| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 45 | Number 3 | Pages 635-668 | 2004
Journal of Petrology 45(3) © Oxford University Press 2004; all rights reserved.
Geochemical and Isotopic Heterogeneities along an Island Arc–Spreading Ridge Intersection: Evidence from the Lewis Hills, Bay of Islands Ophiolite, Newfoundland
1 MAX-PLANCK-INSTITUT FÜR CHEMIE, ABTEILUNG GEOCHEMIE, POSTFACH 3060, 55020 MAINZ, GERMANY
2 UNIVERSITÄT ZU KÖLN, INSTITUT FÜR MINERALOGIE UND GEOCHEMIE, ZÜLPICHER STR. 49B, 50674 KÖLN, GERMANY
RECEIVED JULY 11, 2001; ACCEPTED AUGUST 8, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
This study focuses on the origin of magma heterogeneity and the genesis of refractory, boninite-type magmas along an arc–ridge intersection, exposed in the Lewis Hills (Bay of Islands Ophiolite). The Lewis Hills contain the fossil fracture zone contact between a split island arc and its related marginal oceanic basin. Three types of intrusions, which are closely related to this narrow tectonic boundary, have been investigated. Parental melts in equilibrium with the ultramafic cumulates of the Pyroxenite Suite are inferred to have high MgO contents and low Al2O3, Na2O and TiO2 contents. The trace element signatures of these Pyroxenite Suite parental melts indicate a re-enriched, highly depleted source with
0·1 x mid-ocean ridge basalt (MORB) abundances of the heavy rare earth elements (HREE). Initial 
Nd values of the Pyroxenite Suite range from -1·5 to +0·6, which overlap those observed for the island arc. Furthermore, the Pyroxenite Suite parental melts bear strong similarities to boninite-type equilibrium melts from island arc-related pyroxenitic dykes and harzburgites. Basaltic dykes split into two groups. Group I dykes have 0·6 x MORB abundances of the HREE, and initial Nd values ranging from +5·4 to +7·5. Thus, they have a strong geochemical affinity with basalts derived from the marginal basin spreading ridge. Group II dykes have comparatively lower trace element abundances (0·3 x MORB abundances of HREE), and slightly lower initial Nd values (+5·4 to +5·9). The geochemical characteristics of the Group II dykes are transitional between those of Group I dykes and the Pyroxenite Suite parental melts. Cumulates from the Late Intrusion Suite are similarly transitional, with Nd values ranging from +2·9 to +4·6. We suggest that the magma heterogeneity observed in the Lewis Hills is due to the involvement of two compositionally distinct mantle sources, which are the sub-island lithospheric mantle and the asthenospheric marginal basin mantle. It is likely that the refractory, boninite-type parental melts of the Pyroxenite Suite result from remelting of the sub-arc lithospheric mantle at an arc–ridge intersection. Furthermore, it is suggested that the thermal-dynamic conditions of the transtensional transform fault have provided the prerequisite for generating magma heterogeneity, as a result of mixing relationships between arc-related and marginal basin-related magmas. KEY WORDS: Bay of Islands ophiolite; transform (arc)–ridge intersection; boninites; rare earth elements, Nd isotopes
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
Large modern arc–back-arc rift systems, such as the Mariana and Lau arc–back-arc system, expose a diversity of magma types. These heterogeneous magma compositions typically occur within a zone of several tens to hundreds of kilometres width, which marks the transition from the rifted arc to the back-arc lithosphere (e.g. Hawkins & Melchior, 1985
; Ewart et al., 1994; Fryer, 1995; Hawkins, 1995), and have often been linked to distinct tectonic and magmatic phases in the evolution of the arc–back-arc system. For example, modern boninites mostly occur in the fore-arc region of a subduction zone. For this reason, boninites have usually been related to a rather extraordinary phase of magmatism coupled with the initiation of subduction and fore-arc spreading (e.g. Hawkins et al., 1984; Pearce et al., 1984, 1992; Bloomer et al., 1995). Boninites have also been reported from back-arc basin locations. These back-arc basin related boninites are suggested to have been generated by shallow mantle melting during the initial stage of back-arc rifting as (1) ascending back-arc diapirs initiate contact melting of the overlying depleted hydrous arc lithosphere (Crawford et al., 1989) or (2) a back-arc spreading centre propagates into refractory hydrated (arc) crust (Falloon & Crawford, 1991; Falloon et al., 1992; Kamenetsky et al., 1997). In both these models the production of the boninites is linked to an ephemeral phase associated with the initial back-arc basin spreading. Several boninites have also been discovered in transform-related tectonic settings, such as the Troodos transform fault (Murton, 1989; Rogers et al., 1989), the northern Tonga ridge (Falloon et al., 1989; Falloon & Crawford, 1991; Danyushevsky et al., 1995) and the Hunter Ridge–Hunter fracture zone (Monzier et al., 1993; Sigurdsson et al., 1993). However, different mechanisms have been invoked for the origin of such transform-related boninites, suggesting that (1) rapid dilation and transtensional transform movements were the major tectonic factors in the formation of boninitic magmas (Murton, 1989; Rogers et al. 1989), or (2) the transform fault acted as a window, allowing the intrusion of a mantle plume—the source of the boninites—into the mantle wedge above a subduction zone (Danyushevsky et al., 1995).
The Lewis Hills (Bay of Islands Ophiolite, BOIO) expose a major tectonic boundary that represents the transition zone between an arc and its marginal oceanic basin. This contact has been interpreted as a fossil transform fault or fracture zone (Karson & Dewey, 1978; Karson, 1984). Several types of intrusions, distinguished by petrology and geochemistry, are closely related to this narrow, 4–5 km wide zone (Karson et al., 1983; Casey et al., 1985; Elthon et al., 1986; Smith & Elthon, 1988; Suhr & Cawood, 2001). We have studied three types of magmatic intrusions from this tectonic boundary. In terms of trace element and radiogenic isotope geochemistry these intrusions are highly heterogeneous, spanning the range from back-arc basin type to refractory, boninite-type magma compositions. The subject of this study is the petrogenesis of these intrusions and their relation to the specific setting of the fracture zone separating island arc from oceanic lithosphere, or alternatively, two oceanic spreading centres from one another. We suggest that the magma heterogeneity is due to the involvement of two very different mantle source regions: (1) asthenospheric depleted upper mantle, which gave rise to back-arc basin-type basalts (BABB); (2) lithospheric sub-island arc lithosphere, which generated refractory, boninite-like melts. Arc- and back-arc-derived melts were contemporaneous and have presumably mixed, giving rise to transitional geochemical characteristics. Following the model of Kurth et al. (1998), it is suggested that the transform (arc)–ridge intersection of Lewis Hills has provided the prerequisite for generating magma heterogeneity and the genesis of refractory, boninite-type melts within an overall back-arc basin setting. This model may also apply to various transform–ridge intersections where geochemically distinct lithospheres are juxtaposed along a fracture zone.
|
GEOLOGICAL SETTING AND PREVIOUS WORK |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
The BOIO (Fig. 1) is one of the best preserved and well-studied Caledonian ophiolites, and several models for its origin have been proposed (Williams, 1973
; Karson & Dewey, 1978; Malpas, 1979; Searle & Stevens, 1984; Casey et al., 1985; Elthon, 1991; Jenner et al., 1991; Cawood & Suhr, 1992). After nearly 30 years of intense study the tectonic evolution of the BOIO now appears to be well constrained. The BOIO is a supra-subduction zone ophiolite (Elthon, 1991; Jenner et al., 1991) composed of two significantly different lithospheres: the eastern, Bay of Islands (BOIC) marginal basin and the western, island arc-related Coastal Complex [CC; terminology following Karson & Dewey (1978)]. Geochemical and isotope studies on crustal and mantle rocks from the BOIC suggest an overall depleted mid-ocean ridge basalt (MORB)-type source with a weak slab influence (Suen et al., 1979; Richard & Allègre, 1980; Kharas-Khumbatta, 1988; Komor & Elthon, 1990; Elthon, 1991; Jenner et al., 1991; Batanova et al., 1998; Suhr & Batanova, 1998; Suhr et al., 1998). Initial Nd(t) values range from +6 to +9 (Jacobsen & Wasserburg, 1979; Jenner et al., 1991). In contrast, geochemical and isotope analyses of basalts and trondhjemites from the arc-related CC suggest a significant re-enrichment of the CC source by slab-derived components, as reflected in, for example, low initial Nd(t) values ranging from -1 to +1 (Jenner et al., 1991). Furthermore, the CC island arc was determined to be much older (505 +3/-2 Ma) than the BOIC marginal basin (484 ± 5 Ma and 486 ± 2 Ma; Dunning & Krogh, 1985; Jenner et al., 1991).
|
The southernmost massif of the BOIO, the deeply eroded Lewis Hills (Karson, 1977
, 1984; Karson & Dewey, 1978; Fig. 2), has a unique geological structure when compared with the northern massifs, as well as other ophiolites, as it contains the magmatic contact between the split CC island arc and its related BOIC marginal basin (Karson & Dewey, 1978; Cawood & Suhr, 1992; Kurth et al., 1998). Structurally, this contact has been interpreted as a fossil fracture zone (Karson & Dewey, 1978; Karson, 1984; Suhr & Cawood, 2001). According to both the tectonic model of Cawood & Suhr (1992) and geochemical evidence (Kurth et al., 1998), the Lewis Hills are the site where the spreading ridge of the back-arc basin related BOIC abuts older, arc-related lithosphere along a transform (arc)–ridge intersection (Fig. 3). Following structural arguments (Karson & Dewey, 1978; Karson, 1984; Suhr & Cawood, 2001) and the tectonic model cited above, splitting of the CC island arc occurred parallel to its long axis, in contrast to the rifting situation of the western Pacific region. One important implication of the transform (strike-slip) kinematic setting of the Lewis Hills is that the transition from the arc to the back-arc lithosphere occurs abruptly over a distance of 4–5 km, which is referred to as the Mount Barren Complex (Fig. 2).
|
|
Geological units of the Lewis Hills
The Lewis Hills can be roughly divided into three structural zones (Fig. 2).
Eastern Lewis Hills (ELH). The ELH shows strong petrological, geochemical, age and isotopic affinities to the northern BOIC exposures (Karson, 1977; Karson & Dewey, 1978; Sassen et al., 1997; Kurth et al., 1998; Sassen, 2000; Suhr & Cawood, 2001). Initial Nd values of gabbros from the ELH range from +7 to +8 (Sassen et al., 1997). Unlike the northern massifs, the ELH exposes two strikingly different mantle domains: the BOIC-related Wheelers Brook (WB) peridotite and the CC arc-related Springers Hill (SH) harzburgites (Fig. 2). The two mantle units can be clearly distinguished in terms of mantle structure, geochemistry and degree of depletion (Suhr et al., 1991; Batanova et al., 1998; Suhr & Batanova, 1998; Suhr & Edwards, 2000). Equilibrium melts of the SH harzburgite have a strong refractory signature (Suhr & Edwards, 2000). Furthermore, pyroxenitic dykes occurring within the SH harzburgite have been related to boninite-type parental melts (Edwards, 1995).
Western Lewis Hills (WLH). The WLH is mainly composed of undeformed hornblende gabbros, minor basaltic dykes and plagiogranites. The WLH also contains a large ultramafic exposure that has been reinterpreted as representing an uplifted arc-related basement (Fig. 2; Suhr & Cawood, 2001). Initial Nd values of plutonic rocks from the WLH range from -2 to +2, and its U–Pb age was determined to be 503·7 ± 3·2 Ma and 500·6 ± 2·0 Ma (Kurth et al., 1998). Thus, age and isotope data from the WLH confirm its genetic link to the northern exposures of the CC [equivalent to the Little Port Complex of Williams (1973); Fig. 1].
Mount Barren Complex (MBC). The 4–5 km wide transform assemblage of the MBC (Fig. 2) is structurally complex, reflecting a protracted history of transform-related deformation and magmatism (Karson & Dewey, 1978; Karson et al., 1983; Suhr & Cawood, 2001). Dominant amphibolite- to granulite-facies metagabbros have near-vertical, planar-deformed structures with shallow plunging lineations, suggesting strong strike-slip displacements with a tectonic slip direction normal to the spreading ridge of the BOIC (Karson, 1977; Karson & Dewey, 1978; Casey et al., 1983; Rosencrantz, 1983; Suhr, 1992). These rocks are interpreted as having been deformed in a transform fault. Numerous transform-parallel mafic to ultramafic plutonic bodies occur within or at the eastern boundary of the MBC (Karson & Dewey, 1978; Karson et al., 1983; Suhr & Cawood, 2001). These intrusions have been emplaced syn- to postkinematically and were attached to the transform assemblage along a highly transtensional transform fault (Suhr & Cawood, 2001). Abundant basaltic dykes have intruded the MBC. Based on geological and geochemical evidence, these have been divided into three suites (Casey et al., 1985; Elthon et al., 1986). Gabbros and metagabbros from the MBC have initial Nd values between +1·5 and +4·3 (Kurth et al., 1998).
Transform-parallel intrusives of the MBC
A prominent type of the intrusive plutonic rocks forms several north–south highly elongated, massive bodies, which are up to 1 km wide and at least 4·5 km long (Fig. 2). Lithologically, these rocks consist of mainly olivine–clinopyroxene–orthopyroxene-bearing assemblages that are characterized by the late appearance of plagioclase (Suhr & Cawood, 2001). Because of the massive occurrence of the pyroxenites, which locally grade into wehrlite, dunite, websterite and lherzolite, the rock suite has been referred to as the ‘Pyroxenite Suite’ (Suhr & Cawood, 2001). The pyroxenites are cut by basaltic dykes, and are generally undeformed, except in the southern Lewis Hills, where they are more deformed (Suhr & Cawood, 2001). Another type of plutonic intrusion forms a series of weakly deformed to undeformed bodies aligned along the eastern boundary of the MBC (the ‘Late Intrusion Suite’; Karson et al., 1983; Suhr & Cawood, 2001; Fig. 2). These are mafic to ultramafic in composition (e.g. gabbro, troctolite, wehrlite, dunite). The U–Pb apatite age (485 ± 1 Ma; Kurth et al., 1998) and Nd-isotope signature [Nd(485 Ma) = +7 to +8; Sassen et al., 1997] of gabbros from different bodies of the Late Intrusion Suite connect them to the BOIC source.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
In this study, samples were selected from: (1) three north–south elongate intrusions of the Pyroxenite Suite; (2) one intrusion of the Late Intrusion Suite; (3) basaltic dykes cutting the MBC and WLH (Fig. 2).
Petrographic descriptions
Pyroxenite Suite
The rocks of this study (samples 95-69 to 95-94 and L686, Table 1 ) are olivine clinopyroxenite, wehrlite, dunite and (olivine) websterite. The samples chosen for isotope analysis are three websterites (95-73, 95-75 and 95-93) and one clinopyroxenite (L686). The samples are undeformed to moderately deformed, lack plagioclase and contain only minor spinel. Microstructurally, the rocks generally have granular textures with interlocking anhedral mineral grains. In some cases, euhedral grains and magmatic twinning in clinopyroxene have been preserved. Some samples show only a weak undulatory extinction and there is no evidence for post-magmatic recrystallization. The absence of significant strain indicators suggests that the primary, magmatic texture of the rocks is still preserved. The size of the grains varies from 1 mm to a few centimetres. All samples have undergone alteration during greenschist-facies metamorphism. Alteration is sometimes pervasive, involving the formation of amphibole and chlorite at clinopyroxene grain boundaries and inside clinopyroxene grains, and the serpentinization of olivine.
|
|
Late Intrusion Suite
The samples (95-42, 95-45 and 95-107) belong to a mostly undeformed plagioclase wehrlitic intrusion. The dominant rock type in the interior parts of the intrusion is a poikilitic plagioclase wehrlite. At the margins, lithologies become more heterogeneous and include wehrlites, dunites, pyroxenites and gabbros, some of which cross-cut each other. These marginal rocks are more deformed than the rocks from the central part. Petrographically, the three samples selected for analysis are an olivine-bearing gabbro (95-42), an olivine-bearing clinopyroxenite (95-107) and a clinopyroxenite (95-45). In two of the samples (95-45 and 95-107) the clinopyroxenes have been partially replaced by green hornblende during static metamorphism. At least two of the three samples have been plastically deformed to variable extents, as reflected in the formation of lenticular clusters of plagioclase or clinopyroxene neoblasts. However, an original magmatic, poikilitic fabric is still preserved. Rare, interlocking grain boundaries and magmatic twins in clinopyroxene have been preserved.
Basaltic dykes of the Mount Barren Complex and Western Lewis Hills
The majority of the basaltic dyke samples stem from the MBC, although some were also obtained from the WLH. The basalts from the MBC can be divided into deformed, pre- or syntectonic basalts (hornblende– plagioclase-bearing assemblages: M11, M13, M32, M43, M53, M71, M76 and M1232), and undeformed, post-tectonic basalts (clinopyroxene–plagioclase assemblages: M7, M40, M1169 and M1170). The undeformed basaltic dykes of the WLH have clinopyroxene–plagioclase–olivine–hornblende-bearing assemblages (WL1, WL3, WL5, WL9, WL19, WL59 and WL64). Microstructurally, the undeformed dykes of the MBC and WLH are generally fine grained with granular to ophitic textures. The dykes from the WLH contain the mineral assemblage clinopyroxene + plagioclase ± olivine ± amphibole ± opaque minerals. The brown–reddish hornblendes of the WLH dykes are presumably of magmatic origin. Olivine and magmatic amphiboles are absent in the undeformed dykes of the MBC, which are composed of clinopyroxene and plagioclase. Some of the dykes contain clusters of phenocrystic olivine + plagioclase (WL1 and WL3) or isolated phenocrysts of plagioclase and clinopyroxenes (M7, M40, M1169 and M1170). The primary mineral assemblage of the undeformed samples is locally replaced by greenschist-facies minerals. The deformed, syntectonic dykes of the MBC consist of green hornblende and plagioclase. No primary minerals have been found. The hornblendes are elongated, and plagioclases show a preferred lattice orientation, as reflected in their uniform extinction, indicating that the metamorphism of the basalts occurred syntectonically. All of the dykes have been affected to varying extents by static greenschist-facies metamorphism.
Relative time of emplacement of the intrusive rocks
The deformed basaltic dykes of the MBC are pre- to synkinematic intrusions that are interpreted to have formed prior to or during active transform deformation (Karson & Dewey, 1978). Because active deformation within a transform fault finishes at the ridge–transform intersection, the largely undeformed samples from the Pyroxenite Suite and Late Intrusion Suite, together with the undeformed basaltic dykes of the MBC, are interpreted to have been generated near the BOIC ridge–transform intersection (Karson & Dewey, 1978; Karson et al., 1983; Suhr & Cawood, 2001). The age of a Late Intrusion Suite gabbro (485 Ma) is identical to the age determined for the northern massifs of the BOIC (485 Ma). This implies that the emplacement of the low-strain Late Intrusion Suite was contemporaneous with the main constructional phase of the BOIC spreading ridge. Based on geological relationships, the emplacement of the Pyroxenite Suite and Late Intrusion Suite overlapped in time (Suhr & Cawood, 2001). The timing of dyke emplacement within the generally undeformed rock units of the WLH is rather difficult to determine. However, the dykes sampled can be traced from the arc-related hornblende gabbros to the tectonically uplifted mantle unit, which indicates that all these dykes were emplaced after the tectonic uplift of the mantle. As the tectonic ascent of the mantle unit is thought to be associated with the arc rifting and opening of the BOIC marginal basin (Suhr & Cawood, 2001), the ascent of the mantle unit (and also the emplacement of the dykes) may have occurred simultaneously with, or subsequently to, the deformation of the MBC.
Mineral and whole-rock chemistry
Whole-rock and mineral data obtained from the Pyroxenite Suite and the Late Intrusion Suite samples are presented in Tables 1 and 2.
|
|
Mineral chemistry of Pyroxenite Suite and Late Intrusion Suite
Olivines from the Pyroxenite and Late Intrusion Suites have Mg number [= 100Mg/(Mg + Fe)] ranging from 75 to 89, and NiO between 0·09 and 0·23% (Fig. 4a). For comparison, data are shown for pyroxenitic dykes from the arc-related Springers Hill (SH) mantle section (Edwards, 1991
). These latter dykes have been interpreted as early precipitates of highly refractory, possibly boninitic, melts (Edwards, 1995). Their pyroxenitic and plagioclase-free lithologies are very similar to those of the Pyroxenite Suite and are strikingly different from those of the vast majority of the BOIC magmatic rocks. This petrological similarity possibly indicates a genetic relationship between the SH dykes and the Pyroxenite Suite. The fields for olivines from present-day mid-ocean ridges (MOR) and BOIC cumulates are also shown for comparison. The NiO content of the Pyroxenite Suite and Late Intrusion Suite olivines varies little with decreasing Mg number, consistent with pyroxene-dominated crystallization. The olivines from the Pyroxenite Suite have relatively low Mg numbers and low NiO contents when compared with those of the SH dykes. Given an assumed genetic relationship between the SH dykes and Pyroxenite Suite, the parental melts of the SH dykes could represent the more primitive precursor for the more fractionated parental melts of the Pyroxenite Suite samples.
|
Figure 4b–d shows plots of the Na2O (0·01–0·18%), Al2O3 (1·1–3·8%) and TiO2 contents (0·06–0·14%) of clinopyroxenes versus their respective Mg number. The clinopyroxenes are compared with those from present-day MOR and BOIC cumulates, the SH dykes and the so-called BOIC refractory group, which are shown in the various fields. The BOIC refractory group comprises clinopyroxenes from ultramafic cumulates and pyroxenitic dykes, which occur locally in the northern massifs of the BOIC. These refractory cumulates and dykes derive from parental melts that are compositionally different from MOR or BOIC cumulates and have been related to refractory, or boninite-type magmas (Bédard, 1991
, 1994; Edwards, 1995; Bédard & Hébert, 1996; Varvalvy et al., 1996, 1997; Suhr et al., 1998). In all three plots, clinopyroxenes from samples of the Pyroxenite and Late Intrusion Suites fall entirely in the field for the BOIC refractory group, which possess some of the lowest Na2O, TiO2 and Al2O3 concentrations reported from BOIC clinopyoxenes (Fig. 4b–d). The clinopyroxenes from the SH dykes have lower Al2O3 and TiO2 concentrations than those of the Pyroxenite Suite (Fig. 4c and d). This observation is consistent with the assumption—derived from the olivine data—that the samples from the Pyroxenite Suite are evolved to some extent. Another striking feature is that Al2O3 in clinopyroxenes from the Pyroxenite and Late Intrusion Suites strongly increases as Mg number decreases (Fig. 4d). Such strong Al2O3 enrichment cannot be explained by simple fractionation of a single parental melt. The enrichment factor, which may be defined as the quotient of Al2O3 between the most and the least fractionated sample, is 3·4 (3·77% and 1·12% Al2O3). Even if Al2O3 was perfectly incompatible in the fractionating assemblage, this factor requires more than 50% fractionation of the original liquid. This amount of fractionation is clearly inconsistent with other data, especially with the FeO–MgO variation.
Figure 5 shows the chondrite-normalized rare earth element (REE) patterns for clinopyroxenes from three samples of the Pyroxenite Suite. Their low REE abundances distinguish them clearly from the clinopyroxenes from ELH cumulates, which are shown in the shaded field (Sassen, 2000), and clinopyroxenes in equilibrium with average MORB.
|
Whole-rock trace element data and parental melts for Pyroxenite and Late Intrusion Suites
Chondrite-normalized REE patterns for whole rocks are displayed in Fig. 6a. Whole-rock analyses of the Pyroxenite Suite samples have light REE (LREE)-depleted patterns with flat heavy REE (HREE). These patterns are similar to those of the clinopyroxenes from the same samples (Fig. 5), which directly reflect the clinopyroxene-rich mode of these rocks. REE patterns of two Late Intrusion Suite samples resemble those of the Pyroxenite Suite samples, and mirror the large amount of modal clinoproxene (60% and 90%) in these rocks. In contrast, the third Late Intrusion Suite sample (olivine gabbro, sample 95-42) contains only 10% modal clinopyroxene, which is the main host for the REE in this rock. Therefore, the whole-rock REE abundances are very low. The REE pattern is relatively flat and shows a marked positive Eu anomaly, which is due to the high plagioclase content (19%). As the plagioclase/liquid partition coefficients for the LREE are relatively high (Phinney & Morrison, 1990
), plagioclase contributes significantly to the LREE budget of such an olivine-rich rock (70%).
|
Chondrite-normalized REE patterns of calculated equilibrium melts are presented for Pyroxenite Suite (Fig. 6b) and Late Intrusion Suite samples (Fig. 6c), along with those of BOIC basalts, equilibrium melts from the arc-related mantle units (SH–WL harzburgite) of the Lewis Hills and equilibrium melts for the SH boninite-type pyroxenitic dykes from Edwards (1995)
. To estimate the trace element characteristics of parental melts for the samples, equilibrium melts were calculated using whole-rock analyses, combined with the mineral modes (Table 1 ) and a set of crystal/melt partition coefficients (Table 4). For the Pyroxenite Suite samples, we also used clinopyroxene analyses, together with their corresponding clinopyroxene/liquid partition coefficients (e.g. Cawthorn, 1996; Kelemen et al., 1997; Ross & Elthon, 1997; Table 3) to calculate equilibrium melts.
|
|
A critical problem in using minerals and/or cumulate rocks to make inferences about melt compositions is that postcumulus processes may have overprinted the original cumulus processes. The most significant postcumulus process that alters the original chemistry of a cumulate rock or mineral is the ‘trapped melt effect’ (e.g. Barnes, 1986
; Meyer et al., 1989; Bédard, 1994; Cawthorn, 1996; Ross & Elthon, 1997). During the last stage of solidification of a cumulate a certain amount of liquid can become isolated from the main melt reservoir and may, during further cooling, crystallize within the cumulate framework as overgrowth on pre-existing minerals and/or as newly formed intercumulus minerals. As a consequence, the cumulate may be significantly enriched in those elements (for example, REE) that were excluded from the original cumulus phases. However, the relative trace element enrichment of a mineral is strongly dependent on the mineral mode and whether or not the matrix is buffered (Meyer et al., 1989; Ross & Elthon, 1997). Buffered rock matrices are those that have a high bulk partition coefficient for a given incompatible element. For example, clinopyroxene crystallizing from a trapped interstitial melt in an olivine-dominated matrix may be buffered with respect to Mg and Fe, but not with respect to Ti or REE (Meyer et al., 1989). Similarly, it has been proposed that Rb is buffered in an anorthositic matrix but not in a pyroxenitic matrix (Cawthorn, 1996). The relative enrichment of incompatible elements such as Ti or REE in clinopyroxene is particularly significant in the case of a dunitic cumulate matrix, which has very low bulk partition coefficients for such elements. In contrast, clinopyroxene-rich matrices have relatively high bulk partition coefficients for HREE and Ti (when compared with dunitic matrices), and therefore intercumulus clinopyroxene will show only a slight trace element enrichment compared with the original cumulus composition. The Pyroxenite Suite samples chosen for this study are mainly composed of clinopyroxene with variable amounts of orthopyroxene (Table 1). Because of the high clinopyroxene contents (=buffered matrices), we assume that the trace element geochemistry of the clinopyroxenes was not substantially altered by trapped interstitial melt. Furthermore, we have no evidence for non-equilibrium effects. In each of the three samples, we analysed several large (0·4–15 mm) cumulus pyroxenes, and on the scale of the thin sections, we did not detect any grain-to-grain (and rarely core-to-rim) variation in the analysed trace elements. However, we cannot exclude the possibility that postcumulus annealing and re-equilibration effects may have destroyed pre-existing intra-grain trace element variations as a result of the trapped melt effect.
Parental melt compositions can also be estimated using whole-rock composition. This method may overcome some problems associated with clinopyroxene analyses, as postcumulus re-equilibration effects may alter the original magmatic signature of the clinopyroxene. Following the method described by Bédard (1994), we calculated equilibrium melts for the Pyroxenite Suite and Late Intrusion Suite samples. This method assumes that the minerals in a cumulate rock crystallized in equilibrium with a single parental melt at the same temperature, and the trapped melt component is considered as a separate phase with an assumed partition coefficient of unity. Because the amount of a trapped melt is difficult to assess, it was fixed at 5% for each sample. This value is the upper limit determined from model calculations applied to (ultra-)mafic cumulates from the Lewis Hills (Sassen, 2000) and North Arm Mountain (Bédard, 1994). We suggest that this is also a reasonable value for the Pyroxenite Suite samples, as we have no petrological (e.g. a high amount of interstitial clinopyroxenes or other phases) or chemical (e.g. intragrain zonation of trace elements) evidence for higher amounts of trapped melt component.
Figure 6b shows that calculated equilibrium melts for the Pyroxenite Suite samples derived from clinopyroxene analyses (grey triangles) closely match those obtained using whole-rock analyses (black squares). Thus, we suggest that both methods yield reliable results for Pyroxenite Suite parental melts. The modelled parental melts have relatively flat REE patterns, which are similar to those calculated for the SH boninite-type pyroxenitic dykes, but shifted to higher concentrations. Regarding absolute element abundances, the REE patterns of the Pyroxenite Suite parental melts are transitional between those of the BOIC basalts (light grey shaded field) and those of the arc-related mantle melts (dark grey shaded field; Fig. 6b). The equilibrium melts of the Late Intrusion Suite lie on, or plot close to, the field for the Pyroxenite Suite samples (Fig. 6c).
The MORB-normalized trace element patterns for Pyroxenite Suite and Late Intrusion Suite whole rocks are shown in Fig. 7a. Except for the olivine gabbro from the Late Intrusion Suite, the Pyroxenite Suite and Late Intrusion Suite samples have similar patterns and show characteristic spikes in their element distribution. The LREE are depleted relative to the HREE, and the high field strength elements (HFSE) such as Nb and Zr have negative anomalies. Thorium is enriched relative to the LREE in the Pyroxenite Suite samples. The olivine gabbro (sample 95-42) from the Late Intrusion Suite has very low absolute trace element abundances, reflecting the high modal olivine content, and possesses a pronounced negative Zr anomaly. The LREE and middle REE (MREE) distribution of the trace element pattern is roughly U-shaped, except for Nd and Eu. The relatively high Eu abundances are related to the high modal plagioclase abundances.
|
MORB-normalized trace element patterns of melts in equilibrium with the Pyroxenite Suite and Late Intrusion Suite samples are shown in Fig. 7b, along with those for BOIC basalts, equilibrium melts from the arc-related mantle units (SH–WL harzburgite) and equilibrium melts for the SH boninite-type pyroxenitic dykes from Edwards (1995)
. To avoid confusion, caused by the large number of samples, Pyroxenite Suite equilibrium melts derived via clinopyroxene analysis have been omitted from Fig. 7b. The Pyroxenite Suite and Late Intrusion equilibrium melts have relatively uniform characteristics, namely: (1) LREE-enriched to flat patterns with 0·1–0·3 times MORB abundances for the REE; (2) negative anomalies in Nb and Zr; (3) relative enrichments in Th (in the Pyroxenite Suite melts; Fig. 7b). Thus, the geochemical characteristics of the Pyroxenite Suite and Late Intrusion Suite equilibrium melts are similar to those of basalts erupted in supra-subduction zone settings (e.g. Saunders & Tarney, 1984; McCulloch & Gamble, 1991; Pearce & Peate, 1995; Peate et al., 1997). Pyroxenite Suite and Late Intrusion Suite parental melts have significantly lower trace element abundances when compared with BOIC basalts, and closely resemble equilibrium melts of the SH boninite-type pyroxenitic dykes (Fig. 7b). Furthermore, the overall trace element patterns of the Pyroxenite and Late Intrusion Suites, having LREE-enriched to flat patterns and negative Zr anomalies, are similar to the refractory equilibrium melts from the arc-related SH–WL mantle (Fig. 7b). In terms of the generally low trace element abundances, the Pyroxenite Suite and Late Intrusion Suite samples also match those of recent boninites from various parts of the world (Fig. 7c). However, the latter are often characterized by positive Zr anomalies (e.g. Hickey & Frey, 1982; Pearce et al., 1992).
Geochemistry of basaltic dykes
The major and trace element data for the basaltic dykes are listed in Table 3. All of the dykes have been affected to varying extents by (static) greenschist- or (syntectonic) amphibolite-facies metamorphism. Although all samples show a strong positive correlation between SiO2 and (Na2O + K2O) (Kurth, 2000), a remobilization of Na and K during metamorphism cannot be excluded. Based on the SiO2 contents, these samples can be classified as basalts and basaltic andesites (Hughes, 1982). In an AFM diagram, the basalts follow a tholeiitic trend (Kurth, 2000).
Figure 8a and b shows the TiO2 and Zr contents, respectively, of the basalts versus their Mg number. Fields for basalts from the BOIC and basaltic dykes from the MBC [‘normal’ and ‘depleted’ suites of Casey et al. (1985) and Elthon et al. (1986)] are shown for comparison. The basalts can be split into two groups, one with high TiO2 and Zr contents and the other with low TiO2 and Zr contents for a given Mg number. In the following discussion we will refer to the high TiO2 and Zr basalts as ‘Group I basalts’ and to the low TiO2 and Zr basalts as ‘Group II basalts’. Group I basalts fall within the field of the ‘normal suite’ of Casey et al. (1985) and Elthon et al. (1986), whereas Group II basalts plot together with the ‘depleted suite’ of those workers. The ‘normal suite’ has previously been associated with spreading-related magmatism, similar to that found in the northern massifs of the ophiolite (Elthon et al., 1986). In comparison, the ‘depleted suite’ has been attributed to processes related to melting within an oceanic fracture zone (Karson et al., 1983; Elthon et al., 1986).
|
The geochemical differences between the two groups are most evident in their extended MORB-normalized trace element patterns (Fig. 9). To minimize the effects of fractional crystallization on the trace element contents, basalts with comparable Mg numbers, in the range of 55–67, are shown. The patterns of the Group I basalts show increasing REE abundances from the HREE (
0·5–0·7 times MORB abundances) to the LREE (1–1·6 times MORB abundances). The high field strength element Nb exhibits negative anomalies (NbN/LaN < 1), whereas Th shows positive anomalies (ThN/NbN > 1; ThN/LaN > 1). In general, the trace element characteristics of the Group I basalts closely match those of BOIC lavas (Fig. 9). The Group II basalts, in contrast, have REE and HFSE abundances significantly lower than those of MORB and BOIC lavas (HREE 0·3 times MORB abundances; Fig. 9). With respect to the HREE, Group II basalts seem to be transitional between Group I–BOIC basalts and the parental melts of the Pyroxenite Suite and Late Intrusion Suite samples (Fig. 9). Group II basalts from the MBC have nearly flat trace element patterns, whereas those of the WLH show enrichments of the LREE (Fig. 9). The normalized trace element patterns of Group II basalts are relatively depleted in Nb (NbN/LaN <1) and enriched in Th (ThN/NbN >1; ThN/LaN >1).
|
Nd–Sr isotope geochemistry
Nd and Sr isotopic compositions of the whole-rock samples, together with calculated initial
Nd and initial 87Sr/86Sr ratios, are presented in Table 5. The ages of the Pyroxenite Suite, Late Intrusion Suite and basaltic dykes are not precisely known. However, upper and lower bounds on the ages can be constrained from the known ages of the WLH and ELH, which are 502 Ma and 485 Ma, respectively (Kurth et al., 1998). For the undeformed samples of the MBC (including the Pyroxenite and Late Intrusion Suites, Group II basalts and one sample of the Group I basalts M40), the younger age, which is the accretion age of the BOIC, is more likely to be the correct age (see section ‘Relative time of emplacement of the intrusive rocks’). For the deformed Group I basaltic dykes from the MBC and the undeformed dykes of the WLH, age constraints are limited, so any age in the range 485–502 Ma is possible. As one can see in Table 5, calculations based on either the maximum or minimum ages give very similar initial Nd values (Table 5).
|
The initial
Nd values of the samples, together with previously published data for the BOIC, Little Port Assemblage (LPA; Fig. 1), WLH and MBC, are summarized in Fig. 10. The Nd values calculated at 485 Ma for the Pyroxenite Suite samples are low (-1·5 to +0·6) and overlap those of the arc-related Little Port Assemblage and WLH. The samples from the Late Intrusion Suite have Nd(485 Ma) values lying between +2·9 and +4·6. These values are similar to those of samples from the MBC, and are thus transitional between the low initial Nd values of the arc-related LPA–WLH and high values of the back-arc related BOIC. With the exception of one sample, all the deformed dykes from the MBC (which belong to Group I) have similar initial Nd values, ranging from +6·3 to +6·9, irrespective of whether an age of 485 Ma or 502 Ma is taken (Table 5). In comparison, the undeformed dykes from the MBC (which belong to Group II) have lower initial Nd values, lying between +5·4 and +5·7. The basaltic dykes from the WLH have initial Nd values of +5·4 and +7·5 (Group I), and +5·5 and +5·9 (Group II), respectively.
|
Initial 87Sr/86Sr ratios lack a well-defined correlation with initial 143Nd/144Nd ratios (not shown). The majority of the samples are shifted to higher Sr isotope ratios, at nearly constant Nd isotopic composition. This feature is commonly observed for ophiolites and orogenic lherzolites (Richard & Allègre, 1980
) as well as for rocks from the BOIC (Jacobsen & Wasserburg, 1979). It has been explained in terms of seawater contamination (Hawkesworth & Elderfield, 1978). Because there is generally a poor correlation between the trace element concentrations of the samples of this study and their 87Sr/86Sr ratios, and all of the samples have been altered during greenschist-facies metamorphism, we attribute the higher 87Sr/86Sr ratios to contamination by seawater during metamorphism. This implies that the Sr isotopic compositions of the samples do not reflect magmatic values. In contrast, the Nd isotopic compositions should not be affected by alteration unless water/rock ratios were very high (>105; Jacobsen & Wasserburg, 1979). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
Parental melts of Pyroxenite and Late Intrusion Suites
In terms of petrology, geochemistry and Nd isotopic compositions, samples from the Pyroxenite Suite and Late Intrusion Suite are strikingly different from the majority of the BOIC and MOR cumulates and basalts (Figs 4–7 and 10).
The predominance of ortho- and clinopyroxene (high-SiO2 phases) over olivine in relatively unfractionated rocks (high Mg numbers in minerals) suggests that the parental melt of the Pyroxenite Suite had a rather high SiO2 content. In addition, the absence of plagioclase, or its late appearance in the fractionating assemblage, may be due to a high water content in the melt, which delays plagioclase crystallization (Yoder & Tilley, 1962; Ohnenstetter & Brown, 1996), or due to high Ca/Al ratios in the melt (Irvine, 1970). The parental melts of the Pyroxenite Suite appear to be fractionated mantle melts, as indicated by their non-primitive Mg number (89–71) and low Ni contents (0·23–0·08%) in olivine (Fig. 4a). We suggest that the samples from the Pyroxenite Suite evolved by fractional crystallization from more primitive melts, which possessed similarities to the parental melts of the SH pyroxenitic dykes. These latter dykes are characterized by (1) higher NiO contents in olivine (Fig. 4a), (2) higher Mg number, lower Al2O3 and TiO2 contents in clinopyroxene (Fig. 4c and d), and (3) lower REE abundances in their calculated parental melts (Figs 6b and 7b). However, parental melts of the Pyroxenite Suite and SH pyroxenitic dykes cannot be simply related by fractional crystallization to a single parental melt. This is because parental melts of the SH pyroxenitic dykes are depleted in REE by a factor of 2 relative to those of the Pyroxenite Suite (Fig. 6b), which would require an extremely high degree of fractionation (>50%) if derived from a single magma. Instead, the variability of REE abundances suggests an origin for the Pyroxenite Suite involving a heterogeneous source region and/or magma mixing. For example, mixing of a low-REE melt (such as a refractory arc-derived equilibrium melt; Fig. 6b) and a high-REE melt (such as represented by the BOIC basalts; Fig. 6b) could account for the range in REE abundances seen in the Pyroxenite Suite and SH pyroxenitic dykes.
Source heterogeneity or magma mixing is further supported by mineral data from the Pyroxenite Suite. In particular, the strong Al2O3 increase with decreasing Mg number in clinopyroxenes (Fig. 4d) suggests magma mixing and/or the presence of compositionally distinct parental melts. For example, mixing of a low Al2O3 and a high Al2O3 melt could account for the large range in Al2O3. Similarly, mixing of melts with different Si/Al ratios is a plausible process for generating the Al variability.
Although the samples of the Pyroxenite Suite are slightly fractionated, as implied by their Mg numbers in olivine (Fig. 4a), high MgO contents in the parental melt are inferred from high Mg numbers of mafic phases in some less differentiated Pyroxenite Suite samples and—assuming a genetic relationship—especially in samples of the SH pyroxenitic dykes. The Na2O, TiO2 and Al2O3 contents of clinopyroxenes from the Pyroxenite Suite and Late Intrusion Suites overlap those of the BOIC refractory group (Fig. 4). The Al2O3 contents of clinopyroxenes in the most primitive samples (Mg number <88) from the Pyroxenite Suite and Late Intrusion Suites range from 1 to 2·5 wt %. Such Al2O3 contents are far lower than those of clinopyroxenes from MOR and BOIC cumulates (Fig. 4c). Using an Al2O3–clinopyroxene/liquid partition coefficient of 0·3 for melts in equilibrium with a spinel peridotite paragenesis (Baker & Stolper, 1994), the low Al2O3 of the clinopyroxenes requires 3–8% Al2O3 in the parental melts of the Pyroxenite and Late Intrusion Suites. Such melt Al2O3 contents are, however, well below the 12–13% expected in melts of fertile mantle peridotite (Hirose & Kushiro, 1993; Baker & Stolper, 1994). At constant pressure conditions, the Al2O3 content in the melt decreases as the degree of partial melting increases (Jaques & Green, 1980; Hirose & Kushiro, 1993; Baker & Stolper, 1994; Kogiso et al., 1998). Therefore, we suggest that a refractory mantle peridotite source depleted by extensive basalt extraction is the most reasonable candidate for the low-Al melts.
An alternative explanation is that the parental melts of the Pyroxenite and Late Intrusion Suites come from greater depths than those from the BOIC, because melt Al2O3 has been found to decrease with increasing pressure at a given degree of melting (Hirose & Kushiro, 1993). Because the pressure effect on melt Al2O3 content is small up to 30 kbar (Hirose & Kushiro, 1993), parental melts of the Pyroxenite and Late Intrusion Suites need to be formed at depths of 100 km to explain their low Al2O3 contents. This would imply that the Pyroxenite Suite and Late Intrusion Suite parental melts were derived from much greater depths than the BOIC lavas, which is unlikely.
The preferred hypothesis—that the low-Al melts derive from mantle already depleted in basaltic components—is further supported by low concentrations of incompatible elements such as Ti or REE in clinopyroxenes from the Pyroxenite Suite (Figs 4c and 5) as well as in parental melts of the Pyroxenite and Late Intrusion Suites (Fig. 6b and c).
The strong depletion of clinopyroxene Al2O3 and most incompatible elements (Figs 6b and 7b) argues for very high degrees of source depletion. However, the relatively flat trace element patterns for the Pyroxenite Suite and Late Intrusion Suite samples are surprising for melts that are obviously strongly depleted (HREE are 2–5 times chondrite; Fig. 6b and c). Furthermore, the selective enrichment of Th, which is a large ion lithophile element (LILE) (Fig. 7b), together with the low initial Nd values observed (Fig. 10) require that the previously depleted magma source suffered an enrichment event. Positive LILE and negative HFSE are known to be characteristic of supra-subduction zone magmas (e.g. Saunders & Tarney, 1984; McCulloch & Gamble, 1991). In the context of a supra-subduction zone environment, the initial low Nd values indicate a re-enrichment of the mantle wedge by slab-derived components, which carry the fingerprint of crustal rocks or, generally, a reservoir that has evolved with a low Sm/Nd ratio. For example, Nd values <0 are observed in various arc-related basalts (Sunda–Banda arc and Lesser Antilles arc), which have been taken as one line of evidence for the involvement of terrigenous sediments in their genesis (e.g. Hofmann & White, 1982; White & Dupré, 1986; Hawkesworth et al., 1991; Peate et al., 1997).
In summary, the parental melts of the Pyroxenite and Late Intrusion Suites are inferred to have high Si and Mg and low Al contents. The corresponding trace element signature suggests derivation from a highly depleted or refractory source region. Furthermore, initial Nd values, as well as the selective enrichment of LILE and relative depletion of HFSE, require a strong supra-subduction zone influence. Such melts are found in modern and ancient subduction zone settings and are collectively termed boninites (e.g. Crawford et al., 1989) (see Fig. 7c). Although no boninitic lavas have been described from the BOIC, several studies have found a boninite-like signature in some cumulate rocks (Bédard, 1991, 1994; Edwards, 1995; Varvalvy et al., 1996, 1997) and in ultramafic exposures from the transition zone and mantle (Suhr et al., 1998, 2003; Suhr, 1999).
Source region of the Pyroxenite Suite and Late Intrusion Suite parental melts
The very depleted trace element signature of the Pyroxenite Suite and Late Intrusion Suite parental melts (Figs 6b, c and 7b) indicates a refractory mantle source. The arc-related SH harzburgite (Fig. 2) would be a candidate refractory source, as equilibrium melts of the SH mantle have a strong refractory signature (Suhr & Edwards, 2000; Figs 6b and 7b). Furthermore, the SH equilibrium melts show strong Zr anomalies, which are similar to those found in the Pyroxenite Suite and Late Intrusion Suite parental melts (Fig. 7b). Negative Zr anomalies are a common feature of many arc-related volcanics (e.g. McCulloch & Gamble, 1991).
The Nd isotopic characteristics also provide an argument for a genetic link between the island arc and the Pyroxenite and Late Intrusion Suites. Initial Nd values of the arc-related CC (-2 to +2) are significantly lower than those of the back-arc related BOIC (+6 to +9; Fig. 10). Samples with transitional initial Nd values are derived from the MBC arc–back-arc transition zone (Fig. 10). Initial Nd values of the Pyroxenite Suite (-1·5 to +0·6) overlap those of the CC island arc, whereas those of the Late Intrusion Suite (+2·9 to +4·6) correspond to the transitional values found in the MBC (Fig. 10). Furthermore, the Pyroxenite and Late Intrusion Suites are spatially associated with arc lithosphere, as both types of intrusion are restricted to the MBC arc–back-arc transition zone (Fig. 2).
In principle, the depleted trace element signature of the Pyroxenite Suite and Late Intrusion Suite parental melts could have had different origins. For example, (1) original MORB-type melts could have been transformed into refractory melts by chromatographic fractionation within a highly depleted, residual mantle (Navon & Stolper, 1987; Godard et al., 1995; Suhr et al., 1998; Suhr, 1999), or (2) refractory melts may have been formed in the advanced stages of progressive, near-fractional, decompression melting of oceanic mantle (Duncan & Green, 1987). During fractional melting of adiabatically upwelling mantle, a range of magmas are produced, including highly depleted compositions in the uppermost part of the mantle column (Langmuir et al., 1992; Sobolev & Shimizu, 1993). Studies of abyssal peridotites (Johnson et al., 1990), MORB-related cumulates (Ross & Elthon, 1993) and melt inclusions in MORB-hosted olivines and plagioclases (Sobolev & Shimizu, 1993; Sobolev, 1996; Nielsen et al., 1995; Sobolev et al., 2000) have shown that refractory melts can remain chemically isolated during their ascent through the oceanic upper mantle. However, ultra-depleted magma compositions are rare in melt inclusions and cumulates in the oceanic lithosphere, and are absent in oceanic lavas. Therefore, compositionally diverse melt fractions most probably undergo mixing within the mantle or in crustal magma chambers, producing normal typical MORB magmas (Langmuir et al., 1992). In both models mentioned above, melting of adiabatically upwelling mantle has proceeded so far that the uppermost part of the mantle section becomes a highly depleted mantle residue, reflecting the increasing extent of melting upwards in the melting regime. The higher mantle temperatures are, the greater the extent of melting. Therefore, high degrees of melting are expected for hot mantle at moderate to fast spreading ridges, which leave a highly refractory mantle residue at shallow levels. In contrast, ascending mantle beneath thick lithosphere or adjacent to a fracture zone probably does not upwell adiabatically, as a result of conductive cooling near the cold lithosphere (Langmuir et al., 1992). Consequently, melting may stop at depths greater than the base of the crust (Langmuir et al., 1992). Such a cold-wall effect near an oceanic fracture zone will tend to suppress shallow-level melting of the mantle, and preserve a lower extent of depletion (Fox & Gallo, 1984).
Geochemical studies have shown that the BOIC-related mantle of the ELH (Wheelers Brook mantle; see Fig. 2) is one of the least depleted residues within the spectrum of BOIC mantle rocks (Batanova et al., 1998; Suhr & Batanova, 1998). Because the Lewis Hills is the only massif that formed adjacent to the fracture zone contact between the older, colder CC arc and the younger BOIC oceanic basin (Fig. 3), the lower degree of depletion observed there has been related to the thermal cold-wall effect adjacent to the fracture zone (Suhr & Batanova, 1998). Consequently, a highly depleted BOIC mantle residue, which is a prerequisite for the formation of refractory melts, is absent in the ELH. Therefore, the BOIC mantle is not a likely source for the refractory melts of the Pyroxenite and Late Intrusion Suites. Instead, the CC sub-island arc mantle, similar to that of the SH harzburgite, is the most probable candidate to have produced these refractory melts.
Basaltic dykes and source region of Group I dykes
Two types of basaltic dykes—Group I and Group II—can be distinguished on the basis of their trace element abundances (Figs 8 and 9). The geochemical characteristics of Group I dykes favour a depleted MORB-type source. In particular, REE, Ti and Zr abundances are similar to those of MORB (Fig. 9) and initial Nd values lie between +5·4 and +7·5. However, in contrast to these MORB characteristics, the presence of negative Nb and positive Th anomalies indicates a supra-subduction zone setting. Overall, Group I basalts resemble back-arc basin basalts geochemically (e.g. Pearce et al., 1984; Saunders & Tarney, 1984; Hawkins, 1995), and are also similar to BOIC basalts (Fig. 9). Therefore, Group I and BOIC basalts are considered to share common mantle sources. For a similar range of Mg number (55–67), Group II dykes have significantly lower abundances of the incompatible elements (HREE 0·3 times MORB) than found in Group I dykes (HREE 0·6 times MORB; Figs 8 and 9), which are also nearly as low as those of the depleted parental melts of the Late Intrusion and Pyroxenite Suites (Fig. 9). The geochemical difference, at similar Mg number, between the two types of basaltic dykes suggests that the two types cannot be simply related to each other by fractional crystallization of a common magma, as at least 50% fractionation of a MORB-type magma (assuming a fractionating mineral assemblage of 10% olivine, 45% plagioclase and 45% clinopyroxene) would be required to increase its HREE budget by a factor of two.
Source region of Group II dykes
The Group II basalts clearly have lower abundances of REE and HFSE than basalts from Group I and the BOIC (Fig. 9). Abundances of the MREE and HREE are lower than those found in MORB, and are nearly as low as those in the refractory Pyroxenite Suite parental melts (Fig. 9). Their trace element abundances are thus transitional between those of the Pyroxenite Suite and those of BOIC or back-arc basin related melts (Fig. 9). The low trace element abundances imply that Group II basalts originate from a very depleted source.
This marked source depletion can also be inferred from chondrite-normalized Zr–Ti and Yb–Ti relationships (Fig. 11). These elements were selected as model elements because they are relatively unaffected by alteration, such as during secondary enrichment via subduction zone fluids. Figure 11a and b shows the chondrite-normalized Zr–Ti and Yb–Ti covariations of the Group II basaltic dykes. For comparison, the corresponding covariations are also shown for back-arc basin related melts (field 1) and refractory, arc-related melts (field 2) of the Lewis Hills. Field 1 contains the Group I basaltic dykes and calculated parental melts of Lewis Hills gabbros (Sassen, 2000). The latter have been genetically related to the BOIC, based on geochemical and isotopic evidence (Sassen et al., 1997). Field 2 contains the equilibrium melts of the SH harzburgite, and the parental melts of the Pyroxenite and Late Intrusion Suites. In addition, two melting curves are illustrated, showing the compositions of equilibrium melts derived from a MORB-type mantle source (curve 1) and a refractory mantle source (curve 2).
|
The starting composition for modelling mantle melts was a MORB-type mantle (McCulloch & Bennett, 1994
) with Nd(485 Ma) = +8·9, calculated using the model of Goldstein et al. (1984). The unfractionated trace element concentration of an average sediment composition from the Mariana arc with Nd(485 Ma) = -2·2 (Plank & Langmuir, 1998) was used as the compositional analogue for the subduction component. Mariana average sediment was chosen because its composition is well constrained (Plank & Langmuir, 1998), and its Sr isotopic composition is relatively unradiogenic, which is required by the unradiogenic Sr isotopic composition of the majority of the LH gabbros (Sassen, 2000). About 1% sediment addition to the MORB-type mantle yields Nd values that fall in the range of those observed in the BOIC. The melt compositions of the mixed mantle were calculated using a non-modal batch melting equation and the modes of Johnson et al. (1990). The mineral–melt partition coefficients used are listed in Table 4. Melts were calculated for 1, 5, 10, 15, 20, 25 and 30% melting (curve 1 represents batch melting).
The composition of the residue at clinopyroxene-out conditions (extent of melting F = 0·26) was taken as the refractory mantle source. This estimate of the extent of prior source depletion is similar to published estimates for depleted supra-subduction zone peridotites (e.g. Pearce & Peate, 1995; Parkinson & Pearce, 1998). This depletion may be related to sea-floor spreading at an oceanic ridge or earlier arc magmatism. The solid mode of the refractory residual mantle (0·71 olivine, 0·27 orthopyroxene, 0·02 spinel) was calculated using the expression (xi - Fpi)/(1 - F), where xi is the proportion of mineral i the solid, pi is the proportion of mineral i entering the melt, and F is the degree of melting. About 0·5% sediment addition to this refractory mantle yields Nd values that fall in the range of those observed in the CC island arc. The melt compositions of the mixed mantle were then calculated using equations for non-modal batch melting and a melt mode of 12% olivine, 86% orthopyroxene and 2% spinel (Kostopoulos, 1991). Melts were calculated for 1, 2, 4, 6, 8, 10 and 15% melting (curve 2 represents batch melting). For comparison, trends for non-modal fractional melting were additionally calculated (dotted curves in Fig. 11a) utilizing the equation
 |
Figure 11 shows that to achieve the low Zr, Ti and Yb concentrations seen in the Group II dykes, >30% equilibrium batch melting of a MORB-type source would be required. Such high degrees of partial melting would result in melts that were no longer basaltic in composition. Additionally, oceanic basalts segregate from their mantle sources at much lower degrees of melting (e.g. McKenzie, 1984). For a given degree of melting, fractional melting results in lower trace element concentrations but does not match those of the Group II dykes (Fig. 11a). Therefore, it is more reasonable to postulate a second-stage melting of a previously depleted mantle, as illustrated by curve 2 (Fig. 11). According to this melting curve, the depleted melts for the Group II dykes could result from remelting of a refractory source by 2–10% melting.
Because of the arguments outlined above, we do not consider that the Group II depleted melts could arise from (1) remelting of a refractory BOIC mantle, or (2) chemical interaction with strongly depleted residual BOIC mantle. Instead, the refractory sub-arc lithospheric mantle is the most likely source for such depleted melts, as already suggested for the Pyroxenite Suite. However, initial Nd values (+5·4 to +5·9) of the Group II dykes are significantly higher than those of the arc-related rocks and the Pyroxenite Suite (-2 to +2; Fig. 10), suggesting that some additional component must be involved in the magma genesis of Group II dykes. For example, the near linear relationship between the different magma suites, as seen in Fig. 11, could also result from magma mixing. That magma mixing may play an important role in the genesis of MBC intrusives is supported by their mineral chemistry, especially the strong Al2O3 increase with decreasing Mg number in clinopyroxenes from the Pyroxenite Suite (Fig. 4d). Mixing of magmas having different Al2O3 contents and/or Si/Al ratios has already been inferred to account for such Al2O3 enrichment. The intermediate initial Nd isotopic compositions of the Group II dykes, Late Intrusion Suite and MBC-related rocks (Fig. 10) may also support such a mixing relationship, for instance, between a refractory arc-derived melt (with low initial Nd values of -2 to +2) and a BOIC-derived melt (with high initial Nd values of +6 to +8).
However, our attempt to prove the mixing hypothesis is hampered by the fact that more than two compositional end-members are involved in magma mixing. In particular, the sub-island arc lithospheric mantle may be highly heterogeneous (see also the grey field in Fig. 7b), and thus will produce several compositionally distinct refractory melts that may be involved in mixing. Furthermore, refractory melts such as the Pyroxenite Suite may themselves represent mixtures, as deduced from their mineral chemistry (Fig. 4d).
Two refractory melt compositions were chosen for modelling such mixtures, which mark the known compositional ‘end-members’ of the arc lithosphere: (1) the primitive equilibrium melt composition derived from the SH harzburgite [sample LH 73 from Suhr & Edwards (2000)] and (2) a more evolved parental melt from the Pyroxenite Suite (95-73). Both are assumed to have a fixed Nd isotopic composition. However, this is a simplification, because the Nd isotopic composition of the SH harzburgite is at present not known. However, if we accept that the SH harzburgite is part of the mantle underlying the CC island arc, both should have a similar Nd isotopic composition. Samples from the CC analysed up to now have a relatively homogeneous Nd isotopic composition (initial Nd values range from -2 to +2; Fig. 10). Initial Nd values from the Pyroxenite Suite, which we suggest is genetically related to the arc mantle source, lie within this range (-1·5 to +0·6). Therefore, we use the lowest measured Nd isotopic composition of the Pyroxenite Suite (95-73), and assume that this is also the Nd isotopic composition of the SH equilibrium melt. The BOIC end-members are represented by two parental melts from the Lewis Hills gabbros (Sassen, 2000), with a primitive and a more evolved composition. Figure 12a–c presents the results of modelling such mixtures using initial 143Nd/144Nd ratios and the Nd, Zr and Ti contents of the samples. The numbers on the two mixing curves denote the proportions of BOIC melts added to the refractory melts.
|
The Group II dykes match the mixing curves fairly well in Fig. 12a–c, supporting the interpretation that these samples could have been produced by the mixing of refractory and BOIC-derived melts. It should be noted that the Group II dykes from the MBC and WLH lie on separate mixing curves, which suggest that different refractory melt end-members were involved in mixing. Mixing of geochemically distinct end-members could explain the differences in the overall trace element patterns of MBC and WLH Group II dykes; in particular, the distinct LREE abundances (Fig. 9). It is also evident from Fig. 12 that changes in the Nd isotopic composition of the mixtures are greatest for smaller contributions of the BOIC-type melts, whereas the trace element contents change more slowly. For example, the addition of
30% of a BOIC-type melt with highly radiogenic Nd to a refractory melt with unradiogenic Nd will result in a mixed melt that has trace element abundances similar to those of the refractory melt, but its Nd isotopic composition will be dominated by the isotopic signature of the BOIC melt. Furthermore, Fig. 12a–c shows that samples from the Late Intrusion Suite could also represent hybrid compositions, in agreement with their transitional initial Nd values (see Fig. 10).
An alternative scenario for the generation of transitional magma compositions is melting of a heterogeneous source, which might be exposed in the transition zone between the rifted arc and the back-arc basin. Such heterogeneous lithosphere, for example, has been inferred in the Lau and Mariana arc–back-arc area, where the region of rifted arc lithosphere extends over several tens to hundreds of kilometres (e.g. Fryer, 1995; Hawkins, 1995). However, the structure of the arc– back-arc boundary, as exposed in the Lewis Hills, indicates that the juxtaposition of the two geochemically distinct lithospheres occurred within a strike-slip kinematic or fracture zone setting (Karson & Dewey, 1978). Such a tectonic setting would imply that the transition from the arc to back-arc lithosphere occurs abruptly over a distance of 4–5 km [Fig. 13; see also Fig. 4 in Kurth et al. (1998)]. Consequently, a wide zone of heterogeneous lithosphere, which is compositionally transitional between the arc and back-arc basin lithosphere, is lacking in the Lewis Hills.
|
Genesis of heterogeneous magma compositions at a fracture zone (arc)–ridge intersection
In view of the number of geochemically and isotopically heterogeneous magma compositions observed, including highly refractory melt such as boninites, the question arises of how such magma diversity could be generated within a fracture zone setting. In such a setting, the high-temperature spreading ridge abuts against cooler arc lithosphere, and these two distinct lithospheres are juxtaposed along a steeply dipping transform–fracture zone contact (Fig. 13). As the island arc was at least
15 Myr older than the back-arc basin at the time, a relatively large temperature contrast is expected to have existed along the contact. By analogy with extant oceanic fracture zones, the degree of mantle melting along a spreading ridge, as well as the melt production rate, is expected to be lower as one approaches the transform–ridge intersection (Fox & Gallo, 1984; Langmuir & Bender, 1984; Phipps Morgan & Forsyth, 1988). This suggestion is confirmed by the observation that BOIC-related harzburgites from the Lewis Hills (Wheelers Brook mantle; Fig. 2) were formed by lower degrees of melting when compared with mantle peridotites from the northern massifs of the BOIC (Suhr & Batanova, 1998). This ‘transform effect’ has been attributed to perturbations of mantle upwelling flow lines at a transform offset at depth (Phipps Morgan & Forsyth, 1988) or/and increased shallow-level upper-mantle cooling (Fox & Gallo, 1984; Ghose et al., 1996). Consequently, it has been suggested that the development of larger magma chambers will be more or less suppressed towards the transform–ridge intersections, depending on the thermal state of the two lithospheres (Fox & Gallo, 1984). Furthermore, some portions of spreading ridge-derived melts may migrate along-axis towards the transform (Phipps Morgan & Forsyth, 1988).
The presence of boninite-type melts, from which the Pyroxenite Suite has crystallized, indicates that the refractory sub-arc mantle was reactivated and contributed to magma genesis. Remelting of refractory arc mantle requires a high-temperature thermal regime and/or a lowering of the peridotite solidus; for example, by a decrease in hydrostatic pressure or by the introduction of a hydrous phase (Hickey & Frey, 1982; Cameron, 1985; Duncan & Green, 1987; Van der Laan et al., 1989). Such a high thermal regime seems, at first sight, incompatible with a fracture zone setting.
Kurth et al. (1998) has presented a model for the deep continuation of the arc–ridge intersection. In this study we use a refined version of this model to put constraints on the formation of boninite-type melts and, more generally, heterogeneous magma compositions at an arc (transform)–ridge intersection. We suggest that it was the extensional fracture zone setting that provided the prerequisite conditions for generating boninitic magmas in the Lewis Hills. The transform tectonic setting provides the framework for the genesis of boninites because: (1) refractory, arc-type mantle and back-arc basin-type mantle are juxtaposed along a steep contact, which allows for an abrupt transition between the two lithospheres (Fig. 13); (2) the back-arc ridge, which abuts the arc lithosphere at the arc (transform)–ridge intersection, continuously supplies heat necessary to melt the (hydrated) sub-arc lithosphere; (3) the transform tectonics may have ruptured and weakened the arc lithospheric mantle; (4) extensional tectonics allow for rapid ascent of refractory mantle and pressure-release melting; (5) as a result of the thermal constraints near the arc (transform)– intersection, a large melt-homogenizing magma chamber is lacking in the vicinity of the plate boundary.
Figure 13 outlines the most important features of the fracture zone setting and its implications for magma genesis. Because of the reduced melt production near the plate boundary, magmatic flow lines from beneath the arc lithosphere are focused towards the arc (transform)–ridge intersection to feed the gap created by the diverging plates. As a result, heat will be continuously supplied to the arc lithospheric wall. The latter is supposed to be deeply ruptured by strike-slip dominated lithospheric deformation. Furthermore, the refractory sub-arc mantle has a lowered density compared with less depleted asthenospheric mantle. All these factors could have enhanced the probability that uprising asthenospheric mantle may detach and incorporate portions of the refractory arc lithosphere. Heating and decompression melting of entrained refractory arc lithosphere beneath the transtensional transform fault is supposed to have produced the boninite-type magmas of the MBC. Generally, heterogeneous mantle domains may feed several small and isolated magma chambers that develop in the vicinity of the arc–ridge intersection (Fig. 13). Within these magma chambers, compositionally distinct magma batches undergo mixing, possibly superimposed upon crystal fractionation, whereas others may escape mixing before they erupt along the plate boundary along the MBC.
An important aspect of this model is that the formation of boninite-type magmas is linked to a continuous, transform-related magmatic process and associated with a well-established back-arc basin spreading centre. Therefore, the formation of boninite-type magmas occurs simultaneously with the eruption of back-arc basin basalts. This is in contrast to most other models for boninite petrogenesis in which the formation of boninites is restricted to ephemeral magmatic phases of a subduction system; for example, (1) the initial stages of subduction (e.g. Hawkins et al., 1984; Crawford et al., 1989; Pearce et al., 1992; Stern & Bloomer, 1992; Bloomer et al., 1995); (2) arc splitting and initiation of back-arc spreading (Crawford et al., 1981; Pearce et al., 1984; Flower & Levine, 1987); (3) propagating spreading centres (Falloon & Crawford, 1991; Falloon et al., 1992; Kamenetsky et al., 1997).
Does the model apply also for other boninitic rocks in the BOIC?
The vast majority of crustal- and mantle-derived rocks from the BOIC are dominated by MORB-type REE geochemistry and petrology (Suen et al., 1979; Casey et al., 1985; Siroky et al., 1985; Komor & Elthon, 1990; Elthon, 1991; Jenner et al., 1991; Suhr et al., 1998). However, plutonic rocks with boninitic affinity occur at several locations in the northern massifs of the BOIC. These are exposed in the lower-crustal position and/or transition zone of the Blow Me Down and North Arm Mountain massifs (Bédard, 1994; Bédard & Hébert, 1996; Suhr et al., 1998), and also occur as dykes in the mantle section of the latter massif (Varvalvy et al., 1996, 1997). These occurrences are locally restricted, and may represent a rather minor component in the BOIC magmatic evolution. This seems not to be not true for refractory, boninitic melts that have formed numerous, highly depleted mantle dunites found in all three northern massifs (Suhr et al., 2003). These refractory melts are interpreted to have formed at a late stage of the dying BOIC spreading centre (Suhr et al., 2003). A late-stage event is supported by field evidence (Suhr et al., 2003) and the fact that the refractory signature otherwise would have been erased by mixing with contemporaneous MORB-type melts (Suhr et al., 1998).
The formation of boninite-type dykes in the North Arm Mountain massif has been related to fore-arc magmatism (Varvalvy et al., 1997). Such a setting was also implied by Suhr et al. (2003) although not explicitly stated. Modern fore-arcs are very complex, but sparse data indicate that the fore-arc could be built up of rock successions consistent with ophiolite lithologies as exposed, for example, in the Oman or Troodos ophiolite (e.g. Bloomer & Hawkins, 1983; Bloomer et al., 1995). Such rock successions have generally been interpreted as pre-arc basement, generated during the initiation of subduction and incipient arc volcanism (e.g. Pearce et al., 1984; Stern & Bloomer, 1992; Bloomer et al., 1995). However, generation during initiation of subduction is probably not appropriate for the BOIC. This is because an older arc complex (CC) is in magmatic contact with the BOIC marginal basin (Jenner et al., 1991; Kurth et al., 1998), implying established island arc lithosphere at the time of formation of the BOIC spreading ridge. Furthermore, the BOIC has a well-developed sheeted dyke complex (Rosencrantz, 1983). In addition, the crustal sequence overlying the mantle rocks has generally the well-defined undisturbed layered structure typical of oceanic crust (e.g. Casey & Karson, 1981; Casey et al., 1981; Girardeau & Nicolas, 1981). Thus, the BOIC back-arc basin was most probably generated at an established spreading ridge. True fore-arc spreading, in terms of establishing a stable spreading ridge, has actually not been observed in modern subduction zones (see Bloomer et al., 1995), although tectonic models exist for this process (Stern & Bloomer, 1992; Shervais, 2001). We suggest that a strong argument against a fore-arc setting is that the BOIC lacks boninite-type lavas, and crustal plutonic rocks having a boninite-type petrology and geochemical signature are rarely exposed. In contrast, the genesis of boninitic magmas is a major component in the evolution of modern fore-arcs (e.g. Bloomer et al., 1995; Shervais, 2001).
Based on the above considerations and given geological–structural analyses (Karson & Dewey, 1978; Girardeau & Nicolas, 1981; Casey et al., 1983; Suhr, 1992), we suggest that the BOIC was formed in an overall back-arc basin setting along a well-established spreading ridge. The model proposed here and elsewhere (Kurth et al., 1998) offers the possibility that boninite-type, refractory melts occur not only at the arc–ridge contact, but also off-axis. For example, detached slivers of lithospheric sub-arc mantle might be carried away from the arc root, stay as an entrained portion in the asthenospheric mantle, and later undergo decompression melting in an off-axis position. Off-axis melting of such slivers could possibly explain the presence of boninitic melts in the northern massifs. Furthermore, this model suggests that boninite-type magmas occur simultaneously with back-arc basin related magmas. This might be an important criterion to distinguish, for example, the MBC boninite-type Pyroxenite Suite parental melts from late-stage refractory melts, which were probably formed when MORB-type magmatism had already ceased (Suhr et al., 1998, 2003).
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
The Lewis Hills preserve a major tectonic, geochemical and age boundary that is unique to the BOIO. It represents the narrow transition zone (MBC) between a CC island arc and its BOIC back-arc basin, characterized by intense and multiple magmatism. In this study we present new geochemical and isotope data for three types of the MBC intrusions. These are highly heterogeneous in terms of trace element abundances and isotope geochemisty. We suggest that this heterogeneity is due to the involvement of compositionally distinct mantle sources—the lithospheric island arc mantle and the asthenospheric BOIC mantle source.
Parental melts of the Pyroxenite and Late Intrusion Suites are inferred to be rich in MgO and poor in TiO2, Al2O3 and Na2O. Their very low trace element abundances, when compared with those of MORB and BOIC basalts, indicate that the mantle source of the parental melts of the Pyroxenite and Late Intrusion Suites is strikingly different from those of the BOIC back-arc basin, and had been strongly depleted by previous melt extraction processes. Thus, the Pyroxenite and Late Intrusion Suites share geochemical characteristics expected for partial melts of a refractory mantle source (e.g. boninites). Additionally, the lack of LREE depletion, the selective enrichment in LILE and low initial Nd values (+1·5 to +0·6 for the Pyroxenite Suite, and +2·9 to +4·6 for the Late Intrusion Suite) imply the geochemical imprint of subduction zone derived fluids bearing a crustal component. The REE patterns of these melts strongly resemble equilibrium melts of the CC island arc related SH mantle, exposed in the southern part of the Lewis Hills. This similarity, coupled with the low, CC arc-typical initial Nd values of the Pyroxenite Suite, establishes a link with the sub-island mantle of the CC island arc. Therefore, it is likely that the Pyroxenite Suite and Late Intrusion Suite parental melts derive from remelting of the CC sub-arc lithospheric mantle at an arc–ridge intersection.
In contrast, Group I basaltic dykes have trace element abundances and initial Nd values (+5·4 to +7·5) consistent with derivation from the BOIC back-arc basin source. The initial Nd values (Nd = +5·4 to +5·9) of the Group II basalts are slightly lower but are broadly similar to those of Group I. However, the trace element patterns of the Group II basaltic dykes are transitional between those of the refractory arc-related and the BOIC back-arc basin melts. Island arc and back-arc basin derived magmas were probably mixed in small magma chambers in the vicinity of the arc–ridge intersection, giving rise to the transitional geochemical characteristics of the Group II basalts. A mixing relationship may possibly also be responsible for the relatively high initial Nd values observed in the Late Intrusion Suite when compared with those of the remainder of the CC island arc lithosphere. However, we cannot exclude the possibility that some proportions of depleted melts may have also originated from a depleted BOIC source.
Whatever the origin of the depleted Group II basaltic dyke magmas is exactly, we suggest that an extensional transform setting and arc–ridge intersection were the prerequisites for generating boninite-type melts and magma heterogeneity in the Lewis Hills. More generally, the model presented here and elsewhere (Kurth et al., 1998) provides a mechanism for the genesis of heterogeneous magma compositions at transtensional transform faults, where variously depleted lithospheres are juxtaposed at a transform–ridge intersection.
|
APPENDIX: ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
Major and trace elements
Major and some trace element (Sc, V, Ti, Cr and Ni; for basalts additionally Y and Zr) analyses of whole rocks were determined on glass pellets using X-ray fluorescence spectroscopy (Table 1 ). Measurements were performed at the Institut für Mineralogie und Geochemie, Universität zu Köln, using a Philips PW 2400 instrument. The remainder of the trace elements were determined by either inductively coupled plasma atomic emission spectrometry (ICP-AES) or inductively coupled plasma mass spectrometry (ICP-MS). Analyses by ICP-AES were performed (for basaltic samples: M11, M13, M1170, WL3 and WL59) at the Universität zu Köln, after separation of the trace elements by standard cation exchange chromatography in an HCl medium using Dowex AG 50W-X8 resin. The samples were previously digested in concentrated HF–HNO3 and brought into solution in 1N HCl. The analytical precision of the method is <3% (2
), as determined by analyses of an in-house standard.
Analyses by ICP-MS were conducted at (1) Memorial University in Newfoundland (on Pyroxenite Suite samples 95-73, 95-75, 95-93 and L686, and basaltic samples M7, M40, M1169, M1232, WL1 and WL64), and (2) Geologisches Institut an der Universität zu Köln (on Late Intrusion Suite samples 95-42, 95-45 and 95-107, and basaltic samples M43 and M53). The analytical procedure at the Memorial University followed the methods described by Jenner et al. (1990) and Longerich et al. (1990). The detection limits are 6–50 ppb for REE, 10 ppb for Nb and 20 ppb for Zr. The analytical precision of the method is 3–7% (2; M. Tubrett, personal communication, 1997), as determined from multiple analyses of international reference standards. The accuracy of the method is illustrated in Fig. A1. For most elements, the accuracy is <20% expressed in terms of the deviation of the measured values from the recommended values (Govindaraju, 1994; Dulski, 2001). At low concentrations (e.g. ultramafic UB-N standard), measured values for Nb were slightly higher than the literature values. The Nb concentrations of the Pyroxenite Suite samples are very low and are close to Nb detection limits (10 ppb). Therefore, we additionally conducted Nb analyses on two samples (95-73 and 95-93) using multi-ion counting spark-source mass spectrometry (MIC-SSMS). For information about the analytical procedure and instrumental set-up, the reader is referred to Jochum et al. (1997) and Pfänder et al. (1999). The results show that the Nb ICP-MS values for samples 95-73 (0·004 ppm) and 95-93 (0·008 ppm) are lower than those obtained using the MIC-SSMS (95-73: 0·024 ppm and 95-93: 0·023 ppm), probably as a result of Nb loss during sample solution for ICP-MS measurements (Pfänder et al., 1999). The MIC-SSMS data for the re-measured samples are preferred (detection limit for Nb is <5 ppb; Pfänder et al., 1999) and therefore used for data presentation.
|
Samples measured at the Geologisches Institut, Universität zu Köln, were digested in Savillex Teflon vessels using concentrated HF–HNO3. The vessels were heated to
100°C for at least 16 h. The resultant mixture was taken to dryness and then heated to 100°C in a few drops of HClO4 for 12 h. After the mixture was evaporated again to dryness, HCl was added and the residue was evaporated at 140°C. The samples were then dissolved in HNO3. After this treatment (‘open system decomposition’), visible precipitates of presumably spinel were still present in some of the samples; in particular, the cumulates from the Late Intrusion Suite. Therefore, these samples were additionally dissolved under pressure using the Anton Paar–Perkin-Elmer microwave digestion technique. For this purpose, concentrated HF–HClO4 was added to the samples, and the sealed Teflon vial (‘Liner’) was placed on the rotor of the microwave oven. The samples were then digested in two steps under pressure, first for 2 h at 200°C, and then for 45 min at 260°C. Pressure values, measured within the Liner, were Pmax = 30 bar and Pmax = 60 bar. After this treatment, the mixture was evaporated to dryness, refluxed in HCl and taken again to dryness. Then the samples were completely dissolved under pressure in 1 ml HNO3 at 200°C. Measurements of all samples, including those containing the spinel residues, together with procedural blanks (prepared with each batch of samples), were carried out using a Perkin-Elmer SciExi ELAN 6000 instrument. Before measurement, the samples were first diluted with ultrapure H2O to 50 ml total volume. Then, for final sample preparation, 100 µl of a 1000 µl/ml Rh–Re solution was added as an internal standard to 5 ml aliquots of the sample solutions, and the resultant mixture was diluted (with 2% HNO3) to 10 ml total volume. More detailed information concerning instrumental operating settings and the general calibration conditions is available on request. The analytical precision for the method is generally better than 5% (2; H. Kasper, personal communication, 2003). At very low element concentrations, the analytical precisions are between 2 and 20% (1) for REE, and between 25 and 30% for Zr and Th, as determined from multiple analyses of international reference ultramafic standards (DTS-1, PCC-1, UB-N; Schönbeck, 2001). The detection limits were between 2 and 14 ppb for REE and Th, and 30 ppb for Zr. Nb values for ultramafic standards should be considered with caution because of the high detection limit for this element (60–80 ppb). Nb values obtained for ultramafic samples 95-42, 95-45 and 95-107 were below the detection limit and were therefore discarded. The accuracy of the method is shown in Fig. A2. For most elements, the deviation of the measured values from the accepted literature values (Govindaraju, 1994; Dulski, 2001) is <20%. The measurements of the Late Intrusion Suite samples, which were dissolved by both open-system and microwave decomposition techniques, yielded nearly identical values. Data for each Late Intrusion Suite sample represent the average of six analyses, and for the remaining samples, three analyses.
|
The major element compositions of minerals were determined at the Institut für Mineralogie und Geochemie, Universität zu Köln, using a CAMECA CAMEBAX microprobe. Instrumental conditions were 20 kV acceleration voltage, 22 nA beam current and counting times between 10 and 40 s for olivine, and 15 kV acceleration voltage, 22 nA beam current and counting times between 5 and 20 s for pyroxenes. The calibration was based on natural mineral standards, synthetic oxides and alloys. Data for each sample represent the average of typically 4–10 analyses (2–4 analyses per grain). The analytical precision, as determined from multiple analyses of the standards, was better than 2·5% (1
) for high concentrations (e.g. MgO, CaO, SiO2), <4% for Al2O3, <11% for TiO2 and <28% for Na2O. The accuracy, expressed as the deviation (in %) of the measured average concentration from the standards, was <2% for Al2O3, MgO, CaO and SiO2, and 10–30% for Na2O and TiO2.
Trace element determinations on minerals were conducted at the Eidgenössische Technische Hochschule in Zürich, using a Perkin-Elmer Elan 6000 LA-ICP-MS instrument in combination with an excimer laser (193 nm, ArF). The instrumental conditions and analytical procedure have been descibed by Longerich et al. (1996) and Günther et al. (1997a, 1997b). The detection limits were in the range of 5–10 ppb for Sr, Y, Nb and REE, except for Nd, Sm, Gd, Dy, Er and Yb, for which the detection limits were between 10 and 25 ppb. The accuracy, expressed as the deviation (in %) of the measured average concentration from the recommended literature value (Pearce et al., 1997), is better than 2·5% for most elements, except Gd, Yb and Zr, for which it was 5%. The analytical precision of the method for each element is better than 4% (1), as determined by repeated measurements of the NIST 612 glass standard. Data for each sample represent the average of five analyses (2–3 analyses per grain).
Sr–Nd isotopic analyses
The Sr–Nd isotopic compositions, as well as the Sr and Nd concentrations, were determined from a single solution using mixed 85Rb–84Sr and 149Sm–150Nd spikes added to about 100 mg (or 1 g for pyroxenitic samples) of whole-rock powder. The samples were digested in concentrated HF–HNO3 and brought into solution in 6N HCl. Separation of Sr and REE as a group was performed by standard cation exchange chromatography in HCl medium using Biorad AG 50W-X12 resin. Sm and Nd were then separated from the REE using either reversed phase chromatography with hydrogen di-(2-ethylhexyl) phosphoric acid (HDEHP) immobilized on a Kel-F support in HCl medium, or by cation exchange using 
-hydroxyisobutyric acid (HIBA) as eluent. The average procedural blank for Nd was determined to be 25 pg. The isotope analyses of Sr and Nd were carried out at the Max-Planck-Institut für Chemie in Mainz by thermal ionization mass spectrometry on a Finnigan MAT 261 multicollector mass spectrometer operating in static mode. Rb, Sm and Nd were run using double Re filaments, whereas Sr was loaded on W filaments with TaF5 activator. 143Nd/144Nd ratios were corrected for fractionation using 146Nd/144Nd = 0·7219. The La Jolla Nd standard yielded 143Nd/144Nd = 0·511844 ± 21 (n = 31; 2ext.), with an external reproducibility corresponding to 0·4 units. Initial 143Nd/144Nd ratios are expressed in Nd units [Nd = (143Nd/144Ndsample/143Nd/144NdCHUR - 1) x 104] relative to the CHUR values at 485 Ma and 502 Ma (CHUR, today: 143Nd/144Nd = 0·512638; 147Sm/144Nd = 0·1967; Jacobsen & Wasserburg, 1979).
Sr isotope ratios were normalized using 86Sr/88Sr = 0·1194. The NIST NBS-987 standard yielded 87Sr/86Sr = 0·710228 ± 26 (n = 16, 2ext.). The errors on Sm and Rb concentrations are <0·15% (Sm) and 1–7% (Rb).
|
ACKNOWLEDGEMENTS |
|---|
This study is mainly based on Ph.D. studies of M.K. and A.S, supervised by H. A. Seck. Both authors wish to thank G. Suhr, who introduced us to the geology and the different concepts of the Bay of Islands Ophiolite. He also contributed substantially to field work. He is also thanked for useful comments on the manuscript. M.K. and A.S. further are grateful to A. W. Hofmann for providing free access to laboratory and mass spectrometry facilities and financial support during the preparation of the manuscript. W. Abouchami is thanked for her careful supervision in the laboratory, and H. Feldmann, I. Raczek and M. Klein for their technical assistance with mass spectrometry, microprobe and X-ray fluorescence analyses. D. Günther is acknowledged for providing access to a laser ablation– inductively coupled mass spectrometry (LA–ICP-MS) instrument. We also wish to thank K. Mezger for his constant help and review of an earlier version of the manuscript. H. U. Kasper (Geologisches Institut, Cologne) is thanked for his kind support and providing ICP-MS analyses. Thanks are also due to A. Gölden and T. Schönbeck for their help in the laboratory in Cologne. The manuscript benefited greatly from constructive reviews by J. Bédard, J. A. Pearce and D. Kamenetsky. This study was financially supported by the Deutsche Forschungsgemeinschaft under grants Se 274/28-1 and 274/29-1/2 and the ‘Graduiertenstipendium’ of the Universität zu Köln.
|
FOOTNOTES |
|---|
* Corresponding author. Telephone: +49-221-470-2304. Fax: +49-221-470-5199. E-mail: Michaela.Kurth{at}uni-koeln.de
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTING AND PREVIOUS...RESULTSDISCUSSIONSUMMARY AND CONCLUSIONSAPPENDIX: ANALYTICAL METHODSREFERENCES |
|---|
Anders, E. & Grevesse, N. (1989). Abundance of the elements: meteoric and solar. Geochimica et Cosmochimica Acta 53, 197–214.[CrossRef][Web of Science]
Baker, M. B. & Stolper, E. M. (1994). Determining the composition of high-pressure mantle melts using diamond aggregates. Geochimica et Cosmochimica Acta 58, 2811–2827.[CrossRef][Web of Science]
Barnes, S. J. (1986). The effect of trapped liquid crystallization on cumulus mineral compositions in layered intrusions. Contributions to Mineralogy and Petrology 93, 524–531.[CrossRef][Web of Science]
Batanova, V. G., Suhr, G. & Sobolev, A. V. (1998). Origin of geochemical heterogeneity in the mantle peridotites from the Bay of Islands ophiolite, Newfoundland, Canada: ion probe study of clinopyroxenes. Geochimica et Cosmochimica Acta 62, 853–866.[CrossRef][Web of Science]
Beattie, P. (1993). Uranium–thorium disequilibrium and partitioning on melting of garnet peridotite. Nature 363, 63–65.[CrossRef]
Bédard, J. H. (1991). Cumulate recycling and crustal evolution in the Bay of Islands Ophiolite. Journal of Geology 99, 225–249.[Web of Science]
Bédard, J. H. (1994). A procedure for calculating the equilibrium distribution of trace elements among the minerals of cumulate rocks, and the concentration of trace elements in the coexisting liquids. Chemical Geology 118, 143–152.[CrossRef][Web of Science]
Bédard, J. H. & Hébert, R. (1996). The lower crust of the Bay of Islands ophiolite, Canada: petrology, mineralogy, and the importance of syntexis in magmatic differentiation in ophiolites at ocean ridges. Journal of Geophysical Research 101, 25105–25124.[CrossRef]
Bloomer, S. H. & Hawkins, J. W. (1983). Gabbroic and ultramafic rocks from the Mariana trench; an island arc ophiolite. In: Hayes, D. E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands; Part 2. Geophysical Monograph, American Geophysical Union 27, 294–317.
Bloomer, S. H., Natland, J. & Fisher, R. L. (1989). Mineral relationships in gabbroic rocks from fracture zones of the Indian Ocean Ridge: evidence for extensive fractionation, parental diversity and boundary layer crystallization. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications 42, 107–124.
Bloomer, S. H., Taylor, B., MacLeod, C. J., Stern, R. J., Fryer, P., Hawkins, J. W. & Johnson, L. (1995). Early arc volcanism and the ophiolite problem: a perspective from drilling in the Western Pacific. In: Taylor, B. & Natland, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union 88, 1–30.
Cameron, W. E. (1985). Petrology and origin of primitive lavas from the Troodos ophiolite, Cyprus. Contributions to Mineralogy and Petrology 89, 239–255.[CrossRef][Web of Science]
Casey, J. F. & Karson, J. A. (1981). Magma chamber profiles from the Bay of Islands ophiolite complex. Nature 292, 295–301.[CrossRef]
Casey, J. F., Dewey, J. F., Fox, P. J., Karson, J. A. & Rosencrantz, E. (1981). Heterogeneous nature of oceanic crust and upper mantle: a perspective from the Bay of Islands Ophiolite Complex. In: Emiliani, C. (ed.) The Sea, Vol. 7. New York: John Wiley, pp. 305–338.
Casey, J. F., Karson, J. A., Elthon, D., Rosenkrantz, E. & Titus, M. (1983). Reconstruction of the geometry of accretion during formation of the Bay of Islands ophiolite complex. Tectonics 2, 509–528.[Web of Science]
Casey, J. F., Elthon, D. L., Siroky, F. X., Karson, J. A. & Sullivan, J. (1985). Geochemical and geological evidence bearing on the origin of the Bay of Islands and Coastal Complex ophiolites of western Newfoundland. Tectonophysics 11, 1–40.
Cawood, P. A. & Suhr, G. (1992). Generation and obduction of ophiolites: constraints from the Bay of Islands Complex, western Newfoundland. Tectonics 11, 884–897.[Web of Science]
Cawthorn, R.G. (1996). Models for incompatible trace-element abundances in cumulus minerals and their application to plagioclase and pyroxenes in the Bushveld Complex. Contributions to Mineralogy and Petrology 123, 109–115.[CrossRef][Web of Science]
Crawford, A. J., Beccaluva, L. & Serri, G. (1981). Tectono-magmatic evolution of the West Philippine–Mariana region and the origin of boninites. Earth and Planetary Science Letters 54, 346–356.[CrossRef][Web of Science]
Crawford, A. J., Falloon, T. J. & Green, D. H. (1989). Classification, petrogenesis and tectonic setting of boninites. In: Crawford, A. J. (ed.) Boninites and Related Rocks. London: Unwin Hyman, pp. 1–49.
Danyushevsky, L. V., Sobolev, A. V. & Falloon, T. V. (1995). North Tongan high-Ca boninite petrogenesis: the role of Samoan plume and subduction zone–transform fault transition. Journal of Geodynamics 20, 219–241.[CrossRef][Web of Science]
Dulski, P. (2001). Reference materials for geochemical studies: new analytical data by ICP-MS and critical discussion of reference values. Geostandards Newsletter 25, 87–125.[Web of Science]
Duncan, R. A. & Green, D. H. (1987). The genesis of refractory melts in the formation of oceanic crust. Contributions to Mineralogy and Petrology 96, 326–342.[CrossRef][Web of Science]
Dunn, T. & Sen, C. (1994). Mineral matrix partition coefficients for orthopyroxene, plagioclase and olivine in basaltic to andesitic systems: a combined analytical and experimental study. Geochimica et Cosmochimica Acta 58, 717–733.[CrossRef][Web of Science]
Dunning, G. R. & Krogh, E. (1985). Geochronology of ophiolites of the Newfoundland Appalachians. Canadian Journal of Earth Sciences 22, 1659–1670.[Web of Science]
Edwards, S. J. (1991). Magmatic and fluid processes in the upper mantle: a study of the Bay of Islands ophiolite complex, Newfoundland. Ph.D. thesis, Memorial University of Newfoundland, St. John's.
Edwards, S. J. (1995). Boninitic and tholeiitic dykes in the Lewis Hills mantle section of the Bay of Islands ophiolite: implications for magmatism adjacent to a fracture zone in a back-arc spreading environment. Canadian Journal of Earth Sciences 32, 2128–2146.[Web of Science]
Elthon, D. (1987). Petrology of gabbroic rocks from the Mid-Cayman Rise spreading center. Journal of Geophysical Research 92, 658–682.
Elthon, D. (1991). Geochemical evidence for formation of the Bay of Islands ophiolite above a subduction zone. Nature 354, 140–143.[CrossRef]
Elthon, D., Casey, J. F. & Komor, S. C. (1982). Mineral chemistry of ultramafic cumulates from the North Arm Mountain Massif of the Bay of Islands Ophiolite: evidence for high-pressure crystal fractionation of oceanic basalts. Journal of Geophysical Research 87, 8717–8734.
Elthon, D., Casey, J. F. & Komor, S. C. (1984). Cryptic mineral-chemistry variations in a detailed traverse through the cumulate ultramafic rocks of the North Arm Mountain Massif of the Bay of Islands Ophiolite, Newfoundland. In: Gass, I. G., Lippard, S. J. & Shelton, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications 78, 89–103.
Elthon, D., Karson, J. A., Casey, J. F., Sullivan, J. & Siroky, F. X. (1986). Geochemistry of diabase dikes from the Lewis Hills, Bay of Islands ophiolite: evidence for partial melting of oceanic crust in transform faults. Earth and Planetary Science Letters 78, 89–103.[CrossRef][Web of Science]
Ewart, A., Bryan, W. B., Chappell, B. W. & Rudnick, R. L. (1994). Regional geochemistry of the Lau–Tonga arc and backarc systems. In: Hawkins, J., Parson, L., Allan, J., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results 135. College Station, TX: Ocean Drilling Program, pp. 385–425.
Falloon, T. J. & Crawford, A. J. (1991). The petrogenesis of high-calcium boninite lavas dredged from the northern Tonga ridge. Earth and Planetary Science Letters 102, 375–394.[CrossRef][Web of Science]
Falloon, T. J., Green, D. H. & McCulloch, M. T. (1989). Petrogenesis of high-Mg and associated lavas from the north Tonga Trench. In: Crawford, A. J. (ed.) Boninites and Related Rocks. London: Unwin Hyman, pp. 357–395.
Falloon, T. J., Malahoff, A., Zonenshain, L. P. & Bogdanov, Y. (1992). Petrology and geochemistry of back arc basin basalts from Lau Basin spreading ridges at 15, 18 and 19S. Mineralogy and Petrology 47, 1–35.[CrossRef][Web of Science]
Flower, M. F. J. & Levine, H. M. (1987). Petrogenesis of a tholeiite–boninite sequence from Ayios Mamas, Troodos ophiolite: evidence for splitting of a volcanic arc? Contributions to Mineralogy and Petrology 97, 509–524.[CrossRef][Web of Science]
Fox, P. J. & Gallo, D. G. (1984). A tectonic model for ridge–transform–ridge plate boundaries: implications for the structure of oceanic lithosphere. Tectonophysics 104, 205–242.[CrossRef][Web of Science]
Fryer, P. (1995). Geology of the Mariana Trough. In: Taylor, B. (ed.) Backarc Basins: Tectonics and Magmatism. New York: Plenum, pp. 237–279.
Ghose, I., Cannat, M. & Seyler, M. (1996). Transform fault effect on mantle melting in the MARK area (Mid-Atlantic Ridge south of the Kane transform). Geology 24, 1139–1142.
Girardeau, J. & Nicolas, A. (1981). The structures of two ophiolite massifs, Bay-of-Islands, Newfoundland: a model for the oceanic crust and upper mantle. Tectonophysics 77, 1–34.[CrossRef][Web of Science]
Godard, M., Boudinier, J. L. & Vasseur, G. (1995). Effects of mineralogical reactions on trace element redistributions in mantle rocks during percolation processes: a chromatographic approach. Earth and Planetary Science Letters 133, 449–461.[CrossRef][Web of Science]
Goldstein, S. L., O'Nions, R. K. & Hamilton, P. J. (1984). A Sm–Nd isotopic study of atmospheric dusts and particulates from major river systems. Earth and Planetary Science Letters 70, 221–236.[CrossRef][Web of Science]
Govindaraju, K. (1994). Geostandards Newsletter 18, Special Issue, 158 pp.
Günther, D., Frischknecht, R., Heinrich, C. A. & Kahlert, H.-J. (1997a). Capabilities of an argon fluoride 193 nm excimer laser for ablation inductively coupled plasma mass spectrometry microanalysis of geological materials. Journal of Analytical Atomic Spectrometry 12, 939–944.[CrossRef][Web of Science]
Günther, D., Frischknecht, R., Müschenborn, H.-J. & Heinrich, C. A. (1997b). Direct liquid ablation: a new calibration strategy for laser ablation–ICP-MS microanalysis of solids and liquids. Fresenius Journal of Analytical Chemistry 359, 390–393.[CrossRef]
Hart, S. R. & Dunn, T. (1993). Experimental cpx/melt partitioning of 24 trace elements. Contributions to Mineralogy and Petrology 113, 1–8.[CrossRef][Web of Science]
Hauri, E. H. & Hart, S. R. (1994). Constraints on melt migration from mantle plumes: a trace element study of peridotitic xenoliths from Sava'i, Western Samoa. Journal of Geophysical Research 99, 24301–24321.[CrossRef]
Hauri, E. H., Wagner, T. P., Hart, S. R. & Grove, T. L. (1993). Experimental and natural partitioning of U, Zh, and Pb between cpx, garnet and basaltic melt: implications for Th/U, Ce/Pb and Pb isotope evolution in the upper mantle. EOS Transactions, American Geophysical Union 74, 340–341.
Hawkesworth, C. J. & Elderfield, H. (1978). The strontium isotopic composition of interstitial waters from Sites 245 and 336 of the Deep Sea Drilling Project. Earth and Planetary Science Letters 40, 423–432.[CrossRef][Web of Science]
Hawkesworth, C. J., Hergt, J. M., Ellam, R. M. & McDermott, F. (1991). Element fluxes associated with subduction related magmatism. Philosophical Transactions of the Royal Society of London, Series A 335, 393–405.[CrossRef]
Hawkesworth, C. J., Gallagher, K., Hergt, J. M. & McDermott, F. (1993). Trace element fractionation processes in the generation of island arc basalts. Philosophical Transactions of the Royal Society of London, Series A 342, 179–191.
Hawkins, J. W. (1995). Evolution of the Lau Basin—insights from ODP Leg 135. In: Taylor, B. & Natland, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union 88, 125–173.
Hawkins, J. W. & Melchior, J. T. (1985). Petrology of Mariana Trough and Lau Basin basalts. Journal of Geophysical Research 90, 11431–11468.[CrossRef]
Hawkins, J. W., Bloomer, S. H., Evans, C. A. & Melchior, J. T. (1984). Evolution of intra-oceanic arc–trench systems. In: Carlson, R. L. & Kobayashi, K. (eds) Geodynamics of Back-arc Regions. Tectonophysics 102, 175–205.
Hébert, R., Serri, G. & Hékinian, R. (1989). Mineral chemistry of ultramafic tectonites and ultramafic to gabbroic cumulates from the major oceanic basins and Northern Apennine ophiolite (Italy)—a comparison. Chemical Geology 77, 183–207.[CrossRef][Web of Science]
Hébert, R., Constantin, M. & Robinson, P. T. (1991). Primary mineralogy of Leg 118 gabbroic rocks and their place in the spectrum of oceanic mafic igneous rocks. In: Von Herzen, R. P., Robinson, P. T., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 118. College Station, TX: Ocean Drilling Program, pp. 3–18.
Hickey, R. L. & Frey, F. A. (1982). Geochemical characteristics of boninite series volcanics: implications for their source. Geochimica et Cosmochimica Acta 46, 2099–2115.[CrossRef][Web of Science]
Hirose, K. & Kushiro, I. (1993). Partial melting of dry peridotites at high pressures, determination of compositions of melts segregated from peridotite using aggregates of diamonds. Earth and Planetary Science Letters 114, 477–489.[CrossRef][Web of Science]
Hofmann, A. W. (1988). Chemical differentiation of the earth: the relationship between mantle, continental crust, and oceanic crust. Earth and Planetary Science Letters 90, 297–314.[CrossRef][Web of Science]
Hofmann, A. W. & White, W. M. (1982). Mantle plumes from ancient oceanic crust. Earth and Planetary Science Letters 57, 421–436.[CrossRef][Web of Science]
Hughes, C. J. (1982). Igneous petrology. In: Developments in Petrology, Vol. 7. Amsterdam: Elsevier, 551 pp.
Irvine, T. N. (1970). Crystallization sequences in the Muskox intrusion and other layered intrusions. Geological Society of South Africa, Special Publication 1, 441–476.
Jacobsen, S. B. & Wasserburg, G. J. (1979). Nd and Sr isotopic study of the Bay of Islands ophiolite complex and the evolution of the source of mid-ocean ridge basalts. Journal of Geophysical Research 84, 7429–7445.
Jaques, A. L. & Green, D. H. (1980). Anhydrous melting of peridotite at 0–15 kb pressure and the genesis of tholeiitic basalts. Contributions to Mineralogy and Petrology 73, 287–310.[CrossRef][Web of Science]
Jenner, G. A., Longerich, H. P., Jackson, S. E. & Fryer, B. J. (1990). ICP-MS—a powerful tool for high-precision trace-element analysis in earth sciences: evidence from analysis of selected U.S.G.S. reference samples. Chemical Geology 83, 133–148.[CrossRef][Web of Science]
Jenner, G. A., Dunning, G. R., Malpas, J., Brown, M. & Brace, T. (1991). Bay of Islands and Little Port complexes, revisited: age, geochemical and isotopic evidence confirm suprasubduction-zone origin. Canadian Journal of Earth Sciences 28, 1635–1652.[Web of Science]
Jochum, K. P., Laue, H. J., Seufert, H. M., Dienemann, C., Stoll, B., Pfänder, J., Flanz, M., Achtermann, H. & Hofmann, A. W. (1997). Multi-ion counting-spark source mass spectrometry (MIC-SSMS): a new multielement technique in geo- and cosmochemistry. Fresenius Journal of Analytical Chemistry 359, 385–389.[CrossRef]
Johnson, K. T. M., Dick, H. J. B. & 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.
Kamenetsky, V. S., Crawford, A. J., Eggins, S. & Mühe, R. (1997). Phenocrysts and melt inclusion chemistry of near-axis seamounts, Valu Fa Ridge, Lau Basin: insight into mantle wedge melting and the addition of subduction components. Earth and Planetary Science Letters 151, 205–233.[CrossRef][Web of Science]
Karson, J. A. (1977). Geology of the northern Lewis Hills, western Newfoundland. Ph.D. thesis, University of New York, Albany, 474 pp.
Karson, J. A. (1984). Variations in structure and petrology in the Coastal Complex, Newfoundland: anatomy of an ocean fracture zone. In: Gass, I. G., Lippard, S. J. & Shelton, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications 13, 131–146.
Karson, J. A. & Dewey, J. F. (1978). Coastal Complex, western Newfoundland: an early Ordovician oceanic fracture zone. Geological Society of America Bulletin. 89, 1037–1049.
Karson, J. A., Elthon, D. L. & DeLong, S. E. (1983). Ultramafic intrusions in the Lewis Hills, Bay of Islands ophiolite complex, Newfoundland: implications for igneous processes at oceanic fracture zones. Geological Society of America Bulletin 94, 15–29.
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]
Kharas-Khumbatta, N. (1988). Geochemistry and petrology of the gabbroic rocks from Blow Me Down Mountain, Bay of Islands Complex; composition of the parental magma in liquids in equilibrium with gabbroic minerals. Masters thesis, University of Houston, TX.
Kogiso, T., Hirose, K. & Takahashi, E. (1998). Melting experiments on homogeneous mixtures of peridotite and basalts: application to the genesis of ocean island basalts. Earth and Planetary Science Letters 162, 45–61.[CrossRef][Web of Science]
Komor, S. C. & Elthon, E. (1990). Formation of anorthosite–gabbro rhythmic phase layering: an example at North Arm Mountain, Bay of Islands Ophiolite. Journal of Petrology 31, 1–50.
Komor, S. C., Elthon, D. & Casey, J. F. (1985). Mineralogic variation in a layered ultramafic cumulate sequence at the North Arm Mountain Massif, Bay of Islands Ophiolite, Newfoundland. Journal of Geophysical Research 90, 7705–7736.
Komor, S. C., Elthon, D. & Casey, J. F. (1987). Petrology of a leucogabbroic interval within basal layered gabbros at North Arm Mountain, Bay of Islands Ophiolite. Contributions to Mineralogy and Petrology 95, 278–300.[CrossRef][Web of Science]
Kostopoulos, D. K. (1991). Melting of the shallow upper mantle: a new perspective. Journal of Petrology 32, 671–699.
Kurth, M. (2000). Petrographie, Geochemie, Isotopie und Alter von Kumulaten und Basalten des Lewis Hills Massif (Bay of Islands Opiolith, Neufundland): Hinweis auf den Inselbogenursprung des westlichen Lewis Hills und die Implikationen für die tektonische und magmatische Geschichte des Bay of Islands Ophioliths. Ph.D. thesis, University of Cologne, 245 pp.
Kurth, M., Sassen, A., Suhr, G. & Mezger, K. (1998). Precise ages and isotopic constraints for the Lewis Hills (Bay of Islands Ophiolite): preservation of an arc–spreading ridge intersection. Geology 26, 1127–1130.
Langmuir, C. H. & Bender, J. F. (1984). The geochemistry of oceanic basalts in the vicinity of transform faults: observations and implications. Earth and Planetary Science Letters 69, 107–127.[CrossRef][Web of Science]
Langmuir, C. H., Klein, E. M. & Plank, T. (1992). Petrological systematics of mid-ocean ridge basalts: constraints on melt generation beneath ocean ridges. In: Phipps Morgan, J., Blackman, D. K. & Sinton, J. M. (eds) Mantle Flow and Melt Generation at Mid-ocean Ridges. Geophysical Monograph, American Geophysical Union 71, 183–280.
Longerich, H. P., Jenner, G. A., Fryer, B. J. & Jackson, S. E. (1990). Inductively coupled plasma-mass spectrometric analysis of geological samples—a critical evaluation—based on case-studies. Chemical Geology 83, 105–118.[CrossRef][Web of Science]
Longerich, H. P., Jackson, S. E. & Günther, D. (1996). Laser ablation inductively coupled plasma mass spectrometric transient signal data acquisition and analyte concentration calculation. Journal of Analytical Atomic Spectrometry 11, 899–904.[CrossRef][Web of Science]
Malpas, J. (1979). Two contrasting trondhjemite associations from transported ophiolites in Western Newfoundland: initial report. In: Barker, F. (ed.) Trondhjemites, Dacites, and Related Rocks. Amsterdam: Elsevier, pp. 465–487.
McCulloch, M. T. & Bennett, V. C. (1994). Progressive growth of the Earth's continental crust and depleted mantle: geochemical constraints. Geochimica et Cosmochimica Acta 58, 4717–4738.[CrossRef][Web of Science]
McCulloch, M. T. & Gamble, A. J. (1991). Geochemical and geodynamical constraints on subduction zone magmatism. Earth and Planetary Science Letters 102, 358–374.[CrossRef][Web of Science]
McKenzie, D. P. (1984). The generation and compaction of partially molten rock. Journal of Petrology 25, 713–765.
Meyer, P. S., Dick, H. J. B. & Thompson, G. (1989). Cumulate gabbros from the southwest Indian Ridge 54°S–7°16'E: implications for magmatic processes at a slow spreading ridge. Contributions to Mineralogy and Petrology 103, 44–63.[CrossRef][Web of Science]
Monzier, M., Danyushevsky, L. V., Crawford, A. J., Bellon, H. & Cotten, J. (1993). High-Mg andesites from the southern termination of the New Hebrides island arc (SW Pacific). Journal of Volcanology and Geothermal Research 57, 193–217.[CrossRef][Web of Science]
Murton, B. J. (1989). Tectonic controls on boninite genesis. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications 42, 347–377.
Navon, O. & Stolper, E. (1987). Geochemical consequences of melt percolation: the upper mantle as chromatographic column. Journal of Geology 95, 285–307.[Web of Science]
Nielsen, R. L., Crum, J., Bourgeois, R., Hascall, K., Forsythe, L. M., Fisk, M. R. & Christie, D. M. (1995). Melt inclusions in high-An plagioclase from the Gorda Ridge: an example of the local diversity of MORB parent magmas. Contributions to Mineralogy and Petrology 122, 34–50.[CrossRef][Web of Science]
Ohnenstetter, D. & Brown, W. L. (1996). Compositional variation and primary water contents of differentiated interstitial and included glasses in boninites. Contributions to Mineralogy and Petrology 123, 117–137.[CrossRef][Web of Science]
Ozawa, K., Meyer, P. S. & Bloomer, S. H. (1991). Mineralogy and textures of iron–titanium oxide gabbros and associated olivine gabbros from hole 735B. In: Von Herzen, R. P., Robinson, P. T., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 118. College Station, TX: Ocean Drilling Program, pp. 41–73.
Parkinson, I. J. & Pearce, J. A. (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]
Pearce, J. A. & Peate, D. W. (1995). Tectonic implications of the composition of volcanic arc magmas. Annual Review of Earth and Planetary Sciences 23, 251–285.[CrossRef][Web of Science]
Pearce, J. A., Lippard, S. J. & Roberts, S. (1984). Characteristics and tectonic significance of supra-subduction zone ophiolites. In: Kokelaar, B. P. & Howells, M. F. (eds) Marginal Basin Geology; Volcanic and Associated Sedimentary and Tectonic Processes in Modern and Ancient Marginal Basins. Geological Society, London, Special Publications 16, 77–93.
Pearce, J. A., Van der Laan, S. R., Arculus, R. J., Murton, B. J., Ishii, T., Peate, D. W. & Parkinson, I. J. (1992). Boninite and harzburgite from Leg 125 (Bonin–Mariana Forearc): a case study of magma genesis during the initial stages of subduction. In: Fryer, P., Pearce, J. A., Stokking, L. B., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. College Station, TX: Ocean Drilling Program, pp. 623–659.
Pearce, N. J. G., Perkins, W. T., Westgate, J. A., Gorton, M. P., Jackson, S. E., Neal, C. R. & Chenery, S. P. (1997). A compilation of new and published major and trace element data for NIST SRM 610 and NIST SRM 612 glass reference materials. Geostandards Newsletter 21, 115–144.[CrossRef][Web of Science]
Peate, D. W., Pearce, J. A., Hawkesworth, C. J., Colley, H., Edwards, C. M. H. & Hirose, K. (1997). Geochemical variations in Vanuatu arc lavas: the role of subducted material and a variable mantle wedge composition. Journal of Petrology 38, 1331–1358.[CrossRef][Web of Science]
Pfänder, J., Jochum, K. P., Sassen, A., Stoll, B., Maissenbacher, P. & Murmann, M. (1999). Determination of Y, Zr and ultra-low Nb concentrations in geological material by multi-ion counting spark-source mass spectrometry (MIC-SSMS). Fresenius Journal of Analytical Chemistry, Conference Contributions 376–380.
Phinney, W. C. & Morrison, D. A. (1990). Partition coefficients for calcic plagioclase—implications for Archean anorthosites. Geochimica and Cosmochimica Acta 54, 1639–1654.
Phipps Morgan, J. & Forsyth, D. W. (1988). Three dimensional flow and temperature perturbations due to a transform offset: effects on oceanic crustal and upper mantle structure. Journal of Geophysical Research 93, 2955–2966.
Plank, T. & Langmuir, C. H. (1998). The chemical composition of subducting sediment and its consequences for the crust and mantle. Chemical Geology 145, 325–394.[CrossRef][Web of Science]
Richard, P. & Allègre, C. (1980). Neodymium and strontium isotope of ophiolite and orogenic lherzolite petrogenesis. Earth and Planetary Science Letters 47, 65–74.[CrossRef][Web of Science]
Rogers, N. W., MacLeod, C. J. & Murton, B. J. (1989). Petrogenesis of boninitic lavas from the Limassol Forest Complex, Cyprus. In: Crawford, A. J. (ed.) Boninites and Related Rocks. London: Unwin Hyman, pp. 288–313.
Rosencrantz, E. (1983). The structure of sheeted dikes and associated rocks in North Arm Mountain massif, Bay of Islands ophiolite complex, and the intrusive process at oceanic spreading centers. Canadian Journal of Earth Sciences 20, 787–801.[Web of Science]
Ross, K. & Elthon, D. (1993). Cumulates from strongly depleted mid ocean ridge basalt. Nature 365, 826–829.[CrossRef]
Ross, K. & Elthon, D. (1997). Cumulus and postcumulus crystallization in the oceanic crust: major and trace element geochemistry of Leg 153 gabbroic rocks. 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. 333–350.
Sassen, A. (2000). Geochemistry, petrology and isotope geochemistry of magmatic intrusions in the Lewis Hills Massif, Bay of Islands Ophiolite: implications for magma diversity at an island arc–spreading intersection. Ph.D. thesis, University of Cologne, 244 pp.
Sassen, A., Suhr, G., Galer, S. J. G. & Seck, H. (1997). Boninitic–tholeiitic cumulate associations in the Lewis Hills Massif, Bay of Islands Ophiolite (BOIO). Abstract Supplement 1, Terra Nova 9, 333.
Saunders, A. D. & Tarney, J. (1984). Geochemical characteristics of basaltic volcanism within back-arc basins. In: Kokelaar, B. P. & Howells, M. F. (eds) Marginal Basin Geology; Volcanic and Associated Sedimentary and Tectonic Processes in Modern and Ancient Marginal Basins. Geological Society, London, Special Publications 16, 59–76.
Schönbeck, T. (2001). Erprobung einer neuen Methode zur Bestimmung geochemisch relevanter Spurenelemente in ultramafischen Gesteinen. M.S. thesis, University of Cologne, 130 pp.
Searle, M. P. & Stevens, R. K. (1984). Obduction processes in ancient, modern and future ophiolites. In: Gass, I. G., Lippard, S. J. & Shelton, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications 13, 303–320.
Shervais, J. W. (2001). Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems 2, paper number 2000GC000080.
Sigurdsson, I. A., Kamenetsky, V. S., Crawford, A. J., Eggins, S. M. & Zlobin, S. K. (1993). Primitive island arc and oceanic lavas from the Hunter Ridge fracture zone. Evidence from glass, olivine and spinel compositions. Mineralogy and Petrology 47, 149–169.[CrossRef][Web of Science]
Siroky, F. X., Elthon, D. L., Casey, J. F. & Butler, J. C. (1985). Major element variations in basalts and diabases from the North Arm Mountain Massif, Bay of Islands Ophiolite: implications for magma chamber processes at mid-ocean ridges. Tectonophysics 116, 41–61.[CrossRef][Web of Science]
Smith, S. E. & Elthon, D. (1988). Mineral compositions of plutonic rocks from the Lewis Hills Massif, Bay of Islands Ophiolite. Journal of Geophysical Research 93, 3450–3468.
Sobolev, A. V. (1996). Melt inclusions in minerals as a source of principal petrological information. Petrology 4, 209–220.
Sobolev, A. V. & Shimizu, N. (1993). Ultra-depleted primary melt included in an olivine from the Mid-Atlantic Ridge. Nature 363, 151–154.[CrossRef]
Sobolev, A. V., Hofmann, A. W. & Nikogosian, I. K. (2000). Recycled oceanic crust observed in ‘ghost plagioclase’ within the source of Mauna Loa lavas. Nature 404, 986–990.[CrossRef][Medline]
Stern, R. J. & Bloomer, S. H. (1992). Subduction zone infancy: examples from the Eocene Izu–Bonin–Mariana and Jurassic California arcs. Geological Society of America Bulletin 104, 1621–1636.
Suen, C. J., Frey, F. A. & Malpas, J. (1979). Bay of Islands Ophiolite suite, Newfoundland: petrologic and geochemical characteristics with emphasis on rare earth element geochemistry. Earth and Planetary Science Letters 45, 337–348.[CrossRef][Web of Science]
Suhr, G. (1992). Upper mantle peridotites in the Bay of Islands Ophiolite, Newfoundland: formation during the final stages of a spreading centre? Tectonophysics 206, 31–53.[CrossRef][Web of Science]
Suhr, G. (1999). Melt migration under oceanic ridges: inferences from reactive transport modelling of upper mantle hosted dunites. Journal of Petrology 40, 575–599.[CrossRef][Web of Science]
Suhr, G. & Batanova, V. (1998). Basal lherzolites in the Bay of Islands Ophiolite: origin by detachment-related telescoping of a ridge-parallel melting gradient. Terra Nova 10, 1–5.[CrossRef][Web of Science]
Suhr, G. & Cawood, P. A. (1993). Structural history of ophiolite obduction, Bay of Islands, Newfoundland. Geological Society of America Bulletin 105, 399–410.
Suhr, G. & Cawood, P. A. (2001). The southeastern Lewis Hills (Bay of Islands Ophiolite), geology of a deeply eroded, inside-corner, ridge–transform intersection. Geological Society of America Bulletin 113, 1025–1038.
Suhr, G. & Edwards, S. J. (2000). Contrasting mantle sequences exposed in the Lewis Hills Massif: evidence for the early, arc-related history of the Bay of Islands Ophiolite. In: Dilek, Y., Moores, E. M., Elthon, D. & Nicolas, A. (eds) Ophiolites and Oceanic Crust: New Insight from Field Studies and Ocean Drilling Program. Geological Society of America, Special Paper 349, 433–442.
Suhr, G., Calon, T. & Dunsworth, S. M. (1991). Origin of complex upper mantle structures in the southern Lewis Hills (Bay of Islands ophiolite, Newfoundland). Canadian Journal of Earth Sciences 28, 774–787.[Web of Science]
Suhr, G., Seck, H. A., Shimizu, N., Günther, D. & Jenner, G. (1998). Infiltration of refractory melts into the lowermost oceanic crust: evidence from dunite- and gabbro-hosted clinopyroxenes in the Bay of Islands Ophiolite. Contributions to Mineralogy and Petrology 131, 136–154.[CrossRef][Web of Science]
Suhr, G., Hellebrand, E., Snow, J. E., Seck, H. A. & Hofmann, A. W. (2003). Significance of large, refractory dunite bodies in the upper mantle of the Bay of Island Ophiolite. Geochemistry, Geophysics, Geosystems 4, paper number 2001GC000277.
Sun, S. S. & McDonough, W. F. (1989). Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications 42, 313–345.
Tait, S. R. (1985). Fluid dynamic and geochemical evolution of cyclic unit 10, Rhum, Eastern Layered Series. Geological Magazine 122, 469–484.[Abstract]
Tiezzi, L. J. & Scott, R. B. (1980). Crystal fractionation in a cumulate gabbro, Mid-Atlantic Ridge, 26°. Journal of Geophysical Research 85(B10), 5438–5454.
Van der Laan, S. R., Lower, M. F. J. & Koster van Groos, A. F. (1989). Experimental evidence for the origin of boninites: near-liquidus phase relations to 7·5 kbar. In: Crawford, A. J. (ed.) Boninites and Related Rocks. London: Unwin Hyman, pp. 112–147.
Varvalvy, V., Hébert, R. & Bédard, J. H. (1996). Interactions between melt and upper mantle peridotites in the North Arm Mountain massif, Bay of Islands ophiolite, Newfoundland, Canada: implications for the genesis of boninitic and related magma. Chemical Geology 129, 71–90.[CrossRef][Web of Science]
Varvalvy, V., Hébert, R., Bédard, J. H. & Laflèche, M. R. (1997). Petrology and geochemistry of pyroxenite dykes in upper mantle peridotites of the North Arm Mountain Massif, Bay of Islands Ophiolite, Newfoundland: implications for the genesis of boninitic and related magmas. Canadian Mineralogist 35, 543–570.[Web of Science]
White, W. M. & Dupré, B. (1986). Sediment subduction and magma genesis in the Lesser Antilles: isotopic and trace element constraints. Journal of Geophysical Research 91, 5927–5941.
Williams, H. (1973). Bay of Islands map-area (12G), Newfoundland. Geological Survey of Canada 72-34, 7 pp.
Yoder, H. S. & Tilley, C. E. (1962). Origin of basalt magmas: an experimental study of natural and synthetic rock systems. Journal of Petrology 3, 342–532.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  V. Peterson, J. G. Ryan, and 1997-1998 Research Experiences for Undergraduates Petrogenesis and structure of the Buck Creek mafic-ultramafic suite, southern Appalachians: Constraints on ophiolite evolution and emplacement in collisional orogens Geological Society of America Bulletin, March 1, 2009; 121(3-4): 615 - 629. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||