Description
Fish identification on the market can be very challenging for both consumers and experienced inspectors in the cases of fish sold as fillets (Acutis et al., 2019). In this field, reference, and gold standard analyses to identify animal species generally require rather long processing times, but quick decision-making is fundamental in preventing and counteracting frauds (Grassi et al., 2018).
This study investigated the performance of NIR spectroscopy as a fast and non-destructive method to distinguish between two very similar flatfish species, namely the Guinean sole (Synaptura cadenati) and European plaice (Pleuronectes platessa). Fifty fillets of each species were analysed using three near-infrared (NIR) instruments: the handheld SCiO (by Consumer Physics) and MicroNIR (by VIAVI), and the benchtop MPA (by Bruker). All the collected spectra were processed by applying the same chemometric approach, i.e., pre-processed and used to build PLS-DA classification models, whose performances were evaluated and compared. All the three instruments provided very good results, showing high accuracy: both SCiO and MicroNIR reached 94.1 % accuracy, while MPA spectrometer reached 90.1 %. Moreover, a thorough interpretation of actual chemical signals, as recorded by the three NIR instruments, was provided.
The good results in classification obtained by combining NIR spectroscopy and simple chemometric modelling techniques suggest a direct applicability of the method, also using cheap portable instruments, both in the context of real-world marketplaces and in official control plans.