Summary of previous works

Over the course of the last century, and particularly since the emergence of the modern plate tectonics theory, numerous studies have aimed to reconstruct the evolution of the Mediterranean basins in the context of the Alpine orogeny. The main concepts used in these reconstructions (as well as in the present study) correspond to continental drift, microplate rotations and migration of subduction zones. They were introduced as early as 80 years ago in the outstanding tectonic synthesis of Argand (1924) (Figure 2a). In an early attempt to reconstruct the tectonic evolution of the Mediterranean region, Carey (1958) has demonstrated that the unbending of arcuate mountain belts throughout the western Tethys yielded the reassembly of Spain, Balearic Islands, Corsica, Sardinia, Sicily and Italy in the northwestern margin of the Tethys Ocean (Figure 2b). Subsequent reconstruction models have generally followed Carey’s ideas concerning the palaeo-position of these terranes (Smith 1971, Alvarez et al. 1974, Boccaletti & Guazzone 1974, Biju-Duval et al. 1977, Cohen 1980) (Figure 2c,d), although the significance of the Cenozoic deformation was not always recognised (e.g., Smith 1971).

Figure 2. Examples of earlier reconstructions

Examples of earlier reconstructions

Examples of earlier reconstructions showing the western Mediterranean prior to the opening of Late Cenozoic extensional basins. (a) Argand (1924); (b) Carey (1958); (c) Alvarez et al. (1974); (d) Boccaletti and Guazzone (1974). Ap = Apennines; Ba = Balearic Islands; Be = Betic; Ca = Calabria; Co = Corsica; Ka = Kabylies; Sa = Sardinia; Si = Sicily.


During the 1970s, it was suggested that the western Mediterranean basins are relatively late tectonic features formed progressively since the Late Oligocene (Dewey et al. 1973, Alvarez et al. 1974, Biju-Duval et al. 1977). Since then this conclusion has been strongly supported by evidence of Oligocene and Miocene extensional deformation and syn-rift deposits on the margins of the western Mediterranean basins (e.g., Cherchi & Montadert 1982, Rehault et al. 1985, Bartrina et al. 1992 and many others). Thus, the western Mediterranean basins are essentially different from the Eastern Mediterranean; where the remnants of Mesozoic oceanic crust (Neotethys) are probably preserved below the sediments (de Voogd et al. 1992, Ben-Avraham et al. 2002). Mesozoic oceanic crust is not found on the floor of the western Mediterranean basins. However, the existence of consumed oceanic basins in this region is implied by reconstruction models (e.g. Dercourt et al. 1986, Ricou et al. 1986) and by the occurrence of ophiolitic complexes within the adjacent fold-and-thrust belts (Ricou et al. 1986, Knott 1987).

The configuration of extensional basins surrounded by continuous mountain belts has been commonly interpreted as the result of back-arc extension (Boccaletti & Guazzone 1974, Biju-Duval et al. 1977, Cohen 1980, Ricou et al. 1986). In most reconstructions, it has been stressed that back-arc extension led to drifting of continental blocks and to large-scale block rotations (Dewey et al. 1973, 1989, Alvarez et al. 1974, Biju-Duval et al. 1977). Corsica and Sardinia underwent counterclockwise rotations during the opening of the Ligurian Sea, whereas the opening of the Valencia Trough was accompanied by clockwise rotations of the Balearic Islands (Montigny et al. 1981, Pares et al. 1992). The Kabylies and the Calabrian blocks, which had been deformed and metamorphosed in the Alpine orogen, migrated to their present locations during the opening of the Algero-Provençal Basin and the Tyrrhenian Sea, respectively (Alvarez et al. 1974, Cohen 1980, Dewey et al. 1989). Boccaletti and Guazzano (1974) have further suggested a southward and eastward migration of subduction arcs during the formation of the western Mediterranean basins. Similar ideas explained the existence of extensional back-arc basins in convergent margins (Dewey 1980). In the western Mediterranean, it led to the recognition of subduction rollback as an important driving mechanism in the tectonic evolution of the region (Rehault et al. 1985, Malinverno & Ryan 1986, Royden 1993a, Lonergan & White 1997).

The mechanism of subduction rollback has been discussed by Elsasser (1971) Molnar and Atwater (1978), Dewey (1980) and Royden (1993b). These authors have suggested that subduction rollback is the result of a negative buoyancy of the subducting slab relative to the asthenosphere (Figure 3), obtained when the subducting slab is cold and dense, as in the case of oceanic slabs older than ~50 Ma (Molnar & Atwater 1978). It results in a vertical sinking of the subducting lithosphere beneath the asthenosphere, which can lead to a regressive motion of the subduction hinge (Lonergan & White 1997) (Figure 3). As rollback occurs, it produces a potential vacant region, which can either be supported by convergence that matched or exceeds the rates of the retreating hinge, or by back-arc extension in the overriding lithosphere (Royden 1993b). In the works of Rehault et al. (1985), Malinverno and Ryan (1986), Royden (1993a) and Lonergan and White (1997), the evolution of the western Mediterranean basins has been mainly attributed to the rollback of a NNW dipping subduction zone. We note that our reconstruction largely reflects ideas previously presented in these works.

Figure 3. Simplified cross section

Simplified cross section

Simplified cross section showing the evolution of subduction rollback (modified after Lonergan & White 1997). (a) P and R are two components of the vertical negative buoyancy (F) of the subducting slab. If the subducting slab is cold and dense, the component R cannot be supported by the mantle asthenosphere, and the subduction zone is pulled backward; (b) back-arc extension forms when the rate of subduction rollback (Vr) exceeds the rate of convergence (Vc).