Empirical Analysis of Deforestation Drivers and Their Relative Contributions in the Niger-Benue Confluence Zone

Understanding the relative contributions of proximate and underlying deforestation drivers is essential for designing targeted policy interventions in the ecologically sensitive Niger-Benue confluence zone, which spans portions of Kogi, Niger, and Kwara States. This study empirically quantifies the relative contributions of agricultural expansion, fuelwood extraction, charcoal production, and infrastructure development to forest loss over a ten-year period using a multi-variate statistical framework applied to remotely sensed land cover change data. Annual deforestation layers were derived from Landsat and Sentinel-2 composite imagery processed through Google Earth Engine, generating a pixel-level time series of forest cover change. A spatially explicit regression model incorporated explanatory variables including population density, road proximity, market access, household poverty indices, and agricultural commodity prices. An agent-based model was used to simulate deforestation trajectories under alternative policy scenarios. Results attribute 48% of forest loss to smallholder agricultural expansion, 27% to charcoal production, 14% to infrastructure corridors, and 11% to urban and peri-urban sprawl. Cassava price increases were significantly associated with accelerated forest conversion in accessible areas. Infrastructure expansion showed a multiplier effect on subsequent agricultural encroachment. Scenario simulations suggest that enforcement of a minimum 30-metre forest buffer along waterways would reduce projected ten-year deforestation by 19%. The study provides an evidence base for spatially targeted conservation investment and recommends differentiated interventions calibrated to the dominant driver in each ecological zone. Keywords: deforestation drivers, Niger-Benue confluence, spatial regression, remote sensing, policy scenarios