Biometrics and Population Ecology

Poster

  • Discrete-Time Models of Animal Movement: Mule Deer As a Case Study*
  • Meghan P. Keating; Perry J. Williams; Kelley M. Stewart; Levi J. Heffelfinger; Cody J. McKee
    Water availability is a critical driver of habitat suitability for wildlife living in arid environments. Understanding movement behavior and identifying areas of attraction (or repulsion) is critical for wildlife management and conservation. We examined how water availability influenced mule deer (Odocoileus hemionus) behavior in the Mojave National Preserve in southern California. We fit discrete-time movement models to > 900,000 GPS locations collected from 170 marked individuals. We first developed a movement model that permitted estimation of mule deer attraction to known water sources. We then extended our model to estimate the location of features of the landscape that animals were attracted to, but for which we lacked information (e.g., unmapped water sources). Our results suggested attraction to water affected mule deer movement behavior at the individual scale, even when animals were not using water sources. We were also able to provide a spatial surface describing the probability of attraction across the landscape to unknown features. Our results augment the inference obtained from resource selection functions or occupancy models by providing inference on how features of the landscape affect individual movement behavior. Our results also provide information to generate a spatial map for managers that describes areas that appear to be driving animal movement via attraction and repulsion, along with the associated uncertainty of those locations, even when we lack information on why animals are attracted to those locations.

  • Population Estimates and Abundance of Birds in the Loky Manambato Region of Madagascar*
  • Giovanni T. Walters; Brandon P. Semel; Dehlan Estes; Dean F. Stauffer; Aylett Lipford; Angelo Andrianiaina; Tamby Ranaivoson; Faramalal Vololonirina; Ando N. Rakotonanahary; Dimbisoa Rasolomanana; Sarah M. Karpanty
    Madagascar, 1 of the world’s 8 biodiversity hotspots, is home to 223 native bird species. These birds have been minimally studied, and knowledge of population abundance and densities is lacking. With ongoing habitat fragmentation and climate change, knowing baseline population sizes is necessary to understand future anthropogenic impacts. The diversity of forest types within Madagascar’s Loky-Manambato Protected Area offered a unique opportunity to study a diversity of birds (152 species) typically found in disparate forested areas across the island. Our research provides the first density and abundance data for the birds of the Loky-Manambato Protected Area, determines the species composition of bird communities in a wide variety of forest fragments, and examines the relationship of forest types and disturbance to species densities and composition. We conducted line transect surveys along 28 transects (1-3 km long, total effort = 1,358 km) in 9 forest fragments and in forest/grassland matrix from 2016-2018 and determined densities using the program Distance. Relationships between bird densities and environmental factors were determined using ordination methods and generalized linear models in R. Environmental factors examined include average NDVI values for the wet and dry seasons, sampling year, forest fragment ID, canopy height, canopy cover, vegetation stratification, tree density, tree basal area, tree crown volume, 4 vegetation diversity indices, and 4 disturbance measures. We ranked resulting models for individual species using AICc values. A majority of the bird species were not tolerant to disturbances, but common cisticola (Cisticola cherina), Madagascar kestrel (Falco newtoni), Namaqua dove (Oena capensis), and Madagascar manikin (Lepidopygia nana) abundances all indicated a moderate-high level of tolerance for habitat that had been browsed heavily or burned. This study provides baseline population data for monitoring efforts in the region and provides and improves our understanding of Madagascar’s avian ecology.

  • Habitat Utilization-Availability of Nilgai in South Texas*
  • Megan Granger; Clay Hilton; Scott Henke; Tyler Campbell
    Nilgai antelope (Boselaphus tragocamelus) are bovids that are endemic to India and portions of Pakistan and Nepal. They were introduced into South Texas in the 1930’s and now have a free-roaming population of approximately 33,000 individuals. Past studies suggest that nilgai in Texas prefer sparse forest and coastal prairie habitats. Nilgai have a broad diet consisting of 66% grasses, 25% herbaceous species, and 15% browse. Further studies of nilgai habitat use and avoidance is needed to effectively manage the populations in South Texas. Using data collected from aerial nilgai surveys in 2017-2020 on the East Foundation’s El Sauz Ranch, we will use the Neu et al. (1974) method to analyze utilization-availability data. A chi-squared test will be used to test the hypothesis that nilgai utilize each habitat type in exact proportion to its occurrence within the study area. This method uses a Bonferroni statistic which will allow us to estimate whether observations of nilgai occurs more or less frequently than expected given the amount of each habitat type available on the property. The results will give us a better understanding of which habitat types nilgai prefer and which they avoid. With nilgai populations continuing to increase and expand, insight into their habitat utilization will be essential in the management of the species in Texas.

  • Occupancy and Co-Occurrence of Forest Wildlife in Southern Illinois*
  • Justin J. Remmers; Clayton K. Nielsen; Damon B. Lesmeister
    Two of the most influential factors affecting forest wildlife distributions are habitat and intraspecific interactions. Midwestern temperate forests offer diverse landscapes and ecosystems to explore such factors; however, research across large geographic regions that also incorporates a wide array of taxonomic groups remains scarce. To address this gap in the literature, we conducted camera-trapping surveys and are using occupancy modeling to (1) identify areas of high biodiversity and model the influence of habitat on species richness, (2) investigate temporal and spatial partitioning between eastern gray squirrels (Sciurus carolinensis) and fox squirrels (S. niger), and (3) assess co-occurrence patterns between predator and prey species. During January-April 2008-2010, we deployed 3-4 cameras to 357 sites (n=1,188 total camera locations) across a 16,058 km2 region of southern Illinois, USA, that was comprised of a continuum of forest cover. Thirty-three variables were measured at each camera site and encompassed both microhabitat variables (e.g., total basal area, stem density, coarse woody debris) and macrohabitat variables (e.g., landcover type, patch size). We collected >100,000 unique photographic captures of endothermic animals and observed 29 different wildlife species or taxonomic groups. We are currently applying novel occupancy modeling approaches to these data including: a multi-species, hierarchical model with 2 model variants; a multi-season, multivariate, co-occurrence model evaluating occupancy changes through time; and a single-season, multi-species, co-occurrence model capable of incorporating >2 species. Additionally, we will use a kernel density analysis to examine activity patterns between species. We hypothesize decreased distance to anthropogenic features (e.g., roads, buildings, municipalities) will decrease species richness, increase local extinction rates for species sensitive to distance, and increase local colonization rates for species resilient to distance. Furthermore, we expect microhabitat variables to be positively correlated with mammal occupancy. Our research will provide managers with information to forward wildlife conservation in forested Midwestern landscapes.

  • Reproductive Rates, Kitten Survival, and Den Site Selection of Bobcats in the Black Hills, South Dakota*
  • Erin Morrison; Christopher Rota; Chad Lehman; Brady Neiles
    The bobcat (Lynx rufus) is an important furbearer in South Dakota. However, management of bobcats can be difficult because of their elusive nature and lack of demographic information. Matrix projection models can be used by managers to determine how population growth rates may vary in response to management actions, but gaps in key demographic rates remain. In particular, there is a lack of information regarding reproduction, kitten survival, and cause specific mortality of kittens. Additionally, kitten survival relies on sufficient breeding habitat and conserving this habitat is fundamental to maintaining a stable population over time. This study will provide the remaining vital rates necessary to build a bobcat population model, estimate population growth rate, and elucidate what environmental characteristics are associated with bobcat den site selection. My objectives for this study are to: (1) Obtain estimates of reproductive rates for bobcats in the Black Hills, South Dakota; (2) obtain estimates of annual survival rates and cause specific mortality for bobcat kittens; and (3) determine resource selection of den sites. I will locate dens of VHF collared adult female bobcats using ground triangulation. Once dens are located, I will obtain estimates of reproductive rates by directly observing the number of kittens in the dens. To evaluate kitten survival, I will fit bobcat kittens with VHF telemetry collars. I will then locate kittens weekly. If there is a mortality, I will investigate the mortality and determine species specific predation using DNA sequencing. Once kittens depart dens, I will evaluate den site selection by comparing environmental variables at den sites to environmental variables at random sites within a female bobcat’s home range. At each den site I will measure a range of environmental variables including terrain ruggedness, vertical cover, and distance to roads.

  • Space Use in Free-Ranging Canids: Are Gonadal Hormones Required for Territory Maintenance?
  • Eric M. Gese; Patricia A. Terletzky
    Fertility control among carnivores has been used to reduce depredations on livestock and wild neonates, for population control, to modify behavior, to inhibit genetic introgression, and to reduce human-wildlife conflicts. Fertility can be accomplished via surgical or chemical sterilization, endocrine perturbation, and immunocontraception. While there is considerable knowledge concerning the techniques used to sterilize carnivores, there is little information concerning how the presence or absence of gonadal hormones influences space use. We examined territorial fidelity, home-range size and overlap, and survival rates of free-ranging coyotes (Canis latrans) with gonadal hormones present (tubal ligated females, vasectomized males) versus gonadal hormones absent (spayed females, neutered males). We examined home range data from 169 sterilized coyotes (83 females, 86 males) monitored in eastern North Carolina. Across sexes and hormonal treatment, there were no differences in home-range size and percent home-range overlap of coyotes with gonadal hormones present (tubal ligation, vasectomy) versus gonadal hormones absent (spay, neuter). There was no difference among average annual home-range fidelity rates between females with gonadal hormones present (0.90 ± 0.16) versus absent (0.86 ± 0.25). Similarly, average annual home-range fidelity rates for males was not different between males with gonadal hormones present (0.82 ± 0.19) versus absent (0.90 ± 0.20). We estimated annual survival rates for 179 sterilized coyotes monitored for 135,263 radio-days. There was no difference among survival rates between females with gonadal hormones present (0.80 ± 0.15) versus absent (0.79 ± 0.16). Annual survival rates of males were also not different between males with gonadal hormones present (0.75 ± 0.13) versus absent (0.85 ± 0.15). The absence of gonadal hormones did not appear to influence home-range size, overlap, and fidelity, as well as survival rates of free-ranging coyotes.

  • Grazing Effects of Bison on Native Plant and Arthropod Communities*
  • Molly Koeck; Heather Mathewson; Adam Mitchell; Donald Beard
    ABSTRACT Texas Parks & Wildlife (TPWD) manages a remnant herd of Southern Plains Bison (Bison bison) at Caprock Canyons State Park (CCSP) in Briscoe County, Texas. Today the bison population in CCSP is approximately 250 individuals. In summer 2018, CCSP officials established 20, 3×3-m2 grazing enclosures at random locations but with consideration of accessibility of the bison to the area. The enclosures are used to restrict bison grazing, creating a control area for vegetation and arthropods to persist without disturbance from bison. The impact of bison grazing on the biotic and abiotic processes can be evaluated through the use of these enclosures. Our goal is to identify the effects of bison grazing at CCSP on native plant and arthropod communities. We will use a paired design that compares communities in ungrazed enclosures to adjacent grazed enclosures. We will ensure that enclosures have similar soil and topography. At both grazed and ungrazed areas, we will survey vegetation characteristics (plant diversity, litter depth, Daubenmire cover classes, and height) as well as collect samples for analysis of nutritive quality for bison. We will also measure arthropod characteristics (diversity, abundance, and occupancy) within the same area using pitfall traps and bee bowls to target detritivores and pollinators, respectively. The results of this study be used to produce a TPWD management plan for CCSP that improves plant and arthropod diversity on landscapes grazed by bison

  • Survival of Headstarted Eastern Hellbenders Released in Ohio Streams
  • Noah Skinner; Matt Kaunert; Helena Littler; Andrew Travers; Ryan Brown; Julia Golias; Christine Hanson; Greg Lipps; Viorel Popescu
    The Eastern Hellbender (Cryptobranchus a. alleganiensis) represents an ancient lineage of long-lived (>25 years), large-bodied (up to 2.5 ft.), stream-obligate salamander endemic to the Eastern United States. Eastern hellbender populations have recently undergone rapid, enigmatic declines throughout their range; mechanisms driving declines are poorly understood, but often speculated to be associated with land use change, high sedimentation rates, degraded water quality, introduction of non-native predators, illegal collection and novel pathogens. Eastern Hellbenders are Endangered in Ohio, and state management agencies are implementing various conservation strategies such as headstarting and artificial nest box placement to augment population and breeding habitat. The goal of this study was to evaluate the survival of headstarted hellbenders released in Ohio streams. In August 2018, captive-reared sub-adult Hellbenders tagged with 12.5 mm PIT tags (2-3 year old, n = 205) were released across eight sites within three tributaries of the Ohio River in southeastern Ohio. We used amplified PIT-tag readers (BioMark HPR Plus) to monitor reintroduced Hellbenders during repeated surveys (5-9 repeats per site) in August-October 2018 and June-October 2019. Using CJS capture-recapture methods for the 2018 data, we estimated that apparent survival in the short term (up to 3 months post-release) was high (0.66 – 0.93) depending on the release site. We repeated surveys in 2019, and estimated abundance of animals persisting 3 sites with sufficient recaptures using POPAN methods. We found that 32 (95%CI 26-41) individuals persisted at the release sites for 1 year post-release (0.16; 0.13-0.20 of released animals), and there were differences between sites (one site had 0.35 survival). These results indicate that headstarting can be a successful conservation strategy for Eastern Hellbenders. We recommend careful consideration of release sites, and periodic augmentation with successive releases to build populations at sites where animals are successful.

  • Understanding the Disturbance Human Recreation Causes for Wildlife Using Multiple Dynamic Agents within an Individual Based Modeling Framework
  • Soraida Garcia; Alex Cohen; Shadi Atallah; Patrick A. Zollner
    Human recreation within outdoor landscapes is growing, and the extent and intensity of the impact of such activities upon the wildlife are increasing. For example, the presence of humans may increase risk averse behavior by wildlife, restricting the access of wildlife to essential resources, and reducing foraging, thereby negatively impacting breeding. Ultimately, the impacts that recreationists can have on wildlife include directly or indirectly altering population structure and community composition. Unfortunately, understanding the impacts of recreating humans upon wildlife is a complex challenge that is dependent upon the wildlife species and human activity types. Our understanding of human-wildlife relationships can be improved by combining results from empirical studies with simulation models to extrapolate mechanisms to a broader range of circumstances and investigate their implications. Accordingly, we developed an IBM modeling framework, that enables both dynamic virtual human and wildlife agents to change their actions based upon their state as a consequence of their interactions with their environment and other virtual agents such as birders and bikers. We use this framework to model the disturbance of birds, in the Lawrence Creek Forest Unit (LCFU) of Fort Harrison State Park, by human recreation. We parameterize the model with human recreation data collected through an intercept survey of recreationists at the park and bird data from published studies. We simulate 3 scenarios of alternative trail spatial configurations within the LCFU. Our results indicate that birders influence the rates of disturbance of birds and the energy provided to nesting birds. Interestingly, simulations with imposed static human behavior rather than dynamic human agents estimated a much lower impact of birders on the disturbance of birds. This comparison illustrates the value of simulating both human and wildlife agents in dynamic ways and demonstrates a better approach to understanding the conflicts between human recreation and wildlife.

  • Survival and Recruitment Rates of Repatriated Southern Fox Squirrel Population
  • John Huang; John Holloway; Jayme Waldron; Shane Welch
    Life history theory provides a framework for conservation biologists to examine how specific population parameters (e.g., survival, fecundity, and recruitment) interact with environmental perturbations (e.g., habitat loss and fragmentation) to drive population declines that ultimately result in a species imperilment. The objective of this study was to estimate survival and recruitment of a repatriated population of the southern fox squirrel (Sciurus niger niger) on a South Carolina sea island. Long-term studies assessing survival and recruitment of repatriated populations are scarce. Our study population was established in 2016 and 2017 from five donor populations. We used radio-telemetry and a capture-mark-recapture (CMR) study design to monitor our population for over 4 years. We expected that our estimates of survival and recruitment would support the hypothesis that southern fox squirrels exhibit a slow life history relative to the eastern gray squirrel (Sciurus carolinensis), an un-imperiled, syntopic congener. Specifically, we expected to find evidence of differential survival among age cohorts, with high adult survival and low pup and juvenile survival. We used Bayesian analysis to estimate the mean annual survival (0.561, [95%Credible Interval:0.153 – 0.884]), capture probability (0.567, [0.187 – 0.954]), annual local recruitment post-translocation (0.183, [0.080 – 0.327]). Southern fox squirrels exhibited traits of a slow life history (e.g., low recruitment rate). To better inform conservation of the southern fox squirrel, more information is needed on their survival and recruitment rates.

  • Occupancy, Abundance, and Population Genetics of Bobcats in Western Maryland*
  • Kevin J. Lamp; Berna Yalcinkaya; Angela Holland; Jacob M. Haus; Kyle McCarthy; Greg Shriver; Harry Spiker; Lisette P. Waits; Jacob L. Bowman
    Monitoring bobcat (Lynx rufus) distribution, abundance, and genetics are essential for their continued conservation. In Maryland, little is known about bobcat demographics but confirmed sightings east of historical bobcat range have managers interested in learning more about the species’ status. We aimed to estimate bobcat occupancy and abundance in western Maryland using two noninvasive sampling methods: camera trapping and genetic sampling. On 3 study areas, we overlaid a 5 × 8 grid network of 40 cells, each cell 5.5km2 in size. In January – March 2019 and 2020, we placed 1 camera in each cell and visited cameras weekly to collect photo data and re-bait the sites. We used a single season site occupancy model to estimate bobcat occupancy within our camera grid for each year. Additionally, we used a repeated count model to estimate bobcat abundance. To collect fecal samples, we surveyed transects consisting of hiking trails, closed roads, lightly used roads, and off-road vehicle trails from May – August 2019 and 2020. We used mitochondrial DNA to identify species and microsatellite loci to identify individual bobcats for use in spatially explicit capture recapture models. Camera trapping from 2019 resulted in 126 detections in 41 of 120 cells over 6,360 trap-nights. The 2020 camera trapping data is currently being analyzed. In 2019 we completed 5 surveys of 555km of transect, collecting 816 fecal samples for species and individual DNA analysis. Species identification yielded 243 bobcat samples and individual identification is still in progress. We will continue fecal sampling during May – August 2020. This research will not only address the lack of knowledge surrounding bobcats in Maryland, but it will offer state managers multiple options to monitor bobcat distribution and demographics into the future.

  • Carnivore Responses to Human Disturbance in Harenna Forest, Ethiopia*
  • Phillys N. Gichuru; Marcella J. Kelly; Matt Thornton; Chrystina Parks
    Bale Mountains National Park (BMNP), containing Harenna forest, is the largest and most biologically diverse protected area. It supports approximately 26% of mammals and 57% of bird species endemic to Ethiopia. The area also supports economy for ~100,000 people who depend on wild coffee planted in the forest, livestock grazing, and subsistence crop farming. However, there is no current data on mammalian wildlife status or responses to human disturbance. We report on the first camera-trapping study in BMNP, providing insight into species distributions and wildlife-human relationships. We also use occupancy modeling to quantify six carnivores’ responses to humans. We deployed 50 cameras stations, 0.5-2.0 km apart, from December 2015-September 2016. Preliminary analysis showed positive correlation in trap-rates between spotted hyenas (Crocuta crocuta) and humans and human-related animals (cows, dogs, donkeys, goats). We are analyzing occupancy/detection for each species and using 2-species, co-occurrence analyses for carnivore pairs and between carnivores and humans or human-related animals. Results will reveal landscape/habitat features influencing carnivore occupancy and will quantify potential interaction between target species by estimating species interaction factors (SIFs). We also will determine whether the SIF is mediated by habitat such as distance to forest edge and whether carnivore occupancy is higher at stations with and without domestic species and humans. We anticipate dwarf mongoose(Helogale parvula, common genet (Genetta genetta), honey-badger (Mellivora capensis), spotted hyena(Crocuta crocuta) and white-tailed mongoose (Ichneumia albicauda) will co-occur with humans. These species are opportunistic foragers and can take advantage of anthropogenic food that is increasingly available on the landscape due to habitat modification, leading to reduction in natural prey abundance. In contrast, we expect leopards (Panthera pardus)prefer closer proximity to water, grass and tree cover, and will avoid humans. Our study provides a baseline of information from which to build conservation plans for biodiversity of the area.

  • Southeastern Fox Squirrel Occupancy and Habitat Associations of the Piedmont and Coastal Plain Regions of Virginia*
  • Marissa H. Guill; W. Mark Ford; Jesse De La Cruz; Marc Puckett; Scott D. Klopfer; Brandon Martin
    In Virginia, fox squirrel (Sciurus niger) populations are still present in the Delmarva Peninsula and west of the Piedmont into the Appalachians. However east of the Appalachians, particularly in the lower Piedmont and Coastal Plain, fox squirrels are rare and patchily distributed, especially the Southeastern subspecies Sciurus niger niger. Regionally, formerly suitable habitat has been subjected to fragmentation and degradation of mixed pine-hardwood forests and bottomland hardwoods by conversion to agriculture and plantation forestry, as well as decades of fire suppression. From fall 2019 to the present, we have conducted continual camera-trapping and nest box grid surveys for fox squirrels on the Big Woods/Piney Grove complex in the Coastal Plain and Fort Pickett in the lower Piedmont – two areas in Southeastern Virginia believed to contain the largest and most intact pine savanna and mixed-pine hardwood forest patches. To date, our observations indicate a low level of site occupancy and daily detection probability of southeastern fox squirrels at the Big Woods/Piney Grove complex, ᴪ = 0.23, ρ = 0.06 across pine savanna, pine plantations, mixed pine-hardwood and bottomland habitats whereas over numerous surveys, fox squirrels appear to be absent from Fort Pickett. Results from our pilot survey will be used to conduct future focal live-trapping to catch and radio-collar fox squirrels for home range and habitat use analyses at Big Woods/Piney Grove complex. Continued documentation of absence at Fort Pickett may lead to potential re-introduction efforts.

  • Population Genetic Structure of Bobcats in South Dakota Using Harvested Samples to Inform Management*
  • Stuart C. Fetherston; Chad Lehman; Lisette P. Waits; Jennifer R. Adams; Robert C. Lonsinger
    A primary objective of state wildlife management agencies is to establish sustainable harvest levels for game species. An important component of sustainable management practices is the identification of appropriate management units for monitoring and establishing defensible harvest levels. Across their range, bobcats (Lynx rufus) are an ecologically and economically important species. In South Dakota, bobcat harvest was restricted to west of the Missouri River between 1977 and 2011. Since then, bobcat harvest has been permitted in select counties east of the Missouri River, but this harvest has been restricted (i.e., shorter season, bag limit of 1) relative to western South Dakota (i.e., longer season, no bag limit). Despite their importance, little is known about the genetic structure of bobcat populations in South Dakota. Our study is using tissue sampled from n = 1,162 bobcats harvested across the state, from 2013 to 2018, to infer population genetic structure and inform management. We are using 19 microsatellite loci and a sex identification marker to assign individuals to genetically distinct populations using program STRUCTURE. We will calculate standard measures of population genetic diversity (e.g., allelic richness, heterozygosity, inbreeding coefficient) and population differentiation (e.g., FST). Our results will provide additional information about bobcat populations in the state and inform harvest management of the species. Additionally, our results will be extended to investigate landscape connectivity for bobcats.

  • Evaluating Methods for Detecting Pangolins in Central and West Africa to Inform Conservation Actions*
  • Ichu G. Ichu; Dana J. Morin; Daniel W.S. Challender; Stephen F. Spears
    Tropical pangolins in Central and West Africa, and their parts (mainly scales), are trafficked in high volumes, increasingly to Asia. There is an urgent conservation need to mitigate this threat by protecting population strongholds in the region. However, little is known about the local distribution of the three species, including the white-bellied pangolin (Phataginus tricuspis), black-bellied pangolin (P. tetradactyla), and giant pangolin (Smutsia gigantea), which is needed to inform pangolin conservation efforts. Confirming species presence throughout their suspected ranges using standard field sampling methods is challenging and will likely require specific methods and protocols for each species. Our main objective is to identify efficient, rapid assessment methods to document pangolin occurrence at potential strongholds in Central and West Africa. To accomplish this objective, we will first evaluate the effectiveness of three sampling methods: camera-traps, eDNA, and semi-structured interviews, which have demonstrated variable success in detecting the three tropical African pangolin species. We will implement this study at Campo Ma’an National Park (CMNP), Cameroon, where each species is known to occur. We will use the accuracy and detection rates estimated for each method and species to identify an optimal standardized protocol for rapid assessment surveys in protected areas. This protocol will guide future monitoring efforts throughout Central and West Africa to inform conservation actions to protect pangolin populations.

  • The Myriad Effects of Landscape Composition and Island Accessibility on Predation Management on a Fragmented Landscape
  • Yulan Lu; John H. Porter; Brian M. Scharle; Raymond D. Dueser
    Effective predation management requires an understanding of both landscape composition and permeability to predator movement. The raccoon (Procyon lotor) is a significant predator on beach-nesting and colonial waterbirds on the U. S. Atlantic and Gulf Coasts. This study investigated the dual influences of land cover composition (Li) and island accessibility (Mi) on raccoon distribution and abundance on 31 island and marsh surfaces along the seaward margin of the southern Delmarva Peninsula. Eleven islands previously were classified as resident/source islands and 20 as transient/sink islands. We combined 11 years of USDA-WS raccoon trapping data (1,453 captures) on 11 islands with high-resolution (30-m pixel) NOAA C-CAP land cover maps for the year 2010 to determine a cover-class association for each of 7,431 trap stations. Eleven cover classes were represented. We tallied the area Aix (in hectares) for each cover class x found on surface i. Based on capture data, station GPS locations and cover-class information, we calculated the observed capture rate (rx) for each cover class, and estimated land cover suitability (Li) and island accessibility (Mi) for each surface. Given rxand Aix, we then calculated the projected numbers of captures (CI) for each surface. Both Li and Ciwere higher for woody islands than for grassy and for source islands than for sink. Mi varied independently of both cover composition and island type. These results suggest a clear guideline for predation management: Focus trapping on grassy islands to maximize protection of high quality shorebird habitat, but focus trapping on woody habitats to maximize capture rates. Trapping on grassy islands has been observed to both accentuate their sink function and provide direct protection for shorebirds. Trapping on woody source islands potentially may reduce overall raccoon abundance and disrupt the flow of dispersers between islands, but the benefits to shorebirds are less certain.

  • Habitat Suitability for Striped Skunks in a Spatially Heterogeneous Landscape
  • Katelyn Amspacher; Agustin Jimenez; Clay Nielsen
    Striped skunks (Mephitis mephitis) are generalist mesopredators found throughout most of North America. Previous studies of striped skunk habitat preferences reveal varying trends throughout their large distributional range, but no spatial models of striped skunk habitat suitability currently exist. Southern Illinois is a unique region to examine habitat suitability for striped skunks given the high interspersion of row-crop agriculture, urban development, and forest cover. We developed a species distribution model (SDM) to examine striped skunk habitat suitability in southern Illinois. We incorporated 234 occurrence locations collected via camera traps during 2007-2010 and 108 occurrence locations collected using radiotelemetry during 2018-2020. We incorporated 1 km2 land cover data from the National Land Cover Database and an index of human modification of the landscape. We built models using SDM Toolbox and MaxEnt and tested multiple regularization multipliers and feature class types. Our final model (OER = 0.3465, AUC = 0.700) included a 1.5 regularization multiplier with a hinge feature class. Land cover and human modification explained 93.8% and 6.2% of variation in the model, respectively. Highest habitat suitability for striped skunks was found in areas with forest and developed open space with moderate human modification, and lowest habitat suitability in cultivated crops and woody wetlands with either very low or high human modification. Forested land provides natural food and shelter resources for striped skunks, thus a highly suitable cover type. Resources are likely augmented by human activity in developed open space to create a similarly suitable habitat. Cultivated crops provide few shelter opportunities for striped skunks and are often managed to exclude many insects, a prominent food source for the species. Similarly, flooding in wooded wetlands limits available dry shelter opportunities, decreasing habitat suitability. Our model indicates that striped skunks are a synanthropic species that regularly inhabit both natural and anthropogenic habitats.

  • Resource Selection, Habitat Suitability and Connectivity for a Recovering Bobcat Population in Ohio
  • Madeline Kenyon; Ryan Brown; Marissa Dyck; Maddy Back; Viorel Popescu
    Species distributions and resource selection are shaped by biotic and abiotic factors as well as anthropogenic disturbance. Bobcats (Lynx rufus) have seen a spectacular resurgence in the US Midwest after a century of absence. However, much of the US Midwest is dominated by agriculture, urbanized areas and dense road network, which pose interesting questions regarding space use, habitat selection and habitat connectivity. In this study, we took a two-pronged approach to evaluate 2nd and 3rd order habitat selection and connectivity and potential dispersal pathways for bobcats in Ohio. To model 2nd order selection (habitat suitability), we used bobcat presences collected by citizen science data between 1990 and 2018, through a multitude of methods: trail camera, incidental trapping, or observations. We used logistic regression to model habitat suitability using presences and random pseudo-absences and found that at the population level bobcats avoided agricultural lands and selected for areas with higher proportion of forest and open natural habitat and avoided agricultural land. For the 3rd order habitat selection, we used a GPS telemetry dataset (2012-2015, points collected every 12 hours) of 20 bobcats in SE Ohio. We created 95% kernel density home ranges, and for each individual bobcat, we created random points within the home range boundary equal to the number of occurrences. We implemented Resource Selection Functions with weighted distributions with bobcat as a blocking factor and found that bobcats selected for areas with lower road density, higher canopy cover, and farther from high traffic roads, but were often found near low traffic roads. Our results show that the bobcat population may remain limited to the forested regions of Ohio due to their preference of forest and avoidance of agriculture. Across the Midwest, areas dominated by an intensive agricultural matrix present challenges for bobcat population persistence and expansion.

  • Environmental and Landscape Variables Influencing Occupancy of Coyote and Fisher in Western Maryland*
  • Berna Yalcinkaya; Kevin Lamp; Angela Holland; Kyle Mccarthy; Greg Shriver; Harry Spiker; Jacob L. Bowman
    Species diversity of mesocarnivores is greater than larger carnivores in the order Carnivora and mesocarnivores have more diversity in their ecology and behavior. Mesocarnivores are also indicators of environmental change. Determining the status and ecological effects of mesocarnivores is crucial for wildlife managers when making conservation and harvest-related decisions. To better understand the ecology of mesocarnivores, biologists need to know which environmental and landscape variables affect the occurrence of these species. We are investigating relationships between coyote (Canis latrans) and fisher (Pekania pennanti) occupancy and environmental/landscape variables in western Maryland. To get occurrence data of these species, we set camera traps using a 5 x 8 trapping grid of 40 cells (2.34 km x 2.34 km or 5.5 km²) covering 220 km2. We used this camera trapping design at Potomac-Garrett State Forest (7,689 ha), Savage River State Forest (22,033 ha), and Green Ridge State Forest (18,615 ha) in western Maryland during January-March 2019 and 2020. We will calculate environmental and landscape variables that we hypothesized would influence presence of fisher and coyote in the grid cells using geographic information system (GIS) and FRAGSTATS. We will model occupancy of coyote and fisher using these variables. Our findings will allow determination of environmental and landscape variables most suitable for these species and inform management and restoration of habitats to keep these species’ populations in balance at large scales.

  • Seasonal Migration and Habitat Selection of Mule Deer Across the Navajo Nation*
  • Hannah Manninen; Clayton Nielsen; Jessica Fort; Jeffrey Cole; Guillaume Bastille-Rousseau
    Mule deer (Odocoileus hemionus) populations have been declining throughout their range in the western United States since the 1980s. Habitat loss, overgrazing, disease, and predation contribute to the decline of mule deer populations. Navajo Nation, the largest sovereign tribe in the United States, encompassing 71,000 square km in New Mexico, Arizona, and Utah, has experienced a 53% decline in their mule deer population over the past decade. Lack of information on mule deer ecology on Navajo Nation led to the creation of the Navajo Mule Deer Project, with the Nation beginning to collar mule deer in February 2018. We are using GPS collar data from >80 mule deer to analyze long-distance movements, stopover sites, population-level migration routes, and arrival and departure dates of spring and fall migrations. Movements are being analyzed using net-squared displacement to determine if individuals are migrants, residents, nomadic, or dispersers. We are using Brownian Bridge Movement Models to determine stopover locations and population-level migration routes. Stopover sites will be classified as the highest 25% quartile in the utilization distribution. Population-level migration routes are being determined by combining utilization distributions and classifying the top 25% as high-use corridors. We are using normalized difference vegetation index data to determine if forage quality differs between stopover sites and movement corridors. Daily movements and habitat selection also will be analyzed. Our project will deliver valuable information to researchers and managers so more informed decisions can be made to recover mule deer populations, such as restoring habitat and conserving corridors.

  • Estimating Population Abundance and Growth of Elk in North Carolina Using Spatially Explicit Capture-Recapture Methods*
  • Jessica L. Braunstein; Joseph D. Clark; Justin M. McVey
    The once-abundant eastern subspecies of elk (Cervus elaphus canadensis) previously occupied the Carolinas but was extirpated by the mid-1800s as a result of habitat loss and overexploitation. In an effort to restore elk to their previous range, 52 elk were reintroduced into Great Smoky Mountains National Park (GRSM) in North Carolina during 2001 and 2002. Since their reintroduction, elk numbers have increased and their range has extended beyond GRSM boundaries. Our research objectives include estimating population abundance, survival, recruitment, and growth of elk in North Carolina and developing a reproducible and cost-effective sampling protocol that can be implemented by state wildlife managers for future population monitoring. We are using a spatially explicit capture-recapture (SECR) framework based on fecal DNA to estimate elk population abundance and growth in the region. We walk a series of straight-line transects throughout the region and collect elk pellets encountered along these transects. We conducted simulations using pilot data and developed a sampling design of 410 total transects with the number of transects decreasing with increasing elk density and transect length increasing in low density areas. The final design included a total of 280 500-m x 2-m transects and 130 1500-m x 2-m transects. We have completed one winter field season in January-March of 2020 and recorded over 600 pellet groups. We will conduct sampling over 2 more winter field seasons (2021 and 2022), and fecal DNA from pellet samples will be used to identify individual elk. These data, combined with spatial and home-range data, will be incorporated into SECR models to estimate elk densities across the region which can then be used to obtain an abundance estimate. Furthermore, the multi-year dataset will be used to estimate population growth, recruitment, and survival using open spatially explicit capture-recapture models.

  • Evaluating Density and Spatiotemporal Dynamics of Five Sympatric Ungulates*
  • Jennifer Foca; Mark Boyce
    Factors influencing ungulate distributions include population densities, interspecific interactions, seasonality, and changes in human use. Management strategies that alter population densities (i.e. culling, translocations) influence space use and interactions, but most studies fail to account for the variation of ungulate densities across the landscape. We are evaluating spatial covariance in ungulate densities in Elk Island National Park (EINP) and Cooking Lake -Blackfoot Provincial Recreation Area (BPRA) using trail cameras and aerial census data. EINP is home to plains bison (Bison bison), wood bison (B. bison athabascae), elk (Cervus elaphus), mule deer (Odocoileus hemionus), white-tailed deer (O. virginianus), and moose (Alces alces). Due to the fenced perimeter, wildlife movement is restricted and predation is limited. Active population management is necessary to prevent populations of elk and bison from becoming hyperabundant. BPRA has different management strategies compared to EINP and bison are not present. These adjacent areas provide a unique opportunity to evaluate density estimation methods for trail cameras and compare ungulate spatial dynamics under alternative management strategies. Aerial counts have been used in EINP for many years, but data on deer are limited and demographic data for all species are insubstantial. Trail cameras offer a cost effective, safer alternative to aerial surveys. We are using trail camera data to (i) estimate densities of each ungulate species, (ii) calculate demographic ratios, and (iii) evaluate spatial variation in density during and after bison removals.

  • Estimating Abundance of Black Bears and Mountain Lions Using Camera Surveys in West Texas*
  • Jamie L. Cooper
    Large carnivores act as keystone species as they play an essential role in maintaining the balance of ecosystems. Estimates of large carnivore population sizes are therefore useful not only for their management, but also for management of their prey species. Difficulties arise when studying these carnivores as they are often elusive, few in numbers, and have large ranges, but ever-evolving methods using remote cameras and spatial data may provide practical options for obtaining population estimates. Limited research on the population estimates of large carnivores in West Texas has been published, thus, our objective was to evaluate the use of camera surveys to estimate abundance of black bears (Ursus americanus) and mountain lions (Puma concolor) in the Davis Mountains. We established a 342-km2 grid with 36 remote cameras to survey for these species. We placed cameras at baited sites on likely travel corridors, such as mountain saddles and game trails. We surveyed for two 12-week periods (summer 2018 and spring 2019) and collected approximately 835,000 photographs. In summer 2018, we collected 9 independent photos of mountain lions and 25 independent photos of bears. In spring 2019, we collected 39 and 21 independent photos of mountain lions and black bears, respectively. We will use these photos to estimate abundance and evaluate the use of camera surveys for monitoring large carnivore populations in West Texas. These results will be useful for conservation and management specialists interested in non-invasive techniques for acquiring population estimates.

  • Developing a Sampling Framework for Elk Sightability in Eastern Kentucky*
  • Calvin C. Ellis; Michel T. Kohl; Michael J. Chamberlain; Kyle Sams; Gabriel Jenkins; John Hast; Jonathan L. Fusaro
    Elk (Cervus canadensis) were reintroduced to eastern Kentucky in 1999 after being completely removed from the eastern United States due to over-harvesting and habitat degradation. Today, they number approximately thirteen-thousand individuals, representing the largest elk herd east of the Mississippi. They are found throughout eastern Kentucky which is composed of a variety of habitats including conifer and hardwood forests, agricultural fields, and grasslands. Elk are known to select areas composed primarily of forest with hard mast production throughout the winter to maintain a sustainable food source. In addition, strip mining was historically popular throughout eastern Kentucky, but many mines have become inactive since leading to forest succession. Thus, the large proportion of forests on the landscape which elk inhabit make surveying for elk difficult. Thus, we are attempting to develop a framework for estimating elk abundance in eastern Kentucky. To first do this, we are using a previously developed resource selection analyses of elk GPS data to identify high use areas. Given this information, we will identify viewing locations from roads and high elevation to sight elk in these selected areas. This modeling process is designed to serve as the basis for a new field effort that better estimates elk abundance in Kentucky using remote game cameras and field observations. Further sampling will need to be conducted using this model in order to examine how other factors, such as time of day, affect elk sightability.

  • Does Water Availability to Coyotes Shift Dietary Preferences in the West Desert of Utah?
  • Ashley E. Hodge; Eric M. Gese; Bryan M. Kluever
    Historically, the kit fox (Vulpes macrotis) was the most abundant carnivore in the West Desert of Utah. At present, the coyote (Canis latrans) has become the area’s most dominant predator. Positive correlations between kit fox density and prey abundance have been revealed, as well as coyotes being the predominant source of mortality for kit foxes. However, the impact of this mortality on kit fox density remains unclear. The U.S. Army Dugway Proving Ground (DPG), Utah, has been a principal study site examining if the addition of artificial water sources may be a contributing factor to the observed influx of coyotes. Kit foxes have adapted to arid desert conditions by meeting energy requirements through preformed water (i.e., water from ingested prey). Coyotes reportedly need free-standing water, but in its absence they must triple their wet prey biomass consumption. Based on energy requirements, in theory, one could remove water sources thereby forcing the coyotes out of the area and releasing intraguild competition, thus assisting in kit fox population recovery. In 2012, researchers removed several DPG water resources, yet coyote home ranges and survival remained unaffected. Using a before-after control-impact (BACI) design, we examined ≥2,000 coyote scats for any potential dietary shifts towards higher mass prey using scats collected between 2010 and 2013 during the water manipulation. Based on previous literature, we originally predicted that coyotes in the water removal areas would increase jackrabbit in their diet to compensate for the lack of free drinking water. However, results do not support this prediction and suggest that coyotes can live in the absence of water without changing their energy requirements. This project will contribute to an extensive portfolio of long-term carnivore research in the West Desert, Utah.

  • Mapping a Landscape of Risk for White-Tailed Deer in a Multi-Carnivore System*
  • Robert S. Alonso; Dana J. Morin; David C. McNitt; Michael J. Cherry; Marcella J. Kelly
    The western mountains of Virginia have experienced a relatively recent coyote colonization, the restoration of bobcat and bear populations, and a decrease in the white-tailed deer population. Previous carnivore diet research in the region indicated that white-tailed deer were the most commonly shared diet item among these three carnivores with a coyote-bobcat dietary niche overlap of 0.73, coyote-bear overlap at 0.69, and bobcat-bear overlap at 0.47. However, this previous work could not distinguish between predation and scavenging events. All three of these predators have been documented as effective fawn predators, and to a lesser degree, adult deer predators, as well as efficient scavengers of white-tailed deer carrion. As such, it is apparent that white-tailed deer are likely an important diet item for these carnivores. Although we have not assessed direct effects of predation on deer populations in our multi-carnivore system, ecological theory behind the “landscape of fear” suggests that the mere presence of carnivores on the landscape could have non-consumptive effects on deer population dynamics via changes in behavior. These effects may vary through space and time depending on perceived risk across the landscape. To better understand the spatial processes behind predation, scavenging, and general ranging patterns of the deer predators in our study system, we deployed GPS collars on 10 bobcats (2017-2019), 29 coyotes (2011 – 2013; 2017-2020), and 28 black bears (2016-2019). As a first look of an on-going analysis, we utilize our GPS data to map carnivore landscape use and infer potential areas of risk for white-tailed deer. Our preliminary results suggest that deer must navigate a temporally dynamic landscape of predation risk that is mediated by variation in activity patterns and resource selection of carnivores. Seeking safer conditions may have nutritional consequences for deer as areas with greater forage quality also have relatively high predation risk.

     

    Poster
    Location: Virtual Date: Time: -