One of the successful ways in breeding program and determining breeding methodology is information about the type of gene action involved in the inheritance of a character in different wheat generations. The objective of this study was to estimate gene effects for grain yield traits in wheat (Triticum estivum L.) in four selected crosses, involving five parents, their first and second filial generations (F1, F2) and first back crosses generations (BC1). The gene effects were estimated on the basis of generation mean analysis, using an additive-dominant model with three and six-parameters. The adequacy of the additive-dominance model with three-parameters was tested using the Scaling test and Chi-square (χ2) test. The significant Scaling tests and Joint-scaling test indicated the presence of digenic epistasis for the most of studied traits. The application of six-parameter models revealed that genetic effects for certain traits were depending upon the particular cross and differed on an overall basis. The significant type of non-allelic interactions which was observed in most of cross combinations was additive×additive (i) and dominance×dominance (l). Duplicate gene interactions were also observed, which are quite difficult to exploit in breeding programs. From these observations it is suggested that selection for the improvement of some traits should be effective in advanced segregating generations due to non-allelic gene interaction, while in cross combination where epistasis did not make a significant contribution among the generation means, selection in early segregating generations would be effective.
Key words: Epistasis, Gene action, Wheat, Yield
|
