**4. Discussion**

Overall, our results show that, despite the warmer and more rapid temperature fermentation process used to produce the sauerkraut analyzed here, the bacterial community is in line with that of more traditional, colder fermented cabbage products. Over the first 48 hours of fermentation, the microbial community of sauerkraut experienced a precipitous drop in the number of bacterial taxa present, likely due to the strong selective pressures of high salinity and acidity in the fermentation environment. Over the remainder of the fermentation period, LAB remained the dominant organisms present in the community. Both patterns are indicative of successful fermentation.

— — Perhaps more surprising were the relationships between the microbial communities of the starting ingredients, the fermentation environment, and the fermenting sauerkraut. The major LAB found in fermenting sauerkraut were present only in extremely low levels in the starting ingredients, which may suggest that only trace amounts of LAB are necessary to initiate fermentation. It is also possible that the abundance of fermentative sauerkraut LAB found around the production facility—especially in the air—might contribute to the inception of the fermentative community, acting as a starter culture. The presence of LAB in the environment may also be a direct result of sauerkraut being fermented within it. These hypotheses require further investigation.

Previous studies have used culture- and sequencing-based methods to elucidate the fermentative microbial community of sauerkraut. Culture-based methods have shown that the major LAB involved in sauerkraut fermentation are *E. faecalis*, *L. mesenteroides*, *L. brevis*, *P. cerevisiae*, and *L. plantarum;* while sequencing-based methods highlight the *Lactobacillus* and *Leuconostoc* species in addition to *Weissella* [11,13]. Our results using 16S rRNA sequencing paralleled these expectations and expanded on previous knowledge, identifying *Leuconostoc, Lactobacillus*, and Enterobacteriaceae in addition to a variety of LAB not previously detected, such as *Lactococcus*.

– *cà muối* Our results are also in line with the canonical microbial communities of other fermented vegetable foods. Xiong et al. found that *Lactobacillus* and *Leuconostoc* species were the primary bacteria in the fermentation of Chinese sauerkraut, *pàocài* [20]. Numerous studies have shown that the kimchi bacterial community is dominated by *Weisella, Lactobacillus,* and *Leuconostoc* species [21–23]. A study of traditional Vietnamese fermented vegetables, such as mustard and beet ferment (*dua muoi*) and fermented eggplant (*cà muối*), found a predominance of *Lactobacillus* species in fermentation [24].

Our results suggest that warmer and more rapid production can yield fermented sauerkraut with a similar microbial community to sauerkraut produced by traditional fermentation methods. This may mean that a quick-fermented process is a viable option for industrial production of fermented cabbage foods. This may be of interest to commercial producers, as it would allow them to speed and scale-up production without sacrificing the integrity of the fermentative bacterial community, which is central to the purported probiotic benefits of sauerkraut and other fermented foods.

While the analyzed communities were roughly similar to previously published sauerkraut data, we cannot yet claim that the products are identical or that the production processes are interchangeable. There are multiple metrics—physical, sensory, and nutritive—that were not investigated as part of this study and could possibly vary between the two types of sauerkraut. We anticipate that diminished appearance, shelf life, taste, and nutritive value of warmer fermented sauerkraut could negatively impact its commercial viability. Therefore, additional studies and measurements of these qualities are required before widespread commercial implementation of this fermentation technique.

**Author Contributions:** M.A.Z. and P.B. jointly designed this study and collected all samples from the production facility. D.J.C, J.I.W., and M.A.Z. extracted and prepared DNA for sequencing. W.H.S. and D.J.C. performed all analyses. M.A.Z., W.H.S., D.J.C., J.I.W., and P.B. prepared the manuscript. All authors reviewed and approved its final version.

**Acknowledgments:** Research activities associated with this study were funded in part by the COBRE Center for Computational Biology of Human Disease (NIH P20 GM109035). Sequencing was conducted at the Rhode Island Genomics and Sequencing Center, a Rhode Island NSF EPSCoR research facility supported in part by the NSF ESPCoR Cooperative Agreement #EPS-1004057. M.A.Z was supported by a Karen T. Romer Undergraduate Teaching and Research Fellowship, and D.J.C. was supported by the National Science Foundation Graduate Research Fellowship (Grant No. 1644760).

**Conflicts of Interest:** The authors declare no conflict of interest.
