Abstract:

In order to discriminate among red wines according to their geographical origins, the GC fingerprint profiles of
85 samples from 10 main producing countries such as France, South Africa and Australia were cluster analyzed by using a
chromatographic analysis and data management system for traditional Chinese medicines (TCMs) and statistical analysis
software SPSS 19.0. A classification model was established for discriminant analysis of the grographic origin of red
wines. The results showed that among red wine samples from different countries, there were differences in GC fingerprints
mainly in terms of the relative peak area and the number of peaks, suggesting that each sample from different origins has
characteristic GC peaks. The results of cluster analysis showed that 10 geographic origins could be classified effectively.
Two commercial red wines were analyzed by using this method. One of them might be from France and the other might be
from Italy. The established method proves to be effective at discriminating among red wines according to their geographical
origins based on GC fingerprint profiles.

Key words: red wine, gas chromatography, fingerprint, geographical origin traceability