Accepted Manuscript

Clinopyroxene and bulk rock Sr–Nd–Hf–Pb isotope compositions

of Raivavae ocean island basalts: Does clinopyroxene record early
stage magma chamber processes?

Takashi Miyazaki, Takeshi Hanyu, Jun-Ichi Kimura, Ryoko

Senda, Bogdan Stefanov Vaglarov, Qing Chang, Yuka Hirahara,
Toshiro Takahashi, Hiroshi Kawabata, Tomoki Sato

PII: S0009-2541(17)30696-4
DOI: https://doi.org/10.1016/j.chemgeo.2017.12.015
Reference: CHEMGE 18586
To appear in: Chemical Geology
Received date: 3 April 2017
Revised date: 11 December 2017
Accepted date: 20 December 2017

Please cite this article as: Takashi Miyazaki, Takeshi Hanyu, Jun-Ichi Kimura, Ryoko
Senda, Bogdan Stefanov Vaglarov, Qing Chang, Yuka Hirahara, Toshiro Takahashi,
Hiroshi Kawabata, Tomoki Sato , Clinopyroxene and bulk rock Sr–Nd–Hf–Pb isotope
compositions of Raivavae ocean island basalts: Does clinopyroxene record early stage
magma chamber processes?. The address for the corresponding author was captured as
affiliation for all authors. Please check if appropriate. Chemge(2017), https://doi.org/
10.1016/j.chemgeo.2017.12.015

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

Clinopyroxene and bulk rock Sr–Nd–Hf–Pb isotope compositions of Raivavae

ocean island basalts: Does clinopyroxene record early stage magma chamber

processes?

Takashi Miyazaki1*, Takeshi Hanyu1, Jun-Ichi Kimura1, Ryoko Senda2, Bogdan

PT
Stefanov Vaglarov1, Qing Chang1, Yuka Hirahara3, Toshiro Takahashi4, Hiroshi

Kawabata5, and Tomoki Sato1

RI
SC
1
Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Yokosuka 237-
0061, Japan
2
Faculty of Social and Cultural Studies, Kyushu University, Fukuoka 819-0395, Japan
NU
3
Chiba Institute of Technology, Chiba 275-0016, Japan
4
Department of Geology, Niigata University, Niigata 950-2181, Japan
5
Interdisciplinary Science Unit, Kochi University, Kochi 780-8520, Japan
MA

Takashi Miyazaki: tmiyazaki@jamstec.go.jp

Takeshi Hanyu: hanyut@jamstec.go.jp
D

Jun-Ichi Kimura: jkimura@jamstec.go.jp

Ryoko Senda: senda@scs.kyushu-u.ac.jp
E

Bogdan Stefanov Vaglarov: bog@jamstec.go.jp

PT

Qing Chang: qchang@jamstec.go.jp

Yuka Hirahara: yuka8685@wine.ocn.ne.jp
Toshiro Takahashi: toshirot@geo.sc.niigata-u.ac.jp
CE

Hiroshi Kawabata: hiroshik@kochi-u.ac.jp

Tomoki Sato: t-sato@jamstec.go.jp
AC

*Corresponding author : Takashi Miyazaki

2-15 Natsushima-cho,
Yokosuka 237-0061 Japan

Email address: tmiyazaki@jamstec.go.jp

 ACCEPTED MANUSCRIPT

Abstracts

We investigated the Sr–Nd–Hf–Pb isotope compositions of clinopyroxene separates and

strongly leached bulk rock powders from the basalt lavas on Raivavae Island in the South

Pacific. These data allowed us to explore the magma chamber processes of the two basalt

groups: the high-μ-type (μ = 238U/204Pb, HIMU) Rairua basalts erupted at ~9 Ma, and the

PT
depleted mantle (DM)-type Anatonu basalts erupted at ~6 Ma. We assessed different

analytical techniques for clinopyroxene sample treatment, including leaching, sample

RI
digestion, chemical separation, and determination of parent/daughter ratios for use in age

SC
corrections. The age-corrected initial Sr–Nd–Hf isotope compositions of clinopyroxene

samples solely overlap on the bulk rock compositions of each basalt type. The initial Pb
NU
isotope compositions show significant differences between the clinopyroxenes and bulk

rock samples in both the Rairua and Anatonu basalts. The clinopyroxenes have less
MA

radiogenic Pb than their hosts, reflecting the magma chamber processes. The overall

variations in the bulk rock Pb isotopes are explained by mixing of melts from the HIMU-
D

and DM-type sources in the plume. In detail, early stage Rairua magma was more
E

influenced by DM melts from which Rairua clinopyroxene was crystallized. Subsequent

PT

supply of dominant HIMU melts drove the bulk rock magma chemistry more HIMU-like.

The Anatonu DM magma was largely from depleted mantle that differs from the Pacific
CE

MORB source. Subordinate contribution of a Pacific alkali-basalt-like component is

suggested from the clinopyroxene chemistry of Anatonu perhaps from impregnated

AC

pyroxenite in the lithospheric mantle. (245 words)

6775 words in the main text excluding reference.

598 words in Figure and Table captions

 ACCEPTED MANUSCRIPT

Keywords: Clinopyroxene, Sr–Nd–Hf–Pb isotopes, Isotopic inhomogeneity, Raivavae

Island

the high-μ-type (μ = 238U/204Pb, HIMU); depleted mantle (DM) Ocean island basalts
(OIBs); parent/daughter (P/D) ratios; rare earth element (REE) ; mid-ocean ridge basalts
(MORBs); depleted MORB source mantle (DMM); Poly Tetra- Fluoroethylene-co-

PT
Perfluoro Alkyl vinyl ether (PFA); National Institute of Standards and Technology
(NIST); Standard deviation (SD); Standard error (SE); Relative deviation (RD)

RI
SC
NU
MA
E D
PT
CE
AC
 ACCEPTED MANUSCRIPT

1. Introduction

Basalt lavas vary in composition because of differences in the processes of mantle melting,

fractionation, mixing, and assimilation that take place in magma chambers, along with

differences in source materials in the mantle and crust (Wilson, 1989). Ocean island

basalts (OIBs) are no exception; the importance of shallow magma chamber processes

PT
has been highlighted with respect to mineralogy and petrology (Kawabata et al., 2011),

isotope geochemistry (Hanyu et al., 2011; Hofmann, 2003; Jackson et al., 2009), and

RI
volatile geochemistry (Kendrick et al., 2015; Lassiter et al., 2002). Nevertheless, isotopic

SC
and elemental variations (trends) of OIBs have been attributed, in many cases, to the

variations in the source composition and mantle processes (Jackson et al., 2007; Jackson
NU
et al., 2010; Stracke et al., 2003; Stracke et al., 2005; Willbold and Stracke, 2006).

Recently, numerical modeling of melt-melt mixing in shallow magma chambers has

MA

raised questions about the origin of the geochemical trends, which have previously been

used to identify heterogeneous mantle components (e.g., Rudge et al., 2013; Stracke and
D

Bourdon, 2009). Decoding the shallow magma chamber processes is therefore important
E

for any further discussions on source composition and mantle processes.

PT

Before examining the magma chamber processes, identifying alteration-free

isotopic and elemental compositions of an OIB remains a challenge. This is partly because
CE

of seawater alteration, including interactions with high-temperature brines (Kent et al.,

1999: Lassiter et al., 2002), and potentially due to assimilation with seawater-altered
AC

oceanic crust (Bezard et al., 2015). Even in the absence of crustal assimilation, the effect

of seawater alteration has been detected in the Sr–Nd–(Hf)–Pb isotope analyses of bulk

rock submarine samples (Hirahara et al., 2015; Miyazaki et al., 2015). Various leaching

techniques have been used to remove the effects of this alteration (e.g., Nobre Silva et al.,

2009, 2010; Todd et al., 2015; Weis et al., 2005). Typically, hydrochloric acid is used to

dissolve and remove the altered components from sample powders; however, even if
 ACCEPTED MANUSCRIPT

strong leaching is applied, it is difficult to confirm the complete removal of the altered

materials (e.g., Hanyu and Nakamura, 2000).

Moreover, there is controversy around whether a “leached” or “un-leached”

sample is more suitable for determining parent/daughter (P/D) ratios for use in the

calculation of initial isotope ratios. Thompson et al. (2008) concluded that samples used

PT
in the Nd and Hf isotope analyses of highly and moderately depleted submarine mafic

rocks older than ~50 Ma should be leached with 6 M HCl, and that the same leached

RI
samples should then be used for the P/D ratio measurements. In contrast, Nobre Silva et

SC
al. (2010) found that the P/D ratios measured from leached powders are not reliable

because of the resultant trace element loss and rare earth element (REE) fractionation that
NU
occurs during the acid leaching process. Therefore, they concluded that P/D ratios should

be determined from elemental concentration ratios measured from un-leached sample

MA

powders (Nobre Silva et al., 2010; 2013). For both views, the timings of lava

emplacement and alteration are usually not the same, and thus, the uncertainty persists.
D

To overcome this leaching problem, clinopyroxene is a suitable OIB constituent

E

that can be studied because it is resistant to alteration and is crystallized early in the
PT

magma differentiation process, and therefore, should record a less fractionated magma

history (Hanyu et al., 2011; Hanyu et al., 2014; Hanyu and Nakamura, 2000; Jackson et
CE

al., 2009). Hanyu and Nakamura (2000) reported less radiogenic 87Sr/86Sr compositions

from clinopyroxene separates than from leached bulk rocks from Polynesian OIBs.
AC

Jackson et al. (2009) found an 87Sr/86Sr and 143Nd/144Nd isotopic disequilibrium between

clinopyroxenes and host rocks, and used this to infer isotopic heterogeneity in the mantle

source. These studies also suggested the importance of age corrections, which should be

made not only for bulk rocks but also for clinopyroxene separates.

The P/D ratios of clinopyroxene greatly differ from those of bulk rocks. The most

suitable age correction is to use the P/D ratios determined from the same clinopyroxene
 ACCEPTED MANUSCRIPT

solutions used for the isotope measurements. Hanyu et al. (2011) determined P/D ratios

from the trace element analysis of clinopyroxene. Later Hanyu et al. (2014) determined

the P/D ratios from the same clinopyroxene solutions that were used for the Sr–Nd–Hf

isotope analyses. However, they determined the Th/Pb and U/Pb ratios from the Sr–Nd–

Hf solutions, not from the solutions used for the Pb isotope analysis. The problem with

PT
this method was the incomplete recovery of Th from the Pb isotope digests using a HF–

HNO3 mixture (e.g., Yokoyama et al., 1999).

RI
In this study, we focus on the Sr–Nd–Hf–Pb isotope compositions of

SC
clinopyroxene phenocrysts in the OIBs on Raivavae, one of the Austral Islands in the

South Pacific (Fig. 1). The bimodal isotopic variation that characterizes the bulk rock
NU
basalts on this island suggests the presence of multiple mantle sources (Lassiter et al.,

2003; Maury et al., 2013). An on-land geological survey, coupled with K–Ar dating,
MA

revealed two volcanic units on Raivavae island: the Rairua (7.4 – 10.6 Ma) and the

Anatonu (5.4 – 6.4 Ma) basalts (Maury et al., 2013). The Rairua volcano is a more than
D

200-m-thick pile of submarine autoclastic basaltic breccias, and the Anatonu shield
E

volcano is mostly composed of up to 400-m-thick pile of basaltic and intermediate lava

PT

flows (Maury et al., 2013). The Rairua basalts show typical high-μ (μ= 238U/204Pb, HIMU)

isotopic signatures, similar to those of the Old-Rurutu and Tubuai basalts. The Anatonu
CE

basalts have far more unradiogenic Pb than typical HIMU. They still contain more

radiogenic Pb than typical mid-ocean ridge basalts (MORBs) formed from depleted
AC

MORB source mantle (DMM). The Anatonu basalts also have more similarities with

DMM than with the Rairua basalts in terms of Nd–Hf isotopes. In this paper, the Rairua

and Anatonu basalts are denoted as being HIMU- and depleted mantle (DM)-type,

respectively.

The isotopic compositions of the clinopyroxene separates and the strongly

leached bulk rock powders are examined, with the goal of identifying the magma chamber
 ACCEPTED MANUSCRIPT

processes that control the geochemical variations observed in the Raivavae basalts. In

order to achieve this, the analytical protocols used to determine the P/D ratios and Sr–

Nd–Hf–Pb isotope compositions are first examined. This method is then applied to the

Rairua (~9 Ma) and Anatonu (~6 Ma) basalts. Finally, the origin of the isotopic variations

found in these basalts and clinopyroxenes, caused by the interplay between basalt melts

PT
from different mantle sources that occurred in shallow magma chambers, is discussed.

RI
2. Samples

SC
Raivavae Island is located in the Cook-Austral volcanic chain in the South Pacific (Fig.

1). The origin of the Cook-Austral chain is still a matter of debate (Maury et al., 2013).
NU
Following the idea by Duncan and McDougall (1976) and Chauvel et al. (1997), this chain

is believed to consist of at least three overlapping hotspot trends: the Tubuai trend, Atiu
MA

trend, and Rarotonga trend (Duncan and McDougall, 1976; Chauvel et al., 1997). Among

them, the Tubuai hotspot trend, which includes the islands of Rurutu (Old-Rurutu),
D

Tubuai, and Raivavae, is distinguishable from the other hotspot trends by the isotopic
E

composition of its basalts, which have HIMU signatures.

PT

The Rairua and Anatonu basalts are alkali-basalt to basanite in composition.

These two units are clearly distinguishable based on isotopic composition (Lassiter et al.,
CE

2003; Maury et al., 2013). The phenocrysts in the basalts are typically 1‒3 mm in size.

Some of these basalts contain abundant olivine and clinopyroxene phenocrysts (up to
AC

30 % in volume) and rare plagioclase. Clinopyroxenes are Ti-augite, characterized by

complex zoning in their cores and thin Ti-rich rims.

3. Analysis

For the precise trace element composition and isotope ratio measurements of Sr–Nd–Hf–

Pb in the clinopyroxene separates, particular care must be taken with laboratory blanks,
 ACCEPTED MANUSCRIPT

the choice of resins, and considerations of potential matrix effects that may occur during

the separation of the elements. A description of our analytical protocols has been provided

below.

3.1. Reagents

PT
TAMAPURE®-AA 10 grade (<10 ppt impurities), 38% HF and 68% HNO3, and AA-100

grade (<100 ppt impurities), 47% HBr and 70% HClO4 (Tama Chemical Co., Ltd.,

RI
Kawasaki, Japan), reagents were used without any additional purification. De-ionized

SC
water and HCl were purified via a two-bottle Teflon® still from Milli-Q® water

(Millipore®, Molsheim, France) and EL grade (for electronic industry) HCl (Kanto
NU
Chemical Co., Inc., Tokyo, Japan), respectively.
MA

3.2. Sample preparation and leaching procedure

3.2.1. Bulk rock

For the bulk rock chemical analyses, rock samples, wrapped in plastic bags, were roughly
E D

crushed with a steel hammer and only fresh chips were handpicked. The rock chips were
PT

soaked in a warm water bath to remove any surface and immersed salts, then rinsed (first

with acetone and then with de-ionized water) in an ultrasonic bath. The dried rock chips
CE

were then pulverized in a vibration mill (Multi-Beads Shocker, Yasui Kikai Co., Osaka,

Japan) using a polycarbonate capsule with a 50 cm3 inner volume and a high-purity
AC

alumina rod to prepare the bulk rock powders. The samples were further leached prior to

the Sr, Nd, Pb, and Hf isotopic analyses following the methods set out by McDonough

and Chauvel (1991). Approximately 2 g of each powder sample was weighed into acid-

washed 30 mL Savillex® Poly tetra-Fluoroethylene-co-perfluoro Alkyl vinyl ether (PFA)

jars. 20 mL of 6 M HCl was then added, and the samples were kept at 100 °C for 2 h.

After the supernatant was decanted, 20 mL of 6 M HCl and 1 mL of concentrated HF

were added and kept at 100 °C for a further 30 min. Weis et al. (2005) tested the leaching
 ACCEPTED MANUSCRIPT

method put forth by McDonough and Chauvel (1991), and showed the negligible effect

of HF on isotope ratios. The supernatant was then decanted, and the residues were rinsed

three times using two-bottle-distilled Milli-Q® water. The resulting leached rock powder

was then placed on a hot plate at 45 °C until completely dried and was then analyzed.

Takamasa and Nakai (2009) reported significant Pb contamination during

PT
pulverization using an alumina hand mill. We tested the effect of Pb contamination using

a strongly leached powder of BHVO-2 basalt (United States Geological Survey: USGS).

RI
A sample of the powder was further pulverized by the alumina vibration mill and

SC
compared to the same powder sample before pulverization. The Pb concentration of

BHVO-2 is 1.30 ppm (Weis et al., 2005) and is best suited for the test because the average
NU
Pb concentration of the unleached Raivavae basalts is 1.92 ppm (0.86–3.75 ppm). The

results showed no detectable difference, indicating that our alumina mill was free from
MA

Pb contamination (Table S1). We also tested for the same contamination effect for Sr–

Nd–Hf isotopes and obtained the same results (not shown).

D

We applied the leaching method detailed by McDonough and Chauvel (1991) to

E

the bulk rock analyses, with a modification to the sample/reagent ratio. We increased the
PT

sample amount from 0.3–0.4 g to 2 g, with the corresponding amount of reagent increased

from 10 to 20 mL. To check the validity of this increased sample/reagent ratio, the results
CE

are compared to published Pb isotope ratios of leached BHVO-2 basalt. Weis et al. (2005)

applied the leaching methods used by Weis and Frey (1991, 1996) and McDonough and
AC

Chauvel (1991) for BHVO-2 basalt. Todd et al. (2015) also reported leached BHVO-2

results. Our results are identical to their values, which are indicative of effective leaching

(Table S1). The effects of leaching on Sr–Nd–Hf isotopes have been examined for highly

altered submarine basalt samples and a number of results have been reported (Hirahara et

al., 2015).
 ACCEPTED MANUSCRIPT

3.2.2. Clinopyroxene

For the clinopyroxenes samples, rock chips were broken with a jaw crusher into fragments

using mesh sizes of #9‒16. Coarse clinopyroxene fragments were handpicked and further

crushed using mesh sizes down to #60‒80. Pure clinopyroxene fragments were separated

from other mineral and glass fragments with a hand magnet and a methylene iodide heavy

PT
liquid (MIHL). They were then purified again by handpicking under a binocular

microscope to remove any impurities; approximately a 99 % purity was achieved. The

RI
final clinopyroxene separates were rinsed in ethanol and cleaned with 1 M HCl at room

SC
temperature for 12 h prior to digestion (weak leaching). The clinopyroxene samples were

not pulverized in order to prevent any contamination of their low abundance elements,
NU
particularly, U, Th, and Pb.

In order to check the effective removal of contamination by the weak leaching,

MA

three clinopyroxene separates were strongly leached using the modified leaching method

detailed by McDonough and Chauvel (1991) (see above). The resultant Sr–Nd–Hf–Pb
D

isotope ratios are given for weakly and strongly leached clinopyroxene separates in Table
E

S2. Because of considerable isotopic variations found only in Pb isotopes, we further

PT

assessed effects of Pb contamination from MIHL. Four clinopyroxene separates without

using MIHL were also strongly leached and analyzed for Pb isotopes along with their
CE

leachates (Table S2).

AC

3.3. Sample digestion

Approximately 0.1 g of the leached bulk rock sample powders were used for the Sr–Nd–

Hf isotopic analyses and P/D element ratio measurements. The powders were digested in

an HClO4/HF mixture (with a ratio of 1 to 3), sealed in the Savillex® PFA jars, and kept

at 110 °C for more than two days. The digests were then completely dried using stepwise

heating from 80 °C to 195 °C to ensure evaporation of HF and, subsequently, HClO 4, in

 ACCEPTED MANUSCRIPT

order to prevent CaF precipitation, which co-precipitates with the REEs (e.g., Yokoyama

et al., 1999). The digested samples were dissolved again in a mixture of HClO4 and HCl

(with a ratio of 1 to 3) and kept overnight at 110 °C to ensure the digestion of any invisible

CaF. The samples were then completely dried (again, using stepwise heating from 80 °C

to 195 °C). The digested samples were then dissolved in 6 M HCl and kept overnight at

PT
110 °C. After drying the samples at 90 °C, the digested samples were dissolved in 1 mL

of 1 M HCl containing 0.1 M HF. The digested solutions were quantitatively used for

RI
Rb/Sr, Sm/Nd, Lu/Hf, Th/Pb, and U/Pb ratio measurements. The clinopyroxene separates

SC
(0.03–0.15 g) were also used for the same digestion procedure.

We applied a separate decomposition procedure for the Pb isotopes. The primary

NU
reason for this was because it is inappropriate to use HClO4 with the anion exchange resin.

We used a conventional anion exchange resin for the Pb separation because a Pb-specific
MA

resin increases the risk of incomplete recovery (< 95 %), especially for a small amount of

Pb (Horwitz et al., 1994). Incomplete recovery causes the isotopic fractionation of Pb

D

(Baker et al., 2004). The Pb-specific resin also increases the number of procedural blanks
E

(approximately 150 pg), as shown in the literature, even with the use of a Sr-specific resin
PT

for Pb (Deniel and Pin, 2001). These issues are not problems for the bulk rock analyses

but are for clinopyroxene separates, as they contain small amounts of Pb in a limited
CE

amount of sample. For the Pb isotopes, 0.1–0.4 g of the leached bulk rock powders and

0.4–0.9 g of the clinopyroxene separates were digested in a concentrated mixture of HNO3

AC

and HF (with a ratio of 1 to 4) and kept at 110 °C in tightly sealed Savillex® PFA jars for

at least two days. The samples were then heated step by step from 80 °C to 115 °C until

completely dried. They were then dissolved again in an HNO3-HF mixture and kept at

110 °C overnight. The samples were then completely dried again using stepwise heating

from 80 °C to 115 °C. Next, the samples were dissolved in 8 M HBr and kept overnight

at 110 °C. After drying at 90 °C, they were dissolved again in 8 M HBr and dried at 90 °C.
 ACCEPTED MANUSCRIPT

Finally, the samples were dissolved in 0.5 M HBr for Pb separation, and then followed U,

Th, and Pb concentration analyses.

In order to determine the Th/Pb and U/Pb ratios from the clinopyroxene solutions

used for Pb isotope analysis, we applied the following procedure. First, the clinopyroxene

digested samples, in 0.5 M HBr, were placed in the centrifuge and the Pb-bearing HBr

PT
supernatant solutions were separated from the precipitates (mainly CaF). Fluoride

precipitation commonly occurs when there is a high concentration of Ca in the

RI
clinopyroxene separates. The separated fluoride precipitates were then entirely digested

SC
using 1 mL of HClO4 and heated at 160 °C for at least one day. The digested samples

were then dissolved in 3 % HNO3. If any undigested fluoride remained in the samples,
NU
the HClO4 digestion process was re-applied. The procedural blanks for Th and U during

this fluoride decomposition were 2.8 and 0.51 pg, respectively. These values were
MA

subtracted in accordance with the amount of HClO4 used. We confirmed that no Pb blanks

were detected during the fluoride decomposition. The Th, U, and Pb contents of the HBr
D

solutions and fluoride digests were analyzed and summed to obtain the P/D ratios. This
E

procedure is particularly important for obtaining accurate Th compositions because Th

PT

co-precipitates with fluoride and the REEs (Yokoyama et al., 1999).

CE

3.4. Element analysis for P/D ratio measurement

For the P/D ratio determinations, standard In and Bi solutions were added to all sample
AC

solutions as internal standards. The element concentrations of Rb, Sr, Sm, Nd, Lu, Hf, Pb,

Th, and U were determined by solution nebulization quadrupole inductively coupled

plasma-mass spectrometry (ICP-MS) (Agilent 7500ce; Agilent Technologies, Tokyo,

Japan) at JAMSTEC. The repeatability and laboratory bias’ were better than 1 % and 3 %,

respectively (Chang et al., 2003).

 ACCEPTED MANUSCRIPT

3.5. Element separation

The chemical separation procedures used were conventional and have been described in

detail elsewhere (Hanyu et al., 2011; Hirahara et al., 2009 and 2012; Kimura and Nakano,

2004; Miyazaki et al., 2009 and 2012; Takahashi et al., 2009). In short, the Sr–Nd–Hf

separation was done using an AG50W-X8 cation ion exchange resin (Bio Rad, California,

PT
USA), along with Sr-spec and Ln-spec resins (Eichrom Tec. Inc., Illinois, USA). After

the initial separation, using 1 mL of AG50W-X8 resin, Sr, Nd, and Hf were purified using

RI
0.05 mL of Sr-spec resin, 0.3 mL of Ln-spec resin, and 1 mL of Ln-spec resin, respectively.

SC
The Pb separation was performed using 0.1 mL of AG1-X8 anion exchange resin (Bio

Rad, California, USA). Parts of the column separation procedures were conducted with a
NU
fully automated open-column chemical separation system named COLUMNSPIDER,

developed by JAMSTEC and HOYUTEC Co., Ltd. (Kawagoe, Japan) (Miyazaki et al.,
MA

2012). The total procedural blanks for Sr, Nd, and Hf were less than 69, 8, and 16 pg,

respectively. The total procedural blanks for Pb was 10 pg for the analysis of
D

clinopyroxene separates with strong leaching (with and without MIHL for clinopyroxene
E

separation), and 39 pg for other clinopyroxene and bulk rock samples. Pb is the most
PT

sensitive element for blank contamination because the blank/sample ratio of Pb is the

highest. The maximum blank effects for bulk rock powder and clinopyroxene separate
CE

samples were calculated using the measured Pb isotope ratios of the procedural blanks,

which gave the following ratios: 206Pb/204Pb = 18.3408 ± 0.1146, 207Pb/204Pb = 15.5659 ±
AC

0.0909, and 208Pb/204Pb = 38.1879 ± 0.2248 (errors in 2SE, n = 84). The maximum blank
206
effects on the measured Pb/204Pb, 207
Pb/204Pb, and 208
Pb/204Pb isotope ratios were

0.0013, 0.0001, and 0.0010, respectively, for the bulk rock samples, and 0.0048, 0.0002,

and 0.0043, respectively, for the clinopyroxene separates. All the differences are within

the size of the plot symbols for all figures (see below).
 ACCEPTED MANUSCRIPT

3.6. Isotope ratio measurement

The Sr and Nd isotope ratios were measured with a thermal ionization mass spectrometer

(TIMS) using Triton TI (Thermo-Finnigan, Bremen, Germany) at JAMSTEC. The

measured Sr and Nd isotope ratios were normalized to 86Sr/88Sr = 0.1194 and 146Nd/144Nd

= 0.7219, respectively, to correct for mass fractionation. The mean 87Sr/86Sr value in the

PT
standard reference materials from the National Institute of Standards and Technology

(NIST) SRM 987 was 0.710234 ± 0.000014 (errors given by two-standard deviations

RI
(2SD), n = 39) and the mean value in the JNdi-1 standard was 0.512097 ± 0.000009 (2SD,

SC
87
n = 14) during analytical sessions. The Sr/86Sr and 143
Nd/144Nd data in this paper are

reported without normalization.

NU
The Hf isotope ratios were measured using a multiple-collector (MC)-ICP-MS

(Neptune; Thermo Scientific, Bremen, Germany) at JAMSTEC. During the measurement,

MA

173
Yb and 175Lu peaks were monitored in order to correct for interferences of 176
Yb and
176 176
Lu on the Hf peak, and were always undetectable. The instrumental mass
D

179
fractionation was determined by the Hf/177Hf ratio and normalized to be 0.7325.
E

176
Repeated measurements of the JMC-475 Hf standard solution gave a mean Hf/177Hf
PT

ratio of 0.282145 ± 0.000013 (2SD, n = 67) during the analyses. We reported 176Hf/177Hf

ratios normalized to the reference JMC-475 value of 0.28216 (Patchett et al., 2004).
CE

The Pb isotope ratios were also determined using the MC-ICP-MS. The Pb

isotope mass fractionation was corrected using Tl as an external standard (Baker et al.,
AC

2004; Weis et al., 2006). Additional mass-dependent, inter-element fractionations were

also corrected by applying a standard bracketing method using NIST SRM 981 as a
206
standard (Kimura et al., 2006 and 2015). Repeated measurements of Pb/204Pb,
207
Pb/204Pb, and 208
Pb/204Pb using the SRM 981 yielded values of 16.9301 ± 0.0041,

15.4832 ± 0.0040, and 36.6732 ± 0.0116 (2SD, n = 84), respectively. We reported

206
Pb/204Pb, 207
Pb/204Pb, and 208
Pb/204Pb ratios normalized to reference NIST SRM 981
 ACCEPTED MANUSCRIPT

values of 16.9416, 15.5000, and 36.7262 (Baker et al., 2004), respectively. All the

analytical results are given in Tables 1 and 2. The analytical results of the JB-2 basalt

standard (Geological Survey of Japan: GSJ), obtained during the analytical session, are

given in Table S3.

PT
3.7. Major element analyses

We also analyzed major element composition of the bulk rock powders and the

RI
clinopyroxene separates for representative samples. Major element compositions were

SC
determined by X-ray fluorescence (XRF) spectrometry using Simultix 12 (RIGAKU,

Osaka, Japan) at JAMSTEC (Tani et al., 2006). Analytical repeatability and laboratory
NU
bias were better than 2% RD (relative deviation from the reference values) and 1–5% SD.

Major element analysis of clinopyroxenes separates was made on polished clinopyroxene

MA

cores by laser ablation (LA)-ICP-MS using a 200-nm ultraviolet femtosecond LA system

OK-Fs2000K (OK Lab., Tokyo, Japan) connected with an Element XR (Thermo

D

Scientific, Bremen, Germany) sector field ICP-MS at JAMSTEC. Repeatability and

E

laboratory bias were better than 2 % SD and smaller than 10 % RD, respectively, for most
PT

elements. Those for trace K2O and P2O5 were <8 % SD and <20 % RD, respectively

(Kimura and Chang, 2012). All the results are given in Table 3.
CE

4. Results and Discussions

AC

In this section, we first discuss the effect of blank and laboratory contamination. The

effect of age correction and seawater alteration are then evaluated, and then possible

magma chamber processes are discussed based upon the differences in isotopic

compositions observed between clinopyroxenes and their host bulk rocks. Finally, the

origins of the two types of OIBs (the HIMU-type Rairua and DM-type Anatonu basalts)

are discussed based upon their internal variations and genetic relationship.
 ACCEPTED MANUSCRIPT

4.1. Pb blanks and effect of leaching

Considerable variations between clinopyroxene and bulk rock pairs are found in Pb

isotopes, whereas Sr-Nd-Hf isotopes showed no significant difference (Tables 1 and 2).

We thus carefully evaluated the effect of analytical method and blanks for Pb isotopes.

PT
First, we measured Pb concentration in MIHL. The result showed 0.06–0.14 ppb. Even if

1% clinopyroxene weight equivalent MIHL remained on the surface or in the clacks of

RI
clinopyroxene, only 1.25 pg of Pb over 0.9 g of decomposed clinopyroxene affects Pb

SC
isotope ratios. This level is far smaller than the total procedural blank of 39 pg during the

sample analyses. Therefore, Pb blank from MIHL is negligible.

NU
Second, we examined different leaching methods, weak and strong leaching. The

effect of MIHL separation was doubly checked using strong leaching without MIHL
MA

separation. All but one analytical result showed no difference between weak leaching

with MIHL separation and strong leaching without MIHL separation procedures (Table
D

S2). An exception was clinopyroxene sample RAV-18, which showed large differences in
E

206
Pb/204Pb (0.3931), 207Pb/204Pb (0.0190), and 208Pb/204Pb (0.3225) between weakly and
PT

strongly leached clinopyroxene. (Table S2). The weakly leached RAV-18 clinopyroxene
207
plot between the strongly leached clinopyroxene and its leachate on Pb/204Pb–
CE

206
Pb/204Pb and 208
Pb/204Pb–206Pb/204Pb plots (not shown). Pb blank in the chemical

reagents used for leaching (28 ml of 6M HCl and 3.15 ml of 38% HF were used) was 234
AC

pg, whereas the column separation blank after leaching was only 2 pg. Pb isotope ratios

of the reagent blank were 206Pb/204Pb = 18.6963 ± 0.0097, 207Pb/204Pb = 15.7392 ± 0.0079,
208
and Pb/204Pb = 38.8420 ± 0.0198 (2SE, n = 30). The mixing line drawn between the

two RAV-18 clinopyroxenes and the leachate did not point to the isotope ratios of the

reagent blank.

The large Pb isotope variations found in RAV-18 clinopyroxene samples indicate

 ACCEPTED MANUSCRIPT

insufficient leaching. We have discarded Pb isotope data from RAV-18 clinopyroxene in

the following discussions. Samples other than RAV-18 showed sufficient removal of Pb

contamination even with weak leaching after MIHL separation (Table S2). Considerable

differences were observed between clinopyroxene samples and their bulk rock pairs in
206
Pb/204Pb (+0.0982 to –0.0419) and Pb/204Pb (+0.0397 to –0.0351) ratios. The
208

PT
differences should indicate heterogeneity in the magma chambers.

RI
4.2. Age correction

SC
The initial Sr–Nd–Hf–Pb isotope compositions of the Rairua samples were calculated

with an age of 9 Ma, and those from Anatonu with an age of 6 Ma, assuming average K-
NU
Ar ages (Maury et al., 2013). The age-corrected initial isotope ratios of the strongly

leached bulk rock powders are listed together with the measured isotope ratios in Table
MA

1. The radiogenic ingrowth was corrected using the measured Rb/Sr, Sm/Nd, U/Pb, Th/Pb,

and Lu/Hf ratios. The maximum differences between the measured and the age-corrected
D

isotope ratios were 0.000012, 0.000008, 0.000001, 0.0667, 0.0031, and 0.0980 for
E

87
Sr/86Sr, 143
Nd/144Nd (0.09 εNd), 176
Hf/177Hf (0.18 εHf), 206
Pb/204Pb, 207
Pb/204Pb, and
PT

208
Pb/204Pb, respectively. All differences were within the typical 2SD repeatability of the

JB-2 for the Sr, Nd, and Hf isotopes (Table S3). Significant differences were only
CE

observed in the 206Pb/204Pb, 207Pb/204Pb and 208Pb/204Pb ratios.

The measured and initial isotope ratios of the clinopyroxene separates are listed
AC

in Table 2. The initial isotope ratios of Sr, Nd, and Hf were calculated using the measured

Rb/Sr, Sm/Nd, and Lu/Hf ratios from the clinopyroxene separates. The maximum

differences between the measured isotope ratios and the age corrected initial isotope ratios

were 0.000003, 0.000010, and 0.000001 for 87

Sr/86Sr, 143
Nd/144Nd (0.05 εNd), and
176
Hf/177Hf (0.16 εHf), respectively. All differences were within the typical 2SD

repeatability of the JB-2 (Table S3).

 ACCEPTED MANUSCRIPT

Two sets of U/Pb and Th/Pb ratios were tested for use in the Pb isotope age

corrections of the clinopyroxene separates. The first set used the P/D ratios determined

from the solutions used for Sr–Nd–Hf isotope measurements. The second set used the

P/D ratios determined from the solutions for Pb isotope measurements. The maximum

differences between the measured and the age-corrected isotope ratios in the first set were

PT
0.0423, 0.0020, and 0.0885 for 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb, respectively. The

differences for the second set were 0.0479, 0.0022, and 0.0817, respectively. In both cases,

RI
significant radiogenic ingrowth was shown for the 206Pb/204Pb and 208Pb/204Pb ratios. The

SC
initial 206Pb/204Pb and 208Pb/204Pb ratios calculated by the two tests were different beyond

the analytical repeatability of JB-2 (Fig. 2). In practice, the differences between the two
NU
datasets do not affect any discussions in this paper. We use the second dataset because the

P/D ratios were all from the same sample solutions used for Pb isotope analysis and
MA

therefore are internally consistent.

An analytical repeatability of ~1 % (see above) in the P/D ratio determinations

D

should also be incorporated into the errors in the initial isotope ratios; however, such small
E

errors are negligible for the age correction calculations (e.g., the differences range from
PT

0.0479 to 0.0484 for 206Pb/204Pb and from 0.160 to 0.162 for εHf). The effects are therefore

not considered.
CE

4.3. Effects of seawater alteration

AC

Figure 3 compares the age-corrected Sr(i) ((i) denotes the initial isotope ratio at the time

of eruption), Nd(i), and Hf(i) for the leached bulk rock and clinopyroxene samples. Figure

4 shows the same for the Pb(i) isotope compositions. Seawater alteration may modify the

Sr and Nd isotopic compositions of basalts toward radiogenic and unradiogenic values,

respectively (Bryce et al., 2005; Staudigel et al., 1984). In particular, the Sr isotope

composition is easily altered by seawater contamination because of the high Sr content in

 ACCEPTED MANUSCRIPT

87
the seawater. Modern seawater is known to have an Sr/86Sr ratio higher than 0.7089

(Richter et al., 1992). A low 143Nd/144Nd ratio of 0.51234–0.51245 has been reported in

Pacific seawaters (Ling et al., 1997). For example, differences of up to 365 and 240 ppm

in the 87Sr/86Sr and 143Nd/144Nd ratios, respectively, were observed between the leached

and unleached basaltic samples from ODP Site 797 (Hirahara et al., 2015).

PT
In terms of seawater alteration, the bulk rock samples, including groundmass,
87
are expected to show more radiogenic Sr/86Sr ratios than those in clinopyroxene

RI
phenocryst (Hanyu et al., 2011; Hanyu and Nakamura, 2000; Jackson et al., 2009). Most

SC
of the leached bulk rock samples have Sr(i) values identical to those of the clinopyroxene

samples, within the analytical error. Even though some pairs of bulk rock and
NU
clinopyroxene samples in the Rairua basalts show subtle difference in quantities of Sr(i),

the bulk rock samples tend to have lower Sr(i) than the clinopyroxene (Fig. 3b). Moreover,
MA

strongly leached and weakly leached clinopyroxene samples do not show a systematic

change in Sr(i). This suggests that seawater alteration is not a major cause of isotopic
D

variation, and that the subtle differences in Sr(i), if any, may reflect the isotopic
E

heterogeneity in the magmas. Regarding εNd(i), no differences are observed between

PT

clinopyroxenes and bulk rocks beyond the analytical 2SD errors (Fig. 3b). The isotopic

composition of Hf is neither usually affected by seawater alteration, nor by hydrothermal

CE

alteration in the deep sea (Hirahara et al., 2015; Miyazaki et al., 2015). The εHf(i) values

of the leached bulk rock samples are identical to those of the clinopyroxene samples (Fig.
AC

3d), although the 176Hf/177Hf ratio of global seawater after 27 Ma is known to range from

0.282717 to 0.282965 (David et al., 2001).

After 25 Ma, the Pacific seawater had similar Pb isotope ratios (206Pb/204Pb =

18.6–18.8, 207Pb/204Pb = 15.58–15.7, and 208Pb/204Pb = 38.4–38.8) to the sediments in the

206
Pacific Ocean basin (Ling et al., 1997; Miyazaki et al., 2016). The Pb/204Pb(i),
207
Pb/204Pb(i), and 208Pb/204Pb(i) ratios of the Rairua and Anatonu bulk rock samples do
 ACCEPTED MANUSCRIPT

not show shifts toward the values of the Pacific seawater but the clinopyroxene samples

occasionally do (Fig. 4a and 4b). Therefore, Pb isotopes in the strongly leached bulk

rocks are not affected by seawater alteration, which is consistent with our previous work

(Hanyu et al., 2011).

PT
4.4. Initial Sr–Nd–Hf–Pb isotope compositions

The bulk rock samples of the Rairua basalts (9 Ma; HIMU-type) have fairly uniform εNd(i)

RI
and εHf(i) values, with only slight variations in Sr(i). The Rairua clinopyroxene samples

SC
entirely overlap with the bulk rock samples in terms of Nd(i), εHf(i), and Sr(i) values,

within the analytical error (Fig. 3a–3f). There was no systematic variation observed
NU
between the pairs of the bulk rock and clinopyroxene samples from the same basalt lava.

The bulk rock samples from the Anatonu basalts (6 Ma; DM-type) also entirely overlap
MA

with the clinopyroxene samples. Their compositions are more radiogenic in Sr(i), εNd(i),

and εHf(i) than those of Rairua. The Sr(i) values show a wider variation than in Rairua,
D

suggesting the presence of an enriched component (Fig. 3b and 3d).

E

The Pb(i) isotope compositions of the Rairua bulk rock and clinopyroxene
PT

samples are characterized by typical HIMU composition, whereas those of the Anatonu

are closer to the DM composition (Fig. 4a and 4b, see also Fig. 3). The Rairua bulk rock
CE

samples have more radiogenic Pb(i) than the clinopyroxene samples. The two fields

partially overlap, but the latter systematically plots within less radiogenic field (Fig. 4c
AC

and 4d). The Anatonu bulk rock and clinopyroxene samples show a complete overlap in
207
Pb/204Pb(i) ratios. However, the 206
Pb/204Pb(i) and 208
Pb/204Pb(i) ratios of the

clinopyroxene samples plot within slightly more unradiogenic fields than those of the

bulk rock (Fig. 4e and 4f).

4.5. Implication for magma chamber processes

 ACCEPTED MANUSCRIPT

4.5.1. Clinopyroxene composition

Clinopyroxene records chemical composition of an early stage magma chamber before

fractional crystallization takes place and bulk rock compositions evolves. To test this

hypothesis, we calculated clinopyroxene composition in equilibrium with the bulk rock

composition of Raivavae basalts. We used the MELTS algorithm (Ghiorso et al., 2002)

PT
assuming basalt crystallization in a shallow magma chamber at a pressure of 0.1 GPa,

NNO oxygen buffer, and 0.2 wt% H2O in the basalt. The measured Mg#meas. values (Mg#

RI
= Mg/[Mg+Fe] molar ratio) of clinopyroxene samples range from 0.76 to 0.83. Calculated

SC
clinopyroxene Mg#calc. values range from 0.72 to 0.83. These values show linear positive

correlations with a regression line Mg#calc. = 0.98 Mg#meas. (r2 = 0.60), common to both
NU
Rairua and Anatonu samples, indicating the Mg-rich unfractionated nature of the

clinopyroxene samples (Table 3). This assures the following discussions based on the
MA

isotopic composition of clinopyroxene samples which represent early stage magma

chamber composition.
E D

4.5.2. Crustal assimilation

PT

The isotopic difference in Pb, between clinopyroxene and bulk rock samples, implies a

modification to the magma compositions after clinopyroxene crystallization.

CE

Assimilation from oceanic crust, either in the magma chamber or in the magma conduit,

is a possible scenario for the isotopic modification. The isotopic composition of the
AC

oceanic crust beneath Raivavae could be approximated by that of the Pacific MORB. It

has unradiogenic Pb isotopes defining trends parallel to the Northern Hemisphere

Reference Line (NHRL: Zindler and Hart, 1986) in terms of Pb isotopes (Fig. 4a and 4b).

In contrast, the overall trends of Raivavae basalts and the trends defined by the

clinopyroxene and bulk rock samples are oblique to the NHRL. Because they do not point

to the isotopic field of the Pacific MORB, it cannot be the source of assimilation in
 ACCEPTED MANUSCRIPT

Raivavae.

The Pb(i) values of some Anatonu clinopyroxenes are closer to Pacific seawater

than those of their host bulk rock samples. They are less radiogenic in 206Pb/204Pb(i) and
208
Pb/204Pb(i), and slightly radiogenic in 207Pb/204Pb(i) at a given 206Pb/204Pb(i) (Fig. 4c–

4f). The assimilation of brine or a brine-impregnated oceanic crust could explain the

PT
isotopic variations between the clinopyroxene and bulk rock samples. Lassiter et al.

(2002) reported the existence of olivine-hosted melt inclusions with high chlorine

RI
contents, indicating the assimilation of Cl-rich brine or brine-impregnated oceanic crust.

SC
Covariation of Pb and Cl was also reported from high-Cl glass shards found in Hawaii

(Coombs et al., 2004). However, Lassiter et al. (2003) suggested that such assimilation is
NU
unlikely to be the major cause of the observed isotopic variation because Cl-rich melt

inclusions are only observed in one Rairua basalt containing radiogenic Pb isotopes, and
MA

no melt inclusions from Anatonu basalts were enriched in Cl. Moreover, Pb in seawater

is so low in concentration that it is difficult to affect Pb isotope compositions either in

D

altered oceanic crust or in brine (Hirahara et al., 2015). Thus, crustal assimilation
E

scenarios are not considered in the case of Raivavae.

PT

4.5.3. Mantle lithosphere

CE

The depleted components in the Raivavae basalts are, most probably, sourced from

chemically heterogeneous lithospheric or asthenospheric mantle. There is growing

AC

evidence that small-scale enriched components are widely distributed in the depleted

upper mantle (e.g., Machida et al., 2009; Pilet, 2015). Because the Anatonu basalts have

isotopic compositions close to those of the off-ridge alkalic seamounts along the East

Pacific Rise (Graham et al., 1988), Lassiter et al. (2003) and Maury et al. (2013) suggested

that fusible components, such as pyroxenite veins in the upper mantle lithosphere, may

be involved in the basalt source. We know that: (1) the clinopyroxene samples are less
 ACCEPTED MANUSCRIPT

radiogenic in Pb than the hosts in both the Rairua and Anatonu basalts, and (2) the Pb

isotopic compositions of the Anatonu clinopyroxenes shift to the field defined by the

Pacific alkalic seamounts (Fig. 4a and b). These facts suggest that: (1) the melts

containing unradiogenic Pb are from fusible pyroxenite veins in the mantle lithosphere

and originated from impregnated Pacific alkali basalts, and (2) the melts are more strongly

PT
recorded in the clinopyroxenes than in the bulk rocks.

If we assume that an upwelling mantle plume encountered lithospheric mantle

RI
with enriched veins (e.g., Pilet et al., 2011), then the veins would have been selectively

SC
melted and subsequently mixed with the melts derived from main body of the mantle

plume. In this scenario, the early stage magma chambers, which crystallized the
NU
clinopyroxenes, were more heavily influenced by the vein melts. The magma chambers

would have grown by the replenishment of the melts from the main body of the plume
MA

giving rise to more radiogenic isotopic Pb compositions represented by the bulk rocks.
D

4.5.4. HIMU- and DM-type plume source

E

Although lithospheric pyroxenite with Pacific alkali seamount composition can be a

PT

partial source of unradiogenic Pb in the Raivavae basalts, we still need a discrete DM-

type source which should have more unradiogenic 207Pb and radiogenic 208Pb for both the
CE

Rairua and Anatonu basalts (Fig. 4a and 4b). The overall Pb isotopic variations in the

Rairua and Anatonu basalts require at least two major components: one from a HIMU-
AC

type source and another from a DM-type source (Lassiter et al., 2003; Maury et al., 2013).

Therefore, the Rairua magma was largely affected by the HIMU-type melts of plume

origin, in contrast to only a small contribution of DM-type lithospheric melts, whereas

the Anatonu magma is far less affected by the HIMU-type plume melts, with DM-type

mantle melts being dominant. In detail, the offsets between the clinopyroxene and bulk
206
rock samples in Pb/204Pb–207Pb/204Pb space are more robust in the DM-type Anatonu
 ACCEPTED MANUSCRIPT

basalts than in the HIMU-type Rairua basalts (Fig. 4c and 4e).

Moreover, the Anatonu basalts show wider isotopic variations in Sr(i)–Nd(i)

space, relative to the Rairua basalts (Fig. 3b). These observations are again explained by

the model that describes how the Anatonu basalts were predominated by the melts derived

from the DM-type source. The Anatonu DM-type melt is intrinsically heterogeneous due

PT
to mixing between the melt from DM-type plume source and that from enriched veins in

the lithosphere, with Pacific alkalic seamount composition. Therefore, both the HIMU-

RI
type and DM-type mantle sources for Raivavae are most probably derived from a mantle

SC
plume.
NU
5. Conclusion

We determined the age-corrected Sr–Nd–Hf–Pb isotope compositions of strongly leached

MA

bulk rock powders and clinopyroxene separates from the ~9 Ma HIMU-type Rairua and

the ~6 Ma DM-type Anatomu basalt lavas on Raivavae. We also determined the P/D ratios
D

from the same sample solutions used for the Sr–Nd–Hf and Pb isotopic analyses. We
E

detected some inhomogeneity in these isotopes, either in the clinopyroxene or in the bulk
PT

rock samples, and concluded that these findings are not attributable to secondary

alteration after lava emplacement. The overall isotopic variations, defined by the bulk
CE

rock samples, are explained by the two plume source components: the HIMU-type and

DM-type sources. The clinopyroxene samples exhibit even less radiogenic isotopic Pb
AC

compositions than the bulk rock samples, suggesting a temporal change of isotopic

compositions within the magma chambers. Because the overall trends of Raivavae basalts

and the trends defined by the clinopyroxene and bulk rock samples do not point to the

isotopic field of the Pacific MORB, oceanic crust cannot be the source of assimilation in

Raivavae. The early stage magma chambers, in which clinopyroxene phenocrysts

crystallized, would have been predominated by the melts derived from fusible veins in
 ACCEPTED MANUSCRIPT

the depleted upper mantle lithosphere. The magma chambers would have been

subsequently replenished by the melts from the main body of mantle plumes: more

vigorously for the Rairua basalts and far less vigorous in the Anatonu. A detailed

examination of the isotopic composition in both the mineral and bulk rock samples is a

powerful tool in identifying the source mantle components and understanding the

PT
processes of melt supply and mixing that take place in magma chambers.

RI
Acknowledgements

SC
We thank K. Kudo, W. Zhang, H. Higuchi and M. Kanazawa, of JAMSTEC, for their

support in the laboratory work. Thanks also go to Kyoto Fission-Track Co. Ltd. for their
NU
preparation of clinopyroxene separates and offer of MIHL for Pb blank measurement. We

thank for two anonymous reviewers for their constructive comments, and C. Chauvel for
MA

editorial handling. J.-I.K. was funded by JSPS KAKENHI grants JP15H02148 and

JP16H01123. T.M. was funded by JSPS KAKENHI grant JP15K05357 and JP17H02987.
D

We would like to thank Editage (www.editage.jp) for English language editing.

E
PT

References

Baker, J., Peate, D., Waight, T., Meyzen, C., 2004. Pb isotopic analysis of standards and
CE

207
samples using a Pb–204Pb double spike and thallium to correct for mass bias

with a double-focusing MC-ICP-MS. Chem. Geol. 211, 275-303, DOI:

AC

10.1016/j.chemgeo.2004.06.030.

Bezard, R., Turner, S., Davidson, J.P., Macpherson, C.G., Lindsay, J.M., 2015. Seeing

through the effects of crustal assimilation to assess the source composition

beneath the southern Lesser Antilles arc. J. Petrol. 56(4), 815-844, DOI:

10.1093/petrology/egv018.

Bryce, J.G., DePaolo, D.J., Lassiter, J.C., 2005. Geochemical structure of the Hawaiian
 ACCEPTED MANUSCRIPT

plume: Sr, Nd, and Os isotopes in the 2.8 km HSDP-2 section of Mauna Kea

volcano. Geochemistry, Geophysics, Geosystems, 6(9), DOI:

10.1029/2004GC000809.

Chang, Q., Shibata, T., Sinotsuka, K., Yoshikawa, M., Tatsumi, Y., 2003. Precise

determination of trace elements in geological standard rocks using inductively

PT
coupled plasma mass spectrometry (ICP-MS), Frontier Research on Earth

Evolution 1, IFREE Report for 2001-2002: 357-362.

RI
Chauvel, C., McDonough, W., Guille, G., Maury, R., Duncan, R., 1997. Contrasting old

SC
and young volcanism in Rurutu Island, Austral chain. Chem. Geol. 139(1-4), 125-

143, DOI:10.1016/S0009-2541(97)00029-6.
NU
Coombs M.L, Sisson T.W. and Kimura J.-I., 2004.Ultra-high chlorine in submarine

Kilauea glasses: Evidence for assimilation of brine by magma. Earth and Planetary
MA

Science Letters, 217, 297-313, doi:10.1016/S0012-821X(03)00631-9.

David, K., Frank, M., O'Nions, R. K., Belshaw, N. S., Arden, J. W., 2001. The Hf isotope
D

composition of global seawater and the evolution of Hf isotopes in the deep

E

Pacific Ocean from Fe–Mn crusts. Chemical Geology, 178, 23-42.

PT

Deniel, C., Pin, C., 2001. Single-stage method for the simultaneous isolation

of lead and strontium from silicate samples for isotopic measurements.

CE

Anal. Chim. Acta. 426(1), 95-103, DOI: 10.1016/S0003-2670(00)01185-5.

Duncan, R. A., and McDougall, I., 1976. Linear volcanism in French Polynesia. J.
AC

Volcanol. Geotherm. Res., 1, 197-227.

Ghiorso, M. S., Hirschmann, M. M., Reiners, P. W., Kress, V. C., 2002. The pMELTS: A

revision of MELTS for improved calculation of phase relations and major element

partitioning related to partial melting of the mantle to 3 GPa. Geochem. Geophys.

Geosyst. 3, doi: 10.1029/2001GC000217.

Graham, D. W., Zindler, A., Kurz, M. D., Jenkins, W. J., Batiza, R., and Staudigel, H.,
 ACCEPTED MANUSCRIPT

1988. He, Pb, Sr and Nd isotope constraints on magma genesis and mantle

heterogeneity beneath young Pacific seamounts. Contrib. Mineral. Petrol. 99, 446-

463.

Hanyu, T., Nakamura, E., 2000. Constraints on HIMU and EM by Sr and Nd isotopes re-

examined. Earth Planets Space. 52, 61–70.

PT
Hanyu, T., Tatsumi, Y., Senda, R., Miyazaki, T., Chang, Q., Hirahara, Y., Takahashi, T.,

Kawabata, H., Suzuki, K., Kimura, J.-I., Nakai, S., 2011. Geochemical

RI
characteristics and origin of the HIMU reservoir: A possible mantle plume source

SC
in the lower mantle. Geochem. Geophys. Geosyst. 12(2),

DOI:10.1029/2010GC003252.
NU
Hanyu, T., Kawabata, H., Tatsumi, Y., Kimura, J.-I., Hyodo, H., Sato, K., Miyazaki, T.,

Chang, Q., Hirahara, Y., Takahashi, T., Senda, R., Nakai, S., 2014. Isotope
MA

evolution in the HIMU reservoir beneath St. Helena: Implications for the mantle

recycling of U and Th. Geochim. Cosmochim. Acta. 143, 232-252.

D

Hirahara, Y., Takahashi, T., Miyazaki, T., Vaglarov, B.S., Chang, Q., Kimura, J.-I.,
E

Tatsumi, Y., 2009. Precise Nd isotope analysis of igneous rocks using cation
PT

exchange chromatography and thermal ionization mass spectrometry (TIMS).

JAMSTEC Report of Research and Development, Special Issue: 65-72.

CE

Hirahara, Y., Chang, Q., Miyazaki, T., Takahashi, T., Senda, R., Kimura, J.-I., 2012.

Improved Nd chemical separation technique for 143Nd/144Nd analysis in

AC

geological samples using packed Ln resin columns. JAMSTEC Report of Research

and Development, 15: 25-31.

Hirahara, Y., Kimura, J.-I., Senda, R., Miyazaki, T., Kawabata, H., Takahashi, T., Chang,

Q., Vaglarov, B.S., Sato, T., Kodaira, S., 2015. Geochemical variations in Japan

Sea back-arc basin basalts formed by high-temperature adiabatic melting of

mantle metasomatized by sediment subduction components. Geochem. Geophys.

 ACCEPTED MANUSCRIPT

Geosyst. 16(5), 1324-1347, DOI: 10.1002/2015gc005720.

Hofmann, A. W., 2003. Sampling Mantle Heterogeneity through Oceanic Basalts:

Isotopes and Trace Elements in Treatise on Geochemistry, The Mantle and Core,

edited by R. E. Carlson, pp. 61-101, Elsevier, London, UK.

Horwitz, E.P., Dietza, M.L., Rhoadsa, S., Felintoa, C., Galeb, N.H., Houghtonb, J., 1994.

PT
A lead-selective extraction chromatographic resin and its application to the

isolation of lead from geological samples. Anal. Chim. Acta 292, 263-273, doi.

RI
10.1016/0003-2670(94)00068-9.

SC
Jackson, M.G., Kurz, M.D., Hart, S.R., Workman, R.K., 2007. New Samoan lavas from

Ofu Island reveal a hemispherically heterogeneous high 3He/4He mantle. Earth

NU
Planet. Sci. Lett. 264(3–4), 360-374, DOI: 10.1016/j.epsl.2007.09.023.
87
Jackson, M.G., Hart, S.R., Shimizu, N., and Blusztajn, J.S., 2009. The Sr/86Sr and
MA

143
Nd/144Nd disequilibrium between Polynesian hot spot lavas and the

clinopyroxenes they host: Evidence complementing isotopic disequilibrium in

D

melt inclusions. Geochem. Geophys. Geosyst. 10, Q03006,

E

DOI:10.1029/2008GC002324
PT

Jackson, M.G., Hart, S.R., Konter, J.G., Koppers, A.A.P., Staudigel, H., Kurz, M.D.,

Blusztajn, J., Sinton, J.M., 2010. Samoan hot spot track on a “hot spot highway”:
CE

Implications for mantle plumes and a deep Samoan mantle source. Geochem.

Geophys. Geosyst. 11(12), DOI: 10.1029/2010gc003232.

AC

Kawabata, H., Hanyu, T., Chang, Q., Kimura, J.-I., Nichols, A.R.L. Tatsumi, Y., 2011.

The petrology and geochemistry of St. Helena alkali basalts: Evaluation of the

oceanic crust-recycling Model for HIMU OIB. J. Petrol. 52(4), 791-838, DOI:

10.1093/petrology/egr003.

Kendrick, M.A., Jackson, M.G., Hauri, E.H., Phillips, D., 2015. The halogen (F, Cl, Br,

I) and H2O systematics of Samoan lavas: Assimilated-seawater, EM2 and high-

 ACCEPTED MANUSCRIPT

3
He/4He components. Earth Planet. Sci. Lett. 410: 197-209, DOI:

10.1016/j.epsl.2014.11.026.

Kent, A.J.R., Norman, M.D., Hutcheon, I.D., Stolper, E.M., 1999. Assimilation of

seawater-derived components in an oceanic volcano: evidence from matrix

glasses and glass inclusions from Loihi seamount, Hawaii. Chem. Geol. 156(1–

PT
4), 299-319, DOI: 10.1016/S0009-2541(98)00188-0.

Kimura, J.-I., Chang, Q., 2012. Origin of the suppressed matrix effect for improved

RI
analytical performance in determination of major and trace elements in anhydrous

SC
silicate samples using 200 nm femtosecond laser ablation sector-field inductively

coupled plasma mass spectrometry. J. Anal. Atom. Spectrom. 27, 1549-1559.

NU
Kimura, J.-I., Nakano, N., 2004. Precise lead isotope analysis using Multiple Collector-

Inductively Coupled Plasma-Mass Spectrometry (MC-ICP-MS): Analytical

MA

technique and evaluation of mass fractionation during Pb separation. Geoscience

Report of Shimane University, 23, 9-15.

D

Kimura, J.-I., Sisson, T. W., Nakano, N., Coombs, M. L., and Lipman, P. W., 2006. Isotope
E

geochemistry of early Kilauea magmas from the submarine Hilina bench: The
PT

nature of the Hilina mantle component. J. Volcanol. Geotherm. Res. 151, 51-72.

Kimura, J.-I., Miyazaki, T., Vaglarov, B.S., Haraguchi, S., Chang, Q., Gill, J.B., 2015.
CE

Reply to comment by I. Pineda-Velasco, T. T. Nguyen, H. Kitagawa, and E.

Nakamura on “Diverse magmatic effects of subducting a hot slab in SW Japan:

AC

Results from forward modeling”. Geochem. Geophys. Geosyst. 16(9), DOI:

10.1002/2015gc005984.

Lassiter, J.C., Blichert-Toft, J., Hauri, E.H., Barsczus, H.G., 2003. Isotope and trace

element variations in lavas from Raivavae and Rapa, Cook-Austral islands:

constraints on the nature of HIMU- and EM-mantle and the origin of mid-plate

volcanism in French Polynesia. Chem. Geol. 202, 115-138.

 ACCEPTED MANUSCRIPT

Lassiter, J.C., Hauri, E.H., Nikogosian, I.K., Barsczus, H.G., 2002. Chlorine–potassium

variations in melt inclusions from Raivavae and Rapa, Austral Islands: constraints

on chlorine recycling in the mantle and evidence for brine-induced melting of

oceanic crust. Earth Planet Sci. Lett. 202, 525-540.

Ling, H. F., Burton, K. W., O'Nions, R. K., Kamber, B. S., von Blanckenburg, F., Gibb,

PT
A. J., Hein, J. R., 1997. Evolution of Nd and Pb isotopes in Central Pacific

seawater from ferromanganese crusts. Earth and Planetary Science Letters, 146,

RI
1-12.

SC
Machida, S., Hirano, N., Kimura, J.-I., 2009. Evidence for recycled plate material in

Pacific upper mantle unrelated to plumes. Geochim. Cosmochim. Acta. 73, 3028-
NU
3037.

Maury, R.C., Guille, G., Guillou, H, Chauvel, C., Rossi, P., Pallares, C, Legendre, C.,
MA

2013. Temporal evolution of a Polynesian hotspot: New evidence from Raivavae

(Austral islands, South Pacific ocean). Bull. Soc. Géol. France. 184, 557-567.
D

McDonough, W. F., Chauvel, C., 1991. Sample contamination explains the Pb isotopic
E

composition of some Rurutu island and Sasha seamount basalts. Earth Planet. Sci.
PT

Lett. 105, 397–404.

Miyazaki T., Kimura J.-I., and Katakuse M., 2016. Geochemical records from loess
CE

deposits in Japan over the last 210 kyr: Lithogenic source changes and

paleoclimatic indications. Geochemistry Geophysics Geosystems, 17, 2745-2761,

AC

DOI: 10.1002/2016GC006322.

Miyazaki, T., Kimura, J.-I., Senda, R., Vaglarov, B.S., Chang, Q., Takahashi, T., Hirahara,

Y., Hauff, F., Hayasaka, Y., Sano, S., Shimoda, G., Ishizuka, O., Kawabata, H.,

Hirano, N., Machida, S., Ishii, T., Tani, K., Yoshida., T., 2015. Missing western

half of the Pacific Plate: Geochemical nature of the Izanagi-Pacific Ridge

interaction with a stationary boundary between the Indian and Pacific mantles.
 ACCEPTED MANUSCRIPT

Geochem. Geophys. Geosyst. 16(9): 3309-3332, DOI: 10.1002/2015gc005911.

Miyazaki, T., Vaglarov, B.S., Takei, M., Suzuki, M., Suzuki, H., Ohsawa, K., Chang, Q.,

Takahashi, T., Hirahara, Y., Hanyu, T., Kimura, J.-I., Tatsumi, Y., 2012.

Development of a fully automated open-column chemical-separation system—

COLUMNSPIDER—and its application to Sr-Nd-Pb isotope analyses of igneous

PT
rock samples. J. Mineral. Petrol. Sci. 107: 74-86.

Miyazaki, T., Kanazawa, N., Takahashi, T., Hirahara, Y., Vaglarov, B.S., Chang, Q.,

RI
Kimura, J.-I, Tatsumi, Y., 2009. Precise Pb isotope analysis of igneous rocks using

SC
fully-automated double spike thermal ionization mass spectrometry (FA -DS-

TIMS). JAMSTEC Report of Research and Development, Special Issue: 73-80.

NU
Nobre Silva, I.G., Weis, D., Scoates, J.S., Barling, J., 2013. The Ninetyeast Ridge and its

Relation to the Kerguelen, Amsterdam and St. Paul Hotspots in the Indian Ocean.
MA

J. Petrol., 54, 1177-1210.

Nobre Silva, I.G., Weis, D., Scoates, J.S., 2010. Effects of acid leaching on the Sr‐Nd‐Hf
D

isotopic compositions of ocean island basalts. Geochem. Geophys. Geosyst. 11,

E

Q09011, doi:10.1029/2010GC003176
PT

Nobre Silva, I.G., Weis, D., Barling, J., Scoates, J.S., 2009. Leaching systematics and

matrix elimination for the determination of high‐precision Pb isotope compositions

CE

of ocean island basalts. Geochem. Geophys. Geosyst. 10, Q08012,

doi:10.1029/2009GC002537
AC

Patchett, P.J., Vervoort, J.D., Söderlund, U., Salters, V.J.M., 2004. Lu–Hf and Sm–Nd

isotopic systematics in chondrites and their constraints on the Lu–Hf properties of

the Earth. Earth Planet. Sci. Lett. 222(1): 29-41, DOI: 10.1016/j.epsl.2004.02.030.

Pilet, S., 2015. Generation of low-silica alkaline lavas: Petrological constraints, models,

and thermal implications, in Foulger, G.R., Lustrino, M., and King, S.D., eds., The

Interdisciplinary Earth: A Volume in Honor of Don L. Anderson: Geological

 ACCEPTED MANUSCRIPT

Society of America Special Paper 514 and American Geophysical Union Special

Publication 71, p. 281–304, doi:10.1130/2015.2514(17).

Pilet S., Baker, M.B., Muntener, O, and Stolper, E.M., 2011. Monte Carlo simulations of

metasomatic enrichment in the lithosphere and implications for the source of

alkaline basalts. J. Petrol. 52, 1415-1442, DOI: 0.1093/petrology/egr007.

PT
Richter, F.M., Rowley, D.B., and DePaolo,D.J. 1992. Sr isotope evolution of seawater:

the role of tectonics. Earth and Planetary Science Letters, 109, 11-23.

RI
Rudge, J.F., Maclennan, J., Stracke, A., 2013. The geochemical consequences of mixing

SC
melts from a heterogeneous mantle. Geochim. Cosmochim. Acta. 114, 112-143,

DOI: 10.1016/j.gca.2013.03.042.
NU
Staudigel, H., Zindler, A., Hart, S.R., Leslie, T., Chen, C.Y., Clague, D., 1984. The isotope

systematics of a juvenile intraplate volcano: Pb, Nd, and Sr isotope ratios of

MA

basalts from Loihi Seamount, Hawaii. Earth and Planetary Science Letters, 69,

13-29.
D

Stracke, A., Bizimis, M., Salters, V.J.M., 2003. Recycling oceanic crust: Quantitative
E

constraints. Geochem. Geophys. Geosyst. 4(3), DOI: 10.1029/2001gc000223.

PT

Stracke, A., Hofmann, A.W., Hart, S.R., 2005. FOZO, HIMU, and the rest of the mantle

zoo. Geochem. Geophys. Geosyst. 6(5), DOI: 10.1029/2004gc000824.

CE

Stracke, A. and B. Bourdon, 2009. The importance of melt extraction for tracing mantle

heterogeneity. Geochim. Cosmochim. Acta, 73, 218-238,

AC

DOI:10.1016/j.gca.2008.10.015.

Takahashi, T., Hirahara, Y., Miyazaki, T., Vaglarov, B.S., Chang, Q., Kimura, J.-I.,

Tatsumi, Y., 2009. Precise determination of Sr isotope ratios in igneous rock

samples and application to micro-analysis of plagioclase phenocrysts. JAMSTEC

Report of Research and Development, Special Issue: 59-64.

Takamasa, A., Nakai, S., 2009. Contamination introduced during rock sample powdering:
 ACCEPTED MANUSCRIPT

Effects from different mill materials on trace element contamination. Geocem. J.

43: 389-394.

Tani, K., Kawabata, H., Chang, Q., Sato, K. Tatsumi, Y., 2006. Quantitative analyses of

silicate rock major and trace elements by X-ray fluorescence spectrometer:

Evaluation of analytical precision and sample preparation. Front. Res. Earth Evol.,

PT
2, 1-8.

Thompson, P.M.E., Kempton, P.M., Kerr, A.C., 2008. Evaluation of the effects of

RI
alteration and leaching on Sm‐Nd and Lu‐Hf systematics in submarine mafic rocks.

SC
Lithos. 104: 164–176.

Todd, E., Stracke, A., Scherer, E.E., 2015. Effects of simple acid leaching of crushed and
NU
powdered geological materials on high-precision Pb isotope analyses. Geochem.

Geophys. Geosyst., 16, 2276-2302, doi:10.1002/2015GC005804.

MA

Weis, D., Frey, F.A., 1991. Isotope geochemistry of the Ninetyeast Ridge basement

basalts: Sr, Nd, and Pb evidence for involvement of the Kerguelen hot spot. Proc.
D

Ocean Drill. Program Sci. Results, 121: 591–610.

E

Weis, D., Frey, F.A., 1996. Role of the Kerguelen Plume in generating the eastern Indian
PT

Ocean seafloor. J. Geophys. Res., 101(B6), 13: 381 –13,849.

Weis D., Kieffer, B., Maerschalk, C., Pretorius, W., Barling, J., 2005. High-precision Pb-
CE

Sr-Nd-Hf isotopic characterization of USGS BHVO-1 and BHVO-2 reference

materials. Geochem. Geophys. Geosyst. 6, Q02002, doi:10.1029/2004GC000852.

AC

Weis, D., Kieffer, B., Maerschalk, C., Barling, J., de Jong, J., Williams, G.A., Hanano, D.,

Pretorius, W., Mattielli, N., Scoates, J.S., Goolaerts, A., Friedman, R.M.,

Mahoney, J.B., 2006. High-precision isotopic characterization of USGS reference

materials by TIMS and MC-ICP-MS. Geochem. Geophys. Geosyst. 7(8): DOI:

10.1029/2006gc001283.

Willbold, M., Stracke, A., 2006. Trace element composition of mantle end-members:
 ACCEPTED MANUSCRIPT

Implications for recycling of oceanic and upper and lower continental crust.

Geochem. Geophys. Geosyst. 7(4): DOI: 10.1029/2005gc001005.

Wilson, M., 1989. IGNEOUS PTROGENESIS. Unwin Hyman, London, 466 pp,

Yokoyama, T., Makishima, A., Nakamura, E., 1999. Evaluation of the coprecipitation of

incompatible trace elements with fluoride during silicate rock dissolution by acid

PT
digestion. Chem. Geol. 157: 175-187.

Zindler A., Hart S., 1986. Chemical geodynamics. Annual Review in Earth and Planetary

RI
Sciences, 14, 493-571, doi:10.1146/annurev.ea.14.050186.002425.

Figure and Table captions

SC
NU
Fig 1: Map showing the Cook-Austral Islands (upper panel) and Raivavae Island (lower

panel). The Raivavae fracture zone is shown by a thick dashed line in the upper
MA

panel. The volcanic sequences in the lower panel are modified after Maury et al.

(2013). The samples were collected from the localities shown by the filled (HIMU-
D

type) and open (DM-type) circles marked with sample numbers.

E
PT

Fig 2: Δ206Pb/204Pb(i) vs. Δ208Pb/204Pb(i) diagram of the clinopyroxene samples. The Δ

values are calculated from the absolute differences [Δ = (initial Pb isotope ratios
CE

calculated using P/D ratios from the Sr-Nd-Hf isotope solution) – (initial Pb

isotope ratios calculated using P/D ratios from the Pb isotope solution)]. The box
AC

indicates two standard deviations of repeated measurements of JB-2, which gives

an estimate for the analytical reproducibility in the 206Pb/204Pb and 208Pb/204Pb ratio.

Fig 3: Age-corrected Sr(i), Nd(i) and Hf(i) values of the leached bulk rock and

clinopyroxene samples. The black tie lines connect the clinopyroxene (CPX) and

bulk rock (BR) data from the same sample. The red symbols (CPX) show the
 ACCEPTED MANUSCRIPT

strongly leached results, using the same leaching method applied for bulk rock.

The red tie-lines connect to the corresponding weakly leached CPX sample. The

error bars indicate a typical two-standard deviation (2SD) reproducibility of the

JB-2 standard. (a) Sr(i)–Nd(i) plot, (b) close up of the dashed square areas in the

(a), (c)Nd(i)–Hf(i) plots, (d) close up of the dashed square area in (c), (e) close up

PT
of the Rairua results in (d), (f) close up of the Anatonu results in (d). The isotopic

range of the Pacific alkalic seamounts are shown by the dashed lines in (b)

RI
(Graham et al., 1998). The blue arrow in (b) points to the Pacific Ocean seawater,

SC
which plots outside the panel.
NU
Fig 4: Age-corrected Pb(i) values of the leached bulk rock and clinopyroxene samples.

The black tie lines connect CPX and BR from the same sample. The red symbols
MA

(CPX) show the strongly leached clinopyroxene results corresponding to the

respective weakly leached samples. The blue symbols (CPX) show the strongly
D

leached clinopyroxene, without using the MIHL, results corresponding to the

E

respective weakly leached samples. The green (CPX) and red (BR) fields highlight
PT

the isotopic compositions in the Rairua basalts. The blue (CPX) and orange (BR)

fields indicate the isotopic compositions in the Anatonu basalts. The error bars
CE

206
indicate the typical 2SD reproducibility of the JB-2 standard. (a) Pb/204Pb(i)–
207
Pb/204Pb(i) plot, (b) 206
Pb/204Pb(i)–208Pb/204Pb(i) plot, (c) close up of Rairua
AC

results in (a), (d) close up of the Rairua results in (b), (e) close up of the Anatonu

results in (a), (f) close up of the Anatonu results in (b). The isotopic ranges of the

Pacific alkalic seamounts and Pacific MORB are shown by the dashed and solid

lines, respectively, in (a) and (b) (Graham et al., 1998). The isotopic ranges of the

Pacific Ocean seawater are shown by the dashed blue lines in (a) and (b).
 ACCEPTED MANUSCRIPT

Table 1: Measured and age-corrected Sr–Nd–Hf–Pb isotope ratios of the leached bulk

rock samples.

Table 2: Measured and age-corrected Sr–Nd–Hf–Pb isotope ratios of the clinopyroxene

samples.

PT
Table 3: Major element composition of bulk rock and clinopyroxene core samples.

RI
SC
Table S1: Measured Pb isotope ratios of BHVO-2 basalt standard with or without

pulverization using an alumina mill.

NU
Table S2: Measured Sr-Nd-Pb-Hf isotope ratios of the clinopyroxene samples after weak
MA

and strong leaching treatments, and the clinopyroxene samples and leachates

without using the MIHL after strong leaching treatment.

E D

Table S3: Measured Sr–Nd–Hf–Pb isotope ratios of JB-2 basalt standard.

PT
CE
AC
 ACCEPTED MANUSCRIPT

Table 1 Measured and age-corrected Sr–Nd–Hf–Pb isotope ratios of the leached bulk rock samples

Unit Rairua Anatonu

Sample RAV-05LRAV-08LRAV-11LRAV-20LRAV-32LRAV-34LRAV-41LRAV-06LRAV-14LRAV-17LRAV-18LRAV-21LRAV-23LRAV-27LRAV-30LRAV-37LRAV-38LRAV-39L
A 9 9 9 9 9 9 9 6 6 6 6 6 6 6 6 6 6 6
g
e
f
o
r
i
n

PT
i
t
i
a

RI
l
r
a

SC
t
i
o
[
M
NU
a
]
[Measured isotope ratios]
87Sr/ 86Sr 0.702840
0.7028320.7028360.7028460.7028350.7028260.7028720.7029810.7029510.7030230.7029030.7030000.7029450.7029880.7030210.7029360.7028900.702915
MA

2SE 0.0000060.0000060.0000060.0000070.0000060.0000060.0000110.0000070.0000060.0000060.0000060.0000060.0000060.0000070.0000070.0000050.0000070.000007
143Nd/144Nd
0.5129150.5129180.5129230.5129140.5129060.5129070.5129130.5129980.5129990.5129810.5130010.5129850.5129980.5129800.5129750.5129740.5129940.513001
2SE 0.0000050.0000060.0000090.0000070.0000080.0000040.0000110.0000030.0000070.0000070.0000100.0000060.0000060.0000090.0000050.0000090.0000070.000010
εNd 5.57 5.62 5.72 5.55 5.38 5.41 5.51 7.17 7.20 6.85 7.24 6.93 7.18 6.82 6.73 6.71 7.11 7.24
176Hf/177Hf
0.2829660.2829730.2829700.2829660.2829770.2829800.2829710.2830670.2830700.2830590.2830610.2830570.2830730.2830740.2830490.2830690.2830660.283075
2SE 0.0000100.0000050.0000070.0000040.0000070.0000070.0000050.0000060.0000120.0000080.0000070.0000070.0000060.0000150.0000070.0000060.0000060.000005
D

εHf 6.39 6.64 6.53 6.39 6.79 6.90 6.58 9.96 10.07 9.70 9.76 9.62 10.17 10.23 9.35 10.04 9.93 10.25
206Pb/204Pb
20.8604 20.8621 20.9233 20.8584 21.1065 21.1353 20.8313 19.5983 19.7101 19.3091 19.7145 19.6146 19.5290 19.5756 19.6117 19.7221 19.7060 19.4927
E

2SE 0.0009 0.0009 0.0009 0.0011 0.0009 0.0010 0.0016 0.0009 0.0011 0.0009 0.0010 0.0009 0.0009 0.0008 0.0008 0.0009 0.0008 0.0009
207Pb/204Pb
15.7292 15.7286 15.7368 15.7268 15.7560 15.7602 15.7221 15.5922 15.6062 15.5855 15.6015 15.5966 15.5892 15.5922 15.6054 15.6112 15.6001 15.5791
PT

2SE 0.0008 0.0010 0.0011 0.0012 0.0009 0.0010 0.0014 0.0010 0.0013 0.0010 0.0012 0.0011 0.0011 0.0009 0.0007 0.0009 0.0009 0.0012
208Pb/204Pb
40.1662 40.1857 40.2083 40.1698 40.3755 40.3975 40.1872 39.2996 39.3791 39.0464 39.3648 39.2890 39.2341 39.2639 39.3107 39.3633 39.3211 39.1986
2SE 0.0031 0.0037 0.0040 0.0044 0.0031 0.0036 0.0039 0.0035 0.0047 0.0039 0.0041 0.0040 0.0038 0.0034 0.0024 0.0031 0.0030 0.0045
[Age-corrected isotope ratios]
CE

Rb (ppm)14.8 12.6 14.1 11.3 7.63 7.86 2.07 7.01 5.71 8.98 6.9 7.64 23.3 12.3 8.11 17.3 17.5 13.2
Sr (ppm)542 442 788 391 567 534 227 238 176 380 205 260 482 312 232 545 426 421
87Sr/ 86Sr(i)
0.7028300.7028210.7028290.7028350.7028300.7028210.7028690.7029740.7029430.7030170.7028950.7029930.7029330.7029780.7030120.7029290.7028800.702908
Sm (ppm)
5.93 6.12 9.79 5.91 8.45 8.12 4.68 2.29 2.11 2.98 2.21 2.48 1.34 2.31 2.78 3.25 3.04 2.48
AC

Nd (ppm)26.8 27.6 44.3 26 42.2 39.9 20.3 7.91 6.57 11.2 7.53 8.40 5.28 8.00 9.89 12.6 11.1 8.25
143Nd/144Nd(i)
0.5129080.5129100.5129150.5129060.5128980.5129000.5129040.5129910.5129910.5129750.5129940.5129780.5129920.5129730.5129680.5129680.5129880.512994
εNd(i) 5.64 5.69 5.79 5.61 5.46 5.49 5.58 7.19 7.20 6.88 7.25 6.95 7.21 6.84 6.75 6.74 7.13 7.26
Lu (ppm)0.26 0.26 0.24 0.25 0.29 0.28 0.22 0.11 0.11 0.12 0.12 0.13 0.08 0.13 0.13 0.19 0.17 0.16
Hf (ppm)5.43 5.46 8.37 5.25 6.00 5.98 4.15 3.47 2.20 3.36 3.37 3.57 5.44 3.29 3.36 4.60 2.52 3.75
176Hf/177Hf(i)
0.2829650.2829720.2829690.2829650.2829760.2829790.282969618
0.2830660.2830690.2830590.2830600.2830560.2830720.2830740.2830490.2830680.2830650.283074
εHf(i) 6.55 6.81 6.71 6.55 6.95 7.06 6.73 10.07 10.17 9.81 9.87 9.73 10.30 10.34 9.46 10.15 10.02 10.36
U (ppm) 0.32 0.19 0.54 0.13 1.08 0.94 0.07 0.18 0.06 0.31 0.45 0.23 1.10 0.18 0.06 0.94 0.36 0.28
Th (ppm)1.02 0.89 1.35 0.59 4.82 4.12 0.44 0.40 0.12 0.99 0.90 0.47 2.46 0.42 0.22 2.61 0.96 0.82
Pb (ppm)1.08 0.53 1.41 0.64 1.53 1.55 0.18 0.34 0.14 0.69 0.45 0.36 1.77 0.37 0.19 1.92 0.73 0.69
206Pb/204Pb(i)
20.832 20.828 20.888 20.840 21.040 21.078 20.796 19.567 19.686 19.282 19.653 19.576 19.491 19.5461 19.593 19.692 19.676 19.468
207Pb/204Pb(i)
15.728 15.727 15.735 15.726 15.753 15.758 15.720 15.591 15.605 15.584 15.599 15.595 15.587 15.5908 15.605 15.610 15.599 15.578
 ACCEPTED MANUSCRIPT

2 4 4 4 4 4 4 4 3 3 3 3 3 3 3 3 3 3 3
0 0 0 0 0 0 0 0 9 9 9 9 9 9 9 9 9 9 9
8 . . . . . . . . . . . . . . . . . .
P 1 1 1 1 2 3 1 2 3 0 3 2 2 2 2 3 2 1
b 3 3 7 4 7 1 1 7 6 1 2 6 0 4 8 3 9 7
/ 7 4 9 1 7 5 0 6 2 8 5 3 6 1 7 6 5 5
2 3
0

P
b
(

PT
i
)

2SE: two-standard errors.

RI
Table 2 Measured and age-corrected Sr–Nd–Hf–Pb isotope ratios of the clinopyroxene samples

Uni R

SC
t a
ir
u
a
NU
Sa R R R R R R R R R R
mpl A A A A A A A A A A
e V V V V V V V V V V
- - - - - - - - - -
0 0 1 1 1 2 2 3 4 4
MA

5 8 1 1 1 0 0 3 1 1
c c c c c c c c c c
s s s s
* w * w
* *
* *
D

Ag 9 9 9 9 9 9 9 9 9 9
e
E

for
initi
PT

al
rati
o
[Ma
CE

[Me
asu
red
isot
AC

ope
rati
os]
87
Sr 0 0 0 0 0 0 0 0
/86S . . . . . . . .
r 7 7 7 7 7 7 7 7

0 0 0 0 0 0 0 0
2 2 2 2 2 2 2 2
8 8 8 8 8 8 8 9

6 7 2 4 6 3 1 0
4 1 8 2 2 9 2 3

2S 0 0 0 0 0 0 0 0
E . . . . . . . .
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
 ACCEPTED MANUSCRIPT

0 0 0 1 0 1 0 0

6 6 7 3 6 2 5 6
143
0 0 0 0 0 0 0 0
Nd/ . . . . . . . .
144
5 5 5 5 5 5 5 5
Nd 1 1 1 1 1 1 1 1
2 2 2 2 2 2 2 2
9 9 9 9 9 9 9 9
2 2 1 2 2 1 1 2
6 3 7 7 1 0 2 2

2S 0 0 0 0 0 0 0 0
E . . . . . . . .

PT
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0

RI
0 0 0 1 0 1 0 0
7 6 6 1 5 0 7 6

εNd 5 5 5 5 5 5 5 5
. . . .

SC
. . . .
7 7 5 8 6 4 5 6

7 1 9 0 8 6 0 9
176
0 0 0 0 0 0 0 0

Hf/1 . . . . . . . .
NU
77
Hf 2 2 2 2 2 2 2 2
8 8 8 8 8 8 8 8

2 2 2 2 2 2 2 2
9 9 9 9 9 9 9 9
7 7 8 7 5 6 8 8
MA

5 5 6 7 8 9 4 5

2S 0 0 0 0 0 0 0 0
E . . . . . . . .
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
D

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
E

5 4 6 5 6 6 4 5

εHf 6 6 7 6 6 6 7 7
PT

. . . . . . . .
7 7 0 7 1 4 0 0
4 2 9 9 3 9 5 7
206
P 2 2 2 2 2 2 2 2 2 2
CE

b/20 0 0 0 0 0 0 0 0 0 0
4
Pb . . . . . . . . . .
8 8 8 8 8 8 8 7 7 6

5 5 9 8 5 4 4 6 2 2
8 5 6 5 9 6 3 4 0 2
AC

5 2 0 4 2 2 0 7 7 5

2S 0 0 0 0 0 0 0 0 0 0
E . . . . . . . . . .

0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
1 1 1 5 0 1 2 1 2 0

1 1 2 1 9 2 9 1 2 8
207
P 1 1 1 1 1 1 1 1 1 1
b/20 5 5 5 5 5 5 5 5 5 5
4
Pb . . . . . . . . . .
7 7 7 7 7 7 7 7 7 7
3 3 3 3 4 3 3 2 1 1

2 0 4 7 0 1 7 2 6 9
8 2 1 6 0 9 1 2 7 4

2S 0 0 0 0 0 0 0 0 0 0
E . . . . . . . . . .
0 0 0 0 0 0 0 0 0 0
 ACCEPTED MANUSCRIPT

0 0 0 0 0 0 0 0 0 0

0 0 0 3 0 0 2 0 0 0
8 9 9 4 7 9 2 8 7 7
208
P 4 4 4 4 4 4 4 4 4 4
b/20 0 0 0 0 0 0 0 0 0 0
4
Pb . . . . . . . . . .
1 1 2 2 2 1 1 0 0 0
8 7 0 3 0 7 9 8 8 4
2 3 5 5 8 3 1 9 1 1
7 0 3 2 2 0 3 7 2 5

2S 0 0 0 0 0 0 0 0 0 0
E . . . . . . . . . .

PT
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
2 2 2 9 2 2 5 2 2 2

3 5 4 3 1 5 8 4 9 0

RI
[Ag
e-
cor
rect

SC
ed
isot
ope
rati
NU
os]
Rb 0 0 0 0 0 0

(pp . . . . . .
m) 5 3 2 3 1 1
4 5 7 3 8 9
MA

4 4 6 3 8 2

Sr 6 6 1 6 6 5
(pp 8 4 1 3 1 2
m) . . 6 . . .
8 6 9 8 3
D

87
Sr 0 0 0 0 0 0 0 0
/86S . . . . . . . .
r(i) 7 7 7 7 7 7 7 7
E

0 0 0 0 0 0 0 0
2 2 2 2 2 2 2 2
PT

8 8 8 8 8 8 8 9
6 6 2 4 6 3 1 0
1 9 7 2 0 9 1 2

Sm 3 3 7 3 3 2
CE

(pp . . . . . .
m) 5 2 0 3 2 9
1 6 3 6

Nd 1 1 2 1 1 1
(pp 3 2 8 2 1 0
AC

m) . . . . . .
4 4 2 6 7 7
143
0 0 0 0 0 0 0 0

Nd/ . . . . . . . .
144
5 5 5 5 5 5 5 5
Nd( 1 1 1 1 1 1 1 1

i) 2 2 2 2 2 2 2 2
9 9 9 9 9 9 9 9
1 1 0 1 1 0 0 1

7 4 8 8 2 0 2 2

ε 5 5 5 5 5 5 5 5

Nd (i) . . . . . . . .
8 7 6 8 7 5 5 7
1 6 4 5 2 0 3 3

Lu 0 0 0 0 0 0
(pp . . . . . .
m) 1 1 1 1 1 1
 ACCEPTED MANUSCRIPT

3 5 3 3 4

Hf 3 2 6 3 2 2
(pp . . . . . .
m) 2 8 5 0 7 5
1 1 0 1 6
176
0 0 0 0 0 0 0 0

Hf/1 . . . . . . . .
77
Hf 2 2 2 2 2 2 2 2
(i) 8 8 8 8 8 8 8 8

2 2 2 2 2 2 2 2
9 9 9 9 9 9 9 9
7 7 8 7 5 6 8 8

PT
4 4 5 7 7 8 3 4

ε 6 6 7 6 6 6 7 7

Hf (i) . . . . . . . .
9 8 2 9 2 6 2 2

RI
0 9 8 8 9 6 1 3

U ( 0 0 0 0 0 0
(pp P . . . . . .

SC
m) b 0 0 0 0 0 0

s 4 3 5 3 2 2
o
l

u
NU
ti
o

n
)

Th ( 0 0 0 0 0 0
MA

(pp P . . . . . .
m) b 1 1 2 1 1 0
s 8 4 6 3 1 9
o
l
D

ti
o
E

)
PT

Pb ( 0 0 0 0 0 0
(pp P . . . . . .
m) b 1 0 1 0 0 0
s 0 6 0 8 6 4
CE

o
l
u

ti
o
AC

)
206
P ( 2 2 2 2 2 2 2 2 2 2
b/20 P 0 0 0 0 0 0 0 0 0 0
4
Pb b . . . . . . . . . .
(i) s 8 8 8 8 8 8 8 7 6 5
o 2 1 4 3 1 1 1 3 7 7

l 4 4 8 8 1 4 0 7 6 8
u
ti

o
n
)
207
P ( 1 1 1 1 1 1 1 1 1 1
b/20 P 5 5 5 5 5 5 5 5 5 5
4
Pb b . . . . . . . . . .
(i) s 7 7 7 7 7 7 7 7 7 7
 ACCEPTED MANUSCRIPT

o 3 2 3 3 3 3 3 2 1 1

l 1 8 2 5 8 0 6 1 5 7
u
ti

o
n
)
208
P ( 4 4 4 4 4 4 4 4 4 3
b/20 P 0 0 0 0 0 0 0 0 0 9
4
Pb b . . . . . . . . . .
(i) s 1 1 1 1 1 1 1 0 0 9
o 2 0 2 5 2 2 3 3 0 6

PT
l 6 4 4 3 6 0 9 7 4 5
u
ti
o

RI
n
)

U ( 0 0 0 0 0 0
(pp S . . . . . .

SC
m) r 0 0 0 0 0 0

- 3 3 4 2 2 1
N
d
NU
-
H
f

s
o
MA

u
ti
o

n
)
D

Th ( 0 0 0 0 0 0

(pp S . . . . . .
E

m) r 1 1 2 1 1 0
- 7 3 4 3 0 9
PT

N
d
-

H
CE

f
s

o
l
u
AC

ti
o
n

Pb ( 0 0 0 0 0 0

(pp S . . . . . .
m) r 0 0 1 0 0 0
- 9 6 0 8 6 3

N
d
-

H
f
s

o
l
u
 ACCEPTED MANUSCRIPT

ti

o
n
)
206
P ( 2 2 2 2 2 2
b/20 S 0 0 0 0 0 0
4
Pb r . . . . . .
(i) - 8 8 8 8 7 6
N 2 1 5 1 3 7
d 4 6 8 8 8 8
-
H

PT
f
s
o
l

RI
u
ti
o

SC
)
207
P ( 1 1 1 1 1 1
b/20 S 5 5 5 5 5 5
4
Pb r . . . . . .
NU
(i) - 7 7 7 7 7 7
N 3 2 3 3 2 1
d 1 8 2 1 1 5

-
H
MA

s
o
l

u
ti
D

o
n
E

)
208
P ( 4 4 4 4 4 3
PT

b/20 S 0 0 0 0 0 9
4
Pb r . . . . . .
(i) - 1 1 1 1 0 9

N 2 0 3 2 3 9
CE

d 3 7 3 1 8 3
-

H
f
s
AC

o
l
u

ti
o
n

2SE: two-standard errors; NA: not analyzed, *Strongly leached sample, ** Strongly leached sample without using MIHL for clinopyroxene separation

U A

n n
i a
t t
o
 ACCEPTED MANUSCRIPT

S R R R R R R R R R R R R R
a A A A A A A A A A A A A A
m V V V V V V V V V V V V V
p - - - - - - - - - - - - -
l 0 1 1 1 1 2 2 2 2 2 3 3 3
e 6 4 7 8 8 1 3 7 7 7 0 8 9
c c c c c c c c c c c c c
s s s
w * w
* *

PT
* *

A 6 6 6 6 6 6 6 6 6 6 6 6
g

RI
f
o

r
i

SC
n

i
t
i
NU
a
l
r

a
t
MA

o
[
M

a
]
D

[
M
E

e
a
PT

s
u
r
e
CE

d
i
s
o
t
AC

o
p
e
r
a
t
i
o
s
]
8
0 0 0 0 0 0 0 0 0 0 0
7
. . . . . . . . . . .
S 7 7 7 7 7 7 7 7 7 7 7
r 0 0 0 0 0 0 0 0 0 0 0

/ 2 2 2 2 2 2 2 3 3 2 2
8
9 9 9 9 9 9 9 0 0 8 9
6
9 7 5 2 8 3 9 0 0 9 0
 ACCEPTED MANUSCRIPT

S 1 6 5 3 8 1 5 1 9 2 8

2 0 0 0 0 0 0 0 0 0 0 0
S . . . . . . . . . . .
E 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 1 0 0 0
5 5 5 6 6 6 6 1 7 6 6
1
0 0 0 0 0 0 0 0 0 0 0
4
. . . . . . . . . . .

PT
3
5 5 5 5 5 5 5 5 5 5 5
N 1 1 1 1 1 1 1 1 1 1 1
d 2 2 2 2 2 2 2 2 2 2 3

/ 9 9 9 9 9 9 9 9 9 9 0

RI
1
9 8 9 9 8 9 8 9 7 9 0
4
5 6 2 7 0 7 2 0 1 0 2
4

SC
d

2 0 0 0 0 0 0 0 0 0 0 0
S . . . . . . . . . . .
E 0 0 0 0 0 0 0 0 0 0 0
NU
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 1 0 1 0
7 8 8 6 7 9 7 1 8 2 7
MA

ε 7 6 7 7 6 7 6 7 6 7 7

N . . . . . . . . . . .

d 1 9 0 1 8 1 8 0 6 0 2

3 4 7 5 2 6 8 2 5 0 5
1
0 0 0 0 0 0 0 0 N 0 N
D

7
. . . . . . . . A . A
6
2 2 2 2 2 2 2 2 2
H 8 8 8 8 8 8 8 8 8
E

f 3 3 3 3 3 3 3 3 3

/ 0 0 0 0 0 0 0 0 0
PT

1
7 6 7 8 6 7 7 5 5
7
1 1 2 2 2 9 4 9 7
7

H
CE

2 0 0 0 0 0 0 0 0 N 0 N
S . . . . . . . . A . A

E 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
AC

0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
5 5 6 5 4 4 6 7 4

ε 1 9 1 1 9 1 1 9 N 9 N

H 0 . 0 0 . 0 0 . A . A

f . 7 . . 8 . . 6 6

1 5 1 5 1 3 2 9 3
0 3 0 9 2
2
1 1 1 1 1 1 1 1 1 1 1 1 1
0
9 9 9 9 9 9 9 9 9 9 9 9 9
6
. . . . . . . . . . . . .

P 4 5 2 1 5 5 4 5 5 5 5 6 4
b 9 5 9 3 2 4 9 0 4 5 5 6 3
/ 6 5 2 4 7 8 1 9 4 1 7 4 9
2
4 2 6 1 2 9 6 2 3 1 9 1 4
0
 ACCEPTED MANUSCRIPT

P
b

2 0 0 0 0 0 0 0 0 0 0 0 0 0
S . . . . . . . . . . . . .
E 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0
1 1 1 1 1 0 1 1 6 0 1 0 1
0 1 2 0 0 9 2 0 5 8 0 9 2
2
1 1 1 1 1 1 1 1 1 1 1 1 1
0
5 5 5 5 5 5 5 5 5 5 5 5 5
7
. . . . . . . . . . . . .

PT
P 5 5 5 6 6 6 5 5 5 5 6 6 5
b 9 9 8 0 2 0 9 9 9 9 0 0 8
/ 1 9 9 1 0 1 1 2 7 8 4 5 2
2
8 9 7 1 0 9 4 1 8 0 3 2 7

RI
0

P
b

SC
2 0 0 0 0 0 0 0 0 0 0 0 0 0

S . . . . . . . . . . . . .
E 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0
NU
0 0 1 0 0 0 1 0 4 0 0 0 0
9 8 0 9 8 8 0 9 8 7 9 7 9
2
3 3 3 3 3 3 3 3 3 3 3 3 3
0
9 9 9 8 9 9 9 9 9 9 9 9 9
8
. . . . . . . . . . . . .
MA

P 2 2 0 9 2 2 2 2 2 2 2 3 1
b 1 6 4 6 9 3 2 2 6 6 5 0 7
/ 9 3 7 9 1 8 2 9 3 4 7 2 0
2
5 3 9 4 9 1 5 2 9 3 6 8 9
0
D

P
b
E

2 0 0 0 0 0 0 0 0 0 0 0 0 0

S . . . . . . . . . . . . .
PT

E 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 1 0 0 0 0

2 2 2 2 2 2 2 2 2 2 2 2 2
4 5 8 6 4 3 7 5 6 0 5 1 6
CE

[
A
g
e
-
AC

c
o
r
r
e
c
t
e
d
i
s
o
t
o
p
e
r
 ACCEPTED MANUSCRIPT

a
t
i
o
s
]
R 0 0 0 0 0 0 0 0 0 0
b . . . . . . . . . .
( 2 1 1 1 2 1 2 2 2 1
p 9 3 9 2 1 1 1 3 5 0
p 9 1 9 7 2 6 5 3 2 1
m

PT
)

S 4 4 5 4 4 4 4 4 4 4
r 9 2 5 3 4 2 4 4 7 1

( . . . . . . . . . .

RI
p 3 9 5 3 3 6 9 6 3 5
p

m
)

SC
8
0 0 0 0 0 0 0 0 0 0 0
7
. . . . . . . . . . .
S 7 7 7 7 7 7 7 7 7 7 7
r 0 0 0 0 0 0 0 0 0 0 0
NU
/ 2 2 2 2 2 2 2 3 3 2 2
8
9 9 9 9 9 9 9 0 0 8 9
6
9 7 5 2 8 3 9 0 0 9 0
S 0 5 4 2 7 0 3 1 7 1 8
r
MA

(
i
)

S 2 2 2 2 2 2 3 2 2 2
m . . . . . . . . . .
( 7 6 6 2 9 8 6 4 9 6
D

p 8 2 3 6 8 2 9 6 4 6
p
E

m
)
PT

N 8 7 8 6 9 8 1 7 9 7
d . . . . . . 1 . . .

( 4 6 2 5 2 2 . 5 2 4
p 2 2 2 3 9 4 6 3 9
CE

m
)
1
0 0 0 0 0 0 0 0 0 0 0
4
. . . . . . . . . . .
AC

3
5 5 5 5 5 5 5 5 5 5 5

N 1 1 1 1 1 1 1 1 1 1 1
d 2 2 2 2 2 2 2 2 2 2 2
/ 9 9 9 9 9 9 9 9 9 9 9
1
8 7 8 8 7 8 7 8 6 8 9
4
8 8 5 8 2 9 5 2 3 2 3
4

N
d
(

i
)

ε 7 6 7 7 6 7 6 7 6 7 7

N . . . . . . . . . . .

d 1 9 0 1 8 1 8 0 6 0 2
( 2 3 7 4 2 5 8 2 5 0 4
i
 ACCEPTED MANUSCRIPT

L 0 0 0 0 0 0 0 N 0 N
u . . . . . . . A . A
( 1 1 1 1 1 1 1 1
p 0 1 1 1 2 3 6 3
p
m
)

H 1 1 1 1 1 1 2 N 1 N

f . . . . . . . A . A
( 6 4 8 3 6 8 0 8
p 3 3 2 8 9 3 4 2

PT
p
m
)
1
0 0 0 0 0 0 0 0 N 0 N

RI
7
. . . . . . . . A . A
6
2 2 2 2 2 2 2 2 2

H 8 8 8 8 8 8 8 8 8
f 3 3 3 3 3 3

SC
3 3 3
/ 0 0 0 0 0 0 0 0 0
1
7 6 7 8 6 7 7 5 5
7
0 0 1 1 1 7 3 8 6
7
NU
H
f
(
i
)
MA

ε 1 9 1 1 9 1 1 9 N 9 N

H 0 . 0 0 . 0 0 . A . A

f . 8 . . 9 . . 7 7

( 2 5 2 5 0 4 3 8 2
i 0 3 9 8 1
)
D

U ( 0 0 0 0 0 0 0 0 0 0
( P . . . . . . . . . .
E

p b 0 0 0 0 0 0 0 0 0 0

p s 3 1 1 1 1 1 1 1 1 2
PT

m o
) l

u
t
CE

o
n
)

T ( 0 0 0 0 0 0 0 0 0 0
AC

h P . . . . . . . . . .

( b 0 0 0 0 0 0 0 0 0 0
p s 6 4 6 5 5 4 6 5 6 5
p o

m l
) u
t

i
o
n

P ( 0 0 0 0 0 0 0 0 0 0
b P . . . . . . . . . .

( b 0 0 0 0 0 0 0 0 0 0
p s 7 2 3 2 4 2 4 4 5 2
p o
m l
 ACCEPTED MANUSCRIPT

) u

t
i
o

n
)
2
( 1 1 1 1 1 1 1 1 1 1 1 1 1
0
P 9 9 9 9 9 9 9 9 9 9 9 9 9
6
b . . . . . . . . . . . . .
P s 4 5 2 1 5 5 4 4 5 5 5 6 3
b o 7 3 7 1 0 3 6 8 2 3 4 4 9
/ l 0 5 0 2 5 1 7 9 4 1 1 9 3

PT
2
u
0
t
4
i
P o

RI
b n
( )
i

SC
2
( 1 1 1 1 1 1 1 1 1 1 1 1 1
0
P 5 5 5 5 5 5 5 5 5 5 5 5 5
7
b . . . . . . . . . . . . .
P s 5 5 5 6 6 6 5 5 5 5 6 6 5
NU
b o 9 9 8 0 1 0 9 9 9 9 0 0 8
/ l 1 9 9 0 9 1 0 1 7 7 4 5 1
2
u
0
t
4
i
MA

P o

b n
( )
i

( 3 3 3 3 3 3 3 3 3 3 3 3 3
D

0
P 9 9 9 8 9 9 9 9 9 9 9 9 9
8
b . . . . . . . . . . . . .
E

P s 2 2 0 9 2 2 1 1 2 2 2 2 1
b o 0 2 1 3 5 0 8 9 3 3 3 7 2
PT

/ l 3 9 3 3 5 9 3 5 0 0 4 7 6
2
u
0
t
4
i
CE

P o
b n

( )
i
)
AC

U ( 0 0 0 0 0 0 0 0 0 0
( S . . . . . . . . . .

p r 0 0 0 0 0 0 0 0 0 0
p - 2 1 1 1 1 1 1 1 1 1
m N

) d
-
H

f
s
o

l
u
t

i
o
n
 ACCEPTED MANUSCRIPT

T ( 0 0 0 0 0 0 0 0 0 0
h S . . . . . . . . . .
( r 0 0 0 0 0 0 0 0 0 0
p - 6 4 5 4 6 4 6 5 7 4
p N
m d
) -
H
f
s
o

PT
l
u
t
i

RI
o
n
)

P ( 0 0 0 0 0 0 0 0 0 0

SC
b S . . . . . . . . . .

( r 0 0 0 0 0 0 0 0 0 0
p - 4 3 3 2 4 2 4 4 5 1
p N
NU
m d
) -
H

f
s
MA

l
u
t

i
o
D

n
)
E

2
( 1 1 1 1 1 1 1 1 1 1
0
S 9 9 9 9 9 9 9 9 9 9
PT

6
r . . . . . . . . . .
P - 4 5 2 1 5 4 4 5 6 4
b N 7 4 7 0 3 6 9 4 4 1

/ d 4 0 1 9 0 5 0 1 8 3
CE

2
-
0
H
4
f
P s
b o
AC

( l
i u
) t

i
o
n

)
2
( 1 1 1 1 1 1 1 1 1 1
0
S 5 5 5 5 5 5 5 5 5 5
7
r . . . . . . . . . .
P - 5 5 5 6 6 5 5 6 6 5

b N 9 9 8 0 0 9 9 0 0 8
/ d 1 9 9 0 1 0 1 4 4 2
2
-
0
H
4
f
P s
 ACCEPTED MANUSCRIPT

b o

( l
i u
) t

i
o
n

)
2
( 3 3 3 3 3 3 3 3 3 3
0
S 9 9 9 8 9 9 9 9 9 9
8
r . . . . . . . . . .
P - 1 2 0 9 2 1 1 2 2 1

PT
b N 8 3 1 2 0 7 9 3 7 0
/ d 9 5 0 8 7 8 3 1 4 8
2
-
0
H

RI
4
f
P s
b o

( l

SC
i u
) t

i
o
NU
n
)

2SE: two-standard errors; NA: not analyzed, *Strongly leached sample, ** Strongly leached sample without using MIHL for clinopyroxene separation
MA
E D
PT
CE
AC
 ACCEPTED MANUSCRIPT

Table 3 Major element composition of bulk rock and clinopyroxene core samples

U R A

n a n
i i a
t r t

u o
a n
u

S R R R R R R R R R R R R R R
a A A A A A A A A A A A A A A
m V V V V V V V V V V V V V V

PT
p - - - - - - - - - - - - - -
l 0 0 1 2 4 0 1 1 1 2 2 3 3 3
e 5 8 1 0 1 6 4 7 8 3 7 0 8 9

RI
[
B
u

SC
l
k
r
o
c
NU
k
]
S 4 4 4 4 4 4 4 4 4 4 4 4 4 4
i 4 4 2 4 6 4 7 4 5 8 8 5 8 7
MA

O . . . . . . . . . . . . . .

2 8 6 5 6 3 7 5 5 6 1 0 4 2 6
3 2 9 3 0 9 1 0 0 5 7 7 1 1

T 2 2 3 2 2 1 1 2 1 2 2 1 2 2
i . . . . . . . . . . . . . .
O 5 3 5 4 0 8 7 2 7 5 2 7 6 0
D

2 4 2 2 1 0 4 7 9 0 0 2 6 2 2

A 1 1 1 1 1 9 1 1 9 1 1 9 1 1
E

l 4 3 5 3 1 . 0 0 . 6 2 . 6 6

2 . . . . . 4 . . 6 . . 5 . .
O 7 1 6 8 7 9 6 5 0 1 3 1 0 0
PT

3 3 4 1 0 0 5 9 1 1 6 3

F 1 1 1 1 1 1 1 1 1 1 1 1 1 1
e 1 1 2 1 0 2 1 2 3 1 2 2 0 0
O . . . . . . . . . . . . . .
CE

* 9 5 7 7 5 8 1 8 8 0 0 2 8 3
4 1 4 4 1 4 6 0 2 4 8 2 7 2

M 0 0 0 0 0 0 0 0 0 0 0 0 0 0
n . . . . . . . . . . . . . .
AC

O 2 1 2 2 1 1 1 1 1 1 1 1 1 1

0 9 1 0 8 8 5 7 9 5 7 8 5 4

M 8 1 6 9 1 2 1 1 1 6 1 1 6 9

g . 0 . . 1 0 6 7 8 . 2 8 . .
O 1 . 8 8 . . . . . 9 . . 5 3
3 6 9 2 1 1 1 0 1 8 5 7 7 7

9 6 5 6 0 3 2 6

C 1 1 1 1 1 8 9 9 8 1 8 9 1 1
a 3 3 2 3 5 . . . . 0 . . 1 0

O . . . . . 3 9 8 7 . 4 7 . .
2 7 3 5 8 7 2 1 9 6 3 7 3 9
9 1 4 4 6 7 6 3

N 2 2 3 2 1 1 1 1 1 3 2 1 2 2
a . . . . . . . . . . . . . .

2 8 5 9 5 5 5 9 9 3 1 9 5 9 6
O 9 5 6 3 2 4 8 7 3 7 0 7 9 3

K 1 0 1 0 0 0 0 0 0 0 0 0 0 0
 ACCEPTED MANUSCRIPT

2 . . . . . . . . . . . . . .

O 0 9 4 9 4 5 4 6 6 8 8 5 7 6
2 0 8 3 8 3 8 0 1 7 8 2 9 6

P 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 . . . . . . . . . . . . . .
O 4 3 6 3 2 2 2 2 2 3 4 2 3 2

5 3 7 7 9 8 6 2 8 4 7 1 5 6 8

t 1 1 1 1 1 1 1 1 1 1 1 1 1 1
o 0 0 0 0 0 0 0 0 0 0 0 0 0 0

t 0 0 0 0 0 0 0 0 0 0 0 0 0 0
a . . . . . . . . . . . . . .
l 0 0 0 0 0 0 0 0 0 0 0 0 0 0

PT
0 0 0 0 0 0 0 0 0 0 0 0 0 0

M 0 0 0 0 0 0 0 0 0 0 0 0 0 0
g . . . . . . . . . . . . . .
# 5 6 4 6 6 7 7 7 7 5 6 7 5 6

RI
5 2 9 0 5 4 2 0 0 3 5 3 2 2

SC
C
l
i
n
o
NU
p
y
r
o
x
MA

e
n
e
c
o
r
D

e
]
E

S 4 4 4 4 4 4 4 4 4 4 4 4 4 4

i 5 5 3 5 7 6 7 7 7 7 7 8 8 7
PT

O . . . . . . . . . . . . . .

2 1 2 6 1 9 6 3 0 2 6 0 4 8 4

5 7 2 2 4 4 7 1 1 7 5 2 8 9

T 1 1 2 1 1 1 0 0 0 1 1 0 1 1
CE

i . . . . . . . . . . . . . .
O 5 4 3 5 0 2 9 9 8 0 3 9 1 2

2 0 9 6 8 8 1 8 7 5 6 6 9 0 2

A 5 5 6 5 4 3 4 5 4 3 4 3 3 4
l . . . . . . . . . . . . . .
AC

2 5 3 5 9 0 8 4 0 5 8 5 1 4 4

O 8 6 6 3 7 6 6 0 1 8 2 3 8 9

F 6 6 7 6 5 6 6 6 6 7 7 7 7 7
e . . . . . . . . . . . . . .
O 7 1 7 4 9 4 2 9 8 2 4 3 3 9

* 1 8 3 7 1 4 9 7 2 9 0 1 2 1

M 0 0 0 0 0 0 0 0 0 0 0 0 0 0
n . . . . . . . . . . . . . .

O 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 0 2 0 0 2 3 3 4 4 4 5 4 5
7 2 3 7 6 1 1 5 6 4 7 1 3 0

M 1 1 1 1 1 1 1 1 1 1 1 1 1 1
g 5 5 3 5 5 6 7 6 7 6 6 6 6 6

O . . . . . . . . . . . . . .
4 6 9 2 8 9 1 5 3 6 1 6 2 1
5 8 8 0 8 8 3 9 5 7 3 7 1 2
 ACCEPTED MANUSCRIPT

C 2 2 2 2 2 2 2 2 2 2 2 2 2 2

a 5 5 5 5 4 4 3 2 2 2 2 3 2 2
O . . . . . . . . . . . . . .
1 6 0 2 7 4 2 9 7 8 9 0 4 2

2 0 8 5 2 2 3 1 2 7 3 1 8 5

N 0 0 0 0 0 0 0 0 0 0 0 0 0 0
a . . . . . . . . . . . . . .

2 3 3 5 3 2 3 4 4 3 4 4 3 3 3
O 6 1 3 3 8 2 0 0 9 1 6 2 7 7

K 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 . . . . . . . . . . . . . .
O 0 0 0 0 0 0 0 0 0 0 0 0 0 0

PT
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 1 0
3 3 9 6 0 0 4 1 7 4 2 0 5 0

P 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RI
2 . . . . . . . . . . . . . .
O 0 0 0 0 0 0 0 0 0 0 0 0 0 0

5 1 0 1 0 0 0 0 0 0 0 1 0 0 0
0 7 0 9 6 9 8 6

SC
7 8 0 8 8 8

t 1 1 1 1 1 1 1 1 1 1 1 1 1 1
o 0 0 0 0 0 0 0 0 0 0 0 0 0 0
t 0 0 0 0 0 0 0 0 0 0 0 0 0 0

a . . . . . . . . . . . . . .
NU
l 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0

M 0 0 0 0 0 0 0 0 0 0 0 0 0 0
g . . . . . . . . . . . . . .
# 8 8 7 8 8 8 8 8 8 8 8 8 8 7
MA

0 2 6 1 3 2 3 1 2 0 0 0 0 8

M 0 0 0 0 0 0 0 0 0 0 0 0 0 0

E . . . . . . . . . . . . . .
L 7 8 7 8 8 8 8 7 8 7 7 7 7 7
T 5 3 2 2 2 1 0 9 0 8 7 9 8 8
D

S
M
g
E

#
PT

Note: FeO* = Fe normalized to total FeO basis; Mg# = molar Mg/(Mg+Fe); MELTS Mg# = clinopyroxene Mg# calculated by MELTS at 0.1 GPa, NNO buffer, 0.2

wt% H2O.
CE
AC
Figure 1
Figure 2
Figure 3
Figure 4