Acta Botanica Brasilica
Print version ISSN 0102-3306
LEYSER, Gabriela et al. Tree regeneration and relationships with adult trees in a seasonal forest in the Uruguai river valley, Brazil. Acta Bot. Bras. [online]. 2012, vol.26, n.1, pp.74-83. ISSN 0102-3306. http://dx.doi.org/10.1590/S0102-33062012000100009.
Studies related to variation in plant communities have been involved the analysis of regenerative potential of tree species, because saplings may enable forests to restore themselves after natural or human disturbances. The goal of this study was to analyze tree regeneration by identifying composition and abundance, the ecological groups of dispersion, vertical stratification and light requirements for germination and to look for changes in expected richness, diversity and relationships with structure of adult trees. A sapling survey was conducted in 20 sample units of 10 x 10 m where all individual living trees were sampled that had a height > 0.30 m and diameter at soil height < 4.7 cm. Adult trees with a diameter at breast height > 4.7 cm were surveyed in 20 x 20 m sampling units, according to a previous study. We sampled 1,649 living trees (64 species), which generated a total density of 8,245 ind.ha-1. The species with higher absolute density were Gymnanthes concolor Spreng., Trichilia elegans A.Juss. and Calyptranthes tricona D.Legrand. A principal coordinates analysis indicated sharp differences among sampled units and a Mantel test showed no dependence among composition and abundance matrices from adult and regenerative surveys (r = 0.19; p = 0.1). According to the ecological groups, we found that the regenerative component has kept the same proportions of adults, with a higher proportion of zoochorous, light demanding, canopy covering species. Contrary to our initial hypothesis, we found lower expected richness in the regenerative component (p < 0.001) with the same equability in the entire area.
Keywords : ecological groups; forest structure; Mantel test; patchy dynamics; rarefaction curves; regenerative component.