Forest restoration is increasingly promoted to mitigate climate change, conserve biodiversity, and secure food and water sovereignty. Yet many restored forests do not persist in the long term, and the role of land tenure regimes in shaping these outcomes remains poorly understood. We examine restoration reversals (restored forests later deforested) and long-term restoration gains (restored forests that remained intact) across 1.9 million territories in Brazil’s Atlantic Forest from 1985 to 2022. We compare Indigenous lands, Afro-descendant (Quilombola) territories, agrarian-reform settlements, protected areas, and private properties, introducing a statistical matching technique–agglomerative matching–to account for systematic differences between land tenure regimes. We find that Indigenous lands and agrarian-reform settlements exhibit significantly more long-term restoration gains than private properties. Concurrently, and by a smaller margin and on a smaller land area, Indigenous lands and agrarian-reform settlements exhibit higher reversals. These results highlight the relatively low restoration longevity of private properties and emphasize the importance of socio-political conditions in enabling long-term restoration gains.
Land Tenure Regimes Influenced Long-Term Restoration Gains and Reversals Across Brazil’s Atlantic Forest
Year: 2025





























































































































































































































































































































































































































































































































































































































































































































































































































































































































































































































