Fifty years of change in southern Wisconsin forests: Patterns of species loss and homogenization

We use quantitative surveys from the UW Plant Ecology Lab in the 1950's to infer long-term shifts in species composition of southern Wisconsin forests. Succession has shifted tree composition away from oaks (Quercus) and towards more mesic species (Acer spp.). More than 80% of sites lost native understory species richness with an average decline of 22.4%. Species losses are almost twice as high in late successional stands than early successional stands with relatively shade intolerant taxa showing the most conspicuous declines. Rates of native species loss, floristic quality and community homogenization are negatively correlated with measures of forest fragmentation and urbanization. Current land use patterns also influenced rates of species loss, with lack of deer management emerging as a key predictor of species loss. Invasion of both native and exotic species both respond more to surrounding landscape features than to local site factors, but in opposite ways. Declines in recruitment in more fragmented landscapes is more important than extirpation in explaining overall changes in native species richness. Exotic species invasions are strongly dictated by human dominance of the landscape regardless of local site conditions. Thus, both native species diversity and exotic species invasions reflect the surrounding species pool and their respective abundances and dispersal abilities.; Finally, we evaluate how gradients in local site conditions and surrounding landscape composition influence the composition of southern Wisconsin forests and how these relationships have changed over time. Overstory composition in both time periods strongly reflects underlying moisture and nutrient gradients despite significant successional changes. In contrast, the proximal drivers of native herb species composition have shifted from local overstory and edaphic factors to metrics of landscape fragmentation and urbanization. Species persistence and abundance patterns appear to reflect meta-population and meta-community dynamics in fragmented landscapes more than local environmental conditions. These results strongly support the idea that to sustain plant diversity in the face of landscape fragmentation and urbanization, we need to think and manage at landscape scales, protecting larger blocks of habitat from development and re-establishing landscape connectivity between isolated populations.