Combining ability analysis for within-boll yield components in upland cotton (Gossypium hirsutum L.)

Cotton is an important cash crop worldwide, accounting for a large percentage of world agricultural exports; however, yield per acre is still poor in many countries, including Pakistan. Diallel mating system was used to identify parents for improving within-boll yield and fiber quality parameters. Combining ability analysis was employed to obtain suitable parents for this purpose. The parental genotypes CP-15/2, NIAB Krishma, CIM-482, MS-39, and S-12 were crossed in complete diallel mating under green house conditions during 2009. The F-0 seed of 20 hybrids and five parents were planted in the field in randomized complete block design with three replications during 2010. There were highly significant differences among all F-1 hybrids and their parents. Specific combining ability (SCA) variance was greater than general combining ability (GCA) variance for bolls per plant (9.987), seeds per boll (0.635), seed density (5.672), lint per seed (4.174), boll size (3.69), seed cotton yield (0.315), and lint percentage (0.470), showing predominance of non-additive genes; while seed volume (3.84) was controlled by additive gene action based on maximum GCA variance. Cultivar MS-39 was found to be the best general combiner for seed volume (0.102), seeds per boll (0.448), and lint per seed (0.038) and its utilization produced valuable hybrids, including MS-39 x NIAB Krishma and MS-39 x S-12. The parental line CIM-482 had high GCA effects for boll size (0.33) and seeds per boll (0.90). It also showed good SCA with S-12 and NIAB Krishma for bolls per plant, with CP-15/2 for boll size, and with MS-39 for seeds per boll. The hybrids, namely, CP-15/2 x NIAB Krishma, NIAB Krishma x S-12, NIAB Krishma x CIM-482, MS-39 x NIAB Krishma, MS-39 x CP-15/2, and S-12 x MS-39 showed promising results. Correlation analysis revealed that seed cotton yield showed significant positive correlation with bolls per plant, boll size and seeds per boll while it showed negative correlation with lint percentage and lint per seed. Seed volume showed significant negative correlation with seed density. Seeds per boll were positively correlated with boll size and negatively correlated with bolls per plant lint percentage and lint per seed. Similarly, lint per seed exhibited positive correlation with lint percentage and boll size showed significantly negative correlation with bolls per plant. Presence of nonadditive genetic effects in traits like bolls per plant, seeds per boll, lint per seed, seed cotton yield, and lint percentage is indicative of later generation selection or heterosis breeding may be adopted. For boll size, seed volume and seed density early generation selection may be followed because of the presence of additive gene action. The parental material used in this study and cross combinations obtained from these parents may be exploited in future breeding endeavors.