Microbial food spoilage is responsible for a considerable amount of waste and can cause food-borne diseases in humans, particularly in immunocompromised individuals and children. Therefore, preventing microbial food spoilage is a major concern for health authorities, regulators, consumers, and the food industry. However, the contamination of food products is difficult to control because there are several potential sources during production, processing, storage, distribution, and consumption, where microorganisms come in contact with the product. Here, we use high-throughput full-length 16S rRNA gene sequencing to provide insights into bacterial community structure throughout a pork-processing plant. Specifically, we investigated what proportion of bacteria on meat are presumptively not animal-associated and are therefore transferred during cutting via personnel, equipment, machines, or the slaughter environment. We then created a facility-specific transmission map of bacterial flow, which predicted previously unknown sources of bacterial contamination. This allowed us to pinpoint specific taxa to particular environmental sources and provide the facility with essential information for targeted disinfection. For example, Moraxella spp., a prominent meat spoilage organism, which was one of the most abundant amplicon sequence variants (ASVs) detected on the meat, was most likely transferred from the gloves of employees, a railing at the classification step, and the polishing tunnel whips. Our results suggest that high-throughput full-length 16S rRNA gene sequencing has great potential in food monitoring applications.

Microbial food spoilage is responsible for a considerable amount of waste and can cause food-borne diseases in humans, particularly in immunocompromised individuals and children. Therefore, preventing microbial food spoilage is a major concern for health authorities, regulators, consumers, and the food industry.However, the contamination of food products is difficult to control because there are several potential sources during production, processing, storage, distribution, and consumption, where microorganisms come in contact with the product. Here, we conduct the first study that uses high-throughput full-length 16S rRNA gene sequencing to provide novel insights into bacterial community structure throughout a pork processing plant. Specifically, we investigated what proportion of bacteria on meat are not animal-associated and are therefore transferred during cutting via personnel, equipment, machines, or the slaughter environment. We then created a facility-specific transmission map of bacterial flow which revealed previously unknown sources of bacterial contamination. This allowed us to pinpoint specific taxa to particular environmental sources and provide the facility with essential information for targeted disinfection. For example, Moraxella spp., a prominent meat spoilage organism which was one of the most abundant amplicon sequence variants (ASVs) detected on the meat, was most likely transferred from the gloves of employees, a railing at the classification step, and the polishing tunnel whips. Finally, we provide evidence that 1000 sequences per sample provides a reasonable sequencing depth for microbial source tracking in food processing, suggesting that this approach could be implemented in regular monitoring systems. BackgroundAs the world population is expected to rise to 9.8 billion by 2050, the global demand for food will increase by approximately 70 % in order to satisfy human needs. Resolving this issue, while also reducing greenhouse gas emissions and protecting valuable ecosystems is one of the greatest challenges of our era. Food security experts estimate that 46 % of the required additional food demand can be achieved by increasing food production, whereas the remaining proportion needs to be attained through sustaining the productive capacity (34 %) and better food demand management (20 %) (Keating et al.

Traditionally, the microbiological status of meat is determined by culture-based techniques, although many bacteria are not able to grow on conventional media. The aim of this study was to obtain quantitative data on total bacterial cell equivalents, as well as taxa-specific abundances, on carcass surfaces during pig slaughter using quantitative real-time PCR. We evaluated microbial contamination patterns of total bacteria, Campylobacter, Escherichia coli, Lactobacillus group, Listeria monocytogenes, Salmonella, and Pseudomonas species throughout slaughtering and on different carcass areas. In addition, we compared contamination levels of breeding sow carcasses with fattening pig carcasses, and we assessed the efficacy of carcass polishing machines under two water amount conditions. Our results demonstrate that relevant meat-spoilage organisms show similar contamination patterns to total bacteria. The highest bacterial load was detected in the stunning chute (4.08 × 105 bacterial cell equivalents per cm2) but was reduced by 3 log levels after singeing and polishing (P < 0.001). It increased again significantly by a 4.73-fold change until the classification step. Levels of Campylobacter, Lactobacillus, and Pseudomonas species and of E. coli followed a similar trend but varied between 0 and 2.49 × 104 bacterial cell equivalents per cm2. Microbial levels did not vary significantly between sampled carcass areas for any analyzed taxa. Running the polishing machine with a low water amount proved to be less prone to microbial recontamination compared with a high water amount (17.07-fold change, P = 0.024). In the studied slaughterhouse, slaughter of breeding sows did not produce microbiologically safe meat products (>104 cells per cm2) and the implementation of specific hazard analysis critical control point systems for the slaughter of breeding sows should be considered. A larger cohort from different abattoirs is needed to confirm our results and determine whether this is universally valid.

No abstract