Applicability of joint regression and biplot models for stability analysis in multi-environment barley (Hordeum vulgare) trials

Abstract

GGE and AMMI biplot methods with Eberhart and Russell regression model were applied on the set of 18 barley (Hordeum vulgare L.) genotypes grown in 6 environments for quick and relevant method vis-a-vis to delineate genotype by environment interaction, stable genotypes and environmental discrimination. The average grain yield over the locations was depicted as 41.97 q/ha, which ranged from 31.82 (Karnal) to 55.52 q/ha (Bhatinda). The genotype DWRB 91 (47.51 q/ha) exhibited the highest grain yield followed by DWRB 121 (46.35 q/ha), DWRB 123 (46.04 q/ha) and DWRB 128 (44.70 q/ha) over the locations. In Eberhart and Russell model, the genotypes DWRB 124 and PL 880 were found suitable for favourable environments and DWRB 128 for poor environments. In AMMI analysis, IPCA 1 and IPCA 2 altogether captured 74.73% of the interaction mean squares, while in GGE biplot, PC 1 and PC 2 captured 36.51% and 26.44% interaction variation,respectively. The genotypes BH 992, DWRB 121, DWRB 123, RD 2897 and checks BH 902 and DWRB 91 were high yielding and as well as found stable in GGE and AMMI 1 biplot. The test environments Durgapura and Modipuram exhibited different niches, whereas, Hisar, Ludhiana, Bhatinda and Karnal were representative with better discriminating ability. Between biplot models applied, the GGE biplots were clear in visualization for polygon view, genotypic stability and environmental discrimination. The GGE method considered both G+GE for biplot generation and found most suitable for stability analysis.