Neighbourhood effects modify deer herbivory on tree seedlings

As spatially fine-scale foraging decisions are intrinsic to browsing herbivores such as deer, the search and selection of individual plants may be shaped by the density of neighbouring plants. However, how such neighbourhood effects may modify herbivore foraging decisions at different spatial scales is largely unexplored. Our main goal was to examine deer foraging preferences for tree species and seedling sizes, and to ask whether the spatial relationship of unbrowsed and browsed seedlings was indicative of different neighbourhood effects at the spatial scale of seedling neighbourhoods. We used two data sets from an old-growth temperate forest: (1) a seedling inventory and (2) fully mapped transect data, and applied a generalized linear mixed model and spatial point pattern analyses, respectively. We found that neighbourhood effects modified deer foraging behaviour depending on the proximity, density and species identity of neighbours, which, in turn, determined the strength and spatial extent of browsing impacts on tree seedlings. While sycamore maple seedlings experienced the highest levels of deer selectivity, deer were selective for European beech seedlings only within short distances. Beyond these distances, beech seedlings were browsed only opportunistically, as conspecific neighbours ceased to modify deer selectivity. Beech seedlings were also more susceptible to browsing within short distances from silver fir seedlings, indicating an associational susceptibility. As fir seedlings experienced strong and extensive density-dependent deer selectivity both near conspecific and beech neighbours, using such neighbourhood effects may not be a promising tool to prevent deer browsing on tree seedlings at the spatial scale of seedling neighbourhoods.