<< Previous article

Release of subducted sedimentary nitrogen throughout Earth’s mantle

P.H. Barry1,2,

1Fluids & Volatiles Laboratory, Scripps Institution of Oceanography, UC San Diego, La Jolla, California 92093-0244, USA
2Present address: Department of Earth Sciences, University of Oxford, OX1 3AN, UK

D.R. Hilton1

1Fluids & Volatiles Laboratory, Scripps Institution of Oceanography, UC San Diego, La Jolla, California 92093-0244, USA

Affiliations  |  Corresponding Author  |  Cite as

Barry, P.H., Hilton, D.R. (2016) Release of subducted sedimentary nitrogen throughout Earth’s mantle. Geochem. Persp. Let. 2, 148-159.

Geochemical Perspectives Letters v2, n2  |  doi: 10.7185/geochemlet.1615
Received 3 February 2016  |  Accepted 11 April 2016  |  Published 3 May 2016
Copyright © 2016 European Association of Geochemistry

Keywords: nitrogen, nitrogen isotopes, mantle, plume, subduction, recycling, noble gases



top

Abstract


The dynamic process of subduction represents the principal means to introduce chemical heterogeneities into Earth's interior. In the case of nitrogen (N) - atmosphere's most abundant gas - biological-activity converts N2 into ammonium ions (NH4+), which are chemically-bound within seafloor sediments and altered oceanic crust that comprise the subducting slab. Although some subducted N re-emerges via arc-related volcanism (Sano et al., 1998

Sano, Y., Takahata, N., Nishio, Y., Marty, B. (1998) Nitrogen recycling in subduction zones. Geophysical Research Letters 25, 2289-2292.

; Fischer et al., 2002

Fischer, T.P., Hilton, D.R., Zimmer, M.M., Shaw, A.M., Sharp, Z.D., Walker, J.A. (2002) Subduction and recycling of nitrogen along the Central American margin. Science 297, 1154-1157.

), the majority likely bypasses sub-arc depths (150-200 km) and supplies the deeper mantle (Li et al., 2007

Li, L., Bebout, G.E., Idleman, B.D. (2007) Nitrogen concentration and δ 15 N of altered oceanic crust obtained on ODP Legs 129 and 185: insights into alteration-related nitrogen enrichment and the nitrogen subduction budget. Geochimica et Cosmochimica Acta 71, 2344-2360.

; Mitchell et al., 2010

Mitchell, E.C., Fischer, T.P., Hilton, D.R., Hauri, E.H., Shaw, A.M., de Moor, J.M., Sharp, Z.D., Kazahaya, K. (2010) Nitrogen sources and recycling at subduction zones: Insights from the Izu‐Bonin‐Mariana arc. Geochemistry, Geophysics, Geosystems 11, 2.

; Johnson and Goldblatt, 2015

Johnson, B., Goldblatt, C. (2015) The nitrogen budget of Earth. Earth-Science Reviews 148, 150-173.

; Bebout et al., 2016

Bebout, G.E., Lazzeri, K.E., Geiger, C.A. (2016) Pathways for nitrogen cycling in Earth's crust and upper mantle: A review and new results for microporous beryl and cordierite. American Mineralogist 101, 7-24.

). However, the fate of subducted N remains enigmatic: is it incorporated by the shallow convecting mantle - the source of ridge volcanism, or is the deeper mantle - nominally associated with mantle plumes - its ultimate repository? Here, we present N-He-Ne-Ar isotope data for oceanic basalts from the Central Indian Ridge (CIR)-Réunion plume region to address this issue. All on-axis samples with depleted MORB mantle (DMM) affinities (3He/4He = 8 ± 1 RA; Graham, 2002

Graham, D.W. (2002) Noble gas isotope geochemistry of mid-ocean ridge and ocean island basalts: Characterization of mantle source reservoirs. RIMS 47, 247-317.

) have low N-isotopes (mean δ15N = -2.1 ‰), whereas those with plume-like 3He/4He display higher values (mean δ15N = 1.3 ‰). We explain these data within the framework of a new mantle reference model to predict a time-integrated net N regassing flux to the mantle of ~3.4 × 1010 mol/yr, with the plume-source mantle representing the preferential destination by a factor of 2-3. The model has implications for the present-day imbalance between N subducted at trenches and N emitted via arc-related volcanism, the N-content of Earth's early atmosphere, as well as relationships between N2 and the noble gases in mantle reservoirs, including 3He/4He-δ15N relationships in plume-derived lavas.

Figures and Tables

Table 1a Neon and argon isotope systematics of submarine basaltic glasses from the CIR (on-axis) and adjacent (off-axis) Gasitao Ridge, Three Magi Ridges and Abyssal Hill, and olivine separates from dunite xenoliths of Piton Chisny (Réunion).

Table 1b Nitrogen and helium isotope systematics, and relative He-N-Ar abundances of submarine basaltic glasses from the CIR (on-axis) and adjacent (off-axis) Gasitao Ridge, Three Magi Ridges and Abyssal Hill, and olivine separates from dunite xenoliths of Piton Chisny (Réunion).

Figure 1 (a) He-isotopes (3He/4He) versus N-isotopes, showing high (positive) δ15N values in all off-axis samples with >DMM-like 3He/4He values. (b) Histogram of N-isotopes measured along the CIR (in black) and its adjacent ridges (in red) relative to global DMM (green) and OIB (grey) averages (from Cartigny and Marty, 2013). (c) Argon isotopes (40Ar/36Ar) versus N-isotopes. (d) 4He/40Ar* (degassing proxy) versus N-isotopes suggests that high δ15N values are not produced by magmatic degassing (see text).

Figure 2 Extrapolated Ne ((21Ne/22Ne)EX, i.e. air-corrected 21Ne/22Ne values) versus 15N/14N values of CIR basalts, plotted together with binary mixing curves between a pre-solar nitrogen (PSN) component (15N/14N = 0.00227, δ15N = -373 ‰; Marty et al., 2012) and two mantle endmember components reflecting addition of 15N-enriched sedimentary MORB mantle (SMM; δ15N = +5 ‰) to DMM mantle (δ15N = -5 ‰). The curvature of the hyperbolic mixing lines is described by the r-value = (14N/22Ne)DMM/SMM /(14N/22Ne)PSN.

Figure 3 Nitrogen regassing fluxes (mol/yr) into the mantle as a function of time (Ma) since the onset of subduction. The total flux is equal to the sum of DMM and PLM fluxes. The proportion of N regassed into the plume-influenced mantle relative to the DMM (FPLM/FDMM) is shown in boxes at three given time intervals (1.1 Ga, 2.5 Ga, 3.9 Ga).

Table 1a Table 1b Figure 1 Figure 2 Figure 3

View all figures and tables  


Supplementary Figures and Tables

Figure S-1 Bathymetric map of the (on-axis) CIR and the adjacent (off-axis) Gasitao Ridge, Three Magi Ridges, Rodrigues Ridge and Abyssal Hill. Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999), R/V L'Atalante (Dyment et al., 2000), and R/V Hakuho-Maru (Okino et al., 2008); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008; Füri et al., 2011). The circles indicate the location of samples collected during the KNOX11RR (black = CIR on-axis; red = off-axis) and GIMNAUT (blue) cruises. The inset shows the Mascarene Islands (i.e. Réunion, Mauritius, and Rodrigues islands) to the west and the Rodrigues Triple Junction. Map modified from Füri et al., 2011.

Figure S-2 Neon three-isotope plot (20Ne/22Ne versus 21Ne/22Ne) of CIR basaltic glasses and Réunion xenoliths. Three trend-lines are superimposed on the data: (1) the air-solar mixing line; (2) the Réunion line (Hopp and Trieloff, 2005); (3) the DMM (2∏D43) line (Moreira et al., 1998). Uncertainties are at the 1σ level. Only samples with isotopes ratios >1 sigma from air are plotted.

Figure S-3 Ar-isotopes versus Ne-isotopes. Ne-Ar systematics are coupled for the majority of samples, i.e. air-contamination and mantle isotope anomalies are evident in the same samples.

Figure S-4 N-isotopes (δ15N) versus air-normalised He/Ne values (= 4He/20NeS/4He/20NeA). High He/Ne values relative to air (>100) suggest that all CIR basalt samples have undergone minimal air-contamination. Binary mixing trajectories are shown to postulated endmember values for ‘plume = seds’ (δ15N = +1.5‰; (4He/20Ne)Sample/(4He/20Ne)Air > 104) and ‘hybrid’ (δ15N = -2 ‰; (4He/20Ne)Sample/(4He/20Ne)Air > 104).

Figure S-5 (21Ne/22Ne)EX (i.e. air-corrected 21Ne/22Ne values) versus 4He/3He values of CIR basalts and Réunion xenoliths, plotted together with binary mixing curves between a primordial mantle endmember (PRIM) and a DMM-like component. The curvature of the hyperbolic mixing lines is described by r = (3He/22Ne)DMM/(3He/22Ne)PRIM. In addition, data fields are shown for Réunion lavas (Hanyu et al., 2001) and Iceland subglacial basalts (Füri et al., 2010).

Table S-1 Mantle regassing sensitivity to various input parameters.

Figure S-1 Figure S-2 Figure S-3 Figure S-4 Figure S-5 Table S-1

View all supplementary figures and tables  


top

Letter


The CIR-Réunion plume system is a classic modern example of oblique plume-ridge interaction, with the present-day plume centred at Réunion Island and three submarine off-ridge segments, the Rodriguez, Three Magi and Gasitao ridges, connecting with the CIR ridge axis located ~1100 km to the east. These west-to-east trending ridges were formed by volcanism above a channel of Réunion hotspot mantle as the CIR migrated northeast over and away from the plume (Morgan, 1978

Morgan, W.J. (1978) Rodriguez, Darwin, Amsterdam, a second type of hot spot island. Journal of Geophysical Research 83, 5355–5360.

). He-isotopes – the canonical tracer of mantle plume involvement in petrogenesis – are higher (>9 RA) than typical DMM values of 8 ± 1 RA (where RA = air-like 3He/4He) at Réunion Island and along the off-axis ridges. In contrast, the CIR exhibits DMM-like 3He/4He values along axis, except at the point where the projection of the submarine ridges meets the ridge axis ~19.9 °S (Füri et al., 2011

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J.M.D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

; Fig. S-1).

New nitrogen isotope and abundance analyses, along with accompanying neon and argon data, of basaltic glasses from the CIR axis between 16.7 °S and 20.6 °S, and from the off-axis ridges to the west of the CIR (Fig. S-1) are presented in Table 1a and 1b. In addition, we report Ne and Ar isotope and abundance data on olivine separates of a suite of cumulate dunite xenoliths from Réunion Island. Samples were processed by vacuum crushing with released gases analysed using a noble gas mass spectrometer (Barry et al., 2012

Barry, P.H., Hilton, D.R., Halldórsson, S.A., Hahm, D., Marti, K. (2012) High precision nitrogen isotope measurements in oceanic basalts using a static triple collection noble gas mass spectrometer. Geochemistry Geophysics Geosystems 13, Q01019.

). All samples have been analysed previously for He isotopes, major/minor and trace element chemistry (Füri et al., 2011

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J.M.D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

).


Table 1a Neon and argon isotope systematics of submarine basaltic glasses from the CIR (on-axis) and adjacent (off-axis) Gasitao Ridge, Three Magi Ridges and Abyssal Hill, and olivine separates from dunite xenoliths of Piton Chisny (Réunion).
Sample20Ne/20Ne21Ne/20Ne20Ne ×40Ar/36Ar[40Ar] ×
10-9 cm3 STP/g10-9 cm3STP/g
On-Axis




D1-110.10 ± 0.110.0316 ± 0.00030.67 ± 0.012290.6 ± 0.12416 ± 0.3
D1-1 (Dup)-----
D3-19.90 ± 0.100.0291 ± 0.00012.28 ± 0.025296.1 ± 0.3765 ± 0.5
D2-19.85 ± 0.100.0295 ± 0.00032.21 ± 0.030349.4 ± 0.31415 ± 0.9
D8-29.93 ± 0.100.0302 ± 0.00020.76 ± 0.0173520 ± 12724 ± 0.2
D9-2-----
D15-110.34 ± 0.110.0327 ± 0.00040.55 ± 0.0082193 ± 3.31968 ± 2.6
D14-110.34 ± 0.120.0324 ± 0.00080.25 ± 0.0192413 ± 5.81326 ± 0.3
D14-1 (Dup)10.43 ± 0.110.0331 ± 0.00040.25 ± 0.0121514 ± 3.9761 ± 0.7
DR10-1




D13-110.72 ± 0.120.0340 ± 0.00060.20 ± 0.0233519 ± 12722 ± 0.2
D13-1 (Dup)11.32 ± 0.130.0401 ± 0.00070.10 ± 0.0048487 ± 431272 ± 0.5






Off-Axis




Three Magi Ridges




D22-19.89 ± 0.100.0296 ± 0.00021.89 ± 0.011408.5 ± 0.61903 ± 0.8
D22-1 (Dup #1)-----
D22-1 (Dup #2)-----
D26-29.87 ± 0.100.0296 ± 0.00021.15 ± 0.025554.6 ± 0.51268 ± 0.9






Gasitao Ridge




D20-5 9.84 ± 0.230.0294 ± 0.00130.17 ± 0.0201725 ± 12421 ± 0.1
D20-5 (Dup #1)10.05 ± 0.130.0304 ± 0.00090.12 ± 0.0071786 ± 4.8383 ± 0.1
D20-5 (Dup #2)10.05 ± 0.120.0317 ± 0.00100.14 ± 0.0131576 ± 5.2401 ± 0.3
D18-19.89 ± 0.110.0290 ± 0.00040.14 ± 0.006396.1 ± 3.016 ± 0.1






Abyssal Hill




D37-29.86 ± 0.100.0294 ± 0.00011.19 ± 0.022318.4 ± 0.4849 ± 0.7






Réunion Island




CH07-0110.49 ± 0.110.0306 ± 0.00050.30 ± 0.0061294 ± 3.0776 ± 0.2
CH07-0210.00 ± 0.110.0296 ± 0.00040.33 ± 0.007501.2 ± 1.0327 ± 0.2
CH07-0410.09 ± 0.120.0314 ± 0.00140.06 ± 0.007705.1 ± 3.5125 ± 0.1
CH07-0710.26 ± 0.110.0296 ± 0.00080.20 ± 0.006831.3 ± 2.0308 ± 0.1
Download in Excel

Table 1b Nitrogen and helium isotope systematics, and relative He-N-Ar abundances of submarine basaltic glasses from the CIR (on-axis) and adjacent (off-axis) Gasitao Ridge, Three Magi Ridges and Abyssal Hill, and olivine separates from dunite xenoliths of Piton Chisny (Réunion).
Sample[N2] × 10-6cm3 STP/gaδ15N (‰)bN2/Arc3He/4He (R/RA)d4He/40Ar*






On-Axis




D1-15.56-1.93 ± 0.911328.11 ± 0.11-
D1-1 (Dup)5.21-1.25 ± 0.7199.9--
D3-14.97-1.81 ± 1.1368.57.91 ± 0.02-
D2-1---8.19 ± 0.0962 ± 0.2
D8-221.1-3.81 ± 0.511567.08 ± 0.147.5 ± 0.9
D9-210.51.16 ± 0.592677.25 ± 0.08-
D15-138.7-2.34 ± 0.552818.68 ± 0.018.4 ± 0.3
D14-116-2.01 ± 0.382388.46 ± 0.026.9 ± 0.4
D14-1 (Dup)----10 ± 0.4
DR10-148-0.10 ± 0.6347.910.31 ± 0.06-
D13-120.9-1.99 ± 0.5385.68.26 ± 0.033.0 ± 0.1
D13-1 (Dup)14.1-2.68 ± 0.51131-6.0 ± 2.6






Off-Axis




Three Magi Ridges




D22-174.21.71 ± 0.4582.89.40 ± 0.0623 ± 0.1
D22-1 (Dup #1)78.71.80 ± 0.52127--
D22-1 (Dup #2)92.61.74 ± 0.48104--
D26-215.20.89 ± 0.831299.51 ± 0.025.2 ± 0.1






Gasitao Ridge




D20-5 ---8.28 ± 0.0511 ± 1.3
D20-5 (Dup #1)----9.3 ± 0.4
D20-5 (Dup #2)----11 ± 0.6
D18-12.341.14 ± 1.472449.09 ± 0.0631 ± 0.9






Abyssal Hill




D37-2---9.67 ± 0.1718 ± 0.1






Réunion Island




CH07-01---13.95 ± 0.251.3 ± 0.1
CH07-02---13.66 ± 0.221.5 ± 0.1
CH07-04---14.09 ± 0.232.0 ± 0.1
CH07-07---13.58 ± 0.151.5 ± 0.1

a N2 concentration measurements are accurate within 3 %, based on the reproducibility of standards.
b Uncertainties on δ15N are 1σ. Blank subtractions and a comprehensive CO correction have been applied to all δ15N results.
c All N2/Ar uncertainties are less than 10 %. Blank subtractions have been applied to all N2/Ar results.
d Data previously reported in Füri et al., 2011 Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J.M.D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207. .

Download in Excel

Nitrogen isotope results are presented in the δ15N notation (where δ15N = ((15N/14Nsample/15N/14Nair)-1) × 1000) and plotted against He isotopes in Figure 1a, and displayed in comparison to the δ15N database of ocean basalts in Figure 1b. We highlight the following key features of the N-isotope results. First, all on-axis samples (with the exception of D9-2; δ15N = 1.16 ‰) have negative δ15N values. The highest and lowest values are -0.10 and -3.8 ‰ giving an on-axis δ15N mean value of -2.1 ± 1.1 ‰ (1σ; n = 7) or -1.7 ± 1.6 ‰ (n = 8 if D9-2 is included). Notably, sample D9-2 has a 3He/4He ratio (7.25 RA), which falls in the nominal DMM range (8 ± 1 RA; Graham, 2002

Graham, D.W. (2002) Noble gas isotope geochemistry of mid-ocean ridge and ocean island basalts: Characterization of mantle source reservoirs. RIMS 47, 247-317.

) characteristic of most samples on the CIR ridge-axis (Füri et al., 2011

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J.M.D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

). The only on-axis sample (DR10-1) of the present sample suite with a 3He/4He value higher than DMM (3He/4He = 10.31 RA) is from the region where the projection of the off-axis ridge impinges the spreading centre (~19.9 °S): it has a δ15N value of -0.10 ‰ – the second highest value of the on-axis samples. Second, the three off-axis samples – all with 3He/4He >9 RA – have positive δ15N values, ranging from +0.89 to +1.80 ‰; with a mean value of 1.3 ± 0.7 ‰ (1σ). Thus, with the exception of sample D9-2, there is a clear distinction between relatively low δ15N values associated with DMM-like He-isotopes on the ridge axis and relatively high δ15N values associated with plume-like 3He/4He values off-axis. Finally, we point out that with the exception of sample D8-2, all δ15N values, irrespective of location on- or off-axis, are higher than the range nominally associated with DMM (δ15N = -5 ± 2 ‰; Fig. 1b). Thus, samples of this study have experienced enrichment in 15N compared to the majority of basalts erupted at ridge axes worldwide.


Figure 1 (a) He-isotopes (3He/4He) versus N-isotopes, showing high (positive) δ15N values in all off-axis samples with >DMM-like 3He/4He values. (b) Histogram of N-isotopes measured along the CIR (in black) and its adjacent ridges (in red) relative to global DMM (green) and OIB (grey) averages (from Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

). (c) Argon isotopes (40Ar/36Ar) versus N-isotopes. (d) 4He/40Ar* (degassing proxy) versus N-isotopes suggests that high δ15N values are not produced by magmatic degassing (see text).
Full size image | Download in Powerpoint

There are three processes capable of producing high (>DMM) δ15N signatures in CIR basalts: (1) assimilation of existing crust during magma eruption, likely also involving incorporation of air (δ15N = 0 ‰), (2) mass-dependent fractionation related to magmatic degassing (Cartigny et al., 2001

Cartigny, P., Harris, J.W., Javoy, M. (2001) Diamond genesis, mantle fractionations and mantle nitrogen content: a study of δ13 C–N concentrations in diamonds. Earth and Planetary Science Letters 185, 85-98.

), and/or (3) recycling of oceanic sediments and/or oceanic crust (δ15N ~+5 to +7 ‰) into the CIR mantle source region producing the melts (Marty and Dauphas, 2003

Marty, B., Dauphas, N. (2003) The nitrogen record of crust-mantle interaction and mantle convection from Archean to present. Earth and Planetary Science Letters 206, 397– 410.

). The latter possibility implies that δ15N is a feature intrinsic to the mantle source, whereas the former two are predominantly shallow-level phenomena that act to mask primary (source) δ15N signatures.

There is little evidence in either the on- or off-axis N database (Table 1a and 1b) for a correlation between high N-contents – possibly reflecting crustal assimilation and/or air addition – and high δ15N values. For example, there is a 30-fold difference in N-content between off-axis samples D22-1 and D18-1 yet they have indistinguishable δ15N values. Similarly, two of the three highest N-content on-axis samples (D8-2 and D15-1) have the lowest δ15N values of this suite. Both observations are inconsistent with addition of air and/or crustal N, which would act to increase both δ15N and N-content of samples. Furthermore, a plot of 40Ar/36Ar versus δ15N (Fig. 1c) shows no correlation for either on- or off-axis samples: all on-axis samples fall within 2σ of the mean value of -2.1 ‰, yet 40Ar/36Ar values range between 8500 and close to the atmospheric value (298.6; Lee et al., 2006

Lee, J.Y., Marti, K., Severinghaus, J.P., Kawamura, K., Yoo, H.S., Lee, J.B., Kim, J.S. (2006) A redetermination of the isotopic abundances of atmospheric Ar. Geochimica et Cosmochimica Acta 70, 4507-4512.

), i.e. samples with low 40Ar/36Ar values do not have δ15N values closer to air (0 ‰). Finally, off-axis samples are characterised by markedly different 4He/40Ar* ratios – a parameter sensitive to magma degassing due to the factor of ~10 difference in solubility between He and Ar in basaltic melt (Lux, 1987

Lux, G. (1987) The behavior of noble gases in silicate liquids: Solution, diffusion, bubbles and surface effects, with applications to natural samples. Geochimica et Cosmochimica Acta 51, 1549-1560.

). Although 4He/40Ar* ratios vary between 5-31 for samples D18-1 and D26-2, respectively, their δ15N values are virtually identical at 1.14 and 0.89 ‰ (Fig. 1d). Thus, we conclude that the range in δ15N values reported here reflects intrinsic mantle source features, likely related to subduction of 15N-enriched oceanic sediments and/or oceanic crust.

The neon isotope characteristics of CIR lavas are plotted on a traditional three-isotope Ne diagram (Fig. S-2). We note that all samples lie intermediate to air-DMM and air-Réunion mixing trajectories with no clear distinction between on- and off-axis samples. Thus, all CIR basalts comprise three Ne components: air, a primitive/solar (mantle) Ne component and in-situ nucleogenic 21Ne, which has in-grown over time. We subtract the air-derived Ne to yield a mantle 21Ne/22Ne ratio (the so-called extrapolated value – (21Ne/22NeEX); see Supplementary Information for details). This approach allows us to assess mantle neon features of both on- and off-axis samples without compromise of air contamination, and to compare Ne isotopes to corresponding N-isotope variations.

In Figure 2, we plot the extrapolated Ne-isotope values (21Ne/22NeEX) of all basalts against measured δ15N together with endmember compositions for (1) pre-solar nitrogen (PSN), (2) depleted MORB mantle (DMM) and (3) sediment-modified mantle (SMM). Pure DMM has a δ15N value of ~-5 ± 2 ‰, which reflects mixing between N incorporated at the time of planetary accretion – small amounts of primordial N (δ15N " target="_blank" >40 ‰), possibly as low as proto-solar-nebula (PSN) nitrogen (δ15N = -383 ± 8 ‰), and nitrogen introduced into the mantle by long-term recycling of atmospheric (= 0 ‰) and/or sediment-derived (= ~+5 to +7 ‰) components (Marty, 2012

Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.

). SMM nitrogen reflects superimposition and dominance of recently-added sedimentary N relative to pure DMM, as observed at convergent margins such as Costa Rica (Fischer et al., 2002

Fischer, T.P., Hilton, D.R., Zimmer, M.M., Shaw, A.M., Sharp, Z.D., Walker, J.A. (2002) Subduction and recycling of nitrogen along the Central American margin. Science 297, 1154-1157.

). For Ne-isotopes, a DMM 21Ne/22Ne endmember value (0.0594) is derived by extrapolating the MORB trajectory (Sarda et al., 1988

Sarda, P., Staudacher, T., Allègre, C.J. (1988) Neon isotopes in submarine basalts. Earth and Planetary Science Letters 91, 73–88.

) to the 20Ne/22Ne value of Ne-B (= 12.5) and noting the 21Ne/22Ne value at that point. Solar Ne is defined by the solar wind value of 0.03118 (Trieloff and Kunz, 2005

Trieloff, M., Kunz, J. (2005) Isotope systematics of noble gases in the Earth's mantle: possible sources of primordial isotopes and implications for mantle structure. Physics of the Earth and Planetary Interiors 148, 13-38.

). We assume that subduction of oceanic sediments and crust does not introduce neon (or helium) into the mantle (Hilton et al., 1992

Hilton, D.R., Hoogewerff, J.A., Van Bergen, M.J., Hammerschmidt, K. (1992) Mapping magma sources in the east Sunda-Banda arcs, Indonesia: constraints from helium isotopes. Geochimica et Cosmochimica Acta 56, 851-859.

) so that DMM and SMM have the same 21Ne/22NeEX value.


Figure 2 Extrapolated Ne ((21Ne/22Ne)EX, i.e. air-corrected 21Ne/22Ne values) versus 15N/14N values of CIR basalts, plotted together with binary mixing curves between a pre-solar nitrogen (PSN) component (15N/14N = 0.00227, δ15N = -373 ‰; Marty et al., 2012

Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.

) and two mantle endmember components reflecting addition of 15N-enriched sedimentary MORB mantle (SMM; δ15N = +5 ‰) to DMM mantle (δ15N = -5 ‰). The curvature of the hyperbolic mixing lines is described by the r-value = (14N/22Ne)DMM/SMM/(14N/22Ne)PSN.
Full size image | Download in Powerpoint

The coupled δ15N-21Ne/22NeEX characteristics of both sets of basaltic glasses (i.e. on- and off-axis) are compatible with mixing between two distinct endmember compositions (Fig. 2). One endmember, with low δ15N and low 21Ne/22NeEX, is common to both sample suites, with solar gas being the most plausible candidate. In contrast, the second endmember differs between on- and off-axis samples. In the case of CIR on-axis basalts, the endmember has a DMM 21Ne/22NeEX value (0.060) and δ15N = -2.1 ‰ whereas the off-axis samples project to an endmember with a higher δ15N (~1.3 ‰) but the same DMM-like 21Ne/22NeEX value. Notably, both on- and off-axis endmembers lie on the projection between pure DMM and SMM with the proportion of N derived from SMM clearly greater in off-axis samples versus on-axis samples (Fig. 2).

The curvature of the binary mixing trajectories in Figure 2 has important implications for the overall recycling efficiency of N relative to Ne, as it is controlled by the relative 14N/22Ne ratios of the two endmembers (r-value). For both on- and off-axis samples, r ≥ 1000, which indicates either higher relative 14N or lower 22Ne contents in the mixed DMM-SMM versus the PSN endmember: this indicates that N is recycled at least 103 times more efficiently than Ne into the CIR mantle. Preferential deep recycling of N relative to Ne is consistent with recent theoretical and experimental predictions that indicate that N is stabilised as ammonium under subduction redox conditions (fO2 < QFM; Mikhail and Sverjensky, 2014

Mikhail, S., Sverjensky, D.A. (2014) Nitrogen speciation in upper mantle fluids and the origin of Earth's nitrogen-rich atmosphere. Nature Geoscience 7, 816-819.

) and therefore behaves as a large ion lithophile element whereby it can substitute into K-bearing silicate minerals and gain stability in the downgoing slab (Li et al., 2013

Li, Y., Wiedenbeck, M., Shcheka, S., Keppler, H. (2013) Nitrogen solubility in upper mantle minerals. Earth and Planetary Science Letters 377, 311-323.

). Moreover, mass balance arguments suggest Ne is efficiently recycled back to the surface during the subduction process (Holland and Ballentine, 2006

Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.

; Marty, 2012

Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.

).

The CIR data reaffirms the considerable heterogeneity in mantle δ15N signatures (e.g., Marty and Dauphas, 2003

Marty, B., Dauphas, N. (2003) The nitrogen record of crust-mantle interaction and mantle convection from Archean to present. Earth and Planetary Science Letters 206, 397– 410.

). If we assume that on-axis samples can be used to approximate the N composition of the DMM and that off-axis samples best represent the plume mantle (PLM), then CIR data can be used within the framework of a newly-constructed reference model to place constraints on N regassing fluxes into the mantle (i.e. both absolute flux values as well as the relative proportions regassing the DMM and PLM reservoirs). Our model is based on a layered mantle reservoir concept, similar to that described for noble gases by O’Nions and Tolstikhin (1996)

O'Nions, R.K., Tolstikhin, I.N. (1996) Limits on the mass flux between lower and upper mantle and stability of layering. Earth and Planetary Science Letters 139, 213-222.

and Gonnermann and Mukhopadhyay (2009)

Gonnermann, H.M., Mukhopadhyay, S. (2009) Preserving noble gases in a convecting mantle. Nature 459, 560-563.

. It makes the following assumptions: (1) the N regassing flux commences at the onset of subduction and is constant through time; (2) a constant proportion of the total N regassing flux (F) is subducted into each respective mantle reservoir (defined here as FDMM and FPLM) and instantaneously mixes in each reservoir; (3) there is a constant mantle degassing flux of 5 × 109 mol/yr (Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

), derived from the DMM:PLM reservoirs in the proportion 85:15, respectively (Ito et al., 2003

Ito, G., Lin, J., Graham, D. (2003) Observational and theoretical studies of the dynamics of mantle plume–mid‐ocean ridge interaction. Reviews of Geophysics 41, 4.

); (4) the present-day mantle N-content ([N]) is ~0.27 ± 0.16 ppm (Marty, 2012

Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.

; Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

); and (5) the initial mantle δ15N is -40 ‰ (PSN-like; Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

) which is unmodified by any degassing prior to the onset of subduction. The DMM and PLM reservoirs are considered to have evolved to their current δ15N and [N] compositions, given by the CIR data, as a consequence of additions of different amounts of subducted sedimentary N (δ15N = +5 ‰).

A DMM mass of ~9 × 1026 g (e.g., Anderson, 1989

Anderson, D.L. (1989) Theory of the Earth. Blackwell Scientific Publications, Boston, http://resolver.caltech.edu/CaltechBOOK:1989.001.

) corresponds to a total inventory of 17.3 × 1018 mol N in the modern-day DMM. Assuming that the mantle degassing flux (5 × 109 mol/yr; Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

) includes all non-arc related N (i.e. N degassed from the mantle) and is 85 % DMM-derived (4.3 × 109 mol/yr) and 15 % plume mantle-derived (7.5 × 108 mol/yr), then a total of 10.6 × 1018 mol N has been degassed from the DMM source over an estimated 2.5 Ga since subduction commenced (e.g., Kusky et al., 2001

Kusky, T.M., Li, J.H., Tucker, R.D. (2001) The Archean Dongwanzi ophiolite complex, North China Craton: 2.505-billion-year-old oceanic crust and mantle. Science 292, 1142-1145.

). For a total N-content of 27.9 × 1018 mol N before degassing, the present-day N-isotopic composition of -2.1 ‰ must represent admixture between 23.5 × 1018 moles of subduction-derived N (δ15N = +5 ‰), or ~84 % of the total N, and 4.4 × 1018 moles (or 16 %) of PSN (δ15N = -40 ‰). Thus, the time integrated DMM regassing rate (FDMM) is 9.4 x 109 moles/yr.

Using the same approach for the PLM – assumed to be represented by off-axis high 3He/4He plume melts with an average δ15N = 1.3 ‰, then a mass of ~3.3 × 1027 g would contain 64 × 1018 mol N in the modern day plume-influenced mantle. Again assuming a constant mantle degassing flux of 5 × 109 mol/yr (Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

), of which ~15 % (i.e. 7.5 × 108 mol/yr) is derived from this reservoir (Ito et al., 2003

Ito, G., Lin, J., Graham, D. (2003) Observational and theoretical studies of the dynamics of mantle plume–mid‐ocean ridge interaction. Reviews of Geophysics 41, 4.

), then a total of 1.9 × 1018 mol N has been degassed from the plume-influenced mantle over 2.5 Ga. The total N-content of 65.9 × 1018 mol requires 61 × 1018 moles N (~92 % of the total) of subduction-derived N (δ15N = +5 ‰) mixing with 8 % (i.e. 5.4 × 1018 moles) of PSN (δ15N = -40 ‰) to give the present-day N-isotopic composition of 1.3 ‰. Thus, the model predicts a regassing N-flux (FPLM) of ~2.4 × 1010 mol/yr since the Archean (2.5 Ga). Importantly, in order to satisfy the N systematics for both reservoirs, approximately 2.6 times more N (i.e. FPLM/FDMM) must be subducted into the plume-influenced mantle. If subduction is assumed to have initiated earlier in the geological record (e.g., 3.9 Ga; Condie and Pease, 2008

Condie, K.C., Pease, V. (2008) When did plate tectonics begin on planet Earth? Geological Society of America Special Paper 440, Boulder Colorado, USA.

), then the FPLM/FDMM ratio decreases to ~2.2. Conversely, if an estimated modern day N input flux of 7 × 1010 mol/yr (Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

) is adopted, it would take the Earth only 1.1 Ga to evolve to its current N composition, with FPLM/FDMM = 3.2 (Fig. 3; Table S-1).


Figure 3 Nitrogen regassing fluxes (mol/yr) into the mantle as a function of time (Ma) since the onset of subduction. The total flux is equal to the sum of DMM and PLM fluxes. The proportion of N regassed into the plume-influenced mantle relative to the DMM (FPLM/FDMM) is shown in boxes at three given time intervals (1.1 Ga, 2.5 Ga, 3.9 Ga).
Full size image | Download in Powerpoint

Selection of different δ15N endmembers for mantle reservoirs or lower initial solar values as input parameters has only a minor effect on model results (Table S-1). If more extreme N-isotope estimates of -4 ‰ and +3 ‰ are taken to approximate DMM and PLM sources, respectively (Dauphas and Marty, 1999

Dauphas, N., Marty, B. (1999) Heavy nitrogen in carbonatites of the Kola Peninsula: A possible signature of the deep mantle. Science 286, 2488-2490.

), then an identical PLM regassing flux of 2.4 × 1010 mol/yr is calculated for 2.5 Ga with only a marginally higher FPLM/FDMM = 2.8. Alternatively, a significantly higher estimate for the mantle N-content of ~36 ppm (Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

) results in a 2.5 Ga regassing flux of 3.9 × 1012 mol/yr with a FPLM/FDMM = 4. Significantly, all model scenarios point to long-term cycling of subducted N into Earth’s mantle with preferential storage in the deep (plume) mantle (i.e. FPLM/FDMM >2). Thus nitrogen not only breaches the subduction barrier beneath arcs but is also transported into the (deep mantle) source region that supplies high 3He/4He mantle plumes.

Our new reference model has a number of far-reaching implications. First, given the high N recycling efficiencies into DMM (84 %) and PLM (92 %) reservoirs, respectively (this work), any losses of slab-derived N to the atmosphere at volcanic arcs (e.g., by oxidation back to molecular form in the mantle wedge) must be relatively minor. Thus, our model can explain the present-day gross imbalance between the regassing N flux subducted via trenches and the degassing N flux returned to the surface via arc-related volcanism (Hilton et al., 2002

Hilton, D.R., Fischer, T.P., Marty, B. (2002) Noble gases and volatile recycling at subduction zones. In: Porcelli, D., Ballentine, C.J., Wieler, R. (Eds.) Noble Gases in Cosmochemistry and Geochemistry. Mineralogical Society of America 47 319-370.

; Busigny et al., 2011

Busigny, V., Cartigny, P., Philippot, P. (2011) Nitrogen isotopes in ophiolitic metagabbros: A re-evaluation of modern nitrogen fluxes in subduction zones and implication for the early Earth atmosphere. Geochimica et Cosmochimica Acta 75, 7502–7521.

). Second, prior to initiation of subduction-related recycling of N, the total N content of the atmosphere must have been considerably higher. Our model indicates that the early atmospheric N content was ~2.1 x 1020 moles, or ~50 % higher than at the present-day. Such a high atmospheric N-content would lead to N-enhanced greenhouse warming, and can explain the lack of global glaciations in the early Earth due to the faint young Sun (Goldblatt et al., 2009

Goldblatt, C., Claire, M.W., Lenton, T.M., Matthews, A.J., Watson, A.J., Zahnle, K.J. (2009) Nitrogen-enhanced greenhouse warming on early Earth. Nature Geoscience 2, 891-896.

). Third, subduction of sedimentary N into the deep mantle would lead to a close relationship between high δ15N values and high 3He/4He ratios, the canonical geochemical tracer of deep mantle plumes. Recent work on hyaloclastic basaltic glasses in Iceland (Halldórsson et al., 2016

Halldórsson, S.A., Hilton, D.R., Barry, P.H., Füri, E., Gronvold, K. (2016) Recycling of crustal material by the Iceland mantle plume: New evidence from nitrogen elemental and isotope systematics of subglacial basalts. Geochimica et Cosmochimica Acta 176, 206-226.

) reveals a strong coupling between 3He/4He and δ15N signatures, consistent with recycling of crustal material to the mantle plume source. Finally, we highlight remarkably similar recycling efficiencies between N (84-92 %; this work) and heavy noble gases (73-87 %; Holland and Ballentine, 2006

Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.

; Parai and Mukhopadhyay, 2012

Parai, R., Mukhopadhyay, S. (2012) How large is the subducted water flux? New constraints on mantle regassing rates. Earth and Planetary Science Letters 317, 396-406.

) despite fundamentally different subduction mechanisms: N is fixed as NH4+ and bound in recycled sediments and/or oceanic crust, whereas Kr and Xe undergo no such chemical transformation and instead are simply dissolved in pore fluids and/or in unoccupied amphibole A-sites within subducted oceanic crust (e.g., Holland and Ballentine, 2006

Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.

; Jackson et al., 2015

Jackson, C.R., Parman, S.W., Kelley, S.P., Cooper, R.F. (2015) Light noble gas dissolution into ring structure-bearing materials and lattice influences on noble gas recycling. Geochimica et Cosmochimica Acta 159, 1-15.

).

top

Acknowledgements


This work was supported by NSF grants EAR-0651097 and OCE-0726573. The KNOX11RR cruise was funded by UC Ship Funds. We thank Sami Mikhail, Ben Johnson and anonymous reviewers for their constructive and insightful comments, and Graham Pearson for editorial handling.

Editor: Graham Pearson

top

References



Anderson, D.L. (1989) Theory of the Earth. Blackwell Scientific Publications, Boston, http://resolver.caltech.edu/CaltechBOOK:1989.001.
Show in context

A DMM mass of ~9 × 1026 g (e.g., Anderson, 1989) corresponds to a total inventory of 17.3 × 1018 mol N in the modern-day DMM.
View in article


Barry, P.H., Hilton, D.R., Halldórsson, S.A., Hahm, D., Marti, K. (2012) High precision nitrogen isotope measurements in oceanic basalts using a static triple collection noble gas mass spectrometer. Geochemistry Geophysics Geosystems 13, Q01019.
Show in context

Samples were processed by vacuum crushing with released gases analysed using a noble gas mass spectrometer (Barry et al., 2012).
View in article


Bebout, G.E., Lazzeri, K.E., Geiger, C.A. (2016) Pathways for nitrogen cycling in Earth's crust and upper mantle: A review and new results for microporous beryl and cordierite. American Mineralogist 101, 7-24.
Show in context

Although some subducted N re-emerges via arc-related volcanism (Sano et al., 1998; Fischer et al., 2002) the majority likely bypasses sub-arc depths (150-200 km) and supplies the deeper mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in article


Busigny, V., Cartigny, P., Philippot, P. (2011) Nitrogen isotopes in ophiolitic metagabbros: A re-evaluation of modern nitrogen fluxes in subduction zones and implication for the early Earth atmosphere. Geochimica et Cosmochimica Acta 75, 7502–7521.
Show in context

Thus, our model can explain the present-day gross imbalance between the regassing N flux subducted via trenches and the degassing N flux returned to the surface via arc-related volcanism (Hilton et al., 2002; Busigny et al., 2011).
View in article


Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.
Show in context

Figure 1 [...] (b) Histogram of N-isotopes measured along the CIR (in black) and its adjacent ridges (in red) relative to global DMM (green) and OIB (grey) averages (from Cartigny and Marty, 2013).
View in article
It makes the following assumptions: (1) the N regassing flux commences at the onset of subduction and is constant through time; (2) a constant proportion of the total N regassing flux (F) is subducted into each respective mantle reservoir (defined here as FDMM and FPLM) and instantaneously mixes in each reservoir; (3) there is a constant mantle degassing flux of 5 × 109 mol/yr (Cartigny and Marty, 2013), derived from the DMM:PLM reservoirs in the proportion 85:15, respectively (Ito et al., 2003); (4) the present-day mantle N-content ([N]) is ~0.27 ± 0.16 ppm (Marty, 2012; Cartigny and Marty, 2013); and (5) the initial mantle δ15N is -40 ‰ (PSN-like; Cartigny and Marty, 2013) which is unmodified by any degassing prior to the onset of subduction.
View in article
Assuming that the mantle degassing flux (5 × 109 mol/yr; Cartigny and Marty, 2013) includes all non-arc related N (i.e. N degassed from the mantle) and is 85 % DMM-derived (4.3 × 109 mol/yr) and 15 % plume mantle-derived (7.5 × 108 mol/yr), then a total of 10.6 × 1018 mol N has been degassed from the DMM source over an estimated 2.5 Ga since subduction commenced (e.g., Kusky et al., 2001).
View in article
Again assuming a constant mantle degassing flux of 5 × 109 mol/yr (Cartigny and Marty, 2013), of which ~15 % (i.e. 7.5 × 108 mol/yr) is derived from this reservoir (Ito et al., 2003), then a total of 1.9 × 1018 mol N has been degassed from the plume-influenced mantle over 2.5 Ga.
View in article
Conversely, if an estimated modern day N input flux of 7 × 1010 mol/yr (Cartigny and Marty, 2013) is adopted, it would take the Earth only 1.1 Ga to evolve to its current N composition, with FPLM/FDMM = 3.2 (Fig. 3; Table S-1).
View in article
Alternatively, a significantly higher estimate for the mantle N-content of ~36 ppm (Cartigny and Marty, 2013) results in a 2.5 Ga regassing flux of 3.9 × 1012 mol/yr with a FPLM/FDMM = 4.
View in article


Cartigny, P., Harris, J.W., Javoy, M. (2001) Diamond genesis, mantle fractionations and mantle nitrogen content: a study of δ13 C–N concentrations in diamonds. Earth and Planetary Science Letters 185, 85-98.
Show in context

There are three processes capable of producing high (>DMM) δ15N signatures in CIR basalts: (1) assimilation of existing crust during magma eruption, likely also involving incorporation of air (δ15N = 0 ‰), (2) mass-dependent fractionation related to magmatic degassing (Cartigny et al., 2001), and/or (3) recycling of oceanic sediments and/or oceanic crust (δ15N ~+5 to +7 ‰) into the CIR mantle source region producing the melts (Marty and Dauphas, 2003).
View in article


Condie, K.C., Pease, V. (2008) When did plate tectonics begin on planet Earth? Geological Society of America Special Paper 440, Boulder Colorado, USA.
Show in context

If subduction is assumed to have initiated earlier in the geological record (e.g., 3.9 Ga; Condie and Pease, 2008), then the FPLM/FDMM ratio decreases to ~2.2.
View in article


Dauphas, N., Marty, B. (1999) Heavy nitrogen in carbonatites of the Kola Peninsula: A possible signature of the deep mantle. Science 286, 2488-2490.
Show in context

If more extreme N-isotope estimates of -4 ‰ and +3 ‰ are taken to approximate DMM and PLM sources, respectively (Dauphas and Marty, 1999), then an identical PLM regassing flux of 2.4 × 1010 mol/yr is calculated for 2.5 Ga with only a marginally higher FPLM/FDMM = 2.8.
View in article


Fischer, T.P., Hilton, D.R., Zimmer, M.M., Shaw, A.M., Sharp, Z.D., Walker, J.A. (2002) Subduction and recycling of nitrogen along the Central American margin. Science 297, 1154-1157.
Show in context

Although some subducted N re-emerges via arc-related volcanism (Sano et al., 1998; Fischer et al., 2002) the majority likely bypasses sub-arc depths (150-200 km) and supplies the deeper mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in article
SMM nitrogen reflects superimposition and dominance of recently-added sedimentary N relative to pure DMM, as observed at convergent margins such as Costa Rica (Fischer et al., 2002).
View in article


Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J.M.D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.
Show in context

In contrast, the CIR exhibits DMM-like 3He/4He values along axis, except at the point where the projection of the submarine ridges meets the ridge axis ~19.9 °S (Füri et al., 2011).
View in article
All samples have been analysed previously for He isotopes, major/minor and trace element chemistry (Füri et al., 2011).
View in article
Table 1b [...] d Data previously reported in Füri et al., 2011.
View in article
Notably, sample D9-2 has a 3He/4He ratio (7.25 RA), which falls in the nominal DMM range (8 ± 1 RA; Graham, 2002) characteristic of most samples on the CIR ridge-axis (Füri et al., 2011).
View in article


Graham, D.W. (2002) Noble gas isotope geochemistry of mid-ocean ridge and ocean island basalts: Characterization of mantle source reservoirs. RIMS 47, 247-317.
Show in context

All on-axis samples with depleted MORB mantle (DMM) affinities (3He/4He = 8 ± 1 RA; Graham, 2002) have low N-isotopes (mean δ15N = -2.1 ‰), whereas those with plume-like 3He/4He display higher values (mean δ15N = 1.3 ‰).
View in article
Notably, sample D9-2 has a 3He/4He ratio (7.25 RA), which falls in the nominal DMM range (8 ± 1 RA; Graham, 2002) characteristic of most samples on the CIR ridge-axis (Füri et al., 2011).
View in article


Goldblatt, C., Claire, M.W., Lenton, T.M., Matthews, A.J., Watson, A.J., Zahnle, K.J. (2009) Nitrogen-enhanced greenhouse warming on early Earth. Nature Geoscience 2, 891-896.
Show in context

Such a high atmospheric N-content would lead to N-enhanced greenhouse warming, and can explain the lack of global glaciations in the early Earth due to the faint young Sun (Goldblatt et al., 2009).
View in article


Gonnermann, H.M., Mukhopadhyay, S. (2009) Preserving noble gases in a convecting mantle. Nature 459, 560-563.
Show in context

Our model is based on a layered mantle reservoir concept, similar to that described for noble gases by O’Nions and Tolstikhin (1996) and Gonnermann and Mukhopadhyay (2009).
View in article


Halldórsson, S.A., Hilton, D.R., Barry, P.H., Füri, E., Gronvold, K. (2016) Recycling of crustal material by the Iceland mantle plume: New evidence from nitrogen elemental and isotope systematics of subglacial basalts. Geochimica et Cosmochimica Acta 176, 206-226.
Show in context

Recent work on hyaloclastic basaltic glasses in Iceland (Halldórsson et al., 2016) reveals a strong coupling between 3He/4He and δ15N signatures, consistent with recycling of crustal material to the mantle plume source.
View in article


Hilton, D.R., Hoogewerff, J.A., Van Bergen, M.J., Hammerschmidt, K. (1992) Mapping magma sources in the east Sunda-Banda arcs, Indonesia: constraints from helium isotopes. Geochimica et Cosmochimica Acta 56, 851-859.
Show in context

We assume that subduction of oceanic sediments and crust does not introduce neon (or helium) into the mantle (Hilton et al., 1992) so that DMM and SMM have the same 21Ne/22NeEX value.
View in article


Hilton, D.R., Fischer, T.P., Marty, B. (2002) Noble gases and volatile recycling at subduction zones. In: Porcelli, D., Ballentine, C.J., Wieler, R. (Eds.) Noble Gases in Cosmochemistry and Geochemistry. Mineralogical Society of America 47 319-370.
Show in context

Thus, our model can explain the present-day gross imbalance between the regassing N flux subducted via trenches and the degassing N flux returned to the surface via arc-related volcanism (Hilton et al., 2002; Busigny et al., 2011).
View in article


Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.
Show in context

Moreover, mass balance arguments suggest Ne is efficiently recycled back to the surface during the subduction process (Holland and Ballentine, 2006; Marty, 2012).
View in article
Finally, we highlight remarkably similar recycling efficiencies between N (84-92 %; this work) and heavy noble gases (73-87 %; Holland and Ballentine, 2006; Parai and Mukhopadhyay, 2012) despite fundamentally different subduction mechanisms: N is fixed as NH4+ and bound in recycled sediments and/or oceanic crust, whereas Kr and Xe undergo no such chemical transformation and instead are simply dissolved in pore fluids and/or in unoccupied amphibole A-sites within subducted oceanic crust (e.g., Holland and Ballentine, 2006; Jackson et al., 2015).
View in article


Ito, G., Lin, J., Graham, D. (2003) Observational and theoretical studies of the dynamics of mantle plume–mid‐ocean ridge interaction. Reviews of Geophysics 41, 4.
Show in context

It makes the following assumptions: (1) the N regassing flux commences at the onset of subduction and is constant through time; (2) a constant proportion of the total N regassing flux (F) is subducted into each respective mantle reservoir (defined here as FDMM and FPLM) and instantaneously mixes in each reservoir; (3) there is a constant mantle degassing flux of 5 × 109 mol/yr (Cartigny and Marty, 2013), derived from the DMM:PLM reservoirs in the proportion 85:15, respectively (Ito et al., 2003); (4) the present-day mantle N-content ([N]) is ~0.27 ± 0.16 ppm (Marty, 2012; Cartigny and Marty, 2013); and (5) the initial mantle δ15N is -40 ‰ (PSN-like; Cartigny and Marty, 2013) which is unmodified by any degassing prior to the onset of subduction.
View in article
Again assuming a constant mantle degassing flux of 5 × 109 mol/yr (Cartigny and Marty, 2013), of which ~15 % (i.e. 7.5 × 108 mol/yr) is derived from this reservoir (Ito et al., 2003), then a total of 1.9 × 1018 mol N has been degassed from the plume-influenced mantle over 2.5 Ga.
View in article


Jackson, C.R., Parman, S.W., Kelley, S.P., Cooper, R.F. (2015) Light noble gas dissolution into ring structure-bearing materials and lattice influences on noble gas recycling. Geochimica et Cosmochimica Acta 159, 1-15.
Show in context

Finally, we highlight remarkably similar recycling efficiencies between N (84-92 %; this work) and heavy noble gases (73-87 %; Holland and Ballentine, 2006; Parai and Mukhopadhyay, 2012) despite fundamentally different subduction mechanisms: N is fixed as NH4+ and bound in recycled sediments and/or oceanic crust, whereas Kr and Xe undergo no such chemical transformation and instead are simply dissolved in pore fluids and/or in unoccupied amphibole A-sites within subducted oceanic crust (e.g., Holland and Ballentine, 2006; Jackson et al., 2015).
View in article


Johnson, B., Goldblatt, C. (2015) The nitrogen budget of Earth. Earth-Science Reviews 148, 150-173.
Show in context

Although some subducted N re-emerges via arc-related volcanism (Sano et al., 1998; Fischer et al., 2002) the majority likely bypasses sub-arc depths (150-200 km) and supplies the deeper mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in article


Kusky, T.M., Li, J.H., Tucker, R.D. (2001) The Archean Dongwanzi ophiolite complex, North China Craton: 2.505-billion-year-old oceanic crust and mantle. Science 292, 1142-1145.
Show in context

Assuming that the mantle degassing flux (5 × 109 mol/yr; Cartigny and Marty, 2013) includes all non-arc related N (i.e. N degassed from the mantle) and is 85 % DMM-derived (4.3 × 109 mol/yr) and 15 % plume mantle-derived (7.5 × 108 mol/yr), then a total of 10.6 × 1018 mol N has been degassed from the DMM source over an estimated 2.5 Ga since subduction commenced (e.g., Kusky et al., 2001).
View in article


Lee, J.Y., Marti, K., Severinghaus, J.P., Kawamura, K., Yoo, H.S., Lee, J.B., Kim, J.S. (2006) A redetermination of the isotopic abundances of atmospheric Ar. Geochimica et Cosmochimica Acta 70, 4507-4512.
Show in context

Furthermore, a plot of 40Ar/36Ar versus δ15N (Fig. 1c) shows no correlation for either on- or off-axis samples: all on-axis samples fall within 2σ of the mean value of -2.1 ‰, yet 40Ar/36Ar values range between 8500 and close to the atmospheric value (298.6; Lee et al., 2006), i.e. samples with low 40Ar/36Ar values do not have δ15N values closer to air (0 ‰).
View in article


Li, L., Bebout, G.E., Idleman, B.D. (2007) Nitrogen concentration and δ 15 N of altered oceanic crust obtained on ODP Legs 129 and 185: insights into alteration-related nitrogen enrichment and the nitrogen subduction budget. Geochimica et Cosmochimica Acta 71, 2344-2360.
Show in context

Although some subducted N re-emerges via arc-related volcanism (Sano et al., 1998; Fischer et al., 2002) the majority likely bypasses sub-arc depths (150-200 km) and supplies the deeper mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in article


Li, Y., Wiedenbeck, M., Shcheka, S., Keppler, H. (2013) Nitrogen solubility in upper mantle minerals. Earth and Planetary Science Letters 377, 311-323.
Show in context

Preferential deep recycling of N relative to Ne is consistent with recent theoretical and experimental predictions that indicate that N is stabilised as ammonium under subduction redox conditions (fO2 < QFM; Mikhail and Sverjensky, 2014) and therefore behaves as a large ion lithophile element whereby it can substitute into K-bearing silicate minerals and gain stability in the downgoing slab (Li et al., 2013).
View in article


Lux, G. (1987) The behavior of noble gases in silicate liquids: Solution, diffusion, bubbles and surface effects, with applications to natural samples. Geochimica et Cosmochimica Acta 51, 1549-1560.
Show in context

Finally, off-axis samples are characterised by markedly different 4He/40Ar* ratios – a parameter sensitive to magma degassing due to the factor of ~10 difference in solubility between He and Ar in basaltic melt (Lux, 1987).
View in article


Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.
Show in context

Pure DMM has a δ15N value of ~-5 ± 2 ‰, which reflects mixing between N incorporated at the time of planetary accretion – small amounts of primordial N (δ15N " target="_blank" >40 ‰), possibly as low as proto-solar-nebula (PSN) nitrogen (δ15N = -383 ± 8 ‰), and nitrogen introduced into the mantle by long-term recycling of atmospheric (= 0 ‰) and/or sediment-derived (= ~+5 to +7 ‰) components (Marty, 2012).
View in article
Figure 2 Extrapolated Ne ((21Ne/22Ne)EX, i.e. air-corrected 21Ne/22Ne values) versus 15N/14N values of CIR basalts, plotted together with binary mixing curves between a pre-solar nitrogen (PSN) component (15N/14N = 0.00227, δ15N = -373 ‰; Marty et al., 2012) and two mantle endmember components reflecting addition of 15N-enriched sedimentary MORB mantle (SMM; δ15N = +5 ‰) to DMM mantle (δ15N = -5 ‰).
View in article
Moreover, mass balance arguments suggest Ne is efficiently recycled back to the surface during the subduction process (Holland and Ballentine, 2006; Marty, 2012).
View in article
It makes the following assumptions: (1) the N regassing flux commences at the onset of subduction and is constant through time; (2) a constant proportion of the total N regassing flux (F) is subducted into each respective mantle reservoir (defined here as FDMM and FPLM) and instantaneously mixes in each reservoir; (3) there is a constant mantle degassing flux of 5 × 109 mol/yr (Cartigny and Marty, 2013), derived from the DMM:PLM reservoirs in the proportion 85:15, respectively (Ito et al., 2003); (4) the present-day mantle N-content ([N]) is ~0.27 ± 0.16 ppm (Marty, 2012; Cartigny and Marty, 2013); and (5) the initial mantle δ15N is -40 ‰ (PSN-like; Cartigny and Marty, 2013) which is unmodified by any degassing prior to the onset of subduction.
View in article


Marty, B., Dauphas, N. (2003) The nitrogen record of crust-mantle interaction and mantle convection from Archean to present. Earth and Planetary Science Letters 206, 397– 410.
Show in context

There are three processes capable of producing high (>DMM) δ15N signatures in CIR basalts: (1) assimilation of existing crust during magma eruption, likely also involving incorporation of air (δ15N = 0 ‰), (2) mass-dependent fractionation related to magmatic degassing (Cartigny et al., 2001), and/or (3) recycling of oceanic sediments and/or oceanic crust (δ15N ~+5 to +7 ‰) into the CIR mantle source region producing the melts (Marty and Dauphas, 2003).
View in article
The CIR data reaffirms the considerable heterogeneity in mantle δ15N signatures (e.g., Marty and Dauphas, 2003).
View in article


Mikhail, S., Sverjensky, D.A. (2014) Nitrogen speciation in upper mantle fluids and the origin of Earth's nitrogen-rich atmosphere. Nature Geoscience 7, 816-819.
Show in context

Preferential deep recycling of N relative to Ne is consistent with recent theoretical and experimental predictions that indicate that N is stabilised as ammonium under subduction redox conditions (fO2 < QFM; Mikhail and Sverjensky, 2014) and therefore behaves as a large ion lithophile element whereby it can substitute into K-bearing silicate minerals and gain stability in the downgoing slab (Li et al., 2013).
View in article


Mitchell, E.C., Fischer, T.P., Hilton, D.R., Hauri, E.H., Shaw, A.M., de Moor, J.M., Sharp, Z.D., Kazahaya, K. (2010) Nitrogen sources and recycling at subduction zones: Insights from the Izu‐Bonin‐Mariana arc. Geochemistry, Geophysics, Geosystems 11, 2.
Show in context

Although some subducted N re-emerges via arc-related volcanism (Sano et al., 1998; Fischer et al., 2002) the majority likely bypasses sub-arc depths (150-200 km) and supplies the deeper mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in article


Morgan, W.J. (1978) Rodriguez, Darwin, Amsterdam, a second type of hot spot island. Journal of Geophysical Research 83, 5355–5360.
Show in context

These west-to-east trending ridges were formed by volcanism above a channel of Réunion hotspot mantle as the CIR migrated northeast over and away from the plume (Morgan, 1978).
View in article


O'Nions, R.K., Tolstikhin, I.N. (1996) Limits on the mass flux between lower and upper mantle and stability of layering. Earth and Planetary Science Letters 139, 213-222.
Show in context

Our model is based on a layered mantle reservoir concept, similar to that described for noble gases by O’Nions and Tolstikhin (1996) and Gonnermann and Mukhopadhyay (2009).
View in article


Parai, R., Mukhopadhyay, S. (2012) How large is the subducted water flux? New constraints on mantle regassing rates. Earth and Planetary Science Letters 317, 396-406.
Show in context

Finally, we highlight remarkably similar recycling efficiencies between N (84-92 %; this work) and heavy noble gases (73-87 %; Holland and Ballentine, 2006; Parai and Mukhopadhyay, 2012) despite fundamentally different subduction mechanisms: N is fixed as NH4+ and bound in recycled sediments and/or oceanic crust, whereas Kr and Xe undergo no such chemical transformation and instead are simply dissolved in pore fluids and/or in unoccupied amphibole A-sites within subducted oceanic crust (e.g., Holland and Ballentine, 2006; Jackson et al., 2015).
View in article


Sano, Y., Takahata, N., Nishio, Y., Marty, B. (1998) Nitrogen recycling in subduction zones. Geophysical Research Letters 25, 2289-2292.
Show in context

Although some subducted N re-emerges via arc-related volcanism (Sano et al., 1998; Fischer et al., 2002) the majority likely bypasses sub-arc depths (150-200 km) and supplies the deeper mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in article


Sarda, P., Staudacher, T., Allègre, C.J. (1988) Neon isotopes in submarine basalts. Earth and Planetary Science Letters 91, 73–88.
Show in context

For Ne-isotopes, a DMM 21Ne/22Ne endmember value (0.0594) is derived by extrapolating the MORB trajectory (Sarda et al., 1988) to the 20Ne/22Ne value of Ne-B (= 12.5) and noting the 21Ne/22Ne value at that point.
View in article


Trieloff, M., Kunz, J. (2005) Isotope systematics of noble gases in the Earth's mantle: possible sources of primordial isotopes and implications for mantle structure. Physics of the Earth and Planetary Interiors 148, 13-38.
Show in context

Solar Ne is defined by the solar wind value of 0.03118 (Trieloff and Kunz, 2005).
View in article


top

Supplementary Information


Additional Information and Methods


Geochemical Tracers of Volatile Recycling into Earth’s Mantle

Noble gases are powerful tracers of mantle processes, and sensitive indicators of crustal/sediment inputs to the mantle due to the large isotopic contrast of noble gases between mantle and crustal/atmospheric reservoirs. Light noble gases (helium and neon) in Earth's mantle preserve distinct solar-like isotopic compositions, whereas heavy noble gases (argon, krypton and xenon) display values which are intermediate between solar-like and air-dominated isotopic compositions (Holland and Ballentine, 2006

Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.

). Surficial helium is not recycled into the mantle during subduction as it is not gravitationally bound to Earth. The fate of Ne is debated (Sarda, 2004

Sarda, P. (2004) Surface noble gas recycling to the terrestrial mantle. Earth and Planetary Science Letters 228, 49-63.

): however, due to its relatively low abundance in subducted material, subducted contributions are generally considered negligible (Holland and Ballentine, 2006

Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.

). Heavier noble gas compositions in the mantle were long considered to be independent of subduction as they were believed to be returned back into the atmosphere through subduction volcanism before they could be admixed into the mantle (Staudacher and Allègre, 1988

Staudacher, T., Allègre, C.J. (1988) Recycling of oceanic crust and sediments: the noble gas subduction barrier. Earth and Planetary Science Letters 89, 173-183.

; Hilton et al., 2002

Hilton, D.R., Fischer, T.P., Marty, B. (2002) Noble gases and volatile recycling at subduction zones. In: Porcelli, D., Ballentine, C.J., Wieler, R. (Eds.) Noble Gases in Cosmochemistry and Geochemistry. Mineralogical Society of America 47 319-370.

). This led to the hypothesis that Earth’s mantle is insulated from atmospheric heavy noble gas inputs due to an efficient ‘subduction barrier’ for volatiles. However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993

Nagao, K., Takahashi, E. (1993) Noble gases in the mantle wedge and lower crust: an inference from the isotopic analyses of xenoliths from Oki-Dogo and Ichinomegata, Japan. Geochemical Journal 27, 229–240.

; Sarda et al., 1988

Sarda, P., Staudacher, T., Allègre, C.J. (1988) Neon isotopes in submarine basalts. Earth and Planetary Science Letters 91, 73-88.

; Matsumoto et al., 2001

Matsumoto, T., Chen, Y., Matsuda, J.I. (2001) Concomitant occurrence of primordial and recycled noble gases in the Earth’s mantle. Earth and Planetary Science Letters 185, 35–47.

; Sarda, 2004

Sarda, P. (2004) Surface noble gas recycling to the terrestrial mantle. Earth and Planetary Science Letters 228, 49-63.

; Holland and Ballentine, 2006

Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.

; Sumino et al., 2010

Sumino, H., Burgess, R., Mizukami, T., Wallis, S.R., Holland, G., Ballentine, C.J. (2010) Seawater-derived noble gases and halogens preserved in exhumed mantle wedge peridotite. Earth and Planetary Science Letters 294, 163-172.

; Hopp and Ionov, 2011

Hopp, J., Ionov, D.A. (2011) Tracing partial melting and subduction-related metasomatism in the Kamchatkan mantle wedge using noble gas compositions. Earth and Planetary Science Letters 302, 121–131.

; Kendrick et al., 2011

Kendrick, M.A., Scambelluri, M., Honda, M., Phillips, D. (2011). High abundances of noble gas and chlorine delivered to the mantle by serpentinite subduction. Nature Geoscience 4, 807-812.

; Mukhopadhyay, 2012

Mukhopadhyay, S. (2012) Early differentiation and volatile accretion recorded in deep-mantle neon and xenon. Nature 486, 101–104.

; Parai et al., 2012

Parai, R., Mukhopadhyay, S., Standish, J. (2012) Heterogeneous upper mantle Ne, Ar and Xe isotopic compositions and a possible Dupal noble gas signature recorded in basalts from the Southwest Indian Ridge. Earth and Planetary Science Letters 359, 227–239.

; Tucker et al., 2012

Tucker, J.M., Mukhopadhyay, S., Schilling, J.G. (2012) The heavy noble gas composition of the depleted MORB mantle (DMM) and its implications for the preservation of heterogeneities in the mantle. Earth and Planetary Science Letters 355, 244-254.

; Kendrick et al., 2013

Kendrick, M.A., Honda, M., Pettke, T., Scambelluri, M., Phillips, D., Giuliani, A. (2013) Subduction zone fluxes of halogens and noble gases in seafloor and forearc serpentinites. Earth and Planetary Science Letters 365, 86–96.

; Jackson et al., 2013a

Jackson, C.R.M., Parman, S., Kelley, S.P., Cooper, R.F. (2013a) Noble gas transport into the mantle facilitated by high solubility in amphibole. Nature Geoscience 6, 562–565.

;b

Jackson, C.R.M., Parman, S.W., Kelley, S.P., Cooper, R.F. (2013b) Constraints on light noble gas partitioning at the conditions of spinel-peridotite melting. Earth and Planetary Science Letters 384, 178–187.

; Pető et al., 2013

Pető, M.K., Mukhopadhyay, S., Kelley, K.A. (2013). Heterogeneities from the first 100 million years recorded in deep mantle noble gases from the Northern Lau Back-arc Basin. Earth and Planetary Science Letters 369, 13-23.

; Jackson et al., 2015

Jackson, C.R.M., Parman, S.W., Kelley, S.P., Cooper, R.F. (2015) Light noble gas dissolution into ring structure-bearing materials and lattice influences on noble gas recycling. . Geochimica et Cosmochimica Acta 159, 1–15.

; Parai and Mukhopadhyay, 2015

Parai, R., Mukhopadhyay, S. (2015) The evolution of MORB and plume mantle volatile budgets: constraints from fission Xe isotopes in Southwest Indian Ridge basalts. Geochemistry Geophysics Geosystems 16, 719–735.

; Jackson et al., 2016

Jackson, C.R.M., Shuster, D.L., Parman, S.W., Smye, A.J. (2016). Noble gas diffusivity hindered by low energy sites in amphibole. Geochimica et Cosmochimica Acta 172, 65-75.

), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005

Honda, M., Woodhead, J.D. (2005). A primordial solar-neon enriched component in the source of EM-I-type ocean island basalts from the Pitcairn Seamounts, Polynesia. Earth and Planetary Science Letters 236, 597-612.

; Sumino et al., 2010

Sumino, H., Burgess, R., Mizukami, T., Wallis, S.R., Holland, G., Ballentine, C.J. (2010) Seawater-derived noble gases and halogens preserved in exhumed mantle wedge peridotite. Earth and Planetary Science Letters 294, 163-172.

).

The large isotopic contrast between terrestrial nitrogen reservoirs makes N a potentially powerful tracer of volatile recycling between the surface and mantle (Li et al., 2007

Li, L., Bebout, G.E., Idleman, B.D. (2007) Nitrogen concentration and δ 15 N of altered oceanic crust obtained on ODP Legs 129 and 185: insights into alteration-related nitrogen enrichment and the nitrogen subduction budget. Geochimica et Cosmochimica Acta 71, 2344-2360.

; Mitchell et al., 2010

Mitchell, E.C., Fischer, T.P., Hilton, D.R., Hauri, E.H., Shaw, A.M., de Moor, J.M., Sharp, Z.D., Kazahaya, K. (2010) Nitrogen sources and recycling at subduction zones: Insights from the Izu‐Bonin‐Mariana arc. Geochemistry, Geophysics, Geosystems 11, 2.

; Johnson and Goldblatt, 2015

Johnson, B., Goldblatt, C. (2015) The nitrogen budget of Earth. Earth-Science Reviews 148, 150-173.

; Bebout et al., 2016

Bebout, G.E., Lazzeri, K.E., Geiger, C.A. (2016) Pathways for nitrogen cycling in Earth's crust and upper mantle: A review and new results for microporous beryl and cordierite. American Mineralogist 101, 7-24.

). For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986

Javoy, M., Pineau, F., Delorme, H. (1986) Carbon and nitrogen isotopes in the mantle. Chemical Geology 57, 41–62.

; Cartigny et al., 1998

Cartigny, P., Harris, J.W., Phillips, D., Girard, M., Javoy M. (1998) Subduction-related diamonds? The evidence for a mantle-derived origin from coupled δ13C - δ15N determinations. Chemical Geology 147, 147-159.

; Marty and Zimmermann, 1999

Marty, B., Zimmermann, L. (1999) Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: assessment of shallow-level fractionation and characterization of source composition. Geochimica et Cosmochimica Acta 63, 3619–3633.

; Marty and Dauphas, 2003

Marty, B., Dauphas, N. (2003) The nitrogen record of crust-mantle interaction and mantle convection from Archean to present. Earth and Planetary Science Letters 206, 397– 410.

; Cartigny and Marty, 2013

Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.

; Palot et al., 2012

Palot, M., Cartigny, P., Harris, J.W., Kaminsky, F.V., Stachel, T. (2012) Evidence for deep mantle convection and primordial heterogeneity from nitrogen and carbon stable isotopes in diamond. Earth and Planetary Science Letters 357, 179-193.

; Johnson and Goldblatt, 2015

Johnson, B., Goldblatt, C. (2015) The nitrogen budget of Earth. Earth-Science Reviews 148, 150-173.

; this study). This isotopic contrast was likely established early in Earth’s history, reflecting the integrated effects of mantle degassing, late heavy bombardment, and/or hydrodynamic escape of Earth’s primary atmosphere (Marty, 2012

Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.

). Nitrogen is stable as ammonium under subduction redox conditions (fO2 < QFM; Watenphul et al., 2010

Watenphul, A., Wunder, B., Wirth, R., Heinrich, W. (2010) Ammonium-bearing clinopyroxene: a potential nitrogen reservoir in the Earth's mantle. Chemical Geology 270, 240-248.

; Li et al., 2013

Li, Y., Wiedenbeck, M., Shcheka, S., Keppler, H. (2013). Nitrogen solubility in upper mantle minerals. Earth and Planetary Science Letters 377, 311-323.

; Li and Keppler, 2014

Li, Y., Keppler, H. (2014) Nitrogen speciation in mantle and crustal fluids. Geochimica et Cosmochimica Acta 129, 13-32.

; Mikhail and Sverjensky, 2014

Mikhail, S., Sverjensky, D.A. (2014) Nitrogen speciation in upper mantle fluids and the origin of Earth's nitrogen-rich atmosphere. Nature Geoscience 7, 816-819.

) and acts much like a large ion lithophile element, substituting for K-bearing minerals and sedimentary material. Subsequent modifications to mantle N may result from subduction of this material comprised of isotopically distinct N (~+5‰) (Sano et al., 1998

Sano, Y., Takahata, N., Nishio, Y., Marty, B. (1998) Nitrogen recycling in subduction zones. Geophysical Research Letters 25, 2289-2292.

; Busigny et al., 2013

Busigny, V., Lebeau, O., Ader, M., Krapez, B., Bekker, A. (2013) Nitrogen cycle in a Late Archean ferruginous ocean. Chemical Geology 362, 115-130.

; Thomazo and Papineau, 2013

Thomazo, C., Papineau, D. (2013) Biogeochemical cycling of nitrogen on the early Earth. Elements 9, 345-351.

; Halama et al., 2014

Halama, R., Bebout, G.E., John, T., Scambelluri, M. (2014). Nitrogen recycling in subducted mantle rocks and implications for the global nitrogen cycle. International Journal of Earth Sciences 103, 2081-2099.

). For example, modern oceanic sediments are enriched in the heavy isotope of nitrogen (i.e. 15N) relative to the DMM. As a result, subduction of altered oceanic crust and sediments, containing molecularly-bound nitrogen – residual to sub-arc fluid loss – could result in the introduction of high N-isotope signatures into the (deep) mantle (Hofmann and White, 1982

Hofmann, A.W., White, W.M. (1982) Mantle plumes from ancient oceanic crust. Earth and Planetary Science Letters 57, 421-436.

; Marty and Humbert, 1997

Marty, B., Humbert, F. (1997) Nitrogen and argon isotopes in oceanic basalts. Earth and Planetary Science Letters 152, 101-112.

; Halama et al., 2010

Halama, R., Bebout, G.E., John, T., Schenk, V. (2010). Nitrogen recycling in subducted oceanic lithosphere: The record in high-and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636-1652.

). N-isotope enrichments have been observed in subduction-related geothermal gases (Fischer et al., 2002

Fischer, T.P., Hilton, D.R., Zimmer, M.M., Shaw, A.M., Sharp, Z.D., Walker, J.A. (2002) Subduction and recycling of nitrogen along the Central American margin. Science 297, 1154-1157.

) and ocean island basalts (OIBs; Hoffman and White, 1982

Hofmann, A.W., White, W.M. (1982) Mantle plumes from ancient oceanic crust. Earth and Planetary Science Letters 57, 421-436.

; Mohapatra and Murty, 2002

Mohapatra, R.K., Murty, S.V.S. (2002) Nitrogen and noble gas isotopes in mafic and ultramafic inclusions in the alkali basalts from Kutch and Reunion—implications for their mantle sources. Journal of Asian Earth Sciences 20, 867-877.

; Marty and Dauphas, 2003

Marty, B., Dauphas, N. (2003) The nitrogen record of crust-mantle interaction and mantle convection from Archean to present. Earth and Planetary Science Letters 206, 397– 410.

; Fischer et al., 2005

Fischer, T.P., Takahata, N., Sano, Y., Sumino, H., Hilton, D.R. (2005) Nitrogen isotopes of the mantle: Insights from mineral separates. Geophysical Research Letters 32, L11305.

). To date, however, such N-isotope anomalies have rarely been detected in the DMM (Marty and Zimmermann, 1999

Marty, B., Zimmermann, L. (1999) Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: assessment of shallow-level fractionation and characterization of source composition. Geochimica et Cosmochimica Acta 63, 3619–3633.

; Cartigny et al., 2001

Cartigny, P., Jendrzejewski, N., Pineau, F., Petit, E., Javoy, M. (2001) Volatile (C, N, Ar) variability in MORB and the respective roles of mantle source heterogeneity and degassing: the case of the Southwest Indian Ridge. Earth and Planetary Science Letters 194, 241–257.

).

Geological Setting

The CIR separates the African and Indo-Australian plates and extends over ~3000 km from the Carlsberg Ridge (3.0 °N, 66.0 °E) southwards towards the Rodrigues Triple Junction (20.5 °S, 70.0 °E), where it bifurcates into the Southeast and the Southwest Indian ridges. Basalt samples of this study were collected along a segment of the CIR between 16.7 °S and 20.6 °S, located to the north and south of the Marie Celeste Fracture Zone (FZ) and northwest of the Egeria FZ, respectively (Fig. S-1). The CIR axis shoals from north to south along this segment. In addition, several off-axis samples were collected along a perpendicularly offset lineament, which extends westward from 66 °E to 64 °E toward the Mascarene Islands (e.g., Réunion, Mauritius, Rodrigues islands; Parson et al., 1993

Parson, L.M., Patriat, P., Searle, R.C., Briais, A.R. (1993). Segmentation of the Central Indian Ridge between 12 12′ S and the Indian Ocean triple junction. Marine geophysical researches 15, 265-282.

).

During the past ~70 Ma, the Réunion hot spot formed the Deccan Trap flood basalts, the Chagos-Maldive-Laccadive Ridge, the Mascarene Plateau and the Mascarene Islands (Duncan et al., 1989

Duncan, R.A., Backman, J., Peterson, L., Shipboard Scientific Party (1989) Réunion hotspot activitity through tertiary time: Initial results from the ocean drilling program, leg 115. Journal of Volcanology and Geothermal Research 36, 193-198.

, 1990

Duncan, R.A., Backman, J., Peterson, L. (1990) The volcanic record of the Réunion hot spot. Proceedings of the Ocean Drilling Program Scientific Results 115, 3–10.

). At ~34 Ma, the northeastward migrating CIR moved over the Réunion hot spot (Duncan et al., 1990

Duncan, R.A., Backman, J., Peterson, L. (1990) The volcanic record of the Réunion hot spot. Proceedings of the Ocean Drilling Program Scientific Results 115, 3–10.

) and embedded a “fossil” Réunion hot spot mantle component into the sub-ridge mantle. The Réunion hot spot is currently located ~1100 km west of the CIR axis (Figure S-1; Mahoney et al., 1989

Mahoney, J.J., Natland, J.H., White, W.M., Poreda, R., Bloomer, S.H., Baxter, A.N. (1989) Isotopic and geochemical provinces of the Western Indian Ocean spreading centers. Journal of Geophysical Research 94, 4033–4052.

; Murton et al., 2005

Murton, B.J., Tindle, A.G., Milton, J.A., Sauter, D. (2005) Heterogeneity in southern Central Indian Ridge MORB: implications for ridge– hot spot interaction. Geochemistry Geophysics Geosystems 6, Q03E20.

). Using He isotopes ratios (3He/4He) – the canonical tracer of mantle plume involvement in petrogenesis (Courtillot et al., 2003

Courtillot, V., Davaille, A., Besse, J., Stock, J. (2003) Three distinct types of hotspots in the Earth’s mantle. Earth and Planetary Science Letters 205, 295-308.

) – Füri et al. (2011)

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J. M. D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

identified plume material flowing eastward toward the CIR on a trajectory that impinges the ridge at ~19.9 °S. Réunion Island is composed of two volcanoes: Piton des Neiges, which was active between ~2 Ma and 43 ka, and the currently active Piton de la Fournaise, where volcanic activity began ~530 ka ago (Gillot and Nativel, 1989

Gillot, P.Y., Nativel, P. (1989) Eruptive history of the Piton de la Fournaise volcano, Réunion, Indian Ocean. Journal of Volcanology and Geothermal Research 36, 53–65.

). The island of Mauritius is located 250 km east of Réunion Island and 850 km west of the CIR, at the previous site of the Réunion hot spot (~8 Ma) (Morgan, 1981

Morgan, W.J. (1981) Hot spot tracks and the opening of the Atlantic and Indian oceans. In The Sea 7, 443–487.

; Paul et al., 2005

Paul, D., White, W.M., Blichert-Toft, J. (2005) Geochemistry of Mauritius and the origin of rejuvenescent volcanism on oceanic island volcanoes. Geochemistry Geophysics Geosystems 6, Q06007.

). Rodrigues Island is located ~600 km east of Mauritius and formed at 1.5 Ma (McDougall et al., 1965

McDougall, I., Upton, B.G.J., Wadsworth, W.J. (1965) A geological reconnaissance of Rodriguez Island Indian Ocean. Nature 206, 26–27.

); it marks the eastern extent of the Rodrigues Ridge, an east-west trending volcanic ridge that is between 8 to 10 Ma old (Duncan et al., 1990

Duncan, R.A., Backman, J., Peterson, L. (1990) The volcanic record of the Réunion hot spot. Proceedings of the Ocean Drilling Program Scientific Results 115, 3–10.

). Smaller en-échelon volcanic ridges at 19 °S (e.g., the Three Magi Ridges and the Gasitao Ridge) extend the Rodrigues Ridge close to the CIR (Fig. S-1) (Dyment et al., 1999

Dyment, J., Gallet, Y., Magofond 2 Scientific Party (1999) The Magofond 2 cruise: A surface and deep tow survey on the past and present Central Indian Ridge. InterRidge News 8, 25–31.

, 2000

Dyment, J., Hémond, C., Gimnaut Scientific Party (2000) Deep‐sea exploration of the Central Indian Ridge at 19°S. InterRidge News 9, 29–32.

). Various models have been proposed to explain the occurrence of volcanism between the Mascarene Islands and the CIR. First, Morgan (1978)

Morgan, W.J. (1978) Rodriguez, Darwin, Amsterdam, a second type of hot spot island. Journal of Geophysical Research 83, 5355–5360.

proposed that the Rodrigues Ridge was formed by volcanism above a channel of upwelling Réunion hot spot mantle, which was deflected toward the CIR as it migrated northeastward away from the hot spot, and predicted a Réunion signature would be present at the intersection of the CIR with a line projected through Réunion and Rodrigues islands. Mahoney et al. (1989)

Mahoney, J.J., Natland, J.H., White, W.M., Poreda, R., Bloomer, S.H., Baxter, A.N. (1989) Isotopic and geochemical provinces of the Western Indian Ocean spreading centers. Journal of Geophysical Research 94, 4033–4052.

later detected this Réunion-like isotope signature (low 143Nd/144Nd, high 87Sr/86Sr, high 207Pb/204Pb, and high 206Pb/204Pb) in basalts from the Marie Celeste FZ portion of the CIR.

Several detailed geochemical studies have focused on plume-ridge interaction in the region between the Rodrigues Triple Junction and the Marie Celeste fracture zone. Murton et al. (2005)

Murton, B.J., Tindle, A.G., Milton, J.A., Sauter, D. (2005) Heterogeneity in southern Central Indian Ridge MORB: implications for ridge– hot spot interaction. Geochemistry Geophysics Geosystems 6, Q03E20.

conducted a detailed study of this area and reported enrichments in incompatible elements that increased northward, which they interpreted to reflect the presence of enriched mantle originating at the Réunion hot spot, which had subsequently migrated eastward toward the CIR against the direction of motion of the lithosphere. Nauret et al. (2006)

Nauret, F., Abouchami, W., Galer, S.J.G., Hofmann, A.W., Hémond, C., Chauvel, C., Dyment, J. (2006) Correlated trace element - Pb isotope enrichments in Indian MORB along 18–20°S, Central Indian Ridge. Earth and Planetary Science Letters 245, 137–152.

also targeted a suite of basalts collected both on and off the CIR axis between 18 °S and 20 °S for Sr-Nd-Pb isotopes and showed that the most trace element enriched samples display the most radiogenic Pb isotopic compositions. Furthermore, they compared basalt samples with submarine Réunion lavas (Fretzdorff and Haase, 2002

Fretzdorff, S., Haase, K.M. (2002) Geochemistry and petrology of lavas from the submarine flanks of Réunion Island (western Indian Ocean): implications for magma genesis and the mantle source. Mineralogy and Petrology 75, 153-184.

) and concluded that the radiogenic isotope characteristics of the majority of on-axis samples cannot be explained with a Réunion-like endmember. Recently, however, Ulrich et al. (2012)

Ulrich, M., Hemond, C., Nonnotte, P., Jochum, K.P. (2012) OIB/seamount recycling as a possible process for E-MORB genesis. Geochemistry Geophysics Geosystems 13, Q0AC19.

showed that trace element and isotopic enrichments in the 18 °S to 20 °S region of the CIR are consistent with a binary mixture between the regional depleted MORB mantle (DMM) source and a recycled Ocean Iceland Basalt (OIB)/plume component. In contrast, off-axis basalts from the Gasitao Ridge, as well as a single on-ridge sample collected at ~19.9 °S, appeared to record a Réunion source signature. As a result, these authors proposed that Réunion hot spot material flows eastward toward the CIR on a trajectory that impinges the ridge at ~19.9 °S, in agreement with the initial hypothesis put forth by Morgan (1978)

Morgan, W.J. (1978) Rodriguez, Darwin, Amsterdam, a second type of hot spot island. Journal of Geophysical Research 83, 5355–5360.

. Furthermore, they concluded that mantle source enrichment in the vicinity of the Marie Celeste FZ, ~100 km to the north, cannot be related to influx of mantle material from Réunion. Füri et al. (2011)

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J. M. D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

, however, suggested that these enrichments, along with slightly radiogenic He-isotope values could result from a “fossil” Réunion plume component in the region. In addition, these authors showed that the highest 3He/4He values (~12.2 RA) were measured in glasses from off-axis portions of the CIR, and are consistent with flow of hot spot mantle material from Réunion (~100 km to the west) toward the CIR.


Figure S-1 Bathymetric map of the (on-axis) CIR and the adjacent (off-axis) Gasitao Ridge, Three Magi Ridges, Rodrigues Ridge and Abyssal Hill. Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997

Smith, W.H.F., Sandwell, D.T. (1997) Global sea floor topography from satellite altimetry and ship depth soundings. Science 277, 1956-1962.

); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999

Dyment, J., Gallet, Y., Magofond 2 Scientific Party (1999) The Magofond 2 cruise: A surface and deep tow survey on the past and present Central Indian Ridge. InterRidge News 8, 25–31.

), R/V L'Atalante (Dyment et al., 2000

Dyment, J., Hémond, C., Gimnaut Scientific Party (2000) Deep‐sea exploration of the Central Indian Ridge at 19°S. InterRidge News 9, 29–32.

), and R/V Hakuho-Maru (Okino et al., 2008

Okino, K., Ichikawa, Y., Tamaki, T. (2008) Detailed morphology of the Central Indian Ridge between 20°15S and 15°30S: Implication for hot spot ridge interaction. Proceedings of the Japan Geoscience Union (JPGU), Chiba, Japan, Abstract J164 002.

); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008

Füri, E., Hilton, D.R., KNOX11RR Scientific Party (2008) Sampling and surveying ridge‐hot spot interaction on the Central Indian Ridge, 19°S: Cruise KNOX11RR. InterRidge News 17, 28–29.

, 2011

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J. M. D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

). The circles indicate the location of samples collected during the KNOX11RR (black = CIR on-axis; red = off-axis) and GIMNAUT (blue) cruises. The inset shows the Mascarene Islands (i.e. Réunion, Mauritius, and Rodrigues islands) to the west and the Rodrigues Triple Junction.
Full size image | Download in Powerpoint
Sample Collection

Twenty-four submarine pillow basalts were dredged from the ocean floor along the CIR axis (16.7 °S to 20.6 °S), and the adjacent Gasitao, Three Magi, and Rodrigues ridges on our Knox11RR expedition (Füri et al., 2008

Füri, E., Hilton, D.R., KNOX11RR Scientific Party (2008) Sampling and surveying ridge‐hot spot interaction on the Central Indian Ridge, 19°S: Cruise KNOX11RR. InterRidge News 17, 28–29.

, 2011

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J. M. D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

; Fig. S-1). In addition, several off-axis samples were collected along a perpendicularly offset lineament which extends westward from 66 °E to 64 °E toward the Mascarene Islands (e.g., Réunion, Mauritius, Rodrigues islands; Parson et al., 1993

Parson, L.M., Patriat, P., Searle, R.C., Briais, A.R. (1993). Segmentation of the Central Indian Ridge between 12 12′ S and the Indian Ocean triple junction. Marine geophysical researches 15, 265-282.

). Following the cruise, cumulate dunite xenoliths (n = 4) were collected from Piton Chisny on Réunion Island during a land-based field campaign.

Sample Preparation

Selected basalt samples were first ultrasonically cleaned in dichloromethane to remove any organic contaminants from the glass surface. Subsequently, approximately 300 mg of dried fresh glass, free of surficial alteration, phenocrysts, or large vesicles were handpicked using a binocular microscope. The selected glass was then ultrasonically cleaned in a 1:1 acetone-methanol mixture, dried, and transferred to the crushing apparatus.

Isotope Analyses

Neon and argon gases were released from basalts using a custom-built piston-activated in vacuo rock-crusher (see Hahm et al. (2012)

Hahm, D., Hilton, D.R., Castillo, P.R., Hawkins, J.W., Hanan, B.B., Hauri, E.H. (2012) An overview of the volatile systematics of the Lau Basin – Resolving the effects of source variation, magmatic degassing and crustal contamination. Geochimica et Cosmochimica Acta 85, 88-113.

for details). The released gas was purified using a dedicated cryogenic separation line whereby Ne and Ar were extracted and analysed individually. Neon and argon abundances and isotope ratios (as well as He abundances) were measured using a modified VG5440 mass spectrometer equipped with five Faraday cups and Daly photo-multiplier detector (Craig et al., 1993

Craig, H., Marti, K., Wiens, R. (1993) A Static Mass Spectrometer With Triple Collection for Nitrogen and Neon Isotopes. SIO Reference Series. Scripps Institute of Oceanography, La Jolla, California, USA, 93-11, 1–20A.

; Hahm et al., 2012

Hahm, D., Hilton, D.R., Castillo, P.R., Hawkins, J.W., Hanan, B.B., Hauri, E.H. (2012) An overview of the volatile systematics of the Lau Basin – Resolving the effects of source variation, magmatic degassing and crustal contamination. Geochimica et Cosmochimica Acta 85, 88-113.

), operated in peak jumping mode.

Neon results were corrected for procedural blanks and inferences of doubly-charged 40Ar and CO2 with 20Ne and 22Ne, respectively (see Niedermann et al. (1993)

Niedermann, S., Graf, T., Marti, K. (1993) Mass spectrometric identification of cosmic-ray-produced neon in terrestrial rocks with multiple neon components. Earth and Planetary Science Letters 118, 65–73.

for details). 20Ne procedural crusher blanks were (15 ± 8) × 10−12 cm3STP and 40Ar procedural blanks were (6 ± 2) × 10−9 cm3STP, representing less than 10 % of sample yields.

Nitrogen was released using the same crushing mechanism described above for Ne-Ar; however, gases were purified on a separate vacuum clean-up line (see Barry et al. (2012)

Barry, P.H., Hilton, D.R., Halldórsson, S.A., Hahm, D., Marti, K. (2012) High precision nitrogen isotope measurements in oceanic basalts using a static triple collection noble gas mass spectrometer. Geochemistry Geophysics Geosystems 13, Q01019.

for details) prior to inlet into the VG5440 mass spectrometer, which was operated in static triple collection mode. Interfering CO species were monitored and CO corrections were applied to all samples. Procedural N2 crusher blanks were ~4.2 ± 0.5 × 10−6 cm3STP on average, and were run prior to each sample analysis. Subsequent blank subtractions were applied, representing less than 20 % of the sample size. In addition, N2/Ar ratios were measured separately, on an aliquot of the same gas, using a quadrupole mass spectrometer (QMS).

Atmospheric-derived Contamination

Different noble gas species in the mantle are variably affected by atmospheric-contamination – depending, in part, on their relative abundances in the mantle versus Earth’s surface reservoirs. For example, Ne (18.2 ppm) and Ar (0.937 %) are enriched in the Earth’s atmosphere relative to basaltic abundances and therefore basalts are highly susceptible to air-contamination processes. Air-like Ne-isotope contributions are ubiquitous in natural samples (Graham, 2002

Graham, D.W. (2002) Noble gas isotope geochemistry of mid‐ocean ridge and ocean island basalts: Characterization of mantle source reservoirs. RIMS 47, 247–317.

) with infiltration of air into basalts occurring through small micro-fractures, which can develop in glasses during seafloor sampling and/or handling in the laboratory (Ballentine and Barfod, 2000

Ballentine, C.J., Barfod, D.N. (2000) The origin of air-like noble gases in MORB and OIB. Earth and Planetary Science Letters 180, 39–48.

), and/or by assimilation of seawater-derived components (Farley and Craig, 1994

Farley, K.A., Craig, H. (1994) Atmospheric argon contamination of ocean island basalt olivine phenocrysts. Geochimica et Cosmochimica Acta 58, 2509–2517.

).

Nitrogen is the most abundant (~78 %) gaseous species in the atmosphere; however, it occurs only in trace amounts in CIR basalts, and thus the potential for atmospheric contamination is substantial. Samples identified to be air-like with respect to Ne-Ar isotope systematics (Figs. S-2 and S-3) display non-air-like δ15N and N2/Ar values. In addition, the majority of CIR samples have N2/Ar ratios well above the air value, indicating that air contamination is minimal. In Figure S-4, we plot δ15N versus the air-normalised He/Ne values ((4He/20Ne)Sample/(4He/20Ne)Air) – a useful indicator of extent of air contamination. We show that both on-axis and off-axis CIR glasses have markedly (i.e. >2 orders of magnitude) higher 4He/20Ne ratios than air, and thus are not considered to be significantly modified by air-contributions.


Figure S-2 Neon three-isotope plot (20Ne/22Ne versus 21Ne/22Ne) of CIR basaltic glasses and Réunion xenoliths. Three trend-lines are superimposed on the data: (1) the air-solar mixing line; (2) the Réunion line (Hopp and Trieloff, 2005

Hopp, J., Trieloff, M. (2005) Refining the noble gas record of the Réunion mantle plume source: Implications on mantle geochemistry. Earth and Planetary Science Letters 240, 573–588.

); (3) the DMM (2∏D43) line (Moreira and Allègre, 1998

Moreira, M., Allègre, C.J. (1998) Helium-neon systematics and the structure of the mantle. Chemical Geology 147, 53–59.

). Uncertainties are at the 1σ level. Only samples with isotopes ratios >1 sigma from air are plotted.
Full size image | Download in Powerpoint


Figure S-3 Ar-isotopes versus Ne-isotopes. Ne-Ar systematics are coupled for the majority of samples, i.e. air-contamination and mantle isotope anomalies are evident in the same samples.
Full size image | Download in Powerpoint


Figure S-4 N-isotopes (δ15N) versus air-normalised He/Ne values (= 4He/20NeS/4He/20NeA). High He/Ne values relative to air (>100) suggest that all CIR basalt samples have undergone minimal air-contamination. Binary mixing trajectories are shown to postulated endmember values for ‘plume = seds’ (δ15N = +1.5‰; (4He/20Ne)Sample/(4He/20Ne)Air > 104) and ‘hybrid’ (δ15N = -2 ‰; (4He/20Ne)Sample/(4He/20Ne)Air > 104).
Full size image | Download in Powerpoint
Atmospheric Neon Correction

Although Ne-isotopes in basalts are highly susceptible to air-contamination, air-derived Ne contributions can be corrected by assuming that measured neon isotope values represent a binary mixture of atmospheric and mantle-derived neon. Following the methods of Honda et al. (1991)

Honda, M., McDougall, I., Patterson, D.B., Doulgeris, A., Clague, D.A. (1991) Possible solar noble-gas component in Hawaiian basalts. Nature 349, 149 – 151.

, we extrapolate measured 20Ne/22Ne values to “solar” (i.e. Ne-B = 12.5 ) values in order to estimate air-corrected (i.e. 21Ne/22NeEX) values. The following equations are used:

Eq. S-1 
21Ne/22NeEX = (21Ne/22Ne M - 21Ne/22NeA)/f22 + 21Ne/22NeA

where f22 is the proportion of mantle-derived Ne in a sample:

Eq. S-2 
f22 = (20Ne/22NeM - 20Ne/22NeA)/(20Ne/22NeS - 20Ne/22NeA)

and 20Ne/22NeA and 21Ne/22NeA are the isotopic ratios of air, assumed to be 9.8 and 0.029, respectively, whereas 21Ne/22NeM is the measured sample 21Ne/22Ne value and 20Ne/22NeS is the assumed 21Ne/22Ne “solar” (i.e. Ne-B) neon component (=12.5; Trieloff et al., 2000

Trieloff, M., Kunz, J., Clague, D.A., Harrison, D., Allègre, C.J. (2000) The nature of pristine noble gases in mantle plumes. Science 288, 1036–1038.

). Notably, the most air-like sample (e.g., D3-1; highest 20Ne content) cannot be corrected due to a 21Ne/22NeM value that is indistinguishable from air. The Ne–B endmember is obtained from gas-rich meteorites and lunar soils with a 20Ne/22Ne ratio of 12.52 ± 0.18, considered to represent implanted solar neon, now present within Earth’s mantle (Black, 1972

Black, D.C. (1972) On the origins of trapped helium, neon and argon isotopic variations in meteorites – I. Gas-rich meteorites, lunar soil and breccia. Geochimica et Cosmochimica Acta 36, 347–375.

; Trieloff et al., 2000

Trieloff, M., Kunz, J., Clague, D.A., Harrison, D., Allègre, C.J. (2000) The nature of pristine noble gases in mantle plumes. Science 288, 1036–1038.

, 2002

Trieloff, M., Kunz, J., Allègre C.J. (2002) Noble gas systematics of the Réunion mantle plume source and the origin of primordial noble gases in Earth’s mantle. Earth and Planetary Science Letters 200, 297–313.

; Ballentine et al., 2005

Ballentine, C.J., Marty, B., Lollar, B.S., Cassidy, M. (2005) Neon isotopes constrain convection and volatile origin in the Earth’s mantle. Nature 433, 33–38.

; Moreira, 2013

Moreira, M. (2013) Noble gas constraints on the origin and evolution of Earth’s volatiles. Geochemical Perspectives 2, 229-230.

). The primordial, “solar” neon component in Earth’s mantle has also been assumed to be best represented by present-day solar wind (i.e. 20Ne/22Ne = 13.8; Benkert et al., 1993

Benkert, J.P., Baur, H., Signer, P., Wieler, R. (1993) He, Ne, and Ar from the solar-wind and solar energetic particles in lunar ilmenites and pyroxenes. Journal of Geophysical Research – Planets 98, 13147–13162.

; Kallenbach et al., 1997

Kallenbach, R., Ipavich, F.M., Bochsler, P., Hefti, S., Hovestadt, D., Grünwaldt, H., Hilchenbach, M., Axford, W.I., Balsiger, H., Bürgi, A., Coplan, M.A., Galvin, A.B., Geiss, J., Gliem, F., Gloeckler, G., Hsieh, K.C., Klecker, B., Lee, M.A., Livi, S., Managadze, G.G., Marsch, E., Möbius, E., Neugebauer, M., Reiche, K.U., Scholer, M., Verigin, M.I., Wilken, B. Wurz, P. (1997) Isotopic composition of solar wind neon measured by CELIAS/MTOF on board SOHO. Journal of Geophysical Research – Space 102, 26895–26904.

); note, however, that selection of this endmember as opposed to Ne-B will not significantly affect the conclusions in the following discussion. Basalt CIR 21Ne/22NeEX values vary between 0.0437 ± 0.0009 and 0.0582 ± 0.0013 and fall in the range of values observed in previous studies of plume-rift related systems (e.g., Füri et al., 2010

Füri, E., Hilton, D.R., Halldórsson, S.A., Barry, P.H., Hahm, D., Fischer, T.P., Gronvold, K. (2010) Apparent decoupling of the He and Ne isotope systematics of the Icelandic mantle: the role of He depletion, melt mixing, degassing fractionation and air interaction. Geochimica et Cosmochimica Acta 74, 3307–3332.

; Hahm et al., 2012

Hahm, D., Hilton, D.R., Castillo, P.R., Hawkins, J.W., Hanan, B.B., Hauri, E.H. (2012) An overview of the volatile systematics of the Lau Basin – Resolving the effects of source variation, magmatic degassing and crustal contamination. Geochimica et Cosmochimica Acta 85, 88-113.

). On-axis CIR 21Ne/22NeEX values range from 0.0437 ± 0.0011 to 0.0560 ± 0.0023, whereas off-axis samples range from 0.0441 ± 0.0017 to 0.0582 ± 0.0013.

Degassing Fractionation

In addition to air contamination, another process capable of modifying intrinsic mantle volatile features is degassing. Having discussed the importance of identifying air-contaminated samples and applying a correction when necessary, we now focus on additional processes which can modify source features, such as degassing. Degassing can potentially cause both isotope and relative abundance fractionation. For example, relative noble gas abundances (e.g., 4He/40Ar*) can provide useful information about the mode and extent of volatile loss (Marty and Tolstikhin, 1998

Marty, B., Tolstikhin, I.N. (1998) CO2 fluxes from mid-ocean ridges, arc and plumes. Chemical Geology 145, 233–248.

). Using this approach, an air-correction is first applied to the argon data by assuming that all 36Ar is derived from the atmosphere. In this way the measured 40Ar content in each sample can be corrected for the presence of atmospheric Ar to radiogenic Ar (i.e. 40Ar*):

Eq. S-3 
40Ar* = [36ArM] × [(40Ar/36Ar)M - (40Ar/3636)A]

where 36ArM and (40Ar/36Ar)M are the measured 36Ar abundance and argon isotope ratio, respectively, and (40Ar/36Ar)A is the air ratio (= 298.56 ± 0.31; Lee et al., 2006

Lee, J.Y., Marti, K., Severinghaus, J.P., Kawamura, K., Yoo, H.S., Lee, J.B., Kim, J.S. (2006) A redetermination of the isotopic abundances of atmospheric Ar. Geochimica et Cosmochimica Acta 70, 4507-4512.

).

The measured He contents and calculated 40Ar*, from Equation S-3, are then combined to give 4He/40Ar* ratios, which can be used to determine the extent of degassing, as He is more soluble than Ar (SHe/SAr = 9.5; Jambon et al., 1986

Jambon, A., Weber, H.W., Braun, O. (1986) Solubility of He, Ne, Ar, Kr and Xe in a basalt melt in the range 1250–1600°C - Geochemical implications. Geochimica et Cosmochimica Acta 50, 401–408.

; Lux, 1987

Lux, G. (1987) The behavior of noble gases in silicate liquids: solution, diffusion, bubbles and surface effects, with applications to natural samples. Geochimica et Cosmochimica Acta 51, 1549–1560.

) in basaltic magmas (Marty and Tolstikhin, 1998

Marty, B., Tolstikhin, I.N. (1998) CO2 fluxes from mid-ocean ridges, arc and plumes. Chemical Geology 145, 233–248.

). Thus, residual (i.e. basalt) 4He/40Ar* values should increase with increased degassing. Using estimates of upper mantle K/U (Arevalo et al., 2009

Arevalo, R., McDonough, W.F., Luong, M. (2009) The K/U ratio of the silicate Earth: Insights into mantle composition, structure and thermal evolution. Earth and Planetary Science Letters 278, 361-369.

) and Th/U (O'Nions and McKenzie, 1993

O’Nions, R.K., McKenzie, D. (1993) Estimates of mantle thorium/uranium ratios from Th, U and Pb isotope abundances in basaltic melts. Philosophical Transactions of the Royal Society 342, 65–77.

), a production ratio for 4He/40Ar* of ~2 (Jambon et al., 1986

Jambon, A., Weber, H.W., Braun, O. (1986) Solubility of He, Ne, Ar, Kr and Xe in a basalt melt in the range 1250–1600°C - Geochemical implications. Geochimica et Cosmochimica Acta 50, 401–408.

; Marty and Zimmermann, 1999

Marty, B., Zimmermann, L. (1999) Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: assessment of shallow-level fractionation and characterization of source composition. Geochimica et Cosmochimica Acta 63, 3619–3633.

) is assumed as a mantle starting value. In the case of the present sample suite, 4He/40Ar* could not be calculated for two on-axis CIR basalt samples only (D1-1 and D3-1) due to measured air-like 40Ar/36Ar ratios (i.e. insignificant radiogenic 40Ar contributions). Notably, all 4He/40Ar* ratios calculated for CIR basalts are higher than the mantle production ratio of ~2 and range from 3.0 to 62, compared to Réunion xenoliths which fall between 1.3 and 2.0. In general, 4He/40Ar* are low (<20), suggesting moderate but variable amounts of degassing. Two groups are evident: on axis samples, which have 4He/40Ar* ratios <~10, with the exception of sample D2-1 (4He/40Ar* = 62) and off-axis samples, which have 4He/40Ar* ratios >~10 with the exception of D26-2 (4He/40Ar* = 5.2). Higher 4He/40Ar* ratios in off-axis samples (~5-10 x the production ratio) suggest more extensive degassing in off-axis portions of the crust, compared with on-axis samples. In contrast, 4He/40Ar* values close to the mantle production value are measured in Réunion xenoliths, indicating that gas loss was insignificant or very minor.

In order to test if the observed N-isotope variations result from degassing-related modification, we plot 4He/40Ar* versus δ15N (Fig. 1d) and note that two distinct (i.e. a positive and negative) fields are evident. Significantly, both fields have variable 4He/40Ar*, yet exhibit overlapping δ15N values. For example, positive δ15N samples have 4He/40Ar* values that vary by a factor of 10, but all δ15N values are indistinguishable within error. The same observation can be made for the negative δ15N field where 4He/40Ar* values range from 2 to 5: however, δ15N values overlap. These results indicate that N-isotopes variations are independent of 4He/40Ar* and thus we conclude that N-isotopes of CIR basalts have not experienced degassing-related fractionation. Significantly, this observation is consistent with previous compilation studies of various mid ocean ridge (MOR) and OIB settings (e.g., Marty and Humbert, 1997

Marty, B., Humbert, F. (1997) Nitrogen and argon isotopes in oceanic basalts. Earth and Planetary Science Letters 152, 101-112.

; Fischer et al., 2005

Fischer, T.P., Takahata, N., Sano, Y., Sumino, H., Hilton, D.R. (2005) Nitrogen isotopes of the mantle: Insights from mineral separates. Geophysical Research Letters 32, L11305.

), which also suggest that nitrogen isotope variations are independent of degassing. Therefore, nitrogen isotope variability in CIR basalts must be attributed to other processes – either mixing between isotopically distinct components in the mantle source and/or crustal contributions. In the following section, we investigate coupled He-Ne-Ar isotope systematics in order to assess if observed co-variations can be explained by mantle-mixing processes.

Solar Endmembers

Ne-isotope variations of CIR basalts suggest a solar component is present in the CIR mantle, with observed Ne-isotope variations best explained by mixing between DMM and primitive Réunion plume-like mantle endmembers (Fig. S-3). Assuming solar wind samples accurately approximate the proto-solar nebula (PSN; Marty et al., 2011

Marty, B., Chaussidon, M., Jurewicz, A.,Wiens, R., Burnett, D.S. (2011) A 15N-poor isotopic composition for the solar system as shown by Genesis solar wind samples. Science 332, 1533–1536.

; Marty, 2012

Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.

; Füri and Marty, 2015

Füri, E., Marty, B. (2015) Nitrogen isotope variations in the Solar System. Nature Geoscience 8, 515-522.

), there may also be a small, but detectable, primitive (i.e. solar) nitrogen component in the CIR mantle.

For a primordial starting neon component we choose here 21Ne/22Ne that is associated with solar wind implanted Ne–B (20Ne/22Ne = 12.52 ± 0.18). Solar Ne–B as found in meteorites (Black, 1972

Black, D.C. (1972) On the origins of trapped helium, neon and argon isotopic variations in meteorites – I. Gas-rich meteorites, lunar soil and breccia. Geochimica et Cosmochimica Acta 36, 347–375.

) was long regarded as the actual solar composition, but Benkert et al. (1993)

Benkert, J.P., Baur, H., Signer, P., Wieler, R. (1993) He, Ne, and Ar from the solar-wind and solar energetic particles in lunar ilmenites and pyroxenes. Journal of Geophysical Research – Planets 98, 13147–13162.

have shown that it is composed of low energetic solar wind particles (20Ne/22Ne = 13.8 ± 0.1; 21Ne/22Ne = 0.0328 ± 0.0005) and a mass fractionated high energetic solar energetic particles (SEP) component (20Ne/22Ne = 11.2 ± 0.2; 21Ne/22Ne = 0.0295 ± 0.0005), more deeply implanted and thus separable by stepwise etching of the carrier grains. We accordingly calculated 21Ne/22Ne of Ne–B as a mixture of these two subcomponents, yielding 21Ne/22Ne = 0.03118 ± 0.00048. This value is in agreement with, but more precise than, the value given by Black (1972)

Black, D.C. (1972) On the origins of trapped helium, neon and argon isotopic variations in meteorites – I. Gas-rich meteorites, lunar soil and breccia. Geochimica et Cosmochimica Acta 36, 347–375.

of 21Ne/22Ne = 0.0335 ± 0.0015.

He-Ne Isotope Variability

In Figure S-5, the air-corrected 21Ne/22NeEX values of CIR basalts as well as Réunion xenolith samples are plotted against He-isotopes, together with a series of binary mixing trajectories between postulated primordial (PRIM) and DMM end-members. Notably, 21Ne/22NeEX values in Réunion xenoliths range from 0.0325 ± 0.0108 to 0.0513 ± 0.0120, and fall in the range previously measured in Réunion lavas (Hanyu et al., 2001

Hanyu, T., Dunai, T.J., Davies, G.R., Kaneoka, I., Nohda, S., Uto, K. (2001) Noble gas study of the Reunion hot spot: evidence for distinct less‐degassed mantle sources. Earth and Planetary Science Letters 193, 83–98.

; Trieloff et al., 2002

Trieloff, M., Kunz, J., Allègre C.J. (2002) Noble gas systematics of the Réunion mantle plume source and the origin of primordial noble gases in Earth’s mantle. Earth and Planetary Science Letters 200, 297–313.

; Hopp and Trieloff, 2005

Hopp, J., Trieloff, M. (2005) Refining the noble gas record of the Réunion mantle plume source: Implications on mantle geochemistry. Earth and Planetary Science Letters 240, 573–588.

). The 21Ne/22Ne of PRIM (i.e. primitive solar component) is again assumed to be 0.03118 (Trieloff and Kunz, 2005

Trieloff, M., Kunz, J. (2005) Isotope systematics of noble gases in the Earth's mantle: possible sources of primordial isotopes and implications for mantle structure. Physics of the Earth and Planetary Interiors 148, 13-38.

) and the DMM 21Ne/22Ne endmember value is estimated to be 0.0594 by extrapolating the DMM trajectory (Sarda et al, 1988

Sarda, P., Staudacher, T., Allègre, C.J. (1988) Neon isotopes in submarine basalts. Earth and Planetary Science Letters 91, 73-88.

) to the 20Ne/22Ne of Ne-B (= 12.5). In the case of He-isotopes, we assume a DMM endmember 4He/3He ratio of 90,000 (= 8 RA) (Graham, 2002

Graham, D.W. (2002) Noble gas isotope geochemistry of mid‐ocean ridge and ocean island basalts: Characterization of mantle source reservoirs. RIMS 47, 247–317.

), and a primitive (PRIM) He-isotope endmember value of 280 RA (Black, 1972

Black, D.C. (1972) On the origins of trapped helium, neon and argon isotopic variations in meteorites – I. Gas-rich meteorites, lunar soil and breccia. Geochimica et Cosmochimica Acta 36, 347–375.

; 4He/3He = 2580), consistent with the Ne-B component (i.e. both values were measured in gas-rich meteorite samples; Black, 1972

Black, D.C. (1972) On the origins of trapped helium, neon and argon isotopic variations in meteorites – I. Gas-rich meteorites, lunar soil and breccia. Geochimica et Cosmochimica Acta 36, 347–375.

). The curvature of binary mixing trajectories is controlled by the r-value = (3He/22Ne)DMM / (3He/22Ne)PLUME. For reference, we also plot 21Ne/22NeEX versus 4He/3He data fields for previous studies of Réunion lavas (Hanyu et al., 2001

Hanyu, T., Dunai, T.J., Davies, G.R., Kaneoka, I., Nohda, S., Uto, K. (2001) Noble gas study of the Reunion hot spot: evidence for distinct less‐degassed mantle sources. Earth and Planetary Science Letters 193, 83–98.

), and Iceland (Füri et al., 2010

Füri, E., Hilton, D.R., Halldórsson, S.A., Barry, P.H., Hahm, D., Fischer, T.P., Gronvold, K. (2010) Apparent decoupling of the He and Ne isotope systematics of the Icelandic mantle: the role of He depletion, melt mixing, degassing fractionation and air interaction. Geochimica et Cosmochimica Acta 74, 3307–3332.

), which represent the Réunion-plume component as well as an archetypal plume-MOR interaction.

The observed variations between Ne and He isotopes can be described by binary mixing between a primordial mantle endmember (PRIM) and a second endmember similar to DMM. However, unlike Ne-N relationships, CIR basalt and Réunion xenolith samples do not fall on a single Ne-He mixing curve, but rather require r-values between 0.4 and 30 to explain observed variations (Fig. S-5). Apart from one obvious outlier, on-axis CIR samples plot close to PRIM-DMM mixing curves with r-values between 1 and 30, whereas the majority of off-axis samples fall between 1 and 5, with one outlier with an r-value of 30. The fact that all basalt samples require r-values greater than unity, implies a higher 3He/22Ne ratio of the DMM endmember compared to the 3He/22Ne ratio of the primordial mantle component (i.e. (3He/22Ne)DMM > (3He/22Ne)PRIM).

If we assume that the DMM endmember 3He/22Ne ratio is constant throughout the CIR, then lower r-values in off-axis samples suggest a higher relative PRIM 3He/22Ne endmember in the plume source. In this regard, CIR basalts resemble other plume-related glasses (e.g., Iceland; Füri et al., 2010

Füri, E., Hilton, D.R., Halldórsson, S.A., Barry, P.H., Hahm, D., Fischer, T.P., Gronvold, K. (2010) Apparent decoupling of the He and Ne isotope systematics of the Icelandic mantle: the role of He depletion, melt mixing, degassing fractionation and air interaction. Geochimica et Cosmochimica Acta 74, 3307–3332.

). Significantly, all Réunion xenolith samples plot closer to the PRIM endmember compared with CIR basalt samples and overlap with Réunion lavas (Hanyu et al., 2001

Hanyu, T., Dunai, T.J., Davies, G.R., Kaneoka, I., Nohda, S., Uto, K. (2001) Noble gas study of the Reunion hot spot: evidence for distinct less‐degassed mantle sources. Earth and Planetary Science Letters 193, 83–98.

), consistent with a higher 3He/22Ne plume contribution in the admixture. Notably, mixing trajectories for xenolith samples span the entire range observed in CIR basalts (i.e. r-values between ~0.4 and 30), however, this range is bracketed by Réunion lava samples (Hanyu et al., 2001

Hanyu, T., Dunai, T.J., Davies, G.R., Kaneoka, I., Nohda, S., Uto, K. (2001) Noble gas study of the Reunion hot spot: evidence for distinct less‐degassed mantle sources. Earth and Planetary Science Letters 193, 83–98.

), which span an even broader range (Fig. S-5). The single outlying on-axis basalt sample (Fig. S-2; D8-2) falls outside the mixing envelope defined by the series of binary mixing curves between PRIM and DMM endmembers; D8-2 is marked by lower than DMM He-isotopes and thus an additional endmember (with low 3He/4He) is required to account for the He–Ne systematics; radiogenic He (accompanied by insignificant nucleogenic Ne) produced within the crust is a distinct possibility. This observation in agreement with the conclusions of Füri et al. (2011)

Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J. M. D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.

, who explained low He-isotope (i.e. 3He/4He < DMM) samples observed in the ~18.1 ºS portion of the CIR by closed system radiogenic 4He in-growth in a “fossil” Réunion mantle component. Furthermore, this finding is consistent with larger contributions from the Réunion-like (i.e. plume) component, which were identified in the Ne-N isotope systematics of sample D8-2.

An alternative explanation for variations in 21Ne/22Neex values between presumed upper and lower mantle reservoirs is elemental He/Ne heterogeneities in the mantle. For example, Moreira et al. (2001)

Moreira, M., Breddam, K., Curtice, J., Kurz, M.D. (2001) Solar neon in the Icelandic mantle: new evidence for an undegassed lower mantle. Earth and Planetary Science Letters 185, 15–23.

proposed that different mantle reservoirs (i.e. DMM versus PRIM) could have evolved from the same He and Ne isotope compositions but with distinct 3He/22Ne ratios and that subsequent ingrowth of nucleogenic 21Ne and radiogenic 4He have produced the observed range in 21Ne/22Neex and 3He/4He values. Nucleogenic 21Ne is generated by the 18O (α,n) 21Ne and 24Mg (n,α) 21Ne reactions (Wetherill, 1954), resulting in an increase in the 21Ne/22Ne ratio and δ21Ne values over time. Since α-particles are derived from U and Th decay, the production of radiogenic 4He and nucleogenic 21Ne (21Ne*) should therefore be directly coupled, and high 4He/3He (i.e. low 3He/4He) ratios would therefore correlate with high 21Ne/22NeEX values. However, in CIR basalts and Réunion xenoliths, measured 3He/22Ne ratios vary over 2 orders of magnitude (0.02 to 3.93). If mantle heterogeneities were the origin of these variations it would imply that the 3He/22Ne ratio in the mantle source varied by a factor of ~200. However, there is no known mechanism that can preserve such large 3He/22Ne disparities within a single mantle domain for such long durations (i.e. over Earth history). Therefore, we propose that the He–Ne characteristics of CIR basalt and Réunion xenolith samples are better explained by the binary mixing scenario presented in Figure S-5.



Figure S-5 (21Ne/22Ne)EX (i.e. air-corrected 21Ne/22Ne values) versus 4He/3He values of CIR basalts and Réunion xenoliths, plotted together with binary mixing curves between a primordial mantle endmember (PRIM) and a DMM-like component. The curvature of the hyperbolic mixing lines is described by r = (3He/22Ne)DMM/(3He/22Ne)PRIM. In addition, data fields are shown for Réunion lavas (Hanyu et al., 2001

Hanyu, T., Dunai, T.J., Davies, G.R., Kaneoka, I., Nohda, S., Uto, K. (2001) Noble gas study of the Reunion hot spot: evidence for distinct less‐degassed mantle sources. Earth and Planetary Science Letters 193, 83–98.

) and Iceland subglacial basalts (Füri et al., 2010

Füri, E., Hilton, D.R., Halldórsson, S.A., Barry, P.H., Hahm, D., Fischer, T.P., Gronvold, K. (2010) Apparent decoupling of the He and Ne isotope systematics of the Icelandic mantle: the role of He depletion, melt mixing, degassing fractionation and air interaction. Geochimica et Cosmochimica Acta 74, 3307–3332.

).
Full size image | Download in Powerpoint


Table S-1 Mantle regassing sensitivity to various input parameters.
StudyaDMM Endmember δ15N (‰)PLM Endmember δ15N (‰)Initial Mantle Endmember δ15N (‰)bStarting [N] of mantle (ppm)cOnset of subduction (Ga)dRegassing Flux (FDMM) (×109mol/yr)Regassing Flux (FPLM) (×109mol/yr)FPLM/FDMM
 
 
This study-2.11.3-400.271.116.753.33.2



-400.272.59.424.22.6



-400.273.97.315.82.2
 



-40361.1176771064



-40362.578131334



-40363.951720714
 



-3830.271.119.758.33



-3830.272.51126.22.4



-3830.273.98.617.52
 



-383361.1206576533.7



-383362.591133763.7



-383363.960122293.7
 
 
Johnson and Goldblatt, 2015 Johnson, B., Goldblatt, C. (2015) The nitrogen budget of Earth. Earth-Science Reviews 148, 150-173. -1.12.5-400.271.116.854.23.2



-400.272.59.424.12.6



-400.273.97.516.22.1
 



-40361.1181573174



-40362.580132224



-40363.952921244
 



-3830.271.119.758.33



-3830.272.51126.32.4



-3830.273.98.717.42
 



-383361.1207477263.7



-383362.591434023.7



-383363.960222443.7
 
 
Dauphas and Marty. 1999 Dauphas, N., Marty, B. (1999) Heavy nitrogen in carbonatites of the Kola Peninsula: A possible signature of the deep mantle. Science 286, 2488-2490. -43-400.271.11656.53.5



-400.272.58.925.12.8



-400.273.9716.42.3
 



-40361.1168474164.4



-40362.574432694.4



-40363.949121524.4
 



-3830.271.119.558.53



-3830.272.510.926.12.4



-3830.273.98.5172
 



-383361.1206177393.7



-383362.590533923.7



-383363.959622293.7

a The average N-isotope values for the modern DMM and PLM (e.g.,OIB-source) mantle source regions from these studies are used.
b Initial mantle δ15N endmember values of -40 ‰ and -383 ‰ are from Cartigny and Marty (2013) Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366. and Marty et al., 2011 Marty, B., Chaussidon, M., Jurewicz, A.,Wiens, R., Burnett, D.S. (2011) A 15N-poor isotopic composition for the solar system as shown by Genesis solar wind samples. Science 332, 1533–1536. , respectively.
c Initial mantle [N] endmember contents of 0.27 ppm and 36 ppm are from Marty (2012) Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66. and Cartigny and Marty (2013) Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366. , respectively.
d Onset of subduction estimates of 1.1 Ga, 2.5 Ga and 3.9 Ga are based on Cartigny and Marty (2013) Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366. , Kusky et al. (2001) Kusky, T.M., Li, J.H., Tucker, R.D. (2001) The Archean Dongwanzi ophiolite complex, North China Craton: 2.505-billion-year-old oceanic crust and mantle. Science 292, 1142-1145. and Condie and Pease (2008) Condie, K.C., Pease, V. (2008) When did plate tectonics begin on planet Earth? Geological Society of America Special Paper 440, Boulder Colorado, USA. , respectively.
Values in Italics are assumptions, remaining values are model output.

Download in Excel

Supplementary Information References


Arevalo, R., McDonough, W.F., Luong, M. (2009) The K/U ratio of the silicate Earth: Insights into mantle composition, structure and thermal evolution. Earth and Planetary Science Letters 278, 361-369.
Show in context

Using estimates of upper mantle K/U (Arevalo et al., 2009) and Th/U (O'Nions and McKenzie, 1993), a production ratio for 4He/40Ar* of ~2 (Jambon et al., 1986; Marty and Zimmermann, 1999) is assumed as a mantle starting value.
View in Supplementary Information


Ballentine, C.J., Barfod, D.N. (2000) The origin of air-like noble gases in MORB and OIB. Earth and Planetary Science Letters 180, 39–48.
Show in context

Air-like Ne-isotope contributions are ubiquitous in natural samples (Graham et al., 2002) with infiltration of air into basalts occurring through small micro-fractures, which can develop in glasses during seafloor sampling and/or handling in the laboratory (Ballentine and Barfod, 2000), and/or by assimilation of seawater-derived components (Farley and Craig, 1994).
View in Supplementary Information


Ballentine, C.J., Marty, B., Lollar, B.S., Cassidy, M. (2005) Neon isotopes constrain convection and volatile origin in the Earth’s mantle. Nature 433, 33–38.
Show in context

The Ne–B endmember is obtained from gas-rich meteorites and lunar soils with a 20Ne/22Ne ratio of 12.52 ± 0.18, considered to represent implanted solar neon, now present within Earth’s mantle (Black, 1972; Trieloff et al., 2000; Trieloff et al., 2002; Ballentine et al., 2005; Moreira, 2013).
View in Supplementary Information


Barry, P.H., Hilton, D.R., Halldórsson, S.A., Hahm, D., Marti, K. (2012) High precision nitrogen isotope measurements in oceanic basalts using a static triple collection noble gas mass spectrometer. Geochemistry Geophysics Geosystems 13, Q01019.
Show in context

Nitrogen was released using the same crushing mechanism described above for Ne-Ar; however, gases were purified on a separate vacuum clean-up line (see Barry et al. (2012) for details) prior to inlet into the VG5440 mass spectrometer, which was operated in static triple collection mode.
View in Supplementary Information


Bebout, G.E., Lazzeri, K.E., Geiger, C.A. (2016) Pathways for nitrogen cycling in Earth's crust and upper mantle: A review and new results for microporous beryl and cordierite. American Mineralogist 101, 7-24.
Show in context

The large isotopic contrast between terrestrial nitrogen reservoirs makes N a potentially powerful tracer of volatile recycling between the surface and mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in Supplementary Information


Benkert, J.P., Baur, H., Signer, P., Wieler, R. (1993) He, Ne, and Ar from the solar-wind and solar energetic particles in lunar ilmenites and pyroxenes. Journal of Geophysical Research – Planets 98, 13147–13162.
Show in context

The primordial, “solar” neon component in Earth’s mantle has also been assumed to be best represented by present-day solar wind (i.e. 20Ne/22Ne = 13.8; Benkert et al., 1993; Kallenbach et al., 1997); note, however, that selection of this endmember as opposed to Ne-B will not significantly affect the conclusions in the following discussion.
View in Supplementary Information
Solar Ne–B as found in meteorites (Black, 1972) was long regarded as the actual solar composition, but Benkert et al. (1993) have shown that it is composed of low energetic solar wind particles (20Ne/22Ne = 13.8 ± 0.1; 21Ne/22Ne = 0.0328 ± 0.0005) and a mass fractionated high energetic solar energetic particles (SEP) component (20Ne/22Ne = 11.2 ± 0.2; 21Ne/22Ne = 0.0295 ± 0.0005), more deeply implanted and thus separable by stepwise etching of the carrier grains.
View in Supplementary Information


Black, D.C. (1972) On the origins of trapped helium, neon and argon isotopic variations in meteorites – I. Gas-rich meteorites, lunar soil and breccia. Geochimica et Cosmochimica Acta 36, 347–375.
Show in context

The Ne–B endmember is obtained from gas-rich meteorites and lunar soils with a 20Ne/22Ne ratio of 12.52 ± 0.18, considered to represent implanted solar neon, now present within Earth’s mantle (Black, 1972; Trieloff et al., 2000; Trieloff et al., 2002; Ballentine et al., 2005; Moreira, 2013).
View in Supplementary Information
Solar Ne–B as found in meteorites (Black, 1972) was long regarded as the actual solar composition, but Benkert et al. (1993) have shown that it is composed of low energetic solar wind particles (20Ne/22Ne = 13.8 ± 0.1; 21Ne/22Ne = 0.0328 ± 0.0005) and a mass fractionated high energetic solar energetic particles (SEP) component (20Ne/22Ne = 11.2 ± 0.2; 21Ne/22Ne = 0.0295 ± 0.0005), more deeply implanted and thus separable by stepwise etching of the carrier grains.
View in Supplementary Information
In the case of He-isotopes, we assume a DMM endmember 4He/3He ratio of 90,000 (= 8 RA) (Graham, 2002), and a primitive (PRIM) He-isotope endmember value of 280 RA (Black, 1972; 4He/3He = 2580) in order to be consistent with the Ne-B component (i.e. both values were measured in gas-rich meteorite samples; Black, 1972).
View in Supplementary Information


Busigny, V., Lebeau, O., Ader, M., Krapez, B., Bekker, A. (2013) Nitrogen cycle in a Late Archean ferruginous ocean. Chemical Geology 362, 115-130.
Show in context

Subsequent modifications to mantle N may result from subduction of this material comprised of isotopically distinct N (~+5‰) (Sano et al., 1998; Busigny et al., 2013; Thomazo and Papineau, 2013; Halama et al., 2014).
View in Supplementary Information


Cartigny, P., Marty, B. (2013) Nitrogen isotopes and mantle geodynamics: The emergence of life and the atmosphere–crust–mantle connection. Elements 9, 359-366.
Show in context

For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986; Cartigny et al., 1998; Marty and Zimmermann, 1999; Marty and Dauphas, 2003; Cartigny and Marty, 2013; Palot et al., 2012; Johnson and Goldblatt, 2015; this study).
View in Supplementary Information
Table S-1 [...] b Initial mantle δ15N endmember values of -40 ‰ and -383 ‰ are from Cartigny and Marty (2013) and Marty et al., 2011, respectively.
c Initial mantle [N] endmember contents of 0.27 ppm and 36 ppm are from Marty (2012) and Cartigny and Marty (2013), respectively.
d Onset of subduction estimates of 1.1 Ga, 2.5 Ga and 3.9 Ga are based on Cartigny and Marty (2013), Kusky et al. (2001) and Condie and Pease (2008), respectively.
View in Supplementary Information


Cartigny, P., Harris, J.W., Phillips, D., Girard, M., Javoy M. (1998) Subduction-related diamonds? The evidence for a mantle-derived origin from coupled δ13C - δ15N determinations. Chemical Geology 147, 147-159.
Show in context

For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986; Cartigny et al., 1998; Marty and Zimmermann, 1999; Marty and Dauphas, 2003; Cartigny and Marty, 2013; Palot et al., 2012; Johnson and Goldblatt, 2015; this study).
View in Supplementary Information


Cartigny, P., Jendrzejewski, N., Pineau, F., Petit, E., Javoy, M. (2001) Volatile (C, N, Ar) variability in MORB and the respective roles of mantle source heterogeneity and degassing: the case of the Southwest Indian Ridge. Earth and Planetary Science Letters 194, 241–257.
Show in context

To date, however, such N-isotope anomalies have rarely been detected in the DMM (Marty and Zimmermann, 1999; Cartigny et al., 2001).
View in Supplementary Information


Condie, K.C., Pease, V. (2008) When did plate tectonics begin on planet Earth? Geological Society of America Special Paper 440, Boulder Colorado, USA.
Show in context

Table S-1 [...] d Onset of subduction estimates of 1.1 Ga, 2.5 Ga and 3.9 Ga are based on Cartigny and Marty (2013), Kusky et al. (2001) and Condie and Pease (2008), respectively.
View in Supplementary Information


Courtillot, V., Davaille, A., Besse, J., Stock, J. (2003) Three distinct types of hotspots in the Earth’s mantle. Earth and Planetary Science Letters 205, 295-308.
Show in context

Using He isotopes ratios (3He/4He) – the canonical tracer of mantle plume involvement in petrogenesis (Courtillot et al., 2003) – Füri et al. (2011) identified plume material flowing eastward toward the CIR on a trajectory that impinges the ridge at ~19.9 °S.
View in Supplementary Information


Craig, H., Marti, K., Wiens, R. (1993) A Static Mass Spectrometer With Triple Collection for Nitrogen and Neon Isotopes. SIO Reference Series. Scripps Institute of Oceanography, La Jolla, California, USA, 93-11, 1–20A.
Show in context

Neon and argon abundances and isotope ratios (as well as He abundances) were measured using a modified VG5440 mass spectrometer equipped with five Faraday cups and Daly photo-multiplier detector (Craig et al., 1993; Hahm et al., 2012), operated in peak jumping mode.
View in Supplementary Information


Dauphas, N., Marty, B. (1999) Heavy nitrogen in carbonatites of the Kola Peninsula: A possible signature of the deep mantle. Science 286, 2488-2490.
Show in context

Table S-1
View in Supplementary Information


Duncan, R.A., Backman, J., Peterson, L., Shipboard Scientific Party (1989) Réunion hotspot activitity through tertiary time: Initial results from the ocean drilling program, leg 115. Journal of Volcanology and Geothermal Research 36, 193-198.
Show in context

During the past ~70 Ma, the Réunion hot spot formed the Deccan Trap flood basalts, the Chagos-Maldive-Laccadive Ridge, the Mascarene Plateau and the Mascarene Islands (Duncan et al., 1989, 1990).
View in Supplementary Information


Duncan, R.A., Backman, J., Peterson, L. (1990) The volcanic record of the Réunion hot spot. Proceedings of the Ocean Drilling Program Scientific Results 115, 3–10.
Show in context

During the past ~70 Ma, the Réunion hot spot formed the Deccan Trap flood basalts, the Chagos-Maldive-Laccadive Ridge, the Mascarene Plateau and the Mascarene Islands (Duncan et al., 1989, 1990).
View in Supplementary Information
At ~34 Ma, the northeastward migrating CIR moved over the Réunion hot spot (Duncan et al., 1990) and embedded a “fossil” Réunion hot spot mantle component into the sub-ridge mantle.
View in Supplementary Information
Rodrigues Island is located ~600 km east of Mauritius and formed at 1.5 Ma (McDougall et al., 1965); it marks the eastern extent of the Rodrigues Ridge, an east-west trending volcanic ridge that is between 8 to 10 Ma old (Duncan et al., 1990).
View in Supplementary Information


Dyment, J., Gallet, Y., Magofond 2 Scientific Party (1999) The Magofond 2 cruise: A surface and deep tow survey on the past and present Central Indian Ridge. InterRidge News 8, 25–31.
Show in context

Smaller en-échelon volcanic ridges at 19 °S (e.g., the Three Magi Ridges and the Gasitao Ridge) extend the Rodrigues Ridge close to the CIR (Fig. S-1) (Dyment et al., 1999, 2000).
View in Supplementary Information
Figure S-1 [...] Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999), R/V L'Atalante (Dyment et al., 2000), and R/V Hakuho-Maru (Okino et al., 2008); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008; Füri et al., 2011).
View in Supplementary Information


Dyment, J., Hémond, C., Gimnaut Scientific Party (2000) Deep‐sea exploration of the Central Indian Ridge at 19°S. InterRidge News 9, 29–32.
Show in context

Smaller en-échelon volcanic ridges at 19 °S (e.g., the Three Magi Ridges and the Gasitao Ridge) extend the Rodrigues Ridge close to the CIR (Fig. S-1) (Dyment et al., 1999, 2000).
View in Supplementary Information
Figure S-1 [...] Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999), R/V L'Atalante (Dyment et al., 2000), and R/V Hakuho-Maru (Okino et al., 2008); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008; Füri et al., 2011).
View in Supplementary Information


Farley, K.A., Craig, H. (1994) Atmospheric argon contamination of ocean island basalt olivine phenocrysts. Geochimica et Cosmochimica Acta 58, 2509–2517.
Show in context

Air-like Ne-isotope contributions are ubiquitous in natural samples (Graham et al., 2002) with infiltration of air into basalts occurring through small micro-fractures, which can develop in glasses during seafloor sampling and/or handling in the laboratory (Ballentine and Barfod, 2000), and/or by assimilation of seawater-derived components (Farley and Craig, 1994).
View in Supplementary Information


Fischer, T.P., Hilton, D.R., Zimmer, M.M., Shaw, A.M., Sharp, Z.D., Walker, J.A. (2002) Subduction and recycling of nitrogen along the Central American margin. Science 297, 1154-1157.
Show in context

N-isotope enrichments have been observed in subduction-related geothermal gases (Fischer et al., 2002) and ocean island basalts (OIBs; Hoffman and White, 1982; Mohapatra and Murty, 2002; Marty and Dauphas, 2003; Fischer et al., 2005).
View in Supplementary Information


Fischer, T.P., Takahata, N., Sano, Y., Sumino, H., Hilton, D.R. (2005) Nitrogen isotopes of the mantle: Insights from mineral separates. Geophysical Research Letters 32, L11305.
Show in context

N-isotope enrichments have been observed in subduction-related geothermal gases (Fischer et al., 2002) and ocean island basalts (OIBs; Hoffman and White, 1982; Mohapatra and Murty, 2002; Marty and Dauphas, 2003; Fischer et al., 2005).
View in Supplementary Information
Significantly, this observation is consistent with previous compilation studies of various mid ocean ridge (MOR) and OIB settings (e.g., Marty and Humbert, 1997; Fischer et al., 2005), which also suggest that nitrogen isotope variations are independent of degassing.
View in Supplementary Information


Fretzdorff, S., Haase, K.M. (2002) Geochemistry and petrology of lavas from the submarine flanks of Réunion Island (western Indian Ocean): implications for magma genesis and the mantle source. Mineralogy and Petrology 75, 153-184.
Show in context

Furthermore, they compared basalt samples with submarine Réunion lavas (Fretzdorff and Haase, 2002) and concluded that the radiogenic isotope characteristics of the majority of on-axis samples cannot be explained with a Réunion-like endmember.
View in Supplementary Information


Füri, E., Marty, B. (2015) Nitrogen isotope variations in the Solar System. Nature Geoscience 8, 515-522.
Show in context

Assuming solar wind samples accurately approximate the proto-solar nebula (PSN; Marty et al., 2011; Marty, 2012; Füri and Marty, 2015), there may also be a small, but detectable, primitive (i.e. solar) nitrogen component in the CIR mantle.
View in Supplementary Information


Füri, E., Hilton, D.R., KNOX11RR Scientific Party (2008) Sampling and surveying ridge‐hot spot interaction on the Central Indian Ridge, 19°S: Cruise KNOX11RR. InterRidge News 17, 28–29.
Show in context

Twenty-four submarine pillow basalts were dredged from the ocean floor along the CIR axis (16.7 °S to 20.6 °S), and the adjacent Gasitao, Three Magi, and Rodrigues ridges on our Knox11RR expedition (Füri et al., 2008, 2011; Fig. S-1).
View in Supplementary Information
Figure S-1 [...] Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999), R/V L'Atalante (Dyment et al., 2000), and R/V Hakuho-Maru (Okino et al., 2008); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008, 2011).
View in Supplementary Information


Füri, E., Hilton, D.R., Halldórsson, S.A., Barry, P.H., Hahm, D., Fischer, T.P., Gronvold, K. (2010) Apparent decoupling of the He and Ne isotope systematics of the Icelandic mantle: the role of He depletion, melt mixing, degassing fractionation and air interaction. Geochimica et Cosmochimica Acta 74, 3307–3332.
Show in context

Basalt CIR 21Ne/22NeEX values vary between 0.0437 ± 0.0009 and 0.0582 ± 0.0013 and fall in the range of values observed in previous studies of plume-rift related systems (e.g., Füri et al., 2010; Hahm et al., 2012).
View in Supplementary Information
For reference, we also plot 21Ne/22NeEX versus 4He/3He data fields for previous studies of Réunion lavas (Hanyu et al., 2001), and Iceland (Füri et al., 2010), which represent the Réunion-plume component as well as an archetypal plume-MOR interaction.
View in Supplementary Information
In this regard, CIR basalts resemble other plume-related glasses (e.g., Iceland; Füri et al., 2010).
View in Supplementary Information
Figure S-5 [...] In addition, data fields are shown for Réunion lavas (Hanyu et al., 2001) and Iceland subglacial basalts (Füri et al., 2010).
View in Supplementary Information


Füri, E., Hilton, D.R., Murton, B.J., Hemond, C., Dyment, J., Day, J. M. D. (2011) Helium isotope variations between Réunion Island and the Central Indian Ridge (17˚-21˚S): new evidence for ridge-hotspot interaction. Journal of Geophysical Research – Solid Earth 116, B02207.
Show in context

Using He isotopes ratios (3He/4He) – the canonical tracer of mantle plume involvement in petrogenesis (Courtillot et al., 2003) – Füri et al. (2011) identified plume material flowing eastward toward the CIR on a trajectory that impinges the ridge at ~19.9 °S.
View in Supplementary Information
Füri et al. (2011), however, suggested that these enrichments, along with slightly radiogenic He-isotope values could result from a “fossil” Réunion plume component in the region.
View in Supplementary Information
Twenty-four submarine pillow basalts were dredged from the ocean floor along the CIR axis (16.7 °S to 20.6 °S), and the adjacent Gasitao, Three Magi, and Rodrigues ridges on our Knox11RR expedition (Füri et al., 2008, 2011; Fig. S-1).
View in Supplementary Information
This observation in agreement with the conclusions of Füri et al. (2011), who explained low He-isotope (i.e. 3He/4He < DMM) samples observed in the ~18.1 ºS portion of the CIR by closed system radiogenic 4He in-growth in a “fossil” Réunion mantle component.
View in Supplementary Information
Figure S-1 [...] Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999), R/V L'Atalante (Dyment et al., 2000), and R/V Hakuho-Maru (Okino et al., 2008); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008, 2011).
View in Supplementary Information


Gillot, P.Y., Nativel, P. (1989) Eruptive history of the Piton de la Fournaise volcano, Réunion, Indian Ocean. Journal of Volcanology and Geothermal Research 36, 53–65.
Show in context

Réunion Island is composed of two volcanoes: Piton des Neiges, which was active between ~2 Ma and 43 ka, and the currently active Piton de la Fournaise, where volcanic activity began ~530 ka ago (Gillot and Nativel, 1989).
View in Supplementary Information


Graham, D.W. (2002) Noble gas isotope geochemistry of mid‐ocean ridge and ocean island basalts: Characterization of mantle source reservoirs. RIMS 47, 247–317.
Show in context

Air-like Ne-isotope contributions are ubiquitous in natural samples (Graham et al., 2002) with infiltration of air into basalts occurring through small micro-fractures, which can develop in glasses during seafloor sampling and/or handling in the laboratory (Ballentine and Barfod, 2000), and/or by assimilation of seawater-derived components (Farley and Craig, 1994).
View in Supplementary Information
In the case of He-isotopes, we assume a DMM endmember 4He/3He ratio of 90,000 (= 8 RA) (Graham, 2002), and a primitive (PRIM) He-isotope endmember value of 280 RA (Black, 1972; 4He/3He = 2580) in order to be consistent with the Ne-B component (i.e. both values were measured in gas-rich meteorite samples; Black, 1972).
View in Supplementary Information


Hahm, D., Hilton, D.R., Castillo, P.R., Hawkins, J.W., Hanan, B.B., Hauri, E.H. (2012) An overview of the volatile systematics of the Lau Basin – Resolving the effects of source variation, magmatic degassing and crustal contamination. Geochimica et Cosmochimica Acta 85, 88-113.
Show in context

Neon and argon gases were released from basalts using a custom-built piston-activated in vacuo rock-crusher (see Hahm et al. (2012) for details).
View in Supplementary Information
Neon and argon abundances and isotope ratios (as well as He abundances) were measured using a modified VG5440 mass spectrometer equipped with five Faraday cups and Daly photo-multiplier detector (Craig et al., 1993; Hahm et al., 2012), operated in peak jumping mode.
View in Supplementary Information
Basalt CIR 21Ne/22NeEX values vary between 0.0437 ± 0.0009 and 0.0582 ± 0.0013 and fall in the range of values observed in previous studies of plume-rift related systems (e.g., Füri et al., 2010; Hahm et al., 2012).
View in Supplementary Information


Halama, R., Bebout, G.E., John, T., Schenk, V. (2010). Nitrogen recycling in subducted oceanic lithosphere: The record in high-and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636-1652.
Show in context

As a result, subduction of altered oceanic crust and sediments, containing molecularly-bound nitrogen – residual to sub-arc fluid loss – could result in the introduction of high N-isotope signatures into the (deep) mantle (Hofmann and White, 1982; Marty and Humbert, 1997; Halama et al., 2010).
View in Supplementary Information


Halama, R., Bebout, G.E., John, T., Scambelluri, M. (2014). Nitrogen recycling in subducted mantle rocks and implications for the global nitrogen cycle. International Journal of Earth Sciences 103, 2081-2099.
Show in context

Subsequent modifications to mantle N may result from subduction of this material comprised of isotopically distinct N (~+5‰) (Sano et al., 1998; Busigny et al., 2013; Thomazo and Papineau, 2013; Halama et al., 2014).
View in Supplementary Information


Hanyu, T., Dunai, T.J., Davies, G.R., Kaneoka, I., Nohda, S., Uto, K. (2001) Noble gas study of the Reunion hot spot: evidence for distinct less‐degassed mantle sources. Earth and Planetary Science Letters 193, 83–98.
Show in context

Notably, 21Ne/22NeEX values in Réunion xenoliths range from 0.0325 ± 0.0108 to 0.0513 ± 0.0120, and fall in the range previously measured in Réunion lavas (Hanyu et al., 2001; Trieloff et al., 2002; Hopp and Trieloff, 2005).
View in Supplementary Information
For reference, we also plot 21Ne/22NeEX versus 4He/3He data fields for previous studies of Réunion lavas (Hanyu et al., 2001), and Iceland (Füri et al., 2010), which represent the Réunion-plume component as well as an archetypal plume-MOR interaction.
View in Supplementary Information
Significantly, all Réunion xenolith samples plot closer to the PRIM endmember compared with CIR basalt samples and overlap with Réunion lavas (Hanyu et al., 2001), consistent with a higher 3He/22Ne plume contribution in the admixture.
View in Supplementary Information
Notably, mixing trajectories for xenolith samples span the entire range observed in CIR basalts (i.e. r-values between ~0.4 and 30), however, this range is bracketed by Réunion lava samples (Hanyu et al., 2001), which span an even broader range (Fig. S-5).
View in Supplementary Information
Figure S-5 [...] In addition, data fields are shown for Réunion lavas (Hanyu et al., 2001) and Iceland subglacial basalts (Füri et al., 2010).
View in Supplementary Information


Hilton, D.R., Fischer, T.P., Marty, B. (2002) Noble gases and volatile recycling at subduction zones. In: Porcelli, D., Ballentine, C.J., Wieler, R. (Eds.) Noble Gases in Cosmochemistry and Geochemistry. Mineralogical Society of America 47 319-370.
Show in context

Heavier noble gas compositions in the mantle were long considered to be independent of subduction as they were believed to be returned back into the atmosphere through subduction volcanism before they could be admixed into the mantle (Staudacher and Allègre, 1988; Hilton et al., 2002).
View in Supplementary Information


Hofmann, A.W., White, W.M. (1982) Mantle plumes from ancient oceanic crust. Earth and Planetary Science Letters 57, 421-436.
Show in context

As a result, subduction of altered oceanic crust and sediments, containing molecularly-bound nitrogen – residual to sub-arc fluid loss – could result in the introduction of high N-isotope signatures into the (deep) mantle (Hofmann and White, 1982; Marty and Humbert, 1997; Halama et al., 2010).
View in Supplementary Information
N-isotope enrichments have been observed in subduction-related geothermal gases (Fischer et al., 2002) and ocean island basalts (OIBs; Hoffman and White, 1982; Mohapatra and Murty, 2002; Marty and Dauphas, 2003; Fischer et al., 2005).
View in Supplementary Information


Holland, G., Ballentine, C.J. (2006) Seawater subduction controls the heavy noble gas composition of the mantle. Nature 441, 186-191.
Show in context

Light noble gases (helium and neon) in Earth's mantle preserve distinct solar-like isotopic compositions, whereas heavy noble gases (argon, krypton and xenon) display values which are intermediate between solar-like and air-dominated isotopic compositions (Holland and Ballentine, 2006).
View in Supplementary Information
The fate of Ne is debated (Sarda, 2004): however, due to its relatively low abundance in subducted material, subducted contributions are generally considered negligible (Holland and Ballentine, 2006).
View in Supplementary Information
However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Honda, M., Woodhead, J.D. (2005). A primordial solar-neon enriched component in the source of EM-I-type ocean island basalts from the Pitcairn Seamounts, Polynesia. Earth and Planetary Science Letters 236, 597-612.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Honda, M., McDougall, I., Patterson, D.B., Doulgeris, A., Clague, D.A. (1991) Possible solar noble-gas component in Hawaiian basalts. Nature 349, 149 – 151.
Show in context

Following the methods of Honda et al. (1991), we extrapolate measured 20Ne/22Ne values to “solar” (i.e. Ne-B = 12.5 ) values in order to estimate air-corrected (i.e. 21Ne/22NeEX) values.
View in Supplementary Information


Hopp, J., Trieloff, M. (2005) Refining the noble gas record of the Réunion mantle plume source: Implications on mantle geochemistry. Earth and Planetary Science Letters 240, 573–588.
Show in context

Notably, 21Ne/22NeEX values in Réunion xenoliths range from 0.0325 ± 0.0108 to 0.0513 ± 0.0120, and fall in the range previously measured in Réunion lavas (Hanyu et al., 2001; Trieloff et al., 2002; Hopp and Trieloff, 2005).
View in Supplementary Information
Figure S-2 [...] Three trend-lines are superimposed on the data: (1) the air-solar mixing line; (2) the Réunion line (Hopp and Trieloff, 2005); (3) the DMM (2∏D43) line (Moreira and Allègre, 1998).
View in Supplementary Information


Hopp, J., Ionov, D.A. (2011) Tracing partial melting and subduction-related metasomatism in the Kamchatkan mantle wedge using noble gas compositions. Earth and Planetary Science Letters 302, 121–131.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Jackson, C.R.M., Parman, S., Kelley, S.P., Cooper, R.F. (2013a) Noble gas transport into the mantle facilitated by high solubility in amphibole. Nature Geoscience 6, 562–565.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Jackson, C.R.M., Parman, S.W., Kelley, S.P., Cooper, R.F. (2013b) Constraints on light noble gas partitioning at the conditions of spinel-peridotite melting. Earth and Planetary Science Letters 384, 178–187.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Jackson, C.R.M., Parman, S.W., Kelley, S.P., Cooper, R.F. (2015) Light noble gas dissolution into ring structure-bearing materials and lattice influences on noble gas recycling. . Geochimica et Cosmochimica Acta 159, 1–15.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Jackson, C.R.M., Shuster, D.L., Parman, S.W., Smye, A.J. (2016). Noble gas diffusivity hindered by low energy sites in amphibole. Geochimica et Cosmochimica Acta 172, 65-75.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Jambon, A., Weber, H.W., Braun, O. (1986) Solubility of He, Ne, Ar, Kr and Xe in a basalt melt in the range 1250–1600°C - Geochemical implications. Geochimica et Cosmochimica Acta 50, 401–408.
Show in context

The measured He contents and calculated 40Ar∗, from Equation S-3, are then combined to give 4He/40Ar* ratios, which can be used to determine the extent of degassing, as He is more soluble than Ar (SHe/SAr = 9.5; Lux, 1987; Jambon et al., 1986) in basaltic magmas (Marty and Tolstikhin, 1998).
View in Supplementary Information
Using estimates of upper mantle K/U (Arevalo et al., 2009) and Th/U (O'Nions and McKenzie, 1993), a production ratio for 4He/40Ar* of ~2 (Jambon et al., 1986; Marty and Zimmermann, 1999) is assumed as a mantle starting value.
View in Supplementary Information


Javoy, M., Pineau, F., Delorme, H. (1986) Carbon and nitrogen isotopes in the mantle. Chemical Geology 57, 41–62.
Show in context

For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986; Cartigny et al., 1998; Marty and Zimmermann, 1999; Marty and Dauphas, 2003; Cartigny and Marty, 2013; Palot et al., 2012; Johnson and Goldblatt, 2015; this study).
View in Supplementary Information


Johnson, B., Goldblatt, C. (2015) The nitrogen budget of Earth. Earth-Science Reviews 148, 150-173.
Show in context

The large isotopic contrast between terrestrial nitrogen reservoirs makes N a potentially powerful tracer of volatile recycling between the surface and mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in Supplementary Information
For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986; Cartigny et al., 1998; Marty and Zimmermann, 1999; Marty and Dauphas, 2003; Cartigny and Marty, 2013; Palot et al., 2012; Johnson and Goldblatt, 2015; this study).
View in Supplementary Information
Table S-1
View in Supplementary Information


Kallenbach, R., Ipavich, F.M., Bochsler, P., Hefti, S., Hovestadt, D., Grünwaldt, H., Hilchenbach, M., Axford, W.I., Balsiger, H., Bürgi, A., Coplan, M.A., Galvin, A.B., Geiss, J., Gliem, F., Gloeckler, G., Hsieh, K.C., Klecker, B., Lee, M.A., Livi, S., Managadze, G.G., Marsch, E., Möbius, E., Neugebauer, M., Reiche, K.U., Scholer, M., Verigin, M.I., Wilken, B. Wurz, P. (1997) Isotopic composition of solar wind neon measured by CELIAS/MTOF on board SOHO. Journal of Geophysical Research – Space 102, 26895–26904.
Show in context

The primordial, “solar” neon component in Earth’s mantle has also been assumed to be best represented by present-day solar wind (i.e. 20Ne/22Ne = 13.8; Benkert et al., 1993; Kallenbach et al., 1997); note, however, that selection of this endmember as opposed to Ne-B will not significantly affect the conclusions in the following discussion.
View in Supplementary Information


Kendrick, M.A., Scambelluri, M., Honda, M., Phillips, D. (2011). High abundances of noble gas and chlorine delivered to the mantle by serpentinite subduction. Nature Geoscience 4, 807-812.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Kendrick, M.A., Honda, M., Pettke, T., Scambelluri, M., Phillips, D., Giuliani, A. (2013) Subduction zone fluxes of halogens and noble gases in seafloor and forearc serpentinites. Earth and Planetary Science Letters 365, 86–96.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Kusky, T.M., Li, J.H., Tucker, R.D. (2001) The Archean Dongwanzi ophiolite complex, North China Craton: 2.505-billion-year-old oceanic crust and mantle. Science 292, 1142-1145.
Show in context

Table S-1 [...] d Onset of subduction estimates of 1.1 Ga, 2.5 Ga and 3.9 Ga are based on Cartigny and Marty (2013), Kusky et al. (2001) and Condie and Pease (2008), respectively.
View in Supplementary Information


Lee, J.Y., Marti, K., Severinghaus, J.P., Kawamura, K., Yoo, H.S., Lee, J.B., Kim, J.S. (2006) A redetermination of the isotopic abundances of atmospheric Ar. Geochimica et Cosmochimica Acta 70, 4507-4512.
Show in context

In this way the measured 40Ar content in each sample can be corrected for the presence of atmospheric Ar to radiogenic Ar (i.e. 40Ar*): 40Ar* = [36ArM] × [(40Ar/36Ar)M - (40Ar/3636)A] where 36ArM and (40Ar/36Ar)M are the measured 36Ar abundance and argon isotope ratio, respectively, and (40Ar/36Ar)A is the air ratio (= 298.56 ± 0.31; Lee et al., 2006).
View in article


Li, Y., Keppler, H. (2014) Nitrogen speciation in mantle and crustal fluids. Geochimica et Cosmochimica Acta 129, 13-32.
Show in context

Nitrogen is stable as ammonium under subduction redox conditions (fO2 < QFM; Watenphul et al., 2010; Li et al., 2013; Li and Keppler, 2014; Mikhail and Sverjensky, 2014) and acts much like a large ion lithophile element, substituting for K-bearing minerals and sedimentary material.
View in Supplementary Information


Li, L., Bebout, G.E., Idleman, B.D. (2007) Nitrogen concentration and δ 15 N of altered oceanic crust obtained on ODP Legs 129 and 185: insights into alteration-related nitrogen enrichment and the nitrogen subduction budget. Geochimica et Cosmochimica Acta 71, 2344-2360.
Show in context

The large isotopic contrast between terrestrial nitrogen reservoirs makes N a potentially powerful tracer of volatile recycling between the surface and mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in Supplementary Information


Li, Y., Wiedenbeck, M., Shcheka, S., Keppler, H. (2013). Nitrogen solubility in upper mantle minerals. Earth and Planetary Science Letters 377, 311-323.
Show in context

Nitrogen is stable as ammonium under subduction redox conditions (fO2 < QFM; Watenphul et al., 2010; Li et al., 2013; Li and Keppler, 2014; Mikhail and Sverjensky, 2014) and acts much like a large ion lithophile element, substituting for K-bearing minerals and sedimentary material.
View in Supplementary Information


Lux, G. (1987) The behavior of noble gases in silicate liquids: solution, diffusion, bubbles and surface effects, with applications to natural samples. Geochimica et Cosmochimica Acta 51, 1549–1560.
Show in context

The measured He contents and calculated 40Ar∗, from Equation S-3, are then combined to give 4He/40Ar* ratios, which can be used to determine the extent of degassing, as He is more soluble than Ar (SHe/SAr = 9.5; Lux, 1987; Jambon et al., 1986) in basaltic magmas (Marty and Tolstikhin, 1998).
View in Supplementary Information


Mahoney, J.J., Natland, J.H., White, W.M., Poreda, R., Bloomer, S.H., Baxter, A.N. (1989) Isotopic and geochemical provinces of the Western Indian Ocean spreading centers. Journal of Geophysical Research 94, 4033–4052.
Show in context

The Réunion is currently located ~1100 km west of the CIR axis (Figure S-1; Mahoney et al., 1989; Murton et al., 2005).
View in Supplementary Information
Mahoney et al. (1989) later detected this Réunion-like isotope signature (low 143Nd/144Nd, high 87Sr/86Sr, high 207Pb/204Pb, and high 206Pb/204Pb) in basalts from the Marie Celeste FZ portion of the CIR.
View in Supplementary Information


Marty, B. (2012) The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth and Planetary Science Letters 313-314, 56-66.
Show in context

This isotopic contrast was likely established early in Earth’s history, reflecting the integrated effects of mantle degassing, late heavy bombardment, and/or hydrodynamic escape of Earth’s primary atmosphere (Marty, 2012).
View in Supplementary Information
Assuming solar wind samples accurately approximate the proto-solar nebula (PSN; Marty et al., 2011; Marty, 2012; Füri and Marty, 2015), there may also be a small, but detectable, primitive (i.e. solar) nitrogen component in the CIR mantle.
View in Supplementary Information
Table S-1 [...] c Initial mantle [N] endmember contents of 0.27 ppm and 36 ppm are from Marty (2012) and Cartigny and Marty (2013), respectively.
View in Supplementary Information


Marty, B., Humbert, F. (1997) Nitrogen and argon isotopes in oceanic basalts. Earth and Planetary Science Letters 152, 101-112.
Show in context

As a result, subduction of altered oceanic crust and sediments, containing molecularly-bound nitrogen – residual to sub-arc fluid loss – could result in the introduction of high N-isotope signatures into the (deep) mantle (Hofmann and White, 1982; Marty and Humbert, 1997; Halama et al., 2010).
View in Supplementary Information
Significantly, this observation is consistent with previous compilation studies of various mid ocean ridge (MOR) and OIB settings (e.g., Marty and Humbert, 1997; Fischer et al., 2005), which also suggest that nitrogen isotope variations are independent of degassing.
View in Supplementary Information


Marty, B., Tolstikhin, I.N. (1998) CO2 fluxes from mid-ocean ridges, arc and plumes. Chemical Geology 145, 233–248.
Show in context

For example, relative noble gas abundances (e.g., 4He/40Ar*) can provide useful information about the mode and extent of volatile loss (Marty and Tolstikhin, 1998).
View in Supplementary Information
The measured He contents and calculated 40Ar∗, from Equation S-3, are then combined to give 4He/40Ar* ratios, which can be used to determine the extent of degassing, as He is more soluble than Ar (SHe/SAr = 9.5; Lux, 1987; Jambon et al., 1986) in basaltic magmas (Marty and Tolstikhin, 1998).
View in Supplementary Information


Marty, B., Zimmermann, L. (1999) Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: assessment of shallow-level fractionation and characterization of source composition. Geochimica et Cosmochimica Acta 63, 3619–3633.
Show in context

For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986; Cartigny et al., 1998; Marty and Zimmermann, 1999; Marty and Dauphas, 2003; Cartigny and Marty, 2013; Palot et al., 2012; Johnson and Goldblatt, 2015; this study).
View in Supplementary Information
To date, however, such N-isotope anomalies have rarely been detected in the DMM (Marty and Zimmermann, 1999; Cartigny et al., 2001).
View in Supplementary Information
Using estimates of upper mantle K/U (Arevalo et al., 2009) and Th/U (O'Nions and McKenzie, 1993), a production ratio for 4He/40Ar* of ~2 (Jambon et al., 1986; Marty and Zimmermann, 1999) is assumed as a mantle starting value.
View in Supplementary Information


Marty, B., Dauphas, N. (2003) The nitrogen record of crust-mantle interaction and mantle convection from Archean to present. Earth and Planetary Science Letters 206, 397– 410.
Show in context

For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986; Cartigny et al., 1998; Marty and Zimmermann, 1999; Marty and Dauphas, 2003; Cartigny and Marty, 2013; Palot et al., 2012; Johnson and Goldblatt, 2015; this study).
View in Supplementary Information
N-isotope enrichments have been observed in subduction-related geothermal gases (Fischer et al., 2002) and ocean island basalts (OIBs; Hoffman and White, 1982; Mohapatra and Murty, 2002; Marty and Dauphas, 2003; Fischer et al., 2005).
View in Supplementary Information


Marty, B., Chaussidon, M., Jurewicz, A.,Wiens, R., Burnett, D.S. (2011) A 15N-poor isotopic composition for the solar system as shown by Genesis solar wind samples. Science 332, 1533–1536.
Show in context

Assuming solar wind samples accurately approximate the proto-solar nebula (PSN; Marty et al., 2011; Marty, 2012; Füri and Marty, 2015), there may also be a small, but detectable, primitive (i.e. solar) nitrogen component in the CIR mantle.
View in Supplementary Information
Table S-1[...] b Initial mantle δ15N endmember values of -40 ‰ and -383 ‰ are from Cartigny and Marty (2013) and Marty et al. (2011), respectively.
View in Supplementary Information


Matsumoto, T., Chen, Y., Matsuda, J.I. (2001) Concomitant occurrence of primordial and recycled noble gases in the Earth’s mantle. Earth and Planetary Science Letters 185, 35–47.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


McDougall, I., Upton, B.G.J., Wadsworth, W.J. (1965) A geological reconnaissance of Rodriguez Island Indian Ocean. Nature 206, 26–27.
Show in context

Rodrigues Island is located ~600 km east of Mauritius and formed at 1.5 Ma (McDougall et al., 1965); it marks the eastern extent of the Rodrigues Ridge, an east-west trending volcanic ridge that is between 8 to 10 Ma old (Duncan et al., 1990).
View in Supplementary Information


Mikhail, S., Sverjensky, D.A. (2014) Nitrogen speciation in upper mantle fluids and the origin of Earth's nitrogen-rich atmosphere. Nature Geoscience 7, 816-819.
Show in context

Nitrogen is stable as ammonium under subduction redox conditions (fO2 < QFM; Watenphul et al., 2010; Li et al., 2013; Li and Keppler, 2014; Mikhail and Sverjensky, 2014) and acts much like a large ion lithophile element, substituting for K-bearing minerals and sedimentary material.
View in Supplementary Information


Mitchell, E.C., Fischer, T.P., Hilton, D.R., Hauri, E.H., Shaw, A.M., de Moor, J.M., Sharp, Z.D., Kazahaya, K. (2010) Nitrogen sources and recycling at subduction zones: Insights from the Izu‐Bonin‐Mariana arc. Geochemistry, Geophysics, Geosystems 11, 2.
Show in context

The large isotopic contrast between terrestrial nitrogen reservoirs makes N a potentially powerful tracer of volatile recycling between the surface and mantle (Li et al., 2007; Mitchell et al., 2010; Johnson and Goldblatt, 2015; Bebout et al., 2016).
View in Supplementary Information


Mohapatra, R.K., Murty, S.V.S. (2002) Nitrogen and noble gas isotopes in mafic and ultramafic inclusions in the alkali basalts from Kutch and Reunion—implications for their mantle sources. Journal of Asian Earth Sciences 20, 867-877.
Show in context

N-isotope enrichments have been observed in subduction-related geothermal gases (Fischer et al., 2002) and ocean island basalts (OIBs; Hoffman and White, 1982; Mohapatra and Murty, 2002; Marty and Dauphas, 2003; Fischer et al., 2005).
View in Supplementary Information


Moreira, M. (2013) Noble gas constraints on the origin and evolution of Earth’s volatiles. Geochemical Perspectives 2, 229-230.
Show in context

The Ne–B endmember is obtained from gas-rich meteorites and lunar soils with a 20Ne/22Ne ratio of 12.52 ± 0.18, considered to represent implanted solar neon, now present within Earth’s mantle (Black, 1972; Trieloff et al., 2000; Trieloff et al., 2002; Ballentine et al., 2005; Moreira, 2013).
View in Supplementary Information


Moreira, M., Allègre, C.J. (1998) Helium-neon systematics and the structure of the mantle. Chemical Geology 147, 53–59.
Show in context

Figure S-2 [...] Three trend-lines are superimposed on the data: (1) the air-solar mixing line; (2) the Réunion line (Hopp and Trieloff, 2005); (3) the DMM (2∏D43) line (Moreira and Allègre, 1998).
View in Supplementary Information


Moreira, M., Breddam, K., Curtice, J., Kurz, M.D. (2001) Solar neon in the Icelandic mantle: new evidence for an undegassed lower mantle. Earth and Planetary Science Letters 185, 15–23.
Show in context

For example, Moreira et al. (2001) proposed that different mantle reservoirs (i.e. DMM versus PRIM) could have evolved from the same He and Ne isotope compositions but with distinct 3He/22Ne ratios and that subsequent ingrowth of nucleogenic 21Ne and radiogenic 4He have produced the observed range in 21Ne/22Neex and 3He/4He values.
View in Supplementary Information


Morgan, W.J. (1978) Rodriguez, Darwin, Amsterdam, a second type of hot spot island. Journal of Geophysical Research 83, 5355–5360.
Show in context

First, Morgan (1978) proposed that the Rodrigues Ridge was formed by volcanism above a channel of upwelling Réunion hot spot mantle, which was deflected toward the CIR as it migrated northeastward away from the hot spot, and predicted a Réunion signature would be present at the intersection of the CIR with a line projected through Réunion and Rodrigues islands.
View in Supplementary Information
As a result, these authors proposed that Réunion hot spot material flows eastward toward the CIR on a trajectory that impinges the ridge at ~19.9 °S, in agreement with the initial hypothesis put forth by Morgan (1978).
View in Supplementary Information


Morgan, W.J. (1981) Hot spot tracks and the opening of the Atlantic and Indian oceans. In The Sea 7, 443–487.
Show in context

The island of Mauritius is located 250 km east of Réunion Island and 850 km west of the CIR, at the previous site of the Réunion hot spot (~8 Ma) (Morgan, 1981; Paul et al., 2005).
View in Supplementary Information


Mukhopadhyay, S. (2012) Early differentiation and volatile accretion recorded in deep-mantle neon and xenon. Nature 486, 101–104.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Murton, B.J., Tindle, A.G., Milton, J.A., Sauter, D. (2005) Heterogeneity in southern Central Indian Ridge MORB: implications for ridge– hot spot interaction. Geochemistry Geophysics Geosystems 6, Q03E20.
Show in context

The Réunion is currently located ~1100 km west of the CIR axis (Figure S-1; Mahoney et al., 1989; Murton et al., 2005).
View in Supplementary Information
Murton et al. (2005) conducted a detailed study of this area and reported enrichments in incompatible elements that increased northward, which they interpreted to reflect the presence of enriched mantle originating at the Réunion hot spot, which had subsequently migrated eastward toward the CIR against the direction of motion of the lithosphere.
View in Supplementary Information


Nagao, K., Takahashi, E. (1993) Noble gases in the mantle wedge and lower crust: an inference from the isotopic analyses of xenoliths from Oki-Dogo and Ichinomegata, Japan. Geochemical Journal 27, 229–240.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Nauret, F., Abouchami, W., Galer, S.J.G., Hofmann, A.W., Hémond, C., Chauvel, C., Dyment, J. (2006) Correlated trace element - Pb isotope enrichments in Indian MORB along 18–20°S, Central Indian Ridge. Earth and Planetary Science Letters 245, 137–152.
Show in context

Nauret et al. (2006) also targeted a suite of basalts collected both on and off the CIR axis between 18 °S and 20 °S for Sr-Nd-Pb isotopes and showed that the most trace element enriched samples display the most radiogenic Pb isotopic compositions.
View in Supplementary Information


Niedermann, S., Graf, T., Marti, K. (1993) Mass spectrometric identification of cosmic-ray-produced neon in terrestrial rocks with multiple neon components. Earth and Planetary Science Letters 118, 65–73.
Show in context

Neon results were corrected for procedural blanks and inferences of doubly-charged 40Ar and CO2 with 20Ne and 22Ne, respectively (see Niedermann et al. (1993) for details).
View in Supplementary Information


Okino, K., Ichikawa, Y., Tamaki, T. (2008) Detailed morphology of the Central Indian Ridge between 20°15S and 15°30S: Implication for hot spot ridge interaction. Proceedings of the Japan Geoscience Union (JPGU), Chiba, Japan, Abstract J164 002.
Show in context

Figure S-1 [...] Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999), R/V L'Atalante (Dyment et al., 2000), and R/V Hakuho-Maru (Okino et al., 2008); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008; Füri et al., 2011).
View in Supplementary Information


O’Nions, R.K., McKenzie, D. (1993) Estimates of mantle thorium/uranium ratios from Th, U and Pb isotope abundances in basaltic melts. Philosophical Transactions of the Royal Society 342, 65–77.
Show in context

Using estimates of upper mantle K/U (Arevalo et al., 2009) and Th/U (O'Nions and McKenzie, 1993), a production ratio for 4He/40Ar* of ~2 (Jambon et al., 1986; Marty and Zimmermann, 1999) is assumed as a mantle starting value.
View in Supplementary Information


Palot, M., Cartigny, P., Harris, J.W., Kaminsky, F.V., Stachel, T. (2012) Evidence for deep mantle convection and primordial heterogeneity from nitrogen and carbon stable isotopes in diamond. Earth and Planetary Science Letters 357, 179-193.
Show in context

For example, the depleted MORB mantle (DMM) has a lower 15N/14N ratio (by -5 ± 2 ‰) than Earth’s atmosphere (Javoy et al., 1986; Cartigny et al., 1998; Marty and Zimmermann, 1999; Marty and Dauphas, 2003; Cartigny and Marty, 2013; Palot et al., 2012; Johnson and Goldblatt, 2015; this study).
View in Supplementary Information


Parai, R., Mukhopadhyay, S. (2015) The evolution of MORB and plume mantle volatile budgets: constraints from fission Xe isotopes in Southwest Indian Ridge basalts. Geochemistry Geophysics Geosystems 16, 719–735.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Parai, R., Mukhopadhyay, S., Standish, J. (2012) Heterogeneous upper mantle Ne, Ar and Xe isotopic compositions and a possible Dupal noble gas signature recorded in basalts from the Southwest Indian Ridge. Earth and Planetary Science Letters 359, 227–239.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Parson, L.M., Patriat, P., Searle, R.C., Briais, A.R. (1993). Segmentation of the Central Indian Ridge between 12 12′ S and the Indian Ocean triple junction. Marine geophysical researches 15, 265-282.
Show in context

In addition, several off-axis samples were collected along a perpendicularly offset lineament, which extends westward from 66 °E to 64 °E toward the Mascarene Islands (e.g., Réunion, Mauritius, Rodrigues islands; Parson et al., 1993).
View in Supplementary Information
In addition, several off-axis samples were collected along a perpendicularly offset lineament which extends westward from 66 °E to 64 °E toward the Mascarene Islands (e.g., Réunion, Mauritius, Rodrigues islands; Parson et al., 1993).
View in Supplementary Information


Paul, D., White, W.M., Blichert-Toft, J. (2005) Geochemistry of Mauritius and the origin of rejuvenescent volcanism on oceanic island volcanoes. Geochemistry Geophysics Geosystems 6, Q06007.
Show in context

The island of Mauritius is located 250 km east of Réunion Island and 850 km west of the CIR, at the previous site of the Réunion hot spot (~8 Ma) (Morgan, 1981; Paul et al., 2005).
View in Supplementary Information


Pető, M.K., Mukhopadhyay, S., Kelley, K.A. (2013). Heterogeneities from the first 100 million years recorded in deep mantle noble gases from the Northern Lau Back-arc Basin. Earth and Planetary Science Letters 369, 13-23.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Sano, Y., Takahata, N., Nishio, Y., Marty, B. (1998) Nitrogen recycling in subduction zones. Geophysical Research Letters 25, 2289-2292.
Show in context

Subsequent modifications to mantle N may result from subduction of this material comprised of isotopically distinct N (~+5‰) (Sano et al., 1998; Busigny et al., 2013; Thomazo and Papineau, 2013; Halama et al., 2014).
View in Supplementary Information


Sarda, P. (2004) Surface noble gas recycling to the terrestrial mantle. Earth and Planetary Science Letters 228, 49-63.
Show in context

The fate of Ne is debated (Sarda, 2004): however, due to its relatively low abundance in subducted material, subducted contributions are generally considered negligible (Holland and Ballentine, 2006).
View in Supplementary Information
However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Sarda, P., Staudacher, T., Allègre, C.J. (1988) Neon isotopes in submarine basalts. Earth and Planetary Science Letters 91, 73-88.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information
The 21Ne/22Ne of PRIM (i.e. primitive solar component) is again assumed to be 0.03118 (Trieloff and Kunz, 2005) and the DMM 21Ne/22Ne endmember value is estimated to be 0.0594 by extrapolating the DMM trajectory (Sarda et al, 1988) to the 20Ne/22Ne of Ne-B (= 12.5).
View in Supplementary Information


Smith, W.H.F., Sandwell, D.T. (1997) Global sea floor topography from satellite altimetry and ship depth soundings. Science 277, 1956-1962.
Show in context

Figure S-1 [...] Three shades of colours represent bathymetry data: (1) pale colours are bathymetry “predicted” from satellite altimetry (Smith and Sandwell, 1997); (2) intermediate colours represent previous multi-beam bathymetric data from R/V Marion Dufresne (Dyment et al., 1999), R/V L'Atalante (Dyment et al., 2000), and R/V Hakuho-Maru (Okino et al., 2008); (3) bright colours are multibeam bathymetric data collected by R/V Revelle in 2007 (Füri et al., 2008; Füri et al., 2011).
View in Supplementary Information


Staudacher, T., Allègre, C.J. (1988) Recycling of oceanic crust and sediments: the noble gas subduction barrier. Earth and Planetary Science Letters 89, 173-183.
Show in context

Heavier noble gas compositions in the mantle were long considered to be independent of subduction as they were believed to be returned back into the atmosphere through subduction volcanism before they could be admixed into the mantle (Staudacher and Allègre, 1988; Hilton et al., 2002).
View in Supplementary Information


Sumino, H., Burgess, R., Mizukami, T., Wallis, S.R., Holland, G., Ballentine, C.J. (2010) Seawater-derived noble gases and halogens preserved in exhumed mantle wedge peridotite. Earth and Planetary Science Letters 294, 163-172.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Thomazo, C., Papineau, D. (2013) Biogeochemical cycling of nitrogen on the early Earth. Elements 9, 345-351.
Show in context

Subsequent modifications to mantle N may result from subduction of this material comprised of isotopically distinct N (~+5‰) (Sano et al., 1998; Busigny et al., 2013; Thomazo and Papineau, 2013; Halama et al., 2014).
View in Supplementary Information


Trieloff, M., Kunz, J. (2005) Isotope systematics of noble gases in the Earth's mantle: possible sources of primordial isotopes and implications for mantle structure. Physics of the Earth and Planetary Interiors 148, 13-38.
Show in context

The 21Ne/22Ne of PRIM (i.e. primitive solar component) is again assumed to be 0.03118 (Trieloff and Kunz, 2005) and the DMM 21Ne/22Ne endmember value is estimated to be 0.0594 by extrapolating the DMM trajectory (Sarda et al, 1988) to the 20Ne/22Ne of Ne-B (= 12.5).
View in article


Trieloff, M., Kunz, J., Clague, D.A., Harrison, D., Allègre, C.J. (2000) The nature of pristine noble gases in mantle plumes. Science 288, 1036–1038.
Show in context

The following equations are used: 21Ne/22NeEX = (21Ne/22Ne M - 21Ne/22NeA)/f22 + 21Ne/22NeA where f22 is the proportion of mantle-derived Ne in a sample: f22 = (20Ne/22NeM - 20Ne/22NeA)/(20Ne/22NeS - 20Ne/22NeA) and 20Ne/22NeA and 21Ne/22NeA are the isotopic ratios of air, assumed to be 9.8 and 0.029, respectively, whereas 21Ne/22NeM is the measured sample 21Ne/22Ne value and 20Ne/22NeS is the assumed 21Ne/22Ne “solar” (i.e. Ne-B) neon component (=12.5; Trieloff et al., 2000).
View in Supplementary Information
The Ne–B endmember is obtained from gas-rich meteorites and lunar soils with a 20Ne/22Ne ratio of 12.52 ± 0.18, considered to represent implanted solar neon, now present within Earth’s mantle (Black, 1972; Trieloff et al., 2000, 2002; Ballentine et al., 2005; Moreira, 2013).
View in Supplementary Information


Trieloff, M., Kunz, J., Allègre C.J. (2002) Noble gas systematics of the Réunion mantle plume source and the origin of primordial noble gases in Earth’s mantle. Earth and Planetary Science Letters 200, 297–313.
Show in context

Notably, 21Ne/22NeEX values in Réunion xenoliths range from 0.0325 ± 0.0108 to 0.0513 ± 0.0120, and fall in the range previously measured in Réunion lavas (Hanyu et al., 2001; Trieloff et al., 2002; Hopp and Trieloff, 2005).
View in Supplementary Information
The Ne–B endmember is obtained from gas-rich meteorites and lunar soils with a 20Ne/22Ne ratio of 12.52 ± 0.18, considered to represent implanted solar neon, now present within Earth’s mantle (Black, 1972; Trieloff et al., 2000, 2002; Ballentine et al., 2005; Moreira, 2013).
View in Supplementary Information


Tucker, J.M., Mukhopadhyay, S., Schilling, J.G. (2012) The heavy noble gas composition of the depleted MORB mantle (DMM) and its implications for the preservation of heterogeneities in the mantle. Earth and Planetary Science Letters 355, 244-254.
Show in context

However, recent studies suggest that air-like values in the deep mantle may be the direct result of subduction (Nagao and Takahashi, 1993; Sarda et al., 1988; Matsumoto et al., 2001; Sarda, 2004; Holland and Ballentine, 2006; Sumino et al., 2010; Hopp and Ionov, 2011; Kendrick et al., 2011; Mukhopadhyay, 2012; Parai et al., 2012; Tucker et al., 2012; Kendrick et al., 2013; Jackson et al., 2013a;b; Pető et al., 2013; Jackson et al., 2015; Parai and Mukhopadhyay, 2015; Jackson et al., 2016), thus potentially providing an additional mechanism for monitoring recycling (Honda and Woodhead, 2005; Sumino et al., 2010).
View in Supplementary Information


Ulrich, M., Hemond, C., Nonnotte, P., Jochum, K.P. (2012) OIB/seamount recycling as a possible process for E-MORB genesis. Geochemistry Geophysics Geosystems 13, Q0AC19.
Show in context

Recently, however, Ulrich et al. (2012) showed that trace element and isotopic enrichments in the 18 °S to 20 °S region of the CIR are consistent with a binary mixture between the regional depleted MORB mantle (DMM) source and a recycled Ocean Iceland Basalt (OIB)/plume component.
View in Supplementary Information


Watenphul, A., Wunder, B., Wirth, R., Heinrich, W. (2010) Ammonium-bearing clinopyroxene: a potential nitrogen reservoir in the Earth's mantle. Chemical Geology 270, 240-248.
Show in context

Nitrogen is stable as ammonium under subduction redox conditions (fO2 < QFM; Watenphul et al., 2010; Li et al., 2013; Li and Keppler, 2014; Mikhail and Sverjensky, 2014) and acts much like a large ion lithophile element, substituting for K-bearing minerals and sedimentary material.
View in Supplementary Information

top