Beyond the factor-driven paradigm: Spatiotemporal dynamics and system coupling of agricultural transition in Henan Province, China

Abstract

The core of agricultural transition lies in changes to the various components of the agricultural systems and their interactions. Drawing on the social-ecological-technological systems (SETS) framework, this study developed an “element coupling–external manifestation” analytical framework and a corresponding evaluation index system for agricultural transition. Using county-level data from Henan Province spanning 1990 to 2020, this study further examined the spatiotemporal patterns of agricultural transition in Henan Province and identified its key influencing factors through the XGBoost-SHAP and GTWR models. The results show that the average agricultural transition index rose steadily from 0.280 in 1990 to 0.565 in 2020, while regional disparities narrowed. During the study period, the spatial pattern of agricultural transition has shifted from a “higher in the north and lower in the south” model to one featured by multi-point diffusion and cluster-based development, with hypervariable density emerging as the primary source of overall inequality. Government macro-control capability and economic development level are the two factors most strongly associated with agricultural transition, collectively accounting for 63.3% of the total contribution. Both exhibited a U-shaped relationship, and all factors demonstrated significant nonlinearity and spatiotemporal non-stationarity, with government macro-control capability being particularly prominent. By establishing a SETS-based analytical paradigm, this study redefines agricultural transition as a coupling process within the SETS framework and provides theoretical and empirical insights into the diversity of its pathway as well as the nonlinearity and spatiotemporal heterogeneity of its influencing factors.