Genome-wide Marker-based dissection of genetic variability for yield and yield components, and multi-trait selection in Kersting’s groundnut (Macrotyloma geocarpum)

Understanding genetic parameters and trait relationships is essential for effective breeding programs. This study evaluated fifteen yield components in 81 kersting’s groundnut accessions from Benin and Burkina Faso using additive and dominant marker-based relationship matrices and mixed effect models. Structural equation modeling was used to assess trait association patterns, while genotype by yield-trait biplot and multi-trait genotype-ideotype distance index identified superior genotypes. Significant accession and environmental effects were observed for most traits. Broad-sense heritability was moderate for yield ( H ²  = 0.39) and high ( H ²  > 0.60) for other traits. Seed width, length, thickness, 100-seed weight, emergence rate, and days to 50% flowering exhibited moderate narrow-sense heritability ( h ²  = 0.33–0.58), indicating additive gene action amenable to selection. High genetic gains were observed for emergence rate (84.09%), yield/plant (48.00%), and grain yield (44.19%), while low gains were found for number of pods/plant (0.32%), grain filling duration (2.60%) and pod width (3.72%). Grain yield exhibited significant positive associations with yield/plant, number of pods/plant, seeds/pod, 100-seed weight, pod harvest efficiency, and number of branches. These traits can guide indirect selection strategies for enhancing grain yield. Seventeen superior accessions with distinct strengths and weaknesses were identified as candidate parents for kersting’s groundnut breeding programs.