Microbial co-occurrence networks are associated with key flavor compound accumulation in medium- and high-temperature Daqu for strong-flavor Baijiu

Daqu is a key saccharifying and fermenting starter shaping the flavor quality of strong-flavor Baijiu, yet the association between microbial co-occurrence patterns and flavor compound accumulation remains unclear. In this study, ten finished medium- and high-temperature Daqu sample groups from ten selected enterprises in Yibin, each represented by three independent composite samples, were analyzed using high-throughput sequencing, physicochemical analysis, volatile profiling, co-occurrence network analysis, path coefficient analysis , and Mantel tests. Firmicutes, Actinobacteriota, and Proteobacteria dominated the bacterial community, whereas Ascomycota dominated the fungal community. Lactobacillus , Weissella , Bacillus , Aspergillus , and Thermomyces were major genera potentially related to Daqu fermentation, and Wickerhamomyces was identified as a putative keystone genus based on network topology. The sample groups were classified into high-, medium-, and low-interaction groups according to co-occurrence distance. Shorter co-occurrence distance was associated with higher ester and pyrazine accumulation. Path coefficient analysis showed negative associations between co-occurrence distance and esters/pyrazines, with coefficients of −0.962 and −0.726, respectively ( P < 0.05), while Mantel tests supported significant taxon–flavor associations, especially in the low-interaction group. These findings suggest that microbial co-occurrence patterns are closely associated with flavor accumulation in Daqu.