Difference in the Contribution of Driving Factors to Nitrogen Loss With Surface Runoff Between the Hill and Plain Agricultural Watersheds

Identifying the factors and quantifying their contributions to nitrogen (N) loss associated with surface runoff is of great significance to the control of non-point source N pollution. However, the distinct geographical units, such as hills and plains, may lead to great differences in the contribution of these driving factors, which has been rarely investigated. This study developed an effective framework, which simulated the N loss with surface runoff in hills and plains by SWAT and NDP, and analyzed their spatial distribution variations by spatial autocorrelation analysis, and distinguished the contribution of their driving factors by multi-scenario simulation and partial redundancy analysis (pRDA). The framework was instantiated in a hill and a plain agricultural watershed, respectively, in the upper Taihu Lake Basin, China. We found the contribution of fertilization to N loss with surface runoff in the hills (10.4%) was greater than that in the plains (6.4%), which may be due to the N increment effect of paddy land. The contribution of rainfall to N loss with surface runoff in the plains (93.2%) was greater than that in the hills (74.3%), which may be due to the N increment effect of urban land. The developed framework could provide a viable way to study the environmental impacts of natural and anthropogenic drivers in different types of agricultural watersheds, thus offer scientific references for nutrient control measures. Geographical units within a watershed, such as hills and plains, exhibit distinct characteristics that give rise to variations in the processes of nitrogen (N) runoff loss and their associated driving factors. These variations pose challenges to the development of effective control measures for non-point source N pollution. Consequently, we have developed an effective framework capable of simulating N loss with surface runoff, analyzing variations in their spatial distribution, and discerning the external (e.g., rainfall) and internal (e.g., land management, fertilization) contributions of driving factors in the hills and plains, respectively. A notable discovery was that fertilization played a more substantial role in N loss in hills due to increased N levels resulting from paddy land, whereas rainfall emerged as the primary driver of N loss in plains due to elevated N levels associated with urban land. This research contributes to our understanding of the complexities surrounding N runoff loss and provides valuable insights for the development of targeted control measures within diverse geographical units. Contribution of different drivers to N runoff loss was compared between hills and plains Fertilization contributed more to N loss in hills due to N increment of paddy land Rainfall contributed more to N loss in plains due to N increment of urban land