Local Dynamics of Bird Assemblages in Small Forest Habitat Islands in Australia and North America

Abstract

Small isolated forest patches (0.1—7 ha) in eastern New South Wales and southern Wisconsin were surveyed regularly for at least 1 yr between 1977 and 1981. Forests and woodlands in both regions have been cleared extensively during the past century. This analysis examines how fragmentation of forest habitat has affected composition and dynamics of local bird assemblages. Species in forests islands are compared with those in equivalent "control plots" near the edge of a large, continuous forest. Disruption of continuous tracts apparently affects not only birds of the forest interior, but also those occurring along or near the edge. Birds in New South Wales and Wisconsin show common patterns of distribution among forest islands, despite considerable differences in taxonomic relationships, seasonality, and habitat. (1) Area was the best predictor of species richness in islands and control plots. Other habitat variables (except isolation) were relatively uniform among sites and did not contribute consistently to variation in numbers of species. (2) Species common in forest islands also tended to be common in large nearby forests, but bird assemblages in forest islands were statistically different from those in forest control plots. (3) Species assemblages in forest islands generally were more predictable over space and time than were local assemblages in larger forests. Although transients or regular visitors commonly were observed in forest islands, nonresident species occurred even more frequently in control plots. Apparently habitat fragmentation alters the spatial and temporal dynamics of native forest bird communities. This, in turn, may favor some species and leave others at a disadvantage. Differences between species assemblages of forest islands and control plots were most pronounced in Wisconsin. Numbers of forest or forest edge species were much lower in Wisconsin forest islands than in corresponding control plots. Isolation of forest islands was associated negatively with species richness only in Wisconsin. Long—term effects of forest fragmentation in New South Wales might not yet be realized, because relatively large areas of forest remain near existing habitat islands.