Abstract
Density-dependent competition for food reduces vital rates, with juvenile survival often the first to decline. A clear prediction of food-based, density-dependent competition for large herbivores is decreasing juvenile survival with increasing density. However, competition for enemy-free space could also be a significant mechanism for density dependence in territorial species. How juvenile survival is predicted to change across density depends critically on the nature of predator-prey dynamics and spatial overlap among predator and prey, especially in multiple-predator systems. Here, we used a management experiment that reduced densities of a generalist predator, coyotes, and specialist predator, mountain lions, over a 5-year period to test for spatial density dependence mediated by predation on juvenile mule deer in Idaho, USA. We tested the spatial density-dependence hypothesis by tracking the fate of 251 juvenile mule deer, estimating cause-specific mortality, and testing responses to changes in deer density and predator abundance. Overall juvenile mortality did not increase with deer density, but generalist coyote-caused mortality did, but not when coyote density was reduced experimentally. Mountain lion-caused mortality did not change with deer density in the reference area in contradiction of the food-based competition hypothesis, but declined in the treatment area, opposite to the pattern of coyotes. These observations clearly reject the food-based density-dependence hypothesis for juvenile mule deer. Instead, our results provide support for the spatial density-dependence hypothesis that competition for enemy-free space increases predation by generalist predators on juvenile large herbivores.