Alternative partitioning of the genotype-by-environment interaction

Area of Science:

• Agricultural Science
• Genetics
• Biometrics

Background:

• Genotype-by-environment (GxE) interaction is a critical factor in crop breeding and performance.
• Understanding the nature of GxE interaction is essential for optimizing breeding strategies and predicting cultivar performance across diverse environments.
• Existing methods for partitioning GxE interaction may not be suitable for an arbitrary number of genotypes or environments.

Purpose of the Study:

• To examine alternative methods for partitioning genotype-by-environment (GxE) interaction.
• To differentiate GxE interaction into components arising from heterogeneous variances and imperfect correlations.
• To assess the applicability of different partitioning methods based on the nature of genotypes and environments (fixed or random).

Main Methods:

• Two distinct methods for partitioning GxE interaction were investigated.
• Method 1: Based on heterogeneity among environments in scaling genotype differences.
• Method 2: Based on heterogeneity among genotypes in scaling environment differences.

Main Results:

• Both methods successfully separated GxE interaction into components related to heterogeneous variances and imperfect correlations.
• Method 1 is suitable for random genotypes in fixed or random environments, with interactions from genotypic scaling heterogeneity generally being unimportant.
• Method 2 is useful for evaluating fixed genotypes in random environments, with partitioning by genotype-specific interactions proving beneficial.

Conclusions:

• The proposed partitioning methods offer a more nuanced understanding of GxE interaction.
• The choice of partitioning method depends on whether genotypes or environments are considered fixed or random.
• Identifying specific interactions can guide the development of adapted genotypes for particular environments.