Study On The Tracing Technology Of Wolfberry With Geographical Indication Based On Mineral Eelement Content And Stable Isotope Ratio

Because of its edible and medicinal characteristics,Chinese wolfberry is deeply loved by consumers.Chinese wolfberry is a kind of characteristic agricultural product with high additional value,which is one of the geographical indication protected products.As significant correlation between sales price and origin of wolfberry,the adulteration with low cost shoddy or mislabeling the origin of wolfberry always occurs.This phenomenon has greatly affected purchasing confidence of consumers and the development of related industries.Thus,it is necessary to establish an accurate and reliable method for the discrimination of Chinese wolfberry.In this paper,ICP-MS was used to determinate 44mineral elements,and EA-IRMS was applied to determinate the ratio of stable isotope ratios,including carbon,nitrogen,hydrogen,oxygen and sulfur in Chinese wolfberry.Simultaneously,these variables including the content of elements and the ratio of stable isotopes were used to establish the identification model of provenance of Chinese wolffberry with chemometrics.The main result of research could be found as follows:(1)A method for determination of 44 elements in Chinese wolfberryA method for the determination of 44 mineral elements in wolfberry was established by microwave digestion combined with ICP-MS.After optimizing the microwave digestion program and parameters,the experiment was carried out.The results showed the spiking recovery was evaluated at three levels(n=6)in the range of 85.4-108.6%and the range of relative standard deviations(RSD)were 0.53?15.53%.The standard substances,carrot and GSBW10047,were also determined to verify the reliability of the method.The obtained results were in the range of standard certificates and consistent with the standard values.What'more,the values of linear correlation coefficient in the method was greater than 0.9999,the range of method detection limit was 0.009?0.103 mg/kg,and the range of limits of quantification was 0.030?0.343 mg/kg.The established method for the determination of elements in Chinese wolfberry has the characteristics of high efficiency,time-saving,sensitive and accurate,and wide linear range.(2)The discrimination of the origin of wolfberry based on the contents of mineral elements.In this study,ICP-MS was used to determinate the content of 44 mineral elements in Chinese wolfberry from Ningxia,Qinghai and Xinjiang provinces.The discriminant model for this regions was established due to Chinese wolfberry samples Ningxia and Qinghai has the high similarity and its number is large.The results of T-tset showed that 9 mineral elements,Sb,La,Tb,Lu,Al,Sc,V,Cr,Se,had significant difference between Chinese wolfberry samples from Ningxia and Qinghai provinces.Thus,partial least squares-discriminant analysis(PLS-DA)and back propagation-artificial neural network(BP-ANN)were used to establish the discriminant models of Chinese wolfberry samples from Qinghai and Ningxia,respectively.The results showed that in the PLS-DA model,when all wolfberry samples were uesd to establish the model,the sensitivity and specificity of the model were 100%and 97.5%,respectively.When 75%samples of wolfberry samples were used to establish the model,the sensitivity and specificity of the discriminant model were98.6%and 98.4%,respectively.What'more,when the model that was established by 75%samples was applied to predict the remaining 25%samples,the accuracy of the model was up to 100%.In the BP-ANN model,whether all samples were selected to model or 75%samples of samples were randomly selected for modeling,the sensitivity and specificity of the discriminant model were both 100%,the prediction accuracy of the model for 25%samples was 100%.By contrast,It can be found that the sensitivity and specificity of the BP-ANN model were better than that of the PLS-DA model when elements were applied to establish the model.(3)The discrimination of the origin of wolfberry based on the ratio of stable isotopesEA-IRMS was applied to determinate the stable isotope ratio of C,N,H,O and S in Chinese wolfberry samples from Qinghai,Ningxia and Xinjiang.Subsequently,the distribution and correlation of stable isotope ratio in wolfberry from Qinghai,Ningxia and Xinjiang with multiple analysis methods,such as boxplot,one-way ANOVA and principal component analysis(PCA).According to this analysis,linear discriminant analysis(LDA)and support vector machine(SVM)were used to establish the discriminant models.The results showed that?¹³C,?¹⁵N,?~2H and?¹⁸O contributed significantly to the identification of provenance,but?³⁴S was weak relatively in the identification of provenance.The results of PCA showed the cumulative variance contribution rate of the two principal components was up to 75.3%.In LDA model,the overall accuracy of validation set and training set was 89.7%and 93.9%,respectively.In SVM model,the overall accuracy of model verification and model training was 95.9%and 96.9%,respectively.In comparison,when stable isotopes were applied to establish the model,SVM model has better discrimination effect than LDA model.(4)Multivariate including elements and stable isotopes was combined to filter the characteristic factors for origin discriminantThe study was developed deeply to filter the characteristic factors for origin discriminant based on the elements combined with stable isotopes.To avoid isotopic ratios with negative values was deleted,partial least squares-discriminant analysis(PLS-DA)was selected to analyze contribution factors in Chinese wolfberry from different regions.The results showed that the discriminant model was R~2X(CUM)=0.676,R~2Y(CUM)=0.917,Q~2(CUM)=0.861,the values of Q~2were greater than 0.5.This indicated that the effects of the discriminant model was ideal when mineral elements and stable isotope ratio were combined.The discriminant effect of the model was satisfactory.According to the VIP values of these variables,such as Sc,La,Ca,Be,Ga,Sb,Th,?¹³C,V,Cr,Yb,Ba,Er,?¹⁵N,Mn,Zn,were selected because its values were greater than 1.The multiple variables could be used as effective indexes for identification of provenance of Chinese wolfberry.In brief,this paper has practical application significance,provided theoretical basis for further study on the origin of Chinese wolfberry.