Evolution of resident bird breeding phenology in a landscape with heterogeneous resource phenology and carryover effects

It is generally expected that, in environments with pronounced seasonal resource peaks, birds’ reproductive success will be maximised when nestlings’ peak food demand coincides with the timing of high food availability. However in certain birds that stay resident over winter, earlier breeding leads juveniles to join the winter flock earlier, which by the prior residence effect increases their success in breeding territory competition. This trade-off between reproduction and competition may explain why, in certain species, breeding phenology is earlier and asynchronous with the resource. This study extends a previous model of the evolution of breeding phenology in a single habitat type to a landscape with two habitat types: ‘early’ and ‘late’ resource phenology. The offspring’s natal habitat type has a carryover effect upon their competitive ability regardless of which habitat type they settle in to potentially breed. We find that, when the difference in resource phenology between habitats is small (weak carryover effect), breeding phenology in the late habitat evolves to occur earlier and more asynchronously than in the early habitat, to compensate for the competitive disadvantage to juveniles raised there. However if the difference is large (strong carryover effect), then the reproductive cost of earlier breeding outweighs the benefit of the compensation, so instead breeding phenology in the late habitat evolves to become more synchronous with the resource. Recruitment is generally asymmetric, from early to late habitat type. However if the early habitat is less frequent in the landscape or produces fewer offspring, then the asymmetry is reduced, and if there is some natal habitat-type fidelity, then recruitment can have an insular pattern, i.e. most recruits to each habitat type come from that same habitat type. We detail the different scenarios in which the different recruitment patterns are predicted, and we propose that they have implications for local adaptation.