We demonstrated the effect of cross-habitat foraging by sika deer (Cervus nippon) on plant communities under the hypothesis that the intensity of herbivory on a plant community is changed by the presence of a preferable habitat for deer nearby. To investigate this landscape-level effect, we examined two types of forest understory; "adjacent site" was located near agricultural fields where deer prefer to forage, and "remote site" was far from fields. We compared plant community structures between adjacent and remote sites in areas with high deer densities, and found that plant species richness and plant coverage were significantly higher in adjacent sites than in remote sites. We hypothesized that this difference was caused by the lower intensity of browsing at adjacent sites due to the higher use of fields by the deer at these sites. The following four results supported this hypothesis. First, in areas with no deer, plant species richness and plant coverage did not differ significantly between adjacent and remote sites. Second, we demonstrated a lower intensity of herbivory at adjacent sites by experimentally transplanting a preferred plant species, Aucuba japonica. Third, we detected no difference in the number of deer fecal pellets found in adjacent and remote sites, indicating that the difference in browsing intensity between the two sites was not due to differences in the frequency of site use by the deer. Fourth, fecal analysis showed that deer at adjacent sites consumed more graminoids, suggesting that deer at these sites used fields to forage because graminoids were abundant in fields. All of these results support the notion that the intensity of herbivory on forest understorys becomes lower in the presence of agricultural fields nearby. This also implies the importance of the indirect effects at the landscape level in that the two ecosystems are linked by the consumers moving between them.
Keywords: Herbivory; Indirect effects; Large herbivore; Numerical response.