Diatomaceous geothermal soils: a sinter particles (disturbed) with white on the surface above orange and green photic zones (scale bar 0 1 cm); b fibrous, dense diatomaceous soil crust (scale bar 0 3 cm), and this is the general appearance and surface texture of crust samples; and c vertical photic layering (scale bar 0 1 cm). This striation was 

Contexts in source publication

Context 1

... proceeding with ordination, we assessed phylogenetic influence when using the unweighted UniFrac distance rather than a more traditional nonphylogenetic index (Supplemental Fig. 1). We found surprising agreement between UniFrac and the Ochiai distances and similar agreement when compared to the abundance- based Canberra index. Though Canberra allows greater dissimilarity due to abundance differences, both are compara- ble; we used the UniFrac metric for all further multivariate analyses since the small amount of information gained by considering phylogeny is a step toward incorporating community ecology and evolution. Final stress of the two- dimensional NMDS (Fig. 3) was 8.0, below the conventional problematic cutoff (>15). Points are coded by soil type (shape) and depth (grayscale), with replicate points joined by gray lines. There is a predictable grouping by both soil type and by depth, with only one clear exception: two replicate sinter surface particles are not grouped accordingly, but rather with mid-depth crust samples. Four fuzzy set ordinations were produced and are dis- played in Fig. 4. The purpose of FSO is a single- dimensional comparison of community data (UniFrac phylogenetic distance, in this case) with a fuzzy set ordination value ( y -axis); the ordering of the x -axis is categorical (a) and ordinal (b and c), and the degree of separation along the y -axis indicates community agreement with x -axis data [7, 25]. The three combinations tested were: (a) soil type; (b) depth; and (c) the combination of both categorical variables (FSO correlations and P values are reported in Table 1). While the results from these FSOs are convincing on their own, we also formally tested the three sets of groups with permutational MANOVA using . It is important to note that we ran the combination of the two variables as an interaction with , while FSO was run as two different sets for the “ interaction ” FSO (Fig. 4), since this produces a more straightforward ordination. All three combinations tested with were significant ( P values<0.001 in every case; Table 2). R 2 values from are shown in each panel of Fig. 4, and the third combination is clearly the most explanatory of the variation in the data. Soil types show a remarkably consistent ordination pattern in Fig. 4c, with overlap between surface and mid-depth samples and more separation of deep samples from all ...

Context 2

... soil samples were taken from Imperial Meadow (cen- tered around 44°33 ′ 04 ′′ N, 110°51 ′ 05 ′′ W) in the Lower Geyser Basin, Yellowstone National Park, WY, USA. These biologically crusted soils represent a history of influence from alkaline siliceous geothermal water from a single source spring in the Lower Geyser Basin of Yellowstone National Park, WY, USA (YNP Thermal Inventory ID LFMNN020; [29]). Alkaline water from the source spring (pH 0 9.7) contains little N or P but is elevated in several biologically important elements, including As, B, Na, and Si (see [21] for a description of geothermal water chemistry at this site). The spring water occasionally inundates the adjacent soils, but these events appear to be infrequent. Sinter soils ( Fig. 1a) are composed of coarse, white, taphonized siliceous particles adjacent to the spring, while crust soils (Fig. 1b) are a dense, fibrous remnant wet meadow substrate largely composed of diatomaceous residuum and opaline silica. Sinter and crusted soils are within a 3-ha area sur- rounding the source spring. The only vascular plant present across the site is Triglochin maritima L., and its coverage is sparse; this results in a soil surface exposed to full solar radiation, and thus a colored phototrophic microbial community along a depth gradient that is conspicuous in both soil types (Fig. 1c). There are several clear differences between the two soil types, primarily in soil texture and structure; sinter soil particles are relatively more coarse (mean particle diameter 3 mm), while crust soils are primarily composed of taphonized T . maritima root material and diatomaceous frustular residuum, both a product of history under a geothermal wet meadow. This difference lends to a generally dry, aerated surface in sinter soils, and a moist, microbially active surface in crust soils. All samples used in this study were taken in October 2010. Three individual sinter particles (3-mm diameter) were sampled from each of the representative photic depths using sterile forceps. All sinter particles were sampled from within 1 m of one another, with one replicate set containing a particle from each photic depth taken from within a 1-m area. Mid-depth sinter samples consist of an extra set of replicate particles, since sinter particles were more ambiguous in orange versus green coloration (Fig. 1). These were preliminar- ily tested for community differences and were subse- quently lumped into one mid-depth group for final analysis. Likewise, deep-depth crust samples were also separated in preparation and grouped for this analysis. Crust samples were taken from three different locations within the upland crusted soils, each within 10 m of another and approximately 100 m from sinter sampling locations. Crust soils were sampled with a sterilized soil core (2-cm diameter) and were sampled to 5-cm depth. All samples were frozen on dry ice, on site, and ar- chived at − 80 °C until DNA extraction. Total C and N were measured using combustion of 0.2 g soil [35]. Irradiance was measured in both soil types (thawed soil samples at approximately field capacity) using a LI-COR Photometer (Model LI-189; www.licor.com) in full outdoor sunlight (mean 1,408 μ molm − 2 s − 1 ). Soils were tested for pH using a combination of 1:1 water extract for larger samples [11] and color pH ast pH indicator strips (www.emdmillipore. com/chemicals) for spot ...