Fragmentation affects plant community composition over time

Habitat fragmentation can lead to major changes in community composition, but little is known about the dynamics of these changes, or how community trajectories are affected by the initial state of habitat maturity. We use four landscape-scale experiments from different biogeographicregions to understand how plant community composition responds to fragmentation over decades. Within each experiment, we consider first whether plant communities in the most-fragmented treatments diverge in composition from plant communities in the least-fragmented treatments. Second, because communities embedded in different fragments may become more similar to one another over time (biotic homogenization), we asked whether beta diversity - compositional variation across space - declines among fragments over time. Third, we assessed whether fragmentation alters the degree to which temporal change in fragmented landscapes is due to ordered species losses and gains (nestedness) versus species replacements (turnover). For each of these three questions, we contrasted patterns of compositional change in mature communities following fragmentation (disassembly; n = 2 experiments) with patterns in newly-developing plant communities in fragments cleared of vegetation (assembly; n = 2 experiments). In the two studies where communities were disassembling, community composition in the most-fragmented habitats diverged from that in least-fragmented habitats. Beta diversity within a fragmentation treatment did not change over time at any of the four sites. In all four experiments, temporal patterns of compositional change were due mostly to species turnover, although nestedness played a role in the least-fragmented sites in two of the studies. Overall, the impacts on community composition varied among landscape experiments, and divergence may have been affected by the maturity of the plant community. Future comparisons across ecosystems that account for species identities (vs simply richness) will be critical for predicting the effects of fragmentation, managing mature plant communities in remnants, and restoring plant communities where habitat has been lost.