Spatially heterogeneous controls of soil organic carbon in a karst mountainous area of southern China: Insights from interpretable machine learning

Understanding the spatial variability and environmental drivers of soil organic carbon (SOC) is critical for improving carbon management in fragile karst landscapes. This study collected 110 topsoil samples across county Yangshan, southern China, and applied an interpretable machine learning framework combining Random Forest (RF) and SHapley Additive exPlanations (SHAP) to explore the spatial heterogeneity and key environmental controls of SOC. The measured contents ranged from 3.33 to 44.20 g/kg, with a coefficient of variation of 43.5%, indicating moderate variability. RF-based spatial predictions revealed that higher SOC levels were mainly concentrated in the northern and southern subregions associated with clastic rocks, while lower SOC values clustered in central areas dominated by carbonate bedrocks. SHAP analysis indicated that soil physicochemical properties contributed over 53% to SOC, with total nitrogen and cation exchange capacity exerting the strongest influences, particularly in karst zones. Hydrological, vegetation, and terrain-related factors showed moderate importance, especially in high-elevation areas with natural vegetation and complex topography that promoted SOC accumulation. In contrast, climatic variables had relatively weak impacts, with their influences clustered in lowlands dominated by anthropogenic land uses. These findings revealed spatially heterogenous controls on SOC between karst and non-karst landscapes, emphasizing the dominant role of soil properties under shallow, erosion-prone conditions and highlighting the role of topography and vegetation in enhancing SOC stocks in mountainous areas. The integrated use of interpretable machine learning approaches improves the understanding of localized SOC dynamics and provides a valuable reference for precision carbon management and ecological restoration in other environmentally sensitive regions. • RF-SHAP framework effectively identified key environmental drivers on SOC. • Local SHAP values visualized spatial heterogeneity in karst and non-karst areas. • TN and CEC dominated contributions of SOC, especially under karst landscapes. • Hydrology, vegetation, and terrain influenced SOC in high-elevation clastic zones. • Climatic impacts clustered in lowlands associated with anthropogenic disturbance.