Abstract: 【Objective】 Environmental issue caused by unreasonable fertilization has been widely noticed in the paddy field. Yield increase potential and nutrient surface balance in the plant-soil system were analyzed, aiming to achieve scientific fertilizer principle and management strategy from the view of yield productivity and environmental benefit. 【Method】 A total of 735 site-year fertilization experiments were conducted on early rice, middle rice and late rice in Hunan province under the Project of Soil Testing and Fertilizer Recommendation during the period of 2005-2010. Three treatments were selected, control (i.e., CK), farmer’s practice (i.e., FP), and optimum fertilization (i.e., OPT). The yield increment and partial factor productivity (PFP) of nitrogen (N), phosphorus (P) and potassium (K) between the OPT and FP treatments for early, middle and late rice were analyzed. Three groups (i.e. high-yield, high-efficiency and high-yield & efficiency) were divided through screening the 75-100th percent of the yield and/or PFP_N for the OPT treatment. Evaluation of N saving and emission reduction in the paddy soil were combined the empirical model between N rates and N losses (i.e., nitrous oxide emission, N leaching and runoff, and ammonia volatilization). Apparent balance of P for early, middle and late rice were calculated to analyze the P residuals. In addition, the planting areas of early, middle and late rice in Hunan province were applied to assess the regional environmental benefit of N and P. 【Result】 Compared with the CK treatment, rice (average of early, middle and late rice) yield increment of FP and OPT treatments were 41.9% and 52.0%, respectively. And yield increase rate reached 7.0%-8.3% for the OPT treatment as compared with FP treatment for early, middle and late rice. Compared with the FP treatment, the optimum fertilization improved the PFP of N and P fertilizers by 16.8%-19.6% and 5.5%-37.3%, respectively, and slightly decreased (6.4%) in PFP_K since K rates were increased in the OPT treatment. In general, the high-yield group had high yield but also high N rate, and the high- efficiency group got low N rate but also low yield. The high-yield & efficiency group was proper to estimate the potential of yield and efficiency. Under the group of high-yield & efficiency, N fertilizers would be saved by 81 kg N·hm^-2 annually (32 kg N·hm^-2 for late rice), nitrogen loss decreased by 15 kg N·hm^-2, and nitrous oxide emission reduced by 20%. Total N saving potential would be 12.5×10⁴ t, mainly contributing by early rice and late rice, with the emission reduction potential of 2.35×10⁴ t. Optimum fertilization improved the total P uptake of rice from 59-66 kg·hm^-2 (FP treatment) to 63-71 kg·hm^-2 (OPT treatment). However, total P uptake of rice was not improved by increasing P rate. Clearly, when phosphorus rate meets the requirement of rice, surplus P would be residual in soil. There were 46%, 44% and 15% of the samples appeared to P residual for early, middle and late rice, respectively. The ratio of P residual for OPT treatment was obviously lower than that of the FP treatment. Phosphorus residual amounts of early, middle, and late rice for OPT treatment can be dropped to 1.36-5.30 kg P₂O₅·hm^-2, which were 33.7%-48.5% lower than the FP treatment. For the whole province, the total phosphorus residual of early, middle, and late rice would be decreased by 18.14×10³, 3.59×10³ and 5.30×10³ t. In the other hand, no P application was still exist in the farmer practice. Balance fertilization is very important for crop growth and soil nutrient maintaining. 【Conclusion】 Optimized fertilization is an important guarantee for food and environment security. Applying lower nitrogen fertilizer would improve nitrogen use efficiency, decrease nitrogen loss and produce less greenhouse gases. Controlling total phosphorus fertilization rate would reduce the pollution risk of soil phosphorus residual.

Key words: early, middle and late rice, optimized fertilization, yield, nitrogen loss, phosphorus residual