Mammal Diversity, Persistence, and Conservation in India by Krithi K

Home , Conservation in India, List of mammals of India, Nilgiri tahr

Mammal Diversity, Persistence, and Conservation in India

Krithi K. Karanth

Department of Environment

Duke University

Date:______Approved:

______Norman L. Christensen, Supervisor

______Stuart L. Pimm

______Dean L. Urban

______James D. Nichols

Dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Environment in the Graduate School of Duke University

2008

ABSTRACT

Mammal Diversity, Persistence, and Conservation in India

Krithi K. Karanth

Department of Environment Duke University

Date:______Approved:

______Norman L. Christensen, Supervisor

______Stuart L. Pimm

______Dean L. Urban

______James D. Nichols

An abstract of a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophyin the Department of Environment in the Graduate School of Duke University

2008

Abstract

Biodiversity conservation issues are complex and contentious. In this dissertation, I

focus on Indian mammal conservation science, management, as well as policy issues that

shape these factors. I am particularly interested in where and which mammals are extinction

prone, and what factors promote species persistence in human-dominated landscapes. I

examine patterns of extinction, range contraction and current distribution of 25 species of

large mammals in India in Chapters 2 and 3. I apply occupancy models to data from a sub-

continental scale expert opinion survey. I model species occurrence in relation to ecological

and social covariates based on a priori hypotheses about the determinants of mammalian

distribution patterns.

I find that all 25 large mammal species are extinction prone. I find time affects

extinction, and conservation initiatives of the last four decades have allowed some species to

re-colonize some areas. I find protected wildlife reserves are critically important for

persistence of species. Many species with much of their habitat outside existing protected areas will require new protected areas to persist. I find that human population density negatively influences survival probability for species, and human cultural tolerance positively affected persistence of species. Most large-bodied animals, habitat specialists, and rare species had higher extinction probabilities. I find that in addition to protected areas, land use, and human population densities, regionally rooted cultural and religious factors have allowed some species to survive. Conservation strategies must integrate all these factors to ensure the survival of India’s large mammals in the future.

Conservation efforts to protect wildlife in human-dominated landscapes, often requires relocation of people. This policy has rarely been examined in detail. In Chapter 4, I focus on a reserve in India’s Western Ghats of India to assess resettlement experiences of people during and after implementation of a relocation project.

Lastly, the success or failure of conservation policies and management interventions be they for protecting wildlife or addressing needs of local communities, depends substantially on the attitudes of conservation practitioners. In Chapter 5, I examine the attitudes, perspectives and opinions of Indian conservationists towards conservation issues and policies in India.

Dedication

I would like to dedicate this to my grandfather Dr. Shivarama Karanth, the most

amazing person I have known. His accomplishments in so many fields as a writer, novelist, theater and dance artist, and passionate environmental activist, continue to inspire me and are a testament to how extraordinary people can be.

Contents

Abstract ...... iv

List of Tables ...... xi

List of Figures ...... xiii

Acknowledgements ...... xv

1. Introduction and Overview ...... 1

1.1 Introduction ...... 1

1.2 Forest History in India ...... 3

1.3 History of Trophy and Bounty of Mammals in India ...... 4

1.4 Changing Landscapes and Attitudes ...... 6

1.5 Wildlife Protection and Conservation History in British Colonial India ...... 8

1.6 India Today: Conservation Post Independence ...... 12

1.7 Dissertation Outline ...... 14

2. The Shrinking Ark: Large Mammal Extinctions in India ...... 17

2.1 Introduction ...... 17

2.2 Methods ...... 19

2.2.1 A priori predictions ...... 19

2.2.2 Surveyed Species, Sampling Approach, and Data Sources ...... 21

2.2.3 Modeling of Occupancy and Extinction using Covariates ...... 22

2.3 Results ...... 26

2.3.1 Extinction and Elapsed Time ...... 26

2.3.2 Presence and Proportion of Wildlife Protected Areas ...... 36

vii

2.3.3 Proportion of Forest Cover...... 36

2.3.4 Elevation ...... 37

2.3.5 Human Population Density and Cultural Tolerance ...... 37

2.4 Discussion ...... 47

2.5 Conclusions ...... 49

3. Where Are India’s Large Mammals and Why? Patterns and Determinants of Species Occurrence ...... 51

3.1 Introduction ...... 51

3.2 Methods ...... 53

3.2.1 Occupancy Modeling, Expert Opinion Surveys and Study Design ...... 53

3.2.2 Estimation and Modeling of Occupancy and Detection Probability ...... 55

3.2.3 Covariates ...... 57

3.2.4 A priori predictions ...... 59

3.3 Results ...... 60

3.3.1 Proportion of protected area ...... 66

3.3.2 Land Cover ...... 66

3.3.3 Elevation ...... 67

3.3.4 Human Population Density and Cultural Tolerance ...... 68

3.3.5 Estimated Area Occupied by Different Mammals ...... 68

3.4 Discussion ...... 69

4. Making Resettlement Work: The Case of India’s Bhadra Wildlife Sanctuary ...... 85

4.1 Introduction ...... 85

4.2 Materials and Methods ...... 88

viii

4.2.1 Study Site ...... 88

4.2.2 Village Surveys (2002 and 2006) ...... 88

4.3 Results ...... 89

4.3.1 Villages in Bhadra Prior to Relocation ...... 89

4.3.2 Relocation and Resettlement History ...... 90

4.3.3 Resettlement Experiences in M.C Halli and Kelaguru ...... 96

4.4 Discussion ...... 101

4.4.1 Reasons for Success and Emerging Challenges ...... 101

4.4.2 Implications for Bhadra ...... 104

4.4.3 Relocation and Resettlement as a Conservation Tool ...... 105

5. Examining Conservation Attitudes, Perspectives, and Challenges in India ...... 107

5.1 Introduction ...... 107

5.1.1 Background ...... 108

5.2 Methods ...... 114

5.2.1 Survey Methods and Participants ...... 114

5.2.2 Questionnaire Design and Analysis ...... 114

5.3 Results and Discussion ...... 117

5.3.1 General Characteristics ...... 117

5.3.2 Research and Conservation Focus in India ...... 117

5.3.3 Protected Areas, People, and Conservation Threats ...... 119

5.3.4 Conservation Initiatives and Policies ...... 125

5.4 Conclusions ...... 129

6. Conclusion ...... 132

6.1 Chapter Summaries ...... 133

6.2 Future Directions ...... 139

Appendices ...... 141

Appendix 1: Models included in model set for 25 large mammal species ...... 141

Appendix 2: Models included in the model sets for all 20 Species ...... 142

Appendix 3: Models included in the model set for groups of species ...... 142

References ...... 144

Biography...... 144

List of Tables

Table 2.1: Estimated extinction for large mammals in India. Estimated extinction across ψ ψ India Ei= (1- ˆ i ). ˆ i is the occupancy probability for the ith cell. Average extinction across

all cells is E = Σ E i / n, where n is the total number of cells with historic records. Ei100 and Ei 50 refer to average estimated extinction in the last 100 and 50 years respectively……………………………………...... 39

Table 2.2: Best models for predicting occupancy and extinction for each species. Top ranked models where wi is the AIC model weight, N is the number of parameters in the model, and ΔAIC is the difference in values between lowest AIC model and each model……………………………………………………………………………………..40

Table 2.3: Beta coefficients for occupancy that indicate importance of covariates in the top model for all species……………………………………………………………………. ...46

Table 3.1: Environmental and Social Covariates and a priori predictions about their influence on habitat occupancy of large mammals in India ……………………………………….....61

Table 3.2: Best Models for predicting occupancy for each Species. Top ranked models are

shown, wi is the AIC model weight, NPar is the number of parameters in the model, and ΔAIC is the difference in values between lowest AIC model and each model. All models are shown in Appendix 2 and 3……………………………………………………………….62

Table 3.3a: Beta Coefficients for the Top Model for Herbivores………………………….64

Table 3.3b: Beta Coefficients for the Top Model for Carnivores…………………………..65

Table 3.4: Estimated model-averaged occupancy for large mammals in India. The number of cells in which a species was detected = x, and the number of plausible cells within which a ψ species might occur = s. The naïve estimate of occupancy parameter Ψ = x/s. ˆ i is the ψ ψ ψ occupancy probability for the ith cell and Σ ˆ i is the sum of all probabilities. ˆ = Σ ˆ i /s ψ ψ and ˆ IN is Σ ˆ i divided by 1326 (the total number of cells in India………………………70

Table 4.1: Bhadra Wildlife Sanctuary: History and Household Characteristics…………….92

Table 4.2: Project Costs for the Bhadra Resettlement Effort ……………………………..95

Table 4.3: People’s Opinions about the Relocation and Resettlement……………………..97

Table 4.4: Resettlement Package for Households……………………………………….....98 xi

Table 4.5: Comparing Households Resources and Assets………………………………...100

Table 5.1: Demographic Characteristics of Survey Participants…………………………..115

Table 5.2: Summary of Tree Models……………………………………………………...122

Table 5.3: Summary of Participants Comments on Reserves in India…………………….124

Table 5.4: Participants Comments on Conservation Policies and Initiatives……………...128

xii

List of Figures

Figure 1.1: Historic and Current Distribution of Lion …………………...... 8

Figure 1.2: Timeline depicting major events in India’s Wildlife History…………………....11

Figure 2.1: Geographic distribution of 3606 historic records for Tiger (Panthera tigris) in India. Green dots are records pooled into grey cells and red lines demarcate State boundaries .…………………………………………………………………………………………...23

Figure 2.2a: Predicted local extinction for Tiger over 50 years…………………………….28

Figure 2.2b: Predicted local extinction for Tiger over 100 years…………………………...28

Figure 2.3a: Predicted local extinction for Leopard over 50 years……………………….....29

Figure 2.3b: Predicted local extinction for Leopard over 100 years………………………..29

Figure 2.4a: Predicted local extinction for Nilghai over 50 years…………………………..30

Figure 2.4b: Predicted local extinction for Nilghai over 100 years…………………………30

Figure 2.5a: Predicted local extinction for Four-horned antelope over 50 years…………....31

Figure 2.5b: Predicted local extinction for Four-horned antelope over 100 years………….31

Figure 2.6a: Predicted local extinction for Sloth bear over 50 years………………………..32

Figure 2.6b: Predicted local extinction for Sloth bear over 100 years……………………....32

Figure 2.7a: Predicted local extinction for Wolf over 50 years……………………………..33

Figure 2.7b: Predicted local extinction for Wolf over 100 years…………………………....33

Figure 2.8a: Predicted local extinction for Gaur over 50 years……………………………..34

Figure 2.8b: Predicted local extinction for Gaur over 100 years…………………………....34

Figure 2.9a: Predicted local extinction for Chital over 50 years……………………………35

Figure 2.9b: Predicted local extinction for Chital over 100 years…………………………..35

xiii

Figure 3.1a: Predicted Distribution of Nilghai (Boselaphus tragocamelus) and Blackbuck (Antelope cervicapra)…………………………………………………………………………………..75

Figure 3.1b: Predicted Distribution of Chinkara (Gazella bennetti) and Chital (Cervus axis)….76

Figure 3.1c: Predicted Distribution of Sambar (Cervus unicolor) and Muntjac (Muntiacus muntjak)…………………………………………………………………………………....77

Figure 3.1d: Predicted Distribution of Elephant (Elephas maximus) and Wild Pig (Sus scrofa)……………………………………………………………………………………...78

Figure 3.1e: Predicted Distribution of Gaur (Bos gaurus) and Rhino (Rhinoceros unicornis)…....79

Figure 3.2a: Predicted Distribution of Black Bear (Ursus thibetanus) and Sun Bear (Helarctos malayanus)………………………………………………………………………………….80

Figure 3.2b: Predicted Distribution of Sloth Bear (Melursus ursinus) and Brown Bear (Ursus arctos)……………………………………………………………………………………....81

Figure 3.2c: Predicted Distribution of Tiger (Panthera tigris) and Leopard (Panthera pardus)....82

Figure 3.2d: Predicted Distribution of Wild Dog (Cuon alpinus) and Wolf (Canis lupus)…...... 83

Figure 3.2e: Predicted Distribution of Jackal (Canis aureus) and Hyena (Hyaena hyaena)…...... 84

Figure 4.1: Location of Bhadra Wildlife Sanctuary in Western Ghats, India……………...... 90

Figure 5.1: CART model for the question on use of force in protected areas. The text above each split shows the variable that is split, and the condition for the left branch is stated. The original data set has 40 “no” responses and 127 “yes” responses. At the first node, academics (left branch) were equally likely to say “yes” or “no” and non-academics (right branch) overwhelmingly favor “no”. Academics are likely to say “yes” (56%) and non-academics are more likely to say “yes” (82%). Among academics, older participants are more likely to say yes. Among non-academics, those with doctoral degrees are more likely to say “yes”……………………………………………………………………………………..121

Figure 5.2: CART model for question on “success” of Project Tiger. The text above each split shows the variable that is split and the condition for the left branch is stated. At the first node, age was the primary factor. Younger people (< 33.5 years) were more likely to say “no”. Among the older group, at the second node people affiliated with academic institutions and NGOs are less likely to say “yes”. In the older group amateur naturalists and social scientists are more likely to say “yes””……………………………………………..127

xiv

Acknowledgements

Growing up in India, at a very young age I boldly declared that I wanted to also be

called Dr. Karanth like my parents. Luckily, I was unaware of the effort, and commitment

involved. The completion of this dissertation fulfils this bold and naïve declaration, and my

personal desire to complete a Ph.D. Along the way I have interacted with many wonderful people, and benefited enormously from their advice, love, and support.

Norm: thank you for being an extraordinary mentor and advisor. From that very first day of my Ph.D. in Aug 2004, it is your wisdom, demeanor, and support that haves allowed me to complete my dissertation and publish my papers, as well as balance my personal- professional life. Stuart: thank you for pushing me to think “big”, write well, be passionate,

and take our role as conservation ecologists seriously. Jim: thank you for being a wonderful

mentor and collaborator. You have made an invaluable contribution to my intellectual

development as an ecologist by teaching me the importance of using good methods to

address interesting scientific questions. It is has been an absolute joy to work with Hines

and you. Dean: thank you for challenging me to look beyond my questions and methods,

and giving me advice when I needed it the most. All of you have re-in forced my belief that I

need both interesting ideas and appropriate methods to be a good scientist. It has been a

privilege interacting and collaborating with each of you. Most importantly, thank you for

supporting my endeavors to pursue interdisciplinary research in India.

My parents Prathibha and Ullas Karanth have instilled me a passion for my work by

their extraordinary contributions to their own fields. They are incredible parents. Amma, you

are a wonderful mother and terrific role model. You continue to inspire me to give my best.

Appa, thanks for your love and support. Your passion and commitment to conservation has had an unintended consequence, and played a major role in my professional interests and choice of career. I am lucky to share a love of making conservation work in India with you, and I look forward to many more Karanth and Karanth collaborations.

I have benefited tremendously from the support and mentorship of two other advisors. Dr. Michael Binford, my undergraduate advisor, enthusiastically encouraged my early research efforts and cultivated my interest in pursuing interdisciplinary research. Dr.

Lisa Curran, my master’s advisor inspired me with her passion for doing the very best science in relation to real world environmental problems. Her enthusiasm for my work and consistent support has allowed me to overcome challenges that I have faced as a young woman scientist.

I would like to thank other faculty from Duke, whom I have enjoyed interacting with in research and in teaching: Joel Meyer, Randy Kramer, Song Qian, John Terborgh, Ken

Knoerr, and Marie-Lynn Miranda. Other colleagues Jim Hines and Mahesh Rangarajan have supported different aspects of my research.

My favorite aspect of being at Duke has been the amazing friends and colleagues I have met. These wonderful friendships are what sustained me through the highs and lows of completing my degree. Thanks to Sathya Gopalakrishnan, Sean Berthrong, Nicki Cagle,

Libby Davis, Mariana Vale, Valerie Hickey, Nilesh Mistry, Soumya Balasubramanya, Clinton

Jenkins, Claudia Penaloza, Joe Sexton, Iara Diaz, Marion Adeney, Drew and Sara

xvi

Gronewald, Scott Loarie, Jessica Areen, Ariana Sutton-Grier, Varun Swamy, Dawoon Jung,

and Nancy Morgans. I would like to thank members of the Pimm, Christensen, and

Terborgh lab groups for many interesting discussions, and constructive inputs on my work.

I would like to thank other friends and family, who have been with me through this

process. My in laws: Alla and Roman Ostrovsky for their unshakable confidence in me and

love for me, Sumana Shashidhar, Soumya Alva, Girish D. V, Garima Kocchar, Aparna

Shanadi, Sharon Karackattu, Anita Raj, Chris Mangal, Soni Mulmi, Kartheek B.N, Sunita

Thyagarajan, Ajit Thygarajan, and Marsha Ostrovsky. My American family: Fiona, Mel and

Claire Sunquist for all the years at UF and beyond. Linda Goodney for her friendship and

for taking care of Keya so I could work. My charismatic mega fauna- Leia, Obi, and Pips,

whose purrs and company bring great comfort to me every day.

My husband Dennis and our daughter Keya. Dennis thanks for your love, all the

wonderful intellectual discussions, moral support, programming skills, and being my best

friend these last 7 years. Keya, you joined me half way through the Ph.D. and have given my

life so much joy and happiness.

Chapter 2:

I thank wildlife experts N. Akhtar, R. Ali, S. Amu, A. Aiyadurai, Y. V. Bhatnagar, R.

Borges, A. Chandola, S. Chandola, D. Chetry, K. Choudhary, S. Choudhry, H. Dang, S.

Dasgupta, S. Dattatri, A. Dutta, P. S. Easa, D. V. Girish, H. Ghuleria, D. Ghose, S. P. Goyal,

B. Hegde, Hilaluddin, D. Jathanna, B. Jetva, Y. V. Jhala, A. J. T. John Singh, S. Jones, J.

Joshua, K. Kakati, K.U. Karanth, K. Kathju, D. K. Kashyap, R. Kaul, M. Khanduja, J.

xvii

Kulkarni, A. Kumar, N. S. Kumar, S. Kumar, H. Kumara, A. Lobo, P. Mehta, B. Mohanty,

S. Molur, S. Mukherjee, L. Nehemiah, N. Patil, S. Pawar, S. Pradhan, S. Radhakrishna, A.

Rahmani, N. Rajamani, S. Rajesh, S. Ram, G. S. Rawat, G. V. Reddy, V. Rishi, A. D. Roy, P.

K. Sen, K. Sharma, N. Sharma, K. Sathasivam, J. N. Shah, G. Shahabuddin, D. Sharma, V.

Srinivas, G. Sundar, A. Tamim, S. Tiwari, P. Trivedi, H. Tyabji, N. Ved, R. Vyas, R

Wangchuk, T. Wangyal, and their associates who completed the surveys. We thank the

following institutions Bombay Natural History Society; Van Ingen & Van Ingen

Taxidermists; Duke University Library; Harvard Museum of Natural History; American

Museum of Natural History; Carnegie Museum; California Academy of Sciences; Field

Museum of Natural History; Michigan State Museum; University of Kansas Biodiversity

Center; University of Washington Burke Museum; Los Angeles County Museum; Museum

of Natural Sciences; Museum of Vertebrate Zoology; Natural History Museum, D.C;

Museum of Texas Tech University; University of Michigan Museum of Zoology; University

of New Mexico Museum of Southwestern Biology; Museum of Natural History Berkley;

Yale Peabody Museum; Royal Ontario Museum; New Hancock Museum; University of

Manchester Museum; Natural History Museum of Geneva; Hungarian Natural History

Museum; National Museum of Wales (Cardiff); New Castle Museum. We thank J. Van

Ingen, M. Rangarajan, V. Thapar, S. Kapoor, D. Sinh, D. Urban, J. Terborgh, P. M. Kumar,

P. Karanth, A. Ostrovsky, D. Ostrovsky, Duke University, Center for Wildlife Studies, and

Wildlife Conservation Society’s India Program. Karanth received funding from the

Conservation, Food and Health Foundation, Forest History Society, Duke International

xviii

Travel, IDEA Wild, Roger Williams Park Zoo, AZFA Clark Waldram Conservation Fund,

and Cleveland Zoo.

Chapter 3:

I am grateful to the wildlife experts N. Akhtar, R. Ali, S. Amu, A. Aiyadurai, Y. V.

Bhatnagar, R. Borges, A. Chandola, S. Chandola, D. Chetry, K. Choudhary, S. Choudhry, H.

Dang, S. Dasgupta, S. Dattatri, A. Dutta, P. S. Easa, D. V. Girish, H. Ghuleria, D. Ghose, S.

P. Goyal, B. Hegde, Hilaluddin, D. Jathanna, B. Jetva, Y. V. Jhala, A. J. T. John Singh, S.

Jones, J. Joshua, K. Kakati, K.U. Karanth, K. Kathju, D. K. Kashyap, R. Kaul, M. Khanduja,

J. Kulkarni, A. Kumar, N. S. Kumar, S. Kumar, H. Kumara, A. Lobo, P. Mehta, B. Mohanty,

S. Molur, S. Mukherjee, L. Nehemiah, N. Patil, S. Pawar, S. Pradhan, S. Radhakrishna, A.

Rahmani, N. Rajamani, S. Rajesh, S. Ram, G. S. Rawat, G. V. Reddy, V. Rishi, A. D. Roy, P.

K. Sen, K. Sharma, N. Sharma, K. Sathasivam, J. N. Shah, G. Shahabuddin, D. Sharma, V.

Srinivas, G. Sundar, A. Tamim, S. Tiwari, P. Trivedi, H. Tyabji, N. Ved, R. Vyas, R

Wangchuk, T. Wangyal, and their associates who completed the surveys. We thank S. L.

Pimm, D. L. Urban, C.N. Jenkins, M. Vale, D. MacKenzie, N. S. Kumar, P. M. Kumar,

Duke University, Center for Wildlife Studies, and Wildlife Conservation Society’s India

Program. I received funding from the Conservation, Food and Health Foundation, Forest

History Society, Duke International Travel, IDEA Wild, Roger Williams Park Zoo, AZFA

Clark Waldram Conservation Fund, and Cleveland Zoo.

Chapter 4:

I thank D.V. Girish, K. U. Karanth, S. L. Pimm, N. L. Christensen, L.M. Curran, Y.

xix

Kumar, M. Rangarajan, G. Shahabuddin, T. Pandurangaswamy, R. Sathyanarayana, P.

Karanth, and D. Ostrovsky. Yale’s Tropical Resources Institute and FES Summer Internship

Fund (in 2002) and Duke’s International Travel Award (2006) provided financial support.

Yale and Duke Universities institutional review boards approved survey protocols in 2002

and 2006. I appreciate the time that people from Bhadra invested in discussing their lives

and aspirations. I would like to thank the editor and two anonymous reviewers for their

constructive comments.

Chapter 5:

I thank the survey participants, G. Shahabuddin, K. U. Karanth, N. L. Cagle, M.

Rangarajan, C. Clark, J. D. Nichols, P. M. Kumar, N. S. Kumar, and Centre for Wildlife

Studies. Duke University Institutional Review Board reviewed survey protocols. I received

support from a Duke University Graduate School International Travel Award 2005 and

2006.

1. INTRODUCTION AND OVERVIEW

1.1 Introduction

The challenges facing the Indian conservationists include potential species

extinctions, issues of effective protection and scientific management of protected areas and

resolution of human-wildlife conflicts. Due to its of its large and dense human population juxtaposed with a mega biodiversity India provides an ideal setting for examining how other countries may resolve similar conservation problems. India has a well documented colonial and post-colonial wildlife conservation history – periods across which significant changes

occurred in land ownership policy, land use, population declines and range contraction of

wildlife, as well as increased human populations, and cultural change. India presents a

particularly appropriate context for examining these conservation issues in the context of

conserving its rich mammalian fauna.

In this dissertation, I focus on Indian mammal conservation science and

management, as well as policy issues that shape these factors. Globally, large terrestrial

mammals are among the most threatened taxa` in the world, with 25% of species facing

extinction (Ceballos et al. 2005; Schipper et al. 2008). Recent studies suggest that South Asia

harbors the most threatened terrestrial mammals (Schipper et al. 2008). For India, in

particular, conservative estimates suggest that 20% of large mammal species may face

extinction, and several species have already disappeared from over 90% of their original

range (Madhusudan and Mishra 2003). The Indian subcontinent harbors more than 500

mammal species, but also has a ‘modern’ conservation history of regulating land-uses to

protect natural areas that dates back over a century (Blythe 1863; Jerdan 1874; Russell 1900;

Prater 1948; Stebbing 1920; Rangarajan 2001).

In the last 100 years, rapid economic and demographic growth has intensified human

impacts on India’s mammal species and their habitats (Forest Survey of India 2000; Das et

al. 2006). We currently lack basic information on the consequent patterns of range

contraction and the current distribution of mammals. We also need to understand the

ecological and social factors that influence species persistence, if we are to successfully

manage and protect surviving populations of these species in future. Examining the

importance of nature reserves, understanding possible reasons for species persistence under

different land uses, human densities, and human cultural tolerance are critical to species

recovery (Cardillo et al. 2004). I documented the patterns of extinction, range contraction

and current distribution of 25 species of large mammals in India, and examined factors that

influence their occurrence and persistence in Chapters 2 and 3. These chapters are the first to provide accurate extinction and occupancy estimates for large mammal species across the

Indian sub-continent.

Conservation efforts to protect wildlife in human-dominated landscapes such as those in southern Asia, often require relocation of human settlements. The implementation of this policy has rarely been examined in detail. In Chapter 4, I focus on a reserve in India’s

Western Ghats of India to assess resettlement experiences of people during and after implementation of voluntary relocation project.

In the final analysis, success or failure of conservation policies and management interventions, be they for protecting wildlife or addressing needs of local communities, depends substantially on the attitudes of conservation practitioners. In Chapter 5, I examine attitudes and opinions of a representative sample of 167 conservationists in India to gain an understanding of their perceptions, and opinions towards conservation issues in India.

This introductory chapter provides the contextual background that underlies

research questions addressed in Chapters 2 to 5 of this dissertation. I first describe the

history of forest conservation, and trophy and bounty hunting in India. This is followed by a

description of the changes in the land use and human attitudes, and documentation of the

early conservation efforts in India under British colonial rule. Next, I present an overview of

India’s post-colonial conservation efforts. This chapter concludes with an outline of my

dissertation and summaries of data chapters.

1.2 Forest History in India

Variations in the role of forests across time and space, and changes in human cultural

attitudes have greatly influenced in the nature and extent of habitat available for large

mammals. Under the British East India Company’s rule there was little directed forest

management in India and vast tracts of forests remained unmanaged from 1750 - 1850

(Ribbentrop 1900). Although a fledgling forestry system was established in 1806, it

floundered and was abandoned in 1823 (Brandis 1897). Following their victory against rebels

in the first war of Indian independence (1857-1858), the British crown seized political

control and started an official system of timber plantations to provide timber for military and

construction of an extensive railway network (Stebbing 1923 a-c; Rangarajan 1996; Skaria

1998). Initially, these plantations promoted only economically valuable tree species such as

teak (Tectona grandis), sal (Shorea robusta) and sandalwood (Santalum album) (Brandis 1897; Flint

1998). Other species (wattle, rubber, and eucalyptus) were also planted to meet industrial, structural and fuel wood needs of a growing human population (Ribbentrop 1900; Stebbing

1923b). Forest ‘working plans’ modeled after European forestry systems were adopted extensively (Ribbentrop 1900; Rangarajan 1998). These forest management and inventorying efforts centralized the production of large and sustainable timber volumes, generation of

forest cash revenue, and incidentally environmental conservation (Guha 1983; Rajan 1998).

Gradually, forest management transitioned from an unorganized activity to a publicly controlled enterprise with a state monopoly over the commercial value of forests (Skaria

1998).

Government managed forests provided the land base for much of India’s current protected area system. Gradually forest management spread through a network of reserved, protected, and unclassified forests. Forest laws recognized and assigned ownership to the state, to individuals, and to communities. These laws established India’s forest plantations and game sanctuaries. In 1889-90, there were 145,040km2 of reserved forests, and 51,800 km2 of government-protected forests. By 1920, there were 636,912 km2 of government- controlled forests and 332,297 km2 in the native states and under private ownership

(Stebbing 1920). Regulations in state controlled forests sometimes extended into wildlife protection, and some forests were effective at conserving species (Russell 1900; Stebbing

1923a).

1.3 History of Trophy and Bounty of Mammals in India

The British colonial policy encouraged intensive and widespread trophy and bounty hunting of large carnivores, and other species during the 1850s to 1920s in India (Shakespear

1862; Simson 1886; Afalo 1904; Wardrop1923). Carnivore trophy and bounty targets included tigers, leopards, lions, wolves, hyenas, wild dogs, and bears (Sterndale 1887;

Sanderson 1896; Russell 1900; Fenton 1921). Herbivore targets included rhino, wild buffalo, gaur, and several deer species (Baker 1891; Burton 1933). Civil and military colonial officials were encouraged to engage in hunts and exterminate dangerous beasts from large swatches of land to promote agriculture (Dunbar-Brander 1923; Champion 1934). During the years

from 1875- 1925, rewards were paid out for the killing and capture of 80,000 tigers, 150,000

leopards, and 200,000 wolves (Rangarajan 2001).

These colonial policies were superimposed on the complex relationship that local

society had with forests and wildlife in India (Sivaramakrsihanan 1999; Rangarajan 2001).

People often had an ambivalent relationship with crop-raiders; animals were destructive yet provided meat, antlers, and skins (Rangarajan 2001). Some of these species preyed on livestock and, occasionally, on humans (Pollock 1894; Glasfurd 1903; Burke 1920). Local support for these extermination programs varied depending on species. For example, wolves and leopards that preyed on livestock or deer and pigs that raided crops were seen as pests.

Yet some others, such as elephant, Nilghai, Blackbuck, gaur, tiger were culturally and religiously more tolerated (BNHS 1934; Rangarajan 2001).

In many villages, professional hunting castes existed to protect residents, livestock, and crops from wildlife damage (Sivaramakrishnan 1999). Post 1857 local hunters were only able to assist in shoots, and their hunting capacities restricted (Pandian 1998). Extermination programs for large carnivores launched by the colonial government depended on local cooperation, and it was observed that sport targets often did not match villager’s needs

(Rangarajan 1996). Despite established laws restricting use, people often treated government reserved forests as open access resources (Stebbing 1923b). Livestock grazing rules were inconsistent and varied in different forests (Rangarajan 2001). The passage of the keystone law, the 1878 Indian Forest Act, extended the government,s regulatory powers over forest produce exploitation beyond timber to include non-timber forest products, hides, horns, tusks, and skins (BHS 1934; Rangarajan 2001). Punishable offences were expanded to include fishing, shooting, poisoning, and trap setting for certain species and specific seasons.

These restrictions came into conflict with increasing human populations, spread of

cultivation, and unregulated exploitation of forests by people. Consequently conflicts

emerged across the country as forest dwellers and dependents sought to retain their rights of

access and exploitation of forests (Rajan 1998; Rangarajan 1996).

1.4 Changing Landscapes and Attitudes

The above changes in colonial forest and wildlife management policies were

coterminous with massive changes in human numbers, activities and attitudes, with major

consequences for ecological landscape and wildlife habitats. India’s population quadrupled

from 250 million to 1 billion people between 1850 and 2000. Large extents of dry scrub and forested areas were irrigated by major projects and intensively cultivated. Expansion of

agricultural frontiers resulted in shrinking habitats in the Indian plains which forced many

species to become restricted to forests (Russell 1900; Dunbar-Brander 1934; Rangarajan

1998). Tea, coffee and rubber plantations and other agricultural expansions led to the widespread disappearance of several large mammal species from vast regions (Glasfurd 1903;

Stebbing 1920).

Deforestation of any unreserved and unprotected forests occurred on a large scale

(Brandis 1897; Rangarajan 2001). In 1880, 32% of land was cultivated and 32% of land was

under forests. By 1980, 44% of land was cultivated and legal forests had shrunk to less than

20% of land area (Flint 1998). Forests were deliberately burnt for expansion of cultivation

and to improve hunting access (Russell 1900; Stebbing 1923b). A declining ungulate prey

base and proximity of livestock increased carnivore-human conflicts (BNHS 1934), leading

to local extermination of carnivores such as tigers and leopards. Many herbivores competed

with livestock for grazing pastures and palatable forage.

Traditional slash and burn agriculture also severely affected much of the country,

particularly hillside forests of northeastern India (Ribbentrop 1900; BNHS 1934). Reserved

forests protected some forest dwelling species (tiger, wild dog, deer; Rangarajan 1996). Scrub forests often provided fodder for livestock during famines (Brandis 1897; Fenton 1921).

Habitats of scrub and grassland species became fragmented and degraded. For example,

Asiatic lions were originally present in large tracts of open and scrub habitats across India

(Figure 1.1). By the 1930s, habitat loss and sport hunting had restricted lions to only the Gir national park in the western state of Gujarat (Figure 1.1).

There were many other factors that accelerated the disappearance of wildlife. The establishment of railways played a major role in changing the landscape. The railways increased accessibility, and opened up large areas for hunting and cultivation (BNHS 1934).

The expansion of railways accelerated deforestation in many parts of India, with approximately 2000 trees cut for every mile of railroad laid (Rangarajan 2001). Railways opened up vast tracts of land, increased the influx of sportsmen, and promoted an alarming decrease in game (Ali 1934; Champion 1934; Toovey 1987).

Mammals faced intensified poaching pressure due to the arrival of roads, automobiles and fire arms (Jepson 1937). The relaxation of hunting rules during the two

World wars, led to a large increase in guns and gun licenses issued (Champion 1934). By the

1920s, construction of roads and automobiles had opened up large tracts of the country to hunters. Automobiles drastically cut down the search times, requiring less effort, and facilitated nocturnal hunting expeditions (Dunbar- Brander 1934). Trade in mammal body parts also expanded, and hunting became commercialized (Jepson 1937). All of these changes were accompanied by a lack of serious law enforcement, outbreaks of diseases, fires, epidemics, and droughts which hastened the decline of Indian fauna during 1850-1950.

Figure 1.1: Historic and Current Distribution of Lion in India

1.5 Wildlife Protection and Conservation History in British Colonial India

Large-scale extermination efforts and rapidly changing landscape resulted in the widespread disappearance of many large mammals. In 1890, the government announced reduced hunting quotas (Stebbing 1920; BNHS 1934). By 1900, there was a wider interest in protecting species beyond sporting reasons (Pythian-Adams 1934; Rangarajan 2001). Game sanctuaries were established hunting regulations imposed and closed seasons implemented to facilitate recovery of mammal populations. Yet, on ground law enforcement was difficult and ineffective, and hunting was rampant in unprotected and remote areas (Ali 1934;

Champion 1934; Schaller 1967, Gee 1964). Many Indian kings and private landowners were able to protect specific species of interest to them (for example, lion in Gir and tiger in

Mysore; Rangarajan 2001). These native states controlled vast tracts of undisturbed forests and scrub jungle (Ribbentrop 1900). These areas experienced slower conversion of the forests and grasslands into agricultural lands (Sivaramakrishnan 1999) compared to regions directly ruled by Britain. Rulers of native states protected these game reserves, kept out local hunters, monopolizing the hunting opportunities (Ellison 1925). For example, Junagadh state originally established Gir lion sanctuary to protect and allow selective shooting of lions, and today it is the last refuge of the lion in India (Cadell 1934).

By the 1930s, many hunters and naturalists were deploring the rapid disappearance of mammals across India. The restrictions proposed by these conservationists included stricter gun control laws, reducing game hunting licenses, preventing commercial trophy sales, restricting sale of game meat, regulating tanneries, and general trade in animal body parts (especially hides and horns, Monteath 1934). Additional suggestions included reducing hunting rewards, stiffening anti-poaching sentences, protecting stray wild animals, and preventing shooting of wild mammals within a distance of 400 yards from automobiles and carts (Burton 1933; Prater 1948). Other recommendations included increasing public awareness, establishing wildlife protection associations, creating wildlife funds, and establishing new national parks (Champion 1934; Pythian-Adams 1934).

Several wildlife laws (called acts) were passed to protect species (Figure 1.2). These wildlife acts standardized hunting and shooting rules. They identified protected forest boundaries that permitted seasonal hunting of specific species. Fines included monetary penalties and imprisonment. These hunting regulations could not restrict or regulate hunting in native and princely states (Dunbar-Brander 1923). Game animals on private lands were especially vulnerable as the forest service had no jurisdiction on these lands and could not

enforce laws (Champion 1934; Pythian-Adams 1934). Some states passed laws that

prohibited killing of females and young.

The British colonial powers left behind a mixed conservation legacy. From the 1850s

to 1920s, they initially promoted widespread hunting of game animals, but they also set aside

more than 600,000 km2 of government owned forests which could not be easily turned into farm land (Stebbing, 1923). With the expansion of agricultural frontiers, construction of the railways, and establishment of coffee, tea and rubber plantations, many species survived only in these forested areas. They later established hunting reserves that laid foundation for

India’s current protected areas (Stebbing, 1923). Originally, they promoted widespread trophy and bounty hunting of large carnivores, but later passed game preservation laws to regulate hunting. Despite the establishment of reserves and hunting rules, they failed to curtail the disappearance of many species (Champion, 1934), although overall but for the government owned reserved forests they established there probably would have been little room at all for wildlife habitats in India today.

Figure 1.2: Timeline depicting major events in India’s Wildlife History

1.6 India Today: Conservation Post Independence

India is the world’s second most populous country with a population growth rate of

1.3% and economic growth rate more than 8% in the last decade. During last 150 years,

human population density has quadrupled from 80 to 324 people/ km2 (Rangarajan 2007).

Despite economic growth, there is widespread poverty, driven by social disparities: 35% of people live on less than $1 a day and 80% of them on less than $2 a day (UN 2006). The majority (70%) of people live in rural areas. India has been an agrarian country for much of its recorded history and even at present 46% of total land area is cultivated (Rangarajan

2007). Like much of sub-Saharan Africa, South and South-Eastern Asia, a huge proportion

(57%) of the country’s labor force is dependent on agriculture (UN 2006).

India also has ten biogeographic realms (Ref) and is one of 17-mega diversity countries that together support two-thirds of the world’s biological resources (Rodgers and

Panwar 1988; Briggs 2003). Thirty-three percent of the country’s 49,219 plant species are endemic (MOEF 1999). Although, India covers just 2.4% of the earth’s area, it harbors 7.3% of the world’s terrestrial vertebrate species and 89, 451 faunal species (MOEF 2000). India has several charismatic mammal species, including 40% of the world’s tigers, and most of the world’s Asian elephants. Overall, some estimates suggest that 20% of Indian mammal species face imminent extinction, and several have already disappeared from over 90% of their historic range (Madhusudan and Mishra 2003).

In the post colonial period that followed independence, the first effective legislation to protect Indian wildlife was enacted in 1972 – the Wildlife Protection Act (MOEF 2003).

The impetus for this came from the personal commitment of the then Prime Minister Indira

Gandhi, as well as pressure, from Indian and international conservationists (Rangarajan,

2001, 2007). This law banned hunting through severe restrictions as well as ‘commercial’ exploitation of timber and forest products from nature reserves. The Indian government established three categories of wildlife protected areas under this new law, with increasing degrees of protection. These respectively were: game reserves, wildlife sanctuaries and national parks in the 1970-1990 period. In 30 years after independence, land under formal wildlife protection grew from less than 1% to greater than 4% of India’s total land area

(Rangarajan 2006). Wildlife Protection Act sought to protect forests, closed some areas to grazing and collection of forest products, and slowed down the legalization of forest occupation by homesteaders (Saberwal et al. 2001). In 1980-81, the Indian government passed the Forest Conservation Act. This law restricted the diversion of Reserved Forest land for agriculture as well as development projects. Together these laws facilitated the establishment and strict protection of several nature reserves (Karanth 2002). In subsequent decades, conservation efforts have focused on managing the new challenges that have emerged in these reserves (Karanth and Madhusudan 2002; Saberwal and Rangarajan 2003;

Karanth and Karanth 2007; Shahabuddin and Rangarajan 2007). During the 1980s and

1990s, de-centralization of power and failures of the federal and state governments to address local challenges weakened these conservation initiatives and policies.

Today, protected areas cover less than 5% of total land area in India, although they try to protect most of the country’s remaining mammal biodiversity. Indian reserves are under enormous livelihood related but market linked pressure from people residing both inside and surrounding them. These wildlife reserves support many human activities that are

illegal (hunting, fishing, grazing, and collection of forest products) but widespread (MOEF

2000; Forest Survey of India 2000; Das et al. 2006; Karanth et al. 2006). Conflicts between humans and wildlife (crop raiding, livestock killing and rarely killing of humans) do occur, resulting in massive retaliation against many species (Karanth and Madhusudan 2002).

Fragmentation of habitats from both subsistence pressures and economic growth has added a new dimension to the threats to wildlife in the post globalization era. There is increasing concern that India will lose most of its incredibly rich mammalian fauna by the end of this century.

1.7 Dissertation Outline

Against the above ecological and historical background, in this dissertation I examine some key questions of conservation science, management, and policy. This dissertation contains four data papers. Chapters 2 and 3 focus on distribution of large mammal species in

India. Chapter 2 examines extinction and range contraction of 25 large mammal species, and factors that are likely to have driven these processes. I compile and geo-reference 30,000+ records of occurrence of mammal species across 150 years using reliable sources of natural history accounts, taxidermy records, and museum specimen records to reconstruct species- historic occurrence across India.

Using occupancy modeling approaches that overcome major deficiencies of traditional presence or absence survey approaches, I integrate historical and current geo- referenced mammal presence data, to estimate local and regional range contraction and extinction patterns for 25 large mammal species across India. Modeling of habitat occupancy and local extinction processes in relation to covariates such as, time elapsed, presence of

protected areas, amount of forest cover, elevation, human population density, and cultural tolerance of mammals that are likely to drive these processes that affect mammal persistence.

Using these analyses I have generated species-specific extinction rate estimates and range maps. Chapter 2 will be submitted to Conservation Biology/Proceeding of National

Academy of Sciences, USA (PNAS).

Chapter 3 focuses on determining the patterns and determinants of current geographical distribution of 20 species of Indian large mammals. This chapter presents results of testing general and species-specific hypotheses about the relevance of ecological and social covariates to distribution of large mammals. A novel aspect of the chapter is the estimation of species distribution patterns, and associated covariate relationships, using occupancy models that deal probabilistically with false absences. It is the first paper to provide accurate estimates of the distribution of these 20 large mammal species across the

Indian sub-continent. This chapter will be applied in nature and provide information to direct current and future conservation efforts particularly in terms of identifying new areas for on-ground intervention. This chapter is under review in the Journal of Applied Ecology

(Submitted September 2008).

Chapter 4 examines a controversial issue in conservation, the resettlement of people from reserves, using Bhadra Wildlife Sanctuary in India as a case study. Resettlement and relocation efforts worldwide have been criticized for poor planning and execution. Yet, in countries like India where < 5% of land is protected for large conflict prone mammals, if implemented properly, voluntary, incentive driven resettlements could in fact bea practical, win-win solution that meets people’s genuine needs, while greatly enhancing survival of large

mammal species in the face of almost certain extirpation from many reserves. This chapter was published in the journal Biological Conservation (Karanth, 2007. Oct 2007. Vol 139:

315-324).

Chapter 5 examines the attitudes and opinions of the Indian conservation community. Understanding conservation attitudes of different actors from the Indian government, non-governmental organizations, conservation practitioners, scientists, and amateur naturalists is required if we are to make progress on conservation challenges facing

India, and initiate dialog among the people who influence Indian conservation management, policy, and science. This chapter was published in the journal Biological Conservation

(Karanth et al. September 2008. Vol 141: 2357-2367).

I conclude with Chapter 6 which provides recommendations to improve the practice of conservation science, and indicate directions for future research that is critically required.

Each of these chapters arose from my passion and desire to integrate my knowledge of India’s wildlife history with the sciences of conservation ecology and bio-geography. I sought to apply innovative methods in both science and social sciences, and integrate them with rigorous modeling approaches to explore solutions to major conservation problems in a country that has some of the richest wildlife resources, as well as the worst problems, and best potential solutions to address them. The attempt here is to integrate good ecological science and understanding of the social context of conservation in a pragmatic mix that is mostly like to be effective in saving India’s wildlife.

2. THE SHRINKING ARK: LARGE MAMMAL EXTINCTIONS IN INDIA1

2.1 Introduction

Global extinction of species is occurring at an unprecedented rate, (Morrision et al.

2007; Isaac et al. 2008). Large terrestrial mammals are among the most threatened taxa` in

the world, with 25% of species facing extinction and 50% with declining populations

(Ceballos et al. 2005; Schipper et al. 2008). Mammals in South Asia are among the most

endangered (Schipper et al. 2008). Extinction vulnerability is increased by unique life history

traits (body size, litter size), endemism, rarity (low densities), and large habitat requirements

of mammals (Fagan et al. 2001; Brashares 2003; Pimm and Jenkins 2005; Cardillo et al.

2006). Human activities (hunting, harvesting, deforestation, habitat fragmentation, land use change) have resulted in widespread mammal range contractions and extirpations (Ceballos and Ehrlich 2002; Michalski and Peres 2002; Laliberte and Ripple 2004; Morrison et al.

2007). Mammals are critical to maintaining functioning of ecosystems and trophic levels

(seed dispersal, vegetation composition, prey abundance, Terborgh et al. 2001; Wright et al.

2007). Despite their charismatic nature, ecological and economic importance, we lack basic information on the current distribution of many large mammals. Conservation success depends on identifying critical species, and understanding environmental factors that support their persistence in human-dominated landscapes.

We show where mammal extinctions have occurred in India and identify factors that promote persistence of mammals. India has more than 500 mammal species, but many have

1 This chapter was co-authored with J.D. Nichols, K. U. Karanth, J. E. Hines, and N. L. Christensen. 17

disappeared from 90% of their original range (Blythe 1863; Prater 1948; Madhusudan and

Mishra 2003). India’s rich ecological and social history includes the establishment of a

wildlife reserve system (over 100 years ago), and well-documented sport and bounty hunting

(Rangarajan 2001). Changing ecological landscape (deforestation, agricultural expansion), and

rapid economic and demographic growth in the last 100 years have severely affected

mammals and their habitats (Forest Survey of India 2000; Das et al. 2006). We examine

extinction vulnerability of 25 mammals in relation to time, presence and proportion of

protected areas, landscape features, human factors (densities and cultural tolerance), and

species biology.

We compile 30,000+ natural history, taxidermy, and museum records to reconstruct

mammal historic occurrence across India. The reconstruction is partial because we can make

no statement about historic species occurrence in locations for which there are no records.

We conducted an expert-opinion interview survey of more than 100 acknowledged

naturalists to determine current presence or absence of these mammals. Using innovative

occupancy modeling approaches, we integrate historical and current occurrence data to

estimate local and regional extinction of species. Occupancy modeling overcomes problems

posed by imperfect detection probabilities, and allows us to test hypotheses about

determinants of large mammal extinctions (MacKenzie et al. 2006, Ferraz et al. 2007).

Modeling species occurrence relative to covariates (presence of protected areas, forest cover, elevation, human population density, and cultural tolerance of mammals) allows us to identify the key variables that affect mammal persistence, and to generate species-specific extinction estimates and maps.

2.2 Methods

2.2.1 A priori predictions

We evaluated a priori hypotheses about factors influencing probabilities of occupancy

and extinction patterns for 25 large mammals in India. Our predictors of extinction include

factors such as time, habitat - landscape features, likely human impacts, and species biology.

Our basic expectation was of a positive relationship between extinction probability and time elapsed between each historic record and the present. If all local extinctions were permanent, we would expect a simple positive relationship between extinction probability and time

(number of years between 2006 and collection year for each historic record). For example, assume the simple case of time-invariant annual extinction probability (ε). In this case, we

could compute the expected probability of extinction for any number of years, T, as Pr (local extinction in T years) =1− (1− ε)T . However, consideration of the rescue effect (Brown and

Kodric-Brown 1977), re-colonization of locations during the T years, can lead to the event of no extinction for a specific period of observation. We hypothesized that probabilities of such re-colonization might have increased during two periods in Indian history, establishment of forest reserves and hunting restrictions during the 1920s, and establishment of a national protected area system in independent India during early 1970s (Stebbing 1920;

Pythian-Adams 1934; Champion 1934; Rangarajan 2001). These policy changes may have resulted in periods of possible re-colonization, and therefore we fit linear and quadratic logistic models. We generally expected a positive relationship between time elapsed the probability that a species was now extinct. Models in which the beta parameter for time-

squared was negative indicate that the probability of extinction actually decreased with time

elapsed over some period, likely reflecting recolonization.

We expected the presence of wildlife protected areas and the proportion of a cell covered by them to result in higher occupancy rates and lower extinction probabilities for most species (Newmark 1996; Channel and Lomolino 2000; Brashares et al. 2001; Park and

Harcourt 2002; Michalski and Peres 2007). We expected higher proportion of forest cover to be associated with lower extinction probabilities for 13 dense forest-dwelling species (chital,

sambar, muntjac, mouse deer, Nilgiri tahr, gaur, elephant, black bear, brown bear, sloth bear,

wild dog, tiger and leopard). We predicted higher extinction probabilities for grassland and

open forest species because of increased land conversion of such habitats to agriculture

(rhino, wild buffalo, black buck, nilghai, chinkara, four-horned antelope, hyena, and wolf).

We expected lower extinction probabilities for three species (black bear, brown bear, and

Nilgiri tahr) that live at higher elevations because of relative inaccessibility, and lower

likelihood of land conversion of these habitats.

We expected higher extinction probabilities to be associated with higher human

population density (Brashares 2003; Cardillo et al. 2004; Laliberte and Ripple 2004; Pautosso

2007) for all large mammals except 5 species that were either adaptable (wild pig, jackal) or

culturally tolerated (blackbuck, chinkara and nilghai). We expected culturally tolerated species

(nilghai, chinkara, and blackbuck) to have higher persistence probabilities even in the

presence of high human population densities. We expected large bodied herbivores

(elephant, wild buffalo, rhino, and gaur), and large carnivores (brown bear, tiger, and lion) to

be more vulnerable to extinction (Cardillo et al. 2005; Morrison et al. 2007; Schipper et al.

2008). We expected habitat generalists (leopard, jackal, wild pig) to be less vulnerable than

more habitat-specialist species (Nilgiri tahr, brown bear, rhino, and swamp deer). We also

predicted higher persistence probabilities for herbivores compared to carnivores, and lower

persistence rates for crop raiding and livestock killing species (Woodroffe and Ginsberg

1998; Treves and Karanth 2003; Cardillo et al. 2004). We expected endemic species to be

more vulnerable (Ceballos and Brown 1995) in comparison to widely distributed ones. We

predicted that species hunted for sport (tiger, leopard, rhino, wild buffalo, gaur, sambar,

chital, swamp deer, black buck), as well as vermin species (wolf, hyena) to be more

extinction prone (BNHS 1934).

2.2.2 Surveyed Species, Sampling Approach, and Data Sources

We partitioned India into 1326 cell grid, with an average cell size of 2818 km2.

Historical species presence was determined from species location records obtained from 50+ museums, libraries, and taxidermy firms in Asia, Europe, and North America. Imperial and district gazetteers, hunting journals (>150), and all issues of major natural history journals

(Journal of Bombay Natural History Society and Indian Forester) yielded additional records.

These records were used to determine exact geographic locations where species had been either observed or hunted from the year 1850 onwards (few records were between 1760 and

1860). We collected 30,000+ records for over 100 species of carnivores, ungulates, and primates. We selected 25 large mammals with 60+ to 3600+ usable point records (Table

2.1). We pool the individual historic point locations to grid cells to establish species historic presence (Table 2.1, Figure 2.1). Comparing historic and current occupancy patterns allows us to determine local extinctions and range contractions for individual species.

We use expert opinion surveys to determine current occurrence of 25 mammals in

India (Karanth et al. 2008, in review). Occupancy modeling requires only replicated species

presence - absence (in reality species detection - non-detection) records. Therefore, it is

relatively quick and efficient for large-spatial scale studies involving multiple species

(MacKenzie and Nichols 2004). Our occupancy survey conducted in 2006 was based on

elicitations of expert opinions from >100 Indian ecologists, naturalists, and knowledgeable

conservation practitioners. These experts indicated their knowledge of species presence or

absence for all surveyed species in every grid cell. The use of multiple expert opinions

provided the replicates required for estimation of detection probabilities and occupancy

(MacKenzie and Royle 2005, MacKenzie et al. 2006). This approach allows for the inclusion

of missing observations caused by lack of expert opinions or non-response, and permits

unequal sampling effort resulting from varying number of experts responding among

surveyed cells (MacKenzie et al. 2002, 2006). We compiled occupancy data in Excel, and

analyzed using specialized program PRESENCE (v 2.0, Hines 2006).

2.2.3 Modeling of Occupancy and Extinction using Covariates

If species presence or absence is known for each location of interest, we estimate

occupancy as

where x = number of occupied sites and s = total number of potential sites available

(MacKenzie et al. 2006). However, imperfect detections result in x not being unknown. In order to estimate x, we need to account for detection probabilities being < 1 (MacKenzie et al. 2002, Nichols and Karanth 2002). We use multiple observers as ‘replicates’, and

incorporated environmental and social covariates to simultaneously model detection probabilities and occupancy (MacKenzie et al. 2006).

Figure 2.1: Geographic distribution of 3606 historic records for Tiger in India. Green dots are records pooled into grey cells and red lines demarcate State boundaries.

We focus on occupancy of species in a subset of grid cells across India for which we found historic records of species occurrence. We estimate the probability that a historically occupied site is still occupied in 2006. The complement of this “survival” probability thus represents the probability of local extinction. In summary, because we condition present occurrence on locations with historic records, estimates of current occupancy are estimates of the complement of local extinction probability.

We used the maximum likelihood approach to draw inferences about occupancy from detection-non detection data (MacKenzie et al. 2006). We ran 115 models representing different ecological hypotheses about the factors influencing local probabilities of extinction, and detection (Appendix 1). We fit models to the data and calculated model weights using

Akaike’s Information Criterion (AIC, Burnham and Anderson 2002). We used model averaging in situations where there were multiple models with substantial weight to derive weighted averages of probability of occupancy (Burnham and Anderson 2002). The AIC weights sum to 1 for all candidate models, and represent relative measures of the appropriateness of a given model relative to others in model set.

We model occupancy and detection probabilities as functions of covariates using logit link functions (MacKenzie et al. 2006). We select a set of environmental covariates, which are most likely to influence the current distribution of these species, based on our prior knowledge of relevant ecological and social factors. These covariates include elapsed time, presence of protected areas in a cell, proportion of forest cover, elevation, and human influences. For every record, time since collection is calculated. For cells with records from multiple years, we randomly selected a record and its associated year. Information on protected areas is from the World Database on Protected Areas (2007). We use topographic maps, expert opinions, and prior knowledge to improve accuracy of these data. Proportion of cell covered by a protected area has five categories (0%, 1-25%, 26-50%, 51-75%, and 76-

100%). We created two measures: (1) presence or absence of a protected area in a cell, and

(2) proportion of the cell occupied by a protected area (Table 2.1). We use forest cover in a cell is derived from Global Land Cover Facility 2000 (Bartholomé and Belward 2005),

further refined using data from Joshi et al. (2006), and Roy et al. (2006). We derive elevation data from CGIAR-CSI SRTM 100m Digital Elevation Data (2004). We calculated the average elevation in every grid cell, transformed this value to range between 0 and 10. We developed two covariates to represent human influence on species persistence - human population density, and “human cultural tolerance for large mammals”. Human population density data are derived from LandScan Global Population Database 2000 (Dobson et al.

2000, Bhaduri et al. 2002), and log transformed. We use personal knowledge of human hunting patterns, diets, and cultural tolerance or reverence for species in different regions across India (Karanth, U. K., pers obs.; Datta, 2007), to develop a “human cultural tolerance” variable. This variable divided Indian states intro 3 groups, from “most tolerant” to “least tolerant”. The western states of Rajasthan and Gujarat were classified as most tolerant, while seven north-eastern hill states, Chhattisgarh and Jharkhand were classified least tolerant and all other states as medium tolerant.

ψ Our approach to inference yields unconditional estimates of occupancy, ˆi for each

cell i, (for the cells with records of historic species occurrence). Since the historic records

were collected at different points in time, we use coefficient values from top models to

calculate extinction probabilities at two specific times (t=50 and t=100 years ago) for every

cell. This allows us to standardize estimates for different grid cells at the same point in time,

ψ and meaningfully compare cells across India. We calculate local extinction (E) as (1- ˆi ).

These values reflect the probability that a species is locally extinct in a given cell i. Estimates are averaged over all cells to calculate the overall local extinction for each species across the entire historical range. Beta coefficients from the models with highest weights are used to

calculate local extinction probabilities for every species in specific grid cells, and then

ψ ψ averaged across cells to determine species extinction probabilities (E i= 1- ˆ i ) where ˆ i is

the occupancy probability for the ith cell. Average extinction across all cells is = Σ E i / n, where n is the total number of cells with historic records.

2.3 Results

For 25 large mammal species, we use model averaging (2 to 10 top models) to estimate individual site occupancy (Table 2.2). Since these estimates represent the complement of local extinction for sites occupied historically at various times, the point estimates themselves are difficult to interpret. We focus on covariate relationships, particularly how the results match our predictions on factors influencing local extinction probability for mammal species in India.

2.3.1 Extinction and Elapsed Time

Time (number of years between 2006 and collection year for each historic record) was included in the top ranked occupancy model for 20 species, and lower ranked models for four other species. We expected the probability of extinction generally to increase with elapsed historical time, which was true for 11 species (Table 2.3). For example, tiger (Figure

2.2 a, b), leopard (Figure 2.3 a, b), nilghai (Figure 2.4 a, b) and four-horned antelope (Figure

2.5 a, b). For other species, the relationship between extinction and time was negative, and the time-squared term was frequently required in the model as well. For example, sloth bear

(Figure 2.6 a, b), wolf (Figure 2.7a, b), gaur (Figure 2.8 a, b), and chital (Figure 2.9a, b).

We interpret the choice of latter models for the relationship between extinction and time as some evidence of re-colonization. Re-colonization of areas by mammals might have

increased during two periods in Indian history, establishment of forest reserves and hunting

restrictions during the 1920s, and establishment of a national protected area system in

independent India during early 1970s (Champion 1934; Rangarajan 2001).

The actual estimates correspond to the period between each specific historic record

and the year 2006 when present occurrence was determined. We included time (and time

squared) in the modeling as a way of dealing with the relevance of elapsed time to extinction

probability. This was possible because the ‘time related’ covariates provided us means of

scaling these probabilities to periods of 100 years and 50 years. These estimates indicate the

probability of extinction for a given species in a particular cell. Averaging these estimates

provides us with the proportion of cells where species has gone extinct (Table 2.1).

We estimated overall extinction probabilities for 25 species in cells with historic

records at two time periods of 50 and 100 years ago (Table 2.1). For all these species, estimated extinction probability was ≥ 0.20 in both periods (Table 2.1). For all species, during the 100 years, we find extinction probability values to range between 0.20 and 0.96, and 50 years values range between 0.25 and 0.93. We expected higher average extinction over past 100 years than during past 50 years, and this was true for 13 species (Table 2.1,

Figures 2.2 a - 2.9b). In other cases, the differences were small, and be attributed to re- colonization processes.

Figure 2.2a: Predicted local extinction for Tiger over 50 years.

Figure 2.2b: Predicted local extinction for Tiger over 100 years.

Figure 2.3a: Predicted local extinction for Leopard over 50 years.

Figure 2.3b: Predicted local extinction for Leopard over 100 years.

Figure 2.4a: Predicted local extinction for Nilghai over 50 years.

Figure 2.4b: Predicted local extinction for Nilghai over 100 years.

Figure 2.5a: Predicted local extinction for Four-horned Antelope over 50 years.

Figure 2.5b: Predicted local extinction for Four-horned Antelope over 100 years. 31

Figure 2.6a: Predicted local extinction for Sloth Bear over 50 years.

Figure 2.6b: Predicted local extinction for Sloth Bear over 100 years. 32

Figure 2.7a: Predicted local extinction for Wolf over 50 years.

Figure 2.7b: Predicted local extinction for Wolf over 100 years. 33

Figure 2.8a: Predicted local extinction for Gaur over 50 years.

Figure 2.8b: Predicted local extinction for Gaur over 100 years. 34

Figure 2.9a: Predicted local extinction for Chital over 50 years.

Figure 2.9b: Predicted local extinction for Chital over 100 years. 35

2.3.2 Presence and Proportion of Wildlife Protected Areas

Top ranked models for 17 species included either the protected area presence

covariate (3 species), and or the proportion of cell covered by protected area (14 species). All

large mammals except swamp deer thus had either protected area presence or proportion as

covariates in additional selected models (Table 2.2). The beta parameters for these covariates

were positive for most species except nilghai, brown bear, and hyena (Table 2.3), indicating

that protected areas, both in terms of their presence, and proportion of land they occupy are

critical to the survival of 22 species we studied. These results broadly support our initial predictions, especially for large carnivores and other forest dwelling mammal species.

2.3.3 Proportion of Forest Cover

Forest cover was present in the top ranking model for nine species, and in additional highly ranked models of seven other species. Beta coefficients were positive for three species suggesting that forest cover is associated with persistence of these species. Negative coefficients for six species (Table 2.3) suggest that forest cover alone is not sufficient to ensure species survival. As predicted, forest cover was not in the top ranked model for some species (swamp deer, wild buffalo, gaur, rhino, nilghai, and four-horned antelope). Contrary to our predictions, forest cover was not in the top models for some species that we predicted (chital, muntjac, gaur, elephant, black bear, brown bear, and wild dog). Signs of beta parameters differed for species (Table 2.3), indicating that forest cover presence is associated with persistence of some species and extinction for others. Overall, our results suggest that the presence of higher proportion of forest cover is important for 16 mammal species, broadly agreeing with our predictions.

2.3.4 Elevation

Elevation was included in the top ranking model for 11 species, and in additional top

ranked models for nine other species (Table 2.2). The beta coefficients were positive for six

species suggesting that higher elevations are associated with persistence of these species.

Negative coefficients for nilghai, chinkara, rhino, sloth bear, and tiger suggest that elevation is less important for these species (Table 2.2 and 2.3). Our predictions about the importance of high elevation habitats for species persistence were supported for black and brown bear, but not for Nilgiri tahr. These species are restricted entirely to higher elevations. This anomaly may be due to no lower elevation historic records for this species or the use of average elevation values for relatively large grid cells.

2.3.5 Human Population Density and Cultural Tolerance

Human population density (log-transformed value) emerged as an important covariate negatively influencing species persistence in the top ranking occupancy models for

14 species, and in additional top models for 9 other species (except wild pig and Nilgiri tahr,

Table 2.2 and 2.3). We expected higher extinction probabilities to be associated with higher human population densities for most species. Our predictions were correct for most species.

For five species, that we predicted this covariate was not in the top model or had a negative value (wild pig, jackal, blackbuck, chinkara and nilghai). This is likely due to the result of greater adaptability of wild pigs and jackals as well as complete religious protection enjoyed by the three antelope species over large parts of northern India. We find some other species

(four-horned antelope, black bear, wolf, and wild dog) have positive beta coefficients. The deer benefit perhaps from protection like the antelope species above, while the carnivores

come into less contact with people. The beta parameters were negative or not in the top model for most herbivores, and similar mixed results for carnivores. Overall, our results suggest that human population density does affect species persistence, some herbivores adapt better to people than carnivores because of cultural tolerance. However, it is important to note that human density was a factor in our analysis at a large spatial scale – essentially asking if it mattered for species persistence in a cell of 2800 km2. It is likely that a more sensitive spatial scale to measure the influence of human population density might be at the level of sites or reserves with much smaller cell sizes.

Human cultural tolerance was included as an important covariate in the top ranking occupancy models for 12 species (Table 2.2) and included in additional supported models of

9 other species (exceptions was mouse deer, Nilgiri tahr, black bear, brown bear, hyena and lion). Our a priori predictions were met in the case of culturally tolerated species (nilghai, chinkara, and blackbuck), and extended to some other species (four-horned antelope, and jackal) with negative beta coefficients. We find that some species are more culturally tolerated (positive beta coefficients for chital, muntjac, gaur, elephant, wild dog, tiger, and leopard), thus agreeing with our predictions.

Table 2.2 continued

We expected large bodied animals to be more vulnerable to extinctions and these

predictions were met in the case of wild buffalo, rhino, gaur, brown bear, tiger, and lion except for elephant (Table 2.1) which enjoys a high degree of cultural tolerance, and was one of earliest species to be protected by the British which continues today (Dunbar-Brander

1923). However, we also find higher extinction vulnerability in smaller species such as swamp deer, mouse deer, and four-horned antelope, with plausible explanation being their inherently low ecological low densities. Habitat generalist species (leopard, jackal, wild pig) persisted better compared to habitat specialists (Nilgiri tahr, brown bear, rhino, and swamp deer) as per our predictions. Overall, contrary to our predictions, herbivores (extinction probabilities ranged from 0.25 to 0.93) were just as vulnerable as carnivores (extinction probabilities ranged from 0.20 to 0.96). More specifically, however, culturally tolerated or forest dwelling herbivores, as well as adaptable generalist carnivores, were less vulnerable to extinctions.

2.4 Discussion

Determining factors that drive mammalian extinctions and species persistence are critical to successful conservation efforts. Most studies of animal distribution and persistence, however, use methods that distinguish true species absence from simple non- detection during survey. This underestimates true occupancy, which confounds the analyses of spatial data to determine the impact of various ecological and environmental factors on species persistence, survival, and recovery. India’s rich wildlife history allowed us to combine historic records (hunting, taxidermy, and museum) of past large mammal occurrence with a current an expert opinion survey to examine extinction patterns for 25 large mammals in

India over time. By applying modern occupancy modeling, we overcame problems of imperfect detections. There were several novel features in this analysis as we quantified and examined determinants of species-specific extinction patterns.

As predicted, the time-related covariates were important for 24 species. All species were prone to extinction (values ranged between 0.20 and 0.96). Scaled extinction probability was higher or the same over 100 years for 13 species, the same for 4 species, and higher over

50 years for 8 species. Replicate sampling through time would have permitted the use of multi-season occupancy models that permit the direct estimation of time-specific probabilities of local extinction and colonization (MacKenzie et al. 2003, 2006). In the absence of the needed replicate data for all but the terminal year of the analysis (2006), we dealt with time using logistic and quadratic models. Re-colonization of areas supported by conservation initiatives of the last four decades, hunting restrictions (post 1930), and establishment of formal protected areas (post 1970), affects our inferences about temporal variation.

As we predicted, protected wildlife reserves both in terms of presence and proportion of land occupied in the cell are critically important for persistence of 22 species.

This is strongly so for forest-dwelling species, and all carnivores in general. Species with much of their habitat outside existing protected areas (mouse deer, blackbuck, nilghai, chinkara, four-horned antelope, sloth bear, jackal, wolf, wild pig), and species for which only a tiny part of their historic range was protected (swamp deer, wild buffalo, gaur, elephant,

Nilgiri tahr, rhino, black bear, lion, wild dog, tiger), will require new protected areas to persist. We find proportion of forest cover to be important for persistence of 16 species (the

only exceptions were grassland or open habitat species).Higher elevation to be important for

persistence of 11 species and present in top models of 20 species. Human population density

negatively influences survival probability for 14 species and present in models of 23 species.

Human cultural tolerance positively affected persistence of 12 species and present in models

of 19 species. Exceptions were species that were rare localized endemics (Nilgiri tahr, swamp

deer, lion), widely distributed but low density species (mouse deer, four-horned antelope),

generalist species that adapt to human dominated landscapes to some extent (wild pig, jackal,

wolf, leopard), and culturally tolerated or protected species (blackbuck, nilghai, chinkara).

These human cultural factors affect the persistence of both carnivores and herbivores.

Overall, most large-bodied animals, habitat specialists, and rare species had higher extinction probabilities.

2.5 Conclusions

Overall, we find that all 25 large mammal species we studied are extinction prone.

We find time affects extinction, and conservation initiatives of the last four decades (hunting restrictions post 1930, and establishment of formal protected areas post 1970) has allowed some species to re-colonize some areas. We show that protected nature reserve areas are critical for the survival of most of Indian large mammals. India’s current fragmented network of relatively small protected areas (average size < 300 km2), do have high carrying

capacities for large mammals (Karanth et al. 2004). However, given the overall patterns of

species occupancy and extinction we found in this study, creation of new protected areas and

interconnection of more protected areas is necessary for the persistence of most of these species in the long term. This is particularly so for in the context of long-term changes in

climate and human demography. We find that in addition to presence of protected areas, land use, and human population densities, regionally rooted cultural and religious factors have allowed some species to survive. Conservation strategies must integrate all these factors to ensure the survival of India’s large mammals in the future.

3. WHERE ARE INDIA’S LARGE MAMMALS AND WHY? PATTERNS AND DETERMINANTS OF SPECIES OCCURRENCE1 3.1 Introduction

Many animal species now face large-scale range contraction and extinction (Channell

and Lomolino 2000 a, b; Laliberte and Ripple 2004; Cardillo et al. 2005). Human activities

continue to extirpate animal populations and fragment their habitats, thus shaping current

species persistence patterns (Woodroffe and Ginsberg 1998; Sanderson et al. 2003).

Determining where species occur and which species are threatened is a fundamental step in

their conservation. Examining the importance of nature reserves, understanding species

persistence in different land cover - land use types, human densities, and human cultural

tolerance are critical to improving mammal conservation efforts (Habib et al. 2003; Cardillo

et al. 2004).

Large mammals are often keystone species that maintain ecosystem stability and

biodiversity (Terborgh 1988), but they are particularly vulnerable to local extirpations leading

drastic range contractions or even extinctions. Geographical endemism, inherent rarity

exacerbated by human impacts, have now pushed 25% of terrestrial mammal species closer

to extinction (Ceballos et al. 2005). Many species have disappeared from over 50% of their historical ranges (Ceballos and Ehrlich 2002). Their relatively larger body sizes, unique

habitat requirements, and life history traits increase vulnerability of mammals to extinction

(Brown et al. 1996; Brashares 2003; Michalski and Peres 2005). Mammals generally occur at

1 This chapter was co-authored with J. D. Nichols, J. E. Hines, K. U. Karanth, and N. L. Christensen. It is currently under review in the Journal of Applied Ecology. 51

low densities, and they are difficult to observe or tag. This makes it difficult to document

their occurrence and distribution from field surveys particularly at larger spatial scales. The resulting lack of basic information on individual species distributions significantly hinders mammal conservation (Gaston 2000; Brashares et al. 2001; Parks and Harcourt 2002).

Assessments of large scale distributions of mammal species at the regional or sub- continental scale are usually impossible to obtain from direct field surveys of animals or even their signs. Therefore, most surveys (Channell and Lomolino 2000b; Sanderson et al. 2003;

Ceballos et al. 2005) use questionnaires, static range maps or other forms of expert opinion as basic data. We note that even these types of presence-absence surveys are in fact affected by the problem of species being present but going unreported. However, typically such studies tend to ignore this issue, resulting in estimates of habitat occupancy being biased low and its relationship with habitat covariates being poorly understood.

Indian subcontinent harbors more than 500 mammal species, has been occupied by humans for over 50,000 years (Wells 2002), but also has a ‘modern’ conservation history of regulating land-uses to protect natural areas for over a century (Blythe 1863; Prater 1948;

Rangarajan 2001). A countrywide wildlife reserve system was set up in the 1970’s. In the last

100 years, rapid economic and demographic growth has intensified human impacts on mammal species and their habitats (Forest Survey of India 2000; Das et al. 2006; Karanth et al. 2006). Conservative estimates suggest that 20% of Indian large mammals face extinction, and several have disappeared from > 90% of their original range (Madhusudan and Mishra

2003). This highlights the need for basic information on the current distributions of mammals in India and their ecological and social determinants.

We synthesized results from a sub-continental scale expert opinion based survey with

current occupancy modeling approaches to estimate the present geographic ranges of 20 large mammal species in India. We modeled species occurrence in relation to associated ecological and social covariates elucidate causes underlying the observed species persistence patterns. These covariates (protected areas, landscape characteristics, and human influences)

were incorporated directly into the occupancy models in order to test our hypotheses about

plausible determinants of species distribution patterns. We expected presence of protected

wildlife reserves to have positive effects on persistence of many species. We developed

species specific predictions for different forest and other land cover types. We expected

presence of protected areas to have positive effect and human population density to have a

negative effect on the persistence of many species, whereas human cultural tolerance wa

expected to have differential effects on herbivores and carnivores. We generated species-

specific occupancy estimates, tested our hypotheses by confronting our model-based

predictions with the data, and developed distribution maps to identify priority conservation

targets.

3.2 Methods

3.2.1 Occupancy Modeling, Expert Opinion Surveys and Study Design

We used occupancy modeling using data from expert opinion surveys to determine

the current distribution of 20 Indian mammals. Occupancy (Ψ) refers to the proportion of

sites occupied by a species (or the probability that a site in a given area of interest is occupied

by a species). Occupancy surveys are relatively quick and efficient for large-scale studies

involving many species, as they require only basic presence- absence information

(MacKenzie and Nichols 2004). These surveys require less effort and are less expensive than surveys directed at abundance estimation, and they can work well even for rare species

(MacKenzie et al. 2004). Occupancy surveys have been widely used at various geographic scales to study the distribution of many taxa — amphibians, birds, insects and mammals

(Karanth and Nichols 2002; Bailey et al. 2004; Dorazio et al. 2006; Ferraz et al. 2007).

Modern occupancy surveys use temporal or spatial replication to deal with situations where species may be present but are not always seen and recorded by observers (detection probabilities < 1, MacKenzie et al. 2002, 2003). This approach allows for the inclusion of missing observations and permits unequal sampling effort (MacKenzie and Royle 2005;

MacKenzie et al. 2006).

We used occupancy modeling with expert opinion surveys to determine the occurrence and distribution of 20 mammals in India. We used a grid-based sampling approach, and divided the country into 1326 grid cells (average cell size 2818 km2). This cell size was chosen because it allowed us to collect accurate data on species presence-absence as well as covariates. For each species we used existing information to select a subset of cells for which species occurrence was a possibility. Our motivation was to exclude from consideration areas in which the species had never occurred historically and areas that are biogeographically and ecologically plausible (e.g., Prater 1948; Menon 2003). Given the scale of the survey, for logistical reasons our occupancy survey was based on elicitations of opinions of experts based on their experience in the field rather than our own field surveys.

Survey forms were completed by wildlife experts (>100) in India between January and

August 2006. Experts chosen to complete the survey were people very knowledgeable about

particular regions and species. Expert opinions elicited from Indian scientists, naturalists, and conservation practitioners yielded presence – absence information for all 20 species on a countrywide scale. Experts indicated their knowledge of species presence or absence for all species in every grid cell. The use of multiple expert opinions permits estimation of detection probabilities (MacKenzie and Royle 2005). The number of observers (experts) ranged between 2 and 37 per cell. The use of multiple independent observers provided the replicate presence-absence data needed to estimate occupancy. Data were compiled in Excel, and program PRESENCE (v 2.0, Hines 2006) was used to estimate occupancy and detection probabilities for all the species.

3.2.2 Estimation and Modeling of Occupancy and Detection Probability

We can estimate occupancy using multiple approaches (MacKenzie et al. 2002, 2003;

Royle and Nichols 2003). If the number of occupied sites was known with certainty, we would estimate occupancy as

where x = number of occupied sites and s = total number of sites (MacKenzie et al. 2006).

However, x is not known with certainty.

In order to estimate x, we need to account for detection probabilities <1 (Nichols and Karanth 2002). Detectability is a function of the total number of individuals in the surveyed area as well as their behaviors, sizes, and activity patterns, and can be affected by general habitat characteristics and weather conditions (Buckland et al. 2001; Bailey et al.

2004). Therefore, differences in detection probabilities among species will reflect variation in all of these factors (Williams et al. 2002; Royle 2005). Failure to incorporate detection

probability will underestimate occupancy, produce misleading inferences about effects of causal factors, and typically produce inaccurate estimates of dynamic processes (colonization, extinction) that affect occupancy (MacKenzie et al. 2003; Tyre et al. 2003; Gu and Swihart

2004). Here, we used multiple observers, and incorporate covariates into modeling efforts.

Simultaneous modeling of occupancy and detection, allows us to test a wide range of models

(MacKenzie and Nichols 2004; MacKenzie et al. 2006).

We used the maximum likelihood approach to draw inferences about occupancy from detection-non detection data (MacKenzie et al. 2006). This approach was used to fit multiple models to the data, allowing us to compare multiple models representing different hypotheses about the processes that generated the data. The models were ranked in order of parsimony, and model weights calculated using Akaike’s Information Criterion (AIC,

Burnham and Anderson 2002). This approach focuses on the differences in AIC between different models and compares each model to the model with lowest AIC (ΔAIC; Burnham and Anderson 2002, 2004). Models with ΔAIC < 2 have the most substantial empirical support, models with ΔAIC between 4 and 7 have less support, and models with ΔAIC >

10 have very little support (MacKenzie et al. 2006). We used model averaging in situations where there were multiple models with substantial weights. Under model-averaging, AIC weights of candidate models were used to derive weighted averages of individual parameter estimates (Burnham and Anderson 2002). The AIC weights sum to 1 for all members in a model set, and represent relative measures of the appropriateness of a given model relative to other models in the model set.

ψ Our approach to inference yields unconditional estimates of occupancy, ˆ i for each

cell, i, within the area deemed plausible for species occurrence. These values reflect the

probability that cell i is occupied by the focal species. These estimates can then be summed over all cells to estimate the overall proportion of plausible land area occupied by the

ψˆ = ψˆ species: * ∑ i . This value can then be used to estimate the overall proportion of land i

ψ = αψ within India that is occupied by the species: ˆ IN ˆ * , where α = the proportion of land

in all of India that was assigned to be plausible for the species. This proportion, α, was

estimated by dividing the number of cells used in our analysis by the total number of cells

(1326) found in the Indian subcontinent.

3.2.3 Covariates

In most field situations, occupancy and detection probabilities are not constant

across all sample units (Royle 2005) but are influenced by site characteristics. For example,

patch size and isolation (in Ferraz et al. 2007), and habitat variables such as elevation and

vegetation type (in Bailey et al. 2004) may influence occupancy and/or detection. We model

occupancy and detection probabilities as functions of covariates using logit link functions

(MacKenzie et al. 2006). Using the logit link, the logit of the probability of a site being

occupied is expressed as

logit (Ψ i) = β0 + β1 xi1 + β2 xi2 + ……. + βu xiu ,

which is a linear function of the u covariates associated with site i, with one intercept

term β0 and u + 1 regression coefficients that need to be estimated(MacKenzie et al. 2006).

We selected covariates that we thought would be most likely to influence the distribution of large mammals in India. These include protected area designation, landscape characteristics, and human influence. Information on protected areas is from the World

Database on Protected Areas (2007), and data were refined using topographic maps, and expert opinions from India. The data were used to create two measures - presence or absence of a protected area in a cell, and proportion of cell occupied by a protected area

(Table 3.1). For most models, we used the proportion of cell occupied by a protected area

(park prop covariate in model).

For landscape characteristics, we used presence or absence of forest cover in a cell, land cover-land use data, and elevation. Land cover-land use data were derived from Global

Land Cover Facility 2000 (Bartholom é and Belward 2005) and refined using Joshi et al.

(2006) and Roy et al. (2006). We consolidated land cover and land use categories from 23 to

10, and used the total number of pixels in each of these 10 categories for every grid cell).

Simplifying land cover-land use categories was necessary to build a reasonable model set and to ease the interpretation of results (Table 3.1). Elevation data were obtained from CGIAR-

CSI SRTM 100m Digital Elevation Data (2004). We calculated the average elevation in every grid cell, divided by the maximum, and transformed this to range between 0 and 10.

To measure human influence, we used human population density, and “human cultural tolerance” towards mammal species. Human population data were derived from

LandScan Global Population Database 2000 (Dobson et al. 2000; Bhaduri et al. 2002). We calculated the total number of people for every cell, and log transformed this variable. We developed a “human cultural tolerance” variable from personal observations, socio-cultural

knowledge and hunting patterns of local communities in the different states of India

(Rangarajan 2001; Datta 2007; Karanth, U. K., pers obs.). We grouped these States from

most tolerant to least tolerant (Table 3.1). The western states of Rajasthan and Gujarat were

classified as most tolerant, seven north-eastern hill states, Chhattisgarh and Jharkhand were

classified least tolerant, and all other states as medium tolerant.

We ran general models, species group models, and individual species based models.

General models included both additive (models 4, 17-25) and multiplicative models (5)

(Appendix 2). We divided the species into two groups based on general habitat preferences

(Appendix 3). Group 1 had 14 species that are believed to prefer more closed habitat types

(deciduous, evergreen, temperate, and scrub-grassland areas). Group 2 had six species

believed to prefer more open habitat types (scrub-grassland, salt pan, barren, and cultivated

areas). We also ran occupancy models tailored to hypothesized habitat preferences of

individual species. For each species, we estimated the total area currently occupied by

summing occupancy across all cells.

3.2.4 A priori predictions

Modeling covariates enabled us to evaluate predictions of a priori hypotheses about factors influencing probabilities of occupancy and detection. For protected area covariates, we predicted higher occupancy for 12 species that prefer dense habitat cover (Table 3.1).

With regard to land cover covariates, we selected covariates that are ecologically relevant

(Table 3.1, Appendix 2), and divided the species into two groups. Group 1 included 14 species that are predominantly found in closed forest habitat types. Group 2 included 6 species that are predominantly found in open scrub and grassland areas. We also predicted

elevation to affect occupancy, with higher elevation predicted to yield increased occupancy for black, sun, and brown bears and reduced occupancy for the remaining species (Table

3.1). We expected human population density to have a negative relationship with occupancy for most species. The exceptions were 5 species, namely wild pig and jackal that adapt to people, and nilghai, chinkara, and blackbuck that are culturally tolerated by people (only in some areas). For the cultural tolerance covariate, we expected higher occupancy for herbivores to accompany greater tolerance. Models include effects of covariates on occupancy, detection, and both occupancy and detection. Detection probability was predicted to be higher in protected areas, higher in some land cover types for particular species, and lower at higher elevations. Details on predictions are in Table 3.1.

3.3 Results

The global model 13 with all covariates was the top model for three species (chital, leopard, and tiger, Table 3.2). Among the other species, seven had the general models as their top models (general model 12g for blackbuck and wild pig, model 12f for brown bear, sloth bear and tiger, model 12h for jackal and wolf). Seven species (sambar, muntjac, nilghai, chinkara, rhino, black bear, and sun bear) had a top group based model. Two species had individual species models as their top model (gaur and elephant). Model averaging was used to derive weighted averages of individual site occupancy estimates for 14 species, as there were multiple models with substantial AIC weights (Table 3.2).

3.1 Proportion of protected area

Proportion of protected area was included in at least one of the top models for 19

species except hyena (Table 3.2). The beta parameters were positive for most species except

blackbuck, elephant, black bear, and wolf (Table 3.3a and 3.3b), indicating that the

proportion of land set aside as park is important to species occupancy. This agreed with our predictions for forest dwelling species, but extended the inference to open habitat species as well.

3.3.2 Land Cover

Different combinations of land cover types appeared important for different species

(Table 3.2). For 12 species, the models with all land cover types were included in their top models. Sambar, muntjac, nilghai, chinkara, rhino, black bear, sun bear, and wild dog had top group based models with four or seven land cover types. Two species, gaur, and elephant

had top individual based models with three to five land cover types.

Evergreen forest was in the top model for 19 species except nilghai. The beta

parameters for evergreen forest were positive for 15 species and negative for four species

(chinkara, sloth bear, and hyena). These values agreed with our predictions (Table 3.1) for 17

species, the exception being sloth bear. Temperate forest was not in the top model for four

species, namely sambar, nilghai, gaur, and elephant. The beta parameters for temperate

forests were positive for nine species, and negative for seven species. Importantly, the beta

parameters suggest that temperate forests were very important for wild dog, jackal, brown

bear, sloth bear, tiger, and chital (Tables 3.3a and 3.3b), and agreed with our predictions.

Deciduous forests were in the top model of all 20 species. The beta parameters for

deciduous forests were positive for 18 species and negative for two species (elephant and black bear). These values agreed with our predictions for most species. The beta parameter for scrub-grassland habitat was important for all 20 species with coefficient values positive for 12 species, and negative for nine species. The signs of these coefficients agreed with our predictions, but also extended to black and brown bears.

The beta parameters for snow were negative for five species, not part of the top models for 10 species, and positive for five species (Tables 3.3a and 3.3b). Our predictions held for most species. The beta parameters for salt pan areas were positive for seven species

(especially important for blackbuck, chinkara, and wild pig), negative for three species, and not in the top model for ten species (Tables 3.3a and 3.3b). This agreed with most of our predictions. The beta parameters for barren areas were positive for eight species, negative for five species, and not in the top model for seven species (Tables 3.3a and 3.3b). The beta parameters for cultivated areas were positive for 11 species, negative for four species, and not in the top model for five species (Tables 3.3a and 3.3b). Inferences were consistent with most of our predictions.

3.3.3 Elevation

Elevation (transformed and re-scaled) was included in the top models for only 4 species (for chital, hyena, tiger, and leopard) (Tables 3.3a and 3.3b). Elevation was also included in other models that received support for sambar, rhino, black bear, brown bear, and sun bear. This agreed with our predictions for some species.

3.3.4 Human Population Density and Cultural Tolerance

Human population density (log-transformed) was included in the top model for 14 species, in additional supported models for three species (wild pig, rhino, and wild dog). This covariate was not in the top models for blackbuck, black bear, and hyena. The beta parameters were negative for five herbivores (chital, sambar, muntjac, nilghai, and chinkara), and three carnivores (brown bear, wolf, and leopard). The beta parameters were positive for the other six species. This suggests that human population density does affect species occupancy, with some species adapting better to people than others. This agreed with many of our predictions (especially for herbivores).

Human cultural tolerance was included in the top models for 15 species except rhino, brown bear, sloth bear, sun bear, and hyena. The beta parameters were negative for six species, and positive for nine species (Tables 3.3a and 3.3b). Among herbivores chital, sambar, muntjac, gaur, and elephant had positive coefficients. Among carnivore’s black bear, wild dog, wolf, and tiger had positive coefficients. This is perhaps because these species have less contact with people (reside in parks), or are feared by people. Our predictions were accurate for some species, but did not separate out into herbivores and carnivores.

3.3.5 Estimated Area Occupied by Different Mammals

We estimated summed occupancy, and developed current distribution maps for all species (Table 3.5, Figures 3.1a-e, 3.2a-e). For all 20 species the model estimated proportion of cells occupied by a species was higher than the naïve estimate, especially so for sambar, blackbuck, rhino, wild pig, gaur, elephant, black bear, brown bear, sloth bear, sun bear, wolf, wild dog, tiger, and leopard (Table 3.4) Figures 1a, 1d-e, 2a-2d) . The widest occurring

ψ ψ species were jackal (total ˆ IN = 0.78, Figure 2e), and wild pig (total ˆ IN = 0.91, Figure

ψ ψ 3.1d). The most restricted species were rhino ( ˆ IN = 0.04, Figure 1e), and sun bear ( ˆ IN =

0.03, Figure 3.2a). Among herbivores the occupancy values ranged from 0.3 to 0.8, although most species had values between 0.3 and 0.5 (Table 3.5). Among herbivores the most vulnerable species appear to be rhino, gaur, muntjak, and elephant. For carnivores, occupancy values ranged more widely from 0.3 to 0.8 (Table 3.4). Among carnivores, the most vulnerable species appear to be sun bear, brown bear, black bear, wild dog, and tiger.

3.4 Discussion

Range maps exist for most species in most parts of the world and are widely used to investigate hypotheses that deal with geographical ecology and macroecology. The information used to develop such range maps varies widely with respect to such issues as data type (museum records, expert opinion, field survey data) and time scale (accumulation of data over various time periods). Such maps almost never consider the issue of non detection, the idea that a species may be present in an area and go undetected by the method used to develop the map. In comparing the naïve estimate of occupancy to model-derived estimates, we find that failure to incorporate detection probability substantially underestimates overall occupancy for most species (Table 3.4). Thus, most studies that have relied on traditional presence versus absence surveys to map distributional ranges are likely to provide underestimates of true species distributions. In addition, relationships between habitat and other covariates and species occurrence are obscured by analyses that fail to deal with detection probability (MacKenzie et al. 2006).

Table 3.4: Estimated model-averaged occupancy for large mammals in India. The number of cells in which a species was detected = x, and the number of plausible cells within ψ which a species might occur = s. The naïve estimate of occupancy parameter Ψ = x/s. ˆ i is ψ ψ ψ the occupancy probability for the ith cell and Σ ˆ i is the sum of all probabilities. ˆ = Σ ˆ i /s ψ ψ and ˆ IN is Σ ˆ i divided by 1326 (the total number of cells in India).

Species Common x/s Naïve ψ ψ ψ Σ ˆ i ˆ ˆ IN Name Estimate of Ψ Cervus axis Chital 481/1009 0.48 564.96 0.56 0.43 Cervus unicolor Sambar 539/1124 0.48 721.82 0.64 0.54 Muntiacus muntjak Muntjac 430/1096 0.39 505.41 0.46 0.38 Antilope cervicapra Blackbuck 342/970 0.35 558.87 0.58 0.42 Boselaphus tragocamelus Nilghai 518/883 0.57 583.56 0.66 0.44 Gazella bennetti Chinkara 387/883 0.44 437.93 0.50 0.33 Sus scrofa Wild pig 987/1229 0.80 1113.78 0.91 0.84 Bos gaurus Gaur 167/799 0.21 252.04 0.31 0.19 Elephas maximus Elephant 198/963 0.20 340.36 0.35 0.26 Rhinoceros unicornis Rhino 13/163 0.08 50.17 0.31 0.04 Ursus thibetanus Black bear 130/301 0.43 204.20 0.68 0.15 Ursus arctos Brown bear 71/170 0.42 88.27 0.52 0.07 Melursus ursinus Sloth bear 464/1116 0.42 762.37 0.68 0.57 Helarctos malayanus Sun bear 18/131 0.14 43.33 0.33 0.03 Canis aureus Jackal 963/1229 0.78 1030.58 0.83 0.78 Hyaena hyaena Hyena 588/1029 0.57 684.05 0.66 0.51 Canis lupus Wolf 575/1094 0.52 844.47 0.77 0.64 Cuon alpinus Wild dog 211/1106 0.19 330.99 0.30 0.25 Panthera tigris Tiger 249/1189 0.21 433.52 0.36 0.33 Panthera pardus Leopard 647/1229 0.52 904.21 0.73 0.68

Our approach used occupancy modeling combined with data from an expert opinion survey. In addition to providing reasonable species distribution maps, our occupancy modeling permitted us to investigate some of the determinants of species occurrence on the

Indian subcontinent. We estimated the ‘probability of occupancy’ for the 20 mammal species for every cell for which species occurrence was deemed plausible. It should be noted that these probabilities of occupancy do not indicate the proportion of a cell that is occupied, but instead the probability of occurrence for a given species in a particular cell.

Summing these estimates provided us with an estimate of the proportion of all cells occupied for each species.

Many species that were widespread in India just 100 years ago are now restricted to protected areas (Prater 1948). Range contractions have occurred widely (Karanth et al. in prep). Given this background, that the proportion of land set aside in nature reserves

(protected areas) is important for the present persistence of all 20 large mammal species is a critical finding. Protected areas are demonstrated to be important for the persistence of forest-dwelling and open habitat species, wide ranging and regionally restricted species, and for both herbivores and carnivores. Despite their relatively small sizes (average size < 300 km2) and fragmented status, India’s protected areas have relatively high carrying capacities for large mammals (Karanth et al. 2004), and are critical for conserving these large mammals.

These results agree with other studies that showed strong linkages between protected areas and species persistence (Newmark 1995, 1996; Brashares et al. 2001; Gurd et al. 2001).

Therefore, connecting and expanding India’s fragmented nature reserves will enhance

prospects of species survival and persistence. Strengthening existing legal and social support mechanisms necessary should be the top priority for Indian conservation policy makers.

All of the ten land cover and land use types were relevant for the 20 species. Most species occur in a combination of land cover types. All forest types (evergreen, deciduous and temperate) were important to species occupancy, as well as scrub-grassland areas. The

matrix of these land cover types supports species and additionally cultivated areas, snow,

barren and salt pan areas were important to some species. Elevation was relevant to four species (chital, hyena, wild dog, and tiger). As Indian conservation planners establish, expand or prioritize conservation in the future, the ability of the overall reserve system to meet needs of a variety of species requires this sort of consideration of species-specific requirements.

For several species such as the blackbuck, chinkara, wild pig, hyena, jackal, sloth bear, wolf, and leopard a substantial proportion of their current habitat occurs outside

protected areas (Figures 3.1a-b, 3.1d, 3.2b-e). As the ongoing land-use changes in India

accelerate on account of demographic growth, economic and agricultural development, it

will be necessary to protect both open forest and scrub habitat species by establishing

conservation reserves. The planning process for meeting this goal can be initiated rapidly by

using the findings of our study. For some other species like muntjac, gaur, elephant, rhino,

wild dog, and tiger, protected nature reserves are their last refuge (Figures 3.1c-e, 3.2c-d).

Improving protection and monitoring of these areas will be critical to future survival of these

species. Therefore, long term conservation planning will need to strengthen the current

reserve network, and to invest in linking and expanding key reserves and habitat areas. There

appears to be immense scope for doing so through a network of reserves that are

permissible under Indian laws, which permit a variety of local management regimes but can

be used to minimize macro-scale land use changes.

Our study shows that variations in human activities, as well as cultural tolerance of

animals, are important correlates with the persistence of mammals in India. We found that human population density and cultural tolerance are important influences in determining the current occurrence of all 20 species we surveyed. Human population density affected both

carnivores and herbivores. Our data show that such human influences affect distribution of

mammal species in different ways. Species such as blackbuck, chinkara, nilghai, jackal, wild

pig, and leopard are ecologically capable of adapting to human impacts, and actually appear

to do relatively well in human-dominated landscapes. Some species have benefited from

regionally rooted religious or cultural tolerance. These include nilgai, blackbuck, and chinkara

in parts of Northern and Western India, to some extent gaur in central and southern India,

and elephants across their entire range, except the northeastern hills, exemplify this

phenomenon unique to the Indian subcontinent.

We note that persistence patterns of species being primarily linked to human dietary

dependence on them was not clearly discernable from our data, unlike other studies

conducted in Africa, South America and Asia (Alvard et al. 1997; Madhusudan and Karanth

2002; Michalski and Peres 2005; Corlett 2007). Our results support previous studies

(Newmark et al.1995; Woodroffe and Ginsberg 1998; Parks and Harcourt 2002; Treves and

Karanth 2003; Cardillo et al. 2004) that although human demography and land use are

important factors, cultural tolerance plays a very important role in survival of wildlife in

India despite much higher human population densities. It is noteworthy that virtually all across India there has been a total ban on hunting for the last 30 years, and this has reinforced this cultural ethos. Therefore, management policies such as a commercial or licensed hunting may be inappropriate in this context as conservation tools, despite their use in other parts of the world.

India is undergoing enormous economic and demographic growth. As human pressures continue to increase, many mammal species will depend on the presence of protected areas with strong enforcement if they are to survive. This protection of nature reserves must be further strengthened through social acceptability rooted in traditional conservation ethos. Future conservation strategies must carefully harness a blend of protected areas, people’s attitudes, and ongoing land use changes if large mammals are to survive into the future.

Figure 3.1a: Predicted Distribution of Nilghai (Boselaphus tragocamelus) and Blackbuck (Antelope cervicapra)

Figure 3.1b: Predicted Distribution of Chinkara (Gazella bennetti) and Chital (Cervus axis)

Figure 3.1c: Predicted Distribution of Sambar (Cervus unicolor) and Muntjac Muntiacus muntjak)

Figure 3.1d: Predicted Distribution of Elephant (Elephas maximus) and Wild Pig (Sus scrofa)

Figure 3.1e: Predicted Distribution of Gaur (Bos gaurus) and Rhino (Rhinoceros unicornis)

Figure 3.2a: Predicted Distribution of Black Bear (Ursus thibetanus) and Sun Bear (Helarctos malayanus)

Figure 3.2b: Predicted Distribution of Sloth Bear (Melursus ursinus) and Brown Bear (Ursus arctos)

Figure 3.2c: Predicted Distribution of Tiger (Panthera tigris) and Leopard (Panthera pardus)

Figure 3.2d: Predicted Distribution of Wild Dog (Cuon alpinus) and Wolf (Canis lupus)

Figure 3.2e: Predicted Distribution of Jackal (Canis aureus) and Hyena (Hyaena hyaena)

4. MAKING RESETTLMENT WORK: THE CASE OF INDIA’S BHADRA WILDLIFE SANCTUARY1

4.1 Introduction

Nature reserves protect a mere 12% of the world’s land surface (Rodrigues et al.

2004). They are located disproportionately in remote, cold, or very dry places, and protect

only 2% of biodiversity rich tropical forests. Moreover, people who live in tropical forests of

Asia do so at high densities, either within or near nature reserves (Sanderson and Redford

2003). In India, 5 million people live inside nature reserves, and 147 million depend on the

resources the reserves provide according to one source (Kutty and Kothari 2001). Human

densities around these nature reserves often exceed 300 people per km2 (Rodgers et al. 2003).

Despite laws that prohibit hunting, fishing, collection of forest products, agriculture and livestock grazing, these practices are widespread in many reserves (Kothari et al. 1989; Forest

Survey of India 2000; Madhusudan and Karanth 2000; Karanth 2002; Karanth and

Madhusudan 2002; Madhusudan 2004; Barve et al. 2005; Das et al. 2006; Karanth et al. 2006;

Kumar and Shahabuddin 2006). Consequently, conflicts between people and wildlife resulting in livestock loss, crop damage, and injures to people are widespread (Madhusudan and Mishra 2003; Treves and Karanth 2003).

In India, village resettlements to promote conservation date back to the 1960s

(Rangarajan and Shahabuddin 2006). To reduce human pressures on reserves, the Indian government has relocated or tried to relocate people (Shahabuddin and Shah 2003; Sharma

2003; Ganguly 2004; Shahabuddin et al. 2005; Kabra 2006; Rangarajan and Shahabuddin

1 This chapter was published by the Journal Biological Conservation in October 2007 (Vol 139: 315-324).

2006; Karanth and Karanth 2007). Such relocation is highly controversial, is widely perceived to be ineffective, but documented case histories are few. This paper’s objective is to provide one.

Authorities have sometimes forcibly evicted ‘illegal’ settlers in some cases, whereas in other cases, people have voluntarily moved out of nature reserves. Where and if relocation of people is appropriate are questions subject to intense debates in the conservation community. Conservation related resettlement has been widely practiced in Africa, South, and South East Asia, and in North America (Brockington and Igoe 2006). There is criticism of such efforts because of their perceived negative impacts on relocated people (Sato 2000;

Brechin et al. 2002; Wilshusen et al. 2002; Brockington 2003; Chatty and Colchester 2003;

Schmidt-Soltau 2003; Brosius 2004; Rangarajan and Shahabuddin 2006).

Relocated people appear to have lost connections to their culture, history, and identity (Schama 1996; Karisson 1998; Jacoby 2001). They often face loss of economic security and social injustice (Shyamsundar and Kramer 1997; Brockington 2002; Wilshusen et. al. 2003; Bolaane 2004). Other dominant social groups often overpower relocated people

(Bolaane 2004; Rangarajan and Shahabuddin 2006). Governments and non-governmental organizations involved in resettlement efforts rarely consult the people to be moved (Chatty and Colchester 2003; Shahabuddin et al. 2005; Brockington and Igoe 2006). They often lack the experience and the expertise to facilitate such major projects leading to poor planning and hasty execution. Following relocation, governments and non-governmental organizations fail to assess the long-term impacts on relocated people and people have to fend for themselves (McLean and Straede 2003; Rangarajan and Shahabuddin 2006).

As a result, in some cases people have objected to resettlement, sometimes violently, and occasionally moved back into the reserves (Brockington 2004; Brockington and Igoe

2006). In many reserves, there are efforts to prevent relocation. This issue has left the

conservation community deeply divided, and many people are hesitant to promote

resettlement as a conservation tool (Chatty and Colchester 2003; Sanderson and Redford

2003; Brosius 2004; Rangarajan and Shahabuddin 2006). However, the fragile nature of

biodiversity in many nature reserves, ongoing conflicts and the demand from people for

better living standards necessitates the conservation community examine relocation as a

possible conservation solution (Karanth 2002; Karanth and Karanth 2007).

This paper examines resettlement of people out of Bhadra Wildlife Sanctuary, a

biodiversity rich nature reserve in southern India (Karanth 1982; Karanth et al. 2006). People

living inside this reserve faced intense wildlife conflicts: animals raided crops, killed livestock,

and injured and killed people (Karanth 1982, 2003; Madhusudan 2003). They also lacked

basic amenities (no access to electricity and running water, quality health care and schools,

transportation, and communication). Reserve management policies and construction of a

reservoir limited infrastructure development and services in the villages. In the 1970s, some

people voluntarily asked the state governments for help to settle outside the reserve, leading

to the initiation of a voluntary resettlement program (Sreekantaiah and Subramanya 1992;

Karanth 2005). As a result, people from 11 villages moved outside the reserve by 2002.

In 2002, I interviewed relocated households, and key individuals from the government and non-governmental organizations about their experiences. In 2006, I conducted a second follow-up survey to examine changes in people’s lives after relocation and the project’s overall impact. The literature on resettlement has few case studies, they provide limited details, and show a poor understanding of the relocation process

(Brockington and Igoe 2006). This case study provides an opportunity to understand how

relocated people cope, the role of governments and non-governmental organizations, as well

as the costs and time involved.

4.2 Materials and Methods

4.2.1 Study Site

Bhadra Wildlife Sanctuary (Bhadra from here onwards) is located in India’s Western

Ghats (75°15' to 75°50' E and 13°25' to 13°50' N) and covers an area of 492 km2 (Figure

4.1). Vegetation includes dry and moist deciduous forests, evergreen forests, and Shola

grasslands that are unique to this biodiversity hotspot (Karanth 1982). The reserve has > 300

bird species and several threatened mammals: tigers (Panthera tigris), leopards (Panthera

pardus), and elephants (Elephas maximus).

An early British record describes one village in the reserve with 88 people and 186 cattle, occupying an area of 4.19 km2 (Anonymous, unpublished report 1917). During 1956-

66, a major irrigation reservoir was constructed. The reservoir limited access and provision

of basic amenities to some villages (Karanth 1982; Sreekantaiah and Subramanya 1992).

During the monsoons, several bridges and roads would be washed away leaving many

villages isolated for months. Schools did exist in a few villages but with teachers who were

rarely present (Sreekantaiah and Subramanya 1992). There was no hospital or functional

primary health center. In the 1970s, faced with many hardships some families wanted to

move voluntarily outside the reserve (Sreekantaiah and Subramanya 1992).

4.2.2 Village Surveys (2002 and 2006)

I interviewed households in the local languages Kannada and Tulu. The survey

questions were semi-structured and mainly close-ended. The surveys compiled demographic

data, obtained people’s attitudes, and opinions about their relocation experience. The survey

also questioned households on livelihood activities (such as collection of forest products,

grazing, fishing and hunting), along with wildlife conflicts that they experienced. This paper

presents responses by male and female respondents from the resettled households.

I conducted the first survey in July to August 2002 at the time of relocation. I

interviewed 42% to 70% of all households in the 11 villages in Bhadra and the resettlement

villages of M.C. Halli and Kelaguru. These two resettlement villages are in rural areas located

30-50 km from Bhadra. In 2002, I interviewed 257 of 419 (61%) relocated households and

interviewed individuals from the government and non-governmental organizations involved

in the resettlement effort. Repeated attempts were made to interview all relocated

households, but the transitional period in 2002 prevented us from interviewing households

that were temporarily missing in the reserve and resettlement villages. Two villages in the reserve, Pardeshappanamatta and Muduguni, were not part of the relocation effort. I

conducted a second survey from March to April 2006 and interviewed 232 of the 419 (55%) relocated households. In 2006, I found that 75 households were absent in M.C Halli during the follow-up survey. In 2002, the average time households had lived in M. C. Halli was 95 days and Kelaguru was 23 days. In 2006, the average time households had lived in M. C.

Halli was 4.6 years and Kelaguru was 4.7 years.

4.3 Results

4.3.1 Villages in Bhadra Prior to Relocation

In 2002, approximately 4000 people lived in 13 villages inside Bhadra and 63% of households from all villages were interviewed (Karanth 2003; Karanth et al. 2006). These villages were remarkably homogeneous across several major socio-economic and demographic characteristics with few significant differences among them (Table 4.1; additional details in Karanth 2003; Karanth et al. 2006). Some households had legal status, but many had occupied forestland illegally, converting it to agricultural fields and plantations

Figure 2.1 Location of Bhadra Wildlife Sanctuary in India

(Karanth 2003). Households harvested crops, grazed livestock, and collected forest products in the reserve (Karanth 2003; Karanth et al. 2006; Table 4.1). There were intense confrontations between people and wildlife. Tigers and leopards killed 12% of livestock

(Sreekantaiah and Subramanya 1992; Madhusudan 2003). Crop raiding elephants and ungulates destroyed 15% of crops annually (Karanth 2003).

4.3.2 Relocation and Resettlement History

In the 1970s, faced with many hardships some villages wanted to move outside the reserve if offered adequate compensation (Sreekantaiah and Subramanya 1992). The

Karnataka government initially proposed the resettlement project in 1974 but no plans were outlined until 1987 (Y. Kumar, personal communication 2002). From 1987 to 1992, the

government initiated surveys to determine eligible households and outlined project details.

There was a huge time lag by the time the government was able to outline plans and raise

funds required. The delay in execution required a significant increase in funds and more land

was required as the number of people, households, and villages in the reserve had grown. In

1996, the Karnataka government released funds and requested additional funding from the

Central government.

In the 20 years since the original request was made by Hebbe village much had

changed inside the reserve. The number of villages and number of households had increased due to in migration and births (Karanth 2003). At this time, relocation was welcomed by some villages, opposed by one village while the majority of villages adopted a wait and watch policy. In 1999, Madla village filed a case in Karnataka High Court opposing the relocation, and some people set fires that burnt 25 km2 of the reserve. In 2001, the High Court

dismissed the case and judge recommended rapid completion of the project. At the same

time, approval and consent for the project was gained from many households due to efforts

of the local conservation non-governmental organizations (Bhadra Wildlife Conservation

Trust, Wildlife First! and Nature Conservation Guild), the forest, and revenue departments.

Table 4.1: Bhadra Wildlife Sanctuary: History and Household Characteristics

History of 1. Reserved forest in 1912 - 1950 Establishment 2. Jagara Wildlife Sanctuary in 1951 (area 200 km2) 3. Bhadra Wildlife Sanctuary in 1974 (Shyamsunder 1974) 4. Project Tiger Site in 1998 Villages 11 out of 13 relocated Cultural Hindu (97%) and Islam (3%). 12 castes (numerically dominant were Composition Shettys, Gowdas and Schedules Castes, composition varied significantly between villages, chi-squared test p < 0.0001). Households Average time in reserve 42 years (Range 3 to 70 years, similar across villages, p > 0.05). Average household size 5 (Range 1 to 15, varied significantly among villages, p < 0.03). Males headed 81% (Range 63% to 90%). Household heads born outside the reserve was 62%. Income Agriculture primary source of income for 96% of households. Other sources: wage labor, non-timber forest products, livestock (mainly cattle), home-gardens, hotels, and shops. Land Tenure Households with legal land rights was 68% (varied significantly between villages p < 0.005). Households renting additional land 35% (similar across villages, p > 0.05). Households encroaching forestland 32% (varied significantly between villages, p < 0.0000001). Encroachment was higher in remote villages and in villages with no park guards (Karanth 2003; Karanth et al. 2006). Agricultural Average time household had farmed 32 years (Range 1 to 90 years, Practice varied significantly between villages, p < 0.03, N = 222). Average distance to field 0.4 km (Range 0 to 3 km, differed among villages p < 0.025, N = 219). Average plot size (owned and/or encroached) 1.3 hectares (Range 0.2 to 8 hectares, N = 228). Crops grown rice, coffee, betal nut, and rubber. Other Major Livestock were cattle, sheep, pigs, and goats (grazed in the reserve by 94 Livelihood - 100% of households). Cattle owned by 80% of households and Sources average per household was 8 (Range 1 to 50 per household). Wildlife hunted Sambar (Cervus unicolor), muntjak (Muntiacus muntjak), wild pig (Sus scrofa), hare (Lepus nigricollos), giant squirrel (Ratufa indica), monitor lizard (Varanus flavescens), and jungle fowl (Gallus lafayetii). People hesitant to discuss gun ownership and hunting practice (as it is illegal inside reserves as per Indian law). Licensed guns owned by 7% of households and unlicensed guns hard to establish. Fishing pursued by 74% of households (using dynamite, traps, nets). Human- Crop raiding affected 93% of households (N= 230). wildlife Livestock predation affected 80% of households (N = 201). The Conflicts average cattle lost was 2 per household/ year (varied between villages, N = 201, p= < 0.00001). Livestock predation rate 12% to 25%. Note: N= 257 except where indicated.

From 1998 to 2001, the non-governmental organizations made a concerted effort to improve their relationship with people, and build consensus for relocation among differing factions and the different villages. These organizations continually sought to address people’s concerns, fears, and problems inside the reserve, and ensure that the forest-revenue departments continued to move forward with plans for resettlement. The forest department constructed and improved bridges, roads, and a better water supply system for villages inside the reserve (Y. Kumar, Divisional Forest Officer, personal communication 2002). Over time, the continued support from the local non-governmental organizations and forest-revenue departments convinced people, and many people who were unsure about relocating changed their minds (D.V. Girish, personal communication 2002; Y. Kumar, personal communication 2002). Households felt reassured when they received compensation for land and houses from the government and chose to move. Project implementation began with involvement of village representatives, and non-governmental organizations, and the forest and revenue departments of Chikmagalur and Shimoga districts. Despite a decade long delay in the execution of the project, the land and money remained available due to efforts of the

Project Tiger steering committee (U. Karanth, personal communication 2002).

Official implementation began in October 2001, and households received land deeds in the two-resettlement villages M. C Halli and Kelaguru. The government provided every household with land, and an individual housing site. Originally, 334 hectares of land were set aside for resettlement in the village of M. C. Halli. The delays in the project led to the encroachment of some of this land by people already living in M. C. Halli village. This required purchase of additional land at a second site, the village of Kelaguru. The first village, M.C. Halli, covering an area 304 hectares became home to 373 families (with 373 agricultural plots and individual housing sites). Many households in the reserve had

encroached forestland, cultivating crops and constructing houses and did not possess legal

land deeds (Sreekantaiah and Subramanya, 1992; Karanth, 2003). Although these households

were legally not eligible for compensation, the government decided to provide many of them

land and houses in M. C. Halli. In M. C. Halli, the Rajiv Gandhi Housing Corporation

constructed 203 houses (between November 2001 and June 2002) for poorer landless

villagers at a cost of 42,000 rupees (~ US $900) per house. Wealthier households were

required to construct their own houses. The second village Kelaguru consisted of 186

hectares of revenue land allocated to the remaining families.

Land provided in M.C. Halli was best suited for growing crops such as sugarcane and rice, while in Kelaguru land was best suited for coffee. Therefore, households in Kelaguru received more land compared to the households in M. C. Halli. Households received financial compensation for dismantling their houses in the reserve, and transportation costs to the resettlement villages. All households received a free housing site free, and a subsistence allowance for the first six months following relocation. This allowance helped defray costs of food, fuel wood, and fodder. Both villages have electricity, pumped water available to every house, as well as access to public transportation, better communication facilities, several markets, schools, hospitals, and major cities. People in M.C. Halli were actively involved in the decision-making process (their suggestions on the design and construction of houses were incorporated) and requests such as gas stoves accommodated by the revenue department.

Recent developments (2002-2005) include a public water system, electricity, phones, television, solar lights, and gas cooking stoves in both villages. All households received housing compensation and dismantled their houses in the reserve (Y. Kumar, personal communication 2002). In addition, there was a new school constructed M.C. Halli and health

camps are held regularly (D.V. Girish, personal communication 2004). School enrollment

has increased with the active participation by the children.

The cost of land acquisition was originally proposed as 7.02 crore Rupees (~ US $

1.59 million) in 1992 and revised to 13.15 crore Rupees (~ US $ 2.98 million) in 1999

(Karnataka Forest Department 1999). The relocation and resettlement cost was originally

proposed as 5.74 crore Rupees (~ US $ 1.30 million) in 1992 and revised to 8.07 crore

Rupees (~ US $ 1.83 million) in 1999 (Karnataka Forest Department 1999). Costs associated

with different project components are in Table 4.2.

Table 4.2: Project Costs for the Bhadra Resettlement Effort

Costs in Crore Rupees Estimated in Revised in 1999 1992 Land Acquisition 1. Value of land, buildings, trees and 2.92 crores 8.63 crores (US $) plants 0.93 crore 3.32 crores(US $) 2. Solatium 3.17 crores 1.19 crores (US $) 3. Contingency costs and administrative overhead 7.02 crores 13.15 crores TOTAL ($ 1.59 million) ($ 2.98 million)

Resettlement Costs 1. Payment for displaced families 1.65 crores 1.98 crores 2. Site costs 0.15 crores 0.26 crores 3. Infrastructure construction costs 1.46 crores 3.03 crores (schools, roads, community center, water, electricity, health center) 4. Transportation costs 0.11 crores 0.40 crores 5. Cash assistance for land development 0.15 crores 0.18 crores 6. Contingencies and administrative 0.88 crores 0.58 crores overhead costs 7. Environmental costs 8. Monthly subsistence allowance per 0.10 crores 0.15 crores family (6 months) 1.23 crores 1.48 crores TOTAL 5.74 crores 8.07 crores ($ 1.30 million) ($ 1.83 million)

4.3.3 Resettlement Experiences in M.C Halli and Kelaguru

During the interviews, people discussed good and bad aspects of their experiences,

as well as the uncertainties they faced (details in Table 4.3). There were differences among

individuals, households, and villages with respect to their willingness to relocate. Some villages were more willing to relocate, and within villages, some households were more eager

to relocate. Larger landholding households were often the least willing to relocate. These

households felt insufficiently compensated, and thought that relocation decreased their

social status and dominance over other households. These wealthier households also

expressed frustration that all households now enjoy the same facilities (phone, television etc.)

and access to better resources.

There are differences between the two-resettlement sites and people’s experiences in

the two villages. In M.C. Halli, all households received fertile and well-irrigated land.

Therefore, these households were able to cultivate their first crop within 3 months of

moving. Adequate water supply allowed households to diversify crop production (sugarcane,

rice, fruits, and vegetables) within 3 to 6 months of settling in M. C. Halli. In Kelaguru,

households received larger plots with soil suitable for growing coffee. Since coffee takes

longer to grow, many households in Kelaguru had to find employment outside to support

themselves. These households are also vulnerable to fluctuating coffee market prices.

Initially, some households from both villages adopted a double-farming approach, they

continued to cultivate crops in both the reserve and new villages, and this ensured that

people did not face much loss of livelihood security. This caused significant stress between

them and the forest and revenue departments who were trying to complete the relocation

and resettlement process (Y. Kumar and D.V. Girish, personal communication 2002).

Table 4.3: People’s Opinions about the Relocation and Resettlement

Positive Aspects 1. Fertile land provided (especially in M. C. Halli) with access to ample irrigation facilities 2. Diversify crop production and increase in number of harvests/year. 65% of households cultivate sugar cane and 51% of households in M.C. Halli cultivate rice 3. Absence of crop raiding elephants and ungulates 4. Absence of livestock predation and loss of human life 5. Increase in comfort due to availability of electricity, pumped water, gas stoves, television, and telephones 6. Proximity to towns and access to markets 7. Increase in contact with family and friends due to better communication facilities. 8. Improvement in the quality of education for the children and no disruptions. One 9. centrally located school compared to three schools in Bhadra that required children walking 2-12 km in the reserve. 10. Proximity to hospitals and improved health care. In Bhadra, people walked many kilometers to reach buses to take them to a hospital. 11. More time available to pursue other professions and hobbies. 12. Households headed by women or lower castes did not face discrimination.

Negative Aspects 1. Limited access to firewood and costs associated with gas stoves. 2. Limited livestock, many sold livestock due to limited grazing land around new villages and absence of freely available fodder. People switched to owning goats and did not like purchasing milk. In M.C. Halli livestock ownership decreased from 87% to 40% of households. In Kelaguru livestock ownership decreased from 100% to 40% of households. 3. Ability to find work and reluctance to work as labor in the new villages 4. Time required to adapt to new environment and tensions with people who previously lived in M. C. Halli. 5. Some landless households given one acre were dissatisfied about the plot size and their inability to acquire additional land. 6. Importance of money- now required to purchase many basic commodities that they had unlimited access to and could barter in the reserve (especially non-timber forest products and fuel wood).

In M. C. Halli, households have successfully grown many crops and sold them in local markets (Table 4.5). Some households have opened restaurants, grocery stores, and shops that serve the new community. Women united to form an association that tackled and successfully dismantled an illicit liquor shop that opened in M. C. Halli (Pandurangaswamy,

personal communication 2002). There have been several weddings between families from different villages. Initially, people in M. C. Halli felt some hostility from surrounding village households that had encroached relocation land set aside by the government but overcame these challenges. People from M. C. Halli contested elections, and were elected to the local

Panchayat (D.V. Girish, personal communication 2005). In the short and mid-term, households in Kelaguru have faced greater hardship than in M. C. Halli and it will be important to follow how they cope in the long-term.

Table 4.4: Resettlement Package for Households

Relocation M.C. Halli Kelaguru Village Land (in Hectares) 0.4 to 2.02 0.81 to 4.04

Compensation Money Yes for all households Yes for all households

Housing Sites 40 X 50, 50 X 80, 60 X 90 50 X 80, 60 X 90

House built 203 constructed by Rajiv Gandhi 46 self constructed Corporation at cost $900 per house (for landless households) 170 self constructed Crops Grown Sugar cane, Rice, Vegetables Coffee, Pepper

Access to Towns 10 km to towns 12 to 18 km to towns

Facilities Schools, colleges, primary health Schools, primary health center, and hospitals. Electricity, center and hospitals. gas stoves, solar lights, pumped Electricity, gas stoves, solar water, phones and television lights, pumped water

There has been a decrease in overall approval for the project from 71% to 52%. In

2002, 70% of households in M. C. Halli and 81% in Kelaguru considered the relocation project a success and indicated that it improved their quality of life. In 2006, 53% of households in M. C. Halli and 41% in Kelaguru considered the relocation project a success and indicated that it has improved their quality of life. There are several possible reasons for

this. Most people agreed that their standard of living had improved. They are also happy with all the facilities and access to resources (Table 4.5). Many observed that it was taking them time to get used to the new environment. People acknowledged the significant improvement in their assets (Table 4.5). This increase in assets was more evident for M. C.

Halli households than for Kelaguru (Table 4.5). During the follow-up survey 75 households were absent in M.C Halli. These households were able to sell or rent their property and move to other towns or villages. People acknowledged the improvement in lifestyle and significant economic benefits. There is a gradation in people’s satisfaction with the project as supported by Tables 4.3 and 4.5. This is not captured by a single question on overall satisfaction with the project. The various nuances and dimensions presented in Tables 4.3 and 4.5 are not captured by the use of a single statistic in either year. There was a tendency among people to emphasize the negatives aspects during the 2006 survey. Their dissatisfaction focused mainly on the inadequate financial compensation provided. Their statements regarding amount of compensation provided to them often contradicted the official government records.

I observed that in 2002, people’s statements were more forthright and less influenced by external views. In 2006, people’s opinions appeared influenced by living in the new village society (particularly in M.C. Halli). In the follow-up survey, it was evident that people had become used to speaking to reporters and others due to the publicity that this project has generated. People often asked if I would provide them financial compensation for participating in the interviews.

Table 4.5: Comparing Households Resources and Assets

Characteristics M.C. Halli ( 200 Households) Kelaguru ( 32 Households) Houses 91% of households have built 90% of households have houses. built houses and all are self- 69% of houses built by Rajiv constructed. Gandhi Corporation and 31% self- constructed. Time in new village Average: 4.57 years Average: 4.7 years

Plot Size Average: 0.72 hectares Average: 1.86 hectares

Compensation Average: 46,044 Rupees (US $ 996) Average: 57,143 Rupees ( US $ 1236)

Family Structure 81% were single families and 19% 78% were single families joint families and 22% joint families

Household Facilities in 100% have access to schools, health 100% have access to the Relocated Villages care, electricity, water, towns, roads schools, health care, and transportation, markets and electricity, water, towns, phones roads and transportation, markets and phones

Households in the 4% had access to schools 3.5% had access to schools reserve 0% had access to health care center 0% had access to health care 1% had access to electricity center 96% had access to water 3% had access to electricity 5% had access to towns 100% had access to water 4% had access to roads and 0% had access to towns transportation 0% had access to roads and 9% had access to markets transportation 1% had access to phones 0% had access to markets 17% had injury or death due to 0% had access to phones confrontations with wildlife 97% had injury or death due 81% lost livestock to tigers, to confrontations with leopards and wild dogs wildlife 63% lost crops to elephants and 97% lost livestock to tigers, wild pigs leopards and wild dogs 18% hunted wildlife and 39% fished 97% lost crops to elephants and wild pigs 0% hunted wildlife and 36% fished

100

Percentage increase in Cycle: Increased from 15% to 34% Cycle: Increased from 19% households owning asset Two wheelers: Increased from 10% to 39%, from 2002 to 2006 to 15% Two wheelers: No change Three wheelers: Increased from 0% 13% to 1% Three wheelers: No change Four wheelers: Increased from 1% 0% to 1.5% Four wheelers: Decreased Tractor: Decreased from 3.5% to < from 10% to 6% 1% Tractor: Decreased from Plough: No change 7% 10% to 0% Pump set: No change 6% Plough: No change 20% Refrigerator: Increased from 1% to Pump set: Increased from 3.5% 3% to 6% Stoves: Increased from 2% to 54% Refrigerator: No change 0% Phones: Increased from 3% to 20% Stoves: Increased from 0% Radio: Decreased from 64% to 63% to 48% Television: Increased from 4% to Phones: Increased from 0% 58% to 16% Radio: No change 81% Television: Increased from 6% to 29%

4.4 Discussion

4.4.1 Reasons for Success and Emerging Challenges

The Bhadra relocation is now complete and considered a success by the government, non-governmental organizations, and many relocated people. There are many important reasons for this success. There was no forcible eviction of people from Bhadra, with some people voluntarily choosing to relocate and others choosing to relocate once they received fair compensation. The hardships (especially wildlife conflicts) faced by people living in the reserve, the opportunities available to them (particularly their children), facilities, and equitable land tenure appear to be the strongest reasons for their willingness to relocate.

Active participation by the non-governmental organizations and government ensured that the households received adequate support throughout the process.

Positive aspects include active consultation of households, choice of suitable relocation sites (particularly M.C. Halli). The distribution of fertile plots and irrigated land to

101

all families, appropriate housing, financial compensation, provision of adequate health care, schools, transportation and communication, access to electricity, water and markets- all facilitated the project’s success. Equitable distribution of land titles and other development aid (houses, water supply, electricity, solar lights, and gas stoves) convinced people to relocate. In contrast, people relocated from Indian reserves such as Kuno and Gir received poor agricultural land and limited developmental aid (Sharma 2003; Ganguly 2004; Kabra

2006). Importantly, both the non-governmental organizations and the government continually interacted with people and actively sought their opinions. Following relocation, loss of economic security did not affect people in M. C. Halli, but in Kelaguru, the long-term consequences for people are still unknown. Financial compensation and other support were provided to people during the transitional phase unlike other relocations (Rangarajan and

Shahabuddin 2006). The resettlement package offered to households in Bhadra cost three to four times more than packages offered in other reserves (Table 4.4; Karanth 2005).

People in Bhadra are primarily cultivators and post relocation did not have to alter their livelihoods. They remain agrarian with access to better facilities, with opportunities to improve crop productivity and diversify production (Table 4.4 and 4.5). They were not nomadic or hunter-gatherer communities that are more dependent on the reserve, and perhaps less able to undergo rapid socio-cultural transformations that are often required by such resettlement efforts. In Kanha and Gir National Parks, relocated people were unable to adapt from nomadic hunter-gatherers to settled agriculturalists and wage laborers

(Rangarajan and Shahabuddin 2006). People from Bhadra are not indigenous groups or tribal communities that lived in the reserve for many generations. Some people moved into the reserve as labor for British teak plantations about 85 years ago, and most migrated from neighboring towns and cities in the last 30 years (Karanth, 2003). No single social group was

102

able to exert dominance in both M. C. Halli and Kelaguru (Karanth 2003, 2005), and people

integrated well into the larger society. In fact, women and weaker groups in particular seem

to have benefited from the move. In Kuno and Gir, dominant social groups overpowered

people from marginal social groups (Kabra 2003; Rangarajan and Shahabuddin 2006).

This project’s success is attributable to the commitment of key individuals in the

non-governmental organizations working together with the government. In many Indian

reserves, only the government carried out relocation and resettlement projects. In Bhadra,

non-governmental organizations were actively involved and ensured that the government

delivered on its promises to the people. These organizations negotiated and resolved

problems that arose and ensured that compensation efforts proceeded smoothly. Regional

differences also play a role in success or failure of resettlement projects. The southern Indian

state of Karnataka is a state with greater social safety services than North or Central Indian

states where relocations have taken place (Rangarajan and Shahabuddin, 2006). Hence,

perhaps regional political and social conditions also contributed towards facilitating the

planning and execution of this project.

The Bhadra project appears to have overcome many hurdles, but it also faces new

challenges. Differences between the two-resettlement villages have emerged in the last 4

years. Resources available in M.C. Halli are better than in Kelaguru. In this case, due to the

encroachment of land in M.C. Halli there was no choice but to move some households to

Kelaguru. It is best to provide similar resources to all households and ensure equity among

them. It will be important to follow how households (particularly in Kelaguru) cope

economically in the long run. Hurdles include the slow distribution of housing compensation and delays in provision of facilities to households in Kelaguru. It has taken households in both relocation villages longer than the 6 months to settle and re-establish living in the new

103

villages. This suggests that households should receive financial support up to a year

following relocation. Other problems include absence of alternatives to forest resources

(forest products, fodder, and grazing land). Concerted efforts to avoid or minimize delays

have to be made as this only increases frustrations of people. Prolonging the process leaves

people vulnerable, as they deal with pressures of moving to and establishing themselves in a

new environment.

4.4.2 Implications for Bhadra

The resettlement of people will have major implications for the reserve. Karanth et

al. (2006) examined forest disturbance and impacts of human activities in the reserve. This

study estimated that human activities in six of these 13 villages disturbed 8 to 10% of the

forest. The households in the reserve owned ~4000 livestock and grazed all their livestock in

the reserve. Removal of livestock will allow regeneration of vegetation, recovery of grazed

areas, and improve forage available to wild herbivores. In 2002, all households collected fuel

wood inside the reserve (Karanth 2003). Relocation will significantly decrease the number of

trees lopped, cut, and burnt (Karanth et al. 2006). Fuel wood consumption rate varied

between 2190 kgs per week in the smallest village Karvani and 22140 kgs per week in the

largest village Madla (Karanth et al. 2006). Households collected several forest products in

the reserve. The non-timber forest products include four species of bamboo, honey, savige

(Sterculia villosa), nellikai (Emblica officinalis), wate huli (Artocarpus lacoocha), seegakai (Acacia concinna), soap nut (Sapindus emarginatus), wild fruits, and edible mushrooms (Karanth 2003;

Karanth et al. 2006). The households also sold two non-timber forest products (seegakai and soap nut) commercially in markets. Details on quantities collected, market price, and income generated are in Karanth et al. (2006). The relocation will significantly decrease poaching and will support the recovery of several commonly hunted species (see Table 4.1; Karanth 2003).

104

The relocation of villages will cause an overall decrease in forest disturbance, as well as

diminished impacts of grazing, hunting, and collection of forest products. I expect this to

promote regeneration of several plant species and recovery of animal populations.

4.4.3 Relocation and Resettlement as a Conservation Tool

At present, there are few well-documented studies of resettlement in India or,

indeed, worldwide. There is limited understanding of the social injustice issues and impacts

that relocation has on the lives of people. Relocation has been attempted in many Indian

reserves (Gir National Park, Sariska Wildlife Sanctuary, Kanha National Park, Nagarahole

National Park), and most did not involve long-term follow up of relocated people

(Rangarajan and Shahabuddin 2006; Karanth and Karanth 2007). The Bhadra case study is

certainly a step in the right direction compared to the overall poor record of relocations from reserves in India. This case study illustrates the complex nature of issues involved and provides several valuable insights for conservation practice. It is clear that the multitude of socio-economic, cultural, political, and ecological aspects involved requires the active engagement of relocated people and involvement of more than one agency. The conservation community must not assume that people living in reserves want to remain there and are reluctant to leave. Rather, as in this case, people are willing to relocate if provided fair compensation and adequate socio-economic opportunities. This case study shows that co-operation among non-governmental organizations and governments can be effective and help resolve complex conservation issues in an equitable and ethical manner.

Such projects require sustained financial investments with support for relocated people extended to the medium and long run. People must be able to regain or surpass the economic security and social stability they experienced before relocation. Overall, the resettlement has had a mixture of beneficial and detrimental outcomes, and is claimed a

105

success by some interest groups. The project has many aspects that need to be emulated, and

others that must be avoided.

In countries like India that are under enormous economic growth and demographic

pressure, nature reserves play a critical role in protecting biodiversity. With less than 5% total

land area classified as protected, it is important to delineate some nature reserves as fully protected while allowing different levels of access and use in others. In fully protected reserves, the conservation community must work on developing long-term solutions that may require relocation and resettlement of people. Conservation efforts that require resettlement of people from reserves will need to provide suitable alternatives and must not be the only solution. Relocation of people from reserves will not eliminate conflicts that exist with the surrounding landscape. Successful conservation initiatives will also require greater participation and involvement of local communities that live around these reserves.

For resettlement to be considered a viable conservation tool, solutions must be developed with active involvement of local people and require substantial long-term investments. The conservation community will need to assess the needs of wildlife and people in an objective independent manner on a case-by-case basis. Targeted efforts to conserve biodiversity and improve human welfare can be effective in specific cases such as Bhadra. In densely populated countries like India, I submit that resettlement presents a workable conservation solution for sustaining nature reserves and improving human lives if carried out in a socially just and equitable manner.

106

5. EXAMINING CONSERVATION ATTITUDES, PERSPECTIVES, AND CHALLENGES IN INDIA1

5.1 Introduction

Biodiversity conservation issues are often contentious and complex. Conservation

debates and disagreements among Indian conservationists mirror those of their global

counterparts. Indian conservationists are divided about rights of people living in reserves

versus protection of wildlife, conservation-development initiatives in and around reserves,

effectiveness of charismatic species as umbrella species, state versus local stakeholder control

of reserves and natural resources, and specific conservation laws and policies (Karanth and

Madhusudan 2002; MOEF 2005; Madhusudan et al. 2006; Shahabuddin and Rangarajan

2007; Karanth 2007). Differing visions of how India’s wild landscapes need to be managed

are at the center of these debates (Saberwal and Rangarajan 2003; Sivaramakrishnan 2003;

Karanth 2003; Karanth and Bhargav 2005; Rangarajan and Shahabuddin 2006; Sekhsaria

2007). These differences in opinion have polarized conservationists and hindered

conservation efforts. Understanding the basis for these divergent views and differences of

opinion is a vital first step to building consensus, addressing emerging challenges, and

developing solutions that promote effective biodiversity conservation in India and elsewhere.

Attitude and opinion surveys are widely used to understand why people make

decisions, and behave in certain ways. Surveys elicit respondents’ beliefs about important

issues that might influence their behavior and actions (that may or may not be directly

1 This chapter was co-authored with R. Kramer, S. Qian and N. L. Christensen and published in the Journal Biological Conservation in September 2008 (Vol 141: 2357-2367). 107

observable). Attitudes play a major role in acceptance of environmental policies or

management actions by the public, and conservationists in particular (Winter et al. 2005).

Examining people’s attitudes is important for formulating policies and management actions,

and generating public awareness (Gillingham and Lee 1999; Kaltenborn et al. 2006; Soto et

al. 2001).

Individuals’ worldview, familiarity with and knowledge of issues, and socio-economic

characteristics affect their opinions and attitudes. Factors such as age, gender, education, and

income level often influence people’s support for particular issues (Kaczensky et al. 2004;

Kleiven et al. 2004; Pratt et al. 2004; McFarlane et al. 2006). Other factors found to influence

conservation attitudes include personal environment, value of open spaces, and experiential

events (Kellert 1991). Conservation surveys in particular, can help us understand attitudes

and opinions towards conservation measures and policies. Examples include landowners’

attitudes towards conservation schemes (Brook et al. 2003; Winter et al. 2005), public

perceptions of biodiversity protection (Kramer et al. 2008), and community attitudes

towards management regimes or particular species (Soto et al. 2001; Bandara and Tisdell

2002; Caro et al. 2003; McFarlane et al. 2006).

5.1.1 Background

India is the world’s second most populous country with a population growth rate of

1.3%, and economic growth rate of 9.2%. In the last 150 years, population density has

quadrupled from 80 to 324 people/ km2 (Rangarajan 2007). Despite rapid economic growth, there is widespread poverty- 35% of people live on less than $1 a day, and 80% of people

live on less than $2 a day (UN, 2006).

108

The majority (70%) of people live in rural areas. India has been an agrarian country

for much of its recorded history, and at present 46% of total land area is cultivated. In

comparison, China has only 10% of land under agriculture (Rangarajan 2007). Population

growth coupled with rapid economic growth is exerting enormous pressure on India’s

limited natural resources. Like much of sub-Saharan Africa, South, and South-Eastern Asia,

a huge proportion (57%) of the country’s labor force is dependent on agriculture (UN 2006).

India has ten biogeographic realms, and is one of the world’s 17-mega diversity countries that together support two-thirds of the world’s biological resources (Rodgers and

Panwar 1988; Briggs 2003). Among the flora, 33% of the country’s 49,219 plant species are endemic to India (MOEF 1999). Although it covers just 2.4% of the world's area, India accounts for 7.3% of the world’s terrestrial vertebrate species with 89 451 faunal species

(MOEF 2000). India has several charismatic species, including 40% of the world’s tigers, and most of the world’s Asian elephants. Both tigers and elephants are ‘umbrella species’, whose protection is thought to conserve other species and habitats. Current population estimates are 1,500 – 3,000 tigers, and 10,000 – 20,000 elephants. Overall, conservative estimates suggest that 20% of Indian mammals face imminent extinction, and many have disappeared from over 90% of their historic range (Madhusudan and Mishra 2003)

The British left behind a mixed conservation legacy. From the 1850s to 1920s, they promoted widespread hunting of game animals, but they also set aside more than 600 000 km2 of government forests (Stebbing 1923). With the expansion of agricultural frontiers,

construction of the railways, and establishment of plantations, many species survived only in

these areas. These hunting reserves and government forests laid the foundation for India’s

109

protected areas (Stebbing 1923). The game preservation laws passed by the British

standardized the hunting and shooting rules, but failed to curtail the disappearance of many

species (Champion 1934). Post independence, the first successful legislation to protect

Indian wildlife was enacted in 1972 – the Wildlife Protection Act (MOEF 2003). The

impetus for this came from the personal commitment of Prime Minister Indira Gandhi,

international pressure, and alarm raised by prominent conservationists (Rangarajan 2001).

This act banned hunting, and ‘commercial’ exploitation of timber and forest products from

nature reserves in India.

The Indian government established many types of protected areas (national parks,

sanctuaries) in the 1970s through 1990s. In 30 years, land under formal protection grew

from less than 1% to greater than 4% of total area (Rangarajan 2006). In 1980-81, the Indian

government passed the Forest Conservation Act that prevented the diversion of Reserved

Forest land for agriculture and development projects. Wildlife laws sought to protect forests,

closed some areas to grazing, collection of forest products, and slowed down the legalization

of forest occupation by homesteaders (Saberwal et al. 2001). Both laws facilitated the establishment and strict protection of several nature reserves (Karanth 2003). In subsequent decades, conservation efforts have focused on managing the new challenges that have emerged in these reserves (Karanth and Madhusudan 2002; Saberwal and Rangarajan 2003;

Karanth and Karanth 2007; Shahabuddin and Rangarajan 2007). During the 1980s and

1990s, de-centralization of power and failures of the federal and state government to address local challenges weakened conservation schemes.

110

Today, protected areas cover less than 5% of total land area in India, and they

protect most of the country’s remaining biodiversity. Indian reserves are under enormous

pressure from people residing both inside and surrounding them. These reserves support

many human activities such as hunting, fishing, grazing, and collection of forest products are

widespread (MOEF 2000; Forest Survey of India 2000; Das et al. 2006; Karanth et al. 2006).

Human wildlife conflicts (crop raiding, livestock and human depredation) do occur, resulting

in retaliation against some species (Karanth and Madhusudan 2002). Fragmentation of

habitats has added to wildlife threats. There is increasing concern that India will lose most of

its rich fauna, and flora by the end of this century.

Project Tiger (launched in 1973), is one of India’s most important conservation measures. This policy aimed to protect this endangered cat, and provide federal funding and

oversight to create tiger reserves in important habitats. The tiger became a flagship for new

policies, and initiated widespread on-the-ground conservation initiatives (Rangarajan 2001).

The Indian government launched Project Elephant in 1992. This effort focused on

protecting elephants, providing financial and technical support to Indian states, and

conserving important elephant habitats and corridors. Both initiatives were developed by the

Indian government to oversee protection of these species, and their habitats on a

countrywide scale. However, the resulting population estimates for each species and long-

term effectiveness in conserving these species on-ground have come under intense criticism.

The Scheduled Tribes and Other Traditional Forest Dwellers Recognition Act, 2006

(from here on referred to as the Forest Rights Act) seeks to recognize and vest forest rights

to forest dwelling people. Currently, Scheduled Tribes comprise 8% of India’s population

111

(Rangarajan 2005). Many of the tribal members depend heavily on forests for their

livelihood, and this has resulted in intense conflicts with wildlife in many reserves. The

Tribal Act is an attempt to resolve these issues. This Act provides families access to and

ownership of minor forest produce, grazing and cultivation rights, and access to traditional

seasonal resources. The Act attempts to restore forest rights to these people on ancestral land, some of which consolidated under forest areas during the colonial and post independence periods (Kothari and Pathak 2005; MOEF 2005). It legalizes land occupied by each nuclear family (residing inside government forests prior to 2005, Rangarajan 2005). The

Indian Parliament passed the Act in 2006. India’s Supreme Court is reviewing a case challenging the passage of this Act. Recently, there has been a decision to exclude tiger reserves from the jurisdiction of this Act (Ramnath 2008).

Supporters of the Act argue that it will correct past injustices that some forest dwelling tribes were subjected to (Kothari 2005; Prabhu 2005). The Act has potential to provide secure land tenure, and livelihood security to many marginalized groups of Indian society (Sekhsaria 2007; Ramnath 2008). Critics of the Act believe that it will result in significant loss of forest cover (Jayakrishnan 2005). A major provision of this Act is that it will override provisions of the Wildlife Protection Act (1972), the Forest Conservation Act

(1980), and the National Forest Policy (1988), and dismantle many of country’s current

wildlife protection measures (Karanth and Bhargav 2005; Madhusudan 2005; Mohanty 2005;

Sahgal et al. 2005; Bhargav 2007). Critics state it will be hard to separate deserving members from others who have no rights to these lands, and will reverse effectiveness of Indian conservation policies (Goenka 2005; Karanth and Bhargav 2005; Krishnaswamy 2005;

112

Mohanty 2005; Ramnath 2008). There is also concern that the Act will be used to front

vested development initiatives such as mines and roads (Ramnath 2008). For additional

discussion on the Act see Bhargav 2007, Krishnaswamy 2005, Sekhsaria 2007, and Ramnath

2008.

In the last two decades, India has witnessed many wildlife crises. In 2005, tigers went

locally extinct from Sariska Tiger Reserve. Sariska reserve had been cited as an example of a tiger reserve where there was successful local participation in the conservation of the tiger,

and its habitat. In response to widespread public outcry, the Government of India instituted

the Tiger Task Force Committee in 2005. This committee assessed the situation in Sariska,

overall conservation problems in India, and developed strategies and guidelines to mitigate

the situation.

Against this background of conservation history and policy in India, we examined

attitudes and perspectives of modern day Indian conservationists on specific conservation

issues and policies. We questioned participants on the quality of biodiversity research and

dissemination of results, as well as perceived bias in conservation and research approaches

(scale, species, and ecosystems). We focused on several questions. What are participants’ attitudes toward conservation and research efforts in India? Are protected areas effective in conserving India’s biodiversity? Should “force” be used to protect reserves? Can people live inside reserves? Do charismatic species like tigers and elephants serve as effective umbrella species? We examined their attitudes towards specific conservation initiatives like Project

Tiger, Project Elephant, Forest Rights Act, and Tiger Task Force Report. We used classification and regression tree (CART) models to examine if participant characteristics

113

such as age, education, role, affiliation, gender, and years of conservation experience may

influence their attitudes.

1.2 Methods

5.2.1 Survey Methods and Participants

We conducted a sample survey to investigate the attitudes and perspectives of natural

and social scientists, and conservation practitioners in India. We focused on urban and semi-

urban individuals in India who are actively involved in conservation. We pre-tested the

survey with five conservationists in India to check for wording clarity, and revised the survey questions based on their recommendations. We utilized a “snowball-sampling” framework – starting with an initial list of 50 well-known conservationists. These participants recommended other individuals suitable to participate in the survey. We administered the survey questionnaire by postal and electronic mail to 312 people from January to August

2006.

5.2.2 Questionnaire Design and Analysis

The questionnaire had 32 questions. Participants used a Likert scale to score their responses to indicate their level of agreement to particular questions. The Likert score scale

ranged from 1 to 5 (with 1 representing strongly disagree to 5 representing strongly agree).

We elicited narrative responses to a number of open-ended questions. A series of close-

ended questions were used to obtain socioeconomic data. Participants provided details on

their age, gender, academic degree, affiliation, role, and years of involvement in conservation.

We grouped educational background into four categories: bachelors, masters, doctoral, and

other. Occupational groupings were academia, governmental, and non-governmental 114

institutions, and independent. Participants classified themselves as natural scientists, social scientists (including conservation education and advocacy), conservation managers, amateur naturalists, and other (law, media etc). Conservation experience was divided into four intervals: 1 to 10 years, 11 to 20 years, 21 to 30 years, and more than 30 years. Questions assessed participants’ opinions on the quality of biodiversity research and dissemination of results, as well as perceived bias in conservation and research approaches (scale, species, and ecosystems).

Table 5.1: Demographic Characteristics of Survey Participants

Characteristics of Participants Percentage 1. Gender Male 83% Female 17% 2. Average Age 39.9 years (range 23 to 79) 3. Highest Academic Qualification Bachelors 21% Masters 42% Doctoral 34% Other 3% 4. Conservation Experience 1 – 10 years 41% 11 – 20 years 33% 21 – 30 years 17% > 31 years 9% 5. Affiliation Government 19% Non-governmental Organization 52% Academic Institution 25% Independent 4% 6. Primary Focus Natural Science 40% Social Science, Policy and Education 16% Conservation Management and 19% Advocacy Amateur Naturalist 9% Other 16%

115

We compiled questionnaire responses in Excel (Microsoft Office 2003), and analyzed

the dataset in R (2.6) software. We used classification and regression tree (CART) models

(implemented in package rpart) to assess the relationship between participant characteristics,

and their attitudes towards conservation management and policies in India. Many disciplines

such as ecology (Taverna et al. 2004), economics (Wezel and Potharst 2007), and medicine

(Hess et al. 1999) use CART as an exploratory data analysis tool or as a model building

method. CART recursively partitions data to yield models known as tree-based models

(Brieman et al. 1984). CART models allowed us to explore and uncover structure in the data,

screen variables, and capture non-additive behavior. Tree-based models are easier to

interpret than linear models, and allow us to model interactions among predictors without

pre-specifying them (Cirincione and Gurrieri 1997). The resulting tree identifies a series of

best predictors (depicted as nodes and branches) for the dependent variable. Qian and

Anderson (1999) suggest that CART models can identify important factors that explain the

variance in a response variable. We used CART to identify participant characteristics that

might influence their responses to important conservation policies, and measures. In our

case, CART allowed us to explore relationships with a small sample size, and identify the

most important factors. We used classification models and assessed model performance by

misclassification rate. The misclassification rate is the proportion of responses incorrectly classified by the fitted model. The rpart package in R reports this statistic using the percentage of improvement over the “root misclassification rate". This rate is the error based on a model in which every individual has the same probability of responding a particular

way. For example, the model presented in Figure 5.1 used a dataset with 167 participants,

116

among whom 40 answered “no” and 127 answered “yes”. Without using any predictor (i.e.,

all individuals exhibit the same probability of responding “yes”), the logical prediction of the

answer from a future participant would be “yes”. This prediction is correct 76% (127/167)

of the time. The root misclassification rate is 24%. The measure of a classification model

“improvement” is the improvement over the root misclassification rate.

5.3 Results and Discussion

5.3.1 General Characteristics

We received 167 responses from the 312 individuals who were contacted, with a

response rate of 54%. Ninety-two percent of participants stated that they were actively

involved in conservation. Details on survey respondents’ gender, age, academic qualification,

conservation experience, affiliation, and primary focus are in Table 5.1. The typical

participant was male, 40 years old, had a master’s degree, worked for a non-governmental

organization, and had an average of 11 to 20 years of experience. Participants worked in 28

of India’s 30 States and Union Territories. Sixty-five percent of participants felt that some

states were more effective in biodiversity conservation efforts. These states were Karnataka,

Kerala, Maharashtra, Tamil Nadu, and Uttaranchal (four of these five states occur in southern India).

5.3.2 Research and Conservation Focus in India

Survey participants reported that they focused their research and conservation efforts in more than 15 different land cover-habitat types. The four most common habitats were deciduous forests (23%), evergreen forests (18%), scrub areas (12%), and grasslands (11%).

Research and conservation efforts were conducted at different scales, with local (park level)

117

being the most common (33%). The majority of participants (68%) worked solely on terrestrial species, while some (5%) worked solely on aquatic species, and the rest (27%) worked on both. Participants worked with more than nine major taxonomic groups, and 78 individual species of animals and plants. The most common groups were mammals (34%), birds (21%), and flora (19%). Fifty-three percent of participants felt that there was a taxonomic bias in species studied by researchers. The three most commonly listed species were tigers, elephants, and leopards. Many participants felt that terrestrial species (3.7/5.0

Likert score), and terrestrial ecosystems (3.7/5.0 Likert score) were given too much emphasis. Concurrently, they also felt that aquatic species (3.5/5.0 Likert score), and aquatic ecosystems (3.6/5.0 Likert score) were given too little emphasis.

Participants felt that the quality of biodiversity research in India was average relative to global standards (2.8/5.0 Likert score). The two major reasons chosen by participants for constraints in reaching global standards of research were generally poor state of science education in the country (37%), and insufficient allocation of funds to science education and research (33%). Participants suggested many reasons why Indian researchers lagged behind their international counterparts. Many stated that they had conflicts with the forest and wildlife departments. Conflicts arose from participants not being allowed to conduct research inside reserves (denied permits, or given permits that did not accommodate their sampling framework). This was corroborated in a recent paper by Madhusudan et al. (2006).

Some felt that the bureaucracy and corruption in the system were too pervasive to overcome. Participants also stated that there were very limited educational opportunities in ecology and conservation in India. They felt that there are relatively few research institutions

118

with mediocre faculty, and no financial support to pursue research. Some observed that

educational opportunities were often restricted to English speakers, and there was no

support or educational material in the regional languages. The lack of co-operation and

communication between governments and non-governmental organizations, as well as

limited funding from Indian granting agencies were also suggested as reasons for mediocre

research.

The majority of participants (81%) stated that they obtained scientific information from both primary and secondary sources. Participants expressed their views and disseminated their results through many outlets (radio stations, television channels, magazines, and newsletters). Participants published their results in 49 scientific journals.

There was wide consensus that research results were disseminated poorly, and rarely accessible to the public and policy makers. Participants suggested that there is a general disinterest in conservation issues among the public, government, business, and political entities. Participants highlighted that researchers collaborated with international media and audiences to further their personal profile, but did not engage policy makers and local communities once results were available. Many participants suggested that this might be improved by increasing the number of popular articles written, wider use of the internet, and opening forums for public debates.

5.3.3 Protected Areas, People, and Conservation Threats

Many participants agreed that India’s current protected areas are effective in conserving biodiversity (3.4/5.0 Likert score). The majority of participants (90%) stated that the proportion of land under protected area status in India is insufficient to conserve

119

biodiversity, and that the proportion needed to increase. The suggested increase ranged from

6% to 50% of total land area (average was 15% and median was 10%). There were more

than 15 types of conservation threats categorized for India. The six most serious threats

identified by participants were habitat fragmentation, human-wildlife conflicts, deforestation,

land conversion, mining, and hunting of wildlife. Suggestions for improving conservation

efforts involved a variety of approaches. Some participants (34%) felt that this required

changes in forest and wildlife management expertise, infrastructure, and better

accountability. Others (24%) felt that it was important to involve local communities in

sharing benefits from reserves. Other suggestions included partial removal of people from

nature reserves (16%), and complete removal of people from reserves (11%).

The use of “force” to protect reserves from illegal human activities was acceptable to

many participants (76%). Participants’ suggested several enforcing agencies: forest and

wildlife personnel, setting up a new specialized park protection force, local police and law

enforcement agencies, non-governmental organizations, and guards from the army. Tree- based modeling allowed us to examine the effects of participant characteristics (age, gender, degree, affiliation, role, conservation involvement, and years of experience) on their “yes” or

“no” response to this question about the use of force. The resulting tree (Figure 5.1) suggests that affiliation is the primary factor, with age and degree as the secondary factors.

Both academics and non-academics are likely to approve of the use of force to protect reserves, but the percentage is much higher for non-academics (Table 5.2).

120

Figure 5.1: CART model for the question on use of force in protected areas. The text above each split shows the variable that is split, and the condition for the left branch is stated. The original data set has 40 “no” responses and 127 “yes” responses. At the first node, academics (left branch) were equally likely to say “yes” or “no” and non-academics (right branch) overwhelmingly favor “no”. Academics are likely to say “yes” (56%) and non-academics are more likely to say “yes” (82%). Among academics, older participants are more likely to say “yes”. Among non-academics, those with doctoral degrees are more likely to say “yes”.

121

122

123

Table 5.3: Summary of Participants Comments

People • Relocate people from important reserves, especially in places where people want to relocate outside, and provide them with suitable alternatives (13). • Promote conservation education programs, and create awareness among local communities that live in/around reserves (10). • Communities that have historic-cultural ties to the reserves should be allowed access to them (6). • Provide alternative livelihood options, and make people less dependent on reserves (4). • Lifestyles and standards of living of forest dwellers have changed, and are incompatible with wildlife conservation (3). • Develop secure property rights institutions, and link conservation benefits to local livelihoods (3). Management and Research • Strict enforcement of laws and protection by the wildlife and forestry departments (12). • Set aside core and inviolate areas and prevent human access, but allow access in buffer areas (5). • Requires strong support of the government (3). • Increase collaboration with NGOs, and involve them in decision making (3). • Integrate unprotected areas rich in biodiversity using creative solutions (3). • Involve communities in benefit sharing (ecotourism etc.), and involve them in protection (2). • Promote research, establish wildlife monitoring programs, and adapt management decisions accordingly (2). Participatory Approaches • The need for involvement must come from within the local communities, and they must derive benefits (17). • Rights and responsibilities must be outlined at the outset, and regulatory mechanisms within the community must be in place (10). • Works if resource use patterns and conservation priorities do not seriously compromise each other (8). • Need regulatory agencies, and must involve local NGOs. Local organizations and individuals with long-term commitments to the area are most effective, and can engage local communities (5).

124

We asked participants if people could sustainably live inside protected areas in the

long term. Most participants (61%) did not agree, although some (32%) felt this could be

sustainable if appropriate guidelines were established. A second tree model identified

affiliation as the primary factor affecting views towards people living inside reserves.

Participants who worked for the government, NGOs, and others generally felt this was not

feasible, whereas academics were supportive of people living inside reserves (Table 5.2).

Participants distinguished fully protected zones and extractive zones in reserves. The

majority (63%) felt that both types of zones were important, although some (35%) felt that fully protected zones were more important. Representative comments of participants on the effectiveness, and importance of protected areas are summarized in Table 5.3.

We asked participants whether they believed participatory conservation approaches

could work. Many (> 84%) participants were familiar with participatory approaches to

conservation. Among all participants, 44% said “yes” (participatory approaches could work),

38% said “under certain conditions”, and 18% said “no”. The tree model identified

affiliation as the primary factor influencing these responses. Respondents affiliated with

NGOs and academic institutions were more likely to support participatory approaches

(Table 5.2). The most common observations and comments of participants on the use of

participatory approaches in conservation are in Table 5.3.

5.3.4 Conservation Initiatives and Policies

We examined attitudes towards four major conservation policies and initiatives in

India. They are Project Tiger, Project Elephant, Forest Rights Act, and the Tiger Task Force

Report. Project Tiger is one of India’s most important conservation initiatives. The majority

125

(99%) of survey participants were aware of Project Tiger. Seventy-eight percent of

respondents believed that the tiger is an effective umbrella species and 60% of participants

believed that Project Tiger was a complete “success”. We modeled the effect of participant

characteristics on “yes” or “no” responses to the success of this initiative. The resulting tree

(Figure 5.2) divides the participants into two groups determined primarily by age. Younger people (<33.5 years) are more likely to say “no” (53%) and older people are more likely to say, “yes” (66%, see Table 5.2). Most (89%) participants were familiar with Project Elephant.

Sixty-three percent of participants believed that the elephant is an effective umbrella species, and 51% of participants said considered Project Elephant a success. The tree model for responses about this initiative identified age as the primary factor, with younger participants

(< 43.5 years) more likely to say “no” (54%) and older people (> 43.5 years) are more likely to say, “yes” (60%, see Table 5.2).

At the time of our survey in 2005, the Forest Rights Act was under debate in the

Indian Parliament. Most participants (81%) were familiar with the Act, and many felt that the bill required changes (63%). The tree model identified academic degree as the primary factor influencing opinion about this Act. Those with doctoral degrees were more inclined to want the act “changed” (Table 5.2). Representative comments of participants’ on the Forest

Rights Act are in Table 5.4.

We also examined participants’ attitudes towards the Tiger Task Force committee report. Among all participants, 38% thought the report was a “step forward”, 27% were

“unfamiliar” with it, 20% thought it was “no different”, and 16% thought it was a “step backward” from previous reports. Tree models suggest that conservation involvement is the

126

primary factor influencing opinion about this Task Force. Among those involved in

conservation, 40% thought the report was a “step forward” from previous efforts of the

Indian government (Table 5.2). A summary of participants’ comments and observations on

the Tiger Task Force Report are in Table 5.4.

Figure 5.2: CART model for question on “success” of Project Tiger. The text above each split shows the variable that is split, and the condition for the left branch is stated. At the first node, age was the primary factor. Younger people (< 33.5 years) were more likely to say “no”. Among the older group, at the second node people affiliated with academic institutions and NGOs are less likely to say “yes”. In the older group amateur naturalists and social scientists are more likely to say “yes”.

127

Table 5.4: Participants Comments on Conservation Policies and Initiatives

Forest Rights Act

Positive Comments • Provides rights to forest dwellers, many of whom were marginalized and dispossessed historically (14). • Could establish and revise reserve boundaries and forest cover. Identify land that is converted and in use by people, and settle property rights permanently (4). Negative Comments • The Act will be misused, spells doom for protected areas and wildlife. It disregards ecological consequences, will fragment reserves, increase encroachment and human-wildlife conflicts. Ultimately will decimate India’s biodiversity (26). • The Act is part of a political agenda. Pays lip service to both forest dwelling groups and conservation without establishing criteria for serving either (14). • Attempt to straightjacket and group people- cannot represent 500+ forest dwelling groups with diverse historical, geographical, cultural, and socio-economic backgrounds. Disregards the changing needs of communities, their desire to integrate with society and live like everyone else. Relocate and resettle those that want to move, and provide them with appropriate resources (15). • Contradicts provisions of the Wildlife Protection Act, and requires further debate (11). Tiger Task Force Report

Positive Comments • Recommendations are sound, but implementation is going to be a challenge (14). • Pools together regional and local knowledge from across India, and encouraged inputs from the Indian conservationists (11). • Information collected is exhaustive, and available to the public (7). Negative Comments • Lacks depth, divisive, recommendations are contradictory and vague, and written by people with little scientific expertise (31). • Anthropocentric and the focus is not the tiger, biased towards forests and forest dwellers (17). • Fails to address primary purpose of reviewing current tiger crisis, lacks long term vision, and fails to provide conservation road map for tiger conservation (13). • Forest and wildlife departments are expected to do much (manage parks, enforce laws, oversee tourism, work with local communities, monitor wildlife, and conduct research). Does not hold them accountable for their failures and negligence particularly in places like Sariska (8). • Disregards successful conservation efforts, ignores the role that NGOs play, and the need for collaboration with them (4).

128

5.4 Conclusions

Our study provides a number of insights about attitudes, and opinions of conservationists in India. Qualitative and quantitative responses of the participants demonstrated that many conservationists were aware of challenges, and problems facing

Indian conservation. Although sample surveys provide a general and broad assessment of conservation attitudes, we recognize that additional insights may be gained from in-depth interviews and other qualitative research approaches to understand the plurality of factors influencing attitudes of conservation researchers and practitioners.

The use of tree models provided insights into participant characteristics that may influence conservationist’s attitudes. The survey highlighted differences in awareness and interest levels among different respondent groups, as well as differences in conservation attitudes. For questions related to protected areas (use of force and people living in parks), affiliation was the primary response predictor. Academics were more likely to say “no” to the use of force in protecting reserves, and “yes” to people living inside protected areas.

Those participants engaged in “hands-on” conservation (e.g., reserve managers, people in forest and wildlife departments) tended to be more skeptical that parks can achieve their conservation goals with people living in them. Perhaps these differences in factors reflect a more academic view of human behavior versus a more practical view informed by firsthand experience with people and protected areas. With regard to Project Tiger and Elephant, older participants were more likely to think that these initiatives were successful. This is perhaps because older participants have more experience with traditional conservation programs, while younger participants are more aware of newer research and monitoring

129

tools (camera trapping versus pugmark census of tigers). Academic degree was the primary factor in predicting participants’ opinion of the Forest Rights Act. Participants with doctoral degrees were more likely to think it needed revision. There is considerable controversy surrounding this Act, and its potential impacts. Perhaps those higher education levels are more aware of the issues surrounding the controversy. With respect to the Tiger Task Force

Report, among those involved in conservation, some (40%) thought it was a step forward.

This may be because this high profile report has helped refocus national attention on conservation needs and challenges in India.

Our study has highlighted the differences, and commonalities in opinion that exist among Indian conservationists (although our survey primarily focused on urban and semi- urban conservationists). India faces new conservation challenges with a rapidly growing economy, and the influence of globalization. Our participants suggested several ways to improve conservation research and management efforts (Table 5.4). These included relocation of people (with compensation) from important reserves, stronger enforcement of existing wildlife laws, and greater involvement of local communities. Formation of committees to address crises (like the Tiger Task Force), or debates on laws (like the Forest

Rights Act) require active participation of conservationists, and increasing awareness among the public. Understanding the views of different actors and agencies can lead to diminished conflict, more constructive debates, and provide improved opportunities for conservation.

Finally, debates on biological conservation reflect different viewpoints that are shaped by a variety of stakeholders’ characteristics, particularly education levels, work environment, and age. This implies that consensus among conservationists about

130

conservation strategies and policies will depend on forging alliances among those with quite different backgrounds. This research is a first step toward identifying sources of varying attitudes and opinions. Our research also suggests that the Indian conservation community may benefit from workshops and roundtable discussions comprised of conservationists with diverse backgrounds in the search for solutions for the country’s increasingly complex challenges surrounding protected areas, nearby human communities, and tensions between conservation and development.

131

6. CONCLUSION

This dissertation provides new insights on factors influencing extinction and

distribution of mammals, as well as conservation policies and debates in India. The

challenges facing the Indian conservationists are much like those facing global

conservationists. They include species extinctions, issues of effective protection and

scientific management of protected areas, and resolution of human-wildlife conflicts. India presented a particularly appropriate context for examining these conservation issues in the context of conserving biodiversity in a human-dominated landscape.

We need to identify patterns of extinction, occurrence and distribution that are based on good science. We also need to understand the ecological and social factors that influence species persistence, if we are to successfully manage and protect surviving populations of these species in future. Examining the importance of nature reserves, understanding possible reasons for species persistence under different land uses, human densities, and human cultural tolerance are critical to species recovery.

In this dissertation, I focus on Indian mammal conservation science, management, as well as policy issues that shape these factors. Recent estimates suggest that 25% of global mammal species face extinction and, South Asia in particular, harbors the most threatened terrestrial mammals (Ceballos et al. 2005; Schipper et al. 2008). I find that protected areas and human cultural tolerance have a major role in promoting persistence of species in human-dominated landscapes. Conservation efforts to protect wildlife in human-dominated landscapes such as those in southern Asia, often require relocation of human settlements. I find that conservation policies that involve relocation and resettlement of people from

132

wildlife reserves, can be carried out equitably and present a potential solution to protecting biodiversity in India. Our ability to succeed in developing and implementing conservation policies and management interventions, be they for protecting wildlife or addressing needs of local communities, depends substantially on the attitudes of conservation practitioners.

6.1 Chapter Summaries

Mammal conservation efforts are hindered by poor distribution data and range maps.

Many Indian mammals face range contraction and extinction, but population status cannot be evaluated without an understanding of current geographic ranges. In Chapter 2, I examine extinction of 25 large mammals in India. In Chapter 3, I focus on current distribution of 20 large mammals in India. For both chapters, I estimated the extinction and current geographic ranges of large Indian mammals by applying occupancy models to data from a sub-continental scale expert opinion survey. We modeled species occurrence in relation to ecological and social covariates (protected areas, landscape characteristics, and human influences) based on a priori hypotheses about the determinants of mammalian extinction and distribution patterns. I used maximum likelihood inference and information-theoretic model selection procedures to discriminate among multiple models representing different hypotheses about the processes that generated the data. I also generated species-specific extinction and occupancy estimates. I mapped extinction and current distribution to identify priority conservation targets.

Determining factors that drive mammalian extinctions and species persistence is critical to successful conservation efforts. Most studies of animal distribution and persistence, however, use methods that do not tease out true species absence from simple

133

non-detection during survey. This underestimates true occupancy, which confounds the

analyses of spatial data to determine the impact of various ecological and environmental

factors on species persistence, survival, and recovery. Occupancy models incorporate

detection probability along with relevant biological and socio-cultural covariates to accurately

predict species extinction and occurrence. There were several novel features in this analysis

as we quantified and examined determinants of species-specific extinction patterns.

In Chapter 2, I find all 25 large mammal species are extinction prone (values ranged

between 0.20 and 0.96). As predicted, the time-related covariates were important for 24 species. Scaled extinction probability was higher or the same over 100 years for 17 species,

and higher over 50 years for 8 species. Replicate sampling through time would have

permitted the use of multi-season occupancy models that permit the direct estimation of

time-specific probabilities of local extinction and colonization (MacKenzie et al. 2003, 2006).

In the absence of the needed replicated data for all but the terminal year of the analysis

(2006), we dealt with time using linear and quadratic models.

In both Chapters 2 and 3, as predicted I find protected wildlife reserves (both in

terms of presence and proportion of land occupied in the cell) are critically important for persistence of species that protected areas are critical for the current persistence of species

This is strongly so for forest-dwelling species, and all carnivores in general. Species with

much of their habitat outside existing protected areas, and species for which only a tiny part

of their historic range was protected, will require new protected areas to persist. India’s

current fragmented network of relatively small protected areas (average size < 300 km2), do have high carrying capacities for large mammals (Karanth et al. 2004). However, given the

134

overall patterns of species occupancy and extinctions we found in this study, creation of new

and interconnection of more protected areas is necessary for the persistence of most of these

species in the long term. This is particularly so for in the context of long-term changes in

climate and human demography.

In Chapter 2, I find proportion of forest cover to be important for persistence of 9

species, and higher elevation to be important for persistence of 11 species. In Chapter 3, I

find found evergreen forest to be important for 19 species, temperate forests important for

16 species, deciduous forests important for all 20 species, and elevation was important for

four species.

In both Chapters 2 and 3 I find that human population density negatively influences

survival probability for species, and human cultural tolerance positively affected persistence

of species. Exceptions were species that were rare localized endemics, generalist species that

adapt to human dominated landscapes to some extent (wild pig, jackal, wolf, leopard), and

culturally tolerated or protected species (blackbuck, nilghai, chinkara). Overall, most large-

bodied animals, habitat specialists, and rare species had higher extinction probabilities.

In both Chapters, I find that in addition to presence of protected areas, land use, and

human population densities, regionally rooted cultural and religious factors have allowed

some species to survive. Conservation strategies must integrate all these factors to ensure the survival of India’s large mammals in the future.

The relocation and resettlement of people from nature reserves is a controversial issue in the conservation community. The perceived poor success rate of resettlement efforts, combined with availability of few well-documented studies, warrants a detailed

135

examination of this issue. In Chapter 4, I have analyzed a relocation and resettlement project in India’s Bhadra Wildlife Sanctuary. I examine the relocation experience of 419 households who moved to two villages located outside the reserve. I interviewed 61% of relocated households in 2002 and 55% relocated households in 2006. In 2002, 71% of households were satisfied with the relocation effort with respect to their quality of life. In 2006, 52% of households were satisfied with the quality of life. Four years after relocation, all households have access to electricity, water, schools, health care, transportation, and communication facilities. Many households have increased their income and assets. Yet, there were differences between the two resettlement villages, with one of them faring better in terms of economics, hardships, and uncertainty. This chapter draws out insights important for improving conservation practices related to resettlement efforts. It documents short to mid- term successes and challenges that affect the communities involved. I submit that in specific contexts, relocation may be a potential viable conservation tool. Successful conservation resettlement requires substantial financial support to meet people’s socio-economic needs, active consultation of the people involved, and partnerships of committed non- governmental and governmental organizations.

In Chapter 5, I examined the attitudes, opinions and perspectives in the Indian conservationists. Polarized debates on the effectiveness of protected areas and role of people inside them, charismatic species as conservation foci, and on specific policy initiatives are common among Indian and global conservationists. I surveyed Indian conservationists about the conservation effectiveness of protected areas and charismatic species, as well as status of conservation and research efforts. I expected differences among people based on

136

professional affiliation and educational background. I examined participants’ opinions on

conservation policies like Project Tiger and Elephant, the Forest Rights Act, and the Tiger

Task Force Report.

Participants ranked Indian research efforts as average, and identified a bias towards

terrestrial species and ecosystems. Ninety-percent of participants considered reserves to be

effective, many (61%) participants felt that the situation of people living inside reserves is unsustainable, and many (76%) felt the use of force to protect reserves from illegal human activities is acceptable. Classification and regression tree models for these questions suggested that non-academics were more likely than academics to agree with these positions.

On the success of Project Tiger and Elephant, older participants were more likely to think these initiatives were a success. Many (63%) participants felt the Forest Rights Act needed revision, particularly if they had doctoral degrees. Sixty-two percent of participants did not think Tiger Task Force was effective. Overall, participants’ professional affiliation, age, and academic degree were important predictors of their attitudes of conservation initiatives.

Chapter 5 provides interesting insights about attitudes and opinions of conservationists in India. India faces new conservation challenges with a rapidly growing economy, and the influence of globalization. Our participants suggested several ways to improve conservation research and management efforts. These include relocation of people

(with compensation) from important reserves, stronger enforcement of existing wildlife laws, and greater involvement of local communities. Formation of committees to address crises

(like the Tiger Task Force) or debates on laws (like the Forest Rights Act) require active participation of conservationists and increasing awareness among the public. Understanding

137

the views of different actors and agencies can lead to diminished conflict, more constructive debates, and provide improved opportunities for conservation.

Each of these research questions addressed in Chapters 1 to 5 arose from my passion and desire to integrate India’s wildlife history with the conservation ecology and social sciences. My attempt here is to integrate good ecological science and understanding of the social context of conservation in a pragmatic mix that is mostly like to be effective in saving

India’s wildlife. I apply innovative methods in both science and social sciences, and integrate them with rigorous modeling approaches to explore solutions to major conservation problems in a country that has some of the richest wildlife resources.

The research questions addressed in my dissertation (species extinctions, human resettlement, conservation policies), are of global conservation interest. Although, I focus on mammals, these techniques can be easily applied to other taxa, and in other geographical regions. The emphasis is on using innovative modeling approaches which builds on existing knowledge (historic and current) to identify where species are threatened, and what factors promote their persistence. India is at a critical juncture- species are experiencing widespread declines in numbers, and ranges are rapidly shrinking. We have to factor in the needs of 1.2+ billion people, and develop creative solutions to save India’s biodiversity. Other countries, although not under the same demographic and resource pressure, are also experiencing significant threats to biodiversity that will require integrating conservation science, policy and management efforts.

My dissertation suggests conserving biodiversity requires integrating sound scientific approaches with effective management decisions, and realistic conservation policies. I find

138

that human attitudes and involvement, whether they be local communities or conservation

practitioners, will be critical to long term success in our fight to save our planet’s biodiversity. Conservation will succeed if it is driven by good science, supported by effective

management and policy solutions, and involve people.

6.2 Future Directions

Growing up in India, I was aware of the deep connections between human

communities and natural ecosystems. As a scientist, I am committed to understanding these

connections fully, particularly amid fast social and ecological change. My work in India

engages global environmental issues (biodiversity loss, social vulnerability, land cover

change). I am committed to finding place-based conservation strategies sensitive to local

ecologies, and socioeconomic systems. I am broadly interested in human-environment

interactions, particularly focusing on protected areas, and their relationship to people and

biodiversity. I intend to pursue research themes that are extensions of my doctoral and

master’s research. I want to focus on (1). Protected areas and their relationship to people

living around them; (2). Biodiversity patterns and factors that drive the persistence of

wildlife; (3). Societal impacts of conservation interventions and policies. I am particularly

interested in human-wildlife interactions, land use change, livelihood dependence of local

communities, resettlement of people, and tourism impacts. I focus on protected areas

because they remain the foundation of most global conservation efforts. My dissertation

results suggest that parks and human cultural attitudes are critical to sustaining India’s

biodiversity, I will explore this issue further examining conservation strategies in protected

areas that focus on benefits (ecosystem services, tourism) from parks. My research has

139

challenged appealing, but simplistic notions of local communities as stewards of environment. My experiences have convinced me that communities are heterogeneous, deeply connected to extra local sociopolitical processes and market forces that shape their decisions about the environment. Therefore in addition to studying patterns of direct resource use by local communities, I want to examine impacts of conservation interventions

(resettlement, land purchase). Additionally, I aim to explore the issue of human cultural tolerance, how people’s varying attitudes across India affect persistence of wildlife in and outside parks, and how this can be used to build local and regional support for conservation efforts.

140

Appendices

Appendix 1: Models included in model set for 25 large mammal species

Models Occupancy and Detection Constant 1. Ψ (.) p (.) Protected Areas 2a. Ψ (park) p (park) 2b. Ψ (parkprop) p (parkprop) People 3a. Ψ (ppl) p (ppl) 3b. Ψ (ppl+park) p (ppl+park) 3c. Ψ (ppl+parkprop) p (ppl+parkprop) 3d. Ψ (ppl+elv) p (ppl+elv) 3e. Ψ (tol) p (tol) 3f. Ψ (tol+park) p (tol+park) 3g. Ψ (tol+parkprop) p (tol+parkprop) 3h. Ψ (tol+elv) p (tol+elv) 3i. Ψ (ppl+tol+park) p (ppl+tol+park) 3j. Ψ (ppl+tol+parkprop) p (ppl+tol+parkprop) 3k. Ψ (ppl+tol+elv) p (ppl+tol+elv) 3l. Ψ (ppl+tol+parkprop+elv) p (ppl+tol+parkprop+elv) Elevation 4a. Ψ (elv) p (elv) 4b. Ψ (elv+park) p (elv+park) 4c. Ψ (elv+parkprop) p (elv+parkprop) 4d. Ψ (elv+fc) p (elv+fc) Forest Cover 5a. Ψ (fc) p (fc) 5b. Ψ (fc+park) p (fc+park) 5c. Ψ (fc+park+ppl) p (fc+park+ppl) 5d. Ψ (fc+park+tol) p (fc+park+tol) 5e. Ψ (fc+park+elv) p (fc+park+elv) 5f. Ψ (fc+parkprop) p (fc+parkprop) 5g. Ψ (fc+parkprop+ppl) p (fc+parkprop+ppl) 5h. Ψ (fc+parkprop+tol) p (fc+parkprop+tol) 5i. Ψ (fc+parkprop+elv) p (fc+parkprop+elv) 5j. Ψ (fc+ppl) p (fc+ppl) 5k. Ψ (fc+ppl+tol) p (fc+ppl+tol) 5l. Ψ (fc+ppl+elv) p (fc+ppl+elv) 5m. Ψ (fc+tol) p (fc+tol) 5n. Ψ (fc+tol+elv) p (fc+tol+elv) 5o. Ψ (fc+ppl+park+elv) p (fc+ppl+park+elv) 5p. Ψ (fc+ppl+parkprop+elv) p (fc+ppl+parkprop+elv) 5q. Ψ (fc+tol+park+elv) p (fc+tol+park+elv) 5r. Ψ (fc+tol+parkprop+elv) p (fc+tol+parkprop+elv) 5s. Ψ (fc+tol+park+elv+ppl) p (fc+tol+park+elv+ppl) 5t. Ψ (fc+tol+parkprop+elv+ppl) p (fc+tol+parkprop+elv+ppl) Time Time + Models 2a to 5t, and Time + Time Squared + Models 2a to 5t Note: Covariates - park = park presence/absence, park prop= % age of cell covered, fc = forest cover presence/absence, elv = elevation (transformed), ppl = Log (human population density), tol = cultural intolerance, time = time (transformed). For every model additional variations with constant occupancy [Ψ (.) p (covariates)] and constant detection [Ψ (covariates) p (.) were run].

141

Appendix 2: Models included in the model sets for all 20 Species

Models Occupancy and Detection Constant 1. Ψ (.) p (.) Protected Areas 2a. Ψ (park) p (park) 2b. Ψ (park prop) p (park prop) 2c. Ψ (park+fc) p (park+fc) 2d. Ψ (park * fc) p (park * fc) People 3a. Ψ (ppl) p (ppl) 3b. Ψ (tol) p (tol) Elevation 4a. Ψ (elv) p (elv) 4b. Ψ (elv+park prop) p (elv+park prop) Land-cover 5. Ψ (evergreen) p (evergreen) and 6. Ψ (temperate) p (temperate) Land-use 7. Ψ (deciduous) p (deciduous) 8. Ψ (scrub-grassland) p (scrub-grassland) 9. Ψ (barren) p barren) 10. Ψ (cultivated) p (cultivated) 11. Ψ (rural-urban) p (rural-urban) 12a. Ψ (10lc) p (10lc) 12b. Ψ (10lc+park prop) p (10lc+park prop) 12c. Ψ ( 10lc+ppl) p (10lc+ppl) 12d. Ψ (10lc+tol) p (10lc+tol) 12e. Ψ (10lc+ppl+tol) p (10lc+ppl+tol) 12f. Ψ (10lc+ppl+park prop) p (10lc+ppl+park prop) 12g. Ψ (10lc+tol+park prop) p (10lc+tol+park prop) 12h. Ψ (10lc+tol+park prop+ppl) p (10lc+tol+park prop+ppl) Global Model 13. Ψ (10lc+tol+park prop+elv+ppl) p (10lc+tol+park

prop+elv+ppl)

Note: Models 2c, 4b, 12a – 12h are additive, Model 2d is multiplicative, and Model 13 is the global model.

142

Appendix 3: Models included in the model sets for groups of species

Group Occupancy and Detection Group 1 4lc = Evergreen + Deciduous + Temperate + Scrub/Grassland 14a. Ψ (4lc) p (4lc) 14b. Ψ (4lc+park prop) p (4lc+park prop) 14c. Ψ (4lc+ppl) p (4lc+ppl) 14d. Ψ (4lc+tol) p (4lc+tol) 14e. Ψ (4lc+elv) p (4lc+elv) 14f. Ψ (4lc+ppl+tol) p (4lc+ppl+tol) 14g. Ψ (4lc+ppl+park prop) p (4lc+ppl+park prop) 14h. Ψ (4lc+tol+park prop) p (4lc+tol+park prop) 14i. Ψ (4lc+tol+park prop+ppl) p (4lc+tol+park prop+ppl) 14j. Ψ (4lc+tol+park prop+ppl+elv) p (4lc+tol+park prop+ppl+elv) Group 2 7lc = Evergreen +Deciduous + Scrub/Grassland + Barren + Cultivated + Temperate + Salt Pan 15a. Ψ (7lc) p (7lc) 15b. Ψ (7lc+park prop) p (7lc+park prop) 15c. Ψ (7lc+ppl) p (7lc+ppl) 15d. Ψ (7lc+tol) p (7lc+tol) 15e. Ψ (7lc+elv) p (7lc+elv) 15f. Ψ (7lc+ppl+tol) p (7lc+ppl+tol) 15g. Ψ (7lc+ppl+park prop) p (7lc+ppl+park prop) 15h. Ψ (7lc+tol+park prop) p (7lc+tol+park prop) 15i. Ψ (7lc+tol+park prop+ppl) p (7lc+tol+park prop+ppl) 15j. Ψ (7lc+tol+park prop+ppl+elv) p (7lc+tol+park prop+ppl+elv) Group 1: Black bear, brown bear, chital, elephant, gaur, leopard, muntjac, rhino, sambar, sloth bear, sun bear, tiger, wild dog, wild pig Group 2: Blackbuck, chinkara, hyena, jackal, nilghai, wolf

143

References

Aflalo, A. F. 1904. The Sportsman’s Book for India. Horace Marshall and Sons. UK.

Ali, S. 1934. Hyderabad state. In The Preservation of Wildlife in India. BNHS, Diocesan Press. Madras, India.

Alvard, M. A., Robinson, J. G., Redford, K. H., Kaplan, H. 1997. The sustainability of subsistence hunting in the Neotropics. Conservation Biology. 11: 977-982.

Bailey, L. L., Simons, T. R., Pollock, K. H. 2004. Estimating site occupancy and species detection probabilities for terrestrial salamanders. Ecological Applications. 14: 692- 702.

Baker, S. A. 1891. Wild Beasts and Their Ways. Macmillan and Co. London, UK.

Bandara, R., Tisdell, C. 2003. Comparison of rural and urban attitudes to the conservation of Asian elephants in Sri Lanka: empirical evidence. Biological Conservation. 110: 327-342.

Bhargav, P. 2007. Four hectares of forest: Correcting history or destroying collective future. SANDEE Newsletter. 14:9-10.

Bartholomé E., Belward, A.S. 2005. GLC2000: a new approach to global land cover mapping from Earth observation data. International Journal of Remote Sensing. 26: 11051- 11077.

Barve, N., Kiran, M. C., Vanaraj, G., Aravind, N. A., Rao, D., Shaanker, R. U., Ganeshaiah, K. N., Poulsen, J. G. 2005. Measuring and mapping threats to a wildlife sanctuary in southern India. Conservation Biology. 19: 122-130.

Bhaduri, B., Bright, E., Coleman, P., Dobson, J. 2002. LandScan: Locating people is what matters. Geoinformatics. 5:34-37.

Blythe E. 1863. Catalogue of the mammalia in the museum Asiatic society. Savielle and Cranenburgh. Bengal Printing Committee Ltd. Calcutta, India.

BNHS, 1934. The Wild Animals of India. Bombay Natural History Society. Bombay, India.

Bolaane, M. 2004. The impact of game reserve policy on the River BaSarwa bushmen of Botswana. Social Policy and Administration. 38: 399-417.

Brandis, D. 1897. Forestry in India. Origins and early developments. Oriental Institute. Poona, India. 144

Brashares, J. S., Arcese P., Sam, M. K. 2001. Human demography and reserve size predict wildlife extinction in West Africa. Proceedings of the Royal Society of London. 268: 2473-2478.

Brashares, J. S. 2003. Correlates of extinction in West Africa. Conservation Biology. 17: 734- 743.

Brechin, S. R.., Wilshusen, P. R., Fortwangler, C. L., West, P. C. 2002. Beyond the square wheel: toward a more comprehensive understanding of biodiversity conservation as a social and political process. Society and Natural Resources. 15: 41-64.

Brieman, L, Friedman, J.H., Olshen, R.A., Stone, C.J. 1984. Classification and Regression Trees. Wadsworth and Brooks/Cole. Monterey, USA.

Briggs, J. 2003. The biogeographic and tectonic history of India. Journal of Biogeography. 30: 381-388.

Brook, A., Zint, M., Young, R. de. 2003. Landowners’ response to an endangered species act listing and implications for encouraging conservation. Conservation Biology. 17: 1638-1649.

Brockington, D. 2002. Fortress Conservation. The Preservation of the Mkomazi Game Reserve, Tanzania. James Curry. Oxford, U. K.

Brockington, D. 2003. Injustice and conservation: is local support necessary for sustainable protected areas? Policy Matters. 12: 22-30.

Brockington, D. 2004. Community conservation, inequality and injustice. Myths of power in protected area management. Conservation and Society. 2: 411-432.

Brockington, D., Igoe, J. 2006. Eviction for conservation. A global overview. Conservation and Society. 4: 424-470.

Brosius, J. P. 2004. Indigenous peoples and protected areas at the World Parks Congress. Conservation. Biology 18: 609-612.

Brown, J. H., Stevens, G. C., Kaufman, D. M. 1996. The geographic range: size, shape, boundaries, and internal structure. Annual Review of Ecology and Systematics. 27: 597–263.

Brown, J. H., Kodric-Brown, A. 1977. Turnover rates in insular biogeography. The effect of immigration on extinction. Ecology. 58: 445-449. Burke, W. S. 1920. The Indian Field Shikar Book. Thacker, Spink and Co. Calcutta, India.

145

Burton, R. G. 1933. The Book of the Tiger. Hutchinson & Co Ltd. London, U. K.

Buckland, S. T., Anderson, S. T., Burnham, K. P., Laake, J. L., Borchers, D. L., Thomas, L. 2001. Introduction to Distance Sampling. Oxford University Press. Oxford, UK.

Budhendra, B., Bright, E., Coleman, P., Dobson, J. 2002. LandScan: Locating people is what matters. Geoinformatics. 5: 34–37.

Burnham, K. P., Anderson, D. R. 2002. Model Selection and Multimodal Inference: A Practical Information‐Theoretic Approach. Springer-Verlag. New York, USA.

Burnham, K. P., Anderson, D. R. 2004. Multimodal inference: understanding AIC and BIC in model selection. Sociological Methods and Research. 33: 261-304.

Cadell, P. 1934. The Indian Lion. In The Preservation of Wildlife in India. BNHS, Diocesan Press. Madras, India.

Chatty, D., Colchester, M. 2003. Introduction: Conservation and Mobile Indigenous Peoples. In Chatty, D., Colchester, M (eds). Conservation and Mobile Indigenous Peoples, Displacement, forced settlement and sustainable development. Berghan Books. New York, USA.

Cardillo, M., Purvis, A., Sechrest, W., Gittleman, J. L., Beilby, J., Mace, G. M. 2004. Human population density and extinction risk in the world’s carnivores. Public Library of Science Biology. 2: 909–914.

Cardillo, M., Mace, G. M., Jones, K. E., Bielby, J., Bininda-Emonds, O. R. P., Sechrest, W., Orme, C. D. L., Purvis, A. 2005. Multiple causes of high extinction risk in large mammal species. Science. 309: 1239–1241.

Cardillo, M., Mace, G. M., Gittleman, J. L., Purvis, A. 2006. Latent extinction risk and the future battle grounds of mammal conservation. Proceedings of the National Academy of Sciences. 103:4157-4161.

Caro, T., Mulder, M. B., Moore, M. 2003. Effects of conservation education on reasons to conserve biological diversity. Biological Conservation. 114: 143-152.

Champion, F. W. 1934. With a Camera in Tiger Land. Chatto and Windus. London, U. K.

Channell, R., Lomolino, M. 2000a. Trajectories to extinction: spatial dynamics of the contraction of geographic ranges. Journal of Biogeography. 27: 169-179.

Channell, R., Lomolino, M. 2000b. Dynamic biogeography and the conservation of endangered species. Nature. 403: 84-86.

146

Cirincione, C., Gurrieri, G. A. 1997. Research methodology: Computer-Intensive Methods in the Social Sciences. Social Science Computer Review. 15: 83-97.

Ceballos, G., Brown, J. H. 1995. Global patterns of mammalian diversity, endemism, and endangerment. Conservation Biology. 9: 559-568.

Ceballos, G., Ehrlich, P. R. 2002. Mammal population losses and the extinction crisis. Science. 296: 904-907.

Ceballos, G., Ehrlich, P. R., Soberon, J., Salazar, I., Fay, J. P. 2005. Global mammal conservation: what must we manage? Science. 309: 603–607.

CGIAR- CSI. 2004. SRTM data V1, International Centre for Tropical Agriculture (CIAT). ( http://srtm.csi.cgiar.org).

Corlett, R. T. 2007. The impact of hunting on the mammalian fauna of tropical Asian forests. Biotropica. 39: 292-303.

Das A., Krishnaswamy, J., Bawa, K. S., Kiran, M. C., Srinivas, V., Kumar, N. S., Karanth, U. K. 2006. Prioritization of conservation areas in the Western Ghats, India. Biological Conservation. 133: 16-31.

Datta, A. 2007. Protecting with people in Namdapha: threatened forests, forgotten people. In Shahabuddin, G., Rangarajan, M (eds). Making Conservation Work. Permanent Black. New Delhi, India.

Dobson, J. E., Bright, E. A., Coleman, P. R., Durfee, R.C., Worley, B. A. 2000. LandScan: A global population database for estimating populations at risk. Photogrammetric Engineering and Remote Sensing. 66: 8410-857.

Dorazio, R. M., Royle, J. A. 2005. Estimating size and composition of biological communities by modeling occurrence of species. Journal of American Statistical Association. 100: 389-398.

Dorazio, R. M., Royle, J. A., Soderstrom, B., Glimskar, A. 2006. Estimating species richness and accumulation by modeling species occurrence and detectability. Ecology 87: 842-854.

Dunbar-Brander, A. A. 1923. Wild Animals in Central India. Edward Arnold and Co. London, U. K.

Dunbar-Brander, A. A. 1934. The Central Provinces. In The Preservation of Wildlife in India. BNHS, Diocesan Press. Madras, India.

147

Ellison, B. C. 1925. The Prince of Wales’s Sport in India. William Heinemann Ltd. London, U. K.

Fagan, W. F., Meir, E., Prendergast, J., Folarin, A., Karieva, P. 2001. Characterizing population vulnerability for 758 species. Ecology Letters. 4: 132-138.

Fenton, L. L. 1921.The Rifle in India. Being the Sporting Experiences of an Indian Officer. W Thacker and Co. London, India.

Flint, E. P. 1998. Deforestation and land use in Northern India with a focus on Sal (Shorea robusta) forests 1880-1980. In Groves, R. H., Damodaran, V., Sangwan, S (eds). Nature and the Orient: The Environmental History of South and Southeast Asia. Oxford University Press. New Delhi, India.

Ferraz, G., Nichols, J.D., Hines, J.E., Stouffer, P.C., Bierregaard, R.O. Jr., Lovejoy, T. E. 2007. A large-scale deforestation experiment: Effects of patch area and isolation on Amazon birds. Science. 315: 238-241.

Forest Survey of India. 2000. State of the Forest Report. Forest Survey of India. Dehradun, India.

Ganguly, V. 2004. Conservation, Displacement and Deprivation. Maldhari of Gir Forest of Gujarat. Indian Social Institute. New Delhi, India.

Gaston K. J. 2000. Global patterns in biodiversity. Nature 405: 619-625.

Gee, E. P. 1964. The Wild Life of India. Collins. U. K.

Gillingham, S., Lee, P. 1999. The impact of wildlife-related benefits on the conservation attitudes of local people around the Selous Game Reserve, Tanzania. Environmental Conservation. 26:218-228.

Glasfurd, A. I. R. 1903. Leaves from an Indian Jungle. Gathered During Thirteen Years of a Jungle Life in the Central Provinces, the Deccan, and Berar. The Times Press. Bombay, India.

Goenka, D. 2005. A non-solution. Seminar. 552: 35-36.

Gu, W., Swihart, R. K. 2004. Absent or undetected? Effects of non-detection of species occurrence on wildlife-habitat models. Biological Conservation. 116: 195- 203.

Guha, R. 1983. Forestry in British and Post-British India: An historical analysis. Economic

148

And Political Weekly. 17: 1882-96.

Gurd, B. E., Nudds, T. D. Rivard, D. H. 2001. Conservation of mammals in eastern North American wildlife reserves: How small is too small. Conservation Biology. 15:1355-1363.

Habib, L.D., Wiersma, Y. F., Nudds, T. D. 2003. Effects of errors in range maps on estimates of historical species richness of mammals in Canadian national parks. Journal of Biogeography. 30: 375-380.

Hines, J. E. 2006. Program Presence (version 2). http://www.mbr- pwrc.usgs.gov/software/doc/presence/presence.html

Hess, K. R., Abbruzzese, M. C., Lenzi, R., Raber, M. N., Abbruzzese, J L., Abbruzzese, J L. 1999. Classification and regression tree analysis of 1000 consecutive patients with unknown primary carcinoma. Clinical Cancer Research. 5: 3403-3410.

Isaac, N. J. Turvey S.T., Collen, B., Waterman, C., Baillie, J.E. 2007. Mammals on the EDGE: Conservation Priorities Based on Threat and Phylogeny. PLoS ONE 2: e296. doi:10.1371/journal.pone.0000296.

Jacoby, K. 2001. Crimes Against Nature: Squatters, Poachers, Thieves and the Hidden History of American Conservation. University of California Press. Berkeley, USA.

Jayakrishnan, P. V. 2005. Is there a need for this bill? Seminar. 552: 23-29.

Jepson, S. 1937. Big Game Encounters. Critical Moments in the Lives of Well-Known Shikaris. H. F. and G. Witherby, Ltd. London, U. K..

Jerdan, T. C. 1874. A Handbook of Mammals of India. Reprint by Mittal Publications. India.

Joshi, P. K. K., Roy, P. S., Singh, S., Agrawal, S., Yadav, D. 2006. Vegetation cover mapping in India using multi-temporal IRS Wide Field Sensor (WiFS) data. Remote Sensing of Environment. 103: 1100-1112. Kabra, A. 2003. Displacement and Rehabilitation of an Adivasi settlement- Case of Kuno Wildlife Sanctuary, Madhya Pradesh. Economic and Political Weekly. 38: 3073-3087.

Kabra, A. 2006. Wildlife Protection. Reintroduction and Relocation. Economic and Political Weekly. 41: 1309-1310.

Kaczensky, P., Blazic, M., Gossow, H. 2004. Public attitudes towards brown bears (Ursus arctos) in Slovenia. Biological Conservation. 118: 661-674.

149

Kaltenborn, B. P., Bjerke, T., Nyahongo, J. W., Williams, D. R. 2006. Animal preferences and acceptability of wildlife management actions around Serengeti national park. Biodiversity and Conservation. 15: 4633-4649.

Karanth, K. K. 2003. Forest use and human-wildlife conflicts in Bhadra Wildlife Sanctuary, India. Tropical Resources Bulletin. 22: 44-54.

Karanth, K. K. 2005. Bhadra Wildlife Sanctuary: Addressing relocation and livelihood concerns. Economic and Political Weekly. 12: 4809-4811.

Karanth, K. K. 2007. Making resettlement work: The case of India’s Bhadra Wildlife Sanctuary. Biological Conservation. 139: 315-324.

Karanth, K. K., Curran, L. M., Reuning-Scherer, J. D. 2006. Village Size and Forest Disturbance in Bhadra Wildlife Sanctuary, Western Ghats, India. Biological Conservation. 128: 147-157.

Karanth, K. U. 1982. Bhadra wildlife sanctuary and its endangered ecosystem. Journal of Bombay Natural History. 79: 79-86.

Karanth, K. U. 2002. Nagarahole: Limits and Opportunities in Wildlife Conservation. In Terborgh, J., van Schaik, C., Davenport, L., Rao, M (eds). Making Parks Work: Strategies for Preserving Tropical Nature. Island Press. Washington, D.C., USA.

Karanth, K. U., Madhusudan, M. D. 2002. Mitigating human-wildlife conflicts in southern Asia. In Terborgh, J., van Schaik, C., Davenport, L., Rao, M (eds). Making Parks Work: Strategies for Preserving Tropical Nature. Island Press. Washington, D.C., USA.

Karanth K. U., Nichols J. D. 2002. Monitoring Tigers and their Prey. Centre for Wildlife Studies. Bangalore, India.

Karanth, K. U., Bhargav, P. 2005. De-fragmenting nature. Seminar. 552: 59-62.

Karanth, K. U., Karanth, K. K. 2007. Free to Move: Conservation and Resettlement in the Western Ghats of Karnataka, India, in: Redford, K. H., Fearn, E (eds). Protected Areas and Human Displacement: A Conservation Perspective. Working Paper 29. Wildlife Conservation Society. New York, USA.

Karanth, U. K., Nichols, J. D., Kumar, S. N., Link, W. A., Hines, J. E. 2004. Tigers and their prey: Predicting carnivore densities from prey abundance. Proceedings of the National Academy of Sciences. 101: 4854-4858.

150

Karisson, B. G. 1998. Ecodevelopment in practice: The Buxa tiger reserve and forest people. Economic and Political Weekly. 34: 2087-2092.

Karnataka Forest Department. 1999. Report on Rehabilitation in Bhadra Tiger Reserve. Bangalore, India.

Kleiven, J., Bjerke, T., Kaltenborn, B. P. 2004. Factors influencing the social acceptability of large carnivore behaviors. Biodiversity and Conservation. 13: 1647-1658.

Kothari, A. 2005. Bungle in the jungle. Seminar. 552: 63-79.

Kothari, A., Pathak, N. 2005. Forest and tribal rights. Frontline 22. India.

Kothari, A., Pande, P., Singh, S., D., Variava, D. 1989. Management of National Parks and Sanctuaries in India: A Status Report. Indian Institute of Public Administration. New Delhi, India.

Kramer, R.A., Sills, E.O., Pattanayak, S. K. 2008. National Parks as Conservation and Development Projects: Gauging Local Support. In Karachepone, N (ed.). Conserving and Valuing Ecosystem Services and Biodiversity. Earthscan Press. London, U. K.

Krishnaswamy, M. 2005. One Step Forward Two Steps Back. Economic and Political Weekly. 47.

Kumar, R., Shahabuddin, G. 2006. Effects of biomass extraction on vegetation structure, Diversity and composition of an Indian dry tropical dry forest. Environmental Conservation. 32: 1-12.

Kutty, R., Kothari, A. 2001. Protected Areas in India: A Profile. Kalpavriksh Publications. New Delhi, India.

Laliberte, A. S., Ripple, W. J. 2004. Range contractions of North American carnivores and ungulates. BioScience. 54: 123-138.

LandScan Global Population Data. 2000. (http://www.ornl.gov/gist)

MacKenzie, D. I., Nichols, J. D. 2004. Occupancy as a surrogate for abundance estimation. Animal Biodiversity and Conservation. 27: 461-467.

MacKenzie, D. I., Royle, J. A. 2005. Designing occupancy studies: general advice and allocating survey effort. Journal of Applied Ecology. 42: 1105-1114.

151

MacKenzie, D. I., Nichols, J. D., Lachman, G. B., Droege, S., Royle, J. A., Langtimm, C. A. 2002. Estimating site occupancy rates when detection probabilities are less than one. Ecology. 83: 2248-2255.

MacKenzie, D. I., Nichols, J. D., Hines, J. E., Knutson, M. G., Franklin, A. D. 2003. Estimating site occupancy, colonization and local extinction when a species is detected imperfectly. Ecology. 84: 2200-2207.

MacKenzie, D. I., Bailey, L. L., Nichols, J. D. 2004. Investigating species co-occurrence patterns when species are detected imperfectly. Journal of Animal Ecology. 73: 546– 555.

MacKenzie, D.I., Nichols, J.D., Royle, J.A., Pollock, K.H., Bailey, L.A., Hines, J.E. 2006. Occupancy Modeling and Estimation. Academic Press. San Diego, USA.

Madhusudan, M. D. 2003. Living with large wildlife: livestock and crop depredation by large mammals in the interior villages of Bhadra Tiger Reserve, southern India. Environmental Management. 31: 466-475.

Madhusudan, M. D. 2004. Recovery of wild large herbivores following livestock decline in a tropical Indian wildlife reserve. Journal of Applied Ecology. 41: 858-869.

Madhusudan, M. D. 2005. Of Rights and Wrongs. Economic and Political Weekly. 40: 4893- 4895.

Madhusudan, M. D., Karanth, K. U. 2000. Hunting for an answer: is local hunting compatible with wildlife conservation in India? In Robinson, J. G., Bennet, E. L (eds). Hunting for Sustainability in Tropical Forests. Columbia University Press. New York, USA.

Madhusudan, M. D., Mishra, C. 2003. Why Big, Fierce Animals are Threatened: Conserving Large Mammals in Densely Populated Landscapes. In Saberwal, V., Rangarajan, M (eds). Battles Over Nature: Science and Politics of Conservation. Permanent Black. New Delhi, India.

Madhusudan M. D., Karanth, U. K. 2005. Local hunting and large mammal conservation.In Jhala Y. V., Chellam R., Qureshi, Q (eds). Wildlife conservation, research and management. Wildlife Institute of India. Dehradun, India.

Madhusudan, M. D., Shankar, K., Kumar, A., Mishra, C., Sinha, A., Datta, A., Rangarajan,M., Chellam, R., Shahabuddin, G., Sankaran, R., Singh, M., Ramakrishnan, U., Rajan, P. D. 2006. Science in the wilderness: the predicament of scientific research in India’s wildlife research. Current Science. 91: 1015-1019.

McFarlane, B. L., Stumpf-Allen, C. G., Watson, D. O. 2006. Public perceptions of natural

152

disturbance in Canada’s national parks: The case of the mountain pine beetle (Dendroctonus ponderosae Hopkins). Biological Conservation. 130: 340-348.

McLean, J., Straede, S. 2003. Conservation, relocation and Paradigms of Park and People Management- a case study of Padampur Villages and Royal Chitwan National Park, Nepal. Society and Natural Resources. 16: 509-526.

Menon, V. 2003. A Field Guide to Indian Mammals. Dorling Kindersley India Pvt. Limited. Delhi, India.

Michalski, F., Peres, C. A. 2005. Anthropogenic determinants of primate and carnivore local extinctions in a fragmented forest landscape of southern Amazonia. Biological Conservation. 124: 383-396.

Michalski, F., Peres, C. A. 2007. Disturbance-mediated mammal persistence and abundance- area relationships in Amazonian forest fragments. Conservation Biology. 21:1626-40.

MOEF. 1999. National Policy and Macro level Action Strategy on Biodiversity Ministry of Environment and Forests, Government of India. New Delhi, India.

MOEF. 2000.Annual Report 2000-2001. New Delhi: Ministry of Environment and Forests, Government of India. New Delhi, India.

MOEF. 2003. The Wildlife Protection Act of 1972. As amended up to 2003. Wildlife Trust of India. Nataraj Publishers. Dehra Dun, India.

MOEF. 2005. Joining the Dots, Report of the Tiger Task Force. New Delhi: Ministry of Environment and Forests, Government of India. New Delhi, India.

Mohanty, B. 2005. View from the field. Seminar. 552: 30-34.

Monteath, G. 1934. The Bombay Presidency. In The Preservation of Wildlife in India. BNHS, Diocesan Press. Madras, India.

Morrison, J. M., Sechrest, W., Dinerstein, E., Wilcove, D. S., Lamoreux, J. L. 2007. Persistence of large mammal faunas as indicators of human impact. Journal of Mammalogy. 88: 1363-1380.

Newmark W. D. 1995. Extinction of mammal populations in western North American national parks. Conservation Biology. 9: 512-526.

Newmark, W. D. 1996. Insularization of Tanzanian parks and the local extinction of mammals. Conservation Biology. 10: 1549-1556.

153

Nichols, J. D., Karanth, U. K. 2002. Statistical concepts: estimating absolute densities of tigers using capture-recapture sampling. In Karanth U. K., Nichols J. D (eds). Monitoring Tigers and their Prey: A Manual for Researchers, Managers, and Conservationists in Tropical Asia. Centre for Wildlife Studies. Bangalore, India.

Pandian, M. S. S. 1998. Hunting and colonialism in the nineteenth-century Nilgiri hills of south India. In Groves, R. H., Damodaran, V, Sangwan, S (eds). Nature and the Orient: The Environmental History of South and Southeast Asia. Oxford University Press. New Delhi, India.

Parks, S. A., Harcourt, A. H. 2002. Reserve size, local human density, and mammalian extinctions in U.S protected areas. Conservation Biology. 16: 800-808.

Pautosso, M. 2007. Scale dependence of the correlation between human population presence and vertebrate and plant species richness. Ecology Letters. 10: 16-24.

Pimm, S. L., Jenkins, C. N. 2005. Sustaining the variety of life. Scientific American. 293: 66- 73.

Pollock, A. J. O. 1894. Sporting Days in Southern India: Being Reminiscences of Twenty Trips in Pursuit of Big Game, Chiefly in Madras Presidency. Horace Cox, and Windsor House. London, UK.

Prabhu, P., 2005. The right to live with dignity. Seminar. 552: 14-19.

Pratt, D. G., Mac Millan, D. C., Gordon, I. J. 2004. Local community attitudes to wildlife utilization in changing the economic and social context of Mongolia. Biodiversity and Conservation. 13: 591-613.

Prater, S. H. 1948. The book of Indian animals. Bombay Natural History Society. Bombay, India.

Pythian-Adams, E. G. 1934. Mysore. In The Preservation of Wildlife in India. BNHS, Diocesan Press. Madras, India.

Qian, S. S., Anderson, C. W. 1999. Exploring factors controlling the variability of pesticide concentrations in the Williamette river basin using tree-based models. Environmental Science and Technology. 33: 3332-3340.

Rajan, R. 1998. Imperial environmentalism or environmental imperialism? European forestry, colonial foresters and the agendas of forest management in British India 1800 - 1900. In Groves, R. H., Damodaran, V., Sangwan, S (eds). Nature and the Orient. Oxford University Press. New Delhi, India.

154

Ramnath, M. 2008. Surviving the Forest Rights Act Between Scylla and Charybdis. Economic and Political Weekly. 43: 37-42.

Rangarajan, M. 1996. Fencing the Forest. Conservation and Ecological Change in India’s Central Provinces 1860 - 1914. Oxford University Press. New Delhi, India.

Rangarajan, M. 1998. Production, Desiccation and Forest Management in the Central Provinces 1850 – 1930. In Groves, R. H., Damodaran, V., Sangwan, S (eds). Nature and the Orient: The Environmental History of South and Southeast Asia Oxford University Press. New Delhi, India.

Rangarajan M. 2001. India’s Wildlife History. Permanent Black. New Delhi, India.

Rangarajan, M. 2005. Fire in the Forest. Economic and Political Weekly. 40: 4888-4890.

Rangarajan, M. 2006. Ideology, the environment, and policy: Indira Gandhi. India International Centre Quarterly. 33: 50-64.

Rangarajan, M. 2007. Retreating elephant, vanishing tiger, the environmental legacies ofIndia and China. In: Pande, I. (Ed.). India at Sixty. Harper Collins. Delhi, India.

Rangarajan, M., Shahabuddin, G., 2006. Displacement and relocation from protected areas: Towards a biological and historical synthesis. Conservation and Society. 4: 359-378.

Ribbentrop, B. 1900. Forestry in British India. Calcutta Government Press. Calcutta, India.

Rodrigues, A. S. L. , Andelman, S. J., Bakarr, M. I., Boitani, L., Brooks, T. M., Cowling, R. M., Fishpool, L. D. C., da Fonseca, G. A. B., Gaston, K. J., Hoffmann, M.., Long, J. S., Marquet, P. A., Pilgrim J. D., Pressey, R. L., Schipper, J., Sechrest, W., Stuart, S. N., Underhill, L. G., Waller, R. W., Watts, M. E. J., Yan, X. 2004. Effectiveness of the global protected area network in maintaining species diversity. Nature. 428: 640- 643.

Rodgers, W. A., Panwar, H. S. 1988. Planning a Wildlife Protected Area Network in India. (vol. I). Wildlife Institute of India. Dehra Dun, India.

Roy, P.S., Joshi P. K., Singh S., Agrawal S., Yadav D., Jegannathan C. 2006. Biome mapping in India using vegetation type map derived using temporal satellite data and environmental parameters. Ecological Modeling. 1107: 148–158.

Royle, J. A. 2005. Site occupancy models with heterogeneous detection probabilities. Biometrics. 62: 97-102.

Royle, J. A., Nichols, J. D. 2003. Estimating abundance from repeated presence-absence data or point counts. Ecology. 84: 777-790.

155

Russell, C. E. M. 1900. Bullet and Shot in Indian Forest, Plain and Hill. Thacker and Co. Ltd. Bombay, India.

Saberwal, V., Rangarajan, M., Kothari, A. 2001. People, Parks and Wildlife. Towards Coexistence. Orient Longman Private Limited. New Delhi, India.

Saberwal, V., Rangarajan, M. 2003. Introduction. In: Saberwal, V., Rangarajan, M (eds). Battles Over Nature: Science and Politics of Conservation. Permanent Black. New Delhi, India.

Sahgal, B., Raman, L., Basu, B., Shah, S. 2005. Citizens of the forest. Seminar. 552: 44-49.

Sanderson, G. P. 1896. Thirteen years among the wild beasts of India. W. H Allen and Co. London, UK.

Sanderson, S. E., Redford, K. H. 2003. Contested relationships between biodiversity conservation and poverty alleviation. Oryx. 37: 1-2.

Sanderson, E. W., Jaiteh, M., Levy, M. A., Redford, K. H., Wannebo, A. V., Woolmer, G. 2003. The human footprint and the last of the wild. BioScience. 52: 891-904.

Sato, J. 2000. People in Between: Conversion and Conservation of Forest Lands in Thailand. Development and Change. 31: 155-177.

Schama, S. 1996. Landscape and Memory. Fontana Press. London, UK.

Schmidt-Soltau, K. 2003. Conservation-related Resettlement in Central Africa: Environmental and Social Risks. Development and Change. 34: 525-551.

Schaller, G. S. 1967. The Deer and the Tiger. University of Chicago Press. Chicago. USA.

Schipper et al. 2008. The status of the world's land and marine mammals: diversity, threat, and knowledge. Science. 322: 225- 30.

Sekhsaria, P. 2007. Conservation in India and the need to think beyond ‘Tiger vs Tribal’. Biotropica. 39: 575-577.

Shahabuddin, G., Kumar, R., Shrivastava, M. 2005. Pushed over the edge: village relocation from Sariska. Economic and Political Weekly. 40: 3563-3564.

Shahabuddin, G., Shah, A. 2003. Relocation of people from wildlife areas: Socio-economic and ecological issues. Economic and Political Weekly. 38: 4945-4946. 156

Sharma, A. 2003. Displacement from wildlife protected areas: Implications for conservation and livelihoods. The case of Kuno Wildlife Sanctuary, Madhya Pradesh. Social Change. 33: 89-114.

Shakespear, H. 1862. The Wild Sports of India. Smith, Elder and Co. London, UK.

Shyamsundar, P., Kramer, R. 1997. Biodiversity Conservation- At what cost? A study of households in the vicinity of Madagascar’s Mantandia National Park. Ambio. 26: 180-184.

Shyamsunder, S. 1974. Government of Karnataka Notification of Bhadra Wildlife Sanctuary (AFD. 25. FWL). Karnataka Government. Bangalore, India.

Sivaramakrishnan, K. 2003. Reflections on community involvement in conservation. In: Saberwal, V., Rangarajan, M (eds). Battles Over Nature: Science and Politics of Conservation. Permanent Black. New Delhi, India.

Simson, F. B. 1886. Letters on Sport in East Bengal. Taylor and Francis. London, UK.

Skaria, A. 1998. Timber conservancy, desiccation and scientific forestry: The Dangs 1840s- 1920s. In Groves, R. H., Damodaran, V., Sangwan, S (eds). Nature and the Orient: The Environmental History of South and Southeast Asia. Oxford University Press. New Delhi, India.

Soto, B., Munthali, S. M., Breen, C. 2001. Perceptions of the forestry and wildlife policy by the local communities living in the Maputo Elephant Reserve, Mozambique. Biodiversity and Conservation. 10: 1723-1738.

Stebbing, E. P. 1920. The Diary of a Sportsman Naturalist in India. John Lane, the Bodely Head. London, UK.

Stebbing, E. P. 1923a. The Forests of India. Volume I. John Lane, London, UK.

Stebbing, E. P. 1923b. The Forests of India. Volume II. John Lane, London, UK.

Stebbing, E. P. 1923c. The Forests of India. Volume III. John Lane, London, UK.

Sterndale, R. A. 1887. Seonee; Camp Life on the Satpura Range. Thacker, Spink and Co. Calcutta, India.

Sreekantaiah, G. N., Subramanya, S. 1992. Bhadra Wildlife Sanctuary Acquisition, Rehabilitation and Resettlement Project Report. Karnataka Forest Department. Bangalore, India.

157

Taverna, K., Urban, D. L., McDonald, R. I. 2004. Modeling landscape vegetation patterns in response to historic land use: a hypothesis-driven approach for the North Carolina Piedmont, USA. Landscape Ecology. 20: 689-702.

Terborgh, J. W. 1988. The big things that run the world- a sequel to E. O. Wilson. Conservation Biology. 2: 402-403.

Terborgh, J. W., Lopez, L., Nuñez, P., Rao, M., Shahabuddin, G., Orihuela, G., Riveros, M., Ascanio, R., Adler, G. H., Lambert, T. D., Balbas, L. 2001. Ecological meltdown in predator-free forest fragments. Science. 294: 1923-1926.

Treves A., Karanth, U. K. 2003. Human-carnivore conflict and perspectives on carnivore management worldwide. Conservation Biology. 17: 1491-1499.

Toovey, J. 1987. Tigers of the Raj. Pages from Shikar Diaries 1894 to 1949 of Colonel Burton, Sportsman and Conservationist. Alan Sutton Publishing. Gloucester, UK.

Tyre A. J., Tenhumberg B., Field S. A., Niejalke, D., Parris K., Possingham, H. P. 2003. Improving precision and reduction bias in biological surveys, by estimating false negative error rates in presence-absence data. Ecological Applications. 13: 1790- 1801.

UN, 2006. Human Development Report. Beyond Scarcity: Power, Poverty, and the Global Water Crisis. United Nations Development Programme. Palgrave Macmillan. New York, USA . Wardrop, A.E. 1923. Days and Nights with Indian Big Game. MacMillan and Co. London, UK.

Wells, S. 2002. The journey of man: A genetic odyssey. Princeton University Press. New Jersey, USA.

Wezel, van, M., Potharst, R. 2007. Improved customer choice predictions using ensemble methods. European Journal of Operational Research. 181: 436-452.

Williams, B. K., Nichols, J. D., Conroy, M. J. 2002. Analysis and Management of Animal Populations. Academic Press. San Diego, USA.

Wilshusen, P., Brechin, S. R., Fortwangler, C., West, P. C. 2002. Reinventing a square wheel: Critique of a resurgent "protection paradigm" in international biodiversity conservation. Society and Natural Resources. 15: 17-40.

Wilshusen, P. R., Brechin, S. R., Fortwangler, C. L., West, P. C., 2003. Conservation and Development at the turn of the Twenty-First Century. In Brechin, S. R., Wilshusen,

158

P. R., Fortwangler, C. L., West, P. C (eds). Contested Nature. Promoting International Biodiversity with Social Justice in the Twenty-first Century. State University of New York Press. Albany, USA.

Winter, S. J., Esler, K. J., Kidd, M. 2005. An index to measure the conservation attitudes of landowners towards Overberg Coastal Renosterveld, a critcally endangered vegetation type in the Cape Floral Kingdom, South Africa. Biological Conservation. 126: 283-394.

Wright, J. S., Stoner, K. K., Beckman, N , Corlett, R. T., Dirzo, R., Muller-Landau, H., Nuñez-Iturri, G., Peres, C. A., Wang, B. C. 2007. The Plight of Large Animals in Tropical Forests and the Consequences for Plant Regeneration. Biotropica. 39: 289- 291.

Woodroffe, R., Ginsberg G. R. 1998. Edge effects and extinctions of populations inside protected areas. Science. 280: 2126-2128.

World Database on Protected Areas. 2003 and 2007. {http://www.unep-wcmc.org/wdpa}. World Conservation Union (IUCN) and UNEP-World Conservation Monitoring Center (UNEP-WCMC).

Yoccoz, N. G., Nichols, J. D., Boulinier, T. 2001. Monitoring of biological diversity in space and time. Trends in Ecology and Evolution. 16: 446-453.

159

Biography

Krithi K. Karanth was born on March 3rd, 1979 in Mangalore, India. Her parents

Prathibha and Ullas Karanth encouraged her to be independent, and to discover a passion.

Her fascination for nature began at a very young age. She was lucky to see her first wild leopard and tiger in Nagarahole National Park (Karnataka, India) at less than 3 years of age.

Her holidays were spent with her father, getting to watch wildlife, and visiting parks in India.

She got to see her father tag and radio-track tigers when she was 8 years old. Her favorite

subjects in school were biology, history, and geography. She never imagined that she could combine all of her interests. She had a wonderful childhood growing up in Mysore and

Bangalore, going to the same school from grade 1 to 12 (Demonstration Multipurpose

School) with many wonderful teachers. She grew up with some intentions to pursue a career in law, veterinary medicine, or history.

Krithi moved to America in 1997 to pursue undergraduate degree at the University of Florida. She spent the first year majoring in Microbiology, and found that she disliked chemistry intensely. Thanks to some wise council from her parents, she switched her major to Environmental Science, and loved its interdisciplinary breadth. She met her undergraduate mentor Dr. Mike Binford who encouraged her early research efforts from 1999-2001.

Her undergraduate honors thesis analyzed land cover and land use change in Thailand using remote sensing and socio-economic variables. She graduated from UF in 2001 with dual degrees in Environmental Science (highest honors) and Geography (high honors), with a minor in

Economics. For her graduate education, she decided to pursue research that involved field

work.

160

Krithi pursued a Master of Environmental Science degree at Yale with Dr’s. Lisa

Curran and Oswald Schmitz. She came up with a project to exam human impacts inside

Bhadra Wildlife Sanctuary in India. Her initiation into the field was dramatic! She had a car accident on the second day, and lost 6 weeks of field time as her fractured left leg was immobilized. She used the recuperation time to add additional component to her study- resettlement of people from Bhadra. She completed her fieldwork and graduated from Yale in 2003. She published several papers from this research Karanth 2003, Karanth 2005,

Karanth et al. 2006 (see below).

Upon graduation, Krithi spent the following year working with Dr. Jim Nichols at

USGS Patuxent Wildlife Research Center. Her research focused on avian distribution patterns in North America (Karanth et al. 2006), and she also learned new methods.

Conversations with Dr. Ullas Karanth highlighted the huge gap in current knowledge about distribution of mammals in India. She decided to use newly developed occupancy modeling techniques to examine extinction and current distribution patterns of large mammals in

India.

In 2004, Krithi began her doctoral career at Duke. Her dissertation research draws on the fields of conservation biology, landscape ecology, biogeography, history, geography, and the social sciences. She has enjoyed collaborating with people in India and the U.S.

Along with her interest in working on large mammal distribution patterns in India, she

developed two more projects. She examined conservation attitudes and opinions in India.

For this paper she collaborated with Dr’s Randy Kramer, Song Qian, and Norm

Christensen. The results of this paper were recently published (Karanth et al. 2008). Her

161

other project on resettlement in Bhadra was an extension of her master’s work. She felt that

it was important to follow up on the relocated people as most other research efforts on relocation have not done so. The results from research on resettlement work conducted in

2002 and 2006 were published as a sole author paper (Karanth 2007). She was awarded

Nicholas School Dean’s first award for outstanding research paper by a graduate student in

May 2008.

Krithi is an avid traveler, yearns to see more of the world, and has visited 15

countries. She wants to see more of India and the U.S. She speaks 4 Indian languages. On

Dec 25th 2003, she married a fellow Yalie, Dr. Denis Ostrovsky in Bangalore, India. While at

Duke, Krithi also gave birth to her daughter Keya in March 2007, and has enjoyed juggling motherhood and research. Krithi also enjoys pottery, painting, dancing, and rediscovering nature with Keya.

Krithi will be a post doctoral fellow with Dr. Ruth Defries at Columbia University from January 2009, and will continue research in India.

Publications (while at Duke)

Karanth, K. K., Kramer, R, Qian, S, Christensen, N. L. 2008. Conservation Attitudes, Perspectives And Challenges in India. 2008. Biological Conservation 141: 2357- 367.

Karanth, K. K. 2007. Making Resettlement Work: The Case of India’s Bhadra Wildlife Sanctuary. Biological Conservation 139: 315-324. (First Nicholas School Dean’s Award for Outstanding Student Paper 2008)

Karanth, U. K., Karanth, K. K. 2007. Free to Move: Conservation and Voluntary Resettlement in The Western Ghats of Karnataka, India. Protected Areas and Human Displacement: A Conservation Perspective. Editors K. H. Redford and E. Fearn. 29: 48-59.

Karanth, K.K., Curran, L.M., Reuning-Scherer, J.D. 2006.Village Size and Forest

162

Disturbance in Bhadra Wildlife Sanctuary, Western Ghats, India. Biological Conservation 128: 147-157.

Karanth, K. K., Nichols, J. D., Sauer, J. R., Hines, J. E. 2006. Comparative Community Dynamics of North American Bird Communities Across Edge and Interior Habitats. Journal of Biogeography 33: 674-682.

Karanth, K. K. 2005. Addressing Relocation and Livelihood Concerns: Bhadra Wildlife Sanctuary. Economic and Political Weekly 40: 4809-4811.

Karanth, U. K., Karanth, K. K. Conservation and Voluntary Resettlement in Karnataka, India. Conservation and Society (accepted).

Karanth, K. K., Nichols, J. D., Hines, J. E., Karanth, U. K., Christensen, N. L. Where are India’s Large Mammals and Why? Patterns and Determinants of Species Occurrence. Journal of Applied Ecology (in review).

Funding (Awards and Grants)

First Dean’s Award (Nicholas School of Environment) for Outstanding Research Paper 2008 Sigma Xi (Duke Chapter) Service 2006 - 2007 Preparing Future Faculty Fellow, Duke University 2006 - 2007 PEO International Peace Scholarship for Women 2005 - 2006, 2004 - 2005 Teaching Fellowship, Duke University 2004 - 2008

F.K. Weyerhaeuser Forest History Society Fellowship 2007 - 2008 Duke International Travel Award 2007 AZFA Clark Waldram Conservation Fund 2006 - 2007 Duke International Travel Award 2006 Conservation, Food and Health Foundation 2005 - 2006 Conservation and Research Small Grants Program, Cleveland Zoo 2005 - 2006 Sophie Danforth Conservation Fund, Roger Williams Park Zoo Grant 2005 - 2006 Duke International Travel Award 2005 IDEA Wild Research Fund 2005

163