Bayesian approach increases accuracy when selecting cowpea genotypes with high adaptability and phenotypic stability.

Area of Science:

• Agricultural Science
• Biostatistics
• Genetics

Background:

• Selecting high-performing crop varieties is crucial for food security.
• Cowpea (Vigna unguiculata) adaptability and stability are key breeding objectives.
• Bayesian methods offer a robust framework for statistical inference in agriculture.

Purpose of the Study:

• To validate the efficacy of a Bayesian approach for selecting upright cowpea genotypes.
• To assess cowpea adaptability and phenotypic stability.
• To compare informative versus minimally informative a priori distributions in Bayesian analysis.

Main Methods:

• Six randomized block trials were conducted to evaluate 17 upright cowpea genotypes.
• Grain yield data were analyzed using the Eberhart and Russell method within a Bayesian framework.
• Meta-analysis concepts informed the informative a priori distributions.
• Bayes factors were employed to compare prior distribution efficiencies.

Main Results:

• The Bayesian approach successfully identified upright cowpea genotypes with superior adaptability and phenotypic stability.
• Informative a priori distributions demonstrated higher accuracy compared to minimally informative ones.
• Bayes factors confirmed the superiority of informative priors in this selection process.

Conclusions:

• Bayesian analysis is a powerful tool for selecting cowpea genotypes with desired agronomic traits.
• The choice of a priori distributions significantly impacts the accuracy of Bayesian selection.
• Informative priors enhance the precision of adaptability and stability assessments in cowpea breeding.