A method of choosing locations based on genotype × environment interaction for regional trials of rice

An ideal test location with particular environmental conditions should generate a relatively large genotype × environment interaction (GE) value in regional trials. A method of selecting test locations based on GE was proposed. Using the statistic D j , the ranking of ability to discriminate the best genotypes for locations could be obtained. The locations among the group with the smallest D j value were deleted from the whole set step by step, and the number of locations was reduced to a rational level at which most (over 85%) of the GE sum of squares of trials were maintained. A comparative evaluation based on data collected from Zhejiang Province (China) rice regional trials showed that the new method was more feasible than two other methods without loss of accuracy and reliability.