Grouping barley genotypes by regression-based and ANOVA-based clustering methods in multienvironments trials

Mohtasham Mohammadi

Abstract


Abstract
Multi-environmental trials are conducted to test of newly improved genotypes and to select the most favorable genotype(s). The genotype
environment (GE) interaction often conflicts the selection trend and the cluster analysis is efficient statistical tool in extracting useful
information on GE interaction nature. Fourteen barley genotypes and two local check cultivars including Izeh and Gachsaran were tested
for yield stability in four locations across three years. The trial of each environment was laid out in a randomized complete block design
with four replications. Combined analysis of variance revealed significant differences for genotype, environment and GE interaction.
Clustering of genotypes based on method 1 (intercept and slope of linear regression model) showed three distinct groups while using line
slopes (method 2) indicated two genotypic clusters. The determination coefficients of regression model were high and so, using these
clustering methods was useful in this dataset. According to dendogram of genotype main effect plus GE interaction (method 3), and
dendogram of GE interaction (method 4), there were nine and eight genotypic groups, respectively. In conclusion, genotypes G1 (3805
kg ha-1), G2 (3690 kg ha-1) and G6 (3591 kg ha-1) were found to be the most favorable genotypes, and could be recommended as good
new cultivars for national release. Such an outcome could be regularly applied in the future to exploration barley genotypes and other
crops based on regression or analysis of variance models in the other areas of the world.


Full Text:

PDF

Refbacks

  • There are currently no refbacks.