Running out of gas: Zircon 18O-Hf-U/Pb evidence for Snowball Earth preconditioned by low degassing
Affiliations | Corresponding Author | Cite as | Funding information
- Share this article
Article views:12,683Cumulative count of HTML views and PDF downloads.
- Download Citation
- Rights & Permissions
Abstract
Figures and Tables
 Figure 1 Schematic diagram of the age evolution of a zircon grain, showing how combined Hf and U/Pb isotope information can be used to determine the time window (∆THf-U/Pb) during which the δ18Ozircon value is acquired, based on the conceptual model proposed here. The Hf model age constrains the time that parent material separated from the mantle, beginning its journey of reworking via weathering, erosion, and incorporation into melts. This journey ends when the zircon crystallises, as represented by its U/Pb age. Longer ∆THf-U/Pb windows may be affected by mixing and make it difficult to identify the timing when the δ18Ozircon signature was acquired. We screen zircons by ∆THf-U/Pb in order to calculate the time evolution of δ18Ozircon. |  Figure 2 Time evolution of δ18Ozircon calculated for different values of ∆THf-U/Pb as selection criterion (see also Supplementary Video S-1). The very short values of ∆THf-U/Pb restrict the total number of grains in the dataset to a small number (<1000 grains if ∆THf-U/Pb <300 Ma using the New Crust Hf model age, shown here), leading to large uncertainties, although some of the first order features are still evident (e.g., a maximum ~2 Ga and minimum ~0.8 Ga). For longer ∆THf-U/Pb windows, the uncertainties are much reduced and general patterns are stable across a range of window sizes. For very large ∆THf-U/Pb (e.g., >1000 Ma; see Supplementary Video S-1) the δ18Ozircon curve represents a long-term accumulated signal that is difficult to tie to specific geologic events because of the long integration window. The grey band represents ±2 s.d. (s.d.: standard deviation of the mean). Blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The green bar covers the period of the Lomagundi event. Also see Figure 3. |  Figure 3 δ18Ozircon calculated for ∆THf-U/Pb <400 Ma, along with timing of major carbon cycle anomalies observed in the geologic record. The grey band represents ±2 s.d. (s.d.: standard deviation of the mean). Red bars indicate periods of supercontinent assembly (Cawood et al., 2013). Dark blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The dark green bar covers the period of the Lomagundi event. Note that the δ18Ozircon values are plotted using the midpoint age of the ∆THf-U/Pb window, but the actual δ18Ozircon values integrate processes (e.g., weathering conditions) occurring over longer periods of time, characterised by a mean ∆THf-U/Pb duration of 225 Ma. To illustrate the effect on the comparison between the timing of geological events and the δ18Ozircon curve, lighter shaded areas around periods of geological events reflect uncertainty of the δ18Ozircon curve associated with plotting the curve for the midpoint between the Hf and U/Pb ages, given as half of the mean ∆THf-U/Pb (112.5 Ma). |
| Figure 1 | Figure 2 | Figure 3 |
Supplementary Figures and Tables
 Table S-1 New data sources with coupled δ18O, U/Pb age and Hf model age information in addition to Dhuime et al. (2012). |  Figure S-1 Results as in Figure 2 in the main text, but using the traditional TDM Hf model age instead of the New Crust Hf model age. Time evolution of δ18Ozircon calculated for different values of ∆THf-U/Pb as selection criterion (see also Supplementary Video S-2). The very short values of the selection criterion ∆THf-U/Pb restrict the total number of grains in the dataset to a small number (<400 grains if ∆THf-U/Pb <300 Ma and using the TDM Hf model age). Blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The green bar covers the period of the Lomagundi event. |  Figure S-2 Results as in Figure 3 in the main text, but using the traditional TDM Hf model age instead of the New Crust Hf model age. Note that, due the difference in the calculation procedure, the New Crust Hf model age for single grains provides on average 150 Ma shorter ages than the TDM Hf model age. Dark blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The dark green bar covers the period of the Lomagundi event. To illustrate the effect on comparison between the timing of geological events and the δ18Ozircon curve, lighter shaded areas around periods of geological events reflect uncertainty of the δ18Ozircon curve associated with plotting the curve for the midpoint between the Hf and U/Pb ages, given as half of the mean ∆THf-U/Pb (140 Ma). |  Figure S-3 Results as in Figure 3 in the main text, but with analysis of variance in ages for the population of zircons with ∆THf-U/Pb <400 Ma. The black line shows δ18Ozircon using the weighted mean zircon age of the selected data, when age is weighted using the same Gaussian filtering procedure as applied to the δ18Ozircon data in Figure 3 in the main text. As expected, the weighted age (black line) plots very similarly to the actual ages used in Figure 3. The dashed blue and red lines show δ18Ozircon plotted against time considering variance in the ages of the zircon population, specifically the weighted ±1 standard deviation of the age using the same Gaussian filtering as applied to the δ18Ozircon data for the midpoint ages (blue: minus 1 s.d.; red: plus 1 s.d.). Weighted variance in ages ranges from 66 to 130 Ma and does not change systematically over time, so we consider the single average value reported in the text (225 Ma) as representative of the time resolution of the record for a ∆THf-U/Pb window <400 Ma. |
| Table S-1 | Figure S-1 | Figure S-2 | Figure S-3 |
 Video S-1 This supplementary video shows the evolution of the δ18O time series with increasing ∆THf-U/Pb time window as selection filter for the New Crust Hf model age. The video also provides for each time step the number of included data points and the mean ∆THf-U/Pb for data points fulfilling the selection criterion. | Video S-2 This supplementary video shows the evolution of the δ18O time series with increasing ∆THf-U/Pb time window as selection filter for the traditional TDM Hf model age. The video also provides for each time step the number of included data points and the mean ∆THf-U/Pb for data points fulfilling the selection criterion. | Table S-2 Applied data including δ18O, hafnium and lutetium isotopic data, the uranium-lead-isotope age and references. (a) New compiled data. (b) Dhuime et al. (2012) data. |
| Video S-1 | Video S-2 | Table S-2 |
top
Letter
Over geologic time, Earth’s climate is regulated by the fluxes of carbon to and from the coupled ocean-atmosphere system, specifically by the balance between solid Earth degassing, the consumption and release of CO2 via rock weathering, and the burial and oxidation of organic carbon (Berner, 2004
Berner, R.A. (2004) The Phanerozoic Carbon Cycle: CO2 and O2. Oxford University Press, New York.
). For most of Earth’s history, the sources and sinks of atmospheric CO2 have remained in close balance, maintaining the planet’s equable climate (Walker et al., 1983Walker, J.C.G., Klein, C., Schidlowski, M., Schopf, J.W., Stevenson, D.J., Walter, M.R. (1983) Environmental evolution of the Archean-early Proterozoic Earth. In: Schopf, J.W. (Ed.) Earth's earliest biosphere: Its origin and evolution (A84-43051 21-51). Princeton University Press, Princeton, New Jersey, 260–290.
; Berner and Caldeira, 1997Berner, R.A., Caldeira, K. (1997) The need for mass balance and feedback in the geochemical carbon cycle. Geology 25, 955–956.
). The large carbon cycle anomalies observed in the sedimentary record of the Archean and Proterozoic eras represent puzzling deviations from this general stability (Hoffman et al., 1998Hoffman, P.F., Kaufman, A.J., Halverson, G.P., Schrag, D.P. (1998) A Neoproterozoic snowball earth. Science 281, 1342–1346.
; Goddéris et al., 2003Goddéris, Y., Donnadieu, Y., Nédélec, A., Dupré, B., Dessert, C., Grard, A., Ramstein, G., François, L.M. (2003) The Sturtian ‘snowball’ glaciation: fire and ice. Earth and Planetary Science Letters 211, 1–12.
; Bekker and Holland, 2012Bekker, A., Holland, H.D. (2012) Oxygen overshoot and recovery during the early Paleoproterozoic. Earth and Planetary Science Letters 317–318, 295–304.
; Lyons et al., 2014Lyons, T.W., Reinhard, C.T., Planavsky, N.J. (2014) The rise of oxygen in Earth's early ocean and atmosphere. Nature 506, 307–315.
). Understanding the causes of these events requires information about the carbon cycle at the time, which remains scarce. Geochemical evidence from marine sediments, such as records of 87Sr/86Sr or δ13C, is subject to multiple interpretations (e.g., Kump, 1989Kump, L.R. (1989) Alternative modeling approaches to the geochemical cycles of carbon, sulfur, and strontium isotopes. American Journal of Science 289, 390–410.
; Goddéris et al., 2017Goddéris, Y., Le Hir, G., Macouin, M., Donnadieu, Y., Hubert-Théou, L., Dera, G., Aretz, M., Fluteau, F., Li, Z.X., Halverson, G.P. (2017) Paleogeographic forcing of the strontium isotopic cycle in the Neoproterozoic. Gondwana Research 42, 151–162.
). For example, the evolution of seawater 87Sr/86Sr over geological time (McArthur et al., 2012McArthur, J.M., Howarth, R.J., Shields, G.A. (2012) Sr isotope time series. In: Gradstein, F., Ogg, J., Schmitz, M., Ogg, G. (Eds.) The Geological Timescale 2012. Elsevier, Oxford, Amsterdam, Waltham, 127–144.
) may be influenced by changes in degassing and hydrothermal activity, continental weatherability, and the isotopic composition of rocks undergoing weathering (Kump, 1989Kump, L.R. (1989) Alternative modeling approaches to the geochemical cycles of carbon, sulfur, and strontium isotopes. American Journal of Science 289, 390–410.
). Without independent constraints, marine isotopic records cannot distinguish between these possibilities, each with different implications for the carbon cycle.Zircons may provide information about carbon cycle fluxes that is independent of the multiple factors affecting seawater composition. The Hf isotopic composition of a zircon reflects the time when the grain’s unmixed parent material separated from the depleted mantle reservoir (Griffin et al., 2002
Griffin, W.L., Wang, X., Jackson, S.E., Pearson, N.J., O'Reilly, S.Y., Xu, X., Zhou, X. (2002) Zircon chemistry and magma mixing, SE China: in-situ analysis of Hf isotopes, Tonglu and Pingtan igneous complexes. Lithos 61, 237–269.
; Hawkesworth and Kemp, 2006Hawkesworth, C.J., Kemp, A.I.S. (2006) Using hafnium and oxygen isotopes in zircons to unravel the record of crustal evolution. Chemical Geology 226, 144–162.
). In contrast, the U/Pb age of a zircon records the last time the mineral experienced temperatures above the closure threshold for the Pb system (>1000 °C), and thus the crystallisation age (Mezger and Krogstad, 1997Mezger, K., Krogstad, E.J. (1997) Interpretation of discordant U‐Pb zircon ages: An evaluation. Journal of metamorphic Geology 15, 127–140.
). The difference between the Hf and U/Pb ages, here termed ∆THf-U/Pb, represents the time interval during which zircon parent material could have been affected by weathering and magmatic processes (Fig. 1), leaving an imprint on δ18Ozircon (Valley et al., 2005Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S. (2005) 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contributions to Mineralogy and Petrology 150, 561–580.
). The contribution from even small amounts of altered crustal material can increase δ18Ozircon values above the primary mantle signature of ~5.3 ± 0.3 ‰, because altered crust is isotopically enriched as a result of weathering by meteoric fluids (Savin and Epstein, 1970Savin, S.M., Epstein, S. (1970) The oxygen and hydrogen isotope geochemistry of clay minerals. Geochimica et Cosmochimica Acta 84, 25–42.
; Gregory and Taylor, 1981Gregory, R.T., Taylor, H.P. (1981) An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophiolite, Oman: Evidence for δ18O buffering of the oceans by deep (> 5 km) seawater‐hydrothermal circulation at mid-ocean ridges. Journal of Geophysical Research: Solid Earth 86, 2737–2755.
; Bindeman et al., 2016Bindeman, I.N., Bekker, A., Zakharov, D.O. (2016) Oxygen isotope perspective on crustal evolution on early Earth: A record of Precambrian shales with emphasis on Paleoproterozoic glaciations and Great Oxygenation Event. Earth and Planetary Science Letters 437, 101–113.
). The extent of isotopic enrichment for a given zircon grain will depend on the amount of weathering-related contamination of the parent magma, which is related to the alteration state of the crustal material (Valley et al., 2005Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S. (2005) 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contributions to Mineralogy and Petrology 150, 561–580.
). More intense chemical weathering and clay formation on the continents will produce more weathered sedimentary material, with heavier δ18O, increasing the potential for elevating δ18Ozircon (Valley et al., 2005Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S. (2005) 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contributions to Mineralogy and Petrology 150, 561–580.
; Payne et al., 2015Payne, J.L., Hand, M., Pearson, N.J., Barovich, K.M., McInerney, D.J. (2015) Crustal thickening and clay: Controls on O isotope variation in global magmatism and siliciclastic sedimentary rocks. Earth and Planetary Science Letters 412, 70–76.
; Bindeman et al., 2016Bindeman, I.N., Bekker, A., Zakharov, D.O. (2016) Oxygen isotope perspective on crustal evolution on early Earth: A record of Precambrian shales with emphasis on Paleoproterozoic glaciations and Great Oxygenation Event. Earth and Planetary Science Letters 437, 101–113.
). Thus, a grain’s δ18Ozircon value should preserve some information about the weathering conditions during that grain’s ∆THf-U/Pb time interval. While petrologic processes and magma chamber conditions such as oxygen fugacity may also influence δ18Ozircon by changing fractionation factors (Valley, 2003Valley, J.W. (2003) Oxygen isotopes in zircon. Reviews in Mineralogy and Geochemistry 53, 343–385.
), we propose that, when averaged over grains from around the world, δ18Ozircon reflects globally averaged weathering conditions during a given ∆THf-U/Pb interval.Figure 1 Schematic diagram of the age evolution of a zircon grain, showing how combined Hf and U/Pb isotope information can be used to determine the time window (∆THf-U/Pb) during which the δ18Ozircon value is acquired, based on the conceptual model proposed here. The Hf model age constrains the time that parent material separated from the mantle, beginning its journey of reworking via weathering, erosion, and incorporation into melts. This journey ends when the zircon crystallises, as represented by its U/Pb age. Longer ∆THf-U/Pb windows may be affected by mixing and make it difficult to identify the timing when the δ18Ozircon signature was acquired. We screen zircons by ∆THf-U/Pb in order to calculate the time evolution of δ18Ozircon.
We have compiled published δ18O-Hf-U/Pb zircon data from predominantly detrital zircons – including in total 4444 grains with coupled information for all three isotope systems (i.e. all three analyses made on the same zircon crystal). These data span 4.4 Ga of Earth’s history. We have calculated ∆THf-U/Pb for each grain and evaluated the δ18O of the zircons that fall within a given ∆THf-U/Pb window. We argue that shorter windows are most likely to provide information about crustal weathering conditions during a given ∆THf-U/Pb window. Long windows may incorporate mixing of material with different Hf-derived mantle extraction ages (Fig. 1) and also reflect the accumulated effects of reworking over long periods of time, potentially subject to long-term tectonic controls on the extent of low temperature contribution to the δ18O signature. Even for short ∆THf-U/Pb, any individual zircon’s δ18O will be influenced by the tectonic and petrologic history of that grain’s parent magma, which is likely specific to its regional origin. Thus, when plotting data from all individual grains, a very large spread is observed in δ18Ozircon values for any given U/Pb crystallisation age (Valley et al., 2005
Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S. (2005) 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contributions to Mineralogy and Petrology 150, 561–580.
; Payne et al., 2015Payne, J.L., Hand, M., Pearson, N.J., Barovich, K.M., McInerney, D.J. (2015) Crustal thickening and clay: Controls on O isotope variation in global magmatism and siliciclastic sedimentary rocks. Earth and Planetary Science Letters 412, 70–76.
, 2016Payne, J.L., McInerney, D.J., Barovich, K.M., Kirkland, C.L., Pearson, N.J., Hand, M. (2016) Strengths and limitations of zircon Lu-Hf and O isotopes in modelling crustal growth. Lithos 248–251, 175–192.
). By aggregating many hundreds to thousands of grains from detrital sources, we seek to extract the average conditions that are most likely to represent the state of crustal alteration for defined time slices, rather than the specific history of individual zircons.We have calculated a moving average of δ18Ozircon over time, using a Gaussian filter. Specifically, for each time t, beginning at 4 Ga and proceeding in 50 Ma increments, we begin by calculating a weighting factor for all zircon data points. The weighting factor is determined by each zircon’s age with respect to the normal distribution centred at time t, with standard deviation of 100 Ma, thus giving zircons closer to t a higher weight and effectively giving zero weight to zircons more than a few hundred million years from t. For zircon age, we average the Hf model age and the U-Pb age of each grain, to reflect the midpoint of the assumed processing time during which the grain’s parent material acquired its δ18O composition. We estimate a representative δ18Ozircon value for time t by calculating a mean weighted δ18Ozircon for all data points. The standard deviation of the mean δ18Ozircon for time t is similarly calculated with the Gaussian weighting (±2 standard deviation shown by grey envelopes in the Figures and the Supplementary Videos).
We include only zircons that fall within a given ∆THf-U/Pb window, while varying the size of the ∆THf-U/Pb window from 50 Ma to 1500 Ma (see Supplementary Videos, where the number of grains considered and the mean ∆THf-U/Pb for each time window is reported). Figure 2 shows examples, with average zircon δ18O plotted versus time for ∆THf-U/Pb ranging from 50 to 450 Ma. For low ∆THf-U/Pb (<~300 Ma) the number of total data points is low (<1000 grains total), leading to larger uncertainties (Fig. 2). There is thus a tradeoff between focusing on short time windows (over which there are too few data from zircon grains to provide a conclusive picture) versus longer time windows (which average over longer periods of time, reducing the temporal resolution for relating to specific geologic events). For ∆THf-U/Pb of 350–500 Ma, uncertainties are reduced and distinct patterns emerge in the data, with first order minima and maxima that remain stable across a wide range of ∆THf-U/Pb. This stability suggests these features are not artefacts of the data analysis. In the discussion that follows, we focus on the time-evolution of δ18Ozircon calculated for grains with ∆THf-U/Pb < 400 Ma, which is representative of the first order patterns and still sufficiently short to relate to major changes in the environmental state of the Earth system. We primarily report results based on the New Crust Hf age model (Dhuime et al., 2011
Dhuime, B., Hawkesworth, C., Cawood, P. (2011) When continents formed. Science 331, 154–155.
; Payne et al., 2016Payne, J.L., McInerney, D.J., Barovich, K.M., Kirkland, C.L., Pearson, N.J., Hand, M. (2016) Strengths and limitations of zircon Lu-Hf and O isotopes in modelling crustal growth. Lithos 248–251, 175–192.
). For comparison, the same calculations are also presented in the Supplementary Information (Figs. S-1 and S-2) using the traditional TDM Hf model age, revealing basically the same picture.Figure 2 Time evolution of δ18Ozircon calculated for different values of ∆THf-U/Pb as selection criterion (see also Supplementary Video S-1). The very short values of ∆THf-U/Pb restrict the total number of grains in the dataset to a small number (<1000 grains if ∆THf-U/Pb <300 Ma using the New Crust Hf model age, shown here), leading to large uncertainties, although some of the first order features are still evident (e.g., a maximum ~2 Ga and minimum ~0.8 Ga). For longer ∆THf-U/Pb windows, the uncertainties are much reduced and general patterns are stable across a range of window sizes. For very large ∆THf-U/Pb (e.g., >1000 Ma; see Supplementary Video S-1) the δ18Ozircon curve represents a long-term accumulated signal that is difficult to tie to specific geologic events because of the long integration window. The grey band represents ±2 s.d. (s.d.: standard deviation of the mean). Blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The green bar covers the period of the Lomagundi event. Also see Figure 3.
In comparing the trends in δ18Ozircon with the timing of geological events (Fig. 3), it is important to note that the δ18Ozircon values aggregate effects over a time interval between the Hf model age and the U-Pb age. The timescale in Figure 3 represents average values focused at the midpoint of this interval (see above), but the weathering signature is likely to have been acquired over a longer period of time, so each time point on Figure 3 actually reflects a range of time. The size of the prescribed ∆THf-U/Pb selection time window puts a maximum limit on this duration (in the case of Fig. 3, 400 Ma). In practice, the actual ∆THf-U/Pb values for individual grains are often shorter than the maximum prescribed window. In the case of Figure 3, the mean ∆THf-U/Pb is 225 Ma. This reflects the typical time over which the effects of weathering are expected to have left an imprint in the δ18Ozircon, and thus a characteristic uncertainty on the time axis of Figure 3. To facilitate comparison, this duration is illustrated by the lighter shading on the timing of geological events; this shading does not reflect uncertainty on the timing of these events themselves but rather reflects the temporal resolution of the δ18Ozircon curve. The mean ∆THf-U/Pb for zircons actually varies somewhat as a function of time. Supplementary Information Figure S-3 shows δ18Ozircon curves accounting for this variability by calculating the weighted standard deviation of the age at each time t. As expected, the variance in age (2 standard deviation) is ~0.1–0.2 Ga and is similar over the 4 Ga record, indicating that the single average value of 225 Ma in Figure 3, while a simplification, captures the characteristic uncertainty in the time dimension.
Figure 3 δ18Ozircon calculated for ∆THf-U/Pb <400 Ma, along with timing of major carbon cycle anomalies observed in the geologic record. The grey band represents ±2 s.d. (s.d.: standard deviation of the mean). Red bars indicate periods of supercontinent assembly (Cawood et al., 2013
Cawood, P.A., Hawkesworth, C.J., Dhuime, B. (2013) The continental record and the generation of continental crust. GSA Bulletin 125, 14–32.
). Dark blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The dark green bar covers the period of the Lomagundi event. Note that the δ18Ozircon values are plotted using the midpoint age of the ∆THf-U/Pb window, but the actual δ18Ozircon values integrate processes (e.g., weathering conditions) occurring over longer periods of time, characterised by a mean ∆THf-U/Pb duration of 225 Ma. To illustrate the effect on the comparison between the timing of geological events and the δ18Ozircon curve, lighter shaded areas around periods of geological events reflect uncertainty of the δ18Ozircon curve associated with plotting the curve for the midpoint between the Hf and U/Pb ages, given as half of the mean ∆THf-U/Pb (112.5 Ma).In common with prior studies, we see relatively little variability in δ18Ozircon until ~2.5 Ga, considering the patterns of long-term evolution over 4 Ga (Valley et al., 2005
Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S. (2005) 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contributions to Mineralogy and Petrology 150, 561–580.
; Payne et al., 2015Payne, J.L., Hand, M., Pearson, N.J., Barovich, K.M., McInerney, D.J. (2015) Crustal thickening and clay: Controls on O isotope variation in global magmatism and siliciclastic sedimentary rocks. Earth and Planetary Science Letters 412, 70–76.
). After 2.5 Ga, we observe that the minimum and maximum in the δ18Ozircon record coincide with two of the most pronounced carbon cycle perturbations in the geologic record. Specifically, the rise towards the highest δ18Ozircon in the record coincides with the Lomagundi carbon isotope excursion (ca. 2.22–2.07 Ga) (Bekker and Holland, 2012Bekker, A., Holland, H.D. (2012) Oxygen overshoot and recovery during the early Paleoproterozoic. Earth and Planetary Science Letters 317–318, 295–304.
), while the lowest δ18Ozircon in the record corresponds to the Neoproterozoic Snowball Earth events and the framing Kaigas, Gaskiers, and Vingerbreek glaciations (ca. 0.75–0.55 Ga) (Germs and Gaucher, 2012Germs, G.J.B., Gaucher, C. (2012) Nature and extent of a late Ediacaran (ca. 547 Ma) glacigenic erosion surface in southern Africa. South African Journal of Geology 115, 91–102.
; Hofmann et al., 2015Hofmann, M., Linnemann, U., Hoffmann, K.-H., Germs, G., Gerdes, A., Marko, L., Eckelmann, K., Gärtner, A., Krause, R. (2015) The four Neoproterozoic glaciations of southern Namibia and their detrital zircon record: The fingerprints of four crustal growth events during two supercontinent cycles. Precambrian Research 259, 176–188.
). Moreover, the Archean global glaciations (ca. 2.45–2.24 Ga) (Gumsley et al., 2017Gumsley, A.P., Chamberlain, K.R., Bleeker, W., Soderlund, U., de Kock, M.O., Larsson, E.R., Bekker, A. (2017) Timing and tempo of the Great Oxidation Event. Proc Natl Acad Sci U S A 114, 1811–1816.
) also coincide with a local minimum in δ18Ozircon. Periods of supercontinent assembly (Cawood et al., 2013Cawood, P.A., Hawkesworth, C.J., Dhuime, B. (2013) The continental record and the generation of continental crust. GSA Bulletin 125, 14–32.
) appear to occur at similar times to peaks in δ18Ozircon (Fig. 3) and mature supercontinents with periods of lower δ18Ozircon, except during Gondwana-Pangea formation and breakup of the last 0.7 Ga.The Lomagundi carbon isotope excursion (Karhu and Holland, 1996
Karhu, J.A., Holland, H.D. (1996) Carbon isotopes and the rise of atmospheric oxygen. Geology 24, 867–870.
; Bekker and Holland, 2012Bekker, A., Holland, H.D. (2012) Oxygen overshoot and recovery during the early Paleoproterozoic. Earth and Planetary Science Letters 317–318, 295–304.
; Lyons et al., 2014Lyons, T.W., Reinhard, C.T., Planavsky, N.J. (2014) The rise of oxygen in Earth's early ocean and atmosphere. Nature 506, 307–315.
) occurs as δ18Ozircon rises towards the highest values in the geological record (Fig. 3), suggesting the possibility of intense weathering at the time. High weathering intensity may have resulted from high mantle degassing due to vigorous convection (Condie et al., 2001Condie, K.C., Des Marais, D.J., Abbott, D. (2001) Precambrian superplumes and supercontinents: a record in black shales, carbon isotopes, and paleoclimates? Precambrian Research 106, 239–260.
, 2016Condie, K.C., Aster, R.C., van Hunen, J. (2016) A great thermal divergence in the mantle beginning 2.5 Ga: Geochemical constraints from greenstone basalts and komatiites. Geoscience Frontiers 7, 543–553.
; Grenholm and Scherstén, 2015Grenholm, M., Scherstén, A. (2015) A hypothesis for Proterozoic-Phanerozoic supercontinent cyclicity, with implications for mantle convection, plate tectonics and Earth system evolution. Tectonophysics 662, 434–453.
), or to the onset of oxidative weathering following the rise of atmospheric O2 (Bekker and Holland, 2012Bekker, A., Holland, H.D. (2012) Oxygen overshoot and recovery during the early Paleoproterozoic. Earth and Planetary Science Letters 317–318, 295–304.
; Planavsky et al., 2012Planavsky, N.J., Bekker, A., Hofmann, A., Owens, J.D., Lyons, T.W. (2012) Sulfur record of rising and falling marine oxygen and sulfate levels during the Lomagundi event. Proceedings of the National Academy of Sciences 109, 18300–18305.
). Abundant Al-rich shales and quartz-rich sandstones at the time have been cited as evidence for intense weathering conditions (Bekker, 2014Bekker, A. (2014) Lomagundi Carbon Isotope Excursion. In: Amils, R., Gargaud, M., Cernicharo Quintalla, J., Cleaves, H.J., Irvine, W.M., Pinti, D., Viso, M. (Eds.) Encyclopedia of Astrobiology. Springer, Berlin, Heidelberg, 1–6.
), and the extremely enriched δ13C values in carbonates have been explained by large amounts of organic carbon burial, resulting from enhanced biological productivity facilitated by the release of phosphorous (a limiting nutrient) during rock weathering (Bekker and Holland, 2012Bekker, A., Holland, H.D. (2012) Oxygen overshoot and recovery during the early Paleoproterozoic. Earth and Planetary Science Letters 317–318, 295–304.
; Harada et al., 2015Harada, M., Tajika, E., Sekine, Y. (2015) Transition to an oxygen-rich atmosphere with an extensive overshoot triggered by the Paleoproterozoic snowball Earth. Earth and Planetary Science Letters 419, 178–186.
). The zircon record is consistent with this hypothesis, lending some confidence to interpretation of δ18Ozircon as reflecting global weathering conditions.The minimum in δ18Ozircon coinciding with the beginning of the Neoproterozoic Snowball Earth events points to a decreased imprint of weathered crust in parent magmas at these times. We suggest this signal is indicative of low weathering intensity on the continents. In simplistic terms, if glaciations were caused by drawdown of atmospheric CO2 via enhanced silicate weathering, as suggested in prior studies (Goddéris et al., 2003
Goddéris, Y., Donnadieu, Y., Nédélec, A., Dupré, B., Dessert, C., Grard, A., Ramstein, G., François, L.M. (2003) The Sturtian ‘snowball’ glaciation: fire and ice. Earth and Planetary Science Letters 211, 1–12.
; Donnadieu et al., 2004Donnadieu, Y., Goddéris, Y., Ramstein, G., Nédélec, A., Meert, J. (2004) A ‘snowball Earth’ climate triggered by continental break-up through changes in runoff. Nature 428, 303–306.
; Pierrehumbert et al., 2011Pierrehumbert, R.T., Abbot, D.S., Voigt, A., Koll, D. (2011) Climate of the Neoproterozoic. Annual Review of Earth and Planetary Sciences 39, 417–460.
), the opposite would be expected, namely higher weathering intensity and thus elevated δ18Ozircon immediately preceding and during glaciation. Global weathering fluxes should balance degassing fluxes over timescales >1 Ma, because of climate-dependent weathering feedbacks (e.g., Berner and Caldeira, 1997Berner, R.A., Caldeira, K. (1997) The need for mass balance and feedback in the geochemical carbon cycle. Geology 25, 955–956.
). Thus, we propose that the minima in δ18Ozircon reflect times of low solid Earth CO2 degassing, which preconditioned the Earth system for glaciation by forcing a state of low atmospheric pCO2. Low degassing flux has similarly been suggested as an underlying cause of icehouse climate states based on compilations of zircon abundance over the past ~0.7 Ga (McKenzie et al., 2016McKenzie, N.R., Horton, B.K., Loomis, S.E., Stockli, D.F., Planavsky, N.J., Lee, C.-T.A. (2016) Continental arc volcanism as the principal driver of icehouse-greenhouse variability. Science 352, 444–447.
).The data presented here average over long periods of time (100s of millions of years to yield a sufficient number of grains in each time interval to extract a robust pattern). As a consequence, the results do not preclude an immediate triggering of glaciation by changes in weathering fluxes and CO2 drawdown, for example by enhanced weathering of continental crust, large igneous provinces, or submarine basalt (Goddéris et al., 2003
Goddéris, Y., Donnadieu, Y., Nédélec, A., Dupré, B., Dessert, C., Grard, A., Ramstein, G., François, L.M. (2003) The Sturtian ‘snowball’ glaciation: fire and ice. Earth and Planetary Science Letters 211, 1–12.
; Donnadieu et al., 2004Donnadieu, Y., Goddéris, Y., Ramstein, G., Nédélec, A., Meert, J. (2004) A ‘snowball Earth’ climate triggered by continental break-up through changes in runoff. Nature 428, 303–306.
; Gernon et al., 2016Gernon, T.M., Hincks, T.K., Tyrrell, T., Rohling, E.J., Palmer, M.R. (2016) Snowball Earth ocean chemistry driven by extensive ridge volcanism during Rodinia breakup. Nature Geoscience 9, 242–248.
). Relatively low background levels of atmospheric pCO2, resulting from low degassing fluxes, could have set up the system for the onset of Snowball Earth, which then occurred as an immediate result of enhanced weathering events without requiring anomalously intense weathering, since pCO2 can be more easily reduced to the levels required for runaway glaciation when the background concentrations are low. Thus, we suggest that enhanced weathering, as may have occurred during supercontinent breakup, might have been the immediate trigger, but that a long-term trajectory of low degassing set the stage for globally extensive glaciation.The subsequent increase in δ18Ozircon as glaciations continued (i.e. leading up to 0.5 Ga) might have been influenced, at least in part, by intense weathering after glaciations (e.g., Hoffman et al., 1998
Hoffman, P.F., Kaufman, A.J., Halverson, G.P., Schrag, D.P. (1998) A Neoproterozoic snowball earth. Science 281, 1342–1346.
) and submarine basalt weathering (Gernon et al., 2016Gernon, T.M., Hincks, T.K., Tyrrell, T., Rohling, E.J., Palmer, M.R. (2016) Snowball Earth ocean chemistry driven by extensive ridge volcanism during Rodinia breakup. Nature Geoscience 9, 242–248.
).Whether similar conditions prevailed at the time of the earlier Archean glaciations remains unclear from the zircon data (e.g., see suggested mechanisms for the onset of Palaeoproterozoic glaciations; Teitler et al., 2014
Teitler, Y., Le Hir, G., Fluteau, F., Philippot, P., Donnadieu, Y. (2014) Investigating the Paleoproterozoic glaciations with 3-D climate modeling. Earth and Planetary Science Letters 395, 71–80.
), and we emphasise that the long integration window of the zircon data preclude interpretation of trends in Figure 3 in terms of individual episodes of glaciation such as specific Snowball Earth events.Elevated values of δ18Ozircon at times of supercontinent assembly (i.e. immediately preceding the red bars in Fig. 3) may be related to tectonically enhanced exchange of sediments with magma reservoirs (Spencer et al., 2014
Spencer, C. J., Cawood, P.A., Hawkesworth, C.J., Raub, T.D., Prave, A.R., Roberts, N.M.W. (2014) Proterozoic onset of crustal reworking and collisional tectonics: Reappraisal of the zircon oxygen isotope record. Geology 42, 451–454.
). Incorporation of sediments into melts can occur within a 100 Ma time window (Payne et al., 2015Payne, J.L., Hand, M., Pearson, N.J., Barovich, K.M., McInerney, D.J. (2015) Crustal thickening and clay: Controls on O isotope variation in global magmatism and siliciclastic sedimentary rocks. Earth and Planetary Science Letters 412, 70–76.
), short enough for the signal to propagate into the zircon record. The zircon record at times of continental assembly may also reflect increased degassing during collision (Bickle, 1996Bickle, M.J. (1996) Metamorphic decarbonation, silicate weathering and the long‐term carbon cycle. Terra Nova 8, 270–276.
; Kerrick and Caldeira, 1998Kerrick, D.M., Caldeira, K. (1998) Metamorphic CO2 degassing from orogenic belts. Chemical Geology 145, 213–232.
), generating high weathering fluxes and thus more altered crustal material.As each supercontinent matured, solid Earth degassing rates and sediment exchange may have diminished, decreasing δ18Ozircon values. More broadly, we speculate that the overall peak in δ18Ozircon around 2 Ga and the decrease towards the present day may be attributed to diminished overall mid-ocean ridge mantle degassing rates with time due to depletion of carbon in the mantle reservoir undergoing degassing (Hofmann, 1988
Hofmann, A.W. (1988) Chemical differentiation of the Earth: the relationship between mantle, continental crust, and oceanic crust. Earth and Planetary Science Letters 90, 297–314.
), if little subducted carbon is recycled into the convecting mantle (Kelemen and Manning, 2015Kelemen, P.B., Manning, C.E. (2015) Reevaluating carbon fluxes in subduction zones, what goes down, mostly comes up. Proceedings of the National Academy of Science of the USA 112, E3997–4006.
).In summary, based on the 4444 coupled δ18O-Hf-U/Pb data compiled in this study, we have identified peaks and valleys in the δ18Ozircon record, particularly between 2.5 and 0.5 Ga, that can be pinned to <250 Ma time intervals on the basis of ∆THf-U/Pb. We suggest that changes in continental weathering can explain these observed variations in a manner that is mechanistically consistent with the alteration of crustal material leaving a characteristic isotopic signature in zircon parent material. Inferred changes in continental weathering over time help to explain first order geologic events including the Neoproterozoic Snowball Earth episodes, which we argue based on the zircon record were preconditioned by low solid Earth degassing over the long-term. More rigorous testing of these hypothesised relationships will require better understanding of the links between increases in δ18Ozircon, the propagation of altered crust isotopic signature into parent magmas, and global continental weathering conditions.
Additional data collection and analysis, for example at higher temporal resolution, could help to refine understanding of relationships between δ18Ozircon and weathering conditions, including identifying the optimal ∆THf-U/Pb that captures the effect of geological events on the δ18Ozircon signal.
It might be possible to scale δ18Ozircon with changes in mantle degassing rate, given sufficient information about fractionation during weathering and incorporation of this signal into crustal melts, opening the possibility of extending the evaluations considered in this study to semi-quantitative interpretations. But with present constraints, the proxy remains qualitative, and indeed many factors other than continental weathering are likely to influence the δ18O of zircons. Nonetheless, even without a further leap of quantitatively linking δ18Ozircon to carbon fluxes, we propose that coupled δ18O-Hf-U/Pb data from zircons have the promise to illuminate links between mantle dynamics, plate tectonics, and weathering processes, helping to unravel the processes regulating Earth’s carbon cycle and climate over geologic time.
top
Acknowledgements
This work was supported by the German Science Foundation DFG (Cluster of Excellence ‘CliSAP’, EXC177, Universität Hamburg), the National Natural Science Foundation of China (grant no. 41422205 and 41730101), and the US National Science Foundation (grant nos. 1338329). The authors thank Wagner de Oliveira Garcia and Thorben Amann for help copying the data and making the videos. T.M. Gernon and one anonymous reviewer are thanked for constructive comments, and E. Oelkers for editorial handling.
Editor: Eric Oelkers
top
Author Contributions
J.H. and G.L. designed the study. J.H. collected the data. J.H., G.L., A.J.W. conducted the research and wrote the manuscript.
top
References
Bekker, A. (2014) Lomagundi Carbon Isotope Excursion. In: Amils, R., Gargaud, M., Cernicharo Quintalla, J., Cleaves, H.J., Irvine, W.M., Pinti, D., Viso, M. (Eds.) Encyclopedia of Astrobiology. Springer, Berlin, Heidelberg, 1–6.
Show in context Abundant Al-rich shales and quartz-rich sandstones at the time have been cited as evidence for intense weathering conditions (Bekker, 2014 ), and the extremely enriched δ13C values in carbonates has been explained by large amounts of organic carbon burial, resulting from enhanced biological productivity facilitated by the release of phosphorous (a limiting nutrient) during rock weathering (Bekker and Holland, 2012; Harada et al., 2015).
View in article
Bekker, A., Holland, H.D. (2012) Oxygen overshoot and recovery during the early Paleoproterozoic. Earth and Planetary Science Letters 317–318, 295–304.
Show in contextThe large carbon cycle anomalies observed in the sedimentary record of the Archean and Proterozoic eras represent puzzling deviations from this general stability (Hoffman et al., 1998; Goddéris et al., 2003; Bekker and Holland, 2012; Lyons et al., 2014).
View in article
Specifically, the rise towards the highest δ18Ozircon in the record coincides with the Lomagundi carbon isotope excursion (ca. 2.22–2.07 Ga) (Bekker and Holland, 2012), while the lowest δ18Ozircon in the record corresponds to the Neoproterozoic Snowball Earth events and the framing Kaigas, Gaskiers, and Vingerbreek glaciations (ca. 0.75–0.55 Ga) (Germs and Gaucher, 2012; Hofmann et al., 2015).
View in article
The Lomagundi carbon isotope excursion (Karhu and Holland, 1996; Bekker and Holland, 2012; Lyons et al., 2014) occurs as δ18Ozircon rises towards the highest values in the geological record (Fig. 3), suggesting the possibility of intense weathering at the time.
View in article
High weathering intensity may have resulted from high mantle degassing due to vigorous convection (Condie et al., 2001, 2016; Grenholm and Scherstén, 2015), or to the onset of oxidative weathering following the rise of atmospheric O2 (Bekker and Holland, 2012; Planavsky et al., 2012).
View in article
Abundant Al-rich shales and quartz-rich sandstones at the time have been cited as evidence for intense weathering conditions (Bekker, 2014 ), and the extremely enriched δ13C values in carbonates has been explained by large amounts of organic carbon burial, resulting from enhanced biological productivity facilitated by the release of phosphorous (a limiting nutrient) during rock weathering (Bekker and Holland, 2012; Harada et al., 2015).
View in article
Berner, R.A. (2004) The Phanerozoic Carbon Cycle: CO2 and O2. Oxford University Press, New York.
Show in context Over geologic time, Earth’s climate is regulated by the fluxes of carbon to and from the coupled ocean-atmosphere system, specifically by the balance between solid Earth degassing, the consumption and release of CO2 via rock weathering, and the burial and oxidation of organic carbon (Berner, 2004).
View in article
Berner, R.A., Caldeira, K. (1997) The need for mass balance and feedback in the geochemical carbon cycle. Geology 25, 955–956.
Show in context For most of Earth’s history, the sources and sinks of atmospheric CO2 have remained in close balance, maintaining the planet’s equable climate (Walker et al., 1983; Berner and Caldeira, 1997).
View in article
Global weathering fluxes should balance degassing fluxes over timescales >1 Ma, because of climate-dependent weathering feedbacks (e.g., Berner and Caldeira, 1997).
View in article
Bickle, M.J. (1996) Metamorphic decarbonation, silicate weathering and the long‐term carbon cycle. Terra Nova 8, 270–276.
Show in context The zircon record at times of continental assembly may also reflect increased degassing during collision (Bickle, 1996; Kerrick and Caldeira, 1998), generating high weathering fluxes and thus more altered crustal material.
View in article
Bindeman, I.N., Bekker, A., Zakharov, D.O. (2016) Oxygen isotope perspective on crustal evolution on early Earth: A record of Precambrian shales with emphasis on Paleoproterozoic glaciations and Great Oxygenation Event. Earth and Planetary Science Letters 437, 101–113.
Show in context The contribution from even small amounts of altered crustal material can increase δ18Ozircon values above the primary mantle signature of ~5.3 ± 0.3 ‰, because altered crust is isotopically enriched as a result of weathering by meteoric fluids (Savin and Epstein, 1970; Gregory and Taylor, 1981; Bindeman et al., 2016).
View in article
More intense chemical weathering and clay formation on the continents will produce more weathered sedimentary material, with heavier δ18O, increasing the potential for elevating δ18Ozircon (Valley et al., 2005; Payne et al., 2015; Bindeman et al., 2016)
View in article
Cawood, P.A., Hawkesworth, C.J., Dhuime, B. (2013) The continental record and the generation of continental crust. GSA Bulletin 125, 14–32.
Show in context Figure 3 [...] The grey band represents ±2 s.d. (s.d.: standard deviation of the mean). Red bars indicate periods of supercontinent assembly (Cawood et al., 2013).
View in article
Periods of supercontinent assembly (Cawood et al., 2013) appear to occur at similar times to peaks in δ18Ozircon (Fig. 3) and mature supercontinents with periods of lower δ18Ozircon, except during Gondwana-Pangea formation and breakup of the last 0.7 Ga.
View in article
Condie, K.C., Aster, R.C., van Hunen, J. (2016) A great thermal divergence in the mantle beginning 2.5 Ga: Geochemical constraints from greenstone basalts and komatiites. Geoscience Frontiers 7, 543–553.
Show in contextHigh weathering intensity may have resulted from high mantle degassing due to vigorous convection (Condie et al., 2001, 2016; Grenholm and Scherstén, 2015), or to the onset of oxidative weathering following the rise of atmospheric O2 (Bekker and Holland, 2012; Planavsky et al., 2012).
View in article
Condie, K.C., Des Marais, D.J., Abbott, D. (2001) Precambrian superplumes and supercontinents: a record in black shales, carbon isotopes, and paleoclimates? Precambrian Research 106, 239–260.
Show in context High weathering intensity may have resulted from high mantle degassing due to vigorous convection (Condie et al., 2001, 2016; Grenholm and Scherstén, 2015), or to the onset of oxidative weathering following the rise of atmospheric O2 (Bekker and Holland, 2012; Planavsky et al., 2012).
View in article
Dhuime, B., Hawkesworth, C., Cawood, P. (2011) When continents formed. Science 331, 154–155.
Show in context We primarily report results based on the New Crust Hf age model (Dhuime et al., 2011; Payne et al., 2016).
View in article
Donnadieu, Y., Goddéris, Y., Ramstein, G., Nédélec, A., Meert, J. (2004) A ‘snowball Earth’ climate triggered by continental break-up through changes in runoff. Nature 428, 303–306.
Show in contextIn simplistic terms, if glaciations were caused by drawdown of atmospheric CO2 via enhanced silicate weathering, as suggested in prior studies (Goddéris et al., 2003; Donnadieu et al., 2004; Pierrehumbert et al., 2011), the opposite would be expected, namely higher weathering intensity and thus elevated δ18Ozircon immediately preceding and during glaciation.
View in article
As a consequence, the results do not preclude an immediate triggering of glaciation by changes in weathering fluxes and CO2 drawdown, for example by enhanced weathering of continental crust, large igneous provinces, or submarine basalt (Goddéris et al., 2003; Donnadieu et al., 2004; Gernon et al., 2016).
View in article
Germs, G.J.B., Gaucher, C. (2012) Nature and extent of a late Ediacaran (ca. 547 Ma) glacigenic erosion surface in southern Africa. South African Journal of Geology 115, 91–102.
Show in context Specifically, the rise towards the highest δ18Ozircon in the record coincides with the Lomagundi carbon isotope excursion (ca. 2.22–2.07 Ga) (Bekker and Holland, 2012), while the lowest δ18Ozircon in the record corresponds to the Neoproterozoic Snowball Earth events and the framing Kaigas, Gaskiers, and Vingerbreek glaciations (ca. 0.75–0.55 Ga) (Germs and Gaucher, 2012; Hofmann et al., 2015).
View in article
Gernon, T.M., Hincks, T.K., Tyrrell, T., Rohling, E.J., Palmer, M.R. (2016) Snowball Earth ocean chemistry driven by extensive ridge volcanism during Rodinia breakup. Nature Geoscience 9, 242–248.
Show in context As a consequence, the results do not preclude an immediate triggering of glaciation by changes in weathering fluxes and CO2 drawdown, for example by enhanced weathering of continental crust, large igneous provinces, or submarine basalt (Goddéris et al., 2003; Donnadieu et al., 2004; Gernon et al., 2016).
View in article
The subsequent increase in δ18Ozircon as glaciations continued (i.e. leading up to 0.5 Ga) might have been influenced, at least in part, by intense weathering after glaciations (e.g., Hoffman et al., 1998) and submarine basalt weathering (Gernon et al., 2016).
View in article
Goddéris, Y., Donnadieu, Y., Nédélec, A., Dupré, B., Dessert, C., Grard, A., Ramstein, G., François, L.M. (2003) The Sturtian ‘snowball’ glaciation: fire and ice. Earth and Planetary Science Letters 211, 1–12.
Show in context The large carbon cycle anomalies observed in the sedimentary record of the Archean and Proterozoic eras represent puzzling deviations from this general stability (Hoffman et al., 1998; Goddéris et al., 2003; Bekker and Holland, 2012; Lyons et al., 2014).
View in article
In simplistic terms, if glaciations were caused by drawdown of atmospheric CO2 via enhanced silicate weathering, as suggested in prior studies (Goddéris et al., 2003; Donnadieu et al., 2004; Pierrehumbert et al., 2011), the opposite would be expected, namely higher weathering intensity and thus elevated δ18Ozircon immediately preceding and during glaciation.
View in article
As a consequence, the results do not preclude an immediate triggering of glaciation by changes in weathering fluxes and CO2 drawdown, for example by enhanced weathering of continental crust, large igneous provinces, or submarine basalt (Goddéris et al., 2003; Donnadieu et al., 2004; Gernon et al., 2016).
View in article
Goddéris, Y., Le Hir, G., Macouin, M., Donnadieu, Y., Hubert-Théou, L., Dera, G., Aretz, M., Fluteau, F., Li, Z.X., Halverson, G.P. (2017) Paleogeographic forcing of the strontium isotopic cycle in the Neoproterozoic. Gondwana Research 42, 151–162.
Show in contextGeochemical evidence from marine sediments, such as records of 87Sr/86Sr or δ13C, is subject to multiple interpretations (e.g., Kump, 1989; Goddéris et al., 2017).
View in article
Gregory, R.T., Taylor, H.P. (1981) An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophiolite, Oman: Evidence for δ18O buffering of the oceans by deep (> 5 km) seawater‐hydrothermal circulation at mid-ocean ridges. Journal of Geophysical Research: Solid Earth 86, 2737–2755.
Show in context The contribution from even small amounts of altered crustal material can increase δ18Ozircon values above the primary mantle signature of ~5.3 ± 0.3 ‰, because altered crust is isotopically enriched as a result of weathering by meteoric fluids (Savin and Epstein, 1970; Gregory and Taylor, 1981; Bindeman et al., 2016).
View in article
Grenholm, M., Scherstén, A. (2015) A hypothesis for Proterozoic-Phanerozoic supercontinent cyclicity, with implications for mantle convection, plate tectonics and Earth system evolution. Tectonophysics 662, 434–453.
Show in context High weathering intensity may have resulted from high mantle degassing due to vigorous convection (Condie et al., 2001, 2016; Grenholm and Scherstén, 2015), or to the onset of oxidative weathering following the rise of atmospheric O2 (Bekker and Holland, 2012; Planavsky et al., 2012).
View in article
Griffin, W.L., Wang, X., Jackson, S.E., Pearson, N.J., O'Reilly, S.Y., Xu, X., Zhou, X. (2002) Zircon chemistry and magma mixing, SE China: in-situ analysis of Hf isotopes, Tonglu and Pingtan igneous complexes. Lithos 61, 237–269.
Show in context The Hf isotopic composition of a zircon reflects the time when the grain’s unmixed parent material separated from the depleted mantle reservoir (Griffin et al., 2002; Hawkesworth and Kemp, 2006).
View in article
Gumsley, A.P., Chamberlain, K.R., Bleeker, W., Soderlund, U., de Kock, M.O., Larsson, E.R., Bekker, A. (2017) Timing and tempo of the Great Oxidation Event. Proc Natl Acad Sci U S A 114, 1811–1816.
Show in context Moreover, the Archean global glaciations (ca. 2.45–2.24 Ga) (Gumsley et al., 2017) also coincide with a local minimum in δ18Ozircon.
View in article
Harada, M., Tajika, E., Sekine, Y. (2015) Transition to an oxygen-rich atmosphere with an extensive overshoot triggered by the Paleoproterozoic snowball Earth. Earth and Planetary Science Letters 419, 178–186.
Show in context Abundant Al-rich shales and quartz-rich sandstones at the time have been cited as evidence for intense weathering conditions (Bekker, 2014 ), and the extremely enriched δ13C values in carbonates has been explained by large amounts of organic carbon burial, resulting from enhanced biological productivity facilitated by the release of phosphorous (a limiting nutrient) during rock weathering (Bekker and Holland, 2012; Harada et al., 2015).
View in article
Hawkesworth, C.J., Kemp, A.I.S. (2006) Using hafnium and oxygen isotopes in zircons to unravel the record of crustal evolution. Chemical Geology 226, 144–162.
Show in context The Hf isotopic composition of a zircon reflects the time when the grain’s unmixed parent material separated from the depleted mantle reservoir (Griffin et al., 2002; Hawkesworth and Kemp, 2006).
View in article
Hoffman, P.F., Kaufman, A.J., Halverson, G.P., Schrag, D.P. (1998) A Neoproterozoic snowball earth. Science 281, 1342–1346.
Show in context The large carbon cycle anomalies observed in the sedimentary record of the Archean and Proterozoic eras represent puzzling deviations from this general stability (Hoffman et al., 1998; Goddéris et al., 2003; Bekker and Holland, 2012; Lyons et al., 2014).
View in article
The subsequent increase in δ18Ozircon as glaciations continued (i.e. leading up to 0.5 Ga) might have been influenced, at least in part, by intense weathering after glaciations (e.g., Hoffman et al., 1998) and submarine basalt weathering (Gernon et al., 2016).
View in article
Hofmann, A.W. (1988) Chemical differentiation of the Earth: the relationship between mantle, continental crust, and oceanic crust. Earth and Planetary Science Letters 90, 297–314.
Show in context More broadly, we speculate that the overall peak in δ18Ozircon around 2 Ga and the decrease towards the present day may be attributed to diminished overall mid-ocean ridge mantle degassing rates with time due to depletion of carbon in the mantle reservoir undergoing degassing (Hofmann, 1988), if little subducted carbon is recycled into the convecting mantle (Kelemen and Manning, 2015).
View in article
Hofmann, M., Linnemann, U., Hoffmann, K.-H., Germs, G., Gerdes, A., Marko, L., Eckelmann, K., Gärtner, A., Krause, R. (2015) The four Neoproterozoic glaciations of southern Namibia and their detrital zircon record: The fingerprints of four crustal growth events during two supercontinent cycles. Precambrian Research 259, 176–188.
Show in contextSpecifically, the rise towards the highest δ18Ozircon in the record coincides with the Lomagundi carbon isotope excursion (ca. 2.22–2.07 Ga) (Bekker and Holland, 2012), while the lowest δ18Ozircon in the record corresponds to the Neoproterozoic Snowball Earth events and the framing Kaigas, Gaskiers, and Vingerbreek glaciations (ca. 0.75–0.55 Ga) (Germs and Gaucher, 2012; Hofmann et al., 2015).
View in article
Karhu, J.A., Holland, H.D. (1996) Carbon isotopes and the rise of atmospheric oxygen. Geology 24, 867–870.
Show in context The Lomagundi carbon isotope excursion (Karhu and Holland, 1996; Bekker and Holland, 2012; Lyons et al., 2014) occurs as δ18Ozircon rises towards the highest values in the geological record (Fig. 3), suggesting the possibility of intense weathering at the time.
View in article
Kelemen, P.B., Manning, C.E. (2015) Reevaluating carbon fluxes in subduction zones, what goes down, mostly comes up. Proceedings of the National Academy of Science of the USA 112, E3997–4006.
Show in contextMore broadly, we speculate that the overall peak in δ18Ozircon around 2 Ga and the decrease towards the present day may be attributed to diminished overall mid-ocean ridge mantle degassing rates with time due to depletion of carbon in the mantle reservoir undergoing degassing (Hofmann, 1988), if little subducted carbon is recycled into the convecting mantle (Kelemen and Manning, 2015).
View in article
Kerrick, D.M., Caldeira, K. (1998) Metamorphic CO2 degassing from orogenic belts. Chemical Geology 145, 213–232.
Show in context The zircon record at times of continental assembly may also reflect increased degassing during collision (Bickle, 1996; Kerrick and Caldeira, 1998), generating high weathering fluxes and thus more altered crustal material.
View in article
Kump, L.R. (1989) Alternative modeling approaches to the geochemical cycles of carbon, sulfur, and strontium isotopes. American Journal of Science 289, 390–410.
Show in context Geochemical evidence from marine sediments, such as records of 87Sr/86Sr or δ13C, is subject to multiple interpretations (e.g., Kump, 1989; Goddéris et al., 2017).
View in article
For example, the evolution of seawater 87Sr/86Sr over geological time (McArthur et al., 2012) may be influenced by changes in degassing and hydrothermal activity, continental weatherability, and the isotopic composition of rocks undergoing weathering (Kump, 1989).
View in article
Lyons, T.W., Reinhard, C.T., Planavsky, N.J. (2014) The rise of oxygen in Earth's early ocean and atmosphere. Nature 506, 307–315.
Show in context The large carbon cycle anomalies observed in the sedimentary record of the Archean and Proterozoic eras represent puzzling deviations from this general stability (Hoffman et al., 1998; Goddéris et al., 2003; Bekker and Holland, 2012; Lyons et al., 2014).
View in article
The Lomagundi carbon isotope excursion (Karhu and Holland, 1996; Bekker and Holland, 2012; Lyons et al., 2014) occurs as δ18Ozircon rises towards the highest values in the geological record (Fig. 3), suggesting the possibility of intense weathering at the time.
View in article
McArthur, J.M., Howarth, R.J., Shields, G.A. (2012) Sr isotope time series. In: Gradstein, F., Ogg, J., Schmitz, M., Ogg, G. (Eds.) The Geological Timescale 2012. Elsevier, Oxford, Amsterdam, Waltham, 127–144.
Show in context For example, the evolution of seawater 87Sr/86Sr over geological time (McArthur et al., 2012) may be influenced by changes in degassing and hydrothermal activity, continental weatherability, and the isotopic composition of rocks undergoing weathering (Kump, 1989).
View in article
McKenzie, N.R., Horton, B.K., Loomis, S.E., Stockli, D.F., Planavsky, N.J., Lee, C.-T.A. (2016) Continental arc volcanism as the principal driver of icehouse-greenhouse variability. Science 352, 444–447.
Show in context Low degassing flux has similarly been suggested as an underlying cause of icehouse climate states based on compilations of zircon abundance over the past ~0.7 Ga (McKenzie et al., 2016).
View in article
Mezger, K., Krogstad, E.J. (1997) Interpretation of discordant U‐Pb zircon ages: An evaluation. Journal of metamorphic Geology 15, 127–140.
Show in context In contrast, the U/Pb age of a zircon records the last time the mineral experienced temperatures above the closure threshold for the Pb system (>1000 °C), and thus the crystallisation age (Mezger and Krogstad, 1997).
View in article
Payne, J.L., Hand, M., Pearson, N.J., Barovich, K.M., McInerney, D.J. (2015) Crustal thickening and clay: Controls on O isotope variation in global magmatism and siliciclastic sedimentary rocks. Earth and Planetary Science Letters 412, 70–76.
Show in contextMore intense chemical weathering and clay formation on the continents will produce more weathered sedimentary material, with heavier δ18O, increasing the potential for elevating δ18Ozircon (Valley et al., 2005; Payne et al., 2015; Bindeman et al., 2016)
View in article
Thus, when plotting data from all individual grains, a very large spread is observed in δ18Ozircon values for any given U/Pb crystallisation age (Valley et al., 2005; Payne et al., 2015, 2016).
View in article
In common with prior studies, we see relatively little variability in δ18Ozircon until ~2.5 Ga, considering the patterns of long-term evolution over 4 Ga (Valley et al., 2005; Payne et al., 2015).
View in article
Incorporation of sediments into melts can occur within a 100 Ma time window (Payne et al., 2015), short enough for the signal to propagate into the zircon record.
View in article
Payne, J.L., McInerney, D.J., Barovich, K.M., Kirkland, C.L., Pearson, N.J., Hand, M. (2016) Strengths and limitations of zircon Lu-Hf and O isotopes in modelling crustal growth. Lithos 248–251, 175–192.
Show in contextThus, when plotting data from all individual grains, a very large spread is observed in δ18Ozircon values for any given U/Pb crystallisation age (Valley et al., 2005; Payne et al., 2015, 2016).
View in article
We primarily report results based on the New Crust Hf age model (Dhuime et al., 2011; Payne et al., 2016).
View in article
Pierrehumbert, R.T., Abbot, D.S., Voigt, A., Koll, D. (2011) Climate of the Neoproterozoic. Annual Review of Earth and Planetary Sciences 39, 417–460.
Show in context In simplistic terms, if glaciations were caused by drawdown of atmospheric CO2 via enhanced silicate weathering, as suggested in prior studies (Goddéris et al., 2003; Donnadieu et al., 2004; Pierrehumbert et al., 2011), the opposite would be expected, namely higher weathering intensity and thus elevated δ18Ozircon immediately preceding and during glaciation.
View in article
Planavsky, N.J., Bekker, A., Hofmann, A., Owens, J.D., Lyons, T.W. (2012) Sulfur record of rising and falling marine oxygen and sulfate levels during the Lomagundi event. Proceedings of the National Academy of Sciences 109, 18300–18305.
Show in contextSavin, S.M., Epstein, S. (1970) The oxygen and hydrogen isotope geochemistry of clay minerals. Geochimica et Cosmochimica Acta 84, 25–42.
Show in context The contribution from even small amounts of altered crustal material can increase δ18Ozircon values above the primary mantle signature of ~5.3 ± 0.3 ‰, because altered crust is isotopically enriched as a result of weathering by meteoric fluids (Savin and Epstein, 1970; Gregory and Taylor, 1981; Bindeman et al., 2016).
View in article
Spencer, C. J., Cawood, P.A., Hawkesworth, C.J., Raub, T.D., Prave, A.R., Roberts, N.M.W. (2014) Proterozoic onset of crustal reworking and collisional tectonics: Reappraisal of the zircon oxygen isotope record. Geology 42, 451–454.
Show in contextElevated values of δ18Ozircon at times of supercontinent assembly (i.e. immediately preceding the red bars in Fig. 3) may be related to tectonically enhanced exchange of sediments with magma reservoirs (Spencer et al., 2014).
View in article
Teitler, Y., Le Hir, G., Fluteau, F., Philippot, P., Donnadieu, Y. (2014) Investigating the Paleoproterozoic glaciations with 3-D climate modeling. Earth and Planetary Science Letters 395, 71–80.
Show in context Whether similar conditions prevailed at the time of the earlier Archean glaciations remains unclear from the zircon data (e.g., see suggested mechanisms for the onset of Palaeoproterozoic glaciations; Teitler et al., 2014), and we emphasise that the long integration window of the zircon data preclude interpretation of trends in Figure 3 in terms of individual episodes of glaciation such as specific Snowball Earth events.
View in article
Valley, J.W. (2003) Oxygen isotopes in zircon. Reviews in Mineralogy and Geochemistry 53, 343–385.
Show in context While petrologic processes and magma chamber conditions such as oxygen fugacity may also influence δ18Ozircon by changing fractionation factors (Valley, 2003), we propose that, when averaged over grains from around the world, δ18Ozircon reflects globally averaged weathering conditions during a given ∆THf-U/Pb interval.
View in article
Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S. (2005) 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contributions to Mineralogy and Petrology 150, 561–580.
Show in context The difference between the Hf and U/Pb ages, here termed ∆THf-U/Pb, represents the time interval during which zircon parent material could have been affected by weathering and magmatic processes (Fig. 1), leaving an imprint on δ18Ozircon (Valley et al., 2005).
View in article
The extent of isotopic enrichment for a given zircon grain will depend on the amount of weathering-related contamination of the parent magma, which is related to the alteration state of the crustal material (Valley et al., 2005).
View in article
More intense chemical weathering and clay formation on the continents will produce more weathered sedimentary material, with heavier δ18O, increasing the potential for elevating δ18Ozircon (Valley et al., 2005; Payne et al., 2015; Bindeman et al., 2016)
View in article
Thus, when plotting data from all individual grains, a very large spread is observed in δ18Ozircon values for any given U/Pb crystallisation age (Valley et al., 2005; Payne et al., 2015, 2016).
View in article
In common with prior studies, we see relatively little variability in δ18Ozircon until ~2.5 Ga, considering the patterns of long-term evolution over 4 Ga (Valley et al., 2005; Payne et al., 2015).
View in article
Walker, J.C.G., Klein, C., Schidlowski, M., Schopf, J.W., Stevenson, D.J., Walter, M.R. (1983) Environmental evolution of the Archean-early Proterozoic Earth. In: Schopf, J.W. (Ed.) Earth's earliest biosphere: Its origin and evolution (A84-43051 21-51). Princeton University Press, Princeton, New Jersey, 260–290.
Show in contextFor most of Earth’s history, the sources and sinks of atmospheric CO2 have remained in close balance, maintaining the planet’s equable climate (Walker et al., 1983; Berner and Caldeira, 1997).
View in article
top
Supplementary Information
Selection of Data
In addition to a previously published compilation of coupled δ18O-Hf-U-Pb data from zircons (Dhuime et al., 2012), data were compiled from 18 other literature sources reporting detrital zircon analyses (Table S-1). Data were included for zircons with values for δ18O, U/Pb-age, and Lu and Hf isotopes measurements all on the same grain, with the latter allowing calculation of Hf model ages. Datasets were copied from supplementary source files of the given publications or from tables provided in the publications.
In most cases, U/Pb age was reported in the source publication, though in some cases ages were not reported and were calculated based on published isotope ratios. In latter cases, data were restricted to results showing discordance less than 10 % for the 207Pb/206Pb, 206Pb/238U and 207Pb/235U age.
176Lu/177Lu and 176Hf/177Hf isotope ratios were used to calculate the New Crust and the TDM Hf-model ages following the approach and data provided by Payne et al. (2016), with λ = 1.865 10-11 a-1 (the decay constant for 176Lu), 176Hf/177Hf = 0.283251 for the CHUR-like composition, and 176Lu0/177 HfDM = 0.0384 for the today’s depleted mantle (DM is depleted mantle). The New Crust and TDM Hf model ages were newly calculated with these parameters for all compiled data, using the published isotope ratios.
In 22 and 92 cases, the difference between the Hf model ages (traditional TDM and New Crust, respectively) and the U/Pb-age was negative and set to zero. Reported rim-data from Zeh et al. (2014) were removed.
Table S-1 New data sources with coupled δ18O, U/Pb age and Hf model age information in addition to Dhuime et al. (2012).
| Source Reference | Number of data points |
| Canile et al. (2016) | 170 |
| Castillo et al. (2016) | 167 |
| Davis et al. (2015) | 128 |
| Ge et al. (2014) | 37 |
| Herve et al. (2013) | 149 |
| Hollis et al. (2014) | 184 |
| Iizuka et al. (2013) | 442 |
| Jiang et al. (2015) | 158 |
| Li et al. (2012) | 257 |
| Meinhold et al. (2014) | 29 |
| Pankhurst et al. (2016) | 161 |
| Partin et al. (2014) | 130 |
| Wang et al. (2012) | 125 |
| Yang et al. (2015) | 324 |
| Yin et al. (2012) | 131 |
| Zeh et al. (2014) | 132 |
| Zhang et al. (2016) | 143 |
| Zhang et al. (2014) | 201 |
| Count new sources | 3068 |
| Dhuime et al. (2012) | 1376 |
| Sum of all used data points | 4444 |
Differences in the Hf Model Age Calculations
The differences between the TDM and New Crust Hf model ages result in slightly younger ages, of on average 150 Ma, for the New Crust Hf model ages. The location of discussed minima and maxima in the time series therefore shift slightly. The main conclusions if comparing results applying both Hf model age types do not change, i.e. the minima and maxima as discussed in the main text are still within the time window of relevance for given geological events. Results based on the traditional TDM Hf model age calculation are shown in Figures S-1 and S-2.
Figure S-1 Results as in Figure 2 in the main text, but using the traditional TDM Hf model age instead of the New Crust Hf model age. Time evolution of δ18Ozircon calculated for different values of ∆THf-U/Pb as selection criterion (see also Supplementary Video S-2). The very short values of the selection criterion ∆THf-U/Pb restrict the total number of grains in the dataset to a small number (<400 grains if ∆THf-U/Pb <300 Ma and using the TDM Hf model age). Blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The green bar covers the period of the Lomagundi event.
Figure S-2 Results as in Figure 3 in the main text, but using the traditional TDM Hf model age instead of the New Crust Hf model age. Note that, due the difference in the calculation procedure, the New Crust Hf model age for single grains provides on average 150 Ma shorter ages than the TDM Hf model age. Dark blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The dark green bar covers the period of the Lomagundi event. To illustrate the effect on comparison between the timing of geological events and the δ18Ozircon curve, lighter shaded areas around periods of geological events reflect uncertainty of the δ18Ozircon curve associated with plotting the curve for the midpoint between the Hf and U/Pb ages, given as half of the mean ∆THf-U/Pb (140 Ma).
Figure S-3 Results as in Figure 3 in the main text, but with analysis of variance in ages for the population of zircons with ∆THf-U/Pb <400 Ma. The black line shows δ18Ozircon using the weighted mean zircon age of the selected data, when age is weighted using the same Gaussian filtering procedure as applied to the δ18Ozircon data in Figure 3 in the main text. As expected, the weighted age (black line) plots very similarly to the actual ages used in Figure 3. The dashed blue and red lines show δ18Ozircon plotted against time considering variance in the ages of the zircon population, specifically the weighted ±1 standard deviation of the age using the same Gaussian filtering as applied to the δ18Ozircon data for the midpoint ages (blue: minus 1 s.d.; red: plus 1 s.d.). Weighted variance in ages ranges from 66 to 130 Ma and does not change systematically over time, so we consider the single average value reported in the text (225 Ma) as representative of the time resolution of the record for a ∆THf-U/Pb window <400 Ma.
Supplementary Videos
The supplementary videos show the evolution of the δ18O time series with increasing ∆THf-U/Pb time window as selection filter for both approaches to calculate the Hf-model age. The videos also provide for each time step the number of included data points and the mean ∆THf-U/Pb for data points fulfilling the selection criterion.
Video S-1 This supplementary video shows the evolution of the δ18O time series with increasing ∆THf-U/Pb time window as selection filter for the New Crust Hf model age. The video also provides for each time step the number of included data points and the mean ∆THf-U/Pb for data points fulfilling the selection criterion.
Video S-2 This supplementary video shows the evolution of the δ18O time series with increasing ∆THf-U/Pb time window as selection filter for the traditional TDM Hf model age. The video also provides for each time step the number of included data points and the mean ∆THf-U/Pb for data points fulfilling the selection criterion.
Supplementary Data Table
Table S-2 can be viewed and downloaded in Excel below.
Table S-2 Applied data including δ18O, hafnium and lutetium isotopic data, the uranium-lead-isotope age and references. (a) New compiled data. (b) Dhuime et al. (2012) data.
Supplementary Information References
Castillo, P., Fanning, C.M., Hervé, F., Lacassie, J.P. (2016) Characterisation and tracing of Permian magmatism in the south-western segment of the Gondwanan margin; U–Pb age, Lu–Hf and O isotopic compositions of detrital zircons from metasedimentary complexes of northern Antarctic Peninsula and western Patagonia. Gondwana Research 36, 1–13.
Davis, W.J., Ootes, L., Newton, L., Jackson, V., Stern, R.A. (2015) Characterization of the Paleoproterozoic Hottah terrane, Wopmay Orogen using multi-isotopic (U-Pb, Hf and O) detrital zircon analyses: An evaluation of linkages to northwest Laurentian Paleoproterozoic domains. Precambrian Research 269, 296–310.
Dhuime, B., Hawkesworth, C.J., Cawood, P.A., Storey, C.D. (2012) A Change in the Geodynamics of Continental Growth 3 Billion Years Ago. Science 335, 1334–1336.
Ge, R., Zhu, W., Wilde, S.A., He, J. (2014) Zircon U–Pb–Lu–Hf–O isotopic evidence for ≥3.5Ga crustal growth, reworking and differentiation in the northern Tarim Craton. Precambrian Research 249, 115–128.
Hervé, F., Calderón, M., Fanning, C.M., Pankhurst, R.J., Godoy, E. (2013) Provenance variations in the Late Paleozoic accretionary complex of central Chile as indicated by detrital zircons. Gondwana Research 23, 1122–1135.
Hollis, J.A., Carson, C.J., Glass, L.M., Kositcin, N., Scherstén, A., Worden, K.E., Armstrong, R.A., Yaxley, G.M., Kemp, A.I.S. (2014) Detrital zircon U–Pb–Hf and O isotope character of the Cahill Formation and Nourlangie Schist, Pine Creek Orogen: Implications for the tectonic correlation and evolution of the North Australian Craton. Precambrian Research 246, 35–53.
Iizuka, T., Campbell, I.H., Allen, C.M., Gill, J.B., Maruyama, S., Makoka, F. (2013) Evolution of the African continental crust as recorded by U–Pb, Lu–Hf and O isotopes in detrital zircons from modern rivers. Geochimica et Cosmochimica Acta 107, 96–120.
Jiang, X.-Y., Li, X.-H., Collins, W.J., Huang, H.-Q. (2015) U-Pb age and Hf-O isotopes of detrital zircons from Hainan Island: Implications for Mesozoic subduction models. Lithos 239, 60–70.
Li, X.-H., Li, Z.-X., He, B., Li, W.-X., Li, Q.-L., Gao, Y., Wang, X.-C. (2012) The Early Permian active continental margin and crustal growth of the Cathaysia Block: In situ U–Pb, Lu–Hf and O isotope analyses of detrital zircons. Chemical Geology 328, 195–207.
Meinhold, G., Morton, A.C., Fanning, C.M., Howard, J.P., Phillips, R.J., Strogen, D., Whitham, A.G. (2014) Insights into crust formation and recycling in North Africa from combined U–Pb, Lu–Hf and O isotope data of detrital zircons from Devonian sandstone of southern Libya. Geological Society, London, Special Publications 386, 281–292.
Pankhurst, R.J., Hervé, F., Fanning, C.M., Calderón, M., Niemeyer, H., Griem-Klee, S., Soto, F. (2016) The pre-Mesozoic rocks of northern Chile: U–Pb ages, and Hf and O isotopes. Earth-Science Reviews 152, 88–105.
Partin, C.A., Bekker, A., Sylvester, P.J., Wodicka, N., Stern, R.A., Chacko, T., Heaman, L.M. (2014) Filling in the juvenile magmatic gap: Evidence for uninterrupted Paleoproterozoic plate tectonics. Earth and Planetary Science Letters 388, 123–133.
Payne, J.L., McInerney, D.J., Barovich, K.M., Kirkland, C.L., Pearson, N.J., Hand, M. (2016) Strengths and limitations of zircon Lu-Hf and O isotopes in modelling crustal growth. Lithos 248–251, 175–192.
Wang, X.-C., Li, X.-h., Li, Z.-X., Li, Q.-l., Tang, G.-Q., Gao, Y.-Y., Zhang, Q.-R., Liu, Y. (2012) Episodic Precambrian crust growth: Evidence from U–Pb ages and Hf–O isotopes of zircon in the Nanhua Basin, central South China. Precambrian Research 222–223, 386–403.
Yang, C., Li, X.-H., Wang, X.-C., Lan, Z. (2015) Mid-Neoproterozoic angular unconformity in the Yangtze Block revisited: Insights from detrital zircon U–Pb age and Hf–O isotopes. Precambrian Research 266, 165–178.
Yin, Q.Z., Wimpenny, J., Tollstrup, D.L., Mange, M., Dewey, J.F., Zhou, Q., Li, X.H., Wu, F.Y., Li, Q.L., Liu, Y., Tang, G.Q. (2012) Crustal evolution of the South Mayo Trough, western Ireland, based on U-Pb ages and Hf-O isotopes in detrital zircons. Journal of the Geological Society 169, 681–689.
Zeh, A., Stern, R.A., Gerdes, A. (2014) The oldest zircons of Africa—Their U–Pb–Hf–O isotope and trace element systematics, and implications for Hadean to Archean crust–mantle evolution. Precambrian Research 241, 203–230.
Zhang, H.-F., Wang, J.-L., Zhou, D.-W., Yang, Y.-H., Zhang, G.-W., Santosh, M., Yu, H., Zhang, J. (2014) Hadean to Neoarchean episodic crustal growth: Detrital zircon records in Paleoproterozoic quartzites from the southern North China Craton. Precambrian Research 254, 245–257.
Zhang, H.-F., Zhang, J., Zhang, G.-W., Santosh, M., Yu, H., Yang, Y.-H., Wang, J.-L. (2016) Detrital zircon U–Pb, Lu–Hf, and O isotopes of the Wufoshan Group: Implications for episodic crustal growth and reworking of the southern North China craton. Precambrian Research 273, 112–128.
Figures and Tables
Figure 1 Schematic diagram of the age evolution of a zircon grain, showing how combined Hf and U/Pb isotope information can be used to determine the time window (∆THf-U/Pb) during which the δ18Ozircon value is acquired, based on the conceptual model proposed here. The Hf model age constrains the time that parent material separated from the mantle, beginning its journey of reworking via weathering, erosion, and incorporation into melts. This journey ends when the zircon crystallises, as represented by its U/Pb age. Longer ∆THf-U/Pb windows may be affected by mixing and make it difficult to identify the timing when the δ18Ozircon signature was acquired. We screen zircons by ∆THf-U/Pb in order to calculate the time evolution of δ18Ozircon.
Figure 2 Time evolution of δ18Ozircon calculated for different values of ∆THf-U/Pb as selection criterion (see also Supplementary Video S-1). The very short values of ∆THf-U/Pb restrict the total number of grains in the dataset to a small number (<1000 grains if ∆THf-U/Pb <300 Ma using the New Crust Hf model age, shown here), leading to large uncertainties, although some of the first order features are still evident (e.g., a maximum ~2 Ga and minimum ~0.8 Ga). For longer ∆THf-U/Pb windows, the uncertainties are much reduced and general patterns are stable across a range of window sizes. For very large ∆THf-U/Pb (e.g., >1000 Ma; see Supplementary Video S-1) the δ18Ozircon curve represents a long-term accumulated signal that is difficult to tie to specific geologic events because of the long integration window. The grey band represents ±2 s.d. (s.d.: standard deviation of the mean). Blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The green bar covers the period of the Lomagundi event. Also see Figure 3.
Figure 3 δ18Ozircon calculated for ∆THf-U/Pb <400 Ma, along with timing of major carbon cycle anomalies observed in the geologic record. The grey band represents ±2 s.d. (s.d.: standard deviation of the mean). Red bars indicate periods of supercontinent assembly (Cawood et al., 2013
Cawood, P.A., Hawkesworth, C.J., Dhuime, B. (2013) The continental record and the generation of continental crust. GSA Bulletin 125, 14–32.
). Dark blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The dark green bar covers the period of the Lomagundi event. Note that the δ18Ozircon values are plotted using the midpoint age of the ∆THf-U/Pb window, but the actual δ18Ozircon values integrate processes (e.g., weathering conditions) occurring over longer periods of time, characterised by a mean ∆THf-U/Pb duration of 225 Ma. To illustrate the effect on the comparison between the timing of geological events and the δ18Ozircon curve, lighter shaded areas around periods of geological events reflect uncertainty of the δ18Ozircon curve associated with plotting the curve for the midpoint between the Hf and U/Pb ages, given as half of the mean ∆THf-U/Pb (112.5 Ma).Back to article
Supplementary Figures and Tables
Table S-1 New data sources with coupled δ18O, U/Pb age and Hf model age information in addition to Dhuime et al. (2012).
| Source Reference | Number of data points |
| Canile et al. (2016) | 170 |
| Castillo et al. (2016) | 167 |
| Davis et al. (2015) | 128 |
| Ge et al. (2014) | 37 |
| Herve et al. (2013) | 149 |
| Hollis et al. (2014) | 184 |
| Iizuka et al. (2013) | 442 |
| Jiang et al. (2015) | 158 |
| Li et al. (2012) | 257 |
| Meinhold et al. (2014) | 29 |
| Pankhurst et al. (2016) | 161 |
| Partin et al. (2014) | 130 |
| Wang et al. (2012) | 125 |
| Yang et al. (2015) | 324 |
| Yin et al. (2012) | 131 |
| Zeh et al. (2014) | 132 |
| Zhang et al. (2016) | 143 |
| Zhang et al. (2014) | 201 |
| Count new sources | 3068 |
| Dhuime et al. (2012) | 1376 |
| Sum of all used data points | 4444 |
Figure S-1 Results as in Figure 2 in the main text, but using the traditional TDM Hf model age instead of the New Crust Hf model age. Time evolution of δ18Ozircon calculated for different values of ∆THf-U/Pb as selection criterion (see also Supplementary Video S-2). The very short values of the selection criterion ∆THf-U/Pb restrict the total number of grains in the dataset to a small number (<400 grains if ∆THf-U/Pb <300 Ma and using the TDM Hf model age). Blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The green bar covers the period of the Lomagundi event.
Figure S-2 Results as in Figure 3 in the main text, but using the traditional TDM Hf model age instead of the New Crust Hf model age. Note that, due the difference in the calculation procedure, the New Crust Hf model age for single grains provides on average 150 Ma shorter ages than the TDM Hf model age. Dark blue bars represent periods of glaciations in the late Archean and in the Neoproterozoic. The dark green bar covers the period of the Lomagundi event. To illustrate the effect on comparison between the timing of geological events and the δ18Ozircon curve, lighter shaded areas around periods of geological events reflect uncertainty of the δ18Ozircon curve associated with plotting the curve for the midpoint between the Hf and U/Pb ages, given as half of the mean ∆THf-U/Pb (140 Ma).
Figure S-3 Results as in Figure 3 in the main text, but with analysis of variance in ages for the population of zircons with ∆THf-U/Pb <400 Ma. The black line shows δ18Ozircon using the weighted mean zircon age of the selected data, when age is weighted using the same Gaussian filtering procedure as applied to the δ18Ozircon data in Figure 3 in the main text. As expected, the weighted age (black line) plots very similarly to the actual ages used in Figure 3. The dashed blue and red lines show δ18Ozircon plotted against time considering variance in the ages of the zircon population, specifically the weighted ±1 standard deviation of the age using the same Gaussian filtering as applied to the δ18Ozircon data for the midpoint ages (blue: minus 1 s.d.; red: plus 1 s.d.). Weighted variance in ages ranges from 66 to 130 Ma and does not change systematically over time, so we consider the single average value reported in the text (225 Ma) as representative of the time resolution of the record for a ∆THf-U/Pb window <400 Ma.
Video S-1 This supplementary video shows the evolution of the δ18O time series with increasing ∆THf-U/Pb time window as selection filter for the New Crust Hf model age. The video also provides for each time step the number of included data points and the mean ∆THf-U/Pb for data points fulfilling the selection criterion.
Video S-2 This supplementary video shows the evolution of the δ18O time series with increasing ∆THf-U/Pb time window as selection filter for the traditional TDM Hf model age. The video also provides for each time step the number of included data points and the mean ∆THf-U/Pb for data points fulfilling the selection criterion.
Table S-2 can be viewed and downloaded in Excel below.
Table S-2 Applied data including δ18O, hafnium and lutetium isotopic data, the uranium-lead-isotope age and references. (a) New compiled data. (b) Dhuime et al. (2012) data.