Non-destructive fluorescence spectroscopy as a tool for discriminating between olive oils according to agronomic practices and for assessing quality parameters

Abstract

A non-destructive fluorescence method combined with chemometric algorithms has been developed for discriminating between olive oils. The excitation-emission matrices (EEMs) of two olive oil varieties (Arbosana and Oliana) from two crop seasons, which had undergone two different irrigation treatments (control irrigation strategy and regulated deficit irrigation (RDI)), were recorded. EEMs were analysed using parallel factor analysis (PARAFAC), followed by linear discriminant analysis (LDA) incorporating three PARAFAC components. This analysis was able to discriminate between olive oils according to crop season (100% of predictions in the validation set were correct) and variety (100% of predictions were correct). Moreover, good discrimination (80% of correct predictions) was also achieved when examining olive oils belonging to the same variety but submitted to two different irrigation treatments. Further, the olive oil quality parameters obtained using conventional methods were compared with those obtained using unfolded partial least squares (U-PLS). Good correlation coefficients were obtained for Rancimat hours (r = 0.87), K270 (r = 0.75) and total polyphenol content (r = 0.94).