Ageing Status Characterization of Chinese Rice Wines Using Chemical Descriptors Combined with Multivariate Data Analysis
Fei Shen,Fangzhou Li,Dongli Liu,Huirong Xu,Yibin Ying,Bobin Li
DOI: https://doi.org/10.1016/j.foodcont.2011.11.019
IF: 6
2012-01-01
Food Control
Abstract:Wine ageing status identification is of great commercial and scientific interest, as wine quality and value are closely related to the organoleptic characteristics developed during the ageing process. In this study, Chinese rice wines from three well-known wineries (“Guyuelongshan”, “Kuaijishan” and “Pagoda”) were analyzed for 21 chemical parameters, including six conventional parameters, five sugars, lactic acid and nine macro-elements. Then the experimental data were subjected to multivariate statistical analysis to predict and classify samples of different ageing status (3, 9, 15, 21, and 33 months). Systematic differences between samples were revealed by a two-way analysis of variance (ANOVA) and principal component analysis (PCA). Discrimination model built by forward stepwise linear discriminant analysis (LDA) based on the 16 selected parameters achieved 88.5% accuracy in leave-one-out (LOO) cross-validation. The most five discriminant variables were Zn, Mn, alcohol, Cu and Al, respectively. When the discrimination was performed on the samples from each winery, the classification accuracy in LOO cross-validation was 97.7%, 91.1% and 78.0%, respectively. The results demonstrated that these chemical parameters have the potential to enable the authentication of ageing status of rice wine.