Abstract:
Research dealing with the occurrence and measurement of genotype X environment interactions of yield and other quantitative characters in economic crops has been reviewed. The conventional analysis of variance approach mainly considers the average performance of all genotypes in a trial grown over a range of environments covering different locations and years. This analysis of variance method has been further extended to incorporate a regression analysis of G X E interactions. With regression analysis approach, the significant G X E interaction sum of squares is partitioned into linear and non linear components, thus indicating whether G X E interaction is a linear function of environment or a more complex function. Different workers have shown that the performance of individual genotypes are linear functions of the environmental values to which they collectively subscribe. When the genotype response across environments is predominantly linear, the regression coefficient (b) parameter for each genotype provides a measure of its adaptation or phenotypic stability. On the other hand, when the non-linear (deviation) component is large, another parameter (deviation mean squares i.e. S d) is required to specify the stability of genotypes across environments .A stable genotype is defined as one with bi=1.0 and S di=0.
Page(s):
21-29
DOI:
DOI not available
Published:
Journal: Nucleus, Volume: 13, Issue: 3, Year: 1976