The effects of forest fragmentation on tree-hole nesting solitary wasps and bees were investigated at a site 90 km north of Manaus, Brazil. Wasp and bee faunas were monitored in continuous terra firme forest, forest fragments of 1, 10 and 100 ha, natural gaps in continuous forest and deforested areas. These habitats were studied in terms of abundance, richness, diversity and similarity. The wasps and bees were monitored monthly during June 1988 through June 1990 by means of a trap-nests technique. A total of 1529 nests of wasps of 24 species and 405 nests of bees of 14 species were collected. The number of species of wasps and bees varied little among the habitats. The genus Trypoxylon Latreille, 1796 (Sphecidae) accounted for 79% of wasp nests and Centris Fabricius, 1804 (Anthophoridae) for 56% of bee nests. Wasps showed an overall preference for cleared areas and fragments of 1 ha, whereas bees showed an overall preference for continuous forest and natural gaps. Nevertheless, some species of wasps showed a preference for nesting in continuous forest and some bees a preference for deforested areas. Species found in deforested areas also nested in small size forest fragments. This shows that species occurring predominantly in cleared áreas can also colonize small forest fragments. The diversity of wasps and bees was greater in continuous forest. The composition of wasp and bee faunas of continuous forest was different from that of altered habitats. The similarity between the natural gaps and cleared areas was the smallest. The data suggest that the forest adapted bees are more sensitive to habitai fragmentation than the corresponding species of wasps. It is concluded that the preexisting cavity nesting solitary wasp and bee communities were altered by the forest fragmentation. However, it appears that at least in part, the natural variation in continuous forest could be responsible for the resulte obtained from this study.
Amazonia; forest fragmentation; solitary bees; solitary wasps; trap-nests