The initial composition and structure of microbial community determined the yield and quality of Baijiu during the spontaneous fermentation

doi:10.21203/rs.3.rs-2744347/v1

Download PDF

Research Article

The initial composition and structure of microbial community determined the yield and quality of Baijiu during the spontaneous fermentation

https://doi.org/10.21203/rs.3.rs-2744347/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 25 May, 2023

Read the published version in International Microbiology →

You are reading this latest preprint version

The microbiota during pit mud fermentation is one of the most important factors in Baijiu brewing since it determines the yield and quality. However, the effects of the microbial community at the initial stage of Baijiu fermentation remain unknown. Herein, we used high-throughput sequencing to explore the microbial diversity and distribution at the initial and late stages of Baijiu fermentation in individual pit mud workshops. At the initial fermentation stage, the bacterial community had a more significant effect on Baijiu quality than the fungal community. The richness and evenness of the bacterial community in the high-yield pit mud workshop were lower, as well as the Bray–Curtis dissimilarity, during Baijiu fermentation. Lactobacillus was the dominant genus and the biomarker in high-yield pit mud and the only genus composing the association network. For the high Baijiu yield rate, the fungal community tended to maintain a simple association network with a few species as the core microbiota. Rhizopus and Trichosporon were the biomarkers according to the association network during Baijiu fermentation. Finally, Lactobacillus and Rhizopus might be used as bio-indicators for Baijiu quality in the initial fermentation stage. Therefore, we provided new insights into microbiota interactions during fermentation and the effects of the initial microbiota on the final Baijiu quality.

Baijiu spontaneous fermentation

bacterial community

fungal community

association network

biomarkers

The traditional Chinese liquor, Baijiu, is a distillate fermented from grains and has a history of thousands of years [1]. Hundreds of different Baijiu types are produced by various processes in different regions of China [2]. The brewing process of Baijiu is unique and complex [3]. Normally, it uses various grains, such as sorghum, corn, wheat, and rice and daqu, as raw materials. After grinding, the daqu and steamed grains are mixed at a certain proportion, then spontaneously fermented for 20–40 days under sealed solid-state conditions in a pit mud workshop. Finally, the Baijiu is collected by distillation under atmospheric conditions [4, 5] (Fig. 1).

Microbiota structure and diversity are crucial to Baijiu fermentation yield and flavor, providing different aromatic compounds and enzymes. The early stage of fermentation is critical for the yield and flavor of Baijiu because microbiota structure and diversity provide microbial enzymes that accumulate to ensure the formation of flavor precursors, especially via the enrichment of fermentation microorganisms [6]. The bacteria and fungi communities involved in daqu [7] and the pit mud spontaneous fermentation [8] have been extensively studied. However, the effects of the microbiota at the initial stage of fermentation on Baijiu production have not been elucidated. Although the types differed, Lactobacillales and Saccharomycetes were the domain orders in Baijiu fermentation [1, 9]. Previous studies found that, although the bacterial community in the fermentation process comes from daqu, the composition of bacteria in grains before and after fermentation significantly differed from those in daqu and pit mud [10]. However, other studies have shown that the bacterial communities of daqu and fermented grains have similar structures at phylum and genus levels [11]. Fungi community structures changed greatly during Baijiu fermentation [12]. Most studies found that the microbial community tended to be stable after the pit mud fermentation for 3–5 days [1, 12]. Therefore, insights into the changes in microbial structures in the initial and late stages of pit mud fermentation would be beneficial to improve Baijiu quality.

In the present study, we used high-throughput sequencing to investigate the diversity and distribution of microbial communities from individual spontaneous pit mud fermentation. We screened the microbiota and its structure changes in Baijiu's initial and late fermentation stages. Our current results might help us understand the effects of pit mud microbiota on Baijiu fermentation quality.

2.1 Experimental design and sample collection

The experimental site was located at a light flavor Baijiu distillery with the sorghum as main fermentation raw materials in Zhuxianzhuang Town, Anhui Province, China (N: 33°37'; E: 117°5'). We processed two fermentation batches in different pit mud workshops (L and H). Samples were collected in September and October 2021 from the middle and four corners of pit mud workshops. All grains samples were collected in the initial (day 0) and late (day 30) fermentation stages and transported to the laboratory on ice. In the end, 20 samples were treated three times with liquid nitrogen and stored at -80 ℃ as soon as possible for further analysis.

2.2 DNA extraction, amplification, and sequence processing

The DNA from 20 samples was extracted using the E.Z.N.A. ®Stool DNA Kit (D4015, Omega, Inc., USA) following the manufacturer’s instructions. This reagent was designed to uncover trace amounts of DNA from samples and has been shown effective for isolating DNA from most bacteria. Nuclear-free water was used as a blank control. Total DNA was eluted in 50 µL of elution buffer and stored at -80°C until PCR using LC-Bio Technology Co., Ltd (Hang Zhou, Zhejiang Province, China).

The V3 and V4 hypervariable regions were selected for generating amplicons and subsequent taxonomy analysis for 16S rDNA. Regions were amplified using the following primers: 341F (forward primer: 5’-CCTACGGGNGGCWGCAG-3’) and 805R (reverse primer: 5’-GACTACHVGGGTATCTAATCC-3’) [13]. The ITS2 regions were selected for generating amplicons and subsequent taxonomy analysis for ITS rDNA. Regions were amplified using the following primers: ITS1FI2 (forward primer: 5'-GTGARTCATCGAATCTTTG-3') and ITS2 (reverse primer: 5'-TCCTCCGCTTATTGATATGC-3') [14]. The 5' ends of primers were tagged with specific barcodes per sample and sequencing universal primers. PCR amplification was performed in a total volume of 25 µL reaction mixture with 25 ng of template DNA, 12.5 µL PCR Premix, 2.5 µL of each primer, and PCR-grade water to adjust the volume. The amplification conditions for prokaryotic 16S fragments consisted of an initial denaturation at 98 ℃ for 30 s; 32 cycles of denaturation at 98 ℃ for 10 s, annealing at 54 ℃ for 30 s, and extension at 72 ℃ for 45 s; final extension at 72 ℃ for 10 min. PCR products were confirmed with 2% agarose gel electrophoresis. Throughout the DNA extraction process, ultrapure water, instead of a sample solution, was used as a negative control to exclude the possibility of false-positive PCR results. PCR products were purified by AMPure XT beads (Beckman Coulter Genomics, Danvers, MA, USA) and quantified by Qubit (Invitrogen, USA). Amplicon pools were prepared for sequencing, and the size and quantity of the amplicon library were assessed on the Agilent 2100 Bioanalyzer (Agilent, USA) and the Library Quantification Kit for Illumina (Kapa Biosciences, Woburn, MA, USA), respectively. Libraries were sequenced on the NovaSeq PE250 platform.

2.3 Sequence processing and taxonomic affiliation

Usearch10 [15] and Vsearch 2.8.1 [16] data analysis pipelines were used for 16S and ITS rRNA analysis. Forward and reverse reads were joined, assigned to samples based on barcodes, and truncated by removing barcode and primer sequences. Quality filtering was performed on joined sequences. Sequences that did not fulfill the following criteria were discarded: no ambiguous bases and expected errors per base rate > 0.01; dereplicated and singletons sequences (size < 8 nt). Then, sequences were clustered into amplicon sequence variants (ASVs) using the Unoise3 sequence variants algorithm [17, 18]. Chimeric sequences were simultaneously removed, and effective sequences were used in the final analysis. The Ribosomal Database Project (RDP, http://rdp.cme.msu.edu/) and UNITE (https://unite.ut) database were used to assign taxonomy to the aligned sequences with a confidence of 0.8 [19].

2.4 Data analysis and Baijiu yield calculation

All statistical analyses were performed using MicrobiomeAnalyst [20] (https://www.microbiomeanalyst.ca/) and R software (version 4.1.2, https://www.r-project.org/). Alpha diversity was evaluated using the Richness, Simpson, and Faith’s PD diversity indices based on the relative abundance table using the ‘vegan’ R package (v. 2.5-6). Comparisons between different samples were performed using one-way analysis of variance (ANOVA) with Tukey's multiple comparison test. Principal coordinate analysis (PCoA) plots were generated from a Bray-Curtis dissimilarity matrix of ASVs, and Permutational multivariate analysis of variance (PERMANOVA) was performed via the ‘vegan’ R package. The differences in β-diversity were analyzed based on the Bray-Curtis dissimilarity index. For taxonomic analysis, the ‘ggplot2’ R package [21] was used to draw stacking bar charts at the phylum and genus levels to check the relative abundance, and Proteobacteria was classified into four groups based on annotation: Alpha, Beta, Gamma, Delta, and Epsilon. The correlation network analysis was performed to construct and display the Spearman correlation coefficient of indicators using the ‘indicspecies’ R package [22] and Gephi (https://gephi.org). Indicators with p < 0.05 and relative abundance > 0.05% were selected as the ASVs for constructing association networks, with 0.8 as the correlation coefficient cut-off. The linear discriminant analysis effect size (LEfSe) method [23] was used to compare the major microbial composition differences between two pit mud workshops in the initial and late fermentation stages.

The Baijiu yield rate was calculated at a specified ratio of grains (1200 kg) input to liquor output: Baijiu yield (%) = alcohol output/grain input × 100% [24].

3.1 Overall analysis of sequencing data

In total, 1,495,488 high-quality counts were filtered from the bacterial 16S rRNA V3–V4 sequences and 834,952 from the fungal ITS2 sequences. The 16S rRNA and ITS sequences were clustered into 888 and 225 ASVs, respectively. The coverage estimator of bacterial and fungal ASVs was more than 99% in all samples, suggesting that the obtained sequences could represent microbial diversity. Although the counts and ASVs of the two pit mud workshops fluctuated, the Wilcoxon test found no significant difference between the groups (Table 1). The Baijiu yield rate of the L workshop was very low (1.14 ± 0.28%), while the yield of the H workshop (13.70 ± 2.33%) was more than 10 times higher.

Table 1

Microbial sequencing counts, ASVs, and coverage of different *Baijiu* fermentation stages. n: number of samples.
Samples	Counts		ASVs		Coverage		Baijiu yield rate (%)
Samples	Bacteria	Fungi	Bacteria	Fungi	Bacteria	Fungi	Baijiu yield rate (%)
L							1.14 ± 0.28
LS (n = 5)	72633.8 ± 3112.1	43860.4 ± 19756.1	222.6 ± 98.9	44.2 ± 22.5	99.907 ± 0.028	99.96 ± 0.077
LE (n = 5)	74557.2 ± 4026.5	40259.2 ± 15829.9	449.8 ± 59.8	74.4 ± 16.1	99.921 ± 0.034	99.946 ± 0.033
H							13.70 ± 2.33
HS (n = 5)	74967 ± 2890.9	48559.8 ± 7947.9	188 ± 23.1	29.4 ± 9.5	99.929 ± 0.011	99.989 ± 0.004
HE (n = 5)	76939.6 ± 2751.4	34311 ± 10887.1	168.2 ± 13.7	64.2 ± 8.0	99.931 ± 0.004	99.934 ± 0.024

3.2 Microbial diversity in different pit mud workshops at different Baijiu fermentation stages

The Richness, Simpson, and Faith's PD indices were used to determine the differences in microbial community composition between different pit mud workshops at the initial and late fermentation stages. The bacterial Richness, Simpson, and Faith's PD indices of the L group showed an increasing trend after fermentation, while Richness and Faith’s PD indices significantly increased. The bacterial alpha diversity of the H group was stable during Baijiu fermentation. Regarding the fungal alpha diversity, Richness and Faith's PD indices of all pit mud workshops significantly increased at the late fermentation stage (Fig. 2A). There were no significant differences in the Simpson index before or after Baijiu fermentation in the two pit mud workshops (Fig. 2B).

The PCoA of the Bray–Curtis distance matrix was used to assess the microbial communities during Baijiu fermentation (Fig. 3). The horizontal (PC1 axis) and vertical (PC2 axis) coordinates were the PCs contributing to microbial community differences among all samples. Collectively, these PCs contributed to 88% of the variation in bacterial communities and 68% in fungal communities for all samples (Fig. 3). This result indicated that the different Baijiu fermentation stage was the major influencing factor (bacterial p = 0.001 and fungal p = 0.046 in the PERMANOVA by Adonis).

The bacterial communities of the L workshop were significantly different before and after Baijiu fermentation, while they were similar in the H workshop. Moreover, the LE group showed evident heterogeneity. The Bray-Curtis dissimilarity significantly increased at the end of fermentation in the L workshop (LE). In contrast, in the H workshop, the Bray-Curtis dissimilarity (HE) significantly decreased (Kruskal-Wallis test, p = 1.07e^− 06) compared to before fermentation (Fig. 3A). In different workshops, the fungi communities did not significantly fluctuate before and after Baijiu fermentation. Significant clustering of fungi communities in the PC2 axis was found in the H workshop. Although not significant, the fungal dissimilarity also showed an increasing trend in the L workshop and a decreasing trend in the H workshop at the initial and late Baijiu fermentation stages (Fig. 3B).

3.3 Changes in microbial community composition for different pit mud workshops and Baijiu fermentation stages

Based on the RDP database, the bacterial and fungal ASVs were classified and clustered at the phylum and genus levels. The relative abundance of the top 10 phyla and genera is represented by bar plots (Fig. 4). In different pit mud workshops, the composition and relative abundance of taxa varied considerably before and after Baijiu fermentation. In the initial fermentation stage, Gammaproteobacteria (70.8 ± 6.9%) was the dominant bacterial phyla in the L workshop, while Firmicutes (91.7 ± 2.9%) presented the highest relative abundance in the H workshop. Although Firmicutes were the main phylum in the two workshops at the Baijiu fermentation late, many Gammaproteobacteria and Actinobacteria were also observed in the L workshop (Fig. 4A, Table S1). Regarding the fungal community, the main phylum was Ascomycota (LS: 97.2 ± 0.5%, HS: 86.6 ± 9.6%, LE: 96.6 ± 2.7%, HE: 99.0 ± 0.8%) in all groups at different fermentation stages. The HS group also presented a high relative Mucoromycota abundance (13.0 ± 9.5%) (Fig. 4B, Table S1).

At the genus level, abundant Stenotrophomonas (69.4 ± 8.4%) was found in the LS group, as well as Arthrobacter (13.5 ± 1.6%) and Bacillus (7.0 ± 0.8%). In contrast, Lactobacillus (91.0 ± 3.2%) accounted for the majority, and Stenotrophomonas (6.6 ± 2.4%) accounted for a small amount in the HS group. In the late fermentation stage, besides a high relative abundance of Lactobacillus (53.9 ± 24.4%), Streptomyces, Bordetella, and Olivibacter appeared as low-abundance taxa in the LE group. The HE group presented a high Baijiu yield rate, and Lactobacillus (99.2 ± 0.2%) was the dominant genus (Fig. 4A, Table S1). Regarding the fungal community, Saccharomycetes_unassigned was the main taxa (LS: 67.7 ± 10.5%, HS: 49.3 ± 18.9%, LE: 71.6 ± 31.6%, HE: 84.6 ± 9.1%) in all groups at different fermentation stages. The relative abundance of Aspergillus in the L workshop (low Baijiu yield rate) was significantly higher than that in the H workshop at the initial and late Baijiu fermentation stages (Fig. 4B, Table S1).

3.4 Association networks of microbial indicators in different pit mud workshops and Baijiu fermentation stages

The ASVs-based indicators correlation networks of different groups were constructed to examine microbiota community relationships. ASVs with relative abundance > 0.05% and correlation coefficient > 0.8 are presented in Fig. 5. The red-to-yellow color transition represents the correlation degree. The node size represents the average closeness centrality of ASVs, and the same node color represents the same phylum or class-level species. The network properties analyzed included average path length, average clustering coefficient, and modularity (Table 2).

In the L workshop, late fermentation presented a higher clustering coefficient than initial fermentation (bacteria: 0.167 vs. 0.371, fungi: 0 vs. 0.425), indicating that more bacterial and fungal communities clustered at the late fermentation stage. Meanwhile, the higher average path length (bacteria: 2.091 vs. 3.28, fungi: 1.429 vs.1.905) indicated the less compact local interconnection and incompact bacterial communities. The network's modularity also decreased in the late fermentation stage (bacteria: 0.624 vs. 0.53, fungi: 0.375 vs. 0.302). In the H workshop, the bacterial community composed a simple network of a single Lactobacillus at the late fermentation stage with an average clustering coefficient of 0. Meanwhile, the network modularity increased (0.031 vs. 0.32), and the average path length reduced (1.528 vs. 1.286). The association network of fungal communities at the late Baijiu fermentation stage showed complexity and diversity dominated by Saccharomycetes, including Cyberlindnera, Zygosaccharomyces, Kazachstania, and Candida. The average clustering coefficient (0 vs. 0.257) and average path length (1.528 vs. 1.286) increased while modularity decreased (0.612 vs. 0.221) (Fig. 5, Table 2).

Table 2

Microbial correlation network properties during *Baijiu* fermentation.
Correlation network properties	Bacteria				Fungi
Correlation network properties	LS	LE	HS	HE	LS	LE	HS	HE
Nodes	21	111	9	6	6	24	10	17
Edges	24	762	18	5	4	63	7	42
Average clustering coefficient	0.167	0.371	0.548	0	0	0.425	0	0.257
Average path length	2.091	3.28	1.528	1.286	1.429	1.905	1.538	2.257
Modularity	0.624	0.53	0.031	0.32	0.375	0.302	0.612	0.221

Overall, the microbial association networks of the H workshop were simpler, and the value of network properties was smaller than the L workshop. At the initial Baijiu fermentation stage, the bacterial network in the L workshop showed high complexity compared to the H workshop, including more nodes and edges, and higher average path length and modularity, while the fungal network was the opposite. At the late Baijiu fermentation stage, the microbial community in the L workshop formed a complex network with multiple species, while it was a simple network with fewer species in the H workshop (Fig. 5, Table 2).

3.5 LEfSe of microbial communities during Baijiu fermentation

The LEfSe is efficient for high-efficiency biomarkers discovery and enables the detailed identification of abundant features by characterizing potential discriminating taxa between different biological groups [25]. Herein, the LEfSe was used to reveal microbial community differences between the two workshops during fermentation (Fig. 6). Except for Streptomyces and Rummeliibacillus, the bacterial genus community changed greatly at the initial and late fermentation in the L workshop. Meanwhile, Lactobacillus was the only biomarker in the H workshop (high Baijiu yield rate) (Fig. 6A). In the initial fermentation, the L workshop did not present important fungal biomarkers, and many unassigned species occupied the main biomarkers. In contrast, Rhizopus in the early stage and Rhizopus and Trichosporon in the late stage of fermentation were the main biomarkers in the H workshop (Fig. 6B).

The fermentation microbiota plays an essential role in Baijiu production via some interaction and regulatory mechanisms [8]. As important microbial components for Baijiu brewing, the bacterial community can regulate production and improve quality [10]. Moreover, the fungal community is related to flavor formation during Baijiu fermentation [26]. Hence, understanding the diversity, structure, and dynamics of microbial communities is important for improving Baijiu quality and yield [1]. Herein, we used high-throughput sequencing to reveal the diversity, structure, and association of microbial communities and the effects of microbial succession on Baijiu yield.

4.1 Effects of microbial diversity on Baijiu fermentation and yield rate in different pit mud workshops

In the present study, we used the Richness, Simpson, and Faith's PD indices to assess the richness and evenness of microbiota. The bacterial alpha diversity, especially Richness and Faith’s PD indices of the L workshop (low yield rate), significantly increased at the late fermentation stage, contrary to the decreasing trend of the alpha diversity index of the H workshop (high yield rate). Our study also indicated that Simpson's index of the L workshop was significantly higher than the H workshop at the initial fermentation stage (Fig. 2). Although the raw materials and fermentation process of Baijiu can vary, we found that fermentation was mainly represented by bacterial diversity reduction [1, 9]. Examining the bacterial alpha diversity of different workshops made it possible to judge the Baijiu yield rate. Interestingly, we found that the trends in bacterial alpha diversity of different pit mud did not affect fungal alpha diversity.

Our findings also showed that the beta diversity distance of the L workshop significantly changed with the initial fermentation stage, and the distance was highly heterogeneous. At the same time, the Bray-Curtis dissimilarity showed that the changes of the bacterial structure significantly increased in the L pit mud (Fig. 3A). In contrast, the bacterial structure of the H workshop (high yield rate) changed little before and after fermentation, while the dissimilarity index significantly decreased (Fig. 3B). The structure of the microbial community in pit mud is related to Baijiu quality and yield. For example, bacteria's co-occurrence patterns determined the fermentation process's quality and stability [27]. Here, we showed that the high bacterial structural heterogeneity and dissimilarity were the reasons for the Baijiu yield decrease. In contrast, the beta diversity of fungi in different Baijiu fermentation workshops did not significantly change.

4.2 Effects of microbial community composition on Baijiu fermentation and yield rate in different pit mud workshops

Furthermore, we showed that, although Firmicutes and Proteobacteria were the main bacterial phyla as in previous studies [9, 28], the relative abundance of bacteria at the initial fermentation stage significantly affected the post-fermentation Baijiu yield. A high production rate was observed in the H workshop with Firmicutes as the dominant phylum at the beginning of Baijiu fermentation, while the L workshop had Gamma proteobacteria as the main phylum and a low yield. Besides, the low-yield workshop had more phyla with high relative abundance, such as Actinobacteria, Bacteroidetes, Betaproteobacteria, and Alphaproteobacteria (Fig. 4A). During Baijiu fermentation, Ascomycota was the dominant phylum in both workshops with the relative abundance majority, consistent with previous studies [29, 30]. Additionally, high Mucoromycota abundance was also observed in the H workshop during the initial fermentation stage (Fig. 4B). Succession of Mucoromycota has been reported to have a high relative abundance in medium-temperature daqu [31] and was correlated with daqu quality [32]. Also, Mucoromycota had a strong protease activity, allowing the digestion of proteins in dregs into amino acids and further reaction with reducing sugars to form various aroma substances [1].

Based on statistical analysis at the genus level, Lactobacillus and Stenotrophomonas were the dominant bacterial groups, and Saccharomycetes_unassigned and Aspergillus were the dominant fungal groups in the high Baijiu yield workshop at the initial fermentation stage (Fig. 4). These genera were also detected in daqu and other liquor fermentations [7, 31, 33, 34], indicating that the main microbiota was formed in the initial Baijiu fermentation stage or daqu culture. The lactic acid from Lactobacillus can maintain the acidic brewing environment and balance the microbial brewing system in liquor production [35]. Saccharomycetes produce various lytic enzymes for synthesizing substrates for Baijiu fermentation and further forming flavor compounds [1, 36]. Meanwhile, the pit mud with Arthrobacter and Bacillus as the main bacterial groups at the initial fermentation stage had low Baijiu yield. Arthrobacter is a strictly aerobic bacterium [37], and Bacillus has been isolated from daqu, which could provide flavor substances for liquor [7, 36]. However, the community composition in Baijiu fermentation is low [30]. Our current results indicated that the disturbance of bacterial community composition in the initial stage might not form the functional microbiota structure in the pit mud environment, resulting in liquor fermentation failure. We also showed that the composition of bacterial genera in the initial stage could predict the quality and yield of Baijiu fermentation.

4.3 Effects of different indicators association network on fermentation quality and Baijiu yield

The relationships of microbial diversity and correlation-based network complexity were assessed at the initial and late stages of Baijiu fermentation in pit mud workshops with different yields. First, bacterial association networks for the high Baijiu yield workshop lost their complexity, whereas those with low yields gained complexity. During Baijiu fermentation, the nodes (9 to 6) and edges (18 to 5) in the bacterial association network of the high-yield workshop decreased, as well as the average clustering coefficient (0.548 to 0) and average path length (1.528 to 1.286), the opposite network properties of the low-yield workshop (Fig. 5A). The complexity of the correlation network could predict the function [38, 39]. Different processing might also significantly disturb the stability of microbial community networks [40]. Our current results indicated that the network function of single Lactobacillus formed in the late fermentation stage was more specific, which might interact with the microenvironment of the workshop by producing lactic acid through anaerobic fermentation to improve the Baijiu yield rate [35].

Second, contrary to bacterial association networks, fungal communities formed more complex networks during Baijiu fermentation, but the fungal microbiota with a high Baijiu yield rate still tended to maintain a simple network with a few species in the core. In the initial fermentation stage, the fungal community network of the high-yield workshop had higher complexity. In contrast, fewer species at the class level, fewer nodes and edges, and lower average clustering coefficient and modularity were observed in the late fermentation stage (Fig. 5B). Saccharomycetes, such as Zygosaccharomyces and Cyberlindnera, can generate ethyl acetate and 2-phenyl ethanol during fermentation, which play an important role in the liquor style and aroma characteristics of Baijiu [41]. Mucoromycota, such as Rhizopus, and Eurotiomycetes, such as Aspergillus, can produce diastase and organic acids, promoting daqu fermentation [31].

Previous ecological studies have suggested that microbial network complexity can reflect the ecosystem's multifunctionality [42, 43]. The succession rate of microbial communities during Baijiu fermentation can affect the flavor profile by obtaining or losing microbial network complexity [44]. Our current findings contributed to a better understanding of the influence of bacterial and fungal community network complexity on Baijiu fermentation quality and yield.

4.4 Biomarkers during Baijiu fermentation in different pit mud workshops

Moreover, we used LEfSe to detect microbial biomarkers between the two pit mud workshops during the Baijiu fermentation. The bacterial biomarker in the high-yield workshop was Lactobacillus during the whole Baijiu fermentation, with relative abundance significantly higher than that in the low-yield workshop, and as the only species in the late Baijiu fermentation (Fig. 4, 5, and 6, and Table S1). Lactobacillus is indispensable in regulating other microorganisms and Baijiu quality [35], which might be used as bio-indicators to evaluate Baijiu quality at the initial fermentation stage. In contrast, bacterial biomarkers significantly differed in the initial and late fermentation stages in the low-yield workshop. Stenotrophomonas, Arthrobacter, and Bacillus were biomarkers whose relative abundance was significantly higher in the initial fermentation stage. Although they are common bacterial taxa in daqu and pit mud [7, 45], complex initial taxa significantly decreased the Baijiu yield in our study. These results suggested that the presence or high relative abundance of Stenotrophomonas, Arthrobacter, and Bacillus might be related to the decrease of Baijiu quality and yield at the initial fermentation stage. At the late fermentation stage, Streptomyces, Bordetella, Rummeliibacillus, Olivibacter, and other genera were the biomarkers. Although Streptomyces have been reported in other studies [46], most genera had low relative abundance and were not found in other Baijiu fermentation studies.

Regarding the fungal community, Rhizopus and Trichosporon were biomarkers in the high-yield pit mud workshop in the initial and late Baijiu fermentation stages. Surprisingly, they were neither the most abundant in the fungal community nor the highest modularity class in the association network. However, Rhizopus and Trichosporon were at the center of the network during Baijiu fermentation and played a vital role in connecting the network (Fig. 5). Their relative abundance is positively correlated with multiple fragrances, including tetradecane, valeric acid, oleic acid, ethyl butyrate, propyl acetate, and methylhexanoic acid [31]. Although the fungal community had a complex correlation network during Baijiu fermentation, no biomarkers were found in the low-yield pit mud workshop. Generally, the Rhizopus enrichment at the initial fermentation stage might be used as a fungal bioindicator for Baijiu quality.

In the present study, we evaluated the microbial diversity, structure, correlation, and succession in high and low-yield pit mud samples at the initial and late Baijiu fermentation stages. We showed that the microecosystem comprised a microbial community that significantly affected Baijiu yield and quality. The high-yield pit mud had lower alpha diversity, a more stable microbiota structure, and a simpler correlation network, which were significant in bacterial community properties. Additionally, the association network analysis showed that the interspecies interactions within microbial communities were markedly changed and reflected the ecosystem functionality, such as alcohol production during Bajiu fermentation. Furthermore, as the biomarkers in high-yield pit mud, Lactobacillus and Rhizopus played an indispensable role in regulating other microorganisms and Baijiu quality and might be used as indicator species to estimate the fermentation quality at the initial stage. These findings represented scientific evidence to support the practical experience that the initial structure of the microbial community determines yield and quality during spontaneous Baijiu fermentation.

Competing interests

No conflict of interest exits in the submission of this manuscript, and manuscript is approved by all authors for publication. I would like to declare on behalf of my co-authors that the work described was original research that has not been published previously, and not under consideration for publication elsewhere, in whole or in part. All the authors listed have approved the manuscript that is enclosed.

Authors' contributions

Dongjuan Li and Fan Chang wrote the main manuscript text and Fengan Jia and Lingling Wang prepared fiqures. All authors reviewed the manuscript.

Funding

This study was supported by the Anhui Key Project of Scientific Research in Universities (KJ2020A1146).

Liu C, Gong X, Zhao G, Soe Htet MN, Jia Z, Yan Z, et al. 2021. Liquor Flavour Is Associated With the Physicochemical Property and Microbial Diversity of Fermented Grains in Waxy and Non-waxy Sorghum (Sorghum bicolor) During Fermentation. Front. Microbiol. 12: 618458.
Zheng X-W, Han B-Z. 2016. Baijiu (白酒), Chinese liquor: History, classification and manufacture. Journal of Ethnic Foods 3: 19–25.
Wang YC. 2014. The 2nd Chinese Liquor Technology Symposium. China Light Industry Press, Beijing.
Shen, Y. 2014. Encyclopedia of Production and Technology of Chinese Baijiu. China Light Industry Press, Beijing.
Zhang, S. Y, Xu, D. F. 2011. Skill of Luzhou-Flavour Liquor. China Light Industry Press, beijing.
Ding X, Wu C, Huang J, Zhou R. 2015. Changes in Volatile Compounds of Chinese Luzhou -Flavor Liquor during the Fermentation and Distillation Process: Volatile compounds in liquor…. Journal of Food Science 80: C2373–C2381.
Jiang Q, Wu X, Xu Y, Zhang Y, Wang Z, Shen L, et al. 2021. Microbial composition and dynamic succession during the Daqu production process of Northern Jiang-flavored liquor in China. 3 Biotech 11: 224.
Xu J, Wu H, Wang Z, Zheng F, Lu X, Li Z, et al. 2018. Microbial dynamics and metabolite changes in Chinese Rice Wine fermentation from sorghum with different tannin content. Sci Rep 8: 4639.
Wang H, Huang Y, Huang Y. 2021. Microbiome diversity and evolution in stacking fermentation during different rounds of Jiang-flavoured Baijiu brewing. LWT 143: 111119.
Wang X, Du H, Xu Y. 2017. Source tracking of prokaryotic communities in fermented grain of Chinese strong-flavor liquor. International Journal of Food Microbiology 244: 27–35.
Fang RS, Dong YC, Chen F, Chen QH. 2015. Bacterial Diversity Analysis during the Fermentation Processing of Traditional Chinese Yellow Rice Wine Revealed by 16S rDNA 454 Pyrosequencing. Journal of Food Science 40: 152–159.
Xiao C, Lu Z, Zhang X, Wang S, Xu Z. 2018. Succession of the fungal community on fermented grains of Luzhou-flavor baijiu through fermentation. Chinese Journal of Applied and Environmental Biology 24: 1081–1086.
Takahashi S, Tomita J, Nishioka K, Hisada T, Nishijima M. 2014. Development of a Prokaryotic Universal Primer for Simultaneous Analysis of Bacteria and Archaea Using Next-Generation Sequencing. PLoS ONE 9: e105592.
Karlsson I, Friberg H, Steinberg C, Persson P. 2014. Fungicide Effects on Fungal Community Composition in the Wheat Phyllosphere. PLoS ONE 9: e111786.
Edgar RC. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461.
Rognes T, Flouri T, Nichols B, Quince C, Mahé F. 2016. VSEARCH: a versatile open source tool for metagenomics. PeerJ 4: e2584.
Edgar RC. 2018. Updating the 97% identity threshold for 16S ribosomal RNA OTUs. Bioinformatics 34: 2371–2375.
Knight R, Vrbanac A, Taylor BC, Aksenov A, Callewaert C, Debelius J, et al. 2018. Best practices for analysing microbiomes. Nat Rev Microbiol 16: 410–422.
Bacci G, Bani A, Bazzicalupo M, Ceccherini MT, Galardini M, Nannipieri P, et al. 2015. Evaluation of the Performances of Ribosomal Database Project (RDP) Classifier for Taxonomic Assignment of 16S rRNA Metabarcoding Sequences Generated from Illumina-Solexa NGS. J. Genomics 3: 36–39.
Dhariwal A, Chong J, Habib S, King IL, Agellon LB, Xia J. 2017. MicrobiomeAnalyst: a web-based tool for comprehensive statistical, visual and meta-analysis of microbiome data. Nucleic Acids Research 45: W180–W188.
Wickham H. 2016. ggplot2. Springer International Publishing, Cham.
Cáceres MD, Legendre P. 2009. Associations between species and groups of sites: indices and statistical inference. Ecology 90: 3566–3574.
Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett WS, et al. 2011. Metagenomic biomarker discovery and explanation. Genome Biol 12: R60.
Zhou H, Zhang S, Zonghua AO, Song C, Zhanyuan SU, Liu H. 2015. Research Progress in Liquor-receiving Technology of Baijiu(Liquor) Production. Liquor-Making Science & Technology 03: 105–107.
Afgan E, Baker D, Batut B, van den Beek M, Bouvier D, Čech M, et al. 2018. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Research 46: W537–W544.
Li Y, Cheng Y, Wang H, Hu X, Wang L, Huang Y. 2022. Diverse structure and characteristics of the fungal community during the different rounds of Jiang-flavoured Baijiu production in Moutai town. LWT 161: 113313.
Zheng Y, Hu X, Jia Z, Bodelier PLE, Guo Z, Zhang Y, et al. 2020. Co-occurrence patterns among prokaryotes across an age gradient in pit mud of Chinese strong-flavor liquor. Can. J. Microbiol. 66: 495–504.
Song Z, Du H, Zhang Y, Xu Y. 2017. Unraveling Core Functional Microbiota in Traditional Solid-State Fermentation by High-Throughput Amplicons and Metatranscriptomics Sequencing. Front. Microbiol. 8: 1294.
Gan S-H, Yang F, Sahu SK, Luo R-Y, Liao S-L, Wang H-Y, et al. 2019. Deciphering the Composition and Functional Profile of the Microbial Communities in Chinese Moutai Liquor Starters. Front. Microbiol. 10: 1540.
Gong Y, Ma N, Tang H. 2022. Analysis of microbial community diversity and physicochemical factors in pit mud of different ages based on high-throughput sequencing. Can. J. Microbiol. cjm-2022-0125.
Liu Y, Li X, Li H, Zhang H, Shen X, Zhang L, et al. 2022. Taorong-type Baijiu starter: Analysis of fungal community and metabolic characteristics of middle-temperature Daqu and high-temperature Daqu. PLoS ONE 17: e0274881.
Ma S, Luo H, Zhao D, Qiao Z, Zheng J, An M, et al. 2022. Environmental factors and interactions among microorganisms drive microbial community succession during fermentation of Nongxiangxing daqu. Bioresource Technology 345: 126549.
Zuo Q, Huang Y, MinGuo. 2020. Evaluation of bacterial diversity during fermentation process: a comparison between handmade and machine-made high-temperature Daqu of Maotai-flavor liquor. Ann Microbiol 70: 57.
Wang X, Du H, Zhang Y, Xu Y. 2018. Environmental Microbiota Drives Microbial Succession and Metabolic Profiles during Chinese Liquor Fermentation. Appl Environ Microbiol 84.
Yang F, Zhang Q, Liu Y, Li J, Wang L, Chen J. 2020. Lactic acid biosynthesis pathways and important genes of Lactobacillus panis L7 isolated from the Chinese liquor brewing microbiome. Food Bioscience 36: 100627.
Zhang R, Wu Q, Xu Y. 2013. Aroma characteristics of Moutai-flavour liquor produced with Bacillus licheniformis by solid-state fermentation. Lett Appl Microbiol 57: 11–18.
Daengbussadee C, Laopaiboon L, Kaewmaneewat A, Sirisantimethakom L, Laopaiboon P. 2021. Novel Methods Using an Arthrobacter sp. to Create Anaerobic Conditions for Biobutanol Production from Sweet Sorghum Juice by Clostridium beijerinckii. Processes 9: 178.
Chen W, Wang J, Chen X, Meng Z, Xu R, Duoji D, et al. 2022. Soil microbial network complexity predicts ecosystem function along elevation gradients on the Tibetan Plateau. Soil Biology and Biochemistry 172: 108766.
Landi P, Minoarivelo HO, Brännström Å, Hui C, Dieckmann U. 2018. Complexity and stability of ecological networks: a review of the theory. Population Ecology 60: 319–345.
Ye G, Fan J, Hu H-W, Chen J, Zhong X, Chen J, et al. 2022. Short-term cellulose addition decreases microbial diversity and network complexity in an Ultisol following 32-year fertilization. Agriculture, Ecosystems & Environment 325: 107744.
Zha M, Sun B, Wu Y, Yin S, Wang C. 2018. Improving flavor metabolism of Saccharomyces cerevisiae by mixed culture with Wickerhamomyces anomalus for Chinese Baijiu making. Journal of Bioscience and Bioengineering 126: 189–195.
Wagg C, Schlaeppi K, Banerjee S, Kuramae EE, van der Heijden MGA. 2019. Fungal-bacterial diversity and microbiome complexity predict ecosystem functioning. Nat Commun 10: 4841.
Qiu L, Zhang Q, Zhu H, Reich PB, Banerjee S, van der Heijden MGA, et al. 2021. Erosion reduces soil microbial diversity, network complexity and multifunctionality. ISME J 15: 2474–2489.
Tan Y, Zhong H, Zhao D, Du H, Xu Y. 2019. Succession rate of microbial community causes flavor difference in strong-aroma Baijiu making process. International Journal of Food Microbiology 311: 108350.
He G, Huang J, Wu C, Jin Y, Zhou R. 2020. Bioturbation effect of fortified Daqu on microbial community and flavor metabolite in Chinese strong-flavor liquor brewing microecosystem. Food Research International 129: 108851.
Zheng X-W, Yan Z, Robert Nout MJ, Boekhout T, Han B-Z, Zwietering MH, et al. 2015. Characterization of the microbial community in different types of Daqu samples as revealed by 16S rRNA and 26S rRNA gene clone libraries. World J Microbiol Biotechnol 31: 199–208.

No competing interests reported.

Download PDF

Journal Publication

published 25 May, 2023

Read the published version in International Microbiology →

Editorial decision: Major revision
05 Apr, 2023
Editor assigned by journal
31 Mar, 2023
Submission checks completed at journal
31 Mar, 2023
First submitted to journal
27 Mar, 2023

You are reading this latest preprint version

The initial composition and structure of microbial community determined the yield and quality of Baijiu during the spontaneous fermentation

Status:

Journal Publication

Version 1

Abstract

Figures

1 Introduction

2 Materials And Methods

2.1 Experimental design and sample collection

2.2 DNA extraction, amplification, and sequence processing

2.3 Sequence processing and taxonomic affiliation

2.4 Data analysis and Baijiu yield calculation

3 Results

3.1 Overall analysis of sequencing data

3.2 Microbial diversity in different pit mud workshops at different Baijiu fermentation stages

3.3 Changes in microbial community composition for different pit mud workshops and Baijiu fermentation stages

3.4 Association networks of microbial indicators in different pit mud workshops and Baijiu fermentation stages

3.5 LEfSe of microbial communities during Baijiu fermentation

4. Discussion

4.1 Effects of microbial diversity on Baijiu fermentation and yield rate in different pit mud workshops

4.3 Effects of different indicators association network on fermentation quality and Baijiu yield

4.4 Biomarkers during Baijiu fermentation in different pit mud workshops

5. Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1