Microbes at the center of networks are keystone species during early colonization of a natural ecosystem.

Background/Question/Methods

Microorganisms are key drivers of global ecosystem processes including organic matter decomposition, nutrient cycling, maintenance of community biodiversity, and facilitation of organismal health. When there is a dysbiosis in the microbial community, a plethora of ecosystem problems can occur ranging from invasive weed growth and leaf litter accumulation to declines in primary productivity. Yet, our understanding of how natural and healthy microbial communities assemble lags far behind our knowledge of the importance of these communities. Central taxa (taxa with many connections to other microbes within microbial networks) are theorized to be keystone species that disproportionately structure microbial communities, yet this is based purely on placement within a correlative network of microbial occurrences and laboratory manipulations rather than empirical field tests of microbial roles in natural communities. To address this gap in understanding of microbial community assembly, our research experimentally investigates soil microbiomes from a local imperiled ecosystem to ask what characteristics of early microbial colonizers influence the soil microbiome’s community structure. We use microbiome sequencing, culture collections, network analysis, and field experiments to identify ‘central’ and ‘peripheral’ microbes within microbiome networks and experimentally evaluate their effects on natural microbial community assembly after an ecosystem disturbance.

Results/Conclusions

Our experiment demonstrated that early colonization of soils by highly connected central microbes drove significantly higher prokaryotic richness (22% increase) after disturbance than what resulted from less-connected, peripheral early colonizers (F4,310 = 2.52, p = 0.04). Communities initiated by central microbes also showed less variance in primary succession community assembly compared to communities initiated with less connected microbes (F4,100 = 5.21, p < 0.001). This demonstrated that central microbes lead to more predictable early succession patterns and are actively structuring their communities. Inoculations with central microbes also matched community succession patterns in naturally occurring microbiomes. Taken together, these results demonstrate for the first time that central microbes within networks are acting as keystone microbes that enhance biodiversity and structure microbiomes. Our work validates the use of network approaches to identify keystones in natural communities and provides foundational knowledge of microbial assembly important for advancing habitat restoration in the future.