Résumé
The activation of the hypothalamic-pituitary-adrenal axis is a neuroendocrine response to external and internal changes that animals face on a predictable or unpredictable basis. Across species, variation in glucocorticoid production has been related to such changes. In this study, we investigated the predictable, seasonal sources of variation in the levels of fecal glucocorticoid metabolites (fGCM) in a large natural population of mandrills (Mandrillus sphinx) in Southern Gabon. Materials and methods: Using five years of regular behavioral monitoring and hormone analyses performed on 1,233 fecal samples collected on 99 individuals of both sexes and all ages and General Linear Mixed Models, we studied the three main seasonal predictors of fGCM concentrations: (i) weather conditions, (ii) number of adult males, and (iii) female reproductive status. These three predictors all vary seasonally in mandrills. Results: We first showed an increase in fGCM concentrations during the short dry season while controlling for other factors. Pregnant females, which include the large majority of adult females at this time of the year, mainly drove this increase, although a combination of other small-magnitude, season-related effects linked to climatic events and demographic changes also partly explained this seasonal trend. Indeed, fGCM concentrations increased with both low temperatures (and low rainfall) and high numbers of adult males present in the group. These seasonal changes, while correlated, held true throughout the studied years and when restricting our analyses to a given season. Finally, we found that older mandrills showed on average higher fGCM concentrations than younger ones and that medium-ranked females exhibited the highest levels of fGCMs. Discussion: The observed patterns suggest that plasticity in mandrills' metabolism in the form of glucocorticoid production allows them to adjust to predictable changes in climatic, demographic and physiological conditions by mobilizing and redirecting energetic resources toward appropriate, calibrated seasonal responses.