Genotypic stability and adaptability in tropical maize based on AMMI and GGE biplot analysis

Area of Science:

• Agronomy
• Genetics
• Biometrics

Background:

• Evaluating hybrid performance across diverse environments is crucial for crop improvement.
• Additive Main Effect and Multiplicative Interaction (AMMI) and Genotype x Genotype-Environment Interaction (GGE) biplot models are widely used for stability and adaptability analysis.
• Understanding the predictive accuracy of these models is essential for reliable hybrid selection.

Purpose of the Study:

• To assess the phenotypic and genotypic stability and adaptability of 45 double-cross hybrids.
• To compare the accuracy of AMMI and GGE biplot models in predicting hybrid performance.
• To determine the best model for selecting stable and adaptable hybrids.

Main Methods:

• A complete diallel cross was performed using 10 single-cross hybrids.
• The resulting 45 double-cross hybrids were evaluated across 15 locations in Brazil.
• AMMI and GGE biplot analyses were conducted, considering random effects and employing cross-validation.

Main Results:

• When random effects were considered, best linear unbiased predictors (BLUPs) from AMMI and GGE analyses showed higher accuracy.
• The GGE biplot retained more genotype-environment (GE) and genotype plus genotype-environment (G + GE) interaction in its graphical analysis compared to AMMI 1.
• Despite better graphical representation, GGE biplots were less accurate than AMMI 1 graphs based on cross-validation results.

Conclusions:

• Graphical retention of GE and G + GE is not solely indicative of predictive accuracy for hybrid stability and adaptability.
• The AMMI 1 graph demonstrated superior predictive accuracy over the GGE biplot in this study.
• Careful consideration of cross-validation is necessary when selecting between AMMI and GGE models for hybrid performance evaluation.