ABSTRACT

The Paleozoic Middle East terranes, neighboring the present-day Arabian and Levant plates, are shown by most authors to consist of ten major tectonic units: (1 and 2) the Helmand and Farah terranes of Afghanistan, southwest Pakistan and southeast Turkmenistan; (3 to 6) the Alborz, Central Iran (Lut, Yazd and Tabas) and Sanandaj-Sirjan terranes of Iran, and Northwest Iran (possibly extending into eastern Turkey); (7 and 8) the Pontides and Taurides terranes of Turkey; and (9 and 10) the Greater and Lesser Caucasus terranes between the Caspian and Black seas (Armenia, Azerbaijan, Georgia and southwest Russia). Published plate-tectonic reconstructions indicate that all ten terranes may have broken off from either: (1) the Gondwana Supercontinent in the mid-Silurian as part of the Hun Superterrane; or (2) the Pangea Supercontinent during the mid-Permian - Triassic as part of the Cimmeria Superterrane. To the north of Gondwana and Pangea, three successively younger Tethyan oceans evolved: (1) Proto-Tethys (Cambrian - Devonian); (2) Paleo-Tethys (mid-Silurian - Mesozoic); and (3) Neo-Tethys (mid-Permian - Cenozoic).

Two regional Paleozoic unconformities in the Arabian Plate are generally linked to major regional-scale structural events, and commonly correlated to the Caledonian and Hercynian orogenies. These orogenies took place many thousands of kilometers away from the Arabian Plate and are considered unlikely causes for these unconformities. Instead, the breakaway of the Hun and Cimmeria superterranes are considered as alternative near-field tectonic sources. The older unconformity (middle Paleozoic event), represented by a mid-Silurian to Middle Devonian hiatus in North Arabia (Iraq and Syria), reflects an episode of epeirogenic uplift, which might be related to the mid-Silurian rift of the Hun Superterrane. The younger mid-Carboniferous Arabia-wide angular unconformity involved compressional faulting and epeirogenic uplift, and might be related to the earliest phase of subduction by the Paleo-Tethyan crust beneath Cimmeria (Sanandaj-Sirjan and nearby regions) before it broke off. Based on our review and regional considerations, we assign the Helmand, Farah, Central Iran, Alborz, Sanandaj-Sirjan, Northwest Iran, Lesser Caucasus, Taurides and Pontides to Cimmeria, whereas the Greater Caucasus is considered Hunic.

INTRODUCTION

During the past decade, our general knowledge on the geochronological evolution, paleopositions, paleotectonic compositions and paleogeographic outlines of the Phanerozoic supercontinents has significantly improved (e.g. Dalziel, 1997; Stampfli et al., 2001, 2002; Lindsay, 2002; Cocks and Torsvik, 2002; Lawver et al., 2002; Stampfli and Borel, 2002; Veevers, 2003; Pesonen et al., 2003; von Raumer et al., 2002, 2003; Golonka, 2004; Scotese, 2004; Torsvik and Cocks, 2004). Yet today, many uncertainties persist in reconstructing the geological evolution of the regions adjoining the supercontinents, especially for the Paleozoic Era (Cocks and Torsvik, 2002; Torsvik and Cocks, 2004; Scotese, 2004). These regions are in themselves vast, and consist of numerous plate-tectonic units that are sometimes referred to as microplates, terranes, blocks, structural domains, and sometimes grouped into superterranes. The Middle East is a typical example of a border-region that consists of a complex mosaic of tectonic units (Figures 1 and 2).

We identify the Middle East terranes, bordering the present-day Arabian and Levant plates, in Afghanistan, Iran, western Pakistan, Turkey, southeast Turkmenistan and the Caucasus (Armenia, Azerbaijan, Georgia and southwest Russia) (Figures 1 and 2). Several of these terranes are strongly deformed and stacked within a wide tectonic belt between the Eurasian, Arabian and Indian plates. Although their boundaries are generally traced along well-preserved or/and reactivated Paleozoic fault systems, in some cases the borders remain unclear. Correlation of the sedimentary core complexes, however, suggests that all of these terranes share a common ancestry during some time in the Paleozoic Era.

The Middle East terranes were affected by the evolution of the Paleozoic Tethyan oceans, the Hun (Hunic or Intermediate) and Cimmeria (Cimmerian) superterranes, and the Gondwana and Pangea supercontinents (Figures 3 to 11; e.g. Sengör, 1990; Stampfli, 1996; von Raumer, 1998; Cocks and Torsvik, 2002; von Raumer et al., 2002, 2003; Stampfli and Borel, 2002; Stampfli et al., 2001, 2002; Torsvik and Cocks, 2004; Natal’in and Sengör, 2005; Xypolias et al., 2006). At least three major Paleozoic rift episodes occurred along the margins of Gondwana and Pangea (Figures 4 to 11). The first was in the Early Ordovician when Avalonia broke off from Gondwana. This episode was unlikely to have influenced the Middle East region, which was located about 6,000 km away (Figure 4). The second involved the mid-Silurian breakaway of the Hun Superterrane (Figures 6 and 7), the detailed reconstruction of which is unresolved due to insufficient paleomagnetic and paleontological data. This episode is relevant to our review because parts of the superterrane may have involved the Middle East terranes. The third episode is the mid-Permian - Triassic breakaway of several Middle East Cimmerian terranes from Gondwana, by then a part of Pangea (Figures 10 and 11).

In most Paleozoic reconstructions, the Middle East region is interpreted as part of the passive margin of Gondwana and Pangea until the mid-Permian - Triassic, when Cimmeria started to rift away, causing the opening of the Neo-Tethys Ocean (e.g. Sharland et al., 2001; Stampfli et al., 2001). In addition, two regional unconformities are recognized. The first corresponds to a mid-Silurian to Middle Devonian hiatus – “middle Paleozoic hiatus” – that is sometimes correlated with the Caledonian Orogeny (e.g. Buday, 1980) (Figures 3, 5 to 7). The second unconformity represents a “mid-Carboniferous hiatus”, and is often correlated to the Hercynian Orogeny (e.g. Berberian and King, 1981) (Figures 3 and 9). These correlations do not provide a satisfactory plate-tectonic model for the Paleozoic evolution of the Middle East region because it was located far away from these two orogenies.

Our paper starts with a brief global review of the largest and relatively well-constrained Paleozoic plate-tectonic units and seaways. We adopt the conventions of the ICS (International Commission on Stratigraphy; Gradstein et al., 2004) by not capitalizing informal qualifiers “late, middle, mid, early, etc.” except where defined (Ordovician and Devonian). After setting the global stage, we discuss the smaller and less constrained units of the Middle East, many of which have unfamiliar names and interpretations. Most of the illustrated global reconstructions follow Cocks and Torsvik (2002) and Torsvik and Cocks (2004), as the involved units are adequately represented. We have also considered the global reconstructions by von Raumer (1998), Stampfli and Borel (2002), von Raumer et al. (2002, 2003), Stampfli et al. (2001, 2002), Scotese (2004), Natal’in and Sengör (2005) and others. Our principal objective is to present the Paleozoic plate-tectonic framework and nomenclature for the Middle East, which can form a basis for further refinements.

GLOBAL PALEOZOIC PLATE-TECTONIC SETTING

Gondwana, Laurussia and Pangea Supercontinents

The global plate-tectonic configuration during the Paleozoic was dominated by three supercontinents: Gondwana, Laurussia and Pangea. Gondwana, the largest supercontinent on Earth from the Late Cambrian to mid-Carboniferous (Figures 2 to 8), consisted of several present-day plates including Arabia, Africa, most of Antarctica and Australia, India, Madagascar and most of South America, with numerous small terranes attached to their margins (Courjault-Radé et al., 1992; Cocks, 2001; Stampfli et al., 2001; Cocks and Torsvik, 2002; Stampfli and Borel, 2002; von Raumer et al., 2002, 2003; Fortey and Cocks, 2003; Scotese, 2004; Avigad et al., 2005).

During the Paleozoic Era, Laurussia was assembled out of three large plates (Avalonia, Baltica and Laurentia, Figures 3 and 5) and several island arcs in a series of orogenic phases (McKerrow et al., 2000; Stampfli and Borel, 2002; Scotese, 2004) (Figures 3 to 7). The collisional assembly of Laurussia caused the Caledonian Orogeny, an event that was redefined by McKerrow et al. (2000) to apply to the closure of the Iapetus Sea (Figures 3, 5 to 7).

Gondwana and Laurussia remained separated by the Rheic Ocean until the mid-Carboniferous (c. 325–310 Ma) when they collided during the Hercynian Orogeny to form the Pangea Supercontinent (Stampfli and Borel, 2002; Scotese, 2004; Figures 3, 8 and 9, Torsvik and Cocks, 2004). In the late Carboniferous (Pennsylvanian) and Permian, Pangea was enlarged with the amalgamation of the Kazakh, Siberia, Kara and other terranes (Figure 10). The combination of Laurussia with these terranes would later in the Mesozoic form the Laurasia Supercontinent, a term that is easily confused with Laurussia.

Three Tethyan Oceans

The names of the Paleozoic oceans that separated the supercontinents are not unique and vary to reflect somewhat different interpretations. The seaway that opened to the north of the Paleozoic Middle East terranes is called the Paleo-Tethys Ocean by some authors (e.g. Sharland et al., 2001; Bykadorov et al., 2003). Others refer to it as the Proto-Tethys (or Asiatic Ocean) and reserve the term “Paleo-Tethys” for the ocean that opened in the mid-Silurian along the trailing edge of the Hun Superterrane (Figures 3, 7 and 8; e.g. Ziegler et al., 2001; Stampfli et al., 2001; Stampfli and Borel, 2002; von Raumer et al., 2002, 2003). Most authors adopt the Neo-Tethys for the ocean that opened in mid-Permian -Triassic along the African-Arabian margin of Gondwana (e.g. Stampfli and Borel 2002; also Meso-Tethys of Metcalfe, 1999; Pindos Ocean of Golonka, 2004; Figure 10). Following Stampfli and Borel (2002), we adopt Proto-Tethys for the early Paleozoic ocean to distinguish it from the late Paleozoic Paleo-Tethys and Neo-Tethys oceans (Figures 310).

The interpretation of the lateral extent and evolution of the Tethyan oceans can vary. Hünecke (2006), for example, argued that the Middle-Late Devonian ocean between Gondwana and Laurussia was not as large as depicted by Stampfli and Borel (2002) and Torsvik and Cocks (2004). Stampfli and Borel (2002) and von Raumer et al. (2002, 2003) interpreted that in the Late Ordovician-early Silurian, the eastern branch of the Proto-Tethys Ocean might have closed when Serindia terranes (North China and Tarim) amalgamated with Gondwana. Several authors interpreted the initial opening of the Neo-Tethys Ocean in early rather than mid-Permian (Vannay, 1993; Garzanti and Sciunnach, 1997; Garzanti et al., 1994, 1996a, b, 1999; Stampfli and Borel, 2002; Angiolini et al., 2003), or to have started north of Australia in the Carboniferous and extended diachronously westwards into the Permian (Stampfli, 2000).

Iapetus Sea and Rheic Ocean

Two more Paleozoic seaways are significant for our review (Figures 3 to 7). The Iapetus Sea, which separated Laurentia, Avalonia and Baltica in the early Paleozoic, closed in the late Silurian when these terranes joined to form Laurussia (Figures 3 to 5). The Rheic Ocean (also Rheic-Mauritania, Rhenohercynian or Hercynian-Rheic) opened in the Cambrian along Avalonia’s northerly trailing edge (Figures 3 to 7). During the Devonian-early Carboniferous (c. 420–320 Ma), Gondwana drifted towards Laurussia, closing the Rheic Ocean (Figures 7 and 8, Torsvik and Cocks, 2004). In the mid-Carboniferous, the Hercynian Orogeny occurred along a front between northwest Africa and southeast North America and closed the Rheic Ocean (Figure 9; incorrectly referred to as the Iapetus Sea in Al-Husseini, 2004).

Hun Superterrane

Plate reconstruction of the mid-Silurian to mid-Permian northern margin of Gondwana fall into two general models (Figures 6 and 7). Whereas both show the breakaway of various terranes from Gondwana, they differ in detail and lateral extent – particularly near the Middle East region.

In the first model, following Stampfli et al. (2001) and Stampfli and Borel (2002), the ribbon-like Hun Superterrane extended from westernmost Iberia (in Spain) to Qiantang (Figure 6). This superterrane is also referred to as the Hun Composite Superterrane because it is divided into (Figure 6): (1) the northern Hun Cordillera terranes (also European Hunic terranes); and (2) the southern Hun Gondwana terranes (also Asiatic Hunic terranes) (von Raumer, 1998; Stampfli et al., 2001; von Raumer et al., 2002, 2003; Stampfli and Borel, 2002; Schulz et al., 2004). This division reflects the separate evolution of the two sets of terranes after they docked along Laurussia in the late Paleozoic. The Hun Superterrane rifted away from Gondwana in the mid-Silurian (possibly in different phases that lasted into the Devonian) and then drifted towards Laurussia, with which it collided in the Devonian-Carboniferous (Figures 3, 6 to 8). Interpretations of late Silurian paleocurrents indicate that the Panthalassic (north of the Proto-Tethys) waters did not mix with those of the Paleo-Tethys (Johnson et al., 2004), thus suggesting that the Hun Superterrane remained consolidated until at least the Early Devonian.

From west to east the Hun Cordillera terranes included: Ossa-Morena, Channel, Saxo-Thuringian, Moesia, Istanbul, Pontides, Ligerian, Moldanubian, Moravo-Silesicum, Helvetic, South Alpine, Penninic, Austro-Alpine, Carpathian and North Tarim (Figure 6; Stampfli et al., 2001). The Hun Gondwana terranes included: Iberia, Armorica, Cantabria, Aquitaine, Alboran, Intra-Alpine terranes (Adria, Carnic, Austro-Carpathian), Dinaric-Hellenic, Karakum-Turan, Pamirs, South Tarim, Qiangtang, North and South China, and Annamia terranes (Stampfli et al., 2001; Stampfli and Borel, 2002; the latter two easternmost terranes are shown in Figure 1, but not in Figure 6). In the Devonian, the Kazakh terranes may also have been close to the Hunic Cordillera Superterranes (Stampfli and Borel, 2002). The remaining adjoined with Pangea were Apulia, Hellenides-Taurides, Menderes-Taurus, Sanandaj-Sirjan, Northwest and Central Iran, Helmand, North and South Tibet.

In the second model, several terranes broke off and had drifted some distance away from the northwestern margin of Gondwana by the Early Devonian (Figure 7; Torsvik and Cocks, 2004). They also formed a ribbon-like superterrane that vaguely resembles the western part of the Hun Superterrane (compare Figures 6 and 7). The breakaway group included: Rheno-Hercynian, Armorica (includes Iberia), Adria, Pontides, Hellenic and Moesia (Figures 2 and 6). The Rheno-Hercynian and Perunica are considered as separate terrane between the Hunic terranes and Baltica (in Laurussia) (Kríz et al., 2003; Torsvik and Cocks, 2004). Following the breakaway event, terranes adjacent to Gondwana were Apulia, Taurides, Sanandaj-Sirjan, Northwest and Central Iran, Helmand, South Tibet (Lhasa) and North Tibet (Qiangtang) terranes.

Some differences and confusion occur when comparing the two models in detail. The term Qiangtang (also Qangtang) is a synonym for North Tibet in Torsvik and Cocks (2004; Figure 2) and positioned next to South Tibet (Lhasa). In contrast, Stampfli et al. (2001) show Qiangtang as Hunic but North Tibet as Gondwanan (Figure 6). Other confusing terms are Karakum and Karakoram (also spelled as Karakorum). Karakum and neighboring Mangyshlak of Torsvik and Cocks (2004) are equivalent to the Karakum-Turan terrane (Figure 1). Karakum-Turan was not attached to Gondwana in the late Paleozoic (L. Angiolini, written communication, 2006) and probably Hunic (Figure 6). Karakoram is located in northern Pakistan (Gaetani, 1997; Figure 1), which belonged to Cimmeria (L. Angiolini, written communication, 2006). Further studies of the Cambrian-Ordovician rocks in Karakorum, based on the works of Gaetani et al. (1996), Gaetani (1997), Quintavalle et al. (2000) and Rolland et al. (2002), may provide new insights for its early Paleozoic paleoposition. Turan is often mentioned as a plate, but the Russian term ‘plate’ differs in meaning from ‘tectonic plate’, causing some further confusion (Laz’ko, 1975). It remains unclear whether Karakum and Turan formed one or several terranes.

Cimmeria Superterrane

In the mid-Permian - Triassic, Cimmeria started rifting away from Pangea and closing the Paleo-Tethys Ocean to the north, while opening the Neo-Tethys Ocean in its wake (Figure 10). Less clear is which terranes were Cimmerian, or Hunic or possibly neither. Torsvik and Cocks (2004, Figure 10) show Cimmeria to consist of Apulia, Taurides, Sanandaj-Sirjan, Northwest and Central Iran, Helmand and North Tibet (Qiantang). They place South Tibet (Lhasa) to the north of India, however noting that it is not constrained by paleomagnetic or faunal content. Stampfli et al. (2001) and Stampfli and Borel (2002) (Figures 1, 2 and 6) included in Cimmeria: Apulia, Hellenides-Taurides, Menderes-Taurus, Sanandaj-Sirjan, Northwest and Central Iran, Helmand, South and North Tibet. A comparison indicates that several Middle East terranes (Northwest and Central Iran, Taurides and Sanandaj-Sirjan) are considered Cimmerian by both groups of authors.

In contrast to the somewhat generalized Cimmeria of some authors (e.g. Sharland et al., 2001; Stampfli et al., 2001; Stampfli and Borel, 2002; Torsvik and Cocks, 2004), Sengör (1990, Figure 11) showed the Cimmeria breakaway event in substantial detail and to consist of three ribbons. He divided North Tibet into East and West Qiangtang and considered the former as the leading Cimmerian ribbon. The “Intermediate” ribbon consisted of East Pontides, Dzirula Massif, Artvin/Karabagh, Sanadaj-Sirjan, Northwest Iran (including Alborz), Central Iran (Yazd, Tabas, and Lut), Farah, central Pamirs (China) and West Qiangtang. The trailing third ribbon included Helmand and South Tibet. The latter two ribbons connected to Australia, and the Neo-Tethys Ocean consisted of several seaways. Significantly for our paper, Sengör’s model intepreted a subduction zone along the northeast front of Sanandaj-Sirjan and other northerly terranes, a subject that will be discussed later. The Intermediate ribbon of Natal’in and Sengör (2005) is generally comparable with the Cimmerian Superterrane.

PALEOZOIC OROGENIES AND THE ARABIAN PLATE

In most of the pre-Permian Paleozoic reconstructions (Figures 4 to 9), the Arabian Plate is generally depicted inland from the Tethyan margins of Gondwana, or later Pangea. Until the mid-Permian (Figures 10 and 11), it is shown as bounded by the Middle East terranes and, for the most part, at latitudes of about 30° to 60° south. Two regional hiatuses that were associated with polar glaciations occurred in the Late Ordovician Hirnantian Stage (Figures 3 and 5; Vaslet, 1990; Abed et al., 1993) and in the late Carboniferous - early Permian (Figures 3 and 9; Osterloff et al., 2004).

Two structurally significant unconformities have been recognized in the Arabian Plate (Figure 3). The mid-Silurian (Wenlock) to Middle Devonian hiatus is regionally manifested in Syria and Iraq, and possibly other parts of the Middle East (Brew and Barazangi, 2001; Al-Hadidy, 2007). Because of its age, it was correlated to the Caledonian Orogeny by some authors (e.g. Buday, 1980). In southeast Arabia (Oman), distinct hiatuses occur in the mid-Silurian (Wenlock) to earliest Devonian, and in the Middle Devonian to mid-Carboniferous (Millson et al., 1996; Droste, 1997; Osterloff et al., 2004). It would therefore appear that parts of Arabia could have been uplifted as highlands, most probably sometime between the mid-Silurian and Middle Devonian. These highlands may be related to pre-rift thermal swelling or post-rift isostatic rebound associated with the breakaway of the Hun Superterrane (Figure 6), rather than the Caledonian Orogeny (Figure 5).

The mid-Carboniferous unconformity is sometimes correlated to the Hercynian Orogeny, and the term “Hercynian unconformity” is adopted in regional and local studies by numerous authors (e.g. Stöcklin and Setudehnia 1972; Berberian and King, 1981). In Saudi Arabia, the angular pre-Unayzah unconformity correlates to the mid-Carboniferous hiatus (Figure 3, c. 325–310 Ma, Al-Husseini, 2004; at least Serpukhovian, Bashkirian and early Moscovian, Gradstein et al., 2004). The associated differential structural relief is manifested by broad epeirogenic swells (many 100s of kilometers in lateral extent) and laterally extensive (100s of kilometers) upthrown blocks (several 100s of meters), bounded by transpressional to reverse faults (Wender et al., 1998; Al-Husseini, 2004). The Hercynian Orogeny appears to have been too distant to account for the severity and style of this in deformation in Arabia (Figure 8 and 9).

An alternative to correlating the Hercynian Orogeny to the mid-Carboniferous Arabian unconformity is considered in the interpretation shown in Figure 11 of Sengör (1990). This Early Triassic reconstruction shows a SW-oriented subduction zone of the Paleo-Tethys oceanic crust beneath parts of Cimmeria. Next to the subduction complex, the Podataksasi Arc (a name Sengör derived from the initial letters of Pontides, Dzirula, Adzharia-Trialeti, Artvin-Karabagh and Sanandaj-Sirjan; Figure 11) was mainly a Carboniferous episode of orogenic deformation, metamorphism, and arc-type, calk-alkaline magmatism. This interpretation is based on a detailed study of successions in the involved terranes (see Sengör, 1990). Natal’in and Sengör (2005) included the Podataksasi Arc in the so-called Silk Road Arc, which stretched during the late Paleozoic-early Mesozoic from the Caucasus through north Iran and the Pamirs to China.

Further westwards, Xypolias et al. (2006) extended the subduction-arc model from the Pontides to the Hellenic terrane (External Hellenides) and northeast Greece. U-Pb dating of zircon from a granitic orthogneiss in the Kithira Island (southern Greece) yielded a late Carboniferous age of 324–323 Ma. Taken together with other geochronological data from the Aegean region it provides evidence for a restricted period of plutonism between 325–300 Ma (Xypolias et al., 2006). These authors concluded that northeast Greece (Cycladic and Palegonian basements) and northwest Turkey (Menderes Massif in the Taurides terrane and Sakarya Zone in the Pontides terrane) formed part of Cimmeria.

A reviewer (written communication, 2006) noted that a subduction complex does not transmit compressional horizontal stresses across broad regions and, moreover, a SW-directed compression appears inconsistent with the NS-trending grain of the fault-bounded Arabian structures. He pointed out that the initiation of subduction is commonly associated with a strong pulse of trench suction leading to roll-back, both processes exerting a strong extensional pull on the continental margin overlying the evolving subduction zone. As an analog to the Hercynian Orogeny’s influence in Arabia, he suggested the present-day Indian Plate. It is piercing into the anisotropic assembly of Central South Asia, with its effect reaching even the distant east coast of Asia. He concluded that a far-field relationship to Hercynian orogenic forces still cannot be completely excluded in Arabia.

Besides the data supporting a compressional mid-Carboniferous pulse due to a subduction-arc complex (Sengör, 1990, Figure 11; Xypolias et al., 2006), the concerns raised by the reviewer can be addressed. Whereas subduction complexes are indeed driven by slab-pull and associated with back-arc extension in the continental margin, this regime develops after an initial compressional stage. The pre-subduction stage involves first rupturing the entire oceanic crust (10 or more kilometers thick) and initiating subduction along a new thrust zone. The horizontal forces required to fracture the brittle crust, then bend and push down the oceanic plate are not only compressional but of regional significance. In some cases, the thrust geometry is reversed and the compressional force is great enough to push the oceanic crust above the continental margin resulting in an obduction. It is considered here that the mid-Carboniferous event was an early pre-subduction compressional phase, while the mid-Permian -Triassic was the extensional one.

The relationship between a SW-directed compression and the NS-trending Arabian uplifted fault blocks was one of mid-Carboniferous transpression along pre-existing NS-oriented fault systems. The Arabian basement manifests a fault system with a NS-, NE- and NW-trending grain that was established in the late Proterozoic and Early Cambrian (Al-Husseini, 2000, 2004). We argue that a SW-directed compressional pulse would have caused the pre-existing Arabian basement-cored structures to be dislocated in a right-lateral transpressional style.

In summary, it seems likely that two regional angular unconformities in the Arabian Plate were related to plate-tectonic events that occurred in the vicinity of the Middle East terranes. The mid-Silurian to Middle Devonian unconformity (middle Paleozoic event instead of Caledonian Orogeny) may have involved the uplift of the northern Arabia margin (and Oman) in Gondwana. The uplift may have occurred along the newly formed Paleo-Tethys Ocean prior to, during or possibly after the breakaway of the Hun Superterrane. The mid-Carboniferous unconformity (instead of Hercynian unconformity) may have resulted from deformation caused by a compressional subduction-initiation phase along the outer margin of Cimmeria before it broke off in the Middle Permian-Triassic.

DISCUSSION OF THE MIDDLE EAST TERRANES

In this section, we discuss the various plate-tectonic interpretations of the individual Middle East terranes (Figures 1 and 2). We highlight conflicting interpretations and suggest preferred interpretations where data and regional considerations allow. The terrane-by-terrane sections follow from present-day northwest to southeast (Figure 1).

The Pontides and Taurides Terranes

Turkey (Figure 1) presently occupies the active collision zone between the Arabian and Eurasian plates (Bird, 2003), and interpretations of its Paleozoic history vary from a single terrane to several. Scotese (2004) positioned Turkey next to the Levant and Egypt (Figures 1 and 2) throughout the Paleozoic and early Mesozoic, and in the Cretaceous showed it drifting northwards until it collided with Eurasia in the middle Cenozoic (c. 30 Ma). Similar single-terrane models involving only Mesozoic rifting were adopted by others (e.g. Grabowski and Norton, 1995; Sharland et al., 2001).

Göncüöglü and Kozlu (2000) separated Turkey into the northern Pontides and southern Taurides by a Paleozoic ocean that closed in the Carboniferous. They correlated post-collisional granitoids and suggested that the Taurides was originally Gondwanan. As discussed earlier, Sengör (1990, Figure 11) considered the Pontides (and Kersehir) and Taurides (Menderes-Taurus) terranes to be parts of two Cimmerian ribbons with the intervening Neo-Tethys (Inner Taurides) Ocean.

Several authors, however, interpreted the Pontides as Hunic and the Taurides as Cimmerian (Figures 6 and 7; Stampfli et al., 2001; Cocks and Torsvik, 2002). Moreover, based on a detailed analysis of foraminiferal paleobiogeography and plate tectonic review, Kalvoda (2002) concluded that Turkey was Laurussian, rather than Gondwanan. An investigation of the Carboniferous (Viséan) foraminiferal and algal paleobiogeography suggests that the Taurides was located along the northwestern border of the Paleo-Tethys (Okuyucu and Vachard, 2006).

Therefore it appears that the Paleozoic paleopositions of the main Pontides and Taurides terranes of Turkey remain unresolved in the literature. As discussed above, we favor the interpretation of these two terranes as Cimmerian and specifically within the regional context of the mid-Carboniferous subduction-arc complex (Figure 11, Sengör, 1990; Xypolias et al., 2006).

The Caucasian Terranes

The Greater Caucasus terrane is presently located south of the Russian Platform (Gamkrelidze, 1997; Tawadros et al., 2006) (Figures 1 and 2, i.e. Baltica), and its Paleozoic sedimentary complexes crop out in the central Greater Caucasus Mountains (Ruban, 2006). Paleontological data from Silurian (Ludlow) carbonates (bivalve and ammonoid assemblages, Bogolepova, 1997), Pennsylvanian paleobotanical data (Anisimova, 1979), and middle and upper Paleozoic paleomagnetic data (Shevljagin, 1986) suggest that the Greater Caucasus was not a part of Baltica, as traditionally proposed (e.g. Laz’ko, 1975; Bykadorov et al., 2003). The faunal and floral assemblages, as well as the lithostratigraphic architecture, are similar to those of Hunic Perunica and Carnic Alps (Central and Alpine Europe, Figures 2 and 7). Moreover, its lower Silurian mainly clastic and volcaniclastic succession resembles that of the Hunic margin of the Paleo-Tethys.

In the Middle-Late Devonian (until Famennian) about 4,500 m of mixed volcaniclastics and volcanic rocks were deposited in the Greater Caucasus (Kizeval’ter and Robinson, 1973). The volcanic activity may have been due to tectonism between the Greater Caucasus and other Hunic terranes. Alternatively, the magmatic activity may have been associated with the closure of the Rheic Ocean (Stampfli and Borel, 2002; Figure 6). We therefore follow Tawadros et al. (2006) in assigning the Greater Caucasus to the Hun (probably Cordillera) Superterrane.

The early Paleozoic location of the Greater Caucasus before the Hunic breakaway is uncertain. Tawadros et al. (2006) positioned it along the African-Arabian margin of Gondwana, but without constraining data. In the mid-Paleozoic, it was located near the easternmost extremity of the Hun Cordillera terranes with westward strike-slip dislocation along the northern Paleo-Tethys Shear Zone in the Carboniferous - Middle Triassic, and eastward dislocation in the Late Triassic - Early Jurassic (not depicted in figures in this paper). Such a late Paleozoic to Mesozoic shear zone may have stretched along the southern margin of Laurussia and connected with an intra-Pangean shear zone (Arthaud and Matte, 1977; Swanson, 1982; Rapalini and Vizán, 1993; Lawver et al., 2002; Stampfli and Borel, 2002; Bykadorov et al., 2003; Vai, 2003; Garfunkel, 2004; Natal’in and Sengör, 2005; Ruban and Yoshioka, 2005; Tawadros et al., 2006).

Stampfli and Borel (2002) positioned Kazakhstan (or parts of it) along the easternmost part of the Hun Superterrane suggesting proximity to the Greater Caucasus in the Devonian. Available paleontological data does not support this suggestion. The trilobite species Paciphacops occurs in the upper Silurian to Lower Devonian strata and its distribution encompasses the circum-Pacific (Merriam, 1973; Wright and Haas, 1990; Ramsköld and Werdelin, 1991; Edgecombe and Ramsköld, 1994). Its presence in Kazakhstan (Maksimova, 1968) and absence in Europe suggests the former was located on the margin of the Panthalassic Ocean, i.e. too far to be Hunic. This is also confirmed with other paleontological data (Blodgett et al., 1990; Campbell, 1977; Chlupác, 1975; Kobayashi and Hamada, 1977; Maksimova, 1972; Ormiston, 1972; Perry and Chatterton, 1976; Pedder and Oliver, 1990; Pedder and Murphy, 2004).

The Lesser Caucasus (Transcaucaus) terrane is presently located south of the Greater Caucasus, and north of Turkey and Iran (Figures 1 and 2). Interpretations based chiefly on paleomagnetic and paleontological data (Lordkipanidze et al., 1984; Gamkrelidze, 1986), indicate that it was apparently a separate terrane. It appears to have drifted northwards together with Cimmeria (“Iran-Afghan” microcontinent of Gamkrelidze). In the absence of conflicting evidence, we assign the Lesser Caucasus to Cimmeria. We also conclude that the paleopositions of the Caucasus along the margin of Gondwana or within the two superterranes remain unconstrained.

East Turkey, Northwest Iran and Alborz Terranes

The Eastern Turkey, Northwest Iran and Alborz regions are inconsistently interpreted in published reconstructions. Sengör (1990) interpreted Eastern Turkey as a Neo-Tethyan accretionary prism. Northwest Iran is considered Cimmerian and similarly depicted by several authors (e.g. Sengör, 1990; Sharland et al., 2001), but is sometimes referred to as the Alborz terrane by others (Stampfli et al., 2001; Torsvik and Cocks, 2004). In this review we consider Northwest Iran and Alborz as separate terranes (Figures 1 and 2).

Based on paleobiogeographic studies, Kolvoda (2002) suggested that the Alborz terrane was a part of the late Paleozoic Laurussia Supercontinent. Angiolini and Stephenson (in press), based on a re-examination of early Permian (Asselian-lower Sakmarian) brachiopods of the lower Permian Dorud Formation in the Alborz Mountains and a new study of palynomorphs from the same formation, also concluded that there is little affinity with Gondwana and the peri-Gondwanan region. Brachiopod fauna shows affinities with those of Baltica (Urals and of the Russian Platform), and to a lesser extent to the Trogkofel Limestone (Carnic Alps) in the west. The palynomorph assemblage is completely different from those recorded from the Asselian-Sakmarian Granulatisporites confluens Biozone, which is ubiquitous in the Gondwana region. L. Angiolini (2007, written communication) and coworkers, based on their studies and published data, concluded that the Alborz, Northwest and Central Iran remained adjacent to one another throughout most of the Paleozoic. This is reflected by the continuity and common evolution of their Paleozoic sedimentary rocks, and uniform distribution of biota. They attribute the similarity of the fossil record to the Urals to surface currents and the low latitudinal position of the Iranian terranes.

In summary, Eastern Turkey may not have been a Paleozoic terrane. The Alborz, Northwest and Central Iran terranes were apparently adjacent to one another. Their paleobiogeographic signature suggests a Laurussian affinity, but in the absence of more definitive data we follow most authors and assign them to the Cimmerian Superterrane.

Sanandaj-Sirjan Terrane

The Sanandaj-Sirjan terrane (Figures 1, 2, 4 to 11) was attached to the Zagros Mountains (and the Arabian Plate) until it broke off as part of Cimmeria in the mid-Permian - Triassic (Berberian and King, 1981; Sengör, 1990; Grabowski and Norton, 1995; Stampfli et al., 2001; Sharland et al., 2001; Scotese, 2004). Most authors show the Paleozoic position of Sanandaj-Sirjan adjacent to the Zagros Suture, effectively implying that today it occupies the same approximate position as 250 million years ago.

This seems remarkable as it was involved in the opening and closing of the Neo-Tethys Ocean. During the opening it may have subducted the Paleo-Tethys (Sengör, 1990), and during its closing, the Neo-Tethys (Ghasemi and Talbot, 2005).

Central Iran

Sengör (1990) divided the Central Iran microplate into the Lut, Tabas and Yazd blocks (Figure 11). Other authors consider Central Iran and Lut as synonyms (Stampfli and Borel, 2002; von Raumer et al., 2002, 2003; Torsvik and Cocks, 2004; Scotese, 2004; Golonka, 2004); or two neighboring terranes: Central Iran and Lut, or Yazd and Lut (e.g. Sharland et al., 2001; Stampfli et al., 2001). We adopt Sengör’s Central Iran terrane and follow others by considering it as Cimmerian.

Zagros Mountains and Makran Region

The Zagros Mountains region in southwest Iran forms a part of the Miocene-Pliocene collision zone between the Arabian and Eurasian plates (Figure 1). This region was a part of the Arabian Plate from the late Neoproterozoic to the present-day (Berberian and King, 1981; Sepehr and Cosgrove, 2004). During the Permian-Triassic (Figures 10 and 11), the opening of the Neo-Tethys Ocean along the Zagros Suture Zone was accompanied by normal faulting and horsts and graben systems (Sepehr and Cosgrove, 2004).

South of the Zagros Mountains, the Makran region in Iran and Pakistan (Figure 1) consists of the Inner Makran ophiolites and the Cenozoic Makran and Saravan accretionary prisms (McCall, 1997, 2002, 2003). This region is associated with the NE-directed subduction of the Gulf of Oman oceanic crust (a remnant of the Neo-Tethys Ocean) beneath Iran. The Makran core may have amalgamated with Central Iran and Sanandaj-Sirjan during the Triassic (McCall, 2003). Therefore, Makran may have formed a part of Mesozoic Cimmeria.

Helmand and Farah Terranes

Afghanistan, western Pakistan and southeast Turkmenistan are cored by the southern Helmand and northern Farah terranes and considered Cimmerian (Figures 1 and 2; Sengör, 1990; Sharland et al., 2001; Stampfli et al., 2001; Golonka, 2004). Scotese (2004) adopted Sengör’s (1990) model showing Helmand and Farah formed parts of two Permian-Triassic ribbons (Figure 11). Together with Karakoram in north Pakistan, the Farah and Helmand terranes are considered Cimmerian.

CONCLUSIONS

Recent publications that interpreted the Paleozoic tectonic units of the Middle East and their paleopositions were reviewed in the global context of supercontinents and exhumed vast oceans, to individual terranes. Adjoining the Arabian and Levant plates, ten Paleozoic Middle East terranes were apparently involved in the evolution of the Gondwana and Pangea margins and the Hun and Cimmeria superterranes. The Cimmerian terranes that broke off from Gondwana in mid-Permian - Triassic appear to have been: (1 and 2) Turkey’s northern Pontides and southern Taurides; (3 to 6) Alborz, Central Iran (Lut, Tabas and Yazd), Sanandaj-Sirjan and Northwest Iran; (7 and 8) Helmand and Farah of Afghanistan, western Pakistan and southeast Turkmenistan; and (9) the Lesser Caucasus. The Greater Caucasus may have been Hunic.

The Caledonian and Hercynian orogenies occured far away from Arabia. Correlation between these two orogenies and deformations in Arabia can be misleading. They imply that far-field stresses were transmitted many thousands of kilometers from the orogenic fronts to Arabia’s crust. The terms Caledonian and Hercynian should not be applied to the tectonic evolution of Arabia. Instead two significant and more proximal tectonic events were identified as possible near-field sources of regional deformation. The mid-Silurian breakaway of the Hun Superterrane is identified as a candidate that may be related to the mid-Silurian to Middle Devonian (middle Paleozoic) uplift in North Arabia and possibly Oman. The initiation of subduction, which could have preceeded the mid-Permian - Triassic breakaway of Cimmeria, is considered a possible force for the regional mid-Carboniferous faulting and epeirogenic deformation in Arabia.

Middle East plate-tectonic models require much more data and investigations if they are to be firmly constrained. The first step is to adopt common boundaries and names for the terranes, not only for the Middle East, but also of those in Asia and Europe (Figures 1 and 2). The second step requires constructing a regional tectono-stratigraphic framework that crosses from the interior of the Arabian Plate and its outer margins (Oman, Zagros, North Iraq, Syria and Southeast Turkey) to the ten and possibly more Middle East terranes. The framework requires correlating stratigraphic rock units that are much better constrained by age (biostratigraphy), paleontology and tectonics. Additionally, paleomagnetic and age data, together with the descriptions and interpretations of volcanic rocks could better clarify many aspects of the tectonic events.

ACKNOWLEDGEMENTS

We thank Lucia Angiolini, Manuel Berberian, Arthur Boucot, Robin Cocks, Anton Koopman, Joerg Mattner and other colleagues for their important comments and suggestions for improving the manuscript. Suggestions by Henrique Zerfass on the Late Paleozoic shear zones, and assistance with literature by Nico Janssen are highly appreciated. We recognize that many of the plate reconstructions beyond the Middle East region also remain unresolved but beyond the scope of this review. We therefore welcome any comments, improvements and discussion of the data and interpretations presented in this paper. The final design and drafting of the graphics by Heather Paul is appreciated.

ABOUT THE AUTHORS

Dmitry A. Ruban is a Senior Lecturer at the Geology and Geography Faculty of the Southern Federal University (Rostov-na-Donu, Russia). He received a Diploma in Geology from the Rostov State University in 2000, and a PhD (Russian equivalent) from the same university in 2004. Dmitry joined the faculty in 2003, and now teaches various courses to students, leads summer field practice, and supervises diploma theses. His broad research interests include Phanerozoic stratigraphy, paleobiology (fossil diversity and mass extinctions), paleoenvironmental reconstructions, tectonics, and geoconservation. He has conducted field work in the Caucasus, Rostov Dome, Donbass, and the Urals. Dmitry pays special attention to interregional comparisons and correlations. He has participated in a number of collaborative projects with geologists from Azerbaijan, Brazil, Poland, Canada, Japan, and USA and has published 10 papers in international journals and about 100 papers in Russian scientific media. Dmitry is also a permanent book reviewer of the German journal “Zentralblatt für Geologie und Paläontologie, Teil II”. He is a member of the Swiss Association of Petroleum Geologists and Engineers. ruban-d@mail.ru

Moujahed Al-Husseini founded Gulf PetroLink in 1993 in Manama, Bahrain. Gulf PetroLink is a consultancy aimed at transferring technology to the Middle East petroleum industry. Moujahed received his BSc in Engineering Science from King Fahd University of Petroleum and Minerals in Dhahran (1971), MSc in Operations Research from Stanford University, California (1972), PhD in Earth Sciences from Brown University, Rhode Island (1975) and Program for Management Development from Harvard University, Boston (1987). Moujahed joined Saudi Aramco in 1976 and was the Exploration Manager from 1989 to 1992. In 1996, Gulf PetroLink launched the journal of Middle East Petroleum Geosciences, GeoArabia, for which Moujahed is Editor-in-Chief. Moujahed also represented the GEO Conference Secretariat, Gulf PetroLink-GeoArabia in Bahrain from 1999-2004. He has published about 30 papers covering seismology, exploration and the regional geology of the Middle East, and is a member of the AAPG, AGU, SEG, EAGE and the Geological Society of London. geoarabi@batelco.com.bh

Yumiko Iwasaki is a PhD candidate in the Earth and Environment Sciences Department at the City University of New York, USA. She received a BA in Geology from Hunter College (CUNY), New York in 1996. Yumiko’s research, in association with the Division of Paleontology (Invertebrates) at the American Museum of Natural History, New York, took her to Bolivia, Peru and Morocco for fieldwork. She has also worked in the collections and helped translate Japanese scientific literature for staff. Yumiko has taught geology at Brooklyn College (CUNY) (1999–2002) and at Hunter College as an Adjunct Professor (2003–2005). Her research interests include Devonian phacopid trilobite phylogeny, Devonian biogeography, and Devonian stratigraphy of Bolivia. iwasaki@amnh.org