Natural forest regrowth under different land use intensities and landscape configurations in the Brazilian Atlantic Forest

Natural forest regrowth is considered the most cost-effective strategy to promote large-scale forest restoration, but regrowth trajectories and their consequences for nature and people can be highly variable. This uncertainty may obstruct devoting land, time and resources in promoting this restoration approach. Process-based models allow to simulate forest regrowth under different scenarios and may thus support decision making to manage areas in ways that maximize forest restoration based on natural forest regrowth. In the present study, we used the process-based model LandClim to assess the following questions:1) How does land use intensity affect forest growth, aboveground biomass (AGB) and tree species diversity in agricultural landscapes in the long term? 2) How do different land use configurations affect the structure and expansion of second-growth forests? We based our study on six 4 x 4 km agricultural landscapes, dominated by pastures and sugarcane, in the Brazilian Atlantic Forest region. We parameterized and validated LandClim for the study region by comparing model outputs with published estimates and information from field data collected across the six landscapes. We then simulated natural forest regrowth under different land use intensities in the six landscapes and analyzed how land use intensity and landscape configuration affect AGB, tree species diversity and the spatial dynamics of second growth forests. Our results showed negative effects of increased land use intensity on landscape biomass accumulation due to limited forest cover expansion. Landscapes dominated by sugarcane had lower forest regrowth potential than pasture-dominated landscapes due to lower likelihood of abandonment, and limited rates of seed availability and seedling growth. Land use intensity and landscape configuration were also important factors for tree diversity changes. However, temporal patterns differed, and while AGB reached a plateau after 100-150 years, tree diversity peaked between 30 and 70 years. Thus, the effects of land use intensity and landscape configuration on forest AGB and tree diversity take decades to be fully expressed, highlighting the importance of a long-term commitment in restoration projects.