Cultivar discrimination of Spanish olives by using direct FTIR data combined with linear discriminant analysis

Abstract

Attenuated total reflection Fourier‐transform infrared spectroscopy (ATR‐FTIR), followed by linear discriminant analysis (LDA) of spectral data, was used to discriminate olive fruits according to their cultivar. For this purpose, the spectral data of 136 olives coming from 17 cultivars, collected at different Spanish locations, were recorded. Up to 24 frequency regions were selected on the spectra, which corresponded to a peak or shoulder. The normalized absorbance peak areas within these regions were used as predictors. Although a good resolution was achieved among all categories, a second LDA model was also constructed to improve cultivar discrimination. With both models, evaluation set samples were correctly classified with assignment probabilities higher than 95%. Thus, it is demonstrated that FTIR followed by LDA of the spectral data presents a high potential to discriminate olives from a high number of different cultivars.Practical applications: Since virgin olive oil (VOO) quality can be affected by different parameters, such as the varietal and geographic origin, among others, the production of good‐quality VOOs should start with raw materials that possess well‐defined quality standards. Thus, it is necessary to develop analytical methods able to control raw material quality, and also to control its origin since olive oil producers have to include in their manufactured products both the genetic variety (monovarietal oils) and the geographical origin of the olives. In this work, a simple and quick ATR‐FTIR method, capable of predicting the cultivar of several olive samples, after spectral data treatment with LDA, has been developed. With the proposed method, it could be possible to establish in a reliable way to perform cultivar prediction of unknown raw materials arising from VOO samples.ATR‐FTIR data of Spanish olives were used to predict the olive cultivar using LDA.