| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology Volume 42 Number 1 Pages 39-54 2001
© Oxford University Press 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
DEPARTMENT OF EARTH SCIENCES, FACULTY OF SCIENCE, CHIBA UNIVERSITY, 1-33 YAYOI-CHO, INAGE-KU, CHIBA 263-8522, JAPAN
Received November 16, 1999; Revised typescript accepted June 23, 2000
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL BACKGROUNDGEOLOGICAL AND PETROGRAPHIC...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
The Nikanbetsu peridotite complex, Hokkaido, Japan, is composed of mainly fertile lherzolite, which shows several lines of evidence for incipient partial melting in the spinel–plagioclase facies. There are petrological, textural and mineral chemical variations in plagioclase-free and -bearing lherzolites from the base to the top of the complex within the total thickness of 1400 m. Two-pyroxene and spinel symplectites occur only at the base of the complex. Mass-balance calculations on their bulk compositions suggest that they lost the Al component from pyropic garnet. The Wo content of orthopyroxene cores continuously increases, whereas the Al content decreases from the base upward. Ca–Na zoning patterns of plagioclase in the plagioclase lherzolites characteristically change from W-shaped patterns at the base to oscillatory patterns in the upper part of the complex. These lines of petrological, textural and mineral chemical evidence indicate that incipient partial melting occurred everywhere in the complex, with an increase in the degree of melting from the base toward the top, in proportion to a monotonous rise of the equilibrium temperature from 1100°C to 1250°C. The systematics of plagioclase zoning provides evidence for simultaneous incipient partial melting, melt migration, decompression and melt crystallization in the ascending upper-mantle rocks.
KEY WORDS: melt migration; oscillatory zoning; partial melting; plagioclase lherzolite; symplectite
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDGEOLOGICAL AND PETROGRAPHIC...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Incipient melting and melt migration are important processes in the upper mantle for the generation of primary magma. Samples of primary melt from the upper mantle are found as melt (glass) inclusions in mantle xenoliths or mantle-derived minerals, such as olivine (e.g. Schiano et al., 1995
). These melt samples usually show a wide range of compositions in term of major and trace elements, which may be caused by several processes during melt migration, such as mixing between melt fractions and/or reaction with wall-rock peridotites (Kamenetsky et al., 1998). In addition to melt inclusions, direct information on mantle melting processes can be obtained through detailed investigation of tectonically emplaced peridotites, because they record various processes that occurred from their first formation to their final ascent to the Earth’s surface.
The Horoman complex has been well investigated by many workers (Niida, 1984; Frey et al., 1991; Takahashi, 1991a, 1991b, 1992; Takazawa et al., 1992, 1999; Ozawa & Takahashi, 1995; Morishita & Arai, 1999), and is regarded as an example of upwelling mantle materials from depth within the garnet stability field or deeper. Ozawa & Takahashi (1995) revealed a diapir structure in the Horoman complex from detailed investigations of P–T history of plagioclase lherzolite samples from the bottom to the top of the complex. They described petrological, mineralogical, and textural variations from the Lower Zone to the Upper Zone, whose division was defined by Niida (1974, 1984). They concluded that the essential differences between the two zones are caused by different final upwelling P–T paths from the garnet stability field. The Lower Zone is inferred to have taken a lower-temperature trajectory probably corresponding to the outer part of the diapir, and the Upper Zone to have taken the higher-temperature trajectory, probably corresponding to the inner part of the diapir. A nearly adiabatic decompression path for the Upper Zone suggests that the rocks passed just below the dry solidus. However, some petrological evidence, the occurrence and chemical composition of isolated plagioclase grains and plagioclase-rich segregation veins, indicates that incipient melting occurred in the spinel–plagioclase stability field for the Upper Zone (Takahashi, 1997).
Unlike the Horoman complex, little research has been carried out on the Nikanbetsu complex, which is located 9 km to the east of the Horoman complex. Because of its dominantly plagioclase lherzolite character, the Nikanbetsu complex has been regarded as a fragment of the Upper Zone of the Horoman complex (Niida & Katoh, 1978; Komatsu et al., 1982). Recently, Takahashi (1997) suggested that the Nikanbetsu complex might represent a much hotter portion of the Horoman diapir, based on textural and chemical characteristics of common plagioclase-rich segregation veins and isolated plagioclase grains. Important information about the generation of incipient melt and melt migration processes in a mantle diapir can therefore be obtained through the investigation of these two complexes.
To clarify this point, it is necessary to compare the Horoman and Nikanbetsu complexes on the basis of the textural and mineral chemical characteristics of the plagioclase lherzolite in detail. In this paper, the geological and petrographical contrasts between the Horoman and Nikanbetsu complexes are summarized first. Data on bulk compositions and the results of mass-balance calculations on two-pyroxene and spinel symplectitic aggregates from the Horoman and Nikanbetsu complexes are then presented and the origin of the symplectite from the Nikanbetsu complex is discussed to examine the hypothesis that it represents a garnet pseudomorph, modified by low-degree partial melting. Additionally, the textures and mineral compositions within the Nikanbetsu peridotites are described to clarify their origin. Lastly, several types of chemical zoning in plagioclase from the Nikanbetsu plagioclase lherzolite are described in detail and are discussed in the light of incipient melting and melt migration processes in the upper mantle.
|
GEOLOGICAL BACKGROUND |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDGEOLOGICAL AND PETROGRAPHIC...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
The Horoman and Nikanbetsu peridotite complexes, Hokkaido, Japan, are located at the southern end of the Hidaka belt, which consists of two zones, that is, the Main Zone and the Western Zone, and has metamorphic and igneous rocks of Paleogene to early Miocene ages (e.g. Saeki et al., 1991
). Both complexes belong to the Main Zone and occupy the tectonically lowest mantle horizon of the Hidaka belt (Komatsu et al., 1989) (Fig. 1a). The peridotite complexes are in fault contact with mylonitized granulite facies crustal rocks, such as gneiss, schist and/or metagabbro.
|
|
GEOLOGICAL AND PETROGRAPHIC CONTRASTS BETWEEN THE HOROMAN AND NIKANBETSU COMPLEXES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDGEOLOGICAL AND PETROGRAPHIC...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Horoman complex
The Horoman complex is
8 km x 10 km in size and is stratified with a total thickness of 3000 m. Niida (1984) divided the complex into Lower Zone and Upper Zone on the basis of the structure of layer boundaries and abundance of gabbro layers. Sharp layer boundaries and the occurrence of gabbro layers are more common in the Upper Zone than in the Lower Zone. Ozawa & Takahashi (1995) concluded that the differences between the two zones were caused by different final P–T upwelling paths; the path was higher in temperature for the Upper Zone than for the Lower Zone. However, there is a transition zone between them in terms of texture as well as P–T history (Ozawa & Takahashi, 1995; Takahashi, 1997). The peridotites of the Horoman complex can be divided into a dominant Main Harzburgite–Lherzolite suite (MHL) of residual facies, a subordinate Spinel-rich Dunite–Wehrlite suite (SDW) of cumulus facies from a magma derived from the MHL, and a minor Banded Dunite–Harzburgite suite, which occurs only in the Upper Zone, considered to be of cumulus origin from a different magma of high-Mg character (Takahashi, 1991a, 1991b, 1992). The MHL is composed of harzburgite, symplectite-free lherzolite, symplectite-bearing lherzolite and plagioclase lherzolite, which show gradual changes in terms of modal and bulk chemical compositions. The modes of clinopyroxene and orthopyroxene, bulk Al2O3 and CaO contents decrease from plagioclase lherzolite through symplectite-bearing lherzolite, symplectite-free lherzolite to harzburgite (Obata & Nagahara, 1987; Takazawa et al., 1992; Yoshida & Takahashi, 1997). There are several differences in the MHL peridotites between the Lower Zone and the Upper Zone, with respect to texture, grain size and mineral assemblage of thin plagioclase-bearing seams (0·5–0·8 mm in thickness), which are juxtaposed parallel to the foliation, from plagioclase lherzolite. For instance, the grain size of plagioclase in plagioclase-bearing seams gradually increases from 0·10 mm at the bottom of the Lower Zone to 0·23 mm at the top of the Upper Zone (Takahashi, 1997) (Fig. 2b–d).
|
Lower Zone
In the 2000 m thick Lower Zone, symplectite-bearing and -free lherzolites and harzburgite predominate over plagioclase lherzolite (Ozawa & Takahashi, 1995
; Takazawa et al., 1999). All peridotites are typically characterized by mylonitic to porphyroclastic textures with large orthopyroxene and clinopyroxene porphyroclasts (Sawaguchi & Takagi, 1997; Takahashi, 1997). Elliptical two-pyroxene and spinel symplectites surrounded by fine-grained aggregates commonly occur in the lherzolites (Takahashi & Arai, 1989) (Fig. 2a). On crossing the boundary with the plagioclase lherzolite, a thin film of plagioclase rimming spinel grains can be observed in some symplectitic aggregates and in the surrounding fine-grained aggregates (Takahashi & Arai, 1989; Ozawa & Takahashi, 1995). Near the boundary with the symplectite-bearing lherzolite, the plagioclase lherzolite contains relic symplectitic aggregates in seam-like fine-grained aggregates (Fig. 2b). The mode of occurrence of plagioclase (0·1–9·6% modal) in the plagioclase lherzolite is restricted to the thin seams (Fig. 2c). These plagioclase-bearing seams usually consist of plagioclase, olivine, orthopyroxene and spinel, and do not contain clinopyroxene (Takahashi, 1997). They are interpreted as a deformed subsolidus reaction product among two pyroxenes and spinel which were derived from decompression reactions between garnet and olivine (Ozawa & Takahashi, 1995) (Fig. 2a–c). Thick cumulus dunite layers of the SDW (up to 15 m) commonly occur within residual harzburgite. Gabbro is rare.
Upper Zone
The Upper Zone, with a total thickness of up to 1000 m, is dominated by plagioclase lherzolite (Ozawa & Takahashi, 1995; Takazawa et al., 1999). The volume ratio of lherzolite is lower than in the Lower Zone. Two-pyroxene and spinel symplectites rarely occur in lherzolite and the grain size of the constituent minerals of the symplectite is coarser than in the Lower Zone. The peridotites are characterized by tabular equigranular to equigranular textures with minor smaller porphyroclastic grains of pyroxenes. Three types of plagioclase (4·5–11·0% modal) can be observed in plagioclase lherzolite from the Upper Zone (Takahashi, 1997). The first type of plagioclase occurs as one of the minerals forming a seam-like aggregate, parallel to the foliation (Fig. 2d), which always contains a significant amount of clinopyroxene (to 11·0% modal). The second type of plagioclase occurs outside the seam-like aggregate, that is, at the grain boundaries among pyroxenes and olivine as isolated grains ranging from 0·20 to 0·57 mm in size. Isolated plagioclase is often mantled by orthopyroxene films. The third type is plagioclase-rich veins cutting the host plagioclase lherzolite, and this type of plagioclase can often be observed in the Upper Zone except for its lowest portion (up to 120 m above the boundary with the Lower Zone). These veins are oriented oblique to the foliation at an angle of 25–35°, parallel to the lineation. They are locally fringed by thin orthopyroxenite zones and have plagioclase-absent depletion zones (tens of millimetres in thickness). A single thick dunite layer has never been observed and the cumulus dunite of the SDW always occurs as thin layers alternating with gabbro, which corresponds to GB II of Niida (1984) and Type II of Takazawa et al. (1999).
Nikanbetsu complex
The Nikanbetsu complex is 2 km x 1 km in size and its total thickness is estimated to be up to 1400 m (Takahashi, 1997). Throughout the complex, the foliation strikes N70°E–N86°W and dips 78°N–86°N, except for the northern part, where it dips steeply to the south (Figs 1 and 3). The stratigraphical base of the complex is exposed in the southwestern margin. The complex is mainly composed of plagioclase lherzolite, with up to 11% modal plagioclase. Small amounts of lherzolite and harzburgite, frequently alternating with gabbro dykes (several to tens of centimetres thick), occur in the plagioclase lherzolite.
|
Peridotites from the southwestern portion, corresponding to the lower part of the complex, show tabular equigranular to equigranular textures (Mercier & Nicolas, 1975), except in the marginal part of the complex where they show partly porphyroclastic texture. The plagioclase lherzolite from this area exhibits features similar to those of the Upper Zone of the Horoman complex, such as plagioclase-bearing seam-like aggregates that contain clinopyroxene, isolated plagioclase and spinel rimmed by plagioclase (Fig. 4a and b). Symplectitic fine-grained aggregates associated with plagioclase are observed only in the seam-like aggregates from the plagioclase lherzolite from the lower part of the complex.
|
Two-pyroxene and spinel symplectitic aggregates are very rare and I have found two symplectitic aggregates without or almost free of the coarser-grained margin (Fig. 5a and b), which are commonly observed in the Horoman complex, from the same lherzolite horizon near the base of the complex (Fig. 1). These symplectites are composed of relatively coarse-grained minerals and are similar to those from the Upper Zone of the Horoman complex (Fig. 5). The Nikanbetsu symplectitic aggregates have retained their spherical shape, which indicates that they have not experienced strong deformation as observed in the Horoman complex.
|
Peridotites from the central to upper part of the complex show protogranular texture with coarser grains than those from the top of the Upper Zone of the Horoman complex. Plagioclase occurs in a fashion similar to that of the Upper Zone of the Horoman complex; however, interstitial isolated plagioclase grains, which locally are in contact with each other, and plagioclase-rich veins oblique to the foliation (Fig. 3), which are usually several tens of centimetres long and 1 cm in thickness, are more common. In addition, plagioclase occurs in small monomineralic lenses (0·5 mm in size) or in lens-shaped intergrowths with pyroxenes. Interstitial plagioclase from the central to upper part of the complex often cuts and replaces primary coarse orthopyroxene grains (Fig. 4c and d). Coarse plagioclase grains in plagioclase-rich lenses from the top of the complex are cut by small amounts of interstitial calcic plagioclase. The plagioclase-rich veins, having a plagioclase-free zone of a few centimetres on both sides, are composed of mainly coarse plagioclase and orthopyroxene, olivine and clinopyroxene ± Ti-pargasite. Such plagioclase-rich veins are frequently found in several zones of the plagioclase lherzolite and are sometimes connected to each other to make a network structure (Figs 1 and 3). Gabbro dykes (several tens of centimetres in thickness) have irregular outlines and consist of plagioclase (44% modal), clinopyroxene (37% modal), olivine (17% modal) and orthopyroxene (2–3% modal), and accessory opaque minerals (1% modal). These gabbro dykes, which correspond to GB I of Niida (1984) and Type I of Takazawa et al. (1999), preserve magmatic textures.
The bulk-rock compositions of the Nikanbetsu and Horoman peridotites overlap, except for the depleted harzburgite member, which is lacking in the Nikanbetsu complex (Yoshida & Takahashi, 1997; N. Takahashi & H. Yoshida, unpublished data, 1999).
|
MINERAL CHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDGEOLOGICAL AND PETROGRAPHIC...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Minerals were analysed with a JEOL-8900 electron microprobe analyser at the Chemical Analysis Center of Chiba University. Analytical procedures were described by Nakamura & Kushiro (1970)
. ZAF correction procedures were used. Representative analyses are given in Table 1, together with detailed analytical procedures and mineral calculation assumptions.
|
Two-pyroxene and spinel symplectite
Bulk chemical compositions of the symplectite were determined using averaged mineral compositions and modal amounts of each constituent mineral of a given symplectite, which were obtained from the images of two-dimensional elemental mapping for Ca, Al and Fe (Table 2). For comparison with the Nikanbetsu symplectite, three samples from the Horoman complex were examined with the same method (Fig. 5). The bulk composition of Nikanbetsu symplectite is similar to those from the Upper Zone (88082301) and the top of the Lower Zone (L53) of the Horoman complex, but is different from that from near the bottom of the Lower Zone (H703) in terms of SiO2 content.
|
Garnet compositions were estimated using mass-balance calculations as described by Obata et al. (1997), which involve subtraction of an olivine component from the bulk composition of the symplectite until the residual composition reached garnet stoichiometry. This calculation is based on the assumption that the compositions reflect a mixture of garnet and olivine, if they have not experienced any chemical modification during or after garnet breakdown. The symplectite from near the bottom of the Horoman complex (H703), with low SiO2 content, reached a pyropic garnet composition with this method (Table 2). However, the other two samples of the Horoman (88082301 and L53) and one sample from the Nikanbetsu complex have Al contents too low to represent a pyropic garnet after the calculation.
Orthopyroxene
Orthopyroxene from the Nikanbetsu plagioclase lherzolite is characterized by a high concentration of Al and Ca in the core (>0·18/6 oxygens or 4·4 wt % Al2O3; >2·7 mol % Wo or >1·4 wt % CaO) and a monotonous or stepwise decrease in Al content from core to rim. An Al-poor area also appears along the margin protruded into by plagioclase. Orthopyroxene grains from the symplectite-bearing lherzolite also exhibit this type of zoning pattern. An M-shaped Al zoning pattern, which is common in the Lower Zone of the Horoman complex (Ozawa & Takahashi, 1995), has not been observed in the Nikanbetsu complex.
In Fig. 6, the average Al and Wo contents in the orthopyroxene core from the plagioclase lherzolite are plotted against the distance from the base of the Horoman (Ozawa & Takahashi, 1995) and Nikanbetsu complexes. The Al and Wo contents from the base of the Nikanbetsu complex are similar to those from the top of the Horoman complex. In the Nikanbetsu complex, the Wo content continuously increases from the base to the top, whereas Al content slightly decreases. The stratigraphic variation of Wo content indicates that the Nikanbetsu complex preserves a temperature gradient that is possibly continuous from the Horoman complex.
|
Spinel
Although spinel also occurs as isolated grains, most spinel is associated with plagioclase. Some spinel grains occurring in plagioclase lherzolite, which contains symplectitic aggregates, and in symplectite-bearing lherzolite from the southwestern portion, show Al–Cr zoning (Fig. 5c). It is characterized by monotonous decrease in cr-number [Cr/(Cr + Al + Fe3+] from 0·33 to 0·29 toward the rim. Spinel showing this type of zoning occurs only in the vicinity (up to 10 cm) of symplectitic aggregate with or without plagioclase. Except for these samples, spinel grains with plagioclase rinds are homogeneous in terms of cr-number. However, TiO2 content tends to vary from grain to grain (e.g. 0·2–0·5 wt %), even for grains with similar cr-number (0·31–0·32).
Spinel from plagioclase lherzolite has a large compositional variation in TiO2 content, from 0·04 to 0·6 wt % (Fig. 7). On the other hand, spinel grains from plagioclase-free peridotites show lower Ti concentrations. Spinel occurring in the symplectite exhibits the most Al-rich composition (cr-number = 0·16–0·19).
|
Olivine
Forsterite and NiO contents of olivine in some plagioclase lherzolite and plagioclase-free lherzolite and harzburgite from the Nikanbetsu complex plot within the Horoman trend, which is considered to be a trend for residual peridotites in the upper mantle (Takahashi et al., 1987; Takahashi, 1991a, 1991b, 1997) (Fig. 8). The forsterite content (Fo) is in the range 89·0–90·9 for plagioclase lherzolite, 90·9–91·1 for lherzolite and 91·2–91·8 for harzburgite. The NiO content ranges from 0·36 to 0·37 wt % for plagioclase lherzolite, and from 0·37 to 0·38 wt % for plagioclase-free lherzolite and harzburgite. However, olivine from some plagioclase lherzolites shows low NiO contents, ranging from 0·33 to 0·35 wt %. Olivine occurring in the wall of plagioclase-rich veins is characterized by low contents of Fo and NiO, which is on the Fe-rich extension of the trend for the Nikanbetsu plagioclase lherzolite (Fig. 8).
|
Plagioclase
Plagioclase in plagioclase lherzolite has obvious Na–Ca zoning, which is characterized by an increase in 100 x Ca/(Ca + Na + K) (An) toward the rim (Takahashi, 1997). Isolated plagioclase from near the base to the central part of the complex exhibits a W-shaped An zoning profile, which is characterized by a slight decrease in An content in the core and a subsequent monotonous increase in An content towards the rim (Takahashi, 1997) (Fig. 9a). In contrast, plagioclase from the central to upper part of the complex often shows oscillatory zoning patterns such as saw-toothed and step patterns (Fig. 9b–d). This type of zoning pattern in plagioclase has never been reported before in upper-mantle rocks. As shown in Fig. 9b and c, some plagioclase grains have a marked zoning pattern at their margin in addition to obvious oscillatory zoning in the core. They are characterized by a monotonous increase in An content from 65 to 77–80, followed by a decrease to 73 and finally a drastic increase again toward the rim, to 80. The An zoning profile for a coarse plagioclase grain cut by calcic interstitial plagioclase exhibits a similar pattern (Fig. 9e).
|
Plagioclase in gabbro dykes (type I) preserving magmatic textures often exhibits oscillatory zoning patterns that are similar to those from surrounding plagioclase lherzolites (Fig. 9f).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDGEOLOGICAL AND PETROGRAPHIC...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Origin of two-pyroxene and spinel symplectite from the Nikanbetsu complex
The origin of two-pyroxene and spinel symplectite from the Horoman complex can be explained as a decompressional reaction product between garnet and olivine (Takahashi & Arai, 1989
; Ozawa & Takahashi, 1995; Obata et al., 1997). Obata et al. (1997) examined symplectites from the middle of the Lower Zone (1600–1700 m above the base) and suggested that their bulk compositions reflect mixtures of garnet and olivine. The bulk of the Nikanbetsu symplectite, however, cannot have been generated by the process described by Obata et al. (1997), even though the symplectite is spherical (Fig. 5a and b), as the simple mass-balance approach as described above (Table 2) does not result in stoichiometric garnet. Symplectites from the uppermost level of the Lower Zone (L53) and the Upper Zone (88082301) of the Horoman complex, which are composed of coarser-grained minerals than the typical symplectite from the Lower Zone (H703), exhibit similar composition to that of the Nikanbetsu symplectite. The mismatch is primarily caused by their orthopyroxene-rich modal compositions (Table 2). In other words, they represent a depletion in spinel, that is, in the Al component, in comparison with the composition of a mixture of garnet and olivine.
Through detailed investigation of plagioclase-bearing seams in the Horoman complex, Takahashi (1997) concluded that incipient melting occurred in fertile plagioclase lherzolite from near the boundary of the Lower Zone to the top of the Upper Zone during final ascent of the complex. Accordingly, plagioclase-bearing seams from the Horoman Lower Zone do not have clinopyroxene, which indicates that subsolidus reaction among pyroxenes and spinel consumed all the clinopyroxene to produce olivine and plagioclase. Small amount of clinopyroxene appears in those seams near the boundary with the Lower Zone and the modal volume of clinopyroxene in the seams dramatically increases to 10%, with increasing distance from the boundary up-section, in spite of a higher-temperature trajectory for these peridotites. Therefore, plagioclase-bearing seams in the Horoman Upper Zone cannot be explained by only subsolidus reaction processes in a closed system, but melting and melt migration processes in an open system must also be considered. One possibility is that incipient melting reduced the Al component from the symplectites from the uppermost part of the Horoman Lower Zone and the Upper Zone.
As discussed below, the Nikanbetsu complex has possibly taken a higher-temperature path than the Horoman Upper Zone. This suggests that the Nikanbetsu complex could have experienced considerable modal and chemical modification by melting at higher temperatures during ascent of the peridotite, such as dissolution of the symplectite. This is supported by petrological evidence that Al–Cr zoning in spinel occurs only in the vicinity of the plagioclase-bearing or -free aggregates where symplectitic textures are preserved (Fig. 5c). If symplectite was formed by simple decompressional breakdown of garnet as observed in the Horoman Lower Zone, Cr-rich spinel was first produced in the outer portion of the symplectite and finally most Al-rich spinel was formed in the inner portion during the decompression process (Ozawa & Takahashi, 1995). Cr-rich spinel having an Al-enriched rim may be formed by migration of Al from the inner portion of the symplectite to the initial Cr-rich spinel occurring outside the symplectite during the low-degree melting and melt migration. The extent of Al enrichment is up to several centimetres, a phenomenon that has not been observed in the Horoman complex. This idea is also consistent with the fact that the M-shaped Al zoning profiles in orthopyroxene, suggesting nearly adiabatic subsolidus decompression through the garnet field (Ozawa & Takahashi, 1995), have not observed in the Nikanbetsu complex. Furthermore, the slight upward decrease in Al content in the core region of orthopyroxene throughout the Nikanbetsu complex (Fig. 6b) can be explained by the interpretation that relatively high degrees of melting reduced the Al content of the orthopyroxene, but relatively low degrees of melting did not. The reason why symplectites are very rare in the Nikanbetsu complex and occur only in the area near the base is that the basal part of the peridotite experienced the lowest degree of melting, as discussed below, and the remaining symplectites do not preserve garnet pseudomorph compositions, as a result of the dissolution of aluminous spinel and the redistribution of Al during melting.
Relationships between the Nikanbetsu and Horoman complexes
As discussed above, the occurrence of two-pyroxene and spinel symplectites at the base of the Nikanbetsu complex requires that the peridotites ascended from depth within the garnet stability field or deeper. Several lines of evidence from geological, petrological, textural and mineral chemical differences between the two complexes support the notion that the Nikanbetsu complex represents a hotter portion than the Horoman peridotites within a mantle diapir; incipient partial melting took place in the former. The Wo content of orthopyroxene cores in plagioclase lherzolites from the Nikanbetsu complex continuously increases from the base to the top (Fig. 6a). The minimum temperature is estimated to be higher than 1100°C at the base and 1250°C from the centre to the top, on the basis of the Wo content of the orthopyroxene (Wells, 1977; Lindsley, 1983). The Nikanbetsu complex records a continuous temperature gradient of >150°C within the total thickness of 1400 m. The range of the estimated minimum temperature, 1100–1250°C, for the Nikanbetsu complex is higher than for the Upper Zone of the Horoman complex, which ranges from 1100 to 1150°C (Ozawa & Takahashi, 1995).
Origin of the Nikanbetsu plagioclase lherzolite
The bulk-rock compositions suggest that the plagioclase lherzolite in the Nikanbetsu complex is similar to that of the Lower Zone of the Horoman complex, which was regarded as residual peridotite (Yoshida & Takahashi, 1997; N. Takahashi & H. Yoshida, unpublished data, 1999). The textural and chemical features of plagioclase suggest that it is of magmatic origin and not subsolidus. Spinel grains with plagioclase rinds do not have a Cr-enriched rim, but are homogeneous in cr-number, which also indicates that the spinel was not a subsolidus product. Higher TiO2 contents and cr-number of spinels in plagioclase lherzolite than in lherzolite suggest that the spinel was crystallized or influenced by the trapped melt (Fig. 7). However, the variation of the TiO2 content of the spinel from the Nikanbetsu plagioclase lherzolite is not large when compared with the range in abyssal plagioclase-bearing peridotites (0·10–1·45 wt %; Dick & Bullen, 1984). Furthermore, they are characterized by restricted range of cr-number (0·23–0·36), in contrast to the wide range of the abyssal plagioclase peridotites (0·1–0·6). The spinel chemistry thus suggests that the Nikanbetsu plagioclase lherzolite is not simply a mixture of depleted peridotite and impregnated melt (e.g. Obata & Nagahara, 1987), but represents partially molten fertile peridotite, although small-scale incipient melt migration has occurred. The occurrence of olivine with low NiO content, ranging from 0·33 to 0·35 wt %, in the walls of segregation veins within the residual plagioclase lherzolite indicates that they were formed by low degrees of crystallization from the partial melt, which coexisted with the host peridotite.
The increase in An content towards the rim in plagioclase was interpreted to have occurred as a result of decompression processes (Ozawa & Takahashi, 1995). The occurrences of plagioclase showing obvious Ca–Na zoning patterns such as W-shaped and oscillatory patterns, which is a new observation in mantle rocks, constrain incipient partial melting and melt migration processes in the upper mantle. The preservation of these zoning patterns requires that incipient partial melting, melt migration, decompression and crystallization took place at the same time in the Nikanbetsu complex. These processes could have occurred simultaneously when the adiabatically ascending diapir had just passed the solidus. The different zoning patterns in the core should be caused by different types of melt–crystal geometry in the host peridotite (Fig. 10). Plagioclase showing the W-shaped pattern suggests that rapid decompression, which caused the marginal Ca enrichment, took place after the partial crystallization of a trapped partial melt. The W-shaped pattern can be observed mainly in the base of the complex, where the lowest estimated temperature was obtained, as discussed above. This also supports the idea that the relatively low-temperature condition at the base of the complex led to lower degrees of melting, which caused formation of more isolated melt pockets than in the rocks from higher in the complex (Fig. 10). The oscillatory zoning pattern indicates that incipient melting and small-scale melt migration occurred several times during decompressional ascent of the complex. There is a possibility that individual incipient melts had different compositions (Chalot-Prat & Arnold, 1999). This idea is supported by two lines of petrological evidence; one is the different TiO2 contents in spinels in a given sample and the other is the obvious marginal zoning of An content in plagioclase, ranging from 73 to 80 (Fig. 9b, c and e). This Ca-enriched melt may have formed and migrated into the almost frozen mantle materials at shallower levels.
|
Melt segregation processes in the upper mantle
The Nikanbetsu complex possibly represents the hottest, almost central portion of a mantle diapir, where significant partial melting took place. Two melt migration processes can be observed in the Nikanbetsu complex. One is grain boundary migration, which formed an oscillatory zoning pattern in plagioclase, as discussed above. The other is crack-transportation migration, which is recognized as oblique plagioclase-rich veins (Nicolas, 1986; Takahashi, 1997). Plagioclase demonstrates a characteristically systematic chemical variation within veins. The An content of the core of plagioclase decreases from the margin, where its value is similar to that of isolated plagioclase in the surrounding peridotite, to the centre, where it ranges from 60 to 54 (Takahashi, 1997). The observation of vein walls free of plagioclase and the chemical composition of plagioclase in the segregation veins suggest that cracks sucked partial melt from the surrounding partially molten peridotite (Boudier & Nicolas, 1977; Nicolas, 1986). The occurrence of the same pattern of plagioclase zoning in the Nikanbetsu gabbro (type I) suggests that the gabbro dykes represent larger equivalents of the veins and some type I gabbro was crystallized in the plagioclase facies, not only in the garnet facies (Takazawa et al., 1999). In the case of the Nikanbetsu complex, the depleted peridotites occur only around cracks, now represented by plagioclase-rich veins and gabbro dykes. These observations indicate that melt segregation processes proceeded mainly in the vicinity of fractures, although the host peridotite may have been melted by up to 10%, as estimated from the total volume of plagioclase and associated interstitial minerals.
|
ACKNOWLEDGEMENTS |
|---|
The author extends her sincere thanks to Professor S. Arai (Kanazawa) for his discussions and suggestions, and for critically reading the manuscript; and Professors N. Shimizu (WHOI) and K. Ozawa (Tokyo) for their discussions and suggestions. The critical comments by two anonymous reviewers were very helpful in improving the manuscript. This work was supported by the Japanese Society for the Promotion of Science (Nos. 09740390 and 11740290).
|
FOOTNOTES |
|---|
*Telephone: +81-43-290-2839. Fax: +81-43-290-2859. E-mail: natsuko{at}earth.s.chiba-u.ac.jp
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDGEOLOGICAL AND PETROGRAPHIC...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Boudier, F. & Nicolas, A. (1977). Structural controls on partial melting in the Lanzo peridotites. In: Dick, H. J. B. (ed.) Magma Genesis. Oregon, Department of Geology and Mineral Industries, Bulletin 96, 63–78.
Chalot-Prat, F. & Arnold, M. (1999). Immiscibility between calciocarbonatitic and silicate melts and related wall rock reactions in the upper mantle: a natural case study from Romanian mantle xenoliths. Lithos 46, 627–659.
Dick, H. J. B. & 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.[Web of Science]
Frey, F. A., Shimizu, N., Leinbach, A., Obata, M. & Takazawa, E. (1991). Compositional variation within the lower layered zone of the Horoman peridotite, Hokkaido, Japan: constraints on models for melt-segregation. Journal of Petrology (Special Issue) 211–227.
Kamenetsky, V. S., Eggins, S. M., Crawford, A. J., Green, D. H., Gasparon, M. & Falloon, T. J. (1998). Calcic melt inclusions in primitive olivine at 43° MAR: evidence for melt–rock reaction/melting involving clinopyroxene-rich lithologies during MORB generation. Earth and Planetary Science Letters 160, 115–132.[Web of Science]
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).
Komatsu, M., Miyashita, S. & Arita, I. (1986). Composition and structure of the Hidaka metamorphic belt, Hokkaido: historical reviews and present status. Monograph of Association for Geological Collaboration in Japan 31, 189–203 (in Japanese with English abstract).
Komatsu, M., Osanai, Y., Toyoshima, T. & Miyashita, S. (1989). Evolution of the Hidaka metamorphic belt, northern Japan. In: Daly, J. S., Cliff, R. A. & Yardley, B. W. D. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications 43, 487–498.
Lindsley, D. H. (1983). Pyroxene thermometry. American Mineralogist 68, 477–493.[Abstract]
Mercier, J. C. & Nicolas, A. (1975). Textures and fabrics of upper mantle peridotites as illustrated by basalt xenoliths. Journal of Petrology 16, 454–487.
Morishita, T. & Arai, S. (1999). Evolution of the Horoman peridotite complex and its implications for the origin of heterogeneous mantle. Ofioliti 24, 137–138.
Nakamura, Y. & Kushiro, I. (1970). Compositional relations of coexisting orthopyroxene, pigeonite and augite in a tholeiitic andesite from Hakon volcano. Contributions to Mineralogy and Petrology 26, 265–275.[Web of Science]
Nicolas, A. (1986). A melt extraction model based on structural studies in mantle peridotite. Journal of Petrology 27, 999–1022.
Niida, K. (1974). Structure of the Horoman ultramafic massif of the Hidaka metamorphic belt in Hokkaido, Japan. Journal of the Geological Society of Japan 80, 31–44.
Niida, K. (1984). Petrology of the Horoman ultramafic rocks in the Hidaka metamorphic belt, Hokkaido, Japan. Journal of Faculty of Science, Hokkaido University, Series IV 21, 185–194.
Niida, K. & Katoh, T. (1978). Ultramafic rocks in Hokkaido. Monograph of Association for Geological Collaboration in Japan 21, 61–81 (in Japanese with English abstract).
Obata, M. (1980). The Ronda peridotite: garnet-, spinel-, and plagioclase-lherzolite facies and the P–T trajectories of a high-temperature mantle intrusion. Journal of Petrology 21, 533–572.
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.
Obata, M., Morisita, R. & Tanaka, K. (1997). The microstructure of symplectite in the Horoman peridotite and its formation processes. Memoirs of the Geological Society of Japan 47, 163–171 (in Japanese with English abstract).
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.
Saeki, K., Shiba, M. & Itaya, T. (1991). K–Ar ages of metamorphic and igneous rocks from the southern Hidaka metamorphic belt. Journal of Mineralogy, Petrology and Economic Geology 86, 177–178.
Sawaguchi, T. & Takagi, H. (1997). Inverted ductile shear movement of the Horoman peridotite complex in the Hidaka metamorphic belt, Hokkaido, Japan. Memoirs of the Geological Society of Japan 47, 193–208.
Schiano, P., Clocchiatti, R., Shimizu, N., Maury, R. C., Jochum, K. P. & Hofmann, A. W. (1995). Hydrous, silica-rich melts in the sub-arc mantle and their relationship with erupted arc lavas. Nature 377, 595–600.
Takahashi, E., Uto, K. & Schilling, J. G. (1987). Primary magma compositions and Mg/Fe ratios of their mantle residues along Mid Atlantic Ridges 29°N to 73°N. Technical Report ISEI, Okayama University A-9, 1–12.
Takahashi, N. (1991a). Evolutional history of the uppermost mantle of an arc system: petrology of the Horoman peridotite massif, Japan. In: Peters, T., Nicolas, A. & Coleman, R. G. (eds) Ophiolite Genesis and Evolution of the Oceanic Lithosphere. Dordrecht: Kluwer Academic, pp. 195–205.
Takahashi, N. (1991b). 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.
Takahashi, N. (1997). Incipient melting of mantle peridotites observed in the Horoman and Nikanbetsu peridotite complexes, Hokkaido, northern Japan. Journal of Mineralogy, Petrology and Economic Geology 92, 1–24.
Takahashi, N. & Arai, S. (1989). Textural and chemical features of chromian spinel pyroxene symplectites in the Horoman peridotites, Hokkaido, Japan. Science Report of Institute of Geoscience, University of Tsukuba, Section B 10, 45–55.
Takazawa, E., Frey, F. A., Shimizu, N., Obata, M. & Bodinier, J. L. (1992). Geochemical evidence for melt migration and reaction in the upper mantle. Nature 359, 55–58.
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.
Wells, P. R. A. (1977). Pyroxene thermometry in simple and complex systems. Contributions to Mineralogy and Petrology 62, 129–139.[Web of Science]
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).
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. MORISHITA, S. ARAI, and D. H. GREEN Possible Non-melted Remnants of Subducted Lithosphere: Experimental and Geochemical Evidence from Corundum-Bearing Mafic Rocks in the Horoman Peridotite Complex, Japan J. Petrology, February 1, 2004; 45(2): 235 - 252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. OZAWA Thermal History of the Horoman Peridotite Complex: A Record of Thermal Perturbation in the Lithospheric Mantle J. Petrology, February 1, 2004; 45(2): 253 - 273. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||