The Study Of Geographical Tracing Of Flatten-Shaped Tea Based On Stable Isotope Ratios And Mineral Element Fingerprints

Tea plant is perennial cash crop, whose quality has a close relationship with the climate conditions of producing area and processing method. The exist variation of the element content and stable isotope ratio in the flatten-shaped teas from different areas. Building the geographical origin discriminant model based on the element content and stable isotope ratio differences can provide a theoretical basis for the origin protection of tea products.The present study focused on the flatten-shaped tea with the similar appearance and processing procedures. According to the ionomic differences in flatten-shaped teas from different producing areas, the key indexes of Xihu Longjing tea were stepwise selected to establish a discrimination model and compared with several models using different methods. Main results for this study are as follows:（1） The stable isotope ratios in flatten-shaped tea produced from different areas showed the difference followed with the different producing areas. For the δD and δ18O in samples, maybe influenced by the land effect and the elevation effect; A positive correlation was found betweenδD and δ18O; The ratio of ⁸⁸Sr/⁸⁶Sr was increased following the latitude’s increasing of the four provinces. Compared the recognition ability among those models of FLDA, decision tree and BP-ANN, all the recognition ability were above 76%. As applied by external validation samples, decision tree model was suitable for geographical origin traceability of different regions, while the accuracy rate was 92%. So the decision tree model was more suitable for geographical tracing of Xihu longjing tea.（2） The means of REEs contents were in the range of 286.1⁹54.7 μg·kg-1, and enriched in the light fraction, while diluted in heavy fraction. The REEs contents in teas showed following increasing trend in Qingchuan of Sichuan, Zhejiang, Rizhao of Shandong. Significant correlationships were observed between the REEs with adjacent atomic number from Rizhao of Shandong and Liping of Guizhou, except for those from Zhejiang. Distribution patterns of REEs were different in these areas, especially for Ce and Eu. The allocation model of Ce in teas showed positive anomaly from Liping of Guizhou, but negative anomaly from Rizhao of Shandong, Qingchuan of Sichuan and Zhejiang. The allocation model of Eu in teas showed negative anomaly from Liping of Guizhou, Rizhao of Shandong and Zhejiang, but positive anomaly from Qingchuan of Sichuan. The cross-validated accuracy rates of linear discriminant analysis（LDA） were 73.33%, 94.12%, 85.45% and 83.33% in Shandong, Sichuan, Zhejiang and Guizhou, respectively. The content characteristics and distribution patterns were different in four tea producing areas, especially for Ce and Eu; LDA analysis can distinguish the flat teas from the four provinces.（3） Four data mining methods, including orthogonal partial least square discriminant analysis（OPLS-DA）, fisher linear discriminant analysis（FLDA）, decision tree 5.0（DT5.0） and back propagation artificial neural network（BP-ANN） were used to establish a model for geographical origin discrimination. All four methods showed the accuracy rate with over 92% under validation using training dataset. In cross-validation, FLDA exhibited the accuracy rate, with 92.17%, slightly higher than DT5.0 and BP-ANN. While in validation by independent dataset, OPLS-DA showed the best accuracy rate with 93.33%. These results confirmed that the multi-element fingerprint combined with chemo metric method is an effective way in geographical tracing of Longjing tea. FLDA and OPLS-DA are suitable for the recognition of geographical origin of Longjing tea.（4） Potential biomarkers existed in different flattened tea from Xihu, Zhejiang area outside of Xihu and other areas, respectively. Potential biomarkers between Xihu Longjing tea and flatten-shaped tea producing from Shandong, Sichuan and Guizhou were Sn, Pb, Cu, Mg, ²⁰³Tl/²⁰⁵ Tl and ⁹⁵Mo/⁹⁷ Mo, respectively. With respect to Xihu Longjing tea and Zhejiang flatten-shaped tea, the potential biomarkers were La, B, Li, ¹⁴³Nd/¹⁴⁶ Nd, ²⁰⁶Pb/²⁰⁷ Pb, ⁹⁵Mo/⁹⁷ Mo, δ13C, )Sr/Ca and Ba/Ca. ⁹⁵Mo/⁹⁷Mo was an unique potential biomarkers, which could be used alone to distinguish Xihu Longjing tea. However, its actual effect must be verified in future studies.