Evaluation of bread wheat (Triticum aestivum L.) genotypes in Ethiopia through multi-environment trial analysis using factor analytic mixed models

Abstract

Bread wheat is a vital crop for Ethiopia’s food security, with expanding cultivation driven by variety development efforts and strategic national initiatives. This study presents a multi-environment trial (MET) analysis to evaluate the genetic performance of 140 unique bread wheat genotypes, using data from 20 trials conducted across Ethiopia’s major wheat-growing regions. Each trial was designed using a row-column design (RCD) arranged in a rectangular array of plots. Four traits —grain yield (GYLD), hectoliter weight (HLW), days to heading (DTH), and thousand kernel weight (TKW) were considered in the analysis. We fitted a factor analytic linear mixed model (FAMM) based on a one-stage approach. Spatial models were fitted to account for field trends, while factor analytic (FA) model was used to model genotype-by-environment (G × E) effects. Model comparisons showed that the FA model consistently outperformed the diagonal (DIAG) model across all traits, offering greater flexibility capturing the variation in G × E effects. Heritability estimates were generally high, especially for DTH, which also showed strong genetic correlations across environments highlighting its reliability as a selection trait. In contrast, yield-related traits exhibited complex G × E patterns and variable heritability, reflecting their polygenic nature and environmental sensitivity. While the highest-yielding check variety, Balcha, retained its lead in GYLD, three genotypes—EBW202104, EBW202110, and EBW193155—demonstrated stable and competitive performance across environments for multiple traits, particularly grain quality and maturity. This makes them strong candidates for varietal release and registration for commercial production in Ethiopia. Future research should integrate genomic and environmental data.