Microbial biogeography of the eastern Yucatán carbonate aquifer

ABSTRACT Constraining the spatial distribution of microorganisms and their ecological interactions is crucial for informing biogeochemistry. To that end, we explore horizontal and vertical patterns of microbial biogeography in the eastern Yucatán carbonate aquifer by examining the relative abundance of microbial taxa via 16S rRNA gene sequencing. As one of the largest anchialine groundwater systems on Earth, the density-stratified Yucatán aquifer consists of a meteoric lens overlying saline groundwater. The myriad sinkholes (cenotes) of the eastern peninsula lead into a vast network of subsurface conduits. Several studies describe microbial communities within specific regions of the aquifer, yet fundamental questions remain regarding the ecology and distribution of biogeochemically relevant microbes. Our analysis demonstrates that this aquifer hosts a distinct microbiome from nearby seawater, with regionalism observed across cave systems and vertical water column zones. We apply novel software to construct taxonomic co-occurrence networks at different scales and categorize highly connected groups of taxa into potential niches. Our network analysis approach suggests that ubiquitous, metabolically flexible taxa such as the family Comamonadaceae act as ecological linchpins across several niches, often directly or indirectly co-occurring with taxa capable of anammox (e.g., Gemmataceae), methanotrophy (e.g., Methyloparacoccus), or organoheterotrophy. Furthermore, communities from a deep, pit-like cenote open to the surface show the strongest niche partitioning between water column zones, differing from those encountered throughout the mostly dark and oligotrophic aquifer system, including another deep pit cenote with no direct surface opening. Our results suggest that members of a core microbiome could modulate different biogeochemical regimes depending on location, acting as reservoirs of metabolic potential in disparate environments of this groundwater system. IMPORTANCE The extensive Yucatán carbonate aquifer, located primarily in southeastern Mexico, is pockmarked by numerous sinkholes (cenotes) that lead to a complex web of underwater caves. The aquifer hosts a diverse yet understudied microbiome throughout its highly stratified water column, which is marked by a meteoric lens floating on intruding seawater owing to the coastal proximity and high permeability of the Yucatán carbonate platform. Here, we present a biogeographic survey of bacterial and archaeal communities from the eastern Yucatán aquifer. We apply a novel network analysis software that models ecological niche space from microbial taxonomic abundance data. Our analysis reveals that the aquifer community is composed of several distinct niches that follow broader regional and hydrological patterns. This work lays the groundwork for future investigations to characterize the biogeochemical potential of the entire aquifer with other systems biology approaches.

Table S1: Number of quality-controlled reads per sample.Table S2: Sample metadata.Table S3: Phylum-level abundance table.Table S4: ASV-level taxonomic abundance data rarefied to a sampling depth of 9,957.Table S5: Global network ASV cluster prevalence table.Table S6: Global network ASV cluster per-sample abundance table.Table S7: Global network ASV cluster memberships for each ASV-level taxon.Table S8: List of taxa node ASV cluster membership for each cave system.Saved as an Excel table with each tab corresponding to one region.

Supplemental Figures
All supplemental figures are included in this document unless otherwise specified.

Figure S1 :
Figure S1: Geochemical variables across study sites.Figure S2: Interactive global network (separate HTML file), filled by ASV cluster.Figure S3: Regional co-occurrence networks, filled by ASV cluster.Figure S4: Relative abundance of regional network ASV clusters.Figure S5: Co-occurrences of the unclassified Comamonadaceae bin with network nodes across the global and regional networks.Figure S6: Relative abundance of selected taxa putatively capable of sulfur cycling.

Figure S2 :
Figure S1: Geochemical variables across study sites.Figure S2: Interactive global network (separate HTML file), filled by ASV cluster.Figure S3: Regional co-occurrence networks, filled by ASV cluster.Figure S4: Relative abundance of regional network ASV clusters.Figure S5: Co-occurrences of the unclassified Comamonadaceae bin with network nodes across the global and regional networks.Figure S6: Relative abundance of selected taxa putatively capable of sulfur cycling.

Figure S3 :
Figure S1: Geochemical variables across study sites.Figure S2: Interactive global network (separate HTML file), filled by ASV cluster.Figure S3: Regional co-occurrence networks, filled by ASV cluster.Figure S4: Relative abundance of regional network ASV clusters.Figure S5: Co-occurrences of the unclassified Comamonadaceae bin with network nodes across the global and regional networks.Figure S6: Relative abundance of selected taxa putatively capable of sulfur cycling.

Figure S4 :
Figure S1: Geochemical variables across study sites.Figure S2: Interactive global network (separate HTML file), filled by ASV cluster.Figure S3: Regional co-occurrence networks, filled by ASV cluster.Figure S4: Relative abundance of regional network ASV clusters.Figure S5: Co-occurrences of the unclassified Comamonadaceae bin with network nodes across the global and regional networks.Figure S6: Relative abundance of selected taxa putatively capable of sulfur cycling.

Figure S5 :
Figure S1: Geochemical variables across study sites.Figure S2: Interactive global network (separate HTML file), filled by ASV cluster.Figure S3: Regional co-occurrence networks, filled by ASV cluster.Figure S4: Relative abundance of regional network ASV clusters.Figure S5: Co-occurrences of the unclassified Comamonadaceae bin with network nodes across the global and regional networks.Figure S6: Relative abundance of selected taxa putatively capable of sulfur cycling.

Figure S6 :
Figure S1: Geochemical variables across study sites.Figure S2: Interactive global network (separate HTML file), filled by ASV cluster.Figure S3: Regional co-occurrence networks, filled by ASV cluster.Figure S4: Relative abundance of regional network ASV clusters.Figure S5: Co-occurrences of the unclassified Comamonadaceae bin with network nodes across the global and regional networks.Figure S6: Relative abundance of selected taxa putatively capable of sulfur cycling.

Figure
Figure S1: Geochemical variables across study sites.A. Conductivity profiles of each site reported in mS/cm.Points are shaped based on the water column zone and colored by Cave group.

S1: Geochemical variables across study sites. A
. Conductivity profiles of each site reported in mS/cm.Points are shaped based on the water column zone and colored by Cave group.Ionic composition of water samples plotted by cave.Seawater concentrations measured in controls are show by dotted vertical lines.All values are reported in milliequivalents per liter (mEq/L), but sulfate is shown at 10X exaggeration for more direct comparison.Symbols reflext reported water column zone for each sample. B:

S2: Interactive global network, filled by edge betweenness cluster (EBC). Open
the attached file "SuppFig2.html" in a web browser to probe individual pairwise relationships.Refer to Figure 4B in the main text for a static version of this figure.Refer to the main text for discussion.