|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 JAGUAR ( PANTHERA ONCA ) POPULATION DYNAMICS, FEEDING ECOLOGY, HUMAN INDUCED MORTALITY, AND CONSERVATION IN THE VRZEA FLOODPLAIN FORESTS OF AMAZONIA By EMILIANO ESTERCI RAMALHO A DISSERTATION PRESENTED TO THE GRADUATE SCHOO L OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Emiliano Esterci Ramalho
3 To my mother and father
4 ACKNOWLEDGMENTS I thank my parents for th eir love and unconditional support of my work. I thank Mariana Mieko Sakamoto, Paulo Faiad, Joana Macedo, and Daniel Rocha for their his assistance, friendship, wisdo m, and incredible knowledge of the Vrzea. I thank participating in my jaguar resea rch project. I thank all the communities of Mamirau and Aman Sustainable Development Reserves for their trust and willingness to participate in my research and provide critical information for the success of this study. I thank Jos Mrcio Ayres for his trust in my potential to execute this study and create a long term jaguar research project in Mamirau Reserve I thank Leandro Castello for my first opportunity to work in Amazonia and all the learning and opportunities that came from my internship with h im. I thank Ana Rita Alves, Helder Queiroz, and Joo Valsecchi for their continued support of my research throughout the years. I thank Jim Nichols, Katie Sieving, Mel Sunquist, and Brian Child for their great courses at University of Florida, wisdom, and willingness to be a part of my committee. It was an honor for me to have such brilliant minds be a part of my PhD. I thank Jim Hines, Claudia Penaloza, and Catherine Langtimm for the invaluable assistance in the analysis of the data. I thank Brandy Jo Petr onio for her friendship and grammar review of this dissertation. I thank the crew of boat Gaivota for transporting me safely during field work. I thank the whole staff of Mamirau Sustainable Development Institute for their assistance throughout the years
5 This work was made possible by funding and logistical support from Mamirau Sustainable Development Institute, the Brazilian Ministry of Science Technology and I nnovation (MCTI), the Coordenao de Aperfeioamento de Pessoal de Nvel Superior (CAPES ), Fu lbright, Wildlife Conservation Society, Panthera, and University of Florida.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 THE JAGUAR CONSERVATION PROBLEM IN BRAZIL A VALUES PROBLEM ................................ ................................ ................................ .............. 15 The Policy Sciences Method for Solving Problems ................................ ................. 19 The Jaguar Conservation Problem in Brazil (Problem Orien tation) ........................ 22 Definition of the Conservation Problem ................................ ............................ 22 Current Goal of the Policy Process ................................ ................................ .. 23 Historical Trends and Conditions (Analysis of the Problem) ............................. 23 Habitat Loss ................................ ................................ ................................ ..... 23 Direct Killing of Jagua rs ................................ ................................ .................... 28 Decrease of Prey Populations ................................ ................................ .......... 33 Jaguar Population Vigor ................................ ................................ ................... 34 Future Projections ................................ ................................ ............................ 35 The Social Context ................................ ................................ ................................ .. 35 The Decision Process ................................ ................................ ............................. 39 Research and Planning (Intelligence Function) ................................ ................ 39 Recommending and Debating Policies (Promotion Function) .......................... 41 Creating, Imple menting and Enforcing Rules (Prescription, Invocation and Application Functions) ................................ ................................ ................... 41 Evaluation and Termination of Rules (Appraisal and Termination Functions) .. 43 Current Solutions, Recommended Solutions, and Alternative Solutions ................. 43 Bans on Hunting and Trade of Jaguars and Prey ................................ ............. 43 Protected Areas ................................ ................................ ................................ 45 Better Management of Livestock ................................ ................................ ...... 46 Techniques for Adverse Conditioning of Predators ................................ .......... 47 Corridors ................................ ................................ ................................ ........... 48 Translocations ................................ ................................ ................................ .. 49 Compensation Schemes ................................ ................................ .................. 49 Ecotourism ................................ ................................ ................................ ....... 50 Sports Hunting ................................ ................................ ................................ .. 51 Environmental Education ................................ ................................ .................. 52 Discussion ................................ ................................ ................................ .............. 52
7 2 JAGUAR ( PANTHERA ONCA ) POPULATION DYNAMICS AND ACTIVITY PATTERNS IN A SUSTAINABLE USE RESERVE IN THE VRZEA FLOODPLAIN FORES TS OF BRAZILIAN AMAZONIA ................................ .......... 69 Methods ................................ ................................ ................................ .................. 72 Study Site ................................ ................................ ................................ ......... 72 Vrzea Floo dplain Forests ................................ ................................ ......... 72 Mamirau Sustainable Development Reserve ................................ ........... 74 Field Methods ................................ ................................ ................................ ... 79 Camera trap surveys ................................ ................................ ................. 79 Foot snare live captures ................................ ................................ ............. 81 Data Analysis ................................ ................................ ................................ ... 82 Population density ................................ ................................ ...................... 82 Survival and recapture probabilities ................................ ........................... 85 Activity patterns ................................ ................................ .......................... 87 Results ................................ ................................ ................................ .................... 88 Density ................................ ................................ ................................ ............. 88 Survival and Recapture Probabilities ................................ ................................ 89 Population Structure and Reproduction ................................ ............................ 90 Activity Patterns ................................ ................................ ................................ 91 Discussion ................................ ................................ ................................ .............. 92 3 ESTIMATING LARGE CARNIVORE MORTALITY FROM HUNTING USING CAPTURE RECAPTURE MODELS: THE CASE OF JAGUARS IN THE AMAZON FLOODPLAIN FORESTS ................................ ................................ ..... 114 Met hods ................................ ................................ ................................ ................ 119 Study Sites ................................ ................................ ................................ ..... 119 Characterizing Hunting Events ................................ ................................ ....... 122 Estim ating Total Number of Jaguars Killed using CR Methodology ............... 124 Important Assumptions for Estimation of Total Number of Jaguars Killed ...... 127 Results ................................ ................................ ................................ .................. 128 Distribution of Hunting Events among Environments ................................ ..... 128 Seasonality of Hunting Events ................................ ................................ ........ 129 Hunting Pressure on Jaguars and Pumas ................................ ...................... 130 Hunting Pressure on Males and Females ................................ ....................... 130 Opportunistic Versus Intentional Hunting ................................ ....................... 130 Hunting Method ................................ ................................ .............................. 131 Motive of Hunt ................................ ................................ ................................ 131 Activity of the Hunter ................................ ................................ ...................... 132 Consumption of Meat ................................ ................................ ..................... 132 Estimates of Total Number of Jaguars Killed ................................ .................. 133 Discussion ................................ ................................ ................................ ............ 134 Conclusions ................................ ................................ ................................ .......... 140
8 4 THE IMPORTANCE OF CAIMANS AND ARBOREAL M AMMALS IN THE DIET OF THE JAGUAR ( PANTHERA ONCA ) IN THE VRZEA FLOODPLAIN FORESTS OF AMAZONIA. ................................ ................................ .................. 155 Methods ................................ ................................ ................................ ................ 156 Study Area ................................ ................................ ................................ ...... 156 Collection and Analysis of Scats ................................ ................................ .... 157 Results ................................ ................................ ................................ .................. 159 Discussion ................................ ................................ ................................ ............ 161 APPENDIX: BIBLIOGRAPHY REVIEW METHODS ................................ .................... 171 LIST OF REFERENCES ................................ ................................ ............................. 176 B IOGRAPHICAL SKETCH ................................ ................................ .......................... 195
9 LIST OF TABLES Table page 1 1 The conceptual framework of the social process (adapted from: Lasswell, 1971; Clark & Wallace, 1998) ................................ ................................ ............. 56 1 2 The seven decision functions that constitute a policy process (adapted from: Lasswell, 1971; Clark & Wallace, 1998) ................................ ............................. 57 1 3 Number of known jaguar subpopulations per biome, average size of subpopulations per biome, and estimated total population of jaguars per biome. ................................ ................................ ................................ ................. 58 1 4 Biome's original area, percentage of Brazil's area in each biome, area and percentage of habitat lost, area and percentage of biome remaining, total number of protected areas (conservation units (CUs) and indigenous ............... 59 1 5 Th e social context of the jaguar conservation problem in Brazil. It includes the groups of participants: small farmers and traditional communities (SFTC); large scale farmers (LSF); NGOs, research institutes, and universities ............. 60 2 1 Survey years, periods, field method used (camera trap or foot snare), area covered by trap array, and effort (reported as trap nights) ................................ 96 2 2 Effort and capture rates per method and combined ................................ ............ 97 2 3 Posterior summaries of model parameters for the jaguar surveys in Mamirau Reserve based on data from 24 jaguars. N is the number of jaguar exposed to sampling and D is the density per 100 km, is the scale .............. 98 2 4 Model selection statistics for the full set of candidate models ............................. 99 2 5 Results of likelihood ratio tests used to test hypotheses related to survival and recapture probabilities. Model parameters are transients (M2), gender (g), level of flooding (low/high flooding), time (t), constant (.) ........................... 100 2 6 Model ranking of CJS mark recapture models used to estimate apparent from 2005 2010. Only models used in model averaging with more than 0.05 .. 101 2 7 resident and transient jaguars between sampling periods. Values shown are weighted average estimates, with standard error (SE), lower (LCI) and upper 102 3 1 Characterization of all reported hunting events, and by species ....................... 142 3 2 Characteristics of hunting events by environmen t type ................................ .... 144
10 3 3 Results from Capture analysis for the estimation of total number of jaguars killed. Minimum number of animals killed, total number of captures, number of sampling occasions, bes t model selected by Capture ................................ .. 146 4 1 List of jaguar prey identified from scats, average prey body weight, estimated density of prey, abundance, biomass of prey species population available (Bio mass available=average weight of prey species x abundance) ................. 166 4 2 Weight, density, and abundance of the most important prey species in study area and the estimated consumption of each species by jaguars during the three months of low water level (September November). ................................ 168 A 1 Jaguar peer reviewed publications and book chapters produced in Brazil ....... 172
11 LIST OF FIGURES Figure page 1 1 Flowchart representing proximate causes of jaguar population decline in Brazil (light grey area) and the factors that contribute to the aggravation of these causes (darker grey area) ................................ ................................ ......... 67 1 2 Remaining original vegetation of Brazilian biomes in 2008. Data source: MMA, INPE and IBGE ................................ ................................ ........................ 68 2 1 Location of all camera trap surveys conducted to date to estimate jaguar density (white circles), Ecoregions within the jaguar present distribution (other colors), and extent of Amazonia (red line) ................................ .............. 103 2 2 Location and extent of the Vrzea floodplain forests of Amazonia ................... 104 2 3 Smaller frame shows location of Mamirau Sustainable Development Reserve within Brazil. In l arger frame red line represents the limits of the Reserve ................................ ................................ ................................ ............ 105 2 4 White uakari monkey. photo: Luiz Claudio Marigo ................................ ............ 106 2 5 Water level dynamics during the period of this study. Water level is presented in meters above sea level (masl) ................................ ................................ ...... 107 2 6 Felids sniffing the homemade lure used in this study during a preliminary came ra trap survey in the study site. Panthera onca (A), Leopardus pardalis (B), and Leopardus wiedii (C) ................................ ................................ ........... 1 08 2 7 State space area of 1,079 km determined by a 15 km buffer (black line) around the trap array used to survey the jaguar population of Mamirau Reserve (black circles), potential home range centers (green pixels). White areas represent non habitat ................................ ................................ .............. 109 2 8 Jaguar density per year with SD of the posterior ................................ .............. 110 2 9 Apparent survival rates of resident (black circles) and transient (white circles) jaguars ................................ ................................ ................................ .............. 111 2 10 Activity patterns of jaguars in Mamirau Reserve according to the number of independent photo captures recorded per one hour period of the day (n=111) 112 2 11 Number of observed inde pendent photo captures recorded per period of the day (black bars) and expected number of captures based on availability (n=111). Chi squared = 7.78, df = 3, p value = 0.05 ................................ ......... 113
12 3 1 Smaller map shows location of the study area within Brazil. Larger map shows the limits of Mamirau and Aman Sustainable Development Reserves (white lines) and the area covered during the hunting survey (white shaded area) ................................ ................................ ................................ .... 147 3 2 Area surveyed in Mamirau and Aman Reserves (white shaded area) and villages visited during survey (black circles) ................................ ..................... 148 3 3 Number of large cat hunting events repo rted per decade until the 1990s, and per year between 2000 and 2010 (n=179) ................................ ........................ 149 3 4 Number of jaguar hunting events reported per month (n=79) (designated by bars), and mean monthly water level (MMWL, designated by line) in the study area from 1990 2008 (data from Ramalho et al. 2009) ........................... 150 3 5 Number of jaguar hunting events observed and expected per season (n=187) 151 3 6 Number of puma hunting events reported per month (n=17) (designated by bars), and mean monthly water level (MMWL, designated by line) in the study area from 1990 2008 (data from Ramalho et al. 2009) ........................... 152 3 7 Number of puma hunting events observed and expected per season (n=38) .. 153 3 8 Number of jaguar and puma hunting events reco rded per season of the year and environment type (n=182 and 37, respectively) ................................ ......... 154 4 1 Location of all diet studies conducted to date (green circles), Ecoregions within the jaguar present distr ibution (other colors), and extent of Amazonia (red line) ................................ ................................ ................................ ........... 169 4 2 Smaller frame shows location of Mamirau Sustainable Development Reserve within Brazil. In larger frame red line represents th e limits of the Reserve. Dashed yellow ellipse represents location where samples were collected ................................ ................................ ................................ ........... 170 A 1 Number of peer reviewed publications related to the jaguar per year. .............. 174 A 2 Number of jaguar related peer reviewed publications per country .................... 175
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy JAGUAR ( PANTHERA ONCA ) POPULATION DYNAMICS, FEEDING ECOLOGY, HUMAN INDUCED MORTALITY, AND CONSERVATION IN THE VRZEA FLOODPLAIN FORESTS OF AMAZONIA By Emiliano Esterci Ra malho August 2012 Chair: Martin Main Major: Wildlife Ecology and Conservation The jaguar ( Panthera onca ) is the largest felid of the Americas, and its historical range has been drastically reduced in the last century to approximat ely 46% of its original size This drastic reduction and continuing decline of jaguar populations has been associated with the compound effects of habitat loss, direct killing of jaguars, and depletion of prey populations. The Amazon Forest has been recogn ized as the most important region for the long term survival of the jaguar. However, as human settlements, hunting pressure, and deforestation rates increase, the Amazon and jaguars may have their survival compromised. The Vrzea Flooded F orest is an impor tant ecosystem in the Amazonian Biome because of its rich soils and abundance of resources, but overexploitation makes it the most critically endangered environment in Amazonia. Studies previous to this one indicate that the Vrzea can have high densities of jaguar for at least part of the year and also that they can be important breeding and weaning grounds for jaguars in Amazonia but knowledge on jaguar ecology and the impact of people on jaguars in this environment are still very limited
14 In the first c hapter of this dissertation I make a thorough review of jaguar knowledge and conservation in Brazil, and apply the policy sciences approach to solving problems to understand why current jaguar conservation ac tions are not being effective in preventing jagu ar population declines in Brazil. In the second chapter I estimate jaguar population density and survival in Mamirau Sustainable Development Reserve using spatially explicit capture recapture models, and inves tigate if the co existence of jaguars and peop le inside Mamirau Reserve caused significant changes in jaguar population parameters over the course of six years of monitoring. In the third chapter I characterize the hunting of jaguars by local people in Mamirau and Aman Reserves, and estimate the to tal number of jaguars hunted in these two sites using closed population capture recaptures model and interviews. In the fourth and final chapter I investigate the feeding behavior of the jaguar in Mamirau Reserve and compare it to other environments.
15 CH APTER 1 THE JAGUAR CONSERVAT ION PROBLEM IN BRAZI L A VALUES PROBLEM The ecological and life history traits of large carnivores (e.g., low density, dietary needs, large home ranges, territoriality, and low fecundity) make them particularly prone to extinct ion, especially where intense conflict with humans for food and space result in direct persecution (Inskip & Zimmerman 2009). Understanding ecology and behavior has been the focus of most carnivore related studies in the last 2 3 decades but the need for e cological knowledge of many species in specific conservation contexts remains unfulfilled (Karanth & Chellam 2009). At the same time, there has been increasing support in the literature to the importance of understanding the human dimensions of carnivore c onservation to improve the design of conservation policies (Weber & Rabinowitz 1996; Clark et al. 1996; Kellert 1996; Clark et al. 2001; Treves & Karanth 2003; Inskip & Zimmerman 2009; Treves 2009; Karanth & Chellam 2009). Reflecting on diverse policy and management experiences Ludwig et al. (1993:36) are human problems that we have created at many times and in many places, under a variety of political, social, and economi conservation problem, local human density has been found to be strongly associated with carnivore extinctions. Direct killing by humans has been identified as the most important cause of mortality in practical ly every large carnivore studied to date (inside and outside protected areas), and the sources of current threats are virtually all anthropogenic (Crawshaw 1995; Woodroffe & Ginsberg 1998; Woodroffe 2000; Cardillo et al. 2004; Andren et al. 2006; Adams et al. 2008; Obbard & Howe 2008; Robinson et al. 2008). This has led to the general realization that large carnivore conservation will
16 only be successful if the human dimensions associated with the problem are carefully considered (Clark et al. 1996; Kellert 1996; Weber & Rabinowitz 1996; Clark et al. 2001; Treves & Karanth 2003; Inskip & Zimmerman 2009). The jaguar ( Panthera onca ) is the largest felid of the Americas, the third largest felid in the world, and the only species of the genus Panthera in the New World. The historical range of the jaguar extended from Southwestern United States to Southern Argentina (Guggisberg 1975), but has been drastically reduced since European settlement to approximately 46% of its original size (Sanderson et al. 2002). This drastic reduction and continuing decline of populations in the last century have been associated with the compound effects of human actions: habitat loss, commercial hunting (for parts or trophy), retaliatory hunting, and depletion of prey populations (Fig ure 1 1; Doughty & Myers 1971; Smith 1976; Emmons 1987; Medellin et al. 2002; Sanderson et al. 2002; Caso et al. 2008). Clark et al. (2001) describe the decision process surrounding jaguar conservation as highly fragmented, under organized, complex and ine ffective. The authors attribute these undesired characteristics to the lack of biological knowledge, the lack of a unified conservation strategy across the species range (i.e., stakeholders have different goals and problem definitions), understaffed, under qualified, and bureaucratic government wildlife agencies susceptible to political change, and the difficulty of guaranteeing long term funding. This scenario has arguably changed for the better with the increase in biological knowledge, efforts of researc hers, non g overnmental organizations, and government wildlife agencies to establish protected areas (PAs), and the proposal of national, regional, and range wide strategies for the conservation of the jaguar
17 (Appendix 1 1; Quigley & Crawshaw 1992; Sanderso n et al. 2002; Rabinowitz & Zeller 2010; Paula et al. 2011). Additionally, there has been general improvement of wildlife agencies, more stable economies, bans on commercial hunting and trade of jaguar and most prey species, and an increase of the number o f PAs in Latin American countries (FAO, 2011). These improvements, however, have had limited success in stopping the decline of jaguar populations (Sanderson et al. 2002; Caso et al. 2008; Paula et al. 2011). ent range and, therefore, has a major role to play in the conservation of the jaguar (Sanderson et al. 2002). The country has stepped up to this responsibility, pioneering ecological research on the jaguar in the late 1970s (Schaller & Vasconcelos 1978; Sc haller 1979), creating more PAs than any reviewed publications and book chapters on jaguars in the last 40 years, which represents ~39% of all research publications on the species (see Appendix for results and methodology of bibliography review methods). Despite this relatively large number of publications, human dimensions aspects have only been approached in a few studies, and have been limited in analyzing local perce ptions of people in relation to jaguars (Conforti & Azevedo 2003; Zimmermann et al. 2005; Santos et al. 2008). The jaguar conservation problem in Brazil is still being approached in a technical as ).
18 Under this solution oriented approach the jaguar conservation problem is defined by the main proximate ecological factors that are reducing populations and threatening the populations. This definition, however, is inadequate because it only identifies symptoms whose origins are not analyzed and which will vary greatly according to ecological, political, cultural, social and economic characteristics (Figure 1 1). Solving the problem goals, identifying and choosing alternatives, committing resources, and implementing The policy sciences provide a means fo r researchers, policy makers, practitioners and other conservation professionals to understand and participate in the social and decision processes pertaining to natural resources realistically, comprehensively, practically, and constructively (Lasswell 19 71; Clark et al. 2000). It provides the system of interest, and a means to clarifying and achieving the common interest of stakeholders involved (Clark et al. 2000). T his interdisciplinary problem oriented approach to solving problems has already been demonstrated for large carnivores (Clark et al. 1996; Clark et al. 2000), but has never been used in jaguar conservation. The objective of this paper is to define the jagu ar conservation problem in Brazil using the policy sciences problem oriented approach to solving problems (Lasswell 1971; Clark et al. 1996; Clark et al. 2000), to improve the understanding of the social context and decision making processes involved, to e valuate whether or not the current goal of the policy process is adequate in relation to the common interest of
19 stakeholders, and to evaluate past, present and proposed conservation actions. Additionally, based on this analysis, I suggest strategies to imp rove jaguar conservation efforts in Brazil. The Policy Sciences Method for Solving Problems The policy sciences method of solving problems is a problem oriented, contextual, comprehensive, and interdisciplinary approach that is applicable to any context wh ere people interact (Lasswell & MacDougal 1992). The method provides a conceptual framework that has four dimensions: problem orientation, social process mapping, decision process mapping and observational standpoint (Lasswell 1971; Clark et al. 2000). Emp irical data pertaining to each of these dimensions is collected, organized and analyzed to create a realistic model of the policy system of interest. Problem orientation is a strategy to address problems and create solutions. It is of stakeholders and to define the goal of the policy process. The interests of participants will vary, and defining the goal must be inclusive and encompass as many views as possible. Thi s task is usually addressed after an analysis of the social context of the problem (Clark et al. 2000). The second task is to describe the history and trends of the problem, using empirical data on the biophysical and cultural context of the problem, and a ny other relevant processes. It implies identifying the status of key elements of the problem in relation to the desired status of those elements, given the goal(s) identified in task one. The third task is a description of the physical, biological and soc ial conditions that have influenced, permitted, or caused the trends. The fourth task is to predict future trends based on present and past conditions assuming that there are no new management interventions in present conditions. The fifth and final task i s achieved by
20 comparing the predicted future trends (task four) to the desired goals of the policy process (task one), and then creating, appraising and selecting alternative strategies to make up for the discrepancy between predicted and desired states. T Wallace 1998). Social process mapping is a way of understanding any particular social contex t (Lasswell 1971). Policy sciences use conceptual categories to describe any social context (Table 1 1). The rationale is that each participant has a different perspective of the policy process and interacts with other participants in specific situations w here they use their assets (or base values), through various strategies, to achieve desired outcomes (goals), which have specific effects over the policy process and over other participants. The resulting map of the social context clarifies which participa nts are being benefitted and which are being deprived of their desires under the current policy scenario, and provides a frame of reference to understand how management actions may influence participants in the future. The decision making process is concer ned with who makes decisions and how natural resources are used. Decision process mapping is the analysis of the decision making process that is part of any policy process. It consists of seven interlinked functions: intelligence, promotion, prescription, invocation, application, termination, and appraisal (Table 1 2; Lasswell 1971; Lasswell & MacDougal 1992). By understanding the decision process decision makers can maintain good practices and correct an ineffective process (Clark & Brunner 1996). The info rmation for this work was gathered through a combination of interviews and personal interactions with stakeholders, as well
21 as reports and publications from government agencies, newspaper articles, scientific literature, and publically accessible online da tabases. resulting from personality, disciplinary training, experiences, epistemologi cal Department of Wildlife Ecology and Conservation at the University of Florida. I have be en involved in biological research and conservation of the jaguar since 2002, having specialized in the ecology of jaguars in the Vrzea Floodplain Forests of Brazilian Amazonia. In my history as a jaguar researcher and conservationist, I have transitioned from an almost preservationist perspective, where I thought jaguars and humans should be separated in space for jaguars to have a chance of survival, to a human based conservation perspective, where I can no longer see the survival of healthy natural jagu ar populations without the involvement of stakeholders and consideration of their values and goals within this policy process. Although this background necessarily inserts some bias to my viewpoints, my intention is to analyze this process as an independen t policy analyst. My main motivation for this work is the possibility of contributing to the conservation of the beautiful, ecologically and culturally important, jaguar, and simultaneously achieving of the expectations of the stakeholders involved in th is process. I believe jaguars have all of the features and historical conditions to be a model for the conservation of carnivores.
22 The Jaguar Conservation Problem i n Brazil (Problem Orientation) Definition of the Conservation Problem In Brazil the jaguar h as been classified as vulnerable to extinction in virtue of the steady decline of jaguar populations over the last century, despite large scale management actions to protect natural environments, the jaguar, and biodiversity in general, and the consistent increase of scientific knowledge on the jaguar (Machado et al. 2008; Paula et al. 2011; Appendix 1 1). Historically the jaguar occupied the entire extension of Brazil and occurred in all six major continental biomes of the country but today it is found onl y in five of the six Brazilian biomes and populations are severely reduced and isolated in at least three of these five biomes (Table 1 3; Figure 1 1). It is estimated that only 55% of the remaining natural areas are adequate to sustain jaguar populations (Ferraz et al. 2011). The main proximate causes of this decline have been habitat loss, commercial hunting, retaliatory hunting, and depletion of prey populations (Paula et al. 2011). The more than one hundred scientific publications and book chapters publ ished about jaguars in Brazil in the last four decades seem to have had limited effect on jaguar conservation (Appendix 1 1; Weber & Rabinowitz 1996; Sanderson et al. 2002; Rabinowitz & Zeller 2010; Caso et al. 2008; Paula et al. 2011). Even major national and international management actions such as the national ban on commercial hunting of wildlife in Brazil in 1967 (Brazilian Fauna Protection Law 5197/67), the international protected status of the jaguar since 1973 (CITES 1973), and the increase in the n umber of PAs (Rylands & Brandon 2005), have been unsuccessful in stopping the decline of jaguar populations in Brazil.
23 Current Goal of the Policy P rocess The current goal of the policy process surrounding jaguar conservation in Brazil has been determined by a panel of researchers, non governmental organizations and occurs in the next 1 Historical Trends and Conditions (Analysis of the P roblem) Our analysis of the jaguar conservation problem is organized in four parts that focus on the main proximate causes of jaguar decline: habitat loss, direct killing of j aguars, depredation of prey populations, and loss of population vigor. For each of these threats I identify how they influence jaguar population and what conditions have allowed these threats to persist or aggravate. Habitat Loss Habitat loss has been clai med to be responsible for major jaguar declines even in the most preserved natural areas of the most fragmented biomes of Brazil, such as the Atlantic forest (Leite et al. 2002; Mazzolli & Hammer 2008). As >37% of the natural habitats of the country have b een converted to other land uses and the rate of habitat conversion is still high in all biomes (Table 1 4), habitat loss is arguably the most pressing issue in the conservation of the jaguar in Brazil. This loss of natural habitats is the compounded resul t of a myriad of human activities or of factors that have been created or exacerbated by human actions (Figure 1 2). Brazil is divided into six continental biomes: Amazonia, the Atlantic Forest, the Caatinga, the Cerrado, the Pantanal, and the Pampa (Figur e 1 1). Biomes, as defined by the Brazilian Institute of Geography and Statistics (IBGE), are groups of plants and
24 animals constituted by continuous vegetation types identifiable at a regional scale, with similar geologic and climatic conditions, and share d history of change, resulting in a characteristic biological diversity. Until the 1900s the jaguar inhabited all biomes but a few decades ago it was extirpated from the Pampa biome (Fontana et al. 2003). The historical causes of habitat loss and the condi tions that promoted it in each biome are distinct. The Amazon forest is the largest tropical forest on our planet, covering 5,300,000 (Dirzo & Raven 2003) and a fifth o f the freshwater that runs from continents into the borders and constitutes the biome Amazonia. Amazonia remained practically intact to habitat loss until the early 1970s, but since then has experienced habitat loss at dramatic rates (Fearnside 2005). Most deforestation in this biome has been caused by large scale cattle farmers and soybean producers motivated by tax incentives, government subsidized credit, inflation (i.e., de forestation enabled claim of land and land speculation, and cutting forest for cattle pasture was the cheapest way to do it), and growth in the international market for soybean and beef (Fearnside 2005). Government investment in infrastructure such as high ways, railroads, and waterways, has also played its part, as it accelerates human migration to remote areas, increases clearing of established properties, and opens frontiers for investing timber profits in cattle ranches and soybean plantations (Fearnside 2005). Logging, including selective logging, also increases the susceptibility of the forest to fire and further contributes to habitat loss (Nepstad et al. 2004). The large portion of Amazonia that remains, >82%,
25 can be associated to the high logistical costs of exploring natural resources in this environment due to its large extension and intricate river system, to the large number and area of PAs, and more recently to efforts of the government to slow deforestation rates in the agricultural frontier are as to the south and east, known as the arch of deforestation (Figure 1 4.771/65) which requires that 80% of rural private properties in this biome be maintained in its natural state. The Pantan al is one of the largest continuous wetlands in the world covering >150,000 km of the floodplain of the upper Rio Paraguay and its tributaries. The patchy landscape of this biome is a mix of grasslands (31%), woodlands (22%), bush savanna (14%), marshes ( 7%), semi deciduous forests (4%), gallery forests (2.4%), and floating mats (2.4%; Harris et al. 2005). The main ecological factor influencing this environment is the flood pulse (Junk & Silva 1999; Oliveira & Calheiros 2000), which follows an annual, mono modal cycle with an amplitude of 2 to 5 m and duration of 3 to 6 months. The main human activity in this biome is cattle ranching and > 95% of the area is privately owned (Quigley & Crawshaw 1992; Soisalo & Cavalcanti 2006). Replacement of forest and savan na habitats by exotic grass species for cattle ranching, and burning to renew pastures, which often leads to uncontrolled fires, have resulted in most of the habitat loss in this biome (Harris et al. 2005; Alho 2008). Furthermore, cattle ranching is also b ecoming increasingly competitive, and intensive and irrigated agriculture spreading inside the floodplain are main factors of concern. Current PAs constitute less than 5% of the region and offer little help for conservation of the Pantanal (Table 1 4). The large area of natural habitat remaining in the Pantanal, >84%, is thought to be
26 linked to the difficulties of implementing extensive agriculture in a seasonally flooded environment (Alho & Lacher 1991). The Atlantic forest is one of the most highly threat ened tropical forests in the world and is the Brazilian biome with the smallest portion (12%) of natural habitats remaining (Table 1 3), of which more than 70% is private property (Leite et al. 2002). Because of the human colonization path, the Atlantic fo rest has a much earlier history of habitat loss than other biomes, which probably explains its critical condition. The conversion of habitat in this biome has been closely related to the economic exploitation of different story, such as Pau Brasil tree ( Caesalpinia echinata ) in the 16th century, sugar cane in the 18th century, cattle ranching from colonization to present, coffee in the 19th and 20th centuries, and more recently, the expansion of urban areas and eucalyptus ( Eucalyptus sp.) plantations (Dean 1997). Even today, despite severe legal restrictions on deforestation, the rate of forest loss is still high, approaching 0.25% or 350 km per year (Fundao SOS Mata Atlntica and INPE 2011). As a consequence of this long history of degradation, the Atlantic forest is highly fragmented (Figure 1 1). The Cerrado biome is the second largest biome of Brazil (Table 1 4). It covers grasslands, and gal lery and dry forests (Eiten 1977; Ribeiro et al. 1981). The Cerrado is also the second most threatened biome of Brazil with >48% of its natural habitats lost and only 2.5% of its area inside PAs. The major causes of habitat loss in the Cerrado in the last three decades have been the expansion of the agricultural frontier and increment of the production of soy, maize, and beef (Klink & Moreira 2002; Klink &
27 Machado 2005). The production of charcoal is also a major contributor to habitat loss in this biome an d the imminent expansion of sugar cane plantations is a clear threat (Carvalho et al. 2009). The growth of international markets for soybean and beef, this biome have p rovided the right conditions for this trend. In the Cerrado, differently from Amazonia, land owners are only required to maintain 20% of their properties in its natural state. The Caatinga consists primarily of xeric shrubland and dry thorn forest that cov er much of northeastern Brazil. It is the third most degraded and highly fragmented Brazilian biome with >45% of its natural habitats altered by human activities (Figure 1 1; Castelli et al. 2004). Habitat loss in the Caatinga also has a long history. The introduction of cattle and goats by Europeans in the early 1500s rapidly devastated the native plant species that lacked resistance to intensive grazing, and in the early sixteenth century most of the forests were destroyed for timber and for cattle ranchi ng, leaving mostly open scrub forest (Coimbra Filho & Cmara 1996; Leal et al. 2003). Current threats include slash and burn agriculture, which converts remnant vegetation to new and short lived cropland, harvesting of firewood, and continuous depredation of the vegetation by cattle and goat herds, which are now estimated to number more than 10 million animals (Medeiros et al. 2000). The Pampa is one of the smallest Brazilian biomes, occupying only 2.1% of the country. Grasslands, with sparse shrub and tree formations, are the dominant vegetation (Berreta 2001). Livestock production, mainly cattle and sheep, have been the main economic activity in the region. Habitat loss in the Pampa biome in the last 40
28 years has been the result of a strong expansion on ag ricultural activities, primarily due to the increase in production of corn, soybeans, wheat, and rice (Overbeck et al. 2007). The cultivation of exotic trees (e.g. for pulp production) has also increased and taken its portion of the biome as a result of in centives from both private industries and the government, and new projects will increase this area in the near fu ture (Overbe ck et al. 2007). Cultivated pastures and the introduction of exotic species of grass have also taken its toll of the natural grassl ands of this biome. Direct K illing of J aguars Commercial hunting was a critical issue for jaguar conservation in the 1960s due to an unfortunate trend in the fashion industry, which created a large demand for spotted cat skins and transformed hunting of ja guars into a lucrative occupation for rural people (Doughty & Myers 1971; Smith 1976). Professional hunters were killing approximately 15,000 jaguars per year during that period in Brazilian Amazonia alone (Smith 1976). Back then, a jaguar was worth as muc h as US $130 to hunters and in the larger regional markets, such as Belem and Manaus, skins would sell for up to US $180 (Doughty & Myers 1971; Smith 1976). Updating these values to current buying power this would be equivalent to approximately US$ 505 976 The jaguar had an easily accessible, positive, economic value to stakeholders. Today, despite their protected status (Brazilian Fauna Protection Law 5197/67, CITES 1973), illegal commercial hunting of jaguars still occurs due to market demand for huntin g jaguars as trophy (i.e., there are people that want to hunt jaguars and are willing to pay large sums of money for it), buying jaguar parts as souvenirs (i.e. pelts, skull and teeth), meat for food, and raising cubs as pets. In 2010, a group of 11 people were arrested in the south of Brazil where they organized jaguar trophy hunts in three
29 states, encompassing the Pantanal and Atlantic Forest biomes, including hunts inside Iguau National Park (R. Morato, personal comment). Their clients included hunters from Brazil, Europe and other Latin American countries and each hunt sold for US$ 1,500. Jaguar parts are also still collected as ornaments and trophies. It is common to find jaguar skulls and pelts on the walls of rural households in Amazonia and there is also a black market demand from urban centers (E. Ramalho, personal comment). Jaguar meat, although usually distributed among neighbors and family, may also be Colombia ja guar meat is sold for about US $1.5/kg (Balaguera Reina & Gonzalez Maya 2008). The capture of cubs occurs occasionally, usually during hunting of game with dogs. The mother is either killed or chased off and the cubs are kept as pets by the hunter, given o ut to neighbors or sold. In Mamirau Sustainable Development Reserve, for example, in 2004 a female jaguar cub was sold by a riverside tradesman for less than US $20 to a local farmer (E. Ramalho, unpublished data). This cub was raised into adulthood in th the age of five. Cultural historical motives are also a large contributor to people killing jaguars. As the largest terrestrial predator in Brazil, capable of taking prey much larger than them, including humans, jaguars have historically been feared and killed by indigenous people and rural stakeholders. For indigenous people the jaguar has a multitude of cultural, cosmological and ecological meanings that are not negative, although t hey can be related to fear (Whitehead & Right 2004). For some indigenous people, however, this special relationship does not necessarily affect the decision of killing a jaguar if it is felt
30 that the jaguar is threatening a tribe member, or to prevent futu re livestock depredation (P. Constantino, personal comment). It is difficult to assess the impact of indigenous communities on jaguar populations before colonization, but these communities probably exerted small hunting pressure on wild cats. Jaguars were occasionally killed for cultural rituals but were generally feared and respected (Smith 1976). Despite the effort of colonizers to dissociate indigenous cultures to the jaguar, the perception and cultural importance of jaguars is unlikely to have changed ( Fausto 2004). The jaguar has historically been revered by these cultures as symbols of power and beauty (Saunders 1998; Luna & Amaringo 1999; Whitehead & Right 2004). However, as indigenous communities evolve within the contemporary world and change their economic activities, it is expected that their impact on wildlife, including jaguars, may change. Livestock, for example, was not a subsistence activity for indigenous communities before colonization, who therefore had no motive to kill jaguars in retaliat ion for livestock losses. After the introduction of livestock, it is likely that indigenous farmers kill jaguars to prevent or retaliate for depredation in the same way that traditional livestock farmers do. Their impact on jaguars today will be associated with livestock depredation. Even after European settlers arrived, hunting pressure on jaguars and other spotted cat populations in Amazonia was relatively small and concentrated around urban centers, agricultural frontiers, and small human settlements tha t were sparsely distributed along rivers in the Amazon basin (Smith 1976). Although there are no estimates of harvest rates for this period, the impact of hunting was likely small because demand for spotted cat skins was relatively low and consequently the re was no economic incentive to pursue these animals. Furthermore, human density was
31 low and hunting was locally aggregated, leaving large portions of the jaguar population under little to no pressure. Consequently, there was a large source and few sinks. The arrival of European settlers to the new world probably also brought along their deep rooted fear of large carnivores and their cultural bias towards eliminating predators (Clark et al. 1996). Although jaguar attacks on people have been recorded in man y areas throughout Brazil (CENAP, unpublished data) they are rare events and usually related to people approaching jaguars deliberately or by accident, mostly caused by animals cornered during hunting (Almeida 1976; Paula et al. 2008). The first official r ecord of a predatory attack on a human occurred in June of 2008, in the Pantanal biome (Paula et al. 2008). In this case a fisherman was attacked and killed while sleeping in a tent in on the banks of the Paran River, Mato Grosso state. His body was carri ed a couple hundred meters and he was partially eaten. Different from other large cats, there are no reported cases of jaguars that have developed man eating habits, but Paula et al. (2008) highlight that if habituation of jaguars to people (i.e., using ba its for tourism) continues to occur, predatory attacks on humans may become an issue. On top of this cultural import, contemporary rural stakeholders in general have a negative perception of the jaguar associated to real or perceived negative effects that jaguars may have on their livelihood, mainly: predatory attacks on people, economic losses due to depredation of livestock and dogs, and competition for game (Conforti & Azevedo 2003; Zimmerman et al. 2005; E. Ramalho, unpublished data). Negative symbolism associated with large carnivores in general, such as viciousness and ferociousness, also contribute to this perception (Leopold 1949; Kellert 1991; Kellert et
32 al. 1996). These perceptions often lead to negative attitudes towards jaguars, where most of rur al stakeholders do not support, or want no part, in jaguar conservation, and where the ultimate result is the persecution and deliberate killing of jaguars (Carvalho & Pezutti 2010; Hunting chapter). The overlap between the diet of the jaguar and that of s ubsistence hunters (Jorgenson & Redford 1993) and the consequent deduction that jaguars deplete prey populations also motivates rural stakeholders to kill jaguars. Jaguars are also killed for pleasure (trophy/sport hunting), status, or both, and bounties a re still offered in most areas where jaguars kill livestock. This retaliatory killing of jaguars, as a form of control of depredation of domestic animals, is one of the main factors contributing to jaguar population decline in Brazil (Crawshaw 2003). The p erceived value of jaguars to stakeholders has been reported to be very distinct between biomes (Santos et al. 2008), age group, and rural and urban populations (E. Ramalho, unpublished data), and will also probably vary according to social, cultural, histo rical and economic factors. Regardless of the predominantly negative view of most rural stakeholders towards jaguars (Conforti & Azevedo 2003; Zimmerman et al. 2005), there has been increasing support from the general public to support the conservation of the jaguar. Positive perceptions and attitudes towards jaguars can be attributed to positive symbolism associated with large carnivores, such as beauty, strength, intelligence, courage, and endurance, or a general affection for nature, understanding of the ecological role of large predators, or moral and ethical beliefs, as observed by many authors (Leopold 1949; Lopez 1978; Rolston 1981; 1985; Kellert 1985, Kellert et al. 1996). These positive values, however, are difficult to quantify economically and are usually ignored or undervalued (Bishop 1978; Usher 1986;
33 Rasker & Hackman 1996), while negative values associated with livestock depredation are easily measured by stakeholders and therefore receive more attention. The consumption of jaguar meat is not un usual among rural communities in Amazonia (E. Ramalho unpublished data) and the Atlantic Forest (Rocha Mendes et al. 2005) biomes, and has been reported in the Colombian Choc (Balaguera Reina & Gonzalez Maya 2008), but we found no recent records of these events in other Brazilian biomes. However, even in Amazonia, jaguars do not represent an important food resource and hunters will seldom, if ever, go out of their household with the intention of hunting a jaguar to eat. Because jaguars occur in low densiti es, are difficult to track, and are dangerous to hunters and dogs, actively hunting them for food is not cost effective. The consumption of jaguar meat is usually associated with hunting of jaguars for other motives or during chance encounters (e.g., durin g fishing expeditions) and is driven by the protein needs of rural dwellers whose main source of protein is fish and game. Decrease of P rey P opulations The jaguar is an opportunistic predator with a rather flexible feeding ecology, consuming over 85 differ ent species of prey, from snakes to tapirs ( Tapirus terrestris ; Seymour 1989). However, in most environments studied to date, jaguar populations seem to depend on medium to large sized terrestrial mammals to survive (Novack et al. 2005). This dependence m akes them vulnerable because medium and large sized terrestrial mammals are less resilient to habitat loss, and because these animals are also the preferred game species of subsistence and commercial hunters (Jorgenson & Redford 1993; Robinson & Bennett 20 00). Prey depletion by subsistence hunting has been pointed as a major threat to jaguar survival range wide (Emmons 1987; Sanderson et al. 2002) including in Brazil (Guix 1997; Leite & Galvo 2002).
34 Subsistence hunting is a critical activity for indigenou s and rural communities outside urban areas because wildlife is a major source of protein and fat. At the same time, subsistence hunting has been considered the main cause of wildlife population declines in Latin America (Redford 1992), and has increased i n recent years as the result of human population growth, easier access to undisturbed natural habitats, improvement of hunting technology, and scarcity of alternative protein sources (Robinson et al. 1999). The increase of rural populations is usually foll owed by a maximize immediate harvest success instead of long term conservation goals (Stephens & Krebs 1986; Robinson & Redford 1991; Alvard 1993). Commercial hunting, although p resently illegal in Brazil, also contributes to defaunation, as there is demand for bushmeat inside communities and local markets, and law enforcement is scarce. Currently, one of the most important conservation issues for jaguars in Amazonia may be the im plementation of the commercial harvest of black caiman ( Melanosuchus niger ), especially in the varzea floodplain forests where black caiman eggs constitute an important food source for jaguars (Ramalho 2006; Ramalho & Magnusson 2008, Silveira et al. 2010). The depletion of prey populations may also contribute to generating more conflicts between rural stakeholders and jaguars by increasing instances of livestock depredations (i.e., reducing availability of prey to jaguars may result in more livestock depred ation and more direct killing of jaguars to protect livestock. Jaguar Population V igor Analyses of the genetic structure of the jaguar have concluded that there has been historical connectivity between jaguar populations across broad geographical areas, wi th few barriers to gene flow on a continental scale (i.e., the Amazon river, the Andean
35 mountain chain, and a apparent barrier in Central America; Eizirik et al. 2001; Ruiz Garcia et al. 2006). However, reduction and isolation of jaguar populations as a co nsequence of all the threats described in the previous sections contributes to the decrease of genetic diversity of subpopulations within biomes, as well as drift induced differentiation among local fragments, as shown by Haag et al. (2010) for the Atlanti c Forest biome. This reduction of genetic diversity due to inbreeding depression, has been shown to have harmful effects on development, survival and growth rate of species in captivity and in the wild, and may leave small jaguar populations at the mercy o f stochastic forces that lead to extinction: demographic, genetic, environmental and catastrophes (Schaffer 1983). Future P rojections If the current conditions persist jaguar populations will continue to decrease in Brazil, and the jaguar will eventually b ecome extinct in more Brazilian biomes. The most threatened jaguar populations are in the Atlantic forest and Caatinga, where subpopulations are small (in both biomes subpopulations average less than 40 individuals, and total population is less than 200 in dividuals Table 1 3), isolated (i.e., biomes are largely fragmented), and poorly protected (not enough parks and inefficient law enforcement). The causes of this decline will continue to be habitat loss, retaliatory hunting, and depletion of prey populat ions. The Social Context We have identified six major groups of stakeholders directly involved in the jaguar conservation problem in Brazil (Table 1 5). Small to medium sized farmers and traditional communities (SFTC) include people that live in rural are as with properties
36 <900 ha, and <500 heads of cattle or other livestock herd (this is the classification used by IBGE www.ibge.gov.br). They may be single families, a community of families, or a tribe. They are usually poor and live inside or in the vici nity of jaguar habitat. Their livelihoods may be directly affected by a jaguar (i.e., depredation of livestock, attack or perceived threat of attack on humans), and/or their livelihoods depend on jaguar habitat (i.e., extractivist activities, converting na tural habitats for pasture or plantations) and/or prey. Large scale farmers (LSF) include livestock farmers and crop producers with properties >900 ha and/or >500 heads of cattle or other livestock herd (this is the classification used by IBGE). These stak eholders are usually wealthy and/or politically powerful agricultural businessman, and national or international corporations (i.e., Monsanto). Non governmental organizations, research institutions, and universities (NRU) include researchers, conservationi sts and their funders. Government wildlife agencies (GWA) represent the Brazilian government environmental agencies directly involved in the jaguar policy process: ICMBIO, CENAP and IBAMA. The general public (GP) includes the national and international urb an populations that do not interact with jaguars on a daily basis or never interact. Trophy hunters and outfitters (THO) are all sport hunters and outfitters, and any other individual or organization, involved in the activity of hunting animals for sport. They are currently prohibited by law to exercise this activity in Brazil, with the exception of a few private properties in the south of the country which have obtained special permits to hunt specific game species, but not the jaguar. By mapping the socia l milieu of jaguar conservation in Brazil (Table 1 5), I observed that stakeholders can be further grouped into two categories: direct interaction
37 stakeholders (DIS), and indirect interaction stakeholders (IIS). DIS are people whose livelihoods involve, or depend, on the direct interaction with jaguars, their natural habitat, and their prey. Three stakeholder groups fit these attributes: SFTC, LSF, and THO. Their goals involve improving their livelihoods of at least three of these four base values: wealth, well being, power, and respect. As strategies to achieve their goals SFTC and LSF kill jaguars to prevent future loses of livestock (wealth less economic losses to depredation equals more profits) and/or potential jaguar attacks on local people (well bei ng people feel safer). They also hunt legally for subsistence (well being wildlife is an important source of protein for many SFTC), or illegally, for commercial or recreational purposes (wealth and well being some people profit economically from sel ling wildlife, which is most cases is complementary subsistence activity for SFTC; others hunt for leisure), and convert natural habitats to pasture and croplands, legally, and illegally, to increase agricultural profits (wealth). They use their political power to pressure the government for more management rights over natural resources (power and respect they demand legal rights to hunt wildlife commercially and to be able to convert larger areas of natural habitat within their properties). The goal of T HO is to have the right to hunt wildlife (power), including the jaguar, because it is an activity that gives them pleasure (well being), and develop hunting enterprises, because they can generate profit (wealth). Because none of these activities are curren tly allowed by law, some of them hunt or promote hunting illegally in their property or in state lands, taking advantage of the incapability of the government to enforce the law. The strategies of DIS (i.e., converting natural habitats to pasture or cropla nds, illegal killing of jaguars, and over exploiting or illegal hunting of prey populations) are considered to
38 be the main proximate causes of the declining trend of jaguar populations, habitat, and prey populations. IIS are stakeholders whose livelihoods do not generally involve, or depend, on the direct interaction with the jaguar, their habitat, or their prey. The other three stakeholder groups fit these attributes: NRIU, GWA, and GP. Their main goal is to reverse the current trend of decline of the jag uar population of Brazil. Most of these stakeholders have pleasure in knowing jaguars still exist, that they are protected in the wild, and will be around for the next generations to appreciate (well being). And some of them, mainly NRIU and GWA, also unde rstand the ecological and cultural importance of the jaguar. Their strategies, however, are mainly coercive, restricting management rights of natural resources for DIS (i.e., ban on hunting of jaguars and wildlife, creation of reserves, restricting the por tion natural areas within private properties that can be converted to other land uses) without giving stakeholders alternatives to compensate restrictions. These strategies are guided by a technical rationalist biological rationale that does not take into consideration the goals and value demands of DIS. This discrepancy between the goals of DIS (who are assumed to be responsible for a large part of the jaguar conservation problem) and the strategies of IIS seem to be a central obstacle for the effective solution of the problem, since the strategies of DIS only benefit DIS, and the strategies of IIS only benefit IIS, both depriving the other stakeholders group of achieving their goals. The question is now: who is going to be the bigger man and change stra tegies to encompass other stakeholder goals and value demands?
39 The Decision Process Research and Planning (Intelligence Function) biology and conservation have been th e responsibility of NRU and GWA. Since the 1970s, when the first field studies on jaguar ecology were conducted in the Pantanal biome (Schaller & Vasconcelos 1978), considerable advances have been made with regards to scientific information on the jaguar. Our review of jaguar scientific literature in Brazil resulted in 145 research publications, including peer reviewed publications, thesis and dissertations, books, and book chapters (see Appendix 1 1 for details on literature review methods). Nonetheless, i mportant scientific information to guide management decisions is lacking in all biomes. The ecology and behavior of the jaguar is still poorly understood in Brazil. Diet has been the ecological aspect most studied to date, being the subject of research on 17 (48.6%) of the 35 studies on jaguar ecology and behavior. But research on jaguar feeding habits has been concentrated in the Atlantic forest and Pantanal biomes, and is scant in the other three biomes. Scientific information on movement, home range siz e, and habitat use, and on populations parameters and structure are have only been conducted to some extent in the Pantanal and the Atlantic forest. In the other biomes home range sizes are unknown and jaguar density has only been estimated in one or two s ites per biome, being difficult to estimate population sizes. Most studies that involve jaguar conservation propose actions to improve the status of jaguar populations (15; 55.6%), but there are no studies in any of the five biomes that have actually empir ically tested a conservation strategy proposed. Status and distribution studies are abundant in the Atlantic forest biome (7; 25.9%) but are
40 practically non existent everywhere else. And, despite the drastic impact of habitat loss on jaguar populations, on ly two studies involve evaluating the impact of habitat loss on jaguars in Amazonia and the Atlantic forest biomes. The conflict between human and jaguars has also received smaller attention than expected based on the impact of this interaction on the jagu ar (22; 20.2%). The impact of jaguar livestock depredation has concentrated in the Atlantic forest and Pantanal biomes (4 studies in each), but has also been studied in Amazonia and the Cerrrado. These studies however use different methods and units, which make them hard to compare. Direct hunting of jaguar is another crucial, yet neglected topic. To this day there is only one study which has actually tried to estimate the number of jaguars killed by local people, and that was in Amazonia (Carvalho & Pezutt i 2010). The human dimensions of the jaguar conservation problem have only been approached in three publications, all of which looked at local perceptions about the jaguar (Conforti & Azevedo 2003; Zimmerman et al. 2005; Santos et al. 2008), but not at the ir goals, values demands, or interactions with other stakeholders. Planning activities have been realized at a national level by NRIU and GWA. In 2007, NGO Jaguar Conservation Fund (JCF) organized the first national meeting of jaguar researchers in Brazil and in 2009, the Brazilian government initiated a jaguar meeting, organized by the Brazilian government agency CENAP (National Center of Research and Conservation of Mammalian Carnivores)/ICMBIO (Chico Mendes Institute of Conservation of Biodiversity) in partnership with NGO Panthera, and supported by NGO Instituto Pr
41 Conservation of Nature) Cat Specialist Group and the Cons ervation Breeding Specialist Group (CBSG), brought together 37 researchers and policy makers that study or have studied jaguars in Brazil with the objective of evaluating the current status and trend of jaguar populations in the country, and the production of a National Action Plan for the conservation of the species. Recommending and Debating Policies (Promotion Function) Recommending and debating policies related to jaguars has only been done at local scales (i.e., inside a few protected areas), but has n ever been done at regional or national scale. I am aware that it is unrealistic to imagine a meeting of all groups of stakeholders identified, from all biomes, at one location, at one point in time, to discuss and decide on alternative policies to solve th e jaguar conservation problem in Brazil. But what I have shown by mapping the social context is that the goals, values demands, strategies, and interactions of stakeholders involved can, at least initially, be represented by information from scientific lit erature or professional experience, giving a much clearer view of how different policies will affect stakeholders and the policy process as a whole. Although this social map is a model, it improves the decision process and provides a frame of reference for adapting to more specific situations where policies to recuperate jaguar populations need to be implemented. Creating, Implementing and Enforcing Rules (Prescription Invocation a nd Application Functions) Creating rules at regional and national scales (i. e., laws) is a task of the government and its wildlife agencies, but this function has been historically limited by a lack of scientific knowledge on important aspects of jaguar ecology in all biomes, as show in the intelligence function section. Until ver y recently, almost all management
42 actions that contributed to the conservation of the jaguar in Brazil had not been specifically designed for the conservation of the jaguar. The only formal management action that was created, implemented and enforced speci fically designed for jaguars was the international ban on hunting and trade of jaguars, enacted by the inclusion of the jaguar in appendix 1 of CITES (1973). Interestingly, this action was created and implemented before any solid scientific information on jaguars was available in Brazil or anywhere else. The National Action Plan for the conservation of the jaguar in Brazil (Paula et al. 2011) is the first recovery plan designed for the jaguar in Brazil. The action plan contains a valuable and unprecedented compilation of information on the status, trends and threats to the jaguar, its habitats, and its prey, in all 5 biomes where the species exists in Brazil. It also compiles a prioritized list of conservation actions proposed by the participants to revert t he declining trend of jaguar populations in each Brazilian biome. Building corridors to connect jaguar sub populations, and a formal recognition of the jaguar as a natural symbol of Brazil by the Brazilian government, are the only two current conservation actions that are being formally undertaken by the GWA and NRIU to specifically address the conservation of the jaguar in Brazil. Corridors are under implementation in the Atlantic Forest biome and the Caatinga biomes in Brazil, led by the NGO Institute of Ecological Research (IP) and CENAP (R. Morato, personal comment). In both cases implementers have involved local stakeholders through public meetings, and have taken into consideration their goals and value demands to improve the chances of success of the corridors. In all these cases, however, stakeholders are consulted after the conservation action has already been decided by GWA and NRIU.
43 Meetings only have the function of adapting the strategy to the demands of local stakeholders. Evaluation and Termin ation of Rules ( Appraisal and Termination Functions) Although it is commonly accepted that protected areas, and the bans on hunting and trade of wildlife, have had a positive impact on the conservation of jaguars, natural habitats, and prey populations, th ere has been no quantitative evaluation of the impact, or other measure of success, of these strategies on jaguar populations. How many jaguar where killed before the ban on hunting versus how many jaguar are killed today? Have protected areas had a positi ve impact on jaguar and prey populations? What is the impact of subsistence hunting and illegal hunting on jaguar and prey populations inside and outside protected areas? Because evaluation of conservation measures have not been promoted it is difficult to determine which rules to terminate. Current Solutions, Recommended Solutions, and Alternative Solutions Bans on Hunting and Trade of Jaguars and Prey Bans on hunting of wildlife in Brazil in 1967 (Brazilian Fauna Protection Law 5197/67) and the inclusion of the jaguar in appendix 1 of CITES, banning hunting and international trade of jaguar parts in 1973 (CITES 1973), have had a substantial impact on commercial hunting of jaguars in Brazil, and are thought to have reduced the number of jaguars killed in B razilian Amazonia by half (Smith 1976). However, it is nave to believe that these bans can effectively protect jaguars in Brazil given that GWA do not have enough staff or financial resources to regulate direct killing of jaguars for commercial or other m otives. Additionally, the diversity of municipal, state, and federal
44 competencies of GWA, under different economic, social, and political pressures throughout Brazil, make it even harder to enforce these bans (Crawshaw 2003). Although these strategies help ed address the killing of jaguars for commercial reasons, the ban on hunting also reduced the jaguar to a zero or negative social economic value to rural stakeholders (SFTC and LSF) because they cannot profit from the commercial harvest of jaguars anymore; however, at the same time, they still feel threatened physically by jaguars and have a financial burden from livestock losses from jaguar depredation. Today, this negative social economic value associated with the jaguar is the most frequent motivation of SFTC and LSF for killing jaguars in Brazil and other countries, wherever livestock farmers and jaguars coexist (CENAP, unpublished data; Sanderson et al. 2002). This negative social economic value is created by the proximity of jaguars and people, associa ted with anthropogenic imbalances in the environment (i.e. habitat loss, decrease of prey populations), lack of information about the species, natural variations in prey availability, and poor management of livestock (especially calves), all of which usual ly lead to jaguars approaching properties and killing livestock. This negative value is easily estimated by the stakeholder (i.e., the value of the livestock), and because stakeholders have to cope with the loss themselves, they choose the cheapest and fas test solution to stop and prevent depredation future depredation, which is to kill the jaguar. The ban on hunting of wildlife also decreased the pressure on jaguar prey populations, but outside PAs animals are hunted almost indiscriminately, and often insi de PAs too (Leite & Galvo 2002), due to the limitations of GWA in enforcing the law. The impact of subsistence hunting on wildlife is also of major concern and it is still
45 highly controversial if subsistence hunting is sustainable or not (see section on d epletion of prey populations). Protected A reas So far, the main strategy of the Brazilian government to mitigate habitat and biodiversity loss has been the creation of PAs (Peres 2005; Silva et al. 2005). PAs in Brazil have functioned as effective barriers to habitat loss (Silva 2005), and remain a corner stone of conservation worldwide, being credited with saving wildlife populations from regional and range wide extinction (Terborgh et al. 2002; Woodroffe & Ginsburg 1998), despite deficiencies in managemen t and implementation, and criticism for imposing societal goals on local people (West & Brockington 2006). Indigenous territories (IT) have also contributed to the conservation of natural habitats, especially in Amazonia, where they encompass over a fourth 4). Unfortunately, this relative success has not been enough to protect natural habitats, jaguars, and prey population because many PAs only exist on paper and most have inadequate law enforcement. Additionally, PAs do not pro tect large portions of most biomes (Table 1 4), and most habitat loss is expected to occur in private properties outside PAs (Soares et al. 2006). Furthermore, the actual success of this strategy in protecting jaguar and prey populations is controversial ( Chapter 3), although local people and researchers frequently report higher abundances of both inside PAs (E. Ramalho, personal comment). To prevent unnecessary and unwanted habitat conversion outside PAs, the Brazilian government created the Brazilian for est code (Law 4.771/65) in 1965, a federal law that determines the extent and specific areas of a private property that must be maintained in natural state. These areas are denominated permanent protected areas
46 (APPs). The extent of a private property that must be assigned as an APP varies from 20% in the Cerrado, to 80% in Amazonia. Riparian areas along waterways must be protected, independently of biome, but with varying extents relative to the width of the waterway (e.g., 30 m for streams narrower than 1 0 m). This legislation has also been successful to some extent, but compliance with minimum legal requirements are highly variable (Resque et al. 2004) and difficult to enforce due to poor land titles management. Of great concern for jaguar conservation is the current proposal of Senator Aldo Rebelo, representative of agricultural producers, to change the Brazilian forest code. The core of his proposal contends giving amnesty to landowners that have destroyed natural habitats illegally (i.e., over the allow ed limits as explained in the previous paragraph), and establishes new, less restrictive, rules for determining APPs. His proposal has already been approved in the House of Representatives by a great majority of deputies and is soon to be voted on in the Senate. While it has been acknowledged by all sides of this debate that the forest code needs to be updated, it is imperative that this legislation is not changed to allow larger portions of private properties to be converted to other land uses, as this wo uld result in a accelerated reduction of natural habitats in all biomes, and may seal the fate of small jaguar populations in the most fragmented biomes of Brazil: Atlantic forest, Cerrado, and Caatinga. Better Management of Livestock Improvement of lives tock management practices has been cited in numerous publications as an effective and inexpensive way to reduce livestock depredation by jaguars, and, consequently, human jaguar conflict (Quigley & Crawshaw 1992;
47 Crawshaw & Quigley 2002; Hoogesteijn et al. 2002; Azevedo & Murray 2007). Suggestions include concentrating births in a shorter period of time to allow better management and protection of calves, and maintain more vulnerable age classes away from areas of higher predator occurrence (Crawshaw 2004); moving cattle herds away from jaguar core areas (Azevedo & Murray 2007). Improving livestock management practices has been shown to be a successful strategy to reduce livestock depredation by jaguars and it is even suggested that private farms, with adequ ate management, could be successful wildlife sanctuaries (Hoogesteijn & Chapman 1997; Hoogesteijn et al. 2002). It is also cheaper than other methods of reducing livestock depredation, like the techniques for adverse conditioning of predators described bel ow. But the issue is that being cheaper does not mean that livestock owners will agree to do it, or comply to do it. The number of farmers actually willing to change their management practices to prevent depredation is, unfortunately, very small. The simpl e reason is that it is even cheaper, less time and energy demanding to kill jaguars than it is to change management practices. We can understand this very easily by making an analogy to urban stakeholders. What is the reaction of most people from larger ci ties when they are asked to not use their car because of global warming? Even if you give the people a reasonable alternative public transportation system, it still implies leaving your house earlier, having to walk to the station, stay in line, buy a tick if they have the option of driving. Techniques for Adverse Conditioning o f Predators Different methods for reducing the frequency of livestock depredation through adverse conditioning have been tes ted in Brazil, such as electric fences, nauseating
48 substances put in carcasses of depredated livestock, toxic collars, electronic devices with strong lights and loud sounds, dogs and llamas to guard sheep, and fireworks, but the high cost of most of these have been prohibiting to most livestock owners (Crawshaw 2004). Corridors The solution that is in vogue for jaguar conservation at national (Leite et el. 2002; Cullen 2006; Haag et al. 2010) and range wide scales (Rabinowitz & Zeller 2010) is the creation of corridor of habitat to connect jaguar subpopulations. This strategy has been was proposed by IBAMA in 1996 (Ayres et al. 1997), but only recently has it started to be applied specifically for jaguar conservation. Corridors are currently under implementa tion in the Atlantic Forest and the Caatinga biomes in Brazil, led by the NGO Institute of Ecological Research (IP) and CENAP, respectively, and in Central and South America by NGO Panthera. Theoretically, corridors allow the exchange of individuals betwe en patches of natural habitat, facilitate gene flow between subpopulations and reduce chances of stochastic extinction (Fahrig & Merriam 1994), as well as the potential for deleterious genetic effects resulting of inbreeding depression (Brown et al. 2004). However, not only the effectiveness of corridors in facilitating animal movement between habitat patches remains controversial (Rosenberg et al. 1997; Beier & Noss 1998; Bennett 2003), but the financial, political, and logistical viability of using corrid ors as a single species conservation strategy over large scales, such as Brazil, or continental and multinational scales, such as Latin America, has never been evaluated for large carnivores. In fact, Cullen et al. (2005), in simulations of the viability o f jaguar subpopulations in the Atlantic forest biome, show that corridors may have a negative
49 effect on connected subpopulations if these subpopulations are not effectively protected. This is worrisome since most PAs in Brazil are ineffectively protected ( Soares et al. 2006). Translocations Translocation of jaguars serves the same purpose of corridors, that is, to facilitate gene flow between subpopulations and reduce chances of stochastic extinction (Fahrig & Merriam 1994), as well as the potential for del eterious genetic effects resulting of inbreeding depression (Brown et al. 2004). The few cases of jaguar translocation described in literature have reported translocated individuals being killed shortly after release (Rabinowitz, 1986; Crawshaw, 1995). Oth er attempts in Brazil have been inconclusive due to inadequate monitoring after release (Crawshaw, 2003). On the other hand, experiments with pumas ( Puma concolor ) in the United States indicate that translocations may be successful with sub adult individua ls in dispersion age, as these animals have a higher probability of remaining at target site if conditions are favorable (Crawshaw, 2003). Compensation S chemes Compensation schemes have been applied unsystematically, and informally, in a few locations in Brazil, namely in the Pantanal and Cerrado biomes. The rationale of this strategy is that by compensating rural people from losing livestock from jaguar depredation that these stakeholders will agree, and comply, with not killing jaguars. To a certain exte nt this line of thought is adequate in relation to some human values. When rural people lose livestock they lose wealth, and by financially compensating them for their loss you give them back the wealth they lost, and they go back to the status quo, as if there was no jaguar attack. Therefore, there is no more reason to kill jaguars. In
50 none of these cases was there was any scientific evaluation of the effectiveness of this strategy, but its discontinuity in Brazil and lack of scientific support indicate in efficiency or inadequacy as a strategy to reduce conflicts between SFTC and LSF, and jaguars. Farmers consulted by Crawshaw (2003), when asked about solutions to the livestock problem, cite financial compensation for losses as one of the preferred methods to deal with livestock depredation by jaguars. This is a logical and have to spend more money, more time, or more energy) in order to avoid economic losses. However, Cra wshaw (2003) points to a few caveats related to this strategy. First, if the strategy is implemented by the government at a local level it will generate a justified dissatisfaction of other stakeholders in similar situations in other areas, which could imp licate in more antipathy for jaguar conservation and continuation or increase of illegal control of jaguars. Second, funding to cope with compensations must be self sustainable or it will be inevitably doomed to failure. Finally, there must be a multi inst itutional technical body to attest the veracity of declared depredations and application of compensations that, given the size of continental dimensions of Brazil and the small staff of GWA, seems like an unattainable task. Ecotourism Dalponte (2002) sugge sts a program that integrates research, education and tourism. The main obstacle for jaguar related ecotourism is the sightings themselves, which are have only been shown to be frequent enough to allow tourism in some areas of the Pantanal. Studies have to be conducted to evaluate the viability of tourism. Does it generate enough profit to compensate depredation losses, change livestock
51 management practices, or change stakeholder activities from livestock farming to ecotourism ? Sports H unting Hunting large carnivores as a strategy to maintain populations at target levels, reduce losses to local stakeholders, and build public support for carnivore conservation has been found to be largely lacking in support from scientific data (Treves 2009). There are also n umerous arguments against hunting of large carnivores based on ethical, functional and economic grounds (Rutberg 2001; Knight 2003; Peterson 2004; Campbell & Mackay 2009). There have been, however, non experimental attempts to use research captures as mean s of generating funds for research and conservation, although these are undocumented. Crawshaw (2003) mentions that a frequent solution proposed by farmers in Brazil (anim als that have acquired the habit of eating livestock). Despite the inevitable vociferous opposition from a large part of society (i.e., WGOs and GP) he believes that his option should not be discarded without well controlled experiments of its efficacy, as there is plenty of scientific support showing the efficiency of sport hunting as a tool to man a ge wildlife. Hunting of prey, on the other hand, has been used with success, as conservation strategy for jaguars in Mexico (Rosas Rosas & Valdez 2010). Their s trategy involved the creation of an economic alternative for local farmers, in this case the commercial sports hunting of white tailed deer ( Odocoileus virginianus ), in exchange for the support of the farmers in not killing jaguars in retaliation for lives tock losses.
52 Environmental Education Environmental education has also been proposed as a conservation action in many publications (Crawhaw 1995; Dalponte 2002; ) but the success of this action is seldom measured. Discussion Because jaguars are large and ecologically sensitive predators with extensive area requirements, it is unlikely that PAs will be enough to conserve viable jaguar populations in the long run, unless jaguars can move and survive outside the boundaries of PAs (Soul & Noss 1998). Hoogest eijn et al. (2002) suggest that informal protection, stakeholders accepting jaguar within their properties, may be the most important factor in jaguar conservation. Through a distinct approach, I come to a similar conclusion. The comprehensive problem ori ented approach used in my analysis of the jaguar conservation problem in Brazil allowed me to observe that there is a disconnect between the goals of DIS and the actions proposed, and rules prescribed by GWA and NRIU. I contend that this is the result of a traditional technical rationalist approach to conservation that only views the conservation problem through its proximate causes, but pays little attention to the social context and the decision process of the problem. In the case of the jaguar, although researchers and managers generally acknowledge the importance of DIS in the conservation of jaguars, they usually view DIS as a means to an end, rather than integral participants of the policy process, having goals and value demands that should be accounte d for. I believe that, to a large extent, this is the reason why jaguar conservation in Brazil has been ineffective.
53 As a first step to improve the jaguar policy process in Brazil, I suggest a more comprehensive goal, one that would appeal to all stakehold ers instead of just a portion of them. The current goal, as defined by Paula et al. (2011), does not take into consideration the goals and values of the DIS, who are responsible for most habitat loss, illegal killing of jaguars, and depletion of prey popul ations. It is no surprise, but rather an expected human behavior, that they do not want to participate or collaborate with stakeholders responsible for management, NRIU and GWA, in the conservation of the jaguar. Policy makers must understand that it is no t enough to acknowledge that DIS must participate in jaguar conservation, as has been done extensively in literature (Weber & Rabinowitz 1996). If jaguars are to survive in the long run it is imperative that managers genuinely understand that their goals a nd values are as valid as any of the other stakeholders, and should be incorporated into management knowledge before policy makers define goals, suggest actions, and prescribe rules. Based on this rationale I suggest an alternative goal for this policy pro cess: to reverse the trend of where the species still occurs, and, at the same time, reduce and/or compensate economic losses and threats of attacks to stakeholders who interact with jaguars on a daily basis, and empirically evaluate competitive alternative economic uses of natural resources (i.e., habitat and wildlife) to substitute inappropriate natural habitat conversion or retaliatory killing of jaguars. The poli cy process for jaguar conservation is still incipient, and, more importantly, knowledge deficient. Our analysis of jaguar literature in Brazil shows that jaguar ecology is poorly understood, there have been no management experiments to evaluate
54 proposed co nservations actions, or strategies already implemented. There needs to be more research in all biomes aimed at filling gaps in knowledge necessary for the decision process. The National Plan for the conservation of the jaguar is an important step in gather ing this information and orienting future research, but it remains to be seen if goals will be completed and orientations will be followed. Evaluation of conservation actions need to be put into practice as soon as new actions are implemented and current e fforts to create corridors should be carefully monitored in relation to economic and time costs, versus effectiveness. Also, other alternatives should be evaluated, especially those that include the goals of DISs, and not only those of IIS. Strategies are being proposed for large scales without trial runs. After a proposal of strategies, it is the duty of those that have proposed it to test it, before making it a large scale prescription. It is not fair for society to demand from SFTC and LSF to cope with losses due to depredation by jaguar on their own (Crawshaw 2003), and if we do that we cannot criticize the strategies they use to deal with those losses. There is increasing understanding among DIS of the importance of conserving jaguars and the will to support it as long as the damage caused by jaguars is solved or at least reduced (Crawshaw 2003). It is up to conservationists to take this opportunity, if not for the sake of all stakeholders, for the sake of jaguars. I conclude that the jaguar conservati on problem is, first of all, a values problem and that the process can be greatly improved if strategies are designed to improve the
55 livelihood off all stakeholders, instead of looking at the problem from a biased biological perspective.
56 Table 1 1. The co nceptual framework of the social process (adapted from: Lasswel l, 1971; Clark & Wallace, 1998) Categories Definition Participants All individuals, groups, or institutions that can affect and/or be affected by the policy process. The analyst of the proce ss should include participants that he/she feels should be involved in the policy process but that are not currently involved. Perspectives The way participants view the policy process (demands, expectations and identifications) and the direction they want the process to go (their desired goals). What does each participant want? Situations The situations, events, where participants interact (e.g., meetings, workplace, etc.). Base values The assets that participants use to achieve their desired go als. Lasswell (1971) identifies 8 base values that can be used in any social process: 1) power to be able to make and carry out decisions 2) enlightenment to have knowledge 3) wealth to have money or its equivalent 4) well being to have health, physical and ps ychic 5) skill to have special abilities 6) affection to have family, friends, and warm community relationships 7) respect to show and receive deference 8) rectitude to have ethical standards Strategies The strategies participants use to achieve their desi red goals. Outcomes The outcomes achieved under the current policy process. Which participants are achieving their desired goals and which are not? Effects The effect the current policy process has on the participants desired outcomes.
57 Table 1 2. The seven decision functions that constitute a policy process (adapted from: Lasswel l, 1971; Clark & Wallace, 1998) Decision function Definition Intelligence (research and planning) Information relevant to decision making is collected, analyzed, and di stributed. Planning and prediction take place. Goals are clarified. Promotion (debating and recommending) Active advocacy debate about what to do. Recommendations are made and alternatives are debated based on desired goals of participants. Prescrip tion (creating rules) Policies or guidelines are formulated and enacted. Demands are crystallized. These rules must be specified, communicated, and approved by participants. Invocation (implementation of rules) Rules are put into practice and applied i n actual cases. Application (dispute resolution) Deviations from the rules are resolved and implementation continues. There must be enforcement as well as continuous approval, or disapproval of behavior. Appraisal (review) An assessment of performanc e. Efforts are evaluated and responsibility for success or failure is determined. Termination Terminating rules that are not having the desired outcome, or that have already achieved their goal, and compensating participants who are adversely affected b y termination.
58 Table 1 3. Number of known jaguar subpopulations per biome, average size of subpopulations per biome, and estimated total population of jaguars per biome. Numbers in parenthesis represent subpopulation estimates that may be considered par tially disconnected from the main popu lation of Amazonia Biome # subpopulations Avg. sub pop. size Pop. size Amazonia 1 (4) >10000 (473) >10000 Atlantic Forest 8 21 169 Caatinga 5 35 178 Cerrado 11 86 949 Pantanal 1 >5000 >5000 Source: Paula et al. 2 011.
59 Table 1 4. Biome's original area, percentage of Brazil's area in each biome, area and percentage of habitat lost, area and percentage of biome remaining, total number of protected areas (conservation units (CUs) and indigenous territories (ITs)), ar ea and percentage of biome protected Biome Amazonia Cerrado Atlantic Forest Caatinga Pampa Pantanal Total Original area 1 (km) 4,196,943 2,036,448 1,110,182 844,453 176,496 150,355 8,514,877 Pertecentage of Brazil (%) 49.3 23.9 13.0 9.9 2.1 1.8 100 Hab itat lost 2,3 (km) 744,584 986,247 977,172 383,297 95,308 22,969 3,209,577 Perc. of habitat lost (%) 17.7 48.4 88.0 45.4 54.0 15.3 37.7 Area of biome remaining (km) 3,452,359 1,050,201 133,010 443,182 81,188 127,386 5,287,326 Perc. of biome remaining (%) 82.3 51.6 12.0 52.5 46.0 84.7 62.1 Number of CUs 1 219 189 418 75 13 7 867 Area inside CUs (km) 1 1,070,061 165,227 101,762 62,631 5,851 4,400 1,409,932 Perc. inside CUs (%) 25.5 8.1 9.2 7.4 3.3 2.9 16.6 Number of ITs 409 NI NI 36 NI NI NI Area inside ITs (km) 4 991,951 85,388 5,104 2,185 24 2,561 1,087,213 Perc. inside TIs (%) 23.65 4.20 0.46 0.26 0.01 1.71 12.8 Area inside protected areas (CUs + ITs) (km) 2,062,012 250,615 106,866 64,816 5,875 6,961 2,497,145 Perc. inside protected areas ( CUs + ITs)(%) 49.1 12.3 9.6 7.7 3.3 4.6 29.3 source: Ministrio do Meio Ambiente (MMA) http://www.mma.gov.br/estruturas/sbf_dap_cnuc2/_arquivos/uc_por_biomacnuc_02junho2011_119_1.pdf; habitat loss in Amazonia up to 2010, source:
60 Table 1 5. The social context of the jaguar conservation problem in Brazil. It includes the groups of participants: small farmers and traditional communities (SFTC); large scale farmers (LSF); NGOs, research institutes, and universities (NRIU); government and wildlife agencies (GWA); the general public (GP); and trophy hunters and outfitters (THO). And their goals within the policy process, their strategies to achieve their goals, the outcomes of their strategies, and the effects that their actions have on other participants. Goals, strategies, outcomes and effects refer to the use, gain, or loss of the eight base human values as defined by Lasswell (1971) power, enlight en ment, wealth, well being, skill, af fection, respect, and rectitude G oals (What are the participants goals?) Strategies (actions) & Assets (How do participants use their assets, or b ase values, to achieve their goals?) Outcomes (What are the outcomes of the participants actions?) Effects (How do outcomes affect participants and the policy process?) Small farmers and traditional communities (SFTC) (1) to stop losing livestock to depre dation by jaguars (wealth and well being). (2) to be safe from potential jaguar attacks (well being). (3) to maintain their legal right to hunt game species for subsistence within their properties (power, well being, respect). (4) if corridors are imple mented in their properties, or in the vicinities of their properties, they want to their region to be formally recognized by the government as a jaguar conservation region. (respect, wealth, and well being). (a) They kill jaguars (power) to prevent, or ret aliate, depredation (1), and when they feel physically threatned (2). (b) They hunt game species (power) for subsistence. (c) They use their crucial role (power) in the establishment of corridors to guarantee their recognition (i.e., a green stamp on the ir produce)(4). Outcomes of (a): Jaguar populations may support the harvest and remain stable, decrease, or go extinct depending on jaguar population size, connectivity, and intensity of persecution. And/or move away from properties and communities. Depred ation decreases locally or regionally, temporarily or for a long period, or ceases. People encounter jaguars and their signs less often, or never. Outcomes of (b): Game populations may support harvest, decrease, or go extinct, depending on population s ize, connectivity, and harvest rates. And/or move away from properties and communities. Game species become harder to encounter. Effects of outcomes of (a): SFTC are benefited because economic losses are reduced and they feel safer. They gain wealth and well being. NRIU, GWA, and GP are contrary to the illegal killing of jaguars by SFTC. To the majority of these participants the killing is unjustified. They lose respect and well being. Effects of outcomes of (b): If game population decrease and/or becom e harder to encounter SFTC that depend on meat for protein may starve, switch to secondary game species, switch to alternative subsistence activities, or move to another rural location or urban areas. They lose well being and wealth.
61 Table 1 5. Continued G oals (What are the participants goals?) Strategies (actions) & Assets (How do participants use their assets, or base values, to achieve their goals?) Outcomes (What are the outcomes of the participants actions?) Effects (How do outcomes affect particip ants and the policy process?) Depredation of livestock by jaguars may increase if prey populations are negatively affected by subsistence hunting. Outcomes of (c): GWA formally recognizes SFTC in the region where corridors are created as jaguar conser vation agricultural properties If jaguar depredation on livestock increases as a result of the decrease of natural prey SFTC loose wealth and well being, and more conflict may lead to more killing of jaguars. Effects of outcomes of (c): SFTC feel that the y are part of the conservation process and that their demands are being taken into account. They gain respect, and their formal recognition by the government may help selling their product and increase wealth and well being. Large scale farmers (LSF) (1) to stop losing livestock to depredation by jaguars (wealth). (a) They kill jaguars (power) to prevent, or retaliate, depredation (1). (b) They destroy natural vegetation and transform it into pasture and crop land (2) (power and wealth). Outcomes of (a) : Jaguar populations may support the harvest and remain stable, decrease, or go extinct depending on jaguar population size, connectivity, and intensity of persecution. And/or move away from properties and communities. Effects of outcomes of (a): LSF are benefited because economic losses are reduced. They gain wealth and well being. Effects of outcomes of (b): Reduction and fragmentation
62 Table 1 5. Continued G oals (What are the participants goals?) Strategies (actions) & Assets (How do participants use their assets, or base values, to achieve their goals?) Outcomes (What are the outcomes of the participants actions?) Effects (How do outcomes affect participants and the policy process?) (2) to have legal rights to remove a larger percentage of the n atural vegetation of their properties to expand pasture and crop land (power, wealth, and well being). (3) if corridors are implemented in their properties, or in the vicinities of their properties, they want to their region to be formally recognized by the government as a jaguar conservation region. (respect, wealth, and well being). (c) They use their crucial role (power) in the establishment of corridors to guarantee their recognition (i.e., a green stamp on their produce)(4). Depredation decreases loc ally or regionally, temporarily or for a long period, or ceases. Outcomes of (b): Natural habitats are reduced and become fragmented. Jaguar and prey populations decrease or go extinct. Outcomes of (c): GWA formally recognizes LSF in the region where cor ridors are created as jaguar conservation agricultural properties. of natural habitats may increase depredation of livestock by jaguars. LSF loose wealth and more jaguars are killed. Effects of outcomes of (a) and (b): NRIU, GWA, and GP are contrary to th e illegal killing of jaguars and the destruction of natural habitats by LSF. To the majority of these participants the killing and the destruction of natural habitat is unjustified. They lose respect and well being. Effects of outcomes of (c): LSF feel th at they are part of the conservation process and that their demands are being taken into account. They gain respect, and their formal recognition by the government may help selling their product and increase wealth and well being.
63 Table 1 5. Continued G oals (What are the participants goals?) Strategies (actions) & Assets (How do participants use their assets, or base values, to achieve their goals?) Outcomes (What are the outcomes of the participants actions?) Effects (How do outcomes affect participants and the policy process?) Trophy hunters and outfitters (THO) (1) to have legal rights to hunt jaguars and other wildlife for sport (well being). (2) to have legal rights to keep jaguar, and other wildlife parts as suveniers (well being). (3) to be able to profit from trophy hunting (wealth and well being). (a) They hunt illegally (1)(2)(power and wealth). (b) They create illegal hunting enterprises (3)(power, wealth, and skill). Outcomes of (a) and (b): Jaguar and prey populations may support the harve st and remain stable, decrease, or go extinct depending on population size, connectivity, and intensity of persecution. And/or move away from properties. Effects of outcomes of (a) and (b): NRIU, GWA, and GP are contrary to the illegal killing of jaguars b y SFTC. To the majority of these participants the killing is unjustified. They lose respect and well being. Hunters are able to hunt. They gain well being. And outfitters profit from the illegal hunting. They gain wealth and well being. NGOs, research i nstitutions and universities (NRIU) (1) to reverse the trend of decline of jaguar populations threat status in all biomes of Brazil where the species still occurs in the next 10 years (well being). (a) They conduct research and work shops (skill and wealth) to gather information on jaguar biology and conservation (enlightment). (b) They write scientific and non scientific publications and reports (skill) to share information on jaguar biology and conservation with other Outcomes of (a): More and better information is available for the scientific community and policy makers. Outcomes of (b): Other participants are better informed of jaguar biology and conservation. Outcomes of (c): Some management actions implemented or proposed by GWA reflect NRIUs perspective of the policy process. Effects of outcomes of (a): NRIU gain scientific knowledge to plan future research and propose better management actions. They gain enlightment. Effects of outcomes of (b): GWA have more information on which to base management actions. They gain enlightment. Management actions have a
64 Table 1 5. Continued G oals (What are the participants goals?) Strategies (actions) & Assets (How do participants use their assets, or base values, to achiev e their goals?) Outcomes (What are the outcomes of the participants actions?) Effects (How do outcomes affect participants and the policy process?) participants (enlightment). (c) They use their knowledge (enlightment), technical abilities (skill), res ources (wealth), recognizement as serious and ethical institutions (respect and rectitude), to pressure the government to implement management actions to revert the decline of jaguar and prey populations, and their habitat. higher potential of being succe ssful. SFTC and LSF usually have no access to this information or find it of little use to achieve their goals. The GP receives most of this information, usually through media outlets (i.e., TV, internet, newspapers). The GP gains enlightment and well bei ng. SFTC gain little enlightment and rarely well and a portion of these participants feels pleasure (well being) for receiving information about the jaguar biology, research and conservation. Effects of outcomes of (c): Management actions proposed and imp lemented do not take into consideration the goals of SFTC, LSF and THO. They cannot hunt jaguars or their prey without approval of GWA (power), jaguars depredation continues to cause them economic losses and jaguars cannot be used for commercial purposes ( wealth and well being), THO are deprived of their pleasure of hunting and
65 Table 1 5. Continued G oals (What are the participants goals?) Strategies (actions) & Assets (How do participants use their assets, or base values, to achieve their goals?) Outcom es (What are the outcomes of the participants actions?) Effects (How do outcomes affect participants and the policy process?) revenue (well being and wealth), SFTC remain concerned about jaguar attacks (well being), and SFTC feel disrespect for being deprived of rights over their natural resources without their consent (respect). Government and Wildlife Agencies (GWA) (1) to reverse the trend of decline of jaguar populations threat status in all biomes of Brazil where the speci es still occurs in the next 10 years (well being). a)(b)(c): same as NRIU (d) They created a law (power) that protects natural areas within private properties, the permanent protected areas (APPs). (e) They banned hunting of the jaguar and trade of jagua r parts (power). (f) They banned commercial hunting and trade of wildlife (power). (g) They created and implemented protected areas (power, wealth, and skill). Outcomes of (a)(b)(c): same as NRIU Outcomes of (d) and (g): More natural areas are protecte d inside and outside private properties. More populations of jaguar and prey are protected. Outcome of (e): Commercial hunting of jaguars and trade of jaguar parts decreased. Hunting pressure over jaguar populations decreases. Outcomes of (f): Commercial hunting and trade of wildlife decreases. Hunting pressure over game populations decreases. Effects of outcomes of (a)(b)(c): same as NRIU Effects of outcomes of (d): Mainly LSF, but to some extent SFTC, lose decision right over a portion of their propert y where they cannot modify the natural habitat for pasture or crops. They lose power, wealth and respect. Effects of outcomes of (e)(f): Commercial hunting of jaguar and other wildlife is no longer a viable economic activity for SFTC. They lose power.
66 T able 1 5. Continued G oals (What are the participants goals?) Strategies (actions) & Assets (How do participants use their assets, or base values, to achieve their goals?) Outcomes (What are the outcomes of the participants actions?) Effects (How do outco mes affect participants and the policy process?) (h) They use GIS methods to monitor deforestation (skill and wealth). (i) They establish corridors to link subpopulations (power, wealth, skill). (j) They created compensation schemes to refund livestoc k farmers for livestock lost to depredation by jaguars (wealth and skill). Outcomes of (h): Deforestation can be monitored and participants responsible for deforestation can be held accountable. Outcomes of (i): Viability of small subpopulations may incre ase. Outcomes of (j): Retaliatory hunting of jaguars decreases and/or more cattle depredations are reported. Deaths that are not jaguar related be reported as jaguar depredation. Effects of outcomes of (g): The livelihood of SFTC living inside protected areas improved because wildlife is more abundant. They gain power, wealth, respect, and well being. Effects of outcomes of (g) and (h): Deforestation rates decreased. LSC lose power and wealth. Effects of outcomes of (i): GWA's chances of effectively imp lementing corridors increases. Effects of outcomes of (j): SFTC and LSF receive compensation. They gain wealth. General public (GP) (1) jaguars to be protected and available for future generations (well being). (a) Pressure the government to take action to protect the jaguar (power). (b) Donate resources (wealth) to NRIU to enable research and conservation actions. Outcomes of (a): May lead to implementation of management action or law enforcement efforts. Outcomes of (b): More resources are available f or research and conservation. Effects of outcomes of (a): GWA is pressured to act, but also gains political support. GWA gains power. Effects of outcomes of (b): NRIU has more resources for research and conservation actions. They gain power and wealth.
67 Figure 1 1. Flowchart representing proximate causes of jaguar population decline in Brazil (light grey area) and the factors that contribute to the aggravation of these causes (darker grey area)
68 Figure 1 2. Remaining original vegetation of Brazilian biomes in 2008. Data so urce: MMA, INPE and IBGE
69 CHAPTER 2 JAGUAR ( PANTHERA ONCA ) POPULATION DYNAMIC S AND ACTIVITY PATTE RNS IN A SUSTAINABLE USE RESERVE IN THE VRZE A FLOODPLAIN FORESTS OF BRAZILIAN AMAZONIA Estimates of abundance and other demographic par ameters are crucial in determining trends in population dynamics and identifying parameters responsible for those trends. Demographic information is desired in the decision making process of conservation and management of wildlife, but is still limited or non existent for most species including many endangered ones (IUCN 2011). This shortage of demographic information is mainly due to the logistical and financial constraints associated with sampling animal populations at the required spatial and temporal sc ales (e.g., monitoring the jaguar ( Panthera onca ) population of Amazonia), and the inability of current sampling methods to detect all individuals even within a limited survey area (i.e., imperfect detection; Williams et al. 2002). Demographic parameters a re particularly difficult to estimate for large felids because they occur at relatively low densities, have large home ranges, and are typically difficult to detect due to their elusive and cryptic nature. The m anagement of large felids, and inference on p opulation dynamics, are thus often hindered by limited or unavailable information on demographic parameters. The lack of demographic information for large felids started to change after Karanth (1995) proposed the use of camera traps associated with closed population capture recapture (CR) models as a method to estimate abundance and density of tigers ( Panthera tigris ) and showed its potential use for other individually marked species (method further developed in Karanth & Nichols 1998, 2000, 2002). Since t hen, there has been wide use of this methodology to estimate density of many carnivores ( Trolle & Kry 2003, 2005 ; Maffei et al. 2005 ; Di Bitetti et al. 2006, 2008 ; J ackson et al.
70 2006 ; Dillon & Kelly 2007, 2008; Kelly et al. 2008 ; Cullar et al. 2006). Mo re recently recapture approach (Pollock 1982, Pollock et al. 1990) to estimate other population parameters such as survival, growth rates, and recruitment, and to make better inference about po pulation dynamics (Karanth & Nichols 2006). The methodology developed by Karanth (1995) became particularly popular for estimating density of jaguar populations and has been used in over 83 surveys in at least 50 different locations since 2003 (Wallace et al. 2003; Maffei et al. 2011). These surveys, however, only cover a small portion of the 137 Ecoregions 1 2 1) which is thought to be the most important area for the conservation of the jaguar (Sanderson et al. 2002, Caso et al. 2008, Paula et al. 2011), and the majority of surveys have not been conducted over sufficiently long enough periods of time to observe population dynamics and allow estimation of other population parameters. The closed population capture recapture method so far used to obtain jaguar abundance and density estimates has relevant weaknesses Wide ranging animals like the jaguar have large home ranges and are highly mobil e which means that jaguars occurring in th e border regions of trap arrays will move in and out of the survey area during the survey even when we restrain the survey to a short period of time thereby violating the critical assumption of population closure. Th i s movement of individuals can be view ed as a form of temporary emigration and it leads to heterogeneity in capture probabilities (i.e., individuals with center of activity in the vicinity of the trap array will have lower ex posure to trapping compared with individuals whose center of activity is 1 Ecoregions are defined by Olson et al. (2001) as relatively large units of land containing a distinct assemblage of natural c ommunities and species, with boundaries that approximate the original extent of natural communities prior to major land use change.
71 located inside the trap array), negatively biasing detection probability and positively biasing abundance estimates (Kendall et al. 1997, Kendall 1999). T o convert abundance estimates to density it is necessary to calculate the effective trapping area (ETA) of the survey. Traditionally, this has been achieved using ad hoc approaches based on estimates of boundary strip width, usually half or the full mean maximum distance moved (MMDM) by individuals captured during the survey (Karanth & Nichols 1998, 20 02). The MMDM is used as a surrogate of the home range size radius, which is added as a buffer to the trap array to estimate ETA. This is viewed as the weak link in this methodology because this approach has no theoretical mechanism to link abundance with the survey area to estimate density (William et al. 2002), and ETA may vary with different methods, yielding different density estimates for the same abundance To deal with these issues formal model based procedures have been developed to estimate density directly from capture history data and the auxiliary spatial information from the location where individuals are captured (Efford 2004 ; Borchers & Efford 2008 ; Royle & Young 2008; Royle et al. 2009) Thes e procedures have been developed under likelihood (Borchers & Efford 2008; Efford et al. 2009a) and Bayesian analysis frameworks (Royle & Young 2008; Royle et al. 2009a, b) and use hierarchical models to condition the encounter history data to an underlyin g point process that describes the distribution of individuals in space (Efford 2004; Royle & Young 2008; Borchers & Efford 2008; Royle et al. 2009a) Recently, it has been shown that density estimates for jaguars and other species were consistently overes timated by the MMDM method when
72 compared to model based spatially explicit capture recapture (SECR) methods (Gerber et al. 2011; Noss et al. 2012 ) In this study we used a SECR approach under a Bayesian analysis framework to estimate jaguar density from 20 05 2009 in Mamirau Sustainable Development Reserve, a Vrzea Floodplain Forest site in the Brazilian Amazon with a relatively high human density (1.55 people/km) and high human induced mortality of jaguars (Ramalho 2012 Chapter 3). We also estimate jag uar survival for the period and test the prediction that jaguar survival and density are stable in Mamirau Reserve despite high human induced mortality because of the large number of immigrants and abundance of prey. Additionally we describe jaguar activi ty patterns in the study area. Methods Study Site Vrzea Floodplain Forests Floodplains can be briefly defined as wetlands that periodically transition between by t he lateral overflow of rivers or lakes and/or by direct precipitation or groundwater; the resulting physico chemical environment causes the biota to respond by morphological, anatomical, physiological, phenological, and/or ethological adaptations and produ ces characteristic community structures floodplains fringe the Amazon River and its large tributaries along most of their course, with the exception of the estuary, covering an area of approximately 300,000 km (Junk 1997). These river floodplains are seasonally inundated by the large and predictable monomodal flood pulse of the Amazon River and its tributaries (e.g., the average annual amplitude of the flood pulse in Mamirau Sustainable Development Reserve is
73 >10 m, Ramalh o et al. 2009). Amazonian floodplains have been categorized as two main types based on their hydrological characteristics. When inundated by alluvial (muddy) white water rivers, which are nutrient and sediment rich (e.g., Amazon River), they are called Vr zea. When inundated by black water rivers (e.g., Rio Negro), which are nutrient and sediment poor, they are called I gap. Vrzea forests cover approximately 180,000 km (2.6%) of the Amazon basin (Fig 2 2) and are crucially important to Amazonia due to t he abundance of fish, and their role as breeding grounds for many species of fish, birds, mammals and reptiles (Bayley & Petrere 1989 ; Goulding 1996; Thorbjarnarson & Da Silveira 2000 ). These areas are also very fertile due to the constant renewal of soil nutrients caused by annual flooding, which makes Vrzea forests the most productive environments of Amazonia (Morn, 1990). These attributes and the proximity to rivers (the main transport routes for local people) have historically favored human occupation of the Vrzea forests by people resulting in the most densely human populated environment in Amazonia (Ayres 1993). In the Vrzea floodplain forests, as in other floodplain environments, the variation in the water level dictates most ecological processes As the water level rises, the nutrient rich waters invade the floodplains, replenishing the soil with nutrients, restricting the terrestrial habitat, and expanding the aquatic habitat. Fish and other aquatic organisms reproduce during flooding, taking ad vantage of the lower density of predators, which have migrated or are confined to small islands of dry land, and the abundance of food, such as seeds from dispersing trees. The trees of the Vrzea forests also take advantage of the high water level to disp erse their seeds using water
74 and fish as dispersal agents. When the water re ce d e s the aquatic organisms become restricted and concentrated in lakes, channels, and other water bodies, or migrate into the main course of the larger rivers. In this low water s eason Vrzea forests become accessible to the terrestrial fauna, and attractive to predators, which find an abundance of prey concentrated in small water ways. It is in the areas surrounding these bodies of water, in the interface between the aquatic and terrestrial environment, that most predato r prey interactions occur (Junk 1993). Most terrestrial animals and predators reproduce during this period in which food is abundant to them. This cycle allows for a great variety of organisms to occur in the same area, but demands that plants and animals have a large range of morphological, anatomical, physiological and ethological adaptations, to survive (Junk 1993). Mamirau Sustainable Development Reserve This study was conducted in Mamirau Sustainable Developm ent Reserve (hereafter, Mamirau Reserve), located in the western portion of Brazilian Amazonia, Fig. 2 3 ). Mamirau Reserve is delimited by the Japur an d Amazon R ivers, and the Auati paran channel, and encompasses an area of 11,240 km of Vrzea forests (6.25% of the total area of the Vrzea ecosystem in Amazonia). It is the largest protected area exclusively dedicated to protecting this type of environm ent The climate in the region is tropical humid with average annual precipitation of 2,373 mm (Ayres 1993). Mamirau Reserve was originally created as an Ecological Station in 1984 by the Brazilian Environmental Agency (SEMA) and in 1990 its administratio n was transferred to Amazonas state government (decree n 12,836 of March 9th 1990). Mamirau
75 Ecological Station was created mainly in response to a proposal from biologist Jos Mrcio Ayres to create a protected area of approximately 2,500 km, primarily to protect the endangered white uakari monkey ( Cacajao calvus calvus ; Fig. 2 4) and its habitat (Queiroz 2005). However, the area ultimately designated for conservation was 11,240 km, almost five times larger than that requested, due to the environmentall y favorable national political climate in Brazil in the late 1980s and the increasing world wide concern about global warming and loss of biodiversity (Esterci & Ramalho 2007). This rare and paradoxical circumstance, where a government creates a protected area that is actually much larger than solicited, although often welcome and at first view positive, resulted in a caveat. The objective of an Ecological Station is to preserve nature and to be a pristine natural area for the realization of research and ed ucational activities (Brasil 2000). People are not allowed to live in, visit, or use natural resources from inside an Ecological Station. However, Mamirau Ecological Station encompassed a crucial system of Vrzea forest lakes and other water ways with fis hery stocks that supplied hundreds of thousands of people in the region, and also contained the households and subsistence territories of approximately 5,000 local people distributed over 60 villages, whose livelihoods were completely dependent upon the na tural resources of the area. The creation of such a large Ecological Station was clearly inappropriate given the socio economic characteristics of the region and the function of this type of protected area (Esterci & Ramalho 2007). Facing this dilemma, Ayr es and other researchers proposed the creation of a sustainable development reserve, a new category of protected area that they judged more adequate and viable for the area. This type of reserve was based on the sustainable use model of protected areas tha t was
76 rapidly gaining popularity over the fortress conservation model, and was based on the rationale that for effective conservation to occur local people had to participate in the management of the protected areas and benefit from the conservation of the natural resources within it. Mamirau Reserve was the first sustainable development reserve to be created in Brazil (Queiroz 2005; Esterci & Ramalho 2007). Sustainable development reserves are defined in the Brazilian National System of Conservation Units (SNUC) as natural areas inhabited by traditional human populations whose existence is based in sustainable systems of natural resource exploitation, developed through generations and adapted to the local ecological conditions, and that play a fundamental role in the protection of nature and maintenance of biological diversity. The objective of this category of protected area is to promote the conservation of biodiversity, and, at the same time, to secure the conditions and means necessary for reproduction, improvement of quality of life, and sustainable exploitation of natural resources by traditional local people, as well as to value, conserve and improve upon the knowledge and natural resource management techniques developed by these populations. The env ironment and the ecology of animals and plants in Mamirau Reserve, as in other Vrzea forest areas, are largely determined by the flood pulse, as the water level of the Amazon River and its tributaries in the region of Mamirau Reserve can fluctuate >13 m in a year, with an average annual fluctuation of >10 m (Fig. 2 5; Ramalho et al. 2009). The variation of altitude of the terrain within the Vrzea forests of Mamirau Reserve, and the consequent difference in level and period of flooding of the area, crea ted distinct terrestrial environments with characteristic vegetation structure
77 and composition. Ayres (1993) identifies three main types of environment and defines them as follows. The High Restinga (HR) represents the higher elevation terrain that is floo ded for up to 4 months per year by a water column of up to 2.5 m. These areas, although structurally similar to Terra Firme forests (upland forests) have a distinctive tree community. Most frequently encountered botanic families reported by Ayres were Anno naceae (16.4%), Euphorbiaceae (10.5%), Leguminosae (7.8%), Apocynaceae (7.4%), Lecythidaceae (6.0%), and Lauaraceae (5.2%). Some of the largest tree species in Amazonia, such as the samaumeira ( Ceiba pentandra ) and the assacu ( Hura crepitans ), are found in this environment. The Low Restinga (LR) represents intermediate elevation terrain with a generally open understory. This environment covers most of Mamirau Reserve and can be flooded for up to 6 months by a water column of up to 5 m. In the LR, Euphorbia ceae is the most frequent botanical family (18.8%) followed by Leguminosae (16.0%), Lecythidaceae (7.0%), Myrtaceae (5.8%), and Annonaceae (5.5%). Some of the most frequent tree species found by Ayres in this environment were the mututi branco ( Pterocarpus amazonicus ), the mat mat ( Eschweilera albiflora ), and the piranheira ( Piranhea trifoliate ). Palms are rare in both of the restingas. The Chavascal (CH) is the lowest terrestrial environment in the Vrzea forests of Mamirau Reserve. The CH is a swampy e nvironment with low vegetation and a dense understory, and can be flooded for up to 8 months per year by a water column of 7 m or more. The most abundant plants in the CH are the bamboos called tabocas ( Bambusa spp.), the munguba ( Pseudombax munguba ), the piranheira, the imbabas ( Cecro pia sp.), and apu species ( Ficus spp.). The palm jauari ( Astrocaryum jauari ) is also frequently found.
78 Because Mamirau Reserve is a seasonally inundated island in the middle of two large rivers, animals that live inside it have to be well adapted to swimming and/or climbing trees, and to survive the steep annual fluctuation in resource availability caused by flooding. These peculiar environmental characteristics are responsible for the presence of some endemic species, but a lso to a lower density and diversity of terrestrial species in general. Primate diversity is lower than in the surrounding Terra Firme forest, but Mamirau Reserve encompasses most of the distribution of the white bald headed uakari monkey and the entire d istribution of the endemic blacked headed squirrel monkey ( Saimiri vanzolinii ). Threatened and charismatic top predators such as the jaguar, the black caiman ( Melanosuchus niger ), the pirarucu ( Arapaima gigas ), and the Amazon river dolphin ( Inia geoffrensi s ) are abundant. Mamirau Reserve also holds a diverse fish and bird fauna with at least 340 species of each group (Queiroz & Peralta 2011). The human population of Mamirau Reserve lives primarily along the margins of the main Rivers (e.g., Amazon River a nd Japur River), on smaller channels, and along the margins of some lakes. The main economic and subsistence activities of local people are agriculture, hunting, fisheries, and harvesting of wood and other non timber products. The main source of protein c omes from fishing. The human population of Mamirau Reserve is estimated at 9,733 people (0.87 people/km), distributed in 1684 households over 181 villages (MSDI 2011). This study was conducted during 2005 2010 in an area of ~566 km around Mamirau Lake, a well protected area of Mamirau Reserve where jaguar prey populations are abundant and have been protected since the creation of the Reserve in
79 populations are estimated to be concentrated at 230 individuals/km of margin (Da Silveira 2002), brown throated three toed sloth ( Bradypus variegatus ) density is estimated to be over 200 individuals/km (Queiroz 1995), and red howler ( Allouata seniculus ) occur at 35 individuals/km (F Pain unpublished data). This area receives regular human activity from local fisherman and tourists from the Uakari ecotourism lodge. Field Methods Camera trap surveys Jaguars were surveyed with camera traps in the low water season of years 2005 2008 in a total of four surveys (Table 2 1). The low water season in Mamirau Reserve extends from September to December, after which the water starts to rise (Ramalho et al. 2009). C amera traps were installed along trails created by humans and wildlife, near the margin of lakes, and in other locations that maximized the probability of jaguars being photo captured (Karanth & Nichols 2002) Selection of camera trap locations was based on signs of jaguar presence (e.g., tracks, scats, carcasses of prey, and scratches on trees) and my own experience or that of local people in identifying jaguar travel paths in the study site. At each location two camera traps were set on opposite sides of the target path, separated by 3 5 m. Each camera trap pair composed a camera tra p station. Camera traps were programmed to take photographs 24 h per day with a 30 sec interval between photos, and to record date and time on each photograph Camera traps were inspected for malfunction, batteries and film at 3 7 day intervals. Both digit al and conventional film cameras were used and included the following models: Camtrakker (Cam Trak South Inc., GA. USA) models Original (film) and Environmental
80 unit (film), Bushnell (Bushnell Corp., KS, USA) model Trail Master (digital), Tigrinus (Tigr inus lnc., SC, Brazil) models Convencional (film) and Digital (digital). In total four camera trap surveys were conducted at approximately yearly intervals (Table 2 1). Each camera trap station was also equipped with a homemade lure of sardine and eggs whi ch was placed in a small container in the center of the station at equal distance from the two camera traps. The objectives of using the lure were to increase the chances of a jaguar being photographed, and to position jaguars at a central position between the camera traps for a longer period of time to improve the quality of pictures and chances of identification of individuals. Although there is little information on the effect of lures on jaguar camera trap surveys, this homemade lure worked well to draw jaguars and other felids to a central position in camera trap stations in preliminary surveys at the study site (Ramalho 2006; Fig. 2 6). Additionally all supplies to make the lure were easily accessible in local markets and the cost of equipping stations with the lure were minimal, approximately $ 0.05 USD per station night. In using lures, however, I made two important assumptions. First I assumed that the lure was only effective at a small range (i.e., <100 m) and therefore did not cause animals to disp lace their home ranges (i.e., the lure did not draw animals from outside the effective sampling area). In a survey conducted by Gerber et al. (2011) the use of lure did not affect permanent immigration or emigration, abundance and density estimation, maxim um movement distances, or temporal activity patterns of Malagasy civets ( Fossa fossana ), but did provide more precise population estimates by increasing the number of recaptures. Second, although the odor of the lure changed with time I assumed that its
81 ef fectiveness in attracting the interest of jaguars did not change over the 3 7 day period within which lures were not replaced. The number of camera trap stations used in each survey varied from 5 to 17, and effort varied between 735 and 2,695 trap nights ( Table 2 1). Trap polygon area varied between 3.6 and 81 km (Table 2 1). In surveys conducted in 2007 and 2008 cameras were moved every 30 days in blocks. Camera trap stations were placed between 0.7 and 1.4 km apart. It is also important to reiterate that surveys were conducted in the low water season, when terrestrial habitats are extensively available, and that density is likely to change during the high water season (May August; Ramalho et al. 2009) when virtually all terrestrial habitats of Mamirau Re serve are under water Foot snare live captur es Jaguars we also physically captured with foot snares also during the low water season, from October to December, in years 2008 2010 (Table 2 1). Assistance in captures was provided by local knowledge and a pr ofessional trapper (D. Simpson, www.wildlifecaptureinternational.com ). Foot snares followed the design described by Frank et al. (2003) for African lion ( Panthera leo ) with minor modifications. Snar es consisted of an approximately 1 m long and 5 mm diameter stainless steel aircraft cable with ~5 cm loops at both ends made with swaged aluminum ferrules and a 19 mm angle iron lock used to keep the snare tight on the foot of the animal after the snare was sprung. Snares were fired using a modified Aldrich spring powered throw arm and were anchored to the ground or to a tree. Captured jaguars were immobilized with Telazol (tiletamine zolazepan, Fort Dodge do Brasil), or a combination of Telazol and ketam ine hydrochloride. Telazol was administered via an intramuscular shot using a 3 ml Daninject dart propelled by a CO2
82 rifle, an air rifle, or a blowpipe. Additional Ketamine was administered also via an intramuscular shot using a syringe. Dosage for Telazol was 6 mg/kg, and for ketamine 1 mg/kg. After immobilization individuals were fitted with a VHF or GPS/ARGOS telemetry collar made by Telonics, USA (Mesa, AZ) or Telemetry Solution (Concord, CA) and released. GPS locations where collected by the Telonics GPS/ARGOS collar every 5 hours. VHF collars were monitored on the ground for 5 day periods, at 7 day intervals, and a one hour search was conducted by plane every two months to find collared animals that could not be found on ground searches. All GPS colla rs produced by Telemetry Solutions (n = 6) stopped operating properly shortly after deployme nt or in the lab. Data Analysis Population density Historically, the estimation of abundance and density of carnivores in camera trapping studies has been done usin g ad hoc or heuristic methods based on closed population capture recapture estimators of population size applied to individual encounter histories. Although this approach is adequate for estimating the population size exposed to sampling, the effective sam ple area of the trapping array is unknown because conventional methods used to estimate effective sample area are not formally linked to the observed encounter history data. To address this issue, spatial capture recapture models have been developed by con ditioning the encounter history data to an underlying point process that describes the distribution of individuals in space in the context of a multinomial observation model where each individual can only be captured in one trap per sampling occasion (Effo rd 2004; Royle & Young 2008; Borchers & Efford 2008; Royle et al. 2009a). Royle et al
83 (2009b) describe a hierarchical modeling framework for inference from spatial capture recapture data for methods wherein the traps function independently of one another, allowing individuals to be captured multiple times within a sampling occasion. To estimate the density of jaguars in this study I used the R software package SPACECAP 1.0 (Singh et al. 2010), which was specifically developed for estimating animal densities from camera trap surveys using the spatial capture recapture models developed by Royle et al. (2009b). The models used in SPACECAP are based on point process models where it is supposed that, in a population of N individuals, each individual has a center of activity si = (s1j, s2j; i = 1, 2, ..., N), over which their movements are concentrated. It is assumed that these activity centers are independent, uniformly distributed over some region S, the state space of a binomial point process, and that the locat ion of activity centers does not change during the survey. S is defined as an area large enough to contain the trap array and also assure that all individuals outside S have a zero probability of being captured in the trap array. The basic inference proble m is to estimate the number of activity centers per unit area of S, which is equivalent to estimating N under the point process model. This uniform point process model represents a prior distribution for individual activity centers. To model the overlap of activity centers with the trap array, it is assumed that the assumed to work independently so that individuals can be encountered by more than one trap within a sampling occasion. It is also assumed that the probability of an individual i being encountered by trap j (i.e., the juxtaposition of si and xj) is a decreasing function of the distance between the trap and the individuals activity center,
84 plus one or more paramete activity center and trap j is represented by dij = || si xj ||. Because the location and number of activity centers is unknown, the method of data augmentation is used to define the paramete r space (Royle & Dorazio 2008). Data augmentation is the physical augmentation of the n observed encounter histories with individuals that includes the actual N individuals as a subset. M must be large enough so as not to truncate the posterior of N. The result is that the model for the augmented data is a zero inflated version of the model when N is known. In this model N is replaced by 1 population of size N that was ex posed to sampling by the trap array. To implement these models, SPACECAP requires three input files: potential home range centers, trap deployment details, and animal capture details. The potential home range centers file contains spatial information of a grid of equally spaced points generated to encompass the whole extent of the trap array plus an additional area surrounding it, where the researcher believes all individuals potentially detectable by the trap array are contained. Each point in this grid is a potential home range center and the area covered by the grid is the state space (S). The trap deployment details file contains spatial information on the location of each trap plus information on the operational status of each trap (i.e., if a trap was operational or not) at each sampling occasion. The of each capture.
85 For this study, S was defined as the area covered by the trap array plus a 15 km buffer around i t (Fig. 2.7), with a total area of 1,079 km (i.e., excluding non habitat). Each potential home range center within that area was spaced 250 m apart, a 0.0625 km pixel. In defining the model for the analysis, we considered the possibility of trap response to capture, used the spatial capture recapture model with a half normal used 10,000 iterations with 1,000 initial burn in values, thinning rate of 1, and data augme ntation of 300. Survival and recapture probabilities The primary interest in this section of the study was to estimate annual apparent and p throughout the study (20 05 2010). During this period there were no obvious anthropogenic or environmental events (e.g., increase in the number of jaguars killed by local people, or decline in abundance of main prey) in the study site that that would indicate significant changes i and p would be different between years). However, an unusually high flood in 2009, the the Vrzea in se arch of higher ground in the neighboring Terra Firme forests. Every year the terrestrial habitat of Mamirau Reserve is severely reduced by flooding, but in 2009 all terrestrial habitat was flooded. To evaluate this hypothesis, I included the level of floo categorical variable indicating normal flooding and high flooding (low high Flood) and as a continuous variable in meters above sea level (Flood).
86 I was also interested in as main reasons. The first is based on the theory of sexual segregation, which predicts that female behavior will al ways have the objective of increasing the chances of survival of offspring, while males will behave in a manner that favors their chances of reproducing, even when such behaviors increase personal risk (Main 2008). Under this rationale, females would stay in the Vrzea during annual flooding periods, raising their cubs in a semi arboreal aquatic lifestyle for a portion of the year, while males would migrate out of the Vrzea during flooding to avoid the scarcity of food and terrestrial habitat and maintain their physical condition in preparation for the next reproductive period. The second is that males are well known for having larger ranges than females, which increases the likelihood of encounters with people which could lead to higher mortality. This not ion is corroborated by the higher number of male jaguars killed by local people in comparison to females in the study site (Ramalho 2012 Chapter 3). To estimate apparent survival and recapture rates we used Cormack Jolly Seber (CJS) CR models (Cormack 19 64; Jolly 1965; Seber 1970) implemented in program MARK 6.2 (White & Burnham 1999). To test our hypotheses we used a set of candidate models that represented the effect of time (i.e., sampling year), flood level (i.e., highest level of flooding during the year of the survey), and sex, on survival; and the effect of sampling method (i.e., camera traps or snares), and sampling effort on recapture used as an objective mean s of model selection (Burnham & Anderson 2002). Goodness of fit was assessed on a fully time dependent model using program
87 RELEASE from within MARK (White & Burnham 1999). Survival and recapture probabilities were model averaged using Akaike weights (wi) to include model uncertainty in the estimates of parameter precision (Buckland et al. 1997). We also incorporated the presence of transients in our models given that 13 (54.2%) of the 24 individuals captured in our surveys were only captured once. As defin which then permanently emigrates from the sample, such that it is no longer available val and not true survival, individuals that emigrate permanently will appear to have died, and thus lower estimates of apparent survival of resident animals which are available for recapture in subsequent surveys. To account for the presence of transients we used a survival of residents and transients separately. Activity patterns As cameras operate intermittently throughout the surveys and 24 h per day it is usually assu med that the number of photographic captures by time period reflects an 2004). To quantify the activity patterns of jaguars in Mamirau Reserve we counted the numbe r of pictures of jaguars per hour period of the day, and by separating the day into four time periods: dawn, day, dusk, night. Dawn and dusk were comprised of two hour periods with centers around sunrise and sunset respectively. Day was comprised of an 11 hour period and night of a 9 hour period. To evaluate if jaguars were more likely to be photo captured during any of these periods, I used a contingency test using the chi squared statistic (Zar 1984). To increase the sample size for this analysis, I used
88 photographs from a survey conducted in the study site during the dry season of 2011, which was not used for the estimation of density and survival. Results The combined effort using cameras and snares was 10,668 trap nights for the six years of study in La ke Mamirau. The surveys yielded 94 observations of jaguars and allowed identification of 24 adult jaguars, of which 10 were male and 14 were female for a male:female sex ratio of 1:1.4. In total, camera traps identified 22 individuals and foot snares iden tified an additional 2 individuals out of a total of 9 captures, all of which were adult jaguars (Table 2 2). The combined rate of capture during this study was one jaguar per 113 trap nights of effort, and cameras were more effective in detecting jaguars than were snares. Cameras recorded one jaguar per 98 trap nights, whereas snares captured one jaguar per 227 trap nights (Table 2 2). Density The estimated number of jaguars within the 1,079 km state space area considering only appropriate habitat (Fig. 2 7) varied between 125 48.96 and 252 47.83 jaguars (estimated mean posterior SD) with an estimated abundance of 193 (Table 2 3). Population density in the surveyed area was high com pared to most published reports, and varied between 11.60 4.54 and 23.37 4.43 jaguars/100 km with an average density of 17.84 jaguars/100 km (Table 2 3; Fig. 2 8). In the five intervals between surveys co nducted in years 2005 and 2010, the population decreased during 2005 2007, had a sharp increase between 2007 and 2008, and then maintained densities intermediate to these two time frames during 2009 and 2010 (Table 2 3). The posterior mean of (the data augmentation parameter that represents the ratio of the
89 number of animals actually present within S to the maximum allowable number set by us during analysis) was lower than 1 for all years (Table 2 3), indicating the data augmentati on number M = 300 was adequate and did not truncate our estimates of population size and density Survival and Recapture Probabilities The saturated model (group structure (males and females) and time p) fitted the data reasonably well, yiel ding a = 1.29, and we adjusted the AIC scores by this measure of over dispersion (White & Burnham 1999). The most parsimonious flood) p(effort)}, where transients and level of flooding inf luenced apparent survival, and effort affected recapture probabilities (Table 2 4). This model was substantially better than the other 39 models in the candidate set (AICc weight = 0.34). I formally tested our hypotheses by comparing more general models wi th reduced models using the likelihood ratio test (LRT). I found no evidence of variation in survival associated with gender or time (Table 2 5). I did however find support for variation in survival dependent upon flood level ( = 6 .126, df = 2, P = 0.047), and also for differences in survival rates between residents and transients ( = 4,864, df = 1, P = 0.027; Table 2 5). I did not find evidence of time or sampling method variation in recapture probabilities (Table 2 5), but found strong support for the influence of sampling effort ( = 4.496, df = 1, P = 0.034). I obtained estimates of survival and recapture probabilities for residents and transients for each sampling period by model averaging the estimates of the best ranked models in our candidate data set (i.e. models with QAICc weight > 0.05)(Table 2 6). For model averaging, I also excluded models that were failing to estimate
90 parameters (i.e., models with singular values because of parameters estimated at the boundary). Model averaged survival estimates for resident jaguars varied between 0.59 0.21 and 0.78 0.20 with an average of 0.66 over the entire study (Table 2 7). The highest survival rate for residents was observed between 2008 2009 surveys, which was the period when flooding was highest during the study (Fig. 2 5). For transients survival estimates were much lower, varying between 0.34 0.26 and 0.40 0.17, with an average of 0.38 over the entire study (Table 2 7). Flooding appeared to have a negative effect on survival of transients, but large variance in survival estimates give weak inference to this conclusion (Table 2 7; Fig 2 9). Population Structure a nd Reproduction Evidence of transients in the population of jaguars in Mamirau Reserve was inferred from the high number of adult animals that were captured during only one survey period (e.g. Pradel et al. 1997). The mean p roportion of residents to transients was roughly 1:1 (i.e., 51.5% of the 24 adult jaguars observed were transients), which was estimated by dividing the average survival of animals identified as transients by the average survival of residents (for details see Pradel et al. 1997). A similar proportion was observed in the raw data, with 11 individuals captured in surveys during more than one year, and 13 individuals captured during only one survey. Assuming individuals caught in only one year were transients, the sex ratio of residents and transients was quite different. For residents, the male:female ratio was 1:2.7 (3 males and 8 females), whereas for transients it was 1:0.9 (7 males and 6 females). Fourteen individual female jaguars were photographed or phy sically captured during this study. At least 5 of these females, 4 of which were residents, were pregnant
91 or with small cubs. This indicates that jaguars are reproducing and rearing their cubs in the Vrzea Forests of Mamirau Reserve. Furthermore, one fem ale was photographed or physically captured during multiple years and provided evidence of producing multiple litters during the study period. This jaguar was documented as pregnant with one cub at heel in 2006, was not observed pregnant or with a cub in 2 007 and 2008, was not captured on film during 2009, and then was physically captured in December 2010 and determined pregnant. Upon capture, this female was outfitted with a GPS telemetry collar and monitored for 12 months until December 2011, a period dur ing which she is believed to have successfully raised her cub. During the year she was monitored with GPS telemetry she remained in Mamirau during the entire year, including during the flooded season. It is also relevant to report that she was captured in the same area over all years. This female, therefore, demonstrated strong site fidelity. Her cub (also a female) also demonstrated strong site fidelity and established a home range in the same area as her mother, where she was observed in 2007, 2008, and 2010. Activity Patterns Jaguars were photographed during all hours of the day and night, characterizing a cathemeral behavior, but were most active during daylight hours (63.1% of observations). The distribution of activity observed reflected a bimodal pat tern of behavior with peaks of crepuscular activity during early morning (between 5:00 and 8:00 hr), and during midday (between 12:00 and 16:00 hr). Jaguar were observed more frequently than expected during the day and dawn, and less than expected in dawn as dusk (chi squared = 7.78, df = 3, p value = 0.05; Fig. 2 11).
92 Discussion The results of this study reveal that, at 17.84 jaguars/100 km, jaguars can live in very high densities in the Vrzea Floodplain Forests of Amazonia during the low water season. D ensities of more than 12 jaguars/100 km have never been reported prior to this study (Maffei et al. 2011). High density of jaguars are thought to be associated with high density of prey (Nuez 2012), a flexible cat social system (Harmsen et al. 2009), and /or mutual avoidance (Nuez 2006). The high density of jaguars observed in this study may be made possible by the abundance, concentration, and accessibility of prey in Mamirau Reserve during the low water season. Mamirau Reserve has one of the largest d ensities of caiman in Amazonia, and also has high densities of brown throated three toed sloth and red howler monkeys, all of which are important prey for jaguars in Mamirau (Ramalho 2012 Chapter 4). The relationship between prey abundance and large car nivore abundance has also been demonstrated for the tiger in India (Sunquist et al. 1999; Karanth et al. 2004). Evidence of a flexible social structure is supported by the high degree of overlap indicated by the large number of jaguars captured in our rela tively small sample area. The high density of jaguars, however, seems unlikely to be maintained during one whole year. As water rises in the Vrzea a large portion of the jaguars may migrate, crossing the Amazon or Japur Rivers in search of higher dry gro und. This lateral migration and changes in density with flooding are observed in many animal populations living in the Vrzea (Junk et al.1989). However, the single adult female that was monitored via a GPS collar did remain in the Vrzea throughout both t he low and high water periods. This female had a cub at the beginning of the high water period and large rivers such as the Amazon may present a formidable obstacle to migrating with a
93 cub. Therefore, females with offspring may remain resident and survive by living a semi arboreal existence during high water periods. This is consistent with sexual segregation theory, which predicts females will sacrifice foraging opportunities in lieu of security for offspring (Main et al. 1996, Main 2008). However, in the absence of a larger sample size and information on male migratory patterns, it is not possible to make conclusive statements regarding migratory patterns of male and female jaguars in response to seasonal flooding in the Vrzea. Density and survival did not show clear trends over the period of this study and suggest the jaguar population in the study area remains fairly stable (Table 2 3) despite high levels of human activity and human induced mortality (Ramalho 2012 Chapter 3). Estimates of average app arent survival revealed survival rates among resident jaguars to be approximately 66%, but survival rates of transient jaguars were considerably lower and ranged between ~30 40% (Fig. 2 9). There was no evidence of variance in survival related to gender, d espite reports of higher numbers of male jaguars being killed by people in the study area (Ramalho 2012 Chapter 3). These data support the idea that large felid populations may be able to withstand high human induced mortality if prey populations are ade quate (Karanth & Stith 1999). However, the relatively high percentage of transients in the population (~50 %), indicates that there is a large number of new individuals coming into the population every year and suggests that immigrants may play an importan t role in maintaining the jaguar population of Mamirau Reserve. Approximately 73 jaguars are killed by people every year in Mamirau Reserve (Ramalho 2012 Chapter 3), and immigrant jaguars may play an important role in replacing this source of mortality
94 The social organization of jaguars is similar to that of other large cats with males occupying larger home ranges than females, and male home ranges overlapping with those of multiple females (Sunquist & Sunquist 2002). In areas where ranges of individua ls of the same sex overlap, they tend to avoid each other by not using the same area at the same time (Nuez 2006). Having said that, home ranges may overlap extensively even among individuals of the same sex. This is particularly true during the reproduct ive season when multiple male jaguars have been observed to accompany one female at the same time (Almeida 1976; Hoogesteijn & Mondolfi 1992). This behavior has also been reported by several local people in Mamirau Reserve (E. Ramalho, unpublished data). The observations made in this study corroborate the social organization and land tenure system expected for jaguars and confirms that jaguars are reproducing in the Vrzea during the low water season, which would explain the temporary tolerance of a high d ensity of individuals. The activity pattern of the jaguar in Mamirau Reserve was cathemeral, but mostly diurnal, with peaks of activity at dawn and around 13:00 h. Higher activity of the jaguar has been found to coincide with periods of higher activity of their main prey (Harmsen et al. 2011), when prey are thought to be more vulnerable. The jaguar has four main prey species in Mamirau: the spectacled caiman, the three toed sloth, the lesser tamandua (Tamandua tetradactyla), and the red howler monkey (Ram alho 2012 Chapter 4). The activity patterns of jaguars may match periods of higher activity of sloths and red howlers. However, the lack of information on the activity pattern of these species in the study area makes it hard to establish this relationshi p. Three toed sloths have been found to be highly active during the day in some sites (Sunquist & Montogomery 1973)
95 but not in others (Castro Vsquez et al. 2010). The lesser tamandua is more active during the night, which could explain part of their activ ity during the night. Establishing these relationships between activity of prey and predators is further confounded by the fact that camera traps were located on trails, which may be good foraging areas for the arboreal prey, but may not represent foraging for spectacled caiman. This study has demonstrated that Vrzea floodplain forests are breeding and rearing grounds for jaguars in Amazonia and as such the protection of these areas should be a conservation priority. This demonstrated also that high jaguar densities may be maintained in the Vrzea despite high levels of human activity and human induced mortality of jaguars if prey populations are abundant and there is a stable source of immigrants. The abundance of prey is likely due in part to the fact tha t none of the most important prey species of the jaguar in Mamirau Reserve are intensively used by local people as a protein source (Ramalho Chapter 4). Although reproduction was confirmed during this study, information is not available on reproductive success for this population and, consequently, it is not known whether annual reproduction replaces adult mortality and if population densities of jaguars would remain stable without a steady influx of immigrants. However, under current scenarios, our data suggest jaguar populations are relatively stable in Mamirau Reserve, despite the ineffectiveness of protected areas in preventing human induced mortality of jaguars (Carvalho & Pezzuti 2010; Ramalho 2012 Chapter 3).
96 Table 2 1 Survey years, periods, field method used (camera trap or foot snare), area covered by trap array, and e ffort (reported as trap nights) Year Period Method Trap polygon (km) Effort (trap night) 2005 8 July 2005 20 October 2005 Camera trap 53.9 2352 2006 16 July 2006 5 Decem ber 2006 Camera trap 3.6 735 2007 19 July 2007 28 January 2008 Camera trap 81.9 2583 2008 25 July 2008 9 December 2008 Camera trap and foot snare 16.6 2695 2009 12 September 2009 29 November 2009 Foot snare 11.3 1778 2010 10 November 2010 10 De cember 2010 Foot snare 8.4 525
97 Table 2 2. Effort and captur e rates per method and combined Camera Snare Camera + Snare Total effort (trap nights) 8146 2522 10668 Effort per capture (trap nights) 98 229 113 Capture per trap*night 0.010 0.004 0.009 Capture per 100 trap*night 1.02 0.44 0.88 Individuals captured 22 9 24
98 Table 2 3. Posterior summaries of model parameters for the jaguar surveys in Mamirau Reserve base d on data from 24 jaguars. N is the number of jaguar exposed to sampling and D is the density per 100 km, is the scale parameter of a bivariate normal encounter function, is the baseline detectability for an individual whose ac tivity center is located precisely at a trap, and is t he data augmentation parameter Year of survey 2005 2006 2007 2008 2009 2010 Average Parameter Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD N 181.161 73.692 161.931 68.946 125.118 48. 963 252.182 47.829 214.658 74.520 219.873 65.880 193.000 87.170 D 16.790 6.830 15.008 6.390 11.596 4.538 23.372 4.433 19.894 6.906 20.390 6.110 17.887 8.079 0.694 0.222 1.023 0.243 0.377 0.094 0.255 0.122 0.297 0.268 0.568 0.316 0.003 0.002 0.008 0.005 0.029 0.013 0.739 1.560 0.040 0.039 0.017 0.170 0.591 0.241 0.522 0.222 0.404 0.159 0.807 0.154 0.070 0.244 0.720 0.215
99 Table 2 4. Model selection statistics for t he full set of candidate model s Models where we applied an alternative optimization procedure based on simulated annealing. ** Models where confidence intervals were derived using profile likelihood. Model Q AIC c AIC c Q AIC c weight Model Likelihood Num. parameters Deviance flood) p(effort)* ** 83.889 0.000 0.336 1.000 6 27.862 fl ood) p(.)* ** 85.921 2.032 0.122 0.362 5 32.358 flood) p(method)* ** 86.920 3.030 0.074 0.220 6 30.892 86.977 3.087 0.072 0.214 5 33.414 flood) p(method,effort)* ** 87.232 3.342 0.063 0.188 8 25.990 (M2) p(.) 87.377 3.487 0.059 0.175 3 38.484 87.903 4.014 0.045 0.134 6 31.876 flood) p(method)* ** 88.414 4.525 0.035 0.104 3 39.521 89.088 5.199 0.025 0.074 3 40.195 89.165 5.2 75 0.024 0.072 4 37.978 89.700 5.811 0.018 0.055 7 31.115 90.027 6.138 0.016 0.047 2 43.348 flood) p(effort)* 90.301 6.412 0.014 0.041 4 39.115 90.602 6.712 0.012 0.035 4 39.415 90.776 6.887 0.011 0.032 3 41.883 90.847 6.958 0.010 0.031 6 34.820 flood) p(method,effort)* ** 90.978 7.089 0.010 0.029 6 34.951 91.133 7.244 0.009 0.027 5 37.570 9 1.616 7.726 0.007 0.021 5 38.053 flood) p(.) 91.683 7.794 0.007 0.020 3 42.790 91.840 7.950 0.006 0.019 3 42.946 92.155 8.266 0.005 0.016 3 43.262 92.237 8.347 0.005 0.015 3 43.343 93 .224 9.335 0.003 0.009 5 39.661 94.024 10.135 0.002 0.006 6 37.996 94.080 10.190 0.002 0.006 6 38.052 94.407 10.518 0.002 0.005 4 43.221 flood) p(effort)* 94.497 10.608 0.002 0.005 6 38.470 94.960 11.071 0.001 0.004 5 41.397 flood) p(method)* ** 95.183 11.294 0.001 0.004 6 39.155 95.833 11.943 0.001 0.003 9 31.831 96.049 12.160 0.001 0.002 10 29.177 ** 96.582 12.692 0.001 0.002 7 37.996 97.288 13.398 0.000 0.001 6 41.260 98.654 14.764 0.000 0.001 12 25.682 99.961 16.072 0.000 0.000 5 46.398 101.174 17.285 0.000 0.000 6 45.146 (g,flood) p(g,method,effort)* 104.748 20.859 0.000 0.000 12 31.776 118.380 34.491 0.000 0.000 17 27.676 123.725 39.835 0.000 0.000 18 28.967
100 Table 2 5. Results of likelihood ratio tests used to test hypotheses related to sur vival and recapture probabilities. Model parameters are transients (M2), gender (g), level of flooding (low/high fl ooding), time (t), constant (.) Hypotheses Reduced model General model df p value Survival Gender influenc es survival rates 0.913 2 0.633 There is no time variation in survival during this study 6.221 3 0.101 High floods will influences survival flood) p(.) 6.126 2 0.047* Transients are present and have different survival than residents 4.864 1 0.027* Recapture probability No time variation in recapture probabilities 7.369 4 0.117 Sampling effort affects recapture probabilities (M2,low/high flood) p(.) flood) p(effort) 4.496 1 0.034* Sampling method affects recapture probabilities flood) p(.) flood) p(method) 1.466 1 0.226 Significant results
101 Table 2 6. Model ranking of CJS mark recapture models used to estimate apparen (p) for jaguars in Mamirau Reserve from 2005 2010. Only models used in model averaging with more than 0.05 Q AICc weight and fully estimable parameters (i.e., no singular values) were used for model averaging. Model selection statistics presented are: quasi c), delta Q AIC AIC), QAIC weight, model likelihood, num ber of parameters, and deviance Models where we applied an alternative optimization procedure bas ed on simulated annealing was applied Model QAIC c QAIC c QAIC c weight Model Likelihood Num. Par Deviance 86.977 0.000 0.283 1.000 5 33.414 87.377 0.400 0.231 0.819 3 38.484 87.903 0.927 0.178 0.629 6 31.876 89.088 2.111 0.098 0.348 3 40.195 89.165 2.188 0.095 0.335 4 37.978 90.027 3.050 0.062 0.218 2 43.348 flood) p(effort)* 90.301 3.324 0.054 0.190 4 39.115
102 Table 2 7. Estimated model averaged survival ( ) and recapture probabilities ( p ) for resident and transient jaguars between sampling periods. Values shown are weighted average estimates, with standard error (SE), l ower (LCI) and upper ( UCI) 95% confidence intervals 2005 2006 2006 2007 2007 2008 2008 2009 2009 2010 Estimate SE LCI UCI Estimate SE LCI UCI Estimate SE LCI UCI Estimate SE LCI UCI Estimate SE LCI UCI Residents 0.59 0.21 0.21 0.89 0.61 0.1 9 0.25 0.88 0.78 0.20 0.26 0.97 0.67 0.17 0.31 0.90 p 0.85 0.11 0.50 0.97 0.85 0.11 0.51 0.97 0.78 0.10 0.53 0.92 0.66 0.20 0.25 0.92 Transients 0.40 0.17 0.14 0.73 0.38 0.15 0.15 0.68 0.39 0.15 0.16 0.68 0.34 0.26 0.05 0.83 0.40 0.18 0.13 0.7 5 p 0.72 0.16 0.35 0.92 0.91 0.10 0.49 0.99 0.91 0.10 0.51 0.99 0.79 0.11 0.50 0.93 0.62 0.23 0.20 0.92
103 Figure 2 1. Location of all camera trap surveys conducted to date to estimate jaguar density (white circles), Ecoregions within the jaguar pre sent distribution (other colors), an d extent of Amazonia (red line) This study
104 Figure 2 2. Location and extent of the Vrzea floodplain forests of Amazonia
105 Figure 2 3. Smaller frame shows location of Mamirau Sustainable Development Reserve within Brazil. In la rger frame red line repre sents the limits of the Reserve
106 Figure 2 4. White uakari monkey. photo: Luiz Claudio Marigo
107 Figure 2 5. Water level dynamics during the period of this study. Water level is presented i n meters above sea level (masl)
108 Figure 2 6. Felids sniffing the homemade lure used in this study during a preliminary camera trap survey in the study site. Panthera onca (A), Leopardus pardalis (B), and Leopardus wiedii (C) C B A C
109 Figure 2 7. State space area of 1,07 9 km determined by a 15 km buffer (black line) around the trap array used to survey the jaguar population of Mamirau Reserve (black circles), potential home range centers (green pixels). White areas repres ent non habitat
110 Figure 2 8. Jaguar density pe r year with SD of the posterior
111 Figure 2 9. Apparent survival rates of resident (black circles) and tr ansient (white circles) jaguars
112 Figure 2 10. Activity patterns of jaguars in Mamirau Reserve according to the number of independent photo captures recorded per one hour period of the day (n=111)
113 Figure 2 11 N umber of observed independent photo captures recorded per period of the day (black bars) and expected number of captures based on availability (n=111) C hi squared = 7.78, d f = 3, p value = 0.05
114 CHAPTER 3 ESTIMATING LARGE CAR NIVORE MORTALITY FRO M HUNTING USING CAPT URE RECAPTURE MODELS: TH E CASE OF JAGUARS IN THE AMAZON FLOODPLAI N FORESTS Hunting here defined as the legal or illegal pursuit and/or trapping of animals by humans with the int ent of killing them for food, management, sport, or trade has been reported to be the greatest source of mortality for many large carnivore species, and has caused the decline of large carnivore populations inside and outside protected areas worldwide (N oss et al. 1996; Woodroffe & Ginsburg 1998; Woodroffe 2001; Andren et al. 2006; Adams et al. 2008; Obbard & Howe 2008; Robinson et al. 2008). Hunting accounted for 53% (28 individuals) of total mortality of radio collared pumas ( Puma concolor ) in two popul ations in Washington State, USA (Cooley et al. 2009). Gray wolf ( Canis lupus ) populations have been consistently reduced by hunting in eastern Asia (Reading et al. 1998), Europe (Sidorovich et al. 2003) and the USA (Carbyn 1987). Between 77 and 85% of 99 r adio collared grizzly bears ( Ursus arctos ) were killed by people in the interior mountains of Northwestern USA (McLellan et al. 1999). Poaching accounted for 75% of radio collared Amur tigers monitored in the province of Primorski Krai, in Russia, between 1976 and 2005 (Goodrich et al. 2008). In Laikipia, Kenya, 17 out of 18 monitored lions ( Panthera leo ) were killed in retaliation for livestock depredation, with an estimated population decrease of 4% per year (Woodroffe & Frank 2005). In Western South Afri ca 60% of recorded leopard ( Panthera pardus ) deaths resulted from hunting (Balme et al. 2009). Hunting is thus a critical issue in the conservation of large carnivores. The jaguar ( Panthera onca ) is the third largest felid in the world, and the only repre sentative of the genus Panthera in the American continent. The jaguar is currently
115 classified as Near Threatened by the International Union for Conservation of Nature (IUCN) due to the declining trend of most jaguar populations and extensive reduction of t 2008). Despite the international ban on trade and export of jaguar parts that resulted from the inclusion of the jaguar in Appendix 1 of the Convention on International Tr ade in Endangered Species of Wild Fauna and Flora (CITES) in 1973 and national bans on hunting of jaguars in most countries after the 1960s, the declining trend of jaguar populations and continuing contraction of its range have persisted (Sanderson et al. 2002; Caso et al. 2008; Paula et al. 2011). Hunting is recognized as one of the most important threats to the survival of the jaguar (Ojeda & Mares 1982; Brown1983; Melquist 1984; Swank & Teer 1989; Sanderson et al. 2002; Caso et al. 2008; Paula et al. 20 11). Commercial motivation for hunting in the 1960s and 1970s, due to the high international demand for jaguar skins, promoted the overexploitation of jaguar populations range wide. In the late 1960s a high quality jaguar coat could be sold for as much as US $20,000 in New York, and 31,105 jaguar skins were imported into the United States from 1968 to 1970 (Myers 1973). During the early 1970s, approximately 15,000 jaguars were killed every year in Brazilian Amazonia alone to support the international skin t rade (Smith 1976; Fitzgerald 1989). Hunting related to fur trade was also an important mortality factor in Argentina until the country joined CITES in 1980 (Altrichter et al. 2006), and overhunting is considered the main factor responsible for the decline and decimation of jaguar populations in Venezuela (Hoogesteijn & Mondolfi 1987).
116 The national bans on jaguar hunting in most countries where jaguars occur and the international ban on trade of jaguar body parts are thought to have considerably reduced the number of jaguars killed for commercial motives (Smith 1976). Hunting, however, continues to be one of the main causes of decline of jaguar populations. Hunting is a major source of jaguar mortality in Iguau National Park in Southern Brazil, where at leas t 70 jaguars were killed by hunters between 1995 and 2002 (Crawshaw 1995, 2002; Conforti & Azevedo 2003) and also in Northern Misiones, in Argentina (Paviolo et al. 2008). In the semi arid Chaco of Argentina, high mortality from hunting is motivated by a d (Altrichter et al. 2006). In Brazil, hunting continues to be one of the main causes of decline of jaguar populations in all biomes of the country (Paula et al. 2011). In Amazonia, jagua rs are hunted for subsistence and illegal trade, given that law enforcement is difficult in isolated rural villages, which are consequently not subject to control (Reina and Gonzalez Maya 2008), and in Mamirau and Aman Sustainable Development Reserves, h unting of jaguars occurs regularly (Valsecchi, 2005). Despite the evidence that hunting can have a major impact on jaguar populations, few studies have estimated jaguar mortality from hunting or characterized the hunting events to determine when, where, ho w and why people hunt jaguars. Mortality of jaguars from hunting has been estimated by the number of radio collared animals killed by hunters (e.g., Crawshaw 1995) or through interviews with local people (e.g., Michalski & Peres 2006). The first method is logistically complex, time consuming and financially prohibitive for most research budgets. The second only allows researchers to access the minimum number of jaguars killed because
117 interviewees may forget, have no knowledge of, or consciously omit the oc currence of jaguar hunting events due to the illegality of this activity and the association of researchers to law enforcement. The minimum number of jaguars reported killed is an unknown fraction of the total number of jaguars killed and thus has limited application, serving only to set a minimum boundary on the real number of jaguars killed. Carvalho and Pezzuti (2010) were the first to use a statistical approach to go beyond the minimum number reported in interviews and attempt to estimate the true numbe r of jaguars killed. Using a first order jackknife (Manly 1997), they obtained an estimate of mortality of jaguars from hunting that was almost twice the minimum number of jaguars reported killed in their interviews. Their method, however, does not allow e stimation of the uncertainty associated with their estimates, nor does it allow the use of different models that take into account different detection probabilities assumptions. Fortunately, the problem of estimating the number of jaguars killed is very si milar to the methodological challenge of estimating the size of wildlife populations. In both cases a survey method yields a count statistic (e.g., minimum number of jaguars killed reported in interviews, or minimum number of jaguars identified during a ca mera trap survey) which represents an unknown fraction (p) of the population of interest. In the case of jaguar mortality from hunting, the estimate represents an unknown fraction of the total number of jaguars killed. For these count statistics to be usef ul it is necessary to estimate p so that the real parameter of interest can be estimated (i.e., total number of jaguars killed or abundance). Capture recapture (CR) models were specifically designed for estimating p and, subsequently, the real parameter of interest, abundance (or in the case of this study the total number of jaguars killed). This methodology is
118 robust, well developed and tested, and has shown to be effective for estimating abundance and density of large cats in many environments (Karanth 19 95; Karanth & Nichols 1998, 2000; Silver et al. 2004; Maffei et al. 2004; Soisalo & Cavalcanti 2006; Maffei et al. 2011). The similarities in the methodological challenge of estimating abundance using field surveys and estimating total number of jaguars ki lled using interviews suggests that CR models may also allow estimation of p for the case of mortality from hunting. The objectives of this study were t o characterize jaguar and puma hunting events, to estimate jaguar mortality from hunting using CR method ology in two Sustainable Development Reserves located in the Western portion of Brazilian Amazonia, and to evaluate the effectiveness of sustainable development reserves in controlling hunting of jaguars. In characterizing jaguar and puma hunting events I tested the following hypotheses: Hunting of large cats has a seasonal variation in its distribution and occurs more frequently during months and season when the water level is higher and terrestrial habitat is less available (i.e., the flood season, months of May July). The jaguar is hunted more frequently than pumas because jaguars occur at high density in the Vrzea flooded forest (Chapter 2) and are known for outcompeting and excluding pumas from areas closer to water bodies in areas where they co occur Male jaguars and pumas are hunted more frequently than females because male large cats usually have larger home ranges and disperse greater distances. Most hunting events will be opportunistic because killing of jaguars is often associated with livestoc k depredation and chance encounters near margins of lakes, streams and other water bodies where both humans and large cats search for food.
119 Methods Study Sites This study was conducted in two neighboring Sustainable Development Reserves (SDRs) in the Weste rn portion of Brazilian Amazonia, Mamirau and Aman SDRs (Fig. 3 1). SDRs are defined in the Brazilian National System of Conservation Units (SNUC) as natural areas inhabited by traditional human populations whose existence is dependent upon sustainable s ystems of natural resource exploitation and that play a fundamental role in the protection of nature and maintenance of biological diversity. The objective of this category of protected area is to promote the conservation of biodiversity and secure and imp rove the quality of life of traditional human populations (Brazil 2000). The climate in the region is tropical humid with a stable average monthly temperature around 26C and an average annual precipitation of 2,373 mm (Ayres 1993) with 64% of the annual p recipitation occurring during December May. The flood pulse of the Amazon River causes river levels in the region to fluctuate an average of 10.6 meters annually (Ramalho et al. 2009). The flooding caused by this fluctuation may inundate the entirety of Ma mirau Reserve and also the western portion of Aman Reserve. Mamirau is located at the confluence of the Japur and Amazon Rivers, but entirely encompassed within the Amazon River floodplain (Fig. 3 Created in 1990 Mamirau was the first SDR to be created in Brazil (Queiroz 2005, Esterci & Ramalho 2007). The Reserve covers an area of 11,240 km, and is the only official protected area in Brazil exclusively protecting the Vrzea ecosystem. The Vrzea is a wetland that is seasonally i nundated by the large and predictable monomodal flood pulse of the Amazon River and its tributaries (e.g., the average
120 annual amplitude of the flood pulse in Mamirau Sustainable Development Reserve is >10 m, Ramalho et al. 2009). Inundation forces the Vr zea to periodically transition between terrestrial and aquatic phases, resulting in a biota that has morphological, anatomical, physiological, phenological, and/or ethological adaptations to this flooding regime (Junk et al. 1989). Most frequently encounte red botanic families are Annonaceae, Euphorbiaceae, and Leguminosae (Ayres 1993). Some of the most frequent tree species found are Pterocarpus amazonicus, Eschweilera albiflora, and Piranhea trifoliate. Because Mamirau is a seasonally inundated island in the middle of two large rivers, animals that live inside the Reserve have to be well adapted to swimming and/or climbing trees and able to survive the steep annual fluctuation in resource availability caused by flooding. These peculiar environmental charac teristics are responsible for the presence of some endemic species but also for a lower density and diversity of terrestrial species in general. Primate diversity is lower than in the surrounding Terra Firme, but Mamirau Reserve encompasses most of the di stribution of the white bald headed uakari monkey ( Cacajao calvus calvus ) and the entire distribution of the endemic blacked headed squirrel monkey ( Saimiri vanzolinii ). Threatened and charismatic top predators such as the jaguar, the black caiman ( Melanos uchus niger ), the pirarucu ( Arapaima gigas ), and the Amazon river dolphin ( Inia geoffrensis ) are abundant. Mamirau Reserve also holds a diverse fish and bird fauna with at least 340 species of each group (Queiroz & Peralta, 2011). Aman is located between the black, nutrient poor, waters of the Negro River and the white, nutrient rich, alluvial waters of the Amazon River, also near the confluence of
121 the Japur and Amazon Rivers (Fig. 3 covers a mosaic of thr ee main ecosystems Terra Firme, Igap, and Vrzea covering an area of 23,500 km (for a detailed description of forest types see Prance 1978, 1980). The Terra Firme ecosystem encompasses all non floodable areas of the Amazon Forest and is the predomina nt ecosystem in Aman covering approximately 84% of the Reserve. The diversity and density of tree species is higher in the Terra Firme in comparison to the Igap and the Vrzea. The most abundant tree species in the Terra Firme environments of Aman are E schweleira coreacea, Iryanthera paraensis, Virola calophylla, Iryanthera sp., and Iryanthera juruensis (Souza 2006). Igap, like the Vrzea, is a seasonally inundated environment. The difference is that the Igap is flooded by black water rivers, while the Vrzea is flooded by white water rivers (Ayres, 1993). Igaps cover 9% of Aman and 6% of Vrzea. The mosaic of forest types found in Aman make it one of the most bio diverse forests in the world (Sears & Marn 2001), but detailed information on the faun a of the Reserve is still lacking. However, many charismatic, endangered, and ecologically important species are common in Aman, including the lowland tapir ( Tapirus terrestris ), Amazonian manatee ( Trichechus inunguis ), the harpy eagle ( Harpyia harpij a ), black faced uakari ( Cacajao melanocephalus ), giant river otter ( Pteronura brasiliensis ), pink river dolphin ( Inia geoffrensis ), and both the jaguar and the puma. The human populations of Mamirau and Aman live primarily along the margins of the main Rivers, on smaller channels, and along the margins of Lake Aman (Fig. 3 1). The main economic and subsistence activities of local people in both Reserves are agriculture, hunting, fisheries, and harvesting of wood and other non timber products.
122 But the i ntensity of these activities is markedly different. In Mamirau the principle source of protein comes from fishing, whereas hunting is the major source of protein in Aman. The human population in Mamirau is 9,733 people (0.87 people/km), distributed in 1684 households over 181 villages. In Aman there are 3,653 people (0.15 people/km), distributed in 612 households over 84 villages (MSDI 2011). This study covered only a portion of both Reserves and was conducted in areas with the largest concentration o f villages (Fig. 3 1). Characterizing Hunting Events To characterize large cat hunting events, I interviewed local people using a semi structured interview in which I recorded information on the hunting event and on the large cats killed. Villages visited were selected based on a logistically viable route through the study areas that would allow us to survey the largest number of settlements during the survey. Households were selected at random and people from the same household were not interviewed. The fi rst step I took after approaching a household was to explain the purpose of the study, to assure every interviewee that all information provided would only be used for the objectives of this study and that anonymity would be preserved for all participants. I started the interview asking subjects if they knew of any large cat events or stories. Events and stories could be a hunting event, but could also be a sighting, a depredation of livestock event, an attack on a person, a scat found, or an animal calling The idea of this first question was to get the interviewee more comfortable with the questionnaire and with the interviewer. During the interviews, for each hunting event I recorded the date (or month, or season, when interviewee could not remember date) name of location (e.g., Lake
123 Vrzea, Terra Firme, Igap, or river), whether the event occurred during day or night, the activity of the hunter at the moment of the event method used in the hunt, motivation/reason to hunt, if hunter was a livestock owner, if meat was consumed, and if parts were collected. The name of the hunter was obtained whenever possible but ideally it was given spontaneously by the interviewee. Altho ugh giving the name of the hunter was delicate information given the illegality of killing a large cat, this information was often key in identifying animals during analysis. For each large cat reported killed, I recorded the species, whether the animal wa s melanistic or not, sex, age class (i.e., cub, sub adult, adult), and reproductive state (pregnant or lactating). I also recorded any other information regarding the event that could facilitate the identification of that event and its distinction from oth er events (e.g., a jaguar was killed on the same day as the soccer tournament). Information provided also enabled me to determine whether the large cat), intentio accidental (i.e., the hunter killed a large cat involuntarily, such as might occur by mistaking a large cat for another species when hunting). To evaluate if hunting events had a season al distribution I compared the observed number of hunting events reported for each season with the expected number of hunting events based on the number of months within each season. For example, flood season lasts three months (25% of the year), and if se ason of the year had no effect on hunting events I would expect 25% of hunting events to occur within this season. To determine if there was a significant difference between observed and expected values I used Pearson's Chi squared test (Zar 1984). I also used Pearson's Chi squared test to
124 compare frequencies of hunting events for jaguar and puma, for males and females, and for opportunistic and intentional hunts. Estimating Total Number of Jaguars Killed using CR Methodology Capture recapture methodology ( CR) was originally designed to allow estimation of population size because the imperfect detection provided by available field survey methods did not allow a total count of animals in the population of interest (Karanth 1995; Karanth & Nichols 1998, 2002). In this study, my field survey method is represented by the interviews, and instead of abundance I am interested in the total number of jaguars killed. Like most field survey methods used to count the number of animals in a population (e.g., line transect s, point counts, camera traps, etc.) interviews used in this study to estimate the total number of jaguars killed are unlikely to detect all animals that were in fact killed (i.e., the probability of detection is less than one). Consequently, interviews yi eld an imperfect count, or a count statistic, that represents an unknown fraction of the total number of jaguars killed during hunting events. As in the estimation of abundance, to estimate the true number of animals hunted it is necessary to estimate this unknown proportion, the detection probability (p). When the count statistics of the survey method yield numbers of marked (captured) and unmarked (not captured) individuals, such as those resulting from camera trap surveys, abundance can be estimated thro ugh closed population CR models. The resulting data of a camera trap survey is made of individual capture l was not captured. For example, if animal #1 has a capture history of 00101, it means that it was
125 photo captured in the third and fifth sampling occasions, but was not photo captured in the first, second and forth sampling occasions. Each capture history allows estimation of the probability of observing that capture history based on p (Karanth & Nichols 1998). In this study, I was interested in estimating the total number of jaguars that were s reported by an interviewee as having been killed. I also replaced sampling occasions in the columns of capture histories, by interviewees. Therefore in my adapted capture histories, if a hunted jaguar had a capture history of 00101, it means that it was not detected in the interviews with interviewees 1, 2, and 4, but was detected in the interviews with interviewees 3 and 5. Setting up the data this way allows the estimation of p much the same way that it would be done in a study designed to estimate abun dance. To estimate the parameter of interest p from the individual capture histories I used software Capture (Otis et al. 1978; Rexstad & Burnham 1991). The software calculates p under different models that reflect different assumptions for p. There are two basic steps to this process. The first step is to create a model that states the probability of the data observed based on the parameter(s) of interest (i.e., p), which is a product of the probabilities of all capture histories recorded. The second ste p is to select the estimate of the parameter that maximizes the likelihood function. Both steps are done by Capture. The software calculates abundance, or in this case number of jaguars killed during hunting events, for each model and ranks them indicatin g the one that best suites the data set. Available methods include: Mo, no variation of p between individuals or sampling occasions; Mh, different p for each individual but constant over time (i.e.
126 interviewee); Mb, difference in p between captured and not captured individuals; Mt, difference in p between sampling occasions (Karanth & Nichols 1998). Capture can also compute estimates of abundance under models that allow two sources of variation in p (Karanth & Nichols 1998). In this study I assumed that hun ting events have different capture probabilities because they have different characteristics which make them more or less likely to be detected (e.g., a jaguar shot while depredating livestock near a village may have higher detectability than a jaguar shot in the forest). Therefore Mh is the most logical model to be use for estimates of the number of jaguars killed. Additionally, Otis et al. (1978) showed in simulations that when p varies by individual that Mo provides estimates of abundance that are signif icantly negatively biased. So even when Mo or another model was selected I also used Mh when both models were adequate for the data. For the analysis of the data I considered each interviewee a capture occasion and each group of 48 interviewees as a survey block, with a total of three blocks. These blocks were combined to form one final matrix with all identified jaguars killed making up the rows and the combined 48 sampling occasions making up the columns (Karanth & Nichols 2002). Interviewees of the three blocks were combined so that the first interviewee (i.e., sampling occasion) in all blocks formed sampling occasion number one, the second interviewee in all blocks combined formed sampling occasion two, and so forth. To estimate the area covered by this survey (i.e., area from which jaguars were harvested) I used the area defined by villages as their area of use within the Reserves plus a 9 km buffer around villages (Fig. 3 1, Fig 3 2). This 9 km distance is suggested to
127 be the furthest away from a villag e local people will explore non timber forest products in Amazonia (Peres & Lake 2003). Important Assumptions for Estimation of Total Number o f Jaguars Killed Surveys that use closed population CR models must adhere to several important assumptions for the analysis to be valid. Since I adapted the CR methodology for estimating jaguar mortality from hunting using an interview survey, I adapted CR assumptions accordingly. Assumption 1, and perhaps the most important assumption, is that I was able to accuratel y identify recaptures of the same individual from the information provided by interviewees. To guarantee this assumption was met I considered that a recapture occurred only when two or more events reported by different interviewees had matching information on most of the following characteristics of the hunting event: (1) species killed, (2) age class, sex, and whether the animal killed was melanistic, (3) name of hunter, (4) date (month, year, season) of event, (5) location and environment where event occu rred, (6) method of hunt, motive for the hunt, activity of the hunter, and whether the meat was consumed. Additional information given by the interviewee was also used to match events. Given the relative rarity of jaguar hunting events at any one location, the local fuss created by most jaguar hunting events, and the method of matching information described above, I believe that this assumption was well met in this study. Assumption 2 is that interviewees provided accurate information about hunting events t o the best of their knowledge and did not make up jaguar hunting events that did not exist. Because hunting of jaguars is illegal in Brazil, It would be unexpected that a person would describe a jaguar hunting event that did not occur and that could the
128 in terviewee himself and/or a neighbor. The omission of information may have affect this assumption, but the error associated with the omission of information was presumably minimized by interviewing multiple individuals from the same village. Assumption 3 w as of population closure, that is, no jaguars were killed during the period used to calculate the estimates of number of jaguars hunted. Although information characterizing hunting events included dates previous to January 2009, estimating jaguar mortality from hunting using CR methodology required defining a discrete sampling period that met the above assumption of a closed population. This assumption was met by defining an 18 month period (January 2009 to mid July 2010) that ended on the day of the first interview, which ensured that no jaguars killed during the interview period were included in the sample population. Results The survey was conducted over a 12 day period from the 15th to the 26th of July of 2010 and included 53 villages in Mamirau and Ama n Reserves (Fig. 3 2). A total of 144 people were interviewed and information on 257 large cat hunting events was reported. Of these events 239 were reported as confirmed kills and 18 events that a jaguar was wounded but escaped. Events dated from the 196 0s to 2010, with most of reported events occurring during 2009 for both species (Fig. 3 3). Most hunting events were reported as happening during the day (80%, n=243) (Table 3 1). Parts of dead animals (e.g., pelt, skull, teeth, claws, or fat) were collect ed as souvenirs, for medicinal purposes, or illegally traded in 130 events (61%, n=214). Distribution of Hunting Events a mong Environments Hunting events were reported in three types of environment: Vrzea, Terra Firme, and river. Most events reported occu rred in the Vrzea (56%, n=247), followed by Terra
129 Firme, (43%) (Table 3 2). Events in which the large cat killed was in a river were rare, only being reported three times. In all three cases that a large cat was killed in a river the species hunted was a jaguar, and in the two events where sex was identified individuals killed were females. One was killed in the Japur River and the other in the Amazon River. Seasonality o f Hunting Events Distribution of reported jaguar hunting events throughout the year had a marked seasonal distribution, being more frequent in the first two months of the flood season (May and June; Fig. 3 4), which represented 45.6% of the total number of events where month was identified (n=79). Seasonality was also observed when events were grouped into seasons (Table 3 1; Fig. 3 5). The number of events reported in each season (i.e., observed events) was significantly different from expected under a null hypothesis of no seasonal variation ( = 22.52, df = 3, p < 0.01). Jaguar hunting events occurred almost twice as often as expected during the flood season indicating a concentration of events during this season. On the other hand, the frequency of events was almost half of the expected during the rising season. Hunting events during d rought and lowering season occurred at approximately the same frequency as expected. Seasonality in the distribution of puma hunting events was not observed in the analysis by month, probably due to the small number of events recorded during this survey (F ig. 3 6). When grouped into seasons, the distribution of puma hunting events was similar to that of the jaguar (Fig. 3 7), with events occurring more than expected during the flood season and less than expected during the rising season. However,
130 difference s between observed and expected number of events per season was not significant ( = 5.78, df = 3, p = 0.12) Hunting Pressure on Jaguars and Pumas Jaguars were reported hunted significantly more frequently than pumas ( = 60.66, df = 1, p < 0.01). Of all events, 83% were related to the jaguar and only 17% to the puma (n=257). Most jaguars were killed in the Vrzea (63.6%, n = 206), while pumas were killed more frequently in the Terra Firme (80.5%, n = 41). This pattern was consistent throughout all season s, with the exception of the lowering season when jaguars and pumas were killed in approximately the same proportion in the Terra Firme and the Vrzea (Fig. 3 8). Hunting Pressure on Males and Females Males from both species were killed more often than fem ales, with an overall ratio of 1.7:1.0 (Table 3 1). Males and females were killed in similar proportions for both species, but the ratio of males to females killed was slightly higher for pumas, 1.9:1.0, than for jaguar, 1.7:1.0. Males were hunted more fre quently than females in the Vrzea and Terra Firme, but the ratio of males to females hunted was higher in the Vrzea, 2.1:1.0, when compared to the Terra Firme, 1.4:1.0. This pattern was maintained when jaguar were analyzed alone, but not for pumas. Pumas where hunted in the Vrzea only five times, and all individuals killed were males. Opportunistic Versus Intentional Hunting The majority of reported events was identified as opportunistic (57.7%, n = 234), intentional hunting events accounted for 96 event s (41%), and events which occurred by accident were reported only three times. Most hunting events of both species were opportunistic, but the proportion of intentional hunts was higher for jaguars, 45.1%, than
131 for pumas, 20.5%. Opportunistic events were m ore common in the Vrzea, but proportions were similar between Vrzea and Terra Firme. Hunting Method Shotguns were the most common method used to kill large cats, and were used in almost 80% of all reported hunts (Table 3 1). Shotgun events were associate d with hunting dogs in 67 cases (35.1%, n = 191). The second most frequently used method was a harpoon in association with another weapon (usually a large club), which accounted for 13.6% of all kills. Although both species were killed with shotguns, only jaguars were killed with harpoons (Table 3 1). Harpoons are used by local people to fish and consist of a wood or bamboo spear with a metal tip that is attached to the spear by a fine rope, and that comes off once the harpoon hits its target but remaining attached to the spear by the rope. When people harpoon a jaguar, they usually throw one or more harpoons at the jaguar (which is usually in the water) and tie the spear to a tree, long stick or other weapon. Shotguns were the most common weapon used to kill large cats in the Vrzea and Terra Firme, but harpoons were rarely used in the Terra Firme. Shotgun traps and other methods (e.g., callback) were also rarely used. Motive o f Hunt The main reported motive for hunting large cats in the Vrzea and the Terra Firme was depredation of livestock, which was reported for 120 events (47.1%)(Table 3 1). Of these events, the most frequently predated species of livestock was the pig, 62 event s (57.7%), followed by cattle, 21 events (17.5%), chicken, 14 events (11.7%), dog, 13 events (10.8%), sheep, 8 events (6.7%), duck, 5 events (4.2%), and buffalo, 2 events
132 which was reported for 111 times (43.5%). This category of answer encompasses cases in which the hunter had no apparent reason for hunting the animal, but also cases where the interviewee did not know or did not want to say the motive of the hunt. Unfo rtunately, these two cases cannot be separated. Other motives reported were attacks on humans, commerce (i.e., illegal trade of meat or body parts), and others (e.g., medicinal purposes). Depredation was the most frequent motive for hunting jaguars and pum as. Most frequently depredated livestock species, however, differed between jaguars and pumas. Jaguars depredated on pigs in 58% (n=100) of cases, followed by cattle, 18%, while pumas were divided between chicken, 35.7%, sheep, 28.6%, and pigs, 28.6% (N=14 ). Activity of t he Hunter The most frequent activity of the hunter at the time of the hunt was hunting large cats, reported in 87 cases (39.5%, n=220), followed by subsistence hunting of other animals, 56 cases (25.5%), fishing, 45 cases (20.5%), and other activities (e.g. traveling, herding, resting at home), 32 cases (14.5%). Hunting large cats was the most frequent activity of hunters in both environments, followed by subsistence hunting in the Terra Firme and fishing in the Vrzea. Hunter activity diffe red between events relating to jaguars and pumas. While in most jaguar related cases hunters where out specifically to hunt jaguars (43.8%), in puma hunts in most cases hunters where hunting other species for subsistence when the event occurred (60%). Cons umption of Meat The meat of large cats killed was consumed and/or sold for consumption in 96 cases (43.05%, N=223). Consumption of large cat meat was reported in a relatively high
133 proportion of cases for both species, but jaguar meat was consumed more ofte n than that they eat this kind of meat, it is likely that meat of large cats is actually consumed more than reported. Estimates of Total Number of Jaguars Killed To estimate the number of jaguars killed, I used data from January 1st 2009 to July 14th 2010 to guarantee population closure and more accurate estimates, since detection rates were expected to decrease with time. When I analyzed that for this whole period, 18.5 mont hs, model Mh, where capture probabilities are heterogeneous between jaguars killed, was selected by Capture as the best model for the data. The statistical test in program Capture confirmed population closure (z= 1.144, P=0.126) and the estimated number of jaguars killed was 108 (SE=16.85; p hat = 0.021), with a 95% confidence interval of between 86 and 154 individuals (Table 3 3). In the analysis of year 2009 only, model Mh was also selected and population closure was also confirmed (z= 0.217, P=0.414). Th e number of jaguars estimated to have been killed in 2009 was 73 (SE=15.55, p hat=0.018), with a 95% confidence interval of between 54 and 118 individuals (Table 3 3). In the analysis of the portion of 2010 that was surveyed, model Mo was selected by Captu re. However, given the characteristics of hunting events Mh is a more appropriate model for the data of this study because hunting events have different capture probabilities depending on their characteristics. This was also a reasonable step given that th e model criteria should only a small difference in AIC between these two models. For data from 2010 Capture indicated lack of population closure (z= 1.875, P=0.030). The number of jaguars estimated to have been killed in 2010 was 29 and 33, with Mo and Mh respectively (SE=3.28 and 5.03, p
134 hat = 0.035), with a 95% confidence interval of between 25 and 38 individuals for Mo and between 27 and 48 for Mh (Table 3 3). The minimum number of jaguars killed in 2009 and 2010 was respectively 38 and 24, and the minim um number for the period was 62 animals (Table 3 3). These represent 52.1%, 72.1%, and 57.1%, of the total number of jaguars killed estimated using CR models. Discussion Amazonia is recognized as the most important biome for the long term conservation of N eotropical felids (Oliveira 1994) due to its large area, connectivity, preservation status, low human density, and proportion of area inside reserves. These attributes help maintain wild cat populations, but do not impede hunting, especially in the Vrzea Aman Reserve), where most of the human population of Amazonia is concentrated (Goulding et al. 1996). Both jaguars and pumas are frequently hunted by local people in the Vrzea floodpl ain forests and neighboring Terra Firme forests of central Brazilian Amazonia. However, hunting pressure on the jaguar is of particular concern since the hunting events involving the jaguar represented 82.8% of all events reported, 4.8 times more than puma s. This finding is corroborated by the data from the community based Fauna Use Monitoring System of Mamirau Sustainable Development Institute, which found similar results for the same period and the same area (J. Valsecchi, unpublished data). This large d ifference in the number of jaguars versus pumas killed by local people, however, does not necessarily indicate preference, or higher dislike, of hunters for jaguars but is most likely a result of the larger abundance of jaguars in the floodplain
135 and its tr ansition zone with the neighboring Terra Firme (Chapter 2), and consequently a higher encounter rate of jaguars and people, higher frequency of conflicting interactions, hence larger number of jaguars killed. The higher encounter rate can be exemplified by the camera trap surveys conducted in Mamirau Reserve, which have recorded over 100 pictures of jaguars in the last six years, but only one of puma (Emiliano Ramalho, unpublished data). At the same time, the only preference that was mentioned by interview ees was that puma meat tasted better than that of jaguars, so that if there was preference it would be to hunt pumas. This higher abundance of jaguars in the floodplain is also corroborated by other studies which have reported that jaguars tend to exclude or reduce abundance of pumas in area closer to water bodies (Crawshaw & Quigley 1991). In summary, jaguars are hunted more frequently because they are in conflict with people more frequently. Jaguar hunting events were markedly seasonal, not only when obse rved by the number of events per month, but also when comparing number of events per season, corroborating hypothesis 1. The seasonality of hunting events observed, with most hunting events occurring during the flood season, may be explained by the large v ariation of the water level in the region due to the flood pulse regime of the Amazon River and its tributaries, which causes the flooding of large extents of forest including virtually the whole area of Mamirau Reserve. Flooding reduces the terrestrial h abitat and forces the lateral migration of terrestrial animals to higher ground (Junk 1989). Flooding also causes a natural seasonal variation in prey availability because spectacled caiman ( Caiman crocodiles ) are scattered in the growing aquatic habitat, eggs of black caiman ( Melanosuchus niger ) spectacled caiman are not available, and
136 access to other important prey, such as the sloth, becomes limited because area for foraging is restricted by flooding. This may force at least a portion of the jaguar popul ation to migrate from the Vrzea to the Terra Firme in search of dry ground and prey. This lateral migration also causes them to cross paths with people and their livestock. This finding has great importance in directing future conservation actions in Amaz onia and optimizing resource allocation, since optimal efforts to reduce conflict, hence number of jaguars killed, should be focused in the flood season when most jaguars are killed. The higher frequency of males killed is most likely the result of higher encounter rates between jaguars and people. Males are expected to move larger distances than females throughout the year and this may include lateral migration from the Vrzea to the Terra Firme during flooding. This would also be consistent with sexual se gregation theory, which predicts that males in polygynous species use areas during non breeding periods where they can maximize body condition in preparation to compete for mates, whereas females select areas that maximize offspring security (Main et al. 1 996, Main 2008). Female jaguars have been documented raising offspring in the Vrzea (Chapter 2), and have been documented to remain in the Vrzea during flooding (E. Ramalho, unpublished data), which likely restricts their movement patterns during floodin g and reduces encounter rates with people. Hunts were opportunistic in almost 60% of cases, which indicates that most hunting events were not directly related to a depredation event but rather were carried out as a preventive measure against depredation of livestock, attacks on people, for food, and/or status. On the other hand, the main motivation for killing large cats was
137 depredation of livestock. Pigs, cattle, chicken and dogs being the most often predated livestock species. Poor management of livestock is common in Amazonia, and improving management of at least pigs and cattle could have a significant impact on the number of jaguars killed. Although none of the interviewees identified subsistence hunting as a motivation for hunting large cats, the meat of kills was consumed in over 40% of reported events. This contradicts other studies which have indicated that jaguars and pumas are rarely killed by subsistence hunters (Yanez et al. 1986, Cunningham et al. 1995, Hoogesteijn et al. 1996, Novack et al. 200 5). It also indicates that large cat meat is considered by many local people as an opportunistic source of protein. The estimate of total number of jaguars killed from January 2009 to mid July 2010 (n = 108, 95% CI: 86 154) raises questions about the effec tiveness of current conservation measures in protecting large cats in Brazilian Amazonia. The first issue well guarded Amazonian protected areas, Mamirau and Aman SDRs. Under the co management of Mamirau Sustainable Development Institute and the Amazonas State Government these Reserves have one of the largest law enforcement budgets and logistical support of any protected area in Amazonia. Mamirau Institute has a lso implemented various capacity building and environmental education actions with the local people during the last 15 years since the creation of the Reserves and local people have had the chance to interact with researchers and participate in research. It is therefore clear that sustainable use protected areas, under current models of management, are not effective at preventing the hunting of jaguars.
138 On perhaps more positive notes, females were killed less frequently than males, and the density of jagu ars in the Vrzea portion of the study area has been maintained in the last five years, despite the heavy hunting pressure. Hunting may be shifting the male to female ratio in the Vrzea, which is relatively low in Mamirau, 0.79:1.0 (Chapter 2), but does not seem to be influencing population size. This may be explained by the compensatory mortality theory which predicts that the more frequent harvest of adult males, which comprised more than 60% of all hunting events recorded in this study, will reduce int ra specific competition and trigger positive density dependent responses in reproduction and survival of offspring and females (Connell 1978). Although SDRs have the goal of maintaining wildlife populations, they were not designed to reduce hunting of jagu ars, and consequently are not effective for that goal. Other mechanisms have been put in place for that purpose. The national ban on hunting of jaguars in Brazil is a coercive measure; it prohibits people from hunting jaguars because the Federal government believes it is important to conserve jaguar populations. This measure was effective at reducing commercial hunting of jaguars because it was associated with another coercive measure, the international ban on trade of jaguar parts. Commercial hunting of ja guars was thus effectively controlled in Brazil, not only because it became nationally illegal, but also because market demand was drastically reduced, hence economic motivation for hunters to pursue jaguars was also reduced. However, as shown by the large number of jaguars killed in two well protected Reserves, these measures have little effect on the current motivations for killing jaguars in the Brazilian portion of Amazonia.
139 The national ban on hunting of wildlife in Brazil since 1967, and the inclusion of the jaguar in Appendix 1 of CITES in 1973 are also ineffective towards contemporary hunting of jaguars in Amazonia. Although they have had an important role in stopping commercial hunting and trade (Smith 1976) in the past, and are still effective toda y, they are inadequate against current hunting of jaguars, simply because the main motivations for hunting these large cats have changed. When these two conservation actions were created and implemented, the main motivation for killing jaguars was commerci al, meaning that people went out of their way and actively searched for jaguars to kill and make a profit. Today, the main motivations for killing jaguars in Amazonia are the economic loss imposed on local people by livestock depredations. Also, fear of ja guars, which are culturally portrayed as dangerous, treacherous and powerful animals, may also play a role as suggested by large number of opportunistic hunting events that had no clear motivation. Other conservation actions, such as payment of local peopl e by the Amazonas State Government for environmental services (e.g., Bolsa Floresta), which started in September of 2007, also have little chance to affect hunting since there is no accountability from local people on their performance in conservation, inc luding whether they killed jaguars or not. Of even greater concern should be the possibility of increased hunting pressure on jaguar populations associated with the proposed changes in the forest code of Brazil, which are currently being voted in congress and may reduce the amount of protected areas in Amazonia considerably, which itself is a threat to large carnivores. Most hunting of jaguars today seems to occur in the Vrzea floodplain forests and neighboring Terra Firme forests, where most of the human population of Amazonia is
140 concentrated. And the only reason that hunting pressure has not caused the extinction of jaguars in these areas seems to be the input of individuals from the continuous inhabited Terra Firme forests, and probably increased surviva l of females due to their ecology. If roads are built and affect these sources and if prey in the Vrzea decreases populations of Amazonia. Conclusions The use of CR models to estimate total number of animals killed is a promising method to obtain estimates of mortality from hunting on endangered charismatic species like jaguars and pumas. Careful attention should be taken, however, in guaranteeing that assumptions are met, e specially in the identification and matching of hunting events reported by different interviewees. Using such models provide an objective and robust method for estimating the extent of mortality from hunting and identifying the need for conservation strate gies to address such losses. It is clear to me that jaguars and pumas are ineffectively protected from hunting in Brazilian Amazonia, despite protected areas, the national ban on hunting, the international ban on trade, environmental education, and payment for ecosystem services. New conservation measures that take into account the current motives for killing jaguars and pumas must be designed and implemented if hunting pressure on these large cats is to be reduced along the Amazon River floodplain forests and neighboring Terra Firme. In doing so, it is imperative that the human values of stakeholders involved, especially those that interact with jaguars and pumas directly (i.e., local people), be taken into consideration and that all stakeholder groups part icipate as wholly as possible in the conservation process (Chapter 1). Their
141 participation and consideration of how this conservation process influences their human values is indispensable to the success of new conservation actions.
142 Table 3 1. Characteri zation of all reported hunting events, and by species All events P. onca P. concolor N % N % N % Total 256 212 82.8 44 17.2 Environment 247 206 41 Vrzea 139 56.3 131 63.6 8 19.5 Terra Firme 105 42.5 72 35.0 33 80.5 River 3 1.2 3 1 .5 Season 225 187 38 Drought 52 23.1 42 22.5 10 26.3 Rising 51 22.7 44 23.5 7 18.4 Flood 102 45.3 88 47.1 14 36.8 Lowering 20 8.9 13 7.0 7 18.4 Sex 190 159 31 Male 119 62.6 99 61.1 20 64.5 Female 71 37.4 60 37.0 11 35.5 Male:Female ratio 1.7 1.7 1.8 Method of hunt 241 200 41 shotgun 124 51.5 100 50.0 24 58.5 shotgun and dogs 66 27.4 52 26.0 14 34.1 harpoon 33 13.7 33 16.5 0 0.0 shotgun trap 9 3.7 9 4.5 0 0.0 bush knife 5 2.1 4 2.0 1 2.4 other 4 1.7 2 1.0 2 4.9 Type of hunt 233 195 38 opportunistic 134 57.5 106 54.4 28 73.7 intentional 96 41.2 88 45.1 8 21.1 accidental 3 1.3 1 0.5 2 5.3 Motive of hunt 254 211 43 depredation 120 47.2 104 49.3 16 37.2 attack on human 13 5.1 10 4.7 3 7.0 commerce 8 3.1 8 3.8 0.0 other 3 1.2 3 1.4 0.0 not identified 110 43.3 86 40.8 24 55.8 Activity of hunter 219 185 34 hunting large cats 87 39.7 81 43.8 6 17. 6 subsistence hunting 55 25.1 35 18.9 20 58.8 fishing 45 20.5 41 22.2 4 11.8 other 32 14.6 28 15.1 4 11.8
143 Table 3 1. Continued All events P. onca P. concolor N % N % N % Meat consumed 222 183 39 yes 96 43.2 81 44.3 15 38 .5 no 126 56.8 102 55.7 24 61.5
144 Table 3 2. Characteristics of hun ting events by environment type Vrzea Terra Firme River N % N % N % 139 56.3 105 42.5 3 1.2 Species 139 105 3 P. onca 131 94.2 72 68.6 3 100.0 P. concolor 8 5.8 33 31.4 Season 131 87 3 Drought 27 20.6 22 25.3 2 66.7 Rising 30 22.9 19 21.8 Flood 64 48.9 36 41.4 1 33.3 Lowering 10 7.6 10 11.5 Sex 105 73 2 Male 71 67.6 42 57.5 Female 34 32.4 31 42.5 2 100.0 Male:Female ra tio 2.1 1.4 Species*Sex 100 49 2 P. onca male 66 66.0 28 57.1 P. onca female 34 34.0 21 42.9 2 100.0 1.9 1.3 5 24 P. concolor male 5 100.0 14 58.3 P. concolor female 10 41.7 1 Method of hunt 130 101 3 shotgun 70 53.8 52 51.5 shotgun and dogs 26 20.0 37 36.6 harpoon 25 19.2 3 3.0 3 100.0 shotgun trap 4 3.1 5 5.0 bush knife 3 2.3 2 2.0 other 2 1.5 2 2.0 Type of hunt 129 94 3 opportunistic 78 60.5 49 52.1 3 100.0 intentional 49 38.0 44 46.8 accidental 2 1.6 1 1.1
145 Table 3 2. Continued Vrzea Terra Firme River N % N % N % Motive of hunt 138 104 3 depredation 60 43.5 55 52.9 attack on h uman 7 5.1 6 5.8 commerce 7 5.1 other 3 2.2 not identified 61 44.2 43 41.3 3 100.0 Activity of hunter 122 86 2 hunting large cats 46 37.7 37 43.0 subsistence hunting 26 21.3 27 31.4 fishing 33 27.0 8 9.3 2 100. 0 other 17 13.9 14 16.3 Meat consumed 121 95 1 yes 67 55.4 28 29.5 1 100.0 no 54 44.6 67 70.5
146 Table 3 3. Results from Capture analysis for the estimation of total number of jaguars killed. Minimum number of animals killed, total number of captures, number of sampling occasions, best model selected by Capture, detection rates (p hat), estimated total number of jaguars killed, standard error of estimate (SE), 95% confidence inte rval (95% CI), and closure test Period Animals captured (minimum number of jaguars killed) Total number of captures Sampling occasions Model selected p hat Estimated number of jaguars killed SE 95% CI Closure test z value p value 2009 38 63 48 M h 0.018 73 15.55 54 118 0.217 0.414 2010 24 48 48 M o & M h 0.035 29/33 3.28/5.03 25 38/27 48 1.875 0.030 2009 2010 62 111 48 M h 0.021 108 16.85 86 154 1.144 0.126
147 Figure 3 1. Smaller map shows location of the study area within Brazil. Larger map shows the limits of Mamirau and Aman Sustainable Devel opment Reserves (white lines) and the area covered during the hun ting survey (white shaded area) Lake Aman
148 Figure 3 2. Area surveyed in Mamirau and Aman Reserves (white shaded area) and villages visite d during survey (black circles)
149 Figure 3 3. Number of large cat hunting events reported per decade until the 1990s, and per yea r between 2000 and 2010 (n=179)
150 Figure 3 4. Number of jaguar hunting events reported per month (n=79) (designated by bars), and mean monthly water level (MMWL, designated by line) in th e study area from 1990 2008 (data from Ramalho et al. 2009)
151 Figure 3 5. Number of jaguar hunting events observed and expected per season (n=187)
152 Figure 3 6. Number of puma hunting events reported per month (n=17) (designated by bars), and mean mont hly water level (MMWL, designated by line) in the study area from 1990 2008 (data from Ramalho et al. 2009)
153 Figure 3 7. Number of puma hunting events observed and expected per season (n=38)
154 Figure 3 8. Number of hunting events recorded per seas on of the year and environment type for jaguars (A) and puma (B) (n=182 and 37, respectively) A B
155 CHAPTER 4 THE IMPORTANCE OF CAIMANS AND ARBOREAL MAMMALS IN THE DIET OF THE JAGUAR ( PANTHERA ONCA ) IN THE VRZEA FLOODPLAIN FORESTS OF AMAZONIA. Felids are exclus ively carnivorous animals, and, as such, their survival, and a large portion of their ecology and behavior, depends on the prey species they consume. In turn, felid population size and density, population structure, and social behavior are largely determin ed by prey abundance and biomass (Pierce et al., 2000; Carbone & Gittleman 2002; Karanth et al. 2004). Therefore, information about diet is fundamental to understanding the ecology and behavior of felids in a specific habitat and should form the basis for the development of sound conservation actions. Jaguar (Panthera onca) feeding habits have been investigated in various regions the Atlantic Forest (Crawshaw 1995; Facu re & Giaretta 1996; Leite 2000; Garla et al. 2001; Crawshaw et al. 2003), in the Cerrado (Silveira 2004), in the Pantanal (Schaller & Vasconcelos 1978; Dalponte 2002), and in the Caatinga (Olmos 1993). Feeding habits of the jaguar were also described in su b humid forests (Aranda & Snchez Cordero 1996; Aranda 1993, 1994) and in dry deciduous forests (Nuez et al. 2000) in Mexico, in the Peruvian Amazon (Emmons 1987), in the Llanos of Venezuela (Polisar et al. 2003; Scognamillo 2003), in Paraguayan Chacos (T aber et al. 1997), in humid subtropical forests in Belize (Rabinowitz & Nottingham 1986) and Costa Rica (Chinchilla 1997), and in low tropical forests in Guatemala (Novack et al. 2005). These studies have shown that the jaguar is an opportunistic predator, consuming prey in the proportion of their availability (Rabinowitz & Nottingham 1986; Emmons 1987). They have also shown that the jaguar has great ecological adaptability (Rabinowitz &
156 Nottingham 1986), consuming over 85 different species of prey (Seymour 1989). Despite this large number of prey reported as being consumed by jaguars, studies conducted to date show that terrestrial mammals of large and medium size are the most frequently consumed prey species in most environments (Oliveira 2002), although other mammals, reptiles and birds may also be important items of the jaguar diet (Emmons 1987; Ramalho 2006; Da Silveira et al. 2010). Despite the large number of diet studies that have been conducted, there is a lack of information regarding diets of jagu ars in Amazonia, where jaguar feeding habits are largely unknown (Fig. 4 1). The Vrzea Floodplain Forests (hereafter, Vrzea) of Amazonia are an area of high density of jaguars (Ramalho 2012 Chapter 2), which suggests abundant prey populations. The Vrz ea, however, is a seasonally flooded environment, where only animals that are arboreal and/or have good swimming capacity are able to survive (Ayres 1993). The flooding dynamics makes Mamirau Reserve, which is a protected Vrzea Floodplain Forest site in Amazonia, an unlikely home to terrestrial mammals. In two years of transect surveys conducted in the Reserve, terrestrial mammals were observed only once (Santos 1996), and in five years of camera trap surveys, with the notable exception of jaguars, terres trial mammals were never observed (E. Ramalho, unpublished data). Consequently, the most commonly reported prey species of the jaguar are not available in Mamirau Reserve and the objective of this study was to au Reserve. Methods Study Area Mamirau Sustainable Development Reserve (hereafter, Mamirau Reserve), is located in the western portion of Brazilian Amazonia, approximately 30 km northwest
157 67 2). The Reserve is delimited by the Japur and Amazon Rivers, and the Auati paran channel, and encompasses an area of 11,240 km of Vrzea forests (6.25% of the total area of the Vrzea ecosystem in Amazonia). It is the largest protected a rea exclusively dedicated to protecting this type of environment. The climate in the region is tropical humid with average annual precipitation of 2,373 mm (Ayres 1993). This study was conducted in a 566 km area in the Southern part of the Reserve, around Lake Mamirau (Figure 4 2). For a more detailed description of the study area see Ramalho (2012 Chapter 2). Collection and Analysis of Scats Scats and carcasses of prey were collected opportunistically along the margins of lakes and trails during the lo w water season which occurs between September December (Ramalho et al. 2009). Jaguar scats were identified by size and width and tracks found near the place where the sample was collected. Scats of adult jaguars are generally larger than 19 mm in width (Fa rrel et al. 2000). I was confident that these scats were from jaguars and not from pumas ( Puma concolor ) because during this period I recorded jaguars ~100 times using camera traps and food snares (Ramalho 2012 Chapter 2), but never recorded pumas. Carca sses of prey found were identified to species in the field or brought back to Mamirau Institute for identification when necessary. Identification of remains was conducted using the same procedure described below for scats. For caiman (i.e., Melanosuchus n iger or Caiman crocodilus ), total length of animal was measure when possible or estimated from the size of the head.
158 All scats were dried in the sun, stored in hermetical containers and frozen until analysis. For the analysis of content, scats were sifted in running water and then dried. Fur, bones, nails, scales, feathers and other undigested remains were separated and analyzed macroscopically, with magnifying lenses and a microscope when necessary. Identification was done through comparison with a refere nce fauna collection from Mamirau Reserve. Caiman and other reptile scales were identified to species by a group of herpetologists from the Federal University of Amazonas with extensive experience in Amazonian reptiles, or when identification was not poss ible, categorized as unidentified caiman species. To determine the importance of each prey species in the diet of the jaguar, I calculated several parameters. The frequency (F) that each prey species was identified in scats was calculated by dividing the number of scats a prey species was identified by the total number of scats collected. To facilitate comparison with other studies, I also calculated percent occurrence (Po) for each prey species by dividing the number of individuals of each prey species ob served in the sample of scats (n) by the total number of individuals of all species found in the sample (T), multiplied by 100 (Po = n/T x 100) (Ackerman et al., 1984). To estimate the biomass contributed to jaguar diet be each prey species indentified in scats, I obtained average weights of all prey species and used a correction factor developed by Ackerman et al. (1984), represented by the linear relation Y = 1.98 + 0.035X (where Y = biomass of the prey species consumed, and X = biomass of the prey in kil ograms). I did not use this correction factor for prey species with average body weights of <2 kg because I assumed that the occurrence of these small species in a scat represented a whole individual (Ackermann et al. 1984). The
159 relative biomass of each pr ey species consumed was then calculated using equation (Equation 4 1) Results A total of 78 jaguar scats were collected in Mamirau Reserve between 2004 and 2010. From these scats a total of 142 individual prey, belonging to 10 species and several combined species categories from classes Mammalia, Reptilia, and Aves (Table 4 1). Evidence of plant and invertebrate ( Pomacea spp.) were also documented. An average of 1.68 prey species was obser ved per scat. Mammals were the most frequently consumed prey class representing 55% of all prey items observed, and also contributed the most to the total biomass consumed, 52.1% (Table 4 1). All identified species of mammals consumed were arboreal, with t he exception of one instance of domestic cattle ( Bos taurus ). Reptiles were also important representing 42.1% of all prey items observed and 47.2% of the biomass. These two classes alone represented ~97% of all prey items and >99% of the total biomass. The two species of prey most frequently consumed were the brown throated three toed sloth ( Bradypus variegatus ), present in 53% of samples, and the spectacled caiman ( Caiman crocodilus ), present in 41% (Table 4 1). These two species represented 30% and 24% of all items and 33% and 31.7% of the total biomass, respectively. It is important to note, however, that 13 caiman samples could not be identified to species, but given the ratio of spectacled caiman consumed by jaguars in comparison to black caiman (32:1), it is likely that these unidentified caiman were
160 spectacled caiman. If this was the case, spectacled caiman would be the most frequently consumed prey (34%) and also represent the largest part of the biomass (44.6%) consumed by jaguars in Mamirau Reserve The lesser tamandua ( Tamandua tetradactyla ) and the red howler monkey were also found relatively frequently, and represented 8% and 7% of all prey items respectively. Birds were present in only four samples and were insignificant diet components. Remains of freshwater snails from the genus Pomacea were found in four samples and vegetable matter in two. We believe that the presence of snails in scats is associated with the consumption of caiman and other reptiles which may eat snails. Consumption of vegeta ble matter, often grass, is a common behavior in felids and is thought to help the digestive system and elimination of feces. Vegetable matter could also have been ingested by prey and appear in the sample because of that, rather than the jaguar intentiona lly eating it. Carcasses were also used in identifying prey of jaguars. All carcasses found were from the same species identified in scats and included 5 spectacled caiman, 5 black caiman, and 2 brown throated three toed sloths. Caiman carcasses were eas ily identified because the hard armor and head of caiman were not consumed. All carcasses of spectacled caimans were adult individuals with an average total length of 1.4 m. Carcasses of black caimans included one juvenile and one adult individual, with to tal lengths of 1.5 and 3 m, respectively. Both black caiman carcasses were found on the same day and were 5 m apart. Both black caiman carcasses were fresh when discovered, and the signs of struggle in the flooded grass a few meters from where the carcasse s were found suggests these prey were killed and not eaten as carrion.
161 Carcasses of the lesser tamandua and of red howler monkeys were not found, probably because jaguars consume the whole animal. In the case of sloths, the carcass can be identified from t heir claws, which are not consumed. Discussion The diet of the jaguar in Mamirau Reserve is almost entirely composed of reptiles and arboreal mammals, which represented >95% of all prey items consumed. The diet of the jaguar in Mamirau is largely depende nt on four species of prey: the brown throated three toed sloth, the spectacled caiman, the lesser tamandua, and the red howler monkey. Although all of these species have been reported in jaguar diets in other environments, they usually represent a small percentage of prey consumed. Arboreal mammals and reptiles together have never been reported to represent more than 32.5% of prey items, and, although reptiles are usually more important in jaguar diets in flooded environments, rarely in other environments do reptiles contribute to in the Amazonian Vrzea differ radically from diets of jaguars reported from any other environment. These results can be explained by the a bundance of aquatic and arboreal prey and the absence of terrestrial prey at Mamirau Reserve. Whereas terrestrial mammals are rare, caimans are abundant. Total caiman density in Lake Mamirau, the center of the study area, is estimated at 23.7 individuals / km, or 230 caimans/km of lake margin (Da Silveira 2002). This equates to a total population estimate of 13,340 caimans in Lake Mamirau, and the surrounding lakes and other waterways in the immediate study area. The most abundant species of caiman in th is area is the black caiman, which represents 81% of the total caiman population. Spectacled caimans are less abundant
162 and have been estimated to occur at densities of 4.4 individuals/km, or 43 individuals/km in that same area and compose the other 19% of the caiman population (Da Silveira, 2002). Based purely on abundance, it would be expected that black caimans might occur more frequently in the diet of jaguars, but evidence of predation on this species was only found once. Spectacled caiman are found on land much more frequently and for longer periods of time than black caimans, making this species more available and vulnerable to jaguar predation. Spectacled caiman are also much smaller than black caimans (Rebello & Magnusson 2003), which should make th em easier and safer for jaguars to capture. Jaguars are also known to eat the eggs of both species of caimans. A survey of caiman nests in the same area of this study reported that between 12 27% of all caiman nests surveyed were predated by jaguars (Ramal ho 2006; Da Silveira et al. 2010). However, jaguars do not typically eat the egg whole and the presence of eggs in the diet of jaguars is undetectable in scats and cannot be compared to other diet items. Arboreal mammals usually do not represent a large po The greatest representation of arboreal mammals reported in the diet of jaguars was 14% in the Atlantic Forest in Brazil (Garla 2001). Even in areas where the abundance of arboreal mammals is greater than reported for Mamirau R eserve, arboreal mammals represent a small percentage of jaguar diets than found in this study. In the Llanos of Venezuela, for example, the density of red howler monkeys can reach >100 individuals/km but no evidence of this prey species was reported in t he diet of jaguars from this area (Scognamillo et al. 2003). Instead, terrestrial mammals represented 83% of the identified prey. Presumably, the concentration of terrestrial mammals in jaguar
163 diets represents a higher encounter rate and less difficulty in capture, which results in a better cost/benefit ratio in energetic terms. The brown throated three toed sloth occurs at high densities in Mamirau Reserve and its population is estimated to be over 100,000 animals (Queiroz, 1995), which is ~43% of the tot al individuals of all other species (Table 4 2). This high density may explain why three toed sloths appear so frequently in the diet of jaguars in Mamirau Reserve. Two toed sloths were also documented in jaguar scats but in only two samples. This may be because of their lower abundance and nocturnal habits, which may make them less vulnerable to jaguar predation. The only other jaguar diet study that describes the consumption of sloths was conducted in an Atlantic Forest site in southeast Brazil by Garla et al. (2001), who reported three toed sloths in 3% of the scats analyzed (n=101), or 2,1% of the prey identified (n=142). The solitary, silent and arboreal behavior of three toed sloths, their camouflaged pelt and small biomass, together with the availabi lity of terrestrial mammals, are probably important factors that explain why this species is not consumed more frequently in other regions. In Mamirau Reserve, however, jaguars may actively hunt sloths. Jaguars in this area are active mostly during the da y (Ramalho 2012 Chapter 2), which is also when three toed sloths are most active (Sunquist & Montgomery 1973). Once detected, sloths are probably easy prey for jaguars because of their limited mobility and lack of defenses. The only other species that c onstituted more than 5% occurrence or biomass were the lesser tamandua and the red howler monkey. The lesser tamandua is a nocturnal species and generally occurs in low densities. Red howler monkeys are consumed at approximately the same frequency as the t amandua, but are reported to occur at higher
164 densities and live in social groups of 4 10 animals (Boubli et al. 2008). Red howler monkeys are much more agile than either the lesser tamandua or the sloth, which would make it more difficult to catch and may explain why this species wa often in scats. The abundance of prey, especially the concentration of spectacled caiman during low water periods, may be responsible for the high densities of jaguars reported in this area (Ramalho 2012 Chapter 2). An adult jaguar has to consume a minimum of 34 g of meat per kg of cat biomass per day to survive (Altman & Dittmer 1973). The average weight of jaguars in Mamirau Reserve is ~50 kg (based on weights of 4 adult males and 5 females, E. Ramalho, unpubl ished data), which means that each jaguar needs to eat roughly 1.7 kg of meat per day or 620.5 kg per year. Based on reported jaguar densities of ~17 jaguars/km in Mamirau during the low water season (Ramalho 2012 Chapter 2), jaguars are estimated to consume ~12,848 kg of prey (Table 4 2). Based on estimates of prey biomass in the region (Table 4 2), this represents only 0.9 % of the available prey population and only 1.3 of standing biomass of prey, which indicates jaguars are not being limited by fo od resources in Mamirau during the low water months of September November. The results of this study give further support to the importance of reptiles, especially caiman, in the diet of the jaguar in seasonally flooded environments. It also raises atten tion to the key role that arboreal mammals, the three toed sloth in particular, may have in the maintenance of jaguar populations in the Vrzea Floodplain Forests of Amazonia. These findings are, in part, encouraging because these prey species are
165 generall y not highly sought or consumed by local people, and, in turn this should facilitate their conservation in the densely human populated Vrzea.
166 Table 4 1. List of jaguar prey identified from scats, average prey body weight, estimated density of prey, abu ndance, biomass of prey species population available (Biomass available=average weight of prey species x abundance), number of times each prey was identified (n), percent of scats which contained that species of prey, percent occurrence (Po), biomass repre sented in scats (Contributed biomass), percent of total biomass represented by contributed biomass ( % Biomass=relative biomass*100) Prey Weight (kg) Density (ind/km) Abundance Biomass available n % scats P o Contributed Biomass (kg) % Biomass Reptiles Spectacled caiman Caiman crocodilus a,f 16.8 4.4 2,494 24,940 32 0.41 0.24 82.2 31.7 Black caiman Melanosuchus niger a,f 16.8 19.2 10,846 108,460 1 0.01 0.01 2.6 1.0 Crocodile tegu Crocodilurus amazonicus c 1 1 0.01 0.01 2.0 0.8 Northern caiman lizard Dracaena guianensis c 2 1 0.01 0.01 2.0 0.8 Unidentified caiman Caiman sp. a 16.8 13 0.17 0.10 33.4 12.9 Unidentified chelonian Chelonia sp. 1 0.01 0.01 Unidentified lizards and snakes Squamata spp. 4 0.05 0.03 Unidentified reptile 4 0.05 0.03 Total 13,340 133,400 57 0.42 122.1 47.2 Mammals Brown throated three toed sloth Bradypus variegatus b 2.98 212 119,992 357,576 41 0.53 0.30 85.5 33.0 Lesser tam andua Tamandua tetradactyla c 4.5 3 1,698 7,641 11 0.14 0.08 23.5 9.1 Red howler monkey Alouatta seniculus b 5.31 38 21,508 114,207 9 0.12 0.07 19.5 7.5 Two toed sloth Choloepus didactylus b 6 88 49,808 298,848 2 0.03 0.01 4.4 1.7 Unidentified squirrel monkey Saimiri sp. c 0.95 72 40,752 38,714 1 0.01 0.01 2.0 0.8 Cattle Bos taurus 30 1 0.01 0.01 3.0 1.2 Unidentified mammals 10 0.13 0.07 Total 233,758 816,987 75 0.55 134.9 52.1 Birds Black bellied Whistl ing Duck Dendrocignia autumnalis e 0.7 1 0.01 0.01 2.0 0.8
167 Table 4 1. Continued Prey Weight (kg) Density (ind/km) Abundance Biomass available n % scats P o Contributed Biomass (kg) % Biomass Unidentified birds 3 0.04 0.02 Total 4 0.03 2.0 0.8 Others Unidentified snail Pomacea sp. 4 0.05 Vegetable matter 2 0.03 Total 6 Grand Total 247,098 950,387 136 259.0 a Mean weight of caiman consumed by jaguars during the study, based on total length of the carcasses found; b Queiroz 1995; c Valsecchi 2005; d R. Cintra (INPA), unpublished data; f Da Silveira 2002
168 Table 4 2. Weight, density, and abundance of the most important prey species in st udy area and the estimated consumption of each species by jaguars during the three months of low water level (September November). Estimates are calculated assuming a jaguar population density of 17 jaguars/100 km (Ramalho 2012 Chapter 2), which would r epresent a jaguar population size of 96 adult jaguars within the study area Prey Weight (kg) Density (ind /km) Population Size Biomass B iomass consumed % of biomass consumed Number of individual prey % of prey population consumed Caiman crocodilus 10 .0 43.0 2,494 24,940 4,667 18.7 278 11.1 Bradypus variegatus 4.2 212.0 119,992 503,966 4,858 1.0 1,157 1.0 Tamandua tetradactyla 4.6 3.0 1,698 7,811 1,340 17.2 291 17.2 Alouatta seniculus 5.5 35.0 19,810 108,955 1,104 1.0 201 1.0 Choloepus didactlyus 6 .0 88.0 49,808 298,848 250 0.1 42 0.1 Melanosuchus niger 10.0 19.2 1,111 11,114 147 1.3 9 0.8 Saimiri sp. 1.0 72.0 40,752 38,714 118 0.3 124 0.3 Total 472.2 235,665 994,349 12,484 1.3 2,101 0.9
169 Fig ure 4 1. Location of all diet studies conducted t o date (green circles), Ecoregions within the jaguar present distribution (other colors), and extent of Amazonia (red line)
170 Figure 4 2. Smaller frame shows location of Mamirau Sustainable Development Reserve within Brazil. In larger frame red line rep resents the limits of the Reserve. Dashed yellow ellipse represents location where samples were collected
171 APPENDIX BIBLIOGRAPHY REVIEW METHODS To assess the current knowledge base on jaguars I used the major biomes of Brazil (Amazonia, Caatinga, Cerrado Atlantic Forest, Pantanal, and Pampas) as the management regions and their populations as the units. To evaluate knowledge within these biomes I used five web based search engines: Thomson Reuters (formerly ISI) Web of knowledge (via University of Florid a www.isiwebofknowledge.com), Peridicos CAPES (www.periodicos.capes.gov.br), Scielo (www.scielo.org), Google Scholar mer ged these results with the compilation of Inskip an d Zimmerman (2009 ; available from: www.jaguarnetwork.org/Jaguar%20Bibliography%20Updated.pdf). I also used the cit ed references of the publications found to search for other publ ications not encountered in the first search.
172 Table A 1. Jaguar peer reviewed publications and book chapters produced in Brazil Research Category Research Sub category Count Freq. w/ category Overall Freq. Amazonia Atlantic forest Caatinga Cerrado Pantanal Ecology and behavior Diet 18 51.4 16.4 3 6 2 3 7 Movement, home range s ize, and habitat use 10 28.6 9.1 3 3 1 1 6 Population parameters and structure 10 28.6 9.1 1 5 2 1 4 Reproduction 2 5.7 1.8 0 0 0 1 2 NI 1 2.9 0.9 0 1 0 0 0 Sub total 35 31.8 5 10 3 4 16 Conservation Proposed measures and threats 15 75.0 13.6 4 7 1 2 4 Population viability analysis 1 5.0 0.9 0 1 0 0 0 Habitat loss and fragmentation 1 5.0 0.9 1 0 0 0 0 Habitat suitability model 1 5.0 0.9 0 2 0 0 0 Management experiment 0 0.0 0.0 0 0 0 0 0 NI 3 15.0 2.7 2 2 0 0 0 Sub total 20 18.2 4 10 1 2 4 Conflict Depredation 12 54.5 10.9 1 4 0 1 4 Hunting 8 36.4 7.3 5 2 0 0 0 Local perception 3 13.6 2.7 1 2 1 1 2 Attack 1 4.5 0.9 0 0 0 0 1 NI 0 0.0 0.0 0 0 0 0 0 Sub total 22 20.0 6 8 1 2 7 Veterinary and pathology 35 31.8 1 2 0 1 0 Status and distribution 11 10.0 1 7 1 0 1
173 Table A 1. Continued Research Category Research Sub category Count Freq. w/ category Overall Freq. Amazonia Atlantic forest Caat inga Cerrado Pantanal Method 11 10.0 0 2 0 0 1 Biology and morphology 9 8.2 0 0 0 1 2 Genetics 8 7.3 0 2 0 1 0 Total 110 12 24 4 6 21 Overall frequency 19.0 38.1 6.3 9.5 33.3
174 Figure A 1. Number of peer reviewed p ublications related to the jaguar per year.
175 Figure A 2. Number of jaguar related peer reviewed publications per country
176 LIST OF REFERENCES Ackerman, B. B., F. G. Lindzey, and T. P. Hemker. 1984. Cougar Food Habits in Southern Utah. Journal of Wildl ife Management, 48: 147 155. Adams, L. G., R. O. Stephenson, B. W. Dale, R. T. Ahgook, and D. J. Demma. 2008. Population dynamics and harvest characteristics of wolves in the Central Brooks Range. Alaska Wildlife Monographs 170: 1 25. Alho, C. J. R. 2008. Biodiversity of the Pantanal: response to seasonal flooding regime and to environmental degradation. Brazilian Journal of Biology 68: 957 966. Alho, C. J. R., and T. E. Lacher Jr. 1991. Mammalian conservation in the Pantanal of Brazil. Pages 280 294 in M. A. Mares and D. J. Schmidly, editors. Latin American mammalogy: history, biodiversity and conservation. University of Oklahoma Press, Norman. Almeida, A. E. 1976. Jaguar hunting in the Mato Grosso. 1st edition. Stanwill Press, England. Altman P. L., and D. S. Dittmer. 1973. Biology data book, vol 2. Federation of American Societies for Experimental Biology, USA. c prey choice by Piro hunters of Amazonian Peru. Human Ecology 21: 355 387. Andren, H., et al. 2006. Survival rates and causes of mortality in Eurasian lynx ( Lynx lynx ) in multi use landscapes. Biological Conservation 131: 23 32. Aranda, M. 1993. Hbitos alimentarios del jaguar (Panthera onca) en la Reserva de la Biosfera de Calakmul, Campeche. Pages 231 238 in R. A. Medelln, and G. Ceballos, editors. Avances en el Estudio de los Mamferos de Mxico. Asociacin Mexicana de Mastozoologia, A. C., Mxico, D. F. Aranda, M. 1994. Importancia de los pecaries (Tayassu spp.) en la alimentacin del jaguar (Panthera onca). Acta Zool. Mex. (n.s.) 62:11 22. Aranda, M., and V. Snchez Cordero. 1996. Prey spectra of jaguar (Panthera onca) and Puma (Puma concolor) in tro pical forests of Mexico. Studies of Neotropical Fauna & Environment 31: 65 67. Ayres, J. M. 1993. As Matas de Vrzea do Mamirau. MCT CNPq PTU, Sociedade Civil Mamirau, Braslia. Ayres, J. M. 1993. As Matas de Vrzea do Mamirau. MCT CNPq PTU, Sociedade C ivil Mamirau, Braslia.
177 Ayres, J. M., G. A. B., Fonseca, A. B. Rylands, H. L. Queiroz, L. P. De S. Pinto, D. E. Masterson, R. Cavalcanti. 1997. Abordagens Inovadoras para Conservao da Biodiversidade do Brasil: Os Corredores Ecolgicos das Florestas Neot ropicais do Brasil Verso 3.0 Programa Piloto para a Proteo das Florestas Neotropicais, Projeto Parques e Reservas. Ministrio do Meio Ambiente, Recursos Hdricos e da Amaznia Legal (MMA), Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renovveis (IBAMA), Braslia. Azevedo, F. C. C., and D. L. Murray. 2007. Spatial organization and food habitats of jaguar ( Panthera onca ) in a floodplain forest. Biological Conservation 137: 392 402. Balaguera Reina, S. A., and J. F. Gonzalez Maya. 2008. Occasional jaguar hunting for subsistence in Colombian Choc. Cat News 48: 5. Bayley, P. B., and M. Petrere. 1989. Amazon fisheries: assessment methods, current status and management options. Pages 385 398 in D. P. Dodge, editor. Proceedings of the intern ational large river symposium. Canadian Special Publication of Fisheries and Aquatic Sciences. Beier, P., and R. F. Noss. 1998. Do habitat corridors provide connectivity? Conservation Biology 12: 1241 1252. Bennett, A. F. 2003. Linkages in the landscape: t he role of corridors and connectivity in wildlife conservation. IUCN, Gland, Switzerland, and Cambridge, United Kingdom. Berreta, E. 2001. Ecophysiology and management response of the subtropical grasslands of Southern America. Pages 939 946 in Gomide, J. A., W. R. S. Mattos, and S. C. Silva, editors. Proceedings of the XIX International Grassland Congress, So Pedro, Piracicaba, Brazil. Bishop, R. C. 1978. Endangered species and uncertainty: the economics of a safe minimum standard. Journal of Agricultural Economics 60: 10 18. Borchers, D. L., and M. G. Efford 2008. Spatially explicit maximum likelihood methods for capture recapture studies. Biometrics 64:377 385 Boubli, J. P., Di Fiore, A., Rylands, A.B. & Mittermeier, R.A. 2008. Alouatta seniculus. In: IUC N 2011. IUCN Red List of Threatened Species. Version 2011.2.
178 Buckland, S. T., K. P. Burnham, and N. H. Augustin. 1997. Model selection: an integral part of inference. Biometrics 53:603 618. Burnha m, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information theoretic approach. Springer Verlag, New York, New York, USA. Campbell, M., and K. J. Mackay. 2009. Communicating the role of hunting for wildlife managem ent. Human Dimensions of Wildlife 14: 21 36. Carbone, C., and J. L. Gittleman. 2002. A common rule for the scaling of carnivore density. Science 295:2273 2276 Cardillo, M., A. Purvis, W. Sechrest, J. L. Gittleman, J. Bielby, and G. M. Mace. 2004 Human popu Biology 2: 909 914. DOI:10.1371/journal.pbio.0020197. Carvalho, E. A. R., and J. C. B. Pezzuti. 2010. Hunting of jaguars and pumas in the Tapajs Arapiuns Extractive Reserve, Brazilian Amaz onia. Oryx 44: 610 612. Carvalho, F. M. V., P. de Marco, and L. G. Ferreira. 2009. The Cerrado into pieces: Habitat fragmentation as a function of landscape use in the savannas of central Brazil. Biological Conservation 142: 1392 1403. Caso, A., C. Lopez Gonzalez, E. Payan, E. Eizirik, T. de Oliveira, R. Leite Pitman, M. Kelly, and C. Valderrama. 2008. Panthera onca. In IUCN 2011. IUCN Red List of Threatened Species. Version 2011.1. Available from http:www.iucnredlist.org. Downloaded on 22 August 2011. Cas telletti, C. H. M., J. M. Silva, M. Tabarelli, A. M. M. Santos. 2004. Quanto ainda resta da Caatinga? Uma estimativa preliminar. Pages 91 100 in Silva, J. M., M. Tabarelli, M. Fonseca and L. Lins, editors. Biodiversidade da Caatinga: reas e aes priorit rias para a conservao. Ministrio do Meio Ambiente, Braslia. Castro Vsquez, L., M. Meza, T. Plese, and S. Moreno Mora. 2010. Activity patterns, preference, and use of floristic resources by Bradypus variegatus in a tropical dry forest fragment, Santa C atalina, Bolvar, Colombia. Edentata 11:62 69. Chetkiewicz, C. B., and G. Raygorodetsky. 1999. Jaguar reconnaissance survey in Mamirau and Aman sustainable development reserves. Unpublished report to Sociedade Civil Mamirau, Tef, AM, Brazil. Chinchilla F. A. 1997. La dieta del jaguar (Panthera onca), el puma (Felis concolor) y el manigordo (Felis pardalis) (Carnvora: Felidae) en el Parque Nacional Corcovado, Costa Rica. Ver. Biol. Trop. 45(3): 1223 1229.
179 CITES (Convention on International Trade in End angered Species of Wild Fauna and Flora). 1973. Signed at Washington, D.C., on 3 March 1973. Amended at Bonn, on 22 June 1979. Clark, T. W., and R. D. Brunner. 1996. Making partnerships work in endangered species conservation. Endangered Species Update 13: 1 4. Clark, T.W., A. P. Curlee, and R. P. Reading. 1996. Crafting effective solutions to the large carnivore conservation problem. Conservation Biology 10: 1036 1045. Clark, T.W., A. R. Willard, C. M. Cromley, editors. 2000. Foundations of Natural Resourc es Policy and Management. Yale University Press, New Haven and London. Clark, T.W., and R. L. Wallace. 1998. Understanding the human factor in endangered species recovery. Endangered Species Update 15: 2 9. Clark, T.W., M. Stevenson, K. Ziegelmayer, and M. Rutherford, editors. 2001. Species and Ecosystem Conservation: An Interdisciplinary Approach. Yale School of Forestry and Environmental Studies, Bulletin Series 105. Coimbra Filho, A. F., and I. de G. Cmara. 1996. Os limites originais do bioma Mata Atln tica na regio Nordeste do Brasil. Fundao Brasileira para Conservao da Natureza, Rio de Janeiro (in Portuguese). Conforti V. A., and F. C. C. de Azevedo. 2003. Local perceptions of jaguars ( Panthera onca ) and pumas ( Puma concolor ) in the Iguau Natio nal Park area, south Brazil. Biological Conservation 111: 215 221. Cormack, R. M. 1964. Models for capture recapture. Biometrika 51:429 438. Crawshaw Jr., P. G., H. B. Quigley. 1991. Jaguar spacing, activity and habitat use in a 370. Crawshaw, P. G. 1995. Comparative ecology of ocelot (Felis pardalis) and Jaguar (Panthera onca) in a protected subtropi cal forest in Brazil and Argentina. Ph.D. Thesis University of Florida :1 189. Crawshaw, P. G. 1995. Compartive ecology of ocelot ( Felis pardalis ) and jaguar ( Panthera onca ) in a protected subtropical forest in Brazil and Argentina. Ph.D. Thesis, Univers ity of Florida, USA, 205 pp. Crawshaw, P. G. 2003. A personal view on the depredation of domestic animals by large cats in Brazil. Natureza e Conservao 1(1): 13 16. Crawshaw, P. G. 2004. Depredation of Domestic Animals by Large Cats in Brazil. Human Dime nsions of Wildlife 9: 329 330.
180 Crawshaw, P. G., and Quigley H. B. 2002. Jaguar and puma feeding habits in the Pantanal, Brazil, with implications for their management and conservation. Pages 223 235 in Medelln, R. A., et al., editors. El Jaguar en el nuev o milenio: una evaluacin de su estado, deteccin de prioridades y recomendaciones para la conservacin de los jaguares en Amrica. Universidad Nacional Autnoma de Mxico & Wildlife Conservation Society, Mxico city. Crawshaw, P.G., J. K. Mhler, C. Indru siak, S. M. C. Cavalcanti, M. Leite Pitman, and K. M. Silvius. 2003. Ecology and Conservation of the jaguar (Panthera onca) in Iguau National Park, Brasil. In: Silvius, MK; Bodmer, RE & Fragoso, JMV. 2004. People in Nature. Columbia University Press. Nova York. Cullar, E., L. Maffei, R. Arispe, and A. J. Noss. 2006. limit of their range: activity patterns and density estimates from camera trapping in Bolivian dry forests. Studies on Neotropical Fauna and Environment 41(3):1 69 178. Cullen Jr., L. 2006. Jaguars as landscape detectives for the conservation of Atlantic Forests in Brazil. PhD thesis, University of Kent, Canterbury, UK. Cullen Jr., L., K. C. Abreu, D. A. de Sana, A. F. D. Nava. 2005. Jaguars as landscape detective s for the upper Paran River Corridor, Brazil. Natureza & Conservao 3: 147 161. Da Silveira, R. 2002. Conservao e manejo do jacar au (Melanosuchus niger) na Amaznia Brasileira. In: Larriera, A.; Verdade, L. M. (Eds.). La Conservasin y el Manejo de Caimanes e Cocodrilos de Amrica Latina. Volume II. Piracicaba: C. N. Editoria. p. 61 78. Da Silveira, R., E. E. Ramalho, J. B. Thorbjarnarson, W. E. Magnusson. 2010. Depredation by jaguars on caimans and importance of reptiles in the diet of jaguar. Jour nal of Herpetology 44:418 424. Da Silveira. 2002. Conservao e manejo do jacar au (Melanosuchus niger) na Amaznia Brasileira. Pages 61 78 in A. Larriera, and L. M. Verdade, editors. La Conservasin y el Manejo de Caimanes e Cocodrilos de Amrica Latin a. C. N. Editoria, Vol II., Piracicaba. Dalponte, J. C. 2002. Dieta del jaguar y depredacion de ganado en el norte de Pantanal, Brasil. Pages 223 235 in R. A. Medellin, C. Chetkiewicz, A. Rabinowitz, K. H. Redford, J. G. Robinson, E. Sanderson, and A. Tabe r, editors. Jaguars in the new millennium. A status assessment, priority detection, and recommendations for the conservation of jaguars in the Americas. UNAM/WCS.Mexico D. F.
181 Dalponte, J. C. 2002. Dieta del jaguar y depredacin de ganado en el norte del Pa ntanal, Brasil. Pages 209 221 in Medellin, R., et al., editors. Medelln, R. A., et al., editors. El Jaguar en el nuevo milenio: una evaluacin de su estado, deteccin de prioridades y recomendaciones para la conservacin de los jaguares en Amrica. Univer sidad Nacional Autnoma de Mxico & Wildlife Conservation Society, Mxico city. Dean, W. 1997. With Broadax and Firebrand: The Destruction of the Brazilian Atlantic Forest. University of California Press, Berkeley. Di Bitetti, M. S., A. Paviolo, and C. D. de Angelo. 2006. Density, habitat use and activity patterns of ocelots (Leopardus pardalis) in the Atlantic Forest of Misiones, Argentina. Journal of Zoology 270:153 163. Di Bitetti, M. S., A. Paviolo, C. D. de Angelo, and Y. E. Di Blanco. 2008. Local and continental correlates of the abundance of a Neotropical cat, the ocelot (Leopardus pardalis). Journal of Tropical Ecology 24:189 200. Dillon, A. and M. J. Kelly. 2007. Ocelot Leopardus pardalis in Belize: the impact of trap spacing and distance moved on d ensity estimates. Oryx 41:469 477. Dillon, A. and M. J. Kelly. 2008. Ocelot home range, overlap, and density using camera trapping and radio telemetry. Journal of Zoology 275(4):391 398. Dirzo, R., and P. H. Raven. 2003. Global state of biodiversity and lo ss. Annual Review of Environment and Resources 28: 137 167. Doughty, R. W., and N. Myers. 1971. Notes on Amazon wildlife trade. Biological Conservation 3: 293 297. Efford, M. 2004. Density estimation in live trapping studies. Oikos 106:598 610. Eiten, G. 1 977. Delimitao do conceito de Cerrado. Arquivos do Jardim Botnico, Rio de Janeiro 21: 125 134. Eizirik, E., J., Kim, M. Menotti Johnson. 2001. Phylogeography, population history and conservation gene tics of jaguars ( Panthera onca Mammalia, Felidae). Molecular Ecology 10: 65 79. Emmons, L. H. 1987. Comparative feeding ecology of felids in a Neotropical rainforest. Behavioral Ecology and Sociobiology 20: 271 283. Emmons, LH. 1987. Comparative feeding e cology of felids in a Neotropical rainforest. Behav. Ecol. Sociobiol., 20: 271 283. Esterci, N., and E. E. Ramalho. 2007. Mamirau Protegendo um sistema rico e ameaado. Cincia Hoje 40:50 54.
182 Facure, K. G., and A. Giaretta. 1996. Food habits of carnivor es in a coastal Atlantic forest of southeastern Brazil. Mammalia 60(3):499 502. Fahrig, L., and G. Merriam. 1994. Conservation of fragmented populations. Conservation Biology 8: 50 59 FAO (Food and Agriculture Organization of the United Nations). 2011. Sta te of the Farrell, LF; Roman, J; Sunquist, ME. 2000. Dietary separation of sympatric carnivores identified by molecular analysis of scats. Mol. Ecol. 9: 1583 1590. Fausto, C. 2004. A Blend of blood and tobacco: Shamans an d jaguars among the Parakan of Eastern Amazonia. Pages 157 178 in N. L. Whitehead, and R. Wright, editors. In Darkness and Secrecy: The Anthropology of Assault Sorcery and Witchcraft in Amazonia. Duke University Press, Durham, NC. Fearnside, P. M. 2005. D eforestation in Brazilian Amazonia: History, Rates, and Consequences. Conservation Biology 19: 680 688. Ferraz, K., R. C. de Paula, B. Beisiegel, E. Moraes Jr., S. M. C. Cavalcanti, and T. G. de Oliveira. 2011. Adequabilidade ambiental dos biomas brasileir os ocorrncia da ona pintada ( Panthera onca ). Pages 269 310 in R. C. Paula, A. Desbiez, S. M. C. Cavalcanti, editors. Plano de Ao para a Conservao da ona pintada no Brasil Anlise da viabilidade populacional e adequabilidade ambiental. Srie Esp cies Ameaadas. ICMBio, Braslia. Rio Grande do Sul, Porto Alegre. Frank, L. D., D. Simpson and R. Woodroffe. 2003. Foot snares: an effective method for capturing African lions. Wildlife Society Bulletin 31:309 314. Fundao SOS Mata Atlntica e o INPE (Instituto Nacional de Pesquisas Espaciais). 2011. Atlas dos remanescentes florestais da Mata Atlntica perodo 2008 2010. Fundao SOS Mata Atlntica, So Paulo, SP. Garla, R. C., E. Z. Setz, and N. Gobbi. 2001. Jaguar (Panthera onca) food habits in Atlantic rain forest of southeastern Brazil. Biotropica 33(4):691 696. Gerber, B. D., S. M. Karpan ty, and M. J. Kelly. 2011. Evaluating the potential biases in carnivore capture recapture studies associated with the use of lure and varying density estimation techniques using photographic sampling data of the Malagasy civet. Population Ecology 54:43 54. Goulding, M., N. J. H. Smith, D. J. Mahar. 1996. Floods of fortune: ecology and economy along the Amazon. Columbia University Press, New York, USA.
183 Guggisberg, C. A. W. 1975. Wild cats of the world. Taplinger, New York. Guix, J. C. 1997. Cat communities i n six areas of the state of So Paulo, southeastern Brazil, with observation on their feeding habits. Grupo Estudos Ecolgicos, Srie Documentos 5: 16 38. Haag, T., A. S. Santos, D. A. de Sana, R. G. Morato, L. Cullen Jr., P. G. Crawshaw Jr., C. de Angelo, M. S. di Bitetti, F. M. Salzano, E. Eizirik. 2010. The effect of habitat fragmentation on the genetic structure of a top predator: loss of diversity and high differentiation among remnant populations of Atlantic Forest jaguars ( Panthera onca ). Molecular E cology 19: 4906 4921. Harmsen, B. J., R. J. Foster, S. C. Silver, L. E. T. Ostro, and C. P. Doncaster. 2011. Jaguar and puma activity patterns in relation to their main prey. Mammalian Biology 76:320 324. Harmsen, B. J., R. J. Foster, S. Silver, L. Ostro, and C.P. Doncaster. 2009. Spatial and temporal interactions of sympatric jaguars (Panthera onca) and pumas (Puma concolor) in a Neotropical forest. Journal of Mammalogy 90:612 620. Harris, M. B., W. Tomas, G. Mouro, C. J. Silva, E. Guimares, F. Sonoda, E Fachim. 2005. Safeguarding the Pantanal wetlands: threats and conservation initiatives. Conservation Biology 19: 714 720. Hoogesteijn, R, and C. A. Chapman. 1997. Large ranches as conservation tools in the Venezuelan Llanos. Oryx 31: 274 284. Hoogesteijn R, E.O. Boede, and E.Mondolfi. 2002.Observaciones de la depredacon de bovinos por jaguares en Venezuela y los programas gubernamentales de control. Pages 183 197 in Medelln, R. A., et al., editors. El Jaguar en el nuevo milenio: una evaluacin de su es tado, deteccin de prioridades y recomendaciones para la conservacin de los jaguares en Amrica. Universidad Nacional Autnoma de Mxico & Wildlife Conservation Society, Mxico city. Hoogesteijn, R., and E. Mondol 1992. The jaguar. Armiantos Editores, C aracas. Inskip, C., and A. Zimmermann. 2009. Review Human felid conflict: a review of patterns and priorities worldwide. Oryx 43: 18 34. IUCN (International Union for Conservation of Nature and Natural Resources). 2011. The IUCN Red List of Threatened Spec ies. Version 2011.2. Available from http:www.iucnredlist.org. Downloaded on 10 November 2011. Jackson, R. M., J. D. Roe, R. Wangchuk, and D. O. Hunter. 2006. Estimating snow leopard population abundance using photography and capture recapture techniques. W ildlife Society Bulletin 34(3):772 781.
184 Jolly, G. M. 1965. Explicit estimates from mark recapture data with both death and immigration stochastic models. Biometrika 52:225 247. Jorgenson, J., and K. H. Redford. 1993. Humans and big cats as predators in the Neotropics. Pages 367 390 in Dunstone, N. and M. L. Gorman, editors. Mammals as predators. Claredon Press, Oxford, United Kingdom. Junk, W. J. 1993. Wetlands of tropical South Amrica. Pages 679 739 in Whigham, D., Dykyjov, D., and Hejny, S., editors. We tlands of the World: Inventory, ecology and management. Vol.1. Kluwer Academic Publishers, Netherlands. Junk, W. J., and C. da Silva. 1999. O conceito do pulso de inundao e suas implicaes para o Pantanal de Mato Grosso. Pages 17 28 in II Simpsio sobre recursos naturais e scio econmicos do Pantanal: Manejo e Conservao. Empresa Brasileira de Pesquisa Agropecuria (Embrapa), Corumb, Brasil (in Portuguese). Junk, W. J., editor. 1997. The Central Amazon Floodplain: Ecology of a Pulsing System. Ecologic al studies Vol. 126, Springer, Berlin. Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood pulse concept in river floodplain systems. Pages 110 127 in D. P. Dodge, editor. Proceedings of the International Large River Symposium (LARS). Canadian Spe cial Publication of Fisheries and Aquatic Sciences, Ottawa, Canada. Karanth, K. U, J. D. Nichols, N. S. Kumar, and J. E. Hines. 2006. Assessing tiger population dynamics using photographic capture recapture sampling. Ecology 87:2925 2937. Karanth, K. U, J. D. Nichols, N. S. Kumar, W. A. Link, and J. E. Hines. 2004. Tigers and their prey: predicting carnivore densities from prey abundance. Proceedings of The National Academy of Sciences (USA) 101:4854 4858. Karanth, K. U. 1995. Estimating tiger (Panthera tig ris) density from camera trap data using capture recapture models. Biological Conservation 71:333 338. Karanth, K. U., and B. M. Stith. 1999. Prey depletion as a critical determinant of tiger population viability. Pages 100 113 in J. Seidensticker, S. Chri stie, and P. Jackson, editors. Riding the tiger: tiger conservation in human dominated landscapes. Cambridge University Press, Cambridge, UK. Karanth, K. U., and J. D. Nichols. 1998. Estimation of tiger densities in India using photographic captures and re captures. Ecology 79:2852 2862. Karanth, K. U., and J. D. Nichols. 2000. Ecological status and conservation of tigers in India. WCS, US Fish and Wildlife Service. Centre for Wildlife Studies, Bangalore, India.
185 Karanth, K. U., and J. D. Nichols. 2002. Monit oring tigers and their prey: a manual for researchers, managers and conservationists in tropical Asia. Centre for Wildlife Studies, Bangalore, India. Karanth, K. U., and R. Chellam. 2009. Carnivore conservation at the crossroads. Oryx 43: 1 2. Karanth, K. U ., J. D. Nichols, N. S. Kumar, W. A. Link, and J. E. Hines. 2004. Tigers and their prey: predicting carnivore densities from prey abundance. Proceedings of the National Academy of Sciences (USA) 101:4854 4858. Kellert, S. R. 1985f. Social and perceptual fa ctors in endangered species management. Journal of Wildlife Management 49: 528 536. Kellert, S. R. 1991. Public views of wolf restoration in Michigan. Transactions of the North American Wildlife & Natural Resources Conference 56: 152 161. Kellert, S. R. 1 996. The value of life: biological diversity and human society. Island Press, Washington, D. C. Kellert, S. R., M. Black, C. R. Rush, and A. J. Bath. 1996. Human Culture and Large Carnivore Conservation in North America. Conservation Biology 10: 977 990. Kelly, M. J., A. J. Noss, M. S. Di Bitetti, L. Maffei, R. Arispe, A. Paviolo, C. D. de Angelo, and Y. E. Di Blanco. 2008. Estimating puma densities from camera trapping across three study sites: Bolivia, Argentina, Belize. Journal of Mammalogy 89(2):408 41 8. Kendall, W. L. 1999. Robustness of closed capture recapture methods to violations of the closure assumption. Ecology 80:2517 2525. Kendall,W. L., J. D. Nichols, and J. E. Hines. 1997. Estimating temporary emigration and breeding proportions using captur e design. Ecology 78:563 578. Klink, C. A., A. G. Moreira. 2002. Past and current human occupation, and land use. Pages 69 88 in Oliveira, P.S. and R. J. Marquis, editors. The Cerrados of Brazil: Ecology and Natural His tory of a Neotropical Savanna. Columbia University Press, New York. Klink, C. A., and R. B. Machado. 2005. Conservation of the Brazilian Cerrado. Conservation Biology 19: 707 713. Knight, J. 2003. Waiting for Wolves in Japan. Oxford University Press, Oxfo rd. Lasswell, H. D. 1971. A pre view to policy sciences. American Elesrvier, New York.
186 Lasswell, H. D., and M. McDougal. 1992. Jurisprudence for a free society: Studies in law, science, and policy. New Haven Press, New Haven, Connecticut. Leal, I. R., A. V icente, and M. Tabarelli. 2003. Herbivoria por caprinos na caatinga: uma estimativa preliminar. Pages 695 715 in I. R. Leal, M. Tabarelli, and J. M. C. Silva, editors. Ecologia e conservao da Caatinga. Editora Universitria, Universidade Federal de Perna mbuco, Recife, Brasil (in Portuguese). Leite, M. R. P. 2000. Relaes entre a ona pintada, ona parda e moradores locais em trs unidades de conservao da floresta atlntica do estado do Paran, Brasil. Dissertao de mestrado, Universidade Federal do Pa ran. Leite, M. R. P., and F. Galvo. 2002. El jaguar, el puma y el hombre en tres reas protegidas del bosque atlntico costero de Paran, Brasil. Pages 237 250 in Medelln, R. A., et al., editors. El jaguar en el nuevo milenio. Fondo de Cultura Econmica Universidad Nacional Autnoma de Mxico, Mxico, D.F. and Wildlife Conservation Society, New York (in Spanish). Leite, M. R. P., R. L. P. Boulhosa, F. Galvo, and L. Cullen Jr. 2002. Conservacon del jaguar en las reas protegidas del bosque Atlntico de la costa de Brasil. Pages 25 42 in Medelln, R. A., et al., editors. El Jaguar en el nuevo milenio: una evaluacin de su estado, deteccin de prioridades y recomendaciones para la conservacin de los jaguares en Amrica. Universidad Nacional Autnoma de M xico & Wildlife Conservation Society, Mxico city. Leopold, A. 1949. A Sand County Almanac. Oxford University Press, New York Lopez, B. H. 1978. Of wolves and men. Charles Scribner's Sons, New York. Ludwig, D., R. Hilborn, and C. Walters. 1993. Uncertaint y, resource exploitation, and conservation: lessons from history. Science 260: 36. Luna, L. E., and P. Amaringo. 1999. Ayahuasca visions: The religious Iconography of a Peruvian Shaman. North Atlantic Books, USA. Machado, A. B. M, G. M. Drummond, A. P. Pagl ia, editors. 2008. Livro vermelho da fauna brasileira ameaada de extino. 1st edition. Grfica e Editora Roma Ltda., Belo Horizonte, Brazil. Maffei, L., A. J. Noss, E. Cullar, and D. I. Rumiz. 2005. Ocelot (Felis pardalis) population densities, activity and ranging behavior in the dry forests of eastern Bolivia: data from camera trapping. Journal of Tropical Ecology 21(3):349 353 Maffei, L., A. J. Noss, S. C. Silver, and M. J. Kelly. 2011. Abundance/Density case study: jaguars in the Americas. Pages 11 9 and K. U. Karanth, editors. 2011. Camera traps in animal ecology: Methods and analyses. First edition. Springer, New York, USA.
187 Maffei, L., E. Cullar, and A. Noss. 2004. One thousand jaguars (Panthera onca) in Boli Iya National Park. Journal of Zoology 262:295 304. Main, M. B. 2008. Reconciling competing ecological explanations for sexual segregation in ungulates. Ecology 89:693 704 Main, M.B., V.C. Bleich, and F.W. Weckerly. 1 996. Sexual segregation in ungulates: new directions for research. Journal of Mammalogy 77:449 461. Mazzolli, M., and M. L. A. Hammer. 2008. Qualidade de ambiente para a ona pintada, puma e jaguatirica na Baa de Guaratuba, Estado do Paran, utilizando os aplicativos Capture e Presence. Biotemas 21: 105 117. Medeiros, L. P., R. N. Giro, E. S. Giro, and J. A. Leal. 2000. Caprinos. Empresa Brasileira de Pesquisa Agropecuria (EMBRAPA), Centro de Pesquisa Agropecuria do Meio Norte, Teresina, Brasil (in Por tuguese). Medelln, R. A., et al., editors. 2002. El Jaguar en el nuevo milenio: una evaluacin de su estado, deteccin de prioridades y recomendaciones para la conservacin de los jaguares en Amrica. Universidad Nacional Autnoma de Mxico & Wildlife Con servation Society, Mxico city. Miranda, E. E., C. A. Carvalho, C. A. Spadoto, M. C. Hott, O. T. Oshiro, and W. A. Holler. 2008. Alcance Territorial da Legislao Ambiental e Indigenista. Embrapa, Campinas. http:www.alcance.cnpm.embrapa.br (acessed 18 June 2011). Morn, E. F. 1990. Ecologia Humana das populaes da Amaznia. Editora Vozes, Rio de Janeiro. MSDI (Mamirau Sustainable Development Institute). 2011. Censo Demogrfico das Reservas de Desenvolvimento Sustentvel Mamirau e Aman 2011. MSDI, Tef Amazonas, Brazil. Nepstad, D. C., P. Lefebre, U. L. da Silva, J. Tomasella,, P. Schlesinger, L. Solorzano, P. Moutinho, D. Ray, and J. G. Benito. 2004. Amazon drought and its implications for forest flammability and tree growth: a basin wide analysis. Gl obal Change Biology 10: 704 717. Noss, A. J., B. Gardner, L. Maffei, E. Cullar, R. Montao, A. Romero Muoz, R. Sollman, A.F O'Connell. 2012. Comparison of density estimation methods for mammal populations with camera traps in the Kaa Iya del Gran Chaco l andscape. Animal Conservation. doi: 10.1111/j.1469 1795.2012.00545.x Novack, A. J., Main, M. B., Sunquist, M. E. and Labisky, R. F. 2005. Foraging ecology of jaguar ( Panthera onca ) and puma ( Puma concolor ) in hunted and non hunted sites within the Maya Bio sphere Reserve, Guatemala. Journal of Zoology 267:167 178.
188 Novack, J. N., M. B. Main, M. E. Sunquist, and R. F. Labisky. Foraging ecology of jaguar ( Panthera onca ) and puma ( Puma concolor ) in hunted and non hunted sites within the Maya Biosphere Reserve, G uatemala. Journal of Zoology 267: 167 178. Nuez, R. 2006. Patrones de actividad, movimiento y mbito hogareo del jaguar y del puma en la reserva de la biosfera Chamela Cuixmala, Jalisco. MSc thesis, Universidad Nacional Autonoma de Mexico, Mexico. Nuez, R. 2012. Estimating jaguar population density using camera traps: a comparison with radio telemetry estimates. Journal of Zoology 285:39 45. Nunez, R., B. Miller, and F. Lindzey, F. 2000. Food habits of jaguars and pumas in Jalisco, Mexico. Journal of Zoo logy, 252:373 379. Nuez, R., B. Miller, and F. Lindzey. 2002. Ecologa del jaguar en la reserva de la Biosfera Chamela Cuixmala, Jalisco, Mxico. Pages 107 126 in R. A. Medelln et al., editors. El Jaguar en el NuevoMileno.Una evaluacionde suestado, detec cion de prioridades y recomendaciones para la conservacin de los jaguars en America. Universidad Nacional Autonoma de Mexico and Wildlife Conservation Society, Distrito Federal Mexico, Mexico City. dance: a conceptual framework. In pages 71 editors. Camera traps in animal ecology: methods and analyses. Springer Obbard, M. E., and E. J. Howe. 2008. Demography of black bears in hunted and un h unted areas of the boreal forest of Ontario. Journal of Wildlife Management 72: 869 880. Oliveira, M. D., and D. F. Calheiros. 2000. Flood pulse influence on phytoplankton communities of the south Pantanal floodplain, Brazil. Hydrobiologia 427: 101 112. Oli veira, T. 2002. Ecologia comparativa de la alimentacon del jaguar y del puma en el Neotrpico. Pages 265 288 in R. A. Medellin, C. Chetkiewicz, A. Rabinowitz, K. H. Redford, J. G. Robinson, E. Sanderson, and A. Taber, editors. Jaguars in the new millenniu m. A status assessment, priority detection, and recommendations for the conservation of jaguars in the Americas. UNAM/WCS.Mexico D. F. Olmos, F. 1993. Notes on the food habits of Brazilian 'Caatinga' carnivores. Mammalia 57(1): 126 130. Olson, D. M. et al. 2001. Terrestrial Ecoregions of the world: A new map of life on earth. Bioscience 51:933 938.
189 Overbeck, G. E., S. C. Muller, A. Fidelis, J. Pfadenhauer, V. D. Pillar, C. C. Blanco, I. I. e South Brazilian Campos. Perspectives in Plant Ecology, Evolution and Systematics 9: 101 116. Paula, R. C., A. Desbiez, S. M. C. Cavalcanti, editors. 2011. Plano de Ao para a Conservao da ona pintada no Brasil Anlise da viabilidade populacional e adequabilidade ambiental. Srie Espcies Ameaadas. ICMBio. Braslia. Paula, R. C., M. F. Campos Neto, and R. G. Morato. 2008. First Official Record of Human Killed by Jaguar in Brazil. Cat New 49: 31 32. Peres, C. A. 2005. Why we need mega reserves in Ama zonia. Conservation Biology 19: 728 733. Peterson, M. N. 2004. An approach for demonstrating the social legitimacy of hunting. Wildlife Society Bulletin 32: 310 321. Pierce, B. M., V. C. Bleich, R. T. Bowyer. 2000. Social organization of mountain lions: Do es a Land Tenure system regulate population size? Ecology 91:1533 1543. Polisar, J. I.; Maxit, D.; Scognamillo, L.; Farrell, M. E.; Sunquist, J. F. e Eisenberg. 2003. Jaguars, pumas, their prey base, and cattle ranching: ecological interpretations of a man agement problem. Biological Conservation, 109: 297 310. Pollock, K. H. 1982. A capture recapture design robust to unequal probability of capture. Journal of Wildlife Management 46:757 760. Pollock, K. H., J. D. Nichols, C. Brownie, and J. E. Hines. 1990. S tatistical inference for capture recapture experiments. Wildlife Monographs 107:1 97. Pradel, R., J. E. Hines, J. D. Lebreton, and J. D. Nichols. 1997. Capture recapture Biometrics 53:60 72. Queiroz, H. L. 1995. Preguias e Guaribas: Os Mamferos Arborcolas do Mamirau. Sociedade Civil Mamirau e CNPq (DUP). Srie Estudos de Mamirau, volume 2, Rio de Janeiro. Queiroz, H. L. 2005. A Reserva de Desenvolvimento Sustentvel Mamirau. Revista de Estudos avana dos IEA USP. Dossi Amaznia II 19(54):183 203. Queiroz, H. L., and N. Peralta. 2011. Protected areas in the Amazonian Vrzea and their role in its conservation. Pages 465 484 in W.J. Junk, M. T. F. Piedade, F. Wittmann, J. Schngart, and P. Parolin, edito rs. Amazonian Floodplain Forests: Ecophysiology, Biodiversity and Sustainable Management Ecological Studies 210. 1 ed. Springer, Dordrecht.
190 Queiroz, HL. 1995. Preguias e Guaribas Os Mamferos Arborcolas do Mamirau. Srie Estudos de Mamirau, volume 2. Sociedade Civil Mamirau e CNPq (DUP). Rio de Janeiro, 176p. Quigley, H. B., and P. G. Crawshaw Jr. 1992. A conservation plan for the jaguar Panthera onca in the Pantanal region of Brazil. Conservation Biology 61: 149 157. Rabinowitz, A. 1986. Jaguar pr edation on domestic livestock in Belize. Wildlife Society Bulletin 14: 170 174. Rabinowitz, A., and K. A. Zeller. 2010. A range wide model of landscape connectivity and conservation for the jaguar Panthera onca Biological Conservation 143: 939 945. Rabino witz, AR & Nottingham, BG. 1986. Ecology and behaviour of the jaguar ( Panthera onca ) in Belize, Central America. Journal of Zoology. London. 210: 149 159. Ramalho, E. E. 2006. Uso do habitat e dieta da ona pintada ( Panthera onca ) em uma rea de Vrzea, Re serva de Desenvolvimento Sustentvel Mamirau, Amaznia Central, Brasil. Page 50. Programa de Ps Graduao em Biologia Tropical e Recursos Naturais, do convnio INPA/UFAM. Instituto Nacional de Pesquisas do Amaznas (INPA), Manaus. Ramalho, E. E., and W. E. Magnusson. 2008. Uso do habitat por ona pintada ( Panthera onca ) no entorno de lagos de vrzea, Reserva de Desenvolvimento Sustentvel Mamirau, AM, Brasil. Uakari 4(2): 33 39. Ramalho, E. E., J. Macedo, T. M. Vieira, J. Valsecchi, J. Calvimontes, M. Ma rmontel, H. L. Queiroz. 2009. Ciclo hidrolgico nos ambientes de Vrzea da Reserva de Desenvolvimento Sustentvel Mamirau Mdio Rio Solimes, Perodo de 1990 a 2008. Uakari 5(1) :61 87. Rasker, R., and A. Hackman. 1996. Economic development and the conse rvation of large carnivores. Conservation Biology 10: 991 1002. Reblo, G. H., and W. E. Magnusson. 1983. An analysis of the effect of hunting on Caiman crocodilus and Melanosuchus niger based on the sizes of confiscated skins. Biological Conservation 26:9 5 104. Redford, K. H. 1992. The empty forest. BioScience 42: 412 422. Resque, F. Jr., A. Alencar, and J. H. Benatti. 2004. Avaliao da abrangncia das reas de Proteo Permanente (APP) em relao a Reserva Legal em ambientes de floresta da Amaznia III LBA ScientificConference, Brazilia.
1 91 Ribeiro, J. F., S. M. Sano, and J. A. Silva. 1981. Chave preliminar de identificao dos tipos fisionmicos da vegetao do Cerrado. Pages 124 133 in Anais do XXXII Congresso Nacional de Botnica. Sociedade Botnic a do Brasil, Teresina, Brasil (in Portuguese). Robinson, H. S., R. B. Wielgus, H. S. Cooley, and S. W. Cooley. 2008. Sink populations in carnivore management: cougar demography and immigration in a hunted population. Ecological Applications 18: 1028 1037. Robinson, J. G., and E. L. Bennett, editors. 2000. Hunting for Sustainability in Tropical Forests. Columbia University Press, New York, USA. Robinson, J. G., and K. H. Re dford. 1991. Neotropical wildlife use and conservation. University of Chicago Press, Chicago, IL. Robinson, J. R., K. H. Redford, and E. L. Bennett. 1999. Wildlife Harvest in Logged Tropical Forest. Science 284: 595 596. Rocha Mendes, F., S. Mikich, G. V. Bianconi, and W. A. Pedro. 2005. Mamferos do municpio de Fnix, Paran, Brasil: etnozoologia e conservao. Revista Brasileira de Zoologia 22: 991 1002. Rolston, H., III. 1981. Values in nature. Environmental Ethics 3: 113 128. Rolston, H., III. 1985. Va luing wildlands. Environ mental Ethics 7: 23 48. Rosas Rosas, O. C., and R. Valdez. 2010. The role of landowners in jaguar conservation in Sonora, Mexico. Conservation Biology 24: 366 371. Rosenberg, D. K., B. R. Noon, and E. C. Meslow. 1997. Biological cor ridors: form, function, and efficacy. BioScience 47: 677 688. Royle, J. A. and K. V. Young. 2008. A hierarchical model for spatial capture recapture data. Ecology 89:2281 2289. Royle, J. A., and R. M. Dorazio. 2008. Hierarchical Models and Inference in Eco logy: The Analysis of Data from Populations, Metapopulations and Communities. Academic Press, Bullington, MA, USA. Royle, J. A., J. D. Nichols, K. U. Karanth, and A. M. Gopalaswamy. 2009a. A hierarchical model for estimating density in camera trap studies. Journal of Applied Ecology 46:118 127. Royle, J. A., K. U. Karanth, A. M. Gopalaswamy, and N. S. Kumar. 2009b. Bayesian Inference in Camera Trapping Studies for a Class of Spatially Explicit Capture Models. Ecology 90(11):3233 3244.
192 Ruiz Garcia, M., E. Pa yn, A. Murillo, D. Alvarez. 2006. DNA microsatellite characterization of the jaguar ( Panthera onca ) in Colombia. Genes & Genetic Systems 81: 115 127. Rutberg, A. T. 2001. Why state agencies should not advocate hunting or trapping. Human Dimensions of Wild life 6: 33 37. Rylands, A. B., and K. Brandon. 2005. Brazilian protected areas. Conservation Biology 19: 612 618. Sanderson, E. W., K. H. Redford, C. L. B. Chetkiewicz, R. A. Medellin, A. R. Rabinowitz, J. G. Robinson, and A. B. Taber. 2002. Planning to sa ve a species: The jaguar as a model. Conservation Biology 16: 58 72. Santos, F. R., A. T. A. Jcomo, L. Silveira. 2008. Humans and jaguars in five Brazilian biomes: Same country, different perceptions. Cat News 4: 21 25. Santos, P. M. 1996. Uso e plano de gesto da fauna silvestre numa rea de vrzea amaznica: a Estao Ecolgica Mamirau (Amazona, Brasil). Dissertao de Mestrado. Faculdade de Cincias da Universidade de Lisboa. pp. 104. Saunders, N. J. 1998. Icons of Power: Feline Symbolism in the Americ as. Routledge, London. Schaffer, M. L. 1983. Determining minimum viable population sizes for the grizzly bear. International Conference on Bear Research and Management 5: 133 139. Schaller, G. B. 1979. On the status of jaguar in the Pantanal. Report to Ins tituto Brasileiro Desenvolvimento Florestal, Braslia. Schaller, G. B., and J. M. C. Vasconcelos. 1978. Jaguar predation on capybara. Z. Saeugetierk 43: 296 301. Scognamillo, D; Maxit, IE; Sunquist, M; Polisar, J. 2003. Coexistence of jaguar ( Panthera onca ) and puma ( Puma concolor ) in a mosaic landscape in the Venezuelan llanos. J. Zool., Lond. 259: 269 279. Seber, G. A. F. 1970. Estimating time specific survival and reporting rates for adult birds from band returns. Biometrika 57:313 318. Seymour, K. L. 19 89. Panthera onca. Mammalian Species 340: 1 9. Silva, J. M. C., A. B. Rylands, G. A. B. Fonseca. 2005. The fate of the Amazonian Areas of Endemism. Conservation Biology 19: 689 694. Silva, M. 2005. O programa Brasileiro de unidades de conservao. Megadive rsidade. Desafios e oportunidades para a conservao da biodiversidade no Brasil 1(1): 22 26.
193 Silveira, L. 2004. Ecologia comparada e conservao da ona pintada ( Panthera onca ) e ona parda (Puma concolor), no Cerrado e Pantanal. Tese de doutorado. Univer sidade de Braslia. Silveira, R., E.E. Ramalho, J. Thorbjarnarson, and W. Magnusson. 2010. Depredation by jaguars on caimans and importance of reptiles in the diet of jaguar. Journal of Herpetology 44: 418 424. Singh, P., A. M. Gopalaswamy, A. J. Royle, N. S. Kumar, and K. U. Karanth. 2010. SPACECAP: A program to estimate animal abundance and density using Bayesian spatially explicit capture recapture models. Version 1.0. Wildlife Conservation Society India Program, Centre for Wildlife Studies, Bangalure, India. Smith, N. J. 1976. Spotted cats and the Amazon skin trade. Oryx 13: 362 371. Soares, B. S., D. C. Nepstad, L. M. Curran, G. C. Cerqueira, R. A. Garcia, C. A. Ramos, E. Voll, A. McDonald, P. Lefebvre, and P. Schlesinger. 2006. Modeling conservation in the Amazon basin. Nature 440: 520 523. Soisalo, M. K., and S. M. C. Cavalcanti. 2006. Estimating the density of a jaguar population in the Brazilian Pantanal using camera traps and capture recapture sampling in combination with GPS radio telemetry. Biol ogical Conservation 129: 487 496. Soul, M., and R. Noss. 1998. Rewilding and biodiversity: Complementary goals for continental conservation. Wild Earth 8(3):18 28. Stephens, D., and J. Krebs. 1986. Foraging theory. Princeton University Press, Princeton, N J. Sunquist, M. E., and G. G. Montgomery. 1973. Activity patterns and rates of movement of two toed and three toed sloths ( Choloepus hoffmanni and Bradypus infuscatus ). Journal of Mammalogy 54(4) :946 954. Sunquist, M., and F. Sunquist. 2002. Wild Cats of the World. The University of Chicago Press, USA. Sunquist, M., U. K. Karanth, and F. Sunquist. 1999. Ecology, behaviour and resilience of tiger and its conservation. Pages 5 18 in J. Seidensticker, S. Christie, and P. Jackson, editors. Riding the tiger. Ca mbridge University Press, Cambridge. Swank, W. G., and J. G. Teer. 1989. Status of the jaguar 1987. Oryx 23(1):14 21. Taber, A. B.; Novaro, A. J.; Neris, N. e Colman, F. H. 1997. The food habits of sympatric jaguar & puma in the Paraguayan Chaco. Biotrpi ca 29: 204 213.
194 Terborgh, J., N. Pitman, M. Silman, H. Schichter, and P. Nuez V. 2002. Maintenance of tree diversity in tropical forests. Pages 1 17 in D. J. Levey, W. R. Silva, and M. Galetti, editors. Seed dispersal and frugivory: ecology, evolution and conservation. CAB International, Wallingford, UK. Thorbjarnarson, J., and R. Da Silveira. 2000. Secrets of the flooded forest. Natural History 109:70 79. Treves A. 2009. Hunting to conserve large carnivores. Journal of Applied Ecology 46:1350 1356. Treves A., and K. U. Karanth. 2003. Human carnivore conflict: Local solutions with global applications. Conservation Biology 17: 1489 1490. Trolle, M., and M. Kry. 2003. Density estimation of ocelot ( Leopardus pardalis ) in the Brazilian Pantanal using capture recapture analysis of camera trapping data. Journal of Mammalogy 66:13 21. Trolle, M., and M. Kry. 2005. Camera trap study of ocelot and other secretive mammals in the northern Pantanal. Mammalia 69(3 4):405 412. UNEP (United Nations Environment Programme ), and ACTO (Amazon Cooperation Treaty Organization). 2008. GEO Amaznia: Perspectivas do Meio Ambiente na Amaznia. Usher, M. B., editor. 1986. Wildlife conservation evaluation. Chapman & Hall, New York. Valsecchi, J. A. 2005. Diversidade de Mamferos e U so da Fauna nas Reservas de Desenvolvimento Sustentvel Mamirau e Aman. Dissertao de mestrado, Museu Paraense Emlio Goeldi e Universidade Federal do Par. Belm, 2005. Wallace, R. B., H. Gmez, G. Ayala, and F. Espinoza. 2003. Camera trapping for jagu ar ( Panthera onca ) in the Tuichi valley, Bolivia. Mastozoologa Neotropical 10(1):133 139. Weber, W., and A. Rabinowitz. 1996. A global perspective on large carnivore conservation. Conservation Biology 10: 1046 1054. White, G. C., and K. P. Burnham. 1999. Program MARK: survival estimation from populations of marked animals. Bird Study 46:120 138. Williams, B. K., J. D. Nichols, and M. J. Conroy. 2002. Analysis and management of animal populations. Academic Press, San Diego, California, USA. Zar, J. H. 1984. Biostatical analysis. Prentice Hall, Englewood Cliffs, New Jersey.
195 BIOGRAPHICAL SKETCH E m i l i a n o Esterci Rama lho was born in 1978 in Rio de Janeiro, Brazil, where he lived until the age of 23. He completed h i s undergraduate degree in Biological Sciences, majoring in Ecology, in t he Federal University of Rio de Janeiro (UFRJ) in 2003 H e concluded his m aster s in the National Institute of Research of Amazonia (INPA) in 2006 In 2007 he was awarded a CAPES/Fulbright scholarship to pursue his PhD in the United States, where he studied in t he University of Florida in the Department of Wildlife Ecology and Conservation He has been involved in jaguar research and conservation since 2003, and ha s been studying he jaguar in Amazonia since 2004. He is currently coordinating jaguar research activities in M amirau and Aman Sustainable Development Reserves, and he is also part of a non governmental organization called Pr Carnvoros.