Effects of grazing by white-tailed deer (Odocoileus virginianus) on deciduous forest plant communities in southern Ontario, Canada: A multiple-scale evaluation

From 1992 to 1998, large-scale field studies were conducted at 10 forest sites in the Carolinian or deciduous forest eco-region of southwestern Ontario, Canada, to examine the effects of high deer populations on forest plant communities at multiple scales. Longer-term studies (1992–1998) focused on one forest site, Rondeau Provincial Park, were aimed at determining: (1) the effects of high deer populations on the trajectories of plant communities, (2) the effects of experimental manipulations of deer populations and abiotic variables on the recovery trajectories of over-grazed forest plots, and (3) whether seed bank composition of different plant functional groups provided a barrier to recovery of plant communities.; Constrained multivariate analyses of plant community and environmental data showed that light was the single most important abiotic variable associated with variation in plant community composition and forest structure when intensely deer grazed sites were compared with undisturbed reference sites defined as those with low deer densities. High deer densities caused changes in forest structure, primarily reductions in understorey and canopy woody stem densities, and these changes were significantly related to changes in light availability (increased light levels) and possibly nitrogen regimes. Seasonal shifts in the biomass of forest-floor plant communities observed in undisturbed sites differed when compared with sites with high deer densities. Patterns of nitrogen allocation in the herbaceous layer also changed. The nitrogen content (% N dwt) of herbaceous biomass was lower in the intensely grazed (disturbed) sites compared to the reference sites. However the amount of N per unit area was similar across all study sites. Mean total soil nitrogen content at each site was significantly correlated to mean tree density x DBH (diameter at breast height) for those sites.; In summary, this research demonstrated that, in order to better predict the changes in plant communities during and after disturbances, alterations in abiotic conditions, community dynamics and interactions via feedbacks must be taken into account. (Abstract shortened by UMI.)...