Helium isotope, C/3He, and Ba‐Nb‐Ti signatures in the ...... · RESEARCH ARTICLE 10.1002/2014GC005625 Helium isotope, C/3He, and Ba-Nb-Ti signatures in the northern Lau Basin:

RESEARCH ARTICLE10.1002/2014GC005625

Helium isotope, C/3He, and Ba-Nb-Ti signatures in the northernLau Basin: Distinguishing arc, back-arc, and hotspot affinitiesJohn Lupton1, Ken H. Rubin2, Richard Arculus3, Marvin Lilley4, David Butterfield5, Joseph Resing5,Edward Baker5,6, and Robert Embley1

1NOAA Pacific Marine Environmental Laboratory, Newport, Oregon, USA, 2Department of Geology and Geophysics, SOEST,University of Hawaii, Honolulu, Hawaii, USA, 3Research School of Earth Sciences, Australian National University, Canberra,ACT, Australia, 4School of Oceanography, University of Washington, Seattle, Washington, USA, 5NOAA Pacific MarineEnvironmental Laboratory, Seattle, Washington, USA, 6Now at Joint Institute for the Study of Atmosphere and Ocean,University of Washington, Seattle, Washington, USA

Abstract The northern Lau Basin hosts a complicated pattern of volcanism, including Tofua Arc volca-noes, several back-arc spreading centers, and individual ‘‘rear-arc’’ volcanoes not associated with thesestructures. Elevated 3He/4He ratios in lavas of the NW Lau Spreading Center suggest the influence of a man-tle plume, possibly from Samoa. We show that lavas from mid-ocean ridges, volcanic arcs, and hotspotsoccupy distinct, nonoverlapping fields in a 3He/4He versus C/3He plot. Applied to the northern Lau Basin,this approach shows that most of Lau back-arc spreading systems have mid-ocean ridge 3He/4He-C/3Hecharacteristics, except the NW Lau spreading center, which has 3He/4He-C/3He similar to ‘‘high 3He’’ hot-spots such as Loihi, Kilauea, and Yellowstone, but with slightly lower C/3He. Niua seamount, on the northernextension of the Tofua Arc, falls squarely in the arc field. All the NE Lau rear-arc volcanoes, including therecently erupting West Mata, also have arc-like 3He/4He-C/3He characteristics. Ba-Nb-Ti contents of the lavas,which are more traditional trace element indicators of mantle source enrichment, depletion, and subduc-tion input, likewise indicate arc and hot spot influences in the lavas of the northern Lau Basin, but in a moreambiguous fashion because of a complex prior history. This verifies that 3He/4He-C/3He systematics are use-ful for differentiating between mid-ocean ridge, arc, and hotspot affinities in submarine volcanic systems,that all three of these affinities are expressed in the northern Lau Basin, and provides additional support forthe Samoan plume influence in the region.

1. Introduction

The northern Lau Basin has several active back-arc spreading centers driven by the rapid subduction of thePacific Plate along the Tofua Arc [Hawkins, 1995; Bevis et al, 1995; Pelletier et al., 1998, 2001; Zellmer and Tay-lor, 2001]. In the northeastern part of the basin, the Fonualei Rift and Spreading Center (FRSC), the Manga-tolu Triple Junction (MTJ), and NE Lau Spreading Center (NELSC) form a long back-arc extensional zoneseparated by lateral offsets (Figure 1). Farther west, the Northwest Lau Spreading Center (NWLSC) andRochambeau Rifts (RR) form another major back-arc spreading system which terminates in the south againsta long transform fault called the Peggy Ridge (PR). While the NWLSC and NELSC have a simple linear ridge-crest-type geometry, the Rochambeau Rifts consist of a complex series of rifts, pull-apart basins, and vol-canic centers with no clear zone of spreading. At its southeastern end, the Peggy Ridge breaks into a seriesof short rift-segments known as the Lau Extensional Transform Zone (LETZ) [Martinez and Taylor, 2006]. Theoverall opening rate of the northern Lau Basin has been estimated to be �160 mm/yr [Pelletier et al., 1998],making it the fastest opening back-arc basin. The spreading rates for the RR, the NWLSC, and the NELSChave been estimated to be 110, 75, and 120 mm/yr, respectively [Bird, 2003]. Thus, the northeastern andnorthwestern Lau spreading centers each have approximately equal spreading rates of �100 mm/yr.

The NWLSC-RR extensional zone has attracted considerable attention due to the elevated 3He/4He signatureof the erupted lavas in this region. While mid-ocean ridge basalts have fairly uniform 3He/4He ratios of about7–9 Ra (where R 5 3He/4He and Ra 5 Rair 5 1.39 3 1026) [Graham, 2002], the Rochambeau Rifts have muchhigher 3He/4He similar to that found at oceanic hotspots. Poreda and Craig [1992] were the first to reportelevated 3He/4He values in samples from the Rochambeau Rifts ranging from 11 up to 22 Ra, much higher

Special Section:Assessing Magmatic,Neovolcanic, Hydrothermal,and Biological Processes alongIntra-Oceanic Arcs andBack-Arcs

Key Points:� Establishes C-He systematics as valid

tool for volcanic systems� C-He technique is applied to the

northern Lau Basin� Arc, back-arc, and hotspot affinities

are all present in the N Lau Basin

Correspondence to:J. Lupton,[email protected]

Citation:Lupton, J., K. H. Rubin, R. Arculus, M.Lilley, D. Butterfield, J. Resing, E. Baker,and R. Embley (2015), Helium isotope,C/3He, and Ba-Nb-Ti signatures in thenorthern Lau Basin: Distinguishing arc,back-arc, and hotspot affinities,Geochem. Geophys. Geosyst., 16, 1133–1155, doi:10.1002/2014GC005625.

Received 22 OCT 2014

Accepted 11 MAR 2015

Accepted article online 21 MAR 2015

Published online 28 APR 2015

VC 2015. American Geophysical Union.

All Rights Reserved.

LUPTON ET AL. HELIUM AND CARBON IN NORTHERN LAU BASIN 1133

Geochemistry, Geophysics, Geosystems

PUBLICATIONS

http://dx.doi.org/10.1002/2014GC005625

http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1525-2027/specialsection/OCEANARC1/





http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1525-2027/

http://publications.agu.org/

than the values typical of mid-ocean ridge (MOR) or back-arc basin basalts. More recently, Lupton et al.[2009] and Hahm et al. [2012] reported helium isotope ratios from the NWLSC and RR ranging from 11 to 28Ra, all clearly higher than typical depleted mantle or MOR-type values. These elevated helium isotope ratiosare not confined to the Rochambeau Rifts, but extend southward along the NWLSC down to the PeggyRidge [Lupton et al., 2009]. In addition to the elevated 3He/4He, Lupton et al. [2012] and Hahm et al. [2012]have detected a hotspot-type neon isotopic signature trapped in the NWLSC-RR basalts, distinct from theneon found in MOR basalts. Peto et al. [2013] measured the complete spectrum of noble gases on four sam-ples from the Rochameau Rifts and found mantle plume-type neon, argon, and xenon isotopic signatures.These findings all point to the presence of a hotspot or ocean island basalt (OIB) component along theNWLSC-RR spreading center. It has been postulated that this is due to the intrusion of the Samoan hotspotcomponent into the northern Lau Basin through a tear in the downgoing Pacific Plate [Natland, 1980; Turner

172°W174°W176°W178°W

14°S

16°S

18°S

20°S

22°S

0 Km 100

This workHahm et al.

TM proj. CM 175°W

spreadingcenters

0 m

-3700 m

Savai’i

NWLSCy Ridg

Pegg

e

RR

MTJ NEL

CSC

FRS

CLS

C

LETZ

ESL

C

Tofu

a

Vol

cani

c

Arc

Niuafo’ou

Tongatapu

Vava’u

Tofua

Lakeba

Vanua Balavu

and deeper

Depth scale

7.7

7.3

7.2

9.0 8.67.3

7.3

10.9

12.114.1

22.618.2

17.8

18.5

12.717.614.2

28.111.2

10.4

7.9

7.6

7.77.7

7.7

8.6

8.5

9.5

9.6 9.38.9

8.9

9.1

8.1

8.1

8.58.58.59.5

Figure 1. Overview map of the Lau Basin showing locations of the samples used in this study. White circles are this work, red-filled circles aresamples from Hahm et al. [2012]. Solid lines denote spreading centers and plate boundaries. Numbers next to the sample locations arehelium isotope ratios expressed as R/Ra, where R 5 3He/4He and Ra 5 Rair 5 1.4 3 1026. Abbreviations are as follows: Northwest Lau Spread-ing Center (NWLSC), Rochambeau Rifts (RR), Mangatolu Triple Junction (MTJ), Northeast Lau Spreading Center (NELSC), Fonualei Rift andSpreading Center (FRSC), Lau Extensional Transform Zone (LETZ), Central Lau Spreading Center (CLSC), and Eastern Lau Spreading Center(ELSC). Dashed box indicates area covered by Figure 2. The bold-dashed line denotes extent of influence of the OIB or hotspot componentinto the northern Lau Basin, based on helium isotope ratios. Data are listed in Table. 2.

Geochemistry, Geophysics, Geosystems 10.1002/2014GC005625


and Hawkesworth, 1998; Lupton et al., 2009]. Certainly Samoa is a ‘‘high-3He’’ hotspot, since helium isotoperatios from 12 up to 33.8 Ra have been reported for the Samoan Islands [Farley et al., 1992; Jackson et al.,2007]. In contrast to the NWLSC-RR extensional zone, the spreading centers of the NE Lau Basin (the NELSCand MTJ) have typical mid-ocean-ridge-type helium isotope ratios [Poreda and Craig,1992; Honda et al.,1993; Hilton et al.,1993; Lupton et al., 2009]. Thus, the elevated 3He/4He ratios appear to be confined to theNWLSC-RR system as shown in Figure 1.

The map of helium isotope ratios indicates that the intrusion of a 3He-rich OIB or hotspot type signatureinto the northern Lau Basin is confined to the Rochambeau Rifts and NWLSC, with all the samples east andsouth of Niuafou’ou having MORB-like 3He/4He ratios or lower (Figure 1). However, it has been argued onthe basis of trace element ratios and radiogenic isotope signatures that an OIB-type component, probablyfrom Samoa, is present in other parts of the northern Lau Basin, but not necessarily everywhere. Regelouset al. [2008] used the fact that Nb/Zr ratios decrease systematically with distance from Savaii to argue thatSamoan hotspot influence extends for 400 km into the northern Lau Basin down to the Fonualei Rift. Wendtet al. [1997] made a similar assertion based on a combination of Zr/Nb and Pb/Ce ratios in the northeastLau Basin, and Pearce and Stern [2006] used Nb/Yb ratios to trace mantle flow into the northern Lau Basin.

The use of radiogenic isotopes to trace ingress of the Samoan plume [e.g., Tian et al., 2011] is complicatedby the fact that Samoan lavas contain a mixture of mantle sources, including FOZO-A, EM1, EM2, HIMU, anddepleted MORB mantle DMM [Jackson et al., 2014; Nebel and Arculus, 2015]. For example, Sr, Nd, and Pb iso-topes in Niuafou’ou Island lavas appear to be derived in part from a component similar to Uo Mamae(Machias) Seamount lavas, somewhat distinct from the main Samoan Islands [Falloon et al., 2007; Regelouset al., 2008]. In general Lau Basin lavas plot, more or less along a mixing line between the CLSC and Samoain the traditional Sr, Nd, and Pb isotope diagrams [Regelous et al., 2008]. Because samples from Niuafou’ou,Mangatolu Triple Junction (MTJ), and the NELSC fall intermediate between Samoa and the CLSC on radio-genic isotope plots, it has been argued that these sites all have a Samoan component, and that the bound-ary of influence of the Samoan hotspot should be moved to the east of MTJ and the NELSC (see Figure 1)[Regelous et al, 2008; Caulfield et al., 2012]. Samoan lavas range from FOZO-like eNd and eHf to significantlylower values: both eNd and eHf decrease systematically from north to south in Lau Basin lavas. Pearce et al.[2007] used eNd to trace the penetration of enriched mantle into the northern Lau, showing that northeastLau has lower eNd than the CLSC and ELSC farther south. Hf is postulated to be less mobile than Nd andthus should be even better at tracing mantle flow. Caulfield et al. [2012] mapped eHf in the Lau Basin show-ing again that the MTJ and NELSC have lower values (eHf 5 12–13) compared to the Fonualei Rift and theCLSC (eHf 5 15–16) farther south.

NE Lau Region

Mata Taha2160m

Mata Ua2105m

Mata Fa2315m

Mata Nima2670m

Mata Ono2370m

Mata Fitu2310m

Mata Tolu1820m

WestMata1165m

EastMata1265m

Mata Volcanoes

South

(Niuatahi)

NiuaNorth

Figure 2. Nested maps showing detail of the NE Lau Basin and of the Mata volcanoes. Filled circles indicate locations where vent fluid samples were collected for this study.



The trace element and radiogenic isotope data, taken as a whole, are persuasive evidence for at least aFOZO-type component, plausibly from Samoa, in the northeast Lau that is not confined to the Rochambeauand NWLSC farther west. This component could also be derived from other hot spot tracks that have previ-ously traversed the area [Jackson et al., 2010]. However, the fact remains that the elevated helium isotoperatios that are a clear marker of hotspot influence are confined to the RR-NWLSC, with all of the 3He/4Heratios in lavas east of Niuafou’ou having MORB-like helium (R/Ra� 8). It has been proposed that plume man-tle flowing into the northern Lau Basin has been previously degassed or ‘‘pre-conditioned’’ by meltingbeneath the Samoan plume while still preserving its Nd-Hf signature, thus explaining the lack of a enrichedhelium signature [Pearce et al., 2007]. The discrepancy between the helium and trace/metal/radiogenic iso-tope data can then only be explained by a somewhat unlikely model in which two very different types ofmantle intrusion into the northern Lau have occurred, with the mantle ingress into the northeastern Lauhaving suffered from degassing and volatile depletion, while the incursion into the RR-NWLSC has not.

Both the Fonualei Rift and the CLSC lavas are derived from strongly depleted mantle sources. The CLSClavas plot at the most depleted end of the spectrum of global MORB with regard to light rare earth deple-tion [Jenner and O’Neill, 2012; Nebel et al., 2013]. Additions of enriched components, such as those fromplume-derived or subducted slab-derived sources, would be expected to be detectable at low mass addi-tions to either of the source regions of the CLSC and Fonualei Rift. Whereas the CLSC has no sign of a sub-duction zone component [Jenner et al., 2012], the Fonualei Rift has a strong subduction overprint [Kelleret al., 2008]. It therefore seems uncertain whether the lower eNd and eHf of the Fonualei Rift compared withthe CLSC is the result of the subduction overprint rather than selectively degassed Samoan plume ingress.

Nebel and Arculus [2015] have recently combined new Hf-Nd measurements with previously publishedhelium data [Lupton et al., 2009], showing that samples from the RR-NWLSC can be explained as a mixturebetween DMM (similar to CLSC) and only a single FOZO Samoan component, with no evidence for the EM2,HIMU, and EM1 components that have distinctively lower eNd and eHf than found in the Lau Basin. Thereare some complications, however. While Samoan lavas have low Cu contents, Jenner et al. [2012] foundhigh Cu contents in the NWLSC, but not the Rochambeau Rifts. This suggests that the RR-NWLSC samplesmay be the result of three component mixing between DMM, a high-Cu mantle source beneath the NWLSC,and a low-Cu Samoan plume component.

The NE Lau Basin is perhaps the most complex part of the northern Lau Basin (Figures 1 and 2). It isbounded on the east by the Tofua Arc, and on the west by the NELSC, a typical back-arc system. The regionbetween these is populated by a series of what we are refer to as ‘‘rear-arc’’ volcanic centers [Rubin et al.,2013], including Niuatahi (formerly called Volcano O), the Mata volcano group, and areas of recent volca-nism in between these (Figure 2). These volcanoes provide a critical sampling opportunity westward fromthe volcanic front of the active Tongan Arc toward the NELSC and MTJ, and closest possible within the LauBasin to the Samoan plume ingress. Recent submarine eruptions have occurred and volcanic productssampled at both West Mata volcano and along the southern NELSC [Resing et al., 2011; Baker et al., 2011;Rubin et al., 2012]. Most of these volcanic centers are host to active hydrothermal systems [German et al.,2006; Kim et al., 2009; TN-234, 2009; RR1211, 2012]. Niua submarine volcano (so-named by Nautilus Mineralsand formerly referred to as Volcano ‘‘P’’) falls on the northern extension of the Tofua Arc, and for this reasonmight be expected to have subduction zone or arc-type affinities. The NELSC, on the other hand, might beexpected to have a typical back-arc or MOR signature. Although the arc versus back-arc provenance of therear-arc volcanoes remains in question, these volcanoes have been proposed to have an arc-provenancebecause of their high apparent magma flux through the Holocene, the relatively magma-starved nature ofthe adjacent segment of the Tofua arc, and the arc magma trace element affinities of lavas erupted there[Rubin et al., 2013]. On the other hand, as the cartoon in Figure 3 shows, due to the mixing regime in themantle, these rear-arc volcanoes might be expected to have a mixture of inputs from the arc proper andfrom the upper mantle or asthenosphere.

In this paper, we address the question of the provenance of the submarine volcanoes of the northern LauBasin using helium isotopes in combination with C/3He ratios, as well as Nb-Ba-Ti systematics of volcanicrocks. This paper expands upon the original work by Marty and Jambon [1987]. By examining publishedbaseline measurements, we first establish that mid-ocean ridges, arc volcanoes, and hotspots occupy dis-tinct, nonoverlapping fields in a plot of 3He/4He versus C/3He. We then use this approach to investigate theprovenance of various volcanic centers and back-arc spreading centers of the northern Lau Basin, including



the rear-arc volcanoes of the NE LauBasin. Using He-C measurements ofvolcanic glass, vent fluid, and water-column samples and Ba, Nb, and Tiin volcanic rocks, we found thatMOR, arc, and hotspot affinities areall present in the northern LauBasin.

2. Sampling and AnalyticalMethods

A variety of samples were used inthis study, including volcanic rocks,seafloor hydrothermal fluids, andsamples of water-column hydro-thermal plumes. Samples were col-lected on several seagoingexpeditions: voyages SS11/2004 and

SS07/2008 of the Australian Marine National Facility R/V Southern Surveyor [Arculus, 2004, 2008], expeditionTN-234 of the R/V Thomas Thompson [TN-234, 2009], expeditions KM1024 and KM1129a of the R/V KiloMoana, and expedition RR1211 of the R/V Roger Revelle [RR1211, 2012]. Rock samples were collected byboth dredging and grab sampling by ROV.

For the gas analysis of volcanic rocks, subsamples were selected in order to optimize the purity of the glassand to minimize the occurrence of phenocrysts. Small chips of volcanic glass from the chilled exterior mar-gin of the samples (� 300 mg) were cleaned in distilled water and acetone, and then loaded into stainlesssteel ‘‘crusher’’ tubes. Blanks were run on each of the loaded tubes before proceeding with mass spectrome-ter analysis. Gases were released by crushing the samples to �100 microns size in vacuo, a process whichreleases the gases trapped in the vesicles. Condensable gases (mainly H2O and CO2) were removed on a U-trap held at 2195� C. The ‘‘non-condensible’’ gases were first exposed to a hot Ti getter, followed by expo-sure to a charcoal finger at 2195�C. The remaining gases, mainly He and Ne, were then exposed to a lowtemperature trap at �38�K, which traps 98% of the Ne while leaving the He in the gas phase. The purifiedsample, consisting essentially of pure helium, was then admitted to the mass spectrometer for analysis. Inthe second phase of the analysis, the trap temperature was raised to 150�K, releasing the neon whichwas then admitted to the mass spectrometer for analysis. After completion of the mass spectrometer analy-sis, the temperature of the U-trap was raised to 278�C, thereby releasing the CO2 fraction while retainingH2O. The CO2 fraction was then transferred into a calibrated volume where the pressure was measured witha precision capacitance manometer. The CO2 and helium were both released by crushing, and thus theresulting carbon/helium ratios reflect the gas composition contained in the vesicles.

Samples of hydrothermal vent fluids were collected using special titanium alloy gas-tight bottles having aninternal volume of �150 cc. Each bottle was pumped to a high-vacuum prior to each submersible dive, andthe sample was collected by first inserting the Ti bottle snout into the vent orifice and then opening thebottle valve by depressing the bottle trigger cylinder. After each dive, the samples were processed on a sea-going high vacuum line. The contents of the gas-tight bottle, consisting of a mixture of fluid and gas, weredropped into an evacuated flask containing sulfamic acid powder. A metal bellows pump was used topump the released gases through a U-trap held at 260�C into a calibrated volume. The total gas contentwas determined by measuring the gas pressure at a known temperature with a precision capacitancemanometer. Multiple splits of the dry extracted gases were then sealed into glass ampoules. For generalgas analysis, pyrex ampoules were used, while samples for helium and rare gas analysis were sealed intoampoules constructed of aluminosilicate glass with low helium permeability. At the end of the extraction,the water frozen in the U-trap was melted and then combined with the water remaining in the extractionflask. The water was then weighed to determine the total sample weight and then saved in Nalgene bottlesfor subsequent analysis of Mg and other fluid properties. This method avoids any problems of phase

volcanic arcrear-arc

seamountback-arc

ridge

intraplatevolcano

asthenosphere

lithosphere

2

13

4

Figure 3. Schematic diagram showing the principal mantle inputs to back-arc basinvolcanic systems. Solid arrows show mantle flow and tectonics. Dotted arrows indi-cate melt extraction. Numbers denote mantle melting zones: 1 mantle input, 2 sub-duction input, 3 component interaction, 4 back-arc melting column. After similarfigure in Pearce and Stern [2006].



separation within the gas-tight bottle since the entire sample of fluid and gas is treated as a whole. Adetailed description of the extraction process including a diagram of the seagoing extraction line is given inLupton et al. [2006].

Water-column samples were collected with a standard CTD-bottle-rosette sampling system. Samples forhelium analysis were drawn from the rosette bottles on deck and sealed into copper tubing using a specialhydraulic crimping system [Young and Lupton, 1983]. All sample preparation and mass spectrometric analy-ses of rare gases were carried out at the Helium Isotope Laboratory, NOAA Pacific Marine EnvironmentalLaboratory, Newport, Oregon. Helium and neon concentrations and helium isotope ratios were determinedusing a 21 cm radius, dual-collector, sector-type instrument specially designed for helium isotope analyses[see Lupton, 1990]. The measurements were standardized using marine air and also a precisely known geo-thermal standard (the MM gas from Yellowstone Park, 16.5 Ra) [Lupton and Evans, 2004]. For rock and ventfluid samples, the precision for the helium isotope determinations averaged about 0.5% (1 r) in the3He/4He ratio, or about 0.05 Ra. The helium blanks averaged 1.3 3 10210 cc STP He, in every case less than1% of the sample size. The neon blanks averaged 4.7 3 10211 cc STP Ne. The precision for helium isotoperatios on seawater samples averaged about 0.2% in d3He, where d3He is the percentage deviation of3He/4He from the atmospheric ratio.

Carbon dioxide concentrations were determined by a variety of methods depending on the type of sample.As described above, for rock samples CO2 was determined manometrically with the assumption that >90%of the dry condensable gas is CO2. Thus our CO2 concentrations for rock samples are uncertain to about10%. For vent fluid samples, precision splits of the extracted gases in glass ampoules were sent to the Uni-versity of Washington where CO2, N2, O2, CH4, and H2 were determined by gas chromatography. For water-column seawater samples, samples were drawn from CTD rosette bottles and Total CO2 was determined bystandard methods [DOE, 1994]. Changes in CO2 were calculated by subtracting the regional background,which varies with depth.

In some of the discussion which follows, we have also plotted 3He/4He and C/3He ratios for rock samplesrecently published by Hahm et al. [2012] as a useful comparison with our results (Figure 1). As describedabove, we extracted helium and CO2 by crushing a single sample which releases the gases in the vesicles.The helium was then measured by mass spectrometry and the CO2 by manometry. Hahm et al. [2012]extracted the helium by crushing as we did but analyzed CO2 by manometry on a separate split of the sam-ple using a stepped heating technique in which 700–900�C and 1000–1200�C fractions were analyzed sepa-rately. These fractions were assumed to be derived from the vesicles and glass matrix, respectively, andthese two CO2 concentrations are reported separately in their data tables. In deriving C/3He ratios from theHahm et al. [2012] data, we used the CO2 concentration from their 700–900�C fraction and divided it by thecorresponding 3He concentration, assuming that this would give a consistent C/3He ratio for the vesicle gas.When data from the two laboratories are compared within a given geographic province, the data fieldslargely overlap. However, upon closer inspection, the Hahm et al. [2012] C/3He ratios appear to be higher(approximately a factor of 2) than our results on the average (see Figures 8 and 9). This may be due to thedifferent analytical technique used by Hahm et al. [2012], namely that they released helium by crushing andCO2 by stepped heating on different splits of the sample.

Ba and Nb were analyzed by X-ray florescence spectroscopy of pressed powder whole rock pellets and Tiwas measured on fused glass disks at the University of Hawaii using standard methods. The instrument wascalibrated against rock standards with a range of Ba, Nb, and Ti abundances (e.g., BIR-1, BCR-1, AGV-1,BHVO-1) and checked against standards run as unknowns to have precision and accuracy of better than10% for Ba and Nb, and 2.5% for Ti.

3. Helium Isotope and C/3He Variations by Tectonic Setting

3.1. Rock Samples and Submarine Hydrothermal FluidsIn order to use the combination of 3He/4He ratios and C/3He ratios to distinguish between various mantlesignatures in the northern Lau Basin, it is first necessary to examine the variability of these ratios for basicmantle reservoirs. We did this using an empirical approach by assessing the 3He/4He-C/3He for representa-tive samples from mid-ocean ridges (MOR), arcs, and hotspots. Figure 4a shows a semilog plot of 3He/4He(R/Ra) versus C/3He for MORs, submarine arc volcanoes, and subaerial arc volcanoes. To define the mid-



ocean ridge field, we have plotted values for both N-type and E-type MOR basalts [Marty and Zimmermann,1999; Marty and Tolstikhin, 1998] and for MOR hydrothermal fluids from a variety of mid-ocean ridges. Witha few exceptions, the basalt and vent fluid data agree and define a MOR field in which 3He/4He varies from7.0 to 9.5 Ra, and C/3He varies from 2 3 108 up to 3 3 109. For volcanic arcs, we first consider the valuescompiled by Hilton et al. [2002] for subaerial volcanic arcs. In this case, each data point represents an aver-age for several volcanoes in each of various localities (see Table 1). With the exception of the Philippines,which have somewhat higher 3He/4He (average 7.34 Ra), most subaerial arc volcanoes fall between 4.4 and6.1 Ra and have C/3He between 8 3 109 and 6 3 1010 (Figure 4a),[Hilton et al., 2002]. Thus subaerial arcshave much lower 3He/4He ratios and much higher C/3He ratios compared to MORBs, defining a field distinctfrom the MOR field. For completeness, we also consider submarine arc volcanoes. For this we have plottedvalues for submarine hydrothermal fluids collected along the Mariana and Kermadec arcs (see Table 1). Sur-prisingly, the submarine arc volcanoes define a third field with 3He/4He varying from 6.5 to 8.2 Ra and C/3Hefrom 8 3 109 to 8 3 1010. Thus while submarine arc volcanoes exhibit approximately the same elevatedC/3He as the subaerial arcs, they have distinctly higher 3He/4He ratios. This difference may be due to theincorporation of U- and Th-rich continental-type crustal material into the melts along subaerial arcs whichhas had the effect of lowering the 3He/4He ratio. Another factor is that continental arcs are subaerial andmore evolved and thus more likely to have been affected by degassing.

The higher C/3He in both subaerial and submarine arcs compared to mid-ocean ridges is attributed to theaddition of carbon in arc melts, likely from marine carbonates and organic materials that are subductedwith the downgoing slab and incorporated into the melt at depth [Hilton et al., 2002]. Additional evidencefor the presence of marine carbonates in volcanic arcs has come from measurements of carbon isotopes.Although we have not measured d13C as part of the present study, other studies have found d13C in CO2 inarc volcanoes ranging from 25 to 11& [Sano and Williams, 1996; van Soest et al., 1998, Lupton et al., 2006],similar to marine carbonates (�0&) [Hoefs, 1980], and heavier than mid-ocean ridges (213 to 24&) [Kelleyet al., 2004]. The lower 3He/4He along arcs is attributed to the fact that magma genesis in subduction zonesinvolves melting of the degassed downgoing slab (perhaps R/Ra 5 1) which is mixed with fresh mantlematerial, resulting in 3He/4He ratios lower than DMM [Poreda and Craig, 1989].

Finally, we examine the 3He/4He and C/3He signatures for mantle plumes or hotspots. In this case, we focuson so-called 3He-rich hotspots by plotting values for Loihi seamount, Kilauea fumarole gases, YellowstonePark gases, and the island of Reunion (Table 1). As shown in Figure 4b, these hotspots define a large fieldin the plot, with 3He/4He varying from 13 to 32 Ra and C/3He varying by over an order of magnitude from

Figure 4. (a) Plot summarizing 3He/4He ratio versus C/3He for various mantle provinces, including mid-ocean ridges (black-filled symbols),submarine arc volcanoes (blue), and subaerial arc volcanoes (green). 3He/4He is expressed as R/Ra. Crosses indicate average values forMORBs and for subaerial arcs from Hilton et al. [2002]. (b) Similar plot including values for hotspot volcanoes such as Loihi, Kilauea fumar-ole, Yellowstone Park gases, and Reunion. Data from Hilton et al. [1998, 2002], Marty and Zimmermann [1999], Lupton et al. [2006, 2008], Lil-ley et al. [2003], Rison and Craig [1983], Kurz et al. [1983], Trull et al. [1993], Exley et al. [1986], Craig and Lupton [1976], Naughton et al.[1973], Craig et al. [1978b], and Staudacher et al. [1990]. Data and references are listed in Table 1.



Table 1. Background 3He/4He and C/3He values for various mantle provinces3He/4He (R/Ra) C/3He Referencesa

Subaerial arcs averageAndes 4.62 4.55E110 1Alaska-Aleutians 6.08 2.37E110 1Lesser Antilles 5.86 1.55E110 1Central America 6 9.70E109 1Indonesia 4.83 2.20E110 1Italy 4.38 2.38E110 1Japan 5.15 1.06E110 1Kamchatka-Kuriles 6.05 1.13E110 1New Zealand 6 1.26E110 1Philippines 7.34 8.80E109 1

Average subaerial arc flux 5.4 1.57E110 1Average MOR Flux 8.06 2.20E109 1

MOR basaltsEPR 13�N 8.1 1.26E109 2EPR 13�N 8.8 4.60E108 2EPR 6�N 7.9 9.50E108 2EPR 18�N 8.5 9.20E108 2EPR 18�N 9.1 1.08E109 2EPR 18�N 8.7 5.40E108 2EPR 21�S 7.6 2.55E109 2EPR 21�S 8.1 4.90E108 2Red Sea (E-MORB) 8.2 2.49E109 3b

Red Sea (E-MORB) 8.2 2.73E109 3b

Red Sea (E-MORB) 9.6 3.02E109 3b

Red Sea (E-MORB) 8 6 1 2.92E109 3b

SE Indian Ridge 8.9 4.90E108 2Indian Ridge Triple Junction 8 4.80E108 2Indian Ridge Triple Junction 8.9 1.12E109 2Indian Ridge Triple Junction 8.1 2.20E108 2EPR 13�N 8 6 1 6.75E108 3b

EPR 13�N 8 6 1 1.09E109 3b

EPR 13�N (E-MORB) 8 6 1 4.26E108 3b

EPR 13�N (E-MORB) 8 6 1 2.00E108 3b

EPR 13�N 8 6 1 5.37E108 3b

MAR 14�N (E-MORB) 8 6 1 2.32E109 3b

MAR 30�N 8 6 1 1.32E109 3b

MAR 30�N 8 6 1 1.11E109 3b

MAR 30�N 8 6 1 9.30E108 3b

MAR 30�N 8 6 1 7.84E108 3b

MAR 36�N (E-MORB) 8 6 1 5.07E109 3b

MAR 36�N (E-MORB) 8 6 1 4.72E109 3b

MAR 36�N (E-MORB) 8 6 1 1.49E108 3b

MAR 36�N 8 6 1 3.18E109 3b

MAR 36�N (E-MORB) 8 6 1 3.65E109 3b

MAR 36�N (E-MORB) 8 6 1 2.86E109 3b

MOR vent fluidsJdFR Axial - Inferno 8.2 2.35E109 4c

JdFR Axial - Virgin Mound 8.2 2.15E109 4c

JdFR Axial - Hell 8.2 2.40E109 4c

SEPR - Pagoda 9.02 8.50E108 4c

SEPR - DragonTeeth 8.8 2.00E109 4c

SEPR - Brandon 8.2 2.00E109 4c

SEPR - Kasuga 8.6 5.40E109 4c, 7Endeavour - Hulk 8.1 9.40E108 4c, 7Endeavour - Hulk 8 2.60E109 4c, 7Endeavour - Bastille 8.1 1.80E109 4c, 7Endeavour - Dante 8.3 1.60E109 4c, 7Endeavour - Puffer 8 1.80E109 4c, 7EPR 9�N - X Vent 8.5 8.30E109 4c

EPR 9�N - M Vent 8.5 1.40E109 4c

EPR 9�N - A Vent 8.4 1.60E109 4c, 7EPR 9�N - Bio9 Vent 8.6 1.06E109 4c

Submarine arc vent fluidsMariana - E Diamante 7.6 9.40E109 4c

Mariana - E Diamante 7.8 1.10E110 4c

Mariana - Maug 7.14 1.50E110 4c

Mariana - Kasuga #2 7.8 2.20E110 4c



6 3 108 to 2 3 1010. Within this suite of samples, hotspots have distinctly higher 3He/4He ratios compared tothe other mantle reservoirs, and exhibit a large variation in C/3He that overlaps both the MORB and arc fields.

An important consideration is possible fractionation of carbon versus helium, which can affect C/3He ratiosfor both rocks and fluids. While helium isotope ratios are fairly immune to fractionation effects, CO2 and

Table 1. (continued)3He/4He (R/Ra) C/3He Referencesa

Mariana - Ruby 7.7 2.90E110 4c

Mariana - Seamount X 7.9 2.80E110 4c

Mariana - Forecast 8 5.00E109 4c

Mariana - NW Rota 8.3 9.00E109 5Mariana - Daikoku 7.4 6.40E110 5Mariana - NW Eifuku 7.3 5.45E110 5, 6Mariana - Nikko 6.8 3.95E110 5Kermadec - Giggenbach 7.4 2.50E109 5Kermadec - Volcano 1 6.6 1.68E110 5Kermadec - Monowai 7.2 6.60E109 4Kermadec - Brothers NW 7.1 7.00E109 4Kermadec - Brothers Cone 7.2 2.00E110 4Kermadec - Volcano 19 6.9 5.40E109 4Kermadec - MacCauley 7.2 2.60E110 4Kermadec - Clark 6.9 4.10E109 4Kermadec - Healy 7.2 1.60E110 4Kermadec - Tangaroa 7.1 1.40E110 4

HotspotsLoihi basalt 28.2 1.75E110 8Loihi basalt 31.9 5.40E109 9, 10Loihi basalt 26.7 4.94E109 9, 10Loihi basalt 30.1 1.79E109 9, 10Loihi xenolith 23.9 2.90E109 10, 11Loihi xenolith 24 2.50E109 10, 11Loihi vent fluid 25.3 6.40E108 12Loihi vent fluid 24.3 5.90E108 12Loihi vent fluid 25.2 9.60E108 12Loihi vent fluid 24.5 6.90E108 12Loihi vent fluid 24 7.30E108 12Loihi vent fluid 17.7 8.20E108 12Loihi vent fluid 25.5 6.60E108 12Loihi vent fluid 23.5 7.90E108 12Loihi vent fluid 20.5 3.60E109 12Loihi vent fluid 20.2 3.50E109 12Loihi vent fluid 23.8 3.70E109 12Loihi vent fluid 24.1 3.80E109 12Loihi vent fluid 23.5 3.40E109 12Loihi vent fluid 17.8 4.60E109 12Loihi vent fluid 24.7 2.70E109 12Loihi vent fluid 12.4 2.90E109 12Loihi vent fluid 25.6 5.40E108 13c

Loihi vent fluid 25.2 6.70E108 13c



Kilauea Fumarole gas 15 3.40E109 14, 15Yellowstone gas 12.1 2.70E109 16Yellowstone gas 15.6 2.10E109 16Reunion rock 13.1 2.20E109 10, 17Reunion rock 13.4 4.70E109 10, 17Reunion rock 13 1.50E110 10, 17Reunion rock 13.4 6.50E109 10, 17Reunion rock 14.3 1.30E110 10, 17Reunion rock 13.4 1.78E110 10, 17

aReferences 1, Hilton et al. [2002]; 2, Marty and Zimmermann [1999]; 3, Marty and Tostikhin [1998]; 4, Lupton, Lilley, and Butterfield(unpublished data); 5, Lupton et al. [2008]; 6, Lupton et al. [2006]; 7, Lilley et al. [2003]; 8, Rison and Craig [1983]; 9, Kurz et al. [1983]; 10,Trull et al. [1993]; 11, Exley et al. [1986]; 12, Hilton et al. [1998]; 13, Lupton, Lilley, and Wheat (unpublished data); 14, Craig and Lupton[1976]; 15, Naughton et al. [1973]; 16, Craig et al. [1978b]; 17, Staudacher et al. [1990].

bThis reference did not list actual 3He/4He measurements, but stated that all the values fell within the normal MORB range of 8 6 1 Ra.cThese samples were all collected and analyzed in the same manner as the other samples discussed in this paper. Samples were col-

lected with Ti gas tight bottles. Helium was analyzed by mass spectrometry at NOAA/PMEL in Newport, OR. CO2 was analyzed by gaschromatography at U. of Washington.



helium have much different solubilities inmagma and in water, and thus carbon/heliumratios in volcanic rocks and vent fluids can beaffected by fractionation. It is likely that thiseffect contributes to the scatter we see inC/3He ratios, e.g., Figure 4. For volcanic rocks,bubble nucleation and vesicle formation dur-ing magma ascent can potentially affect theCO2/helium ratio. Furthermore, many rocksamples appear to have suffered preferentialhelium loss leading to elevated C/3He ratios.Although this is less of a problem for basalticsamples, samples with higher silica contentsuch as basaltic andesites, andesites, anddacites are prone to helium loss due to higherhelium diffusivity. Several investigators haveused the 4He/40Ar* ratio to identify samplesthat have been compromised by degassing(40Ar* is the nonatmospheric or mantle-derived 40Ar fraction) [Burnard et al., 2004;Shaw et al., 2004]. These studies showed thatthe 4He/40Ar* ratio increases as samples suffergas loss, and undegassed samples wouldideally have 4He/40Ar* of 2 to 4, which is theestimated mantle ratio [Graham, 2002; Bur-

nard et al., 2004]. However, argon isotope ratios are not always available and, in the case of the presentstudy, argon isotopes were not measured on any of our Lau Basin samples. Instead of using the 4He/40Ar*ratio, we have used 4He concentrations to identify rock samples which have suffered serious gas loss. Byplotting C/3He versus 4He concentration for specific localities, we found that rock samples with helium con-tent less than �5 3 1027 cc STP/g (2.2 3 1028 mM/kg) have elevated C/3He and thus appear to have beencompromised by helium loss (Figure 5). Therefore, we have eliminated most of these low [4He] samplesfrom consideration.

In the case of helium and CO2 in hydrothermal vent fluids, these fluids contain a finite proportion of dis-solved atmospheric gases, including He, Ne, and Ar. Assuming that virtually all of the Ne in the vent fluids is

Figure 6. Results for Total CO2, [3He] and [4He] for a suite of water-column samples collected over West Mata volcano in 2008 illustrating how dilute water-column samples can be usedto estimate end-member 3He/4He and C/3He ratios. (a) Total CO2 and [3He] profiles versus depth. The dashed line indicates the total CO2 values for background (nonplume) waters inthis region which increase slightly with depth. The difference between these background values and the values in the plume is then the excess CO2. (b) [3He] versus [4He] yielding a3He/4He slope of 7.19 Ra. (c) Excess CO2 versus [3He] yielding a C/3He slope of 8.4 3 109.

Figure 5. Plot of C/3He versus 4He content for rock samples in the LauBasin showing that for low [4He] the C/3He ratios become unreliable,tending to be anomalously high due to helium loss. The plot includes allthe samples considered for this project. Samples for which [4He]� 5 3

1027 cc STP/g (shaded region) were for the most part rejected for ourstudy. The exceptions were a few samples from West Mata and from theFonualei Rift with low [4He] which were used because the C/3He agreedwith vent fluid results from the same site.



Table 2. Vent Fluid, Rock, and Water-Column Plumes Data Used in This Study

Site SampleDate

Month–Year Sampler Site DescriptionDepthmeters Latitude Longitude

Vent T(�C)

TotalGas

(mM/Kg)He

(mM/Kg)He/Ne

airaR/Ra

Empiricalb

R/RaCorr

CorrectedcMg

(mM/Kg)CO2

(mM/Kg) CO2/3He

Vent fluid samplesMaka J2–416-13 May 2009 GTB 2 Maka 1525 215.422 2174.284 315 81.4 5.97E-03 528.1 8.65 8.66 4.4 69.5 9.7E108Maka J2–416-14 May 2009 GTB 11 Maka -

mk1491525 215.422 2174.284 316 83.5 6.03E-03 669.5 8.67 8.68 4.6 66.2 9.1E108

Maka J2–416-19 May 2009 GTB 7 Maka 1525 215.422 2174.284 81.8 6.03E-03 555.8 8.65 8.66 5.0 66.7 9.2E108NELSC J2–415-17 May 2009 GTB 6 SW-NELSC

worm mkr1615 215.383 2174.245 21 4.9 1.34E-04 11.9 7.81 7.98 52.6 4.1 2.8E109

West Mata J2–414-7 May 2009 GTB 11 Kohu 1187 215.094 2173.748 31 19.0 1.89E-04 11.1 7.15 7.32 60.4 18.0 9.7E109West Mata J2–414-16 May 2009 GTB 9 Prometheus 1174 215.094 2173.748 62 - 78 99.5 5.42E-04 37.7 7.16 7.21 49.8 60.0 1.1E110West Mata J2–414-20 May 2009 GTB 5 Prometheus

smoke1174 215.094 2173.748 64.3 2.48E-04 21.4 7.17 7.26 50.9 27.3 1.1E110

West Mata J2–414-21 May 2009 GTB 7 Prometheus,above fire

1174 215.094 2173.748 41.7 1.86E-04 14.3 7.15 7.28 52.4 18.6 1.0E110

West Mata J2–417-23 May 2009 GTB 11 Hades 1202 215.095 2173.749 50 29.3 1.57E-04 10.1 7.12 7.31 49.5 16.0 1.0E110West Mata J2–417-24 May 2009 GTB 9 Hades 1202 215.095 2173.749 50 23.6 1.24E-04 6.4 7.06 7.35 49.9 12.9 1.1E110West Mata J2–418-13 May 2009 GTB 7 Hades 1201 215.095 2173.749 55 - 95 61.1 2.81E-04 8.1 7.09 7.31 48.5 26.9 9.8E109West Mata J2–418-17 May 2009 GTB 2 Hades 1201 215.095 2173.749 55 - 95 46.4 2.50E-04 15.0 7.13 7.26 47.2 23.2 9.4E109Fonualei Q323MWc3 Sept. 2012 Major Fonualei

South1547 217.535 2174.567 250 10.0 1.94E-04 10.3 7.72 7.92 42.6 8.4 4.1E109

Niuatahi(Volcano O)

Q324g3c3 Sept. 2012 GTB 3 Volcano O 1244 215.376 2174.002 22 24.0 1.42E-04 10.1 7.16 7.34 52.4 23.1 1.6E110

Niuatahi(Volcano O)

Q324g5c3 Sept. 2012 GTB 5 VolcanoO cone

1247 215.376 2174.002 10.9 4.36E-05 3.0 6.85 7.48 51.8 9.6 2.3E110

Niua South Q326g7c3 Sept. 2012 GTB 7 Niua South 215.164 2173.574 23.3 2.28E-05 5.5 6.83 7.15 20.0 21.2 9.9E110Niua South Q333g3c4 Sept. 2012 GTB 3 Niua South 1150 215.164 2173.573 274 112.0 8.61E-04 35.9 7.23 7.28 12.6 110.5 1.3E110Niua North Q330g4c3 Sept. 2012 GTB 4 Niua North 754 215.081 2173.555 79.1 6.37E-04 50.1 7.10 7.13 46.7 71.5 1.1E110Niua North Q330g14c7 Sept. 2012 GTB 14 Niua North 748 215.081 2173.555 133.0 1.34E-03 92.3 7.14 7.16 49.0 115.5 8.7E109Mata Ua Q328g5c3 Sept.2012 GT 5 Mata Ua 2366 215.017 2173.788 360 4.6 2.74E-05 2.3 6.76 7.58 51.0 3.6 1.4E110Mata Ua Q328g6c5 Sept. 2012 GTB 6 Mata Ua 2366 215.017 2173.788 360 159.3 2.29E-03 103.3 7.38 7.40 1.6 154.5 6.6E109Mata Fitu Q329g7c3 Sept. 2012 GTB 7 Mata Fitu 2585 214.913 2173.781 15.5 6.76E-05 5.2 7.11 7.47 45.5 13.4 2.0E110Mata Fitu Q329g12c3 Sept. 2012 GTB 12 Mata Fitu 2627 214.914 2173.779 32.7 5.23E-05 9.3 7.25 7.45 34.7 31.2 5.9E110Mata Tolu Q331g13c3 Sept. 2012 GTB 13 Mata Tolu 1817 215.005 2173.794 270 14.1 9.94E-05 8.0 7.21 7.45 46.1 12.9 1.3E110Mata Tolu Q331g18c3 Sept. 2012 GTB 18 Mata Tolu 1840 215.005 2173.793 242 21.0 2.26E-04 20.7 7.20 7.29 41.4 18.9 8.4E109

Mata Tolu Q331g19c3 Sep-12 GTB 19 Mata Tolu 1840 215.005 2173.793 242 47.2 6.74E-04 40.4 7.18 7.23 25.2 41.6 6.2E109West Mata Q332g8c3 Sept.2012 GTB 8 West Mata 1176 215.094 2173.748 diffuse 7.9 5.43E-05 3.7 6.83 7.32 55.3 6.6 1.3E110West Mata Q332g9c3 Sept. 2012 GTB 9 West Mata 1176 215.094 2173.748 diffuse 7.0 4.31E-05 3.8 6.89 7.38 55.3 5.9 1.4E110

Rock samplesWest Mata J2–413R13v6 May 2009 ROV Prometheus

scoria/spatter1174 215.094 2173.748 33.3 7.05 7.24 9.5E109

West Mata J2–418R20v9 May 2009 ROV Akel’sAfi scoria

1198 215.094 2173.749 22.6 7.00 7.28 8.1E109

Maka J2–416bio22v14 May 2009 ROV H-therm site,older pillow

1542 215.422 2174.284 11.7 7.97 8.62 6.5E108

Maka J2–416bio22v28 May-09 ROV H-therm site,older pillow

1542 215.422 2174.284 22.0 8.24 8.59 2.8E108

NELSC J2–415R07v11 May 2009 ROV sheet/lobate 1626 215.390 2174.250 14.1 8.11 8.66 5.5E109NELSC J2–415R20v12 May 2009 ROV North end

lobate flow1657 215.382 2174.242 20.5 8.17 8.54 3.2E109

West Mata KM1024 D15 R01 Dec. 2010 Dredge 79.2 7.28 7.39 6.6E109Fonualei RSCd ND-37.2 Nov. 2004 Dredge small cone on

overlapperridge

1600 217.017 2174.500 1.69E-05 11.9 7.13 7.57 0.553 3.1E109

Fonualei RSCd ND-47.1 Nov. 2004 Dredge cone onpropagator

2620 217.633 2174.567 8.56E-06 17.8 7.43 7.73 0.399 4.3E109

Fonualei RSCd ND-56.1 Nov. 2004 Dredge cone 1520 217.017 2174.517 1.11E-05 12.0 6.58 6.98 0.308 2.8E109Fonualei RSCd ND-60.1 Nov. 2004 Dredge termination of

central ridge1840 216.750 2174.517 4.45E-06 37.9 7.76 7.91 0.152 3.1E109

Fonualei RSCd ND-45.1 Nov. 2004 Dredge cone oncentral ridge

1890 217.467 2174.583 2.72E-06 35.4 7.52 7.67 0.021 7.4E108

Mangatolu TJd ND-70.1 Nov. 2004 Dredge axial ridge crest 2500 215.867 2174.850 1.71E-05 6624 7.22 7.22 0.054 3.1E108Mangatolu TJd ND-69.1 Nov. 2004 Dredge cone 2220 215.600 2174.817 3.18E-04 5456 7.33 7.33 3.283 1.0E109Rochambeau

RiftsNLD-8 Apr.2008 Dredge 2522 214.70 2175.97 6.57E-05 11.22 11.22 0.022 9.5E108

RochambeauRifts

NLD-11-1 Apr. 2008 Dredge 2660 214.71 2175.98 2.14E-05 10.88 10.88 0.004 5.3E108

RochambeauRifts

NLD-13-1 Apr. 2008 Dredge 2081 214.83 2175.97 5.12E-05 28.10 28.10 0.009 2.0E108

RochambeauRifts

NLD 13b Apr. 2008 Dredge 2081 214.83 2175.97 2.42E-05 28.12 28.12 0.005 2.3E108

Rochambeau Rifts NLD-14-2 Apr. 2008 Dredge 2096 214.84 2176.00 2.52E-04 17.85 17.85 0.058 4.1E108Rochambeau

RiftsNLD-18-2 Apr. 2008 Dredge 2295 214.83 2176.07 1.55E-05 12.27 12.27 0.003 5.8E108

RochambeauRifts

NLD-20 Apr. 2008 Dredge 2096 214.94 2176.20 1.05E-04 18.55 18.55 0.038 6.3E108

RochambeauRifts

NLD-24 Apr. 2008 Dredge 1834 215.11 2176.41 7.45E-05 14.31 14.31 0.023 6.9E108

RochambeauRifts

NLD-25-1 Apr. 2008 Dredge 2038 215.10 2176.29 8.12E-05 14.92 14.92 0.030 7.9E108

RochambeauRifts

NLD 27-2 Apr. 2008 Dredge 2021 215.07 2176.15 9.40E-04 15.36 15.36 0.229 5.1E108



an atmospheric component derived from seawater, we have corrected the 3He/4He ratio by subtracting outthe atmospheric component using the formula Rcorr/Ra5 [(R/Ra)X21]/(X21) where R 5 3He/4He, Ra5

Rair 5 1.39 3 1026 and X 5 (He/Ne)/(He/Ne)air [Craig et al., 1978a]. We applied the same correction to ourresults for rock samples using the He/Ne ratio. We have also eliminated vent fluid samples that have (He/Ne)/(He/Ne)air� 20, since that implies that �5% or more of the helium is atmospheric in origin.

In summary, while fractionation effects contribute to variations in C/3He ratios, we have attempted to mini-mize these effects by filtering out samples that have clearly suffered gas loss. Although considerable varia-tions in C/3He ratios remain, the empirical approach we have adopted defines clear signatures for MORBs,submarine arcs, subaerial arcs, and hotspots by plotting 3He/4He versus C/3He.

3.2. Water-Column PlumesGases trapped in seafloor volcanic rocks and dissolved in hydrothermal vent fluids usually provide the bestestimates of the magmatic 3He/4He and C/3He signature in a given locality. However, in cases where directseafloor samples are not available, it is still possible to estimate 3He/4He and C/3He if an active hydrothermalsystem is generating strong water-column plumes. We have done this for plumes over East Mata, WestMata, Niua, Niuatahi, and Mata Ua. As an example, Figure 6 shows how a suite of CTD rosette samples

Table 2. (continued)

Site SampleDate

Month–Year Sampler Site DescriptionDepthmeters Latitude Longitude

Vent T(�C)

TotalGas

(mM/Kg)He

(mM/Kg)He/Ne

airaR/Ra

Empiricalb

R/RaCorr

CorrectedcMg

(mM/Kg)CO2

(mM/Kg) CO2/3He

RochambeauRifts

NLD-31-1 Apr. 2008 Dredge 1279 215.12 2176.27 2.52E-05 15.57 15.57 0.004 2.9E108

RochambeauRifts

NLD-32 Apr. 2008 Dredge 1509 215.34 2176.27 8.57E-05 17.84 17.84 0.006 1.2E108

RochambeauRifts

NLD-37 Apr. 2008 Dredge 1811 215.47 2176.48 8.97E-06 18.24 18.24 0.005 1.0E109

NWLSC NLD-38 Apr.2008 Dredge 1995 215.731 2177.203 8.11E-05 13.81 1.23 7.9E108NWLSC NLD-40 Apr. 2008 Dredge 2200 215.593 2177.095 8.12E-05 12.72 1.23 8.6E108NWLSC NLD-41b Apr. 2008 Dredge 1971 215.674 2177.161 1.46E-04 13.18 1.55 5.8E108NWLSC NLD-42 Apr. 2008 Dredge 2108 215.802 2177.268 3.52E-05 15.65 0.21 2.8E108NWLSC NLD-42b Apr. 2008 Dredge 2108 215.802 2177.268 3.14E-05 15.46 0.27 4.0E108NWLSC NLD-43 Apr. 2008 Dredge 1986 215.800 2177.282 3.15E-05 13.71 0.04 7.5E107NWLSC NLD-44 Apr. 2008 Dredge 2033 215.958 2177.478 7.45E-05 17.05 0.71 4.0E108NWLSC NLD-45-1 Apr. 2008 Dredge 2165 215.915 2177.422 1.31E-05 15.67 0.05 1.8E108NWLSC NLD-45b Apr. 2008 Dredge 2165 215.915 2177.422 1.07E-05 15.65 0.21 9.0E108NWLSC NLD-46 Apr. 2008 Dredge 1995 215.886 2177.395 7.37E-05 17.61 0.92 5.1E108NWLSC NLD-47 Apr. 2008 Dredge 2140 215.895 2177.368 5.65E-05 14.14 1.25 1.1E109NWLSC NLD-48 Apr. 2008 Dredge 2139 215.796 2177.265 2.95E-05 15.94 0.31 4.7E108NWLSC NLD-50 Apr. 2008 Dredge 2059 215.985 2177.523 2.17E-05 16.49 0.18 3.7E108NWLSC NLD-51-1 Apr. 2008 Dredge 2129 216.039 2177.577 7.68E-05 14.22 1.51 9.9E108NWLSC NLD-52 Apr. 2008 Dredge 2192 216.054 2177.665 1.07E-05 7.37 0.10 8.8E108Dugong

VolcanoNLD-35 Dredge 215.459 2175.702 4.01E-06 8.99

DugongVolcano

NLD-35-2 Dredge 215.459 2175.702 1.40E-04 8.95

ILSCd DR03-x T2 Dredge 2438 219.079 2176.068 5.30E-05 7.72 1.36 2.4E109ILSCd DR04-1 Dredge 3254 219.333 2176.166 1.82E-05 9.63 0.19 7.7E108ELSCd DR05-1 Dredge 2938 219.441 2175.959 1.61E-04 9.25 2.32 1.1E109ELSCd DR06-3 Dredge 2636 219.62 2175.986 2.24E-04 8.90 2.83 1.0E109ELSCd DR07-1 Dredge 2672 219.783 2176.034 2.73E-04 9.06 3.16 9.2E108ELSCd DR08-1 Dredge 2660 219.96 2176.092 1.86E-04 8.76 2.52 1.1E109ELSCd DR52-1 Dredge 2640 220.05 2176.134 7.76E-05 8.86 1.11 1.2E109ELSCd RC31 Dredge 2778 220.247 2176.111 1.22E-04 8.99 2.28 1.5E109ELSCd DR11-2 Dredge 2700 220.317 2176.143 1.17E-05 8.80 0.14 9.9E108ELSCd DR53-1 Dredge 2700 220.32 2176.139 1.67E-05 8.98 0.10 4.9E108ELSCd DR10-2 Dredge 2650 220.382 2176.161 3.34E-05 8.46 0.19 4.9E108ELSCd DR41-1a Dredge 2482 220.52 2176.204 6.35E-05 8.43 2.24 3.0E109ELSCd DR41-1b Dredge 2482 220.52 2176.204 3.43E-05 8.51 1.03 2.5E109ELSCd DR50-1 Dredge 2154 220.762 2176.191 4.94E-05 8.50 0.52 9.0E108ELSCd RC46 Dredge 2205 221.25 2176.339 2.18E-05 8.05 0.27 1.1E109ELSCd RC71 Dredge 1816 221.694 2176.441 2.87E-05 8.10 0.58 1.8E109

Water-column plume samplesEast Mata CTD rosette 215.1 2173.7 7.14 3.6E109Niua CTD rosette 215.1 2173.6 6.90 9.8E109West Mata 2008 CTD rosette 215.09 2173.75 7.19 8.4E109Niuatahi V10B-11 2010 CTD rosette 215.38 2174.00 7.36 1.4E110

aHe/Ne/(He/Ne)air.bHelium isotope ratio expressed as R/Ra, where R 5 3He/4He, and Ra 5 Rair 5 1.4 3 1026.cHelium isotope ratio corrected for air addition using Rcorr/Ra 5 [(R/Ra)X 21]/(X21) where X 5 He/Ne/(He/He)air.dThese samples were analyzed by D. Graham and C. Goddard at the NOAA/PMEL helium laboratory in Newport, Oregon. The results have been reported to the the Marine Geo-

sciences Data System and can be accessed at: http://www.marine-geo.org/tools/search/Files.php?data_set_uid=9748.



http://www.marine-geo.org/tools/search/Files.php?data_set_uid=9748

collected over West Mata volcano in 2008 were used to estimate end-member values for the West Matahydrothermal system. The slope of a plot of 3He concentration versus 4He concentration provides an esti-mate of the 3He/4He ratio of the pure hydrothermal helium end-member (Figure 6b). In order to estimateC/3He using water-column samples, since background CO2 concentrations vary with depth, we first cor-rected the measured CO2 values by subtracting the average CO2 for background (non-plume) samples inthe region. An estimate of the C/3He ratio in the hydrothermal end-member is then determined from theslope of the plot of excess [CO2] versus [3He] (Figure 6c). This exercise generated estimates of 7.19 Ra and8.4 3 109 for 3He/4He and C/3He, respectively, in the West Mata plume, which are in good agreement withvalues of 7.25–7.30 Ra and 8–21 3 109 we obtained directly from volcanic rocks and vent fluids collectedfrom West Mata in 2009 (see Table 2).

4. Helium Isotope and C/3He Results by Geographic Location

4.1. Northeast Lau BasinHaving established the broad fields of variation in 3He/4He and C/3He associated with the basic mantle reser-voirs, i.e., MORs, volcanic arcs, and hotspots, we now use this categorization to evaluate the sources of mag-matic gases in the NE Lau Basin. As shown in Figure 2, the NE Lau Basin is complex in terms of its tectonicsand volcanism. Although not currently volcanically active, the large volcanic edifice of Niua in the eastern part

Figure 7. Plot of 3He/4He versus C/3He for rock, vent fluid, and water-column samples from the NE Lau Basin, including the NELSC, MakaVolcano, Niua, and the Mata volcanoes. Shaded ellipses indicate fields for mid-ocean ridge and arc volcanoes. Sample locations shown inFigures 1 and 2. Data are listed in Table 2.



of the basin is likely to have arc-like 3He/4He and C/3He, since it falls on the structural northern extension ofthe Tofua Arc. Somewhat to the west, the NELSC, a backarc spreading center, might be expected to haveMOR or enriched MORB affinities (e.g., based on rock compositions erupted there; Falloon et al., [2007]). At thesouthern end of the NELSC is Maka, an axial high volcanic edifice that rises above the average depth of thesouthern NELSC. In the almost 80 km between Niua and the NELSC, this easternmost part of the Lau basin ispopulated with several individual volcanic centers of unknown provenance (arc versus backarc), includingNiuatahi (the large caldera formerly called Volcano O), West Mata, East Mata, and the so-called seven northernMatas (Taha, Ua, Tolu, Fa, Nima, Ono, and Fitu) [Embley et al., 2009; Rubin et al., 2013].

Figure 7 summarizes our 3He/4He versus C/3He results for the aforementioned volcanoes in the NE LauBasin. It includes 3He/4He and C/3He values for volcanic rocks from the NELSC, Maka, and West Mata, andfor vent fluids from the NELSC, Maka, West Mata, Niua North, Niua South, Niuatahi, Mata Ua, Mata Tolu, andMata Fitu. Finally, we have included estimates derived from water-column plumes for East Mata, Niua, Niua-tahi, and Mata Ua.

The results confirm that 3He/4He and C/3He ratios are useful for distinguishing MOR versus arc affinities insuch volcanic systems. For Maka, the axial high on the southern end of the NELSC, both the volcanic rockand vent fluid results fall squarely in the MORB field. Samples from the NELSC farther north also showMORB-like signatures, although with C/3He ratios slightly higher than typical MORBs. The remaining samples

Figure 8. Plot of 3He/4He versus C/3He for rock samples from the NW Lau Spreading Center (NWLSC), the Rochambeau Rifts, and thePeggy Ridge. All results are for volcanic rocks. Results published by Hahm et al. [2012] are shown as red-filled symbols. Shaded ellipsesindicate fields for mid-ocean ridges, arc volcanoes, and hotspots. Sample locations shown in Figure 1, and results are listed in Table 2.



from the NE Lau Basin all have lower 3He/4He and higher C/3He which places them within the volcanic arcfield. As mentioned above, Niua might be expected to be the most ‘‘arc-like’’ since it is farthest to the eastand lies on the track of the Tofua Arc. Surprisingly, West Mata, the northern Matas, and even Niuatahi have3He/4He-C/3He signatures that are indistinguishable from Niua. In fact all of the volcanoes in the central andeastern part of the NE Lau Basin have very similar 3He/4He ratios that fall in a narrow range between 7.1and 7.4 Ra, at the lower end of the MORB field. There is a rather large discrepancy in the C/3He ratios for the2 vent fluids collected on Niua South (C/3He 5 1.3 and 8.9 3 1010) (see Table 2). These samples were col-lected on separate dives and at separate locations on the volcano. The Niua South sample with the higherC/3He was collected from a black smoker at 1150 m depth venting 315�C fluid, which is close to the boilingpoint at that depth, while the other Niua South sample was from a lower temperature vent. This suggeststhat phase separation may have played a role in producing the discrepancy in the Niua South C/3He ratios.

4.2. Northwest Lau Back-Arc Basin and Peggy RidgeAs mentioned in the Introduction, the Northwest Lau Backarc system is unique in that the volcanic rocks inthis region have apparently been affected by the introduction of a distinct ocean island (OIB) or hotspotcomponent, as evidence by the elevated 3He/4He ratios and the presence of mantle-type neon isotopes in

Figure 9. Plot of 3He/4He versus C/3He for samples from other back-arc spreading centers in the northern Lau Basin, including the FonualeiRift and Spreading Center (FRSC), the Mangatolu Triple Junction (MTJ), the Central Lau Spreading Center (CLSC), and the Eastern LauSpreading Center (ELSC). A result for Niuafou’ou submarine volcano is also shown. Results published by Hahm et al. [2012] are shown asred symbols. All results are for rock samples except for a single vent fluid sample from the FRSC. Shaded ellipses indicate fields for mid-ocean ridges and arc volcanoes. Sample locations shown in Figure 1, and results are listed in Table 2.



Table 3. XRF Data for Samples in Figure 10

Sample IGSN Location Nb (ppm) Ba (ppm) Ti (ppm) Ba/Nb

J2–415-R1 KHR00000H NELSC 30 290 6478 9.7J2–415-R2 KHR00000I NELSC 28 273 6114 9.8J2–415-R3 KHR00000J NELSC 28 285 6205 10.1J2–415-R4 KHR00000K NELSC 4 117 4325 26J2–415-R6A KHR00000M NELSC 27 257 6084 9.6J2–415-R7 KHR00000N NELSC 35 360 6569 10J2–415-R8 KHR00000O NELSC 35 354 6599 10J2–415-W16 KHR00000T NELSC 9 176 4962 19J2–415-R18 KHR00000U NELSC 35 365 6660 10J2–415-R20 KHR00000V NELSC 33 320 6690 9.7J2–416-R7 KHR000012 NELSC 5 110 6024 20J2–413-R2 KHR000002 W Mata 6 123 2392 20J2–413-R13 KHR000008 W Mata 7 127 2664 19J2–414-R12 KHR00000B W Mata 7 123 2674 18J2–414-R27 KHR00000E W Mata 6 94 2304 16J2–417-R2 KHR000012 W Mata 6 103 2402 16J2–417-R9 KHR000018 W Mata 6 107 2471 19J2–418-R18 KHR00001I W Mata 6 126 2694 20J2–418-R1 KHR00001A W Mata 6 137 2149 23J2–418-R5 KHR00001E W Mata 7 137 2331 21J2–418-R11 KHR00001H W Mata 7 125 2618 18J2–420-R1 KHR00001M W Mata 6 123 2160 22J2–420-R2 KHR00001N W Mata 6 124 2219 20J2–420-R5 KHR00001Q W Mata 7 118 2513 18J2–420-R16 KHR000020 W Mata 6 116 2465 20J2–420-R17 KHR000021 W Mata 6 105 2240 18KM1024-D12-Rock 1 KHR00004G W Mata 5 119 2057 22KM1024-D15-Rock 1 KHR000051 W Mata 6 110 2300 18KM1024-D15-Rock 4 KHR000054 W Mata 7 126 2422 19RR1211 Q332-R03 KHR000107 W Mata 7 133 2663 19RR1211 Q332-R03 KHR000109 W Mata 7 143 2891 20KM1024-D14-Rock 2 KHR00004S E Mata 9 184 2481 21KM1024-D14-Rock 7 KHR00004X E Mata 9 171 2542 19KM1129a-D2-Rock 2 KHR00013c E Mata 14 273 2478 19KM1129a-D2-Rock 4 KHR00013f E Mata 13 262 2358 20KM1024-D16-Rock 1 KHR000059 Mata Taha 12 144 2481 12KM1024-D16-Rock 3 KHR00005B Mata Taha 12 139 2572 11KM1024-D16-Rock 4 KHR00005C Mata Taha 12 143 2482 12KM1024-D16-Rock 6 KHR00005E Mata Taha 12 145 2542 12KM1024-D18-Rock 1 KHR00005G Mata Ua 7 138 2299 19KM1129a-D10-Rock 1 KHR000166 Mata Ua 4 123 1877 31KM1129a-D10-Rock 2 KHR000167 Mata Ua 6 216 3966 37KM1129a-D10-Rock 5 KHR00016a Mata Ua 6 106 2158 19KM1129a-D10-Rock 6 KHR00016b Mata Ua 4 118 1938 30RR1211 Q325-R01 KHR000088 Mata Ua 7 148 2419 20RR1211 Q325-R02 KHR000089 Mata Ua 10 283 4752 29KM1024-D20-R01 KHR00005O Mata Tolu 9 145 2422 17KM1129a D07-Rock 1 KHR00015a Mata Tolu 13 222 2300 16KM1129a D07-Rock 2 KHR00015b Mata Tolu 13 201 2360 16KM1129a-D8-Rock 1 KHR00015c Mata Tolu 14 224 2209 16KM1129a-D8-Rock 7 KHR000162 Mata Tolu 8 141 2482 17KM1129a-D9-Rock 1 KHR000164 Mata Tolu 18 245 4119 13RR1211 Q331-R16 KHR000103 Mata Tolu 8 130 2649 16KM1024-D21-Rock 1 KHR00005T Mata Fa 17 187 5395 11KM1024-D21-Rock 2 KHR00005W Mata Fa 17 196 5330 12KM1024-D21-Rock 4 KHR00005U Mata Fa 17 179 5350 10KM1024-D22-Rock 1 KHR00005X Mata Ono 7 124 2664 17KM1024-D22-Rock 2 KHR00005Y Mata Ono 5 104 1876 20KM1024-D23-Rock 1 KHR000063 Mata Fitu 6 108 1906 17KM1024-D23-Rock 4 KHR000066 Mata Fitu 6 101 1997 17RR1211 Q329-R01 KHR00009B Mata Fitu 8 146 3327 17KM1129a-D4-Rock 1 KHR000147 Niua 18 362 1924 21KM1129a-D4-Rock 5 KHR00014b Niua 17 359 1928 21KM1024-D08-Rock 10 KHR00003V Niuatahi 6 237 3504 41KM1024-D24-Rock 4 KHR00006D Niuatahi 6 235 3478 39KM1024-D24-Rock 6 KHR00006F Niuatahi 6 233 3443 41



the erupted basalts [Poreda and Craig, 1992; Lupton et al., 2009, 2012; Hahm et al., 2012]. One possibility isthat mantle material from the Samoan hotspot has intruded into the northern Lau Basin through a tear inthe downgoing slab and has been incorporated into the seafloor volcanic rocks erupted along this backarcspreading system [Natland, 1980; Turner and Hawkesworth, 1998; Lupton et al., 2009].

Figure 8 shows results for the Rochambeau Rifts and NWLSC plotted on the R/Ra versus C/3He graph.These results are all from the analysis of seafloor basalts, and they include samples published by Hahmet al. [2012] in addition to our measurements. Two basalt samples from the Peggy Ridge reported byHahm et al. [2012] are also shown (see Figure 1 for location). As mentioned in section 2, the Hahm et al.

Figure 10. Plots of Ba-Nb-Ti chemistry for Lau Basin volcanic rocks, including the NE Lau Spreading Center (NELSC) as blue diamonds, Niua as purple circles, Niuatahi as orange circles,and the Mata volcanoes green triangles except for W. Mata, shown as green squares. Data are listed in Table 3. Literature data are shown by fields: orange is Niuatahi [Park et al., 2015];light green is the Tofua arc [Caulfield et al., 2012; Keller et al., 2008]; gray-blue is Fonualei spreading center [Keller et al., 2008; Escrig et al., 2012]; light blue is the Mangatolu Triple Junction(MTJ) [Keller et al., 2008; Tian et al., 2011]; and pink is various NW Lau spreading centers and islands, NWLSC, Rochambeau Rifts, Peggy Ridge, and Niuafo’ou Is.) [Tian et al., 2011]. Shadedfields in some plots are for global MORB [Rubin and Sinton, 2007; Jenner and O’Neill, 2012; Gale et al., 2013], arcs/back-arcs [Pearce et al., 2005; Pearce and Stearn, 2006], and Samoa [Jack-son et al., 2014]. (a) Whole rock Nb (ppm) versus Ti (ppm) shows a range of mantle wedge source depletion/enrichment conditions using mildly incompatible elements that are not influ-enced by subduction zone inputs. Note two discrete trends in the new data (constant Ti at variable Nb pointing to Niua on the Tofua Arc, and highly variable Nb with moderatelyvariable Ti, pointing to the NELSC data). Literature data form a third, less steep trend, mostly varying in Ti. (b) Ba/Ti versus Nb/Ti diagram after Pearce and Stern [2006], using Ti as a proxyfor Yb (see text). Note location of most Mata data between the three gray fields. (c) and (d) Ba/Nb versus 1/Nb at two different scales (Figure 10c is optimized to view the new data andFigure 10d is optimized for the regional data). Lavender outline field in Figure 10d is Eastern Lau Spreading Center and Valu Fa ridge data [Escrig et al., 2009]. Note location of most Matadata between the three gray fields in Figure 10c and the locus of all new data in Figure 10d (lower left corner) at very high Nb compared to regional arc and back-arc data elsewhere inthe basin.



[2012] results appear to have slightly higher C/3He values on the average versus our results, likely due todifferent methods of measuring CO2 content. Despite this difference, the results show a consistent patternin which 3He/4He ratios are elevated above the MORB field similar to that for hotspots (see Figure 5b).Surprisingly, the samples from the Rochambeau Rifts and NWLSC have C/3He values in the range of 108

up to 2 3 109, similar to MORBs and lower than typical hotspots. Although Jenner et al. [2012] found thatthe NWLSC basalts are distinct in having elevated Cu compared to the Rochambeau Rift basalts, we donot see a difference in 3He/4He-C/3He, i.e., the NWLSC and RR samples results occupy overlapping fields inFigure 8. Our results suggest that the NWLSC-RR basalts represent mixing between DDM and a hotspot-type end-member with elevated 3He/4He that is distinct from the 3He-rich hotspots such as Hawaii, Yel-lowstone, and Reunion. This hotspot end-member would then appear to have 3He/4He of >30 Ra and arather low C/3He of �3 3 108 (Figure 8). While several different mantle components are present inSamoan lavas [Jackson et al., 2014], recent Hf-Nd isotope measurements indicate that only DMM and a sin-gle FOZO Samoan component are present in the NWLSC-RR sample suite [Nebel and Arculus, in press]. Lit-tle is known about C/3He in Samoan lavas, and additional measurements will help to determine whetherthe hotspot component in the NW Lau Back-arc is from Samoa or instead represents a new mantle plumeend-member. The two samples from the Peggy Ridge have elevated 3He/4He (10.4 and 10.9 Ra) indicatingthat this 3He-ich component has penetrated southward to the Peggy Ridge. This conclusion is further sup-ported by the fact that the NWLSC samples are virtually indistinguishable from the Rochambeau Rift sam-ples (see Figure 1).

4.3. Other Back-Arc Spreading Centers in the Northern Lau BasinHelium isotope and C/3He ratios are also available from other localities in the Lau Basin, including the Cen-tral Lau Spreading Center (CLSC), the Eastern Lau Spreading Center (ELSC), the Mangatolu Triple Junction(MTJ), and the Fonualei Rift and Spreading Center (FRSC) (see our data in Table 2 and Hahm et al. [2012]).These are all back-arc spreading centers that might be expected to have 3He/4He and C/3He values that fallin the MORB field, and they do. Hahm et al. [2012] also analyzed a rock sample from Niuafou’ou Volcano, asubmarine volcano in the northern Lau Basin situated northeast of Niuafou’ou Island proper (Figure 1). Asshown in Figure 9, the samples from the CLSC and ELSC all fall generally within the MORB field on the plotof 3He/4He versus C/3He, verifying that these typical back-arc spreading centers have MORB-likecharacteristics.

The other back-arc samples deviate slightly from this. Samples from the Fonualei Rift have slightly lower3He/4He and somewhat higher C/3He with the result that they fall in the region between typical MORBs andsubmarine arc volcanoes. This includes the hydrothermal vent fluid collected on the Fonualei Rift, whichhas a 3He/4He versus C/3He signature similar to that of the Fonualei Rift rocks. Keller et al. [2008] docu-mented that the Fonualei Rift has a strong subduction overprint, and thus it is not surprising that ourFonualei samples show some arc influence.

Although the MTJ can be considered to be the northward extension of the FRSC, the two rock samples col-lected along the southern leg of the triple junction fall on yet another part of the plot. These MTJ rock sam-ples have MORB-like C/3He values but 3He/4He ratios of 7.2–7.3 Ra, at the low end of the MORB field. Ourresults are thus consistent with those of Keller et al. [2008], who found that the MTJ has trace element signa-tures characteristic of back-arc basin basalts, in contrast to the samples from the Fonualei Rift. The samplefrom Niuafou’ou Volcano reported by Hahm et al. [2012] had 3He/4He of 7.7 Ra, placing it at the lower endof the MORB field. Other studies have found 3He/4He ratios slightly lower than typical MORBs in this region.Hilton et al. [1993] reported 6.9 Ra for a rock sample on the northern leg of the MTJ, while Honda et al.[1993] reported 6.2 and 7.3 Ra for two rock samples collected right at the triple junction.

5. Ba-Nb-Ti Results

Ba-Nb-Ti analyses of NE Lau Basin and N. Tofua arc lavas were also conducted and compared to regional lit-erature data as a means to compare more ‘‘traditional’’ tracers of mantle wedge composition (i.e., trace ele-ment signatures that are unrelated to the subduction zone and instead reflect relative source enrichmentand depletion and those that reflect fluxes off the subducting slab). Figure 10 and Table 3 show the resultsof these analyses. A more complete suite of major and trace element abundances in these lavas have beenanalyzed and will be presented elsewhere, but are not relevant to the discussion presented here. We focus



now on Ba-Nb-Ti chemistry because, much like the C-He systematics, they are sensitive indicators of the tec-tonic settings and magma sources that are relevant to this study.

As detailed in Pearce and Stern [2006], variations of moderately incompatible elements (e.g., Ti, Y, Yb) aswell as more incompatible high field strength (HFS) elements (Nb, Ta, Hf) in suprasubduction zone lavas pri-marily track variations in the mantle source that are unrelated to subduction zone inputs. Thus they areideal for investigating variations in depleted (e.g., MORB-like) and enriched (e.g., plume-sourced) compo-nents in the mantle wedge that are reflected in magmas produced from it. Ba, on the other hand, is a fluidmobile incompatible lithophile element that is arguably the most sensitive indicator of subduction fluidaddition to the mantle wedge [Pearce and Stern, 2006]. Very high Ba/Nb is characteristic of subduction fluidinputs to arc and back arc volcanoes at low Nb abundance [e.g., Pearce and Stern, 2006] and significantlylower but similar Ba/Nb occurs in MORB [e.g., Rubin and Sinton, 2007; Jenner and O’Neill, 2012; Gale et al.,2013] and ocean island basalts, OIB (e.g., Samoa) [Jackson et al., 2014]. MORB and OIB are distinguished bylow and high Nb abundances, respectively.

Nb and Ti variations in Lau basin lavas (Figure 10a) show the complexity of mantle compositions there: theMata volcanoes form two discrete trends, from low Nb and Ti compositions common in many subductionszones to (a) high Nb/low Ti values like those found at Niua (the northernmost Tofua arc volcano) and to (b)high Nb and higher Ti, values like those found on the NELSC. NELSC lavas also have Ba-Nb characteristicsthat are indistinguishable from Samoa (see Figure 10c). West Mata sits at the low Nb-Ti intersection of thesetwo trends but with significant internal variability. These two trends, plus a third, shallower trend forregional literature data elsewhere in the basin and on the arc that have more variable Ti and relatively lessvariable Nb, collectively indicate that there are multiple sources and conditions of enrichment and deple-tion in the basin that affect trace element abundances in the lavas, as has been discussed by multipleauthors [e.g., Hergt and Hawkesworth, 1994; Hawkins, 1995; Falloon et al., 2007; Escrig et al., 2009, 2012; Tianet al., 2011; Caulfield, et al., 2012; Nebel and Arculus, 2015].

Subduction zone fluxes now and in the past have superimposed compositional variations on the afore-mentioned mantle wedge composition. Pearce and Stern [2006] use a plot of Ba/Yb versus Nb/Yb to distin-guish the relative proportion of subducted Ba to mantle Ba in the mantle source, using the fact that mostMORB and hotspot lavas form an array of Ba/Yb and Ba/Yb values that reflect limited Ba/Nb variationseven at wide variations in Ba, Nb, and Yb concentrations. Lavas at subduction zones (e.g., right on the arc)form a parallel array at higher Ba/Nb. Many backarc lavas sit between the two arrays, reflecting the rela-tive amount of subduction input to the Ba budget of the source [Pearce and Stern, 2006]. Figure 10b is ananalogous diagram using Ti as a proxy for Yb (we do not have Yb data). Y can also be used as a Yb proxyand the resulting plot is similar (not shown), but Y data are not available for some of the literature sour-ces, so we have used Ti instead. The one additional consideration for using Ti in this manner is to evaluateif oxide mineral fractionation has affected Ti abundances (e.g., in very differentiated magmas). This pro-cess does not affect Ti in the samples discussed here, as indicated by positive relationships of Ti vs. Si,and Ti vs. Fe (not shown), both of which instead indicate that Ti is incompatible, just like Yb and Y. Fieldsfor arcs, MORB, the nearby Samoan hotspot, and various geographic domains of the Lau Basin and Tofuaarc from the literature are also shown (see figure caption for data sources). Niua and Niuatahi sit in thearc field (as does Tofua data from the literature), consistent with their location on or very near the arc aswell as with their C-He systematics (Figure 7). The NELSC sits on the extension of the MORB-OIB array.Data for the Mata volcanoes span the entire range between arcs and the MORB-OIB array and overlapwith the bottom of the arc array. The substantial range within the volcanoes on Figures 10a and 10b,even within a single volcano, (West Mata, shown as green squares, which is best-sampled of the Matasand second closest to the arc), indicates that the subduction input to their mantle sources is variablyexpressed because it overprints a variably enriched/depleted mantle wedge. The offset of the Mata volca-noes and even of Niuatahi to higher Nb/Ti than other nearby parts of the Lau basin (e.g., the FonualeiSpreading Center, Spreading Centers in the NW Lau Basin, and the Mangatolu Triple Junction) indicatethat the NE Lau basin is significantly more enriched in HFS elements in the mantle source, despite havingvery low Ti (which usually indicates very depleted sources). The complexity of the source chemistry isbeyond the scope of this discussion, but illustrates the potential difficulties of using traditional subductioninput tracers like Ba/Nb [Pearce and Stern, 2006] to deduce subduction inputs, and illustrates the utility ofapplying C-He systematics to the problem, which tends to see through some of this source complexity.



The Ba/Nb versus 1/Nb plots (Figures 10c and 10d) show the net effect of the competing processes ofwedge depletion/enrichment and subduction inputs for NE Lau basin lavas. The data lie along a broad posi-tive correlation in Figure 10c and sit between the fields for the various tectonic settings, further illustratingthat these data do not conform as simply to tectonic discriminants as do the C-He data. The array of Ba-Nbdata has the shape of a mixing relationship, between the most subduction-fluid enhanced samples (Niua-tahi and several northern Mata lavas) and the trace element enriched mantle source sampled by lavas onthe southern NELSC [Rubin et al., 2009], which sit in the Samoa field rather than the MORB field. Niua,although squarely on the arc physically, has too much Nb and too low Ba/Nb for the global systematics tofall in the arc field on this plot, as do all of the Mata volcanoes, including West Mata [Rubin et al., 2009; Res-ing et al., 2011]. By contrast, the Tofua arc in general and both the Fonualei and Eastern Lau Spreading Cen-ters (both further south in the basin) do largely overlap the arc fields (Figure 10d), by virtue of their muchlower Nb and relatively higher Ba than the NE Lau basin lavas. Lavas from the enriched sources found in theNW Lau basin sit below the arc field and overlap with the MORB and hotspot fields by virtue of much higherNb relative to the Tofua arc and much lower Ba/Nb compared to the NE Lau lavas. The "ambiguous" posi-tion of the NE Lau lavas thus reflect the complex mantle history in the young and rapidly evolving NE Lauregion [Embley et al., 2009], which samples mantle of locally variable depletion and enrichment, overprintedby subduction zone inputs [e.g., Escrig et al., 2012; Falloon et al., 2007; Rubin et al., 2013]. Interestingly, theC-He systematics are able to see through this mantle source complexity in the NE Lau and more clearly indi-cate the subduction zone inputs.

6. Summary and Conclusions

An examination of representative samples from various provinces has shown that samples from mid-oceanridges, arc volcanoes, and hotspot volcanoes define separate nonoverlapping fields in a plot of 3He/4He ver-sus C/3He. Arc volcanoes have lower 3He/4He and higher C/3He compared to MOR samples, and hotspotsdefine a third field with elevated 3He/4He and C/3He spanning both the MOR and arc values.

Having defined the general nature of 3He/4He versus C/3He variations, we applied this approach to the sub-marine volcanic systems of the Lau Basin. The NE Lau Basin is of particular interest because multiple ‘‘rear-arc’’ volcanoes of unknown provenance occur between the NELSC, a typical back-arc system, and Niua, avolcano located on the northward extension of the Tofua Arc proper. Although samples from Niua and theNELSC fall in the expected 3He/4He versus C/3He fields, the rear-arc volcanoes, including Niuatahi and all ofthe Mata volcanoes, all fall in the arc field. This suggests that arc-type melts are being recirculated and aredominating both the arc and the rear-arc regions of the NE Lau Basin.

Most of the other back-arc systems, including the MTJ, the CLSC and the ELSC, properly fall in the MOR field.The Fonualei Rift samples have slightly lower 3He/4He and slightly higher C/3He, with the result that thesesamples fall between the MOR and arc fields, indicating some arc affinity, consistent with the arc overprintevident in the trace metals. Finally, samples from the NWLSC and the Rochambeau Rifts, which are thoughtto have a Samoan hotspot component, do not exactly plot in the hotspot field. Instead, they have elevated3He/4He but C/3He similar to MORBs, lower than most hotspot volcanic systems. Our data suggest that theNWLSC-RR lavas may be the results of mixing between DMM and a distinct hotspot end-member character-ized by 3He/4He of� 30 Ra and C/3He of �3 3 108.

Unlike C-He data, most Ba-Nb data in general lie along a broad inverse correlation that sits between the fieldsdefined by volcanoes in ridge, arc, and hotspot tectonic settings elsewhere, despite the fact that Ba/Nb is con-sidered to be one of the most sensitive indicators of subduction flux inputs to arc and backarc magmas. Thisindicates that these data do not conform as simply to tectonic environment as do the C-He data, because ofa complex pre-subduction mantle history that is recorded in Nb-Ti variations. Thus the dispersion in the Ba-Nb-Ti data is derived from a complex regional mantle history involving variable depletion, enrichment, andsubduction inputs. Interestingly, the C-He systematics are able to see through this mantle source complexity.

These results show that helium isotope ratios, in conjunction with C/3He, can be used to differentiatebetween mid-ocean ridge, arc, and hotspot affinities in submarine volcanic systems. The volcanic systemsof the northern Lau Basin exhibit a complex mixture of all three of these affinities, indicating that the mantlemixing regime there is very complex.



ReferencesArculus, R. J. (2004), Voyage Summary SS11/2004 Northern Lau Vents Expedition. [Available at http://www.marine.csiro.au/nationalfacility/

voyagedocs/2004/Summary_SS11-2004.pdf.]Arculus, R. J. (2008), Voyage Summary SS07/2008 Northern Lau Vents Expedition. [Available at http://www.marine.csiro.au/nationalfacility/

voyagedocs/2008/MNF-SS07-08_sum.pdf.]Baker, E. T., J. E. Lupton, J. A. Resing, T. Baumberger, M. D. Lilley, S. L. Walker, and K. H. Rubin (2011), Unique event plumes from a 2008

eruption on the Northeast Lau Spreading Center, Geochem. Geophys. Geosyst., 12, Q0AF02, doi:10.1029/2011GC003725.Bevis, M. et al. (1995), Geodetic observations of very rapid convergence and back-arc extension at the Tonga arc, Nature, 374, 249–251.Bird, P. (2003), An updated digital model of plate boundaries, Geochem. Geophys. Geosyst., 4(3), 1027, doi:10.1029/2001GC000252.Burnard, P., D. Graham, and K Farley (2004), Fractionation of noble gases (He, Ar) during MORB Mantle melting: A case study on the South-

east Indian Ridge, Earth Planet. Sci. Lett., 227, 457–472.Caulfield, J., S. Turner, R. Arculus, C Dale, F. Jenner, J. Pearce, C. Macpherson, and H. Handley (2012), Mantle flow, volatiles, slab-surface tem-

peratures and melting dynamics in the north Tonga arc: Lau back-arc basin, J. Geophys. Res., 117, B11209, doi:10.1029/2012JB009526.Craig, H., and J. E. Lupton (1976), Primordial neon, helium, and hydrogen in oceanic basalts, Earth Planet. Sci. Lett., 31, 369–385.Craig, H., J. E. Lupton and Y. Horibe (1978a), A mantle helium component in circum-Pacific volcanic gases, Hakone, the Marianas, and Mt.

Lassen, in Terrestrial rare gases, proceedings of the U.S.-Japan Seminar on Rare Gas Abundance and Isotopic Constraints on the Origin andEvolution of Earth’s Atmosphere, edited by E. C. Alexander, Jr. and M. Ozima, pp. 3–6, Adv. Earth Planet. Sci., 3, 3–16.

Craig, H., J. E. Lupton, J. A. Welhan, and R. Poreda (1978b), Helium isotope ratios in Yellowstone and Lassen Park volcanic gases, Geophys.Res. Lett., 5, 897–900.

DOE (1994), A Handbook of Methods for the Analysis of the Various Parameters of the Carbon Dioxide System in Sea Water, version 2, editedby A. G. Dickson, and C. Goyet, ORNL/CDIAC-74.

Embley, R. W., S. G. Merle, J. E. Lupton, J. Resing, E. T. Baker, M. D. Lilley, R. J. Arculus, and P. V. Crowhurst (2009), Extensive and diverse sub-marine volcanism and hydrothermal activity in the NE Lau Basin, EOS Trans. AGU, 90(52), Fall Meet. Suppl., Abstract V51D-1719.

Escrig, S., A. B�ezos, S. L. Goldstein, C. H. Langmuir, and P. J. Michael (2009), Mantle source variations beneath the Eastern Lau SpreadingCenter and the nature of subduction components in the Lau basin: Tonga arc system, Geochem. Geophys. Geosyst., 10, Q04014, doi:10.2019/2008GC00281.

Escrig, S., A. B�ezos, C. H. Langmuir, P. J. Michael, and R. Arculus (2012), Characterizing the effect of mantle source, subduction input andmelting in the Fonualei Spreading Center, Lau Basin: Constraints on the origin of the boninitic signature of the back-arc lavas, Geochem.Geophys. Geosyst., 13, Q10008, doi:10.1029/2012GC004130.

Exley, R. A., D. P. Mattey, D. A. Clague, and C. T. Pillinger (1986), Carbon isotope systematics of a mantle ’hotspot’: A comparison of LoihiSeamount and MORB glasses, Earth Planet. Sci. Lett., 78, 189–199.

Falloon, T. J., L. V. Danyushevsky, T. J. Crawford, R. Maas, J. D. Woodhead, S. M. Eggins, S. H. Bloomer, D. J. Wright, S. K. Zlobin, and A. R.Stacey (2007), Multiple mantle plume components involved in the petrogenesis of subduction-related lavas from the northern termina-tion of the Tonga Arc and northern Lau Basin: Evidence from the geochemistry of arc and backarc submarine volcanics, Geochem. Geo-phys. Geosyst., 8, Q09003, doi:10.1029/2007GC001619.

Farley, K. A., J. H. Natlund, and H. Craig (1992), Binary mixing of enriched and undegassed (primitive?) mantle components (He, Sr, Nd, Pb)in Samoan lavas, Earth Planet. Sci. Lett., 111, 183–199.

Gale A., C. A. Dalton, C. H. Langmuir, Y. Su, and J.-G. Schilling (2013), The mean composition of ocean ridge basalts, Geochem. Geophys. Geo-syst., 14, 489–518, doi:10.1029/2012GC004334.

German, C. R., E. T. Baker, D. P. Connelly, J. E. Lupton, J. Resing, R. D. Prien, S. L. Walker, H. N. Edmonds, and C. H. Langmuir (2006), Hydro-thermal exploration of the Fonualei Rift and Spreading Center and the Northeast Lau Spreading Center, Geochem. Geophys. Geosyst., 7,Q11022, doi:10.1029/2006GC001324.

Graham, D. W. (2002), Noble gas isotope geochemistry of mid-ocean ridge and ocean island basalts; characterization of mantle source res-ervoirs, in Noble Gases in Geochemistry and Cosmochemistry, Reviews in Mineralogy and Geochemistry, edited by D. Porcelli, R. Wieler, andC. Ballentine, pp. 247–318, Mineral. Soc. Am., Washington, D. C.

Hahm, D., D. R. Hilton, P. R. Castillo, J. W. Hawkins, B. B. Hanan, and E. H. Hauri (2012), An overview of the volatile systematics of the LauBasin: Resolving the effects of source variation, magmatic degassing and crustal contamination, Geochim. Cosmochim. Acta, 85, 88–113.

Hawkins, J. W. (1995), The geology of the Lau Basin, in Backarc Basins: Tectonics and Magmatism, edited by B. Taylor, pp. 63–138, Plenum,N. Y.

Hergt, J. M., and C. J. Hawkesworth (1994), Pb-, Sr-, and Nd-isotopic evolution of the Lau Basin: Implications for mantle dynamics duringbackarc opening, Proc. Ocean Drill. Program Sci. Results, 1(35), 505–517.

Hoefs, J. (1980), Stable Isotope Geochemistry, 208 pp., Springer, N. Y.Hilton, D.R., K. Hammerschmidt, G. Loock, and H. Friedrichsen (1993), Helium and argon isotope systematics of the central Lau Basin and

Valu Fa Ridge: Evidence of crust/mantle interactions in a back-arc basin, Geochim. Cosmochim. Acta, 57, 2819–2841.Hilton, D. R., G. M. McMurtry, and F. Goff (1998), Large variations in vent fluid CO2/3He ratios signal rapid changes in magma chemistry at

Loihi seamount, Hawaii, Nature, 396, 359–362.Hilton, D. R., T. P. Fischer, and B. Marty (2002), Noble gases and volatile recycling at subduction zones, Rev. Mineral. Geochem., 47, 319–370.Honda, M., D. B. Patterson, I. McDougall, and T. J. Falloon (1993), Noble gases in submarine pillow basalt glasses from the Lau Basin: Detec-

tion of a solar component in backarc basin basalts, Earth Planet. Sci. Lett., 120, 135–148.Jackson, M. G., M. D. Kurz, S. R. Hart, and R. K. Workman (2007), New Samoan lavas from Ofu Island reveal a hemispherically heterogeneous

high 3He/4He mantle, Earth Planet. Sci. Lett., 264, 360–374.Jackson, M. G., S. R. Hart, J. G. Konter, A. A. P. Koppers, H. Staudigel, M. D. Kurz, J. Blusztajn, and J. M. Sinton (2010), Samoan hot spot track

on a ‘‘hot spot highway’’: Implications for mantle plumes and a deep Samoan mantle source, Geochem. Geophys. Geosyst., 11, Q12009,doi:10.1029/2010GC003232.

Jackson, M. G., S. R. Hart, J. G. Konter, M. D. Kurz, J. Blusztajn, and K. A. Farley (2014), Helium and lead isotopes reveal the geochemicalgeometry of the Samoan plume, Nature, 514, doi:10.1038/nature13794.

Jenner, F. E., and H. St. C. O’Neill (2012), Analysis of 60 elements in 616 ocean floor basaltic glasses, Geochem. Geophys. Geosyst., 13,Q02005, doi:10.1029/2011GC004009.

Jenner, F. E., R. J. Arculus, J. A. Mavrogenes, N. J. Dyrix, O. Nebel, and E. H. Hauri (2012), Chalcophile element systematics in volcanic glassesfrom the northwestern Lau Basin, Geochem. Geophys. Geosyst., 13, Q06014, doi:10.1029/2012GC004088.

Keller, N. S., R. J. Arculus, J. Hermann, and S. Richards (2008), Submarine back-arc lava with arc signature: Fonualei Spreading Center, north-east Lau Basin, Tonga, J. Geophys. Res., 113, B08S07, doi:10.1029/2007JB005451.

AcknowledgmentsWe thank the captains and crews ofthe R/V Southern Surveyor, the R/VThomas Thompson, the R/V Kilo Moana,and the R/V Roger Revelle for theirassistance in the collection of thesamples for this study. The teams forthe ROPOS and Jason ROVs made itpossible to obtain high-quality ventfluid samples. L. Evans and R. Greenecollected samples at sea and carriedout the laboratory analytical work. Wethank S. Merle for generating thecruise reports for the TN-234, KM1024,and RR1211 expeditions and forassistance with the figures, andS. Glancy for assistance with samplepreparation for XRF analysis. We thankM. Jackson, J. Pearce, A. Colin, and oneother reviewer for constructive reviewsof the manuscript. Funding wassupplied by the NOAA Vents (nowEarth-Ocean Interactions) Program andby the Joint Institute for the Study ofthe Atmosphere and Ocean (JISAO)under NOAA Cooperative AgreementNA10OAR4320148. This is PMELpublication No. 4224, JISAOcontribution No. 2375 and SOESTcontribution 9287.



http://www.marine.csiro.au/nationalfacility/voyagedocs/2004/Summary_SS11-2004.pdf

http://www.marine.csiro.au/nationalfacility/voyagedocs/2004/Summary_SS11-2004.pdf

http://www.marine.csiro.au/nationalfacility/voyagedocs/2008/MNF-SS07-08_sum.pdf

http://www.marine.csiro.au/nationalfacility/voyagedocs/2008/MNF-SS07-08_sum.pdf

http://dx.doi.org/10.1029/2011GC003725

http://dx.doi.org/10.1029/2001GC000252

http://dx.doi.org/10.1029/2012JB009526

http://dx.doi.org/10.2019/2008GC00281

http://dx.doi.org/10.1029/2012GC004130

http://dx.doi.org/10.1029/2007GC001619

http://dx.doi.org/10.1029/2012GC004334

http://dx.doi.org/10.1029/2006GC001324

http://dx.doi.org/10.1029/2010GC003232

http://dx.doi.org/10.1038/nature13794

http://dx.doi.org/10.1029/2011GC004009

http://dx.doi.org/10.1029/2012GC004088

http://dx.doi.org/10.1029/2007JB005451

Kelley, D. S., M. D. Lilley, and G. L. Fruh-Green (2004), Volatiles in submarine environments: food for life, in The Subseafloor Biosphere at Mid-Ocean Ridges, Geophys. Monogr. Ser., 144, edited by W. Wilcock et al., pp. 167–190, AGU, Washington, D. C.

Kim, J., S. K. Son, J. W. Son, K. H. Kim, W. J. Shim, C. H. Kim, and K. Y. Lee (2009), Venting sites along the Fonualei and Northeast Lau spread-ing centers and evidence of hydrothermal activity at an off-axis caldera in the northeastern Lau Basin, Geochem. J., 43, 1–13.

Kurz, M. D., W. J. Jenkins, S. R. Hart, and D. Clague (1983), Helium isotopic variations in volcanic rocks from Loihi Seamount and the Islandof Hawaii, Earth Planet. Sci. Lett., 66, 388–406.

Lilley, M. D., D. A. Butterfield, J. E. Lupton, and E. J. Olson (2003), Magmatic events can produce rapid changes in hydrothermal vent chem-istry, Nature, 422, 878–881.

Lupton, J. E. (1990), Water-column hydrothermal plumes on the Juan de Fuca Ridge, J. Geophys. Res., 95, 12,829–12,842.Lupton, J., and L. Evans (2004), The atmospheric helium isotope ratio: Is it changing?, Geophys. Res. Lett., 31, L13101, doi:10.1029/

2004GL020041.Lupton, J., et al. (2006), Submarine venting of liquid carbon dioxide on a Mariana Arc volcano, Geochem. Geophys. Geosyst., 7, Q08007, doi:

10.1029/2005GC001152.Lupton, J., M. Lilley, D. Butterfield, L. Evans, R. Embley, G. Massoth, B. Christenson, K.-I. Nakamura, and M. Schmidt (2008), Venting of a sepa-

rate CO2-rich gas phase from submarine arc volcanoes: Examples from the Mariana and Tonga-Kermadec arcs, J. Geophys. Res., 113,B08S12, doi:10.1029/2007JB005467.

Lupton, J. E., R. J. Arculus, R. R. Greene, L. J. Evans, and C. I. Goddard (2009), Helium isotope variations in seafloor basalts from the North-west Lau Backarc Basin: Mapping the influence of the Samoan hotspot, Geophys. Res. Lett., 36, L17313, doi:10.1029/2009GL039468.

Lupton, J. E., R. J. Arculus, L. J. Evans, and D. W. Graham (2012), Mantle hotspot neon in basalts from the Northwest Lau Back-arc Basin, Geo-phys. Res. Lett., 39, L08308, doi:10.1029/2012GL051201.

Martinez, F., and B. Taylor (2006), Modes of crustal accretion in back-arc basins: Inferences from the Lau Basin, in Back-Arc Spreading Sys-tems: Geological, Biological, Chemical and Physical Interactions, Geophys. Monogr. Ser., 166, edited by D. M. Christie et al., pp. 5–30, AGU,Washington, D. C.

Marty, B., and A. Jambon (1987), C/3He in volatile fluxes from the solid Earth: Implications for carbon geodynamics, Earth Planet. Sci. Lett.,83, 16–26.

Marty, B., and I. Tolstikhin (1998), CO2 fluxes from mid-ocean ridges, arc and plumes, Chem. Geol., 145, 233–248.Marty, B., and L. Zimmermann (1999), Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: Assessment of shallow-level fractionation and char-

acterization of source composition, Geochim. Cosmochim. Acta, 63, 3619–3633.Natland, J. (1980), The progression of volcanism in the Samoan linear volcanic chain, Am. J. Sci., 280, 709–735.Naughton, J. J., J. H. Lee, D. Keeling, J. B. Finlayson and G. Dority (1973), Helium flux from the Earth’s mantle as estimated from Hawaiian

fumarolic degassing, Science, 180, 55–57.Nebel, O., & Arculus R.J., (2015), Selective ingress of a Samoan plume component into the northern Lau backarc basin. Nature Communica-

tions, DOI: 10.1038/ncomms7554.Nebel, O., R. J. Arculus, P. A. Sossi, F. E. Jenner, and T. H. E. Whan (2013), Iron isotopic evidence for convective resurfacing of recycled arc-

front mantle beneath back-arc basins, Geophys. Res. Lett., 40, 5849–5853, doi:10.1002/2013GL057976.Park, J.-W., I. H. Campbell, J. Kim, and J.-W. Moon (2015), The role of late sulfide saturation in the formation of a Cu- and Au-rich magma:

Insights from the platinum group element geochemistry of Niuatahi–Motutahi lavas, Tonga rear arc, J. Petrol., 56, 59–81, doi:10.1093/petrology/egu071.

Pearce, J., and R. Stern (2006), Origin of back-arc basin magmas: Trace element and isotope perspectives, in Back-arc Spreading Systems:Geological Biological, Chemical, and Physical Interactions, Geophys. Monogr. Ser. 166, edited by D. M. Christie et al., pp. 63–86, AGU, Wash-ington, D. C.

Pearce, J. A., R. J. Stern, S. H. Bloomer, and P. Fryer (2005), Geochemical mapping of the Mariana arc-basin system: Implications for thenature and distribution of subduction components, Geochem. Geophys. Geosyst., 6, Q07006, doi:10.1029/2004GC000895.

Pearce, J. A., P. D. Kempton, and J. B. Gill (2007), Hf-Nd evidence for the origin and distribution of mantle domains in the SW Pacific, EarthPlanet. Sci. Lett., 260, 98–114.

Pelletier, B., S. Calmant, and R. Pillet (1998), Current tectonics of the Tonga-New Hebrides region, Earth Planet. Sci. Lett., 164, 263–276, doi:10.1016/S0012-821X(98)00212-X.

Pelletier, B., et al. (2001), Newly identified segments of the Pacific: Australia plate boundary along the North Fiji transform zone, EarthPlanet. Sci. Lett., 193, 347–358, doi:10.1016/S0012-821X(01)00522-2.

Peto, M. K., S. Mukhopadhyay, and K. A. Kelley (2013), Heterogeneities from the first 100 million years recorded in deep mantle noble gasesfrom the Northern Lau Back-arc Basin, Earth Planet. Sci. Lett., 369, 13–23.

Poreda, R., and H. Craig (1989), Helium isotopes in circum-Pacific volcanic arcs, Nature, 338, 473–478.Poreda, R. J., and H. Craig (1992), He and Sr isotopes in the Lau Basin mantle: Depleted and primitive mantle components, Earth Planet. Sci.

Lett., 113, 487–493.Regelous, M., S. Turner, R. J. Falloon, P. Taylor, J. Gamble, and T Green (2008), Mantle dynamics and mantle melting beneath Niuafo’ou

Island and the northern Lau back-arc basin, Contrib. Mineral. Petrol., 156, 103–118.Resing, J. A., et al. (2011), Active submarine eruption of boninite in the northeastern Lau Basin, Nat. Geosci., 4, 799–806, doi:10.1038/

NGEO1275.Rison, W., and H. Craig (1983), Helium isotopes and mantle volatiles in Loihi Seamount and Hawaiian island basalts and xenoliths, Earth

Planet. Sci. Lett., 66, 407–426.RR1211 (2012), Final Cruise Report for Submarine Ring of Fire 2012: Northeast Lau Basin. [Available at http://oceanexplorer.noaa.gov/explo-

rations/12fire/logs/summary/srof12-cruisereport-final.pdf.]Rubin, K. H. and J.M. Sinton (2007), Inferences on mid-ocean ridge thermal and magmatic structure from MORB compositions, Earth Planet.

Sci. Lett., 260, 257–276.Rubin, K.H., R. W. Embley, D. A. Clague, J. A. Resing, P. J. Michael, N. S. Keller, and E.T. Baker (2009), Lavas from active boninite and very

recent basalt eruptions at two submarine NE Lau Basin sites, Eos Trans. AGU, 90(52), Fall Meet. Suppl., Abstract # V43I-05.Rubin, K. H., S. A., Soule, W. W., Chadwick, D. J. Fornari, D. S., Clague, R.W. Embley, E. T. Baker, M. R. Perfit, D. W. Caress, and R. P. Dziak.

(2012), Volcanic eruptions in the deep sea, Oceanography, 25,142–157.Rubin, K. H., R. W. Embley, R. J. Arculus, and J. E. Lupton (2013), Magmatically greedy reararc volcanoes of the N. Tofua segment of the

Tonga Arc, Abstract V13I-04 presented at2013 Fall Meeting, AGU, San Francisco, Calif.Sano, Y., and S. N. Williams (1996), Fluxes of mantle and subducted carbon along convergent plate boundaries, Geophys. Res. Lett., 23,

2749–2752.



http://dx.doi.org/10.1029/2004GL020041

http://dx.doi.org/10.1029/2004GL020041

http://dx.doi.org/10.1029/2005GC001152

http://dx.doi.org/10.1029/2007JB005467

http://dx.doi.org/10.1029/2009GL039468

http://dx.doi.org/10.1029/2012GL051201

http://dx.doi.org/10.1038/ncomms7554

http://dx.doi.org/10.1002/2013GL057976

http://dx.doi.org/10.1093/petrology/egu071

http://dx.doi.org/10.1093/petrology/egu071

http://dx.doi.org/10.1029/2004GC000895

http://dx.doi.org/10.1016/S0012-821X(98)00212-X

http://dx.doi.org/10.1016/S0012-821X(01)00522-2

http://dx.doi.org/10.1038/NGEO1275

http://dx.doi.org/10.1038/NGEO1275

http://oceanexplorer.noaa.gov/explorations/12fire/logs/summary/srof12-cruisereport-final.pdf

http://oceanexplorer.noaa.gov/explorations/12fire/logs/summary/srof12-cruisereport-final.pdf

Shaw, A. M, D. R Hilton, C. G. MacPherson, and J. M Sinton (2004), The CO2-He-Ar-H2O systematics of the Manus back-arc basin: Resolvingsource composition from degassing and contamination effects, Geochim. Cosmochim. Acta, 68, 1837–1856.

Staudacher, T., P. Sarda, and C. J. Allegre (1990), Noble gas systematic of Reunion Island, Chem. Geol., 89, 1–17.Tian, L., P. R. Castillo, D. R. Hilton, J. W. Hawkins, B. B. Hanan, and A. J. Pietruszka (2011), Major and trace element and Sr-Nd isotope signa-

tures of the northern Lau Basin lavas: Implications for the composition and dynamics of the back-arc basin mantle, J. Geophys. Res., 116,B11201, doi:10.11029/2011JB008791.

TN-234 (2009), Final Cruise Report for Northern Lau Basin Response Cruise. [Available at http://www.pmel.noaa.gov/eoi/laubasin/docu-ments/tn234-NELRC-CruiseReport-final.html.]

Trull, T., S. Nadeau, F. Pineau, M. Polv�e and M. Javoy (1993), C-He systematics in hotspot xenoliths: Implications for mantle carbon contentsand carbon recycling, Earth Planet. Sci. Lett., 118, 43–64.

Turner, S., and C. Hawkesworth (1998), Using geochemistry to map mantle flow beneath the Lau Basin, Geology, 26, 1019–1022.van Soest, M. C., D. R. Hilton, and R. Kruelen (1998), Tracing crustal and slab contribution to arc magmatism in the Lesser Antilles island arc

using helium and carbon relationships in geothermal fluids, Geochim. Cosmochim. Acta, 62, 3323–3335.Wendt, J. I., M. Regulous, K. D. Collerson, and A Ewart (1997), Evidence for a contribution from two mantle plumes to island-arc lavas from

northern Tonga, Geology, 25, 611–614.Young, C., and J. E. Lupton (1983), An ultratight fluid sampling system using cold-welded copper tubing, Eos Trans. AGU, 64, 735.Zellmer, K. E., and B. Taylor (2001), A three-plate kinematic model for Lau Basin opening, Geochem. Geophys. Geosyst., 2, 1020, doi:10.1029/

2000GC000106.



http://dx.doi.org/10.11029/2011JB008791

http://www.pmel.noaa.gov/eoi/laubasin/documents/tn234-NELRC-CruiseReport-final.html

http://www.pmel.noaa.gov/eoi/laubasin/documents/tn234-NELRC-CruiseReport-final.html

http://dx.doi.org/10.1029/2000GC000106

http://dx.doi.org/10.1029/2000GC000106

Documents

Helium isotope, C/3He, and Ba‐Nb‐Ti signatures in the ...... · RESEARCH ARTICLE 10.1002/2014GC005625 Helium isotope, C/3He, and Ba-Nb-Ti signatures in the northern Lau Basin: