Author Information

George C.J. Fernandez

Abstract

Genotype x Environment interactions results from the changes in the magnitude of differences among genotypes (non-crossover or quantitative interactions) or changes in the relative ranking of the genotypes (crossover or qualitative interactions) in different environments. Non-crossover interactions are usually associated with variance heterogeneity and non-additivity. The analysis of variance combined with joint regression analysis failed to differentiate between the crossover and non-crossover interactions. Tedious computations are necessary in comparisons of all possible pairs of genotypes in all possible pairs of environments in the crossover detection tests. Therefore, differentiating the non-crossover interaction caused by variance heterogeneity and non-additivity from crossover interaction by simple but effective methods such as exploratory data analysis should be carried out before assessing the stability in GEl studies. The effectiveness of the four graphical methods i) variance heterogeneity diagnostic plot (Box et al. 1978), ii) transformable non-additivity diagnostic plot, (Box et al. 1978) iii) Emerson and Hoaglin's (1983) non-additivity diagnostic plot, and iii) Gabriel's bi-plot (1971) in detecting non-crossover interactions resulting from variance heterogeneity and nonadditivity are presented in this paper. Baker's (1990) three simulated and the spring wheat data sets were used to evaluate the effectiveness of these four graphical techniques.

Keywords

Crossover and non-crossover interactions; variance heterogeneity; transformable non-additivity diagnostic plot; bi-plot; exploratory data analysis.

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS
 
Apr 28th, 3:00 PM

ANALYSIS OF GENOTYPE X ENVIRONMENT INTERACTION BY GRAPHICAL TECHNIQUES

Genotype x Environment interactions results from the changes in the magnitude of differences among genotypes (non-crossover or quantitative interactions) or changes in the relative ranking of the genotypes (crossover or qualitative interactions) in different environments. Non-crossover interactions are usually associated with variance heterogeneity and non-additivity. The analysis of variance combined with joint regression analysis failed to differentiate between the crossover and non-crossover interactions. Tedious computations are necessary in comparisons of all possible pairs of genotypes in all possible pairs of environments in the crossover detection tests. Therefore, differentiating the non-crossover interaction caused by variance heterogeneity and non-additivity from crossover interaction by simple but effective methods such as exploratory data analysis should be carried out before assessing the stability in GEl studies. The effectiveness of the four graphical methods i) variance heterogeneity diagnostic plot (Box et al. 1978), ii) transformable non-additivity diagnostic plot, (Box et al. 1978) iii) Emerson and Hoaglin's (1983) non-additivity diagnostic plot, and iii) Gabriel's bi-plot (1971) in detecting non-crossover interactions resulting from variance heterogeneity and nonadditivity are presented in this paper. Baker's (1990) three simulated and the spring wheat data sets were used to evaluate the effectiveness of these four graphical techniques.