Understanding how microbiomes influence the systems they inhabit

This perspective piece is a product of a working group from the United States Geological Survey’s John Wesley Powell Center for Synthesis and Analysis.

Like Comment

The review in Nature Microbiology is here: https://go.nature.com/2Qdyizh

The Powell Center funds groups of researchers to come together in Fort Collins, CO USA and spend a week or so synthesizing data and ideas with the intent to advance a particular area of earth systems science. Our authorship group was composed of post-doctoral researchers, graduate students, professors and federal scientists working at the interface of ecosystem and microbial ecology. We initially looked into the existing literature to ask how the field was currently addressing research at the interface of these two disciplines. This resulted in two meta-analyses, one led by Dr. Jenny Rocca (Duke University) and one led by Dr. Raven Bier (Stroud Water Research Center). Dr. Rocca’s manuscript evaluated how frequently the relative abundance of a protein encoding gene or transcript correlated with the process rate that gene or transcript was thought to influence (Rocca et al. 2015). Dr. Bier’s manuscript looked at the response of microbial community membership (i.e. 16S amplicon libraries) to an environmental perturbation (Bier et al. 2015). As a group we suspected that the dominant research at the intersection of these disciplines had a more correlative than causal focus. However, we were all genuinely surprised by the findings of both analyses. Whereas each of these manuscripts reveals a series of interesting findings and is worth a read in and of themselves, what surprised our working group was that both synthesis papers found that the relationship between measurable microbial characteristics and ecosystem processes were commonly assumed but rarely tested. More often than not the researchers would measure both the microbial and ecosystem components but not test for a relationship between the two. Furthermore, when studies did test for a correlation, the correlations were more commonly not significant than significant. These findings were surprising but also provided a sort of fascinating challenge to see if we could provide a way forward that would be useful to researchers at various career stages.

The Powell Center does several things to foster collaboration. First they provide all participants with bicycles. Fort Collins is a very bike friendly town and biking between dinner, the hotel, and meeting site provided a unique chance to interact in an unstructured way. Second they encourage one day outside of the meeting site to stimulate discussion. We took a day, during our first meeting, to hike in Rocky Mountain National Park – a short drive from Fort Collins. After discussing ideas on the drive and hike in – we rested and ate lunch on the shores of a sub-alpine lake. It was hear that the current manuscript was conceived. Armed with rock alone, group member Paul del Giorgio illustrated the discussion ideas by sketching (with rock, on rock, – a “lithograph”?) the first iteration of what is now Figure 2. 

A sub-set of the paper's authors discussing the intersection of ecosystem and microbial ecology. This picture was taken at the initial meeting, moments before the first iteration of what is now Figure 2 in the current manuscript. Notice the improvised writing implement in Paul del Giorgio's (center, back to lake) right hand.

Figure 2 illustrates how the two disciplines (microbial and ecosystem ecology) interact and identifies three categories that contain the majority of commonly measured microbial metrics. Our goal was to move research at the intersection of microbial and ecosystem ecology from correlative and descriptive studies towards more mechanistic and causal studies. This start led to a special session on the topic at an annual meeting of the Ecological Society of America, a long-series of additional discussions, reiterations, revisions, and eventually the manuscript published today in Nature Microbiology (Hall et al. 2018).

I also want to highlight an important participant to this work who passed away during its preparation. During our second in-person meeting (at which we began to draft the current manuscript) we were joined by Dr. Diana Nemergut. Diana participated in our group as a vibrant and creative scientist who effortlessly meshed with an established working group, added critiques and contributions, rode bikes through the streets of Fort Collins, raised the bar on our collective dance skills, and came ready to do it all again the next day. She was a true light whose clarity of thought and enthusiasm is represented in the work published here and has provided (me and many others) an important model of how to create and push the scientific discipline of our choice a bit farther each day. Thank you Diana.

Below are the full citations and links to each of the manuscripts mentioned above.

Rocca, J. D., Hall, E. K., Lennon, J. T., Evans, S. E., Waldrop, M. P., Cotner, J. B., … Wallenstein, M. D. (2015). Relationships between protein-encoding gene abundance and corresponding process are commonly assumed yet rarely observed. ISME Journal, 9(8). https://www.nature.com/articles/ismej2014252

Bier, R. L., Bernhardt, E. S., Boot, C. M., Graham, E. B., Hall, E. K., Lennon, J. T., … Wallenstein, M. D. (2015). Linking microbial community structure and microbial processes: An empirical and conceptual overview. FEMS Microbiology Ecology, 91(10). https://academic.oup.com/femsec/article/91/10/fiv113/2467433

Hall, E. K., Bernhardt, E. S., Bier, R. L., Bradford, M. A., Boot, C. M., Cotner, J. B., … Wallenstein, M. D. (2018). Understanding how microbiomes influence the systems they inhabit. Nature Microbiology, 3(9), 977–982. https://www.nature.com/articles/s41564-018-0201-z

Ed Hall

Assistant Professor, Colorado State University