Evaluation and Selection of Ethiopian Bread Wheat Varieties Using Linear Mixed Model

Abstract

Ethiopia is a country with diverse agro ecological settings which make the
yield of crop variety fluctuate and make selection of best variety difficult. This
study was conducted to select best bread wheat variety, variety that has
higher performance with relatively stable performance across varying
environments, by evaluating the varieties in terms their average performance
and stability across the test environments. The yield and stability of the
performance of 17 bread wheat genotypes across 18 test environments was
evaluated using linear mixed model. The yield performance of genotypes was
evaluated using different types of BLUPs. The stability of genotypes was
evaluated using different methods under mixed model assumption. Genotypes
such as ETBW5798, ETBW5800 and ETBW5879 were generally identified to
have higher average yield with relatively stable performance across the test
environments whereas Digelu, ETBW5875 and ETBW5899 were generally
found to have poor performance in terms of average yield and stability as
well. Currently very efficient experimental designs for agricultural
experiment such as IBD, and Lattice designs have been
introduced and such designs are not suitable for analysis using the usual fixed effect model and are better fitted using linear mixed model. So, a future multi environmental trial study that uses such designs must use the linear
mixed model.