Multi-trait selection for yield, earliness, and seed size in a cowpea (Vigna unguiculata) MAGIC Population: AMMI, WAASB, and MTSI analyses across six environments in Burkina Faso

---

Résumé

Cowpea is widely cultivated legume in Burkina Faso. However, rainfall fluctuation and soil poverty are major constraints to its production. This study aimed to identify superior cowpea genotypes combining favorable agronomic traits and earliness. To achieve this, 52 MAGIC genotypes and two checks were planted using an alpha lattice design with two replications at Kamboinsé, Farako-Bâ, and Saria during the 2019 and 2022 rainy seasons. The likelihood ratio test applied to the restricted maximum likelihood model showed a significant genotype × environment interaction for grain yield, hundred-seed weight, and days to 95% maturity. Heritability on genotype-mean basis (h²mg), reflecting genotypic efficiency in multi-environment trials, was low for grain yield (0.28) and high for maturity (0.67) and hundred-seed weight (0.77). AMMI analysis revealed significant mean squares for the first four interaction principal components, with the first two axes explaining 71.8% of the variance for maturity, 69.9% for yield, and 63.7% for HSW. A combined statistical analysis of stability revealed stable yield performance in G46, G29, G16, and G53 stable earliness in G12, and stable seed size in G20, G22, and G65. Based on a multi-trait selection index, genotypes G16, G58, G65, G46, G33, and G12 emerged as stable ideotypes combining high grain yield and early maturity, G56 and G29 excelled for seed size. Overall, G16 (MAGIC061) and G46 (MAGIC228) were confirmed as the most promising candidates.

Mots-clés

Agronomic performance, Genotype, Cowpea production, Multi-Environment Trials, Trait stability.