Assessment of Sour Taste Quality and Its Relationship with Chemical Parameters in White Wine: A Case of Koshu Wine

Abstract

This study aimed to clarify the sensory characteristics of Koshu wine, which is the most popular white wine produced in Japan, by identifying descriptive terms for sour taste, a primary aspect of wine flavor. A sensory evaluation generated 56 terms related to sour taste quality. Some terms were categorized on the basis of the timing of perception— immediately after sipping, holding in the mouth, and after swallowing—while others were classified as expressing “temporal change”, “overall impression terms”, or “metaphorical terms”. From these, 12 terms—“fresh”, “stand out”, “sharp”, “soft”, “round”, “gentle”, “bright”, “duration”, “crisp”, “intensity”, “mild”, and “calm”—were selected, with definitions and reference standards (materials or examples that represent each characteristic) established. A trained sensory panel evaluated 16 Koshu wines, revealing significant differences in all sour taste quality terms except “duration”. The evaluation of “duration” may require improvement. Correlation analysis indicated that pH was strongly associated with “sharp” sour taste immediately after sipping, while titratable acidity and pH correlated with “round” and “gentle” sour taste when the wine was held in the mouth. Total acidity was linked to the duration of sour taste. Applying the sour taste quality terms determined from this study will enable the quantification of the sour taste quality of wines.

1. Introduction

Koshu (Vitis vinifera) is a Japanese indigenous grape cultivar. Its origin is attributed to a cross between the European wine grape species Vitis vinifera and wild Asian grape species (V. davidii or a closely related species) during its transmission to Japan via the Silk Road [1]. Currently, Koshu is the most extensively cultivated grape used for white wine production in Japan. In recent years, owing to improvements in winemaking techniques, Koshu wines have gained increasing recognition in international wine competitions, drawing global attention. Understanding the characteristics of Koshu is critical for advancing Japan’s wine industry. Koshu is known to have a low pH and a high proportion of tartaric acid relative to malic acid [2,3]. More recently, Koshu wines have been produced using a variety of winemaking techniques, including sur lie aging and skin fermentation. This diversification in production styles appears to have introduced qualitative differences in perception of sour taste. Because sour taste is a key factor in defining the characteristics of Koshu, a comprehensive analysis of its sour taste profile is necessary.

Previous research on sour taste can be broadly categorized into two main perspectives: the physiological mechanisms of sour taste perception (the receptor side) and the chemical contributors to sour taste (the stimulus side). Regarding the former, the mechanisms underlying the perception of sour taste have been partially elucidated, and in 2018, the sour taste receptor OTOP1 was identified, revealing its selective response to H⁺ ions. Additionally, research on the pathways through which acid stimuli are transmitted from taste cells to the nervous system and, ultimately, the brain, as well as the brain’s response to the acid stimuli, has been conducted [4,5,6]. Moreover, it has been reported that salivation and the buffering capacity of saliva influence the perception of sour taste [7,8]. On the other hand, regarding the latter, research on the contributors of sour taste has shown that various chemical parameters, pH, titratable acidity, the acid taste index (titratable acidity–pH) [9], and total acidity [10], as well as chemical parameters, including the concentration and composition of organic acids, play a role in the perception of sour taste [11]. The relationship between the receptor and the contributor of sour taste has been investigated through sensory evaluations. Although it is generally believed that lower pH and higher titratable acidity increase the perception of sourness, Plane et al. [9] found that neither parameter alone could fully explain differences in sour taste. Their sensory tests showed that a simple index, calculated by subtracting pH from titratable acidity, had a stronger correlation with perceived sourness. This empirical formula was proposed based on straightforward correlation analysis, without using complex statistical methods. However, the sour taste perceived when consuming complex food products, consisting of multiple components, is also influenced by other basic tastes such as sweet and bitter tastes, and related taste substances. Consequently, it is difficult to fully understand sour taste on the basis of a single attribute alone.

When we eat food, even a sour taste can evoke different impressions. Therefore, it is important to consider the “quality of taste”. A broad survey of research on taste quality indicates that, in addition to the basic tastes (such as sweetness and sourness), the concepts of “sub-attributes” and “additional attributes” play a significant role.

Sub-attributes are descriptors that provide more specific characteristics to the basic tastes. Paravisini et al. [12] demonstrated that 13 bitter compounds could be sensorially distinguished at the same bitter taste intensity, revealing qualitative differences in bitter taste perception. Wang et al. [13] developed descriptive terms to characterize the qualities of astringent taste (e.g., velvet and silk) and structured these terms into a systematic sensory wheel. Gawel et al. [14] proposed definitions and reference standards for different astringent taste qualities, enabling the quantification of astringent taste characteristics.

Another key concept in the study of taste quality is “additional attributes,” which refer to a range of influential factors, including aftertaste and cultural context. It is well known that aftertaste perception following food consumption influences sensory evaluation and preference, with factors such as duration, taste type, and sensory characteristics being the key aspects of assessment. Prescott et al. [15] investigated differences in taste perception between Japanese and Australian individuals, reporting that while there were no cross-cultural differences in taste intensity perception, hedonic evaluation varied across cultures, suggesting that cultural factors such as dietary experience influence preferences. These differences in sensory attributes are also linked to emotional responses and ultimately affect purchasing decisions, highlighting the importance of research on taste perception attributes.

Sour taste is closely associated with food spoilage, making it highly perceptible and significantly influencing preference. Spinelli et al. [16], investigating sour taste preferences among adult panels in the United States and Italy, demonstrated that individuals could be classified into three distinct groups on the basis of their preference for sour taste intensity. They found that approximately 63–70% of adults exhibited a decrease in preference as sour taste intensity increased, whereas 11–12% showed an opposite trend, with preference increasing with sour taste intensity. These findings suggest considerable variations in sour taste preference among adults, who constitute the primary consumer base for wine. Therefore, sour taste attributes play a crucial role in understanding sour taste perception and contribute to shaping individual preference.

In this study, the term “sour taste quality” encompasses both sub-attributes and additional attributes of sour taste, including immediate sensory aspects perceived during tasting and post-swallow perceptions that emerge after the wine is swallowed. However, research on sour taste quality is still in its infancy, and no systematic studies on the sour taste quality in wine have been reported to date. On the other hand, several sensory evaluation studies of sake have come up with descriptive terms, such as “fresh sour taste”, “mild sour taste”, and “crisp sour taste” in Japan [17,18,19]. Hayakawa et al. [20] developed a lexicon for describing the sensory attributes of coffee. They proposed six sour taste-related terms and their definitions, including “acidic”, “pure acidic”, “sharp acidic”, “mild acidic”, “rounded acidic”, “bright acidic”, and “acidic aftertaste”. These terms likely represent different qualities of sour taste. However, no reference standards have been established for learning sour taste quality, and its quantification remains insufficiently developed. When drinking wine, not only the intensity of the sour taste but also its qualitative aspects, such as “fresh”, “mild”, and “aftertaste”, can be perceived. Indeed, wine tasting notes frequently include expressions that describe sour taste quality. Prescott et al. [15] reported that Japanese individuals exhibit a higher tolerance for sour taste than Australians, suggesting that Japanese consumers may be more sensitive to variations in sour taste quality. Nevertheless, compared to English-language studies, research on Japanese wine tasting terminology remains limited, and the relationship between sour taste qualities and chemical parameters has yet to be elucidated.

This study aimed to clarify the qualities of sour taste in Koshu wines. First, terms used to describe the sour taste of wine were derived and systematically organized. Next, definitions and reference standards for selected terms describing sour taste qualities were developed, and a trained panel was established on the basis of these reference standards. Using this trained panel, sour taste qualities in wine were quantified, and differences or commonalities in sour taste qualities between wines were examined. Furthermore, the relationship between sour taste qualities and chemical parameters was elucidated. This study is a fundamental investigation of whether differences exist in sour taste qualities. Koshu wine, which has a mild aroma [21], was considered an ideal subject for this study because the influence of its aroma on the perception of sour taste could be minimized, allowing for a more focused evaluation of sour taste qualities.

2. Materials and Methods

2.1. Wine Samples

Twenty-two dry Koshu wines and one dry Chardonnay wine were used in this study. Twenty of the 22 Koshu wines were commercial wines categorized as dry. These wines were selected based on a sensory screening to ensure the absence of unpleasant odors, carbonation, or a notably sweet taste. Eight were produced using the sur lie method, and one was made through skin fermentation. Not all the wines were barrel-aged, and the wine vintage was from 2020 to 2022. The remaining Koshu wine and the Chardonnay wine were experimental wines. The experimental wines were produced by conventional winemaking methods using Koshu and Chardonnay grapes harvested from the experimental vineyard of the Institute of Enology and Viticulture, University of Yamanashi, in 2021.

2.2. Sensory Analysis

A summary of the methods used in this study is shown in

Table 1. Summary of methods used in this study.

	Purpose	Method	Number of Panelists	Wine Samples Used
1: Deriving a sour taste vocabulary	To derive a sour taste vocabulary	Tasting followed by writing out terms to describe the sour taste	9 (4 men and 5 women)	First day: Two commercial Koshu wines and one experimental Koshu wine (including one wine produced using the sur lie method) Second day: One experimental Koshu wine and one experimental Chardonnay wine
2: Selecting evaluation terms	To select evaluation terms	Organizing the terms derived in (1) by positioning them in a two-dimensional space. Discussing the meanings of the terms to reach a common understanding.	11 (5 men and 6 women)	Not applicable
3: Creating definitions and reference standards for evaluation terms	To create and validate definitions and reference standards for the evaluation terms	Step 1: Discussing appropriate definitions and reference standards for the evaluation terms in (2) while tasting several reference standards. Step 2: Confirming definitions and reference standards for each evaluation term	Step1: 3 (1 man and 2 women) Step2: 10 (5 men and 5 women)	Tartaric acid and malic acid solutions, carbonated drinks, juices, and white wines
4: Trial test	To assess the flow of sensory evaluation methods and the between-panelist agreement	Rating the intensity of sour taste qualities using an 11 cm line scale.	10 (5 men and 5 women)	Three commercial Koshu wines (including one wine produced using the sur lie method)
5: Test	To measure the sour taste qualities of wines and evaluate differences between wine samples	Rating the intensity of sour taste qualities using an 11 cm line scale.	13 (7 men and 6 women)	Sixteen commercial Koshu wines (including six wines produced using the sur lie method and one wine produced through skin fermentation)

. The panel consisted of men and women participants aged between 23 and 40 years.

2.2.1. Panel

The sensory panel comprised 14 individuals (7 men and 7 women), including researchers and students at the Institute of Enology and Viticulture, University of Yamanashi. All panelists had extensive wine tasting experience and were familiar with Koshu wine. They could recognize the five basic tastes (sweet, salty, bitter, sour, and umami). In addition, they were trained to evaluate the intensity of sour taste using tartaric acid solution at different concentrations (0.1, 0.15, 0.22, and 0.34 g/L). Tartaric acid was selected as the reference substance because it is the major organic acid present in wine, and its use is also recommended in training materials published by the Japanese Society for Sensory Evaluation. The selected concentrations were based on the citric acid levels recommended in ISO 8586-1 [22] for sour taste training. Participation was on a voluntary basis, and prior to the experiments, all panelists were fully informed of the study’s objectives and procedures. Because the University of Yamanishi did not have an established process for ethical review at the time of this study, guidelines from the Japanese Society of Sensory Evaluation were followed. They provided written consent by signing a consent form.

2.2.2. Procedure

Deriving a Sour Taste Vocabulary

This task was performed by a panel of 9 individuals (4 men and 5 women) for two days in a meeting room. On the first day, two commercial Koshu wines and one experimental Koshu wine were used; the wines for this session were randomly selected from both conventionally vinified wines and those made using the sur lie method. On the second day, one experimental Koshu wine and one experimental Chardonnay wine were used. Wine samples stored at 4 °C were served in standard ISO tasting glasses marked with three-digit numbers, in a randomized order. The panelists were instructed to sip and spit out a wine sample and to write as many words related to the sour taste as possible. In addition, a space was provided for comments on aroma. This made the panelists aware of having to distinguish between taste and aroma. They were asked to rinse their mouths with water between evaluations. In addition, all terms were written out in Japanese.

Selecting Evaluation Terms

This task was performed by a panel of 11 individuals (5 men and 6 women). The terms written out by the panelists were sorted by their meaning by positioning them in a two-dimensional space. Some terms were categorized on the basis of the timing of perception: immediately after sipping, holding in the mouth, and after swallowing. We also categorized some terms as terms that express temporal changes, overall impression terms, and metaphoric terms. Furthermore, the final terms used in the sensory evaluation were determined from those that were used frequently and those that could be commonly understood by the panelists.

Creating Definitions and Reference Standards for Evaluation Terms

This task was performed by a panel of 3 individuals (1 man and 2 women). The definitions and reference standards for the evaluation terms were developed using tartaric acid and malic acid solutions, carbonated drinks, juices, and white wines (

Table 2. Twelve evaluation terms and their definitions and reference standards.

Term (Japanese)	Definition	Reference Standard
Immediately after sipping
Fresh (Sukkiri)	Sour taste is perceived to be fresh immediately after sipping	Ramune (Japanese carbonated drink)
Stand out (Kiwadatsu)	Sour taste that stands apart from other tastes	Koshu wine with 1 g/L malic acid added
Sharp (Surudoi)	Sour taste is perceived to be powerful and to occur shortly after sipping the sample	Lemon juice
Soft (Yawarakai)	Sour taste that penetrates smoothly without imparting discomfort	Tasting commercial soft water (Ca 24, Mg 11, K 7.0, Na 8.0 mg/L) after commercial hard water (Ca 470, Mg 75, K 2.8, Na 9.4 mg/L)
Holding in the mouth
Round (Marui)	Sour taste perceived to be mild and not pungent	Peach juice
Gentle (Yasashii)	Sour taste perceived to be mild and to slowly become apparent after holding in the mouth	Lemonade
After swallowing
Bright (Sawayaka)	Sour aftertaste perceived to be clear	Ramune (Japanese carbonated drink)
Duration (Jizoku)	Length of time sour taste is felt	No suitable reference standard
Temporal change
Crisp (Kire)	Sour taste that occurs shortly after sipping and does not persist long after swallowing	6 g/L malic acid solution
Overall impression
Intensity (Kyou-jyaku)	Intensity of sour taste	0.1, 0.15, 0.22, 0.34 g/L tartaric acid solution
Mild (Maroyaka)	Sour taste is perceived to be smooth and to have a pleasant mouthfeel	Peach juice
Calm (Odayaka)	Sour taste that does not stand out and blends well with other tastes	No suitable reference standard

). The validity of the definitions and the reference standards was confirmed by a panel of 10 individuals (5 men and 5 women) who reviewed and discussed the definitions and reference standards for each term.

Trial Test

This test was participated in by a panel of 10 individuals (5 men and 5 women) in a sensory room equipped with individual booths and set at a controlled temperature (20–22 °C). Black ISO wine glasses were used to avoid visual bias. The wine was stored at 4 °C before the sensory evaluation, poured into the wine glass, and served to the panelists with a plastic lid. By the time the panelists rated the wine, the temperature of the wine in the wine glass had increased to approximately 10 °C. Evaluations were conducted between 10:00 am and 12:00 noon or between 3:00 pm and 4:00 pm. Three commercial Koshu wines were used in this trial test. The panelists were instructed to rate the presented wines from left to right. The order in which the wines were presented was randomized for each panelist. The panelists were asked to move the plastic lid slightly and take wine into their mouths without smelling it. There were three major questions in the evaluation. In Question 1, the panelists were asked to assess the intensity and duration of the sour taste (Supplementary Figure S1). The panelists evaluated these two terms first because they are commonly used evaluation terms and are easy to evaluate. In Question 2, they were asked to evaluate 10 sour taste qualities: “fresh”, “stand out”, “sharp”, “soft”, “round”, “gentle”, “bright”, “crisp”, “mild”, and “calm”. Questions 1 and 2 were evaluated by a check method on an 11 cm line scale on the paper (with a vertical line 0.4 cm inward from both ends). The lines for Question 1 were labeled strong and weak for the “intensity” item and short and long for the “duration” item. The line for Question 2 was not labeled. Question 3 showed nine metaphoric terms obtained through the term derivation. The panelists were asked to select the corresponding terms (multiple answers possible). Intensity and duration in Question 1 were assessed by taking the wine into the mouth once (tasting). For Questions 2 and 3, the tasting could be repeated multiple times.

The trial test was conducted to confirm the appropriateness of the evaluation sheet, instructions, and sensory procedure, as well as to ensure panelist understanding of the defined attributes. It also helped determine that evaluating five to six samples per session was optimal to avoid fatigue. These findings informed the design of the main test.

Test

The test was participated in by a panel of 13 individuals (7 men and 6 women). Sixteen commercial Koshu wines were evaluated twice in three sessions. Five or six Koshu wines were evaluated per session. We ensured that the order in which the wines were served was not the same on the first and the second visits. The evaluation process was similar to the trial test. There was an interval of at least 5 min between sessions. After each session, the experimenter cleared the samples and prepared them for the next session, while the panelists rinsed their mouths with water and filled in their names and the date on the evaluation sheet for the following session. Although no strictly standardized break time was set between sessions, the panelists were allowed to rest sufficiently before proceeding to the next session.

2.3. Chemical Parameters

Wine pH was measured with a pH meter (Model F-71, Horiba Ltd., Kyoto, Japan). Titratable acidity was determined by neutralization titration with 0.1 N NaOH to pH 8.2 endpoint using an automatic titrator (COM-1700, Hiranuma Sangyo Co., Ltd., Ibaraki, Japan), and is expressed as g/L of tartaric acid equivalent. Titratable acidity–pH value was derived by subtracting the pH value from the titratable acidity (expressed in g/L) as a numerical operation. The concentrations of organic acids (citric, tartaric, malic, succinic, lactic, and acetic acids) were analyzed with a high-performance liquid chromatography (HPLC) system. The HPLC system (Shimadzu Co., Kyoto, Japan) consisted of a system controller (SCL-10AVP), a pump (LC-20AD), a column oven (CTO-20A), a degasser (DGU-20A3), an electrical conductivity detector (CDD-20A), and a Chromatopac data processor (C-R8A). Two Shodex RSpack KC-811 columns (300 mm × 8 mm I.D., Showa Denko, Tokyo, Japan) were connected in series and maintained at 40 °C. The mobile phase was a 5 mM p-toluenesulfornic acid aqueous solution, delivered at a flow rate of 0.8 mL/min. Total acidity is the proton equivalence of the amount of organic acid anions of citric, tartaric, malic, succinic, lactic, and acetic acids [10]. Ethanol concentration was analyzed on a gas chromatograph–flame ionization detector (GC 2014, Shimadzu, Co., Ltd., Kyoto, Japan) equipped with a packed column (polyethylene glycol 600, Shinwa Chemical Industries, Co., Ltd., Kyoto, Japan), following the official analytical method specified by the National Tax Agency of Japan. Glucose and fructose were determined by a Biosystems Y15 Enology Automatic Analysis System (Biosystems, S.A., Barcelona, Spain), according to the manufacturer’s protocol.

2.4. Data Analysis

The statistical software JMP 13 (SAS Institute Inc., Cary, NC, USA) was used for statistical analysis. Analysis of variance (ANOVA) and multiple comparison tests (Tukey–Kramer HSD test) were conducted to examine differences in sour taste quality among wines. These analyses were performed using all data obtained from the sensory evaluation. A hierarchical cluster analysis with the Ward criteria was performed to categorize wines on the basis of sour taste quality. Correlation analysis was performed to determine the relationship between sour taste qualities and chemical parameters, using average values. Additionally, a hierarchical cluster analysis with bootstrap resampling was also performed using R software (version 4.5.1; R Foundation for Statistical Computing, Vienna, Austria) to statistically evaluate the stability of clusters. Panel performance was evaluated using R and the SensoMineR package (version 1.27; Agrocampus Ouest, Rennes, France) [23]. For this purpose, F-values, p-values, mean square errors (MSE), intraclass correlation coefficients (ICC), and Cronbach’s α were calculated for each of the 12 sour taste qualities. Furthermore, attribute-wise F-values for each of the 13 panelists were obtained. Panelist-wise ICC, 95% confidence intervals, p-values, inter-rater agreement, and Distance-from-Target values were also computed to assess individual panelist reliability and accuracy.

3. Results and Discussion

3.1. Creating and Categorizing a Sour Taste Quality Vocabulary

A sensory evaluation using white wine led to the derivation of 60 Japanese terms describing sour taste quality. These terms were translated into their closest English equivalents in nuance, resulting in 56 English terms, as some Japanese terms shared the same English translation. The English terms and their frequencies are presented in

Table 3. Terms describing sour taste quality and their frequencies.

Classification	Term (Frequency)
Immediately after sipping	Fresh (3), Stand out (3), Sharp (3), Pungent (2), Stimulating to the tongue (3), Tangy (1), Clear (1), Burnished (1), Soft (3)
Holding in the mouth	Round (3), Gentle (4), Adequate (2), Volume (1), Robust (1), Delicate (1), Faint (3), Flat (2)
After swallowing	Light (1), Smooth (2), Bright (1), Duration (4), Continue (2), Remain (14), Not remain (2), Long (8), Short (2)
Temporal change	Crisp (2), Gradually stimulating (1), Disappear midway (1), Weakly stimulating midway (1), Gradually spreading (1), Come after (4), Even (1)
Overall impression	Strong (17), Weak (8), Not strong (2), Acidic (6), Mild (8), Calm (4), Depth (1), Inharmonious (2), Pleasant (2), Moderate (1), Piquant (1), Unpleasant (5)
Metaphoric terms	Fruit (3), Fruity (3), Citrus (4), Natsumikan (a type of Japanese citrus) (3), Lemon (1), Ume (Japanese apricot) (2), Ginger (1), Ramune (a Japanese carbonated drink) (1), Tartaric acid (3), Malic acid (2), Green (2)

. Additionally, explanations to facilitate a precise understanding of each term are provided in Supplementary Table S1. Furthermore, from the perspective of understanding the perception of sour taste, the derived terms were categorized on the basis of the timing of perception during the tasting process. Specifically, perceptions occurring immediately after sipping the sample, while holding the sample in the mouth, and after swallowing were classified as “immediately after sipping”, “holding in the mouth”, and “after swallowing”, respectively. Terms conveying nuances of temporal change were separately categorized under “temporal change”. Terms representing overall impression were grouped under “overall impression”, while metaphorical expressions referring to fruits or citrus notes were classified as “metaphorical terms”.

Most of the terms were also found in English-language wine tasting comments, suggesting the use of shared descriptive terminology across cultures [24]. However, some metaphorical terms, such as natsumikan (a type of Japanese citrus), ume (Japanese apricots), and ramune (a Japanese carbonated drink), were uniquely Japanese terms.

Immediately after sipping wine, the panelists used such terms as “fresh”, “sharp”, and “pungent”, suggesting that these sensations were caused by the rapid and strong induction of sour taste perception. Furthermore, it is thought that the term “stand out” was used when sour taste perception exceeded other taste perceptions. On the other hand, the term “soft” was used when no rapid induction of sour taste perception was felt. The terms “round” and “gentle” were used to describe the sour taste quality that could be felt while wine was held in the mouth. Unlike immediately after sipping, no terms expressing rapid and strong sour taste perception were derived when the wine was held in the mouth. After swallowing the wine out, the term “light”, expressing the lack of a rich and heavy taste, and the associated term “bright” were also used. In addition, the terms “duration”, “continue”, “remain”, “long”, and “short” were used to express the length of time. The term “crisp” was used to express the temporal change in sour taste. Crisp is the sour taste sensation that occurs shortly after taking the sample into the mouth; it does not persist long after swallowing and is used as the term to describe the taste characteristics of wine and beer [25]. The terms “gradually stimulating”, “gradually spreading”, and “even” were used to express changes in sour taste intensity and its spread in the mouth. The terms “disappear midway” and “come after” were used to express changes in the sense of taste over time. The terms “strong”, “weak”, “mild”, “calm”, and “depth” were used to express the overall impression of sour taste. The terms “citrus” and “Japanese apricot”, which are fruits having a sour taste, and “tartaric acid” and “malic acid”, which are the main organic acids in grapes and wine, were derived. The term “green”, often used to describe the taste of malic acid, was also derived. These terms were categorized as metaphoric terms.

Of the 56 derived terms, 12 that were commonly understood and frequently used by the panelists were selected as the descriptors for evaluating sour taste quality. The terms “strong” and “weak” were consolidated into a single descriptor, “intensity”, to encompass both meanings. Furthermore, the definitions and reference standards for the 12 selected terms were established (Table 2). Through training using these definitions and reference standards, the panelists became capable of accurately understanding and distinguishing sour taste qualities.

3.2. Differences in Sour Taste Quality of Koshu Wine

Table 4. Average intensities of sour taste qualities of 16 Koshu wines.

Term			Wine No.																Range of Means
	p	F	W1	W2	W3	W4	W5	W6	W7	W8	W9	SL1	SL2	SL3	SL4	SL5	SL6	O1
Fresh	0.0014	2.52	7.05 a	5.75 a	5.79 a	5.71 a	6.43 a	5.69 a	6.52 a	7.13 a	7.22 a	6.41 a	6.63 a	7.07 a	6.91 a	6.69 a	6.90 a	6.75 a	5.69–7.22
Stand out	<0.001	6.70	6.73 abcd	5.49 abcde	7.05 abc	3.60 e	7.50 a	4.88 de	6.73 abcd	6.81 abcd	7.23 ab	5.89 abcde	6.46 abcd	5.59 abc	6.37 abcd	5.05 cde	5.27 bcde	4.24 e	3.60–7.50
Sharp	<0.001	5.72	5.92 abc	5.07 abcde	6.25 ab	2.88 e	6.40 ab	3.71 de	5.79 abcd	5.75 abcd	6.97 a	4.67 bcde	6.01 ab	4.81 abcde	5.78 abcd	4.76 abcde	4.70 abcde	3.62 de	2.88–6.97
Soft	<0.001	4.30	6.21 abcd	5.46 bcd	4.56 d	7.18 ab	5.09 cd	6.75 abc	5.99 abcd	5.12 bcd	4.46 d	6.67 abc	5.55 abcd	6.46 abcd	5.52 abcd	5.95 abcd	6.36 abcd	7.46 a	4.46–7.46
Round	<0.001	4.88	5.07 bc	5.28 bc	4.34 c	7.44 a	4.55 c	6.80 ab	5.01 bc	4.81 bc	4.10 c	6.11 abc	4.73 bc	5.52 abc	5.54 abc	6.20 abc	5.98 abc	7.48 a	4.10–7.48
Gentle	<0.001	3.64	5.55 ab	5.70 ab	4.70 b	7.52 a	4.54 b	6.16 ab	5.77 ab	5.11 b	4.88 b	6.50 ab	4.99 b	5.94 ab	6.18 ab	5.94 ab	6.30 ab	7.39 a	4.54–7.52
Bright	<0.001	2.96	6.81 ab	5.27 b	5.24 b	6.30 ab	6.12 ab	6.01 ab	6.51 ab	7.41 a	6.88 ab	6.49 ab	6.40 ab	6.70 ab	6.75 ab	7.03 ab	7.14 a	7.26 a	5.24–7.41
Duration	0.804	0.68	5.77 a	5.76 a	6.52 a	5.58 a	5.73 a	5.41 a	6.10 a	5.76 a	6.55 a	5.91 a	5.75 a	5.86 a	5.80 a	6.36 a	5.64 a	5.46 a	5.41–6.55
Crisp	0.007	2.17	6.80 a	5.34 a	5.98 a	4.87 a	6.55 a	5.52 a	6.01 a	6.62 a	6.37 a	5.60 a	6.41 a	5.57 a	6.15 a	5.25 a	6.84 a	5.52 a	4.87–6.84
Intensity	<0.001	4.72	6.60 abc	5.91 abcd	7.18 a	4.27 d	7.02 ab	5.14 bcd	6.52 abc	6.47 abc	7.49 a	5.86 abcd	6.76 abc	6.12 abcd	6.92 ab	6.23 abc	5.87 abcd	4.96 cd	4.27–7.49
Mild	<0.001	3.99	5.52 abc	5.36 bc	4.60 c	6.84 ab	5.02 bc	6.66 ab	5.90 abc	5.31 bc	4.57 c	6.20 abc	5.11 bc	5.62 abc	5.75 abc	6.17 abc	6.52 abc	7.39 a	4.57–7.39
Calm	<0.001	2.92	5.57 abc	5.66 abc	4.67 bc	6.83 ab	4.55 c	5.95 abc	5.55 abc	5.15 bc	4.85 bc	6.32 abc	5.33 abc	6.01 abc	5.82 abc	6.41 abc	6.16 abc	7.35 a	4.55–7.35

Different letters indicate significant differences among Koshu wines according to the Tukey–Kramer HSD test (p < 0.05).

shows the average intensities of the 12 sour taste qualities of 16 Koshu wines. One panelist conducted only one of the two repeated evaluations for a single wine, resulting in 12 missing data points and the exclusion of their data from the statistical analysis. The corresponding standard deviation values are provided in the Supplementary Material (Table S2) to support the interpretation of statistical differences. All sour taste qualities except “duration” show a significant difference in the average intensity (p < 0.05). This study demonstrated qualitative differences in sour taste among wines and showed that the sour taste qualities can be quantified through the establishment of a trained panel. Among the attributes, “stand out,” “sharp,” “soft,” and “round” exhibited strong inter-rater agreement (ICC > 0.75) and acceptable to high internal consistency (Cronbach’s α > 0.8), indicating that the panel was able to evaluate these characteristics across wines with both reliability and consistency (Supplementary Material, Table S3). In contrast, “duration” did not differ significantly among the wines (p = 0.804), and its ICC was negative. In the present panel training, “duration” was understood only through its definition, without standardizing the evaluation scale by providing specific time references. This lack of standardization may have led to inconsistencies in evaluation scales among the panelists, ultimately resulting in the absence of significant differences among wines. For the evaluation of “duration”, the time-intensity method, which measures the duration of taste perception, may be more suitable than the line scale method. Similarly, while “fresh” and “crisp” showed moderate performance, the relatively low ICC and Cronbach’s α values suggest potential ambiguity or a lack of clarity in the evaluation criteria. Notably, “crisp”, one of the key taste attributes in beer, is evaluated using the time-intensity method and specialized sensors [26]. In sensory evaluations conducted by beer manufacturers, samples are often swallowed rather than spat out, which may contribute to a more accurate perception of temporal characteristics. These methodological differences suggest that incorporating time-based evaluation techniques could enhance the assessment of similar attributes in wine tasting.

Panelist-specific performance analyses (Table S4: F-values; Table S5: ICC, inter-rater agreement, and Distance-from-Target) revealed variability in discrimination ability and consistency. Several panelists exhibited low F-values across multiple attributes, indicating limited ability to discriminate between wines. Additionally, panelists 4, 6, and 8 showed low ICCs and inter-rater agreement, along with higher Distance-from-Target values, suggesting a deviation from the overall panel consensus and inconsistency in their evaluations. These findings highlight the need for further training, particularly in terms of interpretation and use of intensity scales for specific sourness-related attributes. Additional calibration of intensity scaling and reinforcement of shared definitions for each attribute would be essential to improve both the reliability and discriminatory performance of the panel.

3.3. Relationship Between Sour Taste Quality Expression

Table 5. Correlation coefficients between sour taste quality expressions.

	Immediately After Sipping				Holding in the Mouth		After Swallowing		Temporal Change	Overall Impression
	Fresh	Stand Out	Sharp	Soft	Round	Gentle	Bright	Duration	Crisp	Intensity	Mild	Calm
Immediately after sipping
Fresh	1.000
Stand out	0.362	1.000
Sharp	0.425	0.963 **	1.000
Soft	–0.156	–0.828 **	–0.887 **	1.000
Holding in the mouth
Round	–0.307	–0.942 **	–0.965 **	0.909 **	1.000
Gentle	–0.180	–0.892 **	–0.893 **	0.896 **	0.938 **	1.000
After swallowing
Bright	0.818 **	–0.106	–0.074	0.270	0.206	0.268	1.000
Duration	0.161	0.507 *	0.604 *	–0.663 **	–0.593 *	–0.518 *	–0.131	1.000
Temporal change
Crisp	0.590 *	0.723 **	0.703 **	–0.490	–0.648 **	–0.604 *	0.307	0.066	1.000
Overall impression
Intensity	0.444	0.934 **	0.980 **	–0.879 **	–0.931 **	–0.872 **	–0.055	0.664 **	0.654 **	1.000
Mild	–0.171	–0.863 **	–0.899 **	0.928 **	0.959 **	0.923 **	0.342	–0.622 *	–0.482	–0.876 **	1.000
Calm	–0.077	–0.904 **	–0.875 **	0.903 **	0.932 **	0.957 **	0.359	–0.473	–0.602 *	–0.835 **	0.923 **	1.000

Correlation between sour taste qualities was analyzed by Pearson’s correlation test. Single asterisk (*) and double asterisks (**) indicate significant correlations at 5% and 1% levels, respectively.

shows the correlation coefficients between various sour taste qualities. Strong correlations were frequently observed among the different taste qualities. In particular, “stand out” and “sharp”, perceived immediately after sipping, exhibited a strong positive correlation. This result suggests that the “sharp” sour taste may accentuate the other tastes, resulting in these two sensations occurring simultaneously. On the other hand, “stand out” and “sharp” showed a strong negative correlation with “soft”. Although “stand out” and “sharp” are not the exact opposites of “soft”, they are considered to be understood as “soft” sour taste when the sour taste is perceived as smooth with no unpleasant sharpness. Additionally, “soft” demonstrated a strong positive correlation with “round” and “gentle”, both of which were perceived when the wine sample was held in the mouth. Fukushima [27] reported that in a corpus analysis of Japanese sake tasting comments, “soft”, “round”, and “gentle” frequently co-occur. This aligns with the findings of the present study regarding the relationship between sour taste quality and its intensity. Therefore, these sour taste qualities are likely to be perceived simultaneously and with similar intensity. “Bright”, which is perceived after swallowing, was strongly correlated with “fresh”, which is perceived immediately after sipping. These terms frequently co-occur in product descriptions in Japan, suggesting that the sour taste quality perceived immediately after sipping influences the sour taste quality experienced after swallowing. In terms of overall impression, “intensity” exhibited a positive correlation with “stand out”, “sharp”, “duration”, and “crisp”, indicating that the strength of the acidic stimulus experienced immediately after sipping, combined with the duration of sour taste remaining after swallowing, contributes to the overall perception of sour taste “intensity”. Conversely, “mild” and “calm” showed a strong negative correlation with “stand out” and “sharp” while displaying a strong positive correlation with “soft”, “round”, and “gentle”. These findings suggest that when sour taste does not produce a prominent stimulus immediately after sipping or when held in the mouth, it is generally perceived as “mild” or “calm”. Regarding the relationship between “mild” and “calm”, Mora et al. [28] conducted a sensory evaluation of wine through a principal component analysis and found that both terms were oriented in the same vector direction. This suggests that they may represent qualities that occur simultaneously and with similar intensity. Therefore, further investigation is necessary to re-examine the distinguishability of “mild” and “calm” in the context of sour taste. Based on the outcome, a decision should be made as to whether these potentially synonymous terms should be treated as distinct descriptors or integrated into a single attribute for future evaluations.

3.4. Classification of Koshu Wine Based on Sour Taste Quality

Hierarchical cluster analysis of the parameters of sour taste qualities was conducted, resulting in the classification of 16 Koshu wines into four distinct clusters (Figure 1). Koshu wines in Cluster 1 had an intensity of approximately 6 for all sour taste qualities, with slightly higher values for “stand out” and “sharp”. Koshu wines in Cluster 2 were characterized by high intensities for “stand out”, “sharp”, and “duration”. In contrast, Koshu wines in Cluster 3 displayed high intensities for “soft”, “round”, and “gentle”. Koshu wines in Cluster 4 had an intensity of approximately 6 for all sour taste qualities. In summary, Cluster 1, by a balanced sour taste quality profile with moderate sharpness; Cluster 2, by pronounced “stand out”, “sharp”, and “duration” sour taste qualities; Cluster 3, by a “mild” and “calm” sour taste qualities; and Cluster 4 is characterized by a balanced sour taste quality profile. These findings demonstrate the capability of classifying Koshu wines on the basis of their sour taste qualities. Until now, sour taste has typically been compared on the basis of its intensity. The capability of comparing differences in sour taste qualities is expected to help winemakers describe the characteristics of their wines in greater detail while also aiding consumers in selecting wines that best match their preferences. Interestingly, all sur lie Koshu wines were classified into Clusters 1 and 4, with balanced sour taste quality profiles. These data indicate that the production method may affect the sour taste quality of the wine is important and worthy of further investigation.

3.5. Relationship Between Sour Taste Quality and Chemical Parameters

Table 6. Correlation coefficients of sour taste qualities with chemical parameters.

	Immediately After Sipping				Holding in the Mouth		After Swallowing		Temporal Change	Overall Impression
	Fresh	Stand Out	Sharp	Soft	Round	Gentle	Bright	Duration	Crisp	Intensity	Mild	Calm
pH	–0.349	–0.584 *	–0.700 **	0.667 **	0.583 *	0.430	–0.031	–0.620 *	–0.314	–0.723 **	0.539 *	0.394
Titratable acidity	0.162	0.524 *	0.614 *	–0.643 **	–0.538 *	–0.571 *	–0.097	0.683 **	0.237	0.622 *	–0.556 *	–0.551 *
Titratable acidity–pH	0.202	0.571 *	0.671 **	–0.692 **	–0.584 *	–0.591 *	–0.084	0.722 **	0.266	0.683 **	–0.593 *	–0.566 *
Total acidity	0.247	0.458	0.518 *	–0.535 *	–0.459	–0.442	0.035	0.746 **	0.145	0.530 *	–0.461	–0.425
Citric acid	–0.331	–0.381	–0.356	0.110	0.321	0.350	–0.037	0.124	–0.463	–0.360	0.247	0.202
Tartaric acid	0.368	0.587 *	0.612 *	–0.535 *	–0.544 *	–0.461	0.136	0.835 **	0.365	0.647 **	–0.510 *	–0.440
Malic acid	–0.006	0.092	0.161	–0.345	–0.119	–0.162	–0.016	0.268	–0.116	0.165	–0.150	–0.170
Succinic acid	–0.104	0.382	0.425	–0.503 *	–0.525 *	–0.505 *	–0.503 *	0.614 *	–0.025	0.466	–0.631 **	–0.448
Lactic acid	0.322	0.200	0.183	0.124	–0.116	–0.064	0.215	–0.065	0.391	0.117	–0.018	–0.029
Acetic acid	0.061	–0.157	–0.210	0.266	0.207	0.123	0.174	–0.063	–0.173	–0.177	0.222	0.120
Tar/Mal	0.254	0.205	0.183	0.078	–0.142	–0.093	0.117	–0.044	0.376	0.138	–0.102	–0.076
Ethanol	–0.118	0.117	0.037	–0.262	–0.151	–0.225	0.029	–0.064	0.135	0.013	–0.124	–0.293
Sugar	0.092	–0.364	–0.331	0.402	0.460	0.430	0.291	–0.280	–0.182	–0.343	0.492	0.490

Correlation between sour taste quality and pH, acidity, or organic acid concentration was analyzed by Pearson’s correlation test. Single asterisk (*) and double asterisks (**) indicate significant correlations at 5% and 1% levels, respectively.

shows the correlation coefficients of sour taste qualities with chemical parameters. “Stand out”, “sharp”, and “soft” showed significant correlations with pH, titratable acidity, and titratable acidity–pH. In particular, pH exhibited a strong negative correlation with “sharp”. It is generally known that acidity is perceived as oral pH decreases [7]. Specifically, wines with low pH cause a rapid decrease in oral pH upon ingestion, which may be perceived as “sharp”. Conversely, “soft” showed a strong positive correlation with pH, suggesting that this sour taste quality is perceived when the decrease in oral pH occurs gradually. Furthermore, pH and titratable acidity–pH showed higher correlation values than titratable acidity, suggesting that the soft sour taste quality perceived immediately after sipping is more strongly influenced by pH. Like “soft”, “round” showed significant correlations with pH and titratable acidity–pH. In this study, “round” was defined as the absence of strong stimulation. Wines with low pH are perceived as “sharp” immediately after sipping; however, when such stimulation is absent, the sour taste may be perceived as “round”. “Gentle” showed a negative correlation with titratable acidity and titratable acidity–pH. In this study, “gentle” was defined as a sour taste that develops slowly, which may be related to saliva secretion. Tenuta et al. [29] reported that acidic beverages with higher titratable acidity resulted in a longer recovery time of salivary pH. This suggests that higher titratable acidity induces greater saliva secretion. Conversely, lower titratable acidity may result in less saliva secretion, causing the spread of sour taste in the mouth to occur more gradually, perceived as “gentle”. “Duration” showed strong correlations with titratable acidity, titratable acidity−pH, and total acidity, with the strongest correlation observed with total acidity. Notably, total acidity has rarely been addressed in acidity-related studies, but its importance was highlighted in this study. Furthermore, total acidity has been shown to be associated with buffering capacity [30], which resists changes in pH, suggesting that this effect influences the duration of sour taste. When total acidity is high, buffering capacity increases, prolonging the time required for neutralizing acids by saliva, thereby enhancing the perception of “duration”. Regarding “intensity”, significant correlations were observed with pH, titratable acidity, and titratable acidity–pH, with the strongest correlation found with pH. Previous studies have established that increased titratable acidity enhances sour taste intensity [7,9], and titratable acidity is widely recognized as an indicator of sour taste. However, one study has suggested that sour taste is more strongly correlated with pH than with titratable acidity [31]. The debate on which of these parameters more accurately reflects the sour taste continues. This study revealed that the sour taste quality perceived immediately after sipping correlates more strongly with pH than with titratable acidity. In contrast, the sour taste quality perceived when the wine is held in the mouth or after swallowing had a stronger correlation with titratable acidity than with pH. These findings suggest that the discrepancies among previous studies may be attributed to differences in the timing of sour taste evaluation by the panelists.

We also investigated the relationship between organic acids and sour taste qualities and found that tartaric acid concentration was correlated with several sour taste attributes. Sass-Kiss et al. [32] reported that a high tartaric acid concentration imparts a sharp taste to wine. We found a positive correlation between “sharp” and tartaric acid concentration, supporting the previous finding by Sass-Kiss et al. Tartaric acid is a stronger acid than other organic acids and has a significant impact on pH. Therefore, it is possible that tartaric acid enhances the perception of “sharp” indirectly by lowering the pH. On the other hand, malic acid is known to be associated with such descriptors as “green” and “fresh” [33,34,35]. However, we did not find a significant correlation between malic acid and any of the sour taste qualities. The wines used in this study had a higher concentration of tartaric acid than malic acid, except for one sample. Therefore, to clarify the relationship between malic acid and sour taste quality, further investigation using wines with higher malic acid concentration is necessary. Regarding succinic acid concentration, a negative correlation was observed with “soft”, “round”, “gentle”, “bright”, and “mild”. Succinic acid is known to contribute not only to the sour taste but also to the bitter and salty tastes. We offer a plausible explanation for these negative correlations; at higher concentrations, succinic acid may enhance bitter taste and salty taste perceptions, thereby suppressing sour taste qualities, such as “soft”. However, taste interactions are complex and depend on taste intensity. Keast and Breslin [36] demonstrated that bitter taste enhances sour taste at moderate intensities but suppresses it at high intensities. Similarly, salty taste enhances sour taste at moderate intensities, but they mutually suppress each other at high intensities. Because succinic acid concentration in wine is lower than tartaric, malic, and lactic acid concentrations, the bitter and salty tastes derived from succinic acid may not be as pronounced as the sour taste. Further investigations are required to elucidate the extent to which succinic acid suppresses sour taste perception. On the other hand, while sweet taste is known to suppress sour taste, no correlation was found between sugar concentration and sour taste qualities (Table 6). Blesić et al. [37] reported that sweet taste is perceptible in wine when sugar concentration exceeds 3.79 g/L. In this study, except for O1, the wines had significantly low sugar concentrations (The average sugar concentration of all wines except O1, was 0.837 g/L, Supplementary Materials, Table S6). On the basis of these findings, we concluded that low sugar concentration does not influence sour taste quality. In the case of O1, although a distinct sweet taste was not perceived during the wine selection process, its sugar concentration was relatively high (14.3 g/L). This could be attributed to the presence of astringent compounds, such as polyphenol compounds derived from skin fermentation, which might have masked the sweet taste perception

4. Conclusions

In this study, 56 terms related to sour taste quality were derived, and 12 were selected for further analysis. Definitions and reference standards were developed for these 12 terms, which were then used to establish a trained sensory panel. The panelists classified 16 Koshu wines into four clusters based on sour taste quality, indicating that sour taste perception involves not only intensity but also qualitative aspects. This suggests that wines can be categorized by differences in sour taste quality. The development of a flavor wheel for Koshu’s sour taste could provide valuable guidance for consumers when selecting Koshu wines.

The sour taste descriptors were classified according to the timing of their perception during tasting, and their relationships with various acid-related parameters were examined. Sour qualities perceived immediately after sipping wine correlated strongly with pH, while those perceived while holding the wine in the mouth were related to titratable acidity and pH. Sourness perceived after swallowing showed the strongest correlation with total acidity. These findings suggest that both the degree of pH decrease and the amount of saliva needed to restore oral pH contribute to sour taste perception.

Succinic acid showed a negative correlation with five sour taste descriptors. As succinic acid can also impart bitter and salty tastes, the possibility of taste–taste interactions should be considered. These interactions vary depending on taste intensity, suggesting that sour taste perception in wine is influenced not only by acid concentration and composition but also by the presence of other taste components.

Finally, climate change is affecting grape acid composition, which may lead to an increase in acid supplementation during winemaking. The relationships identified in this study between chemical parameters and sour taste qualities provide a basis for optimizing wine taste profiles to meet desired sensory characteristics. This study confirmed that even among wines made from the same grape cultivar, differences in sour taste quality can be perceived, suggesting that future comparisons among different grape cultivars may offer deeper insight into varietal characteristics.