Carnivore Conservation

Contributed Oral Presentations


Contributed paper sessions will be available on-demand for the duration of the conference, then again at the conclusion of the conference.


Community Responses of African Carnivores to Prescribed Burning
Laura C. Gigliotti; Goncalo Curveira-Santos; Rob Slotow; Craig Sholto-Douglas; David S. Jachowski
Fires are common in many ecosystems worldwide and are also widely-used for management. Although the responses of herbivores to fire have been well-studied, less research has addressed how carnivores respond to fire. In particular, little is known about how fires affect multiple carnivore species simultaneously, and how changes in prey abundance as a result of fire have the potential to influence carnivore coexistence or suppression. We studied community-level responses to fire in the Mun-Ya-Wana Conservancy in South Africa. We used prescribed burning to experimentally increase prey densities and monitored carnivore intensity of use using camera traps with a Before-After-Control-Impact study design. We analyzed the camera trap data using community N-mixture models to understand how individual species, as well as large carnivores and small carnivores as a whole, change intensity of use in response to burning. We found that although the intensity of use of several important prey species increased following burns, the responses of carnivores to prescribed burning were variable. Lions (Panthera leo), honey badgers (Mellivora capensis), and side-striped jackals (Canis adustus) increased intensity of use in burned area following the fires, whereas intensity of use for all other large and small carnivores was not affected by burning. Our results indicate that fire does not promote carnivore coexistence by creating conditions for all carnivores to increase use of burned areas, but that it also likely does not result in numerical suppression of subordinate predators. Instead, fires might cause a suppression of opportunities for subordinate predators because they need to avoid lions, rather than take advantage of increased hunting opportunities in recently burned areas. In systems with a high diversity of carnivore species, complexity in species-specific responses to fire is to be expected, and understanding this complexity is critical for planning prescribed burns.
Density and Activity Patterns of Pallas’s Cats in Central Mongolia
Stefano Anile; Claudio Augugliaro; Clay K. Nielsen
Although the Pallas’s cat is listed as Near Threatened, few robust estimates of population density or activity patterns exist for the species. Furthermore, little research of Pallas’s cats has been conducted on unprotected lands, which encompass most of the species’ distributional range. We estimated density and activity patterns of Pallas’s cats on unprotected lands in central Mongolia during two periods (May – August and September -November) in 2019. We obtained 484 Pallas’s cat images from 153 detections during 4266 camera-days. We identified Pallas’s cats using pelage markings and identified 16 individuals from 64 detections. We used spatially explicit capture-recapture models to estimate population density at 15.2 ± 4.8 individuals/100 km2. Pallas’s cat activity was consistent between the two survey periods (~0.50), with cats mainly active during crepuscular hours in the first period and strictly diurnal in the second. We provide the first estimation of a Pallas’s cat population density using camera-trapping. Compared to other methods used, densities were relatively high on our study area, likely due to a combination of highly-suitable habitat and abundant prey. Seasonal shifts in activity patterns by Pallas’s cats likely indicated an adaptive response to reduced risk of depredation by raptors. We recommend August to November as the best time for camera-trapping surveys for Pallas’s cats given their high daily activity and easiest interpretation of images used for individual identification collected during this time. We also suggest future camera trapping surveys for Pallas’s be mindful of potential camera trap avoidance through time.
Fine-Scale Movement Decisions by a Large Carnivore Inform Conservation Planning in Human-Dominated Landscapes
Justin P. Suraci; Barry A. Nickel; Christopher C. Wilmers
Developed landscapes are increasingly important movement habitat for many large carnivore populations, despite fragmentation and heightened anthropogenic risks. The availability of vegetation cover, which provides concealment from humans and promotes hunting success for some species, is a key factor mediating large carnivore use of human-dominated landscapes. Restoring or modifying networks of vegetation patches may therefore provide an important tool for enhancing the connectivity value of developed areas but requires understanding how vegetation patch networks are functionally linked by carnivore movement decisions, which occur at scales considerably finer than those typically addressed by connectivity analyses. Here we use high-resolution data on vegetation cover and five-minute GPS locations of pumas (Puma concolor) to investigate the factors driving puma movement decisions when navigating heavily fragmented landscapes and apply these results to enhancing puma connectivity through human-dominated landscapes. Puma movements between vegetation patches were driven by patch size, vegetation type, and spatial arrangement relative to sources of anthropogenic risk (buildings). Pumas avoided buildings but accepted higher building densities as patch area increased or inter-patch travel distances decreased. Pumas also selected for shrub over tree cover and patches with relatively little neighboring vegetation, indicating preferences for high concealment and for edge habitats known to increase hunting success. We applied these inferences to the comparison of multiple habitat restoration scenarios for increasing connectivity across a degraded wildlife corridor. Network-based connectivity modeling revealed that the strategic placement of vegetation patches can substantially reduce resistance to puma movement across an otherwise high resistance landscape by diversifying movement options, particularly when paired with road crossing enhancements. Our results reveal the factors mediating large carnivore use of human-dominated landscapes as movement habitat and provide a highly generalizable tool for increasing movement potential through the manipulation of vegetation cover.
Dracula’s Menagerie: A Multi-Species Occupancy Analysis of Lynx, Wildcats and Wolves in Transylvania
Marissa A. Dyck; Ruben Iosif; Barbara Promberger; Viorel D. Popescu
Loss of biodiversity is a huge concern in the current Anthropocene epoch. Therefore, it is essential to put forth effective and efficient conservation strategies. In order to do this, we need to better understand the species and systems we aim to protect. Terrestrial carnivores, in particular, are some of the most imperiled species today, due to their large home range requirements, high metabolic demands, and sensitivity to habitat fragmentation. Changes in carnivore populations can cause widespread ecosystem effects through top-down regulation. We assessed the distribution and interactions of three carnivore species in the Romanian Carpathians. Romania houses one of the last fully intact carnivore guilds in Europe, making it an ideal system to assess innate interactions between carnivore species. We used data from 64 camera traps distributed throughout Transylvanian forests to assess occupancy and co-occurrence of Eurasian lynx (Lynx lynx), European wildcat (Felis silvestris) and gray wolf (Canis lupus). We modeled marginal and conditional occupancy as a function of environmental and anthropogenic covariates to explore intraguild interactions between species and how their interactions and distributions may be altered in response to anthropogenic change. Both lynx and wolf occupancy were highly influenced by road density, while wildcat occupancy was most contingent upon altitude. Based on our assessment of conditional occupancy, we found little evidence of exploitative competition among carnivores. Understanding the distribution of species relative to environmental and anthropogenic variables can provide valuable insight for management decisions in Romania and areas with analogous species aiming to promote coexistence between humans and wildlife.
Challenging the Assumed Superiority of Camera- Versus Capture-Based Surveys for Assessing Occupancy: A Case Study with a Cryptic Forest Mustelid
Kirstin E. Fagan; Daniel J. Harrison; Erin M. Simons-Legaard; Tyler F. Woollard
Occupancy analysis using camera traps has become a popular alternative to live-trapping for cryptic forest carnivores because of lower costs and the ability to more easily account for imperfect detection. Live-trapping, however, allows collection of critical information on demographics and spatial use post-release that can enhance inferences about the effects of habitat features on occupancy. We compare approaches using grid-based camera-trapping simultaneous with systematic live-trapping to estimate patterns of occupancy by American martens (Martes americana). We demonstrate that live-trapping may be as effective as camera-trapping for detection and estimation of asymptotic proportion of area occupied (PAO) by martens. We used a non-spatial occupancy framework to estimate method-specific detection probabilities (p) and compared model estimates of PAO, excluding subsequent telemetry data. Grid cells (n = 62) were the average size of female home ranges in the study area (250 ha). Detection histories generated using a single station (method1 = one camera-trap, method2 = one live-trap) resulted in models that overestimated p, particularly for camera-traps, and underestimated occupancy (ψ). Including data from all stations (method1 = one camera-trap, method2 = average of 5 live-traps) resulted in less biased estimates. Detection probability was greater for live-trapping than cameras when cells contained >2 live-traps and equivalent when cells contained 2 traps. Cumulative probability of detection was equivalent for both methods. In post-hoc comparisons of all available detection data, the combination of live-trapping and telemetry data had a lower false-negative rate (0.02) than live-traps (0.38) and cameras (0.29) alone. Our results suggest live-trapping protocols informed by the life history characteristics of the target species should not be assumed to be plagued by imperfect detections. Further, the process of capturing individuals not only permits the collection of demographic information, but also subsequent spatial information that may reduce Type II error with estimating species occupancy.
Genetic Structure of a Reintroduced Fisher Population in West Virginia
Caroline Harms; Rich Rogers; John Edwards; Chris Ryan; Amy Welsh
Fisher (Pekania [Martes] pennanti) is a midsized furbearer species reintroduced to the state of West Virginia in 1969. Objectives of this study were to assess the genetic structure of the WV population(s) and compare structure to surrounding state populations as well as to the source population in New Hampshire. Sample collection involved contacting WV trappers during the trapping seasons for donations as well as state wildlife biologists for archival samples. Genetic analysis methods included modified DNA extraction and PCR amplification of 10 microsatellite loci. Current results are for samples originating from West Virginia, Virginia, New Hampshire and New York. Our results from assignment clustering software determined the most likely number of populations was two and grouped West Virginia/Virginia (WV/VA) as a population (n = 28) and New Hampshire/New York (NH/NY) as a population (n = 27). Both populations were considered significantly (p < 0.0001) out of Hardy-Weinberg Equilibrium. Effective population size analyses found NH/NY had Ne = 89.4 and WV/VA had Ne = 7.5. In conclusion, WV currently does not group with individuals from its source population (NH) of 50 years ago, likely due to founder effects of small reintroduction size (n = 23). Genetic diversity (He) is higher for the NH/NY population, likely due to higher gene flow in the northeastern range of fisher. Fisher have been extirpated from VA and these chance VA fisher samples may be representative of migrant West Virginia fisher released in 1969. These results may encourage a modern translocation of diverse individuals to the state.
Corridor Connectivity Drives Tiger Meta-Population Dynamics in the Terai-Arc Landscape, India
Suvankar Biswas; Supriya Bhatt; Debanjan Sarkar; Bivash Pandav; Samrat Mondol
Terai-Arc landscape (TAL) holds 22% of the Indian tiger (Panthera tigris tigris) population across 15000 km2 of potential tiger habitat along the foot hills of Himalayas spanning over 13 protected and non-protected areas. This landscape has 13 previously identified tiger dispersal corridors and is currently under significant threat of habitat fragmentation. To assess the functionality of these corridors, field survey, non-invasive genetic sampling and remote sensing GIS tools were used. 1608 large carnivore feces were collected and tiger samples were identified to understand their distribution and habitat use in TAL. Unique individuals were identified using 13 microsatellite loci to assess population structure. Hence, direction and magnitude of gene flow were estimated as a measurement of functional connectivity. Additionally, corridor connectivity map was created using tiger occurrence points and other habitats variables. 219 unique tiger individuals were identified which is 34% of the tiger population in TAL. It was found that currently, western, central and eastern parts of TAL form three genetic sub-populations and four out of 13 corridors are non-functional. On other hand, six new functional corridors were identified for tiger dispersal giving rise to 15 existing functional corridors. But corridor connectivity analysis confirmed 12 corridors as critical, where any further interruptions could restrict future tiger movement and hinder the source-sink dynamics.
This study confirms majority of tiger corridors are functional. However, presence of these critical corridors outside the PAs make them more vulnerable to further degradation. Therefore, long term survival of tiger in this fragmented landscape would only sustain by protecting the source populations and practicing effective management strategies.
Precautionary Approaches to Set Baselines for Future Management of Coyotes in An Urban County
Rachael E. Urbanek; Rebecca J. Buteau; Christopher Dumas
As coyotes (Canis latrans) expand their range into urban areas, monitoring human-coyote conflicts and discussing future management options can avoid reactionary decisions that often accompany public conflict. We surveyed 4,000 taxpayers from New Hanover County, North Carolina, to assess human-coyote interactions, public attitudes, conflict (PCI2), and willingness to pay regarding 3 coyote management methods: no management (NM); public education (PE); and trap with euthanasia (TE). Most (62%) respondents believed the coyote population was increasing yet >58% of respondents had not seen or heard a coyote recently. Forty-five percent of respondents had interacted with coyotes and 67% of those interactions were with coyotes behaving naturally. PE was the most acceptable management method countywide (P < 0.0001). NM was acceptable for female respondents, younger respondents, members of animal rights groups, and those who considered coyotes to be native, whereas TE was acceptable for respondents from opposite demographic segments (P < 0.0001). For all demographic segments, conflict was low for PE (PCI2 = 0.13-0.25) but high for NM (PCI2 = 0.34-0.57) and TE (PCI2 = 0.33-0.48). Countywide, attitudes toward each management method differed (P < 0.002) and were influenced by perceptions of whether a coyote would experience an unnatural or inhumane death; family, personal, and pet safety; public ability to participate in management; and whether the coyote population would decrease. When coyotes are behaving naturally, resident support decreased for PE and TE as potential management costs increased while support for NM increased. However, when coyotes were aggressive toward humans, support for NM was sparse and support for PE and TE were high, regardless of cost. Seeking public opinion early will likely benefit natural resource managers by increasing public support for decisions, reducing conflict, and facilitating a more proactive approach to coyote management.
The First Range-Wide Meta-Analysis of Coyote Diet
Alex J. Jensen; David S. Jachowski
In the last 100 years the geographic distribution of coyotes (Canis latrans) has expanded dramatically to include nearly all of North America. As an adaptable carnivore, coyotes eat a wide variety of foods, including those of societal interest like white-tailed deer (Odocoileus virginianus), protected species, and livestock. Despite much research on coyote diet at the population level, little is known about what drives diet across their range. Our aim was to conduct the first range-wide meta-analysis of coyote diet to better understand how factors such as coyote size, presence of wolves, environmental productivity, and human population density influence diet. Based on data from 148 study areas in 100 publications, we used multivariate linear modeling to assess trends in occurrence of nine food groups (cervids, feral hogs (Sus scrofa), lagomorphs, small mammals, livestock, birds, arthropods, vegetation, and anthropogenic items) as well as dietary diversity. Overall, small mammals occurred most often (41%), followed by vegetation (34%), cervids (21%), rabbits (18%), arthropods (14%), birds (9%), livestock (4%), anthropogenic items (2%), and feral hogs (1%). Our statistical models suggest that larger coyotes ate more ungulates, less small mammals, and had less diverse diets than smaller coyotes. Where wolves were present, coyotes ate more ungulates and less arthropods, vegetation, and small mammals, though few studies differentiated ungulate predation from scavenging. Environmental productivity was positively related to cervids and vegetation in diets, while human population density was negatively related to arthropod use, and positively related to rabbit use. As a relatively recently established carnivore across much of eastern North America, our results provide important insight into differences in ecological niche use and the associated impact coyotes are likely having across their range. Further, as coyotes appear poised to begin colonizing South America, research which can be used to predict their effects in new areas is imperative.
Seasonal and Daily Shifts in Behavior and Resource Selection: How a Carnivore Navigates Costly Landscapes
Hance Ellington; Erich Muntz; Stanley Gehrt
The dynamic environmental conditions in extreme systems likely have a strong influence on how species use the landscape. Animals must respond to risks and benefits that shift daily and seasonally across challenging landscapes. Animals must balance these dynamic changes to landscape risk with the underlying resources provided by that landscape. One way to balance the seasonal and daily changes in the costs and benefits of a landscape is through behaviorally-explicit resource selection and temporal partitioning. Here, we test whether resource selection of coyotes (Canis latrans) in Cape Breton Highlands National Park, Nova Scotia, Canada is behaviorally-explicit and whether they modify selection in a manner that reduces cost and maximizes benefits as these values vary both daily and seasonally. We used GPS data and Hidden Markov Models to determine space use of coyotes and estimate three types of movement behavior: encamped, foraging, and traveling. We then used integrated step-selection analysis to investigate different behaviorally-explicit resource selection across times of day (diurnal, crepuscular, and nocturnal) and season (snow-free and snow). We drew two main conclusions: 1) coyotes shifted foraging behavior throughout the day and seasonally; and 2) coyotes altered behavior and resource choices throughout the day and seasonally to minimize movement cost on the landscape. By examining resource selection across three axes (behavior, time of day, and season), we have a more nuanced understanding of how a predator balances the cost and benefits of a stochastic environment.


Location: Virtual Date: Time: -