Contrasting the Genetic Patterns of Microbial Communities in Soda Lakes with and without Cyanobacterial Bloom

Abstract

Soda lakes have high levels of sodium carbonates and are characterized by salinity and elevated pH. These ecosystems are found across Africa, Europe, Asia, Australia, North, Central, and South America. Particularly in Brazil, the Pantanal region has a series of hundreds of shallow soda lakes (ca. 600) potentially colonized by a diverse haloalkaliphilic microbial community. Biological information of these systems is still elusive, in particular data on the description of the main taxa involved in the biogeochemical cycling of life-important elements. Here, we used metagenomic sequencing to contrast the composition and functional patterns of the microbial communities of two distinct soda lakes from the sub-region Nhecolândia, state of Mato Grosso do Sul, Brazil. These two lakes differ by permanent cyanobacterial blooms (Salina Verde, green-water lake) and by no record of cyanobacterial blooms (Salina Preta, black-water lake). The dominant bacterial species in the Salina Verde bloom was Anabaenopsis elenkinii. This cyanobacterium altered local abiotic parameters such as pH, turbidity, and dissolved oxygen and consequently the overall structure of the microbial community. In Salina Preta, the microbial community had a more structured taxonomic profile. Therefore, the distribution of metabolic functions in Salina Preta community encompassed a large number of taxa, whereas, in Salina Verde, the functional potential was restrained across a specific set of taxa. Distinct signatures in the abundance of genes associated with the cycling of carbon, nitrogen, and sulfur were found. Interestingly, genes linked to arsenic resistance metabolism were present at higher abundance in Salina Verde and they were associated with the cyanobacterial bloom. Collectively, this study advances fundamental knowledge on the composition and genetic potential of microbial communities inhabiting tropical soda lakes.

Keywords: alkaline lakes; arsenic; metagenomic; nitrogen; saline lakes; sulfur.

FIGURE 1

Relative abundances of the main bacterial phyla/classes in Salina Preta and Salina Verde lakes. Data are shown based on bacterial 16S rRNA gene sequences retrieved from the metagenomes (A) and functional metagenome annotations (B) (see section “Materials and Methods” for details). V: Salina Verde, P: Salina Preta, M: morning, A: afternoon, S: surface, B: bottom.

FIGURE 2

Functional distribution of the metagenomes annotated using the SEED subsystems. Statistically significant differences were found between Salina Preta and Salina Verde lakes (A), and between the bottom and surface water layers in Salina Preta (B) (see section “Results” for details).

FIGURE 3

Distribution of functional annotations involved in the C, N, S, and As cycles. The heatmap displays the relative abundance (row z-scores) of functional annotations across all samples (triplicate per site). ^∗Functional annotations that differentially segregated across sites were identified by random forest analysis with Boruta feature selection (average z-scores of 1,000 runs > 4). Circles are proportional to the relative abundance of each functional annotation in all samples. V: Salina Verde, P: Salina Preta, M: morning, A: afternoon, S: surface, and B: bottom.

FIGURE 4

Relative abundances (%) of the main bacterial phyla/classes in Salina Preta and Salina Verde lakes linked to Nitrogen (A), Sulfur (B), and Arsenic (C) metabolisms. Data obtained using the SEED genomic approach and the FOCUS tool. This information at order level can be seen in Supplementary Table 2.