Spatial distribution of prokaryotic communities in hypersaline soils

Abstract

Increasing salinization in wetland systems is a major threat to ecosystem services carried out by microbial communities. Thus, it is paramount to understand how salinity drives both microbial community structures and their diversity. Here we evaluated the structure and diversity of the prokaryotic communities from a range of highly saline soils (EC_1:5 from 5.96 to 61.02 dS/m) from the Odiel Saltmarshes and determined their association with salinity and other soil physicochemical features by analyzing 16S rRNA gene amplicon data through minimum entropy decomposition (MED). We found that these soils harbored unique communities mainly composed of halophilic and halotolerant taxa from the phyla Euryarchaeota, Proteobacteria, Balneolaeota, Bacteroidetes and Rhodothermaeota. In the studied soils, several site-specific properties were correlated with community structure and individual abundances of particular sequence variants. Salinity had a secondary role in shaping prokaryotic communities in these highly saline samples since the dominant organisms residing in them were already well-adapted to a wide range of salinities. We also compared ESV-based results with OTU-clustering derived ones, showing that, in this dataset, no major differences in ecological outcomes were obtained by the employment of one or the other method.

Figure 1

Relative abundances of major phyla detected in the saline soil samples studied. (A) Mean abundance of major phyla in the complete dataset. Error bars indicate standard error of the mean. (B) Mean abundance values for each site and depth shown over a map of the sampling locations at the Odiel Saltmarshes (Huelva, SW Spain). Abundances are the mean of 6 replicates, except for the surface samples for sites 2 and 4, of which there are 5 replicates.

Figure 2

Distribution of ESVs across all samples, grouped by phylum. Points represent individual ESVs and their size indicate their mean abundance in the complete dataset. Phyla are sorted by median prevalence of their ESVs. The distribution of the data and its probability density are shown by violin plots.

Figure 3

RDA ordination plot including the soil variables that best explain community data, as determined by forward selection. It is based on Hellinger-transformed abundance data. It was significant at α = 0.05 according to Monte Carlo permutation tests. S: total sulfur, EC_1:5: electrical conductivity measure in an 1:5 w/v soil to water extract, WC: water content, P: phosphate, Mn: manganese and Al: aluminium.

Figure 4

Correlation of soil parameters to ESVs relative abundance. All shown correlations are significant at α = 0.05. The complete list of correlation coefficients and ESVs identity and taxonomy from this plot can be found in Supplementary Data S2.