WHCNS-Veg Modelling of N2O, NH3 and NO3- Dynamics in a Vegetable Production System under Different Fertilization and Irrigation Regimes

Greenhouse vegetable production in China not only increases farmers' income, but also increases the risk of nitrogen losses due to excessive water and fertilizer input. Nitrogen losses, including the potent greenhouse gas nitrous oxide (N2O), are driven by water content, soil temperature and pH; regulated by available organic carbon and inorganic nitrogen (N); and affected by management. Therefore, a process-based model was applied to explain the complex interaction of the factors affecting N losses in the form of N2O, NH3 and NO3- from a greenhouse vegetable production system in a northeast suburb of Beijing, China. We designed four treatments: two equal N input treatments with one flooding (FP) and the other drip irrigation (FPD) and two equal water input treatments (drip irrigation) with one 100% chemical N input (FPD) and the other 50% N input (OPTD). The last one was CK treatment (flooding without chemical N). We calibrated the WHCNS-veg model using year-round measurements of soil temperature, N2O emission, NH(3)volatilization, NO3- distribution and yields for greenhouse cucumber-tomato cultivation under farmers' practice (flooding + 100% chemical N, FP). Then, we validated the model using the data sets under drip irrigation (70% of flooding amount + 100% chemical N, FPD), reduced chemical N by 50% (drip + 50% chemical N, OPTD) and CK treatment. The WHCNS-veg model was able to capture the above processes under different treatments. Annual N2O emissions were 5.47 and 3.76kg N ha(-1) for the cucumber and tomato seasons under FP, respectively. Compared to FP, drip irrigation (FPD) decreased N2O emissions by 19.0% and 45.5% in the two seasons, respectively. Compared to FPD, applying a lower rate of N (OPTD) further reduced N2O emissions by 13.7% and 40.5%, respectively. According to the model simulation, N2O emission was mainly controlled by nitrification/denitrification in the cucumber/tomato seasons, respectively. Compared to FP, drip irrigation (FPD) increased NH3 volatilization by 54.2% in the cucumber season, while in the tomato season, there were no significant differences inNH(3) volatilization under the three fertilizer treatments. The nitrate leaching levels were 48.5 and 81.0 kg N ha(-1) for the two seasons under FP treatment. Drip irrigation (FPD) decreased NO3- leaching by 20.6% in the cucumber season. Drip irrigation (FPD) and/or reducing chemical N (OPTD) did not compromise vegetable yields. In all, WHCNS-veg performed well in simulating N2O, NH3 and NO3- dynamics from the greenhouse vegetable field, which means that the model can be used to manage water and nitrogen precisely in greenhouse vegetable production systems by scenario analysis, and drip irrigation and/or lower N input can be applied in this area to secure yield and reduce N losses.