In the unglaciated portion of southern Ohio, variation in forest composition is associated with a complex topographic and geologic mosaic. However, a shift in composition has been documented in oak-dominated forests throughout their range. Thus, despite the differences between individual oak-dominated forest stands in southern Ohio, one might expect to find similar regeneration patterns within their closed understories. I examined the variation in the regeneration layer of four, second-growth temperate forest fragments in the Bluegrass Region of southern Ohio that share similar land use histories, but differ in bedrock, topography, and overstory composition. At each site, I evaluated densities of woody seedlings in 30, circular (19.6-m2) plots. I also collected environmental data such as soils, slope, aspect, and canopy openness (using hemispherical photography) from each plot. Soils were analyzed for texture, pH, organic matter, N, P, and water holding capacity. The immediate overstory neighborhood was quantified using the basal area of trees > 2 cm dbh, measured in 0.01-ha plots nested around each seedling plot. Vegetation patterns were described using ordination (nonmetric multidimensional scaling). Relationships involving environmental data were explored via vector fitting.
Ordination distinguished four woody seedling assemblages that corresponded to the four study sites. However, the sites all shared similar relative abundance distributions. Further, the understory layers of three of these sites were dominated by shade tolerant species (mainly red maple and sugar maple). Oaks were present in the regeneration layer of the sites, but less abundant than maples. Sites also differed in environmental characteristics, but they shared similar relationships between vegetation community patterns and primary environmental drivers, especially nutrients, water stress, and light. On a cautionary note, there were also differences in species assemblages between the burn units established at each site. Therefore, when comparing vegetation response to burn treatments within and across such heterogeneous forest stands, it may be necessary to explore responses among tolerance or regeneration guilds rather than individual species.