The sources and transmission routes of microbial populations throughout a meat processing facility

Microbial food spoilage causes substantial food waste and poses a public health risk. Reducing food loss and preventing microbial contamination are crucial goals. Animal-derived products like pork are particularly vulnerable to spoilage due to their high water content and nutrient availability, creating ideal conditions for psychrotolerant organisms. Biofilm formation on processing equipment further complicates control. Traditional culture-dependent methods are insufficient for characterizing complex microbial communities and tracking transmission routes. Next-generation sequencing offers a powerful alternative for mapping microbial ecosystems in food processing environments. This study investigated the hypothesis that a significant portion of microbial contamination on meat originates from sources other than the animal itself during the cutting process. The researchers aimed to identify key contamination points and create a transmission map to guide targeted disinfection strategies.

Several studies have highlighted the challenges of controlling microbial contamination in meat processing. Culture-dependent methods reveal overall hygiene status, but fail to characterize complex microbial communities and population flows. Recent research employing next-generation sequencing has shown its suitability for mapping microbial ecosystems in various food industry sectors. Previous work by the authors indicated that many bacteria on slaughter pigs are not animal-associated, suggesting transfer during processing. This study built on these findings to explore the sources and routes of microbial transmission within a pork processing facility.

Samples were collected from twelve pigs at different stages of processing in an Austrian slaughterhouse (Sticking, Singeing, Polishing, Evisceration, Classification, Truck). Environmental samples were also collected from various surfaces and equipment. Total bacterial cell equivalents (BCE) were determined using 16S rRNA gene qPCR, along with aerobic mesophilic counts (AMC), Enterobacteriaceae (EB), and Pseudomonadaceae (PS) counts using ISO reference methods. Full-length 16S rRNA gene sequencing (PacBio Sequel) was performed on 133 samples, with a subset also sequenced using Illumina MiSeq for comparison. Sequence processing and analysis were performed using DADA2 for ASV identification and the Genome Taxonomy Database for taxonomic classification. SourceTracker software was used to identify the source of microbial contamination on meat samples by comparing meat samples (sinks) with environmental and animal samples (sources). Beta diversity analysis using Bray-Curtis distances examined changes in community structure. A simulation study was performed to evaluate the effect of sequencing depth on SourceTracker performance.

Bacterial cell counts varied significantly across sampling positions. Alpha diversity decreased from the beginning to the end of the processing line, with a transient increase after polishing. Beta diversity analysis revealed two major shifts: one during singeing and another between classification and trucking. The 50 most abundant ASVs showed heterogeneous distributions. SourceTracker analysis indicated that the main sources of bacterial contamination on meat were the polishing tunnel (whips, nozzles, and water), employee gloves, and a railing at the classification step. *Moraxella* spp., a major spoilage organism, was linked to these sources. A substantial portion (31.6%) of bacterial contamination on meat was attributed to an unknown source. Sequencing depth simulation showed that datasets with 1000 sequences per sample provided results comparable to deeper sequencing in terms of beta diversity and SourceTracker classification.

This study demonstrates the value of high-throughput, full-length 16S rRNA gene sequencing for identifying microbial contamination sources and transmission routes in meat processing. The findings highlight the importance of equipment hygiene (polishing tunnel) and worker practices (gloves) in minimizing contamination. The identification of specific taxa linked to particular sources allows for targeted disinfection strategies. The study shows that while WGS offers strain-level resolution, full-length 16S rRNA sequencing provides sufficient resolution for environmental monitoring and is more cost-effective. Further investigation into the unknown contamination source is warranted.

This research provides a comprehensive analysis of microbial transmission within a pork processing plant, pinpointing critical contamination points and enabling targeted interventions. The use of full-length 16S rRNA gene sequencing coupled with SourceTracker software proved highly effective. Future research should focus on identifying the unknown contamination sources, optimizing sampling and bioinformatics standardization, and exploring the application of these methods across other food production systems.

The study was conducted in a single slaughterhouse, limiting the generalizability of the findings. The unknown source of contamination may reflect limitations in sampling or sequencing depth. The SourceTracker analysis assumes source-sink relationships but does not account for the directionality of microbial transfer. The study focused on bacterial communities and did not address other microbial groups (fungi, viruses).