Genetic gains for grain yield in two selection phases of a wheat breeding program.

Evaluation and selection of experimental lines over several environments is critical component of wheat breeding programs before release of cultivars to growers. Fifteen-year (1989-90 to 200405) data from three consecutive trials (A, B and Microplot) of the Cereal Crops Research Institute, Pirsabak (Nowshera) were analyzed as two independent selection phases to estimate selection differentials, genetic gains and realized heritability for grain yield. Wheat lines tested in A-trials ranged from 108 to 378 year -1 with selection intensity of 10 to 30% vs 72 to 198 year -1 in B-trials with selection intensity of 6 to 22%. Selection differentials and genetic gains were positive for each pair of years during the 15-year period. Averaged across 15-years, mean selection differential, genetic gain and realized heritability were 8.9 vs 5.3%, 6.0 vs 7.2%, and 0.63 vs 0.56 under Phase– I and Phase-II, respectively indicating more selection efficiency in Phase-I. The selected wheat lines out yielded the check cultivars throughout the 15-year period in A-trials, while the checks surpassed the selected lines in 12 of the 15-years in B-trials. An upward trend in grain yield ha -1 was generally followed by a decline both under A and B-trials. The statistical procedure used is effective for estimating genetic improvement for important traits in multi-stage crop breeding programs.