Use of landscape metrics to predict avian nest survival in a fragmented midwestern forest landscape

Abstract

Habitat fragmentation fundamentally affects trophic interactions and ecosystem function. Understanding how the landscape matrix modulates such interactions can improve our understanding of fragmentation ecology. Studies of breeding birds provide clear examples of the consequences of habitat fragmentation, but the landscape context of these effects are unclear. We sampled avian nesting success in 12 moderately-large forest patches (>250 ha) embedded in different types of landscapes in southern Illinois, USA. We then evaluated eight models that predicted the probability of nest success and brood parasitism by Brown-headed Cowbirds. These models incorporated landscape composition (% grassland, % agriculture, fragmentation), year and seasonal effects, conspecific density, predator density, and combinations of these variables. Temporal factors (stage of nesting cycle, seasonal effects, annual variation) had the most effect on nesting success; landscape factors had little influence on nesting success. The rate and intensity of brood parasitism were significantly influenced by the amount of grassland for the Wood Thrush, but not for the Acadian Flycatcher. Fine-scale management of the matrix surrounding the patches may dictate the local abundance and movements of nest predators and parasites. Other major nest predators may prefer the forest interior and at least partially compensate for the lower abundance of nest predators that depend upon the matrix. Overall, landscape metrics were weak predictors of avian nesting success in complex landscapes that have diverse predator communities.