Lithoautotrophy in the subsurface


Tel.: +1 (509) 373-0891; Fax: +1 (509) 376-9650; e-mail:


If microorganisms can carry out primary production within the Earth's crust, then the biosphere might not be totally dependent on surface-based photosynthesis. Potential chemical energy from purely geochemical sources within the earth can support growth of a number of known microorganisms, chiefly strict anaerobes, such as methanogens, homoacetogens, and sulfate-reducers. (Chemo)lithoautotrophic microorganisms have been detected in sedimentary systems, but they have not been shown to carry out primary production in situ, at least not without some dependence on surface-based photosynthesis. Microbial communities within igneous rock formations might, of necessity, be based on in situ primary production. Evidence has emerged for the presence of microorganism in basalt below the sea floor, but data on in situ activity are not yet in hand. Microbial communities have been observed, within continental flood basalts and granitic plutons, which appear to be based on in situ primary production by anaerobic bacteria. Geochemical measurements have confirmed that in situ activity is lithoautotrophic. This evidence for subsurface lithoautotrophic microbial ecosystems, which are not dependent on surface organisms, may have profound implications for life on the early Earth, and on other planets, including Mars.


The concept of bacteria living deep underground now seems well-accepted in the scientific world. However, investigators are only beginning to understand the functional roles of these organisms in the Earth's biosphere. One question of current interest is whether subsurface microorganisms carry out primary production underground, or whether all living things are ultimately dependent on primary production occurring at the Earth's surface. This article reviews recent findings which bear direct and indirect evidence for this question, and concludes by discussing some of the wider implications of subsurface primary production.

2Types of subsurface ecosystems

Conceptually, at least, subsurface ecosystems might be divided into ‘detrital systems’ and ‘productive systems’. Most of what we know about subsurface microbiology, so far, has been learned in detrital systems, although there is growing evidence for several productive systems. In practice, the dividing line between the two types may be somewhat blurred, but properties of the two extremes (or ‘end-members’ in geological parlance) can be described, at least for purposes of this discussion.

Detrital systems conform to the standard paradigm of ecology, in which primary production is accomplished by photosynthetic organisms, at the Earth's surface. All other organisms act as consumers of the photosynthate, secondary consumers, or recyclers of photosynthetic energy. This paradigm is extended into the subsurface by burial of photosynthetically produced organic matter by sedimentation and other processes of the rock cycle. In a detrital subsurface ecosystem, microbial metabolism is primarily heterotrophic, and is based upon consuming buried organic matter. Such a system would seem somewhat limited, since, with increasing time, the organic matter would be depleted, and the remainder would become more recalcitrant and less available, due to compaction, cementation and decreased porosity.

The alternative to a detrital system is a subsurface ecosystem in which primary production occurs, using in situ energy sources of geochemical origin. Rather than being based on photosynthesis, subsurface primary production must be based on chemolithoautotrophy. That is, both the energy source and the electron sink must be inorganic chemicals, and inorganic carbon is converted to organic carbon. Chemolithoautotrophy is a well-known physiological trait of diverse microorganisms, as shown in Table 1. In surface ecosystems, autotrophic organisms are assumed to function primarily as recyclers, conserving energy that would otherwise be lost during decomposition of organic matter. For instance, an autotrophic methanogen consumes hydrogen and carbon dioxide, which were produced from fermentation of organic acids, yielding methane, and conserving the potential (photosynthetic) energy that would otherwise be lost. In a productive subsurface system, the same organism could function as a primary producer, if its substrates had an abiotic geochemical source. This has the potential to lead to ecosystems in which life is based on geochemically derived or ‘terrestrial energy’[1], rather than solar energy. This potential for subsurface primary production has been noted by several reviewers (e.g., [1–4]) but only recently has evidence become available to show that this phenomenon actually occurs.

Table 1.  Some redox couples known to support chemolithoautotrophy by microorganisms
Microbial processElectron donorElectron acceptorPhotosynthesis dependent?Deep subsurface geochemical source?
Sulfide oxidationS2−O2X 
Anaerobic sulfide oxidationS2NO3  
Sulfur oxidationSoO2X 
Thiosulfate oxidationS2O2−3O2X 
Iron oxidationFe2+O2X 
Anaerobic iron oxidationFe2+NO3  
Manganese oxidationMn2+O2X 
Methane oxidationCH4O2X 
Anaerobic methane oxidationCH4SO2−4 X
Hydrogen oxidationH2O2X 
Sulfur reductionH2So X
Sulfate reductionH2SO2−4 X
MethanogenesisH2CO2 X
AcetogenesisH2CO2 X

As noted above, if (chemo)lithoautotrophy occurs in the subsurface (we can dismiss (photo)lithoautotrophy) both an electron donor and an electron acceptor must be supplied from geochemical sources. Table 1 lists a number of redox couples which are known to support lithoautotrophy. Those processes which require the most oxidized electron acceptors, such as oxygen and nitrate, might occur in the subsurface, but would be limited to near-surface or young groundwater environments. This is because oxygen is a product of photosynthesis, and must be transported from the surface. The remaining lithoautotrophic processes are more likely to take advantage of in situ subsurface geochemical resources, and they conspicuously feature hydrogen as an electron donor.

Hydrogen is widespread in the deep subsurface, and has a number of geochemical sources (e.g., [5]). The most relevant to this discussion are the reactions of igneous rocks, which are the main components of the Earth. These reactions form a continuum from fast high-temperature reactions of magma outgassing and magma-water interactions, to slow, low-temperature reactions that take place during weathering of igneous rocks. Igneous rocks are formed by extrusion of melted material from the Earth's mantle into the crust. Rocks with greater content of silica are classified as felsic, while those with greater content of magnesium and ferrous iron are classified as mafic. Those with the highest magnesium and iron content are called ultramafic (e.g., [6]). The greater the content of reduced metals, the more potential redox energy a rock should contain. The slow oxidation of ultramafic rocks by groundwaters is known as serpentinization (e.g., [7]).


Although serpentinization is often assumed to require elevated temperatures (several hundred degrees) and geological time scales, we have recently demonstrated that microbially significant quantities of hydrogen can be formed during room temperature reactions of igneous rocks over a few hours or days [8]. Even some rather felsic rocks produced hydrogen, if strictly anaerobic conditions were maintained.

Of course, while productive subsurface ecosystems would depend upon primary production by lithoautotrophs, we know of no reason why heterotrophic organisms would not be present as well. Conceivably, complex subsurface ecosystems could be supported by in situ primary production. However, since rates of metabolism are extremely low in the subsurface [9, 10], metabolism will be primarily directed towards maintenance, rather than growth (e.g., [11]) and geochemical signatures, such as isotope ratios, might reflect only the primary energy reaction.

3Where to look?

The vast majority of subsurface microbiology studies have been carried out on samples obtained from sedimentary formations, and there are good reasons for this. Most subsurface sediments contained significant populations of microorganisms at the time of deposition. Sediments often contain large concentrations of organic matter, so they can support microbial metabolism. The extensive pore space in many sedimentary rocks provides ample habitat volume for subsurface microorganisms. Because sediments are typically porous media, slow fluxes of groundwater are able to distribute nutrients through the formations. These factors make sedimentary formations likely habitats for subsurface microorganisms, and indeed, such organisms are often found to thrive in them. Because of these same characteristics, however, one might hypothesize that any active microbial ecosystem in a sedimentary environment is likely to be a detrital ecosystem.

Igneous rocks, at first glance, might seem unlikely to provide a habitat for subsurface microorganisms. Because they are formed from molten magma, one would expect that they contain little or no organic matter. (However, there have been sometimes controversial reports of small amounts of kerogenous organic matter within igneous rocks [12, 13].) Because they are crystalline rocks, and usually not porous, they provide little habitat volume. Any microorganisms must be confined to fractures, cooling joints, and rubble zones (which form between successive lava flows), as must groundwater and associated nutrient fluxes. Nevertheless, although the habitat density may be lower than sedimentary rocks, when the sheer volume of igneous rocks in the Earth's crust is considered, these fracture networks can form a vast total volume. As discussed above, reactions between water and igneous rock in fracture networks can result, under the correct conditions, in production of energy gases, which might be used by lithoautotrophic microorganisms. Because of these characteristics, one might hypothesize that any active microbial ecosystem in an igneous rock habitat is likely to be based on in situ primary production.

4The evidence

4.1Field evidence: lithoautotrophy in sedimentary systems

We have hypothesized that active microbial ecosystems in sedimentary formations will be detrital. Yet there have been reports of lithoautotrophy in sedimentary systems. Is the hypothesis incorrect?

4.1.1Karstic environments

A number of investigators have reported evidence for lithoautotrophy in caves located in karstic terrains. Karst is a chemical sediment, composed mainly of limestone, which is readily dissolved by acids in groundwater. This results in extensive caverns and underground streams (e.g., [6, 14]). These caves and streams conduct oxygen, from the atmosphere, into the subsurface. In several places, anaerobic water, containing reduced gases has been found to enter cave systems from some deeper source. This results in a chemical environment that is analogous to that at geothermal hot springs. In several caves, sulfide- and sulfur-oxidizing bacteria have been found to form mats, in which lithoautotrophic carbon fixation occurs [15–17]. In at least one case, invertebrate fauna have been found to graze on these mats, and the isotopic composition of their biomass showed that the animals depend entirely on bacterial lithoautotrophic carbon fixation [18]. However, the source of the reduced gases, which provide energy for these systems, has not been determined. To understand whether this is true subsurface primary production, perhaps the scale of the system should be redefined to include this unknown energy source. If the reduced sulfide is a product of microbial sulfate-reduction, coupled to fermentation of organic matter in a deeper aquifer, then the lithoautotrophic mats in the karst are recycling and conserving photosynthetic energy, as their relatives do in surface sediments. If the source of reduced gases is volcanic activity, then the system would be analogous to deep-sea hydrothermal vents. However, since O2, produced at the surface by photosynthetic organisms, is required as an electron acceptor, these karst systems (as well as the deep-sea vent organisms) are ultimately dependent on photosynthetic processes, and do not constitute an independent subsurface ecosystem.

4.1.2Oil bearing sediments

Lithoautotrophic microorganisms have been detected in several very deep sedimentary formations. Thermophilic hydrogen-oxidizing bacteria have been detected in fluids produced from formations over 3 km deep [19–22]. These include hyperthermophilic sulfidogenic archaea, methanogenic archaea, and thermophilic sulfidogenic and metal-reducing bacteria. The possibility that these organisms are an artifact of well drilling or sea-water injection (which is used to enhance oil recovery) cannot be completely ruled out [23, 24]. However, some lithoautotrophs have been recovered from wells which were not affected by water injection. The physiology of microorganisms isolated from these environments is often consistent with the conditions at depth, in terms of optimal salinity and temperature, which is elevated due to geothermal gradients.

Core samples may be less likely to contain contaminating organisms than borehole fluids, and chemical tracers can be used to detect contamination. Lithoautotrophs were also isolated from core samples taken from 2800 m deep shale-sandstone formations, and which tracer methods indicated were uncontaminated. In these samples, lithoautotrophs were found to outnumber heterotrophs, at least as counted by laboratory growth-based assays [25]. These organisms included hydrogen-oxidizing metal-reducing and sulfidogenic bacteria [26].

Unfortunately, no in situ measurements of metabolic activity are available for deep sedimentary formations, so the ecological role of the recovered microorganisms, which are often facultatively heterotrophic, can only be inferred. It may be that the apparent high incidence of lithoautotrophs in deep sediments is an artifact of laboratory methods used to recover microorganisms. Alternatively, it may be that hydrogen is an important electron donor in these formations. Produced from organic matter by fermentation or thermal cracking, hydrogen would be transported through the low-porosity environments of the deep subsurface more readily than many other nutrients. Because these hydrocarbon-bearing sediments are rich in fossil organic matter, we can assume that they are ultimately detrital ecosystems.

In summary, it is evident that lithoautotrophy occurs in sedimentary subsurface habitats. However, as mentioned above, it can be seen already that there is not always a clear-cut distinction between detrital and productive systems. The reader will have noticed that this discussion belabors the point of independence from surface photosynthesis. What difference does this make? Is it really a useful distinction? We will return to this point in the final section below.

4.2Field evidence: lithoautotrophy in igneous rock ecosystems

Relatively few investigations have studied microorganisms in subsurface igneous rock formations. According to the hypothesis above, however, active microbial ecosystems in igneous rock might be expected to be based on in situ primary production.

4.2.1Sub-seafloor ultramafic rocks

The most reactive igneous rocks are ophiolite suites, which are ultramafic rocks produced from magma at mid-ocean spreading centers. This is also the location of the well-known chemolithoautotrophic (but oxygen-dependent) communities on the sea-floor. There has been considerable speculation that a ‘deep hot biosphere’ might exist within the basalt, in the hydrothermal convection cell of the hot springs (e.g., [27]). Such a hypothetical system would, of necessity, be based on anaerobic chemolithoautotrophy. This would require life at temperatures above 200°C, which most investigators currently believe to be unlikely [1]. Thermophilic anaerobic lithoautotrophs have been detected on the walls of ‘black smoker’ chimneys, but not in the very hot fluids emitted from the vents. There is currently no geochemical or isotopic evidence for active lithoautotrophy below the hot springs. Lilley et al. [28]reported anomalous isotope compositions for methane emitted from an unsedimented hydrothermal system which were consistent with microbial lithoautotrophy, but attributed the gases to pyrolysis of organic matter buried below the basalt.

Though conditions below mid-ocean spreading centers may lie outside the feasible range for living organisms, the rocks produced there are steadily transported outward, and temperatures decrease concomitantly. Zones of active serpentinization have been observed in cooler ophiolites, and it seems feasible that this could support microbial communities below the sea floor. Plumes of dissolved methane have been detected in the ocean above these serpentinization zones [29, 30], as might be expected from a bacterial system. The origin of this methane has been explained by abiotic geochemical models [31]though it may be useful to reexamine these sites as our understanding of microbial interactions with igneous rocks improves.

Unfortunately, while it is difficult to obtain samples from the subsurface that are useful for microbiology, it is even more difficult to do so at the bottom of the ocean! Nevertheless, several researchers have recently reported encouraging results from studies of basalt cores obtained by the Ocean Drilling Program [32]. In these cores, basaltic glass was altered along natural fractures, through which sea water entered and reacted with the rocks, over long periods of time. At the interface between unaltered glass and the alteration materials, where contaminants could not have penetrated during sampling, evidence was found which suggested in situ bacterial activity. The evidence included bacterium-shaped etched channels in the glass, putative bacteria which were stained by DNA-specific fluorescent dyes [32], detection of DNA [33], and enrichment of potassium in the altered glass, which may be an indicator of biomass [34]. No evidence for lithoautotrophy in these samples has been obtained so far, but one might predict aerobic iron-oxidation near the sea floor, and anaerobic hydrogen-oxidation in deeper strata.

In summary, no firm evidence is yet available for lithoautotrophic microbial ecosystems below the sea floor. However, the few observations that are currently available appear to provide encouraging circumstantial support. Ophiolite suites with associated serpentinization zones are also found on the continents (e.g., [35]), though they have not been examined for the presence of microorganisms, it may prove fruitful to do so.

4.2.2Continental basalt formations

Mafic rocks contain less reduced iron than ultramafic rocks, and are correspondingly less reactive. Nevertheless, they can react with water, under appropriate conditions, to produce microbially available hydrogen. We have recently reported evidence for a subsurface lithoautotrophic microbial ecosystem (SLME) within the Columbia River Basalt group (CRB) [8]. The CRB is a series of continental flood basalts which form a layered structure up to 3 km deep and covering over 300 000 km2 in western North America. Confined aquifers between the basalt flows contain anaerobic, reducing water, indicating extensive water-rock interactions, as well as abundant dissolved hydrogen, methane, sulfide and bacterial cells. Dissolved inorganic carbon (DIC) is progressively depleted with increasing methane concentration, suggesting that lithoautotrophic methanogens consume bicarbonate to produce methane. The stable isotope composition of the DIC is consistent with this model. These observations suggest that microbial lithoautotrophic metabolism occurs in situ in the CRB. If the system were heterotrophy-based, DIC would increase, rather than decrease with increasing methane.

The ferrous silicate minerals in basalt react with water, under appropriate conditions, to generate hydrogen gas [36]. This reaction is inhibited by molecular oxygen and promoted by low pH, high temperature, and greater reacting surface area. Since meteoritic water is initially aerobic, if H2 production is to occur, groundwaters must be isolated from communication with the atmosphere, and remain in contact with subsurface rocks long enough for all the O2 to be removed. These properties should allow some general predictions about the sorts of aquifers where SLMEs might be found. The rock matrix should contain significant quantities of ferrous-silicate minerals. Depth below surface should be adequate to preclude diffusion of oxygen from the atmosphere. Groundwater flow rates should be low, or flow paths long, to ensure sufficient water rock interaction. The two currently proposed SLMEs are in deep confined aquifers with groundwater ages of several tens of thousands of years. Future research might allow more quantitative description of SLME habitat. For instance, volcanic ash deposits might form a high surface-area aquifer with relatively fast reaction rates, compared to bulk igneous rock. However, no subsurface microbiology investigations have yet examined such formations.

A number of important questions remain unanswered in these systems. What happens when the fresh rock surfaces become oxidized? What are the exact mineral reactions that occur, and what are the limiting factors? Do microorganisms actively mine basalt, or passively consume products of abiotic reactions? What sorts of competitive or syntrophic interactions occur between the different types of microorganisms observed?

It should be noted that, while primary production occurs in sulfidogenic basalt microcosms (Landau, N. and Stevens, T., unpublished), the stable isotope signatures measured in sulfidogenic CRB groundwaters do not conform to the lithoautotrophic pattern, though no known source of organic carbon is present. These aquifers warrant further study, especially since they may be analogous to sub-seafloor basalts, which would also contain elevated sulfate, due to the presence of seawater.

Microbial communities from the CRB can be grown in laboratory microcosms containing basalt as the sole electron donor, as can some well-characterized pure cultures of bacteria [36]. These microcosms provide an opportunity to carry out manipulative experiments to test hypotheses about SLMEs.

4.2.3Continental granite formations

Granites are felsic igneous rocks with rather low iron content, yet they can contain enough ferrous silicate to react with water to form hydrogen, though less than mafic or ultramafic rocks [8]. It seems possible that SLMEs could exist in large granitic plutons. In fact, natural gas deposits with anomalous isotopic and hydrocarbon composition, apparently of microbial origin, have been observed within the Canadian and Fenno-Scandian shields [37, 38]. Microorganisms within granite formations in Sweden have been studied extensively [39, 40]. Recent results from this formation reveal that anaerobic lithoautotrophs are abundant in deep granitic groundwaters, and geochemical measurements indicate that these organisms are active in situ [56]. Accumulating evidence indicates that a lithoautotrophy based ecosystem may exist within this deep granite groundwater system.

To summarize, accumulated evidence seems to indicate that SLMEs have the potential to exist in aquifers within most kinds of large igneous rock bodies. Indeed, the strongest field evidence, so far, comes from formations composed of some of the less reactive igneous rocks. This suggests that prospects are good for, and studies are certainly warranted of, similar systems in ultramafic subsurface environments.

5Consequences and implications of SLMEs

Are SLMEs merely a scientific novelty, or do they have broader significance that makes them worthy of attention? And what difference could it make, if SLMEs could truly function independently of the surface biosphere? The following discussions are highly speculative, but illustrate why it may be important to study these subsurface phenomena, and perhaps outline some research questions for the future.

5.1Economic considerations

Certainly, microorganisms may have impacts on economically important activities that take place in their habitat. This has motivated most SLME investigations, to date.

Current evidence suggests that SLMEs may contribute to natural gas formation, in certain settings. If this should be confirmed, it could lead to a better understanding of this important resource, and possibly of certain other mineral resources.

Several countries have planned high-level nuclear waste repositories in geological formations which may contain SLMEs. Microbial activity in these settings may have important long-term controls on corrosion of containment vessels, and migration of radionuclides.

A variety of groundwater pollutants are present in igneous rock formations around the world. It is possible that active SLMEs could degrade or transform these contaminants, or affect their mobility. For instance, many of the anaerobic microorganisms known to occur in SLMEs can carry out reductive dehalogenation of solvents, or reductive transformation of heavy metals. It remains to be seen whether the in situ rates of metabolism in a SLME are great enough to interdict contaminant plumes. If so, SLMEs could form a stable long-term barrier to waste migration.

5.2Evolution of the biosphere

If SLMEs can truly function without input from the surface, it may mean that the depth of the biosphere is limited only by increasing geothermal temperature. This could have several important implications for the history of early life on Earth [41]. SLMEs could be a model for how ecosystems functioned before the evolution of photosynthesis. It is interesting that the energy currency of reduced iron and sulfur compounds that appear to support SLMEs is reminiscent of the abiotic chemistry invoked in the origin-of-life theories that are collectively known as ‘iron-sulfur hypotheses’ (e.g., [42]).

The ability to survive in the subsurface, without input from the surface, could also confer on organisms the ability to survive cosmological events which sterilize the surface of the planet. Many such events are believed to have occurred during the early history of the Earth [43, 44], yet the geological record suggests that relatively advanced microorganisms may have been present soon afterwards [45–47]. Interestingly, the oldest known biogenic signatures, in 3.8 Ga carbonates, seem to indicate lithoautotrophic methanogenesis [47]. Perhaps the earliest continuously habitable biosphere was in the deep subsurface, sandwiched between geological heat from below, and impact heat from above [41].

5.3Possible life on Mars

The surface of Mars is uninhabitable by, and probably lethal to, any known organism (e.g., [48]). The temperature and pressure are too low for water to exist as liquid at the surface, though water is abundant on the planet. Yet geological outwash features, observed from orbit, suggest that large quantities of water, possibly as liquid, exist in the subsurface [49]. Mechanisms have been suggested which could allow a hydrologic cycle to exist in the subsurface (e.g., [50]). Thus, it seems possible that there could be conditions in the deep subsurface of Mars where chemolithoautotrophic microorganisms could exist [3]. SLMEs in igneous rock formations on Earth are the only known examples of functioning ecosystems that could survive on Mars today, if they existed there. Water, basalt, and inorganic carbon should be abundant, in the Martian subsurface, though the depths at which pressures and temperatures reach permissible levels for life may be more than 2 km [51].

Earlier in the history of the solar system, around the time of the origin of life, conditions on Earth and Mars were more similar. If life arose on one planet, it is not unreasonable to suspect that it arose on the other as well. As the surface of Mars became uninhabitable, any life that was there must have perished, or retreated into the subsurface [52]. If life exists there today, it is most likely in the deep subsurface.

Several recently recognized meteorites of Martian origin are known on the Earth, and have been the subject of much recent study. These are subsurface igneous rocks ejected from Mars by cosmic impacts. Some investigators have even suggested that one may contain evidence of ancient microbial life [53]. Whether or not these hypotheses are eventually proved, Earthly SLMEs are an exact analog of the proposed biological system, and are the positive control by which to evaluate putative Martian life.

Though the connection remains speculative, an understanding of SLMEs and the geochemical signatures they leave behind in rocks may be highly significant to Martian exploration. During the upcoming decade, several spacecraft are planned to land and conduct investigations of Mars. Though planned experiments do not focus on life detection per se, a number of geological and chemical observations could shed light on the possibility that life ever existed there. A better understanding of microorganism-rock interactions could provide possible targets for some of these remote investigations.

5.4Life elsewhere in the solar system?

Are there possible abodes for life beyond Mars? Liquid water is assumed to be the primary requirement for life [52], and probably does not exist on Mercury, Venus, or the gas giant planets. However, though we know relatively little about them, there are several other planets where liquid water, and hence life, may be possible.

The Galilean moons of Jupiter – Io, Europa, Callisto, and Ganymede – appear to be tectonically active and may have enough interior heat to posses liquid water. The surfaces of the latter three planets are composed largely of water ice. Europa, in particular, is hypothesized to contain a subsurface ocean of liquid water [54]. These planets have rocky cores that could supply reduced minerals to act as electron donors for microorganisms. The Galileo spacecraft is currently exploring the Jovian system, and accumulating new information about these worlds [55]. Much more must be learned, but it remains feasible that SLME type life could survive within one or more of these planets.

Titan, the largest moon of Saturn, is often mentioned as a possible location for extraterrestrial life, or at least interesting pre-biotic chemistry [54]. Little is known of this planet, which is hidden beneath an atmosphere containing methane and ammonia. Upcoming exploration of the Saturn system by the Cassini spacecraft will include ‘Huygens’, an autonomous probe that will be dispatched to Titan to learn more about this remote, enigmatic planet.

While it is far too early to reach conclusions about any of the above hypotheses, it is reasonable to begin formulating methods for testing them. The emerging understanding of the ecology of the terrestrial subsurface may provide direction in the search for life in distant space and time.


SLMEs have frequently been invoked in the literature, but only recently has evidence been found that they exist in nature. The potential for SLMEs in the Earth's crust is widespread, since presumably suitable igneous bodies are ubiquitous. Field evidence for SLMEs is restricted to only a few locations, however, mostly because of the difficulty and expense of observing the deep subsurface. Continuing and future studies, including field measurements and study of laboratory microcosms, should contribute to the emerging understanding of the ecological functions and biogeochemical activities of SLMEs. Such new information could lead to techniques for more simply detecting the past or present activity of SLMEs in the Earth's crust, or elsewhere in the solar system.


I thank many colleagues for stimulating discussions on this topic, including P. Long, C. McKay, J. McKinley and K. Zahnle. Preparation of this review was supported by the Subsurface Science Program, Office of Health and Environmental Research, U.S. Department of Energy (DOE). Pacific Northwest Laboratory is operated for DOE by Battelle Memorial Institute under Contract DE-AC06-76RLO 1830.