Noninvasive measures of physiological stress are confounded by exposure

Abstract Glucocorticoids and glucocorticoid metabolites are increasingly used to index physiological stress in wildlife. Although feces is often abundant and can be collected noninvasively, exposure to biotic and abiotic elements may influence fecal glucocorticoid metabolite (FGM) concentrations, leading to inaccurate conclusions regarding wildlife physiological stress. Using captive snowshoe hares (Lepus americanus) and simulated environmental conditions, we evaluated how different realistic field conditions and temporal sampling constraints might influence FGM concentrations using an 11-oxoetiocholanolone-enzyme immunoassay. We quantified how fecal pellet age (i.e., 0–6 days), variable summer temperatures, and precipitation affected FGM concentrations. Fecal pellet age had a strong effect on FGM concentrations (βAge = 0.395, s.d. = 0.085; β2Age = −0.061, s.d. = 0.012), which were lowest at the beginning and end of our exposure period (e.g., meanday6 = 37.7 ng/mg) and typically highest in the middle (meanday3 = 51.8 ng/mg). The effect of fecal pellet age on FGM concentrations varied across treatments with warm-dry and cool-wet conditions resulting in more variable FGM concentrations relative to control samples. Given the confounding effects of exposure and environmental conditions, if fresh fecal pellet collection is not an option, we encourage researchers to develop a temporally consistent sampling protocol to ensure all samples are exposed to similar environmental conditions.