Predictors of resistance to brood parasitism within and among reed warbler populations

In the coevolutionary arms race between avian brood parasites and their hosts, several adaptations have evolved on both sides, the most prominent and important host defense being rejection of the parasitic egg. In the present study, we investigated possible predictors of egg rejection in 14 populations of reed warblers Acrocephalus scirpaceus across Europe differing in risk of parasitism by the common cuckoo Cuculus canorus, providing a test of factors associated with geographic variation in host resistance to parasitism. In a binomial general linear mixed model procedure, we quantified the possible influence of host clutch size, cuckoo parasitism in population (yes/no), height of the nest above ground, height of vegetation in the vicinity of the nest and distance to nearest vantage point on rejection of an experimentally added nonmimetic cuckoo sized egg. In addition, we entered "population" into the models as a random factor. Rejection rate varied significantly among populations (range 4.8-68.9%). The most parsimonious model, based on selection by the Akaike information criterion, included cuckoo parasitism in the population (yes/no) and host clutch size; rejection rate was the highest in parasitized populations, and individuals laying larger clutches were the best rejecters. Furthermore, rejecters tended to breed in higher vegetation than acceptors. These findings suggest that spatial variation in the level of host resistance to brood parasitism may depend on current and/or past selection pressure due to the parasite and individual differences linked to abilities for egg rejection.