Tree species composition and similarity in semideciduous forest fragments of southeastern Brazil |
| |
Authors: | Karin dos Santos Luiza Sumiko Kinoshita |
| |
Affiliation: | a Post-Graduate Program in Plant Biology, Department of Botany, Institute of Biology, State University of Campinas, CP 6109, Campinas, SP 13083-970, Brazil b Department of Botany, Institute of Biology, State University of Campinas, CP 6109, Campinas, SP 13083-970, Brazil |
| |
Abstract: | This study investigates the influence of fragment size on tree species composition, species richness and on individual groups of species within 11 semideciduous forest fragments in southeastern Brazil. We compared same-sized samples of 500 trees from 10 fragments <100 ha, allocated to three size categories, with three 500-tree samples collected in different areas within a large forest, used as a reference forest. The variation in species richness, in the proportion of species/individuals in dispersal modes, and in the proportion of rare species was not related to fragment size. Nonetheless, comparisons between the mean values of these variables of each fragment size categories, using 95% confidence intervals, showed a wider range of variation within the category of small-sized fragments than in the other size categories. NMS ordination plot also suggested the absence of a relationship between fragment area and tree species composition. However, multiple regressions using the scores from the NMS ordination as response variables, and area and disturbance intensities as predictor variables, suggested that the interaction of area and disturbance might be a good predictor of species composition. Pre-existing environmental heterogeneity and geographical proximity also appeared to play a role in the variations of forest composition among the fragments studied. Our results demonstrated the high conservation value of small forest fragments. |
| |
Keywords: | Habitat fragmentation Fragment area Conservation Disturbance Species loss NMS ordination |
本文献已被 ScienceDirect 等数据库收录! |
|