Combining ability analysis to identify suitable parents for heterosis in seed cotton yield, its components and lint % in upland cotton

Abstract

Combining ability and heterosis were studied in a 6 × 6 diallel cross to see the nature of gene action in Upland cotton ( Gossypium hirsutum L.) during 2002 to 2004. Analysis of variance revealed highly significant differences among all the F 1 and F 2 hybrid means and their respective six parental values for all the traits examined. In both generations, the mean squares due to general combining ability (GCA) and specific combining ability (SCA) were also highly significant. SCA genetic variances were greater than GCA and more important for the traits, i.e. boll weight, boll number and seed cotton yield per plant, showing the predominance of non-additive gene action. Lint % in both generations and boll weight in F 2s only were controlled by additive type of gene action due to maximum GCA variances. Cultivar CIM-1100 was found to be the best general combiner and its utilization produced valuable hybrids with desirable SCA in both generations. F 1 and F 2 hybrids, viz., CIM-1100 × CRIS-9, CIM-1100 × FH-682, CIM-1100 × BH-36 and CIM-109 × CIM-1100 as high × low and low × high parents performed well in SCA determination, outstanding mean performance and heterosis. Better SCA effects associated with useful heterosis were more pronounced for yield traits. In F 1 hybrids, maximum heterosis was observed for seed cotton yield followed by boll number, boll weight and lint %. The heterosis over better parent was +3.13 to +65.63% for bolls per plant, +0.75 to +24.40% for boll weight, +0.82 to +115.22% for seed cotton yield and +0.27 to +3.88% for lint %. Involvement of CIM-1100 in most of the F 1 and F 2 hybrids resulted in the synthesis of superior genotypes for most of the traits studied. Inbreeding depression was elevated in good performing hybrids and was the highest for seed cotton yield. Highest yielding F 1 hybrids yielded lesser in the subsequent generation due to over-dominance and inbreeding depression, whereas moderate yielding F 1 hybrids were found more stable even passing through process of segregation due to additive gene action. The combined performance of F 1 and F 2 hybrids could be a good indicator to identify the most promising populations to be utilized either as F 2 hybrids or as a resource population for further selection.