Skip to main content
NCBI home page
Food Chemistry: X logoLink to Food Chemistry: X
. 2026 Feb 21;34:103692. doi: 10.1016/j.fochx.2026.103692

Unraveling volatile and microbial dynamics of Pukeng tea with different storage times using metabolomics, chemometrics, and microbiome analysis

Fengke Lin a, Yuedi Xu a, Siyu Lin a, Ziqi Zhao a, Jingran Zhao a, Chunsong Cheng b, Binsheng Luo b,⁎
PMCID: PMC12950448  PMID: 41777631

Abstract

Pukeng tea (PKT), a traditional Chinese dark tea, has been consumed for centuries, yet its volatile and microbial dynamics remain unclear. This study integrated metabolomics, chemometrics, and microbiome analysis to explore PKTs with 3–20 years of storage. HS-SPME-GC–MS identified 189 volatiles, mainly alcohols, aldehydes, and ketones. PCA and PLS-DA revealed distinct metabolite patterns, with 46 differential volatiles, such as 1-butanol and 1-penten-3-ol, characterized as potential discriminants among PKT samples. Microbiota analysis showed 11 dominant bacterial genera, shifting from Firmicutes in early storage to Actinobacteriota in later stages, while Aspergillus dominated fungal communities. Correlation analysis revealed significant associations between dominant microbes such as Staphylococcus and Saccharopolyspora and aroma-active volatiles, suggesting microbial contributions to PKT's evolving flavor. This study provides the first integrated characterization of volatile and microbial diversity in PKT, offering insights into quality control, product authentication, and functional microbe discovery for the sustainable development of traditional dark teas.

Keywords: Fermented tea, Storage time, Microbial diversity, Volatile composition, Metabolomics, Chemometrics, Microbiome analysis

Highlights

  • •

    Metabolomics, chemometrics, and metagenomics were firstly applied to Pukeng tea.

  • •

    Volatile and microbial profiles varied in Pukeng tea with different storage times.

  • •

    Differential volatiles were identified as potential biomarkers of storage age.

  • •

    Key genera detected included Staphylococcus, Saccharopolyspora, and Aspergillus.

  • •

    Significant correlations between microbes and volatiles were revealed.

1. Introduction

Camellia sinensis-derived tea is among the most widely consumed non-alcoholic beverages worldwide, prized for both its sensory qualities and health-promoting effects (Hayat et al., 2015). According to differences in processing, teas are traditionally classified into six major categories: green, white, yellow, oolong, black, and dark teas (Wang, Han, et al., 2022). Among these, dark teas are characterized by post-fermentation and suitability for prolonged storage, during which complex biochemical reactions and microbial transformations occur, leading to the formation of diverse metabolites, including volatile compounds (Feng et al., 2023; Zhang et al., 2013). Understanding the microbial communities and metabolic profiles involved is essential for elucidating the development of unique flavors, bioactive properties, and overall quality evolution in dark tea during long-term storage (Zheng et al., 2015).

Pukeng tea (PKT), a type of dark tea, is primarily produced in Shitan Town, Qingxin District, Qingyuan City, Guangdong Province. As a local geographical indication product, it carries significant ecological, economic, and cultural value (Liao et al., 2020). Historical records indicate that PKT tea had already been consumed locally before the 1830s, reflecting its long tradition of use (Liao et al., 2020). Our ethnobotanical investigations further reveal that local communities regard long-term storage as essential for shaping the flavor and functional qualities of PKTs, which are commonly consumed at 3, 5, 8, 10, 15, or even 20 years of aging. Despite its long-standing use, research on PKT remains exceptionally scarce, with few studies addressing its chemical composition and microbial ecology.

Volatile compounds are key contributors to tea aroma and flavor, directly influencing consumer perception and product evaluation, and their profiles are strongly affected by storage duration (Wang et al., 2025; Zhai et al., 2022). Untargeted metabolomics using headspace solid-phase microextraction coupled with gas chromatography–mass spectrometry (HS-SPME-GC–MS) has emerged as a powerful tool for comprehensive volatile profiling (Liang et al., 2024). Combined with chemometric approaches such as partial least squares discriminant analysis (PLS-DA), this approach enables the identification of differential metabolites associated with specific variables, such as storage time (Liang et al., 2024; Wang et al., 2025). For instance, HS-SPME-GC–MS metabolomics has been used to identify 131 volatile compounds in green tea across different storage periods, with 16 compounds, including methyl salicylate and linalool, identified via PLS-DA as potential markers for predicting storage duration (Wang et al., 2025). Although preliminary studies on PKT have reported changes in approximately 40 volatiles, such as alcohols, aldehydes, and ketones, during processing and storage (Liao & Zong, 2021; Mu & Wang, 2025). a comprehensive untargeted volatile metabolomics study has yet to be conducted. Moreover, differential volatiles across varying storage periods remain unexplored.

In addition to volatiles, microbial activity plays a critical role in shaping tea quality during long-term storage (Huang et al., 2024). Advances in high-throughput sequencing now enable high-resolution characterization of microbial communities in fermented teas (Chen et al., 2017). For instance, amplicon sequencing-based microbial community analysis has been used to track microbial succession during the pile fermentation of Fuzhuan brick tea, revealing dynamic shifts in key genera such as Aspergillus and Bacillus (Le et al., 2023). Furthermore, as demonstrated in studies on oolong tea, the integration of microbiome profiling with HS-SPME-GC–MS-based metabolomics and chemometrics has proven effective for linking microbial taxa with differential metabolites, thereby facilitating the characterization of functional microorganisms (Huang et al., 2023; Yang et al., 2025). However, to our knowledge, such integrative approaches have not yet been applied in dark tea research, and the microbial community of PKTs has not been reported.

In this study, HS-SPME-GC–MS-based untargeted metabolomics, chemometric analysis, and microbiome profiling were employed to investigate the volatile and microbial dynamics of PKTs during long-term storage. Correlations between microbial taxa and volatile metabolites were further explored. This study provides insights into the functional microbes potentially responsible for key aroma-forming processes, aids in the identification of biomarkers for authenticating PKTs with precise aging times to ensure quality control, and supports the sustainable development of traditional tea products.

2. Materials and methods

2.1. Tea sample collections

PKTs, stored for 3, 5, 8, 10, 15, and 20 years, were obtained from Guangdong Huosheng Food Co., Ltd. (Shitan Town, Qingxin District, Qingyuan City, Guangdong Province, China). These samples were correspondingly labeled as PKT3, PKT5, PKT8, PKT10, PKT15, and PKT20. To minimize variability associated with primary growth conditions, all raw tea leaves were sourced from the same plantation in Shitan Town. The teas were produced in accordance with the Local Agricultural Standard of Qingxin District, Qingyuan City (DB441827). This standard is jointly issued by the Qingxin District Agriculture and Rural Affairs Bureau and the Qingxin District Market Supervision Administration. The traditional manufacturing process involves harvesting fresh leaf, steaming to inactivate endogenous enzymes, re-rolling, pile fermentation, drying, and extended storage. After collection, samples were transferred into sterile tubes, sealed, and stored at −80 °C until further analysis.

2.2. Volatile analysis by HS-SPME-GC–MS metabolomics

Tea samples were ground into fine powder under liquid nitrogen. For each homogenized sample with different storage times, five independent aliquots (each approximately 800 ± 0.2 mg) were weighed and transferred into separate 20 mL headspace (HS) vials for solid-phase microextraction (SPME). The extraction was carried out at 60 °C, with a preheating time of 15 min, an incubation time of 30 min, and a desorption time of 4 min. A 50/30 μm DVB/CAR/PDMS fiber (2 cm, Supelco, Bellefonte, PA, USA) was used during the HS-SPME procedure. Gas chromatography-mass spectrometry (GC–MS) analysis was performed on an Agilent 7890 gas chromatograph coupled to a 5977B mass spectrometer (Agilent, Santa Clara, CA, USA). A DB-Wax capillary column (30 m × 250 μm × 0.25 μm, Agilent) was employed for volatile separations. Samples were injected in splitless mode using helium as the carrier gas at a constant flow rate of 1.0 mL min−1 and a front inlet purge flow of 3.0 mL min−1. The oven temperature was initially held at 40 °C for 4 min, ramped to 245 °C at 5 °C min−1, and maintained for 5 min. The injector, transfer line, ion source, and quadrupole temperatures were set at 250, 250, 230, and 150 °C, respectively. Mass spectra were acquired in an electron impact (EI) mode at 70 eV, with a scan range of m/z 20–400. Data processing was performed using Chroma TOF 4.3× software, including baseline calibration, peak alignment, deconvolution, and integration. Retention indices (RIs) of detected volatiles were calculated relative to a series of n-alkanes (C7-C40, Sigma-Aldrich, St. Louis, MO, USA). Volatile compounds were tentatively identified by comparing mass spectra with the NIST library (similarity scores ranging from 0 to 1000), and compounds with similarity above 700 were retained (Li, Du, et al., 2025). The tentative annotations were further cross-validated by comparing the calculated RIs with reference RIs reported for a DB-Wax column in the NIST Chemistry WebBook (https://webbook.nist.gov/chemistry/). Odor descriptors of the identified volatiles were retrieved from databases such as Flavornet (https://www.flavornet.org/), FlavorDB2 (https://cosylab.iiitd.edu.in/flavordb2/), and FooDB (https://foodb.ca/). The relative abundance of each identified volatile was determined by expressing its peak area as a percentage of the total ion chromatogram.

2.3. DNA extraction and amplification

Total microbial DNA from each tea sample was extracted using the OMEGA Soil DNA Kit (Omega Bio-Tek, Norcross, GA, USA) based on the manufacturer's instructions. After extraction, DNA quality was evaluated by 0.8% agarose gel electrophoresis, and DNA concentration was determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Sunnyvale, CA, USA). For bacterial communities, the V5-V7 regions of the 16S rRNA gene were amplified using the forward primer 799F (5’-AACMGGATTAGATACCCKG-3′) and the reverse primer 1193R (5’-ACGTCATCCCCACCTTCC-3′). For fungal communities, the ITS region was amplified with the forward primer ITS1F (5’-CTTGGTCATTTAGAGGAAGTAA-3′) and the reverse primer ITS2 (5’-GCTGCGTTCTTCATCGATGC-3′). Each PCR reaction (25 μL) contained 0.25 μL Q5 high-fidelity DNA polymerase, 5 μL 5× reaction buffer, 5 μL 5× high-GC buffer, 2 μL dNTPs (10 mM), 2 μL template DNA, 1 μL forward primer (10 μM), 1 μL reverse primer (10 μM), and 8.75 μL nuclease-free water. The amplification program consisted of an initial denaturation at 98 °C for 5 min; 28 cycles of denaturation at 98 °C for 30 s, annealing at 53 °C for 30 s, and extension at 72 °C for 45 s; followed by a final extension at 72 °C for 5 min. PCR products were verified by 1% agarose gel electrophoresis and purified using the Agencourt AMPure XP nucleic acid purification kit (Beckman Coulter, MA, USA).

2.4. High-throughput sequencing and data processing

The purified PCR products were quantified using the Quant-iT PicoGreen dsDNA Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA), and sequencing libraries were constructed with the NEBNext Ultra II DNA Library Prep Kit (New England Biolabs, Ipswich, MA, USA). The quality of the constructed libraries was evaluated by quantification with the Qubit 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA), fragment size detection using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA), and concentration determination with the ABI StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The libraries were sequenced on the Illumina NovaSeq 6000 platform (Illumina, San Diego, CA, USA) using a paired-end 2 × 250 bp strategy with the NovaSeq 6000 SP Reagent Kit (500 cycles), conducted by Shanghai Biotree Biotech Co., Ltd. (Shanghai, China). Raw sequences that passed the initial quality screening were demultiplexed into samples based on index and barcode information, after which barcode sequences were removed. Paired-end reads were subjected to quality control and filtered according to sequencing quality. Cutadapt (v4.9) was used to remove primer sequences, and DADA2 was subsequently applied for filtering, denoising, merging, and chimera removal. After chimera removal, preliminary amplicon sequence variants (ASVs) were obtained, and singletons (ASVs with a count of 1) were further discarded to generate the final ASV set for taxonomic analysis. The SILVA database was used for bacterial annotation, while the UNITE database was employed for fungal annotation. Rarefaction curves were generated based on the analysis of observed features.

2.5. Statistical analysis

Principal component analysis (PCA) and PLS-DA were performed to evaluate the volatile profiles among PKT samples using SIMCA-P 14.1 software (Umetrics, Umeå, Sweden). PCA was conducted with unit variance scaling, whereas pareto scaling was applied for PLS-DA. Analysis of variance (ANOVA) was conducted using IBM SPSS Statistics 27 software (IBM Corp., Armonk, NY, USA). Volatiles with a variable importance in projection (VIP) score > 2.0 from PLS-DA and a p-value <0.05 in ANOVA were considered significant differential compounds for discriminating PKT samples. The relative abundances of these differential compounds were visualized in a heatmap using GraphPad Prism 10 (GraphPad Software Inc., San Diego, CA, USA). To explore the possible correlations between selected metabolites and microbial taxa, data normality was assessed using the Shapiro-Wilk test, followed by calculation of Pearson correlation coefficients based on their relative abundances using IBM SPSS Statistics 27 (Analyze → Correlate → Bivariate). The resulting correlation coefficients were visualized as a correlation heatmap using GraphPad Prism 10. To further elucidate the associations between differential volatile components and microbial communities, redundancy analysis (RDA) was performed using the vegan package in R (version 4.5.2).

3. Results and discussion

3.1. Volatile dynamics in PKTs across different storage years

The volatile profile of PKT samples stored for six distinct durations was comprehensively characterized using HS-SPME-GC–MS-based untargeted metabolomics. A total of 189 volatile compounds were identified and assigned to 13 chemical classes, highlighting the complex chemical composition underlying the aroma of aged PKTs. These included 31 alcohols, 26 ketones, 17 aldehydes, 17 ethers, 14 heterocyclic compounds, 14 monoterpenoids, 14 aromatic hydrocarbons, 13 phenols, 12 furans, 10 esters, 7 alkanes, 4 sesquiterpenoids, and 10 other compounds, together accounting for 79.19–83.51% of the total volatile fraction. The number of volatiles detected in this study was substantially higher than that reported for other dark teas, such as Liupao tea (53 volatiles) and Pu-erh tea (43) (Liang et al., 2025; Rong et al., 2023), highlighting the chemical richness of aged PKTs. Detailed information on the identified volatiles, including names, odor attributes, and relative abundances across storage years, is provided in Table S1.

At the class level, the five most abundant groups were alcohols (34.29–45.25%), aldehydes (8.01–16.61%), furans (4.76–8.08%), ketones (5.31–8.08%), and heterocyclic compounds (1.63–4.64%). In contrast, alkanes and sesquiterpenoids consistently accounted for less than 1% of the total. Significant variations were observed among different storage durations (Fig. 1). Alcohols dominated in all samples, peaking at 45.25% in PKT20. Aldehydes were relatively abundant in mid-term storage, reaching 16.61% in PKT10, but declined sharply thereafter. Furans and ketones exhibited moderate fluctuations, both peaking in PKT8 (8.08%). Heterocyclic compounds were most prominent in early-stage samples, particularly PKT3 (4.64%), while monoterpenoids were more abundant in long-stored samples such as PKT10, PKT15, and PKT20 (3.19–3.48%). These dynamic shifts reflect continuous biochemical transformations that shape the evolving aroma profile of PKTs.

Fig. 1.

Fig. 1

Open in a new tab

Distribution of volatile compound classes and their relative abundances in PKTs with different storage durations. PKT3-PKT20 denote Pukeng tea samples stored for 3, 5, 8, 10, 15, and 20 years, respectively.

At the individual compound level, the most abundant volatiles included 1-penten-3-ol (3.35–22.64%), 1-butanol (2.90–30.04%), hexanal (2.80–6.84%), 2-ethylfuran (1.13–4.50%), and cis-2-pentenol (0.22–3.76%) (Table S1). Notable variations in individual components were also detected across different storage years. For example, 1-penten-3-ol was most abundant in younger teas (22.64% in PKT5) but decreased to 3.35% in PKT20, whereas 1-butanol increased markedly from 2.91% in PKT8 to 30.04% in PKT20. Such pronounced shifts may be partly driven by the dynamics of microbial communities during long-term storage, which are known to modulate enzymatic activity and biochemical transformations in tea matrices (Liang et al., 2025). Previous studies have also identified 1-penten-3-ol and 1-butanol as dominant volatiles in fermented teas such as Fuzhuan brick tea (Xu et al., 2007), although their relative proportions were generally lower than those reported in the current study. These discrepancies may arise from differences in tea processing, storage conditions, or microbial succession (Cao et al., 2018; Xu et al., 2023). Among the identified volatiles, many (114 out of 189) were odor-active. Notably, 23 of the 30 most abundant compounds have been documented to have sensory attributes. For instance, 1-penten-3-ol imparts butter, pungent, green, and vegetable notes, while hexanal contributes grassy and fatty aromas (Table S1).

To further illustrate inter-sample variation, multivariate chemometric analyses including PCA and PLS-DA were performed (Fig. 2). PCA demonstrated good explanatory and predictive power, with cumulative R2X(cum) and Q2(cum) values exceeding 0.5 (Table S2). The PLS-DA model was similarly robust, as indicated by high cumulative R2Y(cum) and Q2(cum) values (Table S2), along with a negative y-intercept Q2 values from permutation tests (Fig. S1). As shown in Fig. 2, both PCA and PLS-DA score plots revealed clear separation among PKT samples, reflecting significant differences in volatile profiles across storage years. PKT8 clustered closer to PKT3 and PKT5, indicating relatively similar profiles in teas with shorter storage durations. In contrast, PKT10 and PKT15 diverged further, while PKT20 occupied a distinct position on the score plot, signifying the most differentiated volatile composition. These results indicate the potential impact of storage time on volatile composition, with prolonged storage leading to progressively greater differentiation. However, differences in primary growth conditions, such as temperature and humidity, may also partly contribute to the biochemical variations among PKT samples. Such findings are consistent with observations in other tea types, including white tea and Pu-erh tea, where volatiles were also significantly differed across storage durations (Rong et al., 2023; Wang, Wang, et al., 2022). Altogether, this study, for the first time, comprehensively presents the volatiles of PKT using untargeted metabolomics and demonstrates significant differences in volatile profiles among PKT samples with different storage times, as confirmed by chemometric analyses. The dynamic shifts in these volatiles may be associated with the evolving aromatic character of PKTs across different storage durations.

Fig. 2.

Fig. 2

Open in a new tab

Difference of volatile compositions across PKTs with different storage years using PCA (A) and PLS-DA (B) analyses. PKT3-PKT20 denote Pukeng tea samples stored for 3, 5, 8, 10, 15, and 20 years, respectively.

3.2. Differential volatile compounds in PKTs during long-term storage

To identify key volatiles that significantly distinguish PKTs across storage durations, a chemometric strategy integrating VIP scores derived from the multi-group PLS-DA model (Fig. 2B) and p-values from ANOVA was applied. This integrative approach is widely used to characterize significant metabolites contributing to overall group discrimination (Ji et al., 2023; Lin et al., 2025; Lin & Long, 2023). In the present study, compounds with VIP > 2 and p < 0.05 were considered significantly dynamic constituents across all storage years (Lin & Long, 2023). A total of 46 compounds were identified as key differential metabolites, primarily alcohols, aldehydes, and ketones (Table 1). To further characterize metabolic shifts relative to the early storage stage, pairwise PLS-DA analyses were performed by comparing the metabolomic profiles of PKT3 with those of each subsequent storage time point (Fig. S2). This complementary analysis identified 67 differentially abundant volatiles across all pairwise comparisons, with VIP values ranging from 2.00 to 14.83 (Table S3). Among these, 24 compounds, such as 1-butanol and 1-penten-3-ol, overlapped with those identified in the multi-group analysis, suggesting their possible involvement in the changes of the PKT volatilome during long-term storage.

Table 1.

Differential compounds responsible for the volatile variations in pkts with different storage timesa.

no.b compound RIc class VIP valued p valuee
8 1-butanol 1151 alcohols 8.32 < 0.001
10 1-penten-3-ol 1165 alcohols 7.77 < 0.001
37 hexanal 1356 aldehydes 6.50 < 0.001
17 1-hexanol 1452 alcohols 5.98 < 0.001
21 1-octen-3-ol 1254 alcohols 5.18 < 0.001
13 1-pentanol 1321 alcohols 4.59 < 0.001
139 isophorone 1603 ketones 4.24 0.014
180 2,5,5-trimethyl-1-hexen-3-yne 1115 others 4.23 0.014
78 dihydroactinidiolide 1131 esters 4.05 < 0.001
185 3,5,5-trimethyl-1-hexene 1384 others 3.71 < 0.001
88 3,4-dimethoxytoluene 1405 ethers 3.70 < 0.001
34 2-methylbutanal 1595 aldehydes 3.70 < 0.001
40 2-hexenal 1078 aldehydes 3.47 < 0.001
98 2-ethylfurane 911 furans 3.46 < 0.001
112 3,5-dimethylpyrazine 1210 heterocyclic compounds 3.35 < 0.001
36 pentanal 976 aldehydes 3.35 < 0.001
15 cis-2-pentenol 1180 alcohols 3.34 < 0.001
82 2,3-dimethylanisole 915 ethers 3.07 < 0.001
188 benzoic acid 1486 others 2.95 < 0.001
102 2-pentylfuran 2319 furans 2.86 < 0.001
144 2-hydroxy-5-methoxyacetophenone 1617 ketones 2.77 < 0.001
26 dichloroethanol 1800 alcohols 2.69 0.002
86 1,3-dimethoxybenzene 1497 ethers 2.67 < 0.001
6 3-pentanol 1741 alcohols 2.66 < 0.001
127 1-penten-3-one 951 ketones 2.50 < 0.001
137 3,5-(E,E)-octadien-2-one 1228 ketones 2.50 < 0.001
109 methylpyrazine 1324 heterocyclic compounds 2.46 0.003
39 heptanal 1261 aldehydes 2.44 < 0.001
7 1-methoxy-2-propanol 1317 alcohols 2.43 < 0.001
168 p-ethylguaiacol 1399 phenols 2.43 < 0.001
134 6-methyl-5-heptene-2-one 1585 ketones 2.43 0.003
150 limonene 1874 monoterpenoids 2.42 < 0.001
111 2,5-dimethylpyrazine 1018 heterocyclic compounds 2.40 0.003
18 trans-3-hexenol 1514 alcohols 2.37 0.002
133 2,6,6-trimethylcyclohexanone 1334 ketones 2.34 < 0.001
20 trans-2-hexenol 1307 alcohols 2.31 < 0.001
149 3-carene 1179 monoterpenoids 2.30 < 0.001
116 trimethylpyrazine 1189 heterocyclic compounds 2.28 < 0.001
128 2-heptanone 1139 ketones 2.26 0.007
35 3-methylbutanal 1276 aldehydes 2.25 < 0.001
166 o-cresol 967 phenols 2.24 0.004
45 2,4-heptadien-1-al 1485 aldehydes 2.13 < 0.001
189 dimethyl sulfone 1767 others 2.09 < 0.001
152 terpinolene 1886 monoterpenoids 2.08 < 0.001
25 1-methylcycloheptanol 2027 alcohols 2.06 < 0.001
73 minacide 1995 esters 2.04 < 0.001
Open in a new tab
a

Orders of differential compounds were ranked according to their VIP values. Additional details, including odor descriptions and specific relative abundances, are provided in Table S1.

b

Compound numbers correspond to those listed in Table S1.

c

Retention indices (RIs) were determined on a DB-Wax column using a series of C7-C40n-alkanes as references.

d

Variable importance in projection (VIP) values were obtained from the PLS—DA model in Fig. 2.

e

p values were calculated by analysis of variance (ANOVA).

Among the differential compounds identified by the multi-group model, 1-butanol was the most discriminative with a VIP value of 8.32, followed by 1-penten-3-ol (7.77), hexanal (6.50), and 1-hexanol (5.98), indicating their significant contributions to distinguishing PKTs with different storage years. To visualize the potential temporal dynamics of these differential volatiles, their relative abundances were illustrated in a heatmap (Fig. 3), revealing marked variations across samples of different storage years. Several compounds showed a decreased trend in relative abundance with increasing storage years, including 1-penten-3-ol, cis-2-pentenol and 1-penten-3-one. These volatiles are associated with grassy, green, or mushroom-like notes (Table S1), suggesting that the fresh and roasted aroma gradually weakens during long-term storage. In contrast, compounds such as 1-methoxy-2-propanol, 1-butanol, benzoic acid, and limonene displayed an overall increasing trend. Notably, limonene has been reported to positively correlate with storage duration in Tuo teas, showing a steady rise between 2 and 25 years (Chen et al., 2024), while benzoic acid has also been found to accumulate in aged black tea (Zhang et al., 2023), both of which are consistent with our observations. Certain microbes, including solvent-producing species such as Clostridium acetobutylicum, are known to synthesize 1-butanol, and this compound has been reported to increase significantly in Fuzhuan brick tea after fermentation (Xu et al., 2007). These previous findings suggest that microbial metabolism may contribute to its accumulation in PKTs during storage. In addition, some volatiles, such as p-ethylguaiacol, 1-methylcycloheptanol, 2-heptanone, and 3-carene, presented complex fluctuating patterns, e.g., an initial rise followed by a decline after 5 or 8 years, or even a decline followed by a later increase, implying that PKTs undergo transitional stages of aroma enhancement and attenuation during long-term storage. Although odor characteristics for 36 out of 46 (78.26%) differential compounds were available in odor databases, others like 2,5,5-trimethyl-1-hexen-3-yne, 2,6,6-trimethylcyclohexanone, and 3,5-(E,E)-octadien-2-one lacked sensory annotation, highlighting the need for further investigation. Overall, this study provides the first systematic identification of differential volatiles in PKTs with different storage years using chemometric approaches. These metabolites not only offered a chemical explanation for volatile variations among samples with different storage times but also served as potential biomarkers for authenticating PKT products with defined storage periods, thereby ensuring quality control.

Fig. 3.

Fig. 3

Open in a new tab

Dynamic changes in differential volatiles, as characterized by PLS-DA analysis, during PKT storage. PKT3-PKT20 represent Pukeng tea samples stored for 3, 5, 8, 10, 15, and 20 years, respectively.

3.3. Dynamics of microbial communities in PKT samples

The observed feature analysis revealed that the bacterial and fungal rarefaction curves for different PKT samples increased with sequencing depth and eventually approached saturation (Fig. S3), indicating that the sequencing depth was sufficient to capture the majority of microbial diversity present in the communities (Lin et al., 2020). To investigate dynamic changes of microbial communities across PKT samples with different storage durations, the sequences were analyzed at both the phylum and genus levels (Fig. 4). At the phylum level, seven bacterial phyla were identified: Actinobacteriota, Firmicutes, Proteobacteria, Deinococcota, Gemmatimonadota, Planctomycetota, and Bacteroidota (Fig. 4A). Among these, Actinobacteriota (21.82–78.57%) and Firmicutes (13.05–77.83%) were consistently dominant across all groups, together comprising 89.36–99.65% of the total bacterial communities, highlighting their central roles in PKTs. During the earlier storage period (3–8 years), the community was primarily composed of Firmicutes (56.92–77.83%), while Actinobacteriota accounted for a smaller proportion (21.82–42.45%). With prolonged storage, Actinobacteriota markedly increased, exceeding 60% in PKT10, PKT15, and PKT20, while Firmicutes declined to 13.05–29.25%. Proteobacteria exhibited elevated abundances only in PKT10 (8.35%) and PKT15 (10.62%), while other phyla (Deinococcota, Gemmatimonadota, Planctomycetota, and Bacteroidota) were sporadically detected at very low abundances (0–0.23%). Firmicutes, Actinobacteriota, and Proteobacteria have been commonly reported as dominant in other teas such as Fuzhuan brick tea, Qingzhuan brick tea, raw Pu-erh tea, or white tea (Fu et al., 2016; Wang et al., 2023), underscoring their importance in tea product ecosystems.

Fig. 4.

Fig. 4

Open in a new tab

Microbial community analysis of PKTs across different storage durations. Panels A-D depict the relative abundances of bacteria (C, D) and fungi (E, F) at the phylum and genus levels. PKT3-PKT20 correspond to Pukeng tea samples stored for 3, 5, 8, 10, 15, and 20 years, respectively.

At the genus level, 133 bacterial genera were annotated, with 29, 40, 50, 51, 51, and 50 detected in PKT3, PKT5, PKT8, PKT10, PKT15, and PKT20, respectively, indicating variations in the detected microbial diversity among PKT samples. The number of genera identified in this study exceeded those reported for Liupao tea (65) (Liang et al., 2025) and dark hawk tea (48) (Yang et al., 2024), but was considerably lower than that observed in Sichuan dark tea (1052) (Yan et al., 2021), reflecting marked differences among dark tea types. Among these genera, 11 dominant genera (relative abundance >1% in at least one group) were identified (Fig. 4B). The five most abundant were Staphylococcus (0.01–77.49%), Saccharopolyspora (6.44–31.90%), Rhodococcus (0–19.58%), Sciscionella (0–22.97%), and Turicibacter (0–18.97%). Staphylococcus predominated in the early stages (77.49%, 56.83%, and 72.28% in PKT3, PKT5, and PKT8, respectively) but declined dramatically in later stages to only 0.01% in PKT20. This genus is widely distributed in fermented foods, including various teas such as ripened Pu-erh tea (Ma et al., 2017; Yang et al., 2025), Liupao teas (Wang et al., 2021), and Sichuan dark tea (Yan et al., 2021). Although Staphylococcus has been reported to contribute to flavor, aroma, and color development in fermented meats (Fan et al., 2025), its specific sensory roles in tea samples remain to be elucidated. In contrast, Saccharopolyspora was relatively abundant throughout all PKT samples (6.44–31.90%), peaking in PKT10. Rhodococcus and Sciscionella showed marked fluctuations among samples, with their highest relative abundances observed in PKT20 at 19.58% and 22.97%, respectively. Turicibacter was undetectable in PKTs with storage times shorter than 20 years but emerged prominently in PKT20, reaching a relative abundance of 18.97%. Its occurrence may be associated with the development of increasingly anaerobic conditions during storage and the availability of specific metabolites generated in earlier stages (Lin et al., 2023). To our knowledge, this genus has not previously been reported in tea samples. Other genera, such as Bacillus and Actinopolyspora, were consistently detected throughout the storage period but exhibited fluctuating abundances. For example, Bacillus and Actinopolyspora fluctuated in different samples, and both peaked in PKT15 at 8.49% and 4.50%, respectively.

Compared to bacterial communities, fungal communities displayed significantly lower diversity at both the phylum and genus levels (Fig. 4). Only two phyla, Ascomycota and Basidiomycota, were identified, with Ascomycota overwhelmingly dominant (>99%) in all PKT groups (Fig. 4C), consistent with previous findings in several Chinese post-fermented teas such as dark green tea and Fuzhuan tea (Cui et al., 2025; Yan et al., 2021). A total of 110 genera were annotated, with 31, 30, 46, 30, 35, and 28 detected in PKT3, PKT5, PKT8, PKT10, PKT15, and PKT20, respectively. However, all samples were dominated by Aspergillus (87.76–99.27%) (Fig. 4D). Similar trends were observed in Fuzhuan tea (91.16%), although abundances were lower in Qingzhuan (54.89%), Tianjian (64.11%), and Liupao (47.43%) teas (Cui et al., 2025). Several Aspergillus species, such as A. niger and A. cristatus, have been reported to enhance taste and aroma in dark tea, green tea, and lightly fermented sour tea (Cai et al., 2023; Cai et al., 2025; Li, Zhang, et al., 2025), suggesting that Aspergillus plays a crucial role in tea fermentation and sensory quality, potentially serving as a key quality marker for PKTs. Given the overwhelming dominance of Aspergillus, the sequences were further analyzed at the species level, revealing several dominant species, including A. penicillioides, A. hordei, A. ruber, and A. restrictus (Fig. S4). These fungi may serve as functional candidates for PKT production; however, their isolation, taxonomic confirmation, and functional validation are necessary. Overall, this study is the first to characterize the microbial communities of PKT samples and their dynamics across storage durations, providing a foundation for understanding their functional roles and for guiding quality control in PKT production.

3.4. Correlations between microbial communities and volatile compositions

Microbial communities are widely recognized to be closely associated with volatile production in fermented products, including dark teas, thereby influencing their quality (Assad et al., 2023). To explore the possible relationships between microbes and volatiles in PKTs, Pearson correlation coefficients was determined based on the dynamic changes in microorganisms and volatile metabolites across different storage years. Dominant microorganisms (11 bacterial genera and one fungal genus) and key volatiles (differential metabolites and the 30 most abundant compounds) were selected for heatmap construction (Fig. 5).

Fig. 5.

Fig. 5

Open in a new tab

Associations between dominant microbial genera and volatile constituents based on Pearson's correlation analysis. Dotted lines demarcate different metabolite classes. Color values above and below zero indicate positive and negative correlations, respectively, between microbial genera and volatile compounds.

As shown in Fig. 5, both positive and negative correlations were observed between specific microbial taxa and volatile compounds in PKTs. Staphylococcus exhibited strong positive associations with more than half of the volatiles, including alcohols (e.g., 1-penten-3-ol, 1-pentanol, dichloroethanol), ethers (e.g., 2,3-dimethylanisole, 1,3-dimethoxybenzene), and heterocyclic compounds (e.g., methylpyrazine, 2,5-dimethylpyrazine), suggesting its key role in the accumulation of these metabolites. Staphylococcus is prevalent in fermented foods and known for its contributions to the formation and development of aromatic volatiles such as alcohols (Søndergaard & Stahnke, 2002), which supports its positive contributions to the aroma-active metabolites in the current study. Saccharopolyspora was positively correlated with volatiles such as hexanal, heptanal, 2-hexenal, and dihydroactinidiolide, while showing negative correlations with several compounds, including methylpyrazine and 1,3-dimethoxybenzene. The positive association with these aldehydes may be attributed to the amine oxidases produced by Saccharopolyspora, which catalyze the oxidative deamination of amino compounds, generating corresponding aldehydes and thereby promoting the accumulation of these flavor-active volatiles (Liu et al., 2022).

The bacteria of Rhodococcus, Sciscionella, and Turicibacter showed similar association patterns in terms of volatile production, while Bacillus, Actinopolyspora, and Noviherbaspirillum exhibited a distinct pattern. Specifically, Rhodococcus, Sciscionella, and Turicibacter contributed to the production of metabolites such as 1-methylcycloheptanol, 1-butanol, ethylbenzene, and benzoic acid and Bacillus, Actinopolyspora, and Noviherbaspirillum were positively associated with trans-3-hexenol, 1-octen-3-ol, and terpinolene. These genera were also negative to the production of several volatiles, such as heterocyclic compounds 3,5-dimethylpyrazine and trimethylpyrazine. Several Rhodococcus, such as Rhodococcus sp. VLD-10, are capable of producing benzoic acid and its derivatives (Yellamanda et al., 2016), which partly contributed to the observed positive associations with this compound. Although Bacillus spp. are known for their ability to produce high levels of ketones among total volatiles (Li et al., 2015), weak or even negative associations between this genus and ketones were observed in the present research, which may result from antagonistic interactions with other coexisting microbes (Stocki et al., 2025). Other dominant bacterial genera, namely Natribacillus and Virgibacillus, also showed both positive and negative correlations with volatile metabolites. While these two genera have been reported in fermented foods such as kimchi (Oh et al., 2025) or Yacai (Zhang et al., 2024), they have not been previously identified in tea microbiota. The present findings suggest that the presence of Natribacillus and Virgibacillus may contribute to the unique sensory characteristics of PKTs. The exclusive fungal genus Aspergillus, widely distributed in fermented foods and teas, also showed strong correlations with the production of several volatiles, such as 2-methylbutanal and 2,3-dimethylanisole, consistent with its recognized role in flavor development in post-fermented teas (Li, Du, et al., 2025).

To further examine the pairwise associations identified by Pearson correlation analysis, RDA was additionally conducted to evaluate microbe-volatile relationships at the community level (Fig. 6). The RDA ordination revealed clear associations between microbial succession and volatile profiles during storage, and the positioning of major microbial genera generally supported the correlation patterns observed in the heatmap. For instance, Staphylococcus clustered in the direction of several volatiles, including 1,3-dimethoxybenzene, dichloroethanol, 3,5-dimethylpyrazine and o-cresol, in agreement with their positive Pearson correlations. Similarly, Saccharopolyspora, Noviherbaspirillum, and Pseudomonas were positively associated with volatiles such as hexanal, 2-hexenal, and dihydroactinidiolide (Fig. 6), further supporting the Pearson correlation results. These complementary analyses suggest that microbial community dynamics collectively drive volatile evolution during PKT storage. Overall, the integrated application of Pearson correlation and RDA provided complementary evidence at both pairwise and community levels, revealing complex microbe-volatile associations and offering ecological insights into volatile formation, which may inform the targeted selection of aroma-active microorganisms for PKT industrial production.

Fig. 6.

Fig. 6

Open in a new tab

Correlations between dominant microbial genera and volatile constituents based on redundancy analysis (RDA). Compound numbers are consistent with those reported in Table 1 and Table S1.

4. Conclusion

This study presents the first comprehensive analysis of the volatile and microbial profiles of PKTs over a storage period of 3–20 years, integrating HS-SPME-GC–MS-based untargeted metabolomics, chemometrics, and microbiome analysis. In total, 189 volatile compounds were identified and classified into various groups, such as alcohols, aldehydes, furans, and ketones. Among these, 46 volatiles, like 1-butanol and 1-penten-3-ol, exhibited significant changes among samples and were identified as potential biomarkers for distinguishing storage durations. Microbial community structure shifted markedly with prolonged storage: early-stage samples were dominated by Firmicutes especially Staphylococcus, whereas extended storage led to increased abundance of Actinobacteriota, including Saccharopolyspora, Rhodococcus, Sciscionella, and Turicibacter. Fungal communities showed lower diversity, being overwhelmingly dominated by Aspergillus, with several species (e.g., A. penicillioides, A. ruber) potentially acting as functional contributors. Correlation analyses revealed complex positive and negative associations between key microbial genera (e.g., Staphylococcus, Saccharopolyspora, Rhodococcus) and aroma-active volatiles (e.g., 1-butanol, 1-penten-3-ol, hexanal), highlighting the microbial contributions to PKT's evolving flavor. Overall, these findings advance the understanding of the chemical and microbial diversity of PKTs across different storage durations, offering valuable markers for product authentication, quality control, and the discovery of functional microbes. Moreover, this study highlights the effectiveness of combining metabolomics, microbiome profiling, and chemometrics to explore metabolite–microbe interactions in traditionally fermented teas, thereby supporting their sustainable development and industrial application.

CRediT authorship contribution statement

Fengke Lin: Writing – original draft, Visualization, Software, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation. Yuedi Xu: Investigation, Data curation. Siyu Lin: Investigation, Data curation. Ziqi Zhao: Investigation, Data curation. Jingran Zhao: Investigation, Data curation. Chunsong Cheng: Writing – review & editing, Validation, Funding acquisition. Binsheng Luo: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Funding acquisition, Conceptualization.

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.

Acknowledgments

This study was supported by the National Natural Science Foundation of China (32300325 and 32400306), the “Xuncheng Talents” Program of Jiujiang City (No. JJXC2023136), the Key Research Projects in Jiangxi Province (20223BBH8007 and 20232BBG70014), and the Natural Science Foundation of Tianjin (No. 25JCQNJC00730).

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.fochx.2026.103692.

Appendix A. Supplementary data

Supplementary material

mmc1.docx (675.2KB, docx)

Data availability

Data will be made available on request.

References

  1. Assad M., Ashaolu T.J., Khalifa I., Baky M.H., Farag M.A. Dissecting the role of microorganisms in tea production of different fermentation levels: A multifaceted review of their action mechanisms, quality attributes and future perspectives. World Journal of Microbiology and Biotechnology. 2023;39(10):265. doi: 10.1007/s11274-023-03701-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Cai M., Huang L., Dong S., Diao N., Ye W., Peng Z., Fang X. Enhancing the flavor profile of summer green tea via fermentation with Aspergillus niger RAF106. Foods. 2023;12(18):3420. doi: 10.3390/foods12183420. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cai M., Peng Z., Xu P., Yu M., Diao N., Cao Y.…Fang X. Comprehensive analysis of the flavor and color characteristics of light-fermented sour tea mediated by Aspergillus niger RAF106. Food Chemistry. 2025;481 doi: 10.1016/j.foodchem.2025.143866. [DOI] [PubMed] [Google Scholar]
  4. Cao L., Guo X., Liu G., Song Y., Ho C.-T., Hou R., Zhang L., Wan X. A comparative analysis for the volatile compounds of various Chinese dark teas using combinatory metabolomics and fungal solid-state fermentation. Journal of Food and Drug Analysis. 2018;26(1):112–123. doi: 10.1016/j.jfda.2016.11.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen G., Chen C., Lei Z. Meta-omics insights in the microbial community profiling and functional characterization of fermented foods. Trends in Food Science & Technology. 2017;65:23–31. doi: 10.1016/j.tifs.2017.05.002. [DOI] [Google Scholar]
  6. Chen H., Liu Y., Zhang X., Chu J., Pu S., Wang W.…Xiong L. “Age” of tea: The impact of long-term storage on the aroma of Tuo tea and age prediction. Food Research International. 2024;187 doi: 10.1016/j.foodres.2024.11431. [DOI] [PubMed] [Google Scholar]
  7. Cui P., Li J., Yao T., Gan Z. Fungal community composition and function in different Chinese post-fermented teas. Scientific Reports. 2025;15(1):8514. doi: 10.1038/s41598-025-93420-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Fan Y., Badar I.H., Liu Q., Xia X., Chen Q., Kong B., Sun F. Insights into the flavor contribution, mechanisms of action, and future trends of coagulase-negative staphylococci in fermented meat products: A review. Meat Science. 2025;221 doi: 10.1016/j.meatsci.2024.109732. [DOI] [PubMed] [Google Scholar]
  9. Feng X., Chen M., Song H., Ma S., Ou C., Li Z.…Pan Y. A systemic review on Liubao tea: A time-honored dark tea with distinctive raw materials, process techniques, chemical profiles, and biological activities. Comprehensive Reviews in Food Science and Food Safety. 2023;22(6):5063–5085. doi: 10.1111/1541-4337.13254. [DOI] [PubMed] [Google Scholar]
  10. Fu J., Lv H., Chen F. Diversity and variation of bacterial community revealed by MiSeq sequencing in Chinese dark teas. PLoS One. 2016;11(9) doi: 10.1371/journal.pone.0162719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hayat K., Iqbal H., Malik U., Bilal U., Mushtaq S. Tea and its consumption: Benefits and risks. Critical Reviews in Food Science and Nutrition. 2015;55(7):939–954. doi: 10.1080/10408398.2012.678949. [DOI] [PubMed] [Google Scholar]
  12. Huang Y., Liu H., Wang J., Zhang R., Zhang Y., Liu Z., Pang Y., Yang C., Nie J. Impact of the long-term storage on flavor quality of Liupao tea using sensory evaluation combined with metabolomics analysis. Food Research International. 2024;198 doi: 10.1016/j.foodres.2024.115386. [DOI] [PubMed] [Google Scholar]
  13. Huang Y., Liu H., Zhang X., Wu Y., Liu Z., Pang Y., Liu R., Yang C., Nie J. Impact of storage time on non-volatile metabolites and fungal communities in Liupao tea using LC-MS based non-targeted metabolomics and high-throughput sequencing. Food Research International. 2023;174 doi: 10.1016/j.foodres.2023.113615. [DOI] [PubMed] [Google Scholar]
  14. Ji R., Garran T.A., Luo Y., Cheng M., Ren M., Zhou X. Untargeted metabolomic analysis and chemometrics to identify potential marker compounds for the chemical differentiation of Panax ginseng, P. quinquefolius, P. notoginseng, P. japonicus, and P. japonicus var. major. Molecules. 2023;28(6) doi: 10.3390/molecules28062745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Le M., Zhong L., Ren Z., An M., Long Y., Ling T. Dynamic changes in the microbial community and metabolite profile during the pile fermentation process of Fuzhuan brick tea. Journal of Agricultural and Food Chemistry. 2023;71(48):19142–19153. doi: 10.1021/acs.jafc.3c04459. [DOI] [PubMed] [Google Scholar]
  16. Li M., Du S., Xiao Y., Wu Y., Zhong K., Huang Y.…Gao H. Enhancing the quality of dark tea through fermentation with Aspergillus niger: Unveiling aroma and taste characteristics. Food Microbiology. 2025;128 doi: 10.1016/j.fm.2025.104721. [DOI] [PubMed] [Google Scholar]
  17. Li X., Mao Z., Wu Y., Ho H., He Y. Comprehensive volatile organic compounds profiling of Bacillus species with biocontrol properties by head space solid phase microextraction with gas chromatography-mass spectrometry. Biocontrol Science and Technology. 2015;25(2):132–143. doi: 10.1080/09583157.2014.960809. [DOI] [Google Scholar]
  18. Li X., Zhang C., Li S., Wang H., Yu P., Shao H.…Jin F. Unveiling key aroma-active compounds of two distinct aroma-types oriental melon: An integrated sensomics approach utilizing GC×GC-O-TOF-MS, GC-MS, aroma recombination and omission experiments. Food Chemistry: X. 2025;28 doi: 10.1016/j.fochx.2025.102517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Liang J., Wu H., Lu M., Li Y. HS-SPME-GC-MS untargeted metabolomics reveals key volatile compound changes during Liupao tea fermentation. Food Chemistry: X. 2024;23 doi: 10.1016/j.fochx.2024.101764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liang Y., Liu Z., Pang Y., Li M., Zheng S., Pan F., Guo C., Wu Q., Chen T., Li Q. Effects of storage durations on flavour and bacterial communities in Liupao tea. Food Chemistry. 2025;470 doi: 10.1016/j.foodchem.2024.142697. [DOI] [PubMed] [Google Scholar]
  21. Liao Z., Zong W. Effects of refining processes on the processing quality of Pukeng tea. Journal of Tea Business. 2021;43(1):37–40. [Google Scholar]
  22. Liao Z., Zong W., Zheng X., Lin J., Huang C., Wu Y. Pukeng tea from Qingxin District, Qingyuan City. Guangdong Tea Industry. 2020;4:21–23. [Google Scholar]
  23. Lin F., Cai F., Luo B., Gu R., Ahmed S., Long C. Variation of microbiological and biochemical profiles of Laowo dry-cured ham, an indigenous fermented food, during ripening by GC-TOF-MS and UPLC-QTOF-MS. Journal of Agricultural and Food Chemistry. 2020;68(33):8925–8935. doi: 10.1021/acs.jafc.0c0325. [DOI] [PubMed] [Google Scholar]
  24. Lin F., Long C. GC-TOF-MS-based metabolomics correlated with bioactivity assays unveiled seasonal variations in leaf essential oils of two species in Garcinia L. Industrial Crops and Products. 2023;194 doi: 10.1016/j.indcrop.2023.116356. [DOI] [Google Scholar]
  25. Lin F., Recchia M.J., Clark T.N., Kennelly E.J., Linington R.G., Long C. Metabolite profiling and characterization of potential anticancer constituents from Garcinia subfalcata using UPLC-IMS-QTOF-MS. Food Chemistry. 2025;465 doi: 10.1016/j.foodchem.2024.141900. [DOI] [PubMed] [Google Scholar]
  26. Lin T.-C., Soorneedi A., Guan Y., Tang Y., Shi E., Moore M.D., Liu Z. Turicibacter fermentation enhances the inhibitory effects of Antrodia camphorata supplementation on tumorigenic serotonin and Wnt pathways and promotes ROS-mediated apoptosis of Caco-2 cells. Frontiers in Pharmacology. 2023;14 doi: 10.3389/fphar.2023.1203087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Liu S., Yao H., Sun M., Zhou Z., Mao J. Heterologous expression and characterization of amine oxidases from Saccharopolyspora to reduce biogenic amines in Huangjiu. LWT - Food Science and Technology. 2022;169 doi: 10.1016/j.lwt.2022.113963. [DOI] [Google Scholar]
  28. Ma Y., Duan S., Zhang D., Su X., Zhang D., Lv C., Zhao M. Microbial succession and the dynamics of chemical compounds during the solid-state fermentation of Pu-erh tea. Applied Sciences. 2017;7(2):166. doi: 10.3390/app7020166. [DOI] [Google Scholar]
  29. Mu X., Wang J. A preliminary study on the change trend of main aroma components in Pukeng tea during storage. Guangdong Tea Industry. 2025;2:2–6. [Google Scholar]
  30. Oh Y.J., Kim J.Y., Kwon M.-S., Lee S., Choi S.-P., Choi H.-J. Virgibacillus saliphilus sp. nov. and Virgibacillus salidurans sp. nov., isolated from kimchi. Journal of Microbiology. 2025;63 doi: 10.71150/jm.2501001. [DOI] [PubMed] [Google Scholar]
  31. Rong Y., Xie J., Yuan H., Wang L., Liu F., Deng Y.…Yang Y. Characterization of volatile metabolites in Pu-erh teas with different storage years by combining GC-E-nose, GC-MS, and GC-IMS. Food Chemistry: X. 2023;18 doi: 10.1016/j.fochx.2023.100693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Søndergaard A.K., Stahnke L.H. Growth and aroma production by Staphylococcus xylosus, S. carnosus and S. Equorum-a comparative study in model systems. International Journal of Food Microbiology. 2002;75(1–2):99–109. doi: 10.1016/S0168-1605(01)00729-2. [DOI] [PubMed] [Google Scholar]
  33. Stocki M., Stocka N., Borowik P., Dudzińska M., Staszowska A., Okorski A., Oszako T. Use of volatile organic compounds produced by Bacillus bacteria for the biological control of Fusarium oxysporum. Forests. 2025;16(8):1220. doi: 10.3390/f16081220. [DOI] [Google Scholar]
  34. Wang C., Han J., Pu Y., Wang X. Tea (Camellia sinensis): A review of nutritional composition, potential applications, and omics research. Applied Sciences. 2022;12(12):5874. doi: 10.3390/app12125874. [DOI] [Google Scholar]
  35. Wang J., Zhang J., Chen Y., Yu L., Teng J., Xia N., Wei B., Xiao S., Huang L. The relationship between microbial dynamics and dominant chemical components during Liupao tea processing. Food Bioscience. 2021;43 doi: 10.1016/j.fbio.2021.101315. [DOI] [Google Scholar]
  36. Wang Z., Ahmad W., Zhao S., Zhu A., Huo S., Chen Q. Temporal analysis of non-targeted flavor component variation in green tea storage and rapid prediction of storage duration. Food Chemistry. 2025;464 doi: 10.1016/j.foodchem.2024.141898. [DOI] [PubMed] [Google Scholar]
  37. Wang Z., Wang Z., Dai H., Wu S., Song B., Lin F., Huang Y., Lin X., Sun W. Identification of characteristic aroma and bacteria related to aroma evolution during long-term storage of compressed white tea. Frontiers in Nutrition. 2022;9 doi: 10.3389/fnut.2022.1092048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wang Z., Wang Z., Dai W., Huang Y., Lin F., Dai H., Zhou Z., Lin X., Lin Z., Sun W. The relationship between bacterial dynamics, phenols and antioxidant capability during compressed white tea storage. LWT - Food Science and Technology. 2023;174 doi: 10.1016/j.lwt.2022.114418. [DOI] [Google Scholar]
  39. Xu W., Zhao Y., Lv Y., Bouphun T., Jia W., Liao S., Zhu M., Zou Y. Variations in microbial diversity and chemical components of raw dark tea under different relative humidity storage conditions. Food Chemistry: X. 2023;19 doi: 10.1016/j.fochx.2023.100863. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Xu X., Mo H., Yan M., Zhu Y. Analysis of characteristic aroma of fungal fermented Fuzhuan brick-tea by gas chromatography/mass spectrophotometry. Journal of the Science of Food and Agriculture. 2007;87(8):1502–1504. doi: 10.1002/jsfa.2874. [DOI] [Google Scholar]
  41. Yan K., Yan L., Meng L., Cai H., Duan A., Wang L.…Abbas M. Comprehensive analysis of bacterial community structure and diversity in Sichuan dark tea (Camellia sinensis) Frontiers in Microbiology. 2021;12 doi: 10.3389/fmicb.2021.735618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Yang Y., Peng J., Li Q., Song Q., Cronk Q., Xiong B. Optimization of pile-fermentation process, quality and microbial diversity analysis of dark hawk tea (Machilus rehderi) LWT - Food Science and Technology. 2024;192 doi: 10.1016/j.lwt.2023.115707. [DOI] [Google Scholar]
  43. Yang Z., Xie Y., Zhu Y., Lei M., Chen X., Jin W., Fu C., Yu L. Unraveling the flavor formation process of mellow and thick-type ripened Pu-erh tea through non-targeted metabolomics and metagenomics. Food Chemistry: X. 2025;27 doi: 10.1016/j.fochx.2025.102424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Yellamanda B., Vijayalakshmi M., Kavitha A., Reddy D.K., Venkateswarlu Y. Extraction and bioactive profile of the compounds produced by Rhodococcus sp. VLD-10. 3 Biotech. 2016;6(2) doi: 10.1007/s13205-016-0576-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Zhai X., Zhang L., Granvogl M., Ho C.T., Wan X. Flavor of tea (Camellia sinensis): A review on odorants and analytical techniques. Comprehensive Reviews in Food Science and Food Safety. 2022;21(5):3867–3909. doi: 10.1111/1541-4337.12999. [DOI] [PubMed] [Google Scholar]
  46. Zhang H., Meng F., Zhou Y., Luo S., Yang Q., Li Y., Liu D. Effects of industrial-scale pickling processes on the dynamic changes in the physicochemical indicators and microbial diversities of Yacai. Food Chemistry: X. 2024;23 doi: 10.1016/j.fochx.2024.101637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zhang L., Zhang Z., Zhou Y., Ling T., Wan X. Chinese dark teas: Postfermentation, chemistry and biological activities. Food Research International. 2013;53(2):600–607. doi: 10.1016/j.foodres.2013.01.016. [DOI] [Google Scholar]
  48. Zhang S., Li Q., Wen S., Sun L., Chen R., Zhang Z., Cao J., Lai Z., Li Z., Lai X. Metabolomics reveals the effects of different storage times on the acidity quality and metabolites of large-leaf black tea. Food Chemistry. 2023;426 doi: 10.1016/j.foodchem.2023.136601. [DOI] [PubMed] [Google Scholar]
  49. Zheng W., Wan X., Bao G. Brick dark tea: A review of the manufacture, chemical constituents and bioconversion of the major chemical components during fermentation. Phytochemistry Reviews. 2015;14(3):499–523. doi: 10.1007/s11101-015-9402-8. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary material

mmc1.docx (675.2KB, docx)

Data Availability Statement

Data will be made available on request.

Articles from Food Chemistry: X are provided here courtesy of Elsevier

RESOURCES