Species-specific microbial dynamics are associated with molecular compositional restructuring of free metabolites during Light-flavor Baijiu fermentation

Microbial succession during Jiupei fermentation is closely associated with biochemical changes that contribute to Baijiu quality formation; however, the fermentation-induced restructuring of the non-volatile metabolome and its species-level microbial associations remain poorly understood. We hypothesized that species-level microbial succession would be associated with pathway-specific changes in the non-volatile metabolome, and that bacterial and fungal diversity would show contrasting association patterns with different metabolite classes during fermentation. To test these hypotheses, we integrated physicochemical analysis, PacBio SMRT sequencing, and untargeted metabolomics to characterize microbial succession and metabolite dynamics during Light-flavor Baijiu Jiupei fermentation. A total of 1392 peaks were detected, with 134 metabolites annotated, including amino acids, organic acids, carbohydrates, alcohols, and nucleotides. Fermentation was accompanied by marked temporal shifts in the non-volatile metabolite profile, with galactose metabolism and starch and sucrose metabolism identified as the main enriched pathways. Fungal diversity (Shannon index) showed predominantly positive correlations with many metabolites across multiple classes, whereas bacterial diversity showed predominantly negative correlations with most metabolites. At species level, Lactobacillus homohiochii showed the largest number of positive correlations with amino acids (14 edges), alcohols (8 edges), and organic acids (7 edges), while Saccharomyces sp. was positively correlated with 19 amino acids and derivatives. These species-specific correlation patterns provide a correlation-based view of microbe-metabolite dynamics during Jiupei fermentation and generate testable hypotheses for future culture-based and mechanistic studies.