Search

The evolution of the global carbon and silicon cycles is thought to have contributed to the long-term stability of Earth's climate1-3. Many questions remain, however, regarding the feedback mechanisms at play, and there are limited quantitative constraints on the sources and sinks of these elements in Earth's surface environments4-12. Here we argue that the lithium-isotope record can be used to track the processes controlling the long-term carbon and silicon cycles. By analysing more than 600 shallow-water marine carbonate samples from more than 100 stratigraphic units, we construct a new carbonate-based lithium-isotope record spanning the past 3 billion years. The data suggest an increase in the carbonate lithium-isotope values over time, which we propose was driven by long-term changes in the lithium-isotopic conditions of sea water, rather than by changes in the sedimentary alterations of older samples. Using a mass-balance modelling approach, we propose that the observed trend in lithium-isotope values reflects a transition from Precambrian carbon and silicon cycles to those characteristic of the modern. We speculate that this transition was linked to a gradual shift to a biologically controlled marine silicon cycle and the evolutionary radiation of land plants13,14.

The Earth’s inner core grows by the freezing of liquid iron at its surface. The point in history at which this process initiated marks a step-change in the thermal evolution of the planet. Recent computational and experimental studies1,2,3,4,5 have presented radically differing estimates of the thermal conductivity of the Earth’s core, resulting in estimates of the timing of inner-core nucleation ranging from less than half a billion to nearly two billion years ago. Recent inner-core nucleation (high thermal conductivity) requires high outer-core temperatures in the early Earth that complicate models of thermal evolution. The nucleation of the core leads to a different convective regime6 and potentially different magnetic field structures that produce an observable signal in the palaeomagnetic record and allow the date of inner-core nucleation to be estimated directly. Previous studies searching for this signature have been hampered by the paucity of palaeomagnetic intensity measurements, by the lack of an effective means of assessing their reliability, and by shorter-timescale geomagnetic variations. Here we examine results from an expanded Precambrian database of palaeomagnetic intensity measurements7 selected using a new set of reliability criteria8. Our analysis provides intensity-based support for the dominant dipolarity of the time-averaged Precambrian field, a crucial requirement for palaeomagnetic reconstructions of continents. We also present firm evidence for the existence of very long-term variations in geomagnetic strength. The most prominent and robust transition in the record is an increase in both average field strength and variability that is observed to occur between a billion and 1.5 billion years ago. This observation is most readily explained by the nucleation of the inner core occurring during this interval9; the timing would tend to favour a modest value of core thermal conductivity and supports a simple thermal evolution model for the Earth.

Data are presented that support the idea of an oxygenation event in the immediate aftermath of the Marinoan glaciation, pre-dating previous estimates for post-Marinoan oxygenation by more than 50 million years. Macroscopic metazoans first appeared in the fossil record shortly after the termination of the late Cryogenian (Marinoan) glaciation about 635 million years ago. It has been suggested that an oxygenation event at about this time was the driving factor behind the rise of the metazoans, but current estimates suggest that oxygenation occurred between 580 million and 550 million years ago, well after initial animal diversification. New geochemical data from early Ediacaran organic-rich black shales of the basal Doushantuo Formation in South China now suggest that the oxidation event occurred more than 50 million years earlier, in the immediate aftermath of the Marinoan glaciation. The data provide evidence for a significant postglacial oxygenation and support a link between the most severe glaciations in Earth's history, the oxygenation of Earth's surface and the earliest emergence of complex animals. Metazoans are likely to have their roots in the Cryogenian period1,2,3, but there is a marked increase in the appearance of novel animal and algae fossils shortly after the termination of the late Cryogenian (Marinoan) glaciation about 635 million years ago4,5,6. It has been suggested that an oxygenation event in the wake of the severe Marinoan glaciation was the driving factor behind this early diversification of metazoans and the shift in ecosystem complexity7,8. But there is little evidence for an increase in oceanic or atmospheric oxygen following the Marinoan glaciation, or for a direct link between early animal evolution and redox conditions in general9. Models linking trends in early biological evolution to shifts in Earth system processes thus remain controversial10. Here we report geochemical data from early Ediacaran organic-rich black shales (∼635–630 million years old) of the basal Doushantuo Formation in South China. High enrichments of molybdenum and vanadium and low pyrite sulphur isotope values (Δ34S values ≥65 per mil) in these shales record expansion of the oceanic inventory of redox-sensitive metals and the growth of the marine sulphate reservoir in response to a widely oxygenated ocean. The data provide evidence for an early Ediacaran oxygenation event, which pre-dates the previous estimates for post-Marinoan oxygenation11,12,13 by more than 50 million years. Our findings seem to support a link between the most severe glaciations in Earth’s history, the oxygenation of the Earth’s surface environments, and the earliest diversification of animals.

The Late Heavy Bombardment lasted much longer than previously thought, up to 1.7 billion years ago on Earth, with impacts on the Moon and Earth coming mostly from the E-belt-survivor Hungaria asteroids. The Late Heavy Bombardment was a period of time, generally put at about 4.1 billion to 3.8 billion years ago, when the inner planets of the Solar System were subjected to a high-frequency barrage of asteroids. This left its mark on the Moon, but on Earth the craters quickly disappeared owing to tectonic processes and erosion. In the first of two papers on the bombardment, Brandon Johnson and Jay Melosh determine the properties of the asteroids by looking at spherule beds: layers of debris ejected during the impacts. The thickness of spherule layers is expected to vary according to the size of the impactor and the speed at which it hit Earth. This historical record of impacts indicates that the number of projectiles colliding with Earth was substantially higher 3.5 billion years ago than it is today, with a gradual decline in the number of strikes after the Late Heavy Bombardment. Bottke et al. modelled the evolution of an asteroid belt that extended farther towards Mars than the present one. They find that most of the impactors traced by the spherule beds probably originated in this 'E-belt', which was disrupted during migrations of some of the giant planets. The barrage of comets and asteroids that produced many young lunar basins (craters over 300 kilometres in diameter) has frequently been called the Late Heavy Bombardment1 (LHB). Many assume the LHB ended about 3.7 to 3.8 billion years (Gyr) ago with the formation of Orientale basin2,3. Evidence for LHB-sized blasts on Earth, however, extend into the Archaean and early Proterozoic eons, in the form of impact spherule beds: globally distributed ejecta layers created by Chicxulub-sized or larger cratering events4. At least seven spherule beds have been found that formed between 3.23 and 3.47 Gyr ago, four between 2.49 and 2.63 Gyr ago, and one between 1.7 and 2.1 Gyr ago5,6,7,8,9. Here we report that the LHB lasted much longer than previously thought, with most late impactors coming from the E belt, an extended and now largely extinct portion of the asteroid belt between 1.7 and 2.1 astronomical units from Earth. This region was destabilized by late giant planet migration10,11,12,13. E-belt survivors now make up the high-inclination Hungaria asteroids14,15. Scaling from the observed Hungaria asteroids, we find that E-belt projectiles made about ten lunar basins between 3.7 and 4.1 Gyr ago. They also produced about 15 terrestrial basins between 2.5 and 3.7 Gyr ago, as well as around 70 and four Chicxulub-sized or larger craters on the Earth and Moon, respectively, between 1.7 and 3.7 Gyr ago. These rates reproduce impact spherule bed and lunar crater constraints.

To determine the geometric pattern of the supercontinent cycle, a new ‘orthoversion’ model is suggested that matches the geologic evidence better than the traditional ‘introversion’ and ‘extroversion’ models, enabling the calculation of absolute palaeolongitude for the early Earth. Geological models have predicted that the next supercontinent to form — to be called Amasia — will amalgamate either where Pangaea rifted apart, or on the opposite side of the world. An alternative model proposed here places the next supercontinent at right angles to these two positions. This new 'orthoversion' model assumes a 90° shift between supercontinents because the new aggregations would coalesce within the great circle of subduction encircling its relict predecessor. Traditional models of the supercontinent cycle predict that the next supercontinent—‘Amasia’—will form either where Pangaea rifted (the ‘introversion’1 model) or on the opposite side of the world (the ‘extroversion’2,3,4 models). Here, by contrast, we develop an ‘orthoversion’5 model whereby a succeeding supercontinent forms 90° away, within the great circle of subduction encircling its relict predecessor. A supercontinent aggregates over a mantle downwelling but then influences global-scale mantle convection to create an upwelling under the landmass6. We calculate the minimum moment of inertia about which oscillatory true polar wander occurs owing to the prolate shape of the non-hydrostatic Earth5,7. By fitting great circles to each supercontinent’s true polar wander legacy, we determine that the arc distances between successive supercontinent centres (the axes of the respective minimum moments of inertia) are 88° for Nuna to Rodinia and 87° for Rodinia to Pangaea—as predicted by the orthoversion model. Supercontinent centres can be located back into Precambrian time, providing fixed points for the calculation of absolute palaeolongitude over billion-year timescales. Palaeogeographic reconstructions additionally constrained in palaeolongitude will provide increasingly accurate estimates of ancient plate motions and palaeobiogeographic affinities.

The Gamburtsev Subglacial Mountains are the least understood tectonic feature on Earth, because they are completely hidden beneath the East Antarctic Ice Sheet. Their high elevation and youthful Alpine topography, combined with their location on the East Antarctic craton, creates a paradox that has puzzled researchers since the mountains were discovered in 19581. The preservation of Alpine topography in the Gamburtsevs2 may reflect extremely low long-term erosion rates beneath the ice sheet3, but the mountains’ origin remains problematic. Here we present the first comprehensive view of the crustal architecture and uplift mechanisms for the Gamburtsevs, derived from radar, gravity and magnetic data. The geophysical data define a 2,500-km-long rift system in East Antarctica surrounding the Gamburtsevs, and a thick crustal root4 beneath the range. We propose that the root formed during the Proterozoic assembly of interior East Antarctica (possibly about 1 Gyr ago), was preserved as in some old orogens5, 6 and was rejuvenated during much later Permian (roughly 250 Myr ago) and Cretaceous (roughly 100 Myr ago) rifting. Much like East Africa7, the interior of East Antarctica is a mosaic of Precambrian provinces affected by rifting processes. Our models show that the combination of rift-flank uplift, root buoyancy and the isostatic response to fluvial and glacial erosion explains the high elevation and relief of the Gamburtsevs. The evolution of the Gamburtsevs demonstrates that rifting and preserved orogenic roots can produce broad regions of high topography in continental interiors without significantly modifying the underlying Precambrian lithosphere.

Geochemical data from ancient sedimentary successions provide evidence for the progressive evolution of Earth’s atmosphere and oceans1, 2, 3, 4, 5, 6, 7. Key stages in increasing oxygenation are postulated for the Palaeoproterozoic era (~2.3 billion years ago, Gyr ago) and the late Proterozoic eon (about 0.8 Gyr ago), with the latter implicated in the subsequent metazoan evolutionary expansion8. In support of this rise in oxygen concentrations, a large database1, 2, 3, 9 shows a marked change in the bacterially mediated fractionation of seawater sulphate to sulphide of Δ34S

It had been thought that ocean temperatures during the early Archaean (around 3.5 billion years ago) were between 55 and 85 °C. But a recent study made a case for Archaean ocean temperatures no greater than 40 °C. The well-preserved rocks of the Barberton Greenstone Belt in southern Africa contain a geochemical record of the early life and ocean chemistry of 3.2 to 3.5 billion years ago. A new study of oxygen isotope compositions of phosphates in Barberton sediments provides support for the cooler, temperate Archaean ocean around 26 to 35 °C. The findings suggest a well-developed phosphorus cycle and evolved biological activity on the Archaean Earth. It has been thought that ocean temperatures during the early Palaeoarchaean era (around 3.5 billion years ago) were 55–85 °C. But a recent study indicated that the temperatures might be no higher than 40 °C. Here, studies are reported of the oxygen isotope compositions of phosphates in sediments from the 3.2–3.5-billion-year-old Barberton Greenstone Belt in South Africa. The findings indicate a well-developed phosphorus cycle and evolved biological activity in an Archaean ocean with temperatures of 26–35 °C. Oxygen and silicon isotope compositions of cherts1,2,3 and studies of protein evolution4 have been interpreted to reflect ocean temperatures of 55–85 °C during the early Palaeoarchaean era (∼3.5 billion years ago). A recent study combining oxygen and hydrogen isotope compositions of cherts, however, makes a case for Archaean ocean temperatures being no greater than 40 °C (ref. 5). Ocean temperature can also be assessed using the oxygen isotope composition of phosphate. Recent studies show that 18O:16O ratios of dissolved inorganic phosphate (δ18OP) reflect ambient seawater temperature as well as biological processing that dominates marine phosphorus cycling at low temperature6,7. All forms of life require and concentrate phosphorus, and as a result of biological processing, modern marine phosphates have δ18OP values typically between 19–26‰ (VSMOW)7,8, highly evolved from presumed source values of ∼6–8‰ that are characteristic of apatite in igneous rocks9,10 and meteorites11. Here we report oxygen isotope compositions of phosphates in sediments from the 3.2–3.5-billion-year-old Barberton Greenstone Belt in South Africa. We find that δ18OP values range from 9.3‰ to 19.9‰ and include the highest values reported for Archaean rocks. The temperatures calculated from our highest δ18OP values and assuming equilibrium with sea water with δ18O = 0‰ (ref. 12) range from 26 °C to 35 °C. The higher δ18OP values are similar to those of modern marine phosphate and suggest a well-developed phosphorus cycle and evolved biologic activity on the Archaean Earth.

Dozens of studies in the past decade have reported carbon isotope variations in Neoproterozoic carbonate rocks and linked them to perturbations of the global carbon cycle. Paul Knauth and Martin Kennedy have taken a sideways look at the data by concentrating on the oxygen isotope measurements (for over 20,000 samples) that are necessarily obtained as part of the carbon isotope analysis but are often overlooked. They arrive at the striking conclusion that the combined oxygen and carbon isotope systematics are identical to those of well-understood Phanerozoic examples that lithified in coastal pore fluids receiving a groundwater influx of photosynthetic carbon from terrestrial phytomass. Rather than being perturbations to the carbon cycle, widely reported decreases in 13C/12C in Neoproterozoic carbonates are more easily interpreted by analogy to the Phanerozoic examples. And that could suggest a 'greening' of the Earth under a ground-hugging mat of photosynthetic algae, mosses and fungi in the late Precambrian. Such an event, producing oxygen and phytomass, could even be indirectly responsible for the critical transition from the essentially microbial world of the Precambrian to the metazoan world of the Cambrian. The low 13C/12C ratio in some Neoproterozoic carbonates is considered to be evidence of carbon cycle perturbations unique to the Precambrian. Here, all published oxygen and carbon isotope data for Neoproterozoic marine carbonates are compiled. The combined isotope systematics are found to be identical to those of well-understood Phanerozoic examples, suggesting an influx of photosynthetic carbon rather than perturbations to the carbon cycle — and implying an explosion of photosynthesizing communities on late Precambrian land surfaces. Many aspects of the carbon cycle can be assessed from temporal changes in the 13C/12C ratio of oceanic bicarbonate. 13C/12C can temporarily rise when large amounts of 13C-depleted photosynthetic organic matter are buried at enhanced rates1, and can decrease if phytomass is rapidly oxidized2 or if low 13C is rapidly released from methane clathrates3. Assuming that variations of the marine 13C/12C ratio are directly recorded in carbonate rocks, thousands of carbon isotope analyses of late Precambrian examples have been published to correlate these otherwise undatable strata and to document perturbations to the carbon cycle just before the great expansion of metazoan life. Low 13C/12C in some Neoproterozoic carbonates is considered evidence of carbon cycle perturbations unique to the Precambrian. These include complete oxidation of all organic matter in the ocean2 and complete productivity collapse such that low-13C/12C hydrothermal CO2 becomes the main input of carbon4. Here we compile all published oxygen and carbon isotope data for Neoproterozoic marine carbonates, and consider them in terms of processes known to alter the isotopic composition during transformation of the initial precipitate into limestone/dolostone. We show that the combined oxygen and carbon isotope systematics are identical to those of well-understood Phanerozoic examples that lithified in coastal pore fluids, receiving a large groundwater influx of photosynthetic carbon from terrestrial phytomass. Rather than being perturbations to the carbon cycle, widely reported decreases in 13C/12C in Neoproterozoic carbonates are more easily interpreted in the same way as is done for Phanerozoic examples. This influx of terrestrial carbon is not apparent in carbonates older than ∼850 Myr, so we infer an explosion of photosynthesizing communities on late Precambrian land surfaces. As a result, biotically enhanced weathering generated carbon-bearing soils on a large scale and their detrital sedimentation sequestered carbon5. This facilitated a rise in O2 necessary for the expansion of multicellular life.

Komatiites are ancient volcanic rocks, mostly over 2.7 billion years old (from the Archaean era), that formed through high degrees of partial melting of the mantle and therefore provide reliable information on bulk mantle compositions1. In particular, the platinum group element (PGE) contents of komatiites provide a unique source of information on core formation, mantle differentiation and possibly core–mantle interaction2, 3, 4, 5, 6, 7, 8. Most of the available PGE data on komatiites are from late Archaean (~2.7–2.9 Gyr old) or early Proterozoic (2.0–2.5 Gyr old) samples. Here we show that most early Archaean (3.5–3.2 Gyr old) komatiites from the Barberton greenstone belt of South Africa and the Pilbara craton of Western Australia are depleted in PGE relative to late Archaean and younger komatiites. Early Archaean komatiites record a signal of PGE depletion in the lower mantle, resulting from core formation. This signal diminishes with time owing to progressive mixing-in to the deep mantle of PGE-enriched cosmic material that the Earth accreted as the ‘late veneer’ during the Early Archaean (4.5–3.8 Gyr ago) meteorite bombardment.

The isotopic decay of samarium-146 produces neodymium-142, with a half-life of 103 million years. Major geochemical differentiation events such as the formation and crystallization of a magma ocean or the extraction of the first crust early in planetary history can fractionate samarium (Sm) from neodymium (Nd) and produce reservoirs having contrasting 142Nd/144Nd compositions that provide a sensitive monitor of the main silicate differentiation events that took place in the early Earth. Upadhyay et al. have measured low 142Nd/144Nd ratios in ∼1.48 billion-year-old rocks from the Khariar complex in southeastern India. This indicates that enriched Hadean reservoirs may be hidden within the roots of old cratons. The isotope 146Sm decays to 142Nd with a half-life of 103 million years, and therefore variations in the 142Nd/144Nd values of rocks resulting from Sm–Nd fractionation provide a sensitive monitor of the main silicate differentiation events that took place in early Earth. The measurement of low 142Nd/144Nd ratios in approximately 1.48 billion-year-old rocks from the Khariar complex in southeastern India now indicates that enriched Hadean reservoirs may be hidden within the roots of old cratons. The isotope 146Sm undergoes α-decay to 142Nd, with a half-life of 103 million years. Measurable variations in the 142Nd/144Nd values of rocks resulting from Sm–Nd fractionation could therefore only have been produced within about 400 million years of the Solar System’s formation (that is, when 146Sm was extant). The 142Nd/144Nd compositions of terrestrial rocks1,2,3,4,5,6,7 are accordingly a sensitive monitor of the main silicate differentiation events that took place in the early Earth. High 142Nd/144Nd values measured in some Archaean rocks from Greenland1,2,3,4,5 hint at the existence of an early incompatible-element-depleted mantle. Here we present measurements of low 142Nd/144Nd values in 1.48-gigayear-(Gyr)-old lithospheric mantle-derived alkaline rocks from the Khariar nepheline syenite complex in southeastern India8. These data suggest that a reservoir that was relatively enriched in incompatible elements formed at least 4.2 Gyr ago and traces of its isotopic signature persisted within the lithospheric root of the Bastar craton until at least 1.48 Gyr ago. These low 142Nd/144Nd compositions may represent a diluted signature of a Hadean (4 to 4.57 Gyr ago) enriched reservoir that is characterized by even lower values. That no evidence of the early depleted mantle has been observed in rocks younger than 3.6 Gyr (refs 3, 4, 7) implies that such domains had effectively mixed back into the convecting mantle by then. In contrast, some early enriched components apparently escaped this fate. Thus, the mantle sampled by magmatism since 3.6 Gyr ago may be biased towards a depleted composition that would be balanced by relatively more enriched reservoirs that are ‘hidden’ in Hadean crust6, the D′′ layer9,10,11 of the lowermost mantle or, as we propose here, also within the roots of old cratons.

Detrital zircons more than 4 billion years old from the Jack Hills metasedimentary belt, Yilgarn craton, Western Australia, are the oldest identified fragments of the Earth's crust and are unique in preserving information on the earliest evolution of the Earth. Inclusions of quartz, K-feldspar and monazite in the zircons, in combination with an enrichment of light rare-earth elements and an estimated low zircon crystallization temperature, have previously been used as evidence for early recycling of continental crust, leading to the production of granitic melts in the Hadean era. Here we present the discovery of microdiamond inclusions in Jack Hills zircons with an age range from 3,058 +/- 7 to 4,252 +/- 7 million years. These include the oldest known diamonds found in terrestrial rocks, and introduce a new dimension to the debate on the origin of these zircons and the evolution of the early Earth. The spread of ages indicates that either conditions required for diamond formation were repeated several times during early Earth history or that there was significant recycling of ancient diamond. Mineralogical features of the Jack Hills diamonds-such as their occurrence in zircon, their association with graphite and their Raman spectroscopic characteristics-resemble those of diamonds formed during ultrahigh-pressure metamorphism and, unless conditions on the early Earth were unique, imply a relatively thick continental lithosphere and crust-mantle interaction at least 4,250 million years ago.

Odontogriphus omalus was originally described as a problematic non-biomineralized lophophorate organism. Here we re-interpret Odontogriphus based on 189 new specimens including numerous exceptionally well preserved individuals from the Burgess Shale collections of the Royal Ontario Museum. This additional material provides compelling evidence that the feeding apparatus in Odontogriphus is a radula of molluscan architecture comprising two primary bipartite tooth rows attached to a radular membrane and showing replacement by posterior addition. Further characters supporting molluscan affinity include a broad foot bordered by numerous ctenidia located in a mantle groove and a stiffened cuticular dorsum. Odontogriphus has a radula similar to Wiwaxia corrugata but lacks a scleritome. We interpret these animals to be members of an early stem-group mollusc lineage that probably originated in the Neoproterozoic Ediacaran Period, providing support for the retention of a biomat-based grazing community from the late Precambrian Period until at least the Middle Cambrian.

Methanogenic microbes may be one of the most primitive organisms, although it is uncertain when methanogens first appeared on Earth. During the Archaean era (before 2.5 Gyr ago), methanogens may have been important in regulating climate, because they could have provided sufficient amounts of the greenhouse gas methane to mitigate a severely frozen condition that could have resulted from lower solar luminosity during these times. Nevertheless, no direct geological evidence has hitherto been available in support of the existence of methanogens in the Archaean period, although circumstantial evidence is available in the form of approximately 2.8-Gyr-old carbon-isotope-depleted kerogen. Here we report crushing extraction and carbon isotope analysis of methane-bearing fluid inclusions in approximately 3.5-Gyr-old hydrothermal precipitates from Pilbara craton, Australia. Our results indicate that the extracted fluids contain microbial methane with carbon isotopic compositions of less than -56 per thousand included within original precipitates. This provides the oldest evidence of methanogen (> 3.46 Gyr ago), pre-dating previous geochemical evidence by about 700 million years.

The Proterozoic aeon (2.5 to 0.54 billion years (Gyr) ago) marks the time between the largely anoxic world of the Archean (> 2.5 Gyr ago)1 and the dominantly oxic world of the Phanerozoic (< 0.54 Gyr ago). The course of ocean chemistry through the Proterozoic has traditionally been explained by progressive oxygenation of the deep ocean in response to an increase in atmospheric oxygen around 2.3 Gyr ago. This postulated rise in the oxygen content of the ocean is in turn thought to have led to the oxidation of dissolved iron, Fe(II), thus ending the deposition of banded iron formations (BIF) around 1.8 Gyr ago1,2. An alternative interpretation suggests that the increasing atmospheric oxygen levels enhanced sulphide weathering on land and the flux of sulphate to the oceans. This increased rates of sulphate reduction, resulting in Fe(II) removal in the form of pyrite as the oceans became sulphidic3. Here we investigate sediments from the ∼1.8-Gyr-old Animikie group, Canada, which were deposited during the final stages of the main global period of BIF deposition. This allows us to evaluate the two competing hypotheses for the termination of BIF deposition. We use iron–sulphur–carbon (Fe–S–C) systematics to demonstrate continued ocean anoxia after the final global deposition of BIF and show that a transition to sulphidic bottom waters was ultimately responsible for the termination of BIF deposition. Sulphidic conditions may have persisted until a second major rise in oxygen between 0.8 to 0.58 Gyr ago4,5, possibly reducing global rates of primary production and arresting the pace of algal evolution6.

Geological and palaeomagnetic studies indicate that ice sheets may have reached the Equator at the end of the Proterozoic eon, 800 to 550 million years ago, leading to the suggestion of a fully ice-covered 'snowball Earth'. Climate model simulations indicate that such a snowball state for the Earth depends on anomalously low atmospheric carbon dioxide concentrations, in addition to the Sun being 6 per cent fainter than it is today. However, the mechanisms producing such low carbon dioxide concentrations remain controversial. Here we assess the effect of the palaeogeographic changes preceding the Sturtian glacial period, 750 million years ago, on the long-term evolution of atmospheric carbon dioxide levels using the coupled climate-geochemical model GEOCLIM. In our simulation, the continental break-up of Rodinia leads to an increase in runoff and hence consumption of carbon dioxide through continental weathering that decreases atmospheric carbon dioxide concentrations by 1,320 p.p.m. This indicates that tectonic changes could have triggered a progressive transition from a 'greenhouse' to an 'icehouse' climate during the Neoproterozoic era. When we combine these results with the concomitant weathering effect of the voluminous basaltic traps erupted throughout the break-up of Rodinia, our simulation results in a snowball glaciation.

Unlike the familiar Phanerozoic history of life, evolution during the earlier and much longer Precambrian segment of geological time centred on prokaryotic microbes1. Because such microorganisms are minute, are preserved incompletely in geological materials, and have simple morphologies that can be mimicked by nonbiological mineral microstructures, discriminating between true microbial fossils and microscopic pseudofossil ‘lookalikes’ can be difficult2,3. Thus, valid identification of fossil microbes, which is essential to understanding the prokaryote-dominated, Precambrian 85% of life's history, can require more than traditional palaeontology that is focused on morphology. By combining optically discernible morphology with analyses of chemical composition, laser–Raman spectroscopic imagery of individual microscopic fossils provides a means by which to address this need. Here we apply this technique to exceptionally ancient fossil microbe-like objects, including the oldest such specimens reported from the geological record, and show that the results obtained substantiate the biological origin of the earliest cellular fossils known.

Sulphate-reducing microbes affect the modern sulphur cycle, and may be quite ancient, though when they evolved is uncertain. These organisms produce sulphide while oxidizing organic matter or hydrogen with sulphate. At sulphate concentrations greater than 1 mM, the sulphides are isotopically fractionated (depleted in 34S) by 10-40/1000 compared to the sulphate, with fractionations decreasing to near 0/1000 at lower concentrations. The isotope record of sedimentary sulphides shows large fractionations relative to seawater sulphate by 2.7 Gyr ago, indicating microbial sulphate reduction. In older rocks, however, much smaller fractionations are of equivocal origin, possibly biogenic but also possibly volcanogenic. Here we report microscopic sulphides in approximately 3.47-Gyr-old barites from North Pole, Australia, with maximum fractionations of 21.1/1000, about a mean of 11.6/1000, clearly indicating microbial sulphate reduction. Our results extend the geological record of microbial sulphate reduction back more than 750 million years, and represent direct evidence of an early specific metabolic pathway--allowing time calibration of a deep node on the tree of life.

Phosphorus is generally thought to be a limiting nutrient of primary productivity in the oceans, and is important in regulating the redox state of the ocean–atmosphere system. Planavsky et al. use the ratio of phosphorus to iron in iron-oxide-rich sedimentary rocks through time to evaluate the evolution of the marine phosphate reservoir. They find relatively constant phosphate concentrations during the past 542 million years of Earth's history. The data are also indicative of high dissolved phosphate concentrations in the aftermath of the 'snowball Earth' glaciations around 700 million years ago, which could have led to high rates of primary productivity, organic carbon burial and an increase in atmospheric oxygen levels, paving the way for the rise of metazoan life. Phosphorus is a biolimiting nutrient that is important in regulating the redox state of the ocean–atmosphere system. Here, the ratio of phosphorus to iron in iron-oxide-rich sedimentary rocks through time has been used to evaluate the evolution of the marine phosphate reservoir. Phosphate concentrations have been relatively constant over the past 542 million years of Earth's history, but were high in the aftermath of the 'snowball Earth' glaciations some 750 to 635 million years ago, with implications for the rise of metazoan life. Phosphorus is a biolimiting nutrient that has an important role in regulating the burial of organic matter and the redox state of the ocean–atmosphere system1. The ratio of phosphorus to iron in iron-oxide-rich sedimentary rocks can be used to track dissolved phosphate concentrations if the dissolved silica concentration of sea water is estimated2,3,4,5. Here we present iron and phosphorus concentration ratios from distal hydrothermal sediments and iron formations through time to study the evolution of the marine phosphate reservoir. The data suggest that phosphate concentrations have been relatively constant over the Phanerozoic eon, the past 542 million years (Myr) of Earth’s history. In contrast, phosphate concentrations seem to have been elevated in Precambrian oceans. Specifically, there is a peak in phosphorus-to-iron ratios in Neoproterozoic iron formations dating from ∼750 to ∼635 Myr ago, indicating unusually high dissolved phosphate concentrations in the aftermath of widespread, low-latitude ‘snowball Earth’ glaciations. An enhanced postglacial phosphate flux would have caused high rates of primary productivity and organic carbon burial and a transition to more oxidizing conditions in the ocean and atmosphere. The snowball Earth glaciations and Neoproterozoic oxidation are both suggested as triggers for the evolution and radiation of metazoans6,7. We propose that these two factors are intimately linked; a glacially induced nutrient surplus could have led to an increase in atmospheric oxygen, paving the way for the rise of metazoan life.

There was a significant oxidation of the Earth's surface around 2 billion years ago (2 Gyr)1,2,3,4. Direct evidence for this oxidation comes, mostly, from geological records of the redox-sensitive elements Fe and U reflecting the conditions prevailing during weathering1,2,3. The oxidation event was probably driven by an increased input of oxygen to the atmosphere arising from an increased sedimentary burial of organic matter between 2.3 and 2.0 Gyr5. This episode was postdated by the final large precipitation of banded iron formations around 1.8 Gyr1,2. It is generally believed that banded iron formations precipitated from an ocean whose bottom waters contained significant concentrations of dissolved ferrous iron, and that this sedimentation process terminated when aerobic bottom waters developed, oxidizing the iron and thus removing it from solution1,2. In contrast, I argue here that anoxic bottom waters probably persisted until well after the deposition of banded iron formations ceased; I also propose that sulphide, rather than oxygen, was responsible for removing iron from deep ocean water. The sulphur-isotope record supports this hypothesis as it indicates increasing concentrations of oceanic sulphate, starting around 2.3 Gyr6, leading to increasing rates of sulphide production by sulphate reduction. The increase in sulphide production became sufficient, around 1.8 Gyr, to precipitate the total flux of iron into the oceans. I suggest that aerobic deep-ocean waters did not develop until the Neoproterozoic era (1.0 to ∼0.54 Gyr), in association with a second large oxidation of the Earth's surface. This new model is consistent with the emerging view of Precambrian sulphur geochemistry and the chemical events leading to the evolution of animals, and it is fully testable by detailed geochemical analyses of preserved deep-water marine sediments.

Oil is generally thought to be geologically young, as it is thermodynamically unstable when subjected to elevated temperatures over long periods in open systems1,2. Indeed, almost all petroleum production comes from rocks younger than 400 million years (ref. 3). Although the oldest known oil occurs in rocks 1,650 Myr old4, suitable source rocks were abundant in older geological successions5 and circumstantial evidence suggests that some of these generated hydrocarbons early in their history6. Here, we report the discovery of oil preserved in fluid inclusions in sandstones dating back ∼3,000 Myr. Most inclusions lie within healed microfractures confined to individual detrital quartz grains, indicating that their oil was emplaced before Archaean or Palaeoproterozoic metamorphism sealed all voids and thus came from older (in some cases Archaean) sources. The fluid inclusions apparently acted as inert pressure vessels that protected the oil from subsequent degradation by circulating fluids and mineral catalysts. Because of its great age, this oil can potentially yield valuable information about the size and diversity of the early biosphere, particularly if it contains molecular fossils (biomarkers) of the primordial organisms from which it was derived.

Phosphorites of the late Neoproterozoic (570 ± 20 Myr BP) Doushantuo Formation, southern China, preserve an exceptional record of multicellular life from just before the Ediacaran radiation of macroscopic animals. Abundant thalli with cellular structures preserved in three-dimensional detail show that latest-Proterozoic algae already possessed many of the anatomical and reproductive features seen in the modern marine flora. Embryos preserved in early cleavage stages indicate that the divergence of lineages leading to bilaterians may have occurred well before their macroscopic traces or body fossils appear in the geological record. Discovery of these fossils shows that the early evolution of multicellular organisms is amenable to direct palaeontological inquiry.

The fossil Kimberella quadrata was originally described from late Precambrian rocks of southern Australia1. Reconstructed as a jellyfish2, it was later assigned to the cubozoans (‘box jellies’), and has been cited as a clear instance of an extant animal lineage present before the Cambrian3,4,5,6,7. Until recently, Kimberella was known only from Australia, with the exception of some questionable north Indian specimens8. We now have over thirty-five specimens of this fossil from the Winter Coast of the White Sea in northern Russia. Our study of the new material does not support a cnidarian affinity. We reconstruct Kimberella as a bilaterally symmetrical, benthic animal with a non-mineralized, univalved shell, resembling a mollusc in many respects. This is important evidence for the existence of large triploblastic metazoans in the Precambrian and indicates that the origin of the higher groups of protostomes lies well back in the Precambrian.

One of the most fundamental enigmas of the Earth's palaeoclimate concerns the temporal and spatial distributions of Precambrian glaciations. Through four billion years of Precambrian history, unequivocally glacial deposits have been found only in the Palaeoproterozoic and Neoproterozoic record. Nonetheless, some of these deposits are closely associated with tropical— rather than just polar—palaeolatitudinal indicators such as carbonate rocks, red beds, and evaporites. These observations are quantitatively supported by palaeomagnetic results indicating a ~5° latitude for Neoproterozoic glaciogenic rocks in Australia. Similarly reliable palaeolatitudes for the older, Palaeoproterozoic glaciogenic rocks have not yet been obtained, as such deposits commonly suffer from poor preservation and secondary magnetic overprinting. The Archaean–Palaeoproterozoic 'Transvaal Supergroup' on the Kaapvaal craton in South Africa is, however, exceptionally well preserved, and is thus amenable to the palaeomagnetic determination of depositional palaeolatitudes. Within this supergroup the ~2.2 billion-year old Ongeluk lavas are a regionally extensive, largely undeformed and unmetamorphosed, extrusive volcanic succession, which conformably overlies glaciogenic deposits (the Makganyene diamictite). Here we report a palaeomagnetic estimate of 11 ± 5° depositional latitude for the lavas, and hence for the underlying contemporaneous glacial rocks. The palaeoclimate enigma is thus deepened; a largely ice-free Precambrian world was apparently punctuated by two long ice ages, both yielding glacial deposits well within tropical latitudes.

LATEST Archaean and earliest Palaeoproterozoic times (from 2.6 to 2.2 billion years ago) have generally been viewed as a largely quiescent period of Earth history; the geological record indicates the very slow deposition of pelagic and chemical sediments1,2, and bears only a limited record of magmatic and tectonic activity3–5. Such quiescence is consistent with the contention that the Earth's main banded iron formations (BIFs)—finely laminated chemical sedimentary rocks, rich in iron oxide—formed slowly as oxygen abundances in the oceans gradually increased, thus reducing the capacity of sea water to retain dissolved iron6–10. Here we show that a large igneous province, comprising >30,000km3 of dolerite, basalt and rhyolite, accompanied deposition of a Hamersley Province BIF 2,449 ±3 million years ago. This observation indicates that Hamersley BIFs formed during a major tectono-magmatic event and were deposited very much faster than previously thought, at similar rates to (or faster than) modern pelagic sediments. Thus the largest Palaeoproterozoic BIFs, rather than simply reflecting a gradual increase in the oxygen content of the oceans during a period of tectonic quiescence, are more likely to have formed as a result of an increased supply of suboxic iron- and silica-rich sea water upwelling onto continental shelves during a pulse (or pulses) of increased submarine magmatic and hydrothermal activity.

Fluids trapped as inclusions within minerals can be billions of years old and preserve a record of the fluid chemistry and environment at the time of mineralization. Aqueous fluids that have had a similar residence time at mineral interfaces and in fractures (fracture fluids) have not been previously identified. Expulsion of fracture fluids from basement systems with low connectivity occurs through deformation and fracturing of the brittle crust. The fractal nature of this process must, at some scale, preserve pockets of interconnected fluid from the earliest crustal history. In one such system, 2.8 kilometres below the surface in a South African gold mine, extant chemoautotrophic microbes have been identified in fluids isolated from the photosphere on timescales of tens of millions of years. Deep fracture fluids with similar chemistry have been found in a mine in the Timmins, Ontario, area of the Canadian Precambrian Shield. Here we show that excesses of (124)Xe, (126)Xe and (128)Xe in the Timmins mine fluids can be linked to xenon isotope changes in the ancient atmosphere and used to calculate a minimum mean residence time for this fluid of about 1.5 billion years. Further evidence of an ancient fluid system is found in (129)Xe excesses that, owing to the absence of any identifiable mantle input, are probably sourced in sediments and extracted by fluid migration processes operating during or shortly after mineralization at around 2.64 billion years ago. We also provide closed-system radiogenic noble-gas ((4)He, (21)Ne, (40)Ar, (136)Xe) residence times. Together, the different noble gases show that ancient pockets of water can survive the crustal fracturing process and remain in the crust for billions of years.

It is unknown when life first appeared on Earth. The earliest known microfossils (approximately 3,500 Myr before present) are structurally complex, and if it is assumed that the associated organisms required a long time to develop this degree of complexity, then the existence of life much earlier than this can be argued. But the known examples of crustal rocks older than 3,500 Myr have experienced intense metamorphism, which would have obliterated any fragile microfossils contained therein. It is therefore necessary to search for geochemical evidence of past biotic activity that has been preserved within minerals that are resistant to metamorphism. Here we report ion-microprobe measurements of the carbon-isotope composition of carbonaceous inclusions within grains of apatite (basic calcium phosphate) from the oldest known sediment sequences--a approximately 3,800-Myr-old banded iron formation from the Isua supracrustal belt, West Greenland, and a similar formation from the nearby Akilia island that is possibly older than 3,850 Myr. The carbon in the carbonaceous inclusions is isotopically light, indicative of biological activity; no known abiotic process can explain the data. Unless some unknown abiotic process exists which is able both to create such isotopically light carbon and then selectively incorporate it into apatite grains, our results provide evidence for the emergence of life on Earth by at least 3,800 Myr before present.

THE late Archaean greenstone terranes of Western Australia contain abundant komatiites (high-MgO lavas) hosting magmatic sulphide deposits rich in nickel, copper and platinum-group elements. Thermal erosion and assimilation of sulphidic sea-floor sediments has been proposed as a mechanism by which the komatiites were brought to sulphide saturation1–4. Such models have important implications not only for the genesis of these sulphide ores, but also for interpreting the magnitude and extent of isotopic heterogeneity in the Archaean mantle. Here we report that massive, matrix and disseminated sulphide ores and a komatiite from Western Australia yield a magmatic Re–Os isochron age of 2,706 ± 36 Myr and a near-chondritic initial 187Os/188Os ratio of 0.10889 ± 0.00035, whereas a proposed sulphidic sedimentary contaminant has an extremely radiogenic 187Os/188Os of 1.0983 at 2,706 Myr. These data demonstrate that the ore-forming komatiites were derived from the upper mantle without significant contamination by radiogenic crust either before eruption or during turbulent flow at the surface. Thus, ground melting and assimilation of sulphidic sediments may not be as important in ore genesis as current theories suggest.

The evolution of non-photosynthetic sulphide-oxidizing bacteria was contemporaneous with a large shift in the isotopic composition of biogenic sedimentary sulphides between 0.64 and 1.05 billion years ago. Both events were probably driven by a rise in atmospheric oxygen concentrations to greater than 5-18% of present levels--a change that may also have triggered the evolution of animals.

A new interpretation of fossilized soils (palaeosols) suggests that at least some Ediacaran (625–542 million years ago) organisms lived on land; thus these Ediacaran fossils were not animals, but a fungus-dominated terrestrial biota that predated vascular plants by about 100 million years. Ediacaran fossils —542 to 635 million years old — occur worldwide in a variety of sedimentary deposits, generally interpreted as shallow to deep marine origin. They have been regarded as early animal ancestors of the Cambrian evolutionary explosion of marine invertebrate phyla, as giant marine protists and as lichenized fungi. Here, Gregory Retallack raises doubts over the assumption that Ediacaran organisms were marine: a new interpretation of fossilized soils ('palaeosols') from South Australia suggests that at least some lived on land. Retallack's suggestion that some Ediacaran fossils were large sessile organisms of cool, dry soils is compatible with observations that Ediacaran fossils are similar in appearance and preservation to lichens and other microbial colonies of biological soil crusts, rather than marine animals or protists. Ediacaran (635–542 million years ago) fossils have been regarded as early animal ancestors of the Cambrian evolutionary explosion of marine invertebrate phyla1, as giant marine protists2 and as lichenized fungi3. Recent documentation of palaeosols in the Ediacara Member of the Rawnsley Quartzite of South Australia4 confirms past interpretations of lagoonal–aeolian deposition based on synsedimentary ferruginization and loessic texture5,6. Further evidence for palaeosols comes from non-marine facies, dilation cracks, soil nodules, sand crystals, stable isotopic data and mass balance geochemistry4. Here I show that the uppermost surfaces of the palaeosols have a variety of fossils in growth position, including Charniodiscus, Dickinsonia, Hallidaya, Parvancorina, Phyllozoon, Praecambridium, Rugoconites, Tribrachidium and ‘old-elephant skin’ (ichnogenus Rivularites7). These fossils were preserved as ferruginous impressions, like plant fossils8, and biological soil crusts9,10 of Phanerozoic eon sandy palaeosols. Sand crystals after gypsum11 and nodules of carbonate12 are shallow within the palaeosols4, even after correcting for burial compaction13. Periglacial involutions and modest geochemical differentiation of the palaeosols are evidence of a dry, cold temperate Ediacaran palaeoclimate in South Australia4. This new interpretation of some Ediacaran fossils as large sessile organisms of cool, dry soils, is compatible with observations that Ediacaran fossils were similar in appearance and preservation to lichens and other microbial colonies of biological soil crusts3, rather than marine animals1, or protists2.

The 'Late Heavy Bombardment' was a phase in the impact history of the Moon that occurred 3.8-4.0 Gyr ago, when the lunar basins with known dates were formed(1,2). But no record of this event has yet been reported from the few surviving rocks of this age on the Earth. Here we report tungsten isotope anomalies, based on the Hf-182-W-182 system (half-life of 9 Myr), in metamorphosed sedimentary rocks from the 3.7-3.8-Gyr-old Isua greenstone belt of West Greenland and closely related rocks from northern Labrador, Canada. As it is difficult to conceive of a mechanism by which tungsten isotope heterogeneities could have been preserved in the Earth's dynamic crust-mantle environment from a time when short-lived Hf-182 was still present, we conclude that the metamorphosed sediments contain a component derived from meteorites.

The snowball Earth hypothesis contends that most of the planet was covered with thick ice for millions of years, about three-quarters of a billion years ago. A dramatic warming event, thought to have been fuelled by high atmospheric carbon dioxide levels, eventually defrosted Earth. A new isotopic analysis of rocks that were laid down after the initial glaciation challenges the idea of a hard snowball Earth. The data are consistent with a rather low partial pressure of atmospheric CO2, not the orders of magnitude above current levels that would provide the extreme warming necessary for escape from snowball Earth. This supports the idea that the initial glaciation was much less severe than has been proposed. The snowball Earth hypothesis postulates that the planet was entirely covered by ice for millions of years in the Neoproterozoic era, in a self-enhanced glaciation caused by the high albedo of the ice-covered planet. In a hard-snowball picture, the subsequent rapid unfreezing resulted from an ultra-greenhouse event attributed to the buildup of volcanic carbon dioxide (CO2) during glaciation1. High partial pressures of atmospheric CO2 ( ; from 20,000 to 90,000 p.p.m.v.) in the aftermath of the Marinoan glaciation (∼635 Myr ago) have been inferred from both boron and triple oxygen isotopes2,3. These values are 50 to 225 times higher than present-day levels. Here, we re-evaluate these estimates using paired carbon isotopic data for carbonate layers that cap Neoproterozoic glacial deposits and are considered to record post-glacial sea level rise1. The new data reported here for Brazilian cap carbonates, together with previous ones for time-equivalent units4,5,6,7,8, provide estimates lower than 3,200 p.p.m.v.—and possibly as low as the current value of ∼400 p.p.m.v. Our new constraint, and our re-interpretation of the boron and triple oxygen isotope data, provide a completely different picture of the late Neoproterozoic environment, with low atmospheric concentrations of carbon dioxide and oxygen that are inconsistent with a hard-snowball Earth.

THE volume of early Archaean crust that still survives today is very small (less than 1% of the present continental volume). This has been interpreted as indicating that crustal growth did not begin until about 4.0 Gyr ago, before which the silicate Earth remained well mixed and essentially undifferentiated. But the existence in some early Archaean rocks1–6 of inferred initial 143Nd/144Nd ratios higher than that of the bulk Earth suggests that by 3.8 Gyr ago the volume of the crust was as large as about 40% of the present value3,7,8. Given the apparently low rate of recycling in the Hadean era (that is, before 4 Gyr ago)7,8, consideration of 147Sm–143Nd systematics then suggests that primordial differentiation of the Earth may have begun ~4.5 Gyr ago. Here we present evidence for early differentiation, based on measurements of 143Nd/144Nd and 142Nd/144Nd ratios in a ~3.8-Gyr-old supracrustal rock from Isua, West Greenland. Coupled 146,147Sm–142,143Nd systematics suggest that fractionation of Sm/Nd took place 4.44–4.54 Gyr ago, owing to extraction of a light-rare-earth-element-enriched primordial crust.

ONE of the main problems in understanding the origin of the Witwatersrand Basin mineralization concerns the source of the prodigious concentrations of gold and uranium in the basin; another problem is the age and nature of the rocks in the source area. Recent data1,2 indicate that the source area was largely 3,200–2,900 Myr old, becoming progressively younger as the basin evolved. Granites adjoining the basin are commonly overprinted by pervasive and vein-related hydrothermal alteration, which resulted in an enrichment of both gold and uranium3,4. Here we report uranium-lead data demonstrating that these hydrothermally altered granites were intruded sporadically, both before and during the 360-Myr period5 constraining Witwatersrand sediment deposition, and that hydrothermal alteration was associated with granitoid intrusion and subsequent cooling. Thus, 'fertilization' of the crust occurred during basin evolution, a feature that supports the idea19 that gold and uranium mineralization occurred simultaneously with sediment deposition.

The widely held view that the Archaean climate around 3.5 billion years ago was remarkably warm — with ocean temperatures perhaps as high as 80 °C — has been questioned on the grounds that the established method of estimating ancient ocean temperature (from the oxygen isotope ratios of sedimentary deposits) is subject to significant uncertainty. Michael Hren et al. have adopted a different approach to estimating ocean temperature, based on an analysis of both oxygen and hydrogen isotopes of 3.4 billion-year-old Buck Reef Chert sediments from South Africa. The isotopes sampled are consistent with formation in waters no warmer than about 40 °C, suggesting that Earth's early oceans may have been far cooler than previously thought. The study of stable oxygen isotope ratios (δ18O) of Precambrian cherts suggests that ocean temperatures during the Archaean era (about 3.5 billion years ago) were between 55 °C and 85 °C, but uncertainty about the δ18O of the primitive ocean has led to considerable debate regarding this conclusion. Here, a combined analysis of oxygen and hydrogen istopes sampled from 3.42-billion-year-old Buck Reef Chert rocks in South Africa indicates that the ancient ocean was much cooler than previously thought. Stable oxygen isotope ratios (δ18O) of Precambrian cherts have been used to establish much of our understanding of the early climate history of Earth1,2,3 and suggest that ocean temperatures during the Archaean era (∼3.5 billion years ago) were between 55 °C and 85 °C (ref. 2). But, because of uncertainty in the δ18O of the primitive ocean, there is considerable debate regarding this conclusion. Examination of modern and ancient cherts indicates that another approach, using a combined analysis of δ18O and hydrogen isotopes (δD) rather than δ18O alone, can provide a firmer constraint on formational temperatures without independent knowledge of the isotopic composition of ambient waters4,5. Here we show that δ18O and δD sampled from 3.42-billion-year-old Buck Reef Chert rocks in South Africa are consistent with formation from waters at varied low temperatures. The most 18O-enriched Buck Reef Chert rocks record the lowest diagenetic temperatures and were formed in equilibrium with waters below ∼40 °C. Geochemical and sedimentary evidence suggests that the Buck Reef Chert was formed in shallow to deep marine conditions, so our results indicate that the Palaeoarchaean ocean was isotopically depleted relative to the modern ocean and far cooler (≤40 °C) than previously thought2.

The cause of the most marked changes in the evolution of life, which define the first-order stratigraphic boundary between the Precambrian and the Phanerozoic eon, remains enigmatic and a highly topical subject of debate. A global ocean anoxic event, triggered by large-scale hydrogen sulphide (H(2)S) release to surface waters, has been suggested by Wille et al., on the basis of two data sets from South China and Oman, to explain the fundamental biological changes across the Precambrian/Cambrian (PC/C) boundary. Here we report a new precise SHRIMP U-Pb zircon age of 532.3 +/- 0.7 million years (Myr) ago (Fig. 1) for a volcanic ash bed in the critical unit that reflects the ocean anoxic event, the lowermost black shale sequence of the Niutitang Formation in the Guizhou Province, South China. This age is significantly younger than the precise PC/C boundary age of 542.0 +/- 0.3 Myr ago, approximately 10 Myr younger than the extinction of the Ediacaran fauna, and thus challenging the view of a major ocean anoxic event having been responsible for the major changes in the direction of evolution at the PC/C boundary.

The world's main sources of non-alluvial diamonds are found in ultrapotassic kimberlite1 or lamproite2 diatremes (pipes filled with explosive volcanic debris), most of which have Phanerozoic ages and are located in stable Precambrian cratons. Diamond exploration has therefore tended to focus on such deposits. Microdiamonds are known to occur in metamorphic rocks such as gneiss3 and eclogite4 that have equilibrated deep in the mantle and were then tectonically transported to the surface, but such deposits are thought to have little commercial potential. Here we report a new type of diamond occurrence from the Dachine region in French Guiana for which the host rock is volcaniclastic komatiite—an unusual type of volcanic rock whose composition and origin are quite unlike those of kimberlite and lamproite. These komatiites form part of a Proterozoic island-arc sequence, a tectonic setting distinct from that of all other currently exploited diamond deposits. The discovery of diamonds in volcaniclastic komatiite has implications not only for diamond exploration, but also provides strong evidence that these komatiite magmas originated at depths of 250 km or greater within the Earth.

The first 600 million years or so of the Earth's history (the Hadean Eon) remain poorly understood, largely because there is no rock record dating from that era. Ancient zircons from that time period are, however, resistant minerals that can potentially provide insights into the conditions on the Earth at that time. Mark Harrison and colleagues now present an examination of inclusion assemblages in Hadean zircons from Jack Hills, Western Australia, which constrain the magmatic formation conditions to about 700°C and 7 kbar. This result implies a near-surface heat flow about three to five times lower than estimates of global Hadean heat flow. As the only site of magmatism on modern Earth that is characterized by heat flow of a quarter of the global average is above subduction zones, the authors suggest that the magmas from which the Jack Hills Hadean zircons crystallized, formed largely in an underthrust environment, perhaps similar to modern convergent margins. An examination of inclusion assemblages in Hadean zircons from Jack Hills (Western Australia) is presented, which constrain the magmatic formation conditions to about 700 °C and 7 kbars. This result implies a near-surface heat flow about 3–5 times lower than estimates of global Hadean heat flow. As the only site of magmatism on modern Earth that is characterized by heat flow of about one-quarter of the global average is above subduction zones, it is suggested that the magmas from which the Jack Hills Hadean zircons crystallized were formed largely in an underthrust environment, perhaps similar to modern convergent margins. The first ∼600 million years of Earth history (the ‘Hadean’ eon) remain poorly understood, largely because there is no rock record dating from that era. Detrital Hadean igneous zircons from the Jack Hills1, Western Australia, however, can potentially provide insights into the conditions extant on our planet at that time. Results of geochemical investigations2,3,4,5,6,7,8,9,10,11,12,13 using these ancient grains have been interpreted to suggest the presence of a hydrosphere2,3,4,7,8 and continental crust9,10 before 4 Gyr. An underexploited characteristic of the >4 Gyr zircons is their diverse assemblage of mineral inclusions14,15,16,17. Here we present an examination of over 400 Hadean zircons from Jack Hills, which shows that some inclusion assemblages are conducive to thermobarometry. Our thermobarometric analyses of 4.02–4.19-Gyr-old inclusion-bearing zircons constrain their magmatic formation conditions to about 700 °C and 7 kbar. This result implies a near-surface heat flow of ∼75 mW m-2, about three to five times lower than estimates of Hadean global heat flow. As the only site of magmatism on modern Earth that is characterized by heat flow of about one-quarter of the global average is above subduction zones, we suggest that the magmas from which the Jack Hills Hadean zircons crystallized were formed largely in an underthrust environment, perhaps similar to modern convergent margins.

Comparisons of genome sequence data in closely and distantly related modern organisms can be used for the computational reconstruction of ancient protein sequences that may have existed in related but now extinct types. These proteins can then be 'resurrected' in the laboratory. This has now been achieved for a group of 25 ancestral elongation factors from bacteria across an estimated span of 3 billion years. These ancient proteins display a near linear increase in thermostability travelling back in geological time, suggesting that the environment supporting ancient life was initially hot, then cooled progressively by about 30 °C during that period. This pattern is corroborated by the palaeotemperature trend inferred for the geologic record. Using phylogeny to reconstruct proteins inferred to have existed in the ancestry of organisms is a powerful and provocative way to discover the evolutionary processes of life. A protein from various stages in the history of life has been reconstructed to show that life originated in a hot environment, and has tracked the cooling of Earth through time. Biosignatures and structures in the geological record indicate that microbial life has inhabited Earth for the past 3.5 billion years or so1,2. Research in the physical sciences has been able to generate statements about the ancient environment that hosted this life3,4,5,6. These include the chemical compositions and temperatures of the early ocean and atmosphere. Only recently have the natural sciences been able to provide experimental results describing the environments of ancient life. Our previous work with resurrected proteins indicated that ancient life lived in a hot environment7,8. Here we expand the timescale of resurrected proteins to provide a palaeotemperature trend of the environments that hosted life from 3.5 to 0.5 billion years ago. The thermostability of more than 25 phylogenetically dispersed ancestral elongation factors suggest that the environment supporting ancient life cooled progressively by 30 °C during that period. Here we show that our results are robust to potential statistical bias associated with the posterior distribution of inferred character states, phylogenetic ambiguity, and uncertainties in the amino-acid equilibrium frequencies used by evolutionary models. Our results are further supported by a nearly identical cooling trend for the ancient ocean as inferred from the deposition of oxygen isotopes. The convergence of results from natural and physical sciences suggest that ancient life has continually adapted to changes in environmental temperatures throughout its evolutionary history.

The recent discovery of diamond and graphite inclusions in zircon grains formed over 4 billion years ago in the Jack Hills meta-sedimentary belt in Western Australia has given geologists a glimpse of Earth's earliest known carbon reservoir. New ion microprobe analyses of the carbon isotope composition of these inclusions reveal low carbon isotopic ratios, which could reflect deep subduction of biogenic surface carbon. Though this is not unambiguous evidence for life on Earth as early as 4,250 million years ago, as low carbon isotope values can also be produced by certain inorganic chemical reactions. The recent discovery of diamond graphite inclusions in the Earth's oldest zircon grains from the Jack Hills metasediments in Western Australia provides a unique opportunity to investigate Earth's earliest known carbon reservoir. This paper reports ion microprobe analyses of the carbon isotope composition of these diamond-graphite inclusions and finds low carbon isotopic ratios, which may reflect deep subduction of biogenic surface carbon. But such carbon isotope values may also be produced by inorganic chemical reactions. The recent discovery of diamond–graphite inclusions in the Earth’s oldest zircon grains (formed up to 4,252 Myr ago) from the Jack Hills metasediments in Western Australia1 provides a unique opportunity to investigate Earth’s earliest known carbon reservoir. Here we report ion microprobe analyses of the carbon isotope composition of these diamond–graphite inclusions. The observed δ13CPDB values (expressed using the PeeDee Belemnite standard) range between -5 per mil and -58 per mil with a median of -31 per mil. This extends beyond typical mantle values of around -6 per mil to values observed in metamorphic and some eclogitic diamonds that are interpreted to reflect deep subduction of low-δ13CPDB biogenic surface carbon. Low δ13CPDB values may also be produced by inorganic chemical reactions2, and therefore are not unambiguous evidence for life on Earth as early as 4,250 Myr ago. Regardless, our results suggest that a low-δ13CPDB reservoir may have existed on the early Earth.

Information about the past composition of the Earth's atmosphere on geological timescales is hard to come by. So the debut of a new stable isotope proxy for ancient atmospheric condition is a notable event. The proxy, the triple oxygen isotope composition of sulphate from ancient evaporites and barites, exhibits variable negative oxygen-17 anomalies over the past 750 million years. The anomalies track atmospheric oxygen and in turn reflect the partial pressure of carbon disoide via a stratospheric ozone/carbon dioxide/oxygen photochemical reaction network. In line with modelling results, the proxy data point to a high-carbon dioxide atmosphere in the Early Cambrian compared to earlier eras. Significantly, the oxygen-17 anomalies of barites from Marinoan cap carbonates (∼ 635 million years ago) display a distinct negative spike, suggesting that carbon dioxide was still high when barite was precipitating in the cap carbonate sequences. This supports the Neoproterozoic 'snowball' Earth hypothesis and/or massive methane release after the Marinoan glaciation. Historical information about the atmosphere is scarce. Huiming Bao and colleagues show that the triple oxygen isotope composition of sulphate from ancient evaporites and barites exhibits variable negative oxygen-17 isotope anomalies over the past 750 million years. Understanding the composition of the atmosphere over geological time is critical to understanding the history of the Earth system, as the atmosphere is closely linked to the lithosphere, hydrosphere and biosphere. Although much of the history of the lithosphere and hydrosphere is contained in rock and mineral records, corresponding information about the atmosphere is scarce and elusive owing to the lack of direct records. Geologists have used sedimentary minerals, fossils and geochemical models to place constraints on the concentrations of carbon dioxide, oxygen or methane in the past1,2,3,4. Here we show that the triple oxygen isotope composition of sulphate from ancient evaporites and barites shows variable negative oxygen-17 isotope anomalies over the past 750 million years. We propose that these anomalies track those of atmospheric oxygen and in turn reflect the partial pressure of carbon dioxide ( ) in the past through a photochemical reaction network linking stratospheric ozone to carbon dioxide and to oxygen5,6. Our results suggest that was much higher in the early Cambrian than in younger eras, agreeing with previous modelling results2. We also find that the 17O isotope anomalies of barites from Marinoan (∼635 million years ago) cap carbonates display a distinct negative spike (around -0.70‰), suggesting that by the time barite was precipitating in the immediate aftermath of a Neoproterozoic global glaciation, the was at its highest level in the past 750 million years. Our finding is consistent with the ‘snowball Earth’ hypothesis7,8 and/or a massive methane release9 after the Marinoan glaciation.

Changes in environmental conditions at the Precambrian–Cambrian transition (around 542 million years ago) have been suggested as a possible explanation for the apparent rapid increase in abundance of multicellular organisms known as the 'Cambrian explosion'. The nature of the environmental changes is still a matter of debate, however. Wille et al. now report molybdenum isotope signatures of black shales from two stratigraphically correlated sample sets with a depositional age of about 542 million years. With the help of a box model of the oceanic molybdenum cycle, they find that intense upwelling of hydrogen sulphide-rich deep ocean water best explains the observed early Cambrian molybdenum isotope signal. This suggests that the early Cambrian animal radiation may have been triggered by a major change in ocean circulation, following a long period during which the ocean was stratified, with sulphidic deep water. This paper reports molybdenum isotope signatures of black shales from two stratigraphically correlated sample sets with a depositional age of around 542 million years. The findings suggest that the Early Cambrian animal radiation may have been triggered by a major change in ocean circulation, terminating a long period during which the Proterozoic ocean was stratified, with sulphidic deep water. Animal-like multicellular fossils appeared towards the end of the Precambrian, followed by a rapid increase in the abundance and diversity of fossils during the Early Cambrian period, an event also known as the ‘Cambrian explosion’1,2,3. Changes in the environmental conditions at the Precambrian/Cambrian transition (about 542 Myr ago) have been suggested as a possible explanation for this event, but are still a matter of debate1,2,3. Here we report molybdenum isotope signatures of black shales from two stratigraphically correlated sample sets with a depositional age of around 542 Myr. We find a transient molybdenum isotope signal immediately after the Precambrian/Cambrian transition. Using a box model of the oceanic molybdenum cycle, we find that intense upwelling of hydrogen sulphide-rich deep ocean water best explains the observed Early Cambrian molybdenum isotope signal. Our findings suggest that the Early Cambrian animal radiation may have been triggered by a major change in ocean circulation, terminating a long period during which the Proterozoic ocean was stratified, with sulphidic deep water.

The oxygenation of the Earth's atmosphere is thought to have occurred in two steps near the beginning and the end of the Proterozoic eon, around 2,500 to 550 million years ago. The oxidation state of the ocean between these two steps and the timing of deep ocean oxygenation, however, remain poorly known. Scott et al. now use molybdenum and total organic carbon data from black shales to track the redox state of the ocean at this time. Molybdenum is an essential participant in nutrient cycling, and its availability is highly sensitive to Earth's redox state. The results provide a new narrative for the historical texture of Earth's oxygenation, and will be of relevance for the study of the events that presaged the appearance of animals on Earth. Molybdenum and total organic carbon data from black shales is used to gain insights into the redox state of the ocean. The data suggests mild oxidative weathering of the continents before ∼2,200 Myr ago, but weathering becomes more persistent and vigorous at ∼2,150 Myr ago, 200 million years after the initial rise in atmospheric oxygen. Limited availability of molybdenum after 1,800 Myr ago may have acted as a negative nutrient feedback limiting the spatial and temporal extent of sulphidic conditions. Biogeochemical signatures preserved in ancient sedimentary rocks provide clues to the nature and timing of the oxygenation of the Earth’s atmosphere. Geochemical data1,2,3,4,5,6 suggest that oxygenation proceeded in two broad steps near the beginning and end of the Proterozoic eon (2,500 to 542 million years ago). The oxidation state of the Proterozoic ocean between these two steps and the timing of deep-ocean oxygenation have important implications for the evolutionary course of life on Earth but remain poorly known. Here we present a new perspective on ocean oxygenation based on the authigenic accumulation of the redox-sensitive transition element molybdenum in sulphidic black shales. Accumulation of authigenic molybdenum from sea water is already seen in shales by 2,650 Myr ago; however, the small magnitudes of these enrichments reflect weak or transient7 sources of dissolved molybdenum before about 2,200 Myr ago, consistent with minimal oxidative weathering of the continents. Enrichments indicative of persistent and vigorous oxidative weathering appear in shales deposited at roughly 2,150 Myr ago, more than 200 million years after the initial rise in atmospheric oxygen1,2. Subsequent expansion of sulphidic conditions after about 1,800 Myr ago (refs 8, 9) maintained a mid-Proterozoic molybdenum reservoir below 20 per cent of the modern inventory, which in turn may have acted as a nutrient feedback limiting the spatiotemporal distribution of euxinic (sulphidic) bottom waters and perhaps the evolutionary and ecological expansion of eukaryotic organisms10. By 551 Myr ago, molybdenum contents reflect a greatly expanded oceanic reservoir due to oxygenation of the deep ocean and corresponding decrease in sulphidic conditions in the sediments and water column.

The observation of non-mass-dependent sulphur isotope ratios in sedimentary rocks more than ∼2.4 billion years old and the disappearance of this signal in younger sediments is taken as evidence for the transition from an anoxic to oxic atmosphere around 2.4 Gyr ago. But now, the preservation of a non mass-dependent signal that differs from that of preceding and following periods in the Archean is demonstrated. The findings support the original idea of an anoxic early atmosphere before 2.4 Gyr ago, and at the same time identifies variability within the isotope record that suggests changes in pre-2.4 Gya atmospheric pathways for non-mass-dependent chemistry and in the ultraviolet transparency of an evolving early atmosphere. The evolution of the Earth’s atmosphere is marked by a transition from an early atmosphere with very low oxygen content to one with an oxygen content within a few per cent of the present atmospheric level. Placing time constraints on this transition is of interest because it identifies the time when oxidative weathering became efficient, when ocean chemistry was transformed by delivery of oxygen and sulphate, and when a large part of Earth’s ecology changed from anaerobic to aerobic1. The observation of non-mass-dependent sulphur isotope ratios in sedimentary rocks more than ∼2.45 billion years (2.45 Gyr) old and the disappearance of this signal in younger sediments is taken as one of the strongest lines of evidence for the transition from an anoxic to an oxic atmosphere around 2.45 Gyr ago1,2,3,4,5. Detailed examination of the sulphur isotope record before 2.45 Gyr ago also reveals early and late periods of large amplitude non-mass-dependent signals bracketing an intervening period when the signal was attenuated5,6,7,8,9. Until recently, this record has been too sparse to allow interpretation, but collection of new data has prompted some workers8 to argue that the Mesoarchaean interval (3.2–2.8 Gyr ago) lacks a non-mass-dependent signal, and records the effects of earlier and possibly permanent oxygenation of the Earth’s atmosphere. Here we focus on the Mesoarchaean interval, and demonstrate preservation of a non-mass-dependent signal that differs from that of preceding and following periods in the Archaean. Our findings point to the persistence of an anoxic early atmosphere, and identify variability within the isotope record that suggests changes in pre-2.45-Gyr-ago atmospheric pathways for non-mass-dependent chemistry and in the ultraviolet transparency of an evolving early atmosphere.

The hypothesis that the establishment of a permanently oxygenated atmosphere at the Archaean-Proterozoic transition (approximately 2.5 billion years ago) occurred when oxygen-producing cyanobacteria evolved is contradicted by biomarker evidence for their presence in rocks 200 million years older. To sustain vanishingly low oxygen levels despite near-modern rates of oxygen production from approximately 2.7-2.5 billion years ago thus requires that oxygen sinks must have been much larger than they are now. Here we propose that the rise of atmospheric oxygen occurred because the predominant sink for oxygen in the Archaean era-enhanced submarine volcanism-was abruptly and permanently diminished during the Archaean-Proterozoic transition. Observations are consistent with the corollary that subaerial volcanism only became widespread after a major tectonic episode of continental stabilization at the beginning of the Proterozoic. Submarine volcanoes are more reducing than subaerial volcanoes, so a shift from predominantly submarine to a mix of subaerial and submarine volcanism more similar to that observed today would have reduced the overall sink for oxygen and led to the rise of atmospheric oxygen.

Micrometeorites-submillimetre-sized particles derived from asteroids and comets-occur in significant quantities in deep sea sediments, and the ice sheets of Greenland and Antarctica. The most abundant micrometeorites are cosmic spherules, which contain nickel-rich spinels that were crystallized and oxidized during atmospheric entry, therefore recording the oxygen content in the uppermost atmosphere. But the use of micrometeorites for detecting past changes in the flux of incoming extraterrestrial matter, and as probes of the evolution of the atmosphere, has been hampered by the fact that most objects with depositional ages higher than 0.5 Myr show severe chemical alteration. Here we report the discovery of unaltered cosmic spherules in a 1.4-Gyr-old sandstone (red bed) from Finland. From this we infer that red beds, a common lithology in the Earth's history, may contain substantial unbiased populations of fossil micrometeorites. The study of such populations would allow systematic research on variations in the micrometeorite flux from the early Proterozoic era to recent times (a time span of about 2.5 Gyr), and could help to better constrain the time when the atmospheric oxygen content was raised to its present level.

The globular microfossils from the Neoproterozoic Doushantuo Formation in China are arguably one of the most significant fossil finds in the past decade. They were thought to be animal embryos, based on size and the presence of reductive cell division. If the attribution is correct, 600-million-year-old fossilized cells would provide an important window into early animal evolution. But they may not be embryos at all. The recent discovery of reductive cell division in a modern sulphur bacterium, and their direct association with phosphate minerals, find exact parallels in with the Doushantuo microfossils. Such an accumulation of animal embryos has always been seen as problematic, and no plausible phosphatization mechanism has been offered. The simplest explanation, therefore, is that these are the fossils of giant sulphur bacteria. In situ phosphatization1 and reductive cell division2 have recently been discovered within the vacuolate sulphur-oxidizing bacteria. Here we show that certain Neoproterozoic Doushantuo Formation (about 600 million years bp) microfossils, including structures previously interpreted as the oldest known metazoan eggs and embryos3,4,5,6,7,8,9,10, can be interpreted as giant vacuolate sulphur bacteria. Sulphur bacteria of the genus Thiomargarita have sizes and morphologies similar to those of many Doushantuo microfossils, including symmetrical cell clusters that result from multiple stages of reductive division in three planes. We also propose that Doushantuo phosphorite precipitation was mediated by these bacteria, as shown in modern Thiomargarita-associated phosphogenic sites, thus providing the taphonomic conditions that preserved other fossils known from the Doushantuo Formation.

The terrestrial sediment record indicates that the Earth's climate varied drastically in the Precambrian era (before 550 million years ago), ranging from surface temperatures similar to or higher than today's to global glaciation events. The most continuous record of sea surface temperatures of that time has been derived from variations in oxygen isotope ratios of cherts (siliceous sediments), but the long-term cooling of the oceans inferred from those data has been questioned because the oxygen isotope signature could have been reset through the exchange with hydrothermal fluids after deposition of the sediments. Here we show that the silicon isotopic composition of cherts more than 550 million years old shows systematic variations with age that support the earlier conclusion of long-term ocean cooling and exclude post-depositional exchange as the main source of the isotopic variations. In agreement with other lines of evidence, a model of the silicon cycle in the Precambrian era shows that the observed silicon isotope variations imply seawater temperature changes from about 70 degrees C 3,500 million years ago to about 20 degrees C 800 million years ago.

Oxygenation of the Earth's surface is thought to have occurred in two main steps. The first, about 2,300 million years ago, saw a significant increase in atmospheric and surface ocean oxygen levels and has been widely studied. Much less is known about the second step, which took place around 800–540 million years ago and appears to have been associated with the evolution of complex animals. Now carbon and sulphur isotope records from sediments in the Sultanate of Oman have been used to construct a record of the amount of oxygen in the ocean during the Ediacaran period (635–542 million years ago). The data suggest that there were three distinct stages of oxidation within this period: the second stage involved an increase in oxygen in the deep ocean and appears to have been associated with the evolution of complex animals. Reconstruction of the amount of oxygen in the ocean during the Ediacaran period using carbon- and sulphur-isotope records identifies three distinct stages of oxidation over this interval. Complex animals may have evolved during the second stage, indicating that this event may have played a key role in the evolution of eukaryotic organisms. Oxygenation of the Earth’s surface is increasingly thought to have occurred in two steps. The first step, which occurred ∼2,300 million years (Myr) ago, involved a significant increase in atmospheric oxygen concentrations and oxygenation of the surface ocean1,2. A further increase in atmospheric oxygen appears to have taken place during the late Neoproterozoic period3,4 (∼800–542 Myr ago). This increase may have stimulated the evolution of macroscopic multicellular animals and the subsequent radiation of calcified invertebrates4,5, and may have led to oxygenation of the deep ocean6. However, the nature and timing of Neoproterozoic oxidation remain uncertain. Here we present high-resolution carbon isotope and sulphur isotope records from the Huqf Supergroup, Sultanate of Oman, that cover most of the Ediacaran period (∼635 to ∼548 Myr ago). These records indicate that the ocean became increasingly oxygenated after the end of the Marinoan glaciation, and they allow us to identify three distinct stages of oxidation. When considered in the context of other records from this period7,8,9,10,11,12,13,14,15, our data indicate that certain groups of eukaryotic organisms appeared and diversified during the second and third stages of oxygenation. The second stage corresponds with the Shuram excursion in the carbon isotope record16 and seems to have involved the oxidation of a large reservoir of organic carbon suspended in the deep ocean6, indicating that this event may have had a key role in the evolution of eukaryotic organisms. Our data thus provide new insights into the oxygenation of the Ediacaran ocean and the stepwise restructuring of the carbon6,16,17 and sulphur cycles3,18,19 that occurred during this significant period of Earth’s history.