Poster Session IV

Poster
ROOM: CC, Ballroom C
SESSION NUMBER: 83
 

1 Arctic Ground Squirrels as Vital Signs of Tundra/Alpine Habitats – Predictors of Occupancy and Abundance in Denali National Park, Alaska
Nigel Golden; Toni Lyn Morelli
Increasing atmospheric greenhouse gases are causing global warming, particularly in sensitive climates like the Arctic. It is critical to understand the complexities of ecosystem responses to climate change, and the subsequent social and economic consequences, especially in the rapidly warming Arctic. The Arctic ground squirrel (Urocitellus parryii) (AGS) is considered an indicator of the health of vulnerable alpine and tundra habitat to climate change, yet little is known about the basic population dynamics of the species, let alone its response to climate change. The objective of this research is to determine the population structure of AGS, characterize environmental associations with abundance and genetic diversity, and ultimately to identify how vulnerable AGS is to climate change. Using a point-transect sampling design, we used the unmarked software package in R, to estimate the density of AGS by modeling the detection probability as a function of distance to the observer. Our estimated summer densities were low (27.9 animals/km2) and elevation was one of the best predictors of ground squirrel occupancy whereas the detection probability was dependent on percent cloud cover. Our current preliminary analysis has addressed how arctic ground squirrels are distributed and their density in Denali National Park. We are currently scheduled to begin capture-mark-recapture and to collect tissue samples in the summers of 2018 and 2019. This will allow us to update the analysis to address our inadequate understanding of AGS dispersal ability and ultimately determine which populations in Denali National Park are declining or healthy.
2 Assessing Changes in the Spatial Distribution of Swift Fox in Western Kansas
Ty J. Werdel; Adam A. Ahlers
Swift fox (Vulpes velox) are secretive mesocarnivores and are associated with short-grass prairie ecosystems. Prior to the 1950s, swift fox were extirpated from much of their geographic distribution. Swift fox populations have since expanded back into parts of their historic geographic range, however, ecological factors structuring their current distributions are still unclear. We will quantify how landscape composition (e.g., row-crop agriculture, prairie, cedar encroachment) and ecological factors (e.g., precipitation gradients, intraguild competition) structure the spatial distribution of swift fox populations in Kansas, USA. Additionally, we will develop a predictive map of swift fox occupancy for Kansas based on our most-supported habitat occupancy model. We will use a stratified-random sampling design to select 360 sites within the known geographic range of swift fox in Kansas and position a single camera trap at each site. Cameras will be placed for 28 days at each site between May-August. We will use a multi-season occupancy modeling approach to estimate swift fox detection probability, initial site occupancy, and colonization and extinction rates. Our research will be used to inform current management of swift fox on the eastern periphery of their geographic range. Additionally, our work will provide a contemporary baseline to assess changes in future swift fox distributions in Kansas.
3 Assessing the Negative Impacts of White-Tailed Deer Poaching in the United States and Broadening Availability of the Dna Forensics Tools to State Wildlife Agencies.
Melanie Quain; Jan E. Janecka
Most Americans support legal hunting for food, sport, and conservation. However, illegal hunting activities like poaching pose a negative impact on wildlife populations, their environment, and on the economy through lost revenue. State wildlife agencies rely on license sales and excise taxes on equipment. Hunters, trappers, and anglers provide 70 percent of funding that supports conservation and wildlife management initiatives. Furthermore, indirect spending by hunters during legal hunting activities supports the economy of rural communities. Poachers don’t pay into this system. With an increase in poaching there is potential for the non-hunting public to incorrectly assign their attitudes towards poachers to all hunters. The increasing trend in proposed anti-hunting legislation could sway voters to vote against legal hunting opportunities. This would have a negative impact on hunters, wildlife management, and conservation. In 2010, the Pennsylvania Game Commission (PGC) passed new laws and improved their enforcement to curb wildlife poaching; specifically, the white-tailed deer, Odocoileus virginianus. Nevertheless, the new laws have not substantially reduced poaching. The PGC transitioned to receiving wildlife crime tips through Operation Game Thief. This system has increased efficiency in reporting wildlife crimes. Although there has been a positive effect from this system, more work is needed in order to curb wildlife poaching in deer populations. Our study focuses on assessing the negative impacts of poaching and identifying factors vital to prevention. In addition, we are developing a more broadly applicable forensics panel to identify individual deer remains sampled by wildlife law enforcement officers. We have extracted DNA from 48 tissues provided by the PGC and genotyped 8 microsatellites. The final forensic panel will include a sex-marker and will be made available to other state wildlife agencies to enable successful convictions of poaching suspects. Our project will further aid in prosecuting poaching and supporting legal hunting opportunities.
4 Behavior and Spatial Ecology of a Desert-Adapted Mammal in a Changing Environment
Alexandra Burnett; John Koprowski
The Southwest United States has undergone substantial environmental change in the past century due to increasing temperatures, shifting precipitation patterns and fire regimes, and employment of management strategies for a multitude of land uses. Extreme climate conditions and continued urban expansion particularly threatens the Sonoran Desert, which is among the most biologically diverse deserts on Earth. It is unknown whether desert-adapted species will be more prepared to face future challenges or whether they will be more vulnerable due physiological limitations. The Harris’ antelope ground squirrel (Ammospermophilus harrisii) is endemic to the Sonoran Desert, exhibiting physiological and behavioral adaptations that enable it to maintain diurnal activity throughout the year. With few behavioral studies available, however, scientists have little insight into how this desert species may respond to rapid changes in climate patterns and community composition. To provide a framework for future research with A. harrisii and inform management decisions affecting desert rodents, I conducted a two-year study sampling Harris’ antelope ground squirrels in Southeastern Arizona from Sonoran Desert scrub at ~900-m to desert grassland at ~1100-m. The study area underwent extensive vegetation change over the past century due to climate change and past land use, a widespread pattern observed across arid grasslands worldwide. I fit 15 individuals of mixed sex with radio collars to track movement patterns and space use. I will provide seasonal home range estimates and observed behavioral patterns. I will additionally discuss potential implications of further environmental change on antelope squirrel populations and other small mammals in the Southwest, as well as pathways for future research efforts.
5 Camera Placement for Detecting Bobcats in the Southern Lower Peninsula of Michigan
Mrianda E. Millikin; Gary J. Roloff; Dwayne R. Etter; Melissa Nichols
Some central-Midwestern states are experiencing expansion of bobcats (Lynx rufus) into agricultural regions and some of the states have opened harvest seasons. Currently, southern Michigan is dominated by agriculture with pockets of forest. Recently, public reports of bobcats have become more common in southern Michigan and there is interest in documenting distribution and range expansion. To achieve this, the Michigan Department of Natural Resources implemented a camera trap study during the winters of 2017 and 2018. Our goal was to quantify the probability of detecting a bobcat based on commonalities in land cover and road density data to aid in future camera placement. We deployed 117 cameras between December and April at 64 sites (some sites >1 camera) in and around four southern Michigan state game areas (SGA). Cameras were active for 4 – 9 weeks providing an average of 49.8 (SE = 0.68) exposure days per camera. We detected bobcats at three state game areas and 13 sites. We used the 2016 Coastal Change Analysis Program (C-CAP) cover type raster to measure distances at each camera site to 9 cover types. Road densities were also calculated using Michigan All Roads raster within 1km of each camera site. To predict bobcat occupancy, we used generalized linear mixed effects models (SGA as a random effect) using land cover and road density. Our top-ranking model accounted for 95% of the weight of evidence and included distances to human development, emergent wetlands, open water and grassland/herbaceous/young forest. Only distances to emergent wetlands (Beta= 0.0006, 95% CI = -0.004 – 0.004) were significant. We applied our model to the larger southern Michigan landscape and generated a probability surface to aid in future camera placement.
6 Can a Large Carnivore Subsist on an Abundant, But Seasonally Available Prey Species?
David A. Keiter; John F. Benson; Arthur R. Rodgers; Brent R. Patterson
Abundance of primary prey is a key factor influencing population dynamics and distribution of large carnivores. Furthermore, the degree to which a population of large-bodied predators exploits alternative prey when its primary prey declines is likely an important determinant of population persistence for both predator and prey species. Thus, to better understand predator-prey dynamics and inform management, it is necessary to understand the mechanisms underlying carnivore population change during declines of their primary prey species. Gray wolves (Canis lupus) may present a valuable opportunity to evaluate the ability of large carnivores to switch prey. Wolves are commonly viewed as obligate ungulate predators, but beavers (Castor canadensis) are seasonally important prey for many wolf populations. It is unclear whether wolves can rely primarily on beavers during all seasons to persist when ungulates are scarce or absent. In this study, we will test the hypothesis that wolves are obligate ungulate predators on an island where their primary prey, woodland caribou (Rangifer tarandus), has recently declined, but beavers remain abundant. Our objectives are 1) to evaluate prey-switching by wolves with changes in abundance and availability of caribou and beaver, and 2) to assess how prey availability and consumption rates affect mechanisms underlying wolf population change (i.e. survival, reproduction). To accomplish these objectives, we will estimate summer and winter kill rates on caribou and beaver by investigating clusters of wolf GPS locations, evaluate seasonal and annual change in wolf diet using scat and stable isotope analysis, and measure change in wolf reproduction, survival, and abundance via known-fate survival estimation and non-invasive sampling techniques. Based upon the alternative prey hypothesis, we predict that wolf kill rates and diet will reflect a shift towards increased beaver consumption, and that wolf reproduction and survival rates will decline concurrent with this dietary shift over time.
7 Classifying High Frequency Acceleration Data to Understand the Movement Behaviors of Free Ranging Black Bears, Bobcats, and Coyotes
Robert S. Alonso; David C. McNitt; Dana J. Morin; Marcella J. Kelly
The field of movement ecology is rapidly progressing with improvements to global-positioning-system (GPS) collar technology and the subsequent improvements in movement modeling analysis methods. However, specifically studying behaviors of cryptic wildlife that live in habitats with high topographic variability and dense vegetation still remains challenging as interpreting spatio-temporal data alone limits inference about distinct behavioral states of wild animals. Recent advances in field deployable accelerometer units have opened a new window into interpreting behavioral states of collared animals, further advancing the field of movement ecology. As part of the Virginia Appalachian Carnivore Study located in the central Appalachian Mountains of Bath County, Virginia; we deployed 3-axis accelerometer units along with GPS collars on black bears, bobcats, and coyotes to record the acceleration movements of collared animals combined with GPS location data. Further, we video recorded captive bears, bobcats, and coyotes collared with accelerometer units to classify the video recorded behaviors (e.g. running, walking, resting, and feeding) to create a reference behavioral accelerometry dataset. In turn, we used that reference dataset to “translate” the accelerometry data recovered from collars we deployed in the field. Ultimately, we aim to combine our GPS collar location data with accelerometer data to understand how various behavioral states affect animal movement and habitat/resource use throughout the central Appalachian Mountains.
8 Managing Trophy Hunting of Bighorn Sheep: Should We Use Data, Models Or Opinions?
Marco Festa-Bianchet; Michael Morrissey
A recent invited paper in the Journal of Wildlife Management uses a model to claim that intense selective hunting cannot rapidly lead to the evolution of small horns in bighorn sheep (Ovis canadensis). That model has 4 major flaws: it attributes the entire phenotypic decline measured at Ram Mountain to genetic effects, uses artificially deflated values of Additive Genetic Variance (VA) and Total Phenotypic Variance (VP) even though published values for bighorn sheep are available, underestimates the amount of variance that can be detected by statistical analyses by confusing standard error and standard deviation, and uses an erroneous calculation of segregational variance to estimate how genetic variability may be eroded by selection. All four flaws contribute to results that minimize the importance of evolution. Once these flaws are corrected, the model predicts an evolutionary effect of greater magnitude than that empirically measured at Ram Mountain, confirming the evolutionary implications of intense selective hunting. When the published estimates of additive and phenotypic variance in bighorn sheep horn size are applied, Coulson et al.’s model predicts a decline in horn size of 0.5 standard deviations over just 6 generations at a harvest rate of 50% of ‘trophy’ rams, even when the other 3 flaws are ignored. Intensive monitoring of one population and long-term harvest record of mountain sheep in Alberta and British Columbia indicate that quota-free, intense, phenotype-based regulations for trophy sheep, which prevail in most of Canada, lead to evolution of small horns. Possibly because of a perceived threat to the image of sport hunting, however, opinions continued to be expressed questioning the validity of these results.
9 Co-Occurrence and Interspecific Interactions between Three Carnivore Species in the Starkey Experimental Forest
Rylee Jensen; Joel Ruprecht; Tavis D. Forrester; Darren A. Clark
The Starkey Experimental Forest is a 25,000-acre enclosure in northeastern Oregon established in the 1980’s, primarily used to study the effects of land management and interactions between wild ungulates and cattle. An ongoing study since the late 1980’s has focused on the interaction between elk (Cervus elaphus), mule deer (Odocoileus hemionus), and domestic cattle. However, until 2016, no research has been conducted on the carnivore species inhabiting the enclosure, namely coyotes (Canis latrans), bobcats (Lynx rufus), and cougars (Puma concolor). Coyotes and bobcats are considered dietary generalists and have comparable home range sizes, which suggests some niche overlap. Cougars have been known to kill both coyotes and bobcats, indicating possible effects through intraguild predation. Using co-occurrence occupancy models, this study aims to elucidate whether competition and/or habitat preferences play a role in limiting these species’ space use and daily activity patterns. Data will be collected using camera traps from April to October 2018. Previous surveys from the 2016 season (May-September) and 2017 season (April-September) will also be analyzed for comparison across years. Researching potential competition between these sympatric carnivores and their space use is important to more fully understand how carnivore interactions can limit their distribution and affect their impacts on prey populations.
10 Determining Species Richness and Abundance of Bats in Urban, Forested, and Pasture-Mosaic Habitats in the Southern Ecuadorian Andes Using Different Sampling Methods
Megean A. Myers; Taylor R. Azizeh; Rodrigo Cisneros; David Roon; Lisette Waits
Southern Ecuador faces ongoing degradation and loss of habitat in its main biodiversity hotspot, the Tropical Andes. Since the 1970s, annual deforestation rates have increased from 0.75% to 2.86%, and roughly 46% of southern Ecuador’s original forests were converted to pasture or agricultural land by 2008. Quantifying the impacts of this land conversion on biodiversity is critical for conservation planning. Bats make excellent bioindicators because they are abundant, easy to sample, and occur in many diverse habitats. Bats are also significant ecologically- with important roles in seed distribution, pollination, and regulation of insect populations. This research will assess the current species richness and abundance of bats in urban, pasture, and forested landscapes of the Zamora watershed near Loja, Ecuador. Data collection was initiated in summer 2017 and will continue in summer 2018. Bat communities will be sampled at 12 stream-catchments: three dominantly forest sites, three forest-pasture sites, three pasture sites, and three urban sites. Our methods of sampling will include the use of bat boxes, mist-netting, and acoustic surveys using the Wildlife Acoustics Song Meter SM3BA. We expect there to be greater species richness and abundance in the surrounding forest and pasture habitats compared to urban habitat, and this hypothesis is supported by data collected by mist netting in 2017. According to this 2017 data, bat species richness was greater in forest, forest-pasture, and pasture landscapes than in the urban landscapes. Furthermore, there was a significant decrease in bat activity with increasing levels of human land use. This research provides valuable data on the current status of biodiversity and ecosystem health in this region of the southern Ecuadoran Andes. It will provide important planning information for managers in the Loja area, and could be applied to other areas facing increased loss of biodiversity.
11 Development and Evaluation of Surveys for Occupancy Models of Mesocarnivores in Northern Utah
Eric Ethington; Sierra Bailey
Mesocarnivores, small to midsized carnivore species such as American pine marten (Martes americana) and gray fox (Urocyon cinereoargenteus), often garner far less attention than do larger carnivores, despite superior abundance and species richness in most ecosystems where they occur. In systems like Northern Utah where many large predators, such as brown bears (Ursus arctos) and wolves (Canis lupus), have been extirpated, mesocarnivores may actually function as apex predators and shape prey densities and community structure. However, despite their ecological significance, they are frequently designated as furbearers, vermin, or pests. Thus, it is critical for management to better understand both mesocarnivore species’ occupancy and habitat use within the system. The objectives for our project are to 1) develop and implement snow track surveys for mesocarnivores using double-sampled, one-to-two kilometer transects within randomly-selected mountain valleys and drainages in the Bear River Range of Northern Utah over a three-year time frame. 2) Evaluate snow track-based survey techniques under changing climate conditions with variable snow rates, and 3) develop occupancy models for mesocarnivores within the Bear River Range, including variable habitat use under snow conditions.
12 Development and Validation of a Spatially Explicit Individual-Based Model for African Savanna Elephant Space Use
Stephanie Diaz
Understanding the spatiotemporal dynamics of elephant movement across landscapes is critical for effective conservation of elephant populations in southern Africa. Current tools, however, have limited ability for predicting how elephants will respond to novel landscape conditions. Individual-based modeling can extend current methods and serve as a predictive tool, facilitating understanding of how future scenarios of landscape change may impact elephant movement dynamics. I developed a spatially explicit individual-based model (IBM) to simulate elephant space use for elephants in Chobe National Park (CNP). Known external and internal drivers of elephant movement, including perceived temperature and the time since an individual last visited a water source, are linked to the external environment through behavior-based movement rules. Movement within the model’s environment thus results from the agent interacting with the various landscape attributes. I am utilizing empirical movement data to validate model outputs and have identified characteristics of movement patterns for elephants in CNP, including home range size, total daily displacement distances, and net daily displacement distances. Preliminary model runs were replicated sixty times, and model outputs currently statistically underestimate home range sizes. Ongoing work sets out to understand the reasons for the discrepancies between model and empirical home range sizes and, based on that understanding, modify the model parameters to improve the IBM’s ability to project the movement patterns observed in nature. Once validated, the model will be used to predict how elephant space use will be impacted by changes in surface water distribution induced by the closure of artificial water points, as well as by potential changes in the landscape associated with global climate change.
13 Diet of a Recently Reintroduced Population of River Otters in the Upper Rio Grande River, New Mexico.
Gabriela A. Wolf
North American river otters (Lontra canadensis) were historically abundant throughout North America,but due to overexploitation, they were extirpated for large portions of their historic range. From 2008-2010, New Mexico Game and Fish reintroduced a total of 33 adult river otters in the upper Rio GrandeRiver after a 55-year absence. However, the program was discontinued in 2012 due to potential conflictswith game fish populations as well as several sensitive fish and amphibian species. As part of a largerpopulation assessment, we are conducting a non-invasive latrine survey occurring in the northernstretches of the Rio Grande in New Mexico. Because this is the first river otter diet study in the region,our goal is to describe the consumption frequency of sensitive species and game species in relation toother prey groups (i.e., other fish, crayfish, amphibians etc.). Beginning in January 2018, scat will besystematically collected from latrine sites and frozen. Since seasonal diet shifts have been observed inother regions, surveys will be completed over multiple periods(i.e., winter, spring, summer, and fall) inorder to capture any seasonal variability in diet. In the lab, we will use a fine mesh sieve to isolate preyremains (i.e., bones, scales, and exoskeletons). Recovered remains will be identified using a key of fishscales, otoliths, and bones, as well as crayfish exoskeleton compiled from state agency (i.e., New MexicoGame and Fish) samples. We will use frequency of occurrence within scat to determine diet preferenceand importance for this reintroduced river otter population with hopes of determining if any conflictsexist between the reintroduced river otter and local fisheries management goals. Results from winter,spring, and summer collections will be presented during the 2018 TWS meeting.
14 Do Domestic Dogs Enhance Or Spoil Competition Refugia for Gray Foxes?
Dana J. Morin; Damon B. Lesmeister; Clayton K. Nielsen; Eric M. Schauber
Data from Midwestern states suggest red fox (Vulpes vulpes) and gray fox (Urocyon cinereoargenteus) populations have declined, and a recent assessment of mesocarnivore occupancy across southern Illinois suggested gray fox range contracted while coyotes (Canis latrans) were ubiquitous. Asymmetrical competition among native canids resulting in intraguild killing, spatial displacement, and increased exploitative competition in human-associated habitats, are proposed agents of gray fox population decline. One complication previously unaccounted for in assessing dynamics among canids is the presence of free-ranging domestic dogs (Canis familiaris), common across the region. We analyzed an extensive camera-trap data set collected over three years at 1,181 stations across 16 counties in southern Illinois to evaluate factors influencing species co-occurrence and spatio-temporal overlap among domestic dogs, coyotes, gray foxes and red foxes. Dog occupancy was 0.59 and decreased with distance from structures. Data also lent support to models with an interaction term describing a positive effect of forest cover on dog occupancy on private land but not public land. We found no evidence for species interactions between domestic dogs and coyotes, weak support for a positive interaction between dogs and red foxes. Gray foxes and domestic dogs showed affiliation for the same sites but gray fox detection decreased when dogs were present, suggesting behavioral avoidance or decreased gray fox density. Temporal activity overlapped extensively among the nocturnal native canids while dogs showed greatest activity during the day. We found greatest potential for interaction between gray fox and dogs in the evening, and species interaction factors consistently >1 suggest high potential for domestic dogs to encounter gray foxes and thereby affect fox occupancy and behavior. Interacting competitive pressures from coyotes in forests, red foxes in anthropogenic habitats, and high co-occurrence with free-ranging dogs could exacerbate cumulative impacts to gray fox populations and contribute to the recent decline.
15 Does Risk of Predation Lead to Chronic Stress and Reproductive Suppression in Black-Tailed Prairie Dogs?
Colleen Crill; Jeffrey E. Lane
Black-tailed prairie dogs (Cynomys ludovicianus) are a foundational species in the mixed-grass prairie ecosystem, whose numbers have been declining in Canada over the past two decades. In some predator-prey systems, the risk of predation has been shown to increase hormones associated with the physiological stress response, and drive population declines by limiting reproductive success. I used hair cortisol to investigate whether the perceived risk of predation could be a factor contributing to the local decline of this species. I hypothesized that individuals on the edge of the colony would suffer from high levels of stress, due to a paucity of neighbours available to raise an alarm call. Further, I also expected that the individuals with higher levels of stress, measured through hair cortisol, would have reduced reproductive success compared to less stressed prairie dogs. I found that prairie dogs who live on the edge of the colony perceive a higher risk of predation, but this risk does not translate to significant differences in cortisol or reproductive success. Instead, cortisol varies between individuals, but is present at levels that facilitate an adaptive response to environmental challenges, and appears to be repeatable within individuals over time. These results suggest that the risk of predation, whether actual or perceived, should not be considered a limiting factor in the persistence of the Canadian black-tailed prairie dog population.
16 Early Stages of Restoration of a Western Tallgrass Prairie in Clay County, Minnesota: Impacts on Small Mammals
Donna M. Bruns Stockrahm; Sarah S. Sanderson; Miranda J. Sater; Rachel H. Rusten; Dylan C. Leach; Breanna N. Huynh
Minnesota State University Moorhead (MSUM) was recently awarded a grant from the “Minnesota Environment and Natural Resources Trust Fund” as recommended by the Legislative‐Citizen Commission on Minnesota Resources (LCCMR) to restore old farmland, called the “Houston Property” (HP) owned by the MSUM Regional Science Center (RSC), near Glyndon, MN, and an adjoining former golf course (GC) back to native tallgrass prairie habitat. One main objective of the grant was to live-trap small mammals to determine species composition, abundances, habitat use, and behaviors before, during, and after restoration. During summer 2015, small mammals were live-trapped on HP and GC to monitor “pre-restoration” populations. We re-trapped in the summers of 2016 and 2017 during the early stages of the restoration. Herbicide was applied to the HP during summer 2016 and to the GC in 2017 in preparation for planting a prairie seed mix of forbs/grasses in those 2 respective places in the respective years. In 2017, trapping was conducted in July-August and again in September as part of the grant to integrate restoration efforts into related college courses. Deer mice/white-footed mice (Peromyscus spp.) were very abundant on HP throughout the 2017 season (n=720), probably because of the frequency of over-night trapping and the great improvement in vegetative cover compared to 2016. HP also yielded higher captures of thirteen-lined ground squirrels (Ictidomys tridecemlineatus) (n=23) than during 2016, likely due to higher vegetation cover. GC only had captures of meadow voles (Microtus pennsylvanicus) (n=15), mainly from one grassy area, probably because herbicide had been recently applied to other areas of the GC in summer 2017. Our 2017 findings contrasted sharply with those of 2016, where HP produced low capture numbers and low species diversity soon after herbicide application and re-planting whereas GC had higher abundance and species diversity prior to herbicide application.
17 Ecotoxicology of Bats in Urban, Pasture and Forested Sites of the Southern Ecuadorian Andes
Taylor R. Azizeh; Megean A. Meyers; Rodrigo Cisneros; David A. Roon; Lisette P. Waits
Urbanization and agricultural conversion are the leading global causes of deforestation and biodiversity loss. Latin America is the second most urbanized region in the world with 80% of people living in cities. Agricultural conversion is the main driver of deforestation in tropical and subtropical countries accounting for 80% of deforestation from 2000-2010. This habitat alteration can lead to bioaccumulation of pollutants and other toxins that are harmful to humans and other species. For example, toxins leeched into the ecosystem from agriculture and urbanization have been shown to degrade red and white blood cells and create primary physiological responses such as DNA damage. Insectivorous bats can be valuable bioindicators of human health risks, because these species are at the same trophic level in the food chains as humans. The goal of this research is to evaluate whether urbanization and conversion of land to pasture is negatively affecting the physiology of bats within the southern Ecuadorian Andes. Our study area is in the Zamora watershed near the city of Loja, and bats will be captured by mist netting at 12 stream-catchments: three forest sites, three forest-pasture sites, three pasture sites, and three urban sites. Blood samples will be collected from each bat for genotoxicology analysis. Genotoxic effects will be measured using the Comet assay and a micronuclei assay. We hypothesize higher levels of DNA damage in bats captured in urban and pasture areas compared to bats captured in forested areas. We expect to see the highest levels of DNA damage in insectivorous bats because of the bioaccumulation of pollutants in the insects. This research will provide valuable baseline data on the toxological effects of land conversion on the overall health of bat populations, as well as identify regions with high levels of bioaccumulation that may also pose a risk to human health.
19 Evaluating Potential Effects of Camera Density on Capture and Recapture Rates of Bobcats
Chris Jacques; Tim Swearingen; Edward Davis; Robert Klaver; Charles Anderson; Christopher DePerno; Jonathan Jenks; Robert Bluett
Performance advances among remotely triggered cameras has prompted their widespread use in estimating density (the single parameter of greatest intrinsic importance to wildlife managers studying population dynamics) of elusive and mobile predators such as bobcats (Lynx rufus). Corresponding advancements in statistical methods are available for estimating population density from capture-recapture studies. Nevertheless, reliability of density estimates varies widely between species, due to heterogeneity in numbers and placement of camera stations among studies, and the ability to capture and recapture individuals across multiple camera trap locations. We evaluated potential effects of camera density on capture and recapture rates of bobcats in agriculturally dominated landscapes of west-central Illinois. We deployed 31 camera stations during two sampling intervals (15 May-15 Jun 2016, 20 Apr-20 May 2017) over 1,800 trap nights. Our analyses revealed that effects of camera density on bobcat detection probability was marginally significant (F2,3 = 7.33, P = 0.07, R2 = 0.22); maximum detection (mean = 0.53, SE = 0.08) was associated with moderate camera densities (4-6 cameras/9 km2). Numbers of individual bobcats detected varied (F2,3 = 9.93, P = 0.04) with camera density; moderate and high (8-10 cameras/9 km2) camera densities yielded greater (P ≤ 0.05) numbers of individuals than lower camera densities (1-2 cameras/9 km2) and we detected no more than 4 individual bobcats at any 1 camera station. We documented no differences (F2,3 = 4.21, P = 0.14) in recapture rates between low and high camera densities, though noted positive associations (R2 = 0.71) between recapture rates and increasing camera densities. When photo-capture and recapture rates are a function of camera density, modifying camera trapping techniques by deploying moderate camera densities or repositioning cameras to more productive areas within survey grids may improve capture success in low density bobcat populations.
20 Evaluating the Effects of Roads and Mitigation Efforts on Wildlife Movement and Mortality: 15 Years of Collaboration with Caltrans in the Los Angeles Area
Seth P.D. Riley; Justin L. Brown; Joanne G. Moriarty; Jeff A. Sikich
Roads of all sizes can have significant effects on wildlife, including by causing mortality through vehicle strikes and by creating barriers. The Los Angeles area has an extensive road and freeway system, including some of the busiest freeways in the world. At Santa Monica Mountains National Recreation Area, we have been working closely with Caltrans to understand both the impacts of various roads and the potential value of efforts to mitigate those impacts. Since 2002, we have collaborated on six different projects involving measuring mortality through roadkill surveys, evaluating the use of various road crossing structures, and intensively tracking wildlife, particularly mammalian carnivores, to determine their behavior relative to the road. We have also monitored the effectiveness of mitigation efforts such as clearing culverts, improving fencing, adding one-way gates, and adding a wildlife “sidewalk.” Overall we have found that even smaller highways can form important barriers to movement for species such as bobcats, but that effective underpasses can significantly increase connectivity across them. We have also found that the response to roads can vary by species. For example coyotes regularly cross over the road surface itself, while bobcats depend more on underpasses, which means that the barrier effects may be stronger for bobcats, but that coyotes are more susceptible to vehicle mortality. In terms of mitigation, clearing culverts can significantly increase their use, but failure to consider the specific needs of wildlife can easily render crossing structures of little value. On one major project adding carpool lanes to the 405 Freeway we found that the period of construction itself, often not considered in before-after wildlife road projects, can have a significant impact on wildlife use. Overall, evaluating multiple different aspects of wildlife response, for as long as possible, gives the best picture of road impacts and how to reduce them.
21 Evaluation of Seasonal Variations in Food Composition and Nutritional Value in a Sika Deer Population on a Deteriorated Habitat
Mizuki Kaneko; Kazutaka Takeshita; Kiyoshi Tanikawa; Koichi Kaji
High density deer populations have been sustained in temperate forests despite striking decline in vegetation by severer grazing. Deer populations that have access to alternative food may maintain high density, however, there are limited studies on food habits of deer living in deteriorated habitats. To understand the mechanisms of sustaining high deer density in a scanty vegetation habitat, we evaluated seasonal food composition and nutrition in a sika deer (Cervus nippon) population on Tanzawa Mountains, central Japan, where habitat has been deteriorated by long-term chronic overgrazing. We collected rumen contents of culled deer from May 2015 to June 2016 (n = 128) and evaluated the food composition by point frame method and the nutritional value of the rumen contents by using crude protein (CP) contents. Food composition showed clear seasonal variations, however, no spatial variation and sex-age differences, which were divided into several groups in each season. In winter, twig, bark and dead leaves were staple foods in most groups. In spring, two types: high quality foods such as forbs and deciduous broad leaves and low quality foods such as twig, bark and dead leaves were staple foods. These high quality and low quality foods were also used in autumn. For seasonal level of nutritional value, the winter and autumn CP contents in rumen were higher than requirement for maintenance of body mass and lives, while the spring CP contents in rumen were higher than requirement for growth and lactation. These results suggested that the resources from forest canopy such as dead leaves and evergreen broad leaves supply all year around and the abundance is independent of browsing, which could sustain high density of deer population on the deteriorated habitat.
22 Evaluation of the Tooth Wear and Replacement Method for Aging Male White-Tailed Deer in Iowa
Daniel Adams; Julie Blanchong
Assessing demographic parameters such as population age structure is a key component of many wildlife management programs, with ages of harvested animals being a significant factor in that assessment. The most common method used to estimate ages of harvested white-tailed deer (Odocoileus virginianus) and other cervids is a criterion based on tooth wear and replacement (TWR). Previous studies have shown this method is prone to considerable error because TWR is partially subjective. A more accurate, but more labor intensive and expensive, method to estimate age involves the counting of cementum annuli (CA) of cross-sectioned incisors. Our objective was to identify error rates of TWR age estimates of male white-tailed deer in Iowa. We estimated the age of 404 adult (≥1.5 year old) male white-tailed deer harvested in Iowa (USA; 2015-2016) using both TWR and CA methods and evaluated error rates TWR using CA estimates. We classified deer using TWR into age classes of 1.5, 2.5 and 3.5+ years. Deer 3.5+ were grouped together because it is well-documented that TWR error rates increase with age. Assuming CA age represented the true age, we found that the TWR method was prone to considerable error at all age classes (26% error in 1.5 TWR age classification, 28%-2.5, 18%-3.5+). We also obtained paired age-estimates for 180 deer (≥50 pairs from each TWR age class) to evaluate precision of CA estimates. Our results suggest that while TWR is efficient and cost-effective, it may lead to inaccurate age-structure estimates that could have potential downstream impacts on demographic and epidemiological models. Managers interested in obtaining accurate age information for cervid populations should invest in the CA method, at a minimum, to identify misclassification rates using the TWR method for their area.
23 Factors Affecting Glucocorticoid and Thyroid Hormone Production By Island Foxes on San Miguel and Santa Rosa Islands
Helen Clawitter; Corinne P. Kozlowski; Angela Guglielmino; Juliann Schamel; Stacy Baker; Ashley Franklin; Cheryl S. Asa
The Island fox (Urocyon littoralis) is native to six of the eight Channel Islands of California. The species experienced a population decline in the 1990s but has rebounded, and in 2016 it was removed from the Endangered Species List. As part of the recovery effort, Island foxes undergo an annual census in which individuals are trapped and health parameters are assessed. Here, we document non-invasive measures of stress for foxes trapped on San Miguel and Santa Rosa Islands from 2009-2015. Fecal glucocorticoids (fGC) were quantified to estimate physiological stress; fecal thyroid concentrations (fT3) were assessed as a measure of nutritional stress. Samples defecated overnight in a trap (thus reflecting pre-trapping concentrations), or during handling were collected and sent to the Saint Louis Zoo Endocrinology lab for analysis. Hormones in 743 samples were extracted according to published protocols, and concentrations were quantified using commercially available immunoassays. ANCOVAs were used to assess the relationships between hormone concentrations, month, year and island, age-class, sex, and body condition. Concentrations did not vary between males and females, among years, or in relation to body condition. FGC concentrations were marginally higher in foxes on Santa Rosa than on San Miguel, and higher concentrations were measured in younger animals on both islands. Seasonal trends were also detected; fGC concentrations were highest from June to August and decreased from September through December; no seasonal trends in fT3 concentrations were observed. These results suggest that foxes experience heightened stress in the summer, corresponding to the period of pup weaning. Younger animals appear more impacted than older individuals, possibly due to heightened disease susceptibility. Our study provides the first data on measures of stress for Island foxes. This information may assist recovery efforts by identifying factors correlated with stress and may serve as baseline data for future health monitoring.
24 Forensic Methods for Species and Sex Determination of Unknown Wildlife Carcasses
Amy L. Weber; Elizabeth B. Allmon; Maria S. Sepulveda
Forensic methods used to solve crimes for human victims have been widely researched and documented. Application of these methods to other victims such as wildlife, has not been widely studied to date. Research must be done to confirm application of these methods for use in court cases against poachers. The American court system requires the scientific community to peer-review research and broadly accept the results of this research before data collected by those methods may be admitted as evidence in trial. Research objectives are to confirm efficacy of forensic methods designed for human crime are applicable to wildlife crimes and to provide an alternative method to determine the species and sex of unknown animal remains. Exposed carcasses will be colonized by blowflies, the resultant maggots will be collected. Maggots fill their crops with tissue from the carcass they colonize. The crop stores this tissue as it is ingested faster than it can be digested. This tissue is a great source of DNA and will be used to identify species and sex of the carcass. Extraction of DNA using the cytochrome B gene, followed by PCR and gel visualization. Species will be identified through the DNA fingerprinting method that uses the COI gene and species-specific primers. Sex determination with primers designed from the testes specific SRY 1 gene. The animals included in this study are Raccoon (Procyon lotor), Mute Swan (Cygnus olor), American Beaver (Castor canadensis), Bobcat (Lynx rufus) and Red-tailed Hawk (Buteo jamaicensis). All carcasses will be sexed prior to decomposition. Maggots collected from known carcasses will be analyzed. Maggot species will be visually identified and DNA testing to confirm visual identification. Maggots will then be compared across carcass species to look for trends or variations.
25 Fukushima’s Wildlife: Camera Trap Analysis of a Shuttered Region
Phillip C. Lyons; Matthew T. Hamilton; Thomas G. Hinton; Kei Okuda; James C. Beasley
In recent decades, the interactions between humans, environmental contaminants, and wildlife have been extensively studied. However, there remains a paucity of data on the impacts of chronic radiation exposure on wildlife. The Fukushima nuclear accident, and the subsequent human exclusion zone, provides a unique opportunity to study wildlife populations across a gradient of radioactive contamination and human habitation. Through the use of remote cameras, this study aims to determine the distribution and relative abundance of wildlife in and around the Fukushima Exclusion Zone (FEZ). In 2016, we placed unbaited cameras along game trails or travel corridors at 120 sites equally distributed between upland and lowland habitats and along a gradient of human habitation for two 60-day periods. To produce single-species occupancy results, we utilized the R program unmarked. Our models indicated there may be a positive relationship between wild boar (Sus scrofa) occupancy and radioactive contamination. Additionally, occupancy of Japanese macaques (Macaca fuscata) and Japanese serow (Capricornis crispus) was higher in populated areas. By contrast, occupancy for other species, such as masked palm civets (Paguma larvata) and wild boar, was higher in less populated areas. Importantly, there appears to be no uniform response to either radioactive contamination or human habitation among the FEZ’s species. This data, in conjunction with ongoing abundance analyses and GPS collar data collection, will provide the most comprehensive information to date on the distribution and composition of the mammalian wildlife community in and around the FEZ.
26 Genetic Analysis of Snow Leopard Employing Next Generation Sequencing for Improved Conservation and Management
Safia Janjua; Jeffery L. Peters; Thomas P. Rooney
Despite having the high profile of a charismatic carnivore, information on snow leopard biology, population structure, and genetics is scarce. This is because of its cryptic nature and remote habitat. As a result, we lack sufficient data to identify population numbers, locations of peripheral and core populations, and areas where they are in decline. Such information is needed for the conservation of this apex predator. Molecular tools can provide important insights into conservation in ways traditional field research cannot. Genomic analysis of DNA extracted from non-invasively collected snow leopard samples can benefit conservation efforts. Previously, genotyping errors were common due to the low DNA yield and quality obtained from such non-invasive sources. These errors lead to incorrect inferences, such as misidentification of individuals. Next generation technologies have revolutionized the depth of information we can get from a species’ genome. Here we used ddRAD-seq, a well-established technique for studying non-model organisms, to develop a reference sequence library for snow leopards using blood samples from five Mongolian snow leopards. Our final data set has 4504 loci with median size range of 221 bp. We identified 697 SNPs in five Mongolian snow leopards with very low nucleotide diversity i.e. 0.00032. The probes for DNA capture method using this sequence library are developed which we are using for global analysis for SNP variability in snow leopards. Genetic data from ddRAD-seq is proving useful for identifying snow leopard population conservation strategies.
27 Habitat Requirements and Occurrence of Snowshoe Hares in Northern Lower Michigan
Spencer West
Snowshoe hare (Lepus americanus) are a mid-sized North American lagomorph that range from northern continental U.S. through Canada and Alaska. Their preferred habitat is northern boreal forest with a thick undergrowth, typically dense in coniferous vegetation. In their core range, climate is cold with long winters and high periods of snowpack. Recently, populations at the southern reaches of their range, such as Michigan’s Lower Peninsula, have struggled and decreased due to changing climate and habitat. Warmer winter temperatures have decreased the amount of winter snowpack, reducing access to higher browse in the winter. Reduction of snow also can cause a camouflage mismatch between the white winter coat of the hares and the brown forest surroundings. These impacts have already reduced hare occupancy in the Huron National Forest while reports of hare sightings in the Manistee National Forest have also declined. This project aims to assess the occupancy of snowshoe hares in the Manistee National Forest through the use of camera traps and track surveys. Live trapping, and radio telemetry will be used to provide insight into local habitat requirements, home range information, movement patterns, and dispersal data. These findings will be used to provide management recommendations to improve and increase snowshoe hare populations in Michigan’s Lower Peninsula. Individual hares thus far have found to prefer dense young aspen stands due to a lack of dense coniferous understory.
28 Identifying Cryptic Species of Tarsiers on North Sulawesi, Indonesia
Thalita C P Sumampow; Paul Beier; Faith M. Walker; Myron Shekelle
Eastern tarsiers (Tarsius tarsier complex) are small nocturnal primates endemic to Sulawesi Island and small adjacent islands of Indonesia. The taxon presently includes 11 named species, some of which are listed as Vulnerable, Endangered, and Critically Endangered by IUCN. Several lines of evidence – including the formation of Sulawesi by volcanic and tectonic activity, temporary fragmentation during Pleistocene interglacial periods, and biogeographic and genetic relationships among other species – suggest that at least 5 additional species of eastern tarsier have yet to be described. Efforts to identify these cryptic species are urgently needed so that habitat conversion, pet trade, and cultural activities will not render some species extinct before they are recognized.We will gather data that could support or rule out the existence of cryptic tarsier species on three volcanic islands in Bunaken Marine Park, North Sulawesi, namely Bunaken, Manadotua, and Mantehage, on May-August 2018. We will record tarsier duet calls and use these calls to follow tarsiers to their sleeping sites where we will capture them, collect saliva by buccal swab, take small ear biopsies, and promptly release them. We intend to build a phylogeny by comparing the DNA sequences of the island tarsiers to DNA sequences at Mt. Tumpa on the North Sulawesi mainland (the type locality for Gursky’s spectral tarsier, which has not yet been genetically sampled), and previously published sequences from 14 sites throughout Sulawesi. Each island is known to have a tarsier population and qualifies as a distinct biogeographic setting. Although the Bunaken population may have recently been introduced by humans, the populations on the other volcanic islands are thought to be natural.This study will provide IUCN and the conservation community with strong, new information needed for tarsier conservation, will advance understanding of biogeography of Sulawesi, and will contribute to Indonesian awareness of biodiversity.
29 Impacts of a Rapidly Changing Landscape on Brown Bear and Roosevelt Elk Habitat Use and Movements on Afognak Island, Alaska.
Shannon P. Finnegan
Current timber harvest regimes on Afognak Island, Alaska are leading to large changes of habitat structure and availability, impacting the island’s resident animals such as brown bear Ursus arctos middendorffi and Roosevelt elk Cervus canadensis roosevelti, which are species of important economic and conservation value in the region. I seek to understand how extrinsic and intrinsic factors affect habitat use and resource selection of these two species under an optimality movement paradigm. I will investigate the energetic costs of locomotion associated with changing landscape gradients to understand how areas of timber harvest affect these species movements. In addition, few studies have investigated the role of brown bears as top predators in systems where other large carnivores are absent, I will seek to answer whether this species is a major predator of elk in this system and how they may also be impacting elk habitat use. Over 40 elk and 80 bears will have been fitted with GPS collars and accelerometers after this upcoming 2018 summer field season. Data analysis will consist of species movement modeling and energetic expenditure mapping. Elk scat samples will be collected and tested for nitrogen levels to infer forage quality and an isotope analysis will be conducted with bear tissue samples to infer diet information. This project will lead to novel developments in our understanding on how timber harvest regimes impact upon species movements and physiological responses, and the role of brown bears as top predators in an island ecosystem. The study will develop our knowledge on brown bear and elk habitat use in rapidly changing environments, which will be crucial for informing management and long-term conservation on the greater Kodiak Island Archipelago. Finally this project will develop novel insights on large scale mammal movements, and how varied energetic costs of movement affect individual habitat selection.
30 Impacts of Mountain Pine Beetle Epidemics on Habitat Selection By Moose
Forest P. Hayes; Chad J. Bishop; Joshua J. Millspaugh; Eric J. Bergman; Ragan M. Callaway
Understanding moose (Alces alces) habitat selection and their ability to adapt to habitat alteration is fundamental to successful conservation and management of the species. Breeding populations of moose in the United States are found as far south as Colorado and Utah. At the southern edge of their range, thermal stress is thought to be a larger factor in habitat selection than at more northern latitudes. The availability of thermal cover is impacted by mountain pine beetle (Dendroctonus ponderosae) epidemics which have dramatically impacted forests across the western United States over the past two decades. Despite large scale habitat impacts, little is known about impacts of mountain pine beetles on wildlife. Impacts on ungulates such as moose is even more poorly studied. Our objective is to determine the impacts of mountain pine beetles on habitat selection by moose. We are examining habitat selection by moose using data from GPS collared animals as part of a larger ongoing study in Colorado. We are using a hierarchical habitat model to compare landscape covariates at used and randomly selected available points at population and individual scales. Physical adaptation of moose to cold climates leads to thermal stress at relatively low temperatures. We hypothesize that moose will be particularly responsive to increased thermal stress in these areas as a result of increased energetic demands. This work provides a foundation for understanding the effects of habitat alteration caused by mountain pine beetles on moose and will help determine the need for evaluating these impacts on other species. Understanding habitat selection in these areas is fundamental to efforts to maintain a healthy population of moose as beetle kill areas are not currently incorporated into conservation or management plans.
31 Influence of Acoustic Monitoring Sampling Duration on Detection of Bat Species in Southeastern Ohio
Andrew T. Merkle; Donald P. Althoff; Karen M. Roberts
The use of acoustic monitoring equipment is becoming commonplace for estimating presence-absence and relative activity of bat species. Passive acoustic monitoring often is the first approach for inventorying bats because multiple units can be deployed simultaneously for multiple nights versus the more labor-intensive mist-netting technique. A key component of such sampling efforts is the length of deployment that is likely to ensure high detection rates if, in fact, a species is present. We rotated Wildlife Acoustics bat detectors across 24 sites from late spring to early fall 2015-2016 in southeast Ohio to determine the influence of sampling duration on detection of individual species. The sites represented a range of low-clutter to high-clutter structural habitats. During 2015 we recorded and identified 3,869 call files to species; for 2016, we recorded and identified 3,346 call files. We detected 8 species with >50 call files: big brown bat (Eptesicus fuscus), Eastern red bat (Lasiurus borealis), evening bat (Nycticeius humerali), little brown bat (Myotis lucifugus), Northern long-eared bat (Myotis septentrionalis), tri-colored bat (Perimyotis subflavus), and silver-haired bat (Lasionycteris noctivagans). Based on our sampling frame of 7-consecutive nights of monitoring per site per rotation, we determined that the probability of detection, p, of >0.8 was almost never achieved with less than 6 consecutive nights of monitoring for 7 of the 8 species evaluated. Only the hoary bat, during 2015, was detected with p=1.0 within the first 4 nights of monitoring at sites. Our results indicate that, for southeast Ohio, limiting monitoring at individual sites to fewer than 6 consecutive nights is likely to significantly overestimate species absence when, in fact, the species of interest is present.
32 Influence of Spatial Alignment on Photographic Detection Rates at Remotely Triggered Camera Stations
Edward Davis; Tim Swearingen; Robert Klaver; Charles Anderson; Christopher DePerno; Jonathan Jenks; Robert Bluett; Chris Jacques
Remotely triggered cameras can provide a cost-effective, non-invasive approach for investigating a variety of natural history and conservation concerns for species that are solitary and occur at low population densities. Trail camera performance is influenced by a wide range of factors, though no studies have rigorously evaluated potential sources of sampling bias (e.g., camera type, relative position of cameras) on overexposure (i.e., capturing the flash of one camera by another) events within paired camera station (i.e., 2 camera traps placed perpendicular to animal travel corridors) designs. We evaluated potential effects of camera type (Browning™ Recon Force, Moultrie™ M-880 Series, Reconyx™ HC 600 Hyperfire) and relative camera position (directly aligned vs. offset from one another [i.e., staggered]) on wildlife photographs recorded and overexposure events across 48 camera stations deployed during summer 2017. Total number of wildlife photographs varied by camera model and alignment (model × alignment interaction, F2,42 = 5.56, P = 0.007); Reconyx and Browning cameras detected more wildlife photographs at aligned camera stations whereas Moultrie cameras more wildlife photographs at staggered camera stations. Further, the number of overexposure events varied (F1,46 = 35.24, P ≤ 0.001) between aligned (mean = 3.56, SE = 0.42, n = 25) and staggered (mean = 0.00, SE = 0.46, n = 23) camera stations. Mean percent overexposure for aligned stations was 5.63 (SE = 1.02, range = 23.91). We documented no overexposure events at staggered camera stations and no difference (F2,45 = 0.05, P = 0.95) in numbers of exposure events across camera types. We recommend that future use of paired camera stations for research, inventory, or monitoring of elusive species consider staggering the placement of cameras to minimize overexposure events of target species. Further, wildlife managers should consider evaluating seasonal effects (i.e., winter) on overexposure rates in paired camera station sampling designs.
33 Influences on Foraging Preference Development in an Endangered Species: Implications for a Novel Conservation Strategy
Brigit D. Harvey; Debra Shier; Greg Grether
Captive breeding and reintroduction programs face several challenges as a result of having limited knowledge about the target species’ nutritional ecology, such as the development of unintentional dietary preferences for captive diets and consequential ineffective foraging ability upon release. Additionally, historic landscapes are changing due to the pervasiveness of invasive species that threaten to replace native food sources of endangered populations. A novel management strategy to counteract these two factors is to expand the diet of a target species to also include palatable and nutritious invasive species, thereby increasing the foraging options of target species upon release and decrease the spread of invasive species being consumed. Utilizing the captive breeding program for the endangered Pacific pocket mouse (Perognathus longimembris pacificus, PPM), I am experimentally determining if historic exposure, exposure during early developmental periods, and nutritional quality of 2 invasive plants, Erodium botrys and Bromus rubens, influence foraging preferences of PPM. Preferences are tested using the standard Cafeteria Method design and the nutritional quality of seeds is determined with NIRS analysis. This study will impact nutritional protocol and foraging training of PPM and insights from this management strategy can be directly useful for other captive breeding programs with a similar conflict.
34 Landscape Ecology of the Ocelot in South Texas
Amanda M. Veals; AnnMarie Blackburn; Michael E. Tewes; John H. Young Jr.; Humberto L. Perotto-Baldivieso; Randy W. DeYoung
Wildlife species across the globe are faced with landscapes that are becoming more impermeable to movement as the result of habitat loss and degradation. Conservation of landscape connectivity is therefore essential to sustain wildlife populations, support animal movements, and maintain genetic heterogeneity. Anthropogenic factors, such as urban sprawl and roadways are a major cause of habitat loss. Roadway networks are widespread and significant anthropogenic influences on the landscape that can have profound impacts on wildlife populations. The ocelot (Leopardus pardalis) is an endangered felid in the United States, with remnant populations of < 80-100 individuals in southern Texas. The Lower Rio Grande Valley of southern Texas is one of the fastest growing population centers in the United States. This has led to an increase in road collisions and a decrease in available habitat for ocelots. We aim to model probability of landscape use by ocelots in southern Texas as a function of woody vegetation structure and road permeability at the population level. We will model resource selection from animal location data. In conjunction, we will use circuit theory to identify potential long-distance movement pathways to assess which resistance scenario best represents the modeled habitat use. Our project aims to inform the placement of future road crossing structures to decrease ocelot-vehicle collisions and increase landscape permeability for this endangered species.
35 Large Mammal Distributions Relative to Human Settlements Bordering an Important Wildlife Sanctuary in Borneo
Olivia G. Cosby; Belden Giman; Roslina Ragai; Timothy R. Van Deelen; Terilyn D. Allendorf; William J. McShea
Understanding drivers of habitat selection at the boundaries of protected areas is critical to mitigating human wildlife conflict. In Borneo, remote communities reliant on hunting and harvesting forest products often border protected areas. Because Southeast Asian forests have irregular (3-10 years) synchronized fruiting events, forest wildlife near human settlements may benefit from increased availability of annual and continuously fruiting trees and palms typical of more disturbed secondary forests, but increase their risk of being hunted. This 3-year project examines distributions of large mammals in primary and human-impacted secondary forests within and adjacent to Lanjak Entimau Wildlife Sanctuary (LEWS), in Sarawak, Malaysia. Our goal is to evaluate wildlife shifts between primary and secondary forests relative to annual variation in fruit and seed production and human disturbance. Using 50 remote-sensing cameras and bimonthly phenology surveys, we are tracking fruit and seed production, and large mammals foraging under fruit trees, through 3 annual cycles. Our research focuses on species dependent on fruit consumption as well as predators dependent on small mammals. Here we present our first year of data, relating large mammal shift with phenological patterns observed in fruit trees from 50 vegetation plots (1260 m2 each). Since October 2016, we have documented 44 mammal species, many of whom are vulnerable or critically endangered, including orangutans (Pongo pygmaeus; 30 detections), sun bear (Helarctos malayanus; 111 detections), clouded leopard (Neofelis diardi; 9 detections), Bornean bay cat (Catopuma badia; 3 detections), Hose’s civet (Diplogale hosei; 24 detections), white-fronted langur (Presbytis frontata; 5 detections), and pangolin (Manis javanica; 33 detections). Preliminary findings contribute much needed baseline data for LEWS, a global hotspot for biodiversity. This information will be used to develop best practices for the sustainable use of forest resources with the goal of mitigating risks to species of concern.
36 Muskrat Movements, Survival and Disease
Laken S. Ganoe; W. David Walter
Declines in muskrat (Ondatra zibethicus) harvest has been observed across the United States and in several Canadian provinces, with many states recording larger than 50% decline in muskrat harvest within the past few decades. Harvest rates are historically used as indicators of population status, thus declines in muskrat harvest indicate a decline in the muskrat population. Furthermore, disease has been considered a cause for concern for muskrat population declines, specifically Leptospirosis, Toxoplasmosis, as well as other diseases and parasites. This study was designed to investigate the movements and population status of muskrats in a region of Pennsylvania. To understand muskrat movements, preliminary home range sizes have been estimated for adult muskrats within various habitat types using radio telemetry and capture locations. We used mark-recapture data to determine survival estimates for the first phase of the project. We also analyzed blood from live-trapped muskrats and organs from trapper-harvested muskrats for exposure to various diseases. Expanding the knowledge of muskrat movements, creating baseline population estimates and survival rates, and understanding the exposure of muskrats to different diseases will increase the knowledge of the muskrat and the potential reasons for their population decline.
37 Novel Insights Into Carnivore Biology: Bacterial Gut, Skin, and Oral Microbiomes of Free-Ranging Bobcats
Claudia Wultsch; Angela Benton; Gabriella Sosa Medina; David McNitt; Robert Alonso; Marcella J. Kelly; Konstantinos Krampis
Microbiomes have profound effects on the health, physiology, behavior, and development of their animal hosts. Despite the importance of microbes, little is known about microbiomes in wild carnivore populations. Here, we applied high-throughput 16S rRNA sequencing to investigate gut, oral, and skin microbiomes of free-ranging bobcats (Lynx rufus), a medium-sized felid widespread across North America. Four bobcats were captured in Bath County, VA and microbiome samples were collected noninvasively from sedated animals via swab samples. A total of 412,767 sequences were generated for 11 samples (gut, GM, n = 4; oral cavity, OM, n = 3; skin, SM, n = 4) to examine microbial community composition, diversity, and structure. We found compositional differences in microbiomes across the three body sites with microbiota being dominated by the phyla Bacteroidetes (GM, 24.8%; OM, 25.2%; SM, 20.2%), Firmicutes (GM, 21.9%; OM, 4.4%; SM, 26.2%), Fusobacteria (GM, 32.8%; OM, 28.1%; SM, 7.6%), and Proteobacteria (GM, 16.5%; OM, 36.5%; SM, 27.1%). Interestingly, the skin microbiome was significantly more diverse than the oral and gut microbiomes (Kruskal Wallis test, H = 7.05, P = 0.029). Ordination analysis by non-metric dimensional scaling confirmed that bacterial communities on bobcats are distinct and body-site specific (PERMANOVA, t = 4.71, P = 0.001). We also detected high inter-individual variation in the microbial communities, such that individual bobcats have their own microbial fingerprints. This study provides an alternative perspective on carnivore biology and offers novel baseline information on bobcat microbiomes, which holds vast potential aiding ongoing and future conservation and management efforts.
38 Optimization of Camera Trapping Methods for Surveillance of Eastern Spotted Skunks and Mesopredator Communities in the Appalachian Foothills
Courtney R. Hayes; Kelly Watson; Luke E. Dodd
The global decline of apex predators has allowed mesopredator populations to increase, a phenomenon described by the mesopredator release hypothesis (MRH). Some mesopredator species, however, are of conservation concern, such as the eastern spotted skunk (Spilogale putorius), whose populations have noticeably declined in the past 40 years. To assess camera deployment strategies and survey for the presence of eastern spotted skunks in the Appalachian Foothills, we conducted baited camera trap surveys in Kentucky, a state for which systematic methodological data is lacking. We surveyed 64 sites across 10 counties over more than 1,200 trap days from October 2017 to April 2018, focusing on features associated with mesopredator presence. We detected approximately 400 individual mesopredators of 8 different species, including 3 records of eastern spotted skunks. We evaluated effects of bait type (sardines vs. sardines + fatty acid scent tablet) and deployment duration (2 week vs. 4 week) by comparing mesopredator activity per trap day, species richness per trap day, and species accumulation curves across deployment strategies. We found no significant differences in the effect of bait type nor deployment duration on mesopredator detection rates per trap day or species richness per trap day (P > 0.05). However, accumulation curves tended to reach asymptote more quickly in the duration of the deployment in traps using sardines and scent tablets than in traps using only sardines. Our results suggest the most efficient method (2-week duration with no fatty acid scent tablet) yielded results comparable to more time and resource-intensive options, however species-specific trapping rates need to be taken into consideration when choosing deployment duration as eastern spotted skunks were not recorded in our study until 6-21 days after deployment. The results of this study will allow for targeted, expanded, and more efficient camera trapping efforts throughout the Appalachian Foothills in upcoming studies.
39 Pellets Or Pictures, Which Would You Prefer to Count? a Comparison of Two White-Tailed Deer Population Survey Techniques
Jonathan A. Matthews; Matthew T. Springer; John J. Cox
The ability to accurately measure white-tailed deer (Odocoileus virginianus) population density is a valuable tool for wildlife managers; however, generating accurate estimates can be challenging. Due to varying habitat quality, quantity, and various other external factors, population densities can vary drastically from one area to another. This presents a distinct challenge when dealing with making localize management decisions based on landscape level estimates, so effective and efficient localized population estimates are required to aid in the decision making process. Additionally, many common estimation techniques (i.e. helicopter surveys, FLIR surveys) are expensive to conduct or complicated to perform/analyze, and thus may not be an option. Trail camera surveys of wildlife populations have shown to provide cost-effective estimates with less complex analysis. Furthermore, trail cameras can be used in a multitude of areas where other methods may be unavailable (i.e. urban environments). We sought to test the reliability of a trail camera population estimate method by comparing it to a statistically robust distance sampling method. We used the Jacobson et al. (1997) trail camera method for estimating white-tailed deer populations to estimate the deer populations on 11 farms in western Kentucky during the summers of 2017 and 2018. Concurrently with the trail camera surveys, we performed a pellet based distance sampling method on all farms. Trail camera surveys were analyzed following Jacobson et al. (1997) while distance sampling results were analyzed using Program Distance 7.1. We plan to present results of the analyses from both 2017 and 2018 at the conference. Results from this comparison will provide managers an assessment of accuracy and time relative to obtaining a localized population estimate, which will help managers choose a population estimate technique that best fits their needs and resources, and aid in the management of localized wildlife populations.
40 Persistence in the Face of Change: Effects of Human Recreation on Coyote Habitat Use in an Altered Ecosystem
Morgan J. Morales; Maximilian L. Allen
Human activity and development has led to extirpation of apex carnivores and the creation of altered ecosystems in many regions. As a result, mesocarnivores, such as coyotes, are shifting to occupy the apex carnivore role in some areas. It is important to understand the habitat needs and ecological roles of these new, adaptable apex carnivores, especially in altered and human-dominated ecosystems. We investigated the relationship of coyotes (Canis latrans) with environmental and human recreation variables in an altered eucalyptus ecosystem (San Francisco, California, USA) using camera traps and vegetation surveys. We documented coyotes at eight of our nine camera trap sites, with an average relative abundance index of 86.41 coyote events per 100 days. We tested a set of a-priori models to potentially explain coyote spatial use with generalized linear mixed effect models. Our top model was ‘Humans and Cover’ (wAICc=0.98), which included tree density, abundance of human walkers, joggers, and bikers as explanatory variables. We also tested whether coyotes were temporally avoiding humans, and found low temporal overlap between humans and coyotes (Δ4=0.40). Human activity within the park only ceased between 11 pm and 5 am, which resulted in coyote peak activity occurring around 4 am when no humans were present. This appears to be a shift in coyote temporal activity to avoid humans, as coyotes typically exhibit crepuscular activity cycles. Our results indicate that coyotes were able to adjust to the altered ecosystem, and instead their spatial use and temporal activity was most affected by human recreation.
41 Persistence of Threatened Northern Long-Eared Bats on the Island of Martha’s Vineyard, Ma Amidst a Regional Population Decline From White-Nose Syndrome
Luanne Johnson; Elizabeth Baldwin; Zara Dowling; Jonathan Reichard; Susanna vonOettingen
The federally threatened Northern Long-eared Bat (Myotis septentrionalis) (MYSE) is persisting in multiple locations on Martha’s Vineyard, Massachusetts, despite a confirmed bat death from white-nose syndrome (WNS) in February 2017 and widespread regional population declines. We used a combination of acoustic monitoring, emergence counts, mist netting, and radio tracking to locate 3 summer colonies in 2015 and monitor them in 2016 and 2017. During the summer maternity season in 2015, we captured and radiotagged 8 reproductive females that occupied house and tree roosts and a single female at another site that occupied tree roosts. We confirmed a third maternity colony at a house roost through mist net capture of a volant pup. Summer monitoring in 2016 and 2017 confirmed activity at these colonies and telemetry identified a fourth colony area. We recaptured a banded female at a maternity colony in 2016 and one in 2017 at a different colony, which verified overwinter survival. To monitor for migration behavior in the fall, we radiotagged 2 females in Sept. 2015 and 1 in October 2016, and tracked them to day roosts for 14 – 24 days. These bats remained on the island during monitoring, and one was roosting in a tree as late as 7 November 2016. This fall roosting behavior, coupled with four opportunistic observations of MYSE in February or March at structures, indicated that some MYSE hibernate on the island. Through extensive community outreach, we were successful in locating a hibernacula in a cement cellar in December 2017. We hypothesize that MYSE overwinter in small groups in multiple anthropogenic hibernacula on Martha’s Vineyard, which may provide some protection from the most severe impacts of WNS.
42 Population Demographics of the North American River Otter in Ohio after 13 Years of Harvest.
Sara Adamczak; E. Hance Ellington; Stanley Gehrt
Population demographic information is essential for management of furbearers subjected to harvest. In Ohio, the river otter (Lontra canadensis) was reintroduced during 1986-1993, and a restricted harvest was implemented in 2005. Demographic parameters were determined from carcasses during 2005-2008 harvest seasons, but no demographic information over the past decade has been recorded. Our goal was to determine the population demography of the otter population following a decade of harvest, and compare the current demographic pattern to the population at the time harvest was initiated. Male otter length seems to have a positive trend over time, whereas their weight has a negative (2005-2007: 108.2 cm [n = 207, SD = 8.3], 2016-2017: 109.6 cm [n = 58, SD = 3.9], and 2005-2007: 8.6 kg [n = 235, SD = 1.9], 2016-2017: 7.1 kg [n = 58, SD = 1.6]). Harvested females show a similar trend (2005-2007: 103.9 cm [n = 190, SD = 7.4], 2016-2017: 104.82 cm [n = 41. SD = 3.4] and 2005-2007: 7.3 kg [n = 216, SD = 1.7], 2016-2017: 6.1 kg [n = 41, SD = 1.2]. The sex ratio of the harvested population has maintained a male biased (2005-2007 [239 males, 225 females], 2016-2017 [58 males, 41 females]). Using cementum annuli analysis we found that from 2005-2007, 60% of the harvested individuals were juveniles (< 1 year old), 24% were yearlings (1 - 2 years old), and 16% were adults (> 2 years old). In 2016, 40% of the harvested individuals were juveniles (< 1 year old), 51% were yearlings (1 - 2 years old), and 9% were adults (> 2 years old). These metrics will form the basis for a population model to assess the influence of harvest on population growth.
43 Population Estimate of White-Tailed Deer in the Cleveland Metroparks Based on Microsatellite Genetic Diversity
Karen E. Munroe; Johnny Donovan; Savanah Craig; Pam Dennis
White-Tailed deer (Odocoileus virginianus) populations have dramatically increased in Ohio and throughout the US in the past 50 years. Current management techniques within the urban Cleveland Metropark system mainly rely on culling by trained sharp-shooters based on traditional population size estimates from aerial surveys. The goal of this ongoing study is to use the microsatellite diversity of a White-Tailed deer population to gain a comparative and possibly more accurate estimate of the population size to use in future management decisions. Coalescent theory and associated computer models (e.g. LAMARC) are often used to determine the demographic history of a population including migration and population size. DNA samples were extracted from muscle tissue samples collected from White-Tailed deer culled between 2011-2015. The extracted DNA was amplified using eight microsatellites designed for other ungulates and optimized for this population. Allelic diversity of the eight primers ranged from 5-29 alleles and did not significantly vary between years. Overall, relatedness between culled individuals was statistically significant and yet similar within and between years (r = 0.09-0.12). Overall and individual marker Fst values between years were low (Fst ≤ 0.05) suggesting minimal genetic differentiation between years. Our results show that even though our genetic samples (i.e. individuals) were being removed from the population when they are sampled, that genetic diversity remained constant between years, possibly suggesting that the population size between 2011 and 2015 is remaining constant. However, additional primers are needed to ensure detection of any potential genetic differences between years.
44 Population Genetic Demography of the Tricolored Bat and Implications for the Impact of White-Nose Syndrome
Alynn Martin; Maarten Vonhof; Amy L. Russell
The recent emergence of threats to North American bat conservation has prompted increased population genetics research on high risk species. The tricolored bat is affected by both white-nose syndrome and wind turbine mortality. However, little work has been done regarding the population structure and effective population size of this species. Using mitochondrial sequence and nuclear microsatellite data, we analyzed male and female structure across the sample range of Perimyotis subflavus and estimated the effective population size of their populations. Pairwise FST values indicate that there is one panmictic population based on microsatellite data, while mitochondrial data supports two populations within the sampled range. AMOVA results suggest that females are making short distance movements. These data yield contrasting results for effective population size and size change over time. Mitochondrial data suggest an increase in female effective size for both Appalachian and West populations since the last glacial maximum, while microsatellite data suggest a recent bottleneck. The persistence of the eastern pipistrelle is dependent upon the maintenance of genetic diversity, and calls for the conservation of genetically distinct populations as well as the preservation of hibernacula and swarming locations.
45 Regulations, Restrictions, and Habitat: How Management Shaped the Distribution of Trophy White-Tailed Deer in the Midwest
Rebecca L. Cain; David M. Williams; William F. Porter
Wildlife agencies use white-tailed deer (Odocoileus virginianus) harvest records to monitor the deer population, and these data are used to inform management decisions. Unfortunately, managers lack information comparing the effectiveness of different management strategies at producing high-quality deer. The impact management decisions have on the large, mature males in the population may often go undetected because of the focus on performance indicators of deer population trends and hunter satisfaction. Our objective was to investigate how regulations, hunting restrictions, and habitat shape the distribution of trophy-sized deer. We collected data from the Boone and Crockett records and equivalent record keeping programs of Midwestern states to investigate the variation in trophy white-tailed deer abundance and distribution. We used an N-mixture modeling framework to account for spatial and temporal changes in detection probability and to evaluate factors affecting annual estimates of the number of trophy deer in each county of our study area. We investigated covariates that are important to managers but accounted for variation resulting from environment and climate differences. Our results identified factors that influenced the probability that a trophy deer is harvested in a county. Detection probabilities varied across space with Buffalo County, Wisconsin having the highest probability (p = 0.55). Our findings suggest that season lengths and timing are influential factors in the harvesting of trophy white-tailed deer. The focus on covariates related to management strategies and regulations and the broad spatial extent of our analysis has important implications for policy decisions about deer management.
46 Relating Wolf and Coyote Kill Sites to Resource Selection for White-Tailed Deer in Northern Wisconsin.
Lucas O. Olson; Timothy R. Van Deelen; Daniel Storm; Shawn Crimmins
The trophic impact of wolves on white-tailed deer (Odocoileus virginianus) and forest plants has been studied throughout parts of the Northern Great Lakes Region and Western United States. Little research, however, has been applied to understanding the impact of top predators on agriculture. Our objective is to understand factors influencing white-tailed deer resource use within a forest-agriculture system, and to link patterns of resource selection to the risk of mortality by two sympatric large carnivores. Our study area falls within Ashland, Rusk, and Sawyer counties in Northern Wisconsin where wolves (Canis lupus) and coyotes (Canis latrans) have been well established. 411 white-tailed deer were monitored for cause-specific mortality between 2011-2014 using VHF radiocollars. Monitoring provided 28 wolf kill sites, 35 coyote kill sites, and 17,436 unique telemetry locations. We will use resource selection functions (RSF’s) in a modelling framework to infer resource use patterns among factors influencing selection including forage availability, proximity to crop agriculture, proximity to livestock, terrain features, and risk of mortality by wolves and coyotes. Competitive interaction between two sympatric carnivores may have implications for the risk landscape for white-tailed deer in relationship to crop and livestock agricultural operations. Understanding factors shaping white-tailed deer resource use patterns in forest-agriculture systems is important for management, agricultural revenue, and for grasping dynamic ecosystem relationships.
47 Resource Selection and Movements of Female Mule Deer and White-Tailed Deer during Parturition and Lactation in Western Kansas.
Talesha Karish; David Haukos; Andrew M. Ricketts; Levi Jaster
The abundance and occupied range of mule deer (Odocoileus hemionus) in Kansas have been declining for twenty years. The two predominant hypotheses for the loss of mule deer and concurrent expansion of white-tailed deer (O. virginianus) are changes in land use and competitive dominance of white-tailed deer over mule deer. Despite the popularity and income that stem from hunting revenue, there have been no recent studies that provide critical insight on how to improve management and conservation of the sympatric populations in Kansas. Our objectives were to evaluate seasonal survival rates, cause-specific mortality, resource selection and movements of adult female mule deer and white-tailed deer in western Kansas. We used helicopters to capture and GPS-collar 30 female mule deer and 30 white-tailed deer split evenly between two study sites. Each deer was fitted with a high resolution GPS/VHF collar that recorded bi-hourly locations with an activity sensor to identify mortality events. Each collared deer also received a Vaginal Implant Transmitter digitally linked with the collar to recorded and send an alert for the exact time of parturition. Dominant land-cover (native grassland vs cropland) differed between the study sites and therefore, we predict that resource selection will also differ. We used ArcGIS to measure temporal movements and Brownian Bridge Movement Models to estimate seasonal home ranges. Average daily movements and home range sizes of females were influenced by reproductive stage of does. We used logistic regression to determine habitat selection of birth sites. We expect females will select areas of greater cover for birth sites to reduce predation. Parturition and lactation are energetically demanding of females and their resource needs change. We used logistic regression to compare foraging locations of lactating and non-lactating females throughout the season. This information will be used to improve management of both species in Kansas.
48 Seasonal Survival and Movements of Male Mule Deer and White-Tailed Deer in Western Kansas
Maureen Kinlan; David Haukos; Andrew Ricketts; Levi Jaster
The abundance and occupied range of mule deer (Odocoileus hemionus) in Kansas have been declining for twenty years. The two predominant hypotheses for the loss of mule deer and concurrent expansion of white-tailed deer (O. virginianus) are changes in land use and competitive dominance of white-tailed deer over mule deer. Despite the popularity and income that stem from hunting revenue, there have been no recent studies that provide critical insight on how to improve management and conservation of sympatric populations of both species in Kansas. Our objectives were to evaluate seasonal survival rates, cause-specific mortality, and movements of adult male mule deer and white-tailed deer in western Kansas. We aerially captured and GPS-collared 60 male mule deer and white-tailed deer at two different study sites. Dominant landcover (native grassland vs cropland) differed between the study sites. Each deer was fitted with a high resolution GPS/VHF collar that recorded bi-hourly locations and used an activity sensor to identify mortality events. Known fate models were used to evaluate landscape factors affecting survival and estimate seasonal survival rates. We used ArcGIS to measure temporal movements and Brownian Bridge Movement Models to estimate seasonal home ranges. Survival of male mule deer and white-tailed deer during spring and summer is high. Average daily movements and home range size are influenced by season and landscape. Inclusion of another two years of data will add to the information available to natural resource managers for management of both species in western Kansas.
49 Sexually Dimorphic Differences in the Brain Transcriptome Profile in Amami Spiny Rats (Tokudaia Osimensis): A Rodent Species Where Males Lack the Y Chromosome and Sry
Madison T. Ortega; Sarah A. Johnson; Nathan Bivens; Takamichi Jogahara; Asato Kuroiwa; Scott A. Givan; Cheryl S. Rosenfeld
Within therian mammals, females are generally homozygous (XX), while males are heterozygous (XY). However, Amami spiny rat (Tokudaia osimensis), has neither the Sry gene nor the Y chromosome that direct the undifferentiated gonad to become a testes and result in an increase in testosterone during fetal life that permanently masculinizes the brain. The Sry gene is located on the Y chromosome and is primarily responsible for male sexual differentiation. Both males and females of this species only have a single X chromosome, resulting in an XO genotype. It remains unclear how sexual differentiation occurs in this species. By examining the brain transcriptomic profile in both sexes this species, it will provide better understanding of genes or transcripts that might compensate for absence of the Y chromosome and Sry. The hypothesis being tested is that such genes likely reside on autosomal chromosomes or the X chromosome. Thus, we examined the global brain transcriptomic profile from male and female Amami spiny rats, and identified several genes and select transcripts that showed sexually-dimorphic expression differences. By using quantitative Real Time Polymerase Chain Reaction, we validated that Svs5 and Serpina1b were elevated in the brain of males and females, respectively (p ≤ 0.05). It remains to be determined whether some of the gene and transcript expression differences identified in adult male and female spiny rats might compensate for the absence of Sry and serve as the initial drivers of brain sexual differentiation. These findings could be useful in recovery efforts for this endangered species.
50 Simultaneous GPS Tracking of Hunters and Black-Tailed Deer: Detecting Real-Time Behavioral Responses of an Important Game Species
Alex McInturff; Kaitlyn Gaynor; Justin Brashares
Every year, millions of people in the United States become active participants in wildlife communities by hunting deer. While the consequences of hunting on deer population dynamics have been well studied, we know surprisingly little about how hunting shapes patterns of wildlife movement and behavior. Recent developments in GPS technology now allow the collection of very high-resolution movement and activity data that can shed light on these questions. At the 5,300 acre Hopland Research and Extension Center in northern California, I have collected data from 360 hunters with wearing GPS units set at 5-second intervals. In collaboration with the California Department of Fish and Wildlife (CDFW), I have also fitted 10 legally huntable black-tailed deer (O. hemionus columbianus) bucks and 25 does with GPS collars. Collection of GPS data from both hunters and deer will support three major analyses. First, I will examine whether and how deer alter their behavior and movement during the course of the hunting season. Hunters and wildlife biologists have long suggested that deer respond to – and may even anticipate – hunting seasons by changing behaviors and contracting their ranges. However, a rigorous analysis of how deer alter their behavior, the spatial and temporal extent of these changes, and their attenuation after hunting season will provide important information to biologists and managers. Second, by collecting spatial data on both human predators and prey, I will be able to examine the behavioral responses of deer to hunters in real time at an unprecedented resolution. Finally, the spatial data collected from hunters and their kill sites will support development of a statewide spatial model of hunting pressure that will help CDFW improve the precision and application of its game management strategies.
51 Social System Shapes Differences in Social Behaviors of Equids
Sarah R B King; Kathryn A. Schoencker
Horses (Equus ferus caballus) and burros (E. africanus asinus) have transported humans throughout our civilization. Although they are often considered to be similar, they are descended from species with different social systems and adaptation to aridity. These ancestral differences are likely to be evident in feral species, yet burros on public lands are largely managed in the same way as horses. We used the same methods to examine horses and burros in a free-roaming setting, enabling us to compare the two species directly. We conducted observations at a horse herd management area (HMA) and a burro HMA in western and central Utah respectively. Individual animals were identified with radio collars, freeze marks, and individual color markings, and we located study animals with radio telemetry. We conducted behavioral observations on burro groups in 2016 and 2017, and on horse groups in 2017. Horses and burros had different social groupings: horses were in groups of females and one male, whereas burros tended to be in groups of only females, or females with one or many males. Horses and burros had similar feeding and standing rates, but burros lay down more, and moved less than horses. Burros exhibited more affiliative and agonistic behaviors, and vocalized more than horses. The difference in how horses and burros cluster on the landscape mean that they may have different impacts, and their different behavior means that different management approaches could be considered. Our ongoing study has implications for how these animals are managed by the BLM and could have wider applications for the conservation of threatened equid species.
52 Spatial and Temporal Analysis on White-Tailed Deer Movement Activity Relative to Environmental Conditions
Sandor Dibble; Jonathan Fleming
Patterns of daily movement activity and behavior of ungulates are often difficult to quantify. Camera traps have become a useful method of observing wildlife behavior and have advantages over other common field survey methods by being less invasive and less costly. This ongoing research uses data from 81 camera traps located approximately 200 meters apart in a systematic grid pattern design across 220 hectares of mixed managed forest at Andrews Forestry and Wildlife laboratory in Oktibbeha County, Mississippi. Data for this study includes occurrences of white-tailed deer (Odocoileus virginianus) collected and acquired from the Mississippi State Deer Lab at Mississippi State University from January 1, 2015 to December 31, 2016. Camera locations along with recorded white-tailed deer occurrence data are being used to conduct a geospatial analysis on the active movement periods of white-tailed deer relative to daily environmental conditions including temperature, precipitation, wind speed, moon phase, and time of day. Accumulated counts of deer occurrences are grouped into four-time classes for each 24hr period. A generalized linear mixed effects model for Poisson distributed data accounting for spatial structure and repeated occurrences among cameras is being used in this analysis to test hypotheses regarding the relationship of occurrences and movement activity to daily environmental conditions. Results of this study will provide insights into white-tailed deer movement behavior that will contribute to management and conservation of this culturally and economically important species.
53 Spatial Co-Occurrence of Andean Bears with Puma and Domestic Dogs in Ecuador
Vanessa L. Springer; Angela K. Fuller; Evan G. Cooch
The Andean bear (Tremarctos ornatus) is threatened across its range in South America and endangered in Ecuador, where habitat is increasingly fragmented by conversion of forest for expanding development, grazing and agriculture. The Andean bear is considered an umbrella species for conservation planning, although little is known about its ecology and distribution throughout Ecuador. Andean bear and puma (Puma concolor) are two of the largest native mammal species in the cloud forest in Ecuador, both with large area requirements. Although their diets differ and therefore they are not likely to compete for resources, it is possible that these species practice avoidance of one another via spatial segregation. Similarly, Andean bears may avoid dogs (Canis familiaris) and be displaced from areas where nonnative domestic or feral dogs are present. We conducted a large-scale camera trap survey (101 cameras surveyed July-November 2016) within and outside of the Andean Bear Ecological Corridor in the Metropolitan District of Quito. We examined spatial co-occurrence patterns between Andean bears and two other species, puma and domestic/feral dogs using a two-species conditional occupancy modeling framework. We discuss how knowledge of these co-occurrence patterns can help inform conservation and management of Andean bears.
54 Spatial Ecology and Citizen Science Monitoring of Red Panda in Langtang National Park, Nepal
Manoj Bhusal; Lisette P. Waits; Sophie Gilbert; Ryan Long; Dennis Becker; Sabita Malla; Samundra A. Subba; Saroj Koirala
The Red Panda Ailurus fulgens is an endangered mammal, endemic to the temperate and sub-alpine forests in the Himalaya. Threats including habitat fragmentation, deforestation and poaching have caused rapid decline in the global population size by 50% in the last 54 years. Despite this, very few studies on species distribution and ecology have been conducted. This research aims to address the question of how red pandas are spatially distributed within Langtang National Park (LNP) in Nepal and to evaluate the relationship between species presence and specific habitat variables. Field surveys were conducted from May-Aug 2016 by training local yak herders and youth as ‘Citizen Scientists’ and involving them in data collection. Data on species presence/absence and habitat preferences was collected in 403 sample locations and 110 vegetation plots within 36.6 km line transects (n=37) covering 370 sq. km grids (47.4% of the potential red panda habitat in LNP). A naïve occupancy estimates of 0.4054 was obtained using PRESENCE™ software, and relative abundance was estimated to be 1.23 signs/km. Red panda presence was also associated with elevation, slope, proximity to water, presence of cover trees, and bamboo in LNP. Further analysis will provide estimates on individual site-based occupancy, associated habitat variables and conservation threats at a finer spatial scale. This research is crucial in fostering new scientific information on red panda distribution and ecology, evaluating community based red panda monitoring methods as a citizen science approach, and providing recommendations for the habitat management in LNP.
55 Spatial Ecology of Gray Wolves in the Chernobyl Exclusion Zone
Michael Byrne; Sarah Webster; Stacey Lance; Cara Love; Thomas Hinton; Peter Schlichting; Dmitry Shamovich; Valery Dombrovsky; James Beasley
The Chernobyl Exclusion Zone (CEZ) is a ~4300 km2 area in Belarus and Ukraine that remains heavily contaminated with radiation from the nuclear accident of 1986. As such, human residency remains extremely sparse, and the CEZ has become a refuge for some populations of wildlife. Little is known about the spatial ecology of wildlife populations in the CEZ however, including the potential influence of contamination on movements and habitat use. Gray wolves (Canis lupus) are one species which appear to have benefited from the lack of human disturbance, with estimated population densities in the CEZ exceeding those of uncontaminated reserves in the region. Our objective was to use GPS telemetry to understand the influence of contamination on wolf spatial ecology within the CEZ. We tracked fourteen wolves (5 male, 9 female) with GPS collars from November 2014 – May 2015 (n = 8), and November 2016 – August 2017 (n = 6). All adult wolves (n = 13) maintained ranges within the CEZ for the duration of their respective tracking periods, suggesting movements outsize the CEZ boundary by resident adults are infrequent. One juvenile collared in 2014 exhibited evidence of long-distance dispersal behavior, and traveled 369 km from its home range center over a 21-day period in February 2015, before contact with the GPS collar was lost. This suggests the CEZ may serve as a source for some wildlife populations, and raises questions about potential interactions with populations in uncontaminated areas. Relatively high soil contamination levels (> 750 kBq/m2) comprised on average 57% (range: 5 – 99%) of resident wolf home ranges, suggesting space use of wolves in the CEZ is influenced by resource availability and social factors rather than the distribution of contaminates. This study represents an important step in understanding the effects of nuclear disasters on large carnivores.
56 Spatial Requirements of Plains Bison in the American West
Robert J. Ritson; Nate Bickford; Dustin H. Ranglack
The future of bison (Bison bison) in North America is far from certain despite consideration as a conservation success. Land-use conflict resulting in reduced space availability remains a primary obstacle for restoring this iconic grassland species. The objective of this study was to calculate the spatial needs and habitat requirements of bison herds throughout their range in order to identify suitable areas for future recovery. The home range sizes of six bison herds across a latitudinal gradient of the American west were computed using an autocorrelated kernel density estimator. These included the Texas State Bison Herd, the Wichita Mountains herd in Oklahoma, the Medano-Zapata Ranch herd in Colorado, the Henry Mountains and Book Cliffs herds in Utah, and the American Prairie Reserve herd in Montana. Habitat variables were fitted with a general linear model to predict home range sizes including NDVI, distance from roads, distance from water, distance from fences, average precipitation, and landscape ruggedness. The Utah herds are free-range while the remaining herds are contained in fenced areas of varying sizes. Free-range bison exploit between 600-2000 square kilometers while captive bison use space equal to their reserve size. Average precipitation, distance from roads, and landscape ruggedness were expected to influence space use most strongly. These factors may be broadly associated with forage availability and grassland habitat. Suitable areas for future free-range bison herds vary by spatial context.
57 Spatially-Explicit Analysis of White-Tailed Deer Reporting in the Boone and Crockett Records
Garrett J. Knowlton; Rebecca L. Cain; David M. Williams
For effective white-tailed deer (Odocoileus virginianus) management in the United States, managers need data that accurately reflect the populations they are managing. Biased data can lead to incorrect conclusions about the status of the population and mismanagement. The Boone and Crockett big game records provide a record of trophy white-tailed deer harvest and antler sizes spanning nearly 100 years. Antler size can indicate a healthy deer herd where bucks are surviving long enough with sufficient nutrition to meet the minimum scores of entry in to the record book. Concern about reporting bias in the Boone and Crockett Club white-tailed deer records has raised questions about the use of that data set. To ensure that the inferences drawn by researchers using these data are appropriate, information about this perceived bias is required. The objective of this research was to evaluate spatial and temporal patterns of reporting bias in the Boone and Crockett records for white-tailed deer harvested in the Midwest. Data of white-tailed deer records from 1973—2014 were collected from the Boone and Crockett Club and the equivalent state record books for Michigan, Ohio, Minnesota, Wisconsin, and Iowa for our analysis. To evaluate the patterns of bias within the Boone and Crockett Club records, we compared the number and location of entries in each record book to the Club’s records. Our analysis suggests that over time, the Boone and Crockett record book and state records books of Iowa, Wisconsin, Michigan, and Ohio are becoming more closely related. Spatially there are significantly clustered areas of underreporting to the Boone and Crockett Club. We discuss the possible implications of areas of underreporting within the Boone and Crockett records, and how the Boone and Crockett records could be a useful tool to white-tailed deer managers.
58 Survival and Cause-Specific Mortality of White-Tailed and Mule Deer Fawns in Western Kansas.
Mitchell Kern; Andrew Ricketts; David Haukos; Levi Jaster
Mule (Odocoileus hemionus) and white-tailed deer (O. virginianus) are common sympatric deer species in the Great Plains and western United States that have been exhibiting divergent population trends temporally and spatially. Mule deer populations are declining and contracting to the west while populations of white-tailed deer are increasing and expanding. Limited research has been conducted in Kansas to understand why two similar species are exhibiting vastly different population trends. We employed helicopter net-gun techniques to capture 60 female deer at 2 independent study sites during February in western Kansas. Each of the 30 white-tailed and 30 mule deer does were collared with GPS radio transmitters and received Vaginal Implant Transmitters (VITs). We anticipated an average of 1.5 fawns per doe and capture of 45 VIT-associated fawns of each species during the spring/summer field season. We monitored fawns daily via ground-telemetry to assess bed-site selection, survival, and cause-specific mortality. We compared fawn survival rates and cause-specific mortality between the two sympatric deer species and assessed landscape level factors influencing survival. Additionally, we evaluated habitat characteristics correlated with fawn survival and cause-specific mortalities in western Kansas to provide insight on the dissimilar population trends between white-tailed deer and mule deer.
59 The Hare-Y Truth: an Investigation of Snowshoe Hare Dispersal, Habitat, and Predation Across the Apostle Islands
Jarod Reibel; Paul Keenlance; Jennifer Moore; Joseph Jacquot
The Apostle Islands National Lakeshore, comprised of 21 islands and a section of mainland, is located off the coast of Wisconsin along Lake Superior’s southern shore. Located near the intersection of the boreal and hardwood/hemlock forest range limits, as well as having differing disturbance histories, the islands vary greatly in composition. Snowshoe hare (Lepus americanus) are found on numerous islands and are an important prey species across their geographic range. As a species adapted to snowy winter conditions, changing climate conditions near their southern range boundary is thought to be moving their range limit northward. Due to decreasing snow duration as a result of climate change, hares are seeing an increasing amount of days having a camouflage mismatch with their environment, which results in lower weekly survival rates. With climate change modifying the environment, the ability to adapt is critical for a species survival, which requires genetic variation, and could be limited without dispersing individuals. The future is uncertain for snowshoe hares because of the current rapid change in climate, meaning further management will be key in maintaining populations near their southern range boundary. Unfortunately, little is currently known about the snowshoe populations that inhabit several of the Apostle Islands, which makes management impractical. This project aims to determine if snowshoe hares are dispersing between the Apostle Islands, or if gene flow is restricted, using a noninvasive technique of genotyping collected fecal samples. Additionally, we would like to identify if vegetative characteristics or the presence of certain predators are affecting island populations by conducting vegetation sampling, identifying predator presence with trail cameras, along with estimating hare densities by using the genotyped fecal samples in a mark-recapture framework. Understanding the role of vegetation and predation in limiting snowshoe hare density will help guide future management decisions by the National Park Service.
60 The Impact of Fragmented Landscapes on the Spatial Distribution and Dispersal of Eastern Cottontail Rabbits
Jourdan M. Ringenberg; Dustin H. Ranglack; Nate Bickford
As habitat fragmentation becomes an increasing reality of the modern global environment, understanding how wildlife species are impacted is key to their continued survival. Eastern cottontail rabbits (Sylvilagus floridanus), a widely distributed species throughout the United States, are particularly vulnerable to habitat fragmentation and often occupy relatively small habitat patches among highly modified landscapes. In areas of intense agriculture, modified landscapes are common and often are comprised of cropfields, grassy prairies, and/or woody habitat patches, all of which rabbits must navigate. Our objectives are to investigate rabbit spatial distribution and movement patterns in an agricultural landscape of south-central Nebraska. Study animals are captured and fixed with radio-telemetry collars in eight sites distributed across private farmland south of Kearney, NE. Rabbits are captured in woody or grassy habitat patches and each study site consists of at least three different habitat types. At least one and up to seven locations are determined per week for each rabbit for the duration of a year. Locations are taken randomly across six four-hour time periods to encompass the entire extent of possible movement. In addition to determining locations for all study animals to identify movement patterns, vegetation density, vegetation height, horizontal obstruction, percent canopy cover, and pellet count surveys are conducted biweekly at randomly placed quadrants across all eight sites to identify seasonal growth and rabbit abundance. Understanding how rabbits traverse and utilize fragmented landscapes is essential to their longevity and will help land and wildlife managers promote movement and species conservation.
61 The Influence of Agriculture on Mule Deer in a Fragmented Landscape
Levi Heffelfinger; Laura Warner; David Hewitt; Shawn Gray; Warren Conway; Timothy Fulbright; Randy DeYoung; Louis Harveson
Habitat fragmentation is an ever present and important issue for many wildlife species. Conversion of native rangeland to row crop farming is one of the largest forms of habitat fragmentation. Moreover, as the human population grows, the necessity for more agricultural land for food production will increase. Understanding how species react to such landscape alterations will prove important for conservation and management. Mule deer (Odocoileus hemionus) populations have been stable throughout most of the western United States but have been increasing in the Texas Panhandle, an area of extensive agricultural production. We seek to evaluate the influence of agriculture on several populations of mule deer in the Panhandle. We are currently evaluating how spatial crop use intensity and duration affects population health parameters such as rump fat, body mass, lactation rates, and antler size. Additionally, we are using the known fate and nest survival modules in Program MARK to assess how cropland use, among other habitat and morphometric measurements, influences adult and juvenile survival. We expect that the use of agricultural fields will positively influence rump fat and body mass across all age classes and sexes. Additionally, we hypothesize that lactating females will utilize cropland significantly more than non-lactating females demonstrating the allocation of nutrition to provisioning young. Adult male and female survival will likely be influenced by pre-winter rump fat and adult female survival will be negatively influenced by lactation. We also expect juvenile mule deer survival to be primarily driven by body mass at capture. In the coming months we hope to further understand the relationship between mule deer and their environment in the Panhandle. Creating baseline population measures will aid in establishing an adaptive management plan as mule deer population levels in the Panhandle continue to increase and the rangeland-cropland juxtaposition continues to change.
62 The Road to Recovery: A Study on Menace and Conservational Scheme for Sangai Deer
Manisha Singh; Dr. Neelam Khare
The only wild population of Rucervus eldii eldii or the Sangai deer in found in the North Eastern premise of India’s unique floating national park, the Keibul Lamjao National Park (KNLP), Manipur. The national park is located at the southern end of Loktak Lake, the largest fresh water lake in Eastern India and one of the seven Ramsar sites of international importance.Once reported to be extinct in 1951, the species was rediscovered in a small pocket during a survey conducted by IUCN in 1953. Since then the Government has been working hard to defend and stabilize the existence of species and yielded satisfying results, with population of Sangai escalated from 14 in 1975 to 260 in 2016. But the road to recovery is difficult due to number of menacing threats looming over the species. The construction of Ithai barrage for generation of power (Loktak Multi-Purpose project – 1983) has severely affected the hydrological regime of KLNP. Other threats can be recapitulated as biomass demand and competition with human, depleting habitat condition, poaching and a few more. The newest amidst the peril being the wild boars.Conservation measures are being carried out both in In-situ by habitat manipulation, rescue centre setups, anti- poaching camps and engrossing local community in conservation efforts whilst in ex-situ a captive breeding centre has been constructed in the Langol Reserve forest hosting 15 deer in number at present. Only the authorities of Manipur Zoological Garden are responsible for the care and perpetration of measures to forfend the deer.The above data was collected by field analysis and questionnaire. And hence it can be concluded that the conservational activities have stepped up and the government is also taking substantial measures to bring the species to sturdiness.
63 The Role of Neophobia and Social Rank in Social Learning by Coyotes
Julie K. Young; Laura Touzot; Stacey P. Brummer
Social learning has important ecological and evolutionary consequences but the role of certain factors, such as social rank and neophobia, remain poorly understood. We used three tasks to examine the role of object neophobia (i.e., avoidance of novel stimuli) and social rank in social learning by captive coyotes (Canis latrans). The first task involved individual animals and eliminated object neophobia by familiarizing the subjects to the testing apparatus prior to testing. The second and third tasks used mated pairs to assess social rank, and included object neophobia, but differed in that the third task decoupled the food reward from the testing apparatus. For all tasks we compared performance between coyotes that received a demonstration from a conspecific to control animals with no demonstration prior to testing. Coyotes demonstrated social learning during two of the three tasks, with persistence and dominance being important to individual success. Coyotes that observed a demonstrator performed the first and second tasks more successfully than control coyotes. Most coyotes were unable to perform the decoupled task, including those with demonstrators. These results suggest social learning of food-reward-based tasks occurs in dominant coyotes, but decoupling of the food reward reduces social learning capabilities. This study contributes to understanding the mechanisms underlying how animals gain information about their environment.
64 Trace Mineral Supplies for Regional Productivity of Mammals in Texas
Kaylee Hollingsworth; Rachel Shively; Perry Barboza
Copper (Cu), iron (Fe), and zinc (Zn), are essential trace nutrients for disease resistance and reproduction ultimately arising from bottom-up processes among soils and plants. Trace nutrient supplies must complement food supplies of energy (i.e. carbon) and protein (i.e. nitrogen) to sustain growth of wildlife populations. We compared trace minerals in hispid cotton rats (Sigmodon hispidus) and plants to assess differences in supplies across four ecoregions in Texas for a highly fecund mammal. Mineral stores in the liver were compared with diet by measuring carbon and nitrogen isotopes in the heart. We trapped rats (n = 73) at the end of summer (August – September) after collecting grasses (11 species) and woody browse (9 species) during peak growing season (May-June) across 15 study sites. Liver concentrations of trace minerals were higher (Cu 10 ± 4 ppm; Fe 620 ± 147 ppm; Zn 83 ± 15 ppm) than plant concentrations (Cu 5 ± 3 ppm; Fe 177 ± 215 ppm; Zn 27 ± 19 ppm) and consistent among regions, which indicates tight regulation of stores in cotton rats. Interregional variation was greatest in Cu for woody browse and Fe for grasses, which indicated differences in plant soil interactions among minerals. Diet varied widely in cotton rats, as high as 10 ‰ for δ13 C and 7‰ for δ15 N within one ecoregion. Dietary variation was not related to liver stores of Cu or Fe in rats but Zn stores were positively related to δ13 C ratios. Diverse diets with similar liver stores of trace minerals are consistent with highly productive populations of cotton rats that are often irruptive, even though plant mineral concentrations vary widely. We are testing this hypothesis by measuring mineral stores and diet of white-tailed deer (Odocoileus virginianus), a large seasonal breeder in the same ecoregions.
65 Understanding the Relative Roles of Nutrition and Predation in Regulating Sympatric Ungulate Populations in a High-Desert Ecosystem
Katey Huggler; Matthew M. Hayes; Patrick Burke; Mark Zornes; Daniel Thompson; Kevin L. Monteith
Demography of large ungulates is typified by relatively high and invariable survival of adults. Consequently, survival of young is the demographic that often underpins population trajectories. In systems where ungulates co-occur with predators, predation is commonly the leading cause of mortality among neonates. Moreover, predator avoidance by females during parturition is common among ungulates to minimize risk of predation to newborns; this behavior involves increased use of habitat that provides adequate cover for neonates. If habitats that minimize predation risk are not consistent with those that offer energetic gain, however, such behavior may conflict with the need for females to acquire adequate forage to meet the energetic demands of lactation. Behaviors in response to forage acquisition and predation risk can have important consequences for fitness, and therefore, we aim to link behavior, nutrition, and predation to survival and reproduction in mule deer and elk, two species that differ both in body size and behavioral strategies during parturition. We expect the sensitivity of habitat selection to predation risk by coyotes to vary as a function of body size and time post-parturition. Further, we expect that nutritional condition at the end of winter will influence whether parturient females adopt a risk-prone or risk-averse strategy, and such behavior should be related to neonate survival. Behavior is one of the primary mechanisms by which females cope with constraints on survival and reproduction. Thus, understanding the interacting roles of predation, habitat, and nutrition on behavior is key to identifying the mechanisms underpinning ungulate dynamics. High-desert systems in particular are ubiquitous across the West, therefore, identifying the relative contributions of behavior, nutrition, and predation on dynamics of thriving elk herds alongside of stagnant populations of mule deer in such a system is key to maintaining robust herds of elk and enhancing population growth of mule deer.
66 Using Computer Vision to Help Support Rewilding Decisions
Manoj K. Sarathy
An automated image classifier was implemented to distinguish remote camera trap images containing wildlife from images that are false positive so that conservation biologists can receive data pertinent to their analyses more quickly than was previously possible. Conservation biologists need timely and accurate data to make recommendations on whether or not rewilding — the reintroduction of flora or fauna to a geographical area — would be beneficial to the ecosystem of that area. In the last decade, remote camera traps have become one of the most widely used tools to capture this kind of data. Because the numerous deployed camera traps generate an unwieldy number images and citizen volunteer time to classify images is scarce, the images can languish for months before biologists receive the data they need. An obstacle to quickly classifying these images is that often three quarters of the images are “false positive” images that were not triggered by wildlife. False positive images can take much longer to classify than other images containing animals because every portion of the image must be analyzed to determine whether or not there is an animal. A convolutional neural-network based image classification software system was trained using actual remote camera trap images to separate false positives so that researchers can quickly prioritize images for manual classification and postpone analyzing the false positives. The model created was able to achieve an accuracy of approximately 86.6% on about 50,000 remote camera trap images. This model provided an economical and portable (capable of running on laptops) solution to initial classification of images. By first analyzing images the system identifies as containing animals and postponing analysis of other images, biologists are now able to get their data more quickly and make decisions regarding rewilding sooner.
67 What Data Are Sufficient for Making Informed Decisions to Manage Bobcats?
Tim L. Hiller; Florent Bled; Andrew J. Tyre; Robert Lanka
The application of appropriate scientific and statistical methods for furbearer management has often lagged behind that of large mammalian game species, largely due to limited state resources and the relative difficulty of collecting and analyzing data across large spatial and temporal scales. Development of contemporary statistical population models would allow for more informed and defensible furbearer management decisions, as these decisions are among the most scrutinized and challenged. In particular, questions arise about sustainable levels of harvest, particularly for western bobcats, where pelt prices and harvest effort have been relatively high for several seasons. Many state agencies collect detailed and long-term harvest data on bobcats, providing a unique opportunity to address concerns and ensure sustainable harvest. Although the majority of U.S. state wildlife agencies use harvest-based indices to monitor trends in bobcat abundance, these indices may have unquantified biases. Further, collection of harvest data often varies substantially by state, and temporally within states, raising questions about what data are necessary to make informed decisions in a complex management scenario. Using harvest (e.g., age-at-harvest data, harvest effort) and auxiliary (e.g., demographic) data in a statistical population reconstruction framework, we developed population models at the management area and statewide scales in Wyoming to assess long-term trends in abundance, and to evaluate the ability of different types of harvest data to inform decision making. Our early results suggest that harvest information resolved to age class (juvenile, adult) and sex may not support relative abundance predictions given our data. However, age data resolved to age-in-years may be adequate for predictions of long-term changes in relative abundance. Supplemental information on the effective harvest with regard to the population (obtained either through complementary field studies, external data, or literature) might allow for an estimate of abundance.
68 Wildlife Movement Path Metric Responses to Human Footprint Gradients
Tristan Nunez; Justin Brashares
The fine-scale movement behaviors of wild animals are being revealed at unprecedented resolutions by GPS tracking datasets, enabling inference into how these behaviors change in response to human activities and features. Importantly, analysis of movement paths can provide an important supplementary view to insights gained traditionally from use versus availability frameworks (resource or step selection functions). In particular, this understanding of movement behaviors can shed light on the assumptions underlying resistance surfaces and landscape connectivity analyses used frequently in conservation planning. We draw on ecological and conservation literature to generate hypotheses about potential changes in movement rates, turn angles, tortuosity, first passage times, and other movement path metrics in response to human footprint gradients. These path metrics form a suite of metrics that can help identify behavior-specific responses such as directional movements, hiding, avoidance, foraging, corridor use, denning, and resting. We then evaluate these hypotheses across multiple species, including deer, fisher, mountain lions, elephants, and others, using large high-resolution GPS telemetry datasets for each species. We find that movement metrics generally change with increases in human footprint, but the direction and magnitude of response vary among species. For many species, we find increases in movement rate and reduced tortuosity in high footprint areas. We consider the implications of these changes for the movement assumptions underlying resistance-based connectivity models.

 

Poster
Location: Cleveland CC Date: October 11, 2018 Time: 9:00 am - 5:00 pm