G × E models for yield adaptation and target environment analysis in barley (Hordeum vulgare)

Abstract

The present study was conducted to find out the best stability models out of methods, viz. Eberhart and Russell regression (ER model), yield stability index (YSI), AMMI, unscaled GGE and heritability adjusted GGE (HA-GGE) with target environment delineation in multi-location barley yield trials conducted at 08 locations during rabi, 2016–17. The pooled analysis revealed significant mean squares and large variations were attributed by the location effect (56.38%) followed by G × E (21.06%) and genotypes (7.77%), respectively. The initial two PCs exhibited 30.14 and 20.51 % variations in HA- GGE, which was slightly lower for PC1 and was marginally higher for PC2 than unscaled GGE. The which won where and mean vs. stability of GGE biplots model were useful to judge crossover G × E and in selecting specifically adapted genotypes easily. The YSI concluded based on grain yield and stability value simultaneously, hence found reliable than AMMI stability value. The locations Pantnagar and Modipuram were discriminating for genotypic differences, while the environments Durgapura and Ludhiana were found representative and discriminative for future barley yield trials. The genotype DWRB160 and two-row malt barley checks DWRB123 and RD2849 were found consistent and promising. Therefore, we suggest applying HA-GGE in coordinated barley yield trials to identify representative locations and thereby curtailing evaluation cost.

Keywords

• ER regression
• G × E
• HA-GGE
• Unscaled GGE
• YSI