RESUMO
Microbial contamination in food processing plants can play a fundamental role in food quality and safety. In this study, the microbiota in a dairy plant was studied by both 16S rRNA- and 26S rRNA-based culture-independent high-throughput amplicon sequencing. Environmental samples from surfaces and tools were studied along with the different types of cheese produced in the same plant. The microbiota of environmental swabs was very complex, including more than 200 operational taxonomic units with extremely variable relative abundances (0.01 to 99%) depending on the species and sample. A core microbiota shared by 70% of the samples indicated a coexistence of lactic acid bacteria with a remarkable level of Streptococcus thermophilus and possible spoilage-associated bacteria, including Pseudomonas, Acinetobacter, and Psychrobacter, with a relative abundance above 50%. The most abundant yeasts were Kluyveromyces marxianus, Yamadazyma triangularis, Trichosporon faecale, and Debaryomyces hansenii. Beta-diversity analyses showed a clear separation of environmental and cheese samples based on both yeast and bacterial community structure. In addition, predicted metagenomes also indicated differential distribution of metabolic pathways between the two categories of samples. Cooccurrence and coexclusion pattern analyses indicated that the occurrence of potential spoilers was excluded by lactic acid bacteria. In addition, their persistence in the environment can be helpful to counter the development of potential spoilers that may contaminate the cheeses, with possible negative effects on their microbiological quality.