Introduction

Recent studies to reduce the amount of alcohol produced in wines during fermentation have focused on the use of non-Saccharomyces species. While Saccharomyces cerevisiae converts sugar from must into ethanol and CO₂, non-Saccharomyces yeasts metabolize sugar to by-products other than ethanol such as biomass, glycerol, or succinic acid. Among the species examined, Metschnikowia pulcherrima has been reported to not only produce wines containing less alcohol but also increase glycerol and decrease acetic acid (Comitini et al., 2011; Contreras et al., 2014; González-Royo et al., 2015; Quirós et al., 2014). In fact, researchers have observed that sequential inoculation of M. pulcherrima with S. cerevisiae reduces ethanol by 0.20 % to 3.0 % (v/v) depending on conditions, yeast species and strains, and medium (Aplin & Edwards, 2021; Carbon et al., 2023; Contreras et al., 2014; Edwards & Aplin, 2022; Garciá et al., 2020; Hranilovic et al., 2020; Varela, 2016). More recently, Aplin et al. (2019) noted that M. pulcherrima provided 0.9 % (v/v) ethanol reduction in Syrah wines having an initial sugar (glucose and fructose) concentration of 301 g/L. Carbon et al. (2023) also reported that using M. pulcherrima in a synthetic grape juice medium with 240 g/L of sugar led to a reduction of up to 3 % (v/v) in alcohol content while reaching dryness.

Like S. cerevisiae, non-Saccharomyces yeasts require specific nutrients for growth and metabolism, with a preference for nitrogen-containing compounds such as amino acids and ammonium. Effective nitrogen management is therefore crucial when applying sequential inoculation. Most nitrogen additions are applied practically, without considering the cell's nitrogen needs or the timing throughout wine fermentation (Beltran et al., 2005). Early nitrogen supplementation may result in excessive biomass production, whereas delayed addition can potentially lead to fermentation challenges such as sluggish fermentations and/or the production of off-odors/flavors (Beltran et al., 2005; Bisson, 1999; Gobert et al., 2019; Varela et al., 2004; Vilanova et al., 2007). In addition, nitrogen starvation causes changes in the kinetics of yeast metabolism and intracellular protein content, specifically RNA, glycogen, and trehalose (Schulze et al., 1996; Walker, 1998). Fermentation kinetics are also affected by the timing of nitrogen supplementation (Bely et al., 1990a; Godillot et al., 2022). Therefore, the objective of this study was to evaluate the impact of nitrogen addition timing during sequential inoculation of M. pulcherrima and a commercial strain of S. cerevisiae in the fermentation of Chardonnay grape must, with the aim of optimizing yeast assimilable nitrogen (YAN) management strategies for winemakers.

Materials and methods

1. Preparation of Starter Cultures

M. pulcherrima P01A016 was previously isolated from vineyards located at the Irrigated Research and Extension Center, Washington State University (Prosser, WA, USA) as described by Bourret et al. (2013). S. cerevisiae (Lalvin NBC^TM) was obtained from Lallemand Inc. (Montréal, Quebec, Canada). Both yeasts were cultured using yeast extract-peptone-glucose (YPG) broth/agar (Zott et al., 2010) which contained 10 g/L yeast extract (Becton, Dickinson, and Company, Sparks, MD, USA), 10 g/L peptone and 20 g/L glucose (VWR International, Radnor, PA, USA), as well as 20 g/L agar (Acros Organics, Morris, NJ, USA).

Single colonies of M. pulcherrima were aseptically removed from solidified media and transferred to 10 mL YPG broth. After the incubation at 28 ˚C for 24 h, the culture was transferred to 100 mL. After reaching the end of exponential growth phase, cells were harvested by centrifugation at 2500 × g for 15 min, washed twice with 0.2 M sodium hydrogen phosphate buffer (Na₂HPO₄, pH 7), and resuspended in the buffer before inoculation into grape must. Active dried cultures of S. cerevisiae were prepared according to the manufacturer’s instructions.

2. Vinification

Chardonnay grape juice (25.6 ˚Brix, pH 3.65, 48.4 mg N/L YAN) was obtained from Columbia Crest Winery (Paterson, WA, USA). Soluble solids (˚Brix), the populations of total yeasts and non-Saccharomyces yeasts, and concentrations of SO₂ and YAN were measured as described in the following section. An additional 30 mg/L total SO₂ was added as K₂S₂O₅ while the concentrations of glucose and fructose (1:1) were adjusted to yield 28 ˚Brix before transfer into 3 L Celstir fermentation vessels (Wheaton Science Products, Millville, NJ, USA). Duplicate fermentations were conducted at 20 °C based on a completely randomized design with four treatments; no YAN addition (Treatment A) or YAN added on day 1 (Treatment B), day 3 (Treatment C), or day 5 (Treatment D) (Table 1). Treatment A was established as a control to assess the effects of nitrogen addition. In terms of fermentation progress, day 1 corresponded to the lag phase of the alcoholic fermentation (27.7 ˚Brix), day 3 to exponential growth (26.9 ˚Brix), day 5 to stationary phase (17.2 ˚Brix), and day 8 to middle of the fermentation (9.2 ˚Brix). All experiments were conducted with two independent replicates to ensure reproducibility. According to Scott Laboratories (2022) (https://scottlab.com/fermentation-nutrition-planning), theoretically, a must with 28 °Brix should contain 260 mg N/L YAN to complete fermentation. Therefore, the must, which initially contained 48.4 mg N/L YAN, was supplemented with two separate Fermaid K^TM additions, containing inorganic nitrogen (DAP), organic nitrogen (alpha amino nitrogen), essential nutrients (biotin, magnesium sulfate, folic acid, thiamine, niacin, and calcium pantothenate), and inactivated yeast (Lallemand Inc., 2025), of a calculated 105.8 mg N/L YAN to reduce the risk of sluggish or stuck fermentation. Table 1 shows the timing of additions and inoculation for M. pulcherrima and S. cerevisiae. Musts were inoculated with M. pulcherrima at 10⁵ CFU/mL on day 0, with S. cerevisiae added at 10⁶ CFU/mL on day 4. For the control, S. cerevisiae was only inoculated on day 4. For control fermentations with only S. cerevisiae inoculated (Treatment E), YAN was added on days 1 and 4 after S. cerevisiae inoculation, and this fermentation was carried out concurrently with the other treatments through inoculation with S. cerevisiae on day 4. Treatment E served as an additional control to evaluate alcohol reduction through treatments.

3. Analytical Methods

A spiral plater (Autoplate 4000, Spiral Biotech, Bethesda, MD) was used to monitor yeast culturability using lysine agar (Oxoid, Hampshire, England) for non-Saccharomyces yeasts and Wallenstein Laboratory agar (WL, Becton, Dickinson and Company, Franklin Lakes, NJ, USA) for total yeast populations. Prior to colony counting, all plates were incubated for three days at 28 °C.

Fermentation progress was tracked by measuring soluble solids (°Brix) with a portable density meter (DMA35, Anton-Paar, Graz, Austria). However, °Brix does not directly indicate dryness since it is affected by ethanol and other dissolved compounds. Therefore, fermentation was considered complete when residual sugar fell below 4 g/L, which was confirmed using AimTab™ reducing sugar tablets (Germaine Laboratories Inc., TX, USA) and then HPLC. Molecular SO₂ concentrations were calculated according to Buechsenstein and Ough (1978).

Biomass was measured triplicate using the method described by Schnierda et al. (2014) and used to inform the effects of timing of nitrogen addition. To monitor YAN levels throughout fermentation, amino acids were quantified using the NOPA method (Dukes & Butzke, 1998), and ammonium levels were measured with an ion-selective probe (Denver Instruments, Orville, NY, USA), following Zoecklein et al. (1995). YAN was calculated as the sum of free amino acids and ammonium concentrations.

Residual sugars (glucose and fructose), glycerol, ethanol, and organic acids (acetic, and succinic acids) were measured using an Agilent 1100 HPLC system (Agilent Technologies, Santa Clara, CA, USA) fitted with a refractive index detector with some modifications to Eyéghé-Bickong et al. (2012) in triplicate. Samples were centrifuged and filtered through 0.45 µm polyethersulfone membranes (MilliporeSigma) into crimp-top vials. An Aminex HPX-87H column (300 × 7.8 mm, BIO-RAD, Hercules, CA, USA) maintained at 60 °C was used for separation using a mobile phase of 0.005 M H₂SO₄ and a flow rate of 0.6 mL/min. Instrument controls were performed using ChemStation software (version C.01.10) based on external calibration curves (r² ≥ 0.99).

4. Statistical Methods

Data analyses were carried out using Excel 365 (Microsoft Corporation, Richmond, WA, USA). Data related to yeast population, ammonium, amino acids, soluble solids, and biomass were reported as mean values accompanied by standard deviation from duplicate fermentations. In addition, statistical analyses about biomass, glucose, fructose, glycerol, ethanol, acetic acid, and succinic acid were conducted with ANOVA with mean separations performed by Fisher’s least significant difference at p ≤ 0.05 using XLSTAT (Lumivero, Denver, CO, USA).

Results

1. Changes in yeast population, YAN, soluble solids, and biomass

The Chardonnay must initially contained 48.4 mg N/L YAN and before the inoculation, the indigenous non-Saccharomyces and total yeast populations were found as (1.0 ± 0.6) × 10⁴ CFU/mL, and (1.4 ± 0.6) × 10⁴ CFU/mL, respectively (mean ± standard deviation). Changes in populations, YAN, and soluble solids in Chardonnay ferments inoculated with M. pulcherrima and S. cerevisiae but without additional nitrogen (Treatment A) are presented in Figure 1. Here, non-Saccharomyces populations were initially 4.0 ± 2.8 × 10⁵ CFU/mL on day 1 but reached 1.9 ± 0.0 × 10⁷ CFU/mL on day 4. However, the population of non-Saccharomyces yeasts decreased to 7.0 ± 0.5 × 10⁶ CFU/mL on day 7, whereas total yeast population mainly S. cerevisiae was maintained at 10⁷ to 10⁸ CFU/mL until day 25. The grape must initially contained 48.4 mg N/L YAN which were mostly exhausted by day 4 (Figure 1B). Finally, the fermentation did not reach dryness, even after 34 days.

When YAN was added one day after M. pulcherrima inoculation in the lag phase (Treatment B), the population trends of non-Saccharomyces and total yeast followed a similar pattern to those of Treatment A during the first four days, but YAN consumption was higher (Figure 2). Yeasts in Treatment B depleted nearly all of the YAN, including both preexisting and added sources. Despite this, not all of the ammonium and amino acid sources were consumed following YAN addition. Rather, concentrations decreased to 1.15 mg N/L. From days 4 to 8, the YAN level fluctuated between 1 to 2 mg N/L, while the non-Saccharomyces population decreased from 10⁷ to 10⁶ CFU/mL. YAN consumption was much lower after the second YAN addition. As evidence, on day 9, YAN levels measured 89.2 mg N/L, while YAN measured 79.1 mg N/L on day 19. After day 19, YAN concentration increased by 35.2 mg N/L during the final fermentation stage. Unlike Treatment A, Treatment B reached dryness by day 31 (Figure 2B).

The yeast populations and fermentation characteristics were similar for treatments B and C, but there was a slight difference in nitrogen consumption (Figure 3). While the majority of YAN was consumed after three days in Treatment B, this same level of depletion required only one day for Treatment C. As evidence, nearly all YAN that was added on day 3 in the exponential growth phase was consumed by day 4, leaving only 2.38 mg N/L of YAN (Figure 3B). Following the second addition of YAN, YAN consumption closely resembled treatment B. On day 9, YAN levels measured 89 mg N/L, as compared to 81.9 mg N/L on day 19. After day 19, YAN concentration (primarily ammonium) increased by 33.7 mg N/L towards the final stage of fermentation. Similar to Treatment B, dryness was reached in 31 days.

While the first addition of YAN on day 5 towards the stationary phase did not affect the yeast population in treatment D, it reduced fermentation time compared to treatment A (Figure 4). The non-Saccharomyces yeast population peaked at 10⁷ CFU/mL on day 4 and then declined to 10⁶ CFU/mL on day 7 (Figure 4A). Nearly all of the naturally occurring YAN was metabolized before nitrogen was added, with 1.34 mg N/L measured on day 4 (Figure 4B). Following the addition on day 5, YAN was rapidly metabolized and only 1.32 mg N/L remained on day 7. Since the YAN additions in this treatment were closer together by time, the second added YAN was consumed more slowly than other treatments. However, a greater quantity of YAN was metabolized in the time following the second addition in treatment D compared to other treatments. Following day 19, YAN concentration (predominantly ammonium) rose by 30.0 mg N/L as fermentation reached its final stage, similar to those in treatments B and C. Fermentation in treatment D reached dryness on day 25, six days earlier than treatments B and C.

As previously mentioned, before the commercial S. cerevisiae inoculation, the indigenous non-Saccharomyces and total yeast populations were counted as 1.0 × 10⁴ CFU/mL, and 1.4 × 10⁴ CFU/mL, respectively. After the commercial S. cerevisiae inoculation in Treatment E, total yeast population increased 6.5 × 10⁶ CFU/mL 1 day later and then up to 2.8 × 10⁸ CFU/mL within 24 hours (Figure 5A). Until reaching dryness, total yeast population was around 3.0 ± 1.0 × 10⁸ CFU/mL, after that point, it decreased to 10⁷ CFU/mL. On the other hand, the indigenous non-Saccharomyces yeast population were counted as 5.0 × 10⁴ CFU/mL 1 day later, 1.5 × 10⁵ CFU/mL 2 days later and reached to 7.0 × 10⁵ CFU/mL 4 days later after commercial S. cerevisiae inoculation and then it was not detected after that point. Yeasts consumed nearly all YAN in the must within two days. After the second YAN addition on day 9, YAN was measured at 95 mg N/L but decreased to 89 mg N/L by day 19. Despite reaching dryness on day 19, the YAN concentration increased by an additional 27 mg N/L when the final measurements were taken on day 31 (Figure 5B).

Prior to S. cerevisiae inoculation, treatment C produced 3.33 g/L of biomass, significantly higher than the other treatments on day 4 (data not shown). In contrast, the other treatments produced similar biomass levels, ranging from 2.50 g/L to 2.67 g/L (Figure 6).

2. Wine composition

The impact of nitrogen addition during the first 10 days of fermentation was tracked by monitoring residual sugar, alcohol, glycerol, and organic acid production. Until day 7, ethanol, glycerol production, and sugar consumption were greater in treatments B and C compared to treatments A and D (Figure 7 and 8). After day 7, ethanol and glycerol production in treatment D increased, surpassing the other treatments. For organic acids, the acetic acid level in treatment B was higher than in the other treatments until day 7. However, all treatments produced similar amounts of succinic acid over the first 10 days.

There were few differences in the chemical composition of treatments that received supplemental nitrogen (Table 2). In contrast, treatment A, without YAN addition, differed significantly from the other treatments in concentrations of ethanol, glycerol, fructose, and succinic acid. For Treatments A, B, C, and D, wines containing M. pulcherrima had alcohol contents of 14.2, 16.1, 15.5, and 15.9 % (v/v), respectively, whereas the wine inoculated with only S. cerevisiae as a control fermenter had 16.3 % (v/v) alcohol content. Treatment A produced significantly less ethanol but had a high residual fructose concentration. Furthermore, glycerol production was significantly highest in treatment D (10.5 g/L), compared to treatments A (8.65 g/L), B (9.64 g/L), C (9.78 g/L), and E (10.1 g/L). In addition, acetic acid levels in treatment D (0.786 g/L) were significantly higher than in treatment E (0.627 g/L), while treatments B (0.739 g/L) and C (0.672 g/L) showed no significant difference. Treatment A had a significantly higher succinic acid concentration (0.569 g/L) than treatments B, C, D, and E (0.277, 0.326, 0.371, and 0.379 g/L, respectively).

Discussion

The timing of nitrogen addition is crucial for yeast assimilation. In this study, nitrogen was supplemented with Fermaid K^TM, a complex nutrient containing inorganic nitrogen (ammonium), amino acids, and other micronutrients. Treatment C consumed the supplemented nitrogen within 24 hours, aligning with the findings of Beltran et al. (2005) and Su et al. (2021). In Treatments B and D, the yeasts utilized nitrogen more slowly. By day 4, yeasts had consumed nearly all measurable ammonium in the must. At that point, amino acids were more abundant than ammonium in all treatments but remained below 2.15 mg N/L. Our results confirm that ammonium was preferentially assimilated from the available nitrogen sources, as supported by previous studies (Bell & Henschke, 2005; Brice et al., 2018; Jiranek et al., 1995; Torrea et al., 2011). The addition of Fermaid K^TM enhanced overall nitrogen assimilation, as yeasts in Treatments B, C, and D consumed both preexisting and added sources, consistent with Rollero et al. (2021).

Non-Saccharomyces yeasts assimilated nearly all measured YAN sources (ammonium and amino acids) in Treatments A, B, C, and D before S. cerevisiae inoculation. Gobert et al. (2017) showed that depending on species, these yeasts can metabolize between 66 to 215 mg N/L of YAN. This rapid YAN consumption by non-Saccharomyces yeasts can deplete nitrogen levels, potentially hindering S. cerevisiae growth and fermentation efficiency (Kemsawasd et al., 2015; Medina et al., 2012).

In our study, an initial YAN level of 260 mg N/L was applied to meet the moderate–high nitrogen requirements of the commercial S. cerevisiae strain under high-sugar conditions and to minimize the risk of sluggish or stuck fermentations. This is consistent with the guideline of Bisson and Butzke (2000), who recommend approximately 25 mg N/L per 1 ˚Brix increase. Although elevated YAN could theoretically support the growth of spoilage organisms such as Brettanomyces bruxellensis, Childs et al. (2015) demonstrated that at similar YAN levels (250 mg N/L) in a high-sugar medium, B. bruxellensis was not detected due to the inhibitory effect of high ethanol concentrations (15.6–15.9 % v/v). Likewise, the wines in our study reached 15.5–16.3 % (v/v) ethanol, which likely limited the risk of microbial instability despite the relatively high YAN.

To avoid sluggish or stuck fermentation, a second YAN addition was made on day 8 using Fermaid K^TM. YAN was only partially consumed, in contrast to the efficient uptake observed with earlier additions. The possible reasons of that might decrease cell viability, elevated ethanol concentrations (Varela et al., 2004; Tesnière et al., 2013). In addition, contrary to expectations, the measured YAN levels, particularly ammonium, increased towards the end of fermentation. This rise could be due to yeast autolysis, where nitrogen is released as cells die in the final stage of the stationary phase (Alexandre & Guilloux-Benatier, 2006). During autolysis, intracellular nitrogen is released into the surrounding environment, increasing ammonium and amino acid levels. This rise is consistent with both the findings of this study and those of previous studies (Beltran et al., 2004; Martínez-Rodríguez & Polo, 2000; Martínez-Rodríguez et al., 2001).

The timing of nitrogen addition can significantly influence fermentation duration. In Treatment D, where YAN was supplemented with Fermaid K^TM after S. cerevisiae inoculation, fermentation reached dryness six days faster than in Treatments B and C, where YAN was added before S. cerevisiae inoculation, contrasting with Medina et al. (2012). Godillot et al. (2022) also noted that adding nitrogen during the stationary phase can accelerate fermentation kinetics. Similarly, Su et al. (2021) reported that yeast strains with high nitrogen needs, such as S. cerevisiae Lalvin NBC^TM strain, completed fermentation in less time when YAN additions were made. Supplementation after S. cerevisiae inoculation proved most successful in enhancing both fermentation activity and nitrogen assimilation.

In this study, Chardonnay grape musts with an initial sugar content of 28 ˚Brix reached dryness with YAN addition. In contrast, Treatment A (no YAN addition) did not achieve dryness. This may be attributed to the inactivation of sugar transporters. Nitrogen starvation hindered protein synthesis, likely affecting sugar transport activation during fermentation (Lucero et al., 2002; Mendes-Ferreira et al., 2004; Salmon, 1989). The addition of nitrogen in treatments B, C, D, and E likely supported sugar transport and resulted in greater sugar metabolism.

Treatment C produced significantly more biomass than the other treatments prior to the addition of S. cerevisiae. Treatment B did not show a similar increase in biomass. During the exponential growth phase at the onset of alcoholic fermentation, nitrogen availability is directly related to biomass output (Martínez-Moreno et al., 2012; Varela et al., 2004). Treatment C supports this relationship, while Treatment B does not exhibit the same effect. In addition, as noted by Bely et al. (1990b) and Varela et al. (2004), the late addition of YAN can accelerate fermentation without impacting biomass, as seen in Treatment D.

Wines with YAN addition showed no significant reduction in alcohol content. Treatments B, C, and D reached dryness and produced 15.5 to 16.1 % (v/v) alcohol. The control treatment (Treatment E) containing only S. cerevisiae produced 16.3 % (v/v) alcohol which is not significantly different from the other treatments. The lack of alcohol reduction in Treatments B, C, and D may be due to the high initial soluble solids (28 ˚Brix) and insufficient aeration. Similarly, Hranilovic et al. (2017) observed no significant alcohol reduction in Shiraz wine with 29 ˚Brix, sequentially inoculated with M. pulcherrima and S. cerevisiae. On the other hand, Aplin et al. (2019) reported a 0.9 % (v/v) alcohol reduction in Syrah wines (301 g/L initial sugar and 253 mg N/L YAN) with sequential inoculation of M. pulcherrima and S. cerevisiae.

YAN supplementation before or after S. cerevisiae inoculation impacted glycerol production across treatments B, C, and D. Treatment D produced the highest level of glycerol, 10.5 g/L. This may be due to the differential expression of 17 % of yeast genes, as observed in a study on the timing of nitrogen addition and its effect on gene expression such as GPD1 and ATF2 genes, which are roles in glycerol and acetate ester synthesis, respectively (Du et al., 2012; Godillot et al., 2022).

Succinic acid production can influence assimilable nitrogen and amino acids although its relationship with the timing of nitrogen addition remains unclear (de Klerk, 2010; Torres-Guardado et al., 2024). A previous study indicates that nitrogen availability has an adverse impact on succinic acid production, where lower YAN typically results in increased concentrations (Su et al., 2021). Our study found that treatment A, which had the lowest YAN, produced the highest levels of succinic acid. In contrast, YAN addition in treatments B through E decreased succinic acid production, with no significant impact observed from the timing of supplementation. All treatments except Treatment A had succinic acid levels below 0.5 g/L. In literature, the concentration of succinic acid in wine typically falls within the range of 0.5 to 1.5 g/L depending on the yeast species, and it contributes to salty-bitter notes (Chidi et al., 2015; Contreras et al., 2015; de Klerk, 2010). Therefore, succinic acid was dosed to comparatively evaluate the potential effects of YAN addition and its timing on its production.

Nitrogen addition can also decrease acetic acid production depending on timing of the addition and the amount of nitrogen (Barbosa et al., 2009; Su et al., 2021). Barbosa et al. (2009) reported that nitrogen addition on day 3 decreased acetic acid production, consistent with previous studies showing an inverse relationship between nitrogen availability and acetic acid formation under low to moderate nitrogen conditions but direct effect at high nitrogen level (Barbosa et al., 2009; Hernandez-Orte et al., 2006). In our study, treatment C (YAN supplementation on day 3) produced similar acetic acid levels comparable to the other treatments, contrary to the aforementioned studies. The timing of YAN addition did not significantly affect acetic acid production; however, a significant difference was observed between treatments D and E, which may be attributed to the inoculation with M. pulcherrima. Although a significant difference was observed between treatments D and E, acetic acid concentrations in all treatments remained above the detection threshold of 0.28 g/L but below 0.9 g/L, the level at which it typically imparts sour and bitter flavours (Contreras et al., 2015; Lambrechts & Pretorius, 2000; Ribereau-Gayon et al., 2006).

Overall, YAN supplementation is a key factor in completing alcoholic fermentation, especially in musts with high sugar content. In this study, by day 4, YAN levels were nearly depleted in all treatments, including the control group without YAN addition (Treatment A), as well as in the groups supplemented on day 1 (Treatment B) and day 3 (Treatment C), indicating early and intense nitrogen uptake by yeast. Therefore, a second YAN supplementation is recommended after day 4 to support fermentation. Although two separate YAN additions enabled fermentation completion, the timing of these additions produced distinct effects: addition on day 3 in the exponential growth phase (treatment C) promoted biomass formation, whereas addition on day 5 in the stationary phase (treatment D) shortened fermentation duration and increased glycerol production. These findings highlight the importance of optimizing both the amount and timing of YAN addition to align with specific fermentation objectives, such as optimizing yeast growth, decreasing fermentation duration or modulating metabolite production.

Conclusions

This study evaluated the effects of the timing of YAN addition on Chardonnay fermentation sequentially inoculated with M. pulcherrima and S. cerevisiae. Treatments with YAN addition reached dryness, though fermentation durations varied significantly. After the S. cerevisiae inoculation (treatment D), the addition of YAN significantly reduced the duration of fermentation (20 %) and produced noticeably more glycerol (10.5 g/L). However, there were no significant differences in alcohol reduction across the YAN supplemented treatments, possibly due to osmotic stress encountered by M. pulcherrima at the beginning of fermentation. In summary, while YAN addition did not affect alcohol reduction, it significantly influenced fermentation duration and glycerol production. Future studies could explore how nitrogen timing affects sensory and chemical profiles of wine.

Acknowledgements

The authors gratefully acknowledge Washington State Grape and Wine Research Program (Prosser, WA, USA), the Northwest Center for Small Fruits Research (Corvallis, OR, USA), the School of Food Science at Washington State University (Pullman, WA, USA) and the Ph.D. scholarship of Turkish Higher Education Council (YOK 100/2000, Ankara, Turkey), Lallemand Inc. and Columbia Crest Winery for financial and material support. This work was also supported in part by the USDA National Institute of Food and Agriculture, Hatch project 1016366.