Skip to main content
NCBI home page
Food Chemistry: X logoLink to Food Chemistry: X
. 2025 Sep 1;30:102976. doi: 10.1016/j.fochx.2025.102976

Unraveling the mystery of microecology behind the fermented Chinese sauerkraut (jiangshui): flavor profile, community structure, and functional taxa

Shuai Zhao a,b,c,d, Chang Hao a,b, Linzhe Huang a,b, Yuchen Gao a,b,c, Lei Wang b,c, Jing Lu a,b,c, Yawei Shi b,c,d,⁎
PMCID: PMC12445595  PMID: 40980281

Abstract

In this study, we aimed to elaborate the flavor profile, community structure, and functional taxa of Chinese sauerkraut (jiangshui), and revealed the associations between flavor profiles and core microbiota. We identified the 19 markers to characterize the regional flavor profiles by multivariate statistical analysis, and constructed the synthesis network of 19 volatile compounds. Additionally, Weissella, Schleiferactobacillus, Latilactobacillus, and Pichia were screened as biomarkers for fermented jiangshui, respectively. Neutral Community Model analysis suggested that the assembly of bacterial community was more dominated by stochastic processes, whereas deterministic factors, particularly humidity and temperature, drove the dynamic of fungal community. Furthermore, there was the commensal relationship between P. fermentans and four acid-producing bacteria. Moreover, P. fermentans, L. plantarum, and S. harbinensis played an important role in shaping the unique flavor and taste of jiangshui. This study provided the insights for understanding the mechanism of jiangshui fermentation and regulation of fermentation process.

Keywords: Jiangshui, Fermentation, Microbial community, Flavor profile, Functional taxa

Graphical abstract

Unlabelled Image

Open in a new tab

Highlights

  • •

    19 volatiles were identified to characterize flavor profiles of Shanxi jiangshui.

  • •

    The synthesis network of 19 flavor compounds was predicted and constructed.

  • •

    NCM analysis was used to evaluate the community assembly of jiangshui.

  • •

    There was commensalism between Pichia fermentans and acid-producing bacteria.

  • •

    P. fermentans, L. plantarum and S. harbinensis could improve the jiangshui flavor.

1. Introduction

Chinese sauerkraut (jiangshui) is a traditional fermented food crafted from vegetables such as cabbage, celery, and carrots(Jin et al., 2024). As a delicacy in the northwest of China, the primary production areas of jiangshui include Gansu, Qinghai, Shaanxi, Ningxia, and Shanxi province. Jiangshui is a pale white color liquid with a slightly sour taste, it is conducive to enhance taste and eliminate fishy odors or greasy sensations. As early as in Li Shizhen's “Compendium of Materia Medica”, there was a record stating that “jiangshui has a cooling nature and is effective in dispersing heat, thus relieving thirst and resolving stagnation”, indicating that there is a certain medicinal effect for jiangshui in addition to being a condiment.

The fermentation process of jiangshui involves the synergistic action of various microorganisms, which accumulate a diversity of flavor compounds and nutrients during fermentation. With the advancement of modern biotechnology, the research on the microbial community of jiangshui has attracted widespread attention, mainly focusing on analysis of community structure and diversity(Li, Ju, et al., 2022), isolation and assessment of fermented strains(Guo et al., 2025), and detection of metabolites in jiangshui(Zhang et al., 2023). The bacterial community structure of jiangshui in Nanchang, China was explored via culture-dependent and denaturing gradient gel electrophoresis (DGGE) analysis. The researcher found that diverse lactic acid bacteria (LAB) species belonging to the Lactobacillus and Weissella dominated the bacteria community during jiangshui fermentation, and the potential pathogenic bacteria were detected in jiangshui samples(J. Zhang et al., 2018), which may compromise the safety and quality of jiangshui. Besides, the functional characteristic of core strain in jiangshui was also evaluated. For example, the Lactiplantibacillus plantarum isolated from jiangshui could inhibited the growth of Aspergillus flavus, and could survive and colonize in the gastrointestinal tract, without hemolysin activity(Ou et al., 2022). Additionally, the researchers also explored the metabolites and differences in the main metabolic pathways in jiangshui fermentation, suggesting that 761 different metabolites were detected in jiangshui from different regions, and the structure of community was highly correlated with the components of metabolites(Zhang et al., 2023).

Traditional jiangshui fermentation predominantly depends on microorganisms enriched in raw materials or the environment. This creates a potential risk of coexistence between beneficial and harmful microorganisms. Furthermore, the fermentation process is significantly influenced by local climate and terroir(Niccum et al., 2020), making precise control over fermentation challenging. It is precisely the complex community in fermented jiangshui that brings great challenges to the development of pure-culture fermentation technologies. Notably, it was evident that current research primarily focused on the analysis of community structure and dynamic monitoring of metabolites, while neglecting the association between community assembly and local terroir as well as the quality of jiangshui. Specially, the core functional taxa in jiangshui and the connection between these core taxa and flavor as well as taste remain unclear. Therefore, it is essential to elucidate the microecology behind the fermented jiangshui, and explore the functional taxa of jiangshui fermentation.

In this study, we focused on the fermented jiangshui from Shanxi Province, China, aiming to clarify the structure and assembly of jiangshui microbial community, reveal the associations between core microbiota and jiangshui flavor as well as taste, and identify the functional taxa. This study was conducive to provide the insights for understanding the mechanism of jiangshui fermentation and promote the transformation of jiangshui production from spontaneous fermentation to pure-culture fermentation.

2. Materials and methods

2.1. Materials and analytical reagents

Internal standard (2-octanol) used in GC–MS analysis was purchased from Aladdin Reagent (Shanghai) Co., Ltd. Other analytical grade chemical reagents were purchased from Sinopharm Group (Shanghai, China).

2.2. Collection of jiangshui samples

The jiangshui samples were collected from Jinzhong (JZ), Lvliang (LL), Taiyuan (TY), and Yangquan (YQ) City of Shanxi Province, China, as shown in Fig. 1A. Each sample was collected for three independent biological replicates from October to December 2024. Fig. 1B showed the preparation process of jiangshui. Briefly, the cabbage was thoroughly washed and sectioned into uniform pieces, which were subsequently subjected to thermal pretreatment by immersion in boiling water for 60 s, followed by cooling to room temperature. A starch-based liquid was prepared by incorporating wheat flour (5 % w/v) into boiling water with constant stirring until complete homogenization. The blanched vegetables were transferred to sterilized containers and immersed in the cooled starch-based liquid. Finally, fermented jiangshui from the previous batch was added as starter. During the fermentation process, the jiangshui is stirred once a day, and the period of fermentation lasted around 7 days.

Fig. 1.

Fig. 1

Open in a new tab

Collection of jiangshui samples and the production process of jiangshui. (A) The jiangshui samples were collected from Jinzhong (JZ), Lvliang (LL), Taiyuan (TY) and Yangquan (YQ) in Shanxi Province. (B) Schematic diagram of the jiangshui production process.

2.3. Analysis of physicochemical characteristics and sensory evaluation

The pH of jiangshui samples was detected by pH meter (FB2200, OHAUS, Jiangsu, China). The titratable acidity of jiangshui in different regions was measured with titration by neutralization with 0.1 mol/L NaOH(Akbar et al., 2024). The color of jiangshui samples was measured by chroma meter (CR400, Konica Minolta, Osaka, Japan) and presented by a* (redness), b* (yellowness), and L* (lightness) values. All the above assays were conducted in triplicate. The sensory evaluation of jiangshui was conducted by a panel of 20 sensory assessors referring the published method(Zhao et al., 2022). Ethical permission was given by Research Ethics Committee for sensory evaluation in Shanxi university. All sensory assessors participated in this research have signed the consent form and agreed to carry out this sensory evaluation experiment. Fig. S1 showed the consent form in this sensory evaluation of jiangshui. The samples were stored at 4 °C prior to analysis. The attributes assessed included color, flavor, taste, and overall appearance.

2.4. Analysis of volatile organic compounds by GC–MS

Volatile organic compounds (VOCs) of jiangshui samples were analyzed by headspace solid phase microextraction-gas chromatography–mass spectrometry (HS-SPME-GC–MS). 5 mL of jiangshui and 1.5 g NaCl were taken into 20 mL headspace vial, 10 μL 2-octanol (5 ppb) was added as internal standard, and the headspace vial was placed in the autosampler. VOCs were extracted by SPME at 50 °C for 45 mins and determined by GC–MS system equipped with Agilent DB-wax column (60 m × 250 μm × 0.25 μm). The inlet temperature was 250 °C, and the injection mode was 1:1 split injection, the sample was operated in constant current mode with a flow rate of 1.0 mL/min. The GC temperature was programmed to warm up from 45 °C (holding for 2 mins) to 80 °C at a rate of 4 °C/min, then to 150 °C at a rate of 5 °C/min, and then to 230 °C (10 mins) at a rate of 10 °C/min. The MS was operated using an EI ion source, with a temperature of 230 °C, electron energy 70 eV, and the m/z acquisition range was 20–400 amu. VOCs were identified by comparing the mass spectra and retention indices with data from National Institute of Standards and Technology Mass Spectrometry (NIST2017). The assays were in triplicate and with three independent biological replicates.

2.5. High-throughput amplicon sequencing and data analysis

Total genome DNA of jiangshui samples was extracted using a FastDNA spin kit for soil (MP Biomedicals, USA) following the manufacturer's instructions, and its concentration and purity was evaluated by Nanodrop 2000 spectrophotometer (Thermo Scientific, USA). The V3-V4 variable regions of bacterial 16S rRNA genes and ITS1 region of fungus were amplified used specific primer 341F(5′-CCTAYGGGRBGCASCAG-3′) and 806R(5′-GGACTAC-NNGGGTATCTAAT-3′), ITS1-F(5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS1-R(5′-GCTGCGTTCTTCATCGATGC-3′), respectively. All PCR reactions were carried out with 15 μL of Phusion® High-Fidelity PCR Master Mix (New England Biolabs). PCR reaction procedure as following: 2 μM of forward and reverse primers, and about 10 ng template DNA. Thermal cycling consisted of initial denaturation at 98 °C for 1 min, followed by 30 cycles of denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, and elongation at 72 °C for 30 s. Finally, 72 °C for 5 mins. After that, the mixture PCR products was purified with Universal DNA (TianGen, China). The sequencing libraries were constructed using NEB Next® Ultra DNA Library Prep Kit (Illumina, USA) following manufacturer's description. The library quality was assessed on the Agilent 5400 (Agilent Technologies Co Ltd., USA). And then the constructed DNA library was sequenced on an Illumina platform.

Raw data was imported into the format which could be operated by QIIME2 system. Demultiplexed sequences from each sample were quality filtered, de-noised, merged, and then the chimeric sequences were removed using the QIIME2 dada2 plugin to obtain the feature table of amplicon sequence variant (ASV). The ASVs for 16S or ITS amplicon were taxonomically classified by Greengenes (v13_8) and UNITE databases with a confidence threshold of 0.6. The obtained data had been deposited at NCBI Sequence Read Archive, including PRJNA1265511 (SUB15333724) and PRJNA1265629 (SUB15333953). Diversity metrics were calculated to estimate the microbial diversity within an individual sample by core-diversity plugin within QIIME2. Beta diversity distance measurements were performed to investigate the structural variation of microbial communities across samples.

2.6. Isolation of strains and microbial co-culture

Isolation and screening of strains was carried out in accordance with the published methods(Jia et al., 2020). Briefly, 10 mL of jiangshui samples were homogenized with 90 mL of sterile saline (0.9 % NaCl) in conical flasks and incubated at 180 rpm at 30 °C  for 60 mins. Subsequently, the suspension was serially diluted 10-fold in sterile saline (from 10−1 to 10−6) and spread onto MRS agar (supplemented with 0.1 % nystatin) and YPD agar (supplemented with 0.1 % ampicillin). Following this, all inoculated plates were incubated at 30 °C for 72 h. Colonies exhibiting distinct morphologies and sizes were selected, and the corresponding strains were cultured in MRS or YPD liquid medium. Genomic DNA was extracted using the Genome Extraction Kit (TIANGEN, Beijing, China), followed by PCR amplification of taxonomic marker genes: the 16S rRNA gene (using primers 27F/1492R) for bacteria and the ITS region (using primers ITS1/ITS4) for fungi. The resulting amplicons were sequenced and compared against the NCBI GenBank database. The isolated strains were stored in 25 % glycerol tubes and preserved at −80 °C for long-term culture.

Co-culture of Pichia fermentans and 4 acid-producing bacteria (Schleiferilactobacillus harbinensis, Lactobacillus parafarraginis, Lentilactobacillus parafarraginis, and Lactobacillus plantarum) were conducted following previously established methods(Zhao et al., 2023). The two strains were co-inoculated in 100 mL jiangshui medium (1 % wheat flour, 2 % cabbage) with equal proportions (1.0 %) at a concentration of 107 CFU/mL, and then incubated at 30 °C for 96 h. The mono-cultures served as the control group. The fermentation broth was continuously diluted and then spread onto the MRS agar (supplemented with 0.1 % nystatin) and YPD agar (supplemented with 0.1 % ampicillin), respectively. After incubation at 30 °C for 72 h, the number of colonies was recorded. The fermented jiangshui samples were subjected to sensory evaluation, while the supernatant was collected after filtration for subsequent pH determination. The experiment was performed independently three times in parallel.

2.7. Statistical analysis

One-way analysis of variance (ANOVA) was executed with SPSS v22.0 software (IBM, USA). Orthogonal partial least squares discriminant analysis (OPLS-DA) was done by SIMCA (version 14.1) (UMETRICS, Umeå, Sweden). Random Forest (RF) analysis was caried out via online (https://www.bioincloud.tech). The predicted metabolic of volatile compounds was conducted by website (https://www.metaboanalyst.ca/) and KEGG pathway analysis. The potential KEGG Ortholog (KO) functional profiles of microbial communities were predicted with PICRUST2. Correlation analysis was conducted in RStudio statistical environment (RStudio, Inc. USA; v1.4.1717) by R package ‘psych’ using the Spearman's rank correlation algorithm. Analysis of neutral community model (NCM) was carried out using the ‘vegan’ package in R statistical environment(Zorea et al., 2024). Additional data analysis and image processing were performed using EXCEL (Microsoft, USA) and GraphPad Prism 8.3 (GraphPad Software, USA).

3. Results and discussion

3.1. Analysis of physicochemical parameters of jiangshui samples

Acidity is an important indicator of fermented food, which is closely related to the flavor and quality of product. The titratable acidity and pH of jiangshui samples were measured from TY, LL, YQ, and JZ, as shown in Fig. 2. It was found that there were significant differences in the accumulation of titratable acidity among the samples from different regions (p < 0.05) (Fig.2A). The titratable acidity of the jiangshui in JZ region was the highest (11.19 ± 0.13 g/L), while the titratable acidity of the jiangshui in TY region was the lowest (7.01 ± 0.05 g/L). The titratable acidity of the jiangshui from LL (9.8 ± 0.08 g/L) was higher than that from YQ (8.77 ± 0.05 g/L). Similar results were also reflected in the change of pH value (Fig.2B). The titratable acidity of jiangshui varied among different regions, which might be attributed to the community composition and abundance differences of acid-producing bacteria among samples.

Fig. 2.

Fig. 2

Open in a new tab

Physicochemical parameters of jiangshui samples from JZ, LL, TY and YQ. (A) pH; (B) Titratable acidity; (C) Color of jiangshui samples; (D) Chromaticity index. The data was presented in the form of average value ± standard deviation value, ‘a’ means p < 0.001, ‘b’ means 0.001 ≤ p < 0.01, ‘c’ means 0.01 ≤ p < 0.05, ‘d’ means 0.05 ≤ p < 0.1.

Fig. 2C presented the color of 4 jiangshui samples. It was found that the L* value of the jiangshui from TY, LL, and JZ was lower than that of the samples from YQ (8.02 ± 0.05) (Fig.2D, Table.S1). Additionally, the a* value of the jiangshui from LL was −0.67 ± 0.01, showing a more distinct green tone. The b* value of jiangshui ranged from 1.41 to 2.43, indicating that the color of jiangshui in different regions was relatively close in the yellow-blue color (Fig.2D, Table.S1). The differences of color among the 4 jiangshui samples might be attributed to the variations in types of vegetables and quality of water during the fermentation process. Overall, there were significant differences in titratable acidity, pH and chromaticity among the jiangshui samples from 4 regions, which were related to the variations in the microbial community, raw materials and fermentation crafts.

3.2. Analysis of volatile organic compounds in jiangshui from different aeras

Flavor profile is critical to characterize the quality of fermented food. The VOCs in the jiangshui from TY, LL, JZ and YQ were detected (Fig.S2 and Table.S2). Acids were the main volatile components in the jiangshui from TY and JZ, while alcohols were the main volatile components in the jiangshui from LL and YQ (Fig.3A). Notably, the content of butyric acid in the jiangshui samples from TY was significantly higher than that in others (Table.S2). Butyric acid could contribute to the cheese and buttery aroma, but it could impart an unpleasant sweat-like odor at high concentrations, this might be a typical flavor characteristic for the TY jiangshui sample. Additionally, the concentration of ethyl acetate was the highest in jiangshui from LL (Table.S2), the accumulation of ethyl acetate mainly results from the metabolism of yeasts, which is conducive to provide the floral and fruity aroma(Wang et al., 2024), it was speculated that a higher abundance of yeasts may have been enriched in the LL samples.

Fig. 3.

Fig. 3

Open in a new tab

Analysis of volatile compounds in jiangshui samples from JZ, LL, TY and YQ.

(A) Concentrations of volatile compounds in different jiangshui samples; (B) The screening of marker volatile compounds in different jiangshui samples by OPLS-DA analysis. The volatile compounds with variable importance projection (VIP) greater than 1 and –log10 (FDR adjusted p) greater than 3 was selected; (C) The screening of marker volatile compounds in different jiangshui samples by RF analysis, the top 35 volatile compounds with mean decrease accuracy were selected.

(D) The predicted metabolic network on substrate breakdown and synthesis of 19 characteristic volatile compounds in jiangshui. The blue, yellow, and pink modules represent amino acid metabolism, carbohydrate metabolism, and lipid metabolism, respectively. The red dotted blocks denoted the substrate. The black dotted blocks denoted the characteristic volatile compounds. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

To screen the typical volatile compounds of jiangshui from TY, LL, JZ, and YQ, Orthogonal Partial Least Squares Discriminant analysis (OPLS-DA) and random forest analysis (RF) were performed. For OPLS-DA analysis (FC ≥ 2, p ≤ 0.05, VIP ≥ 1), it was observed that the spatial distance among samples from different regions was relatively substantial (Fig.S3), indicating that there was unique flavor profile for the fermented jiangshui in different regions. This was likely to the differences in environmental factors such as temperature and humidity among regions, which shaped the unique ‘native microbial communities’ and caused the change in accumulation of flavor metabolites. Besides, n-butanol and N-valeric acid with higher VIP scores and FDR values were noticed for distinguishing the flavor characteristics of jiangshui from different regions (Fig.3B). N-valeric acid is mainly accumulated by lactic acid bacteria and could contribute to the sour taste, while n-butanol could provide a fruity or grassy aroma. Additionally, there was higher Mean Decrease Accuracy for isocaprylic acid, 1-nonyl alcohol, and 4-ethylphenol after RF analysis (Fig.3C), indicating that these compounds served as critical discriminators for jiangshui from different regions.

Although differential volatile compounds among the four regions were identified by OPLS-DA and RF analysis, variations in the results between these two analyses were observed, potentially attributable to algorithmic biases inherent in the discriminant methods(Zhao et al., 2022). So, the results of OPLS-DA and RF were integrated. A total of 19 volatile compounds were screened to characterize the differences of chemosensory, including 8 alcohols, 6 acids, 3 phenols, 1 aldehyde, and 1 ester (Table.S3). Obviously, the regional differences in flavor compounds were primarily manifested in the changes of alcohols, acids, and phenols. The accumulation of butyric acid and caproic acid in the TY samples was significantly higher than in other samples (p < 0.05), indicating that these compounds serve as characteristic volatile compounds of TY jiangshui. Similarly, acetic acid and propionic acid were identified as marker volatile compounds associated with the flavor profile of JZ jiangshui due to the significantly higher accumulation (p < 0.05). In addition, 4-ethylphenol, as the characteristic volatile compound in the LL samples, contributed the jiangshui with a slightly sweet and fragrant aroma(Haag et al., 2023). The accumulation of (E) -3-hexene-1-ol in the jiangshui samples of YQ was significantly higher (p < 0.05), contributing to the grassy aroma(Pichersky et al., 2006).

The accumulation of characteristic flavor compounds is closely associated with the community structure and metabolic functions. To understand the synthesis pathways of 19 marker volatile compounds, the predicted metabolic network was constructed, as shown in Fig. 3D. It was noticed that the accumulation of 19 characteristic volatiles involved the glucose metabolism, amino acid metabolism, and fatty acid metabolism. Leucine, valine, phenylalanine, and tyrosine served as key precursor compounds, promoting the synthesis of 3-ethyl-4-methylpenton-1-ol, 4-methylvalerate, isobutyric acid, benzaldehyde, eudinol, 4-ethyl-2-methoxyphenol(Zhao et al., 2024). Lactic acid bacteria and yeast could catalyze the metabolism of leucine and valine to synthesize 3-ethyl-4-methylpenton-1-ol and isobutyric acid by secreting branched-chain aminotransferase and keto acid decarboxylase(Yu et al., 2025). It was speculated that the reason for the higher accumulation of 3-ethyl-4-methylpenton-1-ol and isobutyric acid in the TY region may be attributed to the more efficient leucine and valine metabolism. Moreover, the accumulation of benzaldehyde was higher in the jiangshui samples of LL and YQ, benzaldehyde can be synthesized through phenylalanine ammoniase secreted by Lactobacillus(Zou et al., 2025), speculating that the abundance of Lactobacillus in the LL and YQ might be higher. Isocaprylic acid, enanthic acid, and butyric acid were accumulated by microbial decomposition and metabolism of starch and cellulose present in raw materials(Zhou et al., 2024). Propionic acid is mainly synthesized by Lactobacillus plantarum through heterologous fermentation pathways(Zhou et al., 2024), speculated that there was much higher biomass of L. plantarum in the JZ samples. Furthermore, linolenic acid and linoleic acid is crucial precursors for the synthesis of (E)-3-hexene-1-ol, cis-2-pentene-1-ol, 2-methyl-6-heptene-1-ol, and cis-4-heptene-1-ol, which could be accumulated through lipoxygenase (LOX) pathway after metabolism of yeast or Bacillus (Zhang et al., 2025), a greater abundance of fatty acid-metabolizing microbiota might be present during the fermentation of LL jiangshui. Overall, there were significant differences in the flavor profiles of jiangshui in TY, LL, JZ and YQ regions, and a total of 19 marker volatile compounds were screened to characterize the chemosensory profile of different regions. Because the accumulation of key flavor compounds intrinsically linked to microbial metabolic process(Shen et al., 2021), we supposed that the ‘native microbial communities’ played a pivotal role in formation of flavor profiles, it was essential to explore the microbial community structure and function in fermented jiangshui.

3.3. Analysis of community structure in jiangshui fermentation

The fermentation of jiangshui is mediated by synergistic interactions among diverse microbial strains. Given that variations in microbial community structure directly influence the execution of metabolic functions during fermentation(Gaur et al., 2020). Therefore, the elucidation of region-specific differences in microbial ecology and functional profiles became imperative. To elucidate the region-specific microbiota of jiangshui, the microbial community of jiangshui from JZ, LL, TY, and YQ regions was analyzed via 16S and ITS amplicon sequencing.

Through α-diversity analysis, significant differences were observed in microbial community structures among the 4 regions. For bacterial taxa, the jiangshui of YQ demonstrated markedly higher Shannon index (3.79 ± 0.09) and Simpson index (0.86 ± 0.02) compared to those from JZ, TY, and LL regions (Fig.4A and 4C), indicating that there was the most abundant diversity for bacterial taxa in YQ region. Furthermore, no significant differences were detected for the diversity of bacterial community between JZ and TY samples. This phenomenon may be attributed to the geographical proximity of JZ and TY regions, where analogous climatic and edaphic conditions likely contribute to convergent evolution in the structural configuration of microbial ecosystems. For fungal taxa, jiangshui samples from YQ and LL exhibited significantly higher Shannon index and Simpson index compared to those from JZ and TY (p < 0.05) (Fig.4B and 4D), indicating that the diversity of fungal community in YQ and LL was more abundant during jiangshui fermentation. Notably, fungal taxa played a pivotal role in the biosynthesis of key flavor compounds(Ma et al., 2022). The elevated fungal diversity observed in YQ and LL regions may contribute to the formation of distinct flavor profiles during jiangshui fermentation, suggesting that the potential linkage between the diversity of fungal taxa and the variation of metabolic characteristics. Analysis of β-diversity employing the weighted UniFrac algorithm revealed significant divergence in bacterial community structure among TY, LL, JZ, and YQ jiangshui samples (p < 0.01) (Fig.4E). Interestingly, fungal communities demonstrated spatial autocorrelation patterns, TY and JZ samples, as well as YQ and LL samples exhibited the higher community structural similarity (Bray Curtis>0.75) (Fig.4F).

Fig. 4.

Fig. 4

Open in a new tab

Analysis of microbial community structure in jiangshui. (A) Shannon's diversity of bacterial taxa; (B) Simpson's diversity of bacterial taxa; (C) Shannon's diversity of fungus taxa; (D) Simpson's diversity of fungus taxa; (E) β-diversity of jiangshui bacterial community by PCoA analysis; (F) β-diversity of jiangshui fungal community by PCoA analysis; (G) Relative abundances of bacterial genera in Top 20; (H) Relative abundances of fungus genera in Top 20.

After classification and annotation based on ASV sequences, we focused on the top 20 microbial genera in terms of abundance in 4 different jiangshui samples, as shown in Fig. 4G and 4H. It was observed that Schleiferilactobacillus, Caproicibacter, Lacticaseibacillus, and Clostridium were the dominant bacterial taxa in the jiangshui from TY region, while Lentilactobacillus, Pediococcus, and Acetobacter were the dominant bacterial groups in JZ region. Latilactobacillus, Weissella, Leuconostoc, and Companilactobacillus were all dominant bacterial taxa in the YQ and LL regions, suggested that the bacterial community structure of jiangshui in Shanxi was obviously different from Shaanxi and Henan province(J. Zhang et al., 2018). Additionally, the relative abundance of Latilactobacillus (51.86 ± 4.07 %) and Leuconostoc (26.08 ± 7.00 %) occupied the completely dominant position in the jiangshui samples from LL region, which were much higher than those in the YQ region (p < 0.01). Conversely, the relative abundance of Weissella (35.92 ± 5.42 %) in the jiangshui from YQ region was significantly higher than that in the LL region (16.40 ± 8.70 %) (p < 0.05) (Fig.4G). For fungal groups, Pichia dominated the fermentation process of jiangshui in the TY and JZ regions, with an average relative abundance exceeding 80 %. Notably, the community structure of fungal taxa in the LL and YQ regions was more complex than that in the TY and JZ regions (Fig.4H), which might be caused by the differences in raw materials and fermentation crafts(Zhang, Yu, et al., 2023). Additionally, the changes of raw material pretreatment also play a crucial role in shaping the structure of jiangshui microbial community. When the raw materials are subjected to thermal pretreatment by immersion in boiling water prior to fermentation, which not only reduces the microbial load in raw materials but also mediates the structure of jiangshui microbial community. Overall, the microbial community structure of jiangshui from JZ, LL, TY, and YQ regions appeared the significant differences in diversity and relative abundance, which might be the key factor leading to the unique flavor profiles of jiangshui among 4 regions. Taken together, the microbial community structure of fermented jiangshui in Shanxi province has distinct local characteristics. There was much higher diversity for fungal community in LL and YQ samples. Moreover, Lentilactobacillus, Pediococcus, Leuconostoc, Schleiferactobacillus, and Pichia were identified as biomarkers of jiangshui fermentation.

3.4. Minning of biomarkers in different jiangshui fermentation micro-environment

To further excavate the biomarker microorganisms of jiangshui fermentation micro-environment in JZ、LL、TY、YQ, LEfSe analysis was performed with the P value less than 0.05 and LDA score more than 4 being considered significant(Li et al., 2022), as shown in Fig. S3. It was observed that the biomarkers of jiangshui exhibited significant regional differences. Regarding bacterial communities, Lentilactobacillus and Pediococcus were identified as biomarkers in jiangshui samples from JZ, whereas Latilactobacillus and Leuconostoc were characterized as key species in the jiangshui samples from LL. Additionally, Schleiferactobacillus and Caproicibacter were found to be biomarkers of fermented jiangshui in TY, while Weissella and Companilactobacillus were specific to YQ (Fig.S4A). For fungal taxa, Pichia stood out as an iconic species due to its higher LDA value in the JZ samples, while Fusarium and Zea were identified as biomarkers in LL samples. Furthermore, Aspergillus and Naumovozyma were determined to be the biomarkers in YQ samples (Fig.S4B).

To reveal the variation of microbiota function among different jiangshui samples, the metabolic function of the community was predicted by PICRUST2. It was noticed that carbohydrate metabolism, amino acid metabolism, metabolism of cofactors and vitamins, and metabolism of other amino acids were the dominant metabolic function of jiangshui microbial community. Moreover, there were significantly fluctuation for the abundance of dominant metabolic pathway in JZ, LL, TY, and YQ samples, as shown in Fig. S5 and Table.S4. Furthermore, the relative abundances of carbohydrate metabolism in JZ (12.49 ± 0.01 %) and TY (12.75 ± 0.03 %) samples were significantly higher than that in LL (10.66 ± 0.10 %) and YQ (11.43 ± 0.16 %) (p < 0.05), this was likely to the higher abundance of microorganisms with starch metabolism capacity during LL and YQ jiangshui fermentation. Amino acids are the precursors of flavor compound synthesis, and more abundant amino acid metabolism is conducive to promote the diversification of flavor profile(Zhang et al., 2022). It was observed that the abundance of amino acid metabolism in YQ samples (9.03 ± 0.20 %) was much higher than that in JZ samples (8.11 ± 0.06 %) (p < 0.05), which might be attributed to higher microbial community diversity in fermented jiangshui from YQ, which was consistent with the analysis of community structure. Furthermore, the synthesis of flavor compounds is closely related to the catabolism of amino acids(Falasconi et al., 2020), which might be one of the reasons for the higher diversity of VOCs in YQ jiangshui samples. Additionally, it was also noticed that there were significant differences for the metabolism of cofactors and vitamins, which was associated with stability of community structure and environmental adaptability(Zhang et al., 2023), this was probably caused by the variations of environment and terroir in JZ, LL, TY, and YQ regions. Taken together,

3.5. Analysis of microbial community assembly process

To understand the mechanisms of community assembly, neutral community models (NCM) was used to quantify the neutral processes (stochastic processes) of microbial taxa assembly in JZ, LL, TY and YQ regions during jiangshui fermentation(Sun et al., 2021), as shown in Fig. 5. Elevated R2values derived from model fitting reflect the enhanced congruence between community patterns and neutral model expectations, suggesting an increased predominance of stochasticity in assembly dynamics(Sun et al., 2022).

Fig. 5.

Fig. 5

Open in a new tab

Analysis of the driving forces for microbial community dynamics and the associations with flavor formation in jiangshui. (A) Analysis of neutral community models (NCM) based on bacterial community structure; (B) Analysis of neutral community models (NCM) based on fungal community structure. The dotted line represents the 95 % confidence interval around the model prediction. The R2 value ranges from 0 (no fitting) to 1 (perfect fitting), and represents the goodness-of-fit for the NCM. (C) RDA analysis of the Top 10 bacterial taxa; (D) RDA analysis of the Top 10 fungi taxa. RDA analysis reflects the correlation between the environmental factors and microbial community. The red arrows represent the environmental factors, including P (Average sea level pressure), RH (Relative humidity), DPT (Average dew point temperature) and T (Average temperature), and the blue arrows represent the abundance of microbiota, the circles represent the jiangshui samples from different regions. (E) Correlation analysis between bacteria taxa and volatile compounds. (F) Correlation analysis between fungal taxa and volatile compounds. The orange circles represent positive correlations and the green circles represent negative correlations, ‘*’ represents 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.)

It was found that the higher model-fitted R2 values (JZ: R2 = 0.658, LL: R2 = 0.695, TY: R2 = 0.494, YQ: R2 = 0.566) were noticed for bacterial taxa compared to fungal communities(Fig.5A), indicating that the assembly process of the bacterial community was closer to the prediction of the neutral model, and in the multi-strain fermentation system of jiangshui, the stochastic process was the main driving force for the assembly of bacterial communities in fermented jiangshui from the four regions. Since jiangshui fermentation relies on the accumulation of metabolic products of core functional strains, the fluctuation of functional bacterial taxa under the dominance of stochastic processes was the key factors leading to the stability of jiangshui quality during fermentation. Notably, JZ (R2 = 0.456) and YQ (R2 = 0.426) displayed higher NCM-fitted R2 values compared to LL (R2 = 0.258) and TY (R2 = 0.381) (Fig.5B), suggesting that fungal communities in JZ and YQ were primarily governed by stochastic processes. In contrast, the NCM-fitted R2 values of the fungal communities in LL (R2 = 0.258) and TY (R2 = 0.381) were much lower, indicating that the deterministic process played a more crucial role in the assembly process of fungal community. Particularly, the dominance of deterministic process was most prominent in the LL samples (R2 = 0.258). This might be attributed to local specific environmental conditions, such as pH, temperature or humidity, which mediate the assembly of fungal communities, highlighting the critical influence of environmental selection pressures in shaping microbial diversity.

To clarify the role of environmental factors in mediating the assembly of microbial communities in different regions, the data of climate dynamics was collected for TY, JZ, LL, and YQ from October to December 2024, including dew point temperature, mean sea level pressure, average temperature, and relative humidity, and further quantified the directional effects of environmental gradients on shaping the community structure of jiangshui fermentation through RDA analysis(Huang et al., 2020). As shown in Fig. 5C, relative humidity and dew point temperature contributed to the shaping of community structure by 40.6 % (p < 0.05) and 37.7 % (p < 0.05), which significantly influenced the community assembly of jiangshui fermentation in YQ samples. Moreover, the Weissella, Companilactobacillus, Latilactobacillus, and Leuconostoc exhibited positive correlations with fluctuations in relative humidity, suggesting that relative humidity played a pivotal role for driving bacterial community formation.

Additionally, average temperature, relative humidity, and dew point temperature were positively associated with the fungal genera Cystofilobasidium, Zea, Sporobolomyces, Fusarium, Filobasidium and Aspergillus while negatively correlated with Pichia (Fig.5D). These findings indicated that the assembly of the fungal community was predominantly governed by dynamic variations in environmental temperature. Appropriate temperature could enhance the enzymatic catalytic activity of fungi such as Cystofilobasidium, thereby promoting the growth and reproduction. In contrast, elevated environmental temperature may reduce the enzymatic activity of Pichia, thereby inhibiting its growth(Tian et al., 2022). Moreover, relative humidity and dew point temperature influenced the growth of filamentous fungi by modulating the water balance within fungal cells. Overall, the climatic differences among various regions provide a broader ecological niche for fungal taxa. The assembly of fungal community in jiangshui fermentation was more regulated by environmental factors, such as temperature and relative humidity, while the assembly of bacterial community was more dominated by stochastic process.

3.6. Associations between microbial community and jiangshui flavor profiles

To reveal the relationships between microbial community and jiangshui flavor, the heatmaps of correlation was constructed between the top 20 microbial abundances and volatile compounds in different jiangshui samples (Fig.5E and 5F). It was observed that Lacticaseibacillus, Lentilactobacillus, Loigolactobacillus, Pediococcus, Schleiferilactobacillus, and Acetobacter exhibited the significant positive correlations with acids. This suggested that these bacterial taxa played a crucial role in the accumulation of organic acids during jiangshui fermentation, meanwhile, it also contributed to the sour taste of jiangshui. Moreover, the acid-producing bacteria could reduce the pH in jiangshui fermentation system via the accumulation of organic acids, thereby inhibiting the growth of certain harmful microorganisms and promoting the purification of community in the jiangshui micro-environment via regulation of microbial metabolism, the similar findings had been observed in vinegar fermentation(T. Huang et al., 2022). Moreover, previous study indicated that some microbial metabolism gave priority to organic acids rather than carbohydrates, and the accumulation of organic acids was an important carbon source for microbial growth and metabolism(Wiesenbauer et al., 2025). Additionally, it was found that Clostridium, Caproicibacter, Caproiciproducens, and Klebsiella are positively correlated with phenols and terpenes. Notably, dominant mold taxa, such as Cladosporium, Alternaria, Aspergillus, and Penicillium, showed the significant negative correlations with acids in jiangshui while positively correlated with ester compounds. These results suggested that organic acids accumulated by acid-producing bacteria in jiangshui might inhibited the growth of molds.

Ester compounds are key flavor substances in fermented foods and play a critical role in shaping flavor profile(Que et al., 2023). As shown in Fig. 5F, it can be inferred that fungal taxa may be closely associated with the accumulation of esters during jiangshui fermentation. Furthermore, Pichia exhibited a positive correlation with the accumulation of acids (Fig.5F), speculating that there may be a synergistic interaction between Pichia and acid-producing bacteria. The synergistic interactions between acid-producing bacteria and yeasts could be a key factor for driving the assembly of community.

3.7. Verification of functional taxa for improving the jiangshui flavor by co-culture

According to the above analysis, Pichia and acid-producing bacteria were found to have potential roles in enhancing the flavor and taste of jiangshui. To validate this hypothesis, P. fermentans and 4 strains of acid-producing bacteria (S. harbinensis, Lac. parafarraginis, Len. parafarraginis, and L. plantarum) were isolated, and subjected to mono-culture and co-culture during jiangshui fermentation, as shown in Fig. 6A. After sensory evaluation, it was observed that the scores for color, flavor, and taste of the fermented jiangshui was the lowest when P. fermentans and 4 acid-producing bacteria were inoculated alone, and the color of fermented jiangshui appeared the yellowish-green, and the flavor profile was weak (Fig.6B). However, when P. fermentans was co-cultured with the acid-producing bacteria, the pH values (Co-culture 1: pH = 5.00 ± 0.02，Co-culture 2: pH = 5.17 ± 0.01，Co-culture 3: pH = 5.29 ± 0.02，Co-culture 4: pH = 5.29 ± 0.01) were significantly lower than those in mono-culture (p < 0.05)(Fig.6C), and the taste and flavor of jiangshui were richer and fuller, presenting a higher sensory evaluation score. Particularly, when P. fermentans was co-cultured with L. plantarum and S. harbinensis, the score of sensory evaluation was the highest (Fig.6B), indicating that P. fermentans, L. plantarum, and S. harbinensis played an important role in shaping the unique taste and flavor of jiangshui.

Fig. 6.

Fig. 6

Open in a new tab

Verification of functional taxa for improving the jiangshui flavor by co-culture. (A) The schematic diagrams of mono-culture and co-culture of Pichia fermentans and 4 strains of acid-producing bacteria (Schleiferilactobacillus harbinensis, Lactobacillus parafarraginis, Lentilactobacillus parafarraginis, and Lactobacillus plantarum); (B) Scores of sensory evaluation;

(C) Changes of pH value in mono-culture and co-culture; Fig. (D) to (H) showed the change of biomass for Pichia fermentans and 4 strains of acid-producing bacteria in mono-culture and co-culture; (D) Pichia fermentans; (E) Lactobacillus plantarum; (F) Lactobacillus parafarraginis; (G) Lentilactobacillus parafarraginis; (H) Schleiferilactobacillus harbinensis.

To further characterize the interaction relationship between P. fermentans and 4 acid-producing bacteria, the biomass was determined during mono-culture and co-culture (Fig.6D to 6H). Notably, when P. fermentans was co-cultured with 4 acid-producing bacteria, its biomass was significantly reduced compared to that observed under mono-culture condition (p < 0.05), indicating that the growth of P. fermentans was inhibited in the co-culture system (Fig.6D). Conversely, the biomass accumulation of S. harbinensis, Lac. parafarraginis, Len. parafarraginis, and L. plantarum increased substantially after co-culture with P. fermentans (Fig.6E to 6H). This suggested a commensalism relationship between P. fermentans and 4 strains of acid-producing bacteria (S. harbinensis, Lac. parafarraginis, Len. parafarraginis, and L. plantarum). The dominance of acid-producing bacteria in co-culture may be closely associated with the accumulation of organic acids. Since acid-producing bacteria taxa could efficiently metabolize carbohydrates via glycolysis(Zhao et al., 2023), leading to the accumulation of organic acids such as lactic acid and acetic acid, thereby inhibiting the growth of P. fermentans via changing the pH of fermentation environment(Bangar et al., 2022). Previous studies have demonstrated that the affinity of specific transport proteins in lactic acid bacteria for pantothenic acid was significantly greater than that of P. fermentans. Consequently, the intake of pantothenic acid by P. fermentans might be reduced during co-culture with acid-producing bacteria, thereby restricting its biomass accumulation(Salvatore et al., 2025). The structural stability and functional characterization of microbial communities often arises from the interactions among multiple microbial species within the community(Bajic & Sanchez, 2020). As a multi-strain fermented delicacy in the northwest of China, the intricate interaction mechanisms within the microbial community of jiangshui warrant in-depth exploration in the future.

4. Conclusion

Taken together, this study systematically elucidated the microecology behind the fermented Shanxi jiangshui from flavor profile, community structure, and functional taxa. The characteristic volatile compounds and biomarkers were identified for the fermented Shanxi jiangshui. NCM analysis revealed that stochastic processes dominated the bacterial community assembly, whereas deterministic environmental factors (notably humidity and temperature) governed the fungal dynamics, highlighting the connection between jiangshui fermentation microecology and local terroir. Additionally, we also confirmed the relationship of commensalism between P. fermentans and acid-producing bacteria, and illustrated P. fermentans, L. plantarum, and S. harbinensis were conducive to improve the flavor and taste of fermented jiangshui. Overall, these findings provided the insights for understanding the mechanism of jiangshui fermentation and regulating the process of jiangshui fermentation. Future research will focus on the analysis of microbial interactions among functional strains and development of synthetic microbiome in jiangshui fermentation.

CRediT authorship contribution statement

Shuai Zhao: Writing – review & editing, Writing – original draft, Visualization, Supervision, Project administration, Funding acquisition, Data curation. Chang Hao: Visualization, Methodology, Investigation, Formal analysis. Linzhe Huang: Visualization, Investigation. Yuchen Gao: Supervision. Lei Wang: Supervision. Jing Lu: Supervision. Yawei Shi: Supervision, Project administration, Formal analysis.

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.

Acknowledgement

This study was financially supported by Fundamental Research Program of Shanxi Province (No. 202403021222016), Opening Project of Xinghuacun College of Shanxi University (Shanxi Institute of Brewing Technology and Industry) (No. XCSXU-KF-202407), Scientific and Technological Innovation Programs of Higher Education Institutions in Shanxi (No. 2024L005).

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.fochx.2025.102976.

Appendix A. Supplementary data

Supplementary material

mmc1.docx (3.8MB, docx)

Data availability

Data will be made available on request.

References

  1. Akbar M., Ali N., Imran M., Hussain A., Hassan S.W., Haroon U., Munis M.F.H. Spherical Fe2O3 nanoparticles inhibit the production of aflatoxins (B1 and B2) and regulate total soluble solids and titratable acidity of peach fruit. International Journal of Food Microbiology. 2024;410 doi: 10.1016/j.ijfoodmicro.2023.110508. [DOI] [PubMed] [Google Scholar]
  2. Bajic D., Sanchez A. The ecology and evolution of microbial metabolic strategies. Current Opinion in Biotechnology. 2020;62:123–128. doi: 10.1016/j.copbio.2019.09.003. [DOI] [PubMed] [Google Scholar]
  3. Bangar S.P., Suri S., Trif M., Ozogul F. Organic acids production from lactic acid bacteria: A preservation approach. Food Bioscience. 2022;46 doi: 10.1016/j.fbio.2022.101615. [DOI] [Google Scholar]
  4. Falasconi I., Fontana A., Patrone V., Rebecchi A., Garrido G.D., Principato L., Morelli L. Genome-assisted characterization of Lactobacillus fermentum, Weissella cibaria, and Weissella confusa strains isolated from Sorghum as starters for sourdough fermentation. Microorganisms. 2020;8(9) doi: 10.3390/microorganisms8091388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Gaur G., Oh J.-H., Filannino P., Gobbetti M., Van Pijkeren J.-P., Ganzle M.G. Genetic determinants of Hydroxycinnamic acid metabolism in Heterofermentative lactobacilli. Applied and Environmental Microbiology. 2020;86(5) doi: 10.1128/aem.02461-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Guo C., Sun Y.X., Chen H.J., Yin G.Y., Song Y.Z. Identification and assessment of Pichia kudriavzevii YS711 isolated from “Jiangshui” with the capacity for uric acid metabolism. Microbiological Research. 2025;298 doi: 10.1016/j.micres.2025.128200. [DOI] [PubMed] [Google Scholar]
  7. Haag F., Frey T., Hoffmann S., Kreissl J., Stein J., Kobal G., Krautwurst D. The multi-faceted food odorant 4-methylphenol selectively activates evolutionary conserved receptor OR9Q2. Food Chemistry. 2023;426 doi: 10.1016/j.foodchem.2023.136492. [DOI] [PubMed] [Google Scholar]
  8. Huang L.B., Bai J.H., Wen X.J., Zhang G.L., Zhang C.D., Cui B.S., Liu X.H. Microbial resistance and resilience in response to environmental changes under the higher intensity of human activities than global average level. Global Change Biology. 2020;26(4):2377–2389. doi: 10.1111/gcb.14995. [DOI] [PubMed] [Google Scholar]
  9. Huang T., Lu Z.M., Peng M.Y., Chai L.J., Zhang X.J., Shi J.S., Xu Z.H. Constructing a defined starter for multispecies vinegar fermentation via evaluation of the vitality and dominance of functional microbes in an autochthonous starter. Applied and Environmental Microbiology. 2022;88(3) doi: 10.1128/aem.02175-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Jia Y., Niu C.T., Lu Z.M., Zhang X.J., Chai L.J., Shi J.S., Li Q. A bottom-up approach to develop a synthetic microbial community model: Application for efficient reduced-salt broad bean paste fermentation. Applied and Environmental Microbiology. 2020;86(12) doi: 10.1128/AEM.00306-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Jin D.X., Jin Y.X., Zhang W., Cao W., Liu R., Wu M.G., Yu H. Exploring Jiangshui-originated probiotic lactic acid bacteria as starter cultures: Functional properties and fermentation performances in pear juice. Food Bioscience. 2024;61 doi: 10.1016/j.fbio.2024.104982. [DOI] [Google Scholar]
  12. Li P.Y., Ju N., Zhang S.Z., Wang Y.Y., Luo Y.L. Evaluation of microbial diversity of Jiangshui from the Ningxia Hui autonomous region in China. Food Biotechnology. 2022;36(2):173–190. doi: 10.1080/08905436.2022.2054818. [DOI] [Google Scholar]
  13. Li Y.L., Liu S.P., Zhang S.Y., Liu T.T., Qin H., Shen C.H., Mao J. Spatiotemporal distribution of environmental microbiota in spontaneous fermentation workshop: The case of Chinese baijiu. Food Research International. 2022;156 doi: 10.1016/j.foodres.2022.111126. [DOI] [PubMed] [Google Scholar]
  14. Ma D., Li Y., Chen C.C., Fan S.C., Zhou Y., Deng F.M., Zhao L.Y. Microbial succession and its correlation with the dynamics of volatile compounds involved in fermented minced peppers. Frontiers in Nutrition. 2022;9 doi: 10.3389/fnut.2022.1041608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Niccum B.A., Kastman E.K., Kfoury N., Robbat A., Wolfe B.E. Strain-level diversity impacts cheese rind microbiome assembly and function. Msystems. 2020;5(3) doi: 10.1128/mSystems.00149-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ou D.X., Ling N., Wang X.H., Zou Y.Y., Dong J.J., Zhang D.F., Ye Y.W. Safety assessment of one Lactiplantibacillus plantarum isolated from the traditional Chinese fermented vegetables-Jiangshui. Foods. 2022;11(15) doi: 10.3390/foods11152177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Pichersky E., Noel J.P., Dudareva N. Biosynthesis of plant volatiles: Nature’s diversity and ingenuity. Science. 2006;311(5762):808–811. doi: 10.1126/science.1118510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Que Z.L., Jin Y., Huang J., Zhou R.Q., Wu C.D. Flavor compounds of traditional fermented bean condiments: Classes, synthesis, and factors involved in flavor formation. Trends in Food Science & Technology. 2023;133:160–175. doi: 10.1016/j.tifs.2023.01.010. [DOI] [Google Scholar]
  19. Salvatore M.M., Maione A., Buonanno A., Guida M., Andolfi A., Salvatore F., Galdiero E. Biological activities, biosynthetic capacity and metabolic interactions of lactic acid bacteria and yeast strains from traditional home-made kefir. Food Chemistry. 2025;470 doi: 10.1016/j.foodchem.2024.142657. [DOI] [PubMed] [Google Scholar]
  20. Shen D., Shen H., Yang Q., Chen S., Dun Y., Liang Y., Zhao S. Deciphering succession and assembly patterns of microbial communities in a two-stage solid-state fermentation system. Microbiology Spectrum. 2021;9(2) doi: 10.1128/Spectrum.00718-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Sun P., Huang X., Wang Y., Huang B.Q. Protistan-bacterial microbiota exhibit stronger species sorting and greater network connectivity offshore than nearshore across a Coast-To-Basin continuum. Msystems. 2021;6(5) doi: 10.1128/mSystems.00100-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Sun X., Zhang X.K., Zhang G.S., Miao Y.J., Zeng T.X., Zhang M., Huang L.F. Environmental response to root secondary metabolite accumulation in Paeonia lactiflora: Insights from rhizosphere metabolism and root-associated microbial communities. Microbiology. Spectrum. 2022;10(6) doi: 10.1128/spectrum.02800-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Tian W., Zhang H.Y., Guo Y.H., Wang Z.Y., Huang T.S. Temporal and spatial patterns of sediment microbial communities and driving environment variables in a shallow Temperate Mountain river. Microorganisms. 2022;10(4) doi: 10.3390/microorganisms10040816. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Wang F., Zhao P., Du G., Zhai J., Guo Y., Wang X. Advancements and challenges for brewing aroma-enhancement fruit wines: Microbial metabolizing and brewing techniques. Food Chemistry. 2024;456 doi: 10.1016/j.foodchem.2024.139981. [DOI] [PubMed] [Google Scholar]
  25. Wiesenbauer J., Gorka S., Jenab K., Schuster R., Kumar N., Rottensteiner C., Kaiser C. Preferential use of organic acids over sugars by soil microbes in simulated root exudation. Soil Biology & Biochemistry. 2025;203 doi: 10.1016/j.soilbio.2025.109738. [DOI] [Google Scholar]
  26. Yu P., Wang J., Lao F., Shi H.M., Xu X.B., Wu J.H. Investigation on sweaty off-flavors in small mill sesame oil and its formation mechanism via molecular sensory science, preparative gas chromatography, and microbiomics. Food Chemistry. 2025;463 doi: 10.1016/j.foodchem.2024.141224. [DOI] [PubMed] [Google Scholar]
  27. Zhang D.D., Yu H., Yang Y.C., Liu F., Li M.Y., Huang J., Yan Q.Y. Ecological interactions and the underlying mechanism of anammox and denitrification across the anammox enrichment with eutrophic lake sediments. Microbiome. 2023;11(1) doi: 10.1186/s40168-023-01532-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Zhang H., Xiang S., Zhai R., Li X., Hu M., Wang T., Pan L. Analysis of microbial and metabolic diversity in Jiangshui from Northwest China. Food Science and Technology. 2023;43 doi: 10.1590/fst.107222. [DOI] [Google Scholar]
  29. Zhang H.X., Tan Y.W., Wei J.L., Du H., Xu Y. Fungal interactions strengthen the diversity-functioning relationship of solid-state fermentation systems. Msystems. 2022;7(4) doi: 10.1128/msystems.00401-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Zhang J., Wu S.S., Zhao L.H., Ma Q.L., Li X., Ni M.Y., Zhu H.L. Culture-dependent and-independent analysis of bacterial community structure in Jiangshui, a traditional Chinese fermented vegetable food. LWT- Food Science and Technology. 2018;96:244–250. doi: 10.1016/j.lwt.2018.05.038. [DOI] [Google Scholar]
  31. Zhang X., Zheng Y.R., Liu Z.M., Su M.Y., Wu Z.J., Xu X.M. Integrated analysis of characteristic volatile flavor formation mechanisms in probiotic co-fermented cheese by untargeted metabolomics and sensory predictive modeling. Food Research International. 2025;211 doi: 10.1016/j.foodres.2025.116379. [DOI] [PubMed] [Google Scholar]
  32. Zhao C., Zhang Y.X., Li S.S., Lin J.Y., Lin W.F., Li W.X., Luo L.X. Impacts of aspergillus oryzae 3.042 on the flavor formation pathway in Cantonese soy sauce koji. Food Chemistry. 2024;441 doi: 10.1016/j.foodchem.2024.138396. [DOI] [PubMed] [Google Scholar]
  33. Zhao S., Niu C., Yang X., Xu X., Zheng F., Liu C., Li Q. Roles of sunlight exposure on chemosensory characteristic of broad bean paste by untargeted profiling of volatile flavors and multivariate statistical analysis. Food Chemistry. 2022;381:132115. doi: 10.1016/j.foodchem.2022.132115. [DOI] [PubMed] [Google Scholar]
  34. Zhao S., Niu C.T., Wang Y.H., Zheng F.Y., Liu C.F., Wang J.J., Li Q. The facilitation between Staphylococcus carnosus M43 and Zygosaccharomyces rouxii Y-8, and as starter on the quality of broad bean paste. Food Bioscience. 2023;55 doi: 10.1016/j.fbio.2023.103019. [DOI] [Google Scholar]
  35. Zhou Z.P., Wang X.Y., Duan C.Y., Liu Z.J., Wang Y.F., Zhong Y.J., Wang T. A synergistic fermentation system of probiotics with low-cost and high butyric acid production: Lactiplantibacillus plantarum and Clostridium tyrobutyricum. Food Bioscience. 2024;62 doi: 10.1016/j.fbio.2024.105152. [DOI] [Google Scholar]
  36. Zorea A., Pellow D., Levin L., Pilosof S., Friedman J., Shamir R., Mizrahi I. Plasmids in the human gut reveal neutral dispersal and recombination that is overpowered by inflammatory diseases. Nature. Communications. 2024;15(1) doi: 10.1038/s41467-024-47272-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Zou J.H., Shen H.Y., Li A.P., Wang X.P., Yang H.G., Cheng J.R., Tang D.B. The effect of lactic acid bacteria as a starter on the microbial community and flavors of the fermented beef-soybean paste. Food Chemistry. 2025;484 doi: 10.1016/j.foodchem.2025.144328. [DOI] [PubMed] [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 (3.8MB, docx)

Data Availability Statement

Data will be made available on request.

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

RESOURCES