Biomathematics & Statistics Scotland
The effects of growing barley cultivars together as mixtures have long been a subject of research at JHI–Dundee. Mixtures grown together have been shown to generally enhance yield and reduce disease, but there remain questions about which mixtures have satisfactory malting quality.
A trial of seven winter barley cultivars and all possible equal proportion 2-, 3-, 4-, 5- 6- and 7-component mixtures was grown at JHI–Dundee in a three-replicate split-plot design with and without disease control. Yield, thousand grain weight (TGW), and disease were recorded and grain samples were malted, giving measurements of several quality traits: cell wall modification (CWM); homogeneity (HOM); hot water extract (HWE); fermentability (FERM); soluble nitrogen (SN%); predicted spirit yield (PSY) and milling energy (ME). Mixed models were used to explore the contribution of individual cultivars to the quality of the mixture and to detect mixtures that were significantly different from the average of the individual cultivars.
Biplots proved a useful graphical display to visualise these effects. Principal components analysis summarises the relationships among the quality, yield and disease scores, providing new axes which can be used to display variation amongst the mixtures. The figure below shows a biplot of the monocultures and two-component mixtures. Mixtures that are located far from the midpoint of the corresponding monocultures indicate interactions. For example, cultivar A has the lowest TGW, while G has the highest among the monocultures, but mixture AG has a higher TGW still. Mixtures involving D tended to score higher for quality traits such as PSY, while those involving F generally had poor quality scores.
Further details from: Christine Hackett
Article date 2009
