1 is near vertical and minimum stress direction 3 strikes N80ºE).
This section summarizes the results from the four drill sites of Leg 161 (Fig. 2, Fig. 3, Fig. 4) that aid in better understanding the tectonic history and origin of the Alboran Basin.
Site 976 (Shipboard Scientific Party, 1996b) was drilled in the West Alboran Basin. It is about 110 km east of the Strait of Gibraltar and 8 km northeast of DSDP Site 121, drilled during Leg 13 in 1970 (Ryan, Hsü et al., 1973). A seismic survey (Klaus and Shipboard Scientific Party, 1996) was carried out from JOIDES Resolution to verify site location and Site 976 drilled at the intersection of JOIDES Resolution single-channel seismic profiles and MCS line ALB-39 (Fig. 2), 8 km northeast of DSDP Site 121 (see Fig. 16). Site 976 lies on a basement high correctly predicted as being formed of metamorphic rocks exhumed during the rifting; it composes the most prominent crustal structure in the Alboran Basin (Fig. 4). The basement was penetrated to a maximum depth of 260 m at Hole 976B and to 85 m at Hole 976E. A total of 343.19 m of high-grade metamorphic rock was cored in the two holes (Fig. 3, Fig. 5), averaging 27.8% recovery of hard rocks.
A full suite of logging tools was run in Holes 976B and 976E (quad-combo, FMS, BHTV, and geochemical), yielding information on unrecovered intervals, on the basement/sediment transition in Hole 976B, and on fault-gouge zones within the basement. Thermal gradient and in-situ sediment conductivity measurements from logging (ADARA and water-sampling temperature probe [WSTP] data), conducted in both holes, indicated a heat-flow of 102 mW m-2, which is in good agreement with average heat-flow values in the West Alboran Basin (Polyak et al., 1996).
Drilling in Holes 976B and 976E sampled the entire sedimentary cover above the metamorphic basement in a sector wherein seismic images suggest that Messinian seismic Unit II is practically absent. The stratigraphic sequence consists of 650 m of late Serravallian to Holocene marine sediments (Fig. 3), including hemipelagites, and muddy turbidite and gravity-flow facies (Alonso et al., Chap. 4, this volume). The Miocene/Pliocene boundary is ~572 mbsf, and the Pliocene Pleistocene boundary is about 360 mbsf. On the basis of new biostratigraphic data (from Nannofossil Zones [NN]), Siesser and de Kaenel (Chap. 16, this volume) refined the ages of the two major hiatuses identified in the sedimentary sequence in Hole 976B by Shipboard Scientific Party (1996b). They determined that the hiatuses are within the late Miocene (lower Tortonian [NN9 to NN10], from about 10.2 Ma to 8.6 Ma;) and between the early Pliocene and late Pliocene (Zanclean and Piacenzian [NN13 to NN16] from about 5 Ma to 2.5 Ma). The late Tortonian and Messinian? (NN11) and earliest Pliocene (NN12) intervals, according to these authors, have a total thickness of about 66 m and 57 m, respectively; they consist of open-marine fine-grained sediments. Average sedimentation rates are low for the late Miocene (15 m/m.y.), and particularly high for the earliest Pliocene (453 m/m.y.), late Pliocene (340 m/m.y.), and Pleistocene-Holocene (208 m/m.y.; Shipboard Scientific Party, 1996b).
In both holes, uppermost Serravallian marine deposits (NN7, 11-12 Ma) overlie the brecciated basement (lithostratigraphic Unit IV; Shipboard Scientific Party, 1996b). In Hole 976B, the basal sediments (lithostratigraphic Unit IV, equivalent to seismic Unit IV; Fig. 3) consist of interspersed pebbles to sand-sized clasts of metamorphic rocks, similar to the underlying basement lithologies, with mixed deep- and shallow-water fauna. In Hole 976E, basal sediments above the basement are formed of a 15-cm-thick sandy-silty claystone interval. The Serravallian sand includes significant volcanic components, as deduced from a high plagioclase content, relict pyroclastic textures in sand grains, and a high percentage of zeolite alteration minerals. It has therefore been interpreted as epiclastic deposits likely associated with a previous or near-coeval pulse of volcanism in the basin (Marsaglia et al., Chap. 3, this volume).
Subsidence analyses (Fig. 6) at Site 976 and at two commercial wells (Granada-D1 in the Granada Basin, and Andalucía-A1, Fig. 4; see Fig. 18) provide some constraints on the subsidence and uplift history across the basin (Rodríguez-Fernández et al., Chap. 5, this volume). The backstripped diagram at Hole 976B has been calculated using porosity, density, sedimentary, and biostratigraphic data from Shipboard Scientific Party (1996b); backstripping at the commercial wells consider paleobathymetric corrections from new sedimentary and biostratigraphic studies on well samples. Steep sections of the diagram at Site 976 reveal two sudden periods of subsidence in the late Serravallian (from 11 to 10.7 Ma) and during the late Pliocene and Holocene (from 2.5 Ma onwards), and uplift of the metamorphic basement high during the late Pliocene (from 5 to 2.5 Ma). According to these results, total subsidence at Site 976 is on the order of 900 m during the late Serravallian (subsidence rate of about 3 km/m.y.) and on the order of 1200 m during the late Pliocene to Holocene (subsidence rate of about 480 m/m.y.).
In order to maximize drilling results, regional seismic reflection data from the vicinity of Site 976 have been pooled for postcruise research. To this end, seismic analyses have been carried out on the JOIDES Resolution lines (Lines 2S-1 and 2S-2) and MCS reflection profiling (5-8 s, TWT penetration) around Site 976 (Fig. 4, Fig. 7, Fig. 8, Fig. 16).
Wireline logging and biostratigraphic data from Hole 976B have been tied to JOIDES Resolution line 2S-1 by synthetic seismograms (Tandon et al., 1998). This study has distinguished several key reflectors within the sedimentary cover: the boundaries between seismic Subunits Ia (Pleistocene-Holocene), Ib (late Pliocene), and Ic (early Pliocene), and reflectors M and R3 (Fig. 3, Fig. 7).
The structural map in Figure 4, drawn from a dense grid of MCS lines, shows the complex structural pattern of the Alboran Basin, comprising extensional and contractional features. Interpreted MCS reflection profiles cutting across the entire basin are presented in the back-pocket foldout (see Fig. 16, Fig. 17, Fig. 18, Fig. 19). The structural high (referred to as the Site 976 High hereafter) is located in a fault-bounded basement forming a rounded-arrowhead outline pointing west. It trends northeast-southwest at the site position, turning northwest-southeast toward the south and east-west toward the north (Fig. 4). The whole length of its convex flanks is normal-faulted, bounding a complex graben filled with lower Miocene (seismic Unit VI, late Aquitanian?-Burdigalian, 22?-19 Ma) to Pleistocene sequences. The shape of the entire graben, which comprises the main depocenter in the West Alboran Basin and a northern prolongation flanking the Spanish shelf, is also arcuate, paralleling the trend of the Site 976 High from southeast of Almería to the Xauen Bank (Fig. 4). Remarkably, this main depocenter is the only one in the Alboran Sea basin in which the basal seismic Unit VI occurs (Fig. 4, Fig. 16, Fig. 17). In those areas where seismic Unit VI reaches its maximum thickness, that is, in the central West Alboran Basin, an extensive province of mud diapirs has developed. According to seismic images, the diapirism began by the middle Miocene and large diapirs occupy a huge part of the depocenter; the diapirs are rooted in the basal olistostromic seismic Unit VI. Diapirism likely resumed by the Pliocene and the top of some diapirs then reached nearly to the seafloor (Comas et al., 1992, 1993; Maldonado et al., 1992; Pérez-Belzuz et al., 1997).
In the northern branch of the major-graben depocenter (hereafter called the Málaga Graben; Fig. 4, Fig. 8, Fig. 9, Fig. 10, Fig. 16, Fig. 17) sediments attain up to 5 km in thickness just facing the drill site (according to depth conversions); notably, toward the southwest, sediment thickness is up to 7-8 km in the center of the West Alboran Basin and up to 4 km to the south near Xauen Bank in the southern Alboran Sea (Soto et al., 1996; Chalouan et al., 1997).
The contour map of the top of the basement (Fig. 10) reveals the complex and structurally asymmetric character of the Málaga Graben. Both flanks of the graben correspond to intricate interference of former submarine or formerly emerged surfaces, and low- and high-angle normal faults, oblique transfer faults, and tops of rollover structures affecting the basement (de la Linde et al., 1996). Seismic section across the Málaga Graben through Site 976 (Fig. 8, Fig. 16), and further to the east (Fig. 17), show that individual reflectors within the sedimentary cover diverge with different attitude toward the steeply dipping border faults, indicating that faulting was active at both flanks of the graben at somewhat different times. At places, the contact between the sedimentary cover and the basement likely accommodated normal faults (Fig. 7, Fig. 8, Fig. 16). The seismic images suggest that the northern flank of the graben was relatively uplifted at some time between the late Tortonian (age of reflector R3) and the Pliocene, as Pliocene sediments overlie middle Miocene deposits (seismic Unit IV) near the Spanish coast.
Comas and Soto (Chap. 25, this volume) interpret the structure in the northern flank of the Málaga Graben (Fig. 10, Fig. 16) as likely comprising horsetail faults of a listric-fan rooted on deeper listric (intra-crustal) detachments (oriented with the top to the south-southeast), considering that the controlling low-angle master fault (southeast-dipping detachment system) is located in the northern basement slope of the graben. These authors suggest that the Site 976 High has derived from a initial (early Miocene) rollover anticline, subsequently faulted by counter faults, therefore representing a central high (following Gibbs, 1984) in the Alboran Basin.
Single-channel Line 2s-1 (Fig. 7) images normal faulting on the crest and flanks of the Site 976 High, consistent with data indicating brittle deformation of the basement (see below). In the northwest flank of the high, individual reflectors within sediments of seismic Unit IV are synsedimentary normal faulted, confirming that extension continued till the middle Miocene, as demonstrated by drilling (Langhian age of lithostratigraphic Unit 4; Shipboard Scientific Party, 1996b). According to seismic images, Messinian Unit II is missing at the top of the basement high and further to the southeast. The M-reflector, either an erosional or nondepositional surface, correlates with the base of early Pliocene seismic Subunit Ic (Tandon et al., 1998). The anticline attitude of Miocene and early Pliocene (seismic Subunit Ic) sediments, as well as the unconformity at the top of Subunit Ic, indicate that the whole packet (basement through Unit II) was upwarped during Subunit Ic times (i.e., during the early Pliocene).
Statistical measurements on processed FMS and BHTV electrical images from Hole 976B (de Larouzière et al., Chap. 24, this volume) indicate that low-angle surfaces within the basement, probably representing the metamorphic foliation, tend to dip to the west; steeper surfaces, interpreted as brittle fractures or faults, are mostly east dipping. Narrow intervals with sharp changes in hole geometries can be ascribed to active faults, also recognizable in BHTV images. Mapping of these faults (up to 20 active faults at Hole 976B) leads de Larouzière et al. (Chap. 24, this volume) to propose a present-day stress regime in the area with a relative east-west direction of extension (maximum stress direction 1 is near vertical and minimum stress direction 3 strikes N80ºE).
The upper 124 m of cored basement at Hole 976B consists of biotite-sillimanite-plagioclase high-grade schist, with visible porphyroblasts of garnet and andalusite. This metapelite sequence has interlayered calcite and dolomite marble and associated calc-silicate rocks as reaction bands, which developed along the metapelite-metacarbonate contacts (López Sánchez-Vizcaíno and Soto, Chap. 18, this volume). Metacarbonate and associated rocks are particularly abundant in the upper part of the cored section (Fig. 5). This metapelite sequence overlies banded pelitic gneiss with large porphyroblasts of coarse-grained K-feldspar, cordierite, and andalusite. Much of the gneiss is migmatitic, with centimeter-thick veins and segregations of weakly foliated or unfoliated felsic material (leucosomes) containing large crystals of cordierite, biotite, andalusite, and sillimanite. The facts that leucosome geometries largely follow pre-existing planes of anisotropy, that the volume of leucosome veins (<20%) is low with respect to the whole rock volume, and that the common occurrence of mafic selvages (melanosomes) are associated with them, suggest that the low degree of melting was controlled by strong, compositional foliation (Soto and Platt, in press).
Major and trace compositional trends of the gneiss are similar to those of the high-grade schist (Spadea and Prosser, Chap. 28, this volume), indicating a local source for melting and relatively low mobility of melts during migmatite formation (Soto and Platt, in press). Both rock units are cut by peraluminous, leucogranitic dikes. The high-grade schist and the gneiss show differences in their petrological and structural evolution; at Hole 976B, they are separated by a significant fault zone (Fig. 5), according to BHTV and FMS images (Comas and Soto, Chap. 25; de Larouzière et al., Chap. 24, both this volume).
Two systematically identifiable sets of ductile fabrics and structures can be distinguished in the high-grade schist, referred to as D1 and D2 by Shipboard Scientific Party (1996b). The earlier fabric, S1, is characterized by compositional layering and biotite-rich laminae and oriented biotite. S1 has been intensively affected by 1-50 mm-scale tight, asymmetrical D2 folds. Biotite is recrystallized in the fold hinges and tends to lie parallel to the axial surface, as do elongated mats of fibrolite, the two together defining an axial planar fabric, S2. In much of the high-grade schist, D2 has been so strong that the main foliation is in fact a composite fabric formed by the transposition and modification of S1. D2 was followed by static crystallization of andalusite, plagioclase, a second garnet, and K-feldspar. The gneissic foliation in relation to the deformational history of the schist is uncertain. On the basis of textural relationships, it is a composite foliation, which may have initially formed at about the same time as D2 in the high-grade schist but was subsequently heavily modified by compaction or further deformation during partial melting (Soto and Platt, in press).
The P-T-t evolution of the high-grade schist and gneiss is shown in Figure 11A (Soto and Platt, in press). The most complete mineral assemblages in these rocks are:
As reaction textures among the mineral phases, and in particular between andalusite and sillimanite, are always present in the rocks, the above-mentioned mineral assemblages should be considered disequilibrium assemblages. Nevertheless, the high-grade schist assemblage reflects lower peak-temperatures than the gneiss assemblage. Thermobarometric estimates on the high-grade schist give approximate P-T conditions of 3.5-5.5 kbar at 650°-700° C for the matrix assemblages (Soto and Platt, in press; Prosser et al., Chap. 20, this volume). The gneiss achieved higher temperature during its metamorphic evolution, with approximate melting conditions of 700°-750°C, at 6 kbar to 3 kbar (Soto et al., Chap. 19, this volume).
The first stages of the P-T evolution of the high-grade schist are only preserved in the cores of garnet porphyroblasts and are characterized by prograde evolution through the temperature range 500º-600°C, accompanied by a slight drop in pressure, from 10.5 kbar to 8 kbar. Subsequently, the main deformation event took place along a decompression P-T path, concomitantly with a temperature rise, achieving final conditions of 650º-700°C and 3-4 kbar. Overall, a temperature increase of 50°C to 100°C occurred during decompression, from more than 8 kbar to less than 4 kbar (Soto and Platt, in press).
The decompression P-T paths of the high-grade schist and gneiss were followed by a cooling path of up to T < 500º-600°C and P < 2-3 kbar. During this evolution, crystallization of residual granitic melts occurred under overstepped conditions in the andalusite stability field. Ar/Ar dating on muscovite and biotite indicates that cooling took place at 20.0 ± 0.2 Ma (±2) for muscovite and 19.2 ± 0.7 Ma (±2) for biotite (Kelley and Platt, Chap. 22, this volume). Apatite fission-track analyses on the same rocks indicates cooling below the apatite partial annealing zone (60°-120°C) at 18.3 ± 1.0 Ma (± 2; Hurford et al., Chap. 21, this volume). All these data suggest that during the early Miocene the basement cooled rapidly from 426 ± 22°C (the estimated closure temperature of muscovite) to 60°C in a period of between 0.5 m.y. and 2.9 m.y., which gives an average minimum cooling rate of 126°C/m.y.
Platt et al. (in press) present the results of thermal calculations carried out to explain the P-T path determined by Soto and Platt (in press) for the high-grade schist sampled at Site 976. As mentioned above, the essential features of this path (Fig. 11A) are that the rocks reached pressures of up to 10.5 kbar at ~500°C relatively early in their history, corresponding to a burial depth of about 40 km, followed by decompression and heating to a peak temperature of 650º-700°C at a pressure of about 3-4 kbar, and then by cooling at low pressure.
Variables considered in the modeling included the thickness and thermal gradient of the post-orogenic lithosphere, the radiogenic heat production in the thickened crust (assumed to be 60 km thick); the time gap (pause) between the end of contractional tectonics and the start of extension, the effects of convective removal or delamination of lithospheric mantle at depths of 125, 75, or 62.5 km, and the rate of extension.
The only combinations of variables that produce modeled P-T paths (Platt et al., in press) with the observed characteristics involve high radiogenic heat production combined with a significant post-contractional pause (to produce high temperatures in rocks initially at 40 km depth), removal of mantle lithosphere below 62.5 km (to produce further heating during decompression), exhumation by extension at a very high rate (about 4.5 km/m.y., to delay the onset of cooling until the rocks reached shallow depths), and final exhumation and cooling at a rate of 4 km/m.y. (to satisfy radiometric and petrological constraints). This results in a total time of about 9 m.y. for exhumation from 40 km depth to the surface. Apatite fission-track analysis (Hurford et al., Chap. 21, this volume) indicates that the rocks cooled below 60°C at 18.3 ± 1.0 Ma, so the thermal modeling results suggest that extension started at about 27 Ma.
Faulting, brecciation, and fracturing characterize a brittle deformation at the top of the basement in both Holes 976B and 976E. Intervals of polymictic breccia and clay-rich fault-gouges in Holes 976B and 976E mark zones of brittle faulting and their widespread occurrence largely determined the low total recovery (Fig. 5).
Comas and Soto (Chap. 25, this volume) report that brittle deformation was mainly a result of fault-related cataclasis, as most of the breccia and fault gouge intervals are cataclasites. The cataclastic breccias consist of consolidated rocks with angular and highly variably sized clasts (<60%) from nearby basement lithologies, a matrix formed by comminution of the metamorphic clasts, and a dolomite-rich cement with abundant displacive and replacive textures. The fault gouges comprise angular fragments of basement rocks in a clay-rich, dark matrix (>70%). Cataclastic structures in breccias and gouges include cataclastic foliation, cataclastic flows, and cataclastic lineation. Extensive internal microfracturing (splintered grains) in metamorphic clasts has been interpreted by these authors as being caused by tectonic extension, decompression and fast exhumation of the metamorphic basement. Open fracturing in breccias is indicated by sediment-fill fractures. Neptunian dikes and sills filled with late Serravallian sediments in the uppermost basement cores from Holes 976B and 976E prove that dilation and brittle tensional fracturing in the top of the basement occurred in a submarine environment. Sediment injections into open fractures and pervasive dolomitization in cataclastic fabrics indicate fluid-assisted processes for brittle deformation, probably including heat advection associated with the tectonic regime (Comas and Soto, Chap. 25, this volume). A major breccia and gouge interval (20-25 m thick) at Hole 976B, coinciding with a noticeable change in basement lithology, corresponds to a thick fault zone separating the higher sequence of high-grade schist and marble from the lower part of the section comprising mainly pelitic and migmatite gneiss (Fig. 5). According to seismic interpretation, this cataclastic zone is probably a westward-dipping brittle fault (Fig. 7).
Sites 977 (Shipboard Scientific Party, 1996c) and 978 (Shipboard Scientific Party, 1996d) were drilled in the East Alboran Basin, south of Cabo de Gata. Site 977 lies the south of Al-Mansour seamount at a secondary depocenter bounded by the Yusuf Ridge, and Site 978 is just 24 km north of Site 977, to the north of Al-Mansour Seamount. Both sites lie in the same east-west-trending, 70-km-wide, main depocenter that connects the Alboran and South Balearic Seas and that opens to the South Balearic Basin (Fig. 2, Fig. 4). One single-channel seismic profile (Line 3S; Fig. 12) was acquired by the JOIDES Resolution (Klaus and Shipboard Scientific Party, 1996) to verify site location, and Site 977 drilled at the intersection of this line and MCS line Conrad 823 (Fig. 19).
Sites 977 and 978 sampled 598.5 m and 485 m, respectively, of equivalent sedimentary sequences. The topmost 213 m of sediments at Site 978 were deliberately washed out (because of the tectonic interest in penetrating as deep as possible), but a continuous sequence from 213 to 698 mbsf of sediments was recovered downhole.
The lowermost fine-grained sediments drilled at Site 977 (at about 530 mbsf) are earliest Pliocene in age (NN-12b; Siesser and de Kaenel, Chap. 16, this volume). Scraping from calcareous cement surrounding pebbles of gravel sampled in the lowermost cores at Site 977 (see below) yielded nannofossil assemblages containing middle and upper Miocene fauna (NN7-NN11; Siesser and de Kaenel, Chap. 16, this volume).
At Site 978, the Miocene/Pliocene boundary is proposed at about 620 mbsf and the lowermost cores at this site, just underlying the gravel interval (see below), have been assigned to the Messinian or latest Tortonian (NN11; Siesser and de Kaenel, Chap. 16, this volume). Iaccarino and Bossio (Chap. 42, this volume) assigned an early Messinian (pre-evaporitic interval) age to foraminifers in these cores. However, they leave open the possibility of early Messinian faunas having been reworked; therefore, they suggest the interval may be latest Messinian in age. The Pliocene/Pleistocene boundary is at about 266 mbsf at Site 977 and at about 223 mbsf at Site 978 (Fig. 3).
A brief early Pliocene hiatus at about 490 mbsf at Site 977 (between 5 and 4 Ma, lasting less than 1.0 m.y.) is reported by Siesser and de Kaenel (Chap. 16, this volume); and interpreted by Tandon et al. (1998) as a seismic discontinuity. Average sedimentation rates at this site are 400 m/m.y. for the lowermost cores just below the early Pliocene hiatus, 98 m/m.y. for the late-early Pliocene to late Pliocene, and 154 m/m.y. for the Pleistocene-Holocene (Shipboard Scientific Party, 1996c). Average sedimentation rates at Site 978 are 159 m/m.y. for the late Miocene, 120 m/m.y. for the early Pliocene, 111 m/m.y. for the late Pliocene, and 127 m/m.y. for the Pleistocene (Shipboard Scientific Party, 1996d).
According to the late Messinian age assigned to the 74 m of sediments underlying the gravel bed at Site 978 (Iaccarino and Bossio, Chap. 42, this volume), the probable age of this gravel is Pliocene or latest Messinian. Logging data, acquired only at Site 977 (quad-combo and FMS), provided excellent images of the gravel interval in the lower part of the hole (downhole from 531 mbsf), in which the core recovery was extremely low. No postcruise work on these logging data has been performed yet.
Pebbles from the aforementioned gravel interval sampled at Sites 977 and 978 consist of a collection of mixed volcanic rocks, accompanied by minor metasedimentary (chert, dolomite) and metamorphic (quartzite, quartz-schist, schist) pebbles. Some of the volcanic clasts are coated by a calcareous sedimentary cement suggesting that the gravel was derived from a partly cemented sandy conglomerate. Volcanic lithotypes in the gravel clast encompass basalt, andesite, dacite, rhyolite, meta-basalt, meta-rhyolite, and meta-dacite.
Hoernle et al. (Chap. 27, this volume) have carried out the 40Ar/39Ar dating and the major- and trace-element geochemical analysis of the volcanic pebbles. The samples range from basalts to rhyolites, belonging to the tholeiitic, calc-alkaline, and shoshonitic series. Plagioclase phenocrysts from dacites yield an isochron age of 6.4 ± 0.3 m.y., and a mean apparent age of 6.1 ± 0.3 m.y. Sanidine phenocrysts from rhyolites give an apparent age of 9.25 ± 0.02 m.y. Amphibole phenocrysts from basalts yield apparent ages ranging from 8.8 ± 2.9 m.y. to 11.8 ± 1.6 m.y., with an isochron age of 9.95 ± 0.64 m.y. and a mean apparent age of 9.90 ± 0.40 m.y. Plagioclase phenocrysts from dacite give apparent ages from 11.1 ± 1.3 m.y. to 13.8 ± 0.3 m.y. Other plagioclase crystals from dacite indicate an apparent age of 12.1 ± 0.2 m.y. Pebble samples, ranging from basalt to rhyolite, are attributed to tholeitic, calc-alkaline, high-K calc-alkaline, and shoshonitic series. Basalts and basaltic andesites have been divided into two groups on the basis of their rare-earth element (REE) pattern: (1) the LREE (lithophile and light REE) -depleted groups, with patterns similar to those observed in normal (N) and enriched (E) MORB (Mid-Oceanic Ridge Basalts), and (2) the LREE-enriched group, with patterns characteristic of OIB (Ocean-Island Basalts). The andesites, dacites, and rhyolites generally denote LREE-enrichment ([La/Yb]N 
1), while the more evolved rock types have negative Eu anomalies, reflecting depletion of the MREE (medium REE). Among the basalt samples, the LREE-enriched basalts are characterized by a relative enrichment in Th and U and depletions in Rb, Ba, K, Zr, and Ti. In contrast, the LREE-depleted basalts show a relative enrichment in Rb, U, K, Pb, and Sr and a relative depletion in Th, Nb, and Ta. These basalt groups also display distinct immobile incompatible element ratios, with the LREE-enriched group having lower Ta/Nd, Sm/Nd, and Zr/Nb, but higher Ta/Yb, Th/Yb, Sm/Yb, Gd/Yb, and La/Nb than the LREE-depleted group.
According to Hoernle et al. (Chap. 27, this volume), fractionation of plagioclase, clinopyroxene, titanomagnetite, olivine, apatite, and amphibole can explain most of the observed variations in the major and compatible trace elements. On the basis of the presence of xenocrysts and the large range in Pb concentration in dacites and rhyolites (1.5 to 66 ppm, occasional samples having Pb of >15.5 ppm), they suggested that crustal contamination was an important process conditioning the chemistry of the most evolved magmas. Hoernle et al. (Chap. 27, this volume) conclude that the calc-alkaline affinities and the incompatible element systematics are characteristics of subduction zone volcanism, and that the immobile trace-element data indicate that the Miocene (6 to at least 12 Ma) asthenosphere beneath the Alboran region contained both depleted (MORB-type), peridotitic component and an ocean-island (OIB) or plume-type (HIMU) component, possibly in the form of garnet-pyroxenite layers.
On the basis of the sediment ages constrained at Sites 977 and 978, seismic interpretation around the sites confirms that many bathymetric depressions and post-Messinian depocenters in the eastern Alboran Sea region are very recent wrench-related transtensional structures (Fig. 4). The M-reflector (base of the Pliocene-Pleistocene sequence, sampled at Sites 977 and 978) embodies a key reference for structural interpretations (Fig. 16, Fig. 17, Fig. 18, Fig. 19).
JOIDES Resolution Line 3S (Fig. 12) and Conrad line 823 (Fig. 19) image the Yusuf right-lateral strike-slip fault system (Mauffret et al., 1987). The Yusuf Basin, containing sediments up to 2000 m thick, is bounded by the Yusuf Fault to the northeast. The Yusuf Fault, with an apparent vertical throw of >2 km, is interpreted as being the master-fault of the system. The rhomboid-shaped Yusuf Basin is a negative flower-structure (pull-apart-type, transtensive basin), and the adjacent Yusuf Ridge (900 m high) has been interpreted as a positive flower-structure (compressive feature) within the Yusuf strike-slip system (Alvarez-Marrón, Chap. 26, this volume).
Striking images appear in the northeastern half of Line 3S. The M-reflector (placed at the bottom of Hole 977A, Fig. 12) images a channeled erosional-surface cutting a faulted and slightly folded sedimentary sequence of unknown age. Over the M-reflector, seismic Subunit Ic sampled at Site 977 (68 m thick interval, early Pliocene) seems to be part of a thicker-to-the-northeast lower Pliocene sequence in the East Alboran Basin. At the northeastern end of the seismic line, the lower Pliocene reflections represent ponding or onlapping sediments upon the erosional M-reflector surface; however, these reflections (Subunit Ic) show synsedimentary folding very near Site 977. Therefore, the reported brief hiatus within the lower Pliocene cores at Site 977 (Siesser and de Kaenel, Chap. 16, this volume) may be associated with contemporaneous uplifting of the antiform limb in Subunit Ic: the antiform seems to be related to the culmination of a steep reverse-fault in the underlying sediments. In the eastern flank of the Yusuf Ridge, the lower Pliocene sequence is missing, and the M-reflector and underlying reflections are uplifted and dragging up on the steep-faulted ridge flank. This uplifting does not involve the upper Pliocene (seismic Subunit Ib) or the Pleistocene sediments (seismic Subunit Ia). The main conclusion to be drawn from the seismic images and Site 977 results is that the fault that limits the Yusuf Ridge to the northeast was active before or during the early Pliocene, but not later on. The fault forming the southern scarpment of the Yusuf Ridge (the Yusuf Fault), however, was active till the Present, as demonstrated by the considerable uplifting of the Pleistocene-Holocene sequences near the top of the southwest flank of the ridge ("perched basins" in Fig. 19).
A significant angular unconformity can be noted in the northern end of Line 3S, at a depth of about 3.5 s TWT, between the Pliocene-Pleistocene sequence and sediments (more than 1.7 s TWT thick) underlying the M-reflector (Fig. 12). Although no sample information exists here for sediments beneath the M-reflector, MCS seismic correlation in the region suggests that the high-reflectivity reflector at about 3.7 s TWT may be correlated with reflector-R3 (base of seismic Unit III, late Tortonian, Fig. 3). Hence, sediments below the gravel interval (M-reflector) sampled at Site 977 probably belong to seismic Unit IV, i.e., to the middle Miocene-early Tortonian sequence of the East Alboran Basin. Images from Conrad Line 823 (Fig. 19) confirm the presence of an unconformity below the Pliocene sequence sampled by drilling, and sediments up to 2 s TWT thick underlying the M-reflector.
Site 979 (Shipboard Scientific Party, 1996e) was drilled south of the Alboran Ridge in the northern South Alboran Basin (Fig. 2). The basin corresponds to a fault-bounded depocenter with variable sedimentary thickness, probably partially overlying, or interlayering, volcanic rocks similar to those forming the Alboran Ridge (Fig. 4; Fig. 17, Fig. 18). Site 979 sampled middle-to-upper Pliocene (NN16A to NN19A, seismic Subunit Ib) and Pleistocene/Holocene open-marine (seismic Subunit Ia), hemipelagic monotonous deposits to a total depth of 580.9 mbsf, where drilling had to stop to allow time for logging operations before the end of Leg 161 (Fig. 3). The Pliocene/Pleistocene (NN19A/NN 19B) boundary is at ~345 mbsf. Average sedimentation rates at Site 979 are comparable to the other sites (184-191 m/m.y., late Pliocene; 200 m/m.y., Pleistocene), but below a short hiatus within the late Pliocene (Siesser and de Kaenel, Chap. 16, this volume), biostratigraphic ages suggest a dramatic increase in sedimentation rate to 696 m/m.y. downsection.
Quad-combo logging data were acquired from 60.0 to 277 mbsf and the entire logged section confirms the homogeneity of the sampled sediments. Heat-flow estimation from thermal logging at Site 979 is 79 mW m-2, in contrast to that reported by Polyak et al. (1996) in the South Alboran Basin, which reached average heat-flow values of 118 ± 8 mW m-2. However, Shipboard Scientific Party (1996e) note that calculating heat flow using the shallowest measurement at 20.5 mbsf and its delta temperature (in situ minus mudline) of 2.04°C provides a heat-flow value of 111 mW m-2, close to that reported from nearby sites. This suggests that there may have been a recent change in bottom-water temperature at the drilled site (Shipboard Scientific Party, 1996e).
Sedimentary strata are largely horizontal in the drilled section, with some cored dipping-bed intervals suggesting synsedimentary slumping within the recovered Pliocene-Pleistocene sequence. Nevertheless, seismic images across Site 979 (Fig. 17, Fig. 18) indicate a zone of Pleistocene-to-Holocene deformation (post-M-unconformity) and tilting imaged as a series of folds and faults extending from the southern flank of the Alboran Ridge to the adjacent basin floor. The lowermost sediments sampled at Site 979 (middle Pliocene) lie a few dozen meters above a major angular unconformity, seismically correlated to the M-unconformity (equivalent to the gravel interval) cored at Sites 977 and 978 (Shipboard Scientific Party, 1996e). Line Conrad 825 (Fig. 17) clearly evidences the existence of post-Messinian compressional structures in the Alboran Basin. At the southern flank of the Ridge, the attitude of the Pliocene-to-Holocene syntectonic sediments suggests that folds developed above a near-vertical reverse fault, and the northern flank of the Ridge shows comparable fault-related uplifting of syntectonic Pliocene to Holocene sediments, but folding is absent. The volcanic Alboran Ridge represents a positive, antiformal flower-structure related to a left-lateral strike-slip system that extends along more than 120 km to the Xauen Bank (comprising folded sediments), and aligns with the Jebha Fault (Fig. 4; Fig. 16, Fig. 17, Fig. 18, Fig. 19).
Seismic images (Fig. 17) and the age of sediments at Site 979 confirms that later uplifting of the Alboran Ridge occurred from the late Pliocene to the Holocene. The high sedimentation rate at this site indicates that uplifting was coeval with active subsidence. No significant volcanic or volcaniclastic material was sampled at Site 979, evidencing that late Pliocene-to-Pleistocene uplift of the Ridge was not accompanied by nearby volcanic activity. This implies that volcanism at the Alboran Ridge was not active at these times.
