Multiple drivers influence tree species diversity and above-ground carbon stock in second-growth Atlantic forests: Implications for passive restoration.

Second-growth forests (SGF) are critical components for limiting biodiversity loss and climate change mitigation. However, these forests were established after anthropic disturbances such as land use for planting, and in highly human-modified landscapes. These interventions can decrease the ability of biological communities to recover naturally, and it is necessary to understand how multiple drivers, from local scale to landscape scale influence the diversity and carbon stock of these forests in natural regeneration. For this, we used data from 37 SGF growing on areas previously used for eucalyptus plantations in the Brazilian Atlantic Forest, after the last cut cycle. For each SGF, the forest tree species diversity was calculated based on the Hills number, and we also calculated the above-ground carbon stock. Then, we evaluated the influence of multiple environmental factors on these indexes: soil properties, past-management intensity, patch configuration, and landscape composition. Little influence of soil properties was found, only soil fertility negatively influenced above-ground carbon stock. However, past-management intensity negatively influenced tree species diversity and carbon stock. The isolation of other forests and tree species propagules source distance (>500 ha) also negatively influenced the diversity of species. This is probably due to the favoring of tree pioneer species in highly human-modified landscapes because they are more tolerant of environmental changes, less dependent on animal dispersal, and have low carbon stock capacity. Thus, areas with higher past-management intensity and more isolated areas are less effective for passive restoration and may require intervention to recover tree diversity and carbon stock in the Atlantic Forest. The approach, which had not yet been applied in the Atlantic Forest, brought similar results to that found in other forests, and serves as a theoretical basis for choosing priority areas for passive restoration in the biome.