We report the discovery of an extinct deep-sea brine pool in the Red Sea's Hume Deep, providing evidence that rift-shoulder pools are ephemeral on millennial timescales, and that the microbially-enriched metalliferous deposits produced during the pool's active phase persist into the stratigraphic record. Now devoid of brine, the Hume Deep retains signatures of activity, including a well-preserved Lagerstätte death assemblage, seafloor staining consistent with past inhabitation by extremophile microbes, and enrichment of metals associated with their metabolisms. Elemental analyses of seafloor sediments demonstrate that the high metal concentrations in the Hume Deep match those of the active NEOM brine pool, situated 90 km north. In the NEOM pool, metagenomics and metatranscriptomics, coupled with scanning electron microscopy and energy-dispersive spectroscopy, suggest that microbes drive metal enrichment via microbe-mineral interactions up to 117× higher than background values. These results suggest a role for microbial processes in metalliferous sediment formation in deep-sea brine pools. Uranium-thorium dating suggests the Hume pool was last active between two and 16 thousand years ago. We propose that rift-shoulder brine pools fill and drain in response to tectonically controlled hydrothermal circulation. Our findings suggest that: (a) microbes inhabiting ancient Red Sea brine pools radically enhanced the metalliferous content of the stratigraphy within and above the basin's Miocene evaporites, and (b) that the elemental composition of other salt giant basins might have been similarly microbially mediated. With implications for understanding early life on Earth, our results provide a framework for interpreting deep-sea metal enrichments in modern and ancient extremophile environments.
Plain Language Summary
Metals like iron, potassium, manganese, copper, and molybdenum are known to accumulate in deep-sea brine pools, usually through interactions between sediments, seawater, and hydrothermal fluids. Our new research in the deep Red Sea reveals a surprising biological mechanism behind this metal enrichment. We found that microbes living in the oxygen-free brine help concentrate these elements in sediments—sometimes to levels more than one hundred times higher than those on the normal seafloor. This discovery matters for two reasons. First, many of these metals are essential for clean energy technologies, and understanding how they accumulate could help guide future resource exploration. Second, these brine pools offer a rare glimpse into how early life on Earth may have functioned. We show that sharp rises in manganese levels in ancient oceans may not have been caused by oxygen alone, but could also have come from microbial activity. Our findings give scientists a new tool for identifying early life and interpreting the chemical clues left behind in ancient rocks.
1 Introduction
Metalliferous marine sediments have long been recognized as geochemically distinctive deposits, but have received renewed attention because of their potential relevance to critical raw materials and the global energy transition (Bostrom & Peterson, 1966; Meylan et al., 2012). Such deposits are often found in association with deep-sea brine pools (Bäcker & Richter, 1973; Zierenberg, 1990) and can host valuable concentrations of transition elements—manganese, copper, iron, and molybdenum—plus other rare-earth elements (Boström & Peterson, 1969; Meylan et al., 2012). For brevity, we will refer to this portfolio of commercially-important elements (Mn, Cu, K, Fe, Mo) as “metals and trace elements.” As a consequence of the alternative energy revolution, including initiatives like Net Zero 2050, metals and trace elements are receiving sharp attention; global demand is rapidly outstripping supply (Jones, 2023; Vysetti, 2023).
Anoxic brine pools are more common in the deep sea than once thought, but their discovery is limited by sparse exploration and their isolated, patchwork distribution. Notable examples occur in the Gulf of Mexico, the Mediterranean, and the Red Sea, and many are thought to have formed through the dissolution of subsurface evaporites deposited during past episodes of basin desiccation (Camerlenghi, 1990; Craig, 1966; Delaunay et al., 2023; Joye et al., 2005; Pensa, 2024; Pensa, Baby, et al., 2025; Pensa, Huertas, & Afifi, 2025; Purkis et al., 2025).
The elemental composition of deep-sea sediments associated with brine pools can be set by a variety of mechanisms, ranging from hydrothermal processes to microbial metabolisms. One mechanism driving enriched metal concentration in seafloor sediments is precipitation from hydrothermal fluids (Backer & Schoell, 1972; Corliss, 1971; Corliss et al., 1978; Cronan, 1979; Rona, 1978; Spooner & Fyfe, 1973; Weiss, 1977). The Thetis Deep, located on the Red Sea's central spreading axis, provides one such example of a hydrothermal brine system (Pierret et al., 2010). In contrast, microbial processes offer another means of enriching metals through three predominant pathways—biomineralization, bioaccumulation/biosorption, and biologically mediated redox reactions (Ams et al., 2013; Dupraz et al., 2009; Gadd, 2010; Harris, 2017; Krumbein, 1983; Wang et al., 2013). The prokaryotes involved in these metal-concentrating mechanisms thrive in anoxic hypersaline environments, once thought inhospitable to life, but previous studies reveal that anoxic deep-sea brine pools actually harbor oases of halophilic microbes (Antunes et al., 2011; Danovaro et al., 2010; Minch et al., 2024; Purkis et al., 2022; Rothschild & Mancinelli, 2001).
Extremophile microbes—prokaryotes living in extreme physical and/or chemical conditions—which inhabit deep-sea brine pools can enrich sediments with metals to concentrations orders of magnitude higher than in the surrounding oxygenated seabed (Wang et al., 2013). Despite such elevated concentrations, these metalliferous sediments may be volumetrically inconsequential because modern brine pools are rare and small. For example, by our calculation, the pools so far discovered in the Red Sea occupy less than 0.1% of the area of the 440,000 km2 basin. But are modern conditions a good analog for the geological past? Perhaps not. We posit that brine pools were both large and abundant when, (a) their host “salt giant” basins initially formed, and then, (b) when those basins reflooded after desiccation events. We explore this theory through a conceptual model of the genesis of brine pools in the Red Sea rift (Figure 1) and contend that it broadly applies to brine pool formation in other salt basins throughout geological time.
2 The Origin of Brine Pools in the Red Sea
Red Sea rifting initiated in the Oligocene. Via a connection to the Mediterranean, the proto-Red Sea was first flooded by normal marine waters about 24.5 mya (Segev et al., 2017) (Figure 1a). Then, in the Mid Miocene, the basin was isolated from the Mediterranean, initiating a salinity crisis, which deposited evaporite sequences that locally exceed 2.5 km in thickness (Heaton et al., 1995; Petrovic et al., 2023). Evaporite-rich sequences deposited during this crisis include the Mansiyah Salt, deposited 14.0–13.2 mya (Hughes et al., 1999; Tubbs et al., 2014), overlain by the Ghawwas Formation, a siliciclastic unit interbedded with anhydrite which was deposited 13.2–6.2 mya (Baby et al., 2024; Hughes & Johnson, 2005; Pensa, 2024). Given that the Gulf of Aqaba only formed 14–11 mya (Bosworth et al., 2018; Stockli & Bosworth, 2018), it is likely that evaporites in this pull-apart basin can predominantly be ascribed to the Ghawwas Formation and more recent evaporitic episodes, with only minor contribution from the older Mansiyah Salt.
The precise origin of the Red Sea and Gulf of Aqaba evaporites is still an open question. Solar evaporation of seawater likely contributed to their formation, but Pensa (2024) estimated that >50 evaporation cycles would be required to account for the thickness of salt observed in the Red Sea, which can locally exceed 2.5 km. So many cycles seem unlikely, and would need to be coupled to an exceedingly rapid rate of basin subsidence (Heaton et al., 1995). Whereas the radiogenic strontium (Sr) value of the salt supports evaporation of Miocene seawater (Chakraborty et al., 2025), several thin horizons in the Mansiyah and Ghawwas display conspicuously low radiogenic Sr isotope ratios, suggesting the influence of fluids which interacted with the mantle (Pensa, 2024; Pensa, Baby, et al., 2025; Pensa, Huertas, & Afifi, 2025). Low radiogenic Sr isotope ratios support two additional mechanisms of salt production: via hydrothermal processes (Hovland et al., 2015) and via mantle serpentinization (Debure et al., 2019). Regardless of how salt was deposited, our conceptual model assumes that Red Sea brine pools, replete with extremophile microbes, episodically occupied the evaporitic basin during its 7.8 mya desiccation phase (Figure 1b).
By the Late Miocene (5 mya), a connection between the Red Sea and the Indian Ocean was established, reflooding the basin (Mitchell et al., 2021; Pensa, Baby, et al., 2025). At that time, we assume that the surface of the Ghawwas Formation was directly exposed to the inflowing seawater, again producing large quantities of brine, which are capable of hosting assemblages of metal-enriching microbial communities (Figure 1c). As the deposition of hemipelagic mud into the basin would have swiftly sealed the exposed evaporites from the overlying seawater, we envision this enrichment episode to have been brief but spatially extensive, because of the pervasive extent of Miocene evaporites, which likely exceeded 150,000 km2 in area five million years ago. Our model suggests that the resulting metal-enriched sedimentary horizon was later buried under the 200–300 m overburden of Quaternary sediment which now caps the Miocene stratigraphy (Baby et al., 2024; Mitchell et al., 2015; Pensa, 2024; Rowlands et al., 2014).
Today, albeit small and rare, brine pools remain a consistent feature of the modern Red Sea (Figure 1d) and form in one of two settings. First, brine pools occupy many of the “deeps” along the rift's spreading axis (Backer & Schoell, 1972; Laurila et al., 2015; Swallow & Crease, 1965; Zierenberg, 1990), though it should be noted that not all deeps boasting metalliferous sediments contained brine, as theorized by Pierret et al. (2010). The second setting for modern Red Sea brine pools occurs along the rift shoulders, where fault blocks enable modern seawater to percolate into the subsurface, dissolve underlying Miocene evaporites, and subsequently return to the seabed as brine. This brine accumulates in topographic depressions, forming brine pools. The NEOM brine pools which we recently discovered in the Gulf of Aqaba (Duchâtellier et al., 2024; Minch et al., 2024; Purkis et al., 2022, 2025) are one such example, as are the Thuwal Seeps (Batang et al., 2012) and the Afifi Pool (Duarte et al., 2020), located in the central and southern Red Sea, respectively.
Framed by our understanding of the geological history of the Red Sea, we reexamine the NEOM brine pools, located in the Aragonese Deep, and quantify the microbial enrichment of metals and trace elements to understand the ability of microbes to concentrate potentially vast metalliferous deposits earlier in the basin's history. Next, we examine a newly discovered extinct brine pool in the Hume Deep, located in the northern Red Sea, which retains evidence of microbial metalliferous enrichment, despite the brine having drained away. While we note that these brine pools have formed in tectonically distinct provinces—the Gulf of Aqaba is a pull-apart basin versus the Red Sea rift—we believe this comparison is valuable. The processes we describe can more broadly be associated with the salinity crises and subsequent reflooding of other major rift basins, such as the Michigan Basin in the Silurian (Cohee, 1965), the North Sea in the Permian (Brunstrom & Walmsley, 1969), the South Atlantic in the Jurassic (Nürnberg & Müller, 1991), as well as the Miocene salinity crises in the Mediterranean (Rouchy & Caruso, 2006), Persian/Arabian Gulf (Teller et al., 2000), and Caspian Basin (Van Baak et al., 2019). Our results also hold importance for the deep-time geological record. Mechanistic understanding of the formation of chemical biosignatures in well-constrained modern extreme environments provides an important process-based understanding that is critical to the accurate reconstruction of conditions that fostered the rise of early life on Earth (Anbar & Knoll, 2002; Fischer et al., 2016; Johnson et al., 2013; Lyons et al., 2014). Our work is timely because demand for metals required for the energy transition has renewed interest in marine mineral resources (Jones, 2023; Vysetti, 2023). Although current exploration efforts primarily target polymetallic nodules, cobalt-rich ferromanganese crusts, and seafloor massive sulfides, we contend that deep-sea metalliferous sediments will become increasingly important.
3 Materials and Methods
3.1 Sampling the NEOM and Hume Brine Pools With a Remotely Operated Vehicle
This study considers two brine pools. First, the active NEOM pool in the Gulf of Aqaba (Purkis et al., 2022), which we discovered in 2022 at a depth of 1,770 m in the Aragonese Deep pull-apart basin (Figure 2). Our second site, the Hume pool, is in the northern Red Sea rift, offshore the Midyan Peninsula off the southwest coastline of Tiran Island. The Hume pool is situated at a depth of 1,370 m, in the Hume Deep, another pull-apart basin. Like the NEOM pool, this site bears all the hallmarks of an active brine pool (Figure 3)—a seabed depression containing a rich Lagerstätte of calcareous metazoans, rimmed by a radically discolored “beach,” directly akin to the brightly colored microbial zones which surround the NEOM pool (Figure 3c). The only component lacking from this site is the brine itself, which we presume to have drained away prior to discovery. Thus, we consider our second “Hume Deep” site to be an extinct pool. We will build evidence for this interpretation via comparison between the elemental compositions of the sediment accumulating on the regular bathyal Red Sea seabed, versus those that accumulate in the active and extinct brine pools.
We sampled our five study sites; an active pool (two surface samples and three subsurface), a potentially extinct pool (five surface samples), and three regular seabed sites (five surface samples), during 2020, 2022, and 2023 research expeditions aboard the R/V OceanXplorer. Using a Mariner XL Argus remotely operated vehicle (ROV) equipped with two 7-function hydraulic manipulators and high-definition cameras, we collected surficial sediments from each location, which were placed into sealed vessels at the seabed, so as to prevent loss of fine material and contamination during the ascent of the vehicle. The ROV position was determined by integrating its acoustically tracked position relative to the dGPS-equipped mother ship with the vessel's GPS coordinates (measured to be cm-precision), providing continuous georeferenced navigation at the seabed, with <0.5 m precision. Each of the 12 surficial samples were collected from the sediment-water interface down to a depth of 10 cm beneath the seabed, and weighed at least 50 g. Sample site locations are reported in accompanying Data Set S1.
To supplement our surficial sediment samples from the active pool, we also extracted four additional samples from a core taken from the NEOM brine pool analyzed by Purkis et al. (2022, 2025), for which robust 14C age models have already been established (Figure S1 in Supporting Information S1). This record extends from the present day back to 1.6 kyr (±0.21 kyr), enabling us to audit the elemental composition of the stratigraphy beneath the NEOM pool. Prior to subsampling, we used X-ray fluorescence (XRF) scans and sediment facies observations to guide interval selection, targeting horizons with elevated Mn and Fe counts that were most likely to capture phases of enhanced metalliferous enrichment (Figure S1 in Supporting Information S1). The four downcore intervals which we sampled explicitly avoided turbidite layers, so as to ensure the geochemical record reflected in situ depositional conditions rather than episodic disturbance.
3.2 Geochemistry, Geochronology, and Genetics
We conducted a multi-step quantitative sequential leaching protocol to quantify the concentrations of metals and trace elements associated with different compositional fractions in our sediment samples, including sedimentary organic matter, carbonate minerals, and iron-manganese oxides (Figure S2 and Table S1 in Supporting Information S1). Analyses were conducted using triple quadrupole inductively coupled plasma mass spectrometry (Agilent 8900 ICP-QQQ) as previously described by Pollier et al. (2025), measuring manganese (Mn), iron (Fe), molybdenum (Mo), copper (Cu), and potassium (K). All concentrations were normalized to mass loss for each step in the sequential leaching protocol. Accuracy of the method was assessed through analysis of multiple certified reference materials and SPEX CertiPrep solutions, with average percent recovery always greater than 90% for all analytes of interest. Full methodological details and assessment of accuracy can be found in Supporting Information S1. Then, we used metagenomic and metatranscriptomic analyses to determine the composition, metabolic pathways, and activity levels of the extremophile bacterial and archaeal communities. This was achieved using total mRNA and total eDNA extraction, short-read whole genome and transcriptome sequencing, and metagenomic binning to recover bacterial metagenome assembled genomes (MAGs). Protein prediction of these MAGs was used to identify microbial metabolic pathways, and transcriptome reads were utilized to quantify the expression of these pathways relating to enriched metals. Scanning electron microscopy (SEM) was used to image the microbes, and energy-dispersive X-ray spectroscopy (EDS) to quantify the elemental enrichments on their surfaces. Lastly, to establish when the Hume Deep pool last contained brine, we determined the age of 14 echinoderm skeletons from the pool's Lagerstätte using uranium-thorium geochronology. All digestions were filtered through 0.22 μm PFA filters and spiked with a calibrated 229Th/233U/236U tracer for isotope dilution mass spectrometry. Spike calibration and external normalization were conducted using certified reference materials IRMM-35 (Th) and CRM-112a (U). A detrital 230Th/232Th activity ratio of 0.6 ± 0.2 was applied to correct for excess 230Th. The measured skeletons were collected along a transect across the seabed depression containing the Lagerstätte, with each sample separated by at least 25 m. Full methodological details for all these analyses is provided in our accompanying Supplemental Information.
4 Results
4.1 Depositional Setting of the Active NEOM and Extinct Hume Deep Brine Pools
The NEOM pool in the Gulf of Aqaba is active (presently brine filled) and situated at 1,770 m in the Aragonese Deep (Figure 2b). Our second site, the Hume Deep pool in the northern Red Sea, is extinct, now devoid of brine, and lies at a depth of 1,370 m offshore Tiran Island (Figure 2c). Both pools situate atop active transform faults associated with pull-apart basin mechanics. The Arona fault passes beneath the NEOM pool and the Hume fault beneath the Hume pool (Ribot et al., 2024).
The 10,000 m2 NEOM pool is surrounded by several small brine “ponds,” each only a few m2 in area. Our coring of this area suggests that the shallow subsurface of at least the eastern side of the Aragonese Deep is saturated with brine (Purkis et al., 2022), akin to a terrestrial water table. Configured as such, any seabed depression deeper than the “brine table” fills with brine, turning seabed depressions into brine pools—for example, Figures 3a and 3b. We hypothesize that the brine filling the Aragonese Deep (and subsequently, the NEOM pools) is sourced from the interaction of seawater in the subsurface with the underlying sediments and evaporitic formations along transform faults.
A belt of dense microbial communities surrounds each of the NEOM pools. This belt, varying in width from 1 to 3 m, neatly divides into three distinct zones based on the seabed coloration induced by the resident microbial assemblages—the “gray shoreward,” “orange,” and “gray brineward” zones (Purkis et al., 2022) (Figure 3c).
Despite the lack of brine, the Hume Deep site displays two visual hallmarks which suggest it to be an extinct pool. First, a seabed depression is ringed by a belt of discolored sediment (Figures 4a and 4b), mimicking the microbial zones surrounding the active NEOM pools (Figure 3c). In addition, both depressions are filled with an abundant death assemblage of calcareous metazoans—a Lagerstätte—explored in the following section.
4.2 Brine Pools as Lagerstätte Traps
Many brine pools act as Lagerstätte traps (Bäcker & Richter, 1973; Purkis et al., 2022), though, in rare cases, taphonomic processes can degrade shells and skeletons over time (Parsons-Hubbard et al., 2011). When exposed to anoxic brine, metazoans (animals) swiftly asphyxiate, die, and their shells and skeletons accumulate beneath the brine. The bed of the NEOM pool is veneered by a “hash” of fish bones and scales, plus pteropods, foraminifera, gastropods, sponge spicules, brachiopods, serpulid worms, and piles of intact echinoid skeletons. The brine also effectively preserves organic matter; abundant seagrass blades float on the surface of the NEOM pool (Figure 3d), despite it being located nearly 2 km beneath the photic zone, similar to the observations of Stuhr et al. (2025).
The now-brineless depression in the Hume Deep is filled with the same death assemblage as the NEOM pool (Figure 4). Echinoid skeletons have accumulated there in a similar composition, along with the same characteristic calcareous hash (Figure 5). Skeletal grains are strikingly more abundant at the Hume and NEOM sites compared to the regular bathyal seabed of the Gulf of Aqaba. In surficial samples collected by ROV, the ratio of skeletal to non-skeletal grains is approximately equal (1:1.1 for Hume and 1:1.0 for NEOM). In stark contrast, regular bathyal seabed samples are dominated by siliciclastic grains, with approximately 99 non-skeletal grains for every skeletal one (Figure 5). The two pools evidently serve as highly effective passive-entrapment Lagerstätte traps. The active NEOM brine pool and extinct Hume Deep pool are identical in this regard.
4.3 Hume Brine Pool Activity Spanned at Least 14 kyr
We presume that the abundant echinoid death assemblage present in the Hume depression (Figure 5) represents individuals that wandered into the pool during its active phase(s), asphyxiated and died in the anoxic conditions, and became incorporated in the Lagerstätte. This assumption seems reasonable, given that dead echinoids are exceedingly rare on the regular seabed which we surveyed. There, we only encountered three such skeletons in over 75 km of ROV transects recorded across the bathyal seabed of the northern Red Sea and Gulf of Aqaba. By contrast, many tens of echinoid skeletons can be found in each square meter of the brine pool Lagerstätte. To quantify this disparity in abundance, we weighed the proportion of echinoid skeletal fragments in 10-g seabed samples collected from the beds of the NEOM and Hume pools, and also from the regular seabed. The mass of echinoid skeletal fragments per 10-g sample at the NEOM site is 3.91 g ± 1.21 g over five replicates. For the Hume site, that value is 5.22 g ± 2.87 g. Meanwhile, the mass of echinoids in the regular seabed away from these two depressions, but in the same depth range, is only 0.01 g ± <0.01 g.
The ages of 14 echinoid skeletons collected from the extinct Hume brine pool ranged between 1.99 and 16.17 kyr BP (Figure 5d, Data Set S1). At the 95% confidence interval, the minimum uncertainty of these ages is ±0.25 kyr, and the maximum is ±1.52 kyr. Assuming the echinoids were killed by the anoxic conditions, these ages suggest that the Hume Deep hosted brine for at least part of the 1.99–16.17 kyr BP interval. Although these ages cannot determine whether the depression remained continuously brine-filled throughout this period, the concentration of skeletal remains within a single topographic depression may be more consistent with prolonged occupancy than with repeated episodes of drainage and refilling. The data are insufficient to distinguish conclusively between these scenarios, and open-system behavior of echinoid calcite further limits the precision of our dating, even after detrital 230Th correction (see accompanying Data Set S1). Nonetheless, it is clear that the longevity of the active and inactive phases of this pool is measured in millennia, as opposed to centuries.
4.4 Microbe-Mineral Interactions in and Around Brine Pools
A conspicuously stained band of sediment rims both the active NEOM and extinct Hume brine pools (Figures 3 and 4). For the case of the NEOM pool, extremophile brine-associated microbes occupy this zone (Minch et al., 2024; Purkis et al., 2022). Despite the absence of brine and living microbial mats in the extinct Hume pool, the seabed retains its discoloration, indicating that the residue, potentially linked to microbial activity, pigments the seabed. Sequential leaching of sediments revealed that the NEOM and Hume brine pools possess distinct chemical signatures, differentiating them from the surrounding seabed. The concentrations of five elements (Fe, K, Mn, Cu, Mo), all of which have been recognized as being associated with extremophile microbial activity (Dong et al., 2022), displayed substantial enrichments in the organic matter fraction of the sediments associated with the NEOM and Hume Deep brine pools. In the surface sediments, total iron (Fe) concentrations were enriched in brine pool-associated organic matter by a factor that ranged between 14 and 41×, potassium (K) by factors between 1–16×, manganese (Mn) by factors between 60 and 117×, copper (Cu) by 7–48×, and molybdenum (Mo) by 25–41×, when compared to organic matter associated with sediments from the regular seabed (Figure 6b, note log y-axis, Table S2 in Supporting Information S1).
The active microbial communities inhabiting the NEOM pool afforded a unique opportunity to investigate associations between total Fe, K, Mn, Mo, and Cu enrichments in the organic matter and microbial activity. Scanning electron microscopy (SEM) revealed abundant filamentous structures embedded within a dense matrix that is consistent with extracellular polymeric substances (EPS; Figure 7a), although we cannot exclude the possibility that some of these features represent authigenic ferromanganese precipitates. Energy dispersive electron spectroscopy (EDS) analyses confirmed that Fe, K, and Mn were highly enriched on the filamentous structures (Figure 7b), whereas Mo and Cu—though identified by sequential leaching and ICP-QQQ analyses to occur at concentrations of 0.008 wt % and 0.038 wt %, respectively (Table S2 in Supporting Information S1)—remained below EDS detection thresholds (∼0.1 wt %). ICP-MS analyses supported the results of SEM-EDS analyses, and showed that concentrations of Fe, K, and Mn reached 11.3 wt %, 6.4 wt %, and 1.5 wt %, respectively, in the organic matter associated with the active NEOM pool, well within EDS detection capability (Table S2 in Supporting Information S1; Figure 6). Non-microbial substrates returned zero counts for all five elements. Our EDS analyses demonstrate that elevated concentrations of Fe, K, and Mn are spatially associated with the filamentous structures and their surrounding matrix, even within extinct brine pools. Although the precision of the EDS measurements precludes a definitive determination for Mo and Cu, our combined approach of sequential leaching and ICP-QQQ analysis confirms these elemental enrichments occur in the sedimentary organic matter, and metagenomic and metatranscriptomic data clearly advocate that these elements are likewise microbially linked.
