Combining multi-criteria decision analysis with agro-morphological-biochemical-molecular traits of interest for use in breeding in promising common bean breeding lines (Phaseolus vulgaris L.)

Abstract

Common bean is one of the most important legumes in the world. This study analyzed 26 promising breeding materials of Phaseolus vulgaris L. for 12 quantitative and 15 qualitative traits to identify elite lines for possible registration of new varieties. High variability was observed for most quantitative traits, with coefficients of variation (CV) from 8.87 to 67.04% and Shannon-Weaver diversity index (H’) values from 1.21 to 1.54. Qualitative traits had H’ values from 0.54 to 1.47. Seed yield per plant and total soluble solids showed considerable CVs (>42%), indicating significant variability among breeding lines. The H’ value also illustrated diversity among quantitative traits, with 100 seed weight showing the highest diversity (H’=1.54). Factor analysis of mixed data (FAMD) and hierarchical clustering on principal components (HCPC) grouped the breeding materials into two clusters and subgroups. The number of seed colors and the main seed color were key differentiators. Using a Composite Performance Index (CPI) from multi-criteria decision analysis (MCDA), the breeding materials were ranked, and five elite breeding lines were selected: KIS6B, KIS23B, KIS22B, KIS8B and KIS7B (CPI = 44.5–53.1). These finings offer valuable insights into the diversity of promising common bean breeding lines and provide a solid basis for the final selection and registration of new varieties.