INTRODUCTION
The intricacies of the deep subsurface have triggered many investigations into the microbial habitability sustaining itself in the rock-hosted subsurface. The chemical drivers that govern the transition from terrestrial surface to subsurface life are not fully understood, especially in near-subsurface zones where the system may behave as a hybrid between surface and subsurface processes. Field sites spanning terrestrial, lacustrine, hot springs, and marine environments have gained traction for molecular biological investigation due to unique, extreme conditions mimicking potential environmental settings of early life on Earth as well as analogs to those on other planetary systems. While marine systems have unveiled the extent of microbial subsurface habitability in oceanic settings, research aimed at understanding accessible terrestrial environments has gained considerable interest because of more uniquely limiting constraints (low nutrient access, low O2 availability, and elevated temperatures) toward microbiological habitability.
The large biodiversity of microbial life in the deep biosphere remains to be explored through terrestrial subsurface investigations. Terrestrial sites can offer further geologic and hydrologic variability than marine environments due to the presence of unconstrained fluid sourcing and lithology in surface soils and bedrock. Even more, water-rock interactions demonstrated in terrestrial subsurface environments distinguish the chemical disparity of chemolithotrophic processes that sustain ecosystems of low biomass utilizing restricted sources of energy. The contemporary outlook of an "extreme" environment encompasses some variation of a chemical or hydrological endpoint for microorganisms to exist primarily due to physiological constraints. However, new insights into shallow and deep subsurface research may not necessitate harsh geochemical gradients or temperatures. Instead, restricted access to nutrients, vital for microbial activity, makes terrestrial subsurface studies of interest to investigate how moderately warm and/or acidic environments can still be uniquely constraining to microbiota.
In order to study subsurface microbiology, many continental deep subsurface observatories have been established. Reaching depths of up to ∼2.4 km, the Sanford Underground Research Laboratory in South Dakota has investigated the geochemistry and DNA sequences of up to 10,000-year-old groundwater to understand the energetics of chemolithotrophic organisms. These subsurface laboratories extend across multiple continents and countries, including the United States, Finland, France, Japan, and the United Kingdom. Another exists in Canada at the Kidd Creek Observatory, where fracture waters at 2.4 km in depth were acquired in order to cultivate microorganisms from the fluids and determine the metabolic energetic pathways from parallel geochemical analysis. Onstott et al. have established the presence of a microbial community in a gold mine in the deep subsurface from fluids obtained up to 3 km below the surface. These fluids proved microbiota indigenous to the host rock environment and further demonstrate the potential for life driven by equilibration of water-rock interactions. Microorganisms that exist in the secluded hematite iron formation of the Soudan Underground Mine State Park were examined with metagenomic analyses to determine genes representing diverse metabolic strategies from a low biomass system to overcome nutrient limitations.
These established observatories in the deep subsurface offer novel investigations into niche environmental conditions. However, the scope of which microorganisms survive in shallow subsurface environments from young meteoric groundwater infiltration from the surface versus chemical energy sources originating in local settings remains insufficiently answered. Studies have utilized deep fracture fluids or ancient aquifers to constrain life in the subsurface, yet only limited research has considered the influence of meteoric water penetration into the near subsurface. The role of young meteoric groundwater infiltration as a transportation mechanism for nutrients introduces questions regarding the greater potential for life, the connectivity of the surface to the subsurface, and the relation of surface processes to subsurface microorganisms.
An additional consideration for deep observatories is the cost and accessibility of boreholes for sampling and inquiry. Due to expenses from boreholes and gaining access to subsurface environments, numerous questions remain unanswered from both marine and terrestrial subsurface ecosystem exploration. Therefore, easier access into the subsurface is advantageous not only to establish consistent research regarding the microbial communities existing in the subsurface but also to monitor geochemical variables influencing subsurface habitability with greater control over experimental design and statistical power.
This paper discusses the advantages of studying subsurface microbiology at the Edgar Experimental Mine, an easily accessible portal to the subsurface with ample opportunities to unveil and experiment with the subsurface biosphere. Once an active economic mining site, it was repurposed for teaching demonstrations by Colorado School of Mines for mining engineering practices. Four boreholes drilled from previous mining teaching practices are the focus of this study, located centimeters to meters apart along the same wall of our field site. Meteoric groundwater infiltrating the fractures of the mine leak out of the boreholes, offering the opportunity to study the subsurface microbiology of meteoric water influenced by young fluid-rock equilibration (Fig. 1). Additionally, the periodic (∼2 weeks) to seasonal recharge of water makes it possible to study temporal influences through incubation and recharge periods. Here, we apply genomic, geochemical, and isotopic approaches to better define the variable microbial composition of borehole fluids and their relationship to surface influences from meteoric penetration into the subsurface.
RESULTS
Borehole geochemistry.
Over the 4-year sample collection period, the subsurface boreholes remained relatively uniform in their recorded analyte and physical chemistry measurements. While boreholes minorly fluctuate in analyte (major anion and cation) concentration, no discrete patterns were observed. Despite a lack of change, each borehole hosts nuanced geochemical profiles, providing niche microenvironments for microbial communities. Fluid recharge in the boreholes was unpredictable regarding volume and temporal repetition. Borehole 2 was able to consistently fill with greater than 1 liter of water each sampling period, whereas all other boreholes required months or years before refilling with fluid (Fig. 1). The remaining boreholes did not refill with predictable volumes of water at each sampling period, further emphasizing a complex fracture pathway for fluid transport. Physical parameters and chemical composition remained relatively homogenous among boreholes. Water temperature ranges varied seasonally, generally fluctuating between 10.6 and 13.5°C and reaching 17.6°C in summer. Dissolved oxygen varied most strongly from 1.97 to 7.7 mg/liter, while pH was generally uniform within individual boreholes but varied among all samples ranging from pH 5.30 to 7.16.
Ion chromatography (IC) and inductively coupled plasma atomic emission spectroscopy (ICP-AES) results of borehole fluids demonstrated weakly variable subsurface fluid chemistry. Concentrations of anions (putative electron acceptors for microbial growth as well as phosphorous sourcing) tend to be very low or undetected, allowing for selective substrate availability of microbial communities. Fluoride concentrations across all boreholes ranged from below detection or from 0.15 to 0.33 ppm. Nitrite concentrations were undetected except for three samples that recorded values of 0.19, 0.27, and 0.37 ppm. Nitrate remained undetected or reported values ranging from 0.45 to 1.27 ppm. Chloride concentrations varied unexpectedly within individual boreholes with concentrations between 1.59 and 22.08 ppm but spiked as high as 474.85 ppm. Sulfate was the most concentrated analyte among all boreholes and consistently remained high (usually greater than 1,000 ppm). Analytes chosen from IC and ICP analysis were selected to demonstrate greatest variability among the borehole fluids (Table S1 in the supplemental material). These analytes were evaluated with a principal-component analysis (PCA) to observe the relationship of geochemical measurements collected from the individual boreholes (Fig. 2). Of interest, principal component 2 appears to suggest a strong delineation driven by metals concentrations. In particular, manganese and zinc were the highest notable metal concentrations with maximum values of 27.15 mg/liter and 18.46 mg/liter, respectively. There is a linear relationship between the dissolved manganese and zinc concentrations in the borehole fluids (Fig. 3). This highlights the distinction between the metal-rich fluids found in boreholes 2 and 3 and more dilute fluids demonstrated in boreholes 6 and 8. Additionally, this relationship between the boreholes shows the consistency of borehole fluid chemistry despite sampling period and how this can serve as an indicator for variation in microbial community composition.
The δ18O and δ2H signatures for borehole fluids were compared to the global meteoric water line (GMWL) along with samples from other subsurface systems or observatories (Fig. 4). This trendline serves as an investigative tool for understanding if the subsurface fluids have undergone some extent of low-temperature water-rock interactions or remain isotopically conserved to reflect meteoric recharge from the surface. Samples are plotted above the GMWL with slightly enriched δ2H values. Edgar Mine is located at greater elevation and drier climate, which often causes enriched δ2H values of meteoric water as a result from the rainout effect.
Microbial community analysis.
16S rRNA gene sequence analyses were performed on filtered water to determine the microbial community composition within each borehole. Several samples from each borehole were taken to contrast composition over time. The fluid samples were dominated by Bacteria (98.5%) with only a small presence of Archaea (1.5%) belonging predominantly to samples from borehole 2. All boreholes were collectively dominated by the phylum Proteobacteria (39.2%) and also contained sequences for divisions of Firmicutes (16.1%) and Bacteroidota (13.0%) known to represent a diverse group of microorganisms.
Ordination via principal-coordinate analysis (PCoA) with a weighted UniFrac distance matrix was performed to determine statistical similarity of microbial communities from each borehole (Fig. 5). Ordination suggested strong clustering as a result of grouping by borehole location. Additional soil samples were collected from the surface of the mine to confirm the distinctive compositions of borehole microbiota. An Adonis test was run on the dissimilarity matrix to verify sampling source as distinct and significant (R2 = 0.44, P = 0.001). Weighted UniFrac distance matrices were also tested by Adonis for the number of days since the initial extraction of fluid from the boreholes, the number of days between sampling events, and the season in which samples were collected. Results demonstrated the number of days since the initial extraction period as significant (R2 = 0.38, P = 0.001) as well as seasonality (R2 = 0.36, P = 0.001) and the number of days between sampling events as modestly significant (R2 = 0.073, P = 0.047). These values suggest that each borehole is host to unique microbial community compositions despite their relatively close proximity and similarity of host rock. The inherent covariance in our study between days since initial extraction and seasonality suggests that further work is needed to decipher these two signals. Nonetheless, a clear shift in community composition occurs over time, predominantly in boreholes 2 and 3. However, the source of this change requires additional sampling to account for unsampled seasons and how microbial diversity may change with even more time allowed to pass within the boreholes.
To further investigate the nuances in microbial community composition and shifts occurring over the course of the experiment, differential abundance tests revealed the top 15 most abundant genera in the boreholes and soil samples (Fig. 6 and 7). Boreholes 2 and 3 appear to host similar microbial communities with subtle changes in taxa present and changes over the course of the sampling period. The boreholes share a large relative abundance of Desulfosporosinus, a potential sulfate-reducing bacteria, and Sulfurifustis, a potential sulfur-oxidizing bacteria, from each sampling occurrence. Boreholes 2 and 3 exhibit the same relationship where an increased relative abundance of Desulfosporosinus leads to a dramatic decrease in the relative abundance of other microbial members within the top 15 most abundant genera of fluid samples from boreholes 2 and 3. This stark contrast in microbial composition is demonstrated from the last two sampling periods, after 948 days have passed into the experiment (Fig. 6). In borehole 2, this shift appears to drastically change the microbial composition where the relative abundance of Desulfosporosinus continues to increase, and there is little relative abundance represented from additional microorganisms. In contrast, borehole 3 displays a shift in composition at this same sampling instance, but Desulfosporosinus decreases in relative abundance while other taxa increase (Fig. 6). Additionally, this shift does not last through the final sampling period, and Desulfosporosinus increases in relative abundance similar to earlier fluid samples. The other taxa present in the borehole fluid samples are indicators of metal-rich environments, including genera such as Candidatus Nitrotoga and Chelatococcus.
Comparatively, the microbial community composition of boreholes 6 and 8 appear to be fixed without any distinctive influence as observed in boreholes 2 and 3. However, the genera observed in boreholes 6 and 8 juxtapose the microbial communities observed in boreholes 2 and 3, where the microorganisms represented do not indicate metal-rich environments.