Dynamics of microbiota and antimicrobial resistance in on-farm dairy processing plants using metagenomic and culture-dependent approaches

On-farm dairy processing plants, which are situated close to farms and larger dairy processing facilities, face unique challenges in maintaining environmental hygiene. This can impact various stages of dairy processing. These plants operate on smaller scales and use Low-Temperature-Long-Time (LTLT) pasteurization, making them more susceptible to microbial contamination through direct and indirect contact. Antimicrobial-resistant bacteria found on dairy farms pose risks to human health by potentially transferring resistance via dairy products. Our study aimed to investigate microbial distribution and antimicrobial resistance at four key stages: the farm, pre-pasteurization, post-pasteurization, and processing environments. We assessed microbial distribution by quantifying indicator bacteria and conducting metagenomic analysis. Antimicrobial resistance was examined by identifying resistance phenotypes and detecting resistance genes in bacterial isolates and metagenomes. Our results showed that the indicator bacteria were detected at all stages of on-farm dairy processing. We observed a significant reduction in aerobic microbes and coliforms post-pasteurization. However, contamination of the final dairy products increased, suggesting potential cross-contamination during post-pasteurization. Metagenomic analysis revealed that Pseudomonas, a representative psychrotrophic bacterium, was predominant in both the farm (24.1 %) and pre-pasteurization (65.9 %) stages, indicating microbial transfer from the farms to the processing plants. Post-pasteurization, Pseudomonas and other psychrotrophs like Acinetobacter and Enterobacteriaceae remained dominant. Core microbiota analysis identified 74 genera in total, including 13 psychrotrophic bacteria, across all stages. Of the 59 strains isolated from these plants, 49 were psychrotrophic. Antimicrobial resistance analysis showed that 74.6 % (44/59) of isolates were resistant to at least one antibiotic, with cefoxitin, ampicillin-, amoxicillin-, and ticarcillin-resistant bacteria present at all stages. Identical antimicrobial resistance patterns were observed in isolates from serial stages of the same farm and season, suggesting bacterial transmission across stages. Additionally, 27.1 % (16/59) of isolates carried plasmid-mediated resistance genes, which were also detected in the metagenomes of non-isolated samples, indicating potential antimicrobial resistance gene transmission and their presence in uncultured bacteria. These findings reveal the persistence of antimicrobial-resistant psychrotrophic bacteria in on-farm dairy processing plants, which pose potential health risks via dairy consumption. Our study underscores the importance of both culture-dependent and cultureindependent methods to fully understand their distribution and impact.