Of the many causes of species loss, habitat destruction is the least unambiguous, because the species-area relationship indicates how many species will remain after a fraction of the habitat is destroyed. More ambiguous are the species losses underway where habitat remains intact. We investigated patterns of species loss in understory plant communities in intact forests in northern Wisconsin, relying on baseline data collected by the Plant Ecology Laboratory at the University of Wisconsin during the 1940s and 50s. At 62 sites, we examined changes in canopy composition over the 50 year interval. We also examined understory turnover, extinction, and colonization rates, and asked if these rates were influenced by initial species richness and shifts in canopy composition. We found about half of all stands sampled in 1950 maintained the same forest cover type in 2000, regardless of the initial cover type. Turnover rate per fifty years ranged from 0.102 to 0.370 (mean = 0.236). Extinction rates (mean = 0.290) exceeded colonization rates (mean = 0.145) at most sites (paired t = 7.96; df = 61; P < 0.001). ANCOVA revealed that initial forest cover type and the degree to which canopy composition changed did not affect turnover or extinction rates, but the number of species present in 1950 was positively correlated with both rates. ANCOVA indicated colonization rates were not influenced by initial forest cover type, the degree to which canopy changed, or initial species richness. The reported turnover, extinction, and colonization rates probably underestimate the degree to which forest understory communities change. The results indicate that forest understory herb communities in this region are not in equilibrium. Because extinction rates exceed colonization rates at most sites, species losses are underway.

Key words: colonization rates, community ecology, disequilibrium, extinction rates, passive sampling model, species loss, turnover