Unraveling the Dynamics of Cropland Abandonment in Guangdong Province, China: A Social‐Ecological Systems Perspective

Land abandonment, a significant contributor to cropland loss, has garnered growing global attention. However, the identification of abandoned land remains challenging due to limitations in high‐resolution data and the absence of universally applicable methodologies. Moreover, current research mainly focuses on mapping abandonment patterns, yet it lacks a comprehensive theoretical framework for analyzing the underlying causes, processes, and impacts. To address these gaps, this study introduces a place‐based methodology that integrates remote sensing technology with machine learning techniques to identify abandoned land, while also employing the social‐ecological systems (SES) framework to analyze the mechanisms and outcomes of cropland abandonment. Guangdong Province—a region in southern China characterized by complex topography and diverse socioeconomic conditions—was selected as the case study. The findings revealed that: (1) Between 2018 and 2022, abandoned croplands were predominantly concentrated in the underdeveloped hilly regions of Guangdong Province, accounting for about 12% of the total cropland annually. (2) The proportion of the primary industry and agricultural employment, abundance of cultivated land, population outflow, and effective irrigated area were positively correlated with cropland abandonment, whereas population density and road density were negatively correlated. (3) The most significant impacts of land abandonment on ecosystem services, characterized by a decline in food production and nutrient cycling, were observed in the coastal and estuarine areas, particularly in the Pearl River Delta. These findings highlight the need for targeted policy interventions to mitigate the impacts of cropland abandonment and promote sustainable land use strategies that balance agricultural productivity with ecological conservation.