Nitrate-reducing microorganisms prevent souring of an oil field produced water storage pond

Abstract

Nitrate addition for mitigating sulfide production in oil field systems has been studied in laboratory settings and in some subsurface oil reservoirs. To promote water recycling and reuse associated with oil reservoirs produced by hydraulic fracturing, high-salinity produced waters are temporarily stored in surface ponds prior to subsequent reinjection into the subsurface. In this study, nitrate was added directly to a storage pond to prevent sulfide accumulation. DNA sequencing of pond water over a 4-week period revealed a decrease in the proportion of sulfate-reducing microorganisms following nitrate application. Sulfate levels remained stable during this period, whereas nitrate and nitrite fluctuated in the days following the nitrate addition. Metagenome-assembled genomes (MAGs) reconstructed from the pond water microbiome highlighted different organisms with genes for organoheterotrophic and lithoheterotrophic nitrate reduction, whereas genes associated with sulfide production via sulfate or thiosulfate reduction were barely detected. Within those MAGs, genes for acetate metabolism were observed, consistent with acetate decreasing substantially in the pond water in the presence of nitrate. After nitrate was consumed an increase in relative abundance of putative autotrophic microorganisms was observed (e.g. Arhodomonas, Guyparkeria, and Psychroflexus), corresponding to a drop in total inorganic carbon measurements in the storage pond. This trial offers an overview on microbial processes taking place in storage pond environments in response to nitrate addition.

Introduction

The presence and influence of microorganisms in oil fields can be detrimental in many instances related to biofouling, undesired plugging, microbially influenced corrosion, and souring (Enning and Garrelfs 2014, Basafa and Hawboldt 2019, Nicoletti et al. 2022). Certain microbial populations can also be beneficial in these settings and offer cost-effective solutions to different problems. Of particular interest is controlling sulfide production (souring) by adding nitrate to stimulate nitrate-reducing microorganisms (NRM) that combat the activity of sulfate-reducing microorganisms (SRM) and other sulfidogens (Hubert and Voordouw 2007, Kumaraswamy et al. 2011, An et al. 2017, Suri et al. 2017, Carlson and Hubert 2019). Oxidation of organic or inorganic electron donors coupled to nitrate reduction by organotrophic or lithotrophic NRM is more thermodynamically favorable than coupling the same oxidative processes to sulfate reduction; hence, nitrate injection can block the activity of SRM that use the same electron donors (Carlson and Hubert 2019). Furthermore, some lithotrophic NRM can couple the reduction of nitrate to sulfide oxidation, effectively catalyzing the opposite of the souring reaction (Carlson and Hubert 2019). Nitrate reduction can lead to an increase in nitrite, resulting in sulfate reduction being inhibited (Nemati et al. 2001, Greene et al. 2003).

Produced water storage ponds allow temporary retention of a mixture of water, residual hydrocarbons, and organic and inorganic compounds following separation of water from produced oil and gas (Lyman et al. 2018). This strategy allows for produced water to be reused and reinjected into the subsurface to avoid reliance on freshwater. The reuse of produced water involves its storage during a prolonged period before its reuse, and the water can quickly turn sour if not properly maintained (Shi et al. 2023). Common constituents of produced water microbiomes are members of the genera Halanaerobium, Halomonas, Idiomarina, and Marinobacter, as summarized in Kashani et al. (2024). Members of the genus Halanaerobium have been shown to be sulfidogenic (via thiosulfate reduction), while members of the three other genera are notable for potential roles in souring prevention in response to nitrate addition (Suri et al. 2017, Nicoletti et al. 2020). Preventing biofouling and souring within produced water storage ponds is therefore critical to ensuring the integrity of injection wells used in hydraulic fracturing. While the addition of oxidizing agents has been used by industry to remove sulfide and control bacteria, their effect is not consistent (Shi et al. 2023) such that the use of other techniques is required. In light of this, nitrate was added directly to a produced water storage pond to investigate souring control over a 4-week period. Geochemical analyses coupled with amplicon and metagenomic sequencing enabled monitoring a microbial succession within the pond microbiome in response to nitrate addition.

Materials and methods

Sample collection and handling

Samples were collected from a produced water storage pond associated with hydraulic fracturing operations in the Permian Basin (TX, USA). The produced water originated from multiple wells, with geochemistry similar to what has been reported previously (Jiang et al. 2022). Water was collected 1 day before nitrate was added (day −1), on the day of the nitrate addition (day 0) and then periodically over the course of 28 days after the nitrate addition. Nitrate dosing to the pond was calculated based on stochiometry related to available carbon sources, with a 20%–30% margin. In total, sampling was performed on days −1, 0, 1, 3, 7, 14, 21, and 28, consisting of collecting 2 l that was frozen immediately and then transported to the laboratory for further analyses.

Chemical measurements

Samples were thawed and 4-ml aliquots were passed through a 0.2-µm polyethersulfone filters (VWR; Mississauga, Ontario, Canada) and stored at −20°C until analyses. Sulfate, thiosulfate, nitrate, and nitrite levels were assessed by ion chromatography (Dionex ICS-5000) using an analytical column (AS23) with 8 mM Na₂CO₃/1 mM NaHCO₃ eluent at a flow rate of a 1 ml/min. The absence of sulfide production in the pond (personal communications) and the time elapsed during shipment to the laboratory did not allow for confident laboratory-based sulfide measurements. Chromeleon software was used for visualization. Peak area calibration used Na₂SO₄ standard solutions. Volatile fatty acid levels were measured using high-performance liquid chromatography in an Ultimate 3000 RSLC system with a 5 mM H₂SO₄ mobile phase at a flow rate of 0.6 ml/min and a temperature of 60°C using a Bio-Rad Aminex HPX-87H column. Colleagues at the Canada Excellence Research Chair in Materials Engineering analyzed organic and inorganic carbon levels within each sample. To perform those analyses, a total organic carbon (TOC) analyzer from Shimadzu (Kyoto, Japan) was operated with the following parameters: the flow gas (purified air) is passed at 150 ml/min through an oxidation catalyst-filled combustion tube and heated to 680°C. Results were converted from mg/L to mM concentrations of carbon. Inorganic carbon concentrations were obtained by substracting the organic carbon concentration from the total carbon values.

DNA extraction and 16S rRNA gene sequencing

For further DNA analyses, all samples were passed through a 0.1-µm polyethersulfone membrane filter using an autoclaved, UV-sterilized filtration unit. A total of 400 ml was filtered for each sample. Filters were stored within Powersoil PowerBead tubes (Qiagen, Hilden, Germany) and kept at −20°C until DNA extraction was performed. Upon thawing, genomic DNA was extracted using the DNeasy PowerSoil extraction kit (Qiagen) following the manufacturer’s instructions. Resulting DNA concentrations were measured using a Qubit fluorometer (Qiagen, Hilden, Germany). DNA was used for both 16S rRNA gene amplicon sequencing (described later) and shotgun metagenomic sequencing (see the “Metagenomic sequencing and analyses” section).

The V4 hypervariable region of the 16S rRNA gene was amplified using primers 515F (GTGYCAGCMGCCGCGGTAA) and 806R (GGACTACNVGGGTWTCTAAT) (Wei et al. 2023). Polymerase Chain Reaction (PCR) with a total volume of 25 µl included: 2 × KAPA HiFi Hot Start Ready Mix (KAPA Biosystems, Wilmington, MA, USA), a final concentration of 0.1 mM for each primer and 7.5 µl of DNA sample. PCR included an initial denaturation at 95°C for 5 min followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 45 s and extension at 72°C for 1 min. This was followed by a final extension for 5 min at 72°C. Triplicate PCR products were pooled prior to cleanup and indexing following the manufacturer’s instructions for sequencing using Illumina’s v3 600-cycle (paired-end, 2 × 250 bp) reagent kit on an in-house MiSeq benchtop sequencer (Illumina, San Diego, CA, USA). Prior to sequencing, all DNA extraction blanks and PCR reagent blanks were confirmed to be negative for amplification with the absence of bands on an agarose gel. Quality-controlled reads from eight different samples were merged and subsequently dereplicated to construct an amplicon sequence variant (ASV) table using DADA2 on R (DADA2 2024). Relative abundance calculations were based on the number of unrarefied reads per library, which ranged from 7645 to 102 326 reads. R and GraphPad Prism were used for visualization of the data. Relative abundance cut-offs are indicated in the figure legends (1% in Fig. 2, no cut-off in Fig. 3).

Metagenomic sequencing and analyses

Shotgun sequencing of genomic DNA was performed on the day 21 sample (Novogene, Sacramento, CA, United States). Adapters were removed, the last 151 bp of the last read was trimmed, PhiX spike-in sequences (PhiX control V3 library, Illumina) were filtered out, and low-quality ends were clipped off using BBDuk. Assembly was performed with Megahit (Li et al. 2015). Reads were mapped using Bowtie2 (Langmead and Salzberg 2012, Li et al. 2015). Binning of the contigs into metagenome-assembled genomes (MAGs) used Metabat (Kang et al. 2015). Quality assessment of the dereplicated bins was performed with CheckM “lineage_wf” (v.1.1.2) with MAG taxonomy assigned using GTDB-Tk “classify_wf” (v.1.5.0) (Parks et al. 2015, Olm et al. 2017, Chaumeil et al. 2020). Only bins with <15% redundancy and >50% completeness were retained for annotation and further study. DRAM (v.1.2) was used to annotate MAGs for functional potential (Shaffer et al. 2020), and phyloFlash allowed for taxonomic identification from the metagenomic reads (Gruber-Vodicka et al. 2020).

Results

Microbial activity in the produced water storage pond

Levels of nitrate, nitrite, thiosulfate, and sulfate in pond water were measured to assess the effectiveness of nitrate addition for souring prevention. Initially, a drop in sulfate of 4.9 mM prior to nitrate addition (Fig. 1) was consistent with sulfate reduction and the risk of souring, despite direct evidence of sulfide production not being observed by the facility operator. Following addition of nitrate (day 0), sulfate remained at a constant level, ∼12 mM, with no further sulfate reduction observed throughout the monitoring period. While thiosulfate can also be a precursor to souring, it was not detected in any of the samples suggesting that it was not present or lower than the 0.1 mM detection limit of the ion chromatography method used for analysis.

Nitrate consumption was observed within 1 week of its addition. Upon its addition at day 0, nitrate levels were ∼1.5 mM, and quickly decreased the next day (day 1) to 0.7 mM, eventually reaching 0.1 mM by day 7 (Fig. 1). Nitrite increased initially from 0.7 to 0.9 mM within 24 h of nitrate addition, and then decreased rapidly to reach a concentration of 0.1 mM after 3 days.

Water samples were also tested for volatile fatty acids to assess microbial activity within the pond. Screening for acetate, butyrate, formate, lactate, propionate, and succinate revealed that only acetate could be detected. Acetate levels remained stable at 0.7 mM before and after the initial addition of nitrate, and then started to decrease at 3 days post-addition, dropping to 0.4 mM. Acetate was no longer detected after 14 days. These acetate dynamics were reflected in a general decrease in TOC that was most pronounced over the first 7 days, with a corresponding increase in total inorganic carbon (TIC), consistent with acetate-driven organotrophic nitrate reduction during this period (Fig. 1). TIC then gradually decreased between 7 and 28 days, dropping from 3.1 to 1.7 mM. These changes were more gradual than the sharp changes in TOC and TIC before and after nitrate addition (days −1 and 0).

Pond water microbial community composition

Prior to nitrate addition (−1 days), the pond water microbial community was characterized by Halanaerobium and Marinobacter ASVs with relative abundances of 38% and 45%, respectively (Fig. 2A; Table S1). While a decrease in the relative abundance of Halanaerobium was observed over time following nitrate addition, levels of Marinobacter remained relatively high. Other genera also decreased in abundance in the days following nitrate addition, notably Fusibacter, Geotoga, and Izimaplasma. Groups showing an increase in abundance included Halomonas, Sulfurospirillum, Desulfuromonas, and members of the E6aC02 family. Halomonas exhibited the most notable enrichment in abundance during the monitoring period, increasing from 0.4% (pre-nitrate addition) to 13.6% after 14 days. Similarly, lineages that were below detection in early stages and then increased in relative abundance in later stages of the monitoring period were Alcanivorax, Aliifodinibius, Idiomarina, Methylophaga, Psychroflexus, members of the JGI069 class, and members of the Rhodobacteraceae and Arcobacteraceae families.

Despite the apparent sulfate consumption observed prior to nitrate addition to the storage pond (Fig. 1), putative SRM were detected in low relative sequence abundance in the amplicon libraries. Fig. 2C shows SRM reaching maximum levels of 0.4%–0.6% in the pond during an early stage in the monitoring. This signal declined below detection limit by the end of the monitoring period (i.e. <0.007% relative sequence abundance based on 13 967 reads in the day 28 amplicon library).

To better understand how nitrate metabolism prevents souring within the storage pond, metagenomic sequencing was performed after 21 days, i.e. the point at which amplicon sequencing suggested high representation of putative nitrate-reducing microorganisms (NRM) such as Halomonas and Idiomarina (Fig. 2A and B). Metagenomic reconstruction resulted in 19 MAGs with <15% redundancy and >50% completeness (Table 1). Out of the 13 taxonomic groups of ASVs identified to the genus level that were abundant (>1%) in the amplicon libraries (Fig. 2A), 8 were observed at similarly high abundance in quality-controlled metagenomic reads using phyloFlash (Fig. 2B). The most notable exception to this broad congruence is the discrepancy of Halanaerobium relative abundance being 28.9% based on amplicon sequencing compared to 8.7% in the metagenomic library. Metagenomic analysis also suggests that the decreasing abundance of Halanaerobium over 21 days may be more rapid and pronounced than the amplicon time course analysis reveals (Fig. 2).

MAGs for which taxonomy was assigned with GTDB-Tk enabled tracking corresponding ASVs in the amplicon libraries in order to link metabolic potential from MAG analysis with temporal dynamics in the pond microbiome from ASV analysis (Fig. 3). Groups experiencing early enrichment after nitrate addition (days 0–7) included Bacteroidales S143-32 (MAG 3, 17), Marinobacter (MAG 13, 14, 15), and Sulfurospirillum (MAG 18). Flexistipes (MAG 5) decreased in relative abundance between days 0 and 21, but increased from 0.2% to 1.1% between days 21 and 28 (Fig. 3). Later in the incubation (days 14–21), amplicon sequencing highlights increases in Balneolales (MAG 2), Arhodomonas (MAG 4), Guyparkeria (MAG 6), Halomonas (MAG 7, 8, 9), Idiomarina (MAG 10), and Psychroflexus (MAG 16).

Metabolic pathway reconstruction within MAGs enriched following nitrate addition

Nitrate reduction

MAGs of interest with respect to nitrogen metabolism (Fig. 4) are affiliated with Arhodomonas aquaeolei (MAG 4), Halomonas (novel species; MAG 7), Halomonas alimentaria (MAG 9), Marinobacter segnicrescens (MAG 14) and Sulfurospirillum (novel species; MAG 18), Bacteroidales (novel family S143-32; MAG 3), and Sulfurovaceae (novel genus TCS-49; MAG 3). Genes for the reduction of nitrate to dinitrogen (denitrification) were observed in A. aquaeolei (MAG 4) and H. alimentaria (MAG 9; Fig. 4). Potential for dissimilatory nitrate reduction to ammonia (DNRA) was observed in Bacteroidales (MAG 3), A. aquaeolei (MAG 4), Halomonas (MAG 7), and Sulfurospirillum (MAG 18; Fig. 4). Additional potential for nitrate reduction only to nitrite (i.e. narGHI or napAB genes) was also observed in M. segnicrescens (MAG 14) and Sulfuvoraceae (MAG 19) with genes for further nitrite reduction (complete denitrification or DNRA) potentially not captured (Fig. 4) due to these MAGs having 93%–99% estimated completeness (Table 1).

Lithotrophic metabolism

Genes involved in the oxidation of sulfide to sulfite (asrAB and cysJI) were seen within Bacteroidales S143-32 (MAG 3), A. aquaeolei (MAG 4), H. alimentaria (MAG 9), and Sulfurospirillum (MAG 18). S143-32 represents a family (MAG 3) within the Bacteroidales order reported at very low relative abundance within the metagenome (0.2%). Potential for sulfide oxidation to elemental sulfur (fccAB genes) was observed only within Sulfurovaceae TCS-49 (MAG 19) (Fig. 4). Genes for hydrogen metabolism (hypF) were found within Balneolales (MAG 2), Guyparkeria (MAG 6), Halomonas (MAG 7), and Izemoplasmataceae (MAG 11), and genes for [NiFe] hydrogenases were found in A. aquaeolei (MAG 4), Flexistipes (MAG 5), and Halomonas (MAG 7).

Carbon metabolism

Internalization of acetate can be mediated by the cation/acetate symporter encoded by the actP gene, found within Balneolales M1805 (MAG 2), Bacteroidales S143-32 (MAG 3), A. aquaeolei (MAG 4), Guyparkeria (MAG 6), Halomonas (MAG 7), Halomonas ventosae (MAG 8), Idiomarina bentica (MAG 10), Halanaerobiales M55B127 (MAG 12), Marinobacter persicus (MAG 13), Marinobacter (MAG 15), Psychroflexus solarius (MAG 16), and Sulfurospirilum (MAG 18; Figure S1). Of those MAGs, genes for acetate metabolism via an acetyl-CoA synthetase are present within Bacteroidales S143-32 (MAG 3), A. aquaeolei (MAG 4), Guyparkeria (MAG 6), and Sulfurovaceae TCS-49 (MAG 19). Genes for acetate kinase and phosphate acetyltransferase were found within A. aquaeolei (MAG 4), Halomonas (MAG 7), and M. persicus (MAG 13; Figure S1).

Carbon fixation via the Calvin–Benson–Bassham cycle is evidenced in genomes of Bacteroidales S143-32 (MAG 3), Guyparkeria (MAG 6), Izemoplasmataceae M55B118 (MAG 11), and Halanaerobiales M55B127 (MAG 12). Nearly complete sets of genes for the incomplete reductive citrate cycle were found within MAG 1 Desulfuromonadaceae (MAG 1), Bacteroidales S143-32 (MAG 3), H. ventosae (MAG 8), H. alimentaria (MAG 9), Psychroflexus salarius (MAG 16), Bacteroidales PWNQ01 (MAG 17), and Sulfurovaceae TCS-49 (MAG 19). Arhodomonas aquaeolei (MAG 4) was the only genome with a nearly complete set of genes for the reductive citrate cycle (Figure S1).

Discussion

Nitrate addition for souring control has been studied in different contexts in oil systems, most often in laboratory settings (Hubert and Voordouw 2007, Suri et al. 2017, Kamarisima et al. 2018, Wu et al. 2018, Nicoletti et al. 2022). Here, nitrate amendment directly to a produced water storage pond prevented the increase of SRM in a topsides setting (Fig. 2), and sulfate levels remained stable (Fig. 1). The monitoring period of 28 days mimicks the period to which the produced water usually sits in the storage pond before being reused for hydraulic fracturing operations. In addition to preventing souring in the pond, topsides removal of organic electron donors helps to minimize reservoir souring following reinjection of this water into the subsurface. Nitrate levels in the pond decreased in the days following its addition, coinciding with nitrite detection, consistent with the activity of NRM. Bacterial lineages associated with nitrate metabolism increased in relative abundance over the first 8 days of the monitoring period (Arhodomonas, Halomonas, and Sulfurospirillum; Figs 2B and 3), and five different MAGs with nitrate-related metabolism were recovered. Accordingly, sulfate levels remained stable after nitrate was added to the pond (Fig. 1), highlighting the efficacy of nitrate addition for preventing souring in surface storage pond settings.

The transience of nitrite accumulation in the pond indicates that nitrate reduction to nitrogen (denitrification) or ammonia (DNRA) achieved the observed souring control. Biocompetitive exclusion implies that NRM utilize electron donors that would otherwise be available for SRM to couple to sulfide production. In principle, DNRA achieves greater souring control by this mechanism, since it is an eight-electron transfer (NO₃⁻ → NH₄⁺) compared to denitrification being a five-electron transfer (NO₃⁻ → N₂) and thus consumes more substrate per nitrate reduced. Acetate is a common electron donor used by sulfate-reducers (Skyring 1988, Glombitza et al. 2015) as well as by many organotrophic nitrate-reducing bacteria (Kłodowska et al. 2018). MAGs from the storage pond with potential to use acetate and reduce nitrate were affiliated with A. aquaeolei, Bacteroidales S143-32, and Sulfurovaceae TCS-49 (Figs 2, 3, and 4). Arhodomonas aquaeolei is reported to be a halophile that was first isolated from a petroleum reservoir in 1993 and reportedly has potential for DNRA and denitrification (Adkins et al. 1993). Bacteroidales MAG 3 belongs to the novel family S143-32 that was first isolated from a hydrothermal vent (Reysenbach et al. 2020), but has not been extensively investigated. Sulfurovaceae TCS-49 has genes for acetate and nitrate metabolism, and members of this lineage are well known for using reduced sulfur compounds as electron donors. The presence of genes involved in sulfur compound oxidation in Sulfurovaceae MAG 19 indicates the potential for souring control to alternatively occur via direct removal of sulfide.

Indeed, sulfide generation (souring) can also be reversed by lithotrophic nitrate-reducers that utilize sulfide as an electron donor. Pathways for both sulfide oxidation to sulfite and nitrate reduction to nitrogen (denitrification) were found within A. aquaeolei (MAG 4) and H. alimentaria (MAG 9) (Figs 3 and 4). While the relative abundance of Halomonas started increasing at day 4 as nitrate levels in the storage pond were dropping (Figs 1 and 3), Halomonas sequence abundance in the pond continued to increase once nitrate was below detection limits. Halomonas alimentaria has been shown to be involved in hydrocarbon biodegradation (Olajire and Essien 2014) and has also been implicated in denitrification coupled to souring control (González-Domenech et al. 2010, Perman et al. 2022). These earlier observations could also be due to lithotrophic sulfide oxidation by members of this genus, based on the pattern observed here. Genes for both sulfide oxidation and DNRA were found within Bacteroidales S143-32 (MAG 3) and Sulfurospirillum (MAG 18). Relative abundances of both of these groups increased in the early stage of monitoring when nitrate was still detected in the pond, then slowly decreased as nitrogen compounds and organic carbon levels declined (Figs 1, 3, and 4).

Microorganisms need not be limited to only one metabolic mechanism for contributing toward souring control. Sulfurospirillum spp. have been shown to control souring in laboratory settings owing to the ability to couple DNRA to either organotrophic or lithotrophic metabolism (Hubert and Voordouw 2007, Ross et al. 2016, Hillman et al. 2022). While this potential was not uncovered in the Sulfurospirillum MAGs discovered in this study, three other MAGs possessed genes for both organotrophic and lithotrophic metabolism as well as genes for nitrate reduction. These belonged to A. aquaeolei (MAG 4), Sulfuvoraceae TCS-49 (MAG 19), and Bacteroidales S143-32 (MAG 3). Despite these wide arrays of metabolic genes, the increase in relative abundance within the initial monitoring stage for MAG 3 (S143-32) suggests its role in the pond was organoheterotrophic nitrate reduction based on the observed acetate consumption being coupled with nitrate reduction (Figs 1, 3, and 4), thereby preventing acetate-driven sulfate reduction (i.e. biocompetitive exclusion of SRM by NRM). The increase in relative abundance of ASVs affiliated with A. aquaeolei (MAG 4) during later stages of the monitoring period may indicate a role in autotrophic carbon fixation (Figs 3 and 4), given the decline in TIC measured at the same time (Fig. 1).

Monitoring the biogeochemistry, microbial diversity and genomic potential of a nitrate-treated storage pond revealed enrichment of nitrate-reducing populations and mitigation of sulfate reduction over a 4-week period. Putative nitrate-metabolizing organisms, Arhodomonas, Halomonas, and Sulfurospirillum, have been found in other produced water studies (Hubert et al. 2007, Mnif et al. 2009, Mbow et al. 2024). Potential for coupling nitrate reduction (through DNRA or denitrification) to either biocompetitive exclusion by organotrophic nitrate-reducers or sulfide oxidation by lithotrophic nitrate reducers, as well as inhibition of sulfate-reducing bacteria via nitrite production, is widely reflected in the genomic potential of the pond microbiome that emerged following nitrate addition. This study demonstrates nitrate treatment as a promising strategy for souring management in topsides oil field-produced water storage ponds. For industry operations, these findings highlight the importance of sufficient topsides nitrate addition to storage ponds in order to exhaust carbon sources that would otherwise be available for sulfidogenic bacteria to metabolize after nitrate is fully consumed in the storage pond. Additional nitrate dosing may thus be needed if produced water is stored in topsides pond environments for a prolonged period (Shi et al. 2023).

Acknowledgments

The authors would like to acknowledge that this work was done on the traditional territories of the Blackfoot Confederacy (Siksika, Kainai, Piikani), the Tsuut'ina, the Îyâxe Nakoda Nations, the Métis Nation (District 5 and 6), and all people who make their homes in the Treaty 7 region of Mohkínsstsisi (now named Calgary). The authors thank Brian Baillie for performing the measurements of organic and inorganic carbon, and Whitney England for bioinformatics support. This research was enabled in part by access and support provided by Prairies DRI the Digital Research Alliance of Canada through the Resource Allocation Grant to Dr Srijak Bhatnagar.

Author contributions

Gabrielle Scheffer (Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Methodology, Project Administration, Resources, Software, Supervision, Validation, Visualization, Writing—original draft), Jayne Rattray (Conceptualization, Methodology, Writing—review & editing), Paul Evans (Conceptualization, Funding Acquisition, Project Administration, Writing – review & editing), Wei Shi (Conceptualization, Data Curation, Methodology, Project Administration, Writing— review & editing), Srijak Bhatnagar (Resources, Software, Writing—review & editing), and Casey R. J. Hubert (Conceptualization, Formal Analysis, Funding acquisition, Project Administration, Resources, Supervision, Validation, Writing – original draft, Writing—review & editing)

Conflict of interest

None declared.

Funding

This work was funded by Chevron, by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) to C.R.J.H., and by scholarship funding to G.S. from NSERC (Alexander Graham Bell scholarship program) and the University of Calgary Eyes High program.

Data availability

All amplicon and metagenomic data are available under the National Genomics Data Center bioproject PRJCA024061. All codes and software used can be found at the following address: https://github.com/gabriellescheffer/Metagenomics_DigitalResearchAllianceCanada.

References