UC Davis

Temperature has a known potential to influence glucocorticoid concentrations obtained from fecal samples in nonhuman primates. Studies reliant on hair cortisol estimates obtained using samples from outdoor subjects, however, may not control for temperature. This omission is despite the general utility of hair as a sample matrix with relatively longer periods of accrual time. We examined these dynamics in rhesus macaques (Macaca mulatta); importantly, this population of rhesus macaques exhibits covariation between season and breeding behavior. Thus, we also examined temperature relative to contributions of social behavior associated with breeding seasons, which may coincide with climatic shifts. We utilized 1921 hair samples from 580 subjects across six large outdoor-housed mixed-sex groups at the California National Primate Research Center to quantify the capacity for warmer or cooler outdoor temperatures to influence hair cortisol concentrations. We found that colder maximum temperature estimates over the days preceding hair sampling were associated with elevated hair cortisol concentrations, relative to warmer periods. Temperature contributed similarly in a model with a reduced data set (1418 samples) which included breeding-associated social behaviors. Consortship behavior was associated with hair cortisol without temperature, but was not associated with temperature included. Aggression was associated with cortisol, with or without the inclusion of temperature. Outdoor temperature is an important confound or covariate to account for statistically or via careful study design. Inclusion is especially important among research projects reliant on hair cortisol from outdoor-housed primates and spanning multiple seasons.