Utilisation of Elemental Fingerprinting Techniques and Isotopic Analysis to Authenticate Salmon Sold in Supermarket Seafood Supply Chains

In salmon products, fraud most commonly arises from the substitution of one species, production method, or geographic provenance for another, often after processing stages during which salmon's morphological characteristics have been removed. As the chemical composition of salmon flesh is fundamentally shaped by both diet and the surrounding environment in which salmon is grown or harvested, both X-Ray fluorescence and stable isotope analyses (δ¹³C and δ¹⁵N) were utilised in combination with statistical analyses, including random forest machine learning methods, to determine acceptability for use as a provenance testing method.

Random forest modelling demonstrated strong capacity to discriminate salmon samples by supplier (n = 10) and country of origin (n = 6), driven by rubidium (Rb), silicon (Si), δ¹³C, δ¹⁵N and light elements (LE). Modelling between suppliers within the same geographic location showed limited capability. The combined use of X-ray fluorescence and stable isotope analysis provides a cost-effective, efficient, and relatively non-destructive approach for salmon provenance authentication, with wide-ranging applications across the seafood supply chain.