The western margin of Iberia comprises three segments (the Tagus Abyssal Plain segment, the Iberia Abyssal Plain segment, from Estremadura Spur to Vasco da Gama Seamount, and the segment that lies west of Galicia Bank), which appear to have experienced progressive breakup from south to north during the Early Cretaceous (Pinheiro et al., 1992). Geological and geophysical studies of each of these segments have provided data leading to a conceptual model for the nature of the ocean/continent transition on this nonvolcanic rifted margin. These studies and the subsequent model are outlined below.
In the Tagus Abyssal Plain, Pinheiro et al. (1992) showed magnetic models that indicate seafloor spreading began about 136 Ma (Valanginian in the time scale of Harland et al., 1990). They used seismic refraction, seismic reflection, and magnetic profiles to show that the oceanic crust adjacent to the OCT is unusually thin (2 km) and that a transitional region lies between the thinned continental crust and the thin oceanic crust; this region, although not truly oceanic (for example, it has no seafloor-spreading magnetic anomalies), has a magnetization far stronger than is usually associated with continental crust. This may indicate the presence of intrusive and extrusive material within the crust of the transitional region. The thin oceanic crust is underlain by a 7.6- to 7.9-km/s layer, which is probably serpentinized peridotite.
Whitmarsh, Miles, and Mauffret (1990) and Whitmarsh et al. (1993) studied the middle segment off Iberia (between 39° and 41°30'N) using seismic refraction and reflection profiles, gravity, and magnetics. They found that the oceanic crust adjacent to the OCT is thin (4 km) and that the OCT is underlain by a layer of about 7.6 km/s. A regional magnetic anomaly chart and modeling of a magnetic anomaly profile across the Iberia Abyssal Plain (Whitmarsh, Miles, and Mauffret, 1990), strongly suggest that seafloor spreading began about the time of anomaly M3 (130 Ma, Barremian), but that crust to the east of M3 was weakly magnetized and probably of continental origin. This analysis appears to be confirmed by a deep-towed magnetometer profile that also conclusively indicates that a transitional region exists between the thin oceanic crust and the thinned continental crust that, although not formed by seafloor spreading, has a relatively high magnetization (R.B. Whitmarsh and P.R. Miles, unpubl. data). Modeling of a single east-west gravity profile across the Iberia Abyssal Plain, which was constrained by a multichannel seismic reflection profile and by seismic refraction profiles, appears to confirm the existence of thinned continental crust adjacent to unusually thin oceanic crust, both of which are underlain by a continuous layer having a velocity of about 7.6 km/s. Two possible explanations for this unusual layer have been proposed (Whitmarsh et al., 1993). First, it may represent underplated material emplaced at the time of rifting and breakup, as in White and McKenzie's model (1989); although such material is commonplace on volcanic rifted margins, the essential absence of synrift volcanism on the west Iberia margin, the restriction of the layer to only the parts of the OCT where the crust is thinnest (unlike, for example, the more widespread occurrence of underplated material and seaward-dipping reflector sequences off the western Rockall Plateau; Fowler et al, 1989), and the relatively high velocity associated with the layer all strongly suggest that this explanation is unlikely to be the correct one. Second, the layer may represent serpentinized upper mantle peridotite; serpentinization might have occurred immediately after lithospheric breakup and during the onset of seafloor spreading because of the relatively easy access of seawater, probably thermally driven, to the upper mantle through the thin crust, in an analogous fashion to the serpentinization of the uppermost mantle, known to have occurred in oceanic fracture zones (Calvert and Potts, 1985). Last, the Iberia Abyssal Plain at this latitude also is underlain by a smooth acoustic basement in multichannel seismic reflection profiles between the apparently most-seaward tilted continental rift block to the east and a highly linear basement ridge to the west. One might speculate that this smooth basement represents volcanic rocks, thereby also providing an explanation for the high magnetic anomalies in the transitional region between the thin oceanic crust and the weakly magnetized thinned continental crust. Should tentative extrapolations of basement morphology be correct, the basement ridge represents the counterpart in the Iberia Abyssal Plain of a peridotite ridge drilled off Galicia Bank (Boillot, Winterer, Meyer, et al., 1987; Beslier et al., 1993).
The western margin of Galicia Bank, the third of the three segments, has been studied with seismic refraction and reflection profiles and also has been sampled extensively with dredges, submersibles, and by drilling (Horsefield, 1992; Mauffret and Montadert, 1987; Boillot, Winterer, Meyer, et al., 1987; Boillot et al., 1988). A seismic refraction model across the margin shows a thinned continental crust at the OCT adjacent to a moderately thinned (5-km) oceanic crust, which thickens rapidly to the west (Horsefield, 1992). A layer having a 7.2- to 7.3-km/s velocity, which underlies the thinned continental crust, may represent either crustal underplating (Horsefield, 1992) or serpentinization of the upper mantle. In August 1992, a continuous gravity profile was obtained along an east-west seismic refraction line across the whole OCT. A complete crustal-density model across the OCT, constrained by seismic velocities, has now been computed (J.C. Sibuet, pers. comm., 1992). Because crust that formed during the Cretaceous constant-magnetic-polarity interval abuts the OCT at this margin, magnetic anomalies cannot be used to date the beginning of seafloor spreading. However, the recognition of the reversed-polarity paleomagnetic interval M0 in cores from below the breakup unconformity at Site 641 (Ogg, 1988) indicates that breakup occurred about 120 Ma (Aptian). Sampling has shown unequivocally that a north-south basement ridge, which appears to coincide with an abrupt ocean/continent boundary, is composed of serpentinized peridotite.
The cumulative results from studies of these three segments of the west Iberia margin suggest that the following features are characteristic of the OCT in this region and may exist elsewhere in similar settings:
Before Leg 149, the following working hypothesis (Whitmarsh et al., 1993) seemed to account for the abnormally thin crust. Continental rifting before breakup (i.e., before the onset of seafloor spreading) often develops over tens of millions of years, yet rarely involves extension of more than 100 to 200 km at a single margin. In this situation, the slow rate of extension results in conductive cooling of the ascending mantle diapir. Consequently, much less melt is produced than in the case of adiabatic decompression (Bottinga and Allègre, 1978). Fracture zone (FZ) seismic velocity structures and thin FZ crust have been explained by a poor magma supply, caused by the cooling effect of the adjacent older oceanic plate (Whitmarsh and Laughton, 1976; Langmuir and Bender, 1984). An analogous mechanism may explain the thin oceanic crust at the OCT, where the poor magma supply was caused by the adjacent cooler continental lithosphere. However, this situation is temporary because a spreading axis continually recedes from the two adjacent continental lithospheric plates. Hence, the "steady-state" production of melt characteristic of a slow-spreading ridge takes some time to develop once continental breakup has occurred. Were magma production a passive response to the upwelling of asthenospheric material and, indirectly, to the separation rate of the overlying plates (as in the model of White and McKenzie, 1989), then mantle melting should have increased at breakup to keep pace with plate separation at a rate that was faster than the original horizontal stretching of continental crust alone. The thin oceanic crust at the OCT appears to represent the earliest product of seafloor spreading at the west Iberia margin. The width (a few tens of kilometers) of this region thus should indicate the time needed to establish steady-state melt production (a few million years). In fact, the oceanic crust may gradually thicken oceanward.
The above conceptual model of the OCT off Iberia was developed by Whitmarsh et al. (1993; Fig. 3A). The elements of this model were derived principally from the observations summarized in this section. Independent gravity modeling of a gravity profile across the Iberia Abyssal Plain, constrained by seismic refraction and reflection structures, led to a structure that is consistent with the broad features of the model, in particular the existence of a wide zone of 7.6 km/s and 3.26 Mg/m3 at the base of the crust (Fig. 3B). The 7.6-km/s layer might be serpentinized peridotite or, alternatively, the result of underplating. We can explain the presence of a serpentinized peridotite layer by the downward percolation of seawater through the thin, faulted oceanic or continental crust to the underlying mantle around the time of breakup, similar to what happens at fracture zones. Underplating is a less likely explanation on this margin because synrift volcanics are absent on land and because the 7.6-km/s velocity is relatively high for underplated material and can be found only where the continental crust is thinnest.
Several aspects of our model need further elaboration and investigation. One is the existence of a basement peridotite ridge (so far, detected with certainty only off Galicia Bank) and its emplacement mechanism. Another is the fact that the seaward edge of the thinned continental crust may be densely intruded by dikes (e.g., as observed in the southern Red Sea; Voggenreiter et al., 1988) or even be covered in places by lava flows. At present, our only evidence of this off Iberia is a region of unusually strongly magnetized continental crust in magnetic models of profiles across the Tagus and Iberia abyssal plains that coincides with a smooth diffractive acoustic basement (Pinheiro et al., 1992). Nevertheless, Srivastava et al. (1988) also postulated the existence of a similar feature when modeling profiles across the north-eastern Newfoundland margin.
Clarification and testing of aspects of the above model constituted the chief reason for drilling the Leg 149 sites.