The main reverse fault (at SP 390) intersects the PR at a significant angle of about 60° (Fig. 9B). In the portion of the sedimentary section located east of the ridge and below the depth of the intersection point of the reverse fault and the ridge, sedimentary layers are flat (SP 0-200). This implies that the eastern prolongation of the major reverse fault occurs along the contact between the PR and the sedimentary cover (Fig. 11).
We propose a model of crustal and sedimentary compressive deformation focused by the presence of a serpentinized PR, in which the basic intraplate compressive motion occurs along a main intracrustal fault, accommodating the crustal shortening within the brittle crust (Fig. 11). In fact, the compressive motion could correspond to the reactivation of a large normal fault (Fig. 4), which limits the PR and was active at the end of rifting during the uplift of the ridge as suggested by Beslier et al. (1993). This crustal fault probably separates the lower crustal serpentinized peridotites already evidenced from refraction measurements (Pinheiro, 1994; Whitmarsh et al., 1990) from the highly serpentinized peridotites that compose the ridge. In addition, the seawater circulation might increase the degree of serpentinization along the fault and, consequently, the formation of lubricant (e.g., clay or talc) and the probability of further reverse motion along it. During the Betic compression, the reverse fault follows the interface between the ridge and the sediments already deposited up to the top of the ridge. Sediments infilling depressions remain undisturbed. However, sediments lying above the ridge are deformed and the portion of the reverse fault located within the sediments merges with the sea bottom of that time. The dip of the reverse fault is a function of the physical properties of the sediments, which explains why the fault trend makes a significant angle of about 60°, with the portion of the fault lying along the ridge flank. West of the main reverse fault, the deformation front corresponds to a very small reverse fault, which implies a slight shear motion near the top of the ridge.
The existence of a normal fault that was reactivated as a reverse fault during the Betic orogeny is supported by ODP drilling results. Within Holes 897C and 897D, 180 m of clayey conglomerates with serpentinite, sandstone, dolomite, limestone, claystone with serpentinite, massive serpentinized peridotite with shear bands have been drilled (Shipboard Scientific Party, 1994). Site 897 peridotites under- went a late shear deformation event, leading to the development of foliation and shear bands in the serpentinite and local brecciation of the serpentinite. These tectonic breccias occur throughout the cores of both holes and are locally extremely sheared or fractured. The shear features may document a deformation involving the upper portion of the basement during the last stages of rifting that may have trapped sediments within the shear zone. Site 897 is located west of the deformation front. No sign of Tertiary shear motion appears in the cores. However, both the extensively serpentinized upper portion of the ultramafic section and the underlying claystone with serpentinite breccias are appropriate candidates for the location of shear motion as suggested by the model.
Peridotites emplaced at the OCT are characterized by a high degree of serpentinization (Girardeau et al., 1988). Because the process of serpentinization results in a lubricant effect, PRs emplaced at the OCT are potentially weak zones for further tensional or compressive motions. Strain deformations will be primarily localized along the PR flanks. Therefore, under significant compressive motion, the final result could be the duplication of transitional crust and the formation of Ophiolites as suggested by Whitmarsh et al. (1993).
Seismic profiles acquired in the Iberia Abyssal Plain show that up to two parallel reverse faults exist there (Beslier et al., 1993; Whitmarsh and Miles, 1995). This may mean that two overlapping PRs lie parallel to the margin and correspond to uplifted portions of lower crustal peridotite with highly serpentinized material at their surface. However, the absence of serpentinized ridges does not mean that a lower crustal peridotite layer is absent, as attested by refraction results obtained west of Iberia on the Galicia Margin (Horsefield, 1992; Sibuet et al., 1995; Whitmarsh et al., 1993), Iberia Abyssal Plain margin (Pinheiro, 1994; Whitmarsh et al., 1990), and Tagus Abyssal Plain margin (Pinheiro, 1994; Pinheiro et al., 1992), but this only means that no PR has been uplifted during the last stages of rifting of these continental margins.
This statement is only valid in faint compressive settings such as the Betic, west of Iberia. It is anticipated that for larger compressive motions, as in the north Galicia area during the Pyrenean phase of compression, reverse faults could also develop along previous zones of weakness, as well as along normal faults that were previously active during rifting and that bound tilted fault blocks located over extremely thinned continental crust.
In a reverse manner, the existence of a slight compressive motion could help to locate the OCT, if serpentinized PRs are present. PRs probably exist in the Tagus Abyssal Plain, for example, where refraction and magnetic data are insufficient to precisely locate the OCT (Pinheiro, 1994). Here, well defined synsedimentary reverse faulting has been observed on several seismic profiles acquired in this area. We suggest that these Miocene reverse faults are associated with a crustal motion that occurs along serpentinized PRs emplaced at the OCT. However, further geophysical work is needed to confirm such a hypothesis.