| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 45 | Number 2 | Pages 439-455 | 2004
© Oxford University Press 2004; all rights reserved
Re–Os and Lu–Hf Isotope Constraints on the Origin and Age of Pyroxenites from the Beni Bousera Peridotite Massif: Implications for Mixed Peridotite–Pyroxenite Mantle Sources
DEPARTMENT OF GEOLOGICAL SCIENCES, DURHAM UNIVERSITY, SOUTH ROAD, DURHAM DH1 3LE, UK
RECEIVED NOVEMBER 15, 2002; ACCEPTED AUGUST 16, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
A suite of pyroxenites from the Beni Bousera peridotite massif, northern Morocco, have been analysed for Re–Os and Lu–Hf isotopic compositions. Measured sections of the massif indicate that pyroxenite layers make up between 1 and 9% by volume of the total outcrop. Clinopyroxenes from two Cr-diopside pyroxenites have unradiogenic Hf isotope compositions (
Hfi 
-7·7 to -8·5) whereas those of the Al-augite suite are more radiogenic (Hfi 9·4 to 25·6). In general, the Nd–Hf isotope compositions of the pyroxenites lie close to the mantle array. One garnet pyroxenite lies significantly below the mantle Hf–Nd isotope array such that it requires an ancient history characterized by high Lu/Hf and Sm/Nd but low Lu/Hf relative to Sm/Nd. As with the Sm–Nd and Rb–Sr systems, parent–daughter and isotopic ratios for the Lu–Hf system have been recently decoupled by a partial melting event associated with transfer of the massif from mantle to crust. This created highly fractionated Sm/Nd and Lu/Hf ratios in many rocks and the pyroxenites can be referred to as ‘residual’. The near-solidus extraction of a siliceous melt from the pyroxenites is also a possible explanation for the orthopyroxene-rich margins to numerous pyroxenite layers, via reaction with peridotite. Pyroxenite Os isotope compositions are much more radiogenic than their host peridotites. In contrast to the non-systematic Nd and Hf model ages, a large portion of the pyroxenite Re–Os model ages cluster between 1·2 and 1·4 Ga, within error of the model ages defined by many Ronda pyroxenites and close to the precise 1·43 ± 0·07 Ga Lu–Hf isochron defined by clinopyroxenes from the peridotites. The Re–Os system thus seems to have been more robust to late-stage melting events that decoupled Sm/Nd and Lu/Hf isotope systematics in the pyroxenites. In contrast to pyroxenites measured from Ronda, some Beni Bousera pyroxenites have relatively radiogenic Os isotope compositions at high Os concentrations (0·18 to >2 ppb), comparable with values reported for some cratonic pyroxene-rich xenoliths. In contrast to cratonic eclogites, most pyroxenites analysed here and those reported in the literature lie close to the mantle Nd–Hf isotope array. The Nd–Sr–Pb–Hf isotopic compositions and stable isotope characteristics of these pyroxenites reflect signatures from recycled oceanic crust and sediment. Hence, mixing of such material, if present within the convecting mantle, with peridotite, could account for some of the heterogeneity seen in oceanic basalts. Small amounts of pyroxenite incorporated into peridotite can also produce the radiogenic Os isotope signatures evident in the source of oceanic basalts. However, these observations alone do not require pyroxenite to be an integral part of the convecting upper-mantle magma source region. The spectrum of Nd, Hf and Os isotope compositions also makes them a suitable component to explain some of the isotopic characteristics of the source regions of ultrapotassic magmas. KEY WORDS: osmium isotopes; hafnium; pyroxenites
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
Pyroxenite layers within orogenic peridotite massifs provide direct evidence of mantle heterogeneity and have been used as key pieces of evidence for ‘marble-cake’ mantle models (Allègre & Turcotte, 1986
; Kellog, 1992). Some recent models to explain the detailed major, trace element and isotopic systematics of mid-ocean ridge basalts (MORB) have advocated contributions from pyroxenites to the melting regime (e.g. Prinzhofer et al., 1989; Langmuir et al., 1992; Lundstrum et al., 1995; Hirschmann & Stolper, 1996; Blichert-Toft et al., 1999a). In addition, recent petrogenetic models for continental ultrapotassic volcanics involve melting of veined lithospheric mantle, in which the vein component is pyroxene-rich (Foley, 1992; Carlson & Nowell, 2001). Because of their potential significance in such geodynamic and petrogenetic models, it is important to improve characterization of the elemental and isotopic systematics of pyroxenites. This will help to constrain the distinguishing geochemical criteria that might indicate a contribution from pyroxenite to magma sources and in turn will allow further testing of marble-cake mantle models.
This study focuses on the Beni Bousera peridotite massif, northern Morocco, as an example of an orogenic peridotite massif with relatively abundant pyroxenites. Previous work on this peridotite massif has revealed that the pyroxenites probably originate from a variety of sources and are likely to be variable in age (Loubet & Allègre, 1982; Kornprobst et al., 1990; Pearson et al., 1993; Kumar et al., 1996; Blichert-Toft et al., 1999a). In general, the age of the pyroxenite layers has not been well constrained, yet this information is important in the context of the applicability of the massif to ‘marble-cake mantle’ or ‘plum-pudding’ mantle models. In this study, we have analysed the Lu–Hf and Re–Os isotopic compositions of a suite of well-characterized pyroxenites (Pearson et al., 1993) from the Beni Bousera peridotite massif. Our objectives were to try to further constrain the timing of formation of pyroxenite formation and to evaluate the Hf and Os isotopic characteristics of such rocks, in terms of them being a possible component in the sources of magmas originating from the oceanic and continental lithospheric mantle.
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
The Beni Bousera peridotite massif is situated in the Rif mountains of northern Morocco and is part of the Betic–Rif orogenic belt (Fig. 1). The massif is surrounded by migmatitic graphite–sillimanite–garnet gneisses (kinzigites), that are part of a lower-crustal assemblage exposed with the peridotite body. The tectonic setting and emplacement age of the Beni Bousera massif is identical to that of the Ronda massif in southern Spain (Fig. 1). The two peridotite bodies are compositionally very similar, differing mainly in their degree of mineralogical equilibration during emplacement. Both massifs contain pyroxenite layers with graphitized diamonds (Pearson et al., 1989
; Davies et al., 1993) and hence ultimately originate from the diamond stability field. It is possible, indeed likely, that Beni Bousera and Ronda were derived from very similar portions of lithospheric mantle underpinning the Betic–Rif region prior to lithospheric delamination and extension. Further evidence to support this idea will be presented below.
|
|
ABUNDANCE OF PYROXENITE LAYERS AND THEIR MINERALOGICAL ZONATION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
Detailed descriptions of the field occurrence, mineralogy and petrology of the Beni Bousera pyroxenite suite have been given by Kornprobst et al. (1969)
, Pearson (1989), Pearson et al. (1993), Kumar et al. (1996) and Pearson & Nixon (1996). Here, we concentrate on a few salient features that we believe are relevant to petrogenetic models for the pyroxenites. Carefully measured sections within the peridotite massif indicate that locally pyroxenites can constitute between 1 and 9% by thickness of the section (Pearson & Nixon, 1996; Fig. 2). The value of 9% is probably a minimum for the section concerned because of the likely thinning of some layers to extents that render them undiscernible in the field. Despite the high abundance of pyroxenites in some places, a value of 1–3% is the favoured estimate for the massif as a whole (Kornprobst, 1969; Allègre & Turcotte, 1986; Pearson et al., 1993). These lower values are similar to the abundance of pyroxenites estimated from measured sections in the Horoman peridotite (Takazawa et al., 1999). In terms of frequency distribution, thinner layers (<20 cm) are much more abundant than thick layers (20–260 cm thick; Fig. 2; Allègre & Turcotte, 1986) as observed for Horoman pyroxenites (Takazawa et al., 1999). In the thinner layers (<50 cm) the cumulative frequency distribution approximates to log-normal. For thicker layers this relationship breaks down. As noted by Allègre & Turcotte (1986), the number of very thin layers (<1 cm) is low, probably because of their destruction by convection/diffusive re-equilibration and also the difficulty in identifying them in the field.
|
A notable feature of the Beni Bousera pyroxenites is the mineralogically zoned character of some layers compared with a more homogeneous mineralogy in others (Kornprobst, 1969
; Pearson et al., 1993; Kumar et al., 1996). This mineralogical zonation has been previously ascribed to combinations of high-pressure crystal fractionation within magmatic veins and melt-rock reaction with the surrounding peridotite. Recently, experiments that melt eclogite/peridotite ‘sandwiches’ have been carried out that shed new light on the likely origin of the mineralogical variation within these veins (Yaxley & Green, 1998).
Near-solidus partial melts of pyroxenite and eclogite are siliceous and hence highly reactive towards the host peridotite (Yaxley & Green, 1998). Migration of siliceous melts from pyroxenites into peridotite will increase the modal proportion of orthopyroxene in the wall rock and may result in mantling of the pyroxenitic residue by orthopyroxenite (Yaxley & Green, 1998). Symmetrical, thick orthopyroxenite margins tend to be present on garnet pyroxenite layers that are moderately light rare earth element (LREE) depleted, suggesting low fractions of melt loss (Pearson, 1989; Pearson et al., 1993). The highly LREE-depleted garnet pyroxenites indicate greater extents of partial melting and tend not to have orthopyroxene-rich margins in many cases. This is consistent with larger degrees of melting resulting in less SiO2-rich melts. The occurrence of orthopyroxenite margins to websterite and garnet clinopyroxenite layers is widespread at Beni Bousera (Kornprobst, 1969; Pearson et al., 1993) and could probably originate in this way rather than as a result of high-pressure crystal fractionation as proposed previously by Pearson et al. (1993). None the less, crystal fractionation at upper-mantle depths is likely to have operated during the intrusion and crystallization of the layers in the peridotite host.
|
SAMPLES AND ANALYTICAL TECHNIQUES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
Samples
Samples were selected from the extensive pyroxenite suite analysed by Pearson et al. (1993)
and the reader is referred to this work for more petrological detail; a summary is provided in Table 1. Briefly, three websterites belonging to the Cr-diopside pyroxenite suite were studied together with four garnet pyroxenites and one websterite from the Al-augite pyroxenite suite. Major and trace element geochemistry of the whole rocks and minerals have been presented by Pearson et al. (1993) and Pearson & Nixon (1996). All layer thicknesses exceeded 10 cm and two garnet pyroxenites were in excess of 2 m thick (Table 1). All samples analysed here were taken from the centre of layers. The letter M denotes a sample from the layer centre, taken as part of a sample set collected across a given layer. GP147 M contains graphite pseudomorphs after diamond, up to 15% by volume. This observation defines the ultimate derivation of the massif to be from within the diamond stability field (Pearson et al., 1989; Pearson & Nixon, 1996).
|
All clinopyroxenes within the pyroxenites contain abundant exsolution lamellae of orthopyroxene. In addition, clinopyroxenes within the garnet pyroxenites commonly contain exsolved blebs of garnet. Discrete garnet crystals frequently contain 10 µm needles of rutile; however, this phase was not observed as an interstitial phase in the samples analysed. Bulk isotopic compositions for garnet pyroxenites are calculated using a mode of 60% clinopyroxene and 40% garnet (Table 2). This ratio is typical of modes estimated visually and calculated from mineral and bulk major element compositions. The coarse grain size and modal heterogeneity within layers makes it very difficult to define a mode for any given layer and the 60:40 ratio is the best estimate for any given garnet clinopyroxenite. Not all samples analysed for Re–Os isotopes in this study have corresponding Lu–Hf isotope analyses. This simply reflects the lack of remaining sample for adequate separation of pure mineral separates.
|
Analytical techniques
Techniques used in mineral separation and picking have been described by Pearson et al. (1993)
. In contrast to the leaching procedure employed by Pearson et al. (1993), we used only a 40°C 6 M HCl ultrasonic leach to clean minerals for Hf analysis because of concern that the dilute HF–HCl leaching procedure used for Sr–Nd isotope analysis (Pearson et al., 1993) might leach Hf from the separates.
Splits (10 mg) of the leached mineral separates were taken for trace element analysis by inductively coupled plasma mass spectrometry (ICP-MS) following established procedures (Ottley et al., 2003), except that we used ultra-pure acids and ultra-clean work environments throughout. The final solutions were diluted to 50 ml of 3·5% HNO3 and run directly on the ICP-MS system. Repeat analyses of standards show that Lu/Hf is reproducible to 1·5% (1 x relative standard deviation) using this procedure. Given the current disagreement on the half-life of 176Lu, this level of parent–daughter isotope ratio precision is more than adequate for our purposes.
The bulk (95%) of the leached mineral separate was used for Lu–Hf isotope analysis. We used a simple two-column pre-concentration procedure that employs a 5 ml cation separation as the first step, using 1N HF–1N HCl to elute Hf and 6N HCl to elute Nd, followed by a mixed sulphuric acid–H2O2 anion column for final purification of the Hf. The procedure provides a rapid, low blank method for the analysis of Sr, Nd and Hf isotopes in geological samples in two column steps (Dowall et al., 2003). Hf blanks were 60 pg and are insignificant for the levels of Hf analysed here. Measurements were made on a ThermoFinnigan Neptune plasma ionization multi collector mass spectrometer. Whole rocks and clinopyroxenes were analysed using an ESI PFA-50 nebulizer with quartz, cyclonic Scott-type double pass spray chamber. Twelve analyses of the JMC-475 standard during this session gave a 176Hf/177Hf value of 0·282150 ± 7 (2 S.D.; i.e. 26·5 ppm external reproducibility). Garnets were analysed using a Cetac Aridus desolvating nebulizer and a high-sensitivity skimmer cone that produced a sensitivity of 470 V/ppm Hf at an uptake rate of 80 µl/min, for the analytical session in question. Nine analyses of the JMC-475 standard during this session gave a 176Hf/177Hf value of 0·282148 ± 3 (2 S.D.; i.e. 11 ppm external precision). 176Hf/177Hf values were corrected to an accepted value of 0·282160 (Blichert-Toft et al., 1997; Nowell et al., 1998). The long-term average for the JMC 475 standard on the Durham Neptune system is 0·282155 ± 9 (n = 195; Nowell et al., 2003a) and is within 17 ppm of the accepted value. Full details of mass spectrometry procedures, sensitivity and instrumental performance have been given by Nowell et al. (2003a).
Whole-rock Re–Os chemical procedures followed the methods of Pearson & Woodland (2000). Samples were analysed by negative thermal ionization mass spectrometry (N-TIMS) on a ThermoFinnigan Triton mass spectrometer. All analyses were carried out on the secondary electron multiplier, via ion-counting, in peak-hopping mode using a Ba(OH)2 activator solution. Using this procedure, our long-term mean 187Os/188Os value for 161 runs of the University of Maryland College Park standard at signal sizes equivalent to those of the samples was 0·11383 ± 32 (2 S.D.; 2·8 per mil) and is within error of the value of 0·113791 ± 3 produced from static Faraday runs of large loads by Walker et al. (1997). The mean 189Os/188Os over this period is 1·21976 ± 192 (2 S.D.; equates to 1·6 per mil).
|
Lu–Hf AND Sm–Nd ISOTOPE SYSTEMATICS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
Element partitioning and isotope systematics
Measured whole-rock vs calculated bulk isotopic compositions
For comparative purposes, bulk compositions have been calculated for the garnet clinopyroxenites. This allows the bulk Nd–Hf isotope compositions to be easily compared with the websterites and with basaltic rocks. We use calculated bulk compositions rather than the measured whole-rock compositions presented in Table 2 for various reasons. First, there is evidence for late-stage, grain-boundary, LREE–HFSE (high field strength element) enriched metasomatic phases in orogenic peridotites (e.g. Reisberg et al., 1989
; Bodinier et al., 1996). This can result in measured whole rocks having less radiogenic Nd isotope compositions than their clinopyroxene, in assemblages that contain no primary LREE-enriched phases (Pearson, 1989). Second, whole rocks show considerable evidence of serpentinization and other types of alteration. This can also be a source of LREE-enriched crustal material. Crustal input to the whole-rock powders is clearly shown to be the case for Sr isotopes by detailed leaching studies (Zindler et al., 1983; Pearson et al., 1993). Hence, it is likely that the measured whole-rock pyroxenites will not faithfully record the Nd or Hf isotope composition of the pyroxenite and we prefer to use calculated bulk isotopic compositions for the garnet pyroxenites. In practice, the measured whole-rock compositions for the rocks studied here are either within error, or close to within error (0·5 epsilon units) of the calculated bulk compositions (Table 2). In some cases, the measured whole rocks are further off the mantle Nd–Hf isotope array than the calculated bulk rocks. The small differences between measured and calculated bulk Nd–Hf isotope characteristics in no way alter the conclusions arrived at below. The following discussion will refer to the calculated bulk values when discussing pyroxenite bulk isotopic compositions. Because of the likely open-system behaviour of the measured whole rocks, they will not be included in any isochron regressions.
Mineral equilibria and isotopic compositions
Lu/Hf and Sm/Nd are partitioned between garnet and clinopyroxene in the manner expected, i.e. garnets have considerably greater Lu/Hf and Sm/Nd than their coexisting pyroxenes. Garnet 176Lu/177Hf ratios are predictably high (0·41–1·42; Table 2). Clinopyroxenes equilibrated with garnet have correspondingly low 176Lu/177Hf (0·0032–0·016) compared with clinopyroxenes from garnet-free assemblages (0·012–0·02). 147Sm/144Nd values of some clinopyroxenes are among the highest ever measured in mantle clinopyroxenes and are testament to the extreme LREE depletion of some of the pyroxenites, as noted by Loubet & Allègre (1982) and Pearson et al. (1993). This feature led those workers to propose a late-stage partial melting event to explain the LREE-depleted nature of many pyroxenites. The presence of orthopyroxenite reaction rims in many of the Beni Bousera pyroxenite layers supports this idea.
Garnets from garnet pyroxenites have more radiogenic measured Hf than their coexisting clinopyroxenes, as expected. There is evidence of some minor, late-stage disturbance of both Lu–Hf and Sm–Nd systems from the small to moderate differences in initial isotopic compositions at the preferred emplacement age of 22·5 Ma (Table 2) that is reflected in some very anomalous isochron ages (see below). Clinopyroxenes from the two Cr-diopside pyroxenites have unradiogenic Hf isotope compositions (Hfi -7·6 to -8·4; Table 2) whereas those of the Al-augite suite are more radiogenic (Hfi +9·6 to +26·3). This range extends up Hfi = +42 when the data of Blichert-Toft et al. (1999a) are considered. These values are much more restricted than the values obtained for Archaean eclogite xenoliths (cpx Hfi up to 166; Jacob et al., 2002) and garnet–spinel (alkremite) mantle xenoliths from kimberlites (garnet Hfi up to +24 000; Nowell et al., 2003a).
Beni Bousera pyroxenites show much greater Sr–Nd isotopic heterogeneity than their host peridotites (Pearson et al., 1993). In contrast, the range of calculated bulk pyroxenite Hf isotope compositions is considerably more restricted than the large range shown by clinopyroxenes from the peridotites (Hfi 14–209; Pearson & Nowell, 2003). This is because the peridotites have surprisingly radiogenic Hfi values that are much more variable than their Ndi values (+3·2 to + 14·9; Pearson et al., 1993).
Measured Hfi values for the websterite clinopyroxenes and calculated bulk Hfi values for the garnet clinopyroxenites, when combined with their Nd isotopic compositions, can be compared with the composition of oceanic basalts that scatter about the so-called Hf–Nd ‘mantle array’ (Fig. 3). Deviation from the mantle array can be denoted using the 
Hfi notation of Beard & Johnson (1993). Beni Bousera pyroxenites scatter around the mantle Hf–Nd isotope array (Fig. 3). The garnet pyroxenites analysed by Blichert-Toft et al. (1999a) plot close to the mantle array, with the Hfi values close to zero. Three garnet pyroxenites (this study) plot at varying distances off the mantle array. GP139 and GP147 have Hfi values of -3·2 to -4·3, and plot between the fields of MORB and ocean island basalt (OIB). More extreme is GP37, which plots well to the right of the mantle Nd–Hf array with a Hfi value of -13·7. This is one of the largest deviations below the mantle array so far observed for a mantle sample. The radiogenic Hfi value indicates that Lu/Hf was supra-chondritic but the displacement below the mantle array indicates lower levels of Lu/Hf fractionation compared with Sm/Nd fractionation, relative to the mantle array. In contrast to the garnet-bearing pyroxenites, the garnet-free Al-augite websterite GP236 plots above the mantle array with a Hfi value of +6·3, lying at the outer edge of the OIB field.
|
Although the two Cr-diopside websterites have unradiogenic Nd and Hf isotope compositions, relative to Bulk Earth, they plot on an extension of the mantle array, in the field occupied by upper-crustal rocks (Vervoort et al., 1999
). These are the only pyroxenite compositions occupying the ‘enriched’ part of the Nd–Hf isotope diagram.
Inter-mineral Sm–Nd and Lu–Hf isochrons
The essentially bi-mineralic nature of the garnet pyroxenites combined with their high-temperature evolution provide the opportunity to obtain two-point inter-mineral isochrons for the Beni Bousera pyroxenites (Polvé, 1983; Kumar et al., 1996). Recently, Blichert-Toft et al. (1999a) produced six Lu–Hf isochrons for the Beni Bousera pyroxenites. In this study, we have produced an additional three Lu–Hf and Sm–Nd clinopyroxene–garnet isochrons. The combined isochron regressions for all datasets are presented in Table 3. All uncertainties on ages, including means, are quoted at the 95% confidence limit and the Lu–Hf ages of Blichert-Toft et al. (1999a) have been recalculated to a 
value of 1·865 x 10-11 (Scherer et al., 2001). Six Sm–Nd garnet–clinopyroxene isochrons (excluding GP37) give a mean of 20·9 ± 4·0 Ma (2 S.D.), within error of the mean of eight Lu–Hf isochrons (excluding M5-15) of 24·1 ± 8·6 (2 S.D.; Table 3; Fig. 4). The variability of the two-point isochron ages, in particular the large disagreement between the Sm–Nd and Lu–Hf isochron ages for the most aberrant samples, e.g. GP37 (Table 3), indicates that additional, open-system processes were affecting Sm–Nd and Lu–Hf equilibrium in these rocks during or after cooling. The best agreement between Sm–Nd and Lu–Hf isochrons is shown by samples GP147 (Lu–Hf isochron = 20·8 ± 2·5 Ma; Sm–Nd isochron = 19·1 ± 1·1 Ma) and M5-101 (Lu–Hf isochron = 25·3 ± 1·2 Ma; Sm–Nd isochron = 24·0 ± 4·3 Ma). Unfortunately, the isochron ages for these two samples are significantly discrepant and it does not seem sensible to use the level of agreement between Lu–Hf and Sm–Nd systems as an indication of accuracy. The most precise isochron is provided by the Lu–Hf isochron for GP139, where the extreme Lu/Hf fractionation between garnet and clinopyroxene provides an age of 22·5 ± 1·1 Ma. Blichert-Toft et al. (1999a) produced a similarly precise, but older garnet–clinopyroxene isochron age of 25·3 ± 1·2 Ma for sample M5-101 (Table 3). However, the garnet and clinopyroxene in this layer did not show the extreme Lu/Hf fractionation of GP139. Because the GP139 mineral pair show the most extreme Lu/Hf fractionation, we argue that they will be the least readily disturbed by secondary processes and hence might provide the best estimate of when the Beni Bousera massif passed through the Lu/Hf blocking temperature for the garnet–clinopyroxene assemblage. We acknowledge that the GP139 isochron age is almost within error of the age produced by Blichert-Toft et al. (1999a) and so, in reality, either age could be valid.
|
|
|
Re–Os ISOTOPE SYSTEMATICS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
Although Os isotope studies of orogenic lherzolite massifs have been carried out previously (Reisberg et al., 1991
; Kumar et al., 1996; Roy-Barman et al., 1996; Becker et al., 2001), only two individual pyroxenite layers from Beni Bousera have been analysed for both Re and Os. We present an additional seven analyses, including two Cr-diopside pyroxenites. Kumar et al. (1996) made three analyses of a single, complex zoned layer from the northern area of Beni Bousera. These analyses are included in our plots.
Os contents of the Beni Bousera pyroxenites are mostly in the range from 0·03 to 0·6 ppb (Fig. 5). One Cr-diopside websterite (GP30) has 2·2 ppb Os. This total range is similar to that found by Roy-Barman et al. (1996) but those workers did not note which petrogenetic groups their pyroxenites belonged to. Typical common Os concentrations measured in Beni Bousera pyroxenites are significantly higher than those found for Ronda (Reisberg et al., 1991). Several Beni Bousera pyroxenites have relatively radiogenic Os isotope compositions at high Os concentrations (0·18 to >2 ppb), comparable with values reported for some cratonic pyroxene-rich xenoliths (Carlson & Irving, 1994). For the Beni Bousera pyroxenites, the high Os contents suggest crystallization in a sulphur-saturated environment. This is confirmed by the presence of abundant sulphide inclusions within many of the clinopyroxenes.
|
Re contents and hence Re/Os ratios in the pyroxenites are highly variable (Fig. 5). Re values as low as 0·035 ppb (GP37) are lower than those reported from basaltic magmas (Shirey & Walker, 1998
), whereas the value of 2·73 ppb for GP147 is considerably higher than for typical magmatic rocks. Elevated Re contents have been reported for pyroxenites from other massifs. A whole-rock Re content of 3·5 ppb has been reported for a pyroxenite from the Horoman peridotite by Saal et al. (2001), whereas Roy-Barman et al. (1996) reported a Re concentration of 12·6 ppb for a garnet separate from a Lherz pyroxenite layer. This could provide support for the idea that garnet has a very high partition coefficient for Re (Righter & Hauri, 1999) and that the pyroxenites may have formed by accumulation of garnet. However, the Re content of some garnetiferous pyroxenites is very low (e.g. GP37) and Re abundance does not directly correlate with either whole-rock Yb or garnet content. These observations argue against a strong garnet control on Re. A possible explanation for the lack of strong garnet control in the Beni Bousera garnet pyroxenites is that garnet may be exsolved from a higher-T aluminous pyroxene in some pyroxenites, whereas it is a liquidus phase in others. An additional complicating factor is that Beni Bousera pyroxenites contain abundant sulphide which may also account for, or contribute towards the high Re contents of the whole rocks. Micro-sulphide inclusions in the garnets may also be present.
As observed previously for orogenic peridotites (Reisberg et al., 1991; Roy-Barman et al., 1996; Saal et al., 2001) the Os isotopic compositions of the pyroxenites are generally significantly more radiogenic than those reported for the host peridotites (Table 3; Fig. 6). Although the most radiogenic 187Os/188Os ratios for pyroxenites are those reported from the Ronda massif (Reisberg et al., 1991), the values for Beni Bousera and Ronda pyroxenites overlap (Fig. 6). Additional sampling and analysis would probably reveal very similar isotopic ranges given the similarity of Nd–Sr isotope systematics and emplacement ages of the two massifs. High 187Os/188Os in the Beni Bousera pyroxenites is supported by high 187Re/188Os such that a positive correlation is defined on a Re–Os isochron diagram with a slope equating to an age of 980 ± 330 Ma. This relationship and its significance will be addressed below in more detail. The high 187Os/188Os values of the pyroxenites are significantly more radiogenic than modern-day, uncontaminated oceanic basalts (e.g. Hauri & Hart, 1993; Marcantonio et al., 1993; Widom & Shirey, 1996; Widom et al., 1999; Fig. 6).
|
Os, of Beni Bousera pyroxenites compared with their host peridotites and pyroxenites from the Ronda massif (Reisberg et al., 1991 |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
The age and evolution of the Beni Bousera pyroxenites
Lu–Hf and Sm–Nd isochrons
An 40Ar/39Ar plateau age of 21·5 ± 1·7 Ma was obtained from a plagioclase separate from a sillimanite–garnet gneiss surrounding the Beni Bousera massif (Pearson et al., 1993
) and is taken to indicate the age of cooling through the 300–400°C K–Ar blocking temperature, corresponding to the final stages of crustal emplacement of the peridotite body into the crust. The high equilibration temperatures recorded by the pyroxenite assemblages (>900°C for rim compositions; Pearson, 1989; Pearson & Nixon, 1996) imply rapid cooling of the massif. Hence, it is likely that any garnet–clinopyroxene isochron for the Sm–Nd or Lu–Hf systems will ideally record the timing of removal of the peridotite body from the mantle into the crust, followed by rapid exhumation. As such, the ages are likely to approximate to the massif emplacement age. The variation of the Lu–Hf and Sm–Nd isochron systematics was discussed above. In general, the Lu–Hf isochrons provide the most precise ages (Table 3).
The two-point 22·5 ± 1·1 Ma Lu–Hf isochron for GP139, or the older 25·3 ± 1·2 Ma isochron for M5-101 (Table 3; Blichert-Toft et al., 1999a) are consistent with the higher blocking temperature for the Lu–Hf system, whereas the younger Lu–Hf isochron ages for GP147 and GP37 are not. The similarity of the GP139 isochron to the precise Sm–Nd isochron obtained for the Ronda massif (Zindler et al., 1983) leads us to take the relatively precise Lu–Hf isochron age of 22·5 Ma for GP139 as the ‘emplacement age’ of the Beni Bousera massif. The small difference between the 40Ar/39Ar plateau age and the Lu–Hf isochron age implies cooling rates of the order of 400°C/Myr if the blocking temperature for Lu–Hf in the garnet–clinopyroxene system is of the order of 800°C. If the older age of 25·3 Ma is viewed as more reliable, this decreases cooling rates by almost a factor of two. The errors involved do not allow much certainty to be attached to these estimates. The 22·5 ± 1·1 Ma Lu–Hf age for GP139 is in closer agreement with the Sm–Nd age of 21·5 ± 1·8 Ma determined for a Ronda garnet pyroxenite (Zindler et al., 1983). The similarity of these ages indicates the approximate synchroneity of emplacement of the two massifs into the crust.
Re–Os whole-rock isochrons and model age systematics
Pearson et al. (1993) noted the complex Sm–Nd isotope systematics of the Beni Bousera pyroxenites. Nd model ages (relative to depleted mantle or CHUR) are extremely variable and suggest that the pyroxenite suite as a whole could not have been derived simultaneously from any isotopically homogeneous source with their present Sm/Nd ratios. The same observation can be made for the Lu–Hf system because model ages for calculated bulk pyroxenite compositions (from garnet–clinopyroxene pairs), or from clinopyroxene in websterites, are extremely variable.
Loubet & Allègre (1982) and Pearson et al. (1993) suggested that a recent partial melting event may have disrupted the parent–daughter ratios. The recent nature of this event means that insufficient time has elapsed to allow its expression isotopically, thereby decoupling parent–daughter and isotopic ratios. This results in considerable variation in model age and whole-rock isochron systematics. The evidence for this event is the extreme LREE depletion shown by some pyroxenite layers, together with the orthopyroxene-rich margins of numerous layers, which may document the extraction of a dacitic near-solidus melt, leaving a residual pyroxenite. This partial melting event, particularly if of a non-equilibrium nature, affecting pyroxenes that had probably partially exsolved orthopyroxene, may contribute to the extreme inter-mineral Lu/Hf–Sm/Nd fractionations observed between garnet and clinopyroxene that differ significantly from experimental values (Blichert-Toft et al., 1999a).
Regression of the Beni Bousera pyroxenite whole-rock Re–Os data does not produce a line that has a high probability of fit. A model 3 regression (assuming scatter as a result of assigned errors and variation in initial Os isotope ratio) of all the Beni Bousera pyroxenite data, including those of Kumar et al. (1996), gives an age of 980 ± 330 Ma (2
). Initial ratio variation is likely to be highly correlated with Re/Os, such that artificial trends can be generated in samples that are unrelated to each other. These trends are a particular danger when evaluating low-probability-of-fit regressions such as those that can be made with the pyroxenite data. More useful information can be obtained by examining the Re–Os model age systematics.
Although variable, five of the nine whole-rock pyroxenites [including the layer studied by Kumar et al. (1996) as one sample] have Re–Os model ages ranging between 1 and 1·4 Ga (Table 4). The melts from which the pyroxenites originally crystallized may not have had Os isotopic ratios that fell exactly on the mantle evolution curve. This is likely given their complex Nd–Hf isotope systematics and highly varied oxygen isotope compositions. This will have little effect on the model ages calculated for extremely radiogenic pyroxenites, but could be important for samples with relatively unradiogenic compositions such as GP30 and GP194 M. Interestingly, four out of seven pyroxenites from the Ronda massif (Reisberg et al., 1991) have whole-rock Re–Os model ages in this range. The Re–Os model age systematics are much more coherent than Sm–Nd and Lu–Hf model ages in the pyroxenites. This observation suggests that the Re–Os isotopic system might be more robust to disturbance from a late-stage partial melting event than the Lu–Hf and Sm–Nd systems. Re and Os may be relatively unfractionated by the extraction of a low-degree, S-undersaturated melt from the pyroxenites. Any small fractionation produced by low degrees of partial melting of the pyroxenites may be insufficient to disturb Re/Os significantly and hence results in only minor alteration of the calculated model ages. In contrast, the presence of residual garnet during this partial melting event will have a greater effect on the fractionation of Sm/Nd and Lu/Hf. The much steeper intersection of the pyroxenite Os isotope evolution curves with the Primitive Mantle evolution curve, compared with the shallow-angle intersections of the Hf and Nd isotope evolution curves, means that minor variations in Re/Os will not greatly affect the Re–Os model age.
|
Whether the Re–Os model ages reflect the timing of pyroxenite formation or some later, major Re–Os fractionation event is debatable. Kumar et al. (1996)
suggested, on the basis of three portions of the same pyroxenite layer yielding model ages in the 1·2–1·3 Ga range, that this represented the formation age of some of the pyroxenite layers. Our more extensive dataset supports the significance of the 1·2–1·3 Ga age in the evolution of the Beni Bousera massif, especially when considered with the 1·4 ± 0·7 Ga Lu–Hf isochron age of the peridotite clinopyroxenes (Pearson & Nowell, 2003), the 1·2 Ga Re–Os model age of a very high-Os, low-Re peridotite (Pearson et al., 2004b) and the 1·4–1·6 Ga Sm–Nd model ages of the surrounding kinzigite crustal units (Polvé, 1983). The coincidence of these ages suggests that a major melting event took place in the peridotites, approximately coincident with differentiation in the overlying crust. This melting event was probably responsible for the removal of the peridotite body from the convecting mantle, into the depleted lithospheric mantle, where the peridotites evolved highly radiogenic Hf isotopic compositions, characteristic of ancient lithospheric mantle (Fig. 3; e.g. Pearson & Nowell, 2004a; Pearson et al., 2003). The pyroxenites may have intruded the peridotites during this initial melting event. Although the Os isotopic compositions of the pyroxenites could have evolved to their present-day values if the peridotites had been their source, the extreme oxygen isotopic compositions of many of the pyroxenites (Table 4; Pearson et al., 1991) rule out any genetic relationship. Hence, pyroxenite formation, as magmatic veins, may have merely been triggered by the large-scale crust–mantle differentiation but the source for many pyroxenites seems to have been from crustal precursors based on oxygen isotope systematics. The significance of this differentiation event regionally is indicated by the similarity in the Ronda pyroxenite Re–Os model ages and by the c. 1·3 Ga melting age indicated by the Ronda peridotite Re–Os and Sm–Nd isotope data (Reisberg et al., 1989; Reisberg & Lorand, 1995). Furthermore, the age systematics of minerals and whole rocks in the different systems applied, together with the petrological similarities between the massifs, strongly indicates that the Beni Bousera and Ronda peridotite bodies were originally derived from a contiguous portion of the asthenosphere, differentiated into the lithospheric mantle c. 1·4–1·3 Gyr ago and then emplaced and exhumed as two separate portions into the crust at 22 Ma.
Re–Os and Lu–Hf isotope constraints on the origin of the pyroxenites
Previous models for the genesis of the Beni Bousera pyroxenites have suggested multiple origins for the different pyroxenite layers (Kornprobst, 1969; Polvé & Allègre, 1980; Allègre & Turcotte, 1986; Kornprobst et al., 1990; Pearson et al., 1993; Kumar et al., 1996). The complex and diverse Re–Os and Lu–Hf isotope systematics found in this study support this notion. If the pyroxenite layers are viewed as oceanic crust thinned by mantle convection and diffusion (Allègre & Turcotte, 1986) then there should be a simple relationship between increasing pyroxenite age and decreasing thickness. Our dataset, combined with that of Kumar et al. (1996), does not show any such relationship and thus we discount the notion that the layers simply represent thinned oceanic crust in favour of models that involve high-pressure crystal–liquid equilibria. This latter origin probably involved derivation of some pyroxenites from recycled oceanic crustal protoliths, as suggested by available oxygen and sulphur isotopic data (Pearson et al., 1991, 1993).
The Hf–Nd isotope systematics of the pyroxenites can be evaluated in terms of recycling models. The relatively low Lu/Hf of MORB compared with their Sm/Nd ratios means that, if no fractionation occurs during subduction, recycled ancient MORB will generate Hf–Nd isotope characteristics that will evolve below the mantle array with time (Fig. 7). Three of the garnet pyroxenites plot below the mantle Hf–Nd array (Figs 3 and 7). GP139 and GP147 plot close to, or within the field occupied by 1–2 Gyr old subducted normal MORB (N-MORB). As such, the initial Nd–Hf isotopic compositions of these two garnet pyroxenites could have originated via evolution from subducted basic–ultrabasic crustal protoliths or high-pressure cumulates. Initial Nd–Hf isotopic compositions for GP37 are significantly outside the field for isotopically evolved subducted MORB. The isotopic composition of this sample cannot easily be generated by any melt, even allowing for 1–3 Gyr of isotopic evolution. However, compilation of measured Lu/Hf and Sm/Nd systematics in mantle minerals (Pearson et al., 2004b) shows that clinopyroxene can have Lu/Hf significantly below chondritic values while retaining moderately high Sm/Nd, such that if GP37 originally crystallized as a high-T pyroxenite and subsequently exsolved garnet on cooling, its long-term isotopic evolution could evolve to the bulk composition observed at 22·5 Myr ago. This indicates the likelihood that crystal fractionation occurring at upper-mantle depths was an important process in forming the pyroxenites.
|
GP188 and GP101 have enriched Hf–Nd isotope signatures, outside the OIB field, but plot on the mantle array. The most likely explanation for these characteristics is that the pyroxenites have incorporated a significant amount (2–5%) of subducted sediment into their source, as proposed by Pearson et al. (1993)
on the basis of high 7/4 Pb isotope systematics combined with unradiogenic Nd and radiogenic Sr isotopic compositions. In the case of Nd–Hf isotopes, the sediment can be constrained to be of turbiditic, or possibly pelagic turbiditic character, rather than true pelagic sediment or red clay (e.g. Vervoort et al., 1999). This is because the high Lu/Hf of pelagic or red clays generates distinctively high Hf isotopic compositions at a given Nd (Fig. 7).
The anomalous oxygen and sulphur isotopic compositions of the pyroxenites indicate a role for recycled oceanic crustal protoliths (Pearson et al., 1993). The relatively high Os contents of some of the pyroxenites combined with their very variable Re contents clearly indicate that the pyroxenites cannot be metamorphosed MORB (e.g. Roy-Barman et al., 1996) because of the low Os contents of most MORB. It is possible that some layers containing corundum might represent metamorphosed aluminous oceanic crustal cumulates (Kornprobst et al., 1990) but many have the petrological and geochemical characteristics of high-pressure crystal fractionation products superimposed upon their recycled crustal isotopic signatures. The high Os abundances reported here suggest that their parental melts must have been sulphur saturated.
Implications for mixed peridotite–pyroxenite source regions
The attraction of involving pyroxenite in the sources of mantle-derived magmas is two-fold: (1) it increases the amount of melt at a given P–T condition as a result of the lower solidi of most pyroxenites [see summary by Hirschmann & Stolper (1996)]; (2) because garnet is stable on the pyroxenite solidus well into the spinel-peridotite stability field (e.g. Irving, 1974), a ‘residual garnet’ elemental signature can be generated by shallow melting. Here we will concentrate on the isotopic character of likely pyroxenite components in mantle source regions, with reference to the Beni Bousera pyroxenites.
The numerous studies conducted on the Beni Bousera and Ronda massifs have not reached agreement on whether the massif was isolated in the lithospheric mantle for 1·3 Gyr, or remained as a fragment of ancient depleted mantle, ‘foundered’ in the asthenosphere. We show above that long-term residence of the massif in the lithospheric mantle is most likely. However, this conclusion does not affect our purpose here. We aim to characterize the isotopic signatures of ancient pyroxenitic mantle material that could have remained as discrete heterogeneities within the convecting or lithospheric mantle to act as potential components in mantle-derived magma sources. Because we have no convincing direct samples of recycled material from within the convecting mantle, the Beni Bousera pyroxenites are probably our best analogues. As such their geochemical characteristics can be used to constrain models relating to the petrogenesis of oceanic basalts (e.g. Hauri, 1996) and of potassic igneous rocks thought to originate from partial melting of veined continental lithospheric mantle (e.g. Carlson et al., 1996; Carlson & Nowell, 2001). Isotopic signatures that may be indicative of pyroxenite contributions to either oceanic or continental magma sources, based on observations from Beni Bousera pyroxenites and other orogenic massif pyroxenites, are as follows.
- Radiogenic Os isotope compositions. Pyroxenites have 187Os/188Os ratios that are almost exclusively more radiogenic than their host peridotites or the range shown by uncontaminated oceanic basalts (Fig. 6).
- Variable oxygen isotopic compositions. Pearson et al. (1993) reported variable
18O values that were lighter (4·9
) and heavier (9·4) than the typical mantle value of 5·2. These values have been subsequently confirmed by laser-fluorination methods (D. P. Mattey & D. G. Pearson, unpublished data, 1994). Some pyroxenites have oxygen isotopic compositions that are indistinguishable from typical mantle.
- Nd and Sr isotopic compositions are very variable and can be similar to peridotite values, more depleted (GP37), or considerably more enriched.
- Hf isotope compositions are also variable and range from within the MORB–OIB field to considerably more radiogenic values.
- Combined Nd–Hf isotope systematics can be distinctive, even if the Nd and Hf values are within the respective ranges of oceanic basalts. One Beni Bousera pyroxenite plots well below the mantle Nd–Hf isotope array, with low Hf. This type of signature is rare in the Beni Bousera samples analysed so far. Ancient garnet-bearing xenoliths from cratonic areas have particularly extreme Hf–Nd isotope signatures that scatter both well above and well below the mantle array (Jacob et al., 2002; Nowell et al., 2003b).
Any one of these signatures in isolation is not particularly distinctive, but a combination of several features provides a strong indication of the possible presence of pyroxenitic material in the source regions (e.g. Carlson et al., 1996; Carlson & Nowell, 2001). The prominence of any of these chemical signatures in a mantle-derived magma obviously depends on the extent and nature of mixing between pyroxenite- and peridotite-derived melts and the relative concentrations and abundance of pyroxenite in the mantle.
It is unlikely that partial melts of pyroxenite layers can be easily extracted from a peridotite matrix because such melts are siliceous and will react with the surrounding peridotite (Yaxley & Green, 1998). At high pressures, the pyrope–omphacite–orthopyroxene thermal divide prevents the mixing of siliceous liquid produced from eclogite/pyroxenite melting with the nepheline-normative picritic liquids produced from metasomatized lherzolite. The siliceous liquids react with and metasomatize the surrounding lherzolites. Eventually, residual phase compositions in the eclogite/pyroxenite and metasomatized lherzolite converge but a modally heterogeneous, refertilized peridotite results (Yaxley & Green, 1998). This refertilized mantle can then produce nepheline-normative melts at the solidus that retain the isotopic memory of the heterogeneous source mixture. Veining within the asthenosphere may be on a much finer scale than the decimeter scale most evident in the lithosphere as sampled by massifs and xenoliths. It is possible that pyroxenite veins may become intimately mixed into the peridotite, making a fertile peridotite composition. This would simplify the problem of extracting melts from a mineralogically zoned source. The intimate physical and diffusional mixing of garnet pyroxenites with spinel-facies peridotites at Beni Bousera and Ronda produces Fe-rich, ‘extra-fertile’ garnet peridotite with olivine Mg-numbers as low as 84–86 (Pearson et al., 1995b). Melts of this material will be strongly influenced by the garnet-pyroxenite ingredient in the mixture and could give Fe-rich melts with radiogenic Os. Such a source would appear simply as Fe-rich, rather than pyroxenitic and may be suitable for the origin of Fe-rich picrites (e.g. Gibson et al., 2000).
Although potentially complex, the processes involved in melting heterogeneous mantle ultimately involve net transfer and mixing of pyroxenite components into peridotite. This ‘mixed’, refertilized source can then remelt. Because we do not have samples of the near-solidus pyroxenite melt that would mix with a peridotite, and because there is abundant evidence in the Beni Bousera and other massifs for physical mixing of pyroxenite into peridotite, we chose to crudely model the interaction by simple mixing of the two end-members. Given that derivative small-degree melts from the pyroxenites are likely to be higher in incompatible elements than their source, the estimates so derived are likely to be overestimates of the required pyroxenite-derived mass flux for Sr, Pb and Nd but underestimates for Os if mixing occurs via melt interacting with peridotite. An additional reason for modelling simple solid–solid mixing is that the measured sections documenting pyroxenite abundance in the Beni Bousera massif (Fig. 2) allow some quantitative bounds to be placed on the mass-balance effects of pyroxenite–peridotite mixtures in a lithospheric mantle environment such as that from which the massif was derived.
In terms of element balance for radiogenic isotope systems, input of 10% ‘typical’ pyroxenite into a fertile peridotite produces an increase in Sr and Nd abundance of 10%, increases Pb abundances by 30%, Hf abundances by 50% and decreases Os abundances by 10% (Fig. 8). Whatever mixing scenario is favoured, it is clear that significant elemental flux enters the ‘mixed’ peridotite from the pyroxenite and this will affect isotopic systematics (Fig. 8). This, in turn, constrains some of the likely isotopic variations to be expected when ancient recycled materials contribute to magma source regions.
|
Becker (2000)
has noted that if the Re/Os and U/Pb isotopic characteristics of subducted eclogites and blueschists are used in mixing calculations to simulate mixing between oceanic crustal material and mantle, excessive amounts (70–90%) of 0·5–2 Gyr old recycled material are required in the source. One reason for this is the very low Os abundances of subducted metabasalts (typically 5 ppt). Although the Os isotopic compositions of the Beni Bousera pyroxenites are not as elevated as values predicted for ancient MORB, they have considerably higher Os concentrations than MORB/metabasalts (by a factor of 10–100), and so have a much more dramatic effect on mixing relationships with peridotite (Fig. 8). This greatly alleviates some of the mass-balance problems identified by Becker (2000). Os isotopic variation remains highly dependent on which pyroxenite composition is used (Fig. 8). Very large amounts (>70%) of a relatively unradiogenic, moderately low-Os pyroxenite such as GP37 are required to elevate the Os isotopic composition above 1% of the original peridotite value. In contrast, high-Os pyroxenites such as GP147, with a radiogenic Os isotope composition, can create >1% variation in Os isotopic composition (i.e. within the range of OIB magmas) by mixing in 5–10% by mass. This mass fraction is within the range observed in the Beni Bousera massif, although this is not necessarily a constraint on the mixing relations in a particular petrogenetic scheme.
Although the largest changes in elemental concentrations are observed for the Hf mass balance over the 1–10% pyroxenite mixing range modelled in Fig. 8, the precise effects on the Hf isotopic composition of the mix are difficult to predict. This is because of the extreme Hf isotopic and elemental abundance variability of the pyroxenites and peridotites. Peridotite–pyroxenite mixtures could lie anywhere within a polygon defined by the extremities of the pyroxenite–peridotite fields of Fig. 3 if the peridotite end-member was a Beni Bousera peridotite. Extreme Nd–Hf isotope compositions (Hf and Nd commonly being > +50 and lying well above and below the mantle array) have been observed for ancient eclogites and alkremites sampled from the lithospheric mantle (Jacob et al., 2002; Nowell et al., 2003b) and indicate the potential variation available for veined melting models within the ancient lithospheric mantle.
In contrast to cratonic eclogites, most pyroxenites analysed here and by Blichert-Toft et al. (1999a) lie close to the mantle Nd–Hf isotope array. Hence, mixing of such material with convecting mantle peridotite, less variable in its Hf isotopic composition than the Beni Bousera peridotites, could account for the more coherent (with respect to the mantle Nd–Hf isotope array) heterogeneity seen in oceanic basalts (MORB and OIB). The coherency of the mantle Nd–Hf isotope array suggests a minimal role for ancient recycled materials with the extreme, diverse isotopic characteristics of ancient (3 Ga) eclogites. Moreover, the coherency of the Nd–Hf isotopic systematics in oceanic basalts suggests that ancient subducted MORB alone is unlikely to be the sole recycled ingredient in their source regions, except for HIMU basalts (Fig. 7), and indicates the likely addition of continental or continent-derived material (Fig. 7; e.g. Blichert-Toft et al., 1999a). The Beni Bousera pyroxenites have a spectrum of radiogenic and stable isotopic characteristics that include combinations of recycled oceanic crustal and sedimentary signatures (Pearson et al., 1993; Tables 2 and 4). This suggests, whether they evolved in the lithospheric mantle or not, that these pyroxenites provide perhaps the closest analogy that we have for any proposed pyroxenitic component in oceanic mantle magma source regions.
|
ACKNOWLEDGEMENTS |
|---|
We thank Chris Ottley for assistance with ICP-MS measurements, and Gareth Davies and Peter Nixon for field assistance and collaboration on other aspects of Beni Bousera. The instrumentation used in this study was funded by HEFCE/NERC grant DUPEEQ to D.G.P. Helpful and detailed reviews by L. Reisberg, J. Blichert-Toft, F. Frey and R. Carlson, and editorial comments by M. Wilson considerably improved the quality and focus of this paper.
|
FOOTNOTES |
|---|
* Corresponding author. E-mail: d.g.pearson{at}durham.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGABUNDANCE OF PYROXENITE LAYERS...SAMPLES AND ANALYTICAL...Lu-Hf AND Sm-Nd ISOTOPE...Re-Os ISOTOPE SYSTEMATICSDISCUSSIONREFERENCES |
|---|
Allègre, C. J. & Turcotte, D. L. (1986). Implications of a two component marble-cake mantle. Nature 323, 123–127.[CrossRef]
Beard, B. L. & Johnson, C. M. (1993). Hf isotope composition of late Cenozoic basaltic rocks from northwestern Colorado, U.S.A.: new constraints on mantle enrichment processes. Earth and Planetary Science Letters 119, 495–509.[CrossRef][Web of Science]
Becker, H. (2000). Re–Os fractionation in eclogites and blueschists and the implications for recycling of oceanic crust into the mantle. Earth and Planetary Science Letters 177, 287–300.[CrossRef][Web of Science]
Becker, H., Shirey, S. B. & Carlson, R. W. (2001). Effects of melt percolation on the Re–Os systematics of peridotites from a Paleozoic convergent margin. Earth and Planetary Science Letters 188, 107–121.[CrossRef][Web of Science]
Blichert-Toft, J. (2001). On the Lu–Hf isotope geochemistry of silicate rocks. Geostandards Newsletter 25, 41–56.[Web of Science]
Blichert-Toft, J., Chauvel, C. & Albarède, F. (1997). Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector-multiple collector ICP-MS. Contributions to Mineralogy and Petrology 127, 248–260.[CrossRef][Web of Science]
Blichert-Toft, J., Albarède, F. & Kornprobst, J. (1999a). Lu–Hf isotope systematics of garnet pyroxenites from Beni Bousera, Morocco: implications for basalt origin. Science 283, 1303–1306.
Blichert-Toft, J., Frey, F. & Albarède, F. (1999b). Hf isotope evidence for pelagic sediments in the source of the Hawaiian basalts. Science 285, 879–882.
Bodinier, J.-L., Merlet, C., Bedini, R. M., Simien, F., Remaidi, M. & Garrido, C. J. (1996). Distribution of niobium, tantalum, and other highly incompatible trace elements in the lithospheric mantle: the spinel paradox. Geochimica et Cosmochimica Acta 60, 545–550.[CrossRef][Web of Science]
Carlson, R. W. & Irving, A. J. (1994). Depletion and enrichment history of sub-continental lithospheric mantle: Os, Sr, Nd and Pb evidence for xenoliths from the Wyoming craton. Earth and Planetary Science Letters 126, 457–472.[CrossRef][Web of Science]
Carlson, R. W. & Nowell, G. M. (2001). Olivine-poor sources for mantle-derived magmas: Os and Hf isotopic evidence from potassic magmas of the Colorado Plateau. Geochemistry, Geophysics, Geosystems 2, 2000GC000128.
Carlson, R. W., Esperanca, S. & Svisero, D. P. (1996). Chemical and Os isotopic study of Cretaceous potassic rocks from southern Brazil. Contributions to Mineralogy and Petrology 125, 393–405.[CrossRef][Web of Science]
Chauvel, C. & Blichert-Toft, J. (2001). A hafnium isotope and trace element perspective on melting of the depleted mantle. Earth and Planetary Science Letters 190, 137–151.[CrossRef][Web of Science]
Davies, G. R., Nixon, P. H., Pearson, D. G. & Obata, M. (1993). Tectonic implications of graphitised diamonds from the Ronda peridotite massif, southern Spain. Geology 21, 471–474.
Dowall, D. P., Nowell, G. M. & Pearson, D. G. (2003). A 2-column procedure for the pre-concentration of Sr, Nd and Hf for isotopic analysis by plasma ionisation multi-collector mass spectrometry. In: Holland, J. G. & Tanner, S. D. (eds) Plasma Source Mass Spectrometry: Applications and Emerging Technologies. Cambridge: Royal Society of Chemistry, pp. 321–337.
Foley, S. F. (1992). Vein-plus-wall-rock melting mechanisms in the lithosphere and the origin of potassic alkaline magmas. Lithos 28, 435–453.[CrossRef][Web of Science]
Gibson, S. A., Thompson, R. N. & Dickin, A. P. (2000). Ferropicrites: geochemical evidence for Fe-rich streaks in upwelling mantle plumes. Earth and Planetary Science Letters 174, 355–374.[CrossRef][Web of Science]
Hauri, E. & Hart, S. R. (1993). Re–Os isotope systematics in HIMU and EMII ocean island basalts. Earth and Planetary Science Letters 114, 253–271.
Hauri, E. H. (1996). Major-element variability in the Hawaiian mantle plume. Nature 382, 415–419.[CrossRef]
Hirschmann, M. M. & Stolper, E. M. (1996). A possible role for garnet pyroxenite in the origin of the ‘garnet signature’ in MORB. Contributions to Mineralogy and Petrology 124, 185–208.[CrossRef][Web of Science]
Irving, A. J. (1974). Geochemical and high pressure experimental studies of garnet pyroxenite and pyroxene granulite xenoliths from the Delegate pipes, Australia. Journal of Petrology 15, 1–40.
Jacob, D. E., Bizimis, M. & Salters, V. J. M. (2002). Lu–Hf isotope systematics of subducted ancient oceanic crust: Roberts Victor eclogites. Geochimica et Cosmochimica Acta 66S1, A360.
Kellog, L. H. (1992). Mixing in the mantle. Annual Review of Earth and Planetary Sciences 20, 365–388.[CrossRef][Web of Science]
Kornprobst, J. (1969). Le massif ultrabasique des Beni Bouchera (Rif Interne, Maroc). Contributions to Mineralogy and Petrology 29, 283–322.
Kornprobst, J., Piboule, M., Roden, M. & Tabit, A. (1990). Corundum-bearing garnet clinopyroxenites at Beni Bousera (Morocco): original plagioclase-rich gabbros recrystallized at depth within the mantle? Journal of Petrology 31, 717–745.
Kumar, N., Reisberg, L. & Zindler, A. (1996). A major and trace element and Sr, Nd and Os isotopic study of a thick pyroxenite layer from the Beni Bousera ultramafic complex of N. Morocco. Geochimica et Cosmochimica Acta 60, 1429–1444.[CrossRef][Web of Science]
Langmuir, C. H., Klein, E. M. & Plank, T. (1992). Petrological constraints on melt formation and migration beneath mid-ocean ridges. In: Phipps Morgan, J., Blackman, D. & Sinton, J. L. (eds) Mantle Flow and Melt Generation at Mid-Ocean Ridges. Geophysical Monograph, American Geophysical Union 71, 183–280.
Loubet, M. & Allègre, C. J. (1982). Trace elements in orogenic lherzolites reveal the complex history of the upper mantle. Nature 298, 809–811.[CrossRef]
Lundstrum, C., Gill, J., Williams, Q. & Perfit, M. (1995). Mantle melting and basalt extraction by equilibrium porous flow. Science 270, 1958–1961.
Marcantonio, F., Zindler, A., Elliott, T. & Staudigel, H. (1995). Os isotope systematics of La Palma, Canary Islands; evidence for recycled crust in the mantle source of HIMU ocean islands. Earth and Planetary Science Letters 133(3–4), 397–410.[CrossRef][Web of Science]
Nowell, G. M., Kempton, P. D., Noble, S. R., Fitton, J. G., Saunders, A. D., Mahoney, J. J. & Taylor, R. N. (1998). High precision Hf isotope measurements of MORB and OIB by thermal ionisation mass spectrometry: insights into the depleted mantle. Chemical Geology 149, 211–233.[CrossRef][Web of Science]
Nowell, G. M., Pearson, D. G., Ottley, C. J., Dowall, D. P. & Schwieters, J. (2003a). Long-term performance characteristics of a plasma ionisation multi-collector mass spectrometer: the ThermoFinnigan Neptune. In: Holland, J. G. & Tanner, S. D. (eds) Plasma Source Mass Spectrometry. Special Publication of the Royal Society of Chemistry, London 307–320.
Nowell, G. M., Pearson, D. G., Jacob, D. J., Spetsius, Z. V., Nixon, P. H. & Haggerty, S. E. (2003b). The origin of alkremites and related rocks: a Lu–Hf, Rb–Sr and Sm–Nd isotope study. Extended Abstracts, 8th International Kimberlite Conference, Victoria, FLA 0271.
Ottley, C. J., Pearson, D. G. & Irvine, G. J. (2003). A routine procedure for the analysis of REE and trace elements in geological samples by ICP-MS. In: Holland, J. G. & Tanner, S. D. (eds) Plasma Source Mass Spectrometry: Applications and Emerging Technologies. Cambridge: Royal Society of Chemistry, pp. 221–230.
Pearson, D. G. (1989). The petrogenesis of pyroxenites containing octahedral graphite and associated mafic and ultramafic rocks of the Beni Bousera peridotite massif, N. Morocco. Ph.D. thesis, University of Leeds, 412 pp.
Pearson, D. G. & Nixon, P. H. (1996). Diamonds in young orogenic belts: graphitised diamonds from Beni Bousera, N. Morocco, a comparison with kimberlite-derived diamond occurrences and implications for diamond genesis and exploration. Africa Geoscience Reviews 3, 295–316.
Pearson, D. G. & Nowell, G. M. (2003). Dating mantle differentiation: a comparison of the Lu–Hf, Re–Os and Sm–Nd isotope systems in the Beni Bousera peridotite massif and constraints on the Nd–Hf composition of the lithospheric mantle. Geophysical Research Abstracts 5, 05430.
Pearson, D. G. & Woodland, S. J. (2000). Carius tube digestion and solvent extraction/ion exchange separation for the analysis of PGEs (Os, Ir, Pt, Pd, Ru) and Re–Os isotopes in geological samples by isotope dilution ICP-mass spectrometry. Chemical Geology 165, 87–107.[CrossRef][Web of Science]
Pearson, D. G., Davies, G. R., Nixon, P. H. & Milledge, H. J. (1989). Graphitized diamonds from a peridotite massif in Morocco and implications for anomalous diamond occurrences. Nature 338, 60–62.[CrossRef]
Pearson, D. G., Davies, G. R., Nixon, P. H., Greenwood, P. B. & Mattey, D. P. (1991). Oxygen isotope evidence for the origin of pyroxenites in the Beni Bousera peridotite massif, North Morocco; derivation from subducted oceanic lithosphere. Earth and Planetary Science Letters 102, 289–301.[CrossRef][Web of Science]
Pearson, D. G., Davies, G. R. & Nixon, P. H. (1993). Geochemical constraints on the petrogenesis of diamond facies pyroxenites from the Beni Bousera peridotite massif, north Morocco. Journal of Petrology 34, 125–172.
Pearson, D. G., Snyder, G. A., Shirey, S. B., Taylor, L. A., Carlson, R. W. & Sobolev, N. V. (1995a). Archaean Re–Os age for Siberian eclogites and constraints on Archaean tectonics. Nature 374, 711–713.[CrossRef]
Pearson, D. G., Davies, G. R. & Nixon, P. H. (1995b). Orogenic ultramafic rocks of UHP (diamond facies) origin. In: Coleman, R. G. & Wang, X. (eds) Ultrahigh Pressure Metamorphism. Cambridge: Cambridge University Press, pp. 456–510.
Pearson, D. G., Canil, D. & Shirey, S. B. (2004a). Mantle samples included in volcanic rocks: xenoliths and diamonds, In: Turekian, K. & Holland, H. D. (eds) Treatise on Geochemistry, Volume 2: The Mantle and Core. Amsterdam: Elsevier (in press).
Pearson, D. G., Irvine, G. J., Ionov, D. A., Boyd, F. R. & Dreibus, G. E. (2004b). Re–Os isotope systematics and Platinum Group Element fractionation during mantle melt extraction: a study of peridotite xenoliths from the N. Lesostho and S. Namibian kimberlites, the Vitim volcanic field and massif peridotites from Beni Bousera. Chemical Geology Special Volume: Highly Siderophile Elements (in press).
Polvé, M. (1983). Les isotopes du Nd et du Sr dans les lherzolites orogeniques: contribution à la détermination de la structure et de la dynamique du manteau supérieur. Ph.D. thesis, Université Paris 7.
Polvé, M. & Allègre, C. J. (1980). Orogenic lherzolite complexes studied by 87Rb–87Sr: a clue to understand the mantle convection processes. Earth and Planetary Science Letters 51, 71–93.[CrossRef][Web of Science]
Prinzhofer, A., Lewin, E. & Allègre, C. J. (1989). Stochastic melting of the marble-cake mantle: evidence from local study of EPR at 12 degrees 50 N. Earth and Planetary Science Letters 92, 189–196.[CrossRef][Web of Science]
Ravizza, G. & Turekian, K. K. (1992). The osmium isotopic composition of organic-rich marine sediments. Earth and Planetary Science Letters 55, 3741–3752.
Reisberg, L. C. & Lorand, J. P. (1995). Longevity of sub-continental mantle lithosphere from osmium isotope systematics in orogenic peridotite massifs. Nature 376, 159–162.[CrossRef]
Reisberg, L. C., Zindler, A. & Jagoutz, E. (1989). Sr and Nd isotopic compositions of garnet and spinel bearing peridotites in the Ronda ultramafic complex. Earth and Planetary Science Letters 96, 161–180.[CrossRef][Web of Science]
Reisberg, L. C., Allègre, C. J. & Luck, J.-M. (1991). The Re–Os systematics of the Ronda ultramafic complex in southern Spain. Earth and Planetary Science Letters 105, 196–213.[CrossRef][Web of Science]
Righter, K. & Hauri, E. H. (1998). Compatibility of rhenium in garnet during mantle melting and magma genesis. Science 280, 1737–1741.
Roy-Barman, M. & Allègre, C. J. (1995). 187Os/186Os in oceanic island basalts: tracing oceanic crust recycling in the mantle. Earth and Planetary Science Letters 129, 145–161.[CrossRef][Web of Science]
Roy-Barman, M., Luck, J. M. & Allègre, C. J. (1996). Os isotopes in orogenic lherzolite massifs and mantle heterogeneities. Chemical Geology 130, 55–64.[CrossRef][Web of Science]
Saal, A. E., Takazawa, E., Frey, F. A., Shimizu, N. & Hart, S. R. (2001). Re–Os isotope in the Horoman peridotite: evidence for refertilisation? Journal of Petrology 42, 25–37.
Scherer, E. E., Mezger, K. & Munker, C. (2001). Calibration of the lutetium–hafnium clock. Science 293, 683–687.
Shirey, S. B. & Walker, R. J. (1998). The Re–Os isotope system in cosmochemistry and high-temperature geochemistry. Annual Review of Earth and Planetary Sciences 26, 423–500.[CrossRef][Web of Science]
Takazawa, E., Frey, F. A., Shimizu, N., Saal, A. & Obata, M. (1999). Polybaric petrogenesis of mafic layers in the Horoman peridotite complex, Japan. Journal of Petrology 40, 1827–1851.[CrossRef][Web of Science]
Vervoort, J. D., Patchett, P. J., Blichert-Toft, J. & Albarède, F. (1999). Relationships between Lu–Hf and Sm–Nd isotopic systems in the global sedimentary system. Earth and Planetary Science Letters 168, 79–99.[CrossRef][Web of Science]
Walker, R. J., Morgan, J. W., Beary, E. S., Smoliar, M. I., Czamanske, G. D. & Horan, M. F. (1997). Applications of the 190Pt–186Os isotope system to geochemistry and cosmochemistry. Geochimica et Cosmochimica Acta 61, 4799–4807.[CrossRef][Web of Science]
Widom, E. & Shirey, S. B. (1996). Os isotope systematics in the Azores: implications for mantle plume sources. Earth and Planetary Science Letters 142, 451–465.[CrossRef][Web of Science]
Widom, E., Hoernle, K. A., Shirey, S. B. & Schmincke, H. U. (1999). Os isotope systematics in the Canary Islands and Madeira: lithospheric contamination and mantle plume signatures. Journal of Petrology 40, 279–296.[CrossRef][Web of Science]
Yaxley, G. M. & Green, D. H. (1998). Reactions between eclogite and peridotite: mantle refertilisation by subduction of oceanic crust. Schweizerische Mineralogische und Petrographische Mitteilungen 78, 243–255.[Web of Science]
Zindler, A., Staudigel, H., Hart, S. R., Endres, R. & Goldstein, S. (1983). Nd and Sr isotopic study of a mafic layer from Ronda ultramafic complex. Nature 304, 226–230.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. M.D. Day, D. G. Pearson, C. G. Macpherson, D. Lowry, and J.-C. Carracedo Pyroxenite-rich mantle formed by recycled oceanic lithosphere: Oxygen-osmium isotope evidence from Canary Island lavas Geology, June 1, 2009; 37(6): 555 - 558. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-L. Bodinier, C. J. Garrido, I. Chanefo, O. Bruguier, and F. Gervilla Origin of Pyroxenite-Peridotite Veined Mantle by Refertilization Reactions: Evidence from the Ronda Peridotite (Southern Spain) J. Petrology, May 1, 2008; 49(5): 999 - 1025. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Luguet, D. Graham Pearson, G. M. Nowell, S. T. Dreher, J. A. Coggon, Z. V. Spetsius, and S. W. Parman Enriched Pt-Re-Os Isotope Systematics in Plume Lavas Explained by Metasomatic Sulfides Science, January 25, 2008; 319(5862): 453 - 456. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. Thompson, A. J. V. Riches, P. M. Antoshechkina, D. G. Pearson, G. M. Nowell, C. J. Ottley, A. P. Dickin, V. L. Hards, A.-K. Nguno, and V. Niku-Paavola Origin of CFB Magmatism: Multi-tiered Intracrustal Picrite-Rhyolite Magmatic Plumbing at Spitzkoppe, Western Namibia, during Early Cretaceous Etendeka Magmatism J. Petrology, June 1, 2007; 48(6): 1119 - 1154. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. S. C. Simon, R. W. Carlson, D. G. Pearson, and G. R. Davies The Origin and Evolution of the Kaapvaal Cratonic Lithospheric Mantle J. Petrology, March 1, 2007; 48(3): 589 - 625. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Porreca, J. Selverstone, and K. Samuels Pyroxenite xenoliths from the Rio Puerco volcanic field, New Mexico: Melt metasomatism at the margin of the Rio Grande rift Geosphere, December 1, 2006; 2(7): 333 - 351. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. MONTANINI, R. TRIBUZIO, and R. ANCZKIEWICZ Exhumation History of a Garnet Pyroxenite-bearing Mantle Section from a Continent-Ocean Transition (Northern Apennine Ophiolites, Italy) J. Petrology, October 1, 2006; 47(10): 1943 - 1971. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. M. Harrison, J. Blichert-Toft, W. Muller, F. Albarede, P. Holden, and S. J. Mojzsis Heterogeneous Hadean Hafnium: Evidence of Continental Crust at 4.4 to 4.5 Ga Science, December 23, 2005; 310(5756): 1947 - 1950. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. THOMPSON, C. J. OTTLEY, P. M. SMITH, D. G. PEARSON, A. P. DICKIN, M. A. MORRISON, P. T. LEAT, and S. A. GIBSON Source of the Quaternary Alkalic Basalts, Picrites and Basanites of the Potrillo Volcanic Field, New Mexico, USA: Lithosphere or Convecting Mantle? J. Petrology, August 1, 2005; 46(8): 1603 - 1643. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. RANKENBURG, J. C. LASSITER, and G. BREY The Role of Continental Crust and Lithospheric Mantle in the Genesis of Cameroon Volcanic Line Lavas: Constraints from Isotopic Variations in Lavas and Megacrysts from the Biu and Jos Plateaux J. Petrology, January 1, 2005; 46(1): 169 - 190. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. NOWELL, D. G. PEARSON, D. R. BELL, R. W. CARLSON, C. B. SMITH, P. D. KEMPTON, and S. R. NOBLE Hf Isotope Systematics of Kimberlites and their Megacrysts: New Constraints on their Source Regions J. Petrology, August 1, 2004; 45(8): 1583 - 1612. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||