Ecosystem Resilience and Land-Use Governance: Remote Sensing Insights for Environmental Management in a Tropical Dry Forest.

Seasonally Dry Tropical Forests (SDTFs), such as the Brazilian Caatinga, are highly vulnerable to climate change and chronic anthropogenic disturbances. Despite extensive land-use conversion, understanding how vegetation productivity and resilience vary across different land governance regimes remains a critical gap. Here, we integrated two decades (2004-2024) of remote sensing data to evaluate Net Primary Productivity (NPP) and Water Use Efficiency (WUE). We applied interannual climate anomaly analyses (Z-scores) and adapted a trend-based classification framework to assess ecosystem trajectories. These biophysical metrics were cross-referenced with the spatial boundaries of Conservation Units, Indigenous Lands, Quilombola Territories, and private properties, including a distance-decay edge effect analysis. While long-term temporal trends for NPP and WUE remained relatively stable over the two decades, standardized anomaly analyses revealed a strong, statistically significant inverse coupling between productivity and water-use efficiency (r = - 0.741, p < 0.001), primarily mediated by surface moisture retention (LSWI) rather than immediate precipitation alone. Spatial analyses demonstrated that Conservation Units act as the main ecological refugia, maintaining peak restoration within a 0-2 km internal buffer. Notably, Indigenous Lands and Quilombola Territories exhibited remarkable edge-to-core resilience, maintaining homogeneous conservation levels that effectively buffer external degradation. To safeguard the Caatinga's structural integrity against desertification, public policies must move beyond reactive drought subsidies towards proactive landscape management that legally empowers traditional communities and incentivizes conservation on private lands.