Exploring the Capability of LC‐MS and GC‐MS Multi‐Class Methods to Discriminate Virgin Olive Oils from Different Geographical Indications and to Identify Potential Origin Markers

Looking for a strategy to authenticate the declared origin of commercial extra virgin olive oils (EVOOs), 126 samples from six different Mediterranean geographical indications (GIs) are analyzed by means of two different platforms [LC‐ESI‐QTOF MS (in positive and negative polarity) and GC‐APCI‐QTOF MS (in positive mode)] combined to chemometrics. The sample treatment and chromatographic/detection conditions (in both platforms) are chosen to enable the comprehensive characterization of the complete minor fraction of the oils within a single run. Noticeable discrimination among the six evaluated GIs [Priego de Cordoba and Baena (Spain), Kalamata (Greece), Toscano (Italy), and Ouazzane and Meknes (Morocco)] is achieved building two‐class PLS‐DA models which consider the data coming from both platforms. The contribution of a few thousand molecular features to the statistical models is evaluated in depth and several compounds are pointed out as possible origin markers, describing characteristic compositional patterns for each GI in the evaluated crop year. The complementarity of the different analytical approaches is discussed and diverse strategies are used to identify the classifiers. Practical Applications: Protected Designation of Origin (PDO) and Protected Geographical Indication (PGI) labels are important tools to promote high quality EVOOs, assuring the connection to a particular territory and the characteristic combination of natural and human factors which make possible to obtain unique oils. In this context, it is imperative to furnish the control labs with innovative tools and methods which are able to provide extensive information about the EVOOs’ minor fraction (of unquestionable importance regarding its overall quality) in just one run and to give the chance to find and identify (and validate) origin markers. The utility of validated classifiers to authenticate the belonging (or not) of an EVOO to a particular GI is clear. The consumers’ confidence can be perceptibly undermined if the geographical name is used on products not having the expected qualities or if the production specifications are sometimes not followed by producers. The capability of LC‐MS and GC‐MS multi‐class methods to discriminate virgin olive oils from different geographical indications (crop season 2016‐2017) and to identify potential origin markers is demonstrated.