Decades of Faba Bean (Vicia faba L.) Breeding for Better Grain Yield and Seed Size has Inadvertently Reduced G × E Interaction and Increased Inter-Temporal Performance Stability

Thirteen faba bean varieties including 11 released between 1977 and 2007 and two promising genotypes were evaluated at seven contrasting environments in the central and southeastern highlands of Ethiopia during the main cropping seasons of 2007/2008 and 2008/2009. The objectives of the study were to evaluate temporal genetic progresses made over three decades of breeding in patterns of G × E interaction and performance stability of the varieties developed in due course for grain yield and seed size of faba bean. The study was conducted using a randomized complete block design with four replications. Regression coefficients (bi) of genotypes over years of release as a stability parameter showed a steadily but smoothly decreasing trend at the rate of 9.0 × 10−3 year−1 (r = −0.60; P ≤ 0.01), indicating that varietal performance stability increased with time. Trend analysis based on AMMI stability value (ASV), genotypic selection index (GSI), Shukla stability variance (σi2), and Kang’s rank sum (KRS) values of grain yield also revealed an increasing yield stability over the years of release. For seed size, GSI and KRS values decreased with time. The coefficient of variability (CV) for grain yield and seed size also tended to temporally decline while sustainability index (SuI) increased across the year of releases indicating that there was no performance stability sacrificed to achieve the greater yield potential with larger seed size. Therefore, this data support that selection of new genotypes that yield well at multiple environments, specifically genotypes with large seed size and resistance to disease as a method to increase performance stability.