Diversity and stratification of archaea in a hypersaline microbial mat

Abstract

The Guerrero Negro (GN) hypersaline microbial mats have become one focus for biogeochemical studies of stratified ecosystems. The GN mats are found beneath several of a series of ponds of increasing salinity that make up a solar saltern fed from Pacific Ocean water pumped from the Laguna Ojo de Liebre near GN, Baja California Sur, Mexico. Molecular surveys of the laminated photosynthetic microbial mat below the fourth pond in the series identified an enormous diversity of bacteria in the mat, but archaea have received little attention. To determine the bulk contribution of archaeal phylotypes to the pond 4 study site, we determined the phylogenetic distribution of archaeal rRNA gene sequences in PCR libraries based on nominally universal primers. The ratios of bacterial/archaeal/eukaryotic rRNA genes, 90%/9%/1%, suggest that the archaeal contribution to the metabolic activities of the mat may be significant. To explore the distribution of archaea in the mat, sequences derived using archaeon-specific PCR primers were surveyed in 10 strata of the 6-cm-thick mat. The diversity of archaea overall was substantial albeit less than the diversity observed previously for bacteria. Archaeal diversity, mainly euryarchaeotes, was highest in the uppermost 2 to 3 mm of the mat and decreased rapidly with depth, where crenarchaeotes dominated. Only 3% of the sequences were specifically related to known organisms including methanogens. While some mat archaeal clades corresponded with known chemical gradients, others did not, which is likely explained by heretofore-unrecognized gradients. Some clades did not segregate by depth in the mat, indicating broad metabolic repertoires, undersampling, or both.

FIG. 1.

Relative proportions of the three domains of life seen in the GN pond 4 mat and a comparison of bacterial and archaeal species richnesses. (A) The pie chart shows the relative proportions of each domain as detected in 557 16S rRNA sequences obtained with universal primers from core samples of the pond 4 mat taken in 2001. The archaea are further subdivided into the two main phylogenetic clades, the Euryarchaeota and the Crenarchaeota. (B) Measured (Sobs) and estimated (SChao1) (32) species richnesses versus sequence identity are shown for bacterial and archaeal sequences found with all primer pairs in the 2001 mat bulk samples. Rarefaction was used to compensate for the differences in numbers of observed bacteria (1,584) and archaea (282).

FIG. 2.

Species richness and diversity of archaea vary by depth in the mat. (A) The observed (Sobs) and estimated (SChao1) (32) species richnesses of archaea decrease with depth in the mat. Rarefaction was used to compensate for differences in the numbers of sequences per layer (see Table S3 in the supplemental material) in the 817 archaeal sequences from the 2005 mat core samples. (B) The number of 99% OTUs versus depth for the Euryarchaeota and Crenarchaeota. Euryarchaeota dominate the diversity of archaeal species in the mat and are almost the exclusive archaea in the top 2 mm. Crenarchaeota are consistently present in the mat below 3 mm and begin to dominate diversity below 27 mm. (C) Simpson's reciprocal diversity index (32) indicates the highest diversity in the top 3 mm of the mat, at which point it decreases rapidly to a depth of 5 mm and remains relatively constant through to the deepest sample. Good's coverage index (32) shows approximate inverse proportionality to diversity and indicates that all layers of the mat may benefit from sequencing beyond what was attempted in this study. Indices are shown for 99% OTUs computed on the 817 archaeal sequences from the 2005 mat core samples.

FIG. 3.

Phylogenetic trees of members of the Euryarchaeota and Crenarchaeota of the pond 4 mat. Bootstrap support values (percent) are indicated at the bases of branches. Branches with less than 70% bootstrap support were collapsed to better reveal the statistically supported structure of the trees. GenBank sequences reported elsewhere that are present in each clade are indicated with accession numbers along with the distance to the closest sequence in this study. Clones show total numbers of clones from all samples and primer pairs, numbers of clones from primers 4FA/515F/333FA/1391R, and numbers of clones from primer 23FA/1391R. The ICDs were obtained by the identification of the largest uncorrected distance in the sequence distance matrix calculated by the ARB neighbor-joining distance matrix calculation function applied to the sequences of interest, masked for hypervariability, within a specific clade. (A) The Euryarchaeota. The Euryarchaeota clade names were selected in order of analysis except for GNMethanos (cluster with classic methanogens), GNHalos (cluster with the halobacteria), and GNThrm, which cluster with sequences in the Greengenes taxonomy Thermoplasmata/E2/terrestrial and Thermoplasmata/E2/aquatic group (III). (B) The Crenarchaeota. The Crenarchaeota clade names were selected based upon similarity to previously published sequences and are prefixed as follows: DV, deep-sea hydrothermal vent; MB, marine benthic; MV, mud volcano. MBCrenUn is a synthetic clade formed from all of the nonclade sequence singletons and doubletons found within the MBCren branch of the tree shown in B.

FIG. 3.

Phylogenetic trees of members of the Euryarchaeota and Crenarchaeota of the pond 4 mat. Bootstrap support values (percent) are indicated at the bases of branches. Branches with less than 70% bootstrap support were collapsed to better reveal the statistically supported structure of the trees. GenBank sequences reported elsewhere that are present in each clade are indicated with accession numbers along with the distance to the closest sequence in this study. Clones show total numbers of clones from all samples and primer pairs, numbers of clones from primers 4FA/515F/333FA/1391R, and numbers of clones from primer 23FA/1391R. The ICDs were obtained by the identification of the largest uncorrected distance in the sequence distance matrix calculated by the ARB neighbor-joining distance matrix calculation function applied to the sequences of interest, masked for hypervariability, within a specific clade. (A) The Euryarchaeota. The Euryarchaeota clade names were selected in order of analysis except for GNMethanos (cluster with classic methanogens), GNHalos (cluster with the halobacteria), and GNThrm, which cluster with sequences in the Greengenes taxonomy Thermoplasmata/E2/terrestrial and Thermoplasmata/E2/aquatic group (III). (B) The Crenarchaeota. The Crenarchaeota clade names were selected based upon similarity to previously published sequences and are prefixed as follows: DV, deep-sea hydrothermal vent; MB, marine benthic; MV, mud volcano. MBCrenUn is a synthetic clade formed from all of the nonclade sequence singletons and doubletons found within the MBCren branch of the tree shown in B.

FIG. 4.

Phylogenetic clades versus depth in the mat for the sectioned core samples from 2005. (A) The Euryarchaeota. The height of the three-dimensional cones represents the percentage of total sequences (817) for each depth sample point for each euryarchaeal clade (solid filled circles represent 0%). The order of the clades was selected to minimize visual interference among the data points. Although most numerous and diverse at the top of the mat, the Euryarchaeota are present in all layers sampled. (B) The Crenarchaeota. As in A, the height of the cones is the percentage of total sequences from the 2005 sectioned core samples. The Crenarchaeota show the highest abundance in the deeper layers of the mat. The largest crenarchaeal clade, MBCrenUn, is composed of many species but does not readily segregate by depth distribution or phylogenetic bootstrap support. (C) Additional detail showing minority euryarchaeal clades found in the top 5 mm of the mat, including the mat methanogen clade GNMethanos.

FIG. 4.

Phylogenetic clades versus depth in the mat for the sectioned core samples from 2005. (A) The Euryarchaeota. The height of the three-dimensional cones represents the percentage of total sequences (817) for each depth sample point for each euryarchaeal clade (solid filled circles represent 0%). The order of the clades was selected to minimize visual interference among the data points. Although most numerous and diverse at the top of the mat, the Euryarchaeota are present in all layers sampled. (B) The Crenarchaeota. As in A, the height of the cones is the percentage of total sequences from the 2005 sectioned core samples. The Crenarchaeota show the highest abundance in the deeper layers of the mat. The largest crenarchaeal clade, MBCrenUn, is composed of many species but does not readily segregate by depth distribution or phylogenetic bootstrap support. (C) Additional detail showing minority euryarchaeal clades found in the top 5 mm of the mat, including the mat methanogen clade GNMethanos.