Journal of Petrology

Journal of Petrology Advance Access originally published online on February 25, 2005
Journal of Petrology 2005 46(7):1393-1420; doi:10.1093/petrology/egi020

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 46/7/1393    most recent egi020v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (3) Request Permissions Google Scholar Articles by ZEH, A. Articles by MILLAR, I. L. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

© The Author 2005. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oupjournals.org

Provenance and Magmatic–Metamorphic Evolution of a Variscan Island-Arc Complex: Constraints from U–Pb Dating, Petrology, and Geospeedometry of the Kyffhäuser Crystalline Complex, Central Germany

ARMIN ZEH^1,*, AXEL GERDES², THOMAS M. WILL¹ and IAN L. MILLAR³

¹ MINERALOGISCHES INSTITUT, AM HUBLAND, D-97074 WÜRZBURG, GERMANY
² INSTITUT FÜR MINERALOGIE, SENCKENBERG ANLAGE 28, D-60054 FRANKFURT AM MAIN, GERMANY
³ BRITISH ANTARCTIC SURVEY, c/o NERC ISOTOPE GEOSCIENCES LABORATORY, KINGSLEY DUNHAM CENTRE, KEYWORTH, NOTTINGHAM NG12 5GG, UK

RECEIVED JUNE 10, 2004; ACCEPTED JANUARY 28, 2005

	ABSTRACT

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
The Kyffhäuser Crystalline Complex, Central Germany, forms part of the Mid-German Crystalline Rise, which is assumed to represent the Variscan collision zone between the East Avalonian terrane and the Armorican terrane assemblage. High-precision U–Pb zircon and monazite dating indicates that sedimentary rocks of the Kyffhäuser Crystalline Complex are younger than c. 470 Ma and were intruded by gabbros and diorites between 345 ± 4 and 340 ± 1 Ma. These intrusions had magmatic temperatures between 850 and 900°C, and caused a contact metamorphic overprint of the sediments at P–T conditions of 690–750°C and 5–7 kbar, corresponding to an intrusion depth of 19–25 km. At 337 ± 1 Ma the magmatic–metamorphic suite was intruded by granites, syenites and diorites at a shallow crustal level of some 7–11 km. This is inferred from a diorite, and conforms to P–T paths obtained from the metasediments, indicating a nearly isothermal decompression from 5–7 to 2–4 kbar at 690–750°C. Subsequently, the metamorphic–magmatic sequence underwent accelerated cooling to below 400°C, as constrained by garnet geospeedometry and a previously published K–Ar muscovite age of 333 ± 7 Ma. With respect to P–T–D–t data from surrounding units, rapid exhumation of the KCC can be interpreted to result from NW-directed crustal shortening during the Viséan.

KEY WORDS: contact metamorphism; U–Pb dating; hornblende; garnet; Mid-German Crystalline Rise; P–T pseudosection

	INTRODUCTION

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
Reconstruction of geodynamic processes related to the formation and destruction of present and ancient island-arc complexes is a general problem in geology (e.g. Geist et al., 1994; Oncken, 1997; Oh et al., 2004). Such reconstructions require a comprehensive knowledge of the timing and duration of magmatic intrusions and metamorphic events, and the pressure–temperature–deformation–time (P–T–D–t) evolution of the magmatic and metamorphic rocks. Furthermore, information about the age and provenance of related metasediments is of interest, because they can give important insights into the pre-island-arc environment, and may allow correlations with surrounding rock units (Geist et al., 1994). Investigation of contemporaneous and/or tectonically transposed magmatic and metamorphic rocks can serve as an ultimate test for the suitability of published geothermobarometers, geospeedometers, and geochronometers.

In this paper we present new geochronological and petrological results from magmatic and metasedimentary rocks from the Kyffhäuser Crystalline Complex (KCC), which forms the northeasternmost outcrop of the Mid-German Crystalline Rise (MGCR). The latter extends for over 350 km from SW to NE and is also exposed in the northern Vosges, the Odenwald Mountains, the Spessart Mountains, and in the Ruhla Crystalline Complex (Fig. 1). The MGCR is a zone of medium- to high-grade metamorphic gneisses and granitoids, which separates the very low- to low-grade metasediments and volcanic rocks of the Northern Phyllite Belt and the Rhenohercynian domain in the NW from low- to medium-grade rocks of the Saxothuringian domain in the SE (Fig. 1). In a wider plate tectonic context, the MGCR is interpreted to be part of a suture zone between the Eastern Avalonia Terrane, which forms the basement underlying Belgium, Southern England and Wales, and the Saxothuringian microterrane that forms part of the so-called Armorican Terrane Assemblage (Tait et al., 1997).

View larger version (58K): [in this window] [in a new window]   Fig. 1. Location of the Variscan Kyffhäuser Crystalline Complex within Europe and a simplified geological map of the study area with sample localities. NPB, Northern Phyllite Belt.

 
The MGCR is not a coherent belt, but consists of at least two contrasting rock units (I and II; Fig. 1), which are assumed to have formed in different geotectonic environments and were juxtaposed during Variscan collision (Zeh, 1996; Oncken, 1997; Hansch & Zeh, 2000). Rocks of unit I are interpreted to represent relics of a Variscan magmatic arc, which formed at the northwestern margin of the Saxothuringian domain (upper plate), whereas rocks of unit II comprise a metasediment–metabasite succession, which is assumed to have been deposited in the Rhenohercynian ocean NE of the Saxothuringian terranes (Oncken, 1997). The Rhenohercynian ocean closed during oblique Variscan convergence between 360 and 320 Ma, and rocks of unit II were accreted to the base of the overriding Saxothuringian island arc (Oncken, 1997; Altherr et al., 1999). The central part of the magmatic arc was eroded as a result of accretion during oblique collision, leading to the recently suggested threefold division of the MGCR. The central part of the MGCR, which comprises the Spessart, Ruhla and eastern part of the Odenwald crystalline complexes (Böllstein Odenwald), now exposes underplated rocks of unit II (Fig. 1), whereas the southwestern and northeastern parts, exposed in the western Odenwald and the Kyffhäuser Mountains, respectively, represent parts of the former Saxothuringian magmatic arc (Oncken, 1997).

The geotectonic model of Oncken (1997) is broadly based on structural, geophysical, petrological and geochronological data from the southwestern section of the MGCR, comprising the Odenwald, Spessart and Ruhla Crystalline complexes. At present, little is known about the timing of granitoid intrusions and metamorphic–magmatic processes of the northeastern continuation of the MGCR, including the KCC. Therefore, all geotectonic models that include this part of the Variscides are highly speculative.

Thus, to support or disprove present models related to the formation and destruction of the MGCR we present new geochronological data from the KCC and combine them with petrological results and geospeedometry studies. These data will allow us to set tight constraints on the timing of metasediment deposition, magmatic intrusions and metamorphic overprint, and to reconstruct geodynamic processes in this part of the Variscan orogenic belt. For this purpose we present high-precision U–Pb monazite and zircon data and comprehensive petrographic descriptions, and test several conventional geothermobarometers on dioritic, basaltic and metapelitic rocks. Furthermore, P–T pseudosections in the model system CaO–Na₂O–K₂O–TiO₂–MnO–FeO–MgO–Al₂O₃–SiO₂–H₂O are employed to derive P–T paths for individual metapelitic rocks. Finally, geotectonic implications of our detailed P–T–t results are discussed in a wider geotectonic context.

	GEOLOGICAL SETTING

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
The KCC forms a small basement outcrop of 1·2 km² at the northern margin of the Kyffhäuser mountains, confined by the Kyffhäuser and Kelbra faults in the NE and NW (Fig. 1). Along these faults, the KCC is uplifted by nearly 1000 m against Upper Permian to Cenozoic sediments to the north. In the south the KCC is overlain by Stefanian molasse sediments, which were deposited at about 300 Ma (Menning et al., 1996). The KCC can be subdivided into (Fig. 1): (1) a magmatic–metamorphic complex in the west; (2) the Borntal Intrusive Complex in the centre; (3) the Bärenkopf granite in the east. The magmatic–metamorphic complex can be further subdivided from north to south into a hornblende gabbro complex, a biotite–plagioclase gneiss unit and a diorite–gneiss complex.

Hornblende gabbro complex
This unit is predominantly made up of hornblende gabbro intercalated with granodiorite–gneiss. The gabbro contains the mineral assemblage hornblende + plagioclase. Generally, three types of hornblende gabbro can be distinguished. Gabbro type HGn (normal type) contains nearly round plagioclase crystals up to 4 mm in size, which are intergrown with and overgrown by hornblende crystals and aggregates up to 10 mm in size. Gabbro type HGf is fine grained (grain size <2 mm) and contains columnar to isometric hornblende crystals surrounded by, and intergrown with plagioclase. In contrast, gabbro type HGc is very coarse grained (pegmatitic) and contains hornblende crystals up to 15 cm long, which overgrow plagioclase crystals and aggregates. All three types grade continuously into each other. The gabbro is truncated by metre- to decametre-scale dykes, bodies and lenses of granodiorite–gneiss. Locally, the granodiorite–gneiss is cut by thin leucogranite dykes (Zeh & Katzung, 1994). The hornblende gabbro is usually massive. Mylonitic shear zones occur close to the contacts with the granodiorite–gneiss and the biotite–plagioclase gneiss unit. The shear zones commonly dip steeply north or south. In contrast to the surrounding gabbro, the granodiorite–gneiss contains a pronounced foliation defined by aligned biotite and quartz ribbons.

Biotite–plagioclase gneiss unit
In the south, the hornblende gabbro complex is bordered by a biotite–plagioclase gneiss unit. The predominant biotite–plagioclase gneiss is intercalated with lenses and layers of marble, calcsilicate rocks, amphibolite, hornblende gneiss, biotite–hornblende gneiss and minor metapelite. All rocks are well foliated, and locally show migmatitic structures. The foliation dips steeply towards the north or south, parallel to the orientation of the shear zones observed in the gabbro complex. Schüller (1957) suggested that the biotite–plagioclase gneiss unit represents a Cadomian granite, which was deformed during the Variscan orogeny. However, because of the occurrence of marbles and calcsilicate rocks, Seim (1960, 1967) argued that the biotite–plagioclase gneiss unit represents a succession of metagreywackes and metapelites, intercalated with limestones and marlstones. The associated amphibolites and hornblende gneisses are assumed to represent basaltic dykes, which were strongly deformed together with the surrounding sediments (Seim, 1967).

Diorite–gneiss complex
In the south, the biotite–plagioclase gneiss unit grades into a diorite–gneiss unit, which predominantly consists of dioritic gneisses with the mineral assemblage hornblende + biotite + plagioclase. Locally, these gneisses contain abundant K-feldspar and/or retrograde epidote and chlorite. Commonly, rocks of the diorite–gneiss complex show a foliation, parallel to that observed in the biotite–plagioclase gneiss unit and the gabbro complex. Locally, massive diorite bodies with relic pseudo-ophitic textures occur (Neumann, 1968). The diorite–gneiss complex is cut by numerous unfoliated to weakly foliated, steep and flat lying leucogranite dykes (Fig. 1).

Borntal Intrusive Complex (BIC)
The rocks of the magmatic–metamorphic complex abut against the BIC to the east (Zeh & Katzung, 1994). This complex consists of porphyritic granite, grey granite, syenite and a schlieric, coarse-grained hornblende diorite. The porphyritic granite contains K-feldspar phenocrysts up to 3 cm in size, which are heterogeneously distributed in a groundmass of plagioclase, quartz and biotite. The porphyritic granite commonly contains metre-sized bodies of fine-grained diorite–gneiss, which can be related to the nearby diorite–gneiss complex. Contacts between the porphyritic granite and the diorite–gneiss bodies are sharp. The porphyritic granite grades more or less abruptly into the syenite, which itself is intruded by a grey, equigranular granite. The contact between the syenite and the grey granite is generally sharp. In the Borntal valley, decametre sized bodies of a schlieric, coarse-grained hornblende diorite occur in contact with porphyritic granite and syenite. In one outcrop the diorite is cross-cut by dykes of the grey granite. With the exception of the schlieric diorite, all rocks of the BIC show a pronounced foliation, which commonly dips steeply north or south, indicating that the magmatic rocks were deformed after their intrusion.

Bärenkopf granite
This granite forms the easternmost unit of the KCC. It shows a steeply NW dipping foliation, is commonly equigranular and medium grained, but locally contains K-feldspar porphyritic domains. In some places the Bärenkopf granite is cross-cut by east–west-trending muscovite-bearing pegmatite dykes. Muscovites from these dykes have yielded a K–Ar age of 333 ± 7 Ma (Neuroth, 1997).

	GEOCHRONOLOGY

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
With the exception of the 333 ± 7 Ma K–Ar muscovite cooling age of Neuroth (1997) nothing is known about the timing of metamorphism and magma intrusions in the KCC. In addition, the deposition age of the metasediment precursors is unknown. To fill this gap in our knowledge, new U–Pb ages are presented.

Samples
To constrain the timing of the magmatic intrusions, single zircon grains and small multi-grain fractions were analysed by U–Pb isotope dilution technique, in combination with thermal ionization mass spectrometry (TIMS) (Appendix A). Zircons were selected from a hornblende gabbro sample (HGn: hornblende gabbro complex), a diorite–gneiss sample (DG3: diorite–gneiss complex), from the Bärenkopf granite (sample BK), as well as from the schlieric diorite (sample BCD) and grey granite (sample BCG) from the BIC (Table 1, Fig. 1). Zircon grains of all magmatic rocks are euhedral, prismatic to needle-like, and invariably contain round to lenticular melt inclusions. For U–Pb dating, only clear, inclusion-free, euhedral zircon grains or fragments were selected.

View this table: [in this window] [in a new window]   Table 1: Co-ordinates of samples from the Kyffhäuser Crystalline Complex

 
In addition, three monazite fractions and two single zircon grains from a garnet–sillimanite–cordierite gneiss (sample ST) within the biotite–plagioclase gneiss unit were analysed, to constrain the timing of metamorphism and investigate the provenance of the metasediments. The analysed monazite grains are clear, yellow, and tabular with rounded edges. The zircon grains from the metasediment invariably have round cores overgrown by thin, euhedral, transparent rims of a later zircon generation, assumed to be of metamorphic origin. To obtain age information from the zircon cores, the overgrowths were abraded for about 12 h before the zircons were dissolved (Appendix A).

Geochronological results
All zircon grains of the magmatic rocks, and the monazite grains from the metasediment, yield Variscan ages (Fig. 2). Six zircon analyses from the diorite–gneiss sample DG3 define a discordia with an upper intercept at 345 ± 3·4 Ma and a lower intercept at 108 ± 86 Ma. Only one zircon fraction from the diorite–gneiss sample is concordant at 339 ± 1 Ma (Table 2). Two zircon fractions from the hornblende gabbro sample HGn yield a concordant age of 340·7 ± 1·1 Ma, which is identical to an age of 340·3 ± 1·4 Ma defined by two discordant and one concordant monazite analyses obtained from the metasediment sample ST (Fig. 2, Table 2). Four zircon fractions from the grey granite gneiss (sample BCG), and two zircon fraction from the schlieric diorite (sample BCD) of the BIC, gave identical concordant ages of 337·1 ± 1·0 Ma and 336·5 ± 0·5 Ma, respectively. The same age of 337·0 ± 2·7 Ma was obtained from three zircon fractions from the Bärenkopf granite (sample BK; Fig. 2f).

View larger version (50K): [in this window] [in a new window]   Fig. 2. (a–f) U–Pb concordia diagrams showing the results of zircon and monazite (c) dating (±2). (g) Results of U–Pb analyses of two abraded zircon cores from the metasediment sample ST. (0*) is the upper intercept age if the results are forced by a discordia through zero (recent Pb loss) and through 340 Ma (340*; indicates metamorphic zircon overgrowth and/or Pb loss as a result of contact metamorphism; for further explanation see text).

 

View this table: [in this window] [in a new window]   Table 2: Results of U–Pb dating

 
In contrast to the magmatic rocks, the cores of the two single zircon grains selected from the metasediment sample ST gave Pre-Variscan ages (Fig. 2g). One grain (ST-4) plots close to concordia at about 475 Ma, and the second grain (ST-5) is more discordant at >550 Ma. Assuming recent Pb loss, minimum upper intercept ages of 471 ± 5 Ma and 569 ± 10 Ma are estimated for grains ST-4 and ST-5, respectively. Assuming a Variscan zircon overgrowth at 340 Ma, upper intercept ages of 482 ± 17 Ma (ST-4) and 658 ± 24 Ma (ST-5) are obtained (Fig. 2g), setting an upper age limit for these grains.

Discussion of the age results
U–Pb analyses of zircons from magmatic rocks from the western part of the KCC gave ages between 345·1 ± 3·4 Ma and 340·7 ± 1·1 Ma, which are significantly older than ages obtained from the magmatic rocks of the BIC (337·1 ± 1·0 and 336·5 ± 0·5 Ma) and the Bärenkopf granite (337·0 ± 2·7 Ma). The age sequence is in agreement with the field relationships, indicating that the magmatic–metamorphic complex was intruded by magmatic rocks of the BIC and the Bärenkopf granite. From the age data, in combination with field relationships, the following intrusion sequence can be derived: (1) diorite–gneiss 345·1 ± 3·4 Ma; (2) hornblende gabbro 340·7 ± 1·1 Ma, (3) porphyritic granite (BIC), (4) syenite (BIC) and (5) schlieric diorite (BIC) all 336·5 ± 0·5 Ma; (6) grey granite (BIC) 337·1 ± 1·0 Ma; (7) Bärenkopf granite 337·0 ± 2·7 Ma. The age data clearly indicate that all magmatic rocks in the KCC intruded within a very short time span of between 3 and 10 Myr.

The U–Pb zircon ages presented here are identical to granite and diorite crystallization ages obtained from other parts of the Mid-German Crystalline Rise. They agree, within error, with Pb–Pb single zircon evaporation ages and U–Pb zircon ages of 337 ± 4 and 340 ± 1 Ma obtained from granites and diorites in the Ruhla Crystalline Complex, respectively (Zeh et al., 2000, 2003), and are within the range of K–Ar and ⁴⁰Ar/³⁹Ar hornblende ages between 345 and 330 Ma obtained from deformed diorites, quartz diorites, granodiorites, granites and amphibolites of the western part of the Odenwald Crystalline Complex (Kreuzer & Harre, 1975, recalculated; Hellmann et al., 1982; Schubert et al., 2001). Our zircon ages also show that the gabbros in the KCC were intruded c. 20 Myr later than the Frankenstein gabbro in the northwestern Odenwald Crystalline Complex (SZE: 362 ± 7 Ma, Kirsch et al., 1988).

The nearly concordant 470 Ma U–Pb zircon age (Fig. 2g) provides evidence that the metasedimentary rocks in the KCC, represented by the biotite–plagioclase gneiss unit, were deposited after the Early Ordovician. The 470 Ma old zircon cores could result from magmatic activities related to the break-up of the Avalonian–Cadomian orogenic belt, which formed the northern Gondwanan margin during the Neoproterozoic (e.g. Nance & Murphy, 1994). This is in agreement with sedimentological and geochronological results obtained by several other researchers (e.g. Linnemann et al., 2000; see discussion by Zeh et al., 2001, 2003). In contrast, zircon ages between 550 and 670 Ma (grain ST-5) can be related to magmatic activities during formation of the Avalonian–Cadomian orogenic belt (e.g. Nance & Murphy, 1994). In the MGCR, similar ages were obtained from detrital zircons from the Ruhla Crystalline Complex (Zeh et al., 2001, 2003).

The good agreement between the monazite age of the garnet–cordierite gneiss sample ST (340·3 ± 1·4 Ma), and the magmatic crystallization ages obtained from the hornblende gabbro (340·7 ± 1·1 Ma) and diorite–gneiss (345·1 ± 3·4 Ma) provides evidence that the sedimentary rocks in the KCC underwent a contact metamorphic overprint, caused by the gabbro and diorite intrusions.

	PETROGRAPHY AND MINERAL CHEMISTRY

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
To obtain information about the geodynamic processes during the Variscan magmatic–metamorphic evolution, metapelitic rocks from the biotite–plagioclase gneiss unit (samples ST and TT) and gabbroic or dioritic rocks from the hornblende gabbro complex (sample HGn and HGf), the diorite–gneiss complex (sample DG3 and DG4) and the BIC (sample BCD) were investigated in detail.

Biotite–plagioclase gneiss unit
Sample TT (garnet–biotite–cordierite gneiss)
This sample contains the minerals garnet, cordierite, biotite, chlorite, muscovite, plagioclase, quartz, ilmenite, rutile, zircon, monazite and sulphides. In thin section, two textural domains can be distinguished: a plagioclase-rich domain (TTp) and a biotite-rich domain (TTb). In both domains plagioclase is unzoned, with anorthite contents [X_an = Ca/(Ca + Na + K)] ranging from 0·28 to 0·31. Only a few plagioclase grains show increasing anorthite contents from 0·28 to 0·33 from core to rim, and some of them contain spatially separated garnet and biotite inclusions (Fig. 3a and b).

View larger version (114K): [in this window] [in a new window]   Fig. 3. Photomicrographs of the metasediment samples TT (a and b) and ST (c–f). (a) Garnet porphyroblasts (grt) are either occluded in plagioclase (pl) or occur along plagioclase grain boundaries (grt2) together with biotite (bt). (b) Garnet enclosed by pinitized cordierite (crd), and in contact with biotite, chlorite (chl) and a few muscovite (ms) grains. (c) Isolated garnet and biotite crystals enclosed in plagioclase, which is surrounded by biotite and a few retrograde chlorite and muscovite grains. (d) Pinitized cordierite is surrounded by plagioclase, quartz and fibrous sillimanite (sil). (e) Perthitic K-feldspar clast (ksp) is resorbed and surrounded by muscovite, biotite and cordierite. (f) Statically recrystallized K-feldspar crystals containing sillimanite needles are in contact with muscovite.

 
Garnet in both domains invariably forms euhedral porphyroblasts with diameters ranging from 0·01 to 0·65 mm. Garnet is observed in four distinct textural positions. Garnet 1 is completely enclosed in plagioclase (Fig. 3a); garnet 2 occurs along plagioclase grain boundaries, generally in contact with biotite (Fig. 3a); garnet 3 is completely surrounded by biotite; garnet 4 is surrounded by cordierite and/or retrograde chlorite (Fig. 3b). In all textural domains, garnet can contain rutile and ilmenite inclusions. Garnet 1 is nearly unzoned (Fig. 4), whereas all other garnets show an increase of the almandine [X_alm = Fe/(Fe + Mg + Ca + Mn)] and spessartine [X_sps = Mn/(Fe + Mg + Ca + Mn)] components, and in X_Fe [Fe/(Fe + Mg)] from core to the rim (Fig. 4). In contrast, the grossular component [X_grs = Ca/(Fe + Mg + Ca + Mn)] is almost constant or increases slightly from 0·03 to 0·04 from core to rim (e.g. Grt 1, Fig. 4). The X_Fe and X_sps values in the cores of garnet 2, 3 and 4 are higher than in garnet 1 and increase with decreasing garnet size (Fig. 4). Only the core composition of the largest garnet 2 is almost identical to that of garnet 1. This indicates that garnet 2, 3 and 4 experienced a diffusive Fe–Mg–Mn cation exchange during the retrograde evolution, which in most cases reset their core compositions. In contrast, garnet 1, which was shielded from diffusive exchange by the surrounding plagioclase (Fig. 3a), and the cores of the largest garnet grains of generation 2 (Fig. 4) still show the original growth zonation patterns, and thus can be used to obtain peak metamorphic conditions (see below).

View larger version (44K): [in this window] [in a new window]   Fig. 4. Zonation patterns of garnet and Xan of coexisting plagioclase observed in different domains of the metapelite samples TT and ST. ‘core’ and ‘rim’ indicate apparent closure temperatures obtained by garnet–biotite geothermometry (Kleemann & Reinhardt, 1994) of garnet cores and rims.

 
Cordierite rarely occurs in the biotite domain TTb. It is always pinitized, contains round garnet inclusions, and is locally replaced by chlorite and rare muscovite (Fig. 3b). Biotite occurs as inclusions in plagioclase (biotite 1), along plagioclase grain boundaries (biotite 2), and defines the foliation in the biotite-rich domains (biotite 3). Locally, biotite 2 and 3 are intergrown with chlorite. Biotites show the following compositional variations: biotite 1: X_Fe = 0·54–0·55, Al^VI = 0·28–0·34 p.f.u. (per formula unit), Ti = 0·09–0·15 p.f.u.; biotite 2: X_Fe = 0·54–0·56, Al^VI = 0·23–0·29 p.f.u., Ti= 0·13–0·18 p.f.u.; biotite 3: X_Fe = 0·54–0·57, Al^VI = 0·25–0·28 p.f.u., Ti = 0·15–0·16. Muscovite is extremely rare in cordierite-bearing assemblages.

Rutile and ilmenite are common accessory phases, which are observed as inclusions in garnet and in the matrix. Ilmenite contains small amounts of ferric iron and manganese; rutile is nearly pure TiO₂. Further accessory phases are zircon, monazite, apatite and sulphides.

Sample ST (garnet-bearing sillimanite–cordierite gneiss)
In this sample two domains can be distinguished on a hand-specimen scale: a massive biotite–plagioclase gneiss domain (STm), which contains the minerals sillimanite, garnet, cordierite, biotite, plagioclase, minor muscovite, quartz, ilmenite, zircon and monazite, and a mylonitic domain (STd), which contains sillimanite, cordierite, garnet, biotite, muscovite, plagioclase, K-feldspar, quartz, zircon, and monazite.

Plagioclase in domain STm has X_an = 0·24 and is generally unzoned, although in some grains, X_an decreases from 0·28 to 0·24 from core to rim (Fig. 5). Plagioclase in domain STd is unzoned, with X_an ranging from 0·24 to 0·26. Perthitic K-feldspar occurs only in domain STd, either as small polygonal grains (Fig. 3f) or as clasts surrounded and/or completely replaced by muscovite (Fig. 3e). Integrated microprobe analyses gave the following K-feldspar components: 85–92 mol % orthoclase (Or), 9–15 mol % albite (Ab), 1 mol % celsian and less than 0·2 mol % anorthite (An) (Table 3). Fibrous sillimanite occurs in both domains and forms needles enclosed in either large plagioclase crystals (Fig. 3d) or statically recrystallized K-feldspar (Fig. 3f). Sillimanite is almost pure, with a low Fe₂O₃ content of 0·3–0·4 wt %.

View larger version (15K): [in this window] [in a new window]   Fig. 5. Plagioclase zonations observed in the metasediment samples TT and ST, and in the hornblende gabbro samples HGn and HGf.

 

View this table: [in this window] [in a new window]   Table 3: Mineral analyses of samples HGn, HGf, DG3, DG4, BCD, TT and ST

 
Garnet is rare, in particular in domain STd, and occurs as small, weakly zoned crystals of up to 0·12 mm in diameter. The mineral is found only as inclusions in plagioclase and rarely as resorbed inclusions in cordierite in domain STm or in cordierite in domain STd, but never in contact with other minerals (Fig. 3c). Even though garnet in domain STm is surrounded by plagioclase, the garnet compositions vary with respect to Fe, Mg and Mn (Fig. 4, right column). Garnet in domain STd has higher X_Fe and X_sps than garnet in domain STm (Fig. 4).

Cordierite in both domains is always pinitized (Fig. 3d). Biotite is widespread in both domains and defines the foliation. It occurs as individual flakes enclosed in plagioclase (Bt1, Fig. 3c) and cordierite (Bt2) in domain STm. Biotite 3 is present in the mylonitic shear zones, where it occurs together with muscovite, cordierite and sillimanite, and rarely with K-feldspar (Fig. 3e). In some places, biotite is intergrown with and replaced by chlorite along the foliation. The compositional range of biotite 1 is narrow, with X_Fe = 0·64–0·66, Al^VI = 0·35–0·41 p.f.u and Ti = 0·15–0·23 p.f.u., whereas those of biotites 2 and 3 are larger: X_Fe = 0·58–0·67, Al^VI = 0·36–0·56 p.f.u and Ti = 0·07–0·23 p.f.u. (Table 3).

In domain STd, muscovite is either intergrown with biotite or formed at the expense of K-feldspar (Fig. 3e). In domain STm, rare muscovite occurs in retrograde domains, parallel to the foliation (Fig. 3c). Muscovite has Si contents between 3·008 and 3·036 p.f.u, and paragonite components between 7·8 and 9·7 mol %. Monazite, ilmenite, apatite and zircon are common accessory phases and are usually enclosed in biotite and plagioclase, locally in cordierite.

Hornblende gabbro complex
From the gabbro two samples were investigated; a medium-grained hornblende gabbro (sample HGn) and a fine-grained gabbro (sample HGf). Both samples contain hornblende, plagioclase, chlorite, epidote, sericite, magnetite, ilmenite, titanite, calcite, apatite and zircon. Sample HGf also contains cummingtonite, biotite and sulphides, and sample HGn rutile (Fig. 6). In both samples, zoned Ca-amphiboles occur (Fig. 7a and c). These show sharply decreasing Ti, Al, K and Na contents from core to rim. In sample HGf the X_Mg ratio [(Mg/(Mg + Fe²⁺)] increases from core to rim, whereas in sample HGn, X_Mg decreases abruptly at the core–rim boundary and then increases again (Fig. 7c). According to the nomenclature of Leake et al. (1997), the hornblende cores in sample HGn are pargasites, and those in sample HGf are tschermakites (Fig. 7a). In both samples, the amphibole cores are overgrown by a rim of an Al-rich magnesio-hornblende (hbl2 in Figs 6 and 7), and in sample HGn additionally by a magnesio-hornblende with lower Al contents (hbl3) (Fig. 7a and c). Some Ca-amphiboles in sample HGn show exsolusions of rutile needles parallel to the hornblende b–c plane (Fig. 6a). Cummingtonite was rarely found in sample HGf, where it forms cores overgrown by hornblende hbl2 (Figs 6e and 7c). Cummingtonite has X_Mg ratios of 0·64–0·66, and low Ti (0·009–0·011 p.f.u.), Al (0·2–0·29 p.f.u.), K (0–0·005 p.f.u.) and Na (0·03–0·06 p.f.u.) contents.

View larger version (116K): [in this window] [in a new window]   Fig. 6. Photomicrographs of hornblende gabbro sample HGn (a and b) and HGf (c–e), and the diorite–gneiss sample DG3 (f). (a) Zoned hornblende crystal showing a core (hbl1) with rutile exsolution lamellae (rt) parallel to the hornblende b–c plane. The core is overgrown by a rutile-free amphibole rim (hbl2), which occurs in contact with chlorite (chl). (b) Hornblende 2 overgrows a zoned, angular plagioclase grain (pl). (c) Ti-rich hornblende (hbl1) with biotite inclusions (bt) is overgrown by a Ti-poor hornblende (hbl2), which contains inclusions of magnetite (mag) and ilmenite (ilm). (d) Zoned plagioclase is overgrown by hornblende 2. (e) Hornblende 2 overgrows cummingtonite (cum) and plagioclase. (f) Hornblende crystal occurs in a groundmass of K-feldspar (kfs), quartz (qtz), plagioclase and biotite.

 

View larger version (34K): [in this window] [in a new window]   Fig. 7. Composition (a and b) and zonation patterns (c) of amphiboles observed in gabbros and diorites of the KCC. (a and b) Composition of amphiboles observed in different domains of the hornblende gabbro samples HGf and HGn and in the diorite samples DG3, DG4 and BCD, shown in the diagram of Leake et al. (1997). m, magmatic stage; s, solidus stage; ss, sub-solidus stage. The ferric iron in (c) was obtained using the method outlined by Spear & Kimball (1984).

 
In both samples hbl2 occludes angular plagioclase crystals (Fig. 6b and d), which have abruptly decreasing anorthite contents from core to rim (from 91 to 37 mol % for HGn; from 51 to 44 mol % for HGf; Fig. 5). Epidote occurs in retrograde domains of both samples, commonly associated with chlorite and, in places, with magnetite. It has pistacite components [X_ps = Fe³⁺/(Fe³⁺ + Al)] of 0·23–0·26 in sample HGn and 0·28–0·29 in sample HGf. Biotite was observed only in sample HGf, where it traces the foliation. It has X_Fe ratios of 0·36–0·40 and Ti contents of 0·15–0·18 p.f.u. Chlorite occurs in both samples and shows the following compositional variations (HGf: ; HGn: . Ilmenite in sample HGn forms inclusions in hornblende hbl1 and hbl2, whereas magnetite occurs either along hornblende grain boundaries or in matrix domains, in places surrounded by epidote. In sample HGf, ilmenite forms inclusions in hbl1 and is intergrown with magnetite in hbl2 and in the surrounding matrix. In some HGf domains magnetite overgrows pyrite. The ilmenite composition, expressed by the end-members ilmenite (ilm), haematite (hem) and pyrophanite (pyp), is ilm_61–87hem_6–33pyp_6–7 in sample HGn, and ilm_82–83hem_8–11pyp_7–8 in sample HGf. Rutile forms either needles in hornblende hbl1 or large crystals on grain boundaries in sample HGn.

Diorite–gneiss complex
Two samples from the diorite–gneiss complex were investigated (DG3 and DG4). Sample DG3 is strongly foliated and contains hornblende, biotite, plagioclase, K-feldspar, quartz, epidote, allanite, chlorite, titanite, apatite, zircon, ilmenite and magnetite. Sample DG4 is more massive and free of quartz and K-feldspar.

Both samples contain angular plagioclase crystals, with anorthite contents of 36–37 mol % (sample DG3) and 46–50 mol % (sample DG4). The angular plagioclase crystals are, in particular in sample DG3, surrounded and/or truncated by round recrystallized plagioclase grains, indicating a dynamic to static recrystallization after deformation. K-feldspar was observed only in the matrix of sample DG3 (Fig. 6f), and has integrated compositions of Or_{0·88–0·92}Ab_{0·08–0·11}. Amphibole crystals in sample DG4 occur enclosed in plagioclase and in the matrix (Fig. 8e and f). They have a tschermakitic composition with little compositional variation (Fig. 7b). In contrast, amphiboles in sample DG3 are magnesio-hornblendes with more variable compositions (Fig. 7b, Table 3). Biotite in sample DG4 occurs enclosed in plagioclase and in the matrix, whereas biotite in sample DG3 traces the foliation and/or is intergrown with hornblende. Biotite in the two samples has the following compositional variations (Table 3): in sample DG4, X_Fe = 0·48–0·50, Al^VI = 0·31–0·35 p.f.u., Ti = 0·26–0·32 p.f.u.; in sample DG3, X_Fe = 0·45–0·46, Al^VI = 0·34–0·38 p.f.u., Ti = 0·26–0·27 p.f.u. Chlorite and epidote occur in the retrograde domains of both samples. Epidote in sample DG3 has X_ps values of 0·28–0·29, and 0·25–0·27 in sample DG4. The chlorites have the following compositional range: for sample DG3, Fe_{1·72–1·79}Mg_2·4–3·06Mn_0·04–0·1Al_{2·15–2·38}Si_{2·73–2·89}O₁₀(OH)₈; for sample DG4, Fe_{1·71–1·75}Mg_2·5–3·18Mn_{0·02–0·09}Al_{2·10–2·42}Si_{2·71–2·91}O₁₀(OH)₈.

View larger version (117K): [in this window] [in a new window]   Fig. 8. Photomicrographs of the diorite samples BCD (a–d) and DG4 (e and f). (a) Domain BCD1: clinopyroxene crystals (cpx) occur in a groundmass of plagioclase (pl) and hornblende (hbl). (b) Statically crystallized hornblende and plagioclase crystals forming a rim around domain BCD1. (c) Domain BCD2: euhedral hornblende crystals are present in a matrix of K-feldspar (kfs), quartz (qtz) and plagioclase (pl), which is strongly sericitized. (d) Domain BCD3: epidote aggregate (ep) in contact with plagioclase, quartz, patchy and zoned hornblende and chlorite (chl). (e) Angular plagioclase surrounded by hornblende and biotite aggregates. (f) Plagioclase crystals with inclusions of hornblende, biotite (bt) and magnetite.

 
Schlieric diorite of the Borntal Intrusive Complex (BCD)
This rock consists of melanocratic hornblende-rich and leucocratic feldspar-rich domains, which grade continuously into each other. In a few melanocratic domains clinopyroxene is present together with plagioclase and amphibole (domain BCD1, Fig. 8a). These domains, which are of centimetre scale, are commonly surrounded by a centimetre-thick amphibole rim (Fig. 8b). The leucocratic domains contain hornblende, biotite, plagioclase, K-feldspar, quartz, titanite, zircon and a little magnetite (domain BCD2, Fig. 8c), and in some places epidote and chlorite (domain BCD3, Fig. 8d).

Amphibole forms xeno- to idiomorphic crystals up to 1·5 cm in size (Fig. 8b and c). In the domains BCD1 and BCD2 the amphiboles are generally brown and unzoned, whereas those in the retrograde domain BCD3 commonly have greenish rims or are patchy. From domain BCD1 to BCD3 the amphiboles show a compositional trend characterized by increasing X_Mg ratios (from 0·70 to 0·82) and silica contents (Si from 6·9 to 7·8 p.f.u), whereas the sodium (0·29 to 0·05 p.f.u.), potassium (0·12 to 0·01 p.f.u.), and titanium contents (0·09 to 0·01 p.f.u.) decrease. This trend is well reflected by an amphibole zonation profile from domain BCD3 (Fig. 7c). This profile indicates that magnesio-hornblende, which is typical for domain BCD1 and BCD2, was overgrown by actinolite in domain BCD3 (Fig. 7a and c).

Clinopyroxene has a diopsidic composition, with X_Mg ratios between 0·82 and 0·84 and low aluminium contents between 0·034 and 0·048 p.f.u. (Table 3). Plagioclase in domain BCD1 is unzoned (X_an = 0·37–0·39), whereas plagioclase in domain BCD2 shows decreasing anorthite contents from core to rim (X_an= 0·37 to 0·29). Microcline K-feldspar observed in domain BCD2 (Fig. 8c) has the following compositional range: Or_{0·85–0·90}Ab_{0·08–0·15}. The anorthite components range between 0·2 and 0·5 mol %, and the celsian component between 0·9 and 1·3 mol %. Biotite occurs in domain BCD2, commonly intergrown with hornblende. With very rare exceptions biotite is transformed into chlorite, with the following compositional variations: . Epidote has X_ps between 0·22 and 0·23.

	INTERPRETATION OF THE PETROGRAPHIC OBSERVATIONS

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
The combination of intergrowth relationships, mineral compositions and mineral zonation patterns in the investigated samples (see above and Figs 3–8) allows the reconstruction of assemblage sequences, which can finally be used to calculate P–T conditions using minerals or mineral zones from distinct domains of a single rock, and to obtain P–T paths through quantitative phase diagrams. In the following, the equilibrium assemblages inferred for the investigated rock samples are characterized using observations from single thin sections.

Sample TT (garnet–biotite–cordierite gneiss)
The occurrence of biotite and euhedral, chemically unzoned garnet inclusions in the core of zoned plagioclase (garnet 1; Figs 3a and 4), and the presence of ilmenite and rutile inclusions in garnet, indicates that the oldest assemblage formed in the garnet–biotite gneiss sample TT is

This assemblage was subsequently transformed into the assemblage

This is concluded from the observations that garnet occurs together, or even in contact, with biotite, ilmenite and rutile along plagioclase grain boundaries in plagioclase-rich domains (garnet 2; Fig. 3a) and is overgrown by biotite and cordierite in biotite-rich, cordierite-bearing domains (garnet 3; Fig. 3b). In the latter two domains, the originally euhedral garnet grains are partially resorbed and are generally round (Fig. 3b). Pinitized cordierite in contact with a few chlorite and muscovite grains provides evidence for a final retrograde alteration, which led to the formation of the assemblage

Sample ST (garnet-bearing sillimanite–cordierite gneiss)
In this sample two domains were distinguished (STm and STd), which show textural and mineralogical variations (see above). In both sample domains plagioclase and quartz are predominant and ilmenite occurs as a minor phase.

Sample STm
The occurrence of rounded garnet, angular biotite and fibrous sillimanite inclusions in the same, or in spatially separated but chemically identical, plagioclase crystals (Fig. 3c) indicates that the oldest assemblage preserved in sample STm is

A few resorbed garnet grains in cordierite, and the general absence of garnet together with cordierite and sillimanite throughout sample STm (Fig. 3d) indicate that garnet was consumed during cordierite growth and finally reacted out, leading to the formation of the assemblages

Sparse muscovite and chlorite flakes, which can locally be observed along foliation planes, are interpreted to be formed during the retrograde alteration.

Sample STd
A single resorbed garnet grain enclosed in cordierite in a sillimanite–K-feldspar domain (Fig. 3f) indicates that garnet was consumed during cordierite formation. Thus the oldest assemblage preserved in sample STd is

Muscovite around resorbed K-feldspar clasts (Fig. 3e), which occurs in contact with cordierite and sillimanite (Fig. 3e and f), indicates that muscovite was formed at the expense of K-feldspar in the assemblage

Pinitized cordierite in contact with chlorite and locally with muscovite supports the conclusion that domain STd was finally affected by a retrograde overprint, leading to the formation of the assemblage

Sample HGn (normal-grained hornblende gabbro)
Observed and measured amphibole and plagioclase zonations (Figs 5 –7) provide evidence that sample HGn underwent three crystallization events. The first event is reflected by Ti-rich pargasites (Hbl1) and anorthitic plagioclase (Pl1: X_an = 0·91) forming the cores of the respective minerals (Figs 5, 6a and 7). The second event is reflected by Al-rich magnesio-hornblende overgrowths (Hbl2) intergrown with a more albitic plagioclase (Pl2: X_an = 0·37; Fig. 6b), and the third event by a Ti-poor, Al-poor magnesio-hornblende (Hbl3), which locally occurs in direct contact with serizitized plagioclase, chlorite and epidote, and also with calcite, titanite and magnetite. Thus, we conclude that the assemblage sequences in sample HGn are

This sequence probably reflects the magmatic (HGn1), solidus (HGn2) and subsolidus stages (HGn3) (see below).

Sample HGf (fine-grained hornblende gabbro)
Magnesio-hornblende (Hbl2) overgrowths around tschermakitic hornblende (Hbl1), cummingtonite and zoned plagioclase [Pl1 (i.e. core): X_an = 0·51; Pl2 (i.e. overgrowth): X_an = 0·44; Figs 5, 6c–e and 7] provide evidence for at least two amphibole and plagioclase crystallization events in sample HGf. In addition, biotite and ilmenite inclusions in hornblende 1, as well as magnetite (with pyrite cores) overgrown by hornblende 2, indicate that the following assemblages were successively formed:

Finally, sample HGf underwent a retrograde overprint, leading to the formation of chlorite at the expense of biotite and amphibole, of sericite at the expense of plagioclase, and of a little epidote. Thus the final assemblage was

Again, the assemblage sequence inferred for sample HGf is interpreted to reflect the magmatic (HGf1), the solidus (HGf2) and the sub-solidus stage (HGf3).

Samples DG3 and DG4 (diorite–gneiss)
Amphibole, plagioclase and biotite in both samples show little chemical variation (Fig. 7) and are closely intergrown. In addition, both samples contain only a little retrograde epidote and chlorite. Thus we conclude that the predominant mineral assemblages were formed during a single event that may be related to magmatic crystallization (see below). The following assemblages were observed (all with biotite and plagioclase):

Sample BCD (schlieric diorite)
The microstructural observations presented above indicate three domains in sample BCD (Fig. 8a–d), which contain the following assemblages (all with plagioclase):

From the macroscopic schlieric structure, and the overgrowth relationships described above, it seems likely that the quartz-free clinopyroxene domain (BCD1) represents a relic rock type, which was mixed with a K-feldspar–quartz-bearing rock (BCD2) during the magmatic stage. Finally, domain (BCD2) underwent a retrograde overprint under sub-solidus conditions, as supported by the formation of actinolite rims around magnesio-hornblende (Fig. 7), and the occurrence of actinolite in contact with retrograde chlorite and epidote.

	GEOTHERMOBAROMETRY

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
Metapelitic rocks
Several conventional geothermobarometers [computer software ‘Thermobarometry’ v. 2.1 of Spear & Kohn (1999; available at http://ees2.geo.rpi.edu/MetaPetaRen/Frame-software.html/)] were employed to obtain metamorphic peak conditions from the metapelitic rock samples TT and ST (Fig. 9). For P–T estimates, the chemical compositions of garnet and biotite inclusion in plagioclase were used (assemblages STm1 and TT1). Analyses of these minerals are listed in Table 3. Furthermore, apparent garnet–biotite temperatures for garnet rims were calculated for comparison (see Fig. 4).

View larger version (26K): [in this window] [in a new window]   Fig. 9. Peak P–T estimates using different calibrations of conventional geothermobarometers for the metasediment samples TT (a) and ST (b). The calculations were carried out with the computer software GTB of Spear & Kohn (1999, available at http://ees2.geo.rpi.edu/MetaPetaRen/Frame-software.html/), using the mineral compositions shown in Table 3. (1–5) Garnet–biotite geothermometer: (1) Perchuk & Lavrent'eva (1981); (2) and (3) Indares & Martignole (1985); (4) Holdaway et al. (1997); (5) Kleemann & Reinhardt (1994); (6) and (7) GRISP geobarometer of Bohlen & Liotta (1986); (8) and (9) GASP geobarometer: (8) Newton & Haselton (1981); (9) Koziol (1989).

 
In addition, P–T pseudosections in the model system CaO–Na₂O–K₂O–TiO₂–MnO–FeO–MgO–Al₂O₃–SiO₂–H₂O (CNKTiMnFMASH) were constructed for samples TT and STm, and in the model system CNKMnFMASH for sample STd (Fig. 10) using the software package THERMOCALC v3.1 (Powell & Holland, 1988; Holland & Powell, 1990, 1998), and employing the procedure and activity–composition relationships as outlined by White et al. (2001) and Zeh et al. (2004). For calculations we used chemical compositions obtained by XRF analyses of a representative hand specimen of sample STm, and chemical compositions reconstituted from mineral modes and compositions obtained from samples TT and STd. Finally, the P–T pseudosections were contoured for X_alm and X_sps, to constrain the observed garnet zonation patterns.

View larger version (94K): [in this window] [in a new window]   Fig. 10. P–T pseudosection calculated in the model system CaO–Na2O–K2O–TiO2–MnO–FeO–MgO–Al2O3–SiO2–H2O for sample TT (a–c) and STm (d), and in CaO–Na2O–K2O–MnO–FeO–MgO–Al2O3–SiO2–H2O for sample STd (e) using THERMOCALC v. 3.1 with the dataset HP98 (Holland & Powell, 1998), and activity models mentioned in the text. It should be noted that all phase fields labelled ‘liq’ do not contain H2O as a free phase. The black circles labelled 1–3 represent assemblages observed in the respective samples. Bold continuous line indicates inferred P–T path segments. (b and c) Contoured P–T pseudosections for sample TT, showing the change of Xalm (b) and Xsps (b) during the retrogression.

 
Conventional geothermobarometry yielded peak P–T conditions of 690–780°C at 7 kbar for sample TT, and of 700–740°C at 5–6 kbar for sample STm (Fig. 9, Table 4). These P–T conditions conform with stability ranges for mineral assemblages as predicted by the P–T pseudosections, showing that at c. 730°C and 6–7 kbar the following observed mineral assemblages are stable: for TT1, bt–grt–pl–qtz–rt–ilm–(liq); for STm1, bt–grt–sil–ilm–pl–qtz–(liq); for STd1, bt–grt–sil–ilm–pl–kfs–qtz–(liq) (Fig. 10).

View this table: [in this window] [in a new window]   Table 4: Results of geothermobarometry

 
The phase diagrams in Fig. 10 also predict that cordierite assemblages occur only at pressures <5 kbar, and that garnet will be consumed during decompression whereas cordierite is formed. Thus we conclude that the observed cordierite assemblages and the garnet resorption patterns in samples TT and ST (assemblages TT2 and STm2) result from a post-peak metamorphic pressure decrease to <5 kbar. The presence of chlorite and/or muscovite could be explained by a subsequent temperature fall to <580°C at <4 kbar. This P–T evolution conforms with that obtained by comparison between the predicted and observed assemblage sequence STd1 and STd3, which requires a P–T path from 730°C at 6 kbar to 700°C at 4 kbar and to <650°C at 4 kbar (Fig. 10e). It is also in agreement with the observed increase of X_alm and X_sps toward the rims of the garnet grains in sample TT (Fig. 10b and c).

Gabbros and diorites
To infer P–T information for the gabbro and diorite samples several calibrations of the Al-in-hornblende geobarometer (Hammarstrom & Zen, 1986; Hollister et al., 1987; Schmidt, 1992), the Ti–Al-in-hornblende geothermobarometer (Ernst & Liu, 1998) and the hornblende–plagioclase geothermometer (Holland & Blundy, 1994) were employed using the composition of coexisting Ca-amphibole and plagioclase from the domains defined above, and hornblende Fe³⁺ calculation procedures as recommended by the respective researchers. Application of the geothermobarometers used involves some restrictions, which will be discussed below. The Al-in-hornblende geobarometers of Hammarstrom & Zen (1986), Hollister et al. (1987) and Schmidt (1992) are calibrated for rocks with the buffered mineral assemblage hornblende + biotite + plagioclase + K-feldspar + quartz + sphene + Fe–Ti-oxides + melt + fluid phase, which is present in the samples DG3 and BCD (except melt and fluid phase). In contrast, samples HGf, HGn and DG4 are quartz and K-feldspar free, and sample HGn is additionally biotite free (Table 4). The Ti–Al-in-hornblende geothermobarometer is calibrated using mid-ocean ridge basalts (MORB), which have no, or only a very little, normative quartz and orthoclase, but normative magnetite and ilmenite (Ernst & Liu, 1998). These requirements are generally fulfilled for samples HGn, HGf and DG4, even though these rocks do not have a MORB composition, but rather are island-arc lithologies (A. Zeh, unpublished data, 2004). Hammarstrom & Zen (1986), Hollister et al. (1987) and Schmidt (1992) suggested that their Al-in-hornblende geobarometers are temperature independent, whereas Ernst & Liu (1998) showed that the Al content in the amphiboles investigated depends on both temperature and pressure (see Fig. 12). The two equilibrium reactions of the hornblende–plagioclase geothermometer of Holland & Blundy (1994) (see Fig. 11) can be used on silica-saturated rocks (samples DG3 and BCD), and one reaction is applicable for silica-undersaturated rocks (samples HGn, HGf and DG4). The calibrations used and the results are shown in Table 4 and Figs 11 and 12.

View larger version (23K): [in this window] [in a new window]   Fig. 11. Results of P–T estimates on gabbros (a and b) and diorites (c–e) of the KCC using the Al-in-hornblende geobarometer of Schmidt (1992), and the hornblende–plagioclase geothermometer of Holland & Blundy (1994): T1, edenite + 4quartz = tremolite + albite; T2, edenite + albite = richterite + anorthite. It should be noted that samples HGf, HGn and DG4 are not buffered by quartz and K-feldspar and, thus, pressure results are shown by white squares. Reaction T1 is shown for comparison. m, magmatic stage; s, solidus stage.

 

View larger version (28K): [in this window] [in a new window]   Fig. 12. Results of P–T estimates on gabbros and diorites of the KCC using the Al–Ti-in-hornblende geothermobarometer of Ernst & Liu (1998). m, magmatic stage; s, solidus stage; ss, sub-solidus stage (for further explanation see text).

 
For the quartz–K-feldspar-free samples HGn and HGf P–T conditions of c. 850–900°C at 5–8 kbar were estimated using the composition of Hbl–Pl cores (assemblages HGn1 and HGf1), and 750–770°C at 4–7 kbar using Hbl–Pl rims (assemblages HGn2 and HGf2). Similar P–T conditions of 800°C at 6·5 kbar were obtained for sample DG4, and the buffered assemblage in sample DG3 (770°C at 4–6 kbar). In contrast, the buffered assemblage in rock BCD yielded lower P–T conditions of 700–740° at 2–3 kbar. The lowest P–T conditions of 600°C at 2·5 kbar and 480°C at 2 kbar were obtained with the Al–Ti-in-hornblende geothermobarometer for the retrograde, sub-solidus domains HGn3 and BCD3, respectively (Fig. 12).

Discussion of the petrological results
The petrological results indicate that the metasediments (samples ST and TT) of the magmatic–metamorphic complex experienced peak P–T conditions of 690–770°C at 5–7 kbar, which are, within error, identical to the P–T conditions inferred from solidus assemblages of the gabbro samples HGn and HGf (Hbl2-Pl2: T = 720–780°C, P = 4–7 kbar), and from the diorite samples DG3 and DG4 (matrix assemblages: T = 750–810°C, P = 4–7 kbar) (Table 4, Figs 9 –12). Al-in-hornblende geobarometry on all solidus assemblages yields, within error, identical pressures of about 4–7 kbar (Fig. 11), independently of whether the rocks contain a buffered mineral assemblage or not. In contrast, the Al–Ti-in-hornblende geothermobarometer of Ernst & Liu (1998) yields unrealistically high pressures for assemblage HGf2 of about 10 kbar (Fig. 10). This is in disagreement with the results obtained from all other rocks (Figs 9 and 11), and also with the field relationships, which indicate that samples HGn and HGf must have undergone an identical magmatic–metamorphic evolution.

The P–T data for all rocks from the magmatic–metamorphic complex support the model implied by the geochronological investigations (see above); that is, that the metasediments of the KCC were affected by a contact metamorphic event at c. 341 Ma, as a result of gabbro and diorite intrusion. The high contact metamorphic temperatures of 690–770°C can be explained by the fact that the metasediments represent only a small rock volume within the voluminous gabbro and diorite bodies (Fig. 1).

The P–T data in combination with the textural observations and mineral chemical data indicate that the predominant domains in the dioritic rocks reflect the solidus stage (except the few retrograde sub-solidus domains), whereas the solidus stage in the hornblende gabbros is represented only by the composition of a few mineral rims (Hbl2–Pl2). The predominant core assemblages (Hbl1–Pl1) in the hornblende gabbro reflect instead the magmatic stage, as supported by the estimated temperatures between 850 and 900°C (Table 4).

The results of Al-in-hornblende geobarometry lead to the interpretation that the pressure during the magmatic stage was higher (c. 8 kbar) than during the solidus stage (c. 4–7 kbar). This conforms with the results obtained by Al–Ti-in-hornblende geothermobarometry from sample HGn, but not with the results obtained from sample HGf (Fig. 12). Thus, it remains ambiguous whether or not the obtained pressure difference has any geological significance, especially as samples HGn, HGf and DG4 do not have the required buffering assemblages. Thus, there are several degrees of freedom that may be responsible for the different Al and Ti contents measured in the investigated amphiboles [see discussion by Schmidt (1992)].

The large variation of the hornblende composition in the buffered diorite sample DG3, which results in pressure variations from 4 to 6 kbar (Al-in-Hbl geobarometer, Table 4), could be explained by successive sub-solidus re-equilibration. This re-equilibration was perhaps triggered by sub-solidus deformation, which caused the formation of the observed foliation and of recrystallized plagioclase grains. Thus, it is likely that the highest pressure conditions obtained from sample DG3 (c. 6·0 kbar) represent the solidus conditions. These pressures are identical to that obtained from the nearby undeformed diorite sample DG4 (6 kbar), whose amphiboles show little compositional variation (Figs 7 and 8e, f). Sub-solidus re-equilibration in sample DG3 could also account for the inconsistent results obtained with the two calibrations of the Hbl–Pl geothermometer of Holland & Blundy (1994). Figure 11 shows that, apart from sample DG3, the hornblende–plagioclase reactions (T1) and (T2) yield results consistent with that obtained with the Al-in-hornblende geobarometer (Fig. 11). This consistency is unexpected, as samples HGf, HGn and DG4 are silica undersaturated, and thus reaction (T1) should give unrealistic temperatures [see discussion by Holland & Blundy (1994)].

Considering P–T–t results, field relationships, and the petrographic observations on the gabbros, diorites and metasediments of the magmatic–metamorphic complex, we conclude that magmatic crystallization and contact metamorphic overprint occurred at pressures of 5–7 kbar at c. 341 Ma. This corresponds to a depth of c. 19–26 km, assuming that the lithostatic pressure gradient was 270 bar/km.

Subsequently, at c. 337 Ma, the magmatic–metamorphic complex was intruded by magmatic rocks of the BIC, as constrained by field relationships and geochronological results. Pressures of 2–3 kbar obtained from the schlieric diorite (Fig. 11, Table 4) indicate that the intrusions of the BIC occurred at c. 7–11 km, which is shallower than the level inferred from the western magmatic–metamorphic complex. The combination of the P–T data from the magmatic–metamorphic complex (6 ± 1 kbar at 341 ± 1 Ma) and from the BIC (2–3 kbar at 337 ± 1 Ma) provides evidence that the KCC underwent rapid exhumation from 22 ± 3 km to 9 ± 2 km within 4 ± 2 Myr, which requires average exhumation rates between 0·12 and 0·93 cm/year.

Fast exhumation conforms with the petrological results obtained from the metapelitic rocks, which provide evidence for a nearly isothermal decompression from 5–7 kbar at 730°C to 4 kbar at 700°C, followed by nearly isobaric cooling to 580°C at 3·5 kbar (Fig. 10). The retrograde evolution inferred from the metasediments conforms with the P–T conditions of 600°C at 2·5 kbar and 480°C at 2 kbar, obtained by Al–Ti-in-hornblende geothermobarometry from the retrograde domains HGn3 and BCD3 of the gabbro and diorite samples (Fig. 12, Table 4).

	GEOSPEEDOMETRY

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
To constrain the cooling history of the KCC, in addition to the results obtained from geochronology and petrology, the T–t history was reconstructed by garnet geospeedometry, using the approach of Ehlers et al. (1994), on coexisting garnet and biotite in the garnet–biotite gneiss sample TT. The garnet–biotite gneiss is well suited for this approach, because it contains garnet grains of very different sizes, which are invariably enclosed in biotite. These are prerequisites of the technique outlined by Ehlers et al. (1994). In the biotite domain (TTb) the garnet radii range from 11 to 275 µm. Furthermore, garnet was almost unzoned at the metamorphic peak (see above), and biotite is the only other Fe–Mg phase, except for sparse chlorite and cordierite. However, other restrictions inherent to their technique, namely perfectly spherical and pure Fe–Mg garnet, are not entirely fulfilled and, of course, are very unlikely to be met by nature. Effects on the results caused by CaO in garnet can probably be neglected, as the grossular content is rather low and nearly constant (Fig. 4). However, the observed inverse relationship between garnet size and spessartine contents might lead to a systematic error, which increases with decreasing garnet size. To perform the cooling rate calculations, it is first necessary to determine the apparent garnet closure temperatures, T'_c(y). This was done using the Kleemann & Reinhardt (1994) garnet–biotite geothermometer (see Figs 4 and 13). The same biotite composition was used in all closure temperature calculations.

View larger version (28K): [in this window] [in a new window]   Fig. 13. (a, c) T'c(y) vs I' diagram; (b, d) T'c(y) vs ln s(y) diagram; (e) T–t diagram; where y is the sectioning position in the garnet (centre y = 0; rim y = 1), T'c(y) is the apparent garnet closure temperature, T'c(0) corresponds to the actual closure temperature; I' is the apparent garnet radius, and s(y) is the cooling rate per million years. •, garnets used for estimation of the cooling rate history; , cut off-centre garnets; x, garnets whose radii are inferred to be underestimated. The regression results are shown by the bold black lines; the obtained regression functions are given. (a, b) cooling rate history estimated according to model 1 (continuous change of cooling rates); (c, d) cooling rate history estimated according to model 2 (two-stage cooling history). (e) T–t histories inferred by integration of the cooling rate functions shown in (b). The star at about 640°C indicates the change from function (1) to (2) in model 2 (for further explanation see text).

 
Figure 13a–d shows a series of T'_c(y) vs I' and T'_c(y) vs ln s(y) diagrams (for symbol explanation see Fig. 13 caption), which can be used to assess the quality of the input data (see Ehlers et al., 1994). If all garnets were cut exactly through their respective centres, all data points in Fig. 13a–d should lie on an exponential curve or a straight line. As is obvious from these diagrams, not all data points fulfil this requirement, as would be expected using natural samples. Garnets cut off-centre will always lie above the exponential curve (Fig. 13a and c), because their apparent closure temperature T'_c(0) is underestimated (Ehlers et al., 1994). In contrast, a wrong estimate of the apparent garnet radius I', as a result of non-spherical garnet shape and/or retrograde garnet resorption, may lead to an over- or underestimation of T'_c(0).

Regardless of the algorithm used to fit the data in Fig. 13a–d, there cannot be a unique solution that satisfies all data points. In contrast, there will always be garnets that plot either above or below such a regression curve. Apart from errors in the estimated apparent closure temperatures and garnet radii, another possibility to explain this would be a change in cooling rate during exhumation of the rock. If this was the case, the data would not be expected to fit a single curve. As drastically varying cooling rates cannot be excluded during the exhumation of the garnet–biotite gneiss, we model the cooling history of this rock by assuming (1) a continuous change in the cooling history, and (2) a two-stage cooling history. In the context of model 1, all garnet data must be fitted by a single function; in the context of model 2, garnets smaller and larger than 70 µm are fitted separately.

The results are shown in Fig. 13b and d. The cooling rate increases constantly with falling temperatures in model 1, whereas in model 2, the cooling rates increase rapidly at temperatures above some 640°C, but much more slowly below this temperature. The results of our modelling approach are shown in a temperature–time diagram (Fig. 13e). It is obvious from this diagram that, regardless of the assumptions used, the garnet–biotite gneiss from the KCC must have cooled very rapidly from 690°C to 350°C. Model 1 yields a time span of some 1 Myr and model 2 of c. 1·4 Myr. Assuming that the investigated garnets were formed at the metamorphic peak (>690°C) in the KCC at c. 341 Ma (U–Pb monazite age), the rocks must have cooled to 350°C by about 340 Ma.

This very fast cooling suggested by the results of garnet geospeedometry conforms with the P–T–t results presented above, and with a K–Ar muscovite cooling age of 333 ± 7 Ma obtained from a pegmatitic dyke cross-cutting the Bärenkopf granite (Neuroth, 1997). Assuming a granite crystallization temperature of 750°C at 337 ± 3 Ma (U–Pb zircon), and cooling to below 400°C [the suggested muscovite closure temperature of Kirschner et al. (1996)] at 333 ± 7 Ma, linear cooling rates between 25 and >350°C/Myr (average 90°C/Myr) are possible. These are within the range inferred by garnet geospeedometry (c. 200°C/Myr).

Regardless of the assumptions made, the geospeedometry modelling results presented above indicate that cooling of the KCC rocks was not constant, but point to an increase in cooling rates with falling temperature (Fig. 13). This might correspond to a predominantly erosion-driven exhumation history (England & Thompson, 1984) and is consistent with our petrological results, which indicate an almost isothermal decompression followed by near-isobaric cooling (Figs 10 and 14). The change of the cooling history of the western magmatic–metamorphic complex as suggested by garnet speedometry (model 2 in Fig. 11c, d and e) could be explained by either the magmatic intrusion of the BIC at 337 Ma and/or a decreasing exhumation rate.

View larger version (29K): [in this window] [in a new window]   Fig. 14. Synopsis of the P–T–t data obtained from rocks of the KCC. 1U–Pb zircon and monazite ages (this study); 2K–Ar muscovite age of Neuroth (1997).

 
Geotectonic implications
Detrital zircon cores indicate that metasediments of the KCC were deposited either during the Early Ordovician (470 Ma) or later. Successions comprising Ordovician and younger sediments and volcanoclastic rocks are well known from the Northern Phyllite Belt (Fig. 1) exposed in the southeastern Harz mountains c. 30 km north of the KCC (see Burmann, 1973; Schwab & Jacob, 1996), and are widespread in the Saxothuringian domain, SE of the Mid-German Crystalline Rise (e.g. Linnemann & Buschmann, 1995). Thus the metasediments of the KCC are likely to represent contact metamorphosed equivalents of these lithologies.

The P–T–t results imply that the metasediments of the KCC underwent a contact metamorphic overprint at a depth of 19–26 km (5–7 kbar), caused by gabbro and diorite intrusions at c. 341 ± 1 Ma. By 337 ± 1 Ma the metasediments, together with the solidified magmatic rocks, were uplifted to 7–11 km (2–3 kbar), and then intruded by a new suite of magmatic rocks. Subsequently, the magmatic–metamorphic rocks of the KCC underwent accelerated cooling to temperatures below 400°C by 333 ± 7 Ma. Final exhumation of the present-day KCC outcrop occurred until c. 300 Ma, when the eroded basement rocks were overlain by molasse sediments of the Mansfeld beds.

The inferred intrusion and cooling ages of the KCC conform well to those obtained from other crystalline complexes in the Mid-German Crystalline Rise, e.g. from the western Odenwald Crystalline Complex (Kreuzer & Harre, 1975; Hellmann et al., 1982; Schubert et al., 2001) and from the Ruhla Crystalline Complex (Zeh et al., 2001, 2003), indicating that the Mid-German Crystalline Rise was affected by widespread magmatic activities during the Viséan [for more details see the discussion by Zeh et al. (2003)]. Based on geochemical and geochronological data from the Odenwald Crystalline Complex, Altherr et al. (1999) suggested that island-arc magmatism during the Viséan occurred during closure of the Rhenohercynian ocean, which graded from a ‘pure’ SE-directed subduction at 360 Ma into collision at 330–340 Ma. Such a collision scenario is also likely to explain the inferred magmatic–metamorphic evolution of the KCC, and the structural–metamorphic evolution of the adjacent Northern Phyllite Belt.

Results from K–Ar dating of Ahrendt et al. (1996) indicate that metamorphic white micas from the Northern Phyllite Belt were formed and/or structurally reset at about 340 Ma. This interpretation conforms with petrological and structural investigations by Theye & Seidel (1993) and Jacob & Franzke (1992), respectively, who showed that different rock units of the Northern Phyllite Belt underwent a low-grade metamorphism at peak P–T conditions of 3–4 kbar at 320°C, accompanied and followed by complex stacking as a result of NW-directed crustal shortening. This indicates that rocks from the Northern Phyllite Belt underwent a structural–metamorphic overprint at the same time as the KCC was affected by magmatism, uplift, deformation and cooling. Thus, by analogy, we conclude that accelerated exhumation and cooling of the Kyffhäuser rocks between 341 ± 1 and 337 ± 1 (333 ± 7) Ma may also be a result of NW-directed crustal shortening, owing to progressive collision between the Rhenohercynian and Saxothuringian domains. It may be that collision caused underplating of ‘light’ material underneath the MGCR and Northern Phyllite Belt and thus generated an isostatic, buoyant uplift. A more detailed interpretation of the exhumation mechanism of the KCC (e.g. orthogonal convergence vs subvertical strike-slip shear) cannot be given here, as structural data, which could provide evidence about shear senses and directions, are not available so far. Nevertheless, the age constraints inferred for the provenance and magmatic–metamorphic evolution give a clear hint that the KCC formed part of the Saxothuringian upper plate (unit I; Fig. 1) and, thus, could well be an equivalent of the rocks exposed in the western part of the Odenwald Crystalline Complex as suggested by Oncken (1997).

However, it should be noted that the magmatic–metamorphic evolution inferred for the KCC is not coherent for all c. 340 Ma island-arc-related units within the Mid-German Crystalline Rise (unit I; Fig. 1). In contrast to the KCC, metamorphic rocks within the central part of the Odenwald Crystalline Complex (unit II sensu Krohe, 1991) underwent an anti-clockwise P–T evolution with peak P–T conditions of >700°C at 4 kbar (Will & Schmädicke, 2003). As discussed by Will & Schmädicke (2003) this P–T evolution is closely related to the intrusion of voluminous magmatic rocks, which probably caused magmatic loading. The geochronological data of Todt et al. (1995) indicate that this ‘dynamic regional contact metamorphic event’ occurred at 336–337 Ma, within the range of the age data obtained from the KCC.

	SUMMARY AND CONCLUSIONS

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
(1) Detrital zircon U–Pb ages of 470 Ma and 560–660 Ma provide evidence that the KCC contains sediments that are as young as Early Ordovician, and are derived from magmatic rocks formed during formation (>550 Ma) and break-up (470 Ma) of the Avalonian–Cadomian orogenic belt.

(2) Field relationships in combination with geochronological and petrological results indicate that metasediments of the KCC were intruded by gabbros and diorites at 341 ± 1 Ma. These intrusions caused a high-grade contact metamorphic overprint with peak temperatures between 690 and 750°C at 5–7 kbar, corresponding to intrusion depths of 19–26 km. By 337 ± 1 Ma, the entire sequence was exhumed to a crustal level of 7–11 km, intruded by a new suite of granites and diorites, and cooled quickly to below 400°C at 333 ± 7 Ma.

(3) Rapid exhumation and cooling of the magmatic–metamorphic rocks conforms with results obtained by garnet geospeedometry, and can be explained by crustal shortening and stacking at c. 340 Ma, owing to collision between the Rhenohercynian and Saxothuringian domains. This interpretation agrees well with P–T–t and structural results from the KCC and the adjacent Northern Phyllite Belt.

(4) Petrological results show that gabbroic and dioritic rocks are capable of preserving domains from which detailed information about their magmatic, solidus and sub-solidus P–T evolution can be obtained. Furthermore, Al-in-hornblende geobarometry (e.g. Schmidt, 1992) and hornblende–plagioclase geothermometry (Holland & Blundy, 1994) give reliable P–T information for gabbros and diorites, even if the assemblages are not buffered. Finally, our results demonstrate that P–T pseudosections for complex model systems (e.g. CNKMnTiFMASH) are well suited to infer detailed P–T paths from high-grade metamorphic rocks.

	APPENDIX

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
Appendix A: isotope dilution U–Pb analysis of zircon and monazite
Zircon and monazite were prepared at the NERC Isotope Geosciences Laboratory in Keyworth, Nottingham, UK, using standard crushing and heavy mineral separation techniques. Selected grains were separated under alcohol. All zircon fractions were abraded (Krogh, 1982), and then washed in 4N HNO₃ and H₂O to remove all traces of pyrite. Monazite grains were only washed in 1N HNO₃ for 20 min and then in H₂O. U and Pb were extracted using the methods of Krogh (1973) and Corfu & Ayres (1984). Fractions were spiked with a mixed ²⁰⁵Pb + ²³⁵U isotopic tracer prior to digestion and chemistry (Krogh & Davis, 1985). U and Pb were loaded together onto outgassed single Re filaments with silica gel. Isotope analyses were performed at the NERC Isotope Geosciences Laboratory, Keyworth, on a VG 354 mass spectrometer using a Daly photomultiplier ion counting system, supplemented in part by multiple Faraday cup collection of the larger ion beams. Pb isotope ratios were corrected for initial common Pb in excess of laboratory blank using the common Pb evolution model of Stacey & Kramers (1975). The laboratory blank during the period of analysis was 3 pg. Ages were calculated using the decay constants recommended by Steiger & Jäger (1977). Data reduction was carried out using PBDAT (Ludwig, 1989). The data were plotted using ISOPLOT (Ludwig, 1990).

Appendix B: microprobe and XRF analysis
Electron microprobe analyses were carried out with a CAMECA SX-50 microprobe at the Mineralogical Institute, University of Würzburg, Germany, with three independent wavelength-dispersive crystal channels. Instrument conditions were 15 kV acceleration voltage, 15 nA specimen current, and 20 s integration time for all elements except for Fe (30 s). Natural and synthetic silicates and oxides were used for reference, and matrix corrections were carried out by the PAP program supplied by CAMECA. Point analyses were performed with a 5 µm beam diameter for K-feldspar and muscovite, and a 1 µm beam diameter for all other minerals. For perthitic K-feldspar and hornblende cores with rutile exsolution, a 20 µm beam diameter was used, and subsequently 5–20 analyses were integrated to obtain the original mineral compositions. Detection limits (1) for a typical silicate analysis were 1% relative for each element. Bulk compositions used for phase diagram calculations were analysed by conventional X-ray fluorescence spectrometry using a Phillips PW 1480 spectrometer, and lithium tetraborate fusion discs.

	ACKNOWLEDGEMENTS

 
A.Z. thanks Lutz Gebhardt and Joachim Franzke for stimulating discussions in the field, and Karine David, Jane Evans and Steve Noble (NIGL Keyworth) for their support with the isotope analyses. We also thank Fernando Corfu (University of Oslo), Kurt Stüwe (University of Graz) and Onno Oncken (GFZ Potsdam) for helpful reviews.

---

^* Corresponding author. Telephone: +49-931-888-5415. E-mail: armin.zeh{at}mail.uni-wuerzburg.de

	REFERENCES

TOP ABSTRACT INTRODUCTION GEOLOGICAL SETTING GEOCHRONOLOGY PETROGRAPHY AND MINERAL... INTERPRETATION OF THE... GEOTHERMOBAROMETRY GEOSPEEDOMETRY SUMMARY AND CONCLUSIONS APPENDIX REFERENCES

 
Ahrendt, H., Franzke, H. J., Marheine, D., Schwab, M. & Wemmer, K. (1996). Zum Alter der Metamorphose in der Wippraer Zone/Harz–Ergebnisse von K/Ar-Altersdatierungen an schwach metamorphen Sedimenten. Zeitschrift der Deutschen Geologischen Gesellschaft 147, 39–56.

Altherr, R., Henes-Klaiber, U., Hegner, E. & Satir, M. (1999). Plutonism in the Variscan Odenwald (Germany): from subduction to collision. International Journal of Earth Sciences 88, 422–443.[CrossRef][Web of Science]

Bohlen, S. R. & Liotta, J. J. (1986). A barometer for garnet amphibolites and garnet granulites. Journal of Petrology 27, 328–344.

Burmann, G. (1973). Das Ordovizium der nördlichen Phyllitzone. Teil II: Wippraer Zone. Zeitschrift für Geologische Wissenschaften 1, 9–43.

Corfu, F. & Ayres, L. D. (1984). U–Pb ages and genetic significance of heterogeneous zircon populations in rocks from the Favourable Lake area, northwestern Ontario. Contributions to Mineralogy and Petrology 88, 86–101.[CrossRef][Web of Science]

Ehlers, K., Powell, R. & Stüwe, K. (1994). Cooling rate histories from garnet + biotite equilibrium. American Mineralogist 79, 737–744.[Abstract]

England, P. C. & Thompson, A. B. (1984). Pressure–temperature–time path of regional metamorphism. I. Heat transfer during the evolution of regions of thickened continental crust. Journal of Petrology 25, 894–928.[Abstract/Free Full Text]

Ernst, W. G. & Liu, J. (1998). Experimental phase-equilibrium study of Al- and Ti-contents of calcic amphibole in MORB—a semiquantitative thermometer. American Mineralogist 83, 952–969.[Abstract]

Geist, E. L., Vallier, T. L. & Scholl, D. W. (1994). Origin, transport, and emplacement of an exotic island-arc terrane exposed in eastern Kamchatka, Russia. Geological Society of America Bulletin 196, 1182–1194.

Hammarstrom, J. M. & Zen, E. (1986). Aluminium in hornblende: an empirical igneous geobarometer. American Mineralogist 71, 1297–1313.[Abstract]

Hansch, R. &, Zeh, A. (2000). Metabasites from the Ruhla Crystalline Complex: evidence for distinct pre-Variscan, plate-tectonic environments within the Mid-German Crystalline Rise. Chemie der Erde 60, 1–25.[Web of Science]

Hellmann, K.-N., Lippolt, H. J. & Todt, W. (1982). Interpretation der Kalium–Argon-Alter eines Odenwälder Granodioritporphyritganges und seiner Nebengesteine. Aufschluss 33, 155–164.

Holdaway, M. J., Mukhopadhyay, B., Dyar, M. D., Guidotti, C. V. & Dutrow, B. L. (1997). Garnet–biotite geothermometry revised: new Margules parameters and a natural specimen data set from Maine. American Mineralogist 82, 582–595.[Abstract]

Holland, T. J. B. & Blundy, J. (1994). Non-ideal interactions in calcic amphiboles and their bearing on amphibole–plagioclase thermometry. Contributions to Mineralogy and Petrology 116, 433–447.

Holland, T. J. B. & Powell, R. (1990). An enlarged and updated internally consistent thermodynamic dataset with uncertainties and correlations: the system K₂O–Na₂O–CaO–MgO–MnO–FeO–Fe₂O₃–Al₂O₃–TiO₂–SiO₂–C–H₂–O₂. Journal of Metamorphic Geology 8, 89–124.[Web of Science]

Holland, T. J. B. & Powell, R. (1998). An internally consistent thermodynamic data set for phases of petrological interest. Journal of Metamorphic Geology 16, 309–343.[CrossRef][Web of Science]

Hollister, L. S., Grissom, G. C., Peters, E. K., Stowell, H. H. & Sisson, V. B. (1987). Confirmation of the empirical correlation of Al in hornblende with pressures of solidification of calc-alkaline plutons. American Mineralogist 72, 231–239.[Abstract]

Indares, A. & Martignole, J. (1985). Biotite–garnet geothermometry in the granulite facies: the influence of Ti and Al in biotite. American Mineralogist 70, 272–278.[Abstract]

Jacob, G. & Franzke, H. J. (1992). Die tektonische Entwicklung der Wippraer Zone im Unterharz. Zentralblatt für Geologie und Paläontologie 51–61.

Kirsch, H., Kober, B. & Lippolt, H. J. (1988). Age of intrusion and rapid cooling of the Frankenstein gabbro (Odenwald, SW Germany) evidence by ⁴⁰Ar/³⁹Ar and single zircon ²⁰⁷Pb/²⁰⁶Pb measurements. Geologische Rundschau 77, 693–711.

Kirschner, D. L., Cosca, M. A., Hunziker, J. C. & Masson, H. (1996). Staircase ⁴⁰Ar/³⁹Ar spectra of fine-grained white mica: timing and duration of deformation and empirical constraints on argon diffusion. Geology 24, 747–750.[Abstract/Free Full Text]

Kleemann, U. & Reinhardt, J. (1994). Garnet–biotite thermometry revisited: the effect of Al^vi and Ti in biotite. European Journal of Mineralogy 6, 925–941.[Abstract/Free Full Text]

Koziol, A. M. (1989). Recalibration of the garnet–plagioclase–Al₂SiO₅–quartz (GASP) geobarometer and applications to natural parageneses. American Mineralogist 70, 493.

Kreuzer, H. & Harre, W. (1975). K/Ar-Altersbestimmungen an Hornblenden und Biotiten des kristallinen Odenwaldes. Aufschluss, Sonderband 27, 71–77.

Krogh, T. E. (1973). A low contamination method for the hydrothermal decomposition of zircon and extraction of U and Pb for isotopic age determinations. Geochimica et Cosmochimica Acta 37, 485–494.[CrossRef][Web of Science]

Krogh, T. E. (1982). Improved accuracy of U–Pb zircon ages by the creation of more concordant systems using an air abrasion technique. Geochimica et Cosmochimica Acta 46, 637–649.[CrossRef][Web of Science]

Krogh, T. E. & Davis, L. (1985). The production and preparation of ²⁰⁵Pb for use as a tracer for isotope dilution analysis. Carnegie Institution of Washington, Yearbook 74, 416–417.

Krohe, A. (1991). Emplacement of synkinematic plutons in the Variscan Odenwald (Germany) controlled by transtensional tectonics. Geologische Rundschau 80, 391–409.

Leake, B. E., Woolley, A. R., Arps, C. E. S., Birch, W.D., Gilbert, M. C., Grice, J. D., et al. (1997). Nomenclature of amphiboles. European Journal of Mineralogy 9, 623–651.[Abstract/Free Full Text]

Linnemann, U. & Buschmann, B. (1995). Die cadomische Discordanz im Saxothuringikum (oberkambrisch–tremadocische overlap-Sequenzen). Zeitschrift für Geologische Wissenschaften 23, 707–727.

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 zircon geochronology and basin development (Central European Variscides, Germany). In: Franke, W., Haak, W., Oncken, O. & Tanner, D. (eds.) Orogenic Processes: Quantification and Modelling in the Variscan Belt. Geological Society, London, Special Publications 179, 131–153.

Ludwig, K. R. (1989). PBDAT: a computer program for processing Pb–U–Th isotope data, version 1.20. US Geological Survey Open-File Report 88-542.

Ludwig, K. R. (1990). Isoplot: a plotting and regression program for radiogenic isotope data, version 2.03. US Geological Survey Open-File Report 88-557.

Menning, M., van Amerom, H. W. J., Drozdzewski, G. & Weyer, D. (1996). Integrative Kalibrierung der mittel- und westeuropäischen Karbon-Zeitskala. Exkursionsführer und Veröffentlichung der Gesellschaft für Geologische Wissenschaften, Claustal–Zellerfeld 197, 59–61.

Nance, D. R. & Murphy, B. J. (1994). Contrasting basement isotopic signatures and the palinspastic restoration of peripheral orogens: example for the Neoproterozoic Avalonian–Cadomian belt. Geology 22, 617–620.[Abstract/Free Full Text]

Neumann, W. (1968). Das Grundgebirge des Kyffhäusers (Nordthüringen)—Lithostratigraphie, Tektonik, Regionalmetamorphose und Migmatisation. Geologie (Beiheft 63), 1–79.

Neuroth, H. (1997). K/Ar-Datierungen an detritischen Muscoviten—‘Sicherungskopien’ orogener Prozesse am Beispiel der Varisziden. Ph.D. thesis, University of Göttingen, 134 pp.

Newton, R. C. & Haselton, H. T. (1981). Thermodynamics of the garnet–plagioclase–Al₂SiO₅–quartz (GASP) geobarometer. In: Newton, R. C., Navrotsky, A. & Wood, B.R. (eds) Thermodynamics of Minerals and Melts. New York: Springer, pp. 131–147.

Oh, C. W., Kim, S. W., Ryu, I., Okada, T., Hyodo, H. & Itaya, T. (2004). Tectono-metamorphic evolution of the Okcheon metamorphic belt, South Korea: tectonic implications in East Asia. Island Arc 13, 387–402.[CrossRef][Web of Science]

Oncken, O. (1997). Transformation of a magmatic arc and an orogenic root during oblique collision and its consequences for the evolution of the European Variscides (Mid-German Crystalline Rise). Geologische Rundschau 86, 2–20.

Perchuk, L. L. & Lavrent'eva, I. V. (1981). Experimental investigation of exchange equilibria in the system cordierite–garnet–biotite. In: Saxena, S. K. (ed.) Kinetics and Equilibrium in Mineral Reactions. New York: Springer, pp. 199–240.

Powell, R. & Holland, T. J. B. (1988). An internally consistent thermodynamic dataset with uncertainties and correlations: 3. Applications to geobarometry, worked examples and a computer program. Journal of Metamorphic Geology 6, 173–204.[Web of Science]

Schmidt, M. W. (1992). Amphibole composition in tonalite as a function of pressure: an experimental calibration of the Al-in-hornblende barometer. Contributions to Mineralogy and Petrology 110, 304–310.[CrossRef][Web of Science]

Schubert, W., Lippolt, H. J. & Schwarz, W. (2001). Early to Middle Carboniferous hornblende ⁴⁰Ar/³⁹Ar ages of amphibolites and gabbros from the Bergsträsser Odenwald. Mineralogy and Petrology 72, 113–132.

Schüller, A. (1957). Zur Grundgebirgsgeologie des Kyffhäusers. Geologica 11, 107–114.

Schwab, M. & Jacob, G. (1996). Wippraer Zone–Taunus–Hunsrück. Ein struktureller Vergleich. Geologie 12, 5–25.

Seim, R. (1960). Über Marmor-Kalksilikatfelse im Kyffhäuserkristallin. Geologie 9, 151–155.

Seim, R. (1967). Der Para-Biotit-Plagioklasgneis des Kyffhäuserkristallins, seine metatektischen Derivate und seine Einlagerungen. Geologie (Beiheft 56), 1–185.

Spear, F. S. & Kimball, K L. (1984). RECAMP—A Fortran IV program for estimating Fe³⁺ contents in amphiboles. Computers in Geology 10, 317–325.

Stacey, J. S. & Kramers, J. D. (1975). Approximation of terrestrial lead isotope evolution by a two-stage model. Earth and Planetary Science Letters 26, 207–221.[CrossRef][Web of Science]

Steiger, R. H. & Jäger, E. (1977). Subcommission on geochronology: convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Science Letters 36, 359–362.[CrossRef][Web of Science]

Tait, J. A., Bachtadse, V., Franke, W. & Soffel, H. C. (1997). Geodynamic evolution of the European Variscan fold belt: paleomagnetic and geological constraints. Geologische Rundschau 86, 585–598.

Theye, T. & Seidel, H. (1993). Chloritoid, carpholite and sudoite in very low-grade metamorphic rocks of the Wippra Metamorphic Zone (Harz, Germany). Neues Jahrbuch für Mineralogie, 73–96.

Todt, W. A., Altenberger, U. & von Raumer, J. F. (1995). U–Pb data on zircons for the thermal peak of metamorphism in the Variscan Odenwald, Germany. Geologische Rundschau 84, 466–472.[CrossRef][Web of Science]

White, R. W., Powell, R. & Holland, T. J. B. (2001). Calculation of partial melting equilibria in the system Na₂O–CaO–K₂O–FeO–MgO–Al₂O₃–SiO₂–H₂O (NCKFMASH). Journal of Metamorphic Geology 19, 139–153.[CrossRef][Web of Science]

Will, T. M. & Schmädicke, E. (2003). Isobaric cooling and anti-clockwise P–T paths in the Variscan Odenwald Crystalline Complex. Journal of Metamorphic Geology 21, 469–480.[CrossRef][Web of Science]

Zeh, A. (1996). Die Druck–Temperatur–Deformations-Entwicklung des Ruhlaer Kristallins (Mitteldeutsche Kristallinzone). Geotektonische Forschungen 86, 1–214.

Zeh, A. & Katzung, G. (1994). Zur Geologie des Kyffhäuser-Kristallins. Neues Jahrbuch für Geologie und Paläontologie, 368–384.

Zeh, A., Cosca, M., Brätz, H., Tichomirowa, M. & Okrusch, M. (2000). Simultaneous horst–basin formation and magmatism during Late Variscan transtension: evidence from ⁴⁰Ar/³⁹Ar and ²⁰⁷Pb/²⁰⁶Pb geochronology in the Ruhla Crystalline Complex. International Journal of Earth Sciences 89, 52–71.[CrossRef][Web of Science]

Zeh, A., Brätz, H., Millar, I. L. & Williams, I. S. (2001). A combined zircon SHRIMP and Sm–Nd isotope study on high-grade paragneisses from the Mid-German Crystalline Rise: evidence for northern Gondwanan and Grenvillian provenance. Journal of the Geological Society, London 158, 983–994.[Abstract/Free Full Text]

Zeh, A., Williams, I. S., Brätz, H. & Millar, I. L. (2003). Different age response of zircon and monazite during the tectono-metamorphic evolution of a high grade paragneiss from the Ruhla Crystalline Complex, Central Germany. Contributions to Mineralogy and Petrology 145, 691–706.[CrossRef][Web of Science]

Zeh, A., Klemd, R., Buhlmann, S. & Barton J. M. (2004). Pro- and retrograde P–T evolution of granulites of the Beit Bridge Complex (Limpopo Belt, South Africa): constraints from quantitative phase diagrams and geotectonic implications. Journal of Metamorphic Geology 22, 79–95.[CrossRef][Web of Science]

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

This article has been cited by other articles:

				J. C. Schumacher Metamorphic Amphiboles: Composition and Coexistence Reviews in Mineralogy and Geochemistry, October 1, 2007; 67(1): 359 - 416. [Full Text] [PDF]

				U. Linnemann, A. Gerdes, K. Drost, and B. Buschmann The continuum between Cadomian 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, northeastern Bohemian Massif, Germany) Geological Society of America Special Papers, January 1, 2007; 423(0): 61 - 96. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 46/7/1393    most recent egi020v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (3) Request Permissions Google Scholar Articles by ZEH, A. Articles by MILLAR, I. L. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Online ISSN 1460-2415 - Print ISSN 0022-3530

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

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