Search

This paper presents an exploratory multivariate approach for analysis of malting barley quality data. By using principal component analysis (PCA) and partial least squares regression (PLSR) complex malting quality data are combined into functional factors which are used for malting barley quality characterisation. Fifty barley samples were used in this investigation, representing 15 spring barley and 10 winter barley varieties grown at two locations in Denmark. The samples were micro-malted and mashed and analysed for 13 quality parameters according to official methods of the European Brewery Convention. These data were combined and reduced into a few latent (functional) factors using PCA by which it is demonstrated that the modification of β-glucan plays a major role in both spring and winter barleys. Additionally, the spring barley and winter barley samples display different covariate latent structures, mainly in the nitrogen and diastatic power patterns. It is furthermore shown that graphic display as facilitated by exploratory data analysis, can be utilized in order to evaluate genotype-environmental interactions by considering the position and movements of the individual objects (genotypes in this instance) in the score plots. Thus, in contrast to the classical analysis of variance, the samples can be individually evaluated and the corresponding loadings can be used to validate the genetic and environmental effect of a given sample in a quality perspective.Several of the investigated malting quality parameters are highly intercorrelated. This fact is utilized by applying PLSR to barley and malt data for the prediction of wort quality in order to exclude the mashing step. This approach was successful for the modification-dependent wort parameters, extract, wort β-glucan and viscosity. [Copyright &y& Elsevier]