Stress-associated metabolites vary with both season and habitat across populations of a climate sentinel species

ABSTRACT Relating physiological stress to habitat quality could refine conservation efforts. Habitat quality, which is often inferred from patch occupancy or demographic rates, might be measured in a more timely and nuanced way using metrics of physiological stress. To understand whether stress-associated hormones vary with metrics of habitat quality, we measured fecal glucocorticoid metabolite (FGM) levels in the American pika (Ochotona princeps), a small mammal with well-defined habitat (talus), which can vary in quality depending on the presence of subsurface ice features. In spring and fall 2018, we collected feces noninvasively from pika territories in taluses “with” or “without” subsurface ice to capture seasonal variation in FGM between habitat types. We used linear mixed effects models to explore the interactions among season, habitat metrics (including subsurface ice status), and subsurface temperature as predictors of FGM. We found support for interacting effects on FGM levels, which covaried with season, elevation, putative ice presence, graminoid to forb ratio, graminoid cover, and measures of acute subsurface heat exposure. However, only one subsurface temperature metric differed according to putative presence of subsurface ice. Our results contribute to the growing evidence that FGMs might be developed as a tool to assess habitat quality.