Innovative Approaches for Furbearer Research (hosted by TWS)

Information-driven decision making is critical for the development of wildlife management strategies. Due to the complex role of many furbearer species in the ecosystem, either as a predator or prey species, information on these species not only aids in furbearer management decisions, but in strategies for other wildlife and fisheries species.  Unfortunately, furbearers are often difficult to monitor and study due to their often-cryptic nature, elusive behavior, and, for some species, comparatively low densities. This is frequently exasperated when working over large geographic areas in remote regions. However, advancements in technology and analytical tools have increased the ability to for management agencies and researchers to monitor furbearer populations and address existing knowledge gaps. Improvements to existing trapping technology and leveraging the skills of experienced trappers have provided tools to measure responses to management actions. These recent innovations will be highlighted to showcase modern tools available to students, wildlife managers, and researchers. This symposium is being conducted in conjunction with the Trapping Matters Workshop, which includes training on advancements in trapping technology, including trapping Best Management Practices, to aid in furbearer research, surveys and monitoring efforts. The workshop, combined with this symposium, will stress how wildlife research and management incorporate contemporary science and modern technology.

8:00AM 1) Do You Have to See It to Believe It? Incorporating Imperfect Detection into River Otter Track Surveys
  Shawn Crimmins
Monitoring furbearer populations is a challenging task for wildlife management agencies, as such species tend to occur at low densities, be cryptic in nature, and occur in areas that may be difficult to access. This is especially true for river otters, which has led many management agencies to attempt to use various forms of track surveys as a monitoring tool. Here, I discuss three different forms of river otter track surveys conducted in Wisconsin (aerial surveys, road transects, bridge-site surveys) and highlight how different analytical approaches can be used to account for imperfect detection in each. In each case I accounted for imperfect detection through a different modeling process that was suited to the unique data structure. Regardless of the data collection process, imperfect detection was present and affected inferences regarding river otter distribution, relative abundance, and habitat selection. I demonstrate that a variety of data collection and analytical approaches can be used with track surveys for river otters and discuss how future monitoring frameworks may take detection probabilities into account to improve monitoring and management programs.
8:20AM Partnering with Trappers to Monitor Elusive Species
  Nathan Roberts
Capturing elusive mesocarnivores for research can be logistically challenging. To make robust inference from information collected from captured animals, such as GPS collars, it is critical to have an adequate sample size and spatial distribution of samples. Some species, such as coyotes and fox, are wary and cautious animals that are difficult to catch. Others, such as bobcat and lynx, occur in naturally low densities and can be widely dispersed. We describe two case studies in Wisconsin were researchers partnered with avocational trappers to catch coyotes and bobcats to equip with GPS collars as part of two larger studies. By partnering with the trapping public, a large sample size was obtained over a large geographic area. This approach proved to be an economically efficient strategy as well as a unique opportunity to directly involve stakeholder in research efforts.
8:40AM Mischievous Muskrats; Survey Techniques for Survival and Health Investigations in Pennsylvania
  Laken Ganoe, Justin Brown, Matt Lovallo, Michael Yabsley, Mark Ruder, Duane Diefenbach, David Walter
Declines in muskrat (Ondatra zibethicus) harvest have been observed throughout their North American range. Potential causes of the declines are unknown, but theories may include habitat loss or degradation, predation, and disease. Recent data on muskrat diseases are scant. We used collaborative efforts between trappers, the Southeastern Cooperative Wildlife Disease Study (SCWDS), and agency personnel to address this knowledge gap. This project used two main approaches to assess muskrat survival and health; 1) muskrat parasite, pathogen, and disease investigation using a retrospective study, combined with passive and active surveillance, and 2) a telemetry study to characterize survival and movement of muskrats in Pennsylvania. Pennsylvania furtakers were willing and active to participate, donating over 600 muskrat carcasses during the 2019 trapping season. We examined the carcasses for lesions and parasites and tested for specific diseases of historic significance as determined by the retrospective study (e.g. Tyzzer’s disease and tularemia). For passive surveillance, we reviewed all muskrat diagnostic records (n=27) from SCWDS for cause of disease. To increase trap success, an experienced muskrat trapper trained technicians on use and placement of six live-trap types. Ultimately, only the double-door Tomahawk traps were successful catching muskrats (n=17) at a series of ponds in an urban area in central Pennsylvania. We surgically-implanted muskrats with radio transmitters and monitored weekly. We determined fall home ranges, hourly movements, rest-site use, and overwinter survival rates. Transmission range of the implants were extremely low (>100m) suggesting future studies would benefit from the greater spatial and temporal resolution provided by GPS technology. Trapping and disease survey efforts in the first year of the study have provided valuable insight into successful trapping techniques and potential avenues of data collection and collaboration for future muskrat investigations.
9:00AM Analysis of Black Bear Feeding Behavior Using Video-Enabled GPS Collars
  Marcella Kelly, Robert Alonso, David McNitt
Diet analyses of field-collected scats provide information on frequency of occurrence of diet items in predator scats, but results often are difficult to interpret. For example, remains of large prey species like white-tailed deer are more easily detected due to high amounts of indigestible parts, while more digestible items such as fruits, vegetation, fungus, small mammals, and birds can be harder to detect. Thus, frequency of occurrence data is as relative measures that may miss important diet components. To address potential for missed food items, and to distinguish fawn versus adult deer feeding events, we combined GPS telemetry data with video enabled collars that collected GPS locations every 20 minutes and simultaneously recorded 20 seconds of video. We affixed the camera collars to 5 male and 5 female American black bears (Ursus americanus) from late May to early August 2018. Video data was highly informative of species eaten by bears (both plant and animal), time and duration of feeding, and changes in diet with seasonal availability of food items on the landscape. Preliminary results indicate bears primarily ate vegetation and deer (fawns and occasionally adults) in spring (May to mid-June), and switched to blueberries, blackberries, bear corn, and other types of mast and vegetation as it became available in late June through August. Although sample sizes were small, males appeared to forage more on deer and less on vegetation than females, especially in spring, and instead spent more time finding or following potential mates and more time mating than females.
9:20AM Assessing Mesocarnivore Response to Severe Drought through Long-Term Population Monitoring
  Jody Tucker, David Green, Sean Matthews
From 2012-2015 California experienced an extreme drought, unprecedented in severity and duration in the last ~1200 years. This represents a major disturbance event with the potential to profoundly affect biological communities, yet few studies have evaluated the response of forest-dependent wildlife to a severe drought. In this study we assessed the effects climactic and vegetation change over a 14 year period (2002-2015) on four mesocarnivore species in the southern Sierra Nevada Mountains; fisher (Pekania pennanti), Pacific marten (Martes caurina), ringtail (Bassariscus astutus), and gray fox (Urocyon cinereoargenteus). These four species have been shown to exhibit a variety of competitive interactions due to similarities in body size, diet, and habitat requirements. We analyzed detection information from a long-term carnivore monitoring network across ~12,000 km2 in the southern Sierra Nevada. Monitoring occurred by repeated visits within and across years at fixed monitoring arrays of track plates, remote sensor cameras, and hair snares. Each monitoring array was resurveyed 3-5 times/year over an average of 8 out of 14 years resulting in a total of ~36,484 visits over the study period. We used dynamic occupancy models to investigate how changes in climate (precipitation, snowpack, and temperature) and vegetation (canopy cover, land cover type) influenced detectability, site occupancy, colonization, and persistence of each species over time. This this long-term data set provides a unique opportunity to investigate the effects of climate and vegetation change over time on a mesocarnivore guild at a landscape scale.
09:40AM Break
10:10AM Comparing and Contrasting Trends in Harvest and Non-Harvest Based Indices of Furbearer Abundance
  Javan Bauder, Maximilian Allen, Thomas Benson, Craig Miller, Kirk Stodola
Accurately monitoring furbearer populations at broad spatio-temporal extents is an important challenge in furbearer management and managers often use metrics derived from harvest (e.g., trapper harvest surveys) and non-harvest data (e.g., roadkill and spotlight surveys) as broad-scale population indices. Variation in these metrics may be driven by diverse climatic and landscape factors yet harvest-based metrics may also be driven by socio-economic factors (e.g., pelt price), which may be independent of population size. Comparing harvest and non-harvest based metrics across furbearer species can provide insights into potential sources of variation in these metrics and the degree to which they reflect species abundance. We compared statewide trends in harvest and non-harvest based metrics for six furbearers in Illinois, raccoon (Procyon lotor), skunk (Mephitis mephitis), opossum (Didelphis virginiana), coyote (Canis latrans), red fox (Vulpes vulpes), and gray fox (Urocyon cinereoargenteus), from 1977­‑2018. Harvest metrics included harvest per trapper license, harvest per successful trapper, and total annual harvest. Non-harvest metrics included annual roadkill, spotlight, and archery deer hunter surveys. We also evaluated the effects of fur price for each species. The degree of correlation between harvest and non-harvest metrics varied across species and non-linear trends were present across metrics and species. Fur prices generally declined for canids and raccoon while remaining comparatively stable for skunk and opossum. The concordance between non-harvest metrics and per-trapper harvest suggest red and gray fox have declined in Illinois while raccoon and coyote may have increased. Harvest metrics were not consistently correlated with fur price across species. The varying degree of concordance among metrics highlights the importance of understanding how different factors (e.g., socio-economic, climatic, ecological) influence furbearer metrics to provide a better understanding of the extent to which these metrics reflect true population abundance.
10:30AM Overcoming Challenges of Furbearer Research in the Southwestern United States
  Nicholas Forman, James Cain
Despite advances in methods to monitor cryptic species through non-invasive sampling, furbearer population dynamics are often under studied due to logistic restraints and competing priorities of management agencies. In New Mexico, many furbearing species occur across multiple habitat types present in the state, thus precluding statewide, ad hoc application of density estimates from the literature to derive population estimates. Traditional track surveys and other index based methods for monitoring have been found to be unreliable for describing trends in abundance. We assessed data collected through incidental ‘captures’ of furbearer species that occurred while conducting non-invasive scat collection and remote camera surveys of mountain lion (Puma concolor) populations in New Mexico to determine the feasibility of incorporating such data into a furbearer monitoring program. Camera trapping over six months in the mountains of north-central New Mexico resulted in 2,619 detections of eight furbearer species, the most common species being bobcat (Lynx rufus, n = 262), coyote (Canis latrans, n = 1,500), and grey fox (Urocyon cinereoargenteus, n = 672). Of the 262 photo events of bobcat, a minimum of 13 individuals were detected on multiple occasions. A total of 89 bobcat ‘captures’ were recorded while surveying for mountain lions with scat detection dogs at four study areas in New Mexico, with a maximum of eight captures for a single individual. The increased resolution of data collected via remote cameras over that from track surveys, including the ability to identify individuals or positively identify multiple individuals, allows for additional methods of statistical analysis such as spatial-capture recapture methods or integrated population models. By implementing studies that generate incidental captures with little or no detriment to the sampling efficacy of the target species, in this case mountain lions, sufficient data can be produced to monitor population trends for certain furbearer species.
10:50AM A Comparison of Two Methods of Diet Analysis for the North American River Otter (Lontra canadensis) in North Carolina, USA
  Charles Sanders, Stephen Spear, Colleen Olfenbuttel, Chris DePerno
The river otter (Lontra canadensis) is an important furbearer in North Carolina and a significant predator of riverine and estuarial fish and mollusk species. We examined the diet of harvested river otters throughout North Carolina using traditional methods of examining and identifying stomach contents and using metabarcoding DNA analysis to identify species. We collected stomachs at necropsy and kept them frozen until processing. Stomach contents were weighed and then identified using visual marks, scales, bones, and other visual metrics. We used dichotomous keys, guides, and expert consultations to identify individual specimens to the most precise level possible. We used a genetic metabarcoding approach to compare to the morphological analysis. Metabarcoding DNA analysis is useful to identify food items that are not readily observed or distinguishable in the stomach. Specifically, we extracted and amplified two separate 16S mitochondrial regions for each stomach sample using next-generation sequencing methods. We compiled a genetic reference database of potential invertebrate and vertebrate prey species and classified each sequenced amplicon to species. We compared and contrasted results between the two methods overall, by region, and by river basin. From 2011-16, we collected and analyzed stomachs from 540 harvested river otters. Of these 27% (147) were empty. We observed fish, crayfish, invertebrates, amphibians, snakes, mammals, and birds in the stomachs. A better understanding of the diet of river otters throughout the state will benefit the management of otters and fisheries.
11:10AM Using Fecal DNA and GPS Telemetry to Understand the Consumption of White-Tailed Deer By Carnivores in Western Virginia
  Robert Alonso, Dana Morin, David McNitt, Marcella Kelly
Previous carnivore diet research in western Virginia indicated that white-tailed deer (Odocoileus virginianus) were the most commonly shared prey item among bears (Ursus americanus, 40%), bobcats (Lynx rufus, 36%), and coyotes (Canis latrans, 72%). However, these standard frequency of occurrence methods, are difficult to interpret. Remains of large prey species like white-tailed deer are easily detected due to high amounts of indigestible parts, and the same individual deer can be found in numerous scats, including across predator species. In addition, variable space use patterns among predator species can affect the detectability of deer consumption for each species, making across species comparisons problematic. As such, frequency of occurrence measures do not directly estimate the number of deer consumed across an area, but are instead vague relative measures of diet. This can lead to ambiguous interpretations about predator resource requirements and impacts to prey populations. We combined GPS telemetry data with scat collection from a network of established transects to estimate the number of deer consumed by each predator. Using mitochondrial DNA, we identified both the carnivore species that deposited the scat and the number of scats containing deer DNA. We incorporated GPS telemetry data for each predator species within the detection model allowing for variable distance of travel from a carcass for each predator. This approach is a step forward in quantifying predator diets, understanding and interpreting frequency of occurrence measures, and will lead to a deeper understanding of predator-prey dynamics.
11:30AM Panel Discussion

Organizers: Nathan Roberts, Colleen Olfenbuttel
Supported by: TWS Hunting, Trapping and Conservation Working Group

Location: Reno-Sparks CC Date: October 1, 2019 Time: 8:00 am - 11:50 am