Fishers (Pekania pennanti) are a medium-sized carnivore of conservation concern in portions of the western United States. Fishers in the southern Sierra Nevada of California are federally endangered and other populations in the western United States have been evaluated for federal listing. Small and isolated populations are the most immediate and challenging threat to fisher persistence in Washington, Oregon and California. Species distribution, habitat associations, and population size are critical information needs for species recovery. Distribution patterns and habitat associations have been evaluated using presence-only and expert opinion models. These models, however, frequently fail to meet assumptions, increasing calls for analyses in detection-non detection frameworks. We modeled the distribution, abundance, and habitat relationships of fisher across Washington, Oregon, and California. We developed a hierarchical model of detection-non detection data using occupancy and integrated modeling. We found that the use of bait and surveying in the winter had positive influence on the detection of fishers. We also found that canopy cover, the standard deviation of canopy cover, and stand age influenced patterns in fisher abundance. Our results will be a useful decision-support tool to promote species recovery in the face of stochastic events and a changing climate.

Using a combination of fine-scale GPS data and directed experiments, we evaluated the perceived risk and foraging behavior of Humboldt martens (Martes caurina humboldtensis), an endangered subspecies of Pacific marten. We used GPS collars to collect location data every five minutes (15M:4F, northern California, southern Oregon). We compared vegetation characteristics between resting sites, movement paths, and giving up density (GUD) experiments with diminishing bait returns. We stratified GUD experiments equally within five treatments differing by stand age and complexity. We used remotely triggered cameras to quantify marten visits and categorize vigilance, predicting these would correlate with individuals’ GPS locations (mean±SD: recent clear cuts 3.7±2.8%, regenerating forest 4–20 years 48.3±34.3%, forest >20 years 23.6%±34.9%, riparian 24.4±15.5%). Rest sites often occurred in large structures or rock piles near a high density of downed logs. Within home ranges, martens’ GPS locations were similar to available. One study area, an industrially managed landscape, only had stands <100 years old. Dissimilarly between methods, martens visited baited GUD experiments within riparian more than other landcover types. Visit length did not differ by strata (9.6±16 minutes). Vigilance decreased as the number of visits increased. Multiple methods enhanced our understanding of habitat use and behavior.

In 2016, the Southern Sierra Nevada Fisher Conservation Strategy was released, representing an enormous multi-year effort to map habitat of federally endangered Pacific fishers (Pekania pennanti) and develop recovery guidance. Yet at the same moment, the southern Sierra Nevada was undergoing rapid change. The unprecedented 2012-2016 California drought and subsequent severe megafires produced widespread tree mortality, undoubtedly reshaping fisher habitat—but to an unknown extent. We used 15,666 fisher locations collected from 2002 to 2022 to develop a habitat model for fishers in the Google Earth Engine environment. We used this model to produce a continually-updating 36-year time series of habitat change for southern Sierra Nevada fishers at 30-m resolution. Our work not only provides a much-needed update to the 2016 Conservation Strategy habitat map but will continue to update in the future as new changes to the landscape inevitably occur. Our maps can be used to assess disturbance effects on habitat change, thus providing support for management decisions. In the face of rapid landscape changes from fire and drought in the western US, new modeling approaches are needed to map changes to species habitat in real-time, so managers have the most up-to-date information possible when making decisions.

Rodents provide a unique glimpse into the environmental conditions at specific geographic ranges and timescales and play an important role as primary consumers. Investigating their diet using thermal ionization mass spectrometry reveals resource availability across periods of precipitation fluctuation. Museum specimens and live small mammals were sampled for this study to analyze stable isotopes in correlation with Palmer Drought Severity Indices. Small mammal trapping occurred at the Jack and Laura Dangermond Preserve in Lompoc, California for two consecutive summers to infer diet under varying levels of drought severity. Animal tissues were analyzed in collaboration with the University of New Mexico’s Center for Stable Isotopes to determine the diet of small mammals on the Dangermond Preserve and a historical analysis of Neotoma specimens from Central California. Variability in the ratios of stable isotopes indicates changes in resource availability during periods of drought across many decades of California’s history. Diet changed in response to drought on spatially explicit scales and is subject to habitat-specific pressures as well. Understanding the previous and current ecological niches of native rodents may further our understanding of resource use and conservation efforts.

Preferred Session: Ecology and Conservation of Mammals Type of paper: oral presentation Paper Title: Summary of USGS research on the American badger (Taxidea taxus) in San Diego County Devin Adsit-Morris, Western Ecological Research Center, U.S. Geological Survey,4165 Spruance Road, San Diego, CA 92101, dadsit-morris@usgs.gov, (619)840-0127; Co-authors Cheryl Brehme; Philip Gould; and Robert Fisher Abstract: The American Badger is a wide-ranging mid-sized predator associated with grassland and upland habitats and is a target species for monitoring regional-scale connectivity under the San Diego Multiple Species Conservation Plan. Due to habitat loss, fragmentation, and road mortality, badgers were considered at risk of loss from the region. Previous to 2011, there was little information of their presence within San Diego County. Since then, we have employed multiple methods for detecting badgers including canine scent surveys, sign surveys (burrows, digs, tracks), infrared cameras, public outreach, roadkill reports, and individual identification using facial markings. Canine surveys found 203 scat samples with 25 confirmed to be badger. Over 300 sign surveys produced 80 confirmed badger occurrences, outreach produced over 200 reports with 110 confirmed to be badger, and 32 road mortalities were documented. We theorize that badgers in the County are at low densities and operating on a large spatial scale. Distribution models and least-cost paths were developed using badger data, slope, soils, and vegetation. Models and roadkill data have enabled us to model high-quality habitats and identify roads of concern. Future plans include refining habitat suitability models, conducting radiotelemetry, and identifying of locations for road passages to facilitate safe road crossings.

In California, highly pathogenic avian influenza (HPAI) H5N1 virus infections resulted in the deaths of a free-ranging bobcat (Lynx rufus), six mountain lions (Puma concolor), two fishers (Pekania pennanti) and two raccoons (Procyon lotor) between October 2022 and April 2023. Three of the animals, a fisher and two raccoons, were reported to have had neurological signs and were either euthanized or died in care. The cadavers of the other eight animals were recovered after mortality sensors from their radio collars were activated. Necropsies were conducted, and tissues were submitted for histopathological analysis. HPAI H5N1 virus was detected in the brains with associated encephalitis. Genotyping showed that the virus belonged to the HPAI GsGd H5 clade 2.3.4.4b of viruses found in wild birds. Further analysis demonstrated that the viruses were not closely related in eight of the 11 cases, suggesting that the virus was not transmitted between these individuals. However, viral sequences were identical in three mountain lions indicating potential for mammal-to-mammal transmission. While mammals primarily become infected through consumption of infected birds, these results indicate infections in mammals should be monitored to detect the emergence of mutations that may increase species spillover and zoonotic potential of HPAI H5N1 viruses.

The Humboldt marten (Martes caurina humboldtensis) is a species of conservation concern in northern California and Oregon. Historically, these populations were threatened by unregulated fur trapping and timber harvest leading to a 90% decline in their distribution. Contemporary populations are small and isolated and continue to be threatened by a loss of habitat from timber harvest and increased wildfire risk. Due to their small body size, martens are also at risk of interspecific competition and intraguild predation. Despite efforts to better understand their contemporary distribution and habitat associations, little is known about the abundance and interspecific relationships of Humboldt martens across their current range. To fill these information gaps, we deployed non-invasive hair snares and remote cameras across three sampling grids in the North Coastal California EPA. We developed a spatial capture recapture model to estimate marten abundance and density, and two single-species occupancy models to estimate the space use of fishers and bobcats. We then tested the effects of habitat and carnivore space use on marten density. Given that this is the first effort to evaluate how biotic and abiotic variables influence the distribution and variation of marten density, we offer insights for management and future survey efforts.

It’s important to understand species-habitat relationships over time for effective adaptive management of rare species. We report results from 10-years of track tube monitoring of the two largest extant populations of the federally endangered Pacific pocket mouse (PPM) on Marine Corps Base Camp Pendleton (MCBCP). Total area occupied by PPM has ranged from 117 to 335 ha, with a recent estimate of 165 ha. Dynamic occupancy and structural equation modeling have shown PPM spatial and temporal dynamics are positively associated with open ground, forb and perennial herb cover, while negatively associated with non-native grass cover (NNG), which is strongly influenced by rainfall and fire. MCBCP have used these results to support fire and vegetation management resulting in largely positive PPM responses and colonization of new areas. Due to regulatory concerns about habitat maintainence in PPM occupied areas, management actions have been largely limited to sparsely occupied or unoccupied habitats or after major population declines. Inclusion of habitat maintenance in PPM occupied areas may promote resiliency and reduce population stochasticity, chances of population decline and loss of genetic diversity. We discuss concurrent research on phenology, climate, and diet; and show the power of this program in supporting a monitoring-management feedback loop.