Microbial Community Succession and Flavor Compound Formation in Sesame-Flavored Baijiu from Zaopei

Abstract

The succession of microbial communities during the fermentation process in sesame-flavored Baijiu cellars profoundly influences the flavor profile of the liquor. However, the key factors driving microbial succession in these cellars remain unclear. This study focuses on the fermentation process of sesame-flavored Baijiu Zaopei in traditional Tongcheng cellars. Samples were collected from the surface, middle, and bottom of the cellar, categorized by fermentation time. Various techniques were employed to analyze the physicochemical properties (including moisture, ethanol, total acid, starch, and reducing sugars), flavor compounds (volatile substances and amino acids), and microbial communities (bacteria and fungi) of the Zaopei during fermentation. A total of 68 flavor compounds were detected, with 16 key flavor compounds and 16 amino acids identified. Microbiologically, the Lactobacillus genus dominated in the later stages of fermentation, while the Issatchenkia species were the predominant fungi. Correlation analysis indicated that environmental factors play a significant role in driving microbial community succession. Acetobacter, Staphylococcus, Pichia, Bacillus, and Kroppenstedtia species may contribute to the synthesis of key flavor compounds. The relative contents of acetic acid, 2-phenylethyl ester, and Benzenepropanoic acid ethyl ester were influenced by multiple microbial groups, suggesting a synergistic fermentation effect. PICRUSt2 predictions revealed significant differences in 41 KEGG pathways at level 2 and 293 pathways at level 3 across different fermentation intervals. These pathways are primarily associated with amino acid, ester, and nucleotide metabolism, as well as bacterial transcription, translation, and signal transduction. This research provides a theoretical foundation for understanding the fermentation mechanisms of sesame-flavored Baijiu.

1. Introduction

Chinese Baijiu is renowned for its profound cultural heritage, unique aroma, and vast consumer base. Microbial fermentation serves as the primary source of its distinctive flavors [1]. Based on aromatic characteristics, Baijiu is categorized into 12 flavor profiles, including strong, light, sauce, rice, mixed, feng, te, dong, chi, sesame, fuyu, and laobaigan [2]. Sesame-flavored Baijiu is one of these varieties, celebrated for its unique sesame aroma. Its production process comprises eight key steps: high-temperature daqu making, high-nitrogen raw material preparation, high-temperature stacking, high-temperature fermentation, multi-microbe co-fermentation, multi-base liquor blending, layered distillation, and long-term storage [3]. The production of sesame-flavored Baijiu is mainly concentrated in the northern regions [4]. This type of Baijiu is renowned for its unique roasted sesame aroma, with sulfur compounds (such as 2-furfurylthiol) being its core flavor components. It holds cultural heritage value in traditional brewing processes, while also driving industry upgrades through technological innovation [5]. Although sesame-flavored Baijiu is often regarded as a high-end or specialty product due to its unique flavor, a multi-tiered market structure has emerged, ranging from boutique to mass consumption, through graded production and market segmentation [6,7,8]. Zaopei, a seasoning compound, is directly produced during Baijiu fermentation and forms within the cellar. Typically, fresh Zaopei is a mixture of grains, rice husks, and Daqu powder [9]. During Baijiu fermentation, Zaopei plays a crucial role in environmental circulation and information transmission. It effectively promotes material cycling and energy flow, while also efficiently conveying information, thereby fulfilling a pivotal function [10]. Any alteration in factors during Zaopei fermentation can trigger significant chain reactions, ultimately altering the flavor characteristics of Baijiu [11]. Studies have shown that temperature, acidity, moisture, and reducing sugars all influence the microbial composition during fermentation [12].

Research on the flavor variations in sesame-flavored Baijiu has been conducted. Zheng et al. demonstrated that methionine and Ethyl caproate significantly contribute to the aroma of sesame-flavored Baijiu using liquid–liquid extraction combined with GC-MS and GC-O methods [13]. Li et al. employed comprehensive two-dimensional gas chromatography/time-of-flight mass spectrometry (GC × GC-TOF-MS) and identified 509 volatile components, among which Hexanoic acid ethyl ester, Octanoic acid ethyl ester, and dimethyl trisulfide were the most abundant, making the largest contribution to the aroma of sesame-flavored Baijiu [14]. Wu et al. found that, although sulfur-containing compounds are rare, they play a key role in the roasted aroma of sesame flavor, with benzyl mercaptan making a significant contribution to the overall aroma of sesame-flavored Baijiu [15]. Recent studies indicate that brewing materials significantly impact the production of flavor compounds during Baijiu fermentation [16]. The production process of sesame-flavored Baijiu differs from that of traditional strong-flavored Baijiu in that it involves a stacking fermentation process. By controlling temperature and humidity, it promotes the growth of beneficial microorganisms while inhibiting contaminants, thereby selecting a microbial community that is conducive to aroma production and fermentation, ensuring the formation of its unique flavor and quality [17]. Currently, the research on the mechanisms of flavor compound formation is not yet mature. The biosynthetic pathways of key flavor compounds contributing to sesame aroma, particularly sulfur-containing compounds, have not been fully elucidated, especially the lack of quantitative models for the interaction between microbial metabolism and the Maillard reaction. Furthermore, there is insufficient research on the thresholds and synergistic effects of aroma-active substances [4], for instance, the “enhancement-antagonism” relationship between dimethyl trisulfide and other esters still requires experimental validation [18]. During alcoholic fermentation, the succession of microorganisms in the Zaopei plays a crucial role in the synthesis of key flavor compounds and esters in Baijiu. Therefore, studying the changes in microorganisms, physicochemical factors, and flavor metabolites in Baijiu Zaopei during fermentation is of significant importance for revealing the brewing patterns of sesame-flavored Baijiu.

Due to limited understanding of the succession of microorganisms and the flavor compounds during the fermentation process in the production of Baijiu, it has been challenging to find reasonable solutions to the issues of high acidity and insufficient characteristic flavors that arise during production. This study utilized headspace solid-phase microextraction-gas chromatography-mass spectrometry (HS-SPME-GC-MS) and an amino acid analyzer to investigate the dynamic changes in aroma compounds (volatile substances) and taste compounds (amino acids) during Zaopei fermentation. Combining high-throughput sequencing technology and OTU clustering analysis, the composition and succession of bacterial and fungal communities during Zaopei fermentation were studied. Spearman correlation analysis was employed to examine the relationships between key microorganisms, physicochemical factors, and flavor metabolites during Zaopei fermentation. PICRUSt2 was used to predict the functions of the Zaopei bacterial community, providing a theoretical basis for the production of sesame-flavored Baijiu.

2. Materials and Methods

2.1. Sample Collection

In this experiment, samples were collected from the fermentation pits of old Tongcheng liquor (Tongcheng, China ) that had been in continuous use for years. Each fermentation pit can be subdivided into the surface layer, middle layer, and bottom layer. A total of 15 sampling points were collected from the four corners of the tank as well as the centers of the surface, middle, and bottom layers, and these samples were thoroughly mixed. All samples were placed in six sterile bags and labeled according to the fermentation days: 0 days, 5 days, 10 days, 20 days, 30 days, 35 days, and 40 days, labeled as zp_0, zp_5, zp_10, zp_20, zp_30, zp_35, and zp_40, respectively. Three parallel samples were collected for each group, and after collection, they were stored in a freezer at −80 °C (for microbial sequencing) and −4 °C (for physicochemical testing) [19].

2.2. Analysis of Flavor Substances

In this study, we employed headspace solid-phase microextraction coupled with gas chromatography-mass spectrometry (HS-SPME-GC-MS, CA, USA) to analyze the flavor compounds in the Zaopei samples. A 0.5 g sample was placed into a 20 mL headspace vial, to which 8 mL of saturated NaCl solution (Guangzhou, China) and 1 μL of 2% Pentyl acetate (internal standard) solution (Guangzhou, China) were added. After magnetic stirring at 50 °C for 10 min, the extraction was performed. The chromatographic separation of metabolites was achieved via a DB-5 MS capillary column (60 m × 1 mm × 0.32 μm). The gas chromatography conditions were modified from the original method: the injection port temperature was set at 250 °C, with a carrier gas flow rate of 1.0 mL/min. The temperature gradient was as follows: initial temperature at 40 °C, held for 2 min, then increased at 3 °C/min to 64 °C, followed by an increase at 5 °C/min to 180 °C, and finally at 10 °C/min to 250 °C, held for 5 min. Helium was used as the carrier gas with a split ratio of 10:1. The mass spectrometry conditions were as follows: junction temperature at 250 °C, ionization voltage at 70 eV, ion source temperature at 250 °C, quadrupole temperature at 150 °C, and a mass scan range of m/z 30–500 amu [14,20]. After detection, the unknown compounds in the obtained data were compared with the National Institute of Standards and Technology (NIST) database for qualitative analysis. Only compounds with a match score exceeding 80% were selected, and mass spectral analysis was conducted in conjunction with the retention times of n-alkanes. Finally, compounds meeting the established criteria were screened and standardized to ensure data consistency and reliability, providing a foundation for subsequent in-depth analysis.

2.3. Measurement of Physical and Chemical Parameters

In this study, we monitored the fermentation process of the Zaopei by measuring five key parameters: moisture content, ethanol concentration, total acidity, starch, and reducing sugars. Moisture content was determined using the gravimetric method, where 10 g of the sample was dried in an oven (Shanghai, China) at 100–105 °C until a constant weight was achieved. Total acidity was assessed by titrating the sample with a 0.1 mol/L NaOH standard solution, using phenolphthalein as an indicator. Starch and reducing sugars were quantified using the Fehling’s reagent method [21].

2.4. Measurement of Amino Acids

A 1.0 g sample was placed into a 50 mL centrifuge tube, to which 10 mL of 80% ethanol (C₂H₅OH) was added and mixed thoroughly. The mixture was then centrifuged, and 1 mL of the supernatant was transferred to a 1.5 mL sample vial after filtration through a 0.22 μm membrane. The sample was analyzed in triplicate using an amino acid analyzer (Tokyo, Japan) [22].

2.5. High-Throughput Sequencing

We used the DNeasy PowerSoil Kit (Dusseldorf, Germany) to extract DNA from microbial samples of the sludge. The extracted DNA was subjected to agarose gel electrophoresis to assess the integrity and size distribution of the DNA bands. The V3-V4 region, typically ranging from 400 to 500 bp, is commonly used for amplification. The primers 338F (5′—ACTCCTACGGGAGGCAGCAG—3′) and 806R(5′—GGACTACHVGGGTWTCTAAT—3′) were used for amplification of this region. The concentration of the PCR products was determined using the QuantiFluor-ST Blue Fluorescence Quantification System (Promega, Madison, WI, USA) based on fluorescence intensity. Finally, sequencing of the total genomic DNA was performed by Shanghai Majorbio Tech using the Miseq Benchtop Sequencer (2 × 300 bp; Illumina MiSeq PE 300, San Diego, CA, USA).

2.6. Data Analysis

The raw sequences were processed and assembled using the Flash 1.2.7 and Fastp 0.20.0 software for quality control, and chimeric sequences were removed using the Uchime software. Sequences were clustered into operational taxonomic units (OTUs) based on 97% similarity using Qiime2 software [23]. Representative sequences of OTUs from bacteria and fungi were compared against the Silva 138 16S rRNA database and the Unite ITS database, respectively [19,24]. Alpha and Beta diversity indices were calculated to analyze the community composition differences between samples. Partial least squares discriminant analysis (PLS-DA) was performed using SIMCA 14.1 software (UMETRICS, Sweden). Spearman correlation coefficients between the relative abundance of key bacterial and fungal microbiota and major flavor compounds and physicochemical factors were calculated using R version 3.6.3 [25,26]. To further explore the relationships between fermentation parameters, microbiota, and flavor compounds [27], Bray–Curtis dissimilarity matrices of fermentation parameters were calculated to generate principal coordinate analysis plots. Functional prediction of 16S rRNA was performed using PICRUSt 2 2019 (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States). Data processing and significance analysis were conducted using Origin 2022 64 Bit (OriginLab Corporation, Northampton, MA, USA) and SPSS 26 (SPSS Inc., Chicago, IL, USA).

3. Results and Discussion

3.1. Changes in Physicochemical Properties During Zaopei Fermentation

The physicochemical properties of Zaopei include moisture content, ethanol, total acidity, starch, and reducing sugars. As shown in Figure 1A, moisture serves as a crucial medium for microbial reactions, influencing microbial growth metabolism and the progression of Maillard reactions [28]. Upon the initial stage, the moisture content was measured at 52.78%, with an increase during the early fermentation stages compared to the later stages. This phenomenon is attributed to the vigorous metabolic activities of microorganisms utilizing substrates like starch, leading to the production of additional moisture. The Zaopei contains a substantial amount of water-retaining rice husks, which help maintain moisture levels above 58% even in the later stages of fermentation. Starch and reducing sugars are vital nutritional sources for microorganisms and substrates for Maillard reactions [29]. As shown in Figure 1B, starch in the Zaopei can be degraded by microorganisms into reducing sugars, providing energy for other microbial growth and metabolism. The starch content decreased during the early fermentation stages, dropping from 19.16% to 8.49% over five days, with a slower consumption rate in the later stages, decreasing to 5.08% by day 40. This is due to the rapid growth and reproduction of dominant microorganisms like Aspergillus and Rhizopus in the early fermentation stages, which quickly decompose starch substrates [30,31]. As fermentation progresses, increased acidity inhibits fungal metabolic activities, leading to a gradual decline in fungal dominance [9]. As shown in Figure 1C, during the fermentation process of sesame-flavored Baijiu, the reducing sugar content exhibited a noticeable downward trend, decreasing from 2.98% to 1.28% by day 10, and then stabilizing in the later stages. In the later stages of fermentation, some microorganisms continue to decompose starch to produce reducing sugars, resulting in a dynamic equilibrium of reducing sugar content in the Zaopei, which remains relatively constant. Ethanol content showed a gradual increase, reaching its maximum in the later stages of fermentation. As shown in Figure 1D, ethanol content showed a gradual increase, reaching its maximum in the later stages of fermentation. In the later stages, microorganisms consumed a substantial amount of nutrients, and yeast content no longer increased, leading to the peak ethanol content [32]. As shown in Figure 1E, acidity is an important indicator in the fermentation process, closely related to the taste of the liquor. Appropriate acidity can also inhibit the growth of spoilage microorganisms in the Zaopei. Acidity increased from 1.23 mol/g to 3.63 mol/g, then slightly decreased, and finally rose to 3.72 mol/g.

3.2. Flavor Compound Analysis in the Fermentation Process of Sesame-Flavored Zaopei

The characteristic flavor compounds produced during the fermentation of sesame-flavored Baijiu have been extensively studied [33,34]. To investigate the changes in flavor compounds during the fermentation process of sesame-flavored Zaopei, GC-MS was used to analyze the volatile flavor metabolites. As shown in Figure 2A and Table S1, a total of 68 volatile organic compounds were detected. With the progression of fermentation, the overall content and diversity of volatile organic compounds significantly increased. These included 46 lipid compounds, seven alcohols, four aldehydes, four phenols, and seven acids. Among them, ester compounds, which contribute fruity aromas, are the primary components of the flavor of Chinese Baijiu, accounting for approximately 67% of the volatile substances in sesame-flavored Zaopei. Alcohol compounds are produced through the degradation of glucose and amino acids [35], and they play a key role in the formation of ester compounds [36,37], acting as precursors in the flavor compound formation pathway. Acid compounds, mainly the terminal products of microbial metabolic activities, are closely related to alcoholic fermentation and play a crucial role in shaping the distinctive flavor of Baijiu [35].

Principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) were used to examine the differences in volatile components at various fermentation stages. The PCA plot (Figure 2B) shows a clear distinction between the sample at time point zp_0 and the other groups. PLS-DA was used to assess the major aroma components, with the results shown in Figure 2C. A 200-times permutation test was performed to verify the reliability of the PLS-DA model, and the model was confirmed to meet reliability standards based on the relevant parameters and statistical characteristics, with R² and Q² values of 0.276 and −0.761, respectively (Figure 2D). During the investigation of flavor differences in ZP, variable importance projection (VIP) analysis was employed to accurately identify key variables. Flavor compounds with VIP > 1 and p < 0.05 were considered key flavor compounds that represent the flavor characteristics of Zaopei. As a result, 21 primary flavor compounds were identified (Figure 2E). Sixteen of these were key aromatic substances, including Ethyl Oleate, 1-Butanol, 3-methyl-, Hexyl n-valerate, Decanal, Pentanoic acid ethyl ester, and others.

The concentration of flavor compounds alone cannot be used to define the characteristic aroma of Zaoepi. The Odor Activity Value (OAV) is typically used to identify the primary components of the aroma (Table S2). Previous studies have calculated OAV by dividing the concentration by the aroma threshold, and components with an OAV ≥ 1 are considered to have a noticeable effect on the overall aroma, while those with an OAV ≥ 10 are deemed to contribute significantly to the aroma profile [38]. Based on the reported aroma thresholds and descriptions, the OAVs for the flavor compounds in Zaopei were calculated. The results showed that 16 aroma compounds had an OAV ≥ 1, with Hexanoic acid ethyl ester, Octanoic acid ethyl ester, Benzenepropanoic acid ethyl ester, 4-Ethyl-2-methoxyphenol, and Hexadecanoic acid ethyl ester all having OAVs greater than 10. Among them, Hexanoic acid ethyl ester and Octanoic acid ethyl ester had OAVs greater than 100, indicating that these compounds are likely key flavor substances in sesame-flavored Baijiu Zaopei, aligning with the flavor profile of sesame-flavored Baijiu [14]. Among these, Hexanoic acid, ethyl ester, and Octanoic acid, ethyl ester play crucial roles in imparting a rich, mellow flavor, significantly contributing to the unique profile of traditional Chinese fermented Baijiu [39].

3.3. Amino Acids in the Fermentation Process of Zaopei

Free amino acids are individual amino acids that are not incorporated into peptide chains or proteins. They can serve as precursors for other flavor compounds [40]. Free amino acids can react with other compounds, such as sugars and aldehydes, leading to the formation of a variety of complex flavor compounds. For instance, the Maillard reaction involves the interaction between free amino acids and reducing sugars, producing compounds with a range of flavors, including roasted and nutty notes [41]. A comparison of total amino acids and essential amino acids (EAA) revealed significant differences between the two (p < 0.05). Sixteen amino acids were detected in different Zaopei samples, with the EAAs including Val, Met, Ile, Leu, Phe, Thr, and Lys, as shown in Figure 3A,C. The total content of FAA and EAA exhibited a trend of first increasing and then decreasing over the fermentation cycle. This trend may be due to the initial increase in amino acids, which are typically generated by the hydrolysis of proteins by microbial proteases. Later in fermentation, as microbial growth and metabolism intensify, amino acids are utilized as precursors for flavor compound production [42]. Based on their taste characteristics, free amino acids can be classified into sweet (Gly, Ala, Ser, Thr), bitter (Pro, Val, Met, Ile, Leu, Tyr, Phe, His, Lys, Arg), and umami (Asp, Glu) categories [43]. Notably, the content of bitter amino acids exceeded the combined total of umami and sweet amino acids, as shown in Figure 3B. Among the samples from different fermentation periods, Ala was found to have the highest concentration (Figure 3C). Ala is widely recognized for contributing a sweet flavor to foods [44]. Amino acid analysis further revealed a continuous accumulation of amino acids during Zaopei fermentation, which is beneficial for the synthesis of Baijiu-related compounds [45]. The composition of amino acids in each sample is shown in Table S4.

3.4. Microbial Community Composition and Differences in Zaopei Samples

The microbial species abundance and diversity of 21 Zaopei samples were evaluated using the ACE index, Chao index, Shannon index, Simpson index, and Coverage, under the condition that OTU similarity was greater than or equal to 97%, as shown in Table S3. Table S5 includes the total number of reads obtained and the number of OTUs for each sample. Higher Shannon and ACE indices corresponded to lower Simpson indices, indicating greater richness and even distribution of microbial community diversity. Based on the rarefaction curve, the number of sequences retrieved was sufficient to reach saturation, as shown in Figure S1, confirming that both bacterial and fungal samples are suitable for bioinformatics analysis.

The bar chart of community composition reflects the dominant species and their relative abundance in different samples. During the initial fermentation stage of Zaopei, the predominant bacterial phyla were identified as Firmicutes, Actinobacteria, Proteobacteria, Bacteroidota, Cyanobacteria, Acidobacteriota, Patescibacteria, Myxococcota, Nitrospirota, and Chloroflexi (Figure 4A). On day 10, the Firmicutes reached their highest relative abundance (99.8%) during the fermentation process. The relative abundance of Proteobacteria showed a declining trend throughout the fermentation stages, decreasing from 19.5% on day 0 to 0.02% on day 40. At the genus level(Figure 4B), Lactobacillus dominated on day 5, accounting for 91.2%, and maintained dominance until the later stages of fermentation, although with a slight decrease. This reduction could be attributed to the increased ethanol content in the later stages, which may have hindered the growth of lactic acid bacteria [46]. Genera such as Bacillus, Acetobacter, Thermoactinomyces, Staphylococcus, Weissella, and Kroppenstedtia showed high abundance initially, but their relative abundance decreased significantly by day 5. The microbial richness of Zaopei was highest on day 0 and lowest on day 10. During fermentation, the microbial species richness in Zaopei gradually decreased, accompanied by microbial succession. Lactobacillus is a facultative anaerobe, while Bacillus is an obligate aerobe. In the Zaopei samples, the Ascomycota phylum was the predominant fungal group (99.6%) (Figure 4C). Dominant fungal genera included Issatchenkia, Saccharomyces, Aspergillus, Kodamaea, and Wickerhamomyces. Issatchenkia remained the most dominant genus throughout the fermentation process, reaching a relative abundance of 83.9%. The relative abundance of Saccharomyces and Aspergillus was higher in the early fermentation stages, but as the fermentation cycle progressed, Issatchenkia gradually replaced them as the dominant genus. The dynamic succession of microbial communities is closely related to the interaction of fermentation environmental factors. In the early stage of fermentation, high moisture content and sufficient substrates provide conditions for the rapid proliferation of aerobic and facultative anaerobic organisms from the genera Bacillus, Acetobacter, as well as fungal yeasts from the genera Saccharomyces and Aspergillus. Bacillus accelerates starch degradation by secreting amylase, while yeasts dominate the ethanol conversion of carbohydrates, leading to early ethanol accumulation. As the fermentation process progresses, the acidic environment and high ethanol concentration significantly inhibit aerobic bacteria and acid-sensitive bacteria, while promoting the competitive advantage of acid-tolerant and ethanol-tolerant facultative anaerobic bacteria from the genus Lactobacillus and fungi from the genus Issatchenkia. Lactobacillus produces lactic acid through homofermentation, further lowering the pH and inhibiting contaminating microorganisms, while Issatchenkia, with its strong ethanol tolerance and esterase activity, dominates the synthesis of ester compounds in the later stages [47]. Furthermore, substrate consumption and the accumulation of metabolic byproducts trigger nutritional competition among microorganisms, leading to a decrease in community diversity and the formation of a stable symbiotic system centered around the phylum Firmicutes and the phylum Ascomycota. This succession mechanism shapes the flavor profile of sesame-flavored Baijiu.

3.5. Correlation Between Microbial Communities, Physicochemical Properties, and Volatile Components

RDA (redundancy analysis) was used to predict the impact of environmental factors on microbial dynamics during fermentation. This method helped elucidate the relationships between environmental factors, samples, and major microbial communities. The distance between points represents the degree of difference in community structure; the longer the radius, the greater the influence of the corresponding factor on the microbial community. The angle between the vectors indicates the degree of correlation between the factors. As shown in Figure 5A,B, different colors and shapes of points represent various Zaopei fermentation stages. RDA revealed that the moisture, acidity, and alcohol content in the Zaopei had a consistent impact on the bacterial community, while starch and reducing sugar content showed a consistent effect on the microbial communities as well. Samples from fermentation days 5, 10, 20, 30, 35, and 40 clustered together, whereas the day 0 sample was distinct from the others. Among the top five microbial genera in relative abundance, Lactobacillus was negatively correlated with starch and reducing sugars and positively correlated with moisture, acidity, and alcohol. RDA results suggested that moisture, acidity, and alcohol content were the most influential physicochemical factors on the bacterial microbiota in Zaopei. In fungi, the RDA results were similar to those of bacteria. During fermentation, starch and reducing sugars showed distinct interaction patterns with microbial communities compared to the effects of moisture, acidity, and alcohol. This finding aligns closely with the dynamic changes in physicochemical factors throughout fermentation, reflecting the complex relationships within the fermentation system. Further analysis through Spearman’s correlation identified the top 10 dominant genera significantly affected by environmental physicochemical parameters. The study found that, in the bacterial community, moisture and acidity showed a significant positive correlation with Saccharopolyspora, Weissella, Staphylococcus, Corynebacterium, and Acetobacter, while starch and reducing sugars exhibited a significant negative correlation with these genera. Alcohol content was significantly negatively correlated with Bacillus and Kroppenstedtia but positively correlated with Lactobacillus. In the fungal community, six dominant genera were significantly correlated with fermentation parameters (p < 0.05). Moisture content showed a significant positive correlation only with Kodamaea. Acidity was positively correlated with Alternaria and Issatchenkia and negatively correlated with Aspergillus. Starch and reducing sugars were positively correlated with Aspergillus and Saccharomyces and negatively correlated with Issatchenkia and Alternaria. Alcohol content was significantly negatively correlated with Thermomyces and Pichia but positively correlated with Wickerhamomyces. Environmental factors, including moisture, acidity, and alcohol content, significantly influence the microbial community dynamics. Early-stage dominant microbial groups such as Saccharomyces, Bacillus, and Acetobacter were gradually replaced by Lactobacillus and Issatchenkia as alcohol levels increased, leading to a reduction in microbial diversity [48,49].

3.6. Correlation Between Bacterial and Fungal Communities and Amino Acids and Volatile Flavor Compounds

Spearman’s correlation coefficient was used, along with cluster heatmaps and correlation network diagrams, to reveal the potential correlations between dominant microbial communities and key volatile flavor compounds during Zaopei fermentation. The correlation results between the microbial communities and major volatile compounds at different fermentation times are shown in Figure 6A,B. During the fermentation process, 14 microbial species were significantly correlated with five key volatile compounds: Hexanoic acid, ethyl ester, Octanoic acid, ethyl ester, Benzenepropanoic acid, ethyl ester, 4-Ethyl-2-methoxyphenol, and Hexadecanoic acid, ethyl ester. Among them, Hexanoic acid was significantly negatively correlated with the family Dipodascaceae. Octanoic acid, ethyl ester showed significant positive correlations with Dipodascaceae, Weissella, Staphylococcus, Corynebacterium, and Acetobacter, and a significant negative correlation with Issatchenkia. Benzenepropanoic acid, ethyl ester was significantly positively correlated with Thermomyces, Kroppenstedtia, Bacillus, Clostridium_sensu_stricto_1, Weissella, Staphylococcus, Corynebacterium, Saccharopolyspora, Aspergillus, and Acetobacter, but negatively correlated with Lactobacillus. 4-Ethyl-2-methoxyphenol showed significant positive correlations with Thermomyces, Kroppenstedtia, Bacillus, Pichia, and Acetobacter, but negatively correlated with Lactobacillus. Hexadecanoic acid, ethyl ester had significant positive correlations with Dipodascaceae, Kroppenstedtia, Pichia, Weissella, Staphylococcus, Corynebacterium, and Acetobacter, and a significant negative correlation with Issatchenkia. Further analysis revealed that the relative content of acetic acid, 2-phenylethyl ester, and Benzenepropanoic acid, ethyl ester was influenced by multiple microbial communities, suggesting that these compounds result from the synergistic fermentation of various microbes. During the fermentation process of Zaopei, the synthesis of certain key flavor compounds, such as ethyl hexanoate, involves both direct enzymatic synthesis and indirect synergistic interactions between microorganisms. For instance, in ethyl hexanoate, the esterification of ethanol with carboxylic acids is catalyzed by microbial esterases. Most yeast genera, such as Saccharomyces, show a positive correlation with ester compounds, promoting the production of these lipid substances.

As shown in Figure 7A,B, the amino acids in the Zaopei were significantly correlated with 14 dominant microbial genera. Among them, Aspergillus, Saccharopolyspora, Corynebacterium, Weissella, Acetobacter, and Staphylococcus showed significant negative correlations with most amino acids. Alternaria was significantly positively correlated with His, Asp, Thr, Tyr, Ala, Ile, Leu, Val, Gly, Arg, Met, Glu, and Lys. Pichia was significantly positively correlated with His, Asp, Thr, Val, Met, and Lys. Lactobacillus was significantly positively correlated with Pro and Ser. Saccharopolyspora was significantly positively correlated with Phe. Free amino acids, serving as nitrogen sources for microbial growth and precursors for flavor compounds, significantly promote microbial growth [50]. In summary, the Spearman correlation analysis confirmed the relationships between key microbial taxa, flavor compounds, and amino acids. The results indicated Acetobacter was significantly associated with 10 major flavor compounds, particularly Benzenepropanoic acid, ethyl ester and Hexadecanoic acid, ethyl ester (p < 0.001). The abundance of Acetobacter is closely linked to Baijiu flavor, with higher levels observed in lower-alcohol environments. Additionally, Acetobacter exhibited a negative correlation with sulfur-containing compounds, suggesting its role in reducing undesirable sulfurous notes by promoting the conversion of volatile sulfur compounds into less volatile forms [51,52]. Bacillus showed a strong correlation with Benzenepropanoic acid, ethyl ester and Phenol, 4-Ethyl-2-methoxyphenol (p < 0.01), highlighting its role in flavor compound synthesis. Wang found that Bacillus licheniformis, in synergy with the fermentation system, enhances glucose metabolism and facilitates the production of ethanol and tetramethylpyrazine, key contributors to Baijiu aroma [53]. The 16S rDNA sequencing confirmed that yeasts dominate the fungal community, playing a crucial role in fermentation while also influencing the quality and characteristics of Baijiu with distinct flavor profiles [54]. During Baijiu fermentation, amino acids contribute not only to flavor synthesis and quality enhancement but also to fungal community assembly and yeast metabolic activities, which are essential for physiological processes [45].

3.7. Functional Prediction of the Bacterial Community in the Fermentation Pit Using PICRUSt2

To analyze the differences in microbial functions of Zaopei over different time periods, the Zaopei in the Baijiu cellar was subjected to PICRUSt 2 functional prediction [55]. In this study, 46 KEGG pathways were identified at level 2, as shown in Figure S3. The top 15 most abundant KEGG level 2 pathways are shown in Figure 8A, primarily related to amino acid, ester, nucleotide metabolism, and the transcription, translation, and signal transduction of bacteria. Further analysis at KEGG level 3 identified 383 pathways, with 293 showing significant differences among multiple groups (Figure S4). The top 15 most abundant KEGG level 3 pathways are illustrated in Figure 8B. Notably, during the early fermentation on day 5, the function related to Phenylalanine, tyrosine, and tryptophan biosynthesis (ko 00400) was significantly active, gradually decreasing as fermentation progressed. In contrast, the function related to pyrimidine metabolism (ko 00240) dominated in the later stages, possibly associated with microbial proliferation and nucleic acid metabolism. To study the main pathways and their primary relationship with bacterial microorganisms in the Zaopei, the Spearman correlation heatmaps in Figure 8C,D show the correlation between 15 metabolic pathways and the top 10 dominant microbial genera. At KEGG level 2, Lactobacillus, the dominant genus in Zaopei, was significantly positively correlated with lipid metabolism, replication and repair, transcription, translation, all of which are associated with microbial proliferation and the metabolism of esters. Bacillus showed significant correlations with transport and catabolism, the digestive system, and amino acid metabolism, which may be related to Bacillus’s strong amylase and protease activities in protein degradation [56]. This provides substrates for later flavor precursors such as alcohols and aldehydes. It is noteworthy that genera such as Kroppenstedtia and Thermoactinomyces are significantly positively correlated with the MAPK signaling pathway (ko04011), aromatic amino acid biosynthesis (ko00400), and caffeine metabolism (ko00232), suggesting their key role in the production of complex flavor compounds such as pyrazines and phenolics. In contrast, the significant negative correlation of these genera with pyrimidine metabolism may reflect their metabolic adaptive shifts in the later stages of fermentation. In summary, the microbial community in the cellar drives the dynamic formation of flavor compounds in the pomace by synergistically regulating pathways such as carbon and nitrogen metabolism and secondary metabolite synthesis. The functional differentiation of key genera such as lactic acid bacteria and Bacillus provides a theoretical basis for elucidating the microbial mechanisms underlying the characteristic flavors of strong-aroma Baijiu.

4. Conclusions

This study systematically explores the dynamic changes and interactions of physicochemical properties, flavor compounds, amino acids, and microbial communities during the fermentation process of sesame-flavored Zaopei. The results indicate that the rapid consumption of starch and reducing sugars in the early stages of fermentation promotes active microbial metabolism, leading to an overall increase in moisture, ethanol, and acidity, with the accumulation of acidity effectively inhibiting the metabolic activity of fungi in the later stages. Flavor analysis shows that Hexanoic acid ethyl ester, Octanoic acid ethyl ester, Benzenepropanoic acid ethyl ester, 4-Ethyl-2-methoxyphenol, and Hexadecanoic acid ethyl ester are the main volatile components of Zaopei, with high OAV (>10) indicating their crucial role in the rich flavor of sesame-flavored Baijiu. The dynamic changes in amino acids reveal that Ala content is predominant during different fermentation periods, contributing sweetness to Zaopei. Microbial community analysis reveals that Lactobacillus and Issatchenkia dominate in the later stages of fermentation, with their abundance showing a significant positive correlation with moisture, acidity, and ethanol content, while the early dominant microbial community. The metabolic activities of Bacillus and Acetobacter lay the foundation for the formation of flavor precursors. Correlation analysis further confirms that the succession of microbial communities is closely related to the metabolism of characteristic flavor compounds such as Benzenepropanoic acid ethyl ester and amino acids, with genera such as Lactobacillus and Kroppenstedtia influencing the flavor profile by regulating ester synthesis and phenolic production, respectively. Functional predictions indicate that microorganisms drive the dynamic formation of flavor compounds through the synergistic regulation of carbon and nitrogen metabolism, lipid metabolism, amino acid metabolism, and secondary metabolite synthesis pathways. In summary, the dynamic changes in physicochemical parameters during the Zaopei fermentation process, the succession of microbial communities, and the synergistic effects of their metabolic functions collectively shape the unique flavor characteristics of sesame-flavored Baijiu, providing a theoretical basis for optimizing traditional fermentation processes. Future research should further explore microbial metabolism in the fermentation pit using metabolomics, focusing on the synthesis of key flavor compounds and amino acids. Additionally, identifying genes associated with the production of characteristic flavor compounds will be crucial for a deeper understanding of the molecular mechanisms of Baijiu fermentation.