Quantitative understanding of the irrigated rice ecosystem and yield potential
1994
Kropff, M.J. | Casman, K.G. | van Laar, H.H. (International Rice Research Inst., Los Banos, Laguna (Philippines))
A systems simulation approach is used to evaluate the physiological basis and crop management requirements for achieving rice yields higher than present yield frontiers in tropical environments. A general crop simulation model is parameterized for rice and used to evaluate yield differences between varieties grown with different N management and in different environments to understand the physiological basis of the system. The model explain accurately variations in productivity on the basis of differences in radiation, temperature, leaf N content, and leaf area index (LAI) throughout the growing season. The model predicts that higher N inputs could raise dry season yields from the current 7 t/ha to more than 9 t/ha. The simulations suggest that the yield potential of current short-duration varieties is similar to the yield potential of the first semidwarf variety, IR8. For yields above 10 t/ha, a simplified yield formation model predicts that leaf senescence must be delayed, the grain-filling period must be longer, and there must be more grains per sq m. Theoretically, 15 t/ha could be obtained in a tropical environment with a growth duration of 134 d, a grain-filling period of about 40 d, high LAI, and high leaf N concentration. The simulation model confirms these figures. F1 hybrids show advantages in several physiological traits that are required for raising yield potential. Optimum crop management is crucial for realizing higher yield potentials. A rice crop would need at least 300 kg of N in above-ground biomass for a 15 t/ha yield with normal N concentration in the grains. Much of this N has to be taken up late in the growing season to meet the N requirements of a larger sink and a longer grain-filling period and to support the extended active leaf area duration due to delayed senescence
显示更多 [+] 显示较少 [-]