The tectono-metamorphic development of the southern Iberia Abyssal Plain segment of the west Iberia margin has been investigated at three distinct scales: in thin section, in hand specimen, and at the scale of individual seismic reflection profiles. Similar studies have been carried out off Galicia Bank (e.g., M. Beslier, unpubl. data; Boillot et al., 1995b; Brun and Beslier, 1996; Reston et al., 1996).
The development of the mantle section at the microscopic and hand-specimen scales is revealed mostly by the study of the peridotite cores from Sites 897, 1068, and 1070. The tectono-metamorphic evolution of the Site 897 peridotites was described by Beslier et al. (1996). Here, four stages were recognized, high-temperature (900°-1000°C) ductile shearing, limited partial melting, subsolidus reequilibration in the plagioclase field at <1 GPa, and mylonitic shearing at 700°C under high deviatoric stress and low pressure. The rocks appear to have undergone a continuum of deformation at decreasing temperature under coeval increasing deviatoric stress and decreasing pressure. Shear deformation was a major mechanism of stretching and thinning of the lithosphere. At the present time, only preliminary results from Sites 1068 and 1070 are available. At Site 1068, the Subunit 1B serpentinized peridotites are foliated but commonly only weakly so. The average dip of the foliation is 43° in a northwest to west-southwest direction. The foliation is suggested to have formed under high-temperature conditions (Shipboard Scientific Party, 1998c). At Site 1070, the serpentinized peridotites are locally weakly foliated with discrete layers of pyroxenite or unfoliated. Olivine relicts show dislocation lamellae and some kink bands, indicating a high-temperature upper mantle deformation. Pyroxenes are mostly undeformed. The serpentinization operated, at least in part, after intrusion of the igneous mafic phase. In thin section, high-temperature deformation features are rarely seen. The high-temperature foliation is moderately inclined in the peridotites. The gabbroic veins are moderately to weakly inclined. The apparent absence of high-temperature deformation suggests that these mantle rocks did not undergo intense deformation during their exhumation.
The absence of significant mylonitic deformation in the mantle rocks at Sites 1068 and 1070 and the presence of ultramylonitic shear bands only locally at Site 897 contrast with the intense ductile deformation observed in the Site 900 metagabbro and in the upper amphibolites of Site 1067. Nevertheless, the mafic cores from Hobby High exhibit a similar metamorphic history from granulite to amphibolite to greenschist facies. The Site 1067 and 1068 gabbros and tonalite veins were intruded and experienced granulite to amphibolite facies metamorphism at ~270 Ma (late Hercynian) (R. Rubenach, N. Froitzheim, P. Wallace, M. Fanning, and R. Wyzsoczanski, unpubl. data). This was followed by greenschist facies metamorphism and then very low-grade metamorphism (fig. 35 in Shipboard Scientific Party, 1998b). The Site 1067 metatonalite lenses or veins exhibit mylonitic microstructures indicative of deformation under greenschist facies conditions (Manatschal et al., in press). Consistent shear-sense indicators and crystallographic preferred orientation in dynamically recrystallized quartz layers provide evidence for strongly noncoaxial (simple shear) deformation. The Site 1067 amphibolites similarly experienced retrograde metamorphism under amphibolite to greenschist facies conditions dominated by hydration reactions (Gardien et al., in press). The Site 900 metagabbros were strongly sheared in granulite to high-amphibolite facies conditions that were followed by intense fluid-assisted extension under greenschist facies conditions (Cornen et al., 1996b).
All basement cores exhibit late stage low-temperature brittle deformation accompanied by veining, and in the case of the peridotite cores, intense serpentinization that, at least at Site 1070, appears to decrease with depth. Using stable isotope chronology of flow and deformation, two episodes of fluid infiltration through serpentinized peridotite have been distinguished (Skelton and Valley, 2000). The first episode, at temperatures above 175°C, was pervasive and coeval with the serpentinization; the second episode occurred at 50°-150°C, was "structurally focused," and accompanied mantle exhumation. These authors therefore conclude that upper mantle serpentinization occurred before exhumation. The serpentinized peridotite may have formed weaknesses exploited by faulting and because of its low permeability (and density?), may have inhibited melt migration to the top of basement.
The effects of low-temperature hydrothermal fluids, especially their association with the fault at Site 1068, have been extensively studied by Beard and Hopkinson (2000), Beard (Chap. 2, this volume), and Hopkinson and Dee (in press). Beard and Hopkinson (2000) showed that the Site 1068 fault was host to a hydrothermal system rooted in serpentinization reactions at depth. The serpentinites and breccias exhibit a zonation, which reflects the mixing of seawater with a fluid whose composition is controlled by serpentinization reactions. Hopkinson and Dee (in press) also studied the complex multistage hydrothermal mineralization associated with the fault and showed how very late stage aragonite clusters replace serpentinite (both fractally and nonfractally). The clusters are thought to result from incursions of reactive seawater in and around the fault in response to pressure gradients. The latter may have been short-lived high-flowrate events most likely generated by tectonism. It has even been suggested that the basal sediments and basement(?) at Sites 897 and 1070 remain overpressured today (Ask, [N1]; Karig, 1996).
Studies of the low-temperature alteration of Leg 149 cores were carried out by Agrinier et al. (1996) and Gibson et al. (1996a).
A persistent feature of most of the cores that sampled the sediment/igneous-metamorphic basement interface was the presence of mass-flow deposits (olistostromes and breccias) at the base of the sedimentary section (Sites 897, 899, 1068, and 1070), slumped and fractured deposits (Site 1069), and tectonic breccias within the igneous-metamorphic basement, sometimes localized as narrow shear zones (Sites 897, 900, 1067, 1068, and 1070). These phenomena are a clear sign of brittle failure of the upper basement rocks at a late stage in the rifting process; the sediments involved are Late Jurassic(?) to Early Cretaceous in age. The mass flow deposits of Site 897 and Site 1068 have been studied in detail by Comas et al. (1996) and Gibson et al. (1996b) and by St. John (Chap. 1, this volume), respectively.
Many authors have made tectonic interpretations of seismic reflection profiles, aided by other geophysical and ODP results, to infer how rifting of the west Iberia margin developed and to propose more general models. This has been done either off Galicia Bank (e.g. Boillot et al., 1995b, 1989; Hoffmann and Reston, 1992; Krawczyk and Reston, 1995; Manatschal and Bernoulli, 1999a; Pickup, 1997; Reston et al., 1995, 1996; Sibuet, 1992b; Sibuet et al., 1995) or in the southern Iberia Abyssal Plain (e.g., Krawczyk et al., 1996; Manatschal et al., in press; Whitmarsh et al., 2000). Another model was developed from analogue modeling (Brun and Beslier, 1996). Wilson et al. (1996, in press a) studied the distribution of synrift sediments off Galicia Bank and in the southern Iberia Abyssal Plain. Wilson et al. (in press a) emphasize that published identifications of synrift intervals have not demonstrated thickening of sedimentary units or divergence of seismic reflections toward footwalls. They therefore conclude that rifting lasted <5 m.y. (probably from the late Berriasian to early Valanginian). Although this conclusion has important implications for, among others, models that attempt to estimate the amount of synrift melting, it is important to realize that it is not conclusive because it is based on negative evidence (i.e., the lack of seismic observations of synrift sediment packages). This lack could be explained, for example, by the inability of seismic profiles to resolve any thin synrift sediments that resulted from a relatively low rate of sedimentation or by the collapse and resedimentation of unstable synrift sediments and the resulting destruction of the original characteristic seismostratigraphic geometry (Wilson et al., in press a).
Here, we consider just the southern Iberia Abyssal Plain segment of the margin. Recently, new seismic refraction results have been published that indicate the thickness and extent of thinned continental crust there (Chian et al., 1999; Dean et al., 2000). This enabled Whitmarsh et al. (2000) to revisit the interpretation of the depth section of profile Lusigal-12 by Krawczyk et al. (1996). Several important new conclusions were reached. First, the ODP cores indicate that all tilted fault blocks on reflection profiles consist of thinned continental (and not oceanic) crust. Second, it is now apparent that the thinned continental crust under Site 901 is only ~6 km thick and that this crust thins westward to ~3 km immediately east of Hobby High. Third, in detail, it appears that reflector H, just east of Hobby High, and the corresponding fault-offset reflector M on the crest of the High represent a tectonic crust/mantle boundary as argued by Brun and Beslier (1996). It is also apparent that several reflections interpreted as low-angle normal faults cut down into the uppermost mantle, thereby indicating that the mantle lay in the brittle regime toward the end of the rifting process. Direct evidence of the tilting of a continental fault block was obtained at Site 1065. Here, downhole Formation MicroScanner logs show that the site was tilted 15° to the southeast in the Middle to Late Jurassic and then 15° to the east in post-Late Jurassic (Early Cretaceous?) time (Basile, 2000).
Although the initial stages of rifting may well be considered to have been dominated by pure shear of the whole lithosphere, it is also clear that in the final stages of rifting leading to breakup, a series of low-angle normal or detachment faults dominated crustal deformation. The model of Brun and Beslier (1996) suggests that initially the ductile lower crust acts as a decoupling (décollement) zone between brittle upper crust and strong upper mantle. As extension proceeds, the original ductile lower crust is preferentially thinned (Brun and Beslier, 1996) and/or the brittle-ductile transition descends because the original upper crust has become thinner, causing the ductile zone between brittle crust and strong mantle to thin also (T.J. Reston, pers. comm., 2000). Eventually, however, the model of Whitmarsh et al. (2000) suggests that the thinned crust and uppermost mantle together lie within the brittle domain because low-angle, and even high-angle, faults are seen on seismic profiles to penetrate the uppermost mantle.