Analysis and comprehensive evaluation of cultivated land multifunction in grain-producing areas of Central China: based on social network analysis and zone-based machine learning approach

Cultivated land multifunction (CLMF) is essential for ensuring food security and rural revitalization. Central China plays a key role in national food security while facing increasing pressure from rapid urbanization. Using statistical data from 76 cities during 2010–2023, this study evaluated CLMF with a projection pursuit model based on the real-coded accelerating genetic algorithm (RAGA-PPM), analyzed its spatial association through the gravity model and social network analysis (SNA), and identified optimal spatial zoning and driving factors using the geographical optimal zones-based heterogeneity model (GOZH) from four dimensions: natural conditions, agricultural development, industrial development, and urbanization. The results show that: (1) CLMF exhibited a clear provincial gradient, following the pattern of ‘Hunan > Hubei > Anhui > Henan > Jiangxi’. (2) Significant spatial correlations existed among cities, with Hubei Province occupying the network core and Huanggang showing the highest degree centrality. (3) GOZH revealed that precipitation was the dominant driver of CLMF spatial heterogeneity in 2015, whereas the urbanization rate was the primary zoning factor in 2010, 2020, and 2023, indicating that natural and socio-economic factors jointly shaped CLMF at different development stages. Overall, urbanization has gradually replaced natural conditions as the dominant driver of CLMF spatial differentiation.