Efeitos do isolamento e da perda de área de floresta sobre comunidades insulares de aranhas, Amazônia Central, Brasil

Abstract

The tropical forests are the terrestrial ecosystems with the largest number of species, but the high rates of deforestation have reduced these environments, modifying original landscapes. The results are the arising of forest fragments that consists of remnant areas of native vegetation surrounded by a matrix of altered habitat. Less area of forest, different shapes and degrees of isolation characterize these fragments. With the Landscape Ecology, these variables have been measured through indexes and have been correlated with the biotic variables. In this sense, we have evaluated the relationship among the landscape metrics and the abundance, richness and community composition of spiders. Study area was the Reserva Biológica do Uatumã and surrounded area, where there are islands, formed after the construction Balbina Hydroelectric, and the continuous forest, adjacent the lake. The study was realized in both environments, in 20 islands and seven sites in a continuous forest, between July and October/2006. The night visual search was standardized in one hour and during the day was done by beating. The size, the perimeter-area relation and the proximity index of each island were calculated using Arc View and Fragstats programs. The effects of those variables on richness and abundance of spiders were tested using multiples regressions. The spider’s community was represented by two axis of ordination – NMDS, which were used in multivariate multiple regressions. A comparison between islands and continuous forest was realized too. A total of 9.550 spiders were collected, being 2.187 were adults, belonging to 33 families, 106 genus and 334 species. The landscape metrics did not affect spiders' abundance, while the proximity index affected the richness. The first axis of ordination was significantly influenced by the size and the proximity index, while the perimeter-area relation affected the second one. Spiders' abundance was significantly higher in the islands and at contrary to richness. The community composition was also different between these environments. Similarity was greater in continuous forest than islands. We conclude that the community composition of spiders was affected by variables, rather than abundance and richness of species. Considering that some species become more abundant in modified habitat, the richness may be the same and the species completely different along the time.