SIMULATION TO SUPPORT EVALUATION OF THE PRODUCTION POTENTIAL OF RICE VARIETIES IN TROPICAL CLIMATES

ABSTRACT We investigated how simulation can support evaluation of varietal performance in multilocation trials of the International Network for Genetic Evaluation of Rice (INGER). Potential yields were simulated as a function of solar radiation and temperature for three rice genotypes of varying crop duration in five environments and compared with observed yields. Crop simulation can support INGER by increasing the efficiency of its germplasm evaluation program in multilocation trials. Reference levels of potential production of rice maturity groups can be established for different weather patterns. This permits a fast and cheap first approximation of the production of new varieties at new locations. Yields from 59 field trials were compared with such levels; half of them yielded below 65% of the simulated potential production. The gap reflects the magnitude of possible improvements.