Barley malting quality: are we selecting the best?

This paper presents an analysis of several key barley malting traits from 2 mapping populations grown as a series of multi-environment field trials. The analysis is based on a mixed model that includes terms for all sources of variation in the trait data, namely varieties, variety by trial interaction, variation between plots in the field, and variation between samples in the laboratory processes required to measure the traits. The base-line model is extended to accommodate variance heterogeneity and correlated effects for many of the terms in the model. The results show that accounting for these phenomena may cause substantial changes in variety rankings. The methods of design and analysis proposed in this paper for barley quality trait data do not have common usage so if adopted in future there is an opportunity for substantial improvements in response to selection.