Revealing microbial community composition and disinfection efficacy evaluation in wholesale pork markets through 16S rDNA sequencing

ABSTRACT

This study has employed 16S rDNA analysis to elucidate the characteristics of the microbial community in the surroundings, facilities, and meat samples collected from a pork wholesale market with the aim of revealing bacterial contamination. In addition, the efficacy of sodium hypochlorite disinfectant was assessed by comparing the bacterial communities in the viable samples before and after disinfection. The bacterial abundances in meat, chopping boards, and knives were positively correlated (r > 0.95, p < .05), suggesting potential cross-contamination risks during market transactions. The species richness indices for the external transport vehicles, pork hooks, and floors were significantly different from those of the samples taken after decontamination (p < .05). Linear discriminant analysis Effect Size (LEfSe) analyses showed that Firmicutes, Proteobacteria, and Chloroflexi were significantly more abundant in the transaction period, whereas Bacteroidetes levels decreased compared to the clean-up period. NaClO disinfection altered the composition of bacterial communities.

KEYWORDS:

• Pork
• wholesaler
• 16S rDNA amplicon sequencing
• bacterial contamination
• NaClO disinfection

1. Introduction

Pork is a popular meat in China because of its excellent nutritional content and flavour, accounting for 62.9% of total meat consumption in 2018 (Zhang et al., 2021). However, pork is the most frequently contaminated food, contributing to serious meat safety concerns (Wilson et al., 2020; Wu et al., 2021). It is possible that this microbial contamination originated from the farms, or in the abattoirs during processing for retail sale (Demaître et al., 2020; Kang et al., 2014; Vojkovska et al., 2015). For example, a recent survey of retail markets indicated that Salmonella contamination was present in pork (73.1%), with 30% of these samples also contained Escherichia coli (Zhang et al., 2018). Pseudomonas spp., Brochothrix thermosphacta, and lactic acid bacteria have also been detected in retail pork at levels ranging 5.3 to 6.4 log CFU/g (Andritsos et al., 2012). 16S rDNA amplicon sequencing has also identified Staphylococcus aureus, Pseudomonas spp., and E. coli as the dominant contaminants in retail pork (Peruzy et al., 2020). These previous studies focused on the retail supply chain; however, the degree of microbial contamination at the wholesale market level has not been extensively investigated.

In China, 70–80% of food is distributed through wholesale markets, which is a critical sector which food safety risks can be controlled (Xiao et al., 2021; Zhou et al., 2022). The meat wholesale market handles pork portioning and distribution and connects the slaughterhouse and grocery store. Hog carcasses are split in half in slaughterhouses and transported to wholesale markets using cold chain vehicles. Once the market opens, carcasses are transported by carts from vehicles to stalls, where they are portioned and sold to retailers. Usually, pork wholesale markets conduct operations in regard to trading from 12 am to 8 am and they disinfect floors with sodium hypochlorite (NaClO) after closing (~9 am). NaClO is recommended by the World Health Organization (WHO) for disinfecting surfaces owing to its broad-spectrum antimicrobial activity and is widely used in food production facilities (Sato et al., 2019). However, whether bacterial contamination associated with meat transaction facilities is altered following NaClO decontamination has not yet been evaluated. This information enables the identification of potential microbial risks in pork products.

Although traditional bacterial culture and plate methods are the gold standards for microbiological diagnosis, these approaches generally focus on mapping the distribution of specific environmentally transmitted pathogenic bacteria, such as Salmonella, E. coli, S. aureus (Cobo Díaz et al., 2021; O’Brien, 2002). Traditional bacterial culture and plate methods are time-consuming, largely underestimate the abundance of species, and have limitations in characterizing complex microbial communities (Cobo Díaz et al., 2021; Peruzy et al., 2020). High-throughput sequencing is a technique widely used to assess microbial community structures in different environments (S. Xu et al., 2016) and enables the identification of organisms belonging to previously unknown taxa. For example, 16S rDNA gene sequencing has been used to characterize the microbial composition of pork during storage (Bassey et al., 2021; Li et al., 2019; Zhao et al., 2022).

The focus of the current study was to assess the effect of disinfection with NaClO on the bacterial species composition using 16S rDNA amplicon sequencing. These results not only assess the effectiveness of decontamination but also expand our knowledge on the routes of microbial transmission.

2. Materials and methods

2.1. Sampling

Samples were collected from a pork wholesale market in Yuhang District, Hangzhou, China, which processes approximately 10,000 pigs daily. In this market, regular disinfection operations were implemented following operational closure. This involved spraying a NaClO solution with a chlorine concentration of 500 mg/L over the surfaces of the surroundings and facilities. In total, 104 samples were collected from different sites during daily market activities (Table 1). A total of 62 samples were collected during the market transaction period at 2 a.m. and were obtained from external transport vehicles (BEV, n = 6), local transport vehicles (NLV, n = 6), carts (NC, n = 12), pork hooks (NPK, n = 3), meat (NM, n = 10), knives (NK, n = 10), chopping boards (NCB, n = 10), and floors (NF, n = 5). Disinfection was performed after the market closed at 11 a.m., 42 disinfected samples were collected from external transport vehicles (DEV, n = 6), local transport vehicles (DLV, n = 6), carts (DC, n = 12), pig hooks (DPK, n = 3), chopping boards (DCB, n = 10), and floors (DF, n = 5). We attempted to gather samples before and after the disinfection process, except for meat and knives, which were only present during pork transactions. These samples were examined to determine the attributes of the bacterial communities. Samples were collected using a sterile absorbent gauze pre-moistened with phosphate -buffered saline (1 × PBS, pH 7.2). When a surface with a sufficient size was available (e.g. chopping boards), ~ 1 m² was sampled by swabbing first horizontally, then vertically, and finally by diagonally turning the swab around each direction. For other surfaces, such as knives, individual units were swabbed. Single-use disposable gloves were used during sampling and the gloves were changed after each sample was collected to avoid cross-contamination. Sample swabs were placed in tubes containing 15 mL PBS, placed in a cooling box containing ice packs, and transported to the laboratory within 2 h for sample processing.

Table 1. Sample information in the pork wholesale market.

Sampling time	Sampling sites	Symbol	No. of samples	ASV	Reads
Night	External transport vehicle	NEV	6	1142	48513
	Local transport vehicle	NLV	6	1192	48451
	Carts	NC	12	1372	50977
	Pork hook	NPH	3	679	39739
	Meat	NM	10	966	53194
	Chopping board	NCB	10	976	51945
	Knives	NK	10	1018	51633
	Floor	NF	5	833	53888
	Total	—	62	5019	398340
FDay	External transport vehicle	DEV	6	1207	51817
	Local transport vehicle	DLV	6	1259	49768
	Carts	DC	12	1154	52642
	Pork hook	DPH	3	496	48098
	Chopping board	DCB	10	894	50612
	Floor	DF	5	315	43909
	Total	—	42	3557	296846

2.2. DNA extraction

Swab samples were individually added to tubes containing 15 mL PBS and centrifuged at 10,000 rpm for 5 min. The supernatants were filtered through 0.22 μm polycarbonate membranes (GE Osmonics, Minnetonka, MN, U.S.A.) and stored at −80°C. DNA was extracted using a TIANamp Stool DNA Kit (Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. Final DNA concentration and purity were measured using a Nanodrop One spectrophotometer (Thermo Fisher, Pittsburgh, PA, U.S.A.), and DNA quality was confirmed by 1% agarose gel electrophoresis. Each sample was processed in duplicate and stored at −20°C until Polymerase Chain Reaction (PCR) analysis.

2.3. PCR amplification and sequencing

Polymerase chain reaction (PCR) was performed by amplifying the V3-V4 hypervariable region of 16S rDNA using the universal bacterial-specific primers 16S-341F (5’-CCTAYGGGRBGCASCAG-3’) and 16S-806 R (5’-GGACTACNNGGGTWTCTAAT-3’), where the barcode is an eight-base sequence unique to each sample. The cycling reaction parameters were as follows: 5 min at 95°C and 27 cycles of 30 s at 95°C, 30 s at 55°C, 45 s at 72°C and a final extension for 10 min at 72°C. PCR amplification was performed in a total volume of 25 μL of the reaction mixture, containing 25 ng of template DNA, 12.5 μL of PCR premix, 2.5 μL of each primer and PCR-grade water to adjust the volume. Amplification was carried out using an ABI GeneAmp 9700 instrument (Thermo Scientific) with FastPfu polymerase (Transgen Biotech, Beijing, China) and 10 ng of template DNA, according to the manufacturer’s protocol. The pooled PCR products were purified by 2% agarose gel electrophoresis. Bands corresponding to 300 bp were excised and purified using the GeneJET Gel Extraction Kit (Thermo Fisher Scientific). The purified amplified fragments were subjected to equimolar and paired-end sequencing (2 × 250) using an Illumina PE250 platform (Shanghai, China). Raw data were uploaded to the NCBI BioProject database and we obtained the following ID number: PRJNA823853.

2.4. Sequence data analysis

Raw data generated from 16S rDNA MiSeq sequencing were analyzed by Biozeron Biotechnology (Shanghai, China). Amplicon sequence quality control was performed using the Divisive Amplicon Denoising Algorithm2 (DADA2) to identify indel mutations and substitutions, which were processed with QIIME2 (ver. 2020.11) (Callahan et al., 2016). Amplicon sequence variants (ASVs) were generated using paired sequence merging and chimera filtering. To obtain the species taxonomic information corresponding to each ASV, high-performance clustering, alignment, and search algorithms (UCLUST, ver. 1.2.22q) was used to analyze the phylogenetic affinities of representative ASV sequences in the Silva (SSU132) database, with a confidence threshold of 0.8. Rarefaction analysis, based on Mothur v.1.35.1 was conducted to reveal the diversity. Alpha and beta diversities were calculated by random normalization of the same sequences. The Chao1 index was selected for the alpha diversity analysis, and the complexity of species diversity was calculated using QIIME (De Filippis et al., 2018). Principal coordinate analysis (PCoA) was used in order to explore the effects of variable factors, and ADONIS was used to measure the Bray-Curtis distances. Phyla and ASVs and relative abundance levels > 1% and 0.1%, respectively, were defined as predominant and sorted for further comparison.

Linear discriminant analysis Effect Size (LEfSe) analysis was used to determine significant differences in the relative abundance of bacteria before and after disinfection. LEfSe analysis of microbial abundance was performed by pooling all samples between the N (transaction period) and D (clean-up period) groups to differentiate microbial communities. Additionally, LEfSe analysis was performed at different locations before and after the disinfection was performed. The chi-square and Kruskal-Wallis rank-sum tests were used to investigate the differences resulting from the distribution of multiple populations. Linear discriminant analysis (LDA) effect size (LEfSe) was applied to discover the biomarker panels that possessed organismal features for differentiating the microbial communities specific to a particular treatment (Li et al., 2020).

Microbial functions were predicted based on 16S rDNA sequences using PICRUSt2 (Langille et al., 2013). The Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2, v2.2.0-b) (https://github.com/picrust/picrust2) program based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to predict the functional pathways used by the microbiota in different samples. The representative ASVs in FASTA format and a Biological Observation Matrix (BIOM) table of the abundance of each ASV across each sample were used as inputs for PICRUSt2. A BIOM table was generated using the Python package, Biom (v.2.1.7). The minimum mapping rate of ASV representative sequences to representative sequences in the KEGG database was set to 0.6, a value below this meant they did not participate in the functional prediction. Functional differences among cultivars were analyzed using the Kruskal-Wallis H test in the R project Vegan package. STAMP software (v2.1.3) was used for statistical analyses and visualization of the identified pathways.

2.5. Statistical analysis

GraphPad Prism 8 software (San Diego, CA, U.S.A.) was used for data analysis and generation of the figures. A significant difference was established at p < .05. Correlation analyses were performed using the Pearson correlation coefficient, calculated with the rcorr function in the R package Hmisc.

3. Results

3.1. Taxonomic profiles for sample surfaces

A total of 6,362,065 high-quality validated gene sequences clustered into 14,780 ASVs were obtained from the environmental and meat samples. The total numbers of ASVs and reads were higher during market transactions, particularly for the external transport vehicles, local transport vehicles, and carts. The number of reads in the floor samples decreased after disinfection was performed (Table 1). Sparse curves indicate sufficient sequencing data and species abundances in the sample. As shown in Figure S1, the dilution curves flattened with increasing sequencing depth, suggesting that the amount of sequencing data was reasonable.

The microbiota on the surfaces of the different samples varied, with 31 phyla, 75 classes, 179 orders, 287 families, 749 genera, and 1798 species. The stacked bar chart shows the dominant phyla and genera in each sample. At the phylum level, the primary phyla identified were Proteobacteria (56.9%), Bacteroidota (17.7%), and Actinobacteria (13.1%), with different distributions on the various sample surfaces. Proteobacteria, Bacteroidota, Actinobacteriota, and Firmicutes were the dominant phyla in all samples; interestingly, Deinococcota was dominant only in vehicle samples (Figure 1a). At the genus level, we also identified 30 genera as the most abundant colonizers; Psychrobacter, Acinetobacter, and Chryseobacterium possessed average abundance levels of 29.7, 13.6 and 6.4%, respectively (Figure 1b). Our findings are consistent with the literature (Li et al., 2019; Tassou et al., 2021) in that Acinetobacter is the most commonly found bacterium in pork samples (Emamjomeh et al., 2023).

Figure 1. Relative abundance (%) of the bacterial community at (a) phylum and (b) genus level in the pork wholesale market. The multi-colored stack bar graphs display the relative abundance of bacteria in each sample. Night and day represent before and after disinfection samples, respectively.

3.2. Differences in taxonomic profiles following NaClO decontamination

Alpha-diversity analyses showed that the greatest species abundance was found in transport vehicles, whereas samples from the internal environment of the wholesale market, such as meat, chopping boards, and knife surfaces, had relatively small species abundance. As expected, the average values of the species richness indices for the external transport vehicles, pork hooks, and floors were significantly different from those of the samples taken after decontamination (p < .05). Floor samples showed the lowest alpha diversity after the decontamination was performed (Figure 2a). Bray – Curtis distance-based PCoA plots were generated in order to examine the differences in microbial community structure before and after disinfection of these samples. There were no significant differences in the microbiological composition between the meat and knife samples (p > .05), indicating potential cross-contamination (Figure 2b). The disinfection process had a significant effect on the microbiological composition of external transport vehicles, local transport vehicles, carts, pork hooks, chopping boards, and floors (Figure 2c–h). The vehicle samples were significantly separated from the other samples in coordinate space, indicating significant differences in their microbial community structures (Figures S2 and S3). These results indicated that the NaClO disinfectant was effective in altering the taxonomic profiles of the microorganisms in the pork wholesale market in terms of abundance and composition.

Figure 2. Analysis of the alpha diversity and beta diversity, (a) Chao1 index of bacterial flora in each sampling site. (b) PCoA analysis of meat and knife samples, (c) external transport vehicles, (d) local transport vehicles, (e) carts, (f) pork hook, (g) chopping boards, and(H) floor samples for the two sampling periods.

Differences in bacterial composition between groups N and D were identified using LEfSe analyses, which showed that bacteria were classified into four distinct taxa at the phylum level; Firmicutes, Proteobacteria, and Chloroflexi were found to be significantly more abundant in the N (transaction period) group, whereas Bacteroidetes levels decreased when compared to those in the D (clean-up period) group (Figure 3a). When a logarithmic LDA score of 3.0 was used as the cut-off point, 23 differential genera were identified, with Escherichia-Shigella being significantly higher in group D and Methylobacterium being the most abundant genus in group N (Figure 3b). Listeriaceae were found to be abundant on the chopping board, and Enterobacterales made a large contribution to the floor samples (Figure S4-S9). In addition, we counted pathogen- and spoilage-associated species at the genus level. The presence of potential pathogens and spoilage-associated bacteria identified at the genus level differed between the meat trading period (37.65%) and following decontamination (1.13%). Meat samples contained the highest abundance levels of pathogen contamination (12.9%), and knives contained the highest abundance levels of spoilage-associated bacteria (0.24%). The relative abundance levels of potential pathogens and spoilage-associated bacteria on floors were decreased from 0.01 to 0% and 0.37 to 0.15%, respectively, after market decontamination (Figure 4a,c). The criteria used to designate specific microbes as pathogens or spoilage-causing bacteria (Figure 4a,c) were based on previous studies (Table 2). At the species level, the pathogenic and spoilage bacteria of concern (Table 2) indicated a high abundance of Brochothrix thermosphacta, Pseudomonas fragi, E. coli, and S. aureus in the samples taken during market transactions. These became less dominant after the market closed, and the subsequent decontamination occurred (Figure 4b,d). B. thermosphacta and P. fragi were the most prevalent spoilage bacteria on the knives when the market was on trading (Figure 4d).

Figure 3. LEfSe analysis diagram of each dominant taxa between the N group (before disinfection) and the D groups (after disinfection). (a) Cladogram of microbial communities (b) LDA score of size differentiation using a threshold of 3 in groups D and N.

Figure 4. Distribution of pathogenic and spoilage bacteria on different sample surfaces. Percentage of abundance of (a) pathogenic and (c) spoilage bacteria at the genus level. The most abundant species of (b) pathogenic and (d) spoilage bacteria.

Table 2. The cardinal symptoms and corruption phenomena caused by potential pathogens and spoilage-associated bacteria.

	Species	Cardinal symptom/corruption phenomenon	References
Potential pathogens	Escherichia coli	Urinary tract infection, diarrhea	(Velez et al., 2021)
	CampylobacterUnclassified	Nausea and vomiting	(Brooks et al., 2004; Tong, Ma, et al., 2021)
	Staphylococcus aureus	Diarrhea and vomiting	(Mahros et al., 2021)
	Streptococcus.	Meningitis, and septicaemia	(Cheung et al., 2008; Lim et al., 2019)
Spoilage-associated bacteria	Brochothrix thermosphacta	Tolerance to high salts and low-pH; produce unpleasant compounds.	(Illikoud et al., 2019)
	Pseudomonas fragi	Slime formation	(Ercolini et al., 2010; Wickramasinghe et al., 2021)
	Lactobacillus sakei	Produce acid and gas, and under low-oxygen and acidic conditions	(Correia Peres Costa et al., 2019)
	Weissella viridescens	Produce some volatile spoilage compounds, sulfurous, rancid and fishy off-odors	(Kameník et al., 2015)
	Shewanella putrefaciens	Produce some volatile spoilage compounds and musty off-odor	(Yi et al., 2022)

3.3. Bacterial cross-contamination in the pork wholesale market

We examined the cross-contamination profiles of each sample surface, as shown in Figure 5a, Psychrobacter was higher at group N than group D samples, especially at meat and chopping board surfaces, where the relative abundance of Psychrobacter was significantly higher than at other sample surfaces (Figure 5a). Correlations between bacterial contamination during the two sampling periods were examined using Spearman’s correlation analysis. Bacterial abundance in meat was highly positively correlated with chopping boards (r = 0.99, p < .05), suggesting that these tow sample types were potential sources of bacterial cross-contamination. During the clean-up period, the correlation between DLV, DC, and DPH was only found to be high in Group D (r > 0.7) and it decreased significantly in the remaining samples. The abundance of bacteria in the floor samples showed only a low correlation with other samples, indicating that NaClO was an effective decontaminant. However, targeted disinfection measures may not be sufficient (Figure 5b).

Figure 5. (a) Heatmap of bacterial abundance on the surfaces. Correlation analysis of bacterial taxons over all of the samples. (b) Pearson correlation values are indicated within the cells.

3.4. Microbial function prediction

PICRUSt2 was used to obtain the relative KEGG pathway abundance information derived from 16S rDNA amplicon sequencing data, which revealed several pathways were significantly different in terms of NaClO disinfection (Group D) and without disinfection (Group N) (Figure 6a). RNA degradation (tadA) and glycosphingolipid biosynthesis (gloB) were enhanced in Group D. Biosynthetic pathways for streptomycin (strA, strB) and vancomycin (vraS) displayed an increase after disinfection. Correlations were found between the streptomycin gene (strB) and novobiocin biosynthesis genes (SQD, sqdB) (Figure 6b). This suggests that NaClO promotes antibiotic resistance. Group N samples that were not exposed to the disinfectant yielded higher gene expression levels of S. aureus (isdG and isdl), which coincided with the abundance determinations for this bacterium.

Figure 6. PICRUSt2 prediction analysis. (a) Influence of NaClO disinfection on bacterial abundance assessed using KEGG pathways of bacterial communities, only pathways that were significantly affected are presented. (b) Network plots of correlation analysis between genes associated with these pathways, only correlations ≥ 0.7 were presented.

4. Discussion

Over the past ten years, the application of high-through put sequencing (HTS) methods, specifically targeting the 16S rDNA gene, has facilitated the analysis of microbial composition and detection of pathogenic and spoilage bacteria (Cauchie et al., 2020; Zhao et al., 2015). HTS provides a reliable method for comprehensively describing the dynamics of bacterial communities and analyzing the correlations between microbial metabolic pathways and spoilage bacteria (Charmpi et al., 2020; Xu et al., 2022; Yang et al., 2016; Zhong et al., 2021). In this study, NaClO disinfection altered the bacterial community structures, and heterogeneous taxonomic profiles were found between the surface types and sampling times. B. thermosphacta, P. fragi, E. coli, and S. aureus were the most abundant bacterial species found during the meat transaction period. However, this species become less dominant after decontamination, and the most abundant bacteria causing spoilage in pork were B. thermosphacta, Enterobacteriaceae, and Pseudomonas. Their metabolic activities can lead to a sour flavor, discoloration, gas and slime production, and a decrease in pH (Bassey et al., 2021; Ercolini et al., 2010; Stellato et al., 2016; Wickramasinghe et al., 2021; Zhao et al., 2015, 2022). Pseudomonas, Streptococcus, and Staphylococcus were found to be the dominant genera in another study of fresh pork samples from retail markets (Cheung et al., 2008; Zhang et al., 2021). In this study, Pseudomonas, Streptococcus and Staphylococcus were also detected. Among them, Pseudomonas fragi was detected as a relatively high abundance spoilage-associated bacteria during the meat trade. Staphylococcus spp. and Streptococcus spp. were also detected in 11 and 14 species, respectively. Staphylococcus can generate enterotoxins and biofilms, and certain strains can withstand high temperatures and develop resistance to antibiotics (Kadariya et al., 2014; Khatoon et al., 2018). Pseudomonas is a taxonomic group of great interest in the meat industry because of its prominent role in meat spoilage and is among the most frequently reported taxa found after sanitation of processed surfaces from all types of food production chains (Zhang et al., 2021). Its persistence is most likely due to its ability to grow at low temperatures, to form biofilms and tolerance to biocides (Møretrø & Langsrud, 2017). P. fragi is a lipolytic and proteolytic species of the Enterobacteriaceae family that produces volatile organic compounds and other undesirable metabolites, such as biogenic amines. These contribute to meat spoilage. Brochothrix and Pseudomonas are abundant in the aerobic microbial population of fresh meat, and Brochothrix is associated with sourcing rather than with putrefaction (Illikoud et al., 2019; Peruzy et al., 2020).

The knive and cutting board samples used for meat partitioning exhibited a significant presence of spoilage and pathogenic bacteria, similar to the meat samples, suggesting that meat partitioning could potentially cause cross-contamination (Figure 5b and S2). Bacteria can thrive on these surfaces, resulting in the transmission of foodborne pathogens to humans (Del Blanco et al., 2017). Therefore, it is imperative to implement appropriate measures to prevent the spread of microbes during the process of dividing meat portions and disinfecting surroundings and equipment in wholesale market. The effectiveness of disinfectants depends on factors such as material and protein load (Naïtali et al., 2009; Wang et al., 2020). The wooden chopping board surface exhibits greater roughness and can significantly increase the thickness and compactness of the biofilm, thereby increasing its resistance to disinfectants (Zhu et al., 2021). Under unhygienic conditions, a high load may reduce the efficacy against Gram-positive bacteria (Bessems, 1998; Tong, Hu, Chen, Li, Li, et al., 2021).

Our gene functional annotation analysis using PICRUSt2 suggests that NaClO disinfection during the clean-up period increased the abundance of genes related to RNA degradation and glycosphingolipid biosynthesis. RNA was used as the indicator of cell viability, and the expression of target genes was induced only in viable cells. Previous studies have revealed that the destruction of the cell membrane leads to the extrusion of macromolecules (e.g. ATP, proteins, and RNA) (Xu et al., 2017). Bacterial glycosphingolipids are structural elements tightly linked to environmental stress survival (Johansson et al., 2020). This indicates that NaClO provides selective pressure for bacteria (Tong, Hu, Chen, Li, Aifeng, et al., 2021). A previous study reported that the EC 3.4.16.4 serine-type d-Ala-d-Ala carboxypeptidase is significantly more abundant in pork samples and is a crucial component in regard to vancomycin resistance (Aráoz et al., 2000). The PWY-3781 aerobic respiration I (cytochrome c) genes are abundant in poultry and can contribute to human infections (Nearing et al., 2019).

Culture-dependent analyses coupled with typing of recovered isolates using molecular techniques are widely used to identify routes of microbial contamination between surfaces (Cobo Díaz et al., 2021). In contrast, 16S rDNA methods are culture-independent, convenient, rapid, and accurate and can be used to study microbial ecology and evolution (Ercolini, 2013; Knight et al., 2018). However, it is not possible to determine whether the functional genes detected were possessed by living bacteria, and it cannot be accurately determined whether these genes are functioning as described for the human gut microbiota (Lagier et al., 2016). Culture-dependent and-independent methods can complement each other in the study of microbial populations in food. On the other hand, the correlation analysis is a statistical measure often used in studies to show an association between variables. The limitations of correlation analysis are its assumption of a linear association and its sensitivity to a range of observations. The value of the correlation coefficient is also influenced by the measurement error. Additional PCR or culture procedures must be conducted to validate the findings of correlation analysis (Camara, 2018; Janse et al., 2021).

5. Conclusion

This is the first study to characterize the differences in bacterial contamination during the transaction and clean-up periods of a large wholesale pork market. NaClO disinfectants altered the bacterial community compositions. Chopping boards and meat have also been shown to be potential sources of bacterial cross-contamination. Functional predictions revealed that RNA degradation, glycosphingolipid levels, and antibiotic biosynthesis were enhanced during the clean-up period characterized by NaClO disinfection. This study demonstrates that bacterial community similarities and functional variations are highly correlated with NaClO disinfection.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by Natural Science Founding of Zhejiang Province [LZ24C200004], Natural Science Founding of China [32302233], Key Research and Development Program of Zhejiang Province [2022C02049], Ministry of Agriculture and Rural Affairs [14215033], and Walmart Foundation [UA2020-152, UA2021-247].