A Gradient Analysis of Southern Wisconsin Forests

Abstract

While study of southern Wisconsin forests has led to numerous conceptual advances in ecology, a clear understanding of variation in these factors in relation to environmental factors and successional processes has failed to emerge. The present study was undertaken to (1) determine the principal factors controlling or closely correlated with forest composition, (2) formulate a multidimensional gradient representation using these factors, and (3) reexamine forest composition and dynamics using this representation. It was also expected that this study might help place in perspective the extensive early research on Wisconsin forests and subsequent methodological advances in gradient analysis and ordination. Principal component analysis was used to represent compositional variation of southern Wisconsin forests. Analysis of environmental variation within the resultant ordination suggested two gradients to account for much of the variation observed. These were a moisture—nutrients gradient based on soil texture and a dynamics gradient based on tree size—class data. These gradients, reformulated, serve as axes in a two dimensional scheme of forest characterization. Within this characterization scheme, early successional forests on mesic to xeric sites respectively are dominated. by Quercus rubra, Quercus alba, Quercus macrocarpa, and Quercus velutina. On mesic sites succession proceeds through stands dominated by Ulmus rubra to dominance by Acer Saccharum. Dry—mesic stands appear to have dominance shared at climax by Q. alba, Carya ovata, and Prunus serotina. On extreme xeric sites. Q. macrocarpa and Q. velutina tend to form stable stands. Thus, the successional sequence becomes progressively shorter toward the xeric end of the moisture—nutrient gradient. The original continuum of Curtis and McIntosh was an abstract gradient implying no direct relationship to either site condition or successional status. Use of such factors in the present characterization allows a more precise conceptualization of functional relationships between forest types. In addition the methodology illustrates the complementary usage of mathematical ordination as an exploratory tool and gradient analysis as both a means of representing the relationships discovered and as a framework within which additional critical questions can be formulated and addressed.