Spatio-Temporal Pattern of Cropland Abandonment in Hilly and Mountainous Areas: A Case Study of Jiangxi, China

Cropland abandonment is common and widely distributed in hilly and mountainous areas. Clarifying the current situation and development of cropland abandonment can provide reference for the rational and classified management of cropland abandonment in hilly and mountainous areas. Taking Jiangxi Province as the study area, and using the Google Earth Engine and Landsat data, the scale and years of abandoned cropland from 2002 to 2020 were calculated by using the random forest classifier and rules for identifying cropland abandonment. The spatio-temporal pattern of cropland abandonment at the county level was analyzed. The results indicated that the overall accuracy of land use classification was over 90%. The cropland abandonment rate ranged from 3% to 5.5% from 2002 to 2020, while the cropland abandonment rate was highest in 2013 and showed a downward trend after 2017. Among the years, the area of first-time abandoned cropland was the largest in 2005. The distribution of the cropland abandonment rate was low in the middle and north, but high in the surrounding area and the south. A notable positive spatial correlation was observed in the cropland abandonment rate, with a gradual intensification of spatial clustering. The LISA cluster map revealed a significant north-south disparity, exhibiting an incremental trend over time in the characteristics of the “High-High” cluster in the Southeastern Mountainous Area and the “Low-Low” cluster in the Poyang Lake Hilly Plain in Jiangxi. The results of this study can provide data for extracting spatial information and analyzing the driving factors of cropland abandonment in hilly and mountainous areas, and they can also provide a basis for the development of policies for the utilization and classification management of abandoned cropland.