Predicting resilience and stability of early second‐growth forests

Identifying deforested areas with high potential for natural forest recovery can be used as an aid for ecological restoration projects at large-scale. However, accurate predictions that infer the resilience (i.e. recovery rate after deforestation) and stability (i.e. the ability of the ecosystem to maintain its functions) of early second-growth forests are scarce at a regional scale. Here, we investigated the effect of climate, soil and topography on the resilience and stability of 165 early second-growth forests throughout the Brazilian Atlantic Forest. We also built prediction maps of potential resilience and stability to identify where reforestation could be optimized in the early stages of forest succession. We assessed the resilience and stability through an interannual plant primary productivity time series using a normalized difference vegetation index. Our analysis reveals that resilience was mainly associated with isothermality (i.e. diurnal temperature oscillation relative to the annual temperature oscillation) and precipitation of the warmest quarter. In turn, stability was mainly associated with the probability of bedrock occurrence, annual precipitation and precipitation seasonality. The prediction maps show a spatial pattern in which potential resilience and stability increase from north to south of the Atlantic Forest. Forest restoration can be optimized in regions with high potential resilience and stability, such as an isolated area on the north coast in the Bahia state and the southern region. However, restoration may require active practices and management in regions with low potential for both ecosystem properties, such as the north inland in the Bahia and Minas Gerais states. This ecosystemic approach can help achieve Atlantic Forest restoration commitments.