9628937 Harrison Theoretical models predict that interspecific interactions, combined with limited dispersal, may sometimes produce stable spatial patterns in population densities. This study focuses on a herbivorous insects that tends to form dense, but very localized, populations within larger areas of its host plant. Previous work has established that this insects, the western tussock moth, is a very poor disperser, and that its spatial distribution does not appear to be explained by variable habitat quality. The present study examines two non-exclusive ways that the observed pattern could arise from interactions between the moth and its natural enemies: (1) Predator satiation, in which a poorly mobile, generalist predator (an ant) inflicts higher mortality on prey where prey are dense than where they are sparse, and (2) Predator diffusion, in which mobile, specialized natural enemies (two parasitoids) "spill over" around the edges of dense prey populations, creating surrounding zones in which predator-prey ratios, and therefore mortality, are elevated. Preliminary field results indicate both these mechanisms may be operating, but not whether they are strong enough to suppress the spread of outbreaks. This study uses field experiments combined with simulation and analytic modeling to test these explanations for spatial population patterns. Understanding the factors that govern population densities remains a central issue in ecology. In particular, discerning the causes of epidemic densities of pest species, should greatly enhance the ability to control future outbreaks and thus lessen the ecological and economic damage from such events.