Journal of Petrology Advance Access published online on July 7, 2008
Journal of Petrology, doi:10.1093/petrology/egn032
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Sr and Pb Isotope Micro-analysis of Plagioclase Crystals from Skye Lavas: an Insight into Open-system Processes in a Flood Basalt Province
Northern Centre for Isotopic and Elemental Tracing, Department of Earth Sciences, University of Durham, Durham Dh1 3LE, UK
Received October 31, 2007; Revised typescript accepted June 10, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
Crystals in many magmatic rocks are heterogeneous in terms of their isotope composition. Detailed crystal-isotope stratigraphy (CIS) studies have shown that crystals act as reliable recorders of their magma source composition and of the pathways of magma interaction, and successfully identify the end-members involved in magmatic systems. In this paper we have analysed cores and rims and single plagioclase crystals from samples of the Skye Volcanic Centre, part of the British Palaeogene Igneous Province. The isotope analyses reveal that plagioclase crystals from different volcanic samples through the lava succession in Skye (Skye Main Lava Series, Preshal More Basalts and Big-Feldspar Lavas) have larger Sr and Pb isotopic ranges than the isotopic ranges found in hundreds of whole-rock analyses. The large Sr and Pb isotopic variation in plagioclase crystals in Skye is due to variable degrees of crustal contamination of the parental magmas during fractional crystallization and during the ascent of such magmas to the surface. The isotope variation of plagioclase crystals in the early lavas (Skye Main Lava Series) reflects assimilation–fractional crystallization processes dominated by the lower crust, whereas plagioclase crystals from later erupted magmas (Big-Feldspar Lavas and Preshal More basalts) show assimilation of upper crust superimposed on previous lower crustal contamination processes. The variability in Sr and Pb isotopes documented in different crystals within the same rocks shows that the crystals have been aggregated in magmas from different sites of storage and differentiation during the ascent of the magmas to the surface.
KEY WORDS: micro-analysis; plagioclase; strontium isotopes; lead isotopes; flood basalts; micro-mill; Skye
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
Bulk isotopic analyses of whole-rocks have been used traditionally to determine the sources of magmatic rocks. However, the isotopic heterogeneity observed in crystal components in magmatic systems demonstrates that whole-rock compositions only represent an ‘averaged’ isotopic composition that does not reveal all the isotopic end-members involved in the magmatic system (Tepley et al., 1999
; Davidson et al., 2007). As such, quantitative models based on whole-rock geochemical variations, no matter how complex, are necessarily incomplete. Crystals within many magmatic rocks are heterogeneous in terms of their isotope composition (e.g. Geist et al., 1988; Tepley et al., 1999; Gagnevin et al., 2005). This heterogeneity has been recognized in crystals of different types (e.g. plagioclase, alkali feldspars, clinopyroxene; Geist et al., 1988) and within crystals of the same type (i.e. core to rim variations; e.g. Tepley et al., 1999; Davidson et al. 2001; Gagnevin et al., 2005). Isotopic variations observed within single crystals reflect growth in an open system (Davidson et al., 1998) in which the magma changes its isotope composition in response to processes such as magma mixing and/or interaction with the crust. Detailed crystal-isotope stratigraphy (CIS) studies (Knesel et al., 1999; Tepley et al., 1999, 2000; Davidson et al., 2001, 2007; Morgan et al., 2007) have shown that crystals can act as reliable recorders of their magma source composition and of the pathways of magma interaction, and be successfully used to identify the end-members involved in the magmatic system. Until recently, CIS studies have focused on systems from single large volcanoes, which commonly contain rocks laden with crystal cargoes (e.g. Davidson et al., 2007). The larger magmatic systems that feed flood basalt eruptions commonly result in aphyric lavas at the surface and have generally been thought to represent far simpler systems than their volcanic arc counterparts. Nevertheless, these flood basalt systems sample the entire lithosphere and it has now been shown that where crystal populations are identified they can record shallow, crustal-level processes (Ramos et al., 2005). Moreover, detailed studies of flood basalt provinces highlight their complex internal architecture (e.g. Jerram & Widdowson, 2005), and findings of lavas and dykes with large crystal populations, termed ‘giant plagioclase basalts’ (e.g. Higgins & Chandrasekharam, 2007), provide an important opportunity to probe through the plumbing systems of large igneous provinces. The aim of this study is to address the isotopic variation present in crystals from a magmatic system associated with the eruption of flood basalts. We integrate detailed core–rim and single-crystal analyses with whole-rock isotope results to unravel the complex differentiation history of the magmas and the contamination pathways occurring at different times and places during magma ascent through the lithosphere. Our study provides a useful comparison with that of the Columbia River Flood Basalts (Ramos et al., 2005). Here we focus on the British Palaeogene Igneous Province (BPIP), formerly known as the British Tertiary Igneous Province (BTIP), and extend consideration to Sr and Pb isotope variations. Figure 1 illustrates different scenarios of how isotopic profiles of crystals might be used to constrain the histories of magmas ascending through isotopically distinct crustal layers. A single rock with a given isotope composition may represent a mixture of two or more different end-members (Fig. 1). Using core–rim isotopic profiles and single-crystal isotope analyses compared with whole-rock compositions it is, in principle, possible to distinguish between these end-members to (1) better constrain the composition of the mantle sources and (2) constrain open-system differentiation pathways (Fig. 1). We adopt this approach for selected basaltic and intermediate composition volcanic samples from the Isle of Skye igneous centre, a part of the BPIP. We have analysed the intra-crystal and inter-crystal isotope heterogeneity (Sr and Pb) of plagioclase crystals in detail and combined these data with whole-rock isotopic compositions from this study and from the literature. We highlight in this study that by studying core–rim and single-crystal isotope variations in a few crystals from samples of a complex magmatic system it is possible to obtain more information about magmatic processes and differentiation pathways than by studying the whole-rock composition of hundreds of lavas.
|
), can be used to understand processes such as crystallization and contamination of magmas derived from a source mantle (M) ascending through isotopically distinct crustal components (C1 and C2). (a) The isotopic composition of a whole-rock is the average composition of the mantle-derived magma and crustal contaminants C1 and C2. (b) Two-stage contamination during crystallization of crystal with a core-to-rim isotope composition evolving from an uncontaminated composition in the core to a compositions similar to C2 in the intermediate zones and to C1 in the rim. (c) Crystal that grows from a host-magma that has been contaminated by crustal components C1 and C2 during its ascent to the surface and prior to crystallization. Thus, core and rims show the same isotope composition. (d) Crystallization takes place in the transition zone between upper and lower crust, where the host-magma becomes progressively contaminated by a hybrid mid-crustal component (C1 + C2) The crystal core shows an uncontaminated isotope composition whereas the rim shows an isotope composition similar to the mixed crustal contaminant. (e) Crystallization takes place in the upper crust from magmas that have been previously contaminated by C2 during their ascent through the lower crust. The magma also becomes contaminated by C1 while crystallizing in the upper crust. The cores of the crystal have a composition closer to C2 and the rim close to C1. |
THE SKYE IGNEOUS CENTRE |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
The Skye lavas, and lavas from other BPIP volcanic centres (e.g. Mull), are examples of magmas that have traversed a significant vertical section of continental crust (e.g. Thompson et al., 1980
). The crystalline basement below Skye is composed of Archaean Lewisian metamorphic rocks consisting of granulite-facies gneiss in the lower crust and amphibolite-facies gneiss in the upper crust (Hamilton et al., 1979; Weaver & Tarney, 1980, 1981). These lithologies are overlain in places by Torridonian metasediments (Dickin & Exley, 1981). Lewisian lower crustal gneisses became strongly depleted in K, Rb, U and Th during ancient metamorphism, compared with Lewisian upper crust, resulting in less radiogenic Sr and Pb isotope ratios (Weaver & Tarney, 1981). The overlying Torridonian sediments have similar Pb isotope ratios to Lewisian upper crust but have a much larger range in Sr isotope compositions and are more radiogenic compared with Lewisian upper and lower crust. In contrast, the Hebridean mantle at the time of lava eruption was characterized by unradiogenic Sr isotope ratios and radiogenic Pb isotope ratios broadly similar to those of the Lewisian upper crust (e.g. Moorbath & Thompson, 1980). The large contrast in Sr and Pb isotopic composition between the mantle source for the Skye lavas and the upper and lower crust (Thompson et al., 1980; Dickin, 1981) allows us to observe and document crustal contamination processes and place them in a spatial context (i.e. depth). By investigating core-to-rim isotope variations within crystals in the Skye basaltic lavas we aim to constrain better the contribution of the different crustal end-members to the mantle-derived magmas before and during the growth of each crystal.
Magmatism
The Isle of Skye volcanic centre is located in the NW of Scotland (Fig. 2); it forms part of the BPIP and, like the Mull igneous centre, was active for a protracted period during the onset and main phase of flood volcanism (Jerram & Widdowson, 2005). The magmatic activity on Skye occurred between c. 60·5 Ma and 55·9 Ma (Hamilton et al., 1998; Chambers & Pringle, 2001; Emeleus & Bell, 2005) and is associated with the opening of the North Atlantic Ocean. Extensive eruption of lavas of basaltic and intermediate composition occurred, forming a >1·5 km thick lava succession (England, 1994; Williamson & Bell, 1994), named the Skye Main Lava Series (SMLS; Dunham, 1970; Thompson et al. 1972). The SMLS varies in extent of fractionation from alkali basalts to benmoreites and from hy-normative basalts to trachytes (Thompson et al., 1972). The parental magmas of the SMLS lavas were picritic, formed by variable degrees of partial melting of a spinel-lherzolite mantle source (Thompson et al., 1980; Ellam, 1992; Scarrow & Cox, 1995; Scarrow et al., 2000). The majority of the magma batches underwent fractional crystallization to varying extents at the base of the crust, producing a range of basaltic and hawaiitic parental magmas that penetrated the crust through dyke swarms (Thompson et al., 1972, 1980, 1982). Thompson et al. (1982) demonstrated that the SMLS lavas partially crystallized at the base of the lower crust (Fig. 3) and underwent no further fractional crystallization between the Moho (9 ± 1·5 kbar; Thompson et al., 1982) and the surface. However, the trace element and isotope composition of the lavas indicate that the magmas underwent crustal contamination, mostly by the granulite-facies Lewisian lower crust (Thompson et al., 1986). The SMLS lavas, when mapped in detail, form single lava flows of small volume with many features akin to those of shield volcanic centres with lava tubes, flow lobes and tumuli (Single & Jerram, 2004), indicating that the volcanic centre on Skye was a long-lived feature, and that crustal magma reservoirs were developed early in the evolution of the Skye complex.
|
18O values (labelled in 
), defining the areas affected by hydrothermal alteration related to the emplacement of the intrusive complexes in the south–central part of Skye (Forester & Taylor, 1977
|
A second magmatic episode generated tholeiitic magmas of the Preshal More basalts (PMB) (Thompson et al., 1982
), which were emplaced as dykes cutting the SMLS (Mattey et al., 1977) and as a lava succession lying unconformably upon the SMLS. The PMB magmas, in contrast to the SMLS, underwent fractional crystallization within the upper crust (Moorbath & Thompson, 1980; Thompson et al., 1982) at lower pressures compared with the SMLS (Fig. 3). The compositional change from SMLS to PMB has been related to shallowing of the melting region as a result of the thinning of the lithosphere at the beginning of the opening of the North Atlantic (Thompson & Gibson, 1991; Ellam, 1992; Saunders et al., 1997, and references therein).
Towards the end of the period of flood basalt volcanism (c. 58 Ma; Emeleus & Bell, 2005), intrusive igneous complexes developed including gabbros (Cuillin Centre) and granites (Western and Eastern Red Hills) (e.g. Emeleus, 1982; Emeleus & Gyopari, 1992). The earliest Cuillin magmas are similar in composition to the SMLS; however, the bulk of the Cuillin Centre has similar composition to PMB magmas, suggesting that PMB dykes cutting the SMLS are lateral injections of tholeiitic magma from the Cuillin Centre (Dickin, 1981; Walker, 1993). Other types of lava not a prori belonging to the SMLS and PMB are also found in the northern parts of the Isle of Skye; for example, the Fairy Bridge Basalts (Thompson et al., 1980), which are mafic, porphyritic, and include large plagioclase phenocrysts. They have aphyric margins of hawaiitic–mugearitic composition and a central part similar in composition to the SMLS. These lavas were named Big-Feldspar Lavas by Thompson et al. (1980).
Crustal contamination history: theories and models
Many studies of the BPIP (e.g. Moorbath & Thompson, 1980; Dickin, 1981; Thompson et al., 1982) have shown that magmas ascending through the lithosphere beneath Skye became contaminated to different extents by the lower and upper Lewisian crust. Initial work (Carter et al., 1978) identified two different contaminants, granulite-facies lower crust and amphibolite-facies upper crust, by their distinct isotopic characteristics. Early models (Carter et al., 1978) established that the lavas from Skye and Mull were contaminated by granulite-facies lower crust, but the granites from Skye were contaminated by amphibolite-facies upper crust (Dickin et al., 1984). More detailed models in subsequent studies based on combined whole-rock Pb, Sr and Nd isotopic compositions (Moorbath & Thompson, 1980; Dickin, 1981) showed that the SMLS became selectively contaminated, mainly by the Lewisian lower crust (granulite-facies) and to a lesser extent with the Lewisian upper crust (amphibolite-facies; Thompson et al., 1986; Morrison et al., 1985). Moorbath & Thompson (1980) and Thompson et al. (1982) explained the variability in isotopic and trace element composition of the Skye lavas as generated by ‘selective’ fusion of minor veins of leucogneiss present within the granulite-facies lower–intermediate crust. Thirlwall & Jones (1983) suggested that crustal melts were incorporated into the magmas in variable amounts, generating the trace element and isotopic variability of the SMLS. Kerr et al. (1995), in agreement with earlier work, also suggested that the isotopic and trace element variation in the Palaeogene Mull lavas from the lower half of the Mull Plateau Group, which were erupted earlier than the SMLS, was due to contamination of the magmas by addition (
5%) of felsic and pegmatitic material present as veins in the granulitic lower Lewisian crust. The MgO-rich magmas with high liquidus temperatures assimilated higher amounts of these veins, and therefore became more contaminated compared with the more evolved magmas with lower liquidus temperatures (Kerr et al., 1995). The contamination mechanism that Kerr et al. (1995) proposed, based on the model of Huppert & Sparks (1985), was ‘assimilation by turbulent ascent’ (ATA) of the magmas as they passed through the continental crust via thin, poorly connected dyke- and sill-like magma chambers. Kerr et al. (1995) considered that simple AFC (combined assimilation plus fractional crystallization) models could not explain the compositional signatures of the Palaeogene Mull lavas.
The Sr and Pb isotope compositions of the PMB were explained as mantle-derived magmas which were either uncontaminated or contaminated by the upper crust (Dickin, 1981). A similar origin, with additional fractionation, has been suggested for the Eastern Red Hill granites (Dickin et al., 1984; Morrison et al., 1985; Bell & Williamson, 2002). Consequently, the PMB were included within the same group as the granites in Pb isotope diagrams in subsequent review papers (e.g. Dickin et al., 1984; Bell & Williamson, 2002). The trends in isotopic composition of the different rock types from Skye and their crustal contaminants have been modelled using numerous crustal assimilation models, varying from AFC (DePaolo, 1981), to ATA (Kerr et al., 1995) and to energy constrained-recharged assimilation fractional crystallization (EC-RAFC; Bohrson & Spera, 2003; Fowler et al., 2004). These models indicate that the Skye magmas were derived from a relatively small number of parental magmas exhibiting limited compositional variation. The wide compositional range of the derivative magmas was formed by various combinations of fractional crystallization, magma recharge and assimilation of anatectic melts (Fowler et al., 2004). As indicated above, such models are necessarily incomplete if they underestimate the number of components involved in crustal contamination and the extent of their variation. Here, we attempt to assess the full extent of contributing crustal end-members to a single magmatic product via micro-sampling. To assess the full extent of the interaction between the magmas and the crust we have focused on the more differentiated magmas, from which plagioclase has crystallized.
|
SAMPLING AND SAMPLE DESCRIPTIONS |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
Sample selection
The sampling for this study was carried out in Western Skye, away from the hydrothermally altered regions associated with the Cuillins and Eastern–Western Red Hills granitic complexes (Fig. 2). We collected SMLS samples of basaltic and intermediate (hawaiitic) composition, tholeiitic PMB lavas and a Big-Feldspar composite dyke sample similar in composition to the Big-Feldspar Lavas described by Thompson et al. (1980
). From all the samples collected, we selected a subset of 12 samples (Table 1) for whole-rock isotopic analyses. These samples were chosen based on their composition and position within the stratigraphic sequence of the Skye volcanic centre (Fig. 4) along with the presence of an adequate crystal population to make the detailed crystal isotope analyses approach feasible. Because of the time-intensive nature of the intra-crystal isotopic studies, we have selected only a few crystals per sample (Table 1). For these, we have analysed the isotopic compositions of cores and rims and mineral separates. In a subset of crystals we have analysed complete core-to-rim profiles to explore the isotopic variation within single crystals. In one sample, the Big-Feldspar dyke, we have micro-milled: (1) host glass found at the edges of the dyke; (2) three single plagioclase crystals hosted in the glassy edges; (3) cores and rims of other plagioclase crystals. For this type of isotope micro-sampling study we need lavas that have already fractionated plagioclase, which in our case are lavas with basaltic and intermediate composition (<8 wt % MgO; Table 2). The contamination model presented here is appropriate for these intermediate Skye lavas. To assess the contamination history of the most primitive lavas on Skye (12–13 wt % MgO; Mattey et al., 1977; Kerr, 1998) would require an isotope study of olivine-hosted melt inclusions, which is beyond the scope of this study.
|
|
|
Sample description
The SMLS samples from this study are mainly characterized by a porphyritic texture (Fig. 4) with large plagioclase phenocrysts (0·5–3 cm) of composition An45–55 set in a microcrystalline matrix of plagioclase and olivine (Fo72) crystals. Some of the plagioclase crystals in the SMLS intermediate samples display sieve textures. The PMB lavas from this study are characterized by a small percentage (
10%) of highly anorthitic An88–89 cumulate plagioclase crystals (Fig. 4) hosted in a microcrystalline matrix (0·5–4 mm). The Big-Feldspar composite dyke has a central part with a similar texture and composition to the SMLS samples but a glassy rhyolitic margin hosting plagioclase crystals of a more sodic composition An8–4 (Fig. 4). Generally these sodic feldspars (0·5 cm) within the chilled margins have rounded edges and occur in cumulate clots. The Big-Feldspar composite dyke glassy margins also contain biotite, apatite and amphibole. In the subsequent discussion, the central part of the composite dyke (sample SKLF10A) is included in the SMLS sample group whereas the glassy margin of the composite dyke is referred to as Big-Feldspar dyke sample (SKLF10G) and is discussed separately because of its compositional and textural differences. |
ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
The micro-milling technique: sampling and chemistry procedures
The in situ micro-mill sampling, carried out on feldspars and groundmass, was performed following the techniques outlined by Charlier et al. (2006
) using a New Wave micro-mill at the Arthur Holmes Isotope Geology Laboratory (AHIGL), which forms part of NCIET, Durham University, UK. Thick sections (100 µm) were prepared for each sample so as to have sufficient depth for micro-milling both phenocrysts and the groundmass of the rock samples. Before micro-milling the samples, electron microprobe analysis (EMPA) was carried out using a Cameca SX50 with a 15 nA current and 20 kV voltage, at SCT-Serveis Científico Tècnics, University of Barcelona, Spain.
The micro-milled samples were weighed and dissolved following procedures described in the following sections and presented in detail by Charlier et al. (2006). Following dissolution, aliquots were taken from each sample solution to measure the Rb/Sr, U/Pb and Th/Pb ratios necessary for making age corrections to the measured Sr and Pb isotope ratios.
Total procedural blanks
Total procedural blanks (TPB) for Sr and Pb (including milling, drying down on a gold boat, sample dissolution and column elution procedure) averaged 8 ± 1·2 pg (2SD) for Sr (n = 13) and 99·5 ± 31·6 pg (SD) for Pb (n = 12). Pb blanks monitored, including dissolution and Pb elution, averaged 9 ± 1·6 pg (2SD). The 87Sr/86Sr isotopic composition of the TPBs was also monitored periodically by combining the equivalent of 60 TBPs to yield sufficient Sr (>500 pg) for a precise and accurate thermal ionization mass spectrometry (TIMS) analysis. The average 87Sr/86Sr composition of the laboratory blank was 0·710853 ± 0·000194 (2SD). Although the TPBs are very low in Sr and Pb, all the samples were corrected for blank contribution.
Aliquoting procedure
To avoid issues of sample heterogeneity, trace element and isotope compositions were determined on the same sample solution by removing a 10% aliquot of the solution. The remaining solution (90%) was used to separate the Sr and Pb fractions for isotope ratio analyses. The aliquoting procedure has been described by Font et al. (2007).
The accuracy and uncertainties of the aliquoting procedure were tested by measuring the Rb/Sr, U/Pb and Th/Pb ratios in 10% aliquots taken from four BHVO-1 standard (Abbey, 1982; Flanagan, 1976; Govindaraju, 1994) solutions, which were prepared by weighing 50 mg, 10 mg, 6 mg and 2 mg of the BHVO-1 standard. These standard samples were dissolved using the same dissolution procedure as for the micro-milled samples and three 10% aliquots were taken from each standard solution. The measured Rb/Sr ratios of the aliquots, with an average value of 0·02321 ± 0·0004 (2SD), are within error of the accepted BHVO-1 Rb/Sr ratios [0·0238 ± 0·0104 (2SD)] and the reproducibility of the measurements is 1·7%. The U/Pb and Th/Pb ratios of the BHVO-1 standard aliquots are 0·2150 ± 0·0229 (2SD) and 0·6144 ± 0·0364 (2SD), respectively. The reproducibility of the measurements is 16·4% (U/Pb) and 16·2% (Th/Pb). Both ratios, U/Pb and Th/Pb, are generally higher than the accepted BHVO-1 standard values; 0·1615 ± 0·0265 (2SD) and 0·4462 ± 0·0722 (2SD) respectively (Flanagan, 1976; Abbey, 1982; Govindaraju, 1994). The higher U/Pb and Th/Pb ratios obtained in this aliquoting study and the high uncertainties in the measurements could be related to the low concentrations of U and Th in the BHVO-1 standard and the small size of the aliquots used for the experiment.
Column chemistry: Sr and Pb fraction separation
The Sr separation procedure used in this study was based on the micro-Sr column chemistry method described by Charlier et al. (2006), designed for samples with small amounts of available Sr. Two extra steps were added to the already established column chemistry technique to be able to separate the Pb fraction using the same micro-columns. These involved two washing steps using 100 µl UpA 2·5M HCl before the Pb fraction was collected in 500 µl of UpA 8M HCl. The Pb fraction was then dried down and then dissolved in 3% UpA HNO3 for analysis.
Whole-rock Sr and Pb column chemistry
The Sr and Pb fractions of the whole-rock samples were also separated using the Sr spec micro-columns following the same method as for the micro-milled samples. The total procedure blanks for Sr and Pb were all <50 pg and reflect the less rigorous procedures employed for high abundance whole-rock Sr and Pb determination.
87Sr/86Sr ratio analysis: thermal ionization mass spectrometry
Sr isotope ratios of the micro-milled samples were measured by TIMS using a ThermoElectron Triton system at AHIGL. The Sr fractions were loaded onto Re filaments using a TaF5 activator to enhance ionization of small Sr samples or standards (<12 ng; Charlier et al., 2006). Large Sr standards (600 ng) used to monitor the performance of the system were loaded onto Ta filaments using 1 µl H3PO4. 87Sr/86Sr ratios were measured using a static multicollection routine. An analysis consisted of 18 blocks of 10 cycles with an integration time of 4 s per cycle. 87Sr/86Sr and 84Sr/86Sr ratios were corrected for mass fractionation using an exponential law and 86Sr/88S ratio of 0·1194. During the period of this study (2005–2006) 206 analyses of the international Sr standard NBS987 were carried out on load sizes ranging from 3 ng to 600 ng to monitor and document the system's performance. The overall average 87Sr/86Sr and 84Sr/86Sr ratios for the 206 NBS987 measurements were 0·710263 ± 13 and 0·056483 ± 2, respectively. The agreement in the average 87Sr/86Sr and 84Sr/86Sr ratios across the range in standard size is excellent (<0·0014% and 0·005%, respectively) and the values for each load size are also well within error of the TIMS values reported and recommended by Thirlwall (1991).
Sr and Pb isotope ratio analyses: plasma ionization multicollector mass spectrometry
The Sr fractions of the whole-rock samples and the Pb fractions of both the micro-milled plagioclases and whole-rock samples were measured for isotope ratios by plasma ionization multicollector mass spectrometry (PIMMS) using the AHIGL ThermoElectron Neptune instrument. The basic analytical method for both Pb and Sr on the Neptune comprises a static multi-collection routine of 1 block of 50 cycles with an integration time of 4 s per cycle; total analysis time is 3·5 min.
After chemistry, Sr whole-rock samples were taken up in 1 ml of 3% HNO3 and introduced into the Neptune using an ESI PFA50 nebulizer and a dual cyclonic–Scott Double Pass spray chamber. With this sample introduction setup, and the normal H skimmer cone, the sensitivity for Sr on the Neptune is typically 60 V total Sr ppm–1 at an uptake rate of 90 µl min–1 (Nowell et al., 2003). Prior to analysis a small aliquot was first tested to establish the Sr concentration of each sample by monitoring the size of the 84Sr beam (88Sr was too high in the non-diluted aliquot to measure directly) from which a dilution factor was calculated to yield a beam of c. 20 V 88Sr. Instrumental mass bias was corrected for using a 88Sr/86Sr ratio of 8·375209 (the reciprocal of the 86Sr/88Sr ratio of 0·1194) and an exponential law. The whole-rock samples were analysed in a single session during which the average 87Sr/86Sr value for NBS987 was 0·710245 ± 0·000019 (26·1 ppm 2SD; n = 7).
Following chemistry whole-rock Pb aliquots were taken up in 1 ml of 3% HNO3. Prior to analysis each sample was tested on the Neptune to determine its Pb concentration and thereby calculate the appropriate amount of Tl spike to add to obtain a Pb/Tl ratio of 12, which simultaneously minimizes the tails from 205Tl onto 204Pb and from 206Pb onto 205Tl. After spiking with Tl each sample was introduced into the Neptune using an ESI PFA50 nebulizer and a dual cyclonic–Scott Double Pass spraychamber. With this setup, and the normal H skimmer cone, the sensitivity for Pb on the Neptune is typically 100 V total Pb ppm–1 at an uptake rate of 90 µl min–1 (Nowell et al., 2003). Pb mass bias was corrected externally using the 205Tl/203Tl ratio of the admixed Tl spike and an exponential law. The 205Tl/203Tl used for correcting the Pb ratios is determined for each analytical session by minimizing the difference in offset between the session average Pb ratios (all ratios) and the Galer (1997) triple spike Pb isotope values. The Tl isotope ratio is calculated to give the best fit to all the Pb isotope ratios of Galer (1997) simultaneously. Whole-rock samples were analysed over one analytical session during which the 205Tl/203Tl ratio used for mass bias correction was 2·3883. Average 206Pb/204Pb, 207Pb/204Pb and 208Pb/204Pb for the NBS981 Pb standard for the session were 16·9403 ± 0·0007, 15·4967 ± 0·0006, 36·7151 ± 0·0020 (all 2SD; n = 19).
Micro-milled plagioclase samples were also taken up in 1 ml of 3% HNO3 but, because of the small amount of recovered Pb, samples were introduced into the Neptune using an Aridus desolvating spray chamber together with an ESI PFA50 nebulizer. The H skimmer cone was also exchanged for the higher sensitivity X skimmer cone. The sensitivity for Pb with the Aridus and X skimmer cones was 1100 V total Pb ppm–1 at an uptake rate of 90 µl min–1. As with the whole-rock samples, the Pb concentrations of each plagioclase sample were tested to determine the amount of Tl spike needed to obtain a Pb/Tl ratio of close to 12. Pb mass bias was corrected for externally in a similar manner to that for the whole-rock samples. Micro-milled plagioclase samples were analysed over two analytical sessions during which the 205Tl/203Tl ratios used for mass bias correction were 2·3884 and 2·3886. Average 206Pb/204Pb, 207Pb/204Pb and 208Pb/204Pb for the NBS981 Pb standard for the two sessions were 16·9407 ± 0·0013, 15·4970 ± 0·0014, 36·7188 ± 0·0036 (all 2SD; n = 17) and 16·9404 ± 0·0014, 15·4972 ± 0·0016, 36·7174 ± 0·0034 (all 2SD; n = 10), respectively.
Trace element analyses: inductively coupled plasma-mass spectrometry (ICPMS)
Aliquots of the dissolved sample solutions were analysed for trace element concentrations by inductively coupled plasma-mass spectrometry (ICPMS) using a Thermo Electron Element II system at AHIGL, Durham University, UK. The aliquots were taken up in 1 ml 3% UpA HNO3 and introduced into the Element II using a 100 µl/min microflow nebulizer and a dual cyclonic Scott double-pass spray chamber. A 1 ng/ml In solution was used to tune the system for sensitivity until the 115In peak yielded >106 counts per second. A 1 ng/ml Ce solution was then used to optimize the oxide production rate by adjusting, primarily, the nebulizer gas-flow rate to give a Ce/CeO ratio of between 0·03 and 0·05. The elements of interest and isotopes of each element used for concentration measurement are 85Rb, 88Sr, 208Pb, 232Th and 238U. Other isotopes measured for each of the elements were used for background interference assessments. Concentration calibration was established using multi-element solutions of 50–1000 pg/ml made up from RomilTM (Cambridge, UK) 1000 µg/ml standards. The TPBs for the analysed elements were <25 pg. The trace element concentrations of the sample solutions were corrected for blank and instrument drift. The trace element concentrations of the samples were calculated using the aliquot weight and original sample mass.
|
ISOTOPE DATA |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
Skye Main Lava Series samples
The 87Sr/86Sri values of the whole-rock SMLS samples (Table 2) overlap the previously published 87Sr/86Sri SMLS field (Moorbath & Thompson, 1980
), which plots between the Hebridean mantle and the Lewisian lower crust (Fig. 5). However, the 87Sr/86Sri field of the plagioclase cores and rims and plagioclase separates (Table 3) from SMLS samples extends towards Lewisian upper crust 87Sr/86Sri ratios (Fig. 5). Detailed 87Sr/86Sri profiles across plagioclase crystals from some of the SMLS samples show correlation between 87Sr/86Sri and An content (Fig. 6a and c). However, other crystals show poor or no correlation (Fig. 6b and d).
|
|
|
Pb isotope compositions of whole-rock and micro-milled plagioclase crystals are given in Tables 2 and 3, respectively. Because of the low Pb concentrations of the plagioclase crystals from the SMLS samples, only mineral separates and a few cores and rims could be analysed (Table 3). 208Pb/204Pbi vs 206Pb/204Pbi plots differentiate very well between the Lewisian lower and upper crust end-members (Fig. 7; Dickin, 1981
). According to the published data, the Lewisian lower and upper crust (granulite and amphibolite facies, respectively) are characterized by 206Pb/204Pbi ratios in the range of 13–16·5 for the lower crust and 13–18 for the upper crust. The Lewisian upper crust has higher 208Pb/204Pbi ratios compared with the lower crust, because of the relative depletion of Th in the lower crust. In contrast, Hebridean mantle is more radiogenic in 206Pb/204Pbi than both Lewisian upper and lower crust but it also has high 208Pb/204Pbi similar to Lewisian upper crust (Fig. 7). The SMLS whole-rock samples plot within a well-correlated array between the Hebridean mantle and the Lewisian lower crust (Fig. 7). The cores and rims of the plagioclase crystals and plagioclase fragments show large variations in Pb isotope ratios spanning the entire field previously defined by the SMLS whole-rocks (Fig. 7; Dickin, 1981). Similiar trends are displayed in plots of 208Pb/204Pbi vs 87Sr/86Sri (Fig. 8).
|
|
Preshal More Basalt samples
Whole-rock PMB samples have 87Sr/86Sri very close to proposed mantle values and plot at the lower end of the PMB field, which ranges between the mantle and Lewisian lower crust 87Sr/86Sri ratios (Figs 5 and 8
). The cores and rims of the plagioclase crystals from the PMB samples, however, have higher 87Sr/86Sri ratios compared with their whole-rocks (Fig. 5). The range in 87Sr/86Sri displayed by the cores and rims of the plagioclase crystals is similar to the range shown by the PMB whole-rock 87Sr/86Sri values (Thompson et al., 1982).
The PMB whole-rocks have 208Pb/204Pbi and 206Pb/204Pbi ratios (Table 2) close to the Hebridean mantle value (Fig. 7). Because of the small size and the low Pb concentration in the PMB plagioclase crystals, it was not possible to measure Pb isotopes in those crystals. In Figs 7 and 8, the PMB field has been delimited according to PMB whole-rock data from this study and data available in the literature (Dickin, 1981). In review papers (e.g. Bell & Williamson, 2002), the PMB Pb isotope field is larger than the field outlined here, as it has been grouped together with data from the Cuillin gabbro layered intrusion (Dickin et al., 1984) because they are considered to be geochemically related.
Big-Feldspar dyke sample
Host glasses from the Big-Feldspar dyke (SKLF10G) are characterized by 87Sr/86Sri ratios between 0·704 and 0·706 overlapping the signature of the Lewisian lower crust. Cores and rims of feldspars from the Big-Feldspar dyke are more radiogenic in 87Sr/86Sri than the host glasses and plot between Lewisian lower and upper crust (Fig. 5). In 208Pb/204Pbi vs 87Sr/86Sri space (Fig. 8), albitic plagioclase core–rim pairs, together with single micro-milled plagioclase phenocrysts and host glasses, lie outside the SMLS field and are more radiogenic in both 208Pb/204Pbi and 87Sr/86Sri. However, single micro-milled plagioclase feldspars have similar 206Pb/204Pbi to the host glasses and core–rim pairs (Fig. 7), but are highly radiogenic in 208Pb/204Pbi, with ratios trending towards Lewisian upper crust.
|
INTERPRETATION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
Isotopic variation
The observed large ranges in the Sr and Pb isotope ratios of feldspar core–rim pairs from different Skye samples suggest that different extents of lower and upper crustal contamination occurred during crystallization. 87Sr/86Sri variations observed among feldspars in all samples are larger than the total variation observed in the host whole-rocks (Fig. 5). The 87Sr/86Sri isotopic range of the SMLS and PMB whole-rocks lies between Hebridean mantle and Lewisian lower crust, indicating that the lower crust is the main contaminant in these magmas. However, there is a core and rim of plagioclase from the SMLS samples that have more radiogenic 87Sr/86Sri (Fig. 5: note that low Pb contents mean this core–rim pair cannot be analysed for Pb isotopes or plotted in Figs 7 and 8
), suggesting that their host magmas were contaminated either by Lewisian upper crust or by Torridonian sediments. In contrast, PMB plagioclase crystals have 87Sr/86Sri between Hebridean mantle and Lewisian lower crust (Fig. 5). The small range in 87Sr/86Sri of the PMB and the similarity between 87Sr/86Sri of the PMB plagioclase core–rim pairs and the Hebridean mantle suggests that assimilation of Lewisian lower crust was less than in the SMLS in which plagioclase shows a significant range in core–rim 87Sr/86Sri (Fig. 5). In Pb isotope space and Pb vs Sr isotope space (Figs 7 and 8) the SMLS and PMB lavas and plagioclase crystals lie along a trend between Hebridean mantle and Lewisian lower crust. However, albitic feldspars and host glasses from the Big-Feldspar dyke rhyolitic margin have Sr and Pb isotopic ratios that plot outside the main trend defined by the SMLS and PMB. The core–rim pairs and the single micro-milled albitic plagioclases from the Big-Feldspar dyke have more radiogenic Sr and Pb isotope ratios than both the SMLS and PMB plagioclases, suggesting a larger input of Lewisian upper crust (Fig. 8). Given the difference in whole-rock compositions, this might suggest that the parental magmas are ubiquitously modified in the lower crust, whereas a second stage of differentiation to produce rarer, highly evolved, compositions took place in the upper crust.
Crustal contamination and fractional crystallization model
Researchers such as Thompson et al. (1982) and Kerr et al. (1995) suggested that simple bulk assimilation of lower crust was not the mechanism of contamination of the Palaeogene lavas in NW Scotland, but instead that these magmas preferentially assimilated the most fusible pegmatitic/granitic veins present within the Lewisian crust (e.g. Moorbath & Thompson, 1980; Morrison et al., 1985; Kerr et al., 1995). These pegmatitic/granitic veins represent pre-Palaeogene partial melts of Lewisian crust (Weaver & Tarney, 1980, 1981). In this study, we model the isotopic variations observed in plagioclase crystals from the SMLS, PMB and Big-Feldspar dyke by using simple AFC equations (DePaolo, 1981) and assimilation of (1) pegmatitic/granitic Lewisian crustal material [pegmatitic/granitic veins crossing granulite- and amphibolite-facies rocks reported by Kerr et al. (1995)] and (2) the average compositions of Lewisian lower and upper crust (Dickin, 1981). The composition of the contaminant is expected to be very heterogeneous isotopically and so the model presented is only indicative of possible scenarios that could have occurred during the evolution of the Skye magmas. The contamination process of the magmas by the pegmatitic/granitic veins was by melting of the wall-rocks followed by mixing, not dissolution of the wall-rocks in the magma. As a parental magma composition for the SMLS magmas, we used sample SK940 (labelled in Fig. 8 as P), which is a relatively uncontaminated SMLS lava reported by Moorbath & Thompson (1980), Dickin (1981) and Thompson et al. (1982). We use SKLF08 from this study as the parental magma composition for the PMB samples. We recognize that because the SMLS and PMB isotopic fields have significant ranges, other parental compositions can also be used for the AFC calculations. Parental and crustal contaminant compositions chosen in this study are adequate to model the compositional ranges of our samples.
Observed Sr and Pb isotopic variations from the various samples have been modelled in this study using simple AFC model calculations (Fig. 8; DePaolo, 1981). The AFC models used here are merely to illustrate open-system contamination processes. The EC-AFC [energy-constrained AFC model of Bohrson & Spera (2003)] is clearly thermodynamically more rigorous. However, more parameters are needed for this model, and therefore more degrees of freedom. The simpler AFC analysis here shows that contamination is not simply a bulk-rock phenomenon, but needs consideration of the roles of various crystal cargoes. So although it may be satisfying to model contamination from bulk-rock and magma end-members using the available algorithms, our point is that this approach is not wholly realistic.
We have attempted to characterize, in detail, the nature of the contaminants during ascent of the magmas through the crust. It is well known that fractional crystallization occurred deep in the lower crust (9 kbar) for the SMLS, in the intermediate crust for PMB, and within the upper crust for later magmas (Fig. 4; Thompson et al., 1982). The SMLS field can be modelled using an AFC approach starting from the parental magma composition P (Fig. 8a) and assimilating Lewisian lower crust (AFC1; Fig. 8b) at ratios of assimilation to fractional crystallization r 0·6. The Pb and Sr isotopic composition of the core–rim pairs and single plagioclase micro-milled crystal from the Big-Feldspar dyke can be modelled by an AFC curve starting from parental magma P by assimilating relatively small amounts of Lewisian pegmatitic/granitic veins found within Lewisian gneiss samples [P42, P43 and P50; reported by Kerr et al. (1995)]. The ratio of assimilation to fractional crystallization of these AFC models (2–4) are between r 0·3 and r 0·09 (Fig. 8b). The isotopic signature of one of the single plagioclase micro-milled crystals with the highest 208Pb/204Pb ratio can be modelled from parental magma composition P with an AFC curve calculated by assimilating Lewisian upper crust at an AFC rate of r 0·3 (AFC5; Fig. 8b). The large range in 87Sr/86Sri of the host glass of the Big-Feldspar dyke sample suggests recharge of new magma, which was previously contaminated by Lewisian lower crust, followed by mixing with the existing host magma present in the magma reservoir (see black arrow in Fig. 8b and c). The two melts did not fully homogenize before cooling, generating large Sr isotopic ranges (Fig. 8).
Similar AFC curves can be calculated using a PMB parental magma composition (Fig. 8c). The entire PMB field can be modelled by an AFC curve with a ratio of assimilation of Lewisian lower crust to fractional crystallization of r 0·2 (AFC6; Fig. 8c). As in the previous SMLS model, the core–rim pairs and whole plagioclase micro-milled crystals from the Big-Feldspar dyke can be modelled by an AFC curve at low AFC r values (r 0·03) indicating very small amounts of assimilated pegmatitic/granitic crustal material. The high 208Pb/204Pb ratios of the single plagioclase micro-milled crystals in the Big-Feldspar dyke sample (Fig. 8d) can be also modelled starting from a composition that represents a melt already contaminated by the lower crust that assimilated upper crustal material. This could be achieved with r values of r0·5. The fact that the core–rim pairs and whole micro-milled plagioclase crystals from the Big-Feldspar dyke have a large range in 208Pb/204Pb but a similar range in 206Pb/204Pb and 87Sr/86Sri suggests that the crystals formed from the same host magma, which was variably contaminated by Lewisian upper crust, generating the large range in 208Pb/204Pb. The majority of the AFC curves modelled for different parental magma compositions represent crustal assimilation of less than 20%, in agreement with previous crustal assimilation assessments for the Skye and Mull Palaeogene lavas (Thompson et al., 1982; Kerr et al., 1995).
Analysing the Pb isotope variation within single crystals from the SMLS and Big-Feldspar dyke samples (Fig. 9) in detail shows that crystal growth and crustal contamination processes can be determined by looking at the isotopic variation in the core, rim and intermediate zones of crystals together with whole-rock values (Fig. 9a). 208Pb/204Pbi and 206Pb/204Pbi ratios in Fig. 9a progressively decrease from core to rim, suggesting growth from a magma undergoing progressive contamination by Lewisian lower crust. The large difference in Pb isotope ratios between the core and the rim of the plagioclase crystal and the corresponding whole-rock shows that this sample underwent extensive crustal contamination during differentiation by the addition of more contaminated magma batches into the system. Small variations in 87Sr/86Sri along the plagioclase core–rim profiles (Fig. 6) show that Sr isotope variations, even if small, effectively track contamination by the lower crust, occurring progressively at different levels in multiple episodes. There are other crystals from the SMLS samples (Fig. 9b) that show very little inter-crystal Pb isotopic variation compared with the whole rocks. This suggests that plagioclase crystals grew from the same host magma composition, which was contaminated in the lower crust by a fairly small amount prior to significant plagioclase crystallization. The fact that plagioclase crystals and the host bulk-rock have very similar Pb isotope ratios indicates that the magma erupted immediately following crystallization of plagioclase without further crustal contamination (Fig. 9b).
|
According to 208Pb/204Pbi vs 206Pb/204Pbi systematics, feldspars from the Big-Feldspar sample grew in two stages. Cores and host glasses have very similar 208Pb/204Pbi and 206Pb/204Pbi; however, the rims define a trend showing a range in 208Pb/204Pbi and 206Pb/204Pbi. This could result from recharge of fresh magma into a reservoir containing magma contaminated by lower crust, which mixed with the uncontaminated magma following the crystallization of the cores of the feldspars. The high 208Pb/204Pbi displayed by some of the single micro-milled albitic plagioclases (Figs 8 and 9
c) suggests that these crystals grew during ascent of the magmas through the Lewisian upper crust, becoming progressively more contaminated by it. Our previous AFC modelling, constrained in Sr–Pb space for the same crystals (Fig. 8), demonstrated that high 87Sr/86Sri and large 208Pb/204Pbi variations can be modelled by small additions of pegmatitic/granitic Lewisian material and also by addition of Lewisian upper crust. Therefore, we suggest that the isotope signatures of these crystals preserve the input of both lower and upper crustal assimilants.
The plagioclase crystals that show correlations between 87Sr/86Sri and An content (Fig. 6a and c) along core–rim profiles suggest recharge of high-temperature primitive magmas that became contaminated by the crust during ascent. After initial contamination in the Lewisian lower crust, these magma batches mixed with more evolved and less contaminated magma batches. These continuous changes, observed at such fine scales within the phenocrysts, suggest constant recharge of the magmatic system by new magmas that are variably contaminated by the crust during crystallization. These observations clearly support the finding and models of Huppert & Sparks (1985), Thompson et al. (1986) and Kerr et al. (1995).
Other, more complex models, such as EC-RAFC (Fowler et al., 2004), have been applied to explain assimilation, recharge and fractional crystallization processes that occurred during the evolution of the Skye lavas. Fowler et al. (2004) included a large compilation of whole-rock data in their model, and they concluded that the Skye lavas evolved from a relatively small number of parental magmas. They envisaged that the main contaminant for the early formed magmas is the Lewisian lower crust, whereas later erupted magmas interacted mainly with Lewisian upper crust. Isotopic variations observed in this study suggest that the contamination and differentiation pathways for each magma batch are considerably more complicated, as it is clear from our study that the magmas and their entrained crystals are not in isotope equilibrium.
Overall, the isotopic compositions of the micro-milled plagioclase phenocrysts suggest that the main crustal contaminants are pegmatitic/granitic veins within the lower crust for the majority of the magmas; locally, there is some evidence of contamination by Lewisian upper crust. However, the core-to-rim crystal isotopic variations observed in phenocrysts from different samples (Figs 8 and 9) show, for each specific crystal, which contaminants were assimilated into the host magma at the time of crystal growth. The data also reveal the pathways of crystallization and crustal contamination of the host magmas of such crystals. For example, plagioclase host magmas from the SMLS were mainly contaminated, to different degrees, by melts derived from Lewisian lower crust, but also, to a lesser extent, by Lewisian upper crust. Some SMLS plagioclase crystals show an increase of magma contamination by the crust as they grow from core to rim, whereas other plagioclase crystals show minimal core to rim isotopic variations. Hence, it is clear that very different magmatic processes occurred during the crystallization of such crystals, which grew from variably contaminated magmas erupted during the same volcanic event and are now found in the same rock. Additionally, the composition of the assimilant could have varied. The isotopic signatures of cores and rims of albitic plagioclase phenocrysts hosted in the Big-Feldspar dyke sample suggest that the host magmas were initially mostly contaminated by pegmatitic/granitic material in the lower crust. Further contamination occurred while the magmas were ascending through the upper crust, where a second stage of plagioclase crystallization occurred (Figs 8 and 9). The different crustal contamination signatures revealed by plagioclase phenocrysts and whole-rocks from different Skye volcanic rocks suggest variable magma residence at different depths in the crust, probably because of sill emplacement at different levels (Dickin et al., 1987). This scenario agrees with the fractional crystallization model of Thompson et al. (1982) (Fig. 3) where the SMLS fractionated at the base of the crust (9 kbar) and the PMB and other magmas at shallower depths (between 5 kbar and 1 atm).
Figure 10 is a schematic illustration intended to summarize the different magma pathways and crustal contamination stages reflected by the isotopic signatures of plagioclase crystals and whole-rocks from the Skye lava sequence; it assumes that all crystals are genetically related to the host lavas at some time. Figure 10 shows two scenarios that represent possible end-member differentiation pathways for the Skye magmas. Scenario 1 represents magmas that pause in the lower crust, where plagioclase crystals grow from magmas that have assimilated crustal melts derived from the lower Lewisian crust (SMLS). Scenario 2 represents a more complicated network of pathways of magmas ascending through the crust pausing at different levels in the middle–lower crust, where new plagioclase crystals grow from magmas variably contaminated by melts derived from Lewisian lower-middle and upper crust (PMB and Big-Feldspar samples).
|
The core–rim pairs, core-to-rim profiles, single micro-milled plagioclase and mineral separate isotope data presented in this study have been integrated into the interpretation of the well-known compositional fields delimited by the whole-rock data to enhance understanding of a highly complex volcanic system. The approach shows that isotope variations within crystals can provide more detailed information about magma end-members and crustal contaminants. Our study indicates that even when detailed whole-rock studies of crustal contamination reveal considerable complexity, further significant subtleties are likely to be revealed by isotope analyses at a core–rim and single-crystal scale.
|
SUMMARY |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
- Detailed micro-sampling of the SMLS reinforces conclusions based on whole-rock isotope data that the mantle-derived basalts were variably contaminated by the Lewisian crust through which they ascended.
- Core-to-rim isotopic variations show that single crystals grew in an open-system environment undergoing simultaneous crustal contamination and/or magma recharge. Examples where the cores and rims of crystals are in isotopic equilibrium with each other and the host whole-rock may reflect no contamination (although they are significantly different from the isotopic composition of assumed Hebridean mantle), crystal growth after contamination, or contamination by isotopically identical lithologies, such as antecedent intrusions.
- The fact that crystals with different core-to-rim isotopic characteristics are found in the same rock argues that the crystal cargo has been aggregated from a number of distinct sites of magma storage and differentiation.
- Correlation of core-to-rim isotope profiles with the isotope stratigraphy of the crust through which the magmas ascended allows us to reconstruct the likely sites of differentiation (Figs 9 and 10). In most instances it appears that lower crustal AFC processes dominated. On occasion, evidence is found for upper crustal AFC superimposed on earlier lower crustal processes. We have found no evidence for magmas affected only by upper crustal processes.
- Conclusions (2)–(4) greatly extend our understanding of the mechanisms of magma differentiation within the BPIP (as opposed to simply identifying contributing components). This is a valuable example of the utility of isotope micro-sampling of phenocrysts.
|
ACKNOWLEDGEMENTS |
|---|
We would like to thank an anonymous reviewer, Dennis Geist, Frank Ramos and editor Marjorie Wilson for their helpful reviews and suggestions to improve this manuscript. We would like to thank Dr H. Emeleus for his comments on the first version of this manuscript, Prof. Art Montana and Dr Ambre Luguet for their help during the sampling field trip to Skye, and Professor R.N. Thompson for providing samples and guidance at the start of this study. L.F. was supported by NERC grant NER/A/S/2003/00491.
*Corresponding author. Present address: Iva-Earth Science-Petrology Department, Vrije Universiteit Amsterdam, De Boelelaan 1085, 1081HV Amsterdam, The Netherlands. Telephone: +31 20 5987345. E-mail: laura.font{at}falw.vu.nl
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTHE SKYE IGNEOUS CENTRESAMPLING AND SAMPLE DESCRIPTIONSANALYTICAL METHODSISOTOPE DATAINTERPRETATIONSUMMARYREFERENCES |
|---|
Abbey S. Studies in ‘Standard Samples’ of Silicate Rocks and Minerals 1969–1982. Canadian Geological Survey Paper (1982) 83-15:114–.
Bell BR, Williamson IT. Tertiary Igneous Activity. The Geology of Scotland—Terwin NH, ed. (2002) 4th edn. London: Geological Society. 371–407.
Bohrson WA, Spera FJ. Energy constrained open-system magmatic processes IV: Geochemical, thermal and mass consequences of energy constrained re-charge, assimilation and fractional crystallization. Geochemistry, Geophysics, Geosystems (2003) 4:1–25.[Medline]
Carter SR, Evensen NM, Hamilton PJ, O’Nions RK. Nd- and Sr-isotopic evidence for crustal contamination of continental volcanics. Science (1978) 202:743–747.
Chambers LM, Pringle MS. Age and duration of activity at the Isle of Mull Tertiary igneous centre, Scotland, and confirmation of the existence of subchrons during Anomaly 26r. Earth and Planetary Science Letters (2001) 193:333–345.[CrossRef][Web of Science]
Charlier BLA, Ginibre C, Morgan D, Nowell GM, Pearson DG, Davidson JP, Ottley CJ. Methods for the microsampling and high-precision analysis of strontium and rubidium isotopes at single crystal scale for petrological and geochronological applications. Chemical Geology (2006) 232:114–133.[CrossRef][Web of Science]
Davidson JP, Tepley FJ III, Knesel KM. Isotopic fingerprinting may provide insights into evolution of magmatic systems. EOS Transactions, American Geophysical Union (1998) 79. 185, 189, 193.
Davidson JP, Tepley FJ III, Palacz Z, Meffan-Main S. Magma recharge, contamination and residence times revealed by in situ laser ablation isotopic analysis of feldspar in volcanic rocks. Earth and Planetary Science Letters (2001) 184:427–442.[CrossRef][Web of Science]
Davidson JP, Morgan DJ, Charlier BLA, Harlou R, Hora JM. Microsampling and isotopic analysis of igneous rocks: Implications for the study of magmatic systems. Annual Review of Earth and Planetary Sciences (2007) 35:273–311.[CrossRef][Web of Science]
DePaolo DJ. Trace element and isotopic effects of combined wall-rock assimilation and fractional crystallisation. Earth and Planetary Science Letters (1981) 53:189–202.[CrossRef][Web of Science]
Dickin AP. Isotope geochemistry of Tertiary igneous rocks from the Isle of Skye, N.W. Scotland. Journal of Petrology (1981) 22:155–189.
Dickin AP, Exley RA. Isotopic and geochemical evidence for magma mixing in the petrogenesis of the Coire Uaigneich Granophyre, Isle of Skye, NW Scotland. Contributions to Mineralogy and Petrology (1981) 76:98–108.[CrossRef][Web of Science]
Dickin AP, Brown JL, Thompson RN, Halliday AN, Morrison MA. Crustal contamination and the granite problem in the British Tertiary Volcanic Province. Philosophical Transactions of the Royal Society of London, Series A (1984) 310:755–780.[CrossRef]
Dickin AP, Jones NW, Thirlwall MF, Thompson RN. A Ce/Nd isotope study of crustal contamination processes affecting Paleocene magmas in Skye, Northwest Scotland. Contributions to Mineralogy and Petrology (1987) 96:455–464.[CrossRef][Web of Science]
Dunham AC. The emplacement of the Tertiary igneous complex of Rhum. Geological Journal, Special Issue (1970) 2:23–32.
Ellam RM. Lithospheric thickness as a control on basalt geochemisty. Geology (1992) 20:153–156.
Emeleus CH. The central complexes. In: Igneous Rocks of the British Isles—Sutherland DS, ed. (1982) Chichester: John Wiley. 369–414.
Emeleus CH, Bell BR. The Palaeogene Volcanic District of Scotland. (2005) Keyworth, Nottingham: British Geological Survey.
Emeleus CH, Gyopari MC. British Tertiary Volcanic Province (1992) London: Chapman and Hall. 259 pp.
England RW. The structure of the Skye lava field. Scottish Journal of Geology (1994) 30:33–37.[Web of Science]
Flanagan FJ. Descriptions and Analysis of Eight New USGS Rock Standards. US Geological Survey, Professional Papers 840 (1976) 192.
Font L, Nowell GM, Pearson DG, Ottley C, Willis SG. Sr isotope analysis of bird feathers by TIMS: a tool to trace bird migration paths and breeding sites. Journal of Analytical Atomic Spectrometry (2007) 22:513–522.[CrossRef][Web of Science]
Forester RW, Taylor H. P. Jr. 18O/16O, D/H, and 13C/12C studies of the Tertiary igneous complex of Skye, Scotland. American Journal of Science (1977) 277:136–177.
Fowler SJ, Bohrson WA, Spera FJ. Magmatic evolution of the Skye igneous centre, Western Scotland: Modelling of assimilation, recharge and fractional crystallisation. Journal of Petrology (2004) 45:2481–2505.
Galer SJG. Optimal triple spiking for high precision lead isotope ratio determination. Terra Nova Supplement (1997) 9:441.
Gagnevin D, Daly JS, Poli G, Morgan D. Microchemical and Sr isotopic investigations of zoned K-feldspar megacrysts: insights into the petrogenesis of a plutonic system and disequilibrium processes during crystal growth. Journal of Petrology (2005) 46:1689–1724.
Geist DJ, Meyers JD, Frost CD. Megacrysts bulk rock isotopic disequilibrium as an indicator of contamination processes; the Edgecumbe Volcanic Field, S.E. Alaska. Contributions to Mineralogy and Petrology (1988) 99:105–112.[CrossRef][Web of Science]
Govindaraju K. 1994 Compilation of working values and descriptions for 383 geostandards. Geostandards Newsletter (1994) 18:1–158.
Hamilton PJ, Evensen NM, O’Nions RK, Tarney J. Sm–Nd systematics of Lewisian gneisses: implications for the origin of granulites. Nature (1979) 277:25–28.[CrossRef][Web of Science]
Hamilton MA, Pearson DG, Thompson RN, Kelley SP, Emeleus CH. Rapid eruption of Skye lavas inferred from precise U–Pb and Ar–Ar dating of the Rum and Cuillin plutonic complexes. Nature (1998) 394:260–263.[CrossRef]
Higgins MD, Chandrasekharam D. Nature of sub-volcanic magma chambers, Deccan province, India: Evidence from quantitative textural analysis of plagioclase megacrysts in the Giant Plagioclase Basalts. Journal of Petrology (2007) 48:885–900.
Huppert HE, Sparks RSJ. Cooling and contamination of mafic and ultramafic magmas during ascent through continental crust. Earth and Planetary Science Letters (1985) 74:371–386.[Web of Science]
Jerram DA, Widdowson M. The anatomy of Continental Flood Basalt Provinces: Geological constraints on the processes and products of flood volcanism. Lithos (2005) 79:385–405.[CrossRef][Web of Science]
Kerr AC. On the nature of the parental magma of the Palaeogene Staffa magma sub-type, Isle of Mull, Scotland. Transactions of the Royal Society of Edinburgh (1998) 89:87–93.
Kerr AC, Kempton PD, Thompson RN. Crustal assimilation during turbulent magma ascent (ATA); new isotopic evidence from the Mull Tertiary lava succession, N.W. Scotland. Contributions to Mineralogy and Petrology (1995) 119:142–154.[Web of Science]
Knesel KM, Davidson JP, Duffield WA. Evolution of silicic magma through assimilation and subsequent recharge: evidence from Sr isotopes in Sanidine phenocrysts, Taylor Creek Rhyolite, NM. Journal of Petrology (1999) 40:773–786.[CrossRef][Web of Science]
Mattey DP, Gibson IL, Marriner GF, Thompson RN. The diagnostic geochemistry, relative abundance and spatial distribution of high-Ca, low-alkali tholeiite dykes in the lower Tertiary regional swarm of the Isle of Skye, NW Scotland. Mineralogical Magazine (1977) 41:273–285.[Web of Science]
Moorbath S, Thompson RN. Strontium isotope geochemistry and petrogenesis of the early Tertiary lava pile of the Isle of Skye, Scotland, and other basic rocks of the British Tertiary Province: An example of magma–crust interaction. Journal of Petrology (1980) 21:295–321.
Morgan DJ, Jerram DA, Chertkoff DG, Davidson JP, Pearson DG, Kronz A, Nowell GM. Combining CSD and isotopic microanalysis: Magma supply and mixing processes at Stromboli Volcano, Aeolian Islands, Italy. Earth and Planetary Science Letters (2007) 260:419–431.[CrossRef][Web of Science]
Morrison MA, Thompson RN, Dickin AP. Geochemical evidence for complex magmatic plumbing during development of a continental volcanic centre. Geology (1985) 13:581–584.
Nowell GM, Pearson DG, Ottley CJ, Schweiters J, Dowall D. Long-term performance characteristics of a plasma ionisation multi-collector mass spectrometer (PIMMS): the ThermoFinnigan Neptune. In: Plasma Source Mass Spectrometry: Applications and Emerging Technologies—Holland JG, Tanner SD, eds. (2003) Cambridge: Royal Society of Chemistry. 307–320.
Ramos FC, Wolff JA, Tollstrup DL. Sr isotope disequilibrium in Columbia River flood basalts: Evidence for rapid shallow-level open-system processes. Geology (2005) 33:457–460.
Saunders AD, Fitton JG, Kerr AC, Norry MJ, Kent RW. The North Atlantic Igneous Province. In: Large Igneous Provinces: Continental, Oceanic, and Planetary Flood Volcanism. Geophysical Monograph, American Geophysical Union—Mahoney JJ, Coffin MF, eds. (1997) 100:411–438.
Scarrow JH, Cox KG. Basalts generated by decompressive adiabatic melting of a mantle plume: A case study from the Isle of Skye, NW Scotland. Journal of Petrology (1995) 36:3–22.
Scarrow JH, Curran JM, Kerr AC. Major element records of variable plume involvement in the North Atlantic Province Tertiary flood basalts. Journal of Petrology (2000) 41:1155–1176.
Single RT, Jerram DA. The 3D facies architecture of flood basalt provinces and their internal heterogeneity: examples from the Palaeogene Skye Lava Field. Journal of the Geological Society, London (2004) 161:911–926.
Tepley FJ III, Davidson JP, Clynne MA. Magmatic interactions as recorded in plagioclase phenocrysts of Chaos Crags, Lassen Volcanic Centre, California. Journal of Petrology (1999) 40:787–806.[CrossRef][Web of Science]
Tepley FJ III, Davidson JP, Tilling RI, Arth JG. Magma mixing, recharge and eruption histories recorded in plagioclase phenocrysts from El Chichón volcano, Mexico. Journal of Petrology (2000) 41:1397–1411.
Thirlwall MF. Long-term reproducibility of multicollector Sr and Nd isotope ratio analysis. Chemical Geology (1991) 94:85–104.[Web of Science]
Thirlwall MF, Jones NW. Isotope geochemistry and contamination mechanics of Tertiary lavas from Skye, Northwest Scotland. In: Continental Basalts and Mantle Xenoliths—Hawkesworth CJ, Norry MJ, eds. (1983) Nantwich: Shiva. 186–208.
Thompson RN, Gibson SA. Subcontinental mantle plumes, hotspots and pre-existing thinspots. Journal of the Geological Society, London (1991) 148:973–977.
Thompson RN, Esson J, Dunham AC. Major element chemical variation in the Eocene lavas of the Isle of Skye, Scotland. Journal of Petrology (1972) 13:219–253.
Thompson RN, Gibson SA, Marriner GF, Mattey DP, Morrison MA. Trace element evidence of multistage mantle fusion and polybaric fractional crystallisation in the Paleocene lavas of Skye, NW Scotland. Journal of Petrology (1980) 21:265–293.
Thompson RN, Dickin AP, Gibson IL, Morrison MA. Elemental fingerprints of isotopic contamination of Hebridean Paleocene mantle-derived magmas by Archean sial. Contributions to Mineralogy and Petrology (1982) 79:159–168.[CrossRef][Web of Science]
Thompson RN, Morrison MA, Dickin AP, Gibson IL, Harmon RS. Two contrasting styles of interaction between basic magmas and continental crust in the British Tertiary Igneous Province. Journal of Geophysical Research (1986) 91:5985–5997.
Troll VR, Donaldson CH, Emeleus CH. Pre-eruptive magma mixing in ash-flow deposits of the Tertiary Rum Igneous Centre, Scotland. Contributions to Mineralogy and Petrology (2004) 147:722–739.[Web of Science]
Walker GPL. Basaltic-volcano systems. In: Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications—Prichard HM, Alabaster T, Harris NBW, Neary CR, eds. (1993) 76:3–38.
Weaver BL, Tarney J. Rare earth geochemistry of Lewisian granulite-facies gneisses, northwest Scotland: implications for the petrogenesis of the Archean lower crust. Earth and Planetary Science Letters (1980) 51:279–296.[CrossRef][Web of Science]
Weaver BL, Tarney J. Lewisian gneiss geochemistry and Archean crustal development models. Earth and Planetary Science Letters (1981) 55:171–180.[CrossRef][Web of Science]
Williamson IT, Bell BR. The Paleocene lava field of west–central Skye, Scotland: stratigraphy, paleogeography and structure. Transactions of the Royal Society of Edinburgh: Earth Science (1994) 85:39–75.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. A. JERRAM, K. M. GOODENOUGH, and V. R. TROLL Introduction: from the British Tertiary into the future - modern perspectives on the British Palaeogene and North Atlantic Igneous provinces Geological Magazine, May 1, 2009; 146(3): 305 - 308. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||