Restoration of seasonal semideciduous forests in Brazil: influence of age and restoration design on forest structure

With the high rates of deforestation in tropical regions, the restoration of degraded lands has become an important way for maintaining the diversity of plant communities and for creating wildlife habitats. Evaluating the success of restored areas is essential for improving restoration designs and for successfully restoring such complex ecosystems. In this study, the development of restoration forests with respect to age (5, 9 and 10 years old) and the restoration models used (proportion of pioneer trees) was assessed along the margins of Companhia Energetica do Estado de S?o Paulo (CESP) reservoirs, located in the region of Pontal do Pananapanema, in S?o Paulo state, southeastern Brazil. The overstory (trees >=4.8 cm in DBH) was assessed in nine 900 m^2 permanent plots, and all woody understory regenerating plants (>50 cm in height and <4.8 cm in DBH) were counted and identified in 54 1 m radius subplots. Canopy and grass cover were assessed in the wet and dry seasons. All of the parameters were recorded again 1 year later to evaluate the development of the forests. In general, neither the restoration design nor age appeared to influence forest structure and dynamics, at least at the developmental stage studied here. The floristic complexity and density of regenerating individuals were still fairly low compared to natural forests. The arrival of propagules from other forest remnants was insignificant, suggesting that inadequate seed dispersal and faunal colonization limited species enrichment within the restoration sites. Although the overall results suggested that these restoration forests had reached a structural complexity sufficient to give a start to secondary succession, some factors such as the dominance of pioneer trees in the overstory, the small number of colonizing individuals arriving from external seed sources, the persistence of weedy grasses in the understory and the degree of isolation of the restoration sites may endanger the sustainability of these forests in the long term. Although neither the age nor restoration model produced differences in the forest structure and dynamics, it is possible that such differences may require a longer time to develop. Monitoring restoration sites is essential for understanding a forestʹs trajectory and for guiding management and intervention practices.