Seed-predator satiation and Janzen-Connell effects vary with spatial scales for seed-feeding insects.

Abstract

The Janzen-Connell model predicts that common species suffer high seed predation from specialized natural enemies as a function of distance from parent trees, and consequently as a function of conspecific density, whereas the predator satiation hypothesis predicts that seed attack is reduced due to predator satiation at high seed densities. Pre-dispersal predation by insects was studied while seeds are still on parent trees, which represents a frequently overlooked stage in which seed predation occurs. Reproductive tree density and seed production were investigated from ten Quercus serrata populations located in south-west China, quantifying density-dependent pre-dispersal seed predation over two years by three insect groups. Acorn infestation was nearly twice as high in the low-seed year as that in the high-seed year, with considerable spatio-temporal variation in the direction and magnitude of density-dependent pre-dispersal seed predation evident. Across whole populations of trees, a high density of reproductive trees caused predator satiation and reduced insect attack in the high-seed year. Within individual trees, and consistent with the Janzen-Connell model, overall insect seed predation was positively correlated with seed production in the low-seed year. In addition, there was variation among insect taxa, with positive density-dependent seed predation by Curculio weevils in the high-seed year and moths in the low-seed year, but apparent density independence by Cyllorhynchites weevils in both years. The overall trend of negative density-dependent, pre-dispersal seed predation suggests that predator satiation limited the occurrence of Janzen-Connell effects across Q. serrata populations. Such effects may have large impacts on plant population dynamics and tree diversity, depending on the extent to which they are reduced by counteracting positive density-dependent predation for seeds on individual trees and other factors affecting successful recruitment.