Do deer and raccoons defecate in the right place? Fitness consequences of vertebrate seed dispersal for a deciduous forest herb.

Abstract

Precision of seed placement in a heterogeneous environment is often assumed to select for the evolution of animal-mediated dispersal systems, but this hypothesis has rarely been tested in a multivariate sense. We quantify the microsite fitness benefits of dispersal by white-tailed deer (Odocoileus virginianus) and raccoons (Procyon lotor) for mayapple (Podophyllum peltatum), a shade-tolerant perennial herb, in deciduous forests of southeastern Ohio, USA. Micro-environmental variables were recorded at dung-deposition microsites, at rooting points of mayapple shoots, and at random (control) points in the forest. Fitness was assessed as the degree of overlap in ordinations of microsites by environmental variables. Mayapple occupied a broad sector (56%) of environment space corresponding to low and mid-slope positions, ravines, and proximity to trees. Deer and raccoon defecation placed dung in 71-74 and 86-95% of environment space, respectively, reaching mayapple microsites in 57-60 and 53-54% of cases. Deer placed dung in mayapple environment space 7-9% more often than predicted by random distribution, and raccoons placed dung in mayapple space 0-5% more often, consistent with only a modest degree of directed dispersal. Thus, the precision hypothesis is only weakly supported. The greatest fitness benefit of vertebrate dispersal appears to be the broad distribution of seeds, thereby increasing the probability of randomly reaching a suitable microsite. Imprecise dispersal suggests that secondary mechanisms of seed movement need to be explored in deciduous forest communities.