per branch ((CB)), harvest index ((HI)), flowering completion height ((FCH)), and thousand grain weight ((TGW)) under field conditions across two years. The polygon-view of the accession-by-trait ((AT)) biplot model revealed significant genotypic variation among accessions, with distinct groups forming based on trait performance. In the first year, accession 13 (C-47) exhibited superior performance for GY, HI, and FCH, while accession 18 excelled in CP, CB, and SC. In the second year, similar trends were observed, with accession 13 maintaining its high GY, demonstrating low genotype-by-environment interaction across two years. The perfect accession plot confirmed these results, identifying accessions 13, 16, 23 (Saffire), and 29 (Lesaf-486) as optimal candidates for selection due to their proximity to the ideal trait coordinate. Trait analysis indicated that HI was the most influential determinant of yield performance across both years, followed by TGW in the second year. In contrast, CP and SC did not contribute to yield, suggesting that indirect selection should be approached cautiously, while the vector-length revealed that CB and CP were highly representative and stable across environments, making them reliable indicators. Finally, accession 13 (C-47) was identified as the best accession for yield performance, and could be released as a new cultivar for semi-arid regions.
