Abstract

Variance analysis and graphical biplots were used to understand the nature of genotype × environment interaction (G × E) in a grain yield data set obtained from 39 barley (Hordeum vulgare L.) genotypes grown in 18 environments (a combination of three sowing dates, two crop protection treatments and three years) at Holetta, central highlands of Ethiopia. Genotype × year interaction was much more important than genotype × management interaction. Season-end drought was the environmental variable and time to maturity was the genotypic variable responsible for the high G × year interaction variance. An elite breeding line gave the highest mean yield and was the best under low but not under high season-end drought stress. Sasa, an early maturing landrace, was the best in a year of high season-end drought. Biplots enabled visual identification of compromise genotypes such as 3304 - 11 and 3381 - 04 that yielded reasonably well under both low and high season-end drought conditions. Selection for post-anthesis drought tolerance may result in high and stable yields across years and wider geographical adaptation in Ethiopian barley. The importance of unique landraces for stress situations is ascertained.

Key words: Barley landraces, genotype × environment interaction, GGE biplot, season-end drought, sowing date, insecticidal seed treatment, fungicidal disease control.