LAUSR

The impact of subacute rumen acidosis (SARA) on the rumen bacterial community has been frequently studied in in vivo trials. Here we investigated whether these alterations can be mirrored by using the rumen simulation technique (RUSITEC) as an in vitro model for this disease. We hypothezised that the bacterial community fully recovers after a subacute ruminal acidosis challenge. We combined a PacBio nearly full-length 16S rRNA gene analysis with 16S rRNA gene Illumina MiSeq sequencing of the V4 hypervariable region. With this hybrid approach, we aimed to get an increased taxonomic resolution of the most abundant bacterial groups and an overview of the total bacterial diversity. The experiment consisted of a control period I and a SARA challenge and ended after a control period II, of which each period lasted 5 d. Subacute acidosis was induced by applying two buffer solutions, which were reduced in their buffering capacity (SARA buffers) during the SARA challenge. Two control groups were constantly infused with the standard buffer solution. Furthermore, the two SARA buffers were combined with three different feeding variations, which differed in their concentrate-to-hay ratio. The induction of SARA led to a decrease in pH below 5.8, which then turned into a steady-state SARA. Decreasing pH values led to a reduction in bacterial diversity and richness. Moreover, the diversity of solid-associated bacteria was lower for high concentrate groups throughout all experimental periods. Generally, Firmicutes and Bacteroidetes were the predominant phyla in the solid and the liquid phase. During the SARA period, we observed a decrease in fibrolytic bacteria although lactate-producing and -utilizing families increased in certain treatment groups. The genera Lactobacillus and Prevotella dominated during the SARA period. With induction of the second control period, most bacterial groups regained their initial abundance. In conclusion, this in vitro model displayed typical bacterial alterations related to SARA and is capable of recovery from bouts of SARA. Therefore, this model can be used to mimic SARA under laboratory conditions and may contribute to a reduction in animal experiments.