Evaluating the Success of Deciduous Forest Restoration in Southwestern Ontario, Canada

Forest fragmentation has resulted in reduced richness of native species in northeastern North America. Despite recent large-scale increases in forest cover, studies indicated that understorey herbaceous plant communities may take decades to recover. In 1994 and 1995, we assessed the effectiveness of restoration by comparing the understorey plant composition of 28 former cottage and road sites with less disturbed reference sites at Point Pelee National Park, Ontario, Canada. Sites were restored over a 35-year period. We took into account the effects of selected environmental and landscape variables. There was a significant increase in similarity between restored and reference sites as time-since-restoration (TSR) increased. Overall, there were no signifcant differences in the diversity of native species between restored and relatively undisturbed reference sites suggesting recovery of restored sites may have been unsucessful. However, there was still significant among-site variation in the composition of the native species component of these plant communities. When only restored sites were examined, variation in native species composition was associated with time since restoration, soil moisture, canopy cover, and distance to continuous forest. Native species were assigned vulnerability rankings according to their relative occurence in reference and restored sites. Spring-flowering herbs such as Hepatica acutiloba, Dicentra cucullaria, and Allium tricoccum, with ant or gravity dispersed seeds, were absent from restored sites and were defined as highly vulnerable. In contrast, summer and fall-flowering herbs, with verterbrate and wind-dispersed seeds, dominated restored sites and were assigned lower vulnerability rankings. Species assigned low and intermediate vulnerability rankings had colonized restored sites successfully. These intermediately ranked species should function as indicators of recovery. In contrast, species with high vulnerability rankings had not recovered at all and, because of their limited disperal ranges, may only recolonize restored sites if they are actively reintroduced.