Abstract
Background and goals Chambourcin (Vitis hybrid) red wine grapes, commonly grown in the southern and eastern United States, often exhibit low acidity and high pH at harvest, posing challenges for wine quality. Fermentation with non-Saccharomyces yeast such as Lachancea thermotolerans may influence acidity and other key wine attributes.
Methods and key findings In 2020 and 2021, Chambourcin grapes from Arkansas were processed into wine using four fermentation treatments: Saccharomyces cerevisiae with and without malolactic fermentation (SC-MLF and SC, respectively) and sequential inoculation with L. thermotolerans and S. cerevisiae with and without malolactic fermentation (LT-SC-MLF and LT-SC, respectively). Wine composition and color were evaluated at bottling and during storage at 15°C (0, 6, and 12 mo). In addition, sensory evaluations by consumers (n = 126) and by a professional descriptive panel (n = 7) were conducted on the 2020 wines. Wines fermented with L. thermotolerans (LT-SC) showed lower pH (3.34), higher titratable acidity (TA; 12.28 g/L), and more intense red coloration compared to SC wines (pH 3.64, TA 7.15 g/L). LT-SC wines also had higher lactic acid (5.84 g/L) and total organic acids (12.13 g/L) than SC wines (3.22 and 7.50 g/L, respectively). However, the higher acidity and organic acid levels were negatively correlated with consumer liking of sourness and overall impression, while a higher pH was associated with greater consumer acceptance.
Conclusions and significance Sequential fermentation with L. thermotolerans and S. cerevisiae produced wines with improved acidity and color, but the increased sourness reduced consumer acceptability. Adjustments such as blending or sugar additions at bottling may enhance consumer appeal while retaining the desirable quality attributes imparted by L. thermotolerans.
Introduction
Chambourcin (Seyve-Villard 12-417 × Chancellor) is a Vitis hybrid grapevine grown in the midwestern and eastern United States. Despite being considered one of the best red-wine hybrid varieties due to plant yield, fungal resistance, and economic potential, it can have low acidity and high pH at harvest, especially in warm growing climates (Dami et al. 2006, Prajitna et al. 2007, Robinson et al. 2012, Homich et al. 2016, Pedneault and Provost 2016). Although studies over recent decades have explored methods for growing Chambourcin grapes (Miller et al. 1996, Ferree et al. 2004, Singh et al. 2024), studies on the challenges of Chambourcin wine production are more limited, particularly concerning fermentation techniques. There is a gap in understanding of how different wine production methods—such as yeast selection, including fermentation with non-Saccharomyces yeast such as Lachancea thermotolerans—affect the composition, color, and flavor of Chambourcin wines. In addition to demonstrating the importance of Chambourcin as a red wine variety, such research expands the potential for other hybrid cultivars and Vitis vinifera cultivars that also have attributes challenging for wine production.
Climate changes over recent decades in the U.S. have affected grape production. Since the beginning of the 20th century, near surface air temperatures in Arkansas have risen by 0.3°C, and the period from 2015 to 2019 was the warmest consecutive 5-yr interval (Runkle 2022). In 2023, the United States Department of Agriculture (USDA) plant hardiness zones in the U.S. were reclassified. In Arkansas, the cold hardiness zones that previously ranged from 6b to 8a shifted to the range of 7a to 8b. Thus, the average annual extreme-minimum winter temperatures increased by 2.7°C (5°F), an indication of the warming growth climate (as found on the USDA website: https://planthardiness.ars.usda.gov/).
Rising temperatures influence Chambourcin grape composition prior to harvesting, which can result in low acidity and high pH values at harvest. This can negatively affect wine color and microbial stability, resulting in more addition of sulfur dioxide (SO2) to mitigate spoilage. Furthermore, acid-deficient and high-pH wines directly affect production decisions concerning management of microbial stability and wine flavor, which are largely influenced by individual organic acid concentration (Akin et al. 2008). The primary acids contributing to the total acidity of wine are tartaric, malic, lactic, and citric acids. Tartaric and citric acids confer freshness to a wine, whereas malic acid can contribute harsh sensations, depending on its concentration. Lactic acid—a product of malolactic fermentation (MLF), the conversion of malic acid to lactic acid by the lactic acid bacteria Oenococcus oeni—contributes to a softer mouthfeel in wine (Vicente et al. 2022).
Previous research on Arkansas-grown Chambourcin showed that grapes at harvest had a pH of 3.4 to 3.7 and a titratable acidity (TA) of 6.0 to 7.7 g/L, indicating that high temperatures during grape ripening in Arkansas and other regions of the mid-South can contribute to inconsistent acid and pH levels (Mayfield et al. 2021, Fleming and Threlfall 2024). Furthermore, because hybrid grapes can overcrop, the quality of grape/wine anthocyanins and total phenolics, both of which are pH dependent, can be negatively affected (Mazza et al. 1999, Dami et al. 2006). Although the resulting pH and TA levels could be considered acceptable for V. vinifera varieties, lower pH values are essential for improving the color and phenolic stability of wines made from hybrid grapes, which generally have a lower phenolic content (Prajitna et al. 2007). Such levels in Chambourcin wines and other wines produced from hybrid and native winegrape species may require chemical or physical acidification, such as tartaric acid addition or cation exchange, prior to fermentation to achieve microbial and color stability during aging and after bottling (Frost et al. 2017, Ponce et al. 2018).
An alternative to chemical acidification is blending two or more wines together, a practice commonly used to improve acid balance, wine flavor, and consumer acceptance. Blending different wines often produces sensory and chemical attributes that go beyond “simple averaging effects” (Hopfer et al. 2012). For example, Nandorfy et al. (2023) evaluated the effect of blending on chemical composition, sensory attributes, and consumer liking of wines by blending low-proline wines with high-proline wines. Wines blended with increased proportions of proline showed improved viscosity, increased fruity flavors, and enhanced sweetness, as well as decreased perception of bitterness and astringency, resulting in increased consumer acceptance of the higher-proline blended wine compared to the low-proline wine.
Producers can blend low-acidity wines with higher-acidity wines to enhance desirable attributes such as freshness and more-pronounced fruit flavors. However, not all producers have reliable access to high-acid, low-pH wines for such adjustments. A practical and more recent microbiological tool to overcome acidity challenges is the use of native yeasts, such as non-Saccharomyces species, in mixed inoculation with Saccharomyces species to naturally produce acids that will increase total acid content, decrease pH, and improve flavor.
Although the composition of total organic acids in a wine contributes significantly to wine flavor and perceived balance, few studies have evaluated the contribution of individual organic acids on wine flavor and consumer acceptability. Furthermore, the ways in which non-Saccharomyces yeasts influence wine sensory attributes are not well studied compared to Saccharomyces strains (Chidi et al. 2018). In mixed inoculations with Saccharomyces cerevisiae, L. thermotolerans, a non-Saccharomyces yeast, acts as a biological tool to acidify low-acid red wines by converting sugars, primarily glucose and sucrose, into lactic acid, thereby naturally lowering the pH (Schnierda et al. 2014, Morata et al. 2018). Wines made in a style with higher acid production and lower pH can be appealing for wine producers because acidity contributes to complexity, freshness, color, and microbial stability (Mira de Orduña 2010, Vaquero et al. 2020, Vicente et al. 2021). Furthermore, L. thermotolerans produces a range of aromatic and fruity esters that counter the dairy flavors associated with lactic acid produced during MLF (Morata et al. 2019, Vaquero et al. 2021).
In addition to its acidification effects, L. thermotolerans can produce wines with lower ethanol levels than wines fermented with S. cerevisiae alone. Ethanol levels have been shown to be 0.1 to 1.6% lower in wines produced in mixed inoculation with L. thermotolerans (Morata et al. 2003, 2006, Sgouros et al. 2020, Korenika et al. 2021). L. thermotolerans favors respiratory metabolism, in which pyruvate is converted to carbon dioxide instead of ethanol (Pfeiffer and Morley 2014, Vaquero et al. 2021). Therefore, L. thermotolerans has a weaker ability than S. cerevisiae to produce ethanol in aerobic conditions where high glucose levels are present; this is known as the Crabtree effect (Pfeiffer and Morley 2014).
MLF is another method commonly used in winemaking to address acidity issues, by converting malic acid to the softer-tasting lactic acid. Past studies have investigated co-inoculating L. thermotolerans and O. oeni during alcoholic fermentation to produce microbially stable wines. Although an increase in pH after MLF can be conducive to the growth of spoilage organisms, the reduction of malic acid enhances a wine’s microbial stability (Davis et al. 1985, Morenzoni and Spect 2005). While some L. thermotolerans strains promote MLF, other strains are inhibitory, with more than 6 g/L of produced lactic acid inhibiting MLF (Morata et al. 2018, Snyder et al. 2021, Fleming and Threlfall 2024).
It is important to study compatibility between different species of yeast and bacteria in wine production to better understand their effects on wine composition and sensory attributes, especially concerning hybrid grapes used for wine production. Past studies have evaluated the effect of other non-Saccharomyces yeasts such as Hanseniaspora uvarum and Starmerella baccillaris in combination with different lactic acid bacteria strains to reduce the duration of MLF and improve aromas and flavors in wines produced from V. vinifera varieties Shiraz, Tempranillo, and Barbera (du Plessis et al. 2019, Russo et al. 2020, Urbina et al. 2021).
More recently, the use of L. thermotolerans and Saccharomyces yeasts during the first 10 days of fermentation of Arkansas-grown Chambourcin grapes was investigated; results showed a lower pH and higher TA in must/wine while completing alcoholic fermentation (Fleming and Threlfall 2024). The present study continues this work by evaluating the composition and color of Chambourcin wines at bottling and during storage, as well as by assessing the wines’ sensory attributes when produced in mixed inoculation with L. thermotolerans and S. cerevisiae, with and without MLF.
Materials and Methods
Grape harvest
Chambourcin grapes were grown and hand-harvested in 2020 and 2021 at a commercial vineyard in Arkansas (USDA hardiness zone 6b). For wine production, 181 kg of Chambourcin was harvested on 3 Sept 2020, and 168 kg was harvested on 10 Sept 2021. After harvest in both years, grapes were transported to the University of Arkansas System Division of Agriculture (UA System) Food Science Department in Fayetteville, AR for wine production.
Wine production
Chambourcin grapes were randomized into eight 17.5-kg batches in 2020 and eight 19.5-kg batches in 2021 for wine production. A figure showing the treatments and detailed production methods for Chambourcin wine production in 2020 and 2021 appears in Fleming and Threlfall (2024). In both years, each batch of grapes was passed through a crusher/destemmer, and 15 mg/L SO2 as potassium metabisulfite was added to the must (juice, skin, pulp, and seeds) at crushing.
The initial must composition—including total soluble solids (TSS), pH, and TA as tartaric acid—was analyzed prior to fermentation to determine whether adjustments were needed. The TSS of the musts were adjusted to 21% with table sugar (sucrose) in both years.
The 2020 and 2021 Chambourcin wines underwent the following four treatments in duplicate: SC, inoculation with S. cerevisiae Lalvin Persy™ (Lallemand, Inc.); SC-MLF, inoculation with S. cerevisiae followed by addition of malolactic bacteria O. oeni Lalvin VP41™ (Lallemand, Inc.) after 24 hr; LT-SC, inoculation with L. thermotolerans Level2 Laktia (Lallemand, Inc.) followed by inoculation with S. cerevisiae after 48 hr; and LT-SC-MLF, inoculation with L. thermotolerans followed by inoculation with S. cerevisiae after 48 hr, then addition of malolactic bacteria 24 hr after inoculation with S. cerevisiae (Fleming and Threlfall 2024).
Lalvin Persy™ was used because of its desirable properties, such as the nonproduction of SO2, low volatile acidity production, and strong compatibility with malolactic bacteria. VP41™ MLF bacteria was chosen because it is the MBR™ form of lactic acid bacteria from a Lallemand, Inc. process that subjects the lactic acid bacteria to biophysical stresses so it can withstand direct addition to wine. Laktia™ yeast was selected in the Rioja region of Spain from Tempranillo-fermenting must and has been commercially available since 2018. Lallemand recommends waiting 24 to 72 hr after inoculation with L. thermotolerans to add S. cerevisiae. The 48-hr interval was selected because it would provide sufficient time to see an effect on pH and lactic acid production, as compared to the SC control, with the possibility of observing partial or complete MLF. The MLF bacteria was added 24 hr after S. cerevisiae, according to the manufacturer recommendations.
After the musts reached 0 Brix, the treatments were pressed. The wines were collected in glass carboys fitted with fermentation locks filled with SO2 solution to allow carbon dioxide release and limit oxygen. Wines were racked from the sediment into other carboys as needed during fermentation. When alcoholic fermentation was complete (<3.0 g/L residual sugars), the wines were sparged with nitrogen. The wine treatments without MLF were adjusted to 0.8 mg/L molecular SO2 and kept at 21°C.
For treatments inoculated with malolactic bacteria, when MLF was completed or stalled (as was the case for the LT-SC-MLF treatment in both years), the wines were adjusted to 0.8 mg/L molecular SO2. Final malic acid levels were confirmed by high-performance liquid chromatography (HPLC). Wines were stored at 15°C for ~4 mo until bottling in April 2021 (2020 harvest) and April 2022 (2021 harvest). Wines from each treatment were bottled into 125-, 375-, and 750-mL screw-type glass bottles and stored at 15°C until analysis. All bottles were sparged with nitrogen prior to filling with wine.
In both years, wines were analyzed in duplicate for composition and color attributes at bottling (day 0) and during storage at 6 and 12 mo at 15°C. In addition, descriptive and consumer sensory analyses were done on the 2020 wines at 8- to 12-mo storage at 15°C. For composition analysis, the 125-mL bottles were opened, the color attributes were analyzed, and the wine was placed in 15-mL plastic centrifuge tubes and frozen (−10°C).
Composition attribute analyses
The composition attribute analyses included TSS, pH, TA, glycerol, ethanol, residual sugars, volatile acidity, and organic acids. To minimize any potential effect of tartrate precipitation, frozen wine samples were thawed to room temperature and homogenized prior to analysis. Grape juice TSS were measured using an Abbe Mark II refractometer (Bausch and Lomb, Scientific Instruments). An APERA PH700 pH meter was used to measure the pH of the juice and wines. A Metrohm 862 Compact Titrosampler was used to measure the TA of juice and wine, expressed as g/L tartaric acid. Six grams of sample was added to 50 mL degassed, deionized water and titrated with 0.1 N sodium hydroxide to an endpoint of pH 8.2. Wines were degassed prior to analysis.
Glycerol, ethanol, sugars, and organic acids in the wines were identified and quantified according to the HPLC procedure of Walker et al. (2003) and Fleming and Threlfall (2024), using a Waters HPLC system consisting of a 515 HPLC pump, a 717 plus autosampler, and a 410 differential refractometer detector connected in series with a 996 photodiode array (PDA) detector (Waters Corporation). Citric, tartaric, malic, lactic, acetic, and succinic acids were detected at 210 nm by the PDA detector, and glucose, fructose, ethanol, and glycerol were detected at 410 nm by the differential refractometer detector. Analytes in samples were identified and quantified using external calibration curves based on peak area estimation with baseline integration. Total organic acids were calculated as the sum of citric, tartaric, malic, lactic, acetic, and succinic acids. Total residual sugars were calculated as the sum of glucose and fructose. Results for organic acids, sugars, and glycerol were expressed as g/L, and alcohol was expressed as % v/v (alcohol by volume).
Color attribute analyses
The color attributes—including L* (lightness), hue angle, chroma, red color, brown color, and color density—of the 2020 and 2021 wines were evaluated at bottling (day 0) and during storage at 6 and 12 mo at 15°C. L*, hue angle, and chroma were measured using a ColorFlex system (HunterLab). The ColorFlex system uses a ring and disk set to control liquid levels and light interactions for measurement of translucent liquids in a 63.5-mm glass sample cup with an opaque cover to determine the Commission Internationale de l’Eclairage (CIE) L*a*b* transmission values of L* = 100, a* = 0, and b* = 0 (CIE 1986). The vertical axis L* measures lightness from opaque (0) to transparent (100); +a* red, −a* green, +b* yellow, and −b* blue are measured on the hue circle. Hue angle, calculated as
Absorbance values for red color, brown color, and color density were measured using a VWR spectrophotometer UV-1600PC UV-VIS (VWR International, LLC). Red color of wines was measured at 520 nm, brown color was measured at 420 nm, and color density was calculated as red color + brown color (Iland et al. 2021). In both years, samples of Chambourcin wines were diluted 16.67 times with deionized water prior to spectrophotometric analysis and measured against a blank sample of deionized water using a 1-cm cell.
Sensory attribute analysis
The sensory attributes of the 2020 Chambourcin wines were evaluated by a trained descriptive sensory panel and an untrained consumer sensory panel. To minimize sensory fatigue, the two replications of each wine treatment were combined so that 50% of each replicate was used. Because replication variability is typically minimal, the sensory analysis focused on the treatment differences. All sensory analysis of the wines was completed by 12-mo storage at 15°C (~8-mo storage for the untrained consumer sensory analysis and ~12 mo for the descriptive sensory analysis). All wine samples were served at room temperature (21°C) in plastic cups or wine glasses labeled with three-digit codes. The serving order was randomized among panelists to prevent presentation order bias. Panelists were instructed to cleanse their palate between samples with unsalted crackers and water. Panelists evaluated 30-mL samples of each wine, one at a time, and expectorant cups were provided.
Consumer sensory evaluation
The consumer sensory evaluation was conducted at grape and wine industry conferences and tasting events in Arkansas, Texas, New Mexico, and Canada. Participants represented various levels of wine-tasting expertise. A total of 126 consumers evaluated the Chambourcin wines for “liking” attributes (flavor, sourness, astringency, mouthfeel, and overall impression), using a nine-point hedonic liking scale (1 = dislike extremely, 9 = like extremely), and for Just-About-Right (JAR) attributes (flavor, sourness, astringency, and mouthfeel), using a five-point JAR scale (1 = much too low, 3 = just about right, 5 = much too much). After evaluating each wine, consumers were instructed to list one word to describe what they liked and disliked about each wine. After evaluating all four wines, consumers were asked to rank the wines (1 = most preferred, 4 = least preferred).
The consumer panel was comprised of 51% females, 44% males, and 5% undisclosed. In terms of age, 19% were 21 to 30-yr-old, 15% were 31 to 39, 18% were 40 to 49, 20% were 50 to 59, 21% were 60 to 69, 5% were 70 or older, and 2% were undisclosed. Forty-five percent of the consumers indicated that they currently or previously worked in the grape/wine industry, 48% had never worked in the industry, and 7% were undisclosed or other. Forty-eight percent of the consumers had previously consumed Chambourcin wine, 50% had not, and 2% were undisclosed.
Descriptive sensory evaluation
The descriptive sensory evaluation was conducted at the UA System Sensory Science Center, Fayetteville, AR, by descriptive panelists who work for the Sensory Science Center and have experience evaluating various products at 6 hr/wk. The panel was trained to use the Sensory Spectrum method, an objective method that uses references to describe the intensity of attributes in products (Meilgaard et al. 2007).
During training and practice sessions for this project, the panelists (n = 7) developed a lexicon of sensory terms (with definitions) for the Chambourcin wines through consensus (Table 1). In a 3-hr session, wines were served monadically (one at a time) in random order, then again (i.e., duplicates) after a break. The descriptive panel evaluated the wines for aroma (fruity), aromatics (fruity, bell pepper, and acetone), basic tastes (sour and bitter), and mouthfeel (astringency and finish at three time points: 0, 7.5, and 15 sec), using a scale in which 0 = less of an attribute and 15 = more of an attribute. Basic taste and mouthfeel standards were provided to the panelists.
Lexicon created and used by the trained descriptive sensory panel (n = 7) to evaluate the aroma, aromatics, basic tastes, and mouthfeel attributes of Chambourcin wine made from Arkansas-grown grapes harvested in 2020.
Design and statistical analysis
Statistical analyses were conducted using JMP® Pro statistical software (ver. 17.0.0, SAS Institute, Inc.). A univariate analysis of variance (ANOVA) was used to determine the significance of the main factors. Tukey’s honestly significant difference (HSD) test was used to detect differences among means (p < 0.05). The relationships among composition, color, and sensory data were assessed through multivariate analysis using Pearson’s correlations and pairwise comparisons. The 12-mo composition and color data were averaged and used for analysis of correlation to the sensory data.
Wine composition and color at bottling and during storage (2020 and 2021)
An ANOVA was used to determine the significance of the main factors within each storage period, and Tukey’s HSD test was used to detect differences among means (p < 0.05).
Wine sensory attributes (2020 Chambourcin)
For the consumer and descriptive evaluations, a univariate ANOVA was used to determine the significance of the main effects, and Tukey’s HSD was used to detect significant differences among the means (p < 0.05). For the descriptive sensory analysis, the “panelist” main effect was included in the model to account for error caused by between-panelist variation. The inoculation treatment × panelist interaction was not significant (p > 0.05) for any of the descriptive attributes, indicating that the panelists were consistent in their ratings and thus the training of the descriptive panel was adequate (Biasoto et al. 2014).
Results and Discussion
After harvest in 2020 and 2021, the Chambourcin musts had similar pH (3.4 and 3.5, respectively), TA (7.7 and 7.3 g/L, respectively), and TSS (19.5 and 18.8%, respectively). In both years, TSS were adjusted to 21%. Fleming and Threlfall (2024) evaluated the inoculation treatments daily during fermentation and found that from day 0 to day 5, LT-SC and LT-SC-MLF treatments had reductions in pH and malic acid, while TA and lactic acid increased. To further demonstrate the effect of inoculation treatment type, the composition, color, and sensory analyses of these Chambourcin wines at bottling (0 mo) and during storage (6 and 12 mo) are presented herein.
Wines were analyzed at bottling and during storage for composition attributes (sugars, ethanol, glycerol, pH, TA, and organic acids) and color attributes (L*, hue angle, chroma, red color, brown color, and color density) in both years (Tables 2 to 5 and Figures 1 to 4), and for sensory attributes in 2020 (Tables 6 to 8 and Figures 4 to 6). At bottling, glucose, fructose, and total residual sugars were <3.0 g/L for all wines (data not reported); all wines completed alcoholic fermentation, resulting in ethanol levels between 10.1% and 11.3%. All wines had pH values ≤ 3.75, which is similar to that found for other Chambourcin wines (pH 3.2 to 3.8) (Prajitna et al. 2007, Homich et al. 2016, Mayfield et al. 2021).
Effects of inoculation treatment on composition attributesa of wines produced from Chambourcin grapes grown in Arkansas in 2020 and evaluated at bottling (0 mo) and during storage (6 and 12 mo at 15°C). Treatments were evaluated in duplicate.
Effects of inoculation treatment on composition attributesa of wines produced from Chambourcin grapes grown in Arkansas in 2021 and evaluated at bottling (0 mo) and during storage (6 and 12 mo at 15°C). Treatments were evaluated in duplicate.
Effects of inoculation treatments on color attributesa of wines produced from Chambourcin grapes grown in Arkansas in 2020 and evaluated at bottling (0 mo) and during storage (6 and 12 mo at 15°C). Treatments were evaluated in duplicate.
Effects of inoculation treatment on color attributesa of wines produced from Chambourcin grapes grown in Arkansas in 2021 and evaluated at bottling (0 mo) and during storage (6 and 12 mo at 15°C). Treatments were evaluated in duplicate.
Effect of storage times (at 15°C) on pH of wines produced by inoculation treatments from Chambourcin grapes grown in Arkansas in 2020 and 2021. SC, inoculation with Saccharomyces cerevisiae; SC-MLF, inoculation with S. cerevisiae followed by addition of malolactic bacteria after 24 hr; LT-SC, inoculation with Lachancea thermotolerans followed by inoculation with S. cerevisiae after 48 hr; LT-SC-MLF, inoculation with L. thermotolerans followed by inoculation with S. cerevisiae after 48 hr, then addition of malolactic bacteria 24 hr after inoculation with S. cerevisiae. Means with different letters are significantly different (p < 0.05) for each attribute within a treatment according to Tukey’s honestly significant difference test.
Effect of storage times (at 15°C) on total organic acids, lactic acid, and malic acid of wines produced by inoculation treatments from Chambourcin grapes grown in Arkansas in 2020 and 2021. SC, inoculation with Saccharomyces cerevisiae; SC-MLF, inoculation with S. cerevisiae followed by addition of malolactic bacteria after 24 hr; LT-SC, inoculation with Lachancea thermotolerans followed by inoculation with S. cerevisiae after 48 hr; LT-SC-MLF, inoculation with L. thermotolerans followed by inoculation with S. cerevisiae after 48 hr, then addition of malolactic bacteria 24 hr after inoculation with S. cerevisiae. Means with different letters are significantly different (p < 0.05) for each attribute within a treatment according to Tukey’s honestly significant difference test.
Effect of storage times (at 15°C) on L* of wines produced by inoculation treatments from Chambourcin grapes grown in Arkansas in 2020 and 2021. SC, inoculation with Saccharomyces cerevisiae; SC-MLF, inoculation with S. cerevisiae followed by addition of malolactic bacteria after 24 hr; LT-SC, inoculation with Lachancea thermotolerans followed by inoculation with S. cerevisiae after 48 hr; LT-SC-MLF, inoculation with L. thermotolerans followed by inoculation with S. cerevisiae after 48 hr, then addition of malolactic bacteria 24 hr after inoculation with S. cerevisiae. Means with different letters are significantly different (p < 0.05) for each attribute within a treatment according to Tukey’s honestly significant difference test.
Effect of storage times (at 15°C) on brown color, red color, and color density of wines produced by inoculation treatments from Chambourcin grapes grown in Arkansas in 2020 and 2021. SC, inoculation with Saccharomyces cerevisiae; SC-MLF, inoculation with S. cerevisiae followed by addition of malolactic bacteria after 24 hr; LT-SC, inoculation with Lachancea thermotolerans followed by inoculation with S. cerevisiae after 48 hr; LT-SC-MLF, inoculation with L. thermotolerans followed by inoculation with S. cerevisiae after 48 hr, then addition of malolactic bacteria 24 hr after inoculation with S. cerevisiae. Means with different letters are significantly different (p < 0.05) for each attribute within a treatment according to Tukey’s honestly significant difference test.
Attributes evaluated by a consumer sensory panel (126 panelists) in 2020, using a nine-point hedonic scalea for wines produced from Chambourcin grapes with different inoculation treatments and stored for 8 to 12 mo at 15°C.
Percentage (%) of responses in the consumer sensory analysis (wines evaluated by 126 panelists) using a collapsed five-point just-about-right (JAR)a scale (collapsed to Too low, JAR, and Too much) for wines produced in 2020 from Chambourcin grapes with different inoculation treatments and stored for 8 to 12 mo at 15°C.
Attributes evaluated by a descriptive sensory panel (seven panelists) using a 15-point scalea for wine produced in 2020 from Chambourcin grapes with different inoculation treatments and stored at 15°C for nine months. Treatments were evaluated in duplicate.
Consumers (n = 126) ranked the Chambourcin wines made with different inoculation treatments in 2020. Wines were evaluated at 8- to 12-mo storage at 15°C. Ranking: 1 = most preferred to 4 = least preferred. SC, inoculation with Saccharomyces cerevisiae; SC-MLF, inoculation with S. cerevisiae followed by addition of malolactic bacteria after 24 hr; LT-SC, inoculation with Lachancea thermotolerans followed by inoculation with S. cerevisiae after 48 hr; LT-SC-MLF, inoculation with L. thermotolerans followed by inoculation with S. cerevisiae after 48 hr, then addition of malolactic bacteria 24 hr after inoculation with S. cerevisiae.
Word cloud for descriptors given by consumers (n = 126) to describe the aroma, flavor, and mouthfeel of their first-ranked Chambourcin wines. The wines were made with different inoculation treatments in 2020 and evaluated at 8- to 12-mo storage at 15°C. SC, inoculation with Saccharomyces cerevisiae; SC-MLF, inoculation with S. cerevisiae followed by addition of malolactic bacteria after 24 hr; LT-SC, inoculation with Lachancea thermotolerans followed by inoculation with S. cerevisiae after 48 hr; LT-SC-MLF, inoculation with L. thermotolerans followed by inoculation with S. cerevisiae after 48 hr, then addition of malolactic bacteria 24 hr after inoculation with S. cerevisiae.
Wine composition and color at bottling and during storage (2020 and 2021)
Most Chambourcin wines in 2020 and 2021, fermented in mixed inoculation with L. thermotolerans and regardless of MLF treatment, had lower pH values and higher TA, lactic acid, and total organic acids. The composition attributes of the wines were affected by mixed fermentation with L. thermotolerans, and lactic acid was the predominant acid in most wines at bottling (Tables 2 and 3). TA in the LT-SC and LT-SC-MLF wines was higher (>12 g/L) than is typical for wines.
While the LT-SC-MLF wines struggled to go through MLF (<0.3 g/L considered dry) compared to the SC-MLF wines, SC-MLF wines likewise did not complete MLF. At the time of bottling, the SC-MLF wines had undergone partial MLF; research constraints required the wines to be bottled at this point, therefore completion of MLF was not possible. SC-MLF and LT-SC-MLF wines at bottling had malic acid levels of 0.47 to 0.65 g/L and 1.69 to 3.11 g/L, respectively, and lactic acid values of 3.61 to 3.97 g/L and 6.43 to 6.71 g/L, respectively. Average malic and lactic acid levels of the must prior to fermentation were 4.2 g/L and 1.8 g/L, respectively. It is likely that MLF did not complete in the LT-SC-MLF wines due to high lactic acid levels, a product of L. thermotolerans fermentation. Successful completion of MLF can be hindered when lactic acid values are >6.0 g/L, which can inhibit performance of lactic acid bacteria (Morata et al. 2018, Snyder et al. 2021).
Chambourcin wines at bottling had the following color attributes: L* = 0.55 to 1.98, hue angle = 360.12 to 360.23°, chroma = 2.49 to 9.08, red color = 3.70 to 8.83, brown color = 2.05 to 3.98, and color density = 5.90 to 12.81 (Tables 4 and 5). Wines fermented in mixed inoculation with L. thermotolerans, regardless of MLF, had more red color. The higher lactic acid and lower pH levels in the LT-SC and LT-SC-MLF wines compared to the SC and SC-MLF wines favored higher red color for both years. The effects of storage time on the acidity and color attributes of wines from a given inoculation treatment varied in both years (Figures 1 to 4). Notably, total organic acids increased during storage for the LT-SC and LT-SC-MLF wines in 2020 but decreased during storage for the same wines in 2021. In 2020, the SC-MLF and LT-SC wines darkened, whereas red color and color density decreased for the SC-MLF wines. In 2021, only SC-MLF darkened, but brown color and color density increased for all inoculation treatments. The effect of fermentation treatments on composition and color attributes for the 2020 and 2021 wines will be discussed by attributes.
Ethanol and glycerol
Ethanol levels were not affected by inoculation treatment at bottling or during storage in both years, whereas the effect on glycerol varied (Tables 2 and 3). Although treatments fermented in mixed inoculations with L. thermotolerans and S. cerevisiae produced ethanol levels lower than in wines fermented with S. cerevisiae alone, the difference was not significant. Hranilovic et al. (2021) showed that Merlot wines produced using sequential inoculation with commercial strains of L. thermotolerans (Laktia™ and Concerto™) with S. cerevisiae had 0.9% lower ethanol compared to control wines fermented with S. cerevisiae alone. However, ethanol levels of Merlot produced from mixed inoculations using the Levulia® strain of L. thermotolerans were similar to those of the S. cerevisiae control (Hranilovic et al. 2021).
Glycerol, a by-product of alcoholic fermentation, contributes to body and mouthfeel in wines. Although the average glycerol content of wines produced in the U.S. is ~7.2 g/L, levels can reach 15 g/L (Amerine and Ough 1974, Scanes et al. 1998). In a previous study, wines produced from mixed fermentations with Laktia™ had 10 g/L glycerol, whereas the Concerto™ strain produced wines with up to 11.6 g/L glycerol (Hranilovic et al. 2021).
2020
Inoculation treatment did not affect ethanol at any storage time point (0 mo = 10.85%, 6 mo = 10.57%, and 12 mo = 10.73%). Glycerol levels did not differ among the inoculation treatment groups at 0- (13.08 g/L) or 6-mo storage (12.66 g/L). However, at 12-mo storage, regardless of MLF, glycerol levels were higher in the LT-SC treatments (7.42 to 7.84 g/L) compared to the SC treatments (5.84 to 6.14 g/L) (Table 2).
Glycerol was affected during storage for each inoculation treatment: levels were the same within each treatment group at 0- and 6-mo storage, but decreased at 12-mo storage (data not shown). The decrease could be a result of oxidation of glycerol, which leads to the formation of glyceraldehyde and dihydroxyacetone, two products shown to increase absorbance values in wine at 420 and 520 nm (Laurie and Waterhouse 2006).
2021
Inoculation treatment did not affect ethanol at any storage time point (0 mo = 10.53%, 6 mo = 11.72%, and 12 mo = 11.78%, average of treatments at each storage time). Inoculation treatment affected glycerol levels at 0- and 12-mo storage. At 0- and 12-mo, regardless of MLF, LT-SC wines had higher glycerol levels (13.04 to 13.21 g/L and 12.66 to 12.69 g/L, respectively) compared to SC wines (11.05 to 11.52 g/L and 10.99 to 11.36 g/L, respectively). However, inoculation treatment did not affect glycerol levels at 6-mo storage (10.24 g/L).
In contrast to the results from 2020, glycerol levels generally remained consistent within the SC and SC-MLF treatment groups. However, within the LT-SC and LT-SC-MLF treatment groups, glycerol levels decreased from 0- to 6-mo storage, then increased at 12-mo storage.
pH and TA
Inoculation treatment affected pH and TA at each storage time in both years (Tables 2 and 3). In general, the pH was lower and the TA was higher in the LT-SC wines than in the SC wines, regardless of MLF. Although red table wines have a TA between 4.0 and 7.0 g/L and a pH between 3.3 and 3.6 (Margalit 1997), Chambourcin wines typically have a pH of 3.2 to 3.8 and a TA range of ~6.0 to 8.5 g/L, depending on growing conditions (Prajitna et al. 2007, Homich et al. 2016). Regardless of MLF, the higher TA levels in the LT wines compared to the SC wines were most likely caused by the higher levels of lactic and malic acids in the LT wines compared to the SC wines. Although TA is largely composed of tartaric acid, it also includes citric, malic, lactic, succinic, and acetic acids. In a survey of 277 Australian wines over a period of five vintages, lactic and malic acids together comprised up to 30% of the acid composition in the TA of red wines (Wilkes 2016).
2020
The pH of the SC-MLF wines (3.52 to 3.60) was higher than that of the other treatment groups at each storage time. At 6-mo storage, the SC (pH = 3.52) and SC-MLF (pH = 3.57) wines had a higher pH than the LT-SC (pH = 3.39) and LT-SC-MLF (pH = 3.33) wines. Regardless of MLF, the TA of the SC wines was lower than that of the LT-SC wines at each storage time. TA of the SC and SC-MLF wines ranged from 6.80 to 7.78 g/L, whereas the TA of the LT-SC and LT-SC-MLF wines ranged from 11.53 to 12.73 g/L. Figure 1 shows how pH was affected during storage for each inoculation treatment. For most treatments during storage, pH did not change. However, the pH of the LT-SC wines increased during storage, from 0- (3.29) to 6-mo (3.39). TA did not change within each treatment during storage (data not shown).
2021
Regardless of MLF treatment, the LT wines had lower pH and higher TA values (pH = 3.29 to 3.40, TA = 11.84 to 13.06 g/L) compared to the SC wines (pH = 3.69 to 3.85, TA = 6.30 to 7.50 g/L). Figure 1 shows how pH was affected during storage for each inoculation treatment. pH decreased from 0- to 12-mo storage for the SC and LT-SC-MLF treatments from 3.75 to 3.69 and from 3.40 to 3.29, respectively. TA did not change within each treatment during storage (data not shown).
Organic acids
In both years, most organic acids were affected by the inoculation treatment (Tables 2 and 3). However, tartaric, citric, and succinic acid levels were less affected than lactic, malic, and acetic acid levels. The level of acetic acid was below the sensory threshold for red wines (0.74 g/L) (Corison et al. 1979). Lactic acid, the predominant acid at bottling and during storage for most inoculation treatments, was produced from glucose during fermentation by L. thermotolerans in the LT-SC wines or during MLF in the SC-MLF wines through enzymatic decarboxylation of L-malic acid to L-lactic acid and carbon dioxide (Knoll et al. 2012). Lactic acid production in the LT-SC-MLF wines was a product of glucose metabolism by L. thermotolerans and possibly also through MLF. The predominance of lactic acid in the SC wines, although lower than in the other treatments, was most likely a by-product of yeast metabolism during alcoholic fermentation (Margalit 1997). In general, in both years and regardless of MLF treatment, total organic acid levels were higher in the LT-SC wines compared to the SC wines.
2020
Tartaric acid levels were not affected by inoculation treatment at each storage time, averaging 1.94, 1.79, and 2.86 g/L at 0-, 6-, and 12-mo storage, respectively. At all storage times, LT-SC wines had higher malic and lactic acid levels compared to SC wines, regardless of MLF treatment.
At 0-mo storage, citric and succinic acid levels were affected, with higher levels in the LT-SC-MLF wines (succinic acid = 0.82 g/L, citric acid = 0.46 g/L) than in the SC-MLF wines (succinic acid = 0.70 g/L, citric acid = 0.06 g/L). At 6- and 12-mo storage, succinic (0.90 and 0.40 g/L, respectively) and citric acid (1.06 and 0.35 g/L, respectively) levels were similar among all inoculation treatments. L. thermotolerans has been reported to produce succinic acid in high concentrations in wines, with an average of 0.5 g/L, compared to no succinic acid production in a S. cerevisiae control wine (Binati et al. 2019).
Succinic acid, a product of the tricarboxylic acid cycle, is the primary acid produced during alcoholic fermentation and contributes bitterness and sourness to wines (Boulton et al. 1996). Amerine et al. (1979) reported detection threshold levels for succinic acid at 0.035 g/L, much lower than the levels at bottling for these Chambourcin wines in both years (0.54 to 1.06 g/L). The tricarboxylic acid cycle is favored by L. thermotolerans (Vicenti et al. 2021); thus, higher levels of succinic acid can result in wines fermented by mixed inoculation with L. thermotolerans.
Malic, lactic, and total organic acid levels did not change during storage for any inoculation treatment, with the exception of LT-SC wines, which showed an increase from 0- (10.81 g/L) to 12-mo (16.91 g/L) for total organic acids (Figure 2).
2021
Tartaric levels at 0-mo storage were higher in the SC wines (3.31 g/L) compared to all other wines (2.38 to 2.67 g/L), but no effect was observed at 6- or 12-mo storage. At 6-mo storage, succinic acid levels were lower in the LT-SC wines, regardless of MLF (0.61 to 0.73 g/L), compared to the SC wines (0.97 to 1.06 g/L). At 12-mo storage, citric acid levels were higher in the SC wines (0.28 to 0.31 g/L) compared to the LT wines (0.19 g/L). In general, malic and lactic acid levels were higher in the LT wines, compared to SC wines, regardless of MLF. However, at month 12, malic acid levels were the same among all treatments, averaging 1.14 g/L. Acetic acid levels were higher in the SC and SC-MLF wines at 6-mo storage, but lower at 12-mo storage.
Figure 2 shows how malic, lactic, and total organic acids were affected during storage within each inoculation treatment. Malic acid decreased for the LT-SC and LT-SC-MLF wines from 0- to 12-mo storage. All treatments except SC showed decreases in total organic acids during storage, mainly resulting from decreases in succinic, lactic, malic, and citric acids.
L*, hue angle, and chroma
The effect of inoculation treatment on a wine’s color attributes varied with storage time. The L* (0.55 to 1.98) and chroma (2.49 to 9.08) ranges were lower than those reported by Mayfield (2020) for Chambourcin wines (L* = 7.2 to 8.4, chroma = 19 to 36), whereas the hue angle (360.00 to 360.23°) was within the previously reported range (hue angle = 360°).
2020
The LT-SC wines, regardless of MLF, were generally darker (lower L*) and had lower chroma values compared to the SC wines at each storage time (Figure 3). Lower levels of chroma indicate less saturation, resulting in lower color intensity or concentration (Hernanz et al. 2009, Fan et al. 2023). Despite the difference in L* between the Saccharomyces and non-Saccharomyces treatments, the values were below the range previously reported for Chambourcin wines produced from grapes in Arkansas. From 0- to 12-mo storage, L* decreased for the SC-MLF and LT-SC wines (from 1.98 to 1.35 and 1.06 to 0.87, respectively), possibly an effect of the higher lactic acid levels on wine color (Figure 3). The LT-SC-MLF wines did not darken from 6- to 12-mo storage. Because of its ability to produce L(+) lactic acid during fermentation, L. thermotolerans can reduce wine pH, thereby enhancing wine color (Benito 2020, Vicente et al. 2021).
2021
L* was lower in the SC-MLF wines (0.75) compared to the SC and LT-SC-MLF wines (1.14 and 1.11, respectively) at 6-mo storage. Inoculation treatment did not affect hue angle at any storage time. At 0-mo storage, chroma was higher in the SC-MLF wines (3.56) compared to all other wines (2.49 to 2.84). Figure 3 shows how L* was affected during storage for each inoculation treatment. The SC-MLF wines darkened as L* decreased from 0 to 12 mo. In the SC, LT-SC, and LT-SC-MLF wines, L* decreased from 6 to 12 mo.
Red color, brown color, and color density
In both years, regardless of MLF, the LT-SC wines tended to have higher red color, brown color, and color density than the SC wines (Tables 4 and 5). Red color measurements in both years and all storage times ranged from 3.70 to 8.83, which is similar to the 3 to 6 range measured previously (Auw et al. 1996) for Chambourcin wines produced from grapes grown in Georgia, and the 4 to 6 range measured by Mayfield (2020) for Chambourcin wines produced from grapes grown in Arkansas. The color density of the wines ranged from 5.90 to 12.81, and brown color ranged from 2.05 to 4.90. Previously reported color density values ranged from 7.0 to 8.9 (Mayfield 2020).
More red color was observed in the wines fermented in mixed inoculation with L. thermotolerans regardless of MLF. The higher lactic acid and lower pH levels in the LT-SC and LT-SC-MLF wines compared to the SC and SC-MLF wines favored higher red color for both years. Previous research (Benito et al. 2017) showed that wines produced from mixed fermentation with L. thermotolerans had higher levels of vitisins A and B (malvidin-3-O-glucoside-pyruvate acid and malvidin-3-O-glucoside-4 vinyl, respectively), also known as pyranoanthocyanins, a class of compounds that form stable pigments during alcoholic fermentation and contribute to improved red wine color and color stability during storage (Rentzsch et al. 2007).
In addition, lactic acid production has been shown to improve intensity of color, and decreases in pH can favor the formation of stable pigment complexes that enhance wine color (Morata et al. 2003, 2006, Benito 2020). Due to increases in organic acids, notably lactic acid and the reduction in pH, sequential inoculation with L. thermotolerans and S. cerevisiae can improve wine color intensity. A decrease in pH favors the formation of stable pigment complexes, specifically affecting the flavylium ion, an important component of anthocyanin color intensity (Benito et al. 2017, Benito 2020). Past research has shown ~10% increase in color intensity and total anthocyanin content in wines produced with L. thermotolerans in mixed fermentation with S. cerevisiae, compared to the S. cerevisiae control wine (Benito et al. 2015). Although the color intensity of wines is dependent on the strain of L. thermotolerans used and its anthocyanin absorption abilities (Benito 2018).
Combined use of MLF can also result in color degradation through loss of polymeric pigments, anthocyanin absorption, and a reduction in visitin B formation, which is highly dependent on L. thermotolerans strain variability (Burns and Osborne 2013, 2015, Strickland et al. 2016, Wang et al. 2018, Benito 2020). In contrast to prior research, this study’s wines treated with L. thermotolerans, with or without MLF, had a stronger red color, greater color intensity, and increased polymeric pigments (higher brown color), which form in the first two years after wine production and are mostly derived from monomeric anthocyanins and tannins in the wine (Waterhouse et al. 2016).
2020
Inoculation treatment affected red color, brown color, and color density at 6- and 12-mo storage, with the LT-SC wines, regardless of MLF, having higher red color, brown color, and color density compared to the SC wines (Figure 4). At 12 mo, brown color was higher in LT-SC wines compared to SC and SC-MLF wines, but not LT-SC-MLF wines, suggesting that the use of MLF in wines fermented with L. thermotolerans decreased formation of polymeric pigments. This can be disadvantageous for wines during aging because these pigments are critical for color stability in red wines and are resistant to SO2 bleaching (Burns and Osborne 2015). Red color, brown color, and color density tended to decrease over time within each inoculation treatment during storage, but this effect was not significant.
2021
The LT-SC wines, regardless of MLF, had higher red color, brown color, and color density compared to the SC wines at each storage time point. Brown color and color density increased during storage for all inoculation treatments from 0- to 12-mo storage, whereas red color increased for the SC wines (Figure 4).
Wine sensory attributes (2020)
So that treatment effects could be evaluated, the final wines were not adjusted at bottling (i.e., sugar or acid adjustments), as would typically be done for wines in a commercial setting. Sensory panelists and participants were notified that the wines were not commercially finished and were instructed to focus on the attributes in the ballots. For the consumer sensory evaluation, untrained consumers evaluated liking and JAR attributes, ranked the 2020 Chambourcin wines, and listed one word describing what they liked/disliked about each wine (Tables 6 and 7). For the descriptive sensory analysis, trained panelists evaluated the intensity of 2020 Chambourcin wines’ aroma, aromatics, basic tastes, and mouthfeel (Table 8). In general, the sensory evaluations showed that inoculation with L. thermotolerans resulted in a wine that was more sour and astringent, corresponding with the changes in the acidity attributes of the wines caused by L. thermotolerans. At 12-mo storage, the LT-SC (7.5 g/L) and LT-SC-MLF (7.1 g/L) wines had higher lactic acid (the predominant acid) than the SC (3.3 g/L) and SC-MLF (4.2 g/L) wines. Panelists were sensitive to the sour attribute in the LT-SC and LT-SC-MLF wines, perceiving these wines as more sour than the SC and SC-MLF wines.
Consumer sensory evaluation
Untrained consumers (n = 126) evaluated liking attributes (flavor, sourness, astringency, mouthfeel, and overall impression) and JAR attributes (flavor, sourness, astringency, and mouthfeel) of Chambourcin wines that had been stored for 8 to 12 mo. Inoculation treatments affected all liking attributes of the Chambourcin wine (Table 6). Liking attributes ranged from 3.6 (dislike moderately) to 5.6 (neither like nor dislike). Sensory panelists were sensitive to differences in flavor, sourness, and overall impression between the SC and LT-SC wines, regardless of MLF, and preferred the SC and SC-MLF wines over the LT-SC and LT-SC-MLF wines for all attributes. The higher malic acid, lactic acid, TA, and lower pH values in the LT-SC and LT-SC-MLF wines negatively affected the panelists’ liking of acidity. In addition, the lower ethanol levels in all the treatments, compared to levels typical for red wines, may have enhanced the perception of sourness in the LT-SC and LT-SC-MLF wines.
For all JAR attributes, a higher percentage of consumers scored the SC and SC-MLF wines as JAR compared to the LT-SC and LT-SC-MLF wines (Table 7). For the SC and SC-MLF wines, the JAR percentages were 54% and 52% for flavor, 45% and 52% for sourness, 65% and 62% for astringency, and 58% and 55% for mouthfeel, respectively. In most attributes, the SC-MLF wines had lower JAR percentages than the SC wines, except for sourness, where 52% of panelists rated SC-MLF as JAR compared to the SC wines (45%). This is most likely due to the effect of MLF on reducing malic and tartaric acid levels while improving mouthfeel, aroma, and overall complexity in wines (Lonvaud-Funel 1995, Knoll et al. 2012).
Consumers scored wines made with L. thermotolerans as “too much” in terms of sourness (both LT-SC and LT-SC-MLF = 78%). As is indicated by the JAR results for these wines, higher acidity levels are a notable feature of wines fermented with L. thermotolerans due to production of L(+) lactic acid by the yeast during alcoholic fermentation (Kurtzman 2003, Comitini et al. 2011, Snyder et al. 2021). Moreover, there was a ≥32% difference in consumer JAR of the wines in the “too much” category for sourness intensity of LT-SC and LT-SC-MLF treatments, compared to SC and SC-MLF treatments.
In ranking the wines from most to least preferred, 41% of the consumers ranked the SC-MLF wines first (most preferred), 32% ranked the SC wines second, 40% ranked the LT-SC wines third, and 44% ranked the LT-SC-MLF wines fourth (least preferred) (Figure 5). Although the LT-SC and LT-SC-MLF wines were ranked lowest by most panelists, those panelists who ranked these wines as their highest perceived them as having a “good balance of sourness with flavor,” being “smooth and full bodied,” and “tasted best”; others liked the “sourness” or thought the “tart flavor was just right”. It is important to note this because wines fermented with L. thermotolerans or L. thermotolerans in combination with MLF were acceptable to some consumers.
It is well established that wines with excessive acidity will taste sour for most consumers, whereas wines with too little acidity will taste flabby or dull, with less defined aromas and flavors (Mato et al. 2005, Chidi et al. 2018). Correlation analysis revealed that the lower concentration of acids (specifically, malic and lactic acids) in the SC and SC-MLF wines was correlated with increased liking for most sensory attributes, whereas the same attributes in the LT-SC and LT-SC-MLF wines resulted in decreased consumer liking for flavor, sourness, and overall impression. Achieving a good sugar/acid balance is critical for consumer acceptance but can be a challenge for winemakers, especially in cool-climate grapegrowing regions (Chidi et al. 2018). Sugar is commonly added to wine prior to bottling to create sugar/acid balance, particularly to mediate sourness in high-acid wines (Nordeloos and Nagel 1972, Baldy 1997). Such back-sweetening could be an acceptable solution to the perception of sourness in wines produced with L. thermotolerans.
Flavor, sourness, astringency, mouthfeel, and overall impression liking attributes were negatively correlated (r ≥ −0.93) with TA, glycerol, lactic acid, and total organic acids, whereas the same attributes were positively correlated (r ≥ 0.96) with pH (Supplemental Table 1). Flavor, sourness, and overall impression were negatively correlated (r ≥ −0.96) with malic acid. As glycerol, TA, malic, lactic, and total organic acid levels increased in the wines, consumer liking decreased for most attributes, while higher pH correlated with increased liking.
Malic acid is the main contributor to harsh sourness in wine, with levels >15 g/L considered to be excessive; such levels are sometimes found in grapes grown in cool climates (Gallander 1977). The malic acid levels in the LT wines, regardless of MLF treatment, ranged from 2.1 to 4.3 g/L at time of sensory evaluation, and the lactic acid levels ranged from 4.4 to 7.5 g/L. Typically, finished wines will have either high malic acid or high lactic acid, not both. Although MLF is commonly used to decrease malic acid and its corresponding sourness, the combination of both lactic and malic acid in the LT wine treatments, ranging from 6.5 to 11.8 g/L, created an excessively sour wine, negatively affecting consumer liking. Furthermore, glycerol, typically associated with positive mouthfeel and flavor attributes in wines, contributed to consumer disliking of the LT wines as a result of its strong relationship to astringency, an oral sensation causing dryness or puckering (Breslin et al. 1993). Overall, astringency liking decreased as glycerol levels increased in the LT-SC and LT-SC-MLF wines.
All of the consumer sensory attributes were positively correlated with the color attributes of L*, hue angle, and chroma (≥0.95). Regardless of MLF treatment, the SC wines had higher L* (lighter wine color) and chroma (color intensity) compared to the LT wines. Prior to consumption, consumers judge the quality of food and beverage products primarily by their color (Lawless and Heymann 2010). Wine is no different in that color is a critical quality attribute for the consumer. Higher red color and intensity in young red wines are often positively correlated with flavor and quality (Jackson et al. 1978). Parpinello et al. (2009) showed that most inexpert consumers preferred high color intensity in Italian Novello red wines, a type of wine produced through carbonic maceration and consumed shortly after production. For everyday wine consumers, higher color intensity is related to higher aroma intensity and wine structure.
Lastly, each consumer was asked to provide one word describing what they liked best about each wine treatment. A word-cloud for each treatment was generated to visualize the consumer sensory differences among the four wine treatments (Figure 6). The top three attributes for each treatment, from most to least, were: SC: flavor, fruity, and mouthfeel; SC-MLF: mouthfeel, flavor, and aroma; LT-SC: color, flavor, and mouthfeel; and LT-SC-MLF: color, mouthfeel, and acidity. “Mouthfeel” was listed among the top three of a liked attribute in all the wine treatments. “Cherry” was listed as a liked attribute among all treatments except the LT-SC wine. The term “color” was predominantly listed for the LT-SC and LT-SC-MLF wines, whereas “flavor” was a pronounced descriptor for the SC and SC-MLF wines.
Descriptive sensory evaluation
The descriptive sensory panel (n = 7) evaluated the 2020 Chambourcin wines at 9-mo storage for wine aroma (fruity), aromatics (fruity, bell pepper, and acetone), basic tastes (sour and bitter), and mouthfeel (astringency and finish). Inoculation treatments did not affect aroma, aromatics, basic taste (bitter), or the mouthfeel attributes of astringency and finish at 0 and 7.5 sec. Wines fermented with L. thermotolerans can have fruitier attributes and increased aromatic complexity than their counterparts, due to higher concentrations of volatile aroma compounds, such as ethyl esters and terpenes (Benito et al. 2016, Hranilovic et al. 2018a, Binati et al. 2020, Vejarano and Gil-Calderón 2021). However, the results of the descriptive sensory panel showed no difference among treatments for fruity aroma (5.79 to 6.82) or aromatics (5.51 to 6.63).
Inoculation treatments did affect sourness and the mouthfeel attribute of finish (15 sec) (Table 8). The LT-SC-MLF wine had higher sourness (10.96) than the SC wine (8.20) and the SC-MLF wine (8.35), but not the LT-SC wine (9.98). Wines fermented with L. thermotolerans in combination with MLF were perceived as more sour, as was reported in the consumer evaluations. For finish-15 sec, the LT-SC-MLF wines (5.45) were rated higher than the SC wines (4.53), but not the SC-MLF and LT-SC wines (4.73 and 4.86, respectively). Finish, or aftertaste, refers to the taste, odor, and tactile sensations that remain after a wine is swallowed, possibly caused by volatile compounds that take longer to volatilize after consumption.
Descriptive sourness of the wine was positively correlated to TA and glycerol (r ≥ 0.97) but negatively correlated to pH (r = −0.96) (Supplemental Table 1). TA, glycerol, and pH levels in the LT-SC-MLF wines increased panelists’ sourness ratings for those wine treatments, which was not the case for the SC and SC-MLF wines. Contrary to glycerol’s typically positive effect on mouthfeel in wines, the higher glycerol levels in the LT-SC-MLF wines were associated with higher sourness ratings. This is most likely due to higher glycerol levels commonly reported in wines fermented in mixed inoculation with L. thermotolerans, which also contain higher levels of lactic acid than control wines because of the yeast’s sugar metabolism (Hranilovic et al. 2018b, Fleming and Threlfall 2024). In this case, glycerol did not mask the sourness in the LT wines, regardless of MLF treatment, even though glycerol production was significantly higher than that of the SC and SC-MLF wines. No color attributes were significantly correlated with the descriptive sensory attributes of the wine treatments.
Sensory analyses revealed that LT wines, regardless of MLF, were less liked by consumers and perceived as more sour by descriptive panelists. However, using L. thermotolerans to increase acidity in low-acid wines offers a valuable alternative to chemical or blending adjustments, naturally enhancing the wine’s color, flavor, and microbial stability. Furthermore, winemakers have the option to ameliorate high-acid wines with sugar just prior to bottling to reduce perception of acidity, a practice shown to have a measurable effect on consumers’ perception of acids in wines containing 8.0 to 12.0 g/L tartaric acid (Nordeloos and Nagel 1972).
Conclusion
In 2020 and 2021, the composition, color, and sensory attributes of Chambourcin wines were significantly influenced by co-inoculation with L. thermotolerans and S. cerevisiae, particularly regarding acidity and color. Compared to wines inoculated with S. cerevisiae alone, wines fermented with L. thermotolerans (regardless of MLF treatment) had lower pH; higher TA; higher levels of lactic acid and total organic acid; and increased red color, brown color, and color density, suggesting potential for long-term aging. These wines were darker, with higher red color values, but were perceived as more sour and astringent, attributes negatively correlated with consumer liking. In contrast, wines inoculated with S. cerevisiae alone (regardless of MLF treatment) displayed greater sensory appeal, with higher liking scores across most hedonic categories, including flavor, aroma, and fruitiness.
Overall, mixed fermentation with L. thermotolerans and S. cerevisiae produced wines with distinct advantages, including lower pH, higher acidity, and greater microbial and color stability. To enhance consumer acceptability, particularly addressing sourness, winemakers could adjust wines before bottling by blending with less acidic wines or adding sugar. Blending might alter pH and color, but sugar adjustments would preserve the benefits of the acidity and color characteristics derived from mixed fermentation with L. thermotolerans yeast.
Supplemental Data
The following supplemental materials are available for this article in the Supplemental tab above:
Supplemental Table 1 Pairwise correlation at 12-mo storage between composition and sensory attributes for wines produced from Chambourcin grapes in 2020 with different inoculation treatmentsa.
Data Availability
All data underlying this study are included in the manuscript and its supplemental information.
Footnotes
This research was funded by Lallemand Oenology (Montreal, Canada), with technical support from Scott Laboratories (Petaluma, California). The authors would also like to acknowledge Hindsville Farms (Hindsville, AR) for providing Chambourcin grapes for this project.
Fleming AJ and Threlfall RT. 2025. Effect of Lachancea thermotolerans yeast on Chambourcin (Vitis hybrid) wine quality. Am J Enol Vitic 76:0760012. DOI: 10.5344/ajev.2025.24053
By downloading and/or receiving this article, you agree to the Disclaimer of Warranties and Liability. If you do not agree to the Disclaimers, do not download and/or accept this article.
- Received September 2024.
- Accepted February 2025.
- Published online May 2025
This is an open access article distributed under the CC BY 4.0 license.
References
Vol 76 Issue 1
Jump to section
Related Articles
Cited By...
More from this TOC section
Similar Articles
