Data fusion by ratio modulation of stable isotope, multi-element, and fatty acids to improve geographical traceability of lamb

Abstract There is a growing enthusiasm among consumers for high value lamb, especially the lamb with certain geographical indications, which may result in possible fraudulent labeling due to the opportunity for large economic profits. On such a basis, the aim of this study was to improve the discrimination rate of geographical traceability of lamb. To achieve this objective, lamb samples obtained from four different regions from China and New Zealand were analyzed. The chemical profiles of lamb samples from different origins were characterized by stable isotope analysis, multi-element determination, and fatty-acid profile, and further subjected to a series of statistical analyses, including principal component analysis, hierarchical cluster analysis, and principal component analysis-linear discriminant analysis. Compared with the result obtained by using each method individually, data fusion by ratio modulation of multi-technique clearly improved the discrimination performance for origin traceability. A satisfactory discrimination rate of 100% was obtained after ratio modulation of these techniques. Comprehensively, this study offers proof of concept for a promising method that could be used for lamb provenance authentication and fraud detection.