Adaptability and phenotypic stability of common bean genotypes through Bayesian inference.

Area of Science:

• Agricultural Science
• Genetics and Breeding
• Statistical Modeling

Background:

• Genotype x environment interactions are crucial for common bean (Phaseolus vulgaris) breeding.
• Understanding these interactions aids in selecting genotypes with broad or specific adaptability.
• Bayesian inference offers a robust framework for analyzing complex genetic data.

Purpose of the Study:

• To investigate genotype x environment interactions in common bean using Bayesian inference.
• To evaluate the efficiency of informative versus minimally informative a priori distributions.
• To identify common bean genotypes with high adaptability and phenotypic stability.

Main Methods:

• Six randomized block trials were conducted in Mato Grosso do Sul, Brazil.
• Thirteen common bean genotypes were assessed for grain yield.
• Bayesian inference with informative (meta-analysis concept) and minimally informative (high variance) priors was employed.
• Bayes factors were used for comparing prior distributions and model selection.

Main Results:

• Bayesian inference effectively identified common bean genotypes with high adaptability and stability using the Eberhart and Russell method.
• Informative a priori distributions yielded more accurate results than minimally informative ones, as indicated by Bayes factors.
• Specific adaptable genotypes for favorable environments included EMGOPA-201, BAMBUÍ, CNF 4999, CNF 4129 A 54, and CNFv 8025.
• Genotypes IAPAR 14 and IAC CARIOCA ETE showed specific adaptability to unfavorable environments.

Conclusions:

• Bayesian inference is a powerful tool for common bean breeding, enhancing genotype selection.
• The use of informative a priori distributions significantly improves the precision of genetic analyses.
• This study identified key common bean genotypes suitable for diverse environmental conditions in the region.