Predicting the Effects of Climatic Variation and Elevated CO₂ on Rice Yield in Japan

Abstract

For making a rice growth simulation model (SIMRIW; Horie, 1987) applicable to impact predictions of elevated atmospheric CO₂ concentrations and anticipated global warming on yields, the process to determine the direct effect of CO₂ on the growth and one related to high temperature-induced spikelet sterility were incorporated into SIMRIW. The direct effect of CO₂ was incorporated into the model by giving radiation conversion efficiency to crop biomass as a function of CO₂. On the basis of our CO₂ X temperature experiments with rice in temperature gradient tunnels (TGT) and also on information from the literature, the radiation conversion efficiency was estimated to increase by 24% under doubled CO₂ concentrations. The data on high temperature-induced spikelet sterility obtained in TGT were found to fit a logistic type function with respect to averaged daily maximum temperature over the flowering period.
Although the model gave a climatically potential yield, it could be converted into an actual farmers' yield by multiplying by a technological factor. Rice yield estimated by the model with the technological factor explained well the yearly yield variations in Hokkaido (northern Japan) and in Miyazaki (southern Japan). The model, thus synthesized, was applied to assessment of combined effects of CO₂ and global warming predicted by the GISS model (Hansen et al., 1984) under doubled CO₂ in the atmosphere. It was predicted that in northern Japan, doubled CO₂ and the global warming will increase the average yield of irrigated rice by about 20% and will stabilize the yearly variability, and that, in southern Japan, the global environment change will increase the average yield by about 10% though it will also increase instability in the yearly variability by about 2 times that of present due to the increased possibility of high temperature-induced spikelet sterility.