Study On Modeling The Effects Of Nitrogen Fertilizer Application On Rice Growth

Rice (Oryza sativa L.) is one of the most important grain crops in China. In order to increase rice yield, large amounts of nitrogen fertilizer were input into paddy field. Poor N fertilizer-use efficiency and environmental pollution become prominent problems in China’s rice production. Application nitrogen fertilizer reasonable is important to improve rice yield, ensure national food security and promote sustainable development of agriculture in the future. In this study, we analyzed the effects of nitrogen rates and ratios on the rice growth, through different nitrogen fertilizer treatment. Then, we compared three nitrogen balance model of rice. The results will contribute to improve nitrogen fertilizer management of rice production and national food security.Field experiments with different cultivars and nitrogen fertilizer treatments were conducted in2011and2012, which were used to analyze the effect of nitrogen fertilizer treatments on growth of rice. The experiment results showed that the dynamics of population tiller number, leaf area index, dry matter accumulation and partitioning, yield and yield components were affected by nitrogen fertilizer management. Increasing the amount of nitrogen fertilizer can improve panicle number, leaf area, dry matter and yield of rice, but will reduce1000-grain weight and N fertilizer-use efficiency. Increasing the ratio of the topdressing nitrogen can improve panicle number and N fertilizer-use efficiency, but will reduce the spikelets per panicle and1000-grain weight. For two nitrogen levels, the highest of dry matter and yield were in R3(6:4) treatment. Higher or lower ratios of base fertilizer are not conducive to obtain a higher dry matter and yield. The results showed that reasonable nitrogen fertilizer management can help to improve rice yield and reduce nitrogen loss and environment pollution. Based on the physiological development time (PDT), the rice critical nitrogen concentration models of organs and shoot were developed. The results showed that, the critical nitrogen concentration models of leaf, stem and shoot are fitting power function, while the critical nitrogen concentration of spike model is fitting linear function.In this paper, CERES-Rice, ORYZA2000and RiceGrow models were chosen to compare the differences in process division, parameters setting, model algorithm and estimating performance. The results showed that:1) in the process division, both CERES-Rice and RiceGrow described the nitrogen balance in soil and crop, while ORYZA2000considers on the nitrogen balance in crop more than in soil.2) the algorithms of CERES-Rice and RiceGrow in mineralization, immobilization, denitrification and leaching are similar. While the algorithms in urea hydrolysis, nitrification and ammonia volatilization are different. The algorithms of nitrogen balance in crop are similar in the three models. The difference is just in the critical nitrogen and minimum concentration of organs.3) the parameters in RiceGrow and ORYZA2000are more than CERES-Rice, but most of these parameters are constant values. According to calibrate the three models by different nitrogen treatments, we find that all of the models can simulate total dry matte, total N uptake by shoot and yield well in maturity. In the whole rice growth stage, RiceGrow simulates the panicle weight and total N uptake by shoot better than other models, ORYZA2000simulates panicle weight and shoot biomass better than other models The values of NRMSE for the nitrogen accumulation in different organs and shoot are greater than15%, which indicates more improvement should be done in the future.