Abstract: In order to clarify the succession pattern and potential functions of the microbial communities in fermented grains for Nongxiangxing baijiu, high-throughput sequencing was used to investigate the dynamic change of microbial community structure, and the correlations between microbial community structure and physicochemical indicators were analyzed at different fermentation stages. Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) was used to predict the potential functions of microbial communities in fermented grains. The results showed that Firmicutes, Proteobacteria, Actinobacteriota, Ascomycota, and Basidiomycota were the dominant phyla in fermented grains. The dominant genera were Lactobacillus, Bacillus, Staphylococcus, Thermoactinomyces, Kazachstania, Wickerhamomyces, Penicillium, Candida, Aspergillus, Meyerozyma, Debaryomyces, with Lactobacillus, Kazachstania and Wickerhamomyces being the absolutely dominant ones. The contents of reducing sugar and starch showed a significantly positive correlation with 8 bacterial genera (P < 0.05). Acidity and moisture content showed a significantly negative correlation with 10 bacterial genera. Some fungi were significantly correlated with physicochemical indicators. PICRUSt analysis showed that the relative abundance of functional enzymes was significantly different at different fermentation stages, and the dominant microorganisms in fermented grains participated in glycolysis, acid metabolism and ethanol synthesis to form a complete metabolic chain, indicating that the microbial community succession of fermented grains could be the major reason for the differences in metabolism at different fermentation stages. The findings of this study would provide a theoretical basis for the study of the metabolic mechanisms of microorganisms used for baijiu brewing.

Key words: Nongxiangxing baijiu; fermented grains; microbial community; correlation analysis