小麥品種 育成 方法에 關한 硏究 Ⅱ. 品種環境의 交互作用과 選拔效果

Effect of selection on plant improvement Ⅱ. Genotype environment interaction and gain from selection

A study was made on 11 wheat lines to investigate the G-E interaction and their implication for prediction of the optimal number of replications, locations and years affecting upon the GAIN FROM SELECTION (Gs) for plant improvement on selection, explaining the interpretation of appropriate parameters related to genetic and environmental variances, variance and standard error for each variance compoment, and Gs. There appeared higher significant variations in 2nd order G-E interaction than 1st order G-E interaction, being desirable for plant selection programme. An examination of the relative magnitudes of the estimated variance components indicates that the important ones are σ²e and σ²gsy, which compute numbers of replications and locations as σ²e/σ²gsy, and σ²gsy/σ²gy, respectively. Broad sense heritability (h²) was calculated in grain yield and its components showing relatively high values. In general, Gs is augmented by numbers of replications, locations and years, and will be expanded by strong selection appearing high selection intensity. However, selection is usually conducted in early populations, sometimes which loss desirable genes. This will be protected by several selections as means of increase of number of replications. The most consideration for determining the number of experimental year (Y) is Gs/Y on programme of long or short term selections, no mather how Gs is augmented by increase of experimental years.