Neighborhoods have little effect on fungal attack or insect predation of developing seeds in a grassland biodiversity experiment.

Numerous observational studies have documented conspecific negative density-dependence that is consistent with the Janzen-Connell Hypothesis (JCH) of diversity maintenance. However, there have been few experimental tests of a central prediction of the JCH: that removing host-specific enemies should lead to greater increases in per capita recruitment in areas of higher host density or lower relative phylogenetic diversity. Using spatially randomized plots of high and low host biomass in a temperate grassland biodiversity experiment, we treated developing seedheads of six prairie perennials to factorial applications of fungicide and insecticide. We measured predispersal seed production, seed viability, and seedling biomass. Results were highly species-specific and idiosyncratic. Effects of insect seed predators and fungal pathogens on predispersal responses varied with neither conspecific biomass nor phylogenetic diversity, suggesting that-at least at the predispersal stage and for the insect and fungal seed predators we were able to exclude-the JCH is not sufficient to contribute to negative conspecific density-dependence for these dominant prairie species.