J. Petrology 2009 Tatsumi 1575 603

8/10/2019 J. Petrology 2009 Tatsumi 1575 603

1/29

Tholeiitic vs Calc-alkalic Differentiation andEvolution of Arc Crust: Constraints fromMelting Experiments on a Basalt from theIzu^Bonin^Mariana Arc

Y. TATSUMI * AND T. SUZUKIINSTITUTE FOR RESEARCH ON EARTH EVOLUTION (IFREE), JAPAN AGENCY FOR MARINE^EARTH SCIENCE AND

TECHNOLOGY (JAMSTEC), YOKOSUKA 237-0061, JAPAN

RECEIVED JANUARY 14, 2009; ACCEPTED JUNE 15, 2009ADVANCE ACCESS PUBLICATION JULY 13, 2009

The liquid line of descent (LLD) for a representative basalt fromthe Izu^Bonin^Mariana (IBM) arc was investigated at 0 3 GPain the presence of 049^2 83 wt % H 2 O to constrain the differentiation of arc magmas and the evolution of the arcs crust. This is of interest as differentiated continental crust may form as the middlecrust of the intra-oceanic IBM system.The tholeiitic differentiationtrend, which is often documented in the volcanic sequences of intra- oceanic arcs, is best reproduced by the LLD of an H 2 O-poor ( 0 5 wt %) basalt, whereas the calc-alkalic trend in the IBM rocks, which are likely to form middle crust of intermediate composition, mimics the LLD of a hydrous basalt with 2 5^3 0 wt %H 2 O. Magmatic temperatures estimated for IBM calc-alkalicrocks, however, tend to be higher than those for the hydrous LLD.It may thus be suggested that an alternative mechanism, mixing of basaltic and felsic magmas, could play a major role in calc-alkalicmiddle crust formation in the IBM. Seismic velocities of the inferred middle crustal rocks, obtained based on both the experimental resultsand theoretical calculations, agree well with the observed seismiccrustal structure.

KEY WORDS: andesite; basalt; calc-alkalic: crust; tholeiitic

I N T R O D U C T I O NBasaltic magmas generated by mantle melting transportboth mass and heat to the crust. When basalt magmasreach crustal depths they may produce more evolved

magmas via closed-system fractionation (Sisson & Grove,1993; Kawamoto, 1996; Grove et al ., 2003; Pichavant &Macdonald, 2007), crustal melting (Smith & Leeman,1987; Petford & Atherton, 1996; Fornelli et al ., 2002; Sissonet al ., 2005), or any combination of these two processes(Sakuyama, 1981; Kay & Kay, 1985; Hildreth &Moorbath, 1988; Tatsumi, 2000; Dufek & Bergantz, 2005).Identification of the processes occurring within the arccrust is of critical importance given that evolved magmashave been produced ubiquitously, both in time and space,to create the continental crust that is a differentiated end-member component within the solid Earth (e.g. Rudnick,1995; Albare 'de, 1998; Tatsumi, 2005; Hawkesworth &Kemp, 2006).

It has been well established that a greater abundance of intermediate to felsic magmas is erupted at arcs along con-tinental margins than in oceanic arcs (Gill, 1981; Ewart,1982; Green, 1982; Wilson, 1989; Tatsumi & Eggins, 1995).This general observation leads us to imagine that interme-diate to felsic magmas are not produced voluminously inoceanic arc settings. However, extensive seismic experi-ments in the Izu^Bonin^Mariana (IBM) intra-oceanicarc (Suyehiro et al ., 1996; Takahashi et al ., 1998, 2007, 2008;

Kodaira et al ., 2007 a, 2007b; Sato et al ., 2009) have clearlydocumented the occurrence of middle crust with a P-wavevelocity ( V p ) of 6^7 km/s. This is identical to the V p of both the bulk or average continental crust and plutonicor metamorphic rocks with intermediate compositions

*Correspondingauthor: E-mail: [email protected]

The Author 2009. Published by Oxford University Press. Allrights reserved. For Permissions, please e-mail: [email protected]

JOURNALOFPETROLOGY VOLUME 50 NUMBER 8 PAGES1575^1603 2009 doi:10.1093/petrology/egp044


2/29

(e.g. Christensen & Mooney,1995), providing a compellingreason for concluding that intermediate magmas areactively produced, although not necessarily erupted, atoceanic arcs and for speculating that evolved or differen-tiated continental crust is also being created in such intra-oceanic arc settings. Although intensive approaches have

been adopted in an attempt to determine how intermediatemiddle crust forms at the IBM (Kawate & Arima, 1998;Nakajima & Arima, 1998; Taira et al ., 1998; Takahashiet al ., 2007; Tatsumi et al ., 2008a), lack of quantitativeknowledge of the melting or crystallization regime formagmas or rocks occurring in this arc has impeded a com-prehensive understanding of arc crust evolution.

The primary aim of this study is to investigate the phaserelations and liquid^solid compositions during partialmelting of a representative basalt composition from theIBM under pressure conditions relevant to the IBM deepcrust. Based on these data the differentiation process from juvenile to evolved arc crust is then discussed.

I Z U ^ B O N I N ^ M A R I A N A ( I B M ) A R CTectonic evolutionThe IBM arc^trench system extends for 2800 km, south-ward from Honshu through the Izu Peninsula to Guam(Fig. 1). The IBM subduction zone began as part of ahemisphere-scale foundering of old, dense lithosphere inthe Western Pacific at 50 Ma (Bloomer et al ., 1995; Stern,2004). Its initiation was perhaps aided by mantle downwel-ling at the Indian^Pacific asthenospheric domain bound-ary (Okino et al ., 1999) or by plate convergence (Hallet al ., 2003). Boninite, a characteristic andesite defined bya high MgO content, the absence of plagioclase as both

phenocrysts and groundmass phases, and the presence of clinoenstatite phenocrysts, was the major magmatic prod-uct at this initial stage.This was followed by the generationof low-K tholeiites (Ishizuka et al ., 2006).

IBM arc volcanism continued until about 30 Ma,accompanied until at least 33 Ma by spreading along theWNW^ESE-trending Central Basin Fault in the westernPhilippine Sea (Deschamps et al ., 2002; Taylor &Goodliffe, 2004). The IBM then began to form its firstback-arc basins. Spreading began in the south to form theParece Vela Basin (Fig. 1) and propagated north and south(Okino et al ., 1998). Spreading in the northernmost IBMbegan about 25 Ma and propagated south to form the

Shikoku Basin (Kobayashi et al ., 1995; Okino et al ., 1999).The Shikoku Basin and Parece Vela Basin spreading sys-tems coalesced at 20 Ma, stranding the Kyushu^PalauRidge as a remnant arc (Fig. 1). This back-arc basin spread-ing stopped at 15 Ma and then subduction resumedbeneath SW Japan. This caused the short-lived transformmargin along the northern margin of the Philippine SeaPlate (Nankai Trough, Fig. 1) to become a convergentmargin (e.g. Tatsumi, 2006) and the northernmost IBM to

start colliding with Honshu at about 15 Ma (Itoh, 1988;Itoh & Ito, 1989). A new episode of rifting in the southernIBM began at about 7 Ma, with seafloor spreading form-ing the Mariana Trough back-arc basin from about 3^4Ma (Bibee et al ., 1980; Yamazaki & Stern, 1997).

Crust and mantle structureA wide-angle ocean bottom seismometer (OBS) experi-ment crossing the IBM at 32 8150N (Suyehiro et al ., 1996;Takahashi et al ., 1998) defined an 20 km thick arc crustbroadly composed of three layers (Fig. 2). The upper layer,with a P-wave velocity ( V p ) of 15^5 8 km/s, is composedof sediments and volcanic rocks above a middle-crustallayer ( V p 6 0^6 8 km/s). The lower crust has an upperlayer ( V p 6 8^6 9 km/s) overlying a thick basalcrust layer ( V p 7 1^7 2 km/s). This layered crustal struc-ture has been confirmed by a recent seismic survey acrossthe Mariana arc (Takahashi et al ., 2007, 2008), which alsosuggests, as shown in Fig. 2, the existence of a zone of uppermost mantle with V p of 72^7 6 km/s, significantlyslower than the V p of the normal upper mantle ( 4 7 8 km/s).This crustal structure has also been identified along-strikebeneath the northern and central IBM, although eachcrustal layer varies in thickness (Kodaira et al ., 2007 a,2007b; Sato et al ., 2009).

The Moho discontinuity is a sharp seismological bound-ary that generally separates rocks with V p of 6^7km/sfrom those with V p of about 8 km/s. The sub-Mohomantle beneath the IBM, however, shows lower V p thantypical mantle, causing confusion in the definition of thesub-IBM Moho. The sub-arc Moho in the IBM is identifiedbased on a combination of the distribution of seismicreflectors and the velocity jump across the Moho (from

6 8^7 2 to 7 2^7 6 km/s), and extrapolation of the well-defined normal Moho from beneath the back-arc basinsof the IBM (Takahashi et al ., 2007). Beneath the IBM theV p of the sub-Moho mantle shows cross-arc variations;that is, from 8 0 km/s beneath the back-arc basin to7 2^7 6 km/s beneath the volcanic arc.

The observed seismic structure of the sub-IBM crust andmantle was examined by petrological modeling of arcmagma generation and differentiation (Tatsumi et al .,2008 a). The crust and mantle can be interpreted as petro-logical layers, descending from volcanic rocks with a bimo-dal composition at the surface, through tonalite,amphibolite (the plutonic equivalent of the initial basaltic

arc crust), granulite (originally produced as a restiteduring anatexis of the basaltic arc crust or amphibolite)to harzburgite (Fig. 2).

Magma compositionsMagmatism resumed at 15 Ma and has continued to thepresent day along the entire IBM, resulting in accumula-tion of volcaniclastic rocks, lavas, and plutonic rocks witha wide range of compositions. As emphasized by Tamura

JOURNAL OF PETROLOGY VOLUME 50 NUMBER 8 AUGUST 2009

1576


3/29

Fig. 1. Tectonic setting of the Izu^Bonin^Mariana Arc (IBM), created by subduction of the Pacific plate at the Izu^Bonin^Mariana trenchbeneath the Philippine Sea plate. Behind the IBM back-arc basins have formed, including the Shikoku and Parece Vela Basins and MarianaTrough. The IBM has collided with Honshu at the northern tip of the arc to form the Tanzawa plutonic complex (star), composed mainly of intermediate to felsic plutonic rocks. Similar compositions have also been dredged along the IBM and Kyushu^Palau Ridge (ci rcles).

TATSUMI & SUZUKI ARC MAGMA DIFFERENTIATION

1577


4/29

& Tatsumi (2002) and Tatsumi et al . (2008 a), the major vol-canic products are basaltic in composition, although thereare minor amounts of andesites, which is a characteristicfeature of intra-oceanic arcs. Figure 3 shows the composi-tions of the Quaternary volcanic rocks from the IBM. Itshould be stressed that calc-alkalic andesites in the IBMhave major element compositions broadly similar to thebulk continental crust.

Insights into the lithology of the middle crust of the IBMare provided by exposures of deeper crust on land at thenorthern end of the IBM where it is in collision withHonshu (Fig. 1), and by exposures of deeper crust on thescarps of deep submarine faults. The arc^arc collision atthe northern end of the IBM began at about 15 Ma andexposes the Tanzawa plutonic complex (Fig. 1), which lar-gely consists of tonalites, with minor gabbroic suites(Kawate & Arima, 1998). The complex is characterized bya low-K magma series and has Sr and Nd isotopic compo-sitions comparable with the northern IBM (Kawate &Arima, 1998; Taira et al ., 1998; Tamura & Tatsumi, 2002),and thus was viewed as a potential candidate for a

representative of the middle crust (Taira et al ., 1998;Tatsumi et al ., 2008a). However, sensitive high-resolutionion microprobe (SHRIMP) dating of zircons from thecomplex has yielded ages 5 5 Ma, suggesting that at leastsome plutons postdate collision (Tani et al ., in preparation).Thus, rather than representing the middle crust, this sug-gests that the plutons formed during post-collisional mag-matism. The model for the evolution of IBM arc crustproposed by Tatsumi et al . (2008 a) that assumes Tanzawatonalite as a possible IBM middle crust lithology maytherefore require revision.

Plutonic rocks ranging from gabbro via tonalite to gran-ite have been dredged from fault scarps in the arc and theKyushu^Palau Ridge, a remnant paleo-IBM arc separatedby back-arc rifting in the Shikoku^Parece Vela Basin(Fig. 1). The samples collected to date exhibit chemicaltrends very similar to the Quaternary calc-alkalic volcanicrocks rather than tholeiitic rocks (Fig. 3). If these areindeed part of the IBM deep crust this suggests that it iscomposed of calc-alkalic rocks with compositions similarto average continental crust.

Fig. 2. Generalized P-wave velocity ( V p ) structure of the sub-IBM crust and mantle after Suyehiro et al . (1996), Takahashi et al . (1998, 2007,2008), Kodaira et al . (2007 a, 2007b) and Sato et al . (2009). The IBM arc is characterized by a middle crust with V p of 6 0^6 8 km/s that is identicalto the V p of both the bulk or average continental crust, and plutonic or metamorphic rocks with intermediate compositions. It should be alsostressed that the V p of the uppermost mantle layer is significantly lower than that of the normal mantle. Petrological interpretation of this char-acteristic layered structure i s also shown after Tatsumi et al . (2008 a).


1578


5/29

Fig. 3. Compositions of subaerial volcanic and dredged plutonic rock from the IBM. Two chemical trends, calc-alkalic (CA) and tholeiitic(TH), are recognized within the volcanic rocks; the plutonic rocks have compositions similar to the CA volcanic rocks. IBM calc-alkalic, inter-medate rocks have compositions similar to the average continental crust (Weaver & Tarney,1984; Taylor & McLennan, 1985; Shaw et al ., 1986;Christensen & Mooney,1995; Rudnick & Fountain,1995; Wedepohl, 1995). A basalt represent ing the IBM basalts (filled star) is used as the start-ing material of the melting experiments. An inferred rhyolite, which is produced by crystallization or partial melting of the starting basalt, isalso plotted (open star).


1579


6/29

E X P E R I M E N T SStarting materialBased on a compilation of volcanic rock compositions fromthe IBM, Tatsumi et al . (2008 a) proposed a widely distribu-ted basalt composition (DB in Table 1) as a representativeparental basalt magma. They proposed that this originatesfrom a mantle-derived primary basalt magma via 20%fractional crystallization leaving an olivine cumulate inthe mantle. It is the major basaltic component within theIBM crust, occurring as either basaltic lavas in the uppercrust or as mafic plutons in the lower crust. A fresh basaltfrom Sumisu-jima island (Figs 1 and 3, D18-R04 inTable 1) with a composition similar to the representativebasalt proposed by Tatsumi et al . (2008 a) is used as thestarting material for the experiments in this study.

Five glasses with different water contents were synthe-sized for use in the melting experiments as follows. Thepowdered sample was first sealed in a Pt capsule (4 7 mminner diameter, 0 15 mm wall thickness) with a smallamount of distilled water (0 5, 1 0, 2 0, 2 5 and 3 0wt %)

added in separate runs. The sample was then heated to1350^1400 8C at 0 2 GPa for 30 min, and quenched isobari-cally. The recovered glasses contained 0 49, 1 14, 1 66, 2 43and 2 83wt %, H 2O, measured using Fourier transforminfrared (FTIR) spectrometry (Table 1).

Experimental proceduresMelting experiments were performed in a Kobelco500MPa type internally heated pressure vessel in whichpure Ar gas was used as the pressure medium. The glasssample was placed in a Au 25Pd capsule (2 0 mm inner dia-

meter, 0 15 mm wall thickness), and was hung on Mo wirein the hotspot of a Mo furnace within the pressure vesseland held for 72^195 h at 0 3 GPa and a set temperature(Table 2). The experimental pressure corresponds to thatin the middle to lower crust beneath the IBM. Pressureswere measured with a strain-gauge pressure transducer.Temperatures were monitored with two W 5Re^W 26 Rethermocouples spaced vertically 5 mm apart, and theobserved temperature gradient across the sample was less

Table 1: Compositions of the starting basalt and hydrous glasses

D18-R04 G1 G2 G3 G4 G5 DB

Av. s (64) Av. s (70) Av. s (78) Av. s (66) Av. s (79)

SiO 2 49 30 49 18 0 22 48 85 0 31 48 26 0 24 48 07 0 38 47 40 0 34

TiO 2 0 70 0 66 0 03 0 66 0 04 0 65 0 04 0 65 0 04 0 64 0 03

Al2O3 18 30 17 49 0 16 17 37 0 16 17 00 0 19 17 03 0 19 16 72 0 18

FeO 11 00 10 26 0 11 10 03 0 11 10 11 0 11 9 53 0 18 9 66 0 12

MnO 0 20 0 20 0 02 0 20 0 03 0 19 0 02 0 19 0 02 0 20 0 03

MgO 6 40 6 38 0 09 6 32 0 11 6 26 0 10 6 22 0 13 6 10 0 13

CaO 12 40 12 41 0 16 12 31 0 16 12 27 0 14 12 02 0 18 12 05 0 16

Na 2O 1 60 1 57 0 06 1 53 0 06 1 55 0 06 1 53 0 06 1 55 0 06

K2O 0 10 0 15 0 02 0 15 0 02 0 15 0 02 0 15 0 02 0 15 0 02

H2O 0 49 1 14 1 66 2 43 2 83

Total 100 00 98 79 98 56 98 10 97 82 97 30

H 2 O-free and 100% normalized

SiO 2 49 30 50 03 50 14 50 04 50 39 50 17 50 0

TiO 2 0 70 0 67 0 68 0 67 0 68 0 68 0 8

Al2O3 18 30 17 79 17 83 17 63 17 85 17 70 19 1

FeO 11 00 10 44 10 30 10 48 9 99 10 23 10 2

MnO 0 20 0 20 0 21 0 20 0 20 0 21 0 0

MgO 6 40 6 49 6 49 6 49 6 52 6 46 6 0

CaO 12 40 12 62 12 64 12 72 12 60 12 76 12 1

Na 2O 1 60 1 60 1 57 1 61 1 60 1 64 1 6

K2O 0 10 0 15 0 15 0 16 0 16 0 16 0 2

Total Fe expressed as FeO.The starting basalt (D18-R04) and glass (G1G5) data were determined by X-ray fluorescence and electron microprobeanalysis, respectively. H 2O content in the glass was measured by FTIR. Numbers in parentheses indicate number ofanalytical points. Av, average. DB is a representative IBM basalt (Tatsumi et al ., 2008 a ).


1580


7/29

than 10 8C. At the end of the run the hanging wire was cutwith a surging current, thereby letting the capsule fall tothe cold (5 2508C) bottom of the vessel and quenchisobarically.

Oxygen fugacity ( f O 2) during the melting experimentswas not controlled, because a normal double-capsule

method using a nickel^nickel oxide (NNO) buffer cannotcontrol f O 2 for the long duration of these experimentsbecause of hydrogen leakage through the outer capsuleand dilution of Ni with the Pt capsule (Takagi et al ., 2005;Hamada & Fujii, 2008). Oxygen fugacity was thus esti-mated based on a solution model for coexisting magnetite

Table 2: Experimental conditions, phase assemblages, and proportions (wt %)

Run H 2O T (8C) Duration melt plag cpx opx mt ilm qz RSS

(wt %) (h)

HG379 0 49 1000 170

HG384 0 49 1050 195 13 2 48 17 8 11 1 6 7 0 3 1 0 029

HG394 0 49 1050 (1260) 170 14 3 50 4 18 8 10 6 5 9 0 0 0 000

HG385 0 49 1100 144 20 4 47 8 13 8 12 8 5 2 0 0 0 026

HG393 0 49 1100 (1250) 100 22 9 46 5 17 1 8 8 4 8 0 0 0 011

HG402 0 49 1115 (1260) 117 36 3 39 9 12 4 8 6 2 7 0 0 0 017

HG395 0 49 1125 (1260) 76 38 2 38 4 15 5 5 1 2 7 0 0 0 043

HG392 0 49 1150 (1250) 96 60 9 27 5 4 8 4 3 2 6 0 0 0 026

HG379 1 14 1000 170 17 6 45 7 18 1 11 1 7 1 0 4 0 0 020

HG384 1 14 1050 195 24 4 43 2 16 5 8 8 7 0 0 0 002

HG394 1 14 1050 (1260) 170 24 9 42 2 16 8 9 5 6 6 0 0 0 002

HG385 1 14 1100 144 33 6 39 1 14 7 5 5 7 0 0 0 005

HG393 1 14 1100 (1250) 100 33 3 39 9 14 4 7 3 5 1 0 0 0 125

HG402 1 14 1115 (1260) 117 53 5 30 4 8 6 1 2 1 0 0 0 003HG395 1 14 1125 (1260) 76 53 8 30 9 5 4 8 1 8 0 0 0 013

HG392 1 14 1150 (1250) 96 81 5 17 6 0 0 0 9 0 0 0 644

HG389 1 66 950 (1200) 190 21 3 42 7 16 9 10 2 3 3 5 7 0 0 051

HG379 1 66 1000 170 29 2 36 8 17 6 8 7 7 7 0 0 0 015

HG363 1 66 1050 120 36 8 33 9 16 1 6 4 6 8 0 0 0 001

HG394 1 66 1050 (1260) 170 35 7 34 1 16 1 7 5 6 6 0 0 0 003

HG385 1 66 1100 144 49 9 28 5 12 1 3 7 5 8 0 0 0 008

HG402 1 66 1115 (1260) 117 69 2 21 2 7 5 0 2 0 0 0 059

HG395 1 66 1125 (1260) 76 84 4 15 6 0 0 0 0 0 0 468

HG392 1 66 1150 (1250) 96 87 0 12 2 0 0 0 8 0 0 0 477

HG389 2 43 950 (1200) 190 24 9 39 9 16 4 9 3 2 3 7 2 0 0 037

HG379 2 43 1000 170 41 7 28 3 17 2 5 4 7 5 0 0 0 010

HG391 2 43 1000 (1200) 120 28 38 7 15 6 8 5 5 3 4 0 0 055

HG388 2 43 1050 (1200) 120 52 3 21 6 15 9 4 3 5 9 0 0 0 003

HG381 2 43 1100 72 71 6 15 4 8 9 0 4 2 0 0 0 188

HG392 2 43 1150 (1250) 96 100 0 0 0 0 0 0

HG389 2 83 950 (1200) 190 27 8 37 9 16 1 8 6 2 8 6 9 0 0 061

HG379 2 83 1000 170 43 5 27 4 15 9 5 2 8 0 0 0 002

HG391 2 83 1000 (1200) 120 33 1 34 7 16 1 6 8 6 3 4 0 0 044

HG388 2 83 1050 (1200) 120 63 1 16 4 14 5 0 6 0 0 0 105

HG381 2 83 1100 72 78 6 11 8 5 9 0 3 7 0 0 0 250

HG392 2 83 1150 (1250) 96 100 0 0 0 0 0 0

Numbers in parentheses indicate temperature for 1 h pre-heating before the run at the planned temperature. RSS, residualsum of squares.


1581


8/29

and ilmenite (Spencer & Lindsley, 1981). Although theobserved compositions of magnetite and ilmenite arebeyond the upper limit of their oxygen barometer, we ten-tatively show these estimates in Fig. 4a. The experimental

redox conditions were within a well-defined range of NNO 2 5 to NNO 3 0, which is slightly lower than

the f O 2 during previous experimental runs by Tatsumiet al . (2006), who used a similar assembly (Fig. 4a).

Fig. 4. (a) f O 2 during the melting experiments of this study (stars) and the study of Tatsumi et al . (2006) (dashed line) inferred from

magnetite^ilmenite compositions (Spencer & Lindsley,1981). The f O 2 range of arc magmas (Carmichael & Ghiorso, 1990; Ballhaus et al ., 1991;Carmichael, 1991; Ballhaus, 1993; Parkinson & Arculus, 1999) is also shown. (b^e) Effect of f O 2 on the liquid line of descent based on the exper-imental data of Sisson et al . (2005).


1582


9/29

Mineral and glass compositionsMineral and glass compositions were analyzed using a JEOL JXA-8800 electron-probe micro-analyzer. The exci-tation potential, beam current and time spent on each ele-mental peak were 15 kV, 12 nA and 20s, respectively.A defocused electron beam of 10 mm in diameter was used

for glass, whereas a focused beam was employed for themeasurement of crystalline phases. ZAF correction proce-dures were employed. Modal compositions were calculatedby mass balance based on the compositions of the startingmaterial, minerals and glasses.

R E S U LT SThe modal compositions and the average compositions of minerals and glasses are listed in Tables 2 and 3.

An important criterion indicating that equilibriumhas been reached is that regular and consistent partition-ing of major elements between crystalline phases and meltis achieved. The calculated orthopyroxene melt andclinopyroxene^melt Fe^Mg exchange distribution coeffi-cients ( K D

Fe^Mg ) are 0 20 0 03 (1 s ) and 0 29 0 06 (1 s ),respectively (Fig. 5a). The plagioclase^melt Ca^Naexchange distribution coefficient changes with melt com-position (Fig. 5b). These exchange distribution coefficientsare close to those previously reported (e.g. Gaetani et al .,1993; Sisson & Grove, 1993; Grove et al ., 2003; Sisson et al .,2005; Tatsumi et al ., 2006; Hamada & Fujii, 2008). This,together with constant melt and mineral compositionswithin each run product (Table 3), suggests that the experi-ments achieved equilibrium.

Electron microprobe analyses confirmed Fe loss, rangingfrom 0 5 to 0 7 wt % relative, during the synthesis of

hydrous glass in Pt capsules (Table 1). In the Au^Pd cap-sules used in the long duration ( 4 72 h) melting experi-ments Fe loss was not observed. As a result, the minimumFe content of the quenched glass can be estimated, andfrom mass-balance calculations is 9 8wt %.

For each synthesized sample the melt fraction systemati-cally increases as temperature increases (Fig. 6). The con-centrations of particular elements at a constant meltfraction (e.g. Si and Fe) vary significantly with H 2O con-tent (Fig. 6), as a result of systematic changes in phaseabundance with H 2O content (Fig. 7). Crystallization of plagioclase and orthopyroxene tends to be suppressed,whereas the amount of Fe^Ti oxides increases with increas-

ing H 2O. Lower concentrations of Al 2O 3 and Na 2O, andhigher concentrations of TiO 2 and FeO in H 2O-poormelts may thus be caused by higher and lower abundanceof plagioclase and Fe^Ti oxides, respectively (Figs. 7 and8). Higher abundance of orthopyroxene as a crystallizingphase at higher degrees of partial melting (Fig. 7) couldaccount for the observation that SiO 2 tends to be constantin H 2O-poor systems (Fig. 8). Na 2O data tend to be scat-tered especially at small melt fractions (Fig. 8), suggesting

that the defocused electron beam used here may still causeNa loss during analysis. However, the size of glass poolspresent in those experiments is not large enough to use amore defocused beam.

Major element compositions of experimental melts areplotted as a function of SiO 2 content in Fig. 9. Liquid

lines of descent (LLDs), especially TiO 2, Al2O 3, and FeOvs SiO 2 vary significantly with H 2O content.

D I S C U S S I O NEffect of f O 2 on liquid line of descentThese melting experiments allow the LLD to be definedfor equilibrium not fractional crystallization for a specificbasalt composition in the presence of different amounts of H 2O. To use these results to constrain magmatic processes,however, we have to evaluate the effects of f O 2 on meltcomposition. The reason for this is the difference in f O 2between experiments and natural magmas. The f O 2 of arc magmas are generally estimated to range from NNOto NNO 2 (i.e. NNO 2) (Carmichael & Ghiorso,1990; Ballhaus et al ., 1991; Carmichael, 1991; Ballhaus, 1993;Parkinson & Arculus,1999). For the IBM magmas, similarestimates have been proposed ( NNO 0 2^0 5, Yasudaet al ., 2001; NNO 0^3, H. Shukuno, personal communi-cation). Thus, we suggest that f O 2 during the presentexperiments ( NNO 2^3) is slightly ( 5 2 log unit)higher than the f O 2 in the IBM magmas (Fig. 4a).

Sisson et al . (2005) demonstrated that the amount andcomposition of evolved liquids and coexisting mineralassemblages vary with f O 2 and temperature, with themelt being more evolved at higher f O 2, and the coexistingmineral assemblages more plagioclase and Fe^Ti oxide-

rich, and amphibole-poor. A decrease in f O 2 of 2 logunits can cause a maximum of 0 1 and 0 3 w t %decrease in TiO 2 and MgO, and 0 5 and 0 3w t %increase in Al 2O 3 and FeO at constant SiO 2, respectively(Fig. 4b^e). These effects are shown on the LLD in Fig. 9.

Our data also demonstrate systematic and significantchanges in LLD with varying H 2O content; for example,TiO 2 and FeO contents at a constant SiO 2 decrease,whereas Al 2O 3 increases significantly with increasingH 2O (Fig. 9). Therefore it is suggested that increasing theH 2O content from 0 49 to 2 83wt % has a much greatereffect than that inferred for f O 2.

H 2 O content vs tholeiitic and calc-alkalicdifferentiation trendsThe final goal of this experimental work is to understandthe process involved in the formation of the characteristicmiddle crustal layer of the IBM, which exhibits a seismicvelocity close to that of the average continental crust andthus is likely to have an intermediate composition, andmore generally to understand the genetic relationshipbetween the calc-alkalic and tholeiitic magma series.


1583


10/29

Table 3: Compositions of glass and minerals (wt %)

Run T (8C) SiO 2 TiO2 Al2O3 FeO MnO MgO CaO Na 2O K2O Cr 2O3 Total

0 49% H 2 O

HG379 1000 plag (2) 49 46 0 05 31 22 1 14 0 01 0 15 15 15 2 41 0 07 0 02 99 68

s 0 18 0 00 0 00 0 08 0 01 0 04 0 12 0 05 0 01 0 02

opx (5) 53 10 0 26 2 70 16 37 0 39 23 23 2 4 0 09 0 03 0 00 98 57

s 0 27 0 08 1 06 0 26 0 05 0 72 0 5 0 05 0 01 0 00

cpx 50 69 0 58 2 82 10 20 0 47 15 49 18 7 0 28 0 00 0 02 99 23

qz (1) 97 99 0 11 0 21 0 41 0 01 0 26 0 16 0 02 0 01 0 05 99 23

mt (6) 0 36 6 47 3 30 76 10 0 24 2 88 0 35 0 01 0 03 0 07 89 81

s 0 09 0 17 0 08 0 17 0 01 0 04 0 05 0 02 0 01 0 03

ilm (2) 0 44 25 81 0 93 60 53 0 11 1 97 0 53 0 02 0 02 0 05 90 41

s 0 03 0 15 0 00 0 01 0 00 0 02 0 00 0 02 0 00 0 01

HG384 1050 Melt (4) 67 08 1 25 13 47 6 96 0 19 2 16 6 44 1 51 0 94 0 00 100 00

s 0 47 0 06 0 20 0 18 0 01 0 05 0 18 0 06 0 02 0 00

plag (2) 46 39 0 04 30 69 1 27 0 02 0 15 15 30 2 34 0 07 0 00 96 27

s 0 19 0 01 0 27 0 09 0 02 0 02 0 32 0 05 0 00 0 00opx (4) 50 89 0 32 2 23 14 45 0 54 26 52 1 88 0 03 0 03 0 00 96 89

s 0 53 0 02 0 08 0 37 0 02 0 31 0 19 0 02 0 01 0 00

cpx (5) 47 27 0 71 3 43 10 12 0 43 15 05 18 05 0 26 0 03 0 02 95 37

s 0 41 0 15 0 37 0 50 0 06 0 39 0 84 0 03 0 01 0 01

mt (3) 0 28 5 43 3 52 77 53 0 38 3 25 0 31 0 01 0 03 0 10 90 84

s 0 03 0 04 0 02 0 06 0 02 0 03 0 04 0 01 0 01 0 07

qz (1) 90 30 0 19 0 49 0 49 0 04 0 49 0 70 0 02 0 03 0 00 92 75

HG394 1050 Melt (3) 63 41 1 33 14 13 7 79 0 22 2 59 7 48 2 26 0 80 99 99

s 0 39 0 03 0 34 0 13 0 02 0 08 0 15 0 06 0 03

plag (5) 49 39 0 06 29 47 1 63 0 01 0 27 14 83 2 39 0 06 98 11

s 0 14 0 03 0 26 0 08 0 01 0 05 0 21 0 11 0 01

opx (6) 52 19 0 32 2 41 16 28 0 52 25 05 1 91 0 04 0 01 98 73

s 0 53 0 06 0 41 0 20 0 03 0 50 0 21 0 01 0 01

cpx (10) 50 35 0 52 3 03 11 11 0 40 15 86 16 72 0 21 0 01 98 21

s 0 49 0 09 0 51 0 91 0 03 0 42 1 57 0 02 0 01

mt (1) 0 49 5 65 3 94 75 85 0 39 3 24 0 37 0 00 0 00 89 93

HG385 1100 Melt (5) 59 71 1 49 14 09 10 33 0 16 3 72 8 40 1 48 0 62 0 01 100 00

s 0 37 0 02 0 39 0 13 0 03 0 20 0 23 0 06 0 03 0 01

plag (6) 49 10 0 07 29 99 1 69 0 02 0 42 15 21 2 26 0 07 0 01 98 84

s 0 19 0 04 0 18 0 22 0 01 0 25 0 06 0 10 0 02 0 01

opx (4) 54 17 0 17 1 37 15 27 0 37 24 57 4 05 0 08 0 03 0 01 100 09

s 0 17 0 05 0 17 0 22 0 03 0 62 0 61 0 01 0 01 0 01

cpx (3) 50 23 0 55 3 46 10 57 0 23 15 51 18 16 0 25 0 02 0 00 98 98

s 0 37 0 08 0 12 0 24 0 04 0 10 0 34 0 01 0 01 0 00

mt (3) 0 28 4 65 4 53 76 86 0 22 3 82 0 35 0 01 0 01 0 15 90 88

s 0 01 0 06 0 03 0 27 0 02 0 11 0 04 0 01 0 01 0 02

HG393 1100 Melt (16) 56 20 1 45 14 18 11 74 0 28 4 53 9 28 1 90 0 44 100 00

s 0 27 0 04 0 20 0 13 0 02 0 11 0 10 0 06 0 02

plag (58) 49 73 0 05 29 34 1 76 0 02 0 32 14 50 2 53 0 06 98 31

s 0 65 0 04 0 72 0 27 0 02 0 10 0 46 0 21 0 01

opx (37) 52 37 0 26 3 02 15 05 0 47 25 79 2 05 0 04 0 01 99 06

(continued)


1584


11/29

Table 3: Continued


s 0 24 0 03 0 19 0 35 0 03 0 37 0 22 0 01 0 01

cpx (39) 50 30 0 47 3 53 10 50 0 39 16 00 16 97 0 24 0 01 98 41

s 0 41 0 07 0 33 0 41 0 03 0 43 0 71 0 02 0 01

mt (15) 0 54 3 99 5 21 76 84 0 34 4 06 0 31 0 03 0 01 91 33

s 0 05 0 07 0 19 0 41 0 04 0 11 0 06 0 02 0 01

HG402 1115 Melt (27) 52 68 1 29 14 36 13 68 0 29 5 52 9 98 1 91 0 31 100 01

s 0 31 0 04 0 21 0 15 0 02 0 10 0 14 0 04 0 02

plag (60) 49 43 0 04 30 20 1 69 0 02 0 30 15 32 2 45 0 05 99 50

s 0 92 0 04 0 76 0 22 0 02 0 13 0 61 0 33 0 01

opx (25) 52 55 0 22 3 20 15 46 0 41 25 60 2 17 0 04 0 00 99 65

s 0 29 0 03 0 17 0 21 0 03 0 26 0 14 0 02 0 01

cpx (33) 50 41 0 41 3 85 10 71 0 32 15 93 17 33 0 23 0 01 99 20

s 0 36 0 05 0 40 0 51 0 04 0 54 0 98 0 03 0 01

mt (7) 0 26 3 67 6 57 75 41 0 26 4 52 0 26 0 01 0 01 90 97

s 0 02 0 06 0 13 0 43 0 03 0 06 0 03 0 01 0 01

HG395 1125 Melt (31) 52 14 1 30 14 46 13 66 0 29 5 84 10 10 1 90 0 31 100 00

s 0 28 0 05 0 15 0 15 0 03 0 17 0 10 0 06 0 02

plag (62) 49 56 0 03 29 97 1 43 0 02 0 33 14 60 2 63 0 06 98 63

s 0 62 0 03 0 52 0 13 0 02 0 07 0 41 0 21 0 02

opx (17) 52 02 0 22 3 40 15 04 0 40 25 90 2 17 0 04 0 01 99 20

s 0 40 0 03 0 10 0 31 0 03 0 26 0 20 0 01 0 01

cpx (58) 50 03 0 40 4 27 11 01 0 33 16 56 15 71 0 21 0 01 98 53

s 0 39 0 07 0 60 0 81 0 05 1 03 1 99 0 03 0 01

HG392 1150 Melt (24) 52 14 1 02 14 99 11 59 0 27 6 94 10 95 1 83 0 25 0 01 99 99

s 0 19 0 04 0 16 0 25 0 03 0 13 0 21 0 10 0 02 0 01

plag (102) 47 88 0 03 31 05 1 33 0 02 0 34 16 06 2 10 0 05 0 01 98 87

s 0 84 0 03 0 60 0 14 0 02 0 17 0 59 0 32 0 01 0 01

opx (77) 53 09 0 16 3 35 12 15 0 35 28 18 2 29 0 03 0 02 0 04 99 66s 0 43 0 03 0 44 0 31 0 03 0 43 0 22 0 02 0 01 0 03

cpx (94) 50 56 0 35 4 26 8 51 0 27 17 03 17 88 0 23 0 02 0 05 99 16

s 0 60 0 08 1 20 0 54 0 03 0 48 1 36 0 07 0 01 0 03

1 14% H 2 O

HG389 950 plag (2) 48 36 0 03 29 49 1 89 0 02 0 27 14 77 2 48 0 08 0 02 97 41

s 0 28 0 01 0 52 0 13 0 02 0 11 0 24 0 10 0 01 0 02

opx (1) 53 53 0 28 2 28 13 36 0 74 28 32 1 62 0 02 0 03 0 01 100 19

cpx(3) 49 31 0 7 3 97 8 24 0 46 16 3 18 2 0 26 0 01 0 03 97 48

s 0 54 0 05 0 16 0 22 0 02 0 21 0 57 0 05 0 01 0 01

mt (1) 1 73 10 9 1 31 71 99 0 09 1 35 0 89 0 0 03 0 88 29

qz (1) 94 39 0 16 0 55 0 34 0 0 05 0 17 0 0 03 0 03 95 72

HG379 1000 Melt (3) 68 61 0 87 14 12 5 06 0 11 2 03 5 77 2 70 0 73 0 01 100 01

s 0 56 0 00 0 09 0 09 0 02 0 23 0 02 0 03 0 04 0 01plag (10) 48 02 0 04 31 60 1 30 0 01 0 19 15 85 2 00 0 06 0 01 99 08

s 0 27 0 02 0 36 0 13 0 01 0 11 0 28 0 18 0 02 0 01

opx (13) 52 86 0 31 2 64 13 43 0 39 26 12 1 76 0 02 0 02 0 01 97 56

s 0 69 0 05 0 40 0 53 0 04 0 81 0 29 0 02 0 01 0 01

cpx (7) 50 59 0 52 3 27 8 32 0 28 14 89 19 06 0 27 0 04 0 02 97 26

(continued)


1585


12/29

Table 3: Continued


s 0 69 0 08 0 58 0 28 0 04 0 89 0 69 0 01 0 01 0 02

mt (8) 0 29 4 19 3 47 77 44 0 26 3 41 0 32 0 01 0 02 0 08 89 49

s 0 10 0 11 0 05 0 43 0 03 0 06 0 05 0 01 0 01 0 02

ilm (3) 0 18 22 49 0 99 63 69 0 12 1 97 0 42 0 00 0 02 0 06 89 94

s 0 01 0 08 0 01 0 55 0 01 0 07 0 06 0 00 0 01 0 03

HG384 1050 Melt (5) 63 80 1 15 14 46 6 89 0 22 2 98 7 23 2 63 0 62 0 01 99 99

s 0 21 0 01 0 11 0 08 0 03 0 02 0 08 0 08 0 02 0 02

plag (6) 47 22 0 08 30 61 1 61 0 03 0 31 15 62 1 95 0 07 0 02 97 52

s 0 42 0 04 0 50 0 14 0 02 0 13 0 28 0 03 0 02 0 02

opx (10) 52 76 0 25 2 58 12 16 0 57 28 80 1 68 0 03 0 02 0 01 98 86

s 0 38 0 04 0 35 0 60 0 04 0 47 0 21 0 02 0 01 0 01

cpx (9) 49 82 0 56 3 45 8 45 0 40 16 05 19 16 0 28 0 02 0 01 98 20

s 0 62 0 06 0 17 0 33 0 03 0 43 0 53 0 02 0 01 0 02

mt (8) 0 31 3 80 3 98 76 93 0 39 4 13 0 34 0 01 0 02 0 08 89 99

s 0 05 0 08 0 06 0 41 0 04 0 11 0 05 0 01 0 01 0 06

HG394 1050 Melt (18) 62 10 1 11 14 95 7 61 0 22 3 21 7 67 2 58 0 52 99 99

s 0 14 0 04 0 15 0 13 0 03 0 10 0 07 0 04 0 03

plag (21) 47 83 0 04 30 66 1 53 0 02 0 24 15 60 1 98 0 04 97 94

s 0 43 0 02 0 30 0 11 0 02 0 05 0 28 0 10 0 01

opx (29) 52 27 0 26 3 20 13 03 0 52 27 21 1 89 0 04 0 01 98 43

s 0 45 0 03 0 33 0 54 0 03 0 50 0 24 0 01 0 01

cpx (31) 49 68 0 55 4 02 8 68 0 38 15 40 19 07 0 27 0 01 98 06

s 0 38 0 08 0 60 0 38 0 03 0 42 0 39 0 03 0 01

mt (9) 0 50 3 69 4 31 76 14 0 37 4 01 0 27 0 02 0 00 89 31

s 0 04 0 22 0 11 1 66 0 03 0 08 0 05 0 01 0 01

HG385 1100 Melt (9) 58 78 1 22 14 73 9 13 0 17 4 33 8 84 2 35 0 44 0 01 100 01

s 0 30 0 06 0 17 0 10 0 02 0 06 0 14 0 09 0 02 0 01

plag (5) 47 69 0 11 30 77 2 05 0 03 0 37 15 87 1 94 0 07 0 02 98 92s 0 78 0 08 0 48 0 38 0 02 0 11 0 32 0 11 0 03 0 01

opx (17) 52 71 0 27 3 59 10 90 0 32 29 55 1 91 0 03 0 02 0 01 99 31

s 0 78 0 04 0 56 0 59 0 02 0 65 0 25 0 01 0 01 0 01

cpx (10) 50 11 0 48 3 91 8 48 0 24 16 39 19 17 0 28 0 02 0 02 99 10

s 0 94 0 09 0 55 0 52 0 02 0 60 0 36 0 03 0 01 0 02

mt (6) 0 28 2 81 4 99 77 25 0 22 4 91 0 28 0 01 0 03 0 10 90 88

s 0 10 0 07 0 06 0 34 0 04 0 09 0 03 0 01 0 01 0 02

HG393 1100 Melt (27) 56 36 1 15 14 95 10 51 0 26 4 89 9 40 2 12 0 35 99 99

s 0 29 0 04 0 20 0 12 0 03 0 13 0 09 0 05 0 02

plag (60) 48 87 0 04 29 77 1 84 0 03 0 39 15 13 2 25 0 05 98 37

s 0 94 0 03 0 74 0 19 0 02 0 12 0 58 0 29 0 02

opx (21) 52 83 0 24 3 67 11 47 0 47 28 41 2 02 0 04 0 01 99 16

s 0 56 0 02 0 63 0 52 0 05 0 41 0 24 0 02 0 01cpx (33) 49 71 0 49 4 34 8 96 0 36 15 83 18 44 0 25 0 01 98 39

s 0 47 0 06 0 50 0 39 0 04 0 41 0 60 0 02 0 01

mt (13) 0 50 2 73 5 68 77 11 0 34 4 81 0 30 0 01 0 00 91 48

s 0 02 0 06 0 09 0 31 0 03 0 09 0 05 0 02 0 01

HG402 1115 Melt (30) 52 06 1 05 15 04 12 74 0 27 6 27 10 44 1 89 0 24 100 00

s 0 28 0 04 0 15 0 10 0 03 0 08 0 09 0 06 0 02

(continued)


1586


13/29

Table 3: Continued


plag (51) 48 26 0 03 31 03 1 37 0 01 0 24 16 07 1 96 0 03 99 00

s 1 00 0 02 0 68 0 10 0 02 0 06 0 65 0 32 0 02

opx (35) 52 77 0 17 3 56 13 12 0 39 27 05 2 19 0 03 0 01 99 29

s 0 56 0 04 0 53 0 56 0 04 0 48 0 22 0 02 0 01

cpx (25) 49 90 0 37 4 38 9 49 0 29 15 80 18 66 0 23 0 01 99 13

s 0 60 0 05 0 49 0 44 0 03 0 41 0 49 0 02 0 01

mt (14) 0 22 2 51 7 51 74 51 0 28 5 35 0 13 0 01 0 01 90 53

s 0 02 0 15 0 12 2 46 0 03 0 10 0 03 0 01 0 01

HG395 1125 Melt (30) 51 79 1 02 15 15 12 93 0 27 6 36 10 40 1 85 0 24 100 00

s 0 25 0 05 0 10 0 13 0 03 0 11 0 09 0 05 0 02

plag (43) 48 38 0 02 30 76 1 40 0 01 0 29 15 50 2 21 0 04 98 61

s 0 90 0 02 0 71 0 11 0 02 0 07 0 56 0 29 0 02

opx (47) 52 01 0 18 4 02 13 38 0 37 27 02 2 13 0 03 0 01 99 15

s 0 82 0 03 0 73 0 42 0 03 0 63 0 25 0 02 0 01

cpx (22) 49 35 0 39 4 77 9 68 0 29 15 84 17 97 0 23 0 01 98 53

s 0 83 0 07 0 70 0 42 0 03 0 76 0 63 0 02 0 01

mt (9) 0 28 2 42 7 61 75 81 0 28 5 34 0 12 0 02 0 01 91 89

s 0 03 0 06 0 16 0 30 0 02 0 09 0 03 0 02 0 01

HG392 1150 Melt (13) 50 90 0 83 15 32 11 70 0 25 7 63 11 50 1 65 0 20 0 01 99 99

s 0 35 0 06 0 16 0 12 0 03 0 16 0 09 0 05 0 01 0 01

plag (9) 48 55 0 02 30 57 1 21 0 01 0 31 15 83 2 38 0 05 0 01 98 94

s 0 34 0 03 0 38 0 11 0 01 0 02 0 27 0 17 0 01 0 01

1 66% H 2 O

HG389 950 Melt (1) 70 79 0 22 15 40 2 81 0 14 1 74 5 68 2 57 0 63 0 00 99 99

plag (5) 47 59 0 00 31 89 1 50 0 02 0 14 16 63 1 76 0 04 0 02 99 59

s 0 60 0 01 0 62 0 21 0 01 0 04 0 34 0 17 0 01 0 02

opx (5) 54 06 0 28 3 31 8 95 0 61 30 63 1 17 0 04 0 02 0 02 99 09

s 0 41 0 12 0 40 0 84 0 06 0 49 0 17 0 02 0 01 0 01cpx (6) 49 80 0 62 4 12 7 19 0 40 15 35 20 94 0 27 0 02 0 03 98 74

s 0 64 0 18 0 45 0 70 0 03 0 20 0 51 0 04 0 01 0 02

mt (2) 0 14 0 75 3 18 79 96 0 50 4 66 0 31 0 03 0 02 0 06 89 61

s 0 01 0 00 0 08 0 34 0 04 0 04 0 01 0 02 0 00 0 02

ilm (4) 0 17 8 45 1 31 76 17 0 11 1 05 0 37 0 01 0 02 0 04 87 70

s 0 04 0 58 0 02 0 38 0 02 0 10 0 03 0 02 0 01 0 03

HG379 1000 Melt (6) 64 05 1 00 16 30 5 37 0 14 2 77 7 04 2 83 0 48 0 01 99 99

s 0 27 0 05 0 06 0 04 0 03 0 07 0 13 0 03 0 01 0 01

plag (6) 46 18 0 05 32 41 1 29 0 02 0 12 16 87 1 58 0 04 0 01 98 57

s 0 39 0 01 0 13 0 11 0 02 0 02 0 17 0 05 0 01 0 01

opx (8) 52 82 0 24 3 46 10 74 0 36 27 41 1 59 0 05 0 02 0 02 96 71

s 0 33 0 03 0 88 0 51 0 04 0 88 0 41 0 05 0 01 0 01

cpx (13) 49 40 0 60 3 99 7 67 0 24 15 14 19 78 0 27 0 03 0 01 97 13s 0 59 0 10 0 30 0 42 0 03 0 59 0 85 0 03 0 01 0 01

mt (8) 0 26 3 17 4 29 76 66 0 31 4 64 0 31 0 02 0 03 0 07 89 76

s 0 10 0 07 0 05 0 38 0 02 0 07 0 09 0 02 0 01 0 02

HG363 1050 Melt (13) 60 11 1 02 16 45 6 88 0 23 3 98 8 33 2 59 0 39 0 01 99 99

s 0 36 0 03 0 21 0 15 0 02 0 08 0 13 0 04 0 02 0 02

plag (15) 46 71 0 06 31 68 1 64 0 02 0 32 16 49 1 65 0 05 0 01 98 63

(continued)


1587


14/29

Table 3: Continued


s 0 55 0 03 0 55 0 24 0 02 0 13 0 32 0 12 0 02 0 01

opx (10) 52 61 0 25 3 43 10 86 0 48 28 98 1 98 0 02 0 02 0 01 98 64

s 0 71 0 03 0 50 0 42 0 03 0 72 0 33 0 01 0 01 0 01

cpx (11) 49 42 0 55 4 38 7 97 0 35 15 95 19 71 0 26 0 02 0 01 98 62

s 0 42 0 05 0 28 0 42 0 04 0 45 0 68 0 03 0 01 0 02

mt (11) 0 23 2 59 4 94 77 52 0 42 5 50 0 29 0 01 0 02 0 06 91 58

s 0 06 0 03 0 09 0 43 0 03 0 08 0 05 0 01 0 01 0 02

HG394 1050 Melt (28) 59 89 1 02 16 09 7 42 0 23 4 05 8 35 2 55 0 39 99 99

s 0 27 0 03 0 18 0 12 0 02 0 06 0 10 0 05 0 02

plag (26) 46 54 0 04 31 65 1 51 0 02 0 23 16 61 1 60 0 02 98 22

s 0 51 0 03 0 51 0 08 0 02 0 12 0 50 0 21 0 01

opx (38) 52 88 0 24 3 90 11 43 0 47 28 42 1 81 0 06 0 01 99 22

s 0 92 0 04 0 69 0 57 0 03 0 78 0 28 0 04 0 01

cpx (10) 50 01 0 51 5 31 7 69 0 31 14 95 19 78 0 36 0 02 98 94

s 0 38 0 05 1 81 0 30 0 03 1 00 0 22 0 09 0 02

mt (10) 0 47 2 82 5 10 76 76 0 38 4 98 0 22 0 04 0 00 90 77

s 0 02 0 04 0 10 0 35 0 03 0 11 0 05 0 02 0 01

HG385 1100 Melt (9) 56 34 0 99 16 05 8 66 0 17 5 59 9 62 2 28 0 30 0 01 100 00

s 0 26 0 03 0 12 0 10 0 02 0 07 0 13 0 04 0 02 0 02

plag (8) 46 72 0 04 31 94 1 63 0 01 0 22 16 78 1 58 0 03 0 01 98 96

s 0 28 0 02 0 32 0 12 0 01 0 05 0 19 0 11 0 01 0 01

opx (10) 53 94 0 17 3 65 8 67 0 27 31 46 1 61 0 03 0 02 0 00 99 82

s 0 49 0 03 0 42 0 38 0 02 0 43 0 17 0 03 0 01 0 01

cpx (8) 50 15 0 36 4 42 7 17 0 19 16 54 19 88 0 30 0 03 0 01 99 05

s 0 42 0 04 0 44 0 37 0 03 0 56 0 30 0 03 0 01 0 02

mt (7) 0 27 1 67 6 41 75 01 0 26 6 89 0 29 0 02 0 02 0 07 90 91

s 0 08 0 02 0 05 0 25 0 01 0 06 0 03 0 02 0 01 0 03

HG402 1115 Melt (20) 51 46 0 82 15 93 11 46 0 24 7 15 11 03 1 73 0 19 100 01s 0 33 0 04 0 28 0 11 0 03 0 16 0 07 0 04 0 01

plag (17) 47 62 0 02 31 22 1 48 0 02 0 25 16 50 1 86 0 03 99 00

s 1 12 0 02 0 93 0 09 0 02 0 06 0 76 0 39 0 01

cpx (5) 50 58 0 24 4 04 8 08 0 26 17 14 18 54 0 20 0 01 99 09

s 0 30 0 05 0 23 0 43 0 04 0 18 0 83 0 02 0 01

mt (3) 0 19 1 70 9 06 73 12 0 27 6 65 0 15 0 01 0 00 91 15

s 0 00 0 07 0 17 0 18 0 01 0 15 0 05 0 01 0 00

HG395 1125 Melt (16) 50 98 0 80 16 15 11 54 0 24 7 23 11 19 1 70 0 18 100 01

s 0 27 0 03 0 16 0 13 0 02 0 10 0 08 0 07 0 02

plag (9) 47 57 0 02 31 00 1 40 0 02 0 26 15 74 2 07 0 03 98 11

s 0 64 0 02 0 45 0 09 0 02 0 02 0 19 0 14 0 02

HG392 1150 Melt (17) 50 46 0 74 16 36 11 04 0 23 7 20 12 05 1 73 0 18 0 01 100 01

s 0 32 0 03 0 26 0 11 0 03 0 14 0 09 0 03 0 02 0 01plag (11) 47 15 0 01 31 72 1 11 0 02 0 23 16 94 1 77 0 04 0 00 98 99

s 0 37 0 02 0 22 0 04 0 01 0 01 0 19 0 10 0 02 0 01

2 43% H 2 O

HG389 950 Melt (14) 68 38 0 28 16 26 3 24 0 15 2 28 6 13 2 76 0 54 0 01 100 00

s 0 51 0 03 0 15 0 08 0 02 0 08 0 15 0 07 0 02 0 02

plag (64) 47 14 0 02 32 07 1 35 0 02 0 12 16 83 1 61 0 03 0 01 99 20

(continued)


1588


15/29

Table 3: Continued


s 0 62 0 02 0 39 0 11 0 02 0 13 0 37 0 19 0 01 0 02

opx (19) 54 92 0 07 3 34 6 45 0 73 32 20 1 26 0 02 0 02 0 01 99 02

s 0 37 0 03 0 40 0 35 0 05 0 52 0 44 0 02 0 01 0 01

cpx (39) 50 32 0 37 4 23 6 15 0 41 15 35 21 77 0 29 0 02 0 01 98 92

s 0 93 0 23 0 70 0 56 0 05 0 47 0 26 0 03 0 01 0 02

mt (4) 0 12 0 74 3 63 78 83 0 60 5 68 0 25 0 02 0 03 0 06 89 96

s 0 03 0 02 0 07 0 12 0 03 0 09 0 02 0 01 0 01 0 02

ilm (13) 0 08 7 28 1 44 76 50 0 10 1 25 0 25 0 01 0 03 0 02 86 96

s 0 03 0 66 0 04 1 52 0 03 0 22 0 06 0 01 0 01 0 02

HG379 1000 Melt (7) 60 44 0 89 17 48 5 64 0 21 3 86 8 20 2 88 0 39 0 00 100 00

s 0 52 0 03 0 11 0 07 0 02 0 05 0 14 0 09 0 02 0 00

plag (7) 44 47 0 02 32 22 1 43 0 02 0 18 17 24 1 38 0 04 0 01 97 01

s 0 26 0 02 0 19 0 20 0 02 0 02 0 19 0 07 0 01 0 01

opx (4) 51 68 0 24 4 26 9 28 0 49 29 65 1 85 0 04 0 02 0 01 97 52

s 0 46 0 06 0 81 0 66 0 03 0 84 0 45 0 04 0 01 0 02

cpx (9) 47 32 0 65 4 96 8 05 0 33 15 39 20 23 0 29 0 02 0 01 97 25

s 0 50 0 08 0 44 0 80 0 04 0 20 0 63 0 04 0 02 0 02

mt (4) 0 23 2 39 5 18 75 39 0 45 5 97 0 30 0 01 0 02 0 10 90 04

s 0 02 0 05 0 05 0 37 0 01 0 10 0 06 0 01 0 01 0 02

HG391 1000 Melt (14) 66 16 0 47 16 03 4 08 0 17 3 04 6 96 2 64 0 44 0 02 100 01

s 0 28 0 02 0 11 0 08 0 03 0 09 0 07 0 04 0 02 0 02

plag (5) 47 26 0 03 31 67 1 38 0 02 0 14 17 00 1 43 0 03 0 01 98 97

s 0 60 0 02 0 43 0 09 0 02 0 05 0 43 0 18 0 01 0 01

opx (10) 54 70 0 23 3 82 7 70 0 48 31 30 1 40 0 03 0 03 0 01 99 70

s 0 64 0 06 0 53 0 50 0 07 0 67 0 22 0 02 0 02 0 01

cpx (45) 50 40 0 57 4 45 6 51 0 30 15 42 21 27 0 27 0 02 0 01 99 22

s 0 78 0 12 0 69 0 52 0 03 0 46 0 41 0 03 0 01 0 02

mt (6) 0 22 0 94 4 37 75 41 0 60 6 96 0 26 0 01 0 02 0 05 88 84s 0 12 0 02 0 07 0 26 0 02 0 05 0 03 0 01 0 01 0 05

ilm (25) 0 07 9 55 1 54 74 02 0 10 1 75 0 24 0 01 0 02 0 03 87 33

s 0 02 0 37 0 04 0 50 0 02 0 09 0 05 0 01 0 01 0 02

HG388 1050 Melt (16) 57 61 0 86 17 32 6 95 0 23 4 99 9 28 2 44 0 30 0 02 100 01

s 0 45 0 04 0 07 0 09 0 03 0 05 0 12 0 06 0 01 0 02

plag (48) 46 25 0 03 32 70 1 41 0 02 0 16 17 34 1 46 0 03 0 01 99 41

s 0 78 0 02 0 46 0 09 0 02 0 06 0 49 0 23 0 01 0 01

opx (22) 52 79 0 22 5 12 8 90 0 41 30 91 1 69 0 02 0 02 0 01 100 09

s 0 54 0 03 0 54 0 32 0 03 0 44 0 24 0 01 0 01 0 02

cpx-1 (42) 49 69 0 47 4 78 7 05 0 27 15 90 21 36 0 25 0 02 0 01 99 80

s 0 60 0 06 0 51 0 35 0 02 0 39 0 21 0 02 0 01 0 02

cpx-2 (4) 43 12 0 83 11 40 10 76 0 21 11 32 21 28 0 23 0 02 0 01 99 18

s 0 89 0 08 0 70 0 49 0 04 0 73 0 45 0 03 0 01 0 01mt (12) 0 12 1 83 6 29 76 71 0 44 7 03 0 21 0 01 0 02 0 07 92 73

s 0 02 0 06 0 08 0 28 0 03 0 09 0 08 0 01 0 01 0 02

HG381 1100 Melt (15) 53 36 0 79 17 05 8 84 0 23 6 58 10 99 1 93 0 22 0 01 100 00

s 0 18 0 03 0 30 0 10 0 02 0 14 0 14 0 05 0 01 0 02

plag (7) 44 78 0 03 32 29 1 69 0 01 0 20 17 23 1 15 0 04 0 00 97 42

s 0 31 0 03 0 21 0 18 0 01 0 06 0 24 0 12 0 01 0 00

(continued)


1589


16/29

Table 3: Continued


cpx (6) 46 81 0 46 6 01 7 82 0 20 14 55 20 78 0 25 0 02 0 01 96 91

s 0 72 0 08 0 52 0 56 0 02 0 43 0 19 0 02 0 01 0 01

mt (2) 0 14 1 15 7 81 71 55 0 35 8 13 0 28 0 00 0 03 0 08 89 52

s 0 01 0 01 0 01 0 24 0 02 0 03 0 05 0 00 0 00 0 02

2 83% H 2 O

HG389 950 Melt (13) 67 23 0 36 16 79 3 30 0 16 2 50 6 48 2 69 0 47 0 01 100 00

s 0 51 0 05 0 13 0 08 0 02 0 04 0 06 0 12 0 03 0 01

plag (57) 46 71 0 02 32 52 1 32 0 02 0 11 17 27 1 41 0 03 0 01 99 42

s 0 42 0 02 0 31 0 14 0 02 0 14 0 36 0 15 0 01 0 01

opx (31) 55 15 0 07 3 40 6 06 0 67 32 74 1 08 0 02 0 02 0 01 99 22

s 0 63 0 03 0 61 0 48 0 05 0 51 0 15 0 01 0 01 0 02

cpx (33) 51 12 0 17 3 73 5 61 0 41 15 81 22 06 0 28 0 02 0 01 99 22

s 0 56 0 08 0 37 0 37 0 04 0 35 0 22 0 02 0 01 0 02

mt (4) 0 10 0 82 4 02 78 04 0 58 6 19 0 20 0 01 0 02 0 05 90 03

s 0 02 0 03 0 02 0 50 0 04 0 05 0 02 0 01 0 01 0 02

ilm (17) 0 06 7 57 1 56 76 78 0 09 1 42 0 21 0 00 0 02 0 02 87 73

s 0 02 0 76 0 06 0 79 0 03 0 30 0 05 0 01 0 01 0 02

HG379 1000 Melt (15) 60 15 0 84 17 58 5 58 0 22 4 06 8 44 2 76 0 37 0 01 100 01

s 0 25 0 03 0 08 0 10 0 03 0 05 0 10 0 07 0 01 0 02

plag (12) 44 67 0 02 32 46 1 26 0 02 0 12 17 71 1 19 0 03 0 01 97 49

s 0 31 0 02 0 25 0 05 0 01 0 03 0 32 0 10 0 01 0 01

opx (8) 52 30 0 24 4 21 8 60 0 43 30 58 1 40 0 03 0 02 0 01 97 82

s 0 50 0 04 0 49 0 32 0 03 0 54 0 20 0 01 0 01 0 02

cpx (8) 48 84 0 52 4 60 6 81 0 31 15 57 20 82 0 26 0 02 0 01 97 76

s 0 70 0 08 0 69 0 62 0 02 0 62 0 37 0 02 0 01 0 02

mt (4) 0 24 2 20 5 54 75 25 0 46 6 33 0 26 0 01 0 03 0 07 90 39

s 0 09 0 07 0 07 0 71 0 03 0 06 0 04 0 02 0 02 0 01

HG391 1000 Melt (16) 64 26 0 49 16 84 4 16 0 21 3 55 7 54 2 56 0 38 0 01 100 01s 0 41 0 03 0 13 0 08 0 03 0 04 0 10 0 06 0 02 0 01

plag (47) 46 96 0 02 31 93 1 35 0 02 0 13 17 10 1 36 0 03 0 01 98 91

s 0 56 0 02 0 40 0 07 0 02 0 09 0 40 0 21 0 01 0 01

opx (16) 54 39 0 21 4 11 7 28 0 44 31 75 1 39 0 02 0 02 0 01 99 62

s 0 57 0 05 0 57 0 39 0 04 0 69 0 46 0 02 0 01 0 01

cpx (54) 50 10 0 57 4 80 6 41 0 28 15 36 21 43 0 28 0 02 0 01 99 26

s 0 64 0 14 0 75 0 43 0 03 0 47 0 29 0 04 0 01 0 02

mt (11) 0 15 0 92 4 99 74 25 0 60 7 76 0 21 0 01 0 02 0 04 88 95

s 0 04 0 05 0 07 0 37 0 05 0 07 0 04 0 02 0 01 0 02

ilm (12) 0 04 10 42 1 61 73 55 0 13 2 33 0 09 0 01 0 02 0 03 88 23

s 0 02 0 71 0 04 0 82 0 03 0 23 0 06 0 01 0 01 0 02

HG388 1050 Melt (24) 56 43 0 80 17 79 6 88 0 23 5 67 9 72 2 22 0 27 0 01 100 01

s 0 28 0 04 0 07 0 11 0 03 0 07 0 10 0 07 0 02 0 02plag (48) 45 51 0 01 33 48 1 22 0 01 0 14 18 16 1 07 0 02 0 01 99 63

s 0 37 0 01 0 36 0 08 0 01 0 04 0 30 0 15 0 01 0 02

cpx-1 (24) 50 75 0 32 4 04 6 61 0 29 17 30 19 99 0 20 0 02 0 02 99 54

s 0 32 0 05 0 36 0 48 0 04 0 63 0 99 0 03 0 01 0 02

cpx-2 (4) 41 99 0 92 11 70 11 68 0 18 10 64 21 58 0 23 0 02 0 04 98 98

s 0 52 0 13 0 52 0 54 0 03 0 37 0 36 0 02 0 01 0 03

(continued)


1590


17/29

Calc-alkalic series magmas are clearly dominant in conti-nental arcs and mature arcs with thicker crust(Miyashiro, 1971; Baker, 1974; Gill, 1981), whereas tholeiiticseries magmas characterize magmatism in intra-oceanicarcs. However, tholeiitic rocks have been found to coexistwith calc-alkalic rocks in a single volcanic system in somemature arcs; for example, Mt. Shasta, USA (Baker et al .,1994); Chichontepec, El Salvador (Bau & Knittel, 1993),Aso in SW Japan (Hunter,1998), and Myoko-Kurohime inCentral Japan (Sakuyama, 1981). Furthermore, along thevolcanic front of the NE Japan arc about one-third of theQuaternary volcanoes erupt both tholeiitic and calc-alka-lic magmas (Kawano et al ., 1961). The genetic relationship

between these two magma series is therefore critical tobetter understand andesite genesis and arc crust evolution.It has been repeatedly proposed that the calc-alkalic

trend can be reproduced by differentiation of hydrousbasaltic magmas (e.g. Green & Ringwood, 1967; Sisson &Grove, 1993; Kawamoto, 1996; Grove et al ., 2003;Pichavant & Macdonald, 2007). Hamada & Fujii (2008)examined the LLDs of a basalt similar to the startingmaterial used here but under different pressure^ f O 2 ^H 2Oconditions and demonstrated that the LLDs define calc-alkalic trends in the presence of 4 2 wt % H 2O. Althoughour LLDs cannot be directly compared with those of Hamada & Fujii (2008) because of the higher f O 2 in our

experiments, they also tend to become more calc-alkalicwith increasing H 2O (Figs 9 and 10). It may be thus sug-gested that magmatic differentiation under hydrous condi-tions, such as crystallization of a hydrous basaltic magmaor anatexis of gabbro or amphibolite crust under hydrousconditions, could be a likely process to derive IBM calc-alkalic magmas.

To examine more quantitatively the amount of H 2O thatplays a role in IBM magma differentiation, the LLDs are

compared with the compositions of the IBM volcanicrocks on SiO 2-variation diagrams. TiO 2, Al2O 3 and FeOcontents at a constant SiO 2 content change significantlywith H 2O (Fig. 9), allowing H 2O content in the parental,basaltic materials to be estimated. The compositionaltrends for the IBM calc-alkalic rocks are best reproducedby an LLD in the presence of 2 5^3 wt % H 2O, which isconsistent with the experimental data of Hamada & Fujii(2008).

Tholeiitic differentiation trends observed within theIBM rocks are characterized by an increase in both TiO 2and FeO with increasing degrees of differentiationduring the earlier stages (Figs 3 and 9), which may be

most consistent with an LLD in the presence of 0 49wt %H 2O (Fig. 9). Such an LLD also shows a typical tholeiitictrend in an AFM ternary diagram (Fig. 10). H 2O-poorparental magmas for the IBM tholeiitic series are also con-sistent with the observation that tholeiitic felsic volcanicrocks from the Sumisu caldera volcano (Fig. 1) exhibit pyr-oxene crystallization temperatures higher than 1050 8C(Shukuno et al ., 2006); such temperatures in differentiatedfelsic magmas could be attained only by H 2O-poormagma differentiation (Fig. 11). Analyses of melt inclusionsin olivine and/or plagioclase phenocrysts also provide con-straints on H 2O content in IBM tholeiitic magmas: Saitoet al . (2005) and Hamada & Fujii (2007) analyzed melt

inclusions in tholeiitic basalts from Miyake-jima and Izu-Oshima volcanoes and reported 0 2^2 4 w t % H 2O inmelt inclusions. Because these melt inclusions have basalticcompositions more differentiated than that of the basaltused here, an inferred H 2O content in a parental basaltmagma ( 0 49 wt %) would be consistent with the meltinclusion data.

It should be stressed that partial melting of an H 2O-poor basaltic lower crust, an inverse process of equilibrium

Table 3: Continued


mt (11) 0 09 1 58 7 36 75 53 0 41 8 00 0 18 0 02 0 03 0 07 93 27

s 0 02 0 05 0 13 0 35 0 02 0 12 0 07 0 02 0 01 0 04

HG381 1100 Melt (20) 52 67 0 77 17 22 8 81 0 22 6 68 11 64 1 78 0 20 0 01 99 99

s 0 21 0 05 0 12 0 10 0 02 0 11 0 10 0 05 0 02 0 01

plag (9) 44 37 0 03 32 32 1 71 0 01 0 23 17 41 0 99 0 03 0 01 97 11

s 0 36 0 03 0 43 0 13 0 01 0 13 0 24 0 10 0 02 0 02

cpx (12) 45 89 0 45 6 75 8 50 0 17 13 61 21 32 0 23 0 02 0 01 96 95

s 0 63 0 05 0 27 0 36 0 02 0 28 0 18 0 01 0 01 0 02

mt (5) 0 12 0 99 8 02 71 38 0 34 8 54 0 28 0 02 0 03 0 07 89 79

s 0 01 0 02 0 07 0 26 0 03 0 11 0 01 0 01 0 00 0 02

Numbers in parentheses indicate number of samples analyzed.


1591


18/29

Fig. 5. Element partitioning between (a) orthopyroxene (opx) and clinopyroxene (cpx) and (b) plagioclase (plag) and melt during the experi-ments. These exchange distribution coefficients are similar to those previously reported (e.g. Gaetani et al ., 1993; Sisson & Grove, 1993; Groveet al ., 2003; Sisson et al ., 2005; Tatsumi et al ., 2006; Hamada & Fujii, 2008).


1592


19/29

crystallization of an H 2O-poor basaltic magma, is also apossible process for production of tholeiitic magmas, assuggested by Tatsumi et al . (2008 b) for tholeiitic magmasat Zao volcano, NE Japan. The major observation thatmay lead to this conclusion is that the tholeiitic basaltmagmas have higher 87Sr/86 Sr than the calc-alkalic basal-

tic magmas (0 7042 vs 0 7038) and may thus be derivedfrom a more radiogenic crustal source rather than the lessradiogenic upper mantle. The characteristic trace elementsignatures of the tholeiitic magmas are also consistentwith the presence of plagioclase and amphibole as meltingresidues, suggesting that they are partial melts leavingbehind amphibolitic rather than peridotitic residues.

Origin of calc-alkalic magmas:differentiation of a hydrous basalt?Our experimental data confirm that one possible mecha-nism of calc-alkalic magma production in the IBM is dif-ferentiation of a basaltic magma in the presence of 2 5^3wt % H 2O. If this value is accepted as the H 2O con-tent of the parental basalt, then felsic to intermediatemagmas (67 5^56 6wt % SiO 2 on an H 2O-free basis),which are differentiated from the hydrous parental basaltvia 28^63% partial melting or 72^37% solidification(Fig. 8), should contain H 2O between 4 5 and 10 2wt %(Table 4). Because the H 2O solubility of felsic to intermedi-ate magmas at 0 3 GPa (i.e. at the depth of the middlecrust of the IBM) is 10 wt % (e.g. Zhang, 1999), theinferred hydrous andesitic melts are not oversaturatedwith H 2O. However, such large amounts of H 2O shouldcause H 2O saturation in the magmas during progressivecrystallization and may significantly affect the crystalliz-ing mineral assemblage and consequently have an effect

on the physical properties of the solidified rocks. It is there-fore interesting to examine the lithology of H 2O-rich inter-mediate plutonic rocks and to compare the seismicvelocity of such rocks with the observed seismic velocitystructure of the IBM crust, especially that of the middlecrust with intermediate composition.

Subsolidus phase equilibria for differentiated composi-tions at depths and temperatures relevant to the IBMmiddle crust (10 km and 410 8C) can be obtained using thefree energy minimization algorithm Perple _ X (Connolly,1990, 2005). The temperature profile used here is close tothe moderate temperature gradient across the IBM crust(Tatsumi et al ., 2008 a). f O 2 is fixed at QFM, as phase sta-

bility, and hence the physical properties, are relativelyinsensitive to f O 2 (Behn & Kelemen, 2006). The mineralassemblages obtained under these conditions are listed inTable 4. These mineral assemblages are then used to calcu-late the density ( r ), and P- and S-wave velocity, V p and V s,of the inferred lithologies (Table 4) following the methodof Hacker et al . (2003). The calculated V p (6 3^6 8 km/s) iswithin the range of the observed values (6 0^6 8 km/s), sug-gesting that the differentiation of a hydrous mafic

Fig. 6. Relationship between melt fraction, temperature, H 2O, andselected element abundances in the melt. For each H 2O content inthe starting material the melt fraction systematically increases as tem-perature increases. SiO 2 and FeO in the melt change significantlyeven at a constant melt fraction with changing H 2O content in thestarting material.


1593


20/29

component, either via crystallization differentiation of ahydrous magma or via anatexis of a hydrous gabbro oramphibolite, could be a possible mechanism for the forma-tion of the characteristic IBM middle crust.

An interesting test for this hydrous basalt hypothesiswould be to compare the temperature^composition rela-tionship obtained for the experimentally determined LLD(Fig. 11) with that of the natural calc-alkalic andesites.Detailed petrographic data including the compositions of

coexisting ortho- and clinopyroxenes in the IBM calc-alkalic volcanic rocks are available for basalts to andesitesfrom the Sumisu and Rota volcanoes, both lying astridethe IBM volcanic front (Shukuno et al ., 2006; Y. Tamura& H. Shukuno, unpublished data). It should be stressedthat all of these calc-alkalic rocks are characterized bythe presence of reversely zoned pyroxene phenocrysts.Temperatures obtained by two-pyroxene thermometryapplied to the rims of pyroxene phenocrysts are plotted

Fig. 7. Change in modal compositions of phases in the run products as a function of temperature and H 2O content.


1594


21/29

Fig. 8. Compositional variation of the quenched melt as a function of melt fraction. Filled and open stars indicate the compositions of the start-ing basalt and a rhyolitic melt with 75 wt % SiO 2 inferred by extrapolating the experimental results at 0 49wt % H 2O. Rhyolitic melt can beproduced by 6% partial melting or 94% crystallization of the starting basalt.


1595


22/29

Fig. 9. SiO 2 variation diagrams for quenched melts with different H 2O contents. Calc-alkalic (CA) and tholeiitic (TH) trends observed forIBM rocks (dashed and continuous thick grey lines, respectively) are broadly consistent with liquid lines of descent for higher and lower H 2O.Alternatively, IBM CA trends can be explained by mixing of mafic (filled star) and felsic (open star) melts. The effect of f O 2 on melt composi-tions deduced from the experimental results of Sisson et al . (2005) is schematically shown by arrows (see Fig. 4 and text).


1596


23/29

against the bulk-rock SiO 2 contents in Fig. 11. Because themelt (groundmass), which is likely to have equilibratedwith the pyroxene rims, should have an SiO 2 contenthigher than the bulk-rock, the temperature of naturallyoccurring calc-alkalic melts at a constant SiO 2 content

tends to be higher than that of the experimentally inferredmelt in the presence of 2 5 w t % H 2O (Fig. 11).Although further observations on other volcanoes areneeded, it is suggested that calc-alkalic andesites, at leastin some IBM volcanoes, could not be produced by

Fig. 10. Experimental melt compositions projected onto an FeO ^MgO^Na 2O K 2O diagram. Representative calc-alkalic (CA) and tholeiitic(TH) trends for IBM volcanic rocks and the boundary between the TH and CA series after Kuno (1968) are also shown.

Fig. 11. SiO2 vs temperature relationship along LLDs in the presence of 0 49^2 83wt % H 2O.Temperatures inferred from two-pyroxene ther-

mometry for tholeiitic (TH) and calc-alkalic (CA) are also shown.Temperatures of tholeiitic dacites are consistent with an LLD in the presenceof 0 5 wt % H 2O, whereas those of calc-alkalic rocks are higher than those for an H 2O-rich ( 2 5 wt %) LLD that shows a differentiationtrend similar to the IBM calc-alkalic trend.


1597


24/29

crystallization differentiation of hydrous (H 2O 4 2 5 wt%) basaltic magmas.

Origin of calc-alkalic magmas: mixing of mafic and felsic magmas?An alternative mechanism for the production of calc-alka-lic andesite magmas is mixing between mafic and felsicmagmas. This idea is often proposed to explain the chemi-cal and petrographical characteristics of calc-alkalic rocks(e.g. Eichelberger, 1975; Sakuyama, 1979, 1981; Clynne,1999; Dungan & Davidson, 2004; Tatsumi et al ., 2008b),

although the physical process of magma mixing continuesto be a matter of debate (e.g.Wiebe et al ., 2004). One chem-

ical characteristic that may suggest magma mixingis straight-line differentiation trends for calc-alkalicrocks, and concave or convex trends for tholeiitic rocks,as a result of the fractionation of solid phases with continu-ous compositional changes. The calc-alkalic rocks of the IBM exhibit straight-line differentiation trends(Figs 3 and 9), and thus it is possible that they can beexplained simply by mixing of mafic and felsic end-member magmas.

Table 4: Compositions, subsolidus mineral assemblages, and physical properties for intermediate IBM magmas

Differentiation Mixing

Basalt 27 8 33 1 43 5 63 1 Basalt 2:1 1:1 1:2 Rhyolite

SiO 2 48 00 60 62 57 96 54 74 52 97 49 15 55 69 58 95 62 22 68 75

TiO 2 0 68 0 32 0 44 0 76 0 75 0 70 0 59 0 53 0 48 0 37

Al2O3 17 82 15 14 15 19 16 00 16 70 18 24 16 29 15 32 14 34 12 39

FeO 10 71 2 98 3 75 5 08 6 46 10 97 7 99 6 51 5 02 2 05

MgO 6 23 2 25 3 20 3 69 5 32 6 38 4 46 3 50 2 54 0 63

CaO 12 07 5 84 6 80 7 68 9 12 12 36 9 81 8 53 7 25 4 69

Na 2O 1 56 2 43 2 31 2 51 2 08 1 60 1 61 1 62 1 62 1 64

K2O 0 10 0 42 0 34 0 34 0 25 0 10 0 45 0 63 0 80 1 16

H2O 2 83 10 18 8 55 6 51 4 48 0 50 3 11 4 42 5 72 8 33

Total 100 100 00 100 00 100 00 100 00 100 00 100 00 100 00 100 00 100 00

SiO 2 49 40 67 35 64 40 60 28 56 55 49 40 57 47 61 67 65 99 75 00

TiO 2 0 70 0 36 0 49 0 84 0 80 0 70 0 61 0 56 0 51 0 40

Al2O3 18 34 16 82 16 88 17 62 17 83 18 34 16 82 16 03 15 21 13 52

FeO 11 02 3 31 4 17 5 59 6 90 11 02 8 25 6 81 5 32 2 23

MgO 6 41 2 50 3 56 4 07 5 68 6 41 4 61 3 67 2 70 0 68

CaO 12 42 6 49 7 56 8 46 9 74 12 42 10 12 8 92 7 69 5 12

Na 2O 1 60 2 69 2 57 2 77 2 22 1 60 1 66 1 69 1 72 1 78

K2O 0 10 0 47 0 38 0 37 0 27 0 10 0 47 0 66 0 85 1 26

Total 100 00 100 00 100 00 100 00 100 00 100 00 100 00 100 00 100 00 100 00

Depth (km) 10 10 10 10 10 10 10

T (8C) 410 410 410 410 410 410 410

hornblende 21 1 29 1 36 4 42 3 43 3 36 5 27 7

clinochlore 0 0 0 0 0 0 3 8 1 7 0 3 0 0

quartz 42 4 39 4 32 8 28 9 33 1 39 1 44 8

muscovite 4 0 3 3 3 3 2 4 4 3 5 8 7 5

paragonite 13 0 12 7 12 9 8 3 1 4 4 2 5 2zoisite 9 3 10 1 9 5 14 4 16 2 14 1 11 5

plagioclase 10 2 5 5 5 1 0 0 0 0 0 0 3 2

H2O (wt %) y 1 48 1 65 1 83 2 31 1 88 1 69 1 50

r (g/cm 3) 2 88 2 91 2 95 3 00 3 02 2 97 2 91

V p (km/s) 6 28 6 32 6 39 6 57 6 61 6 46 6 34

V s (km/s) 3 81 3 84 3 84 3 92 3 97 3 92 3 88

Melt fraction (%).yH2O in hydrous phases.


1598


25/29

The petrographic characteristics that are common tocalc-alkalic rocks but not observed in tholeiitic rocks arethe following disequilibrium features (e.g. Sakuyama,1979, 1981; Wada, 1981; Fujinawa, 1988, 1990: Clynne, 1999;Tatsumi et al ., 2008b):

(1) the presence of reversely zoned pyroxene phenocrystswith a lower Mg-number [ 100 Mg/(Mg Fe)]core surrounded by a higher Mg-number rim;

(2) the presence of groundmass pyroxenes with a higherMg-number, which yield higher equilibration tem-peratures than the phenocrysts;

(3) a bimodal distribution in the core compositions of pla-gioclase phenocrysts;

(4) disequilibrium phenocryst assemblages such as Mg-rich olivine and quartz;

(5) patchy groundmass with different colors and/oramount of mafic minerals.

These observations can be also explained by mixing or

mingling of a mafic, high- T magma with a felsic, low- T magma. The mafic magma contains higher Mg/Fe pyrox-enes and higher Ca/Na plagioclases and olivine, whereasthe felsic magma may contain quartz.

If we accept magma mixing as a possible mechanismresponsible for the generation of calc-alkalic rocks, thecompositional range of calc-alkalic volcanic and plutonicrocks from the IBM suggests mixing between a basalticmagma containing 50wt % SiO 2 and a rhyoliticmagma containing 75wt % SiO 2 (Figs 3, 8 and 9). Thebasalt used as a starting material for the experiments inthis study, which has a composition representative of IBMarc basalts and may be produced by fractional crystalliza-

tion of a mantle-derived primary basalt magma, seems tobe a reasonable candidate for the basaltic end-member(Figs 3 and 9). Although our experiments do not yieldfelsic melts with 75% SiO 2, such a felsic composition canbe achieved by extrapolating the relationship betweenmelt fraction and composition (Fig. 8). In this estimation,we tentatively assume that such a felsic melt is a tholeiiticdifferentiation product of an H 2O-poor ( 0 5w t %)basalt. If so, then the composition of a felsic melt with75% SiO 2 can be produced by either 6% partial meltingof a basaltic rock or 94% crystallization of a basalticmagma (Table 4). It should be stressed that the composi-tions of the calc-alkalic intermediate rocks, which mayform the characteristic IBM middle crust layer, plot closeto the mixing line between the inferred basaltic and rhyoli-tic magmas (Figs 3 and 9).

The origin of this felsic end-member magma is still opento debate. One possible mechanism includes self mixing(Couch et al ., 2001) or internal mixing, in which felsicand mafic magmas are essentially co-magmatic; that is, afelsic magma is derived from a basaltic magma via frac-tional crystallization. Alternatively, a felsic end-member

magma could form by anatexis of pre-existing arc crustand mix with a mantle-derived basaltic magma (e.g.Tatsumi et al ., 2008 b), which may be considered as externalmixing. When mantle-derived basaltic magmas are under-plated and/or intruded into the arc crust they transferheat into the overlying and surrounding crust, which can

lead to partial melting of the wall-rocks (e.g. Hildreth,1981; Raia & Spera, 1997; Annen & Sparks, 2002).

Implications for the structure andevolution of the IBM crustThe IBM system is a suitable site for examining arcevolution and continental crust formation, because thestructure of the crust has been extensively surveyed seismi-cally, revealing a well-developed middle crust withV p 6 0^6 8 km/s (Suyehiro et al ., 1996; Takahashi et al .,1998, 2007, 2008; Kodaira et al ., 2007 a, 2007b), values iden-tical to the average V p of continental crust (e.g.Christensen & Mooney, 1995). To understand the processthat creates this distinctive crust and mantle structure, thelithology of the crust and mantle is inferred from magmageneration and differentiation models, for which seismicvelocities are calculated and compared with the observedseismic structure. Tatsumi et al . (2008 a) examined whethera petrological model including mixing of mantle-derivedbasaltic magma with felsic magma could lead to the gene-sis of IBM crust and mantle, and found that the lithologypredicted by the model has a V p consistent with theobserved values (Fig. 2). On the other hand, the new exper-imental results presented here, including the melting^ crystallization regime for a representative IBM basaltcomposition and variable H 2O contents in the magmas,

may provide better constraints on the composition of theIBM crust. Thus, it may be interesting to examine theimplications these results have for the lithology and thephysical properties of the IBM crust.

The basalt end-member is here assumed to be the start-ing material used in this study with 0 5 w t % H 2O,whereas the rhyolite end-member is an experimentallyinferred felsic melt. Accepting either 94% crystallizationof basaltic magma or 6% partial melting of basaltic crust,then the felsic magma could contain 8 3 w t % H 2O.The compositions of the intermediate magmas are thenobtained by mixing the basalt and rhyolite end-membersin proportions of 2:1, 1:1, and 1:2 (Table 4). The mineral

assemblages and V p for these compositions at 10 km and4108C are then calculated using the methods describedabove (Table 4), suggesting that the calculated V p(6 3^6 6 km/s) is close to that observed for the IBMmiddle crust (6 0^6 8 km/s). The major difference betweenthis calculation and that of Tatsumi et al . (2008 a) is theH 2O content of the mixed intermediate magmas, 3 1^5 7and 0 3 wt %, respectively. We consider that the H 2O con-tents used in modelling in this study are more realistic


1599


26/29

that those of Tatsumi et al . (2008 a). Although the above dif-ference results in the difference in the amount of amphibole( 20% vs 36% in H 2O-poorer and -richer rocks, respec-tively), the calculated V p is rather constant at 6 5 km/s.

Petrological modeling by Tatsumi et al . (2008 a) assumed10% partial melting of a basaltic composition for the pro-duction of the felsic magma that mixes with the basalticmagma to form the IBM middle crust with an intermedi-ate composition, and suggested, based on this meltingregime, that the restite after extraction of 10% felsic meltmay be a likely lithology for the characteristic low- V upper-most mantle immediately below the Moho (Fig. 2). Thepresent experiments, on the other hand, suggest 6% par-tial melting for production of a felsic magma. As listed inTable 5, however, the restite composition based onthe experimental results is very similar to that of Tatsumi et al . (2008 a), suggesting that the mechanismproposed by Tatsumi et al . (2008 a), including transforma-tion of a crustal component (i.e. restite) across the trans-parent and permeable Moho, could still be valid forinterpretation of the characteristic crust^mantle structureof the IBM.

C O N C L U S I O N S

It is generally accepted that the calc-alkalic trend can bereproduced by crystallization differentiation of a hydrousbasalt magma (e.g. Sisson & Grove, 1993; Kawamoto,1996; Grove et al ., 2003; Pichavant & Macdonald, 2007;Hamada & Fujii, 2008). Our experimental results confirmthat a parental basaltic magma with 2 5 wt % H 2O dif-ferentiates to produce calc-alkalic liquids. However, wehesitate to generalize this mechanism as the single processfor the genesis of the IBM calc-alkalic andesites, because

the temperatures of the natural andesitic magma tend tobe higher than those of the experimentally obtained calc-alkalic LLD. Instead, production of a felsic melt either bypartial melting of rather H 2O-poor ( 5 0 5 wt %) basalticor amphibolitic crust or by crystallization differentiationof a similar basaltic magma, and subsequent mixing withbasaltic magma, may reasonably account for both thecalc-alkalic trend of the IBM plutonic or volcanic rocksand the disequilibrium petrographic characteristics gener-ally observed in calc-alkalic volcanic rocks.

The V p calculated for an inferred solidified mixedmagma is consistent with the observed value for the char-acteristic IBM middle crust. Furthermore, the restite afterseparation of a felsic melt has a composition similar tothat proposed by Tatsumi et al . (2008 a) and exhibits V pvalues similar to the low- V uppermost mantle beneath theIBM. It may be thus speculated that a juvenile arc crustwith a basaltic composition could evolve into a more differ-entiated, mature crust via transformation of mafic restitesfrom the crust to the upper mantle across a chemicallytransparent and permeable Moho.

A C K N O W L E D G E M E N T S

We thank Alex Nichols, Satoshi Okamura, and SatoruNakashima for analytical assistance, Yohsihiko Tamuraand Hiroshi Shukuno for providing the petrographicdata, and Miki Fukuda for preparing the manuscript andfigures. Constructive comments on the manuscript byAlex Nichols, Trevor Green, two anonymous reviewers,and the editors John Gamble and Marjorie Wilson areappreciated. This work is partially supported by Grant-in-Aid for Creative Scientific Research (19GS0211).

Table 5: Compositions of inferred basaltic, rhyolitic, mixed andesitic magmas and restites

This study Tatsumi et al . (2008 a )

basalt rhyolite andesite 1 restite basalt rhyolite andesite restite

SiO 2 49 40 75 00 62 20 47 76 50 00 75 00 60 00 47 22

TiO 2 0 70 0 40 0 55 0 72 0 80 0 30 0 60 0 86

Al2O3 18 34 13 52 15 93 18 64 19 10 14 00 17 06 19 67

FeO 11 02 2 23 6 63 11 58 10 20 2 00 6 92 11 11

MgO 6 41 0 68 3 55 6 78 6 00 0 20 3 68 6 64

CaO 12 42 5 12 8 77 12 89 12 10 3 00 8 46 13 11

Na 2O 1 60 1 78 1 69 1 59 1 60 4 50 2 76 1 28

K2O 0 10 1 26 0 68 0 03 0 20 1 00 0 52 0 11

Total 100 00 100 00 100 00 100 00 100 00 100 00 100 00 100 00

H2O 0 50 8 33 4 42 0 00 0 10 0 30 0 00

11:1 mixture of basalt and rhyolite.


1600


27/29

R E F E R E N C E SAlbare 'de, F. (1998). The growth of continental crust. In: Vauchez, A.

& Meissner, R. O. (eds) Continents and their Mantle Roots.Amsterdam: Elsevier, pp.1^14.

Annen, C. & Sparks, R. S. J. (2002). Effects of repetitive emplacementof basaltic intrusions on thermal evolution and melt generation inthe crust. Earth and Planetary Science Letters 203, 937^955.

Baker, M. C. W. (1974). Volcano spacing, fractures, and thickness of the lithosphere discussion. Earth and Planetary Science Letters 23,161 163.

Baker, M. B., Grove, T. L. & Price, R. (1994). Primitive basalts andandesites from the Mt. Shasta region, N. California: products of varying melt fraction and water content. Contributions to Mineralogyand Petrology 118, 111^215.

Ballhaus, C. G. (1993). Redox states of lithospheric and asthenosphericupper mantle. Contributions to Mineralogy and Petrology 114, 331^348.

Ballhaus, C. G., Berry, R. F. & Green, D. H. (1991). High pressureexperimental calibration of the olivine^orthopyroxene^spineloxygen geobarometer: implications for the oxidation state of theupper mantle. Contributions to Mineralogy and Petrology 107, 27^40.

Bau, M. & Knittel, U. (1993). Significance of slab-derived partialmelts and aqueous fluids for the genesis of tholeiitic and calc-alka-line island-arc basalts; evidence from Mt. Arayat, Philippines.Chemical Geology 105, 233^251.

Behn, M. D. & Kelemen, P. B. (2006). Stability of arc lower crust:Insights from the Talkeetna arc section, south central Alaska, andthe seismic structure of modern arcs. Journal of Geophysical Research111, doi:10.1029/2006JB004327.

Bibee, L. D., Shor, G. G., Jr, & Lu, R. S. (1980). Inter-arc spreading inthe Mariana Trough. Marine Geology 35, 183^197.

Bloomer, S. H., Taylor, B., MacLeod, C. J., Stern, R. J., Fryer, P.,Hawkins, J. W. & Johnson, L. E. (1995). Early arc volcanism andthe ophiolite problem; a perspective from drilling in the westernPacific. In: Taylor, B. & Natland, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, AmericanGeophysical Union 88, 1^30.

Carmichael, I. (1991). The redox states of basic and silicic magmas: a

reflection of their source regions? Contributions to Mineralogy and Petrology 106, 129^264.

Carmichael, I. & Ghiorso, M. S. (1990). The effect of oxygen fugacityon the redox state of natural liquids and their crystallizing phases.In: Nicholls, J. & Russell, J. K. (eds) Modern Methods of IgneousPetrology: Understanding Magmatic Processes. Mineralogical Society of America, Reviews in Mineralogy 24, 191^212.

Christensen, N. I. & Mooney, W. D. (1995). Seismic velocity structureand composition of the continental crust: a global view. Journal of Geophysical Research 100, 9761^9788.

Clynne, M. A. (1999). A complex magma mixing origin for rockserupted in 1915, Lassen Peak, California. Journal of Petrology 40,105 132.

Connolly, J. A. D. (1990). Multivariable phase-diagramsan algo-rithm based on generalized thermodynamics. American Journal of

Science 290, 666^718.Connolly, J. A. D. (2005). Computation of phase equilibria by linear

programming: A tool for geodynamic modeling and its applicationto subduction zone decarbonation. Earth and Planetary Science Letters236, 524^541.

Couch, S., Sparks, R. S. J. & Carroll, M. R. (2001). Mineral disequili-brium in lavas explained by convective self-mixing in openmagma chambers. Nature 411, 1037^1039.

Deschamps, A., Okino, K. & Fujioka, K. (2002). Late amagmaticextension along the central and eastern segments of the West

Philippine Basin fossil spreading axis. Earth and Planetary ScienceLetters 203, 277^293.

Dufek, J. & Bergantz, G. W. (2005). Lower crustal magma genes is andpreservation: a stochastic framework for the evaluation of basalt^ crust interaction. Journal of Petrology 46, 2167 2195.

Dungan, M. A. & Davidson, J. (2004). Partial assimilative recycling of the mafic plutonic roots of arc volcanoes: An example from the

Chilean Andes. Geology 32, 773^776.Eichelberger, J. C. (1975). Origin of andesite and dacite: evidenceof mixing at Glass Mountain in California and at otherCircum-Pacific volcanoes. Geological Society of America Bulletin 86,1381 1391.

Ewart, A. (1982). Petrogenesis of the Tertiary anorogenic volcanicseries of southern Queensland, Australia, in the light of trace ele-ment geochemistry and O, Sr and Pb isotopes. Journal of Petrology23, 344^382.

Fornelli, A., Piccarreta, G., Del Moro, A. & Acquafredda, P. (2002).Multi-stage melting in the lower crust of the Serre (SouthernItaly). Journal of Petrology 43, 2191 2217.

Fujinawa, A. (1988). Tholeiitic and calc-alkaline magma series atAdatara Volcano, Northeast Japan: 1. Geochemical constraints ontheir origin. Lithos 22, 135^158.

Fujinawa, A. (1990). Tholeiitic and calc-alkaline magma series atAdatara Volcano, Northeast Japan: 2. Mineralogy and phase rela-tions. Lithos 24, 217^236.

Gaetani, G. A., Grove, T. L. & Bryan, W. B. (1993). The influence of water on the petrogenesis of subduction-related igneous rocks.Nature 365, 332^334.

Gill, J. B. (1981). Orogenic Andesites and PlateTectonics. Berlin: Springer.Green, T. H. (1982). Anatexis of mafic crust and high pressure crystal-

lization of andesite. In: Thorpe, R. S. (ed.) Andesites: OrogenicAndesites and Related Rocks. Chichester: John Wiley, pp. 465^487.

Green, T. H. & Ringwood, A. E. (1967). Crystal lization of basalt andandesite under high pressure hydrous conditions. Earth and Planetary Science Letters 3, 481^489.

Grove, T. L., Elkins-Tanton, L. T., Parman, S. W., Chatterjee, N.,Mu ntener, O. & Gaetani, G. A. (2003). Fractional crystallization

and mantle-melting controls on calc-alkaline differentiationtrends. Contributions to Mineralogy and Petrology 145, 515 533.Hacker, B. R., Abers, G. A. & Peacock, S. M. (2003). Subduction fac-

tory;1,Theoretical mineralogy, densities, seismic wave speeds, andH 2O contents. Journal of Geophysical Research 108(B1), doi:10.1029/2001JB001127.

Hall, C. E., Gurnis, M., Sdrolias, M., Lavier, L. L. & Mu ller, R. D.(2003). Catastrophic initiation of subduction following forced con-vergence across fracture zones. Earth and Planetary Science Letters212, 15^30.

Hamada, M. & Fujii, N. (2007). H 2O-rich island arc low-K tholeiitemagma inferred from Ca-rich plagioclase^melt inclusion equili-bria. GeochemicalJournal 41, 437^461.

Hamada, M. & Fujii, T. (2008). Experimental constraints on theeffects of pressure and H 2O on the fractional crystallization of high-Mg island arc basalt. Contributions to Mineralogy and Petrology155, 767^790.

Hawkesworth, C. J. & Kemp, A. I. S. (2006). Evolution of the conti-nental crust. Nature 443, 811 817.

Hildreth, W. (1981). Gradients in silicic magma chambers: implicationsfor lithospheric magmatism. Journal of Geophysical Research 86,10153 10192.

Hildreth, W. & Moorbath, S. (1988). Crustal contr ibutions to arc mag-matism in the Andes of central Chile. Contributions to Mineralogy and Petrology 98, 455^489.


1601


28/29

Hunter, A. G. (1998). Intracrustal controls on the coexistence of tho-leiitic and calc-alkaline magma series at Aso Volcano, SW Japan. Journal of Petrology 39, 1255^1284.

Ishizuka, O., Kimura, J.-I., Li, Y. B., Stern, R. J., Reagan, M. K.,Taylor, R. N., Ohara,Y., Bloomer, S. H., Ishi i, T., Hargrove, U. S.,III & Haraguchi, S. (2006). Early stages in the evolution of Izu^ Bonin arc volcanism: new age, chemical, and isotopic constraints.

Earth and Planetary Science Letters 250, 385^401.Itoh, Y. (1988). Differential rotation of the eastern part of southwest

Japan inferred from paleomagnetism of Cretaceous and Neogenerocks. Journal of Geophysical Research 93, 3401^3411.

Itoh, Y. & Ito, Y. (1989). Confined ductile deformation in the Japan arcinferred from paleomagnetic studies. Tectonophysics 167, 57^73.

Kawamoto,T. (1996). Experimental constraints on dif ferentiation andH 2O abundance of calc-alkaline magmas. Earth and PlanetaryScience Letters 144, 577^589.

Kawano,Y., Yagi, K. & Aoki, K. (1961). Petrography and petrochemis-try of the volcanic rocks of Quaternary volcanoes of northeastern Japan. Science Reports,Tohoku University, Series 3: Mineralogy, Petrology,and Economic Geology 7, 1^46.

Kawate, S. & Arima, M. (1998). Petrogenesis of the Tanzawa plutoniccomplex, central Japan: Exposed felsic middle crust of the Izu^

Bonin^Mariana arc. Island Arc 7, 342^358.Kay, S. M. & Kay, R. W. (1985). Role of crystal cumulates and theoceanic crust in the formation of the lower crust of the Aleutianarc. Geology 13, 461^464.

Kobayashi, K., Kasuga, S. & Okino, K. (1995). Shikoku Basin and ItsMargins. In: Taylor, B. (ed.) Backarc Basins: Tectonics and Magmatism . NewYork: Plenum, pp. 381^405.

Kodaira, S., Sato,T., Takahashi, N., Ito, A., Tamura,Y., Tatsumi, Y. &Kaneda, Y. (2007 a). Seismological evidences for variation of conti-nental growth along the Izu intra-oceanic arc and its implicationfor arc volcanism. Journal of Geophysical Research 112, doi:10.1029/2006JB004593.

Kodaira, S., Sato,T., Takahashi, N., Miura, S., Tamura,Y., Tatsumi,Y.& Kaneda, Y. (2007 b). New seismological constraints on growth of continental crust in the Izu^Bonin intra-oceanic arc. Geology 35,

1031 1034.Kuno, H. (1968). Differentiation of basalt magmas. In: Hess, H. H. &Poldervaart, A. (eds) BasaltsThe Poldervaart Treatise on Rocks of Basaltic Composition,Vol. 2 . NewYork: Interscience, pp. 623^688.

Miyashiro, A. (1971). Processes and patterns of metamorphism in aplate tectonic framework. Geological Society of America, Abstracts withPrograms 3, 648.

Nakajima, K. & Arima, M. (1998). Melting experiments on hydrouslow-K tholeiite: implications for the genesis of tonalitic crust inthe Izu^Bonin^Mariana arc. Island Arc 7, 359^373.

Okino, K.,

J. Petrology 2009 Tatsumi 1575 603

Documents

Transcript of J. Petrology 2009 Tatsumi 1575 603