Abstract: The objective of this study was to expose the change law of the three parameters of the precise and quantitative N application based on the Stanford equation in different growth types of rice. Three levels of N fertilizer application rates of 0, 225, and 300 kg ha⁻¹ N were designed in a field experiment with early (125–135 d), medium (136–145 d), late-maturing (146–155 d) medium japonica and early (156–165 d), medium-maturing (166–175 d) late japonica rice in 2007 and 2008 on the farm of Yangzhou university, Jiangsu province, China. The results were as follows: (1) Under zero N application, soil basal N application (N accumulation for basal yield) increased with the extension of the rice growth duration, in which the parameter of calculating N accumulation for basal yield was the N requirement for 100 kg grains. There were only small differences in N requirement for 100 kg grains between medium, late-maturing medium japonica and early-maturing late japonica rice. For an average yield of 6 t ha⁻¹ (4.87–6.58), the N requirement for 100 kg basal yield was 1.58 kg (1.50–1.63). (2) Under N application, the N requirements for 100 kg grains were rather stable in medium，late-maturing medium Japonica and early-maturing late Japonica rice. With the average yield of 7.5 t ha⁻¹, N requirement for 100 kg grains was 1.86 kg (1.85–1.87) under 225 kg ha⁻¹ N level and 2.01 kg (1.98–2.05) under 300 kg ha⁻¹ N level, respectively. With the average yield of 9 t ha⁻¹, N requirement for 100 kg grains was 1.94 kg (1.91–2.04) under 225 kg ha⁻¹ N level and 2.08 kg (2.01–2.19) under 300 kg ha⁻¹ N level, respectively. With the average yield of 10.5 t ha⁻¹, N requirement for 100 kg grains was 1.95 kg (1.93–2.01)under 225 kg ha⁻¹ N level and 2.09 kg (2.08–2.10) under 300 kg ha⁻¹ N level, respectively. These results suggested that N requirement for 100 kg grains increased with the rising of N application rates. Correlation analysis showed that there existed highly significantly positive conic correlation between yield and N requirement for 100 kg grains, and the correlation coefficients for early-maturing medium japonica and medium-maturing late Japonica were less than those for the other three types. (3)The N use efficiency decreased with the increase of N application rate, and increased with the extension of the growth duration except in the medium-maturing late japonica. N use efficiency was 31.32%, 37.64%, 38.5%, 41.08%, and 38.11% under 225 kg ha⁻¹ N level , and 28.74%, 36.13%, 37.16%, 40.15%, and 39.42% under 300 kg ha⁻¹ N level, for the five types of rice with extended growth duration respectively. (4) The parameters of N application displayed larger differences in early-maturing medium japonica and medium-maturing late japonica rice which were less suitable for local cultivation, compared with the other three types, while these parameters values of the prevailing varieties (headed varieties) were higher and more stable than those of common varieties. These results demonstrated that the changes of the parameters of N application were much larger, it is feasible for the N quantity required for target yield to be calculated accurately by using Stanford equation under the condition of reasonable variety selection and planting location.

Key words: Rice, Different growth type, N application rate, Precise and quantitative N application, N requirement for 100 kg grains