RECURRENT SELECTION FOR GRAIN-YIELD IN DURUM-WHEAT

Considerable improvements of durum wheat (Triticum turgidum L. var. du rum) genotypes with better yield have been achieved with different bre eding methods. Because of several constraints in its application in sm all cereals, recurrent selection in durum wheat has not been used exte nsively. The objectives of this research were to evaluate two cycles o f recurrent selection for grain yield and to assess effects of selecti on on other agronomic and yield-component traits. Nineteen semidwarf d urum wheat genotypes selected for high yield and broad genetic diversi ty, were intercrossed to form the base population. Fifty-five randomly selected St lines from three cycles of selection and a set of four ch ecks were planted in two 1989 and three 1990 environments, respectivel y. Genotype x environment interactions were generally of magnitude rat her than of rank order changes for traits evaluated and therefore a co mbined analysis across environments was performed. Regression analysis indicated an average increase in yield of 0.25 +/- 0.03 Mg ha(-1) per cycle of selection. Cycle 2 (C2) produced Lines with grain yield high er than any line in Cycles 0 (C0) or 1 (C1). Test weight and plant hei ght increased 8.39 +/- 2.01 Kg m(-3) and 2.62 +/- 0.3 cm per cycle, re spectively. Recurrent selection for grain yield did not affect spikele ts per spike and kernels per spike. Kernel weight increased 2.8 mg fro m C0 to C2. When calculated as a percentage of the total variance, the genetic variance of most of the evaluated traits decreased slightly. Genotype x environment variance was the most important component of va riances of five of the evaluated traits. The efficiency of recurrent s election for yield was demonstrated. Indications are that recurrent se lection solely for yield did not have any negative correlated response to unselected traits.