1. Introduction
Closing of the Iapetus Ocean in Late Ordovician–Early Silurian time is generally envisioned to have involved subduction of the oceanic plate underneath the Laurentian margin (Stephens & Gee, Reference Stephens, Gee, Gee and Sturt1985; Grenne, Ihlen & Vokes, Reference Grenne, Ihlen and Vokes1999; Yoshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002; Meyer, Grenne & Pedersen, Reference Meyer, Grenne and Pedersen2003; Cocks & Torsvik, Reference Cocks, Torsvik, Gee and Stephenson2006). Island arcs and active margin magmatic arcs related to this event are well documented in the Appalachians, Scotland and Ireland (e.g. Friedrich et al. Reference Friedrich, Bowring, Martin and Hodges1999a, Reference Friedrich, Hodges, Bowring and Martinb; Van Staal et al. 1999, Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009; Whalen et al. Reference Whalen, McNicoll, van Staal, Lissenberg, Longstaffe, Jenner and van Breeman2006; Zagorevski et al. Reference Zagorevski, Rogers, van Staal, McNicoll, Lissenberg and Valverde-Vaquero2006; Oliver, Wilde & Wan, Reference Oliver, Wilde and Wan2008; Steinhofel, Hegner & Oliver, Reference Steinhoefel, Hegner and Oliver2008). Few of them are present in autochthonous crust, the majority being preserved in allochthons thrust upon the Laurentian continent during the continent collision phases of the Caledonian–Appalachian orogeny (e.g. Heaman, Erdmer & Owen, Reference Heaman, Erdmer and Owen2002; Goodenough, Young & Parsons, Reference Goodenough, Young and Parsons2004; Whalen et al. Reference Whalen, McNicoll, van Staal, Lissenberg, Longstaffe, Jenner and van Breeman2006; Van Staal et al. Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009).
Magmatic arc rocks with Laurentian affinities are also found in the highest nappes of the Scandinavian Caledonides, where they were thrust during the Silurian–Devonian continent–continent collision between Laurentia and Baltica (Grenne, Ihlen & Vokes, Reference Grenne, Ihlen and Vokes1999; Yoshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002; Meyer, Grenne & Pedersen, Reference Meyer, Grenne and Pedersen2003; Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007). The Laurentian origin is indicated by faunal, geochronological and detrital zircon evidence (Bruton & Bockelie, Reference Bruton, Bockelie and Wones1980; Stephens & Gee, Reference Stephens, Gee, Gee and Sturt1985; Pedersen, Bruton & Furnes, Reference Pedersen, Bruton and Furnes1992; Grenne, Ihlen & Vokes, Reference Grenne, Ihlen and Vokes1999; Yoshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002; Meyer, Grenne & Pedersen, Reference Meyer, Grenne and Pedersen2003; Roberts, Melezhik & Heldal, Reference Roberts, Melezhik and Heldal2002; Roberts, Nordgulen & Melezhik, Reference Roberts, Nordgulen, Melezhik, Hatcher, Carlson, McBride and Catalán2007; Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007). These arc rocks are apparently all of island arc type and no occurrences of continental Laurentian crust intruded by subduction-related arc magmas have been described (with the possible exception of the Skattøra migmatite complex of the Tromsø Nappe Complex; Selbekk, Skjerli & Pedersen, Reference Selbekk, Skjerlie and Pedersen2000).
In the Caledonides of East Greenland (Fig. 1) the dominant magmatic rocks are Silurian two-mica granites, inferred to be derived by partial melting of sedimentary cover during crustal thickening attributed to the Laurentia–Baltica continent collision (Hartz et al. Reference Hartz, Andresen, Martin and Hodges2000, Reference Hartz, Andresen, Hodges and Martin2001; Kalsbeek, Jepsen & Jones, Reference Kalsbeek, Jepsen and Jones2001; Kalsbeek, Jepsen & Nutman, Reference Kalsbeek, Jepsen and Nutman2001; White et al. Reference White, Hodges, Martin and Andresen2002; Leslie & Nutman, Reference Leslie and Nutman2003; Andresen, Rehnström & Holte, Reference Andresen, Rehnstrom and Holte2007; Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008). Two recent studies have, however, also described the presence of c. 454 to 420 Ma I-type intrusive rocks in Caledonian Allochthons composed of Laurentian continental crust (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008; Rehnström, Reference Rehnstrøm2010). Further evidence for arc magmatism in the East Greenland Caledonides is documented in the present paper, where we report new U–Pb isotope dilution thermal ionization mass spectrometry (ID-TIMS) and whole-rock major and trace element data from a large plutonic complex in Liverpool Land (Figs 1, 2). We discuss the previously available evidence and evaluate the tectonic implications of the data and their significance in the context of the Caledonian–Appalachian orogen.
2. Geological setting
The East Greenland Caledonides are part of the Early Palaeozoic Caledonian–Appalachian orogenic system exposed in eastern North America, the British Isles, Scandinavia and Svalbard. The East Greenland Caledonides consist of three main lithotectonic units (Fig. 1): (i) parautochthonous foreland basement composed of Archaean and Palaeoproterozoic rocks with a variably preserved cover of Neoproterozoic and Cambrian–Ordovician sedimentary rocks; (ii) allochthonous Archaean to Palaeoproterozoic gneisses overlain, with a depositional contact, by the Mesoproterozoic Krummedal supracrustal sequence, again overlain, either unconformably or (in most places) over an extensional detachment, by a thick package of Neoproterozoic (Eleonore Bay Supergroup) to Lower Ordovician sediments (Higgins et al. Reference Higgins, Elvevold, Escher, Frederiksen, Gilotti, Henriksen, Jepsen, Jones, Kalsbeek, Kinny, Leslie, Smith, Thrane and Watt2004; Andresen, Rehnström & Holte, Reference Andresen, Rehnstrom and Holte2007; Higgins & Leslie, Reference Higgins, Leslie, Higgins, Gilotti and Smith2008); (iii) Devonian continental deposits appearing in fault-controlled basins (Larsen & Bengaard, Reference Larsen and Bengaard1991; Hartz et al. Reference Hartz, Eide, Andresen, Midbøe, Hodges and Kristiansen2002; Larsen, Olsen & Clack, Reference Larsen, Olsen, Clack, Higgins, Gilotti and Smith2008). The Krummedal Sequence and the Eleonore Bay Supergroup are intruded by c. 430–425 Ma leucogranites, interpreted to be derived from partial melting of the Krummedal Sequence metasediments (Hartz et al. Reference Hartz, Andresen, Martin and Hodges2000, Reference Hartz, Andresen, Hodges and Martin2001; Kalsbeek, Jones & Jepsen, Reference Kalsbeek, Jepsen and Jones2001; Kalsbeek, Jones & Nutman, Reference Kalsbeek, Jepsen and Nutman2001; White et al. Reference White, Hodges, Martin and Andresen2002; Leslie & Nutman, Reference Leslie and Nutman2003; Andresen, Rehnström & Holte, Reference Andresen, Rehnstrom and Holte2007; Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008). In the very south of the exposed East Greenland Caledonides (in Renland and Milne Land; Fig. 1) there are Middle Ordovician to Silurian I-type granitoids with geochemical and isotopic signatures typical of arc magmas (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008; Rehnström, Reference Rehnstrøm2010).
Liverpool Land is the most internal Caledonian domain exposed in East Greenland and is isolated from the main area of Caledonian rocks by a cover of Permian and Mesozoic sedimentary rocks in the Jameson Land Basin (Fig. 2). The northern part of Liverpool Land is dominated by a plutonic complex, named here the Hurry Inlet Plutonic Terrane, whereas the southern part comprises eclogite-bearing gneisses of the Liverpool Land Eclogite Terrane (Kranck, Reference Kranck1935; Sahlstein, Reference Sahlstein1935; Coe & Cheeney, Reference Coe and Cheeney1972; Coe, Reference Coe1975; Friderichsen & Surlyk, Reference Friderichsen and Surlyk1981; Cheeney, Reference Cheeney1985; Augland, Andresen & Corfu, Reference Augland, Andresen and Corfu2010). The Hurry Inlet Plutonic Complex is separated from the Liverpool Land Eclogite Terrane by a major, N-dipping shear zone, the c. 400 m thick Gubbedalen Shear Zone, which is characterized by a complex kinematic history. The shear zone has top-up-to-the-S amphibolite to greenschist-facies contractional structures, subsequently superimposed by a zone of localized top-down-to-the-N semi-ductile to brittle extensional structures and an extensional detachment fault (Fig. 2) (Augland, Andresen & Corfu, Reference Augland, Andresen and Corfu2010). The Liverpool Land Eclogite Terrane is interpreted, based on its composition, age relations and the preserved magmatic and metamorphic record, to represent a piece of Baltican crust amalgamated to Laurentia in Mid-Devonian time (Augland, Andresen and Corfu, Reference Augland, Andresen and Corfu2011). By contrast the Hurry Inlet Plutonic Terrane (Corfu & Hartz, Reference Corfu and Hartz2011) is interpreted to represent the tectonostratigraphic level equivalent to the allochthonous Caledonian rocks exposed west of the Jameson Land Basin (Fig. 1). This correlation is based on similarities between partly migmatitic metasedimentary rocks on Liverpool Land and the Krummedal Sequence to the west (Higgins, Reference Higgins and Winchester1988; Johnston et al. Reference Johnston, Gehrels, Valencia and Ruiz2009, Reference Johnston, Hartz, Brueckner and Gehrels2010).
2.a. Hurry Inlet Plutonic Terrane
The Hurry Inlet Plutonic Terrane is dominated by Caledonian granitoid rocks intruding partly migmatitic metasedimentary rocks (Kranck, Reference Kranck1935; Hansen & Steiger, Reference Hansen and Steiger1971; Coe & Cheeney, Reference Coe and Cheeney1972; Coe, Reference Coe1975; Friderichsen & Surlyk, Reference Friderichsen and Surlyk1981; Hansen & Friderichsen, Reference Hansen and Friderichsen1987; Higgins, Reference Higgins and Winchester1988; Johnston et al. Reference Johnston, Gehrels, Valencia and Ruiz2009, Reference Johnston, Hartz, Brueckner and Gehrels2010; this study). The largest of these granitoid intrusions are the Late Ordovician to Silurian Hurry Inlet Composite Pluton (see Section 3.a; Kranck, Reference Kranck1935; Hansen & Friderichsen, Reference Hansen and Friderichsen1987) and the Silurian Hodal-Storefjord Pluton (Fig. 2) (see Section 3.b; Coe & Cheeney, Reference Coe and Cheeney1972; Hansen & Friderichsen, Reference Hansen and Friderichsen1987). The Hurry Inlet Composite Pluton and the Hodal-Storefjord Pluton probably belong to the same batholithic complex, both intruding the Krummedal Sequence metasedimentary rocks. A brittle extensional fault separates the Hodal-Storefjord Pluton from the Triaselv leucogranite to the north (Fig. 2).
The Hurry Inlet Composite Pluton can roughly be divided in three major units visible on a map scale: (i) mainly undeformed granite to monzonite rocks in the west; (ii) mainly undeformed diorite to monzonite in the east; and (iii) foliated granodiorite in between (Coe & Cheeney, Reference Coe and Cheeney1972; Coe, Reference Coe1975). However, in the field and in thin-section at least six different intrusive rock types have been recognized based on texture and modal compositions. The varieties examined here range from monzonite to granodiorite to granite and from equigranular to porphyritic in texture. Some intrusive phases contain more than 20% amphibole and no biotite. Others contain biotite as the only major mafic phase. Hansen & Friderichsen (Reference Hansen and Friderichsen1987) obtained a whole-rock Rb–Sr isochron from six different samples from the Hurry Inlet Composite Pluton of 438 ± 42 Ma (MSWD = 0.99). The Hodal-Storefjord Pluton ranges from diorite to quartz-monzodiorite and most samples are equigranular, although some porphyritic varieties occur close to its margin. Amphibole content of the samples ranges from c. 5 to 20% and biotite content ranges between c. 5 and 15%. Some samples contain up to 13% pyroxene. At one locality a mafic rock (HKD 08-18) comprised of amphibole (80%), olivine (2%), plagioclase (c. 8%) and opaques (10%) occurs as an enclave in the Hodal-Storefjord Pluton. The enclave itself contains inclusions of the Hodal-Storefjord Pluton, testifying to their comagmatic nature.
Major element analyses of 25 samples and trace element analyses of 24 samples reported by Coe (Reference Coe1975) for the Hurry Inlet Composite Pluton, define a calc-alkaline trend in the AFM diagram and plot in the high-K field in the K2O–SiO2 diagram (Fig. 3a, c). All samples except no. 23 are magnesian according to the classification of Frost et al. (Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) and all samples except no. 23 are alkali-calcic to calc-alkalic as defined by the modified alkali–lime index of Frost et al. (Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) (Fig. 3f, g). Most samples are slightly peraluminous, except a few samples with a clear Al-enrichment (Fig. 3e; Maniar & Piccoli, Reference Maniar and Piccoli1989). The samples plot in the volcanic-arc granite field of the (Y + Nb) v. Rb and the Y v. Nb discrimination diagrams of Pearce, Harris & Tindle (Reference Pearce, Harris and Tindle1984) (Fig. 3b, d). In a chondrite-normalized (Thompson, Reference Thompson1982) spider diagram, the trace elements show high Ba and Sr values, a distinct negative trough for Nb and smaller, negative P and Ti anomalies (Fig. 3i).
The Hodal-Storefjord Pluton samples (Table 1) define a calc-alkaline trend in the AFM diagram and all samples except the mafic enclave plot in the high-K field of the K2O–SiO2 diagram (Fig. 3a, c). The samples are magnesian, alkali-calcic to calc-alkalic and metaluminous (Fig. 3f, h) (Maniar & Piccoli, Reference Maniar and Piccoli1989; Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001). The Hodal-Storefjord Pluton samples plot in the volcanic-arc granite field in the (Y + Nb) v. Rb and the Y v. Nb diagrams (Fig. 3b, d) (Pearce, Harris & Tindle, Reference Pearce, Harris and Tindle1984). In a chondrite-normalized (Thompson, Reference Thompson1982) spider diagram the trace elements show elevated Ba and Sr values and a distinct Nb trough (Fig. 3j).
All data determined with a PanAlytical Axios wavelength dispersive X-ray fluorescence (WD-XRF) spectrometer at the Norwegian Geological Survey, Trondheim, Norway.
The mafic enclave (HKD 08-18) has a very low Al2O3 and high Fe2O3 content, and based on the field relations with inclusions of both the mafic rock within a monzodioritic variety of the Hodal-Storefjord Pluton and the monzodiorite within the mafic enclave, magma mingling or mixing must have occurred. The rock has elevated Ni (50 ppm) and Cr (89 ppm) concentrations. This sample has negative Nb and Sr spikes (Fig. 3j). The latter probably reflects the crystallization of plagioclase in the dominant granitoid phase synchronously with mingling or mixing.
Major and trace elements are also reported for a two-amphibole-mica meladiorite (containing c. 25% hornblende, c. 25% cummingtonite, partly altered to actinolite along the grain margins, c. 25% biotite and c. 25% plagioclase) occurring as a xenolith in the Triaselv leucogranite (Fig. 2). The sample (HKD 08-23; Table 1) has an Mg no. of 52 (Fig. 3a) and has high Cr (613 ppm) and Ni (202 ppm) concentrations.
2.b. Rb–Sr isotope data
Rb–Sr analyses of eight samples from the Hurry Inlet Composite Pluton were reported by Hansen & Friderichsen (Reference Hansen and Friderichsen1987). Initial 87Sr/86Sr ratios of 0.708 to 0.711 were calculated at an age of 440 Ma (Fig. 4; Table 2). Two samples from the Hodal-Storefjord Pluton (Hansen & Friderichsen, Reference Hansen and Friderichsen1987), calculated at 424 Ma, yield initial 87Sr/86Sr ratios of 0.707 and 0.708 (Fig. 4; Table 2).
Initial Sr isotopic ratios calculated at 440 Ma for the HICP and 424 Ma for the HSP.
Data from Hansen & Friderichsen (Reference Hansen and Friderichsen1987)
3. Geochronology
Zircons from two phases of the Hurry Inlet Composite Pluton, and one sample from the Hodal-Storefjord Pluton and a dioritic xenolith included in the Triaselv leucogranite have been dated (Fig. 2). The U–Pb analyses were conducted by isotope dilution thermal ionization mass spectrometry (ID-TIMS) at the Department of Geosciences, University of Oslo. The reader is referred to appendix A of Augland, Andresen & Corfu (Reference Augland, Andresen and Corfu2010) for a thorough description of the analytical procedure.
3.a. Hurry Inlet Composite Pluton
Samples PIM 05-Li02 and LEA 06-71A are from medium-grained, equigranular K-feldspar-rich, amphibole and minor clinopyroxene-bearing granitic phases in the southernmost part of the Hurry Inlet Composite Pluton (Fig. 2; Table 3).
The zircon population of LEA 06-71A is relatively homogeneous. Most grains are colourless, with some weakly red and yellow metamict grains. Prismatic grains dominated by 100 and 101 forms are most common, with elongation ratios ranging from 2 to 6. The grains are generally rich in inclusions and seldom contain visible cores. The U content in the analysed fractions ranges from c. 130 to c. 1700 ppm with moderate Th/U ratios of 0.29 to 0.53 (Table 3).
The zircons from PIM 05-Li02 have similar shapes and morphologies to LEA 06-71A, but are more heterogeneous, ranging from colourless to reddish or brown, in part metamict and some contain cores. The U content in the analysed fractions ranges from c. 379 to c. 1034 ppm with moderate Th/U ratios of 0.37 and 0.50 (Table 3).
Of the eight fractions analysed from LEA 06-71A (Fig. 5b), two contain inheritance (nos 5 and 8). Five other analyses define a discordia age, anchored at 0 Ma, with an intercept age of 445 ± 4 Ma (2σ, MSWD = 0.92). Analysis no. 1 is concordant and gives a concordia age of 446 ± 2 Ma (2σ, MSWD = 0.69), concurrent with the upper intercept age of the five point discordia line, and we consider that age as the best estimate of crystallization of this intrusive phase. The eighth fraction yields a concordant analysis (no. 6) that does not fit the line (c. 434 Ma), indicating either some disturbance or growth in response to younger intrusions or hydrothermal effects, or less likely, due to some unknown analytical bias. An alternative explanation is that the older fractions represent antecrysts from an older magmatic pulse in the Hurry Inlet Composite Pluton and the young concordant zircon fraction dates the actual crystallization of this particular phase (Miller et al. Reference Miller, Matzel, Miller, Burgess and Miller2007).
Of the four fractions analysed from PIM 05-Li02, one is concordant and the three remaining are less than 3.4% discordant (Fig. 5a; Table 3). The four analyses of PIM 05-Li02 are collinear and a line anchored at 0 Ma give an upper intercept age of 438 ± 4 Ma (2σ, MSWD = 0.38).
3.b. Hodal-Storefjord Pluton
Sample LEA 06-101 from the Hodal-Storefjord Pluton was collected just north of the NW-trending zone of paragneisses which it intrudes (Fig. 2; Table 3). It is a medium-grained equigranular, plagioclase-rich (c. 50%) diorite, with c. 15% biotite, 10% orthopyroxene, c. 5% K-feldspar, c. 5 % quartz, c. 5% amphibole, c. 3% clinopyroxene, c. 2% opaques and accessory zircon.
The zircon population of this sample is relatively homogeneous, with colourless, prismatic zircon grains dominated by 100 and 101 surfaces. The elongation ratio varies from 2 to 5, and the analysed fractions were generally in the upper range of these values. Most grains contain inclusions and about one third of the grains have smooth sub-rounded terminations. Cores were not observed. U content is low to moderate, varying from 61 to 225 ppm, except for fraction no. 2, with a U content of c. 1650 ppm. Th/U ranges between 0.74 and 0.89 (Table 3).
Six data points are clustered together on the concordia curve (Fig. 5d). One analysis (no. 1) is 2.8% discordant and is omitted from the calculation of the concordia age of 423.7 ± 0.7 Ma (2σ, MSWD = 0.047; 5 points).
3.c. Meladioritic xenolith in the Triaselv leucogranite
Sample HKD 08-23 was collected just north of the Triaselv extensional fault (Table 3; Fig. 2), where it occurs as a meladioritic xenolith in the Triaselv leucogranite. As with sample LEA 06-71 of the Hurry Inlet Composite Pluton, this sample also contains relatively homogeneous zircons, with colourless, prismatic zircon grains dominated by 100 and 101 surfaces. Elongation ratios vary from 2 to 5. The analysed grains contain few inclusions, and no cores were observed. The U content is similar to LEA 06-71 from the Hurry Inlet Composite Pluton, varying from 354 to 2200 ppm. Th/U ranges from 0.28 to 0.39 (Table 3).
The four data points for one tip and three multigrain fractions are clustered close to the concordia curve, all analyses are less than 1.1% discordant and the discordance is attributed to recent Pb loss (Fig. 5c). An upper intercept age including all data points, anchored at 0 Ma, of 425.5 ± 1.2 Ma (2σ, MSWD = 0.22) is interpreted as the crystallization age of the meladiorite (Fig. 5c).
4. Discussion
4.a. Timing and genesis of Ordovician–Silurian intrusives in Liverpool Land
The two intrusive phases of the Hurry Inlet Composite Pluton with ages of 446 ± 2 Ma and 438 ± 4 Ma, the meladioritic xenolith in the Triaselv leucogranite with an age of 426 ± 1 Ma, and the Hodal-Storefjord Pluton with an age of 424 ± 1 Ma, show that the Hurry Inlet Plutonic Terrane underwent repeated intrusions in a period of more than 20 Ma (regardless of whether the oldest phase dated here represents antecrysts from an earlier magmatic pulse or the actual solidification of the studied rock). This pattern confirms the geochronological data of Corfu & Hartz (Reference Corfu and Hartz2011) who obtained ages between c. 440 and 425 Ma for different granitoid plutons in Liverpool Land. However, it appears that the earlier intrusive activity was areally more scattered and much less voluminous than the main event at 426–424 Ma (Fig. 6). There are some indications in the zircon data of sample LEA 06-71A and in some of the samples studied by Corfu & Hartz (Reference Corfu and Hartz2011) with a spread of concordant data points, for the existence of a long-lived crystal mush producing multiple injections of magma (e.g. Miller et al. Reference Miller, Matzel, Miller, Burgess and Miller2007). The alkali-calcic to calc-alkalic (Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001) nature of the Hurry Inlet Composite Pluton and the Hodal-Storefjord Pluton and their typical calc-alkaline trends in the AFM diagram are probably the result of fractional crystallization of magmas with a mixed crustal and mantle source and/or mixing of different amounts of crustal material with juvenile magmas (Barbarin, Reference Barbarin1999). This is supported by the 87Sr/86Sr initial ratios of 0.707–0.711 (Fig. 4; Table 2) (Hansen & Friderichsen, Reference Hansen and Friderichsen1987).
We interpret the described intrusive rocks from the Hurry Inlet Plutonic Terrane to be continental arc granitoids (Fig. 7). This is based on: (i) the prolonged magmatism with a calc-alkaline trend; (ii) 87Sr/86Sr initial ratios between 0.707 and 0.711 from the Hurry Inlet Composite Pluton and Hodal-Storefjord Pluton indicating that the magmas originated from a mixed crustal and juvenile source (Fig. 4); (iii) the magnesian and metaluminous to only slightly peraluminous geochemistry of the analysed rocks; (iv) the abundance of amphibole in most, and of pyroxene in some samples; and (v) the trace element patterns with strongly enriched Ba and Sr concentrations normalized to chondrite values and characteristic Nb-troughs (and also Ti-troughs) (Shand, Reference Shand1943; Thompson, Reference Thompson1982; McCulloch & Gamble, Reference McCulloch and Gamble1991; Barbarin, Reference Barbarin1999; Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001; Murphy, Reference Murphy2007). The calk-alkali to dominantly alkali-calcic compositions of the rocks could be due to a position relatively far from the trench at the time of intrusion, as the alkali–lime index in continental arc granitoids is in many cases controlled by the distance from the trench (e.g. Gill, Reference Gill1981 and references therein).
The presence of the two-amphibole-biotite meladiorite xenolith (426 Ma) with a primitive character (Mg no. = 52; Ni = 202 ppm; Cr = 613 ppm) compared to the Hurry Inlet Composite Pluton, and the mafic enclave in the Hodal-Storefjord Pluton, together with the more metaluminous character of this pluton compared to the Hurry Inlet Composite Pluton, suggest that the late pulses of magmatism (at 425 Ma) in the Hurry Inlet Plutonic Terrane were associated with a more primitive parental magma (Fowler & Henney, Reference Fowler and Henney1996; Barbarin, Reference Barbarin1999; Frost et al. Reference Frost, Barnes, Collins, Arculus, Ellis and Frost2001). If the slightly lower 87Sr/86Sr initial ratios of the Hodal-Storefjord Pluton than of the Hurry Inlet Composite Pluton are representative of their original magmas, this would also indicate that the later pulses of magmatism had a larger juvenile component than the earlier pulses. The Sr isotopic ratios of the Hodal-Storefjord Pluton are also interpreted to be the result of contamination of a juvenile source, or magma, with crustal material represented by the typical Krummedal Sequence (Fig. 4).
4.b. Coeval intrusive suites in the nappes of East Greenland
The combination of Late Ordovician plutonism, with the concluding 425 Ma magmatic episode seen in Liverpool Land, is also observed in Renland and Milne Land further west (Figs 1, 6) (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008; Rehnström, Reference Rehnstrøm2010), and may be a characteristic feature of the magmatic activity in the southernmost part of the East Greenland Caledonides.
Whole-rock and trace element geochemical data of Middle to Late Ordovician granitoids in Renland and Milne Land also show patterns that are very similar to those found in the Hurry Inlet Composite Pluton and the Hodal-Storefjord Pluton, with typical magmatic arc characteristics (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008; Rehnström, Reference Rehnstrøm2010). Like the Hurry Inlet Composite Pluton, they plot in the high-K field in the K2O–SiO2 diagram, in the volcanic-arc granite field of the (Y + Nb) v. Rb discrimination diagram, and display Nb- and Ti-troughs and high Ba values in the chondrite-normalized spider diagram (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008). The Milne Land granitoids have been interpreted to have a mixed mantle and crustal source based on Sr and Nd isotopes (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008). The initial 87Sr/86Sr isotope signatures recorded in these Ordovician rocks typically range between 0.707 and 0.708, only slightly lower than what is observed in the Hurry Inlet Composite Pluton (Table 2).
The younger ‘I-type’ granitoids (as young as c. 420 Ma) from Renland and Milne Land also show an isotopic and chemical trend compatible with a mixed mantle and crustal source (Rehnström, Reference Rehnstrøm2010; Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008). These rocks have Sr initial ratios of between 0.707 and 0.715, with most samples at c. 0.708, and εNd ranging between −3 and −5 for most samples. Geochemically, isotopically and temporally these granitoids are very similar to the Hodal-Storefjord Pluton, and we propose a close correlation between these and the younger magmatic pulses in Liverpool Land.
Ages similar to that of the dioritic xenolith and the Hodal-Storefjord Pluton (c. 425 Ma) are common for leucogranites in the southern part of the East Greenland Caledonides. However, the geochemistry and modal composition of the former are distinctly different from those of the typical two-mica leucogranites in East Greenland. The difference also stands out very clearly in the Sr isotopic signature. The 87Sr/86Sr ratios reported by Hansen & Friderichsen (Reference Hansen and Friderichsen1987) for the Hodal-Storefjord Pluton and Hurry Inlet Composite Pluton are considerably lower (0.706–0.709) than the values reported by Kalsbeek, Jepsen & Nutman (Reference Kalsbeek, Jepsen and Jones2001) for the leucogranites and the Krummedal Sequence which the Hurry Inlet Composite Pluton and Hodal-Storefjord Pluton intrudes. The leucogranites and the metasediments mainly range from 0.72 to 0.74 (Fig. 4) reflecting an origin of the leucogranites by partial melting of the Krummedal metasedimentary rocks (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008). A further difference is noticed in the abundance of xenocrystic zircons with sparse Caledonian zircon growth in the leucogranites (Watt, Kinny & Friderichsen, Reference Watt, Kinny and Friderichsen2000; Kalsbeek, Jepsen & Nutman, Reference Kalsbeek, Jepsen and Nutman2001; Gilotti & McClelland, Reference Gilotti and McClelland2005; Andresen, Rehnström & Holte, Reference Andresen, Rehnstrom and Holte2007), whereas inheritance is much less common or absent in the Hurry Inlet Composite Pluton and Hodal-Storefjord Pluton.
The leucogranites have been interpreted to be the product of post-orogenic collapse (Strachan, Martin & Friderichsen, Reference Strachan, Martin and Friderichsen2001), but it has also been argued that they are synorogenic resulting from contraction and extension at different levels in the crust (e.g. Andresen, Rehnström & Holte, Reference Andresen, Rehnstrom and Holte2007).
4.c. Connections with Ordovician magmatic suites in the Scandinavian Caledonides
The ages of the Ordovician granitoid rocks in East Greenland overlap those of calc-alkaline granitoids in the Helgeland Nappe Complex of the Uppermost Allochthon of the Scandinavian Caledonides (Nordgulen et al. Reference Nordgulen, Bickford, Nissen and Wortman1993; Yoshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002; Meyer, Grenne & Pedersen, Reference Meyer, Grenne and Pedersen2003; Tucker et al. Reference Tucker, Robinson, Solli, Gee, Thorsnes, Krogh, Nordgulen and Bickford2004; Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007). The latter granitoids are thought to have formed in a Laurentian island arc (ensimatic) setting (Stephens & Gee, Reference Stephens, Gee, Gee and Sturt1985; Grenne, Ihlen & Vokes, Reference Grenne, Ihlen and Vokes1999; Roberts, Melezhik & Heldal, Reference Roberts, Melezhik and Heldal2002; Yoshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002; Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007; Roberts, Nordgulen & Melezhik, Reference Roberts, Melezhik and Heldal2007), linked to subduction of Iapetus oceanic crust beneath Laurentia. They are further considered to represent the northern continuation of magmatic arcs preserved in the Irish and Scottish Caledonides and the Appalachians (Van Staal et al. Reference Van Staal, Dewey, Mac Niocaill, McKerrow, Blundell and Scott1998, Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009; Friedrich et al. Reference Friedrich, Bowring, Martin and Hodges1999a, Reference Friedrich, Hodges, Bowring and Martinb; Armstrong & Owen, Reference Armstrong and Owen2001; Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007; Oliver, Wilde & Wan, Reference Oliver, Wilde and Wan2008). Similarly aged magmatic complexes in the Upper Allochthon in Scandinavia are also attributed to ensimatic magmatism above a W-dipping subduction zone close to the Laurentian margin at the latest stages of Iapetus ocean closure (Grenne & Roberts, Reference Grenne and Roberts1998; Grenne, Ihlen & Vokes Reference Grenne, Ihlen and Vokes1999; Meyer, Grenne & Pederesen, Reference Meyer, Grenne and Pedersen2003). As emphasized by Grenne, Ihlen & Vokes (Reference Grenne, Ihlen and Vokes1999), Baltica was rotating rapidly in an anticlockwise manner at the Ordovician–Silurian boundary (Torsvik et al. Reference Torsvik, Smethurst, Meert, Van der Voo, McKerrow, Brasier, Sturt and Walderhaug1996; Torsvik, Reference Torsvik1997), and this could have led to different regimes of contraction and extension parallel to the orogen, resulting in widespread magmatism of variable character (i.e. from extension-related back-arc rifting to convergent margin arc magmas), with certain diachroneities of magmatism along the orogen (e.g. Tucker, Boyd & Barnes, Reference Tucker, Boyd and Barnes1990; Pedersen, Furnes & Dunning, Reference Pedersen, Furnes and Dunning1991; Youshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002; Meyer, Grenne & Pederesen, Reference Meyer, Grenne and Pedersen2003; Corfu et al. Reference Corfu, Torsvik, Andersen, Ashwal, Ramsay and Roberts2006; Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007; Nilsen, Corfu & Roberts, Reference Nilsen, Corfu and Roberts2007).
4.d. Connections with the Newer Granites in the Scottish Caledonides: magma generation at c. 425 Ma related to slab break-off
In terms of geochemistry, isotope signature and age, the younger ‘I-type’ granitoids from the Hurry Inlet Composite Pluton and Hodal-Storefjord Pluton show many similarities with the ‘Newer Granites’ of the Scottish Caledonides (e.g. Rogers & Dunning, Reference Rogers and Dunning1991; Strachan et al. Reference Strachan, Smith, Harris, Fettes and Trewin2002), and a correlation between them as part of a common magmatic arc system is viable (Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008; Oliver, Wilde & Wan, Reference Oliver, Wilde and Wan2008; Steinhofel, Hegner & Oliver, Reference Steinhoefel, Hegner and Oliver2008). This 425 Ma phase of magmatism was temporally linked to the terminal phases of Iapetus closure and oceanic crust-subduction underneath northeastern Laurentia (East Greenland). It was the most voluminous pulse of magmatism recorded in Liverpool Land, and in Renland and Milne Land and it lasted less than 5 Ma. (Fig. 6; this study; Kalsbeek et al. Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008; Rehnström, Reference Rehnstrøm2010; Corfu & Hartz, Reference Corfu and Hartz2011). The brevity and voluminous nature of the magmatic event is consistent with a model of slab break-off (Fig. 7b; Atherton & Ghani, Reference Atherton and Ghani2002; Whalen et al. Reference Whalen, McNicoll, van Staal, Lissenberg, Longstaffe, Jenner and van Breeman2006; Van Staal et al. Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009). Slab break-off would have led to extra heat flow into the lower crust of the upper plate and thus could have provided a heat source for the melting of the Krummedal Sequence and the generation of two-mica granites (Kalsbeek, Jepsen & Nutman, Reference Kalsbeek, Jepsen and Jones2001; Kalsbeek, Jepsen & Jones, Reference Kalsbeek, Jepsen and Nutman2001; Andresen, Rehnström & Holte, Reference Andresen, Rehnstrom and Holte2007; Rehnström, Reference Rehnstrøm2010).
4.e. Links to Caledonian intrusives in the autochthon of East Greenland
A possible subduction-related Caledonian intrusive complex in the Kangerdlugssuaq area, some 400 km southwest of Liverpool Land, may provide an important link between the allochthonous magmatic arc rocks in East Greenland described above and those further south and southwest (present-day Scotland, Ireland and eastern North America). The Batbjerg intrusion is a partly deformed ultrapotassic layered intrusion dated by K–Ar geochronology to c. 440 Ma (Brooks, Fawcett & Gittens, Reference Brooks, Fawcett and Gittins1976; Brooks et al. Reference Brooks, Fawcett, Gittins and Rucklidge1981). An initial 87Sr/86Sr ratio was calculated to 0.705 (Brooks, Fawcett & Gittens, Reference Brooks, Fawcett and Gittins1976). The Batbjerg pluton is intruding basement gneisses thought to be of autochthonous nature and contains screens of Palaeozoic metasedimentary rocks with crinoid fragments (Brooks, Fawcett & Gittens, Reference Brooks, Fawcett and Gittins1976; Brooks et al. Reference Brooks, Fawcett, Gittins and Rucklidge1981). Furthermore the gneisses surrounding the Batbjerg intrusion are intruded by numerous volcanic dykes and sheets of appinitic character that are probably related to the intrusive complex itself (Brooks, Fawcett & Gittens, Reference Brooks, Fawcett and Gittins1976; Brooks et al. Reference Brooks, Fawcett, Gittins and Rucklidge1981).
Intrusive rocks similar to the Batbjerg ultrapotassic complex are present in the Assynt culmination of the Moine Thrust Belt in the Scottish Caledonides. There, alkaline to calc-alkaline intrusive rocks (including pyroxenite to quartz syenites) intrude the (par)autochthonous Archaean basement with its sedimentary cover (Halliday et al. Reference Halliday, Aftalion, Parsons, Dickin and Johnson1987; Goodenough, Young & Parsons, Reference Goodenough, Young and Parsons2004; Styles, Gunn & Rollin, Reference Styles, Gunn and Rollin2004). Especially the c. 430 Ma ultrapotassic Loch Borrelan intrusion (Van Breemen et al. Reference Van Breemen, Aftalion, Pankhurst and Richardson1979; Halliday et al. Reference Halliday, Aftalion, Parsons, Dickin and Johnson1987) shares several characteristics with the Batbjerg intrusion, and a correlation of the two could be justified. Based on the structural setting and their compositions, Goodenough, Young & Parsons (Reference Goodenough, Young and Parsons2004) suggested that these intrusive rocks formed in an ensialic setting in the continental margin of Laurentia, both prior to and in the initial stage of the Scandian phase of the Caledonian orogeny (Halliday et al. Reference Halliday, Aftalion, Parsons, Dickin and Johnson1987).
4.f. Ordovician–Silurian evolution of the northeastern Laurentian margin
The East Greenland Caledonides show little or no evidence of the early Taconian (Grampian, > 460 Ma) contractional events and island arc magmatism present in the North American Appalachians and in the British Caledonides (e.g. Friedrich et al. Reference Friedrich, Bowring, Martin and Hodges1999a, Reference Friedrich, Hodges, Bowring and Martinb; Kinny et al. Reference Kinny, Friend, Strachan, Watt and Burns1999; Van Staal et al. 1999, Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009), respectively. The one exception is an age of 466 ± 9 Ma reported by Kalsbeek et al. (Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008) for a granodiorite from Milne Land, but that age was interpolated from scattered SHRIMP U–Pb analyses with an obvious component of inheritance. It is thus imprecise and not very reliable. It also conflicts with the occurrence of platform carbonates as young as c. 460 Ma in the allochthon of the Central East Greenland Caledonides (Smith & Rasmussen, Reference Smith, Rasmussen, Higgins, Gilotti and Smith2008). These platform carbonates make it seem unlikely that active margin magmatism was initiated prior to the end of passive margin sedimentation, supporting the inference that this part of the East Greenland Caledonides was not affected by an early Taconic (Grampian) orogenic phase. However, the data of Kalsbeek et al. (Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008) clearly showed the existence of Caledonian granitoids older than the common 425 Ma leucogranites. A more reliable age for the same rock as the one dated to 466 ± 9 Ma by Kalsbeek et al. (Reference Kalsbeek, Higgins, Jepsen, Frei, Nutman, Higgins, Gilotti and Smith2008) of c. 454 Ma was obtained by Rhenström (Reference Rehnstrøm2010). This age shows that arc magmatism in East Greenland was synchronous with crystallization of the first arc magmas in what is now the Uppermost Allochthon of the Scandinavian allochthon (Yoshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002). This relatively late initiation of arc magmatism in the East Greenland and Helgeland nappes is consistent with the suggestion that arc magmatism migrated northwards with time (Van Staal et al. Reference Van Staal, Dewey, Mac Niocaill, McKerrow, Blundell and Scott1998, Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009). A plausible hypothesis is that the initiation of arc magmatism in the continental margin, in the areas of East Greenland described here, was synchronous with amalgamation to the Laurentian continent of island arcs (or peri-Laurentian terranes) now preserved in the Uppermost and possibly Upper Allochthons of Scandinavia (Fig. 7a) (Stephens & Gee, Reference Stephens, Gee, Gee and Sturt1985; Yoshinobu et al. Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002; Meyer, Grenne & Pedersen, Reference Meyer, Grenne and Pedersen2003; Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007). According to Yoshinobu et al. (Reference Yoshinobu, Barnes, Nordgulen, Prestvik, Fanning and Pedersen2002) there could have been a switch in the subduction polarity from eastward to westward subduction (in present-day coordinates) as the outboard island arcs ‘arrived’ at the Laurentian margin. A similar mechanism has been proposed by Van Staal et al. (Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009) for the late Taconian stage in the Appalachians. Such a subduction scenario would explain why arc magmatism was not initiated in East Greenland until after 460 Ma.
Such a switch in subduction polarity would also explain the protracted arc magmatism that lasted to c. 425 Ma in the East Greenland nappes, the northern Appalachians, the British Caledonides and the Helgeland Nappe (Fig 7b). The terranes in the latter would have been amalgamated to the periphery of Laurentia, experiencing the last pulse of subduction-related (possibly slab break-off-related) magmatism together with the true continental Laurentian margin before being thrust upon the Baltic Plate during the Scandian continent–continent collision. At present only a few granitoids have been dated to around 425 Ma in the Uppermost Allochthon (Barnes et al. Reference Barnes, Frost, Yoshinobu, McArthur, Barnes, Allen, Nordgulen and Prestvik2007). However, it remains uncertain whether the scarcity of this age group is due to the relatively small number of samples dated from the total volume of intrusive rocks in the Uppermost Allochthon or whether it is a real feature that could indicate that c. 425 Ma magmatism occurred mostly further inland towards the Laurentian foreland. In the British Isles and in the Appalachians, arc magmatism continued into Early Devonian time due to the later arrival of Avalonia to Laurentia (e.g. Oliver, Wilde & Wan, Reference Oliver, Wilde and Wan2008; Van Staal et al. Reference Van Staal, Whalen, Valverde-Vaquero, Zagorevski, Rogers, Murphy, Keppie and Hynes2009).
The analogies and sequence of events proposed here help to reconstruct the northeastern Iapetus palaeogeography supporting a position of the Uppermost Allochthon of the Scandinavian Caledonides close to the southern part of today's exposed allochthonous East Greenland Caledonides prior to continent–continent collision between Baltica and Laurentia.
Acknowledgements
We thank Gunborg Bye Fjeld for assistance with sample preparations and Muriel Erambert for analytical assistance. We highly appreciate the constructive reviews of Ø. Nordgulen and two anonymous reviewers.