Canalization of development reduces the utility of traits as fitness biomarkers: feather fault bars in nestling birds

Biomarkers that predict fitness are instrumental in unravelling mechanisms that link environmental conditions to fitness. However, development is likely to be better canalized for traits with stronger fitness effects. As a consequence, traits that are sensitive to developmental conditions may be poor fitness predictors and vice versa, and we tested for such effects using feather fault bars (translucent areas on feathers associated with developmental stress) in nestling birds. We manipulated developmental conditions (brood size) in free-living jackdaws (Coloeus monedula) and compared the effect on fault bar number between two feather types, tail and wing coverts. Fault bar number in the two feather types correlated poorly, indicating they can be considered different traits. Subsequently, we monitored local survival to investigate how well both traits linked developmental conditions to survival prospects. Tail fault bar number reflected developmental conditions better than the number of fault bars in the wings: when pre-manipulation brood size was small, nestlings had more tail fault bars when brood size was subsequently enlarged, while there was no such manipulation effect on the number of fault bars in wing coverts. In contrast, fault bar number in wing coverts better predicted fledgling local survival compared to tail coverts. We conclude that fault bar number in tail coverts is a biomarker of developmental conditions, while fault bar number in wing coverts is a predictor of survival prospects, but neither trait successfully links developmental conditions to fitness prospects. These findings are in agreement with the expectation that development of traits with a stronger relation to fitness will be better canalized. We discuss three hypotheses that can explain why despite canalization there are traits that are both sensitive to developmental conditions and predict fitness prospects.