Dynamics and causes of cropland Non-Agriculturalization in typical regions of China: An explanation Based on interpretable Machine learning

Abstract

Cropland resources are crucial for food security and economic development. As a populous nation that considers cropland a valuable strategic resource, China has faced challenges of the phenomenon of the cropland non-agriculturalization (CLNA) in recent years. Studying the spatial and temporal patterns of CLNA in different regions and its driving factors is of great significance for formulating and improving cropland protection policies. Based on remotely sensed land use data, the spatial and temporal evolution characteristics of CLNA and its underlying influencing factors in three typical regions of China (Jilin, Henan, and the Guangdong-Hong Kong-Macao Greater Bay Area) from 2000 to 2020 were revealed by using Moran’s index, Lorenz’s curve, and XGBoost-SHAP model. The results of the study show that: (1). There was a certain amount of CLNA phenomenon in all three regions during the past 20 years, and the phenomenon in the Guangdong-Hong Kong-Macao Greater Bay Area (GHM) was the most serious, with the CLNA area exceeding 6 % of the total cropland in the peak period. The average Gini coefficients of Jilin (JL), Henan (HN) and the GHM were 0.39, 0.45 and 0.77, respectively, and the spatial imbalance of CLNA was the most serious in the GHM, which seriously threatened the food security of the region. (2). In the three study areas, socio-economic factors contributed 68 % (JL), 86 % (HN) and 84 % (GHM) to the CLNA, significantly higher than natural factors, and were the dominant factors influencing the CLNA. (3) Although socio-economic factors played a dominant role in non-agriculturalization, the role of natural factors such as elevation and slope should not be overlooked. These natural factors not only directly affect land usability and suitability but also interact with socio-economic factors, jointly shaping the trends and patterns of CLNA. In specific regions, natural constraints may hinder the conversion of cropland to non-agricultural uses, thus playing a protective role. Conversely, in some areas, favorable natural conditions may combine with strong economic development needs, accelerating the transformation of cropland to non-agricultural uses. This interactive effect of socio-economic factors and natural conditions revealed that the CLNA is a complex process influenced by multiple factors. Overall, this study not only provides an in-depth perspective for understanding the spatiotemporal evolution characteristics and complex driving mechanisms of CLNA but also offers a scientific basis for precisely formulating policies for cropland protection and promoting sustainable land use.