Prokaryotic diversity and biogeochemical characteristics of benthic microbial ecosystems at La Brava, a hypersaline lake at Salar de Atacama, Chile

Abstract

Benthic microbial ecosystems of Laguna La Brava, Salar de Atacama, a high altitude hypersaline lake, were characterized in terms of bacterial and archaeal diversity, biogeochemistry, (including O2 and sulfide depth profiles and mineralogy), and physicochemical characteristics. La Brava is one of several lakes in the Salar de Atacama where microbial communities are growing in extreme conditions, including high salinity, high solar insolation, and high levels of metals such as lithium, arsenic, magnesium, and calcium. Evaporation creates hypersaline conditions in these lakes and mineral precipitation is a characteristic geomicrobiological feature of these benthic ecosystems. In this study, the La Brava non-lithifying microbial mats, microbialites, and rhizome-associated concretions were compared to each other and their diversity was related to their environmental conditions. All the ecosystems revealed an unusual community where Euryarchaeota, Crenarchaeota, Acetothermia, Firmicutes and Planctomycetes were the most abundant groups, and cyanobacteria, typically an important primary producer in microbial mats, were relatively insignificant or absent. This suggests that other microorganisms, and possibly novel pathways unique to this system, are responsible for carbon fixation. Depth profiles of O2 and sulfide showed active production and respiration. The mineralogy composition was calcium carbonate (as aragonite) and increased from mats to microbialites and rhizome-associated concretions. Halite was also present. Further analyses were performed on representative microbial mats and microbialites by layer. Different taxonomic compositions were observed in the upper layers, with Archaea dominating the non-lithifying mat, and Planctomycetes the microbialite. The bottom layers were similar, with Euryarchaeota, Crenarchaeota and Planctomycetes as dominant phyla. Sequences related to Cyanobacteria were very scarce. These systems may contain previously uncharacterized community metabolisms, some of which may be contributing to net mineral precipitation. Further work on these sites might reveal novel organisms and metabolisms of biotechnological interest.

Fig 1. Site location and images showing systems studied.

(A) Aerial view of Laguna La Brava indicating the sampling sites. (B) Aerial view of NLM (scale bar 5m). (C) Aerial view of microbialite site. (D) Detail of B, showing NLM (scale bar 0.5m). (E) Top view showing detail of pink mat (PM; scale bar 0.1m). (F) View of black mat (BM, scale bar 0.5m). (G) View from the side of Distichlis spicata (Gramineae), with underground rhizome-associated concretions not visible, scale bar 1m).

Fig 2. In situ depth profiles of oxygen and sulfide.

Microelectrode measurements were obtained during peak photosynthesis (12:00–14:00) Individual profiles of O₂ (squares) and sulfide (triangles) shown. The O₂ peaks and maximum values of sulfide that were observed were higher in microbialites than in the non-lithifying mat. The O₂ penetration was considerably deeper in the mat compared to the microbialites, indicating higher rates of O₂ production and consumption in the latter.

Fig 3. Comparison of prokaryotic diversity by sampling location.

Bulk samples based on 16S rRNA gene sequences of the V4 hypervariable region. Bars indicate the contribution of each phylum to the total diversity. Phyla representing less than 1% of the total diversity are grouped as “minor phyla”.

Fig 4. Prokaryotic diversity by layer in NLM.

(A) Relative abundance of the phyla based on bacterial 16S rRNA gene sequences of the V4 hypervariable region. Phyla that represent less than 1% of total diversity are grouped in “minor phyla”. (B) Functional diversity abundance by layer. Percentage of sequences belonging to Bacteria and Archaea represented layer per layer. Notice the log scale. Functional groups were inferred from literature search of the metabolic capabilities of each classified microorganism present in the sample.

Fig 5. Prokaryotic diversity by layer in BM.

(A) Relative abundance of the phyla based on bacterial 16S rRNA gene sequences of the V4 hypervariable region. Phyla that represent less than 1% of total diversity are grouped in “minor phyla”. (B) Functional diversity abundance by layer. Percentage of sequences belonging to Bacteria and Archaea represented layer per layer. Notice the log scale. Functional groups were inferred from literature search of the metabolic capabilities of each classified microorganism present in the sample.

Fig 6. OTU-level comparison of the sites using 97%-sequence identity.

(A) Venn diagram showing the number of OTUs shared between the sites, the percentages in parentheses represent relative abundance of sequences assigned to the OTUs. The bar graph shows the distribution of shared OTUs: from left to right: 67 OTUs are shared by all four systems, 120 OTUs are shared by three, 229 are shared by two and 457 OTUs are unique (not shared with other systems). (B). Principal Coordinates Analysis (PCoA) based on OTUs in which 94% of the variation is explained by the first two axes.

Fig 7. Canonical Correspondence Analysis (CCA) of the prokaryotic lineages, sampling sites and environmental properties.

Triangles represent response variables (OTU abundances). Arrows represent quantitative explanatory variables (physico-chemical parameters) with arrowheads indicating their direction of increase. Circles represent qualitative explanatory variables (sites). BOD: Biochemical oxygen demand; COD: Chemical oxygen demand; Chla: Chlorophyll a; HN: Hardness; TAlk: Total alkalinity; TOC: Total organic Carbon; NO-3: Nitrate; NO-2: Nitrite; TON: Total organic nitrogen; TP: Total phosphorus; PO43-: Phosphate; SO42-: Sulfate; S: Sulfur; Na+: Sodium; Cl-: Chloride; K+: Potassium; Mg2+: Magnesium; Ca2+: Calcium; DB: Dissolved boron; TB: Total boron; DLi: Dissolved lithium; TLi: Total lithium; SiO2: Silica; DAr: Dissolved Arsenic; TAr: Total Arsenic.