High-pressure rocks in imbricates; Jurassic-Cretaceous subduction

The multiplicity of local terms for the intermediate, imbricate units reflects some variability in lithological content, which in turn may reflect different crustal fragments. The extensive terminology is, in this review, simplified to the most common names reported in figure 5. Two main types of lithological subunits are distinguished, from bottom to top:

Figure 5. Location of the principal unit names found in the literature and used in this review.

Location of the principal unit names found in the literature and used in this review.

Same shade colours and symbols as Figure 4.


- 1) Eclogite-metabasic-gneiss sequence: The imbricates exposing this sequence include the Kerdylion, Sideronero, Kimi formations, in Greece, which find their equivalence in the Mesta, Madan (also Arda2), and Kroumovitsa formations in Bulgaria (Fig. 5). Paragneiss, minor graphitic marbles, and subordinate micaschists screen bodies of metamorphosed gabbros, diorites and granitoids. These rocks contain metamorphosed ultramafic and mafic bodies that locally preserved eclogite-facies parageneses [Kozhoukharova, 1984a; 1984b; Kolceva et al., 1986; Kolčeva and Eskenazy, 1988; Liati and Mposkos, 1990; Sapountzis et al., 1990]. High-pressure mineral assemblages are also preserved in pelitic rocks [Guiraud et al., 1992]. Granulite-facies parageneses formed during retrogression from eclogite to amphibolite-facies have been described in places [Kolceva et al., 1986; Liati and Seidel, 1996; Liati et al., 2002]. The scattered high-pressure rocks were overprinted by regional, amphibolite-facies metamorphism [680-560°C at 0.6-0.3 GPa, e.g. Georgieva et al., 2002] while the strongly mylonitic fabric of the country gneiss tended to be reset by partial melting. Ages of clastic zircons in paragneiss and marbles of this unit, to the north of Xanthi, attest for sedimentation younger than 300-280 Ma and a likely Gondwana source [Liati et al., 2011].

- 2) Gneiss-marble sequences: Structurally higher, lower grade units (Asenitsa and Borovitsa in Bulgaria, Fig. 2 and 5) contain thin marble sequences interlayered with para- and orthogneiss and minor amphibolites. The Asenitsa sequence is overlain by massive and coarse-grained marbles [Ivanov et al., 1984]. No convincing high-pressure and ultra-high-pressure relict has been documented in this essentially metasedimentary unit.

High-pressure metamorphic rocks

Eclogites have recorded various metamorphic histories according to their location and retrogression paths. Highest metamorphic pressures are ca 2 GPa at 700-800°C both in Bulgaria [Kolceva et al., 1986; Kolčeva and Eskenazy, 1988; Janák et al., 2011] and Greece [Liati and Seidel, 1996]. There are typically as many publications as there are outcrops because each rock has its own petrological specificity, including evidence for early, ultrahigh metamorphic pressures (Fig. 6). However, the regional information can be simplified. All high-pressure parageneses and retrogression paths generally document isothermal decompression to about 1 GPa followed by nearly isobaric cooling to the regional amphibolite-facies [0.8-1.1 GPa, 580-750°C, Mposkos, 1989; Machev and Kolcheva, 2008]. While some eclogites went through high-pressure granulite-facies [800°C - 1.5 GPa, Liati and Seidel, 1996; 700ºC at 1.26 GPa, Carrigan et al., 2002] others, as in Eastern Rhodope, went through blueschist-facies metamorphism [Tzontcheff-Bonev, 1992].

Figure 6. Location and references of the ultra-high to high pressure parageneses reported in the Rhodope Metamorphic Complex.

Location and references of the ultra-high to high pressure parageneses reported in the Rhodope Metamorphic Complex.

Same shade colours and symbols as Figure 4.


Ultrahigh pressure conditions inferred from quartz exsolution lamellae in clinopyroxene [Liati et al., 2002] typically reflect crystallization of magmatic pyroxene in the mantle and cannot be extended to the whole region. Indeed, these rocks crop out in the Kimi area in association with ultramafic rocks that are mainly mantle lherzolites and peridotitic cumulates with garnet and clinopyroxene [Mposkos, 2001]. They could be related to the 160 Ma arc. Ultrahigh pressure metamorphism is more convincingly documented by coesite in kyanite-eclogites [Zidarov et al., 1995]. Microdiamond inclusions in garnet from paragneisses [Mposkos and Kostopoulos, 2001; Perraki et al., 2006; Schmidt et al., 2010] indicate that some of these rocks recrystallized within the microdiamond stability field, which is however very sensitive to fluid compositions [e.g. Simakov et al., 2008].

Ages

Many ages have been produced, often without clear description of the tectonic and structural context of the sampled rocks and often with disputable relationships between ages obtained from mineral domains and regional geology / metamorphic history.

Protoliths

Protolith ages define two main groups: Palaeozoic and Jurassic-Cretaceous (Table 1).

The older group is Carboniferous to Permian, from ca. 300 to ca. 250 Ma. Zircon cores of two eclogitic gabbros are 245.6 ± 3.9 [Liati, 2005] and 255.8 ± 2.1 Ma [Liati et al., 2011]. Cores of monazites between 265 and 295 Ma [Bosse et al., 2009] may also witness this magmatic event.

Two gabbros of the Kroumovitsa unit are an exception showing protolith ages >500 Ma and metamorphic rims of 350-300 Ma [Carrigan et al., 2003]. They fall in a time bracket identified also in the Upper Terrane [e.g. Himmerkus et al., 2006], which raises the question as to whether this imbricate is a "Serbo-Macedonian" thrust sheet.

Concordant zircons from orthogneisses indicate Late Jurassic-Early Cretaceous intrusions from ca. 160 to ca. 130 Ma (Table 1). Forty zircon grains dated between 121 and 159 Ma [Bosse et al., 2009] may represent this magmatic event. The 117 Ma oscillatory zircon domain of a garnet-mafic rock (Table 1) may also reflect the protolith age [Liati et al., 2011].

Table 1. Geochronological data: Protolith ages from Intermediate Units

Rock type (location, *=Bulgaria) Age (Ma) Method Reference
    U-Pb zircon  
Gabbro (Bubino*) 572 ± 5   [Carrigan et al., 2003]
Metaplagiogranite (S-Kesebir*) 511 ± 5 mean age [Bonev et al., 2010a]
Amphibolite (S-Kesebir*) 459-434 core [Bonev et al., 2010a]
Metagabbro (S-Kesebir*) 474 ± 6 core [Bonev et al., 2010a]
Orthogneiss (Sidironero) 294 ± 8   [Liati & Gebauer, 1999]
Migmatitic orthogneiss 294.3 ± 2.4   [Liati, 2005]
Migmatitic orthogneiss (Thermes) 291.4 ± 3.4 inherited core [Turpaud & Reischmann, 2010]
Augengneiss (Siroko) 275.8 ± 3.9   [Turpaud & Reischmann, 2010]
Garnet-gneiss (Kimi) 290-247 core [Liati et al., 2011]
Eclogite (NE Komotini) 255.8 ± 2.1   [Liati et al., 2011]
Metagabbro (Drama-Sideronero) 245.6 ± 3.9   [Liati, 2005]
Biotite-gneiss (Sminthi) 164.4 ± 7.1 Pb-Pb Evaporation [Turpaud & Reischmann, 2010]
Biotite-gneiss (N-Drama) 163.4 ± 2.1   [Turpaud & Reischmann, 2010]
Metadiorite (Thermes) 158.7 ± 1.7 Pb-Pb Evaporation [Turpaud & Reischmann, 2010]
Orthogneiss (Bachkovo) 153.5 ± 4.1   [Von Quadt et al., 2006]
Orthogneiss (Zlatograd*) 151.9 ± 2.2 Concordant zircons [Ovtcharova et al., 2004]
Orthogneiss (Kimi) 151.5 ± 2.0 Oscillatory domain [Liati et al., 2011]
Orthogneiss (General Geshevo*) 149.0 ± 0.66 Concordant zircons [Ovtcharova et al., 2004]
Orthogneiss (Thermes) 148.7 ± 5.6 Pb-Pb Evaporation [Turpaud & Reischmann, 2010]
Biotite-gneiss (Echinos) 137.8 ± 5.1 Pb-Pb Evaporation [Turpaud & Reischmann, 2010]
Biotite-gneiss (Paranesti) 136.5 ± 4.3 Pb-Pb Evaporation [Turpaud & Reischmann, 2010]
Orthogneiss (Paranesti) 134.0 ± 3.5 Pb-Pb Evaporation [Turpaud & Reischmann, 2010]
40 zircons (W-Xanthi) 159 to 121   [Bosse et al., 2009]
    Rb-Sr  
Orthogneiss (Kechros) 334 ± 5 Muscovite [Mposkos & Wawrzenitz, 1995]

High-pressure metamorphism

Ages obtained in western, central and eastern Rhodope are similar, which fits a first-order process that goes beyond regional variations. The UHP metamorphism had probably started during the Early Jurassic [older than 170 Ma, Reischmann and Kostopoulos, 2002; Bauer et al., 2007]. The Sm-Nd method applied to a 1.5-1.6 GPa pyroxenite from Eastern Rhodope yielded 119 ± 3.5 Ma, interpreted as the age of the HP metamorphism [Wawrzenitz and Mposkos, 1997]. U-Pb SHRIMP analyses on oscillatory zoned zircon of a garnet-bearing basic rock from Central Rhodope yielded an equivalent, weighted mean age (117.4 ± 1.9 Ma), which was considered as dating the protolith [Liati et al., 2002] but may very well date early metamorphic conditions (no clear core). Metamorphic, zircon rims from a garnet-kyanite paragneiss for which high-pressure conditions have been suggested are dated at 148.8 ± 2.2 Ma, complemented by ages of 147.2 ± 4.7 Ma (paragneiss) and 143.4 ± 3.3 Ma (strongly amphibolitized eclogite) zircon domain ages [Liati, 2005].

Amphibolite-facies recrystallization

Further constraints are provided, in Central Rhodope, by oscillatory zircon domains at 73.5 ± 3.4 Ma interpreted to reflect HP/UHP metamorphism [Liati, 2005] but possibly dating regional, amphibolite-facies metamorphism. This weighted mean age for metamorphism is further supported by the 61.9 ± 1.9 Ma pegmatite vein that intruded these rocks [Liati et al., 2002] and the 65-63 Ma trondhjemitic veins cutting amphibolitized eclogites in eastern Rhodope [Baziotis et al., 2007] where an additional Rb-Sr isochron age of 65.4 ± 0.7 Ma is given for another cross-cutting pegmatite [Liati et al., 2002]. In eastern Rhodope, two syn- to post-deformation granites were dated at ca. 70 Ma [Marchev et al., 2006]. In western Rhodope, zircon rims of a Palaeozoic gabbro are 51.0 ± 1.0 Ma [Liati, 2005]. These dated veins provide solid evidence that amphibolite-facies deformation was waning by ca. 50 Ma. The number of ages distributed with no obvious hiatus between this upper bound and ca 170 Ma (Tables 2 and 3) reflects a protracted residence under evolving eclogite-, granulite- and amphibolite-facies conditions.

Table 2. Geochronological data for (ultra-) high-pressure metamorphism in the Rhodope massif.

Rock type (location) Age (Ma) Method Reference
    U-Pb zircon  
Metapelite (Kimi) 171 ± 1   [Bauer et al., 2007]
Eclogite (Kimi) > ca 160   [Bauer et al., 2007]
Paragneiss (Siroko) 148.8 ± 2.2   [Liati, 2005]
Paragneiss (Siroko) 147.2 ± 4.7 domain [Liati, 2005]
Eclogite (Siroko) 143.4 ± 3.3 domain [Liati, 2005]
Garnet-gneiss (Kimi) 153-139 inner rim [Liati et al., 2011]
Metapelite (Chepelare*) 142-137 monazite cores [Bosse et al., 2010]
Garnet-amphibolite (Kimi) 117.4 ± 1.9 domain [Liati et al., 2002]
Metapelite (Kimi) 160 ± 1 HT overprint [Bauer et al., 2007]
Metapelite (W-Xanthi) 148 to 121 granulitic overprint? [Krenn et al., 2010]
Eclogite (Kimi) ca 115 HT overprint [Bauer et al., 2007]
    Sm-Nd  
Amphibolite (Volvi) 153±13   Kostopoulos, pers.com, 2010
Paragneiss (NW-Xanthi) 140 ± 4   [Reischmann & Kostopoulos, 2002]
Garnet-pyroxenite (Kimi) 119 ± 3.5   [Wawrzenitz & Mposkos, 1997]
    40Ar/39Ar  
Mylonite (Chalkidiki) 142.98 ± 4.89 white mica [Lips et al., 2000]

Table 3. Metamorphic overprint in high-pressure rocks and their country rocks and ages of post deformational granites and pegmatite veins.

Rock type (location, *= Bulgaria) Age (Ma) Method Reference
    U-Pb zircon  
Paragneiss (Siroko) 82.8 ± 1.3   [Liati, 2005]
Pegmatites (Chepelare) around 77   [Bosse et al., 2009]
Eclogite (Kimi) 79 ± 3   [Bauer et al., 2007]
Garnet-gneiss (Kimi) 73.9 ± 0.8   [Liati et al., 2011]
Garnet-amphibolite (Kimi) 73.5 ± 3.4   [Liati et al., 2002]
Pyroxenite (Kimi) 72.9 ± 1.1   [Liati et al., 2011]
Orthogneiss (Kimi) 71.4 ± 1.1   [Liati et al., 2011]
Orthogneiss (Bachkovo*) 55.9 ± 7.2   [Von Quadt et al., 2006]
Garnet-amphibolite (Sideronero) 51.0 ± 1.0   [Liati, 2005]
Granite (Chuchuliga) 68.94 ± 0.4   [Marchev et al., 2006]
Granite (Rozino) 68 ± 15   [Marchev et al., 2006]
Discordant pegmatite (Kimi) 61.9 ± 1.9   [Liati et al., 2002]
    Rb-Sr  
Undeformed pegmatite (Kimi) 65.4 ± 0.7 White mica [Mposkos & Wawrzenitz, 1995]

Tertiary recrystallization

42 Ma zircons in a retrogressed eclogite were interpreted as dating the eclogite-facies event [Liati and Gebauer, 1999] while zircon rims between ca. 40 to ca. 38 Ma in migmatites containing the eclogites and in an adjacent amphibolite have been interpreted as dating the regional, amphibolite-facies metamorphism [Liati, 2005]. These ages (Table 4) are within the range of the numerous K-Ar and 40Ar/39Ar amphibole and mica ages between 50 and 35 Ma reported for many rocks of the intermediate units [Liati and Kreuzer, 1990; Kaiser-Rohrmeier et al., 2004] and with the Eocene age of zircons from discordant leucosomes [Ovtcharova et al., 2002; Peytcheva et al., 2004]. They are also coincident with the many cooling ages measured within this time span all over the Rhodope Metamorphic Complex (Tables 4, 6 and 7) and the ca. 35 Ma age of hydrothermal deposits and volcanic rocks in Eastern Rhodope [Márton et al., 2010]. Since K-Ar and 40Ar/39Ar are reportedly low temperature systems, one cannot exclude that the 42-38 zircons also registered retrogression down to hydrothermally influenced metamorphic conditions of ca. 300°C - 0.3 GPa [Liati and Seidel, 1996]. As such, the many Tertiary ages may simply record cooling from amphibolite to greenschist-facies. Zircon and apatite fission-track ages overlapping K-Ar and 40Ar/39Ar dates (Table 4) demonstrate very fast cooling in Oligocene times [Wüthrich, 2009; Márton et al., 2010].

Table 4. Geochronological data for Tertiary thermal event and cooling in the Intermediate units of the Rhodope massif.

Rock type (location, *=Bulgaria) Age (Ma) Method Reference
    U-Pb zircon  
Synfolial pegmatite (W-Xanthi) 49.6 ± 3.9   [Bosse et al., 2009]
Metagabbro (S-Kesebir*) 49.1 ± 6 rim [Bonev et al., 2010a]
Synfolial pegmatite (W-Xanthi) 48.2 ± 2.2   [Bosse et al., 2009]
Quartz vein (Sminthi) 45.3 ± 0.9   [Liati & Gebauer, 1999]
Eclogite (Thermes) 42.2 ± 0.9   [Liati & Gebauer, 1999]
Migmatitic orthogneiss (Thermes) 42.1 ± 1.0   [Liati & Gebauer, 1999]
Orthogneiss (Thermes) 42.0 ± 1.1   [Liati & Gebauer, 1999]
Leucosome (Thermes) 40.0 ± 1.1   [Liati & Gebauer, 1999]
Leucosome (Sideronero) 39.7 ± 1.2   [Liati, 2005]
Leucosome (Sideronero) 38.1 ± 0.8   [Liati, 2005]
Garnet-amphibolite (Sideronero) 38.1 ± 1.2   [Liati, 2005]
Leucosome 37.08 ± 0.38   [Ovtcharova et al., 2002]
Pegmatite (Sminthi) 36.1 ± 1.2   [Liati & Gebauer, 1999]
    Monazite  
Synfolial pegmatite (W-Xanthi) 54.9 ± 1.7 core [Bosse et al., 2009]
Orthogneiss (Zlatograd*) 47.4 ± 0.66   [Ovtcharova et al., 2004]
Pegmatite (Tchepelare*) 42.1 ± 1.2   [Bosse et al., 2009]
Metapelite (Chepelare*) 42-38 rims [Bosse et al., 2010]
Synfolial pegmatite (W-Xanthi) 38.6 ± 1.1 rim [Bosse et al., 2009]
Leucosome 37.8 ± 1.5   [Ovtcharova et al., 2002]
Orthogneiss (Banite-Gulubovo*) 35.83 ± 0.4   [Peytcheva et al., 2004]
    40Ar/39Ar  
Eclogite (Belopolci*) 45 ± 2 amphibole [Mukasa et al., 2003]
Orthogneiss (Kazak*) 39 ± 1 muscovite [Mukasa et al., 2003]
Gneiss (Pelevun*) 39.28 ± 0.24 muscovite [Márton et al., 2010]
Gneiss (Kremenitz*) 37.28 ± 0.19 muscovite [Márton et al., 2010]
Amphibolite (Ada Tepe*) 36.9 ± 0.16 muscovite [Márton et al., 2010]
Adularia (Kuklitza*) 35.94 ± 0.36 adularia [Márton et al., 2010]
Amphibolite (Ada Tepe*) 36.9 ± 0.16 muscovite [Márton et al., 2010]
Orthogneiss (NW-Pilima) 35.4 ± 0.4 biotite [Moriceau, 2000]
Orthogneiss (Banite-Gulubovo*) 35.35 ± 0.22 biotite [Peytcheva et al., 2004]
Orthogneiss (Pilima) 35.3 ± 0.4 muscovite [Moriceau, 2000]
Gneiss (Davidkovo*) 35.25 ± 0.36 muscovite [Kaiser-Rohrmeier et al., 2004]
Orthogneiss (NW-Pilima) 35.0 ± 0.4 muscovite [Moriceau, 2000]
Adularia (Ada Tepe*) 34.95 ± 0.36 adularia [Márton et al., 2010]
Pegmatite (Tchepelare*) 34.9 ± 0.1 muscovite [Bosse et al., 2009]
Volcanite (Iran Tepe*) 34.62 ± 0.46 amphibole [Márton et al., 2010]
Synfolial pegmatite (W-Xanthi) 34.3 ± 0.2 biotite [Bosse et al., 2009]
Synfolial pegmatite (W-Xanthi) 33.2 ± 0.3 muscovite [Bosse et al., 2009]
Gneiss (Imera) 32.0 ± 0.3 muscovite [Moriceau, 2000]
    Rb-Sr  
Paragneiss (NW-Xanthi) 37 muscovite [Reischmann & Kostopoulos, 2002]
Pegmatite (Banite-Gulubovo*) 35.31 ± 0.25   [Peytcheva et al., 2004]
Paragneiss (NW-Xanthi) 34 biotite [Reischmann & Kostopoulos, 2002]
    K/Ar  
Amphibolites (NW-Xanthi) 95-57 hornblende [Liati & Kreuzer, 1990]
Orthogneiss (E-Kardamos) 42.1 ± 1 muscovite [Krohe & Mposkos, 2002]
Orthogneiss (E-Kardamos) 39.4 ± 1 biotite [Krohe & Mposkos, 2002]
    Fission-track  
Gneisses (Central Rhodope*) ca. 35 several zircons [Wüthrich, 2009]
Migmatites (Starcevo*) 33.2 ± 3.8 apatite [Wüthrich, 2009]
Migmatites (Borovica*) 33.0 ± 4.4 apatite [Wüthrich, 2009]
Gneisses (Central Rhodope*) 35-20 several apatites [Wüthrich, 2009]
Amphibolite (Pelevun*) 25.0 ± 1.5 apatite [Márton et al., 2010]
Gneiss (Kremenitz*) 18.3 ± 1.9 apatite [Márton et al., 2010]
Amphibolite (Ada Tepe*) 14.8 ± 5.1 apatite [Márton et al., 2010]

Protolith compositions

Geochemical analyses document a variety of protoliths. Trace-element and REE ratios of eclogites and associated ultrabasic and basic rocks [Kozhoukharova, 1984a; 1984b; Kolčeva and Eskenazy, 1988; Liati et al., 1990] were suggested to have been derived from mid-ocean ridge basalts with a tholeiitic trend of differentiation and a close relationship to ocean-type lithosphere (dominance of harzburgites and wehrlites with orthopyroxenite veins and dunite). However, the given analyses are mostly T-MORBs (Kostopoulos, pers. comm, 2010). Therefore, they may represent the oceanic floor of attenuated lithosphere, such as a marginal basin that would have retained some continental crust [Barr et al., 1999]. Supporting this hypothesis, trace-element geochemistry and igneous zircon U-Pb ages (SHRIMP II) for some eclogite/amphibolite strata also suggest intrusion of basaltic sills and dykes, of T-MORB affinity, into attenuated continental basement during the Early Triassic [Liati, 2005].

In contrast, amphibole-bearing and biotite orthogneisses are evolved volcanic-arc intrusions [Cherneva and Georgieva, 2005] while some amphibolites derived from basalt to basaltic andesites also define volcanic arc affinity [Liati and Seidel, 1996].

The anatectic para- and orthogneiss, marbles and eclogitic to amphibolitic metabasites of the Kerdylion unit (the lowest Serbo-Macedonian, in Greece, Fig. 5) are comparable to those of the Mesta and Sideronero imbricates.

These protolith compositions demonstrate that the Rhodope metamorphic pile has involved an active margin environment before and during synmetamorphic thrusting. The question is whether subduction occurred below an island arc or a continental margin. This question has importance for any geodynamic reconstruction and will be discussed in the relevant paragraph of this review.