Vegetação e mecanismos de regeneração em fragmento de floresta estacional decidual ripária, Cachoeira do Sul, RS, Brasil

Abstract

Riparian forests are complex ecosystems closely associated with river dynamics and occurring under particular phytogeographic conditions. In the state of Rio Grande do Sul, these forests have been altered and fragmented by human activity. The objective of this study is to analyze the structure and floristic composition of these forests and their regeneration mechanisms (seedling bank, seed rain, and soil seed bank). The study site is a 4 ha fragment located in the municipality of Cachoeira do Sul-RS, low River Jacuí (30o04’36”S; 52o53’09”W). The vegetation was inventoried using strips about 10 m wide that were laid out perpendicular to the river and separated by 50 meters. The strips were divided into 10 x 10 m plots. Within each plot I identified and measured circumference and height of all individuals with girth breast height- GBH15 cm (Classe I). Lianas were also measured, but were identified only according to life form. Species density data were entered into a matrix (70 x 42) that was used in multivariate analysis. The presence of vegetation clustering at the interior of the fragment was tested with TWINSPAN (Two-way indicator species analysis), which identified the existence of three forest subformations (S-F1, S-F2 e S-F3). S-F1 was characterized by a higher influence of flooding and the ground water table was closer to the surface; S-F2 occurred in the central part of the fragment, but presented a strong influence from flooding and inundation; and S-F3, also in the central part of the fragment, experienced the greatest influence from the ground water table. The indicator species of the three forest formations were Sebastiania commersoniana and Eugenia uniflora (S-F1); Gymnanthes concolor, Cupania vernalis and Seguieria aculeata (S-F2); and Casearia sylvestris and Allophylus edulis (S-F3). To evaluate if the regeneration mechanisms also exhibited differences according to the forest formations, and to identify the most effective regeneration process in the ecosystem, I used a different methodology for each mechanism. The seedling bank was sampled in sub-plots 1 x 4 m, placed within 27 plots of Class I, and individuals with height ≥ 20 cm and CBH < 15 cm were identified and measured. Statistical analysis showed that S-F1 was significantly different from S-F2 e S-F3. Using data from two years of study, I calculated rates of mortality, recruitment and change for species in S-F1 and in the middle of the forest fragment (S-F2 and S-F3). I observed that the seedling bank of S-F1 showed lower species richness and change over time than the interior of the fragment. Seed rain was monitored over two years in 30 collectors (1 x 1 m), and I observed no significant difference in the deposition of seeds among the forest formations. However, during the first year of study seed production was higher, principally among tree species. To characterize the soil seed bank, I sampled the first 5 cm of the soil surface (litter and soil). There was no significant difference between the forest formations. Most seeds were herbaceous species, followed by trees, shrubs and lianas. This study revealed that the highest similarity occurred between the vegetation and the seedling bank, followed by vegetation and seed rain, and the least similarity occurred with the seed bank. Therefore, the seedling bank is primarily responsible for the conservation of the fragment, because it depends on seed rain that varies over time. The soil seed bank is important in the first stages after disturbance of the environment, but depends on the seed rain to be more represented qualitatively. For conservation, I suggest twenty two species and also I highly indicate for recovery riparian areas, that species occur under the girth breast height (GBH15 cm) class, and at least under two regeneration mechanism.