**4. Discussion**

Compared to other studies on "minerality", the present analysis of this descriptor is an innovative approach as it combines sensory and chemical characterization.

Previous studies (followed a purely sensory approach to this aspect of wine. Our proposed models integrate chemical and sensory analysis in order to explain the "minerality" descriptor. Our theory is based on the idea that the role that this descriptor plays in sensory analysis cannot be explained by a unique chemical compound. The close examination of the data collected from the study revealed several models that integrate chemical and sensory descriptors.

Our results obtained after having performed the statistical analysis of the chemical composition with the sensory analysis sugges<sup>t</sup> that some volatile chemical compounds are involved in the application of the term "minerality", the origin of which is defined by the plant's metabolism. However, due to the fermentative activity of yeas<sup>t</sup> and bacteria these are transformed into active compounds from the sensory viewpoint; thus, the "minerality" of wine can also be dependent on winemaking techniques and enological itineraries used in the production of wine in the winery and during the process of wine aging, so it is not solely dependent on agroclimatic factors.

Concerning the olfactory spectrum, and based on the results obtained from the descriptive sensory analysis by the panel of expert tasters, different mathematical algorithms with more than reasonable predictive qualities were developed in the mineral description of wine at the olfactory level, which include the following terms or chemical compounds:

With a positive contribution: plant-chlorophyll, oxidation, octanoic acid, ethyl decanoate, isoamyl alcohol and benzyl mercaptan.

With a negative contribution: isobutyric acid, β-phenylethanol, isoamyl acetate and ethyl acetate.

The analysis of the proposed olfactory model revealed a positive association between oxidation and the term "minerality", suggesting that in those samples with less floral and fresh fruit impact, "minerality" is more likely to appear in a sensory characterization. In addition, there is a negative association with the term "oak descriptors" according to the predictive model developed. In the light of these results, it could be concluded that tertiary flavors, such as those created by the oak barrel aging, are negative, but the oxidation process not related to oak barrel aging is positive.

The mathematical model obtained by partial least squares regression for the gustatory "minerality" in white wines suggested a positive relationship with high total acidity levels. In addition, in terms of elements involved in the mathematical relations, the following compounds should be considered:

With a positive contribution: tartaric acid, glucose, fructose (in the form of residual sugars in dry wines or "o ff-dry" type), magnesium, sweetness (sugar), level of total acidity, alcohol (sweetness), body and bitterness.

With a negative contribution: pH, potassium and balance in the mouth.

In red wines, the proposed mathematical model achieved good results for accounting for gustatory "minerality". This model is positively related to organic acids, such as succinic acid with a salty taste and tannin concentration, and negatively related to the feeling of freshness and well-integrated or balanced acidity. The factors to be considered in this case are as follows:

With a positive contribution: alcohol (feeling of warmth), tannin concentration, gustatory depth, and persistence in the aftertaste, succinic acid, and potassium.

With a negative contribution: level of acidity in relation to freshness in the mouth, astringency of tannin and manganese.

As a conclusion, the present procedure, based on partial least squares regression, has demonstrated revealing and promising results. Based on these predictive models, it is possible to focus the technology used in the vineyard and the winery to imprint or to increase the mineral character in wine if required for reasons of marketing, product range, corporate communication, competitiveness or market strategy. The same approach is presumably equally applicable to other descriptors in the food sector that are nowadays not well characterized.

**Author Contributions:** Conceptualization D.M.D. and A.T.P.G.; methodology E.Z.S. and A.T.P.G.; validation E.Z.S. and A.T.P.G.; formal analysis E.Z.S.; investigation E.Z.S., A.T.P.G. and D.M.D.; resources E.Z.S., A.T.P.G. and D.M.D.; data curation A.T.P.G.; writing—original draft preparation E.Z.S.; writing—review and editing A.T.P.G. and D.M.D.; supervision A.T.P.G.

**Funding:** Laboratorios Excell Iberica S.L. and Outlook Wine S.L.

**Conflicts of Interest:** The authors declare no conflicts of interest.
