D.Xu et al Food Microbiology 86 (2020)103340 A A48 First axis 0 1.0 First axis CQ03 isolated from soy auce mo mi exhibited strong capability in ongfprotcaseand tasting amino To the air or and arch Bacillus species that mixture But the bacterial composition was not era th that may ing (Gr lated during th ole pro reshold,the cons this study pro de dra s are also at each ferm tion stage in sufu.The int on b for optim ing the mi tructure of sufu f tion stages and improving produ stud does not nce of oppo contan origineated n sufu quality still un and unstab ence the work reported in this paper. dcdnodefdrstngmixtesreq Acknowledgements The National Key R&D Program of China (2016YFD0400500)and ure the aphy Appendix A.Supplementary data 5.Conclusions In this study.we measured the physicochemical properties of theCQ03 isolated from soy sauce moromi exhibited strong capability in secretion of protease and produced more umami tasting amino acids than other isolates (Hu et al., 2017). Similar like Streptococcus, the re￾lative abundance of Weissella were much higher in the pehtze stages than other stages. This was consistent with increased level of amino nitrogen. The high abundance of Streptococcus and Weissella during the pehtze stages may work together to increase the amino nitrogen con￾tent. Bacillus contain species that are important in food industries and research, Bacillus species produce a number of volatile compounds and play a key role in fermented foods (Li et al., 2019; Tamang et al., 2016). But some species of Bacillus are common pathogens, especially B. cereus, that may cause food poisoning (Granum, 1994). Studies have shown that when B. cereus levels were significantly lower than the toxic threshold, the hazard for consumers were reduced (Andersson et al., 1995; Han et al., 2004). Our study showed that in M3, the abundance of Bacillus dramatically declined to 0.07%. Bacillus species are also re￾nowned for their strong amylase and protease activities (Wei et al., 2013) that partially explains its reduced abundance due to the de￾creased level of starch and protein as fermentation progressed. The previous study showed that though sufu product does not include pasteurization and the pre-fermentation environment is open to air, sufu is a relatively safe product (Shi and Fung, 2000). But the constant presence of opportunistic contaminants are still food safety risks. The first appearance of Bacillus was right after D0, which implied the Ba￾cillus was originated from dressing mixture. Bacillus species were abundant in the first few days of the post-fermentation, this instability made sufu quality still unpredictable and unstable. As the demand of safety and quality increased, suppressing or wiping out the potential pathogen is vital for sufu manufacturing. To increase stability of sufu quality, higher standard of dressing mixture is required. Also, further studies are needed to investigate the specific strains of Bacillus to nur￾ture the functional ones and target killing the B. cereus. Besides Bacillus, studies showed that other common food-borne pathogens including Escherichia coli O157:H7, Salmonella typhimurium, Staphylococcus aureus, and Listeria monocytogenes during sufu fermentation and aging were under the threshold levels. All of these were decreased to non-detect￾able levels after the few months of fermentation (Shi and Fung, 2000). 5. Conclusions In this study, we measured the physicochemical properties of the sufu as well as the fungal and bacterial composition during the whole fermentation process. To the best of our knowledge, this is the first study that explores the microbial successions of sufu in a period of more than 3 months using high-throughput sequencing technology. Our study produced a dataset with sequencing data from 90 samples ob￾tained at 9 different fermentation stages. Trichosporon, Actinomucor and Cryptococcus were the main genera in pre-fermentation while Monascus and Aspergillus were dominant in post-fermentation. This huge shift in fungal composition was correlated with process procedure of pouring dressing mixture. But the bacterial composition was not greatly changed after pouring dressing mixture, the Acinetobacter and Enterobacter were the predominant genera throughout the whole pro￾cess. As no bacterial starter was inoculated during the whole process, this relatively stable bacterial structure may relate to fermentation environment. In conclusion, this study provides a better understanding of the dynamic changes in the microbial community and dominant microbe at each fermentation stage in sufu. The interaction between microbial structure and processing procedures were also discussed and considered, which provides theoretical support for optimizing the mi￾crobial structure of sufu fermentation stages and improving product quality. Declaration of interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ￾ence the work reported in this paper. Acknowledgements The National Key R&D Program of China (2016YFD0400500) and Beijing Postdoctoral Research Foundation (2018-ZZ-120) supported this work. Thanks Dr. Guo Zhuang for his instructive advices and kind help with sequencing data analysis. Thanks to Mr. Austin James Faust and his help with the language editing. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.fm.2019.103340. Fig. 6. Bioplot of RDA for qualitative variables between time spots S and M3 (A), or A48 and S (B). RDA = Redundancy analysis. Pehtze which inoculated with A. elegans for 48 h (A48), salt-pehtze (S), fermentation of sufu for 3 months (M3). D. Xu, et al. Food Microbiology 86 (2020) 103340 8