Conservation and Ecology of Mammals VII

Contributed Paper
ROOM: Room 110 – Galisteo

1:10PM Assessing Site Occupancy of the Oscura Mountains Colorado Chipmunk Using Camera Traps
Ian E. Perkins-Taylor; Jennifer K. Frey
The Oscura Mountain Colorado chipmunk, Tamias quadrivittatus oscuraensis, is a rare subspecies of the Colorado chipmunk endemic to the Oscura Mountains in south-central New Mexico. It is listed as threatened by the state of New Mexico because of its small, isolated habitat and the potential for continuing habitat loss. Knowledge about the ecology of the Oscura Mountains Colorado chipmunk is limited, and most previous studies and monitoring efforts were not published or adequately documented. More robust information about its distribution and ecology is necessary to develop conservation and management recommendations. Our goal was to collect scientifically rigorous data to better understand the ecological factors related to its distribution. We implemented camera trapping surveys in an occupancy modeling framework to assess the ecological factors associated with occupancy and detection probability. Given limited existing information about the natural history of the Oscura Mountains Colorado chipmunk, we first conducted a pilot study to determine appropriate a priori hypotheses regarding which ecological factors were most important. We then collected detection-nondetection data using the camera traps to test these hypotheses using single-season site occupancy models. Our final model showed that areas near an escarpment had an occupancy probability of nearly 100%. In areas far from an escarpment, pinyon woodland sites had an occupancy probability of approximately 30%, while occupancy probability varied greatly by elevation at sites not classified as pinyon woodland (0%-56%). This highlights the importance of pinyon woodlands and escarpment habitat for this taxon. Managing the pinyon woodlands on the Oscura Mountains, particularly in areas near escarpments, will help maintain the habitat that is essential for the chipmunk population. Efforts should be implemented to prevent or minimize the risks of wildfire, disease, and anthropogenic disturbance, and long-term planning must account for the increased level of disturbance predicted under current climate change models.
1:30PM Fight! Fight! Fight! Tail Wounds as Indicators of Agonistic Behavior and Predation Attempts in America Beavers.
Steve Windels
American beavers are highly territorial and agonistic interactions appear common, as residents must defend their territories from neighbors or dispersers seeking their own territories. Beavers use their formidable jaws and incisors as their primary weapon during territorial battles, which leave telltale marks on the hide and tail. Beavers are also subjected to predation by wolves, bears, and other large carnivores and canine punctures from predation events are also observable in the hide and tail. Wounds are difficult to see on the hide without killing and skinning the animal but tail wounds are obvious and easily observed on live-captured beavers. From 2008-2016, tail wounds on live-captured beavers caught in Voyageurs National Park were recorded for >1000 capture events, including many beavers caught >1x. Wounds were classified by type, ranging from superficial nicks on the margins of the tail to large bites that remove sections of tail flesh. Distribution of tail wounds was biased towards the distal portion of the tail. Number of tail wounds increased with beaver age but were similar between sexes. Beavers in lake habitats had >2x more tail wounds than beavers in ponds, suggesting beavers in lakes experience more territorial conflicts than their pond counterparts. As expected, number of tail wounds increased with dispersal distance. Several beavers were documented with what appear to be canine punctures from wolves, some of the first evidence of beavers surviving predation attempts by wolves.
1:50PM Habitat Characteristics of the Hoary Marmot: Assessing Distribution Limitations in Montana
Ben Y. Turnock; Andrea R. Litt; John M. Vore; Christopher M. Hammond
Species that live in ecosystems with extremely different seasonal conditions must balance the constraints of each season to survive. For example, alpine species must accommodate both short growing seasons and long, harsh winters. Hoary marmots (Marmota caligata) are an alpine obligate patchily distributed at or above treeline throughout western North America, north of the 45th parallel. Given that climate change is impacting alpine ecosystems rapidly, this species may be especially vulnerable. We investigated the habitat characteristics of hoary marmots in western Montana to provide a better understanding about habitat selection at the southern extent of this species’ distribution. We conducted occupancy surveys in three study areas along a latitudinal gradient in 2014 and 2015 to assess the importance of specific habitat characteristics to their presence on the landscape. Slope, aspect, and presence of shrubs all were important habitat characteristics. Our results provide evidence that marmots may strike a balance between the environmental conditions they require during summer and winter. Shallow slopes typically accumulate deeper snow in winter that provide the best insulating snowpack. However, southern aspects allow for more snow-free areas in spring, providing a slightly longer growing season than northern aspects. Hoary marmots may be selecting areas with shrubs because shrubs can accumulate deeper snow and the additional insulation can increase subnivian temperatures. Other studies suggest that marmot survival is influenced by snowpack, indicating that marmot distribution may be more closely tied to winter conditions rather than summer conditions. Most studies of hoary marmots occur only during the short growing season, highlighting the difficulty of establishing how habitat selection may limit the distribution of hoary marmots. Effectively conserving, monitoring, and managing alpine obligates under an uncertain climate future will require a closer look at how winter conditions drive habitat selection and distributions on the landscape.
2:10PM Protected Avian Predators Challenge Continued Survival of an Endangered Endemic Forest Mammal
Emily A. Goldstein; Melissa J. Merrick; John L. Koprowski
Peripheral populations are often ecologically distinct and of conservation concern due to isolation, low numbers, and climatic extremes. Targeted management programs are informed via fine-scale analyses of demographic processes in such populations. Population survival rate could be more important than fecundity in species that do not have highly reproductive life histories. The Mount Graham red squirrel (MGRS; Tamiasciurus fremonti grahamensis), the southernmost population of North American red squirrel, is restricted to one mountain range in southeastern Arizona, USA. The population has been monitored intensively, and individuals marked and radio-tagged since 2002. Although MGRS reproduction is similar to congeners, little to no population growth is observed. Here we estimate the survival rates and mortality hazard in MGRS to understand the impact of natural mortality on survival and population function. We estimated survival for 381 marked MGRS and a subset of 135 known-fate individuals using Cormack-Jolly-Seber models. We used hazard models and competing risks regression to assess cause-specific mortality by age class and sex. Across sex and juvenile and adult age classes, individuals had an approximately 10% chance of dying in any given month and a 63% chance of dying each year (annual survival estimate = 0.37). Avian predators were responsible for >65% of all known mortality events. Low juvenile and adult survival and multiple raptor species of conservation concern complicate management for population growth and recovery.
2:30PM Rodent Response to Harvesting Wood for Energy
Sarah Fritts; Chris E. Moorman; Steve M. Grodsky; Dennis W. Hazel; Jessica A. Homyack; Chris B. Farrell; Steven B. Castleberry; Emily H. Evans; Daniel U. Greene
Harvest of low value trees and logging residues for bioenergy reduces downed wood post-harvest with potential indirect negative effects on ground-dwelling wildlife such as rodents. We assessed the influence of woody biomass harvests on rodent abundances using an operational-scale, randomized complete block experimental design study in North Carolina, USA (4 blocks) and Georgia, USA (4 blocks). Each block contained 6 treatments randomly applied to a clearcut harvest, that varied by volume of woody biomass retained, and that represented existing woody biomass harvest recommendations. We live-trapped rodents 2011–2013, calculated the minimum number known alive (MNKA) for each captured species, and used the value as an index of abundance. We compared abundance of the 3 most commonly captured species, deer mouse (Peromyscus spp.), house mouse (Mus musculus), and hispid cotton rat (Sigmodon hispidus), among treatments with generalized linear mixed effects models. We assessed relationships between species’ abundance and measurements of downed woody debris and vegetation in each treatment unit using linear regression. Although abundance varied among treatments in some cases, we did not detect consistent relationships between woody biomass retention treatments and abundance. Volume of downed woody debris in the treatment unit negatively influenced M. musculus, but had varying influences on Peromyscus spp. across years. Downed woody debris groundcover negatively affected Peromyscus spp. in North Carolina in 2012. Litter groundcover negatively influenced Peromyscus spp. and S. hispidus, whereas grass groundcover positively influenced S. hispidus. The lack of consistent relationships between rodent abundance and volume of retained woody debris suggests that the rodent species captured in this study are not affected greatly by current efficiencies of operational woody biomass harvests in southeastern United States loblolly pine plantations. However, focal species were habitat generalists and less common species may have greater sensitivity to biomass harvests.
2:50PM Refreshment Break
3:20PM What to Eat in a Warming World: Do Increased Temperatures Necessitate Hazardous Duty Pay?
Embere Hall; Anna Chalfoun
Contemporary climate change affects nearly all biomes, causing shifts in animal distributions and resource availability. Changes in resource selection may allow individuals to offset climatic stress, thereby providing a mechanism for persistence amidst warming conditions. While the role of predation risk in food choice has been studied broadly, the extent to which individuals respond to thermoregulatory risk by changing resource preferences is an open question. We addressed whether individuals compensated for temperature-related reductions in foraging time by altering forage preferences, using the American pika (Ochotona princeps) as a model species. We tested two hypotheses: 1) Food Quality Hypothesis: Individuals exposed to temperature extremes should increase selection for higher-quality vegetation; and 2) Food Availability Hypothesis: Individuals exposed to temperature extremes should put a premium on foraging quickly, thereby decreasing selection for higher-quality food. We quantified the composition and quality (% moisture, % nitrogen, and fiber content) of available and harvested vegetation, and deployed a network of temperature sensors to measure in situ conditions for 30 individuals. Data were collected in the central Rocky Mountains, July-September, 2015. We found unequivocal support for the Food Quality Hypothesis. Individuals exposed to more extreme daytime temperatures showed increased selection for high-nitrogen and for low-fiber vegetation. By contrast, pikas that experienced warmer conditions did not reduce selection for any of the three vegetation-quality metrics that we assessed, as predicted by the Food Availability Hypothesis. By shifting resource-selection patterns, temperature-limited animals may be able to proximately buffer some of the negative effects associated with rapidly warming environments.
3:40PM Does Lack of Nest-Building Foundations Limit an Arboreal Folivore in Young Forest?
Mark Linnell; James Swingle; Damon Lesmeister; John Bailey; Eric Forsman
Limitations on wild populations have long intrigued wildlife biologists, conservationists, and land managers. The red tree vole (Arborimus longicuadus) is an arboreal Arvicoline endemic to coniferous forests of Oregon and northern California. Red tree voles inhabit tree canopies near their food source of conifer needles, and build reproductive and resting nests on available woody foundations, frequently in tree deformities (e.g. cavities, branch whorls) caused by physical damage or disease. Young trees often lack nest-building foundations, and red tree voles primarily occur in old forest (>80 years). We used a before-after treatment study design to examine our main hypothesis that lack of nest-building foundations limited occupancy of young trees by red tree voles in young forest. We further examined patterns of colonization. In 2015, we added 429 artificial foundations, at a density of two per hectare, to young trees in 17 even-aged patches of young forest (25-40 years old; 7 to 23 ha in size), located adjacent to patches of old forest which we predicted would be a source of migrants. No young trees contained red tree voles in 2015 and 4% contained evidence of old nests. We checked artificial foundations in 2016 and determined 1-year occupancy by searching for red tree voles or their nests. Mean occupancy of artificial foundations in 2016 was 29% (95% confidence interval 18% to 40%), and we captured 47 red tree voles, of which 31 were breeding females. Probability of presence within patches was higher when adjacent artificial foundations were occupied, but was not related to distance to old forest. Our year 1 results broadly supported our hypothesis that lack of nest-building foundations limited occupancy of young trees by red tree voles, particularly for breeding females.
4:00PM Range-Wide Conservation Genomics of Black-tailed Prairie Dogs
John A. Erwin; Robert Fitak; Melanie Culver
The black-tailed prairie dog (Cynomys ludovicianus) once ranged across nearly all of the western portion of North America. As a result of widespread historic persecution, the current population has been reduced to only 2% of its historic abundance. Based on morphology there are two described subspecies C. l. arizonensis and C. l. ludovicianus. C. l. arizonensis once ranged throughout west Texas, southern New Mexico, Arizona, Sonora and Chihuahua; however, its range has been drastically reduced across each state and was extirpated from Arizona by the 1960s. Recently C. l. arizonensis was reintroduced back into Arizona. Thus our objectives included: determining conservation units within the species; examining whether the subspecies determined by morphology are supported by genomic data; and identifying best available populations to serve as a source for future translocations into Arizona. This is the first range-wide genetic study to date, with samples collected from Montana to Chihuahua. To address these objectives we used both nuclear SNPs and the cytochrome B locus of the mitochondria. Using a Genotyping-By-Sequencing (GBS) approach we discovered over 400,000 SNPs. After extensive filtering, a final dataset of over 10,000 SNPs was used for data analysis. The genomic data supports three distinct groups: Chihuahua; Sonora and southern New Mexico; and the entire rest of the range from northern New Mexico through the rest of the Great Plains states. The Great Plains group seemingly matches the C. l. ludovicianus subspecies according to its established range, while the other two groups represent a potential splitting of C. l. arizonensis. For future translocations into Arizona, prairie dogs from southern New Mexico or Sonora likely represent the most genetically similar source for future reintroductions. Future directions have us utilizing a hybrid capture array with museum samples to better determine potential boundaries or hybrid zones between these groups.
4:20PM Landscape Genetics of American Beaver within Coastal Oregon Basins
Jimmy Taylor; Vanessa Petro; Clinton Epps; Tyler Creech; Rachel Crowhurst
American beaver (Castor canadensis) dams provide over-winter habitat for coho salmon (Oncorhynchus kisutch) in the coastal ranges of the Pacific Northwest. While the beaver population in Oregon is considered healthy, most information on beaver dams in coastal Oregon is through anecdotal observations or reports of conflict (e.g., excessive flooding at plugged culverts). Recent research on beaver relocation into western Oregon suggested that survival of relocated beavers was low, beavers moved from release sites, and most relocated beavers did not build dams. Despite these results, some stakeholders believe that beaver relocation can be used to increase critical in-stream habitat for coho. We conducted a landscape genetics study of beavers to determine their genetic structure throughout the Coastal Coho ESU, and to identify if beaver colonies are isolated from stream networks by distance or landscape features. From 2014-2015, we collected 291 individual tissue samples from 23 coastal Oregon basins and genotyped them at 10 microsatellite loci. Watershed boundaries appeared to influence beaver genetic structure at multiple scales with the Coast Range, and slope or distance to water had negligible effects on gene flow. Individual genetic differences suggested that genetic neighborhood distances were <85km, implying that typical dispersal distances were significantly less than that distance. Genetic distances among watersheds were large in some cases (FST 0.02-0.19), implying greater differentiation among subpopulations than in other studies of beavers elsewhere in North America. The explanatory power of all resistance models tested was very low and may have been affected by historical beaver relocation efforts. Therefore, we were unable to model beaver gene flow in a predictive manner. While current beaver relocation may not be necessary, we caution that such actions should be conducted within watershed boundaries.


Contributed Paper
Location: Albuquerque Convention Center Date: September 27, 2017 Time: 1:10 pm - 5:00 pm