The Effects of Flocculant Yeast or Spontaneous Fermentation Strategies Supplemented with an Organic Nitrogen-Rich Additive on the Volatilome and Organoleptic Profile of Wines from a Neutral Grape Variety

Abstract

The effects of spontaneous fermentation and the inoculation of grape must with a flocculant yeast starter culture, together with the supplementation of must with a commercial organic nitrogen compound (ONC), were analyzed. The microbiome during fermentation was tracked, and volatile compounds in the resulting wines were identified and quantified using gas chromatography and mass spectrometry (GC-MS). Volatile compound concentrations were then subjected to statistical analysis. No significant differences in pH, titratable and volatile acidity, and ethanol and lactic acid were observed among the four wines analyzed. However, the musts supplemented with the ONC slightly increased the fermentation rate of the flocculant yeast, and, also, this additive reduced the volume of lees in the spontaneous fermentation and flocculant yeast by 1.2% and 0.6%, respectively. The concentrations of 11 major and 28 minor volatiles were significantly influenced (p-value ≤ 0.05) by the inoculation strategy, while 8 major and 28 minor volatiles were affected by ONC supplementation. This supplementation significantly decreased the Odor Activity Values and, consequently, the values of the odorant series established in wines from spontaneous fermentation. On the contrary, those from flocculant yeast showed a significant increase in all the odorant series except for the waxy series, leading to a more balanced aroma profile. Additionally, lower scores were recorded for the green, creamy, citrus, chemical, and honey series compared to wines from spontaneous fermentation. Overall, the commercial ONC extract contributed to a content increase in volatiles that provided desirable aromatic notes to the wines made with flocculant yeast, although the organoleptic evaluation showed no significant statistical differences in the attributes evaluated at the 95% confidence level.

1. Introduction

Global warming is causing significant changes in climatic conditions, such as rising average temperatures during the grape ripening period, changes in precipitation patterns, and more frequent extreme weather events (e.g., heatwaves, hailstorms, and frosts). These changes cause a lag in the phenology of the vine, leading to a higher accumulation of sugars, a decrease in acidity, and a delay in the accumulation of compounds responsible for the color and odor of grape berries. As a result, wines may have higher ethanol content and reduced acidity, which can contribute to potential microbiological issues or color instability, affecting the composition and quality of grape musts and wines. To mitigate these effects, several tools have been developed, such as the use of non-Saccharomyces yeasts, sequential inoculation, immobilized yeast systems or selected yeast strains in active dry yeast (ADY) format, and nutrient preparations [1,2,3,4]. The introduction of new strains with improved characteristics (e.g., fast fermentation kinetics, high SO₂ tolerance, glutathione production, or aromatic profile enhancement) represents an incremental improvement that contributes to an overall enhancement in wine production. These strains have been selected for their ability to withstand the effects of climate change on grape composition, as well as their ability to produce wines with complex sensory profiles [5]. These sensory properties are mainly the result of yeast metabolism during must fermentation, where sugars and nitrogen are essential for the production of ethanol, carbon dioxide, glycerol, and volatile compounds, which are responsible for most of the wine aromas. The use of ADYs, selected for their stress tolerance or aroma production, is currently a source of innovation in wineries, aiming to avoid or prevent stuck fermentations or to create new wines with different sensory characteristics [5], adapted to market trends.

Yeast-available nitrogen (YAN) in grape must primarily consists of naturally occurring ammonium ions and amino acids, both of which are essential for yeast metabolism [6]. To prevent slow or stuck fermentations, it is often necessary to supplement the must with authorized nitrogen sources. Furthermore, both the source and the concentration of YAN have a direct effect on the diversity and concentration of volatile organic compounds (VOCs), via the keto-acid pool and the Ehrlich pathway. This pathway plays a key role in the formation of aroma compounds in wine by converting amino acids to higher alcohols and their corresponding esters or aldehydes. During fermentation, yeast metabolizes amino acids through transamination and decarboxylation to produce fusel alcohols such as isoamyl alcohol, isobutanol, and phenylethanol. In addition, these alcohols can react with acids to form esters, which impart fruity and floral aromas to the wine [6,7]. Finally, the balance of compounds derived from the Ehrlich pathway has an important influence on the complexity and sensory profile of the wine. These innovations aim to solve some longstanding winemaking problems, such as improving the reproducibility and speed of fermentation by achieving higher ethanol yields and faster completion, while producing wines with personalized flavors according to consumer demand. Inorganic sources of nitrogen, such as diammonium phosphate (DAP), are widely used by winemakers as an additive to achieve adequate levels of YAN, but several authors mention [6,8,9] that their use can lead to the formation of acetic acid, which gives the wine an unpleasant vinegary aroma and sour taste. Alternatively, the effect of adding amino acids to the must has been studied on a laboratory scale, resulting in a decrease in this acid and an improvement of other volatile compounds such as keto acids, their corresponding aldehydes, and higher alcohols and their esters [10,11,12,13,14,15,16,17]; however, the use of pure amino acids in winemaking processes is not included in the European Union’s list of authorized additives.

It is well known that yeast plays a central role in wine fermentation and its aroma [18]. Its growth and efficiency depend on the level of nutrients (fermentable sugars, nitrogen compounds, vitamins, and minerals) in the medium. According to Ugliano et al. [16], the management of available nitrogen in wine fermentation is an effective tool for modulating aromatic composition and wine style, although different yeasts respond in different ways to nitrogen supplementation, in some cases leading to undesirable sensory effects. In addition, recent studies also highlight the possibility of refining the organoleptic profile of a wine by targeting the ideal combination of fermentation temperature with initial and added nitrogen concentrations [19]. The ideal yeast strain for industrial wine fermentations should have good flocculation properties at the end of fermentation, but unfortunately, flocculant wine yeasts are quite rare. Yeast cells with this property adhere to each other and form large aggregates or “flocs” that settle to the bottom of the vessel at the end of the fermentation [17,20,21,22,23]. This natural process provides protective advantages, such as yeast resistance to environmental stress and the creation of a nutrient-rich microenvironment through selective lysis [21]. Strong and complete flocculation at the appropriate stage of the winemaking process is desirable on an industrial scale, as premature flocculation can lead to sluggish fermentations and off-flavors. Ideally, fermenting yeast should not flocculate until the medium is fully fermented. Unfortunately, many industrial yeast strains fail to meet these criteria [24]. The integration of organic nitrogen compounds (ONCs) derived from yeast lee extracts as a winemaking additive is a relatively recent but promising development. The application of these extracts, when combined with selected yeast strains, has the potential to redefine fermentation processes by enhancing yeast performance and aroma compound production. This synergistic approach will allow winemakers to fine-tune fermentation dynamics to address challenges such as nitrogen starvation and the accumulation of toxic metabolites, while maximizing the sensory quality of the final product. Thus, the combined use of yeast-based extracts and the appropriate yeast strain is revealed as a significant and innovative step towards a new concept of oenology.

Despite the increasing availability of yeast-lee-based additives on the market that promise to improve wine quality, scientific information on their effects on volatile metabolite contents and organoleptic properties remains limited, especially when flocculant yeasts are used for fermentation processes. This research aims to contribute to the innovation in winemaking technology by studying the effect of a natural ONC, based on yeast lee extracts, on the volatilome of wines from a non-aromatic Pedro Ximénez white grape variety, fermented by using a flocculant active dry yeast as inoculum or the wild yeasts from the grape must itself.

2. Materials and Methods

2.1. Grape Must and Winemaking Conditions

White grapes of the Pedro Ximénez variety grown in Montilla-Moriles, a warm region of Andalusia, Spain, were harvested in the year 2022, when they reached 23 ± 0.2 °Bx (about 226 ± 1.5 g L⁻¹ sugar content); they were then crushed and pressed with an industrial pneumatic press at operating conditions of 1.8 bar. A volume of 25 L of the must obtained, with a density of 1095 g L⁻¹, equivalent to 1095 kg m³ (approximately 225.1 g L⁻¹ sugars, corresponding to 13.4% v/v potential ethanol) [25], 140 mg L⁻¹ yeast assimilable nitrogen (YAN), 1.07 ± 0.05 g L⁻¹ malic acid, and a pH of 3.5, was used in this study. These variables were analyzed according to the methods described in Section 2.3. This grape must was subjected to a prefermentative treatment by adding 1.5 g L⁻¹ tartaric acid and potassium metabisulfite to obtain a pH of 3.2, 4.0 g L⁻¹ total acidity and 50 mg L⁻¹ total SO₂. Eight Pyrex flasks (2 L each) were filled with 1.75 L each to test four different fermentation approaches run in duplicate. The first (WY) was based on spontaneous fermentation with wild yeasts. The second (WY + Act) included 300 mg L⁻¹ of the activator Fermoplus^® Floral (AEB S.p.A., Brescia, Italy), which increased YAN to 182 mg L⁻¹. The third (AGGLO) used Levulia^® AGGLO (AEB S.p.A., Brescia, Italy), a natural floc-forming Saccharomyces cerevisiae r.f. oviformis yeast, adding a starter culture of 1 × 10⁶ CFUs mL⁻¹ according to the manufacturer’s specifications. The fourth (AGGLO + Act) combined AGGLO with the addition of the same dose of Fermoplus^® Floral. Fermentations were conducted at 20 °C and completed at a density of less than 995 g L⁻¹. Once fermentation was completed, the wines were subjected to spontaneous decantation and stabilization at −2 °C for 20 days prior to analysis.

The additive tested is an authorized nutrient blend recommended for slow or arrested fermentations and is composed of yeast cell walls, yeast autolysates, and L-ascorbic acid. It is rich in organic nitrogen compounds (ONCs) where proline, hydroxyproline, tryptophan, methionine, glutamine, serine, asparagine, alanine, and arginine are the dominant amino acids. In addition, the yeast used was rehydrated according to the manufacturer’s instructions and pre-cultured for 24 h in 250 mL of a synthetic medium containing 50 g L⁻¹ glucose, buffered at pH 3.2 by adding 2.8 tartaric acid and potassium bitartrate 2.4 g L⁻¹ and 200 mg L⁻¹ diammonium phosphate (DAP).

2.2. Microbiological Analysis

To determine the number of viable cells and their evolution during fermentation, samples from the different conditions were streaked at four sampling times: before fermentation began (D0), on day 4 of fermentation (D4), on day 7 (D7), and on day 17 (D17), once fermentation was completed. Serial dilutions were performed to obtain the number of colony-forming units (CFUs) per mL of sample on WLN agar medium (50 g L⁻¹ dextrose, 4 g L⁻¹ yeast extract, 5 g L⁻¹ tryptone, and 0.022 g L⁻¹ bromocresol green; OXOID CM 0501) as a differential culture medium. In order to allow for the growth of the microorganisms at an appropriate temperature, the culture media plates were incubated for 72 h at a temperature of 28 °C. The isolated colonies in the different media were visually characterized, establishing morphological groups. Axenic cultures were later established from the different colonies, and a lysine-containing selective medium was used to distinguish Saccharomyces from non-Saccharomyces yeasts. Isolates were cultured on the lysine agar medium (44.5 g L⁻¹ glucose, 17.8 g L⁻¹ agar, 1 g L⁻¹ lysine, and trace elements in concentrations lower than 1 g L⁻¹; OXOID CM 0191B) and incubated under the same conditions. All analyses were conducted in triplicate.

2.3. Chemical Analysis

Following the protocols of the International Organisation of Vine and Wine (OIV) [25], the oenological variables pH, ethanol, volatile and titratable acidity, reducing sugars, and yeast assimilable nitrogen (YAN) were quantified. Reflectoquant™ (Merck^®, Darmstadt, Germany) was used to measure malic and lactic acids by the reflectometric method. Finally, using an Agilent Cary 60 UV-Vis spectrophotometer (Agilent Technologies, Santa Clara, CA, USA), the absorbances at 280, 420, and 520 nm were measured. All samples were analyzed in triplicate.

2.4. Analysis of Wine Volatiles

This study quantified major volatile compounds (those with levels ≥ 10 mg L⁻¹) and polyols using a gas chromatograph Agilent 6890 (Agilent Technologies, Santa Clara, CA, USA) equipped with a flame ionization detector (FID) and a CP-WAX 57 CB capillary column (60 m; 0.25 mm; 0.4 µm film). Sample preparation involved adding 1 mL of an internal standard solution (4-methyl-2-pentanol, 1018 mg L⁻¹) and 0.2 g of calcium carbonate to 10 mL of the wine sample, followed by ultrasonication, centrifugation, and GC injection of 0.7 µL to the supernatant. Compounds such as methanol and higher alcohols, acetaldehyde, acetoin, ethyl esters, glycerol, and 2,3-butanediol were measured using a calibration table constructed with standard solutions of known concentrations.

Minor volatiles (levels < 1 mg L⁻¹) were analyzed using stir bar sorptive extraction–thermal desorption–gas chromatography–mass spectrometry (SBSE-TD-GC-MS). A coated PDMS stir bar extracted volatiles from a prepared wine solution buffered to pH = 3.2, which were then desorbed and analyzed in a GC-MS system. The oven temperature profile and mass spectrometer settings were optimized for compound identification. All analyses were performed in triplicate, with quantification based on calibration tables of standard solutions of known concentrations. Peaks were quantified using Chemstation software (ChemStation v. 02. 02. 1431, ChemStation International, Inc., Dayton, OH, USA) algorithms that process the chromatographic signal obtained for each compound and calculate its area. The relative area, calculated by dividing this area by the area of the internal standard, is then used in the appropriate calibration equation to calculate the absolute concentration of each compound.

The contribution of volatile compounds to the aroma properties of wine is influenced by their concentration and odor perception threshold (OPT). The Odor Activity Value (OAV), calculated as the ratio between the concentration of a compound and its OPT, is commonly used to assess the impact of a compound on the overall aroma. A more detailed description of the analytical methods for volatiles has been described in a recent paper [26].

2.5. Organoleptic Analysis

The four wines were evaluated using the official OIV tasting sheet [25] by a panel of six males and four females, ages ranging from 24 to 68 years old, from the Department of Agricultural Chemistry, Soil Science, and Microbiology at the University of Córdoba, Spain. All have formal training in oenology, and two are members of a professional tasting panel. The participants were fully informed about the study’s objectives and procedures. Participation was voluntary, and only aggregated, anonymized data were used in the analysis, ensuring compliance with data protection regulations. The panel assessed visual, olfactory, and gustatory attributes, including clarity, color, aroma intensity and quality, and flavor persistence and quality. Wines were rated on a 100-point scale for overall quality. Before analysis, the wines were stored at 4 °C for 24 h, and 30 mL of each was served at room temperature (20 °C) and natural lighting in standardized wine glasses (NF V09-110 AFNOR, 1995) following ISO 3591 standards [25,27]. The glasses were labeled with blind-codes and presented in a randomized order with a one-minute interval between wine tastings.

2.6. Statistical Analysis

The data matrices were pre-processed using two normalization methods: mean centering and autoscaling. Statistical analysis was performed using the Statgraphics software package (Centurion XVI v. 16.1.11, StatPoint Technologies, Inc., Warrenton, VA, USA) and a two-way analysis of variance (ANOVA) with Fisher’s least significant difference (LSD) test. This procedure assessed the dependence of each compound on yeast, activator, and their interaction at a significance level of p < 0.05. The contents of major volatiles and polyols were analyzed by Principal Component Analysis (PCA) using PLS_Toolbox v. 8.5.2 software in MATLAB R2016a v. 9.0.0.341360 (Natick, MA, USA). Minor volatiles were also subjected to the same statistical treatments.

3. Results and Discussion

3.1. Fermentation Kinetics and Microbiota Dynamics During Sampling Stages

The yeast strain Levulia Agglo is a S. cerevisiae r.f oviformis with a neutral killer phenotype (it does not produce any killer toxins itself and is resistant to the killer toxins produced by others) [28]. The use of this strain is recommended for industrial second fermentations or prise de mousse in sparkling wine production, because its agglomerating property reduces the time of the remuage phase from 4–6 weeks to 1–2 weeks compared to standard strains [29]. This tendency of the cells to cohere and adhere to each other facilitates their sedimentation and the wine clarification process. However, yeast strains with this property are not preferred by winemakers in still wine production because they settle quickly, risking incomplete fermentation, off-flavors, and microbial instability. This yeast has other desirable characteristics for industrial winemaking, suitable for aging on the lees due to its easy autolysis, predictable kinetic and complete fermentation at moderate temperatures (13–20 °C), and medium–high nitrogen demand, with an ethanol tolerance of 13.8% v/v, according to the manufacturer’s specifications.

Figure 1 shows the fermentation kinetics over 17 days for the four conditions studied. The density values decreased from 1095 to 995 g L⁻¹ after 17 days in all fermentations, showing similar trends between them, with a gradual decrease in density starting around day 6 in all conditions and converging towards approximately 995 g L⁻¹ by day 16. However, the fermentation kinetics of WY were faster than those of AGGLO, which could be explained by the slower growth rate of flocculent yeasts compared to other yeast strains [30]. Cells within the floc have a smaller surface area, which limits nutrient diffusion and slows growth. In contrast, the AGGLO + Act condition experienced a faster fermentation process between days 6 and 12. The WY + Act condition did not show this trend and instead showed slower fermentation kinetics compared to the other conditions. The faster fermentation kinetics in AGGLO + Act suggest a synergistic effect between the AGGLO yeast and the tested additive, increasing fermentation efficiency. By day 17 of fermentation, all conditions reached similar density levels with only slight differences in the timing and rate of density decline. After this fermentation period, the wines were allowed to settle at 10 °C for 10 days, and the volume of the sediment at the bottom of the glass cylinders was measured. The volume of lees measured for the AGGLO wines was 7.7% of the initial wine volume, and for the AGGLO + Act wines it was lower (7.1%). For WY and WY + Act wines, the volume of lees was 8.3% and 7.1% of the total wine volume, respectively. Therefore, the tested activator caused a reduction of 0.6 and 1.2% in the volume of lees, which reduces the loss of wine volume from the cellar.

The number of viable cells was monitored in all conditions at three different stages of the fermentation process. The microbiological analysis of the uninoculated grape must (D0) revealed a low viable cell count, with a density of 10³ CFU mL⁻¹. By day 4, a significant increase in viable cells was observed, especially in the WY + Act condition, reaching 200 × 10⁶ CFU mL⁻¹, compared to the other fermentations. On day 7, viable cell counts continued to rise in most conditions, with AGGLO + Act exhibiting the highest count at 188 × 10⁶ CFU mL⁻¹. By day 17, the viable cell count had decreased in all conditions except for WY. These results suggest that the ONC-rich additive significantly influences yeast growth, particularly in the early stages of fermentation. Commercially available preparations based on inactive dry yeast preparations contain a soluble fraction consisting of yeast cytoplasmic metabolites and lysed cell walls (proteins, peptides, amino acids and polysaccharides—glucans, mannoproteins, sterols and fatty acids) and an insoluble fraction consisting of an inactive inert carrier, mainly cellulose. This leads to an improvement with a higher number of viable yeasts compared to fermentations carried out without these preparations [31]. In addition, an increase in sterol content within the yeast promotes more active fermentation by increasing membrane permeability, which allows for a greater exchange of substances between the cell and the medium. Furthermore, sterols may act as a survival factor, increasing the reserves that the yeast can use during the decline phase [31]. Analysis of the viable cell cultures revealed six types of CFUs, which were visually classified according to their morphological characteristics such as color, size, and shape. Although this classification does not identify the yeast species present in the different wines, the morphological differences likely reflect the diversity of yeast strains or microbial populations present during the fermentation process. Among these, only one of the six colony types was able to grow on lysine agar, indicating the presence of non-Saccharomyces yeasts. This colony type was only observed in the grape must before fermentation (D0), while the viable cells detected during the remaining days of the fermentation were identified as Saccharomyces yeasts.

3.2. Oenological Parameters

Table 1. Oenological variables analyzed. Mean values and standard deviations. Letters (^{a, b, c}) in the same row indicate statistical differences at the p ≤ 0.05 level according to Fisher’s least significant difference test, represented in the table as homogeneous groups (HGs). Two-way ANOVA: ns = non-significant; *** indicates p ≤ 0.001 (confidence level ≥ 99.9%); ** indicates p ≤ 0.01 (confidence level ≥ 99%); * indicates p ≤ 0.05 (confidence level ≥ 95%). Y: yeast effect; A: activator effect; Y × A: interaction between yeast and activator. WY: wine obtained by spontaneous fermentation; WY + Act: wine obtained by spontaneous fermentation with an activator; AGGLO: wine obtained by fermentation with Levulia^® AGGLO; AGGLO + Act: fermentation with Levulia^® AGGLO supplemented with an activator.

	WY	WY + Act	AGGLO	AGGLO + Act	HG	Y	A	$\mathbf{Y}\times \mathbf{A}$
pH	3.22 ± 0.06 a	3.3 ± 0.1 a	3.22 ± 0.06 a	3.26 ± 0.02 a	1	ns	ns	ns
Titratable acidity (g L−1)	6.1 ± 0.9 a	6 ± 1 a	6 ± 0.3 a	5.8 ± 0.2 a	1	ns	ns	ns
Volatile acidity (g L−1)	0.4 ± 0.1 a	0.45 ± 0.00 a	0.48 ± 0.08 a	0.44 ± 0.06 a	1	ns	ns	ns
Ethanol (% v/v)	13.7 ± 0.1 a	13.7 ± 0.1 a	13.65 ± 0.05 a	13.7 ± 0.2 a	1	ns	ns	ns
Reducing sugars (g L−1)	3.1 ± 0.5 c	1.4 ± 0.2 b	0.9 ± 0.3 a	0.6 ± 0.4 ab	3	***	**	*
Malic acid (mg L−1)	653 ± 5 a	647 ± 30 a	891 ± 10 b	907 ± 7 b	2	***	ns	ns
Lactic acid (mg L−1)	≤0.15	≤0.15	≤0.15	≤0.15	1	ns	ns	ns
Absorbance 420 nm	0.247 ± 0.006 b	0.27 ± 0.01 c	0.245 ± 0.005 b	0.227 ± 0.006 a	3	***	ns	***
Absorbance 520 nm	0.055 ± 0.005 a	0.0646 ± 0.0005 b	0.054 ± 0.001 a	0.051 ± 0.001 a	2	**	ns	**
Absorbance 280 nm (TPI)	9.71 ± 0.05 c	8.22 ± 0.03 a	9.7 ± 0.2 c	9.10 ± 0.05 b	3	***	***	***
SO2 free (mg L−1)	1.4 ± 0.8 a	1.5 ± 0.9 ab	4 ± 2 b	3 ± 1 ab	2	*	ns	ns
SO2 total (mg L−1)	40 ± 1 b	45.1 ± 3 c	29.2 ± 0.3 a	30.0 ± 0.2 a	3	***	**	*
Yeast assimilable nitrogen (YAN)	35 ± 7 b	42 ± 7 c	28 ± 1 a	42 ± 1 c	3	ns	***	ns

shows the values of the parameters quantified in all the wines obtained. Only the absorbances at 280 nm and 420 nm, total SO₂, residual sugars, and YAN showed three homogeneous groups at the p ≤ 0.05 significance level. In addition, this last parameter was the only one dependent on the additive tested at the p ≤ 0.01 level, according to the ANOVA performed. The values of pH, titratable acid, volatile acid, ethanol, and lactic acid did not show statistically significant differences among all four wines, which had a pH of 3.25, a titratable and volatile acidity of 6 and 0.45 g L⁻¹ respectively, and an ethanol content of 13.7% (v/v).

The malic acid content of the wines showed two HGs, the differences being due to the action of the yeast. Fermentations carried out with AGGLO and AGGLO + Act resulted in higher contents of this acid, reaching contents 260 mg L⁻¹ higher than those of WY wines. The malic acid content in white wines is an interesting feature in order to control and preserve their acidity, especially in the context of climate change [32]. However, its excessive accumulation can lead to an overly sour taste, especially in red wines, so a reduction in its content through malolactic fermentation is always recommended, which helps to produce softer, less aggressive red wines from a sensory point of view [33]. The absorbance values at 420 and 520 nm were significantly dependent only on the yeast, while the absorbance at 280 nm (TPI) decreased significantly in the wines obtained with the addition of the ONC-rich additive tested. Thus, wines obtained with the use of additives containing mannoproteins may have a lower color intensity than those obtained without the use of additives containing mannoproteins [31]. One of the proposed reasons is an increase in polyphenol precipitation due to the formation of additional mannoprotein–polyphenol colloidal complexes [34,35]. It has also been shown that yeast cells and yeast walls can adsorb browning and phenolic compounds in white wines [31]. Furthermore, the ascorbic acid content of the additive studied also contributes to this effect due to its antioxidant activity. This results in a lower tendency of wines to browning compared to those not treated with these additives [35]. By contrast, total SO₂ was affected by both factors studied. Thus, AGGLO wines show lower levels of this compound than WY. This decrease is of great interest because the need to diminish its use has become increasingly relevant due to growing concerns of consumer overexposure to sulfites and the health issues this raises [36]. In this context, the low amount of SO₂ added in our experiments does not compromise the normal evolution of yeast development, in agreement with the results of other authors [37]. YAN is influenced only by the tested additive and shows a significant difference with a p-value ≤ 0.001. In both WY + Act and AGGLO + Act, the final YAN levels are the highest. This is consistent with the results obtained in the fermentation kinetics of AGGLO yeast, which shows the low levels of YAN, which are associated with sluggish or stuck fermentations [31]. This behavior is not observed in spontaneous fermentation and highlights the synergistic effect between the AGGLO yeast and the ONC-rich additive.

3.3. Major Volatile Compounds and Polyols

The analysis of the 11 major volatile compounds and three polyols (Table 2) reveals significant differences for acetaldehyde, 2-methyl-1-butanol, acetoin, 2,3-butanediol (levo and meso forms), 2-phenylethanol, and glycerol, considering the two variation factors in the ANOVA analysis identified in two or three HGs at the p ≤ 0.05 level. For 2,3-butanediol (levo meso), 2-phenylethanol, and glycerol, the concentration increases significantly in the two AGGLO wines. However, the profile of these compounds for (AGGLO + Act) tends to resemble that of wines from WY. No significant differences were found for methanol and ethyl lactate, both of which showed one HG, in agreement with other researchers [38]. Also, no differences were observed as an effect of the additive in the contents of 1-propanol, isobutanol, and 3-methyl-1-butanol alcohols, while 2-phenylethanol decreased their levels, in accordance with other studies [39].

Table 2. Major volatile compounds and polyols. Mean and deviation values quantified in wines (mg L⁻¹). Letters (^{a, b, c, d}) in the same row indicate statistical differences at the p ≤ 0.05 level according to Fisher’s least significant difference test, represented in the table as homogeneous groups (HGs). Two-way ANOVA: ns = non-significant; *** indicates p ≤ 0.001 (confidence level ≥ 99.9%); ** indicates p ≤ 0.01 (confidence level ≥ 99%); * indicates p ≤ 0.05 (confidence level ≥ 95%). Y: yeast effect; A: activator effect; Y $\times$ A: interaction between yeast and activator. WY: wine obtained by spontaneous fermentation; WY + Act: wine obtained by spontaneous fermentation with an activator; AGGLO: wine obtained by fermentation with Levulia^® AGGLO; AGGLO + Act: fermentation with Levulia^® AGGLO supplemented with an activator. CAS: identification number assigned by the Chemical Abstracts Service. OPT: odor perception threshold (mg L⁻¹). OS: odorant series.

Compounds	CAS	WY	WY + Act	AGGLO	AGGLO + Act	HG	Y	A	$\mathbf{Y}\times \mathbf{A}$	OPT	OS
Acetaldehyde	75-07-0	86 ± 5 b	66 ± 3 a	65 ± 2 a	66 ± 2 a	2	***	**	***	110 I	1, 6
Ethyl acetate	141-78-6	28.5 ± 0.1 a	31.2 ± 0.3 b	32.5 ± 0.3 c	30.5 ± 0.8 b	3	***	ns	***	7.5 I	6
Methanol	67-56-1	84 ± 6 a	84 ± 6 a	78 ± 3 a	83 ± 3 a	1	ns	ns	ns	50 I	6
1-Propanol	71-23-8	19.9 ± 0.6 b	18.5 ± 0.5 a	18.5 ± 0.7 a	20.9 ± 0.6 b	2	ns	ns	***	830 I	6
Isobutanol	78-83-1	26.8 ± 0.4 a	27.6 ± 0.4 a	39.6 ± 0.8 c	37.7 ± 0.5 b	3	***	ns	**	40 I	6
2-Methyl-1-butanol	137-32-6	38.8 ± 0.5 a	40.2 ± 0.7 b	44.7 ± 0.5 c	45.7 ± 0.5 d	4	***	**	ns	30 I	6
3-Methyl-1-butanol	123-51-3	214 ± 3 a	211 ± 5 a	233 ± 5 b	244 ± 4 c	3	***	ns	*	30 I	6
Acetoin	513-86-0	16 ± 3 a	21 ± 2 b	37 ± 4 c	19.6 ± 0.7 ab	3	***	**	***	150 I	4
Ethyl lactate	97-64-3	14.7 ± 0.2 a	14.8 ± 0.3 a	15.0 ± 0.6 a	14.5 ± 0.3 a	1	ns	ns	ns	150 I	4
2,3-Butanediol (levo)	24347-58-8	258 ± 9 a	289 ± 8 b	447 ± 19 c	268 ± 8 ab	3	***	***	***	-	-
2,3-Butanediol (meso)	5341-95-7	36 ± 2 a	40 ± 3 a	99 ± 9 b	39 ± 2 a	2	***	***	***	-	-
Diethyl succinate	123-25-7	14 ± 2 b	12.9 ± 0.6 b	13 ± 1 b	8 ± 1 a	2	**	**	*	100 I	1
2-Phenylethanol	60-12-8	27 ± 2 a	25 ± 1 a	47 ± 2 c	34.7 ± 0.6 b	3	***	***	***	10 I	9
Glycerol	56-81-5	4090 ± 259 a	4046 ± 118 a	6946 ± 133 c	4473 ± 100 b	3	***	***	***	-	-

1: Fruity; 4: Creamy; 6: Chemistry; 9: Floral. I: OPT from Dumitriu et al., 2024 [40].

Table 2. Major volatile compounds and polyols. Mean and deviation values quantified in wines (mg L⁻¹). Letters (^{a, b, c, d}) in the same row indicate statistical differences at the p ≤ 0.05 level according to Fisher’s least significant difference test, represented in the table as homogeneous groups (HGs). Two-way ANOVA: ns = non-significant; *** indicates p ≤ 0.001 (confidence level ≥ 99.9%); ** indicates p ≤ 0.01 (confidence level ≥ 99%); * indicates p ≤ 0.05 (confidence level ≥ 95%). Y: yeast effect; A: activator effect; Y $\times$ A: interaction between yeast and activator. WY: wine obtained by spontaneous fermentation; WY + Act: wine obtained by spontaneous fermentation with an activator; AGGLO: wine obtained by fermentation with Levulia^® AGGLO; AGGLO + Act: fermentation with Levulia^® AGGLO supplemented with an activator. CAS: identification number assigned by the Chemical Abstracts Service. OPT: odor perception threshold (mg L⁻¹). OS: odorant series.

Compounds	CAS	WY	WY + Act	AGGLO	AGGLO + Act	HG	Y	A	$\mathbf{Y}\times \mathbf{A}$	OPT	OS
Acetaldehyde	75-07-0	86 ± 5 b	66 ± 3 a	65 ± 2 a	66 ± 2 a	2	***	**	***	110 I	1, 6
Ethyl acetate	141-78-6	28.5 ± 0.1 a	31.2 ± 0.3 b	32.5 ± 0.3 c	30.5 ± 0.8 b	3	***	ns	***	7.5 I	6
Methanol	67-56-1	84 ± 6 a	84 ± 6 a	78 ± 3 a	83 ± 3 a	1	ns	ns	ns	50 I	6
1-Propanol	71-23-8	19.9 ± 0.6 b	18.5 ± 0.5 a	18.5 ± 0.7 a	20.9 ± 0.6 b	2	ns	ns	***	830 I	6
Isobutanol	78-83-1	26.8 ± 0.4 a	27.6 ± 0.4 a	39.6 ± 0.8 c	37.7 ± 0.5 b	3	***	ns	**	40 I	6
2-Methyl-1-butanol	137-32-6	38.8 ± 0.5 a	40.2 ± 0.7 b	44.7 ± 0.5 c	45.7 ± 0.5 d	4	***	**	ns	30 I	6
3-Methyl-1-butanol	123-51-3	214 ± 3 a	211 ± 5 a	233 ± 5 b	244 ± 4 c	3	***	ns	*	30 I	6
Acetoin	513-86-0	16 ± 3 a	21 ± 2 b	37 ± 4 c	19.6 ± 0.7 ab	3	***	**	***	150 I	4
Ethyl lactate	97-64-3	14.7 ± 0.2 a	14.8 ± 0.3 a	15.0 ± 0.6 a	14.5 ± 0.3 a	1	ns	ns	ns	150 I	4
2,3-Butanediol (levo)	24347-58-8	258 ± 9 a	289 ± 8 b	447 ± 19 c	268 ± 8 ab	3	***	***	***	-	-
2,3-Butanediol (meso)	5341-95-7	36 ± 2 a	40 ± 3 a	99 ± 9 b	39 ± 2 a	2	***	***	***	-	-
Diethyl succinate	123-25-7	14 ± 2 b	12.9 ± 0.6 b	13 ± 1 b	8 ± 1 a	2	**	**	*	100 I	1
2-Phenylethanol	60-12-8	27 ± 2 a	25 ± 1 a	47 ± 2 c	34.7 ± 0.6 b	3	***	***	***	10 I	9
Glycerol	56-81-5	4090 ± 259 a	4046 ± 118 a	6946 ± 133 c	4473 ± 100 b	3	***	***	***	-	-

1: Fruity; 4: Creamy; 6: Chemistry; 9: Floral. I: OPT from Dumitriu et al., 2024 [40].

The data matrix constructed with the contents of major volatiles and polyols was subjected to a PCA, resulting in a biplot showing the distribution of the four wines obtained (Figure 2). The first two components explain 79.96% of the total variance, clustering the wines into two groups. Those from the AGGLO yeast are on the right, with higher values of PC1 than the WY wines. This indicates a clear difference in the volatile profiles associated with the different yeasts. It is worth noting that the WY + Act wines are grouped close to their counterparts not supplemented (WY), and yet important differences are obtained for AGGLO and AGGLO + Act in their PC2 values, whose separation on the biplot is higher. The wines produced by spontaneous fermentation (WY) show negative values for PC1, which is mainly influenced by compounds such as acetaldehyde and methanol, which show strong negative loadings. The WY + Act samples show some overlap with the WY samples but are slightly shifted along PC2, indicating a moderate differentiation and, consequently, a similar volatile profile based on these volatiles and polyols. The AGGLO wines have positive values for PC1, indicating that they are influenced by volatile compounds with positive loadings on PC1, such as 2,3-butanediol, 2-phenylethanol, glycerol, and acetoin. The AGGLO + Act samples are further shifted along PC2, a shift driven by 1-propanol, isoamyl alcohols, and isobutanol, which have strong positive loadings on PC2 (Table S1. Supplementary Materials).

3.4. Minor Volatile Compounds

Table 3 lists the contents of wines in 32 minor volatiles and their homogeneous groups (HGs), with a significance level of p ≤ 0.05. The results of the two-way ANOVA are also shown to determine the effect of yeast or additive and their interaction on the content of each compound. These compounds include 5 higher alcohols, 15 esters, 4 aldehydes, 1 ketone, 2 lactones, and finally 5 terpenes and nor-isoprenoids, all of which are largely influenced by the dominant yeast strain at each stage of the fermentation process [41,42]. Among these 32 quantified minor volatiles, 28 showed significant differences in their concentrations depending on the type of inoculation used for fermentation, and only hexanol, ethyl heptanoate, hexanal, and γ-nonalactone remained unaffected by this factor. In addition, 27 of the compounds were also influenced by the addition of the rich-ONC additive to the must, and only the esters hexyl acetate and ethyl octanoate, together with the aldehyde decanal, were not dependent on the interaction. As mentioned above, the yeast extract added to the must is composed of yeast cell walls and autolyzed yeast, which increase the content of the medium and long-chain fatty acids in the must. Several compounds listed in Table 3, such as aldehydes and alcohols with six carbon atoms and nonanal, are related to the oxidative degradation of fatty acids, while some carboxylic acids and γ-lactones (γ-nonalactone, γ-decalactone) are related to the degradation of fats [43]. Also, the enrichment in amino acids, provided by the added extract, plays an important role in wine quality because they are precursors of some aroma compounds through deamination or decarboxylation reactions [44]. The high content of esters such as isoamyl acetate or hexyl acetate in wines preserves their fruity character over time due to their interaction with yeast cell walls, which reduces their loss [45]. The results obtained in our study show a remarkable increase in these compounds, together with 2-phenylethyl acetate and ethyl hexadecanoate, when the yeast extract is added to the must and fermented with AGGLO yeast. This increase in ester content confirms the results obtained by other authors after the addition of different yeast extract derivatives to the must [43,45].

The contents of higher alcohols listed in Table 3 for WY wines do not show significant differences compared to those of WY + Act, except for ethyl-1-hexanol, whose content is higher in WY. On the other hand, when the fermentation is carried out with the agglomerant yeast AGGLO, the levels of these compounds increase significantly in the AGGLO + Act wines, showing the different effect of this additive on the different types of inoculation tested and also indicating a high degree of specificity with the yeast strain [46]. At this point, it is crucial to note that the concentration of higher alcohols significantly influences the taste of the wine. When their content is below 400 mg L⁻¹, they contribute to a fuller and rounder wine, enhancing its flavor. However, if their concentration exceeds this threshold, a distinctive off-flavor similar to that of organic solvents may emerge, which negatively impacts the wine’s quality [47]. The statistical analysis performed on the aldehyde, ketone, and lactone contents, shows that they are all grouped in three HGs, except for γ-nonalactone, in four HGs. This last compound and hexanal depend significantly on the addition of the supplement and its interaction with the type of inoculation used for fermentation. In contrast, the decanal content does not depend on the supplement added to the must and its interaction with the type of inoculum. The remaining compounds of this chemical family are dependent on the factors considered and their interaction. The last group of minor volatiles consists of five compounds classified as terpenes and C-13 norisoprenoids derived from grapes, where they are mainly found as non-volatile precursors. All these compounds have four HGs with the exception of geranyl acetone and ethyl dihydrojasmonate, which have three HGs. Only nerolidol is not dependent on the addition of the extract tested, while all the other compounds are significantly dependent on the two factors studied and their interaction. These results are related to a different effect of the extract added to the must depending on the yeast carrying the fermentation.

Table 3. Minor volatile compounds. Mean and deviation values quantified in wines (μg L⁻¹). Letters (^{a, b, c, d}) in the same row indicate statistical differences at the p ≤ 0.05 level according to Fisher’s least significant difference test, represented in the table as homogeneous groups (HGs). Two-way ANOVA: ns = non-significant; *** indicates p ≤ 0.001 (confidence level ≥ 99.9%); ** indicates p ≤ 0.01 (confidence level ≥ 99%); * indicates p ≤ 0.05 (confidence level ≥ 95%). Y: yeast effect; A: activator effect; Y $\times$ A: interaction between yeast and activator. WY: wine obtained by spontaneous fermentation; WY + Act: wine obtained by spontaneous fermentation with an activator; AGGLO: wine obtained by fermentation with Levulia^® AGGLO; AGGLO + Act: fermentation with Levulia^® AGGLO supplemented with an activator. CAS: identification number assigned by the Chemical Abstracts Service. OPT: odor perception threshold (μg L⁻¹). OS: odorant series. n.f.: not found.

	CAS	WY	WY + Act	AGGLO	AGGLO + Act	HG	Y	A	$\mathbf{Y}\times \mathbf{A}$	OPT	OS
Alcohols (5)
Hexanol	111-27-3	2272 ± 168 b	2814 ± 228 b	1265 ± 502 a	3973 ± 502 c	3	ns	***	***	8000 I	3
2-Ethyl-1-hexanol	104-76-7	980 ± 86 c	855 ± 45 b	369 ± 81 a	1097 ± 81 c	3	**	***	***	8000 II	5
Octanol	111-87-5	266 ± 4 b	267 ± 6 b	94 ± 33 a	374 ± 33 c	3	*	***	***	800 III	8
Decanol	112-30-1	69 ± 1 b	71 ± 8 b	24 ± 6 a	69 ± 6 b	2	***	***	***	400 IV	8
Dodecanol	112-53-8	149 ± 15 b	141 ± 10 b	63 ± 17 a	187 ± 17 c	3	*	***	***	1000 V	8
Esters (15)
Ethyl butyrate	105-54-4	61 ± 2 bc	57.4 ± 0.3 b	30 ± 5 a	64 ± 5 c	3	***	***	***	20 I	1
Butyl acetate	123-86-4	53.7 ± 0.6 c	40 ± 3 b	20.6 ± 0.8 a	42.2 ± 0.8 b	3	***	**	***	66 VII	1, 6
Ethyl 2-methylbutyrate	7452-79-1	6.6 ± 0.4 b	6.4 ± 0.9 b	3 ± 1 a	15 ± 1 c	3	***	***	***	18 VIII	1, 2
Ethyl 3-methylbutyrate	108-64-5	17 ± 1 c	13.4 ± 0.5 b	6 ± 1 a	13 ± 1 b	3	***	*	***	3 II	1, 2
Isoamyl acetate	123-92-2	165 ± 2 c	201 ± 3 d	70 ± 12 a	138 ± 12 b	4	***	***	**	30 I	1
Ethyl hexanoate	123-66-0	22 ± 0.2 a	21.7 ± 0.2 a	21 ± 2 a	32 ± 2 b	2	***	***	***	14 II	1, 2
Hexyl acetate	142-92-7	298 ± 12 b	382 ± 4 c	211 ± 10 a	300 ± 10 b	3	***	***	ns	670 IX	1, 2
Ethyl heptanoate	106-30-9	2.2 ± 0.2 b	2.01 ± 0.04 ab	1.9 ± 0.2 a	2.1 ± 0.2 ab	2	ns	ns	*	2.2 VIII	1, 2
Ethyl octanoate	106-32-1	18.5 ± 0.3 a	20 ± 0.4 ab	21 ± 1 b	21 ± 1 b	2	**	ns	ns	5 I	1, 8
2-Phenylethyl acetate	103-45-7	1094 ± 48 b	1094 ± 40 b	902 ± 154 a	1877 ± 154 c	3	***	***	***	250 I	7, 9
Ethyl decanoate	110-38-3	48 ± 2 a	50 ± 2 a	136 ± 8 c	99 ± 8 b	3	***	***	***	200 I	1, 8
Phenethyl hexanoate	101-60-0	0.5 ± 0.07 a	0.54 ± 0.05 a	5.7 ± 0.2 c	1.8 ± 0.2 b	3	***	***	***	250 I	2, 8, 9
Ethyl tetradecanoate	124-06-1	66 ± 3 c	57 ± 3 b	48 ± 6 a	108 ± 6 d	4	***	***	***	4000 I	8
Phenethyl benzoate	94-47-3	13 ± 1 c	9.6 ± 0.3 b	3.7 ± 0.3 a	8.6 ± 0.3 b	3	***	*	***	n.f.	9
Ethyl hexadecanoate	628-97-7	145 ± 16 a	182 ± 13 b	204 ± 6 b	473 ± 6 c	3	***	***	***	2000 I	8
Aldehydes (4)
Hexanal	66-25-1	7.6 ± 0.1 b	6.8 ± 0.8 b	3.9 ± 0.3 a	10.1 ± 0.3 c	3	ns	***	***	5 X	3
Nonanal	124-19-6	5.7 ± 0.5 c	5.4 ± 0.2 c	1.7 ± 0.3 a	4.2 ± 0.3 b	3	***	***	***	2.5 II	5
Decanal	112-31-2	5.9 ± 0.6 c	5.3 ± 0.4 bc	4.7 ± 0.4 ab	4.2 ± 0.4 a	3	**	ns	ns	1.25 II	5, 8
2-Phenyl-acetaldehyde	122-78-1	129 ± 9 c	81 ± 3 ab	69 ± 9 a	86 ± 9 b	3	***	**	***	4 XI	3, 7
Ketones (1)
Acetophenone	98-86-2	n.d. a	n.d. a	14 ± 4 b	56 ± 4 c	3	***	***	***	65 XII	9
Lactones (2)
γ-Nonalactone	104-61-0	69 ± 6 c	50 ± 3 b	33 ± 8 a	84 ± 8 d	4	ns	**	***	30 I	4
γ-Decalactone	706-14-9	44 ± 2 c	26 ± 5 b	11 ± 4 a	40 ± 4 c	3	**	*	***	41 I	1, 4
Terpenes and Norisoprenoids (5)
Limonene	5989-27-5	420 ± 48 d	333 ± 11 c	31 ± 20 a	225 ± 20 b	4	***	**	***	10 II	4, 6
Nerolidol	7212-44-4	11 ± 1 b	0.8 ± 0.1 a	25 ± 3 c	34 ± 3 d	4	***	ns	***	700 II	3, 9
Z-Geranyl acetone	689-67-8	13 ± 1 c	11.7 ± 0.6 c	3.1 ± 0.2 a	6.9 ± 0.2 b	3	***	*	***	60 II	9
Farnesol	4602-84-0	52 ± 5 b	28 ± 2 a	126 ± 9 c	180 ± 9 d	4	***	**	***	20 VI	9
E-Methyldihydro-jasmonate	24851-98-7	36 ± 3 c	31 ± 2 b	11 ± 2 a	37 ± 2 c	3	***	***	***	70 II	9

1: fruity; 2: green fruit; 3: green; 4: creamy; 5: citrus; 6: chemistry; 7: honey; 8: waxy; 9: floral. I: (Dumitriu et al., 2024) [40]; II: (Palenzuela et al., 2023) [26]; III: (Peinado et al., 2006) [48]; IV: (Moreno et al., 2005) [49]; V: (Martín-García et al., 2023) [50]; VI: (Zhu et al., 2019) [51]; VII: (Peinado-Pardo et al., 2016) [52]; VIII: (López de Lerma et al., 2018) [53]; IX: (Gambetta et al., 2014) [54]; X: (Buttery et al., 1988) [55]; XI: (Buttery et al. 1971) [56]; XII: (Leffingwell.com) [57].

Table 3. Minor volatile compounds. Mean and deviation values quantified in wines (μg L⁻¹). Letters (^{a, b, c, d}) in the same row indicate statistical differences at the p ≤ 0.05 level according to Fisher’s least significant difference test, represented in the table as homogeneous groups (HGs). Two-way ANOVA: ns = non-significant; *** indicates p ≤ 0.001 (confidence level ≥ 99.9%); ** indicates p ≤ 0.01 (confidence level ≥ 99%); * indicates p ≤ 0.05 (confidence level ≥ 95%). Y: yeast effect; A: activator effect; Y $\times$ A: interaction between yeast and activator. WY: wine obtained by spontaneous fermentation; WY + Act: wine obtained by spontaneous fermentation with an activator; AGGLO: wine obtained by fermentation with Levulia^® AGGLO; AGGLO + Act: fermentation with Levulia^® AGGLO supplemented with an activator. CAS: identification number assigned by the Chemical Abstracts Service. OPT: odor perception threshold (μg L⁻¹). OS: odorant series. n.f.: not found.

	CAS	WY	WY + Act	AGGLO	AGGLO + Act	HG	Y	A	$\mathbf{Y}\times \mathbf{A}$	OPT	OS
Alcohols (5)
Hexanol	111-27-3	2272 ± 168 b	2814 ± 228 b	1265 ± 502 a	3973 ± 502 c	3	ns	***	***	8000 I	3
2-Ethyl-1-hexanol	104-76-7	980 ± 86 c	855 ± 45 b	369 ± 81 a	1097 ± 81 c	3	**	***	***	8000 II	5
Octanol	111-87-5	266 ± 4 b	267 ± 6 b	94 ± 33 a	374 ± 33 c	3	*	***	***	800 III	8
Decanol	112-30-1	69 ± 1 b	71 ± 8 b	24 ± 6 a	69 ± 6 b	2	***	***	***	400 IV	8
Dodecanol	112-53-8	149 ± 15 b	141 ± 10 b	63 ± 17 a	187 ± 17 c	3	*	***	***	1000 V	8
Esters (15)
Ethyl butyrate	105-54-4	61 ± 2 bc	57.4 ± 0.3 b	30 ± 5 a	64 ± 5 c	3	***	***	***	20 I	1
Butyl acetate	123-86-4	53.7 ± 0.6 c	40 ± 3 b	20.6 ± 0.8 a	42.2 ± 0.8 b	3	***	**	***	66 VII	1, 6
Ethyl 2-methylbutyrate	7452-79-1	6.6 ± 0.4 b	6.4 ± 0.9 b	3 ± 1 a	15 ± 1 c	3	***	***	***	18 VIII	1, 2
Ethyl 3-methylbutyrate	108-64-5	17 ± 1 c	13.4 ± 0.5 b	6 ± 1 a	13 ± 1 b	3	***	*	***	3 II	1, 2
Isoamyl acetate	123-92-2	165 ± 2 c	201 ± 3 d	70 ± 12 a	138 ± 12 b	4	***	***	**	30 I	1
Ethyl hexanoate	123-66-0	22 ± 0.2 a	21.7 ± 0.2 a	21 ± 2 a	32 ± 2 b	2	***	***	***	14 II	1, 2
Hexyl acetate	142-92-7	298 ± 12 b	382 ± 4 c	211 ± 10 a	300 ± 10 b	3	***	***	ns	670 IX	1, 2
Ethyl heptanoate	106-30-9	2.2 ± 0.2 b	2.01 ± 0.04 ab	1.9 ± 0.2 a	2.1 ± 0.2 ab	2	ns	ns	*	2.2 VIII	1, 2
Ethyl octanoate	106-32-1	18.5 ± 0.3 a	20 ± 0.4 ab	21 ± 1 b	21 ± 1 b	2	**	ns	ns	5 I	1, 8
2-Phenylethyl acetate	103-45-7	1094 ± 48 b	1094 ± 40 b	902 ± 154 a	1877 ± 154 c	3	***	***	***	250 I	7, 9
Ethyl decanoate	110-38-3	48 ± 2 a	50 ± 2 a	136 ± 8 c	99 ± 8 b	3	***	***	***	200 I	1, 8
Phenethyl hexanoate	101-60-0	0.5 ± 0.07 a	0.54 ± 0.05 a	5.7 ± 0.2 c	1.8 ± 0.2 b	3	***	***	***	250 I	2, 8, 9
Ethyl tetradecanoate	124-06-1	66 ± 3 c	57 ± 3 b	48 ± 6 a	108 ± 6 d	4	***	***	***	4000 I	8
Phenethyl benzoate	94-47-3	13 ± 1 c	9.6 ± 0.3 b	3.7 ± 0.3 a	8.6 ± 0.3 b	3	***	*	***	n.f.	9
Ethyl hexadecanoate	628-97-7	145 ± 16 a	182 ± 13 b	204 ± 6 b	473 ± 6 c	3	***	***	***	2000 I	8
Aldehydes (4)
Hexanal	66-25-1	7.6 ± 0.1 b	6.8 ± 0.8 b	3.9 ± 0.3 a	10.1 ± 0.3 c	3	ns	***	***	5 X	3
Nonanal	124-19-6	5.7 ± 0.5 c	5.4 ± 0.2 c	1.7 ± 0.3 a	4.2 ± 0.3 b	3	***	***	***	2.5 II	5
Decanal	112-31-2	5.9 ± 0.6 c	5.3 ± 0.4 bc	4.7 ± 0.4 ab	4.2 ± 0.4 a	3	**	ns	ns	1.25 II	5, 8
2-Phenyl-acetaldehyde	122-78-1	129 ± 9 c	81 ± 3 ab	69 ± 9 a	86 ± 9 b	3	***	**	***	4 XI	3, 7
Ketones (1)
Acetophenone	98-86-2	n.d. a	n.d. a	14 ± 4 b	56 ± 4 c	3	***	***	***	65 XII	9
Lactones (2)
γ-Nonalactone	104-61-0	69 ± 6 c	50 ± 3 b	33 ± 8 a	84 ± 8 d	4	ns	**	***	30 I	4
γ-Decalactone	706-14-9	44 ± 2 c	26 ± 5 b	11 ± 4 a	40 ± 4 c	3	**	*	***	41 I	1, 4
Terpenes and Norisoprenoids (5)
Limonene	5989-27-5	420 ± 48 d	333 ± 11 c	31 ± 20 a	225 ± 20 b	4	***	**	***	10 II	4, 6
Nerolidol	7212-44-4	11 ± 1 b	0.8 ± 0.1 a	25 ± 3 c	34 ± 3 d	4	***	ns	***	700 II	3, 9
Z-Geranyl acetone	689-67-8	13 ± 1 c	11.7 ± 0.6 c	3.1 ± 0.2 a	6.9 ± 0.2 b	3	***	*	***	60 II	9
Farnesol	4602-84-0	52 ± 5 b	28 ± 2 a	126 ± 9 c	180 ± 9 d	4	***	**	***	20 VI	9
E-Methyldihydro-jasmonate	24851-98-7	36 ± 3 c	31 ± 2 b	11 ± 2 a	37 ± 2 c	3	***	***	***	70 II	9

1: fruity; 2: green fruit; 3: green; 4: creamy; 5: citrus; 6: chemistry; 7: honey; 8: waxy; 9: floral. I: (Dumitriu et al., 2024) [40]; II: (Palenzuela et al., 2023) [26]; III: (Peinado et al., 2006) [48]; IV: (Moreno et al., 2005) [49]; V: (Martín-García et al., 2023) [50]; VI: (Zhu et al., 2019) [51]; VII: (Peinado-Pardo et al., 2016) [52]; VIII: (López de Lerma et al., 2018) [53]; IX: (Gambetta et al., 2014) [54]; X: (Buttery et al., 1988) [55]; XI: (Buttery et al. 1971) [56]; XII: (Leffingwell.com) [57].

The data matrix built with the contents of the 32 quantified minor volatiles was subjected to a PCA, the results of which are plotted in Figure 3. The first two principal components explain 89.7% of the total variance provided by this matrix (PC1 57.37 and PC2 32.33%), and the scores for each component of the wine samples show a clear clustering of them. In this regard, WY and WY + Act are closely grouped on the positive side of PC1, indicating their high degree of similarity. In contrast, AGGLO is located on the negative side of PC1, while AGGLO + Act shows positive scores for PC1 and PC2, which is related to the higher influence of the supplement added to the must before its fermentation with this flocculant yeast.

The loadings of each volatile compound in the two PCs plotted (Figure 3) allows us to understand the separation of the wine samples. For instance, compounds such as ethyl hexanoate, acetophenone, ethyl hexadecanoate, farnesol, and nerolidol have strong positive contributions to PC2 (Table S2. Supplementary Materials), influencing the separation of AGGLO + Act. Conversely, compounds like phenethyl hexanoate and ethyl decanoate show negative loadings on PC1, which contribute to the separation of AGGLO wines from the other wines. The grouping of WY and WY + Act, both with positive PC1 scores, suggests that the presence of esters such as the acetates of higher alcohols and ethyl esters of 2 and 3-methylbutyrate, E-methyldihydrojasmonate, and ethyl hexanoate significantly influence these samples. As a summary, the tested additive has a different effect on the volatile profile on wines from AGGLO yeast and those obtained by spontaneous fermentation (WY), probably due to the enrichment of the must in the ONC that it provides and its influence on the production of certain esters and higher alcohols [45]. These results underscore the complex interplay between yeast strain, fermentation dynamics, and nutrient supplementation on the volatilome of wines and, consequently, on their aroma.

Flocculant yeast has a higher nitrogen demand for its development [58], which could explain the different effects of ONCs in spontaneous and flocculant yeast fermentations. A possible underlying mechanism is the increased nitrogen flux through the Ehrlich pathway, in which amino acids are degraded to form keto acids, which are then converted into aldehydes and higher alcohols. These compounds can combine with cellular or wine acids, contributing to ester formation. However, while this mechanism is consistent with yeast physiology, additional studies would be needed to confirm its specific role in the interaction between the ONC and flocculant yeast under these conditions.

3.5. Odorant Series

Table 4. Mean and deviation values of odorant series attributed for each compound in wines. Letters (^{a, b, c, d}) in the same row indicate statistical differences at the p ≤ 0.05 level according to Fisher’s least significant difference test, represented in the table as homogeneous groups (HGs). Two-way ANOVA: ns = non-significant; *** indicates p ≤ 0.001 (confidence level ≥ 99.9%); ** indicates p ≤ 0.01 (confidence level ≥ 99%); * indicates p ≤ 0.05 (confidence level ≥ 95%). Y: yeast effect; A: activator effect; Y $\times$ A: interaction between yeast and an activator. WY: wine obtained by spontaneous fermentation; WY + Act: wine obtained by spontaneous fermentation with the activator; AGGLO: wine obtained by fermentation with Levulia^® AGGLO; AGGLO + Act: fermentation with Levulia^® AGGLO supplemented with the activator.

Odorant Series	WY	WY + Act	AGGLO	AGGLO + Act	HG	Y	A	$\mathbf{Y}\times \mathbf{A}$
Fruity (1)	26.7 ± 0.5 b	25.3 ± 0.1 b	15.7 ± 0.7 a	26 ± 1 b	2	***	***	***
Tropical fruit	17.9 ± 0.3 b	17.6 ± 0.2 b	9.5 ± 0.7 a	18 ± 1 b	2	***	***	***
Sweet fruit	6.6 ± 0.2 c	5.9 ± 0.1 b	4.6 ± 0.2 a	6.6 ± 0.2 c	3	***	***	***
Forest fruit	1.2 ± 0.1 b	1.09 ± 0.03 b	0.94 ± 0.04 a	1.4 ± 0.1 c	3	ns	**	***
Green fruit (2)	9.1 ± 0.6 c	7.9 ± 0.1 b	5 ± 0.2 a	8.8 ± 0.4 c	3	***	***	***
Green (3)	34 ± 2 c	22.1 ± 0.9 b	18 ± 1 a	24 ± 2 b	3	***	*	***
Creamy (4)	12.9 ± 0.3 d	8.4 ± 0.2 b	5.8 ± 0.5 a	11.5 ± 0.4 c	4	***	*	***
Citrus (5)	49 ± 4 d	40 ± 1 c	7.5 ± 0.3 a	28 ± 2 b	4	***	**	***
Chemistry (6)	58 ± 5 d	49 ± 1 c	20.2 ± 0.3 a	40 ± 2 b	4	***	**	***
Terpenic	42 ± 5 d	33 ± 1 c	3.1 ± 0.5 a	23 ± 2 b	4	***	**	***
Ethereal	4.61 ± 0.02 a	4.78 ± 0.05 b	4.64 ± 0.03 a	4.7 ± 0.09 ab	2	ns	**	ns
Alcoholic	1.7 ± 0.1 a	1.62 ± 0.05 a	1.55 ± 0.07 a	1.66 ± 0.06 a	1	ns	ns	ns
Fusel	9.1 ± 0.1 a	9.1 ± 0.2 a	10.3 ± 0.2 b	10.6 ± 0.1 c	3	***	ns	ns
Honey (7)	37 ± 2 d	24.7 ± 0.8 b	21 ± 1 a	29 ± 3 c	4	**	ns	***
Waxy (8)	9.4 ± 0.4 a	9.3 ± 0.4 a	9.1 ± 0.2 a	9.3 ± 0.7 a	1	ns	ns	ns
Floral (9)	10.5 ± 0.6 b	8.9 ± 0.4 a	15.1 ± 0.4 c	21.6 ± 1.2 d	4	***	***	***

shows the values of the nine main odorant series grouping the 45 quantified major and minor volatiles. The fruity and chemical series include other sub-series grouping volatiles with specific descriptors and a low perception threshold. Some of the descriptors for each of these sub-series could be tropical fruit: pineapple; sweet fruit: peach; forest fruit: strawberry; terpenic: rose; ethereal: glue; fusel: nail polish; and alcoholic: alcohol. The main series of green fruit and green (or herbaceous) show three HGs, fruity shows two, and the remaining four series show four HGs according to the four wines obtained. Of all the series and subseries considered, only the waxy series and the forest fruit, ethereal, and alcoholic subseries were not dependent on the yeast inoculation strategy studied. In addition, the alcoholic and fusel subseries, together with the honey and waxy series, showed no significant dependence on the supplement added.

In order to better understand and visualize the effects of the two variation factors studied on the aroma profiles obtained from the odorant series, the odorant activity values for each main series and wine were subjected to a Multiple Variance Analysis (MVA), the results of which are presented in Figure 4 as sunray plots. Each ray corresponds to one of the nine series, and the wines from WY show the most regular polygon, with a low value only for the fruity series (1). This profile is negatively affected by the addition of the activator, since the wines WY + Act show a decrease in the OAVs of all the series considered. On the other hand, the wine from AGGLO has the most irregular and small polygon, suggesting that this yeast is not a good aroma producer. However, contrary to WY wine, the effect of the activator increases the OAV of all the series in AGGLO + Act wines, especially the floral (9) and fruity (1) series.

In the case of the fruity seriegs, this may be due to the increase in esters in wines with the activator, which has been linked to a fruity character in other studies. In addition, there may be better preservation of the fruity character not only due to the increase in esters but also due to their adsorption on the cell walls, preventing their loss [45]. Similarly, some studies associate the addition of yeast autolysates, in the form of activators, with the production of terpenes, which are mainly responsible for these aromas [43]. Lastly, some authors suggest that the fruity attributes could be related, at least in part, to the protection of certain aroma compounds from their oxidation by components of the activators, likely during the initial steps of winemaking [59]. Overall, the comparison of these four conditions highlights the interplay between the predominant fermentative yeast behavior and the presence of activators in shaping the wine’s aromatic profile.

The results show that alcoholic fermentation is complete for all the strategies tested and that the wines produced by the AGGLO yeast have lower aroma profiles than those obtained by spontaneous fermentation with the wild yeasts. However, the addition of the authorized additive tested, containing a source of assimilable organic nitrogen, increased their production of volatile compounds with a high aroma activity value in the wines from AGGLO but not for those of spontaneous fermentation. These effects are described for the first time in this research, although similar results have been observed in fermentations carried out with non-Saccharomyces yeasts [7] supplemented with an additive based on yeast lees extracts, and also in fermentations of musts added with a commercial inactive yeast extract enriched with glutathione and a commercial ADY Saccharomyces strain to obtain pink wines [59].

3.6. Organoleptic Evaluation

Figure 5 shows the normalized and scaled values of the scores obtained for clarity, color, aroma frankness, positive aroma intensity, aroma quality, taste frankness, positive taste intensity, taste persistence, taste quality, and overall quality attributes, evaluated by the tasters. According to this plot, wines obtained by spontaneous fermentation of the grape must showed higher values for all the attributes, except for the clarity and color. In contrast, wines obtained with AGGLO yeast showed lower scores for attributes such as aroma frankness, positive taste intensity, and aroma quality. In general, an increase in the values of all attributes is also obtained when the fermentation activator is used in both WY + Act and AGGLO + Act.

On the other hand, AGGLO showed lower values for aroma frankness, positive taste intensity, and aroma quality. Both WY + Act and AGGLO + Act improved the values regarding clarity. This may be due to the protector effect that mannoproteins from the yeast lysis have, decreasing particle size and consequently haze formation [44].

Differences in the scored aroma attributes should be interpreted as a consequence of the differences in the contents of the active volatile compounds quantified. However, when considering the sum of the attributes assessed for sight, smell, taste, overall quality, and total score (

Table 5. Evaluation of organoleptic attributes. Mean score, standard deviation, and homogeneous groups at a confidence level of 95% (p ≤ 0.05). Letters (^a) in the same row indicate significant differences between wines. Abbreviations used for wines obtained by fermentation with WY: wild yeast; WY + Act: the same with the addition of an activator; AGGLO: Levulia^® AGGLO yeast; AGGLO + Act: the same with the addition of an activator.

Attributes	WY	WY + Act	AGGLO	AGGLO + Act	HG	p-Value
Sight	10 ± 2 a	11 ± 2 a	11 ± 1 a	12 ± 1 a	1	0.506
Smell	24 ± 3 a	24 ± 2 a	23 ± 3 a	23 ± 4 a	1	0.740
Taste	34 ± 5 a	36 ± 4 a	33 ± 5 a	33 ± 5 a	1	0.544
Overall quality	9 ± 1 a	9 ± 1 a	9 ± 1 a	9 ± 1 a	1	0.401
Total score	78 ± 8 a	80 ± 8 a	76 ± 8 a	77 ± 10 a	1	0.708

), no significant differences were found between the wines at the 95% confidence level (p ≤ 0.05). Slight differences were observed in the scores for wine produced by spontaneous fermentation, with the activator (WY + Act) obtaining the highest total score (80 points) and AGGLO obtaining the lowest (76 points). These results are similar to those obtained by other authors who attempted to establish relationships between the sensory analysis and the content in active volatile compounds [45].

4. Conclusions

A commercially available additive rich in organic nitrogen compounds was tested in four white wine fermentations, using both spontaneous and flocculent yeast-driven processes. As a control, two of these fermentations were conducted without additive supplementation. No significant differences in standard oenological variables were obtained. The supplement added increased the fermentation rate of the flocculent yeast and reduced the volume of lee formation. Wines obtained by spontaneous fermentation with the tested additive showed a significant reduction in fruity, green fruit, green, creamy, citrus, chemistry, honey, and floral odorant series values compared to those without the additive. In contrast, wines fermented with flocculent yeast and supplemented with the additive showed a significant increase in all the aroma series, except the waxy series, compared to their counterparts.

Volatile compounds analysis suggests a strong interplay between the predominant fermentative yeast behavior and the supplementation of must with the tested additive in shaping the wine’s aroma profile. Depending on the dominant yeast during fermentation, the presence of the additive affects the volatile compounds and, consequently, the odorant series differently. Although volatile differences were observed, the lack of significant sensory differences suggests that the impact on aroma perception may be limited. Further organoleptic studies are required to confirm and extrapolate these results.

This study enhances the understanding of flocculent yeast metabolism and its fermentation process, highlighting its distinct behavior in nitrogen-rich environments compared to wild yeasts. The combination of natural yeast extracts, rich in organic nitrogen compounds, with selected yeast strains improves fermentation efficiency, increasing yield and aroma production—an important advancement for the winemaking industry. Future work should analyze additional nutrient components beyond ONC levels to better understand the effects of ONC addition and yeast inoculation on fermentation. In addition to fermentation dynamics and immediate wine quality, further studies should also examine the genetic identification of the yeast isolated thought the vinification, wine protein stabilization, and potassium acid tartrate (potassium bitartrate) stabilization during storage to build on the findings of this study.