tourism destinations compared to Sri Lanka in 2010 (adapted from WTTC 2010)

Direct contribution Indirect contribution Country Jobs: % of total Jobs: % of total % of GDP % of GDP employment employment Belize 12.0% 10.9% 33.2% 30.1% Botswana 2.4% 3.1% 6.5% 7.6% Costa Rica 4.8% 4.5% 12.3% 11.4% Kenya 5.7% 4.8% 13.7% 11.9% Namibia 4.4% 6.8% 20.3% 27.1% Rwanda 3.3% 2.8% 8.4% 7.3% South Africa 2.7% 3.9% 8.6% 9.0% Sri Lanka 3.4% 3.0% 8.4% 7.5% Tanzania 5.0% 4.2% 13.3% 11.6% Zimbabwe 5.7% 11.7% 4.2% 9.2%

The differences between the impact of nature tourism in Sri Lanka and many of these countries are probably as stark, since less than 20% of tourists who arrive in Sri Lanka visit a national park, although the country has an extensive PA network with charismatic wildlife such as elephants, leopards and bears (SLTDA 2009; Department of Wildlife Conservation website 2010). The corresponding number for South Africa is close to 50%, indicating that Sri Lanka could make much better use of existing PAs for nature-based tourism (South African Tourism 2011; p.58). There is also tremendous

263 Tourism has a direct impact of US$ 9.5 billion in South Africa compared to US$ 1.7 billion for Sri Lanka (WTTC 2010).

324 potential for nature tourism outside national parks. Currently, tourism in forested areas outside PAs is negligible, but, as described in Chapter 2, the Forest Department manages forest reserves covering 23% of the country264—these areas contain many bird species and animals such as elephants and leopards that range widely outside PAs (Survey Department of Sri Lanka 2008).

Southern African countries have convincingly demonstrated that wildlife-based tourism is an economically viable land-use option outside national parks. For example, Bothma et al. (2009; p.149) estimate that, in 2007, wildlife utilisation enterprises outside PAs covered 16.8% of South Africa’s land area, compared to only 6.1% in state PAs265. This success stems from the South African government’s willingness to involve the private sector in conservation, and its practice of considering the broader landscape around PAs, as demonstrated in its initiative to expand the Addo Elephant National Park (Castley et al. 2009). The programme’s overall goal was “the expansion of the park into a mega- biodiversity conservation area, with strong economic links with the surrounding region”, and was undertaken with significant private sector involvement (Castley et al. 2009; p.309). This project demonstrated that establishing conservation-related enterprises are significantly more economical than small stock farming, the area’s main alternative form of land-use. It resulted in direct employment increases of about one job per 100 ha—compared to one job per 367 ha for small stock farming—and contributed over US$ 10 million to the local economy (Castley et al. 2009; p.310).

Namibia has had similar experiences. Barnes and Jones (2009; p.118) calculated that in 2001, 15 – 25 per cent of total freehold farmland was used primarily for commercial game production activities such as game ranching, safari hunting, and non-consumptive wildlife viewing. The conservancy approach266, which is supported by the Ministry of Environment and Tourism, has increased in popularity among freehold farmers—in 2006, at least 22 conservancies existed on freehold land in Namibia, double the number in 1998, covering an area of close to four million ha (Barnes and Jones 2009; p.120). The practice of removing internal fences between properties when creating conservancies has biodiversity advantages, as it enables owners to introduce large

264 In addition to the PA network covering 14% of Sri Lanka’s land area (Ministry of Environment Sri Lanka 2012). 265 This includes national PAs as well as provincial ones. 266 A conservancy is “a group of adjoining wildlife ranches that operate under a cooperative management agreement based on a shared common goal such as wildlife conservation” (Bothma et al. 2009; p.157).

325 species—such as elephant, rhino and lion—that are not viable on smaller parcels of land (Bothma et al. 2009; p.157). Furthermore, as contributing landholders jointly manage conservancies, the operation has access to a wider pool of skills, experience and knowledge.

In addition to these Southern African countries, some Central American nations also have encouraged conservation beyond state PA networks. Costa Rica’s successful system of private reserves that complements its government-managed PA network, includes the Monteverde Cloud Forest Reserve, which has been cited as an example of a successful, financially self-sufficient, privately managed PA (Eagles et al. 2002; p.35). Other countries in the region—such as Belize, Honduras and Guatemala—have also established several private reserves.

Finally, Bhutan is perhaps the best known Asian country to effectively, and sustainably, tap the high-end tourist segment, as described by McIntyre (2011; p.15):

Bhutan’s long-term strategy of controlled tourism with a focus of sustainability and quality has secured the country’s reputation as an exclusive and distinctive destination while ensuring long-term sustainability of the industry and its contribution to the economy… Aware that an unrestricted flow of tourists could negatively impact Bhutan’s pristine environment and unique culture, the government adopted a policy of ‘high-value, low-volume’ tourism in order to control the type and quality of tourism right from the start. “The principle of high-value, low-volume tourism development, guiding tourism’s growth in Bhutan, is highly commendable and has undoubtedly contributed to the unique tourism brand of this country”, said the World Tourism Organisation Secretary General, Mr. Taleb Rifai, during an official visit to the country in January 2011. While nature-based tourism is often the most obvious benefit from conservation landscapes, these areas also provide valuable ecosystem services267. These include benefits such as flood control, carbon sequestration and storm protection that are increasingly being recognised by governments, especially those in Southern Africa (Balmford et al. 2002; p.950). For example, SanParks explicitly acknowledges carbon sequestration benefits from PAs in its annual report (SANParks 2010; p.8):

267 As discussed in Chapter 7.

326 Through the existence and proper management of the approximately 4 million ha of land under the management of SANParks, the organisation continues to contribute immensely to the ‘Green Economy Agenda’. There is research globally that indicates that pristine thicket areas within South Africa hold more carbon compared to transformed or over-grazed land and as national parks are some of the most pristine natural environments of the country, potential for carbon sequestration is definitely higher in these areas. Although these ecosystem services provide valuable benefits to countries, they are usually ignored in traditional economic measures such as GDP. However, as discussed in Chapter 7, there are some options for developing-country governments to obtain monetary benefits for such services through emerging schemes operated on the principles of ‘payment for ecosystem services’. Carbon sequestration is the most obvious, and as discussed previously, countries such as Indonesia and Brazil have signed agreements worth hundreds of millions of dollars in this regard with developed countries such as Norway. Moreover, countries such as Costa Rica and China have developed systems of payments for other ecosystem services such as watershed management and biodiversity (Pagiola 2008; Huang et al. 2009).

Overall, countries such as South Africa, Costa Rica, Bhutan, and Namibia have shown that wildlife and conservation could be an economic form of land-use, and have demonstrated the importance of doing so, especially in developing countries where there is fierce competition for land and need for economic development. As Bond et al. (2004; p.39) point out:

Making conservation pay is particularly important in countries where there is a high demand for land. Secondly, the economic success of wildlife production not only contributes to biodiversity but it legitimizes wildlife as a primary form of land-use in that it provides things that are important in developing countries: economic growth and jobs. This is important because it makes conservation politically palatable. When assessing the profitability and viability of different forms of land-use, policy- makers in the region have realised the importance of identifying and addressing unfair subsidies. For example, Bond et al. (2004; p.35) describe the effects of the unfair subsidisation of the cattle industry in Zimbabwe:

There were direct and indirect subsidies to cattle producers that limited any form of investment in wildlife management. Direct measures included price support, price controls, cheap loans for restocking, provision of fodder in droughts, and policies that provided cheap labour. Indirect

327 subsidies to livestock included massive state investments in the control of endemic livestock diseases, appropriate infrastructure (abattoirs, fencing), research and development. One of the most iniquitous policies in Zimbabwe was the compulsory de-stocking of communal lands that allowed commercial producers to access communally produced livestock at below-market prices. As pointed out in Chapter 7, there are similar subsidies in Sri Lanka, where farmers receive subsidised fertilizer and weedicides, and have free access to water from state- sponsored irrigation schemes (Sivathasan 2010). These types of subsidies should be eliminated where possible, to prevent landowners unfairly favouring some forms of land-use over others.

So, in summary, illustrating the economic and non-economic benefits derived from conservation activities by countries in Southern Africa, Southern and Central America, and parts of Asia, could help supporters of MERs convince high-ranking Sri Lankan government officials that setting aside these areas could be economically and socially beneficial (and therefore politically palatable). These examples could be supported by calculations of the specific potential within the southeast landscape (such as those presented in Chapter 7). ‘Best practice’ case studies could provide insights to what needs to be done to realise this potential, such as the elimination of perverse subsidies (as described above), and legislative reform (as will be discussed shortly).

(b) Deficiencies in key government agencies involved in conservation

Recent studies on Sri Lanka’s government institutions in general, and on those agencies involved in conservation in particular, have identified many deficiencies within these bodies in terms of processes, skills and mind-set. Sri Lankan bureaucrats are susceptible to political interference and are often reluctant to involve the private sector and civil society in policy formulation (ADB 2004; Samaratunge et al. 2008), while an Asian Development Bank review of governance and public management identifies deficiencies in staffing, budgeting and policy development:

The practice of staffing ministries by a generalist bureaucracy—with a practice of unplanned transfers—has contributed to the virtual absence of a continuing institutional memory in the policy-related institutions.

328 … [T]he absence of effective policy management has its consequences on the activities of the planning of interventions and of activities. It has also a direct impact on the country’s budgetary processes—at all levels of the polity. The budgets are not necessarily guided by a policy framework but tend to be governed by historical precedents. The outcomes are budgets that are oriented more towards recurrent expenditure than to investment expenditure. Also, in the absence of coherent policy agendas, development initiatives as well as service delivery interventions tend to be ad hoc (ADB 2004; p.8). A recent World Bank project noted several issues with Sri Lanka’s EIA268 process and institutions responsible for EIA implementation. These included limited staffing capacity, inadequate organisational commitment, poor scoping processes and identification of issues, weaknesses in methodology and analysis, inadequate consideration of alternatives, improper environmental management and monitoring plans, and inadequate financial resources for monitoring (World Bank 2012; p.14-27). It specifically referred to deficiencies within agencies that would be tasked with implementing the proposed MERs: “the Department of Wildlife and the Forest Department have substantial environmental expertise, but lack an organised and dedicated approach to EIA process management” (World Bank 2012; p.14).

A similar study on the Central Environmental Authority that manages the overall legislative framework for the protection, conservation and management of the environment in Sri Lanka, revealed that the agency has weak enforcement tools, lacks a proper monitoring framework and avoids using market-based instruments. The study concluded that the “absence of economic or incentive-based instruments does not allow the Central Environmental Authority to leverage market and community forces and access additional financial resources to complement a relatively weak command-and- control compliance and enforcement system” (AECEN 2006; p.14).

These types of deficiencies in key agencies that would be responsible for implementing MERs—such as the Forest Department and the Department of Wildlife, and supporting agencies such as the Central Environmental Authority—could hamper implementation. The inability, and, at times unwillingness, of these agencies to work with important stakeholders such as other government departments, the private sector and local communities would make it difficult to obtain required buy-in and support. The lack of staff with requisite skills to manage the complex process of establishing these areas, and

268 Environmental Impact Assessment.

329 the extremely limited availability of financial and other resources would be significant constraints. Therefore, Sri Lankan policy-makers should look at lessons learnt in other developing countries—especially those in southern Africa—that have managed to resolve similar issues.

Child et al. (2004) point out that many park agencies in Southern Africa historically suffered from similar problems. Like Sri Lanka, these countries were colonies of western powers, and initial conservation efforts were undertaken by and for a small white minority (Spenceley 2008; p.70). After independence, the quality of staff in most conservation-related agencies deteriorated in terms of skill-set and professionalism and the agencies became inefficient, bureaucratic, overly-centralised, political and suspicious of the private sector (Child et al. 2004; p.126). As budget-funded public institutions, they lacked accountability and a clear set of objectives. Responding to an ever-wider set of stakeholders, they were unable to concentrate on their core competencies as they tried to placate everyone (Child et al. 2004; p.130). However, as available state funding for PAs has declined, there have been many innovative institutional and organisational experiments in Southern Africa to improve the financial viability of responsible agencies, and encourage them to build partnerships with local communities, whose support is essential for successful conservation. A key step in this process has been the creation of park parastatals, whose main characteristics are described by Eagles (2002; p.139):

Generally the trend is for governments to demand that parks earn much higher amounts of their budget from tourism sources. Corresponding to this is the development of forms of management, such as parastatals, that allow for park agencies to function with the efficiencies of a private corporation. A parastatal is an independent corporate body within government that makes its own policy, maintains internal financial operations, and has control over internal reporting and decision structures. Often a government-appointed Board of Directors functions as the overall policy and approving body, sometimes with veto powers held by a Minister. An example of a relatively successful park parastatal is South African National Parks, popularly known as SANParks. Although the agency is still far from perfect, it has a markedly different mind-set and approach from bodies such as Sri Lanka’s Department of Wildlife. SANParks has a clearly defined mission statement and strategic plan, according to which it measures and reports its performance. It publishes annual reports

330 detailing activities of the various divisions involved in delivering its strategy; these reports include audited financial statements. Non-core activities such as the running of lodges and shops associated with the park network have been outsourced to the private sector, enabling the agency to focus on core activities such as biodiversity conservation (Spenceley 2008; p.74). These partnerships generate significant revenue, and in 2006, the agency earned US$ 2.2 million from shops, restaurants and cafeterias alone (Castley et al. 2009; p.402). SANParks has specific divisions to interact with key stakeholders such as private sector operators, local communities and other government agencies (SANParks 2010). The South African government now provides only 20% of SANParks’ funding, a proportion which has been decreasing steadily over the years (Spenceley 2008; p.71; SANParks 2010).

Other park agencies in Southern Africa have created similar semi-independent organisations but these are still less developed than SANParks. For example, the Zambian Wildlife Authority answers to a Board comprised of private-sector representatives. This agency initially faced some difficulties, but “performance is improving and outsourcing slowly accepted; there are some commercial improvements but it is still a long way from viability given decades of deteriorating infrastructure and staff capacity” (Child et al. 2004; p.146). Namibia has also created a tourism parastatal, but Child et al. (2004) point out that this organisation is still largely ineffective.

Some governments in Southern Africa have taken outsourcing one step further, contracting private companies to manage selected national parks. For example, the Zambian Wildlife Authority has outsourced management of Liuwa Plain National Park to a company called the African Parks Network (Fearnhead 2009). Although this venture is still a work in progress, it seems to be on track to deliver expected benefits (Fearnhead 2009). The African Parks Network also manages six parks in several other African countries including Malawi, Rwanda and the Congo (African Parks Network 2011).

These experiences could provide valuable lessons to the Sri Lankan government, showing how to reform conservation-related government departments to help them achieve their objectives. However, as implementing such reforms would take time, it should establish a separate unit—jointly managed by the Forest Department and the Department of Wildlife—to implement MERs. This unit should be allocated a specific

331 budget, with the agreement that these funds would be repaid as soon as MERs become self-financing. The unit should report to a coordinating committee—comprising key stakeholders—that oversees and monitors progress269. After the concept has been successfully established, the government could even consider outsourcing the management of some MERs to private parties in the form of conservation concessions as has been done in areas adjoining South Africa’s Kruger National Park.

A recent World Bank study found that although Sri Lanka has a good environmental framework compared to other developing countries, these laws are not enforced properly:

In comparison to most other developing countries, Sri Lanka has numerous environmental laws, but it lacks effective enforcement. Despite the existence of good EIA guidelines, legislation, and implementation experience over two decades, ensuring environmental sustainability in the development process, as the country moves forward in its quest to attain higher economic status, is a major concern (World Bank 2012; p.9). As I have already addressed reasons for inadequate implementation of environmental legislation270, I now consider how some existing laws should be amended to facilitate MER implementation. This is important, since as the IUCN has pointed out: “legislation needs to establish the legal framework and specific tenurial and other legal instruments necessary to facilitate a range of PA governance and management arrangements, including co-management schemes, private reserves, and community conserved areas” (Barber et al. 2004; p.143). This is especially true of the introduction of novel concepts such as MERs.

The Fauna and Flora Protection Ordinance—the main piece of legislation governing biodiversity conservation in Sri Lanka—is mainly focused on the country’s PA system (Government of Sri Lanka 2005). Although several species, including elephants, are afforded protection throughout the country, provisions to protect their habitat outside

269 Further details about the proposed organsation structure to implement MERs are presented in section 8.3. 270 In the previous section.

332 state reserves are inadequate (Santiapillai and Jayewardene 2004; Government of Sri Lanka 2005). Furthermore, environmental legislation lacks provisions that enable and incentivise conservation outside PAs, especially by the private sector and local communities (Government of Sri Lanka 2005). The government should quickly address these deficiencies, which would hinder successful MER implementation. In drafting revised legislation it could look at laws enacted in countries that have successfully achieved conservation outside state PA networks.

Experiences in Southern Africa—considered a leader in landscape-based conservation beyond state PAs—have shown that a robust legal and institutional framework is a key enabler of conservation-related innovations. For example, devolving management rights to local communities and the private sector was made possible by some key legislation in each of the countries I considered above. Bond et al. (2004; p.40) describe the main characteristics of an enabling framework and the required changes in the mind-set of government officials:

…[W]hat is less recognised is the presence of a regulatory environment that provides for both industry coordination and collective action without imposing burdensome requirements. Also vital is the presence of a large, viable and innovative commercial sector to drive change and product development. The cornerstone of this approach is the transfer of control of wildlife from a few techno-bureaucrats to many independent and experimenting landholders. This implies a major shift in the role of government, which instead of actively managing all wildlife, is responsible for providing a regulatory and institutional framework. This framework should encourage and create incentives for wildlife enterprises but should also control abuse. A common objective must be to keep transaction costs imposed by such a system as low as possible. In Zimbabwe, the Natural Resources Act of 1941 and the Parks and Wildlife Act of 1975 devolved authority of management of natural resources, particularly wildlife, to local communities and the private sector (Bond et al. 2004). The former vested legal rights and responsibilities for controlling the abuse of natural resources to neighbourhood landholders’ associations called ‘Intensive Conservation Areas’, although recourse to the state for appeal or arbitration was permitted (Bond et al. 2004). The Parks and Wildlife Act of 1975 specifically devolved the authority for wildlife occurring on their properties to landholders, who were self-governed through locally constituted landholder committees (Bond et al. 2004; p.40; Child 2009; p.78). In Namibia, the main enabling legislation was the Nature Conservation Ordinance (No. 4

333 of 1975) which granted user-rights to landholders over certain species designated as ‘huntable game’ (Bond et al. 2004; p.39). In South Africa, the Protected Areas Act No. 57 of 2003 made provisions for the state to enter into co-management agreements for PAs with local communities, private individuals or other bodies (Child 2009).

The key lesson for Sri Lankan policy makers from these examples is that legislative reform is usually required to make innovative approaches to conservation—such as MERs—feasible. While drafting the required legislation they could use corresponding laws in countries such as South Africa, Namibia and Zimbabwe as a starting point, and tailor them to the Sri Lankan context.

8.3 Implementing Managed Elephant Ranges

Having outlined the required MERs characteristics and identified potential challenges to successful MER implementation, I now propose a three-step process for implementing MERs in Sri Lanka’s southeast landscape (summarised in Figure 8.6).

STEP 1 STEP 2 STEP 3

Implement plan using an Establish organisation Develop a management adaptive management structure plan approach

• Establish dedicated units • Include essential components • Use an adaptive within lead implementation such as goals; programmes and management approach to agencies, i.e. FD and DWC actions to implement goals; provide managers flexibility rules & regulations; human for dealing with a complex • Establish links with resource and financial plans; socio-ecological systems supporting agencies participatory processes; and such as MERs • Create coordinating public awareness programs • Include three basic elements committee with • Include provisions for of adaptive management representation from key stakeholder input and stakeholder groups − Testing assumptions engagement • Specify roles and − Adaptation • Revise plan through adaptive responsibilities and terms of management process, as per − Learning reference for each group Step 3 2!

Figure 8.6: Proposed process for implementing MERs

334 I now discuss each of these steps in more detail.

Step 1: Establish organisation structure

The first step would be to establish an organisation overseeing MER implementation and monitor progress. This organisation should be structured to achieve the following:

• Establish a lead agency (or agencies) responsible for implementing MERs within the southeast landscape;

• Establish clear linkages within other key government agencies;

• Establish a committee to oversee MER progress, including key stakeholders and comprising people of sufficient stature, authority and knowledge to provide feedback to implementing agencies;

• Establish clear roles and responsibilities for each of the bodies, and the members within the bodies; and,

• Ensure that these bodies have the required resources—human, financial and other—to achieve their goals.

The Forest Department and the Department of Wildlife should be joint lead agencies tasked with implementing MERs—the former because it has jurisdiction over most land within these areas, and the latter because it is responsible for conserving and managing Sri Lanka’s elephants. Given its novelty, the initiative would not easily fit into the organisation structure of either agency, and therefore should be staffed by a joint project team comprising members of both agencies. Ideally this team would sit together within a common office space271.

Other government departments, whose agreement and inputs would be required, should designate a key point person, or people. These include the Department of Land Use Policy Planning and the Survey Department to designate the areas comprising MERs, the Sri Lanka Tourism Development Authority to establish and promote ecotourism, the Climate Change Secretariat to prepare for REDD+ and obtain carbon credits, the

271 This should not be too difficult, as the two departments are located close to each other in the government office complex at Battaramulla, on the outskirts of Colombo.

335 Department of Agriculture to make chena cultivation within MERs more environmentally friendly and productive, and the Central Environmental Authority to obtain required environmental clearances272.

A coordinating committee should oversee the progress of implementing agencies. Ideally, this should be co-chaired by the Directors General of the Forest Department and the Department of Wildlife, given the initiative’s importance to both departments. Committee members should be people with the required skills and stature, drawn from government273, the private sector274, the non-governmental sector, community leaders from areas within and adjoining MERs, local politicians and other stakeholder groups. The committee should choose the schedule and location of its meetings to enable key local stakeholders to participate—ideally, meetings should be held in the southeast landscape.

Roles and responsibilities and terms of reference of each body should be clarified at the outset. The joint project team should facilitate the process by drafting guidelines, but these should be discussed and fine-tuned with input of all the members, at an initial ‘kick-off’ workshop. Subsequently, the leaders of each body should discuss and agree with each team member, his or her specific responsibilities and deliverables; these should be disseminated among the entire group.

The coordinating committee should ensure that the implementation teams are provided with the resources required to achieve the goals of the initiative, ensuring that teams are staffed with people who have the requisite expertise, and have access to adequate financial resources. A process should be worked out to allocate money from existing department budgets; how and when this would be repaid from MER profits should be agreed. The proposed organisation structure is indicated in Figure 8.7.

272 Additional departments could be added to this list, if required. 273 Including the supporting agencies. 274 Especially the tourism industry.

336 Coordinating committee Responsible for overseeing process and Joint-chairs: Forest Department and Department of Wildlife providing feedback to Representatives from key stakeholder groups implementing agencies

Forest Department Department of Wildlife Other agencies

• Dedicated unit to • Dedicated unit to • Sri Lanka Tourism Development implement MERs implement MERs Authority

• Department of Land Use Policy Planning

• Department of Agriculture

• Climate Change Secretariat • Central Environmental Authority

• Other (to be added as appropriate)

Lead agencies Supporting agencies 3!

Figure 8.7: Proposed organisation for implementing MERs

Step 2: Develop a management plan

A clear management plan is essential for any PA. According to the IUCN “the purpose of a management plan is to specify the objectives for which the area is being managed, define legal and operational rules, and lay out programmes and activities that together provide a strategic path for managing the area to achieve stated objectives” (Barber et al. 2004; p.145). IUCN guidelines for creating and using a management plan are summarised in Table 8.5.

337 Table 8.5: ‘Best practices’ associated with the management planning process (adapted from Barber et al. 2004; p.162-165)

Area ‘Best practices’

Essential components • Goal and aims of PA

• Programmes and actions to implement goals

• Description of ecological assets and ecosystem services

• Rules and regulations

• Human resources and other assets (e.g. infrastructure and equipment)

• Financial plan, including available financial assets, projection of revenues and costs

• Provisions for monitoring and amending plan according to the principles of adaptive management

• Participatory processes to obtain stakeholder feedback

• Public awareness and community programmes (e.g. community education)

Creation of the plan • Includes effective mechanisms for stakeholder participation in the creation of the plan

• Approval obtained from relevant parties, especially government authorities

Use of the plan • Management evaluation tied to the delivery of management plan

• Officially published

• Disseminated to all relevant stakeholders

• Constant improvement and amendment according to adaptive management practices

Obtaining stakeholder input is a critical part of the management planning process, as evident from its inclusion in all three areas in Table 8.8. The management plan should identify important stakeholders, and include mechanisms to obtain their feedback. These groups should have opportunities to participate in creating the plan and also provide input at regular intervals while it is being implemented, as part of an adaptive

338 management process 275 . Barber et al. (2004; p.147) describe how inadequate stakeholder involvement could jeopardise the entire management planning process:

…[M]any sophisticated management plans have been developed and never implemented. To avoid this, planners need to ensure an inclusive, participatory process involving all important stakeholders, and, perhaps most importantly, the full range of institutions that hold the real power to either frustrate or facilitate implementation. Often, these institutions will not be those which hold the formal mandate for PA management. A particularly important stakeholder group in the context of Sri Lanka’s southeast landscape are poor farmers. This is because, as pointed out in Chapter 1 (end of section 1.4), the proper implementation of the landscape concept requires that landscapes should be ‘negotiated’ and not ‘planned’ (in a top-down manner), and, in order to achieve this, the needs of marginalised groups (such as poor farmers) should be explicitly taken into account. Barber et al. (2004; p.147) describe how the livelihoods of the poor could be threatened due to inadequate, inequitable management processes:

Asserting that a number of diverse stakeholders have an interest in management of a landscape can easily become an excuse for removing decision making away from a local community, often, in effect, putting resources back in control of interested parties who alienated benefits away from the communities in the first place – stakeholders such as commercial logging companies or forest departments. In this way landscape management and the associated multiple-stakeholder approach can be threats to the livelihoods of the poor. To address this issue, the management plan for each MER should have specific provisions to ensure that the needs of poor farmers are met. These should include:

• Explicit representation of members of local farmers’ associations and NGOs concerned with farmers’ welfare on the coordinating committee who help create the management plan (as recommended in the previous section). • A comprehensive analysis of the needs of farmers conducted across the southeast landscape (i.e. a more in-depth iteration of the stakeholder analysis I described in Chapter 6 that provides farmers more opportunities to participate in the process and provide input regarding issues that would impact their livelihoods). The findings should be consolidated and presented as a clear list of farmers’ needs by area (or even village).

275 Described in more detail in the next section.

339 • Mechanisms to meet the needs identified in the stakeholder analysis (or compensation to the relevant farmers if this is not possible) should be devised through participatory mechanisms such as specific meetings with farmer groups conducted within the southeast landscape. The coordinating committee should approve the final set of mechanisms and ensure that the required funding is set aside within the MER budget.

Step 3: Implement plan using an adaptive management approach

As is evident from discussions in Chapters 5, 6, 7, and 8 about the use and management of natural resources within Sri Lanka’s southeast landscape in general, and the concept of MERs in particular, a number of different factors need to be considered across several dimensions: ecological, economic, social and political. In my stakeholder and economic analyses, there were numerous uncertainties in the data and the methods I used to analyse these data. In short, this is a complex system, of the type described in Chapter 1. This complexity is unsurprising, since, as I have pointed out throughout this thesis, we are considering a broad landscape comprising an entire ecosystem with its various components, and different groups of people who are utilising natural resources within this ecosystem in various ways.

Adaptive management is a useful tool for people responsible for managing such an entity. Plummer and Fennell (2009; p.154) describe it as “an approach to the management of complex systems based on incremental, experiential learning and decision making, buttressed by active monitoring or/and feedback from the effects and outcomes of decisions”. Roux and Foxcroft (2011; p.1) outline the applicability of adaptive management in such complex systems:

Adaptive management is an appealing approach to deal with inherent uncertainty in complex and interactive social-ecological systems. In short, adaptive management is about learning-by-doing in a scientific way, adapting behaviour and overall direction as new information becomes available. It provides a structured way for improving our incomplete understanding through an iterative process of setting objectives, implementing policy decisions and evaluating the implications of their outcomes for future decision making.

340 In essence, adaptive management is the process of treating natural resource management as an experiment such that the practicality of trial and error is added to the rigour and explicitness of the scientific experiment, producing learning that is both relevant and valid. Adaptive management would be a useful tool for MER management, since, while the developing the management plan with stakeholder input would be a good starting point for identifying required actions to manage these areas effectively, all potential issues cannot be anticipated ahead of time. Therefore, an adaptive management approach would provide the implementing team with the flexibility to deal with such unanticipated issues. Barber et al. (2004; p.138-139) describe three basic elements of adaptive management:

• Testing assumptions: systematically trying different interventions to achieve a desired outcome

• Adaptation: systematically using the information obtained through monitoring to take action that improves efficiency and effectiveness of management

• Learning: systematically documenting actions, processes and results so that lessons can be integrated into institution-level decision making and shared with broader practitioners and academic communities

It is evident, therefore, that establishing a robust monitoring and feedback system would be an important prerequisite to implementing an adaptive management approach within MERs—various indicators related to the landscape must be tracked and monitored, and used to alter management strategies and actions (Barber et al. 2004; p.148). A basic set of indicators should be agreed at the planning stage and incorporated into the management plan, while new indicators deemed relevant could be added as required.

341 8.4 Summary of key points in chapter

This chapter brings together the findings of the stakeholder analysis conducted in Chapter 6, and the economic analysis conducted in Chapter 7, to provide insights into how MERs could be implemented within southeast Sri Lanka. First, I outlined the main characteristics of MERs that would sustainably realise their economic potential while addressing needs and concerns of key stakeholder groups. The five characteristics were:

1. Reduced human-elephant conflict within landscape, which would improve poor farmers’ livelihoods, meeting the needs of development-oriented NGOs. This would also help conserve Sri Lanka’s national icon, as per the needs of local and international conservationists and the Sri Lankan public, and help mitigate a major issue for the Department of Wildlife and high-ranking government officials.

2. Improved chena cultivation practices within the landscape, which would help the Forest Department safeguard important natural assets and the ecosystem services they provide, which in turn would benefit many locals living in the landscape. In particular, chena farmers would benefit through increased productivity and more secure land tenure.

3. Opportunities for nature tourism activities, which would provide business opportunities for Sri Lankan (and possibly international) tourism companies, in turn benefitting local businesses. These activities would provide the Forest Department with a source of revenue from land that has previously been ‘unproductive’ from the department’s viewpoint.

4. Credits for carbon sequestration, which would provide an additional source of revenue for the Forest Department.

5. Improved local livelihoods and local support for conservation, which would address equity issues within the landscape and help to obtain vital local support for MERs.

Having described the required characteristics of MERs, I then identified key challenges to implementing these areas, namely: difficulty in convincing government policy- makers of the economic viability of MERs, deficiencies in conservation-related

342 government agencies, and inadequate provisions in Sri Lanka’s legislative framework to encourage conservation outside state PAs. I discussed potential approaches to address these challenges through insights gained from experience in other countries.

Finally, I suggested a three-step process for implementing MERs:

1. Establishing an organisation structure to oversee implementation and monitor progress, comprising of a two lead implementing agencies, supporting agencies and a coordinating committee comprising important stakeholders;

2. Developing a management plan that clearly specifies management objectives, defines legal and operational rules, and lays out required programmes and activities to achieve the stated objectives; and,

3. Implementing the management plan using an adaptive management approach to provide implementing agencies with the required flexibility to manage a complex socio-ecological system such as a MER.

Having described in detail how a landscape-based approach such as MERs could be implemented in southeast Sri Lanka, in my next chapter, I consider possible wider applications of this approach in other parts of Sri Lanka and in other developing nations.

343

This page has intentionally been left blank

344 Chapter 9: Conclusions and implications of research

In this thesis I have explored how a landscape-based approach could be used to mitigate Sri Lanka’s human-elephant conflict, arguably the country’s most important conservation issue. I have made detailed recommendations on how the approach could be implemented within a specific landscape, through insights gained from a series of multi-disciplinary analyses.

In Chapter 1, I provided the rationale for using a landscape-based approach to conservation—given the failures of the ‘traditional’ approach to conservation—in an era of significant worldwide decline of biodiversity. In Chapter 2, I provided background information on Sri Lanka—a developing country whose valuable biodiversity resources are under threat—where I explore how such a landscape-based approach could be used to mitigate human-elephant conflict. In Chapter 3, I introduced the Asian elephant, discussing its ecological, historical, cultural and religious importance across its range in general, and in Sri Lanka in particular. In Chapter 4, I explained why the human- elephant conflict is perhaps the most important conservation-related issue in Sri Lanka, given the conflict’s toll on humans and elephants alike, and showed why current approaches to address it—mostly based on the ‘traditional’ approach to conservation— have largely failed. I then proposed that Managed Elephant Reserves (MERs)—a landscape-based approach—could help mitigate the conflict.

Having introduced the problem and proposed a potential approach to address it, I then explored in detail how this approach could be applied to a specific landscape in southeast Sri Lanka. This type of practical analysis is essential given the unique combination of social, political, economic, and ecological challenges associated with implementing a landscape-based approach in a specific area. Moreover, the multi- faceted issues associated with the complex socio-economic-ecological systems that comprise landscapes require us to adopt a multi-disciplinary outlook. Keeping this in mind, I conducted a series of analyses—using a diverse set of methodologies from several academic disciplines—to develop a comprehensive fact base to draw on when designing a solution for my selected study area. In Chapter 5, I examined this landscape in detail, analysing land-use patterns, local livelihoods, and the area’s human-elephant conflict, which is expected to worsen given the region’s rapid development. In Chapter 6, I conducted a detailed stakeholder analysis to identify needs and concerns of key

345 stakeholders and understand potential roles each group could play in MER implementation. In Chapter 7, I developed an economic model that captures the landscape’s major sources of economic value—especially those associated with its natural assets—and developed alternative scenarios to understand how these sources of value would be impacted by future land-use decisions. Finally, in Chapter 8, I brought together insights from my various analyses to propose how MERs should be established within Sri Lanka’s southeast landscape.

To summarise the implications of my research, I return to my research questions, listed in Chapter 1, namely:

1. What is (and has been) Sri Lanka’s approach to conservation, and what are key issues associated with this approach?

2. What have been the impacts of this approach on elephant conservation and human-elephant conflict?

3. Could a landscape-based approach help conserve elephants while addressing stakeholder needs in southeast Sri Lanka?

- How could resources, especially land, be shared among various stakeholders?

- How could government policymakers (usually the ultimate arbiters regarding land-use) and other relevant stakeholders be convinced to adopt this approach?

- How should the approach be implemented, and what enabling factors are required?

4. What lessons could be drawn from this research in the context of Sri Lanka and other developing countries?

In the following sections, I take in turn these research questions, describing how each has been addressed by my research.

346 9.1 What is (and has been) Sri Lanka’s approach to conservation, and what are key issues associated with this approach?

Sri Lanka has adopted, and largely continues to adopt, a ‘traditional’ approach to conservation. As described in Chapter 2, the country still uses a ‘reductionist’ approach to land-use planning, where numerous government ministries and departments are responsible for managing state-owned land and developing land-use policies for about 80% of Sri Lanka’s countryside. These agencies typically do not consult each other when making land-use decisions, often resulting in destructive land-use practices adjacent to PAs such as the spread of high-intensity agriculture with significant chemical inputs. Consequently, Sri Lanka’s reductionist approach has contributed to losses of habitats of many species, resulting in grave threats to the country’s rich biodiversity. Unsustainable land-use practices near PAs even threaten the biodiversity contained within these parks. Moreover, the lack of an ecosystem-based view in land- use policy has contributed to the loss of valuable ecosystem services that many people—especially the rural poor—depend on for their livelihoods.

Apart from using a reductionist approach, the Sri Lankan government has often perceived local people as apart from nature. For example, the Department of Wildlife and the Forest Department usually consider locals as a threat to the wildlife and other natural resources within state land. This has often resulted in adversarial relationships between local villagers and department officials, as revealed in interviews I conducted in southeast Sri Lanka. The lack of trust between the two groups has often exacerbated human-wildlife conflicts in areas near PAs, as locals believe that animals such as elephants and leopards are the ‘responsibility’ of the Department of Wildlife. This aspect is explored in more detail in the following section.

347 9.2 What have been the impacts of Sri Lanka’s ‘traditional’ approach on elephant conservation and human-elephant conflict?

Sri Lanka’s ‘traditional’ approach to conservation has hampered efforts to address the country’s growing human-elephant conflict. Due to the country’s reductionist approach to land management, the Department of Wildlife, which is responsible for managing the country’s wildlife, only has jurisdiction over a PA network comprising about 14% of Sri Lanka’s land area. This has proved extremely problematic in the case of elephants, which range widely outside these reserves, inevitably coming into conflict with locals. The general lack of co-operation among agencies results in the Department of Wildlife being unable to work effectively with the Forest Department—that manages most elephant habitat outside PAs—to develop an effective approach to mitigating the conflict. The situation has been further exacerbated since the two agencies have different mandates (e.g. the Forest Department has a profit incentive, which the Department of Wildlife lacks), and are now under different ministries (sometimes resulting in bureaucratic infighting).

As described previously, the view that people are apart from nature has led to hostility between government officials and locals. The lack of effective, timely mechanisms to compensate villagers for damages to crops and property, and, on occasion, even lives, has sometimes led to violent confrontations between the two groups, as I witnessed personally during my research trips. Government efforts to separate people and elephants through methods such as elephant drives and the translocation of problem animals have failed because they have not addressed the conflict’s root cause: rapidly dwindling elephant habitat often resulting from poorly planned development efforts and ineffective land management.

Therefore, Sri Lanka’s ‘traditional approach’ to conservation and land management has led to serious human-elephant conflict, claiming the lives of about 75 people and 200 elephants each year, and causing millions of dollars worth of damage. In some areas such as Sri Lanka’s southeast landscape, that have had low conflict levels in the past, the problem is escalating rapidly due to increasing state-led ‘development’ activities.

348 9.3 Could a landscape-based approach help conserve elephants while addressing stakeholder needs in southeast Sri Lanka?

My research indicates that an innovative landscape-based solution developed by conservation biologists could significantly mitigate Sri Lanka’s human-elephant conflict, especially in the country’s southeast.

Research by Fernando et al. (2008)—based on detailed understanding of elephant ranging patterns, resource-use and behaviour—has shown that elephants and farmers practicing traditional chena agriculture could share resources, since the two groups utilise the land at different times; farmers grow crops in the wet season, while elephants feed on regenerating vegetation in abandoned chena plots during the dry season when food is scarce elsewhere. While the authors have presented convincing scientific evidence for their findings, and demonstrated that the approach could work in a few test villages in southeast Sri Lanka, they have gained less traction in establishing MERs. Although this mechanism has been declared as government policy in the National Policy for the Conservation and Management of Wild Elephants in Sri Lanka of 2006, it has not yet been implemented (Government of Sri Lanka 2006; Bhattacharjya 2010).

In my thesis, I have presented a detailed roadmap for implementing MERs. I first identified required MER characteristics through insights gained from a series of analyses to understand the landscape, the main stakeholders involved and economic drivers behind the use of natural resources in the area. MERs should address the root cause of human-elephant conflict (i.e. the lack of adequate food and water resources for elephants). Moreover, the judicious use of properly constructed electric fencing is extremely important, and fences currently on administrative boundaries should be relocated to ecological boundaries (i.e. on the perimeter of state land comprising MERs). Farmers who currently cultivate seasonal chena plots within MERs should be required to use conservation-friendly cultivation methods, and remove fences around their properties after harvesting their crops. This would not only provide elephants access to crucial resources during the dry season, but would also ensure that locals continue to benefit from valuable ecosystem services provided by the landscape’s natural resources. Chena farmers, whose activities are technically illegal, should be provided with a form of legal tenure, such as a permit system. The combination of these steps should reduce human-elephant conflict within the landscape, meeting the needs of several important

349 stakeholders: local rice and chena farmers whose lives, crops and property are currently under threat from marauding elephants, the Department of Wildlife which is under severe pressure to reduce the conflict, conservationists and Sri Lankans living in urban areas concerned with elephant conservation, and politicians and other policy-makers who seek to balance elephant conservation with the needs of locals.

Steps should also be taken to capture major sources of economic value contained within MERs. Low-impact, high-end nature tourism operations should be established, and innovative product offerings that meet customer needs sustainably should be developed. The government should expedite processes that would enable Sri Lanka to participate in emerging global REDD+ schemes. These revenue-generating schemes should be complemented by initiatives to build local support for MERs. For example, a portion of revenues from tourism and carbon credits should be utilised for local development activities, and tourism operators should be encouraged to employ locals, and buy local produce. These economic and equity-related initiatives would meet the needs of several stakeholders: Forest Department officials seeking to generate revenue from land under their purview, tourism authorities promoting Sri Lanka’s natural assets, and tourism companies and local entrepreneurs seeking business opportunities. Efforts to ensure that locals benefit from these activities would help build crucial local support for MERs, and also gain buy-in from local politicians and NGOs seeking to improve local livelihoods.

While implementing MERs, it is important that potential challenges are identified upfront, to the extent possible, and addressed ahead of time. For example, ‘development-focused’ government officials would need to be convinced of the economic viability of these areas. The economic analysis should be shared with them to demonstrate MERs’ significant economic potential, and examples of other countries that have successfully implemented high-end nature tourism and carbon credit programmes should be used to demonstrate the viability of such initiatives. Deficiencies in key implementing agencies such as the Department of Wildlife and the Forest Department should be identified and addressed, but as this process may take time, a separate implementing team with adequate resources should begin establishing MERs. Furthermore, essential legislative and policy changes—such as those incentivising conservation outside PAs—should be set in place.

350 Finally, I proposed a three-step process to implement MERs by establishing an organisation led by the Department of Wildlife and the Forest Department (but with a mandate to collaborate with key stakeholder groups through a coordinating committee), developing a comprehensive management plan with stakeholder input, and putting this plan into practice using an adaptive management approach.

In summary, a landscape-based approach such as MERs would mitigate human-elephant conflict in southeast Sri Lanka if it addresses stakeholder needs and generates economic profits that are shared equitably. In addition, the approach would have tremendous conservation benefits, as MERs would provide habitat for many other species apart from elephants, and would achieve a much better level of biodiversity conservation compared with other less environmentally-friendly land-use types such as irrigated agriculture. An important overall point emerging from my research is that human- wildlife conflicts, which at first seem exclusively a conservation issue that should be dealt with by ‘traditional conservation-oriented’ organisations such as the Department of Wildlife, require a much broader multidisciplinary approach involving multiple actors. Barua (2010; p.56) describe this growing realisation among academics and policy-makers:

The inability of conservation biologists to integrate social scholarship into their own research has become a pressing concern in efforts to conserve biodiversity, and it is increasingly being recognised that mainstreaming conservation into the everyday decisions of the business and public sectors is vital for it to succeed. Conservation requires an interdisciplinary effort, especially when dealing with complex issues such as human-elephant conflict where science, concerns over development, and cultural norms intersect. Conflicts involve multiple actors—international conservation NGOs, national and state governments, agricultural peasants to name a few—each with different ideological standpoints and mediation strategies. In fact, although I began my research with the aim of exploring how a landscape-based approach could be used to mitigate human-elephant conflict, I gradually realised that many of the area’s conservation and social issues stem from fragmented, unsystematic land-use policies of various state agencies; human-elephant conflict is only one of these issues. In the course of my research I discovered several other problems faced by locals due to these policy failures, including the loss of ecosystem services (directly affecting poor locals), inequitable land allocation policies and politically motivated, illegal ‘land grabs’, and exclusion of many locals from the area’s state-sponsored ‘development’

351 activities. Implementing MERs as recommended in my thesis would help address some of these other issues as well.

9.4 What lessons could be drawn from this research in the context of Sri Lanka and other developing countries?

My research methodology and findings have potential applications both in Sri Lanka and beyond. My suggested approach to implementing MERs in southeast Sri Lanka could provide a blueprint for implementing them in other areas with high levels of human-elephant conflict. The overall methodology could have wider applications in other countries with human-wildlife conflict. Finally, my research helps address challenges to applying the concepts of natural capital valuation and the sustainable use of natural resources. I now discuss each of these potential applications.

Establishing MERs in other areas of Sri Lanka

The Sri Lankan government recognises the importance of conserving elephants outside the country’s PA network, addressing this explicitly in its National Policy for the Conservation and Management of Wild Elephants (Government of Sri Lanka 2006; Policy Statement 1):

Conservation of elephants requires large areas, but setting aside sufficient habitat as PAs purely to preserve the current population of elephants which comprises large, highly mobile animals, with home ranges covering hundreds of square kilometres, is almost impossible in Sri Lanka, given its small size and high human population density. This policy also describes the necessity of establishing MERs (Government of Sri Lanka 2006; Policy Statement 1):

Furthermore, given that almost 70% of current elephant range lies outside the PAs, it is imperative that lands other than PAs that could support elephants be integrated into Elephant Conservation Areas (ECAs). ECAs will contain PAs of the Wildlife Department, areas of the Forest Department, and Managed Elephant Ranges (MERs)… MERs are landscapes other than the PAs managed by the Department of Wildlife Conservation (DWC) and forests managed by the Forest

352 Department where elephants will be conserved in-situ. MERs do not indicate any land ownership on the part of the DWC... Although the policy ambitiously envisions that MERs would even encompass private land (apart from areas owned by the Forest Department), the concept still has not been implemented—the Director General of the Department of Wildlife, himself, has admitted that this is the case (Bhattacharjya 2010; p.1):

In 2006, the government had formulated the National Policy for the Conservation and Management of Wild Elephants in Sri Lanka. While the policy was to be revised and reviewed every five years, even at the end of 2010, it is just being taken out of cold storage. “Unfortunately, the policy has not been fully implemented. We have started some of the work and we are committed to its implementation but there are practical problems,” director-general, Department of Wildlife Conservation, Chandrawansa Pathiraja, told the Sunday Times last Friday. My research provides policy makers with a roadmap on how to implement MERs in southeast Sri Lanka. After the concept has been proven to work in this region, it could be extended to develop a set of MERs covering the entire island. Achieving better resource sharing in northeast and northwest Sri Lanka—where human-elephant conflict is more intense than in the southeast—would bring tremendous benefits for farmers, the government, and, of course, elephants. However, land-use patterns in these areas would need to be examined more closely, because large irrigation schemes such as the Mahaweli Development Project have caused significant changes to landscapes in the northeast and northwest. Under this project, which created a series of dams and reservoirs along Sri Lanka’s longest river, large areas of forests were cleared and brought under irrigated agriculture, cultivated by settlers who migrated from other parts of the country (Fernando et al. 2005). In contrast, most people in the southeast landscape have lived and farmed their land for generations, and are generally more used to living with elephants. Therefore, the various context-specific analyses I conducted in the southeast to understand land-use patterns, stakeholder needs and economic sources of value should be performed for these areas so the approach could be adapted to local circumstances.

353 Mitigating human-wildlife conflict elsewhere

As pointed out in Chapter 3, the Asian elephant is threatened throughout its range. Riddle et al. (2010), who conducted a conservation overview of both Asian and African elephants, described habitat loss as a significant threat to Asian elephant conservation:

Habitat loss has been largely responsible for the large-scale decline of the Asian Elephant’s range, and it remains a major threat to elephant conservation in Asia. In the last half-century or so, Asia has seen a rapid growth of its human population and development. This has accelerated habitat loss and most elephant range states have seen declines of 25% to 60% in their forest cover during this period. The problem of habitat loss has been further exacerbated by habitat fragmentation and the absence of integrated and planned development that takes into account conservation needs (Riddle et al. 2010; p.656). Sukumar (2006) describes how human-elephant conflict resulting from habitat loss is posing major issues to policy-makers across the 13 countries that contain Asian elephant populations:

Elephant–human conflict has emerged as one of the biggest conservation issues in Asia and presents an urgent challenge for governments and policy makers. Annually, elephants damage millions of dollars worth of agricultural crops and hundreds of people and elephants are killed as a result of this conflict. The problem of the number of humans killed by elephants is greater in Asia than in Africa. In India alone c. 200 people are killed each year and a further c. 50 people are killed in Sri Lanka. Even in Vietnam and Bangladesh, where there are very few elephants, the conflict can be serious; in Vietnam, for example, small groups of elephants have been in severe conflict with villagers and have been eliminated (Sukumar 2006; p.5). Therefore, it is evident that the challenges associated with maintaining a viable elephant population in Sri Lanka, and the drivers of the country’s human-elephant conflict, are very similar to those in other range states276. Moreover, policy-makers responsible for elephant management are gradually realising that proper land-use planning is an essential step in reducing the conflict. For example, this was one of the main conclusions at a 2006 meeting of Elephant Range States, comprising policy-makers from countries containing Asian elephant populations (Dublin et al. 2006; p.21-22):

276 According to Riddle et al. (2010), however, some Asian elephant populations in Southeast Asian countries face the threat of poaching, a factor that is negligible in Sri Lanka.

354 There is an urgent need for land-use planning that takes into account both elephant conservation and development needs. Where there are justifiable developmental needs, and where habitat loss is inevitable, the conversion of elephant habitat for human use needs to be done in a well-planned manner so as to minimise its adverse impact on elephants and human- elephant conflict resulting from that loss. This can be done using an elephant-specific ‘Environmental Impact Assessment’ that stops poorly planned development and where development is justified, recommends suitable actions to avoid or minimise the adverse impact of the proposed development. Consequently, the landscape-based approach explored in my thesis, and my suggested approach to implementing it, would be of interest to policy-makers throughout elephant range states. While the detailed recommendations would no doubt differ according to each country’s specific context (and specific regions within each country), the multidisciplinary socio-economic analyses I conducted would help them develop customised approaches.

Some of my research findings would also apply to wider human-wildlife conflicts, especially related to species that have large ranges, and therefore cannot be contained within designated PA networks. This group includes large carnivores such as lions, tigers, leopards and wild dogs which have a tendency to kill livestock when their natural prey becomes scarce due to human activities such as habitat reduction and excessive hunting (Woodroffe et al. 2005; Inskip and Zimmermann 2009). Kolowski and Holekamp (2006; p.529 - 530) describe the importance of conservation efforts focused outside PA networks:

Large carnivores, humans and their livestock have coexisted for millennia, but recent decades have seen dramatic increases in the frequency of human–carnivore conflict, resulting mainly from an exponential increase in the human population… Because few of Africa’s existing reserves are large enough to maintain viable populations of wide-ranging predators, conservation of large African carnivores is likely to depend on networks of smaller reserves and private and communal lands, where successful conservation will be closely linked with an ability to resolve human–carnivore conflicts and minimize numbers of carnivores killed by people. To this end, park managers, biologists, and indigenous people must coordinate efforts to understand the circumstances surrounding carnivore-livestock conflict, and combine empirical data with local experience to identify factors that may reduce its frequency.

355 From this description, it appears that different types of human-wildlife conflict have some common causes. The root cause is habitat loss, leading to reduction of food and water resources available for animals—plant species for herbivores, and prey species for carnivores. This in turn, compels animals with large ranges—such as elephants, lions and wolves—to range widely outside PA boundaries in search of food, inevitably coming into contact with rapidly expanding human populations. Countries that have single-mindedly pursued ‘development’ without systematic land-use planning have experienced increasing levels of conflict. Consequently, policy-makers are gradually focusing on improving land-use planning techniques to tackle the root causes of human- wildlife conflict. For example, Lamarque et al. (2009), who conducted a review of the human-wildlife conflict in Africa for the Food and Agriculture Organisation, viewed land-use planning as an essential, long-term approach to mitigate the conflict277:

Land-use planning is a basic human-wildlife conflict management strategy that offers possibly the best chance of overall and long-term success. Unlike strategies of protection and mitigation, it tackles the root of the problem. It is therefore a preventive approach designed to alleviate human- wildlife conflict by creating landscapes in which people and wildlife can co-exist and have as little negative impact on each other as possible. Land-use planning is typically a long-term process that requires government support, legislation and policy changes. It can be extremely expensive to implement, for this reason land-use plans are rarely implemented on a large scale in Africa. On the other hand, land-use plans designed to reduce wildlife losses can be usefully developed and implemented at local level. National land-use planning should be designed through a coordinated approach, involving all government departments, especially those dedicated to wildlife and national parks. Uncoordinated planning could only increase the human-wildlife conflict instead of mitigating it (Lamarque et al. 2009; p.69). The discussion in the above report underscores several points regarding land-use planning. The approach has the best chance of success since it tackles the root of the human-wildlife conflict. It typically requires governments to take a long-term view and often impels legislation and policy changes. National land-use planning requires coordination among the relevant government departments: those involved in environmental conservation, and those involved in development. Although land-use

277 The authors surveyed the various methods currently being used to address the conflict and evaluated the strengths and weaknesses of each method.

356 programmes involve government oversight, they should be implemented locally (and need local support). Finally, they are usually very expensive to implement.

My recommended landscape-based approach demonstrates how many of these suggestions could be implemented. I describe how government officials could be convinced of the long-term economic viability of MERs, and outline required legislative and policy changes. I suggest potential organisation structures and processes to enable relevant government departments to work together to implement MERs. I also demonstrate how requirements of key stakeholders—especially local ones—could be identified and addressed in MER design. While specific solutions would need to be tailored to different areas and different types of human-wildlife conflict, the key steps of my analysis, and the types of methodologies I use, would be helpful to policy-makers.

Helping address challenges to applying the concepts of natural capital valuation and the sustainable use of natural resources

In Chapter 5, I described the need for economic valuation. Turner and Daily (2008; p.26) summarise the basic argument:

In the past, nature conservation and PA policy was justified largely by a combination of separate scientific and ethical “intrinsic value” arguments. But today, with unprecedented and intensifying pressures to deplete natural resources, the traditional arguments in support of ecosystem conservation alone are not sufficient. They do not capture the utter dependence of human well being on natural capital. Despite growing global-level recognition (via the United Nations’ Millennium Development Goals and the Millennium Ecosystem Assessment) that conservation often makes economic sense for society as a whole, decision makers, from individuals to governments, continue to discount inappropriately when choosing between ecosystem conversion or conservation and seem unwilling or unable to provide sufficient finance/investment to match the conservation rhetoric. However, some emerging initiatives to value economic initiatives have been criticised. While academics such as Arne Naess have questioned valuation on moral and ethical grounds (Oksanen 1997; p.544), others have issues with valuation methodologies (Chee 2004; McCauley 2006; Gomez-Baggethun and Ruiz-Perez 2011). A major point of contention among the latter group is that valuation fails to adequately capture the complexities and uncertainties inherent in complicated, interdependent systems such as

357 ecosystems. Chee (2004; p.561) summarises the main arguments related to this view and outlines his proposed alternatives:

The techniques used for valuation suffer from serious limitations and in addition, many ecosystem services are simply not amenable to valuation by the techniques available. Ecosystem management problems are often complex, multi-faceted, socially contentious and fraught with uncertainty. As we have seen, the welfare economic approach to decision-making is too narrowly focused and does not take adequate account of uncertainty and irreversibility. It is therefore imperative to consider more comprehensive approaches for facilitating genuine, substantive stakeholder participation with opportunities for social learning, value formation, problem exploration, risk assessment, analysis of uncertainty, broad-based debate and reconciliation of interests. I believe that this line of argument has merit, especially since, as I have pointed out several times, ecosystems are complex entities, and values of some of their elements could never be fully captured through economic analysis. Consequently, valuation methods would underestimate total ecosystem value in most circumstances278. My point, however, is that since many policymakers, especially those in the developing world, are increasingly making land-use decisions based on economic rationale, and as they routinely ignore ecosystem services when making these decisions, they implicitly set the value of these benefits at zero. Therefore, attaching a value to these services would demonstrate to policymakers that these services have tremendous benefits, especially to people living within the landscape. It is important, however, to make it very clear that these values would almost certainly under-estimate total ecosystem benefits; consequently economic analysis should only be one of several inputs to the decision- making process. For example, when building economic models, non-economic values that are not captured—such as ecosystem limits, equity, and cultural and religious values—should be identified. Moreover a detailed stakeholder analysis should be conducted to understand impacts of land-use decisions on key groups. In fact, as revealed in my stakeholder analysis, people often bring up equity issues as well as moral and ethical reasons to conserve natural resources.

278 In other words, economic valuation would be an underestimate of the total value of ecosystems to humans, for example, since some people derive cultural and religious benefits from nature. ‘Deep ecologists’ like Arne Naess would contend that nature has ‘intrinsic value’ in addition to these anthropocentric benefits.

358 In short, my views are aligned to those of Gomez-Baggethun and Ruiz-Perez (2011; p.624):

Traditional conservation strategies have been shown to be insufficient to reverse biodiversity and habitat loss and we support the environmental movement’s attempt to improve the effectiveness of conservation efforts through novel means. We believe that the idea of ecosystem services is a powerful concept that can advance the ontological position that ecosystems are not only a matter of ethics and aesthetics, but also a basic condition for human life and subsistence. Furthermore, we feel that economic valuation can be a potent information tool when not used as a single decision making criteria (e.g. cost benefit analysis), and if used alongside other valuation methods that capture the non-economic value dimensions of nature. My research has shown how the views of authors such as Gomez-Baggethun and Ruiz- Perez (2011) could be put into practice within specific landscapes.

Although the need to value ecosystem services is increasingly being accepted by policy- makers (World Bank 2012), Turner and Daily (2008) point out that there are constraints to applying the concept, due to what they term “information failures”, “institutional failures” and “market failures”:

Despite growing general awareness of conserved ecosystem benefits, detailed information at scales useful for decision makers on how people benefit from specific services remains deficient. This ‘information failure’ is one reason why conservation investment finances still too low and sometimes ineffective. Another reason is ‘institutional failure’. The beneficiaries of ecosystem service provision are often different and distant from those who gain from ecosystem transformation. Local socio-ecological contexts, including property rights and institutions, are often not given sufficient consideration in conservation programmes, so that legitimacy and equity concerns inhibit uptake. Thirdly, ‘market failure’ occurs because of the public good characteristics of many benefits and their lack of prices. Markets also typically reward short- term values of natural resources (exaggerating the real opportunity costs of conservation) to the detriment of long-term ecological health and human welfare (Turner and Daily 2008; p.26-27). My research helps address some of these constraints. Firstly, my stakeholder and economic analyses provide concrete information on the extent to which people in southeast Sri Lanka benefit from ecosystem services. Since this information is related to a specific landscape that has been receiving significant attention given tensions between rapid development and conservation (as well as growing human-elephant conflict), it

359 provides this information at a scale that is useful for decision-makers. Secondly, my analyses of the landscape’s socio-economic-ecological context, which shed light on the local property rights and institutions related to land-use, addresses the “institutional failure” described by the authors. Finally, the economic model I developed provides a rough estimate of the economic value of the landscape’s natural resources for several land-use scenarios. This, coupled with my analysis of non-economic benefits and costs associated with each scenario, led to my recommendation to establish MERs. If implemented as suggested, these areas would address some “market failures” within the landscape.

The establishment of MERs—which combine elephant conservation with traditional resource-use practices and revenue generation from nature tourism and carbon credits— would promote sustainable resource use within the landscape. Consequently, they address a gap in the literature on the concept of sustainable use, identified by Cooney (2007; p.58): “Despite decades of experience, strong professional networks, and some detailed work on specific issues, there is a glaring lack of accessible, rigorous published case examples establishing a convincing case that sustainable use can be a more effective approach to conservation than strict protection”. The findings of my research in southeast Sri Lanka have potential applications in other tropical biodiversity hotspots such as India, Indonesia, and the Philippines, where there is severe competition for land among various groups such as farmers, conservationists, entrepreneurs and government officials. The principles of sustainable use, combined with a landscape-based approach, could help policy-makers develop and implement practical methods to mitigate catastrophic loss of valuable biodiversity in these areas.

360 9.5 Final thoughts

As pointed out in Chapter 1, the world’s biodiversity is facing a crisis of historic proportions. The problem is especially acute in developing countries like Sri Lanka, which have been designated as biodiversity hotspots, not only for their incredible variety of plant and animal species, but because these species are under severe threat279. These countries are aware of the need to conserve their biological treasures—for example, the Convention on Biological Diversity has been signed by 193 countries280 (Secretariat of the Convention on Biological Diversity 2012). However, they struggle to dedicate additional land as conventional PAs, since there are many competing demands on land, arising from the escalating needs of their growing populations.

Consequently, traditional, PA-focused, conservation approaches—those that ignore ecosystem complexity, and view people as ‘enemies’ of nature—would not work. In fact, as illustrated throughout my thesis, such approaches often put at risk the very resources they are trying to protect. Consequently, there is a need to quickly implement land-use systems that balance conservation needs with those of locals living within landscapes. In the words of the philosopher Jack Turner, “What counts as wilderness is not determined by the absence of people, but by the relationship between people and place” (Dowie 2009; p.19).

Implementing landscape-based approaches that address the deficiencies inherent in traditional approaches are proving complicated, however, given the numerous social, economic and cultural factors that must be considered alongside scientific ones. Their feasibility is often determined by economic and non-economic impacts on various stakeholder groups, and how they react to these impacts. This, in essence, is the overall finding of my research: a complex, multi-faceted issue such as determining land-use, requires a comprehensive, multi-disciplinary approach yielding practical solutions that policy-makers could implement within particular socio-economic contexts. This type of approach must be widely used in order to mitigate catastrophic loss of valuable biological resources in developing countries such as Sri Lanka.

279 To be designated a ‘biodiversity hotspot’ an area must contain at least 0.5% or 1,500 of the world's 300,000 plant species as endemics, and should have lost 70% or more of its primary vegetation, since this being the form of habitat that usually contains the most species, especially endemics (Myers et al. 2000; p.854 - 855). 280 Sri Lanka has been a party to the convention since 1994 (Secretariat of the Convention on Biological Diversity 2012).

361

This page has intentionally been left blank

362 Appendix 1: Problem definition and research overview

A1.1 Global biodiversity loss

According to the Millennium Ecosystem Assessment, species extinctions in the recent past—i.e. due to human influence—are as much as 1,000 times background rates that were typical over Earth’s history (Millennium Ecosystem Assessment 2005; p.3). Moreover, projected future extinction rates are projected to be more than ten times the current rate (Millennium Ecosystem Assessment 2005; p.4), as evident from Figure A1.

Figure A1: Estimated extinction rates—distant past, recent past and future (from Millennium Ecosystem Assessment 2005; p.4)

363 A1.2 Issues with the global protected area network

As pointed out in Chapter 1, the extent to which PAs fulfil their role of protecting biodiversity, depends on how well they meet two objectives (1) ‘representativeness’, or “the need for reserves to represent, or sample, the full variety of biodiversity, ideally at all levels of organisation”, and (2) ‘persistence’, the ability of PAs, once established, “to promote the long term survival of the species and other elements of biodiversity they contain by maintaining natural processes and viable populations by excluding threats” (Margules and Pressey 2000; p.243).

I now evaluate the ability of the global PA network to meet each of these objectives, and consider how climate change could impact both the ‘representativeness’ and ‘persistence’ of PAs.

Evaluating the ‘representativeness’ of protected areas

364 Table A1: Numbers of gap species in the current PA network (from Rodrigues et al. 2004; p.641)

Taxon Median range Current network Current network size (km2) (all PAs) (PAs > 1,000ha and IUCN I – IV) All species Mammals (n=4,835) 247,341 258 (5.5%) 644 (13.5%) Turtles (n=273) 309,172 21 (7.7%) 48 (17.6%) Amphibians (n=5,474) 7,944 913 (16.7%) 1,718 (31.5%) Threatened species Mammals (n=1,063) 22,902 143 (14.0%) 314 (29.6%) Birds (n=1,171) 4,015 232 (19.8%) 437 (37.3%) Turtles (n=119) 167,611 12 (10.1%) 32 (26.9%) Amphibians (n=1,543) 896 411 (26.6%) 767 (49.7%) All species analysed 38,229 1,424 (12.2%) 2,847 (24.5%)

The authors note that the ‘gaps’ identified are likely to be much worse than their baseline finding of 12% because of several reasons: (i) when stricter criteria (column 4) are applied the number of gap species doubles; (ii) the number of covered species is an overestimate because of the unrealistic assumption that species can be protected equally effectively in any part of their range; and (iii) it is likely that other taxa with high levels of endemism, such as plants and insects, are even less well represented, given the tendency for sets of species with smaller range sizes to have higher proportions of gap species (Rodrigues et al. 2004).

Other research has shown that there are significant gaps in coverage of biodiversity-rich habitats. The WWF has divided the terrestrial world into biogeographic realms which are “geographical regions where distinctive assemblages of plants and animals occur” and 14 biomes within these realms according to major plant communities found in them (UNEP-WCMC 2008; p.20). According to a study by the United Nations Environmental Programme, a significant portion of these biomes, especially those in tropical realms, have levels of protection below 10 per cent, the minimum target

365 recommended by the Convention on Biological Diversity281 (UNEP-WCMC 2008; p.21).

The main reason for these gaps is that conservation planning has generally not been systematic and new reserves have often been located in places that do not contribute to the representation of biodiversity (Margules and Pressey 2000). Formal protection is often biased toward areas with least potential for commercial or subsistence use and least in need of conservation action (Pressey et al. 2002). Although this practice, in part, reflects the political and economic implications of setting aside land which may have other productive uses, environmental policy-makers are often not clear about their objectives—they sometimes focus on protecting ‘scenic’ and ‘wilderness’ areas which are remote and rugged, without considering their significance from the point of view of biodiversity (Bengtsson et al. 2003; p.389).

Schmitt et al. (2009) pointed out that the coverage of tropical forests—one of the most important habitat types—is especially poor. According to them, 13.5% of forest cover is protected globally under all IUCN management categories, while median protection of forest cover within eco-regions is 11.2% (Schmitt et al. 2009; p.2124). However, those figures decline to 7.7% for global forest cover and 5.9% for median forest protection within eco-regions if PAs with IUCN management categories I – IV are analysed separately. The authors’ overall conclusion is that “there are still large numbers of forest areas with very low levels of protection in the tropics as well as in the temperate and boreal zones” (Schmitt et al. 2009; p.2129).

281 The CBD target is that at least 10 per cent of each of the world’s ecological regions [should be] effectively conserved’ and ‘areas of particular importance to biodiversity [should be] protected’ (UNEP- WCMC 2008).

366 Evaluating the ‘persistence’ of protected areas

According to Margules and Pressey (2000) the areas surrounding PAs are key to their ‘persistence’ value. Their view is that “if reserves become remnants of natural habitat surrounded by alien habitat such as cropland or pasture, changes brought about by isolation and exposure have implications for the persistence of species within them” (Margules and Pressey 2000; p.248). DeFries et al. (2005) examined the rates of loss of forest cover in the world’s moist and dry tropical forests both within PAs and in areas within a 50 km periphery of their administrative boundaries. They found that loss of forest cover occurred within 25% of PAs located in moist forest and 23% of those in dry forests. Of these, 9% declined by more than 5% forest area. The corresponding figures for the areas surrounding moist and dry forests were 69% and 54% respectively. Of these, 29% declined by more than 5% forest area. These findings reveal that rapid deforestation is occurring in areas surrounding PAs, leading to deterioration in the ‘persistence’ value of these PAs.

The authors believe that their analysis is likely to underestimate the true loss of conservation capacity because they do not account for human activities that do not result in habitat loss (e.g. hunting) and because the satellite analysis cannot distinguish between native forests and forest plantations that are unsuitable for many species. Their overall conclusion is that “despite little or no loss of forest habitat within their administrative boundaries over the past 20 years, reduction in forest habitat surrounding PAs has reduced their capacity to conserve species richness” (DeFries et al. 2005; p.24). In other words the ability of PAs to safeguard the ‘persistence’ of the species within them has been severely eroded.

Another study by Laurance et al. (2012) re-enforced the importance of areas surrounding PAs in preserving the biodiversity contained within them. The authors studied the biodiversity within “a comprehensive data set on changes over the past 20 to 30 years in 31 functional groups of species and 21 potential drivers of environmental change, for 60 PAs stratified across the world’s major tropical regions” (Laurance et al. 2012; p.290). Their findings reveal not only the severe threats to the biodiversity within these reserves but also that many threats to PA persistence originated from areas immediately outside reserves:

367 Our analysis reveals great variation in reserve ‘health’: about half of all reserves have been effective or performed passably, but the rest are experiencing an erosion of biodiversity that is often alarmingly widespread taxonomically and functionally. Habitat disruption, hunting and forest- product exploitation were the strongest predictors of declining reserve health. Crucially, environmental changes immediately outside reserves seemed nearly as important as those inside in determining their ecological fate, with changes inside reserves strongly mirroring those occurring around them. These findings suggest that tropical PAs are often intimately linked ecologically to their surrounding habitats, and that a failure to stem broad-scale loss and degradation of such habitats could sharply increase the likelihood of serious biodiversity declines (Laurance et al. 2012; p.290).

The impact of climate change on both ‘representativeness’ and ‘persistence’ of protected areas

The global climate is changing rapidly as emission rates of anthropogenic greenhouse gases accelerate (IPCC 2007). In the tropics, temperature increases have averaged 0.26ºC per decade since the 1970s (Malhi and Wright 2004). These changes have far reaching implications for the effectiveness of PAs both in terms of representativeness and persistence.

Studies by Hannah et al. (2007) have predicted that climate change will drive species from reserves, thus reducing the ‘representativeness’ of these PAs. Predicting these effects is complicated, because each range shift is dependent on the dispersal ability of that particular species; hence species move independently of one another (Hannah et al. 2007).

Mansourian et al. (2009) describe how climate change will reduce PA ‘persistence’ by causing a range of stresses on their biodiversity. For example, there is a high probability that competing—sometimes invasive—species, more adapted to a new climate, will move in. Climate change is also expected to lead to disease outbreaks, as pest species, better adapted to new conditions, invade PAs. Finally, climate change is also likely to lead to higher incidences of fires in some situations and floods in others (IPCC 2007). Mansourian et al. (2009) believe that these impacts on PAs would be compounded by other stresses caused by humans such as overconsumption, pollution or encroaching urbanization. Therefore, their view is that “biodiversity in PAs that may already be

368 vulnerable because of these human threats may be more quickly or more severely affected by climate change” (Mansourian et al. 2009; p.64).

A1.3 Revised IUCN categories acknowledging human interaction with nature

Given the numerous criticisms of the traditional approach to conservation, in 1992, the IUCN introduced two new PA categories—V and VI—to capture the interaction between humans and nature. This is evident from their definitions, which are provided in Table A2 below.

Table A2: IUCN management categories with detailed descriptions (from Dudley 2008)

Category Title Description

V Protected A PA where the interaction of people and nature over time Landscape/ has produced an area of distinct character with significant Seascape ecological, biological, cultural and scenic value, and where safeguarding the integrity of this interaction is vital to protecting and sustaining the area and its associated nature conservation and other values.

VI Managed Resource Category VI PAs conserve ecosystems and habitats, together Protected Area with associated cultural values and traditional natural resource management systems. They are generally large, with most of the area in a natural condition, where a proportion is under sustainable natural resource management and where low-level non-industrial use of natural resources compatible with nature conservation is seen as one of the main aims of the area.

Phillips (2002; p.10) points out that the interaction between humans and nature is key to Category V areas:

Category V is unique among the six categories by making the core idea the maintenance of environmental and cultural values where there is a direct interaction between people and nature. The focus of management of Category V areas is not on nature conservation per se, but about guiding human processes so that the area and its resources are protected, managed and capable of evolving in a sustainable way – and natural and cultural values are thereby maintained and enhanced.

369 This notion has been adopted by the Conference of Parties to the CBD, which in 2004, adopted a Programme on Protected Areas, including a distinct element on ‘Governance, Participation, Equity and Benefit-Sharing’ (Barber et al. 2004). This included specific goals ‘to promote equity and benefit-sharing’ and ‘to enhance and secure involvement of indigenous and local communities and relevant stakeholders’ (Barber et al. 2004; p.100).

A1.4 Yellowstone National Park as an illustration of the deficiencies with the traditional conservation approach

Yellowstone, the first national park in the USA, plays an important role in the origin myth of PAs. Traditionally, (mostly Western) histories of PAs begin with the establishment of this park in 1872—in fact the World Database of Protected Areas takes Yellowstone as its starting point (Brockington et al. 2008). Yellowstone was the model for many subsequent parks in the late 19th and early 20th centuries in places like Australia, New Zealand and sub-Saharan Africa, and the concepts embodied in this model influence the establishment of PAs even today (Greiber et al. 2009; p.89)

The ‘Wilderness Movement’ of the USA, led by individuals like John Muir and George Marsh, was the driving force in establishing Yellowstone (Kalamandeen and Gillson 2007; p.167). Their emphasis was on ‘pristine’ nature, and humans were assumed to have a negative influence on the landscape. This ideal was instrumental in obtaining Congressional support for the Park’s establishment: “Wilderness, spectacular scenery and geological features provided the primary inspiration for this support, and the vision and management of Yellowstone and its tourists reflected the desire of the American public to enjoy a ‘natural’ or ‘wilderness’ area, unaffected by humans” (Kalamandeen and Gillson 2007; p.167). When the park was first established, it was a square block of 810,000 ha, an area deemed sufficient to preserve the ‘wilderness’ it contained ‘forever’ (Brockington et al. 2008).

This origin myth of Yellowstone and its success in preserving an ‘unspoilt wilderness’ has proved to be problematic for several reasons. Firstly, the history of PAs did not begin with Yellowstone. Mongolia established the Bogd Khan Mountain National Park in 1778, and King Devanampiyatissa of Sri Lanka, established a wildlife sanctuary as

370 far back as the 3rd century B.C. (Milner-Gulland 2004; Jayasooriya et al. 2006). In fact, many scholars contend that setting aside PAs to conserve nature has been done for centuries: “It is an ancient and widespread human practice to set aside areas for the preservation of natural values. The sacred groves of Asia and Africa and royal hunting forests are historical examples” (Margules and Pressey 2000; p.243).

Secondly, despite the park’s relatively large size, it quickly became apparent that it did not cover ranges of large predators such as the grizzly bear. Consequently, the Greater Yellowstone Ecosystem—which was subsequently defined by the range of the grizzly bear—now extends to up to 10.5 million ha (Maginnis et al. 2004). However, such extensions have often been controversial because the areas surrounding the park—most of which are privately-owned—are experiencing some of the fastest population growth in the US (Maginnis et al. 2004).

Finally, Native Americans had inhabited Yellowstone well before white settlers arrived. Even at the time of its designation as a National Park, it bore clear traces of various Native American tribes; indeed it is believed that humans have manipulated the landscape using fire for the past 6,000 years (Kalamandeen and Gillson 2007; p.167). The founders of the park encountered various abandoned Shoshone camps, but because of their commitment to promoting their ideal, they blatantly ignored what they saw, reporting to Congress that the “Yellowstone country was a wilderness that was never trodden by human footsteps” (Kalamandeen and Gillson 2007; p.167). In fact, for a few years after its creation, seven tribes lived, hunted and fished there, but in 1877, due to a policy of ‘strict natural protection’ all ‘Indians’ were relocated to a nearby reservation (Dowie 2009). Park officials believed that the presence of Native Americans, who hunted ‘illegally’ in the park, scared away the tourists. Despite significant evidence to the contrary, the Yellowstone Superintendent made the outrageous claim that “Yellowstone is not Indian country, and no natives lived in the park” (Kalamandeen and Gillson 2007; p.167). As West and Brockington (2006; p.613) have pointed out, “one of the central features of the Yellowstone model was erasure of the social history of land- use on the part of the Native Americans and even the Native Americans themselves”.

Therefore, the entity claiming to be the ‘world’s first national park’ is a stark reminder of the problematic assumptions behind the ‘traditional’ conservation approach that inspired its founding. The problematic isolation approach used to separate ‘pristine

371 nature’ from human influence failed to take into account the fact that the park’s ecosystems are not self-contained, static entities but dynamically interact with the areas surrounding it. The park founders’ desire to maintain the fiction of ‘unspoilt nature’ compelled them to ignore that people had been an integral part of this purported ‘wilderness’ for thousands of years.

A1.3 Acceptability of land-use trade-offs at different spatial scales

As pointed out in section 1.4, trade-offs are only acceptable if balanced at the proper scale, since if attempted at too small a scale, as is often done, both conflict and environmental degradation is often the result. Table A3 illustrates the extent to which trade-offs can be made at different scales without undermining environmental functions.

Table A3: Acceptability of trade-offs at different spatial scales (from Maginnis et al. 2004; p.334)

Unit of scale Acceptability of net losses of environmental functions (goods, services, future options, ecological processes) at a particular unit of scale

Regional/ national (105 Ethically unacceptable, even if possible and economically rational. An km2) example is Lawrence Summers’ option of shipping industrial countries’ toxic wastes to developing countries. Will inevitably lead to greater social inequality and species extinction.

Provincial/ eco- Strongly discouraged in virtually all circumstances. An example is the regional (104 km2) conversion of most peninsular Malaysian lowland forest to estate crops. Directing one provincial administration to pursue industrial development and the other to pursue conservation will raise practical governance issues that will be virtually impossible to resolve democratically. Strong likelihood of social inequity and species loss.

District/ landscape (103 Discouraged in the majority of circumstances. If trade-offs are to be km2) made at this scale, strong justification is needed along with robust mechanisms to balance loss of functions elsewhere. Governance issues are likely to be a problem, and ensuring that social equity is

372 Unit of scale Acceptability of net losses of environmental functions (goods, services, future options, ecological processes) at a particular unit of scale

properly addressed will be a major challenge.

Stand/ site (101 km2) Acceptable in the majority of cases as long as good management practices are pursued and there is a reasonable likelihood of social equity issues being adequately addressed.

A1.4 Examples of applications of landscape approaches by international conservation organisations

Having recognised the pitfalls inherent in the traditional approach to conservation, leading international conservation organisations are increasingly adopting landscape- based approaches. For example the World Wildlife Fund (WWF), the Wildlife Conservation Society (WCS), and the African Wildlife Foundation (AWF) have incorporated the approach into their conservation frameworks. I now consider each of these in more detail.

The WWF uses the concept of ‘eco-regions’ which are defined as “relatively large units of land containing a distinct assemblage of natural communities and species, with boundaries that approximate the original extent of natural communities prior to major land-use change” (Morrison et al. 2009; p.499). The organisation decided to use this approach because it realised that “only at relatively large scales (generally greater than single PAs) could conservation planning and implementation adequately understand and preserve habitats and ecological processes” (Morrison et al. 2009; p.499).

The WCS has developed a framework called the ‘Landscape Species Approach’ to help conservation practitioners plan landscape-scale actions (Didier et al. 2009). It centres on the “careful selection of a suite of focal species, called Landscape Species, defined by their ability to represent all major habitats, management zones and threats at a site, their

373 use of large, heterogeneous areas, and their structural and functional impacts on natural ecosystems” (Didier et al. 2009; p.476).

The AWF uses a landscape-scale conservation planning methodology in eight priority conservation landscapes in Africa, which they call African Heartlands (Henson et al. 2009). This approach incorporates key features of the landscape approach such as focusing on conservation outside PAs (as well as within them) and involving local stakeholders. For example: “The Heartland Program augments PAs and helps to manage the surrounding areas, considering the needs of native species, ecosystem processes and local stakeholders. Such landscapes have the potential to provide economic benefits and ecosystem services that strengthen livelihoods of local people” (Henson et al. 2009; p.508). An interesting application of the AWF’s approach relates to its work in and around Amboseli National Park in Kenya, where the landscape was defined based on a wide-ranging species—the elephant:

The primary determinants of the spatial extent of the landscape are the ranging patterns, key habitat areas, and movement corridors of elephants. By using elephants as the proxy indicator of the conservation landscape extent we were able to plan strategies with stakeholders that would address threats to elephants and other conservation targets (Henson et al. 2009; p.510). As was evident in Chapter 1, this type of approach has particular relevance in my thesis.

374 Appendix 2: Overview of Sri Lanka’s biodiversity resources, land-use and conservation landscape

A2.1 Major vegetation types and forest cover in Sri Lanka

Jayasooriya et al. (2006) propose a classification of fourteen main terrestrial and semi- terrestrial vegetation types constituting nine forest types, four grassland types and the sand dunes. Figure A2 illustrates the distribution of Sri Lanka’s vegetation types.

Figure A2: Sri Lanka’s major vegetation types (from Jayasooriya et al. 2006; p.80)

375 Table A4 shows how forest cover in Sri Lanka has reduced over past decades, at the expense of other types of land uses such as agriculture and human settlements.

Table A4: Forest cover in Sri Lanka over time (% of total land area)

Land use category 1990 2000 2005 2010 Forest excluding rubber 33.0% 29.3% 27.5% 26.6% Other land 65.5% 69.2% 71.0% 71.9% Inland water bodies 1.5% 1.5% 1.5% 1.5% Total 100% 100% 100% 100%

A2.2 Sri Lanka’s biodiversity

As pointed out in Chapter 2, Sri Lanka has exceptional biodiversity resources, and has been included—along with the Western Ghats of India—among 34 global biodiversity hotspots (Mittermeier et al. 2004). Table A5 illustrates the species richness among Sri Lankan fauna.

Table A5: Species richness of fauna in Sri Lanka (from IUCN Sri Lanka 2007)

Taxonomic group Total number of Number of endemics % Endemic species Vertebrate fauna Freshwater fish 82 44 54% Amphibians 106+ 90+ 85% Reptiles 171 101 59% Birds 482 33 7% Mammals 91 16 18%

Invertebrate fauna Bees 148 21 14% Dragonflies 120 57 85% Aphids 84 2 2% Ants 181 unknown Butterflies 243 20 8% Ticks 27 unknown

376 Taxonomic group Total number of Number of endemics % Endemic species Spiders 501 unknown Freshwater Crabs 51 51 100% Land Snails 246 204 85%

A2.3 Socioeconomic Statistics

As pointed out in Chapter 2, Sri Lanka is located in the extremely densely populated region of South Asia, where nearly 24% of the world’s total population lives on less than 4% of the Earth’s land area (Veron 2008; p.9). Table A6 provides data on populations and population densities in South Asian countries, while Table A7 shows that Sri Lanka is the least urbanised country in the region.

Table A6: Population, surface area and population density in South Asian countries in 2007 (from Veron 2008; p.15)

Country Number of inhabitants Land area Population density (thousands) (thousand km2) (inhabitants per km2) Afghanistan 27,145 652.0 42 Bangladesh 158,665 144.0 1,102 Bhutan 658 47.0 14 India 1,169,016 3,287.0 357 Maldives 306 0.3 1,020 Nepal 28,196 147.0 192 Pakistan 163,902 796.0 206 Sri Lanka 19,299 66.0 292

377 Table A7: Percentage of urban population by country, 1950 – 2005, and projections to 2030 (from Veron 2008; p.59)

Country 1950 1975 2005 2030 Afghanistan 5.8 13.3 22.9 36.2 Bangladesh 4.3 9.8 25.7 41.0 Bhutan 2.1 7.9 31.0 56.2 India 17.0 21.3 28.7 40.6 Maldives 10.6 17.3 33.9 60.7 Nepal 2.7 4.8 15.8 30.6 Pakistan 17.5 26.3 34.9 49.8 Sri Lanka 15.3 15.5 15.1 21.4

As discussed in Chapter 2, Sri Lanka has impressive social indicators for a country of its size and income level, and compares favourably with its South Asian neighbours. This is illustrated in Tables A8 to A10.

Table A8: Life expectancy in South Asian countries, 1950 – 2000 (years) (from Veron 2008; p.77)

1950- 1960- 1970- 1980- 1990- 2000- Country 1955 1965 1975 1985 1995 2005 Afghanistan 28.8 32.0 36.1 39.9 41.7 42.1 Bangladesh 37.5 41.2 45.3 50.0 56.0 62.0 Bhutan 36.1 37.7 41.8 47.9 54.5 63.5 India 36.4 43.6 50.7 56.6 60.2 62.9 Maldives 38.9 45.2 51.4 57.1 61.0 65.6 Nepal 36.2 39.4 44.0 49.6 55.7 61.3 Pakistan 43.4 47.7 51.9 56.2 60.8 63.6 Sri Lanka 57.6 62.2 65.0 68.8 70.4 70.8

378 Table A9: Adult literacy rate by sex at age 15 and over in the countries of South Asia (from Veron 2008; p.83)

Country Men Women Overall Female/Male ratio Afghanistan 43 13 28 0.30 Bangladesh 52 33 43 0.63 Bhutan - - - - India 73 48 61 0.66 Maldives 96 96 96 1.00 Nepal 63 35 49 0.56 Pakistan 63 36 50 0.57 Sri Lanka 92 89 91 0.97

Table A10: Development indicators in the countries of South Asia (from Veron 2008; p.82)

283 GNI per capita Human development index 282 Country (US$ PPP ) in World ranking 1980 2004 2004 2004 Afghanistan - - - - Bangladesh 1,870 0.366 0.530 137 Bhutan 1,969 - 0.538 135 India 3,139 0.439 0.611 126 Maldives - - 0.739 98 Nepal 1,490 0.336 0.527 138 Pakistan 2,225 0.388 0.539 134 Sri Lanka 4,390 0.653 0.755 93

282 Purchasing Power Parity 283 Composite index measuring human development, including life expectancy, adult literacy rates, school enrollment ratios and per capita GDP. The closer it is to 1, the better the situation.

379 A2.4 Land-use patterns in Sri Lanka

As discussed in Chapter 2, Sri Lanka is largely an agricultural country, where over one third of the island’s total land area is devoted to agriculture, while another one third has been set aside for forest and wildlife. Table A11 shows Sri Lanka’s land-use patterns.

Table A11: Land-use in Sri Lanka, 2008 (complied from information from Department of Census and Statistics 2011)

Type of land use % of total land in Sri Lanka

Agriculture (main crops) 26 Wildlife reserves 15 Natural forests 23 Other types of land use 36 Total 100

A2.5 Conservation activities in ancient Sri Lanka

As discussed in Chapter 2, Sinhalese Kings—who embraced Buddhist concepts regarding nature and the sanctity of life—engaged in a range of ‘conservation activities’. Table A12 presents some examples.

Table A12: Recorded conservation activities by Sri Lankan Kings (adapated from Nanayakkara 1987; Jayasooriya et al. 2006)

King (period) Recorded conservation activities

Vijaya (543 B.C.) There are references to social tree planting even during the time of Sri Lanka’s first King. They mention that agro-forestry was practiced in homesteads and people depended on forests for a variety of products, including medicine (and therefore avoided destroying them).

Tissa (247 – 207 B.C.) A rock inscription dated 245 B.C. indicates that Buddhist monks used forests extensively for meditation (many forests were situated close to cities and villages). Wildlife was protected.

Dutugemunu (161 – Rules for the protection of forest vegetation and for the use of forest

380 King (period) Recorded conservation activities

137 B.C.) products are referred to in the Rajaratnacari. The Mahawamsa describes the existence of plantation forests.

Kirthi Nissanga (1187 The King proclaimed that no animal should be killed within a radius – 1196 A.D.) of seven gau (35.7 km) of the city of Anuradhapura (Jayasooriya et al. 2006).

During this period forests were valued as part of the national defence system, according to 12th century inscriptions engraved in the rock at Dambulla, which “in describing the strength of the national defence under the King Kirti Nissanga enumerates strongholds in the midst of forests, those of steep hills, and the fortresses surrounded by water” (Nanayakkara 1987; p.16).

Parakrama Bahu III The Rajavaliya described how King Parakrama Bahu, who defeated (1284 – 1310 A.D.) the South Indian invaders and united the country, used the forest to his advantage. “The King and his soldiers knew their forests and terrain very well; they sojourned inside the forest, and thereby were able to defeat the enemy” (Nanayakkara 1987; p.16).

Buwanekabahu V Descriptions of forestry activities during this period are found in the (1373 – 1406 A.D.) Ganegoda Sannasa (Deed), and the Oruwela Sannasa. During this period the King gave away forests to people for Rajakariya, i.e., service performed for the King or State.

A2.5 Land management systems in ancient Sri Lanka

As described in Chapter 2, systems evolved to deal with land fragmentation and to foster more equitable access to resources within villages; some of these practices survive to this day. The land held under the Praveni ownership passes down from the parents to children; each person inherits a Pangu or share of the land. There is no physical partitioning of the land and demarcation of boundaries is by mutual agreement. After several generations, the share becomes a small fraction of the original, and a person may own several small parcels of land scattered in different parts of the village. Since the cultivation of such small parcels of land is often not economically profitable

381 or practicable, their owners often trade the right to cultivate the land to a large landowner in exchange for a part of the harvest. These arrangements are only seasonal and do not involve a change of ownership.

Another system of land tenure is Thattumaru under which each shareholder (Pangukaraya) takes turns in cultivating a field (Rathnayake 2007; p.54-62). If, for some reason, a shareholder does not cultivate the land when it is his or her turn, he or she has to wait for several years for the next opportunity to do so. The system of Thattumaru enables a shareholder to cultivate a reasonably large field at least once in a number of seasons rather than to cultivate a very small extent of land every season.

The Kattimaru system—another system of land tenure—has often been confused with the Thattumaru system. Unlike in the latter system, however, parcels of land are clearly demarcated in the Kattimaru system. Since all the parcels are not equal in fertility, soil quality, farmers shift from one to another on a rotational basis so that all shareholders would have a turn on all parcels (Rathnayake 2007; p.54-62).

However, there have been some issues with these indigenous land management systems. For example, the Thattumaru and Kattimaru systems have often forced small farmers to work as sharecroppers when another co-shareholder was cultivating the land. The insecurity of the tenure and the phenomenon of absentee landlordism can also lead to a decline in soil fertility, as farmers who are cultivating a particular plot during a season attempt to maximize their yields.

A2.5 Sri Lanka’s post-independence land policy

The key land-related policy initiative undertaken by successive governments since independence has been the opening up of land for settlement, especially in the dry zone. These initiatives were essentially a continuation of policies instituted by the colonial government in the decades prior to independence, which culminated in the Land Development Ordinance of 1935. D. S. Senanayake, independent Ceylon’s first Prime Minister, took a keen interest in the developing of peasant agriculture, and under his leadership the government undertook land development and irrigation schemes, especially in the dry zone (Peebles 2006). An example is the Gal Oya scheme, the first

382 major project since the days of the Polonnaruwa Kings (de Silva 1981; p.505). Later governments continued these efforts, which culminated in the Accelerated Mahaweli Development Programme of the 1980s. Most of these schemes involved the settlement of farmers from other parts of the country in newly irrigated lands.

Achieving self-sufficiency in rice and relieving population pressure in the wet zones of the island were among the main objectives of land settlement (Survey Department of Sri Lanka 2008; p.137). In the initial stages, preference was given to large families who were landless, but subsequently, smaller families were preferred. Age, extent of formal education and farming experience were also taken into consideration. Early settlers were given large allotments of 3.2 ha—comprising 2 ha of paddy land and 1.2 ha of highland—but over the years, the unit of allotment has progressively declined to 1.2 ha 284 per family (Survey Department of Sri Lanka 2008; p.137). In settlement programmes, in addition to distributing land, the state also provided infrastructure facilities such as irrigation, roads and other common amenities (Bastian 2003; p.6). In keeping with the provisions of the Land Settlement Ordinance of 1935, the early tenure policy permitted lease in perpetuity to settler families (Survey Department of Sri Lanka 2008; p.137). Present state policies are moving toward a form of freehold tenure, subject to regulatory state controls. Table A13 presents the extent of state land alienation between 1953 and 1985.

Table A13: Land alienation by the state, 1953 – 1985 (from Bastian 2003; p.7)

Type Extent (ha) Major colonisation 175,941 Village expansion 357,239 Highland settlements 13,565 Youth settlement 7,964 Regularisation of encroachments 205,762 Middle class allotments 55,019 Land grants (special provisions) 9,980 Rain-fed farming settlements 5,363 Total 830,833

284 1 ha of paddy land and 0.2 ha of highland.

383 The other major set of land reforms was enacted in the 1970s by the incumbent socialist-oriented coalition government. The Land Reform Law of 1972 created a ceiling of 25 acres285 of paddy land and 50 acres of other land for private ownership; privately owned land in excess of this ceiling was taken over (Peiris 1978). However, ceilings sizes and exceptions to the rule meant that only 563,411 acres were appropriated286. The Land Reform Law of 1975 extended state ownership to land under export oriented plantation crops—mainly tea, rubber and coconut. Until then, this land had been under corporate ownership, managed by a system of privately-owned Agency Houses (Bastian 2003; p.10).

The total extent of land acquired under these two regimes was about 1 million acres (405,000 ha), or 22% of the total area under agriculture in Sri Lanka; about 95% of the acquired land was located in the wet zone (Peiris 1978; p.617). Paddy land was virtually unaffected by the reforms, accounting for less than 2% of the total extent acquired and about 1.4 % of the total extent under paddy in Sri Lanka. In contrast, the acquisitions included over 60% of the country’s tea acreage, 30% of its rubber acreage and 10% of its coconut acreage (Peiris 1978; p.617). Table A14 presents further information regarding amounts of land acquired under land reforms of the 1970s.

Table A14: Extent of land acquired through the 1972 and 1975 land reforms (Peiris 1978; p.617)

Type of crop 1972 (ha) 1975 (ha) Total % of total Tea 56,395 96,150 152,545 38% Rubber 33,412 38,378 71,790 18% Coconut 45,536 2,592 48,129 12% Paddy 7,449 0 7,449 2% Other crops 12,263 32,020 44,284 11% Uncultivated 72,949 0 72,949 18% Total 228,004 169,141 397,146 100%

285 One ha is about 2.47 acres. 286 About 5,600 persons declared that they had assets in excess of the new limits; the total extent declared was 1.2 million acres.

384 Overall, only a small portion of the land acquired was distributed to individuals; most of it was handed over to various state organisations such as the Sri Lanka State Plantation Corporation and Janatha Estate Development Board, as evident in Table A15 (Bastian 2003; p.11).

Table A15: Principal recipients of land under the Land Reform Programme, up to October 1976 (Bastian 2003; p.11)

Approx. extents % of total Recipient alienated (ha) alienated Public sector agencies State Plantation Corporation 105,825 25% USAWASAMA (Up-country Cooperatives 29,583 7% Estates Development Board

JANAWASAMA (People's Estate 94,737 22% Development Board) Tea and Rubber Research Institutes 4,006 1%

Cooperatives Janawasama (Land Reform Cooperative) 19,425 5% Electoral Cooperatives 70,820 17% Special Cooperatives and others 16,187 4%

Peasant Smallholdings 46,539 11% Others 35,815 8% Total 422,937 100%

Unfortunately, over time, productivity of these plantations declined, as politicians began to use them as a means of dolling out patronage to their supporters, especially in terms of employment opportunities (Bastian 2003; p.11). Nationalisation placed a heavy economic burden on the nation, first for the compensation for the nationalised lands and then from the loss of production of the State Plantation Corporation and its mismanagement of plantations (Peebles 2006; p.142). Therefore, the more market- oriented governments of the 1990s started exploring options for the privatization of these estates. The process began in 1992, when ownership was transferred to state- owned Regional Plantation Companies whose management was privatised on a profit

385 sharing basis for five years (Bastian 2003; p.14). Next, 51% of the ownership of Regional Companies was sold to the managing companies. Finally, the remaining shares held by the state were sold through the stock market, and the private sector regained control over one of the most important industries in the country (Bastian 2003; p.14).

A2.6 Governmental and non-governmental organisations within Sri Lanka’s conservation landscape

As discussed in Chapter 2, the Department of Wildlife and the Forest Department are the two main government agencies responsible for managing Sri Lanka’s forests and biodiversity resources. Table A16 presents the vision, mission, responsibilities and organisation structure of these agencies.

Table A16: Vision, mission, responsibilities and organisation structure of the Forest Department and the Department of Wildlife (Ministry of Environment Sri Lanka 2010; p.53-54, 58)

Forest Department Department of Wildlife Vision Sustainably managed forest & tree Conservation of wildlife heritage resources for providing environmental for present and future generations services and meeting needs of forest products for the country while contributing to the national economy and well-being of the people Mission Conserve and develop forest resources in Work with other relevant Sri Lanka to ensure the prosperity of the stakeholders to conserve wildlife nation heritage through professional management Summary of Conserve and maintain adequate forest Conserve the country’s major cover for the conservation of soil, water, indigenous fauna and flora & responsibilities biodiversity, climate moderation and maintenance of its diversity social reasons Fulfil international obligations Build up plantation forests to meet the under Ramsar, CITES and Bonn country’s future timber requirements Conventions Ensure optimum use of forest land Enforce the Fauna and Flora through integrated and multiple land use Protection Ordinance (FFPO) approaches Afford protection to protected Ensure the scientific management and species under FFPO outside PAs sustainable development of the forest resources

386 Forest Department Department of Wildlife Ensure that environmental safeguards are included in all forest development and utilisation activities Establish links among institutions concerned in promoting conservation and forestry development Promote community participation forest resource conservation and management, e.g. through community-based forestry programs Conduct forestry education and extension programs Key divisions Forest inventory and management Elephant conservation Forestry research and education Visitor service management Social forestry & extension Community outreach Planning & monitoring Protection and law enforcement Protection & law enforcement. Planning and progress monitoring Wildlife health management Research and training Accounts and administration

As discussed in Chapter 2, several local NGOs play an important role in helping safeguard Sri Lanka’s biodiversity and environment. Table A17 presents information of some of the major organisations in this field.

Table A17: Key conservation-related NGOs in Sri Lanka (compiled with information from CCR 2009; BECT 2012; Environmental Foundation 2012; GMSL 2012; SLWCS 2012; WPNS 2012)

Organisation Focus Key activities/ projects; significant (date established) achievements Environmental Monitoring implementation of Appeared before the Supreme Court on Foundation environment-related several occasions to present matters of Limited (1981) legislation and litigating environmental concern, especially against abuses by the regarding to sand mining, sound and air government or other parties pollution Influencing environment- Runs centres dealing with environmental related policy injustice reporting and monitoring, Focus areas include pollution, environmental monitoring, EIA PAs, climate change, solid monitoring, environmental legal, policy waste management, and research and advocacy natural resources Conducts seminars, workshops and training courses on issues related to environmental conservation and law,

387 Organisation Focus Key activities/ projects; significant (date established) achievements EIA procedures, state of the environmental for the general public, NGOs, policy makers and community members Produced policy papers, technical papers and briefing papers on key environmental issues Centre for Providing innovative, science- Developed solutions to mitigate human- Conservation based solutions for elephant conflict Research biodiversity conservation Tracked elephant ranging patterns and Creating awareness of the behaviour using radio telemetry links between biodiversity and Conducted conservation education ecological health among communities living near national Promoting conservation parks education by motivating Collaborated with the Department of citizens to participate in Wildlife to develop projects that provide conserving threatened species locals with benefits from conservation and habitats Helping conservationists develop knowledge, skills and attitudes that will aid the practice of environmental conservation Wildlife and Prevent the destruction and Interact regularly with government Nature Protection harmful commercial bodies such as the Department of Society (1894) exploitation of species of wild Wildlife and the Forest Department to animals and plants, and provide them with advice and assistance wherever desirable and on key conservation-related issues such possible, to preserve wildlife as human-elephant conflict and PA intact in natural conditions in administration Sri Lanka. Organise lectures on nature and Negotiate and co-operate with conservation the state and public bodies to Conducted conservation-related establish and support national campaigns such as the successful 1972 reserves and sanctuaries, and campaign to save the Sinharaja Forest assist in their administration. from destructive exploitation (this area has now been declared a World Heritage Site) Sri Lanka Wildlife Conducting sociological, Conducted wetland conservation projects Conservation economical and ecological in areas including the Bellanwila- Society (1997) assessments of the diverse Attidiya Sanctuary, the Bandaragama habitats of threatened species, Wetlands and the Kalpitiya-Karaitive- landscapes, and communities Puttalam Coastal Wetland Complex Developing ecological Through the ‘Saving Elephants by awareness within Sri Lanka Helping People’ project, conducted and internationally. activities to mitigate human-elephant Promoting cooperation conflict such as establishing solar between government powered electric fences along with

388 Organisation Focus Key activities/ projects; significant (date established) achievements institutions and local community development activities communities in order to find strategic long-term solutions to key environmental issues The Biodiversity Mitigating human-elephant Conducted awareness programmes and Elephant conflict and conserving Sri amongst the school children in the rural Conservation Lanka’s biodiversity areas where there is ongoing human- Trust (1988) elephant conflict Organised training programmes for veterinarians with international trainers, e.g., sent staff from the Department of Wildlife and the National Zoological Gardens on foreign training programmes Green Movement A consortium of 153 Conducts various projects in natural of Sri Lanka community-based resource conservation and management, organisations and other groups sustainable agriculture, disaster response that are involved in natural and climate change resource management, environmental conservation, consumer protection, community development, human rights, youth and women’s empowerment and sustainable agriculture. Achieve natural resource- based sustainable development through the empowerment of the poorest sections of the population and conservation of the environment.

389 A2.7 Important environmental laws and key international agreements that have been ratified by the Sri Lankan government

As discussed in Chapter 2, several key pieces of legislation help protect the country’s environment and biodiversity; these are introduced in Table A18.

Table A18: Selected environmental laws in Sri Lanka and the authorities in charge of supervision, regulation and/or enforcement (adapted from Environmental Foundation 2009; p.28-31)

Law Summary of intent/ coverage Authority in charge of supervision, regulation and/or enforcement National Environmental Establishes the Central Environmental Central Environmental Act No. 47 of 1980 (as Authority (CEA) and defines its Authority amended by Acts No. 56 powers, functions and duties. Provides of 1988 and 53 of 2000) overall environmental protection legislation, including licensing procedures, environmental standards and project approval procedures. Fauna and Flora Provides for the conservation of plants Department of Wildlife, Protection Ordinance and animals, which have been declared Director General of No. 2 of 1937 (as as protected species. Empowers the Wildlife amended by Acts No. 49 Minister to declare any area of state of 1993, 12 of 2005) land as a National Reserve or Sanctuary. Forest Ordinance No. 16 Consolidates the laws relating to Forest Department, of 1907 (as amended) forests and to the felling and Conservator General of transportation of timber. Empowers the Forests Minister to declare any area of state land as a Reserved Forest, Conservation Forest or a Village Forest. State Lands Ordinance Provides for how state lands and their Ministry of Agricultural No. 8 of 1947 (as resources, including lakes, rivers and Development, District amended) streams, should be allocated, used and Secretaries managed. Also provides for the declaration of state reservations Coast Conservation Act Identifies coastal zones and regulated Coast Conservation No. 57 of 1981 (as activities within such zones Department, Ministry of amended) Fisheries and Aquatic Resources Marine Pollution Provides for the prevention, reduction, Marine Environment Prevention Act No. 35 and control and management of marine Protection Authority of 2008 pollution in the territorial waters of Sri Lanka, any other maritime zone, the foreshore and the coastal zone of Sri

390 Law Summary of intent/ coverage Authority in charge of supervision, regulation and/or enforcement Lanka. Also provides for the establishment of the Marine Environment Protection Authority National Heritage Provides for the declaration, protection Forestry Department, Wilderness Areas Act and preservation of any area of State Ministry of Agricultural No. 3 of 1988 land with unique ecosystems, genetic Development resources or outstanding natural features as National Heritage Wilderness Areas Soil Conservation Act Provides for the conservation of soil Ministry of Agricultural No. 25 of 1951 (as resources, mitigation of soil erosion Development amended) and the protection of lands against flood and drought Plant Protection Act No. Provides for the prevention of wild Department of Agriculture 35 of 1999 plants, weeds and plant diseases and controls the introduction of new plant species Mahaweli Authority of Established the Mahaweli Authority of Mahaweli Authority of Sri Sri Lanka Act No. 23 of Sri Lanka and provides for the Lanka 1979 (as amended) conservation and maintenance of the physical environment of Mahaweli Areas, including watershed management, soil erosion and the protection of reservation areas.

As discussed in Chapter 2, Sri Lanka has ratified several international agreements related to biodiversity conservation; some these are presented in Table A19.

Table A19: Key international agreements related to the conservation of biodiversity which have been ratified by Sri Lanka (adapted from Ministry of Environment Sri Lanka 2010; p.47)

Sri Lanka’s date of International agreement entry (i.e. date ratified) International Plant Protection Convention 1952 UNESCO Man and the Biosphere Programme 1970

Convention on International Trade in Endangered Species of Wild 1979 Fauna and Flora, 1973 (CITES)

Convention Concerning the Protection of the World Cultural and 1980 Natural Heritage, 1972 (World Heritage Convention)

391 Sri Lanka’s date of International agreement entry (i.e. date ratified)

Convention on the Conservation of Migratory species (CMS) 1979

Convention on Wetlands of International Importance especially as 1990 Waterfowl Habitat, 1971 (Ramsar Wetland Convention)

Principles for a global consensus on the management, conservation and sustainable development of all types of forests, UNCED 1992 (Forest 1992 Principles)

Convention on Biological Diversity, 1992 (CBD) 1994 Cartagena Protocol on Biosafety 2004

392 Appendix 3: Sri Lanka’s imperilled national icon

A3.1 Human-elephant conflict across the range of the Asian elephant

As pointed out in Chapter 3, there has been growing human-elephant conflict across all Asian elephant range states. Some examples are presented in Table A20.

Table A20: Examples of human-elephant conflict reported from Asian elephant range states (from Perera 2009; p.42-46)

Country Reports of human-elephant conflict (HEC) in elephant range states Bangladesh During a period of one year, from June 2001 to May 2002, HEC occurred in 28% of the elephant range and resulted in 38 deaths and 94 injuries to humans, as well as three elephant deaths and damage to crops and households amounting to US$ 86,000. This is particularly significant because the entire country only has 200 elephants, although some may migrate from India and Myanmar. Cambodia There are two main populations of elephants, each comprising about 100 animals. Most incidents of HEC have occurred in the south and southwest regions. They have been increasing steadily, with crop raiding and damage to cottages reaching a frequency of two per month in some areas. India Each year, the HEC results in about 300 human deaths and damage to 10,000 – 15,000 houses and 8 – 10 million ha of crops. Over 200 elephants die due to human-related activities, which include poaching for ivory or meat, poisoning, electrocution and collision with trains. Indonesia The important elephant habitats are in the provinces of South Sumatra and Lampung in the south, Aceh in the north and Riau in the east. The high human population density of Lampung province and the close proximity of settlements to the remaining elephant habitats have resulted in a high degree of HEC. Between June 2000 and September 2002, 717 crop damage incidents were reported; elephants destroyed 21 houses, killed three people and disabled another three persons. Laos Nearly two thirds of the human population in Laos live in rural areas and practice shifting cultivation; many do so within or near the country’s PA network. Increasing levels of HEC are now a social and economic issue. In addition, as many development projects are planned or underway in areas inhabited by elephants the problem is likely to worsen. An on going study in the Nakai Plateau, where a hydroelectric project has inundated a large area of elephant habitat in and adjacent to the Nakai-Nam Theun PA, showed an increase in HEC incidents from 7 to 11 per month over a four year period. Nepal HEC occurs in the areas around elephant habitats, with crop damage being the main consequence. The number of deaths attributed to the conflict during the past 20 years is 66 humans and 18 elephants. Sri Lanka Between 1992 and 2001, wild elephants killed 536 people; of these 75% were men, 13% women and 12% children. On average, the HEC results in

393 Country Reports of human-elephant conflict (HEC) in elephant range states deaths of about 200 elephants and 70 humans each year. A more detailed account of the HEC in Sri Lanka is presented in Chapter 4. Thailand The Kaengkrachan-Kuiburi complex in southwest Thailand contains important elephant habitat; the area contains four PAs with around 150 elephants. Since the complex has over 450 agricultural areas around it, HEC is a common occurrence. In 2005, the value of crops damaged was over 4 million Baht. Vietnam HEC has reached crisis levels in some areas, with 26 people killed over a two-year period at one site.

A3.2 Results of the 2011 census of Sri Lankan elephants

An island-wide elephant census conducted in August 2011 estimated that there are almost 6,000 elephants in Sri Lanka (Hettiarachchi 2011). Table A21 and Figure A3 summarise the key results from this census.

Table A21: Distribution of wild elephants in Sri Lanka according to the 2011 elephant census (data from Hettiarachchi 2011)

Area Number of elephants % of total Northern region 233 4% North western region 1,189 20% Central region 47 1% Mahaweli region (north eastern) 1,751 30% Eastern region 1,573 27% Southern region 1,086 18% Total 5,879 100%

394

Figure A3: Results of 2011 census of wild elephants in Sri Lanka (data from Hettiarachchi 2011)

395 Appendix 4: An approach to mitigate Sri Lanka’s human- elephant conflict

A4.1 The Department of Wildlife’s approach to mitigating human- elephant conflict: methods and costs

As described in Chapter 4, the Department of Wildlife spends about 50 per cent of its working time and allocated budget to mitigate human-elephant conflict. It does this using a variety of methods that include the translocation of problem animals, elephant drives and the establishment of electric fences. Tables A22 and A23 provide some details on elephant drives and elephant translocations conducted in recent years. It is important to note that the drives listed in Table A22 are only small and medium-sized drives. The last major drive was conducted in 2005 – 2006 in the south under the Walawe Left Bank Development Project, taking 1.5 years to complete and costing US$ 1.6 million (Fernando et al. 2011; p.98). Tables A24 and A25 present information on some other activities conducted by the Department of Wildlife to mitigate human- elephant conflict: distribution of thunder flashes to people living in conflict areas and the payment of compensation for deaths, injuries and property damage.

Table A22: Details of elephant drives conducted by the Department of Wildlife by year, 2007 – 2011 (Fernando et al. 2011; p.98)

Year Number of Numbers of Cost (Rs.) Cost (US$) drives elephants chased 2007 4 86 – 91 445,000 4,100 2008 4 242 – 252 5,000,000 45,500 2009 12 341 – 354 3,500,000 31,800 2010 11 438 – 453 5,300,000 48,200 Total 31 1107 – 1150 14,245,000 129,600

396 Table A23: Elephant translocations conducted by the Department of Wildlife by year, 2007 – 2011 (Fernando et al. 2011; p.99)

Year Number of translocations Number of elephant deaths 2007 13 0 2008 10 1 2009 18 1 2010 25 1 2011 2 1 Total 68 4

Table A24: Annual cost and number of ‘elephant thunders’ (thunder flashes) distributed by the Department of Wildlife to people living in areas where there is human-elephant conflict (Fernando et al. 2011; p.98)

Year Number of thunders Cost (Rs.) Cost (US$) 2007 275,000 20,000,000 182,000 2008 275,000 20,000,000 182,000 2009 355,000 25,000,000 227,000 2010 514,000 42,000,000 382,000 2011 514,000 42,000,000 382,000 Total 1,933,000 149,000,000 1,355,000

Table A25: Annual compensation paid by the Department of Wildlife, 2005 – 2010 (Fernando et al. 2011; p.99)

Year For deaths (Rs.) For injuries For property Total (Rs.) (Rs.) damage (Rs.) 2005 4,600,000 490,000 1,730,000 6,820,000 2006 6,090,000 740,000 2,850,000 9,680,000 2007 5,000,000 980,000 5,940,000 11,920,000 2008 5,680,000 1,160,000 7,000,000 13,840,000 2009 6,580,000 1,030,000 9,030,000 16,640,000 2010 11,990,000 1,050,000 25,690,000 38,730,000 Total 39,940,000 5,450,000 52,240,000 97,630,000

397 Appendix 5: Introduction to case study

A5.1 Development activities within Sri Lanka’s southeast landscape

As described in section 5.1, there have been several state-led development projects in southeast Sri Lanka since President Mahinda Rajapaksa—whose political base is in the area—was elected in 2005. Table A26 describes some major projects.

Table A26: Examples of on-going and planned infrastructure development in southeast Sri Lanka (sources included within table)

Development Relevant Estimated Details activities/ projects district(s) cost (US$ milli ons) Hambantota Port Hambantota 7,000 Planned as a service and industrial port (Naalir 2010) providing services such as bunkering, ship repairing and logistics Phase 1 completed in 2012 Phases 2 and 3 to be completed over the following 5 – 10 years Mattala Hambantota 210 Second international airport in Sri Lanka International Land area: 2,000 ha Airport (Sirimane Proposed completion: 2012/2013 2012) Walawe Left Bank Hambantota 36 Phase I: Rehabilitation and upgrading of irrigation scheme 2,900 ha of existing irrigated lands and the development of 1,040 ha of new lands in Walawe Left Bank. Phase 2: 5,150 ha of new lands opened up for agriculture and settling of 9,000 people International Hambantota Not New floodlit international cricket cricket stadium available stadium, which was the venue for several (Dhambarage World Cup cricket matches played in Sri 2010) Lanka in 2011; it has a capacity of 35,000 Administrative Hambantota 16 A five-storied administrative complex to complex (Nizam house all major government offices in the 2011) district (completed in 2011) Rural Hambantota 36 The goal is to increase rural electrification electrification from 65% to 85% of (Sirimane 2009) households in district Oluvil Port Ampara 36 Part of project to develop a chain of (Krishnaswamy coastal harbours

398 Development Relevant Estimated Details activities/ projects district(s) cost (US$ milli ons) 2010) Total area: 165 ha Completed: 2010 Uma Oya Moneragala 16,000 Would construct two reservoirs on the Multipurpose Uma Oya river, generating 90 mega watts project (Sirimane of electricity and irrigating 5,000 ha of 2010) land Milk processing Moneragala 16 Would have the capacity to collect 15,000 plant (Sirimane litres of milk per day; will benefit over 2010) 20,000 farming families Rehabilitation of Hambantota, 25 Rehabilitation of 207 km of roads in all roads (Government Ampara, three districts of Sri Lanka 2005; Moneragala p.90)

Some of the projects listed in Table A26—such as the cricket stadium and administrative complex—have been completed, while others such as the airport and the port are still in progress. The rural electrification schemes and road building programmes have been completed in some areas.

A5.2 Visitation and revenues from parks within Sri Lanka’s southeast landscape

As described in section 5.2, the PAs located within my chosen landscape account for 55% of total visitors and 65% of total revenue from PAs in Sri Lanka. Visitor numbers to parks and revenues from these visitors in 2010 are presented in Table A27 and Table A28 respectively.

399 Table A27: Visitor numbers to Sri Lankan parks, 2010 (Department of Wildlife Conservation 2011); the parks within my study area are highlighted in green

Park Local Foreign Total % of total Yala 179,965 73,580 253,545 40% Horton Plains 144,892 21,846 166,738 26% Uda Walawe 45,752 15,560 61,312 10% Minneriya 27,714 13,479 41,193 7% Kaudulla 16,731 13,085 29,816 5% Bundala 5,920 3,703 9,623 2% Wilpattu 16,156 912 17,068 3% Yala East (Kumana) 19,201 400 19,601 3% Wasgamuwa 19,916 451 20,367 3% Lunugamvehera 3,601 20 3,621 1% Others 7,417 162 7,579 1% Total 487,265 143,198 630,463 100%

Table A28: Visitor revenue from Sri Lankan parks, 2010 (Department of Wildlife Conservation 2011); the parks within my study area are highlighted in green

Park Locals (Rs.) Foreign Other, e.g. Total (Rs.) % of (Rs.) vehicles total Yala 10,274,270 123,850,108 10,793,900 144,918,278 52% Horton Plains 7,763,760 36,646,400 4,593,650 49,003,810 18% Uda Walawe 2,536,968 22,718,483 2,691,850 27,947,301 10% Minneriya 1,604,320 22,602,970 1,721,400 25,928,690 9% Kaudulla 652,670 14,679,538 1,351,875 16,684,083 6% Bundala 228,700 4,183,437 471,925 4,884,062 2% Wilpattu 876,140 1,503,635 641,850 3,021,625 1% Yala East (Kumana) 820,516 445,750 560,850 1,827,116 1% Wasgamuwa 726,380 443,500 577,675 1,747,555 1% Lunugamvehera 140,780 21,583 131,800 294,163 0% Others 299,730 153,836 188,850 642,416 0% Total 25,924,234 227,249,239 23,725,625 276,899,098 100%

400 A5.3 Rice production within Sri Lanka’s southeast landscape

As described in section 5.2, my study area is located mainly in the three districts of Hambantota, Ampara and Moneragala that comprise some of Sri Lanka’s main rice- producing areas. Table A29 presents information on rice production by district for the 2009/2010 Maha season and the 2010 Yala season. Rice production for the Mahaweli and Uda Walawe irrigation schemes are presented separately since they each cut across several districts287.

Table A29: Rice production by district for the 2009/2010 Maha season and the 2010 Yala season (compiled using data from Department of Census and Statistics 2011)

District Total extent % of total Total harvest % of total sown (ha) sown (metric tons) harvest Ampara 129,116 12% 634,187 15% Polonnaruwa 118,613 11% 547,366 13% Kurunegala 125,440 12% 499,502 12% Anuradhapura 104,350 10% 410,562 10% Batticaloa 75,756 7% 267,180 6% Hambantota 50,353 5% 226,125 5% Mahaweli 'H' 40,836 4% 224,360 5% Monaragala 41,376 4% 175,608 4% Uda Walawe 27,385 3% 171,916 4% Trincomalee 39,651 4% 165,606 4% Badulla 39,688 4% 147,427 3% Matale 30,967 3% 122,545 3% Matara 31,482 3% 87,158 2% Puttlam 24,991 2% 76,683 2% Kalutara 27,394 3% 74,772 2% Galle 28,888 3% 70,990 2% Ratnapura 24,948 2% 69,956 2% Kandy 23,958 2% 66,445 2% Gampaha 20,362 2% 61,035 1% Kegalle 15,553 1% 48,811 1% Mannar 8,693 1% 43,995 1%

287 This is the manner in which the numbers are compiled and presented by Sri Lanka’s Department of Census and Statistics.

401 District Total extent % of total Total harvest % of total sown (ha) sown (metric tons) harvest Vavuniya 5,664 1% 28,142 1% Colombo 7,420 1% 21,434 < 1% Nuwara Eliya 8,616 1% 20,380 < 1% Jaffna 8,948 1% 20,117 < 1% Malativu 3,363 < 1% 14,960 < 1% Killinochchi 1,468 < 1% 3,360 < 1% Total 1,065,279 100% 4,300,622 100%

A5.4 Semi-structured interviews conducted

I conducted semi-structured interviews with various stakeholder groups who have a stake in mitigating human-elephant conflict, and consequently could influence MER establishment. This section provides details on the interviews I conducted.

I spoke to five main groups: farmers, government officials, tourism companies, other locals who benefit from nature tourism (e.g. jeep drivers) and non-governmental organisations288. Table A30 presents the numbers of interviews I conducted with each stakeholder group.

Table A30: Numbers of interviews conducted by stakeholder group

Stakeholder group Numbers of interviews conducted Farmers 69 Government officials 12 Tourism companies 8 Other locals 13 NGOs 10 Total 112

288 In Chapter 6, I listed the Sri Lankan public as a stakeholder group. However, I did not conduct semi- structured interviews with this group for several reasons. First, Clem Tisdell and Ranjith Bandara have published results of detailed surveys conducted on this group, which I have used in my stakeholder analysis. Secondly, this stakeholder group is not as influential as some of the other groups, since, by definition, this group excludes members of the public who are part of the other five stakeholder groups. Finally, due to the variety of opinions it would be difficult to get a representative sample, and I did not have time and resources to conduct these types of interviews.

402 I now take each of the groups in turn, describing my interview objectives, and the people I interviewed. In accordance with ethics guidelines of the University of New South Wales, I have excluded names and other information that would enable individuals to be identified.

Farmers

Most of my interviews in this category were with two main farmer groups: chena farmers and paddy farmers. However, I also interviewed a few other locals involved in farming such as those involved in banana cultivation and people who cultivated crops in their home gardens. Table A31 presents details on the interviews I conducted. The aim of these interviews was to get an understanding of:

• Demographics and other basic information (e.g. family size, number of years in area, number of years cultivating the relevant crops, land ownership);

• Livelihoods and sources of income;

• Threats to livelihoods, especially those linked to natural disasters and wild animals (especially elephants);

• Other issues faced, e.g. land tenure, political, social;

• Views on conservation, PAs, state conservation agencies and human-elephant conflict;

• Perceptions on the area’s state-led projects, e.g. road construction, irrigation projects; and,

• Support mechanisms, e.g. community-based mechanisms and/or projects/ activities by NGOs and private sector entities.

403 Table A31: Details of interviews conducted with farmers

Assigned ID Occupation Interview location Date Farmer 1 Chena farmer Yala National Park buffer zone 06/04/11 Farmer 2 Chena and rice farmer Yala National Park buffer zone 06/04/11 Farmer 3 Chena and rice farmer Yala National Park buffer zone 06/04/11 Farmer 4 Chena and rice farmer Yala National Park buffer zone 06/04/11 Farmer 5 Chena and rice farmer Yala National Park buffer zone 06/04/11 Farmer 6 Chena and rice farmer Yala National Park buffer zone 06/04/11 Farmer 7 Chena and rice farmer Yala National Park buffer zone 06/04/11 Farmer 8 Chena farmer Yala National Park buffer zone 06/04/11 Farmer 9 Rice farmer Hambantota, near Bundala National 07/04/11 Park Farmer 10 Rice farmer Hambantota, near Bundala National 07/04/11 Park Farmer 11 Banana farmer Walawe left bank 07/04/11 Farmer 12 Chena farmer Walawe left bank 07/04/11 Farmer 13 Chena farmer Walawe left bank 07/04/11 Farmer 14 Rice farmer Meda Para, Hambantota 25/04/11 Farmer 15 Rice farmer Meda Para, Hambantota 25/04/11 Farmer 16 Chena farmer Meda Para, Hambantota 25/04/11 Farmer 17 Chena and rice farmer Off Meda Para, Hambantota; called 25/04/11 Galwewa in Kaliyapura Farmer 18 Rice farmer Ellala, Hambantota 25/04/11 Farmer 19 Chena and rice farmer Thammannawa 25/04/11 Farmer 20 Rice farmer (other crops Onion fields near Pidurangala (near 28/08/11 in off season) Sigiriya) Farmer 21 Rice farmer (other crops Onion fields near Pidurangala (near 28/08/11 in off season) Sigiriya) Farmer 22 Rice farmer (other crops Onion fields near Ilukwala (near 28/08/11 in off season) Sigiriya) Farmer 23 Rice farmer (other crops 28/08/11 in off season) Ilukwala village (near Sigiriya) Farmer 24 Rice farmer Ilukwala village (near Sigiriya) 28/08/11 Farmer 25 Rice farmer Uda Malala village, off Tissa 04/09/11 Hambantota Road Farmer 26 Rice farmer Tissa Hambantota Road-- 04/09/11 Koholankala Farmer 27 Rice farmer Tissa Hambantota Road-- 04/09/11 Koholankala

404 Assigned ID Occupation Interview location Date Farmer 28 Chena and home garden Tissa Hambantota Road 04/09/11 Farmer 29 Chena and home garden Tissa Hambantota Road 04/09/11 Farmer 30 Chena and home garden Tissa Hambantota Road 04/09/11 Farmer 31 Rice farmer Hambantota Gonnoruwa Road-- 04/09/11 Welipokuna gama Farmer 32 Rice farmer Maha Handilla 04/09/11 Farmer 33 Home garden (chena Ketenwewa village, off Hambantota 05/09/11 crops) Gonnoruwa road Farmer 34 Chena and rice farmer Ketenwewa village, off Hambantota 05/09/11 Gonnoruwa road Farmer 35 Rice and home garden Ketenwewa village, off Hambantota 05/09/11 Gonnoruwa road Farmer 36 Chena and rice farmer Kurudana village, off Hambantota 05/09/11 Gonnoruwa road Farmer 37 Rice and home garden Ketenwewa village, off Hambantota 05/09/11 Gonnoruwa road Farmer 38 Rice and home garden On Hambantota Gonnoruwa road, 05/09/11 near wel yaya near Kurudana Farmer 39 Rice and home garden Thammanna village 05/09/11 Farmer 40 Rice and home garden Thammanna village 05/09/11 Farmer 41 Rice farmer Thammanna village 05/09/11 Farmer 42 Chena and rice farmer Thammanna village 05/09/11 Farmer 43 Chena and rice farmer Fields near Yala National Park, near 06/09/11 Bandu wewa Farmer 44 Chena and rice farmer Fields near Yala National Park, near 06/09/11 Bandu wewa Farmer 45 Chena and rice farmer Fields near Yala National Park, near 06/09/11 Bandu wewa Farmer 46 Chena and rice farmer Fields near Yala National Park, near 06/09/11 Bandu wewa Farmer 47 Chena and rice farmer Fields near Yala National Park, near 06/09/11 Bandu wewa Farmer 48 Chena and rice farmer Fields near Yala National Park 06/09/11 Farmer 49 Chena and rice farmer Fields near Yala National Park 06/09/11 Farmer 50 Chena and rice farmer Fields near Yala National Park 06/09/11 Farmer 51 Chena farmer Kattadi wewa 06/09/11 Farmer 52 Chena and rice farmer Off the Lunugamvehera 21/02/12 Thanamalwila Road Farmer 53 Rice farmer Off the Lunugamvehera 21/02/12

405 Assigned ID Occupation Interview location Date Thanamalwila Road Farmer 54 Rice farmer Off the Lunugamvehera 21/02/12 Thanamalwila Road Farmer 55 Chena farmer Off the Lunugamvehera 21/02/12 Thanamalwila Road Farmer 56 Chena farmer Off the Lunugamvehera 21/02/12 Thanamalwila Road Farmer 57 Banana farmer and other Off the Lunugamvehera 21/02/12 fruits Thanamalwila Road Farmer 58 Paddy and chena farmer Off the Lunugamvehera 21/02/12 Thanamalwila Road Farmer 59 Paddy farmer Near Koholankala 22/02/12 Farmer 60 Paddy and chena farmer Road from Lunugamvehera to 22/02/12 Mattala Farmer 61 Chena farmer Road from Lunugamvehera to 22/02/12 Mattala Farmer 62 Chena farmer Road from Lunugamvehera to 22/02/12 Mattala Farmer 63 Rice farmer Uda Mattala 22/02/12 Farmer 64 Rice and chena farmer Uda Mattala 22/02/12 Farmer 65 Rice and chena farmer Uda Mattala 22/02/12 Farmer 66 Chena farmer Sithulpahuwa road 23/02/12 Farmer 67 Chena farmer Sithulpahuwa road 23/02/12 Farmer 68 Chena farmer Sithulpahuwa road 23/02/12 Farmer 69 Chena farmer Sithulpahuwa road 23/02/12

Government officials

I conducted interviews with officials from government entities that were the most significant stakeholders in the conflict and the suggested resolution: the Department of Wildlife which is responsible for elephant conservation and management in Sri Lanka, and the Forest Department, which has jurisdiction over most elephant habitat outside PAs, i.e. within MERs. I also conducted interviews with other government officials who could be important in the process of implementing the suggested approach, such as

406 those from the Sri Lanka Tourism Development Authority and local government officials. The aim of the interviews was to get an understanding of:

• Department responsibilities, organisation structure and individual’s role;

• Key department policies and initiatives—especially those related to human- elephant conflict and related issues such as land management—and the department’s role in their implementation (if any);

• Interaction with other departments related to human-elephant conflict and related issues such as land management;

• View on current approaches to mitigate human-elephant conflict and issues with these approaches; and,

• Potential approaches that should be explored, and views on concepts such as the landscape approach and conservation through sustainable use.

The exact questions and the time allocation among questions were tailored to the specific interviewee based on his/her department and position within it. Table A32 presents details on interviews I conducted.

Table A32: Details of interviews conducted with government officials

Assigned ID Department Interview location Date Government 1 Department of Wildlife Yala East National Park 23/04/11 Government 2 Local government Office in at the Back of 28/08/11 Beyond Lodge, Sigiriya Government 3 Department of Wildlife Back of Beyond Lodge, 29/08/11 Sigiriya Government 4 Department of Wildlife Bundala National Park 22/02/12 Government 5 Department of Wildlife Maduru Oya National Park 26/02/12 Government 6 Department of Wildlife Maduru Oya National Park 26/02/12 Government 7 Department of Wildlife Home near the Maduru Oya 27/02/12 National Park Government 8 Department of Wildlife Wilpattu National Park 03/03/12 Government 9 Forest Department Headquarters, Battaramulla 06/03/12 Government 10 Forest Department Headquarters, Battaramulla 06/03/12 Government 11 Sri Lanka Tourism Office, Colombo 07/03/12

407 Assigned ID Department Interview location Date Development Authority Government 12 Sri Lanka Tourism Office, Colombo 07/03/12 Development Authority

Tourism companies

I interviewed representatives from tourism companies that are involved in nature tourism in Sri Lanka. Table A33 presents details on the interviews I conducted. The aim of these interviews was to get an understanding of:

• Company business, strategy and interest in nature tourism in Sri Lanka;

• Key company successes;

• Views on industry potential and future growth;

• Issues and barriers to industry growth;

• Views on human-elephant conflict and potential solutions; and,

• View of potential landscape-based approaches outside PAs such as proposed MERs, and, potential company interest and role in such initiatives.

Table A33: Details of interviews conducted with tourism companies

Assigned ID Organisation Interview location Date Tourism operator 1 Jetwing Eco-Holidays Office, Colombo 13/09/11 Tourism operator 2 John Keels Group Office, Colombo 15/09/11 Tourism operator 3 Kulu Safari Yala National Park 19/02/12 Tourism operator 4 Kulu Safari Yala National Park 19/02/12 Tourism operator 5 Kulu Safari Yala National Park 19/02/12 Tourism operator 6 Kulu Safari Yala National Park 19/02/12 Tourism operator 7 Jetwing Eco-Holidays Office, Colombo 25/02/12 Tourism operator 8 John Keels Group Office, Colombo 25/02/12

408 Other locals

I interviewed other locals who benefit from nature tourism in the southeast landscape, and are therefore potential stakeholders regarding the proposed MERs. These included jeep drivers (and owners) who make a living by taking tourists to national parks, and local retailers. Table A34 presents details on the interviews I conducted. The aim of these interviews was to get an understanding of:

• Business and how it benefits from nature tourism;

• Basic information about business, e.g. size, operating model, revenue/profits, seasonality, client base;

• The impact of having PAs within the landscape (on business and well on other locals in the area);

• Business-related challenges faced in the past (e.g. due to terrorism) and the present; suggestions to overcome these challenges, and key people who should be involved;

• Views on the growth of nature tourism in the area; and,

• Views on potential landscape-based approaches outside PAs such as the proposed MERs, and, potential interest and role in such initiatives.

Table A34: Details of interviews conducted with other locals

Assigned ID Occupation/ business Interview location Date Other locals 1 Shop owner Tissamaharama town 21/03/11 Other locals 2 Shop owner Pahatha Mattala 07/04/11 Other locals 3 Jeep driver Minneriya National Park 27/08/11 Other locals 4 Tourist driver Back of Beyond Lodge, Sigiriya 28/08/11 Other locals 5 Hotel employee Back of Beyond Lodge, Sigiriya 29/08/11 Other locals 6 Jeep driver Near Yala NP 06/09/11 Other locals 7 Shop owner Off the Lunugamvehera 21/02/12 Thanamalwila Road Other locals 8 Agricultural advisor Hambantota Weerawila Road, 22/02/12 turnoff to Bundala NP Other locals 9 Villager from colony 5, Hambantota Weerawila Road, 22/02/12

409 Assigned ID Occupation/ business Interview location Date near Koholankala turnoff to Bundala NP Other locals 10 Village priest Hambantota Weerawila Road, 22/02/12 turnoff to Bundala NP Other locals 11 Villager from colony 5, Hambantota Weerawila Road, 22/02/12 near Koholankala turnoff to Bundala NP Other locals 12 Shop owner and chena Off Sithulpahuwa Road 23/02/12 farmer Other locals 13 Shop owner Off the Lunugamvehera 21/02/12 Thanamalwila Road

Non-governmental organisations

I interviewed members of non-governmental organisations who are stakeholders in the human-elephant conflict and the suggested mitgation approach. These included NGOs involved in elephant conservation as well as those interested in improving livelihoods of locals who are affected by the conflict. Table A35 presents details on the interviews I conducted. The aim of these interviews was to get an understanding of:

• Focus of the organisation—major objectives;

• Approach to achieve objectives, key initiatives and projects;

• Key stakeholders of organisation;

• Views on human-elephant conflict: conflict drivers and impact on elephants and humans;

• Views on current approaches to mitigating the conflict and their effectiveness;

• Other solutions that should be tried out; and,

• Views on potential landscape-based approaches outside PAs such as proposed MERs, and, potential NGO interest and role in such initiatives.

410 Table A35: Details of interviews conducted with NGOs

Assigned ID Organisation Interview location Date NGO 1 Centre for Conservation Research Office, Tissamaharama 25/03/11 NGO 2 Environmental Foundation Limited Office, Colombo 20/09/11 NGO 3 UNDP Office, Colombo 5/03/12 NGO 4 Environmental Foundation Limited Office, Colombo 6/03/12 NGO 5 World Bank Office, Colombo 8/03/12 NGO 6 World Bank Office, Colombo 8/03/12 NGO 7 Centre for Conservation Research Colombo 8/03/12 NGO 8 Centre for Conservation Research Colombo 9/03/12 NGO 9 Environmental Foundation Limited Office, Colombo 9/03/12 NGO 10 World Wildlife Fund Colombo 9/03/12

A5.5 World Bank survey and analysis

I had access to raw data from a survey of 800 farmers, conducted in southeast Sri Lanka by the World Bank in 2009. This was given to me in excel format, along with the relevant codes and information. I read these data into the SPSS Statistics Programme, in order to make it easier to conduct the required analysis (e.g. calculate frequencies and descriptive statistics) and stratify the data sample, for example, to separately analyse chena farmers and other farmers. The survey covered 800 households in 11 Grama Niladari divisions located near the Yala National park and around the Tissamaharama town. Table A36 provides details on sample allocation and coverage.

Table A36: Details of sample allocation and coverage of World Bank survey

GN name Number of Households Villages covered surveyed Yodakandiya 60 Yodakandiya Viharamahadevipura 80 Wilamulla, Pustholamulla, Suduwelipelessa, Bandarawatta Kirinda 100 Kirinda, Kirindagama, Karijjawila Magama 50 Magama Andaragasyaya 90 Pattiyawelayaya, Andaragasyaya, Kirinda godana

411 GN name Number of Households Villages covered surveyed Kawantissapura 90 Bogahapelessa, Kawantissagama, Bogahawewa, Samiguliya, Kurukumbiyakanda Julpallama 70 Julpallama, Bogahapelessa, Delikanu- ara, Kodigahawewa Weerahela 70 Kuda gammana 07 (Wam iwura), Kohombagahapelessa Mahasenpura 70 Thanja nagaraya, Tikiriwewa godana, Gonapattiya, Weliaragoda, Sandagirigama Ranakeliya 50 Yodakandiya, Koragaha ulpatha Uddagandara 70 Mahadehiara, Uddagandara, Thambarawa Total 800

These 800 households contained 3,321 individuals. Tables A37 and A38 provide details about their composition.

Table A37: Composition of households in World Bank survey (based on analysis of data from the World Bank Survey)

Category Number of people in category Male 1,718 Female 1,603 Number of children 1,540 Total number of people in survey 3,321

Table A38: Overall household statistics in World Bank survey (based on analysis of data from the World Bank Survey)

Household characteristics Value Number of households 800 % Male 52% % Female 48% Average number of children 1.93 Household size 4.15

412 In Chapter 5, I analysed the income and other characteristics of chena and non-chena households separately. Roughly a third of the households were chena households and the remainder were non-chena households. Statistics for each of these two categories are presented in Table A39.

Table A39: Comparison between chena and non-chena households in World Bank survey (based on analysis of data from the World Bank Survey)

Characteristic Chena Non chena Number of households 252 548 % Male 52% 52% % Female 48% 48% Average number of children 2.2 1.8 Household size 4.8 3.9

Tables A40 and A41 present further information about chena farmers. The former presents information on proportions of farmers within each village who cultivated chena, while the latter indicates the types of crops grown in chenas.

Table A40: Chena cultivation by village in World Bank survey (based on analysis of data from the World Bank Survey)

Village Total in sample Chena cultivated % chena cultivated (by village) Mahasenpura 70 10 14% Ranakeliya 50 25 50% Udahakandara 70 35 50% Yodakandiya 60 26 43% Viharamahadevipura 80 37 46% Kirinda 100 10 10% Magama 50 9 18% Andaragasyaya 90 32 36% Kawantissapura 90 38 42% Julpallama 70 23 33% Weerahela 70 7 10% Total 800 252 32%

413 Table A41: Crop types cultivated in chenas in World Bank survey (based on analysis of data from the World Bank Survey)

Crop type Sinhala name Frequency of famers cultivating crop289 Long beans Me Karal 159 Chilli Miris 90 Bitter-gourd Karawila 68 Green gram Mun Eta 50 Okra Bandakka 35 Cowpea Cowpea 35 Tomatoes Thakkali 35 Maize Iringu 16 Pumpkin Wattakka 15 Brinjals Wambatu 14 Cucumber Pipinna 11 Gingelly Thala 10 Finger Millet Kurakkan 6 Snake-guord Pathola 3 Ash plantain Alu Kesel 3 Centella Gotukola 2 Capsicum Malu Miris 1

A5.4 The elephant population within Sri Lanka’s southeast landscape

As described in section 5.4, my study area contains a significant elephant population as it comprises important elephant habitat—PAs such as the Yala Park Complex, Uda Walawe and Lunugamvehera, as well as surrounding forested areas. However, since the 2011 elephant census did not provide the data at the required level of detail, I estimated the landscape’s elephant population by combining census information with previous figures published by de Silva and de Silva (2007), which are presented in table A42:

289 Each farmer could indicate more than one crop. Many farmers did not answer this question, so the numbers are only indicative of the relative popularity of the crops among those who answered the question.

414 Table A42: Estimates of wild elephant populations contained within several important landscapes (de Silva and de Silva 2007; p.149); areas within my study area are highlighted in green.

Locality Area in km2 Estimate % of total (forest and other) Panama - Wellawaya - Kirinda area including 2,100 900 19% Yala Park Complex Lunugamvehera NP -Wirawila - Bundala area 750 175 4% Uda-walawe NP - Kuda Oya area 950 250 5% Lahugala NP - Heda Oya - Wil Oya area 575 175 4% Galoya NP - Ampara area 1,600 325 7% Sinharaja forest - Deniyaya - Rakwana area 350 15 0% Peak Wilderness - Sri Pada area 150 15 0% Victoria-Randenigala-Rantembe area 450 50 1% Maduru Oya NP area 1,350 350 7% Wasgamuwa - Somawathiya - Minneriya PA 3,350 850 18% complex Kalawewa - Thirappane - Ritigala area 1,950 425 9% Resvehera - Herathgama area 775 200 4% Thabbowa - Karuwalagaswewa - Vanathavillu 975 175 4% area Wilpattu NP area 2,250 350 7% Yodha Wewa - Madhu Road SA areas 975 175 4% Medawachchiya - Vavuniya - Padaviya area 1,350 150 3% Trincomalee - Yan Oya area 1,250 100 2% Killinochichi - Mullaittivu areas 2,350 150 3%

Total 23,500 4,830 100%

The figures in Table A42 indicate that about 30% of Sri Lanka’s elephants live in my study area (i.e. in the areas highlighted in green in the table). Combining this percentage with numbers from the 2011 elephant census (presented in Appendix 4, but reproduced in Table A43 below), I estimate that about 1,700 elephants live in my study area.

415 Table A43: Distribution of wild elephants in Sri Lanka according to the 2011 elephant census (data from Hettiarachchi 2011)

416 Appendix 6: Stakeholder analysis

A6.1 Compensating farmers for damages caused by elephants

As described in Chapter 6, according to a study by (Bandara and Tisdell 2003) most urban dwellers recognise that farmers should be compensated for the damages to their crops and property caused by elephants; in fact 82% of urban respondents were in favour of this (Bandara and Tisdell 2003; p.340). Over 70% believed that all urban dwellers should contribute to farmers’ compensation, while smaller proportions believed that the tourism industry, the private sector and/or the government should pay (Table A44).

Table A44: Urban respondents’ ranking of those who should be major contributors to farmers’ compensation (n=300) (adapted from Bandara and Tisdell 2003; p.340)

Suggested major contributors Percentage of the total number of respondents All city dwellers 71 Those who are interested in elephants 68 Tourist industry 62 The private sector 58 The government 33 Only high income earners in the country 10

A6.2 Using an influence-importance matrix to categorise stakeholders

According to the academic literature, a popular and effective way of categorising stakeholders is to use an ‘influence-importance matrix’, which has two axes: importance, referring to the extent of each stakeholder’s needs and interests in the issue at hand, and influence, referring to the power each stakeholder has over successfully implementing a solution (Grimble 1998; p.176). This methodology results in four key sets of stakeholders, as illustrated in Figure A4.

417 INTEREST

high Subjects Key players

low Crowd Context setters

low high INFLUENCE

Figure A4: Four key stakeholder categories arising from an influence-importance matrix (adapted from Reed et al. 2009; p.1938)

Reed et al. (2009) point out that an influence-importance matrix could help policymakers trying to implement a solution understand how to engage specific stakeholders. For example, ‘key players’ should be closely engaged given their high interest and influence regarding the issue at hand. While ‘context setters’ have significant influence, they have little interest and “because of this, there may be significant risk, and should be monitored and managed” (Reed et al. 2009; p.1938). ‘Subjects’ have significant interest, but they lack the capacity for impact, and often comprise marginalised groups who should be engaged for social equity reasons (Reed et al. 2009; p.1938). The ‘crowd’ have little interest or impact, and therefore there is little need to engage them. However, as pointed out in the next section, it is important to keep

418 in mind that assigning stakeholders in these set categories in often subjective and indeed problematic in some situations290. More details are provided in the next section.

A6.3 Potential issues with my stakeholder analysis and how these issues are addressed

As with any methodology, stakeholder analysis—its theoretical underpinnings and its practical use—has some weaknesses, despite its many benefits (discussed in section 6.1 of Chapter 6). In this section, I consider criticisms of stakeholder analysis and describe how these issues could be addressed (or have been addressed in my analysis), in the context of three main areas that critics focus on: how information is obtained, how stakeholders are categorised and how analysis findings are used.

How information is obtained

The main sources of information for my stakeholder analysis are my semi-structured interviews and the World Bank survey data. As with any interview or survey, these potentially reflect biases and miscommunications between interviewees and interviewers291 (Ritchie and Lewis 2003; p.275 - 276). This issue was mitigated to some extent since both the people who administered the survey and I conducted the interviews in Sinhala—the participants’ language. Furthermore, I was accompanied with a person working for a local NGO operating in the southeast landscape. This arrangement, to some extent, fulfilled what Ritchie and Lewis (2003; p.275) call ‘matching’:

Sharing some aspects of cultural background or experience may be helpful in enriching researchers' understanding of participants' accounts, of the language they use and of nuances and subtexts. The researcher's perceptions here should not be a substitute for the participant's own words,

290 Part of the issues arise from the way this matrix is used. For example, while the four categories identified above may be applicable in many business situations, they may be more problematic in natural resource management. 291 Me in the case of the semi-structured interviews and the Colombo-based staff of the Institute of Policy studies who conducted the World Bank survey.

419 but they can help researchers to make judgements about how to explore issues in more depth. However, interviewees’ responses could potentially have been influenced by what they though the interviewers ‘wanted to hear’ (Marvasti 2004; p.18-19). Moreover, the respondents may have thought that their answers may bring benefits and/or potentially cause them problems, as a person working with a local NGO accompanied me, and the survey was conducted under World Bank auspices 292 . I mitigated this possible perception to some extent through the (mandatory) process of disclosure293, explaining to interviewees at the outset the purpose of the interview, as well as how the results would be used, and emphasising that their participation was entirely voluntary and would not bring them any benefits (or cause them problems, as their identities would remain anonymous). Nonetheless, I do recognize that some sources of bias would remain, and highlighted those likely to be significant. For example, I pointed out in Chapter 5 (section 5.3) that self-estimates of farmers’ incomes were likely to be underestimates. Where possible, I also tried to crosscheck what I was told with findings of other similar research and with newspaper articles, to gauge their reasonableness (or if they were vastly different, to ask clarification questions from subsequent interviewees). This process is described as triangulation294 by Ritchie (2003; p.43):

Triangulation involves the use of different methods and sources to check the integrity of, or extend, inferences drawn from the data. It has been widely adopted and developed as a concept by qualitative researchers as a means of investigating the ‘convergence’ of both the data and the conclusions derived from them. Overall, I tried to identify sources of potential bias, and ask questions in a manner that addressed this dimension to the extent possible. However, when conducting interviews I tried to bear in mind that ultimately there “is no substitute for developing high quality fieldwork skills, having empathy and respect for participants, being reflective about participants’ social worlds as well as one's own, and being able to listen and understand” (Ritchie and Lewis 2003; p.276).

292 Even if people did not know exactly what the World Bank was, they would have been told that it was an international development organisation. 293 According to the guidelines and requirements of the University of New South Wales. 294 The authors also acknowledge that there has been a longstanding debate about the extent to which triangulation offers qualitative researchers a means of verifying their findings, but that the value of triangulation lies in extending our understanding, or adding breadth or depth to our analysis.

420 How stakeholders are categorised

There have been several criticisms of step 2 of stakeholder analysis theory, i.e. ‘differentiating between and categorising stakeholders’. Firstly, the grouping of people into distinct stakeholder groups may be somewhat artificial as there is likely to be a diversity of views within a stakeholder group. Grimble (1995; p.122) points out that:

One potential weakness of stakeholder analysis is that it tends to treat different stakeholder groups as distinct entities. In reality, social groupings are generally not distinct and involve overlaps between groups, e.g. rich farmers and poor farmers. Accordingly, in my analysis, I pointed out the diversity of views within stakeholder categories where relevant—for example among farmers, NGOs and government officials.

Secondly, the process of placing stakeholders on the influence-interest matrix could be problematic. As Billgrena and Holmen (2008; p.560) point out:

By categorizing stakeholders in the framework of grid/group one might assign them qualities that they might not have been aware of themselves… If used as a predetermined straightjacket, there is an obvious risk that world-views and attitudes are taken for granted and that stakeholders are assigned conceptions of nature that they do not have. Moreover the exercise may imply that stakeholder views are fixed, while in fact they often change over time and even according to the particular context, given “the multiple hats that stakeholders wear and the complex networks they belong to” (Chevalier 2001; p.7; Mayers 2005; p.13). This complication has usually been more of an issue in business management literature than in participatory natural resource management and development literature that advocates on-going and evolving involvement of stakeholders beyond stakeholder analysis (Reed et al. 2009; p.1935). In fact, while Billgrena and Holmen (2008; p.560) advocate the use of the influence-interest matrix as an analytical tool, they stress that it should be used “in a conditional, provisional sense, dependent on empirical testing”. Given these criticisms, while I placed the stakeholders on the matrix, I clarified that this was a somewhat subjective exercise and that the positions of the stakeholder groups could (and most probably would) change over time. Also, as is evident in Chapter 8—where I bring together my various analyses—the matrix is not central to how I use my stakeholder analysis to design MERs.

421 How the findings of the analysis are used

As pointed out in section 6.1, a main purpose of stakeholder analysis is to highlight the interests of marginalized groups. However, as Chevalier (2001) points out, differences in power among various stakeholder groups—made explicit through influence-interest matrices—could actually result in the opposite occurring, unless explicit participatory processes are put in place:

By itself, however, stakeholder analysis is not necessarily designed to guarantee these groups stronger representation or empowerment during the research process let alone after (through full involvement in project decision-making or conflict management activities). This is especially the case where stakeholders are ranked according to influence and importance, a strategy that can lead to stakeholder information playing into the hands of the more powerful groups and an even greater underrepresentation of lower-ranked groups (Chevalier 2001; p.6-7). Accordingly, in Chapter 8, I ensured that the views and interests of all stakeholder groups were taken into account in the design of MERs. Moreover, I explicitly included participatory processes within the suggested organisation structure and implementation plans presented at the end of that chapter.

The overall point regarding my use of stakeholder analysis is that it is just one of several tools I use to examine various aspects of my case study. Each tool I use would no doubt have weaknesses as well as strengths, but each provides insights on particular facets of the research questions I posed in Chapter 1; together they have helped me propose a MER design that is practical and equitable, which I present in Chapter 8.

422 Appendix 7: The economic viability of Managed Elephant Ranges

A7.1 Land-use options considered by governments when making forest policy

As discussed in Chapter 7, when considering forest policies, governments need to decide among three stylized options: preservation (i.e. exclude human use); conservation (i.e. permit some human use); and development (i.e. deforestation) (Ninan 2007). Details are presented in Table A45.

Table A45: Land-use options governments consider when making forest policies (adapted from Ninan 2007; p.3)

Option Description Examples

Preservation Precludes human use in a • Forests preserved in original or natural sustainable way state without any human interference (as in IUCN categories I, II)

Conservation Permits human use in a • Sustainable extraction of timber/ non- sustainable way timber forest products

Development Envisages destruction of • Permanent/ settled agriculture forests/ conversion to non- • Establishing human settlements forest uses • Setting up industries, mining, hydroelectric and other development projects

423 A7.1 Bio-prospecting potential in tropical forests

In section 7.4, I omitted discussing bio-prospecting—a direct-use value within TEV— because it is not a significant source of economic value in the landscape I am studying. However, since it could be important in tropical forests in other developing countries (or even within some other landscapes in Sri Lanka), I include a brief discussion of bio- prospecting in this Appendix.

A potentially significant value of forest biodiversity is as a source of genetic information for the development of new agricultural crop varieties, new medicines, or other industrial products and processes (Mullan and Kontoleon 2008). Tropical forests sometimes act as a source of genetic material for modern food crops, as cross-breeding with wild varieties is essential to resist diseases and pests (Pearce 1991). For example, it has been suggested that cross-breeding has already saved sugarcane, banana and cocoa crops from major damage (Pearce 1991). However, Pearce (1991) points out that there are few reliable estimates of the willingness to pay for tropical forest genetic material.

There have been more examples of bio-prospecting in the pharmaceutical industry. An often cited example is an agreement signed in 1991 between INBio, a private, non-profit, scientific organization established by the Costa Rican government, and Merck, a US multinational pharmaceutical corporation (Mullan and Kontoleon 2008). In return for an upfront payment, training assistance and a promise of royalties on future sales of products derived from Costa Rica’s forests, INBio agreed to supply Merck with samples of plants, insects and micro-organisms collected from the wild. Merck thus secured the right to use these samples to develop new pharmaceutical products. However, optimism about the potential of such markets has been dampened more recently, as additional bio- prospecting agreements have been slow to emerge. Mullan and Kontoleon (2008) suggest that this may be partly due to the slow progress of diplomatic efforts to agree an international framework for access and benefit-sharing related to genetic resources. Pearce (1991) also points out that many modern drug manufacturers tend to focus more on the production of synthetic drugs using recent advances in molecular biology and biotechnology, resulting in low willingness to pay for the retention of tropical forests as repositories of potential material for the development of pharmaceuticals.

424 A7.2 Estimates of TEV components related to tropical forests from various studies

There have been numerous studies that estimate the TEV components related to forests. Given the focus of my thesis, I have presented some developing country estimates in Tables A46 to A49, mostly adapted from Mullan and Kontoleon (2008). These provide an indication of the wide range of values that could be obtained in studies conducted in different countries (or even in different areas within the same country).

Table A46: The value of non-timber forest products (NTFPs) (from Mullan and Kontoleon 2008; p.12)295

Location Value of NTFPs (US$/ha/year) Reference(s) Gross Net Guatemala (stock of 787 (Ammour et al. 2000) goods) Sri Lanka (stock of 622 377 (Batagoda 1997) goods) Peru (stock of goods) 700 420 (Peters et al. 1989) Sri Lanka (potential 186 (Batagoda 1997) flow) Brazil (potential flow) 20 (Pinedo-Vasquez et al. 1992) Ecuador (potential flow) 200 (Myers 1988) Philippines Actual flow: 65 (Saastamoinen 1992) Potential flow: 173 India 19 – 55 (Murthy et al. 2005) India 122.5 (Mahapatra and Tewari 2005) India 65 (Verma 2000) India 117 – 144 (Chopra 1993) India 70 (Appasamy 1993) Lao PDR 6 – 8 (Rosales et al. 2005) Cambodia 19 (Bann 1997) Nepal 33 – 115 (Houghton and Mendelsohn 1996) Sri Lanka 14 (Batagoda 1997)

295 Acronyms used in Table A46 to A49 include: TC= Total Cost; WTP = Willingness to pay; CV = Contingency Valuation.

425 Location Value of NTFPs (US$/ha/year) Reference(s) Sri Lanka 13 (Gunatilake et al. 1993) Sri Lanka 2 (Batagoda 1997) Malaysia 8 (Caldecott 1988) Guatemala 30 (Ammour et al. 2000) Venezuela 15 (Melnyk and Bell 1996) Ecuador 77 – 180 (Grimes et al. 1994) Belize 41 – 188 (Balick and Mendelsohn 1992) Mexico 330 (Adger et al. 1995) Mexico 116 (Alcorn 1989) Brazil 79 (Anderson and Ioris 1992) Brazil 97 (Anderson and Ioris 1992) Venezuela 1 (Mori 1992) Peru 67 (Thorbjarnarson 1991) Peru 18 – 24 (Smith et al. 1997) Senegal 0.7 (Padoch and de Jong 1989) Cameroun 6 (Ba et al. 2006) Madagascar 4 (Yaron 2001) Kenya 88 (Kramer et al. 1995) Uganda 11 (Howard 1995) Zimbabwe 21 (Bojö 1993) Zaire 1 – 3 (Wilkie 1989) Cameroun 1 (Ruitenbeek 1988)

The estimates of NTFPs in Table A46 vary widely, from 1 to 420 US$ per ha per year296. Mullan and Kontoleon (2008)—who collated these numbers—raise some possible issues with the relevant studies. Firstly, average values may be overstated because research is more likely to be carried out in locations where NTFPs are important to nearby communities. Less accessible forests would tend to exhibit lower values, as local demand for products would be lower and harvesting costs higher. In addition, as these studies generally measure benefits based on current NTFP use, they do not consider whether current harvests are sustainable, or whether harvests could be increased with alternative management practices.

296 The figure of 420 is an estimate of a stock of NTFPs. The highest value corresponding to a flow of goods is 200.

426 Table A47: Tourism/ recreational values in tropical forests (from Mullan and Kontoleon 2008; p.45)

Location Value Value (US$/trip or Reference(s) (US$/ha/year) US$/household) Uganda (foreign $0.59 with 20 bird $46/person (Naidoo and tourists) species; $1.32 with Adamowicz 2005) 80 bird species Madagascar (foreign $10.73 (TC); $29 $27/trip (TC); $74/trip (Kramer et al. 1995) tourists) (CV; may include (CV; may include existence value) existence value) Madagascar (foreign $360 – 468 Maille and tourists) Mendelsohn (1991) Indonesia (foreign $7.11 per visitor (local van Beukering et al and local tourists) tourists); $12.4 per 2003 visitor (foreign tourists) Malaysia $3 (Bann 1997) Malaysia (foreign $740 (Garrod and Willis tourists) 1997) Thailand (foreign Tourism (Dixon and Sherman tourists) expenditure: $7 – 1990) 35.5/ha/year; consumer surplus: $2.3/ha/year India WTP (Hadker et al. 1997) $2.76/household/year Costa Rica (foreign WTP for '1 level' Bienabe and Hearne and local tourists) increase in scenic (2006) beauty: Costa Ricans - $2.93/year; Foreign tourists - $3.28 Costa Rica, Two $950 and $2,305 $11 and $13 per local Shultz, Pinazzo and forested parks (two sites). visitor, and $23 and Cifuentes (1998) (foreign and local $14 per foreign visitor tourists) Costa Rica, 3 $21 – 25 per visitor Chase et al (1998) national parks (foreign tourists) Costa Rica (foreign $160 (Tobias and tourists) Mendelsohn 1991) Costa Rica (foreign $60/visit (current fees Baldares et al (1990) and local tourists) $30/visit) Bolivia (foreign $2.4-2.8/ha/year Mean WTP: $72 (CB); Ellingson and Seidl tourists) $35 (CV) (2007)

427 Location Value Value (US$/trip or Reference(s) (US$/ha/year) US$/household) Mexico $1 (Adger et al. 1995) Brazil, Atlantic WTP for new parks: (Holmes et al. 1998) Coastal Forest $23 – $89/person WTP for protection of half of remaining forest: $9/person

The values of tourism and recreational activities in tropical forests in the studies summarised in Table A47, which range from a few US$ per hectare each year, to numbers exceeding 2,000 297 . Mullan and Kontoleon (2008)—who collated these numbers—point out that in studies estimating the values per visit for foreign and domestic tourists, recreational values are consistently higher for foreign tourists than for local visitors. This is hardly surprising, however, given the higher average incomes of foreign tourists visiting developing countries.

Table A48: Indirect use values of forests (from Mullan and Kontoleon 2008; p.49)

What is being valued Location Value (US$/ha/year) Reference(s) Cost of soil replacement Guatemala Negligible for soil loss; (Ammour et al. 2000) and preventing soil loss. $12/ha for nutrient loss; $30/ha for NTFPs and ecosystem services Sedimentation effects on Mexico Negligible (Adger et al. 1995) infrastructure Watershed protection Lao PDR $309 – 1,576/ha/year (Rosales et al. 2005) functions All ecosystem services India $4,348/ha/year Verma (2000) Shoreline protection and Malaysia Shoreline protection: Bann (1999) fisheries protection by $845/ha. Fisheries mangrove forest. protection: $526/ha Impacts of RIL on Malaysia $4/ha Shahwahid et al hydroelectricity (1997) Protection of irrigation Malaysia $15/ha for irrigation Kumari (1996) water. Indirect uses of forests Malaysia $20 – 23 million/year Bennett and Reynolds (1993)

297 Their findings related to developing countries are presented in Appendix 7.

428 What is being valued Location Value (US$/ha/year) Reference(s) Drought mitigation Indonesia WTP: $4.10/hh/year; Pattanayak and Estimated profit Kramer (2001) foregone: $6.80 /hh/year Watershed protection Philippines $223 – 455/ha/year Paris and Ruzicka benefits (1991) Fisheries protection Philippines $268/ha Hodgson and Dixon (1988) Flood protection Cameroon $0 – 24/ha Yaron (2001) Flood protection Cameroon $3/ha Ruitenbeek (1989) Watershed protection Kenya $273/ha/year Emerton (1999) Watershed protection Uganda $4.63/ha/year Howard (1995) Replacement costs of soil Turkey $46/ha Bann (1998) nutrients Watershed protection Ireland Watershed protection: - Clinch (1999) functions - water supply. $20/ha (negative Carbon sequestration. amount). Carbon sequestration: $88 per ha. Shelterbelts for crop Northern Rate of return increases Anderson (1987) protection and farm Nigeria from 5% to 13 – 17%. forestry Bee pollination for coffee Costa Rica $361/ha/year Ricketts et al (2004) production Gain in profits to rice and Eastern $3 – 35 per household Pattanayak and coffee production Indonesia Kramer (2001) Carbon sequestration Uganda $5.83/ha/year based on Howard (1995) damage costs; $6.81/ha/year based on replacement costs Carbon sequestration Costa Rica $105/ha/year Bulte et al (2002)

The indirect-use values of forests in Table A48 range from a few US$ per ha per year, to several thousand. As is evident from this table, there have been valuations of ecosystem services other than the three I discussed in detail in Chapter 7 (i.e. watershed protection, biodiversity conservation and carbon sequestration). For example, a study examining the role of wild pollinators found that the value of forest patches as habitat for bees is US$ 361/ha/year, in terms of improved yields on neighbouring coffee farms, i.e. those within 1 km (Mullan and Kontoleon 2008).

429 Table A49: Non-use values of forests (from Mullan and Kontoleon 2008; p.53)

What is being Value Value valued Location (US$/ha/year) (US$/household/year) Reference(s) Increased forest China Increase in forest and Wang et al cover and grassland cover: (2007) biodiversity Beijing - $97/year; Xi’an - $38/year; Ansai - $34/year Biodiversity: Beijing - $0.36 per additional species protected; Xi’an - $0.10; Ansai - $0.04 Protection of China Mean WTP: $1.27 Gong (2004) Guizhou snub- (rural); $3.82 (urban) nosed monkey. Maintain habitat China Value of Panda conservation in Kontoleon and of Giant Panda maintaining natural habitat: $15.40 Swanson (2003) (value to OECD Wolong Reserve: Difference in WTP for citizens) $259/ha/year natural habitat rather than pens: $6.67 Preserving forest Korea Mean WTP: Lee and Mjelde in Korean de- $15.42/person (2007) militarised zone. Value of India WTP for biodiversity Ninan and biodiversity conservation: Sathyapalan conservation to $130.5/household/year (2005) local households WTP of Sri Sri Use values: 0.5% of Gunawardena et Lankans for the Lanka income for peripheral al (1999) forest reserve villages, 0.2% for rural residents and 0.3% for urban residents. Bequest values: 0.4%, 0.1% and 0.2%. Existence values: 0.2%, 0.3% and 0.2% Value of Khao Thailand WTP for existence of (Dixon and Yai national park elephants: $7/person Sherman 1990) WTP of Australia $53/ha/year $15/person Flatley and Australian Bennett (1996) tourists for rainforest in Vanuatu. Preserve South $21.76/ha/year Turpie (2003) biodiversity Africa Existence value Global $4.6/ha/year Payment card: $31 per Kramer and of tropical year; Dichotomous Mercer (1997) rainforests for choice: $21 per year

430 What is being Value Value valued Location (US$/ha/year) (US$/household/year) Reference(s) US citizens Increased Costa WTP for ‘1 level’ Bienabe and biodiversity Rica increase in biodiversity Hearne (2006) protection protection: Costa ican residents - $3.87; Foreign tourists - $6.62 Protection of the Brazil Mean WTP for Mean WTP for Horton et al Brazilian protection of 5% protection of 5% more (2003) Amazon (WTP more of the of the Brazilian of UK and Brazilian Amazon: Amazon in the UK and Italian citizens) $43/ha/year Italy: $42/hh/year Existence value Mexico $0.03-10/ha/year (Adger et al. of Mexican 1995) forests

According to Table A49, non-use values of tropical forests vary from a few US$ per household per year to almost a hundred. Mullan and Kontoleon (2008)—who collated these numbers—point out that in the case of tropical forests, two sets of values are of interest: those expressed by local populations and those expressed by foreigners. For example, people living in countries without tropical forests nevertheless value the continued existence of these forests, expressing willingness to pay ranging from US$7- 42 per person/year for the conservation of biodiversity, or of tropical forests more generally.

A7.3 International climate change mitigation framework and REDD

It is important to understand the current international climate change framework to evaluate how developing countries—which contain most of the world’s tropical forests—could get credit (and the resulting financial support) for preserving their forests. The main instrument in this framework is the Kyoto Protocol of the United Nations Framework Convention on Climate Change (UNFCCC) which commits developed countries to legally binding targets to limit or reduce their greenhouse gas emissions, and provides mechanisms to allow emissions trading (Eliasch 2008; p.101). This is done by allowing developing countries that exceed their emission targets to off-set these

431 excess emissions by investing in emission-reducing projects in developing countries. The convention separates countries into three distinct categories (Eliasch 2008; p.104):

• Annex I: OECD298 countries and economies in transition; • Annex II: OECD countries only; and, • Non-Annex I: mostly developing countries.

The Kyoto Protocol establishes three mechanisms, which collectively constitute a prototype international emissions trading framework or carbon market, as per Table A50:

Table A50: The three Kyoto Mechanisms (adapted from Eliasch 2008; p.109)

Mechanism Explanation/ description

International • Annex I countries that exceed their emission targets are permitted Emissions Trading to buy credits from other Annex I countries that have excess Mechanism (IET) emissions rights to meet their commitment.

Joint • JI is a project-based mechanism for Annex I countries. Under JI, an implementation Annex I country may invest in an emission-reducing project or a Mechanism (JI) project that enhances removals by sinks in another Annex I country.

• The credits earned from a JI project can be counted towards meeting the Kyoto target of the investing country.

• The private sector can be authorised to take part in JI projects.

Clean Development • CDM is a project-based mechanism to assist non-Annex I countries Mechanism (CDM) to achieve sustainable development and to help Annex I countries achieve their emissions reduction and limitation commitments.

• To help meet its Kyoto target, an Annex I country may implement project activities that reduce emissions in non-Annex I countries.

• The tradable credits generated by non-forestry CDM are called Certified Emissions Reductions (CERs). CDM crediting can also take place for forestry activities.

• However, this only applies to afforestation and reforestation activities.Reduced emissions from deforestation and degradation (REDD) activities are excluded from crediting.

298 Organisation for Economic Cooperation and Development.

432 Reduced Emissions from Deforestation and Degradation (REDD)

A key point to note is that reduced emissions from deforestation and degradation (REDD) are currently not allowed under the Kyoto protocol. Venter and Koh (2012; p.137-138) discuss the reasoning behind the concept:

REDD+ is simply a method for putting a price tag on the carbon storage and sequestration services provided by forests. The basic premise of REDD+ is that, without it, a certain amount of carbon dioxide would be emitted due to the loss or degradation of forests. These expected emissions serve as a reference level to measure efforts to protect forests and reduce carbon emissions. If realized emissions are below the reference level, these reductions are considered “additional,” and a corresponding amount of carbon credits may be awarded. These authors go on to describe the more recent addition of a “plus” sign after REDD:

The “+” after REDD comes from more recent discussions that have broadened the mechanism’s scope to also recognize the carbon benefits of forest conservation, improved forest management, and the sequestration potential of afforestation and reforestation. At the same time, expectations for REDD+ to deliver on multiple environmental and societal issues have risen. These “co-benefits” might include conservation of biodiversity and improvements to rural livelihoods (Venter and Koh 2012; p.138). In order to examine the evolution of this concept as well as progress towards establishing viable financial mechanisms for its implementation, we need to consider briefly the history of international REDD negotiations.

The process of incorporating deforestation into global climate change negotiations were set in motion during the 11th conference of parties (COP) in 2005 at Montreal, when the Governments of Papua New Guinea and Costa Rica called for the inclusion of reducing emissions from deforestation and degradation in developing countries (REDD) in the Convention (Cerbu et al. 2011; p.11). REDD was formally accepted in principle at COP 13 at Bali in 2007, and by 2008, during the fourteenth COP in Poznan, its scope had further expanded to include other options and was designated as REDD+, which since has been officially defined as: “Reducing emissions from deforestation and forest degradation in developing countries… and the role of conservation, sustainable management of forests, and enhancement of forest carbon stocks in developing countries” (Agrawal et al. 2011; p.375).

433 Although agreement to move forward with REDD+ was one of the few points of consensus to emerge from the failed negotiations at COP 15 in Copenhagen in 2009, and an international agreement on the broad contours of REDD+ was reached at COP 16 in Cancun in 2010, a global unified financing mechanism seems unlikely given the overall lack of progress of climate-related talks and legislative set-backs in the several major developed countries. Agrawal et al. (2011; p.374) provide further details:

The failure of international negotiations to finalize a new climate treaty in Copenhagen in 2009, combined with blocked legislative processes in the United States, Canada, and Australia that would have created a large demand for REDD+ offsets, greatly diminished the prospects of a near- term, unified global mechanism for REDD+ financing. In spite of positive developments at the sixteenth Conference of Parties (COP) in December 2010 in Cancun, progress on REDD+ is likely to occur through complex, fragmented pathways of international assistance, bilateral and multilateral agreements, and civil society and market-based processes. The only near- term regulatory framework for REDD+ financing is the one under development between tropical states and provinces and the government of California. As noted in Chapter 7, however, despite the lack of a comprehensive, internationally agreed REDD+ financing mechanism, several international and national actors are already involved in helping developing nations implement REDD.

A7.3 Structure of economic model for Sri Lanka’s southeast landscape

In Chapter 7, I discuss the results of the economic model I developed to assess the economic value of natural resources within my chosen landscape. In this section, I present some details regarding my model.

Purpose of model

• The model was developed to achieve two main goals:

1. Estimate the current economic value of the land use within Sri Lanka’s southeast landscape (the base case), and,

2. Develop scenarios of land use within the landscape and assess the economic impact of each scenario.

434 • The model estimates annual benefits, annual costs and required investment costs for each scenario. It also calculates the net present value (NPV) for total benefits minus total costs.

Methodology

The model uses the concept of Total Economic Value (TEV) to estimate the economic value (benefits and costs) of the different land-use scenarios. The concept of TEV captures the benefit or flow of services derived from an environmental resource in economic terms. It is important to keep in mind that TEV is an anthropocentric and relative concept that can only be applied to goods or services over which individuals are willing to make trade-offs (Mullan and Kontoleon 2008). In other words it does not capture the value of services for which trade-offs are not possible, e.g. ‘intrinsic values’ such as cultural and religious values. The components of TEV are presented in Figure A5.

Economic values

Refers to relative values that can be assigned a monetary metric

Altruistic, bequest & Direct use values Indirect use values Option values existence values

Output that is Ecological functions The value of Knowledge of enjoyed directly that support and retaining future continued existence or by consumers protect economic options, either that others will enjoy activity known or benefits of forests & unknown biodiversity

e.g. Wood, e.g. Flood control, e.g. Potential bio- e.g. Contributions to recreation, NFTP, storm protection, prospecting environmental etc. pollination, climate values charities stabilisation, etc.

Figure A5: The components of Total Economic Value (TEV)

435 I did not have the data to estimate ‘option values’, so they are not included. In any case, option values are probably embedded in some of the other numbers.

Structure

• The model is constructed within excel.

- The entire model is in an excel workbook.

- One of the worksheets summarises the results. Others contain the assumptions, land-use scenarios and the calculations related to each of the TEV components.

• The model is constructed so that it is flexible. Most importantly, scenarios can be constructed by changing the values contained in the assumptions page (which is linked to other worksheets).

Scenarios considered

Three main land-use scenarios have been considered:

A. The current state of land-use within the southeast landscape (‘base case’),

B. Current land-use with enhanced nature tourism (developing the region’s nature tourism assets), and,

C. Increased agriculture and development (development that involves some environmental degradation).

Scenarios B and C are intentionally created to emphasize the dichotomy of ‘traditional development’ versus ‘environmentally friendly, nature tourism-led development’. Obviously, there could be various other scenarios in-between, e.g. one that achieves some development while reducing the impact on the environment (a ‘sustainable development’ scenario). Unfortunately, if the government continues on its current path

436 of development, it is much closer to the ‘traditional development’ scenario, so the one that I have chosen is realistic in the current context.

Overview of model

• The summary of benefits, costs and required investment for each of the scenarios are presented in the ‘results’ worksheet. This worksheet also contains the NPV calculations.

• The land use patterns in each scenario are presented in the ‘land use scenarios’ worksheet.

• The assumptions for the entire model are collected in the ‘assumptions’ worksheet. These are linked to all the other relevant tabs, so that a change here will make the relevant changes throughout the entire model.

• The remaining worksheets calculate each of the TEV components. The outputs of these tabs are linked to the ‘results’ worksheet, so a change in any of these worksheets would be automatically reflected here. Moreover, the assumptions used in each of these worksheets in linked to the ‘assumptions’ worksheet, so that changes in this worksheet would automatically be reflected in the relevant worksheets.

• The specific TEV components considered in the model (and the corresponding worksheets) are presented in Table A51:

Table A51: Structure of the model corresponding to TEV components

Component of Sub-components Relevant worksheet(s) TEV Agriculture • Rice Agriculture • Other crops: maize, green Agriculture_yields_prices gram, cowpea, ground nuts Fertiliser subsidies Direct use manioc, ginger values Non-Timber forest • None NTFPs products Nature tourism • Local tourists Nature tourism

437 Component of Sub-components Relevant worksheet(s) TEV • Foreign tourists Park visitation (mainstream) Costs to reduce human- • Foreign tourists (high-end) elephant conflict Carbon • Within parks Value of carbon sequestration • Outside parks (e.g. on Carbon sequestration Forest Department land) Indirect Watershed benefits • Value of water for irrigated Watershed benefits use values paddy and domestic water consumption • Wetlands, mangroves, corals, etc. Bequest Bequest and • None Bequest and existence existence values of values and elephants among existence urban Sri Lankans values

Specific worksheets

‘Results’ worksheet

• This worksheet summarizes the benefits, costs and required investment costs for each of the scenarios (in US$ millions).

− It aggregates the outputs of each of the TEV components from the relevant worksheets.

− These outputs are linked, so a change in any of those tabs is automatically reflected here.

• The net present values of total benefits minus total costs are calculated for each scenario (for 10, 15 and 20 years), using a discount rate of 5%.

• The percentage contributions of each of the components (for each of the three categories of benefits, costs and required investment costs) are presented on the right hand side of the sheet.

438 ‘Land-use scenarios’ worksheet

• This worksheet presents the land use patterns in each scenario considered.

• The data used is calculated from GIS data obtained from the Survey Department of Sri Lanka.

• The land-use categories are those used by the Survey Department.

• The land-use for Scenario C has been constructed by redistributing the land-use among the categories (while ensuring that the total remains the same).

‘Assumptions’ worksheet

• This worksheet presents the assumptions for the entire model in one sheet.

• These are linked to all the other relevant worksheets, so that a change here will make the relevant changes throughout the entire model.

• The ‘tabs used in’ column (column C) specifies the worksheets that each of the assumptions is linked to.

• The values included in this page are those for the current (base) scenario; an indication as to whether the assumption is likely to change according to the different scenarios is included in column titled ‘will vary according to scenario?’ (column D).

• Any further comments on the assumptions are included in the ‘notes’ column (column E).

‘Agriculture’ worksheet

• This worksheet estimates the values of outputs and costs of production of rice, chena crops, and other crops within the landscape.

439 • The rice yields for each of the two growing seasons—Yala and Maha—are the averages for the districts of Hambantota, Moneragala, and Ampara for 2010299.

• The area of land cultivated is from the ‘land use scenarios’ worksheet.

• The ‘chena crops’ chosen are based on the actual data for 2009/2010 from the Department of Census and Statistics.

− I have selected the top six crops grown in the three districts of Hambantota, Moneragala, and Ampara—together they account for 86% of the non-rice crops grown in the three districts, by extent of land cultivated.

− The proportions of cultivated for each crop (by extent) have been distributed among the available land for ‘other crops’ specified in the ‘land use scenarios’ worksheet.

• The ‘other crops’ chosen are based on the actual data for 2009/2010 from the Department of Census and Statistics.

− I have chosen the top four crops grown in the relevant District Secretariat Divisions300—together they account for 87% of the non-rice crops grown in the three districts, by extent of land cultivated.

− The proportions of cultivated for each crop (by extent) have been distributed in a similar manner to ‘chena crops’.

• The 2009 wholesale prices for each of the agricultural commodities have been used. However, since the middlemen take a cut, the farmers do not receive this amount301—therefore a percentage taken by the middlemen has been assumed (based on interviews conducted).

• The costs of production for both rice and other crops are an estimate based on interviews conducted. The amount of fertiliser subsidy per hectare is estimated from data from the Central Bank of Sri Lanka (more details are provided below).

299 These are from the Department of Census and Statistics. 300 The way the data were presented by the Department of Census and Statistics allowed a more precise calculation than for chena crops. 301 The wholesale price is the price middlemen sell the produce to retailers in cities and towns.

440 • The revenues and costs for the various scenarios are calculated on the right hand side of the worksheet, based on deviations from the base case scenario.

• The two worksheets ‘Agriculture_yields_prices’ and ‘Fertiliser subsidies’ contain the backup data and calculations that feed into the ‘Agriculture’ worksheet.

• The ‘Fertiliser subsidies’ worksheet estimates the amount of fertiliser subsidies given for each acre of paddy. It starts from the entire amount for the years 2008 to 2010 provided in the government budget (from information from the Central Bank of Sri Lanka), estimates the amount spent on paddy, and divides this by the total number of hectares of paddy cultivation for the respective years.

‘NTFPs’ worksheet

• This worksheet calculates the value of the non-timber forest products (NTFPs) gathered by villagers from forested land (mostly outside national parks, e.g. on Forest Department land).

• Due to the lack of data, the estimates of NTFP value (US$/ha/ year) are very rough. They are based on averages from studies done elsewhere in the country.

• The revenues and costs for the various scenarios are calculated on the right hand side of the worksheet, based on deviations from the base case scenario. The main variable here is the available forested land outside national parks—this is less in Scenario C.

‘Nature tourism’ worksheet

• This worksheet calculates potential revenues and costs associated with nature tourism.

• The base case includes the receipts from local and foreign tourists for the PAs within the landscape. The data were obtained from the Department of Wildlife.

441 • The daily expenditure of foreign tourists (divided into package and non- package) has been calculated in a previous World Bank study. This expenditure is assumed to take place over the days each of the foreign tourists spends in the national parks. The World Bank study also estimates a multiplier that gives the effects of this expenditure on the broader economy (hence the daily expenditure is adjusted by this multiplier).

• This worksheet also calculates the potential for establishing high-end nature tourism both within and outside parks (on Forest Department land). This provides inputs for Scenario B.

• The section on the bottom left calculates the costs associated with managing nature tourism.

− For the base scenario, this includes the management costs of parks as well as the costs to the local people from animals (e.g. crop raiding by elephants). The costs to reduce human-elephant conflict are actual data from the Department of Wildlife.

− The costs of enhancing the parks for high-end tourism are estimated as input to Scenario B.

• The revenues and costs for the various scenarios are calculated on the right hand side of the worksheet, based on deviations from the base case scenario.

• The two worksheets ‘Park visitation’ and ‘Costs to reduce HEC’ contain the backup data and calculations that feed into the ‘Nature tourism’ worksheet.

‘Value of carbon’ worksheet

• This worksheet calculates the potential value of carbon sequestered both within PAs as well as in forested areas outside parks. The values of carbon stocks and annual flows are estimated.

• Both carbon stocks and flows, within PAs and outside them, are calculated at three price points: 15, 20 and 30 US$ per ton of carbon sequestered

442 (corresponding to low, medium and high price points). The values selected for the final calculation are those obtained using the medium price.

• It is important to keep in mind, however, that the international process for awarding carbon credits for avoided deforestation (i.e. this type of case) is still being worked out. As pointed out in Chapter 7, although there is no international regime for awarding credits, there have been some bilateral efforts, notably between Norway and countries such as Indonesia and Brazil. However, although the Sri Lankan government has been awarded ‘observer status’ by UN-REDD, there have been no concrete discussions regarding payments for carbon sequestered (partly because the country still has considerable work to do to establish a baseline). Therefore, these credits cannot be claimed immediately.

• The average tons sequestered per hectare for different types of vegetation are based on journal articles and reports on Sri Lanka’s REDD potential—Mattsson et al. (2012) and Chokkalingam and Vanniarachchy (2011), for example.

• The main variable among scenarios is the available forested land outside national parks (the land within parks is assumed to be fixed across scenarios).

• The worksheet ‘Carbon sequestration’ contains the backup data and calculations that feed into the ‘Value of carbon’ worksheet.

‘Watershed benefits’ worksheet

• This worksheet calculates the watershed benefits within the landscape which include (a) irrigated paddy and water consumption and (b) wetlands, mangroves, corals, etc. These are based on the major river basins within the landscape.

• The analysis is based on calculations done in a World Bank paper—(World Bank 2009)—which estimated the value of environmental services in the southern province.

• The revenues and costs for the various scenarios are calculated on the right hand side of the worksheet. In Scenario C—which involves a high level of

443 development with few environmental safeguards)—it is assumed that some of the watershed benefits will be eroded.

‘Bequest and existence values’ worksheet

• This worksheet calculates bequest and existence values of elephants among urban Sri Lankans.

• This is based on an estimate by Bandara and Tisdell (2004). Since that study evaluated the value placed on Sri Lanka’s entire elephant population, the figure was adjusted by the proportion of total elephants present in the southeast landscape.

• It is important to keep in mind that these calculations relate only to the value placed on elephants by urban Sri Lankans. Therefore, presumably, the value they place on all the parks and wildlife within the landscape would be much higher, i.e. the calculation is a significant underestimate.

444 Appendix 8: Implementing Managed Elephant Ranges

A8.1 Electric fence design

As noted in Chapter 8, proper electric fence design is critical to ensure that fences are effective in keeping elephants away from human property. Gunaratne and Premarathne (2006) describe how some of the main components of fences—such as the energiser, the solar power system, fence wires and fence posts—should be designed, based a detailed study of fences in several parts of Sri Lanka. Some of their main findings are summarised in Table A52 below.

Table A52: Specifications of well-constructed fences in the Sri Lankan context (adapted from Gunaratne and Premarathne 2006)

Components Use and specifications Energizer The energizer is used for providing energy to the fence wires. A standard energizer can be powered by 230 volts AC302 as well as 12 Volts DC303 batteries. It should incorporate internal adaptive control to maintain effective high-energy different pulses at accepted levels. The energizer unit must incorporate advanced lightening protection. Solar Power System A solar array of 12 volts with 7.5 amperes peak current is required. Power is generated by solar panels. The Solar Controller in the Solar Power System should be capable of handling 10 amperes at 12 volts and should be able to protect the load against short current conditions. Fence wires Galvanized fence wires should be made out of hot dipped galvanized high tensile steel. Minimum zinc coating shall be 260g/m2. Concrete posts The concrete posts used for fencing should be reinforced and made through a process of mechanical vibration so that they are strong enough not to fall over if pushed by an elephant. The distance between two adjacent posts is 20 m. Connecting components These include wire joining clamps, permanent tension springs, wire

302 Alternating current. 303 Direct current.

445 Components Use and specifications tighteners, load out cable, lightning diverters (adjustable), live insulators with anchor clips and porcelain strain insulators.

A8.2 Sustainable agriculture methods

As described in Chapter 8, when implanting MERs, lessons should be drawn from the extensive body of research conducted in many parts of the world on sustainable agriculture. Gomiero et al. (2011; p.15-18) describe a number of different approaches to agriculture management which have been proposed and implemented in this regard. Some of the approaches that may be relevant to the southeast landscape are summarised in Table A53.

Table A53: Potential approaches to achieve sustainable agriculture (adapted from Gomiero et al. 2011; p.15-18)

Approach Description/ details Agro-ecology Agro-ecosystems are communities of plants and animals interacting with their physical and chemical environments that have been modified by people to produce food, fibre, fuel and other products for human consumption and processing. Agro-ecology is the holistic study of agro-ecosystems including all the environmental and human elements. It focuses on the form, dynamics and functions of their interrelationship and the processes in which they are involved. In order to properly study agro-ecosystem functioning and management, integrated scale analysis has to be performed along the multiple scales and dimensions of agro-systems.

446 Approach Description/ details Integrated agriculture Integrated agriculture is a farming method that combines management practices from conventional and organic agriculture. For example, animal manure may be used instead of chemical fertiliser when possible. Integrated pest management is conducted by combining several methods such as using crop rotation, releasing parasitoids, and cultivating pest-resistant varieties; pesticides are used only as the last resort. Weeds could be managed through tillage and cultivation practices, using competitive cultivars, crop diversification; other factors could be used to reduce weed germination, growth, competitive ability, reproduction, and dispersal. Integrated agriculture is not governed by specific regulations but its goal is still to reduce as much as possible both farm management costs and its environmental impact, aiming at the long term sustainability of farming practices. Organic agriculture Although sustainable agriculture practices are adopted by an increasing number of farmers only organic agriculture is regulated by laws and needs to strictly follow a specific set of norms. Such norms, for example, forbid the use of agrochemicals and strictly regulate the use of drugs in animal rearing; they also forbid the use of genetically modified organisms. Organic agriculture has been officially recognized by the European Union in 1991 and by the US federal government in 1995. Organic agriculture aims at preserving soil fertility, reducing soil erosion, and conserving water, biodiversity, the landscape and ecological functionality.

447 Approach Description/ details Transgenic Technology Technological advancements in the field of genetics have made it possible to manipulate gene expression and operate gene transfers from an organism to another. Crops, could, for instance, be engineered to resist pests, improve water use efficiency, cope with drought or salty soil, self-fix nitrogen, or to produce important nutritional elements. According to the International Service for the Acquisition of Agri- biotech Applications, the use of plant transgenics is the fastest adopted crop technology—in 2009 there were 134 million ha of biotech crops, with an underlying 80-fold land increase from 1996 to 2009 and a year-to-year growth of 9 million ha or 7% on average. Developing countries have increased their share of global biotech crops to almost 50% and are expected to continue to significantly increase biotech hectarage in the future. While GMOs304 such as Bt corn and cotton, and herbicide-tolerant soybean have been cultivated in USA since 1990s, most of the European countries are still against their approval for cultivation. In Europe the environmental release of GMOs generated an intense social and political debate concerning the environment and food safety and the ethical acceptability of engineered crops. Consequently, some stakeholders have adopted a precautionary approach to their release. Environmental risks related to GMO crops are based on two main issues: (1) the potential effects of gene flow to non-target organisms, and (2) the probability of gene flow to wild plant relatives. The latter issue could result in the development of resistance in weeds and pests, as has happened with agrochemicals.

When we consider the various options presented in Table A53, it is evident that some techniques are probably too complicated and/or technically and resource-intensive to adopt in the context of MERs. For example, significant effort would be needed to gather the information needed for agro-ecology. Transgenic technology would require significant capital investment, or would require the purchase of genetically modified

304 Genetically Modified Organisms.

448 crops from global multinational firms, which in turn would need to adapt some existing varieties to the Sri Lankan context, making this a very costly option. Moreover, as pointed out in the table, this is a still a somewhat controversial method. The specific requirements associated with organic agriculture may prove to be too onerous to chena farmers. Therefore, it seems that integrated agriculture, which combines management practices from conventional and organic agriculture, could be the most feasible option at present.

A8.3 Tourism survey conducted by the World Bank

An extensive survey of nearly 2,000 tourists was conducted by the World Bank in 2008 (World Bank 2009; p.10). The targeted population included resident and non-resident tourists staying in the hotels and lodges along the southwest coastline of Sri Lanka and near the parks (Bundala, Minneriya, Singharaja Forest Reserve, Uda Walawe, and Yala). The survey was designed to elicit information about the tourist profiles, trip characteristics, satisfaction levels, and the willingness to pay for park-related activities. The interviews were typically conducted following a respondent’s trip into the park to ensure an informed response.

The survey methodology is described as follows (World Bank 2009; p.6 of Annexes):

Although the overall sample stratification was pre-set to be representative of the current population of tourists, the selection of tourists within each category was based on a random interception procedure. In the case of hotels or lodges, permission was obtained from the hotel manager to solicit his/her guests and also to make them aware of the purpose of this survey. Enumerators would then interview every ith tourist out of a total number that would be required per day to achieve a representative sample for the location. In the case of interception at or near the parks, tourists were contacted after they had visited the park, normally at the park gate or in hotels lodges or even restaurants near the site. There were also instances where large groups on tour buses would arrive at the park. In this case, an agreement with the driver was made that upon exiting the park, respondents would be asked whether they would answer a few questions about their experience in the park. In other instances, respondents were traveling individually and not part of a group so interception was not an issue. I was provided access to the raw data from the survey, which I analysed using excel and SPSS. Table A54 provides information of the various categories of tourists—both local and international—who were interviewed.

449 Table A54: Distribution of tourists, by type and location, in the World Bank tourism survey (adapted from World Bank 2009; p.10-11)

Tourist type Number of respondents % International non-park (hotels and lodges) 998 50.0 International non-park (hotels and lodges) 600 30.1 International park 198 9.9 Bundala National Park 43 Minneriya National Park 37 Sinharaja Forest Reserve 42 Uda Walawe National Park 42 Yala National Park 33 Local park 200 10.0 Bundala National Park 40 Minneriya National Park 40 Sinharaja Forest Reserve 40 Uda Walawe National Park 40 Yala National Park 40 Total 1,996 100

As evident from Table A54, the survey included international tourists, who visited PAs and those who did not, as well as local tourists in the same two categories. In order to obtain the feedback from foreign tourists on parks305, I focused on the 198 international tourists who visited parks. Moreover, I only analysed the information related to the three PAs within the southeast landscape—Bundala, Uda Walawe and Yala National Parks. This resulted in a sample size of 118 people.

The detailed information for the relevant variables for each of these three parks is summarised in Tables A55 to A63.

305 This is the group I was especially interested in, since I am looking at the potential for high-end tourism within proposed MERs, and therefore trying to understand the feedback from this customer segment.

450 Table A55: Wildlife species diversity – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 11 6 8 25 21% Bad 8 7 11 26 22% Neutral 12 2 7 21 18% Quite good 7 23 11 41 35% Very good 2 1 2 5 4% Total 40 39 39 118 100%

Table A56: Congestion – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 7 22 12 41 34% Bad 6 10 10 26 22% Neutral 13 5 12 30 25% Quite good 12 3 4 19 16% Very good 2 0 1 3 3% Total 40 40 39 119 100%

Table A57: Knowledge of guide – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 11 10 15 36 31% Bad 8 2 8 18 15% Neutral 5 23 6 34 29% Quite good 8 2 4 14 12% Very good 8 3 5 16 14% Total 40 40 38 118 100%

451 Table A58: Accommodation – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 11 9 11 31 26% Bad 13 7 10 30 25% Neutral 5 17 7 29 24% Quite good 9 6 7 22 18% Very good 2 1 4 7 6% Total 40 40 39 119 100%

Table A59: Diversity of activities – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 13 18 11 42 35% Bad 9 14 13 36 30% Neutral 10 8 7 25 21% Quite good 5 0 5 10 8% Very good 3 0 3 6 5% Total 40 40 39 119 100%

Table A60: Friendliness of staff – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 13 10 7 30 25% Bad 7 4 11 22 18% Neutral 5 9 8 22 18% Quite good 12 17 10 39 33% Very good 3 0 3 6 5% Total 40 40 39 119 100%

452 Table A61: Transport experience – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 9 14 9 32 27% Bad 10 6 9 25 21% Neutral 4 8 10 22 18% Quite good 8 4 6 18 15% Very good 9 8 5 22 18% Total 40 40 39 119 100%

Table A62: Value for money – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 8 10 9 27 23% Bad 10 14 14 38 32% Neutral 9 7 4 20 17% Quite good 5 9 9 23 19% Very good 7 0 3 10 8% Total 39 40 39 118 100%

Table A63: Overall experience – results for Bundala, Uda Walawe and Yala National Parks, numbers of responses (analysis of data from World Bank 2009)

Response Bundala Uda Yala Total % of total Walawe Very bad 9 20 7 36 30% Bad 5 4 10 19 16% Neutral 8 6 9 23 19% Quite good 8 6 10 24 20% Very good 9 4 3 16 13% Total 39 40 39 118 99%

453 A8.3 Results of study commissioned by the Sri Lanka Tourist Board on high-end eco tourists

As mentioned in Chapter 8, the Sri Lanka Tourist Board—in the process of developing an ecotourism development strategy for Sri Lanka—commissioned a study to identify some of the characteristics of high-end eco tourists. The report includes results from a comprehensive survey that was conducted in important ecotourism markets from the Sri Lankan perspective such as US, UK and Australia (Sri Lanka Tourist Board 2003; p.8- 11). The preferences indicated in this survey are summarised in Table A64 below.

Table A64: Preferences of eco tourists as revealed by survey commissioned by the Sri Lanka Tourist Board (Sri Lanka Tourist Board 2003; p.8-11)

Category Indicated preferences

Motivating factors • Physical health and well-being for trip • Discovery, education and intellectual stimulation

• Enjoyment of unspoilt nature and beautiful scenery

• Contact with people from different backgrounds and cultures in a more personalised intimate setting than in mass tourism destinations

• Adventure and excitement

Most popular • Visiting national parks and PAs activities • Hiking/ trekking

• Water-based activities (especially rafting)

• Camping

• Learning about other cultures

Accommodation • Eco tourists tend to prefer other accommodation than conventional hotels, for example that which is more rustic and intimate such as tents, cabins and bungalows

Food preferences • Food is often an important factor for eco tourists, and many are particularly interested in sampling local products and recipes.

• There is increasing attention paid to the origin of food (local specialties) and its quality, i.e. home grown rather than mass- produced

454 References

Adams, W. M., R. Aveling, D. Brockington, B. Dickson, J. Elliott, J. Hutton, D. Roe, B. Vira and W. Wolmer (2004). 'Biodiversity conservation and the eradication of poverty'. Science 306(5699): 1146-1149. ADB (2004). 'Review of Governance and Public Management for Sri Lanka'. Manila, Asian Development Bank. Adger, W. N., K. Brown, R. Cervigni and D. Moran (1995). 'Total Economic Value of Forests in Mexico'. Ambio 24(5): 286 - 296. AECEN (2006). 'Environmental Compliance and Enforcement in Sri Lanka: Rapid Assessment'. Bangkok, Thailand, Asian Environmental Compliance and Enforcement Network. African Parks Network. (2011). 'The African Parks Network website.' Retrieved 31 July, 2011, from http://www.african-parks.org/. Agalawatte, M. B. and P. Abeygunawardena (1994). 'Conservation Farming as an Alternative to Shifting Cultivation in Sri Lanka'. Journal of Sustainable Agriculture 4(2): 65-79. Agarwal, A. and K. Redford (2009). 'Conservation and displacement: an overview'. Conservation and Society 7(1): 1-10. Agrawal, A., D. Nepstad and A. Chhatre (2011). 'Reducing Emissions from Deforestation and Forest Degradation'. Annual Review of Environment and Resources 36: 373–396. Alcorn, J. (1989). An economic analysis of Huastec Mayan Forest Management. Fragile land of Latin America: Strategies for Sustainable Development. J. Brouwer. Boulder, Westview Press: 182–206. Alter, S. (2004). Elephas maximus : a portrait of the Indian elephant. Orlando, Harcourt. Ammour, T., N. Windevoxhel and G. Sencion (2000). Economic valuation of mangrove ecosystems and subtropical forests in Central America. Sustainable Forest Management and Global Climate Change. M. H. I. Dore and R. Guevara. Cheltenham, Edward Elgar: 166-197.

455 Anderson, A. B. and E. M. Ioris (1992). 'Valuing the rain forest: Economic strategies by small-scale forest extractivists in the Amazon estuary'. Human Ecology (Historical Archive) 20(3): 337-336. Anke, H., G. C. Zeri, J. Dietz, A. Freibauer, M. Hüttner, M. Jung, D. Mollicone and M. Scheffler (2008). 'Emissions and removals from land-use, land use change and forestry activities in a post-Kyoto regime – quantitative analysis of a framework for reducing deforestation'. Berlin, Germany, German Federal Environmental Agency. Anonymous (2010). 'India Gives Elephants 'National Heritage Animal' Status', Environmental news service Appasamy, P. P. (1993). 'Role of non-timber forest products in a subsistence economy: the case of a joint forestry project in India'. Economic Botany 47(3): 258-267. Ariyadasa, K. K. (2012). 'Rains ruin the chena cultivator'. Sunday Times Sri Lanka. Colombo. January 1 issue. Aung, U. T. and U. T. Nyunt (2001). The care and management of domesticated Asian elephants in Myanmar. International workshop on the domesticated Asian elephant, Bangkok, Food and Agriculture Organisation of the United Nations. Ba, C. A., J. Bishop, M. Deme, H. D. Diadhiou, A. B. Dieng, O. Diop, P. A. Garzon, B. Gueye, M. Kebe, O. K. Ly, V. Ndiaye, C. M. Ndione, A. Sene, D. Thiam and I. A. Wade (2006). The Economic Value of Wild Resources in Senegal: A Preliminary Evaluation of Non-Timber Forest Products, Game and Freshwater Fisheries. Gland, Switzerland, The World Conservation Union (IUCN). Balick, M. J. and R. Mendelsohn (1992). 'Assessing the Economic Value of Traditional Medicines from Tropical Rain Forests'. Conservation Biology 6(1): 128-130. Balint, P. J. and J. Mashinya (2008). 'CAMPFIRE During Zimbabwe's national crisis: local impacts and broader Implications for Community-Based Wildlife Management'. Society & Natural Resources 21(9): 783-796. Balmford, A., B. Aaron, P. Cooper, R. Costanza, S. Farber, R. E. Green, M. Jenkins, P. Jefferiss, V. Jessamy, J. Madden, K. Munro, N. Myers, S. Naeem, J. Paavola, M. Rayment, S. Rosendo, J. Roughgarden, K. Trumper and R. K. Turner (2002). 'Economic Reasons for Conserving Wild Nature'. Science 297(5583): 950-953.

456 Bandara, A. (2010). 'Hakka patas inserted in vegetables brings slow, painful death to elephant'. The Sunday Times Sri Lanka. Colombo, Sri Lanka. Bandara, H. (2011). 'Produce from the North creates surplus in vegetables'. The Sunday Times Sri Lanka. Colombo, Sri Lanka. March 12, Issue. Bandara, J. M. R. S. (2007). Nature Farming; Integration of traditional knowledge systems with modern farming in rice. Leusden, The Netherlands, Compas. Bandara, J. M. R. S., D. M. A. N. Senevirathna, D. M. R. S. B. Dasanayake, V. Herath, J. M. R. P. Bandara and T. Abeysekara (2008). 'Chronic renal failure among farm families in cascade irrigation systems in Sri Lanka associated with elevated dietary cadmium levels in rice and freshwater fish'. Environmental Geochemistry and Health 30: 465–478. Bandara, J. M. R. S., H. V. P. Wijewardena, Y. M. A. Y. Bandara, R. G. P. T. Jayasooriya and H. Rajapaksha (2011). 'Pollution of River Mahaweli and farmlands under irrigation by cadmium from agricultural inputs leading to a chronic renal failure epidemic among farmers in NCP, Sri Lanka'. Environmental Geochemistry and Health 33: 439–453. Bandara, R. (2004). 'Economic value of conservation: the case of the Asian elephant'. South Asia Economic Journal(5): 283-300. Bandara, R. and C. Tisdell (2002). 'Asian elephants as agricultural pests: Economics of control and compensation in Sri Lanka'. Natural Resources Journal 42(3): 491- 519. Bandara, R. and C. Tisdell (2002). Conserving Asian elephants: economic issues illustrated by Sri Lankan concerns. The economics of conserving wildlife and natural areas. C. Tisdell. Massachusetts, USA, Edward Elgar Publishing: 193- 211. Bandara, R. and C. Tisdell (2003). 'Comparison of rural and urban attitudes to the conservation of Asian elephants in Sri Lanka: empirical evidence'. Biological Conservation 110(3): 327-342. Bandara, R. and C. Tisdell (2004). 'The net benefit of saving the Asian elephant: a policy and contingent valuation study'. Ecological Economics 48(1): 93-107.

457 Bandara, R. and C. Tisdell (2005). 'The history and value of the elephant in Sri Lankan society'. Working Papers in Economics, Ecology and the Environment. Brisbane, University of Queensland. Bandarage, A. (1983). Colonialism in Sri Lanka : the political economy of the Kandyan Highlands, 1833-1886. Berlin ; New York, Mouton. Bandaratillake, H. M. and M. P. S. Fernando (2003). National Forest Policy Review, Sri Lanka. An overview of Forest Policies in Asia. T. Enters, M. Qiang and R. N. Leslie. Bangkok, Thailand, Food and Agriculture Organisation: 250 - 292. Bann, C. (1997). 'An Economic Analysis of Tropical Forest Land Use Options, Ratanakiri Province, Cambodia'. Ottawa, Canada, International Development Research Centre, Economy and Environment Program for Southeast Asia. Barber, C. V., K. R. Miller and M. Boness, Eds. (2004). Securing Protected Areas in the Face of Global Change: Issues and Strategies, IUCN, Gland, Switzerland and Cambridge, UK. Barnes, J. and B. Jones (2009). Game ranching in Nambia. Evolution and innovation in wildlife conservation: parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Barua, M. (2010). 'Whose Issue? Representations of Human-Elephant Conflict in Indian and International Media'. Science Communication 32(1): 55 - 75. Bastian, S. (2003). 'The politics of land reform and land settlment in Sri Lanka'. Colombo, International Centre for Ethnic Studies. Batagoda, B. M. S. (1997). Valuing Rainforests: A Botanical and Ethnobotanical Study of Non-timber Forest Products in the Sinharaja Forest of Sri Lanka, University of East Anglia. Batagoda, B. M. S. (2003). 'The economic valuation of alternative uses of mangrove forests in Sri Lanka'. Netherlands, UNEP. BBC. (2012, May 20, 2012). 'Sri Lanka profile.' Retrieved 31/08/2012. BECT. (2012). 'Biodiversity and Elephant Conservation Trust website.' Retrieved September 3, 2012, from http://elephantsinsrilanka.org/index.htm.

458 Bengtsson, J., P. Angelstam, T. Elmqvist, U. Emanuelsson, C. Folke, M. Ihse, F. Moberg and M. Nystrom (2003). 'Reserves, resilience and dynamic landscapes'. Ambio 32(6): 389-396. Berkes, F. (2004). 'Rethinking community-based conservation'. Conservation Biology 18(3): 621-630. Bhattacharjya, S. (2010). 'Farmers helpless in the face of elephantine fury'. The Sunday Times Sri Lanka. Colombo. Billgrena, C. and H. Holmen (2008). 'Approaching reality: Comparing stakeholder analysis and cultural theory in the context of natural resource management'. Land Use Policy 25: 550–562. Biodiversity Indicators Partnership (2010). 'Official Development Assistance provided in support of the Convention on Biological Diversity', Biodiversity Indicators Partnership,. Birner, R. and H. Gunaweera (2001). 'Between market failure, policy failure and 'community failure': property rights, crop/livestock conflicts and the adoption of sustainable land use practices in the dry zone of Sri Lanka'. Washington, D.C., International Food Policy Research Institute. Bojö, J. (1993). Economic valuation of indigenous woodlands in Zimbabwe. Living with Trees: Policies for Forestry Management in Zimbabwe. Washington D C, World Bank. Bond, I., B. Child, D. d. l. Harpe, B. Jones, J. Barnes and H. Anderson (2004). Private land contribution to conservation in South Africa. Parks in transition : biodiversity, rural development, and the bottom line. B. Child. London; Sterling, VA, Earthscan Publications: xiv, 267 p. Borrini-Feyerabend, G., A. Kothari and G. Oviedo (2004). Indigenous and Local Communities and Protected Areas: Towards Equity and Enhanced Conservation, IUCN, Gland, Switzerland and Cambridge, UK. Bothma, J., H. Suich and A. Spencely (2009). Extensive wildlife production on private land in South Africa. Evolution and innovation in wildlife conservation : parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p.

459 Bradshaw, C. J. A., N. S. Shodhi and B. W. Brook (2009). 'Tropical turmoil: a biodiversity tragedy in progress'. Frontiers in Ecology and the Environment 7(2): 79-87. Brewer, G. (1999). 'The challenges of interdisciplinarity'. Policy Sciences 32: 327 - 337. Brockington, D., R. Duffy and J. Igoe (2008). Nature unbound: conservation, capitalism and the future of protected areas. London; Sterling, VA, Earthscan. Brown, K. (2004). Trade-off analysis for integrated conservation and development. Getting biodiversity projects to work : towards more effective conservation and development. T. O. McShane and M. P. Wells. New York, Columbia University Press: 232-255. Bruun, T. B., A. d. Neergaard, D. Lawrence and A. D. Ziegler (2009). 'Environmental Consequences of the Demise in Swidden Cultivation in Southeast Asia: Carbon Storage and Soil Quality'. Human Ecology 37: 375–388. Caldecott, J. (1988). Hunting and Wildlife Management in Sarawak. Gland, Switzerland and Cambridge, UK, IUCN. Cameron, J. (2005). Focusing on the focus group. Qualitative research methods in human geography. I. Hay. South Melbourne, Vic. New York, Oxford University Press: xxvii, 339 p. Campos-Arceiz, A., S. Takatsuki, S. K. K. Ekanayaka and T. Hasegawa (2009). 'The Human-Elephant conflict in southeastern Sri Lanka: type of damage, seasonal patterns, and sexual differences in the raiding behavior of elephants'. Gajah: Journal of the IUCN/SSC Asian Elephant Specialist Group (31): 5-14. Canadell, J. G. and M. R. Raupach (2008). 'Managing forests for climate change mitigation'. Science 320(5882): 1456-1457. Cannon, T. and P. Davis (1995). Aliya: stories of the elephants of Sri Lanka. Melbourne, Australia; Nevada City, California, USA, Airavata Press; Blue Dolphin Publications. Carrington, R. (1958). Elephants: a short account of their natural history evolution and influence on mankind. London,, Chatto & Windus. Castellanet, C. and C. F. Jordan (2002). Participatory Action Research in Natural Resource Management. New York, USA, Taylor & Francis.

460 Castley, G., M. Knight and J. Gordon (2009). Making conservation work: innovative approaches to meeting biodiversity conservation and socio-economic objectives (an example from the Addo Elephant National Park, South Africa). Evolution and innovation in wildlife conservation : parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. CCR. (2009). 'Centre for Conservation and Research website.' Retrieved September 3, 2012, from http://www.ccrsl.org/. Central Bank of Sri Lanka (2011). 'Central Bank of Sri Lanka Annual Report - 2010'. Colombo, Sri Lanka, Central Bank of Sri Lanka. Cerbu, G. A., B. M. Swallow and D. Y. Thompson (2011). 'Locating REDD: A global survey and analysis of REDD readiness and demonstration activities'. Environmental Science & Policy 14: 168-180. Chape, S., J. Harrison, M. Spalding and I. Lysenko (2005). 'Measuring the extent and effectiveness of protected areas as an indicator for meeting global biodiversity targets'. Philosophical Transactions of the Royal Society B-Biological Sciences 360(1454): 443-455. Chazdon, R. L., C. A. Harvey, O. Komar, D. M. Griffith, B. G. Ferguson, M. Martínez- Ramos, H. Morales, R. Nigh, L. Soto-Pinto, M. v. Breugel and S. M. Philpott (2009). 'Beyond Reserves: A Research Agenda for Conserving Biodiversity in Human-modified Tropical Landscapes'. Biotropica 41(2): 142-153. Chee, Y. E. (2004). 'An ecological perspective on the valuation of ecosystem services'. Biological Conservation 120: 549–565. Chevalier, J. (2001). 'Stakeholder Analysis and Natural Resource Management'. Ottawa, Canada, Carleton University: 14. Child, B. (2009). Community conservation in Southern Africa: rights-based natural resource management. Evolution and innovation in wildlife conservation: parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Child, B. (2009). The emergence of parks and conservation narratives in Southern Africa. Evolution and innovation in wildlife conservation: parks and game

461 ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Child, B., S. McKean, A. Kiss, S. Munthali, B. Jones, M. Mtsambiwa, G. Castley, C. Patton, H. Magome, G. Pangeti, P. Fearnhead, S. Johnson and G. Chilikusha (2004). Park agencies, performance and society in South Africa. Parks in transition: biodiversity, rural development, and the bottom line. B. Child. London; Sterling, VA, Earthscan Publications: xiv, 267 p. Child, G. (2009). The emergence of modern nature conservation in Zimbabwe. Evolution and innovation in wildlife conservation: parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Chokkalingam, U. and S. A. Vanniarachchy (2011). 'Sri Lanka’s REDD+ Potential: Myth or Reality? Forest Carbon Asia Country Profile Report No. 1: Sri Lanka'. http://www.forestcarbonasia.org. Chopra, K. (1993). 'The value of non-timber forest products: an estimation for tropical deciduous forests in India'. Economic Botany 47(3): 251-257. Cincotta, R. P., J. Wisnewski and R. Engelman (2000). 'Human population in the biodiversity hotspots'. Nature 404(6781): 990-992. Clark, T. W. (2002). The policy process: a practical guide for natural resources professionals. New Haven, Yale University Press. Constanza, R., R. d’Arge, R. de Groots, S. Farber, M. Grasso, B. Hannon, K. Limburg, S. Naeem, R. O’Neill, J. Paruelo, R. Raskin, P. Sutton and M. van den Belt (1997). 'The value of the world’s ecosystem services and natural capital'. Nature 387: 253–260. Cooney, R. (2007). 'Sustainable use: concepts, ambiguities, challenges', IUCN Species Survival Commission. Corea, R. (2007). 'Saving Elephants by Helping People: mitigating human-elephant conflicts along the southern boundary of the Lahugala Kitulala National Park for the long-term conservation of the Sri Lankan elephant'. Colombo, Sri Lanka, Sri Lanka Wildlife and Conservation Society. Coria, J. and E. Calfucura (2012). 'Ecotourism and the development of indigenous communities: The good, the bad, and the ugly'. Ecological Economics 73: 47–55.

462 Dalal-Clayton, D. B. and D. Dent (2001). Knowledge of the land: land resources information and its use in rural development. Oxford, England, Oxford University Press. Dawn newspaper (2010). 'Sri Lanka seeks harmony between elephants and humans'. Dawn. Karachi. May 17 issue. de Costa, W. A. J. M. (2008). 'Climate change in Sri Lanka: myth or reality? Evidence from long-term meteorological data'. Journal of the National Science Foundation of Sri Lanka 36 (Special Issue). de Lopez, T. T. (2001). 'Stakeholder Management for Conservation Projects: A Case Study of Ream National Park, Cambodia'. Environmental Management 28(1): 47–60. de Silva, C. R. (2011). A hydraulic civilization. The Sri Lanka Reader: History, Culture, Politics. J. C. Holt. Durham, North Carolina, Duke University Press: 52 - 63. de Silva, K. M. (1981). A history of Sri Lanka. London, Berkeley, C. Hurst; University of California Press. de Silva, M. and P. K. de Silva (2007). The Sri Lankan elephant: its evolution, ecology & conservation. Colombo, WHT Publications (private) Ltd. de Silva, T. (2009). 'A destiny fulfilled'. The Nation Newspaper Sri Lanka. Colombo. April 12 issue. de Silva, T. (2009). 'Kavantissa: Master of strategy'. The Nation Newspaper Sri Lanka. Colombo. March 15 issue. de Soto, H. (2001). 'Dead Capital and the Poor'. SAIS review 21(1): 13-43. DeFries, R., A. Hansen, A. C. Newton and M. C. Hansen (2005). 'Increasing Isolation of Protected Areas in Tropical Forests over the past Twenty Years'. Ecological Applications 15(1): 19-26. Dent, D. L. and L. K. P. A. Goonewardene (1993). 'Resource assessment and land use planning in Sri Lanka: a case study'. London, The International Institute for Environment and Development. Department of Census and Statistics (2011). Household Income and Expenditure Survey: 2009/10. Colombo, Government of Sri Lanka.

463 Department of Census and Statistics (2011). 'Sri Lanka Labour Force Survey Annual Report - 2010'. Colombo, Government of Sri Lanka. Department of Census and Statistics (2011). Statistical Abstract 2010. Colombo, Government of Sri Lanka. Department of Census and Statistics (2011). Statistical Pocket Book 2010. Colombo, Government of Sri Lanka. Department of Census and Statistics (2011). 'The website of the Department of Census and Statistics. Sri Lanka'. The Government of Sri Lanka. http://www.statistics.gov.lk/agriculture/index.htm. Department of Census and Statistics (2012). 'Census of Population and Housing 2011: Preliminary Report'. Colombo, Government of Sri Lanka. Department of Wildlife Conservation (2011). 'Sri Lanka Parks: visitor numbers and income'. Department of Wildlife Conservation Sri Lanka. Department of Wildlife Conservation website. (2010). 'Department of Wildlife Conservation website,.' Retrieved June, 2011, from http://www.dwc.gov.lk/. Dhambarage, C. (2010). 'Sooriyawewa Stadium Opens New Era For Cricket'. Lanka Newspapers. Colombo, The Associated Newspapers of Ceylon Ltd. December 8 issue. Didier, K. A., M. J. Glennon, A. Novaro, E. W. Sanderson, S. Strindberg, S. Walker and S. Di Martino (2009). 'The Landscape Species Approach: spatially-explicit conservation planning applied in the Adirondacks, USA, and San Guillermo- Laguna Brava, Argentina, landscapes'. Oryx 43(4): 476-487. Dirzo, R. and P. H. Raven (2003). 'Global state of biodiversity and loss'. Annual Review of Environment & Resources 28(1): 137-167. Dissanaike, T. (2004). 'Chena and elephants; age-old co-existence '. Daily News. Colombo. November 1 issue. Dissanayake, P. and V. Smakhtin (2007). 'Environmental and social values of river water: Examples from the Menik Ganga, Sri Lanka'. Colombo, Sri Lanka, International Water Management Institute: 15. Dixon, J. A. and P. B. Sherman (1990). Economics of Protected Areas: Approaches and Applications. Washington, DC, East-West Centre.

464 Dowie, M. (2009). Conservation refugees : the hundred-year conflict between global conservation and native peoples. Cambridge, Massachusetts, MIT Press. Dublin, H., A. A. Desai, S. Hedges, J.-C. Vié, C. Bambaradeniya and A. Lopez (2006). 'Elephant Range States Meeting Report, 24 – 26 January 2006, Kuala Lumpur, Malaysia', IUCN Species Survival Commission. Dudley, N. (2008). Guidelines for applying protected area management categories. Gland, Switzerland, IUCN. Dudley, N., S. Stolton, A. Belokurov, L. Krueger, N. Lopoukhine, K. MacKinnon, T. Sandwith and N. Sekhran (2010). 'Natural Solutions: Protected areas helping people cope with climate change'. Gland, Switzerland, Washington DC and New York, USA, The World Bank and WWF. Dunn, K. (2005). Interviewing. Qualitative research methods in human geography. I. Hay. South Melbourne, Victoria, Oxford University Press: xxvii, 339 p. Eagles, P. (2002). 'Trends in Park Tourism: Economics, Finance and Management'. Journal of Sustainable Tourism 10(2): 132-153. Eagles, P. F. J., S. F. McCool and C. D. A. Haynes (2002). 'Sustainable Tourism in Protected Areas: Guidelines for Planning and Management'. Best Practice Protected Area Guideliens. A. Phillips. Gland, Switzerland, IUCN: 183. Economist (2012). 'Carbon prices: A market in need of a miracle'. The Economist. London, UK. March 3 issue. Edirisinghe, D. (2011). 'Human – elephant conflict costly to both sides'. Island. Colombo. April 18 issue. Eliasch, J. (2008). Climate change: financing global forests - the Eliasch review. London, TSO. Emerton, L. and L. D. C. B. Kekulandala (2003). 'Assessment of the Economic Value of Muthurajawela Wetland'. Colombo, IUCN, Sri Lanka. Environmental Foundation (2009). Judges and Environmental Law: a Handbook for the Sri Lankan Judiciary. Colombo, Environmental Foundation Limited. Environmental Foundation. (2012). 'EFL website.' Retrieved September 3, 2012, from http://www.efl.lk/.

465 Farley, J. (2008). 'The Role of Prices in Conserving Critical Natural Capital'. Conservation Biology 22(6): 1399–1408. Fearnhead, P. (2009). Privately managed protected areas. Evolution and innovation in wildlife conservation : parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Fernando, P. (1999). 'Elephants in Sri Lanka: past present and future'. Loris 22(2): 38- 44. Fernando, P., J. Jayewardene, T. Prasad, W. Hendavitharana and J. Pastorini (2011). 'Current Status of Asian Elephants in Sri Lanka'. Gajah 35: 93-103. Fernando, P., M. A. Kumar, A. C. Williams, Eric Wikramanayake, T. Aziz and S. M. Singh (2008). 'Review of human-elephant conflict mitigation measures practiced in South Asia'. Nepal, World Wildlife Fund for Nature. Fernando, P., M. E. Pfrender, S. E. Encalada and R. Lande (2000). 'Mitochondrial DNA variation, phylogeography and population structure of the Asian elephant'. Heredity 84(3): 362-372. Fernando, P., E. Wikramanayake, D. Weerakoon, L. K. A. Jayasinghe, M. Gunawardene and H. K. Janaka (2005). 'Perceptions and patterns of human- elephant conflict in old and new settlements in Sri Lanka: insights for mitigation and management'. Biodiversity and Conservation 14(10): 2465-2481. Fernando, P., E. D. Wikramanayake, H. K. Janaka, L. K. A. Jayasinghe, M. Gunawardene, S. W. Kotagama, D. Weerakoon and J. Pastorini (2008). 'Ranging behaviour of the Asian Elephant in Sri Lanka'. Mammlian Biology 73: 2-13. Fernando, P., E. D. Wikramanayake, D. Weerakoon, H. K. Janaka, M. Gunawardene, L. K. A. Jayasinghe, H. G. Nishantha and J. Pastorini (2006). The future of Asian Elephant conservation: setting sights beyond protected area boundaries. Conservation Biology in Asia. J. A. McNeely, T. M. McCarthy, A. Smith, L. Olsvig-Whittaker and E. D. Wikramanayake. Katmandu, Nepal, Society for Conservation Biology Asia Section and Resources Himalaya Foundation: 252- 260.

466 Fisher, R., S. Maginnis, W. Jackson, E. Barrow and S. Jeanrenaud (2008). Linking conservation and poverty reduction: landscapes, people and power. London; Sterling, VA, Earthscan. Food and Agriculture Organisation (2010). 'Global Forest Resources Assessment 2010: Country Report, Sri Lanka'. Rome, FAO. Frost, P., B. Campbell, G. Medina and L. Usongo (2006). 'Landscape-scale approaches for integrated natural resource management in tropical forest landscapes'. Ecology and Society 11(2): online. Frost, P. G. H. and I. Bond (2008). 'The CAMPFIRE programme in Zimbabwe: Payments for wildlife services'. Ecological Economics 65: 776-787. Garrod, G. D. and K. G. Willis (1997). 'The non-use benefits of enhancing forest biodiversity: A contingent ranking study'. Ecological Economics 21(1): 45-61. Geiser, U. (1995). 'Indigenous resource management and external development interventions in the dry zone of Sri Lanka: from conflicts to strategies'. GeoJournal 35(2): 185-196. Gilpin, A. (2000). Environmental economics: a critical overview. Chichester ; New York, John Wiley. GMSL. (2012). 'Green Movement of Sri Lanka website.' Retrieved September 3, 2012, from http://www.gmsl.lk/. Gollier, C. (2002). 'Discounting an uncertain future'. Journal of Public Economics 85: 149–166. Gomez-Baggethun, E. and M. Ruiz-Perez (2011). 'Economic valuation and the commodification of ecosystem services'. Progress in Physical Geography 35(5): 613–628. Gomiero, T., D. Pimentel and M. G. Paoletti (2011). 'Is There a Need for a More Sustainable Agriculture?'. Critical Reviews in Plant Sciences 30(1-2): 6-23. Government of India (2011). 'Guidelines for ecotourism in and around protected areas'. I. Ministry of Environment and Forests. New Delhi, India. Government of Sri Lanka (2005). 'Fauna and Flora Protection Ordinance No. 2 of 1937 (amended by Acts No. 49 of 1993, 12 of 2005)'. Government of Sri Lanka. Colombo.

467 Government of Sri Lanka (2005). 'A ten year horizon development framework: 2006 - 2016'. Colombo, Ministry of Finance and Planning. Government of Sri Lanka (2006). 'National policy for the conservation and management of wild elephants in Sri Lanka'. Government of Sri Lanka. Colombo. Graham, M., N. Gichohi, F. Kamau, G. Aike, B. Craig, I. Douglas-Hamilton and W. M. Adams (2009). 'The Use of Electrified Fences to Reduce Human Elephant Conflict: A Case Study of the Ol Pejeta Conservancy, Laikipia District, Kenya'. Laikipia Elephant Project Working Paper 1. Nanyuki, Kenya, Laikipia Elephant Project. Green, M. J. B. and E. R. N. Gunawardena (1997). 'Designing an optimum protected areas system for Sri Lanka’s natural forests, Volume 1'. Colombo, Ministry of Forestry and Environment. Greiber, T., M. Janki, M. Orellana, A. Savaresi and D. Shelton (2009). 'Conservation with Justice. A Rights-based Approach'. Gland, Switzerland, IUCN: xiv + 118 pp. Grimble, R. (1995). 'Stakeholder analysis for natural resource management in developing countries'. Narural Resources Forum 19(2): 113 - 124. Grimble, R. (1998). 'Stakeholder methodologies in natural resource management'. Chatham, UK, Natural Resources Institute. Grimble, R. and K. Wellard (1997). 'Stakeholder Methodologies in Natural Resource Management: a Review of Principles, Contexts, Experiences and Opportunities'. Agricultural Systems 55(2): 173 - 193. Grimes, A., S. Loomis, P. Jahnige, M. Burnham, K. Onthank, R. Alarcón, W. P. Cuenca, C. C. Martinez, D. Neill, M. Balick, B. Bennett and R. Mendelsohn (1994). 'Valuing the rain forest: The economic value of nontimber forest products in Ecuador'. Ambio. Stockholm 23(7): 405-410. Groom, M. J., G. K. Meffe and C. R. Carroll (2006). Principles of conservation biology. Sunderland, Mass., Sinauer Associates. Gunaratne, L. H. P. and P. K. Premarathne (2006). 'The effectiveness of electric fencing in mitigating human-elephant conflict in Sri Lanka'. Singapore, Economy and Environment Program for Southeast Asia.

468 Gunatilake, H. M., D. M. A. H. Senaratne and P. Abeygunawardena (1993). 'Role of Non-Timber Forest Products in the Economy of Peripheral Communities of Knuckles National Wilderness Area of Sri Lanka: A Farming Systems Approach'. Economic Botany 47(3): 275-281. Gunatilleke, N., R. Pethiyagoda and S. Gunatilleke (2008). 'Biodiversity of Sri Lanka'. Journal of the National Science Foundation of Sri Lanka 36 (Special Issue): 25- 62. Gunderson, L. H. and C. S. Holling (2002). Panarchy: understanding transformations in human and natural systems. Washington, DC, Island Press. Ha, N. V., S. Kant and V. Maclaren (2008). 'Shadow prices of environmental outputs and production efficiency of household-level paper recycling units in Vietnam'. Ecological Economics 65: 98 - 110. Hadker, N., S. Sharma, A. David and T. R. Muraleedharan (1997). 'Willingness-to-pay for Borivli National Park: evidence from a Contingent Valuation'. Ecological Economics 21(2): 105-122. Hanley, N. and E. B. Barbier (2009). Pricing Nature: Cost–Benefit Analysis and Environmental Policy. Cheltenham, UK, Edward Elgar Publishing Limited. Hannah, L., G. Midgley, S. Andelman, M. Araujo, G. Hughes, E. Martinez-Meyer, R. Pearson and P. Williams (2007). 'Protected area needs in a changing climate'. Frontiers in Ecology and the Environment 5(3): 131-138. Harvey, C. A., O. Komar, R. Chazdon, B. G. Ferguson, B. Finegan, D. M. Griffith, M. Martinez-Ramos, H. Morales, R. Nigh, L. Soto-Pinto, M. Van Breugel and M. Wishnie (2008). 'Integrating agricultural landscapes with biodiversity conservation in the Mesoamerican hotspot'. Conservation Biology 22(1): 8-15. Haviland, C. (2012). 'Speeding and danger in Sri Lanka's safari parks'. BBC. Colombo, British Broadcasting Corporation. 26th April. Henson, A., D. Williams, J. Dupain, H. Gichohi and P. Muruthi (2009). 'The Heartland Conservation Process: enhancing biodiversity conservation and livelihoods through landscape-scale conservation planning in Africa'. Oryx 43(4): 508-519. Hettiarachchi, K. (2009). 'Why do elephants come back home?'. Sunday Times Sri Lanka. Colombo, Sri Lanka. April 5 issue.

469 Hettiarachchi, K. (2011). 'One elephant for every 3,300 persons in Lanka'. Sunday Times Sri Lanka. Colombo, Sri Lanka. September 4 issue. Hoare, R. (2003). 'Fencing and other barriers against problem elephants'. Technical brief series. Nairobi, Kenya, IUCN African Elephant Specialist Group. Hole, D. G., S. G. Willis, D. J. Pain, L. D. Fishpool, S. H. M. Butchart, Y. C. Collingham, C. Rahbek and B. Huntley (2009). 'Projected impacts of climate change on a continent-wide protected area network'. Ecology Letters 12(5): 420- 431. Holmes, T., K. Alger, C. Zinkhan and E. Mercer (1998). 'The effect of response time on conjoint analysis estimates of rainforest protection values'. Journal of Forest Economics 4(1): 7-28. Holt, J. C. (2011). The Sri Lanka Reader: History, Culture, Politics. Durham, North Carolina, Duke University Press. Houghton, K. T. and R. Mendelsohn (1996). 'An economic analysis of multiple-use forestry in Nepal.'. Ambio 25(3): 156-159. Howard, P. C. (1995). 'The Economics of Protected Areas in Uganda: Costs, Benefits and Policy Issues', University of Edinburgh. PhD Thesis. Huang, M., S. K. Upadhyaya, R. Jindal and J. Kerr (2009). 'Payments for Watershed Services in Asia: A Review of Current Initiatives'. Journal of Sustainable Forestry 28(3): 551- 575. Hussen, A. M. (2004). Principles of Environmental Economics, second edition. New York, USA, Routledge Publishing. Imbulana, K. A. U. S., U. Ratnayake, G. B. B. Herath, E. Koncagul and B. R. Neupane (2009). 'Case Study in Walawe Basin Sri Lanka, A contribution to the United Nations World Water Assessment Programme'. Paris, UNESCO. Inskip, C. and A. Zimmermann (2009). 'Human-felid conflict: a review of patterns and priorities worldwide'. Oryx 43(1): 18-34. IPCC (2007). 'Climate Change 2007: Synthesis Report', IPCC: 73. Isaacs, J. C. (2000). 'The Limited Potential of Ecotourism to Contribute to Wildlife Conservation'. Wildlife Society Bulletin 28(1): 61-69.

470 Island newspaper (2006). 'Lunugamvehera reservoir blunders exposed'. Island. Colombo, Sri Lanka. Island newspaper (2011). 'Marketing an emerging tourism destination in Asia: A country perspective of innovation and diversification'. Island. Colombo, Sri Lanka. November 28 issue. IUCN (2000). The IUCN Policy Statement on Sustainable Use of Wild Living Resources. IUCN World Conservation Congress, Amman, Jordan. IUCN (2009). The 2009 Red List of threatened animals. Gland, Switzerland. IUCN Sri Lanka (2007). The 2007 Red List of threatened fauna and flora of Sri Lanka. Colombo, Sri Lanka. IUCN Sri Lanka website. 'http://www.iucn.org/about/union/secretariat/offices/asia/asia_where_work/srilan ka/.' Retrieved Jaunuary, 2011. Jayasooriya, A. H. M., D. Kitchener and M. B. Chadrashekar (2006). 'Portfolio of Strategic Conservation Sites/ Protected Area Gap Analayis in Sri Lanka'. Colombo, Ministry of Environment & Natural Resources. Jayawardene, J. (2001). The care and management of domesticated Asian elephants in Sri Lanka. International workshop on the domesticated Asian elephant, Bangkok, Food and Agriculture Organisation of the United Nations. Jayawardene, J. (2010). 'Human Aspects of Human-Elephant Conflicts'. Gajah: Journal of the IUCN/SSC Asian Elephant Specialist Group (32): 43-44. Jayawardene, O. C. (1990). 'Landmarks in Lanka’s Land Policy'. Economic Review Sri Lanka 15(10): 10-11. Jazeel, T. (2005). '‘Nature’, nationhood and the poetics of meaning in Ruhuna (Yala) National Park, Sri Lanka'. Cultural geographies 12: 199-227. Jones, B. and L. C. Weaver (2009). CBNRM in Namibia: growth, trends, lessons and constraints. Evolution and innovation in wildlife conservation: parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Kaimowitz, D. and D. Sheil (2007). 'Conserving What and for Whom? Why Conservation Should Help Meet Basic Human Needs in the Tropics'. Biotropica 39(5): 567-574.

471 Kalamandeen, M. and L. Gillson (2007). 'Demything "wilderness": implications for protected area designation and management'. Biodiversity and Conservation 16(1): 165-182. Kannangara, N. (2012). 'Mattala Airport Expressway Disturbs The Way Of Life'. Sunday Leader Sri Lanka. Colombo. January 22 issue. Kareiva, P., H. Tallis, T. H. Ricketts, G. Daily and S. Polasky (2011). Natural capital: theory & practice of mapping ecosystem services. New York, Oxford University Press. Kariyawasam, D. (2007). 'Social dynamics at the forest edge: a changing frontier of the national heritage of Sri Lanka'. School of Forestry and Environmental Studies. New Haven, Connecticut, Yale University. PhD thesis. Kelegama, S. (2006). Development under stress : Sri Lankan economy in transition. New Delhi ; Thousand Oaks, Calif. ; London, Sage Publications. Kioko, J., P. Muruthi, P. Omondi and P. I. Chiyo (2008). 'The performance of electric fences as elephant barriers in Amboseli, Kenya'. South African Journal of Wildlife Research 38(1): 52–58. Knight, A. T., R. M. Cowling and B. M. Campbell (2006). 'An operational model for implementing conservation action'. Conservation Biology 20(2): 408-419. Kolowski, J. M. and K. E. Holekamp (2006). 'Spatial, temporal, and physical characteristics of livestock depredations by large carnivores along a Kenyan reserve border'. Biological Conservation 128: 529-541. Kramer, R. A., N. P. Sharma and M. Munasinghe (1995). Valuing Tropical Forests: Methodology and Case Study of Madagascar. Washington DC, World Bank. Krishnaswamy, P. (2010). 'Ampara, Batticaloa districts: Synergetic mechanism clicks with 'Eastern Revival' '. Sunday Observer. Colombo, The Associated Newspapers of Ceylon Ltd. March 14 issue. Krüger, O. (2005). 'The role of ecotourism in conservation: panacea or Pandora’s box?'. Biodiversity and Conservation 14(3): 579-600. Kuruwita, R. (2011). 'Park rangers work to rule'. Lakbima News. Colombo. September 4 issue. Ladduwahetty, R. (2011). 'Solutions to raging human-elephant conflict'. Island Sri Lanka. Colombo. October 23 issue.

472 Lamarque, F., J. Anderson, R. Fergusson, M. Lagrange, Y. Osei-Owusu and L. Bakker (2009). 'Human-wildlife conflict in Africa: Causes, consequences and management strategies'. Rome, Italy, Food and Agriculture Organisation. Land Commissioner General’s Department website. Retrieved January, 2011, from http://www.landcommdept.gov.lk/index.php. Langford, I. H., M. S. Skourtos, A. Kontogianni, R. J. Day, S. Georgiou and I. J. Bateman (2001). 'Use and nonuse values for conserving endangered species: the case of the Mediterranean monk seal'. Environment and Planning A 33(12): 2219-2233. Lanka business online. (2010). 'Sri Lanka tourism industry should focus on niche: officials.' Retrieved 12 October, 2010. Laurance, W. F. (2008). 'Can carbon trading save vanishing forests?'. BioScience 58(4): 286-287. Laurance, W. F., D. C. Useche, J. Rendeiro, M. Kalka, C. J. A. Bradshaw, S. P. Sloan, S. G. Laurance and M. Campbell (2012). 'Averting biodiversity collapse in tropical forest protected areas'. Nature 489: 290-294. Layke, C. (2009). 'Measuring Nature’s Benefits: A Preliminary Roadmap for Improving Ecosystem Service Indicators'. Washington DC, World Resources Institute: 36. Leimgruber, P., J. B. Gagnon, C. Wemmer, D. S. Kelly, M. A. Songer and E. R. Selig (2003). 'Fragmentation of Asia's remaining wildlands: implications for Asian elephant conservation'. Animal Conservation 6: 347-359. Locke, H. and P. Dearden (2005). 'Rethinking protected area categories and the new paradigm'. Environmental Conservation 32(1): 1-10. Lonely Planet. (2012). 'Lonely Planet review for Yala National Park ' Retrieved January 24, 2012, from http://www.lonelyplanet.com/sri-lanka/the-south/yala- national-park/sights/other/yala-national-park. Lorimer, J. (2010). 'Elephants as companion species: the lively biogeographies of Asian elephant conservation in Sri Lanka'. Transactions of the Institute of British Geographers 35(4): 491-506. Lorimer, J. and S. Whatmore (2009). 'After the 'king of beasts': Samuel Baker and the embodied historical geographies of elephant hunting in mid-nineteenth-century Ceylon'. Journal of Historical Geography 35(4): 668-689.

473 Mackenzie, J., P.-L. Tan, S. Hoverman and C. Baldwin (2012). 'The value and limitations of Participatory Action Research methodology'. Journal of Hydrology 474: 11–21. Maginnis, S., W. Jackson and N. Dudley (2004). Conservation Landscapes: Whose landscapes? Whose trade-offs? Getting biodiversity projects to work : towards more effective conservation and development. T. O. McShane and M. P. Wells. New York, Columbia University Press: 321-339. Mahapatra, A. K. and D. D. Tewari (2005). 'Importance of non-timber forest products in the economic valuation of dry deciduous forests of India'. Forest Policy and Economics 7(3): 455-467. Mahaweli Authority of Sri Lanka website. Retrieved January, 2011, from http://www.mahaweli.gov.lk/. Malhi, Y. and J. Wright (2004). 'Spatial Patterns and Recent Trends in the Climate of Tropical Rainforest Regions'. Philosophical Transactions: Biological Sciences 359(1443): 311-329. Manor, J. (1990). The Bandaranaike Legend. Sri Lanka: History and the Roots of Conflict. J. Spencer. New York, Routledge: 599 - 606. Mansourian, S., A. Belokurov and P. J. Stephenson (2009). 'The role of forest protected areas in adaptation to climate change'. Unasylva 60(231): 63-68. Manukulasooriya, M. A. R. (2008). 'Villagers fume after wild elephant kills man '. Daily News Sri Lanka. Colombo, Sri Lanka. September 11 issue. Margules, C. R. and R. L. Pressey (2000). 'Systematic conservation planning'. Nature 405(6783): 243-253. Martín-López, B., M. García-Llorente, I. Palomo and C. Montes (2011). 'The conservation against development paradigm in protected areas: Valuation of ecosystem services in the Doñana social–ecological system (southwestern Spain)'. Ecological Economics 70: 1481-1491. Marvasti, A. B. (2004). Qualitative Research in Sociology: An Introduction. London, SAGE Publications. Mattsson, E., U. M. Persson, M. Ostwald and S. P. Nissanka (2012). 'REDD+ readiness implications for Sri Lanka in terms of reducing deforestation'. Journal of Environmental Management 100: 29 - 40.

474 Mayers, J. (2005). 'Stakeholder power analysis'. London, UK, International Institution for Environment and Development: 24. McCauley, D. J. (2006). 'Selling out on nature'. Nature 443: 27 - 28. McGilvray, D. and M. Raheem (2011). Origins of Sri Lankan Muslims and variety of the Muslim identity. The Sri Lanka Reader: History, Culture, Politics. J. C. Holt. Durham, North Carolina, Duke University Press: 410 - 419. Mcintyre, A. (2008). Participatory Action Research Thousand Oaks, California, Sage Publications, Inc. McIntyre, K. (2011). 'Bhutan: a model for sustainable tourism developments'. Forum: 15-17. McNeely, J. A. and S. A. Mainka (2009). Conservation for a New Era. Gland, Switzerland, IUCN. Meffe, G. K. and C. R. Carroll (1997). Principles of conservation biology. Sunderland, Ma., Sinauer. Meffe, G. K. and S. Viederman (1995). 'Combining Science and Policy in Conservation Biology'. Wildlife Society Bulletin 23(3): 327-332. Melnyk, M. and N. Bell (1996). 'The direct-use values of tropical moist forest foods: The Huottuja (Piaroa) Amerindians of Venezuela'. Ambio 25(7): 468-472. Mendis, C. (2011). 'Going wild!'. Financial Times Sri Lanka. Colombo. August 23 issue. Mendis, R. (2009). 'Experts debate jumbo problem'. The Sunday Leader. Colombo. October 18 issue. Millennium Ecosystem Assessment (2005). 'Ecosystems and Human Well-being: Biodiversity Synthesis'. Washington, DC., World Resources Institute. Miller, T. R., B. A. Minteer and L.-C. Malan (2011). 'The new conservation debate: The view from practical ethics'. Biological Conservation 144: 948–957. Milner-Gulland, E. J. (2004). 'Against extinction: The story of conservation'. Nature 429(6990): 346-347. Ministry of Agricultural Development and Agrarian Services (2007). 'National Agricultural Policy'. Colombo, Sri Lanka. Ministry of Environment and Natural Resources (2003). 'Caring for the Environment 2003-2007: path to sustainable development.'. Battaramulla, Sri Lanka: 152.

475 Ministry of Environment Sri Lanka (2000). 'Initial National Communication under the United Nations Framework Convention on Climate Change: Sri Lanka'. Battaramulla, Sri Lanka, Ministry of Environment Sri Lanka: 122. Ministry of Environment Sri Lanka (2010). 'National Climate Change Adaptation Strategy for Sri Lanka 2011 to 2016'. Battaramulla, Sri Lanka, Ministry of Environment Sri Lanka. Ministry of Environment Sri Lanka (2010). 'Progress Report 2009 and Action Plan 2010'. Battaramulla, Sri Lanka, Ministry of Environment Sri Lanka. Ministry of Environment Sri Lanka (2012). 'Sri Lanka REDD+ Readiness Preparation Proposal'. Battaramulla, Sri Lanka, Ministry of Environment Sri Lanka: 209. Ministry of Forestry and Environment (1999). Biodoversity conservation in Sri Lanka: A framework for action. Battaramulla, Sri Lanka. Miththapala, S. (2009). 'The Uda Walawe Elephant Transit Home - Another Opportunity Missed?'. Gajah: Journal of the IUCN/SSC Asian Elephant Specialist Group (30): 24-28. Mittermeier, R. A., P. Gil, M. Hoffmann, J. Pilgrim, C. G. Mittermeier, J. F. Lamoreux, G. da Fonseca, Cemex S.A. de C.V., Agrupación Sierra Madre. and Conservation International. (2004). Hotspots revisited: earth's biologically richest and most endangered terrestrial ecoregions. Mexico City, CEMEX. Molinos-Senante, M., R. Reif, M. Garrido-Baserba, F. Hernández-Sancho, F. Omil, M. Poch and R. Sala-Garrido (2013). 'Economic valuation of environmental benefits of removing pharmaceutical and personal care products from WWTP effluents by ozonation'. Science of the Total Environment 461: 409–415. Mori, S. A. (1992). The Brazil Nut Industry—Past, Present, and Future. Sustainable Harvest and Marketing of Rain Forest Products. M. Plotkin and L. Famolare. Washington DC, Island Press. Morrison, J., C. Loucks, B. Long and E. Wikramanayake (2009). 'Landscape-scale spatial planning at WWF: a variety of approaches'. Oryx 43(4): 499-507. Muchapondwa, E., F. Carlsson and G. Kohlin (2008). 'Wildlife Management in Zimbabwe: Evidence from a Contingency Valuation Study'. Economic Society of South Africa 76(4): 685-704.

476 Mullan, K. and A. Kontoleon (2008). 'Benefits and costs of forest biodiversity: Economic theory and case study evidence'. Cambridge, UK, Department of Land Economy, University of Cambridge: 82. Murthy, I. K., P. R. Bhat, N. H. Ravindranath and R. Sukumar (2005). 'Financial valuation of non-timber forest product flows in Uttara Kannada district, Western Ghats, Karnataka'. Current Science 88(10): 1573-1579. Mushove, P. and C. Vogel (2005). 'Heads or tails? Stakeholder analysis as a tool for conservation area management'. Global Environmental Change 15: 184–198. Myers, N. (1988). 'Threatened biotas: "Hot spots" in tropical forests '. The Environmentalist 8(3): 187-208. Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca and J. Kent (2000). 'Biodiversity hotspots for conservation priorities'. Nature 403(6772): 853-858. Naalir, M. (2010). 'Rapid development of ports to make Lanka a shipping hub'. Sunday Observer Sri Lanka. Colombo. March 28 issue. Nagendra, H., S. Pareeth and R. Ghate (2006). 'People within parks—forest villages, land-cover change and landscape fragmentation in the Tadoba Andhari Tiger Reserve, India'. Applied Geography 26: 96–112. Naidoo, R. and W. L. Adamowicz (2005). 'Biodiversity and nature-based tourism at forest reserves in Uganda'. Environment and Development Economics 10(2): 159-178. Nanayakkara, V. R. (1987). 'Forest history of Sri Lanka'. The Sri Lanka Forester XVIII(1&2): 15-26. Nelson, A., P. Bidwell and C. Sillero-Zubiri (2003). 'A review of human-elephant conflict management strategies'. Oxford, UK, People and Wildlife Initiative, Wildlife Conservation Research Unit, Oxford University. Neuman, W. L. (2007). Basics of Social Research: Qualitative and Quantitative Approaches. Boston, Massachusetts, USA, Pearson Education nc. Ninan, K. N. (2007). The economics of biodiversity conservation: valuation in tropical forest ecosystems. London; Sterling, VA, Earthscan. Nissan, E. and R. L. Stirrat (1990). The generation of communal identities. Sri Lanka History and the Roots of Conflict. J. Spencer. London, Routledge: 19 - 44.

477 Nizam, A. A. M. (2011). 'Over 200,000 acres of paddy land destroyed by floods in Sri Lanka '. Asia Tribune. Colombo, World Institute for Asian Studies|. January 11 issue. Nizam, A. A. M. (2011). 'President Mahinda Declares Open Hambantota Administration Complex '. Asia Tribune. Colombo, World Institute for Asia Studies. 11. Nizam, I. (2011). 'Wildlife director sacked over government's loss of votes in Wayamba'. Island Sri Lanka. Colombo. July 27 issue. Nongbri, N. (2003). 'Elephant hunting in late 19th century north-east India: mechanisms of control'. Economic and political weekly 38(30): 3189-3199. O'Neill, J. (1997). 'Managing without Prices: The Monetary Valuation of Biodiversity'. Ambio 26(8): 546-550. Odunga, P. O. and S. W. Maingi (2011). 'The Wildlife Tourism Market to Kenya (2002–2003)'. Journal of Travel & Tourism Marketing 28(1): 62-80. Oksanen, M. (1997). 'The Moral Value of Biodiversity'. Ambio 26(8): 541-545. Padoch, C. and W. de Jong (1989). Production and profit in agroforestry: an example from the Peruvian Amazon. Fragile lands of Latin America: strategies for sustainable development. J. O. Browder. Boulder, Colorado, USA, Westview Press: 102–113. Pagiola, S. (2008). 'Payments for environmental services in Costa Rica'. Ecological Economics 65: 712 - 724. Pagiola, S., J. Bishop and N. Landell-Mills (2002). Selling forest environmental services : market-based mechanisms for conservation and development. London; Sterling, VA, Earthscan Publications. Pagiola, S., K. v. Ritter and J. Bishop (2004). 'Assessing the Economic Value of Ecosystem Conservation'. Washington DC, World Bank. Parks, T. H., T. A. Parks and C. Allen (2009). 'The development of a framework for studying ecotourism'. International Journal of Management 26(1): 89-98. Pearce, D. (1991). An economic approach to saving the tropical forests. Economic policy towards the Environment. D. Helm. Oxford, Blackwell Publishers.

478 Pearce, D., G. Atkinson and S. Mourato (2006). Cost-Benefit Analysis and the Environment: Recent Developments. Paris, France, Organisation for Economic Cooperation and Development (OECD). Pearce, D. W. (2001). 'The economic value of forest ecosystems'. Ecosystem Health 7(4): 284-296. Pearce, D. W., D. Moran and IUCN Biodiversity Programme. (1994). The economic value of biodiversity. London, Earthscan. Peebles, P. (2006). The history of Sri Lanka. Westport, CT, Greenwood Press. Peiris, G. H. (1978). 'Land-Reform and Agrarian Change in Sri-Lanka'. Modern Asian Studies 12(Oct): 611-628. Perera, A. (2010). 'No peace in sight in human-vs-elephant war'. IPS news,. Colombo, Sri Lanka, IPS news,. Perera, B. M. A. O. (2007). Status of Elephants in Sri Lanka and the Human-Elephant Conflict. Managing the Health and Reproduction of Elephant Populations in Asia, Bangkok, Thailand. Perera, B. M. A. O. (2009). 'The human-elephant conflict: a review of current status and mitigation methods'. Gajah: Journal of the IUCN/SSC Asian Elephant Specialist Group (30): 41-52. Perera, B. M. K., M. H. J. P. Fernando, Y. M. Wicremasinghe and M. S. Perera (1983). 'Chena development experiment — Paindikulama'. Norwegian Journal of Geography 37(3-4): 187-196. Perrings, C. (2010). 'The economics of biodiversity: the evolving agenda'. Environment and Development Economics 15: 721–746. Peters, C. M., A. H. Gentry and R. O. Mendelsohn (1989). 'Valuation of an Amazonian rainforest'. Nature 339: 655-656. Pethiyagoda, R. (2000). 'The Fauna and Flora Protection Ordinance'. Loris 22(3): 27-38. Pethiyagoda, R. (2000). 'Fauna and Flora Protection Ordinance'. Loris 22(3): 27-38. Phillips, A. (2002). Management guidelines for IUCN category V protected areas : protected landscapes/seascapes. Gland, Switzerland, IUCN--the World Conservation Union.

479 Phua, M. H. and M. Minowa (2005). 'A GIS-based multi-criteria decision making approach to forest conservation planning at a landscape scale: a case study in the Kinabalu Area, Sabah, Malaysia'. Landscape and Urban Planning 71(2-4): 207- 222. Pimmanrojnagool, V. and S. Wanghongsa (2001). A study of street wandering elephants in Bangkok and the socio-economic life of their mahouts. International workshop on the domesticated Asian elephant, Bangkok, Food and Agriculture Organisation of the United Nations. Pinedo-Vasquez, M., D. Zarin and P. Jipp (1992). 'Economic returns from forest conversion in the Peruvian Amazon'. Ecological Economics 6(2): 163-173. Plummer, R. and D. A. Fennell (2009). 'Managing protected areas for sustainable tourism: prospects for adaptive co-management'. Journal of Sustainable Tourism 17(2): 149–168. Prato, T. (1998). Natural Resource and Environmental Economics. Ames, Iowa, Iowa State University Press. Presidential Secretariat (2010). 'Notification of Ministries and the assignment of subjects and functions and Departments, Public Corporations and Statutory Institutions'. Government of Sri Lanka. Colombo. Pressey, R. L., G. L. Whish, T. W. Barrett and M. E. Watts (2002). 'Effectiveness of protected areas in north-eastern New South Wales: recent trends in six measures'. Biological Conservation 106(1): 57-69. Rajapaksa, M. (2010). 'Mahinda Chinthana: A vision for a new Sri Lanka'. Colombo, Sri Lanka. Rajapaksa, R. C. (2007). Captive Breeding of Elephants at Pinnawala Elephant Orphanage in Sri Lanka. Managing the Health and Reproduction of Elephant Populations in Asia, Bangkok, Thailand. Ramsar Convention. (2012). 'The Annotated Ramsar List: Sri Lanka.' Retrieved January 24, 2012, from http://www.ramsar.org/cda/en/ramsar-pubs-annolist- annotated-ramsar-16173/main/ramsar/1-30-168%5E16173_4000_0__. Rangarajan, M., A. Desai, R. Sukumar, P. Easa, Vivek Menon, S. Vincent, S. Ganguly, B. Talukdar, Brijendra Singh, D. Mudappa, S. Chowdhary and A. Prasad (2010).

480 'Securing the Future for Elephants in India'. New Delhi, India, Ministry of Environment and Forests India. Rathnayake, R. M. K. (2007). Rural-urban linkages in regional development: the experience of north central Sri Lanka. Malabe, Sri Lanka, Piyasiri Printer & Company. Reed, M. S., A. Graves, N. Dandy, H. Posthumus, K. Hubacek, J. Morris, C. Prell, C. H. Quinn and L. C. Stringer (2009). 'Who's in and why? A typology of stakeholder analysis methods for natural resource management'. Journal of Environmental Management 90(5): 1933-1949. Riddle, H. S., B. A. Schulte, A. A. Desai and L. van der Meer (2010). 'Elephants - a conservation overview'. Journal of Threatened Taxa 2(1): 653-661. Ritchie, J. (2003). The Applications of Qualitative Methods to Social Research. Qualitative Research Practice: A Guide for Social Science Students and Researchers. J. Ritchie and J. Lewis. London, Sage publications: 24-46. Ritchie, J. and J. Lewis (2003). Generalising from Qualitative Research. Qualitative Research Practice: A Guide for Social Science Students and Researchers. J. Ritchie and J. Lewis. London, Sage publications: 263-286. Robinson, J. G., K. H. Redford and E. L. Bennett (1999). 'Wildlife Harvest in Logged Tropical Forests'. Science 284(5414): 595-596. Rodrigo, M. (2011). 'Don’t stump the ‘Stumpies’ of Hambantota'. Sunday Times Sri Lanka. Colombo. February 27 issue. Rodrigo, M. (2011). 'Villagers block junction demanding solution to Human-Elephant Conflict'. Sunday Times Sri Lanka. Colombo. July 31 issue. Rodrigues, A. S. L., S. J. Andelman, M. I. Bakarr, L. Boitani, T. M. Brooks, R. M. Cowling, L. D. C. Fishpool, G. A. B. da Fonseca, K. J. Gaston, M. Hoffmann, J. S. Long, P. A. Marquet, J. D. Pilgrim, R. L. Pressey, J. Schipper, W. Sechrest, S. N. Stuart, L. G. Underhill, R. W. Waller, M. E. J. Watts and X. Yan (2004). 'Effectiveness of the global protected area network in representing species diversity'. Nature 428(6983): 640-643. Romanelli, A., H. E. Massone and A. H. Escalante (2011). 'Stakeholder Analysis and Social-Biophysical Interdependencies for Common Pool Resource Management:

481 La Brava Wetland (Argentina) as a Case Study'. Environmental Management 48: 462–474. Rosales, R., M. Kallesoe, P. Gerrard, P. Muangchanh, S. Phomtavong and S. Khamsomphou (2005). 'Balancing the returns to catchment management: The economic value of conserving natural forests in Sekong, Lao PDR'. IUCN Water, Nature and Economics Technical Paper No. 5, IUCN--The World Conservation Union, Ecosystems and Livelihoods Group Asia. Roux, D. J. and L. C. Foxcroft (2011). 'The development and application of strategic adaptive management within South African National Parks'. Koedoe 53(2): 1-5. Rozemeijer, N. (2009). CBNRM in Botswana. Evolution and innovation in wildlife conservation: parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Ruitenbeek, H. J. (1988). 'Resettlement of Inhabitants in Korup National Park, Cameroon'. Korup Research Note No. 1. London. Rustad, L. E. (2008). 'The response of terrestrial ecosystems to global climate change: towards an integrated approach'. Science of the Total Environment 404(2-3): 222-235. Saastamoinen, O. (1992). 'Economic Evaluation of Biodiversity Values of Dipterocarp Forests in the Philippines'. Second Meeting of the International Society of Ecological Economics (ISEE). 3. Saayman, M. and A. Saayman (2010). 'Regional development and national parks in South Africa: lessons learned'. Tourism Economics 16(4): 1037–1064. Sahajananthan, S. (1987). 'Forest management in Sri Lanka: A glimpse at the past and the present'. The Sri Lanka Forester XVIII(1&2): 45-51. Salem, B. B. (2003). 'Application of GIS to biodiversity monitoring'. Journal of Arid Environments 54(1): 91-114. Samaranayake, M. R. (2007). 'Participatory community outreach for protected area management and wildlife conservation'. Colombo, Ministry of Environment and Natural Resources. Samaratunga, P. A., M. I. M. Rafeek and R. M. H. Banda (2001) 'Future rice production policy in Sri Lanka: global and national perspectives'.

482 Samaratunge, R., Q. Alam and J. Teicher (2008). 'The New Public Management reforms in Asia: a comparison of South and Southeast Asian countries'. International Review of Administrative Sciences 74(1): 25-46. Sanderson, E. W., K. H. Redford, A. Vedder, P. B. Coppolillo and S. E. Ward (2002). 'A conceptual model for conservation planning based on landscape species requirements'. Landscape and Urban Planning 58(1): 41-56. Sandika, A. L. and S. N. Dushani (2009). 'Growth performance of rice sector: the present scenario in Sri Lanka'. Tropical Agricultural Research & Extension 12(2): 71-76. SANParks (2010). 'SANParks Annual Report 2009/2010'. Johannesburg, South African National Parks. Santiapillai, C. (1997). 'The Asian elephant conservation: a global strategy'. Gajah 18: 21-39. Santiapillai, C., P. Fernando and M. Gunawardene (2006). 'A strategy for the conservation of the Asian elephant in Sri Lanka'. Gajah: Journal of the IUCN/SSC Asian Elephant Specialist Group (25): 91-102. Santiapillai, C. and P. Jackson (1990). 'The Asian Elephant: an action plan for conservation'. Gland, Switzerland, IUCN/SSC Asian Elephant Specialist Group. Santiapillai, C. and R. Jayewardene (2004). 'Conservation of the leopard and other carnivores in Sri Lanka'. Current Science 86(8): 1063-1064. Santiapillai, C., A. Silva, C. Karyawasam, S. Esufali, S. Jayaniththi, M. Basnayake, V. Unantenne and S. Wijeyamohan (1999). 'Trade in Asian elephant ivory in Sri Lanka'. Oryx 33(2): 176-180. Santiapillai, C., S. Wijeyamohan, G. Bandara, R. Athurupana, Naveen, Dissanayake and B. Read (2010). 'An assessment of the human-elephant conflict in Sri Lanka'. Ceylon Journal of Science (Biological Science) 39(1): 21-33. Saparamadu, S. D. (2006). Sri Lanka a wildlife interlude. Dehiwela, Sri Lanka, Tisara Press. Sayer, J., B. Campbell, L. Petheram, M. Aldrich, M. R. Perez, D. Endamana, Z. L. N. Dongmo, L. Defo, S. Mariki, N. Doggart and N. Burgess (2007). 'Assessing environment and development outcomes in conservation landscapes'. Biodiversity and Conservation 16(9): 2677-2694.

483 Sayer, J. and M. P. Wells (2004). The pathology of projects. Getting biodiversity projects to work: towards more effective conservation and development. T. O. McShane and M. P. Wells. New York, Columbia University Press: 321-339. Scheepers, K., L. Swemmer and W. J. Vermeulen (2011). 'Applying adaptive management in resource use in South African National Parks: A case study approach'. Koedoe 53(2): 1-14. Schmitt, C. B., N. D. Burgess, L. Coad, A. Belokurov, C. Besancon, L. Boisrobert, A. Campbell, L. Fish, D. Gliddon, K. Humphries, V. Kapos, C. Loucks, I. Lysenko, L. Miles, C. Mills, S. Minnemeyer, T. Pistorius, C. Ravilious, M. Steininger and G. Winkel (2009). 'Global analysis of the protection status of the world's forests'. Biological Conservation 142(10): 2122-2130. Secretariat of the Convention on Biological Diversity. (2012). 'Website of the Convention on Biological Diversity.' Retrieved August 13, 2012, from http://www.cbd.int/. Shultis, J. D. and P. A. Way (2006). 'Changing Conceptions of Protected Areas and Conservation: Linking Conservation, Ecological Integrity and Tourism Management'. Journal of Sustainable Tourism 14(3): 223 - 237. Siegfried, W. R., G. A. Berm and C. M. Gelderblom (1998). 'Regional assessment and conservation implications of landscape characteristics of African National Parks '. Biological Conservation 84: 131-140. Sirimane, S. (2009). 'Hambantota: way to city hub of Lanka'. Sunday Observer. Colombo, The Associated Newspapers of Ceylon Ltd. September 6 issue. Sirimane, S. (2010). 'Moneragala: in full throttle after kick-start '. Sunday Observer. Colombo, The Associated Newspapers of Ceylon Ltd. March 28 issue. Sirimane, S. (2012). 'Mattala international airport on track '. Sunday Observer. Colombo, The Associated Newspapers of Ceylon Ltd. January 15 issue. Sivathasan, S. (2010). 'Unsustained Paddy Cultivation in Sri Lanka'. Island newpaper. Colombo, Sri Lanka, July 8 issue. SLTDA (2009). 'Annual statistical report of Sri Lankan tourism'. Colombo, Sri Lanka Tourism Development Authority. SLWCS. (2012). 'Sri Lanka Wildlife Conservation Society website.' Retrieved September 3, 2012, from http://www.slwcs.org/.

484 Smith, J., S. Mourato, E. Veneklaas, R. Labarta, K. Reategui and G. Sanchez (1997). 'Willingness to pay for environmental services among slash-and-burn farmers in the Peruvian Amazon: implications for deforestation and global environmental markets'. Joint CSERGE/CIAT/ICRAF Working Paper, CSERGE/CIAT/ICRAF. Sodhi, N. S. (2008). 'Tropical biodiversity loss and people - a brief review'. Basic and Applied Ecology 9(2): 93-99. South African Tourism (2011). '2010 Annual Tourism Report', Government of South Africa. Spenceley, A. (2008). Responsible tourism: critical issues for conservation and development. London ; Sterling, VA, EARTHSCAN. Spencer, J. (1990). Introduction: the power of the past. Sri Lanka: History and the Roots of Conflict. J. Spencer. New York, Routledge: 1 - 16. Sri Lanka Tourism (2009). 'Sri Lanka Tourism strategic plan: 2009 - 2012'. Colombo, Government of Sri Lanka. Sri Lanka Tourist Board (2003). 'Ecotourism Development Strategy of Sri Lanka'. Colombo, Government of Sri Lanka. Sri Lanka Wildlife and Conservation Society. (2011). 'Sri Lanka Wildlife and Conservation Society website.' Retrieved January, 2012. Stem, C. J., J. P. Lassoie, D. R. Lee, D. D. Deshler and J. W. Schelhas (2003). 'Community participation in ecotourism benefits: The link to conservation practices and perspectives'. Society & Natural Resources 16(5): 387-413. Stuart, S. N., J. S. Chanson, N. A. Cox, B. E. Young, A. S. Rodrigues, D. L. Fischman and R. W. Waller (2004). 'Status and trends of amphibian declines and extinctions worldwide'. Science 306(5702): 1783-1786. Sukumar, R. (2006). 'A brief review of the status, distribution and biology of wild Asian elephants'. International Zoo Yearbook(40): 1-8. Sukumar, R. and C. Santiapillai (2006). 'Planning for Asian elephant conservation'. Gajah: Journal of the IUCN/SSC Asian Elephant Specialist Group(25): 9-20. Sunday Times newspaper (2011). 'Chaaya Wild Yala an Epic Adventure Begins'. Sunday Times Sri Lanka. Colombo. November 13 issue.

485 Survey Department of Sri Lanka (2008). National Atlas of Sri Lanka 2007. Colombo, Government of Sri Lanka. Swemmer, L. and S. Taljaard (2011). 'SANParks, people and adaptive management: understanding a diverse field of practice during changing times'. Koedoe 53(2): 1-7. Taylor, R. (2009). The performance of CAMPFIRE in Zimbabwe: 1989 - 2006. Evolution and innovation in wildlife conservation: parks and game ranches to transfrontier conservation areas. H. Suich, B. Child and A. Spenceley. London; Sterling, VA, Earthscan: xvii, 462 p. Thorbjarnarson, J. (1991). An analysis of the spectacles caiman (Caiman crocodilus) harvest program in Venezuela. Neotropical wildlife use and conservation. J. G. Robinson and K. H. Redford. Chicago, University of Chicago Press: 217-235. Tibbotuwawa, M. (2010) 'Better Targeting of Transfers: The Fertilizer Subsidy'. Talking economics August 16 issue. Tietenberg, T. and L. Lewis (2012). Environmental & natural resource economics, 9th edition. New Jersey, USA, Pearson Education, Inc. Tilakaratne, N. and C. Santiapillai (2002). 'The status of domesticated elephants at the Pinnawala Elephant Orphanage, Sri Lanka'. Gajah: Journal of the IUCN/SSC Asian Elephant Specialist Group (21): 41-52. Tipprasert, P. (2001). Elephants and ecotourism in Thailand. International workshop on the domesticated Asian elephant, Bangkok, Food and Agriculture Organisation of the United Nations. Tisdell, C. A. (2005). Economics of Evironmental Conservation, Second Edition. Cheltenham, UK, Edward Elgar Publishing Limited. Tobias, D. and R. Mendelsohn (1991). 'Valuing ecotourism in a tropical rain-forest reserve'. Ambio 20(2): 91-93. Tscharntke, T., A. M. Klein, A. Kruess, I. Steffan-Dewenter and C. Thies (2005). 'Landscape perspectives on agricultural intensification and biodiversity – ecosystem service management'. Ecology Letters 8: 857–874. Turner, R. K. and G. C. Daily (2008). 'The Ecosystem Services Framework and Natural Capital Conservation'. Environmental Resource Economics 39: 25–35.

486 UN Humanitarian Information Center (2005). 'Sri Lanka provinces and districts'. Colombo, United Nations. UN Statistics Division (2008). Official List of MDG Indicators. New York, United Nations. Unanthenna, S. (2010). 'Rajapaksa Style Development'. Sunday Leader. Colombo. May 2 issue. UNEP-WCMC (2008). State of the world’s protected areas: an annual review of global conservation progress. Cambridge, UNEP-WCMC. UNEP-WCMC (2012). 'Protected Areas Statistics'. Cambridge, UK, http://www.wdpa.org/Statistics.aspx. United States Government (1997). 'Asian Elephant Conservation Act'. United States Government. Washington DC, USA. Uragoda, C. G. (1994). Wildlife conservation in Sri Lanka: a history of Wildlife and Nature Protection Society of Sri Lanka, 1894-1994. Colombo, Wildlife and Nature Protection Society of Sri Lanka. Varvasovszky, Z. and R. Brugha (2000). 'How to do (or not to do) a stakeholder analysis'. Health policy and planning 15(3): Health policy and planning. Velázquez, A., E. M. Cué-Bär, A. Larrazábal, N. Sosa, J. L. Villaseñor, M. McCall and G. Ibarra-Manríquez (2009). 'Building participatory landscape-based conservation alternatives: A case study of Michoacán, Mexico'. Applied Geography 29(4): 513-526. Venkataraman, A. B., N. V. Kumar, S. Varma and R. Sukumar (2002). 'Conservation of a flagship species: Prioritizing Asian elephant (Elephas maximus) conservation units in southern India'. Current Science 82(8): 1022-1033. Venter, O. and L. P. Koh (2012). 'Reducing emissions from deforestation and forest degradation (REDD+): game changer or just another quick fix?'. Annals of the New York Academy of Sciences (1249): 137–150. Verma, M. (2000). 'Economic Valuation of Forests of Himachal Pradesh'. London, UK, International Institute for Environment and Development. Veron, J. (2008). 'The demography of South Asia from the 1950s to the 2000s. A summary of changes and a statistical assessment'. Population 63(1): 9-92.

487 Waltner-Toews, D., J. Kay and N.-M. E. Lister (2008). The ecosystem approach: complexity, uncertainty, and managing for sustainability. New York, Columbia University Press. Weeks, P. and S. Mehta (2004). 'Managing People and Landscapes: IUCN’s Protected Area Categories'. Journal of Human Ecology 16(4): 253-263. Weerahewa, J., S. S. Kodithuwakku and A. Ariyawardana (2010). 'Case study #7-11'. Food policy for developing countries: the role of the government in the global food system Ithaca, New York, Cornell University. Weerasinghe, C. (2011). 'First budget with serious proposals'. Daily News. Colombo. November 26 Issue. Weerasuriya, S. (2011). 'Treasure the wonders of ‘The Gathering’'. Sunday Times Sri Lanka. Colombo. October 09 issue. Wenzlhuemer, R. (2008). From coffee to tea cultivation in Ceylon, 1880-1900 : an economic and social history. Boston, MA, Brill Publishers. West, P. and D. Brockington (2006). 'An anthropological perspective on some unexpected consequences of protected areas'. Conservation Biology 20(3): 609- 616. Wickramasuriya, R. C., A. K. Bregt, H. v. Delden and A. Hagen-Zankerc (2009). 'The dynamics of shifting cultivation captured in an extended Constrained Cellular Automata land use model'. Ecological Modelling 220: 2302–2309. Wilkie, D. S. (1989). 'Impact of roadside agriculture on subsistence hunting in the Ituri forest of northeastern Zaire'. American Journal of Physical Anthropology 78(4): 485-494. Wilson, A. J. (1990). The militarisation of Tamil youth. Sri Lanka: History and the Roots of Conflict. J. Spencer. New York, Routledge: 503 - 522. Winchester, H. P. (2005). Qualitative research and its place in human geography. Qualitative research methods in human geography. I. Hay. South Melbourne, Vic. New York, Oxford University Press: xxvii, 339 p. Wittemyer, G., P. Elsen, W. T. Bean, A. C. O. Burton and J. S. Brashares (2008). 'Accelerated human population growth at protected area edges'. Science 321(5885): 123-126.

488 Woodroffe, R., P. Lindsey, S. Romanach, A. Stein and S. M. K. o. Ranah (2005). 'Livestock predation by endangered African wild dogs (Lycaon pictus) in northern Kenya'. Biological Conservation 124: 225–234. Woolf, L. (2011). Leonard Woolf's Ceylon (excerpts from his diaries). The Sri Lanka Reader: History, Culture, Politics. J. C. Holt. Durham, North Carolina, Duke University Press: 269 - 278. World Bank (2009). 'Livelihood survey in the southern region in Sri Lanka'. Unpublished. World Bank (2009). 'Nature-based Tourism and the Human Elephant Conflict in Sri Lanka'. Colombo, Sri Lanka, The World Bank Group: 48. World Bank (2009). 'Valuation of Environmental Services in Sri Lanka: A case study of agriculture and watershed benefits in the Southern Province'. Colombo, Sri Lanka, The World Bank Group: 42. World Bank (2010). 'Project appraisal document: Ecosystems conservation and management project', World Bank. World Bank (2012). 'Moving Beyond GDP: How to factor natural capital into economic decision making'. Washington DC, World Bank. World Bank (2012). 'Sri Lanka: Review of Environmental Impact Assessment System'. Colombo, Sri Lanka, The World Bank Group: 86. WPNS. (2012). 'Wildlife and Nature Protection Society website.' Retrieved September 3, 2012, from http://www.wnpssl.org/. WTTC (2010). 'Travel & tourism economic impact for 2010 (country reports for Botswana, Namibia, South Africa, Sri Lanka, Zimbabwe)', World Travel and Tourism Council. WWF International (2006). 'Species factsheet: Asian elephant'. WWF International. Gland, Switzerland. Yamane, A. (2003). 'Rethinking Vulnerability to Climate Change in Sri Lanka'. 9th International conference on Sri Lanka Studies. Matara, Sri Lanka: 20. Yaron, G. (2001). 'An Economic Analysis of Alternative Land Use Options in the Mount Cameroon Area'. Journal of Environmental Planning and Management 44(1): 85-108.

489 Yatawara, D. (2010). 'A move to solve the human elephant conflict'. Sunday Observer Sri Lanka. Colombo. February 10 issue. Yatawara, D. (2011). 'Survey on elephants vital '. Sunday Observer Sri Lanka. Colombo. April 10 issue. Zubair, L. (2005). 'Modernisation of Sri Lanka's Traditional Irrigation Systems and Sustainability'. Science Technology and Society 10: 161-194.

490