Figure 4.

Environmental, farming and technological parameters contributing to milk microbiota shaping. (A) correlations between the beta-diversity based on bray-Curtis dissimilarities of the total milk dataset (N = 370), and environmental, farming and technological parameters. The parameters were sorted by category as PDO and PDO-driven variables, farming practices associated to each production and milk sample specific descriptors, and then by R2 value, represented by the size of the label, with a threshold set to R2 > 0.2. Dark colors indicate significant correlations with p values below 0.05. Data are presented in column alternately for bacteria and fungi. (B, C) Main contributors to cow’s milk bacterial community shaping. Focus on the milk from the six PDOs with the lowest intra-PDO dispersion (N = 63). (B) Non metric multi-dimentional scaling ordinations based on bray-Curtis dissimilarities. PERMANOVA analyses were performed to test the effects of PDO label (R2 = 0.386, P-value < .001), topography (R2 = 0.220, P-value < .001), and production type (R2 = 0.054, P-value < .001). (C) Box plots showing the relative abundance of the bacterial orders above 10% relative abundance, in the individual samples of cow’s milk across the six PDOs.