Journal of Petrology Advance Access originally published online on November 23, 2006
Journal of Petrology 2007 48(2):395-433; doi:10.1093/petrology/egl065
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Melting and Multi-stage Metasomatism in the Mantle Wedge beneath a Frontal Arc Inferred from Highly Depleted Peridotite Xenoliths from the Avacha Volcano, Southern Kamchatka
1Department of Earth Sciences, Graduate School of Natural Science and Technology, Kanazawa University, Kakuma, Kanazawa 920-1192, Japan
2Department of Earth Sciences, Faculty of Science, Kanazawa University, Kakuma, Kanazawa 920-1192, Japan
3Department of Physical–Chemical Methods of Research and Mineralogy, Institute of Volcanology, Russian Academy of Science, 683006, Petropavlovsk-Kamchatky, Russia
RECEIVED JUNE 15, 2006; ACCEPTED OCTOBER 9, 2006
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONPeridotite xenoliths from the Avacha volcano, Kamchatka, derived from the mantle beneath the volcanic front of the Kamchatka arc, are mainly highly depleted clinopyroxene-poor harzburgites with highly forsteritic olivine (Fo 90–92) and high Cr-number spinel (0·5–0·7). The Avacha peridotites have experienced metasomatism to various extents, with the formation of metasomatic orthopyroxene replacing primary olivine, by infiltration of SiO2-rich fluids. The metasomatic orthopyroxenes can be subdivided into two textural types; (1) radially aggregated prismatic grains (opx II-1); (2) stout grains associated with interstitial glass and metasomatic minerals (opx II-2). The Avacha peridotites exhibit light rare earth element (LREE) enrichment relative to heavy REE (HREE), even in primary lithologies with <2 vol.% of metasomatic orthopyroxene. However, the metasomatism has not altered the chemical characteristics of the Avacha peridotites significantly, and the Fo contents of olivine do not increase with an increase in the Cr-number of spinel. The clinopyroxene in the most depleted primary peridotite, with the highest spinel Cr-number, has the highest CeN (where subscript N indicates normalized to primitive mantle) and relatively low YbN contents. We conclude that the Avacha peridotites are residues of high-degree melting induced by infiltration of an aqueous fluid with relatively high Fe/Mg and LREE/HREE ratios. In situ analysis of olivine grain boundaries by laser-ablation inductively coupled plasma mass spectrometry (ICP-MS) showed significant concentrations of some trace elements (Ce and Ba), even on glass-free boundaries; such grain boundaries can be a repository for large-ion lithophile elements (LILE). Orthopyroxene and Ca-amphibole contribute to the HREE and middle REE (MREE) budget, respectively, of the bulk peridotites. Three types of mantle metasomatism are recorded in the Avacha peridotites. The agent that formed HREE-poor opx II-1 was a slab-derived aqueous fluid rich in SiO2. It was a successor of the fluid influx involved in the initial partial melting event. On the other hand, the metasomatic minerals in highly metasomatized peridotites have high LILE (Th, U and Sr) and LREE contents relative to HREE, indicating involvement of a hydrous melt of adakitic affinity derived from the subducted slab. Glasses associated with opx II-2 are similar in their trace-element characteristics to the host Avacha volcanic rocks, indicating the likelihood that the metasomatic agent was a forerunner of the recent Avacha arc magmatism.
KEY WORDS: Avacha volcano; southern Kamchatka arc; depleted peridotite xenoliths; hydrous melting; mantle metasomatism
|
INTRODUCTION |
|---|
Sub-arc mantle peridotites are expected to be more depleted in magmatic components, on average, than peridotites from other tectonic settings, especially mid-ocean ridges; this is reflected in the strongly depleted nature of some arc magmas compared with mid-ocean ridge basalts (MORB) (e.g. Dick & Bullen, 1984
; Arai, 1994a). The nature of the depleted sub-arc mantle beneath present-day arcs has, however, not been discussed in detail because mantle peridotite xenoliths are much rarer than in other tectonic settings (e.g. Nixon, 1987). Depleted peridotites have been sampled from fore-arc regions (e.g. Parkinson & Pearce, 1998), but typically have been substantially altered at low temperatures.
Peridotite xenoliths from Avacha volcano in the Kamchatka arc, Russia, are mostly harzburgites, highly depleted in magmatic components (i.e. major-element depleted). They have not been subjected to low-temperature alteration and have experienced variable extents of metasomatism by silica-rich fluids and/or melts (Arai et al., 2003a). They provide us with many pieces of information, not only about the genesis of the highly depleted harzburgite protoliths, but also about metasomatism within the supra-subduction zone mantle wedge. The peridotite xenoliths from Avacha represent fragments of the upper mantle beneath an active volcanic front, which are comparatively rare (see Arai et al., 2004).
Metasomatized peridotite xenoliths (mostly harzburgites) from arc settings have been reported from several localities: Cascades, USA (e.g. Ertan & Leeman, 1996), Colorado Plateau, USA (e.g. Smith et al., 1999), Patagonia (e.g. Gorring & Kay, 2000; Kilian & Stern, 2002), Batan Island, Philippines (e.g. Schiano et al., 1995; Arai & Kida, 2000; Arai et al., 2004), Papua New Guinea (e.g. McInnes et al., 2001; Grégoire et al., 2001; Franz et al., 2002), Tallante, SE Spain (Arai et al., 2003b; Beccaluva et al., 2004; Shimizu et al., 2004) and northern Kamchatka (e.g. Kepezhinskas et al., 1995, 1996). The geochemical characteristics of the depleted harzburgites, which are metasomatized to variable degrees, can provide important constraints on mantle-wedge processes (i.e. the depletion of magmatic components by partial melting and re-fertilization by metasomatism).
Salters & Shimizu (1988) suggested that the bulk-rock trace-element signatures of fertile mantle peridotites are controlled by the clinopyroxene trace-element budget; however, this is not the case for more depleted peridotites (e.g. Maury et al., 1992; Grégoire et al., 2001; Franz et al., 2002; Kilian & Stern, 2002). It is well known that ‘grain-boundary components’ in peridotites are rich in incompatible trace elements, such as large ion lithophile elements (LILE) (Zindler & Jagoutz, 1988). Recently, the importance of the grain-boundary components has been highlighted, and their chemical characteristics have been estimated in some peridotite xenoliths (e.g. Condie et al., 2004; Ionov, 2004). However, the methods used have typically been indirect, involving subtraction of calculated bulk-rock compositions, estimated based on modal mineralogy and mineral chemistry, from analyzed bulk-rock compositions. There have, however, been a few direct studies on grain boundaries (Suzuki, 1987; Hiraga et al., 2002, 2003, 2004). The Avacha peridotite xenoliths have advantages for grain-boundary research: the grain boundaries of the primary mantle minerals are devoid of low-temperature alteration and are completely preserved. This makes a good contrast to those fore-arc depleted peridotites (e.g. Parkinson & Pearce, 1998), in which serpentinization has destroyed the original grain boundaries between the mantle minerals. The grain-boundary components are relatively easily detected in the Avacha peridotites because of the low LILE contents in the surrounding mantle mineral grains. The importance of the grain boundary as a sink for LILE is highlighted in the Avacha peridotites because of the extremely small amounts of potentially LILE-rich minerals such as clinopyroxene and amphibole.
|
GEOLOGICAL BACKGROUND |
|---|
The Avacha (=Avachinsky) volcano is located in the southern part of the Kamchatka Peninsula, which is a relatively mature arc (Fig. 1). The Pacific plate is subducting at a relatively rapid rate (70–90 mm/year, Minster et al., 1974
) beneath the southern part of the Kamchatka Peninsula along the Kuril–Kamchatka trench. In the northern part of the Kamchatka Peninsula, young and hot oceanic lithosphere, produced in the Komandorsky Basin, was subducted until 
2 Ma (Hochstaedter et al., 1994). Peridotite xenoliths, which have experienced Na-metasomatism by a slab melt, have been reported from Nb-enriched host basalts in northern Kamchatka (e.g. Kepezhinskas et al., 1995). The dip angle of the subducting plate decreases from 55° to 35° from south to north in southern Kamchatka (Gorbatov et al., 1997), where there are three sub-parallel volcanic chains at distances from the trench of 200 km, 320 km and 400 km, respectively (Tatsumi et al., 1994; Fig. 1). The Avacha volcano is in the frontal chain that forms the volcanic front (VF); the depth to slab here is about 120 km (Gorbatov et al., 1997) and the depth to the Moho is about 37 km (Levin et al., 2002).
|
Avacha is a stratovolcano, rising
2741 m above the sea level. Its volcanic activity began in the late Pleistocene, and is divided into two stages, IAv and IIAv, based on the chemical composition of its ejecta (Braitseva et al., 1998). Andesitic pyroclastic flows and tephra are characteristic of IAv, from 7250 to 3700 years bp, whereas basaltic andesite lavas have been produced in IIAv, from 3500 years bp to the present (Braitseva et al., 1998). All the effusive rocks are calc-alkaline in chemistry; some contain megacrysts of hornblende (Braitseva et al., 1998). Abundant peridotite xenoliths are enclosed in some of the andesitic pyroclastic deposits of IAv (Braitseva et al., 1998) as ejecta occasionally coated by thin lava films. |
SAMPLE DESCRIPTIONS |
|---|
Host rocks of the xenoliths
The host rock of the peridotite xenoliths is a basaltic andesite, containing phenocrysts and microphenocrysts of olivine, plagioclase, amphibole, clinopyroxene and orthopyroxene. Olivine is euhedral to rounded and sometimes includes tiny crystals of chromian spinel. Many of the plagioclase phenocrysts contain abundant melt (glass) inclusions and are strongly zoned. Orthopyroxene exhibits weak pleochroism, and amphibole is euhedral and has strong pleochroism from brown to yellowish brown. The groundmass has an intergranular texture.
Peridotite xenoliths
Most of the xenoliths from Avacha are spinel harzburgites; subordinate pyroxenites (clinopyroxenite and orthopyroxenite), dunite and hornblende-gabbros are also found. Clinopyroxene-bearing hornblendite sometimes coats the peridotite xenoliths, forming a thin selvage as well as cutting them as veinlets (Fig. 2b). Such characteristics have been observed in other sub-arc mantle xenoliths exhumed by arc magmas; for example, those from Iraya volcano, Philippines (Arai et al., 1996) and Oshima-Ôshima volcano, northern Japan (Ninomiya & Arai, 1992). Klügel (1998) found similar hornblendite selvages on peridotite xenoliths from the Canary Islands (oceanic hotspot), and interpreted them as a reaction product between the host magma and mantle peridotite at shallow depths within the lithosphere.
|
The xenoliths are subangular to angular in shape and predominantly 5–6 cm (up to 40 cm) across (Fig. 2). The spinel harzburgites occasionally have thin bands (<2 cm in thickness) of orthopyroxenite (Fig. 2a) to websterite. The grain size of the peridotite xenoliths is variable; some of them are very fine-grained with olivine of less than 1 mm across (down to less than 0·1 mm) (Arai et al., 2003a
). The coarse-grained peridotite xenoliths (with olivine >1–2 mm to 10 mm) are the main focus of this study and are subsequently referred to simply as peridotite xenoliths. The fine-grained peridotites are highly heterogeneous and with complex textures and mineral chemistry, and will be discussed elsewhere. The modal proportions of minerals in the peridotites were determined by point-counting based on 2000–4000 points per thin section (Table 1; Fig. 3). The peridotites are sometimes heterogeneous, and the modal proportion of orthopyroxene in standard size thin sections (48 mmx28 mm) is highly variable.
|
|
The peridotite xenoliths are modally metasomatized to various extents, irrespective of their textures, resulting in the formation of metasomatic orthopyroxene, usually replacing olivine (Arai et al., 2003a
; Fig. 4b, d–g). There is no systematic distribution of the metasomatic orthopyroxene; this occurs mainly as radial aggregates in olivine-dominant parts. All peridotites contain the metasomatic orthopyroxene. In the subsequent discussion, the terms ‘primary peridotite’ and ‘metasomatized peridotite’ are used to refer to peridotites that contain metasomatic orthopyroxene replacing olivine by <2·0 vol.% and 
2·0 vol.%, respectively.
|
The primary peridotites from Avacha are refractory harzburgites that contain only a small amount of primary clinopyroxene (<2 vol.%) and 10–27 vol.% (17 vol.% on average) of primary orthopyroxene (Table 1; Fig. 3). They usually have protogranular to porphyroclastic textures; the grain size of the olivine and orthopyroxene is up to 1 cm. The olivine and orthopyroxene frequently contain lamellae of chromian spinel and clinopyroxene, respectively. Many of the orthopyroxene grains are kinked and have thin lamellae of clinopyroxene (Fig. 4a). Discrete clinopyroxene grains are subhedral to anhedral, and smaller than the olivine and orthopyroxene. Chromian spinel is opaque, subhedral to anhedral and mostly small in size (
1 mm). Almost all the peridotites contain discrete amphibole, which may be a primary phase (Fig. 4h). Tiny grains (10 µm) of monosulfide solid solutions (MSS) are occasionally observed, especially in metasomatized peridotites. Two types of textures are exhibited by olivine-replacing metasomatic orthopyroxene: (1) replacing olivine with ragged boundaries, and sometimes forming radial aggregates (opx II-1) (Fig. 4b); (2) forming stout euhedral to subhedral grains with interstitial amphibole, clinopyroxene and/or highly silicic glass (opx II-2) (Fig. 4e–g). The metasomatized peridotites are often cross-cut by metasomatic orthopyroxenite veinlets, composed of opx II-1 with clinopyroxene, amphibole and pentlandite. Stout grains of opx II-2 sometimes form pockets with interstitial silicic glass, amphibole and clinopyroxene. In addition to the olivine-replacing orthopyroxene, the metasomatized peridotites contain another type of metasomatic orthopyroxene (‘recrystallized orthopyroxene’), formed by recrystallization of primary orthopyroxene porphyroclasts, especially along grain margins or within cracks, into smaller grains. The recrystallization is associated with coarsening of clinopyroxene lamellae (Fig. 4c). Clinopyroxenes and amphiboles closely associated with the metasomatic orthopyroxenes are also of metasomatic origin, having been precipitated from, or modified by, metasomatic fluids or melts. The metasomatic orthopyroxenes are totally free of clinopyroxene exsolution and deformation. The total proportion of orthopyroxene is about 25 vol.% on average in the metasomatized peridotites, significantly higher than that in the primary peridotites (17 vol.% on average; Fig. 3). The metasomatic clinopyroxene and amphibole proportions are negligible in the metasomatized peridotites except in sample 629 (Fig. 3; Table. 1). Silicic glasses are not only accompanied by opx II-2, but are also observed as rare interstitial films along grain boundaries or as inclusions embedded in primary spinel and olivine grains. One of the highly metasomatized peridotites samples (sample 629; Table 1) is especially rich in glass (amount not determined) both as inclusions in spinel and as a film associated with opx II-2. We have found only one grain of phlogopite, exhibiting strong pleochroism and included in olivine in a metasomatized peridotite. Fluid inclusion trails were commonly found in both the primary and metasomatized peridotites (Fig. 4h).
Within the Avacha peridotite xenolith suite, sample 227 is an atypical, highly metasomatized harzburgite, containing a relatively thick orthopyroxenite vein, 2 mm thick, which is considered in detail below. The orthopyroxene vein is composed of radial aggregates of opx II-1 and metasomatic clinopyroxene and amphibole (Fig. 4d). The modal amount of the metasomatic orthopyroxene attains 20%. Pentlandite is also observed and prevalent throughout the peridotite, especially around the vein. Primary orthopyroxene porphyroclasts far from the veins have recrystallized rims and sometimes have coarsened clinopyroxene lamellae and subhedral to anhedral amphibole (Fig. 4c). Minute fluid inclusions are prevalent in the olivine and other minerals.
|
ANALYTICAL METHODS |
|---|
Bulk-rock analysis
Samples for bulk-rock chemical analysis were cut using a diamond saw and washed in an ultrasonic bath to avoid contamination. Chips were selected to be representative of the samples concerned, and any hornblendite selvages and dikelets were carefully removed to allow geochemical characterization free from the effects of the hornblendite. After drying the sample chips for almost 1 day, they were crushed with an iron hammer and ground in an agate mortar. Major-element contents were determined by X-ray fluorescence (XRF) spectroscopy (X-ray spectrometer; System-3270, Rigaku) at Kanazawa University on glass discs made from 0·5 g (±0·5%) powder.
Trace-element contents in bulk-rocks were determined by inductively coupled plasma mass spectrometry (ICP-MS; X7, Thermo Electron) at Kanazawa University. De-ionized water purified by a milli-Q Element A-10 system (Millipore) was used throughout the procedure, and all acids used (HNO3, HClO4, HF) were of Ultrapure grade (Kantokagaku, Japan). Rock powders of 100 mg were weighed into screw-cap Savillex® Teflon® beakers, and 1·0 ml of HNO3, 1· 0 ml of HClO4 and 1· 5 ml of HF were added to each sample. We made three solution samples for each sample to check random errors, and standard deviations (1
) for each analysis are shown in Tables 2 and 3; only analyses with low RSD% (<5%) are presented. The samples were carefully heated to around 110°C (±10°C) for 3–7 days after being agitated in an ultrasonic bath for 10 min. After dissolution, the solutions were diluted with water to 100 µg/g using 100 ml Teflon® PFA volumetric flasks (VIT-LAB). Some spinel grains were insoluble and were removed before analysis, as spinel has a negligible effect on the trace-element budget of the peridotites (e.g. Stosch & Seck, 1980; Bedini & Bodinier, 1999). For quantitative analysis, mixed standard solutions were prepared from XSTC-1, -7, -8 and -13 (SPEX), and 1 ng/g 115In was used for signal-drift correction during each analytical run. Geological reference materials for peridotite (JP-1) and basalt (JB-2) were also analyzed to monitor analytical quality. Determination limits for most elements are 0· 1–100 ppb and 0· 1–10 ppb for volcanic rock and peridotite analysis, respectively. Detailed operational settings for ICP-MS analysis of bulk peridotite samples are shown in Appendix A (see Shirasaka et al., 2004).
|
|
In situ analysis
The major-element compositions of minerals and glasses were determined by electron microprobe (JXA8800, JEOL) at the Center for Co-operative Research of Kanazawa University. Analytical conditions were 20 kV accelerating voltage, 20 nA probe current and 3 µm probe diameter for minerals, and 15 kV accelerating voltage, 10 nA probe current and 3 µm probe diameter for glasses. The analytical conditions for glasses were set to minimize the contamination from surrounding minerals and the loss of alkalis. Ferrous and ferric iron contents in chromian spinel were calculated based on spinel stoichiometry.
Trace-element abundances in the major phases in peridotites (olivine, orthopyroxene, clinopyroxene, amphibole and glass) were determined using a laser ablation system (GeoLas Q-Plus, MicroLas) coupled to an ICP-MS system (Agilent 7500s, Yokogawa Analytical Systems). The laser-spot diameter was 50 µm for clinopyroxenes, amphiboles and glasses, and 100 µm for orthopyroxenes, fluid or melt inclusions and their host olivines. The energy of laser ablation was 5 Hz and 8 J/cm2. The NIST 612 standard glass was used for calibration with Si as an internal standard. We used the Si contents of the host olivine for the calibration of concentrations in the inclusions. Details of the analytical procedures have been described by Ishida et al. (2004) and Morishita et al. (2005). In addition, we also attempted to estimate the trace-element concentrations within olivine grain boundaries in the peridotite xenoliths, as described below in detail. The energy of laser ablation for grain-boundary analysis was 12 Hz and 8 J/cm2 and scan speed was 5 µm/s.
|
BULK-ROCK CHEMISTRY |
|---|
Avacha volcanics
The bulk-rock compositions of two samples of volcanic rocks (samples 578 and 580) obtained as fragments of ejecta in the pyroclastic deposits that contain the studied peridotite xenoliths are listed in Table 2. They are a basaltic andesite and an andesite (SiO2 53· 5 and 59· 5 wt %, on an anhydrous basis) with low K2O contents (0· 62 and 0· 69 wt %). The total alkali contents and the FeO* (=total iron as FeO) to MgO ratio are low, 3· 91 wt % and 1· 13 for sample 578, and 4· 31 wt % and 2· 29 for sample 580, indicating a calc-alkaline affinity. The host rocks in direct contact with the xenoliths are too small in amount for bulk-rock analysis, but are petrographically similar to the basaltic andesite (sample 578) analyzed here.
The rare earth element (REE) and other trace-element compositions of the Avacha volcanic rocks are also listed in Table 2. Primitive mantle-normalized REE patterns are somewhat light REE (LREE)-enriched, with (La/Yb)N (where subscript N indicates normalized to primitive mantle) of around 1· 8 (Fig. 5). The Avacha volcanics are depleted extensively in Nb and Ti relative to other incompatible elements (Fig. 5). The two Avacha volcanic samples show almost the same trace-element pattern (Fig. 5).
|
Peridotite xenoliths
Bulk-rock compositions were determined for 11 peridotite xenoliths from Avacha (Tables 1 and 3). Contents of Al2O3, CaO and SiO2 show negative correlations with MgO (Fig. 6). The MgO contents are high (>43 wt %), and the CaO and Al2O3 contents are low (< 1 wt %), indicating the highly depleted character of the peridotites (Fig. 6), even for the metasomatized peridotites. It should be noted that the metasomatic mineral contents are <6 vol. % in the analyzed metasomatized peridotites, except sample 629 (Table 1), and therefore their presence does not alter the primary bulk-rock major-element compositions significantly. Sample 629 contains appreciably higher amounts of metasomatic orthopyroxene and Ca-amphibole than the other samples (Table 1). The most highly metasomatized peridotite xenoliths, with more than 30 vol.% total orthopyroxene (Fig. 3), are generally small in size and unsuitable for bulk-rock analysis. The SiO2 contents are slightly higher in the Avacha peridotites than in typical abyssal peridotites at a given MgO content (Fig. 6). There are no obvious correlations between the amount of silicic glasses and major-element concentrations in the peridotites, except for sample 629, which contains large amounts of glass and has different chemistry from the other samples (Tables 1 and 3; Fig. 6).
|
Primitive mantle-normalized trace-element patterns of the peridotites are shown in Fig. 7. Except for the highly metasomatized harzburgite (sample 629) (Table 1), the REE abundances are roughly one-tenth or less of the primitive mantle values (McDonough & Sun, 1995
). All of the samples have similarly concave-upward REE patterns with a marked positive slope from Er to Lu, irrespective of the presence or absence of silicic glasses and metasomatic orthopyroxene (Fig. 7). Sample 629, which is rich in glass and metasomatic orthopyroxene, has abundances of some elements 10 times higher than the other peridotites (Fig. 7). There are no obvious correlations between the REEN abundances and modal amounts of clinopyroxene (Fig. 8e–h). Almost half the peridotites are enriched in Sr relative to neighboring elements (Fig. 7b and d). Some LILE, for example, Ba, U, Th and Sr, are detectable but very low in abundance in the peridotites. All the Avacha peridotites exhibit similar REE patterns, which are almost flat or weakly concave downward, similar to those of harzburgites from various localities (e.g. McDonough & Frey, 1989; Takazawa et al., 2000). Niu (2004) reported almost flat REE distribution patterns for abyssal peridotites, in which clinopyroxenes have LREE-depleted REE patterns.
|
|
|
MINERAL AND GLASS CHEMISTRY OF THE PERIDOTITE XENOLITHS |
|---|
Major-element compositions
Representative mineral and glass compositions are listed in Tables 4–6
. Olivines have slightly variable Fo contents from one sample to another (90· 5–92· 0), but show no obvious intra-grain chemical zoning irrespective of the degree of metasomatism. Chromian spinels have high Cr-number [=Cr/(Al+Cr) atomic ratio], from 0· 5 to 0· 7, and low Fe3+/(Al+Cr+Fe3+) atomic ratios (<0· 1). All of the peridotites from Avacha plot within the olivine–spinel mantle array (OSMA), defining a spinel peridotite restite trend (Arai, 1994b); however, their Fo contents do not show an appreciable increase, and may even show a slightly decrease with an increase of spinel Cr-number (Fig. 9). Moreover, the SiO2 contents in the bulk-rock peridotites have no correlation with the Cr-number of chromian spinel, although they show a weak negative correlation with the Fo contents of olivine (Fig. 8a and b). There are no obvious correlations between the modal amounts of opx II, Cr-number of spinel and Fo contents of olivine (Fig. 8c and d), although the formation of metasomatic orthopyroxene is the most prominent modal change induced by the metasomatism.
|
|
|
|
All the clinopyroxenes in the peridotite xenoliths are chromian diopsides with around 0· 5 wt % Cr2O3 and high Mg-number [=Mg/(Mg+total Fe) atomic ratio] varying from 0· 92 to 0· 94 (Table 4). There are no systematic chemical differences between the primary and metasomatic clinopyroxenes within the peridotites. The TiO2 contents of the clinopyroxenes are extremely low (< 0· 10 wt %) and the contents of Al2O3 and Na2O are also low (1· 0–3· 0 wt % and < 0· 5 wt %, respectively) (Table 4).
Orthopyroxenes of both primary and metasomatic origin have a relatively narrow range of Mg-number (0· 91–0· 93) (Table 4). All orthopyroxenes have low Al2O3, Cr2O3 and CaO contents, and opx II-1 has especially low Cr2O3 contents (Table 4, Fig. 10). No obvious correlations are evident between the Mg-number and Al2O3, Cr2O3 and CaO contents. It is noteworthy that opx II-2 is included in the compositional spread of the primary orthopyroxenes (Fig. 10).
|
All the amphiboles analyzed belong to the calcic amphibole group (Leake et al., 1997
), but are variable in chemistry from pargasite through edenite (primary amphibole) to magnesio-hornblende (metasomatic amphibole) (Leake et al., 1997). We refer to them collectively as hornblende. They have relatively high Mg-number (0· 89–0· 91) and Cr2O3 contents (up to 2· 3 wt %). The Mg-number of the hornblendes correlates positively with their TiO2 and Al2O3 contents, and negatively with the Cr2O3 contents. The primary hornblendes are slightly higher in Al2O3 and TiO2, on average, than the metasomatic ones (Table 4).
The phlogopite in sample 375, included in a primary olivine (Fo90· 8), has a low TiO2 content (<0· 1 wt %) and relatively high Mg-number (c. 0· 90). The Na/(Na+K) ratio of this phlogopite is relatively high (c. 0· 23) and similar to those of phlogopites in spinel peridotites (Arai, 1986). The number of Si atoms recalculated on O=22 (anhydrous basis) is around 5· 5, indicating a significant amount of the eastonite molecule, and falls within the range of phlogopite in spinel peridotites (Arai, 1984a).
The sulfide found in the metasomatized peridotites is MSS with an Fe:Ni atomic ratio of around 2:1.
Silicate glasses, both as glass inclusions and interstitial glasses to opx II-2, have variable major-element concentrations with high SiO2 (56–73 wt % on anhydrous basis) and Al2O3 (14–21 wt %) contents (Table 5). The glass inclusions in primary spinels have lower oxide totals (80–95 wt %) in microprobe analyses than the interstitial glasses (around 100 wt %) (Table 5). This result means that the former may include higher amounts of volatile components, such as H2O. A preliminary Fourier transform infrared (FT-IR) analysis of the glass inclusions within spinels shows a broad peak around 3400 cm–1, resulting from the H2O molecule. The glass inclusions have higher SiO2 and Al2O3 and lower MgO and FeO* contents than the interstitial glasses (Table 5), exhibiting high normative quartz (up to 45 wt %) and feldspar (55–70 wt %) contents, with or without normative corundum (0–7 wt %).
Effects of the formation of hornblendite selvages
The host magma, which precipitated hornblende around the peridotite xenoliths (Fig. 2b), could have modified their chemical characteristics. To evaluate the chemical effects of the formation of the hornblendite selvage, we examined a typical hornblendite-coated peridotite xenolith, sample 181, in detail. Olivines have a low Fo content (Fo79· 6) in contact with the hornblendite selvage, but rapidly become more magnesian (Fo91· 2), in the range of primary mantle olivines (Fo90· 5–92· 0), 5 mm away from the hornblendite selvage (Table 6). The rim of a chromian spinel grain adjoining the hornblendite has been changed to magnetite with appreciable amounts of Cr2O3 and Al2O3 (9· 8 wt % and 11· 7 wt %, respectively), whereas the core is unaltered. Chromian spinels are also totally intact within the interior more than 5 mm inward from the hornblendite. Clinopyroxenes in and adjacent to the hornblendite selvage have high TiO2 and Al2O3 contents (up to 0· 5 wt % and 4· 5 wt %, respectively), and low Mg-number (=0· 80). The hornblendes (pargasites; Leake et al., 1997), forming the hornblendite selvages or veins, have low Mg-number (0· 70–0· 77) and Cr2O3 contents (up to 0· 2 wt %), and are rich in TiO2 (up to 2· 0 wt %) (Table 6).
Trace-element compositions
Trace-element compositions of representative minerals and glasses are listed in Table 7, and their primitive mantle-normalized REE and incompatible trace-element patterns are shown in Figs 11–15.
|
|
|
|
|
|
Clinopyroxene
Almost all the clinopyroxenes in the peridotites have similar heavy REE (HREE) abundances although the LREE to middle REE (MREE) patterns exhibit sample-by-sample variations (Fig. 11a and b). They show clear depletion in Zr and Ti relative to adjacent elements in the primary peridotites, and there are no such spikes in clinopyroxenes from the metasomatized peridotites (Fig. 11a and b). The discrete clinopyroxenes both in the primary and metasomatized peridotites have LREE-depleted, HREE-enriched patterns, and MREE are below detection limits in most of them (Fig. 11a and b).
Metasomatic clinopyroxenes within a metasomatically formed orthopyroxenite vein in the highly metasomatized peridotite (sample 227) are extremely enriched in LREE, LILE, Zr and Hf relative to other clinopyroxenes (Figs 11 and 12a). Discrete or coarsened lamellar clinopyroxenes in sample 227, located far from the orthopyroxene vein, have slightly lower concentrations than those in the orthopyroxene vein, although these are higher than the primary clinopyroxenes from the other samples (Figs 11 and 12a).
Orthopyroxene
REE concentrations in orthopyroxenes are low, and only La, Ce and the HREE were above detection limits (Fig. 13). The trace-element patterns have gentler slopes from MREE to HREE for opx II-2 than for the primary orthopyroxenes (Fig. 13a and c). On the other hand, almost all the REE are below detection limits in opx II-1 (Fig. 13b). HREE abundances are almost the same for all orthopyroxenes, although MREE contents are slightly variable. LREE concentrations are very different between the metasomatic orthopyroxenes, especially those in sample 227, and the primary orthopyroxene (Figs 12b and 13). Both orthopyroxenes, primary and metasomatic, observed in sample 227 exhibit especially high concentrations of LILE (Ba, Th and U), Zr and Hf relative to primary orthopyroxenes in the other samples (Fig. 12b). The trace-element patterns are variable for the metasomatic orthopyroxenes, which do not always have high trace-element concentrations (Figs 12b and 13). Compared with the other samples, the trace-element concentrations of orthopyroxene of sample 227 are as distinctive as its clinopyroxene chemical characteristics.
Hornblende
Primitive mantle-normalized trace-element patterns of hornblendes are shown in Figs 12c and 14. They are almost the same as those of clinopyroxene (Figs 11 and 12). The hornblendes have basically the same trace-element patterns but with different abundances. The primary hornblende has distinctly lower (LREE/HREE)N ratios than the selvage hornblende, although the patterns are very similar (Fig. 14). The metasomatic hornblende in the orthopyroxenite vein of sample 227 is rich in LILE, LREE, Zr and Hf and depleted in MREE (Fig. 12c). It is strikingly lower in HREE concentrations than the other hornblendes, which have patterns with gentle negative slopes from MREE to LREE and flat floors from MREE to HREE (Figs 12c and 14).
Inclusions
We examined the trace-element abundances of minute melt or fluid inclusions, which are embedded in olivine porphyroclasts (see Fig. 4h). We analyzed the inclusion-rich parts of olivine and inclusion-free parts as their background (Fig. 15a). Although the nominal trace-element abundances were far lower than the real ones in the inclusions, because of dilution by the background olivine, we found enrichment of Ba, Ce, Sr, Zr, Ti and Y in the inclusions relative to olivine (Fig. 15b).
Glass
The silicic glasses interstitial to opx II-2 (Fig. 4e–g) have trace-element patterns with LILE and LREE enrichment (Fig. 15c). The (La/Yb)N ratios are not so high (2· 16–3· 88), and the patterns are notably similar to those of the Avacha volcanic rocks (Fig. 5).
Mineral chemical correlations
The Cr-number values of spinels show positive correlations with the CeN and SmN contents of clinopyroxenes in the primary peridotites (Fig. 16a and c), and a weak negative correlation with YbN (Fig. 16e). On the other hand, there are no obvious correlations between the REEN of primary clinopyroxenes and the Fo content of olivines (Fig. 16b, d and f). As shown in Fig. 17, the Avacha peridotites also show high concentrations of Ce in clinopyroxene relative to the trend of abyssal peridotite data defined by Hellebrand et al. (2001).
|
|
HREEN contents of orthopyroxene are positively correlated with the Fo content of olivine in the primary peridotites, with and without glasses, but have negative correlations with the Cr-number of spinel (Fig. 16g–l).
Minerals in the hornblendite selvage
Trace-element concentrations of minerals within the hornblendite are different from those of the peridotite minerals (Figs 11c and 14). The primitive mantle-normalized trace-element pattern of the clinopyroxene within the hornblendite selvage is slightly LREE-depleted and flat from MREE to HREE (Fig. 11c). The selvage clinopyroxene has a similar trace-element pattern to some of the primary clinopyroxenes, but the abundances are very different, especially for HREE (3–4 times higher in the former) (Fig. 11c). This is very different from the metasomatic clinopyroxene in the highly metasomatized peridotite (sample 227) (Fig. 12c). Pargasite in the hornblendite selvage has almost the same trace-element pattern as the primary hornblende in the peridotites, although the trace-element abundances are different, especially for HREE (Fig. 14).
Thermobarometry
The two-pyroxene thermometer of Wells (1977) yields similar ranges of equilibrium temperature for the primary peridotites (900–1046°C) and the metasomatized peridotites (890–1100°C) (see Appendix B). We also calculated equilibrium temperatures using the geothermometers of Ballhaus et al. (1991) and Nimis & Taylor (2000), and the values do not show great differences (see Appendix B). Equilibrium pressures were not calculated because of the lack of appropriate barometers. Arai et al. (2003a) favoured conditions in the spinel lherzolite stability field for a highly metasomatized fine-grained peridotite (not discussed in this study) that contains a low-Cr-number (< 0· 2) spinel. We found composite xenoliths of fine-grained and coarse-grained peridotites (Ishimaru & Arai, in preparation), indicating that the coarse-grained peridotites, discussed here, also originated from the spinel lherzolite field.
The redox state of the Avacha peridotites was calculated using the oxygen geobarometer of Ballhaus et al. (1990, 1991). The calculated oxygen fugacity assuming P=1· 5 GPa is high in both primary peridotites [
log(f O2)FMQ=–0· 2 to +1· 2], and metasomatized peridotites [log(f O2)FMQ=0 to +1· 5]. The Avacha primary peridotites are distinctly more oxidized than abyssal peridotites [log(f O2)FMQ=–2· 5 to +0· 5; Bryndzia & Wood, 1990], and are similar in redox state to other sub-arc mantle peridotite xenoliths (Brandon & Draper, 1996; Blatter & Carmichael, 1998; Arai et al., 2006).
|
GRAIN-BOUNDARY CHEMISTRY |
|---|
We examined in situ incompatible trace-element concentrations along olivine grain boundaries of the primary peridotites by LA-ICP-MS. As described by Suzuki (1987
) and Hiraga et al. (2002, 2003, 2004), incompatible trace elements are frequently concentrated on grain boundaries. We made chemical profiles across and along grain boundaries (Fig. 18a) to detect in situ concentrations, if any, relative to background olivine. We designed four line traverses to enhance the sensitivity of the ICP-MS and sampling precision of the grain-boundary components; one was just along the grain boundary, another was across the boundary, and the other two were on both sides of the olivine parallel to the grain boundary (see Fig. 18a). The energy of laser ablation for this analysis was 12 Hz and 8 J/cm2 and the scan speed was 5 µm/s to avoid penetration of the laser beam through the thin section into the epoxy resin or bottom glass slide. We could recognize such penetration by monitoring elements such as Li and B, which are highly concentrated in the glass and resin, respectively. The analytical conditions were set so as to avoid the laser beam penetrating the thin section (<30 µm in thickness) to the resin and bottom glass slide. It took 26–30 s, equivalent to 130–150 shots, to reach the resin when the energy of laser ablation was 5 Hz and 8 J/cm2. We then took 120 shots (about 24 µm sampling depths) at each point in our grain-boundary line analysis with high resolution. The length of each traverse depends on the texture, but was normally 200 µm. As we used a 50 µm laser beam, an analysis on each point (Fig. 18a) was representative of a circular area with a radius of 25 µm around the point. Another advantage of using such a wide beam is to collect signals even from grain boundaries oblique to the thin-section surface.
|
The Avacha peridotite xenoliths sometimes include silicic glass films along the grain boundaries; we analyzed both types of grain boundary, one with glass and the other without. We recognized high concentrations of LILE and LREE, and some HFSE (high field strength elements, e.g. Zr and Nb) relative to olivine at the glass-filled grain boundaries (Fig. 18b), which are probably a contribution from the glass. On the other hand, the traverses across the grain boundaries without visible glasses do not show obviously high concentrations of such elements, but show slightly higher counts in Ce and probably Y than the background olivine (Fig. 18c). The traverses along the glass-free grain boundaries exhibit Ce enrichment relative to the background olivine (Fig. 18e). It is noteworthy that the trace-element concentrations are not homogeneous along the grain boundaries with or without glass (Fig. 18d and e). The glass-free grain boundaries have lower concentrations of incompatible trace elements than the glass-filled ones, but nevertheless have appreciable amounts of Ce relative to the background olivine. Some incompatible elements such as K (Fig. 18b–e) are detectable only on the glass-filled grain boundaries, although some incompatible trace elements (e.g. Ce) are detectable on both types of grain boundaries (Fig. 18).
|
DISCUSSION |
|---|
As shown above (Figs 3, 6 and 7) and discussed by Arai et al. (2003a)
, the Avacha peridotite xenoliths are highly depleted in magmatic components and are modally metasomatized. Here, we discuss their geochemical characteristics in the context of processes of partial melting and metasomatism.
Significance of the grain-boundary components in depleted peridotites from Avacha
The high and low modal amounts of olivine and clinopyroxene, respectively (Fig. 3), and low abundances of HREE (Fig. 7) indicate that the peridotite xenoliths from Avacha have distinctly depleted characteristics. The factors that govern the trace-element characteristics of clinopyroxene-poor harzburgites have not been well constrained in the literature. Contributions of chromian spinel to the incompatible trace-element budget of bulk-rock peridotites are negligible (e.g. Stosch & Seck, 1980; Bedini & Bodinier, 1999). However, Bodinier et al. (1996) claimed that very thin layers of titanium oxides and/or phlogopite exist around spinels, and are capable of storing incompatible trace elements, especially HFSE (Nb, Ta, Zr and Hf). The trace-element compositions of spinel peridotites can be controlled by the following accessory phases in addition to the main minerals: (1) accessory minerals associated with metasomatism, such as amphibole, phlogopite, apatite, and zircon (e.g. Bedini & Bodinier, 1999); (2) fluid inclusions in minerals (e.g. Schiano et al., 1995); (3) the grain boundaries (e.g. Eggins et al., 1998; Condie et al., 2004; Ionov, 2004).
Trace-element patterns of bulk-rock Avacha peridotites, with low modal amounts of clinopyroxene, differ from those of clinopyroxene (Figs 5 and 11), although those of lherzolitic peridotites mimic their clinopyroxenes (Salters & Shimizu, 1988). We calculated the contributions of each mineral phase to the bulk-rock trace-element concentrations for two samples of peridotite (samples 106 and 200), based on the modal proportions and mineral compositions (Fig. 19a and b). Sample 106 contains glasses especially as inclusions in chromian spinel, whereas sample 200 is free from visible glasses. Bulk-rock HREE abundances are dominated by the contribution from pyroxenes, especially orthopyroxene. Figure 19 indicates that there should be some storage of LILE and LREE in phases other than pyroxenes and hornblende although the bulk-rock MREE and HREE concentrations can be largely accounted for by these phases. Ratios of LILE and LREE relative to HREE in glasses are not so high (Fig. 15c), and the glasses could not be a perfect repository for LILE and LREE because of a large deficiency of LILE and LREE relative to HREE in the bulk peridotites (Fig. 19). The glasses, however, control the abundances and patterns of trace elements in the Avacha peridotites to various degrees; for example, the sample 629 peridotite, which is exceptionally rich in glasses, has the highest trace-element contents (Fig. 7).
|
Suzuki (1987
) and Hiraga et al. (2002, 2003, 2004) confirmed that the grain boundaries in mantle peridotites show high concentrations of incompatible trace elements relative to background minerals. We detected the in situ enrichment of the incompatible trace elements Ce, Ba and Y along glass-free grain boundaries (e.g. Fig. 18c), consistent with the observations of Hiraga et al. (2004), who noted that elements with large ionic radii tend to concentrate along grain boundaries. The grain boundaries may be suitable repositories for LILE and LREE, complementary to the mineral and glass phases (Fig. 19). We detected almost exclusively Ba (LILE) and Ce (LREE) along the glass-free grain boundaries. The grain-boundary components as detected in this study are responsible for the metasomatism that produces no visible metasomatic minerals (e.g. Frey & Green, 1974). Minute glass or fluid inclusions in olivine may also contribute to the trace-element budget if they are appreciable in volume.
Partial melting
In general, the Fo and NiO contents of olivine and the Cr-number of chromian spinel increase with increasing degree of partial melting in residual spinel peridotites. Incompatible trace-element concentrations in bulk-rock peridotites and the modal amounts of clinopyroxene also decrease simultaneously. The Avacha peridotite xenoliths are characterized by low modal amounts of clinopyroxene (Table 1; Fig. 3), high Cr-number of spinel (Table 4; Fig. 9) and low concentrations of magmatic components and incompatible trace elements (Tables 1 and 3; Figs 6 and 7), indicating their highly depleted residual character. They are more depleted in magmatic components than abyssal peridotites on average (Figs 6 and 9).
The variations in the Fo content of olivine and Cr-number of spinel in the Avacha peridotites are, however, not concordant with the ordinary residual mantle trend (=OSMA); that is, the Fo contents of olivine do not increase with increasing Cr-number in the spinel (Fig. 9). The Fo contents of olivine are lower at a given Cr-number of spinel in the Avacha peridotites than in peridotite xenoliths from Iraya volcano, the Luzon Arc, Philippines (Arai et al., 2004), which have the characteristics expected of simple restites (e.g. Jaques & Green, 1980) (Fig. 9). There is no correlation between the modal amounts of opx II, Fo contents of coexisting olivine and Cr-number of coexisting spinel (Fig. 8c and d). This indicates that the formation of opx II did not give rise to the high Cr-number of spinel and relatively low Fo contents of olivine. The correlations between the Cr-number of spinel and REEN concentrations in the clinopyroxene of primary peridotites (Fig. 16a, c and e) further constrain their characteristics. For example, the clinopyroxene in sample 729, which shows the highest Cr-number of spinel, has the highest CeN and relatively low YbN contents (Fig. 16). It should be noted that HREEN abundances in orthopyroxenes of the primary peridotites also show negative correlations with the Cr-number of coexisting spinel and positive correlations with the Fo content of coexisting olivine (Fig. 16i–l). Simple extraction of melts cannot produce these chemical characteristics in residual minerals, and consequently influx melting (i.e. melting combined with melt or fluid influx) is the most probable process for the formation of the Avacha primary peridotites. The fluid or melt flux that triggered melting would have been rich in Fe relative to Mg because the Fo contents of primary peridotite olivine decrease with an increase in Cr-number of spinel (i.e. with an increase in the degree of melting). Hirose & Kawamoto (1995) and Matsukage & Kubo (2003) suggested that peridotites with high-Cr-number spinel are residues of high-degree partial melting induced either by an influx of H2O-rich melt (or fluid) or by extremely high temperatures in the case of dry conditions. The former condition is much more likely in the mantle wedge above the subduction zone. The presence of hornblende suggests the hydrous nature of the fluid influx. This was probably an aqueous fluid rich in LREE relative to HREE (Bizimis et al., 2000), in addition to being rich in Fe relative to Mg. The high LREE/HREE ratios of the primary clinopyroxenes relative to clinopyroxenes from abyssal peridotites (e.g. Johnson et al., 1990) are consistent with the involvement of a LREE-rich fluid in partial melting. This melting assisted by a silica-rich fluid is one of the possible explanations for the silica enrichment of the Avacha peridotite relative to abyssal peridotites at a given MgO (Fig. 6d). The Cr-number of spinel shows a weak negative correlation with the HREEN of orthopyroxene (Fig. 16g, i and k). The Fo contents of olivine, on the other hand, have positive correlations with the HREEN of orthopyroxene (Fig. 16h, j and l). In general, HREE have lower incompatibility and lower mobility during alteration than LREE, and therefore, the HREE abundances in clinopyroxene are a good indicator of the degree of melting in abyssal peridotites (Hellebrand et al., 2001). Figure 16 indicates that the HREE contents of orthopyroxene may have the potential to indicate the degree of melting even in clinopyroxene-poor or -free peridotites.
The hydrous fluid involved is inferred to be derived from the subducting oceanic slab because of its high concentrations of incompatible trace elements (see Tatsumi & Kogiso, 1997; Bizimis et al., 2000). It may have become further enriched in incompatible trace elements through percolation through the peridotite of the mantle wedge (e.g. Ionov et al., 2002). The fluid most probably percolated through the residual peridotite just after the initial partial melting in the same sense as envisaged by Niu (2004) for the refertilization of abyssal peridotites.
Metasomatism
Aqueous fluids or H2O-rich melts are important metasomatic agents in the mantle wedge above subduction zones, and can react with peridotite to produce metasomatic orthopyroxene, as described by, for example, Schiano et al. (1995), Smith et al. (1999), Arai & Kida (2000) and Grégoire et al. (2001). In the interpretations of those workers most of the fluid was ultimately derived from a subducting slab and enriched the mantle-wedge peridotite in silica (see Nakamura & Kushiro, 1974). Adakite melts, formed by partial melting of subducted young and/or hot oceanic crust, have been also considered as one of the metasomatic agents in the mantle wedge (Kepezhinskas et al., 1995; Kilian & Stern, 2002). Evidence of carbonatite metasomatism has been reported commonly from peridotite xenoliths exhumed by intra-plate basalts (e.g. Yaxley & Kamenetsky, 1999; Neumann et al., 2002) and rarely from the mantle of an arc tectonic setting (e.g. Laurora et al., 2001). Metasomatized peridotite xenoliths are common in intra-plate tectonic settings, such as the Canary Islands (e.g. Neumann & Wulff-Pedersen, 1997; Neumann et al., 2002), Tanzania (e.g. Jones et al., 1983; Dawson, 2002) and SE Australia (e.g. Yaxley et al., 1997; Varela et al., 1999; Yaxley & Kamenetsky, 1999). In general, the metasomatic agents are rich in incompatible trace elements, especially the LREE.
The modal mineralogical changes attributed to the effects of metasomatism in the Avacha peridotites are the replacement of olivine by orthopyroxene and the formation of hornblendite selvages. The hornblendite may have been precipitated from the host andesite or basaltic andesite magma. This was the latest metasomatic event that occurred en route to the surface, because the selvage is sometimes very thin and completely coats angular or subangular xenoliths (Fig. 2b) with very local chemical effects. We do not consider that hornblendite formation was related to the formation of metasomatic orthopyroxene at the expense of olivine; there are no examples of olivine-replacing metasomatic orthopyroxene associated with hornblendite. Instead, the orthopyroxene in the peridotites is replaced by clinopyroxene and hornblende at the contact with the hornblendite. The calculated trace-element pattern of clinopyroxene in equilibrium with the host magma, using the partition coefficients of Fujimaki et al. (1984), is almost identical to that of the clinopyroxene in the hornblendite selvage (Fig. 11c). It is distinctly different from that of the clinopyroxene in the thick (2 mm) metasomatic vein of opx II-1 in sample 227 (Figs 11 and 12a). Thus, we conclude that the formation of the hornblendite selvage was not related to the formation of metasomatic orthopyroxene within the mantle, but to the recent contact with the host magma that brought the xenoliths to the surface.
It is noteworthy that the metasomatic orthopyroxene (opx II-1) of sample 227 (Fig. 12b) has different trace-element characteristics both from opx II-2 and from opx II-1 in the other samples (Fig. 13). The latter two metasomatic orthopyroxenes are different from each other in chemistry (Figs 10 and 13). This suggests that at least three metasomatic agents with different chemical characteristics were involved in the formation of metasomatic orthopyroxenes in the Avacha peridotites.
Opx II-1, which exhibits complex fibrous textures, may reflect involvement of a low-viscosity metasomatic agent (i.e. an aqueous fluid). The fluid could infiltrate through minute cracks or fractures in the original olivine, resulting in the formation of metasomatic orthopyroxene (opx II-1) with fibrous textures (see Arai et al., 2004). Audétat & Keppler (2004) have shown that the viscosity of fluids is dramatically lowered (by 40 times) by a decrease in dissolved silicate from 40 to 30 wt %. Nakamura & Kushiro (1974) suggested the possibility that H2O as a metasomatic agent could react with olivine to precipitate metasomatic orthopyroxene. Bizimis et al. (2000) demonstrated that the incompatible trace-element compositions of slab-derived fluids are likely to be rich in LREE to MREE and Sr, and poor in HREE and Ti. Aqueous fluids are more enriched in Rb and Pb, and more depleted in Th relative to U than coexisting silicate melts (Keppler, 1996). They are more enriched in trace elements when coexisting with silicic melts than with basaltic melts (Ayers & Eggler, 1995; Adam et al., 1997). However, the concentrations of all trace elements, except Rb and K, are distinctly lower in fluids than in coexisting silicate melts (Ayers & Eggler, 1995; Adam et al., 1997). Adam et al. (1997) thus claimed that aqueous fluids are inappropriate metasomatic agents to cause enrichment of incompatible trace elements. The distinctly low HREE contents (mostly below detection limits) of opx II-1 relative to other orthopyroxenes (Fig. 13) possibly indicate their precipitation from a fluid. This is consistent with the preferential partitionings of HREE into silicate melts with respect to aqueous fluids by analogy with the HREE partitioning between clinopyroxene–melt–fluid (Bizimis et al., 2000). We therefore propose that an aqueous fluid with a small amount of dissolved silicate (SiO2 dominant) component was responsible for formation of opx II-1.
The fluid involved in the formation of opx II-1 was probably a successor of the fluid influx involved in the preceding partial melting event, which depleted the original peridotite protolith. This is consistent with the similarity of the trace-element patterns between the clinopyroxenes associated with opx II-1 (Fig. 11b) and the primary clinopyroxenes (Fig. 11a). The low HREE contents of the former are likely to reflect their precipitation from aqueous fluids (Adam et al., 1997). The apparent non-Fe enrichment of opx II-1 is due to lower fluid/rock ratios in metasomatism than in partial melting.
Opx II-2 was probably formed from a hydrous melt closely associated with the Avacha host magma, because the glass associated with opx II-2 has a similar trace-element pattern to the host magma (Figs 5 and 15b). A viscous melt could not invade fine cracks and fractures, and consequently replaced olivine with opx II-2 in simple morphologies (Fig. 4e; compare Fig. 4d). The metasomatic reaction probably occurred within the mantle before incorporation of the xenoliths in the host magma because, as stated above, there are no relationships between the formation of opx II-2 and the hornblendite selvages. The origin of the metasomatic melt is likely to be closely related to the genesis of the host magmas to the xenoliths, which is beyond the scope of this paper. The silicic melt prevalently invaded the peridotites along cracks and grain boundaries, leaving minute glass inclusions mostly in chromian spinels and less frequently in olivines. The enrichment of LILE and LREE in the silicic glass is consistent with the predicted chemical characteristics of aqueous fluids released from subducting slabs (e.g. Kogiso et al., 1997; Tatsumi & Hanyu, 2003). Such an aqueous fluid could have triggered partial melting of the mantle wedge to produce the relatively SiO2-rich primary magmas (e.g. Hirose, 1997) of the recent Avacha volcanism. These magmas migrated upward, percolating through previously depleted peridotite, and resulted in the precipitation of opx II-2 at the expense of olivine. The similarity of chemistry between the primary orthopyroxene and opx II-2 (Fig. 10) suggests that the migrating melt remained in equilibrium with the surrounding mantle peridotite.
The clinopyroxenes in the highly metasomatized peridotite sample 227 are characterized by relatively high incompatible trace-element concentrations with high LREE/HREE and LILE/HREE ratios (Fig. 12a; compare Fig. 11), suggesting involvement of a metasomatic melt rather than a fluid (e.g. Adam et al., 1997). It is noted that the pyroxenes in this sample share similar chemical characteristics irrespective of their mode of occurrence (Fig. 12), suggesting pervasive metasomatic alteration. As stated above, the metasomatic agent involved in the formation of sample 227 was different in chemistry from the fluid involved in opx II-1 formation in the other metasomatized peridotites. The melt involved may have been of adakitic affinity, judging from the weak positive Sr spike and high LILE/HREE abundances of the metasomatic clinopyroxenes (Fig. 12a), similar to the clinopyroxenes precipitated from adakitic magmas (Yogodzinski & Kelemen, 1998; Tsuchiya et al., 2005). The metasomatic clinopyroxenes in sample 227 are also similar in their trace-element characteristics to those of a highly metasomatized peridotite xenolith from Lihir volcano, Papua New Guinea, although the sample 227 clinopyroxenes have positive spikes at U, Th, Zr, Hf and Ti (Fig. 12a). The metasomatism observed in the Lihir peridotites has been attributed to a hydrous melt derived from a subducting slab (Grégoire et al., 2001; McInnes et al., 2001). We therefore conclude that one of the metasomatic events recorded in the Avacha peridotites was caused by invasion of a melt derived from a subducting slab, possibly of adakitic affinity.
Each metasomatic event may have added the grain-boundary components through infiltration of fluids or melts along the grain boundaries. The metasomatic fluids or melts appear to have been introduced through fractures (e.g. Fig. 4e), and it is most probable that grain-boundary percolation of melts or fluids enriched the grain boundaries soon after formation of the original residual harzburgite protolith as a residue of partial melting (see Niu, 2004). Elemental abundances in the grain-boundary components were variable because of chromatographic mass-dependent fractionation of the fluids or melts during percolation (see Navon & Stolper, 1987).
The temporal relationships between the different metasomatic events are difficult to estimate (Fig. 20). Following the initial partial melting event, the residual peridotites experienced metasomatism, which formed opx II-1 by infiltration of an aqueous fluid from the subducting slab (Fig. 20). Almost simultaneously, an adakitic melt was released from the slab and metasomatically formed the orthopyroxene-rich lithologies (e.g. sample 227) (Fig. 20). A forerunner of the recent Avacha volcanism was caused by melting of the mantle wedge peridotite triggered by a slab-derived fluid (=upwelling of a hydrous asthenospheric mantle diapir). This partial melt invaded the lithospheric mantle to precipitate opx II-2 by reaction with olivine (Fig. 20). Larger-degree partial melts produced the main Avacha magmas, which exhumed fragments of the overlying lithospheric mantle peridotite as xenoliths.
|
|
SUMMARY AND CONCLUSIONS |
|---|
- The Avacha peridotite xenoliths are highly depleted, with low modal amounts of clinopyroxene (< 1· 0 vol.% on average), high Cr-number (>0· 5) spinel, and highly forsteritic olivine (Fo90· 5–92). A slight decrease of the Fo contents of olivine with increasing Cr-number of spinel indicates addition of an FeO*-rich agent on partial melting (flux-induced melting). The influx also enriched the bulk-rock peridotites and constituent clinopyroxene in LREE.
- The HREE concentrations in orthopyroxene are shown to be a good indicator of the degree of depletion of peridotites for the highly depleted harzburgites.
- Orthopyroxene is largely responsible for the HREE characteristics, whereas hornblende and clinopyroxene, if any, contribute to the LREE and MREE budget of the bulk peridotites. The grain boundaries are a repository for LILE and LREE, and are responsible for the metasomatism without visible metasomatic minerals that has been documented from peridotite xenoliths. Grain-boundary glasses may significantly modify the trace-element characteristics of the bulk-rock peridotites from Avacha to various degrees depending on the amount present. The grain-boundary components were added by migrating fluids or melts subsequent to the formation of the residual harzburgite.
- The metasomatic agent was a silicate-bearing aqueous fluid for the generation of opx II-1, and a hydrous silicate melt for opx II-2; the melt was genetically related to the host Avacha magmas. The aqueous fluid was a successor of the initial influx, which triggered the partial melting and depletion of the original peridotite protolith of the harzburgites. These aqueous fluids are the likely products of dehydration reactions in the subducted slab. The agent involved in the metasomatism of sample 227 was a melt of adakitic affinity, with high LREE/HREE ratios and high concentrations of LILE (Ba, Th, U, Pb and Sr).
|
APPENDIX A |
|---|
|
|
APPENDIX B |
|---|
|
|
ACKNOWLEDGEMENTS |
|---|
The samples used in this paper were mainly collected by Dr K. Kadoshima and Mr A. Koyanagi during field research in 2000. The ICP-MS system was installed during the course of the 21st-Century COE project ‘Environmental Monitoring and Prediction of Long- and Short-Term Dynamics of Pan-Japan Sea Area: Construction of Monitoring Network and Assessment of Human Effects’ led by Professor K. Hayakawa, Kanazawa University. Professor K. Tazaki helped us to use the electron microprobe at the Center for Cooperative Research, Kanazawa University, and Drs T. Morishita, J. Uesugi and Y. Shimizu assisted us with microprobe analysis. Discussions with Professor A. Ishiwatari and Dr T. Morishita were helpful. The manuscript has benefited from constructive reviews by Dr L. Franz, Dr T. Hiraga and an anonymous referee. Comments by the editor, Professor M. Wilson, were much appreciated.
*Corresponding author. Tel: +81-76-264-6513, Fax: +81-76-264-6545. E-mail: jaja{at}earth.s.kanazawa-u.ac.jp
|
REFERENCES |
|---|
Adam J, Green TH, Sie SH, Ryan CG. (1997) Trace element partitioning between aqueous fluids, silicate melts and minerals. European Journal of Mineralogy 9:569–584.
Arai S. (1984a) Pressure–temperature dependent compositional variation of phlogopitic micas in upper mantle peridotites. Contributions to Mineralogy and Petrology 87:260–264.[CrossRef][Web of Science]
Arai S. (1984b) Igneous mineral equibria in some alpine-type peridotites in Japan. In Sunagawa I (Ed.). Minerals Science of the Earth(TERRAPUB, Tokyo) pp. 445–460.
Arai S. (1986) K/Na variation in phlogopite and amphibole of upper mantle peridotites due to fractionation of metasomatizing fluids. Journal of Geology 94:436–444.[Web of Science]
Arai S. (1994a) Compositional variation of olivine–chromian spinel in Mg-rich magmas as a guide to their residual spinel peridotites. Journal of Volcanology and Geothermal Research 9:279–293.
Arai S. (1994b) Characterization of spinel peridotites by olivine–spinel compositional relationships: review and interpretation. Chemical Geology 113:191–204.[CrossRef][Web of Science]
Arai S and Kida M. (2000) Origin of fine-grained peridotite xenoliths from Iraya volcano of Batan Island, Philippines: deserpentinization or metasomatism at the wedge mantle beneath an incipient arc? Island Arc 9:458–471.[CrossRef]
Arai S, Kida M, Abe N, Ninomiya A, Yumul GP Jr. (1996) Classification of peridotite xenoliths in calc-alkaline andesite from Iraya volcano, Batan Island, the Philippines, and its genetical implications. Science Reports of Kanazawa University 41:25–45.
Arai S, Ishimaru S, Okrugin VM. (2003a) Metasomatized harzburgite xenoliths from Avacha volcano as fragments of mantle wedge of the Kamchatka arc: implication for the metasomatic agent. Island Arc 12:233–246.[CrossRef]
Arai S, Shimizu Y, Gervilla F. (2003b) Quartz diorite veins in a peridotite xenolith from Tallante, Spain: implications for reaction and survival of slab-derived SiO2-oversaturated melt in the upper mantle. Proceedings of the Japan Academy 79:145–150.
Arai S, Takada S, Michibayashi K, Kida M. (2004) Petrology of peridotite xenoliths from Iraya volcano, Philippines, and its implication for dynamic mantle-wedge processes. Journal of Petrology 45:369–389.
Arai S, Abe N, Ishimaru S. (2006) Mantle peridotites from the Western Pacific. Gondowana Research 10: (in press).
Audétat A and Keppler H. (2004) Viscosity of fluids in subduction zones. Science 303:513–516.[CrossRef][Web of Science][Medline]
Ayers JC and Eggler DH. (1995) Partitioning of elements between silicate melt and H2O–NaCl fluids at 1· 5 and 2· 0 GPa pressure: implications for mantle metasomatism. Geochimica et Cosmochimica Acta 59:4237–4246.[CrossRef][Web of Science]
Ballhaus C, Berry RF, Green DH. (1990) Oxygen fugacity controls in the Earth's upper mantle. Nature 348:437–440.[CrossRef]
Ballhaus C, Berry RF, Green DH. (1991) High pressure experimental calibration of the olivine–orthopyroxene–spinel oxygen geobarometer: implications for the oxidation state of the upper mantle. Contributions to Mineralogy and Petrology 107:27–40.[CrossRef][Web of Science]
Beccaluva L, Bianchini G, Bonadiman C, Siena F, Vaccaro C. (2004) Coexisting anorogenic and subduction-related metasomatism in mantle xenoliths from the Betic Cordillera (southern Spain). Lithos 75:67–87.[CrossRef][Web of Science]
Bedini RM and Bodinier J-L. (1999) Distribution of incompatible trace elements between the constituents of spinel peridotite xenoliths: ICP-MS data from the East African Rift. Geochimica et Cosmochimica Acta 63:3883–3900.[CrossRef][Web of Science]
Bizimis M, Salters VJM, Bonatti E. (2000) Trace and REE content of clinopyroxenes from supra-subduction zone peridotite. Implications for melting and enrichment processes in island arcs. Chemical Geology 165:67–85.[CrossRef][Web of Science]
Blatter DL and Carmichael ISE. (1998) Hornblende peridotite xenoliths from central Mexico reveal the highly oxidized nature of subarc upper mantle. Geology 26:1035–1038.
Bodinier J-L, Merlet C, Bedini RM, Simien F, Remaidi M, Garrido C. (1996) Distribution of niobium, tantalum, and other highly incompatible trace elements in the lithospheric mantle: the spinel paradox. Geochimica et Cosmochimica Acta 60:545–550.[CrossRef][Web of Science]
Braitseva OA, Bazanova LI, Melekestsev IV, Sulerzhitskiy LD. (1998) Large Holocene eruptions of Avacha volcano, Kamchatka (7200–3500 14C years B.P.). Volcanology and Seismology 20:1–27.
Brandon AD and Draper DS. (1996) Constraints on the origin of the oxidation state of mantle overlying subduction zones: an example from Simcoe, Washington, USA. Geochimica et Cosmochimica Acta 60:1739–1749.[CrossRef][Web of Science]
Bryndzia LT and Wood BJ. (1990) Oxygen thermobarometry of abyssal spinel peridotites: the redox state and C–O–H volatile composition of the Earth's sub - occanic upper mantle. American Journal of Science 290:1093–1116.
Condie KC, Cox J, O'Reilly SY, Griffin WL, Kerrich R. (2004) Distribution of high strength and rare earth elements in mantle and lower crustal xenoliths from the southwestern United States: the role of grain-boundary phases. Geochimica et Cosmochimica Acta 68:3919–3942.[CrossRef][Web of Science]
Dawson JB. (2002) Metasomatism and partial melting in upper-mantle peridotite xenoliths from the Lashaine Volcano, Northern Tanzania. Journal of Petrology 43:1749–1777.
Dick HJB and Bullen T. (1984) Chromian spinel as a petrogenetic indicator in abyssal and alpine- type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology 86:54–76.[CrossRef][Web of Science]
Eggins SM, Rudnick RL, McDonough WF. (1998) The compositions of peridotites and their minerals: a laser-ablation ICP-MS study. Earth and Planetary Science Letters 154:53–71.[CrossRef][Web of Science]
Ertan IE and Leeman WP. (1996) Metasomatism of Cascade subarc mantle: evidence from a rare phlogopite orthopyroxenite xenolith. Geology 24:451–454.
Franz L, Becker K-P, Kramer W, Herzig PM. (2002) Metasomatic mantle xenoliths from the Bismarck microplate (Papua New Guinea)—thermal evolution, geochemistry and extent of slab-induced metasomatism. Journal of Petrology 43:315–343.
Frey FA and Green DH. (1974) The mineralogy, geochemistry and origin of lherzolites inclusion in Victorian basanites. Geochimica et Cosmochimica Acta 38:1023–1059.[CrossRef][Web of Science]
Fujimaki H, Tatsumito M, Aoki K. (1984) Partition coefficients of Hf, Zr and REE between phenocrysts and groundmasses. Journal of Geophysical Research 89:B662–B672.[CrossRef]
Gorbatov A, Kostoglodov V, Suárez G, Gordeev E. (1997) Seismicity and structure of the Kamchatka subduction zone. Journal of Geophysical Research 102:17883–17898.[CrossRef]
Gorring ML and Kay SM. (2000) Carbonatite metasomatized peridotite xenoliths from southern Patagonia: implications for lithospheric processes and Neogene plateau metasomatism. Contributions to Mineralogy and Petrology 140:55–72.[CrossRef][Web of Science]
Grégoire M, McInnes BIA, O'Reilly SY. (2001) Hydrous metasomatism of oceanic sub-arc mantle, Lihir, Papua New Guinea, Part 2. Trace element characteristics of slab-derived fluids. Lithos 59:91–108.[CrossRef][Web of Science]
Hellebrand E, Snow JE, Dick HJB, Hofmann AW. (2001) Coupled major and trace elements as indicators of the extent of melting in mid-ocean-ridge peridotites. Nature 410:677–681.[CrossRef]
Hiraga T, Anderson AM, Zimmerman ME, Mei S, Kohlstedt DL. (2002) Structure and chemistry of grain boundaries in deformed, olivine+basalt and partially molten lherzolite aggregates: evidence of melt-free grain boundaries. Contributions to Mineralogy and Petrology 144:163–175.[Web of Science]
Hiraga T, Anderson AM, Kohlstedt DL. (2003) Chemistry of grain boundaries in mantle rocks. American Mineralogist 88:1015–1019.
Hiraga T, Anderson AM, Kohlstedt DL. (2004) Grain boundaries as reservoirs of incompatible elements in the Earth's mantle. Nature 427:699–703.[CrossRef][Medline]
Hirose K. (1997) Melting experiments on lherzolite KLB-1 under hydrous conditions and generation of high-magnesian andesitic melts. Geology 25:42–44.
Hirose K and Kawamoto T. (1995) Hydrous partial melting of lherzolite at 1 GPa: the effect of H2O on the genesis of basaltic magmas. Earth and Planetary Science Letters 133:463–473.[CrossRef][Web of Science]
Hochstaedter AG, Kepezhinskas PK, Defant MJ, Drummond MS, Bellon B. (1994) On the tectonic significance of arc volcanism in northern Kamchatka. Journal of Geology 102:639–654.[Web of Science]
Ionov DA. (2004) Chemical variations in peridotite xenoliths from Vitim, Siberia: inferences for REE and Hf behavior in the garnet-facies upper mantle. Journal of Petrology 45:343–367.
Ionov DA, Bodinier J-L, Mukasa SB, Zanetti A. (2002) Mechanisms and sources of mantle metasomatism: major and trace element compositions of peridotite xenoliths from Spitsbergen in the context of numerical modelling. Journal of Petrology 43:122219–2259.
Ishida Y, Morishita T, Arai S, Shirasaka M. (2004) Simultaneous in-situ multi-element analysis of minerals on thin section using LA-ICP-MS. Science Report of Kanazawa University 48:31–42.
Jaques AL and Green DH. (1980) Anhydrous melting of peridotite at 0–15 kb pressure and the genesis of tholeiitic basalt. Contributions to Mineralogy and Petrology 73:287–310.[CrossRef][Web of Science]
Johnson KTM, Dick HJB, Shimizu N. (1990) Melting in the oceanic upper mantle: an ion microprobe study of diopsides in abyssal peridotites. Journal of Geophysical Research 95:2661–2678.
Jones AP, Smith JV, Dawson JB. (1983) Glasses in mantle xenoliths from Olmani, Tanzania. Journal of Geology 91:167–178.[Web of Science]
Kepezhinskas PK, Defant MJ, Drummond MS. (1995) Na metasomatism in the island-arc mantle by slab melt–peridotite interaction; evidence from mantle xenoliths in the North Kamchatka arc. Journal of Petrology 36:1505–1527.
Kepezhinskas PK, Defant MJ, Drummond MS. (1996) Progressive enrichment of island arc mantle by melt–peridotite interaction inferred from Kamchatka xenoliths. Geochimica et Cosmochimica Acta 60:1217–1229.[CrossRef][Web of Science]
Keppler H. (1996) Constraints from partitioning experiments on the composition of subduction-zone fluids. Nature 380:237–240.[CrossRef]
Kilian R and Stern CR. (2002) Constraints on the interaction between slab melts and the mantle wedge from adakitic glass in peridotite xenoliths. European Journal of Mineralogy 14:25–36.
Klügel A. (1998) Reactions between mantle xenoliths and host magma beneath La Palma (Canary Islands): constraints on magma ascent rates and crustal reservoirs. Contributions to Mineralogy and Petrology 131:237–257.[CrossRef][Web of Science]
Kogiso T, Tatsumi Y, Nakano S. (1997) Trace element transport during dehydration processes in the subducted oceanic crust: 1. Experiments and implications for the origin of ocean island basalts. Earth and Planetary Science Letters 148:193–205.[CrossRef][Web of Science]
Laurora A, Mazzucchelli M, Rivalenti G, Vannucci R, Zanetti A, Barbieri MA, Cingolani CA. (2001) Metasomatism and melting in carbonated peridotite xenoliths from the mantle wedge: the Gobernador Gregores case (southern Patagonia). Journal of Petrology 42:69–87.
Leake BE, Woolley AR, Arps CES, et al. (1997) Nomenclature of amphiboles. Report of the subcommittee on amphiboles of the International Mineralogical Association Commission on new minerals and mineral names. European Journal of Mineralogy 9:623–651.
Levin V, Park J, Brandon M, Lees J, Peyton V, Gordeev E, Ozerov A. (2002) Crust and upper mantle of Kamchatka from teleseismic receiver functions. Tectonophysics 358:233–265.[CrossRef][Web of Science]
Matsukage K and Kubo K. (2003) Chromian spinel during experiments of dry peridotite (KLB-1) at 1· 0–2· 5 GPa. American Mineralogist 88:1271–1278.
Maury RC, Defant MJ, Joron J-L. (1992) Metasomatism of the sub-arc mantle inferred from trace elements in Philippine xenoliths. Nature 360:661–663.[CrossRef]
McDonough WF and Frey FA. (1989) Rare earth elements in upper mantle rocks. In Lipin GA and McKay GA (Eds.). Geochemistry and Mineralogy of Rare Earth Elements. Mineralogical Society of America, Reviews in Mineralogy 21: pp. 99–145.
McDonough WF and Sun S-S. (1995) The composition of the Earth. Chemical Geology 120:223–253.[CrossRef][Web of Science]
McInnes BIA, Grégoire M, Binns RA, Herzig PM, Hannington MD. (2001) Hydrous metasomatism of oceanic sub-arc mantle, Lihir, Papua New Guinea: petrology and geochemistry of fluid-metasomatised mantle wedge xenoliths. Earth and Planetary Science Letters 188:169–183.[CrossRef][Web of Science]
Minster JB, Jordan TH, Molnar P, Haines E. (1974) Numerical modelling of instantaneous plate tectonics. Geophysical Journal 36:541–576.[CrossRef][Web of Science]
Morishita T, Ishida Y, Arai S, Shirasaka M. (2005) Determination of multiple trace element compositions in thin (<30 µm) layers of NIST SRM 614 and 616 using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). Geostandards and Geoanalytical Research 29:107–122.[CrossRef][Web of Science]
Nakamura Y and Kushiro I. (1974) Composition of the gas phase in Mg2SiO4–SiO2–H2O at 15 kbar. Carnegie Institution of Washington Yearbook 73:255–258.
Navon O and Stolper E. (1987) Geochemical consequence of melt percolation: the upper mantle as a chromatographic column. Journal of Geology 95:285–307.[Web of Science]
Neumann ER and Wulff-Pedersen E. (1997) The origin of highly silicic glass in mantle xenoliths from Canary Islands. Journal of Petrology 38:1513–1539.[CrossRef][Web of Science]
Neumann ER, Wulff-Pedersen E, Pearson NJ, Spencer EA. (2002) Mantle xenoliths from Tenerife (Canary Islands): evidence for reactions between mantle peridotites and silicic carbonatite melts inducing Ca metasomatism. Journal of Petrology 43:825–857.
Nimis P and Taylor WR. (2000) Single clinopyroxene thermometry for garnet peridotites. Part I. Calibration and testing of a Cr-in-Cpx barometer and an enstatite-in-Cpx thermometer. Contributions to Mineralogy and Petrology 139:541–554.[CrossRef]
Ninomiya A. Arai S. (1992) Harzburgite fragment in a composite xenoliths from an Oshima-Oshima andesite, the northeast Japan arc. Bulletin of the Volcanological Society of Japan 37:269–273 (in Japanese).
Niu Y. (1997) Mantle melting and melt extraction processes beneath ocean ridges: evidence from abyssal peridotite. Journal of Petrology 38:1047–1074.[CrossRef][Web of Science]
Niu Y. (2004) Bulk-rock major and trace element compositions of abyssal peridotites: implications for mantle melting, melt extraction and post-melting processes beneath mid-ocean ridges. Journal of Petrology 45:2423–2458.
Niu Y, Langmuir CH, Kinzler RJ. (1997) The origin of abyssal peridotites: a new perspective. Earth and Planetary Science Letters 152:251–265.[CrossRef][Web of Science]
Nixon PH. (1987) Mantle Xenoliths(John Wiley, New York).
Parkinson IJ and Pearce JA. (1998) Peridotites from the Izu–Bonin–Mariana forearc (ODP Leg 125): evidence for mantle melting and melt–mantle interaction in a supra-subduction zone setting. Journal of Petrology 39:1577–1618.[CrossRef][Web of Science]
Salters VJM and Shimizu N. (1988) World-wide occurrences of HFSE-depleted mantle. Geochimica et Cosmochimica Acta 52:2177–2182.[CrossRef][Web of Science]
Schiano P, Clocchiatti R, Shimizu N, Maury RC, Jochum KP, Hofmann AW. (1995) Hydrous, silica-rich melts in the sub-arc mantle and their relationship with erupted arc lavas. Nature 377:595–600.[CrossRef]
Shimizu Y, Arai S, Morishita T, Yurimoto H, Gervilla F. (2004) Petrochemical characteristics of felsic veins in mantle xenoliths from Tallante (SE Spain): an insight into activity of silicic melt within the mantle wedge. Transactions of the Royal Society of Edinburgh: Earth Sciences 95:265–276.[CrossRef][Web of Science]
Shirasaka M, Arai S, Ishimaru S, Ishida Y, Shimizu Y, Morishita T. (2004) The solution introduction ICP-MS technique for trace element analysis of rocks. Science Report of Kanazawa University 48:43–71.
Smith D, Riter JCA, Mertzman SA. (1999) Erratum to ‘Water–rock interactions, orthopyroxene growth, and Si-enrichment in the mantle: evidence in xenoliths from the Colorado Plateau, southwestern United States’ [Earth Planet. Sci. Lett. (1999) 45–54]. Earth and Planetary Science Letters 167:347–356.[CrossRef][Web of Science]
Stosch H-G and Seck HA. (1980) Geochemistry and mineralogy of two spinel peridotite suites from Dreiser Weiher, West Germany. Geochimica et Cosmochimica Acta 44:457–470.[CrossRef][Web of Science]
Suzuki K. (1987) Grain boundary enrichment of incompatible elements in some mantle peridotites. Chemical Geology 63:319–334.[CrossRef][Web of Science]
Takazawa E, Frey FA, Shimizu N, Obata M. (2000) Whole rock compositional variations in upper mantle peridotite (Horoman, Hokkaido, Japan): are they consistent with a partial melting process? Geochimica et Cosmochimica Acta 64:695–716.[CrossRef][Web of Science]
Tatsumi Y and Hanyu T. (2003) Geochemical modeling of dehydration and partial melting of subducting lithosphere: toward a comprehensive understanding of high-Mg andesite formation in the Setouchi volcanic belt, SW Japan. Geochemistry, Geophysics, Geosystems 4:1081 doi: 10.1029/2003GC000530.[CrossRef]
Tatsumi Y and Kogiso T. (1997) Trace element transport during dehydration processes in the subducted oceanic crust: 2. Origin of chemical and physical characteristics in arc magmatism. Earth and Planetary Science Letters 148:207–221.[CrossRef][Web of Science]
Tatsumi Y, Furukawa Y, Kogiso T, Yamanaka Y, Yokoyama T, Fedetov SA. (1994) A third volcanic chain in Kamchatka: thermal anomaly at transform/convergence plate boundary. Geophysical Research Letters 21:7537–540.[CrossRef][Web of Science]
Tsuchiya N, Suzuki S, Kimura J, Kagami H. (2005) Evidence for slab melt/mantle reaction: petrogenesis of Early Cretaceous, and Eocene high-Mg andesites from the Kitakami Mountains, Japan. Lithos 79:179–206.[CrossRef][Web of Science]
Varela ME, Clocchiatti R, Kurat G, Schiano P. (1999) Silicic glasses in hydrous and anhydrous mantle xenoliths from Western Victoria, Australia: at least two different sources. Chemical Geology 153:151–169.[CrossRef][Web of Science]
Wells PRA. (1977) Pyroxene thermometry in simple and complex systems. Contributions to Mineralogy and Petrology 62:129–139.[CrossRef][Web of Science]
Yaxley GM and Kamenetsky V. (1999) In situ origin for glasses in mantle xenoliths from southeastern Australia: insights from trace element compositions of glasses and metasomatic phases. Earth and Planetary Science Letters 172:97–109.[CrossRef][Web of Science]
Yaxley GM, Kamenetsky V, Green DH, Falloon TJ. (1997) Glasses in mantle xenoliths from western Victoria, Australia, and their relevance to mantle processes. Earth and Planetary Science Letters 148:433–446.[CrossRef][Web of Science]
Yogodzinski GM and Kelemen PB. (1998) Slab melting in the Aleutians: implications of an ion probe study of clinopyroxene in primitive adakite and basalt. Earth and Planetary Science Letters 158:53–65.[CrossRef][Web of Science]
Zindler A and Jagoutz E. (1988) Mantle cryptology. Geochimica et Cosmochimica Acta 52:319–333.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Facer, H. Downes, and A. Beard In situ Serpentinization and Hydrous Fluid Metasomatism in Spinel Dunite Xenoliths from the Bearpaw Mountains, Montana, USA J. Petrology, August 1, 2009; 50(8): 1443 - 1475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Ishimaru and S. Arai Arsenide in a metasomatized peridotite xenolith as a constraint on arsenic behavior in the mantle wedge American Mineralogist, July 1, 2008; 93(7): 1061 - 1065. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Arai and S. Ishimaru Insights into Petrological Characteristics of the Lithosphere of Mantle Wedge beneath Arcs through Peridotite Xenoliths: a Review J. Petrology, April 1, 2008; 49(4): 665 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Ishimaru and S. Arai Calcic amphiboles in peridotite xenoliths from Avacha volcano, Kamchatka, and their implications for metasomatic conditions in the mantle wedge Geological Society, London, Special Publications, January 1, 2008; 293(1): 35 - 55. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||