Brood-guarding duration in black-browed albatrosses Thalassarche melanophris: temporal, geographical and individual variation

In birds, the period spent brooding or guarding young chicks is highly variable, but such variation has seldom been studied. Previous single-year studies of Antarctic petrels Thalassoica antarctica and grey-headed albatrosses Thalassarche chrysostoma revealed a pronounced seasonal decline in brood-guarding duration and gave rise to the 'synchronisation hypothesis', which suggests that some of the variation in the length of the brood-guarding stage is related to predictable seasonal changes in the risk of chick predation. We tested the predictions of this and three other hypotheses in a two-site, four-year study of the black-browed albatross T. melanophris. The existence of a pronounced seasonal decline in brood-guarding duration was apparent at both sites, and in years of contrasting food availability, providing further support for the 'synchronisation hypothesis'. Alternative explanations for this pattern are that short brood-guarding periods for late-hatched chicks result from a seasonal decline in food availability or from the fact that early nesting birds are of higher individual quality. However, these explanations are at odds with the absence of a seasonal decline in early chick growth or in probability of chick survival. Furthermore, adult quality (measured as past reproductive performance) had a weak and inconsistent effect on the duration of brood-guarding. Weather changes explained some of the variation in brood-guarding, but there were no differences between regions of contrasting climates. Individual pairs displayed a degree of inter-annual consistency in brood-guarding duration and, at least in some years, longer brood-guarding resulted in higher fledging probability. We speculate that a higher investment in brood-guarding increases the cost of reproduction, which counteracts other selective pressures that would otherwise lead to longer brood-guarding durations.