Combining Bioinformatics and Statistical Modeling to Understand the Ecology of the Rumen Microbial Ecosystem in Sheep

We analyzed ruminal metagenomes from 16 sheep that were fed two different diets.  Using Illumina sequencing of raw microbial DNA, we bioinformatically mapped the resulting reads to prokaryotic 16S rDNA sequences to identify known taxa and also sought to discover new taxa. Strikingly, the majority of the microbes found did not map to known taxa. We then used statistical modeling to compare the microbial taxonomic distributions between animals fed a forage-based diet and those fed concentrated grains. There are significant differences between the two groups both in the dominant taxa and in the shape of the taxa abundance curves. Forage-fed animals appear to have a more diverse microbial ecosystem, whereas the concentrate-fed animals have ecosystems dominated by a few taxa. As expected, organisms from methanogenic groups are more prevalent among the forage-fed animals. All of these differences seem to be grounded in a common core of microbial individuals entering rumen, because common organisms from one feed group are present in the other, but at much lower abundance.