The Moroccan Hercynides
le géoparc du jbel bani - tata

The Moroccan Hercynides

By Christian Hoepffner, Abderahmane Soulaimani, Alain Pique

Department of Geology, Faculty of Sciences, University Mohammed V, B.P. 1014 Rabat, Morocco

Department of Geology, Faculty of Sciences, University Cadi Ayad, B.P. S 20, Marrakech, Morocco

21 Rd Point de l’Esplanade, 67 000, Strasbourg, France

Received 5 April 2004; accepted 18 July 2005

Available online 18 November 2005

Abstract

In Northern Africa deformed Paleozic rocks are observed, mainly in Morocco where they are present in two structural domains: (i) the southern domain, i.e. the Anti-Atlas belt. In this northern rim of the West African craton, the Paleozoic sequences are mildly deformed; (ii) the northern Mesetan domain, itself subdivided into several structural subzones, corresponds—with its eastern prolongation in Algeria- to the Hercynian/Variscan belt of Northern Africa.

In the Meseta, the study of the sedimentological history and the structural analysis of the various subzones allow to discuss the tectonic characters of the Hercynian/Variscan orogeny in Northern Africa:

– The deformation was realized through three main events: (i) the Late Devonian, Eovariscan Phase, which developed in the eastern zones of the Moroccan Meseta, northwestern Algeria and in the presently Alpine allochthonous domains of the Rif and Kabylias;

(ii) The Visean Phase, which was restricted at the limit between the Moroccan eastern and western Meseta. The two preceding compressional events were contemporaneous with the development of Late Devonian-Early Carboniferous transtensive sedimentary basins; (iii) the Late Carboniferous Phase, during which the regional shortening affected the entire Hercynian domain, including the Moroccan western Meseta and Anti-Atlas.

– The Anti-Atlas differs fundamentally from the Mesetan domains by: (i) its relatively mild deformation; (ii) the vergence of its regional structures, directed toward the West African shield, whereas the main vergence is toward the west in the Meseta domains.

– In the Anti-Atlas like in the weakly deformed zones of the Meseta, the deformation is heterogeneous, being concentrated along several regional shear zones. Sedimentological and stratigraphic evidences point out the control exerted by these long-lived weakness zones through the sedimentation, and the deformation.

– No direct evidence can be found to invoke the development of a Paleozoic oceanic lithosphere in the considered domains. Therefore, the Hercynides/Variscides of North Africa are considered to be outside the inner zones of the European belt, at the northern border of the Paleo-Gondwana, and probably related to the Appalachian belt of North America.

 2005 Published by Elsevier Ltd.

1. Introduction

More than one century ago, some workers penetrated, with strong difficulties, the Moroccan territory and they began to decipher its then ignored geology. Among them Thomson crossed the High Atlas Mountains in 1888. He described eruptive rocks (the Atlasic Palaeozoic massif) as an island inside the Cretaceous sea, and the red lacustrine

Nfis sandstones, which are folded (Thomson, 1889).

From this onwards, the reality of the Hercynian and the subsequent Atlasic deformations was ascertained.

During the beginning of the 20th century, Gentil performed an extensive field work north of the High Atlas Mountains. ‘‘The structure of the plateaus is characterized by a folded Palaeozoic basement, unconformably covered by transgressive Mesozoic deposits, which remained mostly horizontal. Due to the erosion, we see that a wide chain, developed by the end of Carboniferous times (i.e. the Hercynian chain) has been completely eroded. This structure of western Morocco has to be compared to the Iberian Meseta’’ (Gentil, 1918; translated). Following Gentil, Lecointre (1926) carefully mapped northwestern Morocco while other geologists, most of them belonging to the newly constituted Geological Survey of Morocco, spread the knowledge of the country in Northern Morocco and south of the High Atlas as well. A complete list of the references is given by Morin (1964, 1970, and 1979).

As shown by the ancient authors, the Hercynian (=Variscan) chain of Northern Africa, extending from western Mauritania to Northern Algeria, is best exposed in Morocco where the extended outcrops of Palaeozoic rocks allow to distinguish several structural domains (Fig. 1).

– The Saharian domain: in southern Morocco the Anti-Atlas domain presents Precambrian inliers cartographically surrounded by more or less deformed Palaeozoic rocks. Hercynian folds are known also in the Algerian Ougarta and in western Mauritania. In the extreme south, the Tindouf basin is the undeformed Palaeozoic cover of the West African craton (Reguibat Rise).

– Middle Morocco, where the existence of the Hercynian deformation has been first recognised (Gentil, Lecointre, etc.), is subdivided into two zones: (i) the Meseta, where several massifs made of deformed Palaeozoic rocks are inliers, cartographically separated by a thin and tabular Mesozoic and Cenozoic cover; (ii) the High and Middle Atlas where the Palaeozoic massifs, similar to those of the Meseta, are separated by a thick and folded sedimentary cover. The Atlasic shortening remained weak and correlations between the Palaeozoic massifs of the Meseta and of the Atlas are possible.

– The Rif, a part of the Alpine chain of western Mediterranean: the shortening was important and renders difficult, although not impossible, comparisons between the Rifean Palaeozoic series and the other Palaeozoic massifs of Morocco. The Algerian Alpine Tell belt bears

Palaeozoic blocks, for instance the Kabylias, which are very similar to those of the Moroccan Rif.

The existence of the Hercynian deformation throughout Morocco and adjacent Algeria lead to distinguish several structural zones (Michard, 1976; Pique´ and Michard, 1981, 1989; Michard et al., 1989; Pique´, 1984, 1989, 1994, 2001), which are described below. The definition of the structural zones will in turn allow to present the main characteristics of the Hercynian orogeny in Northern Africa and to tentatively replace the Hercynian belt of North

Africa within the tectonic frame of the Hercynian orogeny of Western Europe.

Fig. 1. Structural subdivisions of North Africa and location of the Paleozoic terranes. 1: Rifian-Tellian belt, a: southern front of the chain; b: Paleozoic allochtonous terranes of the internal zones. 2: Atlasic belt (HA: High Atlas, SA: Saharian Atlas, TA: Tunisian Atlas, MA: Middle Atlas). 3: Paleozoic, a:deformed by variscan orogeny, b: undeformed. 4: West African craton (Archean and Proterozoic massifs). 5: Mauritanides allochtones.

2. The palaeozoic evolution of the Hercynian structural domains

2.1. The Saharian domain              

2.1.1. The palaeozoic sedimentation (Table 1)

Between the northern limit of the West African craton and the Hercynian orogenic domains of central Morocco (see below), the Anti-Atlas (Choubert, 1947, 1963) is a large and open antiform, bent during the Atlasic orogeny.

Its structural grain trends WSW–ENE in its western part, from which it prolongates toward the Zemmour belt, E–W in its central part, and ENE–WSW in the east, where it connects to the Algerian Ougarta (Figs. 1 and 2). It is separated from the High Atlas by the recent Ouarzazate and Souss troughs. Its main characteristics are the wide outcrops of Precambrian rocks, visible in several inliers (‘‘boutonnie`res’’), which are covered by Palaeozoic rocks.

The Precambrian times in Anti-Atlas have been marked by the development of the Eburnian orogeny, at about 2000 Ma and the Pan-African orogeny, at about 700–600 Ma. The oldest granitoids and metamorphic episodes are related to the Eburnian orogeny (Charlot, 1978; Aı¨t

Malek et al. 1998; Thomas et al. 2002; Walsh et al, 2002). The most obvious traces of the Pan-African orogeny are seen in the central Anti-Atlas, where an ophiolitic sequence has been described (Leblanc, 1972, 1975; Saquaque et al., 1989). This ‘‘Pan-African suture’’ is traced southeasterly in Algeria, as far as the Hoggar (Black and Lie´geois, 1993) although its actual signification is currently discussed (Ennih and Lie´geois, 2001).

After the deposition of the latest Proterozoic volcanic series and redbeds (Ouarzazate series, 575–560 Ma; Thomas et al., 2002), at least partly related to crustal extension (Pique´ et al., 1999; Soulaimani et al., 2003), an eastward-directed transgression started at the very beginning of the Cambrian. This event can be observed all over the Anti-Atlas domain. The first sediments were detrital, but open marine conditions rapidly developed, inducing during the Lower Cambrian, at least in the western and central Anti-Atlas, and the development of a shallow platform, more subsident in the western Anti-Atlas (Choubert, 1963; Benziane et al., 1983). During this period, the so-called Lower and Upper Limestones were deposited, separated by purple shales (Lie-de-vin series) representing a brief regressive episode. After a regression (and an emersion?) at the end of the Cambrian, the sea invaded again the domain for a long time at the beginning of the Ordovician.

The region was still a shallow and epicontinental platform.

The Ordovician sediments were typically detrital, being composed of sands and clays, all issued from the Saharian continent (Destombes et al., 1985). At the end of the Ordovician a regional glaciation developed (Ouanaimi, 1998), the Anti-Atlas being located not far from the ice cap that spread over western Africa (Deynoux, 1978). The end of the glaciation occurred at the beginning of the Silurian and it was the cause of a glaciogenic transgression over Morocco. Black shales were deposited in confined marine environments and poorly oxygenated waters. At that time the African shield, eroded since a long time and covered by the marine transgression, did not provide more than very fine argillaceous minerals. Furthermore, during the Lower Devonian, even those fine detrital sediments lack in the Anti-Atlas marine basin, and limestones, often reefal, were deposited onto the carbonate platform.

During the extensional regime that had characterised the first part of the Palaeozoic, the clastics were issued from the West African craton. From the Devonian, the extensional regime changed to a compressive one, expressing the development of the Hercynian orogeny. The Proterozoic axis of the Anti-Atlas rose up and became the source of the clastics (Wendt, 1985; Hassenforder, 1987). As a consequence, the Late Devonian and Carboniferous sequences deposited in the Anti-Atlas are typically detrital: sandstones, clastic limestones and argillites.

Table 1

Simplified paleozoı¨c evolution of the main structural zones of the Northern (Mesetian) domain

1: panafrican basement (Anti-Atlas); 2: sandstones and quartzites; 3: shales and greywackes; 4: shales and psammites; 5: basin deposits (deltaic or deep sea fan sequences); 6: continental redbeds (conglomerates, sandsones); 7: limestones; 8: black shales; 9: volcanoclatic deposits with calk-alkaline volcanism; 10: alkaline to tholeitic volcanism; 11: chaotic deposits (olistostromes, debris flows, rock falls); 12: granites; 13: strong tectonic event: folding and metamorphism; 14: moderate tectonic event: faults, folding; 15: main uncomformities.

1998), the Anti-Atlas being located not far from the ice cap that spread over western Africa (Deynoux, 1978). The end of the glaciation occurred at the beginning of the Silurian and it was the cause of a glaciogenic transgression over Morocco. Black shales were deposited in confined marine environments and poorly oxygenated waters. At that time the African shield, eroded since a long time and covered by the marine transgression, did not provide more than very fine argillaceous minerals. Furthermore, during the Lower Devonian, even those fine detrital sediments lack in the Anti-Atlas marine basin, and limestones, often reefal, were deposited onto the carbonate platform.

During the extensional regime that had characterised the first part of the Palaeozoic, the clastics were issued from the West African craton. From the Devonian, the extensional regime changed to a compressive one, expressing the development of the Hercynian orogeny. The Proterozoic axis of the Anti-Atlas rose up and became the source of the clastics (Wendt, 1985; Hassenforder, 1987). As a consequence, the Late Devonian and Carboniferous sequences deposited in the Anti-Atlas are typically detrital: sandstones, clastic limestones and argillites.

2.1.2. The Hercynian deformation

The Hercynian deformation was weak in the Anti-Atlas, when compared to its effects in central and Northern Morocco, where emplacement of granites, regional metamorphism and penetrative deformation occurred. However detailed examination leads to distinguish several structural zones:

2.1.2.1. The western Anti-Atlas (Figs. 2 and 3).

West of the Bas Draa inlier, the Cambrian rocks are involved in NNE–SSW folds and thrusts, all of them being east-vergent (Choubert, 1963; Mattauer et al., 1972; Soulaimani, 1998; Belfoul et al., 2001). In the westernmost part of the area, near the Atlantic coast, the folds are tight and overturned. At the scale of the outcrop, their axial plane is underlined by a metamorphic foliation with an E–W stretching lineation, compatible with the sense of transport (Belfoul et al., 2001). The intensity of the deformation decreases rapidly eastward and the Cambrian strata are only affected by large tens of meters-scale folds with a westward steeply dipping fracture cleavage, mainly visible in pelitic levels. Further east, in the Goulmine Plain, 15 km from the Atlantic shoreline, the Middle Cambrian strata are only slightly tilted.

In the Bas Draa inlier, the crystalline basement remained practically undeformed in the course of the Variscan orogeny.

By contrast, the base of the sedimentary cover was deformed due to the uplift of the rigid basement along the southeastern limit of the inlier (Soulaimani et al., 1997).

In the Bani area (Fig. 3), disharmonic folds affecting the Cambrian limestones are concentrated in some detachment levels at the base of the cover (Helg et al., 2004). Above, the Cambrian-Ordovician sandstones are affected by kilometre- scale open and box-folds, slightly overturned southeastward and associated with an incipient cleavage in their hinges. The Jbel Rich tens of meters-scale wavelength folds developed in the Devonian sandstones is often asymmetric: their southern flanks are sub-vertical or in reverse position, their axes are subhorizontal and they display a regional en-echelon structure.

South of the Oued Draa, the Upper Devonian shales are the southernmost formation of the Anti-Atlas that is affected by metre-size folds; the Jbel Ouarkziz Carboniferous strata is only slightly tilted toward the south, at the northern border of the Tindouf bassin.

North of the Bas Draa domain, the Lakhsass Plateau area is located between the Ifni and Kerdous inliers. There, the Hercynian structures result from an E–W regional shortening. They are particularly developed along old reactivated fractures in the centre of the area, as on both limbs of the Jbel Inter anticline. The prominent fabric in the cover rocks is a pervasive cleavage, axial-plane to tight or isoclinal upright folds. On both sides of this zone, the deformation decreases laterally toward the inliers around which the unconformable Palaeozoic sedimentary cover is only gently tilted to the centre of the plateau.

Northeast of the Kerdous inlier, the Irherm-Tata area is a large-scale open folded zone where wide synclines cored by Middle Cambrian are separated by narrower uplifted inlier anticlines. The structural pattern of the Variscan folds seems erratic and it would suggest successive folding episodes. However, no refolded axis or cleavage plane is observed in the field. Moreover, there is a regional distribution of the various trends that are controlled by vertical dominant motion of reactive crustal faults like, for instance, the Tata fault (Hassenforder, 1987; Faik et al., 2001; Caritg et al., 2004).

Fig. 3. Geological sections across the north western part of the Anti Atlas domain (see location on Fig. 2).

2.1.2.2. Central and eastern Anti-Atlas. In the Bou Azzer inlier, the crystalline basement did not suffer any noticeable ductile deformation in the course of the Hercynian orogeny.

By contrast, its sedimentary cover was folded, locally affected by an incipient cleavage and, more importantly, it was detached by a decollement from its basement. The resulting Hercynian structure is a large scale NW–SE box anticline. Vertical reactivation of basement fractures is suggested by the concentration of the deformation along the inlier borders, which laterally decreases toward the adjacent Cambrian syncline (Leblanc, 1975).

South of the Tinghir oasis, the Palaeozoic rocks are affected by several thrust faults that dip gently northwards.

After Hindermeyer (1954) and Choubert (1959); Michard et al. (1982) distinguished several allochthonous units overlapping a southern autochthonous sequence apparently undetached from the Sahgro Precambrian basement.

Such southward overthrusting tectonic structures extend to the east, since several thrusts affect the Neoproterozoic cover along the northern side of the Ougnate inlier and in the Tafilalt subsurface.

Generally, in the Anti-Atlas domain, the Hercynian metamorphism is weak, from very low grade to low grade, in the range of 150–300 [1]C (Buggisch, 1988; Soulaimani, 1998). It may be explained by deep sedimentary burial beneath more than 10 km of Paleozoic overburden.

2.2. The northern domains (Fig. 4 and Table 1)

2.2.1. The Sehoul zone

Initially defined by Pique´ (1979, 1982), the Sehoul zone outcrops to the east of Rabat in a small area but it can be followed northward under the Gharb Cenozoic deposits.

Its contact with the Meseta is a faulted zone trending N80–120[1] E (The Rabat Tiflet Fault Zone, RTFZ). The Sehoul zone is made of Cambrian pelites and greywackes, with a minimum exposed thickness of several hundred metres. Sedimentological observations suggest a shallow but subsident deltaic environment (El Hassani, 1990).

The rocks have been subjected to a tectono-metamorphic episode that gave way to the development of E–W trending folds, recumbent and overturned to the south, accompanied by a cleavage contemporaneous with a low-grade regional metamorphism. Deformation and metamorphism increase towards the south, from anchizonal domain (illite crystallinity data, Pique´, 1979) with spaced cleavage to epizonal domain (chlorite and biotite) with slaty cleavage. The metamorphism has been dated at 453 ± 8 Ma (K/Ar on micas; El Hassani et al. 1991). The Tiflet granite emplaced within the slates at 430 ± 3 Ma (Rb–Sr method; Charlot et al., 1973). These data indicate an orogenic episode, coeval with the Caledonian orogeny. This is the only evidence for a Caledonian imprint in Morocco, except in the continental margin off El Jadida, where borehole data show that a Cambrian granodiorite has been mylonitized around 455 Ma (K–Ar method, Kreuzer et al., 1984 in Ruellan, 1985). This suggests that a ‘‘Caledonian’’ beltwraps around the northwest Meseta (El Attari, 2001; Fig. 4).

The folded and metamorphosed Cambrian slates and the Tiflet granite were thrust upon the Meseta prior to the deposition of Late Silurian strata that unconformably lie upon the granite. Elsewhere, the slates are unconformably covered by fluviatile conglomerates of Lower Visean age (Pique´, 1979; El Hassani, 1990).

In the Sehoul zone S.S., the Hercynian deformation was weak and limited to its southern margin, reactivated and also thrust onto the Meseta; it can be connected with the slight thermal event which has been dated at 320 Ma (K/ Ar on mica) in the Sehoul region and in the El Jadida margin.