Introduction: High prevalence of subacute rumen acidosis (SARA) in dairy herds has been reported with large impact on production and welfare. The field diagnosis of SARA is currently unclear and primarily based on point measurements of rumen pH, which are inaccurate. Consequently, SARA cases in the field are often not detected. Thus, other and better markers of SARA are needed. The purpose of this research was to study the feces microbiome during SARA and assess the possibilities of using feces microbial markers as indicators of SARA. Methods: Six lactating, rumen cannulated, Danish Holstein cows were used in a blocked design study including two blocks. In the first block, two cows received control diet and two cows received SARA-challenge diet. In the second block, former control cows received SARA diet while two new cows received control diet. Cows received a total mixed ration (TMR; 24% concentrate) for four weeks before the trial. SARA was induced by gradual substitution of 40% of TMR with grain pellets (50:50 wheat:barley) over 3 days. Full SARA diet was fed for four days. Rumen pH was measured continuously by indwelling probes (eCow). Feces samples were taken at 9 am and 9 pm on last day of the control period, and second and last day of full SARA-feeding. DNA was extracted and the V4 region of bacterial 16S rRNA gene was amplified and subjected to Illumina sequencing. Bioinformatics were performed using QIIME and resultant operational taxonomic units (OTUs) were aligned to Greengenes database at 97% similarity threshold. The differences between phylogenetic structures of microbial communities were calculated based on weighted and unweighted UniFrac distances and tested using permutational multivariate analysis of variance (PERMANOVA). Partial least square discriminant analysis (PLS-DA) was applied to identify the taxa that were correlated to each treatment group. Goodness of fit and predictive value was evaluated by R2Y and Q2 estimates. Taxa with a Variable Importance for the Projection (VIP) below 0.5and relative abundance below 0.3 were excluded from the dataset. Results: In total, 641,335 high quality sequences were obtained resulting in identification of 81 classified bacterial genera belonging to 15 different bacterial phyla. The fecal bacterial communities of SARA and control samples were significantly distinct with P-values of 0.0262 and 0.0001 for weighted and unweighted UniFrac distances, respectively. The proportion of several taxa was significantly higher in SARA samples compared to control during full SARA feeding. These included Lachnospiraceae, Bifidobacteriaceae, YRC22, Roseburia, Treponema, 5-7N15, Parabacteroides , Anaerovibrio, Blautia, Coprococcus, Veillonellaceae, Ruminococcus, Bifidobacterium, and Sutterella. Goodness of fit and predictive value was estimated to R2: 95.3 and Q2: 74.6, respectively. On phyla level, the proportion of Firmicutes, Actinobacteria and Spirochaetes was significantly higher in SARA samples compared to control. Goodness of fit and predictive value was estimated to R2: 87.0 and Q2: 73.2, respectively. Conclusion: Results confirm that intensive grain feeding changes the feces microbiome. The identification of specific taxa characteristic of SARA could provide new knowledge of the pathogenesis and might be useful as future biological markers of the disease. bovine, subacute ruminal acidosis, fecal microbiome, biological marker
Host publication information