Effects of within- and between-patch processes on community dynamics in a fragmentation experiment

Abstract

The effects of the experimental fragmentation of native eucalypt forest on the beetle community were tested, in a controlled, replicated, long-term experiment. Included in our design were three fragment sizes, fragment edge and interior sites, and sites in the surrounding exotic pine plantation matrix. We followed 325 species through 28 sampling periods over seven years, including two years pre-fragmentation. We examined effects of fragmentation on four attributes of community structure: (1) species richness, (2) species composition, (3) relative abundance, and (4) the changes in occurrence of all species individually by the traits of rarity, degree of isolation (dispersal ability), and trophic group. We also considered how changes in these attributes altered community dynamics (turnover). We used both community-level and species-level responses to determine the relative importance of processes acting at the within-patch and between-patch scales. At the within-patch scale there were two findings. (1) There was no evidence of an increase in the extinction rate on fragments, as was hypothesized. Neither species richness nor the occurrence of rare species declined on fragments compared to continuous forest. (2) Edge effects altered species occurrences and abundances on fragments compared to continuous forest. There was evidence of two edge effects, with different penetration distances. Species richness increased at fragment edges in response to a shallowly penetrating edge effect. Species relative abundance and composition changed on fragments in response to a deeply penetrating edge effect, which also caused increases in the occurrences of detritivores and fungivores. At the between-patch scale there were three findings. (1) There was no evidence of a reduction in the colonization rate of fragments. There was no reduction in species richness or in the occurrence of individual species with poor dispersal abilities on fragments compared to continuous forest. (2) The matrix between fragments altered between-patch processes by providing alternative habitat for some species. These species increased in occurrence on fragments compared to continuous forest, supporting the predictions of recent metacommunity theory. However, the matrix did not act as a source of invading species. (3) Turnover was reduced in fragments compared to continuous forest. Thus, the effect of fragmentation was to stabilize community dynamics.