Spatial fungal community diversity pro�les in pit mud samples from fermentation cellars used to manufacture Chinese Strong-�avour liquor


 Background

Few studies to date have sought to characterize the spatial profiles of pit mud microbial communities in fermentation cellars from Chinese strong-flavour liquor distilleries. This study was thus designed to evaluate these eukaryotic communities in pit mud samples via a multidimensional DGGE approach and by assessing associated sample physicochemical properties.
Results

 Penicillium fuscoglaucum, Penicillium glandicola, Aotearoamyces nothofagi, Malassezia restricta, Penicillium lanosocoeruleum, Penicillium crustosum, and Aspergillus tonophilus were detected only in pit mud from the upper cellar wall, while Alternaria alstroemeriae, Trichosporon insectorum, Fusarium equiseti, Calonectria pseudoreteaudii, Penicillium clavigerum, Penicillium compactum, Ascochyta phacae, Metarhizium frigidum, Alternaria burnsii, Fusarium nurragi, and Didymella keratinophila were present only in the middle cellar wall layer. Alternaria zantedeschiae and Ilyonectria cyclaminicola were only present in pit mud samples from the lower cellar wall, while Leptobacillium leptobactrum, Calonectria queenslandica, Aspergillus appendiculatus, and Candida pseudolambica were only detected in pit mud from the cellar bottom. Moisture, pH, PO43−, acetic acid, humus, K+, Mg2+, Ca2+, acetic acid, butyric acid, and caproic acid levels in these different pit mud positions exhibited a rising incremental pattern from the upper wall layer to the bottom layer, whereas lactic acid levels were significantly lower in the bottom pit mud layer relative to other layers.
Conclusions

A clear relationship between fungal community structure and physicochemical variables in different spatial pit mud samples, especially moisture, pH, and NH4+-N were identified as the three most significant factors associated with fungal community through a redundancy analysis.


Background
Chinese strong-avour liquor is a traditional fermented beverage that accounts for roughly 70% of total liquor consumption in China [1]. Owing to its unique avour and brewing approach, strong-avour liquor holds a special status in Chinese culture and history. This liquor is distilled in large rectangular pit cellars (3600 x 2300 mm at the top; 2800 x 1540 mm at the bottom; 2400 mm deep) that serve as fermentation vessels (Fig. 1). The walls of these pits are covered with a speci c type of fermented clay known as pit mud that contains large quantities of functional microorganisms including Clostridium spp., Bacillus spp., and Methanobacterium spp., all of which serve as key mediators of the fermentation process and sources of the aromatic compounds characteristic of Chinese strong-avour liquor [2]. Indeed, the microbes within pit mud are generally accepted to play an essential role in the process of Chinese strongavour liquor fermentation [3]. Given their importance, many studies have analyzed these microbial communities in an effort to better understand the mechanisms whereby these organisms contribute to the liquor production process [4].
Studies of pit mud conducted to date have primarily focused on prokaryotic ora [2]. For example, Liu et al. (2015) used a DGGE approach to explore Clostridium cluster I community diversity in samples of pit mud from cellars of different ages (1, 50, 100, and 400 years), revealing C. ragsdalei, C. ljungdahlii, C. autoethanogenum, and C. kluyveri to be the dominant species therein [3]. Liang et al. (2020) also employed a combination of PCR-DGGE and qPCR approaches to detect higher levels of Clostridium IV species in aged pit mud relative to aging pit mud, which they speculated may be associated with the fact that aged pit mud has a strong aroma whereas aging pit mud does not [5]. Ding et al. (2014)also conducted a nested PCR-DGGE-based study of eubacterial community structures in Chinese strongavour liquor pit mud and found that community diversity was greater in the bottom of the cellar relative to in the cellar walls [6].
Few studies to date have sought to characterize the spatial pro les of pit mud microbial communities in fermentation cellars from Chinese strong-avour liquor distilleries. This study was thus designed to evaluate these eukaryotic communities in pit mud samples via a multidimensional DGGE approach and by assessing associated sample physicochemical properties. In so doing, we aim to improve pit mud quality and consistency, and to facilitate the generation of arti cial pit mud. by exploring pit mud microbial and physicochemical properties. This study is the rst to our knowledge to have explored these multidimensional distributions of fungal communities and physicochemical properties in different spatial positions of pit mud by using PCR-DGGE methods.

DGGE pro ling of fungal communities
We began by characterizing the DGGE ngerprint pro les for pit mud fungal communities (Fig. 2). There were clear differences in the communities present within pit mud samples from the upper wall, middle wall, lower wall, and bottom cellar layers ( Table 1). The Shannon-Wiener index value for the fungal community from the middle wall layer was greater than the corresponding values for the other analyzed pit mud samples, suggesting that maximal fungal diversity was present within this middle wall layer. The evenness index (E) values for these different fungal communities were between 0.961 and 0.996, with these values being higher for samples from the middle wall and cellar bottom relative to other samples. Middle wall pit mud samples also exhibited the highest species richness index value, followed by samples from the bottom of the cellar, with no signi cant differences in these values when comparing samples from the upper or lower cellar wall.  a Numbers are those of bands shown in Fig. 2.
b Most homologous BLAST-derived match.
UPGMA dendrograms were constructed for DGGE pro les based upon Dice coe cient values in order to describe community similarity between pit mud samples from different positions within the fermentation cellar (Fig. 3). Cluster analyses of these fungal pro les revealed that pit mud samples from the upper wall layer formed a group, while the primary microbial populations present in samples from the lower wall layer were similar to those in pit mud samples from the cellar bottom (Fig. 3).

Physiochemical properties
The physicochemical properties of pit mud samples from different cellar positions were next evaluated (  [7]. Levels of NH 4 + -N were higher in the bottom pit mud layer relative to other layers, whereas these levels did not differ signi cantly between the middle and bottom wall pit mud layers, and were lowest in the upper wall layer pit mud samples. In contrast, lactic acid levels exhibited the opposite trend such that these levels were signi cantly lower in the bottom pit mud wall layer.

Relationships between fungal communities and physicochemical variables
A redundancy analysis (RDA) was next conducted to better clarify potential relationships between the 51 detected fungal genera and the 12 analyzed environmental factors (Fig. 5). The rst two component axes explained 77.6% of the variation in fungal composition, with species-speci c environmental correlations for both axes of 48.1% and 78.6%, respectively, indicating that fungal community structure was moderately correlated with these physicochemical variables. An interactive forward selection procedure was used to evaluate these 12 environmental variables, revealing that moisture, pH, and NH 4 + -N contributed signi cantly to community composition (39.5%, 13.8%, and 13.8%, respectively; P < 0.01), whereas the other 8 variables exhibited no signi cant correlations.

Discussion
Chinese strong-avour liquor is prepared through the fermentation of a mixture of sorghum, rice, and wheat known as Zaopei in a rectangular cellar composed of pit mud. This pit mud is an ideal habitat for microbes that are integral to the distillation process, serving as key determinants of the avour of the resultant liquor. The quality of pit mud is thus an important regulator of the quality and taste of the liquor produced.
Pit mud tends to age with increasing cellar usage, and the microbial communities present within this mud vary based upon their spatial location within the walls or bottom of the cellar. A range of sensory descriptions and physicochemical indices have been used to describe pit mud from different locations within these fermentation cellars. For example, pit mud from the bottom of these cellars is often described as smooth, ne, soft, moist, and sooty with an aroma of esters, ammonia, and hydrogen sul de. In contrast, pit mud from the top of these cellars is rough, hard, dry, and light grey with white lumps or aciform crystals and no aroma. While pit mud from the bottom layer can support the production of good-quality liquor, that from the upper layer cannot. As such, studying the microbial communities present within pit mud is essential in order to understand the molecular mechanisms governing the avor and aroma of Chinese strong-avour liquor in an effort to improve the quality of this popular and culturally important beverage.
In prior studies, researchers have utilized both culture-dependent and -independent strategies to determine that bacteria, fungal, archaea, and actinomycetes species are present within pit mud samples, with bacteria and archaea being dominant in this environment [5]. At the family level, common pit mudresident bacteria include haloplasmatacea, Bacillaceae, planococcaceae, synergistaceae, staphylococcaceae, Thermoanaerobacter, and clostridiaceae species. Archaea present within pit mud are largely consistent across regions, and primarily include methanobacteria (Methanobacteriaceae), Methanococcus (Methanococcus), and thermoplasmataceae (thermoplasmata) species [5]. Microbes in the Clostridia class are thought to be primary producers of short-and medium-chain fatty acids including butanoic acid and hexanoic acid, which are directly relevant to the liquor production process [4]. Liu et al. Herein, we explored the structures of fungal communities in multidimensional pit mud environments via a DGGE approach, revealing clear discrimination between the communities present in different locations within the fermentation cellar. Penicillium fuscoglaucum, Penicillium glandicola, Aotearoamyces nothofagi, Malassezia restricta, Penicillium lanosocoeruleum, Penicillium crustosum, and Aspergillus tonophilus were only present in the upper cellar wall pit mud layer, whereas Alternaria alstroemeriae, Trichosporon insectorum, Fusarium equiseti, Calonectria pseudoreteaudii, Penicillium clavigerum, Penicillium compactum, Ascochyta phacae, Metarhizium frigidum, Alternaria burnsii, Fusarium nurragi, and Didymella keratinophila were presented only detected in the middle wall layer. Similarly, Alternaria zantedeschiae and Ilyonectria cyclaminicola were only identi ed in the lower cellar wall pit mud layer, while Leptobacillium leptobactrum, Calonectria queenslandica, Aspergillus appendiculatus, and Candida pseudolambica were only detected in the pit mud found on the bottom of the fermentation cellar. These differences may explain why the quality of strong-avour liquor varies with cellar position. We found that fungal abundance in the upper and middle layers was signi cantly higher than that in the lower wall and bottom layers, potentially due to the lower oxygen levels in these latter two environments, as such oxygen de ciency may have compromised fungal survival [10]. This, in turn, may explain the higher sacchari cation e ciency that is typically detected in the upper and middle Zaopei layers in the context of liquor fermentation.
With respect to pit mud physicochemical properties, we found that moisture, pH, PO 4 3-, acetic acid, Humus, K + , Mg 2+ , Ca 2+ , acetic acid, butyric acid, and caproic acid levels rose with sample position from the upper wall to the bottom of the fermentation cellar, suggesting that organic compounds were gradually degraded with position. The maximal moisture levels in the bottom pit mud layer may be associated with the high levels of Huangshui present in this setting. The higher pH levels lower in the cellar may be linked to the degradation of various acids such as lactic acid [9], and to the synthesis of ammonium nitrogen, consistent with the observed trends in NH 4 + -N levels. The lower acetic acid levels with the upper wall pit mud layer are consistent with less robust prokaryotic metabolism in this location, given that acetic acid is a metabolic end product produced by many bacterial species [4]. The rising lactic acid levels detected from the bottom of the pit to the upper pit may correspond to the different Lactobacillus activity levels in these positions.
Many prior studies have sought to understand the relationship between pit mud physicochemical properties and the microbial communities therein. Meng et al. (2020), for example, found that these properties were signi cantly in uenced by depth within the fermentation cellar [7]. Zhang et al. (2020) found that acid and amino nitrogen concentrations were higher in the bottom pit mud layers relative to other positions, suggesting that these compounds may in uence the overall diversity of the microbial communities found within this bottom layer [4]. We similarly detected a clear relationship between fungal community structure and physicochemical variables in pit mud samples. However, further research will be essential to develop the e cient cultivation strategies necessary to delineate the independent contributions of different fungi to the production of Chinese strong-avour liquor production.

Conclusions
This study explored the multidimensional distributions of fungal communities and physicochemical properties in different spatial positions of pit mud by using PCR-DGGE methods. There were clear differences in the fungal communities present within pit mud samples from the upper wall, middle wall, lower wall, and bottom cellar layers. RDA analysis demonstrated that a clear relationship between fungal community structure and physicochemical variables in different spatial pit mud samples, especially moisture, pH, and NH 4 + -N were identi ed as the three most signi cant factors associated with fungal community through a redundancy analysis. This study provide theoretical basis to design effective strategies to manipulate microbial consortia for better improving pit mud quality in Chinese strongavour liquor production.

Sample collection
Samples of pit mud were obtained from ~20-year-old pits from a well-known liquor manufacturer (Anhui Yingjia Distillery Group Co., Ltd.) located in Luan city, Anhui province, China. Sampling sites are detailed in Fig. 1. Sampling was conducted as per a strati ed random approach [11]. Pit mud wall samples were collected from the center of each wall, with approximately 50 g of mud being collected per position and mixed to yield a composite sample. Samples of pit mud from the cellar bottom were collected from all corners and the center of each pit, and were mixed together. All samples were collected at a depth of ~5 cm. Samples of mixed pit mud from these different sampling sites were separated into small ~100 g samples and stored at -20°C prior to analysis.

Physiochemical property analyses
Pit mud moisture levels were established by drying samples for 3 h at 115℃. Pit mud pH values were established with a Mettler Toledo pH meter after diluting sample 1:4 (w/v) with ddH 2 O. for 3 h. Pit mud ammonium (NH 4 + -N) levels were established via extraction in 10% (w/v) NaCl at 1:10 (w/v) ratio, after which concentrations were measured using a UV spectrophotometer. Acetic acid, butyric acid, and caproic acid were extracted using 15% methanol and quanti ed via gas chromatography (Agilent 7890, US) as described previously [12]. Lactic acid (LA) levels were quanti ed via utra-high-performance liquid chromatography (UPLC, Acquity I-class, Waters, US) as previously reported [1].

DGGE band sequencing
Representative DGGE were excised with a sterile scalpel, and were added to ultrapure water overnight at 4°C to facilitate sample elution. Samples from eluted bands were then again ampli ed with the GC-clamp primers detailed above, After ampli cation, samples were again assessed via DGGE gels to con rm purity. Bands were then re-ampli ed using the same primers without the GC clamp, and were puri ed using a universal PCR puri cation kit (Sangon, Shanghai, China). Cloning and sequencing were then performed by Sangon, and the resultant sequences were compared to ITS sequences in the GenBank (http://www.ncbi.nlm.nih.gov) databases to identify the closest phylogenetic relatives.

Data analysis
Cluster and community diversity analyses were performed with the Quantity One software, with individual DGGE lanes being converted into densitometric pro les. Fungal community Shannon-Wiener index of general diversity (H), the Evenness (E), and the richness (S) values were then calculated based upon relative band intensity with the PAST software package (Palaeontology Statistics, http://folk.uio.no/ohammer/past/). The unweighted pair group method with arithmetic averages (UPGMA) was used for sample clustering.

Declarations
Ethics approval and consent to participate Not applicable. The pro les of the Chinese strong-avour liquor pit mud (A) and the sampling sites of pit mud (B). The bands indicated with numbers were excised and sequenced and the alignment results are listed in Table 2.