FOREST SIZE AND STRUCTURE - COMPETITIVE AND PREDATORY EFFECTS ON SMALL MAMMAL COMMUNITIES

This study investigated how changes in forest fragment size and structure affect small mammal community structure in the Brazilian Atlantic forest. Forest fragments were of three sizes (small: 60-80 ha; medium: 860 ha; large: approximately 36 000 ha) and two disturbance conditions, primary and secondary. Forests were subject to a 17-month capture/recapture program for small mammals ( < 1000 g). The total trapping effort was of 57 120 trap nights, 9520 at each of six forests. During the course of the study, 692 individuals, belonging to 19 small mammal species, were captured 1366 times. The highest number of species (14) and total number of individuals trapped (190) were obtained at the large secondary forest. Both the small and the large primary forests had fewer species (6) and fewer individuals (70 and 57, respectively). The medium-size forests and the small secondary forest had 9 species each. Discriminant analysis indicated that differences in species composition of small mammal communities were not adequately explained by microhabitat preferences, with two exceptions. Species interactions were also poor predictors of community structure and composition, with the noted exception of interactions between the common opossum Didelphis marsupialis and other species. Smallerforest plots, especially among secondary habitat, are dominated by D. marsupialis, which are apparently able to exclude other species from these small mammal communities. It is postulated here that the lack of predators in smaller forests is the factor responsible for the observed lower species richness and diversity of these communities. © 1990.