Integrating Historical Crop Rotation Changes Into Soil Organic Matter Mapping in the Cropland of Southeastern China

The prediction of cropland soil organic matter (SOM) is crucial for understanding ecosystem services such as food production and carbon sequestration. However, the influence of historical agricultural management practices on soil properties is often overlooked in current SOM digital mapping studies. Crop rotation, a critical agricultural practice, significantly influences the spatiotemporal dynamics of SOM. This study investigates the cumulative effects of dynamic crop rotation changes on SOM using digital soil mapping in a multi-cropping region in southeastern China, where 202 topsoil samples were collected. Annual crop rotations from 2019 to 2023 were mapped by integrating Sentinel-2 data with expert knowledge, enabling the extraction of historical crop rotation changes. The added value of these legacy changes in SOM digital mapping was assessed by combining traditional environmental covariates. Results showed that integrating historical crop rotation changes significantly enhanced SOM prediction accuracy, with the coefficient of determination (R²) increased by 14.86% and the root mean square error reduced by 26.43% compared to models using traditional covariates and annual crop rotations alone. Notably, five-year changes in crop rotation emerged as the dominant factor in SOM prediction. The spatial distribution of SOM exhibited distinct heterogeneity, with high-value regions primarily located in regions under relatively stable crop rotation change frequency. In conclusion, this study underscores the importance and effectiveness of incorporating historical crop rotation changes into SOM mapping. The findings highlight the impact of historical agricultural practices on SOM distribution and provide a scientific foundation for enhancing agricultural strategies that promote ecosystem services.