Background/Question/Methods Because ecological change is complex, having many components, and because we usually lack reliable baseline data and ongoing monitoring, it is difficult to track most ecological change. Legacy 1950's data from John Curtis and colleagues, however, provide a uniquely detailed picture of plant community change. Results/Conclusions Intact forest blocks in N Wisconsin lost native species, gained exotics, and homogenized over the past 50 years. Maples, ash, and aspen are replacing pines, hemlock, and yellow birch in the overstory, and exotics, graminoids, and ferns increased greatly in the understory (but these are uncoupled). Shifts across sites and species implicate overabundant deer as a major driver of these changes. More fragmented patches of forest in S Wisconsin are losing native species, gaining exotic species, and homogenizing even faster. Oaks are being replaced by black cherry, maples, and green ash in the overstory, and understories are losing rarer and more light-loving species, reflecting isolation, succession and fire suppression. Landscape factors (patch size, roads, and houses) strongly affect these trends while site factors (like soil) now play less of a role. Exotic earthworms and invasive shrubs also respond to landscape conditions and may be driving additional changes in these forests. Forest fragments may be particularly vulnerable to failed recolonization (the ‘rescue effect'), causing losses to accumulate through time and across the region. Because these changes are diverse, mediated by interactions among species, and cumulative, their influence accumulates through time in a way that is difficult or impossible to reverse. Such hysteresis presents obstacles to sustaining and restoring native forest diversity and implies that we should seek to conserve the landscape and disturbance conditions under which these species evolved.