Effects of nest features and surrounding landscape on predation rates of artificial nests

Artificial nest predation experiments were carried out in northern Italy in woods which varied in size, isolation and surrounding landscape structure. Multivariate analyses, including logistic regression, showed that: (1) size and isolation of woods did not significantly affect predation rates; (2) nests on the edge of woods did not suffer higher predation rates than nests inside the wood; (3) nest camouflage greatly influenced predation rates, suggesting that predators were mainly using visual clues to identify nests; (4) the type of habitat that surrounded the woods emerged as a crucial factor in nest survival and the amount of human settlements in the vicinity of the wood was inversely correlated with nest survival, probably due to predators associated with humans; (5) other habitat variables, which were apparently individually unimportant, were found to have an effect on nesting success, if combined in a single 'suitability index'. It is impossible to generalize about the influence of landscape fragmentation on nest predation because local landscape history and predator guilds, together with the scale of fragmentation, probably interact to determine the suitability of nest sites and their vulnerability to predators.