Agricultural non-point source pollution management in a reservoir watershed based on ecological network analysis of soil nitrogen cycling.

The Miyun Reservoir plays a pivotal role in providing drinking water for the city of Beijing. In this research, ecological network analysis and scenario analysis were integrated to explore soil nitrogen cycling of chestnut and Chinese pine forests in the upper basin of the Miyun Reservoir, as well as to seek favorable fertilization modes to reduce agricultural non-point source pollution. Ecological network analysis results showed that (1) the turnover time was 0.04 to 0.37 year in the NH compartment and were 15.78 to 138.36 years in the organic N compartment; (2) the Finn cycling index and the ratio of indirect to direct flow were 0.73 and 11.92 for the chestnut forest model, respectively. Those of the Chinese pine forest model were 0.88 and 29.23, respectively; and (3) in the chestnut forest model, NO accounted for 96% of the total soil nitrogen loss, followed by plant N (2%), NH (1%), and organic N (1%). In the Chinese pine forest, NH accounted for 56% of the total soil nitrogen loss, followed by organic N (34%) and NO (10%). Fertilization mode was identified as the main factor affecting soil N export. To minimize NH and NO outputs while maintaining the current plant yield (i.e., 7.85e0 kg N/year), a fertilization mode of 162.50 kg N/year offered by manure should be adopted. Whereas, to achieve a maximum plant yield (i.e., 3.35e1 kg N/year) while reducing NH and NO outputs, a fertilization mode of 325.00 kg N/year offered by manure should be utilized. This research is of wide suitability to support agricultural non-point source pollution management at the watershed scale.