Discrimination of oolong tea (Camellia sinensis) varieties based on feature extraction and selection from aromatic profiles analysed by HS-SPME/GC–MS

This study aimed to develop an objective and accurate analytical method to discriminate oolong tea varieties that easily causing adulteration by potential volatile compounds. A total of 75 oolong tea samples of five similar varieties (Tieguanyin, Benshan, Maoxie, Huangjingui and Jinguanyin) were analysed by headspace solid phase microextraction (HS-SPME) coupled with gas chromatography–mass spectrometry (GC–MS). The relative content of 26 major volatile compounds varied significantly according to variety, combined with the results of hierarchical cluster analysis (HCA), indicating that the varietal differences of aromatic profile remain significant for tea cultivars with very close origin. Principal component analysis (PCA) of the aromatic profiles showed that the feature of variety dominated over the other features (like producing region and quality). By stepwise linear discriminant analysis (S-LDA), 18 volatiles with the best discriminating capacity were selected, and 4 discriminant functions (DFs) enabled simultaneously discrimination of the five oolong varieties with 100% correct rate.