Deforestation Edge Effects on Soil Moisture Persistence in the Amazon Basin: Observational Evidence for Lateral Hydrological Degradation and Minimum Viable Restoration Scales

Ali Bin Shahid 

Deforestation in the Amazon basin degrades not only the cleared land but also the hydrological function of adjacent intact forest. Here we combine 10 years of SMAP Level 4 root-zone soil moisture (9 km, 2015 to 2024) with Hansen Global Forest Change data (30 m) to quantify how proximity to deforestation edges affects soil moisture persistence in intact forest across the Amazon basin. We classify 32,011 intact forest cells into six distance bins ranging from edge-adjacent (0 to 9 km) to deep interior (greater than 144 km) and compute post-wet-season dry-down rates and e-folding residence times for each bin. We find a continuous, monotonic gradient: forest within 9 km of a substantial clearing (1 km squared or larger) dries eight times faster than forest more than 144 km from any edge (dry-down rate negative 0.024 versus negative 0.003 cubic metres per cubic metre per month; one-way ANOVA F equals 25.3, p less than 0.000001). The e-folding moisture residence time increases from 13.6 months at the edge to 56.8 months in the deep interior (Spearman rho equals 1.000 across all bins). In a complementary analysis with seven clearing-size bins, we show that the size of the nearest clearing independently predicts neighbor dry-down: forest adjacent to clearings exceeding 1000 km squared dries 46 percent faster than forest adjacent to clearings of 1 to 3 km squared (p equals 0.0004). A critical threshold emerges at approximately 10 km squared, where the largest single step in residence time occurs (17 percent reduction); below this size, effects are moderate, while above it, degradation saturates near the severe levels characteristic of the arc of deforestation. These results demonstrate that deforestation degrades the water-holding capacity of intact forest over distances two orders of magnitude greater than the 100 to 300 m microclimate edge effects documented by the Biological Dynamics of Forest Fragments Project. The findings imply that a restored forest patch must be large enough for its core to lie more than 70 km from any degraded boundary, setting a minimum viable diameter of approximately 150 km for hydrological self-sufficiency. This constraint has direct implications for the design of large-scale restoration programs in the Amazon.