Endocrine phenotype, reproductive success and survival in the great tit, Parus major.

A central goal in evolutionary ecology is to characterize and identify selection patterns on the optimal phenotype in different environments. Physiological traits, such as hormonal responses, provide important mechanisms by which individuals can adapt to fluctuating environmental conditions. It is therefore expected that selection shapes hormonal traits, but the strength and the direction of selection on plastic hormonal signals are still under investigation. Here, we determined whether, and in which way, selection is acting on the hormones corticosterone and prolactin by characterizing endocrine phenotypes and their relationship with fitness in free-living great tits, Parus major. We quantified variation in circulating concentrations of baseline and stress-induced corticosterone and in prolactin during the prebreeding (March) and the breeding season (May) for two consecutive years, and correlated these with reproductive success (yearly fledgling number) and overwinter survival in female and male individuals. In both years, individuals with high baseline corticosterone concentrations in March had the highest yearly fledgling numbers; while in May, individuals with low baseline corticosterone had the highest yearly reproductive success. Likewise, individuals that displayed strong seasonal plasticity in baseline corticosterone concentrations (high in March and low in May) had the highest reproductive success in each year. Prolactin concentrations were not related to reproductive success, but were positively correlated to the proximity to lay. Between-year plasticity in stress-induced corticosterone concentrations of males was related to yearly variation in food abundance, but not to overall reproductive success. These findings suggest that seasonally alternating directional selection is operating on baseline corticosterone concentrations in both sexes. The observed between-year consistency in selection patterns indicates that a one-time hormone sample in a given season can allow the prediction of individual fitness.