Partitioning the colonization and extinction components of beta diversity across disturbance gradients.

Changes in species diversity often result from species losses and gains. The dynamic nature of beta diversity (spatial variation in species composition) that derives from such temporal species turnover, however, has received relatively little attention. Here, we disentangled extinction and colonization components of beta diversity by using the sets of species that went locally extinct and that newly colonized the study sites. We applied this concept of extinction and colonization beta diversity to ground vegetation communities that have been repeatedly surveyed in forests where fire and harvesting were experimentally applied. We first found that fire and harvesting caused no effect on beta diversity 2 yr after the treatments. From this result, we might conclude that they did not alter the ways in which species assemble across space. However, when we analyzed the extinction and colonization beta diversity between pre-treatment and 2 yr after the treatments, both measures were found to be significantly lower in burned sites compared to unburned sites (i.e., the groups of excluded and newly colonized species both showed low beta diversity in the burned sites). These results indicate that the fire excluded similar subsets of species across space, making communities become more heterogeneous, but at the same time induced spatially uniform colonization of new species, causing communities to homogenize. Consequently, the effects of these two processes canceled each other out. The relative importance of extinction and colonization components per se also changed temporally after the treatments. Fire and harvesting showed synergetic negative impacts on extinction beta diversity between pre-treatment and 10 yr after the treatments. Overall, analyses using extinction and colonization beta diversity allowed us to detect nonrandom disassembly and reassembly dynamics in ground vegetation communities. Our results suggest that common practices of analyzing beta diversity at one point in time can mask significant variation driven by disturbance. Acknowledging the extinction-colonization dynamics behind beta diversity is essential for understanding the spatiotemporal organization of biodiversity.