Food security-oriented optimization of rural settlements: coupling GAEZ and PLUS to address cropland productivity loss in Shaanxi, China

Rapid urbanization and rural transformation continue to reshape production and living spaces in rural China. Cropland occupation and reclamation resulting from rural settlement evolution may alter the spatial configuration and production potential of cropland, thereby affecting regional grain production capacity. In practice, China's cropland requisition–compensation balance has primarily emphasized quantitative balance, yet it has not adequately addressed the productivity losses resulting from the occupation of high-quality cropland and compensation with lower-quality land. This study develops an integrated framework to quantify this dual loss in both quantity and quality, and to optimize rural settlement layouts with a firm food security orientation in Shaanxi Province, China. We couple the Global Agro-Ecological Zones (GAEZ) model with the Patch-generating Land Use Simulation (PLUS) model. The GAEZ model was first used to estimate cropland grain production potential (GPP), which was then innovatively incorporated as a core spatial driver into the PLUS model to simulate and optimize the distribution of rural settlements. Results show that 93.43% of new rural settlements were converted from cropland. The occupied cropland (OC) had a higher average GPP than newly reclaimed cropland (NRC), leading to a net reduction of 49.62 × 10 4 tonnes in total GPP. The GPP-informed optimization demonstrated a significant mitigation effect, reducing the net GPP loss by 7.19 × 10 4 tonnes and substantially narrowing the quality gap between OC and NRC. The framework provides a replicable tool for shifting the core of cropland protection policy from a simplistic area balance towards a more secure productivity balance, offering critical insights for food security-oriented spatial planning.