Compositional Changes in Sediment Microbiota Are Associated with Seasonal Variation of the Water Column in High-Altitude Hyperarid Andean Lake Systems

Abstract

The lacustrine systems of La Brava and La Punta, located in the Tilopozo sector in the extreme south of Salar de Atacama, are pristine high-altitude Andean lakes found along the central Andes of South America. This shallow ecosystem suffers from permanent evaporation, leading to falling water levels, causing it to recede or disappear during the dry season. This dynamic causes physicochemical changes in lakes, such as low nutrient availability, pH change, and dissolved metals, which can influence the composition of the microbial community. In this study, we used a metataxonomic approach (16S rRNA hypervariable regions V3 to V4) to characterize the sedimentary microbiota of these lakes. To understand how the water column affects and is structured in the microbiota of these lakes, we combined the analysis of the persistence of the water column through satellite images and physicochemical characterization. Our results show a significant difference in abiotic factors and microbiota composition between La Punta and La Brava lakes. In addition, microbiota analysis revealed compositional changes in the ecological disaggregation (main and isolated bodies) and antagonistic changes in the abundance of certain taxa between lakes. These findings are an invaluable resource for understanding the microbiological diversity of high Andean lakes using a multidisciplinary approach that evaluates the microbiota behavior in response to abiotic factors. IMPORTANCE In this study, we analyzed the persistence of the water column through satellite images and physicochemical characterization to investigate the composition and diversity in High Andean Lake Systems in a hyperarid environment. In addition to the persistence of the water column, this approach can be used to analyze changes in the morphology of saline accumulations and persistence of snow or ice; for example, for establishing variable plant cover over time and evaluating the microbiota associated with soils with seasonal changes in plants. This makes it an ideal approach to search for novel extremophilic microorganisms with unique properties. In our case, it was used to study microorganisms capable of resisting desiccation and water restriction for a considerable period and adapting to survive in ecological niches, such as those with high UV irradiation, extreme drought, and high salt concentration.

Keywords: Salar de Atacama; hyperaridity; hypersalinity; metataxonomics; sediment microbiota.

FIG 1

La Punta–La Brava lake system. Distribution of sediment sampling points for La Punta and La Brava lakes in the main bodies with persistent water columns (blue) and isolated bodies with temporary water columns (red).

FIG 2

Principal coordinates analysis by the unifrac distance of dissimilarity. Left panel, Lake (PERMANOVA, P = 0.00009999), unifrac sediment level (PERMANOVA, P = 0.0289), and right panel, unifrac distance in ecological disaggregation (PERMANOVA, P = 0.0002).

FIG 3

Microbial community composition in La Punta–La Brava lake system. Abundance for lake (La Punta and La Brava) and ecological disaggregation to phylum and genus levels plot top and bottom, respectively.

FIG 4

Differential abundance between La Punta and La Brava. Test to phylum level (Krustall Wallis; P > 0.05).

FIG 5

Differential abundance between main and isolated body per lake. The left panel contains abundant taxa, indicating a significant difference between La Punta and La Brava. The right panel shows a significant difference in abundance between the main body and isolated body for each lake. (Krustall Wallis; P > 0.05).

FIG 6

Relations between microbiota and physical-chemical variables. Spearman rank correlation matrix of the bacterial phylum with >1% abundance in at least one sample. Each box shows the Spearman correlation for each taxon (rows) and physicochemical variable (columns). In addition, the prevalence of each taxon and its relative abundance across the samples are added.