Abstract
The West European collisional Alpine belts are the result of the inversion, initiated in the middle Cretaceous, of the complex western Neotethys and the Atlantic continental rift domains and closure of remnants of Tethys between the North Africa and European cratons. While the kinematics of Africa relative to Europe is well understood, the kinematics of microplates such as Iberia and Adria within the diffuse collisional plate boundary is still a matter of debate. We review geological and stratigraphic constraints in the peri-Iberia fold-thrust belts and basins to define the deformation history and crustal segmentation of the West European realm. These data are then implemented with other constraints from recently published kinematic and paleogeographic reconstructions to propose a new regional tectonic and kinematic model for Western Europe from the late Permian to recent times. Our model suggests that the pre-collisional extension between Europe and Africa plates was distributed and oblique, hence building discontinuous rift segments between the southern Alpine Tethys and the Central Atlantic. They were characterised by variably extended crust and narrow oceanic domains segmented across transfer structures and micro-continental blocks. The main tectonic structures inherited from the late Variscan orogeny localized deformation associated with rifting and orogenic belts. We show that continental blocks, including the Ebro-Sardinia-Corsica block, have been key in accommodating strike-slip, extension, and contraction in both Iberia and Adria. The definition of a new Ebro-Sardinia-Corsica block allows refining the tectonic relationships between Iberia, Europe and Adria in the Alps. By the Paleogene, the convergence of Africa closed the spatially distributed oceanic domains, except for the Ionian basin. From this time onwards, collision spread over the different continental blocks from Africa to Europe. The area was eventually affected by the West European Rift, in the late Eocene, which may have controlled the opening of the West Mediterranean. The low convergence associated with the collisional evolution of Western Europe permits to resolve the control of the inherited crustal architecture on the distribution of strain in the collision zone, that is otherwise lost in more mature collisional domain such as the Himalaya.
Résumé
Les ceintures orogéniques alpines d’Europe occidentale sont le résultat de l’inversion, initiée au Crétacé moyen, de l’Océan Néotéthys occidental et des domaines du rift continental atlantique et de la fermeture des reliquats océaniques de la Téthys entre l’Afrique du Nord et les cratons européens. Si la cinématique de l’Afrique par rapport à l’Europe est bien comprise, celles de microplaques telles que Iberia et Adria, à l’intérieur de la limite de plaque, est encore débattue. Nous examinons les contraintes géologiques et stratigraphiques dans les ceintures orogéniques et les bassins péri-ibériques afin de définir l’histoire de la déformation et la segmentation crustale du domaine européen occidentale. Ces données sont ensuite intégrées avec d’autres contraintes issues de reconstructions cinématiques et paléogéographiques récemment publiées afin de proposer un nouveau modèle tectonique et cinématique régional de l’Europe occidentale, de la fin du Permien à l’actuel. Notre modèle montre que l’extension pré-collisionnelle entre la plaque Europe et la plaque Afrique était distribuée et oblique, à l’origine de segments de rift discontinus entre la partie sud de la Téthys Alpine et l’Atlantique Central. Ces rifts étaient caractérisés par une croûte variablement amincie et des domaines océaniques étroits segmentés à travers des structures de transfert et des micro-continents. Les principales structures tectoniques héritées tardi-varisques ont localisé les domaines d’amincissement et les orogènes. Nous montrons que plusieurs blocs continentaux, y compris le bloc Ebre-Sardaigne-Corse, ont joué un rôle clé dans l’accommodation des mouvements décrochants, de l’extension et de la convergence de l’Ibérie et d’Adria. De plus, nous montrons que l’existence de ce bloc permet d’affiner la relation tectonique entre l’Ibérie, l’Europe et domaine adriatique septentrional dans les Alpes. Au Paléogène, la convergence de l’Afrique referme les domaines océaniques à l’exception du bassin ionien. À partir de ce moment, la collision implique les différents blocs continentaux, permettant un transfert efficace de la déformation de l’Afrique vers l’Europe. La zone a finalement été affectée par le Rift Ouest Européen, à la fin de l’Eocène, qui a en partie contrôlé l’ouverture de la Méditerranée occidentale. La faible convergence associée à l’évolution orogénique de l’Europe occidentale permet de mieux comprendre le contrôle de l’architecture crustale héritée sur la distribution de la déformation dans la zone de collision, qui est autrement mal défini dans des domaines de collision plus matures tels que l’Himalaya.
Introduction
The evolution of the Alpine orogenic belts in the Mediterranean region is understood to result from the accretion, subduction and back-arc extension of fragmented continents or microplates, inherited from the Tethys opening, between Africa and Europe (e.g., Dewey et al., 1973). Tectonic reconstructions of the Alpine collision have attempted to resolve the evolution of tectonic boundaries around two main continental microplates (i.e., Adria and Iberia), assuming that the large-scale kinematics of Africa relative to Europe is reasonably understood (Handy et al., 2010; van Hinsbergen et al., 2020). Controversies on the exact kinematics of Adria and Iberia lead to a different number, size and position of continental blocks and oceanic domains, hence questioning the drivers of the orogenic evolution. The current issues are further challenged by the dismantlement of the initial Alpine orogenic system during the opening of the Western Mediterranean region for which slab dynamics has been well emphasized (Jolivet and Faccenna, 2000; Faccenna et al., 2014).
Only recently, Iberia has been the focus of a number of studies, solving some important issues regarding the timing and amount of extension and strike-slip evolution of the plate boundary between Iberia and Europe. Specifically, the kinematics of Iberia has been subjected to a long lasting debate about the reliability of magnetic anomalies in the North Atlantic and geological evidence in the Pyrenees (Vissers and Meijer, 2012; Barnett-Moore et al., 2016; Nirrengarten et al., 2017). But a consensus has since emerged that the eastward movement of Iberia should be explained by strike-slip and transtension in the interior of the Iberian plate, leading to define an Ebro block. Evidence for an individual Ebro continental block has been shown for decades (Salas and Casas, 1993; Arche and López Gómez, 1996) but the intra-plate segmentation has only recently been considered in geodynamic models (Tugend et al., 2015; Tavani et al., 2018) and plate reconstructions (Nirrengarten et al., 2018; Angrand et al., 2020). These recent models have fundamental implications for the Iberia-Europe boundary, as they argue for the distribution of the left-lateral movement from the late Paleozoic to Early Cretaceous, along the Europe-Iberia plate boundary across two tectonic corridors, in the Pyrenean and Iberian rift-basins. To the south, the Iberia-Africa relative motion has been accommodated along the complex Gibraltar transform zone, which includes the Betic-Rif orogen and north African basins (Sallarès et al., 2011; Schettino and Turco, 2011; Frizon De Lamotte et al., 2015; Daudet et al., 2020). The impact for the reconstruction of the plate boundary between Africa and Europe and farther into the Alpine Tethys and Europe/Adria plate boundary has only been recently tested for the period ranging from the Permian-Triassic to the Early Cretaceous (Angrand et al., 2020).
The present-day configuration of the orogenic system between the European and African plates (Fig. 1a) is characterized by narrow mountain ranges separating small continental blocks or micro-plates that have accommodated both extension/transtension and compression/transpression since the late Paleozoic (Fig. 1b; Angrand et al., 2020). For instance, the deformation between stable Europe and segmented Iberia has been distributed along several orogenic segments: the Pyrenees, which can be extended to the Provence in SE France, between Europe and Ebro (Mouthereau et al., 2014; Teixell et al., 2018; Espurt et al., 2019a), the Basque-Cantabrian Basin between Galicia (West Iberia) and the Bay of Biscay (Cámara, 2017; García-Senz et al., 2019), and the intra-Iberian basins, now inverted in the Iberian Range between South Iberia and Ebro (Aldega et al., 2019; Rat et al., 2019).
In this study, we aim to expand the recently published Angrand et al. (2020) model of the Western Tethys–North Atlantic kinematics, which focused on the late Permian-Triassic geodynamics to the middle Cretaceous. We integrate results from recently published studies of the OROGEN project, coupled with geological, geophysical, and stratigraphic data compiled from the literature in order to build a regional geodynamic and kinematic model of the Mediterranean region, accounting for the Atlantic and Alpine Tethys evolution, from the late Permian to Neogene.
Our manuscript is organized as follow. We first review geological constraints on the tectonic evolution of the complex Europe-Africa plate boundary, from SE France-Gulf of Lion to the Basque-Cantabrian-Pyrenees, Iberian Range-Central System, and the Betics-Alboran domain (Sects. 2 and 3). Considering the size of the area studied and the time span covered by our reconstruction, this first part is inevitably long. Using these geological constraints, we present the crustal segmentation of the West European realm, which is crucial for defining the continental plate geometry of our kinematic reconstruction (Sect. 4). Then, we present the method and the results of our kinematic reconstruction that are shown into successive key time frames (Sect. 5). The outputs of our model is then presented in terms of N-S strain distribution during the Late Mesozoic-Cenozoic (Sect. 6). Finally, we discuss several important, yet unanswered questions, regarding the kinematics of the Africa-Europe plate boundary that can be solved in Iberia (Sect. 7) such as: 1) how the late Variscan shear zones were reactivated both in extension and compression and explain first-order features of the kinematic framework; 2) what are the implications of the Ebro block kinematics on the tectonic evolution of Sardinia-Corsica, and its connection to the Provence and Briançonnais in the Alps (northern Europe/Iberia/Adria plate boundaries), and 3) how the fragmentation of Iberia controlled the tectonic evolution of AlKaPeCa terranes in the south (southern Africa/Iberia/Adria plate boundaries) and the opening of the Western Mediterranean.
Evolution of the Europe-Africa plate boundary since the Late Paleozoic
The post-Variscan geodynamic evolution of western Europe and northern Africa in the Carboniferous-Permian coincided, or slightly predated, the opening of the Neotethys in Permian-Triassic times (Fig. 1b) (Ziegler, 1989, 1990b; Stampfli et al., 2001; Stampfli and Borel, 2002; Schettino and Turco, 2011; Angrand et al., 2020). Permian crustal thinning and magmatism associated with the break-up of Pangea are recognized throughout Europe, including the Oslo Rift, North Sea and the German Permian Basin (Neumann et al., 1992; Van Wees et al., 2000; Glennie et al., 2003; Heeremans et al., 2004; Upton et al., 2004), Massif Central (Bruguier et al., 2003), the Alps and Corsica (Rossi et al., 2009; Cassinis et al., 2012; Ballèvre et al., 2014), the Pyrenees (Vissers, 1992; Denèle et al., 2012; Gretter et al., 2015; Rodríguez-Méndez et al., 2016; Espurt et al., 2019b; Saspiturry et al., 2019) and the peri-Iberian rift basins (Arche and López Gómez, 1996; López-Gómez et al., 2019; Angrand et al., 2020 and references therein) (Fig. 2). The tectonic control of the Tethyan continental rifting by the late Variscan evolution is argued by Permian-Triassic depocentres that are superposed onto older Variscan structures.
The Atlantic opening was preceded by the Late Triassic-Early Jurassic (∼ 201 Ma) Central Atlantic magmatic event (CAMP, Central Atlantic Magmatic Province) (Olsen, 1997; Marzoli et al., 1999; McHone, 2000; Peace et al., 2019a), whose relationship with the geodynamic evolution of western Europe has been recently discussed by Peace et al. (2019a) and Angrand et al. (2020). The extension in the Central Atlantic between North America and Africa started in the Triassic (Leleu et al., 2016), while the breakup and onset of oceanic spreading is commonly considered to have occurred in the Early Jurassic (190 Ma) (Figs. 1b and 2) (Schettino and Turco, 2009; Labails et al., 2010).
In the Lower Jurassic, the Alpine Tethys Ocean opened between Iberia/Europe and Adria/Africa (Figs. 1b and 2) (Puga et al., 2011; Schettino and Turco, 2011; Le Breton et al., 2021). Recent reconstructions in the Betics considered a hyper-extended domain rather than an oceanic crust (Vergés and Fernàndez, 2012; Daudet et al., 2020; Pedrera et al., 2020b). The recognition of MORB-type gabbro dated at 180 Ma in the internal zones of the eastern Betics (Puga et al., 2011) supports the continuation of the Alpine Tethys in the east (named Betic Tethys, Ligurian Tethys, Nevado-Filabrides Ocean, etc.) or a thinned continental crust at the Ocean-Continent transition (Gómez-Pugnaire et al., 2000). However, the precise geometry of the southwestern termination of the Alpine Tethys is debated. Discussions include whether the Alpine Tethys was made of a single or a double branch (e.g., Western and Eastern Ligurian Tethys) of true oceanic to hyper-extended continental crust with variable degree of rifting (see review in Guerrera et al., 2019).
After a phase of stalling in the Lower Jurassic, extension resumed in the southern North Atlantic and the ridge migrated northward during the Upper Jurassic/Lower Cretaceous (see Barnett-Moore et al., 2016; Nirrengarten et al., 2018, for recent review and kinematic models). The E-W extension between Iberia and North America produced strike-slip movement that must be accommodated along the Iberia-Europe plate boundary. The North Atlantic ridge migration then resumed in the Aptian-late Cenomanian and opened the Bay of Biscay at high angle to the Atlantic ridge axis, resulting in the rotation of Iberia. The kinematics of Iberia has long been a matter of debate because of its implications for reconstruction of the Tethys domain. Controversies originally dealt with as to whether the “J” anomaly should be viewed or not as an M0 magnetic anomaly (∼ 120 Ma). The J anomaly is indeed interpreted to reflect magmatic pulses on the Atlantic hyper-extended margin of Iberia instead of a true oceanic anomaly (Bronner et al., 2011; Nirrengarten et al., 2017). Two end-member models have been proposed. Reconstructions that do not consider, or only partially, the J anomaly all reproduce extension in the North Pyrenean Rift, but differ according to the timing and amount of strike-slip partitioned with N-S/NE-SW orthogonal extension along the Iberia-Europe plate boundary (Olivet, 1996; Jammes et al., 2009; Tugend et al., 2014a; Nirrengarten et al., 2018; Angrand et al., 2020). Other restorations that fit the J anomaly imply a 500 km convergence, closing a Jurassic “Tethyan” ocean or hyper-extended margin between Iberia and Europe during the Cretaceous (Srivastava et al., 2000; Sibuet et al., 2004; Vissers and Meijer, 2012). Geological implications of this scenario for the Pyrenees, such as the emplacement of a large basin in the Jurassic and the development of a Cretaceous accretionary prism, are not supported by geological observations. The recent reconstructions by Angrand et al. (2020) emphasize the key role played by the Ebro block between Iberian and Europe. Accounting for the Ebro block indeed allows transfer of deformation from the Atlantic to the Tethys domain by accommodating the required eastward translation to South Iberia. This allows minimizing strike-slip Cretaceous deformation in the Pyrenean rift, while keeping a dominant orthogonal extension during the opening of the Bay of Biscay.
The counter-clockwise rotation of Africa that resulted from the opening of the Central Atlantic in the Early Jurassic caused the closure of the Neotethys and peripheral oceanic basins in eastern Europe such as the Meliata and Pindos oceans (Fig. 1b) (Schmid et al., 2008; van Hinsbergen et al., 2020). By Late Jurassic-Early Cretaceous times the propagation of the southern North Atlantic northwards eventually opened the Bay of Biscay in the late Aptian. The eastward movement of Iberia in turn caused the initiation of convergence in the Alpine Tethys domain (Angrand et al., 2020) at the origin of Eo-Alpine deformation (Handy et al., 2010, 2015). However, the exact implications on the kinematics of the continental blocks surrounding the Alpine Tethys between Europe, Iberia, Ebro and Adria are not well understood.
In Europe, the onset of Africa-Europe convergence in Late Cretaceous times is recorded from Morocco to Central Europe (e.g., Kley and Voigt, 2008). This episode of early contraction is recorded by the stratigraphic record in the Pyrenees and Provence (Fig. 2) (Puigdefàbregas and Souquet, 1986; Beaumont et al., 2000; Vergés et al., 2002; Christophoul et al., 2003; Espurt et al., 2012; Bestani et al., 2016; Ford et al., 2016; Rougier et al., 2016; Angrand et al., 2018). It is followed by widespread contraction inboard Iberia from the Paleogene onward (Fig. 2), for instance, in the Betic Cordillera (Platt and Vissers, 1989; Daudet et al., 2020) and the Iberian Range (e.g., Rat et al., 2019). In the Alps, early deformation is recorded during the Eocene, reaching its climax during the Oligocene-Miocene (Fig. 2) (Ford and Lickorish, 2004; Handy et al., 2010).
The opening of N-S West European Rift in the late Eocene (∼ Priabonian, 34 Ma), parallel to the direction of the regional contraction between Europe and Africa (Bois, 1993; Séranne, 1999; Dèzes et al., 2004), marks a change in the internal dynamics of both the Pyrenean and the Alpine orogenic systems. As the West European Rift progresses southward it was eventually accompanied by, or even promoted opening of the Gulf of Lion (Fig. 2). However, no clear genetic affinity between the West European Rift and the West Mediterranean is yet established. At ∼ 30 Ma, slab rollback-related extension in the Western Mediterranean opens the Gulf of Lion and Valencia basins, followed by extension in the Liguro-Provençal, Alboran (23–20 Ma) and Tyrrhenian basins (∼ 15 Ma) (Fig. 2) (Comas et al., 1992; Lonergan and White, 1997; Séranne, 1999; Jolivet et al., 1999, 2020; Loreto et al., 2020).
Geological constraints on the Europe-Africa plate boundary
France Southeast Basin and Provence
The France Southeast Basin forms a transitional domain between the Provence and the SW French Alps (Fig. 3a). Post-Variscan extension started in the Permian during the initial stages of Pangea break-up as indicated by NNE and ESE trending basins (Espurt et al., 2012; Bestani et al., 2016). Extension then progressively intensified during the Triassic-Jurassic period in response to the Alpine Tethys rifting leading to the formation of the Dauphinois basin in the Early Jurassic (Fig. 2) between the less subsident domains of the Provence High and the Occitan High (paleo-Massif Central) (Lemoine, 1985; Curnelle and Dubois, 1986; Baudrimont and Dubois, 1977). During the Early Cretaceous, the Provence High evolved into the Durance Isthmus, an elongated EW-directed topographic high that separated the South Provence basin from the Vocontian basin (Fig 3a).
The Early Cretaceous rift history in Provence appears to be closely linked to that of the Pyrenees and the opening of the Bay of Biscay (Curnelle and Dubois, 1986; de Graciansky and Lemoine, 1988; Tavani et al., 2018). Based on age correspondence and the position in the external domain of the Western Alps (i.e., west of the Briançonnais and the Penninic Front), it has also been soon suggested that the Vocontian Basin may instead be connected to the termination of the Valaisan domain (Maury and Ricou, 1983; Dercourt and Vrielynck, 1993; Dercourt et al., 2000; Debelmas, 2001; Manatschal and Müntener, 2009). Other reconstructions, in contrast, assume the Valaisan domain is a large oceanic domain in continuity with the Pyrenean rift basin between the Provence and Corsica/Briançonnais, accommodating the eastward motion of Iberia (Trümpy, 1988; Stampfli, 1993; Stampfli and Borel, 2002; Turco et al., 2012).
The Provence fold-thrust belt developed in response to contraction between Sardinia-Corsica and Europe associated with the Campanian-Eocene Pyrenean event. The deposition of a Late Cretaceous foreland succession is reported at this time (Espurt et al., 2012; Bestani et al., 2016). Shortening is thought to have resulted from 1) upper plate transpression above the oblique north-verging subduction of the Ligurian Tethys south of Sardinia-Corsica (Europe) (Lacombe and Jolivet, 2005) or 2) linked to the inversion of the Pyrenean/Valaisan rift shaped by a hyper-extended rift system (Espurt et al., 2012; Bestani et al., 2016). Some alternative models proposed that contraction resulted from the contact between Sardinia-Corsica (Iberia), moving along a left-lateral strike-slip fault corridor until Late Cretaceous, and Adria (Schreiber et al., 2011). None of these reconstructions follow kinematic models assuming a large Pyrenean-Valaisan ocean between the Provence and Sardinia-Corsica (Trümpy, 1988; Stampfli, 1993; Stampfli and Borel, 2002; Turco et al., 2012) because there is no evidence for ophiolite-like mélange or folded deep-marine sediments formed along a presumed suture zone. A total horizontal shortening of ∼ 155 km is estimated between the northern edge of the Provence basins and Sardinia, including ∼ 50 km in the Provence basin (Bestani et al., 2016). This total shortening is in the range of the one estimated in the Pyrenean belt (see below).
The opening of the Oligocene-Miocene Gulf of Lion and Ligurian basin from 32 Ma (Fig. 2) (Gorini et al., 1993; Séranne et al., 1995; Séranne, 1999) resulted in N-S extension, reactivating previous Pyrenean thrusts into E-W normal faults (Bestani et al., 2016). E-W extension related to the opening of West European Rift is also reported in the Forcalquier-Manosque basin (Bergerat, 1987) and the Sainte Victoire area (Lacombe et al., 1992) (Fig. 3a). The Neogene to present-day Alpine contraction (Lacombe et al., 1992; Espurt et al., 2012; Bestani et al., 2016) is superimposed on Late Cretaceous-Eocene Pyrenean compressive structures and reflects a limited amount of shortening. Among the tectonic features that our kinematic reconstruction must take into account are the NNE-SSW to NE-SW major transfer faults (Nîmes, Salon-Cavaillon, Aix, Durance, and Barjols faults, Fig. 3a). These faults intersect the Late Triassic to Early Cretaceous sedimentary cover and have recorded both the Cenozoic extension and inverse reactivation (e.g., Durance Fault) during the Pyrenean and Alpine orogeny (Roure et al., 1992).
Sardinia-Corsica and Gulf of Lion
The Sardinia and Corsica islands form an N-S elongated continental block in the Western Mediterranean Sea that was rifted away from Provence in southern France and rotated by 45–50° counter-clockwise during the opening of the Ligurian-Provençal oceanic basin in the early Miocene, between 20.5 and 15 Ma (Ferrandini et al., 2003; Gattacceca et al., 2007).
Sardinia-Corsica preserves a Variscan Paleozoic basement intruded by late Variscan upper Paleozoic calc-alkaline magmatic rocks and Permian granitoids associated with volcano-sedimentary rocks (Matte, 1991; Paquette et al., 2003; Rossi et al., 2009; Ballèvre et al., 2014). The Alpine Corsica refers to units exposed in northeastern Corsica that comprises HP rocks in a series of Tethys-derived nappes now forming ophiolitic complexes similar to the Schistes Lustrés complex exposed in the Western Alps (Fig. 1) (Lacombe and Jolivet, 2005; Vitale Brovarone et al., 2013; Malusà et al., 2016). As such Alpine Corsica represents the southwestern termination of the Alpine orogenic wedge emplaced between Adria and Europe. It is worth noting the Variscan/Alpine wedge bounding fault in its pre-Neogene orientation lies parallel to the major Nîmes and Durance faults in Provence (Fig. 3a) (Lacombe and Jolivet, 2005) and must be understood as an inherited Variscan structure.
The Mesozoic sedimentary sequence in the northwest and east Sardinia shows depositional environments deepening toward the SE on the former Alpine Tethys margin (Azema, 1977; Philip and Allemann, 1982). Continental break-up associated with opening of Alpine Tethys is outlined by an unconformity dated as Bajocian (∼ 170 Ma) (e.g., Costamagna et al., 2007; Costamagna, 2016). Details of the middle Cretaceous depositional history show similarities between Sardinia and Provence, including the regionally recognized Bauxite deposits from the Albian to the Turonian (110–90 Ma) (Philip and Allemann, 1982). This event temporally correlates with the regional opening of the Bay of Biscay, a relationship that is implicit in some previous publication (Masse and Allemann, 1982). After the deposition of marine carbonates during Coniacian-Santonian, middle Eocene (Lutetian, 48–41 Ma) conglomerates (Cixerri Formation) were deposited in southern Sardinia. They bear evidence of reworked late Albian-Cenomanian carbonate sourced from the eastern Pyrenees (Cherchi, 1976). A tectonic connection with the Pyrenees prior to rotation is therefore suggested.
A late Paleocene to middle Eocene phase of contraction is recognized for Sardinia-Corsica (Dieni et al., 2008). This timing is supported by low-temperature thermochronological constraints from the Variscan basement that document diachronous exhumation events from SE Sardinia to NW Corsica beginning at 67–49 Ma (Malusà et al., 2016). The emplacement of the Alpine wedge in northeast Corsica is also well constrained by remnants of Eocene (post-Lutetian, 41 Ma) foreland basin deposits that are overthrusted by HP units. Intrusions of the Cixerri Formation by andesites topped by late Oligocene-Miocene volcano-sedimentary deposits (Bellon et al., 1977; Barca and Costamagna, 2010) reveal a transition, most likely during the Priabonian, from Pyrenean contraction to extension controlled by the opening of the Algero-Provençal back-arc basin (Fig. 2). The Priabonian extensional event, which marks the southern extension of the West European Rift, appears to predate the rifting in the Gulf of Lion dated as late Oligocene-early Miocene (Gorini et al., 1993; Séranne et al., 1995; Séranne, 1999; Réhault et al., 2012). The geodynamic drivers of these two events are therefore different. The onset of rifting is confirmed in Corsica at 32 Ma by top-to-the-East shearing of metamorphic units (Lahondère and Lahondère, 1988; Fournier et al., 1991; Jolivet et al., 1991, 1998; Caron, 1994; Brunet et al., 2000), the opening of Oligocene to earliest Miocene (Aquitanian) rift basins and exhumation in the upper crust (Zarki-Jakni et al., 2004; Fellin et al., 2005; Cavazza et al., 2007; Danišík et al., 2007; Malusà et al., 2016). As rotation of Sardinia-Corsica occurred in the Aquitanian-Burdigalian (Cherchi and Montadert, 1982), volcanism spread over the whole region both onshore and offshore in the Gulf of Lion (Bellon and Brousse, 1977; Girod and Girod, 1977; Réhault et al., 2012).
Controversies that need to be addressed in our reconstructions are on the pre-Miocene kinematics. Models indeed differ on whether the Sardinia-Corsica domain was part of Iberia (Trümpy, 1988; Stampfli, 1993; Stampfli and Borel, 2002; Turco et al., 2012) or a separate microplate during the Jurassic-Cretaceous until contraction initiated in the Late Cretaceous (Rosenbaum et al., 2002a) or until the Oligocene, after a phase of contraction with southern France in the Eocene (Advokaat et al., 2014).
Pyrenees
The Pyrenees (Fig. 3b) show an exceptional preservation of the pre-collisional architecture of the rifed margin, including remnants of a sub-continental mantle exhumed during Early Cretaceous extension (Lagabrielle and Bodinier, 2008; Jammes et al., 2009; Lagabrielle et al., 2010), which is well imaged at depth by recent seismic tomographic images (Wang et al., 2016; Chevrot et al., 2018). Based on the age of the youngest sediments affected by the synrift HT metamorphism (Albarède and Michard-Vitrac, 1978; Montigny et al., 1986; Vacherat, 2014; Clerc et al., 2015; Saspiturry et al., 2020), the crustal breakup is dated as Early Cretaceous (Fig. 2), but extension was long-lasting likely initiated as a result of Pangea break-up during the late Permian (Lucas, 1985; Saspiturry et al., 2019; Angrand et al., 2020).
In contrast to other peri-Tethyan domains, the stratigraphy of the northern (Europe) and southern (Iberia) margins in the Pyrenees indicate a lack of Early-Middle Jurassic tectonic subsidence (Curnelle, 1983; Brunet, 1986, 1994; Martín-Chivelet et al., 2019). Rifting indeed occurred from Late Jurassic to Early Cretaceous and reached a climax in the Aptian-Albian along E-W-directed rift basins, from the Parentis and the Aquitaine basins in the West (Brunet, 1986, 1994; Bois and Ecors Scientific Party, 1990; Jammes et al., 2009; Angrand et al., 2018) to Organyà in the south (Berástegui et al., 1990; Mencos et al., 2015), including evidence for ubiquitous hyper-extension with sub-continental exhumation in the North Pyrenean basins (Masini et al., 2014; Clerc et al., 2015; Corre et al., 2016; Teixell et al., 2016; Espurt et al., 2019a; Angrand et al., 2021).
The onset of contraction is constrained by the late Santonian-early Campanian age of early thrusting events, exhumation and clastic influx in the northern and southern foreland basins (Puigdefàbregas and Souquet, 1986; Beaumont et al., 2000; Christophoul et al., 2003; Beamud et al., 2011; Whitchurch et al., 2011; Filleaudeau et al., 2012; Mouthereau et al., 2014; Ford et al., 2016; Rougier et al., 2016; Vacherat et al., 2016, 2017; Angrand et al., 2018; Ternois et al., 2019) (Fig. 2). From the late Santonian/early Campanian to middle Miocene times, the Pyrenees accommodated a N-S shortening of ∼ 111 km in the East (excluding the closure of the exhumed mantle domain, Grool et al., 2018), ∼ 100–165 km in the centre (Choukroune et al., 1989, 1990: 100 km; Muñoz, 1992: 147 km; Beaumont et al., 2000: 165 km; Mouthereau et al., 2014: 142 km; Espurt et al., 2019a: 124 km), and ∼ 75–80 km in the west (Teixell, 1998).
Thermochronological ages show a southward-younging pattern, from 70 Ma to about 20 Ma, reflecting the tectonic growth of the Pyrenean mountain chain toward the lower Iberian (Ebro) plate (Fitzgerald et al., 1999; Sinclair et al., 2005; Gibson et al., 2007; Metcalf et al., 2009; Whitchurch et al., 2011; Mouthereau et al., 2014; Vacherat, 2014; Vacherat et al., 2016, 2017; Waldner et al., 2021). Orogenic growth accelerated by the middle-late Eocene (50–40 Ma) as recorded by exhumation rates, foreland flexure migration, sediment accumulation and paleoelevation estimates (Sinclair et al., 2005; Huyghe et al., 2012; Curry et al., 2019; Garcés et al., 2020). Among the possible drivers of exhumation is the increasing strain localisation in mid-crustal levels during accretion of necking and proximal domains of the Iberian margin (Mouthereau et al., 2014; Jourdon et al., 2019; Ternois et al., 2019). More recent studies have emphasized additional post-20 Ma uplift caused by isostatic and tectonic adjustment to density contrasts inherited from the Cretaceous rifting in the West (Fillon et al., 2021) and rifting of the Gulf of Lion in the eastern Pyrenees (Huyghe et al., 2020).
Of particular interest to our reconstructions is the amount and timing of left-lateral strike-slip movement between Europe and Iberia from Late Jurassic to Albian that has been matter of debate for decades. A consensus has recently emerged in reconstructions accounting for strain partitioning between strike-slip and extension over a large region, encompassing the Parentis and Basque-Cantabrian basins, the Iberian basins (Cameros, Maestrat) and the Valencia and Columbretes basins (Jammes et al., 2009; Tugend et al., 2014b; Etheve et al., 2016; Angrand et al., 2020). A major change occurred in the late Aptian when the Bay of Biscay opened (Sibuet et al., 2004). From this time, extension is considered to have occurred mostly orthogonal to the Pyrenean rift axis (e.g., Jammes et al., 2009). For instance, Angrand et al. (2020) estimated 19 km of left-lateral movement during the Aptian-Albian accommodated in the Pyrenees while 259 km were accommodated by transtension since the Permian.
Basque-Cantabrian Basin
The Basque-Cantabrian Basin, located west of the Pyrenean belt (Fig. 3b), represents an inverted hyper-extended Cretaceous extensional basin that developed in response to the incomplete closure of the southern margin of the Bay of Biscay, between the Alvesa Platform and the Landes High (Alonso et al., 1996; Pulgar et al., 1996; Gallastegui et al., 2002; Roca et al., 2011; Pedreira et al., 2015; Quintana et al., 2015; Pedrera et al., 2017).
Following the collapse of the Variscan orogen and Permian Ca-K magmatism, post-orogenic extensional basins formed during the late Permian-Triassic (López-Gómez et al., 2019). Subsidence resumed during the Late Jurassic (Tithonian)-Early Cretaceous after a period of tectonic quiescence associated with deposition of Hettangian-Kimmeridgian (200–160 Ma) marine carbonate platforms (Cámara, 2017; Pedrera et al., 2017; Martín-Chivelet et al., 2019), followed by deposition of Tithonian-Barremian (140–125 Ma) shallow marine sediments. By the late Aptian-early Albian, synrift subsidence increased as revealed by the deposition of 3 km deep-marine sediments. The total synrift sequence (145–100 Ma) is 12 km thick, locally exceeding 20 km (Quintana et al., 2015; Pedrera et al., 2017), and was accommodated by extreme crustal extension leading to sub-continental mantle exhumation (DeFelipe et al., 2017; Pedrera et al., 2017; Ducoux et al., 2019). Despite the Basque-Cantabrian Basin forms a distinct rift segment from the Pyrenees to the West of the Pamplona Fault (Fig. 3b), it shows important similarities, notably the high temperature (HT) metamorphism (Cuevas and Tubía, 1999; DeFelipe et al., 2017), dykes and laccoliths of an alkaline magmatism dated as late Albian (102 Ma) (Ubide et al., 2014; Agirrezabala, 2015).
The Aptian-Albian synrift evolution of the Basque-Cantabrian Basin has soon been interpreted as reflecting deformation in a transtensional regime. It is debated whether strain was partitioned between N-S (or NNW-SSE) extension and NW-SE strike-slip faults (García-Mondéjar et al., 1996; Cámara, 2017), or is not partitioned but rather related to NNE-SSW extension oblique to the E-W inherited Landes High and Duero-Ebro platforms (Pedrera et al., 2020a). As noted by Angrand et al. (2020), distinguishing between the two models may be challenging as the sedimentary cover on which interpretations are based is detached in the Triassic salt layer and thus decoupled from the larger scale crustal tectonics.
In the Basque-Cantabrian Basin, the Ventaniella Fault (Fig. 3b) is part of a NW-SE fault system that acted as left-lateral shear zone during the Late Jurassic-Early Cretaceous and has been subsequently inverted with a right-lateral kinematic during the Cenozoic (De Vicente et al., 2011; Tavani et al., 2011; Cámara, 2017). These faults have a Triassic origin (Tavani and Granado, 2015). Tectonic activity along these faults gets younger northeastward (Ubierna fault: Late Jurassic-Early Cretaceous; Zamanza-Oña fault: Early-Middle Cretaceous; salt tectonics in the center of the basin, Cámara, 2017). Angrand et al. (2020) suggested that the Iberia-Ebro displacement has possibly been distributed along these structures.
Contraction in the Cantabrian belt occurred synchronously with the main continental orogenic phase of the Pyrenees (Quintana et al., 2015; Macchiavelli et al., 2017). Post-Santonian N-S horizontal shortening is in the same order of magnitude as the Pyrenees (97 in the upper crust and 122 km in the middle-lower crust, Quintana et al., 2015).
Despite the low-temperature thermochronology analyses resolve an onset of exhumation at 50–45 Ma in the Cinco-Villas (DeFelipe et al., 2019), the preservation of thick Eocene-Oligocene turbidites indicates the Basque-Cantabrian basin remained a depression during the early stages of rift inversion before a main exhumation in the late Eocene-Oligocene (Fillon et al., 2016).
Iberian Range and Central System
The Iberian Range located south of the Ebro basin (Fig. 3c) is composed of the Aragonese branch (north) and the Castillan-Valencian branch (south), separated by the Teruel high, Montalbán high and the Cenozoic Almazán basin (northeastern Duero basin) (e.g., De Vicente et al., 2011). The Iberian Range is interpreted to result from the tectonic inversion of a series of late Paleozoic-Mesozoic rift basins that delimit the Ebro continental block from South Iberia. Three main Mesozoic rift basins are recognized: the Cameros basin on the northwestern part of the Aragonese branch, the Maestrat basin on the southeastern part of the Iberian Range (e.g., Salas et al., 2001) and southeastward the Columbretes basin positioned offshore in the Valencia Trough (Roca, 1992; Etheve et al., 2018).
A thick series of early Permian (up to 2 km) volcano-sedimentary deposits is documented in the Central System associated with localized subsidence in post-Variscan extensional basins (Sopeña et al., 1988; Arche and López Gómez, 1996). Rapid subsidence is described in the Maestrat basin in the Early and Late Jurassic separated by a period of slow subsidence in the Sinemurian (200–190 Ma) (Van Wees et al., 1998). The latter phase is particularly obvious in the Cameros basin where it forms an asymmetric sag basin containing up to 6.5–8 km of Late Jurassic to Early Cretaceous synrift continental deposits with episodic marine incursions (Guiraud and Seguret, 1985; Alonso and Mas, 1993; Casas-Sainz and Gil-Imaz, 1998; Quijada et al., 2010; Suárez González et al., 2010; Omodeo Salè et al., 2014; García-Lasanta et al., 2017). The postrift sequence consists of 100–500 m of late Albian–Cenomanian sandstones (Utrillas Formation) and 200 m-thick Late Cretaceous marine limestone. The Cameros basin recorded greenschist facies metamorphism at 108–86 Ma (Del Río et al., 2009; Casquet et al., 1992) associated with peak temperatures of 330 °C (Rat et al., 2019) and fluid flow (Omodeo Salè et al., 2014) during the opening of the Bay of Biscay. This event was contemporaneous with metamorphic and volcanic events described in the Pyrenees and Basque-Cantabrian basins (see above). Similar synrift sequences are reported in the Maestrat and Valencia basins, although they are also recognized to be slightly older, Oxfordian-Berriasian (155–142 Ma) and Valanginian-early Albian (139–109 Ma) in age. The subsidence was also lower, characterized by the deposition of 2–4 km of marine carbonates (Van Wees et al., 1998; Salas et al., 2001; Aurell et al., 2019a). In contrast to the Cameros basin, the stratigraphic architecture and depositional environments of the Maestrat and Valencia basins show a clear affinity and polarity towards the Tethys Sea to the East.
Low-temperature thermochronology defines a 60 Ma cooling stage related to the onset of tectonic inversion (Rat et al., 2019). This timing is in agreement with Campanian deformation described in the Maestrat basin (Vergés et al., 2020). Exhumation then increased in the Cameros basin during the late Eocene-Oligocene (40–25 Ma), regionally in good agreement with the timing of exhumation in the Pyrenees. Thrusting north of Cameros basin appears to have been ongoing from 15 to 9 Ma.
The Iberian Range has received a particular attention for its role in accommodating a portion of the eastward motion of Iberia before and during the Early Cretaceous (Tugend et al., 2014b; Nirrengarten et al., 2018; Aldega et al., 2019; Angrand et al., 2020). Angrand et al. (2020) indeed propose to transfer the left-lateral strike-slip movement between Europe and Iberia (∼ 245 km) in the Iberian basins during the Aptian-Albian rather than in the Pyrenean rift basin, which instead accommodates orthogonal extension (only 19 km of strike-slip motion). The High Tagus Fault (Castillan-Valencian branch), which has been active during the Middle-Late Jurassic (Aurell et al., 2019b) and the Rio Grio Fault (north Aragonese branch/north Cameros) active since the Triassic (Aldega et al., 2019) indicate that this domain must be accounted for in tectonic models. A total shortening of ∼ 75 km is estimated across the entire Iberian Range (Guimerà et al., 1996; Salas et al., 2001). During the late Eocene-Miocene, the Iberian Range accommodated ∼ 40 to 50 km of N-S horizontal shortening (Salas et al., 2001; Guimerà et al., 2004; De Vicente et al., 2009; Mouthereau et al., 2014) and the Central System records 25–50 km of NW-SE horizontal shortening during the Cenozoic (Arche and López Gómez, 1996).
Catalan Coastal Range, offshore Valencia Trough and Balearic islands
The Catalan Coastal Range results from the inversion of Mesozoic rift basins that were subsequenty reactivated in extension during the Neogene opening of the Western Mediterranean Sea. The first stages of extension are reflected in Permian-Triassic rift basins and magmatic rocks exposed in the Catalan Coastal Range, Mallorca and Menorca. They were associated with opening of the Neotethys (Galán-Abellán et al., 2013; López-Gómez et al., 2019; Angrand et al., 2020, and references therein). These sediments lie above a Paleozoic basement that have affinity with the Pyrenees, Sardinia and Corsica and slight differences with the Iberian Range. Synrift subsidence started in the Middle Jurassic (Aalenian-Oxfordian; 174–160 Ma), indicating eastern Iberia formed part of the paleomargin of the Alpine (Ligurian) Tethys (Gómez and Fernández-López, 2006).
The Valencia Trough, bounded to the north by the Catalan Coastal Range (Fig. 1a), is interpreted as a NE-SW Oligocene-Miocene extensional basin formed as a result of the opening of the West Mediterranean (Liguro-Provençal basin, Fig. 2) (Bartrina et al., 1992; Roca et al., 1999; Jolivet et al., 2020). However, its connection with the Maestrat basin suggests that the Columbretes basin (southwest Valencia Trough) experienced a Late Jurassic-Early Cretaceous extension (157–105 Ma) (Etheve et al., 2018) and a middle Albian to Maastrichtian slow subsidence (Salas and Casas, 1993; Roca et al., 1994; Salas et al., 2001). The study of the pre-Neogene architecture in the Columbretes basins indeed reveals that this basin, made of 8–9 km thick Mesozoic succession initiated in the Triassic (Roca, 1992), recorded a Late Jurassic/Early Cretaceous extensional phase with hyper-extension recorded during the Albian (Etheve et al., 2018; Roma et al., 2018). The final crustal thickness is estimated at 8–10 km (Roca et al., 1999).
In the Baleares, including Ibiza and Mallorca, the Mesozoic series show the same synrift evolution as described in the Maestrat basin and the Catalan Coastal Range, including a Late Triassic-Early Jurassic rifting (Tethys event) and Late Jurassic-Early Cretaceous rifting (Cameros, Maestrat rifting) with a Middle Jurassic hiatus in between (Etheve et al., 2016; López-Gómez et al., 2019). Stratigraphic and sedimentary constraints therefore collectively indicate that the Balearic promontory lies on the paleomargin of Iberia separated from Ebro by a failed Late Jurassic-Early Cretaceous rift that connects the Valencia to the Maestrat and Cameros basins. Inversion of the Mesozoic rift that formed the Catalan Coastal Range occurred during the lower Lutetian (∼ 50 Ma) (Roca and Desegaulx, 1992; Roca et al., 1999; López-Blanco, 2002; Gaspar-Escribano et al., 2004; Gómez-Paccard et al., 2012; Garcés et al., 2020). A contraction event of the same age is inferred from the upper Oligocene-Miocene unconformity in the Balearic promontory (e.g., Etheve et al., 2016), which likely corresponds to the Pyrenean collision (Fig. 2). Extension started in the early-late Oligocene (Chattian, 28 Ma) in the Catalan-Valencian domain (Roca and Desegaulx, 1992; Roca et al., 1999), preceding the opening of the Gulf of Lion (Anadón et al., 1985; Bois, 1993; Jolivet and Faccenna, 2000) (Fig. 2). Extension appears to have been ongoing until the late Langhian (∼ 14 Ma) in the southern Valencia Trough (Roca and Desegaulx, 1992), that is until the end of Sardinia-Corsica rotation.
The view that shortening affected the southern Valencia basin is however disputed as it requires that the Balearic promontory overthrusts the Valencia basin for which there is limited evidence (Maillard et al., 1992). The late Oligocene-early Miocene extensional event is recognized locally in Ibiza (Etheve et al., 2016) but not in Mallorca, where instead the interval between late Oligocene and mid-Serravallian corresponds to a shortening of about 84 km (Vergés and Sàbat, 1999; Sàbat et al., 2011). This shortening is assumed to balance the extension associated with the opening of Valencia Trough (Vergés and Sàbat, 1999) but the age of the extreme crustal thinning in the Columbretes basin dated to Early Cretaceous (Etheve et al., 2018) casts doubt on the hypothesis of large Neogene extension. In any case, the compression was probably short-lived as extension renewed during the late Tortonian-Serravalian (12–8 Ma).
East of the Balearic Promontary, the North-Balearic-Fracture-Zone (see Maillard et al., 2020, and references therein) forms a transfer zone that accommodated the differential extension between the Valencia basin and the counterclockwise rotation of Sardinia-Corsica and extension in the Liguro-Provençal oceanic basin. The first calc-alkaline magmatic event is dated as late Oligocene-lower Miocene in the Valencia basin (24 Ma; Martí et al., 1992; Fig. 2) coincident with Ca-K magmatism of Aquitanian age recognized in Sardinia. Because at this time Sardinia aligned with the Valencia basin, this Ca-K magmatism must reflect a pre- or early-synrift event (Maillard et al., 2020).
Most of the current controversy are related to the pre-35 Ma reconstruction of the Baleares margin, including the Sardinia-Corsica block. They differ as to whether Baleares-Sardinia-Corsica block is positioned 1) in the lower plate (Iberia), with AlKaPeCa units in the upper plate (Handy et al., 2010; Leprêtre et al., 2018), which requires a polarity reversal as the W-Mediterranean slab retreated, or 2) in the lower plate, together with AlKaPeCa units, as most classically reconstructed (Jolivet and Faccenna, 2000; Rosenbaum et al., 2002a; Lacombe and Jolivet, 2005; Faccenna et al., 2014; van Hinsbergen et al., 2014). It is worth noting, that the solution of Handy et al. (2010) assumes the pre-35 Ma SE-directed subduction of a Western Ligurian basin (also named Betic Ocean or Nevado-Filabrides Basin; Guerrera et al., 2014).
Betic Cordillera and Alboran domain
The Betic Cordillera resulted from Cenozoic subduction along the South Iberian paleomargin and post-orogenic extension associated with opening of the Alboran domain (García-Dueñas et al., 1988; Balanyá et al., 1997; Crespo-Blanc and de Lamotte, 2006) (Fig. 3d). The role of slab dynamics has been emphasized to explain the arcuate shape of the Betic-Rif belt, metamorphic evolution and deep structure of the arc (Lonergan and White, 1997; Wortel and Spakman, 2000; Faccenna et al., 2014). Geodynamic interpretations have varied in terms of timing, amount and vergence of subduction (Lonergan and White, 1997; Vergés and Fernàndez, 2012; van Hinsbergen et al., 2014). Two end-member tectonic reconstructions currently exist. The first one proposes that the Alboran domain must be restored south of the Baleares Promontory in the Paleogene and then travelled for more than 1000 km westwards (van Hinsbergen et al., 2014). Alternatively, the Alboran domain that did not exist before the Neogene (Fig. 2) was part of the Iberia paleomargin that subducted below Africa during the Paleogene; implying a limited convergence of 200–300 km (Vergés and Fernàndez, 2012). Time-temperature paths obtained from the Cretaceous and Miocene Flysch Complex (Daudet et al., 2020) and recent structural balanced cross-sections (Pedrera et al., 2020b) constrain a scenario in which the Alboran basement formed an hyper-extended rift segment of the Atlantic-Tethys that was inverted between Iberia and Africa in the Paleogene. Pedrera et al. (2020b) estimate a shortening of 100–145 km, achieved between the early Eocene and early Miocene. This value falls in the range of the total Cenozoic convergence reconstructed between Iberia and Africa (Macchiavelli et al., 2017).
Above the Variscan basement of the Alboran (AlKaPeCa) domain, which underwent a high-pressure/low-temperature metamorphism during the Paleogene (García-Dueñas et al., 1988; Soto and Azanon, 1994; Augier et al., 2005), lies the less metamorphosed Malaguide complex that preserves continental to coastal Triassic facies (Algarra, 1987; Ortí et al., 2017). In the external and Prebetic domains, the late Permian/Early Triassic series from 270 to 200 Ma (Fig. 2) record rifting on the south-Iberian margin (López-Gómez et al., 2019). This first synrift sequence is intruded during the Jurassic by MORB-type gabbro dated at 180 Ma (Puga et al., 2011), forming the West Ligurian basin or the Betic ocean. Whether this is related to the onset of oceanic spreading or to hyper-extended margin magmatism (Durand-Delga et al., 2000) remains unclear. A second phase of rifting is recorded from Late Jurassic (Bathonian, ∼ 170 Ma) to Early Cretaceous (Albian, 100 Ma) (Hanne et al., 2003; Martín-Chivelet et al., 2019).
Subsidence resumed in the early Cenozoic in the Malaguide Complex from the Danian (66–61 Ma), lower-middle Cuisian (Late Ypresian, 52–50 Ma), and middle Lutetian-Bartonian (45–38 Ma) (Hanne et al., 2003) and from Thanetian (56 Ma) to late Ypresian (50 Ma) associated with the development of a turbiditic basin in the pre-Betic (Martín-Chivelet and Chacón, 2007). The Eocene calcareous turbidites witness the initiation of foreland subsidence (Serrano et al., 1995; Martín-Martín et al., 1997; Maaté et al., 2000; Guerrera et al., 2005), which coincides with the onset of cooling in the Flysch Complex (Daudet et al., 2020). By the early Miocene (Aquitanian, 22 Ma) rapid subsidence occurred but this event was short-lived as indicated by a regional rapid cooling event at 20–15 Ma. This markedly synchronous phase of sediment burial and exhumation suggest a phase of rapid thinning and heating, followed by fast thermal reequilibration and denudation. Subsidence resumed in the early Tortonian (17–15 Ma), during which orogen-parallel extension occurred. From the late Messinian (5.3 Ma) onward, transition from marine to continental sedimentation was associated with regional uplift of up to 1000 m of the intramontane Neogene basins (Iribarren et al., 2009; Janowski et al., 2017) triggered by westward propagation of slab tearing (or detachment) below the Betics (Spakman and Wortel, 2004; Mancilla et al., 2015).
Crustal segmentation of the West European realm
Inverted basins and mountain chains described above form a broad zone (> 1000 km) of intra-continental contraction around Iberia that resulted from the shortening of the variably extended continental crust and closure of oceanic domains during convergence between Europe and Africa. The strong crustal segmentation of the European continent, including Iberia, appears to be inherited from the Mesozoic rifting history itself established on an older highly heterogeneous Paleozoic, mainly Variscan, crustal architecture. The different orogens and inverted basins delimit deformed crustal domains marked by less intensely deformed and undeformed continental bocks (Fig. 4). Smaller amplitude chains (Catalan Coastal Range, Central System) and crustal- to lithospheric-scale faults (Nazaré Fault, Messejana-Plasencia Fault, Fig. 4) further express the segmentation of the Iberian plate into smaller continental blocks (see below the methodology in Sect. 5.1). These different crustal elements are essential to precisely reconstruct the evolution and kinematics of the Southwest European realm.
Iberia
We divide the Iberian plate into three main domains: the Ebro block, the West Iberia block, and the South Iberian block (itself divided into two sub-domains) (Fig. 4). Hereafter, we refer to Iberia as the combination of West and South Iberia blocks. It is delimited from the Ebro continental block by the Iberian Range, a major deformation zone aligned in the NW-SE direction and stretching from the Basque Cantabrian Basin to the Valencia Basin.
The Ebro continental block is delimited by the Pyrenees to the North and the Iberian Range to the South, and by the Basque-Cantabrian Basin and the Catalan Coastal Range to the West and East, respectively (Fig. 4). Recent studies place the Ebro block within the Iberia-Europe plate boundary (Tugend et al., 2015; Angrand et al., 2020). Angrand et al. (2020) estimated a total strike-slip movement between Iberia and Europe of more than 200 km since the Triassic that has been accommodated on both the northern and southern borders of the Ebro block, in the Pyrenees and the Iberian Range, respectively.
Although we consider the Ebro block to be rigid during the Mesozoic, basin analyses (Vargas et al., 2009) and geophysical data (Wehr et al., 2018) beneath the Cenozoic foreland sequence argue for subsidence during the Early Mesozoic rifting. In addition to the north and south-bounding rifts, distributed deformation within the Ebro block is suggested but remains poorly constrained in terms of kinematics as no major structure can be identified within this block. At the scale of the kinematic model and for the sake of simplicity, we here adopt a rigid Ebro block hypothesis with the deformation localized at its borders.
The West Iberia block stretches westwards into the Atlantic continental margin and is delimited by the southern North Atlantic Ocean to the West and by the Basque-Cantabrian basin and Iberian Range to the East (Fig. 4). A N-S strike-slip fault system, along which ∼ 50 km of Cenozoic left-lateral motion is accommodated (De Vicente et al., 2011) is recognized in the West Iberia block. For this reason, we separate the West Iberia block into two sub-domains (Fig. 4). The West Iberia block is delimited to the South by the Central System-Nazaré Fault.
The Nazaré Fault is a major transfer zone across the West Iberian margin (Pereira et al., 2017), that separates the West Iberia block to the north from the South Iberian block (Fig. 4). This is an important feature of the reconstruction as it minimizes the overlap of northwest Iberia (Galicia) with the Flemish Cap (North America), or the gap between southwest Iberia and Newfoundland during the late Paleozoic full-fit (Angrand et al., 2020). However, no consensus currently exists on the amount of displacement along this structure, whilst its extension over 250 km (Pereira et al., 2017) and its connection to the Avalon Fault during the opening of the North Atlantic (Fernández, 2019) suggest it is a major crustal boundary. In our new reconstruction, however, we remained conservative so that we have reduced the absolute strike-slip movement to ∼ 100 km from 145 to 110 Ma with respect to Angrand et al. (2020).
The South Iberian block is delimited to the northwest by the Central System-Nazaré Fault, to the northeast by the Iberian Range, and to the south by the Betic Cordillera (Fig. 4). The Messejana-Plasencia Fault in the South Iberian block is another major strike-slip fault lineament inherited from the late Paleozoic (Pereira and Alves, 2013) and is associated with CAMP-related doleritic magmatism at the Triassic-Jurassic boundary (Cerbiá et al., 2003). Kinematics across this ∼ 530-km long structure is difficult to ascertain. Although it has been proposed to accommodate transtensional deformation throughout the Mesozoic (Pereira et al., 2017), Paleozoic markers are barely displaced across this feature. During the Cenozoic, small-amplitude left-lateral pull-apart basins developed along the Messejana-Plasencia dyke (De Vicente et al., 2011). Therefore, a limited displacement of about 30 km has been estimated in our reconstruction.
Mallorca and Menorca islands (northern Balearic Promontory, BP; Fig. 4) were attached to southeast Ebro until the opening of the Valencia Trough during the Neogene and are not much affected by the pre-Neogene deformation (see Sect. 3.6). Offshore Ibiza, in the Columbretes basin, crustal thinning during the early Cretaceous argues for the eastern propagation of the intra-Iberian rift during the opening of the Bay of Biscay (Roca, 1992; Etheve et al., 2018).
South-East France and Sardinia-Corsica
The France Southeast Basin recorded a succession of tectonic events since the late Paleozoic, including Early Jurassic and Early Cretaceous crustal thinning, Cenozoic contraction and extension (Baudrimont and Dubois, 1977; Lacombe et al., 1992; Lacombe and Mouthereau, 2002; Espurt et al., 2012; Bestani et al., 2015, 2016; see Sect. 3.1). On this basis, we included the rift basins of South-East France, the Provence and the Vocontian basins, into a separated Provence continental domain, which is bounded to the west by inherited NE-SW crustal-scale faults (e.g., Nîmes and Durance faults) and to the north and east by the Alpine orogen.
The similarities between the timing of collision, metamorphic and post-collisional events in the Variscan basement (Rossi et al., 2009), including the late Carboniferous-Permian deformation and volcanism (Maures-Estérel Massif and Corsican Monte Cinto) and paleomagnetic directions between the Provence and the Sardinia-Corsica block (Maures-Estérel-Corsica-Sardinia – MECS – of Edel et al., 2014, 2015), indicate they formed a single block prior the Early Cretaceous rifting. The question of whether Sardinia-Corsica belongs to the European plate or to the Iberian plate is debated. Reconstructions that consider Sardinia-Corsica as part of Europe (e.g., van Hinsbergen et al., 2020) are based on paleomagnetic data in Sardinia that indeed indicate a different rotation history for Iberia at 121 Ma (Advokaat et al., 2014). It further implies a shortening of 400–500 km between Iberia and southern Sardinia of which there is no geological evidence.
Together with sedimentary affinities between Paleozoic sedimentary rocks documented in the eastern Pyrenees, the southern Sardinia, the Montagne Noire, and the SW France (e.g., Padel et al., 2018; see Sect. 3.6), paleomagnetic constraints during the late Paleozoic (Edel et al., 2014) indicate that Ebro, Catalan Coastal Range, Sardinia and Corsica formed a coherent continental block until at least the Triassic (Edel et al., 2014, 2015). We therefore adopt a solution in which Sardinia-Corsica belongs to the Ebro block. Hence, the Maures-Estérel-Sardinia-Corsica domain was part of the same Ebro continental block rather than Europe or Iberia. As such Maures-Estérel-Sardinia-Corsica domain moved eastward in left-lateral strike-slip movement relative to Europe from the late Permian to Early Cretaceous.
Greater Adria
Although focusing on the Iberian realm, our reconstruction aims to be integrated into a large-scale kinematic model that includes Iberia and Greater Adria. Greater Adria as defined by van Hinsbergen et al. (2020) comprises Adria sensu-stricto, Apulia (i.e., southern Adria), and the future Alpine terranes, including AlCaPa, i.e., the Alps, the Carpathian, and the Pannonian regions (Schmid et al., 2008; Handy et al., 2010).
According to paleomagnetic studies in NE Africa and Greater Adria, Muttoni et al. (2001) proposed that Greater Adria has been linked to Africa since the late Permian. These two plates drifted eastwards during the Early Jurassic as the Ionian basin opened (Tugend et al., 2019). During the Cenozoic, parts of the northern Alpine Tethys are accreted into the Apennines (Handy et al., 2015; Marroni et al., 2017) while the northern edge of Adria collided with Europe into the Alpine orogen (Rosenbaum et al., 2002b; Handy et al., 2010, 2015).
Some recent kinematic reconstructions accounted for internal deformation between Adria to the north and Apulia to the south to avoid overlap between Adria and Sardinia-Corsica prior to the opening of the Alpine Tethys (Le Breton et al., 2017, 2021; Müller et al., 2019). But the lack of geological evidence precludes precise quantitative estimates of the internal strain and segmentation within Greater Adria. In order to avoid overlap between Adria and the Balearic Promontory, over 200 km of intraplate shortening along the precursor of the Mid-Adriatic Ridge has been proposed (Müller et al., 2019), which is much larger than in other reconstructions (e.g., ∼ 30 km for van Hinsbergen et al., 2020). In this study, we choose another approach. The kinematics of Adria is reconstructed based on the kinematics of Apulia, which has NE Africa as a reference plate (Muttoni et al., 2001). As a consequence in our reconstruction we assume a left-lateral strike-slip movement of ∼ 100 km between Adria and Apulia.
In the Western Alps, crustal domains have been restored relative to the Penninic Front (Ramsay, 1981; Butler, 1985), which represents the neckline of the European margin (Mohn et al., 2014). In the footwall of the Penninic Front, the crustal domain is deformed only during the orogeny (Ultrahelvetic nappes; Butler, 1985), while in the hanging wall, the Penninic units were deformed during the Tethyan extension and inverted during the Alpine orogeny.
AlKaPeCa
The Neogene evolution of the Western Mediterranean orogenic arcs (Gibraltar and Tyrrhenian) has been related to the tectonics of the AlKaPeCa microcontinent or terrane that originated from the southern European margin (Bouillin et al., 1986) or alternatively from a Mesomediterranean microplate stretching between Africa/Adria and Europe during the Mesozoic. The existence and origin of these two models (see Guerrera et al., 2019, for a recent review of the different published models) are basically related to two different Mesozoic reconstructions of the Alpine Tethys, including its connection to the Central Atlantic. The main difference between those end-member models is linked to the existence or not of an oceanic domain north of AlKaPeCa. Reconstructions that restore AlKaPeCa on the southern margin of Iberia consider a single oceanic domain to the south between Europe and Africa that represents the Alpine Tethys (Bouillin et al., 1986; Lonergan and White, 1997; Lacombe and Jolivet, 2005; Billi et al., 2011; Schettino and Turco, 2011; van Hinsbergen et al., 2014). In a recent reconstruction, Angrand et al. (2020) accounted for the presence of a hyper-extended domain in the Betics, instead of an oceanic spreading center, that actually separates a “Briançonnais-like” rifted block representing Alboran and the Kabylides terranes. The second category of models adopts a solution with two oceanic branches, the Maghrebian Tethys (or East Ligurian Tethys) connected to the Lucanian basin in the east and the Nevado-Filabrides ocean (or Betic ocean) in the north, that is assumed to represent the Alpine Tethys through the connection with a West Ligurian ocean (Guerrera et al., 2005; Michard et al., 2006; Handy et al., 2010; Guerrera and Martín-Martín, 2014; Leprêtre et al., 2018).
Here, we adopt a different approach by suggesting that Alboran and Kabylides terranes are disconnected from the Peloritan and Calabrian blocks. This is supported by the following arguments. Alboran, including the internal parts of the Betics-Rif orogen, and Kabylides shared the same extensional history on the western termination of the Alpine Tethys (see Sect. 3.7), which is also very similar to the tectonic evolution of North Africa (Chalouan and Michard, 1990; Michard et al., 2006). In contrast, the Peloritan-Calabrian domain to the east, once formed the distal margin of Sardinia-Corsica on the northern Alpine Tethys margin (Vitale et al., 2013, 2019; Vitale and Ciarcia, 2013) later rifted from Sardinia during the Neogene due to back-arc extension in the Tyrrhenian Sea (Rossetti et al., 2001; Faccenna et al., 2014). HP metamorphism is reported to have occurred during the Paleogene in Alboran, Kabylides and in Peloritania (see above Sects. 3.2 and 3.7). This indicates these originally dispersed Mesozoic rift units all underwent subduction during Africa-Europe convergence, indicating a similar tectonic evolution during the main Cenozoic compressional event.
Kinematic reconstruction of the Africa-Europe plate boundary: from Tethys and Atlantic rifting to opening of the West-Mediterranean Sea
Implementation of recent kinematic models and reconstruction workflow
In addition to conform with geological constraints regarding the direction and amount of shortening/extension, and timing of deformation all over Iberia and in the South France, our kinematic model presented in Figures 5–12 also builds on the revision and updates of recent reconstructions of Africa, Adria, and Eastern Europe.
From 270 to 100 Ma, our reconstructions update the former kinematic model of Angrand et al. (2020). We remind that this study itself is based on the implementation of previous regional kinematic models of the North Atlantic (Barnett-Moore et al., 2016; Nirrengarten et al., 2018; Peace et al., 2019b), Tethyan domains (Schmid et al., 2008; Handy et al., 2010; van Hinsbergen et al., 2020) and global kinematic reconstructions (Müller et al., 2019). We have applied minor modifications to the initial model by improving the full-fit adjustment between Iberia and North America during the late Permian-Triassic stage to correct the overestimated (and unsupported) strike-slip movement along the Nazaré Fault. In addition, we modified the geometry of the southern termination of the Alpine Tethys as well as the Ionian ocean to better represent the variable degree of crustal stretching and the tectonic evolution of oceanic domains (now consummed or preserved) in the Western Mediterranean domain.
After 100 Ma, the reconstruction of Eastern Europe follows that of van Hinsbergen et al. (2020). Kinematics of Africa and Apulia follows Müller et al. (2019) and Le Breton et al. (2021), but we adopt a solution involving much smaller displacements between Apulia and Adria. The reconstruction of Alboran (Internal Betics) back to 60 Ma is based on the original position of the Cretaceous and Tertiary Flysch units of the Alboran domain which are restored 150 km to the south of present-day subduction front. This solution is supported by tectonic reconstruction of the Betic thrust wedge by Pedrera et al. (2020b). They are both consistent with the amount of convergence estimated between South Iberia and Africa (Macchiavelli et al., 2017). The kinematic evolution of Sardinia-Corsica from 50 Ma onward follows Frasca et al. (2017). From 35 Ma onward, the kinematics of the West Mediterranean is based on Romagny et al. (2020) with the noticeable exception for Alboran that we have placed farther west at 35 Ma (Daudet et al., 2020).
These input models were then updated according to the geological constraints presented in Section 3. These constraints include: (1) age of rifting, mantle exhumation, onset of oceanic spreading in the Atlantic; (2) the present-day position of ophiolite bodies and the timing of rifting, oceanic spreading and subduction in the Tethyan domains (Paleotethys, Neotethys, Pindos, Meliata, Vardar); (3) from 84 to 66 Ma, the closure exhumed mantle domain in the Pyrenean basins; (4) from 66 to ∼ 25 Ma, continental deformation throughout Iberia and most of Western Europe (5) from ∼ 35 Ma onward, continental deformation occurs in the Alps, coeval with the opening of the Gulf of Lion and the West Mediterranean domain.
Our reconstruction workflow relies on the use of GPlates version 2.2 (Müller et al., 2018), an open-source plate tectonic geographic information system software that allows to visualize reconstructed data through time and to build evolving plate tectonic models. Stable continental domains are represented as rigid polygons.
Because of the inherent complexity associated with the reconstruction of highly segmented domains, defining the relationships between moving plate and their reference plate is not straightforward. In practice, we have defined new rigid polygons for Iberia (Galicia, Ebro, SW and SE Iberia), representing the three main continental domains presented in Section 4.1 (SW and SE Iberia being two sub-domains of South Iberia) separated by deformable domains. The plate circuit of Iberian continental blocks has been built as follows.
(1) West Iberia (Galicia) kinematics is based on the reconstruction of the southern North Atlantic, whose evolution is constrained by study of oceanic magnetic anomalies (Nirrengarten et al., 2018).
(2) The Ebro block has a much complex kinematic history. From 270 to 145 Ma, Ebro is attached to Europe, whereas from 145 to 118 Ma, Ebro is moving along with Galicia which movement is driven by the opening of the southern North Atlantic. From 118 to 84 Ma, Ebro motion follows that of Southeasth Iberia to take into account the N-S extension associated with the opening of the Bay of Biscay. Finally, the Ebro block is reconstructed relative to Europe from 66 Ma, once the Pyrenean rift domain is closed.
(3) The southwest and southeast Iberia are two sub-domains of southern Iberia that reflect minor displacement along the Messejana-Plasencia Fault. Southeast Iberia is attached to southwest Iberia, the latter being attached to Galicia.
We present our reconstructions into successive time frames at different spatial scales so as to emphasize 1) the large scale plate configuration between whole Europe and northern Africa highlighting the connection between Atlantic, Alpine Tethys and Neotethys (Figs. 5–8), and 2) a more regional scale reconstruction of the Mediterranean region putting more emphases on the main structures, microplates, strain distribution, and paleogeography (i.e., continental domain versus main rift/orogenic basins) (Figs. 9–12). Results of the kinematic model are synthesized as rotation poles for the main plates in Table 1 (see also Supplementary Material 2 for GPlates files).
Plate reconstruction
Late Permian-Lower Triassic (270–200 Ma)
The breakup of Pangea during the late Paleozoic was accompanied by the closure of the Paleotethys and the opening of the Neotethys to the south (Figs. 5 and 6) (Stampfli et al., 2001; Stampfli and Borel, 2002). The northward drift of the Cimmerian terrane associated with the subduction of the Paleozoic Paleotethys Ocean (e.g., van Hinsbergen et al., 2020), resulted in the opening of back-arc Pindos and Meliata oceanic basins during the Early (∼ 250 Ma) and Late Triassic (Carnian, 220 Ma), respectively (Fig. 5) (Channell and Kozur, 1997; Stampfli et al., 2001).
The westward propagation of the Neotethys into the European lithosphere resulted in distributed extension in Western Europe and North Africa, extending further north in the future North Atlantic domain (Angrand et al., 2020, and references therein; Ziegler, 1990a; Leleu et al., 2016; Tavani et al., 2018; Soto et al., 2019). From early Permian to Middle Triassic, the Arctic rift propagated southwards into a North Atlantic rift, making links in the North Sea basins with the Neotethys ocean (Goldsmith et al., 2003). This is indicated by the development of numerous Permian-Triassic salt-rich basins (Van Wees et al., 2000), and by the acceleration of subsidence in the Betic, Catalonia, and Provence areas (Espurt et al., 2012; Tavani et al., 2018; Angrand et al., 2020) synchronous with the onset of crustal thinning in the Alpine Tethys (Figs. 5 and 9) (Stampfli and Borel, 2002; Schmid et al., 2008).
The exact position of the Alboran and Kabylide terranes is speculative. However, there are several lines of evidence suggesting that Alboran, and by extension the Kabylides, were located south of Iberia, between Adria and Africa, in the center of a rift zone precursor to the Alpine Tethys (Fig. 9). This is corroberated by 1) the occurrence of a Permian-Triassic rift positioned on the south Iberian margin (López-Gómez et al., 2019), north Alboran and Kabylides, including the thick evaporites in the Late Triassic in the external Betic and Pre-Betic (Flinch and Soto, 2017), 2) the facies of Triassic carbonates of the Alpujarride Complex that places the Internal Zone of the Alboran domain in the Alpine Tethys rift (Martin-Rojas et al., 2009), and 3) upper Paleozoic reconstructions that consider the Alboran Paleozoic basement close to the south Iberian paleomargin (Jabaloy-Sánchez et al., 2021). The position of the Peloritani-Calabria terrane is also difficult to precisely assess. According to the two-oceans hypothesis, the Peloritani and Calabria terranes are restored east of the Mesomediterrean microplate on the margin of the Ionian and East Ligurian rift (e.g., Perrone et al., 2006). Here we follow reconstructions that restore Peloritani and Calabria terranes east of Sardinia (present-day coordinates) prior the opening of the Tyrrhenian Sea in the Neogene (Rossetti et al., 2001; Shimabukuro et al., 2012; Vitale and Ciarcia, 2013; Milia and Torrente, 2014). We acknowledge that Calabria may have been slightly rifted from Sardinia later during the Jurassic (Malusà et al., 2016) but this is ill-defined and has not been explicitly taken into account in our reconstruction.
Jurassic (200–145 Ma)
The opening of the Central Atlantic during the Early Jurassic tiggered the northward propagation of the rift in the southern North Atlantic, on the south-western margin of Iberia (Figs. 6 and 10) (Murillas et al., 1990; Leleu et al., 2016). This stage was preceded by the Central Atlantic Magmatic Province (CAMP, Olsen, 1997; Marzoli et al., 1999; McHone, 2000), associated with magmatic intrusions in Iberia (Messejana-Plasencia dyke; Cerbiá et al., 2003), in the Pyrenees (ophitic magmatism; e.g., Azambre et al., 1987), and possibly in the Rif (Gimeno-Vives et al., 2019) predating the opening of the southern Alpine Tethys (also named the Maghrebian Tethys). Although the exact age for the onset of oceanic extension may be debated (Schettino and Turco, 2009; Labails et al., 2010; Kneller et al., 2012), the opening of the Central Atlantic in the Early Jurassic (∼ 190 Ma) (Schettino and Turco, 2009; Labails et al., 2010) is in agreement with the regional geodynamics and magmatic events (CAMP). Changes in the opening directions and rates of the Central Atlantic ocean then controlled the Africa kinematics from the Late Jurassic to Early Cretaceous (Labails et al., 2010).
Rifting and opening of Alpine Tethys initiates rifting between Iberia and Adria (Schmid et al., 2008; Marroni et al., 2017; Le Breton et al., 2021), which occurred by the reactivation of Triassic Neotethyan rift structures (Angrand et al., 2020). The exact timing of continental breakup is determined by the MORB-type gabbro at the Atlantic-Tethys junction dated as Lower to early Middle Jurassic (∼ 190 Ma) (Fernández, 2019) consistent with the age of eclogitized gabbros of ∼ 185 Ma from the Betic ophiolites interpreted as being formed at a slow-spreading ridge or ocean-continent transition at the southwestern termination of the Alpine Tethys (Puga et al., 2011) (Figs. 6d and 10d). In our reconstruction, the Betic rift between Alboran and Iberia is suggested to propagate to the north, but appears interrupted along transfer zones (possibly connected with basins of the Iberian Range) thus forming rift segments in the Alpine Tethys south of the Sardinia-Corsica-Peloritani-Calabrian domain, but not directly connected to it.
The timing of MORB-type magmatism in the Betics and North Africa is consistent, despite slightly older, with the Middle Jurassic (170–161 Ma, Fig. 6) ages for the onset of oceanic spreading in the northern Alpine Tethys as constrained in the Alps (Schaltegger et al., 2002), and in agreement with the first postrift sediments (Bill et al., 2001). This is also indicated by the subsidence associated with the southern Alpine Tethys on the paleo-Calabria margin (Bouillin et al., 1988).
The age of opening of the Ionian basin between Adria and Africa is debated. Continental extension and spreading has been suggested to have started in the late Permian-Triassic (e.g., Stampfli et al., 2001). Despite rifting did occur in the Tethyan realm since the late Permian, there are no firm constraints on oceanic spreading before the Lower Jurassic (∼ 180 Ma, Tugend et al., 2019). The opening of the Ionian basin causes Adria to rotate and move northeastward relative to Africa. Rifting triggered extension in the southern Alpine Tethys and segmentation along continental transfer faults from 180 Ma onward (Figs. 6b and 10b). The eastward drift of Adria relative to Africa at ∼ 160 Ma induced the opening of the Alpine Tethys. The resulted strain accommodated by the differential extension between the Atlantic and the Alpine Tethys along a transform plate boundary between Iberia and Africa. This transform fault temporally coincides with a pause in the north-directed ridge propagation of the Atlantic, as argued by 3D numerical modelling of ridge migration (Jourdon et al., 2020). Coeval with the counterclockwise rotation of Africa and its eastwards displacement, subduction occurs along the northern Neotethys margin and closes the Vardar ocean (Figs. 7c and 7d) (Schmid et al., 2008).
Because the tectonic reorganization during the Early to Middle Jurassic focused east of Iberia, in the Alpine Tethys, this stage defines an apparent tectonic quiescence along the Iberia-Europe boundary in the Pyrenees-Provence. This is also true for Central Iberia that recorded slow subsidence or even uplift between 200 and 160 Ma, contrasting with the previous Permian-Triassic rifting stage between Ebro and Western Iberia (Figs. 6 and 10).
During the Late Jurassic-Early Cretaceous continental rifting occurred between North Africa and the Iberian Range rift. It resulted in transtensional deformation between Ebro and Iberia, which extended farther eastward into the Alpine Tethys (Figs. 6d, 7a, 10d and 11a). This is in agreement with the segmented nature of the East Iberian margin between the ENE-WSW oriented Betics and NE-SW trending Catalan Coastal Ranges at that time (Handy et al., 2010; Frizon De Lamotte et al., 2011; Sallarès et al., 2011; Schettino and Turco, 2011; Vergés and Fernàndez, 2012; Vergés et al., 2020).
Lower to middle Cretaceous (145–94 Ma)
Rift propagation in the southern North Altantic was associated with the spread of extension over Iberia, in the Iberian Range, the Betics, the Basque-Cantabrian and the Valencia (Columbretes) basins, while the Provence and Sardinia-Corsica formed a topographic high and the Pyrenees recorded moderate extension (Figs. 7 and 11).
The opening of the southern North Atlantic induced the eastward displacement of South Iberia relative to the Ebro-Sardinia-Corsica block, which in turn moved eastwards relative to Europe. Extension reactivated earlier rift basins of Permian-Triassic (e.g., Pyrenees, Iberian Range) or Jurassic ages (southern Tethys). The obliquity of extension relative to the direction of Parentis and Pyrenees-Provence rifts caused transtensional opening in the Basque-Cantabrian and Iberia Ranges (Tugend et al., 2015; Aurell et al., 2019a; Rat et al., 2019) (Fig. 7a). Rifting in the Valencia basin (Columbretes) further reveals that extension reached the east of the Ebro block and the eastern Pyrenean rift (Corbières). We suggest extension was associated with the reactivation of ENE-SW trending faults (e.g., Nîmes Fault) inherited from the late Variscan evolution.
After a phase of stalling, the northward propagation of the Atlantic resumed at 147–133 Ma (Fig. 6d) in SW Iberia (Sallarès et al., 2013) and at 139 Ma offshore southwest Galicia margin (Mohn et al., 2015) as shown by mantle exhumation. Oceanic spreading then occurred at 135–133 Ma in SW Iberia (Figs. 7a and 11a) and at 121–112 Ma on the Galicia margin (Bronner et al., 2011; Vissers and Meijer, 2012). Migration of extension to the north (Figs. 17a and 17b) (e.g., Rat et al., 2019) ultimately led to oceanic spreading in the Bay of Biscay and hyper-extension in the Parentis and Pyrenean rift basins at ∼ 120 Ma (Jammes et al., 2009; Lagabrielle et al., 2010; Mouthereau et al., 2014; Tugend et al., 2014a; Tugend et al., 2015). From the late Aptian to Cenomanian, the direction of extension in the Pyrenean basins changed to approximately N-S (Sibuet et al., 2004; Barnett-Moore et al., 2016; Angrand et al., 2020).
The eastward movement of Iberia relative to Adria is suggested to have resulted in the closure of the southern Alpine Tethys during the Early Cretaceous between 125–90 Ma (Angrand et al., 2020; Le Breton et al., 2021) (Fig. 11b). The movement of Iberia is modest at this time but sufficient to start closing the narrow southern Alpine Tethys domain. Such Early Cretaceous contractional event implied by the model seems supported by data from the Kabylides, as there are some evidence of ductile shearing in the gneissic rocks of the paleozoic Kabylides basement dated at 128 Ma (Hauterivian-Barremenian) (Cheilletz et al., 1999). In addition, intra-oceanic subduction associated with emplacement of mantle materials from the Edough Massif of the Kabylides has been suggested (Fernandez et al., 2020). Moreover, geological constraints from the Ligurian units of the North Apennines (Marroni et al., 2017) indicate a limited sedimentary thickness (estimated at 500 m) on the Adria margin from Late Jurassic to Santonian that can hardly be accounted for by postrift subsidence. Although these data are sparse, they argue that Early Cretaceous obduction/subduction is likely to have occurred in the southern Alpine Tethys.
From the late Aptian to the Cenomanian, the southward movement of Ebro due to N-S extension in the Bay of Biscay-Pyrenean basins was responsible for transpression between SW Ebro and the eastern West Iberian block. A modest shortening of ∼ 20 km is estimated on the western Iberian basins (Fig. 11a). Such a contraction is suggested to have been recorded in the Cameros basin between the early Albian and Cenomanian, before the HT metamorphism peak at ca. 100 Ma (Casas-Sainz and Gil-Imaz, 1998) and coeval with a halt of tectonic subsidence (Omodeo-Salé et al., 2017). The axial-plane cleavages documented in the Cameros basin may indeed reflect a contraction phase, much earlier that the collision-related cooling stage at 60 Ma (Rat et al., 2019). To our knowledge there is no evidence for contraction elsewhere in the Iberian Range. At 94 Ma (late Cenomanian; Figs. 7b and 11b), extension reached its maximum in the Pyrenean arm and in the Betics, but appears limited in the Iberian Range despite the high thermal gradients and metamorphism of this age is reported in the Cameros Basin. The regional HT anomaly found close to the Bay of Biscay and Pyrenees suggests control by channelized low-viscosity asthenospheric mantle rising below the Bay of Biscay spreading ridge and directed towards the east.
In the Late Jurassic-Early Cretaceous, a new episode of rifting occurred on the European margin associated with the development of the Valais basin (Schettino and Turco, 2011). While it was previously regarded as a western oceanic branch of the northern Alpine ocean between Adria (or Alcapia of Handy et al., 2010) and Europe, its timing preferably suggests a relationship with the eastward drift of the Iberian plate (Beltrando et al., 2007; Schettino and Turco, 2011) or, more exactly, with the Ebro-Sardinia-Corsica block in our reconstruction (Figs. 7a, 7b, 11a and 11b). The opening of the Valais ocean lasted until the middle Cretaceous (93 Ma) as nappe stacking in the Austro-Alpine domain started (Eo-Alpine orogeny, e.g., Handy et al., 2010) due to the northward movement of Greater Adria (Figs. 7a and 7b) (Müller et al., 2019). It is worth noting that the Pyrenees-Provence rift is not extended further east into the Alpine Tethys but rather jumps into the Vocontian basin, which arguably defined the termination of the Valais ocean. In this case, the Briançonnais block is inferred to be the continuation of the Provence High to the east.
The onset of subduction of the northern Alpine Tethys that responds to the NE-directed movement of Adria (Fig. 7a) is dated at 125 Ma in our model, which is slightly older than in other reconstructions (Schettino and Scotese, 2002: 110 Ma; Handy et al., 2010: 94–84 Ma; van Hinsbergen et al., 2020: 94 Ma). This movement implies transcurrent left-lateral movement along the western margin of Adria, which also limits the amount of subduction below the Adria margin, along the Sardinia-Corsica-Peloritani-Calabria segment, in accord with geological constraints.
Late Cretaceous to Paleogene (84–35 Ma)
The late Santonian-Campanian marks the onset of N-S convergence between Africa, Iberia and Europe, postdating the onset of N-directed drift of Adria. This kinematic change induced the acceleration of closure of Alpine Tethys in the Western Alps and the inversion of most Mesozoic rifted domains in Iberia and Ebro-Sardinia-Corsica (Figs. 7c and 11c). In the west, convergence is accommodated by underthrusting and closure of the exhumed mantle domain in the Pyrenean (Mouthereau et al., 2014; Ford et al., 2016; Teixell et al., 2016) and Basque-Cantabrian basin (DeFelipe et al., 2017; Pedrera et al., 2017). Early syn-orogenic exhumation is suggested by an early cooling episode at 75–70 Ma (Mouthereau et al., 2014; Ternois et al., 2019; Waldner et al., 2021) and a first influx of sediments accumulated in the Pyrenees in pro- and retro-foreland basins (Vergés et al., 1995; Sinclair et al., 2005; Grool et al., 2018).
In the Alpine realm, the Late Cretaceous period, from Turonian to Santonian, corresponds to the Eo-Alpine metamorphic event in the eastern Alps (Villa et al., 2000) associated with the accretion of the Austro-Alpine units to Adria and south-verging subduction of the Alpine Tethys (Schettino and Scotese, 2002; Schmid et al., 2004; Handy et al., 2010; van Hinsbergen et al., 2020). In the Western Alps, the Late Cretaceous Helminthoid Flysch sequence recorded the early Alpine deformation during the latest Cretaceous (Kerckhove, 1969; Decarlis et al., 2014) synchronous with the accretion of the Sesia-Dent Blanche Nappe to Adria (Rubatto et al., 1999; Manzotti et al., 2014; Regis et al., 2014).
During the early Paleogene (∼ 66 Ma) contraction and related exhumation spread over the whole domain from Africa to Europe (Figs. 7d and 11d), as argued in the Pyrenees (Choukroune et al., 1990; Muñoz, 1992; Vergés et al., 1995; Mouthereau et al., 2014; Ford et al., 2016), the Iberian Range (Rat et al., 2019), the Basque-Cantabrian Basin (Pedreira et al., 2007; Cámara, 2017), the Betics (Vergés and Fernàndez, 2012; Daudet et al., 2020), and in the Provence (Espurt et al., 2012; Bestani et al., 2016). In the Betics, early-stage exhumation of high-pressure rocks at 49–28 Ma is reported (Monié et al., 1991; Augier et al., 2005). There is also evidence for pre-Late Oligocene deformation in the Balearic promontory and Valencia basin (Etheve et al., 2016, 2018), and for Eocene exhumation in Sardinia and Corsica (Malusà et al., 2016). We infer that all domains comprising Iberia, the Betics, the Balearic promontory and the Ebro-Sardina-Corsica experienced contraction, uplift and moderate exhumation. The corollary is that a proto-Betic orogenic domain stretching between Europe and Africa, now dismantled by backarc extension should have existed at the emplacement of the Western Mediterranean Sea during the Paleogene (at least Eocene) (Figs. 8a and 12a).
This Paleogene orogenic system dismantled by the opening of the West Mediterranean Sea, was segmented and deformation was likely heterogenous, hence reflecting the geometrically complex margins of the Southern Alpine Tethys domain. It included a rapidly exhuming segment (Pyrenees), an uplifted but slightly deforming domain (Sardinia-Corsica), and zones of localized strain characterized by high-pressure metamorphism like the Betic-Rif domain. Closure of the Betic rift in the west and of the eastern Maghrebian basin between Kabylides and Africa/Sicilia produced the formation of the Mesomediterranean orogen (Fig. 13) that once existed as a coherent orogenic ensemble before the end of Eocene. It is worth mentioning that because deformation occurred on the margins of the AlKaPeCa continental terranes, the basement in the core of these domains were not necessarily affected by significant orogenic exhumation.
The closure of the Alpine Tethys between Corsica-Briançonnais and Adria resulted in the formation of the Western Alps (Figs. 8 and 12). Blueschist-to-greenschist facies units, including Jurassic-Cretaceous metaophiolites/metasediments (Briançonnais) and European basement (Tenda in Corsica), exhumed in the Paleocene-Eocene (58–47 Ma). This first orogenic system was followed by exhumation of eclogitic units of the distal European margin (Monte Rosa, Gran Paradiso, and Dora-Maira), and metaophiolites (Viso) in the late Eocene (44–34 Ma) (Malusà et al., 2011). Further east, deformation along the Dinarides occurred during the latest Cretaceous in association with subduction of remnants of the Vardar ocean (van Hinsbergen et al., 2020). The onset of convergence in the Dinarides system may have started in the middle Cretaceous, as attested by calc-alkaline magmatism (Handy et al., 2015).
Latest Paleogene to recent (35–0 Ma)
From the late Oligocene until the early Miocene, a major geodynamic change occurred in the geodynamics of the Western Europe. As the Alpine orogenic front advances towards the northwest, the West European Rift opened and propagated southward (Bois, 1993; Séranne, 1999; Dèzes et al., 2004; Ziegler and Dèzes, 2006). The first extensional phase (35–23 Ma) is well documented in the Provence, in the eastern Pyrenees, and in the Valencia Trough, coeval with the late activities of the West European Rift (Merle and Michon, 2001; Ziegler and Dèzes, 2006). A change from compression to back-arc extension is recorded in Alpine Corsica at 32 Ma (see Sect. 3), leading to opening of the Tyrrhenian Sea by Apenninic slab retreat eastwards. On the opposite side of the Alpine arc, late tectonic collision of the Western Alps occurred at 35 Ma as indicated by the deposition of Alpine Flysch at the west-migrating deformation front (Ford et al., 2006) (Fig. 12) and exhumation of the Paleozoic basement in the External Massifs (Bellahsen et al., 2014).
The late Eocene extensional event of the West European Rift predates back-arc opening of the Gulf of Lion and Liguro-Provençal basins that occurred during a second rift phase from 23 Ma onward (Figs. 8a, 8b, 12a and 12b). The onset of spreading in the Liguro-Provençal basin between 20.5 and 15 Ma was associated with the rotation of the Sardinia-Corsica block triggered by the eastward retreat of the subduction front toward the Ionian Tethys. This event caused a radical reorganisation of the westernmost Mediterranean Sea as indicated by the distribution of volcanism (e.g., Maillard et al., 2020). Extension and spreading migrated southwestward in the Algerian Mediterranean domain but the driver was different due to the absence of a large slab. The limited convergence between Alboran and the Iberia paleomargin highlighted by recent data (Daudet et al., 2020; Pedrera et al., 2020b) argue that the Alboran Sea opened by the westward retreat of delaminated slab Alboran mantle, leading to dismantlement of the former proto-Betic orogen. From the Tortonian (∼ 11 Ma), slab detachement and slab tearing in the margins of the Alboran Sea, in the Betics and Rif, triggered a regional uplift associated with extension and anorogenic alkaline volcanism. This late stage controlled by asthenospheric flow and magmatism is suggested to account for uplift on the western margin of the Gulf of Lion in eastern Pyrenees (Huyghe et al., 2020) as well as over the whole of Iberia (Conway-Jones et al., 2019).
N-S evolution of the Africa-Iberia-Europe plate boundary
Crustal strain distribution across the African, Iberian and European domains
To gain a better understanding of the role of crustal architecture and segmentation across the Africa-Europe plate boundary, we present a reconstruction of the spatial evolution of crustal strain along a N-S composite transect (Figs. 14 and 15).
During the late Permian-Triassic (270–240 Ma), Tethyan extension is broadly distributed between Africa and Europe and crustal thinning is limited (Figs. 14b and 14c). From the Lower Jurassic to the Early Cretaceous (180–125 Ma), the opening of the Central Atlantic and Alpine Tethys, between the Iberia paleomargin and the Internal Betics (Alboran), is reflected in the Betic rift floored by an attenuated crust or an incipient oceanic basin (Figs. 14d and 14e) (Durand-Delga et al., 2000; Puga et al., 2011; Vergés and Fernàndez, 2012). Extension is also partly accommodated by right-lateral strike-slip movement to the north of Africa, reactivating late Variscan shear zones. During the Late Jurassic-lowermost Cretaceous (125 Ma), oblique extension occurred between Ebro and South Iberia in the Iberian Range rift and in the Betics. From this stage onwards extension focused along the Ebro-Europe plate boundary.
The opening of the Bay of Biscay during the Lower to middle Cretaceous (125–94 Ma) is accompanied by hyper-extension and exhumation of the sub-continental mantle in the Pyrenean basins (Figs. 14e and 15a). This large extension north of Ebro block implies a small amount of shortening (∼ 20 km) in the other side of Ebro, in the northern Iberian Range.
The convergence between Africa and Europe started in the uppermost Cretaceous (Santonian, 84 Ma) and rapidly spread all over Europe in the Paleogene (Kley and Voigt, 2008; Mouthereau et al., 2021). Across Iberia, thrust-related exhumation is reported in the Betics and is indicated by the closure of the exhumed mantle domain in the Pyrenean basins (Fig. 15b). Between 66 Ma and 40 Ma, the advancing contraction led to progressive closure of rift basins in the Pyrenees, the Betics and the Iberian Range (Fig. 15c), characterized by the growth of orogenic topography in these regions.
From the Neogene to present (Figs. 15d and 15e) the internal reorganization of the Africa-Europe convergence, contemporaneous with the opening of the Western Mediterranean, points to a period of increasing control by sub-lithosphere processes through slab retreat and delamination. This led, along our section, to the shift of contraction from the Pyrenees to the Betics, and extension in the Alboran basin as the slab retreated westward. Following recent models that proposed a limited displacement of the Alboran basin, we interpret the back-arc extension in Abloran as the result of delamination and retreat of lithospheric mantle from below the thickened crust of the proto-Betics.
Kinematics of the Africa-Europe convergence along a N-S transect since the Late Mesozoic
In this section, we analyze the implication of our kinematic reconstruction along a N-S transect positioned to explore the relative motions of Africa, Ebro and South Iberia since the late Mesozoic. For each time step, we have projected the motion paths derived from our model onto the N10°E direction between coordinates 2.3°W/28.3°N to the south and 1.2°E/44.9°N to the north, which is taken as the mean direction of contraction since the Late Cretaceous of all blocks relative to Europe (Fig. 16a). Motion paths have been computed in each block by selecting 4 sites in northwest Africa (2.48°W/31.10°N), eastern Iberia (1.24°E/39.31°N), Ebro (0.49°E/41.66°N) and southwest France (2.02°E/47.72°N). We have also calculated shortening and extension between continental blocks by substracting their respective displacement across each orogenic segment: Africa relative to South Iberia (Betic-Rif), South Iberia relative to Ebro (Iberian Range) and Ebro relative to Europe (Pyrenees) (Fig. 16b). Because relative movements are calculated between adjacent continental blocks, the effects of rotation and obliquity with respect to direction N10°E are minimized in Figure 16b.
From 145 to ∼ 90–84 Ma, Africa, South Iberia and Ebro altogether have a left-lateral movement with respect to Europe that is decreasing northward (Figs. 10 and 11). As continental blocks move away from Europe they show cumulated extension on our N-S transect, which is the largest for Africa, decreasing northwards for Iberia and Ebro (dashed curves in Fig. 16). Figure 16b shows a contraction instead of extension across the North Africa-South Iberian orogenic segment. In our reconstruction (Fig. 11) shortening is mainly accommodated between North Africa and Alboran and Kabylies and does not seem to affect South Iberia. Although no clear evidence exists for this event in the field, the Mesorif unconformity (Gimeno-Vives et al., 2020) may in part be due to this transient shortening event.
Since 84 Ma, the total convergence between Africa and Europe is about 300 km of the same order as estimated by Macchiavelli et al. (2017). In details, the convergence of Africa, South Iberia and Ebro relative to Europe appears to be punctuated by periods of contraction at 80–70 Ma, ∼ 50 Ma and after 20 Ma, and stages of tectonic quiescence. Note that instead of quiescence, short intervals of very minor extension are also apparent at ∼ 80 and 50 Ma across Africa-South Iberia boundary (Fig. 16b). They reflect spurious wandering of motion paths. From 84 Ma to 70 Ma, both Ebro and South Iberia converge toward Europe (Fig. 16a) but shortening of 50 km is mostly accommodated at the Ebro-Europe boundary (Fig. 16b) and no convergence is recorded across the South Iberia-Ebro boundary. From 66 to ∼ 35 Ma, the amount of shortening between South Iberia and Ebro remains small because of left-lateral motion at the plate boundary (Figs. 10 and 11) (Anadón et al., 1989). This is coherent with a right-lateral convergent deformation regime in the Iberian Range during the early Cenozoic (Fig. 11) (Parés et al., 1988). From 60–50 Ma and until 20 Ma, the main part of the shortening is accommodated across the Betic-Rif (∼ 50 km across Africa-South Iberia boundary) and the Pyrenees (∼ 80 km across the Ebro-Europe boundary). After 20 Ma, both Ebro and South Iberia kinematically belong to Europe. The remaining N-S convergence (∼ 90–100 km) between Africa and Europe is therefore accommodated between Africa and South Iberia in the Betic-Rif region.
Discussion
Variscan inheritance and late Variscan position of Ebro-Sardinia-Corsica
The fragmentation of Western Europe into several microcontinents partly reflects Pangea breakup in the late Permian as the Neotethys Ocean opened and progressed westward (Ziegler, 1990b; Angrand et al., 2020). Major shear zones are accounted for in our kinematic restoration. They are major fault zones reactived during the Mesozoic rifting inherited for a large part from late Variscan dextral shear zones emplaced during the late Paleozoic (Fig. 17) (Edel et al., 2015; Ballèvre et al., 2018; Molli et al., 2020; Simonetti et al., 2020).
Paleomagnetic constraints (Edel et al., 2015) further suggest that during the late Carboniferous-early Permian (300–290 Ma), Catalonia (Ebro block including the Catalan Coastal Ranges) together with Sardinia-Corsica and Maures-Esterel (Sardinia-Corsica and Provence in our model), were disconnected from Iberia (South Iberia in our model). At this time, the westward migration and clockwise rotation of the Ebro-Sardinia-Corsica block relative to Europe is suggested to have been accommodated along the East Variscan Shear Zone. This is a major tectonic boundary that currently lies in the External Massif disconnecting the Massif Central from allochtonous terranes of Sardinia-Corsica (Ballèvre et al., 2018). In the Western Alps, the anisotropic lithosphere structure inherited from the East Variscan Shear Zone may also have triggered Early Jurassic rifting in the Alpine Tethys and later oceanic spreading south of the Penninic fault.
To the south, this displacement is suggested to be synchronous with the deformation of the Iberian orocline, in West and South Iberia block, along the North Pyrenean Shear Zone (Ballèvre et al., 2018) during the late Carboniferous-early Permian (Edel et al., 2018; Pastor-Galán et al., 2020). These results suggest that both Ebro and Sardinia-Corsica were tectonically distinct from Iberia. The Ebro block should therefore not be restored together with Iberia during the Permian, as commonly assumed in kinematic models that keep Ebro attached to West Iberia (Galicia) (Olivet, 1996; Vissers and Meijer, 2012; Le Breton et al., 2021). This further emphasizes the key role played by the shear zone positioned south of Ebro in the late Carboniferous, here named the Iberian Shear Zone (Fig. 17), that has been reactivated after the late Permian-Triassic and is currently shaping the Iberian Range. The late Variscan segmentation described above can further be extended between Iberia and North Africa and between Iberia and North America (Fig. 17) (Fernández, 2019).
Early Cretaceous reactivation of Tethyan rifted margins: en-échelon opening and contraction in the Alpine domain
A major kinematic change occurred between Europe and Africa during the Late Jurassic-Early Cretaceous (145–125 Ma) as a result of the northward propagation of southern North Atlantic west of Iberia and the Mediterranean region (Fig. 18). This resulted in the eastward drift of Iberia, Ebro-Sardinia-Corsica, and AlKaPeCa that in turn impacted the evolution of the Alpine Tethys, including Adria and European margins (Figs. 10 and 11). The east-directed movement of these blocks had two main consequences for the Alpine Tethys. First, our model predicts that the southern segments of the Alpine Tethys began to close due to comparatively much slower eastward displacement of Adria. We found indeed tectonic arguments in support of an Early Cretaceous closure based on shear zone geochronology and geochemistry from the Kabylides (Cheilletz et al., 1999; Fernandez et al., 2020), and this is indirectly supported on the Adria margin. Northward along the segment of Sardinia-Corsica and Peloritani-Calabria (Fig. 10a) there is no direct evidence for such a contractional event, e.g., Calabria preserves only a few meter of early Cretaceous sediments (Bouillin et al., 1988). As Adria moved in the NE direction, the northwest margin of Adria became highly oblique, thus accommodating transcurrent deformation and preventing large subduction of the Alpine Tethys in this region. The northern margin of Adria however recorded subduction of the Alpine Tethys responsible for Eo-Alpine deformation (e.g., Handy et al., 2010).
Looking closer to the movement of the Ebro-Sardinia-Corsica block relative to Europe, we infer that from Late Jurassic to Early Cretaceous, the progressive rifting from west to east opened en-échelon basins, represented by the Parentis basin, the Pyrenees-Provence basin and further eastward the France Southeast basin. In this context, the relationship between the Pyrenees-Provence rift and the France Southeast basin is of particular interest. The Dauphinois basin that formed during the Early Jurassic stage of Alpine Tethys rifting on the European margin was in part reactivated in the Late Jurassic-Early Cretaceous, hence forming the Vocontian basin. This event was synchronous with the rifting in the Bay of Biscay. All these basins are disconnected en-échelon basins aligned in the EW-direction along the plate boundary between Ebro-Sardinia-Corsica and Europe. They are separated by structural highs, including the Landes High in the western termination of the Pyrenean rift and Durancian uplift between the Provence and the Vocontian basins (Fig. 18). The Vocontian and the Dauphinois basins are suggested to emplaced on the proximal European margin (e.g., Decarlis et al., 2017), now inverted in the Ultra-Helvetic nappes (Ramsay, 1981; Butler, 1985). In this configuration, the Vocontian basin therefore represents the western termination of the Valais Ocean. In our view, the Valais Ocean although tectonically related to the Bay of Biscay is not directly connected to the Pyrenean rift, as commonly suggested (Schmid et al., 2008; Handy et al., 2010; Advokaat et al., 2014; van Hinsbergen et al., 2020). Another important implication is that the Briançonnais forms the eastern promontory (necking domain) of the Provence domain rather that the northeastern promontory of Ebro-Sardinia-Corsica as also commonly suggested (Trümpy, 1988; Stampfli, 1993; Stampfli and Hochard, 2009; Handy et al., 2010; Advokaat et al., 2014; van Hinsbergen et al., 2020). This would be consistent with the slightly deeper sedimentary facies documented in the Briançonnais than in Provence (Laubscher, 1971; Ricou, 1980; Trümpy, 1988). When the Bay of Biscay opened in the late Aptian-Albian (125 and 94 Ma stages; Figs. 18b and 18c), orthogonal extension reactivated the late Variscan North Pyrenean Shear Zone at the origin of the Pyrenean rift, crustal break-up and mantle exhumation. The Albian HT metamorphism affected the Pyrenean rift and the domains close to the spreading centre (Basque-Cantabrian basin, Cameros basin). During this second stage N-S or NE-SW extension was accompanied by the reactivation of NE-SW directed late Variscan inherited faults (e.g., Nîmes Fault, Toulouse Fault). Note that this is a slightly different model than shown in Tavani et al. (2018).
Although our model is a viable solution that explains several different types of constraints, it is still to be refined north of Africa and west of Adria (Maghrebian basin, east Ligurian Tethys and Ionian basin). A careful synthesis of basin architecture and geochronological constraints on Mesozoic mafic rocks in these regions may allow testing implications of our model.
Alpine Tethys and Cenozoic evolution of the West Tethys-Mediterranean region
Because Mesozoic extension was highly oblique to the plate boundary between Africa and Europe, the plate-scale left-lateral shearing induced the opening of narrow basins and distributed extension between the Iberia, Ebro and Adria microplates along pre-existing Permian-Triassic structures (e.g., Angrand et al., 2020). The opening of Alpine Tethys during the Lower Jurassic between the Atlantic and the Neotethys oceans in this oblique kinematic setting did not produce a large oceanic basin. The Alpine Tethys is understood to have been noticeably segmented and floored partially by oceanic crust. Two segments are identified. The western rift segment (Betic ocean/West Ligurian ocean) was positioned to the east of the Iberia paleomargin and arguably extended northward in the eastern Pyrenees during the Late Jurassic-Early Cretaceous reactivation. The ultra-mafic bodies dated at 180 Ma in the eastern Betics originated from a very attenuated crust and/or slow spreading ridge. The eastern branch (Alpine Tethys/East Ligurian basin/eastern Maghrebian basin) was the largest oceanized domain between Adria/Africa (Sicilia) and AlKaPeCa terranes. This is in agreement with an established oceanic subduction which existence is corroborated by fore-arc peridodites associated with the emplacement of the Kabylides (Fernandez et al., 2020). In this framework, the western Maghrebian basin was underlain by a very thin crust, laterally evolving eastward into the eastern Maghrebian basin.
At the beginning of the Africa/Europe convergence (84 Ma), remnants of oceanic crusts of Alpine Tethys and of the Ionian basins origin are positioned in the Betics and west of Adria. In the course of convergence, mountain belts formed by inversion of Mesozoic rifts such as the Pyrenees-Provence/Basque Cantabrians, Betics, Iberian Range, Catalan Coastal Range. The formation of the Mesomediterranean orogen resulted from the inversion of the Betic rift in the west and closure of the eastern Maghrebian basin between Kabylides and Africa/Sicilia in the east (Fig. 13). We infer that most oceanic domains were closed by the late Eocene, with the noticeable exception of the Ionian Sea. The North Africa/Adria slab retreated toward the Ionian basin triggering the opening of the Western Mediterranean Sea. The formation of the Alboran sea is suggested to have occurred by a different mechanism related to the westward migration of a delaminated mantle (e.g., Daudet et al., 2020). This is suggested by the opposite direction of subduction and slab retreat between the Alboran and Algerian basins (westward and southeastward, respectively) imaged by seismic tomography (Spakman and Wortel, 2004; Amaru, 2007).
The different evolution of the Western Mediterranean Sea with respect to the Eastern Mediterranean is likley related to strong segmentation of the southwestern Alpine Tethys, which was variably floored by hyperextended continental crust and oceanic crust between Iberia and Africa. The absence of a well-defined slab beneath both Greater and Lesser Kabylia supports our reconstruction of the southern Alpine Tethys. Indeed, only Lesser (eastern) Kabylia preserves evidence of subducted material, as attested by a small positive seismic velocity anomaly south of Sardinia (see a compilation of seismic tomography depth-slices from Amaru (2007) in Supplementary Material 1). As the Kabylides moved southward relative to the Alpine Tethys subduction (Figs. 7, 8, 11 and 12), the subducted material would now be preserved south of present-day Sardinia. This favors our “hybride-type” model of the southern Alpine Tethys that includes variably extended crust and oceanic crust in both the Betic rift and Maghrebian Tethys, but excludes a continuous AlKaPeCa domain prior to the Cenozoic.
Finally, Oligocene calc-alkaline magmatism in Sardinia (Bellon et al., 1977) may not be simply related to the subduction of the Alpine Tethys below Sardinia. Indeed, our reconstructions do not account for significant subduction of the Alpine Tethys at this time below Sardinia (Figs. 7, 8, 11 and 12). Other mechanisms such as extension or mantle delamination associated with re-melting of continental lithosphere marked by previous subduction may produce similar calc-alkaline magmas (e.g., Richards, 2009).
Conclusions
In this paper we aimed to produce a kinematic reconstruction of the West Mediterranean since the late Paleozoic between Africa and Europe, accounting for the fragmentation of Iberia and its implication for strain distribution during plate convergence. Our main conclusions are:
(1) Iberia appears segmented into three main micro-continents, the West Iberia, the South Iberia and the Ebro-Sardinia-Corsica blocks inherited from late Variscan evolution. The Iberia-Europe plate boundary is therefore represented by two main NW-SE (present-day coordinates) strike-slip systems, in the Pyrenean and Iberian Range basins. These main tectonic features allow partitioning the left-lateral strike-slip kinematics between stable Iberia and Europe along two tectonic corridors, the Pyrenean and the Iberian rifts. These tectonic boundaries have been active throughout the Mesozoic. They accommodated transtensional deformation relative to Europe and therefore are critical to explain how Iberia translated to the east in the Alpine Tethys and during the opening of the southern North Atlantic and Bay of Biscay.
(2) We define a new plate configuration between Europe, Iberia, and Adria. It includes a newly defined Ebro-Sardinia-Corsica block and places the Briançonnais unit in continuity of the Provence promontory. The Vocontian basin is therefore suggested to connect with the Valaisan basin. These basins are both part of a larger en-échelon Early Cretaceous rift system between the south Provence and the Pyrenean regions.
(3) The opening of the Alpine Tethys resulted in the development of hyperextended rift domains and narrow oceanic basins forming discontinuous distal terranes between South Iberia, Ebro-Sardinia-Corsica, AlKaPeCa and Africa. The closure of these rift basins formed an elongated orogen during the Paleogene, dismantled during the opening of the West Mediterranean Sea.
This study shows that in such complex intra-continental geodynamic settings, plate boundaries should not be considered as a continuous and/or unique structural feature. Instead, we have distinguished a number of small stable crustal domains between lithosphere-scale shear zones that shape the Africa-Europe plate boundary. Most of these shear zones show complex architecture characterized by orientation inherited from the late Variscan orogeny. Despite the obvious relationships between tectonics and Mesozoic plate kinematics, firm constraints connecting the late Variscan and Alpine evolution are however lacking. More geochronological data and field-based studies supplemented by novel tectonic constraints from Permian and Triassic rocks, or from Variscan basement with a Permian record are needed to fill this gap.
Data availability
This study is based on data compilation. Data used in this study can be found in the appropriate references. Kinematic model files are available online in the Supplementary Material.
Author contribution
This article was written by Paul Angrand and Frédéric Mouthereau. Paul Angrand carried out the compilation of data, interpretation, and kinematic model and figures, in tight collaboration with Frédéric Mouthereau.
Acknowledgments
This research has been supported by the Centre National de la Recherche Scientifique (CNRS-INSU), the French Geological Survey (BRGM), and Total through the OROGEN project, as a post-doctoral grant to Paul Angrand. We acknowledge the members of the OROGEN scientific project for support and discussions during the postdoc of Paul Angrand, in particular leaders and PIs Emmanuel Masini, Sylvain Calassou, Olivier Vidal, Isabelle Thinon, Laure Moen-Maurel, Mary Ford, Laurent Jolivet, Gianreto Manatschal, as well as former postdocs and phds Gianluca Frasca, Adrien Romagny, Riccardo Asti, and Nicolas Saspiturry. Emmanuel Masini is particularly thanked for continuous discussions during the postdoc of Paul Angrand, which greatly contributed to the development and maturation of this project. Oscar Fernandez is acknowledged for discussions about the evolution of Iberia in the frame of the Atlantic opening.
We thank Editor Olivier Lacombe for his patience during the submission of this manuscript. We thank reviewers Antonio Schettino, Reinoud Wissers, and Charlotte Nielsen for their constructive and helpful comments about our (large) manuscript. Associate Editor David Pedreira is thanked for pointing out mistakes about the Basque-Cantabrian Basin in an earlier version of the manuscript.
Cite this article as: Angrand P, Mouthereau F. 2021. Evolution of the Alpine orogenic belts in the Western Mediterranean region as resolved by the kinematics of the Europe-Africa diffuse plate boundary, BSGF - Earth Sciences Bulletin 192: 42.