Abstract
Questionnaires on farming conditions were retrieved from 2129 dairy farms and clustered, resulting in 106 representative raw cow's milk samples analysed in winter and summer. Substantiating the efficiency of our survey, some farming conditions affected the milk physicochemical composition. Culturing identified several species of lactic acid bacteria (LAB) per milk, whose number increased through 16S ribosomal RNA (rRNA) gene sequencing and shotgun metagenome analyses. Season, indoor versus outdoor housing, cow numbers, milk substitutes, ratio cattle/rest area, house care system during lactation, and urea and medium-chain fatty acids correlated with the overall microbiome composition and the LAB diversity within it. Shotgun metagenome detected variations in gene numbers and uniqueness per milk. LAB functional pathways differed among milk samples. Focusing on amino acid metabolisms and matching the retrieved annotated genes versus non-starter lactic acid bacteria (NSLAB) references from KEGG and corresponding to those identified, all samples had the same gene spectrum for each pathway. Conversely, gene redundancy varied among samples and agreed with NSLAB diversity. Milk samples with higher numbers of NSLAB species harboured higher number of copies per pathway, which would enable steady-state towards perturbations. Some farming conditions, which affected the microbiome richness, also correlated with the NSLAB composition and functionality.