Spatiotemporal evolution and its influencing factors of steep-slope croplands in Southwest China

Steep-slope croplands can be defined as agricultural lands on topographic gradients steeper than 25°. Over half of steep-slope croplands are densely concentrated in the complex mountainous and hilly terrains of Southwest China. They are highly vulnerable marginal lands for national food security and the ecological environment. Spatiotemporal dynamics and trajectories are often required for decision-making on regional land. This study aims to clarify the precise spatiotemporal evolution of these steep-slope croplands. Their underlying influencing factors were also identified after land construction. The density indicator of steep-slope cropland was integrated from high-resolution 30-meter digital elevation models with annual land cover datasets over a 25-year period from 2000 to 2025. A standard grid unit was selected to balance macroscopic regional trends with micro-level spatial heterogeneity. A combination of advanced spatial analyses was employed, including global spatial autocorrelation, cold and hot spot analysis, and land-use transition matrices. A systematic investigation was conducted to explore spatiotemporal patterns and the complex, bidirectional inward and outward conversion of steep-slope croplands in the southwestern region. Furthermore, Random Forest (RF) machine learning was applied at the county administrative scale. The primary driving factors were determined for the static spatial distribution and the dynamic temporal shifts of the croplands among natural and anthropogenic variables. The results reveal that: (1) Steep-slope croplands in Southwest China were widely distributed in diverse geomorphological landscapes at elevations predominantly below 3000 m. There was an average density of 2.71% in steep-slope cropland, indicating a highly significant degree of spatial agglomeration, with the hotspots along major mountain ranges. There was a gradual decline from 2.84% to 2.52% in the overall density of steep-slope cropland in the entire study area during the 2000–2025 period. This macroscopic reduction was characterized by intense spatial heterogeneity, indicating a continuous decline pattern in the northern territories. While there was a highly fragmented, staggered, and interwoven pattern of simultaneous localized expansion and contraction over the central and southern zones. (2) Land-use transition indicated that the outward conversion of steep-slope croplands was directed primarily toward ecological restoration, with the proportions transferred into forests, shrublands, and grasslands accounting for 53.84%, 42.42%, and 3.56%, respectively. In contrast, the reverse inward conversions from the ecological lands into steep-slope agricultural use were sourced from forests (69.94%), followed by shrublands (15.46%) and grasslands (14.40%). The net directional shift conclusively demonstrated that the overarching land-use conversions involved steep-slope croplands during the quarter-century period. Ecological degradation was also reduced in the overall habitat quality of the region. (3) RF predictive modeling demonstrated that topographic constraints and land-use factors consistently ranked at the forefront of variable importance, thus maintaining a stable and dominant influence on steep-slope cropland density over time. Conversely, the explanatory power exerted by standard socio-economic variables and ecological quality indicators was relatively weak and exhibited frequent temporal volatility. Ultimately, qualitative assessments confirm that macro-policies function as the most crucial underlying driving forces behind the spatiotemporal transformations. The findings can provide a fundamental, data-driven reference to optimize territorial spatial patterns for precise, differentiated steep-slope cropland in Southwest China.