Brood‐guarding duration in black‐browed albatrosses <i>Thalassarche melanophris</i> : 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.