Isotope chemometrics determines farming methods and geographical origin of vegetables from Yangtze River Delta Region, China

Abstract

Shanghai city has encountered possible food fraud regarding the geographical mislabeling of vegetables for economic gain. A combination of δ¹³C, δ¹⁵N, δ²H and δ¹⁸O values and partial least squares discrimination analysis and support vector machine (SVM) methods were used for the first time to assess farming methods and determine the origin of vegetables from Shanghai city, Anhui and Zhejiang provinces. The results showed that 65.8% of Shanghai vegetables, 38.2% of Anhui vegetables and 23.6% of Zhejiang vegetables appeared to be grown using green or organic farming methods. The optimal discriminant model was obtained using SVM with a predictive accuracy of 100% for Shanghai vegetables. Zhejiang vegetables had a predictive accuracy of 91.7%, while it was difficult to distinguish Anhui vegetables from Shanghai or Zhejiang vegetables. Therefore, this study provided a useful method to identify vegetable farming methods and discriminate vegetables from Shanghai and Zhejiang.

Keywords: Farming methods; Geographical origin; PLS-DA; SVM; Stable isotopes; Vegetables; Yangtze River Delta Region.