3.2. Yeasts Diversity during the Spontaneous Fermentation
A great diversity of native yeasts was observed (
Figure 2). Non-
Saccharomyces yeasts were abundant in fresh grape juice (day 0 of fermentation), with Pinot, S. Blanc, and C. sauvignon showing the greatest diversity among all grape cultivars. Similarly to our results, Jara et al., (2016) reported several non-
Saccharomyces yeasts in fresh grape juices from different cultivars in the Maule region (Chile), in which
Metschnikowia,
Hanseniaspora, and
Rhodotorula were the most abundant [
51]. Although these genera were also observed in our study, they were not the most abundant in all the grape cultivars we studied, in which a greater diversity was observed.
These difference might be attributed to the different grape cultivars characterized in the studies, while Jara et al., (2016) characterized yeast presence in Petit Verdot, Alicante Bouschet, Torontel, and Mencia cultivars, while we focused on the characterization of Cabernet Sauvignon, Carignan, Chardonnay, Merlot, Pinot, Syrah, and S. Blanc, which present different intrinsic factors, such as nutrient content, that might influence the microbial colonization as well the agronomic practices followed [
72,
73,
74]. In addition, environmental factors, such as rainfall, have been also associated to difference in yeast abundancy [
9,
11,
51]. In grapes, the microbial load is reduced as rainfall increases due to a microorganism wash-out. According to the Environmental Ministry of Chile, no rainfall was reported between March and April 2013, the year during which grape samples were collected for our study. While precipitations were reported for the same months in 2015, during which the study of Jara et al., (2016) took place [
75].
The geographical location of the vineyard from which samples were collected might also influence the microbial composition [
76]. The Palo Alto vineyard is located in one of the central valleys of Chile (35°26′ S 71°40′ W, Chile,
Figure 1), at about 100 m.a.s.l and has a temperature ranging from 17 to 23 °C during the harvest season [
77]. Given these conditions, and that the grape vines were collected towards the end of the harvest season, they might have had a greater sugar content, which favors the establishment of a diverse microbial population [
78]. In addition, the vinery location is characterized by torrid weather with almost no rainfall from October to March, and rainy winters from April to September. However, the entire region is affected by the “La Niña” climatic event, which operates as an eventual factor that specifically affects the climatic conditions associated with rains causing periodic drought. These conditions not only affect the quality and production of the vineyards, but it is also very possible that they affect the types and proportions of native microorganisms in the grapes [
75].
As the spontaneous fermentation progressed, most of the non-
Saccharomyces yeasts decreased their presence in the must, even disappearing on day 7. This might be explained by its low resistance to the harsh environmental conditions, characterized mostly by the lack of nutrients and high ethanol concentrations [
23,
24,
26,
27,
49]. On the contrary, the
Saccharomyces genera that showed a relative abundance from 10% to 40% at day 0 take advantage of these conditions, and therefore become dominant.
In spite of this dominance, some non-
Saccharomyces yeasts genera were also detected after 7 days of fermentation:
Hansenioaspora,
Metschnikowia,
Candida,
Lachancea, and
Torulaspora (
Figure 2). The yeast distribution observed was associated with the type of grape juice studied. For example, in the case of the genus
Hanseniaspora, it was detected in all the grape juices evaluated except in Merlot.
Torulospora was detected in Carignan and Merlot, while
Metshnikowia was detected in C. Sauvignon, Sauvignon Blanc, and Pinot. In the case of the
Candida genus, its presence was found in C. Sauvignon, Merlot, Pinot, and Chardonnay. The presence of these non-
Saccharomyces genera on day 7 of spontaneous fermentation presupposes the peculiarity of being resistant to ethanol. Although non-
Saccharomyces yeasts have low tolerance to ethanol, our results show that some can be isolated once the alcoholic fermentation was completed (
Table 2). Catrileo, Acuña-Fontecilla, and Godoy (2020) reported that
T. delbrueckii YCPUC10, isolated from C. Sauvignon grape juice, was able to persist with ethanol levels of about 11% as the result of an adaptative metabolism [
79]. Furthermore, certain non-
Saccharomyces are able to increase their ethanol tolerance by producing succinic acid [
80] and this might be the case of the
Metschnikowia, Torulospora, Hanseniospora, and
Candida spp. detected at the end of the fermentation. However, more studies are necessary to fully characterize the capability in the isolated strains.
These results are interesting, since they would allow us to investigate local (
terroir-specific) strains that are able to positively impact the sensory profiles of wines, probably in co-fermentations with
S. cerevisiae, or even in monocultures. Increases in ethanol resistance might be advantageous in order to allow for the production of other metabolites associated to the non-
Saccaromyces yeasts. For example,
Metschnikowia spp. is capable of the formation of esters, monoterpenoids, and higher alcohols due to its β-glucosidase activity, which can improve the aromatic profile in wines [
81,
82].
Torulaspora delbrueckii has been associated to the production of mannoproteins and aroma compounds that increase the mouthfeel in red wines [
83,
84].
A yeast-like fungus (
Aureobasidium) was detected with an abundance of approximately 10% in the fresh grape juice samples. Species of
Aureobasidium have been reported as able to survive in grape juice [
23,
85] and have been isolated from different grape cultivars in the Maule region [
51], so their presence is not unusual. However, in terms of their presence at the end of the fermentation process, this might indicate a contamination problem [
23,
86].
The culture-dependent analysis showed a diverse make up of yeasts, although less varied than the molecular analysis. Similarly to what was observed with the molecular analysis, the dominant yeast genus in 7-day fermented grape juice was
Saccharomyces (39%) (
Figure 3). However, a significant diversity of non-
Saccharomyces yeasts was also isolated. The most abundant yeasts were
Hanseniospora (13%), followed by
Candida (11%),
Metchinoskowia (10%),
Lachancea (10%),
Starmerella (7%), and
Torolospura (5%). The yeast-like fungus
Aureobasidium was also isolated (5%). The yeast-like fungus has also been isolated from spontaneous Grignolino grape fermentation [
87]; therefore, their presence is not unusual.
Table 2 shows an abundance of yeasts according to species found in the different musts evaluated;
L. thermotolerans,
M. pulcherrima,
H. uvarum,
T. delbrueckii,
C. oeophila,
Starmerella bacillaris (formerly called C. zemplinina),
C. stellata, and
A. pulluans. The Shannon diversity index calculated for each grape cultivar ranged from 0.17 to 1.15 (data not shown), evidencing, as reported previously, the diversity of yeast within different cultivars even if they are cultivated in the same vineyard [
87].
Similarly to the molecular analysis, single species distribution was also variable:
T. delbrueckii was only isolated from C. Sauvignon fermentations;
A. pullulans from Chardonnay and Syrah;
C. oleophila from Chardonnay and Merlot;
C. stellata from Chardonnay, C. Sauvignon, and Merlot;
St. bacillaris from Carignan, Chardonnay, and Merlot;
M. pulcherrima from C. Sauvignon, Pinot, and S. blanc;
H. uvarum from Carignan, C. Sauvignon, Pinot, and Syrah; and
L. thermotolerans from Carignan, Chardonnay, Merlot, and S. blanc (
Table 2). The uneven distribution of yeast species from a single vineyard might be explained by the many microclimates created within the vineyard [
11,
87]. The non-
Saccharomyes microbiota was dominated by
L. thermotolerans,
H. uvarum, and
M. pulcherrima, representing 54% of the yeast population. In correlation to the molecular characterization, after 7 days, the Merlot fermentations showed the greatest isolate diversity. On the other hand, as expected,
S. cerevisiae was isolated transversally in almost all the fermentations, showing the higher numbers of isolates across all grape cultivar samples (
Table 2,
Figure 3).
The diversity observed in this study is in line with that previously reported for spontaneous grape juice fermentations [
10,
11,
25,
26,
27,
49,
87], in which a number of non-
Saccharomyces yeast were detected and/or isolated primarily at the early stages of the fermentation process. In our study, non-
Saccharomyces yeasts were also isolated after 7 days of fermentation, which indicates that the species have a certain tolerance to the ethanol concentration (13.08 ± 0.03 to 14.2 ± 0.11%
v/
v) and nutrient deprivation (glucose and fructose: 1.70 ± 0.05 to 2.12 ± 0.05 g/L, respectively) conditions reached in 7 days of the spontaneous fermentations (data not shown). Given the scare information available regarding yeast diversity from Chilean grape varieties and
terroirs, the results obtained contribute to broaden the literature in this topic. Furthermore, with the aim of creating signature and different wines, wineries are exploring the possibility of spontaneous and biodynamic fermentation [
88], and therefore, the characterization of the natural microflora and its intrinsic dynamic during the alcoholic fermentation might be crucial to determine correct process that end in stable wines.
3.3. Lactic Acid Bacteria Diversity during the Alcoholic Fermentation
LAB compose a part of the microbiota found in grapes and winery equipment; they are important in the winemaking process, but their identification and isolation have been limited mostly to grapes and during malolactic fermentation [
89]. In our study, the impact of the alcoholic fermentation and the LAB diversity was studied. Five genera were detected in the fresh and fermented grape juices (
Figure 4). In fresh grape juice, the dominant genus was
Leuconostoc, followed by
Lactobacillus,
Lactococcus,
Pediococccus, and
Weisella. Recently, Kačániová et al., (2020) studied the LAB composition in 10 different grape cultivars (grapes, grape juice, and wine), including Pinot, C. Sauvignon, and Merlot, and reported a similar genera make-up, with
Lactobacillus as the dominant genera, followed by
Leuconostoc [
90].
In spontaneous fermentations, the bacterial diversity is more abundant than in an inoculated process [
48]; however, in our study, the alcoholic fermentation negatively affected the LAB abundance, and after 7 days,
Lactobacillus and
Pediococcus decreased significantly (
Figure 4). Reduction of LAB populations is expected as the alcoholic fermentation proceeds, the environment becomes richer in ethanol, and sugars are depleted [
20]. Although these changes occurred after 7 days of fermentation,
Leuconostoc remained as the dominant genera. According to Piettet et al., (2011), LAB resistance to ethanol is not strain dependent, and among the 61 LAB tested by the authors, all showed great resistance to ethanol concentrations ranging from 0.5% to 14%
(v/
v) [
91]. This suggests that an important factor for the reduction in LAB diversity during alcoholic diversity might be the depletion of sugars (20) associated to other factors such as the phenolic content [
92]. Similarly to the yeast diversity, the grapes varieties, cultivation conditions, climate, and geographical location of the vineyard affect the LAB composition. For example, Berbegat et al., (2020) reported that the LAB consortium of Uva di Troi grape juices spontaneously fermented, and was composed of
Oenococcus,
Acetobacter,
Proponibacterium, and
Gluconbacter as the dominant bacteria.
Lactobacillus was also detected but in lower abundance (10%) [
48].
For the culture-dependent analysis, the colonies observed in MRS agar plates were grouped according to their morphology, and these initial observations allowed for the classification of the 58 isolates in three different clusters. Molecular 16S rRNA identification showed that the LAB isolates belonged to the
Leuconostoc and
Lactobacillus genera with three species identified (
Table 3). The most abundant LAB was
Leuconostoc mesenteroides, which was isolated from all fresh and fermented grape juice, while
Fructilactobacillus fructivorans and
Lactobacillus delbrueckii ssp
delbrueckii were only isolated from the Carignan and Pinot fresh grape juice, respectively. Berbegal et al., (2019) reported that a heterogenous consortia of LAB are present in spontaneous most fermentations [
48]. However, in our study, we were able to observe a lesser diversity, represented mostly by Firmicutes, with the
Leuconostoc genera predominant. López-Seijas et al., (2020) reported that
Lactobacillus is the prominent bacteria in Albariño grapes, with
L. plantarum as the dominant species [
93]. In our study, the most abundant species was
L. mesenteroides isolated at diverse time points in the alcoholic fermentation (
Table 3). The lesser diversity might indicate that the spontaneous alcoholic fermentation negatively influences the LAB consortia. This suggests that LAB make-up in the grapes studied shows little tolerance to the conditions reached as ethanol is produced, sugars are depleted, and yeast populations increase. Mendoza, Manc de Nadra, and Farías (2010) reported that, besides the typical conditions achieved during alcoholic fermentation, inhibition might be attributed to the ability of
S. cerevisiae to produce peptides with inhibitory effects [
94]. However, the fact that we were not able to isolate more LAB once the alcoholic fermentation ended, does not indicates that they are not present. According to Capozzi et al., (2021), the concentration of LAB increases 10 to 15 days after the alcoholic fermentation is completed [
95]. In our study, we did not consider keeping the fermented grape juice for a MLF fermentation, at which point dormant bacteria might be able to establish. Furthermore, for the culture-dependent method, the medium used to isolate the LAB did not have any enrichment, and the isolation was based in a morphological similarity among the observed colonies, which might also affect the diversity of the isolated species. Thus, more studies are necessary to complete the diversity of LAB during the spontaneous alcoholic and malolactic fermentations of these Chilean grape cultivars.
An important trait of LAB is the production of organic acids such as lactic and acetic acids. Once the fermentation was stopped, lactic acid reached values ranging from 1.12 ± 0.01 to 1.18 ± 0.02 g/L, with the highest concentration associated with the S. Blanc fermentation (
Table 4); these values are in line with those previously reported for spontaneous wine fermentations [
20]. The higher acetic acid concentration was observed for the Pinot fermentation, which reached 1.17 ± 0.01 g/L, while the rest of the fermentations remained at 1.10 g/L or below. Acetic acid production is expected not only from the native yeasts, but also from the heterofermentative LAB. On the other hand, malic acid decreased in comparison with the initial values, and small amounts were detected once the alcoholic fermentation ended, with values ranging from 0.04 to 0.08 g/L. Ripe grapes might contain between 2 and 6.5 g/L of malic acid; therefore, the presence of small amounts of the acid in the fermented grape juices might be also attributed to the presence of LAB with malolactic activity capacity [
20]. Among the isolated species,
Leuconostoc mesenteroides and
Lactobacillus delbrueckii have been reported as positive for malolactic activity; therefore, their presence might explain the low malic acid values encountered. The degradation of malic acid in wines results in a decrease in acidity, which favors the overall flavor. In our study, the Carignan and Pinot fermentations resulted in a pH about 0.5 units above the other fermentations (data not shown), which correlates with the presence of these malolactic bacteria (
Table 3,
Figure 3).
3.5. Sequential Fermentations
The yeast species
H. uvarum,
L. thermotolerans,
T. delbrueckii,
C. stellata, and
M. pulcherrima have been extensively studied for their use in beer, wine, and the elaboration of other fermented foods [
4,
20,
37,
38,
41,
42,
43,
47,
84,
96,
97,
98,
99,
100,
101], and their beneficial and negative effects in reducing the alcohol content and contributing to the volatile profile and aroma have been reported previously. In contrast, to our knowledge, little information has been reported the
C. oleophila strain isolated in our study. Therefore, the yeast was selected for further experimentation, along with
C. stellata, as a yeast control of the same genera yeast.
Several studies have shown that sequential fermentations initiated by non-
Saccharomyces yeasts, and finished by the inoculation of
S. cerevisiae, result in wines with desirable and distinct attribute profiles, and with complete fermentations (corroborated by sugar depletion) that result in stable wines [
18,
41,
70,
82,
98]. With this in mind,
C. olephila, and
C. stellata were fermented in sequence with the commercial
S. cerevisiae in Carménère grape juice. The grape variety was selected because it is the signature grape in the Chilean wine industry. The variety disappeared from Europe during the middle of the 19th century and reappeared years later in Chile [
71]. The Chilean vineyards in which it is produced have a particular climate and
terroir conditions that enable the production of the variety. Moreover, there is an increasing interest of the regional wine industry to give additional attributes to these signature wines, and the use of non-conventional yeast might aid in this purpose.
Figure 5 shows the growth curves obtained for the different isolates and the
S. cerevisiae strain. As expected, after inoculation, the non-
Saccharomyces isolates increased in population, reaching a peak after 2 days; thereafter, the yeast populations decreased significantly to finish at day 7 of fermentation with values below 5.5 log CFU/mL for all the strains (
Figure 4). The performance observed for the
Candida spp. studied here was similar to those reported when native yeasts from spontaneous Pinot Noir, Coralin, Chardonnay, Resi, Petit Arvine, Etmitage, and Gudel fermentations were studied [
20]. In contrast,
S. cerevisiae, inoculated at day 3, increased its population for 3 days straight and maintained concentrations above 7 log CFU/mL until the end of the fermentation.
The sequential fermentations resulted in lower ethanol concentrations compared to the control fermentation (pure
S. cerevisiae inoculation) (
Table 6). The average ethanol concentrations were 11.8% and 12.5% for
C. stellata and
C. oleophila, respectively (
Table 6). Before the
S. cerevisiae inoculation, ethanol content in the
C. stellata fermentations reached 7.70 ± 0.05% (
v/
v), and 4.55 ± 0.08% (
v/
v) for
C. olephila, which correlates with the low ethanol-producing capacity showed by these strains in the defined medium (
Table 5). Decreases in the non-
Saccharomyces yeast populations were observed once the
S. cerevisae inoculum increased in population. These is expected since
S. cerevisiae is highly effective in ethanol production, and therefore, the ethanol concentrations in the fermented grape juice increase. Domizzo et al., (2007) reported that among the non-
Saccahromyces yeasts characterized in their study, the
Candida spp. showed the greatest population reduction once
S. cerevisiae was inoculated [
102].
The addition of the
S. cerevisiae allowed the wine to complete the sugar utilization, and therefore increased the ethanol content, but concentrations were below 14% (
v/
v). Since Chilean wines are characterized by their strong ethanol content with 14% (
v/
v) as the average concentration, the use of the selected strains in sequential fermentation with
S. cerevisiae might be an alternative for the production of reduced alcohol wines. Reduction of alcohol is of interest given that high concentrations of ethanol mask some wine attributes [
70]. In addition, economical and health trends have influenced this market, and nowadays there a few vineyards that produce lower alcohol content wines.
Acetic acid production was similar for
C. stellata and
C. oleophila, with concentrations of 0.52 and 0.53 g/L, respectively (
Table 7). The ability to produce acetic acid from sugar consumption has been reported before for
Candida spp.; however, the values reported here are higher to those reported previously. Díaz et al., (2013) reported acetic acid concentrations ranging from 0.024 to 0.27 g/L. Although the acetic acid production might be considered a sign of microbial spoilage,
C. stellata have not been reported as spoilage-causing agents. Moreover, the presence of the yeast that are able to persist towards the end of the alcoholic fermentations is associated with the enhancement of the sensory profile of the wine [
103]. Compared to the control fermentations the sequential procedure resulted in higher acetic, malic, and succinic acid concentrations, which is consistent with what was previously reported in the literature [
38,
41,
46,
104].
Succinic acid concentrations (
Table 7) were detected in the range of 1.99 to 2.39 g/L. Although the concentrations fall above the usual range in wine (0.5 to 1.5 g/L), for some red wine varieties, concentrations higher that 3 g/L have been reported without exerting a negative effect in the wine flavor [
41,
70,
105].
Malic acid production was strain specific with 0.08 ± 0.13, 0.06 ± 0.05, and 0.05 ± 0.05 g/L for
C. stellata and
C. oleophila, respectively, and all values were lower than the control fermentation (
Table 7). These values are much lower than those reported for other non-
Saccharomyces cultivated in limited aeration systems [
70,
104]. Some non-
Saccharomyces yeasts have been reported as able to control the acidity of wines by metabolizing organic acids [
106,
107]. However, to our knowledge, no information for malic acid utilization or production for
C. stellata and
C. oleophila in wines has been reported before. However, the values reported here might indicate that the use of the yeast strains could result in less acidic wines, and even more, might help to reduce or eliminate the malolactic fermentation needed for the wines’ stability. However, more studies are necessary to confirm this.
In the sequential fermentations
C. oleophila produced the highest glycerol content (
Table 7), approximately two times higher than the other two sequential fermentations. Glycerol is a non-volatile compound that is formed by sugar consumption during the yeast metabolism. The compound contributes to wine with sweetness and fullness by increasing the density and viscosity of the wine. The glycerol values achieved in the sequential fermentations are higher to those reported for other non-
Saccharomyces yeasts in single, sequential, and co-inoculated fermentations [
20]. The redirecting of ethanol production to other metabolites, including glycerol, has been reported as an approach to reduce the ethanol concentration. The selected yeast strains used in this study showed ethanol yields of 0.18 and 0.40 for
C. oleophila and
C. stellata, respectively (
Table 5), and therefore, the increased glycerol concentration might be attributed to the yeast´s ability to canalize sugar consumption into glycerol production.
C. stellata fermentative potential has been studied not only in wines, but also in beer and vinegar production [
21,
26,
108]. The fructophilic yeast metabolism is characterized by the production of glycerol instead of ethanol, and thus produces wines with reduced alcohol content [
20]. On the other hand, to our knowledge, the use of
C. oleophila for wine production has not been previously reported. Our results suggests that the yeasts have a potential for the production of reduced alcohol wines in sequential fermentations; however, more studies are necessary to determine the effect on aroma and sensory characteristics of the resulting wines.