Predominantly recycled carbon in Earth's upper mantle revealed by He-CO2-Ba systematics in ultradepleted ocean ridge basalts

https://doi.org/10.1016/j.epsl.2020.116646Get rights and content

Highlights

  • Basalts in the Garrett Transform Fault are derived from ultradepleted mantle.

  • This ultradepleted mantle has 3He/4He = 9-10 RA and CO2/3He<3×108.

  • Basalts along the East Pacific Rise are derived from heterogeneous, depleted mantle.

  • This heterogeneous, depleted mantle has 3He/4He = 7-9 RA and CO2/3He 2×109.

  • The higher CO2/3He in EPR basalts shows that most upper mantle carbon is recycled.

Abstract

Basalts erupted within intra-transform spreading centers can be valuable probes of geochemical components in Earth's upper mantle, and provide constraints on the proportions of mantle carbon that are juvenile (primordial) vs. tectonically recycled. Here we present new results for submarine basalt glasses, erupted within the Garrett Transform Fault (GTF) and to its north and south along the East Pacific Rise (EPR). Analyses of 3He/4He, and He and CO2 concentrations in vesicles and glass were performed, through a series of crushing and melting experiments plus FTIR spectroscopy. Trace elements were analyzed by laser ablation ICP-MS.

The GTF basalts provide further tests for the origin of volatile-undersaturated basalts using CO2/Ba and CO2/Nb systematics. CO2 is highly correlated with Ba and Nb in basalts erupted in the transform domain (n = 13, including 5 undersaturated basalts) and does not show the variability expected from mixing between undegassed and variably degassed melts. Rather, the melts appear to originate from a heterogeneous mantle source that was variably depleted through partial melting, and limited mixing of melts is involved in their generation. The CO2/Ba and CO2/Nb weight ratios of 106 ± 8 and 308 ± 27, respectively, are similar to values determined previously for a global suite of undersaturated mid-ocean ridge basalts (MORBs).

The ridge and transform domains have distinct 3He/4He ratios. Along the nearby EPR, 3He/4He = 8.5 - 9.1 RA, while within the Garrett Transform Fault 3He/4He = 9.2 - 10.1 RA. These two basalt populations are also distinct in their Pb-Sr-Nd isotope compositions based on earlier regional studies. The distinct populations result from partial melting of two different mantle source compositions. Melting of depleted mantle containing a small amount (∼1 to 5%) of enriched, ancient heterogeneities occurs beneath the EPR. Melting of ultradepleted mantle (in which the heterogeneities have been removed by earlier melting beneath the EPR) occurs beneath the GTF. This explains the distinction between intra-transform and spreading ridge domains for 3He/4He, if the heterogeneities were enriched in U, Th and He and had low 3He/4He as would be found in tectonically recycled material. The enriched mantle component sampled by the EPR basalts has molar CO2/3He =2×109, and it dominates the CO2/3He ratio generally ascribed to the upper mantle source for mid-ocean ridge basalts. In contrast, the ultradepleted MORB mantle component sampled by the GTF basalts has CO2/3He =3×108 or less. This indicates that most of the carbon in Earth's upper mantle originates from tectonic recycling.

Introduction

The degree to which carbon has been recycled to Earth's deep and shallow mantle is debated. An early view, comparing the concentration and isotope composition of C in peridotites and basalts, was that much of the carbon throughout Earth's mantle was recycled over geologic time given the apparent large flux of carbon from the mantle compared to the mass budget of carbon in the so-called “exosphere”, i.e., the crust+oceans+atmosphere (Javoy et al., 1982). Although the magnitude of that flux was probably overestimated (Walker, 1983; Javoy et al., 1983), the amount of C recycled may still have been significant. Given that the mantle has cooled over geologic time, and the thermal structure of subduction zones has allowed crustal material to be transported to depths greater than the generation depth of volcanic arc magma, recycling of carbon to the mantle seems quite likely to have occurred since at least the Proterozoic (Dasgupta, 2013).

Recent studies of carbon at the Hawaiian hotspot indicate that there is a relatively larger concentration of C (by at least a factor 3 to 8) in its deep mantle source compared to Earth's shallow mantle source of mid-ocean ridge basalts (Tucker et al., 2019). Given that other radiogenic isotope and noble gas evidence shows that the Hawaiian mantle source contains both recycled and primitive mantle components, its enrichment in C may be due either to an abundance of C-rich primitive material or to a significant recycling of surficial C (Tucker et al., 2019). In a study of xenoliths from high-3He/4He ocean island plume localities (Loihi-Hawaii, Réunion, and Kerguelen), Trull et al. (1993) showed that the similar C isotope and C/3He systematics for ocean island and MORB sources were consistent with zoned mantle convection and significant surficial C recycling to the deep mantle, with limited or no C recycling to the upper mantle. Recently, Kelemen and Manning (2015) showed that there seems to be a significant accumulation of C in the continental lithosphere during subduction and plate collision, and argued that the amount of C recycling to the asthenosphere may actually be negligible. But in another recent study, Hirschmann (2018) estimated that >25% of the C that has been outgassed to the exosphere has been recycled to the mantle over geologic time based on the observed difference in CO2/Ba between the mantle and the exosphere.

Mid-ocean ridge basalts are important for assessing the compositional variability of Earth's upper mantle and for providing insight into the deep carbon cycle. Notably, MORBs that are ultradepleted in incompatible elements (UD-MORBs) are sometimes also volatile-undersaturated; that is, their total volatile pressure (determined from the CO2 and H2O content of the glass plus any bubbles that might be present) is less than the sample collection pressure. Olivine-hosted melt inclusions (OHMI) may also retain undegassed magmatic volatile concentrations. Under these circumstances, basalt or OHMI compositions may be used to estimate the C concentration of the mantle source and the volcanic C flux to the exosphere (Saal et al., 2002; Rosenthal et al., 2015; Michael and Graham, 2015; Le Voyer et al., 2017, 2018; Hirschmann, 2018). Although volatile-undersaturated basalt glasses are rare, they are significant because they can represent magma that has not lost any gas since its formation, and aid in the development of chemical proxies for volatile elements such as C that are lost to an unknown amount in other volcanic rocks.

There are at least two complications when utilizing undersaturated basalt glasses to characterize the upper mantle. First, their rare occurrence can make it difficult to obtain a geologic context and understand their relationship to nearby basalts that are not undersaturated. Michael and Graham (2015) suggested from trace element abundances in their study of UD-MORBs that they could be derived by partial melting of an upper mantle source that had undergone an earlier and relatively recent melting event during which 1-2% of melt was removed. About a dozen separate localities along the global mid-ocean ridge system were found to erupt undersaturated basalts; these localities are sometimes associated with atypical tectonic settings, such as near the end of ridge segments or, in a number of cases, they occur at intra-transform spreading centers. A second complication arises from the nearly ubiquitous occurrence of concurrent mixing and crystal fractionation in magma as it ascends and aggregates beneath mid-ocean ridges (Shorttle, 2015). This phenomenon could lead to the eruption of volatile-undersaturated magmas (or OHMI) having ratios of volatile/lithophile elements, such as CO2/Ba or CO2/Nb, that are lower than their mantle source values (Matthews et al., 2017).

The present study investigates a suite of 25 submarine basalts collected from well surveyed regions of the southern East Pacific Rise and within the Garrett Transform Fault. We find that CO2 is highly correlated with Ba and Nb in the transform domain basalts (n = 13, including 5 undersaturated samples). These GTF lavas have CO2/Ba and CO2/Nb ratios (ppm/ppm) of 106 ± 8 and 308 ± 27, respectively, and are indistinguishable from the global UD-MORB suite of Michael and Graham (2015). Furthermore, 3He/4He shows no overlap between EPR and GTF basalt populations. The GTF basalts are derived by partial melting of mantle that underwent a small degree of prior melting beneath the EPR (Wendt et al., 1999). The ultradepleted mantle sampled by the GTF basalts has 3He/4He >9.2 RA and molar CO2/3He 3×108. We propose that the helium isotope and CO2/3He ratios of typical MORB (∼7-9 RA and ∼2×109, respectively) are significantly affected by the presence of tectonically recycled material in their source, and that most of the carbon present in Earth's upper mantle is not juvenile, but has been tectonically recycled from Earth's surface. (The term juvenile is used here to indicate primordial material that has never been present at Earth's surface.)

Section snippets

Background

The Garrett Transform Fault is 130 km long and offsets the superfast spreading East Pacific Rise near 13°30'S (145 mm/y full spreading rate). The transform domain is volcanically active, and basalt, gabbro, and peridotite have all been recovered there (Hébert et al., 1983, 1997; Hékinian et al., 1992, 1995; Cannat et al., 1990; Niu and Hékinian, 1997). Basalts from both the nearby East Pacific Rise and within the Garrett Transform Fault are geochemically well-characterized (Sinton et al., 1991;

Methods and results

Analyses of 3He/4He and He concentrations in vesicles and glass were performed through a series of crushing and melting experiments followed by noble gas mass spectrometry. The CO2 content of vesicles was determined by capacitance manometry, and CO2 and H2O concentrations dissolved in the glasses were measured by FTIR spectroscopy. The basalt glasses were also analyzed for trace element concentrations by laser ablation ICP-MS. Methods follow those described previously (Michael and Graham, 2015;

CO2-Ba and CO2-Nb systematics and the origin of GTF undersaturated basalts

Michael and Graham (2015) refined the CO2/Nb proxy of Saal et al. (2002) and Cartigny et al. (2008) by studying a global suite of undersaturated mid-ocean ridge basalts. They found that another useful proxy element was Ba, and suggested that the depleted upper mantle was characterized by a CO2/Ba ratio of 105 ± 9. Using a simplistic melting model (mean extent of partial melting of 12%) and the mean global MORB Ba concentration of 29.2 ppm (Gale et al., 2013), they estimated the upper mantle [CO2

Summary

He and CO2 were measured in vesicles and glass of East Pacific Rise and Garrett Transform basalts. Total volatile pressures in the basalts reveal the presence of 5 volatile-undersaturated basalts collected from 3 distinct locations within the GTF (a leaky strike-slip fault, Beta Ridge, and adjacent to Alpha Ridge). The CO2-Ba and CO2-Nb systematics in the GTF undersaturated basalts indicate an origin by variable source depletion, with limited or no mixing of melts. The mantle source has ratios

CRediT authorship contribution statement

David W. Graham: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing - original draft, Writing - review & editing. Peter J. Michael: Formal analysis, Investigation, Methodology, Visualization, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Thanks to John Sinton for providing the EPR samples and Craig Lundstrom and Frank Tepley for the Garrett samples. We thank Jackie Dixon and Jonathan Tucker for very constructive and helpful reviews. Fig. 6 was produced using GeoMapApp (http://www.geomapapp.org/), which is supported by the National Science Foundation and internal funding from Lamont-Doherty Earth Observatory. This work was supported by The National Science Foundation through grants OCE15-58798 (DG) and OCE15-58802 (PM).

References (71)

  • R. Hébert et al.

    Ultramafic and mafic rocks from the Garrett Transform Fault near 13°30'S on the East Pacific Rise: igneous petrology

    Earth Planet. Sci. Lett.

    (1983)
  • R. Hékinian et al.

    Volcanic activity and crust-mantle exposure in the ultrafast Garrett transform fault near 13°28'S in the Pacific

    Earth Planet. Sci. Lett.

    (1992)
  • M.M. Hirschmann

    Comparative deep Earth volatile cycles: the case for C recycling from exosphere/mantle fractionation of major (H2O, C, N) volatiles and from H2O/Ce, CO2/Ba and CO2/Nb exosphere ratios

    Earth Planet. Sci. Lett.

    (2018)
  • K.A. Kelley et al.

    Subduction cycling of U, Th, and Pb

    Earth Planet. Sci. Lett.

    (2005)
  • M.D. Kurz et al.

    Correlated helium, neon, and melt production on the super-fast spreading East Pacific Rise near 17°S

    Earth Planet. Sci. Lett.

    (2005)
  • J. Labidi et al.

    Negligible sulfur isotope fractionation during partial melting: evidence from Garrett transform fault basalts, implications for the late-veneer and the hadean matte

    Earth Planet. Sci. Lett.

    (2016)
  • J.J. Mahoney et al.

    Isotope and trace element characteristics of a super-fast spreading ridge: East Pacific Rise, 13-23°S

    Earth Planet. Sci. Lett.

    (1994)
  • S. Mallick et al.

    Constraints on the mantle mineralogy of an ultra-slow ridge: hafnium isotopes in abyssal peridotites and basalts from the 9-25°E Southwest Indian Ridge

    Earth Planet. Sci. Lett.

    (2015)
  • B. Marty

    The origins and concentrations of water, carbon, nitrogen and noble gases on Earth

    Earth Planet. Sci. Lett.

    (2012)
  • B. Marty et al.

    C/3He in volatile fluxes from the solid Earth: implications for carbon geodynamics

    Earth Planet. Sci. Lett.

    (1987)
  • B. Marty et al.

    Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: assessment of shallow-level fractionation and characterization of source composition

    Geochim. Cosmochim. Acta

    (1999)
  • S. Matthews et al.

    Constraining mantle carbon: CO2-trace element systematics in basalts and the roles of magma mixing and degassing

    Earth Planet. Sci. Lett.

    (2017)
  • P.J. Michael et al.

    The behavior and concentration of CO2 in the suboceanic mantle: inferences from undegassed ocean ridge and ocean island basalts

    Lithos

    (2015)
  • W.G.R. Miller et al.

    Estimating the carbon content of the deep mantle with Icelandic melt inclusions

    Earth Planet. Sci. Lett.

    (2019)
  • S. Newman et al.

    VolatileCalc: a silicate melt-H2O-CO2 solution model written in Visual Basic for Excel

    Comput. Geosci.

    (2002)
  • Y. Niu et al.

    Basaltic liquids and harzburgitic residues in the Garrett Transform: a case study at fast-spreading ridges

    Earth Planet. Sci. Lett.

    (1997)
  • D. Porcelli et al.

    The evolution of helium isotopes in the upper mantle and the generation of isotopic anomalies

    Earth Planet. Sci. Lett.

    (2008)
  • A. Prinzhofer et al.

    Stochastic melting of the marble cake mantle: evidence from local study of the East Pacific Rise at 12°50'N

    Earth Planet. Sci. Lett.

    (1989)
  • A. Rosenthal et al.

    Experimental determination of C, F, and H partitioning between mantle minerals and carbonated basalt, CO2/Ba and CO2/Nb systematics of partial melting, and the CO2 contents of basaltic source regions

    Earth Planet. Sci. Lett.

    (2015)
  • C.J. Russo et al.

    Mantle melting and magma supply to the Southeast Indian Ridge: the roles of lithology and melting conditions from U-series disequilibria

    Earth Planet. Sci. Lett.

    (2009)
  • K. Shimizu et al.

    Two-component mantle melting-mixing model for the generation of mid-ocean ridge basalts: implications for the volatile content of the Pacific upper mantle

    Geochim. Cosmochim. Acta

    (2016)
  • K. Shimizu et al.

    Evaluating the roles of melt-rock interaction and partial melting on the CO2/Ba ratios of MORB: implications for the CO2 budget in Earth's depleted mantle

    Geochim. Cosmochim. Acta

    (2019)
  • O. Shorttle

    Geochemical variability in MORB controlled by concurrent mixing and fractional crystallisation

    Earth Planet. Sci. Lett.

    (2015)
  • N. Stroncik et al.

    He, Ne and Ar isotope signatures of mid-ocean ridge basalts and their implications for upper mantle structure: a case study from the Mid-Atlantic Ridge at 4-12°S

    Geochim. Cosmochim. Acta

    (2016)
  • F.J. Tepley et al.

    U-series disequilibria in MORB from the Garrett Transform and implications for mantle melting

    Earth Planet. Sci. Lett.

    (2004)
  • Cited by (9)

    • High-C content and CO<inf>2</inf>/Ba ratio of the Earth's enriched upper mantle

      2023, Geochimica et Cosmochimica Acta
      Citation Excerpt :

      Given that the CO2/Ba ratio of the flux from the mantle into the exosphere (at mid-ocean ridge and ocean island) is higher than the CO2/Ba ratio in the exosphere, this suggests that the material that is returned into the mantle at subduction zones must have a CO2/Ba ratio that is even higher than that of the mantle (i.e., more efficiently recycling of C back into the convecting mantle relative to Ba). This is consistent with studies that have argued for significant recycling of subducted C into the convecting mantle (e.g., Javoy et al., 1982; Marty and Zimmermann, 1999; Sleep and Zahnle, 2001; Dasgupta and Hirschmann, 2010; Farsang et al., 2021; Graham and Michael, 2021). If this is the case, we speculate that our observation of high CO2/Ba ratio and C content in the Pacific E-MORB mantle source would be consistent with greater recycling of C relative to Ba through subduction (e.g., Hirschmann and Stolper, 1996; Donnelly et al., 2004).

    • Carbon stable isotope constraints on CO<inf>2</inf> degassing models of ridge, hotspot and arc magmas

      2022, Chemical Geology
      Citation Excerpt :

      For hotspot samples, the SiO2 content has been used to take into account the increase in CO2 solubility for alkalic melts undersaturated in SiO2 (Dixon, 1997). Carbon concentration and isotope data have been compiled from the following references: Alt et al. (1993), Aubaud (2002), Aubaud et al. (2004, 2005, 2006), Barry et al. (2014), Blank et al. (1993), Cartigny et al. (2001, 2008), Cushman et al. (2004), Des Marais and Moore (1984), Dixon et al. (1988, 1991, 1997, 2002, 2017), Dixon and Clague (2001), Exley et al. (1986), Fine and Stolper, (1985/1986), Gardner et al. (2016), Graham et al. (2018), Graham and Michael (2020), Gribble et al. (1996), Helo et al. (2011), Hochstaedter et al. (1990), Jackson et al. (2015), Javoy and Pineau (1991), Jendrzejewski (1994), Jendrzejewski et al. (1996a, 1996b), Jones et al. (2018, 2019), Kendrick et al. (2014, 2015), Le Roux et al. (2006), Le Voyer et al. (2015, 2017, 2019), Macpherson and Mattey (1994), Macpherson et al. (1999, 2005), Marty and Jambon (1987), Marty and Zimmermann (1999), Mattey et al. (1984, 1989), Michael and Graham (2015), Moore et al. (1977), Newman et al. (2000), Nishio et al. (1998, 1999), Pineau et al. (1976, 2004), Pineau and Javoy (1983), Pineau and Javoy (1994), Ribeiro et al. (2015), Saal et al. (2002), Sakai et al. (1984), Shaw et al. (2004, 2010), Shimizu et al. (2016), Simons et al. (2002), Soule et al. (2012), Standish et al. (2008), Stolper and Newman (1994), Wanless and Shaw (2012) and Workman et al. (2006). The compiled CO2 and H2O concentration data are presented in Fig. 4.

    • The origin of nitrogen in Earth's mantle: Constraints from basalts <sup>15</sup>N/<sup>14</sup>N and N<inf>2</inf>/<sup>3</sup>He ratios

      2022, Chemical Geology
      Citation Excerpt :

      The CO2/3He ratios of depleted MORB is ~5.0 × 108 (Marty and Zimmermann, 1999, Graham and Michael, 2021). Enriched MORB have higher CO2/3He, on average ~ 6.0 × 109 (Graham and Michael, 2021, Marty and Zimmermann, 1999, Péron et al., 2019). The relationship between CO2/3He and enrichment likely results from carbon recycling as an illustration of the marble-cake mantle (Marty and Zimmermann, 1999).

    View all citing articles on Scopus
    View full text