Methods: Software
Fields: Biological Sciences, Engineering, Environmental Science, Metagenomics, Microbiology

Collaborators: David A. C. Beck (eScience & Chemical Engineering), Mila Chistoserdova (Chemical Engineering), Mary Lidstrom (Chemical Engineering, Microbiology)

Nonmetric multidimensional scaling (NMDS) of read counts. NMDS was computed with vegan (Oksanen et al., 2013) using the Bray-Curtis dissimilarity index with a final stress of 0.156. The unamended sediment samples appear grouped at the figure left-top. Trajectories for the first replicate of each HO and LO condition are shown using cyan and magenta lines, respectively, connecting the sampling points. Samples from other replicates are also shown using the same color scheme. OTUs appear as text.

Nonmetric multidimensional scaling (NMDS) of read counts. NMDS was computed with vegan (Oksanen et al., 2013) using the Bray-Curtis dissimilarity index with a final stress of 0.156. The unamended sediment samples appear grouped at the figure left-top. Trajectories for the first replicate of each HO and LO condition are shown using cyan and magenta lines, respectively, connecting the sampling points. Samples from other replicates are also shown using the same color scheme. OTUs appear as text.

We observe the dynamics of bacterial communities in response to methane stimulus in laboratory microcosm incubations prepared with Lake Washington sediment samples. We first measured taxonomic compositions of eleven independent long-term enrichment cultures and determined that, while dominated by Methylococcaceae types, these cultures also contained accompanying types belonging to a limited number of bacterial taxa, both methylotrophs and non-methylotrophs. We then followed the short-term community dynamics, in two oxygen tension regimens (‘high’, 150 mM  and ‘low’, 15 mM), observing rapid loss of species diversity. In all microcosms, a single type of Methylobacter represented the major methane-oxidizing partner in the community. The accompanying members of the communities revealed different trajectories in response to different oxygen tensions. While Methylotenera species were the early responders to methane stimulus under both conditions, under the ‘high’ oxygen condition they were replaced by successions of other bacteria, both methylotrophs and non-methylotrophs. Under the ‘low’ oxygen condition, the Methylotenera partners were more persistent, with the early-responder ecotype being succeeded by a different ecotype. The communities in both conditions were convergent in terms of their assemblage, suggesting selection for specific taxa. Our results support prior observations from metagenomics on distribution of carbon from methane among diverse bacterial populations and further suggest that communities are jointly responsible for methane cycling, rather than a single type of microbe.

Publications:

Software:

Data: