This study aims to reveal the spatial distribution characteristics and loss patterns of soil nitrogen (N) and phosphorus (P) in the Danjiangkou Reservoir area, identify high-risk zones of N and P loss, and evaluate the mitigation effect of the Grain for Green Program on N and P loss.

Based on the 1:250000-scale soil geochemical survey comprising 5251 topsoil sampling points in the reservoir area, Kriging interpolation and ArcGIS spatial mapping were employed to characterize the spatial distribution patterns of soil N and P. A comprehensive risk assessment method for N and P loss was developed using the Soil and Water Assessment Tool (SWAT) model by integrating source intensity, transport risk, and receptor sensitivity, and the spatiotemporal characteristics of N and P loss loads were simulated. Three Grain for Green scenarios, with afforestation proportions of 10%, 30%, and 50%, were established to evaluate their mitigation effects on N and P loss.

The distribution of soil N and P showed a “high in the south and low in the north” pattern, with high-value areas located in the southern mountainous forest regions and low-value areas concentrated in the northern and central agricultural hilly and valley areas. High N/P loss-load zones were mainly located in the northern and central agricultural areas, whereas high-sensitivity zones showed the opposite spatial pattern. The loss loads exhibited significant seasonal variation, with losses during the flood season (May–October) accounting for 80.4% and 82.8% of the annual N and P losses, respectively. The Grain for Green Program significantly reduced N and P loss; when the afforestation proportion reached 50%, the loss loads of total nitrogen (TN) and total phosphorus (TP) decreased by 70.32% and 71.87%, respectively.

Rainfall is the primary driving force of soil N and P loss, whereas land use and agricultural fertilization are the key anthropogenic factors. The Grain for Green Program can reduce the risk of N and P loss by decreasing surface runoff and soil erosion, and these findings provide a scientific basis for nonpoint-source pollution control and ecological management in the reservoir area.