Development of a simulation model for rice [Oryza sativa] panicle blast epidemics, 2: Its verification and sensitive analysis.
1990
Ishiguro K. | Hashimoto A.
Two components of the model of rice panicle blast epidemics previously composed was improved in this report. The gamma-density function was fitted to the potential spore production by an individual lesion during the infectious period. An equation, Y=A(1-e(-B(t-c)))+D, was fitted to the ripening curve of each spikelet, in which y is the weight of a spikelet, t is days after heading, A is the maximum weight of each hulled rice, B is the ripening rate of each hulled rice, C is the time lag from heading to flowering of each spikelets, and D is the weight of a hull, respectively. The model was executed to be verified, using the weather data from 1984-1988 in Koriyama of Fukushima prefecture. The results of verification showed that the general behavior of the model was reasonably similar to that of the real system. The effects of several parameters or functions in the model on its behavior were studied by means of a sensitivity analysis. The percentage of spikelets affected by blast disease represented the behavior of the model. The model was sensitive to the spore production by leaf lesions. The model was also sensitive to the incubation period of a spikelet, although not so sensitive to those of rachis or rachilla. The duration of the wet period of panicles played a critical role in the behavior of the model. The rate of lesion extension toward the proximal direction turned out to be important for the behavior of the model, especially when lesions appeared first on nodes of rachis, or on nodes from which secondary rachilla branches grew. The model was not sensitive to days necessary for a lesion extended from the primary branch to encircle a node of rachis. The model of spore dispersal from leaves to panicles affected the behavior of the model. The results mentioned above will provide useful information to improve the model.
显示更多 [+] 显示较少 [-]