April 2017 Issue Table of Contents
Sedimentary sulfides constitute over 95% of the sulfide on the surface of the planet, and their formation, preservation and destruction largely determines the surface environment. The sulfide in sediments is mainly derived from the products of sulfate-reducing bacteria, which are currently responsible for oxidizing over half the organic matter flux reaching sediments. Pyrite is the mineral overwhelmingly produced. The geochemistry of pyrite, both in terms of its isotopic composition and its trace-element loading, has varied dramatically over geologic time. As such, it is a major source of our current understanding about the nature of the early Earth and of the Earth’s subsequent geochemical and biological evolution.
Keywords: sulfides, sediment, pyrite, microorganisms, evolution
Research into sedimentary sulfides constitutes one of the most exciting areas of contemporary scientific research. This is evident from the thousands of publications describing and reviewing advances in the subject: these papers contribute to our understanding of past, present and future natural worlds. For detailed references, readers are referred to the 3,928 citations listed in recent comprehensive reviews of various aspects of the subject by Rickard (2012) and Rabus et al. (2015).
Sulfides are ubiquitous in modern sediments. Most geologists carry an acid bottle in their rucksacks to test whether rocks contain carbonates; however, if they were to add a drop of acid to most of the sediments of the world—just a few centimetres below the surface—they would be rewarded with the distinctive “rotten egg” smell of H2S. Some 300 million tons of sulfide are produced annually in sediments and associated waters and about 96% of this through the activities of sulfate-reducing microorganisms. Perhaps surprisingly, sedimentary sulfide dwarfs the amount of sulfur produced by volcanic activity.
The sulfate molecule is remarkably stable under ambient conditions, and inorganic reduction to sulfide is kinetically inhibited at temperatures less than ~100 °C. Sulfate-reducing microorganisms bring a small set of enzymes to bear on the process that facilitates the intracellular 8-electron reduction of S(VI) in sulfate to S(−II). This reduction is coupled to the oxidation of organic carbon, ultimately, to CO2 (the “microbial sulfate reduction” equation in Fig. 1). In this process, sulfate reduction is, therefore, coupled as an electron acceptor to the energetics of microbial metabolism, and this is described as dissimilatory sulfate reduction (DSR). This contrasts to assimilatory sulfate reduction, which is a feature of many organisms and refers to the process where the sulfide produced is supplied for biosynthesis (i.e. assimilated) with no energy gain.
Dissimilatory sulfate reduction is also described as sulfate respiration and, although the more familiar oxygen respiration is far more favorable energetically, sulfate is relatively abundant in the modern oceans. There is over 100 times as much dissolved sulfate as dissolved oxygen in average ocean water, and this has resulted in sulfate-reducing microorganisms being amongst the most abundant organisms on Earth. Such organisms are currently so abundant that they oxidize more than half of all the organic matter reaching sediments. This means that sedimentary sulfide formation (the “pyrite formation” equation in Fig. 1) is related to primary productivity (the “primary productivity” equation in Fig. 1), which is the photosynthesis of carbon dioxide and water to produce organic matter and oxygen. The oxygen contents of the oceans and atmosphere are then, ultimately, controlled by the amount of sulfide that is fixed, mostly as pyrite, in sediments. The reverse reaction—the oxidative weathering of pyrite—decreases the proportion of oxygen in the air (the “pyrite oxidation” equation in Fig. 1). Sedimentary sulfides, thus, play a fundamental role in the global sulfur, oxygen, and carbon cycles (Fig. 1) and the formation of sedimentary sulfides largely determines the nature of the surface environment of our planet and how it works.
Of course, it is reasonable to ask if this has always been the case. Sedimentary pyrite has been forming throughout geological time. Figure 2 shows examples of sedimentary pyrite formation from a temporal cross-section of sediments, from framboids found in recent sediments to pyrite from the Eoarchean Era (~3.8 Ga) (Fig. 2). The balanced chemical equations are in Figure 1 allow an accounting of the masses of these compounds in the Earth system. This, in turn, has permitted the construction of great algorithms that describe the variations of these elements in the past, such as for atmospheric oxygen and CO2 (e.g. Berner 1987, 2009). These models can be checked against the amount of pyrite buried in contemporary sedimentary rocks. Indeed, much of what we know about the earliest Earth environment is derived from probes into the geochemistry of sedimentary pyrite. The geochemistry of ancient pyrite has proven to be a powerful proxy for the geochemical evolution of the surface conditions of the planet, particularly the ocean and atmosphere. This itself has led to a deeper understanding of the evolution of life, especially microbial life, over the last 3.8 Ga.
In this review, we describe sedimentary sulfides at the present time and then move on to sedimentary sulfides in the past. Finally, we look at the future of the sedimentary sulfide system and how this will affect the environment.
SEDIMENTARY SULFIDES TODAY
Sulfide in sediments is fixed almost exclusively as pyrite. This is because of the overwhelming abundance of Fe in sediments compared with other metals. Almost is the key word here: in modern sediments, other metastable iron sulfide minerals, such as mackinawite (tetragonal FeS) and greigite (cubic Fe3S4), are occasionally found. Pyrite formation also fixes a variety of trace elements that substitute into the pyrite structure (e.g. Se and As) or that are coprecipitated sulfide minerals, such as chalcopyrite (CuFeS2), sphalerite (ZnS) (Fig. 2C) and as the cobaltian and nickeliferous varieties of pyrite. How these trace elements get concentrated by factors of millions to the levels found in sedimentary pyrite from the exceptionally low concentrations in natural waters has remained a mystery ever since Ed Goldberg’s original discoveries in the 1960s (Table 1).
The total sulfur content of normal, dried marine sediments averages 0.3 wt% and varies up to 2.5 wt%, with most S being concentrated in shallow sediments. Indeed, up to 80% of global microbial sulfate reduction occurs at water depths of <200 m. All of the other sulfur reservoirs in the Earth are dominated by sulfide, mainly combined with Fe(II) in pyrite or pyrrhotite. Our everyday experience of the oxygen-rich Earth’s surface means that sulfate is the most familiar form of sulfur to us. In fact, it is a relatively rare form of sulfur in nature. For the Earth as a whole, for example, less than 0.04% of the total sulfur is in the form of sulfate.
Microbial Production of Sulfide
Sulfate, sulfite, thiosulfate, and elemental sulfur respiration during anoxia is the property of two of the three current domains of life: the Bacteria and the Archaea. Independent of respiration, some bacteria produce sulfide by the disproportionation of sulfite, thiosulfate, or elemental sulfur. Although the capacity to respire sulfur, thiosulfate, and sulfite is widespread, dissimilatory sulfate reduction (DSR) is the main source of biogenic sulfide today.
In DSR, sulfate is the terminal acceptor of electrons from simple organic compounds or hydrogen and the consequent reduction produces sulfide. Until recently, this process was considered to consist of three main enzymatic steps but, recently, a fourth step was identified (Santos et al. 2015). An important ecology for sulfide production in sediments is through consortia of sulfate-reducers and methane-metabolizing microorganisms (methanotrophs) (Figs. 3A, B), which results in some of the highest sulfate-reduction rates recorded in modern sediments (>5 µmol cm−3 d−1). During anaerobic methane oxidation, methanotrophic archaea form zero-valent disulfide, which is then disproportionated by the bacterial partner (Milucka et al. 2012).
The capacity to grow by DSR is not limited to a single modern prokaryotic domain. Most sulfate-reducers belong to the Bacteria, but a few Archaea also display this capability. Sulfate-reducing microorganisms are found in almost all environments on Earth. In marine surface sediments, free-living, single-celled deltaproteobacterial sulfate-reducers are the main source of biogenic sulfide. They far exceed abundances of archaeal or other bacterial sulfate-reducers and may account for more than 108 cells cm−3 of sediment (Ishii et al. 2004). Intriguingly, current genome sequencing of sulfidic biomes reveals a hidden diversity of sulfate-reducers. It seems that many additional, hitherto uncultured, sulfate-respiring microorganisms surely await further discovery.
Myths and Realities
The analytic chemistry of sulfides in modern sediments is complex, and the results are mostly operational. The simplest procedures separate those components that are soluble in dilute mineral acids (acid volatile sulfide, or AVS) and that evolve hydrogen sulfide from those that require oxidizing acids to dissolve. Acid volatile sulfide is often mistakenly equated with FeS, in particular the metastable phases mackinawite (FeS) and greigite (Fe3S4) (Fig. 4). In fact, these latter phases have not been widely reported from marine sediments. They occur more commonly in freshwater sediments. Total reduced sulfur is a relatively robust measurement that reports the sulfur content after reducing all inorganic S in the sediment to S(−II). It is commonly closely correlated with pyrite-S in marine sediments, reflecting the dominance of pyrite as the sulfide mineral in sediments.
One consequence of the disconnect between metastable iron sulfides and AVS is that the classic vertical distribution of sedimentary sulfides—pyrite increases with depth as AVS decreases—is rarely observed. Most pyrite is formed in the top 20 cm of the sediment, but pyrite formation may continue for many millions of years during late diagenesis due to the activities of microorganisms, especially sulfate-reducers, into the deep biosphere.
ANCIENT SEDIMENTARY SULFIDES
Information about the nature of the sedimentary sulfide system in the past comes from two major sources: the geochemistry of sedimentary pyrite, and microbial genomics. Pyrite is an ideal time capsule for probing ancient environments because it is exceptionally stable and only soluble in oxidizing acids. Therefore, pyrite’s chemical and isotopic composition closely reflects its composition when it was formed.
The Smoking Gun
Sulfur occurs naturally as four stable isotopes: 32S, 33S, 34S, and 36S. Of these, 32S and 34S are the most abundant, and measurements of the 34S/32S ratio have been technically accessible since the middle of the 20th century. One of the earliest discoveries was that low temperature, microbial sulfate reduction concentrated the lighter 32S isotope in the product sulfide and that, therefore, the sulfur isotopic composition of pyrite could act as a signature for ancient microbial processes.
In 2000, James Farquhar and colleagues discovered that the less abundant sulfur isotopes in sedimentary pyrite sulfur (pyrite-S) from Archean rocks showed a large deviation from the ratios expected from simple mass-dependent fractionations (Farquhar et al. 2000). It became clear that Archean pyrite commonly showed this effect but that pyrites in younger rocks did not. The pyrite-S was obviously recording a significant change in planet Earth. It appears that the anomalous sulfur isotopic compositions could only happen if there was so little oxygen in the terrestrial atmosphere that the global sulfur cycle was perturbed. Although there had been earlier indications that the Archean atmosphere had been anoxic (reviewed in Canfield 2005), the disappearance of these anomalous sulfur isotope compositions has been described as the smoking gun evidence for an anoxic (i.e. pO2 < 10−5 present atmospheric level) Archean atmosphere (Lyons and Gill 2010).
Further detailed work showed that the timing of this major change, known as the first Great Oxygenation Event (GOE-1), in the composition of the atmosphere could be pinned down to 2.47–2.32 Ga (early Paleoproterozoic). However, recent work suggests that the oxygen content of the atmosphere also varied during the Late Archean and that the ancient atmosphere was intermittently oxygenated (Anbar et al. 2007). The great problem is, of course, the huge swathes of time involved and the relative scarcity of samples from these periods. The vast swings in atmospheric oxygen concentration detailed in the last 500 My of Earth history make the idea of similar variations in the early Earth unsurprising. However, the fact remains that most sedimentary pyrites younger than 2.32 Ga show no anomalous sulfur isotopic signatures whereas those older than 2.47 Ga commonly do. The atmosphere of the early Earth is, therefore, believed to have been generally anoxic.
For some time, geochemists could not understand why, if there was no O2 in the Precambrian atmosphere, the rocks were not full of sulfide minerals. The answer was provided in 1998 by Don Canfield: the intensive sulfide production in sediments today is enabled by the amount of sulfate in the modern oceans; this, in turn, reflects the oxygenated present-day atmospheres (Canfield 1998). This is another counterintuitive idea: significant atmospheric O2 concentrations are required for large amounts of sedimentary sulfides to be produced. In Precambrian oceans, there was limited sulfate and, therefore, the amount of sulfide in the sediments was also limited. Even so, most of the sulfide produced in sediments today is oxidized by sulfur-oxidizing microorganisms. If these organisms were absent from the Earth today, then the whole Earth would bask in the gentle pong of toxic H2S. In the period after GOE-1 (2.3 Ga) and up to the second GOE (GOE-2), which is associated with the explosion in animal diversity ~540 Ma, the O2 content of the atmosphere was still low. Canfield (1998) proposed that the oceans would have often been sulfidic, and oceanic anoxic events would have dominated the environmental history of the Earth during this period. The prevalence of sulfidic conditions had considerable implications for ocean geochemistry and may have inhibited the rapid evolution of multicellular lifeforms.
Evolution of the Sulfur Biome
Sulfate-reducing microorganisms require sulfate to proliferate, and the absence of high sulfate concentrations in the oceans of the past may have meant that dissimilatory sulfate reduction (DSR) was not a dominant microbial process in the early Archean. There is some evidence to support this hypothesis. Sedimentary pyrite enriched in the light sulfur isotope became globally abundant after GOE-1. Before that first major increase in atmospheric oxygen, most sedimentary pyrite did not show this characteristic signature of dissimilatory sulfate reduction. Instead, volcanogenic elemental sulfur and reduced sulfur compounds, such as thiosulfate and sulfite, may have been the primary electron acceptors. Considering the close phylogenetic relationship between sulfate reducers and sulfur disproportionators/respirers, these metabolisms possibly coexisted. Prevalent sulfur disproportionation/respiration in the early Archean is consistent with the fact that thiosulfate, sulfite, and sulfur respiration is more widespread today than DSR (Fig. 5). A phylogenomic reconstruction of the assumed diversity of 4,000 protein families from the Archean also seems to match the geological record, suggesting a shift from sulfite- and thiosulfate-related genes in the early Archean to more sulfate-related genes in the Neoarchean (David and Ahm 2011).
However, there is nothing that upsets a beautiful theory more than an uncomfortable fact. The uncomfortable fact in this case is provided by isotopically light sulfur isotope compositions of 3.5 Ga (Paleoarchean) pyrites from the Dresser Formation in Western Australia. The problem is that widespread isotopically light sulfur in sedimentary pyrite is not found until almost 800 My later. Initially, this light sulfur was reported as a depletion in 34S, and various subsequent attempts were made to explain its occurrence by processes that did not involve DSR. Subsequently, independent analyses showed that none of these alternative theories were tenable. The multiple S isotope systematics were only consistent with DSR and could not be explained by other processes, including sulfur disproportionation. This conclusion was disturbed by later multiple S isotope analyses of other pyrite samples from the 3.5 Ga Dresser Formation, which revealed an unusual depletion in the heaviest sulfur isotope, 36S (Wacey et al. 2015). Although this does not rule out DSR, the logic suggesting that similar isotope fractionation patterns suggest similar enzymatic processes is then not supported.
One idea to resolve the conflict is that on the early Earth, DSR might have been restricted to organisms that thrive at 42–122 °C, the thermophilic sulfate-reducers (cf. Weiss et al. 2016). The 3.5 Ga Dresser Formation pyrites appear to have been formed in an evaporitic basin fed by volcanic sulfate: an environment in which thermophiles might be expected to thrive. Further support for this idea comes from the modeled evolutionary history of the dissimilatory sulfite reductase enzyme. This indicates that DSR probably evolved in the ancestors of thermophilic sulfur-, thiosulfate- and sulfate-reducing archaea (Muller et al. 2015). It is also consistent with the view of Karl Stetter (2006), whose small subunit of ribosomal RNA (SSU rRNA) “Tree of Life” placed hyperthermophiles near the root. However, recent analyses, especially of the genomics of heat-resistant proteins, have cast doubt on this hot-to-cold sequence of evolution, suggesting instead a cold-to-hot sequence, although moderate thermophiles are not excluded (Forterre 2015).
The microfossil evidence for the sulfur organisms that were present in those 3.5 Ga environments shows that they were not, however, equivalent to modern prokaryotes. In some ways, this is not surprising because we did not realize we were sharing our planet with a third domain of life until just 40 years ago (Woese and Fox 1977). It is entirely probable that other domains might have existed in the past but are now entirely extinct. The sulfate-reducing gene set is currently not restricted to a particular domain and, therefore, it is reasonable to speculate that it could have been carried by other domains in the past, including the giant, thick-walled microorganisms which appear to have been involved in sulfide production on the early Earth (cf. Rickard 2012).
After GOE-1, oceanic sulfate concentrations increased and DSR began to dominate the sulfur isotope record. Mesophilic sulfate-reducing bacteria, similar to those dominating marine sulfide production today, likely emerged during, or shortly after, the increase in oceanic sulfate levels between 2.45 Ga and 2.35 Ga. This was followed by several episodes of exchange of DSR genes that fostered a cross-domain dissemination of DSR (Rabus et al. 2015) (Fig. 5).
The Chemical Evolution of Sedimentary Pyrite
Table 1 shows that the average concentrations of trace elements in sedimentary pyrite over geologic time are distinctly different to those in modern sedimentary pyrite. Large et al. (2014) explained this variation by assuming that the concentrations of these elements fixed in pyrite depends to a large extent on how much was available in the sediment porewaters at the time, this being ultimately related to the concentrations of these elements in the oceans and associated waters. The trace element chemistry of sedimentary pyrite might then be used as an effective proxy of past ocean chemistry through geologic time. The results suggest that the Archean oceans were enriched in Fe, Ni, Co, As, Sb, Cu, and Au compared to modern oceans. This Archean ocean composition was controlled by a combination of erosive flux from a continental crust dominated by ultramafic–mafic (komatiite–basalt) rocks and a hydrothermal seafloor flux related to active submarine volcanism (Large et al. 2015a). The high levels of Au and As in marine pyrite from ~3.0 Ga to 2.55 Ga may account for the abundance of both sediment-hosted and orogenic gold deposits of this age around the world. Furthermore, methanogenic bacteria have a relative abundance of Ni- and Co-centered enzymes, and their enrichment in the Archean ocean water may have maintained an atmosphere relatively rich in CH4 (Williams and Frausto Da Silva 1996). This would lead to the potential for the development of microbial consortia such as those formed between methanotrophs and sulfate-reducers in modern sediments (Fig. 3) with the concomitant possibility of the prevalence of organisms with alternative microbial sulfide-producing metabolisms.
In the Paleoproterozoic, following GOE-1 at ~2.4 Ga, there was a gradual decrease in Ni, Co, As, and Au in marine pyrite related to a cessation of komatiitic volcanism. Many bio-essential nutrient trace elements dropped to a minimum through the period 1.7–1.5 Ga, suggesting a low point in productivity and possibly atmospheric O2 for the mid-Proterozoic. The explosion of life at the start of the Phanerozoic is accompanied by one of the most dramatic increases in nutrient supply and productivity recorded by the marine pyrite proxy. Molybdenum, Se, Cu, Ni, Au, and Tl rise by one to two orders of magnitude, peaking at 520–480 Ma (mid-Cambrian to early Ordovician). Following this unprecedented rise in the trace element contents of sedimentary pyrite, Large et al. (2015b) reported five further peaks of trace-element enrichment in sedimentary pyrite at 390 Ma (Middle Devonian), 310 Ma (Middle Pennsylvanian), 240 Ma (Middle Triassic), 150 Ma (Late Jurassic), and 0 Ma, coincident with variations in pO2 in the global atmosphere/ocean system. Large et al. (2015b) interpreted these trace element enrichments in sedimentary pyrite as reflecting nutrient enrichment in contemporary oceans and also noted a coincidence between the corresponding nutrient minima and major mass extinctions of macroscopic life forms in the Phanerozoic.
Conclusions: a sulfidic future?
Of course, the great algorithms describing the biogeochemical cycles of the elements can also be run forwards in time and used to model the fate of our environment over the next decades and centuries. The global biogeochemical sulfur cycle has changed dramatically over the last 200 years. For example, the sulfate content of the oceans has increased by 47% since the industrial revolution due to atmospheric pollution and an increase in the flux of sulfate to the oceans. Death zones, where macroscopic life no longer survives, are spreading in the oceans as multicellular life forms are first suffocated as the oxygen is removed and then poisoned as toxic H2S escapes into the water column in the absence of sulfur-oxidizing bacteria. This process is ultimately powered by global warming, which increases primary productivity (the “primary productivity” equation in Fig. 1) and thereby increases sulfate reduction (the “microbial sulfate reduction” equation in Fig. 1) and sedimentary sulfides. Global warming exacerbates the problem because the oceans become more stratified and the connection between the atmosphere and oceans becomes more restricted. It appears as though humankind needs to be prepared for a sulfidic future.
The increased significance of sulfidic sediments to our future well-being means that we are only seeing the tip of the iceberg of the significance of sedimentary sulfides to future research in the natural sciences. The biogeochemistry of sedimentary sulfides provides a unique probe into how the Earth works and the evolution of both the Earth itself and the life forms that populated it. The burgeoning area of sulfide studies is destined to continue to be a focus for some of the most exciting and rewarding research in the natural sciences.
We thank Ross Large, Rob Raiswell, Michael Böttcher and an anonymous reviewer for their critical comments on an earlier draft of the manuscript.
Anbar AD and 10 coauthors (2007) A whiff of oxygen before the Great Oxidation Event? Science 317: 1903-1906
Berner RA (1987) Models for carbon and sulfur cycles and atmospheric oxygen – application to Paleozoic geologic history.
American Journal of Science 287: 177-196
Berner RA (2009) Phanerozoic atmospheric oxygen: new results using the GEOCARBSULF model. American Journal of Science 309: 603-606
Boetius A and 9 coauthors (2000) A marine consortium apparently mediating anaerobic oxidation of methane. Nature 407: 623-626
Canfield DE (1998) A new model for Proterozoic ocean chemistry. Nature 396: 450-453
Canfield DE (2005) The early history of atmospheric oxygen: homage to Robert A. Garrels. Annual Review of Earth and Planetary Sciences 33: 1-36
David LA, Alm EJ (2011) Rapid evolutionary innovation during an Archaean genetic expansion. Nature 469: 93-96
Farquhar J, Bao H, Thiemens M (2000) Atmospheric influence of Earth’s earliest sulfur cycle. Science 289: 756-758
Forterre P (2015) The universal tree of life: an update. Frontiers in Microbiology 6, doi: 10.3389/fmicb.2015.00717
Ishii K, Mussmann M, MacGregor BJ, Amann R (2004) An improved fluorescence in situ hybridization protocol for the identification of bacteria and archaea in marine sediments. FEMS Microbiology Ecology 50: 203-212
Large RR and 11 coauthors (2014) Trace element content of sedimentary pyrite as a new proxy for deep-time ocean–atmosphere evolution. Earth and Planetary Science Letters 389: 209-220
Large RR and 11 coauthors (2015a) Gold in the oceans through time. Earth and Planetary Science Letters 428: 139-150
Large and 10 coauthors (2015b) Cycles of nutrient trace elements in the Phanerozoic ocean. Gondwana Research 28: 1282-1293
Lyons TW, Gill BC (2010) Ancient sulfur cycling and oxygenation of the early biosphere. Elements 6: 93-99
Milucka J and 9 coauthors (2012) Zero-valent sulphur is a key intermediate in marine methane oxidation. Nature 491: 541-546
Müller AL, Kjeldsen KU, Rattei T, Pester M, Loy A (2015) Phylogenetic and environmental diversity of DsrAB-type dissimilatory (bi)sulfite reductases. ISME Journal 9: 1152-1165
Rabus R and 5 coauthors (2015) A post-genomic view of the ecophysiology, catabolism and biotechnological relevance of sulphate-reducing prokaryotes. Advances in Microbial Physiology, 66: 55-321
Rickard D (2012) Sulfidic Sediments and Sedimentary Rocks. Elsevier, Amsterdam, 816 pp
Santos AA and 6 coauthors (2015) A protein trisulfide couples dissimilatory sulfate reduction to energy conservation. Science 350: 1541-1545
Stetter KO (2006) Hyperthermophiles in the history of life. Philosophical Transactions of the Royal Society B: Biological Sciences 361: 1837-1842
Wacey D, Noffke N, Cliff J, Barley ME, Farquhar J (2015) Micro-scale quadruple sulfur isotope analysis of pyrite from the ~3480 Ma Dresser Formation: New insights into sulfur cycling on the early Earth. Precambrian Research 258: 24-35
Weiss MC and 6 coauthors (2016) The physiology and habitat of the last universal common ancestor. Nature Microbiology 1, doi: 10.1038/nmicrobiol.2016.116
Williams RJP, Frausto Da Silva JJR (1996) The Natural Selection of the Chemical Elements: The Environment and Life’s Chemistry. Clarendon Press, Oxford, 672 pp
Woese CR, Fox GE (1977) Phylogenetic structure of the prokaryotic domain: the primary kingdoms. Proceedings of the National Academy of Sciences of the United States of America 74: 5088-5090