Conservation implications of a lack of relationship between baseline glucocorticoids and fitness in a wild passerine.

The application of physiological measures to conservation monitoring has been gaining momentum and, while a suite of physiological traits are available to ascertain disturbance and condition in wildlife populations, glucocorticoids (i.e., GCs; cortisol and corticosterone) are the most heavily employed. The interpretation of GC levels as sensitive indicators of population change necessitates that GCs and metrics of population persistence are linked. However, the relationship between GCs and fitness may be highly context-dependent, changing direction, or significance, depending on the GC measure, fitness metric, life history stage, or other intrinsic and extrinsic contexts considered. We examined the relationship between baseline plasma corticosterone (CORT) levels measured at two periods of the breeding season and three metrics of fitness (offspring quality, reproductive output, and adult survival) in female Tree Swallows (Tachycineta bicolor). Specifically, we investigated whether (1) a relationship between baseline CORT metrics and fitness exists in our population, (2) whether the inclusion of energetic contexts, such as food availability, reproductive investment, or body mass, could alter or improve the strength of the relationship between CORT and fitness, and (3) whether energetic contexts could better predict fitness compared to CORT metrics. Importantly, we investigated these relationships in both natural conditions and under an experimental manipulation of foraging profitability (feather clipping) to determine the influence of an environmental constraint on GC-fitness relationships. We found a lack of relationship between baseline CORT and both short- and long-term metrics of fitness in control and clipped birds. In contrast, loss in body mass over reproduction positively predicted reproductive output (number of chicks leaving the nest) in control birds; however, the relationship was characterized by a low R (5%), limiting the predictive capacity, and therefore the application potential, of such a measure in a conservation setting. Our results stress the importance of ground-truthing GC-fitness relationships and indicate that baseline GCs will likely not be easily employed as conservation biomarkers across some species and life history stages. Given the accumulating evidence of temporally dynamic, inconsistent, and context-dependent GC-fitness relationships, placing effort towards directly measuring fitness traits, rather than plasma GC levels, will likely be more worthwhile for many conservation endeavours.