Com ruou is a traditional Vietnamese fermented glutinous rice product widely consumed in the Mekong Delta. It is produced using artisanal fermentation starters (banh men) containing mixed microbial communities, including molds, yeasts and lactic acid bacteria (Thanh et al., 2024). The fermentation process remains largely empirical, leading to variability in product quality and safety. Previous studies on Com ruou microbiota have mainly relied on culture-based methods (Anh, 1999; Binh et al., 2015; Dung et al., 2007; Ha, 2012). However, only a small fraction (0.1-1%) of microorganisms can be cultivated under laboratory conditions (Xie et al., 2013), limiting comprehensive characterization of microbial diversity. As a result, the majority of microbial populations and their ecological dynamics during fermentation remain insufficiently understood.
       
Advances in next-generation sequencing (NGS) technologies have enabled culture-independent analysis of microbial communities (Xie et al., 2013). Amplicon sequencing of bacterial 16S rRNA and fungal ITS regions provides detailed insight into taxonomic composition and community shifts during fermentation. Such approaches have been successfully applied to various cereal- and rice-based fermented products. Understanding microbial succession in Com ruou is important for improving fermentation control, ensuring product safety and supporting process standardization while preserving traditional practices. However, information on temporal microbial changes during Com ruou fermentation remains limited. To our knowledge, this is the first study to simultaneously characterize both bacterial 16S rRNA and fungal ITS community dynamics across different stages of traditional Vietnamese Com ruou fermentation using amplicon sequencing, an approach that has also proven useful for investigating microbial succession in fermented food systems reported in recent ARCC journal publications (Nurye and Wolkero, 2023; Rani et al., 2025). Therefore, this study aimed to characterize bacterial and fungal community dynamics during Com ruou fermentation using amplicon sequencing, evaluate diversity changes across fermentation stages and provide preliminary microbiological insight relevant to process management.

Sample collection
 
Com ruou samples were collected in October 2025 from a traditional producer in Trung Thanh village, Co Do district, Can Tho City, Vietnam. The experimental work and laboratory analyses were carried out at Kien Giang University, Vietnam. Sampling was conducted at three fermentation stages: starter (0 h), fermented rice at 36 h and fermented rice at 72 h. Triplicate subsamples were collected at each time point and pooled prior to DNA extraction to obtain one composite sample representing each fermentation stage (Men, CR36H and CR72H). Samples were transported on ice to the laboratory for analysis.
 
Physicochemical and microbiological analyses
 
- The pH of each sample was measured using a calibrated benchtop pH meter (MI150, Martini Instruments, Romania). Measurements were performed in triplicate and expressed as mean ± standard deviation.
- Ethanol concentration was determined by gas chromatography- mass spectrometry (GC-MS) according to AOAC Official Method 972.11 and Vietnamese Standard TCVN 8010:2009. Ethanol content was quantified using an external calibration curve and expressed as percentage (v/v).
- Yeasts and molds were enumerated using the standard plate count method. Serial dilutions were plated on potato dextrose agar (PDA) supplemented with chloramphenicol (100 mg/L) and incubated at 28-30°C for 48 h. Results were expressed as colony-forming units per gram (CFU/g).
 
DNA extraction and amplicon sequencing
 
Genomic DNA was extracted using the QIAamp DNA Stool Mini Kit (Qiagen, Germany). DNA concentration was quantified with the Qubit® dsDNA HS Assay Kit (Invitrogen, USA).
       
Bacterial communities were profiled by amplifying the 16S rRNA V3-V4 region using primers 5'-ACTCCTA CGGGAGGCAGCAG-3' and 5'-GGACTACHVGGGTW TCTAAT-3'. Fungal communities were analyzed using ITS region primers 5'-GTGAATCATCGARTC-3' and 5'-TCCTCCGCTTATTGAT-3'. Amplicon libraries were prepared using the MetaVX™ Library Preparation Kit and validated by agarose gel electrophoresis. Library preparation and Illumina sequencing were performed by Phu Sa Genomics company, Can Tho City, Vietnam according to the MetaVX™ workflow using 2 x 250 bp paired-end reads.
 
Bioinformatic processing
 
Raw sequences were processed using QIIME v1.9.1 (Caporaso et al., 2010). Low-quality reads (length <200 bp, ambiguous bases, or average quality score <Q20) were removed. Chimeric sequences were identified and excluded using UCHIME against the GOLD database (Edgar et al., 2011). High-quality sequences were clustered into operational taxonomic units (OTUs) at 97% similarity using VSEARCH (Rognes et al., 2016). Taxonomic assignment was performed using the RDP Classifier (Wang et al., 2007), with a confidence threshold of 0.8 against the SILVA v138 database for bacterial 16S rRNA and the UNITE ITS database for fungal ITS rRNA.
 
Diversity and statistical analyses
 
Alpha diversity indices (Shannon, Simpson and Chao1) were calculated to assess microbial richness and evenness. To minimize bias caused by unequal sequencing depth among samples, OTU tables were rarefied by random subsampling prior to diversity analyses. Beta diversity was evaluated using Bray-Curtis dissimilarity and visualized by principal coordinates analysis (PCoA). Because triplicate subsamples were pooled prior to sequencing to generate one composite sample for each fermentation stage, diversity comparisons were interpreted descriptively across stages. Raw sequencing data were deposited in the NCBI Sequence Read Archive (SRA) under BioProject accession number PRJNA1372912.

Sequencing quality and data overview
 
Illumina sequencing generated approximately 0.30 million reads per sample with high quality metrics (Q20 >97.9%, Q30 >93%) (Table 1). GC content ranged from 53.75-54.26%, indicating consistent nucleotide composition across samples. The sequencing depth was sufficient for diversity and taxonomic analysis.


 
Community differentiation during fermentation
 
Genus-level Venn analysis showed 75 shared genera across all stages (Fig 1a), while the total number of detected genera increased from 94 (Men) to 129 (CR36H) and 130 (CR72H). This indicates progressive diversification during fermentation.
       
PCoA analysis based on Bray-Curtis dissimilarity revealed clear separation between the starter and fermented samples (Fig 1b, c). The starter formed a distinct cluster, whereas CR36H and CR72H grouped closely, suggesting that major community restructuring occurred within the first 36 h, followed by relative stabilization. Similar early-stage shifts have been reported in rice-based fermentations (Ehrmann et al., 2003; Zhao et al., 2022).
 
Alpha diversity changes
 
Alpha diversity increased during fermentation (Fig 1d). The Shannon index increased from 3.01 (Men) to 3.84 (CR36H) and slightly decreased to 3.62 (CR72H), indicating greater evenness during early fermentation. Similar trends were observed in fungal communities. These results suggest progressive restructuring of microbial diversity during fermentation (Zhang et al., 2016; Hughes et al., 2001).


 
Phylum-level shifts
 
At the bacterial phylum level (Fig 2a), Proteobacteria increased from 62.5% (Men) to 72.01% (CR72H), while Firmicutes decreased from 36.35% to 26.25%. Comparable transitions have been described in cereal and rice fermentations (Zhao and Eun, 2020; Zhao et al., 2019; Yan et al., 2024). For fungi (Fig 2b), Mucoromycota dominated the starter (88.14%) but declined during fermentation, whereas Ascomycota increased correspondingly.  This pattern reflects transition from mold-dominated starter to a mixed fungal community.


 
Genus-level dynamics
 
At the genus level, starter samples were characterized by Cronobacter (30.8%), Pantoea (18.92%), Weissella (16.66%) and Lactococcus (12.68%) (Fig 3a). During fermentation, lactic acid bacteria declined, while Pantoea, Acinetobacter and Bacillus increased and became dominant by 72 h (Pantoea 37.97%; Acinetobacter 17.8%; Bacillus 14.67%). Similar shifts from LAB-dominated communities to Proteobacteria- and Bacillus-enriched populations have been observed in traditional fermentations (Yang et al., 2021). In the fungal community, Rhizopus dominated the starter and declined during fermentation, whereas yeasts including Kodamaea and Wickerhamomyces increased in fermented samples (Fig 3b). Saccharomyces remained at low relative abundance.


 
Integrative interpretation
 
Com ruou fermentation showed a clear early compositional shift (≤36 h), transitioning from LAB-dominated starters (Weissella, Lactococcus; Firmicutes) toward Proteobacteria- rich communities (Pantoea, Acinetobacter) and Bacillus at later stages. This succession coincided with ethanol accumulation (0 to 4.6% v/v) and pH reduction (to 4.11), indicating progressive establishment of an alcoholic fermentation environment. Compared with traditionally matured products (8-12% ethanol), the 72 h stage represents an intermediate yet biologically active phase, with implications for starter control and optimized fermentation windows.
       
From a safety perspective, Cronobacter malonaticus was abundant in the starter but declined markedly during fermentation, likely due to acidification and ethanol stress, suggesting low risk in the final product while emphasizing hygienic starter preparation. Acinetobacter increased at later stages; although some species are opportunistic pathogens, food-associated lineages are genetically distinct from clinical strains and no acute safety concern was indicated, though continued monitoring is advisable.
       
Genus-level shifts were interpreted descriptively, as differential abundance testing across batches was not performed and thus statistical robustness warrants further validation. The fungal transition from Rhizopus arrhizus-dominated starters to yeast-enriched stages (Kodamaea ohmeri, Wickerhamomyces anomalus) reflects functional succession, though amplicon data provide relative rather than absolute activity. Finally, sampling from a single artisanal batch limits generalizability, as microbial profiles may vary among producers and seasons.
       
The increase in Pantoea during fermentation may reflect its adaptation to the changing rice fermentation environment and its association with plant-derived substrates or artisanal starter microbiota. Although its functional contribution could not be confirmed from amplicon data alone, its temporal increase suggests ecological relevance during community succession. In contrast, Acinetobacter requires careful interpretation from a food safety perspective. While some members of this genus are recognized as opportunistic pathogens, detection by amplicon sequencing does not by itself indicate pathogenicity. Nevertheless, its increased abundance at later fermentation stages highlights the importance of hygienic starter preparation and continued microbiological monitoring in traditional fermentation systems, in agreement with observations from related fermented-food studies reported in ARCC journals (Angadi et al., 2021; Ramasamy et al., 2025).

Amplicon sequencing showed stage-dependent microbial shifts during Com ruou fermentation. The starter was dominated by lactic acid bacteria, whereas later stages were enriched in Proteobacteria and Bacillus, coinciding with pH decrease and ethanol increase. At the genus level, Pantoea, Acinetobacter and Bacillus increased during fermentation. The fungal community shifted from Rhizopus arrhizus dominance in the starter to increased abundance of yeasts such as Kodamaea ohmeri and Wickerhamomyces  anomalus. Viable molds and yeasts remained detectable at 72 h with ethanol at 4.595% (v/v), indicating continued fermentative activity. These findings describe temporal microbial trends during Com ruou fermentation.

The present study was supported by Kien Giang University. The authors acknowledge Kien Giang University for providing funding for this project.
 
Disclaimers
 
The views and conclusions expressed in this article are solely those of the authors and do not necessarily represent the views of their affiliated institutions. The authors are responsible for the accuracy and completeness of the information provided and do not accept liability for any direct or indirect losses arising from the use of this content.
 
Informed consent
 
Not applicable. This study did not involve human participants or animal experiments. The research was conducted using traditional fermented rice samples for microbial community analysis.

The authors declare that they have no conflicts of interest regarding the publication of this article. No funding or sponsorship influenced the study design, data collection, analysis, interpretation of results, decision to publish, or preparation of the manuscript.