Mudança na Composição de Espécies Arbóreas em uma Floresta de Terra Firme Explorada Experimentalmente há 25 anos na Amazônia Central

Abstract

The present study investigated recovery of species composition in a Tropical terra firme Forest logged at the Experimental Station of Tropical Forestry 25 years ago. We analyzed 12 long-term permanent plots of 1 ha, submitted to different logging intensities. All trees with DBH ≥ 10 cm were annually measured between 1986-2012 yr. Between 2011 and 2013 we collected and identified all trees included in the monitoring at scientific level. To summarize the species composition we used Principal Coordinates Analysis (PcoA), regarding dissimilarities of quantitative data (abundance). To evaluate differences in species composition among treatments we used multivariate variance analysis (MANCOVA). 500 simulations were performed to evaluate if changes in floristic composition among years followed a directional trajectory or occurred randomly. There were 58 plant families and 709 species. Multivariate analysis of variance showed no difference in floristic composition between years when all treatments and control are considered together. This happens because the trajectories of change between years in controls plots are in the opposite direction of the trajectories compared to the areas explored during the first 11-17 years. However, there was significant difference between the repetition blocks and between treatments. This difference was more evident among the areas that experienced selective logging and control areas. The explored areas showed a tendency to return the composition to the initial year, before logging, higher than expected if plots followed random trajectories (p = 0.024). These changes were mainly caused by the increased abundance of pioneer species after logging, which tended to reduce after an estimated average of 13 years. Little effect of logging was detected on climax species, including commercial species, however the recruitment of these species was low. Control plots had a tendency to follow unidirectional trajectory changes along the monitorament. This suggests that some external factor such as long-term weather may be driving those changes. Those results corroborate that the selectively logged tropical forests may retain significant value for biodiversity conservation. We suggest that 25 years were not enough for recovering species composition, regarding that the first minimum cycle of 25 years guaranteed by law, are not enough to recover species composition.