Incidental nest predation in songbirds: Behavioral indicators detect ecological scales and processes

Incidental predation occurs when secondary prey items are encountered and subsequently consumed, not through directed search for such prey, but through their consequential encounter by a predator engaged in search for primary prey. We developed a mathematical model that examines the relationships between the abundance of primary prey, patch exploitation (i.e., quitting harvest rates), and the rate of incidental predation on secondary prey items. The model's predictions are dependent upon the spatial scale over which a forager integrates foraging costs and thus determines its quitting harvest rate (QHR). At local (i.e., foraging) spatial scales, we predicted that incidental predation should increase with local food abundance. Also at the foraging scale, local food abundance should not influence QHRs, but local predation risk (from higher trophic levels) should increase QHRs. Therefore, we predicted that incidental predation rates should be negatively correlated with QHRs. Over large (i.e., landscape) spatial scales, greater food abundance and predation risk increase QHRs, and we predicted that predation rates should vary inversely with QHR through two complementary mechanisms: foragers use a greater proportion of space and spend more time foraging as quitting harvest rates decrease. We experimentally tested the qualitative predictions of the theory in the field using artificial Veery (Catharus fuscescens) nests depredated by white-footed mice, Peromyscus leucopus, across three spatial scales. We used the technique of giving-up densities to measure QHRs and to determine the scale at which mice integrate different foraging costs. In accord with our predictions, nest predation was positively influenced by the local abundance of food at the foraging scale, and local predation risk to mice and perhaps interference competition from chipmunks resulted in higher giving-up densities and lower nest predation. At the landscape scale, there was an inverse relationship between giving-up densities and nest predation, which was probably the result of large-scale differences in resource abundance between plots. Our study demonstrates how linking theoretical development to the use of empirical behavioral indicators can help determine the relevant ecological scales and processes necessary for understanding predator-prey interactions.