In a number of subduction zones, the overriding plate has been affected by extensional tectonics with the consequent formation of a thinned zone, called back arc basin (e.g. Karig, 1971). The formation of this basin delineates a new structural element, here called arc, which corresponds to the portion of the overriding plate comprised between the basin and the trench. The arc is generally constituted by accretionary and/or magmatic material. In few cases, as in the western Mediterranean, the arc also included fragments of the foreland from which it detached (i.e. the Corsica-Sardinia microplate). The opening of back arc basins has repeatedly occurred in the Mediterranean area (e.g. Royden et al., 1993a,b; Mantovani et al., 1997, 2000a, 2001a), in several circum-Pacific zones and in some Atlantic zones (e.g. Karig, 1971; Uyeda and Kanamori, 1979; Taylor and Karner, 1983).
The main features of subduction zones in the world and of the possibly related extensional basins (where present) are reported in the Table. The geographical locations of these structures are illustrated in Fig. 1. In a number of cases, crustal stretching has occurred in the wake of a trench-ward migrating arc, in connection with trench retreat and slab roll back. Hereinafter, this kind of structure will be mentioned as 'trench-arc-back arc (T-A-BA) system'. In other cases, extensional tectonics has indeed occurred in the overriding plate, but the geometry of the troughs and the extensional trend do not allow to recognize them as typical T-A-BA systems. In some subduction zones, as those of South America and the Java-Sumatra one, a back arc basin never opened up.
Table 1. (View Table) Major parameters of subduction zones in the world and, where present, of the basins lying in the respective overriding plates. Some basins, as the West Philippine, Celebes, Sulu, Solomon, Coral, South Fiji, New Hebrides, Tasman have not been included in the Table since no extensional activity is clearly recognized in these zones for the last 30-40 My (e.g., Hilde and Lee, 1984; Uyeda, 1986; Tamaki and Honza, 1991) and since it is not yet clear to which subduction zones they could be associated (see Uyeda and Kanamori, 1979 and references therein). Age = Age of the subducting lithosphere. Vs = trench normal component of the absolute velocity of the slab. Vo = trench normal component of the absolute velocity of the overriding plate. For subduction zones with active back arc opening (numbers with asterisk), the motion rate of the arc is reported. A positive value indicates that the overriding plate/arc is departing from the subducting plate. Larc = length of the consuming boundary. Ls = slab length. f = slab dip. Time = time span of extensional activity. Vext = extensional rate. Text = extensional trend. The numbers reported in the two columns of references correspond to the following papers: 1) Scholz and Campos, 1995; 2) Peterson and Seno, 1984; 3) Pacheco et al., 1993; 4) Tamaki and Honza, 1991; 5) Yogodzinski et al., 1993; 6) Kusunoki and Kimura, 1998; 7) Jolivet et al., 1994; 8) Brooks et al., 1984; 9) Lee and Lavwer, 1995; 10) Uyeda and Kanamori, 1979; 11) Carlson and Mortera-Gutierrez, 1990; 12) Briais et al., 1993; 13) Tregoning et al., 1998; 14) Carlson and Melia, 1984; 15) Taylor and Karner, 1983; 16) Honza, 1995; 17) Pellettier et al., 1998; 18) Charvis and Pellettier, 1989; 19) Parson and Wright, 1996; 20) Darby and Meertens, 1995; 21) Barker et al., 1984; 22) Papazachos et al., 2000; 23) McClusky et al., 2000; 24) Giardini and Velonˆ, 1991.
The driving mechanism of back arc opening is still matter of debate (e.g. Uyeda and Kanamori, 1979; Taylor and Karner, 1983; Tamaki and Honza, 1991; Taylor, 1995; Mantovani et al., 1997, 2000a; Flower et al., 2001). The fact that this phenomenon is systematically associated with subduction has led many authors to believe that a causal relationship exists between the two processes and that, in particular, back arc extension is a side effect of subduction. A variety of hypotheses has been advanced about the tectonic mechanism responsible for this connection. In this work, we only focus our attention on the three of them which are the most quoted in literature, i.e. the 'slab pull', the 'corner flow'and the 'sea anchor' models.
The slab pull model (Fig. 2b) postulates that back arc extension is driven by the negative buoyancy of the subducted lithosphere with respect to the surrounding mantle (e.g. Molnar and Atwater, 1978; Dewey, 1980; Malinverno and Ryan, 1986; Royden, 1993a,b). This force would cause roll back of the slab, inducing a tensional stress in the overriding plate able to cause extensional deformation.