A test of the predator avoidance hypothesis to explain delayed onset of communal breeding

Abstract

A neglected question in the study of communal breeding concerns why alloparental behaviour begins at variously late stages in the breeding cycle. In group-living corvids, the delay tends to be longer in species that are small and (or) typically have only a small nonbreeder complement. This pattern has been attributed to the relatively poor defensive capabilities of such species and their consequently greater need to minimize predator-attracting traffic to the nest or fledglings. We tested this predator avoidance hypothesis with the Canada Jay (Perisoreus canadensis (Linnaeus, 1766)), a species in which the feeding of young by any nonbreeders in the family group is delayed until the fledgling period. We reasoned that, on Anticosti Island, Quebec (Canada), in the absence of squirrels and other nest predators, nonbreeders might be permitted to feed nestlings as well as fledglings, and that breeders might feed nestlings more frequently (with smaller food loads) than on the mainland. We found no evidence for either prediction and thus no support for the predator avoidance hypothesis but suggest that Anticosti Canada Jays may have had insufficient time to evolve behaviour more appropriate for their predator-free environment. Secondarily, we confirmed that in all observed instances, the nonbreeders were offspring of the breeding pair from previous years and that they therefore failed to provision nestlings in spite of an apparent genetic interest to do so.