Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Earth science

Sea change for the rise of oxygen

A model proposes that falling sea levels shifted the make-up of volcanic gases on the early Earth, triggering a chain of events that may have allowed photosynthesis in the ocean to oxygenate the atmosphere. See Letter p.229

A burning question for those studying the early Earth is whether oxygen production started in the ocean before it first began to accumulate in the atmosphere. A growing body of work convincingly shows that cyanobacteria in the ocean produced oxygen roughly 200 million to 300 million years before the Great Oxidation Event (GOE) — the first appreciable build-up of oxygen in the atmosphere, which occurred about 2.4 billion years ago. But for oxygen to have been produced without accumulating in the atmosphere, efficient sinks or buffers must have existed to consume it as quickly as it was generated. On page 229 of this issue, Gaillard and co-authors1 offer their take on this conundrum. They propose that falling sea levels across growing, stabilizing continents altered the gas composition of volcanoes, and that these tectonic changes weakened oxygen buffers, enabling the GOE.

The idea of pre-GOE oxygen production received a big boost with the recognition2 that organic molecules preserved in roughly 2.7-billion-year-old shaley rocks in Western Australia are biomarkers both of cyanobacterial oxygen production and of oxygen-demanding steroid biosynthesis by eukaryotes (organisms such as plants, animals and fungi). The validity of these biomarkers has since been questioned3,4, but other strong evidence has been found indicating that oxygen was produced5, and at times accumulated at very low levels6,7,8, before the GOE.

Previous models of oxygen buffering postulated the release of chemically reduced volcanic gases, such as hydrogen, that reacted efficiently with oxygen emerging from biological sources. Only when the levels of those oxygen-consuming gases fell did atmospheric oxygenation begin, producing the clear, irreversible evidence now thought of as the smoking gun of the GOE — the permanent disappearance of signatures known as mass-independent fractionations (MIFs) from the sulphur isotopic record of sedimentary minerals such as pyrite9.

A critical subtheme that has gained traction in oxygen-buffering models is the coincident, but probably not coincidental, occurrence of the GOE with one of Earth's great tectonic transitions — a purported spurt in the generation and stabilization of continents and their emergence from the sea. In 2007, Lee Kump and Mark Barley10 noted another important transition that paralleled the rise of oxygen levels and of the continents: a shift from previously dominant submarine volcanoes to subaerial volcanoes — those on land. They argued that the two types of volcanoes had fundamentally different redox states because of differences in the temperature and pressure at which their volatile components degassed from molten rock. Hotter, land volcanism would therefore have resulted in bountiful oxidized gases, including sulphur dioxide (SO2), which are poor oxygen buffers compared with the chemically reduced gases, such as hydrogen and hydrogen sulphide (H2S) resulting from the earlier, mainly submarine eruptions. The GOE followed in response.

Gaillard et al.1 now give us a way to think about continents, volcanoes and the GOE that builds on Kump and Barley's model. On the basis of studies of lava from modern Hawaii, the authors propose that submarine lavas need not be more reduced than those above the sea. They also performed thermodynamic calculations suggesting that the properties of volcanic gases — specifically, the relative concentrations of SO2 and H2S — are little affected by the redox state of lava.

The authors go on to provide a new spin on the volcano theory based on a model of the gas–melt equilibrium of Hawaiian tholeiitic basalt (which is assumed to be a good analogue for the early Earth). Specifically, they attribute increases in the overall amount of sulphur-containing gases released by volcanoes just before the GOE, and the accompanying major shift from reduced H2S towards oxidized SO2, simply to decreases in the average pressure of volcanic degassing. The emergence of continents jutting well above the seas would have produced the ubiquitous subaerial volcanoes that lowered the pressure of eruptions. The argument for continental emergence rests with another model11 asserting that long-term cooling of the mantle allowed for thicker, stronger continental crust, which rose above the oceans about 2.5 billion years ago.

Gaillard and colleagues' theory1 is built largely on theoretical thermal and chemical models that have ample degrees of uncertainty. But another key part of their proposition relies on data: the structure of sulphur's MIF record. The disappearance of MIF signals for sulphur in sedimentary minerals mirrors the rise of atmospheric oxygen at the GOE (Fig. 1), but equally impressive is the large magnitude of MIF signals that occur in the mineral record just before the disappearance. A recent model12 ties these large MIF signals to high ratios of SO2 to H2S roughly 2.7 billion to 2.5 billion years ago, but does not focus on the mechanisms behind the changing gas composition.

Figure 1: Sulphur isotopic data through Earth's history.

a, The isotopic composition of sulphur on Earth can be plotted as Δ33S — a measure of the deviation of a mineral's sulphur isotopic composition from that expected if the variations scaled strictly with the masses of the isotopes, expressed in parts per thousand9 (‰). Values above or below 0‰ indicate the magnitude of 'mass-independent fractionation' (MIF), which is linked to atmospheric chemistry and volcanic-gas releases. Before the Great Oxidation Event (shaded box), atmospheric chemical reactions such as those shown in b led to large Δ33S anomalies following the release of sulphur dioxide (SO2) from volcanoes into the atmosphere. Gaillard et al.1 suggest that these anomalies were caused by an increase in the ratio of SO2 to hydrogen sulphide (H2S) in volcanic emissions, and by an increase in the total amount of sulphur-containing gases released, both of which were the result of more widespread subaerial volcanism. They further propose that a concomitant increase in oceanic sulphate (SO42−) allowed biologically produced oxygen to accumulate in the atmosphere, causing subsequent loss of MIF signals. b, This simplified scheme shows MIF source reactions in the atmosphere following volcanic emission of SO2; indicates light-dependent reactions. The reaction sequences are abridged and depicted as unidirectional to emphasize the main products —sulphate (SO42−) and elemental sulphur (S8) — and pathways from the atmosphere to the ocean.

Gaillard et al. posit that the mechanistic key to atmospheric oxygenation might have been increased delivery of sulphate to the ocean through the same atmospheric, light-dependent reactions that yielded the high-magnitude MIF signals (Fig. 1). Simply put, increased inputs of sulphate (and indeed increased inputs of sulphur-containing gases overall) could have ramped up the production of H2S from sulphate reduction occurring at high-temperature volcanic vents on the sea floor. This H2S might then have reacted with reduced, dissolved iron released from the vents that would otherwise have served as an oxygen buffer. The greater flux of sulphur could even have allowed H2S to accumulate in parts of the ocean13.

Questions certainly remain about the buffering systems and the timing of events in Gaillard and colleagues' model. For example, a peak in the abundance of huge sedimentary iron formations occurs 2.7 billion to 2.5 billion years ago — the same time as MIF signals reached their apex. This relationship indicates that large amounts of reduced iron emerged from hydrothermal vents at that time, despite the putative increase of sulphate in the ocean. What's more, the timing of events in thermal (tectonic) models11 fits only loosely with that of the relevant chemical models and with the timing of the transitional atmospheric oxygenation that occurred before the GOE proper6,7,8. The specific relationships between sulphate availability in the ocean and the processes that buffer iron release from deep-sea vents are open to other interpretations14 that may challenge the authors' emphasis on thermochemical sulphate reduction and attendant H2S production. Also, there is much uncertainty about when continental growth and emergence from the ocean occurred.

Gaillard and colleagues' model falls among a growing number of oxygen-buffering scenarios, including one based on molybdenum13 that is linked to nitrogen bioavailability in the ocean and another that involves nickel15 and its role in methane production (methane can be an important oxygen sink). It may well be that many diverse processes controlled the oxygenation of the atmosphere to varying extents, and that the ties they share to the emergence of continents on a cooling Earth are anything but coincidental.


  1. 1

    Gaillard, F., Scaillet, B. & Arndt, N. T. Nature 478, 229–232 (2011).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Brocks, J. J., Logan, G. A., Buick, R. & Summons, R. E. Science 285, 1033–1036 (1999).

    CAS  Article  Google Scholar 

  3. 3

    Rasmussen, B., Fletcher, I. R., Brocks, J. J. & Kilburn, M. R. Nature 455, 1101–1104 (2008).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Rashby, S. E., Sessions, A. L., Summons, R. E. & Newman, D. K. Proc. Natl Acad. Sci. USA 104, 15099–15104 (2007).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Waldbauer, J. R., Sherman, L. S., Sumner, D. Y. & Summons, R. E. Precambr. Res. 169, 28–47 (2009).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Anbar, A. D. et al. Science 317, 1903–1906 (2007).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Kendall, B. et al. Nature Geosci. 3, 647–652 (2010).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Waldbauer, J. R., Newman, D. K. & Summons, R. E. Proc. Natl Acad. Sci. USA 108, 13409–13414 (2011).

    ADS  CAS  Article  Google Scholar 

  9. 9

    Farquhar, J., Zerkle, A. L. & Bekker, A. Photosynth. Res. 107, 11–36 (2011).

    CAS  Article  Google Scholar 

  10. 10

    Kump, L. R. & Barley, M. E. Nature 448, 1033–1036 (2007).

    ADS  CAS  Article  Google Scholar 

  11. 11

    Flament, N., Coltice, N. & Rey, P. F. Earth Planet. Sci. Lett. 275, 326–336 (2008).

    ADS  CAS  Article  Google Scholar 

  12. 12

    Halevy, I., Johnston, D. T. & Schrag, D. P. Science 329, 204–207 (2010).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Scott, C. T. et al. Geology 39, 119–122 (2011).

    ADS  CAS  Article  Google Scholar 

  14. 14

    Kump, L. R. & Seyfried, W. E. Jr Earth Planet. Sci. Lett. 235, 654–662 (2005).

    ADS  CAS  Article  Google Scholar 

  15. 15

    Konhauser, K. O. et al. Nature 458, 750–753 (2009).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Timothy W. Lyons or Christopher T. Reinhard.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lyons, T., Reinhard, C. Sea change for the rise of oxygen. Nature 478, 194–195 (2011).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing