Repeatability and stability analysis for fiber traits in upland cotton (Gossypium hirsutum L.)

Cotton fiber quality is defined by several traits, which are affected by genotype×environment interaction (GEI). The objectives of this work were to study GEI of fiber traits, the repeatability and the effect of the size of testing environments in detecting cultivar genetic differences by High Volume Instrument (HVI), and the usefulness of two GEI derived stability indices, σ2i and GGE Instability (GGEIN), and one G+GEI derived index, GGE Distance (GGED). A five-year balanced dataset (“5-yr”), included 56 trials, planted in Greece, was divided in one-year (“1-yr”) and two-year (“2-yr”) datasets for each of ten fiber quality traits and lint yield (LY) to estimate repeatability in ranking cultivars using Spearman’s rank correlations along with heritability on broad sense (H). In conclusion, the HVI system applied in two replications and two years represented by 20-24 environments was effective in discriminating cotton cultivars for the majority of quality traits. In particular, lint percentage (L%) and fiber elongation were the most repeatable traits and they could be estimated precisely using 10-12 environments (“1-yr” data). Fiber upper half mean length, strength, uniformity, yellowness (+b), reflectance (Rd) and LY could precisely be estimated by 20-24 environments (“2-yr” data). Fiber grade, micronaire and short fiber index (SFI) were moderately repeatable using “2-yr” data. Neither Shukla stability variance (σ2i) nor GGEIN were repeatable in ranking for stability, whereas GGED index was similar repeatable to mean values but superior in consistency and of higher repeatability for quality traits with high GEI.