Evaluation of multivariate data analysis for marine mussels Mytilus edulis authentication in China: Based on stable isotope ratio and compositions of C, N, O and H

Abstract

Geographical traceability of marine bivalves was crucial to avoid food fraud, to guarantee food quality and to protect the interests of both consumers and importers. In order to evaluate the availability of using stable isotope ratio and content of carbon (C), nitrogen (N), oxygen (O), and hydrogen (H) combined with multivariate analysis for the origin traceability of mussels, 120 samples collected from four major production provinces in China were analyzed. Principal component analysis (PCA) was used for data exploratory analysis, and linear discriminant analysis (LDA) was used to evaluate their performance in terms of classification or predictive ability. Results showed that there was a significant difference (p < 0.05) in the eight variables in mussels from different origins, which proved that these signals were useful for identifying the origins of mussels. Based on the PCA, there was no clear distinction among the mussels from different regions. The LDA gave an overall accuracy rate of 92.8%, cross-validated accuracy rate of 86.7% and predictive accuracy rate of 92.3%. Present findings suggested that the stable isotope ratio and compositions analysis of C, N, O, and H could be potentially applied to the geographical origin traceability of China mussels assisted by multivariate data analysis.