Geothermal soil ecology in Yellowstone National Park
Meadow, James Francis
MetadataShow full item record
Microbial communities in soil are among the most diverse and species-rich of any habitat, but we know surprisingly little about the factors that structure them. Geothermal soils present unique and relatively unexplored model systems in which to address ecological questions using soil microbial communities, since harsh conditions in these soils exert strong filters on most organisms. This work represents two very different approaches to studying soil ecology in geothermal soils in Yellowstone National Park: 1) Arbuscular mycorrhizal fungal (AMF) communities living in the roots of Mimulus gutattus in contrasting plant community types were compared to assess a link in community structure between plants and their AMF symbionts; and 2) soil microbial communities were surveyed across multiple spatial scales in an unstudied diatomaceous biological soil crust in alkaline siliceous geothermal soils, using bar-coded 454 pyrosequencing of 18S and 16S rDNA. Mycorrhizal communities living in plant roots from contrasting community types showed a striking difference in taxon richness and diversity that appears to transcend soil-chemical differences, though robust conclusions are difficult since plant and fungal communities are structured by some of the same confounding soil conditions. Cluster and discriminant analyses were employed to compare drivers of AMF community structure. Eukaryotic and prokaryotic communities in a diatomaceous biological soil crust differ significantly from that of an adjacent sinter soil, and along a photic depth gradient. Along with a description of this unique system, extensive multivariate community analyses were used to address outstanding questions of soil microbial community spatial heterogeneity and the methodologies best suited to the unique assumptions of these datasets. Depending on the intended scope of inference, much detail can be gained by investigation of microbial communities at the aggregate or soil particle scale, rather than through composite sampling. Additionally, beta-diversity patterns are apparent with relatively few sequences per sample.