IntroductionEarly-lactation Holstein cows (n= 261) from 32 herds in three regions (Australia, California, and Canada) were previously categorized using a discriminant analysis model as being at a high (26.1% of cows), medium (26.8% of cows), or low risk (47.1% of cows) of ruminal acidosis. We aimed to investigate if (1) risk of acidosis would be associated with ruminal bacterial taxa and dietary nutrient components, (2) there would be individual or combinations of bacterial taxa associated with acidosis-risk groups, and (3) the abundance of bacterial taxa would be associated with the intake of dietary nutrient components.MethodsDiets ranged from pasture supplemented with concentrates to total mixed rations. Bacteria 16S ribosomal DNA sequences from rumen samples collected < 3 hours after feeding via stomach tube were analyzed to determine bacterial presence. The relative abundance of each bacterial phylum and family was center log transformed and the transformed family data were subjected to two redundancy analysis biplots, one for acidosis risk group and one for region, to identify the 20 best-fit bacterial families from each respective redundancy analysis. A total of 29 unique families were identified when the lists of 20 families were combined from each redundancy analysis, and these 29 families were termed "influential" families." The association of acidosis-risk groups with the abundance of individual influential families was assessed by mixed models. Backward stepwise elimination mixed models were used to determine the bacterial taxa associated with each acidosis-risk group and the dietary nutrients associated with the abundance of the bacterial taxa.Results and discussionHigh-risk acidosis cows were associated with increased abundances of Anaerocella_f and Veillonellaceae and decreased abundances of several bacterial families with different characteristics. Five phyla: Firmicutes [odds ratio (OR) = 7.47 ± 7.43], Spirochaetes (OR = 1.28 ± 0.14), Lentisphaerae (OR = 0.70 ± 0.07), Planctomycetes (OR = 0.70 ± 0.09), and Tenericutes (OR = 0.44 ± 0.15), and nine families were associated with a higher risk of acidosis. Of the nine phyla identified to be of interest based on abundance and strength of association with acidosis-risk groups, all had one or more dietary nutrient that predicted their abundance. Sugar was the most frequently associated nutrient with the nine phyla, and was present in 78% (seven out of nine phyla) of the models; crude protein was present in 56% of models and crude fat was present in 44% of the models. Sugar and crude protein were most associated with the influential families and all but three families had one or more nutrient predictive of their abundance. Ruminal bacterial taxa are associated with ruminal acidosis; dietary sugar and crude protein are vital predictors of these and, thus, of ruminal acidosis risk.
Read full abstract