How do forest roads influence the stress response of black bears and grizzly bears?

Methods

Summary

We are using scat-detection dogs to collect black bear and grizzly bear scats in the Lower Blackfoot Valley, MT. For analyses of cortisol, we want only to collect scats deemed fresh; scats deposited within 72 hours. We will base scat freshness estimates on a set of in-situ environmental variables: moisture, odor, mold, exposure, contents, and weather. We will be collecting ~30g of the scats for cortisol analyses and ~75g for fecal DNA analyses. Cortisol samples will be placed in a vial and DNA samples will be placed in a brown paper bag. We will send fecal DNA samples to the USFS Rocky Mountain Research Station in Missoula, MT to be analyzed for sex, species, and individual ID's. We will conducted cortisol analyses at the University of Montana in Missoula, MT in the Millspaugh lab. The hormone analyses will consist of radio-immunoassays for cortisol metabolite concentrations. 

In 2020 through 2023, we will send hormone samples (cortisol, thyroid, and sex hormones) to the University of Washington in Seattle, WA to be analyzed, We will send DNA samples back to the USDA lab in Missoula, MT.

Challenges

Our biggest challenge with the 2019 pilot study was that we were limited by a small sample size (n=28,n=91). Although, we do plan on increasing our sample size in subsequent years with more intensive surveys. Another challenge associated with the first season is that cortisol analyses alone is generally not enough to get a comprehensive picture of a stress response. Therefore, we will also include thyroid hormone and progesterone analyses in our 2020-2023 analyses. 

Pre Analysis Plan

We evaluated the hypothesis that open roads cause a greater stress response than closed roads. If this is true, then we predict when an individual encounter an open road, we will observe hormonal abnormalities. We believe that when a bear encounters a closed road, the observed hormones will be significantly more stable. To further understand the stress response exhibited to roads, we will evaluate the relationship between thyroid hormones and progesterone and cortisol. It was our hypothesis on this matter that increased progesterone concentrations would be associated with successful reproduction. Given this success, we predict cortisol concentrations will be lowest, and TH should be higher than individuals with low progesterone concentrations and high cortisol. If the latter were the case, an individual should be considered stressed. Lastly, we predict that thyroid hormone levels will be significantly correlated with berry crop and mast success/failure. We tested these hypotheses by developing models to study the relationship between fecal glucocorticoid concentrations, thyroid hormone concentrations, and progesterone in bears as affected by road use restrictions, presence/absence of wolves, active or non-active timber harvest, and the interaction between these variables.

We will analyze fecal cortisol, thyroid, and progesterone metabolites as a function of study area, road type, and species. We will also test for presence/absence of wolves and timber harvest when appropriate. In our analyses, we will log-transform the concentrations to generalize the data. Linear regression analyses will be performed using a generalized linear model and we will use the Akaike Information Criterion corrected (AICc) estimator to select the best fit model of the relationship between our factor and response variables when considering the environmental variables.