Most evolutionary reconstructions of this zone (e.g., Burchfiel, 1980; Burtman, 1986; Royden and Burchfiel, 1989; Fodor et al., 1998) suggest a progressive outward migration of the Carpathian arc (35-6 My), at the expense of a low buoyancy zone of the European foreland through a deformation pattern similar to that shown in Figures 2 and 3. In the internal part of this arc, trans-tensional tectonics took place, with the generation of the Pannonian basin. During the Oligocene-early Miocene (35-17 My), major shear zones allowed the east to northeastward displacement of crustal wedges in the northern part of the Pannonian area, while during the middle-upper Miocene (16-6 My) eastward migration of crustal wedges mainly occurred in the southern Pannonian region (e.g. Royden and Burchfiel, 1989; Fodor et al. 1998).
We advance the hypothesis that the development of this T-A-BA system was connected with an extrusion process, induced by the indentation of the Arabian plate (moving faster than Africa, since the lower Miocene, Hempton, 1987) against the wide orogenic system built up by the closure of the Tethyan ocean and by the consumption of the adjacent African and Eurasian margins. This system was constituted by three parallel belts (Figure 2a): a northern accretionary chain with European affinity (Carpathians-Balkanides-Pontides) and a southern one with African affinity (Dinarides-Hellenides-Taurides), separated by an inner zone (Tethyan belt) constituted by oceanic remnants, metamorphic bodies and crystalline massifs (i.e. the Pelagonian, Aegean and Anatolian) as suggested by a number of authors (e.g. Brunn, 1976; Biju-Duval et al., 1977; Boccaletti and Dainelli, 1982; Burtman, 1986; Royden and Burchfiel, 1989).
It is widely recognized that the indentation of Arabia caused the lateral escape of Anatolia (e.g., Mckenzie, 1972; Dewey and Sengor, 1979). However, we think that this extrusion process did not only involve Anatolia, since this zone still constituted an integral part of the long orogenic system mentioned above. Thus, the overall effect of this extrusion involved the migration and distortion of the whole Tethyan belt and of the adjacent chains, from the eastern Anatolia to the Carpathians, as tentatively reconstructed in Figure 2. This hypothesis is suggested by the fact that the Tethyan belt has maintened its original continuity till to the present (Figure 3c), in spite of the considerable deformation it underwent. In the first phase (Figure 2b), the lateral escape of Anatolia was oriented roughly NWward , guided by a system of major dextral shear zones (Dercourt et al.,1986; Hempton, 1987; Finetti et al.,1988). This displacement of Anatolia and of the whole Tethyan system was accommodated by the lateral NEward extrusion of orogenic wedges in the Carpathian arc, at the expense of a low buoyancy sector of the European foreland, and by a SWward bending of the Aegean arc, at the expense of the Ionian-Levantine old oceanic lithosphere. In the wake of the outward migrating crustal wedges in the Carpathian arc, transtensional deformation took place in the Pannonian area (Figures 2b and 2c).
During this phase, longitudinal shortening also occurred in the Balkanides, accommodated by the NEward lateral escape of crustal wedges, at the expense of the southern Moesian margin (e.g. Stanishkova and Slejko, 1991). Since the proposed pattern (Figure 2b) provides that the NWward displacement of the Balkanides-Carpathian belt is greater than that of the inner Pelagonian massifs, one should expect a left lateral decoupling between these belts. The imprints of this decoupling might be represented by the sinistral shear deformation recognized in the Vardar zone for the period involved (e.g. Brunn, 1960, 1976; Burtman, 1986; Zeilinga de Boer, 1989).
The development of the Carpathian T-A-BA system underwent slowdown/cessation as the arc collided with the continental European domain, with a progressive evolution of this stop from north to south (e.g. Royden, 1993b). At present, minor tectonic activity in this arc only occurs in its southernmost corner, in the Vrancea zone.
The deformation pattern of the Carpathian-Pannonian system appears to be consistent with the dynamic implications of the extrusion model, since the overall shape of the Carpathian arc strongly resembles that of a large crustal wedge, extruded in response to a SE-NW compression. Also, one could note that the outer part of this arc is longer than the internal one, in line with the deformation pattern predicted by the simulation of extrusion processes (e.g., Ratschbacher et al., 1991; Faccenna et al., 1996; Keep, 2000; Sokoutis et al., 2000). The presence of a system of strike slip faults in the Pannonian area is consistent with the proposed driving mechanism, which implies the simultaneous action of a SE-NW compression and of a perpendicular extension.
Instead, the above deformation cannot easily be reconciled with the pure SW-NE extension predicted by the slab pull mechanism. Furthermore, one should consider that arc migration driven by slab pull would imply fragmentation and longitudinal interruption of the initial orogenic belt. Instead, the present configuration of this arc (Figure 2) clearly shows that its various segments have remained in close contact during the evolution.
The timing of this T-A-BA system (e.g., Royden, 1993b) can be plausibly related with the proposed driving mechanism, i.e. the indentation of Arabia. In fact, the most intense tectonic activity in the Carpatho-Balkan-Pannonian area just followed the activation of the tectonic zones which allowed the decoupling of the Arabian promontory from Africa, with particular regard to the Red Sea rifting zone and the Dead Sea fault system (e.g., Hempton, 1987). Instead, the implications of the slab pull mechanism do not provide any precise justification for the onset of slab roll back in the Carpathian arc around the late Oligocene-early Miocene. The consuming process under the Carpathian belt seems to have begun much earlier, at least in the Paleocene (Royden and Baldi, 1988; Dercourt et al., 1986), and thus one should explain why slab roll back has not occurred prior to the Miocene.
Furthermore, we think, on the basis of the arguments already pointed out in the discussion of the Western Mediterranean T-A-BA system, that the remarkable bending that the Carpathian arc underwent cannot easily reconciled with a slab pull driving mechanism.