Mark A. van Zuilen

Reassessing the evidence for the earliest traces of life

The isotopic composition of graphite is commonly used as a biomarker in the oldest (>3.5 Gyr ago) highly metamorphosed terrestrial rocks. Earlier studies on isotopic characteristics of graphite occurring in rocks of the approximately 3.8-Gyr-old Isua supracrustal belt (ISB) in southern West Greenland have suggested the presence of a vast microbial ecosystem in the early Archean. This interpretation, however, has to be approached with extreme care. Here we show that graphite occurs abundantly in secondary carbonate veins in the ISB that are formed at depth in the crust by injection of hot fluids reacting with older crustal rocks (metasomatism). During these reactions, graphite forms from the disproportionation of Fe(II)-bearing carbonates at high temperature. These metasomatic rocks, which clearly lack biological relevance, were earlier thought to be of sedimentary origin and their graphite association provided the basis for inferences about early life. The new observations thus call for a reassessment of previously presented evidence for ancient traces of life in the highly metamorphosed Early Archaean rock record.

Graphite and carbonates in the 3.8 Ga old Isua Supracrustal Belt, southern West Greenland

We present a systematic study of abundance, isotopic composition and petrographic associations of graphite in rocks from the ca. 3.8 Ga Isua Supracrustal Belt (ISB) in southern West Greenland. Most of the graphite in the ISB occurs in carbonate-rich metasomatic rocks (metacarbonates) while sedimentary units, including banded iron formations (BIFs) and metacherts, have exceedingly low graphite concentrations. Regardless of isotopic composition of graphite in metacarbonate rocks, their secondary origin disqualifies them from providing evidence for traces of life stemming from 3.8 Ga. Recognition of the secondary origin of Isua metacarbonates thus calls for reevaluation of earlier interpretations that suggested the occurrence of 3.8 Ga biogenic graphite in these rocks. Thermal decomposition of siderite; 6FeCO3 = 2Fe3O4 + 5CO2 + C, is the process seemingly responsible for the graphite formation. The cation composition (Fe, Mg, Mn, and Ca) of the carbonate minerals, carbon isotope ratios of carbonates and associated graphite and petrographic assemblages of a suite of metacarbonates support the conclusion that multiple pulses of metasomatism affected the ISB, causing the deposition of Fe-bearing carbonates and subsequent partial disproportionation to graphite and magnetite. Equilibrium isotope fractionation between carbonate and graphite in the rocks indicates peak metamorphic temperatures between 500 and 600 ◦ C, in agreement with other estimates of metamorphic temperature for the ISB.

Questioning the evidence for Earth's earliest life—Akilia revisited

It has been argued that apatite crystals containing inclusions of isotopically light graphite in a quartz-pyroxene rock from the island of Akilia, southwest Greenland, represent the earliest (older than 3.85 Ga) traces of life on Earth. Although the age and protolith of this rock have been subjects of vigorous discussions, the occurrence of isotopically light graphite inclusions in Akilia apatite has so far not been debated in the literature. We present here the results of petrographic analysis of 17 different Akilia samples, including the actual sample (G91-26) used in the original study. Our finding that none of the apatite crystals in these samples contain graphite inclusions indicates that the Akilia apatite has no bearing on claims pertaining to a past record of life on Earth.

Calibration of carbonate composition using micro-Raman analysis: application to planetary surface exploration.

Stromatolite structures in Early Archean carbonate deposits form an important clue for the existence of life in the earliest part of Earths history. Since Mars is thought to have had similar environmental conditions early in its history, the question arises as to whether such stromatolite structures also evolved there. Here, we explore the capability of Raman spectroscopy to make semiquantitative estimates of solid solutions in the Ca-Mg-Fe(+Mn) carbonate system, and we assess its use as a rover-based technique for stromatolite characterization during future Mars missions. Raman microspectroscopy analysis was performed on a set of carbonate standards (calcite, ankerite, dolomite, siderite, and magnesite) of known composition. We show that Raman band shifts of siderite-magnesite and ankerite-dolomite solid solutions display a well-defined positive correlation (r(2) > 0.9) with the Mg# = 100 x Mg/(Mg + Fe + Mn + Ca) of the carbonate analyzed. Raman shifts calibrated as a function of Mg# were used in turn to evaluate the chemical composition of carbonates. Raman analysis of a suite of carbonates (siderite, sidero-magnesite, ankerite, and dolomite) of hydrothermal and sedimentary origin from the ca. 3.2 Ga old Barite Syncline, Barberton greenstone belt, South Africa, and from the ca. 3.5 Ga old Dresser Formation, Pilbara Craton, Western Australia, show good compositional agreement with electron microprobe analyses. These results indicate that Raman spectroscopy can provide direct information on the composition and structure of carbonates on planetary surfaces.

Patterns of metal distribution in hypersaline microbialites during early diagenesis: Implications for the fossil record

The use of metals as biosignatures in the fossil stromatolite record requires understanding of the processes controlling the initial metal(loid) incorporation and diagenetic preservation in living microbialites. Here, we report the distribution of metals and the organic fraction within the lithifying microbialite of the hypersaline Big Pond Lake (Bahamas). Using synchrotron-based X-ray microfluorescence, confocal, and biphoton microscopies at different scales (cm-μm) in combination with traditional geochemical analyses, we show that the initial cation sorption at the surface of an active microbialite is governed by passive binding to the organic matrix, resulting in a homogeneous metal distribution. During early diagenesis, the metabolic activity in deeper microbialite layers slows down and the distribution of the metals becomes progressively heterogeneous, resulting from remobilization and concentration as metal(loid)-enriched sulfides, which are aligned with the lamination of the microbialite. In addition, we were able to identify globules containing significant Mn, Cu, Zn, and As enrichments potentially produced through microbial activity. The similarity of the metal(loid) distributions observed in the Big Pond microbialite to those observed in the Archean stromatolites of Tumbiana provides the foundation for a conceptual model of the evolution of the metal distribution through initial growth, early diagenesis, and fossilization of a microbialite, with a potential application to the fossil record.

7.10 Chemical Characteristics of Sediments and Seawater

The transition from an anoxic to oxygenated atmosphere was arguably the most dramatic change in the history of the Earth. This “Great Oxidation Event” (Holland 2006) transformed the biogeochemical cycles of the elements by imposing an oxidative step in the cycles, creating strong redox gradients in the terrestrial and marine realms that energised microbial metabolism. Although much past research was focused on establishing when the rise of atmospheric oxygen took place, recognition that substantial mass-independent fraction (MIF) of the sulphur isotopes is restricted to the time interval before 2.45 Ga and requires an anoxic atmosphere (Farquhar et al. 2000, 2007; Mojzsis et al. 2003; Ono et al. 2003; Bekker et al. 2004) argues the atmosphere became permanently oxygenated at this time (Pavlov and Kasting 2002). A false-start to the modern aerobic biosphere and a “whiff” of atmospheric oxygen (Anbar et al. 2007) may have occurred in the latest Archaean, as reflected in a transient enrichment in the redox-sensitive element molybdenum in marine shales and a reduction in the extent of MIF precisely coincident with the peak in Mo and FeS2 enrichment (Kaufman et al. 2007). Geochemical proxies are imperfect, and an earlier (c. 3 Ga) appearance of atmospheric oxygen is possible (Ohmoto et al. 2006) but disputed (Farquhar et al. 2007; Buick 2008).Ancient rocks record the redox conditions of the oceanatmosphere system through the distribution of iron (Fe) between oxidised and reduced minerals, which can be formulated into a suite of Fe palaeoredox proxies. The balance between Fe and S in a given system reflects the variance in a range of highand low-temperature sources and sinks. Iron can be delivered by hydrothermal, diagenetic or clastic fluxes and can be buried and removed as Fe-oxide phases, Febearing carbonates such as siderite or ankerite, relatively unreactive silicate phases, which often pass through the system in detrital form, or as a constituent of pyrite (FeS2) using sulphide sourced by sulphate reduction. Sulphate is delivered to the ocean primarily from continental weathering, which requires that a surface oxidative cycle exists, and rates of sulphate delivery and Fe removal as pyrite should thus depend on ocean-atmosphere redox. Among other successes, the iron proxies discussed here have proven their value in studies of the 2.5 Ga Mt. McRae Formation and specifically in delineating subtle increases in atmospheric oxygen prior to the Great Oxidation Event, or ‘GOE’. (Anbar et al. 2007; Kaufman et al. 2007; Reinhard et al. 2009). These Fe proxies are our most effective inorganic proxy for ancient euxinia (anoxic and H2S-rich conditions) on the local scale and are an essential independent backdrop for meaningful application of Mo isotopes to address extents of euxinia on ocean scales (Arnold et al. 2004; Gordon et al. 2009). Thus, in addition to being informative on their own, Fe-based palaeoredox indicators are a crucial component of multi-proxy approaches for distinguishing among oxic, anoxic and Fe (II)-rich (ferruginous), and euxinic depositional conditions. The quantity and speciation of highly reactive iron (FeHR) in sediments and sedimentary rocks can provide crucial insight into the redox state of the local depositional environment. The total pool of FeHR consists of mineral phases that have the potential to react with dissolved H2S when exposed on short timescales (within the water column or during earliest diagenesis) plus Fe that has already reacted and is present as FeS2 (Raiswell and Canfield 1998). Such minerals include ferrous carbonates (siderite, FeCO3; ankerite, Ca(Fe,Mg,Mn)(CO3)2), crystalline ferric oxides (haematite, Fe2O3; goethite, FeOOH), and the mixed-valence Fe oxide magnetite (Fe3O4). These phases are separated by means of a well-calibrated sequential extraction scheme described in detail elsewhere (Poulton et al. 2004; Poulton and Canfield 2005; Reinhard et al. 2009). Briefly, ~100 mg of sample powder is first treated with a buffered sodium acetate solution for 48 h to mobilise ferrous carbonate phases. A split of the extract is removed for analysis, the sample is centrifuged, and the remaining supernatant is discarded. The sample is then treated with a sodium dithionite solution for 2 h to dissolve crystalline ferric oxides and processed as before. Finally, the sample is treated with an ammonium oxalate solution for 6 h to mobilise magnetite. All extractions are performed at room temperature in 15 mL centrifuge tubes under constant agitation. The sequential extracts are analysed on an Agilent 7500ce ICP-MS after 100-fold dilution in trace-metal grade HNO3 (2 %). Pyrite iron is calculated separately based on weight percent pyrite sulphur extracted during a 2-h, hot chromous chloride distillation followed by iodometric titration (Canfield et al. 1986), assuming a stoichiometry of FeS2. For measurement of total Fe (FeT), sample powders are ashed overnight at 450 C (in order to remove organic matter but preserve volatile metals, such as rhenium) and digested using sequential HNO3-HFHCl acid treatments (see, for example, Kendall et al. 2009). After digestion, samples are reconstituted in trace-metal grade HNO3 (2 %), diluted, and analysed by ICP-MS In modern oxic sediments deposited across a wide range of environments, FeHR comprises 6–38 % of total sedimentary Fe (i.e. FeHR/FeT 1⁄4 0.06–0.38), with an average value for FeHR/FeT of 0.26 0.08 defining the modern siliciclastic baseline (Raiswell and Canfield 1998). Enrichments in FeHR that are in excess of this detrital background ratio indicate a source of reactive Fe that is decoupled from the siliciclastic flux and thus reflect the transport, scavenging and enrichment (see below) of Fe within an anoxic basin (Canfield et al. 1996; Wijsman et al. 2001). In this context, ratios of FeHR/FeT exceeding the siliciclastic range point to anoxic deposition, and the ratio FePY/FeHR can then be used to establish whether the system was Fe(II)or H2S-buffered. An anoxic system with a relatively small amount of FeHR converted to pyrite indicates a depositional environment in which reactive Fe supply was greater than the titrating capacity of available H2S produced microbially by sulphate reduction, and thus no dissolved H2S was accumulating in pore fluids or the water column. Importantly, this is true even if microbial sulphate reduction and pyrite formation was occurring in the system (Canfield 1989) because the preponderance of Fe precludes the accumulation of free H2S. In contrast, if the vast majority of FeHR is present as pyrite in an anoxic system, euxinic depositional conditions are indicated – a consequence of the C.T. Reinhard (*) Department of Earth Sciences, University of California, Riverside, CA 92521, USA 10 7.10 Chemical Characteristics of Sediments and Seawater 1483

erratum: Reassessing the evidence for the earliest traces of life

This corrects the article DOI: nature00934

Proposed early signs of life not set in stone

Efforts to find early traces of life on Earth often focus on structures in ancient rocks, called stromatolites, that formed by microbial activity. One of the oldest proposed stromatolite discoveries has now been questioned.Whether microbes generated structures in ancient rocks has been questioned.Efforts to find early traces of life on Earth often focus on structures in ancient rocks, called stromatolites, that formed by microbial activity. One of the oldest proposed stromatolite discoveries has now been questioned. Whether microbes generated structures in ancient rocks has been questioned.