Species composition drives ecological stability in dry forests of the Great Lakes-St. Lawrence forest region of Ontario, Canada

The effect of species diversity on the stability of ecosystem functions such as productivity, within all spatial and temporal scales, has long been debated in ecology. Extensive research in grassland ecosystems has suggested a positive diversity effect on stability at the community-level due to the insurance hypothesis, but a negative effect at the population-level due to interspecific competition. The few experiments that have investigated the diversity-stability relationship within forest ecosystems have generated inconsistent results, and general conclusions for these systems are currently lacking. Predicted increases in the frequency of short term drought events in the Great Lakes-St. Lawrence (GLSL) forest region of Ontario, Canada, raise concerns about potential effects of climate change on forest ecosystems. The knowledge gap regarding the importance of diversity and composition effects on tree growth in the deciduous-coniferous forest zone of the GLSL must be addressed to anticipate and mitigate some of the effects of climate change in these forests. My dendrochronological research tested the hypotheses that increasing species diversity increases the community-level stability and decreases the population-level stability to a severe short-term drought event in forest tree assemblages typical of dry forests in the GLSL forest region. The assemblages that were studied spanned a gradient of 15 different combinations of species richness and composition of four tree species. The drought that was studied occurred in 2005 and was characterized by high temperature, low precipitation and reduced tree growth. A total of 63 plots representing replicate assemblages were selected to collect increment cores (n=1193) and reconstruct growth responses to past drought at the community and population-levels. The stability of growth of the assemblages was determined by using crossdated tree-ring measurements to reconstruct basal area increment (BAI) and calculate standard indices of resistance, resilience and productivity of the tree communities and populations to the drought. Statistical analysis identified a significant compositional effect on community-level resistance and productivity. White pine ( Pinus strobus L.) abundance was associated with a lower community resistance and white birch (Betula papyrifera Marsh.) abundance with a higher resistance. Assemblages with greater productivity were often characterized by the high abundance of trembling aspen. There was no overall effect of the assemblage gradient on the population-level stability; however, red pine productivity was higher in combination with trembling aspen ( Populus tremuloides Michx.) and white birch. If pine forests were managed to maintain a component of deciduous species, the capacity to dampen the community-level effects of more frequent drought would increase.