Social‐Ecological Drivers of Cropland Productivity Dynamics for Sustainable Intensification

Abstract Sustainably enhancing cropland productivity is critical for global food security, yet its drivers within complex social‐ecological systems remain poorly understood. Conventional analyses often miss the cross‐scale, non‐linear dynamics that shape productivity, leading to ineffective policies. Here, we deconstruct the drivers of cropland productivity change across China (2001–2020) using a multi‐scale framework integrating machine learning and causal inference. We find that productivity is co‐determined by biophysical properties, urbanization, and cropland use intensity with pronounced non‐linear interactions. Critically, we reveal a scale‐dependent shift in dominant drivers, from biophysical processes at macro‐scales to socioeconomic forces at micro‐scales. Our analysis further uncovers that urbanization reduces cropland productivity sustainability through both direct losses and indirect losses operating via weakened cropland conservation and lower land‐use intensity, consistent with rural labor drain and declining management engagement. However, local land‐ and economic‐urbanization can enhance sustainability in many counties, especially in major farming regions. This points to a misalignment between physical‐resource‐oriented policies and social system‐centered vulnerabilities. Consequently, our results demonstrate that productivity is an emergent property of the entire social‐ecological system, and its sustainability requires an adaptive governance approach capable of managing these systemic, multi‐scale dynamics.