Ecology and conservation of large carnivores in a human-dominated landscape in Eastern Anatolia

Conservation of biodiversity is rapidly changing as a result of increased impact of human activity on the natural world. At the beginning of a new epoch - the Anthropocene - the cumulative effect of population growth and natural resource consumption has left no corner of the planet unaffected by humans. Impacts can be observed on a global scale, such as climate change, ocean acidification, and nitrification and also on a local scale including habitat destruction, community composition, and pollution. These impacts are restructuring ecosystems into novel systems that require creative approaches to conserve ecosystem processes and maintain biodiversity. Large mammalian carnivores represent a clade of organisms that has a varied ability to survive in human dominated landscapes. For my dissertation, I examined community structure, movement, and abundance of brown bears (Ursus arctos arctos), gray wolves (Canis lupus), and Caucasian lynx (Lynx lynx dinniki) in a human dominated landscape in eastern Turkey. From 2013-2016, I surveyed for all medium-large mammal species using remote cameras deployed in a fragmented forest patch near Sarikamiş, Turkey. Occupancy estimates reveal a mammal community dominated by large carnivores, humans and livestock, and lacking a natural prey base. During 2011-2016, I collared 28 bears, 11 wolves and 2 lynx and used species-specific seasonal resource selection functions to assess habitat selection patterns. I found that all three species use of habitat varies between seasons and is strongly linked to elevation and slope. By identifying critical habitat for all three species, I have prioritized a specific area for conservation efforts in the future. To estimate the minimum population size of brown bears in my main study area, during 2013-2015, I used scat detection dogs to collect 1,520 bear scat samples for genetic analysis, and using 8 polymorphic microsatellite loci, I identified 27 unique multilocus genotypes and expected heterozygosity of 0.70 as a proxy of genetic diversity. I also conducted opinion surveys in 2014 and combined results with surveys conducted 2006 and 2010 to understand the perspective of the local community about large carnivores. Lastly, I propose a prioritized list of future conservation plans for large carnivore conservation in the human-dominated landscape of Sarikamiş Forest, eastern Anatolia.

ECOLOGY AND CONSERVATION OF LARGE CARNIVORES IN A HUMAN-DOMINATED LANDSCAPE IN EASTERN ANATOLIA by Mark William Chynoweth A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Biology The University of Utah December 2017 Copyright © Mark William Chynoweth 2017 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Mark William Chynoweth has been approved by the following supervisory committee members: Çağan H. Şekercioğlu , Chair 4/24/2017 Date Approved Frederick R. Adler , Member 4/24/2017 Date Approved , Member Paul Beier 5/15/2017 Date Approved Jack Longino , Member 4/24/2017 Date Approved William Newmark , Member 4/24/2017 Date Approved and by , Chair/Dean of M. Denise Dearing the Department/College/School of and by David B. Kieda, Dean of The Graduate School. Biology ABSTRACT Conservation of biodiversity is rapidly changing as a result of increased impact of human activity on the natural world. At the beginning of a new epoch - the Anthropocene - the cumulative effect of population growth and natural resource consumption has left no corner of the planet unaffected by humans. Impacts can be observed on a global scale, such as climate change, ocean acidification, and nitrification and also on a local scale including habitat destruction, community composition, and pollution. These impacts are restructuring ecosystems into novel systems that require creative approaches to conserve ecosystem processes and maintain biodiversity. Large mammalian carnivores represent a clade of organisms that has a varied ability to survive in human dominated landscapes. For my dissertation, I examined community structure, movement, and abundance of brown bears (Ursus arctos arctos), gray wolves (Canis lupus), and Caucasian lynx (Lynx lynx dinniki) in a human dominated landscape in eastern Turkey. From 2013-2016, I surveyed for all medium-large mammal species using remote cameras deployed in a fragmented forest patch near Sarıkamış, Turkey. Occupancy estimates reveal a mammal community dominated by large carnivores, humans and livestock, and lacking a natural prey base. During 2011-2016, I collared 28 bears, 11 wolves and 2 lynx and used speciesspecific seasonal resource selection functions to assess habitat selection patterns. I found that all three species use of habitat varies between seasons and is strongly linked to elevation and slope. By identifying critical habitat for all three species, I have prioritized a specific area for conservation efforts in the future. To estimate the minimum population size of brown bears in my main study area, during 2013-2015, I used scat detection dogs to collect 1,520 bear scat samples for genetic analysis, and using 8 polymorphic microsatellite loci, I identified 27 unique multilocus genotypes and expected heterozygosity of 0.70 as a proxy of genetic diversity. I also conducted opinion surveys in 2014 and combined results with surveys conducted 2006 and 2010 to understand the perspective of the local community about large carnivores. Lastly, I propose a prioritized list of future conservation plans for large carnivore conservation in the human-dominated landscape of Sarıkamış Forest, eastern Anatolia. iv TABLE OF CONTENTS ABSTRACT ....................................................................................................................... iii LIST OF TABLES ........................................................................................................... viii ACKNOWLEDGEMENTS ................................................................................................ x Chapters 1. INTRODUCTION ...................................................................................................................... 1 1.1 Large Carnivores ....................................................................................................... 1 1.2 Human-Dominated Ecosystems ................................................................................ 3 1.3 Large Carnivore Conservation in the Anthropocene ................................................ 3 1.4 Dissertation Chapter Summary ................................................................................. 5 1.5 References ................................................................................................................. 7 2. IMPROVING THE USE OF CAMERA TRAPS IN ECOLOGY AND CONSERVATION ........................................................................................................... 10 2.1 Abstract ................................................................................................................... 10 2.2 Introduction ............................................................................................................. 11 2.3 Literature Review Methods and Results ................................................................. 12 2.4 Biological Objectives .............................................................................................. 13 2.5 Importance of Camera Trapping for Biodiversity Monitoring and Conservation .. 22 2.6 Future of Camera Trapping in Conservation Biology and Ecology ....................... 24 2.7 References ............................................................................................................... 26 3. LARGE CARNIVORE HYPERABUNDANCE IN AN EMPTY FOREST .............. 42 3.1 Abstract ................................................................................................................... 42 3.2 Introduction ............................................................................................................. 43 3.3 Methods................................................................................................................... 46 3.4 Results ..................................................................................................................... 50 3.5 Discussion ............................................................................................................... 52 3.6 Conclusion .............................................................................................................. 56 3.7 References ............................................................................................................... 57 4. MOVEMENT ECOLOGY AND RESOURCE SELECTION OF THREE LARGE CARNIVORES IN A PREY-DEFICIENT, HIGHLY DEGRADED ECOSYSTEM...... 72 4.1 Abstract ................................................................................................................... 72 4.2 Introduction ............................................................................................................. 73 4.3 Methods................................................................................................................... 76 4.4 Results ..................................................................................................................... 83 4.5 Discussion ............................................................................................................... 85 4.6 Conclusion .............................................................................................................. 88 4.7 References ............................................................................................................... 89 5. MINIMUM POPULATION SIZE AND GENETIC DIVERSITY OF BROWN BEARS WITHIN A FRAGMENTED POPULATION IN EASTERN TURKEY ........ 103 5.1 Abstract ................................................................................................................. 103 5.2 Introduction ........................................................................................................... 104 5.3 Methods................................................................................................................. 106 5.4 Results ................................................................................................................... 110 5.5 Discussion ............................................................................................................. 111 5.6 Conclusions ........................................................................................................... 113 5.7 References ............................................................................................................. 114 6. HUMAN-WILDLIFE CONFLICT AS A BARRIER TO LARGE CARNIVORE MANAGEMENT AND CONSERVATION IN TURKEY ........................................... 124 6.1 Abstract ................................................................................................................ 125 6.2 Introduction .......................................................................................................... 125 6.3 Materials and Methods ......................................................................................... 128 6.4 Results .................................................................................................................. 130 6.5 Discussion ............................................................................................................ 132 6.6 Acknowledgements .............................................................................................. 135 6.7 References ............................................................................................................ 136 7. RECOMMENDATIONS FOR FUTURE CONSERVATION EFFORTS IN HUMANDOMINATED LANDSCAPES: THE CASE STUDY OF SARİKAMİŞ ..................... 137 7.1 Abstract ................................................................................................................. 137 7.2 Introduction ........................................................................................................... 138 7.3 Large Carnivore Management in Sarıkamış-Allahuekber Mountains National Park in the Caucasus Global Biodiversity Hotspot ...................................... 146 7.4 Conclusions ........................................................................................................... 158 7.5 References ............................................................................................................. 160 Appendices A. CONSERVATION OF A NEW BREEDING POPULATION OF CAUCASIAN vi LYNX (LYNX LYNX DINNIKI) IN EASTERN TURKEY ............................................ 166 B. WOLF DIET IN AN AGRICULTURAL LANDSCAPE OF NORTHEASTERN TURKEY ........................................................................................................................ 170 C. ANTHROPOGENIC FOOD RESOURCES FOSTER THE COEXISTENCE OF DISTINCT LIFE HISTORY STRATEGIES: YEAR-ROUND SEDENTARYAND MIGRATORY BROWN BEARS................................................................................... 177 vii LIST OF TABLES Tables 2.1 Camera trap and animal ecology keywords used in the ISI Web of Knowledge literature search at the University of Utah. ....................................................................... 39 2.2. Number of camera trapping articles published in the top ten journals from 1975-2014 based on a systematic search of key terms in ISI Web of Knowledge. ............................ 40 2.3. Proportion of target taxa in camera trapping studies published from 1975-2014 based on a systematic search of key terms in ISI Web of Knowledge. ...................................... 41 3.1. Wild mammal species documented in Sarıkamış forest, eastern Turkey cameratrapping survey conducted 2004-2016 in Sarıkamış forest............................................... 68 3.2. Camera trapping effort in Sarıkamış forest, eastern Turkey conducted 2004-2016 in Sarıkamış forest. ............................................................................................................... 69 3.3. Modeled occupancy and detection probabilities for 2013 camera trapping effort in Sarıkamış forest, eastern Turkey....................................................................................... 70 3.4. Modeled occupancy and detection probabilities for 2014 camera trapping effort in Sarıkamış forest, eastern Turkey. Modeling criteria A is naïve occupancy estimate >0.1, modeling criteria B is modeled detection probability 0.1. ................................................ 70 3.5. Modeled occupancy and detection probabilities for 2015 camera trapping effort in Sarıkamış forest, eastern Turkey....................................................................................... 71 3.6. Modeled occupancy and detection probabilities for 2016 camera trapping effort in Sarıkamış forest, eastern Turkey....................................................................................... 71 4.1. Description and characteristics of environmental variables used to model the probability of occurrence of bears, wolves, and lynx in Sarıkamış forest, eastern Turkey. ............................................................................... 101 4.2. Home range sizes for all bears, wolves and lynx captured in Sarıkamış forest, eastern Turkey from 2011-2014. ................................................................................................. 102 5.1. Summary of all scat samples encountered in a survey conducted from 2013-2015 in Sarıkamış, eastern Turkey using trained scat-detection dogs. ........................................ 122 5.2. Microsatellite marker variability, expected heterozygosity and observed number of alleles of brown bears in Sarıkamış forest, eastern Turkey, 2013. ................................. 123 6.1. Extant carnivore species in Turkey .......................................................................... 126 B.1. Composition of wolf diet in SAM NP and surrounding forest in north-eastern Turkey, based on scats analysis (n=72; May-June 2013). .............................................. 173 C.1. Population level coefficients, estimated standard errors and variance of random coefficients from a mixed conditional logistic regression of movement steps on six different environmental variables during the sendentary, stopover and roaming phases................................................................................................................ 184 ix ACKNOWLEDGEMENTS There are many people who have helped me to reach this achievement and it is impossible to list (or remember) all parties who deserve recognition. Thank you to my PhD advisor, Çağan H. Şekercioğlu, who provided me with guidance, support and positivity. Thank you to my advisory committee, Fred Adler, Paul Beier, Jack Longino, and Bill Newmark whose collective knowledge led me to success. Thank you to those people in my personal life who supported me during this journey, especially my future wife, Laura. She supported me throughout this long arduous passage no matter if I was in the field for months at a time or spending weekends in the lab. Thank you to my lab mates and fellow graduate students at the University of Utah for intellectual discussions, healthy distractions, and new adventures. Lastly, working in a foreign country can be challenging at times. Thank you to KuzeyDoğa staff and volunteers for their tireless efforts through the years and to the people of Sarıkamış for their hospitality. We thank the Christensen Fund, National Geographic Society Education Foundation, UNDP Small Grants Programme, the University of Utah and the Whitley Fund for their financial support. CHAPTER 1 INTRODUCTION Here, I provide a concise description of the three major themes of my doctoral dissertation: (1.1) large carnivores, (1.2) human-dominated landscapes, and (1.3) large carnivore conservation in the Anthropocene, and summarize the chapters of this dissertation (1.4). 1.1 Large Carnivores Large carnivores are a distinct group of mammals defined by body mass and diet; they are naturally rare and typically occupy the upper trophic levels of food webs. Some species are well-studied, and their ecology and natural history are well-known, while other species are among the least studied in the world because they are cryptic, inhabit remote regions of the world, or present challenges to observe or study. Some of these factors contribute to the lack of data and subsequently, to the somewhat contentious issue of how these animals impact the ecosystems they inhabit. Contemporary research suggests that apex predators play a central role in shaping ecosystems through trophic cascades and other top-down controls (Estes et al. 2011, Ripple et al. 2014, Svenning et al. 2016). However, some researchers have questioned the quality of data and subsequent conclusions about the impact of large carnivores on ecosystem function (Mech 2012, 2 Dobson 2014). A limiting factor for many large carnivore studies - due to traits described above - is the small sample size and short duration of most research initiatives (Ripple et al. 2014, Allen et al. 2017). Moving forward, more rigorous and long-term studies in understudied regions of the world will help us understand how large carnivores shape the ecosystems they inhabit (Ford and Goheen 2015). Human-wildlife conflict is another critical barrier to large carnivore conservation and management (Treves and Karanth 2003, Chynoweth et al. 2016). This conflict is a result of different stakeholder opinions, along a broad ideological gradient, about carnivore presence, management, and conservation. These cultural and political attitudes impact large carnivore science and management more than those of most other species. This is particularly challenging because the presence of large carnivores often leads to polarization between people who value these animals as charismatic megafauna and those who view them as a threat to livestock husbandry, game populations, or human wellbeing (Sjölander-Lindqvist et al. 2015). Regardless of personal opinions about large carnivores, a scientific approach is crucial to accurately assess their roles in our changing planet, and we must be cautious not to use science to justify personal beliefs (Mech 2012). The reality is that the science of large carnivores in human-dominated landscapes is undeveloped, and many seemingly fundamental concepts and terms, such as trophic cascade and apex predator are still being defined (Wallach et al. 2015, Ripple et al. 2016). Furthermore, as we struggle to understand large carnivores' relationships and impacts, most populations are in global decline (Ripple et al. 2014), adding to the urgency of taking conservation action. 3 1.2 Human-Dominated Ecosystems Human impact on ecosystems ranges from global change phenomena, such as climate change, alteration of nutrient cycles, and invasive species (Vitousek et al. 1997), to local impacts including water quality, habitat destruction, and defaunation (Dirzo et al. 2014). All these impacts are increasing due to a burgeoning human population, which, coupled with increasing global natural resource consumption per capita, is driving the restructuring of ecological relationships and the creation of novel and hybrid ecosystems (Estes et al. 2011). In these modified landscapes, wildlife species face new challenges and some species thrive while other species become locally extinct. Though research on large carnivore species in human-dominated landscapes is becoming more frequent, most studies maintain a traditional focus on species relationships in natural areas (Kuijper et al. 2016). Furthermore, when biodiversity patterns in human-dominated landscapes are studied, they are usually conducted in North America and Western Europe. Developing nations, where population growth and land alteration are potentially greatest (Bilsborrow and Ogendo 1992), are disproportionately understudied yet harbor some of the world's most important and threatened biodiversity (IUCN 2016). Therefore, as a scientific community, we are data-rich in a small proportion of protected areas and data-poor in human-dominated landscapes, which currently encompass most of the world. 1.3 Large Carnivore Conservation in the Anthropocene The science behind large carnivore ecology, conservation, and management has evolved to incorporate a wide variety of strategies and solutions to achieve established goals. With the technology available today - given unlimited financial resources - an 4 individual or agency can produce comprehensive data to inform scientists and managers about carnivore movement, resource use, density, gene flow, and other important aspects of wildlife ecology. In the coming years, large carnivore research needs to focus more on populations in human-dominated landscapes and novel ecosystems to understand mammal community structure and how large carnivores impact ecosystems. Large carnivores have varying levels of success in these novel ecosystems (Kuijper et al. 2016), depending on species-specific life history traits and habitat requirements. A major advantage for any species in the Anthropocene is the ability to exploit anthropogenic food resources, particularly predictable sources (Oro et al. 2013, Newsome et al. 2015, Cozzi et al. 2016). Waste management, which yields large quantities of such predictable food sources, has become a major issue that is known to affect animal movements and populations (Sutherland et al. 2016). Many large carnivore species, (e.g., bears and wolves) are known to use these and other anthropogenic resources to sustain and even increase populations. One important concept that will be revisited in the final chapter of this dissertation is increasing the social carrying capacity of large carnivores. We are learning that some species are able to coexist with humans and in human-dominated landscapes. However, for successful, sustainable cohabitation to occur, humans must act on a desire to coexist, instead of merely creating the ecological carrying capacity to do so. The general population must understand that the presence of these animals on the landscape will result in some personal losses (e.g., livestock), but also that many tools available to them can reduce human-wildlife conflict and maintain large carnivore populations. 5 1.4 Dissertation Chapter Summary The overall objective of my field work was to understand the basic ecology of three large carnivore species, gray wolf (Canis lupus lupus), Eurasian brown bear (Ursus arctos arctos), and Caucasian lynx (Lynx lynx dinniki), in greatly understudied Caucasus and Iran-Anatolia global biodiversity hotspots in eastern Turkey and to use this information to guide conservation efforts in the region. All three species are widespread outside urban areas throughout Eurasia, and the same species (Canis lupus), related subspecies (Ursus arctos horribilis), or sister taxa (Lynx canadensis) inhabit large regions of North America. To address these objectives and hypotheses, six chapters, written as scientific manuscripts, entail the body of the dissertation. First, a review of camera trapping in ecology and conservation provides a detailed look at how this method is changing the field of conservation biology and is contributing to the conservation of mammals worldwide. This chapter was submitted as my preliminary exam for the PhD program in the Department of Biology at the University of Utah, and has been submitted to the journals Biological Conservation, Conservation Biology, and Environmental Conservation. This review is being revised for resubmission for publication. The third chapter of the dissertation is a conservation ecology camera trap study of the medium-large mammal community in our study region. As a long-term monitoring effort, this study represents a comprehensive summary of the mammal community and the unprecedented high levels of human activity in the Sarıkamış-Allahuekber Mountains National Park. It also documents, for the first time to my knowledge, an ecosystem that supports large carnivores in the absence of a natural prey base. This manuscript will be 6 submitted for publication in the journal Ecology. Third, a resource selection function for all three large carnivore species is developed to quantify how large carnivores are using resources within their home ranges. I used a species-specific, seasonal model to examine what resources drive large carnivore habitat selection in Sarıkamış Forest in eastern Turkey. Results are used to identify the largest tract of suitable habitat for all three species, which will become a priority to be designated as a protected area. This manuscript will be submitted for publication in the journal Conservation Biology. Fourth, I describe a genetic approach to obtaining a minimum population estimate for brown bears in the study area. This work is ongoing and our sample collection will allow us to estimate population size for multiple species and consider important concepts in conservation biology such as gene flow, genetic bottlenecks, subspecies status, and genetic connectivity. Preliminary results identified 27 unique multilocus genotypes which suggests a minimum population size of 27 bears. Fifth, opinion surveys were conducted with 942 people to help understand the opinion of local communities about large carnivores. Results suggest that human perceptions of wildlife are a barrier to conservation and management of wildlife populations in Sarıkamış forest. The research, education, and outreach framework outlined in the manuscript can be used to address human-wildlife conflict across Turkey and guide ongoing conservation efforts of Turkey's existing, and increasingly threatened, large carnivores. This manuscript has been published in the journal Turkish Journal of Zoology. A final chapter discusses the management implications of this work and how a multi- 7 faceted approach can provide the best guidance for large carnivore management and overall biodiversity conservation. As a conclusion, this chapter summarizes some of the main findings from previous chapters and also synthesizes this work to generate solutions for applied conservation in Sarıkamış forest. It also takes a broad approach to apply these results to similar areas around the world that are understudied, human-dominated, and at high risk of biodiversity loss. My overall goal in this dissertation is to provide a comprehensive analysis of large carnivore ecology, conservation and management in a human-dominated landscape in Eastern Anatolia. While this work focuses on one study site in an understudied region, results and conclusions stated here can be applied to many areas that large carnivores inhabit, including species not discussed here. As humans impact more wilderness areas across the planet, there will be more human-wildlife conflict and we will lose more species, unless we use ecology and conservation biology research to guide conservation efforts. 1.5 References Allen, B. L., L. R. Allen, H. Andrén, G. Ballard, L. Boitani, R. M. Engeman, P. J. S. Fleming, A. T. Ford, P. M. Haswell, R. Kowalczyk, J. D. C. Linnell, L. David Mech, and D. M. Parker. 2017. Can we save large carnivores without losing large carnivore science? Food Webs 12: 64-75. Bilsborrow, R. E., and H. W. O. O. Ogendo. 1992. Population-driven changes in land use in developing countries. Ambio 21:37-45. Chynoweth, M. W., E. Çoban, Ç. Altın, and Ç. H. Şekercioğlu. 2016. Human-wildlife conflict as a barrier to large carnivore management and conservation in Turkey. Turkish Journal of Zoology 40:972-983. Cozzi, G., M. Chynoweth, J. Kusak, E. Çoban, A. Çoban, A. Ozgul, and Ç. H. Şekercioğlu. 2016. Anthropogenic food resources foster the coexistence of distinct life history strategies: year-round sedentary and migratory brown bears. Journal of 8 Zoology 300:142-150. Dirzo, R., H. S. Young, M. Galetti, G. Ceballos, N. J. B. Isaac, and B. Collen. 2014. Defaunation in the Anthropocene. Science 345:401-406. Dobson, A. P. 2014. Yellowstone wolves and the forces that structure natural systems. PLoS Biology 12:e1002025. Estes, J. A., J. Terborgh, J. S. Brashares, M. E. Power, J. Berger, W. J. Bond, S. R. Carpenter, T. E. Essington, R. D. Holt, J. B. C. Jackson, R. J. Marquis, L. Oksanen, T. Oksanen, R. T. Paine, E. K. Pikitch, W. J. Ripple, S. A. Sandin, M. Scheffer, T. W. Schoener, J. B. Shurin, A. R. E. Sinclair, M. E. Soule, R. Virtanen, and D. A. Wardle. 2011. Trophic downgrading of planet Earth. Science 333:301-306. Ford, A. T., and J. R. Goheen. 2015. Trophic cascades by large carnivores : a case for strong inference and mechanism. Trends in Ecology & Evolution 30:725-735. IUCN. 2016. The IUCN Red List of Threatened Species. Version 2016-3. http://www.iucnredlist.org/. Kuijper, D. P. J., E. Sahlen, B. Elmhagen, S. Chamaille-Jammes, H. Sand, K. Lone, and J. P. G. M. Cromsigt. 2016. Paws without claws? Ecological effects of large carnivores in anthropogenic landscapes. Proceedings of the Royal Society B 283:1- 9. Mech, D. L. 2012. Is science in danger of sanctifying the wolf? Biological Conservation 150:143-149. Newsome, T. M., J. a. Dellinger, C. R. Pavey, W. J. Ripple, C. R. Shores, A. J. Wirsing, and C. R. Dickman. 2015. The ecological effects of providing resource subsidies to predators. Global Ecology and Biogeography 24:1-11. Oro, D., M. Genovart, G. Tavecchia, M. S. Fowler, and A. Martínez-Abraín. 2013. Ecological and evolutionary implications of food subsidies from humans. Ecology Letters 16:1501-1514. Ripple, W. J., J. A. Estes, R. L. Beschta, C. C. Wilmers, E. G. Ritchie, M. Hebblewhite, J. Berger, B. Elmhagen, M. Letnic, M. P. Nelson, O. J. Schmitz, D. W. Smith, a. D. Wallach, and A. J. Wirsing. 2014. Status and ecological effects of the world's largest carnivores. Science 343:1241484-1241484. Ripple, W. J., J. A. Estes, O. J. Schmitz, V. Constant, M. J. Kaylor, A. Lenz, J. L. Motley, K. E. Self, D. S. Taylor, and C. Wolf. 2016. What is a trophic cascade? Trends in Ecology and Evolution 31:842-849. Sjölander-Lindqvist, A., M. Johansson, and C. Sandström. 2015. Individual and 9 collective responses to large carnivore management: the roles of trust, representation, knowledge spheres, communication and leadership. Wildlife Biology 21:175-185. Sutherland, W. J., S. Broad, J. Caine, M. Clout, L. V. Dicks, H. Doran, A. C. Entwistle, E. Fleishman, D. W. Gibbons, B. Keim, B. LeAnstey, F. A. Lickorish, P. Markillie, K. A. Monk, D. Mortimer, N. Ockendon, J. W. Pearce-Higgins, L. S. Peck, J. Pretty, J. Rockström, M. D. Spalding, F. H. Tonneijck, B. C. Wintle, and K. E. Wright. 2016. A horizon scan of global conservation issues for 2016. Trends in Ecology & Evolution 31:44-53. Svenning, J.-C., P. B. M. Pedersen, C. J. Donlan, R. Ejrnæs, S. Faurby, M. Galetti, D. M. Hansen, B. Sandel, C. J. Sandom, J. W. Terborgh, and F. W. M. Vera. 2016. Science for a wilder Anthropocene: synthesis and future directions for trophic rewilding research. Proceedings of the National Academy of Sciences 113:898-906. Treves, A., and K. U. Karanth. 2003. Human-carnivore conflict and perspectives on carnivore management worldwide. Conservation Biology 17:1491-1499. Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human domination of Earth's ecosystems. Science 177. Wallach, A. D., I. Izhaki, J. D. Toms, W. J. Ripple, and U. Shanas. 2015. What is an apex predator? Oikos 124:1453-1461. CHAPTER 2 IMPROVING THE USE OF CAMERA TRAPS IN ECOLOGY AND CONSERVATION 2.1 Abstract Camera traps are a common tool in animal ecology research, helping answer questions on wildlife presence, abundance, trends, and conservation. Because they document elusive species, capture diurnal and nocturnal animals, and collect data in remote field locations without human presence, these motion-triggered camera traps are an effective, noninvasive biodiversity survey method often used in conservation monitoring. As ongoing technological advances allow cameras to collect continually more photos and video, analysis techniques for large amounts of data are also evolving. However, researchers often use camera traps without defining a specific conservation question or concern or considering alternate, more appropriate methods. In this review, we describe conservation and ecology questions suitable for camera trap studies and their importance for biodiversity monitoring and conservation assessments. By comparing camera traps to other methods, we outline how researchers can match biological questions with appropriate technology. 11 2.2 Introduction Digital camera traps are a relatively low-cost tool for research and management, with negligible impacts on target species or the environment (O'Connell et al. 2011). Over the past decade, use of camera traps in conservation biology, ecology, monitoring and biodiversity assessments has grown significantly (McCallum 2013, Burton et al. 2015) motivating a need to assess how this approach can achieve objectives for a range of research agendas. Without appropriate research design, advance planning, and power analyses, conservation biologists often collect high volumes of data that they are unable to use to either inform their management of vulnerable species and systems or answer the conservation questions that initiated their research (Hebblewhite and Haydon 2010). To understand which conservation goals can be best addressed using camera traps, clear research questions must first be identified. For efficient use of time and resources, researchers must distinguish the reasons for camera trapping programs and choose appropriate study designs and analyses for conservation monitoring programs (Jones et al. 2013). Our review of camera trapping studies from 1975 to 2014 (Appendix A) reveals that publications have increased at an exponential rate (Figure 2.1) around the world (Figure 2.2). Camera traps have been used to study mammals, birds, reptiles, and insects; however, 84% of the 529 scientific papers using camera traps targeted mammals. These studies had at least one of five primary objectives: documenting species presence/richness, quantifying activity patterns, estimating density, evaluating relative abundance, and/or estimating occupancy. In this paper, we discuss the appropriate use of camera traps for each of these objectives. Our goal is to inform conservation biologists of 12 the advantages and disadvantages of camera traps, to assist in appropriate choice of method and study design, and to review the contribution of camera trapping studies to conservation biology. 2.3 Literature Review Methods and Results 2.3.1 Literature Review Methods A systematic search of peer-reviewed literature published between 1975 and 2016 was conducted using search terms related to camera traps and animal ecology (Table 2.1). Every combination of these terms was used in a search of the ISI Web of Knowledge search engine. Each article was reviewed to confirm that it discussed camera traps. The database included: publication year, article title, journal name, target taxa, study country, paper type, primary objective, analysis, and number of camera trap days. 2.3.2 Literature Review Results Our literature search resulted in 529 papers published across 146 journals, with 47 journals having more than two articles (see top 10 journals in Table 2.2). Research was conducted in 70 different countries (Figure 2.2). Target taxa included mammals, birds, herpetofauna, insects and multiple taxa (Table 2.3). The majority of articles covered studies of mammals (84.1%), most of which belonged to the order Carnivora (56.2%), the majority of which were felids (58.8%). Peer-reviewed publications increased exponentially, with the first article published in 1993 and over 130 articles published in 2014. Conservation of species and ecosystems is frequently cited as the main focus, with "conservation" being included in the title or as a keyword in 87% of articles. The primary 13 objective of most studies was documenting species presence (36.7%). Papers also had primary objectives of estimating activity (21.0%), density (19.8%), relative abundance (8.7%), and occupancy (4.2%). Total survey effort, or camera-days, had a median of 1,789 days; however, 40.7% of articles did not report total number of days or survey effort. 2.4 Biological Objectives Camera traps have been used to monitor many species and objectives in animal ecology. Here we focus on the most common study objectives: presence, abundance, density, occupancy and activity. However, even for well-studied species such as tigers, few studies go beyond baseline assessments (Linkie et al. 2010). As this method evolves, conservationists are increasingly using these parameters to test hypotheses and address a range of questions including human impacts on wildlife (Main and Richardson 2002), monitoring biodiversity (Waldon et al. 2011), reproductive ecology (Farhadinia et al. 2009), and nest predation (Bayne and Hobson 1997, Beck and Terborgh 2002, Vilardell et al. 2012). 2.4.1 Documenting Species Presence Documenting species presence or absence--the objective of most camera trap studies-is crucial to discovery (Rovero et al. 2008), rediscovery (Yamada et al. 2010), confirmation (Lhota et al. 2012), and range expansion (Chynoweth et al. 2015) of known or unknown species. Presence is a powerful indicator pertinent to monitoring elusive and endangered species, and photos of these species are invaluable for education and public 14 outreach. Effects of human activity on species and ecosystem dynamics in remote areas (Muhly et al. 2011) and conservation threats, such as the impact of poachers on wildlife populations, can also be monitored (Jenks et al. 2012). Photographic evidence often renders species presence indisputable. However, photos of animals can be misinterpreted or indecipherable, leading to spurious claims of new species (Meijaard et al. 2006). These claims, along with apparent range expansions and rediscoveries, may be a result of lack of baseline information and increasing density of camera traps (Dobson and Nowak 2010). Researchers must acknowledge that nondetection (i.e., absence) is related to detection probability, which is almost always <1, and a few isolated individuals of target species may exist (Tilson et al. 2004). Though several established methods effectively document species presence, comparison studies suggest that camera traps have higher probabilities than hair tunnels (O'Connell et al. 2006, Paull et al. 2012), cubby boxes (O'Connell et al. 2006), patrol observations (Burton 2012), and line-transect surveys (Trolle et al. 2008) for detecting smaller, solitary, and nocturnal species. However, studies incorporating track plates have shown that species richness and recording rates correlate with camera trapping results (Espartosa et al. 2011). In some cases, track plates were more effective (Hackett et al. 2007) and detected more individuals (Rosas-Rosas and Bender 2012). Yet, with technological advances, remote cameras require less maintenance and may be more cost effective than track plates for studies >1 year (Ford et al. 2009). An alternative method for detecting presence is genotyping by scat collection, which has produced consistent (Galaverni et al. 2011) and sometimes better (Harrison et al. 2002) detection probabilities than camera traps. Under proper weather conditions, snowtracking surveys have the 15 highest probability of detection for species active in winter (Gompper et al. 2006). Study design for documenting species presence does not necessarily need to be systematic and can be targeted at specific sites or use species-specific baits to maximize detection probability. If the goal is to produce estimates of abundance or density, a different study objective and more rigorous study design is required. A minimum number of cameras is not required, but having more cameras increases detection probability. To maximize detection, complementary survey techniques should be used (e.g., sign, snow tracking). 2.4.2 Relative Abundance Index (RAI) Presence data can be used to generate a Relative Abundance Index (RAI) by summing detections for each species for all camera traps over all days, dividing it by the total number of camera trap nights, and typically multiplying this fraction by 100 (O'brien 2011). This approach is attractive to conservationists because of its simplicity; however, in recent years it has been scrutinized for being an inappropriate and unreliable method (Sollmann et al. 2013a). This index is known to produce biased estimates based on heterogeneous detection probabilities (Jennelle et al. 2002, Sollmann et al. 2013b), and as a result, its use needs to be justified as the only reasonable alternative to other methods (O'brien 2011). The application of the RAI relies on the assumption that this index is directly related to true species abundance (O'Brien et al. 2003). The majority of RAI studies aim to estimate abundance at a single point in time at a specific site (e.g., protected area), but this index has also been used as an abundance proxy to study a variety of ecological 16 processes including habitat use (Bowkett et al. 2008), human impacts (Kinnaird and O'brien 2012), temporal population dynamics (Jenks et al. 2011), and activity patterns (Ramesh et al. 2012). The main issue confronting RAI is the problem of detectability. Detectability varies among and within species and is considered a major source of bias (Larrucea et al. 2007). Variations in detection probability due to species differences, home range size, study design, and temporal patterns have all been shown to bias RAI estimates (Sollmann et al. 2013b). Calibration of abundances can improve reliability, but requires periodic calibration with independently derived estimates in a double sampling design (O'Brien et al. 2003), rather than from another site or species. Study design for RAI surveys should aim to limit the effect of variation in detection probabilities to account for the main deficiency of this approach (Sollmann et al. 2013b). Once the study area is determined, cameras should be placed at distances smaller than the home range diameter of target species to prevent false negatives. The number of cameras necessary depends on study area extent and target species, but should cover the area uniformly to maximize detection probability. To calculate RAI or trap rate, only species presence data and trapping effort are needed. Therefore, methods described in the previous section (3.1) can be used in lieu of camera traps. Track plates, hair traps, and other signs have produced reliable abundance estimates (Jhala et al. 2011) and may be appropriate in the face of financial and temporal constraints. 17 2.4.3 Density Estimates and Individual Recognition Density estimates are a common objective of camera trapping studies (20% of papers reviewed) and may be the most sought-after population parameter (O'Connell et al. 2011). Density estimates allow for easy comparisons between sites and years or extrapolation to larger areas (Bellan et al. 2013). If individuals can be identified in a population, capture-recapture methods can produce reliable estimates for a study area. Three main types of capture-recapture population models are used to estimate abundance: (i) closed-no birth, death, immigration or emigration (O'brien 2011), (ii) open-losses and recruitments are allowed (Gutiérrez-González et al. 2012), and (iii) spatially explicit-including spatial characteristics such as home range (Gardner et al. 2010). Abundance estimates from these models can be converted to density estimates by estimating the area sampled during the survey (Maffei and Noss 2007). The sampling area for density estimate studies is typically set up in a grid-like system with the outermost trap locations representing the study area boundary. To estimate effective sampling area, the simplest approach is to draw a concave polygon by connecting outermost trap locations. However, this fails to include ingress from outside animals. A more appropriate approach is to estimate a buffer around this polygon. Though no consensus exists on calculating this area, a buffer of mean maximum distance moved (MMDM) of the target species is common. MMDM can be estimated from camera trap data, spatially explicit capture-recapture models (SECR), or estimates based on auxiliary telemetry data. The MMDM method may over-inflate density estimates (Soisalo and Cavalcanti 2006), which has led to the arbitrary but frequently used ½MMDM approach. Neither has a theoretical basis (Obbard et al. 2010). Auxiliary 18 telemetry data, typically available from other studies on target species, is most effective at estimating MMDM (Dillon and Kelly 2008, Núñez-Pérez 2011). Early in camera trapping science, two landmark papers estimated density of tigers by identifying individuals with unique pelage characteristics (Karanth 1995, Karanth and Nichols 1998). This approach has been extended to a variety of species to identify individuals based on spots (Jackson et al. 2006), stripes (Singh et al. 2010), muzzle markings (Mazzolli 2010) and other forms of unique pelage (Caruso et al. 2012). Additionally, capture-recapture methods are possible if animals are captured and tagged with artificial markings such as ear tags or GPS collars (Jordan et al. 2011, Weckel and Rockwell 2013). However, density estimation is not a completely refined analysis technique (Foster and Harmsen 2012). Study design issues related to sampling area, camera spacing, and detection probability may introduce significant biases (Dillon and Kelly 2007). Individual identification is subject to researcher bias (Oliveira-Santos et al. 2010), and efforts have been made to incorporate a more rigorous Bayesian approach to individual identification (Stafford and Lloyd 2011). Bilateral photo identification records from single trap stations can introduce inconsistencies due to bilateral asymmetry in coat patterns, but modeling approaches to combine left- and right-sided photos are being developed (McClintock et al. 2013). A common and simple resolution is to modify study design to include two cameras at each station (Negrões et al. 2012). Several reviews have focused on analysis techniques (Sharma et al. 2010, Obbard et al. 2010, Foster and Harmsen 2012), improving current capture-recapture analysis (Royle et al. 2009), new techniques including Bayesian inferences for arbitrary sample sizes 19 (Gardner et al. 2010), and maximum likelihood approaches (O'Brien and Kinnaird 2011). SECR models use a hierarchical approach to model detection probability and have produced more accurate density measurements in some studies (Kalle et al. 2011, Blanc et al. 2013). These advances in density estimators work only for the relatively few species that can be individually identified by coat patterns. Techniques to estimate density without individual identification have been proposed (Carbone et al. 2001, Rowcliffe et al. 2008, Manzo et al. 2012), but have not been without criticisms (Foster and Harmsen 2012). 2.4.4 Occupancy Reliable density estimates require rigorous study design and knowledge of advanced statistical techniques. An alternative approach is occupancy modeling, an established method to model the probability of a site being occupied by a species (MacKenzie et al. 2006, O'Connell and Bailey 2011). Occupancy uses presence/absence data from independent replicate surveys under the assumption that the population is closed during the survey period. Results provide a probability of occupancy across space based on researchers' definition of the site. In addition, surveys can be conducted over time and space to elucidate how habitat covariates impact species occurrence. A major advantage of occupancy modeling is that it explicitly estimates and models detection probability (Jones 2011). Generally, there is a positive relationship between occupancy and abundance, and occupancy has been used as a proxy for abundance in studies of niche partitioning (Di Bitetti et al. 2010), impact of human disturbance (Mohamed et al. 2013), and predator-prey dynamics (Silva-Rodríguez and Sieving 2012). This approach requires 20 smaller sample sizes and is therefore typically less expensive (MacKenzie et al. 2006). A rich literature exists on modeling species occupancy with a wide variety of presence/absence data (Vojta 2005, MacKenzie et al. 2006) and has been used in a number of camera trap studies (Erb et al. 2012, Gopalaswamy et al. 2012a, Schuette et al. 2013). Because occupancy can be estimated given any type of presence data, methods described in the previous section (3.1) can be substituted for camera traps. Given adequate detection probabilities, site occupancy estimated from camera traps is similar to estimates from cubby boxes, hair traps and track plates (O'Connell et al. 2006). Sign surveys (e.g., scat) have proven effective at generating reliable occupancy estimates (Gopalaswamy et al. 2012a), and may be more effective than camera traps at estimating occupancy at a landscape scale (Karanth et al. 2011). Study design for occupancy models requires a grid system camera array that provides a representative sample of the study area. At least 20 sampling units (grid cells) should be sampled, but occupancy models allow for stations to be shifted between units, given that they are present at each location long enough to collect sufficient data (O'Connell and Bailey 2011). Cameras should be spaced at a distance greater than the minimum of the diameter of the target species' home range. Information on environmental conditions for each grid cell also needs to be collected if researchers choose to include habitat covariates in the occupancy model. 21 2.4.5 Activity Analysis Activity patterns of target species can also be observed using camera trap data (21% of reviewed papers) to elucidate diel and seasonal activity patterns and understand interspecific competition and niche analysis. Camera traps allow researchers to record multiple species over long periods with minimal disturbance (Ramesh and Kalle 2013). Much work has been done with sympatric species, such as felids (Foster et al. 2013) and observation of predator-prey dynamics (Weckel et al. 2006, Ford and Clevenger 2010, Linkie and Ridout 2011). Especially important for conservation, human impact on animal activity, including human-wildlife coexistence, has also been investigated (Carter et al. 2012). However, co-occurrence does not necessarily equal coexistence (Harihar et al. 2013), and camera trapping data may fail to capture important factors that determine species activity and distribution. A presumed benefit of camera traps is that wildlife are not impacted by equipment. However, camera traps should not be considered an entirely nonintrusive approach. The sound (shutter), sight (flash), and smell (human scent) from camera traps may impact the behavior of animals, and even lead to trap shyness (Larrucea et al. 2007). Furthermore, because animal behavior can be difficult to interpret from a single photograph, motivations are often inferred and may lead to false interpretation. Though camera traps enable researchers to gain new insights into the activity patterns of wild animals, other approaches also produce reliable data. Most telemetry collars are now equipped with an activity sensor that monitors dual axis movement of an animal's neck at high temporal resolutions (5 min intervals). Combined with movement data from GPS location, detailed animal activity can be observed. More recently, Crittercams© 22 have been deployed on large mammal species to document previously unknown activity (Şekercioğlu 2013). With increasing amounts of data being collected remotely, the value of firsthand field observations cannot be underestimated. Biologists must ensure they do not become distanced from critical, field-based knowledge of animal ecology (Hebblewhite and Haydon 2010). Study design for activity surveys focuses on documenting temporal and seasonal presence data and therefore should strive to maximize detection probabilities for target species. Camera placement on game trails and other areas frequented by animals may increase captures. To avoid biases associated with detection, study design must aim to have equal detection probabilities between species. 2.5 Importance of Camera Trapping for Biodiversity Monitoring and Conservation Camera traps have become invaluable tools for conservation projects around the world. One of their most valuable features is their ability to deploy a trap that will record thousands of images with limited labor requirements and remain active for months (O'Brien 2016). This is particularly useful in study areas characterized by remoteness, conservation interest and/or lack of data. Camera traps have documented images of new species (Rovero and Rathbun 2006), range extensions (Chynoweth et al. 2015), poaching and other illegal activity (Jenks et al. 2012), or preliminary fauna inventories for understudied areas (Tobler et al. 2008). These images can impact short- and long-term agendas for conservation organizations. In the past decade, much work has been done to improve the scientific rigor of 23 camera trapping studies. Camera trapping science is evolving rapidly, and scientists and practitioners emphasize that carefully executed study designs can yield informative parameters we have described in the sections above. However, the importance of simple, inexpensive camera deployments that can revolutionize conservation projects with budgetary restrictions should also be recognized. It has been suggested that there are two categories of camera trap studies: (1) science, understanding how an ecosystem works, and (2) management, moving an ecosystem from less to more desirable states (Nichols et al. 2011). We assert that conservation outreach and environmental education constitute a third category. While other experts have suggested that photos are the means to an end goal of informing the larger process of science and management (Nichols et al. 2011), we also affirm the value of photographic records of elusive species. For example, an existing conservation project in eastern Turkey initially deployed four camera traps at a study site in 2006. The documentation of an unexpected relative abundance of large carnivores and the scarcity of their prey species has led to national and international support for a largescale monitoring project for mammals, catalyzed the government to designate Turkey's first wildlife corridor, and the project has become a conservation icon in a country experiencing a major biodiversity crisis (Şekercioğlu et al. 2011). The project has since evolved into a more rigorous study with a network of 40 camera traps being systematically deployed over a multiyear period. Camera trap photos and videos are also effective public outreach tools that raise awareness about important study sites, vulnerable species, and conservation priorities of local and global organizations or governmental agencies. A single photo published via social and traditional media can deliver important conservation messages to thousands of 24 people. The authors share camera trap photos and project updates on Facebook and Instagram, where one photo can be viewed by over 5,000 individuals in a five-day period. Public outreach opportunities extend to citizen science approaches in which members of the public deploy cameras or identify species in camera trap photos. Several large-scale camera trapping efforts, such as the Tropical Ecology Assessment and Monitoring Network (TEAM; www.teamnetwork.org; also see Fegraus et al. 2011) and Smithsonian Wild (see http://siwild.si.edu), have already made progress through citizen science efforts. Three case studies outlining successful camera trap projects are available in supplemental material (Appendix B). 2.6 Future of Camera Trapping in Conservation Biology and Ecology In the past two decades, camera trapping has emerged as an important subfield of conservation biology and ecology, and the exponential increase in studies is likely to continue. Research questions on presence/absence and basic ecology of animals are valuable to conservation efforts. However, further development of study designs, analyses and a standardization of reporting camera trap results is needed (Meek et al. 2014). Currently, few studies go beyond baseline assessments (Linkie et al. 2010), but as equipment becomes less expensive, broad scale landscape ecology studies can incorporate camera traps to address novel questions in conservation biology (Erb et al. 2012). Our literature review highlights the benefits of camera traps as a low-cost, low maintenance, and mostly noninvasive monitoring tool for conservation biology research 25 and applied conservation projects. Many studies in our review documented understudied species in remote areas and how significant camera trapping findings contributed to the conservation of species and ecosystems. Use the growing body of literature, conservationists can ensure they are defining questions a priori and making inferences using appropriate analyses and statistical techniques. As more sophisticated studies are designed, camera traps will help shape large-scale conservation agendas, especially across protected areas (Kinnaird and O'brien 2012, Li et al. 2012). Camera trapping has been increasingly discussed as a method for conservation hotspot analyses (Kouakou et al. 2011), monitoring biodiversity (Waldon et al. 2011), comparing human dominated landscapes to natural areas (Cassano et al. 2012), and assessing how animals respond to fluctuations in human activity (Harihar et al. 2009, Mohamed et al. 2013). Camera traps have already helped biologists document unexpected wildlife presence in human-dominated landscapes (Athreya et al. 2013). If global camera trapping efforts can be standardized (Ahumada et al. 2011) and coordinated, camera traps could contribute to a comprehensive global mammal conservation strategy (Rondinini et al. 2011). Furthermore, if data management issues are addressed, meta-analyses of current data could be pursued for regional analysis of abundance and diversity (Ordeñana et al. 2010). This review highlights two recurring issues undermining the main conservation biology questions camera traps aim to address. First, detection probability and survey effort are frequently ignored, but these elements are fundamental to the inferences that can be made from camera trapping data. Scientists, managers and conservationists should be careful when comparing or applying results from camera trapping studies that do not 26 address these issues. Second, camera trapping may not be the most suitable method to address a given conservation biology question. Many reliable methods can document the presence of species and lead to accurate estimations of population parameters. Some of the most successful studies in our review used camera traps in conjunction with other techniques to generate estimates of target species density (Gopalaswamy et al. 2012b) and a more holistic picture of population dynamics (Palomares et al. 2012). Given the current benefits and future prospects of camera traps to address objectives in animal ecology and conservation biology, the surge in peer-reviewed publications over the last few years is to be expected. Our literature review used only the Web of ScienceTM database and did not include gray literature from conservation organizations or government agencies. Inclusion would have increased our sample size of papers, but we believe that we would have reached similar conclusions. In light of the current popularity of camera traps, biologists must carefully define their questions and objectives before field data collection. Camera trapping analyses have evolved to depend heavily on careful study design, without which improper data collection severely limits analyses and subsequent inferences and consequently reduces the conservation benefits of this revolutionary method. 2.7 References Ahumada, J. a, C. E. F. Silva, K. Gajapersad, C. Hallam, J. Hurtado, E. Martin, A. McWilliam, B. Mugerwa, T. O'Brien, F. Rovero, D. Sheil, W. R. Spironello, N. Winarni, and S. J. Andelman. 2011. Community structure and diversity of tropical forest mammals: data from a global camera trap network. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 366:2703-11. Athreya, V., M. Odden, J. D. C. Linnell, J. Krishnaswamy, and U. Karanth. 2013. Big cats in our backyards: persistence of large carnivores in a human dominated 27 landscape in India. PLoS One 8:e57872. Bayne, E. M., and K. A. Hobson. 1997. Comparing the effects of landscape fragmentation by forestry and agriculture on predation of artificial nests. Conservation Biology 11:1418-1429. Beck, H., and J. Terborgh. 2002. Groves versus isolates: how spatial aggregation of Astrocaryum murumuru palms affects seed removal. Journal of Tropical Ecology 18: 275-288. Bellan, S. E., O. Gimenez, R. Choquet, and W. M. Getz. 2013. A hierarchical distance sampling approach to estimating mortality rates from opportunistic carcass surveillance data. Methods in Ecology and Evolution 4:361-369. Di Bitetti, M. S., C. D. De Angelo, Y. E. Di Blanco, and A. Paviolo. 2010. Niche partitioning and species coexistence in a Neotropical felid assemblage. Acta Oecologica 36:403-412. Blanc, L., E. Marboutin, S. Gatti, and O. Gimenez. 2013. Abundance of rare and elusive species: empirical investigation of closed versus spatially explicit capture-recapture models with lynx as a case study. The Journal of Wildlife Management 77:372-378. Bowkett, A. E., F. Rovero, and A. R. Marshall. 2008. The use of camera-trap data to model habitat use by antelope species in the Udzungwa Mountain forests, Tanzania. African Journal of Ecology 46:479-487. Burton, A. C. 2012. Critical evaluation of a long-term, locally-based wildlife monitoring program in West Africa. Biodiversity and Conservation 21:3079-3094. Burton, A. C., E. Neilson, D. Moreira, A. Ladle, R. Steenweg, J. T. Fisher, E. Bayne, and S. Boutin. 2015. REVIEW: Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes. Journal of Applied Ecology 52:675-685. Carbone, C., S. Christie, K. Conforti, T. Coulson, N. Franklin, J. R. Ginsberg, M. Griffiths, J. Holden, K. Kawanishi, M. Kinnaird, R. Laidlaw, A. Lynam, D. W. Macdonald, D. Martyr, C. McDougal, L. Nath, T. O'Brien, J. Seidensticker, D. J. L. Smith, M. Sunquist, R. Tilson, and W. N. Shahruddin. 2001. The use of photographic rates to estimate densities of tigers and other cryptic mammals. Animal Conservation 4:75-79. Carter, N. H., B. K. Shrestha, J. B. Karki, N. M. B. Pradhan, and J. Liu. 2012. Coexistence between wildlife and humans at fine spatial scales. Proceedings of the National Academy of Sciences of the United States of America 109:15360-5. Caruso, N., C. Manfredi, E. M. L. Vidal, E. B. Casanaveo, and M. Lucherinio. 2012. First 28 density estimation of two sympatric small cats, Leopardus colocolo and Leopardus geoffroyi, in a shrubland area of central Argentina. Annales Zoologici Fennici 49:181-191. Cassano, C. R., J. Barlow, and R. Pardini. 2012. Large mammals in an agroforestry mosaic in the Brazilian Atlantic Forest. Biotropica 44:818-825. Chynoweth, M. W., E. Çoban, and Ç. H. Şekercioğlu. 2015. Conservation of a new breeding population of Caucasian lynx (Lynx lynx dinniki) in eastern Turkey. Turkish Journal of Zoology 39:541-543. Dillon, A., and M. J. Kelly. 2007. Ocelot Leopardus pardalis in Belize: the impact of trap spacing and distance moved on density estimates. Oryx 41:469-477. Dillon, A., and M. J. Kelly. 2008. Ocelot home range, overlap and density: comparing radio telemetry with camera trapping. Journal of Zoology 275:391-398. Dobson, A., and K. Nowak. 2010. Does this photo make my range look big? Animal Conservation 13:347-349. Erb, P. L., W. J. McShea, and R. P. Guralnick. 2012. Anthropogenic influences on macro-level mammal occupancy in the Appalachian Trail corridor. PLoS One 7:e42574. Espartosa, K. D., B. T. Pinotti, and R. Pardini. 2011. Performance of camera trapping and track counts for surveying large mammals in rainforest remnants. Biodiversity and Conservation 20:2815-2829. Farhadinia, M. S., A. Mahdavi, and F. Hosseini-Zavarei. 2009. Reproductive ecology of the Persian Leopard, Panthera pardus saxicolor, in Sarigol National Park, northeastern Iran. Zoology in the Middle East 48:13-16. Fegraus, E. H., K. Lin, J. a. Ahumada, C. Baru, S. Chandra, and C. Youn. 2011. Data acquisition and management software for camera trap data: a case study from the TEAM Network. Ecological Informatics 6:345-353. Ford, A. T., and A. P. Clevenger. 2010. Validity of the prey-trap hypothesis for carnivore-ungulate interactions at wildlife-crossing structures. Conservation Biology 24:1679-85. Ford, A. T., A. P. Clevenger, and A. Bennett. 2009. Comparison of methods of monitoring wildlife crossing-structures on highways. Journal of Wildlife Management 73:1213-1222. Foster, R. J., and B. J. Harmsen. 2012. A critique of density estimation from camera-trap data. The Journal of Wildlife Management 76:224-236. 29 Foster, V. C., P. Sarmento, R. Sollmann, N. Tôrres, A. T. a. Jácomo, N. Negrões, C. Fonseca, and L. Silveira. 2013. Jaguar and puma activity patterns and predator-prey interactions in four Brazilian biomes. Biotropica 0:n/a-n/a. Galaverni, M., D. Palumbo, E. Fabbri, R. Caniglia, C. Greco, and E. Randi. 2011. Monitoring wolves (Canis lupus) by noninvasive genetics and camera trapping: a small-scale pilot study. European Journal of Wildlife Research 58:47-58. Gardner, B., J. Reppucci, M. Lucherini, and J. A. Royle. 2010. Spatially explicit inference for open populations: estimating demographic parameters from cameratrap studies. Ecology 91:3376-83. Gompper, M. E., R. W. Kays, J. C. Ray, S. D. LaPoint, D. A. Bogan, and J. R. Cryan. 2006. A comparison of noninvasive techniques to survey carnivore communities in northeastern North America. Wildlife Society Bulletin 34:1142-1151. Gopalaswamy, A. M., K. U. Karanth, N. S. Kumar, and D. W. Macdonald. 2012a. Estimating tropical forest ungulate densities from sign surveys using abundance models of occupancy. Animal Conservation 15:669-679. Gopalaswamy, A. M., J. A. Royle, M. Delampady, J. D. Nichols, K. U. Karanth, and D. W. Macdonald. 2012b. Density estimation in tiger populations: combining information for strong inference. Ecology 93:1741-51. Gutiérrez-González, C. E., M. Á. Gómez-Ramírez, and C. a. López-González. 2012. Estimation of the density of the Near Threatened jaguar Panthera onca in Sonora, Mexico, using camera trapping and an open population model. Oryx 46:431-437. Hackett, H. M., D. B. Lesmeister, J. Desanty-Combes, W. G. Montague, J. J. Millspaugh, and M. E. Gompper. 2007. Detection rates of eastern spotted skunks (Spilogale Putorius) in Missouri and Arkansas using live-capture and noninvasive techniques. The American Midland Naturalist 158:123-131. Harihar, A., P. Chanchani, R. K. Sharma, J. Vattakaven, S. Gubbi, B. Pandav, and B. Noon. 2013. Conflating "co-occurrence" with "coexistence". Proceedings of the National Academy of Sciences of the United States of America 110:E109. Harihar, A., B. Pandav, and S. P. Goyal. 2009. Responses of tiger (Panthera tigris) and their prey to removal of anthropogenic influences in Rajaji National Park, India. European Journal of Wildlife Research 55:97-105. Harrison, R. L., D. J. Barr, and J. Dragoo. 2002. A comparison of population survey techniques for swift foxes (Vulpes velox) in New Mexico. The American Midland Naturalist 148:320. Hebblewhite, M., and D. T. Haydon. 2010. Distinguishing technology from biology: a 30 critical review of the use of GPS telemetry data in ecology. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 365:2303-12. Jackson, R. M., J. D. Roe, R. Wangchuk, and D. O. Hunter. 2006. Estimating snow leopard population abundance using photography and capture-recapture techniques. Wildlife Society Bulletin 34:772-781. Jenks, K. E., P. Chanteap, K. Damrongchainarong, P. Cutter, P. Cutter, T. Redford, A. J. Lynam, J. Howard, and P. Leimgruber. 2011. Using relative abundance indices from camera-trapping to test wildlife conservation hypotheses - an example from Khao Yai National Park, Thailand. Tropical Conservation Science 4:113-131. Jenks, K. E., J. Howard, and P. Leimgruber. 2012. Do ranger stations deter poaching activity in national parks in Thailand? Biotropica 44:826-833. Jennelle, C. S., M. C. Runge, and D. I. MacKenzie. 2002. The use of photographic rates to estimate densities of tigers and other cryptic mammals: a comment on misleading conclusions. Animal Conservation 5:119-120. Jhala, Y., Q. Qureshi, and R. Gopal. 2011. Can the abundance of tigers be assessed from their signs? Journal of Applied Ecology 48:14-24. Jones, J. P. G. 2011. Monitoring species abundance and distribution at the landscape scale. Journal of Applied Ecology 48:9-13. Jones, J. P. G., G. P. Asner, S. H. M. B. And, and K. U. Karanth. 2013. The "why", "what" and "how" of monitoring for conservation. Pages 329-343 in D. W. Macdonald and K. J. Willis, editors. Key Topics in Conservation Ecology. First ed. John Wiley & Sons, Ltd. Jordan, M. J., R. H. Barrett, and K. L. Purcell. 2011. Camera trapping estimates of density and survival of fishers Martes pennanti. Wildlife Biology 17:266-276. Kalle, R., T. Ramesh, Q. Qureshi, and K. Sankar. 2011. Density of tiger and leopard in a tropical deciduous forest of Mudumalai Tiger Reserve, southern India, as estimated using photographic capture-recapture sampling. Acta Theriologica 56:335-342. Karanth, K. U. 1995. Estimating tiger Panthera tigris populations from camera-trap data using capture-recapture models. Biological Conservation 71:333-338. Karanth, K. U., A. M. Gopalaswamy, N. S. Kumar, S. Vaidyanathan, J. D. Nichols, and D. I. MacKenzie. 2011. Monitoring carnivore populations at the landscape scale: occupancy modelling of tigers from sign surveys. Journal of Applied Ecology 48:1048-1056. 31 Karanth, K. U., and J. D. Nichols. 1998. Estimation of tiger densities in India using photographic captures and recaptures. Ecology 79:2852-2862. Kinnaird, M. F., and T. G. O'brien. 2012. Effects of private-land use, livestock management, and human tolerance on diversity, distribution, and abundance of large african mammals. Conservation Biology 26:1026-39. Kouakou, C. Y., C. Boesch, and H. S. Kuehl. 2011. Identifying hotspots of chimpanzee group activity from transect surveys in Taï National Park, Côte d'Ivoire. Journal of Tropical Ecology 27:621-630. Larrucea, E. S., P. F. Brussard, M. M. Jaeger, and R. H. Barrett. 2007. Cameras, coyotes, and the assumption of equal detectability. Journal of Wildlife Management 71:1682- 1689. Lhota, S., B. Loken, S. Spehar, E. Fell, A. Pospěch, and N. Kasyanto. 2012. Discovery of Miller's grizzled langur (Presbytis hosei canicrus) in Wehea Forest confirms the continued existence and extends known geographical range of an endangered primate. American Journal of Primatology 74:193-198. Li, S., W. J. McShea, D. Wang, Z. Lu, and X. Gu. 2012. Gauging the impact of management expertise on the distribution of large mammals across protected areas. Diversity and Distributions 18:1166-1176. Linkie, M., G. Guillera-Arroita, J. Smith, and D. M. Rayan. 2010. Monitoring tigers with confidence. Integrative Zoology 5:342-50. Linkie, M., and M. S. Ridout. 2011. Assessing tiger-prey interactions in Sumatran rainforests. Journal of Zoology 284:224-229. MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and J. E. Hines. 2006. Occupancy Estimation and Modeling: Inferring Patterns and Dynamics of Species Occurrence. Academic, New York, NY. Maffei, L., and A. J. Noss. 2007. How small is too small? Camera trap survey areas and density estimates for ocelots in the Bolivian Chaco. Biotropica 40:71-75. Main, M. B., and L. W. Richardson. 2002. Habitat management response of wildlife to prescribed fire in southwest Florida pine flatwoods. Wildlife Society Bulletin 30:213-221. Manzo, E., P. Bartolommei, J. M. Rowcliffe, and R. Cozzolino. 2012. Estimation of population density of European pine marten in central Italy using camera trapping. Acta Theriologica 57:165-172. Mazzolli, M. 2010. Mosaics of exotic forest plantations and native forests as habitat of 32 pumas. Environmental Management 46:237-53. McCallum, J. 2013. Changing use of camera traps in mammalian field research: habitats, taxa and study types. Mammal Review 43:196-206. McClintock, B. T., P. Conn, R. Alonso, and K. R. Crooks. 2013. Integrated modeling of bilateral photo-identification data in mark-recapture analyses. Ecology:130315134921006. Meek, P. D., G. Ballard, A. Claridge, R. Kays, K. Moseby, T. O'Brien, A. O'Connell, J. Sanderson, D. E. Swann, M. Tobler, and S. Townsend. 2014. Recommended guiding principles for reporting on camera trapping research. Biodiversity and Conservation 23:2321-2343. Meijaard, E., A. C. K. I. T. Chener, and C. Smeenk. 2006. "New Bornean carnivore" is most likely a little known flying squirrel. Mammal Review 36:318-324. Mohamed, A., R. Sollmann, H. Bernard, L. N. Ambu, P. Lagan, S. Mannan, H. Hofer, and A. Wilting. 2013. Density and habitat use of the leopard cat (Prionailurus bengalensis) in three commercial forest reserves in Sabah, Malaysian Borneo. Journal of Mammalogy 94:82-89. Muhly, T. B., C. Semeniuk, A. Massolo, L. Hickman, and M. Musiani. 2011. Human activity helps prey win the predator-prey space race. PLoS One 6:e17050. Negrões, N., R. Sollmann, C. Fonseca, A. T. A. Jácomo, E. Revilla, and L. Silveira. 2012. One or two cameras per station? Monitoring jaguars and other mammals in the Amazon. Ecological Research 27:639-648. Nichols, J. D., K. U. Karanth, and A. F. O'Connell. 2011. Science, conservation, and camera traps. Pages 45-56 in A. F. O'Connell, J. D. Nichols, and K. U. Karanth, editors. Camera Traps in Animal Ecology. Springer Japan, Tokyo. Núñez-Pérez, R. 2011. Estimating jaguar population density using camera-traps: a comparison with radio-telemetry estimates. Journal of Zoology 285:39-45. O'Brien, T. G. 2011. Abundance, density, and relative abundance. Pages 71-96 in A. F. O'Connell, J. D. Nichols, and K. U. Karanth, editors. Camera Traps in Animal Ecology. Springer, Tokyo. O'Brien, T. G. 2016. Camera traps for conservation. Pages 228-241 in L. Joppa, J. G. Robinson, and J. Baillie, editors. Protected Areas. John Wiley & Sons, Ltd, Chichester, UK. O'Brien, T. G., and M. F. Kinnaird. 2011. Density estimation of sympatric carnivores using spatially explicit capture-recapture methods and standard trapping grid. 33 Ecological Applications 21:2908-2916. O'Brien, T. G., M. F. Kinnaird, and H. T. Wibisono. 2003. Crouching tigers, hidden prey: Sumatran tiger and prey populations in a tropical forest landscape. Animal Conservation 6:131-139. O'Connell, A. F., and L. L. Bailey. 2011. Inference for occupancy and occupancy dynamics. Pages 191-205 in A. F. O'Connell, J. D. Nichols, and K. U. Karanth, editors. Camera Traps in Animal Ecology. Springer, Japan. O'Connell, A. F., J. D. Nichols, and U. K. Karanth. 2011. Camera traps in animal ecology. Springer Japan, Tokyo. O'Connell, A. F., N. W. Talancy, L. L. Bailey, J. R. Sauer, R. Cook, and A. T. Gilbert. 2006. Estimating site occupancy and detection probability parameters for meso- And large mammals in a coastal ecosystem. Journal of Wildlife Management 70:1625- 1633. Obbard, M. E., E. J. Howe, and C. J. Kyle. 2010. Empirical comparison of density estimators for large carnivores. Journal of Applied Ecology 47:76-84. Oliveira-Santos, L. G. R., C. A. Zucco, P. C. Antunes, and P. G. Crawshaw. 2010. Is it possible to individually identify mammals with no natural markings using cameratraps? A controlled case-study with lowland tapirs. Mammalian Biology - Zeitschrift für Säugetierkunde 75:375-378. Ordeñana, M. A., K. R. Crooks, E. E. Boydston, R. N. Fisher, L. M. Lyren, S. Siudyla, C. D. Haas, S. Harris, S. A. Hathaway, G. M. Turschak, A. K. Miles, and D. H. Van Vuren. 2010. Effects of urbanization on carnivore species distribution and richness. Journal of Mammalogy 91:1322-1331. Palomares, F., J. A. Godoy, J. V. López-Bao, A. Rodríguez, S. Roques, M. Casas-Marce, E. Revilla, and M. Delibes. 2012. Possible extinction vortex for a population of Iberian lynx on the verge of extirpation. Conservation Biology 26:689-97. Paull, D. J., A. W. Claridge, and R. B. Cunningham. 2012. Effective detection methods for medium-sized ground dwelling mammals: a comparison between infrared digital cameras and hair tunnels. Wildlife Research 39:546-553. Ramesh, T., and R. Kalle. 2013. Activity pattern of Sloth Bear Melursus ursinus (Mammalia: Ursidae) in Mudumalai Tiger Reserve, Western Ghats, India. Journal of Theatened Taxa 5:3989-3992. Ramesh, T., R. Kalle, K. Sankar, and Q. Qureshi. 2012. Spatio-temporal partitioning among large carnivores in relation to major prey species in Western Ghats. Journal of Zoology 287:269-275. 34 Rondinini, C., A. S. L. Rodrigues, and L. Boitani. 2011. The key elements of a comprehensive global mammal conservation strategy. Philosophical Transactions of the Royal Society of London. Series B, Biological sciences 366:2591-7. Rosas-Rosas, O. C., and L. C. Bender. 2012. Population status of jaguars (Panthera onca) and pumas (Puma concolor) in Northeastern Sonora, Mexico. Acta Zoologica Mexicana 28:86-101. Rovero, F., and G. B. Rathbun. 2006. A potentially new giant sengi (Elephant-shrew) from tht Udzungwa Mountains, Tanzania. Journal of East African Natural History 95:111-115. Rovero, F., G. B. Rathbun, a. Perkin, T. Jones, D. O. Ribble, C. Leonard, R. R. Mwakisoma, and N. Doggart. 2008. A new species of giant sengi or elephant-shrew (genus Rhynchocyon) highlights the exceptional biodiversity of the Udzungwa Mountains of Tanzania. Journal of Zoology 274:126-133. Rowcliffe, J. M., J. Field, S. T. Turvey, and C. Carbone. 2008. Estimating animal density using camera traps without the need for individual recognition. Journal of Applied Ecology 45:1228-1236. Royle, J. A., J. D. Nichols, K. U. Karanth, and A. M. Gopalaswamy. 2009. A hierarchical model for estimating density in camera-trap studies. Journal of Applied Ecology 46:118-127. Schuette, P., A. P. Wagner, M. E. Wagner, and S. Creel. 2013. Occupancy patterns and niche partitioning within a diverse carnivore community exposed to anthropogenic pressures. Biological Conservation 158:301-312. Şekercioğlu, Ç. H., S. Anderson, E. Akçay, R. Bilgin, Ö. E. Can, G. Semiz, Ç. Tavşanoğlu, M. B. Yokeş, A. Soyumert, K. İpekdal, İ. K. Sağlam, M. Yücel, and H. Nüzhet Dalfes. 2011. Turkey's globally important biodiversity in crisis. Biological Conservation 144:2752-2769. Sharma, R. K., Y. Jhala, Q. Qureshi, J. Vattakaven, R. Gopal, and K. Nayak. 2010. Evaluating capture-recapture population and density estimation of tigers in a population with known parameters. Animal Conservation 13:94-103. Silva-Rodríguez, E. a., and K. E. Sieving. 2012. Domestic dogs shape the landscape-scale distribution of a threatened forest ungulate. Biological Conservation 150:103-110. Singh, P., A. M. Gopalaswamy, and K. U. Karanth. 2010. Factors influencing densities of striped hyenas (Hyaena hyaena) in arid regions of India. Journal of Mammalogy 91:1152-1159. Soisalo, M. K., and S. M. C. Cavalcanti. 2006. Estimating the density of a jaguar 35 population in the Brazilian Pantanal using camera-traps and capture-recapture sampling in combination with GPS radio-telemetry. Biological Conservation 129:487-496. Sollmann, R., A. Mohamed, H. Samejima, and A. Wilting. 2013a. Risky business or simple solution - Relative abundance indices from camera-trapping. Biological Conservation 159:405-412. Sollmann, R., A. Mohamed, H. Samejima, and A. Wilting. 2013b. Risky business or simple solution - Relative abundance indices from camera-trapping. Biological Conservation 159:405-412. Stafford, R., and J. R. Lloyd. 2011. Evaluating a Bayesian approach to improve accuracy of individual photographic identification methods using ecological distribution data. Computational Ecology and Software 1:49-54. Tilson, R., H. Defu, J. Muntifering, and P. J. Nyhus. 2004. Dramatic decline of wild South China tigers Panthera tigris amoyensis: field survey of priority tiger reserves. Oryx 38:40-47. Tobler, M. W., S. E. Carrillo-Percastegui, R. Leite Pitman, R. Mares, and G. Powell. 2008. An evaluation of camera traps for inventorying large- and medium-sized terrestrial rainforest mammals. Animal Conservation 11:169-178. Trolle, M., A. J. Noss, J. L. P. Cordeiro, and L. F. B. Oliveira. 2008. Brazilian tapir density in the Pantanal: a comparison of systematic camera-trapping and linetransect surveys. Biotropica 40:211-217. Vilardell, A., X. Capalleras, J. Budó, and P. Pons. 2012. Predator identification and effects of habitat management and fencing on depredation rates of simulated nests of an endangered population of Hermann's tortoises. European Journal of Wildlife Research 58:707-713. Vojta, C. D. 2005. Old dog new tricks-innovations with presence-absence information. Journal of Wildlife Management 69:845-848. Waldon, J., B. W. Miller, and C. M. Miller. 2011. A model biodiversity monitoring protocol for REDD projects. Tropiocal Conservation Science 4:254-260. Weckel, M., W. Giuliano, and S. Silver. 2006. Jaguar (Panthera onca) feeding ecology: distribution of predator and prey through time and space. Journal of Zoology 270:060606025751038 Weckel, M., and R. F. Rockwell. 2013. Can controlled bow hunts reduce overabundant white-tailed deer populations in suburban ecosystems? Ecological Modelling 250:143-154. 36 Yamada, F., N. Kawauchi, K. Nakata, S. Abe, N. Kotaka, A. Takashima, C. Murata, and A. Kuroiwa. 2010. Rediscovery after thirty years since the last capture of the critically endangered Okinawa spiny rat Tokudaia muenninki in the northern part of Okinawa Island. Mammal Study 35:243-255. 37 Figure 2.1. Camera trapping studies by year published from 1975-2014 based on a systematic search of key terms in ISI Web of Knowledge. 38 Figure 2.2. The global distribution of camera trapping studies published from 1975-2014 based on a systematic search of key terms in ISI Web of Knowledge. 39 Table 2.1 Camera trap and animal ecology keywords used in the ISI Web of Knowledge literature search at the University of Utah. Camera Trap terms "Camera Trap*" "Game Camera*" "Trail Camera*" "Remote Photography" Animal ecology terms Wildlife Birds Mammals Reptiles Amphibians 40 Table 2.2. Number of camera trapping articles published in the top ten journals from 1975-2014 based on a systematic search of key terms in ISI Web of Knowledge. Journal Oryx Animal Conservation Biological Conservation Journal of Mammalogy Journal of Wildlife Management Biotropica Biodiversity and Conservation European Journal of Wildlife Research Journal of Tropical Ecology Journal of Zoology Number of Articles 48 23 22 17 17 14 13 13 13 12 41 Table 2.3. Proportion of target taxa in camera trapping studies published from 1975-2014 based on a systematic search of key terms in ISI Web of Knowledge. Taxa Percent of Total Articles* Mammal 84.1 Bird (including nest predation) 6.0 Multiple taxa 4.2 Herpetofauna 1.5 Insect 0.6 Within Mammal Order Diversity Percent of Mammal Category Carnivore 56.2 Multiple Orders 29.2 Ungulate 7.6 Rodentia 4.0 Primate 2.2 Marine 0.4 Bats 0.2 Within Carnivore Family Diversity Percent of Carnivore Category Felidae 58.8 Multiple Families 24.4 Canidae 10.8 Ursidae 3.6 Hyaenidae 1.2 Other 1.2 *Percent of total articles does not add up to 100% (96.4%) because review articles were present in the database CHAPTER 3 LARGE CARNIVORE HYPERABUNDANCE IN AN EMPTY FOREST 3.1 Abstract As human population and consequent ecological impact continue to grow, certain wildlife species are increasingly utilizing anthropogenic food sources to sustain and even increase their population sizes. One example is large carnivores, which are known to develop synanthropic behavior and rely on livestock, garbage, and other anthropogenic resources. To examine this phenomenon, we conducted a multiyear camera trap study in Sarıkamış-Allahuekber Mountains National Park and surrounding forest in eastern Turkey to document presence and estimate species-specific single-season occupancy for medium-large mammals in a geographically isolated and heavily degraded forest. Preliminary camera trap efforts began in 2006 with opportunistic sampling. During 20132016 we used a 2 km2 sampling grid to sample approximately 326 km2 of forested area dominated by Scots pine. Camera traps were deployed for a minimum of 45 consecutive days each year during four summer/fall field seasons. We obtained more than 50,000 images and detected 14 species of wild mammals during a total sampling effort of 12,731 camera trap days. Human activity was the most common event captured by cameras and was an order of magnitude more common than that of all other species. Gray wolves and 43 Eurasian brown bears were the most frequent wildlife events. Species-specific single season modeled occupancy estimates ranged across years (2013-2016) from 0.772 - 0.856 for bears, 0.53 - 0.922 for wolves, and 0.372 - 1 for lynx. Natural prey species were rarely captured, implying that these species may be functionally extinct as a natural prey base. Wild boar was the only natural prey species with sufficient data for occupancy modeling, with a range of 0.466 - 0.646 across years. Human activity was ubiquitous across the landscape, with human occupancy estimates ranging from 0.835 - 1 and livestock occupancy ranging from 0.395 - 0.498 across years. Our results suggest that in human-dominated landscapes, the combination of the scarcity of natural prey and the presence of synanthropic carnivores can result in an alternative stable state of an ecosystem in which carnivores increase in abundance while wildlife habitat quality continues to degrade. 3.2 Introduction The growth and impact of human activity has become a dominant force on our planet (Steffen et al. 2007, Barnosky et al. 2012). During the Anthropocene, ecosystems have deviated in structure and function from previous epochs, generating novel and hybrid ecosystems that include anthropogenic impact and activity as a major contributor to ecosystem processes (Hobbs et al. 2006). One major consequence of the Anthropocene is global level defaunation, or the depletion of animals from ecological communities. A concept that has recently drawn more attention, defaunation is an increasing global threat and a major driver of ecological change that has both short- and long-term consequences (Dirzo et al. 2014). 44 Vertebrate megafauna are often the most susceptible to large-scale depletion, due to their large home ranges (i.e., vast habitat requirements), harvest by humans as a potential food source, and direct persecution as a result of human-wildlife conflict (Ripple et al. 2014, 2015). Loss of large vertebrates can result in the disruption of ecological interactions and trophic restructuring of an ecosystem (Estes et al. 2011). For example, extirpation of large mammalian carnivores can lead to hyperabundance of herbivores and negative consequences for producers (Terborgh 2001). Another consequence of defaunation caused by human activity is the creation of empty forests, which are devoid of mammals and maintain the appearance of intact habitat, but many species are functionally extinct (Redford 1992). Large carnivores are one of the most imperiled mammal groups globally, and populations are absent or declining throughout much of their native range (Ripple et al. 2014). Large carnivores experience the same threats as all megafauna, but in addition, are frequently more persecuted than other megafauna because of real and perceived threats to livestock, game species, and people (Frank and Woodroffe 2001, Gross 2008). Therefore, these species have typically been the first trophic level to be removed from an ecosystem. Specialist and obligate carnivores may be the most threatened, as they are often unable to adapt to human-induced changes in ecosystem structure and function. In response to these human-driven changes in their environment, some generalist large carnivores can change behavior to exploit human food sources and use modified habitat (i.e., synanthropy) to sustain and even increase population numbers (Newsome et al. 2015b). Synanthropy in large carnivores is well documented in many species. Coyotes, a well-known synanthropic carnivore, exhibit individual diet specialization in 45 urban environments by altering movement and foraging strategies (Newsome et al. 2015a), and populations can reach higher densities where human food is available (Fedriani et al. 2001). Larger carnivores, including bears (Beckmann and Berger 2003) and wolves (Zlatanova et al. 2014, Newsome et al. 2016), have also demonstrated synanthropic behavior. While some carnivores are able to inhabit-and even thrive-in human-dominated landscapes, anthropogenic food subsidies can cause dramatic changes in animal behavior and trophic cascades (Newsome et al. 2015b). These can decouple predator-prey relationships that exist in more natural systems (Fischer et al. 2012). Furthermore, when carnivores and humans inhabit the same area, there is a higher likelihood of humanwildlife conflict (Messmer 2000), which can decrease social carrying capacity (Breitenmoser et al. 2005) and increase persecution. Recent work has documented large carnivore resource use along wildland-urban interfaces and suggests that tolerance of human activity may be a limiting factor for coexistence (Bouyer et al. 2014, Moss et al. 2016). Globally, these species will increasingly encounter fragmented habitat and humandominated ecosystems (Crooks et al. 2011). In order to gain a deeper understanding of trophic interactions and achieve important conservation goals, humans should be incorporated into trophic ecology (Dorresteijn et al. 2015). In this study, our intent was to implement a biodiversity monitoring program in northeastern Turkey, a historically understudied region of the world that harbors globally important biodiversity (Figure 3.1; Şekercioğlu et al. 2011). To guide ongoing conservation efforts in the region, we applied a multispecies, multiyear camera trapping approach to a fragmented patch of forest in a human-dominated landscape. Our study 46 area is representative of many other regions of the world where little or no biodiversity monitoring is occurring, and consequently, no mechanisms exist to maintain or protect existing biological resources. Our main objective for this study was to document species presence, quantify species richness, generate species' distributions, and evaluate community composition in a patch of fragmented forest surrounding the city of Sarıkamış in eastern Turkey (hereafter "Sarıkamış Forest"). Our ongoing work in Sarıkamış Forest has documented synanthropic behavior (Capitani et al. 2016, Cozzi et al. 2016) and our preliminary camera trapping work showed a lack of natural prey species. We therefore hypothesized that generalist predators (i.e., brown bears and gray wolves) would have higher occupancy estimates than other native species. We also aimed to assess the effectiveness of the Sarıkamış-Allahuekber Mountains National Park, a protected area in the region. Based on our field observations and the small size of the park, we predicted that occupancy estimates within park boundaries would not differ from outside the park boundaries. 3.3 Methods 3.3.1 Study Area Our study was carried out on the Kars-Ardahan high plateau in northeastern Turkey, at the intersection of Caucasus and Irano-Anatolian global biodiversity hotspots (Figure 3.1). The area (c. 550 km2; 40°20'N 42°35'E) ranges between 1900 and 3120 m asl and is composed of fragmented forest in a matrix of agricultural and rangelands. The city of Sarıkamış (population: c. 18,000) is located in the center of the study area and on one of 47 the two paved roads that bisect the forested area. Forest cover consists almost exclusively of Scots pine (Pinus sylvestris Linnaeus, 1753), while understory vegetation is scarce, with consequent scarcity of food resources for browsers. Sarıkamış-Allahuekber Mountains National Park (hereafter SAMNP; Figure 3.2) boundaries cover a total area of 225.1 sq. km., but only include 49.69 sq. km. of forest. Therefore, SAMNP is only comprised of 22.07% forest cover. Total forest cover in the region includes 328.38 sq. km. including a large expanse of forest south of the national park (248.15 sq. km.). These patches of forest represent the southernmost significant forest patch in the region extending south from the extensive forests in the Black Sea Region of Turkey. Human activity in the forest is extensive in both time and space, limited only by harsh winter temperatures, and consists primarily of livestock grazing, harvest of forest products (e.g., fruits, pine cones, mushrooms), and legal and illegal timber extraction. Livestock is abundant in the region with cattle (Bos taurus Linnaeus, 1758), sheep (Ovis aries Linnaeus, 1758), and goats (Capra hircus Linnaeus, 1758) freely roaming rangelands from April to November (Capitani et al. 2016). About 851,445 livestock heads have been registered in the Kars province in 2012 (Ministry of Food, Agriculture and Livestock, Republic of Turkey). A notable feature on the landscape is an unfenced municipal garbage dump 3 km west of Sarıkamıs city. The dump represents a predictable anthropogenic food source, and bears, wolves, and wild boar visit the dump regularly at night (pers. obs.). A portion of the bear population has altered life history strategies to regularly use the dump, while other bears never visit the dump (Cozzi et al. 2016). 48 3.3.2 Survey Methods Small-scale sampling began with preliminary and opportunistic camera trap efforts in 2006. During 2013-16, we followed a standardized protocol to sample the entire forested area (326 km2), using a 2 km2 sampling grid overlaid onto a forest cover map with Arc GIS 9.3 to determine camera stations. Points on the grid were visited to determine if camera trap deployment was feasible given the intense human activity throughout this forest. Theft and vandalism of camera traps were some of the most limiting factors in the completion of this study. We opportunistically targeted forest roads to maximize likelihood of capturing wildlife, but off-road areas were also sampled to determine relative use of dirt roads as movement corridors. If a site was deemed suitable, camera traps (Reconyx HC500/HC600/PC900) were deployed; a single camera was attached to the appropriately sized tree nearest the road. Camera stations were designed to capture medium-large mammals, with cameras secured at knee to waist height and positioned to capture animals approaching at a 45degree angle. Cameras were programmed to take 3 photos with each motion trigger with a 60-second delay. Vegetation that could trigger cameras was removed from the area, and bait was never used in the trap area. Cameras were deployed for a minimum of 45 consecutive days during each of four summer-fall field seasons (2013-2016). National security measures posed a significant barrier to a deploying and checking camera traps on a regular basis. As a result of this and additional permitting issues, we checked cameras sporadically over the course of deployment to replace batteries and download photographs. 49 3.3.3 Data Processing, Statistical Methods, and Occupancy Estimation At the end of each field season, cameras were recovered and data from images were extracted and classified using CAMERABASE software (Tobler 2010). Vandalism and theft caused many cameras to be removed from our final database. We only included camera stations in our analysis that were active for >50% of each sampling season and a minimum of 45 days. After classifying all images, we defined an independent species event as any sequence of photos of a single species within 60 minutes. Occupancy is defined as the proportion of points in the site where a species is expected to occur, does not require individual recognition, and is often a useful surrogate for abundance (Rovero and Marshall 2009). Modeling occupancy allows for heterogeneity in detection probability among survey sites (MacKenzie et al. 2006a). We used single-species single-season occupancy models to estimate occupancy for all species with adequate data. We modeled single-season occupancy (Ψ) and detection probability (p) where p was defined as the probability of observing a species during a survey period if it was present. To satisfy the need for temporally replicated data, we created a detection history of whether each species was observed by a camera trap at each station during each 5-day period throughout the survey resulting in approximately 18 sampling occasions per season (18.0 +/- 2.0). Models were solved by maximum likelihood estimation (MLE) via R statistical software (R Core Team 2016) using the unmarked package (Fiske and Chandler 2011) to estimate the probability species i occurred within the area sampled by a camera station during our survey period (i.e., occurrence), while accounting for incomplete detection (MacKenzie et al. 2002, 2006b). Problems with research permits 50 and political unrest limited when we could access the field site. Therefore, we used the only 3-month period of sampling that was consistent across 2013-16 sampling seasons, which was August 1st - November 1st. We limited occupancy modeling by species based on two factors. Only species that were captured at a naïve occupancy equal to or greater than 0.1 (>10 % of the working cameras) were including in occupancy estimates since studies have shown that occupancy estimates are not accurate for species recorded in less than 15-20% of the cameras (Rovero et al. 2014). Also, species need to be sampled with sufficient detection probability. We excluded species from occupancy modeling with estimated detection probability less than 0.1 for at least 2 of the 4 years (Rovero et al. 2014). 3.4 Results Across all survey years, including preliminary years, a total of 14 mammal species (Table 3.1) were identified from 66,930 images containing wildlife (6,679), humans (24,207), or domestic animals (10,190) during 12,731 trap nights (Table 3.2). Remaining photos were false triggers likely caused by vegetation moving in the wind. During 20132016 survey years, humans (on foot, horse, or vehicle) were the most common event and were captured on 91.3% (+/- 8.5) of the camera stations (Figure 3.3). Domestic species (livestock and dogs) were the second most common species, with cows being most frequent. Brown bears, gray wolves, wild boar, and Caucasian lynx, respectively, were the most commonly captured wildlife species. Natural prey species were rare; boar were present at low capture rates, and Eurasian hare and red squirrels were infrequent. Notably, roe deer were captured in only two surveys (Preliminary: 2009-2012 and 2015), 51 with <5 events in each. Large carnivore and natural prey species have a strong nocturnal activity pattern that contrasts with the strong diurnal pattern of humans and livestock (Figure 3.4). Species-specific occupancy estimates between years for bears, wolves, lynx, boar, livestock, and humans are variable but consistent (Figure 3.5). 3.4.1 National Park Boundaries Based on available data, we were unable to assess the effectiveness of SAMNP due to lack of successful camera trapping efforts within the park boundaries. Camera theft in the national park was a major limitation to generating enough data to compare these two areas. In addition, safety concerns prevented access to the park during several field seasons. In some years, all cameras in the national park were damaged or stolen, while in other years only one or two cameras were retrieved, leading to inadequate data for occupancy modeling. 3.4.2 Community Level Summaries Single-species, single-season occupancy estimates by trophic level reveal a novel mammal community structure characterized by hyperabundance of large carnivores (including the omnivorous brown bear) and absence of natural prey (Figure 3.6). Wild boar was the only natural prey species of large carnivore for which detection histories allowed for occupancy estimates to be generated. Livestock (cows, goats, sheep, donkeys, horses) were pooled together and represent the only herbivores present in the system with high enough detection rates to allow for occupancy modeling. 52 3.5 Discussion Our results suggest that the conditions in Sarıkamış Forest represent an alternative stable state characterized by large carnivore hyperabundance in a human dominated ecosystem. To our knowledge, this is the first time a system with a stable large carnivore population exists with an extremely limited natural prey base. In line with contemporary large carnivore research, our results suggest that the ecological impacts of large carnivores are not straightforward (Allen et al. 2017), especially in human-dominated landscapes (Dorresteijn et al. 2015), and presence of large carnivores is not necessarily equivalent to presence (i.e., function) of apex predators (Ordiz et al. 2013). In terms of conservation goals, caution should be used when applying concepts like umbrella species, indicator species, or ecosystem engineer to large carnivores, particularly generalist predators such as bears and wolves. It is important to recognize the ability of large carnivores to tolerate human activity, exploit anthropogenic food resources, and to coexist with humans in the absence of a natural prey base. 3.5.1 Absence of Natural Prey In many systems with gray wolves and Eurasian lynx, abundant roe deer, red deer, and wild boar are often observed as primary natural prey base. The two deer species are likely absent from Sarıkamış Forest system due to poor quality habitat and hunting pressure. Red deer are locally extinct and roe deer were recorded by camera traps during 2 survey years, but are functionally extinct. Our research group has not had any sightings or reports of red deer during our field work in Sarıkamış Forest since 2004. Wild boar are present, but in low densities, likely due to direct persecution from farmers that consider 53 them a major agricultural pest (Chynoweth et al. 2016). Scat surveys conducted by the authors in forest patches north of the study area reveal presence of roe deer (Chynoweth, unpublished data), but any ingress of deer from northern forests will be met with hyperabundance of large carnivores and high levels of human activity, neither of which are conducive to recolonization. The absence of large herbivores in Sarıkamış Forest supports the concept of the empty forest (Redford 1992), where a forest appears intact with full-grown trees, but large mammals are conspicuously absent. Sarıkamış Forest is a monotypic forest of Scots pine; intensive livestock grazing results in an intact canopy and reduction of understory vegetation to grasses and forbs (Zamora et al. 2001). This prevents regeneration of important browse species for native ungulates. Combined with intense human activity and illegal hunting, Sarıkamış Forest in its current unmanaged state may not contain the necessary resources to support red deer or roe deer. Wild boar, European hare, and red squirrels were detected by our camera traps and represent a small natural prey base for large carnivores, including wolves (Capitani et al. 2016). Wild boar represent a food resource for wolves in many other study areas (Imbert et al. 2016), but occupancy estimates for Sarıkamış are somewhat low, suggesting that wild boar density is low. Low occupancy estimates for European hare and red squirrels are likely slightly biased, due to the small range size and behavior of these species. Nonetheless, our occupancy estimates suggest a deficiency and lack of diversity of natural prey, likely unable to support the hyperabundance of large carnivores in the system. 54 3.5.2 Hyperabundance of Large Carnivores Hyperabundance of large carnivores in Sarıkamış Forest is possible due to two main factors. These are the existence of several predictable anthropogenic food sources available on the landscape, mainly garbage (Cozzi et al. 2016) and livestock (Capitani et al. 2016), and the presence of a large (albeit heavily degraded) forested area as a daytime refuge for wildlife. While humans are heavily utilizing forest areas, coexistence with large carnivores is possible due to complementary diel patterns of activity (Figure 3.4). Other factors influencing carnivore abundance may be species-specific. The large municipal garbage dump has a clear and well-documented effect on the high density of bears in Sarıkamış Forest (Cozzi et al. 2016). Authors routinely observe >10 bears (maximum observed = 33) at the garbage dump during a single night-time visit, with increased numbers during hyperphagia periods prior to bears' hibernation. Impacts of open garbage dumps on bears are well understood, and include foodconditioning and often, increased reproductive rates (Stringham 1989), both of which increase the potential for human-wildlife conflict. As these outcomes illustrate, access to human food sources greatly increases the likelihood of problem bears (Gunther 1994, Huber et al. 2007). The apparent high density of gray wolves in the study area could be a result of multiple factors. Persecution and limited natural prey base has led to decreased pack size and increased number of packs in the study area. Packs in Sarıkamış Forest typically consist solely of a breeding pair, and solitary wolves are common. While wolves are routinely observed at the open garbage dump, this is limited to a single wolf pack with an overlapping territory (pers obs). The variety of predictable anthropogenic food resources 55 in the area has likely resulted in individual level niche specialization (intraspecies variation) within wolf populations and the ability of wolves to subsist on a diversity of prey, including small mammals (Layman et al. 2015, Newsome et al. 2015a, 2016). The Eurasian lynx represents the only obligate carnivore and is likely surviving on small mammals (e.g., European hare and red squirrels). Sarıkamış Forest does contain a breeding population of lynx, most probably the subspecies Lynx lynx dinniki (Chynoweth et al. 2015). However, effective population size is likely low and this species may be at a tipping point. Using coat pattern to identify individuals, we have observed a maximum of 5 individuals in any given year. Given the high densities of sympatric carnivore species, lynx may be experiencing interspecific competition for natural prey with the gray wolf, as well as through competition and kleptoparasitism with brown bears (Krofel and Jerina 2016). 3.5.3 Effectiveness of Sarıkamış Allahuekber Mountains National Park Due to complications of sampling, we were not able to detect any difference between occupancy estimates within and outside the boundaries of SAMNP. However, our anecdotal evidence suggests that this national park offers little to no protection for mammal communities in the region. An important cultural memorial, SAMNP was designated in 2004 because of historical significance; during the Battle of Sarıkamış in World War I, over 60,000 Turkish troops died due to harsh winter weather in the Allahuekber Mountains. The park boundaries include vast expanses of agricultural land and habitat unsuitable for large carnivores. Several villages, a large hydroelectric project, and communal grazing lands lie within the park. The apparent lack of restrictions and 56 absence of enforcement enable people from nearby villages to easily harvest forest products (timber, food, etc.). Because SAMNP was designated for cultural and not environmental reasons, park administration has not effectively taken wildlife management into account. As a result, this National Park cannot claim to function as a protected area for biodiversity and should instead be considered a paper park (Dudley and Stolton 1999) in terms of biodiversity conservation. Our anecdotal evidence of similar or possible lower levels of biodiversity within national park boundaries compared to forested areas outside the park boundaries supports our field observations and bolsters the argument for increasing protected area in larger forest patches within the region. 3.6 Conclusion This work demonstrates an alternative stable state of large carnivore occurrence, a state defined by human activity and anthropogenic food sources. Our results reveal, for the first time to our knowledge, that generalist predators are able to survive in the absence of natural prey. The hyperabundance of large carnivores may have far-reaching impacts throughout the ecosystem, including increased human-wildlife conflict and, thus, a barrier to achieving conservation goals. Moving forward, a logical solution may be to increase the extent of the protected area; however, given the highly degraded habitat and pervasiveness of human activity, this may generate controversy due to uneven distribution of benefits that a protected area can provide (Brockington and Wilkie 2015). Eastern Turkey in general and Sarıkamış specifically are regions where people are reliant on natural resources for food, fuel, and 57 grazing lands. Specific strategies for our study site may include intentional rewilding of this forest fragment with species known to naturally occur in the region, such as roe deer and red deer. As such, this study area may provide an opportunity to move the science of rewilding and relocation forward through hypothesis testing and science-based monitoring of the rewilding concept (Seddon et al. 2014, Svenning et al. 2016). These areas of high human activity can constitute experimental plots to test novel hypotheses related to adaptive management. We can try various approaches in Sarıkamış to learn what is best. This approach may be a prerequisite for large carnivore persistence in human-dominated landscapes. 3.7 References Allen, B. L., L. R. Allen, H. Andrén, G. Ballard, L. Boitani, R. M. Engeman, P. J. S. Fleming, A. T. Ford, P. M. Haswell, R. Kowalczyk, J. D. C. Linnell, L. David Mech, and D. M. Parker. 2017. Can we save large carnivores without losing large carnivore science? Food Webs 12:64-75. Barnosky, A. D., E. a Hadly, J. Bascompte, E. L. Berlow, J. H. Brown, M. Fortelius, W. M. Getz, J. Harte, A. Hastings, P. a Marquet, N. D. Martinez, A. Mooers, P. Roopnarine, G. Vermeij, J. W. Williams, R. Gillespie, J. Kitzes, C. Marshall, N. Matzke, D. P. Mindell, E. Revilla, and A. B. Smith. 2012. Approaching a state shift in Earth's biosphere. Nature 486:52-8. Beckmann, J. P., and J. Berger. 2003. Rapid ecological and behavioural changes in carnivores: the responses of black bears (Ursus americanus) to altered food. Journal of Zoology 261:207-212. Bouyer, Y., V. Gervasi, P. Poncin, R. C. Beudels-Jamar, J. Odden, and J. D. C. Linnell. 2014. Tolerance to anthropogenic disturbance by a large carnivore: the case of Eurasian lynx in south-eastern Norway. Animal Conservation 18:271-278. Breitenmoser, U., C. Angst, J.-M. Landary, C. Breitenmoser-Wursten, J. D. C. Linnell, and J.-M. Weber. 2005. Non-lethal techniques for reducing predation. Pages 49-71 in R. Woodroffe, S. Thirgood, and A. Rabinowitz, editors. People & Wildlife: Conflict or Coexistence? Cambridge University Press, Cambridge, UK. 58 Brockington, D., and D. Wilkie. 2015. Protected areas and poverty. Philosophical Transactions of the Royal Society B: Biological Sciences 370:20140271. Capitani, C., M. Chynoweth, J. Kusak, E. Çoban, and Ç. H. Şekercioğlu. 2016. Wolf diet in an agricultural landscape of north-eastern Turkey. Mammalia 80:329-334. Chynoweth, M. W., E. Çoban, Ç. Altın, and Ç. H. Şekercioğlu. 2016. Human-wildlife conflict as a barrier to large carnivore management and conservation in Turkey. Turkish Journal of Zoology 40:972-983. Chynoweth, M. W., E. Çoban, and Ç. H. Şekercioğlu. 2015. Conservation of a new breeding population of Caucasian lynx (Lynx lynx dinniki) in eastern Turkey. Turkish Journal of Zoology 39:541-543. Cozzi, G., M. Chynoweth, J. Kusak, E. Çoban, A. Çoban, A. Ozgul, and Ç. H. Şekercioğlu. 2016. Anthropogenic food resources foster the coexistence of distinct life history strategies: year-round sedentary and migratory brown bears. Journal of Zoology 300:142-150. Crooks, K. R., C. L. Burdett, D. M. Theobald, C. Rondinini, and L. Boitani. 2011. Global patterns of fragmentation and connectivity of mammalian carnivore habitat. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 366:2642-51. Dirzo, R., H. S. Young, M. Galetti, G. Ceballos, N. J. B. Isaac, and B. Collen. 2014. Defaunation in the Anthropocene. Science 345:401-406. Dorresteijn, I., J. Schultner, D. G. Nimmo, J. Fischer, J. Hanspach, T. Kuemmerle, L. Kehoe, and E. G. Ritchie. 2015. Incorporating anthropogenic effects into trophic ecology: predator-prey interactions in a human-dominated landscape. Proceedings of the Royal Society B: Biological Sciences 282:20151602. Dudley, N., and S. Stolton. 1999. Conversion of "‘Paper Parks'" to Effective Management - Developing a Target. IUCN, WWF, WCPA. Estes, J. A., J. Terborgh, J. S. Brashares, M. E. Power, J. Berger, W. J. Bond, S. R. Carpenter, T. E. Essington, R. D. Holt, J. B. C. Jackson, R. J. Marquis, L. Oksanen, T. Oksanen, R. T. Paine, E. K. Pikitch, W. J. Ripple, S. A. Sandin, M. Scheffer, T. W. Schoener, J. B. Shurin, A. R. E. Sinclair, M. E. Soule, R. Virtanen, and D. A. Wardle. 2011. Trophic downgrading of planet Earth. Science 333:301-306. Fedriani, J. M., T. K. Fuller, and R. M. Sauvajot. 2001. Does availability of anthropogenic food enhance densities of omnivorous mammals? An example with coyotes in Southern California. Ecography 24:325-331. Fischer, J. D., S. H. Cleeton, and J. R. Miller. 2012. Urbanization and the predation 59 paradox: the role of trophic dynamics in structuring vertebrate communities. BioScience 62:809-818. Fiske, I. J., and R. B. Chandler. 2011. unmarked: an R package for fitting hierarchical models of wildlife occurrence and abundance. Journal of Statistical Software 43:1- 23. Frank, L., and R. Woodroffe. 2001. Behaviour of carnivores in controlled and exploited populations. Pages 419-442 in J. Gittleman, R. Wayne, D. Macdonald, and S. Funk, editors. Carnivore Conservation. Cambridge University Press, Cambridge, UK. Gross, L. 2008. No place for predators? PLoS Biology 6:199-204. Gunther, K. A. 1994. Bear Management in Yellowstone National Park, 1960-93. Pages 549-560 Ninth International Association for Bear Research and Management. Hobbs, R. J., S. Arico, J. Aronson, J. S. Baron, P. Bridgewater, V. A. Cramer, P. R. Epstein, J. J. Ewel, C. A. Klink, A. E. Lugo, D. Norton, D. Ojima, D. M. Richardson, E. W. Sanderson, F. Valladares, M. Vila, R. Zamora, and M. Zobel. 2006. Novel ecosystems: theoretical and management aspects of the new ecological world order. Global Ecology and Biogeography 15:1-7. Huber, D., J. Kusak, A. Majic-Skrbinsek, D. Majnaric, and M. Sindicic. 2007. A multidimensional approach to managing the European brown bear in Croatia. Ursus 19:22-32. Imbert, C., R. Caniglia, E. Fabbri, P. Milanesi, E. Randi, M. Serafini, E. Torretta, and A. Meriggi. 2016. Why do wolves eat livestock?: factors influencing wolf diet in northern Italy. Biological Conservation 195:156-168. Krofel, M., and K. Jerina. 2016. Mind the cat: conservation management of a protected dominant scavenger indirectly affects an endangered apex predator. Biological Conservation 197:40-46. Layman, C. a., S. D. Newsome, and T. Gancos Crawford. 2015. Individual-level niche specialization within populations: emerging areas of study. Oecologia 178:1-4. MacKenzie, D. I., J. D. Nichols, G. B. Lachman, S. Droege, J. Andrew Royle, and C. A. Langtimm. 2002. Estimating site occupancy rates when detection probabilities are less than one. Ecology 83:2248-2255. MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and J. E. Hines. 2006a. Occupancy Estimation and Modeling: Inferring Patterns and Dynamics of Species Occurrence. Academic, New York, NY. MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and J. E. Hines. 60 2006b. Occupancy Estimation and Modeling: Inferring Patterns and Dynamics. Elsevier Inc., Burlington, MA. Messmer, T. A. 2000. The emergence of human-wildlife conflict management: turning challenges into opportunities. International Biodeterioration & Biodegradation 45:97-102. Moss, W. E., M. W. Alldredge, and J. N. Pauli. 2016. Quantifying risk and resource use for a large carnivore in an expanding urban-wildland interface. Journal of Applied Ecology 53:371-378. Newsome, S. D., H. M. Garbe, E. C. Wilson, and S. D. Gehrt. 2015a. Individual variation in anthropogenic resource use in an urban carnivore. Oecologia 178:115-128. Newsome, T. M., L. Boitani, G. Chapron, P. Ciucci, C. R. Dickman, J. A. Dellinger, J. V. López-Bao, R. O. Peterson, C. R. Shores, A. J. Wirsing, and W. J. Ripple. 2016. Food habits of the world's grey wolves. Mammal Review 46:255-269. Newsome, T. M., J. a. Dellinger, C. R. Pavey, W. J. Ripple, C. R. Shores, A. J. Wirsing, and C. R. Dickman. 2015b. The ecological effects of providing resource subsidies to predators. Global Ecology and Biogeography 24:1-11. Ordiz, A., R. Bischof, and J. E. Swenson. 2013. Saving large carnivores, but losing the apex predator? Biological Conservation 168:128-133. R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Redford, K. H. 1992. The empty forest. BioScience 42:412-422. Ripple, W. J., J. A. Estes, R. L. Beschta, C. C. Wilmers, E. G. Ritchie, M. Hebblewhite, J. Berger, B. Elmhagen, M. Letnic, M. P. Nelson, O. J. Schmitz, D. W. Smith, a. D. Wallach, and A. J. Wirsing. 2014. Status and ecological effects of the world's largest carnivores. Science 343:1241484-1241484. Ripple, W. J., T. M. Newsome, C. Wolf, R. Dirzo, K. T. Everatt, M. Galetti, M. W. Hayward, G. I. H. Kerley, T. Levi, P. A. Lindsey, D. W. Macdonald, Y. Malhi, L. E. Painter, C. J. Sandom, J. Terborgh, and B. Van Valkenburgh. 2015. Collapse of the world's largest herbivores. Science Advances 1:e1400103. Rovero, F., and A. R. Marshall. 2009. Camera trapping photographic rate as an index of density in forest ungulates. Journal of Applied Ecology 46:1011-1017. Rovero, F., E. Martin, M. Rosa, J. a. Ahumada, and D. Spitale. 2014. Estimating species richness and modelling habitat preferences of tropical forest mammals from camera trap data. PLoS One 9:e103300. 61 Seddon, P. J., C. J. Griffiths, P. S. Soorae, and D. P. Armstrong. 2014. Reversing defaunation: restoring species in a changing world. Science 345:406-412. Şekercioğlu, Ç. H., S. Anderson, E. Akçay, R. Bilgin, Ö. E. Can, G. Semiz, Ç. Tavşanoğlu, M. B. Yokeş, A. Soyumert, K. İpekdal, İ. K. Sağlam, M. Yücel, and H. Nüzhet Dalfes. 2011. Turkey's globally important biodiversity in crisis. Biological Conservation 144:2752-2769. Steffen, W., P. J. Crutzen, and J. R. McNeill. 2007. The Anthropocene: are humans now overwhelming the great forces of nature? AMBIO: A Journal of the Human Environment 36:614-621. Stringham, S. F. 1989. Demographic consequences of bears eating at garbage dumps: an overview. Pages 35-42 in M. Bromley, editor. Bear-people conflicts: Proceedings of a symposium on management strategies. Northwest Territories Department of Renewable Resources, Yellowknife, Northwest Territories, Canada. Svenning, J.-C., P. B. M. Pedersen, C. J. Donlan, R. Ejrnæs, S. Faurby, M. Galetti, D. M. Hansen, B. Sandel, C. J. Sandom, J. W. Terborgh, and F. W. M. Vera. 2016. Science for a wilder Anthropocene: synthesis and future directions for trophic rewilding research. Proceedings of the National Academy of Sciences 113:898-906. Terborgh, J. 2001. Ecological meltdown in predator-free forest fragments. Science 294:1923-1926. Tobler, M. W. 2010. Camera Base Version 1.4. Botanical Research Institute of Texas. Zamora, R., J. M. Gómez, J. A. Hódar, J. Castro, and D. García. 2001. Effect of browsing by ungulates on sapling growth of Scots pine in a mediterranean environment: consequences for forest regeneration. Forest Ecology and Management 144:33-42. Zlatanova, D., A. Ahmed, A. Valasseva, and P. Genov. 2014. Adaptative diet strategy of the wolf (Canis lupus L.) in Europe: a review. Acta Zoologica Bulgarica 66:439- 452. 62 Figure 3.1. The location of Sarıkamış-Allahuekber Mountains National Park and Turkey's first wildlife corridor at the intersection of two global biodiversity hotspots. Turkey is the only country in the world for which >80% is covered by three separate global biodiversity hotspots. 63 Figure 3.2. Our study area on the border of Kars/Erzurum provinces in eastern Turkey including Sarıkamış-Allahuekber Mountains National Park, surrounding human settlements and camera trap stations for all survey years. 64 Figure 3.3. Trapping rate of all species captured every year during a camera-trapping survey conducted 2013-2016 in Sarıkamış forest, eastern Turkey. Species not captured during all years are not included, see Table 3.1 for a full species list. 65 Figure 3.4. Activity patterns of large carnivores, natural prey species, humans, and livestock during a camera-trapping survey conducted 2013-2016 in Sarıkamış forest, eastern Turkey. 66 Figure 3.5. Single-species single-season occupancy estimates with 95% confidence intervals by year for a camera-trapping survey conducted 2013-2016 in Sarıkamış forest, eastern Turkey. Occupancy models for humans in 2013 and 2014 did not converge, resulting in an occupancy estimate of 1 with no associated error. 67 Figure 3.6. Representative mammal community structure for a camera-trapping survey conducted during 2013-2016 in Sarıkamış forest, eastern Turkey. Figure shows the distribution of species along two functional traits: body size (expressed on a log scale) and trophic category. Each circle in the figure represents a species in functional space, with the size of the circle proportional to the estimated occupancy. 68 Table 3.1. Wild mammal species documented in Sarıkamış forest, eastern Turkey camera-trapping survey conducted 2004-2016 in Sarıkamış forest. Order Artiodactyla Artiodactyla Carnivora Carnivora Carnivora Carnivora Carnivora Carnivora Carnivora Erinaceomorpha Genus Capreolus Sus Lynx Ursus Meles Canis Vulpes Martes Felis Erinaceus Species capreolus scrofa lynx arctos meles lupus vulpes foina sylvestris concolor Lagomorpha Rodentia Rodentia Rodentia Lepus Sciurus Chionomys Allactaga europaeus vulgaris nivalis williamsi Carnivora Carnivora Sub species Dinniki Common Name roe deer wild boar Caucasian lynx Eurasian brown bear European badger gray wolf red fox stone marten wildcat southern white-breasted hedgehog European hare red squirrel snow vole Williams jerboa Mammal carcasses discovered in Sarıkamış forest by authors Lutra lutra Eurasian otter Vormela peregusna marbled polecat 69 Table 3.2. Camera trapping effort in Sarıkamış forest, eastern Turkey conducted 20042016 in Sarıkamış forest. Camera trapping days Mean trapping days per camera Successful stations Cameras stolen Pilot Study (2004-2012) 1652 - 2013 2014 2015 2016 1365 112 12 1 3917 76 51 9 3832 1965 127 81 27 24 7 6 Table 3.3. Modeled occupancy and detection probabilities for 2013 camera trapping effort in Sarıkamış forest, eastern Turkey. Modeling criteria A is naïve occupancy estimate >0.1, modeling criteria B is modeled detection probability 0.1. Occupancy Species No. of Events Detection Probability Modeling Criteria Brown bear 45 Naïve Estimate 0.75 Modeled Estimate 0.772 Standard Error 0.13 Lower 95% CI 0.75 Upper 95% CI 0.935 Modeled Estimate 0.17 Standard Error 0.03 Lower 95% CI 0.119 Upper 95% CI 0.238 Livestock 46 0.417 0.42 0.143 0.417 0.696 0.23 0.044 0.155 0.327 Caucasian lynx 8 0.417 0.782 0.44 0.417 0.998 0.039 0.026 0.011 0.135 Eurasian red squirrel 5 0.25 0.357 0.215 0.25 0.778 0.061 0.039 0.017 0.197 European hare 5 0.333 0.858 0.73 0.333 1 0.026 0.024 0.004 0.151 Gray wolf 21 0.5 0.53 0.155 0.5 0.792 0.141 0.035 0.085 0.224 human 278 1 1 0 1 1 1 0 1 1 Wild boar 20 0.421 0.474 0.026 0.421 0.719 0.066 0.017 0.04 0.107 A B Table 3.4. Modeled occupancy and detection probabilities for 2014 camera trapping effort in Sarıkamış forest, eastern Turkey. Modeling criteria A is naïve occupancy estimate >0.1, modeling criteria B is modeled detection probability 0.1. Occupancy Species No. of Events Detection Probability Brown bear 116 Livestock 146 0.455 0.494 0.114 0.455 0.706 0.169 0.029 0.12 0.233 Caucasian lynx 33 0.727 1 0.017 0.727 1 0.165 0.013 0.044 0.195 Eurasian red squirrel 19 0.273 0.311 0.109 0.273 0.551 0.13 0.034 0.076 0.213 European hare Modelled Estimate 0.793 Standard Error 0.092 Lower 95% CI 0.773 Upper 95% CI 0.92 Modelled Estimate 0.18 Standard Error 0.024 Modeling Criteria Naïve Estimate 0.773 Lower 95% CI 0.138 Upper 95% CI 0.231 8 0.136 0.174 0.1 0.136 0.45 0.089 0.044 0.032 0.223 Gray wolf 142 0.909 0.922 0.063 0.909 0.985 0.207 0.022 0.166 0.254 human 1451 1 1 0 1 1 1 0 1 1 58 0.136 0.163 0.026 0.136 0.163 0.039 0.018 0.015 0.095 Wild boar A B 70 Table 3.5. Modeled occupancy and detection probabilities for 2015 camera trapping effort in Sarıkamış forest, eastern Turkey. Modeling criteria A is naïve occupancy estimate >0.1, modeling criteria B is modeled detection probability 0.1. Occupancy Species Detection Probability Modeling Criteria No. of Events Naïve Estimate Modelled Estimate Standard Error Lower 95% CI Upper 95% CI Modelled Estimate Standard Error Lower 95% CI Upper 95% CI Brown bear 130 0.737 0.834 0.122 0.737 0.966 0.114 0.022 0.078 0.164 Livestock 95 0.368 0.404 0.124 0.368 0.651 0.124 0.031 0.074 0.199 Caucasian lynx 34 0.316 0.372 0.133 0.316 0.643 0.197 0.032 0.051 0.179 Eurasian red squirrel 12 0.158 0.229 0.145 0.158 0.597 0.061 0.039 0.017 0.196 European hare 53 0.211 0.215 0.095 0.211 0.453 0.194 0.047 0.118 0.302 Gray wolf 103 0.526 0.545 0.119 0.526 0.755 0.162 0.029 0.114 0.227 human 528 0.842 0.844 0.084 0.765 0.949 0.288 0.026 0.239 0.342 Wild boar 63 0.421 0.474 0.133 0.421 0.719 0.109 0.029 0.064 0.18 A B Table 3.6. Modeled occupancy and detection probabilities for 2016 camera trapping effort in Sarıkamış forest, eastern Turkey. Modeling criteria A is naïve occupancy estimate >0.1, modeling criteria B is modeled detection probability 0.1. Species No. of Events Naïve Estimate Modelled Estimate Brown bear 128 0.833 0.856 Livestock 15 0.292 Caucasian lynx 25 Eurasian red squirrel 16 European hare Occupancy Standard Error Detection Probability Standard Lower Error 95% CI Modeling Criteria Lower 95% CI Upper 95% CI Modelled Estimate Upper 95% CI 0.079 0.833 0.954 0.215 0.026 0.169 0.269 0.434 0.175 0.292 0.757 0.074 0.033 0.03 0.17 0.333 0.376 0.112 0.333 0.606 0.141 0.037 0.083 0.23 0.25 0.424 0.205 0.25 0.793 0.061 0.032 0.021 0.163 28 0.417 0.475 0.121 0.417 0.7 0.092 0.033 0.08 0.211 Gray wolf 74 0.542 0.546 0.103 0.542 0.73 0.278 0.034 0.216 0.349 human 252 0.833 0.836 0.076 0.833 0.938 0.328 0.028 0.275 0.385 Wild boar 21 0.417 0.77 0.302 0.417 0.989 0.053 0.024 0.021 0.126 A B 71 CHAPTER 4 MOVEMENT ECOLOGY AND RESOURCE SELECTION OF THREE LARGE CARNIVORES IN A PREY-DEFICIENT, HIGHLY DEGRADED ECOSYSTEM 4.1 Abstract Conservation and management of wildlife populations is becoming increasingly complex in a world where novel and hybrid ecosystems are emerging from humandominated landscapes. As human impact intensifies, it is important to consider movement and resource selection of wildlife to inform sustainable management decisions in highlymodified landscapes. Resource selection models can identify important habitat for species and guide conservation efforts to increase protected area coverage. To understand how large carnivores are able to coexist with people in heavily modified landscapes, we deployed GPS collars on 16 adult Eurasian brown bears, 7 gray wolves, and 2 Caucasian lynx in eastern Turkey to study their movement ecology. We developed species-specific seasonal resource selection functions to identify high-priority habitat in the area, and to identify suitable habitat for increasing protected area coverage. All 3 species' habitat selection varied between seasons. Brown bears selected for areas closer to paved roads, further from human settlements, and located on steeper slopes throughout the year. During spring, bears preferred lower elevations and more open areas, and during summer 73 and fall, bears preferred higher elevations. Wolves selected for forested areas, areas closer to roads, farther from villages, and steeper slopes throughout the year. Wolves selected for higher elevations during summer and lower elevations during winter. Lynx selected for steeper slopes throughout the year. During summer, lynx selected for forested areas, areas farther from villages, and higher elevations, while during winter, they selected for areas slightly closer to forests. Using predictive maps, we identified important habitat in the area for all three species and propose a new protected area designation in the region. 4.2 Introduction Human activity has become the driving ecological force on our planet, shaping ecosystem structure and function in the new geological epoch, the Anthropocene (Zalasiewicz et al. 2008). The impact of human domination can be detected in geological, chemical, and biological signals, but few are as daunting as what is known as the sixth mass extinction (Barnosky et al. 2011). Part of this process is global defaunation, particularly of large mammals (Dirzo et al. 2014). Large mammals are highly vulnerable to extirpation by humans due to their large size, slow reproductive rates, and vast habitat requirements (Cardillo 2005). Within this group, large carnivores represent a distinct faction, particularly at risk because of their natural rarity and risk of persecution from humans. Nonetheless, some large carnivore species are able to alter their behavior to exploit new resources in a changing environment. Increasingly, wilderness and natural areas are surrounded by a matrix of highly altered, human-modified landscapes (Hobbs et al. 2014). This juxtaposition of habitat can 74 result in a combination of suitable habitat with an abundance of anthropogenic food resources for generalist predators. Anthropogenic food resources result in behavioral changes in predators (Cozzi et al. 2016), which are likely to reduce the effect of trophic cascades (Newsome et al. 2015). A consequence of becoming habituated to human food sources is increased human-wildlife conflict, which may include vehicle collisions and direct persecution via poisoning or illegal hunting. However, human-altered landscapes may also provide benefits, such as highly productive agricultural areas or increased prey availability through livestock presence. This is a new avenue of research in large carnivore ecology, supported by recent studies suggesting limited scientific support for popular concepts in carnivore ecology such as trophic cascades (Haswell et al. 2017, Kuijper et al. 2016, Allen et al. 2017). In our changing world, there is a critical need for more data on large carnivore movement in human-dominated landscapes. Resource selection functions (RSFs) can serve several purposes when examining large carnivore ecology in human-dominated landscapes. Gaining an understanding of how these animals use resources can help scientists and managers understand how human-carnivore coexistence can occur and what conditions must exist for its facilitation (Carter et al. 2012, Oriol-Cotterill et al. 2015). RSFs can also identify patterns in the habitat selection of threatened species to enable their persistence (Dellinger et al. 2013), as well as to identify potential corridors for wide-ranging species (Chetkiewicz and Boyce 2009). We used RSFs to investigate habitat selection and movement patterns of three large carnivore species in eastern Turkey, in a fragmented forest within a human-dominated landscape surrounding Sarıkamış-Allahuekber Mountains National Park (hereafter 75 Sarıkamış Forest). At the intersection of the Caucasus and Iran-Anatolian global biodiversity hotspots, wildlife biology and mammal ecology in this region are mostly unstudied. Similar to many other parts of the developing world, Turkey's biodiversity is globally important and increasingly threatened (Şekercioğlu et al. 2011). We were interested in understanding how Eurasian brown bears (Ursus arctos arctos), gray wolves (Canis lupus lupus), and Caucasian (Eurasian) lynx (Lynx lynx dinniki) are able to persist in a human-dominated landscape largely devoid of natural prey species (Chynoweth et al. in prep). By taking a multispecies approach, we hope that results from this study will help guide conservation efforts for the region, using these three landscape carnivores as umbrella species to increase protected area coverage in the region (Lambeck 1997). Our previous work in the same system documented a unique mammal community structure characterized by the hyperabundance of these three large carnivore species (Chynoweth et al. in prep), an absence of natural prey, and synanthropic behavior of bears (Cozzi et al. 2016) and wolves (Capitani et al. 2016). Importantly, using the data also presented here, we identified two distinct life history traits coexisting within this population of bears: bears that regularly visited the dump and remained sedentary yearround and bears that never visited the dump and migrated (see Cozzi et al. 2016 for details). Based on these results and the goal to inform conservation of large carnivores in Sarıkamış Forest specifically, we used in our analysis only the bears that did not visit the dump. We hypothesized that large carnivores in Sarıkamış Forest would show seasonal variation in habitat selection related to snow cover and associated human presence in the landscape. We hypothesized that all three species would select for steeper slopes and for areas closer to forests. We also hypothesized that wolves and bears would select for areas 76 closer to villages, while lynx, as an obligate carnivore, would select for forested areas and areas away from human activity. 4.3 Methods 4.3.1 Study Area (from section 3.3.1) Our study was carried out on the Kars-Ardahan high plateau in northeastern Turkey, at the intersection of Caucasus and Irano-Anatolian global biodiversity hotspots. The area (c. 550 km2; 40°20'N 42°35'E) ranges between 1900 and 3120 m asl and is composed of fragmented forest in a matrix of agricultural and rangelands. The city of Sarıkamış (population: c. 18,000) is located in the center of the study area and on one of the two paved roads that bisect the forested area. Forest cover consists almost exclusively of Scots pine (Pinus sylvestris Linnaeus, 1753), while understory vegetation is scarce, with consequent scarcity of food resources for browsers. Sarıkamış-Allahuekber Mountains National Park (hereafter SAMNP; Figure 4.1) boundaries cover a total area of 225.1 sq. km., but only include 49.69 sq. km. of forest. Therefore, SAMNP is only comprised of 22.07% forest cover. Total forest cover in the region includes 328.38 sq. km. including a large expanse of forest south of the national park (248.15 sq. km.). These patches of forest represent the southernmost significant forest patch in the region extending south from the extensive forests in the Black Sea Region of Turkey. Human activity in the forest is extensive in both time and space, limited only by harsh winter temperatures, and consists primarily of livestock grazing, harvest of forest products (e.g., fruits, pine cones, mushrooms), and legal and illegal timber extraction. Livestock is abundant in the region with cattle (Bos taurus Linnaeus, 1758), sheep (Ovis 77 aries Linnaeus, 1758), and goats (Capra hircus Linnaeus, 1758) freely roaming rangelands from April to November (Capitani et al. 2016). About 851,445 livestock heads have been registered in the Kars province in 2012 (Ministry of Food, Agriculture and Livestock, Republic of Turkey). A notable feature on the landscape is an unfenced municipal garbage dump 3 km west of Sarıkamıs city. The dump represents a predictable anthropogenic food source, and bears, wolves, and wild boar visit the dump regularly at night (pers. obs.). A portion of the bear population has altered life history strategies to regularly use the dump, while other bears never visit the dump (Cozzi et al. 2016). 4.3.2 Animal Capture From 2012-2014, we captured bears, wolves, and lynx of a variety of ages and fitted satellite transmitters to them. An experienced carnivore biologist from Zagreb University and a wildlife veterinarian from local Kafkas University were present for the captures and the necessary permits were obtained from Turkey's Ministry of Forestry and Water Affairs. Data were added to the online Movebank database and are available upon request. 4.3.2.1 Bears We captured 18 males and 10 females using Aldrich snares in opposing entrances to European-style cubbies baited with fresh sheep carcasses. GPS/GSM radio collars (GPS Plus; Vectronic Aerospace GmbH, Berlin, Germany) were attached to immobilized bears after aging and health assessment. Bears were monitored for a mean duration of 296 days (range: 125-590 days). GPS acquisition rate was >90% for all but one individual; only 78 location fixes with a three-dimensional fix and low Positional Dilution of Position value (PDOP < 10) were included in final datasets for analysis. Data recorded during hibernation were not included in any analysis. 4.3.2.2 Wolves We captured 7 male and 4 female wolves using padded leg hold traps cross-baited with wolf scat, wolf urine and/or rotten liver. GPS/GSM radio collars (GPS Plus; Vectronic Aerospace GmbH, Berlin, Germany) were attached to immobilized wolves after aging and health assessment. Collars were programmed to log a GPS location every 6 hours for 1 year. Wolves were monitored for a mean duration of 307 days (range: 167- 365 days). GPS acquisition rate was >90% for all individuals; only location fixes with a three-dimensional fix and low Positional Dilution of Position value (PDOP < 10) were included in final datasets for analysis. 4.3.2.3 Lynx We captured 2 adult male lynx using live box traps custom designed in Sarıkamış. GPS/GSM radio collars (GPS Plus; Vectronic Aerospace GmbH, Berlin, Germany) were attached to immobilized lynx after aging and health assessment. Collars were programmed to log a GPS location every 4 hours for 1 year. The 2 individual lynx were monitored for 342 and 283 days. GPS acquisition rate was >90% for all individuals; only location fixes with a three-dimensional fix and low Positional Dilution of Position value (PDOP < 10) were included in final datasets for analysis. 79 4.3.3 Home Ranges and Utilization Distributions Utilization distribution (UD) and home range area estimates were calculated using adaptive-kernel density estimators with adehabitat package in R (Calenge 2006). Home range estimates were generated with an ad hoc smoothing parameter using the smallest increment of the reference bandwidth (href) that provided a contiguous 95% kernel home range (i.e., h = 0.5 × href, 0.6 × href,... href-J. Kie, pers. comm.). The number of points used to generate annual utilization distributions ranged from 674 to 16,497, providing robust estimates of kernel density. Home range estimates provide a 95% utilization distribution and a 95% isopleth home range for individual animals at a 30 × 30 m resolution. 4.3.4 Environmental Data We compiled existing environmental data from our study area in a Geographic Information System (GIS). To test the effect of landscape features on animal movement and habitat selection, we created six geo-referenced raster layers that included distance to the nearest village, distance to the nearest paved road, distance to forest cover, altitude, slope, and aspect (Table 4.1). Each layer fully covered the extended study area and was characterized by a cell size of 30 x 30 m. All six variables were retained for further analyses since we did not detect strong correlations (r < 0.37 for any pair). We obtained a land cover map from the Turkey's Ministry of Forestry and Water Affairs at a resolution of 1:25,000, which included two major land cover types: forest and open land, and calculated the distance between each raster centroids and the closest forest patch, the mean distance being 6,056 m (range: 0.1 - 42,727 m). We acquired 80 topographic information on altitude (mean: 1933 m, range: 881 - 3132 m), slope (mean: 16 degrees, range: 0 - 75 degrees), and aspect (mean: 178 degrees, range: 0 - 360 degrees) from an ASTER Global Digital Elevation Map (http://reverb.echo.nasa.gov). We calculated the distance between each raster centroids and the closest road, the mean distance being 5,643 m (range: 0.1 - 26,499 m). Because this study was located in a rural area with very low traffic, we only considered paved national and regional roads (Turkey's Ministry of Forestry and Water Affairs). Finally, we obtained a GIS layer from Turkey's Ministry of Forestry and Water Affairs with the locations of 356 villages (mean density: 1 village/20 km2). Villages were relatively evenly distributed throughout the entire study area with mean distance between raster centroids and villages equal to 2,502 m (range: 0.5 - 11,046 m). 4.3.5 Model Development, Selection, and Predictions We analyzed bear, wolf, and lynx resource selection based on Johnson (1980) "third order selection," or how an animal uses habitat components within its home range. We used a resource selection function approach with a use-availability design (Manly et al. 2002) to examine the relationship between resource use and habitat covariates. Thus, we determined used and available habitat within each animal's 95% kernel home range. Recorded GPS locations were considered to be used locations, because we knew that the animal was present at that location at a given time. Available habitat was determined by systematically sampling habitat at 100 m intervals within each animal's 95% kernel home range. This resulted in a database with all used and available points for each individual of each species with all associated covariate data with a binary response variable of use. 81 Based on our previous work and environmental conditions in the area, our aim was to understand how these three species select for resources in a human-dominated, preydeficient system. To reflect seasonal resource selection among species, we partitioned location data by season from approximate dates of snow cover and food availability, based on authors' multiyear experience and previous work in the study area (Capitani et al. 2016, Cozzi et al. 2016). Wolf and lynx data were partitioned into two seasons: winter (Nov 16 - April 15) and nonwinter (April 16 - Nov 15). Brown bear data were partitioned into three seasons: spring (end hibernation - May 31), summer (June 1 - Aug 31), and fall (Sept 1 - begin hibernation). All bear migration data identified in Cozzi et al. (2016) were excluded from analysis in this manuscript, because bears were moving through large expanses of open habitat to very different forest types; therefore including these points would not inform conservation efforts specific to Sarıkamış Forest. RSFs were created by comparing seasonal bear, wolf, and lynx locations to available location within an individual's annual home range. All covariate data were standardized to enable comparison between the effects of covariates on habitat selection and to aid in model convergence. Individual animals were treated as a random effect to account for interindividual variability. We used generalized linear models (GLMs) with the lme4 package in R and assessed the variance inflation factors (VIF) with the usdm package in R (threshold |r| > 0.50). An information theoretic approach for model selection was used to compare models (Burnham and Anderson 2002). We employed backward stepwise selection for final model selection, using the Aikake Information Criterion, adjusted for small sample sizes (AICc) with the 82 AICcmodavg package (Mazerolle and Mazerolle 2011). To start, all variables were included in every model; then we removed those that were nonsignificant until all variables left were significant at the p =0.05 level. Ninety-five percent confidence intervals were calculated to ensure that each variable was truly significant Final RSF models were input in ArcGIS 10.3.1 to generate population level probability of use for each species at each 30 m resolution cell across the study area. Probabilities were classified into five quantile bins to represent areas of low - high habitat suitability, which represent categories of increasing habitat selection (Johnson et al. 2006). To test the accuracy of our final models, we followed a 10-fold cross validation procedure to examine model performance (Boyce et al. 2002). We split data into 10 equal parts (folds) and kept observation (i.e., used and alternative locations) from the same strata in the same fold. We then fit the model to all data except the ith fold and calculated parameter estimates (β1,… βn). We used the calculated β parameters to estimate w(X) values for the ith fold. We then binned the data based on the deciles of the estimated w(X) values and calculated the spearman correlation coefficient (rs) between the proportion of used locations in each bin and the mean w(X) value in each bin. All statistical analyses were conducted in R (R Core Team 2016). 4.3.6 Prioritizing Conservation Areas We used all validated species-specific seasonal models overlaid in ArcGIS to determine the highest suitability habitat for all three species. Relative probability of habitat use was summed across species and seasons and reclassified into four quantile 83 bins, and the highest category bin was used to represent the most suitable habitat for future conservation efforts. Through this subjective conversion, we assumed that areas identified by this method are the highest quality across seasons and species. We converted the re-binned raster surface into polygons using ArcGIS and selected the single largest contiguous polygon of the highest quantile bin in the region and removed all smaller patches of possible area. We then clipped this single polygon by forest cover to identify areas suitable for protected area status given that deforested areas are typically rangeland. Lastly, we bounded these polygons by a minimum convex polygon to identify a contiguous area that includes all forested area with the highest probability of habitat use for all three species. 4.4 Results We captured 28 bears, 11 wolves, and 3 lynx between 2011 and 2016. Collars were only deployed on adult animals. Equipment failure and mortality (actual and perceived) reduced our sample size to 16 bears, 7 wolves, and 2 lynx. Most animals spent the majority of their time in the main study area (see Cozzi et al. 2016 for an explanation of migratory bear behavior), but notably, two wolves were dispersing individuals and were not retained in the dataset. Mean home ranges for bears were 109 ± 17.67 km2, while mean home ranges for wolves and lynx were highly variable, 1263.1 ± 786.8 km2 and 1296.8 ± 917 km2, respectively, due to small sample size and one wide ranging individual of each species (Table 4.2). 84 4.4.1 Resource Selection Functions Bears selected for sites farther from paved roads, closer to human settlements, and steeper slopes during all seasons. During the spring season, bears preferred lower elevations and more open areas, and during the fall season, bears preferred higher elevations (Figure 4.2). Predictive accuracy for seasonal models using withheld modeltesting data was variable between seasons (spring; r2 =0.833 (p < 0.05), summer; r2 =0.923 (p < 0.001), autumn; r2 =0.948 (p < 0.001)). Wolves selected for forested areas, as well as areas closer to roads, farther from villages, and with steeper slopes during all seasons. They selected for higher elevations during the summer season and for lower elevations during the winter season (Figure 4.3). Predictive accuracy for seasonal models using withheld model-testing data was variable between seasons with the winter season model performing poorly (summer; r2 =0.865 (p < 0.05), winter; r2 =0.347 (n.s.)). Lynx selected for steeper slopes during all seasons. During the winter season, they selected for areas farther from human settlements, closer to forests, and higher in elevation, while during nonwinter seasons, they selected for areas outside of forests (Figure 4.4). Predictive accuracy for seasonal models using withheld model-testing data was variable between seasons with the winter season model performing poorly (summer; r2 =0.929 (p < 0.001), winter; r2 =0.435 (n.s.)). 4.4.2 Prioritizing Conservation Areas All three seasonal RSF models for bears were validated, and summer models for wolves and lynx were validated. By combining all validated seasonal models across 85 species, an area of 161.4 km2 was identified as the highest quality habitat for bears, wolves, and lynx in the region to prioritize future conservation efforts (Figure 4.8). This area includes 102.6 km2 of forest and one human settlement and is bisected by a two-lane highway. SAMNP is the only other protected area in the region with 49.7 km2 total forested area within park boundaries Therefore, protecting this prioritized area would double protected forest (106% increase) and result in protection of 46.4% of the total forest cover in the region. 4.5 Discussion Our study contributes critical information about brown bear, gray wolf, and Eurasian lynx home ranges and resource selection in an understudied region of the world. Speciesspecific seasonal resource selection models identify areas in the landscape with high probability of habitat selection at the population level and can be used to identify suitable habitat for a species. There was some variation in model performance between seasons; bear models performed well for all three seasons, while wolf and lynx summer models performed well, and winter models had low (nonsignificant) model validation scores. Poor model performance for winter models is likely related to the wide-ranging behavior of wolves and lynx during this season. Animals are traveling through open landscapes more often and depending on the assignment of data to any of the 10 folds, some strata during winter season have no used or alternative locations in forest. For this season, we caution the interpretation of due to the poor model performance. All validated model outputs were used to identify an area to be targeted for future conservation efforts, with the overall goal to increase protected area in region. 86 Resource selection by brown bears, gray wolves, and Eurasian lynx varied by season, similar to other studies (Chetkiewicz and Boyce 2009, Latham et al. 2013). Across all three species, elevation had the largest and most varied effect on habitat selection. For wolves and lynx, this is likely related to contrast in human activity patterns between seasons. During the winter, when wolves and lynx are selecting for low elevation areas, harsh weather conditions prevent villagers from accessing the majority of forested area, as well as the vast open rangelands surrounding the forest. Therefore, wolves and lynx can move more freely through low elevation open areas. During the summer season, when wolves and lynx are strongly selecting for high elevation forested areas, human activity is frequent and ubiquitous throughout the forest. Given the lack of natural prey in the area (Chynoweth et al. in prep) and known predation of livestock by wolves (Capitani et al. 2016), these two species are likely forced to travel farther distances in search of prey during the winter season when livestock are in closed, protected pens near villages. This corresponds with anecdotal evidence from villagers who state they frequently see wolves during winter months close to villages preying on dogs and other domestic animals (Chynoweth et al. 2016). Brown bears also demonstrate strong seasonal patterns related to elevation. During the spring season, the importance of primary productivity is likely driving their selection for lower elevations, based on their reliance on herbaceous vegetation as an omnivore (Rode et al. 2001, Robbins et al. 2004) and on the lack of food resources at higher elevations covered in snow. During the summer season, bears begin to select for higher elevation with selection becoming stronger in the fall, likely related to human activity in the forest and food availability similar to results from wolf and lynx models. Den site 87 selection also impacts selection for high elevation in the fall, as these sites are all located in high elevation, steep, and rugged terrain (Chynoweth, pers obs). Distance to forest also had a varied effect on RSFs across season and species. Similar to patterns observed in selection for elevation, the strong selection for forest demonstrated by both wolves and lynx in the summer is probably driven by human activity and the importance of forest cover as a refuge from potential persecution. Wolves also selected for forest cover during the winter season. Both lynx individuals rarely left forested areas during the summer season. However, model results for lynx suggested avoidance of forest cover by lynx during the winter season. This is a product of small sample size (n=2) and individual variation in home range and movement patterns. One lynx was completely confined by forest cover in a well-defined home range, while the other lynx had a much less well-defined territory and also took a 3-month foray out into open rangelands during the winter season, wandering over 40 km from the nearest forest and close to Kars city (human population of ~80,000 people). Model validation results for both wolves and lynx was poor for winter seasons, likely a result of this individual variation. Slope was a significant predictor of habitat selection for all species in all seasons, with positive relationship between slope and probability of use. Field observations by authors suggest that slope may be a proxy for forest quality in Sarıkamış forest due to the heavy grazing that occurs throughout the forest, which has had long-term consequences for forest structure. Steeper areas are more difficult for livestock to access and are therefore less intensely grazed and maintain more understory vegetation than flatter areas. Our prediction that bears and wolves would prefer areas closer to human settlements 88 was not supported, with both species selecting for areas farther from villages during all seasons. However, the distribution of anthropogenic food sources across the landscape is not well known, with the exception of the large municipal garbage dump located 3 km outside of the city of Sarıkamış. However, through our ongoing fieldwork in the region, we have identified the presence of numerous smaller garbage dumps throughout the area and documented the use of these refuse piles by wolves and bears. Furthermore, the presence of livestock, both live and as carcasses, is ubiquitous in this landscape during nonwinter months and likely plays an important role in the diet of wolves (Capitani et al. 2016) and possibly bears. Our results are not surprising, as wolves have been observed using garbage dumps nightly in Italy (Ciucci et al. 1997), and both black and brown bears are known to use open garbage dumps as a food source (Craighead and Craighead F. C 1972, Rogers et al. 2014). 4.6 Conclusion Our results are an example of using resource selection functions across multiple species to prioritize conservation efforts in an understudied region of the world in critical need of increased conservation efforts (Sekercioglu et al. 2011, Şekercioğlu et al. 2011). Understanding the behavior of large carnivores in this human-dominated landscape will also elucidate the mechanisms that allow these species to persist in an alternative stable state in a heavily degraded landscape devoid of natural prey (Chynoweth et al. in prep). While the use of anthropogenic resources by carnivores is well documented, these baseline results are a critical contribution to the expanding field of large carnivore ecology and conservation in human-dominated landscapes. However, more work needs to 89 be done to further understand if this could be an example of an ecological trap (Gates and Gysel 1978), evolutionary trap (Schlaepfer et al. 2002), or simply the persistence of these species beyond a tipping point. Regionally, these results are critical to science-based management of individual species and biodiversity. Results will be used to guide wildlife management, establish new protected areas, and provide solutions to existing humanwildlife conflict. 4.7 References Allen, B. L., L. R. Allen, H. Andrén, G. Ballard, L. Boitani, R. M. Engeman, P. J. S. Fleming, A. T. Ford, P. M. Haswell, R. Kowalczyk, J. D. C. Linnell, L. David Mech, and D. M. Parker. 2017. Can we save large carnivores without losing large carnivore science? Food Webs 12:64-75. Barnosky, A. D., N. Matzke, S. Tomiya, G. O. U. Wogan, B. Swartz, T. B. Quental, C. Marshall, J. L. McGuire, E. L. Lindsey, K. C. Maguire, B. Mersey, and E. a Ferrer. 2011. Has the Earth's sixth mass extinction already arrived? Nature 471:51-57. Boyce, M. S., P. R. Vernier, S. E. Nielsen, and F. K. . Schmiegelow. 2002. Evaluating resource selection functions. Ecological Modelling 157:281-300. Burnham, K. P., and D. R. Anderson. 2002. Model selection and inference: a practical information-theoretic approach. Second ed. Springer-Verlag, New York, New York, USA. Calenge, C. 2006. The package "adehabitat" for the R software: a tool for the analysis of space and habitat use by animals. Ecological Modelling 197:516-519. Capitani, C., M. Chynoweth, J. Kusak, E. Çoban, and Ç. H. Şekercioğlu. 2016. Wolf diet in an agricultural landscape of north-eastern Turkey. Mammalia 80:329-334. Cardillo, M. 2005. Multiple causes of high extinction risk in large mammal species. Science 309:1239-1241. Carter, N. H., B. K. Shrestha, J. B. Karki, N. M. B. Pradhan, and J. Liu. 2012. Coexistence between wildlife and humans at fine spatial scales. Proceedings of the National Academy of Sciences of the United States of America 109:15360-5. Chetkiewicz, C. L. B., and M. S. Boyce. 2009. Use of resource selection functions to 90 identify conservation corridors. Journal of Applied Ecology 46:1036-1047. Chynoweth, M. W., E. Çoban, Ç. Altın, and Ç. H. Şekercioğlu. 2016. Human-wildlife conflict as a barrier to large carnivore management and conservation in Turkey. Turkish Journal of Zoology 40:972-983. Chynoweth, M. W., E. Çoban, and C. H. Sekercioglu. (in prep). Large carnivore hyperabundance in an empty forest. Ecology. Ciucci, P., L. Boitani, F. Francisci, and G. Andreoli. 1997. Home range, activity and movements of a wolf pack in central Italy. Journal of Zoology 243:803-819. Cozzi, G., M. Chynoweth, J. Kusak, E. Çoban, A. Çoban, A. Ozgul, and Ç. H. Şekercioğlu. 2016. Anthropogenic food resources foster the coexistence of distinct life history strategies: year-round sedentary and migratory brown bears. Journal of Zoology 300:142-150. Craighead, J. C., and J. Craighead F. C. 1972. Grizzly bear-man relationships in Yellowstone National Park. International Conference on Bear Research and Management 2:304-332. Dellinger, J. A., C. Proctor, T. D. Steury, M. J. Kelly, and M. R. Vaughan. 2013. Habitat selection of a large carnivore, the red wolf, in a human-altered landscape. Biological Conservation 157:324-330. Dirzo, R., H. S. Young, M. Galetti, G. Ceballos, N. J. B. Isaac, and B. Collen. 2014. Defaunation in the Anthropocene. Science 345:401-406. Gates, J. E., and L. W. Gysel. 1978. Avian nest dispersion and fledging success in fieldforest ecotones. Ecology 59:871-883. Haswell, P. M., J. Kusak, and M. W. Hayward. 2017. Large carnivore impacts are context-dependent. Food Webs 12:3-13. Hobbs, R. J., E. Higgs, C. M. Hall, P. Bridgewater, F. S. Chapin, E. C. Ellis, J. J. Ewel, L. M. Hallett, J. Harris, K. B. Hulvey, S. T. Jackson, P. L. Kennedy, C. Kueffer, L. Lach, T. C. Lantz, A. E. Lugo, J. Mascaro, S. D. Murphy, C. R. Nelson, M. P. Perring, D. M. Richardson, T. R. Seastedt, R. J. Standish, B. M. Starzomski, K. N. Suding, P. M. Tognetti, L. Yakob, and L. Yung. 2014. Managing the whole landscape: historical, hybrid, and novel ecosystems. Frontiers in Ecology and the Environment 12:557-564. Johnson, C. J., S. E. Nielsen, E. H. Merrill, T. L, M. S. Boyce, E. Science, M. Program, N. British, and P. George. 2006. Resource selection functions based on use - availability data : theoretical motivation and evaluation methods. Journal of Wildlife Management 70:347-357. 91 Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65-71. Kuijper, D. P. J., E. Sahlen, B. Elmhagen, S. Chamaille-Jammes, H. Sand, K. Lone, and J. P. G. M. Cromsigt. 2016. Paws without claws? Ecological effects of large carnivores in anthropogenic landscapes. Proceedings of the Royal Society B 283:1- 9. Lambeck, R. J. 1997. Focal Species: A multi-species umbrella for nature conservation. Conservation Biology 11:849-856. Latham, A. D. M., M. C. Latham, K. H. Knopff, M. Hebblewhite, and S. Boutin. 2013. Wolves, white-tailed deer, and beaver: implications of seasonal prey switching for woodland caribou declines. Ecography 36:1276-1290. Manly, B. F. J., L. L. Mcdonald, D. L. Thomas, T. L. Mcdonald, and W. P. Erickson. 2002. Resource selection by animals: statistical design and analysis for field studies. Second ed. Kluwer Academic Publishers, Dordrecht, Netherlands. Mazerolle, M., and M. Mazerolle. 2011. Package "AICcmodavg": model selection and multimodel inference based on (Q) AIC (c). Newsome, T. M., J. a. Dellinger, C. R. Pavey, W. J. Ripple, C. R. Shores, A. J. Wirsing, and C. R. Dickman. 2015. The ecological effects of providing resource subsidies to predators. Global Ecology and Biogeography 24:1-11. Oriol-Cotterill, A., D. W. Macdonald, M. Valeix, S. Ekwanga, and L. G. Frank. 2015. Spatiotemporal patterns of lion space use in a human-dominated landscape. Animal Behaviour 101:27-39. R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Robbins, C. T., C. C. Schwartz, and L. A. Felicetti. 2004. Nutritional ecology of ursids: a review of newer methods and management implications. Ursus 15:161-171. Rode, K. D., C. T. Robbins, and L. A. Shipley. 2001. Constraints on herbivory by grizzly bears. Oecologia 128:62-71. Rogers, L. L., D. W. Kuehn, A. W. Erickson, E. M. Harger, L. J. Verme, and J. J. Ozoga. 2014. Characteristics and management of black bears that feed in garbage dumps, campgrounds or residential areas. Bears: Their Biology and Management, Vol. 3, A Selection of Papers from the Third International Conference on Bear Research and Management, Binghamton, New York, USA. and Moscow, U.S.S.R., IUCN Publications New Series 40:169-175. 92 Schlaepfer, M. A., M. C. Runge, and P. W. Sherman. 2002. Ecological and evolutionary traps. Trends in Ecology & Evolution 17:474-480. Sekercioglu, C. H., S. Anderson, E. Akçay, and R. Bilgin. 2011. Turkey's rich natural heritage under assault. Science 334:1637-1638. Şekercioğlu, Ç. H., S. Anderson, E. Akçay, R. Bilgin, Ö. E. Can, G. Semiz, Ç. Tavşanoğlu, M. B. Yokeş, A. Soyumert, K. İpekdal, İ. K. Sağlam, M. Yücel, and H. Nüzhet Dalfes. 2011. Turkey's globally important biodiversity in crisis. Biological Conservation 144:2752-2769. Zalasiewicz, J., M. Williams, A. Smith, T. L. Barry, A. L. Coe, P. R. Bown, P. Brenchley, D. Cantrill, A. Gale, P. Gibbard, F. J. Gregory, M. W. Hounslow, A. C. Kerr, P. Pearson, R. Knox, J. Powell, C. Waters, J. Marshall, M. Oates, P. Rawson, and P. Stone. 2008. Are we now living in the Anthropocene? GSA Today 18:4. 93 Figure 4.1. Location of Sarıkamış-Allahuekber Mountains National Park and surrounding forest in eastern Turkey. 94 Figure 4.2. Seasonal resource selection by bears in Sarıkamış forest, eastern Turkey. See Table 4.1 for descriptions of variables. 95 Figure 4.3. Seasonal resource selection by wolves in Sarıkamış forest, eastern Turkey. See Table 4.1 for descriptions of variables. 96 Figure 4.4. Seasonal resource selection by lynx in Sarıkamış forest, eastern Turkey. See Table 4.1 for descriptions of variables. 97 Figure 4.5. Predicted probability of bear occurrence in Sarıkamış forest, eastern Turkey during the spring season, the summer season, and the autumn season. 98 Figure 4.6. Predicted probability of wolf occurrence in Sarıkamış forest, eastern Turkey during the winter season and the summer season. 99 Figure 4.7. Predicted probability of lynx occurrence in Sarıkamış forest, eastern Turkey during the winter season and the summer season. 100 Figure 4.8. Prioritized area for future conservation efforts in Sarıkamış forest, eastern Turkey based on resource selection of brown bears, gray wolves, and Eurasian lynx. 101 Table 4.1. Description and characteristics of environmental variables used to model the probability of occurrence of bears, wolves, and lynx in Sarıkamış forest, eastern Turkey. Group Variable Name Abbrev. Resolution (m) Units Environmental Distance to Forest DFOR 30x30 Meters Elevation ELEV 30x30 Meters Slope SLP 30x30 Degrees Aspect ASP 30x30 Degrees Human Distance to Road DROAD 30x30 Meters Distance to Village DVILL 30x30 Meters 102 Table 4.2. Home range sizes for all bears, wolves and lynx captured in Sarıkamış forest, eastern Turkey from 2011-2014. Number 95% Kernel 95% MCP Species Animal_ID Sex Age of Home Range Home Range Locations (km2) (km2) 7.5 630 581 Wolf 2011_09435 M 2112 2011_09436 M 2.5 5938 4732 1446 2013_06823 F 4 1606 43 42 2013_09435 M 5-6 546 1783 666 2013_09436 F 5-6 153 674 125 2014_09435 M 2.5 762 2042 797 2014_09436 M 2.5 769 1632 1592 4 Lynx 2014_13148 M 1390 1966 1902 2014_13151 M 7 1574 132 108 29 Bear 2012_07083 F 12-14 3359 27 324 2012_07949 F 7 745 210 194 2012_11685 M 8 9134 208 163 2012_11686 M 6 1576 108 141 2012_11687 M 6-7 8089 148 43 2013_06089 F 5 5195 38 24 2013_06090 F 10-12 7338 31 117 2013_12262 M >15 7824 137 130 2013_12263 M 4 3053 108 234 2014_11685 M 10 6497 232 31 2014_11686 F 5 3737 31 86 2014_15425 M 8-9 7571 84 120 2014_15426 M 8-10 9440 155 14 2014_15427 F 9-10 16497 17 34 2014_15428 M 6 3635 72 77 2014_15429 M 5-6 3714 142 CHAPTER 5 MINIMUM POPULATION SIZE AND GENETIC DIVERSITY OF BROWN BEARS WITHIN A FRAGMENTED POPULATION IN EASTERN TURKEY 5.1 Abstract Estimating population size of wide-ranging and elusive carnivores is a major challenge for wildlife biologists and conservationists, yet it is also one of the most important population parameter estimates needed to guide management of wildlife. We evaluated a noninvasive method of capture-recapture for Eurasian brown bear (Ursus arctos arctos) density estimation using DNA extracted from scat samples in SarıkamışAllahuekber Mountains National Park in eastern Turkey. This is a highly degraded forest that harbors a bear population that largely relies on anthropogenic food sources. From 2013-2015, we used scat detection dogs from the University of Washington's Center for Conservation Biology's Conservation Canines program to collect scat samples within a sampling grid designed to produce data for spatial capture-recapture modeling. We collected 1,520 bear scat samples across all years and after extracting DNA from 595 samples from the 2013 field season, we identified 157 viable bear samples to genotype using 8 polymorphic microsatellite loci. Logistic constraints were a limiting factor in our ability to generate enough data for capture-recapture analysis; therefore we focused on 104 generating a minimum population estimate in the main study area. Taking a multilocus genotyping approach, our results identified 27 unique multilocus genotypes, which suggests a minimum population size of 27 bears. 5.2 Introduction The growing field of conservation genetics has become a keystone approach in wildlife research. Genetic variability of a species and its populations is an important component for evaluating long-term survival and consequent management approaches (Lacy 1997). Knowledge on genetic processes impacting populations can guide management decisions aimed at conserving or restoring genetic diversity, which is integral to the persistence of populations (Lande and Shannon 1996). Information on connectivity between populations is crucial to counteract the effects of genetic drift, which is particularly important for small or isolated populations (Schwartz et al. 2007). A standard approach in wildlife genetic monitoring is to use short tandem repeats (STRs) also known as microsatellites. STRs are short stretches of DNA made up of core repeats of two to seven nucleotides in noncoding regions of the genome (Allendorf and Luikart 2009). The combination of high polymorphism and neutrality to selection creates an ideal scenario to identify individuals. The use of STRs in wildlife management has become a common and increasingly sophisticated approach to monitor genetic diversity and gene flow in populations (Schwartz et al. 2007). Using STRs, the minimum population size can be determined (number of unique individuals), and mark-recapture analysis can be conducted to determine effective population size (Skrbinšek et al. 2012). The latter is a crucial parameter that is difficult to 105 determine without molecular genetic information in populations of wide-ranging and elusive carnivores. The minimum population size is critical for the sustainable management of wild populations because it may allow for deductions to be made about the current status and viability of a population. STRs can also be used to examine genetic diversity, which is influenced by such parameters as population size, the amount of gene flow to and from other populations and selection over time (Frankham 1996). Populations of wild animals are naturally structured or divided into separate groups where random mating occurs. Large carnivores, in particular, demonstrate spatial division due to isolation by distance and species-specific characteristics (e.g., social behavior and movement; Lowe & Allendorf 2010). Natural barriers such as topographic or water features can limit genetic connectivity; however, anthropogenic barriers such as roads, human activity, and habitat destruction can also cause population fragmentation (Geffen et al. 2004, Proctor et al. 2012). For this reason, knowledge of genetic connectivity can help mitigate human impact on large carnivore populations and guide management actions to increase genetic diversity and ultimately achieve management goals. The aim of this study was to estimate the minimum population size of Eurasian brown bears in a presumably small and isolated subpopulation in northeastern Turkey. In this understudied region, the current population of bears and other wildlife species is unknown. However, previous work suggests a hyperabundance of large carnivores in the system that survives largely on anthropogenic food resources (Chynoweth et al. n.d., Capitani et al. 2016, Cozzi et al. 2016). Based on brown bear densities in forest habitats in Romania (Kalaber et al. 1994) and Georgia (Lexo Gavashelishvili, personal communication), we predict our study area (~550 km2) to have 50-80 brown bears, 106 although comparably drier forest conditions in Sarıkamıs makes it likely that the natural sustainable population is substantially lower. 5.3 Methods 5.3.1 Study Area (from section 3.3.1) Our study was carried out on the Kars-Ardahan high plateau in northeastern Turkey, at the intersection of Caucasus and Irano-Anatolian global biodiversity hotspots. The area (c. 550 km2; 40°20'N 42°35'E) ranges between 1900 and 3120 m asl and is composed of fragmented forest in a matrix of agricultural and rangelands. The city of Sarıkamış (population: c. 18,000) is located in the center of the study area and on one of the two paved roads that bisect the forested area. Forest cover consists almost exclusively of Scots pine (Pinus sylvestris Linnaeus, 1753), while understory vegetation is scarce, with consequent scarcity of food resources for browsers. Sarıkamış-Allahuekber Mountains National Park (hereafter SAMNP) boundaries cover a total area of 225.1 sq. km., but only include 49.69 sq. km. of forest. Therefore, SAMNP is only comprised of 22.07% forest cover. Total forest cover in the region includes 328.38 sq. km. including a large expanse of forest south of the national park (248.15 sq. km.). These patches of forest represent the southernmost significant forest patch in the region extending south from the extensive forests in the Black Sea Region of Turkey. Human activity in the forest is extensive in both time and space, limited only by harsh winter temperatures, and consists primarily of livestock grazing, harvest of forest products (e.g., fruits, pine cones, mushrooms), and legal and illegal timber extraction. Livestock is abundant in the region with cattle (Bos taurus Linnaeus, 1758), sheep (Ovis aries Linnaeus, 1758), and goats (Capra hircus Linnaeus, 1758) freely roaming 107 rangelands from April to November (Capitani et al. 2016). About 851,445 livestock heads have been registered in the Kars province in 2012 (Ministry of Food, Agriculture and Livestock, Republic of Turkey). A notable feature on the landscape is an unfenced municipal garbage dump 3 km west of Sarıkamıs city. The dump represents a predictable anthropogenic food source, and bears, wolves, and wild boar visit the dump regularly at night (pers. obs.). A portion of the bear population has altered life history strategies to regularly use the dump, while other bears never visit the dump (Cozzi et al. 2016). 5.3.2 Survey Methods Scat samples were collected using scat detection dogs trained at the University of Washington's Center for Conservation Biology. These highly specialized dogs are trained to identify particular species and have proven to be remarkably efficient at scat collection (Wasser et al. 2004, 2011, Ayres et al. 2012). A sampling grid designed to produce data appropriate for spatial capture-recapture modeling was used to collect samples in a systematic method (Royle, Chandler, Sollmann, & Gardner, 2014; Figure 5.1). Each 30x30 km grid (black squares) contains fifteen 2x2 km sampling grids. The black grids represent a sampling area that is equivalent to the average home range size of our target species (based on preliminary data from our GPS collar work; see Chapter 4 of this dissertation). Three 2x2 grids within a single black grid were sampled per day, and each black grid was sampled 5 times in a rotating pattern. Overall, each 2x2 square was visited once to collect scat samples. The field team conducted all sampling within a 2-month period to satisfy the assumption of a closed population. 108 5.3.3 Lab Methods Scat samples were collected in plastic bags and frozen (-20 °C) immediately subsequent to collection in the field in Sarıkamıs. Frozen samples were shipped to Boğaziçi University (Istanbul, Turkey) for DNA extraction (Qiagen QIAamp DNA Stool Mini Kit; see also Jackson et al., 2008) following manufacturers' instructions. All samples from 2013 were run at least once (n=595), and samples with unsuccessful extractions are currently being processed a second time for DNA extraction. Extracted DNA samples were shipped to the University of Utah for further analysis. On arrival, samples were amplified by polymerase chain reaction (PCR) using bear specific primer pair G10P (Paetkau and Strobeck 1998) and run on a 2% agarose gel to confirm successful extraction of bear DNA from scat samples. Since extracting DNA from scat samples of various ages and qualities can be challenging and variable (Vynne et al. 2012), identifying nonviable samples early helped limit overall cost of microsatellite amplification and sequencing. Samples containing bear DNA were amplified in a 10 μl reaction containing 2x Qiagen Master Mix (Qiagen Multiplex PCR Kit, Qiagen, USA), 1.3 μM of forward primer, 5.3 μM of reverse primer, 3.3 μM of fluorescent tag, 1 μl of Q mix and 3 μl of template DNA. Cycling conditions were as follows: 95 °C for 15 minutes, 32 cycles of 30 seconds at 94 °C, 90 seconds at 58 °C, 1 minute at 72 °C, and 10 minute final extension at 72 °C. Fluorescent tags NED, PET, 6FAM, VIC were used to enable allele reads on a capillary sequencer. Samples were genotyped using eight polymorphic microsatellite loci which were shown to contain enough variation to identify individual bears across all eight ursid 109 species (Paetkau and Strobeck 1994, 1998): G1A, G10B, G10C, G1D, G10L, G10M, G10P, and G10X. All loci were amplified in three multiplex PCR reactions using Qiagen Multiplex PCR Kit, run on an Applied Biosystems 3730xl capillary sequencer (University of Utah Cores Research Facility), and analyzed with PeakScanner Software. Peaks were read at the tetra level to account for low peak reading on the capillary sequencer and reduce human error. Genotyping errors can result in positive bias if samples from the same individual are assigned different genotypes (Woods et al. 1999, Mills et al. 2000). Conversely, if the markers are not sufficiently variable, too few individuals will be identified resulting in a negative bias. We used a combination of objective (peak height) and subjective (appearance) criteria to quantify fragment size (i.e., length) as genotype scores, and classified microsatellites as tetra-STRs to account for low peaks in many of our samples. 5.3.4 Multilocus Genotype Analysis To estimate the minimum population size for bears, we used multilocus genotyping (MLG). Missing data due to unsuccessful PCRs were identified as an issue early in the process, and we therefore only used loci with 30% or less missing data. We first called MLG by a naïve method, whereby coding missing information as 0 and treating it as another state. This resulted in 122 unique MLGs; however, because of the high degree of uncertainty that this method brought, we tried to resolve this problem by quantify the genetic distance between each individual and collapsing individuals under a defined threshold. To this end, we built a distance matrix between each individual using Nei's genetic distance (Nei 1978), which calculates distance between individuals based on the 110 arithmetic mean of gene identity (the probability of two individuals having the same allele at a particular locus summed over all loci; Nei 1978). To identify the minimum threshold separating two individuals, we used the above distance matrix to build a hierarchical cluster tree based on the nearest and farthest neighbor approaches (also known as single linkage and complete linkage clustering respectively, Legendre and Legendre 1998). In both trees we found the average minimum distance between two individuals to be close to 0.02 (nearest neighbor: 0.02083333, farthest neighbor: 0.01785714). We therefore used 0.02 as a cutoff to collapse individuals. To investigate the effect of our threshold, we also collapsed individuals by an extremely small genetic distance (1.5-8) and did not see different results. When we went over the threshold of 0.02, we started to reduce the number of MLGs. Therefore we believe our approach to collapsing MLGs is appropriate. 5.4 Results Between 2013-2015, we encountered 5,125 scat samples from nine known mammal species, from which we collected 1,520 bear scat samples (Table 5.1). Out of the 595 bear samples collected in 2013 we were able successfully extract DNA from 157 samples. Of these 157 samples, we excluded samples outside the main study area (Figure 5.1) and those that did not have information from at least 5 microsatellite loci, resulting in a final sample size of 126. We experienced varying degrees of success amplifying STRs for the 8 microsatellite loci and had to remove 1 locus (G10M) from further analysis due to high amount of missing data (Figure 5.2). A genotype accumulation curve for 7 microsatellite loci suggests that 5-6 loci provide 111 enough resolution to identify unique MLGs (Figure 5.4). Across 7 loci (mean Hexp: 0.701355176; mean Evenness: 0.819339375; see Table 5.2) we identified 27 MLGs. Based on these 27 MLGs, genotypic diversity according to Shannon's Index was 3.121751, Simpson's Index was 0.947972, and expected genotypic heterozygosity was 0.701355176. Our recapture history for individual MLGs suggests a robust data set to move forward with spatial capture-recapture analysis once more samples are genotyped (Figure 5.5). 5.5 Discussion Our results are the first attempt to estimate minimum population size of brown bears in a region of Turkey using molecular fingerprinting, and as such we encountered some challenging conditions that delayed tissue processing and shipping of DNA samples that significantly slowed progress. As a result, we were forced to work with a smaller number of samples for estimating minimum population size of brown bears in Sarıkamıs forest than initially intended. Therefore, our minimum population estimate is more than likely an underestimate of the true population size, and by including more samples in the future we aim to generate a more accurate estimate of population size. To deal with uncertainties in MLG calls caused by the high amount of missing data we treated individuals that had a genetic distance under 0.02 (according to Nei's distance, Nei 1972) as one. This is a somewhat conservative approach as we only collapsed individuals that were extremely close genetically. However, this method does have the risk of collapsing siblings since MLG profiles of siblings have the risk of being extremely close due to the high amount of missing data. For this reason, and due to samples that did 112 not provide viable bear DNA, our minimum population estimate is most likely lower than the true minimum population. A solution to this problem is to rerun viable DNA samples to fill in data gaps (see Figure 5.2), and also to continue DNA extraction for samples that to date have not provided viable bear DNA. Bases on our results, the bear population in Sarıkamıs forest appears to be a relatively diverse population. By comparing nuclear genetic diversity from other studies based on expected heterozygosity, Sarıkamıs bears are similar to Dinaric and Scandanavian bear populations, more genetically diverse than isolated populations such as in the Italian Apennines and Pyrenees, and slightly less diverse than areas with good conservation status, such as the Carpathian Mountains in Romania (Kocijan et al. 2011). The southernmost patch of forest in the region, Sarıkamıs is a somewhat isolated, but bears have been documented migrating between forest patches (Cozzi et al. 2016), validating results presented here that suggest high genetic diversity. However, ongoing development in the region threatens this connectivity and should be seen as a real threat (Şekercioğlu et al. 2011, Ambarlı et al. 2016). Given our limitations, we were able to generate a robust minimum population estimate of 27 unique MLGs in our study area. This estimate corroborates previous observations in the field, based on ongoing bear capture efforts for GPS collar deployment (see Chapter 4: 25 unique individuals) and observations of bears at the municipal garbage dump (~33 individuals seen in one night, author's pers observation). A minimum population estimate allows us to speculate that the minimum local population density of bears in Sarıkamıs Forest is above 8.23 bears/100 km2. This estimate is similar to values generated from other studies ranging from 2.73 - 13 bears 113 per 100 km2 (Revenko 1995, Miller et al. 1997, Bellemain et al. 2005, Walsh et al. 2010, Morton et al. 2016). However, our estimate of minimum local bear density is still above average, and therefore, we speculate that true bear density in Sarıkamıs is significantly higher than in most regions within the global distribution of brown bears. It is unlikely that bear density in Sarıkamıs surpasses the highest reported local bear density in the Dinaric Mountains of Slovenia at 40 bears/100 km2, which is a result of a long cultural history of supplemental feeding and hunting (Jerina et al. 2013). 5.6 Conclusions Genetic sampling using scat samples was effective in estimating a minimum population size for Eurasian brown bears in Sarıkamıs forest. These preliminary results indicate that this method can be an important tool for carnivore management in this understudied system largely due to the ease of collecting fecal samples compared to the cost of other methods such as large carnivore capture. Moving forward, successful genotyping of more samples will allow for spatially explicit capture-recapture models to be used to estimate effective population size of brown bears, gray wolves, and Eurasian lynx in the study area (Chandler and Andrew Royle 2013). In addition, samples could continuously be collected in the study system to examine long-term trends in survival and fecundity. Genetic sampling is increasingly becoming a reliable alternative to other methods to estimate large carnivore populations and should be considered as a principal approach for studying large carnivore populations in Sarıkamıs forest in the future. 114 5.7 References Allendorf, F. W., and G. Luikart. 2009. Conservation and the genetics of populations. Blackwell Publishing, Oxford, United Kingdom. Ambarlı, D., U. S. Zeydanlı, Ö. Balkız, S. Aslan, E. Karaçetin, M. Sözen, Ç. Ilgaz, A. Gürsoy Ergen, Y. Lise, S. Demirbaş Çağlayan, H. J. Welch, G. Welch, A. S. Turak, C. C. Bilgin, A. Özkil, and M. Vural. 2016. An overview of biodiversity and conservation status of steppes of the Anatolian biogeographical region. Biodiversity and Conservation 25:2491-2519. Ayres, K. L., R. K. Booth, J. a Hempelmann, K. L. Koski, C. K. Emmons, R. W. Baird, K. Balcomb-Bartok, M. B. Hanson, M. J. Ford, and S. K. Wasser. 2012. Distinguishing the impacts of inadequate prey and vessel traffic on an endangered killer whale (Orcinus orca) population. PLoS One 7:e36842. Bellemain, E., J. E. Swenson, D. Tallmon, S. Brunberg, and P. Taberlet. 2005. Estimating population size of elusive animals with DNA from hunter-collected feces: four methods for brown bears. Conservation Biology 19:150-161. Capitani, C., M. Chynoweth, J. Kusak, E. Çoban, and Ç. H. Şekercioğlu. 2016. Wolf diet in an agricultural landscape of north-eastern Turkey. Mammalia 80:329-334. Chandler, R. B., and J. Andrew Royle. 2013. Spatially explicit models for inference about density in unmarked or partially marked populations. Annals of Applied Statistics 7:936-954. Chynoweth, M. W., E. Çoban, and C. H. Sekercioglu. (in Review). Large carnivore hyperabundance in an empty forest. Ecology. Cozzi, G., M. Chynoweth, J. Kusak, E. Çoban, A. Çoban, A. Ozgul, and Ç. H. Şekercioğlu. 2016. Anthropogenic food resources foster the coexistence of distinct life history strategies: year-round sedentary and migratory brown bears. Journal of Zoology 300:142-150. Frankham, R. 1996. Relationship of genetic variation to population size in wildlife. Conservation Biology 10:1500-1508. Geffen, E., M. J. Anderson, and R. K. Wayne. 2004. Climate and habitat barriers to dispersal in the highly mobile grey wolf. Molecular Ecology 13:2481-2490. Jackson, J. V., S. L. Talbot, and S. Farley. 2008. Genetic characterization of Kenai brown bears (Ursus arctos): microsatellite and mitochondrial DNA control region variation in brown bears of the Kenai Peninsula, south central Alaska. Canadian Journal of Zoology 86:756-764. 115 Jerina, K., M. Jonozovič, M. Krofel, and T. Skrbinšek. 2013. Range and local population densities of brown bear Ursus arctos in Slovenia. European Journal of Wildlife Research 59:459-467. Kalaber, L., A. Negrutiu, I. Micu, A. Dogaru, and M. Brateanu. 1994. Distribution and ecology of brown bear in Romania. International Conference on Bear Research and Management 9:173-178. Kocijan, I., A. Galov, H. Ćetković, J. Kusak, T. Gomerčić, and Đ. Huber. 2011. Genetic diversity of Dinaric brown bears (Ursus arctos) in Croatia with implications for bear conservation in Europe. Mammalian Biology - Zeitschrift für Säugetierkunde 76:615-621. Lacy, R. C. 1997. Importance of genetic variation to the viability of mammalian populations. Journal of Mammalogy 78:320-335. Lande, R., and S. Shannon. 1996. The role of genetic variation in adaptation and population persistence in a changing environment. Evolution 50:434-437. Legendre, P., and L. Legendre. 1998. Numerical ecology. Second ed. Elsevier, Amsterdam, The Netherlands. Lowe, W. H., and F. W. Allendorf. 2010. What can genetics tell us about population connectivity? Molecular Ecology 19:3038-3051. Miller, S. D., G. C. White, R. A. Sellers, H. V Reynolds, W. Schoen, K. Titus, V. G. Barnes, R. B. Smith, R. R. Nelson, W. B. Ballard, and C. C. Schwartz. 1997. Brown and black bear density estimation in Alaska using radiotelemetry and replicated mark-resight techniques. Wildlife Monographs 133:3-55. Mills, L. S., J. J. Citta, K. P. Lair, M. K. Schwartz, and D. A. Tallmon. 2000. Estimating animal abundance using noninvasive DNA sampling: promise and pitfalls. Ecological Applications 10:283-294. Morton, J. M., G. C. White, G. D. Hayward, D. Paetkau, and M. P. Bray. 2016. Estimation of the brown bear population on the Kenai Peninsula, Alaska. Journal of Wildlife Management 80:332-346. Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-590. Paetkau, D., and C. Strobeck. 1994. Microsatellite analysis of genetic variation in black bear populations. Molecular Ecology 3:489-495. Paetkau, D., and C. Strobeck. 1998. Ecological genetic studies of bears using microsatellite analysis. Ursus 10:299-306. 116 Proctor, M. F., D. Paetkau, B. N. McLellan, G. B. Stenhouse, K. C. Kendall, R. D. MacE, W. F. Kasworm, C. Servheen, C. L. Lausen, M. L. Gibeau, W. L. Wakkinen, M. a. Haroldson, G. Mowat, C. D. Apps, L. M. Ciarniello, R. M. R. Barclay, M. S. Boyce, C. C. Schwartz, and C. Strobeck. 2012. Population fragmentation and interecosystem movements of grizzly bears in Western Canada and the Northern United States. Wildlife Monographs 180:1-46. Revenko, I. A. 1995. Status of brown bears in Kamchatka, Russian Far East. Ursus 10:11-16. Royle, J. A., R. B. Chandler, R. Sollmann, and B. Gardner. 2014. Spatial capturerecapture. Elsevier, Watham, MA. Schwartz, M., G. Luikart, and R. Waples. 2007. Genetic monitoring as a promising tool for conservation and management. Trends in Ecology & Evolution 22:25-33. Şekercioğlu, Ç. H., S. Anderson, E. Akçay, R. Bilgin, Ö. E. Can, G. Semiz, Ç. Tavşanoğlu, M. B. Yokeş, A. Soyumert, K. İpekdal, İ. K. Sağlam, M. Yücel, and H. Nüzhet Dalfes. 2011. Turkey's globally important biodiversity in crisis. Biological Conservation 144:2752-2769. Skrbinšek, T., M. Jelenčič, L. Waits, I. Kos, K. Jerina, and P. Trontelj. 2012. Monitoring the effective population size of a brown bear (Ursus arctos) population using new single-sample approaches. Molecular Ecology 21:862-75. Vynne, C., M. R. Baker, Z. K. Breuer, and S. K. Wasser. 2012. Factors influencing degradation of DNA and hormones in maned wolf scat. Animal Conservation 15:184-194. Walsh, P., J. Reynolds, G. Collins, B. Russell, M. Winfree, and J. Denton. 2010. Application of a double-observer aerial line-transect method to estimate brown bear population density in southwestern Alaska. Journal of Fish and Wildlife Management 1:47-58. Wasser, S. K., B. Davenport, E. R. Ramage, K. E. Hunt, M. Parker, C. Clarke, and G. Stenhouse. 2004. Scat detection dogs in wildlife research and management: application to grizzly and black bears in the Yellowhead Ecosystem , Alberta , Canada. Canadian Journal of Zoology 82:475-492. Wasser, S. K., J. L. Keim, M. L. Taper, and S. R. Lele. 2011. The influences of wolf predation, habitat loss, and human activity on caribou and moose in the Alberta oil sands. Frontiers in Ecology and the Environment 9:546-551. Woods, J. G., D. Paetkau, D. Lewis, B. N. McLellan, M. Proctor, and C. Strobeck. 1999. Genetic tagging of free-ranging black and brown bears. Wildlife Society Bulletin 27:616-627. 117 Figure 5.1. Black sampling grids (30x30km) are based on home size of target species and will be resampled 5 times to produce data appropriate for capture-recapture analysis. SAM NP, surrounding forests, and the entire proposed wildlife corridor region will be sampled, providing a holistic picture of large carnivore genetics in the region. 118 Percent missing data per locus and population of Data 1 23.81 % 46.83 % 15.08 % 23.81 % 27.78 % 11.11 % 19.84 % 20.63 % 23.61 % Percent Missing Population 40 30 20 10 20.63 % 23.61 % 0 n ea 19.84 % M 11.11 % 1D G B 10 G 27.78 % L 10 G M 10 G 23.81 % P 10 G 15.08 % X 10 G 46.83 % 1A G 23.81 % C 10 G Total Locus Figure 5.2. Percent missing data per locus for all 8 loci originally used for genotyping bears. Loci G10C was identified as having too much missing data to include in further genotyping. 119 Percent missing data per locus and population of Data Population Percent Missing 1 23.81 % 15.08 % 23.81 % 27.78 % 11.11 % 19.84 % 20.63 % 20.29 % 20 10 Total 23.81 % 15.08 % 23.81 % 27.78 % 11.11 % 19.84 % 20.63 % 20.29 % 0 n ea 1D L 10 P 10 X 10 1A B 10 C 10 M G G G G G G G Locus Figure 5.3. Percent missing data per locus for the 7 loci used for genotyping bears, after loci G10M was removed from analysis. 120 Genotype accumulation curve for Data Number of multilocus genotypes 122 120 100% 100 80 60 40 20 0 1 2 3 4 5 6 Number of loci sampled Figure 5.4. Genotype accumulation curve for 7 microsatellite loci used to genotype bears. MLG Figure 5.5. Multilocus genotype (MLG) recaptures for 27 MLGs identified with 7 microsatellite loci for bears. MLG.29 MLG.50 MLG.54 MLG.94 MLG.104 MLG.113 MLG.121 MLG.2 MLG.68 MLG.75 MLG.96 MLG.100 MLG.115 14 MLG.5 MLG.53 MLG.67 MLG.11 MLG.12 MLG.33 MLG.69 MLG.21 MLG.32 MLG.47 MLG.45 MLG.83 MLG.85 MLG.7 count 121 Data: Data N = 126 MLG = 27 1 12 10 8 6 4 2 122 Table 5.1. Summary of all scat samples encountered in a survey conducted from 20132015 in Sarıkamış, eastern Turkey using trained scat-detection dogs. 2013 2014 2015* Species Encounter Collect Encounter Collect Encounter Collect Bear 1379 595 1311 783 187 142 Wolf 326 139 394 326 75 62 Lynx 93 51 141 121 14 13 Wild_Boar 158 78 395 49 31 28 Badger 53 37 56 23 23 19 Marten 86 44 180 47 56 47 Mustelid 0 0 0 0 13 10 Roe deer 3 3 25 17 29 21 Hedgehog 2 0 0 0 0 0 Wildcat 3 3 9 9 0 0 null 44 39 30 28 0 0 blank 0 0 0 0 8 8 Total 2147 989 2542 1404 436 350 *During our 2015 field season, we were arrested during field work and could not continue data collection 123 Table 5.2. Microsatellite marker variability, expected heterozygosity and observed number of alleles of brown bears in Sarıkamış forest, eastern Turkey, 2013. Locus Na 1-Db Hexpc Evennessd G10C 7 0.7734375 0.777486911 0.832836756 G10B 6 0.77198882 0.775613181 0.862228567 G1A 4 0.559082031 0.562009162 0.793323241 G10X 5 0.518294892 0.5211584 0.832395194 G10P 7 0.760044643 0.763452915 0.767265801 G10L 8 0.788256053 0.792177725 0.859223055 G1D 5 0.714 0.71758794 0.788103012 Mean 6 0.697871991 0.701355176 0.819339375 a no. of observed alleles Simpson's Index c observed heterozygosity d evenness b CHAPTER 6 HUMAN-WILDLIFE CONFLICT AS A BARRIER TO LARGE CARNIVORE MANAGEMENT AND CONSERVATION IN TURKEY Chynoweth, M. W., E. Çoban, Ç. Altın, and Ç. H. Şekercioğlu. 2016. Human-wildlife conflict as a barrier to large carnivore management and conservation in Turkey. Turkish Journal of Zoology 40:972-983. Reprinted with permission from the Scientific and Technological Research Council of Turkey. 125 126 127 128 129 130 131 132 133 134 135 136 CHAPTER 7 RECOMMENDATIONS FOR FUTURE CONSERVATION EFFORTS IN HUMAN-DOMINATED LANDSCAPES: THE CASE STUDY OF SARİKAMİŞ 7.1 Abstract Large carnivores are cryptic and opportunistic species that can inhabit humanmodified landscapes at medium-to-high densities without frequent detection. Over the last several decades, several species are rapidly recolonizing large areas in Europe and North America. To properly manage these species in an increasingly dynamic and human-dominated landscape, novel approaches to wildlife conservation and management must be developed. Much research on this issue has been done in developed nations, but more attention needs to be given to vast areas of the developing world that represent hotbeds of global biodiversity. In eastern Anatolia, Sarıkamış-Allahuekber Mountains National Park and surrounding forests is an example of an agrarian landscape dominated by human activity resulting in heavily degraded wildlife habitat. This system may be representative of the majority of existing and potential large carnivore habitat in the world; a long history of human activity has left these areas in a hybrid ecosystem state, one defined by anthropogenic food sources, heavily degraded habitat, and a deficit of wild ungulates and other natural prey species. The condition in eastern Anatolia now sustains a hyperabundance of synanthropic carnivores with no existing plans to monitor 138 or manage these populations. As a solution, a local environmental nonprofit group-the KuzeyDoğa Society-can provide conservation science to inform conservation and management plans in the region. Environmental organizations such as this are often leading conservation efforts in biodiversity hotspots without long-term funding and receiving little support from government entities. Solutions to large carnivore management in human-dominated landscapes that we provide here can be applied to regions experiencing similar conditions around the world. 7.2 Introduction The environmental landscape of today is a mosaic of novel ecosystems in a highly altered matrix (Hobbs et al. 2014). Large tracts of intact wildlife habitat are rare; the majority of the earth's surface has been substantially altered, initiating a modern epoch designated as the Anthropocene (Zalasiewicz et al. 2008, Barnosky et al. 2012, Waters et al. 2016). In order to accomplish global and regional conservation goals, biologists and conservationists are beginning to recognize that biodiversity must be studied and managed in human-dominated landscapes (Martin et al. 2014). The next step is to further develop the tools needed to manage individual species and biodiversity as a whole in these landscapes, which means our traditional view of ecosystem structure and species interactions must also change. Large carnivores represent an ideal focal species to investigate this paradigm shift in conservation and management over time. These species' strong ecological effects through trophic cascades are often mentioned, but the effects of top predators are not wellunderstood and are under considerable empirical and theoretical scrutiny (Allen et al. 139 2017). Effects of large carnivores may not be experienced outside of vast expanses of protected areas that are mostly free of human influence (Ripple et al. 2013, Allen et al. 2017). Historically, these animals were heavily persecuted and eradicated throughout much of their range. Currently, they are recolonizing many human-dominated landscapes. Subsequently, as future recolonization and reintroduction occur, we believe populations will become more synanthropic than ever before, and some large carnivore species will help shape novel ecosystem structure and function. In this paper, we briefly discuss the past and current situation of large carnivore management and conservation and provide a description of potential solutions to the issues surrounding both. As a case study, we focus on the Sarıkamış-Allahuekber Mountains National Park in eastern Turkey, where a comprehensive 4-year large carnivore study has revealed a seemingly novel mammal community structure and synanthropic populations of brown bears, gray wolves, and Eurasian lynx (Chynoweth et al. in prep; Capitani et al. 2016; Cozzi et al. 2016). However, we believe that much of the global range of these species intersects with similar human-dominated agrarian landscapes. The effects of large carnivores on food webs in these human-dominated landscapes is understudied and management suggestions outlined here can be applied to many systems across the globe. 7.2.1 Past Large Carnivore Management Paradigms Early human-predator relationships were largely shaped by the overall relationship between people and their environment. Hunters and shepherds often had a desire to eradicate large carnivores because of competition for prey, threats to livelihood (e.g., 140 livestock), or human well-being. In Europe, significant declines of bears, wolves and lynx began in the Middle Ages and continued into the 20th century, and eradication was directly caused by persecution and the indirect result of habitat loss. When Europeans colonized North America, they continued this pattern to remove large carnivores. As technology improved and became more readily available, humans were able to lead vast campaigns of large carnivore destruction through government-funded campaigns, using firearms and poison. The results of this paradigm are best observed in North America, where vast wilderness areas (occupied at low densities by Native Americans) were settled in a relatively short time period by Europeans. First in Europe, then in North America, wildlife management programs moved towards a more science-based approach, taking into account full ecosystems. Consequently, such programs slowly began to include large carnivore species. However, far from being uniform, wildlife management took the form of region-specific and species-specific management plans. Whether these wildlife management plans have positive or negative impacts on carnivore populations depends on the range of stakeholders' experiences and entrenched views on the impact of large carnivores. Over the last 50 years, attitudes have gradually changed, generally moving away from supporting predator extermination and toward supporting predator conservation. In some regions, these societal changes have facilitated megafauna comebacks. The historically intensive removal of predators was accompanied by broad-scale modification of landscapes including habitat destruction, removal of other plant and animal species, and disruption of many ecosystem functions. Therefore, the effect of large carnivore removal on ecosystems is somewhat unclear. It is a complex question for 141 ecologists, and the paucity of data leaves many of these questions unanswered or with answers that are controversial (Wallach et al. 2015, Ripple et al. 2016, Allen et al. 2017). Many examples from relatively pristine systems, most famously the Greater Yellowstone Ecosystem, suggest that large carnivore removal can result in over abundance of herbivores, subsequent loss of vegetation and loss of critical ecosystem function (Ripple and Beschta 2007). However, large carnivores are inherently difficult to study with a rigorous experimental design given their natural rarity, long generation time, and wide ranges. These factors contribute to a lack of long-term and experimentally-sound studies, which make it very challenging to test ecological hypotheses (Ripple et al. 2014b). In order to successfully move forward with our carnivore management plans, it is important to acknowledge that we no longer live in a natural world, and large carnivores are not the only species impacted by human domination of the world's ecosystems. Humans have modified chemical cycles, caused global climate change, and created a new geologic epoch based on the high impact of our activity. Just as systems differ depending on human activity, the effects of large carnivores will largely depend on the state of the ecosystem they occupy, as well as the types of human activity in that system (Haswell et al. 2017). 7.2.2 Contemporary Large Carnivore Recovery and Persistence In the wake of this new wildlife management paradigm and after the creation of the new field of conservation biology, several regions of the world have experienced large carnivore recovery. This remains a polarizing issue, as opponents of large carnivore recovery maintain attitudes described above, but support for large carnivore recovery 142 grows, in part because large carnivores are considered to be some of the world's most charismatic species. The threat of losing some species at the local or even global level has generated interest from government and private groups to create legislation and enact management plans to conserve these species. Subsequently, over the last several decades, large carnivores have recovered or persisted in some landscapes around the world. One of the best and most well-studied examples is the reintroduction of gray wolves to Yellowstone National Park, USA (Smith and Bangs 2009). Here is an example of one of the largest expanses of protected intact habitat in North America with a near full assemblage of species across trophic levels. Reintroduction of wolves has been reported to alter ecosystem structure through such ecological principles as mesopredator release hypothesis, trophic cascade hypothesis, and behavior-meditiated trophic cascade hypothesis (Ripple and Beschta 2004, 2007, Ripple et al. 2014a). These results are somewhat controversial (Beschta et al. 2014, Winnie 2014), however, and the authors have been criticized for a failure to evaluate alternative hypotheses (Allen et al. 2017). Much attention has been given to the Yellowstone wolf reintroduction project, though it may promote unwarranted support and justification of reintroduction of large carnivores for ecosystem restoration. Europe has also experienced a remarkable recovery of large carnivores (Chapron et al. 2014), mainly as a result of natural recolonizations, but also through managed reintroductions. Conservation success of large carnivores in Europe is largely due to protective legislation and increasing social carrying capacity through influencing public opinion (Chapron et al. 2014). The most important result from these activities is the widespread acceptance of the idea that large carnivores and people can coexist on the 143 same landscape (Lopez-Bao et al. 2015). Whether recovery has been intentional through planned reintroduction (e.g., Yellowstone), occurred naturally through recolonization (e.g., Western Europe), or species have persisted as in Sarıkamış Forest, proponents of large carnivore recovery in our contemporary society typically have three main goals: (i) maintaining stable carnivore populations, (ii) preventing conflict with carnivores (e.g., property damage and competition over game), and (iii) building public support for carnivore conservation. These are the goals our proposed management solutions will strive to achieve in humandominated ecosystems. 7.2.3 Large Carnivore Recovery in Human-Dominated Ecosystems Historically, large carnivore management and conservation has been driven by human needs and desires with little impetus to incorporate ecological information into the decision making process. As the effect of large carnivores continues to be debated, carnivores themselves are recolonizing human-dominated landscapes, and humans are interacting with these species for the first time in novel ecosystems. Conservation of large carnivores faces a new set of challenges in the Anthropocene. The combination of human population growth and expanding human ecological footprint has resulted in the destruction and degradation of wildlife habitat and subsequently, a global decline of species. Though many large carnivore species are highly intelligent and adaptable, those that are able to exploit anthropogenic food resources and use synanthropic behavior to increase fitness are more likely to recover in human-dominated ecosystems than species that do not. Generalist synanthropic species, such as gray wolves and brown bears, are 144 most likely to be successful in these novel ecosystems. However, they are the exception. Currently, of the 295 species of mammal carnivores, 115 species are extinction-prone (extinct, threatened or near threatened with extinction), and most of these are small, tropical specialists (IUCN 2016). 7.2.4 Coexistence with Large Carnivores It is often a challenge for people to co-exist with large carnivores because of competition for prey, threats to livelihood (e.g., livestock), or to human well-being. Of course, these factors are the driving force behind persecution of these animals. As such, the field of conservation biology has recently been interested in the question of coexistence of humans and large carnivores on the same landscape (Bergstrom 2017). Advances in technology and field methods, as well as increased human-wildlife conflict due to increasing human population and habitat encroachment, has stimulated more research focused on this topic. At broad scales, it may appear that humans and large carnivores are not able to regularly use the same locations. Life history traits (e.g., high resource requirement) and human-carnivore conflict has steered large carnivore conservation towards establishing and expanding protected habitat (i.e., protected areas) with low densities of human settlements (Mills 1991, Mech 1996). These protected areas provide resources and space while reducing the likelihood of human-carnivore conflict. But Anthropocene conditions may limit the capability of conservationists to designate large protected areas that satisfy the traditional requirements of large carnivores. At fine spatial scales, some carnivore species are able to coexist in human dominated 145 landscapes. Generalist carnivores of lower body mass are frequently observed in urban and suburban environments (Gehrt et al. 2010) and large-bodied carnivores such as cougars, coyotes, wolves, and bears are also known to be synanthropic. For example, studies of diel scale activity patterns have demonstrated that tigers and humans can coexist (Carter et al. 2012) and that cougars can use riparian corridors to navigate through high-density human settlements (Dickson et al. 2005) In many examples of human-carnivore coexistence, co-adaption (both humans and carnivore change their behavior) has been suggested as a critical component for successful coexistence (Carter and Linnell 2016). Carnivores have been observed changing diets, movement patterns, and range size in response to human presence, but in order for coexistence to occur, humans also need to adapt to the presence of large carnivores. This may involve accepting a tolerable level of risk to increase social carrying capacity enough to ensure long-term carnivore population persistence (Carter and Linnell 2016). As evidence of successful human-carnivore coexistence grows, some conservationists ask the question whether intact wilderness is necessary for these species. To be clear, there is no doubt that protected areas are vitally important to conserve some elements of biodiversity, especially for habitat specialists that make up the majority of species in many ecosystems; however, these ideas have been debated (Lopez-Bao et al. 2015), and studies suggest that for some species of large carnivores, large tracts of protected area are not necessary (Lopez-Bao et al. 2015). Instead, human-carnivore coexistence may be a requisite for some species of large carnivores to persist in the Anthropocene. 146 7.3 Large Carnivore Management in Sarıkamış-Allahuekber Mountains National Park in the Caucasus Global Biodiversity Hotspot An example of large carnivores persisting in a human-dominated landscape is the Sarıkamış-Allahuekber Mountains National Park in Northeastern Turkey. Here, we have observed a hyperabundance of large carnivores in the absence of a significant natural prey base (Chynoweth et al. in prep) coexisting with humans in an agrarian landscape and relying on anthropogenic food sources (Capitani et al. 2016, Cozzi et al. 2016). Wildlife resources are largely left unmanaged, and to our knowledge, few data have been collected for natural resources in the region (Ambarlı et al. 2016). Moving forward, a large carnivore management plan will ensure that large populations of bears, wolves, and lynx will persist, and biodiversity as a whole will be conserved. The goal of this management plan will be to effectively monitor biodiversity in the area and reduce human-wildlife conflict to ensure the persistence of large carnivores. Large carnivores are charismatic species, but due to their magnified potential impacts on rural communities, management and conservation of large carnivores in such areas represent a special challenge. While damage to livestock and crops, competition with hunters, and threats to human safety are historical problems for people living in urban and suburban areas, these concerns are a reality for many rural populations that currently coexist with these animals. In Sarıkamış, human-wildlife conflict is well documented (Chynoweth et al. 2016), but few solutions currently exist. In order to be successful, the management of large carnivores here must include community involvement and needs to emphasize tolerance of large carnivore presence by society. Furthermore, management 147 strategies must be based on scientific data regarding population status and dynamics, ecology, and interaction with other species and humans. Established in 2007 by Dr. Çağan Şekercioğlu, the KuzeyDoğa Society is a volunteerbased environmental nonprofit, nongovernmental organization (NGO) based in Kars, Turkey and poised to make critical contributions to a large carnivore conservation plan in Sarıkamış (Akkucuk and Sekercioglu 2016). As a local organization, KuzeyDoğa Society has a positive reputation in the region and importantly, is able to work with local communities to achieve conservation goals. Overall goals for large carnivore conservation in Sarıkamış will be to reduce human-wildlife conflict, increase social carrying capacity of large carnivore presence, and conserve biodiversity in the immediate area. The single biggest challenge to conserving biodiversity in Sarıkamış and the broader region will be effectively managing human activity. If legislation protecting biodiversity is implemented, substantial financial resources will be needed to generate paid positions for individuals to monitor biodiversity and enforce regulations related to land use and resource extraction. KuzeyDoğa Society has a small existing infrastructure that can be built upon to accommodate additional positions within the organization. In order to maintain local support, any resource management plan, legislation, or law enforcement must allow for local villagers to continue nondestructive land uses and therefore KuzeyDoğa Society is in a good position as an established regional NGO that is recognized locally for benefiting local communities through their past work (Akkucuk and Sekercioglu 2016). 148 7.3.1 Ecological Restoration The unique trophic structure and limited availability of resources in Sarıkamış will shape the plans to restore ecosystem function. Following main paradigms in restoration ecology, the goal of restoration will be to foster an ecosystem that is resilient and selfsustaining while supporting sustainable resources use by communities (Hobbs and Norton 1996). Efforts to restore ecological function in Sarıkamış will be focused on limiting human use of priority areas and reintroductions of native ungulates. There is no baseline ecosystem condition which these efforts are trying to restore. 7.3.1.1 Habitat Restoration Habitat restoration in Sarıkamış should focus primarily on limiting human use and resource extraction in areas identified as critical habitat for large carnivores. Large-scale vegetation restoration is unlikely in the region due to cost, as well as limited expertise and materials. Areas that have been identified as critical habitat (See Chapter 4 of this dissertation) are representative of the entire Sarıkamış Forest, ranging in elevation, geology, and vegetation. In the absence of intense human activity (e.g., livestock grazing), vegetation would gradually be restored, increasing habitat quality for all three target carnivore species, as well as providing critical resources for other wildlife and plant species in the region. A key component of habitat restoration will be designating and maintaining strict nature reserves and wilderness areas. This will be a significant challenge given the complex socio-political condition in the region; however, most land in Turkey is under the ownership and authority of the state. Therefore, designation of legally protected areas 149 can be seen as the principal tool in biodiversity conservation. A network of protected areas currently exists in Turkey, with 14 different types of protected areas across the country covering 7.2% of the country (Ambarlı et al. 2016), which includes SarıkamışAllahuekber Mountains National Park (SAMNP). However, as seen in SAMNP, protected area status is not equivalent to measurable success conserving biodiversity. By designating several types of protected areas in Sarıkamış, biodiversity can be conserved, and local communities can continue to utilize forest resources. Chapter 4 of this dissertation identifies the most suitable habitat for large carnivores in the region. Ideally, this area will be given protected area status which could fall into one (or more) of several categories. The best-case scenario would be to designate a portion of this area as a strict nature reserve (IUCN Category Ia) buffered by wilderness areas (IUCN Category Ib) or a matrix of habitat/species management areas (IUCN Category IV) and protected areas with sustainable use of natural resources (IUCN Category VI). The role of KuzeyDoğa Society is this process will primarily be as an advocacy group raising public awareness and lobbying for political support of legislation for protected area designation. The NGO has demonstrated their effectiveness at this process by creating eastern Turkey's first Ramsar wetland, designating Turkey's first wildlife corridor, connecting SAMNP to the larger, more intact forests of the Black Sea region (Şekercioğlu 2012) and promoting the "Save the Aras River Bird Paradise" campaign aimed to halt the planned Tuzluca Dam project on the Aras river through a change.org campaign (Akkucuk and Sekercioglu 2016). 150 7.3.1.2 Reintroduction of Native Prey Species The hyperabundance of large carnivores, lack of natural prey, and synanthropic behavior of wildlife directs restoration efforts towards the reintroduction of native ungulates into the system. Second only to designating new protected areas, this should be prioritized as a realistic short-term restoration goal in Sarıkamış Forest and SAMNP. Planned reintroductions of ungulates would have two goals: (i) reduce wolf depredation on livestock and (ii) restore native browsing ungulates on the landscape. Red deer (Cervus elaphus) is a potential candidate species especially with existing red deer reintroduction efforts in Turkey, but roe deer (Capreolus capreolus) is the most suitable candidate species for reintroduction based on several factors. Roe deer is an important prey species of wolves and lynx in many other parts of their range (Jędrzejewski et al. 1993, Mattioli et al. 2004). Roe deer have been successfully reintroduced in many other parts of the world, with detailed protocols available on reintroduction logistics (Calenge et al. 2005, Torres et al. 2016). Lastly, while roe deer are functionally extinct in Sarıkamış Forest, they have been captured rarely on camera traps and populations are known to exist in the broader region (Chynoweth et al. in prep). Successful roe deer reintroduction would fill a critical void in available natural prey species for lynx and wolves, two species that prey on livestock and are the source of human-wildlife conflict in the region (Chynoweth et al. 2016, Capitani et al. 2016). In other systems, when wild prey species are present, wolves will preferentially hunt wild ungulates (Imbert et al. 2016, Newsome et al. 2016) and even a small population of native ungulate reintroductions could facilitate wolves' switch to primarily wild ungulate diet (Meriggi and Lovari 1996). Combined with effective damage prevention measures 151 outlined in the following section, human-wildlife conflict would likely decrease as natural prey populations became reestablished. Reintroduction of roe deer will also likely benefit overall habitat restoration efforts in Sarıkamış by restoring important ecosystem effects, such as impacts on primary production, nutrient cycling, disturbance regimes, habitat heterogeneity, and seed dispersal (Hobbs 1996, Svenning et al. 2016). Importantly, a comprehensive ungulate reintroduction program would create opportunities to test important hypotheses in restoration ecology and the science of species reintroductions (Armstrong and Seddon 2008, Seddon et al. 2014). Reintroduction programs would best be initiated by the Ministry of Forestry and Water Affairs and the General Directorate of Nature Conservation and National Parks but KuzeyDoğa Society can contribute to reintroduction of wild ungulates in several ways based on their experience tagging and monitoring large carnivores in Sarıkamış forest. Given their knowledge of the region, the NGO can determine proper release locations and lead the long-term monitoring of released ungulates to determine the success of the program. KuzeyDoğa Society has been working for over 10 years with large carnivores and the local community in Sarıkamış, qualifying them to assess the impact of ungulate reintroduction on carnivore ecology and conservation. Establishing a self-sustaining population of native ungulates will also benefit local communities by providing legal hunting opportunities for residents. Sarıkamış is a rural, agrarian system with a population of people who regularly harvest forest products for sustenance (Chynoweth, pers observation). Hunting in Turkey is strictly enforced through the Terrestrial Hunting Law; however, illegal hunting is still a major cause of population 152 declines throughout the region (Ambarlı et al. 2016). By reintroducing, monitoring, and actively managing a roe deer population in Sarıkamış, residents can eventually be introduced to and participate in an adaptive harvest management plan. Tourism already plays an important role in Sarıkamış' economy (skiing) with existing infrastructure for tourists; in the future, ungulate populations could generate income for local people through hunting tourism, which currently exists in Turkey at low levels. 7.3.2 Human-Wildlife Conflict Human-wildlife conflict is a well-known and intensely examined topic in the field of large carnivore ecology and conservation. Typically, conflict arises when an animal poses a direct and recurring threat to the livelihood or safety of humans or their property. The impact of carnivore losses can be devastating for an individual or family, even if they appear small at the community level (Hill 2004). Globally, the frequency and costs of conflicts are increasing, likely due to growing human populations and ingress of humans into wildlife habitat (Treves and Karanth 2003). These conflicts often lead to the direct and indirect persecution of carnivores, which can lead to population declines or alteration of animal behavior (Ordiz et al. 2013). Public opinion of wildlife species-especially large carnivores-is a critical component in their management and conservation. Human dimensions of wildlife management have thus become a growing field of research with the goal of facilitating a level of human-wildlife coexistence that allows wildlife populations to persist (Madden 2004). Conflict can increase when the interests of local people are not included in management plans, or these people are not empowered to find their own solutions. If the 153 economic and social well-being of local communities is not considered, local support for conservation diminishes, and long-term goals and priorities of conservationists are not met. Therefore, identifying, communicating, and collaborating with stakeholders is essential to conservation of large carnivores, especially outside protected area boundaries (Treves et al. 2006). In Sarıkamış, our multiyear community opinion surveys have revealed humanwildlife conflict to be a critical barrier to large carnivore management and conservation (Chynoweth et al. 2016). Similar to other agrarian systems throughout the world, animal husbandry and farming are significant drivers of the economy. Consequently, livestock depredation and crop damage represents one of the biggest generators of conflict. In order to facilitate successful mitigation, multiple and adaptable tools are required (Madden 2004). KuzeyDoğa Society stands to make the largest contribution to large carnivore conservation by working with the local community to implement mitigation strategies. In the sections below, we outline some key tools the NGO can use to alleviate humanwildlife conflict in Sarıkamış. 7.3.3 Tools to Prevent Human-Wildlife Conflict Each of the target species has varying responsiveness to tools for reducing conflict and limiting persecution of large carnivores, with some tools equally effective for all species. The most effective approach for conservation of all three species will be to incorporate multiple tools into a comprehensive human-carnivore conflict management plan. Given appropriate financial support, KuzeyDoğa Society could effectively manage all mitigation programs by hiring additional full-time staff and become a leader in 154 reducing human-wildlife conflict and increasing carnivore tolerance in the region. Ultimately, a livestock damage compensation program would be most appropriate for the region, a technique known to increase tolerance of large carnivores on the landscape (Dickman et al. 2011). Several prerequisites are necessary to make this program operational. Most important, funding must be made available to supply this program with the necessary funds to compensate for livestock loss. Second, an expert in large carnivore depredation is an absolute necessity for the program. This individual must be properly trained and willing to train others in the community. Finally, cooperation from livestock owners and farmers who agree to be actively engaged in other nonlethal predatory management techniques and would be required to report livestock loss within 24 hours. KuzeyDoğa Society could manage this compensation program while providing expertise in identifying carnivore depredation. Another important tool to reduce human-wildlife conflict in Sarıkamış is to increase the quality of guardian dogs. This was the most common response from community members when they were asked what they needed to protect their livestock (Chynoweth et al. 2016). Most shepherds in the region currently use guardian dogs, but many state that they do not have access to effective dogs, nor the resources to care for dogs properly (e.g., proper food and veterinary care). Guardian dogs have a long history in Turkey and numerous existing groups have access to high quality dogs (Yilmaz et al. 2015). KuzeyDoğa Society could partner with existing dog breeders in other regions and local farmers to establish a dog training and breeding facility in Sarıkamış to produce high quality guardian dogs at subsidized prices for local shepherds. In addition to livestock damage compensation and increasing access to guardian dogs, 155 a suite of other nonlethal techniques exist to reduce human-wildlife conflict from livestock losses (reviewed in Cluff & Murray 1992; Bangs & Shivik 2001; Rigg 2001; Sillero-Subiri & Laurenson 2001; Rigg et al. 2011; Can et al. 2014; Stone et al. 2017). These include separating livestock and carnivores with a physical barrier (e.g., fencing, avoiding dangerous areas, livestock in pens at night), discouraging predators (e.g., electric fences, flandry, turbo flandry, spotlights, klaxon, radioactivated guardbox) and protecting livestock (e.g., guarding animals, harassing, nonlethal/blank ammunition, fitting protective collars). All or some of these nonlethal approaches can organized and distributed to livestock owners by KuzeyDoğa Society as a prerequisite to gaining access to more formal programs such as livestock compensation and access to guardian dogs. Lethal techniques for large carnivore management are routinely used throughout the world, but should represent the last possible option after all nonlethal techniques have been exhausted (Naughton-Treves and Treves 2005). Recent work demonstrates that nonlethal techniques are more effective than lethal techniques and that lethal techniques may actually be counter-productive (Stone et al. 2017). Such steps as direct persecution (e.g., poison, weapons) can result in unstable populations of these animals, which can lead to atypical behaviors, such as relying more heavily on anthropogenic food sources such as livestock (Naughton-Treves and Treves 2005). For example, the disruption of social structure in wolf packs may increase livestock depredation (Imbert et al. 2016). 7.3.4 Waste Management Waste management is the single most important factor in conservation and management of large carnivores in Sarıkamış. The impact of the large, open sky, 156 unmanaged municipal garbage dump here is well-known (Cozzi et al. 2016; Chapter 4). Garbage dumps are widely known to cause habituation of animals. Food-conditioned bears increase human-wildlife conflict and reduce the social carrying capacity for large carnivores on the landscape. For successful conservation management to occur, a plan to close the garbage dump must be initiated and must include methods to accommodate habituated bears, which may include a supplementary feeding program. While the municipal garbage dump presents the most glaring issue, food-conditioned bears are present across the entire study area and have been recorded raiding dumpsters at hotels and picnic sites (Chynoweth, pers obs). Proper disposal of human refuse (garbage, potential food) at smaller sites in bear-proof dumpsters would help alleviate humanwildlife conflict and prevent new generations of bears from becoming food-conditioned. Presence of these bear-proof dumpsters would also help educate the local community about habituated bears, and provide a clear solution. Successful waste management also necessitates the proper disposal of livestock waste (i.e., carcasses). Livestock husbandry is a major component of the local economy, and there are currently no known government regulations regarding disposal of carcasses. Livestock losses that occur while moving herds from one location to another are left behind, and remains from livestock processing plants are dumped at the municipal garbage dump. If these conditions persist, conflict at these disposal locations is inevitable, since people are inadvertently baiting carnivores in these areas frequented by humans. 157 7.3.5 Building Local Capacity and Increasing Social Carrying Capacity for Large Carnivore Conservation The local community in Sarıkamış is one of the most important components to success of a large carnivore conservation program. If social acceptance of large carnivores is low, populations will decline regardless of efforts from biologists to protect habitat or species. One approach to increase social carrying capacity is to involve community members in active research and natural resource management through citizen science and employment. This will encourage people to take ownership of natural resources and thus develop a need to contribute to conservation efforts. Citizen science opportunities for large carnivore management should focus on setting up a broad scale camera trapping effort throughout Sarıkamış forest. Citizen science camera trapping has proven successful in many other areas, and several procedures are available to train community members in camera deployment and data management (McShea et al. 2016, Forrester et al. 2016). By incorporating local people in camera trapping, two main goals would be accomplished. First, and most important, people would learn about the presence and importance of local biodiversity and conservation goals for the area. Second, if properly executed, community members could operate a complete camera-trapping monitoring program, which is an important component for the overall carnivore management plan in Sarıkamış. Also, by learning about camera traps and the reason this equipment is used, we can hope that there would be a decline in the rates of theft, which was a critical barrier in our own work. This citizen science effort would create advocates for mammal conservation and for the organizations facilitating biodiversity conservation (Forrester et al. 2016). 158 Another opportunity to build local capacity is to train individuals to conduct community opinion surveys to understand local perspectives of large carnivores in an effort to reduce human-wildlife conflict. Previous efforts in Sarıkamış suggest that people are very willing to participate in these surveys; however, foreign survey administrators were concerned that responses may not be unbiased as villagers were reluctant to share sensitive or incriminating information with outsiders (Chynoweth et al. 2016), By employing local people to conduct surveys, our results would be less biased and more informative. A major component of this work needs to focus on education and outreach to communicate to the local community the importance of biodiversity and of increasing social carrying capacity of large carnivores in Sarıkamış. The KuzeyDoğa Society has led a few small outreach programs in the region, all of which have been well received. Still more programs need to be implemented to reach people in small villages on the edges of critical habitat. These are the people that likely have the most interaction with large carnivores. Through outreach programs, the KuzeyDoğa Society will become a leader in large carnivore conservation in the broader region and deepen their relationship with the local community as an organization bringing social and economic benefits to the area. As such, facilitating the management approaches described here will become more efficient and effective. 7.4 Conclusions At first glance, the situation in Sarıkamış appears to be somewhat unique in terms of mammal community structure, large carnivore hyperabundance, and degree of 159 synanthropic behavior. But the reality is that these conditions may be more common than previously thought, and may become more common in the future. Sarıkamış Forest is a combination of human encroachment on natural habitat, temperate agrarian system, poor socioeconomic conditions, and being understudied by biologists. How many places like this exist in the world? Are generalist carnivores exhibiting the same behavior? I have had many conversations at academic conferences and with collaborators in an effort to seek out systems with similar parameters and ask the question, "Is Sarıkamış unique, or do other places exist with similar conditions?" Anecdotally, other researchers have shared similar experiences, but rarely have data to share regarding mammal community structure or human-carnivore interactions. In the future, I predict that more systems reflecting conditions in Sarıkamış Forest will be uncovered as generalist predators continue to persist or recover in human-dominated landscapes. Wildlife on our planet is increasingly living out their lives in human-altered landscapes that are continuously threatened by chemical inputs, human activity, and habitat loss. In addition to losing the intrinsic value of biodiversity, humans are also losing the ecosystems services that provide purification processes, pollination of crops, and other critical ecosystem services. Globally, we also know that large carnivore populations are declining and threatened by human activity (Ripple et al. 2014b). Complicating conservation efforts is the fact that the effect of large carnivores on these systems is still largely unknown but often sensationalized (Allen et al. 2017). To alleviate pressures on large carnivores and biodiversity in general in systems such as Sarıkamış, we have identified the management recommendations described above, focused on 160 restoring ecological function, reducing human-wildlife conflict, and increasing social carrying capacity. 7.5 References Akkucuk, U., and C. H. Sekercioglu. 2016. NGOs for environmental sustainability: the case of KuzeyDoga Foundation. Fresenius Environmental Bulletin 25:6038-6044. Allen, B. L., L. R. Allen, H. Andrén, G. Ballard, L. Boitani, R. M. Engeman, P. J. S. Fleming, A. T. Ford, P. M. Haswell, R. Kowalczyk, J. D. C. Linnell, L. David Mech, and D. M. Parker. 2017. Can we save large carnivores without losing large carnivore science? Food Webs 12:64-75. Ambarlı, D., U. S. Zeydanlı, Ö. Balkız, S. Aslan, E. Karaçetin, M. Sözen, Ç. Ilgaz, A. Gürsoy Ergen, Y. Lise, S. Demirbaş Çağlayan, H. J. Welch, G. Welch, A. S. Turak, C. C. Bilgin, A. Özkil, and M. Vural. 2016. An overview of biodiversity and conservation status of steppes of the Anatolian biogeographical region. Biodiversity and Conservation 25:2491-2519. Armstrong, D., and P. Seddon. 2008. Directions in reintroduction biology. Trends in Ecology & Evolution 23:20-25. Bangs, E., and J. Shivik. 2001. Managing wolf conflict with livestock in the northwestern United States. Carnivore Damage Prevention News 3:1999-2002. Barnosky, A. D., E. a Hadly, J. Bascompte, E. L. Berlow, J. H. Brown, M. Fortelius, W. M. Getz, J. Harte, A. Hastings, P. a Marquet, N. D. Martinez, A. Mooers, P. Roopnarine, G. Vermeij, J. W. Williams, R. Gillespie, J. Kitzes, C. Marshall, N. Matzke, D. P. Mindell, E. Revilla, and A. B. Smith. 2012. Approaching a state shift in Earth's biosphere. Nature 486:52-8. Bergstrom, B. J. 2017. Carnivore conservation: shifting the paradigm from control to coexistence. Journal of Mammalogy 98:1-6. Beschta, R. L., C. Eisenberg, J. W. Laundré, W. J. Ripple, and T. P. Rooney. 2014. Predation risk, elk, and aspen: comment. Ecology 95:2669-2671. Calenge, C., D. Maillard, N. Invernia, and J. C. Gaudin. 2005. Reintroduction of roe deer Capreolus capreolus into a Mediterranean habitat: female mortality and dispersion. Wildlife Biology 11:153-161. Can, Ö. E., N. D'Cruze, D. L. Garshelis, J. Beecham, and D. W. Macdonald. 2014. Resolving human-bear conflict: a global survey of countries, experts, and key 161 factors. Conservation Letters 7:501-513. Capitani, C., M. Chynoweth, J. Kusak, E. Çoban, and Ç. H. Şekercioğlu. 2016. Wolf diet in an agricultural landscape of north-eastern Turkey. Mammalia 80:329-334. Carter, N. H., and J. D. C. Linnell. 2016. Co-adaptation is key to coexisting with large carnivores. Trends in Ecology & Evolution 31:575-578. Carter, N. H., B. K. Shrestha, J. B. Karki, N. M. B. Pradhan, and J. Liu. 2012. Coexistence between wildlife and humans at fine spatial scales. Proceedings of the National Academy of Sciences of the United States of America 109:15360-5. Chapron, G., P. Kaczensky, J. D. C. Linnell, M. von Arx, D. Huber, H. Andren, J. V. Lopez-Bao, M. Adamec, F. Alvares, O. Anders, L. Bal iauskas, V. Balys, P. Bed, F. Bego, J. C. Blanco, U. Breitenmoser, H. Broseth, L. Bufka, R. Bunikyte, P. Ciucci, A. Dutsov, T. Engleder, C. Fuxjager, C. Groff, K. Holmala, B. Hoxha, Y. Iliopoulos, O. Ionescu, J. Jeremi, K. Jerina, G. Kluth, F. Knauer, I. Kojola, I. Kos, M. Krofel, J. Kubala, S. Kunovac, J. Kusak, M. Kutal, O. Liberg, A. Maji, P. Mannil, R. Manz, E. Marboutin, F. Marucco, D. Melovski, K. Mersini, Y. Mertzanis, R. W. Mys ajek, S. Nowak, J. Odden, J. Ozolins, G. Palomero, M. Paunovi, J. Persson, H. Poto nik, P.Y. Quenette, G. Rauer, I. Reinhardt, R. Rigg, A. Ryser, V. Salvatori, T. Skrbin ek, A. Stojanov, J. E. Swenson, L. Szemethy, A. Trajce, E. Tsingarska-Sedefcheva, M. Va a, R. Veeroja, P. Wabakken, M. Wolfl, S. Wolfl, F. Zimmermann, D. Zlatanova, and L. Boitani. 2014. Recovery of large carnivores in Europe's modern humandominated landscapes. Science 346:1514-1517. Chynoweth, M. W., E. Çoban, Ç. Altın, and Ç. H. Şekercioğlu. 2016. Human-wildlife conflict as a barrier to large carnivore management and conservation in Turkey. Turkish Journal of Zoology 40:972-983. Chynoweth, M. W., E. Çoban, and C. H. Sekercioglu. (in Review.). Large carnivore hyperabundance in an empty forest. Ecology. Cluff, H. D., and D. L. Murray. 1992. Review of wolf control methods in North America. Proceedings of the Second North America Symposium on Wolves held in Edmonton, Alberta, Canada:491-504. Cozzi, G., M. Chynoweth, J. Kusak, E. Çoban, A. Çoban, A. Ozgul, and Ç. H. Şekercioğlu. 2016. Anthropogenic food resources foster the coexistence of distinct life history strategies: year-round sedentary and migratory brown bears. Journal of Zoology 300:142-150. Dickman, A. J., E. a Macdonald, and D. W. Macdonald. 2011. A review of financial instruments to pay for predator conservation and encourage human-carnivore coexistence. Proceedings of the National Academy of Sciences of the United States of America 108:13937-13944. 162 Dickson, B. G., J. S. Jenness, and P. Beier. 2005. Influence of vegetation, topography, and roads on cougar movement in southern California. Journal of Wildlife Management 69:264-276. Forrester, T. D., M. Baker, R. Costello, R. Kays, A. W. Parsons, and W. J. McShea. 2016. Creating advocates for mammal conservation through citizen science. Biological Conservation 208:98-105. Gehrt, S. D., S. P. Riley, and B. L. Cypher, editors. 2010. Urban carnivores: ecology, conflict, and conservation. The Johns Hopkins University Press, Baltimore, MD. Haswell, P. M., J. Kusak, and M. W. Hayward. 2017. Large carnivore impacts are context-dependent. Food Webs 12:3-13. Hill, C. M. 2004. Farmers' perspectives of conflict at the wildlife-agriculture boundary: some lessons learned from African subsistence farmers. Human Dimensions of Wildlife 9:279-286. Hobbs, N. T. 1996. Modification of ecosystems by ungulates. The Journal of Wildlife Management 60:695-713. Hobbs, R. J., E. Higgs, C. M. Hall, P. Bridgewater, F. S. Chapin, E. C. Ellis, J. J. Ewel, L. M. Hallett, J. Harris, K. B. Hulvey, S. T. Jackson, P. L. Kennedy, C. Kueffer, L. Lach, T. C. Lantz, A. E. Lugo, J. Mascaro, S. D. Murphy, C. R. Nelson, M. P. Perring, D. M. Richardson, T. R. Seastedt, R. J. Standish, B. M. Starzomski, K. N. Suding, P. M. Tognetti, L. Yakob, and L. Yung. 2014. Managing the whole landscape: historical, hybrid, and novel ecosystems. Frontiers in Ecology and the Environment 12:557-564. Hobbs, R. J., and D. A. Norton. 1996. Towards a conceptual framework for restoration ecology. Restoration Ecology 4:93-110. Imbert, C., R. Caniglia, E. Fabbri, P. Milanesi, E. Randi, M. Serafini, E. Torretta, and A. Meriggi. 2016. Why do wolves eat livestock? Factors influencing wolf diet in northern Italy. Biological Conservation 195:156-168. IUCN. 2016. The IUCN Red List of Threatened Species. Version 2016-3. http://www.iucnredlist.org/. Jędrzejewski, W., K. Schmidt, L. Miłkowski, B. Jędrzejewska, and H. Okarma. 1993. Foraging by lynx and its role in ungulate mortality: the local (Białowieża Forest) and the Palaearctic viewpoints. Acta Theriologica 38:385-403. Lopez-Bao, J. V., P. Kaczensky, J. D. C. Linnell, L. Boitani, and G. Chapron. 2015. Carnivore coexistence: wilderness not required. Science 348:871-872. 163 Madden, F. 2004. Creating coexistence between humans and wildlife: global perspectives on local efforts to address human-wildlife conflict. Human Dimensions of Wildlife 9:247-257. Martin, L. J., J. E. Quinn, E. C. Ellis, M. R. Shaw, M. A. Dorning, L. M. Hallett, N. E. Heller, R. J. Hobbs, C. E. Kraft, E. Law, N. L. Michel, M. P. Perring, P. D. Shirey, and R. Wiederholt. 2014. Conservation opportunities across the world's anthromes. Diversity and Distributions 20:745-755. Mattioli, L., C. Capitani, E. Avanzinelli, I. Bertelli, A. Gazzola, and M. Apollonio. 2004. Predation by wolves (Canis lupus) on roe deer (Capreolus capreolus) in northeastern Apennine, Italy. Journal of Zoology 264:249-258. McShea, W. J., T. Forrester, R. Costello, Z. He, and R. Kays. 2016. Volunteer-run cameras as distributed sensors for macrosystem mammal research. Landscape Ecology 31:55-66. Mech, L. D. 1996. Conservation strategies: a new era for carnivore conservation. Wildlife Society Bulletin 24:397-401. Meriggi, a, and S. Lovari. 1996. A review of wolf predation in southern Europe: does the wolf prefer wild prey to livestock? Journal of Applied Ecology 33:1561-1571. Mills, M. G. L. 1991. Conservation management of large carnivores in Africa. Koedoe 34:81-90. Naughton-Treves, L., and A. Treves. 2005. Evaluating lethal control in the management of human-wildlife conflict. Pages 86-106 in R.Woodroff, S. Thirgood, and A. Rabinowitz, editors. People and Wildlife: Conflict or Coexistence? Cambridge University Press, Cambridge, UK. Newsome, T. M., L. Boitani, G. Chapron, P. Ciucci, C. R. Dickman, J. A. Dellinger, J. V. López-Bao, R. O. Peterson, C. R. Shores, A. J. Wirsing, and W. J. Ripple. 2016. Food habits of the world's grey wolves. Mammal Review 46:255-269. Ordiz, A., R. Bischof, and J. E. Swenson. 2013. Saving large carnivores, but losing the apex predator? Biological Conservation 168:128-133. Rigg, R. 2001. Livestock guarding dogs: their current use worldwide. Page IUCN/SSC Canid Specialist Group. Rigg, R., S. Finďo, M. Wechselberger, M. L. Gorman, C. Sillero-Zubiri, and D. W. Macdonald. 2011. Mitigating carnivore-livestock conflict in Europe: lessons from Slovakia. Oryx 45:272-280. Ripple, W. J., and R. L. Beschta. 2007. Restoring Yellowstone's aspen with wolves. 164 Biological Conservation 138:514-519. Ripple, W. J., R. L. Beschta, J. K. Fortin, and C. T. Robbins. 2014a. Trophic cascades from wolves to grizzly bears in Yellowstone. The Journal of Animal Ecology 83:223-33. Ripple, W. J., J. A. Estes, R. L. Beschta, C. C. Wilmers, E. G. Ritchie, M. Hebblewhite, J. Berger, B. Elmhagen, M. Letnic, M. P. Nelson, O. J. Schmitz, D. W. Smith, a. D. Wallach, and A. J. Wirsing. 2014b. Status and ecological effects of the world's largest carnivores. Science 343:1241484-1241484. Ripple, W. J., J. A. Estes, O. J. Schmitz, V. Constant, M. J. Kaylor, A. Lenz, J. L. Motley, K. E. Self, D. S. Taylor, and C. Wolf. 2016. What is a trophic cascade? Trends in Ecology and Evolution 31:842-849. Ripple, W. J., A. J. Wirsing, C. C. Wilmers, and M. Letnic. 2013. Widespread mesopredator effects after wolf extirpation. Biological Conservation 160:70-79. Ripple, Wi. J., and R. L. Beschta. 2004. Wolves and the ecology of fear: can predation risk structure ecosystems? Bioscience 54:755-766. Seddon, P. J., C. J. Griffiths, P. S. Soorae, and D. P. Armstrong. 2014. Reversing defaunation: restoring species in a changing world. Science 345:406-412. Şekercioğlu, Ç. H. 2012. Turkey's first wildlife corridor links bear, wolf and lynx populations to the Caucasus forests. http://newswatch.nationalgeographic.com/2012/02/13/turkeys-first-wildlife-corridorlinks-bear-wolf-and-lynx-populations-to-the-caucasus-forests/. Sillero-Zubiri, C., and K. Laurenson. 2001. Interactions between carnivores and local communities: conflict or co-existence? Carnivore Conservation 45:282-312. Smith, D. W., and E. E. Bangs. 2009. Reintroduction of wolves to Yellowstone National Park: history, values and ecosystem restoration. Pages 92-125 in M. Somers, editor. Reintroduction of Top-order Predators. Wiley-Blackwell, Oxford, UK. Stone, S. A., S. W. Breck, J. Timberlake, P. M. Haswell, F. Najera, B. S. Bean, and D. J. Thornhill. 2017. Adaptive use of nonlethal strategies for minimizing wolf-sheep conflict in Idaho. Journal of Mammalogy 98:33-44. Svenning, J.-C., P. B. M. Pedersen, C. J. Donlan, R. Ejrnæs, S. Faurby, M. Galetti, D. M. Hansen, B. Sandel, C. J. Sandom, J. W. Terborgh, and F. W. M. Vera. 2016. Science for a wilder Anthropocene: synthesis and future directions for trophic rewilding research. Proceedings of the National Academy of Sciences 113:898-906. Torres, R. T., J. Carvalho, C. Fonseca, E. Serrano, and J. M. López-Martín. 2016. 165 Longterm assessment of roe deer reintroductions in North East Spain: a case of success. Mammalian Biology 81:415-422. Treves, A., and K. U. Karanth. 2003. Human-carnivore conflict and perspectives on carnivore management worldwide. Conservation Biology 17:1491-1499. Treves, A., R. B. Wallace, L. Naughton-Treves, and A. Morales. 2006. Co-managing human-wildlife conflicts: a review. Human Dimensions of Wildlife 11:383-396. Wallach, A. D., I. Izhaki, J. D. Toms, W. J. Ripple, and U. Shanas. 2015. What is an apex predator? Oikos 124:1453-1461. Waters, C. N., J. Zalasiewicz, C. Summerhayes, A. D. Barnosky, C. Poirier, A. Ga uszka, A. Cearreta, M. Edgeworth, E. C. Ellis, M. Ellis, C. Jeandel, R. Leinfelder, J. R. McNeill, D. d. Richter, W. Steffen, J. Syvitski, D. Vidas, M. Wagreich, M. Williams, A. Zhisheng, J. Grinevald, E. Odada, N. Oreskes, and A. P. Wolfe. 2016. The Anthropocene is functionally and stratigraphically distinct from the Holocene. Science 351:137-148. Winnie, J. J. 2014. Predation risk, elk, and aspen: reply. Ecology 95: 2671-2674. Yilmaz, O., M. Ertugrul, and R. T. Wilson. 2015. The domestic livestock resources of Turkey: breed descriptions and status of guard and hunting dogs. Research Opinions in Animal & Veterinary Sciences 5:172-177. Zalasiewicz, J., M. Williams, A. Smith, T. L. Barry, A. L. Coe, P. R. Bown, P. Brenchley, D. Cantrill, A. Gale, P. Gibbard, F. J. Gregory, M. W. Hounslow, A. C. Kerr, P. Pearson, R. Knox, J. Powell, C. Waters, J. Marshall, M. Oates, P. Rawson, and P. Stone. 2008. Are we now living in the Anthropocene? GSA Today 18:4. APPENDIX A CONSERVATION OF A NEW BREEDING POPULATION OF CAUCASIAN LYNX (LYNX LYNX DINNIKI) IN EASTERN TURKEY Chynoweth, M. W., E. Çoban, and Ç. H. Şekercioğlu. 2015. Conservation of a new breeding population of Caucasian lynx (Lynx lynx dinniki) in eastern Turkey. Turkish Journal of Zoology 39:1-3. Reprinted with permission from the Scientific and Technological Research Council of Turkey . 167 168 169 APPENDIX B WOLF DIET IN AN AGRICULTURAL LANDSCAPE OF NORTHEASTERN TURKEY Capitani, C., M. Chynoweth, J. Kusak, E. Çoban, and Ç. H. Şekercioğlu. 2016. Wolf diet in an agricultural landscape of north-eastern Turkey. Mammalia 80: 329-334. Reprinted with permission from the De Gruyter. 171 172 173 174 175 176 APPENDIX C ANTHROPOGENIC FOOD RESOURCES FOSTER THE COEXISTENCE OF DISTINCT LIFE HISTORY STRATEGIES: YEAR-ROUND SEDENTARYAND MIGRATORY BROWN BEARS Cozzi, G., M. Chynoweth, J. Kusak, E. Çoban, A. Çoban, A. Ozgul, and Ç. H. Şekercioğlu. 2016. Anthropogenic food resources foster the coexistence of distinct life history strategies: year-round sedentary and migratory brown bears. Journal of Zoology 300(2): 142-150. Reprinted with permission from John Wiley & Sons, Inc. 178 179 180 181 182 183 184 185 186