We examined fecal glucocorticoid (fGC) measures of nutrition and thermoregulatory demands on wild bears in Glacier National Park, Montana, and assessed how these measures changed in samples left in the field. Both ambient temperature and exposure can impact thermoregulation and sample degradation. Bear diets vary markedly with season, affecting body condition and thus fGC. We collected fecal samples during September and October, 2001, when ambient temperatures ranged from 30°C to −5°C. We collected half of each sample immediately and left the other half in its original location for 1–28 days. We used generalized linear models (GLM) to first predict fGC concentrations in fresh samples based on proxies of nutrition, ambient temperature, thermal exposure, and precipitation. These same covariates were then used to predict degradation-based differences in fGC concentrations between the paired sample halves. Variation in fGC was predicted by diet, Julian date, aspect, and the interaction between Julian date and aspect in both fresh and exposed samples. Cumulative precipitation was also a significant predictor of fGC concentrations in the exposed samples, independent of time, indicating that precipitation contributes to sample degradation but not enough to mask effects of other environmental factors on fGC concentrations. Differences between sample halves were only predicted by cumulative precipitation and exposure time; cumulative precipitation decreased, whereas exposure time increased, fGC concentrations in the exposed sample halves. Results indicate that fGC can provide reliable indices of nutrition and thermoregulatory demands in bears and that sample degradation impacts on these relations are minimal and can be virtually eliminated by controlling for cumulative precipitation over the estimated exposure times.

We report the first abundance and density estimates for American black bears (Ursus americanus) in Glacier National Park (NP), Montana, USA. We used data from 2 independent and concurrent noninvasive genetic sampling methods—hair traps and bear rubs—collected during 2004 to generate individual black bear encounter histories for use in closed population mark–recapture models. We improved the precision of our abundance estimate by using noninvasive genetic detection events to develop individual‐level covariates of sampling effort within the full and one‐half mean maximum distance moved (MMDM) from each bear's estimated activity center to explain capture probability heterogeneity and inform our estimate of the effective sampling area. Models including the one‐half MMDM covariate received overwhelming Akaike's Information Criterion support suggesting that buffering our study area by this distance would be more appropriate than no buffer or the full MMDM buffer for estimating the effectively sampled area and thereby density. Our model‐averaged super‐population abundance estimate was 603 (95% CI = 522–684) black bears for Glacier NP. Our black bear density estimate (11.4 bears/100 km2, 95% CI = 9.9–13.0) was consistent with published estimates for populations that are sympatric with grizzly bears (U. arctos) and without access to spawning salmonids. Published 2013. This article is a U.S. Government work and is in the public domain in the USA

Noninvasively collected hair samples have been used in numerous studies to answer questions about the demographic and genetic status and trends of wildlife populations. In particular, these methods are well-suited for researching and monitoring ursid populations, which are typically difficult to study because of their rare and cryptic nature. Recently, researchers have taken increasing advantage of natural bear behaviors to obtain hair samples for genetic analyses by conducting surveys of bear rubs (objects that bears rub against such as trees and power poles). The low quality and quantity DNA in noninvasively collected samples, however, can result in low genotyping success rates, which may be exacerbated by potentially lengthy duration of environmental exposure. We investigated the effects of environmental exposure (sunlight, moisture, and duration of exposure) on genotyping success rates of brown bear Ursus arctos and American black bear Ursus americanus hair samples. We exposed a total of 238 hair samples from one brown bear and one black bear to multiple treatments for either 30-d or 60-d, periods consistent with collection intervals of recent bear rub survey projects. Sample treatments consisted of full or dappled sunlight, kept dry or saturated with water one to two times daily. We genotyped each sample at three microsatellite loci commonly used in noninvasive genetic studies of bear populations. Our results were consistent with predictions, with all three factors significantly reducing genotyping success rates. Based on our results, we recommend that the specific conditions of field exposure be considered when selecting a suite of microsatellite markers for noninvasive genetic sampling projects, and that researchers carefully consider the duration and environmental conditions that hair samples will be exposed to when designing field studies. Limiting exposure to moisture and sunlight by collecting hairs from bear rubs at relatively short intervals and selecting dry and shaded sites should reduce DNA degradation and thus result in higher genotyping success rates.

No abstract