The impacts of climate change on rice yield: a comparison of four model performances

Increasing concentrations of carbon dioxide (CO 2) and other greenhouse gases are expected to modify the climate of the earth in the next 50–100 years. Mechanisms of plant response to these changes need to be incorporated in models that predict crop yield estimates to obtain an understanding of the potential consequences of such changes. This is particularly important in Asia where demographic forecasts indicate that rice supplies worldwide will need to increase by 1.6% annually to the year 2000 to match population growth estimates. The objectives of this paper are (1) to review the major hypotheses and/or experimental results regarding rice sensitivity to climate change and (2) to evaluate the suitability of existing rice models for assessing the impact of global climate change on rice production. A review of four physiologically-based rice models (RICEMOD, CERES-Rice, MACROS, RICESYS) illustrates their potential to predict rice responses to elevated CO 2 and increased temperature. RICEMOD does not respond to increases in CO 2 nor to large increases in temperature. Both MACROS and CERES (wetland rice) responses to temperature and CO 2 agree with recent experimental data. RICESYS is an ecosystem model which predicts herbivory and inter-species competition between rice and weeds but does not respond to CO 2. Its response to increasing temperature also agrees with experimental data.