TRANSMED TRANSECT III, towards a geological cross-section through Western and Central Mediterranean

The TRANSMED III section crosses the Mediterranean region, approximately from W to E, for about 2200 km. It starts with a NW-SE trend crossing the Massif Central, in southern France, the Gulf of Lion, the Provencal basin and the western Sardinian margin. It continues with an E-W direction through Sardinia and part of the Tyrrhenian basin. In the eastern part of the Tyrrhenian basin, the section shows a regional offset along the upper slope of the Campania continental margin. After the offset, the section, crosses the southern Tyrrhenian basin, the Southern Apennines and the Apulian foreland with a SW-NE trend, turning then EW through the southern Adriatic sea, and the Albanides. The easternmost tracts of the section progressively turn to a SW-NE direction, crossing the Serbo-Macedonian massifs, the Balcanides and ending in the Moesian Platform, in northern Bulgaria. From a political point of view, it crosses 5 countries: France, Italy, Albania, Macedonia and Bulgaria. Sixteen scientists from all of these five countries actively contributed to the build up of the cross section that represents the up to date knowledge of the crustal and lithospheric structure along the section trace. From a geological point of view, the section crosses a very complex assemblage of terrains which were repeatedly tectonised both in extensional and compressive regimes during several orogenic cycles (from Proterozoic to Present). This long tectonic history necessarily ended up in the complex crustal and lithospheric structure portrayed in the section. Although most of the main crustal structures of the section formed during the Alpine cycle and the lithospheric plate assemblage is the result of Alpine age plate tectonics, remnants of the Variscan belts and rare relics of previous orogenic cycles are scattered through the section. Labels and colors of faults permit the recognition of such relics. The present day geodynamics along the section is dominated by two active subductions. The continental lithosphere of the Adriatic microplate is, in fact, subducting both eastward and westward underneath the European lithosphere. In particular a westward directed subduction of Adriatic lithosphere underneath the stretched lithosphere of the Tyrrhenian sea occurs along the Apenninic front, continuing through the Calabrian arc . This subduction zone is related, through Sicily and the Sicily channel, to the Maghrebides collisional zone. Seismic evidence for active subduction underneath Italy is provided by the occurrence of subcrustal earthquakes in the northern Apennines (down to some 110 km) and in the Calabrian arc (Fig. 2; down to about 500 km). No subcrustal earthquakes occur under the southern Apennines. The absence of subcrustal seismicity within the continental lithosphere of the Adriatic plate subducting under the southern Apennines could be explained by a rheological control of seismicity. The subducting continental lithosphere could be non-seismogenic at depths larger than about 90 km due to its shallower brittle-ductile transition with respect to the oceanic lithosphere subducting under the Calabrian arc. In the southern Apennines, evidence for active subduction is suggested by local tomographic studies. The dip of the Benioff zone under the Calabrian arc is about 70°. The steep dip of this subduction zone is consistent with observations from worldwide westward-directed subductions. The Apenninic orogen develops on top of this westward-directed subduction. To the east, at the eastern edge of the Adriatic basin, the continental Adriatic lithosphere subducts under the European plate generating the Dinaric and Albanian orogenic belts. More to the south, in correspondence of the Hellenic arc, the Dinaric-Albanian continental subduction merges into the Aegean oceanic subduction, where the Mesozoic? oceanic Ionian lithospere subducts under the subduction complex of the Mediterranean ridge and the stretched european lithosphere of the Aegean basin. Clear Benioff zones are imaged by seismicity under the Hellenic arc. The dip of the slab in this region is markedly lower than that under Calabria. This pattern is typical of eastward-directed subductions, which are quite flattened with respect to those directed to the west. No clear Benioff zones are recognised under the Dinarides and the Albanides, where continental subduction occurs. An explanation similar to that proposed for the absence of seismicity in the continental slab subducting under the southern Apennines could be suggested. Tomographic studies show clearly a fast velocity body under the Albanides. This velocity anomaly is interpreted as the evidence of the active subduction process. Both Apenninic and Dinaric-Albanian-Hellenic orogens are associated to backarc spreading, but they present significant geodynamic differences which are listed below: i) the foreland monocline in the Apennines is regularly steeper (8-10°) than in the Dinarides-Albanides-Hellenides (2-4°); ii) subsidence rates are faster (> 1 mm/yr) in the Apennines foredeep with respect to the Dinarides-Albanides-Hellenides foredeep (< 0.3 mm/yr); iii) the heat flow profiles commonly show lower values for the Apennines foredeep with respect to the Dinarides-Albanides-Hellenides foredeep; iv) the heat flow profiles mark average higher values for the Apennines backarc basins (Provençal, Alboran, Algerian, Tyrrhenian seas) than the Dinarides-Albanides-Hellenides backarc (Aegean sea, western Anatolia); v) gravity profiles indicate a higher degree of compensation in the Apennines backarc basins; vi) mean topography is lower in the Apennines and related backarc basins with respect to the Dinarides-Albanides-Hellenides and associated backarc; vii) the Apennines constitute mainly a thin-skinned accretionary wedge, whereas the Dinarides-Albanides-Hellenides is rather a thick-skinned orogen; however both orogens show along strike variations of the decollement depth. Both Apenninic and Dinaric-Albanian-Hellenic subduction zones and related orogens probably contain the relics of precursor Late Mesozoic-Early Tertiary subduction zones (e.g., the Alps for the Apennines and the Vardar for the Dinarides), later stretched and partly (to completely) dismembered by Tertiary-Quaternary extension. From a quick glance to the TRANSMED III section it appears clearly that the lithospheric structure is controlled mainly by extensional tectonics in the western part of the section and by collision/subduction processes in its eastern part. It should however be recalled that extensional tectonics also dissects the orogens extending from Albania to the Moesian platform in Bulgaria. Cenozoic, mostly Neogene, grabens are widespread in this region (e.g., the Sofia and Pernick grabens in Bulgaria and the Skopje graben in Macedonia). Although the lithospheric signature of extensional tectonics in these regions is partly hidden by the occurrence of the eastward-directed subduction, the influence of extensional tectonics on the lithospheric thickness is evidenced by the lithospheric thinning from the Moesian platform to the southwestern Balkans. The thinning approximately corresponds to the location of the Sofia graben. The lithospheric geometry under Macedonia and Albania is controlled by the subduction of the continental lithosphere of the Adriatic plate. The thickness (about 100 km) of the Adriatic plate along the path of the section is anomalously high, with respect to what observed in the Adriatic basin more to the north (60-70 km). The crustal thickness in the Apulian foreland is around 30 km. Apparently, the lithosphere of the Apulia foreland and of the southern Adriatic basin was only marginally thinned during the Mesozoic rifting which affected the Adriatic plate. The entrance of such a thick continental lithosphere in the westward directed subduction zone below the Apennines most probably induced the Pleistocene uplift of the Apulian foreland which exposed not deformed Apulian platform rocks to erosion in the Apennines foreland (Doglioni et al., 1994). The crustal thickness increases, in the Apennines, from the foreland to the axial part of the chain and is mainly accommodated by the downflexure of the subducting Adriatic slab and by doublings in the sedimentary cover. The lithospheric thickness follows a similar pattern, governed by the downflexure of the Adriatic slab. The crustal and lithospheric geometry from the axial part of the Apennines to the Gulf of Lion is controlled by the Neogene backarc extension related to the “eastward” retreat of the Apenninic subduction zone. The age of the backarc related extension shows a younging from west to east. Extension was mainly active in the Gulf of Lion – Provencal basin area in the Late Oligocene-Langhian period. The shift of active extension from west to east of Sardinia occurred around the end of the lower Miocene or in the Langhian and produced the middle Miocene to Present opening of the Tyrrhenian basin. In this western sector of the section, the European plate, previously affected by the Variscan, Pirenean and Alpine orogenic deformations, shows a boudin-like geometry. Two major basins (the Provencal and the Tyrrhenian basins) are separated by a swell (the Sardinia block). The lithospheric and crustal thickness of the basins and swells is mostly determined by the kinematics of the rifting but a partial control by previous compressive structures can be at places recognised. For example, the 35 km thick crust and the 65-70 km thick lithosphere of Sardinia, although partly dismembered by a N-S Oligocene graben, are controlled by the occurrence of relics of the south to west vergent Variscan edifice. Relics of the Variscan, Pirenean and Alpine edifices are moreover expected to occur within the strongly rotated blocks that characterise the crustal geometry of the above-mentioned basins. Although the geometry of such relics cannot be determined from available data, their existence is clearly testified to by samplings from the Tyrrhenian basin. In the Gulf of Lion, the continental lithospheric thickness decreases from NW (where it is about 90 km thick) to the SE. Within the Provencal basin, a transition from thinned continental lithosphere to intermediate (transitional) to oceanic lithosphere has been revealed by ECORS deep seismic profiles. In the central part of the Provencal basin, the thickness of the Neogene backarc oceanic lithosphere is less than 25 km. More to the SE, towards the western Sardinian margin, first transitional and then thinned continental lithosphere is encountered. The thickness of crust and lithosphere gradually increase to the values outlined above for central Sardinia (respectively 35 km and 65-70 km). The eastern Sardinian margin shows a progressive decrease of the continental lithospheric thickness, which reaches minimum values of about 15 km in the Magnaghi plain. It should be noted that gravity modelling suggests an asymmetric topography of the lithosphere-asthenosphere boundary across Sardinia. Below the western Sardinian margin the increase of the lithospheric thickness toward the center of the Island is less abrupt than below the eastern Sardinia margin. Moreover, below Sardinia, the thickest lithosphere is displaced to the east with respect to location of the thickest crust. The profile crosses the Vavilov plain, characterised by Pliocene oceanic lithosphere with thickness of about 10 km. A new transition to the stretched continental lithosphere of the Campanian continental margin is accompanied to an increase of the lithospheric thickness. Northwest of the Gulf of Lion, the profile crosses continental lithosphere thickened during the Variscan and Pyrenean orogenic events. Crustal thickness remains quite constant (around 30 km) whereas the thickness of the lithospheric mantle shows a minimum in correspondence of the Cantal volcano. This decrease is desumed from thermal models and is tentatively related to thermal erosion of the lithosphere due to the activity of a mantle plume. Figure 3 shows two maps with heat flow values measured respectively onshore and offshore in the Mediterranean area. Along the section, a good correlation between age of geodynamic processes (in particular backarc extension) and heat flow density can be envisaged. Areas characterised by backarc extension show the highest heat flow values: 60-110 mW/m2 in the Gulf of Lion and Provencal basin, and more than 100-200 mW/m2 in the Tyrrhenian Sea. It is clear that heat flow values increase with the decrease of the age of rifting/drifting of the basins. Lower heat flow densities are measured in the axial and eastern part of the Apennines and in the Albania-Macedonian-Bulgarian chains, where values normally lower than 50-60 mW/m2, i.e., consistent with values normally measured in mountain belts, are reported. TRANSMED III has a peculiar orientation with respect to the other sections of TRANSMED project. All the other sections have dominant north – south orientations. Their N-S orientation is justified by two line of reasoning. Firstly, these sections are roughly perpendicular to regional structures. Secondly, they are subparallel to the Neogene direction of relative convergence between Eurasia and Africa. TRANSMED III section, roughly E-W, accomplishes the requirement of perpendicularity to regional structures. Apparently, however, TRANSMED III section has a disputable orientation with reference to the Africa-Eurasia convergence direction. This apparent inconsistency no more holds at a close inspection of Fig. 4, which sketches the Neogene evolution of the Mediterranean area. Although, paradoxically, the extension determining most of the western Mediterranean developed in a context of relative convergence between Africa and Europe, the maximum amount of north-south Africa/Europe relative motion at the Tunisia longitude has been about 135 km in the last 23 Ma, more than five times slower with respect to the eastward migration of the Apennines arc which migrated eastward more than 700 km during the last 23 Ma. The first consequence of this observation is that the eastward migration of the Apennines-Maghrebides arc is not a consequence of the relative N-S relative convergence between Africa and Europe. It is rather the consequence of the Apennines-Maghrebides subduction rollback, which is to be considered a major actor of the Mediterranean geodynamics. In other words, this section, with its “awkward” orientation, permits to adequately describe and analyze this geodynamic aspect. Moreover, TRANSMED III section, one of the longest of the entire project, uniquely shows a spectacular feature of the Mediterranean geodynamics: the double subduction of the Adriatic plate under the European plate. The Adriatic basin is the hinge of this system and links the extensional setting of the western Mediterranean, governed by the westward dipping subduction under the Apennines with the eastern Mediterranean, governed by the eastward directed Dinaric-Albanian-Hellenic subduction. TRANSMED III section permits to appreciate such differences within a single geodynamic setting