Exploration of the differences in aroma characteristics and chemical basis of Jingshan green tea from different tea cultivars

Abstract

Tea varieties play a crucial role in shaping the quality of Jingshan green tea (JSGT). Nevertheless, how the tea cultivars influence the quality of JSGT is still unknown. This study aimed to investigate the flavor substance profiles and their chemical basis in JSGT derived from five distinct tea cultivars, and to establish correlations between them, thereby providing a theoretical reference for cultivar selection based on flavor and health efficacy. Results showed that 9 key aroma-active substances, such as linalool and its oxides, geraniol, (Z)-3-hexen-1-ol were detected from 179 volatile substances. Furthermore, 17 non-volatile compounds, such as EC, _L-Threonic acid and EGC significantly influencing the taste characteristics of JSGT. The antioxidant capacity and α-glucosidase inhibitory activity of the five samples were significantly correlated with catechins and other components. Overall, this study holds important implications for improving JSGT quality, exploring its health benefits, and enhancing market competitiveness.

Highlights

• •Revealed significant differences in aroma, taste and bioactivity among five Jingshan green tea cultivars.
• •Characteristic volatiles and non-volatiles in Jingshan green tea cultivars were confirmed by GC–MS, GC-O-MS and LC-MS.
• •Screened nine key volatile compounds in Jingshan green tea.
• •Determined 17 non-volatile compounds significantly influencing taste profiles.
• •The antioxidant capacity and α-glucosidase inhibitory activity were strongly correlated with catechins.

1. Introduction

Tea is universally recognized as one of the three most prevalent and culturally non-alcoholic beverages (Ray et al., 2021). Green tea is gaining popularity among the different kinds of tea owing to its well-recognized health-enhancing properties as well as its distinct flavor and aroma. (Flaig et al., 2020; Ocieczek et al., 2023; Zhou et al., 2023). JSGT is considered as the quintessential roasted green tea which originated in Hangzhou, Zhejiang Province (Hou et al., 2024). The molecular composition of green tea, which is determined by its fresh leaves, is directly related to its health benefits, sensory qualities and price, thereby influencing consumers' choices and demands. (Shevchuk et al., 2018). The organoleptic characteristics of green tea is in relation to a lot of factors, such as the brewing conditions (Zhang, Feng, et al., 2024), processing circumstances (Wang et al., 2022), planting environments (Wang, Liu, et al., 2024) and tea cultivars (Li, Ran, et al., 2022). However, the specific mechanisms by which tea cultivars influence the quality of JSGT remain to be elucidated.

Variations among cultivars also lead to differences in metabolites, necessitating further analysis through non-targeted metabolomics. As an illustration, Li et al. (2024) investigated the aromatic composition derived from five distinct Lu’an Guapian tea (LAGP) cultivars, namely ‘Longjing43’ (LJ43), ‘Dushanzhong’ (DSZ), ‘Qishanzhong’ (QSZ), ‘Wuniuzao’ (WNZ), ‘Shuchazao’ (SCZ) by headspace-solid phase microextraction-gas chromatograph-mass spectrometry (HS-SPME-GC–MS), solvent-assisted flavor evaporation- gas chromatograph-mass spectrometry (SAFE-GC–MS) and gas chromatography-olfactometry (GC-O). It was discovered that the levels of (Z)-3-hexen-1-ol levels were substantially elevated in the fresh and floral aroma of QSZ compared to other cultivars, while (E)-β-ocimene, contributing the tender chestnut-like aroma, exhibited significantly higher concentrations in LJ43, SCZ, and WNZ. The aromatic profiles of Wuyi rock tea made from 16 distinct tea cultivars, were investigated by Yue et al. (2023) application an electronic nose (E-nose) and HS-SPME-GC–MS. This research identified key volatile compounds, including β-ionone, geraniol, phenethyl alcohol, benzyl acetate and indole, most of which exhibit fruity, floral and woody aromas. However, to our knowledge, very few studies have been conducted to analyze the aroma profiles of JSGT produced from different tea cultivars. The findings of above studies provide an experimental basis for investigating the distinctions in volatile substances of JSGT produced using various tea cultivars.

Exhibiting simplicity, high efficiency and environmental friendliness, HS-SPME is frequently utilized for the characterization of volatile components in tea and other beverages (Liang et al., 2024; Luo et al., 2024). It is often as an extraction approach to the analyze of highly volatile metabolites in tea (Feng et al., 2024). Inversely, SAFE is highly efficient in the extraction process (Lau et al., 2018). Through the integrated use of HS-SPME and SAFE, a more comprehensive extraction of volatile substances in tea can be achieved, and therefore the odor profile of JSGT from different cultivars can be more accurately detected. GC–MS is one of the primary analytical techniques employed in metabolomics for the identification and quantification of volatile metabolites (Baky et al., 2025; Wen, Zhu, et al., 2023). GC-O-MS, which combines the human olfaction with instrumental analyses, has been recognized as a critical approach for the identification of key aroma-active substances among a large number of volatiles in tea (Wang, Xie, et al., 2024). Hence, this research, the representative volatile compounds in JSGT were investigated by employing both SPME-GC–MS and SAFE-GC-O-MS.

Moreover, volatile substances, non-volatile substances including alkaloids, soluble sugars and amino acids also have a substantial impact on tea infusion's flavor and color (Zhang, Sun, et al., 2023). High Performance Liquid Chromatography (HPLC) is marked by its extensive use, high separation efficiency and user-friendly equipment, making it a preferred approach for the detection of catechins and amino acids (Ai et al., 2024). Previous studies utilized peptidomes and targeted metabolomics to compare the taste contributors in Longjing tea made from ‘Longjing43’ and ‘Qunti’ cultivars. The research revealed that key primary metabolites, such as alkaloids, saccharides and phenolic acids, differ between the two cultivars, influencing their flavor profiles (Yan et al., 2025). Liu et al. (2024) analyzed the secondary metabolites of Congou black tea derived from various tea cultivars by ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). The results revealed how the variation of secondary metabolites in Congou black tea derived from various tea tree cultivars were resulting from phenolic acid secondary metabolites and flavonoids.

Furthermore, some studies have revealed that the E-nose has been used as a simple approach to the detection and identification of odor-active substances in tea (Ge et al., 2024). Additionally, the response intensity of the electronic tongue (E-tongue) is affected by the level of non-volatile constituents in the tea infusion. The taste strength of the signals detected by the taste sensors can be quantified using E-tongue, thereby eliminating the influence of human subjectivity (Feng et al., 2024). In a recent study, Qi et al. (2024) employed HS-SPME-GC–MS combined with E-nose and E-tongue to analyze four distinct grades of Jiuqu Hongmei tea, revealing significant differences in volatile compounds among the different grades. These researches proved the feasibility of revealing differences in volatile and non-volatile compounds of JSGT produced using various tea cultivars.

Moreover, the compounds associated with tea flavor, such as tea polyphenols (TPs) and alkaloids, are increasingly recognized as natural antioxidants, α-glucosidase inhibitors and additional health-promoting functions (Li, Zhang, et al., 2022; Ozkan et al., 2024; Qi et al., 2024). Numerous studies have demonstrated that TPs possess diverse biological functions, including antioxidant, anti-obesity, and anti-allergic activities. These compounds exhibit potential for the prevention and treatment of conditions such as diabetes, hypertension, scurvy, and gastrointestinal disorders. (Yan et al., 2020; Chen et al., 2023). Experiments have demonstrated that incorporating active materials derived from tea into food packaging can effectively extend the shelf life of food products. The abundance of antioxidants in tea also highlights its potential applications in natural food preservation and functional packaging. Furthermore, the incorporation of tea or tea extracts as functional ingredients in food preparation is becoming increasingly common (Pourzafar et al., 2023). Thus, it is significant to figure out the bioactive substances present between different tea cultivars.

To date, there is limited academic research, both domestic and international, on the flavor characteristics and health-related efficacy of JSGT produced from different tea cultivars, indicating a notable research gap and substantial potential for further exploration. Therefore, in our research, we selected five tea cultivars named ‘Jingshan1’ (JS1), ‘Jingshan2’ (JS2), ‘Yingshuang’ (YS), ‘Cuifeng’ (CF) and ‘Jiukengzhong’ (JK), planted in the same area. The present study aims to constructed the flavor substance profiles of JSGT from different cultivars using HS-SPME-GC–MS and SAFE-GC-O-MS, elucidate their chemical basis through HPLC and LC-MS analysis, identified chemical markers responsible for these differences, and correlated flavor attributes with chemical composition by E-nose and E-tongue techniques. On this basis, the in vitro antioxidant activity and α-glucosidase inhibitory activity (α-GIA) of the samples were evaluated. This thereby establishes a scientific foundation for choosing different tea cultivars in the production of JSGT and provides a basis for subsequent studies concerning JSGT.

2. Materials and methods

2.1. Tea samples, chemicals and materials

The five tea cultivars of JSGT samples used in this research (JS1, JS2, YS, CF, and JK) were provided by Hangzhou Jingshan Wufeng Tea Co., Ltd. The fresh tea leaves were harvested in March 2024 in the same area of Jingshan Town, Yuhang District, Hangzhou City, Zhejiang Province, according to the criterion of one bud and two leaves. The process and parameters were the same as the research of Hou et al., 2024. The tea samples were stored away from light and placed in a refrigerator at −80 °C awaiting further analysis. The chemicals and materials employed for analysis are detailed in Table S1.

2.2. Color quantitative analysis

2.2.1. Tea infusions preparation

The method for preparing the samples follow the Chinese national criterion GB/T 23776–2018. To begin with, 3 g of each tea samples was precisely measured into the evaluation cup, boiling water was introduced to adjust the final volume of 150 mL, after 4 mins, the infusions and brewed leaves were gathered for further experiments.

2.2.2. Capture images of dry tea samples, tea infusion and brewed tea leaves

The parameters of capturing images were followed the Wei et al., 2023 with little adjustments. In details, digital camera (NIKON Z5, Japan) was set as follows: focal speed, 50 mm, ISO sensitivity. Av aperture, F/9.0, shutter speed 1/200 s and the white balance, 5000-K color temperature.

2.2.3. Color measurements for tea infusion

The colors of dry leaves, infusion and brewed leaves were analyzed with computer vision by ImageJ software (version 1.54, National Institutes of Health, Bethesda, MD, USA) (Sheng et al., 2024). The image analysis was conducted following the methodology described in the aforementioned literature. The L, a*, and b* parameters were determined.

2.3. Extraction and identification of volatile components

2.3.1. HS-SPME-GC–MS detection of volatile substances

For HS-SPME-GC–MS, the method is based on Zhang, Huang, et al. (2024) with appropriate adjustments. Briefly, 0.6 g tea sample was placed into a headspace vial. Subsequently, 1 g of sodium chloride, 5 mL of boiling water, and 10 μL of ethyl decanoate were added and immediately sealed. The vial was subsequently pre-equilibrated 15 min at the temperature of 50 °C. After that, an 80 μm DVB/C-WR/PDMS fibre (Agilent, Santa Clara, CA, USA) was introduced into the vial, which maintained during 40 min within identical settings. HS-SPME was performed by exposing the fibre to the gaseous phase above the tea sample. No direct contact between the fibre and the aqueous solution occurred. Volatile compounds partition from the liquid phase into the vapour phase, and the fibre adsorbed them from the gas phase. Therefore, no liquid water was carried over onto the fibre or into the GC–MS system.

The volatiles extracted by SPME underwent analysis via a Model 8890-5977B GC–MS system (Agilent, Santa Clara, CA, USA) equipped with HP-5MS column (30 m × 0.25 mm × 0.25 μm, J & W, Folsom, CA, USA). The fibre was then introduced into the GC–MS system injector for thermal desorption for 5 min. The injection mode was non-split and injection port temperature was 250 °C. The oven temperature was programmed following the protocol outlined in Zhang, Yang, et al. (2024). Based on the retention times (RT) obtained from the injection detection of n-alkanes (C6 - C40), the retention indices (RI) of all detected volatile components in GC–MS analysis were calculated. The compounds were identified by searching the NIST 2017 standard spectral library, and their identities were further confirmed by comparing the experimental RI values with those reported in public databases. Quantitative analysis was performed using the internal standard method, with relative contents calculated based on the added ethyl decanoate.

2.3.2. SAFE-GC-O-MS detection of volatile substances

For SAFE-GC-O-MS, the method is based on Feng et al. (2024). The 3 g samples brewed 5 min with 150 mL of deionized boiling water. The tea infusion was filtered and cooled after 4 min. Thereafter, 10 μL of ethyl decanoate was employed as the internal standard. The sample was then distilled in a vacuum environment at 10⁻² Pa by a SAFE unit. Both the circulating water bath within the SAFE unit and the temperature of the thermostatic water bath in the sample distillation compartment was precisely controlled at 40 °C. The resulting distillate (approximately 140 mL) underwent three extractions, each with 30 mL of dichloromethane to obtain the final extract. This extract was subsequently dried over anhydrous sodium sulfate and concentrated to 100 μL via nitrogen purging in a 25 °C water bath.

The volatile compounds extracted by SAFE were subjected to GC–MS analysis under identical instrumental conditions as those used for the HS-SPME extracts. All volatile substances were analyzed on the same GC–MS system, a Model 8890-5977B GC–MS system (Agilent, Santa Clara, CA, USA) equipped with an HP-5MS column (30 m × 0.25 mm × 0.25 μm, J & W, Folsom, CA, USA). The temperature profile of the oven followed the protocol reported in Feng et al. (2024). The flow rate of the helium carrier gas (99.99%) was set ae 1.0 mL/min. The mass-selective detector operated in positive ion mode, scanning a mass range of m/z 30 to 350 at 70 eV. The relative content of volatile compounds was calculated using the same method as described in section 2.3.1. GC-O was conducted following the methods outlined by Feng et al. (2024), Zhang, Xia, et al. (2023) with the well training assessors (7 men and 5 women). A sniffer port (ODP3, Gerstel, Germany) was installed on the gas chromatograph.

2.3.3. Calculation of relative odor activity values (ROAV)

In this study, ROAV was employed to assess the impact of individual compounds to the overall odor. The formula is ROAV_i = OAV_i/OAV_max × 100, where OAV_max is the maximum odor activity value of the volatile substance and OAV_i is the OAV of the particular volatile substance. The formula is OAV_i = C_i/OT_i, where C_i is the concentration of volatile substances and OT_i is the odor threshold in water(Zhu et al., 2020).

2.4. Extraction and identification of non-volatile substances

2.4.1. Determination of TPs, free amino acids (FAA) and water extractable (WE)

TPs and FAA were determined using a colorimetric method in accordance with the procedures outlined in GB/T 8313–2018 and GB/T 8314–2013. Similarly, WE was determined following to the standard specified in GB/T 8305–2013.

2.4.2. Quantification of main compounds by HPLC

The catechins and caffeine (CAF) were determined by Waters 2695 HPLC system (Waters, MA, USA) equipped with a Waters Symmetry® C₁₈ HPLC column (4.6 × 250 mm, 5 μm). The treatment of tea samples is the same as 2.3.1. Catechins and CAF were assayed following the methods outlined in ISO standards 14,502–2(ISO 2005). In details, 0.2 g of tea sample in powdered form was combined with 5 mL of 70% methanol in a 10 mL centrifuge tube and was placed in a constant-temperature bath at 70 °C for 10 min. After 10 min, the tube was removed, cooled to room temperature and centrifuged on a 3500 rpm centrifuge for 10 min, then the supernatant was decanted into a 10 mL volumetric flask. The procedure was then conducted once by adding 5 mL of 70% methanol again to this centrifuge tube. The two supernatants were merged and diluted to 10 mL using 70% methanol. Another 2 mL liquid was pipetted into a 10 mL volumetric flask, and the same was brought to volume using 70% methanol solution. The solution was filtered using a 0.22 μL filter membrane and then determined by HPLC. 10 μL of the filtrate was introduced into the column at a fluid delivery rate of 1 mL/min and a column temperature of 35 ± 1 °C. The HPLC elution gradient followed the method described in ISO 14502-2: The HPLC elution gradient followed the method described in ISO 14502-2: 0–10 min, 100% A; 10–25 min, 100–68% A, 0–32% B; 25–35 min 68% A, 32% B, the system was then reset to 100% mobile phase A and equilibrated for 10 min. The flow rate was 1.0 mL/min; 10 μL sample was injected and detected at 278 nm.

L-theanine was determined using High Performance Liquid Chromatography method according to Chinese national standard GB/T 23193–2017. The procedure for the detection of L-theanine was detailed below: 0.2 g of tea powder was added to 50 mL conical flask. Subsequently, 50 mL of deionized water heated to 100 °C was poured into, and the flask was subjected to water-bath heating at 100 °C for 30 min. Afterward, the solution was transferred to a 50 mL volumetric flask. Once brought to ambient temperature, the solution was diluted with pure water, subsequently, take 2 mL of the sample solution and subject it to filtration using 0.45 μm aqueous phase filter membrane and introduced into an HPLC equipped with Symmetry Shield RP18 (5 μm, 4.6 × 250 mm, Waters, MA, USA) HPLC Column for analytical purposes. Inject 10 μL into the HPLC system for analysis each time. The fluid transfer was conducted at 1 mL/min flow rate. Detection occurred at 210 nm. The HPLC analysis was performed followed a previously established method (Qiu et al., 2024). The compounds content was calculated based on the standard curve (Table S2).

2.4.3. LC-MS detection of non-volatile substances

The tea samples were lyophilized and milled into a fine powder. 0.4 g of tea powder was placed into a 10 mL centrifuge tube. 8 mL of 70% methanol (methanol/water, 7/3, v/v) was incorporated, along with the 4-chloro-_DL-phenylalanine as internal reference substance. The sample was ultrasonicated for 60 min in room temperature, left to rest for 1 h, thoroughly mixed and conducted again. 1.5 mL of the supernatant was extracted and centrifuged at 12000 rpm for 10 min. Afterward, 0.1 mL of the supernatant was taken and brought to volume using 3.9 mL 70% methanol and subjected to filtration using a 0.22 μm organic membrane prior to LC-MS detection. Equal volumetric proportions of each tea sample were blended to prepare Quality control (QC) samples. Every single sample was analyzed in triplicate.

The tea samples were detected and assessed through by an ultra-high-performance liquid chromatograph (UltiMate 3000, Dionex, Sunnyvale, CA, USA) coupled with a mass spectrometer (Q-Exactive Focus, Thermo Fisher Scientific, Waltham, MA, USA). Sample isolations conducted with an Acquity HSS T3 column (100 mm × 2.1 mm, 1.8 μm; Waters, MA, USA). The conditions of chromatography and mass spectrometry were aligned with those established in a prior study (Zhang, Zhang, et al., 2024, Xu et al., 2019). The mobile phase included distilled water containing 0.05% formic acid (phase A) and acetonitrile containing 0.05% formic acid (phase B).

2.5. Intelligent sensory judgment

2.5.1. E-nose analysis

The E-nose analysis was followed the study of Wen, Jiang, et al. (2023) with adjustments. 2 g tea powder was transferred into a headspace vial. Subsequently, a sampling needle (HS100, Alpha MOS Company, France) was employed to extract a specific volume of gas, then promptly delivered into the detector of the E-nose system (FOX 4000 model, Alpha M.O·S, France) under 3 s. Volatile compounds separated by each column were monitored using 2 thermal ionization by flame detectors and recorded using the software. The data recording time was configured as 480 s, using hydrogen (99.99%) as the carrier gas, which was produced under a pressure of 3 bar and fluid delivery rate of 1 mL/min. Each sample was detected 3 times.

2.5.2. E-tongue analysis

In this study, the tea broth was measured using an E-tongue, and data acquisition was conducted on a taste sensing system (SA402B, Insent Intelligent Sensor Technology, Inc., Kanagawa, Japan). The experiments adhered to the methodology described by Wu et al. (2024). A taste sensor and a caramic refence electrode are components of the taste sensing system, The taste sensor detects changes in potential by means of a reference electrode containing a contamination-resistant lipid membrane and a solid-state Ag/AgCl electrode. The taste sensors and ceramic electrodes were tested separately in two solutions for 24 h. The first solution contained two components: 30 mM KCl and 0.3 mM TA, and the second solution was a 3.33 M KCl solution. In this experiment, five basic taste indicators were used: bitterness, umami, saltiness, richness, and aftertaste astringency (aftertaste-A). The tea samples were processed following the specification of GB/T23776–2018. For each sample, 35 mL of tea broth was taken and measured four times. The initial measurement was excluded, while the remaining three measurements were kept for analysis.

2.6. Determination of antioxidant capacity and α-GIA

1,1-Diphenyl-2-picrylhydrazyl radical scavenging activity (DPPH) can be used to assess the in vitro antioxidant activity of tea leaves (Mi et al., 2024). The assay was modified from that of Zhang (2024d). In summarizing, 7.8 mL of DPPH solution (0.06 mg/mL) was mixed with 0.2 mL of the sample solution. The mixed liquid was then allowed to incubated under dark conditions for 30 min. Following this step, the absorbance at 517 nm was then determined spectrophotometrically. Methanol mixed with DPPH solution in equal proportions was employed as a blank sample. The formula was calculated as (1).

$$\mathit{DPPH\; inhibition}\%=\frac{A_0-A}{A_0}\times 100$$ (1)

α-GIA was measured following an adapted Li et al. (2022a) method. Briefly, sample preparation, α-glucosidase and substrate (glucoside) were prepared in 0.1 M PBS (pH 6.8). 40 μL of sample extract was incubated with 60 μL of 1 U/mL α-glucosidase working solution was held under incubation conditions of 37 °C for 15 min. After the incubation, 60 μL of substrate was added and maintained at 37 °C for 20 min. Subsequently, the absorbance was recorded at 405 nm through application a microplate reader (Multiskan Ascent, Thermo Fisher Scientific, Waltham, MA, USA) and calculated to obtain α-GIA. α-GIA (%) was calculated as (2).

$$\alpha -\mathit{GIA}\left(\%\right)=\frac{\left(A_C-A_b\right)-\left(A_s-A_{\mathit{sb}}\right)}{A_c-A_b}\times 100\%$$ (2)

In which A_b, A_c, A_s and A_sb denote the absorbance of PBS, the absorbance upon the post-reaction absorbance of PBS with α-glucosidase, the absorbance upon the post-reaction absorbance of tea samples with α-glucosidase and the absorbance upon the post-reaction absorbance of tea samples with PBS, in that order. Each test was repeated three times.

2.7. Sensory evaluation

With reference to GB/T 23776–2018, a review team consisting of 10 experienced tea specialists was formed to make a comprehensive assessment of each tea sample based on five factors, with top score of 100 points.

2.8. Statistical analysis

The obtained LC-MS data were analyzed with MS-DIAL software (version 5.1) after conversion as ABf format by Analysis Base File Converter. Analysis of variance (ANOVA) was conducted with SPSS 27.0 (IBM, Armonk, NY, USA), principal component analysis (PCA) and orthogonal least squares Discriminant (OPLS-DA) analyses were conducted with SIMCA 14.1 (Umeå, Sweden), and plots were drawn using Origin 2018 (OriginLab Co., USA) and TBtools.

3. Results and discussion

3.1. Sensory evaluation of the five distinct cultivars of JSGT

To investigate how different tea cultivars impact the sensory characteristics of JSGT, five tea samples were evaluated and the tea infusion were digitally analyzed (Fig. 1A). As shown in Table S3, The JS1 sample scored the highest, with its aroma characterized as tender, pure and long-lasting. JS2 were the second highest, and their aroma characteristics were tender and pure, but the persistence of the aroma was slightly worse than that of JS1. The aroma of YS and JK was dominated by fresh aroma, yet their tenderness and persistence were comparatively lower than those of JS1 and JS2. CF showed a clear floral aroma, but the tea infusion flavor was slightly poorer compared to JS1 and JS2, resulting in a marginally lower overall score. The analysis of infusion showed that important variations existed in the color of tea infusion of the five samples. Fig. 1 B illustrates that the L* value of CF was the highest (78.59 ± 0.14), which means the brightness of the CF's infusion was the highest. The absolute value of the a* value of JS2 was the largest (−8.11 ± 0.04), which means the infusion of JS2 has more greenish color than other cultivars; and JK has the highest b* value (24.12 ± 0.14), which means that the infusion is more yellowish. The aforementioned sensory and colorimetric results demonstrate significant distinctions in visual appearance, aroma, and taste attributes among different cultivars of JSGT. JS1 and JS2 exhibited superior aromatic qualities, whereas CF displayed pronounced characteristics in liquor brightness and floral notes. The differences in tea infusion color are primarily associated with water-soluble pigments present in the tea leaves, while the distinct flavor profiles are significantly related to the distribution of volatile aroma compounds and taste-active components across different cultivars (Wu et al., 2024; Zhang, Zhang, et al., 2023; Su et al., 2024). This study provides a basis for the sensory quality-oriented selection of tea cultivars.

Fig. 1.

A. Images of dry tea, infusion and brewed leaves. B. 3D scatter plot of L* value, a* value and b* value of the tea infusion of Jingshan green tea from different tea cultivars. (JS1: Jingshan1, JS2: Jingshan2, YS: Yingshuang, CF: Cuifeng, JK: Jiukengzhong). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3.2. Analysis of JSGT of different tea cultivars by intelligent sensory analysis

3.2.1. The results of E-nose analysis

The volatile substances extracted from the samples were analyzed by an E-nose. The PCA plot revealed that tea specimens derived from distinct tea cultivars were separated from each other and the same tea samples were close to each other and partially overlapped (Fig. S1 A). The PCA clearly showed the differences in volatile odor characteristics among the tea samples, indicating that the flavor profiles of different JGST cultivars are distinguishable. This is probably due to the differences in volatile compounds among different tea cultivars. Subsequent studies further incorporate GC–MS to conduct qualitative and quantitative analyses of specific differential volatile compounds, thereby identifying the key aroma-active substances responsible for odor variations between cultivars.

3.2.2. The results of E-tongue analysis

The E-tongue characteristics of the five tea samples are shown in the radargram (Fig. S1 B). JS1 embodies the lowest aftertaste-A. JS2 reflects lower bitterness and has a slightly weaker umami. YS showed the strongest umami than the other cultivars and presented very low bitterness. CF's aftertaste-A was the highest. JK showed higher richness. The results indicated that the E-tongue could effectively differentiate characteristic variations in the basic taste dimensions of JSGT from different cultivars, reflecting systematic distinctions in their taste-related chemical compositions. These differences may be closely associated with variations in the content and proportions of flavor-active substances in the tea leaves, such as catechins, amino acids, and alkaloids. This study found that cultivars of JSGT with higher contents of L-theanine and free amino acids exhibited stronger umami intensity., which is consistent with the findings of Feng et al. (2024).

3.3. Different volatile aroma compounds of JSGT from various tea cultivars

3.3.1. Investigation of the aroma compounds in JSGT of various tea cultivars

In the study, both SAFE and SPME were used to extract the tea volatiles to better illustrate the variation in the aroma profile among different JSGT samples (Table S4). The extracted volatiles were further identified through GC–MS, with the results presented in Fig. 2. A total of 179 volatile substances was characterized in the five samples. Includes 9 acids, 28 alcohols, 21 aldehydes, 22 alkenes, 22 aromatics, 29 esters, 18 heterocyclics, 17 ketones, and 13 others. Among them, 99 compounds were common to all five samples. JS1 and JS2 exhibited the largest percentage of esters, whereas the other three samples had more alcohols than esters. Moreover, the percentage of heterocyclic content of JS1and CF was much higher than that of other cultivars. JSGT from different cultivars exhibited distinct diversity in volatile compound composition, which may be a primary factor contributing to variations in their aroma characteristics. The relatively higher content of heterocyclic compounds in JS1 and CF could be associated with Maillard reactions or thermal transformation processes occurring during processing (Ma et al., 2024).

Fig. 2.

A. UpSet diagram of the volatile profiles detected by GC–MS. B. Percentage of volatiles of different categories of Jingshan green tea from different tea cultivars. C. Effects of different tea cultivars on the contents of volatile compounds of Jingshan green tea, (mean ± Standard deviation). The different letters on the bar chart indicate that the content of compounds was significantly different (p < 0.05). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3.3.2. Screening and analysis of the different flavor substances in JSGT of various tea cultivars

In order to investigate the variations within the five samples, PCA analysis was performed according to the volatile substances obtained from GC–MS. Fig. 3 indicates that tea samples of the identical tea cultivar are grouped closely, in contrast, different tea cultivars show significant separation. The HCA results showed that CF did not cluster with the rest of the tea cultivars, a finding that is in agreement with the PCA plots. This could mean that there is a higher number of unique compounds present in CF, or the concentration of the common chemical constituents is markedly distinct compared with the other samples. The PCA and HCA plots both illustrated indicating the presence of some distinctions in the flavor compounds of JSGT produced using distinct five various tea cultivars.

Fig. 3.

A. PCA score plot of Jingshan green tea from different tea cultivar. B. HCA plot of Jingshan green tea from different tea cultivars. C. OPLS-DA plot of Jingshan green tea from different tea cultivar. D. The validation process for OPLS-DA models of the volatile profiles. E. Heat map analysis of volatiles (VIP > 1) identified by GC–MS. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

In order to enhance the separation of the five samples and demonstrate how the volatiles contribute to their differences, the OPLS-DA model was employed. The model parameters and the results of 200 permutation tests demonstrated a strong ability to explain variance (R²Y), predictive ability (Q²) and there was no overfitting. Basing on the OPLS-DA model, 41 volatile compounds were identified by VIP > 1, p < 0.05. Heat map visualized the change in the content of these substances among different samples. (Fig. 3) Indole was higher in JS1. (Z)-3-hexen-1-ol was mor abundant in JS2. Benzyl alcohol, linalool and (E)-Linalool oxide (furanoid) were higher in YS. Diacetone alcohol and (Z)-3-hexenyl hexanoate were higher in CF. Methyl hexanoate and geraniol were higher in JK. These substances primarily include alcohols, heterocyclic compounds and esters, which have been reported in previous studies to respectively produce floral, fruity, and sweet aromas in green tea (Zhang, Feng, et al., 2024). The sensory evaluation results are consistent with these reported findings, which may explain the the fresh and floral aromas characteristic of JSGT. Variations in their concentrations among different tea cultivars could lead to differentiated aromatic profiles, with some JSGT samples exhibiting more pronounced fresh notes while others display enhanced floral characteristics.

3.3.3. Screening of key volatile substances in JSGT of distinct tea cultivars

Although numerous volatile compounds were present at high concentrations across different cultivars of JSGT, not all high concentration of volatile compounds has a significant influence over the volatile aroma properties of tea (Feng et al., 2024). Therefore, the volatiles were further analyzed by GC-O. According to Table S6, altogether 41 substances were identified and further filtered based on GC-O ≥ 3. Among these, 20 volatiles were common to all five samples. A greater number of compounds exhibit a floral aroma, including linalool, (Z)-3-hexen-1-ol, geraniol, (E)-nerolidol, and others. Conversely, a few compounds emit unpleasant odors, such as (Z)-3-hexenyl butyrate, naphthalene.

The holistic flavor profile of tea is affected by both the content and quantity of the volatile substances and their odor threshold (Huang et al., 2024). Generally, volatile substances with ROAV >1 can be regarded important to the holistic flavor characteristics (Zhang, Sun, et al., 2023). After the screening, 25 volatile substances with ROAV >1 were ascertained: 24 in JK, 19 in CF, 17 in JS1, 16 in JS2, and 11 in YS. Based on the ROAV >1, VIP > 1, and GC-O ≥ 3, 9 key aroma active substances were identified from five samples: (Z)-3-hexen-1-ol, 1-octen-3-ol, (E)-linalool oxide (furanoid), linalool, nonanal, maltol, (Z)-3-hexenyl butyrate, geraniol, and decanoic acid (Fig. 4, Tables S5-S6). Studies have demonstrated that linalool, nonanal, and geraniol are key aroma-active compounds in green tea, which is consistent with our experimental results (Ye et al., 2025). Additionally, the presence of (Z)-3-Hexen-1-ol (floral) and (Z)-3-hexenyl butyrate (woody, green) may contribute to the distinctive fresh and floral aroma of JSGT. Zhang, Zhang, et al. (2024), Zhang, Xia, et al. (2023) and Yan et al. (2024) found that (E)-linalool oxide (furanoid), linalool, nonanal, maltol, (Z)-3-hexenyl butyrate, and geraniol account for the floral, fruity, and sweet flavor in green tea, which is consistent with the results of our study.

Fig. 4.

The contents of key aroma compounds in Jingshan green tea from different tea cultivars (mean ± Standard deviation). The different letters on the bar chart indicate that the content of compounds was significantly different (p < 0.05). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Specifically, linalool exhibited the highest ROAV value in JK. The terpene alcohol is derived from the geranyl pyrophosphate precursors by linalool synthase and is widely recognized as an important source of the floral flavor of green tea (Gao et al., 2024; Mi et al., 2024; Ouyang et al., 2024). (E)-Linalool oxide (furanoid) is formed from linalool through further oxidation (Gao et al., 2024). It plays a key role contributing to the sweet flavor profile of tea. And the content of linalool and its oxides was higher in YS. Nonanal (floral, citrus) can be generated through Maillard reaction and lipid degradation reactions (Sun et al., 2024; Zhu et al., 2024). Its content was similar in CF and YS and higher than other cultivars. Maltol (fruity) and (Z)-3-hexenyl butyrate (floral, fruity, sweet) were higher in JS1. Geraniol (floral) was found to be the most abundant in JK. Among the other volatiles that contribute to the flavor characteristics of JSGT, (Z)-3-hexen-1-ol stands out. This compound can be formed through glycosidic hydrolysis or ester degradation in the course of the withering of fresh tea leaves (Ma et al., 2024). It is responsible for the characteristic greenness of tea at high concentration.

Furthermore, (Z)-3-hexen-1-ol, a crucial aromata substance distinguishing white tea (Zhang, Xia, et al., 2023) and a common flavor substance in green tea (Li, Ran, et al., 2022), exhibited the highest content in JS2 in this study. n-Decanoic acid, which exhibits a grass-like aroma, showed the highest content in YS. 1-Octen-3-ol, considered a major contributor to the mushroom-like flavor in green tea and derived from thermal degradation of unsaturated fatty acids (Wen, Jiang, et al., 2023; Zhai et al., 2023; Zhang, Huang, et al., 2024; Wang, Xie, et al., 2024), was present at higher content in JK than in the other samples.

In summary, differences in the content and composition of these key aroma-active substances collectively contribute to the distinct aromatic profiles of the five JSGT cultivars, thereby revealing the chemical basis for flavor variation among cultivars at the volatile compound level.

3.4. Influence of different tea cultivars on non-volatile substances

3.4.1. Analysis of the catechins, CAF, L-theanine and WE

TPs, catechins, CAF, FAA, _L-theanine, and WE are important components and indicator of tea flavor (Feng et al., 2024; Wang et al., 2022; Qiu et al., 2024; Li, Ran, et al., 2022). For the purpose of explore the chemical basis of flavor differences among distinct cultivars of JSGT, the contents of the above substances were determined (Table S7). Catechins are subdivided to include ester catechins (e.g., (−)-epigallocatechin gallate (EGCG) and (−)-epicatechin gallate (ECG)) and non-ester catechins (e.g., (−)-epicatechin (EC), (−)-epigallocatechin (EGC), and catechin (C)) (Qi et al., 2024). Among them, ester catechins display stronger bitterness and astringency. CAF also contributes significantly to the presentation of bitter and astringent flavors in green tea.

The results showed that the CAF contents of YS, CF, and JK were relatively high and showed no significant differences, whereas the EGCG concentration in JK was significantly higher than all samples. This demonstrates a cultivar-dependent pattern of polyphenol biosynthesis. Moreover, the interplay of CAF with EGCG enhances the bitterness and astringency of EGCG (Qi et al., 2024), causing JK to exhibit a more bitterness flavor. JS1 had the lowest CAF content and the highest theanine content. On the contrary, non-galloylated catechins (EGC and EC) are likely responsible for sweet aftertaste observed in green tea. The contents of EGC exhibits higher in YS and CF. The intensity of the aftertaste is correlated with the levels of these catechins.

FAA are derived from the hydrolysis of proteins in the course of tea leaves withering and mainly give the tea a fresh and umami profile (Yang et al., 2018; Li, Ran, et al., 2022). _L-Theanine, a special amino acid found in tea, has a taste similar to glutamic acid and constitutes approximately 50% of the free amino acid content of tea (Feng et al., 2024). JS1 had the highest level of _L-theanine and exhibited a slightly stronger umami flavor, which aligns with the finding of the sensory assessment.

Studies often use the phenolic ammonia ratio value as an indicator to evaluate the mellow and refreshing taste of green tea, and the ratio value of tea plant cultivars suitable for processing into green tea is usually less than 8. In this experiment, the phenolic ammonia ratios of the five tea plant cultivars were less than 8, which indicated that all of them were suitable for making JSGT. Furthermore, JK exhibited the highest WE content, which further influenced the coarseness of the tea infusion (Li, Ran, et al., 2022).

The results above indicate that significant differences in the composition of key taste-active compounds exist among JSGT samples due to cultivar variation, which directly influences the diversity of their taste profiles. By examining the ratios and interactions among these key components, the manifestations of bitterness, astringency, umami, and sweetness in different JSGT cultivars can be systematically interpreted. This provides a theoretical basis for flavor-driven cultivar selection and the optimization of processing techniques.

3.4.2. Analysis of LC-MS

To further elucidate the chemical composition of distinct cultivars of JSGT we used LC-MS to explore the differential metabolites. In total, 1071 metabolites were identified across the five samples. These metabolites were further examined through multivariate statistical methods. The PCA and HCA results (Fig. 5A, B) revealed that all samples were distinct from each other. The central clustering of QC samples in the score plot demonstrated robust reproducibility throughout metabolite extraction and LC-MS analysis. Moreover, according to the VIP values above 1 derived from the OPLS-DA model (Fig. 5C). After 200 permutation tests, the intercepts of R² and Q² were 0.824 and − 0.577 respectively, indicating that the OPLS-DA model was not overfitting (Fig. 5D). In conjunction with the Human Metabolome Database (https://hmdb.ca/) and previous research (Zhang, Feng, et al., 2024), an aggregate of 42 metabolic compounds was identified, which were shown in Fig. 5E. Including fatty acids, amino acid and their derivatives, phenolic acids, catechins, phenolic glycosides and others (Table S8).

Fig. 5.

A. PCA score plot of Jingshan green tea from different tea cultivars detected by LC-MS. B. HCA plot of Jingshan green tea from different tea cultivars detected by LC-MS. C. OPLS-DA plot of Jingshan green tea from different tea cultivars detected by LC-MS. D. Permutation test of the OPLS-DA mode. E. Heat map analysis of marker metabolites (VIP > 1) identified by LC-MS. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Flavonoids and flavone glycosides are key components of taste-related substances in tea, primarily contributing to its bitterness and richness (Deng et al., 2022). Many metabolites, such as 3-galloyl-(−)-epiafzelechin, EC, EGC and epigallocatechin-3-O-ferulate are recognized as pivotal differential metabolites in green tea (Musial et al., 2020). Among these, the JK samples exhibit the highest levels of 3-galloyl-(−)-epiafzelechin, EC, EGC, taxifolin, and 3,3’-Di-O-galloylepigallocatechin, which may contribute to its strong richness. This is supported by the analysis of E-tongue results.

Amino acids for instance _L-arginine, aspartate, and _L-glutamic acid contribute to umami and brisk flavor profile of green tea (Chen et al., 2022). The JS1, JS2, and YS samples contain higher levels of these compounds, aligning with sensory evaluation results.2’-Deoxyinosine 5′-monophosphate and cis-3-hexenyl hexanoate impart floral and fruity flavor (Yin et al., 2022). JS1 has the highest content of 2’-Deoxyinosine 5′-monophosphate, while CF shows elevated levels of cis-3-Hexenyl hexanoate, which made CF exhibiting distinct floral characteristics.

Integrated analysis of the LC-MS results revealed systematic compositional differences in multiple metabolite classes among different cultivars of JSGT, which directly influence the taste attributes of the tea. By screening and functionally annotating key differential metabolites, a deeper molecular-level understanding of the formation mechanisms underlying flavor and functional variations among cultivars can be achieved, providing a scientific basis for cultivar identification and quality assessment of JSGT.

3.4.3. Correlation analysis involving E-tongue and metabolites

In order to more precisely verify the specific impact of the aforementioned substances on taste, PLSR analysis was employed (Fig. S2). In a PLSR plot, when two variables appear in the corresponding quadrant, it indicates a positive correlation in relation to them. Additionally, the distance of a metabolite relative to the origin of the axis is directly related to its contribution strength, the greater the distance, the more significant its contribution (Yu et al., 2019). The finding revealed that 17 metabolites substantially influenced the taste profiles, including flavonoids and flavone glycosides, catechins, amino acid and derivatives (Fig. S3).

Further Pearson's correlation coefficients (Table S9) revealed that aspartate (ID1767) and glutaric acid (ID1748) exhibit positive correlations with umami attributes, with aspartate played a crucial role in imparting umami. This finding is consistent with studies on umami formation mechanisms in (Lin et al., 2025)). In our research, YS exhibits the highest aspartate contents. The concentrations of _D- (−)-quinic acid (ID 5111), EC (ID10245) and EGC (ID11150), which were higher in CF and JK samples, were strongly associated linked to the richness flavor profile of JSGT. Quinic acid may play a role inthe decrease in bitterness and astringency in tea liquor (Yu et al., 2024). This agrees with our findings. In contrast, the compounds pyrophaeophorbide a (ID22315) and _L-Threonic acid hemicalcium salt (ID1943) exhibit strong positive correlations with bitterness, suggesting their role in increasing bitter taste perception in JS1 samples.

These findings reveal, at the metabolomic level, the material basis underlying taste variations among different cultivars of JSGT. The content and distribution of specific metabolites such as amino acids, catechins and their derivatives, and organic acids collectively influence key sensory attributes of the tea infusion—including umami, bitterness, astringency, and richness. This study provides a scientific basis for further elucidation of the molecular mechanisms governing green tea flavor formation via metabolomics, as well as for the application of key metabolite markers in tea quality evaluation and cultivar discrimination.

3.5. Antioxidant capacity and inhibition of α-glucosidase differences in different tea cultivars

3.5.1. Antioxidant activity

Tea is considered a healthy dietary provider of natural antioxidants because of its rich polyphenol content (Mi et al., 2024). DPPH was used to assess the antioxidant activities of JSGT from various tea cultivars (Fig. 6B). Combining the antioxidant ability and polyphenol content of five tea samples, it was found that YS has higher antioxidant activity, followed by CF and JK. The TPs contents of these three samples were also significantly higher than those of the other two cultivars.

Fig. 6.

A. Mantel test analysis of changes between non-volatile compounds and DPPH, α-glucosidase of Jingshan green tea from different tea cultivars. B. Bar chart of antioxidant capacity and α-Glucosidase inhibitory activity of Jingshan green tea from different tea cultivars. Different superscript letters indicate significant differences (P < 0.05) in the scores of the different tea samples. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

A significant association (r > 0.35, p < 0.01) was observed between tea polyphenol content and the antioxidant activity of JSGT was further confirmed by Mantel analysis (Fig. 6A). In addition, raffinose and 1,6-digalloyl-β-_d-glucopyranose also exhibited significant DPPH radical scavenging activity. Among them, raffinose, a phenolic glycoside, exerts its antioxidant capacity primarily through its phenolic hydroxyl groups, which serve as key reactive sites for free radical scavenging.

The above results elucidate, from a compositional perspective, the material basis for the differences in antioxidant activity among various cultivars of JSGT. The high polyphenol content and the presence of specific antioxidant components, such as raffinose, collectively contribute to the stronger free radical scavenging capacity observed in cultivars like YS. Total polyphenols are prone to degradation, leading to reduced bioactive efficacy; Therefore, novel antioxidant packaging can be employed during subsequent storage to mitigate this loss. This provides a scientific foundation for future antioxidant component-oriented tea cultivar selection and the development of functional JSGT products.

3.5.2. Inhibition of α-glucosidase

Presently, bioactive components such as the phenols or flavonoids from tea are considered a natural carbohydrate inhibitor of the digestive enzymes. Due to the greater number of hydroxyl groups in C and EC, their α-GIA is higher (Li, Zhang, et al., 2022). The α-GIA of the five tea samples is presented in Fig. 6B. JK exhibited the strongest α-GIA, while CF and YS demonstrated a slightly lower α-GIA compared to JK (p < 0.05). The α-GIA of JS1 and JS2 was weaker than other three tea cultivars and did not show a significant difference from one another.

In order to more comprehensively elucidate the association among non-volatile substances and the α-GIA of JSGT, the correlation among the concentrations of key non-volatile compounds and α-GIA in the tea samples was analyzed (Fig. 6). The results indicated a strong correlation (r > 0.35, p ≤ 0.05) among catechins and the α-GIA of the tea samples. Tea samples with higher total catechin content exhibited stronger α-GIA. Moreover, the inhibiting activity of these substances on α-glucosidase was significantly associated with the quantity of hydroxyl groups. Generally, a higher quantity of hydroxyl groups corresponds to greater inhibitory activity (Liu et al., 2025). Compounds such as EGCG, EC, raffinose, and 1,6-digalloyl-β-_d-glucopyranose possess a higher number of hydroxyl groups, allowing for stronger binding to α-glucosidase, which may account for their enhanced α-GIA.

In summary, in addition to TPs and catechins, compounds such as raffinose and 1,6-digalloyl-β-_d-glucopyranose also exhibit significant dual bioactivity, including both antioxidant and α-glucosidase inhibitory effects. and are therefore considered important bioactive constituents contributing to the potential health-promoting effects of JSGT.

4. Conclusion

In summary, different tea cultivars have an important influence on the sensory and chemical attributes of JSGT. Among them, JS1 and JS2 demonstrated superior aromatic quality, whereas CF displayed distinctive characteristics in infusion brightness and floral notes. Based on GC–MS, 9 key aroma-active substances were detected. The concentrations of linalool and its oxides, geraniol, (Z)-3-hexen-1-ol, along with five other volatile substances influenced the floral, fruity, and fresh aromas of the five samples, thereby contributing to the distinct aromatic profile of JSGT. Meanwhile, Significant differences were also observed in the non-volatile constituents among the five tea samples. The TPs and CAF in JS1 and JS2 were substantially reduced compared with YS, CF and JK, resulting in weaker astringency while enhancing the umami and fresh flavor characteristics. Additionally, a total of 17 substances were detected using LC-MS, primarily including flavonoids and flavone glycosides, catechins and FAA, among other categories. Multivariate statistical and correlation analyses revealed that these compounds were significantly associated with sensory attributes such as umami, bitterness, and astringency in the tea infusion. Furthermore, this present research identified that, in addition to TPs and catechins, raffinose and 1,6-digalloyl-β-_d-glucopyranose also exhibit significant antioxidant activity and α-GIA. These highly antioxidative components can serve as natural antioxidants for application in food preservation and the development of functional foods, provided their efficacy is experimentally validated and they meet market regulatory standards. Bioactive substances extracted from tea can also serve as green feed additives to enhance aquaculture practices. By elucidating the key flavor compounds and major functional components of JSGT and clarifying their metabolic pathways, it is possible to directionally cultivate tea plant varieties with specific flavor profiles and functional properties. This approach can not only enhance the economic value of raw materials but also create opportunities for establishing variety-specific brands. Furthermore, by integrating E-tongue, E-nose and GC–MS technologies to visualize the flavor characteristics of tea and establish correlations between flavor attributes and chemical components in JSGT, an industrial-level flavor detection system can be developed, thereby achieving flavor standardization for JSGT. In summary, these findings offer both theoretical frameworks and practical criteria for the future identification of high-quality raw materials for JSGT processing. Nevertheless, the dynamic changes of key aroma and taste compounds during the processing of different JSGT cultivars, as well as the relationship between volatile compounds and bioactivity, still require further investigation.

CRediT authorship contribution statement

Zhiwei Hou: Writing – original draft, Software, Methodology, Funding acquisition, Formal analysis, Data curation, Conceptualization. Qiyue Song: Writing – original draft, Resources, Methodology, Investigation, Data curation. Le Li: Writing – original draft, Methodology, Investigation, Data curation. Huiyuan Zhang: Software, Methodology, Formal analysis. Tiehan Li: Software, Methodology, Formal analysis. Sitong Liu: Methodology, Investigation, Data curation. Yuxuan Chen: Software, Methodology, Formal analysis. Ran Zhang: Methodology, Investigation, Data curation. Huiyan Jia: Supervision, Investigation. Mengqian Lu: Visualization, Supervision, Investigation. Xiangchun Zhang: Writing – review & editing, Visualization, Supervision, Investigation. Jingming Ning: Resources, Project administration, Funding acquisition.

Ethical statements

The sensory evaluation of tea sample in this study has been approved by the Ethics Committee of Zhejiang A&F University (Approval No. SMYX-2026-001). Participants gave informed consent via the statement “I am aware that my responses are confidential, and I agree to participate in this sensory evaluation” where an affirmative reply was required to enter the sensory evaluation. They were able to withdraw from the sensory evaluation at any time without giving a reason.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.