MULTIVARIATE STATISTICAL-ANALYSIS OF 2-DIMENSIONAL NMR DATA TO DIFFERENTIATE GRAPEVINE CULTIVARS AND CLONES

Multivariate statistical methods were applied on two-dimensional NMR d ata (H-1-C-13) of polyphenol extracts from grapevine clones harvested in the Bordeaux region. An analysis of variance detected the most disc riminating NMR spectral correlation volumes which were then used to pe rform principal component, hierarchical clustering and discriminant an alyses. The results showed that clones are divided into three groups a ccording to the cultivar and they can be differentiated inside each cu ltivar. This method was applied to polyphenol extracts from grape seed s and leaves. Nowadays, two-dimensional NMR is the only analytical too l which can differentiate grapevine clones. It could thus be used to c heck the identity of grapevine species, cultivars or clones as well as any other plants which produce polyphenols. Copyright (C) 1996 Elsevi er Science Ltd