NATURAL SELECTION ON AVIAN BREEDING TIME: CAUSALITY, FECUNDITY-DEPENDENT, AND FECUNDITY-INDEPENDENT SELECTION.

Natural selection on breeding time through postfledgling survival was investigated in a population of blue tits, Parus caeruleus, between 1993 and 1996. In particular, selection acting through large brood sizes among early breeders (fecundity-dependent selection) was separated from selection acting through fitness benefits unrelated to the numerical benefits of large brood sizes (fecundity-independent selection). There was evidence that natural selection favored early breeding in 1993 and 1995 (significantly so in 1993), whereas in 1994 there was a tendency for selection to be stabilizing. Analyses of selection gradients further showed that there was direct selection for both early breeding and large brood sizes in 1993. Experimentally delayed pairs produced a lower proportion of surviving young among their fledglings compared to early pairs, indicating that the seasonal fitness trend is a causal one and not caused by differences between early and late breeders in territory or individual quality. There was no evidence for significant heritability for four reproductive traits (clutch sizes, laying dates, hatching dates, and brood size at fledging) although repeatability values ranged from 0.26 to 0.53. The importance of year-specific selective regimes in the evolution of avian breedings times are discussed and a role for population density in this context is proposed.