3.2. General Attributes of Musts and Wines
As pointed out above, the competition between grapevines and cover crops was not severe and this explains the absence of significant differences among treatments for must compositional attributes (
Table 1). Despite the fact that cover crops reduced berry size [
21], berries from the cover crop treatments did not accumulate more sugars (
Table 1). This is likely caused by a greater leaf surface in vines from ST that compensated the greater berry size in this treatment, as reported for other regions with similar climate conditions [
8,
40].
Although not significant, cover crop treatments tended to increase total acidity in Mencía musts, as previously reported for Cabernet Franc [
41]. This trend is more evident in the case of the SC treatment, where musts had lower total soluble solids contents and greater total acidities than the rest of the studied treatments. Finally, organic acids contents were very similar for all treatments, as observed for Tempranillo in La Rioja [
13]. Therefore, our results are in accordance with previous indications that soil management did not affect grape macro-constituents at maturity [
19,
42].
Wine composition was determined at the same time as the tasting sessions, five months after bottling, and soil management did not cause significant differences for any of the wine general attributes (
Table 2), in accordance with the observations in the musts. However, cover crop treatments significantly altered color intensity, color hue, and the concentration of total tannins (
Table 2). Wines from ER showed lower color hue values when compared to those from the other treatments, except for those from SC (
Table 2).
3.3. Amino Acid Composition of Musts and Wines
Although a previous study in another Galician winegrowing region (Ribeira Sacra) reported total amino acid contents in Mencía musts [
43], the current work is the first one to determine the concentrations of individual amino acids in Mencía musts (
Table 3). Independently of the soil management treatment, the total amino acid contents were lower in the current study than those reported for Ribeira Sacra [
43], although a high variability was observed in both studies. However, the concentrations of individual compounds (
Table 3) were within previously reported ranges [
44], except for those of tryptophan, which were slightly greater.
Soil management system did not affect the concentrations of amino acids in Mencía musts (
Table 3), due to the high interannual variability observed. In this sense, other authors did not find significant differences in the amino acid concentrations among Tempranillo musts coming from grapevines grown under soil tillage or cover crop treatments during the first year of their study, while the differences among treatments detected in subsequent years did not coincide between years [
45]. The concentrations of amino acids in musts depend on many factors, including grapevine variety, maturation stage, weather conditions, and management practices [
46], and a combination of these factors might have caused these differences with previous studies on other cultivars and also explain the absence of differences among the treatments considered in the current work.
The NV and SC treatments tended to cause lower (
p-value < 0.1) concentrations of amino acids in Mencía musts (
Table 3). In contrast, ER tended to increase the concentration of amino acids (
p-values < 0.08), such as valine, isoleucine, leucine, and phenylalanine (
Table 3). These modifications on the amino acid composition can be explained by the different water status observed on vines from the different treatments [
21]. Since these amino acids are precursors of wine volatile compounds [
47], these slight alterations might modify the wine aroma.
Independently of soil management, arginine was the most abundant amino acid in Mencía musts (
Table 3). Therefore, this cultivar can be considered an arginine accumulator, similarly to other red grapevine varieties, such as Syrah, Merlot [
48], Garnacha, and Pinot noir [
44]. Moreover, previous studies carried out by our research group proved that other three Galician grapevine varieties—namely Albariño, Godello, and Treixadura—are also arginine accumulators [
30,
31,
49].
In fact, arginine represented about 31.6% of the total free amino acids in Mencía musts, independently of the treatment (
Figure 2a). Other amino acids present at relevant concentrations in the musts from this variety were glutamine (12.6%), alanine (10.3%), glutamic acid (7.2%), and threonine (7.5%). These compounds were also abundant in other Spanish grapevine varieties, such as Monastrell and Verdejo [
48,
50,
51]. Soil management caused very slight differences on the percentages of each free amino acid, although they were greater in the case of minor amino acids (
Figure 2b). The high variability among samples prevented the detection of significant differences among treatments. The ER treatment presented the highest variability for most of these compounds (
Figure 2b). Therefore, soil management slightly altered the Mencía amino acid profile, and this could allow for discerning among management systems [
48].
When considering the eight most abundant amino acids in Mencía musts, representing around 84% of the total amino acid content, PCA was able to separate samples from each treatment (
Figure 3). The two first principal components (PC) explained 93.5% of the total variance in the dataset: PC1 accounted for 73.1% and PC2 for 20.4%. Musts from the ER treatment were located on the positive side of PC1 due to their high concentrations in arginine, threonine, serine, glutamic acid, glutamine, and alanine. Similarly, samples from the ST treatment were located also on the positive side of PC1, although close to the origin. The SC and NV treatments were located on the negative side of PC1; SC was on the positive side of PC2 due to its high concentration of aspartic acid, whereas NV was on the negative side of PC2 due to its high GABA content. These results agree with the fact that the amino acid profile could be a means for discerning vineyard management systems [
48].
Cover crops did not affect the amino acid concentrations of Mencía wines (
Table 4) due to the inter-annual variability on these concentrations, as observed for musts. However, wines from the cover crop treatments tended to have lower amino acid concentrations than those from the ST treatment (
Table 4); however, these trends were not significant at the 95% level. The total and individual concentrations of amino acids in Mencía wines were greater than those detected in red wines from other varieties [
52,
53]. This confirmed the significant effect of grapevine variety on the amino acid concentrations observed in wines and could allow detection of the wine origin [
48,
54].
3.4. Volatile Composition of Mencía Wines
Forty-three volatile compounds were quantified in Mencía wines (
Table 5). According to their chemical characteristics, they were grouped into higher alcohols, other alcohols, acetates of higher alcohols, esters, volatile fatty acids, terpenes, and other compounds.
Previous reports characterized the aroma composition of Mencía wines from other regions within Galicia [
22,
25,
43], so it is expected that the concentrations of volatile compounds observed in wines from the current study would differ from those previously published. For instance, the concentrations of 1-hexanol and benzyl alcohol were greater in wines from our study than in those previously reported [
22].
Soil management slightly altered the concentrations of volatiles in Mencía wines (
Table 5). According to the quantitative data, the total concentration of volatiles in wines from the four treatments ranged from 1616 to 1938 mg L
−1. Wine from ER had the lowest amount while that from the ST treatment had the highest. Methanol contents were greater in samples from ST than in those from the cover crop treatments. The other compound that showed significant differences due to soil management was
trans-linalool oxide (pyran), which appeared at lower levels in wines coming from ER when compared to those from ST (
Table 5). The rest of the volatiles did not show significant differences among soil managements, although we observed trends to greater concentrations of esters and volatile fatty acids in wines from the cover crop treatments. Moreover, 2-phenylethanol, which imparts rose nuances [
23,
55], was present at greater concentrations in wines from NV.
Vineyard-soil management effects on the volatile composition of wines disagree among studies. For instance, significant differences depending on the soil management system for almost all the volatile compounds quantified in Cabernet Sauvignon wines have been found; wines from the soil tillage treatment showed the lowest concentrations of volatiles [
19].
In contrast, higher levels of volatiles in Negroamaro wines coming from the cover crop treatments were detected when compared to those coming from soil tillage [
20]. These discrepancies are due to the different level of competence between cover crops and grapevines in each study, which greatly depends on soil and climate conditions in the study region. Furthermore, an effect of the grapevine cultivar cannot be discarded. In our case, the lack of differences in grapevine water status and the slight alterations in leaf surface and berry size caused the absence of alterations in wine volatile compound concentrations [
21].
In the current study, higher alcohols in wines from all treatments appeared at concentrations greater than the threshold considered to contribute negatively to wine complexity, 300 mg L
−1 [
56], although they were similar to those previously detected in commercial Mencía wines from Ribeira Sacra and Monterrei [
22]. Moreover, higher alcohols constitute more than 95% of the concentration of volatiles in Mencía wines from Valdeorras [
23]. Similarly, other alcohols that impart positive nuances, such as 2-phenylethanol, appeared at concentrations within the range reported for commercial wines from this variety [
22]. Moreover, the use of commercial yeasts could have homogenized the volatile contents in wines from the different treatments.
Only two acetates of higher alcohols were detected in the wines from the current study: Isoamyl acetate and 2-phenylethyl acetate, which did not differ among treatments (
Table 5). They appeared at concentrations similar to those of commercial wines from this cultivar [
22].
Ethyl esters constitute one of the most important groups of aroma compounds in wines because they impart fruity notes. Since yeasts produce them during fermentation as secondary products of sugar metabolism [
57], soil management is not supposed to alter significantly their concentrations, as occurred in this study (
Table 5). However, other studies reported higher concentrations of these compounds in Cabernet Sauvignon wines coming from the cover crop treatments when compared to those coming from soil tillage [
19], while the opposite was found in Negroamaro wines [
20].
Similarly, soil management did not alter the concentrations of volatile fatty acids in Mencía wines (
Table 5). These compounds originate from the metabolism of fatty acids by yeast [
57]. These molecules may contribute negatively to wine aroma, providing rancid and cheese notes; however, concentrations between 4 and 10 mg L
−1 of C6-C10 volatile fatty acids provide mild and pleasant aromas to wines due to synergistic effects [
58]. In this study, all wines had C6-C10 volatile fatty acid contents within this range (
Table 5).
Terpenes are relevant contributors to the wine aroma, and correlations between floral sensory attributes and high levels of some of these molecules, such as α-terpineol and linalool, have been documented [
1]. In our case, soil management did not affect the concentrations of terpenes in Mencía wines, except for that of
trans-linalool oxide (pyran) (
Table 5). Furthermore, wines from ER and SC tended to have greater concentrations of terpenes, which might be caused by the fact that precursors of these compounds abound in berry skin [
23,
59] and, in these two treatments, berries were smaller and, likely, they had a lower flesh:skin ratio [
21]. However, these compounds appeared in concentrations lower than their perception thresholds and their contribution to wine aroma would be not important. In fact, the contents of terpenes detected in the current study were lower than those previously observed in Mencía wines [
22]. Nevertheless, linalool and α-terpineol appeared at concentrations similar to those observed in Mencía wines from Valdeorras [
23].
When considering the OAV as an index of the contribution of a given compound to wine aroma [
33,
34], only 18 compounds had an OAV greater than 0.5 (
Table 6). Among these compounds, the ones with the greatest OAV were 1-propanol, ethyl octanoate, isovaleric acid, isoamyl acetate, and acetaldehyde. Soil management did not alter the OAV of the volatile compounds determined in Mencía wines (
Table 6) due to the high variability among samples. However, the OAV found in the current study were higher than those reported for the same cultivar [
36]. Ethyl esters had a great influence on the aroma of the wines from the current study, and cover crop treatments tended to induce higher OAV for these compounds (
Table 6). Ethyl butyrate, hexanoate, and octanoate (notes to strawberry, apple, and pear, respectively) showed high OAV in the wines of Mencía, in accordance with previous reports [
23].
3.5. Sensory Profiles of Mencía Wines
Figure 4 shows the sensory profiles for Mencía wines from the different soil management systems studied, as averaged for the three years considered.
Visually, wines from the ST had lower violet and cherry reflects than wines from the cover crop treatments (
Figure 4a). Aromatically (
Figure 4b), the wine from SC received the lowest scores for black, red, and ripened fruit. In contrast, wines from ST received the highest marks for these descriptors. Judges perceived floral notes more intensely in wines from the NV and EC treatments. Palate descriptors showed slight differences among treatments (
Figure 4c). Wines from ST received higher marks for persistence, but lower for body and green tannins. Finally, wines from NV, ER, and SC received global quality scores 6.5%, 9.7%, and 19.4%, respectively; lower than wines from ST.
These results seem in contradiction with those previously found [
6,
19], which showed that Cabernet Sauvignon wines coming from the cover crop treatments scored higher than those from the soil tillage control. However, this contradiction can be explained by the different climate conditions on each site, since in our case, grapes from the ST treatment were ripened when harvested in contrast to other studies [
6]. Although Xi et al. [
19] did not report data from musts in their study, wine alcohol in wines from the tilled soil treatment was significantly lower than that of wines from the cover crop treatments, which could lead to an unbalanced wine.
Furthermore, non-trained consumers tasted the same wines from this study and they ranked the wine from the ST treatment in fourth place [
21]. This clearly contrasts with the results from the current sensory analysis, adding information to the debate on the selection of wine tasters [
60].