1 Biological Conservation 109 (2003) Jaguars, pumas, their prey base, and cattle ranching: ecological interpretations of a management problem John Polisar a, *, Ines Maxit b,1, Daniel Scognamillo b,2, Laura Farrell c, Melvin E. Sunquist b, John F. Eisenberg a a Florida Museum of Natural History, University of Florida, PO Box , Gainesville, FL , USA b Department of Wildlife Ecology and Conservation, University of Florida, PO Box , Gainesville, FL , USA c Museum of Comparative Zoology, Harvard University, 26 Oxford Street, Cambridge, MA 02138, USA Received 27 August 2001; received in revised form 29 December 2001; accepted 25 April 2002 Abstract Jaguar and puma depredation on livestock may be influenced by (1) innate and learned behavior; (2) health and status of individual cats; (3) division of space and resources among jaguar and puma; (4) cattle husbandry practices; and (5) abundance and distribution of natural prey. Our study in Los Llanos of Venezuela aimed to establish how all these elements related to cattle being lost to cat depredation. Prey distribution was influenced by forest composition, topographical characteristics, and degree of habitat interspersion. The biomass of natural prey in the study area was adequate to support the resident large cats without a subsidy of livestock. Selective rather than opportunistic hunting by the cats reinforced that conclusion. Puma were responsible for more attacks on livestock than jaguar, frequently in maternity pastures in upland areas of relatively low prey availability. Management recommendations are discussed that may be relevant to other savanna/forest mosaics of South America. # 2002 Elsevier Science Ltd. All rights reserved. Keywords: Jaguar; Puma; Prey; Predator; Cattle ranching; Venezuela 1. Introduction Jaguar (Panthera onca) and puma (Puma concolor) depredation on livestock may be influenced by: (1) innate and learned behavior; (2) health and status of individual cats; (3) division of space and resources among jaguar and puma; (4) cattle husbandry practices; and (5) abundance and distribution of natural prey. Predators select prey based on a cost benefit analysis (search time, handling costs, and energy gained in the context of prey abundance) (Emlen, 1966; MacArthur and Pianka, 1966) and vulnerability factors (Curio, 1976; Taylor, 1976; Temple, 1987). In productive environments, whether homogenous or heterogenous, predators can be expected to be more selective than in * Corresponding author. addresses: (J. Polisar), ifas.ufl.edu (M.E. Sunquist), (L. Farrell). 1 Present address: Louisiana Natural Heritage Program, Louisiana Department of Wildlife and Fisheries, P.O. Box 98000, Baton Rogue, LA , USA. 2 Present address: School of Renewable Natural Resources, Louisiana State University AgCenter, Baton Rouge, LA 70803, USA. unpredictable environments (Emlen, 1966; MacArthur and Pianka, 1966). The value of a patch, in terms of available prey, is usually reduced by predators, stimulating them to search for alternative patches (Charnov, 1976). This predicts roaming among patches in all instances except those where patch values are resilient. These general postulates have to be able to absorb the variation introduced by learned behaviors and individual preferences. Among five intensively monitored female mountain lions (Puma concolor) in Alberta, two never killed bighorn sheep (Ovis canadensis), one killed one sheep, one killed five, and one killed 17, in 1 year killing 8.7% of an early-winter herd (26.1% of its lambs) (Ross et al., 1997). All five cats were healthy, had alternative prey available, and made varying use of those alternatives. The learned ability to handle bighorn sheep, normally more difficult to take than mule deer (Odocoileus hemionus) reduced handling costs for one puma. Risk of injury is a component of potential handling costs (Sunquist and Sunquist, 1989). Preference for certain natural or domestic prey may be transmitted from mother to young (Hoogesteijn and Mondolfi, 1993; Mondolfi and Hoogesteijn, 1986; Quigley and Crawshaw, 1992) /02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S (02)00157-X
2 298 J. Polisar et al. / Biological Conservation 109 (2003) Interactions among predators may influence choice of prey. Seidensticker (1976) commented on the potential effects of social dominance. In areas shared with tigers (Panthera tigris) the behaviorally flexible social subordinate leopard (P. pardus) appeared to allow the dominant tiger first choice of both habitats and prey (Eisenberg and Lockhart, 1972; Seidensticker, 1976). Spatial avoidance of a larger predator is likely to influence diet. Where there is a choice, leopards emphasize smaller prey than tigers (Panthera tigris) (Karanth and Sunquist 1995), but Karanth and Sunquist (2000) found no evidence of spatial exclusion of leopards by tigers. Overlap in carnivore diets may increase when a resource is so abundant that there is little competition (Colwell and Futumaya, 1971; Dailey et al., 1984; Spowart and Thompson Hobbs, 1985). Environments fluctuate; seasonally, annually, with patterns, even erratically. It follows that levels of interspecific competition fluctuate. The present versions of jaguar and puma coexistence are recent and perhaps still in flux (Culver et al., 2000; Morgan and Seymour, 1997). A high proportion of study animals have been killed by ranchers and poachers in every jaguar study to date (Sunquist, in press). Measures that reduce the frequency of large cat depredation on livestock may go a long ways towards maintaining cat populations. The concerns of cattle ranchers are immediate and practical. How can cattle losses to cats be reduced? Roosevelt (1914) observed that ranches in Brazil that possessed abundant native prey experienced fewer jaguar problems. Shaw (1977) hypothesized that the number of cattle taken by puma in Arizona was inversely proportional to the size of the deer herd. Mondolfi and Hoogesteijn (1986) and Hoogestein et al. (1993) hypothesized a similar relationship for jaguar and puma in Venezuela, where the cats exploit a more diverse prey base. Cattle management offers some possibilities. In some areas cattle have been so lightly managed that they resemble wild prey (Hoogesteijn et al., 1993; Hoogesteijn and Mondolfi, 1993; Mondolfi and Hoogesteijn, 1986; Quigley and Crawshaw, 1992; Schaller and Crawshaw, 1980). Hoogesteijn et al. (1993) suggested that losses could be reduced by: (1) excluding cattle from forest; (2) maintaining adequate distance between calving areas and forests; (3) moving calves out of problem areas and replacing them with bulls; and 4) maintaining adequate populations of wild prey. In 1996, we initiated field work on a team project designed to examine all the factors that could contribute to cat-cattle conflicts: (1) ecology and behavior of jaguar and puma; (2) abundance and distribution of natural prey; and (3) cattle management practices. This paper addresses the following questions: Could the natural prey base in the study area support the cats or did they need a subsidy from domestic livestock? What were the dominant components of jaguar and puma diet? How did prey selection relate to prey availability? What scenarios lead to cat/cattle conflicts? What measures can be taken to reduce those conflicts? 2. Methods 2.1. Study area Hato Pin ero is a working 80,000 ha cattle ranch/ wildlife preserve located between and 9 00 N and and W in the southeast corner of the state of Cojedes in north-central Venezuela (Eisenberg and Polisar, 1999; Miller, 1992). The northern boundary of Pin ero lies among hills that rise to 396 m above sea level (Farrell, 1999). The western boundary is formed by the Cojedes and Portuguesa rivers, the southern and eastern boundaries by the Chirgua and Pao rivers (Fig. 1). Smaller streams (can os) run through this basin. The lowest elevations are approximately 65 m above sea level in the open savannas in the southern part of the ranch. The landscape can be characterized as a complex mosaic of interdigitated forests and open areas with vegetation types based on interactions of elevation, substrates, and hydrology. The ratio of open to forested areas is roughly 50:50. Our 63,227 ha study area contained seasonally flooded lowland savanna (39.06%), seasonally flooded semi-deciduous forest (33.90%), dry hillside savannas with chaparral (15.26%), dry hillside semi-deciduous forest (7.88%), pastures in highlands that never flooded (2.86%), evergreen forest (.08%) and mango groves (0.01%), with the remainder developed for human habitation and livestock maintenance. The climate is strongly seasonal, with the majority of the mm of precipitation falling during the wet season between the beginning of May and the end of November. The dry season, from December though April is hotter. Relatively impermeable soils causes surface water to accumulate starting in June and peaking in July and August. The inundation is relatively shallow, with greatest depths occurring in low savannas in the southern portion of the ranch. Forests typically retain pockets of dry land whereas the savannas in the south flood completely for several months. The majority of Pin ero s 14,000 head of cattle are Bos indicus cebu races (nelore, brahma, guzerat, gir). Approximately 420 horses, mules, and burros fulfill working and breeding needs. A herd of approximately 150 water buffalo are maintained in the southern savannas. Many cattle are moved from lowland pastures to higher areas during the wet season. Cows can forage in chest-deep water, but calf mortality is high in flooded conditions. Nursing is hindered and the wet muddy calves are vulnerable to biting insects. Artificial insemination results in a pulse of calving from between
3 J. Polisar et al. / Biological Conservation 109 (2003) Fig. 1. Northwest and north-central Venezuela. Shaded areas show location of Hato Pin ero study area in relation to locations where Smithsonian research projects took place in the 1970s. July and September. The maternity pastures where this takes place are high, well-drained areas Prey abundance and distribution Prey abundance, distribution, and population structure were assessed using vehicle transects, linear foot transects, point counts, night counts, camera trapping, capture mark recapture methods, and opportunistic observations (Buckland et al., 1993; Lancia et al., 1994; Seber, 1982) Habitat characteristics Patterns of forest vegetation were assessed using 35,000 m 2 of quantitative sampling in a vertical profile of Pin ero (Polisar, 2000). Parameters describing physiognomy were recorded at 100 m intervals along the 26 foot transects used for animal observations (Polisar, 2000). Forest types were classified using cluster analyses (SPSS, 1999). Forest composition was tabulated for different forest types, and number of species and percent of individuals in trees, vines, and under story plants that provided food for primary prey extracted for each type (Polisar, 2000). A vegetation map used for areal estimates of habitat types and felid habitat selection analyses using ArcView (1995) was based on Landsat Thematic Mapper imagery, cluster analysis based forest classifications, and ground-truthing with global positioning units Jaguar and puma prey consumption Of five jaguars and six pumas captured, four jaguars were radio-monitored for a total of 53 months and five pumas were radio-monitored for a total of 68 months (Farrell, 1999; Maxit, 2001; Scognamillo, 2001). Telemetry data, and tracks of jaguars and puma in areas not used by collared animals, augmented by observations and camera trapping images accrued during prey survey efforts provided estimates of the study area s cat population. Jaguar and puma scats were collected opportunistically on trails and roads, identified by the presence of tracks or through mitochondial DNA analyses (Farrell et al. 2000), and prey species in scats were based upon hair fibers, teeth, or scales (Farrell, 1999; Maxit, 2001; Scognamillo, 2001). Detailed notes were taken at all cat kills, whether domestic livestock or free-ranging natural prey (Farrell, 1999; Maxit, 2001; Scognamillo, 2001). Age/stage of natural prey kills was determined using indices provided in Dimmick and Pelton (1994) and Ojasti (1973) for white-tailed deer, collared peccary, and capybara (Maxit, 2001; Scognamillo, 2001), and by calibrations of head-length to snout-vent-length and weight obtained from spectacled caiman in Pin ero (Polisar, 2000).
4 300 J. Polisar et al. / Biological Conservation 109 (2003) The minimum annual requirement of killed prey necessary to sustain Hato Pin ero s jaguar and puma population was estimated by multiplying average weights of captured animals by estimated numbers of cats (jaguars: 2.5 adult males and 4 adult females, 2 subadults, and 3 cubs; pumas: 5 adult males, 7 adult females, 3 subadults, 5 cubs) using consumption requirements of 34 g/day/kg of cat for jaguars and 38.5 g/day/kg cat for pumas (extracted from Emmons, 1987; Scognamillo, 2001). We used an estimate of 70% consumption of killed prey based on estimates of 70 79% for pumas (Ackerman, et al. 1986; Hornocker, 1970) 75% for jaguars (Emmons, 1987) and 70% for tigers (Sunquist, 1981). Average weights of captured adult male and female jaguars respectively were 87.5 kg (n=2) and 52 kg (n=2). Average weights of adult male and female pumas respectively were 51 kg (n=2) and 25.5 kg(n=4). Since average jaguar weights (70 kg) were 1.88 times average puma weights (37 kg) (Maxit, 2001), we presumed higher metabolic rates and consumption needs per kg of bodyweight for puma (Emmons, 1987). Subadult consumption requirements were considered equal to adult females of respective species (Sunquist, 1981). Cub requirements were estimated as 25% that amount (Sunquist, 1981), an estimate that averaged size-classes from weaning to subadult Prey biomass estimates Standing crop biomass of prey was derived from abundance estimates and population structure data based on group counts, night counts, transect-based density estimates, and capture mark recapture sampling (Polisar, 2000). Body weights were obtained in single or repeated captures (Polisar, 2000) or in appropriate literature from studies completed nearby (Ayarzagüena, 1983; Brokx, 1972; Eisenberg et al., 1979; Linares, 1998; Ojasti, 1973). Livestock biomass was estimated using figures provided by Ferdinando Corrales (Manager of Agropecuaria San Francisco, father company of Hatos Piñero, Paraima, Sembra, and slaughterhouse; Corrales, 1998). Gross productivity was figured using our standing crop biomass and population structure estimates, combined with values for demographic parameters (average number of litters per year/average number of young per litter/stage specific survival rates) and growth rates that were either local or realistic as possible (Bodmer et al., 1997; Brokx, 1972; Eisenberg et al., 1979; Hayne, 1984; Hellgren et al., 1995; Kleiman et al., 1979; Ojasti, 1973; Ojeda and Keith, 1982; Smythe, 1978; Sowls, 1997; Teer, 1984). 3. Results 3.1. Prey biomass Standing crop biomass of all major food species (excluding livestock) was 374,489 kg, of which 149,988 (40%) was mammalian and 224,501 (60%) was reptilian (Table 1). Annual minimum killing requirements for resident jaguars and pumas were estimated at 10,100 kg and 10,878 kg respectively, or 20,978 kg combined. Minimum killing requirements for both cats combined represented 5.6% of the standing crop: 2.7% for jaguar, 2.9% for puma. Including tapir, terrestrial tortoises, and iguanas, potential food items not actually represented in the diet, standing crop was 470,825 kg (Table 1). The standing crop of cattle was around 4,656,000 kg of which 160, ,000 kg were in the size class most vulnerable to large cats (Table 2). Buffalo constituted around 123,750 kg, and horse, mules, and burros pooled 118,300 kg (Table 2). Biomass estimates, for the entire 63,227 ha study area, are presented as kg/km 2 in Table 3. Percentages of the kg/km 2 pooled along taxonomic groups and the domestic (introduced) wild (recent native) dichotomy are presented in Table 4. Pin ero biomass estimates including and excluding domestic livestock are compared to other sites in the New and Old World in Table 5. An estimate of annual gross productivity of major native mammalian prey is presented in Table 6. Cat killing needs represented a maximum of 29% of gross annual mammalian productivity (Table 6). Average adult weights for deer, capybara, collared peccary, and white-lipped peccary are 40, 46, 23, and 35 kg, respectively (Bodmer et al., 1997; Brokx, 1972; Eisenberg et al., 1979; Ojasti, 1973; Sowls, 1997). Large male deer may weigh slightly over 50 kg, as may large capybara (Brokx, 1972; Ojasti, 1973). The largest caiman kill recorded was around 50 kg. The largest caiman weighed was 75 kg. The largest anaconda weighed was 50 kg. In general, 50 kg was large prey in the region Jaguar and puma use of prey and habitats Despite considerable overlap in prey species, jaguar and puma diets were significantly different whether examined with livestock included or excluded (Table 7) (Maxit, 2001; Scognamillo, 2001). Jaguar showed a preference for large-bodied prey, with puma taking more medium-sized and smaller prey wild prey than jaguar (Maxit, 2001; Scognamillo, 2001). Puma dietary niche breadth exceeded that of jaguar (Maxit, 2001). Jaguars preyed selectively upon capybaras and collared peccaries, took white-tailed deer and caiman less than expected, and killed white-lipped peccaries in proportion to availability (Maxit, 2001; Scognamillo, 2001). The most important wild prey in the jaguar diet in terms of frequency of occurrence and biomass consumed were collared peccary, capybara, and white-lipped peccary (Table 7). Puma took most large prey in proportion to availability, but selected for collared peccaries, and took caiman less than expected according to availability (Maxit, 2001; Scognamillo, 2001). The frequency of
5 Table 1 Natural prey abundance, distribution, and biomass J. Polisar et al. / Biological Conservation 109 (2003) Species Habitat Habitat specific density (km 2 ) of prey Prey abundance in 63,227 ha study area Total prey biomass in study area (kg/63,227 ha) OV All forest BSD & BS & BSV High dry pasture PS Low flooding pasture SI 8.3 1, All habitats pooled 2, ,890 TT All forest BSD & BS & BSV 7.5 1, ,835 TP All occupied habitats (primarily BSD) DA All forest BSD & BS & BSV SF Semi-deciduous forest BSD Dry forest BS Forests pooled DN Dry forest BS Semi-deciduous forest BSD Forests pooled MT All forest HH All occupied habitats ,654 GC All forest BSD & BS & BSV 95 22,300 96,336 CC All occupied aquatic habitats 15, ,827 FWT All occupied aquatic habitats 45, ,674 Species codes as follows: OV Odocoileus virginanus; TT Tayassu tajacu; TP Tayassu pecari; DA Dasyprocta agouti; SF Sylvilagus floridanus; DN Dasypus novemcinctus; MTMyrmecophaga tridactyla; HHHydrochaeris hydrochaeris; GCGeochelone carbonaria; CCCaiman crocodilus; FWT freshwater turtles (Podocnemis voglii, Chelus fimbriatis, P. unifilis). Habitat codes as follows: BSD semi-deciduous forest; BS dry forest; BSV evergreen forest; SI seasonally flooding savanna; and PS high dry pastures. Abundance based upon DISTANCE density estimates except where noted (Polisar, 2000). TP and HH abundance estimates based upon counts in concentration areas, precluding density estimates. CC and FWT abundance estimates based upon night counts adjusted for sighting fraction and mark-recaptures in concentrations that result from dry season low water levels. Water surface area is in constant flux, thus no density estimates are presented (Polisar, 2000). Table 2 Patterns of biomass among livestock at Hato Pin ero Sex/age class Crude numbers Crude weight (kg) Biomass/class (Jul Sep) Biomass/class (Oct Dec) Biomass/class (Jan Mar) Bulls , , ,000 Cows ,940,000 2,940,000 2,940,000 Newborns (0 3 months) ,000 Young calves (3 6 months) ,000 Calves 6 9 months , months , , ,000 Total biomass (max and min) Minimum 4,240,000 4,464,000 Maximum 4,656,000 Biomass of class (most vulnerable to big cats) Minimum 160,000 Maximum 384,000 Sex and age specific numbers, weights, and biomass of cattle (Bos indicus and Bos taurus). Estimated cattle total is 14,000. The peak of parturition is during July, August, and September. Stages most vulnerable to attacks by large cats are shown in italics. In addition to cattle there are: buffalo (Bubalus bubalis) 123,750 kg ( kg); horses and mules 106,750 kg ( kg); young horses and mules 11,100 kg ( kg); and burros 450 kg (3150 kg). Total equidae biomass estimate is 118,300 kg, of which 11,100 would be colts. capybara in puma scats was equal to that of deer, although the large rodents were only one quarter as abundant. The four most important wild prey items in puma diet on bases of both frequency of occurrence and biomass were collared peccary juveniles, deer, capybara, and caiman (Table 7). While both jaguar and puma showed a preference for collared peccaries, pumas were mostly taking juveniles (Maxit, 2001; Scognamillo, 2001). Capybara represented 71% of the sample of wild prey killed by puma and 33% of the jaguar wild prey kill sample. Spectacled caiman were 39% of the sample of wild prey killed by jaguar (Table 7) (Maxit, 2001; Scognamillo, 2001). In terms of numbers of large prey available at any given time, the cumulative hunting pressure upon capybara by big cats was high, and the numbers of large reptiles consumed was less than expected. Analyses of Hato Pin ero s books indicated that 13% of total livestock losses in 10 years could be attributed to large cats. These losses were highest during the peak calving months of August October when on average 20 calves annually were killed by cats. Puma were responsible for 86% of that damage (Maxit, 2001; Scognamillo,
6 302 J. Polisar et al. / Biological Conservation 109 (2003) Table 3 Conversions of biomass estimates for 63,227 ha study area into kg/ km 2 estimates Item kg/632.3 km 2 kg/km 2 Native mammals Capybara 22, Agouti White-lipped peccary Collared peccary 35, White-tailed deer 78, Cottontail rabbit Nine-banded armadillo Giant anteater Subtotal native 149, mammals recorded in jaguar and puma diet Tapir Total native 152, Domestic introduced mammals Cattle 4,656, Buffalo 123, Horses, mules, burros 118, Domestic subtotal 4,898, Total mammalian 5,051, Reptiles Iguana Caiman 167, Freshwater turtles 56, Terrestrial tortoises 96, Total reptilian 321, Table 4 Percentage of total crude mammalian biomass expressed in kg/km 2 represented by select groups Group Pooled group Percent of total mammalian biomass Bovidae (introduced) & Cervidae (native) & Tayassuidae (native) Artiodactyla 96.99% Equidae (introduced) & Tapiridae (native) Perissodactyla 2.40% Agoutidae (native) & Hydrochaeridae (native) Rodentia 0.47% Myrmecophagidae Xenarthra 0.13% Bovidae (introduced) Artiodactyla & Equidae (introduced) Perissodactyla 96.97% Native mammalian prey 3.03% 2001). The results of our study support those trends documented in the ranch ledgers (Table 7). Sixty-nine percent of attacked calves were between 1 and 30 days of age (Scognamillo et al., in press). All five adult puma home ranges included some of the high dry pastures used for calving during the wet season ( % of the home ranges). For three pumas, use of the vegetation type exceeded within-home-range-availability by factors of (Maxit, 2001; Scognamillo, 2001). In contrast, that vegetation type was included in two of four adult jaguar home ranges, and even then scarcely so ( % of those home ranges). There was no evidence of significant temporal segregation between the two cats (within 3-h intervals), and inter-specific spatial overlap was high on a coarse scale (Scognamillo, 2001). On a fine scale, puma avoided the immediate presence of jaguar (Scognamillo, 2001). Both big cats intensively used the forest savanna ecotone (Scognamillo, 2001). Pumas were located significantly more often within the first 500 m of large (>300 ha) forest patches than jaguars, which were located significantly more often within the interior of such patches (Maxit, 2001). 4. Discussion In Nepal, the Serengeti, and the Amazon, large predators kill approximately 8 10% of the standing crop biomass of mammalian prey (Emmons, 1987; Schaller, 1972; Sunquist, 1981). In Hato Pin ero that proportion would require a minimum of 209,780 to 262,225 kg. With wild mammals and reptiles combined there was a minimum of 374,489 kg (Table 1), enough to support as many cats as were present. If potential dietary components apparently being bypassed were factored in, such as red-footed tortoises (Table 1) the margin by which the estimated minimum is exceeded becomes wider. In terms of standing crop prey biomass, the cats were killing approximately 4.46% of all potential wild prey, 5.61% of the species actually being used, and 13.99% of mammals alone. This suggests that, in context of the entire study area, a subsidy from domestic livestock was not necessary to sustain Hato Pin ero s large cats. Biomass ratios indicate that wild mammals could provide approximately 57 71% of the annual requirements, predicting some use of reptiles, which did occur (Table 7). It is fitting to assess the proportion that mammals contribute before the reptiles. Many caiman and turtles were inaccessible to cats, safe in the depths of their aquatic environments. The estimated annual needs of the cats constitute 29% of gross productivity of major native mammalian prey (Table 6). When considering sustainable consumption by humans, Robinson and Redford (1991) and Robinson and Bodmer (1999) suggest a maximum harvest of 20%
7 Table 5 Mammalian biomass in New and Old World study areas J. Polisar et al. / Biological Conservation 109 (2003) Sites Comments kg/km 2 New World Hato Pin ero, llanos, Venezuela Not including small mammals Including livestock 7988 Hato Pin ero, llanos, Venezuela Not including small mammals Excluding livestock 242 Hato Masaguaral, llanos, Venezuela All nonvolant mammals Including livestock 8315 Barro Colorado Island, Panama All nonvolant mammals 2115 Guatopo, Coastal Range, Venezuela All nonvolant mammals 1001 Urucu, Brazilian Terra Firma Amazon All nonvolant mammals 891 Acurizal, Pantanal, Brazil Most nonvolant mammals 380 Old World Wilpattu, Sri Lanka Ungulates only 766 Kanha, India Primarily ungulates 1708 Nagarahole, India Wild & domestic ungulates, primates 15,094 Serengeti Unit, Tanzania Primarily ungulates 4222 Manyara, Tanzania Primarily ungulates 7785 Ngorongoro Crater, Tanzania Primarily ungulates 10,363 Sources are as follows: Hato Masaguaral, Barro Colorado, Guatopo (Eisenberg, 1980); Urucu (Peres, 1999); Pantanal (Schaller, 1983) Wilpattu (Eisenberg and Lockhart, 1972; McKay and Eisenberg, 1974; Eisenberg and Seidensticker, 1976); Kanha (Schaller, 1967; adapted by Eisenberg and Seidensticker, 1976); Nagarahole (Karanth and Sunquist, 1992) Serengeti, Manyara, and Ngorogoro Crater (Schaller, 1972). Table 6 Gross productivity (of major mammalian prey) at Hato Pin ero Species Abundance Standing crop (kg) Annual addition of biomass per 100 animals (kg) Annual gross Productivity (kg) Capybara ,315 22, ,554 White-lipped peccary White-tailed deer , ,887 Collared peccary , ,224 Total 73,363 Presented as increments of biomass added per year. Subtractions (adult mortality and otherwise) not included. Estimated annual kill needed to support resident cats (20,978 kg) is 29% of gross productivity estimate above. of production for long-lived species, and 40% for shortlived species. Last reproduction occurred at over 10 years of age in long-lived species and between 5 and 10 years of age in short-lived species. White-tailed deer are short to medium lived species, with emphasis on short. Few deer live over 10 years and life expectancy in the wild is frequently less than 3 years (Brokx, 1972; Winston, 1991). There were no capybara over 5 years of age in a harvested population in Apure, Venezuela (Lord and Lord, 1988), although Robinson and Redford (1986) estimated age at last reproduction as 9 years using Ojasti (1973). Kleiman et al. (1979) commented on the high reproductive potential of capybara. Using last age of reproduction estimates of 13 years for collared and white-lipped peccaries, Robinson and Redford (1991) classify them as long-lived species. Hellgren et al. (1995) found fecundity rates in collared peccaries over 7 years of age were less than animals between 3 and 7 years of age, roughly equal to animals 2 3 years of age, and greater than animals between 1 and 2 years of age. Some wild females in Texas and Arizona did exceed 10 years in age (Hellgren et al., 1995; Sowls, 1997). Though classified as a long-lived species, Robinson and Redford (1991) commented on the high productivity of peccaries and Robinson and Bodmer (1999) consider a harvest < 40% of production sustainable for both species. The addition of caiman and freshwater turtle production would elevate the total prey production estimate for Hato Pin ero considerably. Production estimates support the assertion that, when the entire study area is considered, resident cats did not require a livestock subsidy. An efficient predator will accept all potential prey encountered when food is scarce or unpredictable, and exercise greater selectivity when food is common and adequate productive patches known (Emlen, 1966; MacArthur and Pianka, 1966; Sunquist and Sunquist, 1989). Thus, diet breadth, in the context of diversity of potential prey, can reflect relative scarcity or abundance of prey. Analyses of jaguar scats from the Peruvian Amazon yielded 40 prey taxa (n=25) (Emmons, 1987). Rabinowitz and Nottingham (1986) recovered 17 taxa from 228 scats in Belize. Analyses from the Chaco of
8 304 J. Polisar et al. / Biological Conservation 109 (2003) Table 7 Numbers of items and proportions of different prey species in jaguar and puma scats and kills at Hato Pinero (data from Scognamillo, 2001) Prey size Prey species Jaguar Puma Scats n (%) Kills n (%) Scats n (%) Kills n (%) Large size prey (>15 kg) Tayassu tajacu 11 (26) 5 (16) 2 (5) 1 (2) Tayassu pecari 5 (12) 0 (0) 1 (2) 0 (0) Hydrochaeris hydrochaeris 9 (21) 6 (20) 4 (10) 15 (30) Odocoileus virginianus (adult) 2 (5) 0 (0) 4 (10) 4 (8) Myrmecophaga tridactyla 4 (10) 0 (0) 0 (0) 0 (0) Caiman crocodilus 0 (0) 4 (13) 0 (0) 0 (0) Livestock 3 (7) 10 (33) 10 (24) 29 (58) Subtotals 34 (81) 25 (83) 21 (50) 49 (98) Medium size prey (1 15 kg) Tayassu tajacu (juvenile) 1 (2) 0 (0) 5 (12) 0 (0) Odocoileus virginianus (juvenile) 0 (0) 0 (0) 0 (0) 1 (2) Procyon cancrivorus 2 (5) 0 (0) 0 (0) 0 (0) Sylvilagus floridanus 1 (2) 0 (0) 3 (7) 0 (0) Dasyprocta agouti 0 (0) 2 (7) 1 (2) 0 (0) Dasypus novemcinctus 0 (0) 0 (0) 1 (2) 0 (0) Caiman crocodilus 3 (7) 1 (3) 4 (10) 0 (0) Podocnemis voglii 0 (0) 2 (7) 0 (0) 0 (0) Subtotals 7 (17) 5 (17) 14 (33) 1 (2) Small prey (<1 kg) Small rodents or marsupials 0 (0) 0 (0) 7 (17) 0 (0) Unidentified birds 1 (2) 0 (0) 0 (0) 0 (0) Subtotals 1 (2) 0 (0) 7 (17) 0 (0) Paraguay yielded 23 taxa (n=106) (Taber et al., 1997). Fifty jaguar scats from the dry forests of Jalisco, Mexico, yielded seven prey species (Nun ez et al. 2000). Based on 42 scats, jaguar diets at Pin ero include approximately 10 taxa (Table 7). Comparing diet breadth among studies has several confounding factors. Innate prey diversity among the study areas varies. Sample sizes were not equal among studies. Nun ez et al. (2000) estimated scats as the minimum to adequately document diet. Anderson (1983) suggested a sample of scats was necessary to calculate food habits of pumas within 10% of actual use patterns. Emmons (1987) obtained diverse taxa with only 25 jaguar scats. The number at which the asymptote is obtained is likely to vary among sites. In Pin ero rank importance of prey items varied little between subsamples of 30 scats and the entire samples of 42 scats for both cats (Maxit, 2001). The diet of jaguars in Pin ero appears to be more specialized than the diet in Peruvian and Belizean rainforest sites (Emmons, 1987; Rabinowitz and Nottingham, 1986). Foraging theory predicts that items will be added to the diet only when the energy gained outweighs the costs invested (Emlen, 1966; MacArthur and Pianka, 1966). In Belize, 54% of jaguar scats contained armadillos (Rabinowitz and Nottingham, 1986). In Pin ero, armadillos outnumber capybara (Table 1). The search time required to obtain armadillos (relatively dispersed in forest habitats) probably equals or exceeds the search time for capybara (relatively concentrated in somewhat predictable habitats, rarely > 500 m from water; Ojasti 1973). Handling costs to capture armadillo may equal handling costs for capybara. Capybara weights are ten times as much as armadillos, which in Pin ero, are practically ignored (Table 7). In the Peruvian Amazon, terrestrial tortoises tied with collared peccary as the numerically most frequent items in the diet (Emmons, 1987, 1989). In Pin ero, where terrestrial tortoises were an order of magnitude more abundant than the larger mammalian prey (Table 1), they were virtually ignored. The large cats in Pin ero have adequate natural prey to make choices, another indication that the natural prey base is adequate. The prey base is adequate when the entire Pin ero study area is considered, but it is far from uniformly distributed (Fig. 2). Pin ero has lower prey diversity than rain forests, but its high horizontal heterogeneity results in patches where prey production is high. Cats move through and between those patches. Some semi-deciduous forests have seasonal concentrations of white-lipped peccaries. Others have resident groups of collared peccaries. Prey abundance is high in lowland forestsavanna mixes and in well-watered small savannas surrounded by forest. The latter contain capybara, caiman, turtles, and deer, and collared peccaries often use the adjacent forest edge.
9 J. Polisar et al. / Biological Conservation 109 (2003) Fig. 2. Comparative encounter rates (observations/km) among 26 linear transects. Sign (which does not directly measure group size) and visual observations independent of group size are pooled. All taxa recorded are included in the figure. This includes preferred prey, less important prey, and some taxa that would rarely be prey. BSD is semi-deciduous forest (Polisar, 2000). The juxtaposition of contrasting productive habitats, which also created productive edges, seemed key in defining desirable jaguar home ranges. In this way, the jaguar may be similar to the tiger, whose prey is most abundant where grasslands and forests form a mosaic and the interdigitation of many different vegetation types supports a rich ungulate community (Sunquist et al., 1999). The ungulate community of the llanos is hardly equal to that of southeast Asia. It is neither cervid nor bovid rich. Larger caviomorph rodents fill some niches occupied by cervids and bovids in the Old World Tropics (Eisenberg and McKay, 1974). Nonetheless, the patterns of prey production across landscapes bear similarities. The areas that had low prey abundance were large open savannas in the far south of the study area, and the high dry pastures, set in hills, that were used for calving (Fig. 2). These pastures were bordered by dry forest, a habitat more frequently occupied by puma than jaguar (Maxit, 2001). Pumas residing in the vicinity of these maternity pastures had pockets of productivity within their ranges, but the immediate vicinity of the pastures had low native prey abundance and diversity. In managing the calving season successfully by moving cattle to higher ground, ranching operations may coincidentally reduce some potential problems with jaguars while increasing the potential for problems with pumas. Prey-poor patches become rich, their wealth in calves. Jaguar preyed selectively upon collared peccaries and capybara. In puma scats, capybara remains were as frequent as deer, although the latter were 4.1 times as abundant. Capybara constituted 71% of the recorded kills of wild prey by puma. Why might capybara be a preferred prey? The profit margin must be high. Capybaras are as large as deer, and larger then peccary, yet appear to have less flight capability and weaker defenses. In Pin ero, jaguars make rounds, following a rough circuit, as they check on productive patches. The marginal value that Charnov (1976) predicted occurs because prey in patches become wary and/or flee in the presence of a predator, particularly after a herd member has fallen victim (Brown et al., 1999). Capybara possess a behavioral constraint that imposes relatively tight site fidelity. They rarely occur more than 500 m from water (Ojasti, 1973). Activity centers are rarely more than 300 m from water (Herrera and MacDonald, 1989). As a consequence, home ranges are very small. In Apure, capybara ranges measured 6 16 ha (Herrera and Mac- Donald, 1989), which would be approximately 20% of the area of collared peccary home ranges in Guarico (Castellanos, 1982). Capybara densities are exceptionally
10 306 J. Polisar et al. / Biological Conservation 109 (2003) high locally (and exceptionally low away from water). Although amphibious like the caiman, capybara spend far more time in terrestrial habitats, including forest. The patch they occupy may be slow to lose its value. Since capybara need to reconvene at water, restoration of the value of the patch they occupy may be more rapid. A cat might decide to visit more frequently, or even stay a while. Handling costs of prey procurement include the physical hazards of capture. Adult caiman, when struggling, represent a risk to dentition. Collared peccary canines approach those of a jaguar in size, and are sharper. Both peccary species rely on groups for vigilance and defense. In Hato Pin ero average jaguar weights were 1.9 times that of puma, a factor reflected in the jaguar s heavier use of large bodied risky prey. Young calves weigh kg, as much as large natural prey, but lack many of the defenses that natural prey possess (Maxit, 2001; Scognamillo, 2001). Although cows do rally to the defense of their calves, the short-term cost of a calf is likely to seem low to a predator. A maternity pasture, in which cows and calves are fenced becomes a patch whose value may never become marginal. Theory would predict no travel from such a patch. Unfortunately some pumas follow the prediction. Although the majority of cats do not make a habit of preying on such situations, some do. It pays them high profits with low costs, until their demise. In Pin ero, the frequency of cattle depredation was inversely related to availability and vulnerability of natural prey and directly related to availability and vulnerability of livestock. There was some coincidence in this. Young calves were not often pastured in the prey-rich well-watered small forest-lined savannas at low elevations. Cattle were virtually absent from some of the most prey rich areas in high-stature semi-deciduous forest due to a lack of suitable forage in those areas. Tables 3 and 4 are revealing. The biomass of the native artiodactyls (Cervidae kg/km 2 and Tayassuidae kg/km 2 ) is roughly equal to that of the introduced Perissodactyls (Equidae kg/km 2 ). As a result, the domestic mammalian biomass (introduced Bovidae and Equidae) is roughly equal to the total artiodactyl biomass (Bovidae, Cervidae, Tayassuidae), both being around 97% (Table 4). Roughly 3% of the mammalian biomass is large native prey (Table 4). The mammalian biomass of the llanos is high, approaching the richest sites of Africa and Asia, and exceeding many productive sites of the Old World Tropics (Table 5; Eisenberg, 1980; Eisenberg and Lockhart, 1972; Eisenberg and Seidensticker, 1976; Karanth and Sunquist, 1992; Mckay and Eisenberg, 1974; Peres, 1999; Schaller, 1967, 1972, 1983). This biomass is in human-facilitated ecological replacements of the grazers that went extinct in the Pleistocene. Approximately 10,000 11,000 years ago the remaining ancient South American ungulates (Toxodontia, Litopterna, and Glyptodonts) and roughly half of the Pliocene s northern immigrants (Proscidea, Perissiodactyla, and Artiodactla) went extinct (Cartelle, 1999; MacFadden and Shockey, 1997; Martin, 1967; Webb, 1978). Bovidae never occurred in South America. Although capybara are grazers, they are small relative to the recent mega-grazers. At present, there are 21 species of ungulates in tropical America (Ojasti, 1983). Most are at least partially dependent upon forest. In proportion to continent areas, by African standards, there would be 55 ungulates in South America (Ojasti, 1983). The savannas of South America were, in some respects, empty when the Spanish arrived, carrying Old World grazers. Feral on the landscape, the bovids and equids multiplied. Managed by humans (immunizations, predator control, forage improvements), their biomass climbed even higher (Crosby, 1986). Before the extinction, the New World tropics had large herbivores, and an associated assemblage of large predators. In the Pleistocene, there were lions (Panthera atrox), jaguars (Panthera onca), pumas (Puma concolor) and sabertooth cats (Smilodon fatalis) in Florida (Morgan and Seymour, 1997). A community of a large sabertooth cat (Homotherium serum), a smaller lighter sabertooth cat (Smilodon gracilis) and a cheetah-like cat (Miracinoyx inexpectatus) approximately mirrored the size class distribution of the present community of lion, leopard, and cheetah in Africa (Morgan and Seymour, 1997). The communities of large mammals in South America are recent. Individual members may be ancient, some more then others, but the communities are very recent. The native herbivore biomass in the llanos is miniscule compared to Africa. With large grazers reintroduced, the biomass surpasses famous grazing grounds in Africa (Table 5). The vast majority of that biomass is in cattle. It is surprising that the problem of livestock depredation by jaguar and puma is not even more severe. The perspective that fossils provide does little to assuage the concerns of ranchers. At present, some puma and fewer jaguars make decisions based on sound energetics that fail to factor in lethal rancher s responses. Cattle ranchers lose when cats take livestock. Ultimately, the cats lose too. Humans introduced bovids into the receptive environment of the New World tropics. In doing so, we created this dilemma, and it thus rests upon us to help the cats make wise decisions Recommendations for ranchers Observations made by ranchers and researchers working in Venezuela and Brazil have resulted in suggestions for promoting the coexistence of cats and cattle (Hoogesteijn et al., 1993, in press; Hoogesteijn and
11 J. Polisar et al. / Biological Conservation 109 (2003) Mondolfi, 1993; Quigley and Crawshaw, 1992; Scognamillo et al., in press). These include: 1. Protect all principal prey of the large cats by preventing poaching; 2. Avoid commercial harvests of capybara and caiman. If harvests are conducted, exert strict control, particularly with capybara; 3. When feasible, impede the ability of cattle to enter forest; 4. Concentrate calving seasons via artificial insemination. A shorter calving season facilitates control; 5. When possible, locate maternity pastures at a distance from cover that cats may prefer; 6. Explore the application of electric fence around maternity pastures, as developed by Scognamillo et al. (in press); 7. If practical, move calves from pastures with chronic depredation problems and replace with older animals, over 1 2 years of age; 8. Move all cattle out of lowland flooding areas before waters rise to avoid isolation and crowding in forest islands amidst flooded savannas; 9. Where possible, stock low flooding savannas with water buffalo (less vulnerable); 10. Keep good clear records of losses from all causes to facilitate planning and decision making; and 11. Do not clear all forests. In our study area, large inter-connected cattle ranches owned by ranchers that appreciated wildlife values allowed large mammals such as tapir, white-lipped peccary, and jaguar to persist in an agricultural mosaic. These ranches constituted partially protected areas, enforced more effectively than some national parks. The natural prey base in Hato Pin ero was judged to be adequate. Cats were not forced to rely on domestic livestock. Outside of these protected ranches prey can be presumed to be less abundant. Where Pin ero lay adjacent to unprotected land, capybaras were less abundant than habitats could support. Wary behavior and nocturnal habits suggested intense poaching pressure. Within Pin ero, wildlife protection was compromised where a small river that ran through the property (by law a public thoroughfare). Capybara were nearly eradicated and caiman far fewer there than in better areas. Camps and cleaned carcasses were evidence of poaching activity. Pin ero s cats had the choice to not take more domestic prey because (in most areas) vigilance had maintained the natural prey. In Pin ero, progressive cattle management had already reduced the potential severity of depredation. (Quigley and Crawshaw, 1992) described how in lightly managed ranches of the Brazilian Pantanal cattle were left in low-lying areas during the wet season. Cattle and large cats were confined to slightly elevated forest islands where the cattle fell prey to jaguar. Pin ero s seasonal cattle drives between lowlands and uplands were geared towards increasing calf survivorship and maximizing forage opportunities. Coincidentally they probably reduced the frequency of jaguar attacks, but may have elevated the likelihood of attacks on calves by puma in upland pastures. From an economic perspective, calves lost to pumas in uplands must have been less than would have been lost by leaving herds in flooded savannas. The persistence of cattle losses, albeit at a relatively low level, in this well-managed setting is testimony that depredation can be reduced but probably never fully eliminated. In the absence of regular herd inspections and written records, calf losses due to injuries, diseases, poor nutrition, or poisonous bites may be mistakenly attributed to cats. Good records clarify the actual proportion of losses that are due to cats, reducing the tendency to make the cats scapegoats for poor management (Hoogesteijn et al., 1993). The recommendation that forests not be cleared may be equivocal. Unrestrained forest clearing might alleviate conflicts between ranchers and wildlife altogether, albeit eliminating all wildlife values. Our findings imply that puma depredation may be less limited by proximity to forest than jaguar depredation. It is clear from this study that forests are a reservoir of prey. Some forests are more productive than others (Fig. 2). In Pin ero, the highest depredation rates occurred in areas where prey abundance and diversity was relatively low. If a ranch has cats and wishes to accommodate them, the more prey-rich forest it has, the less a temptation livestock will be. Any measures that keep cattle from the forest interior are likely to reduce attacks by jaguar. One of the few documented cases of jaguar attacks on cattle during our study took place when a young cow strayed into dense riparian forest. Motivated ranchers might consider locating some excavated water retention ponds specifically for prey. This would elevate prey numbers, direct dry-season spatial distribution of prey, and thus potentially focus cats activities in space. Savannas boasting abundant ponds and bordered by forest were one of Pin ero s richer habitat types in terms of prey (Fig. 2). In these habitats, capybara and caiman frequented the water bodies, deer traversed the ecotone, visiting the ponds to drink, the adjacent forests supported peccary, and attacks on cattle were scarce. Acknowledgements Funding was provided by The National Geographic Society, the Wildlife Conservation Society, the British
12 308 J. Polisar et al. / Biological Conservation 109 (2003) Embassy to Venezuela s Cooperation Fund, the Cat Specialist Group of the Species Survival Commission (IUCN-World Conservation Union), and the Lincoln Park Zoo Scott Neotropic Fund, and Katharine B. Ordway Chair endowment funds via Dr. J.F. Eisenberg. Faunal aspects of the study were assisted in the field by Victor Juan Meires, Orlando Ramirez, Gilson Rivas Fuenmayor, Diego Giraldo, Sandra Melman, Emiliana Isase, Telva Carantona, and Marcus Trepte. Rafael Ortiz and Dr. Francisco Delascio worked on floral aspects. Rafael Hoogesteijn, Roy McBride, Rocky McBride, Tibisay Escalona, Bruno Pampour, the llaneros of Pin ero, and students and staff of Dr. Juhani Ojasti s wildlife management class (UNELLEZ at Guanare) all contributed greatly. Francisco Bisbal performed some scat analyses. His staff provided logistical support. Edgardo Mondolfi served as liaison. Don Antonio Julio Branger provided an excellent study area, and facilitated. Matt Burgess, Kyle Harris, Stephen Taranto, and Richard Owens condensed data. Maria Fernanda Zermoglia, and Rosanna Rivero refined maps. F. Wayne King, Jim Nichols, and Fred Thompson reviewed the dissertation from which this paper originated (Polisar, 2000). References Ackerman, B.B., Lindzey, F.G., Hemker, T.P., Predictive energetics model for cougars. In: Miller, S.D., Everett, D.D. (Eds.), Cats of the World: Biology, Conservation, and Management. The National Wildlife Federation, Washington, D.C, pp Anderson, A.E., 1983, A critical review of literature on puma. Colorado Division of Wildlife Special Report 54. Anon., ArcView. Environmental Systems Research Institute, Redlands, CA. Ayarzagüena, J.S., Ecologia del caiman de anteojos o baba (Caiman crocodilus) en los llanos de Apure (Venezuela). Don ana 10, Bodmer, R., Aquino, R., Puertas, P., Reyes, C., Fang, T., Gottdenker, N., Manejo y uso sustentable de pecaríes en la Amazonı a Peruana. Unio n Internacional para le Conservación de la Naturaleza y los Recursos Naturales, Quito. Brokx, P.A.J., A Study of the Biology of Venezuelan Whitetailed Deer (Odocoileus virginianus gymnotis Wiegmann, 1833), with a Hypothesis on the Origin of the South American Cervids. PhD dissertation, University of Waterloo, Waterloo. Brown, J.S., Laundre, J.W., Gurung, M., The ecology of fear: optimal foraging, game theory, and trophic interactions. Journal of Mammalogy 80, Buckland, S.T., Anderson, D.R., Burnham, K.P., Laake, J.L., Distance Sampling: Estimating Abundance of Biological Populations. Chapman and Hall, London. Cartelle, C., Pleistocene mammals of the Cerrado and Caatinga of Brazil. In: Eisenberg, J.F., Redford, K.H. (Eds.), Mammals of the Neotropics: The Central Neotropics. The University of Chicago Press, Chicago, pp Castellanos, A.H.G., Patrones de movimiento y uso de habitat del baquiro de collar Tayassu tajacu en los llanos centrales de Venezuela. Tesis de Licenciatura, Universidad Central de Venezuela. Charnov, E.L., Optimal foraging, the marginal value theorem. Theoretical Population Biology 9, Colwell, R.K., Futumaya, D.J., On the measurement of niche breadth and overlap. Ecology 52, Corrales, F., Libros de Ganaderia de Hato Pinero. Agropecuaria San Franciso, Valencia, Venezuela. Crosby, A.W., Ecological imperialism: the biological expansion of Europe, Cambridge University Press, Cambridge. Culver, M., Johnson, W.E., Pecon-Slattery, J., O Brien, S.J., Genomic ancestry of the American puma (Puma concolor). Journal of Heredity 91, Curio, E., Ethology of Predation. Springer-Verlag, Berlin. Dailey, T.V., Thompson Hobbs, N., Woodard, T.N., Experimental comparisons of diet selection by mountain goats and mountain sheep in Colorado. Journal of Wildlife Management 48, Dimmick, R.W., Pelton, M.R., Criteria of sex and age. In: Bookhout, T.A. (Ed.), Research and Management Techniques for Wildlife and Habitats. The Wildlife Society, Bethesda, pp Eisenberg, J.F., The density and biomass of tropical mammals. In: Soulé, M.A., Wilcox, B.A. (Eds.), Conservation Biology: an Evolutionary-ecological Perspective. Sinauer Associates, Sunderland, pp Eisenberg, J.F., Lockhart, M., An ecological reconnaissance of Wilpattu National Park, Ceylon. Smithsonian Contributions to Zoology 101, Eisenberg, J.F., McKay, G.M., Comparison of ungulate adaptations in the New World and Old World tropical forests with special reference to Ceylon and the rainforests of Central America. In: Geist, V., Walther, F. (Eds.), The Behavior of Ungulates and its Relation to Management. IUCN, Morges, Switzerland, pp Eisenberg, J.F., O Connell, M.A., August, P.V., Density, productivity, and distribution of mammals in two Venezuelan habitats. In: Eisenberg, J.F. (Ed.), Vertebrate Ecology in the Northern Neotropics. Smithsonian Institution Press, Washington, DC, pp Eisenberg, J.F., Polisar, J.R., The mammal species of northcentral Venezuela. Bulletin of the Florida Museum of Natural History 42, Eisenberg, J.F., Seidensticker, J., Ungulates in Southern Asia: a consideration of biomass estimates for selected habitats. Biological Conservation 10, Emlen, J.M., The role of time and energy in food preference. The American Naturalist 100, Emmons, L.H., Comparative feeding ecology of felids in a neotropical rain forest. Behavioral Ecology and Sociobiology 20, Emmons, L.H., Jaguar predation on Chelonians. Journal of Herpetology 23, Farrell, L., The Ecology of the Puma and the Jaguar in the Venezuelan Llanos. MS thesis University of Florida, Gainesville. Farrell, L., Roman, J., Sunquist, M.E., Dietary separation of sympatric carnivores identified by molecular analysis of scats. Molecular Ecology 9, Hayne, D.H., Population dynamics and analysis. In: Halls, L.K. (Ed.), White-tailed Deer: Ecology and Management. Stackpole Books, Harrisburg, pp Hellgren, E.C., Synatzske, D.R., Oldenburg, P.W., Guthery, F.S., Demography of a collared peccary population in south Texas. Journal of Wildlife Management 59, Herrera, E.A., MacDonald, D.W., Resource utilization and territoriality in group-living capybara (Hydrochoerus hydrochaeris). Journal of Animal Ecology 58, Hoogesteijn, R., Boede, E.O., Mondolfi, E., in. Observaciones sobre la depredacion de jaguares sobre bovinos en Venezuela y los programas de control gubernamentales. In: Medellín, R.A., Chetkiewicz, C., Rabinowitz, A., Redford, K.H., Robinson, J.G., Sanderson, E., Taber, A. (Eds.), Los Jaguares en el Nuevo Milenio.
13 J. Polisar et al. / Biological Conservation 109 (2003) Un Análisis de su Estado, Deteccio n de Prioridades y Recomendaciones Para la Conservacio n del Jaguar en América. Instituto de Ecología, Universidad Nacional Autonoma de México/Wildlife Conservation Society, Me xico, D.F. Hoogesteijn, R., Hoogesteijn, A., Mondolfi, X., Jaguar predation and conservation: cattle mortality caused by felines on three ranches in the Venezuelan Llanos. In: Dunstone, N., Gorman, R.L. (Eds.), Mammals as Predators. Zoological Society, London, pp Hoogesteijn, R., Mondolfi, E., The Jaguar. Armitano Publishers, Caracas. Hornocker, M.G., An analysis of mountain lion predation upon mule deer and elk in the Idaho primitive area. Wildlife Monographs 21, Karanth, K.U., Sunquist, M., Behavioral correlates of predation by tiger (Panthera tigris), leopard (Panthera pardus) and dhole (Cuon alpinus) in Nagarahole, India. Journal of Zoology London 250, Karanth, K.U., Sunquist, M.E., Population structure, density and biomass of large herbivores in the tropical forests of Nagarahole, India. Journal of Tropical Ecology 8, Karanth, K.U., Sunquist, M.E., Prey selection by tiger, leopard, and dhole in tropical forests. Journal of Animal Ecology 64, Kleiman, D.G., Eisenberg, J.F., Maliniak, E., Reproductive parameters and productivity of caviomorph rodents. In: Eisenberg, J.F. (Ed.), Vertebrate Ecology in the Northern Neotropics. Smithsonian Institution Press, Washington, DC, pp Lancia, R.A., Nichols, J.D., Pollock, K.H., Estimating the number of animals in wildlife populations. In: Bookhout, T.A. (Ed.), Research and Management Techniques for Wildlife and Habitats. The Wildlife Society, Bethesda, pp Linares, O., Mamı feros de Venezuela. Sociedad Conservacionista Audubon de Venezuela, Caracas. Lord, R.D., Lord, V.R., Cross checking censuses and a model of the annual cycle of mortality and reproduction in Capybaras (Hydrochaeris hydrochaeris). Studies on Neotropical Fauna and the Environment 23, MacArthur, R.H., Pianka, E.R., On optimal use of a patchy environment. The American Naturalist 100, MacFadden, B.J., Shockey, B.J., Ancient feeding ecology and niche differentiation of Pleistocene mammalian herbivores from Tarija, Bolivia: morphological and isotopic evidence. Paleobiology 23, Martin, P.S., Pleistocene overkill. In: Martin, P.S., Wright, H.E. (Eds.), Pleistocene Extinctions: The Search for a Cause. Yale University Press, New Haven, pp Maxit, I.E Prey Use by Sympatric Puma and Jaguar in the Venezuelan Llanos. MS thesis, University of Florida, Gainesville. Mckay, G.M., Eisenberg, J.F., Movement patterns and habitat utilization of ungulates in Ceylon. In: Geist, V., Walther, F. (Eds.), The Behavior of Ungulates and its Relation to Management. IUCN Publication, Morges, Switzerland, pp Miller, L.E., Socioecology of the Wedge-capped Capuchin Monkey (Cebus olivaceous). PhD dissertation, University of California, Davis. Mondolfi, E., Hoogesteijn, R., Notes on the biology and status of the jaguar in Venezuela. In: Miller, S.D., Everett, D.D. (Eds.), Cats of the World: Biology, Conservation, and Management. National Wildlife Federation, Washington, D.C, pp Morgan, G.S., Seymour, K.L., Fossil history of the panther (Puma concolor) and the cheetah-like cat (Miracinoyx inexpectatus). Bulletin of the Florida Museum of Natural History 40, Nun ez, R., Miller, B., Lindzey, F., Food habits of jaguars and pumas in Jalisco, Mexico. Journal of Zoology 252, Ojasti, J., Estudio bio logico del chigu ire o capibara. Fondo Nacional de Investigaciones Agropecuarias, Caracas. Ojasti, J., Ungulates and large rodents of South America. In: Bourlière, F. (Ed.), Ecosystems of the World. Vol. 3. Tropical Savannas. Elsevier Scientific Publishing Company, Amsterdam, pp Ojeda, M.M., Keith, L.M., Sex and age composition and breeding biology of cottontail rabbit populations in Venezuela. Biotropica 14, Peres, C.A., The structure of nonvolant mammal communities in different Amazonian forest types. In: Eisenberg, J.F., Redford, K.H. (Eds.), Mammals of the Neotropics: The Central Neotropics. The University of Chicago Press, Chicago, pp Polisar, J., Jaguars, Pumas, their Prey Base, and Cattle Ranching; Ecological Perspectives of a Management Issue. PhD dissertation, University of Florida, Gainesville. Quigley, H., Crawshaw, P., A conservation plan for the jaguar Panthera onca in the Pantanal region of Brazil. Biological Conservation 61, Rabinowitz, A.R., Nottingham Jr., B.G., Ecology and behavior of the jaguar (Panthera onca) in Belize, Central America. Journal of Zoology 210, Robinson, J.G., Bodmer, R.E., Towards wildlife management in tropical forests. Journal of Wildlife Management 63, Robinson, J.G., Redford, K.H., Intrinsic rate of natural increase in Neotropical forest mammals: relationship to phylogeny and diet. Oecologia 68, Robinson, J.G., Redford, K.H., Sustainable harvest of Neotropical forest mammals. In: Robinson, J.G., Redford, K.H. (Eds.), Neotropical Wildlife Use and Conservation. The University of Chicago Press, Chicago, pp Roosevelt, T., Through the Brazilian Wilderness. Charles Scribner s Presses, New York. Ross, P.I., Jalkotzy, M.G., Festa-Bianchet, M., Cougar predation on bighorn sheep in southwestern Alberta during winter. Canadian Journal of Zoology 74, Schaller, G.B., The Deer and the Tiger. The University of Chicago Press, Chicago. Schaller, G.B., The Serengeti Lion: a Study of Predator prey Relations. The University of Chicago Press, Chicago. Schaller, G.B., Mammals and their biomass on a Brazilian ranch. Arquivos de Zoologia 31, Schaller, G.B., Crawshaw, P., Movement patterns of jaguar. Biotropica 12, Scognamillo, D., Maxit, I., Sunquist, M., Farrell, L., in. Ecologia del jaguar y el problema de la depredacion sobre ganado en Hato Pin ero, Venezuela. In: Medellín, R.A., Chetkiewicz, C., Rabinowitz, A., Redford, K.H., Robinson, J.G., Sanderson, E., Taber, A. (Eds.), Los jaguares en el nuevo milenio. Un análisis de su estado, detección de prioridades y recomendaciones para la conservacio n del jaguar en América. pressinstituto de Ecología, Universidad Nacional Autonoma de México/Wildlife Conservation Society, México, D.F. Scognamillo, D.G., Ecological Separation Between Jaguar and Puma in a Mosaic Landscape in the Venezuelan Llanos. MS thesis, University of Florida, Gainesville. Seber, G.A.F., The Estimation of Animal Abundance and Related Parameters. Charles Griffin and Company, London. Seidensticker, J., On the ecological separation between tigers and leopards. Biotropica 8, Shaw, H., Impact of mountain lion on mule deer and cattle in northwestern Arizona. Pages. In: Phillips, R.L., Jonkel, C. (Eds.), Proceedings of the 1975 Predator Symposium. Montana Forest and Conservation Experimental Station, Missoula, pp Smythe, N., The natural history of the Central American agouti (Dasyprocta punctata). Smithsonian Contributions to Zoology 257, Sowls, L.K., Javelina and Other Peccaries: Their Biology, Management, and Use. Texas A & M University Press, College Station.
14 310 J. Polisar et al. / Biological Conservation 109 (2003) Spowart, R.A., Thompson Hobbs, N., Effect of fire on diet overlap between mule deer and mountain sheep. Journal of Wildlife Management 49, SPSS, SPSS Base 9.0 User s guide. SPSS Inc, Chicago. Sunquist, M.E., The social organization of tigers (Panthera tigris) in Royal Chitawan National Park, Nepal. Smithsonian Contributions to Zoology 336, Sunquist, M.E., in. History of jaguar research in the Americas. In: Medellín, R.A., Chetkiewicz, C., Rabinowitz, A., Redford, K.H., Robinson, J.G., Sanderson, E., Taber, A. (Eds.), Los Jaguares en el Nuevo Milenio. Un Análisis de su Estado, Deteccio n de Prioridades y Recomendaciones Para la Conservacio n del Jaguar en América. Instituto de Ecología, Universidad Nacional Autonoma de México/ Wildlife Conservation Society, México, D.F. Sunquist, M.E., Karanth, K.U., Sunquist, F.C., Ecology, behavior and resilience of the tiger and its conservation needs. In: Seidensticker, J., Christie, S., Jackson, P. (Eds.), Riding the Tiger: Tiger Conservation in Human-dominated Landscapes. Cambridge University Press, Cambridge, pp Sunquist, M.E., Sunquist, F.C., Ecological constraints on predation by large felids. In: Gittleman, J.L. (Ed.), Carnivore Behavior, Ecology, and Evolution. Cornell University Press, Ithaca, pp Taber, A.B., Novaro, A.J., Neris, N., Colman, F.H., The food habits of sympatric jaguar and puma in the Paraguayan Chaco. Biotropica 29, Taylor, R.J., Value of clumping to prey and the evolutionary response of ambush predators. American Naturalist 110, Teer, J.G., Lessons from the Llanos basin, Texas. In: Halls, L.K. (Ed.), White-tailed Deer: Ecology and Management. Stackpole Books, Harrisburg, pp Temple, S.A., Do predators always capture substandard individuals disproportionately from prey populations? Ecology 68, Webb, S.D., A history of savanna vertebrates in the New World. Part II: South America and the great interchange. Annual Review of Ecology and Systematics 9, Winston, P.M., Odocoileus virginianus. Mammalian Species 388, 1 13.