Density-dependent nest predation – an experiment with simulated Mallard nests in contrasting landscapes

Breeding success is a key element of animal population dynamics. In many taxa including birds, nest success, or the proportion of laid clutches that actually hatch, is mainly determined by predation. Previous research gives an inconsistent picture of the prevalence of density-dependent nest predation and one reason for this is the general lack of well-designed replicated experiments. Using simulated Mallard Anas platyrhynchos nests and a crossover design for 20 lakes in the nemoral and boreal biotic zones, we tested the predictions that nest survival is negatively density-dependent and that nest predation is higher in agricultural than in forested landscapes. Study day and daily abundance of waterfowl, other waterbirds, as well as avian predators were included as covariates in the analysis. Model fitting in program mark revealed a general negative effect of nest density on nest survival. In addition, nest survival rate was higher at forest lakes than at lakes in agricultural landscapes, irrespective of nest density. The only covariate producing model improvement was study day; older nests had higher survival rates than recently initiated ones. This is the first replicated lake-level experimental study showing that nest predation is density-dependent in waterfowl. The pattern was consistent between landscape types, implying that density-dependent nest predation may affect habitat choice and population dynamics over large parts of the Mallard's range.