Journal of Petrology

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (5) Request Permissions Google Scholar Articles by HANGHØJ, K. Articles by STECHER, O. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Journal of Petrology | Volume 44 | Number 11 | Pages 2081-2112 | 2003
© Oxford University Press 2003; all rights reserved

An Isotope and Trace Element Study of the East Greenland Tertiary Dyke Swarm: Constraints on Temporal and Spatial Evolution during Continental Rifting

K. HANGHØJ^*, M. STOREY and O. STECHER

DANISH LITHOSPHERE CENTRE, ØSTER VOLDGADE 10, 1350 COPENHAGEN K, DENMARK

^* Corresponding author. Present address: Woods Hole Oceanographic Institution, G&G, MS# 8, Woods Hole, MA 02543, USA. Telephone: +1 (508) 289 2946. Fax +1 (508) 457 2183. E-mail: khanghoj{at}whoi.edu

RECEIVED SEPTEMBER 4, 2000; ACCEPTED MAY 26, 2003

	ABSTRACT

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
Dykes of the East Greenland Tertiary dyke swarm can be divided into pre- and syn-break-up tholeiitic dykes, and post-break-up transitional dykes. Of the pre- and syn-break-up dykes, the most abundant group (Tholeiitic Series; TS) has major element compositions similar to the main part of the East Greenland flood basalts. A group of high-MgO tholeiitic dykes (Picrite–Ankaramite Series; PAS) are much less common, and are equivalent to some of the oldest lavas of the East Greenland flood basalts. Isotopic compositions of the TS and PAS dykes partly overlap with those for Iceland, but Pb isotopic compositions extend to less radiogenic values than those seen in either Iceland or North Atlantic mid-ocean ridge basalt (MORB). The isotopically depleted source required to account for this isotopic variation is interpreted as subcontinental lithospheric mantle with low ⁸⁷Sr/⁸⁶Sr and ²⁰⁶Pb/²⁰⁴Pb and high _Nd. The post-break-up Transitional Series (TRANS) dykes are isotopically distinct from Iceland and MORB, and are interpreted as the products of contamination of Iceland plume melts with continental crust. Comparison of the Nd–Sr–Pb isotopic and trace element compositions of dykes from different segments of the East Greenland margin indicates that there is no systematic compositional change with distance from the presumed proto-Icelandic plume centre. This suggests that a northward-increasing crustal thickness observed offshore may be attributed to active upwelling rather than a systematic rise in temperature towards the plume centre.

KEY WORDS: isotopes; trace elements; mantle sources; mantle melting

	INTRODUCTION

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
The East Greenland Tertiary Igneous Province is part of the rifted volcanic margin related to the opening of the North Atlantic Ocean (Fig. 1). The province comprises flood basalts, mafic and felsic intrusive centres, and a coast-parallel dyke swarm, which is exposed for more than 500 km along the present coastline. Offshore from the province, a thick seaward-dipping reflector sequence (SDRS), which was drilled by Ocean Drilling Program (ODP) Legs 152 and 163, consists of voluminous subaerially erupted basalt flows (e.g. Larsen & Saunders, 1998). The Greenland–European continent that was rifted during the Tertiary consists of rocks ranging in age from early Archaean to late Mesozoic. In Greenland, the Caledonian front marks the western extent of deformation associated with the Caledonian orogeny as Greenland and Europe collided in Silurian–Devonian time. The basement west of the Caledonian front (Fig. 1) consists of Archaean blocks and Proterozoic mobile belts (e.g. Escher & Watt, 1976; Bridgwater et al., 1978) whereas east of the Caledonian front it consists of deformed Precambrian basement and early Palaeozoic sedimentary and igneous rocks (Henriksen & Higgins, 1976).

View larger version (80K): [in this window] [in a new window]   Fig. 1. Distribution of onshore and offshore basalts in the North Atlantic Igneous Province. ODP drill sites on the Greenland margin are shown. FIGR, Faeroe–Iceland–Greenland Ridge; JMR, Jan Mayen Ridge. Selected sea-floor magnetic anomalies are shown by dashed lines.

 
In the North Atlantic region, active volcanism is at present restricted to the Mid-Atlantic Ridge. Iceland represents an area of high magma supply associated with the thermal and compositional anomaly generally referred to as the Iceland hotspot or mantle plume (Fig. 1). Some plate tectonic reconstructions place the Iceland hotspot beneath central Greenland during the time of break-up, but the timing and exact location for the initial impact of the Iceland plume is subject to different interpretations. Lawver & Müller (1994) proposed that the Iceland plume dates back as far as the Permo-Triassic, and placed the centre of the hotspot beneath central West Greenland at 60 Ma, and beneath central East Greenland around 40 Ma (Fig. 2). Brooks (1973), White & McKenzie (1989) and Saunders et al. (1997) proposed that the arrival and emplacement of the plume head preceded the rift-related volcanism in the North Atlantic, and placed the hotspot under central East Greenland (the Kangerlugssuaq area) before initial break-up at 58–60 Ma.

View larger version (46K): [in this window] [in a new window]   Fig. 2. Map showing the localities studied in this work and the three main segments of the dyke swarm after Myers (1980). Proposed hotspot track of Lawver & Müller (1994) is shown on insert map of Greenland, with numbers indicating location at time in million years.

 
The East Greenland margin is thus potentially an early manifestation of the Iceland mantle plume. Igneous rocks exposed along the margin provide a unique possibility to examine how the proto-Icelandic plume interacted with the lithosphere as break-up proceeded, and whether other mantle sources [e.g. mid-ocean ridge basalt (MORB) source mantle and continental lithospheric mantle] were involved in the melt generation. The East Greenland margin can also potentially provide information about the compositional and thermal structure of the Iceland plume at the time of break-up, because of the good temporal and spatial control and exposure of different structural levels.

This study documents and examines the geochemistry of the East Greenland coastal dyke swarm, a major expression of the break-up-related magmatism. In general, the dykes represent magmas that are equivalent to the flood basalts erupted at the surface. However, as the dykes occur in areas outside the area covered by flood basalts, some dykes may have no preserved eruptive equivalents. This study is complementary to continuing studies of the East Greenland lavas (e.g. Larsen et al., 1996a; Pedersen et al., 1997; Tegner et al., 1998b) by extending, both geographically and temporally, information about the nature of magmatism along the rifted margin before, during and subsequent to break-up. We present new major and trace element data for 115 dykes, and Sr, Nd and Pb isotope data for a representative selection of the dyke samples. The main aim of the chemical and isotopic study is to identify the mantle sources for magmas generated during formation of this volcanic rifted margin, and to place constraints on the temporal and spatial variations in magma chemistry.

	THE EAST GREENLAND DYKE SWARM

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
Field relations and sampling
The coastal dyke swarm of the East Greenland margin was first described by Wager (1935, 1947), and subsequently by Nielsen (1978) and Myers (1980), and most recently by Klausen & Larsen (2002). It consists of an 10–30 km wide zone along the present coastline, where coast-parallel dykes constitute up to >50% of the outcrop. The swarm is most dense from Cape Wandel north of Ammasalik to Nansen Fjord just north of Kangerlugssuaq Fjord, although Tertiary dykes are present north of Nansen Fjord and at least as far south as Skjoldungen, some 400 km south of Ammasalik (Fig. 1). The offshore extension of the dyke swarm is unknown, but as the frequency of dykes generally increases seawards, the most intense zone of dyking is probably located offshore.

Typically the dyke density decreases from more than 50% on the outer coast to less than 5% some 20–30 km inland. Locally, sheeted dykes with dyke densities >90% occur. Associated with the rifting, a large monocline structure [the coast-parallel flexure of Wager & Deer (1938)] developed along the entire margin, causing seaward rotation of fault blocks with resultant landward dip of dykes, as pointed out for the Kangerlussuaq region by Nielsen (1978) and Nielsen & Brooks (1981). Progressive rotation has ensured that dykes on the outer part of the coast are rotated more than dykes further inland, and that early dykes are rotated more than late dykes. The development of the monocline structure was accommodated by Tertiary normal faulting at various scales, ranging from slip along dyke margins to large fault zones with associated cataclasite and pseudotachylyte formation (Karson et al., 1998).

Four localities were chosen for detailed geochemical work, and these are (from south to north) Tasiilaq, Langø, Fladø and I. C. Jacobsen Fjord (Fig. 2). At these localities, collection of representative samples was accomplished through detailed mapping for crosscutting relations and sampling in different parts of the swarm, i.e. the most inland part where the dyke density is less than 5%, the middle part with intermediate dyke densities, and the outer part where dyke densities reach more than 50%.

Analytical methods
Most dykes were sampled close to the margin, i.e. in the most fine-grained portion of the dyke. Weathered surfaces were removed before the samples were crushed in a jaw crusher and then powdered to approximately 200 mesh in an agate shatter box.

The dykes were analysed for major elements by the GEUS (Geological Survey of Denmark and Greenland) laboratory in Copenhagen. SiO₂, TiO₂, Al₂O₃, Fe₂O₃, MnO, MgO, CaO, K₂O and P₂O₅ were determined by X-ray fluorescence (XRF) on fused beads. Na₂O was analysed by atomic absorption spectrometry, and FeO by titration as described by Kystol & Larsen (1999). Some trace element abundances (Zn, Cu, Co, Ni, Sc, V and Cr) were determined by XRF on pressed powder tablets at the Geological Institute at University of Copenhagen.

For additional trace element concentrations, the dykes were analysed by inductively coupled plasma mass spectrometry (ICP-MS) (Fisons PQ2+ PlasmaQuad) at the College of Oceanography and Atmospheric Sciences, Oregon State University, using a digestion procedure involving fusion of the samples to ensure that possible residual phases, such as zircon and chromite, were completely dissolved. Sample solutions were prepared by fusing 200 mg of whole-rock powder with 800 mg of lithium-metaborate flux in graphite crucibles at 1100°C, and dissolving the resulting glass bead in 50 ml of 2N HNO₃. Before analysis, the samples were further diluted with 1% HNO₃ in proportions 1:10. Instrumental drift was corrected using internal standard solutions added to the sample solutions as described by Pyle et al. (1995). Element concentrations for the samples were determined from regression curves for rock standards processed along with the samples [BHVO-1, BIR-1, BR or BE-N, W-2; using recommended values from Govindaraju (1989, 1994) and Cheatham et al. (1993)]. The reproducibility of a monitor sample, which was analysed at least three times for every 20 samples, is 6% or less for most elements.

Thirty-two of the samples were analysed for Sr, Nd and Pb isotope compositions on a VG-354 multi-collector mass spectrometer at the Danish Centre for Isotope Geology at the University of Copenhagen. These were selected to ensure a range of compositions within each group, and to avoid altered samples (on petrographic criteria) and those obviously contaminated with continental crust (e.g. samples with anomalously high SiO₂ contents and La/Nb ratios). Samples selected for isotope analyses were acid-leached following the multi-step HCl-leaching procedure described by Mahoney (1987), with the aim of eliminating all secondary phases. Results are given in Tables 1 and 2.

View this table: [in this window] [in a new window]   Table 1a: Major and trace element analyses of Tholeiitic Series dykes

 

View this table: [in this window] [in a new window]   Table 1b: Major and trace element analyses of PAS dykes

 

View this table: [in this window] [in a new window]   Table 1c: Major and trace element analyses of Transitional Series dykes

 

View this table: [in this window] [in a new window]   Table 1d: Reproducibility of monitor standard 95358

 

View this table: [in this window] [in a new window]   Table 2: Sr–Nd–Pb isotope data

 

	CLASSIFICATION AND CHRONOLOGY OF DYKES

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
This study uses a modified version of the classification proposed by Nielsen (1978), Gill et al. (1988) and Hansen (1997), based on field relations and major element chemistry. The dykes analysed in this study can be divided into three main groups, the Tholeiitic Series (TS), the Picrite–Ankaramite Series (PAS), and the Transitional Series (TRANS) (Fig. 3). Importantly, the distinctions between these groups are not very rigorous. All three groups have a range of compositions, and some dykes are intermediate between the groups. This is especially true for dykes of the Transitional Series and Tholeiitic Series.

View larger version (33K): [in this window] [in a new window]   Fig. 3. TiO2 (wt %) against Mg number (molar Mg/Mg + Fe). Data for Lower Basalts (Fram & Lesher, 1997; Hansen, 1997; Danish Lithosphere Centre (DLC), unpublished data, 1995–1998) and Main Plateau Lavas (Larsen et al., 1989; DLC, unpublished data, 1995–1998). Fields for Tholeiitic Series (TS) and Picrite–Ankaramite Series (PAS) are the originally proposed fields from Gill et al. (1988).

 
Tholeiitic Series (TS)
These dykes, in terms of major elements, are similar to most of the flood basalts and specifically to some lavas from the oldest part of the flood basalt sequence, the Lower Basalts (e.g. Gill et al., 1988; Larsen et al., 1989; Hansen, 1997). The TS dykes in this study are further divided into high Zr/Nb TS dykes and low Zr/Nb TS dykes (Fig. 4) on the basis of their trace element compositions. With the exception of one low Zr/Nb TS dyke, all TS samples are hypersthene normative.

View larger version (13K): [in this window] [in a new window]   Fig. 4. Diagram of La/Yb against Zr/Nb used for the subdivision of the TS dykes.

 
Dykes of the TS constitute more than 70% of the dyke swarm at all localities. They are typically 2–20 m wide and strike roughly parallel to the coast at all localities. They generally dip 60° to 70° landward and they tend to be wider and more steeply dipping in the regions furthest from the coast. Crosscutting relations within the TS dykes are common, and show that the low Zr/Nb TS dykes generally are younger than the high Zr/Nb TS dykes, although they in some cases appear to be contemporaneous.

Petrographically there is no obvious difference between high Zr/Nb TS dykes and low Zr/Nb TS dykes. Both types are typically medium- to coarse-grained dolerites, with ubiquitous plagioclase, clinopyroxene, olivine and Fe–Ti oxide minerals. Low abundances of plagioclase and/or olivine phenocrysts are common, but phenocryst-rich types are rare.

Plagioclase, clinopyroxene and oxide minerals are generally not altered. Partly fresh olivine phenocrysts are very rare, and groundmass olivine is always completely replaced. The most common secondary phases are chlorite, sericite and epidote.

Tholeiitic Picrite–Ankaramite Series (PAS)
These dykes correspond to parts of the Lower Basalts (e.g. Nielsen, 1978; Hansen, 1997; Gill et al., 1998), are relatively rare, and probably constitute less than 1% of the dyke swarm. Most samples of PAS dykes come from the Tasiilaq area, which reflects higher sampling intensity in that area, rather than an increased abundance of PAS dykes. Most of the PAS dykes strike roughly parallel to the coast.

The PAS dykes range in thickness from less than 0·5 m to more than 10 m, and dip between 50° landwards and vertical. Although volumetrically insignificant, the tholeiitic PAS are important in this study because they are amongst the oldest of the dykes. Crosscutting relations show that the PAS dykes are generally older than, or contemporaneous with, the oldest of the TS dykes, and thus belong to the earliest phase of magmatism. All PAS dykes are hypersthene normative.

The PAS dykes typically have olivine phenocrysts and clinopyroxene and/or olivine microphenocrysts in a groundmass of olivine, clinopyroxene, Fe–Ti oxide minerals and plagioclase. The least MgO-rich samples have a doleritic texture with less than 5% phenocrysts. The more MgO-rich dykes typically have more abundant phenocrysts and commonly also have a quench texture in the groundmass defined by composite star-shaped plagioclase grains. Some PAS dykes contain partly resorbed olivine xenocrysts and/or dunite fragments.

Groundmass clinopyroxene and plagioclase is commonly altered, and groundmass olivine is always replaced. Phenocrysts are typically partly replaced with fresh centres. The most common secondary phases are chlorite, sericite and epidote.

The Transitional Series (TRANS)
This group contains both hypersthene normative and nepheline normative dykes, and does not correlate with any known sequence in the flood basalts (Fig. 3). The transitional dykes constitute 20–30% of the dyke swarm, and at all localities they are uncommon in the most inland areas. They are always amongst the youngest in groups of crosscutting dykes. They are usually light grey, coarse-grained dolerites, wider than 5 m, and subvertical with an overall coast-parallel strike. The transitional dykes thus post-date the coastal flexure.

The TRANS dykes typically have a fine-grained doleritic texture with ubiquitous plagioclase, clinopyroxene and Fe–Ti oxide minerals, and occasional olivine and apatite. Some samples have primary biotite and amphibole. Small amounts of plagioclase phenocrysts (<5%) are common.

The TRANS dykes are often very altered with sericitized plagioclase, and completely replaced clinopyroxene. Olivine is never fresh. Some samples are relatively unaltered with fresh plagioclase, clinopyroxene and biotite.

Crosscutting relations show that the relative chronology of the dykes is, from oldest to youngest, PAS, high Zr/Nb TS dykes, low Zr/Nb TS dykes and TRANS dykes. This chronology can only be considered approximate in the sense that the PAS, high Zr/Nb and low Zr/Nb groups occasionally show crosscutting relationships discordant to this general chronology. This is consistent with the observation of intercalated PAS and TS lavas in the Lower Basalts (Nielsen & Brooks, 1981; Gill et al., 1988; Hansen, 1997).

A further constraint on the chronology was provided by Tegner et al. (1998a), who obtained a ⁴⁰Ar/³⁹Ar date of 56·2 ± 0·6 Ma on a TS dyke from the Imilik area between Tasiilaq and Langø (Fig. 2), and by a preliminary ⁴⁰Ar/³⁹Ar age of 52 Ma obtained for a dyke of the Transitional Series from Tasiilaq (R. A. Duncan, personal communication, 2000).

The TS dykes and PAS dykes thus appear to be pre- and syn-break-up, i.e. intrusive equivalents of the Lower Basalts and the main part of the flood basalts [Main Lavas of Larsen et al. (1989)]. The transitional dykes have no known geochemical equivalents in the basalts, and are post-break-up, and perhaps contemporaneous with some of the post-rift gabbroic intrusions along the coast (Bernstein et al., 1998b). The relative chronology is in accordance with previous findings of Nielsen (1978) that the dyke swarm records a general evolution from tholeiitic magmatism to transitional and alkaline magmatism.

	RESULTS

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
Major element chemistry
Major element analyses are given in Table 1, and shown in Fig. 5. The TS dykes have 47–50 wt % silica and 2–5 wt % total alkalis. The alkali contents are generally slightly higher in the low Zr/Nb TS dykes than in the high Zr/Nb TS dykes. MgO contents range from a little less than 5 wt % to 8 wt % with negatively correlated TiO₂, FeO^*, P₂O₅, Na₂O and K₂O. There is a positive correlation between MgO and Al₂O₃ concentrations, which range from 12 to 16%. CaO/Al₂O₃ ratios are relatively constant around 0·75 to 0·85.

View larger version (28K): [in this window] [in a new window]   Fig. 5. Major element compositions of dykes. Oxides are in wt % and chemical analyses are normalized to 100% volatile free with Fe2O3/FeO ratios of 0·15 as suggested by Brooks (1976). Arrows show the direction of the evolution of liquid with simple fractionation, i.e. subtraction of olivine, clinopyroxene, plagioclase and magnetite, using representative mineral compositions from Fram & Lesher (1997) and Hansen (1997).

 
MgO contents for the PAS dykes range from 9 to 25 wt % and show a negative correlation with Na₂O and Al₂O₃ (Fig. 5). For samples with less than 15% MgO, there is also a negative correlation with TiO₂ and FeO^*. Compared with the TS dykes, the PAS dykes are enriched in K₂O, P₂O₅ and TiO₂, i.e. similar or higher concentrations than in the TS dykes for higher MgO contents.

The TRANS dykes are relatively MgO- and TiO₂-poor (3·5–5·5 wt % and 2–3·5 wt %, respectively), and there is no obvious correlation between MgO and TiO₂ (Fig. 5). For similar MgO contents, the TRANS dykes are richer in SiO₂, Na₂O and K₂O than the TS dykes, and poorer in TiO₂, CaO and FeO^*, and they range in composition from tholeiitic basalts and andesites to hawaiites and trachybasalts. There is a negative correlation between MgO and Na₂O, K₂O and P₂O₅. The CaO/Al₂O₃ ratios are lower for the TRANS dykes (0·4–0·75) than for the TS dykes, and there is a positive correlation with MgO (Fig. 5).

Trace element chemistry
Trace element abundances in the TS dykes are illustrated as primitive mantle-normalized trace element variation diagrams in Fig. 6a. The main difference in trace element characteristics between the high Zr/Nb TS dykes and low Zr/Nb TS dykes is that the latter show greater enrichment in the most incompatible elements such as Ba, Nb and the light rare earth elements (LREE). For elements to the right of Ti [Y and middle to heavy REE (MREE–HREE)] the high Zr/Nb TS dykes are slightly more enriched than the low Zr/Nb TS dykes (Fig. 6a). Most of the TS dykes have a positive Ti anomaly and negative K and Sr anomalies relative to primitive mantle.

View larger version (41K): [in this window] [in a new window]   Fig. 6. Incompatible element abundances of dykes normalized to primitive mantle (Sun & McDonough, 1989). (a) All TS dykes with 5–6 wt % MgO. Continuous and dashed black lines show average composition for all low Zr/Nb TS and high Zr/Nb TS dykes, respectively. (b) PAS dykes with 9–15 wt % MgO compared with TS dykes. (c) TRANS dykes with over 5 wt % MgO compared with TS dykes.

 
Figure 6b is a primitive mantle-normalized diagram for PAS dykes with MgO contents between 9 and 15 wt %. For elements more incompatible than Sr, the PAS dykes have trace element contents as high as some of the TS dykes, despite the more primitive nature of the PAS dykes. For elements less incompatible than Sr (including Y and the MREE–HREE), the PAS dykes are less enriched than the TS dykes.

In Fig. 6c mantle-normalized trace element abundances of the least evolved TRANS samples (with MgO >5 wt %) are shown. The pattern for the TRANS dykes is generally similar to that of the low Zr/Nb TS dykes, i.e. they have a relatively steep pattern in the mantle-normalized diagram.

Dykes of all three groups resemble ocean-island basalt (OIB) compositions in that they are enriched in incompatible elements, and in a diagram of Zr/Y against Nb/Y ratios (Fig. 7) most of the dykes plot in the field of enriched Icelandic basalts. Fitton et al. (1997, 1998b) has proposed that this diagram can be used to distinguish between Icelandic and MORB mantle sources.

View larger version (15K): [in this window] [in a new window]   Fig. 7. Nb/Y and Zr/Y variation in East Greenland dykes. The parallel lines mark the limit of Iceland data according to Fitton et al. (1997). Compositions of primitive mantle (PM), MORB and OIB are from Sun & McDonough (1989), and lower crust (LC), bulk crust (BC) and upper crust (UC) are from Taylor & McLennan (1995).

 
Isotope data
A total of three PAS dykes (all from Tasiilaq), 22 TS dykes and seven TRANS dykes were analysed for their Sr–Nd–Pb isotopic compositions. Age-corrected Sr, Nd and Pb isotope data for the dykes are given in Table 2 and shown in Figs 8–11 along with data for North Atlantic MORB and Iceland.

View larger version (18K): [in this window] [in a new window]   Fig. 8. Initial (age-corrected according to Table 2) Nd, i.e. 104 x {[143Nd/144Ndsample (t)/143Nd/144NdCHUR (t)] – 1}, and 87Sr/86Sr compositions for the dykes compared with North Atlantic MORB and Iceland. Error bars (2) for repeat analyses of BCR-1 are smaller than symbol size. Iceland and MORB data for this and following figures are from Sun & Jahn (1978), Dupré & Allègre (1980), Cohen & O'Nions (1982), Condomines et al. (1983), Ito et al. (1987), Shirey et al. (1987), Elliott et al. (1991), Mertz et al. (1991), Nicholson et al. (1991), Nicholson & Latin (1992), Frey et al. (1993), Hémond et al. (1993), Mühe et al. (1993), Hanan & Schilling (1997), Taylor et al. (1997) and Stecher et al. (1999).

 

View larger version (23K): [in this window] [in a new window]   Fig. 9. Pb isotopic compositions for dykes, North Atlantic MORB and Iceland. (a) Initial 207Pb/204Pb against 206Pb/204Pb. (b) Initial 208Pb/204Pb against 206Pb/204Pb. Error bars are 2 for repeat analyses of BCR-1. Data sources for MORB and Iceland are given in Fig. 8.

 

View larger version (19K): [in this window] [in a new window]   Fig. 10. Initial 208Pb/204Pb against 207Pb/204Pb for dykes, North Atlantic MORB and Iceland. Error bars are 2 for repeat analyses of BCR-1. Data sources for MORB and Iceland are given in Fig. 8.

 

View larger version (22K): [in this window] [in a new window]   Fig. 11. Pb, Sr and Nd isotopic compositions for dykes, North Atlantic MORB and Iceland. (a) Initial 87Sr/86Sr against 206Pb/204Pb. (b) Initial Nd against 206Pb/204Pb. Error bars (2) for repeat analyses of BCR-1 are smaller than symbol size. Data sources for MORB and Iceland are given in Fig. 8.

 
The TS and PAS groups have _Nd ranging from +4·96 to +8·46, which is almost entirely overlapping with present-day values from Iceland (Fig. 8). ⁸⁷Sr/⁸⁶Sr ranges from 0·70328 to 0·70442, with the exception of one sample (436707) with ⁸⁷Sr/⁸⁶Sr 0·7071. Most samples thus have slightly higher ⁸⁷Sr/⁸⁶Sr than present-day Iceland basalts, although the PAS dykes and some TS dykes from Tasiilaq do overlap with Iceland compositions (Fig. 8). The low Zr/Nb TS dykes tend to have lower _Nd than the high Zr/Nb TS dykes.

The TRANS dykes show a range in ⁸⁷Sr/⁸⁶Sr from 0·70327 to 0·70676, similar to that of the TS and PAS dykes, but they have much lower _Nd values, ranging from -4·18 to + 2·54, and thus plot below the Iceland and MORB fields in Fig. 8.

Pb isotope data for the TS and PAS dykes show a wide range of compositions and extend to less radiogenic ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb values than reported for MORB and Iceland (Figs 9 and 10), and also to higher ²⁰⁸Pb/²⁰⁴Pb. ²⁰⁷Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb for the TS and PAS dykes show a well-defined trend in Fig. 9a, which projects into the Iceland and MORB fields. On a plot of ²⁰⁸Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb (Fig. 9b) some of the samples fall along a trend that is an extension of the MORB and Iceland fields, although there is considerably more scatter than in ²⁰⁷Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb. Figure 10 shows ²⁰⁸Pb/²⁰⁴Pb vs ²⁰⁷Pb/²⁰⁴Pb, and again most of the dykes fall on a well-defined trend. Dykes that define this trend tend to have slightly elevated ²⁰⁸Pb/²⁰⁴Pb for a given ²⁰⁷Pb/²⁰⁴Pb relative to Iceland.

The TRANS dykes extend to the least radiogenic Pb isotopic compositions measured for the East Greenland dykes (²⁰⁶Pb/²⁰⁴Pb = 15·940–17·411; ²⁰⁷Pb/²⁰⁴Pb = 15·061–15·280; ²⁰⁸Pb/²⁰⁴Pb = 36·076–37·890), and do not overlap with either the Iceland or MORB compositional fields. In terms of ²⁰⁷Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb (Fig. 9a) the transitional dykes fall along, and extend, the trend of the tholeiitic dykes. In Fig. 9b (²⁰⁸Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb) and in Fig. 10 (²⁰⁸Pb/²⁰⁴Pb vs ²⁰⁷Pb/²⁰⁴Pb), the four TRANS samples from Tasiilaq plot along the same trend as the PAS dykes and most of the TS dykes, whereas the two TRANS samples from Fladø and I. C. Jacobsen Fjord plot at higher ²⁰⁸Pb/²⁰⁴Pb.

⁸⁷Sr/⁸⁶Sr and _Nd are plotted against ²⁰⁶Pb/²⁰⁴Pb in Fig. 11. There is no correlation between ²⁰⁶Pb/²⁰⁴Pb and ⁸⁷Sr/⁸⁶Sr for the dykes (Fig. 11a); however, there is a weak negative correlation between _Nd and ²⁰⁶Pb/²⁰⁴Pb for the high Zr/Nb TS dykes (Fig. 11b). In both plots the TS and PAS dykes plot in an array trending towards the radiogenic end-member in Iceland (in terms of Pb isotopes) as defined by Torfajökull (Stecher et al., 1999).

	DISCUSSION

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
Isotopic signature of the PAS and TS dykes
The range in Pb isotopic compositions of the TS and PAS dykes requires that at least three components were involved in their generation: a component with high ²⁰⁶Pb/²⁰⁴Pb; a component with low ²⁰⁶Pb/²⁰⁴Pb and high ²⁰⁸Pb/²⁰⁴Pb; and a component with low ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb.

The high ²⁰⁶Pb/²⁰⁴Pb component
In a diagram of ²⁰⁸Pb/²⁰⁴Pb against ²⁰⁷Pb/²⁰⁴Pb (Fig. 10) most TS and PAS dyke samples fall on a line from the least radiogenic of the dykes towards the most radiogenic end of the Iceland array represented by Torfajökull. That the dykes trend towards this composition, and not the low ²⁰⁶Pb/²⁰⁴Pb end of the Iceland array or MORB, is seen especially clearly in Fig. 11b (_Nd vs ²⁰⁶Pb/²⁰⁴Pb). The TS and PAS dykes furthermore resemble OIB compositions in that they are enriched in incompatible elements, and in a diagram of Zr/Y against Nb/Y ratios (Fig. 7) most of the dykes plot together with enriched Icelandic basalts. Thus both the isotopic and trace element characteristics of the TS and PAS dykes are consistent with the involvement of the enriched Iceland plume in their petrogenesis.

The low ²⁰⁶Pb/ ²⁰⁴Pb components
There are no North Atlantic MORBs with Pb isotopic compositions as unradiogenic as those found in the dykes, and from Figs 8–11 it is apparent that mixing of MORB melts with an Icelandic composition (or a relatively radiogenic dyke composition) cannot explain the observed trends for the dykes. Unradiogenic ²⁰⁶Pb/²⁰⁴Pb signatures in flood basalt provinces are usually attributed to either crustal contamination or a contribution from the subcontinental mantle (e.g. Dickin, 1981; Mahoney et al., 1992). Because of the wide range of ²⁰⁸Pb/²⁰⁴Pb in both crust and lithospheric mantle, either may account for both the low ²⁰⁸Pb/²⁰⁴Pb and high ²⁰⁸Pb/²⁰⁴Pb components.

Crustal contamination
Contamination with upper and lower crust has been argued to be the most plausible explanation for the variation in the isotopic compositions of the Lower Basalts in the East Greenland flood basalt sequence (Fram & Lesher, 1997; Hansen & Nielsen, 1999) and in the continental succession of ODP Legs 152 and 163 (Fitton et al., 1998a; Saunders et al., 1999). In the case of upper-crustal (amphibolite-facies gneiss) contamination as envisaged for some of the Lower Basalts (Hansen & Nielsen, 1999) and the later part of the offshore ODP Leg 152 continental succession (Fitton et al., 1997, 1998a), there is a correlation between SiO₂ and several incompatible element ratios and isotopic compositions, most notably between SiO₂ and ⁸⁷Sr/⁸⁶Sr. For those lavas from the Lower Basalts that have relatively low ⁸⁷Sr/⁸⁶Sr ratios, Fram & Lesher (1997) and Hansen & Nielsen (1999) proposed a lower-crustal silicic contaminant similar to Lewisian and East Greenland granulite-facies gneiss. The lavas of the lower part of the continental series of ODP Leg 152 are also interpreted as contaminated with lower crust, although a mafic rather than silicic contaminant is inferred because of a lack of correlation between SiO₂ and isotopic compositions (Fitton et al., 1998a).

The trace element characteristics for most of the continental sequence of ODP Legs 152 and 163 and for the Lower Basalts are consistent with crustal contamination. For example, they show relative depletion in Nb and Ta, giving rise to relatively high La/Nb ratios (Fitton et al., 1998b; Hansen, 1997).

The East Greenland dykes are in some respects similar in isotopic compositions to many of the contaminated lavas of the Lower Basalts and ODP Legs 152 and 163. For example, they are generally characterized by low ²⁰⁶Pb/²⁰⁴Pb compared with Iceland and North Atlantic MORB (Fig. 12). There are, however, several notable differences indicating that the source of the unradiogenic Pb in the TS and PAS dykes is different from that for the lavas of the Lower Basalts and ODP Leg 152. First, the dykes lack a positive correlation of _Nd with ²⁰⁶Pb/²⁰⁴Pb exhibited by lavas from the Lower Basalts and the continental succession of ODP Leg 152 (Fig. 12c). Such trends are expected from contamination with Archaean or Proterozoic crust (or EM-1 type lithospheric mantle). Furthermore, the TS and PAS dykes lack any significant correlation of isotopic ratios with either SiO₂ or trace element ratios indicative of crustal contamination, such as La/Nb and Nb/U (Fig. 13).

View larger version (22K): [in this window] [in a new window]   Fig. 12. Isotopic compositions of dykes of this study compared with East Greenland flood basalts, SDRS from ODP Leg 152, Iceland, North Atlantic MORB, East Greenland crustal lithologies, West Greenland lamproites, and average Lewisian granulite- and amphibolite-facies gneisses. (a) 207Pb/204Pb against 206Pb/204Pb. (b) 208Pb/204Pb against 207Pb/204Pb. (c) Nd against 206Pb/204Pb. Data sources for MORB and Iceland as Fig. 8. Data for flood basalts from Hansen & Nielsen (1999) and Peate & Stecher (2003). Data for ODP Leg 152 from Fitton et al. (1998a). Basement data from Leeman et al. (1976), Dickin (1981), Taylor et al. (1992) and Kalsbeek et al. (1993). Data for West Greenland lamproites from Nelson (1989).

 

View larger version (29K): [in this window] [in a new window]   Fig. 13. (a) and (b) 87Sr/86Sr against SiO2 and La/Nb; (c) and (d) Nb/U and La/Nb against 206Pb/204Pb; (e) and (f) La/Nb against SiO2 and Nd.

 
If the low ²⁰⁶Pb/²⁰⁴Pb compositions of the dykes are the result of crustal contamination, then the contaminant must have relatively high _Nd (or low Nd/Pb) and incompatible element ratios similar to the uncontaminated magma. The granulite- and amphibolite- facies gneisses described from East Greenland are characterized by low _Nd, and ‘normal’ crustal trace element abundances and ratios (e.g. Nb and Ta depletion), and can therefore not account for the isotopic and trace element characteristics of the TS and PAS dykes by contamination of Iceland plume derived melts.

This point is illustrated in Fig. 14, which shows mixing hyperbolae for ¹⁴³Nd/¹⁴⁴Nd and ²⁰⁶Pb/²⁰⁴Pb compositions resulting from the mixing of the TS dyke sample with the highest ²⁰⁶Pb/²⁰⁴Pb (sample 410660) with a low ²⁰⁶Pb/²⁰⁴Pb granulite-facies gneiss [average Lewisian of Dickin (1981)]. The mixing lines show that although contamination with granulite-facies gneiss can account for the observed variation of the continental succession of ODP Leg 152, such a contaminant with low _Nd cannot easily explain the isotopic compositions of the TS and PAS dykes. If a primitive Icelandic basalt is used instead of the dyke composition in the mixing calculations, much less contamination is needed to shift the isotopic composition, and the bulk mixing hyperbola intersects the dyke data at higher ¹⁴³Nd/¹⁴⁴Nd and lower ²⁰⁶Pb/²⁰⁴Pb. However, bulk mixing still fails to account for the samples with the lowest ²⁰⁶Pb/²⁰⁴Pb, and in this case, also the samples with the highest ²⁰⁶Pb/²⁰⁴Pb. Selective contamination (an effective higher Pb/Nd in the low ²⁰⁶Pb/²⁰⁴Pb end-member) would give a better fit for dyke samples with low ²⁰⁶Pb/²⁰⁴Pb, but would give a worse fit for high ²⁰⁶Pb/²⁰⁴Pb dyke samples (see Fig. 14 for details).

View larger version (22K): [in this window] [in a new window]   Fig. 14. Diagram of 143Nd/144Nd and 206Pb/204Pb with results of mixing calculations. Fine black lines are mixing hyperbolae for (1) bulk mixing of average Lewisian granulite-facies gneiss (Dickin, 1981) with the most enriched (in terms of Pb isotopic composition) of the dyke samples, 410660, (2) bulk mixing of average Lewisian granulite-facies gneiss with a primitive Icelandic basalt (Breddam, 2002), (3) bulk mixing of Lewisian mafic crust (Cohen et al., 1991) with dyke sample 410660, and (4) bulk mixing of Lewisian mafic crust with primitive Icelandic basalt. The compositions of the end-members are given in Table 3. Tick marks indicate amount of contamination in 10% increments. The dotted envelope represents permissible bulk mixing hyperbolae (calculated by a Monte Carlo simulation) for mixing average Lewisian granulite-facies gneiss and 410660, where both end-members are allowed to vary as shown by the white boxes (crust: 143Nd/144Nd = 0·5106–0·5110 and 206Pb/204Pb = 13·5–14·5; dyke: 143Nd/144Nd = 0·5129–0·5130 and 206Pb/204Pb = 18·0–18·5). Dashed lines show selective contamination of 410660 with 15 and 25% average Lewisian crust. The lines are calculated hybridization paths assuming that Pb diffusivity is an order of magnitude greater than Nd diffusivity (Lesher, 1990). The upper trajectory represents the geochemical evolution of the basaltic end-member as hybridization proceeds, and the lower trajectory is the complementary path for the gneissic end-member. Chemical equilibrium is in essence simple binary mixing represented by the intercept with the bulk mixing curve. Arrow shows direction to melts from depleted mantle (Table 3).

 

View this table: [in this window] [in a new window]   Table 3: Compositions used in mixing calculations

 
Mafic lower crust described from the Lewisian has a wide range of isotopic compositions (Cohen et al., 1991). Of the samples described by Cohen et al. (1991) two mafic gneisses have isotopic compositions characterized by low ⁸⁷Sr/⁸⁶Sr and ²⁰⁶Pb/²⁰⁴Pb and high _Nd. Addition of this type of lower crust could account for the observed trend of the dykes. However, this requires more than 80% addition of crust to 410660 to produce the composition of the least radiogenic TS dykes assuming bulk mixing (Fig. 14). If the primitive Icelandic composition is used as the high ²⁰⁶Pb/²⁰⁴Pb end-member, up to 55% bulk contamination is required to reproduce the dyke data.

Because of the heterogeneity of crustal lithologies, crustal contamination of the TS and PAS dykes as an explanation for the observed isotopic variations is possible. It seems unlikely, however, that contamination with silicic upper or lower crust, as represented by amphibolite- and granulite-facies gneisses, is significant. Contamination with relatively depleted mafic crust offers a better explanation because it has the appropriate isotopic composition and lacks the incompatible element signatures characteristic of salic crust. The degree of contamination required, however, to explain the isotopic variation in the dykes is relatively high.

Lithospheric mantle contribution
An alternative hypothesis to crustal contamination is that the source of the low ²⁰⁶Pb/²⁰⁴Pb is the continental lithospheric mantle.

Intuitively this is an attractive model, because depletion is readily understood in mantle rocks. Although part of the subcontinental lithospheric mantle is isotopically enriched, samples of depleted continental lithospheric mantle are fairly common. Xenoliths from East Greenland (Bernstein et al., 1998a; Hanghøj et al., 2001) may be representative of this kind of mantle, and recent Nd isotopic analyses of pyroxene separates from these xenoliths show that their ¹⁴³Nd/¹⁴⁴Nd ratios vary from 0·5122 to 0·5135 (K. Hanghøj, unpublished data, 2001). A model involving a depleted mantle in the petrogenesis is also attractive in terms of mass balance, because small degrees of partial melting produce melts with relatively high abundances of incompatible elements from the depleted mantle, even if the absolute abundances of incompatible elements are low in the mantle rocks.

An example of how depletion of a primitive mantle source by melting controls the subsequent radiogenic isotope characteristics is shown in Table 3. The results indicate that residual mantle rocks with _Nd > +40 can be generated by 15% melting of a primitive mantle 2000 Myr ago, and that the ²⁰⁶Pb/²⁰⁴Pb and ⁸⁷Sr/⁸⁶Sr of the residue will be 16 and 0·7015, respectively. Figure 14 shows the direction of the mixing hyperbola for simple binary mixing of small-degree melts (0·5% batch melting) of this restite with sample 410660. As can be seen from Fig. 14, the depleted mantle has too high _Nd and/or too high ²⁰⁶Pb/²⁰⁴Pb to be a suitable end-member for the dykes. The calculation in Table 3 takes into account the degree of depletion of the primitive mantle, degree of melting of the restite, age of depletion and trace element partition coefficients. Changing these parameters will change the composition of the model low ²⁰⁶Pb/²⁰⁴Pb end-member, but will not significantly affect the coupling of the isotopic systems during mantle melting. If fractional melting is assumed (not shown) the result is a larger degree of depletion of the restite, which is more pronounced for ⁸⁷Sr/⁸⁶Sr and ²⁰⁶Pb/²⁰⁴Pb than for ¹⁴³Nd/¹⁴⁴Nd, and the composition is still too depleted in Nd (and Sr) relative to Pb to be a suitable low ²⁰⁶Pb/²⁰⁴Pb end-member.

Therefore, straightforward depletion of mantle material cannot provide a low ²⁰⁶Pb/²⁰⁴Pb component consistent with the isotopic range of the TS and PAS dykes. Thus, if the low ²⁰⁶Pb/²⁰⁴Pb component is lithospheric mantle, this mantle would have to be modified by metasomatic processes.

Isotopic signature of the transitional dykes
The TRANS dykes have distinctly different isotopic compositions from Icelandic basalts, with relatively unradiogenic lead and neodymium. As for the early TS and PAS dykes, the Pb isotopic compositions of the TRANS dykes plot along a trend indicating the presence of a component with high ²⁰⁶Pb/²⁰⁴Pb and a component with low ²⁰⁶Pb/²⁰⁴Pb and variable ²⁰⁸Pb/²⁰⁴Pb. The low ¹⁴³Nd/¹⁴⁴Nd ratios for the transitional dykes and the lack of correlation between Nd and Pb isotopic compositions (Fig. 11a) imply that the low ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb component is different from that of the TS and PAS dykes. The high ²⁰⁶Pb/²⁰⁴Pb component appears to be similar to the high ²⁰⁶Pb/²⁰⁴Pb component of the TS and PAS dykes, and is assumed to be the Icelandic plume.

The origin of the low ²⁰⁶Pb/²⁰⁴Pb component in the Transitional Series
The main differences between the TRANS and TS dykes are a more pronounced enrichment in incompatible elements and the less radiogenic Nd isotope compositions in the former. The TRANS dykes also extend to slightly lower ²⁰⁶Pb/²⁰⁴Pb.

Figure 13 show variations in silica concentration, Sr isotopic composition and La/Nb and Nb/U ratios. From Fig. 13e, it is evident that the TRANS dykes with the highest SiO₂ contents also have the highest La/Nb indicative of crustal contamination. The TRANS dykes selected for isotope analysis, however, were all selected because of their low La/Nb (0·76–1·04) similar to mantle values (e.g. Sun & McDonough, 1989), and Fig. 13d shows that for these samples, there is no correlation between La/Nb and ²⁰⁶Pb/²⁰⁴Pb. There is, however, a weak negative correlation of La/Nb and _Nd for the same samples (Fig. 13f). Nb/U has been suggested as another ratio sensitive to crustal contamination by Hofmann (1997), who argued that mantle values for Nb/U are >37 and that this ratio decreases with contamination by continental crust. Nb/U values for the dykes are shown versus ²⁰⁶Pb/²⁰⁴Pb in Fig. 13c and, as for La/Nb, there is no correlation with Pb isotopic composition.

If La/Nb is a sensitive indicator of crustal contamination as suggested by Thompson et al. (1983), some of the transitional dykes appear to be generally more affected by crustal contamination than the TS and PAS dykes. Figure 15 shows La/Nb against _Nd when mixing a TS dyke with a crustal contaminant with _Nd of -35 (average Lewisian) and La/Nb of six [average Greenland crust of Wedepohl et al. (1991)]. The mixing hyperbola (assuming bulk mixing) shows that 10–25% assimilation can account for the variation in _Nd and that La/Nb does not change significantly. Similar calculations for Pb (not shown) give a similar result, i.e. the low La/Nb values of the TRANS dykes analysed for isotopes are not inconsistent with contamination with continental crust. Also the Pb and Nd isotopic compositions can be modelled as 10–20% bulk assimilation of continental crust (Fig. 14) where a TS dyke constitutes the isotopically enriched end-member. One preliminary Os isotopic analysis of TRANS dyke 417390 gives ¹⁸⁶Os/¹⁸⁷Os >1 (K. Hanghøj, unpublished data, 2001), which is also consistent with crustal contamination.

View larger version (16K): [in this window] [in a new window]   Fig. 15. Diagram of Nd against La/Nb ratios for TRANS dykes. TS and PAS dykes are shown for comparison. Hyperbola shows results of calculations for mixing a TS dyke (410660) and a crustal contaminant with La/Nb = 6 [average Greenland crust from Wedepohl et al. (1991)] and Nd = -35 [average Lewisian gneiss from Dickin (1981)].

 
The origin of the low ²⁰⁶Pb/²⁰⁴Pb component in the present-day Iceland plume
Several models have been put forward to explain the isotope characteristics of Iceland basalts. Hart et al. (1973) suggested mixing of a plume component and MORB to explain the variation in isotope compositions in Iceland and along the Reykjanes Ridge. Thirlwall (1995), however, pointed out that the Pb isotopic compositions are inconsistent with simple mixing of a plume component and MORB, because binary mixing on diagrams of ²⁰⁷Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb vs ²⁰⁷Pb/²⁰⁴Pb will give rise to linear data arrays between the end-members. He proposed that the offset of Iceland data relative to MORB on these diagrams (see Figs 9 and 10) is due to a relatively recent increase in U/Pb in the Iceland plume source, and ternary mixing of this—possibly somewhat heterogeneous—plume source with MORB source and a depleted mantle source different from MORB mantle. Kerr et al. (1995) raised similar objections to the model of simple mixing between a plume and a MORB source, but interpreted the compositional range of Iceland samples as an inherent feature of the plume. Recently, Hanan & Schilling (1997) and Hanan et al. (2000) have proposed that ternary mixing of plume mantle, depleted MORB mantle and EM1 type mantle may account for the spatial and temporal geochemical variation in Iceland. Fitton et al. (1997) and Kempton et al. (2000) instead proposed a model that involves four geochemical reservoirs, an enriched Iceland plume source, a depleted Iceland plume source, a depleted sheath surrounding the plume and finally shallow MORB mantle. Chauvel & Hémond (2000) suggested, on the basis of trace element chemistry and Pb, Sr and Nd isotope compositions, that the Iceland plume is composed entirely of (heterogeneous) recycled Archaean oceanic lithosphere, where melting of the basaltic portion (+ harzburgite) gives rise to alkali basalts, and melting of the gabbroic portion (+ harzburgite) gives rise to picrites. Skovgaard et al. (2001) found that Os and O isotopic compositions of Icelandic picrites are consistent with a contribution from recycled oceanic lithosphere.

MORB mantle does not appear to contribute significantly to the East Greenland dyke swarm. Instead, the dykes record binary mixing between a lithospheric component and a plume component as discussed above. In Figs 11 and 14, the dyke data trend towards the enriched rather than the depleted portion of the Iceland array, and therefore do not provide evidence for both an enriched and a depleted plume source. However, this may be due to lack of sampling of such a depleted plume component. Because of the relatively thick lithospheric lid during rifting, depleted portions of the plume would melt less, or not at all, leading to magmas dominated by the more fusible enriched plume component. Alternatively, there is no depleted component inherent to the plume, and the most depleted of the Iceland samples may be products of ternary mixing between an enriched plume source, MORB source and a relatively depleted source of old continental lithospheric mantle, mobilized by thermal erosion during break-up. A mechanism for the incorporation of lithospheric mantle into the shallow asthenosphere could be transient heating at the arrival of the Iceland hotspot under Greenland (e.g. Larsen et al., 1996b). Plume-related delamination of continental lithospheric mantle has also been proposed to explain the anomalous trace element and isotopic compositions at the 39°–41° segment of the South West Indian Ridge (Mahoney et al., 1992), for segments of the Shona and Discovery ridges (Douglass et al., 1999), and for the Kerguelen Plateau (Storey et al., 1992).

	MANTLE MELTING CONDITIONS DURING CONTINENTAL RIFTING

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
For simple decompression melting, the depth and degree of melting significantly influences the composition of the mantle-derived liquids. This is in turn dependent on the potential temperature of the mantle, which determines at which depth the solidus is intersected, and on the thickness of the lithosphere, which controls at which depth melting ceases. Quantitatively addressing the melting dynamics in the Greenland Tertiary provides constraints on the temporal and spatial variation of the lithospheric thickness and the thermal structure of the mantle.

Fram & Lesher (1993) showed that there is a general positive correlation of Dy/Yb(N) (N denotes normalization to chondrite) and fractionation-corrected TiO₂ for the North Atlantic province. They explained this in terms of progressive lithospheric thinning where the earliest melts [Lower Basalts with high Dy/Yb(N)] were generated under thick lithosphere, and modern ocean ridge basalts [Iceland and adjacent ridges with low Dy/Yb(N)] under thin lithosphere. Subsequently, trace element data on sections though the SDRS drilled during ODP Leg 152 (Fram et al., 1998) and the East Greenland Plateau Basalts (Tegner et al., 1998b) have been used to refine mantle melting models, offering high temporal resolution for the transition between continental and oceanic magmatism.

For example, Fram et al. (1998), using La/Sm(N) and Lu/Hf ratios in the SDRS, suggested that the lowest units are generated by 4–5% melting of a depleted mantle source at mean pressures greater than 2 GPa, whereas the top units are consistent with 10–12% melting at 2·5–1 GPa. The data are interpreted as the lithospheric lid thinning from 60 km to less than 25 km in less than 5 Myr.

Tegner et al. (1998b) showed that Fe- and Ti-rich lavas from the lower and upper portions of the East Greenland Plateau Basalts plot in two distinct groups in terms of La/Sm(N) and Dy/Yb(N). In the lower part of the succession, La/Sm(N) increases regularly up-section, whereas Dy/Yb(N) decreases. In the upper part of the succession La/Sm(N) is much more variable, and shows a positive correlation with Dy/Yb(N). This is interpreted in terms of falling potential temperature during the eruption of the lower portion, and as variations in the dynamics of melt segregation and in lithospheric thickness in the upper portion. The regular decline in mean pressure of melting coupled with an increase in degree of melting (thinned lid effect) observed in the SDRS off the SE Greenland margin is not observed in the Plateau Basalts. The different REE characteristics observed in the main groups of dykes can likewise be understood in terms of variations in depth and degree of melting. In the following section, the dykes are discussed in the context of a specific mantle melting model similar to that of Tegner et al. (1998b).

Mantle melting modelling
The mantle melting model presented in Fig. 16 is made using the computer algorithm REEBOX first described by Fram & Lesher (1993) and subsequently modified by Fram et al. (1998) and Tegner et al. (1998b). REEBOX simulates dynamic decompression melting assuming a triangular melting regime defined by simple corner flow.

View larger version (50K): [in this window] [in a new window]   Fig. 16. (a) Model curves for pooled mantle melts in terms of La/Sm(N) and Dy/Yb(N). (N) denotes normalization to chondrite. Bold continuous curves show composition of pooled melts for different starting pressures Psolidus. Thin continuous lines contour segregation pressures Psegregation, and dashed curves give mean extent of melting F. Arrows show effects of (1) source depletion, (2) lithospheric thinning or incomplete pooling, and (3) fall in potential temperature. Shaded areas show compositions of Plateau Basalts from Tegner et al. (1998b). (b)–(f) give dyke compositions with superimposed melting model from (a). [Note change of scale in (f).]

 
The source mantle used in the calculation is depleted relative to primitive mantle (Sun & McDonough, 1989) by 0·5% batch melting of a spinel lherzolite. Incremental non-batch melting over a range of pressures and a uniform melt productivity of 1% per 0·1 GPa is assumed. Partitioning coefficients and reaction coefficients are from Baker & Stolper (1994), Green (1994) and Lesher & Baker (1997). The initial mineral mode is 50% olivine, 25% orthopyroxene, 15% clinopyroxene and 10% garnet. For the garnet–spinel transition, reaction coefficients are assumed to vary linearly between 3 and 2·5 GPa, and the spinel–plagioclase transition is likewise assumed to be linear between 1·4 and 1 GPa. The composition of the residue is recalculated after each 0·1 GPa increment of decompression melting and adjusted for pressure-dependent phase transitions (see Fram & Lesher, 1997). Results are shown in Fig. 16 in terms of La/Sm(N) and Dy/Yb(N) ratios.

Because La is more incompatible than Sm during mantle melting, an increase in La/Sm(N) corresponds to a decrease in the mean extent of melting (F). Yb is compatible in garnet relative to Dy, and an increase in Dy/Yb(N) reflects an increase in the proportion of melt derived from a garnet-bearing source, i.e. an increase in mean pressure (P) (Fig. 16a). As discussed by Tegner et al. (1998b), a positive correlation between F and P can be understood in terms of lithospheric control, whereas a negative correlation indicates changes in mantle temperature. A relative enrichment of the source mantle (e.g. primitive mantle) would shift the curves to higher La/Sm(N) and Dy/Yb(N). This implies that progressive depletion of the mantle source also will lead to a positive correlation between La/Sm(N) and Dy/Yb(N). This is seen by the slope of the individual melting curves, which reflects progressive depletion as mantle moves upwards through the melting column.

Results
High Zr/Nb Tholeiitic Series
All TS dykes are shown in Fig. 16b. The high Zr/Nb TS dykes have REE characteristics somewhat similar to the flood basalts (Tegner et al. 1998b), consistent with starting pressures of 3·1–2·6 GPa, segregation pressures of 2·5–1·9 GPa and 2–10% melting. In Fig. 16c–e the high Zr/Nb TS dykes are shown by filled symbols by locality. There is a weak negative correlation between La/Sm(N) and Dy/Yb(N) for Tasiilaq and Langø. Although Tasiilaq shows the greatest range in composition, there is a substantial overlap between all four localities in terms of La/Sm(N). In terms of Dy/Yb(N), however, the localities are offset relative to each other. High Zr/Nb TS dykes from Langø and Fladø have lower Dy/Yb(N) than those from Tasiilaq and I. C. Jacobsen Fjord. Interestingly, within the high Zr/Nb TS group, the oldest dykes in groups of crosscutting dykes have lower La/Sm(N) ratios than younger dykes (not shown). This temporal relationship is similar to what is found for the older portion of the flood basalts (Tegner et al., 1998b).

If the plume was centred under the Kangerlussuaq area or further north and if the plume was a discrete feature, one would expect that the I. C. Jacobsen Fjord and Fladø dykes would show a larger contribution from garnet-bearing mantle than dykes from areas further from the plume centre. Instead, the dyke data are consistent with elevated temperatures in a large region (including all dyke-sampling localities), with local fluctuations perhaps indicating that along-axis variations in mantle upwelling and lithospheric thickness play an important role. Segregation pressures are similar for the various localities, although the largest range is found in Tasiilaq.

Low Zr/Nb Tholeiitic Series
The low Zr/Nb TS dykes generally plot at higher La/Sm(N) than the high Zr/Nb TS dykes, but at similar Dy/Yb(N). With the exception of a few samples, they fall along the 2·7 GPa melting curve at segregation pressures between 2·7 and 2·2 GPa and mean degrees of melting lower than 5% (Fig. 16b–e). There is no systematic change in composition with relative age for this group.

Recalling that the low Zr/Nb TS dykes are younger than the high Zr/Nb dykes, the data are thus consistent with a scenario of temporal cooling of the mantle source (relatively shallow solidus intersection) as suggested by Tegner et al. (1998b) for the Plateau Lavas. The relatively high La/Sm(N) can be explained in two ways. First, it may simply reflect higher segregation pressures and accompanying low degrees of melting, which would indicate a relatively thick lithosphere. Second, it may suggest that the source is heterogeneous, e.g. veined, and that the low Zr/Nb TS dykes represent a relatively larger proportion of the enriched component. Importantly, the second possibility can only be applicable together with lithospheric thickening, unless the mantle source has actually changed to contain an enriched component during the formation of the margin. This is because low-degree melts (high proportion of the more fusible enriched component) from a veined mantle would be diluted by melt fractions from higher in the melting column (small proportion of enriched component) if the thickness of the lithosphere permits melting at similar levels as for the high Zr/Nb TS dykes.

Near-constant solidus intersection pressures and variable segregation pressures are also inferred from LREE-enriched lavas from the upper portion of the Plateau Basalts (Tegner et al., 1998b). This is partly interpreted in terms of lithospheric thickening because of residual mantle accumulation, and partly as variation in melt segregation dynamics. The low Zr/Nb TS dykes are similar to these late-LREE enriched lavas, but appear to be even more enriched. This may be because of lower potential temperature, a slightly more enriched mantle, or a combination of both. Isotopic data for the dykes and the Plateau Basalts (Fig. 12) do not support an isotopically more enriched source for the dykes. If the difference is due to greater enrichment, it thus has a recent origin. Fram & Lesher (1997) suggested recent LREE enrichment of the mantle source for the Lower Basalts on the basis of isotopic and trace element compositions.

Picrite–Ankaramite Series
With the exception of two samples (416401 and 426547), the PAS dykes fall on a well-defined trend of slightly decreasing Dy/Yb(N) with increasing La/Sm(N) (Fig. 16b). The variation in La/Sm(N) is greater and the Dy/Yb(N) ratios are generally higher than for the high Zr/Nb TS dykes, indicating starting pressures of >3·2 GPa to 2·7 GPa, segregation pressures from 2·6 to 2·0 GPa and 2–12% melting.

Assuming that the mantle source for the PAS and TS dykes is the same, the data are in agreement with lower-degree melts segregated from greater depths than the high Zr/Nb dykes [causing the offset to higher La/Sm(N) and Dy/Yb(N)] during a phase of source mantle cooling (resulting in the trend of decreasing F and P, which is also observed for the lower portion of the Plateau Basalts). The PAS dykes are generally older than the high Zr/Nb TS dykes, and the difference in F can be due to a greater lithospheric thickness, i.e. an expression of the lid effect, or to changes in the ‘plumbing’ allowing melts from greater depths to segregate more frequently, i.e. less efficient pooling. Such a change in segregation style may be related to the onset of sea-floor spreading.

Transitional dykes
Most of the TRANS dykes fall outside the array defined by the model melting curves (Fig. 16f). This is not surprising if we recall that the transitional dykes have Nd isotopic compositions more enriched than the TS and PAS dykes, thus excluding the somewhat depleted source mantle used in these calculations as the exclusive source for their geochemical characteristics.

There is overlap in the data for the four localities. The data from Langø, Fladø and I. C. Jacobsen Fjord do not show any correlation between La/Sm(N) and Dy/Yb(N), but the data from Tasiilaq (which is also represented with most data points) show a positive correlation.

The isotopic data discussed above indicate that the isotopic characteristics of the TRANS dykes can be explained by 10–25% crustal contamination of plume melts. Figure 16f show mixing curves between the TS dyke (410660) used for mixing calculations in Fig. 14 and lower and bulk crust (Taylor & McLennan, 1995). The mixing curves show that contamination fails to produce the high Dy/Yb(N) compositions, and that more than 35% crust is needed to explain the remaining data. This amount of bulk assimilation is inconsistent with the isotopic data and implies that some other lithospheric component is needed if the asthenospheric (plume) end-member is the same as for the TS dykes. This lithospheric component could be melts from a garnet-bearing crust or lithospheric mantle, i.e. melting of the lower crust of Taylor & McLennan in Fig. 16f would produce highly LREE-enriched and HREE-depleted liquids. Alternatively, bulk assimilation of crust with higher La/Sm(N) and Dy/Yb(N) than the crustal examples in Fig. 16f could explain the data.

Implications for mantle melting and lithospheric control
Seismic transects show that offshore crustal thickness decreases away from the proposed plume track at the time where sea-floor spreading was initiated (before Chron 24r) (Dahl-Jensen et al., 1997; Holbrook et al., 2001). If the mantle melts by passive upwelling this will simply be an expression of a thermal gradient along the margin. The modelling of the dykes indicates that there was no systematic thermal zonation of the mantle source (plume), i.e. there is no evidence for increased degrees of melting close to the proposed plume centre relative to the more distal parts. This may indicate that the mantle upwelling is not passive, and that melt productivity is determined by active upwelling as well as mantle temperature. Active upwelling can also explain how low to moderate degrees of melting, equivalent to those seen for MORBs, can lead to the anomalously thick crust of the SDRS found offshore (Fram et al., 1998). If active upwelling is assumed under the East Greenland margin, the degree of melting calculated for the dykes would have no bearing on melt productivity, which may increase northwards in a systematic fashion as observed offshore.

When the dykes are compared with the flood basalts, they do extend to lower Dy/Yb(N), which at constant lithospheric thickness (and the same source) would be indicative of lower potential temperature. Recalling the isotopic composition of the dykes and the main part of the flood basalts (Fig. 12), it is evident that the low ²⁰⁶Pb/²⁰⁴Pb component present in the dykes is either lacking or sampled less in the flood basalts. So, although there is no systematic change in apparent temperature and source composition within the dyke swarm (i.e. towards the plume centre), there does appear to be a difference between the dykes and the flood basalts. This difference must be related to source characteristics (because of the difference in isotopic composition), and may or may not reflect differences in melting dynamics. The source difference may be related to the fact that whereas the dyke swarm intruded basement of Archaean and Proterozoic age, the main part of the flood basalts may overlie basement of Caledonian age, which may lack the depleted lithospheric mantle component envisaged as the source of the low ²⁰⁶Pb/²⁰⁴Pb component in the dykes.

The quantitative modelling of REE data for the dykes supports the conclusion of Tegner et al. (1998b) that the mantle source along the rifted margin cooled during flood basalt volcanism. The temporal change from the progressive cooling trend observed in the high Zr/Nb TS dykes, PAS dykes and the lower portion of the Plateau Basalts, to LREE-enriched melts generated at near-constant P_solidus (the low Zr/Nb TS dykes and late upper portion of the Plateau Basalts) is consistent with smaller-degree melts from a possibly heterogeneous mantle segregating at more variable pressures. This indicates increased lithospheric thickness, possibly caused by accumulation of residual mantle, and variability in melt segregation dynamics.

The data for the TRANS dykes are ambiguous and difficult to compare directly with the TS dykes. This is because an additional source or contaminant is required in the genesis of the TRANS dykes as shown by the isotope data, and the calculations of F and P depend on which additional component is chosen.

	SUMMARY AND CONCLUSIONS

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
Dykes of the East Greenland coastal dyke swarm can be divided into three main groups: pre- and syn-break-up tholeiitic picrite–ankaramite dykes (PAS); syn-break-up tholeiitic dykes (TS dykes); post-break-up, transitional-to-alkaline dykes (TRANS dykes).

Pre- and syn-break-up dykes
Of the early dykes, the most abundant group is the TS dykes. This group consists of moderately LREE-enriched tholeiites, with major element compositions similar to the East Greenland flood basalts. This group can be further subdivided into a high Zr/Nb group and a low Zr/Nb group, where the latter is generally more enriched in terms of incompatible elements and Sr–Nd–Pb isotopic compositions. The PAS dykes are much less abundant than the TS dykes. They are equivalent to some of the lavas in the Lower Basalts and have incompatible element characteristics that preclude them as parental to the TS dykes. Isotopic compositions for the TS and PAS dykes partly overlap with those for Iceland, but Pb isotopic compositions extend to less radiogenic values than those seen in either Iceland or North Atlantic MORB. The isotopically depleted source required to account for the isotopic variation of the early tholeiitic dykes is interpreted as subcontinental lithospheric mantle with low ⁸⁷Sr/⁸⁶Sr and ²⁰⁶Pb/²⁰⁴Pb and high _Nd. The different incompatible trace element ratios of the groups are interpreted to represent different degrees and depths of melting within the proto-Icelandic plume.

Comparison between dykes from different segments of the East Greenland margin does not show a systematic compositional change with distance from the presumed proto-Icelandic plume centre, indicating that the systematic changes in crustal thickness offshore may be attributed to active upwelling. Subtle geochemical differences within the dyke swarm can probably be attributed to variations in melting regimes and possibly geochemical heterogeneity of the lithosphere.

Post-break-up dykes
The post-break-up TRANS dykes have more alkaline compositions and are enriched in LREE relative to most of the early tholeiitic dykes. The TRANS dykes are isotopically distinct from Iceland and MORB, both in terms of ²⁰⁶Pb/²⁰⁴Pb ratios (15·9–17·4) and _Nd (–4·18 to +2·54). The isotopic characteristics are interpreted in terms of contamination of Iceland plume melts with continental crust.

The dyke swarm in the context of the North Atlantic Igneous Province
Pre- and syn-break-up dykes from all localities appear to have the enriched (relative to MORB) proto-Icelandic plume as the most significant source component, and there is no evidence of a contribution from a MORB mantle source, or a depleted component inherent to the plume. Instead, the data suggest contribution from an isotopically depleted lithospheric source. In terms of models for the formation of volcanic rifted margins, the data presented here thus lend no support to models that involve substantial melt generation in reservoirs other than enriched plume mantle, e.g. so-called non-plume models, where the primary source would be expected to be MORB mantle (e.g. Mutter et al., 1988; King & Anderson, 1995). Models in which the continental lithospheric mantle is the main source region for flood basalt volcanism (Gallagher & Hawkesworth, 1992) are likewise not supported by the data.

Quantitative modelling of REE data show that the PAS and high Zr/Nb TS dykes can be generated by moderate to low degrees of melting of a slightly depleted mantle at decreasing mean pressures of melting, consistent with a declining temperature during flood basalt volcanism. The low Zr/Nb TS dykes represents lower degrees of melting at near-constant temperatures but at variable segregation pressures, reflecting local variations in segregation dynamics and lithospheric control. REE systematics of the dykes cannot be explained by a systematic rise in temperature towards the plume centre, and instead a model for the formation of the margin involving active upwelling and local lithospheric control is preferred.

	ACKNOWLEDGEMENTS

 
Kent Brooks and Troels Nielsen are thanked for sharing their ideas, knowledge and enthusiasm about the geology of East Greenland. Both have contributed tremendously to this study in innumerable ways. Many colleagues and guests at the Danish Lithosphere Centre provided constructive comments and dicussions; we especially wish to thank Henriette Hansen, who endured particularly numerous discussions and versions of this work, and Chip Lesher for discussions about mantle melting and more. David Christie, Andy Ungerer and Chi Meridith helped with ICP-MS analyses at COAS, Oregon State University. Andy Saunders and Richard Arculus are thanked for constructive and helpful reviews. This work was funded by the Danish National Research Foundation.

	REFERENCES

TOP ABSTRACT INTRODUCTION THE EAST GREENLAND DYKE... CLASSIFICATION AND CHRONOLOGY OF... RESULTS DISCUSSION MANTLE MELTING CONDITIONS DURING... SUMMARY AND CONCLUSIONS REFERENCES

 
Baker, M. B. & Stolper, E. M. (1994). Determining the composition of high-pressure mantle melts using diamond aggregates. Geochimica et Cosmochimica Acta 58, 2811–2827.[CrossRef][Web of Science]

Bernstein, S., Kelemen, P. B. & Brooks, C. K. (1998a). Depleted spinel harzburgite xenoliths in Tertiary dykes from East Greenland: restites from high degree melting. Earth and Planetary Science Letters 154, 221–235.[CrossRef][Web of Science]

Bernstein, S., Kelemen, P. B., Tegner, C., Kurz, M. D., Blusztajn, J. & Brooks, C. K. (1998b). Post-breakup basaltic magmatism along the East Greenland Tertiary rifted margin. Earth and Planetary Science Letters 160, 845–862.[CrossRef][Web of Science]

Breddam, K. (2002). Kistufell: primitive melt from the Iceland mantle plume. Journal of Petrology 43, 345–373.[Abstract/Free Full Text]

Bridgwater, D., Davies, F. B., Gill, R. C. O., Gorman, B. E., Myers, J. S., Pedersen, S. & Taylor, P. (1978). Precambrian and Tertiary geology between Kangerlugssuaq and Angmagssalik, East Greenland. Grønlands Geologiske Undersøgelse, Rapport 83, 1–17.

Brooks, C. K. (1973). Rifting and doming in southern East Greenland. Nature 244, 23–25.[CrossRef]

Brooks, C. K. (1976). The Fe₂O₃/FeO ratio of basalt analyses: an appeal for a standardized procedure. Bulletin, Geological Society of Denmark 25, 117–120.

Chauvel, C. & Hémond, C. (2000). Melting a complete section of recycled oceanic crust: trace element and Pb isotopic evidence from Iceland. Geochemistry, Geophysics, Geosystems 2000-02-14.

Cheatham, M. M., Sangrey, W. F. & White, W. M. (1993). Sources of error in external calibration ICP-MS analysis of geological samples and an improved non-linear drift correction procedure. Spectrochimica Acta 48B, E487–E506.[CrossRef]

Cohen, R. S. & O'Nions, R. K. (1982). The lead, neodymium and strontium isotopic structure of ocean ridge basalts. Journal of Petrology 23, 299–324.[Abstract/Free Full Text]

Cohen, A. S., O'Nions, R. K. & O'Hara, M. J. (1991). Chronology and mechanism of depletion in Lewisian granulites. Contributions to Mineralogy and Petrology 106, 142–153.[CrossRef][Web of Science]

Condomines, M., Grönvold, K., Hooker, P. J., Muehlenbachs, K., O'Nions, R. K., Oskarsson, N. & Oxburgh, E. R. (1983). Helium, oxygen, strontium and neodymium isotopic relationships in Icelandic volcanics. Earth and Planetary Science Letters 66, 125–136.[CrossRef][Web of Science]

Dahl-Jensen, T., Holbrook, W. S., Hopper, J. R., Korenaga, J., Larsen, H. C., Kelemen, P. B., Detrick, R. & Kent, G. (1997). Structure of the upper crust at the SE Greenland volcanic rifted margin out to Chron 21 times. EOS Transactions, American Geophysical Union 78, F668.

Dickin, A. P. (1981). Isotope geochemistry of Tertiary igneous rocks from the Isle of Skye, N.W. Scotland. Journal of Petrology 22, 155–189.[Abstract/Free Full Text]

Douglass, J., Schilling, J.-G. & Fontignie, D. (1999). Plume–ridge interactions of the Discovery and Shona mantle plumes with the southern Mid-Atlantic Ridge (40°–55°S). Journal of Geophysical Research 104, 2941–2962.[CrossRef]

Dupré, B. & Allegrè, C. J. (1980). Pb–Sr–Nd isotopic correlation and the chemistry of the North Atlantic mantle. Nature 286, 17–22.[CrossRef]

Elliott, T. R., Hawkesworth, C. J. & Grönvold, K. (1991). Dynamic melting of the Iceland plume. Nature 351, 201–206.[CrossRef]

Escher, A. & Watt, W. S. (1976). Summary of the geology of Greenland. In: Escher, A. & Watt, W. S. (eds) Geology of Greenland. Copenhagen: Geological Survey of Greenland, pp. 12–15.

Fitton, J. G., Saunders, A. D., Norry, M. J., Hardarson, B. S. & Taylor, R. N. (1997). Thermal and chemical structure of the Iceland plume. Earth and Planetary Science Letters 153, 197–208.[CrossRef][Web of Science]

Fitton, J. G., Hardarson, B. S., Ellam, R. M. & Rogers G. (1998a). Sr-, Nd-, and Pb-isotopic composition of volcanic rocks from the southeast Greenland margin at 63°N: temporal variation in crustal contamination during continental breakup. In: Larsen, H. C., Saunders, A. D. & Wise, S. W., Jr (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 152. College Station, TX: Ocean Drilling Program, pp. 351–357.

Fitton, J. G., Saunders, A. D., Larsen, L. M., Hardarson, B. S. & Norry, M. J. (1998b). Volcanic rocks from the East Greenland margin at 63°N: composition, petrogenesis, and mantle sources. In: Larsen, H. C., Saunders, A. D. & Wise, S. W., Jr (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 152. College Station, TX: Ocean Drilling Program, pp. 331–350.

Fram, M. S. & Lesher, C. E. (1993). Geochemical constraints on mantle melting during creation of the North Atlantic basin. Nature 363, 712–715.[CrossRef]

Fram, M. S. & Lesher, C. E. (1997). Generation and polybaric differentiation of magmas of the East Greenland Early Tertiary flood basalts. Journal of Petrology 38, 231–275.[CrossRef][Web of Science]

Fram, M. S., Lesher, C. E. & Volpe, A. M. (1998). Mantle melting systematics: the transition from continental to oceanic volcanism on the southeast Greenland margin. In: Larsen, H. C., Saunders, A. D. & Wise, S. W., Jr (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 152. College Station, TX: Ocean Drilling Program, pp. 373–386.

Frey, F. A., Walker, N., Stakes, D., Hart, S. R. & Nielsen, R. (1993). Geochemical characteristics of basaltic glasses from the AMAR and FAMOUS axial valleys, Mid-Atlantic Ridge (36°–37°N): petrogenetic implications. Earth and Planetary Science Letters 115, 117–136.[CrossRef][Web of Science]

Gallagher, K. & Hawkesworth, C. J. (1992). Dehydration melting and the generation of continental flood basalts. Nature 358, 57–59.[CrossRef]

Gill, R. C. O., Nielsen, T. F. D., Brooks, C. K. & Ingram, G. A. (1988). Tertiary volcanism in the Kangerlugssuaq region, E. Greenland: trace-element geochemistry of the Lower Basalts and tholeiitic dyke swarms. In: Morton, A. C. & Parson, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publications 39, 161–179.

Govindaraju, K. (1989). Compilation of working values and sample descriptions for 272 geostandards. Geostandards Newsletters 13, 1–113.

Govindaraju, K. (1994). Compilation of working values and sample descriptions for 383 geostandards. Geostandards Newsletters Special Issue 18, 1–158.

Green, T. H. (1994). Experimental studies of trace-element partitioning applicable to igneous petrogenesis—Sedona 16 years later. Chemical Geology 117, 1–36.[CrossRef][Web of Science]

Hanan, B. B. & Schilling, J.-G. (1997). The dynamic evolution of the Iceland mantle plume: the lead isotope perspective. Earth and Planetary Science Letters 151, 43–60.[CrossRef][Web of Science]

Hanan, B. B., Blichert-Toft, J., Kingsley, R. & Schilling, J.-G. (2000). Depleted Iceland mantle plume geochemical signature: artifact of multi-component mixing? Geochemistry, Geophysics, Geosystems 4, 2000-2-14.

Hanghøj, K., Kelemen, P. B., Bernstein, S., Blusztajn, J. & Frei, R. (2001). The Wiedemann Fjord mantle xenoliths, a record of cratonic mantle formation by melt depletion in the Archaean. Geochemistry, Geophysics, Geosystems 2, 2000GC000085.

Hansen, H. (1997). Studies of basalt petrogenesis in contrasting tectonic settings of the North Atlantic Basalt Province. Petrogenesis of plagioclase ultraphyric basalts from Iceland and the Lower Basalts of central East Greenland. Ph.D. thesis, University of Copenhagen.

Hansen, H. & Nielsen, T. F. D. (1999). Crustal contamination in Palaeogene East Greenland flood basalts: plumbing system evolution during continental rifting. Chemical Geology 157, 89–118.[CrossRef][Web of Science]

Hart, S. R., Schilling, J.-G. & Powell, J. L. (1973). Basalts from Iceland and along the Reykjanes Ridge: Sr isotope geochemistry. Nature 246, 104–107.

Hémond, C., Arndt, N. T., Lichtenstein, U. & Hofmann, A. W. (1993). The heterogeneous Iceland plume: Nd–Sr–O isotopes and trace element constraints. Journal of Geophysical Research 98, 15833–15850.[CrossRef]

Henriksen, N. & Higgins, A. K. (1976). East Greenland Caledonian fold belt. In: Escher, A. & Watt, S. W. (eds) Geology of Greenland. Copenhagen: Geological Survey of Greenland, pp. 182–246.

Hofmann, A. W. (1997). Mantle geochemistry: the message from oceanic volcanism. Nature 385, 219–229.[CrossRef]

Holbrook, W. S., Larsen, H. C., Korenaga, J., Dahl-Jensen, T., Reid, I. D., Kelemen, P. B., Hopper, J. R., Kent, G. M., Lizarralde, D., Bernstein, S. & Detrick, R. S. (2001). Mantle thermal structure and active upwelling during continental breakup in the North Atlantic. Earth and Planetary Science Letters 190, 251–266.[CrossRef][Web of Science]

Ito, E., White, W. M. & Göpel, C. (1987). The O, Sr, Nd and Pb isotope geochemistry of MORB. Chemical Geology 62, 157–176.[CrossRef][Web of Science]

Kalsbeek, F., Austrheim, H., Bridgwater, D., Hansen, B. T., Pedersen, S. & Taylor, P. N. (1993). Geochronology of Archaean and Proterozoic events in the Ammasalik area, South-East Greenland, and comparisons with the Lewisian of Scotland and the Nagsugtoqidian of West Greenland. Precambrian Research 62, 239–279.[CrossRef][Web of Science]

Karson, J. A., Brooks, C. K., Storey, M. & Pringle, M. S. (1998). Tertiary faulting and pseudotachylytes in the East Greenland volcanic rifted margin: seismogenic faulting during magmatic construction. Geology 26, 39–42.[Abstract/Free Full Text]

Kempton, P. D., Fitton, J. G., Saunders, A. D., Nowell, G. M., Taylor, R. N., Hardarson, B. S. & Pearson, G. (2000). The Iceland plume in space and time: a Sr–Nd–Pb–Hf study of the North Atlantic rifted margin. Earth and Planetary Science Letters 177, 255–271.[CrossRef][Web of Science]

Kerr, A. C., Saunders, A. D., Tarney, J., Berry, N. H. & Hards, V. L. (1995). Depleted mantle-plume geochemical signatures: no paradox for plume theories. Geology 23, 843–847.[Abstract/Free Full Text]

King, S. D. & Anderson, D. L. (1995). An alternative mechanism of flood basalt formation. Earth and Planetary Science Letters 136, 269–279.[CrossRef][Web of Science]

Klausen, M. B. & Larsen, H. C. (2002). The East Greenland coast-parallel dyke swarm and its role in continental breakup. In: Menzies, M. A., Klemperer, S. L., Ebinger, C. J. & Baker, J. A. (eds) Volcanic Rifted Margins. Geological Society of America, Special Papers 362, 133–158.

Kystol, J. & Larsen, L. M. (1999). Analytical procedures in the Rock Geochemical Laboratory of the Geological Survey of Denmark and Greenland. Geological Survey of Greenland Bulletin 184, 59–62.

Larsen, H. C. & Saunders, A. D. (1998). Tectonism and volcanism at the SE Greenland rifted margin: a record of plume impact and later continental rupture. In: Larsen, H. C., Saunders, A. D. & Wise, S. W., Jr (eds) Proceedings of the Ocean Drilling Program, Scientific Results 152. College Station, TX: Ocean Drilling Program, pp. 503–533.

Larsen, L. M., Watt, W. S. & Watt, M. (1989). Geology and petrology of the Lower Tertiary plateau basalts of the Scoresby Sund region, East Greenland. Geological Survey of Greenland Bulletin 157, 1–164.

Larsen, L. M., Watt, W. S., Tegner, C., Storey, M. & Stecher, O. (1996a). High-Ti and low-Ti magmas within the central East Greenland Tertiary plateau basalts. EOS Transactions, American Geophysical Union 77(Supplement), 826.

Larsen, T. B., Yuen, D. A. & Smedsmo, L. J. (1996b). Thermomechanical modeling of pulsation tectonics and consequences on lithospheric dynamics. Geophysical Research Letters 23, 217–220.[CrossRef][Web of Science]

Lawver, L. A. & Müller, R. D. (1994). Iceland hotspot track. Geology 22, 311–314.[Abstract/Free Full Text]

Leeman, W. P., Dasch, E. J. & Kays, M. A. (1976). ²⁰⁷Pb/²⁰⁶Pb whole-rock age of gneisses from the Kangerlugssuaq area, eastern Greenland. Nature 263, 469–471.[CrossRef]

Lesher, C. E. (1990). Decoupling of chemical and isotopic exchange during magma mixing. Nature 344, 235–237.[CrossRef]

Lesher, C. E. & Baker, M. B. (1997). Near-solidus phase relations of mantle peridotite at 3·6–5 GPa. EOS Transactions, American Geophysical Union 78, F812.

Lundstrom, C. C., Shaw, H. F., Ryerson, F. J., Phinney, D. L., Gill, J. B. & Williams, Q. (1994). Compositional controls on the partitioning of U, Th, Ba, Pb, Sr and Zr between clinopyroxene and haploblastic melts: implications for uranium series disequilibria in basalts. Earth and Planetary Science Letters 128, 407–423.[CrossRef][Web of Science]

Mahoney, J. J. (1987). An isotopic survey of Pacific oceanic plateaus: implications for their nature and origin. In: Mahoney, J. J., Le Roex, A. P., Peng, Z., Fisher, R. L. & Natland, J. H. (1992). Southwestern limits of Indian Ocean Ridge mantle and the origin of low ²⁰⁶Pb/²⁰⁴Pb mid-ocean ridge basalt: isotope systematics of the Central Southwest Indian Ridge. Journal of Geophysical Research 97, 19771–19790.[CrossRef]

Mertz, D. F., Devey, C. W., Todt, W., Stoffers, P. & Hofmann, A. W. (1991). Sr–Nd–Pb isotope evidence against plume– asthenosphere mixing north of Iceland. Earth and Planetary Science Letters 107, 243–255.[CrossRef][Web of Science]

Mühe, R., Devey, C. W. & Bohrmann, H. (1993). Isotope and trace element geochemistry of MORB from the Nansen–Gakkel Ridge at 86° north. Earth and Planetary Science Letters 120, 103–109.[CrossRef][Web of Science]

Mutter, J. C., Buck, W. R. & Zehnder, C. Z. (1988). Convective partial melting. 1. A model for the formation of thick basaltic sequences during the initiation of spreading. Journal of Geophysical Research 93, 1031–1048.

Myers, J. S. (1980). Structure of the coastal dyke swarm and associated plutonic intrusions of East Greenland. Earth and Planetary Science Letters 46, 407–418.[CrossRef][Web of Science]

Nelson, D. R. (1989). Isotopic characteristics and petrogenesis of the lamproites and kimberlites of central west Greenland. Lithos 22, 265–274.[CrossRef][Web of Science]

Nicholson, H. & Latin, D. (1992). Olivine tholeiites from Krafla, Iceland: evidence for variations in melt fraction within a plume. Journal of Petrology 33, 1105–1124.[Abstract/Free Full Text]

Nicholson, H., Condomines, M., Fitton, J. G., Fallick, A. E., Gronvold, K. & Rogers, G. (1991). Geochemical and isotopic evidence for crustal assimilation beneath Krafla, Iceland. Journal of Petrology 32, 1005–1020.[Abstract/Free Full Text]

Nielsen, T. F. D. (1978). The Tertiary dike swarms of the Kangerlugssuaq area, East Greenland; an example of magmatic development during continental break-up. Contributions to Mineralogy and Petrology 67, 63–78.[CrossRef][Web of Science]

Nielsen, T. F. D. & Brooks, C. K. (1981). The East Greenland rifted continental margin: an examination of the coastal flexure. Journal of the Geological Society, London 138, 559–568.[Abstract/Free Full Text]

Peate, D. W. & Stecher, O. (2003). Pb isotope evidence for contributions from different Iceland mantle components in Palaeogene East Greenland flood basalts. Lithos 67, 39–52.[CrossRef][Web of Science]

Pedersen, A. K., Watt, M., Watt, W. S. & Larsen, L. M. (1997). Structure and stratigraphy of the early Tertiary basalts of the Blosseville Kyst, East Greenland. Journal of the Geological Society, London 154, 565–570.[Abstract/Free Full Text]

Pyle, D. G., Christie, D. M., Mahoney, J. J. & Duncan, R. A. (1995). Geochemistry and geochronology of ancient southeast Indian and southwest Pacific seafloor. Journal of Geophysical Research 100, 22261–22282.[CrossRef]

Saunders, A. D., Fitton, J. G., Kerr, A. C., Norry, M. J. & Kent, R. W. (1997). The North Atlantic Igneous Province. In: Mahoney, J. J. & Coffin, M. L. (eds) Large Igneous Provinces. Washington, DC: American Geophysical Union, pp. 45–93.

Saunders, A. D., Kempton, P. D., Fitton, J. G. & Larsen, L. M. (1999). Sr, Nd and Pb isotopes and trace element geochemistry of basalts from the Southeast Greenland Margin. In: Larsen, H. C., Duncan, R. D., Allan, J. F. & Brooks, C. K. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 163. College Station, TX: Ocean Drilling Program, pp. 77–93.

Shimizu, N. & Kushiro, I. (1975). The partitioning of rare earth elements between garnet and liquid at high pressures: preliminary experiments. Geophysical Research Letters 2, 414–416.

Shirey, S. B., Bender, J. F. & Langmuir, C. H. (1987). Three-component isotopic heterogeneity near the Oceanographer transform, Mid-Atlantic Ridge. Nature 325, 217–223.[CrossRef]

Skovgaard, A. C., Storey, M., Baker, J., Blusztajn, J., Hart, S. R. (2001). Osmium–oxygen isotopic evidence for a recycled and strongly depleted component in the Iceland mantle plume. Earth and Planetary Science Letters, 194, 259–275.[CrossRef][Web of Science]

Stecher, O., Carlson, R. W. & Gunnarson, B. (1999). Torfajökull: a radiogenic end-member of the Iceland Pb-isotopic array. Earth and Planetary Science Letters 165, 117–127.[CrossRef][Web of Science]

Storey, M., Kent, R., Saunders, A. D., Hergt, J., Salters, V. J., Whitechurch, H., Sevigny, J. H., Thirlwall, M. F., Leat, P., Ghose, N. C. & Gifford, M. (1992). Lower Cretaceous volcanic rocks on continental margins and their relationship to the Kerguelen Plateau. In: Barbu, E. M. (ed.) Proceedings of the Ocean Drilling Program, Scientific Results, 120. College Station, TX: Ocean Drilling Program, pp. 33–53.

Sun, S.-S. & Jahn, B. (1975). Lead and strontium isotopes in post-glacial basalts from Iceland. Nature 255, 527–530.[CrossRef]

Sun, S.-S. & McDonough, W. F. (1989). Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications 42, 313–345.

Taylor, P. N., Kalsbeek, F. & Bridgwater, D. (1992). Discrepancies between neodymium, lead and strontium model ages from the Precambrian of southern East Greenland: evidence for a Proterozoic granulite-facies event affecting Archean gneisses. Chemical Geology 94, 281–291.[Web of Science]

Taylor, R. N., Thirlwall, M. F., Murton, B. J., Hilton, D. R. & Gee, M. A. M. (1997). Isotopic constraints on the influence of the Icelandic plume. Earth and Planetary Science Letters 148, E1–E8.[CrossRef][Web of Science]

Taylor, S. R. & McLennan, S. M. (1995). The geochemical evolution of the continental crust. Reviews in Geophysics 33, 241–265.

Tegner, C., Duncan, R. A., Bernstein, S., Brooks, C. K., Bird, D. K. & Storey, M. (1998a). ⁴⁰Ar–³⁹Ar geochronology of Tertiary mafic intrusions along the East Greenland rifted margin: relation to flood basalts and the Iceland hotspot track. Earth and Planetary Science Letters 156, 75–88.[CrossRef][Web of Science]

Tegner, C., Lesher, C. L., Larsen, L. M. & Watt, W. S. (1998b). Evidence from the rare-earth-element record of mantle melting for cooling of the Tertiary Iceland plume. Nature 395, 591–594.[CrossRef]

Thirlwall, M. F. (1995). Generation of the Pb isotopic characteristics of the Iceland plume. Journal of the Geological Society, London 152, 991–996.[Abstract/Free Full Text]

Thompson, R. N., Morrison, A. P., Dickin, A. P. & Hendry, G. L. (1983). Continental flood basalts ... arachnids rule OK? In: Hawkesworth, C. J. & Norry, M. J. (eds) Continental Basalts and Mantle Xenoliths. Nantwich: Shiva, pp. 158–185.

Todt, W., Cliff, R. A., Hanser, A. & Hofmann, A. W. (1996). Evaluation of a ²⁰²Pb–²⁰⁵Pb double spike for high-precision lead isotope analyses. In: Basu, A. & Hart, S. R. (eds) Earth Processes: Reading the Isotopic Code. Geophysical Monograph, American Geophysical Union 95, 429–437.

Wager, L. R. (1935). Geological investigations in East Greenland. Part II. Geology of Kap Dalton. Meddelelser om Grønland 105(3), 1–32.

Wager, L. R. (1947). Geological investigations in East Greenland. Part IV. The stratigraphy and tectonics of Knud Rasmussens Land. Meddelelser om Grønland 134(5), 1–64.

Wager, L. R. & Deer, W. A. (1938). A dyke swarm and coastal flexure in East Greenland. Geological Magazine 75, 39–46.

Wedepohl, K., Heinrichs, H. & Bridgwater, D. (1991). Chemical characteristics and genesis of the quartz–feldspathic rocks in the Archean crust of Greenland. Contributions to Mineralogy and Petrology 107, 163–179.[CrossRef][Web of Science]

White, R. S. & McKenzie, D. (1989). Magmatism at rift zones: the generation of volcanic continental margins and flood basalts. Journal of Geophysical Research B94, 7685–7729.

White, R. S. & McKenzie, D. (1995). Mantle plumes and flood basalts. Journal of Geophysical Research 100, 17543–17585.[CrossRef]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				J. HANSEN, D. A. JERRAM, K. McCAFFREY, and S. R. PASSEY The onset of the North Atlantic Igneous Province in a rifting perspective Geological Magazine, May 1, 2009; 146(3): 309 - 325. [Abstract] [Full Text] [PDF]

				M. Storey, A.K. Pedersen, O. Stecher, S. Bernstein, H.C. Larsen, L.M. Larsen, J.A. Baker, and R.A. Duncan Long-lived postbreakup magmatism along the East Greenland margin: Evidence for shallow-mantle metasomatism by the Iceland plume Geology, February 1, 2004; 32(2): 173 - 176. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (5) Request Permissions Google Scholar Articles by HANGHØJ, K. Articles by STECHER, O. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Online ISSN 1460-2415 - Print ISSN 0022-3530

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics