Factors affecting phase change in coupling coordination between population, crop yield, and soil erosion in China’s 281 cities

Abstract

Soil erosion in cropland areas is mainly influenced by agricultural activities and natural conditions. Previous studies have largely focused on the biophysical processes or economic drivers of soil erosion. There have been few attempts to balance the impacts of population, agricultural production, and soil erosion to address the global socioecological predicament facing cropland. We combined the coupling coordination degree model (CCDM) with the Shapley additive explanations (SHAP) method to evaluate the coupling coordination level between population demand, agricultural production, and soil erosion as well as the influence of socioeconomic factors in 281 Chinese cities for the period from 1995 to 2010. Coupling between population, crop yield, and soil erosion was generally moderate across China during 1995–2010. Cities with a GDP in the range of 4.42–241.54 billion could fall into different coupling coordination phases that were identified by K-means clustering. The SHAP results showed that GDP and population density were the most important factors influencing the coordination level, while industrial structure was the key determinant that distinguished the different phases in cities with a similar economic status. Building on research on the evolutionary aspects of system coupling coordination, our study reveals for the first time the probable causes of changes in system coupling coordination via machine learning algorithms, providing a reference for future investigations. Our findings also provide a basis for developing policy recommendations to balance social demands, agriculture, and environmental protection.