| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 45 | Number 2 | Pages 235-252 | 2004
© Oxford University Press 2004; all rights reserved
Possible Non-melted Remnants of Subducted Lithosphere: Experimental and Geochemical Evidence from Corundum-Bearing Mafic Rocks in the Horoman Peridotite Complex, Japan
1 RESEARCH SCHOOL OF EARTH SCIENCES, THE AUSTRALIAN NATIONAL UNIVERSITY, CANBERRA, ACT 0200, AUSTRALIA
2 GRADUATE SCHOOL OF NATURAL SCIENCE AND TECHNOLOGY, KANAZAWA UNIVERSITY, KAKUMA, KANAZAWA 920-1192, JAPAN
3 DEPARTMENT OF EARTH SCIENCES, KANAZAWA UNIVERSITY, KAKUMA, KANAZAWA 920-1192, JAPAN
RECEIVED NOVEMBER 15, 2002; ACCEPTED SEPTEMBER 4, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL BACKGROUND AND SAMPLE...EXPERIMENTAL APPROACHWHOLE-ROCK TRACE-ELEMENT...CONCLUSIONSREFERENCES |
|---|
We report experimental results and whole-rock trace-element characteristics of a corundum-bearing mafic rock from the Horoman peridotite complex, Japan. Coronitic textures around corundum in the sample suggest that corundum was not stable in mafic rock compositions during the late-stage P–T conditions recorded in the complex (P < 1 GPa, T < 800°C). Based on the experimental results, corundum is stable in aluminous mafic compositions at pressures of 2–3 GPa under dry conditions, suggesting that the corundum-bearing mineral assemblages developed under upper-mantle conditions, probably within the surrounding peridotite. Variations in the trace-element compositions of the corundum-bearing mafic rock and related rocks can be controlled by modal variations of plagioclase, clinopyroxene and olivine, suggesting that they formed as gabbroic rocks at low-pressure conditions, and that the corundum-bearing mafic rock was derived from a plagioclase-rich protolith. A complex P–T trajectory, involving metamorphism of the plagioclase-rich protolith at a pressure higher than that at which it was first formed, is needed to explain the origin of the corundum-bearing mafic rocks. They show no evidence for partial melting after their formation as low-pressure cumulates. The Horoman complex is an example of a large peridotite body containing possible remnants of subducted oceanic lithosphere still retaining their original geochemical signatures without chemical modification during subduction and exhumation.
KEY WORDS: Horoman; mafic rock; corundum; experiment; P–T history; recycling
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUND AND SAMPLE...EXPERIMENTAL APPROACHWHOLE-ROCK TRACE-ELEMENT...CONCLUSIONSREFERENCES |
|---|
Subducted oceanic lithosphere is expected to be stretched, boudinaged and thinned by mantle convection leading to a ‘marble-cake’ mantle (Allègre & Turcotte, 1986
). Mafic layers of pyroxenite/eclogite (referred to as ‘mafic rock’ hereafter) that represent recycled oceanic crustal materials in a peridotite host are thought to be an important magma source component in the production of geochemically distinctive magmas (e.g. Takahashi et al., 1998). A basic question is to what extent can subducted oceanic lithosphere retain its original geochemical characteristics during mantle convection. To answer this question, it is desirable directly to determine the geochemical characteristics of recycled subducted oceanic lithosphere that has both mafic and ultramafic lithologies and is a product of high-temperature upwelling after a prior history of high-pressure metamorphism.
Mafic rocks commonly occur in alpine-type peridotite massifs. Some of them have been interpreted to be recycled crustal materials (Polvé & Allègre, 1980; Allègre & Turcotte, 1986; Kornprobst et al., 1990; Morishita et al., 2003b). In the alpine-type peridotite massifs, it is, however, difficult to determine the origin of the mafic rocks because of complete replacement by later mineral assemblages as a result of decompressional metamorphism. Determining the phase relationships and trace-element and/or isotopic compositions of these mafic rocks at upper-mantle conditions would provide strong constraints on the evolutionary history of the mafic rocks (e.g. Kornprobst et al., 1990; Takazawa et al., 1999; Morishita et al., 2003b).
The Horoman peridotite complex, Samani, Japan, is known to have a conspicuous layered structure consisting of peridotite with minor amounts of mafic rocks (e.g. Igi, 1953; Komatsu & Nochi, 1966; Niida, 1984; Takahashi, 1991; Takazawa et al., 1999). One type of the mafic rocks (referred to as Type II) has been interpreted as having originally formed as a cumulate consisting of plagioclase, olivine and clinopyroxene, suggesting a low-pressure origin (Shiotani & Niida, 1997; Takazawa et al., 1999). Takazawa et al. (1999) have proposed a complex P–T history for the Type II mafic rocks, including high-pressure recrystallization after initial formation at lower-pressure conditions, on the basis of geochemical signatures combined with chronological data. Morishita & Arai (2001) have discussed the petrogenesis of corundum-bearing Type II mafic rocks and have suggested that corundum forms by metamorphism and/or partial melting (P > 1·5 GPa) of Type II protoliths during subduction. Thus the Type II mafic rocks in the Horoman complex may represent remnants of recycled crustal materials now preserved in a large peridotite body. Some important questions, however, remain unanswered: (1) under what P–T conditions is corundum stable in aluminous mafic compositions? (2) What are the geological relationships between the corundum-bearing mafic rocks, corundum-free mafic rocks and the surrounding peridotites? (3) What was the geochemical evolution of the mafic rocks during their complex P–T history?
New experiments specifically designed to investigate these highly aluminous mafic compositions and determinations of their whole-rock trace-element characteristics provide important constraints on the evolutionary history of the corundum-bearing mafic rocks from the Horoman complex in the context of recycling of subducted oceanic lithosphere in the mantle.
|
GEOLOGICAL BACKGROUND AND SAMPLE SELECTIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUND AND SAMPLE...EXPERIMENTAL APPROACHWHOLE-ROCK TRACE-ELEMENT...CONCLUSIONSREFERENCES |
|---|
The Horoman peridotite complex is located at the southern end of the Main Zone of the Hidaka metamorphic belt (e.g. Komatsu et al., 1982
), and is 8 km x 10 km in area and more than 3 km in thickness (e.g. Igi, 1953; Komatsu & Nochi, 1966; Niida, 1984; Sawaguchi, 2001). The complex consists of various kinds of layered peridotite with small amounts of mafic rocks (e.g. Komatsu & Nochi, 1966; Niida, 1984). The petrography and mineralogy of the Horoman peridotites have been described in detail by previous workers (e.g. Niida, 1984; Takahashi, 1991; Takazawa et al., 2000; Yoshikawa & Nakamura, 2000, and references therein). Komatsu & Nochi (1966) and Niida (1984) divided the complex into two zones, the Upper and Lower Zones (Fig. 1). The Upper Zone is characterized by an abundance of mafic layers and by sharp lithological boundaries. In contrast, the Lower Zone is characterized by gradational lithological boundaries. Ozawa & Takahashi (1995) examined the systematic changes in mineralogical and petrographical characteristics from the Upper to Lower Zones and the chemical zoning of minerals in the plagioclase peridotites. They suggested that the Horoman peridotite ascended as a mantle diapir from the garnet-peridotite stability field, following a higher-temperature decompression path (1150°C at >2·0 GPa) for the Upper Zone and a lower-temperature path (900°C at >2·0 GPa) for the Lower Zone. We investigated the southern part of the complex (Fig. 1).
|
The mafic rocks in the Horoman complex have been divided into a variety of sub-types (up to five types) (Niida, 1984
; Shiotani & Niida, 1997; Takazawa et al., 1999). In this paper, we basically follow the nomenclature of Takazawa et al. (1999) and refer to them simply as ‘mafic rocks’. A corundum-bearing mafic rock boulder found in the Horoman complex belongs to the Type II mafic rocks (Morishita & Kodera, 1998), characterized by the presence of low-Ti clinopyroxene (Shiotani & Niida, 1997; Takazawa et al., 1999). The petrography of the studied samples has been described in more detail in a separate paper (Morishita & Arai, 2001).
Type II mafic rocks (Fig. 2a) appear exclusively within cumulus peridotite [the SDW suite of Takahashi (1991); Fig. 1] as thin layers (1–200 cm in thickness) concordant with the foliation of the peridotite (Shiotani & Niida, 1984; Takazawa et al., 1999; Morishita & Arai, 2001). Some layers have centimetre-scale internal layering corresponding to a variation of mineral mode. Dunite layers with plagioclase-rich seams characteristically alternate with the SDW peridotites and Type II mafic rocks at intervals of a few centimetres to a few tens of centimetres in the Upper Zone (Fig. 2a). The Type II mafic rocks are more common in the Upper Zone than in the Lower Zone. A Type II mafic rock containing corundum was found as a boulder derived from the complex (Morishita & Kodera, 1998, Fig. 2b).
|
Type II mafic rocks in the Upper Zone have fine-grained (<0·5 mm) equigranular metamorphic textures consisting of plagioclase (An83–91), clinopyroxene (TiO2 <0·4 wt %, Cr2O3 <0·8 wt %) and olivine (Fo 86–89; NiO 0·21–0·25 wt %) with small amounts of amphibole, green spinel {Cr-number [= Cr/(Cr + Al) atomic ratio] <0·1}, brown spinel (Cr-number >0·25) and orthopyroxene. Type II mafic rocks in the Lower Zone show almost the same texture as those in the Upper Zone but are characterized by grain-to-grain chemical heterogeneity (Morishita & Arai, 1997
, 2001). Dunite layers with plagioclase-rich seams have an equigranular texture consisting of relatively fine-grained plagioclase (An90), olivine (Fo 90; NiO 0·28–0·32 wt %), clinopyroxene (TiO2 <0·3 wt %; Cr2O3 <1 wt %) and brown spinel (Cr-number 0·2 and >0·3 for coarse and fine grains, respectively).
The corundum-bearing rock has a layered structure, with olivine-rich (Fo 86; NiO 0·14 wt %) and olivine-poor parts. Associated clinopyroxene is also low in both TiO2 (<0·2 wt %) and Cr2O3 (<0·6 wt %). Amphibole is more abundant in the olivine-rich layer. A zonal arrangement of inner green spinel (Cr-number F < 0·1) and outer plagioclase is observed around large grains of pink corundum (up to 3 cm across), which are sporadically distributed in matrix (Fig. 2b). The corundum sometimes includes saussuritized plagioclase (An > 55). The textural relationships within the corundum-bearing mineral assemblages indicate that the corundum reacted with another phase, probably clinopyroxene, to form the green spinel and plagioclase, indicating that corundum was not stable in the Type II mafic rocks during the later P–T conditions of the Horoman complex (Morishita & Arai, 2001).
|
EXPERIMENTAL APPROACH |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUND AND SAMPLE...EXPERIMENTAL APPROACHWHOLE-ROCK TRACE-ELEMENT...CONCLUSIONSREFERENCES |
|---|
Experimental procedures and analytical techniques
Råheim & Green (1974)
experimentally determined the solidus and liquidus relationships for a Lunar Highland Basalt (gabbroic anorthosite). In their experiments, corundum was found above the solidus at pressure of more than 1·8 GPa. However, their results cannot be directly applied to the corundum-bearing mafic rock in the Horoman complex, because the Ca and Na contents and Mg-number [Mg/(Mg + Fe) atomic ratio], which are critical for garnet and clinopyroxene compositions, are higher in the corundum-bearing mafic rocks than in the starting material in the experiments of Råheim & Green (1974) (Table 1). The composition selected for experiment in this study (aluminous mafic composition) is the corundum-bearing Type II mafic rock obtained by Morishita & Arai (2001) reported in Table 1. The starting material was prepared using the sol–gel procedure (Luth & Ingamells, 1965; Hermann & Green, 2001). The gel was fused to a glass at 1450°C in an Ar atmosphere in a platinum crucible. The chemical composition of the glass was checked by electron microprobe analysis. We confirm there is no serious iron loss during the making of the glass. Another starting material (Mix A) consisting of the glass +10% of a high-pressure assemblage of grossular-rich garnet, clinopyroxene and kyanite, with trace amounts of corundum, was also prepared to determine the growth or disappearance of the seed minerals. The high-pressure seed minerals were prepared in large capacity (100 mg) Ag50Pd50 capsules at 35 kbar, 1150°C, run time 24 h; the kyanite and corundum were usually less than 1 µm in size. Subsequently, the high-pressure mineral seeds were crushed in an agate mortar before mixing with the glass powder. The starting materials were dried overnight at 110°C. Nominally anhydrous experiments were carried out using a 1·27 cm piston-cylinder apparatus at the Research School of Earth Sciences, the Australian National University (ANU). For the high-pressure runs (>3·5 GPa), the starting mix was faintly moistened by breath to enhance the reaction. Samples were enclosed in sealed, graphite-lined, Pt capsules, to prevent Fe loss by contact between the samples and the Pt. The capsule assembly was surrounded by alumina tubes and discs, and placed between composite 60% dense MgO and pyrophyllite inserts (fired at 1050°C). Graphite heaters were used in NaCl–Pyrex sleeves. In all experiments the sample assemblies were heated and pressurized simultaneously. Temperature was controlled by a Eurotherm 904 controller attached to a type B thermocouple (Pt6Rh94/Pt30Rh70), and is accurate to ±10°C and precise to ±1°C. No correction was made for the effect of pressure on thermocouple output e.m.f. Experiments were quenched by terminating the power to the furnace.
|
After the experimental runs, capsules were mounted whole in epoxy, and sectioned longitudinally. Polished run products were examined by reflected light microscopy, and by scanning electron microscopy (SEM) and electron probe microanalysis (EPMA), using a JEOL 6400 system fitted with a Link Energy Dispersive Detector at the Electron Microscopy Unit, ANU. Analytical conditions were 15 kV accelerating voltage and 1 nA beam current, using spot mode (2–3 µm beam diameter). Phases in the run products were usually less than 20 µm across. The experimental conditions and synthesized phase assemblages are presented in Table 2 and plotted in Fig. 3. At sub-solidus conditions, the pyroxenes and garnets exhibit variable degrees of heterogeneity. This might result from the presence of ubiquitous mineral inclusions and the sluggishness of Al diffusion in minerals (e.g. Schairer & Bowen, 1955
). A geochemically distinctive portion, which might be a relict seed mineral, was occasionally recognizable in the analyses of the inner part of the minerals. In these cases, the rims of minerals were analysed as representative of the compositions of the minerals. Mineral analysis demonstrates that total equilibrium was not attained in any of these runs. Corundum in the runs is, however, usually larger than the seed minerals (>2–3 µm) and occurs as equant grains. It is assumed that corundum was stable at the P–T conditions of the experiments. Admittedly, these criteria alone do not rule out the possibility of metastable equilibria. The amount of garnet with respect to clinopyroxene increases with increasing pressure. Furthermore, both garnet and clinopyroxene compositions systematically changed with changing pressure and temperature as discussed below. These results appear to be consistent with those of earlier experiments (e.g. Råheim & Green, 1974).
|
|
Experimental results
The solidus for the aluminous mafic composition was determined by the presence of quenched glass (high-SiO2, -FeO and -Na2O phase). It is usually difficult to determine the precise composition of the quenched glass because of the small size of the melt pools. Corundum appears on the solidus at more than 2·4 GPa. In sub-solidus conditions, only clinopyroxene occurs over the whole P–T range investigated. Garnet appears at pressures greater than 1·8 GPa, whereas plagioclase was found at less than 2·4 GPa. Corundum appears at pressures greater than 2·0 GPa but disappears at more than 3·0 GPa. Kyanite appears above 2·4 GPa, whereas spinel is present at less than 2·0 GPa. A backscattered electron image and X-ray intensity maps of a typical run product at 1200°C, 2 GPa are illustrated in Fig. 4. In runs at pressures of 1·4–1·6 GPa, clinopyroxene is very heterogeneous in terms of its Al2O3 content and Mg-number. This might be caused by the presence of very tiny garnet grains in the clinopyroxene. In runs at pressure conditions of more than 2·8 GPa, it was not possible to obtain clinopyroxene analyses that gave good structural formulae because of the ubiquitous presence of kyanite and/or corundum inclusions. The true clinopyroxene composition would, therefore, have a lower Al2O3 content than the bulk analysis. In this case, the clinopyroxene compositions were corrected by subtraction of the kyanite component, determined in a way such that the calculated formula approaches the ideal for clinopyroxene, i.e. cation total = 4 (O = 6). In such cases, the corrected clinopyroxene data were normalized to 100 wt %. Representative analyses are shown in Table 3. After correction, the mineral compositions systematically change as a function of pressure and temperature (Figs 5 and 6). The grossular content of garnet and the jadeite content of clinopyroxene increase with increasing pressure. The Ca-Tschermak content of clinopyroxene increases to 2·4 GPa but decreases at higher pressure (Fig. 6). The alumina content of the clinopyroxene is very high (up to 20 wt % at 2·0–2·8 GPa). High-alumina clinopyroxene (up to 20 wt % in Al2O3 in the core) has been found in corundum-bearing garnet pyroxenites from the Beni Bousera peridotite massif, Morocco (Kornprobst et al., 1990
). The whole-rock chemical compositions of the corundum-bearing garnet pyroxenites in the Beni Bousera massif are similar to those of the corundum-bearing mafic rocks in the Horoman complex (Morishita et al., 2001).
|
|
|
|
Figure 3, together with the data on the detailed changes in pyroxene and garnet composition, illustrates the mineralogical complexity of the metamorphic reactions that transform the originate aluminous gabbroic mineral assemblage to the eclogitic high-pressure assemblage.
Breakdown of plagioclase usually produces quartz by the reactions
 |
 |
 |
 |
 |
 |
Interpretation of corundum-bearing mineral assemblages and P–T trajectory of the Type II mafic rocks
The mafic rocks in the Horoman complex do not have primary igneous (plutonic) textures but have secondary recrystallization (metamorphic) textures (Takazawa et al., 1999). The existence of a coronitic texture around corundum in the corundum-bearing mafic rocks also indicates that the corundum was not stable in Type II mafic rocks during the later P–T conditions of the Horoman complex, i.e. within the plagioclase-peridotite stability field (P < 1 GPa, T < 800°C: Ozawa & Takahashi, 1995; Takazawa et al., 1996). The experimental results of this study clearly indicate that corundum is stable in aluminous mafic compositions at pressures between 2·0 and 3·0 GPa, under both sub-solidus and super-solidus conditions. This result strongly suggests that corundum-bearing mineral assemblages in the Type II mafic rocks were produced under upper-mantle conditions, probably along with the surrounding peridotite. Furthermore, it is possible that the Type II mafic rocks have never experienced partial melting, because the experimentally determined solidus of the aluminous mafic composition is higher in temperature than the decompression P–T path proposed for the Horoman peridotites (Ozawa & Takahashi, 1995; Takazawa et al., 1996). The Horoman peridotite ascended from the garnet peridotite stability field to the plagioclase peridotite stability field at sub-solidus conditions (Ozawa & Takahashi, 1995; Takazawa et al., 1996; Yoshikawa & Nakamura, 2000); however, Takahashi (1997, 2001) reported that very low-degree partial melting occurred very locally in the plagioclase peridotites within the plagioclase peridotite stability field.
|
WHOLE-ROCK TRACE-ELEMENT COMPOSITIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUND AND SAMPLE...EXPERIMENTAL APPROACHWHOLE-ROCK TRACE-ELEMENT...CONCLUSIONSREFERENCES |
|---|
Sample selections and analytical methods
Portions with and without large corundum–spinel–plagioclase clots were prepared to determine the geochemical characteristics of the corundum-bearing rock because of sample heterogeneity corresponding to the sporadic occurrence of the large mineral clots (Morishita & Arai, 2001
). The sample was sliced with a diamond saw into thin slabs including large mineral clots (crn1), and the corundum-free portion of the rock (crn2) was obtained by eliminating the large mineral clots from the slabs.
Whole-rock major-element compositions of Horoman Type II mafic rocks (± corundum) and related rocks were reported by Morishita & Arai (2001), and define linear trends for CaO and Al2O3 against MgO and SiO2 content (Fig. 7). The corundum-bearing rock and the dunite layers with plagioclase-rich seams plot on the Al-rich and Al-poor extensions, respectively, of the major-element chemical trends defined by the corundum-free Type II mafic rocks (Fig. 7). We selected eight samples from Morishita & Arai (2001) for analyses of trace-element compositions to cover the whole-rock major-element chemical range; the corundum-bearing rock (crn1 and crn2), olivine-rich and olivine-poor rocks alternating with crn1 and crn2, a dunite layer with plagioclase-rich seams (6176001) and ordinary Type II samples (62501B, 62501C and 6176003) (Fig. 7). Although sample 62501B belongs to the Type II mafic rocks as described by Morishita & Arai (2001), it occurs as a centimetre-scale layer alternating with ordinary Type II mafic rock (62501C) and is similar both in texture and geochemistry to the dunite layers with plagioclase-rich seams. These samples, except for the corundum-bearing mafic rock, were taken from the same succession of the Upper Zone of the complex (Fig. 1).
|
Trace-element compositions were determined on fused glasses prepared by mixing rock powders with one-third Li-borate, and analysed by laser ablation (193 nm ArF excimer)–inductively coupled plasma mass spectrometry (Agilent 7500S) (LA–ICP-MS) at the Research School of Earth Sciences, ANU, following the method of Eggins et al. (1998)
. Each analysis was performed by ablating a 230 µm diameter spot at 5 pulses/s to produce an ablation rate of 0·1–0·2 µm/pulse. The NIST 612 glass was used as the primary calibration standard and was analysed at the beginning and end of each batch of <7 unknowns, with a linear drift correction applied between each calibration. Data reduction was facilitated using Ca as the internal standard, based on CaO contents measured by X-ray fluorescence (XRF). The capability of this method for determining trace elements [Sc, Ti, V, Ga, Rb, Sr, Y, Zr, Nb, Cs, Ba, rare earth elements (REE), Hf, Ta, Pb, Th and U] by LA–ICP-MS analysis at ANU in bulk geological materials including peridotites, basalts, granites and sediments prepared as Li-borate glasses has been evaluated by Eggins (2003). More than 95% of the determinations lie within uncertainty of the recommended working values at the 95% confidence level. Impurities in the Li-borate flux contribute at ultra-trace levels to only a few elements, most notably La (Sylvester, 2001; Eggins, 2003). In the studied samples, no selective enrichment of La was found. Three analyses for each sample were averaged and their means are reported in Table 4.
|
Analytical results
There is no apparent difference in trace-element composition between crn1 and crn2 except for Ti, Gd, Pb and Th (Table 4). The relatively high contents of Ti and Gd in crn1 are consistent with the presence of corundum, which would contain trace amounts of these elements (Sutherland et al., 1998
). The measured Th contents are too low to discuss the differences between them. The Pb analyses of the Li-borate glasses by LA–ICP-MS might be unreliable because of an impurity in the Li-borate flux as well as the possibility of contamination during sample preparation and, therefore, we do not place much emphasis on these data. We averaged the data for crn1 and crn2 as reported as corundum-bearing mafic rock (crn3), and compare these averages with other related rocks hereafter (Table 4).
The primitive mantle (McDonough & Sun, 1995)-normalized (PM) trace element patterns of all the samples analysed are characterized by positive anomalies of Eu and Sr, and negative anomalies of Zr and Nb (except for 6176003) (Fig. 8), and are consistent with the data obtained by Takazawa et al. (1999). There is, however, significant layer-to-layer variation. The PM-normalized pattern of the corundum-bearing mafic rock (crn3) is characterized by low REE abundances, especially the heavy REE (HREE), relative to other Type II mafic rocks. The PM-normalized patterns of the dunite layers with plagioclase-rich seams (6176001 and 62501B) are the lowest in middle REE (MREE; Sm, Eu and Gd) and have weak positive anomalies of Eu and Sr. The PM-normalized pattern of 6176003 shows higher concentrations of La, Ce, Nd, Ba and Rb than those of the other samples. The olivine-rich rock alternating with the corundum-bearing rock has the highest HREE content of all the studied rocks.
|
Origin of geochemical variation of the corundum-bearing mafic rock and related rocks: implications for their protoliths
MREE to HREE contents are lower in both the Al-rich and Al-poor varieties than in the ordinary Type II mafic rocks (Fig. 8). Thus, the linear major-element chemical trends in Fig. 7 cannot be explained by simple mixing of an Al-rich end-member (the corundum-bearing mafic rock) and an Al-poor end-member (dunite layers with plagioclase-rich seams). The PM-normalized trace-element pattern of the corundum-bearing mafic rock is similar to that of plagioclase-rich rocks, such as oceanic gabbros, which are characterized by positive anomalies of Sr and Eu and a negative Zr anomaly (Cortesogno et al., 2000
). Furthermore, the geochemical relationships between the corundum-bearing mafic rock and other Horoman Type II mafic rocks are similar to those between troctolite and gabbroic rocks from ophiolites or from the oceanic lower crust (Fig. 8); these relationships are controlled by variations in the modal proportions of plagioclase, clinopyroxene and olivine (e.g. Zimmer et al., 1995; Benoit et al., 1996; Casey, 1997; Kelemen et al., 1997). On the other hand, the PM-normalized trace-element patterns of the dunite layers with plagioclase-rich seams might be explained by accumulation of clinopyroxene (+ olivine) with small amounts of plagioclase and olivine.
We conclude that the corundum-bearing mafic rock was originally formed as a plagioclase-rich cumulate alternating with gabbros (protolith of ordinary Type II mafic rocks) and cumulus peridotites (protolith of dunite layers with plagioclase-rich seams and the SDW suite). This interpretation is consistent with that of Shiotani & Niida (1997), Takazawa et al. (1999) and Morishita & Arai (2001). Shiotani & Niida (1997) and Takazawa et al. (1999) suggested that protoliths of the Type II mafic rocks might be a series of cumulates originally consisting of olivine, plagioclase and clinopyroxene, formed under low-pressure conditions on the basis of their geochemical characteristics. Both Type II mafic rocks and the dunite layers with plagioclase-rich seams occur exclusively in the SDW suite (cumulus peridotite). The major-element compositions of the whole rocks and their constituent minerals gradually change from the SDW peridotites to the Type II mafic rocks through the dunite layers with plagioclase-rich seams (Shiotani & Niida, 1997; Morishita & Arai, 2001). Based on these lines of evidence, the Type II mafic rocks and the dunite layers with plagioclase-rich seams are inferred to have crystallized from a melt that was responsible for the formation of the SDW cumulus peridotites, possibly during the major melting event recorded in the residual peridotites [i.e. the MHL suite of Takahashi (1991)].
As noted above, the PM-normalized trace-element pattern of one ordinary Type II mafic rock (6176003) is the highest in light REE (LREE) content of all the studied rocks. The local enrichment of highly incompatible trace elements might be caused by metasomatism by aqueous fluids because the Horoman peridotites have been interpreted to be metasomatized by aqueous fluids, possibly derived from dehydration of a subducted slab (Hirai & Arai, 1987; Yoshikawa et al., 1993; Yoshikawa & Nakamura, 2000; Matsumoto et al., 2001; Morishita et al., 2003a). If metasomatism occurred in the Type II mafic rocks and related dunitic rocks, the degree of metasomatism was probably variable corresponding to the modal mineralogy. The studied rocks, however, show no apparent systematic difference in their degree of enrichment in highly incompatible elements, which would be expected in a sequence of metasomatized rocks. Thus we infer that the Type II mafic rocks (± corundum) have largely retained their original geochemical signatures, although details of their geochemical characteristics may have been slightly modified by metasomatism.
Discussion
Based on our experimental results and the trace-element compositions of the corundum-bearing aluminous mafic rocks in the Horoman complex, combined with the results of previous studies (Takahashi, 1991, 1992; Takazawa et al., 1992, 1996, 1999, 2000; Ozawa & Takahashi, 1995; Yoshikawa & Nakamura, 2000; Morishita & Arai, 2001), we infer an evolutionary history of the Horoman complex as follows (Fig. 9). (1) The protolith of the Type II mafic rocks as well as the cumulus peridotites (the dunite layers with plagioclase-rich seams and the SDW suite) was originally formed as a series of cumulates at low pressures (<1 GPa; for example, under lower-crustal conditions within the oceanic crust), from a melt related to a major melting event recorded in the residual peridotite (the MHL suite) at 830 Ma. (2) The protoliths of both the Type II mafic rocks and the cumulus peridotites experienced higher-pressure metamorphism (P > 2·0 GPa), probably related to subduction of the oceanic lithosphere. (3) Finally, a diapir of peridotite, now represented by the Horoman massif, ascended from the garnet stability field following the decompression P–T path suggested by Ozawa & Takahashi (1995) and Takazawa et al. (1996).
|
Oceanic lithosphere is formed at mid-ocean ridges by partial melting of mantle peridotite and is recycled into the mantle at subduction zones. Upon subduction, the mafic portion of the oceanic lithosphere will be converted to eclogite, and some of it may subsequently melt. In the example that we have studied, the mafic rocks do not appear to have experienced partial melting after their formation as low-pressure cumulates; they may represent fragments of oceanic lithosphere recycled back to the surface retaining their original composition without melting during subduction and subsequent exhumation. This may be possible only within the shallower part of the asthenosphere in which ductile deformation without melting prevails, although it is not possible to constrain the maximum P–T conditions that the Horoman complex experienced. Alternatively, Yoshikawa & Nakamura (2000)
have suggested that the Horoman peridotites with simple stratification caused by polybaric mid-ocean ridge basalt (MORB) extraction were subsequently incorporated into a supra-subduction zone mantle wedge, by thinning and folding under sub-solidus conditions with the addition of aqueous fluids released from the subducted slab (Yoshikawa & Nakamura, 2000). Yoshikawa & Nakamura (2000), however, did not discuss the relationships between the peridotites and the mafic rocks in detail.
The geochemical evolution of the studied rocks is consistent with the origin of Sr-rich melt inclusions in Mauna Loa olivines (Sobolev et al., 2000). Sobolev et al. envisaged that the Sr-enriched melt inclusions reflect the chemical signature of ‘ghost plagioclase’ derived from former plagioclase-rich gabbros that retained their original chemical characteristics during mantle convection without mixing with other portions of the former oceanic crust.
Morishita et al. (2003b) described the geochemical characteristics of aluminous mafic rocks that sometimes contain corundum and/or sapphirine from the Ronda massif, Spain (Fig. 8) and suggested that these rocks are possible recycled oceanic crustal materials that have retained their low-pressure compositional signature in a centimetre-scale layer structure. Some large peridotite massifs, e.g. the Horoman, Ronda and Beni Bousera massifs, may provide good examples of lithospheric mantle heterogeneity created by mixing of recycled oceanic crust and chemical modification during mantle convection, although only relatively shallow-level processes are represented by these rocks. It remains uncertain, however, whether such heterogeneities could melt individually and their melt fractions remain isolated and distinctive during upward transport through many kilometres of mantle rock (Yaxley & Green, 1998), and subsequently mix with more voluminous melt pools in the shallow mantle or crustal magma reservoirs. In future studies it could be useful to examine the reaction between mantle peridotite and a melt generated by selective fusion of one of the mafic rock types in these complexes, particularly the more iron-rich rock types, which may be more representative of upwelling mantle material containing mafic layers.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUND AND SAMPLE...EXPERIMENTAL APPROACHWHOLE-ROCK TRACE-ELEMENT...CONCLUSIONSREFERENCES |
|---|
Our experimental results have revealed that corundum is stable at upper-mantle conditions (2–3 GPa) in aluminous mafic compositions under dry conditions, suggesting that the corundum-bearing mafic rocks in the Horoman peridotite complex equilibrated at upper-mantle conditions with the surrounding peridotite. Geochemical variations in the corundum-bearing mafic and related rocks probably resulted from variations in the modal proportions of plagioclase, clinopyroxene and olivine, suggesting that they were originally formed as a series of mafic to ultramafic cumulates under low-pressure conditions, of which the corundum-bearing mafic rock was a plagioclase-rich lithology. The protoliths of the low-pressure mafic to ultramafic cumulates experienced higher-pressure metamorphism (P > 2·0 GPa), probably caused by subduction. The Horoman complex is an example of a large peridotite body containing possible remnants of subducted oceanic lithosphere still retaining their original geochemical signatures without significant chemical modification during subduction and subsequent exhumation.
|
ACKNOWLEDGEMENTS |
|---|
We are grateful to Bill Hibberson, Dean Scott, Steve Eggins, Mike Shelley, Charlotte Allen, Jörg Hermann and Linda Hanley for technical assistance with the high-pressure experiments and LA–ICP-MS analyses at RSES (ANU), to Frank Brink for assistance in microprobe analyses at EMU (ANU), and to the Board of Education of Samani Town for kindly permitting us to use the facilities of the Managing Office of Apoi-dake to make some thin sections. Steve Eggins gave us valuable guidance on the LA–ICP-MS analyses of Li-borate glasses. Critical comments from Riccardo Vannucci and two anonymous reviewers improved the manuscript. T.M. thanks Tadahiro Kodera for his assistance in collecting samples. This research was supported by the JSPS Fellowships for Japanese Junior Scientists (T.M.).
|
FOOTNOTES |
|---|
* Corresponding author. E-mail: moripta{at}kenroku.kanazawa-u.ac.jp
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUND AND SAMPLE...EXPERIMENTAL APPROACHWHOLE-ROCK TRACE-ELEMENT...CONCLUSIONSREFERENCES |
|---|
Allègre, C. J. & Turcotte, D. L. (1986). Implications of a two-component marble-cake mantle. Nature 323, 123–127.[CrossRef]
Benoit, M., Polvé, M. & Ceuleneer, G. (1996). Trace element and isotopic characterization of mafic cumulates in a fossil mantle diapir (Oman ophiolite). Chemical Geology 134, 199–214.[CrossRef][Web of Science]
Casey, J. F. (1997). Comparison of major- and trace-element geochemistry of abyssal peridotites and mafic plutonic rocks with basalts from the MARK region of the Mid-Atlantic Ridge. In: Karson, J. A., Cannat, M., Miller, D. J. & Elthon, D. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 153. College Station, TX: Ocean Drilling Program, pp. 181–241.
Cortesogno, L., Gaggero, L. & Zanetti, A. (2000). Rare earth and trace elements in igneous and high-temperature metamorphic minerals of oceanic gabbros (MARK area, Mid-Atlantic Ridge). Contributions to Mineralogy and Petrology 139, 373–393.[CrossRef][Web of Science]
Eggins, S. M. (2003). Laser ablation ICP-MS analysis of geological materials prepared as lithium borate glasses. Geostandards Newsletter 27, 145–162.
Eggins, S. M., Rudnick, R. L. & McDonough, W. F. (1998). The composition of peridotites and their minerals: a laser ablation ICP-MS study. Earth and Planetary Science Letters 154, 53–71.[CrossRef][Web of Science]
Frey, F. A., Shimizu, N., Leinbach, A., Obata, M. & Takazawa, E. (1991). Compositional variations within the lower layered zone of the Horoman peridotite, Hokkaido, Japan: constraints on models for melt–solid segregation. Journal of Petrology, Special Lherzolite Issue 211–227.
Green, D. H. & Falloon, T. J. (1998). Pyrolite: a Ringwood concept and its current expression. In: Jackson, I. (ed.) The Earth's Mantle: Composition, Structure and Evolution. Cambridge: Cambridge University Press, pp. 311–378.
Hermann, J. & Green, D. H. (2001). Experimental constraints on high pressure melting in subducted crust. Earth and Planetary Science Letters 188, 149–168.[CrossRef][Web of Science]
Hirai, H. & Arai, S. (1987). H2O–CO2 fluids supplied in alpine-type mantle peridotites: electron petrology of relic fluid inclusions in olivines. Earth and Planetary Science Letters 85, 416–425.[CrossRef][Web of Science]
Igi, S. (1953). Petrographical studies on the peridotite in the Horoman region at the southern end of the Hidaka mountain range, Hokkaido. Journal of Geological Society of Japan 59, 111–121 (in Japanese with English abstract).
Kelemen, P. B., Koga, K. & Shimizu, N. (1997). Geochemistry of gabbro sills in the crust–mantle transition zone of the Oman ophiolite: implications for the origin of the oceanic lower crust. Earth and Planetary Science Letters 146, 475–488.[CrossRef][Web of Science]
Komatsu, M. & Nochi, M. (1966). Ultrabasic rocks in the Hidaka metamorphic belt, Hokkaido, Japan I—Mode of occurrence of the Horoman ultrabasic rocks. Earth Science 20, 21–29 (in Japanese with English abstract).
Komatsu, M., Miyashita, S., Maeda, J., Osanai, Y., Toyoshima, T., Motoyoshi, Y. & Arita, K. (1982). Petrological constitution of the continental type crust upthrust in the Hidaka belt, Hokkaido. Journal of Mineralogy, Petrology and Economic Geology (Special Issue) 229–238 (in Japanese with English abstract).
Kornprobst, J., Piboule, M., Roden, M. & Tabit, A. (1990). Corundum-bearing garnet clinopyroxenites at Beni Bousera (Morocco): original plagioclase-rich gabbros recrystallized at depth within the mantle? Journal of Petrology 31, 717–745.
Luth, W. C. & Ingamells, O. (1965). Gel preparation of starting materials for hydrothermal experimentation. American Mineralogist 50, 255–258.[Web of Science]
Matsumoto, T., Chen, Y. & Matsuda, J. (2001). Concomitant occurrence of primordial and recycled noble gases in the Earth's mantle. Earth and Planetary Science Letters 185, 35–47.[CrossRef][Web of Science]
McDonough, W. F. & Sun, S.-S. (1995). The composition of the Earth. Chemical Geology 120, 223–253.[CrossRef][Web of Science]
Morishita, T. & Arai, S. (1997). Diversity of occurrence of symplectite in the Horoman peridotite complex of the Hidaka belt, Hokkaido, northern Japan, and its bearing on the P–T history. Memoirs of the Geological Society of Japan 47, 149–162 (in Japanese with English abstract).
Morishita, T. & Arai, S. (2001). Petrogenesis of corundum-bearing mafic rock in the Horoman Peridotite Complex, Japan. Journal of Petrology 42, 1279–1299.
Morishita, T. & Kodera, T. (1998). Finding of corundum-bearing gabbro boulder possibly derived from the Horoman Peridotite Complex, Hokkaido, northern Japan. Journal of Mineralogy, Petrology and Economic Geology 93, 52–63.[CrossRef]
Morishita, T., Arai, S. & Gervilla, F. (2001). High-pressure aluminous mafic rocks from the Ronda peridotite massif, southern Spain: significance of sapphirine- and corundum-bearing mineral assemblages. Lithos 57, 143–161.[CrossRef][Web of Science]
Morishita, T., Arai, S. & Green, D. H. (2003a). Evolution of low-Al orthopyroxene in the Horoman peridotite, Japan: an unusual indicator of metasomatising fluids. Journal of Petrology 44, 1237–1246.
Morishita, T., Arai, S., Gervilla, F. & Green, D. H. (2003b). Closed-system geochemical recycling of crustal materials in the upper mantle. Geochimica et Cosmochimica Acta 67, 303–310.[CrossRef][Web of Science]
Niida, K. (1984). Petrology of the Horoman ultramafic rocks. Journal of Faculty of Sciences, Hokkaido University, Series IV 21, 61–81.
Obata, M. & Nagahara, N. (1987). Layering of alpine-type peridotite and the segregation of partial melt in the upper mantle. Journal of Geophysical Research 92, 3467–3474.
Ozawa, K. & Takahashi, N. (1995). P–T history of a mantle diapir: the Horoman peridotite complex, Hokkaido, northern Japan. Contributions to Mineralogy and Petrology 120, 223–248.[Web of Science]
Polvé, M. & Allègre, C. J. (1980). Orogenic lherzolite complexes studied by 87Rb–87Sr: a clue to understand the mantle convection processes? Earth and Planetary Science Letters 51, 71–93.[CrossRef][Web of Science]
Råheim, A. & Green, D. H. (1974). Experimental petrology of lunar highland basalt composition and applications to models for the lunar interior. Journal of Geology 82, 607–622.[Web of Science]
Sawaguchi, T. (2001). Structural petrology of the Horoman peridotite: deformation history and exhumation process of the Alpine-type peridotite massif. Ph.D. dissertation, Waseda University, 193 pp.
Schairer, J. F. & Bowen, N. L. (1955). The system K2O–Al2O3–SiO2. American Journal of Science 253, 681–746.
Shiotani, Y. & Niida, K. (1997). Origin of mafic layers in the Horoman peridotite complex. Memoirs of the Geological Society of Japan 47, 123–137 (in Japanese with English abstract).
Sobolev, V. V., Hofmann, A. W. & Nikogosian, I. K. (2000). Recycled oceanic crust observed in ‘ghost plagioclase’ within the source of Mauna Loa lavas. Nature 404, 986–989.[CrossRef][Medline]
Sutherland, F. L., Schwarz, D., Jobbins, E. A., Coenraads, R. R. & Webb, G. (1998). Distinctive gem corundum suites from discrete basalt fields: a comparative study of Barrington, Australia, and West Pailin, Cambodia, Gemfield. Journal of Gemmology 26, 65–85.
Sylvester, P. J. (2001). Trace element analysis of fused whole rock glasses by laser ablation ICPMS. In: Sylvester, P. (ed.) Laser-Ablation–ICPMS in the Earth Sciences, Principles and Applications. Ottawa, Ont.: Mineralogical Association of Canada, pp. 147–162.
Takahashi, E., Nakajima, K. & Wright, T. L. (1998). Origin of the Columbia River basalts: melting model of a heterogeneous plume head. Earth and Planetary Science Letters 162, 63–80.[CrossRef][Web of Science]
Takahashi, N. (1991). Origin of three peridotite suites from Horoman peridotite complex, Hokkaido, Japan; melting, melt segregation and solidification processes in the upper mantle. Journal of Mineralogy, Petrology and Economic Geology 86, 199–215.
Takahashi, N. (1992). Evidence for melt segregation towards fractures in the Horoman mantle peridotite complex. Nature 359, 52–55.[CrossRef]
Takahashi, N. (1997). Melt segregation processes observed in the Horoman mantle diapir. Memoirs of the Geological Society of Japan 47, 87–105.
Takahashi, N. (2001). Origin of plagioclase lherzolite from the Nikanbetsu Peridotite Complex, Hokkaido, northern Japan: implications for incipient melt migration and segregation in the partially molten upper mantle. Journal of Petrology 42, 39–54.
Takazawa, E., Frey, F. A., Shimizu, N. & Obata, M. (1992). Geochemical evidence for melt migration and reaction in the upper mantle. Nature 359, 55–58.[CrossRef]
Takazawa, E., Frey, F. A., Shimizu, N. & Obata, M. (1996). Evolution of the Horoman peridotite (Hokkaido, Japan): implications from pyroxene compositions. Chemical Geology 134, 3–26.[CrossRef][Web of Science]
Takazawa, E., Frey, F. A., Shimizu, N., Saal, A. & Obata, M. (1999). Polybaric origin of mafic layers in the Horoman Peridotite Complex, Japan. Journal of Petrology 40, 1827–1851.[CrossRef][Web of Science]
Takazawa, E., Frey, F. A., Shimizu, N. & Obata, M. (2000). Whole-rock compositional variations in an upper mantle peridotite (Horoman, Hokkaido, Japan): implications for melt segregation, migration and reaction. Geochimica et Cosmochimica Acta 64, 695–716.[CrossRef][Web of Science]
Yaxley, G. & Green, D. H. (1998). Reactions between eclogite and peridotite: mantle refertilisation by subduction of oceanic crust. Schweizerische Mineralogische und Petrographische Mitteilungen 78, 243–255.
Yoshida, H. & Takahashi, N. (1997). Chemical behavior of major and trace elements in the Horoman mantle diapir, Hidaka belt, Hokkaido, Japan. Journal of Mineralogy, Petrology and Economic Geology 92, 391–409 (in Japanese with English abstract).[CrossRef]
Yoshikawa, M. & Nakamura, E. (2000). Geochemical evolution of the Horoman Peridotite Complex: implications for melt extraction, metasomatism and compositional layering in the mantle. Journal of Geophysical Research 205, 2879–2901.
Yoshikawa, M., Nakamura, E. & Takahashi, N. (1993). Rb–Sr isotope systematics in a phlogopite-bearing spinel lherzolite and its implications for age and origin of metasomatism in the Horoman peridotite complex, Hokkaido, Japan. Journal of Mineralogy, Petrology and Economic Geology 88, 121–130.
Zimmer, M., Kröner, A., Jochum, K. P., Reischmann, T. & Todt, W. (1995). The Gabal Gerf complex: a Precambrian N-MORB ophiolite in the Nubian Shield, NE Africa. Chemical Geology 123, 29–51.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. FEENSTRA, S. SAMANN, and B. WUNDER An Experimental Study of Fe-Al Solubility in the System Corundum-Hematite up to 40 kbar and 1300{degrees}C J. Petrology, September 1, 2005; 46(9): 1881 - 1892. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||