Biological carbon precursor to diagenetic siderite with spherical structures in iron formations

Abstract

During deposition of Precambrian iron formation, the combined sedimentation of ferrihydrite and phytoplankton biomass should have facilitated Fe(III) reduction during diagenesis. However, the only evidence for this reaction in iron formations is the iron and carbon isotope values preserved in the authigenic ferrous iron-containing minerals. Here we show experimentally that spheroidal siderite, which is preserved in many iron formation and could have been precursor to rhombohedral or massive siderite, forms by reacting ferrihydrite with glucose (a proxy for microbial biomass) at pressure and temperature conditions typical of diagenesis (170 °C and 1.2 kbar). Depending on the abundance of siderite, we found that it is also possible to draw conclusions about the Fe(III):C ratio of the initial ferrihydrite-biomass sediment. Our results suggest that spherical to rhombohedral siderite structures in deep-water, Fe-oxide iron formation can be used as a biosignature for photoferrotrophy, whereas massive siderite reflects high cyanobacterial biomass loading in highly productive shallow-waters.

ARTICLE

Received 26 Sep 2012 | Accepted 19 Mar 2013 | Published 23 Apr 2013

During deposition of Precambrian iron formation, the combined sedimentation of ferrihydrite and phytoplankton biomass should have facilitated Fe(III) reduction during diagenesis. However, the only evidence for this reaction in iron formations is the iron and carbon isotope values preserved in the authigenic ferrous iron-containing minerals. Here we show experimentally that spheroidal siderite, which is preserved in many iron formation and could have been precursor to rhombohedral or massive siderite, forms by reacting ferrihydrite with glucose (a proxy for microbial biomass) at pressure and temperature conditions typical of diagenesis (170 C and 1.2 kbar). Depending on the abundance of siderite, we found that it is also possible to draw conclusions about the Fe(III):C ratio of the initial ferrihydrite-biomass sediment. Our results suggest that spherical to rhombohedral siderite structures in deep-water, Fe-oxide iron formation can be used as a biosignature for photoferrotrophy, whereas massive siderite reects high cyanobacterial biomass loading in highly productive shallow-waters.

DOI: 10.1038/ncomms2770

Biological carbon precursor to diagenetic siderite with spherical structures in iron formations

Inga Khler1, Kurt O. Konhauser2, Dominic Papineau3,4, Andrey Bekker5 & Andreas Kappler1

1 Geomicrobiology Group, Center for Applied Geoscience, University of Tuebingen, Sigwartstrasse 10, 72076 Tuebingen, Germany. 2 Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta, Canada T6G 2E3. 3 Department of Earth and Environmental Sciences, Boston College, 140 Commonwealth Avenue, Devlin Hall 213 Chestnut Hill, Massachussets 02467, USA. 4 Geophysical Laboratory, Carnegie Institution of Washington, Washington DC 20015, USA. 5 Department of Geological Sciences, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2. Correspondence and requests for materials should be addressed to A.K. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 4:1741 | DOI: 10.1038/ncomms2770 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2770

Iron formations (IFs) are marine chemical sedimentary rocks that were precipitated throughout much of the Archaean and Paleoproterozoic before B1.85 Ga (ref. 1). The precursor

sediments, including amorphous ferric oxyhydroxides, such as ferrihydrite (Fe(OH)3), required an oxidative mechanism to transform the dissolved hydrothermally-derived Fe(II) into solid-phase Fe(III). It has been suggested that abiotic UV-light dependent oxidation of Fe(II) led to the precipitation of ferric hydroxide minerals2,3, but later experiments have shown that in seawater containing high concentrations of dissolved Si(OH)4 and

HCO3 the oxidation effects of either UVA or UVC are negligible compared to the precipitation of ferrous iron-silicates4. In contrast, two different roles for bacteria have been inferred in Fe(II) oxidation. The rst is based on the production of O2 by

oxygenic phototrophic cyanobacteria. Their Precambrian ancestors ourished when nutrients were available and their metabolically-produced O2 abiotically oxidized dissolved Fe(II) to

Fe(III), followed by the hydrolysis and precipitation of ferrihydrite5. The second postulated role for bacteria in IF precipitation involves Fe(II)-oxidizing microorganisms, including microaerophilic chemolithoautotrophic bacteria (for example, Gallionella ferruginea) that thrive today in suboxic waters and sediments6. Additionally, anoxygenic photoautotrophic Fe(II)-oxidizing bacteria (photoferrotrophs) may have coupled the C and Fe cycles before the evolution of oxygenic photosynthesis79. A number of experimental studies have conrmed that various purple and green phototrophic bacteria can use Fe(II) as a reductant for CO2 xation10. However, the relative inuence of phototrophic bacteria on the origin of IF is not fully understood, because direct evidence for their involvement in Fe(II) oxidation has not been found and the fate of organic remains and ferrihydrite precipitates on the seaoor are still poorly characterized.

Given that the bulk of the ocean water column was anoxic throughout the Archaean and Paleoproterozoic, ferrihydrite represented a favourable electron acceptor for in situ oxidation of the cellular biomass remains associated with precursor IF sediments. Some Archaean magnetite-rich IF1115 have highly negative d56Fe values that are comparable to those measured from dissolved Fe(II) experimentally produced by cultures of dissimilatory Fe(III)-reducing bacteria16, pointing to the antiquity of anaerobic dissimilatory Fe(III) reduction15,17. Alternatively, abiotic reactions between Fe(III) oxyhydroxides and organic carbon are expected to occur at temperatures above the threshold for microbial life during late-stage diagenesis and metamorphism18, which based on modern microbes is around 120 C (ref. 19). In either case, coupling the reduction of Fe(III) minerals to the oxidation of organic matter provides a simple explanation for the generally low content of organic carbon in IF (o0.5% (refs 1,20)), as well as the light carbon and iron isotopic values of carbonate minerals1,15,21,22.

A number of petrographic studies have described the secondary origin of magnetite in IF2326. Recently, Li et al.27 reported the detection of nanocrystals of apatite in association with magnetite from the 2.48 Ga Dales Gorge IF in Western Australia; the latter had a crystal chemistry identical to modern biogenic magnetite. These considerations and features point to the original presence of biomass in IF precursor sediments, but also indicate that the organic carbon served as an electron donor during Fe(III) reduction. Beside magnetite and haematite, siderite is one of the most abundant Fe-bearing minerals in IF. The origin of siderite has been suggested to involve primary precipitation in the water column through the reaction of Fe2 with bicarbonate from seawater as soon as the solubility product of siderite (controlled by the Fe2 and carbonate/bicarbonate concentrations in the ocean water) is exceeded28,29. Alternatively,

siderite can be formed during the diagenetic (biological or non-biological), and subsequent metamorphic, oxidation of microbial biomass to CO2/bicarbonate coupled to the reduction of Fe(III), followed by reactions of the product Fe2 with the sediment pore-water bicarbonate30. Here, we show that spherical siderite in deep-water IF represents a biosignature for photoferrotrophy, whereas massive siderite reects high cyanobacterial biomass in shallow-water.

ResultsExperimental siderite formation. We experimentally tested whether the diagenetic transformation of biomass and primary Fe(III) oxyhydroxide minerals can form siderite, and whether that siderite is of similar size and morphology to those observed in IF. To this end, we incubated ferrihydrite (as a proxy for primary ferric oxyhydroxides) and glucose (as a proxy for microbial bio-mass) in gold capsules at 170 C and 1.2 kbar (for a more detailed rationale explaining the choice of reactants and for methodological details see Supplementary Information). These chosen conditions simulate the degree of metamorphism in some of the major IF in South Africa and Western Australia that have been exposed to metamorphic grades below the prehnite-pumpellyite facies3133. After 24 h of incubation, haematite, magnetite and siderite were identied as reaction products by their optical properties (Fig. 1). Detailed microscopic analysis showed that the siderite was commonly present as spherical grains characterized by strong bireectance and anisotropism (Fig. 2a,c). As no signicant concentrations of Mg2 and Ca2 were present in our experiments, we deduce that the carbonate spheres are siderite. This has also been independently conrmed by Mssbauer spectroscopy and XRD (data not shown).

Siderite was present in close association with ferrihydrite and haematite, suggesting that the reduction of ferrihydrite or haematite coupled to the oxidation of biomass (for example, glucose) formed siderite, Fe2 , OH , and potentially even H2, according to the following four reactions:

4Fe(OH)3 CH2O-FeCO3 3Fe2 6OH 4H2O

2Fe2O3 CH2O 2H2O-FeCO3 3Fe2 6OH

2Fe(OH)3 2CH2O-2FeCO3 2H2O 3H2

Fe2O3 2CH2O H2O-2FeCO3 3H2

Magnetite

Siderite

Haematite

50 m

Figure 1 | Iron(II) and iron(III) minerals formed during pressure-temperature diagenesis. Reected light microscopy image of a mixture of ferrihydrite and glucose after 1 day of pressure and temperature treatment at 1.2 kbar and 170 C, respectively. Haematite and magnetite have formed along with siderite. All three minerals can be found in close association with each other. Haematite forms large anhedral to subhedral crystals, which show small magnetite inclusions. Siderite is present as the typical grey sub-spherical structures.

2 NATURE COMMUNICATIONS | 4:1741 | DOI: 10.1038/ncomms2770 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2770 ARTICLE

Experiment

Gunflint Iron Formation

a b

Magnetite

Chert

Siderite

Haematite

Siderite

Siderite

50 m 100 m

50 m

c d

Magnetite

Siderite

Chert

50 m

Figure 2 | Spherical siderite structures in diagenesis experiments and Precambrian IFs. Reected light microscopy images of siderite spheres formed in diagenetic experiments from a mixture of ferrihydrite and glucose (a,c) compared with spherical siderite structures in IF (b,d). Ferrihydrite and glucose with an electron ratio of 0.6 was treated at 1.2 kbar and 170 C for 14 days. (a) Shows haematite, magnetite and spherical siderite formed in the diagenetic experiments in comparison to analogous spherical siderite structures in chert from the Gunint IF near Kakabeka Falls, Ontario (b). (c) Ferrihydrite and glucose mixture (electron ratio of 2.4) under crossed polars. Magnetite is surrounded by siderite spheres that show distinct polarization crosses. These spheres are similar to siderite structures in the carbonate facies of the upper and lower Gunint IF, Ontario34.

Comparison with siderite in IFs. Our experiments show that spherical siderite can form from the abiotic diagenetic oxidation of organic carbon coupled to a reduction of ferric iron during simulated late-stage diagenesis (Fig. 2a,c). Figure 2b,d show a comparison of diagenetically formed siderite spheres in our experiments with spherical siderite structures in the 1.88 Ga Gunint IF in Canada3438. Siderite spheres with sizes of tens of microns can be observed in both the experimental and natural mineral assemblages where they appear as individual grains as well as in the form of aggregates when the grains coalesce. Similar structures of spherical siderite have also been described in IFs in the 2.48 Ga Dales Gorge Member, Western Australia39,40 and1.88 Ga Riverton IF, Michigan, MI, USA41, although they are not as common in these successions as in the Gunint IF. Such occurrences suggest that siderite spheres may have been a common feature of IFs before they were recrystallized during late diagenesis and metamorphism to rhombohedral/subhedral siderite grains or formed massive siderite beds. Consistent with this suggestion, it has been shown that in the Gunint IF some siderite rhombohedra contain spheres in their cores and there is a clear progression from siderite spheres to siderite rhombohedra without any trace of spheres left37,42. This progression was also found in the Dales Gorge Member, where siderite spheres and rhombohedra occur in close proximity to each other (Fig. 3c). Note that some spheres already show signs of a recrystallization.

Inuence of initial Fe(III):C ratios on siderite formation. In order to determine how varying levels of biomass might inuence siderite formation and the internal morphology of spheres, we

performed a second series of experiments using four times more glucose. The electron ratio, calculated with the total number of electrons potentially released from carbon in glucose divided by the number of electrons that can potentially be accepted by the Fe(III), was thus increased from 0.6 in the rst experiments to 2.4 (see Supplementary Information). Experiments with an electron ratio of 0.6 simulate an excess of iron as would have occurred in a scenario in which the Fe2 would be oxidized by anoxygenic photoferrotrophs (for example, Chlorobium sp.) or micro-aerophilic chemolithoautotrophs (for example, Gallionella sp.) (Fig. 3): preferential deposition of the dense Fe(III) minerals with low content of organic carbon has been demonstrated experimentally for photoferrotrophs43. In comparison, an electron ratio of 2.4 represents a system with a higher biomass input to the sediment as would be the case if there was either additional electron donors such as H2, or if a microbial metabolism, such as photosynthesis by cyanobacteria, occurred in the water column and was not directly tied to Fe(II) oxidation.

Our diagenesis experiments showed that siderite formed at both electron ratios and that it most commonly had a spherical shape (shown by black and white arrows in Fig. 2a,c) up to 20 mm in diameter. Individual spheres sometimes coalesced to form aggregates. Associated with the siderite were haematite (in experiments with an electron ratio of 0.6) and magnetite (in both experiments). Magnetite was commonly found to be interspersed with haematite grains or as rims around haematite cores. In the experiment with an electron ratio of 2.4, magnetite grains were larger and often surrounded by siderite (Fig. 2c). The sub-spherical siderite crystals in the 2.4 electron ratio experiments showed distinct polarization crosses under crossed polars

NATURE COMMUNICATIONS | 4:1741 | DOI: 10.1038/ncomms2770 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2770

a b

Nutrient flux

O2

Pyrite

Chemocline

Hydrothermal Fe(II)

Corg

Fe(III)/ Corg

0.5 mm

100 m

Siderite

CO2, CH4, H2, Fe(II), NH3+

Fe(III)

c

Siderite

Photosynthetic community of cyanobacteria (stromatolites)

Haematite and magnetite with disseminated siderite

Massive siderite

Black shale

Carbonate Cyanobacteria

Anoxygenic photoferrotrophs

Microaerophilic chemolithoautotrophs

Figure 3 | Origin of massive siderite beds and spherical siderite structures in different depositional environments. Model summarizing the originof shallow- and deep-water IF siderite (a). During periods of enhanced submarine volcanism, hydrothermal uids leached Fe(II) from volcanics and exhaled it to the oceans. Fe(II) upwelling from the deep ocean resulted in the growth of microaerophilic chemolithoautotrophs and anoxygenic photoferrotrophs in the deep-water, photic zone above and below the chemocline, respectively. Due to low-cell densities in the active Fe(II)-oxidation zone and a much smaller Corg reduction potential in sediments with respect to the accumulated Fe(III), haematite and magnetite with disseminated spherical to rhombohedral siderite formed during diagenesis. By contrast, high riverine nutrient uxes to shallow-water environments led to cyanobacterial blooms; this resulted in a C:Fe ratio of 1:4 or more, which is equal to or greater than the reduction ratio in sediments. This led to the formation of large quantities of siderite grains, which coalesced during diagenesis and metamorphism to form massive siderite beds. (b) Massive siderite beds formed in a shallow-water environment in association with stromatolites (with lamina highlighted by pyrite crystals) in the volcanically-dominated ca. 2.7 Ga Michipicoten greenstone belt, Wawa subprovince of the Superior province, Canada (reected light image, Helen IF, MacLeod mine51). (c) In the deeper-water environments, sub-spherical to rhombohedral siderite structures formed during early diagenesis to low-grade metamorphism of IFs as shown in the photomicrograph under cross-nicols of the ca. 2.48 Ga Dales Gorge IF of Western Australia.

(Fig. 2c), suggesting either concentrical or radial crystal growth in siderite spheres34,44. Polarization crosses under crossed polars and radial crystal growth were only observed in capsules with a higher electron ratio, but were also observed in siderite spheres from the Gunint IF (Fig. 2d; see also ref. 45) and from the Riverton IF of the Iron River district, Michigan41. The results thus suggest that these structures can be considered as an indicator for high organic carbon loading into the sediments. Consistent with this suggestion, siderite spheres in IFs often contain organic carbon inclusions42. To our knowledge, polarization crosses under crossed polars have not yet been described for siderite minerals formed by microbial reduction of Fe(III) minerals either in laboratory or in nature. Therefore, by comparison with previously described siderite spheres in IF, our experiments suggest that at least some siderite spheres were formed by abiotic diagenetic formation of siderite via reduction of Fe(III) oxyhydroxides with biomass.

Inuence of silica on siderite formation. To test whether spherical siderite would also form in the presence of Si concentrations similar to those expected for the Archaean oceans, we carried out experiments with Si-coated ferrihydrite and glucose (Si:Fe ratio of 0.08, electron ratio 0.6, for details see Supplementary Information). We found that in the presence of Si, siderite also formed within 14 days and was visible both as individual spheres and as aggregates (Fig. 4).

DiscussionAlthough siderite can form via both abiotic and biotic mechanisms21,17,30, specic petrographic features of spherical siderite in

IF have been suggested to represent remnants of microbial cells based on their shape and internal structure35,36. Heaney and Veblen34 argued for an indirect biological origin of nano-scale siderite spheres from the Gunint IF as distinct cores of apatite were found within the siderite spheres. Based on a simple mass-balance analysis, they inferred that it was the accumulation of thousands of individual organisms that accounted for the relatively high amount of apatite inside of the spheres and for the spheres themselves. However, the described structures in the Gunint IF have a size range between 24115 mm (Kazmierczak et al.35 and Fig. 2b,d), and thus do not resemble the typical size of bacterial cells. In either case, several IF, including the Dales Gorge Member of Western Australia39,40, the Riverton IF in Michigan41, and the Gunint IF in Ontario38, have been reported to contain these distinctive structures. The presence of spherical siderite in IFs with different ages and from varied depositional settings, shallow and deep marine, suggests a common mechanism of their formation.

A key conclusion from our experiments is that the presence of siderite in IFs may allow a rst-order estimate of the initial amount of sedimentary organic matter, and potentially even characterization of the ecosystem that grew in the photic zone during IF deposition. Using the Dales Gorge Member as an example, we calculated that based on measured FeTOTAL

content, a plausible depositional rate of 1 mm per year46, and a post-precipitation reduction efciency of 70% before burial47, the amount of Fe(III) precipitated could have been as high as 45.3 mol per year m2 (for details of calculation see Supple

mentary Information). Using an experimental rate of Fe(II) oxidation of 1.1 10 11 mol per cell by chemolithoautotrophic

bacteria48, such as Gallionella ferruginea (one of the species

4 NATURE COMMUNICATIONS | 4:1741 | DOI: 10.1038/ncomms2770 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2770 ARTICLE

a

Haematite

Siderite

50 m

50 m

b

Figure 4 | Iron(II) and iron(III) minerals formed during pressure-temperature diagenesis in the presence of silica. Reected light microscopy image (a) of minerals formed in diagenetic experiments with a mixture of silica-coated ferrihydrite and glucose (Si:Fe ratio of 0.08, electron ratio 0.6). Siderite is visible as aggregates and as single globular structures. b shows an image of the same area under crossed polars.

believed to have played a prominent role in ferric iron precipitation in some IF49see SOM), it is further possible to estimate the cell density of bacteria growing in the photic zone of early Precambrian seawater, as the rate of Fe(III) precipitation estimated above would necessitate population of chemolithotrophic bacteria on the order of 105 cells per ml (for details of calculation see Supplementary Information) throughout the Fe(II) oxidation zone of the water column with a biologically active depth of 10 m. Similar cell density would have existed if Fe(II) oxidation was primarily driven by photoferrotrophic bacteria.

Putting these numbers into context, we can envision two different scenarios that would have ultimately led to the different diagenetic/metamorphic minerals found in IF (Fig. 3). It has long been recognized that in Neoarchean-early Paleoproterozoic IF-bearing basins with a well-preserved shallow- to deep-water depositional prole, deep-water Fe-oxide facies grade into shallow-water Fe-carbonate facies28. We suggest that during periods of high Fe(II) upwelling, the deep-water, photic zone (that also contained low O2 concentrations due to the activity of cyanobacteria) was populated with Gallionella-type cells above the chemocline and Chlorobium-type cells below the chemocline, but throughout the active Fe(II) oxidation zone, cell densities were on average 105 cells per ml or less. As Gallionella, in

particular, forms Fe(III) with a C:Fe ratio of 1:6, but the reduction of Fe(III) in sediments with dead biomass has a C:Fe ratio 1:4, then excess Fe(III) is accumulated, leading to haematite and magnetite formation with small amount of disseminated siderite found in most deep-water IF, such as the Dales Gorge Member (Fig. 3). This also applies for photoferrotrophic bacteria, as shown by Posth et al.43, due to only partial deposition of the biomass while all precipitated Fe(III) minerals sediment to the ocean oor.

By contrast, in the shallow-water environments, where only a limited amount of upwelling Fe(II) was supplied, the relative numbers of photoferrotrophs/Gallionella were lower (because their electron donor Fe2 is limiting under such conditions), whereas higher riverine nutrient input could have resulted in cyanobacterial blooms. This would have led to the production of ferric iron (via the indirect reaction of O2 with ferrous iron) at a

C:Fe ratio of 1:4 (or even greater). This is equivalent to, or even greater, than the reduction ratio, thereby minimizing the amount of Fe(III) buried. Importantly, the cyanobacteria would have produced biomass that was not tied to Fe(II) oxidation, thereby increasing the overall C:Fe ratio in the sediments. Furthermore, given the possibility that signicant concentrations of H2 were also present (via fermentation) in the water column50, autotrophic photosynthesis decoupled from Fe(II) oxidation would have also added biomass, causing a rise in the C:Fe ratio. Therefore, massive siderite with precursor spherical siderite would have formed in the transition zone to the shallow-water environments. Such siderites have been described in the ca. 2.7 Ga Helen iron formation, Ontario51. Disseminated siderite spheres or, at higher metamorphic grade, rhombohedral to anhedral-subhedral siderite grains, however, would be preserved in deep-water, Archaean Fe-oxide facies IF (Fig. 3c). Consistent with this ecological zonation, stromatolites have been described from massive siderite deposits51 (Fig. 3b), whereas disseminated siderite spheres and, more commonly, rhombohedral to anhedral-subhedral siderite grains were found in the Gunint3438, Riverton41 and Dales Gorge Fe-oxide facies IF39,40.

Although a biological role for the origin of IF has been suggested in the past24, we have, for the rst time, provided experimental evidence that organic biomass in association with Fe(III) mineral precipitates forms spherical siderite structures under diagenetic conditions. This, in turn, can be used to argue for the presence of cellular biomass in association with the primary Fe(III) precipitates in IF, thereby implying that biological activity was indeed involved in IF deposition. Perhaps most intriguingly, the abundance of siderite in IF might also be used to estimate the initial amount of sedimentary organic matter in Precambrian ferruginous marine environments and its predominant style of metabolism. Our study points to diverse microbial communities in the Archaean oceans with the shallow-water environments populated by cyanobacterial communities, whereas photoferrotrophs and microaerophilic chemolithoautotrophs were abundant in the deep-water photic zone.

Methods

Setup and analysis of gold capsule diagenesis simulation experiments. Ferrihydrite (chemically synthesized) and glucose (as proxy for biogenic Fe(III) precipitates and microbial biomass) were weighted, mixed and carefully homogenized before they were lled into 3 cm long gold capsules, which were cut from a 1-m long gold tube with an outer diameter of 2.5 mm and a wall thickness of0.2 mm. The capsules were lled with 100 mg of ferrihydrite and with either 4.23 or16.9 mg glucose. The lower amount of glucose represents an electron ratio of 0.6, which is the ratio of the electrons potentially released from the carbon in the glucose divided by the electrons potentially accepted by the Fe(III) in the ferrihydrite. The larger amount of glucose represents an electron ratio of 2.4. After the lling, the capsules were closed mechanically and sealed with an arc welding. Closed capsules were then placed into rapid quench autoclaves for pressure and temperature treatment at 1.2 kbar and 170 C for 14 days. To stop the reactions, capsules were dropped from the sample chamber into the water-cooled part of the

NATURE COMMUNICATIONS | 4:1741 | DOI: 10.1038/ncomms2770 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2770

autoclave. After incubation, the capsules were either prepared as thick sections and embedded in epoxy resin for identication of the mineral phase content and the mineral spatial arrangement by reected light microscopy, or the material inside the capsules was taken out for Mssbauer spectroscopy and XRD to identify the iron mineral phases.

References

1. Klein, C. Some precambrian banded iron formations (BIFSs) from around the world. Their age, geologic setting, mineralogy, metamorphism, geochemistry, and origin. Am. Mineral. 90, 14731499 (2005).

2. Cairns-Smith, A. G. Precambrian Solution photochemistry, inverse segregation, and banded iron formations. Nature 276, 807808 (1978).

3. Braterman, P. S., Cairns-Smith, A. G. & Sloper, R. W. Photo-oxidation of hydrated Fe2- Signicance for banded iron formations. Nature 303, 163164 (1983).

4. Konhauser, K. O. & Amskold, L. et al. Decoupling photochemical Fe(II) oxidation from shallow-water BIF deposition. Earth Planet. Sci. Lett. 258, 87100 (2007).

5. Cloud, P. Signicance of the Gunint (Precambrian) ora. Science 148, 2735 (1965).

6. Holm, N. The 13C/12C ratios of siderite and organic matter of a modern metalliferous hydrothermal sediment and their implications for banded iron formations. Chem. Geol 77, 4145 (1989).

7. Hartmann, H. Microbial MatsStromatolites (eds Cohen, Y., Castenholz, R.W., Halvorson, H. O. & Boylen, C. W.) 451453 (John Wiley and Sons, 1984).8. Widdel, F. et al. Ferrous Iron oxidation by anoxygenic phototrophic bacteria. Nature 362, 834836 (1993).

9. Kappler, A., Pasquero, C., Konhauser, K. & Newman, D. K. Deposition of banded iron formations by anoxygenic phototrophic Fe(II)-oxidizing bacteria. Geology 33, 865868 (2005).

10. Hegler, F., Posth, N. R., Jiang, J. & Kappler, A. Physiology of phototrophic iron(II)-oxidizing bacteriaimplications for modern and ancient environments. FEMS Microbiol. Ecol. 66, 250260 (2008).

11. Johnson, C. M., Beard, B. L., Beukes, N. J., Klein, C. &Leary, J. M. Ancient geochemical cycling in the earth as inferred from Fe isotope studies of banded iron formations from the Transvaal Craton. Contrib. Mineral. Petr 144, 523547 (2003).

12. Yamaguchi, K. E., Johnson, C. M., Beard, B. L. & Ohmoto, H. Biogeochemical cycling of iron in the ArcheanPaleoproterozoic Earth: constraints from iron isotope variations in sedimentary rocks from the kaapvaal and pilbara cratons. Chem. Geol 218, 135169 (2005).

13. Rouxel, O. J., Bekker, A. & Edwards, K. J. Iron isotope constraints on the Archean and Paleoproterozoic ocean redox state. Science 307, 10881091 (2005).

14. Dauphas, N. et al. Clues from Fe Isotope Variations on the Origin of Early Archean BIFs from Greenland. Science 306, 20772088 (2004).

15. Czaja, A. D. et al. Iron and carbon isotope evidence for ecosystem and environmental diversity in the similar to 2.7 to 2.5 Ga Hamersley Province, Western Australia. Earth Planet. Sci. Lett. 292, 170180 (2010).

16. Johnson, C. M. & Beard, B. L. Biogeochemical cycling of iron isotopes. Science 309, 10251027 (2005).

17. Heimann, A. et al. Fe, C, and O isotope compositions of banded iron formation carbonates demonstrate a major role for dissimilatory iron reduction inB2.5

Ga marine environments. Earth Planet. Sci. Lett. 294, 818 (2010).18. Perry, E. C., Tan, F. C. & Morey, G. B. Geology and stable isotope geochemistry of the Biwabik Iron Formation, Northern Minnesota. Econ. Geol. 68, 11101125 (1973).

19. Kashe, K. & Lovley, D. R. Extending the upper temperature limit for life. Science 301, 934 (2003).

20. Gole, M. J. & Klein, C. Banded iron-formations through much of the Precambrian time. J. Geol. 89, 169183 (1981).

21. Walker, J. C. G. Suboxic diagenesis in banded iron formations. Nature 309,

340342 (1984).

22. Craddock, P. R. & Dauphas, N. Iron and carbon isotope evidence for microbial iron respiration throughout the Archean. Earth Planet. Sci. Lett. 303, 121132 (2011).

23. Han, T. M. Microstructures of magnetite as guides to its origin in some Precambrian iron formations. Fortschritte der Mineralogie 56, 105142 (1978).

24. McConchie, D. Western Australia. Precambrian Iron-Formations (eds Appel,P. W. U. & LaBerge, G. L.) 541597 (1987).25. Morris, R. C. Genetic modelling for banded iron-formation of the Hamersley Group, Pilbara Craton, Western Australia. Precambrian Res. 60, 243286 (1993).

26. Krape, B., Barley, M. E. & Pickard, A. L. Hydrothermal and resedimented origins of the precursor sediments to banded iron formation: sedimentological evidence from the early Palaeoproterozoic Brockman Supersequence of Western Australia. Sedimentology 50, 9791011 (2003).

27. Li, J. L., Konhauser, K. O., Cole, D. R. & Phelps, T. J. Mineral ecophysiological data provide growing evidence for microbial activity in banded-iron formations. Geology 39, 707710 (2011).

28. Beukes, N. J., Klein, C., Kaufman, A. J. & Hayes, J. M. Carbonate petrography, kerogen distribution, and carbon and oxygen isotope variations in an early Proterozoic transition from limestone to iron-formation deposition, Transvaal Supergroup, South Africa. Econ. Geol. 85, 663690 (1990).

29. Winter, B. L. & Knauth, L. P. Stable isotope geochemistry of cherts and carbonates from the 2.0 Ga Gunint Iron Formation: implications for the depositional setting, and the effects of diagenesis and metamorphism. Precambrian Res. 59, 283313 (1992).

30. Johnson, C. M., Beard, B. L., Klein, C., Beukes, N. J. & Roden, E. E. Iron isotopes constrain biologic and abiologic processes in banded iron formation genesis. Geochim. Cosmochim. Acta. 72, 151169 (2008).

31. Miyano, T. & Beukes, N. J. Phase relations of stilpnomelane, ferriannite, and riebekite in very low-grade metamorphosed iron formations. Geol. Soc. So. Africa Trans 87, 111124 (1984).

32. Trendall, A. F. The Hamersley Basin. Iron-Formations: Facts and problems (eds Trendall, A. F. & Morris, R. C.) 69129 (1983).

33. Kaufman, A. J., Hayes, J. M. & Klein, C. Primary and diagenetic controls of isotopic compositions of iron-formation carbonates. Geochim. Cosmochim. Acta 54, 34613473 (1990).

34. Heaney, P. J. & Veblen, D. R. An examination of globular dubiomicrofossils in Precambrian banded iron formations using the transmission electron microscope. Precambrian Res. 49, 355372 (1991).

35. Kazmierczak, J. The eukaryotic nature of Eosphaera-like ferriferous structures from the Precambrian Gunint Iron Formation, Canada: A comparative study. Precambrian Res. 9, 122 (1979).

36. LaBerge, G. L. Possible biological origin of Precambrian iron-formations. Econ. Geol. 68, 10981109 (1973).

37. Carrigan, W. J. & Cameron, E. M. Petrological and stable isotope studies of carbonate and sulde minerals from the Gunint Formation, Ontario: evidence for the origin of early Proterozoic iron-formation. Precambrian Res. 52, 347380 (1991).

38. LaBerge, G. L. Microfossils and Precambrian Iron-Formations. Geol. Soc. Am. Bull 78, 331342 (1967).

39. Trendall, A. F. & Blockley, J. G. The iron formations of the Precambrian Hamersley Group, Western Australia; with special reference to the associated crocidolite. W. Austr. Geol. Surv. Bull 119 p 336 (1970).

40. Ayres, D. E. Genesis of iron-bearing minerals in banded iron formation mesobands in the dales gorge member, hamersley group, Western Australia. Econ. Geol. 67, 12141233 (1972).

41. James, H. L. Iron formation and associated rocks in the Iron River district, Michigan. Geol. Soc. Am. Bull. 62, 251266 (1951).

42. LaBerge, G. L., Robbins, E. I. & Han, T. M. A Model for the Biological Precipitation of Precambrian Iron-Formations, A: Geological Evidence, Precambrian Iron-Formations (eds Appel, P. W. U. & LaBerge, G. L.) 6996 (1987).

43. Posth, N. R., Huelin, S., Konhauser, K. O. & Kappler, A. Size, density and composition of cell-mineral aggregates formed during anoxygenic phototrophic Fe(II) oxidation: Impact on modern and ancient environments. Geochim. Cosmochim. Acta. 74, 34763493 (2010).

44. Floran, R. J. & Papike, J. J. Petrology of the low-grade rocks of the Gunint Iron-Formation, Ontario-Minnesota. Geol. Soc. Am. Bull. 86, 11691190 (1975).

45. Lougheed, M. S. Origin of Precambrian iron-formations in the Lake Superior region. Geol. Soc. Am. Bull. 94, 325340 (1983).

46. Konhauser, K. O. et al. Did bacteria form Precambrian banded iron formations? Geology 30, 10791082 (2002).

47. Konhauser, K. O., Newman, D. K. & Kappler, A. The potential signicance of microbial Fe(III) reduction during deposition of Precambrian banded iron formations. Geobiology 3, 167177 (2005).

48. Emerson, D. & Revsbech, N. P. Investigation of an iron-oxidizing microbial mat community located near Aarhus, Denmark eld studies. Appl. Environ. Microbiol. 60, 40224031 (1994).

49. Planavsky, N. et al. Iron-oxidizing microbial ecosystems thrived in late Paleoproterozoic redox-stratied oceans. Earth Planet. Sci. Lett. 286, 230242 (2009).

50. Croal, L. R., Jiao, Y., Kappler, A. & Newman, D. K. Phototrophic Fe(II) oxidation in the presence of H2: implications for Banded Iron Formations.

Geobiology 7, 2124 (2009).51. Hofmann, H. J., Sage, R. P. & Berdusco, E. N. Archean stromatolites in michipicoten group siderite ore at Wawa, Ontario. Econ. Geol. 86, 10231030 (1991).

Acknowledgements

We thank M Nowak, D Russ and I Gill-Kopp for support with sample preparation and pressure/temperature experiments. U Neumann is acknowledged for advice on reected light microscopy. This work was funded by research grants KA 1736/4-1 and 121 from the German Research Foundation (DFG) to A.K., the Promotionsstipendium im

6 NATURE COMMUNICATIONS | 4:1741 | DOI: 10.1038/ncomms2770 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2770 ARTICLE

Promotionsverbund Einblicke in die Bakterien-Material-Wechselwirkungen to I.K., and the Natural Sciences and Engineering Research Council of Canada (NSERC) to Kurt Konhauser and Andrey Bekker. Furthermore, I.K. would like to thank E Swanner, N Posth and M Eickhoff for helpful advice and discussions. D.P. acknowledges support from NASA (grant numbers NNA04CC09A, NNX08AO16G, NNA09DA81A, and NNX12AG14G), Boston College, and the Carnegie Institution of Washington. We thank the drill core facility of the GSWA for providing access and assistance to samples.

Author contributions

Experiments and analyses were conducted by I.K. A.K. and K.K. designed and developed the project. All authors contributed to the data analysis and writing of the manuscript.

Additional Information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Khler, I. et al. Biological carbon precursor to diagenetic siderite with spherical structures in iron formations. Nat. Commun. 4:1741 doi: 10.1038/ ncomms2770 (2013).

NATURE COMMUNICATIONS | 4:1741 | DOI: 10.1038/ncomms2770 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

& 2013 Macmillan Publishers Limited. All rights reserved.