Across eastern North America, there is a temporal trend from open Quercus forests to closed forests with increased Acer rubrum in the understory. We used a series of Ripley's K(d) analyses to examine changes in the spatial pattern of Quercus and Acer rubrum stems greater than 2.5 cm DBH over 45 yr in a 2-ha mapped stand. Specifically, we asked whether changes over time were consistent with the hypothesis that Quercus is being competitively replaced by Acer rubrum. Both Acer rubrum and Quercus stems are spatially clumped, but have become less clumped over time. Stem mortality from Hurricane Fran (1996) was more clumped in all strata of the forest, at all spatial scales, than expected if damage had occurred to stems at random. Acer rubrum ingrowth occurred more often near established trees (all species) in the midstory, whereas Quercus ingrowth occurred less often near established trees in the midstory. The specific hypothesis that stems of Acer rubrum in the midstory of the forest are associated with a lack of Quercus regeneration was strongly supported. This effect occurred at all spatial scales tested, including scales larger than that at which direct competition for light can occur. Edaphic gradients in the plot are correlated with many of the observed trends at large spatial scales, and our results suggest that the presence of such gradients can generate complex spatial patterns over time.

Nomenclature: Kartesz & Meacham (1999).

Abbreviations: DBH = Diameter at breast height; CSR = Complete spatial randomness.