Introduction

The Eckergneiss Complex (EGC) in the Harz Mountains (Germany) is a high-grade polymetamorphic crustal segment within the Rhenohercynian domain of the Mid-European Variscides (Fig. 1). The high-grade metamorphism of the EGC is a matchless feature within the low-grade Rhenohercynian domain, raising questions about its geological history, which, in turn is important to our understanding of the geodynamic evolution of the Variscan orogenic belt. A conventional lower intercept U–Pb age of about 560 Ma from detrital zircon grains, interpreted to reflect the timing of the main metamorphism (Bauman et al. 1991), suggests that the EGC represents the oldest geological unit of the Harz Mountains. This age has led to controversial speculations about the tectonic setting of the EGC. Bankwitz (1995) suggested that the EGC represents a Cadomian–Caledonian tectonic flake formed during the collision of Gondwana with East Avalonia. Other authors favored a model that explains the EGC as being uplifted upon the intrusive ascent of the magmas of the adjacent Brocken Granite and the Harzburg Gabbronorite (Baumann et al. 1991; Franz et al. 1997). Based on the “Cadomian” (Pan-African) age Franz et al. (1997) further speculated that the EGC belongs to the East Avalonian terrane. Whereas these models agree in assuming that the EGC is an allochthonous enclave within the Rhenohercynian domain, Franzke (2001) has suggested that the EGC may represent a semi-autochthonous outcrop topping a crystalline high located between the Oberharzer Diabaszug (Upper Harz diabase belt) and the Blankenburger Faltenzone (Blankenburg fold zone; Fig. 1b).

Fig. 1
figure 1

a Subdivision of the Mid-European Variscides after Franke (1989). b Geological subdivision of the Upper and Middle Harz (after Mohr 1978): HFZ Harzgeröder fold zone, TGZ Tanne greywacke zone, BFZ Blankenburger fold zone, SiS Sieber syncline, SM South Harz syncline, B Brocken Granite, E Eckergneiss Complex, H Harzburg Gabbronorite, OK Oker Granite, ABZ Acker–Bruchberg zone, ODZ Upper Harz diabase range, SöS Söse syncline, CCF Clausthaler Culm fold zone, ODS Devonian anticline of the Upper Harz. c Simplified geological map of the Eckergneiss Complex with quartzite occurrences and sample localities

Full size image

The present study was initially based on the premise of a “Cadomian” (Pan-African) metamorphic age. The original objectives were to identify the assumed pre-Cadomian provenance of the sedimentary Eckergneiss protolith assemblage, to constrain its deposition age, and, accordingly, the geotectonic position of the EGC by characterizing the age structure of the detrital zircon component of five quartzite occurrences. Zircon is of particular value for provenance studies, since it is chemically and physically resistant and not liable to destruction during the weathering and sedimentary cycle. Furthermore, zircon forms predominantly in felsic-intermediate igneous and metamorphic rocks, and thus records major crust-forming events that characterize potential source areas. However, zircon crystals that have accumulated a certain amount of self-irradiation structural damage or those that are rich in rare earth elements are prone to react with geological fluids, resulting in inward penetrating alteration or recrystallization zones that are usually characterized by a disturbed U–(Th)–Pb system (e.g., Black et al. 1986; Geisler et al. 2001, 2002, 2003a, 2003b, 2003c; Schaltegger et al. 1999; Tomaschek et al. 2003; Vavra et al. 1996, 1999). In addition, more than one growth generation may be preserved in single zircon grains (e.g., Black et al. 1986; Geisler and Schleicher 2000). Based on these grounds, multi-phase zircon crystals are common in metamorphic terranes. We note here that Anthes and Reischmann (2001) mentioned that their “efforts to date the formation age of the EGC or its source using single zircon analyses failed so far”. In consequence, imaging techniques and microanalytical tools must be applied to unravel the internal structures and their age relationships, to obtain geologically interpretable age data. In the present study, we used cathodoluminescence imaging to visualize the internal structures, which were dated in situ by the sensitive high-resolution ion microprobe (SHRIMP) technique. By doing this, we also hoped to be able to date the metamorphic events within the EGC.

Regional setting and basic petrographical features

The EGC is located within the Harz Mountains in central Germany (Fig. 1), which represent a tectonically uplifted segment of the Variscan orogenic belt of central Europe. The Variscan belt is a very complex conglomeration of small crustal blocks of different P–T–t histories that were juxtaposed along strike-slip and detachment systems. The belt is commonly divided into four tectono-stratigraphic units from NW to SE (e.g., Franke 1989): the Rhenohercynian Zone, Mid-German Crystalline Rise (MGCR), the Saxothuringian Zone, and the Moldanubian Zone (Fig. 1a). The Harz Mountains belong to the Rhenohercynian domain and reflect a long evolution of sedimentary, erosional, tectonic, and magmatic processes. The Harz essentially comprises diagenetic and low-grade metamorphosed sedimentary rocks (Friedel et al. 1995) of Ordovician to Upper Carboniferous age. Major basaltic (spilitic) intercalations occur in Middle Devonian and Lower Carboniferous shales and slates, commonly as piles of pillow lavas. Granitoid and gabbroic plutonic rocks form ca. 295-Ma old (Baumann et al. 1991), high-level intrusions within Devonian and Lower Carboniferous country rocks. Significant metamorphic imprint is confined to thermal aureoles around these intrusions.

In Devonian times, the palaeogeographic position of the present Harz area was determined by its affiliation to the Rhenohercynian domain between the Old Red continent (Laurussia) to the north and the MGCR to the southeast. At that time, pelitic, psammitic, and carbonate sediments were deposited, now represented by slates, sandstones, and limestones, which are mainly exposed in the Oberharzer Devonsattel (Upper Harz Devonian Anticline). The Lower Carboniferous constitutes most of the western and central Harz being represented by flysch series (turbiditic graywackes and shales, Fig. 1b) and the Acker–Bruchberg–quartzite (ABQ) (quartzitic sandstone, Acker–Bruchberg zone in Fig. 1b). The sedimentary strata were tectonically overprinted in the course of the last compressional (Asturian) phase of the Variscan orogenic cycle about 300 Ma ago (Mohr 1978).

The EGC represents a 7-km2 large crustal segment that is enclosed by the high level intrusions of the Harzburg Gabbronorite and the Brocken Granite (Fig. 1c). The complex consists of pelitic and siliciclastic metasediments (e.g., quartzites, quartz-(cordierite)-feldspar granofelses, gneisses, hornfelses, and mica schists) with subordinate metabasaltic intercalations. The meta-sedimentary character of almost the entire EGC is indicated, for example, by a psammitic-pelitic, virtually Ca-devoid geochemistry and rare examples of preserved sedimentary depositional structures. These are thin, placer-like laminae of concentrated heavy minerals and cross bedding (Schlüter 1983, unpublished doctoral thesis).

Erdmannsdörffer (1909) has interpreted the EGC as representing a metamorphic equivalent of the ABQ, which he thought was thermally affected by the intrusions of the Brocken Granite and the Harzburg Gabbronorite. However, petrographic and petrological investigations revealed that the EGC experienced multiple metamorphic events, ranging from contact metamorphic to amphibolite and granulite-facies conditions (Franz et al. 1997; Martin-Gombojav 2003, unpublished doctoral thesis; Müller and Strauss 1985; Schlüter 1983, unpublished doctoral thesis). Franz et al. (1997) determined amphibolite- and granulite facies conditions of 560–650°C and P> 0.5 GPa and 720–780°C and 0.67–83 GPa, respectively.

Whereas, the contact metamorphic event can be temporally correlated with the intrusion of the neighboring Harzburg Gabbronorite and Brocken Granite 293–297 Ma ago (Baumann et al. 1991), right at the Carboniferous/Permian border, the regional metamorphic events have not yet confidently been dated by conventional dating methods. Schoell et al. (1973) obtained a Rb–Sr whole rock isochron age of 392±10 Ma Footnote 1 from various metasedimentary lithologies scattered over much of the EGC outcrop. They interpreted this datum to reflect the main metamorphism of the EGC. However, the geological significance of such an isochron remains doubtful. The same doubt applies to the lower intercept U–Pb age of about 560 Ma of Baumann et al. (1991) from euhedral and rounded detrital zircons from metasedimentary rocks within the EGC.

Franz et al. (1997) postulated that the amphibolite facies mineral assemblages were formed during a completely new metamorphic loop. However, we would like to note here that new petrographical observations from various localities of the EGC outcrop (Martin-Gombojav 2003, unpublished doctoral thesis) point to a sequence of metamorphic events in an opposite order than suggested by Franz et al. (1997). Significant in this context are the following observations: (a) poikiloblastic orthopyroxene, which is unaffected by foliation, (b) undeformed symplectites of orthopyroxene and cordierite, obviously at the expense of pyralspite garnet, and (c) no significant alteration or replacement of most of the orthopyroxenes. These features indicate that re-equilibration at granulite-facies temperatures was essentially static and followed earlier penetrative deformation, which seems to have taken place at amphibolite facies temperatures. Amphibolite-facies conditions of 650–700°C and about 0.4 GPa, which were estimated by Schlüter (1983, unpublished doctoral thesis) from kinzingites of the EGC, are in full agreement with the new petrographical observations. Additional indications of dynamic metamorphism under amphibolite-facies conditions are porphyroclasts of critical minerals such as garnet, andalusite, and cordierite (pinite) with a fine-grained foliation wrapped around them, but obviously never so with orthopyroxene. The orthopyroxene–cordierite symplectites as listed under (b) indicate an episode of postdeformational granulite-facies metamorphism. They occur in close proximity to essentially unretrograded, blocky orthopyroxene porphyroblasts, a few of which are bent by tectonic (or intrusive?) pressure.

Sample description

For the present investigation, we have chosen five quartzite samples from different localities (Fig. 1a). Samples 99-20, 0-12, and 0-100 were sampled from the east shore of the Eckertal Stausee, from the Spörenwagen, and from the south Kolför locality, respectively. Sample 99-38 was collected from the Unterer Lobenklee area, which had been mapped by Erdmannsdörfer (1909) as belonging to the ABQ formation. Recent field work and petrographic investigations, however, have shown that the rocks of this locality exhibit high-grade metamorphism and thus clearly belong to the EGC (Martin-Gombojav 2003, unpublished doctoral thesis). Sample 0-40 was sampled from a locality in the Kleines Gierstal.

Some quartzite samples contain significant amounts of K-feldspar and plagioclase. Subordinate components are pinite, chlorite, muscovite, and a mature detrital heavy mineral population of zircon, chromite, garnet and rutile. Monazite was only found in sample 0-12. Sample 0-100 also contains newly formed tourmaline (enclosing ilmenite grains) and titanite. The detrital accessory minerals, which necessarily were deposited as independent grains, are commonly enclosed in random distribution in quartz. This feature indicates a high-metamorphic grade and clearly distinguishes the Eckergneiss quartzites from the quartzitic sandstones of the ABQ formation (Fig. 2). We also note that the detrital chrome spinel grains of all five EGC quartzite samples, analyzed by Martin-Gombojav (2003, unpublished doctoral thesis), have unusually low-Mg/(Mg +Fe2+) values due to low-Mg concentrations. Such unusual chemical composition further discriminates the EGC quartzites not only from the quartzitic sandstones of the ABQ and from Devonian to Carboniferous sandstones and greywackes of the Harz (Fig. 3), but also from Early to Middle Devonian sandstones of the Rhenish Massif (Press 1986). The ABQ, however, also contains low-Mg chromite grains as a subordinate fraction among more abundant chromites with normal Mg-contents (Fig. 3). This indicates that the low-Mg content in the EGC chromites is inherited from a distinct though unknown source, which also contributed to the ABQ rather than being the result of processes after deposition within the EGC protolith. The metamorphism of the EGC, as might be suspected, cannot be responsible for the Mg-deficit. This conclusion is consistent with missing internal, textural features of the chromites that would indicate metamorphic alteration (Tesalina et al. 2003, and references therein).

Fig. 2
figure 2

Photomicrographs with cross-polarizer of a,b a quartzite (sample 99-38) from the Unterer Lobenklee locality of the Eckergneiss Complex and c, d of an Acker–Bruchberg quartzite from the Ilsetal. Note that the detrital zircon crystals (marked by white arrows) in the Ecker quartzite are enclosed in the quartz grains, whereas, they are located along interstitial sites in the Acker–Bruchberg sample. Scale bars denote 500 μm

Full size image
Fig. 3
figure 3

Diagram Cr/(Cr+Al) versus Mg/(Mg+Fe2+) showing the unique chemical composition of detrital chrome spinels from the Eckergneiss quartzites (data from Martin-Gombojav), when compared to the composition of detrital chromites from Devonian to Carboniferous clastic strata of the Harz Mountains (data from Ganssloser 1999; own unpublished data). Fields for chrome spinels from different geodynamic settings are from Irvine (1967) and Dick and Bullen (1984)

Full size image

Most of the quartzite samples of the EGC exhibit a granoblastic texture. The quartz grains are only slightly interlocked. They rarely show undulatory extinction. Some of the quartzites contain interstitial feldspar with a tendency towards poikiloblastic growth between the quartz grains. Only sample 99-20 exhibits flattened platy quartz grains with interlocked grain boundaries. This sample, as well as sample 0-12, is also foliated. The foliation in these samples is macroscopically accentuated by thin pinite-chlorite bands.

Analytical method

Ninety SHRIMP measurements on 76 zircons, separated from the five quartzite samples using heavy liquids and a magnetic separator (Fig. 1), were carried out during two 24-h analytical sessions at the WA Consortium SHRIMP II at the Curtin University of Technology in Perth (Appendix), employing the SHRIMP technique as described in DeLaeter and Kennedy (1998). The primary O −2 beam was focused to a 25 to 40-μm diameter area onto the target zircon. The mass resolution was better than 5,200 in both sessions, which was sufficient to resolve isobaric interferences. Measured U–Pb ratios were corrected following Claoué-Long et al. (1995) from eight to ten regular analyses of the standard zircon CZ3 (Pidgeon et al. 1994) over each 24-h analytical session, resulting in 206Pb/238U calibration errors of 3.5% and 1.6 %, respectively. The isotopic ratios were further corrected for common lead using Broken Hill Pb composition and the 204Pb correction method (Compston et al. 1984). Counting times were reduced from normal SHRIMP practice to achieve rapid acquisition of data at the expense of some measurement precision. Fifty nine detrital grains were measured after random sampling. The measurements on alteration or recrystallization rims (Fig. 5), however, were preselected and were acquired using the normal SHRIMP procedure. This was done because we initially hoped to be able to date the metamorphism of the EGC by analyzing these zones.

In order to obtain a best age estimate, the Concordia age calculation procedure of Ludwig (1998) was applied, which is based on the two U–Pb isotope ratios and their correlation. Only those data with U–Pb ages that overlap the Concordia with their 2σ error ellipse were used to construct the probability plot shown in Fig. 4. However, in those cases, where we obtained more than a single discordant measurement on a single grain, we used the upper or, in one case, the lower discordia intercept age (see Appendix). Ages and errors were calculated using the Isoplot software of Ludwig (2001). All reported errors represent the 1σ a priori error. Decay constants used are those recommended by IUGS (Steiger and Jäger 1977). The new timescale of Gradstein et al. (2004) was consistently applied throughout the text.

Fig. 4
figure 4

Cumulative probability curves of concordant SHRIMP ages (see Appendix; Ludwig 1998) from detrital zircons of a sample 99-20, b sample 99-38, and c of all five samples from the Eckergneiss Complex. Age boundaries are according to Gradstein et al. (2004)

Full size image

Results

In Fig. 4, we have plotted the cumulative probability plots for two quartzite samples from different localities and for all data. It is evident from all three probability plots that the age population is strongly dominated by Proterozoic ages, ranging from about 900 Ma to about 1800 Ma. The limited amount of data collected from zircon crystals of the individual samples (Fig. 4a, b), however, does not allow us to interpret the differences seen in the fine structure of the Proterozoic age population. Archean age components were not found in any of the samples. An unexpected result is that three zircon crystals from two quartzite localities (samples 99-38 and 0-12) yielded statistically indistinguishable Lower Devonian and Silurian ages of 410±10, 419±10, and 436±5 Ma (Appendix). A single grain with an age of 566±6 Ma was also found in sample 99-38. It should also be noted that we have conducted two SHRIMP analyses (core–rim) on each of these grains, yielding ages that are also statistically indistinguishable from each other (Appendix).

Representative cathodoluminescence (CL) images of the investigated detrital zircon crystals are shown in Fig. 5. Based on the internal growth structure, visualized by the CL images, the zircon population may be roughly divided into the following groups: (1) zircon grains with simple oscillatory growth zones (Fig. 5a–c), (2) zircon crystals with bright-CL zones penetrating into older zircon cores (Fig.5d, f), (3) grains with newly grown rims (Fig. 5i), (4) sector-zoned zircon crystals (Fig. 5g), (5) unusual zircon grains showing curved internal structures (Fig. 5h), and (6) relatively homogeneous zircon grains (Fig. 5j). However, we should note that some grains are even more complex and, when considering the internal structure, represent rather mixtures of these groups.

Fig. 5
figure 5

a–j Cathodoluminescence images of representative detrital zircon crystals from the Eckergneiss Complex. SHRIMP spot localities are marked by ellipses. The corresponding 206Pb/238U age is also given. SHRIMP data points 1, 2, and 3 shown in f are discussed in the text

Full size image

First, we note that the Lower Devonian and Silurian crystals are elongated and exhibit simple oscillatory zoning (Fig. 5a, b). These features clearly point to a magmatic origin. Crystals grown during high-grade metamorphic events are usually multi-facetted with an oval shape (e.g., Grauert and Wagner 1975; Kröner et al. 1994). The occurrence of complex internal structures of older zircon crystals, such as those shown in Fig. 5f – j, indicates that these grains underwent a metamorphic imprint involving fluids (e.g., Geisler et al. 2001, 2003a, 2003b, 2003c; Schaltegger et al. 1999; Tomaschek et al. 2003; Vavra et al. 1996, 1999). We note that inwards penetrating recrystallization zones characterized by a bright CL intensity have been produced by hydrothermal treatment of partially metamict zircon samples at temperatures as low as 175°C (Geisler et al. 2001, 2002, 2003a, 2003b, 2003c). In old zircons, bright-CL zones are mainly characterized by a lesser degree of self-irradiation damage or a higher degree of recrystallization, respectively (unpublished Raman spectroscopic data; see also Geisler& Pidgeon 2001). Also, nonequilibrium, secondary structures, such as those seen in Fig. 5h, have recently been observed after treating an almost fully metamict zircon in hydrothermal solutions (Geisler et al. 2004).

As mentioned in the previous section, we originally hoped that some of these complex zircon crystals would yield ages that could be correlated with the inferred metamorphic evolution of the EGC. However, during the course of the work, it became evident that most of these zircon crystals yield Sveconorwegian (Grenvillian) ages, ranging from ca. 980 to 1,200 Ma (Fig. 4), clearly revealing that the ages reflect metamorphic events in the zircon source region rather than in the EGC. Grains outside this age range usually exhibit simple oscillatory zoning typical for magmatic zircon (Fig. 5c–e). Only three analyses of fluid-induced alteration rims around zircon Z12 (Fig. 5f) yielded ages younger than 900 Ma (see analyses Z12-1/2/3 in Appendix). Although the 2σ error ellipses of these analyses overlap or nearly overlap with the Concordia, significant higher 207Pb/206Pb ages clearly indicate that the data points are discordant. The three analyses fall on a reference line between an upper and lower Concordia age of about 960 and 440 Ma, respectively.

Discussion

Time of sedimentation of the Eckergneiss detritus

As a guide to the following discussion, we have schematically summarized the available and reliable geochronological data of the EGC and correlated these data with the known tectono-stratigraphy of the EGC and of the Harz Mountains (Fig. 6). The inferred tectono-metamorphic history of the EGC, derived from such a picture, is also shown. The most important and decisive observation of the present study is the occurrence of detrital zircon crystals of magmatic origin and of Lower Devonian and Silurian age. This observation has several consequences. It implies that sedimentation of the detritus, now incorporated in the EGC, and the high-grade metamorphism must have taken place after 410±10 Ma, as given by the youngest detrital zircon. The lower limit for high-grade metamorphism is constrained by the intrusion of the Harzburg gabbronorite and the Brocken granite at ~295 Ma, which affected the EGC thermally as shown by mineral reactions (Franz et al. 1997) and an isotopically resetted sphene U–Pb age of 296 Ma (Baumann et al. 1991). It follows that any Cadomian metamorphism in the EGC can unambiguously be excluded. The sedimentation of the EGC protolith thus occurred contemporaneously with the deposition of Devonian to Carbonifereous siliciclastic sediments, which make up by far the largest portion of the Palaeozoic outcrop area in the Harz Mountains (Fig. 6). The implications of this paradox, as it might seem, are discussed in below.

Fig. 6
figure 6

Schematic summary of geochronological data and the inferred geological history of the Eckergneiss Complex in comparison with the tectono-stratigraphy of the Harz Mountains (modified after Ganssloser (1999), using data of Mohr (1978); Wachendorf (1986); provenance data are from Haverkamp et al. (1992) and Huckriede et al. (1998)). Age boundaries between series epochs are from Gradstein et al. (2004). For abbreviations see Fig. 1

Full size image

Provenance of the Eckergneiss protolith

The EGC metasediments are a suite of Ca- and alkali-poor metapelites and metapsammites of high compositional maturity. This composition reflects, together with a refractory heavy mineral assemblage of mainly zircon, chrome spinel, rutile, and occasional garnet as major detrital heavy minerals, a passive continental margin provenance (Martin-Gombojav 2003, unpublished doctoral thesis). It is similar to the major Rhenohercynian lithologies, which can be attributed to a passive continental margin setting at the southern rim of the Old Red Sandstone Continent (Franke 2001). As discussed in the previous section, the EGC protolith’s sedimentary sequence must have been deposited between the Lower Devonian and Upper Carboniferous. Taking into account the currently inferred palaeogeographic reconstructions of relevant terranes for this time span (e.g., Torsvik 1998; Pharaoh 1999; Robardet 2003) and the position of the EGC within the Rhenohercynian domain, a number of potential source areas of the detritus of the Eckergneiss protolith have to be considered. In early Devonian time Baltica—the Avalonia microplate with Gondwana affinity—and Laurentia had already been amalgamated in the course of the Caledonian orogeny, now forming Laurussia, i.e., the so called Old Red Continent. At that time, Laurussia was located to the north of the Rhenohercynian Ocean, which was subsequently closed by the northward motion of the Gondwana Supercontinent. The collision of Gondwana with the Old Red Continent in Carboniferous times is recorded by the Variscian orogeny in Central Europe. The clastic strata of the Rhenohercynian domain are characterized in temporal succession by material shed from the active continental margin of the Variscan front in the southeast, i.e., from peri-Gondwana/Gondwana terranes, and by detritus derived from the Old Red continent to the northwest (Haverkamp et al. 1992; Huckriede et al. 1998). Haverkamp et al. (1992) discovered that detrital zircons in Rhenohercynian Lower Devonian sandstones show U–Pb systematics, which hint towards an ultimate derivation from various Scandinavian sources, including south Greenland and the Caledonides on both sides of the Iapetus suture. They consider an origin from “positive” areas within the “Mid-European Caledonides”, i.e., the Danish–North German–Polish Caledonides according to present terminology. A change towards a major contribution of detrital zircons with a Gondwana age provenance is documented by Lower Carboniferous deposits (Haverkamp et al. 1992).

The Gondwana and peri-Gondwana terranes in North Africa, including the Saharan Metacraton (Abdelsalam et al. 2002), North America, and Europe are characterized by volcanism and plutonism between about 510 Ma and 900 Ma ago (e.g., Abdelsalam et al. 2002; Boullier 1991; Bregar et al. 2002; Cahen et al. 1984; Franke 2000; Ingle at al. 2003; Pin 1991; Rocci et al. 1991; Fig. 7e). Cambrian ages were also reported from Cadomian terranes (e.g., Zulauf et al. 1997; Dörr et al. 2002). Tichomirowa et al. (2000) accentuated the absence of Grenvillian ages in detrital zircon populations in various gneisses of the Erzgebirge within the Saxothuringian domain, which exhibits a Gondwana affinity. The authors detected zircon age groups of about 575 Ma, 600–700 Ma, 2.1–2.2 and 2.7–2.8 Ga, suggesting a derivation from the West African Craton.

Fig. 7
figure 7

Comparison of the cumulative probability curve (Ludwig 1998) of SHRIMP ages from detrital zircons of the Eckergneiss Complex with those obtained from age data of magmatic and metamorphic rocks from various potential source areas of a–d the southern Baltic Shield and e North Africa (Gondwana). The inset map shows the lithotectonic domains of SW Scandinavia (after Söderlund et al. 1999, and references therein): SF Svecofennian domain, TIB Transscandinavian Igneous Belt, ES Eastern Segment, SFDZ Sveconorwegian Frontal Deformation Zone, WS Western Segment, O Oslo area, K Kongsberg sector, B Bamble sector, T Telemark sector, R–A Rogalan Agder sector. Data sources: (a) compilation by Å häll and Larson (2000); (b) compilation and references in Söderlund et al. (1999); (c) compilation and references in Åhäll and Larson (2000), metamorphic zircon data are from a compilation of Söderlund et al. (2002); (d) compilation of Bingen et al. (2003), metamorphic zircon data from the R–A sector are from Möller et al. 2002; (e) compilation of Boullier (1991), Cahen et al. (1984), and of Rocci et al. (1991), and Bregar et al. (2002)

Full size image

The North African cratons are dominated by ~2.0 Ga old gneisses and additionally contain Neo-Proterozoic and Archean rocks. Neo-Proterozoic and Archean ages are also known from Greenland, North America (e.g., Nutman et al. 1996; Bevier et al. 1990; Zhao et al. 2002), and the Amazonian craton (Tassinari and Macambira 1999) and are commonly found in provenance studies involving sediments derived from Laurentian (e.g., Karabinos and Gromet 1993; Dickinson and Gehrels 2003; Strachan et al. 1995) and Gonwanan sources (e.g., Fernández-Suárez et al. 2002; Zeh et al. 2001). The strong dominance of Meso-Proterozoic age components in the detrital zircon population of the EGC without a significant Neo-Proterozoic or Archean contributions Footnote 2 (Fig. 4) along with the fact that only one single zircon yielded an age of 566±6 Ma (Fig. 5a) precludes a major derivation of the detritus from Laurentia, Amazonia, Gondwana, and peri-Gondwana. This reasoning applies also to the question of an involvement of the MGCR as a possible detritus source. Zeh et al. (2001) found distinctively Gondwana zircon age characteristics in rocks of the MGCR with the Avalonian–Cadomian orogenic belt and the West African and/or eastern Amazonian cratons as equally dominant sources and an only minor (10%) contribution from the Grenvillian belt. The latter would have to be translated into “Sveconorwegian belt” for its European (Baltica) continuation.

The absence of a significant Archean sediment contribution (<8%) Footnote 3 in the EGC is also in agreement with the conventional U–Pb data of detrital zircon fractions, which point to an upper Concordia intercept of about 1,600 Ma (Baumann et al. 1991). Instead, the observed distribution of Proterozoic ages is entirely consistent with an origin of the detritus from a SW Baltic source (Fig. 7). The SW Baltic provenance becomes evident when comparing the observed age spectrum with the geochronology of different lithotectonic domains within the Baltic Shield (inset map of Fig. 7, see also Nironen 1997; Gaál and Gorbatschev 1987). It is evident from Fig. 7 that the observed Proterozoic age spectrum matches best with the magmatic and metamorphic evolution of the western segment of south Sweden and the Sveoconorwegian domains in south Norway (e.g., Bingen et al. 2003 and references therein). Furthermore, the ages of metamorphic growth zones (Fig. 5i), sector-zoned zircon (Fig. 5g), and recrystallized zircon (Fig. 5h, j; see also Geisler et al. 2004) fall well in the age intervals of 1,120–1,150 Ma and 960–1,030 Ma, which have been determined for early and late Sveconorwegian granulite-facies metamorphism in these areas, respectively (e.g., Bingen et al. 2003; Möller et al. 2002).

Whereas tracing the source of the Meso-Proterozoic zircon crystals is relatively straightforward, the origin of the Silurian or Caledonian component is more of a problem. Caledonian ages are known from those parts of Laurentia and Baltica (e.g., Scotland and north Norway) that were affected by the Caledonian orogeny (e.g., Gee and Sturt 1985), but also from various localities within the Mid-European Variscides (e.g., Anthes and Reischmann 2001; Franke 2001, and references therein). However, the strong SW Baltic provenance of the detritus suggests that the Caledonian zircons of the EGC were derived either from the Caledonian nappes in southern Scandinavia or, more likely, from the geographically closer Danish-North German–Polish Caledonides. One may speculate that one of these Caledonic areas is also a potential source for the unique detrital chrome spinel fraction (Fig. 3). However, a problem is posed by the fact that to the best of our knowledge no significant chromite source is known in the exposed part of the southern and central Baltic Shield as well as in the still existing portions of the severely eroded south Scandinavian Caledonian nappes. The Danish-North German–Polish Caledonides rest deeply subsided under thick post-Silurian sediments and the same applies to the southwestern extension of the Baltic Shield beneath most of Denmark where the Sveconorwegian orogenic belt continues as far west as into the Danish North Sea (Obst et al. 2004). It can thus, likewise, be speculated that a structural high of the concealed Sveconorwegian domain, which was exposed to erosion during the Devonian and/or Carboniferous, might be the missing detritus source for chromites as well as for zircons.

Mode of emplacement and origin of the Eckergneiss allochthon: potential implications for the Variscan geodynamics

Since sedimentation in the Rhenohercynian and in the EGC-protolith was virtually synchronous, the high-grade EGC can hardly be seen as an uplifted segment of the Avalonian basement of the Rhenohercynian domain from beneath the present position of the EGC, as was suggested by Franz et al. (1997). A possible scenario being able to accommodate the reported new age data and other observations together with longer known facts would be that the EGC is a remnant of a tectonic nappe. The root zone of this nappe must be deep-seated enough to allow for the intense and partially dynamic metamorphism.

It would be tempting to search the root zone of the EGC outside the Rhenohercynian domain, where high-grade metamorphic entities are known (e.g., Reinhardt and Kleemann 1994; Willner et al. 2000). Constraining facts for the location of a possible root are: the lateral transport must have taken place prior to ~295 Ma and, more crucial, the sediment supply of the protolith must have partially involved the same sources as certain Rhenohercynian clastic sediments. This reasoning focuses around the question of zircon age spectra and chrome spinel characteristics. Devonian through earliest Carboniferous sediments outside, but also inside (e.g., Harz greywackes, Fig. 6), the Rhenohercynian zone were derived from Cadomian basement to the southeast, i.e., yielded Gondwana age spectra from detrital minerals (Haverkamp et al. 1992; Huckriede et al. 1998). Gondwana zircon age spectra are also reported from the MGCR (Zeh et al. 2001). On the other hand, Baltic provenances are known from certain Lower Devonian to Upper Carboniferous sandstones exposed in the Rhenish Massif and the Harz Mountains (e.g., the ABQ, Fig. 6), in the Carboniferous of the Pennine Basin, and in Carboniferous sandstones of the North Sea (Hallsworth et al. 2000; Haverkamp et al. 1992; Huckriede et al. 1998; Morton et al. 2001). Based on the observed Baltic provenance of the EGC protolith, it follows that the root zone of the EGC cannot be situated too far away from the Old Red continent to the north, i.e., is possibly located inside the Rhenohercynian domain. This hypothesis, which seems to be a paradox at first glance, is supported by the co-occurrence of unusual low-Mg chrome spinels together with a majority of Mg-rich chrome spinels in the clearly Rhenohercynian Lower Carboniferous ABQ. The parallel occurrence of low-Mg chrome spinels, which are the only type of chromite in the EGC (Fig. 3), but a minor component in the ABQ, suggests involvement of an identical though not exclusive source of the detritus of both units and furthermore precludes postdeposition Mg-loss within the EGC.

Within the Harz Mountains and surrounding areas there is no counterpart of the metamorphic EGC lithology and thus no obvious straightforward candidate for a possible root zone. As the entire eastern Harz is a succession of allochtonous units, which were transported as tectonic nappes from the east or southeast (e.g., Franke 2001), the EGC would fit into this general scenario. Its geographical position is within the limits of the Harz allochthonous complex, as shown in Fig. 3 of Franke (2001). The northwestern front of the allochthon is marked by the ABQ. The most likely candidate for a suture, which might be an expression of the EGCs’ root zone, is situated between the MGCR and the Northern Phyllite Zone, which were regarded by Franke (2001) as root of the allochthonous units in the Rhenohercynian Belt. The EGC would in this case barely be just another segment of the stacked allochthon of the eastern and southern Harz Mountains. In a written personal communication, of December 2002, W. Franke classified the EGC as an allochthonous wedge at the base of the Giessen/Harz Nappe.

A remaining problem may be the position of the EGC wedged in between the Brocken Granite to the east and the Harzburg Gabbronorite to the west. The intrusions are certainly autochthonous and related to the Oker Granite intrusion farther west in the parautochthon of the Harz Mountains. This position gave rise to the former view of the EGC being passively lifted into the present position by ascending magma. We suggest that the location of the EGC within the major plutonic complexes of the Harz Mountains is not accidental, but possibly either due to forceful or gravitational subsidence into the igneous complex or as part thereof. Contrary to this model, the present position of the EGC is connected with a thrust plane, which was merely used by magma for the uppermost portion of its ascent. The geometry of the Brocken intrusion could reflect a tectonic imprint of southeastward dipping fault or thrust planes by overlapping the EGC from the southeast with a gently southeastward inclining interface with the EGC (Wachendorf 1986).

The problem of the possible paradox of a Rhenohercynian affinity of the protolith’s sedimentary character, including the zircon and chromite provenances of the EGC in connection with a “non-Rhenohercynian” metamorphic grade, can be solved with respect to the allochthonous nature of the EGC. It can be both Rhenohercynian and medium- to high-grade metamorphic. A possible scenario can be drawn from a geodynamic model developed by Oncken (1997). According to Oncken (1997), the MGCR southeast of the Rhenohercynian (Fig. 1a) is essentially composed of a Lower Carboniferous magmatic arc association, resting on a stack of medium pressure–medium temperature rocks of inferred Rhenohercynian origin. This implies that medium-grade metamorphic rocks of Rhenohercynian provenance exist east of the Harz Mountains concealed beneath the MGCR and, according to Oncken (1997), even still farther east under the Saxothuringian Zone. The author suggested that in the progress of collisional tectonics, the Rhenohercynian passive margin sequence was first underthrust below the upper plate wedge to the east and later accreted under formation of imbricate fans. In line with this scenario, the EGC would be an upthrusted wedge of the concealed eastern extension of the Rhenohercynian, which became incorporated in the nappe movement to the northwest together with the other units of the Rhenohercynian allochthon of the Harz Mountains. According to this mechanism, the EGC would be a klippen-like segment of the southeastern Rhenohercynian that is derived from a relatively deep level of the underthrust Rhenohercynian. The interpretation of the EGC as part of a nappe would demand a slab-like geometry. Indeed, such geometry is indicated by geophysical investigations of Düweke et al. (1976), which figured the maximum thickness of the EGC with ca. 400 m. The high-grade granulite-facies overprinting could then be due to the mainly static influence of the magmatic arc of the MGCR, as suggested by W. Franke in a written communication of December 2002. Nevertheless, new petrographic observations (Martin-Gombojav 2003, unpublished doctoral thesis; see also Sect. 2) and P–T estimates on kinzingites (Schlüter 1983, unpublished doctoral thesis) indicate that the metamorphism in the Eckergneiss occurred at rather low pressures and under progressively higher temperatures up to granulite-facies conditions, which does not fit into a typical P–T scenario expected in a subduction environment.

Before ending the discussion, we should mention that an alternative mode of emplacement of the EGC has to be considered that is based on a concept of strike-slip movements of Variscan terranes of different P–T–t histories, which occurred between about 360 Ma and 320 Ma, as proposed by Krohe (1996). In contrast to the common view that the fault-bounded crustal blocks in the Variscan belt are the result of the amalgamation of continental microplates in the course of the Variscan orogeny (e.g., Franke 1989), which the above model is based on, Krohe (1996) suggested that the fault-bounded entities were created by intraplate deformation of a weak domain of continental lithosphere, which occurs significantly after accretion. In such a tectonic picture, the EGC would represent a fault-bounded complex with an origin, which must be located far east or south east of the present location. We note here that strike-slip displacements can reach distances larger than 500 km (e.g., Tapponier et al. 1990).

Both models would afford metamorphism of both types before commencement of tectonic movement, very likely in short temporal succession and in the described order. However, a problem might be posed by the good preservation of many orthopyroxenes and cordierite in the investigated granulite-facies rocks, which must have been incorporated in the transport. On the other hand, these movements might explain that a minority of the blocky orthopyroxenes of granulite-facies parageneses are tectonically bent.

It is obvious that a detailed knowledge about the timing and the conditions of the metamorphism of the EGC may help to locate the root zone of the EGC, which, in turn, would be another important piece of knowledge needed to understand the complex kinematic pattern of the Variscan Belt. Detailed geochronological and further petrological work is thus envisaged for the next future. For the time being, the location of the root zone remains a matter of speculation.