Yield is a complex trait influenced by multiple components and their genetic behavior. Therefore, this study was designed to understand the complex nature of yield by uncovering relationships among traits and estimating the genetic parameters of lentil germplasm. One hundred and ninety-two lentil (Lens culinaris) germplasm samples were evaluated at Jamma and Enewari (field experiment) and at Debre Birehan agricultural research center (pot experiment) in Amhara Region, Ethiopia, in 2024 and 2025. An alpha lattice design was used for both set of experiment, and data on 12 agronomic traits were collected. Genetic parameter estimations, correlation, path, principal component and cluster analyses were performed in R. The results revealed substantial phenotypic and genetic variation among the evaluated germplasm, with high broad-sense heritability for flowering, maturity, and seed size traits, and moderate heritability for seeds per plant, plant height, harvest index, and yield. Yield exhibited a strong phenotypic correlation with harvest index (r = 0.78 in field and r = 0.95 in pots), biomass (r = 0.77 in the field and r = 0.78 in pots), seeds per plant (r = 0.42 in the field and r = 0.60 in pots), and podding (r = 0.45 in the field and r = 0.69 in pots). Similarly, genotypic correlations were high with harvest index (r = 0.83 in the field and r = 0.96 in pots), biomass (r = 0.75 in the field and r = 0.80 in pots), seeds per plant (r = 0.0.59 in the field and r = 0.58 in pots), and podding (r = 0.39 in the field and r = 0.68 in pot), and both their direct and indirect effects on yield were significantly high. Therefore, indirect selection through traits such as pods per plant, harvest index, biomass, and seeds per plant would be more effective and reliable for improving seed yield than direct selection for yield itself as this is highly affected by environmental variations.
Kefelegn et al. (Tue,) studied this question.