Simulating Soil Carbon Under Variable Nitrogen Application, Planting, and Residue Management

Effective residue management is crucial for maintaining soil organic carbon (SOC) in upland rice systems, particularly under diverse fertilization and planting management practices. This study investigates the impacts of residue management in upland rice fields using the CQESTR model through simulation of SOC dynamics over a 20-year period. The first 10 years served as a spin-up period for carbon pool stabilization in the model, followed by simulations under varying nitrogen (N) application rates and planting date management strategies. Experiments for various N application rates and planting times were conducted during 2018–2019 and 2019–2020. In 2019, 30% and in 2020, 100% of the residue was returned, and these data were used for evaluating model performance. Subsequently, we modeled predictions for residue retention levels of 100%, 70%, 50%, and 30% to assess their effects on SOC. The results indicated a good agreement between the simulated and observed data for model performance evaluation with an MSD value of 9.13. Lack of correlation (0.44) accounted for 5% of MSD, indicating a good agreement between the simulated and observed SOC values. The highest change in SOC was observed at 100% residue return under moderately delayed planting, potentially due to higher crop productivity and residue retention, and moderate climatic conditions. Reduced residue retention gradually declined the SOC stocks, especially under low N input. Delays in planting exacerbated negative impacts, possibly due to low crop productivity and reduced residue return. Despite the limited number of years of data and inconsistent management practices, the overall trends highlight the importance of residue retention under different N fertilization and planting management strategies. This research serves as a preliminary study for sustainable management practices to enhance long-term soil carbon sequestration in upland rice systems in southern Thailand. Long-term evaluations are necessary using the observed data and the CQESTR model application for applicable recommendations.