ON THE COST OF REPRODUCTION IN LONG-LIVED BIRDS - THE INFLUENCE OF ENVIRONMENTAL VARIABILITY

Life history theory suggests that animals should balance their current investment in young against their chances to reproduce in the future. One fundamental prediction from the theory is that long-lived species should be restrictive in any increase of their current investment. It has been suggested that long-lived species, therefore, have evolved a fixed level of investment in young in order to maximize their own adu lt survival. However, recent experimental studies have shown that long -lived seabirds have a flexible reproductive performance and adjust th eir effort in raising young, both according to their own body conditio n and to the need of the chicks. In this study, we present a model of the optimal balance between reproductive effort and adult survival for long-lived birds breeding in a stochastic environment. During poor br eeding conditions, maximum fitness is achieved either by not breeding at all, or by abandoning the brood. Beyond a certain threshold in bree ding conditions, there is a steep increase in reproductive effort and an equally steep decrease in adult survival. The model is applied to t wo hypothetical long-lived seabirds differing in their potential fecun dity. For the genotype with a potentially high fecundity, the model pr edicts a high threshold for breeding (i.e., breeding conditions need t o be very good for the species to attempt breeding); above the thresho ld, the value of reproduction in terms of fitness is high. For the gen otype with potentially low fecundity, the model predicts a low thresho ld for breeding, and the value of reproduction in terms of fitness is low. By increasing clutch size in the model, we examine the optimal re sponse of the two genotypes to an experimental brood size manipulation . For both genotypes, the model predicts that the threshold for breedi ng is lower among controls than among enlarged broods, giving a range of possible outcomes of the experiment depending on breeding condition s. The few studies on brood enlargements in long-lived species carried out so far may support the predictions from the model.