Effects of aroma enhancement fermentation of Zygosaccharomyces rouxii ZR21 and Debaryomyces hansenii DH06 on the sensory properties and consumer liking of Yongchuan douchi

Abstract

Yongchuan douchi is a well-loved condiment. However, the aroma of rapid Yongchuan douchi is inferior to that of traditional Yongchuan douchi. The objective of this study was to improve the aroma quality of rapid Yongchuan douchi and evaluate the effect of aroma enhancement from the perspective of consumers. The aroma characteristics of samples were analyzed by consumers through flash profile (FP) (n = 15) and rate-all-that-apply (RATA) (n = 75). The results showed improvement in the aroma quality of rapid Yongchuan douchi with the two yeast strains, and consumers could perceive the modification of the aroma characteristics. It shows that the douchi aroma and sour aroma of rapid Yongchuan douchi increased significantly after aroma-enhancing fermentation, while the soy sauce aroma, soybean aroma, and musty aroma decreased. Similar results were obtained from FP and RATA. RATA showed that rapid Yongchuan douchi with aroma-enhancing fermentation aroused higher levels of positive emotions, which may increase the acceptance of consumers to rapid Yongchuan douchi. In summary, the aroma quality and consumer preference of rapid Yongchuan douchi were enhanced. The work provides insights into the quality improvement of rapid Yongchuan douchi from the perspective of consumers.

Keywords

Yongchuan douchi halophilic aromatic yeasts aroma enhancement RATA FP

INTRODUCTION

Yongchuan douchi is a famous condiment known in Chongqing China, categorized into traditional and rapid Yongchuan douchi according to the manufacturing process. Yongchuan douchi is abundant in protein and essential amino acids, with strong aroma and delicious taste. Therefore, it is widely loved by the people in Sichuan and Chongqing. Figure 1 shows the process details of traditional Yongchuan douchi and rapid Yongchuan douchi. The koji-making process of traditional Yongchuan douchi is primarily based on Mucor racemosus (Xie et al., 2023), a fungus that grows at low temperatures. Therefore, the production of traditional Yongchuan douchi is limited to winter. In combination with the long fermentation cycle, the traditional process seriously hinders its industrialization and large-scale production. This drives the development of rapid Yongchuan douchi, which uses pure Aspergillus oryzae to make koji. The post-fermentation time of rapid Yongchuan douchi is only 30 days, which greatly improves production efficiency. However, the inferior aroma in rapid Yongchuan douchi can be perceived by consumers (Yang et al., 2020).

The aroma of fermented food is a crucial factor that influences quality and preference of consumers (Seo et al., 2018). A considerable amount of studies has focused on the aroma quality improvement of fermented food (Liu et al., 2022; Lu et al., 2021; Xiao et al., 2021), which is significant for consumers and the industry. The microbial community plays a major role in the formation of aroma compounds (Chen et al., 2021). Compared to rapid Yongchuan douchi, the traditional Yongchuan douchi scored higher in microbial abundance and diversity during the post-fermentation stage. Zygosaccharomyces rouxii (Z. rouxii) and Debaryomyces hansenii (D. hansenii) were involved in the post-fermentation of traditional Yongchuan douchi, but they were not detected in the fermentation of rapid Yongchuan douchi (Li et al., 2020; Suo et al., 2015; Zeng et al., 2019). The difference in microflora may explain the difference in aroma quality between the two types of Yongchuan douchi.

Previous studies have shown that inoculating yeast during fermentation can promote the production of flavor substances, compensate for the lack of aroma, and improve the sensory quality (Gong et al., 2023; Jiang et al., 2021). Zhou reported an effective improvement in aroma by inoculating Z. rouxii and D. hansenii to rapid Yongchuan douchi (Zhou et al., 2022). Z. rouxii is an important microorganism widely applied in traditional fermented foods (Lee et al., 2014), which promotes the formation of alcohols and esters. D. hansenii is associated with caramel, butter, and fruit flavors in fermented soybean products (Jeong et al., 2022). It has been shown to produce aldehydes and alcohols in cheese (Sørensen et al., 2011) and sausage (Wen et al., 2022). According to our preliminary study, aldehydes were the predominant types of compounds in traditional Yongchuan douchi, and esters were the most abundant substances in variety. Therefore, if these substances could be increased or enriched by Z. rouxii and D. hansenii, the aroma quality and the acceptability of consumers could be improved.

The effect of microorganisms in aroma-enhancing fermentation has been an attractive research area. Until now, the majority of previous studies has merely aimed at changes in the quantity and content of volatile substances (Xiao et al., 2021). However, the perception of consumers is one of the criteria used to evaluate the effect of aroma enhancement. In recent years, the consumer-based methods were used to identify discrepancy in samples and seek for an appropriate alternative to descriptive analysis (Ares et al., 2018; Pineau et al., 2022; Wang et al., 2022). In addition to these, the consumer-based methods have a promising application to evaluate perception of consumers. Rate-all-that-apply (RATA) is an affective analysis method derived from check-all-that-apply (CATA) (Ares et al., 2014), in which consumers judge whether the sensory and emotional descriptors in the questionnaire are applicable to describe a given product and assess their strength. Thus, the RATA method can better reflect sensory and emotional variation of aroma enhancement of Yongchuan douchi than the CATA method (Vidal et al., 2018). FP is a free descriptive task based on the vocabulary constructed by consumers. The descriptors generated by the FP method can be used as candidate items in the CATA and RATA questionnaires (Galli et al., 2019), and provide a perspective for translating consumer words into evaluator descriptions. In practical studies, it has been found that combining affective and descriptive sensory methods can guide researchers to design products according to consumer preferences and market needs (Aguiar et al., 2020). Therefore, the combination of RATA and FP methods can be used to explore the aroma enhancement effect in Yongchuan douchi from the perspective of consumers.

We found that the quantity and content of volatile substances in rapid Yongchuan douchi was improved by inoculation with halophilic aromatic yeasts in our previous researches (Zhou et al., 2022). In this study, the aroma enhancement effect of inoculation with halophilic aromatic yeasts was evaluated from the perspective of consumers. The RATA questionnaire was used to analyze sensory changes in samples and emotional changes in consumers. FP was used to generate descriptors that the consumers understood and to differentiate between samples. The aim of this study was to improve the aroma quality of rapid Yongchuan douchi and evaluate the effect of aroma enhancement from the perspective of consumers. This will be complementary to the instrumental and quantitative descriptive analysis (QDA).

MATERIALS AND METHODS

Preparation of start culture strains

Two halophilic aromatic yeasts used in this study, Z. rouxii ZR21 and D. hansenii DH06, were previously isolated in our lab from traditional Yongchuan douchi (Zhou et al., 2022). Z. rouxii was cultured in 8% NaCl (m/v) of Malt Extract Broth (Solarbio, Beijing) at 28 °C for 3 days; D. hansenii was grown in 8% NaCl (m/v) of Yeast Peptone Dextrose Broth (Solarbio, Beijing) at 28 °C for 3 days.

Preparation of douchi

The rapid Yongchuan douchi was obtained from Yongchuan douchi Company (Chongqing, China). The aroma-enhanced douchi was produced from the rapid Yongchuan douchi by the inoculation of halophilic aroma yeasts during secondary fermentation, as shown in Figure 1.

The aroma-enhanced douchi was manufactured according to the following variations of fermentation (Table 1). The halophilic aroma yeasts were suspended in 0.90% (w/v) saline until the inoculation biomass of halophilic aroma yeasts was 10⁶ CFU/mL. The uninoculated rapid Yongchuan douchi was used as a control (sample A). The sample mono-inoculated with Z. rouxii was designated “AZ,” the sample mono-inoculated with D. hansenii was designated “AD,” and the samples co-inoculated were designated “ADZ.” The numbers “1,” “2,” and “3” were used to distinguish the samples with different inoculum amounts. All samples were produced on the same day according to the same production process. The samples were fermented at 25 ± 1 °C in an incubator for 1 month.

Table 1.

Information of aroma-enhanced Yongchuan douchi.

Samples	Strains of vaccination rate (v/m)
Samples	Zygosaccharomyces rouxii	Debaryomyces hansenii
A	–	–
AD	–	1%
AZ	1%	–
ADZ1	0.5%	0.5%
ADZ2	1%	1%
ADZ3	2%	2%

A: uninoculated samples; AD: inoculated with D. hansenii; AZ: inoculated with Z. rouxii; ADZ: inoculated with both yeasts. The numbers “1,” “2,” and “3” are used to distinguish the samples with different inoculum levels.

Before FP and RATA, samples (5 g each) were placed in a 30 mL plastic cup, which were then covered and sealed. The samples were encoded with a three-digit code and then presented to the panelists at 25 °C.

Sensory analysis

The panelists who participate in FP and RATA were recruited from Southwest University by issuing online questionnaires. The questionnaire included basic information such as age, sex, dietary habits, history of allergy to food and the subjects’ interest in Yongchuan douchi. Participants who had consumed Yongchuan douchi in the previous three months were invited to take part in the experiment. Fifteen and seventy-five consumers with no restrictions on the Yongchuan douchi were recruited for FP and RATA respectively. After he official experiment, participants have received financial compensation for their participation after the experiments.

All testing was conducted in individual sensory booths using sensory systems (https://www.cloudsensorylab.com). The trials took place in booths under white fluorescent light and adequate ventilation, at a temperature of 25 °C. At the beginning of the experiment, all samples were placed on trays. During the FP process, the samples were presented simultaneously to the panelists in each booth. During the RATA process, the samples were presented monadically according to a randomized serving order. Each panelist was also provided with a warm clean washcloth to cleanse their nasal cavities and relieve olfactory fatigue. Each session lasted 90 min, and panelists had sufficient time to rest between samples.

Flash profile (FP)

The FP descriptive analysis was based on the methodology described by Wang et al. (2022) and slightly modified. In this experiment, a discussion session was added between the two sessions. The specific process was as follows: (1) first of all, the 15 panelists were requested to list the sensory characteristics of samples, which should be significantly discriminant to allow a ranking and avoid using hedonic terms. In session 1, a total of 22 descriptors were generated. (2) At the beginning of session 2, all the attributes were summed up and discussed individually by panelists. Attributes with the same meaning but different expressions were combined, and the hedonic terms were removed. After discussion, 12 aroma terms were obtained to describe Yongchuan douchi. And, then each panelist could add or replace his/her sensory characteristics, and established a specific and personalized sensory attributes list. (3) Finally, all the samples were presented again, and the panelists were requested to rank the samples in a 15 cm line segment. The scales on the left were “low intensity” and the scales on the right were “high intensity.” In addition, the panelists were allowed to use the same grade for two or more samples if no difference was found between the attribute of the samples.

RATA

The RATA method is based on the method described by Giacalone and Hedelund (2016). The RATA questionnaire contains 18 aroma attributes and 40 emotional terms. The alternative sensory attributes in the RATA questionnaire were picked up from two parts: (1) all the attributes written by panelists in the FP method, (2) the description of Yongchuan douchi in the literature. The classification criteria of emotional terms were described previously (Barker et al., 2021; King and Meiselman, 2010; Nestrud et al., 2016). As stated in King’s research, the consumers were required to categorize the terms according to their emotional perception, such as positive and negative. Terms selected by more than 60% of the consumers were categorized as positive or negative, and terms selected by less than 50% were grouped as no clear classification.

All sensory attributes and emotional terms were ranked using a 3-point scale to rate the strength of the applicable attributes. In the 3-point scale of RATA, 1 = attribute slightly perceived and 3 = intense was used in sensory attributes, 1 = dislike extremely and 3 = like extremely was used in emotional attributes. Terms, which were not applicable to consumers, were left blank. In addition, there was a blank field at the end of the questionnaire for consumers to add any new terms. The position of the terms in the questionnaire was balanced and random in each list.

Data analysis

The frequency of terms was determined by counting the number of consumers who used terms in session 2 of FP. Generalized procrustes analysis (GPA) was applied to FP data to evaluate the consensus between the sensory maps of panelists. In RATA, the average score of sensory attributes and emotional terms were calculated. The one-way analysis of variance (ANOVA) was analyzed to evaluate the difference between the average score of samples (p ≤ 0.05). Principal component analysis (PCA) was carried out to identify the relation between samples and terms. The distance between samples in the GPA and PCA load plots was used to assess sample discrepancy. All statistical analyses were performed using R Version 4.2.0.

RESULTS

FP

After the discussion session of FP, 12 aroma terms were obtained to describe Yongchuan douchi. Figure 2 lists the aroma terms and frequency of Yongchuan douchi based on consumer perspective. The result showed that soy sauce, beany, musty, and douchi were the typical aroma profiles of Yongchuan douchi (the frequency over 9).

Figure 3A presents the load plot obtained by GPA using the FP data to evaluate the differences between the samples. As shown in Figure 3A, the control (sample A) is far away from the other five samples, suggesting that the aroma of sample A was different from the other five samples. Therefore, the halophilic aromatic yeasts influence the aroma of rapid Yongchuan douchi. The samples AD and AZ were distant from each other, indicating that the samples inoculated with different yeasts had different effects on the aroma. The samples AZ deviated from the co-inoculated samples (ADZ1, ADZ2, and ADZ3), probably because the fermentation by inoculating Z. rouxii had a large effect on the certain aroma of samples. The co-inoculated samples (ADZ1, ADZ2, and ADZ3) were closer to AD, indicating that the aromatic effect of the co-inoculated samples might be more similar to the mono-inoculated sample with D. hansenii. The samples ADZ1, ADZ2, and ADZ3 were close to each other, indicating that the aromatic effect of the co-inoculated samples was similar, and the amount of inoculum had little effect on the aroma characteristic. The results of FP showed that consumers can perceive differences in the uninoculated douchi, the mono-inoculated samples, and the co-inoculated samples. However, the process of the FP does not include the identification of preferences. Therefore, it is necessary to combine FP and RATA to evaluate the aromatic effect from the perspective of consumers.

RATA

Sensory attribute analysis

In the RATA data, the missing attribute score is 0 and the existence attribute was 1–3. The sensory attributes were scored according to the perceived intensity, and the average score of the final analysis intensity was obtained. As shown in Table 2, the Chinese herb, umami aromatics, soy sauce, and douchi aromas showed significant differences among the samples (p ≤ 0.05). This result shows that differences between samples can be distinguished based on these four properties. The douchi aroma of the co-inoculated samples improved to different degrees. Among the six samples, the douchi aroma of ADZ2 was the most intense and significantly different from all other samples. Compared with sample A, the score of smoky in ADZ2 and AZ showed signs of increase. It indicates that rapid Yongchuan douchi is endowed with smoke aroma after inoculation with Z. rouxii. A similar conclusion was reached by He et al. (2016). The results of the scores indicate that the sensory attributes changed after aroma-enhancing fermentation. It can be perceived by the consumers.

Table 2.

The strength of aroma attributes and emotion terms from RATA for douchi samples (n = 75).

	A	AD	AZ	ADZ1	ADZ2	ADZ3
Chinese herb	0.853 ± 1.023^b	0.387 ± 0.769^a	0.573 ± 0.947^ab	0.733 ± 0.991^ab	0.56 ± 0.919^ab	0.36 ± 0.71^a
Musty	0.613 ± 0.899^a	0.48 ± 0.875^a	0.587 ± 0.946^a	0.627 ± 0.983^a	0.52 ± 0.906^a	0.56 ± 1.03^a
Smoky	0.333 ± 0.723^a	0.267 ± 0.723^a	0.413 ± 0.824^a	0.333 ± 0.759^a	0.547 ± 0.89^a	0.44 ± 0.889^a
Umami aromatics	0.427 ± 0.903^b	0.467 ± 0.875^b	0.56 ± 0.99^b	0.04 ± 0.346^a	0.373 ± 0.835^ab	0.36 ± 0.765^ab
Rubbery	0.04 ± 0.197^a	0.067 ± 0.38^a	0.067 ± 0.342^a	0.12 ± 0.519^a	0.16 ± 0.57^a	0.067 ± 0.342^a
Fermented	1.587 ± 1.231^a	1.307 ± 1.115^a	1.173 ± 1.178^a	1.173 ± 1.083^a	1.173 ± 1.178^a	1.107 ± 1.214^a
Brunt	0.533 ± 0.859^a	0.56 ± 0.793^a	0.613 ± 0.884^a	0.52 ± 0.906^a	0.707 ± 0.941^a	0.427 ± 0.738^a
Soy sauce	2.027 ± 1.23^b	1.467 ± 1.131^a	1.52 ± 1.143^ab	1.52 ± 1.07^ab	1.4 ± 1.115^a	1.707 ± 1.063^ab
Beany	0.773 ± 1.034^a	0.547±±0.963^a	0.667 ± 0.977^a	0.6 ± 0.854^a	0.667 ± 1.004^a	0.573 ± 0.918^a
Bitter aromatics	0.413 ± 0.755^a	0.533 ± 0.949^a	0.48 ± 0.795^a	0.427 ± 0.825^a	0.573 ± 0.932^a	0.547 ± 0.963^a
Baked	0.24 ± 0.612^a	0.107 ± 0.352^a	0.187 ± 0.562^a	0.187 ± 0.562^a	0.267 ± 0.704^a	0.107 ± 0.452^a
Soybean	0.56 ± 0.919^a	0.653 ± 0.993^a	0.6 ± 0.93^a	0.64 ± 0.88^a	0.547 ± 0.81^a	0.68 ± 0.918^a
Peanut butter	0.173 ± 0.529^a	0.147 ± 0.456^a	0.147 ± 0.456^a	0.147 ± 0.425^a	0.16 ± 0.521 ^a	0.187 ± 0.538^a
Sweet aromatics	0.133 ± 0.475^a	0.187 ± 0.456^a	0.2 ± 0.658^a	0.24 ± 0.75^a	0.187 ± 0.586^a	0.16 ± 0.546^a
Roast	0.133 ± 0.502^a	0.08 ± 0.395^a	0.133 ± 0.475^a	0.093 ± 0.374^a	0.16 ± 0.521^a	0.093 ± 0.336^a
Sour aromatics	0.667 ± 1.07^a	0.573 ± 0.873^a	0.6 ± 0.838^a	0.653 ± 0.878^a	0.693 ± 0.944^a	0.787 ± 1.082^a
Alcohol	0.853 ± 1.036^a	0.52 ± 0.891^a	0.653 ± 0.979^a	0.44 ± 0.74^a	0.653 ± 0.923^a	0.8 ± 1.103^a
Douchi	1.307 ± 0.838^a	1.453 ± 1.142^a	1.493 ± 1.07^a	1.507 ± 1.201^a	2.027 ± 0.93^b	1.56 ± 1.13^ab
Fruity	0 ± 0^a	0.08 ± 0.359^a	0.067 ± 0.3^a	0.093 ± 0.408^a	0.067 ± 0.342^a	0.04 ± 0.197^a
Active	0.56 ± 1.056^b	0.187 ± 0.63^a	0.04 ± 0.257^a	0.16 ± 0.638^a	0.16 ± 0.57^a	0.08 ± 0.427^a
Polite	0.147 ± 0.538^a	0.053 ± 0.28^a	0.093 ± 0.408^a	0.107 ± 0.481^a	0.133 ± 0.577^a	0.053 ± 0.364^a
Good-natured	0.64 ± 1.061^a	0.507 ± 1.032^a	0.467 ± 0.977^a	0.387 ± 0.899^a	0.52 ± 0.978^a	0.347 ± 0.762^a
Bored	0 ± 0^a	0.253 ± 0.699 ^ac	0.267 ± 0.577^ac	0.16 ± 0.494^ab	0.347 ± 0.726^bc	0.467 ± 0.977^c
Steady	0.573 ± 1.068^a	0.6 ± 1.065^a	0.453 ± 0.89^a	0.32 ± 0.774^a	0.373 ± 0.835^a	0.44 ± 0.858^a
Pleasant	0.987 ± 1.257^b	0.28 ± 0.781^a	0.267 ± 0.741^a	0.24 ± 0.675^a	0.2 ± 0.678^a	0.24 ± 0.654^a
Energetic	0.413 ± 0.96^b	0.147 ± 0.586^ab	0.08 ± 0.427^a	0.2 ± 0.697^ab	0.067 ± 0.414^a	0.08 ± 0.427^a
Whole	0.36 ± 0.88^a	0.12 ± 0.544^a	0.16 ± 0.616^a	0.187 ± 0.608^a	0.107 ± 0.481^a	0.147 ± 0.538^a
Disgusted	0.933 ± 1.288^b	0.173 ± 0.578^a	0.133 ± 0.577^a	0.16 ± 0.638^a	0.213 ± 0.684^a	0.147 ± 0.562^a
Guilty	0 ± 0^a	0.027 ± 0.231^a	0.067 ± 0.414^a	0 ± 0^a	0.067 ± 0.342^a	0.053 ± 0.324^a
Calm	0.12 ± 0.544^a	0.227 ± 0.669^a	0.427 ± 0.932^a	0.16 ± 0.546^a	0.413 ± 0.946^a	0.32 ± 0.903^a
Tame	0.4 ± 0.885^a	0.587 ± 0.974^a	0.573 ± 1.002^a	0.64 ± 1.074^a	0.853 ± 1.135^a	0.813 ± 1.123^a
Merry	1.107 ± 1.341^b	0.147 ± 0.586^a	0.16 ± 0.616^a	0.16 ± 0.594^a	0.253 ± 0.68^a	0.2 ± 0.658^a
Enthusiastic	0.413 ± 0.96^b	0.053 ± 0.324^a	0.067 ± 0.342^a	0.107 ± 0.535^a	0.093 ± 0.44^a	0.053 ± 0.226^a
Wild	0.133 ± 0.502^a	0.147 ± 0.538^a	0.067 ± 0.342^a	0.227 ± 0.628^a	0.147 ± 0.512^a	0.133 ± 0.475^a
Glad	0.707 ± 1.171^b	0.147 ± 0.538^a	0.147 ± 0.586^a	0.187 ± 0.63^a	0.173 ± 0.623^a	0.093 ± 0.498^a
Good	0.36 ± 0.91^a	0.173 ± 0.665^a	0.173 ± 0.685^a	0.16 ± 0.616^a	0.093 ± 0.47^a	0.133 ± 0.577^a
Aggressive	0.133 ± 0.502^a	0.427 ± 0.825^ab	0.52 ± 0.978^b	0.467 ± 0.92^ab	0.373 ± 0.785^ab	0.387 ± 0.837^ab
Secure	0.8 ± 1.263^b	0.173 ± 0.578^a	0.227 ± 0.669^a	0.253 ± 0.66^a	0.2 ± 0.615^a	0.2 ± 0.658^a
Pleased	0.64 ± 1.158^b	0.093 ± 0.44^a	0.093 ± 0.44^a	0.053 ± 0.364^a	0.08 ± 0.427^a	0.093 ± 0.44^a
Joyful	0.787 ± 1.222^b	0.067 ± 0.414^a	0.08 ± 0.427^a	0.133 ± 0.577^a	0.04 ± 0.346^a	0.093 ± 0.44^a
Affective	0.333 ± 0.859^b	0 ± 0^a	0.053 ± 0.324^a	0.067 ± 0.414^a	0 ± 0^a	0.027 ± 0.231^a
Quiet	0.08 ± 0.395^a	0.373 ± 0.882^a	0.373 ± 0.866^a	0.267 ± 0.704^a	0.187 ± 0.608^a	0.32 ± 0.774^a
Happy	0.587 ± 1.128^b	0.16 ± 0.616^a	0.12 ± 0.544^a	0.107 ± 0.535^a	0.027 ± 0.231^a	0.053 ± 0.364^a
Daring	0.187 ± 0.672^b	0.04 ± 0.346^ab	0.08 ± 0.427^ab	0.067 ± 0.414^ab	0 ± 0^a	0.013 ± 0.115^ab
Tender	0.4 ± 0.854^a	0.213 ± 0.703^a	0.267 ± 0.794^a	0.2 ± 0.637^a	0.227 ± 0.649^a	0.2 ± 0.615^a
Mild	0.293 ± 0.731^a	0.44 ± 0.842^a	0.533 ± 0.92^a	0.507 ± 0.991^a	0.44 ± 0.858^a	0.493 ± 0.906^a
Free	0.213 ± 0.703^a	0.16 ± 0.521^a	0 ± 0^a	0.173 ± 0.578^a	0.04 ± 0.257^a	0.08 ± 0.395^a
Adventurous	0.373 ± 0.818^a	0.68 ± 1.055^a	0.76 ± 1.195^a	0.693 ± 1.065^a	0.493 ± 0.891^a	0.427 ± 0.808^a
Satisfied	0.733 ± 1.212^b	0.293 ± 0.835^a	0.24 ± 0.714^a	0.347 ± 0.846^ab	0.227 ± 0.649^a	0.173 ± 0.554^a
Worried	0.373 ± 0.912^b	0.12 ± 0.519^ab	0.12 ± 0.492^ab	0.147 ± 0.608^ab	0.053 ± 0.324^a	0.2 ± 0.637^ab
Loving	0.027 ± 0.162^a	0.56 ± 0.874^b	0.493 ± 0.921^b	0.653 ± 1.072^b	0.467 ± 0.844^b	0.56 ± 0.919^b
Understanding	0.293 ± 0.785^a	0.08 ± 0.359^a	0.28 ± 0.727^a	0.307 ± 0.838^a	0.16 ± 0.57^a	0.373 ± 0.835^a
Nostalgic	0.307 ± 0.838^a	0.107 ± 0.421^a	0.147 ± 0.562^a	0.2 ± 0.593^a	0.173 ± 0.529^a	0.213 ± 0.664^a
Friendly	0.573 ± 1.093^a	0.293 ± 0.785^a	0.48 ± 1.005^a	0.307 ± 0.771^a	0.467 ± 0.859^a	0.213 ± 0.622^a
Interested	0 ± 0^a	0.2 ± 0.493^a	0.16 ± 0.616^a	0.16 ± 0.546^a	0.147 ± 0.512^a	0.093 ± 0.374^a
Eager	0.52 ± 1.07^b	0.067 ± 0.414^a	0.173 ± 0.623^a	0.08 ± 0.487^a	0.16 ± 0.57^a	0.027 ± 0.162^a
Warm	0.4 ± 0.9^b	0.107 ± 0.481^a	0.12 ± 0.434^ab	0.107 ± 0.481^a	0.2 ± 0.717^ab	0.133 ± 0.577^ab
Peaceful	0.36 ± 0.816^a	0.56 ± 1.056^a	0.4 ± 0.9^a	0.307 ± 0.735^a	0.547 ± 1.017^a	0.4 ± 0.9^a

Superscripts a, b, c indicate the strength of aroma attributes and emotion terms between samples statistically significant (p < 0.05). RATA: rate-all-that-apply; A: uninoculated samples; AD: inoculated with D. hansenii; AZ: inoculated with Z. rouxii; ADZ: inoculated with both yeasts.

PCA was performed on RATA, and the result of load diagram was showed in Figure 3B. The two-dimensional models F1 and F2 of PCA explain 62.8% of the total variance, and the first and second principal components, F1 and F2, explained 35.4% and 27.4%, respectively. Among them, Chinese herb, umami aromatics, and soy sauce were positively correlated with the first principal component, whereas douchi was negatively correlated with the first principal component. The distance between sensory attributes and samples in the load plot was used to assess correlation. As shown in Figure 3B, the rapid Yongchuan douchi inoculation with yeasts was more correlated with douchi, umami aromatics, soybean, and other properties compared with sample A.

In the PCA results, a shorter distance between two products represents a higher degree of similarity, so deviations can be observed for all samples. There is a long distance between sample A and the aroma-enhanced fermentation sample. There is also a certain distance between the mono-inoculated samples and the co-inoculated samples. It shows that the halophilic aromatic yeast can change the aroma characteristics of rapid Yongchuan douchi, and the different yeasts have different effects. As presented in Figure 3B, it can be clearly observed that sample A is on the right side, while the other samples are distributed on the left side, which once again verified that halophilic aromatic yeast has a significant effect on the aroma of Yongchuan douchi.

Emotional attribute analysis

Sensory characteristics of food affect emotion, and researchers have identified emotion as a key variable in determining food choices (Eertmans et al., 2006). As shown in Table 2, active, bored, pleasant, energetic, merry, enthusiastic, glad, secure and other emotional descriptors showed significant differences between samples (p ≤ 0.05). Sample A also showed significant differences compared with other samples in these properties. Consumers’ attitudes have changed towards aroma-enhanced Yongchuan douchi. The results of the sensory and emotional attributes show that as the sensory properties change, so do the emotion of the consumer.

The results of PCA analysis of the emotional descriptor in RATA are shown in Figure 3C The two-dimensional models F1 and F2 of PCA explain 81.2% of the total variance, and the first and second principal components, F1 and F2, explain 71.2% and 10%, respectively. According to King's classification of emotional terms, the emotional descriptors in this experiment were classified, as shown in Table 3. Sample A was significantly separated from the rest of the samples in the PCA plot, which is consistent with the results of the FP method. The result of PCA plot suggests that aroma-enhanced Yongchuan douchi were more related to positive affective descriptors. For example, ADZ3 was correlated with positive affective terms such as peaceful and calm, while sample A was correlated with negative affective terms such as worried and disgusted. It shows that the halophilic aroma yeasts can change the consumer’s emotional evaluation from negative to positive after aroma-enhancing fermentation. Combined with the results of the sensory questionnaire, we can infer that the sentiment of consumers towards the sample is correlated with the sensory attributes.

Table 3.

Emotion terms classification of RATA.

Positive	Negative	No clear classification
Adventurous, energetic, enthusiastic, friendly, glad, good, good-natured, happy, interested, joyful, loving, merry, nostalgic, peaceful, pleasant, pleased, satisfied, tender, warm, whole, active, affective, calm, secure	Disgusted, bored, worried	Aggressive, daring, eager, guilty, mild, polite, quiet, steady, tame, understanding, wild

The classification criteria of emotional terms were described previously (King and Meiselman, 2010). As stated in King’'s research, the consumers were required to categorize the terms according to their emotional perception, such as positive and negative. Terms selected by more than 60% of the consumers were categorized as positive or negative, and terms selected by less than 50% were grouped as no clear classification. RATA: rate-all-that-apply.

DISCUSSION

Based on the result of FP and RATA sensory analysis, sensory differences were found between different inoculation of aroma-enhanced Yongchuan douchi. In the FP method, the panelists generated a series of descriptors. Among above, douchi, beany, Chinese herb, soy sauce, musty, fermented, and alcohol were mentioned in both FP and RATA. These terms were selected by Zhou et al. (2022) to perform QDA, with the exception of fermented. This indicates that the perception of consumers about Yongchuan douchi is similar to those of trained assessors. In the RATA method, douchi, soy sauce, and fermented are the most frequently used by consumers. While soy sauce, beany, douchi, and musty are the most frequently used sensory attributes in the FP method. The RATA method successfully identified differences between samples among the four attributes of Chinese herb, umami aromatics, soy sauce, and douchi. The result combine FP and RATA showed that soy sauce and douchi were the typical aroma characteristics shared by all douchi samples in this research.

As a result of sensory questionnaire in the RATA method showed that aroma characteristics changed in the aroma-enhanced Yongchuan douchi. It may be due to esters, alcohols, and phenols were produced by Z. rouxii and D. hansenii during the aroma-enhanced fermentation (Zhou et al., 2022). Esters have a smell similar to fruity such as pineapple and apple scent (Niu et al., 2020). The increase of fruity aroma in aroma-enhanced Yongchuan douchi might be owe to the esters, which are produced by Z. rouxii (Yang et al., 2021) and D. hansenii (Wen et al., 2023). Moreover, the aroma-enhanced Yongchuan douchi showed an increase in the sweet aroma might related to alcohols formed by Z. rouxii (Devanthi et al., 2018; Zhang et al., 2022). The phenols, such as 4-vinyl guaiacol, contribute to the smoky aroma. D. hansenii converted ferulic acid into 4-vinyl guaiacol (Mathew et al., 2007), which may account for the increase of smoky aroma. However, the actual relationship between variation in aroma characteristics and variation in substances is unclear. In the future, GC × GC−O−MS, recombination and omission experiments will be considered for more in-depth studies.

As a result of emotional questionnaire in the RATA method showed that the consumers’ attitudes have changed towards aroma-enhanced Yongchuan douchi. In food development, failure is more common than success. Two-thirds of new products never survive their second year on the shelves (Linnemann et al., 2006). Paying close attention to consumer emotion is a promising way to successfully improve product development. Related research shows that sensory characteristics of food are correlated with consumer emotion (King and Meiselman, 2010). Thus, it is necessary to evaluate the variation in consumer emotion. In this research, the number of emotional terms was sufficient to obtain more complete emotional information from consumers (Nestrud et al., 2016). The result of this study provided a consumer perspective, and was of great significance for process optimization of Yongchuan douchi.

It should be pointed out that we merely considered the aroma of Yongchuan douchi itself. Different cooking methods with Yongchuan douchi could provide variation in the result. However, the potential cooking variation of Yongchuan douchi in China is diverse and innovative. For example, twice-cooked pork is one of the most traditional cooking methods. Oils, peppers, and other seasonings interact with Yongchuan douchi at different temperatures, which can enhance or suppress douchi aroma. Therefore, our further research will consider more detailed cooking methods with Yongchuan douchi.

CONCLUSION

The combination of RATA and FP methods can be used to explore the aroma enhancement effect from the sensory and emotional perspective of consumers. The douchi aroma and sour aroma of aroma-enhanced Yongchuan douchi increased significantly while soy sauce aroma, soybean aroma, and musty aroma decreased. In the result of the emotional questionnaire, the aroma-enhanced Yongchuan douchi aroused more positive emotions in consumers than the rapid Yongchuan douchi. Consumers preferred the ADZ2 in all samples. Therefore, co-inoculation with Z. rouxii ZR21 and D. hansenii DH06 is an effective way to improve aroma quality and consumer acceptance of rapid Yongchuan douchi.

Footnotes

ACKNOWLEDGMENTS

The authors would like to acknowledge each of the sensory panelists for their contribution to this study.

DECLARATION OF CONFLICTING INTERESTS

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

FUNDING

This work was supported by the Project of University Innovation Research Group Project of Chongqing (CXQT21007) and Chongqing Science and Technology Specialists Team Project (cstc2021jscx-tpyzxX0009).

ETHICS APPROVAL

There is no human ethics committee at Southwest University. Our samples and yeasts satisfy Chinese food safety standards and are not harmful to humans. Participants were recruited through an online questionnaire and volunteered to participate in the experiment. All participants were fully informed of the study requirements and risks before the experiment began. The appropriate protocols for protecting the rights and privacy of all participants were utilized during the execution of the research.

Participants gave informed consent via the statement “I am aware that my responses are confidential, and I agree to participate in this survey where an affirmative reply was required to enter the survey.” They were able to withdraw from the survey at any time without giving a reason. The products tested were safe for consumption. The study was explained to consumers in the online questionnaire and formal experiments. They were informed that they would participate in the survey using their personal smartphone, and that all data will be de-identified and only reported in the aggregate. All participants acknowledged an informed consent statement in order to participate in the study. They were financially compensated for their participation in the FP and RATA.

ORCID iDs

Hongwei Wang

Huayi Suo

References

Aguiar

LAD

Rodrigues

Queiroz

VAV

, et al. (2020) Comparison of two rapid descriptive sensory techniques for profiling and screening of drivers of liking of sorghum breads. Food Research International 131: 108999.

Ares

Bruzzone

Vidal

, et al. (2014) Evaluation of a rating-based variant of check-all-that-apply questions: Rate-all-that-apply (RATA). Food Quality and Preference 36: 87–95.

Ares

Picallo

Coste

, et al. (2018) A comparison of RATA questions with descriptive analysis: Insights from three studies with complex/similar products. Journal of Sensory Studies 33: e12458.

Barker

Moss

Mcsweeney

(2021) Carbonated emotions: Consumers’ sensory perception and emotional response to carbonated and still fruit juices. Food Research International 147: 110534.

Chen

Huang

, et al. (2021) Microbial diversity and flavor of Chinese rice wine (Huangjiu): An overview of current research and future prospects. Current Opinion in Food Science 42: 37–50.

Devanthi

PVP

Linforth

Onyeaka

, et al. (2018) Effects of co-inoculation and sequential inoculation of Tetragenococcus halophilus and Zygosaccharomyces rouxii on soy sauce fermentation. Food Chemistry 240: 1–8.

Eertmans

Victoir

Notelaers

, et al. (2006) The food choice questionnaire: Factorial invariant over western urban populations? Food Quality and Preference 17: 344–352.

Galli

Baptista

Cavalheiro

, et al. (2019) Lactobacillus rhamnosus GG improves the sensorial profile of camembert-type cheese: An approach through flash-profile and CATA. Lwt 107: 72–78.

Giacalone

Hedelund

(2016) Rate-all-that-apply (RATA) with semi-trained assessors: An investigation of the method reproducibility at assessor-, attribute- and panel-level. Food Quality and Preference 51: 65–71.

10.

Gong

Chen

, et al. (2023) Evaluation and selection of yeasts as potential aroma enhancers for the production of dry-cured ham. Food Science and Human Wellness 12: 324–335.

11.

Huang

Liang

, et al. (2016) Comparing the differences of characteristic flavour between natural maturation and starter culture for Mucor-type douchi. International Journal of Food Science & Technology 51: 1252–1259.

12.

Jeong

Yoo

Jeon

, et al. (2022) Genomic features, aroma profiles, and probiotic potential of the Debaryomyces hansenii species complex strains isolated from Korean soybean fermented food. Food Microbiology 105: 104011.

13.

Jiang

Peng

Zhang

, et al. (2021) Effect of aroma-producing yeasts in high-salt liquid-state fermentation soy sauce and the biosynthesis pathways of the dominant esters. Food Chemistry 344: 128681.

14.

King

Meiselman

(2010) Development of a method to measure consumer emotions associated with foods. Food Quality and Preference 21: 168–177.

15.

Lee

Choi

, et al. (2014) Comparative volatile profiles in soy sauce according to inoculated microorganisms. Bioscience, Biotechnology, and Biochemistry 77: 2192–2200.

16.

Luo

Qiu

, et al. (2020) Microbial community structure and dynamic succession in traditional fermentation of Mucor-type Yongchuan douchi. Food and Fermentation Industries 46: 60–67.

17.

Linnemann

Benner

Verkerk

, et al. (2006) Consumer-driven food product development. Trends in Food Science & Technology 17: 184–190.

18.

Liu

Laaksonen

, et al. (2022) Lactic acid bacteria incubation and aging drives flavor enhancement of goji berry juice. Journal of Food Composition and Analysis 105: 104202.

19.

Yang

, et al. (2021) Bio-augmented effect of Bacillus amyloliquefaciens and Candida versatilis on microbial community and flavor metabolites during Chinese horse bean-chili-paste fermentation. International Journal of Food Microbiology 351: 109262.

20.

Mathew

Abraham

Sudheesh

(2007) Rapid conversion of ferulic acid to 4-vinyl guaiacol and vanillin metabolites by Debaryomyces hansenii. Journal of Molecular Catalysis B: Enzymatic 44: 48–52.

21.

Nestrud

Meiselman

King

, et al. (2016) Development of EsSense25, a shorter version of the EsSense Profile®. Food Quality and Preference 48: 107–117.

22.

Niu

Zhang

Xiao

, et al. (2020) Evaluation of the perceptual interactions between higher alcohols and off-odor acids in Laimao Baijiu by σ–τ plot and partition coefficient. Journal of Agricultural and Food Chemistry 68: 14938–14949.

23.

Pineau

Girardi

Lacoste Gregorutti

, et al. (2022) Comparison of RATA, CATA, sorting and napping® as rapid alternatives to sensory profiling in a food industry environment. Food Research International 158: 111467.

24.

Seo

Lee

Singh

, et al. (2018) Evaluating the headspace volatolome, primary metabolites, and aroma characteristics of koji fermented with bacillus Amyloliquefaciens and Aspergillus oryzae. Journal of Microbiology and Biotechnology 28: 1260–1269.

25.

Sørensen

Gori

Petersen

, et al. (2011) Flavour compound production by Yarrowia lipolytica, Saccharomyces cerevisiae and Debaryomyces hansenii in a cheese-surface model. International Dairy Journal 21: 970–978.

26.

Suo

Zhao

Qian

, et al. (2015) Changes in microbial load and flora during fermentation Yongchuan mucor-type douchi. Food Science 36: 124–131.

27.

Vidal

Ares

Hedderley

, et al. (2018) Comparison of rate-all-that-apply (RATA) and check-all-that-apply (CATA) questions across seven consumer studies. Food Quality and Preference 67: 49–58.

28.

Wang

Feng

Suo

, et al. (2022) Comparison of the performance of the same panel with different training levels: Flash profile versus descriptive analysis. Food Quality and Preference 99: 104582.

29.

Wen

Dong

, et al. (2023) Comparative evaluation of the flavour-promoting role of autochthonous yeast strains on dry sausages. Lwt 184: 115032.

30.

Wen

Yin

, et al. (2022) Technological properties and flavour formation potential of yeast strains isolated from traditional dry fermented sausages in Northeast China. Lwt - Food Science and Technology 154: 112853.

31.

Xiao

Liu

, et al. (2021) The effect of Eurotium cristatum (MF800948) fermentation on the quality of autumn green tea. Food Chemistry 358: 129848.

32.

Xie

Wang

Zeng

, et al. (2023) Whole-genome and comparative genome analysis of Mucor racemosus C isolated from Yongchuan douchi. International Journal of Biological Macromolecules 234: 123397.

33.

Yang

, et al. (2021) Ester synthesis mechanism and activity by Bacillus licheniformis, Candida etchellsii, and Zygosaccharomyces rouxii isolated from Chinese horse bean chili paste. Journal of the Science of Food and Agriculture 101: 5645–5651.

34.

Yang

Suo

Wang

(2020) Application of flash profile method in sensory evaluation of douchi. China Brewing 39: 181–184.

35.

Zeng

Luo

Suo

(2019) Dynamic change rule of microflora in Yongchuan mucor-type fermented soya beans during fermentation. China Condiment 44: 42–45. 64.

36.

Zhang

Liu

Sun

, et al. (2022) Metagenomics-based insights into the microbial community profiling and flavor development potentiality of Baijiu Daqu and Huangjiu Wheat Qu. Food Research International 152: 110707.

37.

Zhou

Wang

Zhang

(2022) Effects of halophilic aromatic yeast on the aroma of rapid Yongchuan douchi by sensory evaluation and GC-MS analysis. Food and Fermentation Industries 48: 59–64.