Abstract:In order to explore nitrogen (N) high-efficiency application technology of rice under rice-fish coculture and achieve the goal of N reduction and efficiency increase, a field plot experiment was conducted under rice-fish coculture with Longliangyou 1206 as planting material, in which four different N application treatments of rice were designed, 1) N0: no N application; 2) CK: local conventional N application, 180 kg/hm2, 50% topdressing at 15d after transplanting, 25% and 25% topdressing at 4th and 2nd leaves emerged from the top, respectively; 3) N1: reducing N application, 120 kg/hm2, 50% topdressing at 15d after transplanting, 25% and 25% topdressing at 4th and 2nd leaves emerged from the top, respectively; 4) N2: reducing and postponing N application, 120 kg/hm2, 50% and 50% topdressing at 4th and 2nd leaves emerged from the top, respectively. Rice growth characteristics, N use efficiency, yield and yield components were measured. Results show that N2 treatment significantly reduces the number of tillers at the critical leaf age for productive tillers and the dry matter accumulation of rice at the inverse 4th leaf stage, but the number of tillers at the critical leaf age for productive tillers reaches more than 88.54% of the effective panicles of the CK treatmenmt, which is enough for tillers. The application of panicle N fertilizer increases the dry matter accumulation, effective panicles and number of grains per panicle under N2 treatment, and significantly reduces the peak seedling of rice and improves the spike rate. The rice yields in the two years are increased by 6.39% and 8.57% respectively, compared with the CK treatment. At the same time, the N2 treatmenmt reduces the residue of hydrolysable N in soil at rice maturity stage, and N harvest index, N use efficiency for biomass production, N use efficiency for grain production, partial factor productivity of applied N and N agronomy efficiency are 5.63%, 12.99 kg/kg, 12.91 kg/kg, 28.45 kg/kg and 7.79 kg/kg higher than the CK treatment (P<0.05), increased by 9.33%, 15.56%, 29.14%, 59.57% and 120.77%, respectively. In conclusion, the postponing and decreasing N fertilizer in rice could significantly reduce the peak seedling of rice and improve the spike rate, the dry matter accumulation, effective panicles, number of grains per panicle, N agronomy efficiency and yield. Under the rice-fish coculture mode after tobacco in southwest China, the total N and hydrolyzed N in soil are more than 1.79 g/kg and 160.02 mg/kg respectively, and the total N are 120 kg/hm2, and the application ratio is 50% topdressing at 4th and 2nd leaves emerged from the top respectively, which could achieve the goal of reducing N fertilization by 33.3% but increasing rice yield by more than 6.39%.