General Location and Physical Geography

A peculiar feature of the geologic history of the Mediterranean region is the intimate association that seems to have existed between the disparate tectonic forces of extension and compression. Within the region are numerous extensional basins that formed coevally with surrounding compression-related mountain belts (Figure 1) (Coward and Dietrich, 1989).

A perfect example of this somewhat odd pairing between extension and contraction is the Betic-Alboran-Rif (BAR) system in the western Mediterranean. The BAR consists of the Betic Mountains in Spain, the Rif Mountains in Northern Morocco, and the Alboran Sea located between the two (Figure 2). The topography of the area is rather subdued with most peaks in the orogenic systems being less than 2km in elevation. Water depths in the Alboran Sea rarely exceed 1km (Watts et al., 1993).

Extension within the Alboran Sea was coeval with south-to-north and north-to-south crustal shortening in the Betic and Rif mountain belts respectively. The following pages provide an overview of the geology within the region and review the various tectonic models that have been proposed in order to explain the kinematic history of the region.

Convergence between Africa and Eurasia likely began some time in the Cretaceous (Dewey et al., 1989). The convergence directions were initially NNE-SSW to N-S before shifting to NW-SE oblique convergence during the Late Miocene time (~10Ma) (Dewey et al. 1989). There has been roughly 200km of N-S convergence in the BAR region from Late Cretaceous-Miocene, followed by ~50km of oblique NW-SE convergence since the Late Miocene (Watts et al. 1993). The convergence between the two plates involved northward subduction of the Tehtyan oceanic crust that resulted in a NE-SW striking collisional orogenic system (Dewey et al. 1989; Zeck, 1999). This collisional system was composed of Paleozoic to Mesozoic marine rocks that record Cretaceous to Paleogene deformation, as well as a few scattered ophiolites (Coward and Dietrich, 1989). The majority of the rocks record low-grade metamorphism (~Greenschist facies) though there are some examples of blueschist, eclogite and granulite metamorphism (Coward and Dietrich, 1989).

Beginning sometime in the Late Oligocene to Early Miocene (~30-20Ma), large extensional basins began forming on the overriding Eurasian plate (Jolivet and Faccenna, 2000). It’s commonly believed that this was due to slab rollback (subduction zone retreat) where the rate of subduction exceeded the rate of convergence between Eurasia and Africa (Malinverno and Ryan, 1986; Royden and Burchfiel, 1989; Royden, 1993; Verges and Sabat, 1999; Zeck, 1999; Jolivet and Faccenna, 2000). These conditions led to a southward migration of the subduction zone, causing extension in the overlying Eurasian plate and a general dispersal of rocks associated with the NE-SW striking collisional system (Royden, 1993;Lonergan and White, 1997; Verges and Sabat, 1999). The subduction zones migrated south carrying the fragments of the former collisional belt as well as the flysch sediments deposited on the Tethyan crust and near the subduction trench (Coward and Dietrich, 1989; Jolivet and Faccenna, 2000). When the retreating subduction zone encountered the buoyant continental crust of the surrounding areas, this material was emplaced onto the continental margins as large nappes (Malvinero and Ryan, 1986; Royden, 1993). It’s not entirely clear however, if this model can adequately explain the kinematic history of the westernmost regions of the Mediterranean (Platt et al. 1998; Zeck, 1999; Jolivet and Faccenna, 2000).

From sedimentary studies and reflection seismology data, extension within the Alboran basin appears to have begun in ~Earliest Miocene (Watts et al. 1993). Seismic reflection data show up to 6-8km of Miocene and younger sediments covering a basement of extended continental crust (Watts et al. 1993). The crust thins from ~35km beneath the Betics to ~15km beneath the center of the Alboran Sea (Watts et al. 1993). The crust of the Alboran Sea is composed primarily of high-grade schists and migmatitic gneisses (Platt et al, 1998). Many of the rocks appear to have undergone significant exhumation (~40km) between 27Ma and 18Ma (Platt et al.1998; Zeck, 1999). Assemblages of metamorphic minerals in the basement rock suggest the initial stage of exhumation involved a significant drop in pressure from 1GPa to ~350MPa (i.e. raising from 40km-13km depths) accompanied by an increase in temperature from 550 degrees Celsius to 675 degrees Celsius (figure 5) (Platt et al., 1998). This is inferred to have occurred over a ~6Ma period from roughly 27-21Ma (Platt et al., 1998). Bioitite Ar/Ar dates and apatite fission track studies show the final stage of exhumation from 13km depth to the surface occurred very rapidly, over a period of roughly 3.5My, from 21-18Ma (Platt et al. 1998). This data has been used as evidence against tectonic models invoking subduction zone retreat (see below).

Both the Betic and Rif orogenic systems have three principal tectonic zones: 1) the Internal Zones; 2) the Flysch Nappes; and 3) the External Zones (Gomez et al. 2000; Lonergan and White, 1997).

The Internal Zones are composed of Paleozoic to Mesozoic rocks that were part of the Late Cretaceous-Paleogene Alpine orogenic system (Lonergan and White, 1997; Verges and Sabat, 1999). These zones are generally metamorphosed to greenschist facies though there are local slivers with amphibolite, granulite, blue-schist and eclogite facies (e.g., the Nevado-Filabride complex of the Betics) (Platt and Vissers, 1989). Many of these rocks record a period of rapid cooling between 27-22Ma (Zeck, 1999). Also present within the internal zones are ultra-mafic peridotites that were emplaced as solid bodies at depths of ~50km and temperatures of ~1000-1200 degrees Celsius (Platt and Vissers, 1989). The emplacement of the internal zones onto the margins of Iberia and Morocco is not well constrained but believed to have occurred during Earliest Miocene times (~23Ma) (Platzman et al., 1993).

The Flysch Nappes consist of Cretaceous to Early Miocene deep-water sediments that were deposited on the Tethyan ocean floor and in basins to the south and west of the Late Cretaceous-Paleogene Alpine subduction zone (Moretti et al. 1991; Lonergan and White, 1997). The rocks are compositionally very mature, suggesting the source area for the sediments was the African Shield (Moretti et al. 1991). During the Oligocene(?)-Early Miocene the Internal Zones were thrust atop the flysch deposits and the two units eventually were emplaced onto the margins of Iberia and Morocco during the Early Miocene (Platzman et al. 1993; Lonergan and White, 1997).

The External Zones consist of Mesozoic to Tertiary sedimentary rocks that were deposited on the shelf and shelf-margins of Iberia and Africa (Lonergan and White, 1997). The external zones are characterized by thin-skinned fold and thrust belts with little to no metamorphism (Coward and Dietrich, 1989). Deformation began in the Early Miocene coincident with the emplacement of the Internal Zones and Flysch Nappes and lasted until the Pliocene (Coward and Dietrich, 1989; Lonergan and White, 1997). Thrusting in the External Zones followed a general pattern of progressive faulting from hinterland to foreland, with displacement vectors showing south, southeast, and southwest directions in the Rif, and north to northwest in the Betics (Lonergan and White, 1997). Parts of the External Rif have experienced up to 200 degrees of Early(?) Miocene counter-clockwise rotation while portions of the External Betics have been rotated clockwise ~130 degrees since the Early Miocene (Platzman et al. 1993). The rotation within these zones cannot be attributed solely to oroclinal bending (Platzman, 1993).

Volcanism in the BAR region can be divided into three general suites based on geochemistry and times of activity (Lonergan and White, 1997; Gomez et al. 2000). The oldest (~18-8Ma) volcanic rocks are calc-alkaline and are present mainly in the center of the Alboran Basin (Lonergan and White, 1997). High-K volcanics are the second broad suite of volcanics that occurred in the BAR region. These rocks have a wide variety of compositions, from shoshonitic to lamproitic, and were erupted between ~9-4Ma (Lonergan and White, 1997). The most recent suite of volcanics are the alkali-basalts that erupted from ~5Ma to ~1Ma (Lonergan and White, 1997). This type of volcanism tends to occur in the more external parts of the Rif and Betic orogenic systems and is often associated with extension (Wilson and Bianchini, 1999; Lonergan and White, 1997).

Many different models have been proposed in an effort to explain how semi-radial thrusting around Gibraltar could have occurred coevally with extension in the Alboran Sea. As more and more geological and geophysical data is collected from the region, the number of viable hypotheses has been reduced. This section reviews some of the more popular models.

Late Cretaceous through Paleogene convergence between Africa and Eurasia produced a zone of thick crust (~Internal Zones) that was located in the area between present-day Morocco and southern Spain. This collisional ridge was underlain by a thick, cold, and gravitationally unstable lithosphere mantle. Removal of the lithospheric root by thermal convective processes in the Late Oligocene-Early Miocene produced a significant increase in the elevation and potential energy of the collisional ridge. The large increase in potential energy led to substantial horizontal stresses that resulted in extension and exhumation of lower crustal rocks (e.g. peridotites within the Internal Zones). The radial horizontal stresses, combined with the continued convergence of Africa and Eurasia, were accommodated by shortening within the External Betic and Rif mountain belts throughout the Miocene.

The delamination model initially proposed by Garcia-Duenas et al. (1992) and Seber et al. (1996) shares many similarities with the convective removal hypothesis of Platt and Vissers (1989). The main difference between the two models centers on how the lithospheric root was removed from the crust. Lithospheric thickening during the Late Cretaceous through Paleogene resulted in a large, gravitationally unstable lithospheric root beneath a NE striking zone of thickened crust. Possibly induced by some convective removal, this root mechanically detached from the crust and was replaced by hot aesthenosphere. The lithosphere peeled back to the west and northwest where it may still be attached beneath portions of Spain (Calvert et al. 2000; see below). The loss of the lithospheric root, along with the inflow of hot aesthenospheric material, resulted in heating of the crust, uplift and extension in the Alboran area. The horizontal forces caused by extension, combined with those forces resulting from the convergence of Africa and Eurasia, were accommodated by shortening within the Betic and Rif orogenic systems.

The ‘retreating-slab subduction’ or ‘slab rollback’ model of Royden (1993) and Lonergan and White (1997) differs significantly from the delamination and convective removal hypotheses. In the area between Spain and Northern Africa the Late Cretaceous-Paleogene subduction zone had a NW-SE to N-S orientation and involved NE to E dipping subduction of the Tethyan oceanic crust. As the rate of subduction (driven by slab-pull forces) exceeded the rate of convergence, the subduction zone began migrating to the west sometime in the Late Oligocene-Early Miocene. As the subduction zone migrated westward, the crust of the former collision ridge (the Internal Zones) was broken up and dispersed while the crust immediately behind this region was extensively thinned. In the northern and southern portions of the subduction zones, the westward movement was slowed as the buoyant continental crust of Iberia and Africa was encountered, while in the central portion of the subducting trench, westward migration continued. This resulted in northward emplacement of the Internal Zone onto southern Iberia and southward emplacement of the Internal Zone onto Northern Africa. This process, similar in some respects to obliquely approaching water waves refracting and shoaling perpendicular to shore, is suggested by paleomagnetic data that shows significant Early (?) Miocene rotations in the Betics and Rifs (Platzman 1993). In both Southern Iberia and Northern Africa the emplacement and compression was accommodated by shortening in the continental margins (i.e. the External Zones), while the continued subduction zone rollback was accommodated in the overriding plate by extension beneath the Alboran Sea. Proponents of the rollback model also point out that the composition and timing of the volcanic activity in the BAR region closely matches the volcanic history of nearby regions that experienced slab rollback (e.g. Calabrian Arc and Tyrrhenian Sea) (Lonergan and White, 1997).

Blanco and Spackman (1993) and Zeck (1999) proposed a model whereby break-off of a NW dipping subducted slab beneath Iberia produced large differences in potential energy and lead to the extension and compression seen within the BAR region. There are two important problems with this model: 1) The total convergence between Africa and Eurasia (<200km) is considerably less than the length of the subducted slab interpreted to have broken off (~450km); and 2) The high P-wave velocity zones in the upper mantle do not correspond to a NW dipping slab (the zone appears to dip to the east; see tomography results below). Andreiux (1971) (as reported in Platt and Vissers, 1989; and Lonergan and White, 1997) proposed a rigid, westward-moving Alboran micro-plate could have caused the radial thrusting around the Gibraltar region. The micro-plate hypothesis was abandoned as subsequent studies showed the large amounts of extension and relatively weak and thin crust that is present within the Alboran Sea.

Recent geophysical work has provided detailed information on the crustal and subcrustal properties of the western Mediterranean. Much of the geophysical data support the existence of shallow, hot, aesthenospheric material beneath the Alboran Sea and portions of the Betic and Rif system. Using data from over 1200 earthquakes, Seber et al. (1996) found a significant gap in earthquake activity beneath the Alboran Sea between depths of 20 and 60km. Intermediate seismic activity is present in the area from depths of 60km to 150km. Beneath the Betics, seismic activity is abundant and present at both shallow and intermediate depths, while the relatively limited amount of seismic activity beneath the Rif is restricted to shallow depths (Figure 7a).

Seismograms from over fifty earthquakes that occurred in the BAR region show there is strong P-wave and S-wave attenuation in the crust and upper mantle of this region (Figure 7b). The seismograms from stations near the Alboran sea showed significant S-wave attenuation (along with moderate P-wave attenuation) in both the crust and along the crust mantle boundary. The zone of high crustal and uppermost mantle attenuation coincides with the region containing the aseismic zone at 20-60km depths.

A large (up to –10mgal) negative Bougher gravity anomaly exists in a horseshoe like pattern around the Betic and Rif mountain belts. Three-dimensional modeling of this anomaly suggests the zone has low density material deep below the surface, which is hypothesized to be hot aesthenospheric material. The heat flow in the Alboran region is consistent with the presence of shallow aesthenospheric material.

Over 7000 events recorded by 96 stations were analyzed using local, regional and teleseismic inversion methods to determine the velocity structure of the lithosphere and upper mantle beneath the BAR region. A low velocity zone exists beneath the Alboran Sea that roughly corresponds to the aseismic, P- and S-wave attenuating zone described by Seber et al. (1996). Beneath this low velocity zone, in an E-W vertical slice, is a NE-SW striking, SE dipping high velocity body that extends from roughly 100km to 350km depth. An additional high velocity body is located at ~600km depth and matches well with the locations of deep earthquakes recorded in the area. Viewing a N-S vertical slice through the region shows the high velocity body appears to be attached to Iberia (in parts) and extends from Spain to ~350km below northern Morocco.

From their own findings and the work of previous studies, Clavert et al. (2000) interpreted the high velocity bodies imaged beneath the BAR region to be delaminated lithosphere. Calvert et al. (2000) noted that the high velocity body could be interpreted as a subducted slab (which would support the slab rollback model), but favored the delaminated lithosphere model because of petrologic work by Platt et al. (1998) which suggested the process of slab rollback would be incapable of producing the PTt paths found in the metamorphic assemblages in Alboran basement rock.

The model shows lithospheric thickening during the Paleogene, leading to delamination beginning during the Late Oligocene and Early Miocene. The lithosphere was removed from beneath the area of the Alboran Sea and slowly peeled back to the west or northwest. The base of the crust was heated as hot aesthonospheric material filled the void created by the removal of the lithosphere. Uplift and extension of the Alboran Sea area followed the delamination while thrusting occurred in the Betic and Rif orogenic systems due to the horizontal forces associated with the extension in the Alboran and the convergence of Africa and Eurasia. The location of the delaminated lithosphere coincides with the location of the intermediate depth earthquakes (100-350km) beneath the Alboran Sea. Many of the shallow earthquakes beneath the Iberian Peninsula might be related to continued, but slowed, delamination.

Currently there are three principal tectonic models that attempt to explain the evolution of the Bar region, each having its own strengths and weaknesses. Listed below are some of the problems that exist with each of the individual models.

1)Using computer modeling an assuming an initial lithospheric thickness of 250km, Platt et al. (1998) argued that the processes associated with slab rollback would be incapable of producing the PTt paths observed in the Alboran basement rock.

2)If the subduction zone was migrating to the west, where is it now? Royden (1993) suggests it may have migrated as far west as the Horseshoe Seamounts, while Lonergan and White (1997) hypothesized migration stopped near present-day Gibraltar.

1)Large (~130 degrees), clockwise rotations about a vertical axis have occurred in thrust sheets within the Betic orogenic system (Lonergan and White, 1997). Counterclockwise rotations about a vertical axis of up to ~100 degrees have occurred in rocks within the Rif orogenic system (Platzman et al. 1993). The rotations in both of the systems are post-Oligocene and are not the result of simple oroclinal bending (Platzman, 1993). Delaminantion and convective removal models do not adequately address these observations.

2)It would seem that removal of the lithosphere should lead to radial deformation - so what occurred on the eastern margin of the BAR region? Looking at this area in plan-view, there is thrusting to the north, south, and west, what processes were acting on the eastern side?

3)Lonergan and White (1997) argue that removal of the lithosphere in the area may not have been able to produce significant potential energy differences. Though this is debatable, it does bring up the point that very little is known about the key process in the delamination and convective removal models.