Regulation of incubation shifts near hatching by giant petrels: a timed mechanism, embryonic signalling or food availability?

In most seabird species incubation shifts shorten when hatching approaches, a behavioural response allowing the chick to be fed soon after hatching. Three mechanisms have been proposed to explain these shortened absences: endogenous timing, a response to embryo signals or a seasonal increase in food availability. I tested these hypotheses by cross-fostering eggs between two sibling species: the northern Macronectes halli and the southern M. giganteus giant petrels; the former normally breeds 5-6 weeks earlier than the latter but both have an incubation period of 60 days. The length of the last trip before hatching was unaffected by the timing of hatching, suggesting that trip length was not related to the increase in food availability with the progress of the season. Furthermore, northern giant petrels reduced their last trip lengths regardless of whether they were incubating their own developed or the southern undeveloped egg (with embryos not developed enough to signal). Parents presumably, therefore, possess endogenous control allowing them to predict hatching date and to reduce trip duration accordingly, regardless of the communication with the embryo. Nevertheless, shorter last trips before hatching in southern giant petrels incubating northern giant petrel eggs suggest that parents use embryo signals for a fine-tuning synchronization of the internal timer to the natural variability in the length of the incubation period. (C) 2004 The Association for the Study of Animal Behaviour