Biomarker of burden: Feather corticosterone reflects energetic expenditure and allostatic overload in captive waterfowl

Allostatic load describes the interplay between energetic demand and availability and is highly context dependent, varying between seasons and life-history stages. When energy demands exceed physiological set points modulated by glucocorticoid hormones, individuals may experience allostatic overload and transition between stages in sub-optimal physiological states. Corticosterone, the major glucocorticoid hormone regulating energy expenditure in birds, is incorporated into growing feathers (CORTf), and it has been suggested that CORTf reflects long-term records of allostatic load during feather growth. However, relationships between allostatic load and CORTf have not been adequately evaluated. If such relationships exist, the use of CORTf to investigate cross-seasonal effects could provide novel insights into impacts of past allostatic load and/or overload events. We tested whether experimental increases in daily workload during two adjacent life-history stages would be reflected in CORTf levels, and examined if CORTf levels reflected either current energetic demand or allostatic load prior to feather growth. Daily workloads in female mallard Anas platyrhynchos ducklings were increased over a 6-week period using physical obstacles and/or carrying back-mounted weights. We measured daily energy expenditure, growth, body mass, and CORTf in growing ducklings. Then, we induced feather moult and reapplied combinations of workload treatments for an additional 6weeks to investigate whether effects of past energetic demands would be detected in future CORTf levels. Ducklings confronted with higher workloads during development had reduced body mass, growth rates and consequently higher daily energy expenditure and CORTf values compared to controls. When ducklings were fully developed, CORTf patterns in birds re-exposed to workload treatments reflected only current, rather than past, energetic demands. However, under allostatic overload conditions, past levels of CORTf were positively associated with CORTf in the subsequent moult. Our study confirms the previously untested assumption that CORTf reflects energetic demand during the period of feather growth in a precocial bird. We show that allostatic overload conditions early in life, which temporarily suppress growth, can be detected using CORTf, an event potentially missed in studies which rely solely on measures of body condition alone. We suggest that CORTf can provide a valuable biomarker of allostatic load and overload conditions during the period of feather growth, but highlight how context should be considered for studies using CORTf to investigate influences of carryover effects. Our study contributes to building a physiological foundation to inform interpretations of ecological patterns using CORTf.