April 7, 2017

Near infrared spectral fingerprinting for confirmation of claimed PDO provenance of honey

Near ultraviolet spectroscopic fingerprints of Corsican honey samples were analysed by an array of chemometric tools in order to authorise their claimed origin. Genuine, unfiltered honeys (n= 373;219 Corsican and 154 non-Corsican) were gathered over the course of two production seasons, the principal goal was to create a particular spectral fingerprint for Corsican honey. After the preliminary data inspection by principal component analysis, the multivariate method analysed for provenance confirmed was partial least squares regression; numerous spectral pre-treatments were investigated. Best PLS discriminant models developed with the adoption of a full cross-validation, a variable selection algorithm and second derivative data pre-treatment gave correct classification results of 90.0% and 90.3% for Corsican and non-Corsican honey samples respectively. Using different calibration and validation samples from the honey collection, highest correct classification values of 90.4% and 86.3% for Corsican and non-Corsican honey samples respectively were acquired again using a variable selection procedure.