Understory vegetation and site factors: implications for a managed Wisconsin landscape

We investigated relationships between edaphic and environmental factors (soil, forest floor, topography, and canopy) and understory vegetation (composition, richness, and Shannon-Wiener diversity index, H ') among 77 plots representing seven major patch types comprising a landscape in northern Wisconsin that has a long history of human management. Sampled patch types included. (1) mature hardwoods; (2) young hardwoods; (3) mature red pine: (4) mature Jack pine; (5) young mixed pine; (6) clearcuts; and (7) open pine barrens. Within each plot, percent cover of understory plant species was estimated, along with canopy cover (%), litter cover (%), coarse woody debris (%), duff depth (O-e and O-a horizons, cm), slope, aspect, and overstory data. Soil samples were taken and analyzed in the laboratory for pH, moisture, organic matter, nitrogen, and carbon. Multiresponse permutation procedures (MRPP) and stepwise multiple linear regression were used to examine differences in variables among patch types and to quantitatively relate the environmental variables to species richness and H '. Differences in species richness were only found between clearcut and hardwood patch types. Of the site variables, canopy cover differed most among patch types. Relationships between diversity and site factors were weak overall, but improved somewhat when patches were separated by overstory characteristics (i.e., coniferous, deciduous, open canopy), with soil factors differing in their relative effects on vegetation according to overstory properties. This suggests that a higher-level control pertaining to the overstory type overrode the effects of local variables on understory plant diversity. Important site variables influencing vegetation were canopy cover, pH, and forest door characteristics. DCA ordination separated patch types according to canopy cover, soil moisture, and duff depth gradients. Species composition showed greater differences among patch types than did quantitative measures of diversity; these differences occurred primarily between patch types that were less qualitatively similar in terms of overstory. Thus, managers should focus on composition rather than numeric diversity measures to maintain landscape-level plant diversity. (C) 2001 Elsevier Science B.V. All rights reserved.