LAUSR

Abstract Overuse of N fertilizer in intensive greenhouse vegetable production systems (GVPS) seriously affects soil and water environment. Quantifying soil N dynamics and plant growth is crucial for developing sustainable N management practices. A long-term greenhouse tomato experiment, including 13 tomato growth seasons (2004–2010) and three different nitrogen (N) management practices (conventional N management, CN; reduced N management, RN; and manure N management, MN), was conducted in this study. Agricultural system model (WHCNS_Veg) was adopted to simulate and analyze the effects of different long-term N fertilization regimens on N loss and tomato yield. The calibration and evaluation of the model showed fairly satisfactory simulation in soil water content, soil nitrate concentration, crop N uptake, and fresh marketable tomato yield; the values of the normalized root mean square error (NRMSE) were 15.4%, 42.9%, 21.2%, and 7.7%, respectively, and the values of index of agreement (IA) were 0.85, 0.62, 0.96, and 0.97, respectively. Different N management practices did not significantly affect the tomato yield. However, the long-term high rates of chemical N application in CN treatment significantly increased soil N accumulation, thereby increasing N loss. The simulated N losses caused by leaching, denitrification, and ammonia volatilization under the reduced N treatments (RN and MN) were decreased by 66.8%, 56.8%, and 83.2% compared to those of CN treatment, respectively. Among different N management practices, manure N management showed a relatively high N use efficiency and maintained a high tomato yield. And the optimal manure N application rate approximately 400 kg N ha−1 yr−1 was recommended to maximize farmers' income. The results not only provide decision supports for local farmers' N management, but also policy making of government for sustainable development of greenhouse tomato production in the region.