Light-flavor Baijiu fermentation is a typical spontaneous solid-state fermentation process fueled by a variety of microorganisms. Mechanized processes have been increasingly employed in Baijiu production to replace traditional manual operation processes, however, the microbiological and physicochemical dynamics in mechanized processes remain largely unknown. Here, we investigated the microbial community succession and flavor compound formation during a whole mechanized fermentation process of light-flavor Baijiu using the conventional dilution plating method, PacBio single-molecule real-time (SMRT) sequencing and headspace solid-phase microextraction coupled with gas chromatography-mass spectrometry. The results showed that largely different fungal and bacterial communities were involved in the soaking and fermentation processes. A clear succession from Pantoea agglomerans to Bacillus (B.) smithii and B. coagulans in dominant bacterial species and from Cladosporium exasperatum to Saccharomyces cerevisiae and Lichtheimia ramosa in dominant fungal species occurred in the soaking processes. In the fermentation process, the most dominant bacterial species was shifted from Pantoea agglomerans to Lactobacillus (La.) acetotolerans and the most dominant fungal species were shifted from Lichtheimia ramose and Rhizopus arrhizus to Saccharomyces cerevisiae. The bacterial and fungal species positively associated with acidity and the formation of ethanol and different flavor compounds were specified. The microbial species exhibited strong co-occurrence or co-exclusion relationships were also identified. The results are helpful for the improvement of mechanized fermentation process of light-flavor Baijiu production.

Daqu is a traditional starter for Baijiu fermentation and is produced by spontaneous fermentation of ground and moistened barley or wheat. The quality of Daqu is traditionally evaluated based on physicochemical and subjective sensory parameters without microbiological analysis. Here, we compared the physicochemical characteristics of qualified (QD) and inferior (ID) Daqu, their microbial communities based on plate counting and PacBio SMRT sequencing of rRNA gene libraries, and their impacts on Baijiu fermentation. The results showed that the glucoamylase and α-amylase activities of QD were significantly higher than those of ID. The counts of yeasts and relative abundances of functional microbes, especially the amylolytic bacterium Bacillus licheniformis and fungi Saccharomycopsis fibuligera and Lichtheimia ramosa, were significantly higher in QD than in ID. The laboratory-scale Baijiu fermentation tests showed that the relative abundances of the amylolytic microbes were higher in the QD than the ID fermentation set, resulting in more efficient fermentation, as indicated by more weight loss and higher moisture content in the former. Consequently, more glycerol, acetic acid, ethanol, and other volatile compounds were produced in the QD than in the ID fermentation set. The results suggest that Daqu quality is determined by, and can be evaluated based on, its microbial community.

Medium-temperature Daqu (MT-Daqu) serves as a crucial saccharifying and fermentation agent in the production of strong-flavor Baijiu. Due to the spatial heterogeneity of solid fermentation, significant differences occurred in the fermentation state and appearance features in different parts of Daqu during fermentation. Currently, the understanding of the underlying mechanism behind this phenomenon remains limited. Here, we analyzed the microbial succession and assembly models and driving factors in different parts of MT-Daqu at the species level based on the PacBio single-molecule real-time sequencing technology. The results showed significantly different bacterial and fungal community compositions, successions, and interaction patterns in different parts of MT-Daqu. The bacterial community composition and succession model in the middle layer were similar to those in the core layer, whereas the fungal community composition and succession model in the surface layer were similar to those in the middle layer. The co-occurrence network analysis showed that microbial interaction is stronger in the middle and core layers than in the surface layer. Analyses based on both niche theory and neutral theory models indicated that deterministic processes predominantly governed the microbial community assembly and these processes played an increasingly important role from the surface to the core layer. Random forest analysis revealed that temperature was the primary endogenous factor driving the bacterial and fungal community assembly. The results of this study contribute to a better understanding of the microbial community in MT-Daqu and are helpful for the quality control of MT-Daqu fermentation.

The arthroconidial yeast-like species currently classified in the asexual genera Geotrichum and Saprochaete and the sexual genera Dipodascus, Galactomyces and Magnusiomyces are frequently associated with dairy and cosmetics production, fruit rot and human infection. However, the taxonomic system of these fungi has not been updated to accommodate the new nomenclature code adopting the "one fungus, one name" principle. Here, we performed phylogenetic analyses of these yeast-like species based on the sequences of the internal transcribed spacer (ITS) region and the D1/D2 domain of the large subunit of the rRNA gene. Two monophyletic groups were recognised from these species. One group contained Dipodascus, Galactomyces, and Geotrichum species and the other Magnusiomyces and Saprochaete species. We thus assigned the species in each group into one genus and selected the genus name Geotrichum for the first group and Magnusiomyces for the second one based on the principle of priority of publication. Five new Geotrichum species were identified from arthroconidial yeast strains recently isolated from various sources in China. The new species are described as Ge. dehoogii sp. nov., Ge. fujianense sp. nov., Ge. maricola sp. nov., Ge. smithiae sp. nov., and Ge. sinensis sp. nov.

Koumiss, a traditional fermented beverage made from mare’s milk, is typically consumed by nomads. Industrialized production of koumiss has been increasingly applied recently due to the increased demand for the beverage and awareness of its potential health benefits. However, it is unknown whether industrial koumiss is comparable to the traditional koumiss in terms of quality. In this study, we compared the microbiological and physicochemical properties in the industrial and traditional koumiss fermentation processes synchronously using culture-dependent and culture-independent approaches. Although Lactobacillus and Kazachstania species were similarly dominant in the bacterial and fungal communities, respectively, in both processes, the microbial counts and diversity in the traditional koumiss were significantly higher than those in the industrial koumiss. Furthermore, the traditional koumiss fermentation consumed more lactose, produced more flavor substances including acetic acid, lactic acid, ethanol, and free amino acids, and reached a lower pH value at the final stage. The physicochemical characters of traditional koumiss were mainly associated with Lactobacillus and Kazachstania species, which, in turn, were positively correlated with each other but negatively correlated with other non-dominant microbes. The starter was the major source of the microbial community of industrial koumiss, whereas both the starter and environment were the major sources of traditional koumiss. Random forest analysis recognized 11 significantly important genera as microbial indicators to distinguish industrial from traditional koumiss. Overall, this study shows that the microbial and physicochemical dynamics during the traditional and industrial fermentation of koumiss differ significantly, and the results obtained are valuable for improving the quality of industrial koumiss.

Flavor components largely depend on microbial activity and environmental conditions during traditional fermented food production. However, the microbial and abiotic contributions to the flavor of Chinese strong-aroma Baijiu (SAB) remain poorly understood. In this study, the composition and functional profiles of the fungal and bacterial communities changed significantly after fourteen days of grain fermentation. Acetilactobacillus jinshanensis, Issatchenkia orientalis and Kazachstani humilis became the dominant species as fermentation proceeded. Pathways related to lipid, protein and secondary metabolite metabolism were enriched during the middle and later stages of grain fermentation. Ethyl caproate, hexyl caproate, and ethyl lactate were identified as the main volatile components in fermented grain. Lactobacillus species were significantly related to volatile components. Compared with fungi, bacterial diversity accounted for 96 % of the variation in volatile components coupled with temperature, acidity and reducing sugar. This work provides insights into the production optimization and flavor enhancement of SAB by optimizing abiotic factors and microbial compositions.