Classification of Marsala wines according to their polyphenol, carbohydrate and heavy metal levels using canonical discriminant analysis

Marsala is a popular Sicilian fortified aged wine with ancient tradition. Nowadays Marsala is exported all over the world and is considered one of the most important dessert wines. The aim of this study was to determine the concentration of carbohydrates, polyphenols and heavy metals in different types of Marsala wines and to achieve statistical classifications by stepwise forward canonical discriminant analysis (CDA). The obtained results provided evidence that different types of Marsala were correctly classified according to their phenolic and carbohydrate compositions. In particular, the residual sugars allowed a good discrimination among Marsalas having similar total sugar contents. CDA, performed using heavy metals as independent variables, showed that Superiore Ambra Secco and Vergine Marsalas were not discriminated, whereas a good separation among Fine Oro Dolce, Superiore Riserva and Fine Ambra Secco wines was obtained. Finally, an overall statistical model showed that the variables with the highest discriminant power were: tyrosol, caffeic acid, procyanidin B1, catechin, quercetin, kaempferol, lactose, rhamnose, zinc, copper and lead.