Characterization of microbiomic and geochemical compositions across the photosynthetic fringe

Abstract

Hot spring outflow channels provide geochemical gradients that are reflected in microbial community compositions. In many hot spring outflows, there is a distinct visual demarcation as the community transitions from predominantly chemotrophs to having visible pigments from phototrophs. It has been hypothesized that this transition to phototrophy, known as the photosynthetic fringe, is a result of the pH, temperature, and/or sulfide concentration gradients in the hot spring outflows. Here, we explicitly evaluated the predictive capability of geochemistry in determining the location of the photosynthetic fringe in hot spring outflows. A total of 46 samples were taken from 12 hot spring outflows in Yellowstone National Park that spanned pH values from 1.9 to 9.0 and temperatures from 28.9 to 92.2°C. Sampling locations were selected to be equidistant in geochemical space above and below the photosynthetic fringe based on linear discriminant analysis. Although pH, temperature, and total sulfide concentrations have all previously been cited as determining factors for microbial community composition, total sulfide did not correlate with microbial community composition with statistical significance in non-metric multidimensional scaling. In contrast, pH, temperature, ammonia, dissolved organic carbon, dissolved inorganic carbon, and dissolved oxygen did correlate with the microbial community composition with statistical significance. Additionally, there was observed statistical significance between beta diversity and the relative position to the photosynthetic fringe with sites above the photosynthetic fringe being significantly different from those at or below the photosynthetic fringe according to canonical correspondence analysis. However, in combination, the geochemical parameters considered in this study only accounted for 35% of the variation in microbial community composition determined by redundancy analysis. In co-occurrence network analyses, each clique correlated with either pH and/or temperature, whereas sulfide concentrations only correlated with individual nodes. These results indicate that there is a complex interplay between geochemical variables and the position of the photosynthetic fringe that cannot be fully explained by statistical correlations with the individual geochemical variables included in this study.

Keywords: Yellowstone National Park; geochemistry; hot spring; microbiome; photosynthetic fringe.

FIGURE 1

Hot spring outflows with depictions of the photosynthetic fringe. Locations above (blue), at (yellow), and below (green) the photosynthetic fringe are indicated by respective diamonds. Black arrows indicate a general flow of water away from the source. (A) The visual representation of the photosynthetic fringe in the outflow of a basic spring, BP, where the white sinter in the outflow meets orange phototrophic mats along the edge of the outflow. (B) The visual representation of the photosynthetic fringe in the outflow of an acidic spring, CF, where the clear outflow meets green phototrophic mats. (C) An example of the start of a hot spring outflow channel from a basic hot spring, BP, before reaching its photosynthetic fringe.

FIGURE 2

A total of 46 samples taken from the outflows of 12 separate hot springs above, at, and below the photosynthetic fringe, as determined by linear discriminant analysis, displayed as functions of (A) pH and temperature, and (B) total dissolved sulfide and temperature. In both panels (A,B), the dashed line represents the photosynthetic limits defined by Cox et al. (2011).

FIGURE 3

Percent relative abundance of the 16S rRNA gene sequencing results to the class level except for Armatimonadota, which is at the phylum level, and unidentified bacteria, which were binned together at the domain level. To focus on abundant features and overarching patterns, classes not occurring at >20% relative abundance when summed over all samples were binned into the “Other” category. Organization of hot spring sites, separated by black bars, follows the order of increasing pH shown in Supplementary Table 1. Within each site, samples are organized down the outflow with above (blue diamond), at (yellow diamond), and below (green diamond) the photosynthetic fringe indicated as in Figure 1.

FIGURE 4

Non-metric multidimensional scaling (NMDS) analysis using the 16S rRNA gene sequencing data from each of the 46 samples. Each point represents the normalized microbial community composition determined in a hot spring sample while the distance between points represents the dissimilarity. Sample point colors (blue, yellow, and green) refer to the position along the photosynthetic fringe (above, at, below, respectively). Geochemical data are added as vectors; vectors that correlate with an ordination axis with a p-value < 0.05 are indicated by an asterisk.

FIGURE 5

Co-occurrence network analysis of commonly occurring ASVs within the 46 hot spring samples. Each node represents an ASV, and nodes were organized into 18 cliques using Louvain’s membership algorithm. The connections between nodes, also known as edges, represent a >0.7 Spearman’s correlation coefficient. In panels (A–D), each node is colored on a gradient of blue to red based on its Spearman’s correlation, from negative one to positive one, respectively, with the following geochemical parameters, (A) pH, (B) temperature, (C) total ammonia, and (D) DOC. Each of the cliques is differentiated by color and number in panel (E).