1. Introduction
The western part of Spitsbergen Island exposes an elongate N–S-trending basement high (Fig. 1) uplifted in the axial zone of an Early Palaeogene fold-and-thrust belt (e.g. Lowell, Reference Lowell1972; Dallmann et al. Reference Dallmann, Andresen, Bergh, Maher and Ohta1993; Braathen, Bergh & Maher, Reference Braathen, Bergh and Maher1995; Bergh, Braathen & Andresen, Reference Bergh, Braathen and Andresen1997). This elevated basement block consists of Precambrian to Early Palaeozoic metamorphic rocks collectively defined in the Svalbard Archipelago as the Hecla-Hoek Formation (Kulling, Reference Kulling1934). Deformation and metamorphism of this formation was originally ascribed solely to the Caledonian Orogeny that consolidated the pre-Devonian basement of Svalbard (e.g. Holtedahl, Reference Holtedahl1926; Harland, Reference Harland1959, Reference Harland, Gee and Sturt1985; Ohta, Reference Ohta1982; Manby, Reference Manby1990; Harland et al. Reference Harland, Scott, Auckland and Snape1992). However, several field studies recognize regional unconformities within the pre-Devonian basement of Svalbard that are often associated with metamorphic and lithological contrasts (Sandford, Reference Sandford1956; Krasilščhikov, Reference Krasilščhikov1979; Birkenmajer, Reference Birkenmajer1975, Reference Birkenmajer, Nairn, Churkin and Stehli1981, Reference Birkenmajer1992; Björnerud, Reference Björnerud1990). In addition, subsequent isotopic studies provide further evidence that the Caledonian basement of Svalbard includes older Proterozoic crustal domains overprinted by the Early Palaeozoic tectonic events (e.g. Peucat et al. Reference Peucat, Ohta, Gee and Bernard-Griffiths1989; Balashov et al. Reference Balashov, Tebenkov, Ohta, Larionov, Sirostkin, Gannibal and Ryungenen1995; Gee, Björklund & Stølen, Reference Gee, Björklund and Stølen1994; Gee et al. Reference Gee, Johansson, Ohta, Tebenkov, Krasilščhikov, Balashov, Larionov, Gannibal and Ryungenen1995; Johansson et al. Reference Johansson, Gee, Björklund and Witt-Nilsson1995; Hellman et al. Reference Hellman, Gee, Johansson and Witt-Nilsson1997; Johansson et al. Reference Johansson, Larionov, Gee, Ohta, Tebenkov, Sandelin, Gee and Pease2004). Thus a growing body of evidence suggests that some of these Proterozoic terranes were likely subjected to both Grenvillian 950–1000 Ma and latest Neoproterozoic 660–620 Ma tectonothermal activity (Peucat et al. Reference Peucat, Ohta, Gee and Bernard-Griffiths1989) prior to Caledonian overprinting of variable extent.
Geological sketch-map of SW part of Wedel Jarlsberg Land (after Czerny et al. Reference Czerny, Kieres, Manecki, Rajchel and Manecki1993, modified). Inset shows location within the Svalbard Archipelago.
In this study, we dated metamorphic monazite from basement rocks in the southern part of the Wedel Jarlsberg Land, immediately north of the Hornsund Fjord (Fig. 1), to better constrain the pre-Caledonian metamorphic history of the Hecla-Hoek Formation. The study area includes an amphibolite-grade polymetamorphic domain subjected to a complex structural and metamorphic evolution and preserving a long history of early tectonothermal events (Balashov et al. Reference Balashov, Tebenkov, Ohta, Larionov, Sirostkin, Gannibal and Ryungenen1995, Reference Balashov, Tebenkov, Peucat, Ohta, Larionov and Sirostkin1996; Manecki et al. Reference Manecki, Holm, Czerny and Lux1998). The tectonometamorphic history of this area seems to be unique compared to that revealed by lower-grade greenschist-facies rocks representing the Hecla Hoek succession in the directly neighbouring basement domains. The study area thus offers a rare opportunity to unravel the enigmatic Proterozoic events which contributed to the pre-Caledonian development of the crystalline basement of Svalbard.
2. Geological setting
A stratified volcano-sedimentary polymetamorphic complex crops out in the SW part of Wedel Jarlsberg Land. This complex consists of a metasedimentary sequence known as the Isbjørnhamna Group and a metavolcanic succession of the Eimfjellet Group (Birkenmajer, Reference Birkenmajer1958; Czerny et al. Reference Czerny, Kieres, Manecki, Rajchel and Manecki1993). The Isbjørnhamna Group is composed of mica schists, paragneisses, calc-silicate rocks and marbles subjected to two metamorphic events. Amphibolite-grade Barrovian metamorphism of these rocks was followed by a retrogressive event under greenschist-facies conditions (Smulikowski, Reference Smulikowski1965; Majka et al. Reference Majka, Czerny and Manecki2004). The younger metamorphism was responsible for partial to complete chloritization of garnet and biotite, disintegration of muscovite, sericitizaton of plagioclase and decomposition of kyanite.
The Isbjørnhamna Group includes three lithostratigraphic subdivisions: the Skoddefjellet, Ariekammen and Revdalen formations, considered to represent sections of a continuous sedimentary succession (Birkenmajer, Reference Birkenmajer1975). The Skoddefjellet Formation consists of mutually layered (at the centimetre to decimetre scale) paragneisses and metapelites. A typical mineral assemblage related to progression of metamorphism in these rocks consists of Q + Pl + Bt + Ms ± Grt ± Chl with increasing amounts of plagioclase (oligoclase) and garnet in paragneisses and mica schists, respectively. Tourmaline, sphene, apatite, allanite, monazite, xenotime, zircon, ilmenite, hematite and magnetite are common accessory phases.
The Ariekammen Formation is distinguished by the presence of carbonate rocks. The formation mostly consists of carbonate–mica schists with irregular layers of mica schists and paragneisses of the Skoddefjellet type. A mineral assemblage characteristic of these rocks comprises C + Q + Bt + Grt + Pl + Ms ± Ep with Ca-enriched plagioclase and the same accessory minerals as those in the Skoddefjellet Formation. In addition, some varieties of schists contain rare mejonite. Notably, a discontinuous horizon of yellow and white calcite marbles occurs in the middle of the formation. Porphyroblasts of garnet up to 6 cm across are abundant at the base of this horizon.
The Revdalen Formation in the uppermost part of the Isbjørnhamna Group represents a uniform sequence of rusty weathered mica schists. These rocks consist of Q + Pl + Bt ± Ms ± Grt ± Chl and accessory phases similar to those found in the Skoddefjellet Formation. Metamorphic zonation is indicated by the local presence of chloritoid, staurolite or kyanite.
In the Isbjørnhamna Group metapelites, the dominant planar structure is a pervasive S1 foliation containing a strong visible L1 lineation. These structures are interpreted to have formed during the older M1 amphibolite-facies metamorphism, as younger, greenschist-facies Caledonian metamorphic overprinting was not associated with a strong deformational overprint of the metapelites. S1 foliation is predominantly expressed by flat, parallel arrangement of muscovite and biotite phyllosilicates and by flattened pophyroblasts of garnet and staurolite. However, in samples where chloritoid is present in the paragenesis, dominant phyllosilicates are muscovite and chlorite. In such samples, prismatic porphyroblasts of chloritoid also lay within the foliation planes.
The contact of the Isbjørnhamna Group with the adjacent greenschist-facies Deilegga and Sofiebogen groups to north is clearly tectonic. A 0.5 km thick high-strain zone making up the strike-slip to oblique-slip sinistral Vimsodden–Kosibapasset shear zone developed under greenschist-facies conditions. The Vimsodden–Kosibapasset shear zone is located directly at the contact of two crustal domains which differ in both metamorphic grade and structural evolution. The northern domain, comprising the Deilegga and Sofiebogen groups, resembles the Vimsodden–Kosibapasset zone in terms of metamorphic grade and structural pattern despite the much lower strain intensity. On the other hand, the margin of the southern domain, corresponding to the Isbjørnhamna Group, shows features of intense mylonitization and retrogression providing evidence for its post-peak metamorphism juxtaposition against the lower-grade crustal blocks to the north.
Initial thermochronological data from the study area were reported by Gayer et al. (Reference Gayer, Gee, Harland, Miller, Spall, Wallis and Winsnes1966), who obtained K–Ar mica ages of 565 and 595 Ma. Subsequent Rb–Sr whole-rock dating of samples from the Isbjørnhamna Group yielded a late Grenvillian age of 936 Ma (Gavrilenko et al. Reference Gavrilenko, Balashov, Tebenkov and Larionov1993). Detrital zircons from the mica schists belonging to the same group dated by the Pb-evaporation method on small populations of grains gave a poorly constrained age of c. 1500 Ma (Balashov et al. Reference Balashov, Tebenkov, Peucat, Ohta, Larionov and Sirostkin1996). The meta-igneous suite of mafic and felsic rocks of the overlying Eimfjellet Group yielded a consistent group of single-zircon Pb-evaporation ages of c. 1160 ± 40 Ma (Balashov et al. Reference Balashov, Tebenkov, Peucat, Ohta, Larionov and Sirostkin1996). Rhyolitic metaconglomerates belonging to the same suite yielded c. 1200 Ma U–Pb detrital zircon ages and a lower intercept age of c. 930 Ma, thought to be the time of regional metamorphism (Balashov et al. Reference Balashov, Tebenkov, Ohta, Larionov, Sirostkin, Gannibal and Ryungenen1995). 40Ar/39Ar step heating of mineral separates yielded a 616 Ma age for hornblende from the Eimfjellt Group and 585–575 Ma ages for biotite and muscovite from the Isbjørnhamna Group (Manecki et al. Reference Manecki, Holm, Czerny and Lux1998).
3. Analytical methods and sampling selection
In order to constrain the timing of amphibolite-facies metamorphism, in situ electron microprobe total-Pb monazite geochronometry was applied to selected metamorphic samples. Monazite was principally chosen because it contains large amounts of Th and U, has minor 204Pb, and exhibits little elemental diffusion under high temperatures (Catlos, Gilley & Harrison, Reference Catlos, Gilley and Harrison2002). Furthermore, monazite is an ideal mineral for dating polyphase tectonometamorphic histories because of its high closure temperature (> 850 °C) and its ability to record multiple metamorphic events (Cherniak et al. Reference Cherniak, Watson, Grove and Harrison2004). Five samples of fine- to medium-grained mica schists from the Skoddefjellet and Revdalen formations were selected for monazite dating (Fig. 2). These rocks underwent amphibolite-facies M1 metamorphism followed by a minor M2 retrogressive event that brought about relatively weak retrogressive changes (e.g. partial chloritization of garnet and biotite).
Sample location within the Isbjørnhamna Group.
Only samples from staurolite and kyanite zones were selected for monazite dating. These samples are metamorphosed at P–T conditions above the so-called ‘allanite window’ (e.g. Wing, Ferry & Harrison, Reference Wing, Ferry and Harrison2003; Janots et al. Reference Janots, Brunet, Goffe, Poinssot, Byrchard and Cemic2007; Krenn & Finger, Reference Krenn and Finger2006) making them ideally suited for determining peak metamorphic age. In samples metamorphosed below the staurolite zone, only allanite was observed in the paragenesis. According to Ferry (Reference Ferry2000), detrital monazite can entirely disappear during metamorphism and the rare-earth elements originally comprised in its grains be preferentially sited in allanite. Metamorphic allanite is stable only under upper-greenschist-facies conditions above the Bt-in isograd, whereas the crystallization of new metamorphic monazite is related to prograde amphibolite-facies metamorphism (Ferry, Reference Ferry2000; Catlos, Gilley & Harrison, Reference Catlos, Gilley and Harrison2002; Wing, Ferry & Harrison, Reference Wing, Ferry and Harrison2003; Gieré & Sorensen, Reference Gieré and Sorensen2004). We emphasize that the amphibolite-facies metamorphic rocks (conditions quantified using Gr–Bt thermometry and GASP barometry: Majka, Czerny & Manecki, Reference Majka, Czerny and Manecki2004) analysed in this study contain monazite but no allanite. Therefore, all dated monazites are interpreted as metamorphic, based on their morphology and internal structure. Detrital monazite, if originally present, likely underwent complete dissolution or recrystallization during M1 metamorphism and actively participated in prograde mineral reactions.
In the studied rocks, monazites occur in micaceous S1 foliation planes and as inclusions in garnet and staurolite porphyroblasts which are synkinematic with the foliation described above (Fig. 3). Monazites from subhedral to anhedral blasts do not exceed about 70 μm in length, and are locally surrounded by aggregates of apatite or apatite–allanite coronas (Majka & Budzyń, Reference Majka and Budzyń2006). Some monazite grains reveal a ‘swiss cheese’-like internal structure characteristic of metamorphic growth. Most of the analysed monazites do not reveal any zoning in BSE imaging. Only sporadically tiny, patchy zoning was observed. Both structural position of analysed monazites and pervasive lack of zoning lead us to conclude that all the monazites likely grew during a single-stage metamorphic event.
(a) Typical subhedral monazite. (b) Typical anhedral monazite. (c) Anhedral monazite with sectoral zoning. (d) Monazite grains enclosed in garnet porphyroblast.
In situ analyses were made of polished thin-sections using the Cameca SX-100 electron microprobe at the Electron Microanalysis Department of the Geological Survey of Slovak Republic in Bratislava. Details of analytical methodology and recalculations are as described by Konečný et al. (Reference Konečný, Siman, Holicky, Janak and Kollarova2004). The age calculation is based on the formulation of Montel et al. (Reference Montel, Foret, Veschambre, Nicollet and Provost1996), which is considered as effective and satisfactory when a single age population of homogeneous monazites is analysed (Williams et al. Reference Williams, Jercinovic, Gonclaves and Mahan2006). Briefly, the model ages are calculated for each analysis and presented in the form of a histogram. From all results, the model weighted average and standard deviation are calculated according to the statistical procedure described by Montel et al. (Reference Montel, Foret, Veschambre, Nicollet and Provost1996). The age is calculated using the Microsoft Excel add-in program DAMON that reads the data, calculates the model and weighted averaged ages, and constructs the histograms and isochrons (Konečný et al. Reference Konečný, Siman, Holicky, Janak and Kollarova2004). The isochrons are not used to determine the age. The fact that all results plot on the very same isochron is additional evidence that supports our interpretation that the population of dated monazites is homogeneous and formed during a single metamorphic event.
The EMP analyses were performed using 100 nA beam current and 15kV accelerating voltage. The beam diameter varied from 1 to 3 μm. Background level was determined using linear fit. The counting time (peak + background) for Si, Al, Ca, P and As was 20 seconds, for REE 25 seconds, for Th and Y 35 seconds, for U 65 seconds and for Pb 150 seconds. The following standards were used for analysed elements: Si – wollastonite, Al – Al2O3, Ca – wollastonite, Pb – PbS, Th – ThO2, U – UO2, P – apatite, As – GaAs2, REE and Y – REE and Y phosphates. Si, Al and As were measured with the use of a TAP crystal; Ca, Pb, U, Th, Y and P were measured with the use of a LPET crystal, and REE with the use of a LLIF crystal. For determination of the content of Si, Al, Ca, P, the Kα line was measured, for La, Ce, Gd, Tb, Tm, Yb, Y, As the Lα line was measured, for Pr, Nd, Sm, Eu, Dy, Ho, Er, Lu the Lβ line was measured, for Pb and Th the Mα line was measured, and for U the Mβ line was measured. ZAF corrections were applied throughout. All errors are reported, depicted and discussed in this paper at the 2σ level (95% confidence limits).
4. Results
Chemical U–Th–total Pb dating performed on all monazite grains (n=61; table of supplementary material available online at http://journals.cambrige.org/geo) span a 130 Ma interval between 580 and 710 Ma (Fig. 4), with an average statistical uncertainty (2σ) of approximately ±36 Ma for a single age determination. The distribution of analytical results on the isochron indicates that all dated monazites belong to the same population (Fig. 5), with the weighted average age being 643 ± 9 Ma. There is no age difference between matrix monazite grains and those forming inclusions in garnet. A uniform age is also characteristic for discrete domains within individual monazite grains revealed with BSE imaging and monazites showing secondary alterations apparent as allanite–apatite coronas (Majka & Budzyń, Reference Majka and Budzyń2006).
Histogram of monazite ages from the Isbjørhamna Group.
Isochron of monazite age from the Isbjørnhamna Group.
5. Discussion
The age of 643 ± 9 Ma obtained for metamorphic monazite from the Isbjørnhamna Group provides the first direct evidence for late Neoproterozoic metamorphism of the Hecla Hoek succession in SW Spitsbergen. Previous geochronological results of K–Ar dating, reported by Gayer et al. (Reference Gayer, Gee, Harland, Miller, Spall, Wallis and Winsnes1966) and Ar–Ar dating of micas and hornblende (Manecki et al. Reference Manecki, Holm, Czerny and Lux1998), indicated a late Neoproterozoic cooling within the same time span. We consider the fact that monazite grains enclosed in garnet yield the same age as matrix monazite grains as strong evidence for a progressive Barrovian metamorphic event having taken place around 640 Ma.
Zircon ages reported by Gavrilenko et al. (Reference Gavrilenko, Balashov, Tebenkov and Larionov1993) and Balashov et al. (Reference Balashov, Tebenkov, Ohta, Larionov, Sirostkin, Gannibal and Ryungenen1995, Reference Balashov, Tebenkov, Peucat, Ohta, Larionov and Sirostkin1996) potentially reveal an older Grenvillian metamorphic event, probably under sub-amphibolite-facies conditions, which took place in the southern part of Wedel Jarlsberg Land. However, no structural or mineral relicts of this event are preserved in rocks of the Isbjørnhamna Group. Essentially, the c. 930 Ma zircon age of the metarhyolites from the Eimfjellet Group is based on a discordia lower intercept (Balashov et al. Reference Balashov, Tebenkov, Ohta, Larionov, Sirostkin, Gannibal and Ryungenen1995) and thus remains poorly constrained. The earlier determined Rb/Sr ages of the Isbjørnhamna rocks (Gavrilenko et al. Reference Gavrilenko, Balashov, Tebenkov and Larionov1993) are based on a small pre-selected population of samples. Furthermore, the Rb–Sr method itself is not fully reliable in the case of rocks subjected to a prolonged polyphase metamorphic evolution.
The c. 640 Ma peak metamorphic age of the Isbjørnhamna Group compares well with the ages of igneous and high-pressure metamorphic events in the Richarddalen Complex of NW Spitsbergen dated at c. 660 and 620 Ma, respectively (Peucat et al. Reference Peucat, Ohta, Gee and Bernard-Griffiths1989). These ages were interpreted to represent the time of magmatic crystallization of a felsic intrusive phase and probably of high-grade (eclogite) metamorphism (Peucat et al. Reference Peucat, Ohta, Gee and Bernard-Griffiths1989). Thus, they were considered as evidence of a significant high-pressure tectonothermal episode in latest Neoproterozoic times. This interpretation received further support from 540–500 Ma Ar/Ar cooling ages for the Richarddalen Complex (Dallmeyer, Peucat & Ohta, Reference Dallmeyer, Peucat and Ohta1990). Nevertheless, the age of high-pressure metamorphism of the Richarddalen eclogites was subsequently reinterpreted as Early to Middle Ordovician and correlated with the Caledonian collisional cycle (Gromet & Gee, Reference Gromet and Gee1998). Accepting the original interpretation of Peucat et al. (Reference Peucat, Ohta, Gee and Bernard-Griffiths1989), our new monazite ages from the Isbjørnhamna Group convincingly support the connection between SW Spitsbergen and NW Spitsbergen recently postulated by Gee & Tebenkov (Reference Gee, Tebenkov, Gee and Pease2004). Although not a favoured model in their interpretation, we note that Gromet & Gee (Reference Gromet and Gee1998) did allow the possibility of two tectonically distinct metamorphic events, a latest Neoproterozoic eclogite facies event followed by a pervasive Ordovician event at mid-amphibolite grade.
Throughout the Neoproterozoic and early Palaeozoic, stratigraphic successions in Svalbard, East Greenland and Scandinavia indicate essentially continuous deposition on a rifted continental margin (e.g. Flood et al. Reference Flood, Gee, Hjelle, Siggerud and Winsnes1969; Kumpulainen & Nystuen, Reference Kumpulainen, Nystuen, Gee and Sturt1985; Henriksen, Reference Henriksen, Gee and Sturt1985). Significantly, the Laurentian Cambro-Ordovician faunal provinciality of Svalbard and East Greenland is in clear contrast with that of Scandinavia. Thus, rifting of a Neoproterozoic supercontinent and the independent development of Laurentian and Baltic passive margins appear substantiated (Gromet & Gee, Reference Gromet and Gee1998). At the same time, Neoproterozoic metamorphism and/or granitoid magmatism are practically unknown from the Scandinavian Caledonides and the underlying basement of the Baltic shield (Gorbatschev, Reference Gorbatschev, Gee and Sturt1985), as well as from East Greenland (e.g. Henriksen, Reference Henriksen, Gee and Sturt1985) and eastern Svalbard (e.g. Gee & Tebenkov, Reference Gee, Tebenkov, Gee and Pease2004). Therefore, if c. 640 Ma metamorphic ages from the Isbjørnhamna Group represent a phase of orogenic development in the Neoproterozoic, they are difficult to reconcile with a quiescent non-tectonic regime inferred from the occurrence of rift-drift sequences on the Baltic and Laurentian passive margins. Instead, our new monazite ages imply an exotic origin of the pre-Devonian basement exposed in SW Spitsbergen and support models of terrane assembly postulated for the Svalbard Archipelago (e.g. Harland & Gayer, Reference Harland and Gayer1972; Harland, Reference Harland, Gee and Sturt1985; Ohta, Dallmeyer & Peucat, Reference Ohta, Dallmeyer and Peucat1989; Gee & Page, Reference Gee and Page1994).
More work needs to be done to better elucidate the tectonothermal evolution of the study area and to better understand its potential origin as an exotic terrane. Speculatively, we note that the time span of 580–710 Ma overlaps ages of many of the Pan-African orogenic belts widespread across the Gondwana supercontinent. However, a Gondwanan derivation of the SW part of Spitsbergen is inconsistent with recent plate reconstructions for the Neoproterozoic, which place Gondwana opposite to the Laurentian passive margin (e.g. Hoffman, Reference Hoffman1991; Torsvik et al. Reference Torsvik, Smethurst, Meert, Van Der Voo, McKerrow, Brasier, Sturt and Walderhaug1996; Dalziel, Reference Dalziel1997). Instead, a more likely solution seems to be the correlation of SW Spitsbergen with the Timanide orogen of the North Urals, which also records Neoproterozoic tectono-magmatic activity (e.g. Kuznetsov et al. Reference Kuznetsov, Soboleva, Udoratina, Hertseva and Andreichev2007). Nevertheless, such speculation contradicts the Laurentian affinity characteristic of the majority of the pre-Devonian basement of Svalbard (with important exceptions of foreign terranes along the west coast of central Spitsbergen) and the postulated large-scale separation between Laurentia and Baltica in the late Neoproterozoic and early Palaeozoic (e.g. Gee & Tebenkov, Reference Gee, Tebenkov, Gee and Pease2004). Therefore, our results call for an essential reconsideration of large-scale tectonic models explaining the assembly of Svalbard terranes.
6. Conclusions
The U–Th–total Pb monazite dating presented here allows the following interpretation of the Isbjørnhamna Group from the southern part of the Wedel Jarlsberg Land north of the Honsund Fjord:
(1) Monazite U–Th–total Pb ages reported herein unequivocally indicate, in line with previous K/Ar and Ar/Ar data (Gayer et al. Reference Gayer, Gee, Harland, Miller, Spall, Wallis and Winsnes1966; Manecki et al. Reference Manecki, Holm, Czerny and Lux1998), that peak amphibolite-facies conditions of a progressive Barrovian-type metamorphic event occurred around 640 Ma.
(2) A preceding Grenvillian metamorphic event, if it occurred, most likely took place under sub-amphibolite-facies conditions.
(3) Later greenschist-facies alteration reported by Manecki et al. (Reference Manecki, Holm, Czerny and Lux1998) can be attributed to the effects of Caledonian tectonism.
(4) Wedel Jarlsberg Land comprises a Proterozoic basement block consisting of a lithostratigraphic profile and metamorphic history unique at the scale of the Svalbard Archipelago. A record of Neoproterozoic orogenic development may suggest an affinity to either Pan-African or Timanide orogens.
(5) The unique characteristics of this basement block confirm its exotic provenance and support terrane models postulated for the Svalbard Archipelago.
Acknowledgements
This research was funded by MEiN research grant no. 2P04D 039 30. The table of supplementary material available online at http://journals.cambrige.org/geo.
