Rapid emergence of subaerial landmasses and onset of a modern hydrologic cycle 2.5 billion years ago


The history of the growth of continental crust is uncertain, and several different models that involve a gradual, decelerating, or stepwise process have been proposed1,2,3,4. Even more uncertain is the timing and the secular trend of the emergence of most landmasses above the sea (subaerial landmasses), with estimates ranging from about one billion to three billion years ago5,6,7. The area of emerged crust influences global climate feedbacks and the supply of nutrients to the oceans8, and therefore connects Earth’s crustal evolution to surface environmental conditions9,10,11. Here we use the triple-oxygen-isotope composition of shales from all continents, spanning 3.7 billion years, to provide constraints on the emergence of continents over time. Our measurements show a stepwise total decrease of 0.08 per mille in the average triple-oxygen-isotope value of shales across the Archaean–Proterozoic boundary. We suggest that our data are best explained by a shift in the nature of water–rock interactions, from near-coastal in the Archaean era to predominantly continental in the Proterozoic, accompanied by a decrease in average surface temperatures. We propose that this shift may have coincided with the onset of a modern hydrological cycle owing to the rapid emergence of continental crust with near-modern average elevation and aerial extent roughly 2.5 billion years ago.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: Triple-oxygen-isotope systematics of ancient and modern terrestrial materials.
Fig. 2: Oxygen-isotopic compositions of shales through time.
Fig. 3: Conceptual palaeohypsometry of Archean and Proterozoic worlds.


  1. 1.

    Taylor, S. R. & McLennan, S. M. The geochemical evolution of the continental crust. Rev. Geophys. 33, 241–265 (1995).

    ADS  Article  Google Scholar 

  2. 2.

    Condie, K. C. Plate Tectonics and Crustal Evolution 3rd edn (Pergamon Press, Oxford, 2013).

  3. 3.

    Belousova, E. A. et al. The growth of the continental crust: constraints from zircon Hf-isotope data. Lithos 119, 457–466 (2010).

    ADS  Article  CAS  Google Scholar 

  4. 4.

    Keller, C. B., Schoene, B., Barboni, M., Samperton, K. M. & Husson, J. M. Volcanic–plutonic parity and the differentiation of the continental crust. Nature 523, 301–307 (2015).

    ADS  Article  PubMed  CAS  Google Scholar 

  5. 5.

    Dhuime, B., Wuestefeld, B. & Hawkesworth, C. J. Emergence of modern continental crust about 3 billion years ago. Nat. Geosci. 8, 552–555 (2015).

    ADS  Article  CAS  Google Scholar 

  6. 6.

    Lee, C.-T. A. et al. Deep mantle roots and continental emergence: implications for whole-Earth elemental cycling, long-term climate, and the Cambrian explosion. Int. Geol. Rev. 60, 431–448 (2017).

    Article  Google Scholar 

  7. 7.

    Hawkesworth, C. J., Cawood, P. A., Dhuime, B. & Kemp, A. I. S. Earth’s continental lithosphere through time. Annu. Rev. Earth Planet. Sci. 45, 169–198 (2017).

    ADS  Article  CAS  Google Scholar 

  8. 8.

    Greber, N. D. et al. Titanium isotopic evidence for felsic crust and plate tectonics 3.5 billion years ago. Science 357, 1271–1274 (2017).

    ADS  Article  PubMed  CAS  Google Scholar 

  9. 9.

    Farquhar, J. & Wing, B. A. The terrestrial record of stable sulphur isotopes: a review of the implications for evolution of Earth’s sulphur cycle. Geol. Soc. Lond. Spec. Publ. 248, 167–177 (2005).

    ADS  Article  CAS  Google Scholar 

  10. 10.

    Bekker, A. et al. Dating the rise of atmospheric oxygen. Nature 427, 117–120 (2004).

    ADS  Article  PubMed  CAS  Google Scholar 

  11. 11.

    Hoffman, P. F. The Great Oxidation and a Siderian snowball Earth: MIFS based correlation of Paleoproterozoic glacial epochs. Chem. Geol. 362, 143–156 (2013).

    ADS  Article  CAS  Google Scholar 

  12. 12.

    Gumsley, A. P. et al. Timing and tempo of the Great Oxidation Event. Proc. Natl Acad. Sci. USA 114, 1811–1816 (2017).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  13. 13.

    Valley, J. W. et al. 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contrib. Mineral. Petrol. 150, 561–580 (2005).

    ADS  Article  CAS  Google Scholar 

  14. 14.

    Bindeman, I. N., Bekker, A. & Zakharov, D. O. Oxygen isotope perspective on crustal evolution on early Earth: a record of Precambrian shales with emphasis on Paleoproterozoic glaciations and Great Oxygenation Event. Earth Planet. Sci. Lett. 437, 101–113 (2016).

    ADS  Article  CAS  Google Scholar 

  15. 15.

    Bleeker, W. The late Archean record: a puzzle in ca. 35 pieces. Lithos 71, 99–134 (2003).

    ADS  Article  CAS  Google Scholar 

  16. 16.

    Gaschnig, R. M. et al. Compositional evolution of the upper continental crust through time, as constrained by ancient glacial diamictites. Geochim. Cosmochim. Acta 186, 316–343 (2016).

    ADS  Article  CAS  Google Scholar 

  17. 17.

    Luz, B. & Barkan, E. Variations of 17O/16O and 18O/16O in meteoric waters. Geochim. Cosmochim. Acta 74, 6276–6286 (2010).

    ADS  Article  CAS  Google Scholar 

  18. 18.

    Pack, A. & Herwartz, D. The triple oxygen isotope composition of the Earth mantle and understanding δ17O variations in terrestrial rocks and minerals. Earth Planet. Sci. Lett. 390, 138–145 (2014).

    ADS  Article  CAS  Google Scholar 

  19. 19.

    Sharp, Z. D. et al. A calibration of the triple oxygen isotope fractionation in the SiO2-H2O system and applications to natural samples. Geochim. Cosmochim. Acta 186, 105–119 (2016).

    ADS  Article  CAS  Google Scholar 

  20. 20.

    Savin, S. & Epstein, S. The oxygen and hydrogen isotope geochemistry of ocean sediments and shales. Geochim. Cosmochim. Acta 34, 43–63 (1970).

    ADS  Article  CAS  Google Scholar 

  21. 21.

    Land, L. S. & Lynch, F. L. δ18O values of mudrocks: more evidence for an 18O-buffered ocean. Geochim. Cosmochim. Acta 60, 3347–3352 (1996).

    ADS  Article  CAS  Google Scholar 

  22. 22.

    Bao, H. M., Cao, X. B. & Hayles, J. Triple-oxygen isotopes: fundamental relationships and applications. Annu. Rev. Earth Planet. Sci. 44, 463–492 (2016).

    ADS  Article  CAS  Google Scholar 

  23. 23.

    Nesbitt, H. W. & Young, G. N. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature 299, 715–717 (1982).

    ADS  Article  CAS  Google Scholar 

  24. 24.

    Knauth, L. P. & Lowe, D. R. High Archean climatic temperature inferred from oxygen isotope geochemistry of cherts in the 3.5 Ga Swaziland Supergroup, South Africa. Geol. Soc. Bull. 115, 566–580 (2003).

    Article  CAS  Google Scholar 

  25. 25.

    Young, E., Yeung, L. Y. & Kohl, I. On the 17O budget of atmosphere. Geochim. Cosmochim. Acta 135, 102–125 (2014).

    ADS  Article  CAS  Google Scholar 

  26. 26.

    Mertanen, S. & Pesonen, L. J. in From the Earth’s Core to Outer Space (ed. Haapala, I.) 11–35 (Springer, Berlin, 2012).

  27. 27.

    Barley, M. E., Bekker, A. & Krapez, B. Late Archean to early Paleoproterozoic global tectonics, environmental change and the rise of atmospheric oxygen. Earth Planet. Sci. Lett. 238, 156–171 (2005).

    ADS  Article  CAS  Google Scholar 

  28. 28.

    Rowley, D. B., Pierrehumbert, R. & Currie, B. A new approach to stable isotope-based paleoaltimetry: implications for paleoaltimetry and paleohypsometry of the High Himalaya since the Late Miocene. Earth Planet. Sci. Lett. 188, 253–268 (2001).

    ADS  Article  CAS  Google Scholar 

  29. 29.

    Flament, N., Coltice, N. & Rey, P. F. A case for late-Archaean continental emergence from thermal evolution models and hypsometry. Earth Planet. Sci. Lett. 275, 326–336 (2008).

    ADS  Article  CAS  Google Scholar 

  30. 30.

    Vlaar, N. J. Continental emergence and growth on a cooling earth. Tectonophysics 322, 191–202 (2000).

    ADS  Article  Google Scholar 

  31. 31.

    Korenaga, J., Planavsky, N. J. & Evans, D. A. D. Global water cycle and the coevolution of the Earth’s interior and surface environment. Phil. Trans. R. Soc. A 375, 20150393 (2017).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  32. 32.

    Miller, M. F. Isotopic fractionation and the quantification of 17O anomalies in the oxygen three-isotope system: an appraisal and geochemical significance. Geochim. Cosmochim. Acta 66, 1881–1889 (2002).

    ADS  Article  CAS  Google Scholar 

  33. 33.

    McLennan, S. M., Hemming, S., McDaniel, D. K. & Hanson, G. N. Geochemical approaches to sedimentation, provenance and tectonics. Geol. Soc. Am. Spec. Pap. 284, 21–40 (1993).

    Google Scholar 

  34. 34.

    Frieling, J. et al. Paleocene–Eocene warming and biotic response in the epicontinental West Siberian Sea. Geology 42, 767–770 (2014).

    ADS  Article  Google Scholar 

  35. 35.

    Zheng, Y. F. Calculation of oxygen isotope fractionation in hydroxyl-bearing silicates. Earth Planet. Sci. Lett. 120, 247–263 (1993).

    ADS  Article  CAS  Google Scholar 

  36. 36.

    Ovey, C. D. Preliminary results from the study of an ocean core obtained by the Swedish Deep-Sea Expedition, 1947-48. J. Glaciol. 1, 370–373 (1950).

    ADS  Article  Google Scholar 

  37. 37.

    Kah, L. C. et al. δ13C stratigraphy of the Proterozoic Bylot Supergroup, Baffin Island, Canada: implications for regional lithostratigraphic correlations. Can. J. Earth Sci. 36, 313–332 (1999).

    ADS  Article  Google Scholar 

  38. 38.

    Campbell, A. S. & Clark, B. L. Miocene radiolarian faunas from southern California. Geol. Soc. Am. Spec. Pap. 51, 1–78 (1944).

    Google Scholar 

  39. 39.

    Retallack, G. J. Lateritization and bauxitization events. Econ. Geol. 105, 665–667 (2010).

    Article  Google Scholar 

  40. 40.

    Ryu, I.-C. Petrography, diagenesis and provenance of Eocene Tyee Basin sandstones, southern Oregon Coast Range: new view from sequence stratigraphy. Isl. Arc 12, 398–410 (2003).

    Article  Google Scholar 

  41. 41.

    McKee, E. H. & Gangloff, R. A. Stratigraphic distribution of archaeocyathids in the Silver Peak Range and the White and Inyo Mountains, western Nevada and eastern California. J. Paleontol. 43, 716–726 (1969).

    Google Scholar 

  42. 42.

    Pecoits, E. et al. U–Pb detrital zircon ages from some Neoproterozoic successions of Uruguay: provenance, stratigraphy and tectonic evolution. J. S. Am. Earth Sci. 71, 108–130 (2016).

    Article  CAS  Google Scholar 

  43. 43.

    Thomson, D., Rainbird, R. H., Planavsky, N., Lyons, T. W. & Bekker, A. Chemostratigraphy of the Shaler Supergroup, Victoria Island, NW Canada: a record of ocean composition prior to the Cryogenian glaciations. Precambr. Res. 263, 232–245 (2015).

    ADS  Article  CAS  Google Scholar 

  44. 44.

    Key, R. M. et al. The western arm of the Lufilian Arc in NW Zambia and its potential for copper mineralization. J. Afr. Earth Sci. 33, 503–528 (2001).

  45. 45.

    Podkovyrov, V. N. et al. Provenance and source rocks of Riphean sandstones in the Uchur-Maya region (east Siberia): implications of geochemical data and Sm-Nd isotopic systematics. Stratigr. Geol. Correl. 15, 41–56 (2007).

    ADS  Article  Google Scholar 

  46. 46.

    Cullers, R. L. & Podkovyrov, V. N. The source and origin of terrigenous sedimentary rocks in the Mesoproterozoic Ui group, southeastern Russia. Precambr. Res. 117, 157–183 (2002).

    ADS  Article  CAS  Google Scholar 

  47. 47.

    Vinogradov, V. I., Veis, A. F., Bujakaite, M. I. & Golovin, D. I. Isotopic evidences of epigenetic transformations and the problem of the age of Precambrian rocks in the Yudoma-Maya trough, eastern Siberia. Lithol. Miner. Resour. 35, 141–152 (2000).

    Article  Google Scholar 

  48. 48.

    Addison, W. D., Brumpton, G. R., Davis, D. W., Fralick, P. W. & Kissin, S. A. Debrisites from the Sudbury impact event in Ontario, north of Lake Superior, and a new age constraint: are they base-surge deposits or tsunami deposits? Geol. Soc. Am. Spec. Pap. 465, 245–268 (2010).

    Google Scholar 

  49. 49.

    Sheppard, S. et al. A new Paleoproterozoic tectonic history of the eastern Capricorn Orogen, Western Australia, revealed by U-Pb zircon dating of micro-tuffs. Precambr. Res. 286, 1–19 (2016).

    ADS  Article  CAS  Google Scholar 

  50. 50.

    Kipp, M. A., Stüeken, E. E., Bekker, A. & Buick, R. Selenium isotopes record extensive marine suboxia during the Great Oxidation Event. Proc. Natl Acad. Sci. USA 114, 875–880 (2017).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  51. 51.

    Rouxel, O. J., Bekker, A. & Edwards, K. J. Iron isotope constraints on the Archean and Paleoproterozoic ocean redox state. Science 307, 1088–1091 (2005).

    ADS  Article  PubMed  CAS  Google Scholar 

  52. 52.

    Rasmussen, B., Bekker, A. & Fletcher, I. R. Correlation of Paleoproterozoic glaciations based on U–Pb zircon ages for tuff beds in the Transvaal and Huronian Supergroups. Earth Planet. Sci. Lett. 382, 173–180 (2013).

    ADS  Article  CAS  Google Scholar 

  53. 53.

    Rasmussen, B., Blake, T. S. & Fletcher, I. R. U–Pb zircon age constraints on the Hamersley spherule beds: evidence for a single 2.63 Ga Jeerinah-Carawine impact ejecta layer. Geology 33, 725–728 (2005).

    ADS  Article  Google Scholar 

  54. 54.

    Hofmann, A., Bolhar, R., Dirks, P. & Jelsma, H. The geochemistry of Archaean shales derived from a mafic volcanic sequence, Belingwe greenstone belt, Zimbabwe: provenance, source area unroofing and submarine versus subaerial weathering. Geochim. Cosmochim. Acta 67, 421–440 (2003).

    ADS  Article  CAS  Google Scholar 

  55. 55.

    Planavsky, N. J. et al. Evidence for oxygenic photosynthesis half a billion years before the Great Oxidation Event. Nat. Geosci. 7, 283–286 (2014).

    ADS  Article  CAS  Google Scholar 

  56. 56.

    Kositcin, N. & Krapež, B. Relationship between detrital zircon age-spectra and the tectonic evolution of the Late Archaean Witwatersrand Basin, South Africa. Precambr. Res. 129, 141–168 (2004).

    ADS  Article  CAS  Google Scholar 

  57. 57.

    Armstrong, R. A., Compston, W., Retief, E. A., Williams, I. S. & Welke, H. J. Zircon ion microprobe studies bearing on the age and evolution of the Witwatersrand triad. Precambr. Res. 53, 243–266 (1991).

    ADS  Article  CAS  Google Scholar 

  58. 58.

    Javaux, E. J., Marshall, C. P. & Bekker, A. Organic-walled microfossils in 3.2-billion-year-old shallow-marine siliciclastic deposits. Nature 463, 934–938 (2010).

    ADS  Article  PubMed  CAS  Google Scholar 

  59. 59.

    Byerly, G. R., Kröner, A., Lowe, D. R. & Todt, W. Prolonged magmatism and time constraints for sediment deposition in the early Archean Barberton greenstone belt: evidence from the Upper Onverwacht and Fig Tree groups. Precambr. Res. 78, 125–138 (1996).

    ADS  Article  CAS  Google Scholar 

  60. 60.

    Mukhopadhyay, J. et al. Dating the oldest greenstone in India: a 3.51-Ga precise U-Pb SHRIMP zircon age for dacitic lava of the Southern Iron Ore Group, Singhbhum Craton. J. Geol. 116, 449–461 (2008).

    ADS  Article  CAS  Google Scholar 

  61. 61.

    Myers, J. S. Protoliths of the 3.8–3.7 Ga Isua greenstone belt, West Greenland. Precambr. Res. 105, 129–141 (2001).

    ADS  Article  CAS  Google Scholar 

  62. 62.

    Liu, D. Y., Nutman, A. P., Compston, W., Wu, J. S. & Shen, Q. H. Remnants of ≥3800 Ma crust in the Chinese part of the Sino-Korean craton. Geology 20, 339–342 (1992).

    ADS  Article  CAS  Google Scholar 

  63. 63.

    Retallack, G. J. & Krull, E. S. Neogene paleosols in the Sirius Group, Dominion, Antarctica. US Antarctic J. 30, 10–14 (1997).

    Google Scholar 

  64. 64.

    Bell, M. A. & Haglund, T. R. Fine-scale temporal variation of the Miocene stickleback Gasterosteus doryssus. Paleobiology 8, 282–292 (1982).

    Article  Google Scholar 

Download references


This work was supported by National Science Foundation (NSF) grant EAR1447337 and by the University of Oregon. N.D. was supported by NSF grant EAR1502591. A.B. was supported by National Sciences and Engineering Research Council (NSERC) Discovery and Accelerator grants. We thank P. Hoffman, S. Mertanen and D. Evans for discussions about Precambrian palaeogeography and environmental changes; K. Johnson for technical help with vacuum lines; and O. Melnik for help with programming.

Reviewer information

Nature thanks C. Hawkesworth and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Author information




I.N.B. conceived the study and wrote the paper; D.O.Z., J.P. and I.N.B. built the purification line and collected the data; N.D.G. and N.D. provided composite shale samples studied previously for titanium isotopes; A.B., A.H. and G.J.R. provided samples; J.S.L. performed major- and trace-element analyses; I.N.B. and N.D. discussed the inversion approach and I.N.B. implemented the model; D.O.Z. contributed to statistical treatment; A.B., N.D., N.D.G. and G.J.R. contributed to discussions on Precambrian environments and crustal evolution. All authors contributed to writing and editing the paper.

Corresponding author

Correspondence to I. N. Bindeman.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Comparison of isotopic and key elemental ratios of the shales studied here, illustrating the relative constancy of the composition of the exposed crust that is undergoing weathering, and in particular proportion of exposed mafic versus silicic rocks.

See, for example, refs 1,33. The δ49Ti data are from ref. 8, which used a large dataset that included many of the samples studied here. The elemental data are from Extended Data Table 3 and ref. 14. Variation in the composition of the exposed crust cannot explain the oxygen-isotope trends that we have identified here.

Extended Data Fig. 2 Distribution density plot for the samples studied here.

Data from Fig. 1.

Extended Data Fig. 3 Calculating the δ′18O and Δ′17O values of the weathering products of shales.

Solving a system of three unknowns (for example, the temperature of alteration and the values of Δ′17O and δ′18O along the MWL) in three equations (Supplementary Information equations (1), (5) and (6)) follows these steps: the results of Supplementary Information equations (1), (5) and (6) are substituted sequentially into each other until one unknown is left. This results in a function with respect to temperature (T, a single parameter) to solve: y = 0 = 36.323T − 23.33T2 + 0.00264T3 − 1.38 × 107. This is a third-order polynomial equation that has three roots; we solve for roots in a realistic temperature range to obtain the temperature and δ18OW. The concave blue curve originating from a point on the MWL shows isotope fractionation between shale and water with the indicated temperatures of equilibration. The grey curve represents a mixing line between detrital and weathering products of the indicated proportions, computed using the CIA.

Extended Data Fig. 4 Evaluating the sensitivity of reconstructed δ18OW and temperature values to variations in the input parameters.

We show here the effects of varying the proportion of quartz (Q) in shales (±20%), the CIA (±10 units), and Δ′17O (D170) (±0.01‰), the latter occurring naturally owing to, for instance, silicification.

Extended Data Fig. 5 Calculated temperature, δ18OW and Δ17OW values plotted against age of shales.

a, δ18OW versus age; b, temperature versus age; c, Δ17OW versus age; d, δ18OW versus temperature. These values are based on the solution of three equations with three input parameters: δ′18Oshale, δ′17Oshale and the proportion of quartz (see Extended Data Fig. 3). Sensitivity analysis is provided in Extended Data Fig. 4. Most measurements yielded solved roots (plus signs within symbols); when equations could not be solved, small variations in the input parameters (Extended Data Table 1) allowed us to find roots. In particular, correcting for secondary silicification (see Extended Data Table 1) by decreasing Δ′17Oshale by 0.01‰ to 0.08‰ along the silicification line allowed us to find roots in all but two cases. Note that the overall calculated δ18OW and temperature ranges agree with modern and recent values for surface, diagenetic, basinal and pore waters measured in drillholes21. Note also that the absolute calculated values for the temperature of water–rock interaction (weathering) and δ18OW depend on the assumed isotopic fractionations in Extended Data Fig. 6; however, given that these values solved within realistic bounds, the fractionations of ref. 19 and the MWL defined in ref. 17 are probably well constrained and calibrated in absolute triple-oxygen-isotope space. The lowest δ18OW and temperature are computed for recent clay samples from Antarctica, and for 2.5–2.2-Gyr-old synglacial Palaeoproterozoic shales, confirming the participation of low-δ18Ow synglacial waters in diagenesis, as proposed previously14. The highest recent temperature and δ18OW values are for Palaeocene–Eocene (55-million-year-old) thermal maximum shales (ref. 34 and Extended Data Table 1). eg, Interquartile range statistics and running averages for the parameters computed in ad.

Extended Data Fig. 6 Quartz/water, illite/water and bulk-shale/water 18O/16O fractionation factors.

The quartz/water and illite/water 1,000lnα (18O/16O) fractionation factors are based on refs 19,35. The bulk-shale/water 1,000lnα (18O/16O) fractionation factors are based on the assumption that bulk shale comprises 70% illite and 30% quartz (that is, Q = 0.3). The blue line (for quartz/water fractionation) corresponds to Supplementary Information equation (2) and equation (9) in ref. 19. The green line (for illite/water fractionation) is the best-fit second-order polynomial with two fit coefficients based on the equation in ref. 35 that includes three fit coefficients. We used these coefficients to solve equations for bulk-shale/water triple-oxygen-isotope fractionations according to the proportion of quartz (determined through X-ray-diffraction; Extended Data Table 2).

Extended Data Fig. 7 Chemical index of alteration (CIA) plotted against time.

CIA = Al2O3/(Al2O3 + CaO + Na2O + K2O)mol. Data are taken from this work (triangles), ref. 14 (diamonds), and literature data (circles).

Extended Data Table 1 Triple-oxygen-isotope analyses of the shale samples used here
Extended Data Table 2 Quantitative XRD data for a selection of studied shales
Extended Data Table 3 New XRF analyses of samples studied here

Supplementary information

Supplementary Information

This file contains Supplementary Methods, Data Availability Statement and Supplementary References.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bindeman, I.N., Zakharov, D.O., Palandri, J. et al. Rapid emergence of subaerial landmasses and onset of a modern hydrologic cycle 2.5 billion years ago. Nature 557, 545–548 (2018). https://doi.org/10.1038/s41586-018-0131-1

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.