Strontium isotopes — A persistent tracer for the recycling of Gondwana crust in the Variscan orogen

Gondwana Research 22 (2012) 262–278

Contents lists available at SciVerse ScienceDirect

Gondwana Research

j ourna l homepage: www.e lsev ie r .com/ locate /gr

Strontium isotopes — A persistent tracer for the recycling of Gondwana crust in theVariscan orogen

Rolf L. Romer ⁎, Hans-Jürgen Förster, Knut HahneDeutsches GeoForschungsZentrum GFZ, Telegrafenberg, D-14473 Potsdam, Germany

⁎ Corresponding author. Tel.: +49 331 288 1405; faxE-mail address: [email protected] (R.L. Romer

1342-937X/$ – see front matter © 2011 International Adoi:10.1016/j.gr.2011.09.005

a b s t r a c t
a r t i c l e i n f o
Article history:Received 15 July 2011Received in revised form 25 August 2011Accepted 5 September 2011Available online 24 September 2011

Handling Editor: R.D. Nance

Keywords:Saxo-ThuringiaPalaeozoicRheic OceanGondwanaVariscan orogenyGeochemistry

Voluminous Cambro-Ordovician quartz-rich sandstones blanket large parts of North Africa and Arabia, aswell as terranes that later became separated from northern Gondwana. These sandstones are the result ofwidespread and intense chemical weathering that gave rise to a strong depletion in Na, Ca, and Sr and, con-comitantly, very high K/Na and Rb/Sr ratios. Because of their high Rb/Sr ratios, these rocks developed throughtime highly radiogenic Sr-isotopic compositions (in part with 87Sr/86Sr0N1.0). This particular geochemicalsignature may provide a geochemical provenance indicator for early Palaeozoic sedimentary rocks, whichmay be used to (i) trace the fate of particular lithological units during orogenic processes and (ii) constrainthe provenance of sedimentary rocks on terranes of disputed palaeogeographic position. Among the Palaeo-zoic sedimentary and volcanic rocks deposited on the Gondwanan shelf of Saxo-Thuringia (Germany), onlyLower Ordovician siliciclastic rocks like those of the Tremadocian Frauenbach Group (Schwarzburg Anticline)are characterized by this weathering-related strong depletion in Na, Ca, and Sr. Metamorphic nappes of theadjacent Erzgebirge consist of lithologies originally deposited on theGondwana shelf.Modeling of the Sr-isotopiccomposition demonstrates that the high Rb/Sr ratios accounting for the highly radiogenic measured 87Sr/86Srvalues are not due to metamorphic element mobility, but represent a primary signature already acquired atthe time of deposition. Thus, the “Frauenbach Sr-isotopic signature” – similar to the geochemical fingerprint –can be traced throughmetamorphism even to high-grade conditions. High 87Sr/86Sr ratios are also present in sili-ciclastic rocks derived from the erosion of the exhumed Variscan orogen. The appearance and disappearance ofthis signature put constraints on the erosion history of the Variscan orogen; its dilution allows a rough estimateof the relative contribution of the high-87Sr/86Sr metamorphic rocks to the erosional debris. Similarly, thehigh Rb/Sr granites of the Erzgebirge may not be the product of extreme fractional crystallization alone, butmay reflect the involvement of protoliths with high Rb/Sr. Granites with radiogenic Sr-isotopic compositionshare – despite geochemicalmodification by differential melting of the source and subsequent fractional crystal-lization – geochemical fingerprints with the Frauenbach Group, in particular increased W, Sn, F, Li, and Rb con-tents and low Sr abundances. Should it turn out that high initial 87Sr/86Sr and highW, Sn, F, and Li of granites arelinked to one particular crustal protolith, both Sn-enriched granites of Variscan Europe and Palaeozoic sedimen-tary rocks with Frauenbach geochemical signatures represent a Gondwana fingerprint.

© 2011 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.

1. Introduction

Voluminous deposits of mature Cambro-Ordovician quartz-richsandstones blanket large parts of North Africa and Arabia, from theAtlantic coast to the Persian Gulf (e.g., Noblet and Lefort, 1990).These chemically mature sandstones and quartzites were mainlyderived from Neoproterozoic Pan-African basement that was affectedby widespread and intense chemical weathering (Avigad et al., 2005),locally resulting in the development of bauxitic laterite and quartzarenite (cf. German et al., 1994; Avigad et al., 2005). Similar maturequartz-rich sandstones are also known from terranes that had been

: +49 331 288 1474.).

ssociation for Gondwana Research.

detached from the northern margin of Gondwana by late Cambrianto early Ordovician extension and rifting that eventually led to theopening of the Rheic Ocean (e.g., Nance et al., 2010). Examples forsuch early Palaezoic sandstones on Gondwana-derived terranes in-clude the Grès Armorican Formation (Armorican Quartzite) of north-western France (Noblet and Lefort, 1990) and the TremadocianFrauenbach Group of the Saxo-Thuringian Zone in Germany (Fig. 1;e.g., Lützner et al., 1986;Mingram, 1998). In contrast, Baltica, Laurentia,and Avalonia did not experience such intense chemical weatheringat that time (Erdtmann, 1991; Giese et al., 1994). Thus, for VariscanEurope, which formed during the early stages of Pangea formationwhen Gondwana collided with Laurussia (Laurentia, Avalonia, andBaltica), the occurrence of such mature clastic sedimentary rocks oflate Cambrian to early Ordovician depositional age and their metamor-phic equivalents provides a tracer for Gondwana provenance.

Published by Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.gr.2011.09.005

mailto:[email protected]

http://dx.doi.org/10.1016/j.gr.2011.09.005

http://www.sciencedirect.com/science/journal/1342937X

Fig. 1. Geological setting of Saxo-Thuringia. A. Tectonic framework of Variscan Europe showing the distribution of basement blocks and major tectonic lineaments (simplified afterFranke, 1989, 2000). B. Subdivision of the Saxo-Thuringian Zone with respect to Variscan overprint (Kroner et al., 2007, 2010). Rocks of the Autochthonous Domain include volcanic,sedimentary, and magmatic rocks of the Cadomian arc (broadly referred to as Cadomian basement) and late Proterozoic and Palaeozoic sedimentary (and volcanic) rocks of theshelf sequence deposited on the Cadomian basement. Rocks of the Allochthonous Domain include the same lithologies as the Autochthonous Domain, forming a stack of nappesthat had experienced contrasting grades of Variscan metamorphism (Mingram and Rötzler, 1999; Rötzler and Plessen, 2010). These two domains are separated by the Wrenchand Thrust Zone that includes the same lithologies as the Autochthonous Domain, but differs from it by showing late-Variscan folding and schistosity that developed during theemplacement of the metamorphic nappes of the Allochthonous Domain (for details see Hahn et al., 2010, Kroner et al., 2010, Kroner and Romer, 2010). C. Geological map of theSaxo-Thuringian Zone (simplified from Linnemann and Schauer, 1999).

263R.L. Romer et al. / Gondwana Research 22 (2012) 262–278

264 R.L. Romer et al. / Gondwana Research 22 (2012) 262–278

Decomposition of feldspar during the profound late Cambrianchemical weathering on Gondwana released, among other elements,Na, K, Al, Si, Ca, Rb, and Sr. Whereas Al, K, and Rb were subsequentlyre-incorporated in clay minerals, Na, Ca, and Sr, along with some Si,were lost. The resulting strong depletion in Na, Ca, and Sr representsa geochemical signature that is characteristic of intense chemicalweathering (e.g., Nesbitt et al., 1982, 1996; Mingram, 1998). Such adepletion in these elements is also present in the quartz-rich sand-stones of the Frauenbach Group, which represent redepositedmaterial derived from the intensely weathered sedimentary coverof northern Gondwana (e.g., Lützner et al., 1986; Mingram, 1998;Linnemann and Romer, 2002; Linnemann et al., 2004, 2010). Inthese rocks, the depletion in Sr resulted in high Rb/Sr values and –

with time – gave rise to anomalously high 87Sr/86Sr ratios (Romeret al., 2008; Romer and Hahne, 2010). This distinctly radiogenicSr isotopic composition in Tremadocian sedimentary rocks repre-sents not only a tracer for Gondwana provenance of tectonic frag-ments of unknown provenance within orogens that formedduring the assemblage of Pangea, but also allows to trace and – tosome extent – quantify the recycling of these rocks during orogenicprocesses, for instane in crustal melts and sedimentary units repre-senting the erosional debris of these orogens.

In the Saxo-Thuringian Zone, the Tremadocian (Lower Ordovi-cian) sedimentary rocks of the Frauenbach Group carry this dis-tinctive geochemical signature (depletion in Ca, Na, and Sr; e.g.,Mingram, 1996, 1998; Rötzler and Plessen, 2010). Due to the deple-tion of Sr, these rocks had already acquired highly radiogenic Sr-isotopic compositions by the time of the Variscan orogeny (at c.340 Ma). We illustrate how this distinctive Sr-isotopic compositioncan be used as (i) an isotopic fingerprint for the protoliths of themetamorphic nappes of the Erzgebirge and (ii) a tracer for the in-volvement of these rocks in the source of the c. 327–318 Ma granitesof this area (Fig. 1). Furthermore, we demonstrate that (iii) this sig-nature – in diluted form – is also present in the erosional debris fromthe Variscan orogen. In fact, the first appearance and disappearanceof this signature in the sedimentary rock record of the forelandbasin, provides – in an indirect way – a glimpse at orographic devel-opment, i.e., the time when orogenic exhumation made rocks withthis particular isotopic signature available for erosion in significantquantities, and the time when erosional leveling of the orogen ren-dered these rocks a subordinate component of the sediment budget.

2. Geological evolution of Saxo-Thuringia: from Palaeozoicsedimentation to Variscan high-grade metamorphism andmagmatism

The Saxo-Thuringian Zone is part of the Variscan orogenic belt,which formed when Gondwana collided with Laurussia (Laurentia,Avalonia, and Baltica), and represents one of the many small terranes(e.g., Saxo-Thuringia, Armorica, Iberia) that lay between these twocontinents during the collision. Common to these terranes is a Neo-proterozoic (Cadomian) basement that is unconformably overlainby sedimentary rocks of the Palaeozoic shelf. Sedimentation startedon the various blocks in the Cambrian or early Ordovician. Neodymi-um crustal-residence ages and the age distribution of detrital zircondemonstrate that these terranes were derived from the northernmargin of Gondana (e.g., Nance and Murphy, 1994; Linnemannet al., 2004, 2010) and were rifted and stretched in the Cambrian toearly Ordovician. Whether these terranes, which remained at thesouthern margin of the Rheic Ocean, drifted as independent micro-plates or represent the stretched leading edge of Gondwana is a mat-ter of debate (cf. Franke, 2000; Matte, 2001; Kroner et al., 2007, 2008,2010; Kroner and Romer, 2010).

The Saxo-Thuringian Zone is distinguished – on basis of contrast-ing Variscan deformation and metamorphic overprint – into anAutochthonous Domain and an Allochthonous Domain, separated

by a Wrench and Thrust Zone (Fig. 1; Kroner et al., 2007, 2010).Rocks of the Autochthonous Domain include remnants of theCadomian basement (N550 Ma) and the sedimentary rocks of theshelf sequence that developed after the opening of the RheicOcean, which started at c. 480 Ma (cf. Linnemann et al., 2000,2010). The Schwarzburg Anticline contains the most completesedimentary record of the opening and closure of the Rheic Ocean.Rocks of the Allochthonous Domain include the same lithologies asthe Autochtonous Domain (Mingram, 1998) that were metamor-phosed and deformed during the Variscan orogeny (e.g., Rötzler,1995; Rötzler and Plessen, 2010). The nappes of the Erzgebirge rep-resent one of the high-grade areas in the Allochthonous Domain.The metamorphic units of the Allochthonous Domain were juxta-posed as nappes along the Wrench and Thrust Zone against theAutochthonous Domain, uplifted and subsequently intruded bypost-kinematic, crustally-derived late-Variscan granites. Erosionaldebris from the Allochthonous Domain has been shed onto theAutochthonous Domain and is also present in the upper part of theSchwarzburg succession (Hahn et al., 2010). Subsequent brittle de-formation of Variscan Europe affected both domains.

2.1. Palaeozoic sedimentary rocks of the Autochthonous Domain

The Schwarzburg Anticline (Fig. 1) contains the most completesuccession of late Neoproterozoic to early Carboniferous sedimentaryrocks in the Saxo-Thuringian Zone. The Neoproterozoic and Cambriandevelopment of the sedimentary succession is interpreted to reflectthe transition from an orogenic Neoproterozoic basement to anextensional regime with bimodal volcanism and transgressive stableshelf systems (Linnemann et al., 2000, 2010; Heuse et al., 2010).Lower and Middle Cambrian siliciclastic debris was deposited inlocal basins. During Upper Cambrian uplift of the margin, these sedi-mentary rocks were largely eroded (e.g., Linnemann et al., 2000,2010). Incipient opening of the Rheic Ocean resulted in the formationof graben systems that were filled with siliciclastic material derivedfrom the uplifted Gondwana margin, which had undergone intensechemical weathering (Noblet and Lefort, 1990; Linnemann et al.,2000, 2010; Avigad et al., 2005). The Lower Ordovician siliciclasticsedimentary sequences of Saxo-Thuringia (Goldisthal, Frauenbach,and Phycodes Groups) are up to 3000 m thick, indicating significantsubsidence (Lützner et al., 1986). During the Middle and Upper Ordo-vician, the sedimentation rate slowed and the Arenigian to AshgillianGräfenthal Group has a total tickness of only 600 m. The sedimentaryrocks of this group include laminated shales with chamositic orehorizons and intercalated quartzites topped by a glaciomarine dia-mectite. The Silurian and Lower Devonian deep water sedimentaryrocks – represented by black shales and carbonates – were depositedat extremely low rates. From the Emsian to Tournaisian, deposition ofpelagic sediments increased progressively in volume. This increase ofclastic sediment input (pre-flysch stage in the older German litera-ture) reflects the onset of Rheic Ocean closure in the Emsian andthe exhumation of the oldest Variscan metamorphic rocks, nowpreserved in the Uppermost Allochthonous Units. These rocks hadreached their metamorphic peak conditions at c. 370–380 Ma (cf.Kreuzer et al., 1989; Rötzler et al., 1999; Klemd, 2010). In the Tournai-sian and Viséan, sedimentation dramatically increased and more than3500 m of shales and greywackes were deposited (Lützner et al.,1986; Hahn et al., 2010). These deposits represent the erosional de-bris from the Variscan nappes that had reached metamorphic peakconditions around 340 Ma and, thereafter, were rapidly exhumed(cf. Rötzler and Plessen, 2010). At the same time, there was an overalldevelopment towards a narrower sedimentation basin that becamefilled with increasingly coarser material. The sedimentary rocks ofthis narrow basin were folded during the final stages of the exhuma-tion of the Variscan nappes (Kroner et al., 2007; Hahn et al., 2010).

SiO2

K2O

/Na 2

O

1

10

0.1

1

10

0.1

1

10

0.1

1

10

0.1

Cs Rb K Na Sr Ca Ba La Eu Y

20

1

10

100

1000

0.140 60 80 100

Un

it (

aver

age)

/Up

per

Cru

st

Phycodes and Gräfen-thal groups

Frauenbach Group

Early Palaeozoicrhyolites

Cadomian basement

Phycodes and Gräfenthal groupsFrauenbach GroupEarly Palaeozoic rhyolitesCadomian basement

Sedimentary rocksMetamorphic rocksVolcanic rocks

BA

C

D

E

Fig. 2. Geochemical characterization of the Palaeozoic sedimentary rocks of the Autochthonous Domain (Schwarzburg Anticline) and the Allochthonous Domain (nappes of theErzgebirge). A. Geochemical variability of sedimentary rocks of the Schwarzburg area highlighting the anomalous composition of rocks of the Frauenbach Group, and showingthat the metasedimentary rocks of the nappes of the Erzgebirge display the same compositional range. B–E. Contrasting composition (normalized to Upper Continental Crust, Taylorand McLennan, 1985) of the various sedimentary and volcanosedimentary units in the Schwarzburg Anticline and their metamorphic equivalents in the Erzgebirge (simplified fromMingram, 1996, 1998; Mingram and Rötzler, 1999). B. Sedimentary rocks of the Phycodes and Gräfenthal groups of the Schwarzburg Anticline and their metamorphic equivalents(albite-bearing schists) of the Erzgebirge. The metamorphic equivalents of these two groups are distinguished by the abundance (Gräfenthal Group) and absence (Phycodes Group)of graphite. C. Sedimentary rocks of the Frauenbach Group of the Schwarzburg Anticline and feldspar-poor (-free) schist of the Erzgebirge. D. Early Palaeozoic rhyolites (typicallyinterfingered with the sedimentary rocks of the Frauenbach Group) and their metamorphic equivalents (muscovite schists and muscovite gneisses). E. Sedimentary rocks of theCadomian basement and red and gray gneisses of the Reitzenhain, Freiberg, and Lauenstein Gneiss Domes (Fig. 1).Data sources: Mingram, 1996, 1998; Mingram and Rötzler, 1999; Falk et al., 2000; Linnemann and Romer, 2002; Drost et al., 2007; Romer and Hahne, 2010.

0

0 10050 150 200Age (Ma.)

300250 350

5

10

15

20

25

30

35

Nu

mb

er o

f ag

es

FL

Age of late-Variscan granitesin the Erzgebirge

Age of lamprophyres in the Saxo-Thuringian Zone

N = 101 ages

EZ

Fig. 3. Histogram showing the age distribution of medium-temperature hydrothermalore deposits of the Erzgebirge. Greisen and skarn deposits directly associated with the


Related to this folding is a gap in the sedimentary record in orogen-proximal depositional areas, such as the Schwarzburg area.

After the opening of the Rheic Ocean, the Palaeozoic clastic sedi-mentary rocks of the Schwarzburg area were derived from a singlecratonic source, i.e., Gondwana. Thus, variations in the mineralogical,geochemical, and isotopic signature do not reflect changes in sedi-ment provenance. Instead, they reflect additions of juvenile material,variable weathering in the source, sorting during transport, and depo-sitional environment. The Proterozoic siliciclastic rocks (turbiditicgreywackes and mudstones) contain quartz, white mica, plagioclase,and chlorite and have large contents of Na, Ca, Sr, and Ba (Romerand Hahne, 2010). The Lower Ordovician Goldisthal Group shows –

stratigraphically upward – a progressive decrease in plagioclase con-tent. This trend culminates in the overlying Frauenbach Group thatis free of feldspar and, concomittantly, shows extreme depletion ofCa, Na, and Sr and an enrichment of K and Rb (Fig. 2). This geochem-ical fingerprint, in conjunction with the high compositional maturityand large heavy-mineral content of the quartzitic sandstones, indi-cates extreme chemical weathering in the source area, as well assorting by transport (e.g., McLennan, 1989; Bauluz et al., 2000). Thepredominance of 550–650 Ma detrital zircon indicates that these sand-stones representfirst-cycle sedimentary rocks (e.g., Avigad et al., 2005;Linnemann et al., 2007). The shales of the overlying Phycodes Groupshow a steady increase in the contents of albite and chlorite and,concomitantly, Na, Ni, and Li (cf. Hahne et al., 1984; Falk et al., 2000).

emplacement of the Late Carboniferous post-kinematic granites are not included. Themagnitude of the peaks does not necessarily respresent the intensity of the ore formingevent, since the selection of deposits for geochronological investigation is influenced bythe occurrence of datable minerals and the economic importance of the deposits,resulting in an overrepresentation of unusual deposits. Field data demonstrate thatmineralization associated with the c. 275 Ma age group is not only the economicallymost important, but also the most widespread.Diagram from Romer et al. (2010a).

2.2. Metamorphic units of the Allochthonous Domain

The Erzgebirge, which is situated in the Saxo-Thuringian Zone atthe northern border of the BohemianMassif, contains ultra-high pres-sure (UHP) and high pressure (HP) gneisses and eclogites, and forms

image of Fig.�2

550

500

450

400

350

300

250

1

2

3

4

Ma

-10 -8

εNd(T)

-6 -4 0 10 20 30 40 50

K2O/Na2O [wt%/wt%]

BituminousSilurian shales

Frauenbach Group

Average shale (K2O/Na2O)(Ronov & Migdisov 1996)

1086

Rb/Sr (ppm/ppm)420

tuffitic

tuffitic

tuffitic shales(Frasnian)

Post-

Variscan

molasse

0.700

87Sr/86Sr275

.725 0.7500.70

87Sr/86Srmeas

0.80 0.90 1.00 0.705

87Sr/86Srstrat

0.710 0.715 0.7200.70

87Sr/86Sr340

0.72 0.74

Fig. 4. Section through the Schwarzburg Anticline showing eNdT, Rb/Sr and K2O/Na2O, and the Sr-isotopic composition for t0=0, t275=275 Ma (age of major hydrothermal min-eralization in the Erzgebirge), t340=340 Ma (age of peak metamorphism in the Erzgebirge), and tstrat=biostratigraphically inferred age of the rocks. The Sr-isotopic compositionfor t325=325 Ma is very similar to the one for t340=340 Ma and is, therefore, not shown separately. The Nd-isotopic variation with stratigraphic age is due to aging of the cratonicsource and the variable admixture of juvenile volcanogenic material (cf. Romer and Hahne, 2010). The trend defined by the samples with the least radiogenic Nd-isotopic compo-sitons (lowest eNdT values) represents the temporal evolution of a reservoir with 147Sm/144Nd≈0.11 (i.e., average felsic crust). Excursions to higher eNdT values reflect variableadmixtures of redistributed volcanic material. The breaks in this trend correspond to changes in sediment source. The oldest sedimentary rocks contain detrital material fromthe Cadomian magmatic arc (Avalonia) and Gondwana. After the opening of the Rheic Ocean, Gondwana was the sole source of siliciclastic material. In the upper part of the sed-imentary section, siliciclastic input is dominated by Permian volcanic rocks. The Rb/Sr and K2O/Na2O data represent average values for the various lithostratigraphic units. Note, theprofoundly weathered rocks of the Frauenbach Group already show distinctly radiogenic 87Sr/86Sr values at the time of the Variscan orogeny. Anomalously low initial 87Sr/86Srvalues are not shown for reasons explained in the text. The chemical weathering that leaves such a strong imprint on Rb/Sr, K2O/Na2O, and the Sr-isotopic composition has no effecton the eNdT data (cf. Hahne and Romer, 2008; Romer and Hahne, 2010). The horizontal gray bands correspond to the following tectonic events: 1: Cadomian orogen; 2: rifting of theGondwana margin preceeding the opening of the Rheic Ocean; 3: peak metamorphic conditions reached in the Uppermost Allochthonous Units; 4: peak metamorphic conditionsreached in the Erzgebirge and the Saxon Granulite Massif.Data source: circles: Romer and Hahne (2010); diamonds: Linnemann et al., (2004).


one of the major metamorphic complexes of the European VariscanBelt (Fig. 1; e.g., Franke, 2000). The Erzgebirge consists of at leastfive tectono-metamorphic units (Rötzler et al., 1998). The high-grade units are separated from each other by mylonitized shearzones that contain mélanges of retrograded rocks (Sebastian, 1995;Rötzler and Plessen, 2010). Dating of metamorphic minerals demon-strated that the various units, which had experienced contrastingmetamorphic conditions and followed different P–T paths, reachedpeak metamorphic conditions at c. 340 Ma (Kröner and Willner,1998; Mingram and Rötzler, 1999; von Quadt and Günther, 1999;Werner and Lippolt, 2000; Tichomirowa, 2003) and were assembledinto the present nappe pile before the emplacement of the post-kinematic late Carboniferous granites at c. 327–318 Ma. The

Notes to Tablea Sample location, sample description, and full chemical characterization is given in Romb Sedimentation age inferred from biostratigraphic correlations using the time scale of Mc 87Sr/86Sr ratios analyzed using dynamic multi-collection (see method section). Values n

with an * has been separated using a 15 ml ion-exchange chromatography and have been anspectrometer, respectively, all other Sr samples were separated using a 3.5 ml ion-exchangespectrometer, respectively.

d Values recalculated for different ages using Rb and Sr contents and λ87Rb=1.42·10−1

over-corrections for in situ 87Sr growth and are – as such – meaningless. They are shown fe The original clay mineralogy of the shales has been altered during diagenetic and anch

tion. Thus, most of the fine-grained clastic sediments are slates rather than shales. The upperather than shales. To stress the common precursor of the investigated rocks, they are here rclaystones.

f Sample Gr-wand-04-1 is a 10-cm-thick tuffite (used as a regional marker horizon), theformations, respectively. For additional information see Romer and Hahne (2010).

mineralogical (for the low grade units) and geochemical similarity ofthe various units with the sedimentary rocks of the Schwarzburg Anti-cline (cf. Fig. 2) has been used to demonstrate that the tectonic repeti-tion within the nappe pile dominantly involved early Palaeozoicsedimentary rocks – and interlayered felsic volcanic rocks – of the pas-sive Gondwana margin (e.g., Mingram, 1996, 1998; Mingram and Röt-zler, 1999; Mingram et al., 2004; Rötzler and Plessen, 2010).

2.3. Late-Variscan magmatism

Late Carboniferous granite and lamprophyre magmatism in theErzgebirge (Vogtland) Zone occurred in two short pulses, onearound 327–318 Ma and one around 305–300 Ma (cf. Förster et al.,

er and Hahne (2010).enning and German Stratigraphic Commission (2002).ormalized to 86Sr/88Sr=0.1194; errors reported as 2σm. Strontium of samples markedalyzed using a VG Sector 54-30 multi-collector or VG Isomass 54 single-collector masschromatography and were analyzed using a MAT262 or a Triton multi-collector mass

1 a−1; strat=biostratigraphic age. Data given in italics represent values with obviousor illustration only. For discussion of the causes of this overcorrection see text.imetamorphic overprint to a monotonous muscovite-chlorite-quartz-feldspar associa-rmost units have been deposited after the Variscan Orogeny and represent claystoneseferred to as shales, although the investigated rocks actually include slates, shales, and

other three samples are oolitic iron ores on top of the Griffelschiefer and Schmiedefeld

Table 1Sr-isotope data for Neoproterozoic to Palaeozoic sedimentary and volcanosedimentary rocks from the Schwarzburg Anticline, Thuringia, Germany.

Samplea Ageb

(Ma)

87Sr/86Sr0c Rb(ppm)

Sr(ppm)

87Sr/86Sr275d 87Sr/86Sr325d 87Sr/86Sr340d 87Sr/86Srstratd

Shale e

RF-91-4* 570 0.711265±20 307 66 0.65858 0.64898 0.64609 0.60183Kaluga-91-3 570 0.790146±11 281 54 0.73120 0.72046 0.71724 0.667722093-13-A 570 0.735657±5 196 119 0.71700 0.71360 0.71258 0.696912074-11 560 0.730346±4 170 126 0.71506 0.71228 0.71144 0.699162070-13 550 0.739419±5 110 77 0.72324 0.72029 0.71940 0.70699TRUC-20-2A 550 0.733031±15 169 127 0.71796 0.71521 0.71439 0.702825431/104* 559 0.730950±20 364 53 0.65316 0.63898 0.63472 0.57506Hei 2/37* 530 0.711914±20 169 303 0.71282 0.71167 0.71133 0.70694Hei 2/32 525 0.717236±10 140 303 0.71200 0.71105 0.71076 0.70723Gl-11* 495 0.751430±30 276 36 0.66459 0.64876 0.64401 0.59487GI-30 495 0.848121±17 272 36 0.76253 0.74693 0.74225 0.6938284/2/06-N 492 0.765276±11 174 67 0.73581 0.73045 0.72884 0.7125282070 495 0.715994±10 487 43 0.58771 0.56433 0.55731 0.48472GLTM-416/8 495 0.798661±8 358 68 0.73903 0.72816 0.72490 0.69115GLHL-543 495 1.077446±14 551 16 0.68737 0.61628 0.59495 0.37421GLHL-630-1 493 0.894212±7 477 38 0.75203 0.72612 0.71834 0.63892GLHL-668-8 493 0.854860±7 500 50 0.74159 0.72095 0.71475 0.65148GLHL-830-9 490 1.018305±40 492 19 0.72499 0.67154 0.65550 0.49488OWS-410-5 490 1.078506±7 474 20 0.81005 0.76113 0.74645 0.59944GO-5 488 0.796743±7 471 89 0.73680 0.72587 0.72260 0.68977ZGI-TB 480 0.734364±12 180 160 0.72162 0.71930 0.71860 0.71209ZGI-TB* 480 0.734413±14 180 160 0.72167 0.71935 0.71865 0.712145225 480 0.749147±3 222 114 0.72709 0.72307 0.72186 0.710595284(+N) 480 0.742321±7 179 117 0.72499 0.72183 0.72089 0.71203Gö 10/3/7 480 0.745230±11 191 106 0.72482 0.72110 0.71998 0.70955UL1/29-N 480 0.756292±5 221 92 0.72908 0.72412 0.72264 0.70873WIR-98-2 480 0.763729±6 296 93 0.72768 0.72111 0.71914 0.700715245+N 470 0.736832±10 178 135 0.72190 0.71918 0.71836 0.711275252* 470 0.742330±20 172 129 0.72723 0.72447 0.72365 0.7164884-5-6-4 470 0.717835±12 0.4 123 0.71780 0.71779 0.71779 0.7177784-5-113A 470 0.727672±26 207 304 0.71996 0.71855 0.71813 0.7144784-5-11-1 445 0.726697±7 170 210 0.71753 0.71586 0.71535 0.7118484/5/10 445 0.742360±10 203 117 0.72271 0.71913 0.71805 0.7105284/3/24 445 0.732056±12 142 102 0.71629 0.71341 0.71255 0.70651Jä-3/4-SM 425 0.708657±3 22 184 0.70730 0.70706 0.70698 0.70656LI-17-04-56B 420 0.798806±13 187 31 0.73048 0.71803 0.71429 0.694344/7/9 415 0.809127±7 165 23 0.72787 0.71306 0.70861 0.686385/26/110 410 0.726659±7 160 156 0.71504 0.71292 0.71229 0.709322/88/318 400 0.732558±8 190 148 0.71802 0.71537 0.71457 0.711391/67/420 393 0.737852±15 186 117 0.71984 0.71656 0.71558 0.712109/8/504 384 0.734836±12 127 88 0.71849 0.71551 0.71462 0.71199BO-99-2 380 0.751726±7 125 46 0.72095 0.71534 0.71365 0.7091684/3/08-1 380 0.707032±8 122 75 0.68861 0.68525 0.68424 0.68155BO-01-1 375 0.743865±13 231 98 0.71717 0.71230 0.71084 0.70743Fi-W-99-1 365 0.763953±4 156 43 0.72286 0.71537 0.71312 0.70938RAT-99-8 362 0.751731±6 153 59 0.72236 0.71700 0.71540 0.71304Mue-4/1 361 0.748712±3 199 83 0.72155 0.71661 0.71512 0.71304BO-20-11 360 0.735684±7 175 112 0.71799 0.71476 0.71379 0.71250Fi-3/5 360 0.731028±7 213 164 0.71632 0.71364 0.71283 0.71176F1-3/16-B 359 0.729335±7 233 206 0.71652 0.71419 0.71349 0.71260Pfaf-IV/8B 359 0.733473±4 224 151 0.71667 0.71361 0.71269 0.71152Pfaf-I/23-N 358 0.735665±4 172 103 0.71675 0.71330 0.71227 0.71103U-90-4 350 0.726945±7 99 99 0.71562 0.71355 0.71293 0.71252U-10/1 342 0.733959±4 144 104 0.71828 0.71542 0.71456 0.71444U-90-5 340 0.726869±6 190 205 0.71637 0.71446 0.71388 0.71388HW.O2S-9 340 0.736395±6 193 138 0.72055 0.71767 0.71680 0.71680Rött-01-1 339 0.749395±6 225 98 0.72339 0.71865 0.71723 0.71732H-W-02F-3 339 0.722882±8 32 101 0.71929 0.71864 0.71844 0.71846EICH-01-3 338 0.758949±9 291 103 0.72695 0.72112 0.71936 0.719601037-N 300 0.748393±10 335 92 0.70715 0.703391203.47-B 300 0.762799±5 420 83 0.70548 0.70026MAN-0104 295 0.78475±4 358 59.1 0.71614 0.71114NESS-20-6 290 0.79755±6 498 57 0.69859 0.69318FB-02-04-13 285 0.763832±9 389 74 0.70429 0.70212

Tuffite and sedimentary iron oresf

Gr-wand-04-1 345 0.725430±9 85 92 0.71496 0.71306 0.71249 0.7122984-5-6-2 470 0.717885±6 0.2 24.0 0.71779 0.71777 0.71777 0.7177284-5-9B 445 0.715577±7 1.2 51.0 0.71532 0.71527 0.71526 0.71516Witt-07-4 445 0.717555±6 0.6 86.0 0.71748 0.71746 0.71746 0.71743



1999, 2007; von Seckendorff et al., 2004; Romer et al., 2007). Themagmatic pulse around 327–318 Ma included the emplacement ofall major granite plutons and subordinate mafic volcanism. Thepulse around 305–300 Ma was volumetrially subordinate and pro-duced more volcanic than intrusive rocks (cf. Förster et al., 1999,2007; Förster and Rhede 2006; Förster and Romer, 2010). The vari-ous granite plutons are generally multi-phase, highly evolved, andchemically and isotopically variable. There are five major composi-tional groups that can be distinguished on the basis of both theirmineralogy and chemical composition (Förster et al., 1999; for loca-tions see Fig. 1 and Supplementary material Fig. A1): (i) F-poor bio-tite granites (e.g., Kirchberg, Niederbobritzsch), (ii) F-poor two-micagranites (e.g., Bergen, Schwarzenberg), (iii) F–P-rich Li-mica gran-ites (e.g., Eibenstock, Ehrenfriedersdorf), (iv) P-poor, F-rich Li-micagranites (e.g., Zinnwald, Altenberg), and (v) medium-F, P-poor bio-tite granites (e.g., Gottesberg, Eichigt). The Li-mica granites are richin both LILE and ore elements (Sn, W) and became even moreenriched in these elements during melt differentiation and late-stage fluid-rock interaction. The essentially coeval formation of A, I,and S-type granites in the same region remained enigmatic inmodels that involved mantle-derived melts modified to variable ex-tents by crustal assimilation. Therefore, Förster and Romer (2010)interpreted the granites to represent crustal melts from the orogeni-cally overthickened Gondwana margin. The melting of Palaeozoicsedimentary rocks and intercalated volcanic rocks readily accountsfor the Nd-isotopic signature of the granites (between mantle andthe quartz-rich siliciclastic sedimentary rocks; cf. Linnemann andRomer, 2002; Romer and Hahne, 2010), whereas the predominantinvolvement of different types of clastic sedimentary rocks accountsfor the contrasting geochemical character of the various plutons. Thispetrogenetic model explains the geochemical variability and theintrinsically high contents of some elements in the granites as aninherited signature from the precursor lithologies. Furthermore, itimplies that the isotopic fingerprints (e.g., Sr, Pb) of the protolithswere inherited as well.

2.4. Post-Variscan hydrothermal overprint

During the post-Variscan rearrangement of the stress field inCentral Europe, largely related to Permian rifting, the opening ofTethys, and various stages of the opening of the Atlantic Ocean, faultzones were repeatedly reactivated and hydrothermal mineral de-posits formed. In the Erzgebirge, there were major pulses of reactiva-tion at c. 275 Ma, 180 Ma, 150 Ma, and 120 Ma (Fig. 3). The formationof these deposits involved the leaching of ore elements (in particularU, Ag, Pb, U–Bi–Co–Ni, Ba–F) from various source rocks and the trans-port of these elements into deformation zones, where they were de-posited (cf. Romer et al., 2010a). The event at c. 275 Ma was themost important one. Since the ore metals were dominantly derivedfrom leaching of granites and possibly their metamorphic surround-ings, the formation of these late- and post-Variscan hydrothermal de-posits is likely to have upset the geochemical and isotopic signature ofthe source rocks of the metals. Later hydrothermal events involvedboth the redistribution of ore elements within older deposits andthe addition of elements previously not present at economically rele-vant levels.

3. Methods and results of Sr-isotope analysis

The Sr-isotope data have been acquired in the course of severalprojects over a period of 12 years, using two different set-ups for Srseparation and three different mass spectrometers. Data for theshales from the Schwarzburg Anticline and the metamorphic unitsof the Erzgebirge are listed in Tables 1 and 2; earlier published datafor the Palaeozoic sedimentary rocks, the metamorphic rocks of the

Allochthonous domain, and the Late-Variscan granites are compiledin Supplementary material Tables A1, A2, and A3.

Strontium has been separated using standard HCl cation-exchange chromatography, using AG50-8X resin, 2.5 N HCl medium,and 15 ml and 3.5 ml resin volume, respectively. Details are given inMingram et al. (2000) for the larger resin volume and Romer et al.(2001, 2005) for the smaller one.

The Sr-isotopic composition was determined on a VG54-30 sector,MAT262, or a Triton mass spectrometer. All reported ratios were nor-malized to 86Sr/88Sr=0.1194. Measurements on the VG54-30 sectorwere performed using triple-jump dynamic multi-collection (Wendtand Haase, 1998). Strontium reference material NBS 987 gave long-time averages of 87Sr/86Sr=0.710246±12 (2 SD; n=16 analyses;Mingram et al., 2000) and 0.710249±4 (2 SD; n=12 analyses;Romer et al., 2005). A few samples have been analyzed on a FinniganMAT262 mass spectrometer using static multi-collection. NBS 987gave over the measurement period 87Sr/86Sr=0.710281±10 (2 SD;n=14 analyses). The statically determined 87Sr/86Sr values havebeen adjusted to a value of 87Sr/86Sr=0.710248 for NBS 987. Mea-surements on the Triton were performed using double-jump dynamicmulti-collection (no amplifier rotation). Routine measurement ofNBS 987 using dynamic multi-collection gave a 87Sr/86Sr value of0.710248±11 (2SD; n=21 analyses). The reported values have notbeen adjusted for this slight instrument bias. It should be noted thatstatic multi-collection (with or without amplifier rotation) does notyield exactly the same value as dynamic multi-collection, even for anew instrument. This point is illustrated in Supplementary materialTable A4.

The Rb and Sr contents used for the recalculation of the Sr-isoto-pic composition were determined by ICP-MS. The reported concen-tration data represent the average of three or four measurementseries, analyzed at different time and together with different –

depending on sample type – certified rock standards (e.g., JA-2, JB-3, SCO-1, TB, and TS). For details on the analytical procedure andlong-time reproducibility of reference material see Romer andHahne (2010).

4. Temporal evolution of the Sr-isotopic composition and thesignificance of apparently anomalous initial 87Sr/86Sr ratios

Several hundred Sr-isotopic analyses exist of Saxo-ThuringianPalaeozoic sedimentary rocks, Variscan metamorphic rocks, and lateVariscan granites (cf. Tables 1 and 2; Supplementary material TablesA1, A2, and A3). Within this huge data set, there are samples yieldinganomalously low apparent initial 87Sr/86Sr values. Two groups ofrocks yield such anomalous values: (i) rocks with very high87Sr/86Sr0 and 87Rb/86Sr values and (ii) rocks that have been over-printed by a secondary event.

(i) Sedimentary rocks from the Schwarzburg area with very radio-genic measured 87Sr/86Sr (N 0.8) and high measured 87Rb/86Srratios typically yield anomalous initial 87Sr/86Sr values, rangingfrom 0.59 to 0.69 (Table 1). Lithologically comparable sedi-mentary rocks below and above these anomalous rocks havegeologically reasonable initial 87Sr/86Sr values in the range of0.706 to 0.712 (Table 1; Fig. 4). Thus, samples with very radio-genic measured 87Sr/86Sr values have calculated apparent ini-tial Sr-isotopic compositions that are 0.02 to 0.12 87Sr/86Srunits lower than expected. Whereas such a deviation is highlysignificant for secondary disturbances for samples with rela-tively unradiogenic measured 87Sr/86Sr and, consequently,low to moderate 87Rb/86Sr values, two additional possibileexplanations exist for samples with high measured 87Sr/86Srand high to very-high 87Rb/86Sr values. First, an overcorrectionof 0.02 to 0.12 87Sr/86Sr units for samples with radiogenic Srisotopic compositions (Table 1) corresponds to 6–40% of the

Table 2Sr-isotope data for metamorphic rocks (former Neoproterozoic to Palaeozoic sediments) from the Erzgebirge, Germany.

Samplea Age(Ma)

87Sr/86Sr0b Rb(ppm)

Sr(ppm)

87Sr/86Sr275c 87Sr/86Sr325c 87Sr/86Sr340c 87Sr/86Srstratc

Slate unit (VLP–VLT)Gö1 485, fbs 0.915917±5 287 24 0.78046 0.75578 0.74837 0.67667Gö4 485, fbs 0.770788±5 183 50 0.72933 0.72178 0.71951 0.69756GI25 485, fbs 0.830758±4 256 47 0.76906 0.75782 0.75444 0.721799/76 485, phyc 0.746120* 211 112 0.72478 0.72089 0.71972 0.70843

Phyllite unit (~2 kbar/300 °C; LP–LT)HS2 485, phyc 0.765475±6 202 61 0.72797 0.72113 0.71908 0.69922HS5 485, phyc 0.742030* 149 107 0.72626 0.72338 0.72252 0.71417

Garnet–Phyllite unit (~9 kbar/470 °C; MP–LT)P38 485, fbs 0.844525±8 207 23 0.74258 0.72400 0.71843 0.66446P3 485, fbs 0.855900±8 334 46 0.77366 0.75867 0.75417 0.71063P4 485, fbs 0.879371±6 292 29 0.76532 0.74453 0.73830 0.67792P4 485, fbs 0.883620* 291 29 0.76996 0.74925 0.74303 0.68286GI3 485, fbs 0.850293±9 223 30 0.76610 0.75075 0.74615 0.70158

Mica Schist–Eclogite unit (N12 kbar/550 °C; HP–LT)61 485, fbs 0.835441±12 209 32 0.76146 0.74798 0.74393 0.70477128 485, fbs 0.826162±5 237 40 0.75905 0.74682 0.74315 0.707623/1 485, fbs 0.882632±11 151 20 0.79711 0.78153 0.77685 0.731582/32 485, fbs 0.830571±5 123 22 0.76724 0.75570 0.75224 0.71872P26 485, phyc 0.745850* 284 158 0.72549 0.72178 0.72067 0.7098998 485, phyc 0.761240* 277 107 0.73192 0.72657 0.72497 0.7094595-2 485, mgn 0.894110* 310 29 0.77303 0.75096 0.74434 0.6802595-8 496, mgn 0.955470* 326 23 0.79492 0.76566 0.75688 0.6654495-5 492, mgn 0.967940* 318 19 0.77836 0.74381 0.73344 0.62824

Gneiss–Eclogite unit (N22 kbar/830 °C; HP–HT)S3.1 485, fbs 0.904031±5 298 29 0.78764 0.76642 0.76006 0.69845E602a 485, fbs 0.957377±7 271 23 0.82391 0.79959 0.79229 0.72165A15c 485, fbs 0.941782±4 382 32 0.80657 0.78192 0.77453 0.70295A11 485, fbs 0.783704±3 187 59 0.74780 0.74126 0.73930 0.7202994-7 ~480, mgn 0.846820* 357 49 0.76429 0.74925 0.74474 0.70256S1a (480), dgn 0.713534±4 71 204 0.70959 0.70887 0.70866 0.70664S1b (480), dgn 0.744831±3 187 95 0.72253 0.71847 0.71725 0.70586

Gneiss Unit (6–8 kbar/620–650 °C; MP–MT)6/1 ~550, gag 0.736490* 188 136 0.72083 0.71798 0.71712 0.70511VOP18 ~550, gag 0.733740* 173 156 0.72118 0.71889 0.71820 0.70857VO1 ~550, gag 0.733530* 157 144 0.72118 0.71893 0.71825 0.7087894-11 ~550, bmt 0.719440* 134 239 0.71309 0.71193 0.71158 0.7067195-9 ~550, bmt 0.726270* 157 181 0.71644 0.71465 0.71412 0.706581/48 ~550, bmt 0.734710* 170 139 0.72086 0.71833 0.71757 0.7069593 ~550, bmt 0.733580* 155 129 0.71997 0.71749 0.71675 0.706317/21a ~550, bmt 0.718410* 102 210 0.71291 0.71191 0.71160 0.70739Vo5/2 ~550, bmt 0.720950* 115 233 0.71536 0.71434 0.71403 0.70975Gru1 ~550, bmt 0.718160* 77 253 0.71471 0.71408 0.71390 0.71125

a All units have reached peak metamorphic conditions at c. 340 Ma. The age of the precursor sediments within the various nappes has been inferred from comparison of theirmajor and trace element chemistry with the Schwarzburg profile (cf. Mingram, 1996, 1998; Mingram and Rötzler, 1999; Rötzler and Plessen, 2010). Precursor lithologies: fbs=sedimentary rocks of the Frauenbach Group; phyc=sedimentary rocks of the Phycodes Group; mgn=rhyolite (now muscovite gneiss); dgn=diamond-bearing gneiss; gag=granitoid augengneiss; bmt=Cadomian basement, exposed in the gneiss domes of Freiberg and Reitzenhain. For sample locations see Mingram (1996). Note, rocks with high87Sr/86Sr values are overrepresented with respect to their abundance in the Erzgebirge as units that have been inferred to be Frauenbach equivalents (Mingram, 1996, 1998)were sampled preferentially to test whether the Sr-isotope signature of the rocks is primary or reflects significant elemental mobility during metamorphism.

b 87Sr/86Sr ratios analyzed using dynamic multi-collection (see method section). Values normalized to 86Sr/88Sr=0.1194; errors reported as 2σm. * unpubl. data from Plessen.c Values recalculated for different ages using Rb and Sr contents and λ87Rb=1.42·10−11 a−1; strat=biostratigraphic age. Full chemical data sets (including Rb and Sr data) are

given in Mingram (1996). Data given in italics represent values with obvious over-corrections for in situ 87Sr growth and are – as such –meaningless. They are shown for illustrationonly. For discussion of the causes of this overcorrection see text.


total correction. An analytical uncertainty of 1–2% in the Rb/Srratio (which is readily obtained by ID-TIMS) for a sample witha measured 87Sr/86Sr ratio of ~1.0, corresponds to an uncer-tainty of 0.003 to 0.006 87Sr/86Sr units in the correction ofin situ 87Sr growth. If the Rb/Sr is determined by ICP-basedmethods (no isotope dilution), however, the uncertainty maybe considerably larger, especially for samples with very lowSr contents (e.g., b20 ppm), since simple rounding to thenext ppm alone may introduce an uncertainty of several per-cent. For such samples, the uncertainty in the Rb/Sr ratio mayeasily reach 10 to 20%, which for samples with measured87Sr/86Sr ratios of ~1.0, corresponds to an uncertainty of 0.03

to 0.06 87Sr/86Sr units in the correction of in situ Sr growth.Thus, for some samples with highly radiogenic Sr-isotopiccompositions, the uncertainty in the Rb/Sr ratio may accountfor the observed overcorrection of in situ 87Sr growth. Note,the uncertainty in Rb/Sr may equally well result in an underes-timation of in situ 87Sr growth, which is not readily apparentsince undercorrection does not result in geologically unreason-able values. Second, all 87Sr/86Sr initial ratios have been calcu-lated using the 87Rb decay constant recommended by IUGS(Steiger and Jäger, 1977). If, however, the true decay constantwere lower, as suggested by Minster et al. (1982) (see also dis-cussions by Begemann et al., 2001; Amelin and Zaitsev, 2002),


all values would have been overcorrected systematically by1.3% (2% according to Nebel et al., 2011). The above two expla-nations apply to all samples. Their effects (excess scatter andunreasonably low apparent initial 87Sr/86Sr values), however,become obvious only for samples with large corrections for insitu 87Sr-growth, i.e., samples with very high 87Rb/86Sr values.

(ii) Fig. 5A shows schematically the Sr-isotopic evolution withtime. The present-day (measured) Sr-isotopic composition isshown at T=0. Earlier Sr-isotopic compositions of the sampleare projected back in time along a line whose slope is definedby the 87Rb/86Sr ratio of the sample and the decay constantof 87Rb. The higher the 87Rb/86Sr ratio, the steeper the lineand the more pronounced the changes in 87Sr/86Sr with time.For a single-stage system, the Sr-isotope evolution wouldevolve along a line extending from the time of formation ofthe rock to the present. A system with changes in Rb/Srthrough time would be represented by a line with a kink ateach disturbance of the Rb/Sr ratio. While the Sr-isotopiccomposition after the disturbance is constrained by the mea-sured Rb/Sr and 87Sr/86Sr values, the pre-disturbance evolu-tion is known only qualitatively. From the Sr data alone,

1.2

1.1

1.0

0.9

0.8

0.7

0.61000 200 300

Age (Ma)

87S

r/86

Sr

400 500

Disturbance with fractionation of 87Rb/86Sr

High 87Rb/86Sr samples

Low 87Rb/86Sr samples

1.2

1.1

1.0

0.9

0.8

0.7

0.61000 200 300

Age (Ma)

87S

r/86

Sr

400

Peak of Variscan metamorphism in Erzgebirge

500

A

C

Emplacement of largeLate-Variscan granites

in Erzgebirge

Formation ofhydrothermal vein deposits

in Erzgebirge

SchwarzburgMetamorphicnappes(Erzgebirge)

Fig. 5. Temporal variation of the Sr-isotopic composition of individual samples. A. Conceptuacomposition of Sr through time for single-stage and poly-stage Sr evolution. Note, any incdisturbance change in 87Sr/86Sr of the sample by growth of radiogenic 87Sr, and (ii) inevitaage of the rock (for granites and sedimentary rocks) or its precursor sedimentary rocks (fodirection of Sr recalculation). For discussion see text. B. Evolution of 87Sr/86Sr in shales frothe Erzgebirge. D. Evolution of 87Sr/86Sr in late Carboniferous granites of the Erzgebirge.might have upset the Rb/Sr systematics of the rock are shown at the appropriate time (vNote that the Sr-isotopic evolution of several granite samples requires a late (200 Ma or yo

however, it is not possible to determine (i) the time of distur-bance, with the exception of the maximum age of the lastdisturbance for samples that yield geologically unreasonablylow apparent 87Sr/86Sr values, (ii) whether there were severaldisturbances, and (iii) the true value of the initial 87Sr/86Srratio of samples with polystage Sr-evolution.

The secondary increase in 87Rb/86Sr results in a higher productionof 87Sr and an accelerated change in 87Sr/86Sr after the disturbance,i.e., the line is steeper after the disturbance (Fig. 5A). Incorrectly ig-noring such a late increase in 87Rb/86Sr and projecting the Sr-isotopiccomposition along a line from the measured 87Sr/86Sr ratio back tothe time of rock-formation (i.e., calculating the “initial” Sr-isotopiccomposition) would result in an overcorrection of in situ 87Sr growthand would yield for many samples geologically unreasonably low ap-parent 87Sr/86Sr values. In contrast, a secondary decrease in 87Rb/86Srresults in an undercorrection of in situ 87Sr growth and, thus, wouldyield anomalously high apparent initial 87Sr/86Sr values. Several sim-ple inferences can be made from this kind of diagram: (i) distur-bances in the Rb–Sr system will be more readily apparent in rockswith high Rb/Sr than in rocks with low Rb/Sr. (ii) For a rock suite

1.2

1.1

1.0

0.9

0.8

0.7

0.61000 200 300

Age (Ma)

87S

r/86

Sr

400


500

1.2

1.1

1.0

0.9

0.8

0.7

0.61000 200 300

Age (Ma)

87S

r/86

Sr

400


500

B

D


in Erzgebirge


in Erzgebirge


in Erzgebirge


in Erzgebirge

Schwarzburg

SchwarzburgMetamorphic nappes(Erzgebirge)Late-Variscan granites(Erzgebirge)

l diagram showing the effect of high and low Rb/Sr (i.e., 87Rb/86Sr) ratios on the isotopicrease in Rb/Sr (by Sr loss or Rb gain) through time (i) results in an accelerated post-bly yields anomalously low apparent initial 87Sr/86Sr values when recalculated to ther metamorphic rocks). Arrowheads indicate direction of Sr evolution (opposite to them the Schwarzburg Anticline. C. Evolution of 87Sr/86Sr in the metamorphic nappes ofData from A, B and C are shown (gray) for comparison. Major geologic events thatertical bars); dark-gray for major events and light-gray for late hydrothermal events.unger) increase of the Rb/Sr value.


characterized by similar initial 87Sr/86Sr and similar low Rb/Sr ratios,samples that had acquired high Rb/Sr ratios by one secondary processwould provide a rough estimate for the time of disturbance, which isgiven by the intersection of the 87Sr/86Sr(t) evolution lines of the twotypes of samples. In addition, these high 87Sr/86Sr samples allow themagnitude of change in 87Rb/86Sr to be constrained. Such an estimateis not possible for multiple disturbances.

Among the data presented here (Tables 1 and 2; Supplementaryma-terial Tables A1, A2, and A3), quartz-rich arkoses, quartzites, and someof the particularly Al-rich shales seem to have lost Sr during later events.Furthermore, many of the Erzgebirge granites with extremely highmeasured 87Sr/86Sr values show anomalous 87Sr/86Sr values forrecalculation ages older than 150 to 200 Ma. The young ages indicatethat low-temperature tectonic reactivation in the Erzgebirge result-ing in the redistribution of metals in hydrothermal vein deposits(cf. Fig. 3) may have also upset the Rb/Sr system of the surroundingrocks.

5. Sr-isotopic composition of sedimentary rocks from theSchwarzburg Anticline

The measured Sr-isotopic composition of the mostly Palaeozoicshales from the Schwarzburg Anticline shows an extreme range in87Sr/86Sr from 0.7086 to N1.07 (Table 1; Figs. 4 and 5B). The excur-sion to high values occurs entirely in samples from the TremadocianFrauenbach Group. The remaining samples yield 87Sr/86Sr valuesranging from 0.7086 to 0.809 (Table 1). Recalculation of the Sr-isotopic composition to 340 Ma (i.e., the time of the Variscan meta-morphism in the Erzgebirge) shows that the Sr-isotopic compositionof the Frauenbach Group sedimentary rocks persists, and so is notdue to Rb/Sr fractionation during the very-low to low grade Variscanoverprint (Table 1). For most samples of the Frauenbach Group, theSr-isotopic compositon at 340 Ma is markedly more radiogenic (upto 0.745) than the Sr-isotopic composition of rocks from overlyingand underlying units (0.710 to 0.722; Table 1) that share the sameVariscan overprint.

The anomalous radiogenic Sr-isotopic composition of the FrauenbachGroup rocks at 340 Ma strongly suggests that its Rb/Sr signatureis a primary feature rather than one that had been aquired duringVariscan low-grade metamorphism and deformation (Fig. 4). Thisinference is supported by the correlation of K2O/Na2O with Rb/Srand the Sr-isotopic composition (Fig. 4). While the correlation be-tween Rb/Sr and 87Sr/86Sr reflects in situ 87Sr-growth, the correla-tion between K2O/Na2O and Rb/Sr is mineralogically controlled.High values for these ratios are not due to high K and Rb contentsalone, but to exceptionally low Na2O and Sr contents, commonlyless than 0.6 wt.% Na2O and 60 ppm Sr, respectively. Shales andslates above and below the Frauenbach Group sedimentary rocksfall within the compositional range characteristic of shales (3.0–6.0 wt.% K2O; 0.5–4.0 wt.% Na2O, 80–300 ppm Sr; see also Ronovand Migdisov, 1996). The fractionation of K from Na and Rb from Sris a signature typically associated with intense chemical weathering(e.g., Nesbitt et al., 1982, 1996) that is known to have occurred inthe source area of the sedimentary rocks of the Frauenbach Group(Linnemann et al., 2000; Avigad et al., 2005). It represents a primarygeochemical fingerprint causing the highly anomalous Sr-isotopiccomposition of these rocks.

6. Sr-isotopic characterization of rocks from the Erzgebirge

6.1. Metamorphic rocks

The analyzed samples originate from tectonic units that experi-enced a wide range of metamorphic conditions, ranging from greens-chist facies to HP-LT (N12 kbar, 550 °C) and UHP-HT (N22 kbar, 830 °C)conditions (cf. Rötzler, 1995; Willner et al., 1997; Rötzler et al., 1998;

Mingram and Rötzler, 1999; Rötzler and Plessen, 2010). The measured87Sr/86Sr values fall into two groups with (i) low 87Sr/86Sr values and(ii) very high 87Sr/86Sr values (cf. Table 2 and Supplementary materialTable A2; Fig. 5C). This bimodal distribution of 87Sr/86Sr closely mimicsthe pattern observed in the Palaeozoic sedimentary rocks of theSchwarzburg Anticline that are the precursor lithologies of the meta-morphic nappes of the Erzgebirge (cf. Mingram, 1996; Mingram,1998; Rötzler and Plessen, 2010). Our Sr-isotopic data demonstrate:(i) all rocks with a Frauenbach signature (low Ca, Na, and Sr) havehigh 87Sr/86Sr values, whereas all others do not. There is a one-to-onecorrelation between major element signature and measured 87Sr/86Sr.(ii) All high 87Sr/86Sr rocks also have high 87Sr/86Sr values at 340 Ma,i.e., at the time of metamorphism. This implies that the Rb/Sr signatureis a signature of the sedimentary protolith and has not been superim-posed during metamorphism. Instead, the close correspondenceof the Sr-isotopic composition of Frauenbach Group sedimentaryrocks and their metamorphic equivalents, i.e., the broad overlap ofthe Sr evolution trend of metamorphosed and unmetamorphosedFrauenbach rocks (Fig. 5C), suggests that the Rb/Sr ratio was notaffected by more than a few percent even during high-grade meta-morphism. Note, the geochemical fingerprinting using the Sr-isotopic composition of one particular lithologic unit (equivalent tothe sedimentary rocks of the Frauenbach Group) would not havebeen possible using the Nd-isotopic composition of rocks of theErzgebirge (Mingram and Rötzler, 1999; Tichomirowa, 2003) thatfall in the same range as those of the Schwarzburg Anticline(Hahne and Romer, 2008; Romer and Hahne, 2010). The eNdT valuesrange from −8 to −10.5 for clastic sedimentary rocks that haveessentially no contribution of juvenile volcanic material, and from−7 to −4 for sedimentary rocks with significant input of suchmaterial.

6.2. Magmatic rocks

The present Sr-isotopic compositions of the granites plot in twogroups (Fig. 5D; Gerstenberger et al., 1983, 1984, 1995; Gerstenberger,1989; Förster and Romer, 2010). One group shows measured 87Sr/86Srvalues ranging from c. 0.710 to 0.810 and yields initial 87Sr/86Srvalues (T=325 Ma) ranging from c. 0.705 to 0.713 (Supplementarymaterial Table A3). The second group shows measured 87Sr/86Srvalues in the range c. 0.750 toN6.1 (Supplementary material TableA3). The recalculation of these radiogenic compositions to the timeof granite emplacement results, for some samples, in geologicallyreasonable initial 87Sr/86Sr values that fall in the same range as theinitial Sr-isotopic compositions of the low-87Sr/86Sr granites. For agreat number of granites, however, the recalculated initial 87Sr/86Srratios have geologically unreasonably low values that can not beaccounted for by uncertainties (and bias) in the determination ofthe Rb/Sr ratio, but require a two-stage Sr evolution, with a second-ary increase in the Rb/Sr ratio of the rock.

7. Discussion

7.1. Extreme Sr signatures in the metamorphic nappes of the Erzgebirge

Rocks that according to their geochemical composition have beenclassified as metamorphic equivalents of the Frauenbach Group showhighly radiogenic Sr-isotopic compositions for 340 Ma (87Sr/86Sr=(0.719) 0.738 to 0.792; Table 2); all rocks that have the geochemicalsignatures of other units show less radiogenic Sr (87Sr/86Sr=0.709 to0.722; Table 2). The highly radiogenic composition of Sr at 340 Maimplies that the high Rb/Sr ratios, which are necessary to produce high-ly radiogenicmeasured Sr-isotopic compositions (87Sr/86SrN0.770), arenot the result of metamorphic elementmobility. Instead, the high Rb/Srrepresents a feature of the protolith. The muscovite gneisses, which arederived from rhyolitic intercalations in the Ordovician sedimentary


rocks of the Frauenbach Group (cf. Mingram, 1996, 1998), are also char-acterized by high 87Sr/86Sr values at 340 Ma (0.733–0.757) and at pre-sent (0.846 to 0.968; Table 2). Our Sr-isotope data (Table 2)demonstrate two features (Figs. 5B and 5C): (i) Metamorphism didnot result in isotopic homogenization. Rocks that had an anomalousSr-isotopic composition before metamorphism retained an anoma-lous Sr-isotopic composition after metamorphism. (ii) Metamor-phism does not necessarily alter the Rb/Sr ratios of rocks to asignificant extent. The effect of sampling-scale, i.e., resetting of theRb–Sr system of minerals while the Rb–Sr system of rocks apparentlyremains unaffected, is well-known (e.g., Roddick and Compston,1977) and is the reason why WR-isochrons (e.g., Krentz, 1985; Sup-plementary material Table A2) yield dates closer to the age of the de-position of the sedimenteray protoliths than to the age of Variscanmetamorphism. The high Rb/Sr values represent a primary featureand the high 87Sr/86Sr340 values are a direct consequence of this.Thus, the Sr-isotopic composition of metasedimentary rocks fromthe Erzgebirge may be used as a geochemical fingerprint in a similarway as the chemical composition of the rocks (Mingram, 1996,1998).

7.2. Sr-isotopic composition of the late-Variscan granites and post-Variscanmedium temperature hydrothermal deposits of the Erzgebirge

The contrasting Rb/Sr ratios among late-Variscan granites, even-tually giving rise to the bimodal distribution of measured 87Sr/86Srvalues, is due to particularly low Sr contents in one group of granitesrather than to a broad variation of Rb content. Recalculation of theSr-isotopic composition results in geologically unreasonably lowrecalculated 87Sr/86Sr values for many granites with high measured87Sr/86Sr values (Fig. 5D and Supplementary material Table A3). Thisis incompatible with a single-stage Sr-isotope evolution, but insteadrequires two ormore evolution stages with increasingly higher Rb/Srratios in later stages.

Gerstenberger (1989) modeled the anomalous Sr-isotopic com-position of the late Variscan granites using a two-step model, assum-ing that (i) all granites evolved from the same magma and, thus, hada similar initial Sr isotopic compositions, and (ii) the 87Rb/86Sr of thesamples was subsequently increased by addition of Rb during “auto-metasomatism”. Such a two-step model gave an age of about 300 Mafor the increase in 87Rb/86Sr for most samples. This age broadly cor-responds with the younger, volumetrically subordinate phase ofmagmatism in the Erzgebirge and, therefore, was thought to be geo-logically relevant.

This model, however, faces several problems. (i) The increaseof the Rb/Sr ratios was thought to have occurred during enrichmentin a long-lived magma chamber when the extraction of the oldergranites at c. 325 Ma resulted in an increase in the Rb/Sr ratio of the re-sidual melt. The Rb/Sr ratio of the residual melt was additionallyincreased by magma internal processes that eventually led to ahigh Rb/Sr melt being parental to the granites emplaced at c.300 Ma. Shallow (200–300 MPa) magma chambers, however, do notpersist for more than 20 Ma. (ii) In the mid-1990s, it was generallyaccepted that granitic magmatism in the Erzgebirge occurred duringtwo major events (e.g., Gerstenberger, 1989; Tischendorf and Förster,1990; Gerstenberger et al., 1995), which produced older and youngergranite suites at about 335–325 Ma and 315–295 Ma, respectively.Extensive U–Pb dating over the last decade, however, showedthat felsic magmatism (with the exception of a few minor bodiesand subvolcanic intrusions) occurred between 327 and 318 Ma(Förster et al., 1999; Romer et al., 2007, 2010b; Förster and Romer,2010; Tichomirowa and Leonhardt, 2010). Furthermore, the recalcu-lated Sr-isotopic composititions of many of the high-87Sr/86Sr0 gran-ites show unreasonably low values for a timemuch younger than theemplacement of the granite. Thus, the proposed model of autometa-somatic increase in Rb/Sr does not work.

Since the high Rb/Sr signature of the granites is not a primary fea-ture, but instead reflects a younger disturbance, the Rb/Sr systematicsof the granites may have been upset much later than originally mod-eled by Gerstenberger (1989). For instance, the disturbance may berelated to the formation of hydrothermal vein-type mineralization(at c. 275 Ma) and the multiphase overprint and metal redistributionin these deposits at c. 180 Ma, c. 150 Ma, and c. 120 Ma (cf. Förster,1996; Romer et al., 2010a). Such a later disturbance is particularlyappealing as some of the elements accumulated in the medium-temperature vein deposits (e.g., U, Pb) have been leached from thegranites (e.g., Barsukov et al., 1996, 2006; Förster, 1999). This leach-ing may have been associated with the decomposition of feldspar toform sericite and clay minerals that concomitantly would haveresulted in the preferential loss of Sr over Rb and an increase of Rb/Sr.Furthermore, if the time of disturbance is not constrained, then nor isthe Sr-isotopic composition of the granites at that time. Thus, the vari-ous granitesmay have already encompassed awide range of Sr-isotopiccompositions at the time of emplacement.

On the assumption that the disturbance of the Rb/Sr system ofthe granites is related to the 275 Ma hydrothermal event and thatsubsequent disturbances affected individual samples to only aminor extent, the Sr-isotopic compositions have been calculatedfor t=275 Ma from the measured 87Sr/86Sr and 87Rb/86Sr ratios(Supplementary material Table A3). The 87Sr/86Sr values obtainedfor t=275 Ma show a bimodal distribution, falling in the ranges c.0.710 to 0.724 and Nc. 0.74 (Supplementary material Table A3).Several of the granites of the second group have 87Sr/86SrT valueseven higher than the Frauenbach Group sedimentary rocks andtheir metamorphic equivalents. This simple modeling of the Sr-isotopic evolution of granites with high measured 87Sr/86Sr values(Fig. 5D) has two important implications. (i) The longer the pre-disturbance Sr-isotopic evolution, i.e., the younger the disturbance(increase) of the Rb/Sr ratio, the smaller are the changes in 87Rb/86Srnecessary to account for the observed Sr-isotopic composition and thehigher were the original Rb/Sr values. (ii) The Sr-isotopic composi-tion of the granites (Fig. 5D) may have already been highly radiogen-ic at a time of the disturbance. The initial Sr-isotopic composition ofthese granites is not constrained. Thus, the high-87Sr/86Sr granitesmay have had different source rocks than the granites that evolvedto relatively unradiogenic Sr-isotopic compositions. Sedimentaryrocks of the Frauenbach Group and their metamorphic equivalents,which have radiogenic 87Sr/86Sr325, may have contributed signifi-cantly to them.

7.3. Frauenbach Group volcanic and sedimentary rocks as granite source

The late Variscan granites have initial Nd-isotopic compositions(typically eNdT –1 to −6; Förster and Romer, 2010) that fall betweenthe ranges defined by the Palaeozoic sedimentary rocks and interca-lated juvenile volcanic rocks of the Gondwana shelf (e.g., Linnemannand Romer, 2002; Linnemann et al., 2004; Romer and Hahne, 2010).Thus, the initial Nd-isotopic composition of the various granitesdoes not necessarily require contributions from mantle and crustalsources, but may also be accounted for by melting of the subductedPalaeozoic cover rocks of the Gondwana shelf (see also Förster andRomer, 2010). Some of the late-Variscan Erzgebirge granites havehigh Rb/Sr ratios and high contents of, e.g., Li, Rb, Cs, and Sn thatcould have been obtained either by residual enrichment followingextensive magmatic fractionation or by melting of source rocks thathad already been enriched in these elements.

Even the least evolved granites belonging to this suite of late-Variscan granites have relatively high SiO2 contents of ~73 wt.%. Themelting of siliciclastic sedimentary rocks and intercalated acidic vol-canic material could account for both the high SiO2 contents of thegranites and the enrichment in incompatible elements. The silici-clastic and volcanic rocks of the Frauenbach Group (and its

Quartz

OrthoclaseAlbite

Shale (Schwarzburg)

Erzgebirge granites

Greywackes, arkoses,and sandstones)(Autochthonous Domain)

950

750

800

850

900

1000

1050

800800

850 900 950

30 kbar20 kbar

10 kbar

2 kbar

5 kbar

XAn

boronfluorinepressure

ternary eutectic or minimum

Fig. 6. Phase relations in the synthetic system quartz–albite–orthoclase–H2O, showingthe pressure dependence of the boundary between quartz and alkali feldspar (cotecticline) and the composition of the minimum melt (ternary eutectic and ternary mini-mum) (thin lines; Johannes and Holtz, 1996). In addition to pressure, the position ofthe cotectic line and the eutectic also depends on the addition of network-modifyingelements, in particular Li, B, and F (e.g., Webster et al., 1997; Thomas et al., 2003)and XAn of albite. The temperature of melting (dotted lines) is shown for 200 MPa(Winkler, 1967). Granitic melts from the Erzgebirge plot generally close to the eutectic;some granite samples show systematic deviations to lower modal quartz. The major-element composition of the sedimentary rocks from Schwarzburg and their felsic andmafic volcanic intercalations is projected into the diagram (data from Linnemann andRomer, 2002; Romer and Hahne, 2010). Note that sedimentary and felsic to mafic vol-canic rocks have compositions that do not produce large amounts of melt, whereas the“mixture” of these rocks, for instance the intercalation of volcanic and sedimentaryrocks, has compositions that allow for larger amounts of melting.


metamorphic equivalents, the feldspar-free schists and muscovite-gneisses of the Garnet–Phyllite, Mica Schist–Eclogite, and Gneiss–Eclogite units, respectively; e.g., Mingram, 1998; Mingram andRötzler, 1999; Rötzler and Plessen, 2010), represent source rockswith high Rb/Sr ratios and high contents of Li, Rb, Cs, and Sn (e.g.,Maaß et al., 1986; Mingram, 1998; Lützner et al., 2001; Rötzler andPlessen, 2010; Romer and Hahne, 2010).

The geochemical fingerprints of source rocks in the late-Variscangranites may be affected by (i) selective melting and assimilation ofthe source rocks and (ii) magma evolution by assimilation and frac-tionation. Melting of the source rocks is not a bulk process. Instead,minerals defining the eutectic point of the rock would have beenpreferentially melted and trace-element signatures bound to theseminerals transferred to the melt (e.g., Zeng et al., 2005; Vásquezet al., 2009). Partial melting does not preserve the characteristicK–Na–Ca signature of the protoliths, but results in K/Na ratios cor-responding to the eutectic of the granitic melt (e.g., Montel andVielzeuf, 1997), and the amount of melt produced largely dependson the breakdown of the hydrous phases muscovite, biotite, andamphibole (e.g., Clemens and Vielzeuf, 1987; Vielzeuf and Montel,1994; Montel and Vielzeuf, 1997). Highly incompatible elementsmay become preferentially partitioned into the melt and enrichedrelative to the protolith. Such behavior might be expected for Sn,Ta, Nb, and W (e.g., Linnen, 1998, 2006; Bhalla et al., 2005), andalso for F, Li, Rb, and Cs (e.g., Webster et al., 1997). Trace-element fin-gerprints that are dominated by the restitic phases will not be trans-ferred to the melt, but remain with the restite. Magma evolutioninvolving fractional crystallization will additionally change themajor and trace-element composition of the melt, in particular forthose elements that are compatible in early crystallizing minerals.The combination of selective mobilization and fractional crystalli-zation camouflage the geochemical link between granite and proto-liths. The most reliable tracers for the prominent contribution ofa particular protolith to the granitic melt are exotic isotopic compo-sitions (as for instance the highly radiogenic Sr-isotopic compositionof the sedimentary and volcanic rocks of the Frauenbach Group) andanomalously low contents of elements that behave incompatible inminerals crystallizing from the melt. Low contents of melt-compati-ble elements would reflect a protolith that also had low contents ofthese elements. In contrast, anomalously high contents of melt-com-patible elements are not perticularly indicative of the protolith sincehigh contents may also be obtained during magmatic processes suchas partial melting and fractionation.

Figs. 6 and 7 show the relationship between likely magmatic andsedimentary protoliths (and their metamorphic equivalents) andlate-Variscan granites. Note that the granites have relatively smallvariations in Na2O and K2O, i.e., albite and orthoclase (Fig. 6),which is due to eutectic melting, and have much higher Na contentsthan most sedimentary rocks, in particular the sedimentary rocks ofthe Frauenbach Group. The selective melting of one particular unitoccurs at a much higher temperature than that of mixtures of sedi-mentary and volcanic rocks. Such a mixture is not only in line withthe Nd-isotopic composition of the granites, but by invoking differ-ent portions of sedimentary and volcanic protoliths and the contrast-ing involvements of different sediment types, would also account forthe range in eNdT and the contrasting trace-element signatures ofthe granites. The siliciclastic sedimentary rocks of the Schwarzburgarea show flat “average crust”-normalized trace element patternsfor the REE and immobile elements and deep minima for Sr and Pbfor all units (Fig. 7A). The flat trends have normalized values slighlyabove 1 because of the lower quartz-content of the sedimentaryrocks. The troughs reflect the low contents of carbonates in thesemarine sedimentary rocks (main Sr carrier) and the decompositionof K-feldspar in the sediment source (with subsequent loss of thePb released from K-feldspar). Geochemical diversity among the sed-imentary rocks is mainly due to the rocks of the Silurian units and the

Ordovician Frauenbach Group (Fig. 7A). The Silurian sedimentaryrocks are rich in organic material and were deposited at a relativelylow sedimentation rate. Therefore, redox-sensitive elements scav-enged from the seawater, such as U, Mo, and V, are particularlyenriched. There is also a weak positive Ba anomaly (Fig. 7A; cf.Romer and Hahne, 2010). In contrast, the rocks of the FrauenbachGroup, which were largely derived from a profoundly weatheredsource, have high contents of Rb, Cs, Sn, W, and Tl and are depletedin Ni (Fig. 7A).

The trace-element pattern of the different late-Variscan granitetypes is highly variable, except for the general depletion of Sc, V, Cr,and Ni that seem to reside in silicate minerals of the restite or havebeen scavenged by sulfides in the melt source (Fig. 7B). To avoidmodifications of the trace-element signatures of the melt source bycrystal fractionation, only the least evolved granites of each grouphave been used for Fig. 7B. The non-mineralized and Sn-specific gran-ites have different “average crust”-normalized trace element pat-terns. The non-mineralized granites have relatively flat patterns forZr, Nb, Ba, and REE (Fig. 7B). They are characterized by a trough inMo, variable peaks in Li, Rb, and Cs, and a saw-tooth pattern forTa–W–Tl–Pb–Bi–Th–U, with local troughs for W and Bi. They haveno or only weak negative Eu-anomalies (Fig. 7B). Most remarkably,they have no or only a relatively small Sr trough. The Sr and W signa-tures indicate that volcanic and sedimentary rocks of the FrauenbachGroup did not significantly contribute to the source of these granites.The Sn-specific granites have irregular patterns that can be addition-ally distinguished on the basis of their HREE signature (Fig. 7B) intothose with HREE andMREE enriched normalized patterns, respective-ly. Common to the Sn-specific granites are deep troughs for Sr, Ba, andEu, distinct peaks in Rb, Ca, Li, and Sn, and a saw-tooth pattern forTa–W–Tl–Pb–Bi–Th–U, with local troughs for Pb and Th.

Fig. 7. Geochemical variation diagrams for (A) sedimentary rocks (mostly shales and slates) of the Schwarzburg area and (B) late-Variscan granites of the Erzgebirge, normalized toaverage continental crust (Rudnick and Gao, 2003). The sedimentary rocks represent average values (full data set in Romer and Hahne, 2010; complemented with unpublished datafor Li, B, Y, Hf, and Bi). The granite data represent the least evolved samples of geochemically and mineralogically different granite types. The least evolved samples (rather thanaverage values) have been used to minimize effects of fractional crystallization on the geochemical fingerprint. Compositional range of sedimentary rocks is shown in B (gray)for reference. For the Schwarzburg area, averages of the following lithological packages are shown: GRA Untere Graptolitenschiefer and Obere Graptolitenschiefer formations;PHY/GRÄ Phycodes and Gräfenthal groups; FRA Frauenbach Group; GOL Goldisthal Formation. For the late Variscan granites, samples from the following intrusions are shown:NBZ Niederbobritzsch; BRG Bergen; TLH Tellerhäuser; SAD Sadisdorf; GOT Gottesberg.


The important feature of the “average crust”-normalized traceelement patterns is the general resemblance of the Sn-specific gran-ites to the sedimentary rocks of the Frauenbach Group and the relatedintercalations of felsic volcanic rocks. This resemblance implies thatthe Sn-specific granites may have received higher contributions ofthese source rocks than the non-mineralized granites. As thesetwo particular source lithologies had distinctly more radiogenic Sr-isotopic compositions at the time of granite formation than theother units of volcano-sedimentary rocks (cf. Fig. 4), the Sn-specific

granites would be expected to have higher inital 87Sr/86Sr values thanthe non-mineralized granites. For the Erzgebirge, however, the exten-sive and repeated post-emplacement hydrothermal mineralization(Fig. 3) makes it impossible to test this hypothesis directly, with thepossible exception of the local occurrence of greisen fluorite withhigh 87Sr/86Sr values (e.g., Höhndorf et al., 1994). Granites from otherVariscan areas that have not been extensively overprinted by laterhydrothermal activity, show a similar bipartite grouping of granitechemistry, with one group having distinctly lower contents of Ca, Sr,


Y, and REE than the other (cf. Siebel et al., 2008; Villaseca et al. 2008).The granites depleted in Sr, Ca, Y, and REE consistently show higher87Sr/86Sr values than the others, e.g., 87Sr/86Sr0=0.712–0.742 vs.0.752–0.919 (SW Bohemian Massif; Siebel et al., 2008) and87Sr/86Sr0=0.727–0.747 vs. 0.760–0.856 (Iberian Massif; Villasecaet al., 2008).

7.4. Recycling of high 87Sr/86Sr rocks in the erosional debris of the Variscanorogen

The upper part of the Schwarzburg profile contains Variscanmolasses, i.e., the erosional debris of the exhumed Variscan orogen.The deposition age of these sedimentary rocks has been estimatedfrom biostratigraphic correlation and the time scale of Menningand German Stratigraphic Commission (2002). There are two dis-tinct excursions to more radiogenic Sr-isotopic compositons atthe time of deposition, a small one in the 362 to 358 Ma sedimen-tary rocks and a larger one starting at c. 338 and 339 Ma (Fig. 4).These excursions reflect the involvement of a sediment sourcewith high 87Sr/86Sr values, such as the metamorphic equivalentsof the sedimentary rocks of the Frauenbach Group. The onset of thisshift to high 87Sr/86Sr values dates the time when this sediment sourcebecame available for erosion, whereas its disappearance dates the timewhen it became subordinate or had disappeared. The two excursions tohigh 87Sr/86Sr values in the sedimentary rocks (Table 1; Fig. 4) corre-spond to the two phases of nappe emplacement in the Saxo-ThuringianZone. The excursion at 362 to 358 Ma corresponds to the exhumationof the Uppermost Allochthonous Units (Hahn et al., 2010; Klemd,2010) that had their metamorphic peak at c. 370 Ma (Klemd, 2010),whereas the peak starting at c. 338–339 Ma corresponds to the exhu-mation of the high-grade metamorphic complexes of the Erzgebirgeand the Saxon Granulite Massif (e.g., Kröner et al., 1995; Kröner andWillner, 1998; Romer and Rötzler, 2001, 2003; Rötzler et al., 2004)that had reached metamorphic peak conditions at 340 Ma. The smalldifference between the age of the metamorphic peak and the age ofthe appearance of high 87Sr/86Sr signatures indicates that the meta-morphic units became available for erosion shortly after orogenicexhumation started. The short duration – less than 5 Ma for the olderexcursion to high 87Sr/86Sr values (the duration of the younger excur-sion is not well known because of an erosional gap) – may indicaterapid leveling of the orogen (see also Roscher and Schneider, 2006;Schneider and Romer, 2010). The re-appearance of the distinctive87Sr/86Sr isotope signature of the Frauenbach Group rocks – althoughin strongly diluted form – in the molasse of the Variscan orogen dem-onstrates the persistance of extreme isotopic compositions that mayremain traceable through an entire orogenic cycle.

7.5. Regional distribution of Palaeozoic and younger high 87Sr/86Sr rocks

Chemical weathering is particularly pronounced in areas with lowrelief and low erosion rate and is promoted by high temperature,organic acids, and abundant water (e.g., White et al., 1999; Kumpet al., 2000; Viers et al., 2000). In the Cambrian and Ordovician, chem-ical weathering was not assisted by land plants, which appearedmuch later in geological time, but may have been supported by amore aggressive and corrosive composition of the atmosphere (cf.Avigad et al., 2005). Early Palaeozoic sedimentary rocks with a geo-chemical signature of profound palaeo-weathering occur throughoutnorthern Gondwana and terranes that had been part of northernGondwana at that time (e.g., Noblet and Lefort, 1990; Beetsma,1995; Mingram, 1998). The rifting of Gondwana resulted in the upliftof the rifted margins and the redistribution of the weathering veneeratop Gondwana crust into the rift basins, where they formed volumi-nous sedimentary packages that were deposited at high sedimentationrates, but still carried the weathering signature (e.g., sedimentary rocksof the Frauenbach Group; Maaß et al., 1986; Lützner et al., 1986). Since

sedimentary rocks with comparable geochemical fingerprints andcorresponding age are absent on other cratons, the particular geo-chemical compositon of these early Palaeozoic sedimentary rocksmay be used as a paleogeographic provenance indicator for terranesof disputed origin.

The geochemical weathering signature is present in Cambrian toLower Ordovician sedimentary rocks deposited on the basement ofnorthern Gondwana (e.g., Beetsma, 1995; Nägler et al., 1995; Mingram,1998; Avigad et al., 2005) and may be present in Cambrian to LowerOrdovician sedimentary rocks derived from Gondwana, but deposited –

in the case of the Meguma Terrane, for example – on crust of differentprovenance (e.g., Romer et al., 2011), particularly during the early stagesof opening of the Rheic Ocen, when the ocean was still narrow (e.g.,Waldron et al., 2009). This geochemical signature persists over awide range of metamorphic overprint (e.g., Mingram, 1998; Mingramand Rötzler, 1999; Rötzler and Plessen, 2010) and may be inherited bygranitic rocks in form of anomalous Rb/Sr, Li, Rb, Cs, and Sn contentsand 87Sr/86Sr ratios.

Early Ordovician sedimentary and volcanic rocks of comparablegeochemical composition to the rocks of the Frauenbach Group andtheir metamorphic equivalents (cf. Mingram, 1996, 1998; Romerand Hahne, 2010) are also known from (i) felsic granulites of theMoldanubian Zone (Bohemian Massif; e.g., Janoušek et al., 2004) —

these (possibly anatectic) granulites, interpreted to be derived frommagmatic precursors, show a bimodal distribution of measured87Sr/86Sr that correlates with the geochemical fingerprint (low con-tents of Ca, Sr, SREE, Zr for rocks with high 87Sr/86Sr); (ii) the Variscanbasement of Armorica and the French Massif Central (e.g., Bernard-Griffiths et al., 1985; Turpin et al., 1990); and (iii) Palaeozoic sedi-mentary rocks from the Iberian Variscan Belt of Spain (Beetsma,1995; Nägler et al., 1995; Villaseca et al., 2008, and referencestherein).

The bipartite geochemical signature of late-Variscan granites, withmore or less distict depletions of Sr, Ca, Y, and REE in combinationwith with contents of Li, Rb, Sn, and Cs in one group of granites andthe bipartite distribution of 87Sr/86Sr values that correlates with thegeochemical fingerprint, indicates that these geochemical finger-prints are inherited from the protoliths. Thus, the distribution of gran-ites with these signatures reflects the distribution of rocks having thegeochemical signature of the Frauenbach Group, i.e., rocks having a“Gondwana” fingerprint. Late-Variscan granites with this signatureare known from the French Massif Central and the Armorican Massif(e.g., Turpin et al., 1990; Stussi 1989), the Iberian Variscan Belt ofSpain (Villaseca et al., 2008), the Meguma terrane of Nova Scotia(e.g., Dostal and Chatterjee, 2000; Dostal et al., 2004), and possiblythe Cornubian Batholith (e.g., Darbyshire and Shepherd, 1994).

The presence of the “Frauenbach” geochemical fingerprint in sed-imentary rocks above the Cadomian basement of central and westernEurope and the persistence of this signature during metamorphismsuggests that this geochemical signature may be used as a tracer ofGondwanan provenance. Close spatial relation between late-VariscanSn-specialized granites and Gondwana-derived sedimentary rockswith Frauenbach signatures may indicate that Frauenbach-typerocks contributed significantly to this granite type and, furthermore,that the recycling of these sedimentary rocks in crustal melts resultsin granites of geochemically distinct composition.

8. Summary

During the early stage of the opening of the Rheic Ocean, the upliftof the rifted Gondwana margin resulted in the erosion of the inten-sively chemically weathered cover of northern Gondwana and thedeposition of this debris in graben structures at the craton margin.Among the Palaeozoic sedimentary rocks of the Gondwana shelf,the redeposited chemically-weathered sedimentary rocks are uniquesince they developed with time an anomalously radiogenic Sr-


isotopic composition (87Sr/86Sr up to 0.9–1.1). Sedimentary rockswith this signature may have been deposited on Gondwanan crust(i.e., the southern margin of the Rheic Ocean) or may have been de-posited in sedimentary fans, in which case they may also have beendeposited on non-Gondwanan crust. A recent example (although ina different plate-tectonic setting) would be the depositional fan ofGanges and Brahmaputra that consists dominantly of Asian debrisdeposited on oceanic crust of the Indian plate.

The early Tremadocian siliciclastic rocks (e.g., Frauenbach Group)represent the only sedimentary unit within the entire Palaeozoic shelfsequence of Saxo-Thuringia with an anomalous Sr-isotopic composi-tion. These rocks had already acquired a distinctive Sr-isotopic com-position at the time of the Variscan orogeny. Thus, this anomaly – incombination with the geochemical composition of the rock (and pos-sibly the age of detrital zircon, e.g., Bahlburg et al., 2010) – provides ageochemical fingerprint that persisted through high-grade metamor-phism. It represents a feature pertinent to the protolith rather than tothe metamorphic overprint. The oldest post-collisional sedimentaryrocks show a distinct increase in 87Sr/86Sr, implying that the meta-morphic equivalents of Frauenbach Group sedimentary rocks becameavailable for erosion. Although the Frauenbach signature in the post-collisional sedimentary rocks is diluted by Sr from other metamorphicunits, the shift in 87Sr/86Sr indicates that about 20% of the Sr in thesesedimentary rocks is derived from sandstones, arkoses, and shales ofthe Frauenbach Group and their metamorphic equivalents. Because ofthe markedly lower Sr contents of this unit, this implies that theserocks possibly account for more than 40–50% of the post-orogenicsedimenary debris. A variably distinct Frauenbach signature alsooccurs in some post-Variscan granites of the Erzgebirge that repre-sent crustal melts. The Frauenbach signature is not restricted to theSaxo-Thuringian Zone. It represents a Gondwana signature thatoccurs in certain early Palaeozoic sedimentary rocks throughout cen-tral and western Europe and may provide palaeogeographic con-straints for sedimentary units deposited on unknown basement orhigh-grade metamorphic units of disputed provenance.

Supplementarymaterials related to this article can be found onlineat doi:10.1016/j.gr.2011.09.005.

Acknowledgments

We thank Birgit Plessen for discussion of various aspects of thetectonostratigraphy of the Erzgebirge and for making available unpub-lished Sr isotope data and rock powders of samples originally presentedby Mingram (1996, 1998). We also thank R. Naumann, H. Rothe, S.Tonn, M. Ospald, and H. Liep (Deutsches GeoForschungsZentrumGFZ) for sample preparation for chemical analysis and C. Schulz(Deutsches GeoForschungsZentrum GFZ) for separation of Sr for iso-topic analysis. We greatly appreciate the help of A. Hendrich(Deutsches GeoForschungsZentrum GFZ) with Figs. 1, 7, and A1.We gratefully acknowledge careful and detailed reviews by FritzFinger (Salzburg) and an anonymous reviewer, and thoughtful edi-torial comments by Damian Nance.

References

Amelin, Y., Zaitsev, A.N., 2002. Precise geochronology of phoscorites and carbonatites:the critical role of U-series disequilibrium in age interpretations. Geochimica etCosmochimica Acta 66, 2399–2419.

Avigad, D., Sandler, A., Kolodner, K., Stern, R.J., McWilliams, M., Miller, N., Beyth, M.,2005. Mass-production of Cambro-Ordovician quartz-rich sandstone as a conse-quence of chemical weathering of Pan-African terranes: environmental implica-tions. Earth and Planetary Science Letters 240, 818–836.

Bahlburg, H., Vervoort, J.D., DuFrane, S.A., 2010. Plate tectonic significance of MiddleCambrian and Ordovician siliciclastic rocks of the Bavarian Facies, Armorican Ter-rane Assemblage, Germany — U–Pb and Hf isotope evidence from detrital zircons.Gondwana Research 17, 223–235.

Barsukov, V.L., Sokolova, N.T., Ivanitskii, O.M., 1996. Distribution of U, Th, and K in granitesof the Aue and Eibenstock massifs, Erzgebirge, Germany. Geochemistry International34, 1041–1056.

Barsukov, V.L., Sokolova, N.T., Ivanitskii, O.M., 2006. Metals, arsenic, and sulphur in theAue and Eibenstock granites, Erzgebirge. Geochemistry International 44, 896–911.

Bauluz, B., Mayayo, M.J., Fernandez-Nieto, C., Lopez, J.M.G., 2000. Geochemistry of Pre-cambrian and Paleozoic siliciclastic rocks from the Iberian Range (NE Spain): im-plications for the source-area weathering, sorting, provenance, and tectonicsetting. Chemical Geology 168, 135–150.

Beetsma, J.J., 1995. The Late Proterozoic/Paleozoic and Hercynian crustal evolution ofthe Iberian Massif, N Portugal: as traced by geochemistry and Sr–Nd–Pb isotopesystematics of pre-Hercynian terrigenous sediments and Hercynian granitoids.PhD Thesis, Frije Universiteit Amsterdam, 223 pp.

Begemann, F., Ludwig, K.R., Lugmair, G.W., Min, K., Nyquist, L.E., Patchett, P.J., Renne,P.R., Shih, C.-Y., Villa, I., Walker, R.J., 2001. Call for an improved set of decay con-stants for geochronological use. Geochimica et Cosmochimica Acta 65, 111–121.

Bernard-Griffiths, J., Peucat, J.J., Sheppard, S.M.F., Vidal, Ph., 1985. Petrogenesis of leu-cogranites from the Southern Armorican Massif: contribution of REE and isotopic(Sr, Nd, Pb and O) geochemical data to the study of source rock characteristicsand age. Earth and Planetary Science Letters 74, 235–250.

Bhalla, P., Holtz, F., Linnen, R.L., Behrens, H., 2005. Solubility of cassiterite in evolvedgranitic melts: effect of T, fO2, and additional volatiles. Lithos 80, 387–400.

Clemens, J.D., Vielzeuf, D., 1987. Constraints on melting and magma production in thecrust. Earth and Planetary Science Letters 86, 287–306.

Darbyshire, D.P.F., Shepherd, T.J., 1994. Nd and Sr isotope constraints on the origin of theCornubian batholith, SW England. Journal of the Geological Society 151, 795–802.

Dostal, J., Chatterjee, A.K., 2000. Contrasting behaviour of Nb/Ta and Zr/Hf ratios in aperaluminous granitic pluton (Nova Scotia, Canada). Chemical Geology 163,207–218.

Dostal, J., Chatterjee, A.K., Kontak, D.J., 2004. Chemical and isotopic (Pb, Sr) zonation ina peraluminous granite plution: tole of fluid fractionation. Contibutions to Miner-alogy and Petrology 147, 74–90.

Drost, K., Romer, R.L., Linnemann, U., Kraft, P., Fatka, O., Marek, J., 2007. Nd–Sr–Pb isoto-pic signatures of Neoproterozoic–Early Paleozoic siliciclastic rocks in response tochanging geotectonic regimes: a case study from the Barrandian area (BohemianMassif, Czech Republic). In: Linnemann, U., Nance, D., Kraft, P., Zulauf, G. (Eds.),The Evolution of the Rheic Ocean: From Avalonian-Cadomian Active Margin toAlleghenian-Variscan Collision: Geological Society of America, Special Paper, 423,pp. 191–208.

Erdtmann, B.D., 1991. The post-Cadomian Early Palaeozoic tectonostratigraphy ofGermany (Attempt at an analytical review). Annales de la Société Géologiquede Belgique 114, 19–43.

Falk, F., Hahne, K., Lützner, H., Gehmlich, M., 2000. Der Beginn siliziklastischer Sedimen-tation im Altpaläozoikum des Saxothuringikums: Sedimentologie und Geochemiedes Richtprofils Goldisthal (Thüringisches Schiefergebirge). Zeitschrift derDeutschen Geolgischen Gesellschaft 151, 365–414.

Förster, B., 1996. U/Pb Datierung an Pechblenden der U-Lagerstätte Aue-Niederschlema(Erzgebirge). PhD Thesis, Univ Giessen, Germany, 212 pp + appendices.

Förster, H.-J., 1999. The chemical composition of uraninite in Variscan granites of theErzgebirge, Germany. Mineralogical Magazine 63, 239–252.

Förster, H.J., Rhede, D., 2006. The Ba–Ta-rich granite of Seiffen (eastern Erzgebirge,Germany): accessory-mineral chemistry, composition, and age of a late-VariscanLi–F granite of A-type affinity. Neues Jahrbuch für Mineralogie, Abhandlungen 182,307–321.

Förster, H.-J., Romer, R.L., 2010. Carboniferous magmatism. In: Linnemann, U., Romer,R.L. (Eds.), The pre-Mesozoic Geology of Saxo-Thuringia — From the CadomianActive Margin to the Variscan Orogen. Schweizerbart Science Publishers, Stuttgart,pp. 287–308.

Förster, H.-J., Tischendorf, G., Trumbull, R.B., Gottesmann, B., 1999. Late-collisional gran-ites in the Variscan Erzgebirge (Germany). Journal of Petrology 40, 1613–1645.

Förster, H.-J., Gottesmann, B., Tischendorf, G., Siebel, W., Rhede, D., Seltmann, R.,Wasternack, J., 2007. Permocarboniferous subvolcanic rhyolitic dikes in thewestern Erzgebirge/Vogtland, Germany: a record of source heterogeneity ofpost-collisional felsic magmatism. Neues Jahrbuch für Mineralogie, Abhandlungen183, 123–147.

Franke, W., 1989. Tectonostratigraphic units in the Variscan belt of central Europe. In:Dallmeyer, R.D. (Ed.), Terranes in the Circum-Atlantic Paleozoic Orogens: GeologicalSociety of America, Special Paper, 230, pp. 67–90.

Franke, W., 2000. The mid-European segment of the Variscides: tectonostratigraphicunits, terrane boundaries and plate tectonic evolution. In: Franke, W., Haak, V.,Oncken, O., Tanner, D. (Eds.), Orogenic Processes: Quantification and Modelling inthe Variscan Belt: Geological Society, London, Special Publications, 179, pp. 35–61.

German, K., Schwartz, T., Wipki, M., 1994. Mineral deposit formation in Phanerozoic sed-imentary basins of north-east Africa: the contribution of weathering. GeologischeRundschau 83, 787–798.

Gerstenberger, H., 1989. Autometasomatic Rb enrichments in highly evolved granitescausing lowered Rb/Sr isochron intercepts. Earth and Planetary Science Letters 93,65–75.

Gerstenberger, H., Haase, G., Habedank,M., 1983. Rb/Sr-Datierungender jüngerenGranitein Ehrenfriedersdorf. ZfI Mitteilungen: Berichte des Zentralinstituts für Isotopen- undStrahlenforschung 76, 125–133.

Gerstenberger, H., Kaemmel, T., Haase, G., Geisler, M., 1984. Zur Charakterisierung derGranite im Westerzgebirge: Rb/Sr-radiogeochronologische Untersuchungen undSpurenelementkonzentrationen. Freiberger Forschungshefte C389, 220–246.

Gerstenberger, H., Haase, G., Wemmer, K., 1995. Isotope systematics of the Variscanpostkinematic granites in the Erzgebirge (E Germany). Terra Nostra 95 (7), 36–41.

Giese, U., Katzung, G., Walter, R., 1994. Detrital composition of Ordovician sandstonesfrom the Rügen boreholes: implications for the evolution of the Tornquist Ocean.Geologische Rundschau 83, 293–308.


Hahn, T., Kroner, U., Melzer, P., 2010. Early Carboniferous synorogenic sedimentation inthe Saxo-Thuringian Basin and the adjacent Allochthonous Domain. In: Linnemann,U., Romer, R.L. (Eds.), The pre-Mesozoic Geology of Saxo-Thuringia — From theCadomian Active Margin to the Variscan Orogen. Schweizerbart Science Publishers,Stuttgart, pp. 171–192.

Hahne, K., Romer, R.L., 2008. Temporal geochemical and isotopic variation in Palaeo-zoic sediments of the Schwarzburg anticline (Saxo-Thuringia). In: Königshof, P.,Linnemann, U. (Eds.), From Gondwana to Laurussia to Pengea: Dynamics of Oceansand Supercontinents. Final Meeting of IGCP 497 and IGCP 499. Abstracts and Pro-gramme Frankfurt/M. and Dresden, pp. 176–178.

Hahne, K., Maaß, I., Lützner, H., Ellenberg, J., Falk, F., Grumbt, E., 1984. C-Isotopenuntersu-chungen zur EntwicklungdesAltpaläozoikums imThüringer Schiefergebirge. ZfIMittei-lungen: Berichte des Zentralinstituts für Isotopen- und Strahlenforschung 85, 116–132.

Heuse, T., Blumenstengel, H., Elicki, O., Geyer, G., Hansch, W., Maletz, G.N., Sarmiento,G.N., Weyer, D., 2010. Biostratigraphy — the faunal province of the southern marginof the Rheic Ocean. In: Linnemann, U., Romer, R.L. (Eds.), The Pre-Mesozoic Geologyof Saxo-Thuringia — From the Cadomian Active Margin to the Variscan Orogen.Schweizerbart Science Publishers, Stuttgart, pp. 99–170.

Höhndorf, A., Kämpf, H., Dulski, P., 1994. Sm/Nd and Rb/Sr isotopic investigationson fluorite mineralization of the eastern Erzgebirge. In: Seltmann, R., Kämpf, H.,Möller, P. (Eds.), Metallogeny of Collisional Orogens. From Gondwana to Laurussiato Pengea: Dynamics of Oceans and Supercontinents. Final Meeting of IGCP 497and IGCP 499. Abstracts and Programme Frankfurt/M. and Dresden, pp. 176–178.

Janoušek, V., Finger, F., Roberts, M., Frýda, J., Pin, C., Dolejš, D., 2004. Decipheringthe petrogenesis of deeply buried granites: whole-rock geochemical constraintson the origin of largely undepleted felsic granulites from the MoldanubianZone of the Bohemian Massif. Transactions of the Royal Society of Edinburgh:Earth Sciences 95, 141–159.

Johannes, W., Holtz, F., 1996. Petrogenesis and Experimental Petrology of GraniticRocks. Springer Verlag. 335 pp.

Klemd, R., 2010. Early Variscan allochthonous domains: the Münchberg Complex,Frankenberg, Wildenfels, and Góry Sowie. In: Linnemann, U., Romer, R.L. (Eds.),The Pre-Mesozoic Geology of Saxo-Thuringia — From the Cadomian Active Marginto the Variscan Orogen. Schweizerbart Science Publishers, Stuttgart, pp. 221–232.

Krentz, O., 1985. Rb/Sr-Altersdatierungen an Parametamorphiten des westlichenErzgebirgsantiklinoriums, DDR. Zeitschrift für Geologische Wissenschaften 13,443–462.

Kreuzer, H., Seidel, E., Schüssler, U., Okrusch, M., Lenz, K.-L., Raschka, H., 1989. K–Ar geo-chronology of different tectonic units at the northwestern margin of the Bohemianmassif. Tectonophysics 157, 149–178.

Kroner, U., Romer, R.L., 2010. The Saxo-Thuringian Zone — Tipp of the Armorican spurand part of the Gondwana plate. In: Linnemann, U., Romer, R.L. (Eds.), The Pre-Mesozoic Geology of Saxo-Thuringia — From the Cadomian Active Margin to theVariscan Orogen. Schweizerbart Science Publishers, Stuttgart, pp. 371–394.

Kröner, A., Willner, A.P., 1998. Time of formation and peak of Variscan HP–HAT meta-morphism of quartz-feldspathic rocks in the Central Erzgebirge, Saxony. Contribu-tions to Mineralogy and Petrology 132, 1–20.

Kroner, U., Hahn, T., Romer, R.L., Linnemann, U., 2007. The Variscan orogeny in the Saxo-Thuringian zone — heterogenous overprint of Cadomian/Palaeozoic Peri-GondwanaCrust. In: Linnemann, U., Nance, D., Kraft, P., Zulauf, G. (Eds.), The Evolution ofthe Rheic Ocean: From Avalonian-Cadomian Active Margin to Alleghenian-VariscanCollision: Geological Society of America, Special Paper, 423, pp. 153–172.

Kroner, U.,Mansy, J.L., Mazur, S., Aleksandrowski, P., Hann,H.P., Huckriede, H., Lacquement,F., Lamarche, J., Ledru, P., Pharaoh, T.C., Zedler, H., Zeh, A., Zulauf, G., 2008. Variscantectonics. In: McCann, T. (Ed.), The Geology of Central Europe: Geological SocietyLondon, pp. 599–664.

Kroner, U., Romer, R.L., Linnemann, U., 2010. The Saxo-Thuringian Zone of the VariscanOrogen as part of Pangea. In: Linnemann, U., Romer, R.L. (Eds.), The Pre-MesozoicGeology of Saxo-Thuringia — From the Cadomian Active Margin to the VariscanOrogen. Schweizerbart Science Publishers, Stuttgart, pp. 3–16.

Kröner, A., Willner, A.P., Hegner, E., Frischbutter, A., Hofmann, J., Bergner, R., 1995. LatestPrecambrian (Cadomian) zircon ages, Nd isotopic systematics and P–T evolution ofgranitoid orthogneisses of the Erzgebirge, Saxony and Czech Republic. GeologischeRundschau 84, 437–456.

Kump, L.R., Brantley, S.L., Arthur, M.A., 2000. Chemical weathering, atmospheric CO2,and climate. Annual Review of Earth and Planetary Science 28, 611–667.

Linnemann, U., Romer, R.L., 2002. The Cadomian orogeny in Saxo-Thuringia, Germany:geochemical and Nd–Sr–Pb isotopic characterization of marginal basins with con-straints to geotectonic setting and provenance. Tectonophysics 352, 33–64.

Linnemann, U., Schauer, M., 1999. Die Entstehung der Elbezone vor dem Hintergrundder cadomischen und variszischen Geschichte des Saxothuringischen Terranes —

Konsequenzen aus einer abgedeckten geologischen Karte. Zeitschrift für Geolo-gische Wissenschaften 27, 529–561.

Linnemann, U., Gehmlich, M., Tichomirowa, M., Buschmann, B., Nasdala, L., Jonas, P.,Lützner, H., Bombach, K., 2000. From Cadomian subduction to Early Palaeozoic rifting:the evolution of Saxo-Thuringia at the margin of Gondwana in the light of single zir-con geochronology and basin development (Central European Variscides, Germany).In: Franke, W., Haak, V., Oncken, O., Tanner, D. (Eds.), Orogenic Processes— Quantifi-cation and Modelling in the Variscan Belt of Central Europe: GeologicalSociety, London, Special Publication, 179, pp. 131–153.

Linnemann, U., McNaughton, N.J., Romer, R.L., Gehmlich, M., Drost, K., Tonk, C., 2004.West African Provenance for Saxo-Thuringia (Bohemian Massif): did Armoricaever leave pre-Pangean Gondwana? — U/Pb-SHRIMP zircon evidence and theNd-isotopic record. International Journal of Earth Sciences 93, 683–705.

Linnemann, U., Gerdes, A., Drost, K., Buschmann, B., 2007. The continuum betweenCadomian Orogenesis and opening of the Rheic Ocean: constraints from LA-ICP-MS

U-Pb zircon dating and analysis of plate-tectonic setting (Saxo-Thuringian Zone, NEBohemian massif, Germany). In: Linnemann, U., Nance, D., Kraft, P., Zulauf, G.(Eds.), The Evolution of the Rheic Ocean: From Avalonian-Cadomian active mar-gin to Alleghenian-Variscan collision: Geological Society of America, SpecialPaper, 423, pp. 61–96.

Linnemann, U., Hofmann, M., Romer, R.L., Gerdes, A., 2010. Transitional stages betweenthe Cadomian and Variscan orogenies: basin development and tectono-magmaticevolution of the southern margin of the Rheic Ocean in the Saxo-Thuringian Zone(North Gondwana shelf). In: Linnemann, U., Romer, R.L. (Eds.), The Pre-MesozoicGeology of Saxo-Thuringia — From the Cadomian Active Margin to the VariscanOrogen. Schweizerbart Science Publishers, Stuttgart, pp. 59–98.

Linnen, R.L., 1998. The solubility of Nb–Ta–Zr–Hf–W in granitic melts with Li and LI+F:constraints for mineralization in rare-metal granites and pegmatites. Economic Geology93, 1013–1025.

Linnen, R.L., 2006. The effect of water on accessory phase solubility in subaluminousand peralkaline granitic melts. Lithos 80, 267–280.

Lützner, H., Ellenberg, J., Falk, F., 1986. Entwicklung der Sedimentationsrate und derAblagerungsprozesse im Altpaläozoikum Thüringens. Zeitschrift für GeologischeWissenschaften 14, 83–93.

Lützner, H., Linnemann, U., Meisel, S., Hahne, K., 2001. Sedimentation und Fazies imgefalteten Palaeozoikum des Thüringischen Schiefergebirges. Schriftenreihe derDeutschen Geologischen Gesellschaft 13, 113–136.

Maaß, I., Hahne, K., Loos, G., Vogler, P., 1986. Isotopen- und elementgeochemischeUntersuchungen an schwach metamorphen Sedimenten als Hilfsmittel zur faziellenund stratigraphischen Einordnung. Isotopenpraxis 22, 131–135.

Matte, P., 2001. The Variscan collage and orogeny (480–290 Ma) and the tectonicdefinition of the Armorica microplate: a review. Terra Nova 13, 122–128.

McLennan, s.M., 1989. Rare earth elements in sedimentary rocks: influence of prove-nance and sedimentary processes. In: Lipin, b.r., McCay, G.A. (Eds.), Geochemistryand Mineralogy of Rare Earth Elements: Reviews in Mineralogy, 21, pp. 169–200.

Menning, M., German Stratigraphic Commission, 2002. A geologic time scale 2002. In:German Stratigraphic Commission (Ed.), Stratigraphic Tables of Germany 2002.

Mingram, B., 1996. Geochemische Signaturen der Metasedimente des erzgebirgischenKrustenstapels. Scientific Technical Report, STR96/04, GeoForschungsZentrumPotsdam. 104 pp.

Mingram, B., 1998. The Erzgebirge, Germany — a subducted part of northern Gondwana:geochemical evidence for repetition of early Paleozoicmetasedimentary sequences inmetamorphic thrust units. Geological Magazine 135, 785–801.

Mingram, B., Rötzler, K., 1999. Geochemische, petrologische und geochronologischeUntersuchungen im Erzgebirgskristallin — Rekonstruktion eines Krustenstapels.Schriftenreihe für Geowissenschaften 9, 1–72.

Mingram, B., Trumbull, R.B., Littmann, S., Gerstenberger, H., 2000. A petrogenetic studyof anorogenic felsic magmatism in the Cretaceous Paresis ring complex, Namibia;evidence for mixing of crust and mantle-derived components. Lithos 54, 1–22.

Mingram, B., Kröner, A., Hegner, E., Krentz, O., 2004. Zircon ages, geochemistry, andNd isoto-pic systematics of pre-Variscan orthogneisses from the Erzgebirge, Saxony (Germany),and geodynamic interpretation. International Journal of Earth Sciences 93, 706–727.

Minster, J.-F., Birck, J.-L., Allègre, C.J., 1982. Absolute age of formation of chondritesstudied by the 87Rb–87Sr method. Nature 300, 414–419.

Montel, J.M., Vielzeuf, D., 1997. Partialmelting ofmetagreywackes, Part II. Compositions ofminerals and melts. Contributions to Mineralogy and Petrology 128, 176–196.

Nägler, T.F., Schäfer, H.-J., Gebauer, D., 1995. Evolution of the Western European conti-nental crust: implications from Nd and Pb isotopes in Iberian sediments. ChemicalGeology 121, 345–357.

Nance, R.D., Murphy, J.B., 1994. Contrasting basement isotopic signatures and thepalinspastic restoration of peripheral orogeny: example from the NeoproterozoicAvalonian-Cadomian belt. Geology 22, 617–620.

Nance, R.D., Gutiérrez-Alonso, G., Keppie, J.D., Linnemann, U., Murphy, J.B., Quesada, C.,Strachan, R.A., Woodcock, N.H., 2010. Evolution of the Rheic Ocean. GondwanaResearch 17, 194–222.

Nebel, O., Scherer, E.E., Mezger, K., 2011. Evaluation of the 87Rb decay constant by agecomparison against the U–Pb system. Earth and Planetary Science Letters 301, 1–8.

Nesbitt, H.W., Markovics, G., Price, R.C., 1982. Chemical processes affecting alkalis andalkaline earths during continental weathering. Geochimica et Cosmochimica Acta44, 1659–1666.

Nesbitt, H.W., Young, G.M., McLennan, S.M., Keays, R.R., 1996. Effects of chemicalweathering and sorting on the petrogenesis of siliciclastic sediments, with impli-cations for provenance studies. Journal of Geology 104, 525–542.

Noblet, C., Lefort, J.P., 1990. Sedimentological evidence for a limited separation betweenAmorica and Gondwana during the Early Ordovician. Geology 18, 303–306.

Roddick, J.C., Compston,W., 1977. Strontium isotopic equilibration: a solution to a paradox.Earth and Planetary Science Letters 34, 238–246.

Romer, R.L., Hahne, K., 2010. Life of the Rheic Ocean: scrolling through the shale record.Gondwana Research 17, 236–253.

Romer, R.L., Rötzler, J., 2001. P–T–t evolution of ultrahigh-temperature granulites fromthe Saxon Granulite Massif, Germany. Part II: Geochronology. Journal of Petrology42, 2015–2032.

Romer, R.L., Rötzler, J., 2003. Effect of metamorphic reaction history on the U–Pb datingof titanite. In: Vance, D., Müller, A., Villa, I. (Eds.), Geochronology: Linking the IsotopicRecord with Petrology and Textures: Geological Society, London, Special Publication,220, pp. 147–158.

Romer, R.L., Förster, H.-J., Breitkreuz, Chr, 2001. Intracontinental extensional magma-tism with a subduction fingerprint: the late Carboniferous Halle Volcanic Complex(Germany). Contributions to Mineralogy and Petrology 141, 201–221.

Romer, R.L., Heinrich, W., Schröder-Smeibidl, B., Meixner, A., Fischer, C.-O., Schulz, C.,2005. Elemental dispersion and stable isotope fractionation during reactive fluid-


flow and fluid immiscibility in the Bufa del Diente aureole, NE-Mexico: evidencefrom radiographies and Li, B, Sr, Nd, and Pb isotope systematics. Contributions toMineralogy and Petrology 149, 400–429.

Romer, R.L., Thomas, R., Stein, H.J., Rhede, D., 2007. Dating multiply overprinted Sn-mineralized granites — examples from the Erzgebirge, Germany. MineraliumDeposita 42, 337–359.

Romer, R.L., Hahne, K., Plessen, B., Förster, H.-J., 2008. Recycling of Palaeozoic crust duringthe Variscan orogeny — the strontium story. In: Königshof, P., Linnemann, U. (Eds.),From Gondwana to Laurussia to Pangea: Dynamics of Oceans and Supercontinents.Final Meeting of IGCP 497 and IGCP 499. Abstracts and Programme Frankfurt/M.and Dresden, pp. 223–225.

Romer, R.L., Schneider, J., Linnemann, U., 2010a. Post-Variscan deformation and hydro-thermal mineralization in Saxo-Thuringia and beyond: a geochronologic review. In:Linnemann, U., Romer, R.L. (Eds.), The Pre-Mesozoic Geology of Saxo-Thuringia —

From the Cadomian Active Margin to the Variscan Orogen. Schweizerbart SciencePublishers, Stuttgart, pp. 347–360.

Romer, R.L., Förster, H.-J., Štemprok, M., 2010b. Age constraints for the late-Variscanmagmatism in the Altenberg–Teplice Caldera (Eastern Erzgebirge/Krušné hory).Neues Jahrbuch für Mineralogie 187, 289–305.

Romer, R.L., Kirsch, M., Kroner, U., 2011. Geochemical signature of Ordovician Mn-richsedimentary rocks on the Avalonian shelf. Canadian Journal of Earth Sciences 48,703–718.

Ronov, a.b., Migdisov, a.a., 1996. Time variations in the abundances of rocks, minerals,and elements in the Russian platform sediment cover. Geochemistry International33 (5), 30–63.

Roscher, M., Schneider, J.W., 2006. Permocarboniferous climate: early Pennsylvanian toLate Permian climate development of central Europe in a regional and global context.In: Lucas, S.G., Cassinis, G., Schneider, J.W. (Eds.), Non-marine Permian Biostratigraphyand Biochronology: Geological Society, London, Special Publication, 265, pp. 95–136.

Rötzler, K., 1995. Die Entwicklung der Metamorphite des Mittel- und Westerzgebirges.Technical Report, STR 95/14, GeoForschungsZentrum Potsdam. 222 pp.

Rötzler, K., Plessen, B., 2010. The Erzgebirge: a pile of ultrahigh- to low-pressurenappes of Early Palaeozoic rocks and their Cadomian basement. In: Linnemann,U., Romer, R.L. (Eds.), The Pre-Mesozoic Geology of Saxo-Thuringia — From theCadomian Active Margin to the Variscan Orogen. Schweizerbart Science Publishers,Stuttgart, pp. 253–270.

Rötzler, K., Schumacher, R., Maresch, W.V., Willner, A.P., 1998. Characterization andgeodynamic implications of contrasting metamorphic evolution in juxtaposedhigh-pressure units of the Western Erzgebeirge (Saxony, Germany). EuropeanJournal of Mineralogy 10, 261–280.

Rötzler, J., Carswell, D.A., Gerstenberger, H., Haase, G., 1999. Transitional blueschist-epidote amphibolite facies metamorphism in the Frankenberg massif, Germany, andgeotectonic implications. Journal of Metamorphic Geology 17, 109–125.

Rötzler, J., Romer, R.L., Budzinski, H., Oberhänsli, R., 2004. Ultrahigh-temperature granu-lites from Tirschheim, Saxon Granulite Massif, Germany: P–T–t path and geotectonicimplications. European Journal of Mineralogy 16, 917–937.

Rudnick, R.L., Gao, S., 2003. Composition of the Continental Crust. Treatise on Geo-chemistry, Vol. 3. Elsevier, Amsterdam, pp. 1–64.

Schneider, J.W., Romer, R.L., 2010. The Late Variscan Molasses (Late Carboniferous toLate Permian) of the Saxo-Thuringian Zone. In: Linnemann, U., Romer, R.L. (Eds.),The Pre-Mesozoic Geology of Saxo-Thuringia — From the Cadomian Active Marginto the Variscan Orogen. Schweizerbart Science Publishers, Stuttgart, pp. 323–346.

Sebastian, U., 1995. Die Strukturentwicklung des spätorogenen Erzgebirgsaufstiegs inder Flöhazone: Ein weiterer Beitrag zur postkollisionalen Extension am Nordrandder Böhmischen Masse. Freiberger Forschungshefte. C461, 114 pp.

Siebel, W., Shang, C.K., Reitter, E., Rohrmüller, J., Breiter, K., 2008. Two distinctive granitesuites in the SW Bohemian Massif and their record of emplacement: constraints fromgeochemistry and zircon 207Pb/206Pb chronology. Journal of Petrology 49, 1853–1872.

Steiger, R.H., Jäger, E., 1977. Subcommission on geochronology: convention on the use ofdecay constants in geo- and cosmochronology. Earth and Planetary Science Letters 36,359–362.

Stussi, J.-M., 1989. Granitoid chemistry and associated mineralization in the FrenchVariscan. Economic Geology 52, 175–192.

Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: Its Composition and Evolution.Blackwell Scientific Publications. 312 pp.

Thomas, R., Förster, H.-J., Heinrich, W., 2003. The behaviour of boron in a peraluminousgranite–pegmatite system and associated hydrothermal solutions: a melt and fluidinclusion study. Contributions to Mineralogy and Petrology 144, 457–472.

Tichomirowa, M., 2003. Die Gneise des Erzgebirges — hochmetamorphe Äquivalentevon neoproterozoischen-frühpaläozoischen Grauwacken und Granitoiden derCadomiden. Freiberger Forschungshefte. C495, 222 pp.

Tichomirowa,M., Leonhardt, D., 2010. Newage determinations (Pb/Pb zircon evaporation,Rb/Sr) on the granites fromAue–Schwarzenberg and Eibenstock,Western Erzgebirge,Germany. Zeitschrift für Geologische Wissenschaften 38, 99–123.

Tischendorf, G., Förster, H.-J., 1990. Acid magmatism and related metallogenesis in theErzgebirge. Geological Journal 25, 443–454.

Turpin, L., Cuney, M., Friedrich, M., Bouchez, J.-L., Aubertin, M., 1990. Meta-igneous originof Hercynian peraluminous granites in N.W. French Massif Central: implications forcrustal history reconstructions. Contributions to Mineralogy and Petrology 104,163–172.

Vásquez, P., Franz, G., Glodny, J., Romer, R.L., Gerdes, A., 2009. Origin of fayalite granitoids:new insights from the Cobquecura pluton, Chile, and its metapelitic xenoliths. Lithos110, 181–198.

Vielzeuf, D., Montel, J.M., 1994. Partial melting of metagreywackes. Part I. Fluid-absentexperiments and phase relationships. Contributions to Mineralogy and Petrology117, 375–393.

Viers, J., Dupré, B., Braun, J.-J., Deberdt, S., Angeletti, B., Ngoupayou, J.N., Michard, A.,2000. Major and trace element abundances, and strontium isotopes in the Nyongbasin rivers (Cameroon): constraints on chemical weathering processes and ele-ments transport mechanisms in humid tropical environments. Chemical Geology169, 211–241.

Villaseca, C., Perez-Soba, C., Merino, E., Orejana, D., Lopez-Garcia, J.A., Billström, K.,2008. Contrasting crustal sources for peraluminous granites of the segmentedMontes de Toledo Batholith (Iberian Variscan Belt). Journal of Geosciences 53,263–280.

von Quadt, A., Günther, D., 1999. Evolution of Cambrian eclogitic rocks in the Erzgebirge: aconventional and LA-ICP-MS U–Pb zircon and Sm–Nd study. Terra Nostra 99/1, 164.

von Seckendorff, V., Timmermann,M.J., Kramer,M.J., Wrobel, P., 2004. New 40Ar/39Ar agesand geochemistry of late Carboniferous-early Permian lamprophyres and relatedvolcanic rocks in the Saxothuringian Zone of the Variscan Orogeny (Germany). In:Wilson, M., Neumann, E.-R., Davies, G.R., Timmermann, M.J., Heeremans, M., Lasen,B.T. (Eds.), Permo-Carboniferous Magmatism and Rifting in Europe: GeologicalSociety of London, Special Publication, 223, pp. 335–360.

Waldron, J.W.F., White, C.E., Barr, S.M., Simonetti, A., Heaman, L.M., 2009. Provenanceof the Meguma terrane, Nova Scotia: rifted margin of early Paleozoic Gondwana.Canadian Journal of Earth Sciences 46, 1–8.

Webster, J.D., Thomas, R., Rhede, D., Förster, H.-J., Seltmann, R., 1997. Melt inclusions inquartz from an evolved peraluminous pegmatite: geochemical evidence for strongtin enrichment in fluorine-rich and phosphorus-rich residual liquids. Geochimicaet Cosmochimica Acta 61, 2589–2604.

Wendt, I., Haase, G., 1998. Dynamic double collector measurement with cup efficiencyfactor determination. Chemical Geology 146, 99–110.

Werner, O., Lippolt, H.J., 2000. White mica 40Ar/39Ar ages of Erzgebirge metamorphicrocks: simulating the chronological results by amodel of Variscan crustal imbrication.In: Franke,W., Haak, V., Oncken, O., Tanner, D. (Eds.), Orogenic Processes: Quantifica-tion andModelling in the Variscan Belt of Central Europe: Geological Society, London,Special Publication, 179, pp. 323–336.

White, A.F., Blum, A.E., Bullen, T.D., Vivit, D.V., Schulz, M., Fitzpatrick, J., 1999. The effectof temperature on experimental and natural chemical weathering rates of granit-oid rocks. Geochimica et Cosmochimica Acta 63, 3277–3291.

Willner, A.P., Rötzler, K., Maresch, V.W., 1997. Pressure-temperature andfluid evolution ofquartzo-feldspathic metamorphic rocks with a relic high-pressure, granulite-facieshistory from the Central Erzgebirge (Saxony, Germany). Journal of Petrology 38,307–336.

Winkler, H.F.G., 1967. Petrogenesis of Metamorphic Rocks, 2nd ed. Springer, New York.Zeng, L., Asimov, P.D., Saleeby, J.B., 2005. Coupling of anatectic reactions and dissolution of

accessory phases and the Sr and Nd isotope systematics of anatectic melts from ametasedimentary source. Geochimica et Cosmochimica Acta 69, 3671–3682.

Documents

Strontium isotopes — A persistent tracer for the recycling of Gondwana crust in the Variscan orogen