Discussion on the MHB evolution

Controls on sedimentation

Because sedimentation in the MHB is mainly marine and because it is contemporaneous with the Tertiary structural development of the Hellenides, tectonic deformation and eustatic sea-level variations have to be considered both as major forcing parameters on sedimentation.

Tectonic control on sedimentation

The existence of syntectonic deposits within this large piggyback basin has been described in previous sections. The tectonic control have an input in the basin sedimentation at various scales, including:

- (i) Flysch and olistolites deposition in small active sub-basins (e.g. late Eocene series in Krania area) at the onset of MHB development;

- (ii) Conglomeratic deposits in association with localized active faults (e.g. synsedimentary listric faults in Mitropoli area);

- (iii) Conglomeratic beds that developed on top of some significant angular unconformities in the basin, such as at the base of Krania, Rizoma, and Eptachorion Formations;

- (iv) The syntectonic fannig system of the Lower Meteora Conglomerates, representative of a major uplift of the source area (Pelagonian basement);

- (v) The eastward migration of the subsidence areas during the early Miocene (Pentalofon to Tsotyli Formations), associated with the tectonic uplift of a structural high (TTS) in the southern MHB.

Eustatism control on sedimentation

The MHB is filled by mainly marine sediments that were deposited at relatively shallow water- depth. Therefore the eustatic sea-level variations have to be considered as a significant driving parameter on sedimentation. This has been outlined by most sedimentological studies [i.e. Ori and Roveri, 1987; Zelilidis et al., 2002].

The high frequency rythmic sequences could be linked to some eustatic Milankovich cycles, notably within the Meteora conglomerates. However, it remains very uncertain to relate the lower frequency cycles to large eustatic events on the basis of correlations with the published eustatic charts [e.g. Haq et al. 1987; Abreu et al., 1998] because of the lack of precise datations in some formations such as in Meteora Conglomerates. For instance, the transition from Eptachorion marls to Pentalofon conglomerates (Lower Meteora Conglomerates) is clearly at least controlled by some tectonic activity [Ferrière et al., 2011] but it could be also enhanced by the large drop of sea-level proposed for the Late Oligocene [Haq et al., 1987]. However the amplitude of this sea-level drop is uncertain and some authors proposed a much smaller drop at that time [Abreu et al., 1998]. Moreover the age determination of this boundary is still questionable, being within the nannofossils biozone NP25 or Aquitanian, according to authors (Fig. 17).

Figure 17. Eustasy vs tectonic controls on MHB evolution

Eustasy vs tectonic controls on MHB evolution

Tectonic control. On the right, tectonic evolution after Vamvaka et al. (2006) assuming an important strike-slip event during Oligocene-Early Miocene. On the left, tectonic evolution after Ferriere et al. (1998, 2004 and 2011) considering a control by the underthrusted units under the MHB: Subduction of the Pindos basin then of the Gavrovo-Tripolitsa thick crust responsible for compressive structures followed by the migration of the main subsidence depocenters and uplifts with development of normal faults and some compressive tectonic structures.

Eustasy. Eustatic curves (from Haq et al. 1987).The eustatic control is difficult to proove as the ages of the lithologic Formations are not accurate enough to be compared with the eustatic charts (see also Fig.12). For us, the eustatic control on the MHB shallow marine sediments is necessarily efficient but with less control than the tectonic one. See text for more discussion.

Discussion on geodynamic interpretations

The first published interpretations on the geodynamic evolution of the MHB are relatively recent (Fig. 18). Firstly, the MHB has been considered as a retro-arc basin [Doutsos et al., 1994] related to the existence of eastward verging structures in the basin (Fig. 18A). However, these tectonic movements are mainly restricted to the late Eocene whereas the main elongated MHB is essentially Oligocene to Early Miocene in age. Moreover, the eastward verging reverse faults are not very developed and they cannot account for the observed eastward shift of the depocenter.

Figure 18. Different geodynamic interpretations

Different geodynamic interpretations

Different geodynamic interpretations. A: Doutsos et al., 1994: retro-arc basin in front of thrusts (Krania, Eptachori thrusts) verging to the East, linked to the main thrust of internal zones on external ones verging to the West (thrusts modified at their base, taking into account the Doutsos et al., interpretative figures). B: Zelilidis et al., 2002: the MHB is supposed to be a graben bounded by normal faults with a moderate strike-slip movement. C: Ferriere et al., 2004: the evolution of the MHB is linked to its piggyback character with different successive tectonic structures. D: Vamvaka et al., 2006: the MHB is mainly a pull-apart basin bounded by major strike-slip faults developed during Oligocene-Early Miocene times.

Alternative models proposed a significant strike-slip component in the development of the MHB [Zelilidis et al., 2002; Vamvaka et al., 2006]. On the basis of the presence of numerous normal faults, few angular unconformities in Oligocene and Miocene series and of interpretation of flower structures from some of the seismic lines, Zelilidis and co-authors [2002] considered the MHB as a strike-slip hemi-graben (Fig. 18B). For other authors [Vamvaka et al., 2006], the MHB is controlled by the subducting slab as described by Ferrière et al. [2004] but also behave as a pull-apart basin during the Oligocene to Early Miocene times (Fig. 18D).

We proposed earlier [Ferrière et al. 1998 ; 2004] an evolution controlled directly by the subduction of the external zones below the internal zones (Fig. 18C): firstly a forearc type of basin (Eocene), then a large crustal-scale piggy-back basin (Fig. 18C). This interpretation is based on: (i) the modifications of the basin geometry and depocenter locations that are driven by some paroxysmal tectonic episodes (e.g. near the Eocene-Oligocene boundary), and (ii) the position of the MHB domain on top of the Pelagonian basement during the downgoing motion of external zones below these internal zones. The displacement of the external zones below the internal zones is evidenced by the existence of the large tectonic windows (Ossa and Olympus windows) exposing the external zones in the inner Pelagonian zone [Godfriaux, 1968].

During the Lutetian to Late Eocene times, the Krania and Rizoma sub-basins are contemporaneous with the westward subduction of the thin (oceanic?) crust of the Pindos Basin. Possibly because of thin crust subduction, it is the upper unit and its irregularities (i.e. Pelagonian Indentor) that control the geometry of Eocene depocenters. Later on, since the basal Oligocene, the subducting crust is thicker and lighter: the continental crust bearing the Gavrovo-Tripolitza Platform. The arrival in subduction of this crust is interpreted as the driving mechanism of the brutal change in basin geometry, of major subsidence in the MHB axis and migration of the depocenter (cf. next section).

Discussion on mechanism of basin development

The driving mechanisms we consider for the tectonic development of the MHB are related to the subduction processes of the Hellenides external zones below the Pelagonian basement (Fig. 19).

The whole MHB is developing on the eastern side of the Pindos Fold and Thrust Belt [e.g. Skourlis and Doutsos, 2003] that can be regarded as the equivalent to an accretionary prism, notably in the early stages of its development. Comparatively with other accretionary complex, the MHB domain has therefore to be regarded as a forearc domain, possibly comparable with for instance the Hikurangi subduction margin in New Zealand, also characterized by the subduction of relatively thick crust [e.g. Nicol et al., 2007]. Such forearc domains are classically located between the highest ridge of the accretionary prism and the volcanic arc. In the present setting of the MHB, the Eocene volcanic arc is not clearly identified. Some late Eocene and Early Oligocene calco-alcaline rocks are exposed in Turkey [Dilek et al., 2009] and in northeastern Greece, but they could be linked to another subducted oceanic area [Intra-Pontide Ocean, Pe-Piper and Piper, 2006; 2007].

Figure 19. Main tectonic mechanisms supposed to have controlled the piggyback MHB evolution.

Main tectonic mechanisms supposed to have controlled the piggyback MHB evolution.

A: Subsidence linked to subduction: the early subsidence areas are linked to the subduction of the thin crust of the Pindos basin. The Krania and Rizoma small basins are located behind the progressively uplifted Pindos accretionary prism, in the fore-arc domain. B: Compressional tectonics: the main compressive tectonic structures (especially in the late Eocene) are linked to the arrival of a thick crust under the basins (transition between Subduction and Collision events). C: Subsidence during the piggyback stage: despite the underthrusting of a thick crust (i.e. Gavrovo crust), subsidence is active and gives rise to the main piggyback basin. The subsidence areas could be linked to basal tectonic erosion: a tectonic slice of Pelagonian basement is pushed to the East along the basal thrust, so that the Pelagonian basement is thinned under the basin. D: Uplift during the piggyback stage: the uplifts (e.g.:uplift of the MHB eastern area) are linked to the stacking of the tectonic sheets (Pelagonian slices and possibly Gavrovo ones) linked to tectonic events along the basal thrust

Accordingly, the initial subsidence of the MHB domain can be attributed to forearc setting, on top of pre-existing dense obducted ophiolites, during the subduction of Pindos thick oceanic basin (Fig. 19A).

The location and geometry of the main MHB basin (Oligocene-Early Miocene) is controlled by its piggy-back behavior. Low angle subduction of the lighter and thicker crust of the Gavrovo- Tripolitza Zone led to the uplift of the Pindos accretionary wedge. This major uplift corresponds to the transition from oceanic subduction to collision (or continental subduction) of the external zones (Fig. 19B). This process is coeval with the development of an elongated sudsiding area in the inner domain, on the back of the highest ridge of the uplifting accretionary wedge: the Oligocene to early Miocene MHB. Because of the eastward motion of the underlying units (Gavrovo-Tripolitza Zone) coeval with sedimentation, this area of subsidence corresponds at that time to a large piggy-back basin (Fig. 19C).

In addition with this process of individualization of a large elongated basin parallel to the Pindos wedge, the subsidence can be driven by basal tectonic erosion (Fig. 19C). Such tectonic erosion is not only an important process for creating subsidence areas in convergent settings, but it is also adapted to account for dominant normal faulting during basin sedimentation (e.g. Chanier et al., 1991; 1999).

The crustal Pelagonian slivers pulled out by tectonic erosion can accumulate farther east below the Pelagonian crust and therefore account for the uplift of the internal zones coeval with the MHB development (Fig.19D). A succession of superimposed crustal duplexes, from Pelagonian material, and then of Gravrovo-Tripolitza material, could be responsible for the initial uplift of the inner Pelagonian zone, leading to the first steps of the progressive exhumation of Olympus and Ossa tectonic windows. Even if the last stages of recent exhumation of these windows could be linked to the Plio-Quaternary general extension within the Hellenic region, this mechanism easily explains the observed structural geometry inside the Olympus window.

As exposed on previous sections, some horizontal striations occur, notably on some of the main faults. We do not believe that the total strike-slip displacement along the main faults was of major importance because no large lateral offsets could be evidenced. Such strike-slip displacements during convergence could be associated with some strain partitioning such as described on most oblique convergent settings [e.g. Cashman et al., 1992; Nicol et al., 2007].