Acessibilidade / Reportar erro

Analysis of structure of forest fragments in the Paulistano Plateau

The variation in the structure of forest fragments in the Paulistano Plateau, under different viewpoints and spatial scales, is described. A total census of the dominant arboreal component (GBH > or = 50 cm) in two areas (0.5 ha and 0.2 ha) inside a forest fragment, the Cidade Universitária "Armando de Salles Oliveira" (CUASO) Reserve, Butantã, São Paulo, São Paulo State, Southeastern Brazil (approximately 23°34'S, 46°43'W) included 227 individuals belonging to 50 species. A total absolute dominance (Do) of 20.1 m².ha-1, a Shannon index (H´) of 3.04 nats.ind-1 and Pielou´s evenness (J´) of 77.79% were found. Inside each area, regions of concentration of pioneer, secondary and climax species were found. The two areas showed differences in species composition and in the numeric contribution of climax species. Analysing 16 studies in the Paulistano Plateau, a significant Spearman correlation was found between inclusion criterion (GBH) and total absolute density (De), between number of individuals (N) and number of species (S), and between N and percentage of species with one individual (S1/S). Values of De, S and S1/S were corrected in order to neutralise the effect of GBH and N, using the residuals of curves adjusted by non-linear regression. After the correction, significant correlations were found among S, S1/S, H´ and J´, variables related to diversity, but not between De and Do or between any of both and any of the variables related to diversity, pointing to an independence between species diversity and physical structure of the analysed forests. The CUASO forest showed low diversity values relative to the other ones, tentatively explained by its history of succession in isolation.

Ombrophilous dense forest; fragments; structure; diversity


Sociedade Botânica de São Paulo Caixa Postal 57088, 04089-972 São Paulo SP - Brasil, Tel.: (55 11) 5584-6300 - ext. 225, Fax: (55 11) 577.3678 - São Paulo - SP - Brazil
E-mail: brazbot@gmail.com