Spatial heterogeneity in post-dispersal predation on Prunus and Uvularia seeds

Abstract
We investigated effects of seed density, distance from parent, and habitat (woods, open field) on post-dispersal predation risk (chiefly by rodents) for seeds of Prunus virginiana (Rosaceae). Additional study of the habitat effect (woods, open field, treefall gap) was made with seeds of Prunus avium (Rosaceae) and Uvularia grandiflora (Liliaceae). Density of Prunus seeds (range 2–40 seeds/group) did not affect predation risk for individual seeds. Distance from parent plants did influence predation risk, which was greatest directly beneath parents. This distance effect primarily comprised a sharp drop in risk within 2 m of parents, a distance too small to generate a “spacing rule” for conspecifics. We found that habitat strongly influenced predation intensity. Rates of removal of Prunus seeds were higher in woods than in open fields, except when overall predation intensity was very low and no pattern could be discerned. Prunus seed removal rates were higher in closed woods than in treefall gaps. Consequently, a Prunus seed will more likely escape predation if dispersed to an open site. In contrast, Uvularia seed removal rates were higher in open fields than in woods but did not differ between closed woods and tree-fall gaps. Predation intensity was spatially patchy between and within experimental arrays, but was consistent over time at some specific points in space, possibly reflecting home ranges of seed predators.