These RobertCharrueC PDF
These RobertCharrueC PDF
Université de Neuchâtel
Suisse
Géologie structurale de
l’Anti-Atlas oriental,
Maroc
par
Charles Robert-Charrue
MARTIN BURKHARD
QUI NOUS A QUITTÉ PRÉMATURÉMENT EN AOÛT 2006.
RESUME
Mots-clés : Anti-Atlas, Maroc ; Tectonique d’inversion ; Orogène varisque ; Paléozoïque ;
Interférence de plis ; Profondeur de détachement ; Déformation interne ; Plis associés aux failles
extensives.
La chaîne de l’Anti-Atlas du Maroc est située sur la bordure nord du craton ouest
africain. Sa forme générale est un vaste anticlinal d’orientation ENE-WSW, formé
en son cœur de boutonnières de socle et sur ses flancs d’une couverture
plissée. Le socle Précambrien recèle les traces des orogènes Eburnéen (2 Ga)
et Pan Africain (600 Ma). Dès la fin du Néoprotérozoïque, un rift initie un
nouveau cycle de Wilson. A l’issue de la sédimentation syn-rift, le rift est avorté,
mais une subsidence régulière permet le dépôt d’une épaisse couverture
Paléozoïque dans un environnement de mer peu profonde. Ce bassin
intracontinental est inversé au Carbonifère moyen, créant ainsi la chaîne plissée
de l’Anti-Atlas. La particularité de cette chaîne est l’absence de décollement
majeur.
Les études géologiques menées dans l’Anti-Atlas oriental ont permis de décrire
sa structure et de caractériser son style structural. Les deux paramètres qui
contrôlent le type de la déformation sont (1) la pré-structuration du socle lors du
rift Néoprotérozoïque – Cambrien et (2) la rhéologie particulièrement
incompétente de la couverture. La compression varisque inverse les anciennes
failles normales du socle. A grande échelle, la couverture accommode la
déformation par drapage, mais elle est également affectée par un
raccourcissement interne qui se traduit par la formation de petites structures
compressives, visibles dans les quelques formations plus compétentes qui
jalonnent la série Paléozoïque. L’analyse des structures a permis de discerner
deux orientations distinctes de déformation. L’une correspond à l’orientation
générale de l’Anti-Atlas, l’autre adopte l’orientation de la chaîne de l’Ougarta
située au sud-est. La superposition de ces deux orientations provoque une figure
d’interférence en dômes et bassins.
Les relations avec les sédiments post-varisques montrent que l’inversion du Haut
Atlas ne provoque pas de déformation mais une légère surrection dans l’Anti-
Atlas oriental. Ce soulèvement combiné à celui induit par une anomalie
thermique récente explique la haute topographie de l’Anti-Atlas oriental.
ABSTRACT
Keywords : Anti-Atlas, Morocco ; Inversion tectonics ; Variscan orogeny ; Paleozoic ; Folds
interference ; Depth to detachment ; Bed-length shortening ; Extensional fault-related folding.
The Anti-Atlas belt of Morocco is situated on the northern edge of the West
African Craton. The Anti-Atlas appears as a huge NE-SW anticlinorium. Locally,
the basement is cropping out as inliers, but the main part is made of a gently
folded Paleozoic cover. The Eburnean (2 Ga) and Pan African (600 Ma)
orogenies left their imprints in the Precambrian basement. A new Wilson Cycle
was initiated by the formation of a rift in Late Neoproterozoic. After syn-rift
sediments, the rift is aborted, but the regular subsidence allowed for the
deposition of a thick pile of Paleozoic sediments in a shallow-water environment.
This intracontinental basin is inverted in mid-Carboniferous, forming the Anti-
Atlas fold belt. The main feature of this chain is the lack of a major décollement
or duplex structure.
Inspection of the structure was carried out during fieldwork to characterize its
structural style. The two principal parameters which control the deformation type
were (1) the pre-structuration of the basement during the Neoproterozoic –
Cambrian rift and (2) the ultraweak mechanical propriety of the cover. Variscan
compression led to the inversion of the basements former normal faulting. On a
larger scale, the cover accommodates the bed-length shortening by the
formation of small-scale compressive structures observable in the rigid members,
which mark the Paleozoic series. The analysis of these structures shows two
distinct orientations of deformation. One corresponding to the Anti-Atlas’ major
trend; the second to the Ougarta chain, situated to the southeast. The
superimposition of these two orientations created an egg-box interference
pattern.
A later extension, attributed to the opening of the Mesozoic High Atlas basin,
reactivated the deep structures, creating normal faults and extensional fault-
related folds in the Paleozoic cover.
Relationship with post-Variscan sediments show that the High Atlas inversion did
not deform the eastern Anti-Atlas. The influence was felt more in terms of a
smooth uplift. This uplift combined with a recent thermal anomaly, are
responsible for the high topography of the eastern Anti-Atlas.
SOMMAIRE
Imprimatur I
Dédicace III
Résumé V
Abstract VI
Sommaire VII
Sommaire des annexes VIII
1. Introduction
1.1. Le projet 1
1.2. Situation géographique et bases géologiques 2
1.3. Buts de l’étude 3
1.4. Méthodes, outils, données 5
1.5. Logistique et approche 6
Références 87
Remerciements 97
B. Articles
- Variscan inversion tectonics and interference pattern in the Tafilalt, Anti-Atlas, Morocco. Soumis
le 14.03.06 au Journal of African earth Sciences. Charles Robert-Charrue & Martin Burkhard.
- Inversion tectonics in the eastern Anti-Atlas of Morocco. Soumis à l’ AAPG Bulletin. Charles
Robert-Charrue & Martin Burkhard.
- Burkhard M., S. Caritg, U. Helg, Ch. Robert-Charrue, A. Soulaimani (2006), Tectonics of the
Anti-Atlas of Morocco, C. R. Geoscience, 338, pp. 11-24.
- Helg U., M. Burkhard, S. Caritg, and C. Robert-Charrue (2004), Folding and inversion tectonics
in the Anti-Atlas of Morocco, Tectonics, 23, TC4006, doi :10.1029/2003TC001576.
- Late Variscan Inversion Tectonics in the Paleozoic Tafilalt basin of South-Eastern Morocco.
Abstract. MIOG 2002.
- The Anti-Atlas fold-belt of Morocco : Thick skin inversion tectonics in the hinterland of the
alleghenian orogeny. Abstract. RST 2004.
- The Anti-Atlas fold belt of Morocco : variscan inversion tectonics and interference pattern of an
“intracratonic” basin. Abstract. SGM 2005.
- Structural study of the eastern Anti-Atlas, the variscan belt of Morocco. Poster. SGM 2005.
D. Figures en format A3
E. Atlas Landsat 7
- NW Africa
- Anti-Atlas
- AA oriental : découpage et légendes de la mosaïque
- Mosaïque de l’Anti-Atlas oriental
NB : Les figures marquées d’un astérisque dans le texte, sont de taille réduite. Elles sont
présentées dans l’annexe D dans un format plus grand.
1
INTRODUCTION
1.1. Le projet
Figure 1. Situation de l’Anti-Atlas du Maroc. Gauche : mosaïque de vue satellite. Droite : carte
géologique simplifiée.
Figure 2. Carte générale du Craton Ouest Africain. D’après Dallmeyer et Lécorché (1991) cité
dans Piqué (2001).
Les études publiées sur la déformation varisque de l’Anti-Atlas oriental sont rares
(A. Piqué dans son ouvrage sur la géologie du nord-ouest de l’Afrique (2001)).
Les premières études faites dans le sud-ouest se sont davantage focalisées sur
les plis et les relations entre les longueurs d’onde différentes observées dans les
bancs marqueurs de la couverture. Cette déformation de type flambage
(buckling), où chaque banc se plisse indépendamment selon ces propres
caractéristiques, n’est pas observée à grande échelle dans l’Anti-Atlas oriental
où la déformation semble affecter l’ensemble de la couverture d’une manière
homogène. Un des buts est donc de caractériser cette différence. L’Anti-Atlas
oriental se poursuit au SE par la chaîne de l’Ougarta (Figures 1 et 2), il diffère
également par la présence de structures extensives qui n’affectent pas la partie
occidentale. L’interférence de ces différents traits géologiques constitue une clef
dans la compréhension de l’édifice. Les objectifs pour cette étude sont de fournir
une image de la géométrie globale de l’Anti-Atlas oriental, d’en définir le style
structural et d’en quantifier la déformation. L’ambition ultime étant de proposer
un modèle dynamique expliquant la chronologie et les paramètres contrôlants la
mise en place de cette partie de la chaîne.
km2. Ces données sont gratuites et disponibles sur internet [3]. Le programme
Surfer 8.01 de Golden Software, Inc. permet la génération de cartes et la
création de vues tridimensionnelles.
Trois lignes sismiques fournies par l’ONAREP (actuellement ONHYM) ont permis
de mieux contraindre le modèle géométrique en profondeur. Une première
interprétation de la ligne RS 8 a été réalisée dans le cadre du travail de diplôme.
Cette interprétation a été revue et corrigée pour le premier article.
Le travail de terrain a été effectué seul. La carte d’accès (Figure 3) montre les
endroits visités lors des différentes sessions de terrain dans l’Anti-Atlas oriental.
A l’exception de la dernière session, les déplacements pour le travail de terrain
se sont faits en Landrover® Defender aménagé, permettant une autonomie de
plusieurs jours, afin d’éviter des trajets inutiles.
L’approche du terrain a été dictée par les recherches. Outre les mesures de la
géométrie générale, on a, dans un premier temps, cherché plus particulièrement
à observer le contact socle - couverture et à définir l’épaisseur des formations
dans les différentes régions de l’Anti-Atlas oriental. Des profils de coupes ont été
tracés perpendiculairement à l’axe des boutonnières, selon les variations de la
structure générale et les possibilités d’accès. Le terrain est très facilement
abordable par les routes nord (Erfoud-Tinejdad-Tineghir-Boumalne) et centrale
2
CADRE GEOLOGIQUE
Etat des connaissances
L’Anti-Atlas est situé sur la bordure nord du craton ouest africain (Figure 2, p. 3).
Cette zone est particulière dans le sens que de nombreux événements
tectoniques affectent cette région au cours des temps géologiques. Sur la base
des compilations paléogéographiques de Stampfli et Borel (2002) (Figure 4),
voici un aperçu des mouvements tectoniques globaux précédents et suivants
l’épisode varisque.
Trias et durant le Jurassique (200 Ma et 180 Ma), la Pangée est disloquée par
l’ouverture de l’océan Atlantique.
Les événements décrits ci-dessus ne constituent qu’un seul cycle de Wilson (rift
- ouverture océanique - fermeture du bassin - collision - rift). D’autres cycles ont
précédé et suivent ce cycle varisque.
Si l’Anti-Atlas est toujours à proximité des limites actives, il faut noter qu’il ne se
situe jamais sur ces limites. Les séries plus marines correspondent à des
niveaux marins élevés. La forme du bassin Anti-Atlas peut être représenté par un
indentation marine ou un golfe. Toujours sur le continent Gondwana, cette zone
enregistre plus ou moins intensément les événements tectoniques, mais ne les
subit pas de manière directe.
ordovicienne qui peut être suivie dans tout l’Anti-Atlas. L’image à grande échelle
du Paléozoïque montre une série de dômes et bassins. D’ouest en est : la
cuvette de Tazzarine, les « bassins » du Mader et du Tafilalt. Cette structure
concorde avec celle de la chaîne de l’Ougarta d’orientation NW-SE. De grandes
structures d’extension tel que le graben de Zagora apparaissent dans les
niveaux rigides de la couverture. Le tout est scellé par des sédiments du Crétacé
ou plus récents. On distingue le plateau Crétacé des Monts Kem-Kem au sud-
est, la Hamada du Guir Mio-Pliocène à l’est, le Crétacé du nord de l’Ougnate et
le bassin Cénozoïque de Ouarzazate sur la bordure NW du Saghro. Au nord-
ouest de l’Anti-Atlas, le Haut Atlas est limité au sud par la faille sud atlasique.
Ces formes géomorphologiques sont visibles sur les vues tridimensionnelles de
la Figure 6.
2.3. Le socle
Avec une impressionnante variété de roches, le socle est le témoin d’une activité
géologique multiphasée qui explique sa complexité. D’un point de vue
« varisque », le socle comprend toutes les roches antérieures aux sédiments
déposés lors du rift Précambrien (début du cycle de Wilson varisque). Il n’en va
pas toute à fait de même en terme de rhéologie et de déformation.
Figure 7. Limites cratoniques de l’Anti-Atlas. SAF : South Atlas Fault ( Faille sud atlasique),
AAMF : Anti-Atlas Major Fault (Accident Majeur), SGA : Saoura gravimetric anomaly (Anomalie
gravimétrique du Saoura).
(Saquaque et al., 1989), cette structure est aujourd’hui décrite comme la limite
australe d’un bassin aulacogène du Néoprotérozoïque inférieur (Hefferan et al.,
2000 ; Ennih & Liégeois, 2001; 2003). Les ophiolities trouvées sur cet accident,
sont en fait charriées sur le craton (et le bassin aulacogène) lors de l’accrétion
Pan-Africaine d’un arc volcanique. Ainsi la limite septentrionale du craton ouest-
africain est la faille sud-atlasique (SAF). La zone intermédiaire peut être qualifiée
de métacratonique (Liégeois et al., 2005). La limite orientale de ce métacraton
semble coïncider avec l’anomalie gravimétrique du Saoura (SGA, Figure 7)
(Bayer et Lesquer, 1978 cités dans Ennih et Liégeois, 2001).
Figure 8. Situation de l’Anti-Atlas par rapport à la chaîne des Appalaches à la fin du Paléozoïque.
Les contours isopaques (en km) sont données pour les bassins sédimentaires paléozoïques qui
n’ont subi que peu ou pas de déformation. Le même code couleur est appliqué à la chaîne
plissée de l’Anti-Atlas afin d’indiquer la profondeur estimée de ce bassin avant son inversion. Les
massifs de socle alléghaniens sont indiqués en bleu. Les parties internes, métamorphiques et
plus anciennes, de la chaîne des Appalaches-Mauritanides-Meseta marocaine sont colorées en
vert et en rose.
Toute la série Paléozoïque est dominée par des sédiments déposés dans un
environnement de mer peu profonde. L’absence de sédiment de bassin, de
talus ou de plate-forme externe exclut une interprétation de marge passive.
L’environnement de dépôt de l’Anti-Atlas s’apparente plutôt à un bassin de type
intracontinental (Burkhard et al., 2006), dû à une subsidence thermique post-rift.
Mais le régime thermique, la structuration pré-rift et les effets d’une compression
lointaine qui contrôlent la formation du bassin (Cloething et al., 1995 ; Ziegler et
Cloething, 2004) sont mal contraints. Le bassin Paléozoïque de l’Anti-Atlas est la
partie tectonisée du gigantesque bassin de Tindouf (Coward et Ries, 2003).
Cambrien
La série de l’Adoudounien (Buggish et Flügel, 1988 ; Algouti et al., 2001) se
dépose après les Conglomérats de base couronnant le super-groupe de
Ouarzazate. L’Adoudounien comprend les Calcaires inférieurs, la série Lie de vin
et les Calcaires supérieurs (Figure 9). Dans le sud-ouest, des signes du rifting
amorcé au Néoprotérozoïque terminal persistent jusqu’à la fin du Cambrien
inférieur (Benssaou et Hamoumi, 2003). Cette série est fortement réduite dans
Ordovicien
L’Ordovicien demeure toujours à tendance détritique. L’Ordovicien inférieur
débute par une puissante série de schistes argileux et se termine par les grès et
quartzites du 1er Bani. L’Ordovicien supérieur ressemble beaucoup à l’inférieur,
mais d’une plus petite épaisseur. Le niveau marqueur qui le couronne est le 2ème
Bani. Il comprend des microconglomérats et des surfaces érodées témoignant
d’une glaciation (Choubert et Faure-Muret, 1983).
Silurien
Tectoniquement, cette période correspond à l’ouverture de l’océan Proto- et
Paléotéthys et à la dérive des terrains Huniques. Après la dominance du
détritisme provenant du craton africain, le Silurien marque le passage à une
sédimentation carbonatée. Ce changement est expliqué par une hausse du
niveau marin qui commence par déposer une épaisse série de black shales,
constituant une importante roche mère en Afrique du nord (MacGregor, 1996 ;
Boote et al., 1998). S’ensuit l’apparition de bancs calcaires qui vont dominer
durant le Dévonien.
Dévonien
Durant cette période, la marge du Gondwana bénéficie d’un calme tectonique.
Les Richs Dévoniens témoignent d’un niveau marin toujours haut. La
composante calcaire est plus développée dans la partie orientale. Les calcaires
Carbonifère
Les sédiments Carbonifères témoignent d’un retour marqué du détritisme. Après
une épaisse série de schistes argileux, la tendance devient gréseuse. Des
figures de chenaux, des marques de courant et des bioturbations décrivent un
environnement deltaïque. On trouve dans la partie nord de l’Anti-Atlas oriental
des figures sédimentaires interprétés comme des tempestites. Les calcaires
couronnant la série dans le sud-ouest sont absents dans la partie orientale.
S’il est vrai que la Meseta (Figure 8, p.17) fait également partie de l’édifice
varisque, des corrélations avec l’Anti-Atlas sont hasardeuses. Malgré sa
proximité actuelle, sa position dans l’orogène varisque a été nettement plus
interne. Elle a subi des déformations bien avant l’Anti-Atlas (Hoepffner et al.,
2005). Aujourd’hui la relation géométrique qui lie l’Anti-Atlas à la Meseta est
masquée par le Haut Atlas.
partie du Haut Atlas (Saddiqi et al., 2005 ; Ayarza et al., 2005). Les modèles
géophysiques (gravimétriques) de Missenard et al. (2006) démontrent un
amincissement lithosphérique (<70 km) orienté NE-SW. Cette anomalie est
oblique par rapport au système atlasique et correspond géographiquement au
volcanisme alcalin Néogène et Quaternaire, ainsi qu’à une bande d’activité
sismique. Initiée par remontée mantellique durant le Miocène (Missenard et
al.,2006), l’anomalie est actuellement davantage entretenue par la convergence
tectonique des plaques Ibéro-Africaine et du craton saharien (Teixell et al.,2005).
3
STRUCTURES
Description des observations
3.1. Introduction
Figure 11*. Carte géologique de l’Anti-Atlas central et oriental. Cette carte n’illustre que les
roches incorporées à la déformation varisque. Ainsi toutes les roches post varisques
apparaissent en noir. Le rouge a été attribué aux roches les plus anciennes, le bleu aux roches
les plus jeunes.
Figure 12. Image satellite Landsat avec la mise en valeur de certains contacts.
Le comportement mécanique des roches est mise en valeur par l’analyse des
structures. Elle sera néanmoins discutée au préalable, puisqu’elle est un des
principaux paramètres qui contrôlent le style de déformation.
Socle
Malgré une composition des plus hétérogènes et une forte anisotropie, les blocs
de socle réagissent de manière homogène, comme des unités rigides avec de
très faibles déformations internes. Ceci est visible sur les images Landsat
(Figures 3 et 12). Mise à part la boutonnière de Bou Azzer, aucune relation
géométrique directe ne lie la surrection des boutonnières et les différents
éléments qui les composent, autant en terme d’âge (Eburnéen, super-groupe de
l’Anti-Atlas ou de Ouarzazate) qu’en terme pétrologique (métamorphique,
plutonique, volcanique ou sédimentaire). Par contre, les failles normales
développées lors du rift précambrien prédéfinissent la structure du rejeu de la
déformation varisque. Elles n’affleurent malheureusement pas dans l’Anti-Atlas
oriental, mais certaines structures dans la couverture fournissent des indices sur
leurs positions. Pour la partie orientale, le rift précambrien est davantage déduit
de l’analyse de la nature des sédiments que de l’observation directe de
structures. Il n’est pas non plus exclu que ces structures extensives soient déjà
héritées d’événements Pan-Africains ou Eburnéens.
Couverture
La couverture paléozoïque est très nettement dominée par de puissantes séries
de faible compétence. Elle est ponctuée de niveaux marqueurs compétents.
Ceci explique la morphologie typique de la couverture sous la forme de
succession de cuestas (Figure 14).
Figure 14. Cuesta au sud de la boutonnière de l’Ougnate. Ici sous le 2ème Bani (Ordovicien
supérieur).
Figure 16. Effet de la mécanique sur un pli forcé ou de drapage (d’après Groshong, 2002). En
rouge : unité rigide, en jaune : série incompétente, en bleu : de compétence intermédiaire.
réduits, car la couverture est elle-même réduite, mais aussi parce qu’ils ont été
masqués par l’inversion du Haut Atlas.
L’analyse du tracé du sommet de l’Ordovicien supérieur (2ème Bani) et celui de
l’Ordovicien inférieur (1er Bani) montre une structure d’un ordre inférieur
(structure majeure secondaire). Une succession de synclinaux et d’anticlinaux
dont les axes sont orientés globalement N-S à NW-SE et plongent vers le sud-
est. Mise en valeur par les membres dominants de l’Ordovicien, cette structure
est également soulignée par les affleurements de socle (resserrement du
Saghro, ensellement Saghro-Ougnate et dépression entre l’Ougnate et
l’affleurement de l’extrémité est) et par les différents « bassins » Dévoniens et
Carbonifères (Cuvette de Tazzarine, « bassin » de Mader et « bassin » du
Tafilalt). Cette structure globale est également visible grâce à la carte géologique
remaniée (Figure 11, p.26).
Figure 17*. Coupes géologiques de l’Anti-Atlas oriental. Image Landsat avec la position des
coupes. Epaisseurs des séries dans les différentes coupes.
Les coupes occidentales (10 à 5) montrent des failles normales très localisées
créant notamment le graben de Zagora. A l’est (coupes 1 à 4), les failles
normales sont plus nombreuses, mais avec des rejets plus faibles.
Les plis sur rampe dans la partie nord des coupes 6 et 7 sont une interprétation.
Ils servent à contre-balancer la déformation particulièrement intense des
sédiments Carbonifères dans cette zone. Le pli sur rampe – chevauchement de
la coupe 8 est lui observé en surface ; sa géométrie en profondeur reste
toutefois interprétative.
Les parties du Haut Atlas sont dessinées à partir des coupes de Frizon de
Lamotte et al. (2000).
La Figure 18 montre les deux cas. Les bancs de calcaire sont plus cassants, la
déformation provoque un chevauchement. Au-dessus de cette structure, la
lithologie est plus incompétente, le banc n’est « que » plissé. La déformation est
diffuse, elle se répartit dans tout le volume de la couche. Lorsque les conditions
d’affleurement permettent l’observation des séries incompétentes, on distingue
une multitude de petits chevauchements presque parallèles au litage (Figure 19).
Figure 19. Plans striés de chevauchements mineurs dans des schistes argileux (Ordovicien
inférieur), illustrant le style caractéristique de déformation interne dans les séries incompétentes.
Sur la bordure nord du Saghro : N 31°24.361’ / W 005°30.192’. La boussole est orientée au
Nord.
Figure 20. Schéma illustrant l’indépendance des petites structures et leur différentes formes en
relation avec leur compétence. A : Structure en « queue de poisson » (Fishtail), illustrée par la
Figure 17. B : Plissement harmonique de type « buckling ». Présence de cisaillement banc sur
banc. C : Plissement d’un banc dominant, présentant des poly- ou des disharmonies. D :
Multitude de microchevauchements sub-parallèles à la stratification (Figure 19). E : Faille inverse
dans un banc massif.
Légendes de la Figure 20 :
A : Plis dans le Dévonien moyen. Vue vers le SE. Bordure ouest du Tafilalt.
N 31°16.800’ / W 004°20.400’. Marteau en bas de l’image.
B : Pli dans le Dévonien moyen. Vue vers le NW. Bordure ouest du Tafilalt.
N 31°15.750’ / W 004°23.400’. Echelle 1 m.
C : Pli dans le Dévonien moyen. Vue vers le NW. Bordure ouest du Tafilalt.
N 31°15.550’ / W 004°23.540’. La falaise fait 200 m de long.
D : Structure majeure secondaire et petits plis dans le Cambrien moyen. Vue vers le SE.
Appendice Cambrien au sud de l’Ougnate. Depuis N 31°17.194’ / W 004°55.591’. Le synclinal au
premier tiers inférieur mesure environ 10 m de large.
E : Plis disharmoniques dans le Dévonien. Vue vers le Nord. Sud de l’Ougnate.
N 31°15.020’ / W 004°48.360’. Marteau au centre.
F + G : (la vue G est à la base de la vue F) Plis dans le Dévonien. Vue vers le NW et NNW. Au
sud de l’Ougnate. N 31°13.220’ / W 004°53.760’. Echelle 1 m.
H : Plis dans le Carbonifère. Vue vers le NE. Jebel Tisdafine (Carbonifère au nord de
l’ensellement Saghro – Ougnate). N 31°27.150’ / W 005°18.950’. Echelle 1 m.
J : Pli-faille dans le 1er Bani. Vue vers le WSW. Bordure NNW du Saghro. Depuis N 31°24.120’ /
W 005°37.640’. L’anticlinal fait 100 m de large.
K+L : (vue générale et détail) Pli dans le Carbonifère. Vue vers le NE. Jebel Asdaf, lambeaux
Carbonifère isolé dans l’Oued Todra à l’ouest de l ‘Ougnate. N 31°25.455’ / W 005°12.368’. K :
250 m de large, L : échelle 1 m.
M : Pli déversé vers le SE dans le Dévonien. Vue vers le NE. Au nord du Saghro. N 31°27.240’ /
W 005°30.950’. Echelle 1 m.
N : Pli couché dans le Dévonien. Vue vers l’est. Au nord du Saghro.
N 31°28.200’ / W 005°31.400’. Marteau en bas à gauche.
Figure 22*. Stéréogrammes des structures mesurées. Jaune : petits plis. Bleu :
chevauchements. Rouge : structures extensives. Vert : structure générale.
Projection de Wulff, hémisphère inférieur.
Légendes de la Figure 22 :
A : Chevauchement dans un banc massif du Dévonien moyen. Vue vers le SE. Centre du
Tafilalt. N 31°19.900’ / W 004°10.050’. Marteau au centre.
B+C+D : Chevauchements dans le 2ème Bani. Sud de l’Ougnate. N 31° 11.250’ / W 004°55.250’.
B : Plan de chevauchement avec un angle aigu par rapport à la stratification, le plan de faille est
matérialisé par une minéralisation siliceuse, vue au SSE, le second plan est à 100 m. C + D :
Chevauchement et crochon dans le toit, vue au SE, bâton 1 m.
E+F+G+H : Chevauchement dans le 1er Bani à l’ouest de Tazzarine. N 30°44.250’ / W
006°04.450’. E+F : Chevauchement et crochon, vue vers le SE, le chevauchement mesure 30 m.
G : Plan strié dans le crochon, vue vers le Sud, boussole orientée. H : Crochon supérieur,
« diffusion » du chevauchement, vue vers l’est, longueur du chevauchement 30 m.
J : Structure Fishtail dans le Dévonien. Sud de l’Ougnate. N 31°15.000’ / W 004°48.375’. Echelle
1 m.
K : Faille inverse et chevauchement dans le Cambrien. Vue vers le NE. Au SE du Saghro (nord
de Tazzarine). Depuis N 30°56.610’ / W 005°31.210’. Palmier dans l’oued.
L : Faille inverse minéralisée et crochon supérieur dans l’Ordovicien inférieur. Vue vers le WSW.
Bordure NNW du Saghro. N 31°24.115’ / W 005°37.525’. Echelle 1 m.
M : Chevauchement et « pli sur rampe » dans le Cambrien moyen. NNW du Saghro. N
31°23.320’ / W 005°37.930’ (sommet). La falaise fait 200 m de long.
Figure 24. Schistosité dans les plis déversés Dévoniens. Vue vers l’est. Au nord de la
boutonnière du Saghro. N 31°28.200’ / W 005°31.400’. Marteau au centre.
Légendes de la Figure 24 :
A : Faille normale dans le sud de l’Ougnate. Vue vers l’ouest. Depuis N 31°16.590’ / W
004°56.025’. Le sommet de gauche est à 4 km. Sans considérer le premier plan, les deux
cuestas (au sud et au nord) sont du Cambrien moyen. Cette répétition est due à une faille
normale qui plonge au nord. On trouve du socle au pied de la falaise de gauche.
B : Faille normale dans l’Ougnate. Vue vers l’est. Depuis N 31°20.945’ / W 005°00.800’.
Landrover en bas à gauche. La faille normale plonge au nord et met en contact le socle et la
cambrien inférieur.
C+D : Faille normale au nord de l’ensellement Saghro-Ougnate. N 31°20.300’ / W 005°16.850’.
C : vue vers l’ouest, le bloc central (à 300 m) est le 1er Bani de l’Ordovicien, il est en contact (au
sud) avec du Cambrien moyen, la faille normale plonge vers le Nord. D : vue vers l’est, la faille
normale est dans l’Ordovicien inférieur, la particularité est que les stries de la faille normale sont
sur-imprimées de stries indiquant un mouvement dextre. Sac à dos pour échelle.
E : Faille normale dans l’Ordovicien inférieur. Vue vers le NE. Bordure nord du Saghro. Depuis N
31°23.995’ / W 005°37.360’. La faille plonge au SE.
F+G+H : Graben à l’est de la cuvette de Tazzarine. Zone N 30°45’ / W 005°13’. F : Limite nord
du graben. Vue au Nord. Au fond les falaises du 1er Bani, plus en avant la faille normale
Figure 26. Vue Landsat et en coupe d’un synclinal lié à une faille normale au sud de la
boutonnière de l’Ougnate.
Une toute autre manière de signaler leur présence sont les synclinaux que les
failles normales provoquent dans les formations du toit, lorsque ces dernières,
de faible compétence, « glissent » sur des roches du mur très rigides. Ces plis
d’entraînement ont souvent été interprétés comme plis varisques. Ces structures
liées à l’extension sont d’autant plus déroutantes que comme tous les plis, elles
présentent du cisaillement chevauchant entre les bancs. Ces synclinaux sont
plus ou moins horizontaux. Cela varie en fonction de la configuration de la
structure majeure. Ils se caractérisent par un flanc limité par la faille
(généralement le plus raide) et un flanc en concordance avec le reste de la série
(Figure 26). Ces structures ont été observées dans le système du rift de la Mer
Rouge (Khalil et McClay, 2002) et sont connues dans d’autres systèmes en
extension (Schlische, 1995). Elles sont également décrites dans des modèles
expérimentaux (Withjack et al., 1990 ; Hardy et McClay, 1999 ; Finch et al.,
2004 ). Les modèles avec une couverture ultra-incompétente semblent le mieux
correspondre à la configuration observée dans l’Anti-Atlas oriental. Selon Xiao et
Suppe (1992), le développement de ces structures est induit principalement par
une géométrie de faille de type listrique.
Toutes les failles normales observées sont très raides, leur plongement est
typiquement compris entre 70 et 80°. L’orientation de leur trace est constante
dans tout l’Anti-Atlas oriental : E-W avec de légères variations de WSW-ENE à
WNW-ESE. Aucune autre orientation n’a été observée. Outre les grabens
symétriques, une majorité des failles normales plongent vers le nord, à
l’exception de la partie nord des boutonnières où les plongements sud dominent.
Figure 28*. Modèle numérique de terrain de l’Anti-Atlas oriental. Données de GTOPO 30.
4
MODELE GEOMETRIQUE
Interprétation des observations
entre la préstructuration et la contrainte est trop obtus) (Bump, 2003 ; Yamada &
MacClay, 2003). (2) Une part importante du raccourcissement horizontal n’est
pas localisée en un seul horizon, mais est accommodée de manière diffuse dans
l’important volume de roche incompétente. (3) Le niveau d’érosion actuel ne
permet pas encore d’observer ces failles inverses. Elles se trouvent « au large »
des boutonnières, sous la couverture. La présence et la localisation de ces
failles inverses sont toutefois trahies par les failles d’extension tardive qui
affectent la couverture au moins jusqu’au 1er Bani.
Figure 29. Principe de trishear (d’après Erslev, 1991). A : Solution classique. Elle nécessite un
décollement basal. Il en résulte un excès de volume de la couverture. B : Le modèle trishear
palie à ces problèmes. C : Répartition des vecteurs de déplacement dans la zone triangulaire.
Figure 30. Modèle évolutif du style structural de l’Anti-Atlas oriental (stades A à C repris de
Bump, 2003). A : Rift et sédiments syn-rift. B : Dépôt des sédiments post-rift. C : Inversion avec
raccourcis dans le socle et drapage de la couverture (trishear). D : Extension. Les failles
normales recoupent la couverture.
Choix de la coupe
Le calcul de la profondeur de détachement a été appliqué sur la coupe 8. Cette
coupe a été choisie, car elle offre le niveau d’affleurement le plus bas dans la
série. Ainsi le minimum d’erreur d’interprétation sur la morphologie du contact
socle-couverture est garanti. La coupe 8 montre un bon développement de la
structure anticlinale majeure. Ceci implique qu’elle coupe un bloc de socle en
son milieu. La coupe 7 passant dans l’ensellement entre les boutonnières du
Saghro et de l’Ougnate, n’aurait pas été significative, puisque ce secteur peut
être interprété comme une zone de transition entre les deux blocs.
Restauration
La Figure 31 illustre les différents stades dans la restauration de la coupe. Afin
d’établir la situation à la fin du paroxysme varisque, il faut tout d’abord restaurer
les failles normales tardives (passage du stade B au stade C). Ce profil est
ensuite remis à plat (D). A cette étape, les couches contiennent encore une
composante importante de raccourcissement interne. Il faut par conséquent les
étirer pour obtenir la configuration à la fin de la sédimentation Paléozoïque. Le
profil A montre l’effet de la surrection Néogène. Ces données proviennent du
modèle développé par Missenard et al. (2006) à partir de données
gravimétriques. Cette composante récente est présente dans le relief actuel et
doit être déduite du profil pour obtenir la configuration réelle du paroxysme
varisque.
Figure 31. Restauration de la coupe 8. A : Surface différentielle due à la surrection Néogène Su,
déduite de Missenard et al. (2006, Figure 6.C, profil oriental). B : Coupe actuelle ; Sx est la
surface en défaut due à l’extension tardive. C : Coupe de la fin de l’événement varisque. Ce profil
contient toutefois encore l’effet de la surrection récente. D : Coupe restaurée montrant la
situation avant l’inversion mais comprenant encore le raccourcissement interne. E : Coupe
restaurée, sans raccourcissement interne.
Figure 32. Détachement provoqué par extension. Définition des éléments géométriques utilisés
dans le calcul de la profondeur de détachement.
La surface en défaut Sx, située sous le niveau régional Hr (Figure 32) résulte
d’une extension. Elle est compensée par la surface déplacée définie par la
longueur du déplacement horizontal Dx multipliée par hx la profondeur du
détachement sous le niveau régional Hr :
Sx
(1) Sx = Dx ⋅ hx ainsi hx =
Dx
(2) Dx = Lx − Wx
Sx
(3) Hx = Hr − hx = Hr −
Lx − Wx
la droite pointillée désigne une profondeur de -16 km. La pente de cette droite
donne l’inverse du déplacement (1/Dx). On obtient une valeur pour Dx de 0.204
km ce qui concorde avec les valeurs mesurées.
Figure 33. Détachement par compression. Cas de figure sans et avec raccourcissement interne
(Dce).
Sc
(1) Sc = Dc ⋅ hc ainsi hc =
Dc
(2) Dc = Lc − Wc
Sc
(3) Hc = Hr − hc = Hr −
Lc − Wc
Appliquée telle quelle, cette formule fournit des valeurs extrêmement profondes
de -111 à -126 km pour le partie Saghro et de -172 à -212 km pour la coupe
entière (Tableau 2). Etant donné les excellentes conditions d’affleurement,
l’absence totale de chevauchement ou d’autres grandes structures est un fait
avéré. L’équilibrage de ces coupes nécessite la considération d’une composante
plus discrète. Le style structural, dicté par une rhéologie particulière, ne permet
pas de changer les valeurs des déplacements horizontaux autrement que par
raccourcissement interne. Cette déformation interne est par ailleurs très bien
documentée par les structures de plus petite taille ne figurant pas sur les
coupes.
Dce
(4) e= ainsi Dce = Lo ⋅ e
Lo
(5) Lc = Lo − Dce
Lc
(6) Lc = Lo − Lo ⋅ e = Lo(1 − e) or Lo =
1− e
Sc
(7) hce =
Dc + Dce
Dc est remplacé par l’équation (2), Dce par la (4) et Lo par la (6). Cette équation
se simplifie ensuite comme suit :
Sc Sc Sc
(7’) hce = = =
Lc − Wc + Lo ⋅ e Lc − Wc + Lc ⋅ e e
Lc1 + − Wc
1− e 1− e
Graphique 2. Profondeur de détachement des différents niveaux sur la coupe entière en fonction
de la déformation compressive.
Figure 34* : Faisceaux dans l’Anti-Atlas oriental. Avec et sans le fond Landsat.
orientale sont caractérisés par une longueur d’onde plus courte et une amplitude
supérieure par rapport à la zone plus à l’ouest. En effet, l’anticlinal séparant le
Tafilalt du Mader est nettement plus étroit que celui séparant le Mader et la
cuvette de Tazzarine (longueur d’onde). Les variations d’amplitude sont
montrées par l’affleurement des roches plus ou moins profondes. Dans la partie
orientale, des roches profondes (Cambriennes au moins) se trouvent à proximité
géographique de roches Carbonifères, alors que dans la partie occidentale, les
affleurements varient du 1er Bani (Ordovicien) au Dévonien (voire même jusqu’au
Silurien uniquement). Cette observation peut être expliquée par les variations
d’épaisseur qui affectent la couverture. Il est évident que si la couverture est plus
fine, la probabilité de voir affleurer des roches plus ancienne est plus grande.
Mais cette constatation est réfutée par l’observation des pendages. Les
pendages sont en effet beaucoup plus forts à l’est et plus plats à l’ouest. Ces
observations démontrent également l’implication du socle dans la déformation
d’orientation ougartienne. L’interférence obtenue dans cette partie de la chaîne
découle donc du croisement de deux déformations toutes deux de type thick
skin.
Les failles plongeant au nord sont majoritaires, on les interprète comme les
principales, alors que les failles qui plongent au sud sont leurs conjuguées. Cette
interprétation est influencée, mais également supportée, par la connaissance de
la configuration du craton.
Malgré le fait que toutes ne recoupent pas toujours l’entière série Paléozoïque,
les failles normales sont attribuées à une extension post-varisque. La rhéologie
particulière de la couverture explique le fait qu’elles n’affleurent pas
systématiquement. Le modèle trishear en extension de la Figure 35 illustre ce
phénomène. Les niveaux marqueurs suffisamment séparés de la faille par des
séries incompétentes, forment des plis de drapage ou plis de revêtement. Ceci
explique l’absence de faille normale dans les sédiments supérieurs de la
couverture. Certains larges plis sont interprétés comme étant le seul résultat des
failles normales, comme par exemple les plis Dévoniens et Carbonifères dans et
sur la bordure sud-ouest du Mader (voir au tiers sud de la coupe 7). Le fait que
ces failles affectent des formations plus jeunes dans l’est qu’à l’ouest, s’explique
par le simple fait que la couverture y est moins épaisse.
Avec l’hypothèse que ces failles réutilisent les anciennes discontinuités du socle,
leur emplacement permet la localisation de ces structures profondes. La forte
inclinaison des failles normales dans la couverture permet une extrapolation
aisée. Toutefois, mise à part la projection, ces failles ne définissent pas
directement le front des blocs inversés lors de l’événement varisque. Tout
comme les failles inverses prennent un raccourcis dans le socle, les failles
normales empruntent un chemin légèrement différent. Elles se développent de
manière verticale et recoupent la couverture (Figure 30.D, p.55), sans utiliser le
chemin créé par l’inversion. La faille inverse étant trop plate en surface
(Faccenna et al., 1995).
est plus grande. On peut en déduire que la taille des blocs de socle est inférieure
dans cette zone et donc que le socle y est plus fracturé qu’ailleurs. Il est possible
que cette intensité de fracture traduise un interférence des structures profondes
des deux orientations (Anti-Atlas et Ougarta).
Cette phase extensive n’affecte que l’Anti-Atlas oriental. Elle est interprétée
comme le témoignage de l’ouverture du bassin adjacent du Haut-Atlas. Ceci
explique le fait que cette phase n’apparaisse pas dans le reste de l’Anti-Atlas.
Les premiers sédiments à se déposer dans ce bassin datent du Trias. Toutefois
comme aucun sédiment Triassique ou Jurassique ne se dépose sur l’Anti-Atlas,
il est possible que cette extension soit uniquement un effet de bord et que cette
extension n’affecte pas forcément l’Anti-Atlas dans les premiers stades de
l’ouverture de ce bassin.
Figure 36*. Topographie et géologie de surface selon les coupes de l’Anti-Atlas oriental. La
coupe 10 comprend également une courbe montrant l’influence de la surrection Néogène
(dessinée d’après Missenard et al. (2006, Figure 6.C, profil oriental).
Le reste du profile est dominé par des failles normales tardives. Une série de
« pseudo-bassins » (p. ex. km 25) ont été anciennement interprétés comme des
Figure 37. Image satellite du Tafilalt avec la position des lignes sismiques RS 8 et RS 10.
5
MODELE DYNAMIQUE
Chronologie des événements
Figure 39. Modèle du détachement médio-crustal pour la coupe 8. Sans exagération verticale.
La couverture comprend le Cambrien, l’Ordovicien et le Silurien. A : Profil avant la compression
avec les failles normales héritées du rift Précambrien. B : Profil au paroxysme de l’inversion
varisque. Les failles normales sont réutilisées mais recoupent le socle et s’aplatissent avant
d’atteindre la couverture. C : Extension. Les failles normales atteignent la couverture. D :
Soulèvement différentiel récent dû à une anomalie thermique (donnée de Missenard et al. 2006,
Figure 6.C).
soient correctement interprétées (Figure 40), cette chronologie relative n’est pas
forcément identique dans l’Anti-Atlas oriental, vu également la proximité qui lie
l’Anti-Atlas oriental à la chaîne de l’Ougarta. D’ailleurs, dans sa partie la plus
orientale, la préférence est donnée pour une orientation ougartienne précédant
la déformation anti-atlasique. En effet, les axes des petits plis d’orientation
ougartienne sont basculés par la structure majeure de la déformation anti-
atlasique (voir le stéréogramme le plus oriental de la Figure 22, p.40). Dans la
région de Tazzarine, l’analyse des deux stéréogrammes de petits plis semble
indiquer le contraire. Les petits plis d’orientation ougartienne ont un axe de pli
horizontal, alors que l’axe des petits plis d’orientation anti-atlasique est basculé
vers l’ouest. Ce basculement étant provoqué par la structure majeure secondaire
attribuée à la déformation ougartienne. Mais ceci peut être expliqué par le fait
que les déformations internes (petits plis) précèdent la déformation de plus
grande envergure (structure majeure). Au sud de la boutonnière de l’Ougnate
ces relations sont brouillées par les synclinaux provoqués par l’extension à venir.
Figure 41*. Modèle tridimensionnel du modèle numérique de terrain. Exagération verticale 10 :1.
Données GTOPO 30.
6
CONCLUSIONS
Apports de l’étude
6.2. Structures
Une phase d’extension tardive crée des failles normales dans la couverture et
provoque localement des synclinaux (anciennement interprétés comme plis
varisques). Ces structures extensives très raides permettent d’extrapoler les
limites des blocs de socle.
6.3. Tectonique
D’après son style structural, l’Anti-Atlas a des analogues tels que les bassins
inversés de l’orogénèse Laramide de Uinta ou de Wind River, décrits par Mitra et
Mounts (1998) ou Stone (1999). Ils ont en commun une déformation de type
trishear, induite par des mouvements verticaux de blocs de socle rigides, forçant
une couverture de faible compétence. L’Anti-Atlas dévie pourtant de ces
exemples, par l’aplatissement en surface des failles inversées. Ceci est rendu
possible grâce aux séries post-Panafricaines et syn-rift qui caractérisent la partie
Enfin, des essais en modèles analogiques, jouant les multiples réactivations des
blocs de socle, en variant la compétence et la puissance de la couverture,
seraient une expérience fort intéressante.
REFERENCES
BARBEY P., F. OBERLI, J.-P. BURG, H. NACHIT, J. PONS, M. MEIER (2004), The
Palaeoproterozoic in western Anti-Atlas (Morocco): a clarification, Journal of
African Earth Sciences, 39, pp. 239-245.
CLOETINGH S., J.D. VAN WEES, P.A. VAN DER BEEK, G. SPADINI (1995), Role of pre-
rift rheology in kinematics of extensional basin formation: constraints from
thermomechanical models of Mediterranean and intracratonic basins, Marine and
Petroleum Geology, Vol. 12, N° 8, pp. 793-807.
ENNIH N., J.-P. LIÉGEOIS (2001), The Moroccan Anti-Atlas: the West African
craton passive margin with limited Pan-African activity. Implications for the
northern limit of the craton, Precambrian Research, 112, pp. 289-302.
ENNIH N., J.-P. LIÉGEOIS (2003), Discussion, The Moroccan Anti-Atlas: the West
African craton passive margin with limited Pan-African activity. Implications for
the northern limit of the craton: reply to comments by E.H. Bouougri,
Precambrian Research, 120, pp. 185-189.
EPARD J.-L., R.H. GROSHONG (1993), Excess area and depth to detachment, The
AAPG bulletin, v. 77, No. 8, pp. 1291-1302.
FACCENNA C., T. NALPAS, J.-P. BRUN, P. DAVY (1995) The influence of pre-
existing thrust faults on normal fault geometry in nature and in experiments,
Journal of Structural Geology, Vol. 17, N° 8, pp. 1139-1149.
FRIZON DE LAMOTTE D., B. SAINT BEZAR, R. BRACÈNE, E. MERCIER (2000), The two
main steps of the Atlas building and geodynamics of western Mediterranean,
Tectonics, Vol. 19, N° 4, pp. 740-761.
GROSHONG R.H., 2002, 3-D structural geology: 2nd printing, Springer-Verlag, 324 p.
JOHNSON K.M., A.M. JOHNSON (2002, A), Mechanical models of trishear-like folds,
Journal of Structural Geology, 24, pp. 277-287.
MACGREGOR D.S. (1996), The hydrcarbon systems of North Africa, Marine and
Petroleum Geology, Vol. 13, N° 3, pp. 329-340.
STAMPFLI , G.M., G.D. BOREL (2002), A plate tectonic model for the Paleozoic and
Mesozoic constrained by dynamic plate boundaries and restored synthetic
oceanic isochrones, Earth and Planetary Science Letters, 196, pp 17-33.
THOMAS R.J., A. FEKKAK, N. ENNIH, E. ERRAMI, S.C. LOUGHLIN, P.G. GRESSE, L.P.
CHEVALLIER, J.-P. LIÉGEOIS (2004), A new lithostratigraphic framework for the
Anti-Atlas Orogen, Morocco, Journal of African Earth Sciences, 39, pp. 217-226.
XIAO H., J. SUPPE (1992), Origin of Rollover, AAPG Bulletin, V. 76, N°. 4, pp.
509-529.
Cartes géologiques
Sites Internet
[1] http://www.geomag.bgs.ac.uk/mercator.html
[2] https://zulu.ssc.nasa.gov/mrsid/
[3] http://edc.usgs.gov/products/elevation/gtopo30/gtopo30.html
Figures Page
____________________
Tableaux Page
Graphiques Page
REMERCIEMENTS
Mes remerciements posthumes à Martin Burkhard qui m’a toujours fait confiance
et encouragé dans toutes mes entreprises. A celui qui m’a inculqué la géologie
structurale, la cartographie, la géologie des Alpes. A celui qui m’a rendu attentif
à l’importance de la géologie dans notre société, aux problèmes
environnementaux et politiques, à la vulgarisation. A celui qui m’a permis de
découvrir un bout du monde.
A l’Institut de Géologie pour les belles années passées. Une thèse étant
également le fruit de tout un environnement. :
Aux étudiants avec qui j’ai partagé TP et excursions : Aline, Pauline, Lucien,
Laurent, Réza, Patrick, Samuel, Tomaso, Roger, Robin, Esther, Laureline, Jean-
Luc, Johann, Jonathan (2x), Vincent (3x), Baptiste, Claude, Gaël, Naomi, Marina,
Mickael, Amstutz, Cédric, Géraldine (2x), Martine, Ivan, Pascal, Pascale, Affoltre,
Christophe, Anne, Nicolas, Ekim, Tiziano, Yannick, Bérénice, Sophie, Laurent,
Rodrigo, Paul, Samuel, Raphaël, Guillaume, Clara-Marine, Clémentine,
Cristobal, Julien, Philipp ... et j’en oublie.
A tous les Marocains, amis de longue date ou d’un instant, pour leur chaleur et
leur hospitalité. Hassane, Abdulla, Redoine, Tarik, Smaïl, Mohammed, Ahmed,
Mansour, Mehdi, Somia, Kaoutar, Isham, Amine, Adnane, Abdulwahed, Aziz, ...
A mes parents
A Rachel
A
DONNEES BRUTES
DE LA FIGURE 22
PETITS PLIS
Stéréogrammes jaunes
Plan de stratification :
Données : Azimut (000-360°) et Plongement (00-90°) de la ligne de plus grande
pente.
Représentation sur les stéréogrammes : Pôle du plan, hémisphère inférieur.
CHEVAUCHEMENTS
Stéréogrammes bleus
Plans de chevauchements :
Données : Azimut et plongement de la ligne de plus grande pente.
Représentation sur stéréogrammes : Grand cercle, hémisphère inférieur.
Stries :
Données : Azimut et plongement de la direction de la strie.
Représentation sur stéréogramme : Direction de la strie, hémisphère inférieur.
Plans de chevauchement
100 32 227 64 064 35 095 30
182 26 056 85 138 45 231 40
078 61 159 30 126 15
Stries
078 28 230 41 082 29 267 36
201 23 074 02 068 34
063 60 332 64 077 13
Plans de chevauchement
074 36 233 40 278 68 050 59
068 33 047 71 051 54
152 35 226 54 273 52
Stries
035 27 207 24 278 68 032 57
032 29 044 69 230 47
215 39 070 51 295 20
Plans de chevauchement
133 27 114 17 101 60 261 18
113 29 123 12 132 33 332 33
Stries
135 27 137 15 116 59 321 08
152 26 140 12 144 33 317 32
Plans de chevauchement
222 25 001 52 018 70 192 53
176 34 182 44 183 50 011 25
Stries
206 35 022 48 038 68 198 49
182 39 200 40 218 42 002 23
Plans de chevauchement
119 65 005 45 321 46 331 40
046 20 332 72 105 23 352 79
343 45 338 59 160 10 325 50
002 31 341 21 339 11 302 24
170 31
Stries
165 61 355 44 002 38 021 28
326 03 279 65 159 14 020 77
299 36 016 55 162 10 296 48
331 28 328 20 337 11 291 23
164 30
Plans de chevauchement
353 19 092 24 339 30 004 35
341 29 013 27 348 21 017 23
348 29 321 13 295 23 327 19
010 49 068 30 349 47 256 20
Stries
342 17 157 09 350 29 330 28
344 28 333 21 346 21 326 15
334 27 334 12 334 21 321 18
338 45 165 01 329 43 324 09
PENDAGES GENERAUX
Stéréogrammes verts
Plans de stratification :
Données : Azimut et plongement de la ligne de plus grande pente.
Représentation sur stéréogrammes : Pôle du plan, hémisphère inférieur.
STRUCTURES EXTENSIVES
Stéréogrammes rouges
Failles :
Données : Azimut et plongement de la ligne de plus grande pente.
Représentation sur stéréogrammes : Grand cercle, hémisphère inférieur.
Stries :
Données : Azimut et plongement de la direction de la strie.
Représentation sur stéréogramme : Direction de la strie, hémisphère inférieur.
Failles
339 40 333 39 328 45 322 41 328 47
Stries
334 40 054 09 334 44 338 40 049 03
Failles
190 53 161 38 137 89 138 55 164 64
Stries
176 71 141 35 225 37 196 51 220 37
B
ARTICLES
Abstract
The structural style of the Anti-Atlas of Morocco is determined by two key parameters: the total
thickness of Paleozoic cover series and the relative abundance of shale vs. competent marker beds. A
late Proterozoic / earliest Cambrian rifting event produces a « Horst and Graben » configuration with
clastic syn-rift deposits of the PII-III (Piqué et al., 1999, Piqué, 2003, Soulaimani et al., 2001). Post-rift
sedimentation during most of the Paleozoic is dominated by shallow marine shale and marl deposits,
alternating with periods of sandstone (Ordovician) and limestone (Devonian, lower Carboniferous)
deposition. Clastic foreland basin sedimentation sets in during the middle Carboniferous and is
terminated by a first tectonic compressional event in late Carboniferous times. This leads to strongly
disharmonic small scale folding and minor thrust faults, by the intervention of diffuse detachments
within thick shale horizons, in a trishear mode (Erslev, 1991, Bump, 2003). In the eastern Anti-Atlas, a
regional gradient in deformation style, intensity and orientation is observed from north to south. A thick
skinned inversion style in the north, is gradually changing southward into a thin skinned detachment
folding style. An egg-box interference pattern is observed inbetween, interpreted to be linked to the
shape of the West-African Craton. A Late Triassic NNW-SSE extension in relation with the onset of the
Atlantic opening, reactivates the strongly compartimentalized, partly inverted basement blocks. Within
the Paleozoic sediments this event leads to extensional fault-related folding above complex normal
faults with stair case geometries. Upper Cretaceous sediments (Hamada du Guir) seal a peneplained
Paleozoic landsurface. The topography of the eastern Anti-Atlas is only very slightly affected by
Cenozoic to Quaternary inversion of the High-Atlas (Frizon de Lamotte et al., 2000). The present height
of the Anti-Atlas chain in general is interpreted as due to large scale thermal uplift from middle Miocene
onward (Saddiqi et al., 2005, Teixell et al., 2005).
Keywords: Anti-Atlas; Inversion tectonics; Thick vs thin skinned tectonics; Paleozoic; Variscan
orogeny; Extensional fault-related folding.
1
1. Introduction
2
Fig. 2. Tafilalt area, eastern end of Anti-Atlas belt, viewed from Landsat 7 (NASA, 2005), false colours ( downloaded from :
https://zulu.ssc.nasa.gov/mrsid/ ). Image description: The Tafilalt Flood plain (shown by the green Palm Tree culture around
Rissani and Erfoud), the Oued Ziz and Oued Rheris (oued : sporadic river) are flooding from the High-Atlas and get dry to
the South, in the center of the view the Erg Chebbi sand dunes. In light green, the Neogene and Cretaceous Hamada seals the
Anti-Atlas which appears as a half-window. The folded and faulted Paleozoic cover is underlined by the competent layers
such as the Devonian Rich in blue, the Ordovician Bani in reddish-orange and the Cambrian sandstones in dark brown. The
structural basement does not crop out in the Tafilalt region. Except some tiny outcrops in the north-east edge of half-window,
it appears on the left side of the view, at latitude of Erfoud, in dark green and belongs to the Ougnat inlier.
3
It was last consolidated during the Panafrican from the stratigraphy of the Tafilalt. Above some
orogeny. Exposed further west in the Bou Azzer thin basal conglomerates, the first competent layers
inlier, the Major Anti-Atlas Suture or Anti- to appear are the Cambrian Green Sandstones.
Atlas Major Fault (AAMF) (named “Accident The Ordovician 1st Bani is very thin compared
Majeur”) was interpreted as the old border of the to the Zagora region and the major Ordovician
african shield (Choubert and Faure-Muret, 1971). cuestas, creating the topographic heights in the
Recent workers, however, considered it as part Tafilalt, can be linked to the 2nd Bani, mostly made
of the “metacratonic” evolution and the northern of quartzites and sandstones. The Silurian shales
boundary of the WAC is now defined further are very weak and outcrops are exceedingly rare.
north, along the South Atlas Fault (Hefferan et al., They are responsible for large valleys inbetween
2000, Ennih and Liégeois, 2001, 2003, Liégeois the two cuestas of the 2nd Bani and the Devonian
et al., 2005), coincident with a major Neogene Richs. However, a single limestone marker bed in
tectonic feature, the southern deformation front of the middle of the Silurian shales is well known
the High-Atlas chain. At the end of Proterozoic, for its echinoderms and some graptolithes at its
a major unconformity is interpreted as the onset base. This bed can be seen as the first member of
of a new rifting cycle (of the Paleozoic era). The a carbonate series that follows in the Devonian,
successive openings of Rheic and Paleotethys in contrast to the first half of the Paleozoic
oceans (Stampfli and Borel, 2002) along the NW series which is mainly detrial. The Devonian
border of the WAC must have left their imprint in series is dominated by fossiliferous blueish dark
the area of the future Anti-Atlas belt. Rifting in carbonates. To the delight of the Filallis (Tafilalt
this area was aborted early, but it is structurally inhabitants), slabs and handicrafts from this area
very important since lead to the preconfiguration are exported worldwide. The Upper Devonian is
of blocks which were inverted during the Variscan marked by a return to detrital sedimentation that
orogeny. Syn-rift sediments of the P II-III unit dominates throughout the Carboniferous where
are mainly volcaniclastics, but the thickness of marker beds are fluvial and/or deltaic sandstones
this unit is very thin in the Tafilalt compared in contrast to the carbonates found further west.
to the western Anti-Atlas. Traditionally, from a The structural style of the cover is dictated by two
structural point of view, the basement englobes key-parameters, namely: the total thickness and
the crystalline basement and the P II-III unit. the incompetent / competent ratio.
Paleozoic cover 2.3. Variscan orogeny
Throughout the Paleozoic, the Anti-Atlas “basin” The closure of Paleotethys leads to the collision
is dominated by shallow marine conditions. In the of Gondwana and Laurussia to form the
western part the series reaches 12 km thickness, supercontinent Pangea. The former normal
while in the Tafilalt it is only about 6 km thick. faults are reactivated and inverted. Because of
The major part of the Paleozoic sediments are its particular incompetent rheology, the Paleozoic
made of fine grained detrial shales, siltstones cover folds. For this reason, the deformation is
and mudstones (Fig. 3). Competent marker beds characterised by a gradient rather than by a major
are subordinate in abundance and thickness, and thrust-front. The timing of the inversion in the
distributed among the series. These relatively Anti-Atlas is hard to constrain since the oldest
thin competent beds appear morphologically as formation sealing the deformed Carboniferous
cuestas and some of these crests can be followed sediments is of Upper Cretaceous age. Dolerites
throughout the Anti-Atlas range. The local name dykes and sills postdate the Variscan folding and
of each crest has given the name to the formations occur between 206 and 195 Ma (Sebai et al.,
(Ordovician Jebel Bani, Devonian Jebel Rich, 1991) . However deformation can be estimated
Carboniferous Jebel Ouarkziz). These formation to late Carboniferous between 320 and 300 Ma
names are not formally accepted lithostratigraphic according to the new paleoreconstructions of
terms but they are very useful, widely known and Stampfli and Borel (2002).
applied by all the workers in the region.
In the studied area, the whole Adoudounian series 2.4. Post Variscan
(local formation name for Cambrian from Basal The next Wilson cycle begins with tholeitic
Conglomerates to Upper Limestones) is lacking magmatism at the end of Triassic, which is the
4
Fig. 3. Schematic lithostratigraphic logs for the Tafilalt area and the whole Anti-Atlas (modified from Soulaimani, 1997 and
augmented from Tafilalt geologic map, Fetah et al., 1986). The whole series (to the end of Paleozoic) reaches 12 km in the
western Anti-Atlas but is decreasing eastward to about 6 km in the Tafilalt area. Some series are reduced but others, like the
infra-Cambrian Adoudounian (from Basal Conglomerates to Upper Limestones) and the Early Cambrian, are lacking. Note
that the real incompetent / competent layers ratio is more important than the sketch suggests.
precursory event to the opening of the Central Jurassic sediments are found in the Anti-Atlas.
Atlantic ocean. The new “Mesozoic High-Atlas The structural basement of the High-Atlas is of
basin” opens just north of eastern Anti-Atlas. Paleozoic age. It is mostly covered by a thick
This extension affects the Anti-Atlas by the deformed Mesozoic series but must include
intrusions of dolerite dykes and sills and by Variscan deformations. So the northern part of
normal faulting. This normal faulting corresponds the eastern Anti-Atlas is hidden under the High-
to a third re-activation of faults inherited from Atlas.
the late Proterozoic (Piqué and Laville, 1996) The actual high topography of the Anti-Atlas (up to
(post-tectonic intrusion of dykes between 206 2500 m) does not result from the Variscan orogeny
and 195 Ma (Sebai et al., 1991)). The Anti- alone. The Cretaceous and Neogene hamada east
Atlas Major Fault is essentially Precambrian, and south of Tafilalt is clearly tilted to the South
but must be reactivated in the Mesozoic since and in the north of Tafilalt the Cretaceous hamada
it produces the graben from Zagora to Taouz in is tilted to the North-East (see Fig. 2, on the DEM
the Paleozoic sedimentary cover. No Permian to
5
and the river patterns on the Landsat view). The to competent horizons within the stratigraphy.
scarp retreat analysis of Schmidt (1992) indicates Below and above a folded or thrust competent
a beginning of uplift in late Eocene times with an layer, the next immediate competent layers may
increase in late Miocene / early Pliocene times. well be unaffected. In terms of structural style,
These geomorphological interpretations fit nicely folding and thrusting occurs within the competent
with structural observations in the High-Atlas and bed which may be folded in one place and thrust
with an analysis of the deformation of Neogene in an other locality. Folds in competent layers are
sediments (Görler et al., 1988). The deformations parallel class 1B folds and class 3 for the weak
of the High-Atlas during Eocene – Oligocene and interbeds, resulting in a characteristic chevron
mainly Pliocene – Pleistocene epochs (Frizon de fold pattern (Ramsay and Huber, 1987). Bedding
Lamotte et al., 2000, Ellouz et al., 2003) just a parallel slickenside surfaces can be found in
few kilometres north, affects also the adjacent folded competent layers indicating flexural slip.
strongly compartmentalized Anti-Atlas basement. Thrust faults are very localised both laterally
This rejuvenated relief could come partly from and vertically within the stratigraphy. The
a Neogene thermal uplift (Anguita and Hernan, identification of contractional faults is limited to
2000, Saddiqi et al., 2005, Teixell et al., 2005). thick competent layers; but wherever observed,
thrusts seem to be rapidly attenuated within
3. Field data adjacent incompetent layers. In the right picture
of Fig. 4, the competent layers are seen to be
3.1. Small scale structures duplicate by a thrust while the weaker beds
above (hammer) are just folded. This example
The dominant structures in the study area are
demonstrates that in absence of competent
apparently regular cuestas. On closer inspection,
marker beds, the layer parallel shortening is hard
however, internal imbrications and small scale
to constrain.
folds such as those shown in Fig. 4 are quite
common. These folds and thrusts are highly In contrast to the apparent ENE-WSW global
localised. The spacing is on the order of 500 trend of the Anti-Atlas (Fig. 1), small scale
to 1000 m, with virtually undeformed regular structures show a clear NW-SE trend as shown
intervals inbetween. Structural perturbations are on the stereogramms of Fig. 4. Note that the
not only laterally localised but seem to be confined apparent dip of the fold axes to the south-east is
Fig. 4. Small scale structures in the cliff of Jebel Amelane (in the North-East of fig. 5). Left picture: folds in the Middle
Devonian; Left stereogramm : poles of bedding planes in folds and folds axes for Jebel Amelane region; Right picture : thrust
fault and layer parallel shortening; Right stereogramm : great circles of thrust-planes and striae for Jebel Amelane region.
6
Fig. 5. Aerial mosaic view of Jebel Amelane / Jebel Mech Idrane (Service Topographique du Maroc), 10 km west of Rissani
(see Fig. 2 for location); Stereogramms of bedding planes and normal faults; Cross-section.
due to the fact that a large majority of the small- are typically some 5 km long, 2 km wide (half
scale structures has been observed on south- wave-length) and are gently plunging to either
verging cuestas (hanging wall syncline). Tilting eastward or westward. Their repetition on a north-
those limbs back to horizontal also restores the south axis is quite intriguing (especially west of
fold axes to a sub-horizontal attitude. This is a Rissani, Fig. 2). These synclinal structures are
strong argument for these small-scale structures well developed only within the northern half of
to be older then the hanging wall synclines. This Tafilalt, mainly because their surface expression
applies to the entire Tafilalt area, which has been is related to the underlying Devonian carbonate
affected by a NE-SW compression throughout. beds of the Rich (see the three northern red
This NE-SW compression direction contrasts squares of Fig. 2). In other words, the synclines
with the overall Anti-Atlas trend, but it fits nicely owe their existence to a particular mechanical
with the structural trend of the Ougarta chain to stratigraphy rather then a geographically distinct
the South-East. part of the Tafilalt basin. The same extension is
expressed in a different way. Ordovician and
3.2. Hanging wall synclines Cambrian sandstones and quartzites are too thick
The synclines shown in Fig.5 have often been and too competent for such gentle fold structures
mistaken as Variscan folds. Strike and dip to develop. Silurian and Carboniferous shales, in
measurements of these meso-scale structures do the other hand, are too incompetent. Hanging wall
not fit with the results obtained from the small- synclines appear discreetly within Carboniferous
scale folds and thrusts discussed above. Synclines sandstones (red square south-east of the sand
7
dunes on Fig. 2). In fact they are extensional Mesozoic sediments of the High-Atlas. The
fault-related folds. contribution of High-Atlas inversion and the
thermal uplift in the exaggeration of this general
A detailed study of the Jebel Amelane region
large scale Anti-Atlas trend is difficult to quantify
revealed a bimodal distribution of the bedding
(Helg et al., 2004). Secondly, the Tafilalt region is
plane orientations (see stereogramm of the
clearly affected by folding along Ougartian NW-
bedding poles, Fig. 5). This is interpreted to be
SE trend. This trend is materialized by the global
a result of kinked or “trisheared” layers on a
antiform axis linking the Ougnat inlier to Taouz,
normal fault as described in the northwestern
separating the synforms of Tafilalt and Mader,
Red Sea (Khalil and McClay, 2002). The fact
and disappearing under the Hamada (Fig. 2). This
that the synclinal axes are plunging to the east,
large scale orientation is identical with the one
is explained by the super-structure. The Jebel
measured in small scale structures found in the
Amelane is located on the eastern flank of a huge
entire Tafilalt (3.1.).
anticlinorium. The most important evidence for
extensional fault-related folding is stratigraphic, While these two orientations are clearly
however. Minor normal faults do appear but distinguished in terms of their geometry and
the major fault zones are mostly obscured by interference patterns, the relative chronology
a thick soft interval of Silurian shales. Major remains open to debate. In the Tafilalt, both
fault zones can be directly observed only where structural trends are of similar, weak overall
the slip places more competent layers side by shortening intensity (estimated to ca. 10%) and
side as for instance at the Jebel Bou Ifarherioun intuitively either structure could be first or second.
(red square, 15 km south of Rissani on Fig. 2). Further west, however clear evidence is found for
The dip of the faults seems to vary according a relative chronology between folds of the Anti-
to the competence, flat in incompetent, steep in Atlas trend preceding the Ougartian trend (Caritg
competent layers. The localisation and repetition et al., 2004). Our own, unpublished observations
of these faults is interpreted as due to pre-existing in the Tata area have since provided additional
faults in the basement. Hanging wall synclines strong evidence for such a relative chronology.
show flexural-slip movement along bedding On the other hand rather than separate these two
plane surfaces, which can be misleading since deformations in time, an alternative explanation
the relative shear sense of such layer parallel slip could be found in the shape of the WAC
is always indicting “thrusting” independent of boundary. The eastern Anti-Atlas is located along
the underlying mechanism of folding in either the northern edge of the WAC metacraton. It is
compression and/or extension. These extensional limited by the South Atlas Fault (ENE-WSW) to
fault-related folds postdate the small scale folds the North and by the Saoura gravimetric anomaly
by folding their axis. (NW-SE) (Ennih and Liégeois, 2001) to the
East. On the very large scale, the arc defined by
3.3. Superstructures
the Ougarta – Anti-Atlas system thus seems to
The eastern Anti-Atlas shows interference of be an inherited feature, mimicking the shape of
two superimposed superstructures. First, the the WAC. The late Paleozoic fold belts would
ENE-WSW anticlinorium trend of the entire have been formed along the boundaries of this
Anti-Atlas belt is materialized by the large scale craton, specifically along its metacratonic borders
shape of basement inliers (Fig. 1). The Hamada which were once again re-mobilized in inversion
half-window apparently sealed to by Upper tectonics during the assembly of Pangea.
Cretaceous sediments to the east, inherited this
ENE-WSW trend (Fig. 1 and 2). In fact the large 4. Seismic data
scale Anti-Atlas superstructure appears more like
a huge monocline. From NNW to SSE rocks get The ONAREP RS8 seismic reflexion line (Fig. 6)
increasingly younger, from basement inliers, crosses the Tafilalt with a north-south orientation
across the folded Paleozoic cover, gradually (see location on fig. 2). Accordingly, this line is
passing into the flat lying basin fill of the Tindouf perpendicular to extensional folding and oblique
basin, partly covered by younger sediments of the with respect to both superstructures : WSW-ENE
Hamada. The northern limit of this superstructure in the northern part and NW-SE in the southern
is defined by the thrust front of the inverted part. The interpretation of this profile has been
8
Fig. 6. Seismic line RS 8 (from ONAREP). Upper : blank; Middle : Surface from field work and geological map,
interpreted at depth; Lower : reconstructed profile before normal faulting.
9
tied with the known surface geology, including
dip measurements and inspired by our knowledge
of local tectonics.
The first 10 kilometres (southern part) display large
scale folds with NW-SE axes. We interpret these
folds to be detached along a major décollement at
the base of the Paleozoic sedimentary pile. Cover
shortening is compensated below the décollement
by inversion of former late Proterozoic normal
faults. The resulting folding structure fits well
with an Ougarta trend observed on the map scale.
The structural style remaining (northern) part of
the transect (from km 10 to 60) is dominated by
late normal faults. A series of pseudo-basins, e.g.
one at km 25, have previously been interpreted
as a syn-sedimentary “thrust top” basin. In their
interpretation, an important early Variscan folding
phase affected the region in Upper Devonian
through early Carboniferous times, obscured
by middle Carboniferous – post folding series.
Such an interpretation is in conflict with field
observations, and no positive evidence for a late
Devonian folding phase has ever been found
anywhere in the Anti-Atlas area. Our interpretation
of the “pseudo-basins” seen in the seismic line is
an extensional context. The geometry observed
is explained as a consequence of a particularly
weak cover and a seismic artefact. First, the
thinning of the footwall layers and the thickening
Fig. 7. Schematic illustration of the proposed kinematical
of the hanging wall layers is observed in many model (inspired from Bump, 2003). South-North
models specially for “ultraweak cover on rigid orientation perpendicular to inliers alignment. But parallel
basement” (Finch, 2004). The image obtained by to the Ougartian trend. a: Late Proterozoic rifting with syn-
discrete-element modelling is very similar to this rift sediments; b: Deposition of Paleozoic cover; c: Variscan
seismic line. Secondly, the seismic line shows the inversion. The thrust fault takes a shortcut in the basement,
the cover reacts in a trishear mode. The Ougartian trend do
Two Ways Time and not the depth. The velocity not appears on this sketch; d: New extension phase (Atlantic
of Carboniferous shales and mudstones must opening). Normal faults short cut the Paleozoic cover, its
be particularly slow (probably less than under dips vary according to the competence of rocks and produce
2000 m/s), while competent layers (Devonian hanging wall synclines.
limestones, quartzites) may attain 3000 m/s or
more. This relationship leads to a pronounced
velocity pull down, which tends to exaggerate the 5. Conclusions
depth of the pseudo basins.
The eastern Anti-Atlas structure is a complex
The third part of Figure 6 is the presumed profile, interference between two trends of Variscan
restored to illustrate the situation at the paroxysm folding which lead to an egg box pattern of
of the Variscan deformation, i.e. without the later basins and domes. This fold interference pattern
normal faults. The overall shortening is minimum is subsenquently cut by a series of normal faults
of 10 %, but this value is measured at regional most likely in response to the opening of the
scale and does not include the layer-parallel Atlantic in Liassic times. The ENE-WSW Anti-
shortening observed at smaller scale. Helg et Atlas global trend is the best developed in terms
al. (2004) argue that such distributed small of deformation and inversion. The Anti-Atlas is
scale deformation add a substantial amount of clearly thick skin. The Paleozoic cover reacts
shortening of similar magnitude.
10
in a “trishear way” (Fig. 7.c) to accommodate Ellouz, N., Patriat, M., Gaulier, J.-M., Bouatmani, R.,
basement shortening which is inverting former Sabounji, S., 2003. From rifting to Alpine inversion:
Mesozoic and Cenozoic subsidence history of some
late Proterozoic normal faults. The deformation Moroccan basins. Sedimentary Geology 156, 185-
gradient decreases strongly from NNW to SSE, 212.
with increasing distance from basement inliers. Ennih, N., Liégeois, J.-P., 2001. The Moroccan Anti-Atlas:
The NW-SE Ougarta trend is materialized the West African craton passive margin with limited
throughout the Tafilalt by small scale folds and Pan-African activity. Implications for the northern
limit of the craton. Precambrian Research 112, 289-
thrust faults. This deformation has a important 302.
thin skin component, detachment along weak Ennih, N., Liégeois, J.-P., 2003. Discussion, The Moroccan
layers; although the underlying cause is equally Anti-Atlas: the West African craton passive margin
thick skin basement inversion. The interference with limited Pan-African activity. Implications for the
between these two trends is thought to be due northern limit of the craton: reply to comments by E.H.
Bouougri. Precambrian Research 120, 185-189.
to the shape of the northern corner of the WAC, Erslev, E.A., 1991. Trishear fault-propagation folding.
and particularly a renewed re-mobilisation of Geology 19, 617-620.
its ”weak” meta-cratonic borders in the late Fetah, M., Bensaïd, M., Dahmani, M., 1986. Carte
Paleozoic times. Former normal and inverted géologique du Maroc 1:200ʼ000 Tafilalt-Taouz, Notes
faults are again reactivated during the Atlantic et Mémoires du Service Géologique du Maroc N°244.
Finch, E., Hardy, S., Gawthorpe, R., 2004. Discrete-element
opening event. This leads to the formation of modelling of extensional fault-propagation folding
new steep normal faults within the Paleozoic above rigid basement fault blocks. Basin research 16,
cover series. The overall weakness of the 489-506.
thick Paleozoic series favours the formation of Frizon de Lamotte, D., Saint Bezar, B., Bracène, R.,
extensional “drape” folding (Fig 7.d). The High- Mercier, E., 2000. The two main steps of the Atlas
building and geodynamics of western Mediterranean.
Atlas inversion and Neogene thermal uplift do Tectonics 19 (4), 740-761.
not significatively deform the eastern Anti-Atlas. Görler, K., Helmdach, F.-F., Gaemers, P., Heissig, K.,
Only a light doming is underlined by the radial tilt Hinsch, W., Mädler, K., Schwarzhans, W., Zucht, M.,
and retreat of cretaceous cuestas. 1988. The uplift of the central High Atlas as deduced
from Neogene continental sediments of the Ouarzazate
province, Morocco. In: V.H. Jacobshagen, Springer-
Acknowledgments Verlag (Ed.), The Atlas System of Morocco. Lecture
Notes Earth Sci. 15, 361-404.
Ongoing support by the “Ministère de lʼIndustrie Haddoum H., Guiraud, R., Moussine-Pouchkine, A., 2001.
et des Mines, Direction de la géologie” by M. Hercynian compressional deformations of the Ahnet-
Mouydir Basin, Algerian Saharan Platform: far-field
Dahmani and M. Haddane and the moroccan
stress effects of the Late Palaeozoic orogeny. Terra
national petroleum agency ONAREP is Nova 13, 220-226.
gratefully acknowledged. Financial support by Hefferan, K.P., Admou, H., Karson, J.A., Saquaque, A.,
Swiss National Science Foundation grants N° 2000. Anti-Atlas (Morocco) role in Neoproterozoic
21-52516.97 / 20-63790.00 and by the “Fonds Western Gondwana reconstruction. Precambrian
Research 103, 89-96.
Matthey-Dupraz” local fund for geology
Helg, U., Burkhard, M., Caritg, S., Robert-Charrue, C.,
students. 2004. Folding and inversion tectonics in the Anti-Atlas
of Morocco. Tectonics 23, TC4006, doi:10.1029/
References 2003TC001576.
Khalil, S.M., McClay, K.R., 2002. Extensional fault-related
Anguita, F., Hernán, F., 2000. The Canary Island origin: folding, northwestern Red sea, Egypt. Journal of
a unifying model. Journal of Volcanology and Structural Geology 24, 743-762.
Geothermal Research 103, 1-26. Liégeois, J.-P., Fekkak, A., Bruguier, O., Errami, E., Ennih,
Bump, A.P., 2003. Reactivation, trishear modeling, and N., 2005. Lower ediacaran age (630-610 Ma) for
folded basement in Laramide uplifts: implications the Sarhro group. Implications for the metacratonic
for the origins of intracontinental faults. GSA Today, evolution of the Anti-Atlas, correlations with the
4-10. Tuareg shield and the evolution of the pan-african
Caritg, S., Burkhard, M., Ducommun, R., Helg, U., Kopp, orogeny in the NW Africa. 4ème Colloque
L., Sue, C., 2004. Fold interference patterns in the Late International 3Ma, Agadir 2005, Abstract Volume, 2.
Palaeozoic Anti-Atlas belt of Morocco. Terra Nova 16, NASA and Earth Satellite Corporation, Landsat 7 images
27-37. download website (https://zulu.ssc.nasa.gov/mrsid/ ),
Choubert, G., Faure-Muret, A., 1971. Epoque hercynienne. last visited 14.03.2006.
In: Tectonique de lʼAfrique, Sciences de la Terre, ONAREP (Office National de Recherches et dʼExploitations
Paris, 6, UNESCO, Paris, France, 353-371. Pétrolières) Morocco. Seismic reflexion line RS 8.
11
Piqué, A., Laville, E., 1996. The central atlantic rifting : Sebai, A., Feraud, G., Bertrand, H., Hanes, J., 1991. 40Ar/
reactivation of Palaeozoic structures ?. J. Geodynamics 39
Ar dating and geochemistry of tholeiitic magmatism
21 (3), 235-255. related to the early opening of the central Atlantic rift.
Piqué, A., Bouabdelli, M., Soulaïmani, A., Youbi, Earth and Planetary Science Letters 104, 455-472.
N., Iliani, M., 1999. Les conglomerates du P III Service Topographique du Maroc, Aerial Photography, ST
(Néoprotérozoïque supérieur) de lʼAnti-Atlas (Sud du 1/40000, Mission 4-9-73, N° 16_10 / 16_11.
Maroc): molasses panafricaines, ou marqueurs dʼun Soulaïmani, A., Le Corre, C., Farazdaq, R., 1997.
rifting fini-protérozoïque ?. C. R. Acad. Sci. Paris, Déformation hercynienne et relation socle/couverture
Sciences de la Terre et des planètes 328, 409-414. dans le domaine du Bas-Drâa (Anti-Atlas occidental,
Piqué, A., 2003. Perspective, Evidence for an important Maroc). Journal of African Earth Sciences 24 (3), 271-
extensional event during the latest Proterozoic and 284.
Earliest Paleozoic in Morocco. C. R. Geoscience 335, Soulaïmani, A., Piqué, A., Bouabdelli, M., 2001. La série
865-868. du PII-III de lʼAnti-Atlas occidental (Sud marocain)
Ramsay, J.G., Huber, M.I., 1987. The techniques of modern : un olistostrome à la base de la couverture post-
structural geology, vol. 2: Folds and Fractures, panafricaine (PIII) du Protérozoïque supérieur. C.R.
Academic Press Ltd. London, pp. 462. Acad. Sci. Paris, Sciences de la Terre et des planètes
Saddiqi, O., Ziad, N., Missenard, Y., Frizon De Lamotte, 332, 121-127.
D., Goffe, B., Leturmy, P., El Mamoun, R., Sebti, S., Stampfli, G.M., Borel, G.D., 2002. A plate tectonic
Poupeau, G., 2005. Age Miocène moyen supérieur model for the Paleozoic and Mesozoic constrained
du soulèvement thermique du Haut-Atlas Central : by dynamic plate boundaries and restored synthetic
apport des traces de fission sur apatite. 4ème Colloque oceanic isochrones, Earth Planet. Sci. Lett. 196, 17-
International 3Ma, Agadir 2005, Abstract Volume, 33.
45. Teixell, A., Ayarza, P., Zeyen, H., Fernandez, M.,
Schmidt, K.-H., 1992. The tectonic history of the Pre- Arboleya, M.-L., 2005). Effects of mantle upwelling
Saharan depression (Morocco) – a geomorphological in a compressional setting: the Atlas Mountains of
interpretation. Geologische Rundschau 81 ( 1), 211- Morocco. Terra Nova 17, 456-461.
219.
12
Inversion tectonics in the eastern
Anti-Atlas of Morocco
Charles Robert-Charrue, Martin Burkhard*
Institut de Géologie et d’Hydrogéologie, Université de Neuchâtel,
11, rue Emile-Argand, CH-2009 Neuchâtel, Switzerland
*Corresponding author. Tel. : +41 32 718 26 52 ; fax : +41 32 7182601.
E-mail address : [email protected]
ACKNOWLEDGMENTS
ABSTRACT
Ten cross sections drawn from field mapping data, existing geological maps and landsat views
characterize the structure of the eastern Anti-Atlas fold belt with its ENE-WSW structural
grain. One of these cross sections (no. 08) is sequentially restored and used to perform depth
to detachment calculations. Late, i.e. Early Mesozoic extensional structures are rooted at a
mid-crustal depth of detachment of -18 to -20 km (-11 to -12 mi). We propose a novel method
of integrating bed internal, layer parallel shortening deformations into depth to detachment
calculations. Balancing late Variscan compressional structures of the eastern Anti-Atlas
requires a substantial amount of 11 to 17 % of layer-parallel shortening within the Paleozoic
cover in an early stage of the Variscan orogeny. The structural style of the Anti-Atlas is dictated
by the reactivation of pre-existing structures within the basement and a predominantly weak
mechanical stratigraphy of the overlying Paleozoic cover rocks. The variscan event is best
described as the inversion of an intracratonic basin, rather than a classical foreland fold-and-
thrust belt.
1
INTRODUCTION
3
sediments with a total thickness of 12 km middle Carboniferous as documented by
(7.5 mi) in the western Anti-Atlas decreasing the youngest sediments involved in folding
to less than 8 km (5 mi) in the study area, i.e. and tilting along the northern border of the
the Mader and Tafilalt. The Anti-Atlas is best Tindouf basin. Anti-Atlas folds are cross-cut
described as an intracratonic basin (Burkhard by the widespread Late Triassic (200 Ma)
et al., 2006); it can be regarded as the intrusions of dykes and sills (Sebai et al.,
northern, tectonized part of the much larger 1991), in relation with Atlantic opening. The
Tindouf basin (Coward and Ries, 2003). peneplained Anti-Atlas folds are sealed only
The basin fill is dominated by an abundance by a generalized marine transgression in
of shales and silts deposited in a shallow Upper Cretaceous times, however. Not far
marine environment (see the stratigraphic log to the north of the Anti-Atlas, rifting in Late
on Figure 3). Competent and morphologically Triassic – Early Jurassic times lead to the
outstanding marker beds are sandstones and opening of the High Atlas trough, filled with
quartzites of the Cambrian and Ordovician. thick series of Triassic to Jurassic sediments
Most of the Silurian is represented by soft (Manspeizer et al., 1978). This rifting event
black shales, very rarely seen in outcrop. lead to some extensional reactivation of
Carbonate plateforms dominated the Anti- faults within the eastern Anti-Atlas. The major
Atlas sea in Lower Cambrian and again event of Cenozoic inversion responsible for
in Devonian times. The return of detrital uplift and folding of the High Atlas chain did
conditions in Upper Devonian times could not lead to any noticeable folding and/or
be interpreted as the onset of convergence faulting within the Anti-Atlas domain. This is
in relation to Variscan tectonics. The documented by a smooth configuration of
Carboniferous is mainly made of shales and the basal Upper Cretaceous unconformity.
rare stiff beds are pericontinental deltaic The Anti-Atlas chain is standing high,
sandstones with an abundance of flute casts however, and it has long been speculated
figures and channel fills. that this renewed uplift and relief could be in
response to alpine shortening deformation at
Variscan orogeny – inversion a crustal scale. Alternatively, however, high
Most if not all the Paleozoic basins of Northern present day topography in both the Anti- and
Africa are involved to some degree in Variscan High Atlas have recently been interpreted as
tectonics in Carboniferous to Permian times being due to a large scale thermal anomaly
(Coward and Ries, 2003). The Anti-Atlas fold located within the upper mantle (Missenard
belt has generally be considered as a fold- et al., 2006).
and-thrust belt developed in the hinterland
of the larger Appalachian – Variscan orogen. RHEOLOGY
It is indeed time equivalent to and bears
some structural resemblance with the The basement is a complex assembly
Alleghenian Valley and Ridge province (Helg of igneous, metamorphic, volcanic and
et al., 2004). The direct link between internal, sedimentary rocks. Despite these differences
metamorphic zones of the Variscan chain in composition and despite the existence of
cropping out north of the South Atlas Fault internal layering and strong anisotropy, in
and the Anti-Atlas belt remains obscure, terms of rheology, basement blocks behave
however. No basal detachment has been as overall rigid units, and display very little
observed anywhere within the Paleozoic internal deformation. This is quite obvious in
cover series of the entire Anti-Atlas. The satellite images such as the Landsat scene
thick skinned involvement of the basement at shown in Figure 2. Normal faults of the Late
very short distances behind the deformation Proterozoic - Middle Cambrian rifting event
front have been used as arguments in favour and possibly older structures, predefine the
of a “basin inversion-” rather than a “frontal structural grain of the variscan inversion.
fold-and-thrust belt-“ deformation style The Paleozoic cover series are quite weak
of the Anti-Atlas (Burkhard et al., 2006). and essentially behave as one soft layer
Anti-Atlas basin inversion is younger than draped over the basement blocks. Details
4
of the cover rheology are shown in Figure 3.
The two major dominant members (labelled
DM in Figure 3) are the basement and the
1st Bani formation (Ordovician sandstones
and quartzites). Remaining stiff members
elsewhere in the stratigraphic column behave
simply as “more dominant members within a
conforming structural lithic unit” (according to
the terminology of Currie et al., 1962). Our
rheology log is based on field observations
and cross-sections, which document the
overall “conforming” geometry, matching
the two major dominant members. On the
smaller scale, however, dominant members
of conforming structural lithic units also
display their own characteristic layer parallel
shortening structures such as wedging
and minor buckle folding (Figure 4), clear
evidence for an important component of bed-
internal layer parallel shortening (LPS).
Figure 5. Basement – cover contact at the southern limit of Saghro inlier (cross section 08), view to North-East
(houses for scale).
Compressional structures
pictures A and B (Figure 6), taken from the
On the outcrop scale, the dominant members same layers of Devonian carbonate illustrate
show isolated meso-scale structures such as two characteristic layer parallel shortening
folds and minor wedging “fishtail structures” features, namely buckle folding and wedging
(Figure 6). The controlling factors are the (fishtail). In the fold case A, the thickness
bed thickness, the abundance of weak layers involves more layers then in case B where
and the relative thickness of the two. The the thrust is limited to one single competent
6
Figure 6. Compressional structures. (A) Fold in Middle Devonian limestones, cross section 03 (hammer in the
centre for scale). (B) Small thrust in Middle Devonian, cross section 03 (hammer for scale). (C).: View to the East,
north-vergent “back” thrusting fault with drag fold in footwall in 1st Bani, on the southern flank of cross section 10
(vertical view is 20 m).
7
bed. The bigger thrust on picture C cuts off cross sections, their role in balancing will be
the whole lower part of the 1st Bani dominant considered however.
member. Made of sandstones and quartzites Bed internal deformations within the thick
this series has very few weak interlayers, and weak zones are more subtle in appearance
reacts as a single layer. The soft layers permit (Figure 7). When looking closer at shales in
the development of a drag fold at the base of good outcrop conditions, they are invariably
the cliff whereas the thick stiffer beds at the affected by a multitude of small thrusts,
top don’t. All these layer parallel shortening generally at a sharp angle with respect to
deformations are too small to appear in the bedding.
Figure 8. Extensional structures. (A) Normal fault. View to the East. The Middle Cambrian sandstones (left)
dipping gently to the south abut basement on the right hand side; the normal fault is dipping steeply to the North;
Ougnate inlier; cross section 06; Landrover lower right for scale. (B) View to the South. Graben structure in the
middle of cross section 08. The cliff in the background is the 1st Bani (Ordovician quartzites) dipping gently to the
South; the normal fault faces the viewer plunging steeply northward, drag folded layers of Devonian carbonates
in the foreground dip very steeply to the North; for the scale : the trail in the foreground is 100 m (328 ft) long.
(C) Drag folds along a normal fault dipping northward (fault plane: 024/68; striae : 043/65); Devonian carbonates
within the oued shown in the centre of picture B; view to the West; stick on the right is one meter long.
8
Overall, compressional layer parallel strike such as the graben in the middle of
shortening structures are best developed in cross sections 07 and 08. Devonian folds
the north and their abundance and intensity of cross section 07 are interpreted as forced
decreases progressively southward. This is “drape” folds resulting from extension in the
interpreted as a regional gradient inbetween underlying graben, but attenuated within the
the supposed internal zones of the Anti-Atlas soft Silurian shales accommodating the fault
orogen to the north and the stable continental throw in a trishear style.
interior to the south. Internal parts of the Anti- The fault-bend fold in the northern part of
Atlas orogen are obscured by, and may have cross sections 06 and 07 is an interpretation.
been substantially shifted in position during, It has been drawn such as to balance
Mesozoic and Cenozoic events of High Atlas the extremely deformed Carboniferous
formation, however. The spatial relationships sediments. However, the thrust in cross
between the Variscan Anti-Atlas belt and section 08 has been observed, but its
the basement blocks found within and to interpretation remains ill constrained at
the north of the High Atlas are a matter of depth. The High-Atlas structures are redrawn
ongoing debate (Hoepffner et al., 2005). according to Frizon de Lamotte et al.(2000).
Extensional structures For calculations of the depth to detachment,
the cross section 08 has been chosen
The eastern Anti-Atlas of the Tafilalt and
because it crops out at the deepest level.
Mader basins is affected by an important
Therefore, interpretation errors are minimal
“late” event of extension. This is in stark
for the geometry of the basement-cover
contrast to the western Anti-Atlas where
contact. For the reconstruction at different
no such extensional structures have been
stages, we first restore the latest normal
identified. Early authors (Choubert et al.,
faults in order to establish the situation at the
1952) depicted the general structure of the
end of the variscan paroxysm. In a second
Tafilalt area as a set of tilted blocks inbetween
step, the variscan folding is restored to a flat
north dipping normal faults. Normal faults
lying configuration. At this stage the layers still
are quite steep and cut mostly straight
contain a component of early Variscan layer-
through the entire Paleozoic cover (pictures
parallel shortening which is compensated by
in Figure 8). Normal faults are responsible
stretching back to a restored configuration at
for a partial collapse of the basement inlier
the end of Paleozoic sedimentation.
antiforms. Locally, extensional fault-related
folds (Withjack et al., 1990; Schlische, 1995;
Finch et al., 2004) and roll over folds (Xiao DEPTH TO DETACHMENT
and Suppe, 1992) developed within weak
Paleozoic cover rocks of the hanging-wall. Depth to detachment is estimated using
Late normal faults are systematically rooted the displaced-area method (Epard and
within the old Late Proterozoic normal Groshong, 1993; Groshong, 1994, 1999).
faults, they were reactivated across the This 2-D cross-section balancing method
compressional footwall shortcut thrust of the (Figures 10 and 11) is based on the uplifted
variscan event. area above regional caused by compression
(excess area Sc) or a downdropped area
below regional caused by an extension (lost
CROSS SECTIONS area Sx). Excess or lost area must be equal
to the displaced area, which is defined by
Cross sections have been constructed at
the depth to detachment times horizontal
the 1:50’000 scale, not allowing to take
displacement. Excess or lost area is
mesoscale deformation structures into
measured above or below a regional horizon
account (Figure 9). Interpretations below and
(Hr in Figures 10 and 11). Such simple area
above topography have been made by simple
balancing methods do neither require any
projection (no change in fault orientation and
detailed information about the shape of the
constant layer thickness). Some features
detachment nor about the kinematical model.
have also been laterally projected along
Bed-internal strains are classically assumed
9
Figure 9. Cross sections.
10
11
Figure 10. Extensional deformation. Definition of
geometric elements used in the calculation of depth
to detachment.
to be zero, but in the following we will develop features (stages C to B in Figure 12). The
a modified displacement area method in differential uplift (Figure 12.A) does not have
order to evaluate the role of layer parallel any influence on the lost area in this stage
shortening. In this study, the excess and lost because this uplift is identical in both profiles.
areas are measured from restored sections However, for the next step backward, the
constructed for different stages of the Anti- differential uplift excess area (Su) must be
Atlas’ evolution through time (Figure 12). In subtracted from the excess area produced
the following discussion, we will concentrate by compressional deformation (stage C)
on cross-section number 08, one of the best because the pre-compression profile D does
constrained and representative sections not include the recent uplift.
across the eastern Anti-Atlas chain. The
present day cross-section has last been Depth to detachment on extensional
affected by Neogene tilting and differential structures
uplift, without any associated faulting. The The depth to detachment calculation for
excess area resulting from this uplift (Su in extension structures is made under the
Figure 12.A) has been removed from section classical assumption of constant bed length.
08 according to the lithospheric model of The area Sx below the regional Hr (Figure
Missenard et al. (2006), which is based on 10) resulting from extensional deformation
gravity data. is compensated by an equal area defined
The youngest faulting event in Triassic times by the horizontal displacement Dx times hx
caused the formation of the so-called Zagora the depth to detachment below the regional
graben as well as an extensional reactivation horizon Hr:
of crustal scale faults throughout the eastern
Anti-Atlas anticlinorium. (1) thus
(3)
14
Table 2. Depth to detachment
on compressional structures
(4) thus
Where the length Dce is the absolute layer-parallel shortening and Lo is the original bed length
(see Figure 11).
The measurable length Lc is defined as the original bed length minus shortening :
(5)
or
(7)
Further, Dc is replaced by equation (2), Dce by equation (4) and Lo by equation (6). This
yields :
(7’)
We examined the influence of layer-parallel to detachment Hce are given in Table 3 and
shortening on the depth to detachment by illustrated in Graphic 2.
varying e from 0 to 0.25, in a range compatible This exercise shows that depth to detachment
with observations made in outcrop and on the calculations applied to weakly folded section
thin section scale. Resulting values of depth such as the Anti-Atlas cross-sections
15
are strongly dependent on layer-parallel to -20 km (-11 to -12 mi). This depth range
shortening. With an assumed moderate is the same as the one initially determined
amount of layer-parallel shortening on the for late extensional structures. Taking into
order of 0.11 to 0.17, the calculated depth account the observed deformation gradient,
to detachment is shifted upward from more the layer-parallel shortening must be higher
than -100 km to a reasonable depth of -18 in the north and decreases southward.
16
CONCLUSIONS folding and widely distributed layer-parallel
shortening of up to 17 %. Extensional re-
The structural style of this inverted activation in Early Mesozoic times (stage C)
intracratonic basin is controlled by two lead to the formation of new set normal faults
properties: within cover rocks, crosscutting the variscan
1. The weak rheology of the cover. This compressional cover structures at a high
allowed for a distributed shortening angle.
deformation within shales and siltstones The depth to detachment calculation and its
and the formation of isolated mesoscale error analysis provide significant informations.
structures within competent layers. The On the extension structures, calculations
weakness of the cover series prevented the yield a detachment depth in the range of -18
development of any strong localization major to -20 km (-11 to -12 mi). Positive deviation
décollements. This soft pile forms a single from this value indicates that the dip of the
structural-lithic unit which adopts a drape fold normal faults must decrease with depth.
allure with respect to the basement structures. Alternatively, a minimal amount of 3% bed
2. Pre-existing inherited structures within the internal stretching could equally well explain
basement. Precambrian rift normal faults the discrepancies in depth to detachment
constitute weak zones (Figure 13, stage obtained from the overall cross section.
A) which are repeatedly reactivated in later Balancing Variscan shortening deformations
events be it compressional or extensional. yields unreasonably high values for depth to
The variscan shortening lead to the inversion detachment in excess of -100 km . Taking a
of Precambrian normal faults. Thrusts faults component of 11 to 17 % distributed layer-
systematically shortcut the footwalls to flatten parallel shortening into account allows to
out within Lower Cambrian shales and never reconcile surface observations and balancing
step up any higher into the Paleozoic cover with a reasonable depth to detachment of -18
(Figure 13, stage B). Shortening in the cover to -20 km.
series is accommodated by an overall drape
Figure 13. Mid-crustal detachment model for cross section 08. No vertical exaggeration. Cover from Cambrian to
Silurian (included) (A) Profile before variscan compression with the rift inherited normal faults. (B) Profile at the
paroxysm of variscan compression. The normal faults play in inversion, but do not reach the Paleozoic cover by
shortcutting the normal faults where they are to steep. (C) Opening of the nearby High-Atlas basin affects Anti-
Atlas. Former faults are reactivated in extension and crosscut the Paleozoic cover. (D) At scale recent uplift effect
from Missenard et al. (2006).
17
REFERENCES CITED
Groshong, R. H., 1994, Area balance, depth to
Benssaou, M., N. Hamoumi, 2003, The lower- detachment, and strain in extension: Tectonics, v. 13,
Cambrian western Anti-Atlasic graben: tectonic control no. 6, pp. 1488-1497.
of palaeogeography and sequential organisation: Groshong, R. H., 1999, 3-D structural geology:
Comptes Rendus Geoscience, 335, p. 297-305. Springer-Verlag, 324 p.
Boudda, A., G. Choubert, 1972, Sur la limite inférieure Helg, U., M. Burkhard, S. Caritg, C. Robert-Charrue,
du Cambrien au Maroc: C. R. de l’Académie des 2004, Folding and inversion tectonics in the Anti-Atlas
Sciences, Paris, Serie II 275 (1), p. 5-8. of Morocco: Tectonics, v. 23, TC4006, doi:10.1029/
Bump, A.P., 2003, Reactivation, trishear modeling, 2003TC001576.
and folded basement in Laramide uplifts: implications Hoepffner, C., A. Soulaimani, A. Piqué, 2005, The
for the origins of intracontinental faults: Geological Moroccan Hercynides: Journal of African Earth
Society of America Today, p. 4-10. Sciences, 43, pp. 144-165.
Burkhard, M., S. Caritg, U. Helg, C. Robert-Charrue, Johnson, K. M., A. M. Johnson, 2002, Mechanical
A. Soulaïmani, 2006, Tectonics of the Anti-Atlas of models of trishear-like folds: Journal of Structural
Morocco: Comptes Rendus Geoscience, 338, p. 11- Geology, 24, p. 277-287.
24. Liégeois, J.-P., A. Fekkak, O. Bruguier, E. Errami, N.
Choubert G., L. Clariond, J. Hindermeyer, 1952, Ennih, 2005, Lower ediacaran age (630-610 Ma) for
Livret-guide de l’excursion C36, Anti-Atlas central the Sarhro group. Implications for the metacratonic
et oriental: Congrès géologique international, XIXe evolution of the Anti-Atlas, correlations with the Tuareg
session-Alger-1952, Série : Maroc N° 11, 89 p. shield and the evolution of the pan-african orogeny in
Coward, M.P., A.C. Ries, 2003, Tectonic development the NW Africa (abs.): 4ème Colloque International
of North African basins : Petroleum geology of Africa; 3Ma, Agadir 2005, Abstract Volume, p. 2.
new themes and developing technologies. T. Arthur, Manspeizer, W., J.H. Puffer, H.L. Cousminer, 1978,
S. MacGregor Duncan and N.R. Cameron, Geological Separation of Morocco and eastern North America:
Society of London. London, United Kingdom, 207, p. A Triassic-Liassic stratigraphic record: Geological
61-83. Society of America Bulletin, v. 89, pp. 901-920.
Currie, J. B., H. W. Patnode, R. P. Trump, 1962, Missenard, Y., H. Zeyen, D. Frizon de Lamotte, P.
Development of folds in sedimentary strata: Geological Leturmy, C. Petit, M. Sérbier, O. Saddiqi, 2006,
Society of America Bulletin, v. 73, p. 655-673. Crustal versus asthenospheric origin of relief of the
Ennih, N., J.-P. Liégeois, 2001, The Moroccan Anti- Atlas Mountains of Morocco: Journal of Geophysical
Atlas: the West African craton passive margin with Research, v. 111, B03410, 13 p.
limited Pan-African activity. Implications for the NASA and Earth Satellite Corporation, < https://
northern limit of the craton: Precambrian Research, zulu.ssc.nasa.gov/mrsid/ >
112, p. 289-302. Accessed June 30, 2006.
Epard, J.-L., R. H. Groshong, 1993, Excess area and Sebai A., G. Feraud, H. Bertrand, J. Hanes, 1991, 40Ar/
depth to detachment: The AAPG bulletin, v. 77, no. 8,
39
Ar dating and geochemistry of tholeiitic magmatism
p. 1291-1302. related to the early opening of the central Atlantic rift:
Erslev, E. A., 1991, Trishear fault-propagation folding: Earth and Planetary Science Letters, 104, p. 455-
Geology, v. 19, p. 617-620. 472.
Finch, E., S. Hardy, R. Gawthorpe, 2004, Discrete- Schlische, R. W., 1995, Geometry and Origin of Fault-
element modelling of extensional fault-propagation Related Folds in Extensional Settings: The AAPG
folding above rigid basement fault blocks: Basin Bulletin, v. 79, no. 11, p. 1661-1678.
research, 16, p. 489-506. Withjack, M. O., J. Olson, E. Peterson, 1990,
Frizon de Lamotte, D., B. Saint Bezar, R. Bracène, Experimental Models of Extensional Forced Folds:
E. Mercier, 2000, The two main steps of the Atlas The AAPG Bulletin, v. 74, no. 7, p. 1038-1054.
building and geodynamics of western Mediterranean: Xiao, H., J. Suppe, 1992, Origin of Rollover, The
Tectonics, v. 19, no 4, p. 740-761. AAPG Bulletin, v. 76, no. 4, p. 509-529.
18
C. R. Geoscience 338 (2006) 11–24
http://france.elsevier.com/direct/CRAS2A/
Tectonics
Abstract
The Anti-Atlas is reviewed and examined in the light of its geodynamic significance as a Palaeozoic basin and fold belt. Short-
ening is accommodated by polyharmonic buckle folding of the cover in a thick-skinned fashion without the development of any
significant thrust/duplex systems. The Anti-Atlas is heavily inverted deep intracratonic basin, rather than a former passive margin of
the Palaeo-Tethys Ocean. Inversion took place in Late Carboniferous to Early Permian times. Main shortening directions changed
from NW–SE to north–south and maybe NE–SW through time, leading to the development of dome and basin patterns on scales
from 100 m to 10 km. To cite this article: M. Burkhard et al., C. R. Geoscience 338 (2006).
© 2005 Académie des sciences. Published by Elsevier SAS. All rights reserved.
Résumé
Tectonique de l’Anti-Atlas marocain. L’Anti-Atlas est revu et examiné sous l’angle de sa signification géodynamique comme
bassin paléozoïque et comme chaîne plissée paléozoïque. Le raccourcissement est accommodé par le plissement polyharmonique
de la couverture, avec une nette implication du socle. Aucun système significatif de chevauchement ni duplex ne s’est développé.
L’Anti-Atlas est un bassin intracratonique fortement inversé plutôt qu’une partie de la marge passive de la Paléotéthys. L’inversion
doit dater du Carbonifère tardif/Permien précoce. La direction du raccourcissement a changé au cours du temps depuis une direction
NW–SE vers une direction nord–sud et peut-être même NE–SW, ce qui conduit à la formation de figures d’interférences de plis en
dômes et bassins aux échelles allant de 100 m à 10 km. Pour citer cet article : M. Burkhard et al., C. R. Geoscience 338 (2006).
© 2005 Académie des sciences. Published by Elsevier SAS. All rights reserved.
Keywords: Folds and folding; Continental contractional orogenic belts; Africa; Morocco; Variscan–Hercynian
Mots-clés : Plis et plissement ; Chaînes orogéniques continentales de contraction ; Afrique ; Maroc ; Varisque–Hercynien
Fig. 1. The Anti-Atlas is shown in its larger context at the End of the Palaeozoic [80]. Isopach contours for total sedimentary thickness are given
in kilometres for those Palaeozoic basins which have not or only weakly been involved in inversion tectonics [102]. The same colour shades are
schematically superimposed onto the Anti-Atlas fold belt in order to illustrate the estimated depth of >10 km of this basin prior to inversion.
Alleghenian basement uplifts are shown in blue (inspired by the Appalachian ‘Blue ridge’). Internal, metamorphic and in part older portions of the
Applachian–Mauritanides–Moroccan Meseta are coloured in green and pink tones [44,73].
Fig. 1. Situation de l’Anti-Atlas par rapport à la chaîne des Appalaches à la fin du Paléozoïque [80]. Les contours isopaques en kilomètres sont don-
nés pour les bassins sédimentaires paléozoïques qui n’ont subi que peu ou pas de déformation [102]. Le même code couleur est superposé à la chaîne
plissée de l’Anti-Atlas afin d’indiquer la profondeur estimée de ce bassin avant l’inversion. Les massifs de socle alléghaniens sont indiqués en bleu
(inspiré du Blue Ridge appalachien). Les parties internes, métamorphiques et plus anciennes, de la chaîne des Appalaches–Mauritanides–Méséta
marocaine sont colorées en vert et en rose [44,73].
to a rejuvenation of relief with summits of 2500 m provinces also have their non-folded time-equivalent in-
and higher [35,47]. Geomorphologists have coined the tracratonic basins further inland: Michigan and Illinois
French term relief appalachien to characterize the pat- on the American side, Taoudenni, Tindouf and others on
tern of deeply eroded fold trains, and there is indeed a the African side. On closer inspection, however, and in
striking similarity between the geomorphic expression stark contrast to the frontal Appalachian chain, the Anti-
of the Appalachian Valley and Ridge and the Anti-Atlas. Atlas fold belt does not easily conform with the stan-
Both chains are external parts of the larger Variscan dard anatomy of foreland fold-and-thrust belts world-
Appalachian–Ouachita–Mauritanides orogen (Fig. 1). wide [77]. The most striking difference is the existence
The relationship between the Anti-Atlas fold belt and of major basement domes at a very short distance behind
the internal parts of this orogen remains to be elucidated the deformation front [76]. Similarities exist with Wind
in terms of tectonic style, timing and geodynamics. River-style basement uplifts of the frontal Rocky Moun-
At first sight, the Anti-Atlas shares many common tains, but in the Anti-Atlas, the basement uplifts occur
features with its American counterpart of the Valley and amidst a tightly folded thick Palaeozoic cover series.
Ridge in general and with the Alleghany Basin in par- The style of cover folding is quite unique too, with a
ticular. Both are located on the craton side of the oro- dominance of upright detachment folds and a complete
gen involving a thick and fairly regular layer cake of absence of any thrusting and duplex structures [47],
mostly shallow marine Palaeozoic sediments. Both fold with the exception of the westernmost parts of the Anti-
M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24 13
Fig. 2. Geologic overview map of the Anti-Atlas, compiled from the geologic map series (1:200 000) of the ‘Service géologique du Maroc’
[105–114]. NW–SE transects, used for the compilation of Fig. 4 are indicated along the southern border. A cross section through the Adrar Zouggar
anticlinorium is shown in Fig. 3.
Fig. 2. Carte géologique de l’Anti-Atlas, synthétisée à partir de la série des cartes 1:200 000 du Service géologique du Maroc [105–114]. Les
transects NW–SE, utilisés pour la compilation de la Fig. 4, sont indiqués le long de la bordure sud. Une coupe à travers l’anticlinorium de l’Adrar
Zouggar est donnée sur la Fig. 3.
Atlas along the Atlantic coast [8,65]. There is no thin 1.1. The basement
skinned basal décollement level and the western Anti-
Atlas does not conform with a foreland fold and thrust The Anti-Atlas basement is a complex assemblage of
belt system in the sense of Boyer and Elliott [15]. crystalline, metamorphic and sedimentary rocks. Note
An exhaustive review of the geology of the Anti- that we use the term basement here in the sense of
Atlas has been presented by Michard [64]. While many the petroleum geologists [58], including all rocks older
authors have interpreted the Anti-Atlas fold belt in than the Gondwana Megacycle. The oldest rocks of
terms of a predominance of strike-slip movements [43, the West-African Craton (WAC) are granitoids, gneisses
61,74,99], recent structural analyses depict the west- and a complex series of metamorphic rocks, attributed
ern Anti-Atlas folds as highly cylindrical frontal folds to the Eburnean orogeny at around 2000 Ma [2,96,
related with a ‘head-on’ collision in a NNW–SSE direc- 98]. The final assembly of most of the African conti-
tion [20,47,91]. nental crust takes place during the Panafrican orogeny,
In this paper the Anti-Atlas system is revisited in lasting roughly from 700 to 600 Ma [46]. Remnants
light of recent data regarding the Pre-Cambrian base- of a Panafrican suture zone are present as a dismem-
ment, the Palaeozoic cover, structural observations bered ophiolite series in the Bou Azzer inlier of the
within the Anti-Atlas belt as well as plate tectonic re- central Anti-Atlas [45,55,81,96] – a structure recently
constructions on a global scale. Questions of particu- reinterpreted as an aulacogen within the WAC [32].
lar interest concern the Palaeozoic sedimentary basin Elsewhere in the Anti-Atlas, the Panafrican event left
history and subsidence mechanisms as well as the evo- a more subtle imprint in the form of strike-slip shear
lution of this Anti-Atlas basin near the border of the zones and thrusts [43]. Post-Panafrican continental ex-
West-African Craton through time. tension is well documented for the entire Anti-Atlas re-
14 M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24
Fig. 3. Cross section through the Adrar Zouggar anticlinorium [27], see Fig. 2 for location. Surface geology is constructed from the 1:200 000
geologic map [110], augmented with our own detailed mapping and measurements of folds and other mesoscale structures [20]. Compare with
Michard [64, (Fig. 20)]. This section is tentatively area-balanced. A minimum determination of map scale cover shortening is 13 to 15% from
mesoscale folding alone. Additional shortening from intra-bed strains could well add another 10% or so of shortening [47]. A resulting ‘likely’
horizontal shortening of 25 km would need a mid-crustal décollement at about 25 km depth.
Fig. 3. Coupe à travers l’anticlinorium de l’Adrar Zouggar [27] ; pour la localisation, voir Fig. 2. La géologie de surface est construite à partir
de la carte géologique à 1:200 000 [110], augmentée par nos propres cartographies et mesures de plis et d’autres structures [20]. À comparer
avec Michard [64, (Fig. 20)]. Cette coupe est approximativement équilibrée en ce qui concerne les aires. Une détermination du raccourcissement
horizontal minimal à l’échelle de la carte est de l’ordre de 13 à 15% par le seul plissement mésoscopique. Quelque 10% de raccourcissement
additionnel par déformation intra-couche sont fort probables. Un raccourcissement total de 25 km nécessite une profondeur de décollement à
25 km de profondeur dans le socle.
gion [72,89] where indications for synsedimentary tec- logical basement/cover relationships, the sedimentary
tonics are found in the clastic series of the Saghro-group cover series is best defined as starting with the first
(PII3 ) [92], the Ouarzazate group (PIII) and, progres- carbonate-bearing units of the lowermost Adoudounian,
sively fading upward within lowermost Cambrian se- following in most places concordantly upon the coarse-
ries. The interpretation of events during the Late Neo- grained PIII conglomerates of the Ouarzazate series.
proterozoic (600–540 Ma) is still a matter of debate, This limit makes a major colour contrast in the field
due in part, at least, to the scarcity of reliable absolute and it is easily visible in satellite images. The PIII
ages. The significance of the basement/cover relation- behaves as a competent basement-like unit, while the
ships and the geodynamic context are also still open well-layered carbonates above it are increasingly de-
to discussion: interpretations range from syn-orogenic, tached and folded at different wavelengths from me-
Late Panafrican molasse series shed in a collision con- ters up to kilometres. Despite some prevailing uncer-
text, to post-orogenic extension and collapse with the tainties, the PIII/Adoudounian transition just above co-
formation of tilted blocks and halfgrabens [55,87,91]. incides with the Latest Proterozoic/Cambrian bound-
In a most recent interpretation [90], based on the type ary [18,38]. There is good evidence for some exten-
and volumes of volcanic rocks associated with the PIII sional tectonics continuing upward into the Earliest
Ouarzazate series, the Late Proterozoic extension event Cambrian (Lower carbonate series of the Adoudounian)
is compared with a basin and range extensional setting, at least in the western Anti-Atlas region [3,9,10]. Lat-
with an important production of lower crustal melts. eral correlations are greatly facilitated by the pres-
Soulaimani and Piqué [88] even go so far as to pro- ence of fossils and characteristic facies assemblages
pose the present-day Anti-Atlas boutonnières as being and the Anti-Atlas cover figures among the best stud-
re-activated former metamorphic core complexes. ied Palaeozoic series worldwide [11,17,29,37,50,64,73,
101]. A schematic compilation in the form of a time
1.2. The cover chart along a strike parallel line is shown in Fig. 4.
Overall, the Anti-Atlas cover series are predomi-
Prior to folding, the Palaeozoic cover series of the nantly deposited in a shallow marine environment. Im-
Anti-Atlas must have represented a fairly regular layer portant platform carbonate build-ups are observed dur-
cake configuration. Overall thickness reaches 10 km ing the Lower Cambrian of the western Anti-Atlas [13,
and more in the westernmost Anti-Atlas near Tiznit 38]. From Middle Cambrian to Late Silurian times,
and decreases to about 6 km and less in the east- sedimentation is dominated by detrital input from the
ernmost Anti-Atlas of the Tafilalt. In terms of rheo- African craton, i.e. from the east and/or the south [17,
M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24 15
Fig. 4. Schematic chronostratigraphy of the Anti-Atlas. Vertical axis is time [40], horizontal axis is a section along strike from west to east (see
Fig. 2 for locations). For each vertical column, stratigraphic data have been collected and projected from northwest (older) to southeast (younger)
transects. Colour coding (see legend) is used to illustrate the dominant character of sedimentation. White is for no deposition and/or erosion.
Informal formation names as well as major events are indicated.
Fig. 4. Schéma chronostratigraphique pour l’Anti-Atlas. L’axe vertical est le temps [40], l’axe horizontal est une coupe ouest–est le long de la
chaîne (voir la Fig. 2 pour la localisation des transects). Pour chaque colonne verticale, les données stratigraphiques ont été compilées et projetées
depuis le nord-ouest (vieux) et le sud-est (jeune). Le code couleur indique le caractère dominant de la sédimentation (voir légende). Les lacunes de
non dépôt ou d’érosion sont laissées en blanc. Les appellations informelles des formations ainsi que certains événements majeurs sont indiqués.
16 M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24
28]. Carbonate sedimentation is resumed at the end of ness changes are subtle at best, and strongly overprinted
the Silurian [50] and lasts for most of the Devonian, be- by later folding. Higher up in the stratigraphic col-
ing combined with clastic inputs, throughout the Anti- umn, subtle lateral facies changes, sedimentary wedges
Atlas region and beyond [100]. The Lower Carbonifer- and minor erosional disconformities have long been
ous [59] is marked by an renewed increase in detrital interpreted as echoing distant tectonic phases such as
input from the east, south and in places from the north Taconic, Acadian, Bretonne and Sudète [64], known
[66,73]. Regional-scale facies and thickness changes from either the Appalachian or the European Variscides.
point toward an open ocean to the west and northwest New palaeogeographic reconstructions (Fig. 6) of the
throughout the Palaeozoic. It is therefore tempting to in- evolution of Gondwana and the reassembly of Pangea
terpret the Anti-Atlas basin as the landward tier of the provide a reference frame against which such statements
former passive margin of Gondwana, facing west and can be tested [83,86,93].
northwestward. In the preserved stratigraphic record, The Late Cambrian hiatus could be interpreted as
the shelf break, continental slope and rise of this ancient related to a new phase of rifting further north and con-
passive margin wedge are missing, however. In com- comitant erosion on a southern rift shoulder. At least
parison with contemporaneous North-African Palaeo- two important Palaeozoic rifting phases have indeed
zoic basins such as Tindouf, Reggane, Béchar, Ahnet, been postulated to occur along the northern border of
Ghadames, Illizi, Hamra, Murzuq [12], the Anti-Atlas Gondwana [93]. Avalonian (and Armorican, etc.) ter-
cover series appears as just another intracratonic basin, ranes or continental fragments are supposed to originate
which happened to be close enough to the continental from the northern margin of Gondwana in Cambrian
edge to be more massively involved in Late Palaeozoic times, leading to the opening of the Rheic Ocean. A sec-
collision tectonics than its neighbours. ond rifting event of Hunic terranes during the Lower
Silurian would have led to the formation of the Palaeo-
1.3. Cover series as a mirror of distant tectonic Tethys ocean [93, (Fig. 3)]. In this proposal, future
events? Avalonian and Hunic terranes represent former active
subduction margins off the northwestern African cra-
Palaeozoic sedimentation of the Anti-Atlas region is ton. Both terranes would have drifted away after two
fairly continuous throughout, and minor disconformities successive stages of back-arc rifting giving way to two
are most easily explained in terms of sequence stratig- (Rheic, Paeothethys) Palaeozoic oceans. This proposal
raphy, i.e. sea-level and/or climatic changes. Probably opens new perspectives in the interpretation of events
the most important climatic events are the Late Ordovi- known in the various Palaeozoic terrains of Morocco.
cian glaciations, recorded within the 2nd Bani micro- The Anti-Atlas–Tindouf basin would not have been any-
conglomerates [39,56,95] and an associated erosional where close to the open ocean prior to Late Silurian
discordance. The following sea-level rise is hold re- times. We see little evidence for rifting phases from
sponsible for the deposition of the thick interval of Sil- Middle Cambrian times onward, however, and the easi-
urian black shales, a major source rock of the North- est interpretation is to assume that none of those rifting
African realm [12,68] and a potential décollement hori- events did take place anywhere close to the Anti-Atlas
zon. The famous Devonian Kellwasser–event at the basin, but rather in a more external position or in a dif-
Frasnian/Famennian boundary [16,54] is probably a ferent place along the northern margin of Gondwana.
meteorite impact that left its imprint in the faunal as- The former passive margin of Gondwana would be en-
semblages. tirely obscured by the future Atlantic rifting and hidden
As illustrated in Fig. 4, there is only one noteworthy below the Mesozoic onshore and offshore basins. Al-
interruption of the Palaeozoic sedimentary cycle in the ternatively, the Anti-Atlas basin could be interpreted
Late Cambrian. Classically, it has been correlated with as the southeastern half of an oblique back-arc basin
a sardic phase of orogeny [64] or with an epeirogenetic that evolved into a passive margin in Silurian–Devonian
uplift [29]. The widespread (although not absolute) na- times, after a (right lateral?) strike-slip departure of Hu-
ture of this hiatus, common to all North-African basins nic terranes.
from Morocco through Tunisia, Algeria to Libya [25] On a local scale, geologists have also long tried
speaks for a general, eustatic, rather than a local, tec- to correlate events between the Meseta, north of the
tonic origin. Crossley and McDougall [26, (p. 160)] present-day High Atlas chain, and the Anti-Atlas [70,
quote evidence for block faulting and marked angular 73,74]. There is no direct evidence for the existence of
unconformities in Algeria; similar evidence is missing any kind of tectonic deformation within the Anti-Atlas
in the Anti-Atlas of Morocco, where some lateral thick- at least up to Middle Carboniferous times. It is thus
M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24 17
quite obvious that the Moroccan Meseta underwent a along Drâa lowland north of the Jbel Ouarkziz [99],
very different tectonic history as compared to that of but field evidence seemingly rules out any significant
the Anti-Atlas, from the Ordovician onward [69,71]. wrenching of any age to occur in this area [20,47,91].
Many parts of the complex assemblage of units in the To the east, the Anti-Atlas belt – or basin – turns gradu-
Meseta have suffered several intense phases of deforma- ally into the Ougarta chain [30,48]. Just as in the case of
tion, metamorphism including the intrusion of granites the Anti-Atlas belt, folding, uplift and erosion prohibit
in Silurian times and from the Lower Carboniferous on- the establishment of original isopach contours for the
ward [49]. In this respect, the Meseta is comparable former Ougarta Basin, but it is quite clear that this intra-
to the internal Appalachian chain, with its Taconic and tratonic chain is localized along a former trough of NW–
Acadian belts. The most important question still open SE orientation. The same structural trend is also present
in this context regards the relative position of the Mo- in neighbouring basins of Reggane and Bechar and it is
roccan Meseta block with respect to the stable African generally admitted that this direction is inherited from
craton. In the light of the most widely accepted plate the Panafrican orogeny [24]. Our compilation in Fig. 1
tectonic models [86], the Meseta might be regarded as further illustrates the location of the different Palaeozoic
a small continental fragment (or terrane) rifted off the basins with respect to the Appalachian–Variscan chain
northern margin of Gondwana during an ill-defined pe- and the subsequent rifting axis of the Atlantic [80]. The
riod in the Lower Palaeozoic, comparable to Avalonia question arises of what was the geodynamic setting of
and the more recently postulated Hunic terranes [93]. these different North-African basins. Is the Anti-Atlas
Such fragments would have been accreted again to Basin a remnant part of a passive margin of Gond-
the African craton only during the latest stages of wana (Fig. 6)? The total thickness of the sedimentary
continent–continent collision in Late Carboniferous [14, pile alone might be used as an argument in favour of a
33]. Such mobilistic proposals have been made as early passive margin setting for the Anti-Atlas with the free-
as 1971 by Schenk [82], but the lack of an identified su- board to the WNW. Other Palaeozoic basins, however,
ture between the Anti-Atlas domain and the Meseta and at hundreds of kilometres inboard, accumulated similar
the similarities in the Lower Palaeozoic stratigraphic total sediment thicknesses in clearly intracratonic set-
record are quoted as evidence against such a sce- tings (Taoudenni > 8 km, Murzuq > 6 km). The same is
nario [49,51,52,71,73,75]; a recent confirmation of the true for large Palaeozoic basins on the American craton
Meseta block being of African affinity is provided by such as the Michigan (ca. 5 km) and to a lesser degree
crustal xenoliths found in Triassic lamprophyres [31]. Williston, Illinois, and Hudson Bay [57]. A major argu-
ment against a passive margin interpretation is the lack
2. Anti-Atlas: what kind of sedimentary basin? of any sedimentary record for deposits from the outer
shelf, talus and continental rise. Despite some general
The Anti-Atlas Basin has been involved in Late tendency to more open marine conditions westward,
Palaeozoic folding, uplift and erosion and it is therefore there are no deep water sediments preserved anywhere
difficult to establish its true initial shape and extent to in the Anti-Atlas (with the exception of some debatable
the west and to the north, where the Alpine High Atlas flysch deposits of Carboniferous age at the northern bor-
is a further obstacle to restorations. East and southeast- der near the High-Atlas front [66]. The determination
ward, the transition towards the neighbouring petroleum of palaeo-water depths within the Carboniferous clastic
bearing Palaeozoic basins of Bechar and Reggane are series of the Anti-Atlas remains a matter of controversy,
somewhat obscured by a thin blanket of Upper Creta- however. Sedimentary structures, such as graded bed-
ceous sediments (Hamada), but compilations from re- ding and intense convolute bedding as well as the pres-
flection seismics reveal the gross trends in the form ence of olistrostromes has been used as an indication for
of isopach contour maps shown in Fig. 1 [6,102]. The turbiditic flysch type sediments laid down in deep wa-
Anti-Atlas actually appears to be the northern half of ter in a compressional context [66]; we could not find
the Tindouf Basin. It is not clear if the two are sepa- any convincing flysch sequences, however, and a shal-
rated from each other by some sort of swell or basement low water origin in a coastal and deltaic environment
high, nor is there any known reason for the localiza- seems an equally appropriate interpretation of the sedi-
tion of the remarkably smooth trend of the Ouarkziz mentary structures [41,59].
chain (Fig. 1), which represents the deformation front A neutral reading of the stratigraphic record of the
of the Anti-Atlas belt to the SSE (Figs. 2 and 3) – Anti-Atlas (Figs. 4 and 5) reveals an eventless sub-
and the northern border of the Tindouf basin. A strike- sidence history very comparable to contemporaneous
slip boundary (of Alpine age) has been proposed to run intracratonic basins of North America [7,53]. The to-
18 M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24
tation in terms of a classic thin-skinned thrust system of detachment. With an excess area of ca. 380 km2 of
organization [15]. basement above regional, a minimum shortening esti-
On the crustal scale, however, the cover shorten- mate of 17 km requires a basal detachment in the lower
ing determined across the western Anti-Atlas fold belt, crust, at ca. −32 km depth (Fig. 3). Note that present-
poses a serious balancing problem: where and how is day Moho depth is thought to be at ca. −35 km [36] at
this shortening accommodated within the basement? least. A more likely estimated horizontal shortening of
Large structural domes, the so-called boutonnières [22] some 25 km, taking into account layer-parallel shorten-
of Proterozoic basement rocks, crop out at a very short ing features and other bed internal strains [47], would
distance behind the south-eastern front of the south- lead to a lesser depth of detachment at mid-crustal lev-
western Anti-Atlas and these basement inliers punc- els of ca. −25 km (Fig. 3).
tuate the tightly folded Palaeozoic cover. The lack of The Appalachians also used to have their basement
any mapped thrusts has led earlier authors to consider problem in the form of the Blue Ridge (Fig. 1), a lin-
even these basement uplifts as just another, deeper level ear belt pre-Cambrian basement rocks of Grenvillian,
of crustal scale folds [22,23] dubbed plis de fonds by i.e. American craton affinity [44,63,79] cropping out
Argand (Fig. 5A–C) [5]. Balancing and rheology con- at some distance behind the thin-skinned frontal Valley
siderations make such a gentle folding interpretation and Ridge fold-thrust belt. While these basement uplifts
improbable [67], however, and we infer that the Anti- have long been considered as more or less autochtho-
Atlas basement uplifts must be associated with a series nous too, crustal scale seismic reflection profiling has
of major crustal scale reverse faults, supposedly in a provided clear evidence of a truly allochthonous na-
Windriver-style well known in the foreland of the Rocky ture of these basement slivers, detached from the former
Mountains [67,76,94]. This uncommon proposal is il- edge of the passive margin and thrust craton-ward by
lustrated in a section across the western tip of the Adrar up to 200 km! [44, plate 1, section C], a tectonic style
Zouggar anticlinorium (Fig. 3) [27,110]. The backbone anticipated by Argand [5, (Fig. 5, case E)], [63]. In com-
of this structure is made of Ordovician quartzites [27] parison, our inversion interpretation of the Anti-Atlas
and drilling in the 1960s has documented the existence basement domes remains very modest and autochtho-
of a basement core at −3.4-km depth, at least 2.5 km nous indeed and the question arises if a Blue Ridge
above the regional basement top (−5 km) as constructed type interpretation would not be more appropriate in the
between the synclinorium to the north and the Tindouf case of the Anti-Atlas? We have not found any positive
basin to the south. Gently plunging fold axis allows the arguments in favour of such an analogy, and we con-
projection of the Jbel Rich (Devonian carbonates) folds clude that the similarities between the Appalachians and
above the top of this anticlinorium. Southward, defor- Anti-Atlas fold belts are not extending far below the ge-
mation ends below the Jbel Ouarkziz triangle structure, omorphic expression of a common Appalachian relief.
while northward, another major basement uplift of the We further propose the structural style of the Anti-Atlas
Tagragra d’Akka boutonnière has led to a complete re- belt being a rather unique combination of basement up-
moval of all cover series. The tightly folded Bani (Or- lifts in a Windriver style that happened to occur below a
dovician) clearly projects above this basement uplift, more than 10-km-thick series of mostly soft, shale dom-
however. The Adrar Zouggar and Tagragra d’Akka an- inated sediments. Tectonic compression of this Palaeo-
ticlinoria nicely illustrate the balancing problem asso- zoic basin leads to a massive inversion of the underlying
ciated with such basement ‘folds’. A conservative un- basement structures and a simultaneous polyharmonic
folding of the well-exposed Devonian and Ordovician buckle folding of the basin fill. Despite its proximity to
marker beds reveals a minimum of 12 to 15% shorten- the Appalachian–Variscan chain, the Anti-Atlas is not
ing ratio – this amounts to some 17 km of minimum easily classified as a foreland fold-thrust belt to this oro-
line-length shortening (Fig. 3). In order to compensate gen. In many respects, it is rather to be considered as a
this shortening at the basement level, we postulate the severely inverted intracratonic basin – related to the oro-
existence of a set of rather steep (40 to 60◦ ) faults, pos- gen on a crustal to lithospheric scale, supposedly with a
sibly Late Proterozoic normal faults that would have floor thrust at mid to lower crustal level. The Anti-Atlas
been strongly inverted. This structural style cannot be belt has a direct connection to the Ougarta–Ahnet fold
mapped in the case of the Adrar Zouggar, but it is com- belts, but other North-African basins suffered a simi-
patible with structural observations made along the bor- lar inversion event at the end of the Palaeozoic [24,42]
ders of more internal, deeply eroded basement inliers in more isolated, clearly intra-cratonic settings. Some
such as the Tagragra de Tata [21,47]. Simple area bal- striking analogies in structural style and complexity also
ancing considerations allow us to estimate the depths exist with the Iberian chain, an intraplate Cretaceous
20 M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24
inversion structure in the southern foreland of the Pyre- – during the Panafrican orogeny, a series of terranes
nees [84]. are accreted to the West African craton on its north-
Interestingly, there are no intracratonic structures of ern and probably western side. While the northeast-
Palaeozoic age known cratonward of the Alleghenian ern suture (Bou Azzer) and terranes to the north-
mountain front of the Appalachians. However, the very east will remain in place, northwestern and western
same craton reacted quite sensitively to Rocky Moun- borders are subsequently reactivated and a series
tain deformations and subduction along its western of terranes or continental fragments will be ripped
border, with the formation of basement uplifts and off again during the Palaeozoic. The southwestern
inversion of many old graben structures hundreds of
Anti-Atlas, however, at the margin of the Saha-
kilometres eastward, well within the North American
ran metacraton [1] remains attached to Gondwana
craton [60].
throughout its Palaeozoic history;
4. Conclusion: plate tectonic setting through time – in Late Proterozoic–Early Cambrian, the Anti-Atlas
(Cambrian to Permian) area is in extension with the formation of many
widely distributed graben and halfgraben struc-
The tectonic evolution of the Anti-Atlas basin and tures, filled in with coarse clastics (PIII) mostly
fold belt in relation to plate tectonics on a global scale of igneous origin. The youngest volcanism is tho-
is illustrated in Fig. 6, and summarized below: leiitic–alkaline and indicates an intracontinental
Fig. 6. Schematic evolution of the Anti-Atlas (star) in comparison with the Appalachian chain through time on a global, plate tectonic scale.
Palaeo-tectonic reconstructions are redrawn and simplified from Stampfli and Borel [93]. Cross sections (cartoons) on the left-hand side (Ap-
palachians) are according to Fichter [34].
Fig. 6. Évolution schématique de l’Anti-Atlas (étoile) en comparaison avec la chaîne des Appalaches à travers le Paléozoïque, à une échelle globale.
Les cartes paléo-tectoniques ont été redessinées et simplifiées à partir de Stampfli et Borel [93]. Les coupes schématiques à gauche (Appalaches)
sont selon Fichter [34].
M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24 21
[18] W. Buggisch, E. Fluegel, The Precambrian/Cambrian bound- [35] D. Frizon de Lamotte, B. Saint-Bezar, R. Bracene, E. Mercier,
ary in the Anti-Atlas (Morocco); discussion and new results, The two main steps of the Atlas building and geodynamics of
in: V.H. Jacobshagen (Ed.), The Atlas System of Morocco; the western Mediterranean, Tectonics 19 (4) (2000) 740–761.
Studies on its Geodynamic Evolution, Springer-Verlag, Berlin– [36] D. Frizon de Lamotte, A. Crespo-Blanc, B. Saint-Bézar, M. Co-
Heidelberg–New York, 1988, pp. 81–90. mas, M. Fernàndez, H. Zeyen, P. Ayarza, C. Robert-Charrue,
[19] M. Burkhard, S. Caritg, U. Helg, C. Robert-Charrue, Forced, A. Chalouan, M. Zizi, A. Teixell, M.-L. Arboleya, F. Alva-
disharmonic multilayer buckle folding in the Late Variscan rez-Lobato, M. Julivert, A. Michard, TRANSMED Transect
Anti-Atlas of Morocco, in: AAPG Annual Meeting, 2001, p. 30. I: Iberian Meseta–Guadalquivir Basin–Betic Cordillera–Albo-
[20] S. Caritg, Géologie structurale dans l’Anti-Atlas occidental du ran Sea–Rif–Moroccan Meseta–High Atlas–Sahara Platform,
Maroc. Implications tectoniques sur les relations entre dômes in: W. Cavazza, et al. (Eds.), The TRANSMED Atlas: The
de socle et couverture plissée en front de chaîne de montagne, Mediterranean Region from Crust to Mantle, Springer, Berlin,
PhD thesis, University of Neuchâtel, 2003. 2004, p. 141 & CD-ROM.
[21] S. Caritg, M. Burkhard, R. Ducommmun, U. Helg, L. Kopp, [37] G. Geyer, E. Landing (Eds.), MOROCCO’95 – The Lower–
C. Sue, Fold interference patterns in the late Paleozoic Anti- Middle Cambrian standard of western Gondwana, Beringeria,
Atlas of Morocco, Terra Nova 16 (1) (2003) 27–37. Special Issue 2, University of Würzburg, Germany, 1995.
[22] G. Choubert, Histoire géologique du Précambrien de l’Anti- [38] G. Geyer, E. Landing, W. Heldmaier, Faunas and depositional
Atlas, Notes Mém. Serv. géol. Maroc 162 (1963) 352. environments of the Cambrian of the Moroccan Atlas region,
[23] G. Choubert, A. Faure-Muret, Anti-Atlas (Maroc), in: Tectonic in: G. Geyer, E. Landing (Eds.), MOROCCO’95 – The Lower–
de l’Afrique, in: UNESCO Earth Sci. Series, 1971, pp. 163– Middle Cambrian Standard of Western Gondwana, Beringe-
175. ria, Special Issue 2, University of Würzburg, Germany, 1995,
[24] M.P. Coward, A.C. Ries, Tectonic development of North pp. 47–119.
African basins, in: T. Arthur, S. MacGregor Duncan, N.R. Ca- [39] J.-F. Ghienne, Late Ordovician sedimentary environments,
meron (Eds.), Petroleum Geology of Africa; New Themes and glacial cycles, and post-glacial transgression in the Taou-
Developing Technologies, Geological Society, London, 2003, denni Basin, West Africa, Palaeogeogr. Palaeoclimatol. Palaeo-
pp. 61–83. ecol. 189 (3–4) (2003) 117–145.
[25] R. Crossley, N. Mcdougall, Lower Palaeozoic reservoirs of [40] F.M. Gradstein, et al., International Stratigraphic Chart, http://
North Africa, in: D.S. MacGregor, R.T.J. Moody, D.D. Clark www.stratigraphy.org/chus.pdf, Stratigraphy I.C.O. Editor,
Lowes (Eds.), Petroleum Geology of North Africa, Geol. Soc. 2004.
London, 1998, pp. 157–166.
[41] J.R. Graham, Wave-dominated shallow-marine sediments in the
[26] R. Crossley, N. McDougall, D.S.E. MacGregor, R.T.J.E.
Lower Carboniferous of Morocco, J. Sediment. Petrol. 52 (4)
Moody, D.D.E. Clark-Lowes, Lower Palaeozoic reservoirs of
(1982) 1271–1276.
North Africa Petroleum geology of North Africa, Geol. Soc.
[42] H. Haddoum, R. Guiraud, A. Moussine-Pouchkine, Hercynian
London, Spec. Publ. 132 (1998) 157–166.
compressional deformations of the Ahnet–Mouydir Basin, Al-
[27] F. Desthieux, Etude tectonique et métallogénique du Jbel Ad-
gerian Saharan Platform: far-field stress effects of the Late
dana, Ordovicien des plaines du Dra, Maroc présaharien, Notes
Palaeozoic orogeny, Terra Nova 13 (3) (2001) 220–226.
Serv. géol. Maroc 268 (38) (1977) 209–236.
[43] B. Hassenforder, La tectonique panafricaine et varisque de
[28] J. Destombes, The Ordovician of the Moroccan Anti-Atlas, in:
l’Anti-Atlas dans le massif du Kerdous (Maroc), PhD thesis,
M.G. Bassett (Ed.), The Ordovician System, Univ. Wales Press
University of Strasbourg, France, 1987.
and Natl. Mus. Wales, 1976, pp. 411–413.
[29] J. Destombes, H. Hollard, S. Willefert, Lower Paleozoic rocks [44] R.D. Hatcher Jr., W.A. Thomas, P.A. Geiser, A.W. Snoke,
of Morocco, in: C.H. Holland (Ed.), Lower Palaeozoic of S. Mosher, D.V. Wiltschko, Alleghanian Orogen, in: R.D.
North-Western and West-Central Africa, Trinity Coll., Dep. Hatcher Jr., W.A. Thomas, G.W. Viele (Eds.), The Geology
Geol., Dublin, Ireland, 1985, pp. 91–336. of North America, The Appalachian–Ouachita Orogen in the
[30] M. Donzeau, J. Fabre, Tectonique des monts d’Ougarta, in: United States, Geol. Soc. Am., 1989, pp. 233–318.
Lexique stratigraphique international, Afrique de l’Ouest : in- [45] K.P. Hefferan, H. Admou, R. Hilal, J.A. Karson, A. Saquaque,
troduction géologique et termes stratigraphiques, N.S.N., 1983, T. Juteau, M.M. Bohn, S.D. Samson, J.M. Kornprobst, Protero-
pp. 118–120. zoic blueschist-bearing melange in the Anti-Atlas Mountains,
[31] J. Dostal, J.D. Keppie, J. Hamilton, E.M. Aarab, J.P. Lefort, Morocco, Precambr. Res. 118 (3–4) (2002) 179–194.
J.B. Murphy, Crustal Xenoliths in Triassic lamprophyre dykes [46] K.P. Hefferan, H. Admou, J.A. Karson, A. Saquaque, Anti-
in western Morocco: tectonic implications for the Rheic Ocean Atlas (Morocco) role in Neoproterozoic western Gondwana re-
suture, Geol. Mag. 142 (2) (2005) 159–172. construction, Precambr. Res. 103 (1–2) (2000) 89–96.
[32] N. Ennih, J.-P. Liégeois, The Moroccan Anti-Atlas; the West [47] U. Helg, M. Burkhard, S. Caritg, C. Robert-Charrue, Folding
African craton passive margin with limited Pan-African activ- and inversion tectonics in the Anti-Atlas of Morocco, Tecton-
ity; implications for the northern limit of the craton, Precambr. ics 23 (2004) 17, TC 4006.
Res. 112 (3–4) (2001) 289–302. [48] Y. Hervouët, G. Duée, Analyse morphostructurale par im-
[33] H. Feinberg, T. Aifa, J.-P. Pozzi, D. Khattach, J. Boulin, agerie satellitaire et coupes structurales modélisées des monnts
Courbes de dérive apparente des pôles magnétiques de d’Ougarta (Sahara occidental, Algérie) : une chaîne hercyni-
l’Afrique et de la Meseta marocaine pendant le Paléozoique, enne chevauchante à plis passifs, Mém. Serv. géol. Algérie 8
C. R. Acad. Sci. Paris, Ser. II 310 (1990) 913–918. (1996) 127–173.
[34] L. Fichter, Geologic History of Virginia, http://Csmres.jmu. [49] C. Hoepffner, M.R. Houari, M. Bouabdelli, Tectonics of the
edu/geollab/vageol/vahist/introduc.html, http://csmres.jmu.edu north African Variscides (Morocco, Western Algeria), an out-
/geollab/vageol/vahist/K-LatPal.html, accessed January 2005. line, C. R. Geoscience 338 (2006).
M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24 23
[50] H. Hollard, Tableaux de corrélations du Silurien et du Dévonien [67] G. Mitra, V.S. Mount, Foreland basement involved structures,
de l’Anti-Atlas, Notes Serv. géol. Maroc (1981) 42. AAPG Bull. 82 (1) (1998) 70–109.
[51] H. Hollard, J.P. Schaer, Southeastern Atlantic Canada, North- [68] A.M. Morabet, R. Bouchta, H. Jabour, An overview of the
western Africa, and Continental Drift; discussion, Can. J. Earth petroleum systems of Morocco, in: D.S. MacGregor, R.T.J.
Sci. 10 (4) (1973) 584–586. Moody, D.D. Clark Lowes (Eds.), Petroleum Geology of North
[52] M.R. Houari, C. Hoepffner, Structures des terrains paléo- Africa, Geological Society, London, 1998, pp. 283–296.
zoiques à la limite sud de la chaîne hercynienne du Maroc (Haut [69] A. Piqué, Variscan terranes in Morocco, Spec. Pap. Geol. Soc.
Atlas oriental), Afr. Geosci. Rev. 7 (1) (2000) 39–53. Am. 230 (1989) 115–129.
[53] P.D. Howell, B.A. Van der Pluijm, Structural sequences and [70] A. Piqué, Géologie du Maroc, Les domaines régionaux et leur
styles of subsidence in the Michigan basin, Geol. Soc. Am. évolution structurale, PUMAG, Marrakech, 1998.
Bull. 111 (7) (1999) 974–991. [71] A. Piqué, Geology of Northwest Africa, Beiträge zur regionalen
[54] M.M. Joachimski, R.D. Pancost, K.H. Freeman, H.C. Ostertag, Geologie der Erde, vol. 29, Gebrüder Bornträger Verlag, Berlin,
W. Buggisch, Carbon isotope geochemistry of the Frasnian– Stuttgart, 2001.
Famennian transition, in: G. Racki, M.R. House (Eds.), Late [72] A. Piqué, M. Bouabdelli, A. Soulaimani, N. Youbi, M. Iliani,
Devonian Biotic Crisis; Ecological, Depositional and Geo- Upper Neoproterozoic PIII conglomerates in the Anti-Atlas,
chemical Records, Elsevier, Amsterdam, 2002. southern Morocco; Pan-African molasses, or indicators of Up-
[55] M. Leblanc, J.R. Lancelot, Interprétation géodynamique du per Proterozoic rifting?, C. R. Acad. Sci. Paris, Ser. IIa 328 (6)
domaine pan-africain (Précambrien terminal) de l’Anti-Atlas (1999) 409–414.
(Maroc) à partir de données géologiques et géochronologiques, [73] A. Piqué, A. Michard, Moroccan hercynides: A synopsis. The
Can. J. Earth Sci. 17 (1) (1980) 142–155. Palaeozoic sedimentary and tectonic evolution at the northern
[56] P. Legrand, Paléogéographie du Sahara algérien à l’Ordovicien margin of West Africa, Am. J. Sci. 289 (1989) 286–330.
terminal et au Silurien inferieur, Bull. Soc. géol. France 174 (1) [74] A. Piqué, J.-J. Cornée, J. Müller, J. Roussel, The Moroc-
(2003) 19–32. can Hercynides, in: R.D. Dallmeyer, J.O.P. Lécorché (Eds.),
[57] M.W. Leighton, D.R. Kolata, Selected interior cratonic basins The West African Orogens and Circum Atlantic Correlatives,
and their place in the scheme of Global Tectonics A Synthe- Springer-Verlag, Berlin, 1991, pp. 229–263.
sis, in: M.W. Leighton, et al. (Eds.), Interior Cratonic Basins, [75] M. Roddaz, S. Brusset, J.-C. Soula, D. Beziat, M. Ben-Abou,
AAPG, Tulsa, 1990, pp. 729–797. P. Debat, Y. Driouch, F. Christophoul, A. Ntarmouchant, J. De-
ramound, Foreland basin magmatism in the western Moroccan
[58] D.S. MacGregor, R.T.J. Moody, D.D. Clark-Lowes, Petroleum
Meseta and geodynamic interferences, Tectonics 21 (5) (2002)
Geology of North Africa, Geological Society London, Spec.
23.
Publ. 132 (1998).
[76] J. Rodgers, Lines of basement uplifts within cratons marginal
[59] B. Mamet, G. Choubert, L. Hottinger, Notes sur le Carbonifère
to orogenic belts, Am. J. Sci. 287 (1987) 661–692.
du Jebel Ouarkziz. Étude du passage du Viséen au Namurien
[77] J. Rodgers, Fold-and-thrust belts in sedimentary rocks. Part 1:
d’après les Foraminifères, Notes Mém. Serv. géol. Maroc 27
Typical examples, Am. J. Sci. 290 (1990) 321–359.
(1966) 4–21.
[78] J. Rodgers, Fold-and-thrust belts in sedimentary rocks. Part 2:
[60] S. Marshak, K. Karlstrom, J.M. Timmons, Inversion of Pro-
Other examples, especially variants, Am. J. Sci. 291 (1991)
terozoic extensional faults: An explanation for the pattern of
825–886.
Laramide and Ancestral Rockies intracratonic deformation,
[79] J. Rodgers, Lines of basement uplifts within the external parts
United States, Geology 28 (8) (2000) 735–738.
of orogenic belts, Am. J. Sci. 295 (1995) 455–487.
[61] M. Mattauer, F. Proust, P. Tapponnier, Major strike-slip fault of [80] M. Sahabi, D. Aslanian, J.-L. Olivet, Un nouveau départ pour
Late Hercynian age in Morocco, Nature 237 (1972) 160–162. l’histoire de l’Atlantique central, C. R. Geoscience 336 (2004)
[62] P. Matte, Variscides between the Appalachians and the Urals; 1041–1052.
similarities and differences between Paleozoic subduction and [81] A. Saquaque, H. Admou, J.A. Karson, K. Hefferan, I. Reuber,
collision belts, in: J.R. Martinez Catalan, et al. (Eds.), Variscan– Precambrian accretionary tectonics in the Bou Azzer-El Grara
Appalachian Dynamics; the Building of the Late Paleozoic region, Anti-Atlas, Morocco, Geology 17 (1989) 1107–1110.
Basement, Geol. Soc. Am., 2002. [82] P.E. Schenk, Southeastern Atlantic Canada, Northwestern
[63] J.H. McBride, J.H. Knapp, Review of seismic reflector sig- Africa, and Continental Drift, Can. J. Earth Sci. 8 (1971) 1218–
natures of crustal deformation in the Appalachian-Caledonide 1251.
Orogen with reference to the Spanish Variscides and the [83] C.R. Scotese, Paleomap project, http://www.scotese.com/
Uralides, in: J.R. Martinez Catalan, et al. (Eds.), Variscan– Default.htm, 2005.
Appalachian Dynamics; the Building of the Late Paleozoic [84] J.L. Simon, Superposed buckle folding in the eastern Iberian
Basement, Geol. Soc. Am., 2002, pp. 281–300. Chain, Spain, J. Struct. Geol. 26 (2004) 1447–1464.
[64] A. Michard, Éléments de géologie marocaine, Notes Mém. [85] L.L. Sloss, Epilog (to Interior Cratonic Basins), in: M.W.
Serv. géol. Maroc 252 (1976) 408. Leighton, et al. (Eds.), Interior Cratonic Basins, AAPG, Tulsa,
[65] A. Michard, J. Sougy, L’orogenèse hercynienne à la lisière 1990, pp. 799–805.
nord-ouest de l’Afrique (Structure des chaînes primaires du [86] A.G. Smith, Gondwana: its shape, size and position from Cam-
Maroc au Sénégal), in : Colloque international CNRS : La brian to Triassic times, J. Afr. Earth Sci. 28 (1) (1999) 71–97.
Chaîne varisque d’Europe moyenne et occidentale, Rennes, [87] A. Soulaimani, Dynamique et interactions Socle/Couverture
France, 1977, pp. 605–640. dans l’Anti-Atlas occidental (Maroc) : Rifting fini-protérozoï-
[66] A. Michard, A. Yazidi, F. Benziane, H. Hollard, S. Willefert, que et orogenèse hercynienne, 1998.
Foreland thrusts and olistostromes on the pre-Sahara margin of [88] A. Soulaimani, A. Piqué, The Tasrirt structure (Kerdous inlier,
the Variscan Orogen, Morocco, Geology 10 (5) (1982) 253– Western Anti-Atlas, Morocco): a Late Pan-African transtensive
256. dome, J. Afr. Earth Sci. 39 (2004) 247–255.
24 M. Burkhard et al. / C. R. Geoscience 338 (2006) 11–24
[89] A. Soulaimani, M. Bouabdelli, A. Piqué, L’extension continen- [101] J. Wendt, Z. Belka, E. Fluegel, Age and depositional environ-
tale au Néo-Protérozoïque supérieur–Cambrien inférieur dans ment of Upper Devonian (Early Frasnian to Early Famennian)
l’Anti-Atlas (Maroc), Bull. Soc. géol. France 174 (1) (2003) black shales and limestones (Kellwasser facies) in the eastern
83–92. Anti-Atlas, Morocco, Facies 25 (1991) 51–89.
[90] A. Soulaimani, A. Essaifi, N. Youbi, A. Hafid, Les marqueurs [102] Tectonic Map of the World, AAPG Foundation, R.W. Wiener,
structuraux et magmatiques de l’extension crustale au Protéro- F.L. Wehr, G.M. Skerlec, I.O. Norton, 1985.
zoïque terminal–Cambrien basal autour du massif de Kerdous [103] P.A. Ziegler, S. Cloetingh, Dynamic processes controlling evo-
(Anti-Atlas occidental, Maroc), C. R. Geoscience 336 (16) lution of rifted basins, Earth Sci. Rev. 64 (1–2) (2004) 1–50.
(2004) 1433–1441. [104] P.A. Ziegler, S. Cloetingh, R. Guiraud, G.M. Stampfli, Peri-
[91] A. Soulaimani, C. Le Corre, R. Farazdaq, Deformation hercyni- Tethyan platforms; constraints on dynamics of rifting and
enne et relation socle/couverture dans le domaine du Bas-Draa basin inversion, in: P.A. Ziegler, et al. (Eds.), Peri-Tethyan
(Anti-Atlas occidental, Maroc), J. Afr. Earth Sci. 24 (3) (1997) Rift/Wrench Basins and Passive Margins, Éditions du Museum
271–284. national d’histoire naturelle, Paris, 2001.
[92] A. Soulaimani, A. Piqué, M. Bouabdelli, La série du PII–PIII
de l’Anti-Atlas occidental (Sud marocain) : un olistostrome à la Geological Maps
base de la couverture post-panafricaine (PIII) du Protérozoïque
supérieur, Earth Planet. Sci. Lett. 332 (2001) 121–127.
[105] D. Alaoui-Mdaghri, M. Bensaid, M. Dahmani, Carte géolo-
[93] G.M. Stampfli, G.D. Borel, A plate tectonic model for the Pale-
gique du Maroc 1:100 000, Bou Izakarn, Notes Serv. géol.
ozoic and Mesozoic constrained by dynamic plate boundaries
Maroc (1992).
and restored synthetic oceanic isochrons, Earth Planet. Sci.
[106] G. Choubert, Y. Ennadifi, Carte géologique du Maroc
Lett. 196 (1–2) (2002) 17–33.
1:200 000, Flanc sud de l’Anti-Atlas occidental et des plaines
[94] D. Stone, Morphology of the Casper Mountain uplift and re- du Dra, Akka-Tafragount-Tata, Notes Serv. géol. Maroc (1970).
lated subsidiary structures, central Wyoming: Implications for [107] G. Choubert, A. Faure-Muret, Carte géologique du Maroc
Laramide kinematics, dynamics, and crustal inheritance, AAPG 1:100 000, Igherm, feuille NH-29-XVI-2, Notes Serv. géol.
Bull. 86 (8) (2002) 1417–1440. Maroc (1983).
[95] O.E. Sutcliffe, J.A. Dowdeswell, R.J. Whittington, J.N. Theron, [108] G. Choubert, A. Faure-Muret, Carte géologique du Maroc
J. Craig, Calibrating the Late Ordovician glaciation and mass 1:100 000, Tafraout ; feuille NH-29-X-3, Notes Serv. géol.
extinction by the eccentricity cycles of Earth’s orbit, Geol- Maroc (1983).
ogy 28 (11) (2000) 967–970. [109] G. Choubert, A. Faure-Muret, Carte géologique du Maroc
[96] R.J. Thomas, L.P. Chevallier, P.G. Gresse, R.E. Harmer, B.M. 1:100 000, Taroudannt, Feuille NH-29-XVI-I, Notes Serv. géol.
Eglington, R.A. Armstrong, C.H. De Beer, J.E.J. Martini, G.S. Maroc (1983).
De Kock, P.H. Macey, B.A. Ingram, Precambrian evolution of [110] M.A. El-Alaoui, Y.C. Chefchaouni, M. Diouri, Carte
the Sirwa Window, Anti-Atlas Orogen, Morocco, Precambr. géologique du Maroc 1:200 000, Flanc sud de l’Anti-Atlas et
Res. 118 (2002) 1–57. des plaines du Drâa, Foum el Hassane-Assa, Notes Serv. géol.
[97] I.R. Vann, R.H. Graham, A.B. Hayward, The structure of moun- Maroc (1963).
tain fronts, J. Struct. Geol. 8 (3–4) (1986) 215–227. [111] S.E.M. Fetah, M. Bensaid, M. Dahmani, Carte géologique du
[98] G.J. Walsh, J.N. Aleinikoff, F. Benziane, A. Yazidi, T.R. Arm- Maroc 1:200 000, Tafilalt–Taouz, ministère de l’Énergie et des
strong, U–Pb zircon geochronology of the Paleoproterozoic Mines, royaume du Maroc (1986).
Tagragra de Tata inlier and its Neoproterozoic cover, western [112] S.E.M. Fetah, M. Bensaid, M. Dahmani, Carte géologique du
Anti-Atlas, Morocco, Precambr. Res. 117 (2002) 1–20. Maroc au 1:200 000, Todrha–Ma’der, ministère de l’Énergie et
[99] R. Weijermars, Estimation of paleostress orientation within de- des Mines, royaume du Maroc (1986).
formation zones between two mobile plates, with Suppl. Data [113] S.E.M. Fetah, M. Bensaid, M. Dahmani, Carte géologique du
9329, Geol. Soc. Am. Bull. 105 (11) (1993) 1491–1510. Maroc au 1:200 000, Zagora–Coude du Drâa–Hamada du Drâa,
[100] J. Wendt, Disintegration of the continental margin of north- ministère de l’Énergie et des Mines, royaume du Maroc (1986).
western Gondwana; Late Devonian of the eastern Anti-Atlas [114] A. Guerraoui, M. Bensaid, M. Dahmani, Carte géologique du
(Morocco), Geology 13 (11) (1985) 815–818. Maroc 1:100 000, Fask, Notes Serv. géol. du Maroc (1997).
TECTONICS, VOL. 23, TC4006, doi:10.1029/2003TC001576, 2004
[1] The late Variscan Anti-Atlas of Morocco shows [Choubert, 1963] of Proterozoic basement rocks, crop
some conspicuous deviations from the standard out at a very short distance behind the southeastern front
anatomy of foreland fold-and-thrust belts. Large of the orogen. Detailed mapping of the Anti-Atlas did not
basement inliers crop out at a very short distance of reveal any evidence for major thrusts within this fold belt.
less than 50 km behind the southeastern front of the Layer-parallel décollements or detachments are required
between different structural levels, but no such décolle-
fold belt, reminiscent of Windriver-style basement
ment has ever been mapped to step up in a ramp-flat
uplifts. In contrast to the Rocky Mountain foreland, geometry. The complete lack of thrusts lead earlier authors
however, the Anti-Atlas basement uplifts punctuate to consider the entire basement uplifts as crustal-scale
tightly folded Paleozoic cover series similar in tectonic folds [Choubert, 1963; Choubert and Faure Muret,
style to the Appalachian Valley and Ridge province. 1971], inspired by the plis de fonds of Argand [1924,
Cover shortening is exclusively accommodated by Figures 5a, 5b, and 5c].
buckle folding, and the Anti-Atlas fold belt lacks any [3] The sedimentary cover of the Anti-Atlas includes the
evidence for duplexing or thrust faults other than the uppermost Proterozoic and an up to 10 km thick Paleozoic
occasional steep reverse fault found near basement series. These mildly deformed and nearly unmetamorphosed
inliers. Basement domes have classically been strata of the Anti-Atlas have received much attention from
considered as the result of vertical tectonics in a stratigraphers and paleontologists [Destombes et al., 1985;
Bertrand-Sarfati et al., 1991; Villeneuve and Cornee, 1994;
horst and graben fashion, or, alternatively, as large
Piqué, 2001]. Special attention has been given to the
‘‘plis de fond’’ [Argand, 1924], basement folds. Proterozoic-Cambrian boundary [Boudda et al., 1979;
Unfolding of a large portion of an Ordovician Bertrand-Sarfati, 1981; Buggisch, 1988b; Latham, 1988;
quartzite marker bed reveals a minimum shortening Geyer and Landing, 1995; Benssaou and Hamoumi, 2003].
of 17% (30 km). Balancing this section at the crustal In contrast to this wealth of stratigraphical and paleonto-
scale indicates a lower crustal detachment level at 18 logical literature, structural publications dealing with the
to 25 km depth. Basement shortening is inferred to be tectonics of the western Anti-Atlas remain scarce [Leblanc,
accommodated through massive inversion of former 1972; Michard, 1976; Soulaimani, 1998; Guiton et al.,
extensional faults, inherited from a Late Proterozoic- 2003].
Lower Cambrian rifting phase. INDEX TERMS: 8005 [4] On the basis of paleomagnetic observations near the
Structural Geology: Folds and folding; 8015 Structural Geology: South Atlas and the Tizi’n’Test faults, a strike-slip origin for
Local crustal structure; 8102 Tectonophysics: Continental the entire Anti-Atlas chain has been proposed [Mattauer et
contractional orogenic belts; 8159 Tectonophysics: Rheology— al., 1972]. This view is widely shared by French authors
crust and lithosphere; 9305 Information Related to Geographic [Leblanc, 1972; Donzeau, 1974; Jeannette, 1981], and
Region: Africa; KEYWORDS: Anti-Atlas, tectonics, balanced cross folding of the sedimentary cover has mostly been explained
section, folding, foreland fold and thrust belt. Citation: Helg, U., as draping over vertically uplifted basement blocks and/or
M. Burkhard, S. Caritg, and C. Robert-Charrue (2004), Folding as wrenching above inherited, subvertical ‘‘zones of weak-
and inversion tectonics in the Anti-Atlas of Morocco, Tectonics, ness’’ [Faik et al., 2002]. Horizontal shortening in the
23, TC4006, doi:10.1029/2003TC001576. central Anti-Atlas has been ignored, or estimated to be on
the order of 5 to 10% at most [Leblanc, 1972; Donzeau,
1974]. Published cross sections of the western Anti-Atlas
1. Introduction depict the basement inliers as horst and graben structures,
indistinct broad basement domes or delimited by vertical
[2] The Anti-Atlas foreland fold belt of southwestern strike-slip faults [Michard, 1976; Michard and Sougy, 1977;
Morocco is part of the Variscan Appalachian-Ouachita- Piqué et al., 1991]. Clearly, these cross sections imply no
Mauretanides chain. It shows some important deviations crustal-scale shortening within the basement, and fold trains
from typical foreland fold and thrust belts [Rodgers, 1990]. within the synclinoria remain unexplained in terms of
Large structural domes, the so-called ‘‘boutonnieres’’ material balance, a question which has never been
addressed in the Anti-Atlas. The striking difference between
Copyright 2004 by the American Geophysical Union. the Devonian ‘‘Jbel Rich’’ folds and the virtually unde-
0278-7407/04/2003TC001576$12.00 formed Jbel Ouarkziz monocline near the southern front of
TC4006 1 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 1. (a) Location of the Anti-Atlas in southwestern Morocco. The polygon represents the study
area. (b) Simplified geological map of the westernmost Anti-Atlas showing the irregular shape and
distribution of the structural domes. These Proterozoic ‘‘boutonnieres’’ consist of pre-Cambrian
crystalline basement and Late Proterozoic volcaniclastic series of variable thickness. The black line A-A0
shows the location of the large-scale cross section of Figure 10. The polygon represents the location of
the 3-D fold model shown in Figure 6. Abbreviations are as follows: AZ, Adrar Zouggar mountain; BR,
Bani range; DP, Drâa plain; JO, Jebel Ouarzemine; JT, Jebel Tayyert; LP, Lakhssass plateau; OM,
Ouarkziz monocline; TS, thin skin part of the ‘‘internal’’ Anti-Atlas.
the Anti-Atlas has been considered as evidence for a major model. Unfolding of this marker bed provides minimum
strike-slip fault [Weijermars, 1993], of a Neogene age. The estimates of horizontal shortening as well as a measure of
apparent ‘‘en echelon’’ arrangement of the Rich folds was regional variations in folding intensity. The consequences of
further used to deduce a right lateral shear sense and amount this tectonic shortening are discussed at the larger, crustal
of shear, assuming a subvertical plunge of fold axes scale of this ‘‘thick skinned’’ foreland fold belt.
[Weijermars, 1993]. In reality, however, fold axes plunges
are very gentle, less than 30°, and the asymmetric fold
pattern of the lower Drâa valley implies very weak, if any, 2. Geological Framework of the Western
wrenching deformations [Soulaimani et al., 1997]; paleo- Anti-Atlas
stress indicators are systematically orientated nearly perpen-
2.1. Basement Preconfiguration
dicular to the fold axial planes, rather than at an inferred
angle of some 50° [Weijermars, 1993]. New observations in [6] The Anti-Atlas fold belt is located at the northwestern
the southwestern Anti-Atlas have recently led to a general border of the West African craton and continues southward
revision of the classic model of vertical and strike-slip into the Mauretanides. Large-scale basement cored domes
tectonics [Soulaimani et al., 1997; Soulaimani, 1998]. with irregular shapes form a continuous area of positive
Convincing evidence exists for southwestward thrusting of structural relief (Figures 1 and 2). The long axis of the Anti-
the Lower Drâa inlier onto the sedimentary cover, and the Atlas belt axis is oriented SW-NE and measures some
dextral strike-slip boundary postulated by Weijermars 700 km.
[1993] is not substantiated by any field evidence. [7] The pre-Cambrian basement has a complex geologic
[5] In this study, we present a detailed structural analysis history. It was consolidated during the Panafrican orogeny
of a large portion of the SW Anti-Atlas. A continuous fold between 620 and 580 Ma [Choubert, 1963; Choubert
train in a competent Ordovician quartzite bed is mapped in and Faure Muret, 1971; Leblanc and Lancelot, 1980;
detail and rendered in a three-dimensional (3-D) structural Hassenforder, 1987]. In detail it shows considerable
2 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 2. Landsat image of the westernmost Anti-Atlas chain. Compare with Figure 1 for major
landmark features such as the basement inliers (dark green) with their ‘‘autochthonous’’ Cambrian cover
in light green and tan colors. Quartz-rich lithologies of the Jbel Bani, the Jbel Rich, and Jbel Ouarkziz
appear in dark purple. Image courtesy of NASA and Earth Satellite Corporation (available at http://
zulu.ssc.nasa.gov/mrsid/mrsid.pl). See color version of this figure at back of this issue.
heterogeneity. It contains the suture zone between the stable unit, separated from the crystalline basement by a major
West African craton and a Proterozoic mobile belt to the unconformity. The geodynamic context of the PII-III is still
north. Remnants of this suture zone are exposed in several a matter of debate [Piqué et al., 1999; Piqué, 2001]. In the
basement domes, for example in the Bou Azzer and in the case of the overlying PIII unit of clastics and volcaniclas-
El Graara inliers. In the central Anti-Atlas this suture zone tics, separated by a minor unconformity from the PII, there
is known as ‘‘accident majeur de l’Anti-Atlas.’’ Southwest is increasing evidence in favor of an extensional tectonic
of this suture zone, Eburnean (approximately 1900 Ma) regime, in relation with a new rifting cycle starting in
augen gneisses, metadolerites, and metamorphic rocks, the the Late Proterozoic [Piqué et al., 1995; Piqué, 2001;
so-called Zenaga and Kerdous Series, crop out in several of Soulaimani et al., 2003]. Extensional structures have been
the structural domes. The Proterozoic rocks are further described within the Proterozoic basement in the western-
characterized by numerous synorogenic granite and grano- most Anti-Atlas [Soulaimani et al., 1997] and large-scale
diorite intrusions. half-grabens are documented in the central Anti-Atlas
[8] The Panafrican orogeny is followed by an important [Azizi-Samir et al., 1990]. The post-Panafrican geological
phase of continental sedimentation, which results in locally history of the Anti-Atlas is now considered as an aborted rift
thick clastic and volcaniclastic deposits, the so-called PII-III [Piqué et al., 1995; Soulaimani et al., 2003].
3 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
4 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 4. A gently inclined fold limb north of Assa. The entire dip slope of this fold limb is formed by
one stratigraphic horizon of Upper Ordivician age: the or6c-quartzites. The foreground of the photograph
shows the Drâa plain corresponding to a lowland eroded into Silurian shales; view is to the north.
provinces are separated by the South Atlas shear zone, a cian Jbel Bani quartzites (Figure 4) and the Lower Devonian
major terrane boundary within the Variscan chains of Jbel Rich limestones, respectively. In the western Anti-Atlas
Morocco [Mattauer et al., 1972]. An overall geodynamic they are separated from each other by about 1000 to 1500 m
interpretation of a complete Canadian-Moroccan transect of Silurian shales and mudstones. The folding style of these
through the entire Appalachian-Meseta-Anti-Atlas system two units is very similar, albeit with different wavelengths
has been presented as early as 1971 [Schenk, 1971]. and amplitudes. Spectacular kilometer-scale Bani folds are
found especially in the western Anti-Atlas between the
towns of Bou Izakarn Assa and Icht (Figure 5). The Jbel
2.4. Anti-Atlas Foreland Fold Belt
Rich folds display smaller wavelengths and amplitudes at
[13] In terms of folding and deformation style, four the 500 m scale. The uppermost structural unit, a thick
structural units with different wavelengths and amplitudes series of Carboniferous sandstones and limestones of the
can be distinguished in the Anti-Atlas. They correspond to Jbel Ouarkziz, is separated from the Jbel Rich folds by a
four distinct stratigraphic levels separated by the thick thick series of Upper Devonian shales. Morphologically, the
incompetent units of the Middle Cambrian, Silurian, and Jbel Ouarkziz defines a very continuous monocline, where
Upper Devonian respectively. None of the fold levels shows the Carboniferous series dip gently south-southeastward
a clear and consistent vergence. into the adjacent Tindouf basin (Figure 2). Some authors
[14] The lowermost unit is represented by the basement (e.g., M. Zizi, personal communication, 2000; see also
domes and their autochthonous cover consisting of Late http://www.onarep.com) consider the lack of folding within
Proterozoic PIII conglomerates and Lower Cambrian lime- the tilted Jbel Ouarkziz as evidence for a Late Devonian age
stones and dolomites of the so-called ‘‘Adoudounian.’’ of folding in the Anti-Atlas, an interpretation inspired by
These basement domes have accessible, mapped amplitudes comparisons with northern and central Morocco and with
of up to 2 km, and the corresponding wavelength is on the the Canadian Appalachians [Piqué and Michard, 1989].
order of several kilometers to tens of kilometers. The true Direct field evidence in the form of an unconformity has
‘‘amplitude’’ of basement ‘‘folds’’ is in excess of 10 km, never been identified in the sedimentary series of the Anti-
however, larger than the total thickness of Paleozoic cover Atlas, where a tectonic interpretation (décollement) is more
series. Locally, thin layers of incompetent shales and generally accepted [Soulaimani, 1998]. While the deforma-
evaporites within the autochthonous Adoudounian series tion front of the Anti-Atlas system is classically seen at the
act as minor detachment horizons, resulting in second-order base (NNW) of the Jbel Ouarkziz, we propose a new
folding at a much smaller scale (10 to 100 m). Toward the interpretation as a triangle structure with a forethrust ending
central Anti-Atlas the ‘‘Lie de vin’’ formation acts as a somewhere blindly SSE of the Jbel Ouarkziz, at depth
distinct local décollement horizon (Figure 2). Fold axis within the Tindouf basin. From this most external tip point
trends within the Adoudounian are strongly influenced by of the Anti-Atlas thrust system, a NNW vergent backthrust
the preconfiguration of the underlying basement domes, and climbs north-northwestward to emerge into the shale series
they often deviate considerably from the large-scale SW-NE separating the Jbel Ouarkziz monocline from the Jbel Rich
Anti-Atlas orientation [Soulaimani et al., 1997; Soulaimani, fold train within the Lower Drâa valley. This interpretation
1998]. Higher up in the stratigraphic column, two structural is corroborated by structural observations, notably the tilting
units are characterized by detachment folds of the Ordovi- of the Jbel Ouarkziz monocline and bed internal layer
5 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 5. Intermediate-scale section across the folded Jbel Bani, in the northwestern sector of the 3-D
fold model shown in Figure 6. A major décollement is inferred within the incompetent Lie de vin
formation of Lower/Middle Cambrian age. The structural style below this décollement is unknown; for
lack of better knowledge, it is depicted here as a flat-lying, undeformed panel of autochthonous
basement-Lower Cambrian unit.
parallel shortening (LPS) deformation features within the outcrop-scale and microscale shortening yield each close to
youngest outcropping Carboniferous limestones [Burkhard 25% of the overall shortening.
et al., 2001]. [17] Qualitatively, layer parallel shortening features are
widely recognized as predating folding and thrusting and
this is very nicely demonstrated in the Jura mountains
3. Anti-Atlas Cross-Section Balancing [Plessmann, 1972; Homberg, 1996].
3.1. Theoretical Considerations [18] For the estimation of the total shortening in the Anti-
Atlas we concentrated on map-scale and outcrop-scale
[15] In areas of tectonic activity, deformation typically
structures. Since there are other important unknowns of
occurs at all scales, ranging from structures in crystal
the shortening at the scale of the orogen we were content
lattices to structures forming mountain ranges [Ramsay
with a qualitative assessment of microscale shortening. In
and Huber, 1983; Mitra, 1992; Mitra, 1994]. Three scales
order to minimize the effects of LPS, we choose a very
of observation are commonly used to describe structures:
competent quartzite horizon, devoid of any LPS features
microscale (<0.01 m), outcrop scale (0.01– 100 m), and map
visible at the outcrop to hand specimen scale. We believe
scale (>100 m). Primary sedimentary structures are used as
that this marker bed deformed essentially by buckling at the
proxies for the qualitative description and quantitative
outcrop to map scale.
determination of strain at all scales.
[16] Cross-section balancing is dealing with map-scale
3.2. Study Area and the Ordovician Quartzite
structures such as folds, faults, and thrusts. Smaller-scale
Marker Bed
structures accommodate additional deformation, which
should be considered in overall estimations of tectonic [19] Our estimation of the shortening is mainly based
shortening. Over the years, various investigators have on measurements and observations made in the Bani
shown that small-scale structures do make a significant mountain range, an area of folded Ordovician rocks between
contribution to total tectonic deformation. Mitra [1994] the towns of Bou Izakarn, Aouinat Torkoz, and Icht
measured strains in sections parallel to the transport direc- (Figure 1b) with excellent outcrop conditions (Figure 2).
tion across the Sevier fold-and-thrust belt. He could dem- The structural style is quite homogeneous, and fold trends
onstrate that failure to include microscale deformation in the are easily recognized.
restoration of regional cross sections results in overestima- [20] We concentrated on a very competent horizon within
tion of the wedge taper for the thrust belt. Another study the so-called ‘‘2nd Bani’’ formation of Ashgillian age. This
examined the contribution of both microscale and outcrop- uppermost competent unit of the Jbel Bani is labeled or6c
scale data to regional deformation using the foreland thrust according to Guerraoui et al. [1997] (Figure 3); on older
belt of the central Appalachians where small-scale struc- maps, however, this thin marker bed is not always distin-
tures account for more than 75% of the shortening in the guished from the immediately underlying or6b or or6s. It
roof sequence of this thrust belt [Smart et al., 1997]. Hogan consists of massive microconglomeratic quartz arenites,
and Dunne [2001] conducted a study in the Upper Devo- interpreted as being of glacial origin. This particular horizon
nian Chemung-Brallier Formation boundary in the western is ideal for the reconstruction of the map-scale structures:
Valley and Ridge province, where map-scale structures (1) It is ubiquitous in the study area and has a relatively
account for about 50% of the total shortening, whereas small thickness of 50 to 90 m [Destombes et al., 1985],
6 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
which is rather thin compared to the average wavelength of this way, we calculated the map-scale shortening for a
the folds. (2) The massive or6c quartzites are much more virtually unlimited number of cross sections.
resistant to weathering and erosion than the Silurian shales [25] In a second phase, sections were extracted from the
immediately above. The top of the or6c horizon is often model and enlarged for those locations where outcrop-scale
very well preserved and forms the present-day surface, often second-order folds had been observed in the field. These
in the form of dip slopes. In consequence, where the or6c- structures were then added ‘‘by hand’’ in order to estimate
quartzites crop out, the topographic contours can directly be their relative contribution to overall shortening.
used as structure contours (Figure 4). Owing to very [26] The assessment of outcrop-scale and microscale
favorable outcrop conditions, large areas could be checked shortening is mainly based on field observations. The
for outcrop-scale structures through the examination of fieldwork had essentially two goals: (1) an assessment of
aerial photographs and verification in the field. (3) By the outcrop-scale structures and the estimation of their
coincidence, the present-day valley floors in the western frequency and distribution and (2) the search for macro-
Anti-Atlas lie close to the inflection point of the folded Jbel scopic indicators of microscale deformation, their descrip-
Bani series. The dip at the inflection point is one of the most tion, and the estimation of their frequency and relative
important pieces of information for the reconstruction of a importance.
fold, especially when considering variations of fold ver- [27] Outcrop-scale deformation features include meter-
gence. Outcrop conditions allow for a straightforward scale minor folds and, more frequently, minor faults with
construction of fold closures below and above topography. displacements of the order of centimeters to decimeters. We
could hardly find any indicators for microscale deformation
3.3. Shortening Estimates Based on Folding features within the quartzites. Axial planes of folds in the
Jbel Bani quartzites are materialized through the develop-
[21] We constructed first a structure contour map of the ment of systematic sets of joints [Guiton et al., 2003] rather
top or6c surface at the 1:50,000 scale, using the most recent than a cleavage, let alone a schistosity. Even shaly inter-
geologic and topographic maps available, with an equi- layers still mostly display primary bedding and the occur-
distance of 25 m. Additional information includes our field rence of LPS feature in the form of a weakly developed
observations, bedding dip measurements, and interpreted pencil cleavage or other tectonic cleavage is rather excep-
aerial photographs. tional. Generally, deformation intensity increases downward
[22] Because of the abundance and the continuity of the in the stratigraphic pile, and the development of cleavage is
outcrops we were able to reconstruct most of the fold limbs more frequently observed near the basement inliers. Higher
and many fold closures quickly and with high accuracy. In a up in the stratigraphy, within limestones of the Devonian
second step, the structure contour lines for the eroded and Jbel Rich and the Carboniferous Jbel Ouarkziz, we identi-
subsurface parts of the folds were added using auxiliary fied layer parallel stylolite peaks as clear indicators of early,
cross sections and classical techniques of down-plunge layer parallel shortening in a SSE-NNW direction, at a high
projection along the measured fold axis direction [Ramsay angle to the strike of this monocline.
and Huber, 1987, p. 356ff ]. Since the stratigraphic thick-
nesses of all the Ordovician formations are very well
known, and not internally deformed to any measurable 4. Results
degree, we were able to reconstruct the hidden parts of
the folds with nearly the same level of accuracy as the 4.1. Map-Scale Shortening
outcropping limbs. [28] The top of the Ordovician or6c horizon has been
[23] The structure contour map was then transformed to a reconstructed for the entire western part of the Jbel Bani
digital 3-D surface with standard digitizing procedures. For (Figures 5 and 6). Our structure model covers an area of
most of our further analyses, we used this 3-D surface in a approximately 1600 km2. It nicely illustrates the low
TIN format, in which a network of small flat, but irregular complexity of this folded structure. Because of favorable
triangles approximates the real surface. In a TIN, the size of outcrop conditions even some of the larger outcrop-scale
the triangles is variable and depends, in our case, on the structures are resolved, especially in the southern quadran-
spacing of the structure contour lines. A TIN is ideal for gle (see Figure 7 for the use of ‘‘quadrangles’’ and ‘‘parts’’).
representing surfaces with features at variable scales since Clearly, folding is the predominant deformation style at the
even smaller structures are adequately represented. The map scale. A few minor tear faults appear in the northern
error stemming from replacing a curve by straight line quadrangle, but they are local phenomena, normally limited
segments (or in 3-D: a curved surface replaced by to one fold limb.
interconnected triangles) is negligible. In our case, even [29] Throughout the area the folds show a high lateral
smaller folds consist of numerous triangles. continuity and cylindricity. Individual fold hinges are near
[24] Using a common geographic information systems horizontal over several kilometers. Plunging fold hinges are
software package (ArcInfo, ArcView) allowed calculation generally found in pairs of en echelon lateral transitions.
of statistics for many of the relevant fold parameters such as The overall orientation of fold axes is very constant at
amplitude, wavelength, symmetry, vergence, and variations approximately N038°.
in the direction of the fold axis. Similarly, the extraction of a [30] Fold limbs near the inflection point have dips
section at a random point in the model and the calculation of between 35° and 75° (Figure 8). Most folds, irrespective
the curve’s length is a more or less automatic procedure. In of wavelength and amplitude are symmetrical and upright.
7 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 6. Three-dimensional rendering of a structure contour map constructed for the top of the or6c
marker horizon in the western part of the Jbel Bani. The small white circle shows the location of the
photograph of Figure 8. Note the decreasing amplitude and wavelength toward the south. See color
version of this figure at back of this issue.
8 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 8. Tight syncline near the town of Targoumait. The width of the small valley is about 60 m. The
or6c horizon forms the limbs of the syncline. This structure is at the transition from outcrop scale to map
scale. It is well resolved in the 3-D model.
9 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 9. Map-scale shortening calculations for different parts of the top or6c-surface. Abbreviations
are; n is number of sections; avg is average shortening; sdev is standard deviation.
10 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
vary between 10% and 20% with an average of 13.2% those observed in the Appalachian Tuscarora Sandstone.
(Figure 9c). In one site near Targoumait (Figure 5), we measured a
grain shape preferred orientation of the detrital quartz grains
4.2. Outcrop-Scale Shortening of Rs = 1.15 within the bedding plane. This shortening is
[35] Field observations and the examination of aerial oriented NW-SE at a high angle to the local fold axis and it
photographs allowed checking different structural positions could well be interpreted as due to an early, prefolding
for outcrop-scale structures. Also, very continuous out- shortening of up to 13%, if it was accommodated by pure
crops, such as the Jbel Oumloulles Anticline west of Assa ‘‘volume loss.’’ In any case, LPS shortening in quartz
(Figure 6, southern quadrangle) allowed checking for arenites seems to be much more restricted than the highly
outcrop-scale structures at several structural positions of variable values reported from shales and other clay-rich
the same fold. lithologies [Kisch, 1991].
[36] The most common outcrop-scale structures are wide [38] In summary, an absolute minimum horizontal short-
flexures of the or6c horizon. Typically, they are character- ening for the studied area, based on the map-scale balancing
ized by very low amplitude to wavelength ratio. Shortening of folds (13%) with the addition of a conservative estimate
associated with these structures is minimal, on the order of a of outcrop-scale shortening (2%), is 15%. It is more difficult
few meters in the case of individual, outcrop-scale struc- to give an estimate for the most likely total tectonic
tures. Their spacing is on the order of hundreds of meters to shortening. It could be as large 30%, twice the minimum
kilometers. More complex outcrop-scale structures such as shortening, if microscale deformation was of the same
tight minor folds and reverse faults are very rare. Such magnitude as reported from comparable Appalachian
structures could add another few tens of meters of shorten- Tuscaroro Sandstone. A more conservative estimate, based
ing per kilometer of cross section at the map scale. Together, on our preliminary thin section observations, is in the range
outcrop-scale shortening is estimated to account for less 15 to 25%.
than 5% total shortening at the map scale.
11 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
a nearly isoclinal east vergent overturned syncline, located is accessible for in situ study. In structural terms, it is
approximately 60 km north of the Jbel Bani. Its axis trends essentially a large continuous monocline, dipping at 15°
N021°, which is in perfect continuity to the northernmost to 30° south-southeastward, forelandward into the Tindouf
Bani folds. On its western side it is bounded by a large basin. This monocline stands in stark contrast to the tightly
thrust. Obviously, this structure accommodates more short- folded Devonian Jbel Rich underneath, from which it is
ening than the open upright folds of the northernmost Jbel separated by the lowland of the Drâa plain, running along a
Bani, and we conclude that the observed trend of increasing thick series of Upper Devonian-lower Carboniferous shales
shortening from southeast to northwest continues north- and siltstones. The Ouarkziz monocline bears all the key
westward outside of our study area [Belfoul et al., 2002]. features of a typical triangle structure [Banks and Warburton,
[42] Folding does not stop at the southeastern border 1986]. Unfortunately, in the absence of seismic lines, we have
of the Jbel Bani either. A general south-southwestward no means to determine the additional shortening accommo-
decrease in deformation intensity is expected, however, dated by underthrusting below this structure other than
and we take this into account in our ‘‘conservative’’ cross minimum estimates based on total structural relief.
section where we let the Bani folds rapidly decrease in
amplitude, wavelength, and shortening toward the south. A
possible indication for the extent of the folding below the 6. Tectonic Implications and Discussion
Drâa plain comes from the Adrar Zouggar and Addana
mountains [Desthieux, 1977] (AZ in Figure 1b). They 6.1. Crustal-Scale Balancing
represent the most external Ordovician outcrops of the [45] Having gained some new information about the
Anti-Atlas chain. Both of these Anticlinoria are basement shortening in the sedimentary cover of the western Anti-
cored domes as documented by drilling in the case of Adrar Atlas, some further implications for the basement are now
Zouggar [Michard and Sougy, 1977]. Unfortunately, this considered. For the illustration of our concepts a section
area is inaccessible for field studies at present. Detailed has been chosen where the interaction between basement
cross sections of the Addana and Adrar Zouggar mountains and sedimentary cover can be best demonstrated. This
[Desthieux, 1977] document folding of the same style, section runs across the central part the Jbel Bani (see
wavelength, and ‘‘intensity’’ as our own observations at Figure 1b), where our 3-D model has been constructed.
the southern border of the Jbel Bani. We will concentrate first on the tectonic implications,
[43] We conclude that the folded Ordovician strata ini- which come from the minimal shortening recorded in the
tially occupied a much larger area than seen today. Folding sedimentary cover. A conservative section has been con-
of the Jbel Bani is not restricted to structural lows where it is structed. For the folded Ordovician only map-scale and
preserved from erosion, but covered also large parts above outcrop-scale shortening were considered. Northwest of
the present day basement inliers. However, as seen in our the outcropping Jbel Bani folds, now eroded from above
model and on the basis of additional information from the basement high, the structural style is assumed to be the
isolated further outcrops, both amplitude and wavelength same geometry as within the 3-D model, i.e., open, upright
generally decrease from northwest to southeast. This folds with steep limbs. In order to account for an observed
decrease is most probably related to regional-scale facies increase in thickness of the Ordovician toward the north-
changes. The thickness of the Middle Cambrian shales, the west [Destombes et al., 1985], folds were drawn with
décollement horizon, also decreases toward the southeast a slightly increasing amplitude and wavelength in this
[Boudda et al., 1979; Benssaou and Hamoumi, 2003], direction.
thereby progressively losing its ability as an efficient [46] Southeast of the model, the Bani folds are depicted
detachment horizon. Toward the southeastern front, the as dying out rapidly below the Drâa plain. The Jbel
spacing between individual folds becomes more pro- Ouarkziz triangle structure is interpreted as stemming from
nounced. This is seen as a consequence of the combined one single thrust stepping up south-southeastward from the
effect of a decrease in layer thickness and an increase in the Middle Cambrian shales to Upper Devonian shales before
ratio of competent versus incompetent beds. ending blindly somewhere below the Tindouf basin in a tip
[ 44 ] Another important unknown in terms of total point, from which a postulated backthrust/décollement ho-
tectonic shortening at the orogen scale is the southwestern rizon must lead back up to the Earth’s surface, most likely
border of the Anti-Atlas, the so-called ‘‘Jbel Ouarkziz’’ (OM into the lower Drâa valley. We hold this backthrust respon-
in Figure 1b). There is strong geomorphic evidence for this sible for the marked disharmony between the folded Jbel
monocline being the surface expression of a major triangle Rich below and the nonfolded Jbel Ouarkziz monocline
structure. The Ouarkziz mountain chain forms a straight above.
ridge with altitudes of up to 500 m (200– 300 m above the [47] The top of the lower Cambrian has been recon-
valley floors) over more than 300 km length, a landmark structed using down-plunge projections and depth to
feature often photographed from space (images available at detachment calculations [Epard and Groshong, 1993;
http://earthobservatory.nasa.gov/Newsroom/NewImages/ Bulnes and Poblet, 1999]. In the western Anti-Atlas the
images.php3?img_id=7285). This ridge is also used over a Adoudounian (Lower Cambrian) consists mainly of massive
large distance as the ‘‘natural’’ boundary between Morocco, platform carbonate series [Destombes et al., 1985; Benssaou
Algeria, and the western Sahara, respectively. There is only and Hamoumi, 2003]. Incompetent horizons are scarce and
one road cut (Assa-Zag) at present, where the Ouarkziz chain sediments up to the Middle Cambrian are strongly coupled
12 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
to the basement (compare Figures 2 and 3). A décollement nearly perpendicular fold axes [Caritg et al., 2003]. Most
within these sediments, at least in the western Anti-Atlas, is authors agree that the intimate relationship of fold axes
very unlikely and not corroborated by any structural obser- directions and basement outlines is due to ‘‘forced’’ folding
vations. Therefore the whole pile from the basement up to of the sedimentary cover, accommodating differential
the top of the Lower Cambrian can be considered as one movements of basement blocks. It is difficult to envisage
structural unit, the ‘‘basement-Cambrian’’ unit in our rheo- these complicated fold pattern as the result of an early, thin
logic terminology. skin folding phase. Thus, at least in the Adoudounian of the
[48] As seen on most classical cross sections of the Anti- western Anti-Atlas, there exists no evidence for folding
Atlas and confirmed by our own field observations, the independent of the main deformation phase, which is clearly
folded Ordovician is clearly decoupled from this basement- thick skinned, involving basement. The tectonic style of
Cambrian unit. Folding is not restricted to structural lows these folds at the lowermost stratigraphic levels is very
between the basement inliers, however, and entire fold similar to the Rocky Mountains of Wyoming [Stone, 1993,
trains can be projected sideways above the larger basement 2002].
domes. This also means that Bani folding at any given [51] Higher up in the stratigraphy, however, within the
location cannot be linked to one specific (assumed) thrust Ordovician Jbel Bani, fold axes directions are very regular
fault in the basement. Shortening in the Ordovician needs to and increasingly independent of the underlying basement
be compared to the overall shortening on a much larger structures. Some minor variations in fold axis directions can
scale, so as to include the entire basement-cover contact. In still be explained as deflections around the basement inliers,
order to balance the section at the crustal scale, an equal albeit strongly attenuated above the Lie de vin detachment
amount of shortening is necessary in the sedimentary cover level. No polyphase deformation history is required to
and the basement. Shortening within the basement is explain these variations in fold axis trends in the western
assumed to take place along widely spaced, mostly hidden Anti-Atlas. Note that further east, near Tata, there is
large-scale reverse faults. Geometric constraints will be evidence for a two-phase history in the form of conspicuous
discussed below. fold interference patterns [Caritg et al., 2003], but even
there, both folding phases are best explained in a thick
6.2. Thin Skin Interpretation skinned fashion, because of reactivation of former basement
faults, in close analogy with Laramide uplifts of the Rockies
[49] With the Ordovician strongly decoupled from the [Stone, 1993; Marshak et al., 2000; Bump, 2003].
basement-Cambrian unit and a shortening increasing toward [52] Paleostress measurements in the western Anti-Atlas
the northwest, the question arises to which degree the fold belt reveal that the local axes of major compressive
observed folding could be the result of a thin skin thrusting- stress are subperpendicular to the fold axes throughout the
folding phase, predating the formation of the basement Jbel Bani. The paleostress field also shows some heteroge-
uplifts. The westernmost part of the Anti-Atlas, west of neity and it is characterized by a predominant local com-
the Bas Drâa inlier, is known as a thin skin domain indeed ponent. It seems therefore that the paleostress field is also
[Belfoul, 1991; Soulaimani, 1998; Piqué, 2001; Belfoul et conditioned by the basement blocks rather than being the
al., 2002]. It is characterized by map-scale thrusts and result of a distant push. In conclusion, an early thin skin
reverse faults within Lower Cambrian units. Accordingly, phase is regarded as a very improbable mechanism to
shortening observed in the Ordovician Jbel Bani might be explain the folding of the Ordovician Jbel Bani.
rooted either here or in even more internal positions to the
west, currently hidden in the Moroccan offshore. We
consider such a distant source for the shortening of the Jbel 6.3. Thick Skin Interpretation
Bani as very unlikely, however. Such a thin skin interpre- [53] A rapid glance at the tectonic map of the SW Anti-
tation still requires two new postulates, which are not Atlas reveals the existence of ‘‘thick skin’’ tectonics, neces-
substantiated by any observations. First, a regional-scale sary to explain the large domes of Proterozoic-Cambrian
basal décollement within Middle Cambrian shales would [Rodgers, 1995]. The central question is not the existence,
have to be increasingly important westward, prior to cutting but the nature and origin of these basement domes, as well as
down into the basement-Cambrian unit, in locations where their relationship with the buckle folding observed in the
neither direct nor indirect evidence for any major basement cover series around and above them. Direct evidence for
ramp is present. Secondly, the basement uplifts of the SW large-scale thrusts within the basement-Cambrian unit is
Anti-Atlas would have to be explained by a second defor- elusive in the western Anti-Atlas. This lack of direct evi-
mation phase, in an essentially ‘‘vertical tectonics’’ fashion, dence should not be used as an argument against the
without any sizeable amount of horizontal shortening. This existence of thrusts or reverse faults at depth, however.
latter view of the Anti-Atlas basement uplifts does not There is a series of explanations why such thrusts have not
provide any explanation for the observed crustal-scale been mapped, nor postulated. The most important reason is
thickening. the present-day erosion level, revealing only a very small,
[50] The fold axis directions within the Adoudounian of topmost 2 km of more than 10 km of structural relief of these
the western Anti-Atlas show considerable heterogeneity basement uplifts. This estimate is based on the known depth
[Soulaimani, 1998]. In many places, they parallel in a to basement in the distant foreland of the Tindouf basin,
conspicuous manner the outlines of the basement domes. where it is located at approximately 8 km [Michard, 1976;
This results in places in fold interference patterns with Bertrand-Sarfati et al., 1991]. Similarly, the preshortening
13 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 10. (a) Large-scale cross section through the entire western Anti-Atlas and corresponding
unfolded section, with a conservatively estimated minimum horizontal shortening of 30 km.
Abbreviations are as follows: Ad, Adoudounian; LC, Lower Cambrian; MC, Middle Cambrian; OR,
Ordovician. Stippled area shows the most competent units (mainly quartz arenites, sandstones, and
conglomerates). (b) Visualization of the ‘‘excess area’’ concept used to estimate the depth to an inferred
middle to lower crustal detachment. The cross-sectional area calculations are based on the assumption of
an initially 10 km deep sedimentary basin, necessary to accommodate the thickness of the Paleozoic
series of the Anti-Atlas chain. A maximum estimate of shortening of the order of 60 km (30 km folding +
15% LPS) requires a midcrustal detachment at about 18 km depth (labeled b1). A shallow detachment at
15 km depth leads to an unreasonably high shortening of 95 km (labeled b2).
top basement level in the southwestern Anti-Atlas region [Soulaimani, 1998]. Here, intense deformation under lower
must have been at some 10 km below sea level in order to greenschist facies lead to the development of a foliation
accommodate the total thickness of Paleozoic sediments, with steeply plunging stretching lineations and a marked
which comprise shallow marine series up to the middle internal deformation of rhyolite pebbles of the PIII; all these
Carboniferous (Visean). deformation features are compatible with a south-southeast-
[54] Another reason why no large-scale thrusts were ward thrust.
found in the SW Anti-Atlas is the presence of widespread [56] At the scale of the entire Anti-Atlas, crustal-scale
thick Middle Cambrian shales. As depicted in Figure 10, we balancing considerations are used to constrain permissible
assume a tectonic style where thrust ramps, breaking basement geometries at depth. One of the key features of the
through the Lower Cambrian Adoudounian, turn to hori- Anti-Atlas is its high structural relief. A conservative
zontal ‘‘flats’’ within incompetent Middle Cambrian shales. estimate of the total thickness of the Paleozoic cover series
Minor thrusts may level out even deeper in the stratigraphy, of the SW Anti-Atlas yields 10 km thickness. Most of this
within thin evaporitic shales of the Adoudounian unit. thickness is preserved today in first-order synclinoria, in
[55 ] An important reverse fault or thrust has been between the basement domes. However, there is good
described in the westernmost Anti-Atlas, at the southern evidence for a similar amount of erosion from atop the
border of the Lower Drâa inlier, where Proterozoic base- westernmost basement inliers (Bas Drâa, Ifni, Kerdous),
ment is thrust toward the SE onto the sedimentary cover which were invariably at lower greenschist facies metamor-
14 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
phic conditions during late Variscan deformation. This Even if we assume a total submap-scale shortening of the
metamorphism is obvious from the deformation style at same magnitude as map-scale shortening, the calculated
the brittle/ductile transition and from our own illite crystal- depth to detachment remains at about 18 km (Figure 10b,
linity measurements [Buggisch, 1988a; Burkhard et al., ‘‘maximum shortening’’ diagram labeled b1). Considering
2001; Helg et al., submitted manuscript, 2004]. In contrast only the minimal shortening (the total observed shortening),
to previous authors, who mostly considered these basement the detachment depth comes to lie at about 26 km depth.
domes as all-time ‘‘horsts,’’ with a reduced thickness of [59] As an alternative to the antiformal stack of thin
total Paleozoic cover, we assume an essentially continuous basement slabs we postulate basement inversion of thick
thickness of sedimentary series across these basement block-like basement parts with the dimension of the present
domes. We therefore assume a similar structural relief to day boutonnieres, uplifted along steep reverse faults. The
exist at the basement cover contact, where it is documented, depth of the detachment horizon again determines the
and at the Devonian/Carboniferous boundary, where it is thickness of such basement blocks.
mostly based on indirect evidence.
[57] The present-day topography and average elevation of
the Anti-Atlas could be at least partly due to Neogene 6.4. Inversion Tectonics
reactivation in relation with High-Atlas inversion tectonics [60] The term inversion is used here in the sense of
[Beauchamp et al., 1999; Frizon-de-Lamotte et al., 2000]. Coward [1994] to describe regions which have experienced
The Anti-Atlas is also the location of Pliocene volcanic a reversal in uplift or subsidence. The Anti-Atlas nicely fits
activity, e.g., in the Jbel Siroua, and broad thermal uplift at a this definition; after an extended period of subsidence
regional scale has to be taken into consideration. Accord- during the Paleozoic, the Anti-Atlas basin is the locus of
ingly, the removal of more than 10 km of Paleozoic sedi- tectonic shortening, uplift, and erosion during the later
ments from above the basement inliers may not be solely the stages of the Variscan orogeny, possibly with some reacti-
result of late Variscan tectonics but the combined effect of vation in Neogene times [Frizon-de-Lamotte et al., 2000].
late Variscan tectonics, Neogene Atlas inversion, and ther- [61] Most authors agree that Variscan basements tectonics
mal hot spot related uplifts and concomitant erosion. The of the Anti-Atlas is dominated by the reactivation of former
total amount of non-Variscan uplift/erosion in the Anti-Atlas weakness zones. In earlier publications these weakness
is ill constrained. Upper Cretaceous sediments of the zones are invariably considered to be subvertical [Michard,
‘‘Hamada’’ clearly seal Variscan folds; their present-day 1976]. Sedimentological evidence for an important post-
elevation is about 400 m above sea level in the eastern Anti- Panafrican extensional phase has accumulated during the last
Atlas but close to sea level near the Atlantic coast. In two decades [Soulaimani et al., 2001; Benssaou and
between, this same marker horizon is completely eroded Hamoumi, 2003]. Decametric fault-bounded graben struc-
from above the SW Anti-Atlas, leaving some room for tures are documented within the Proterozoic basement in the
speculation. Thermal uplift, related to the volcanic activity westernmost Anti-Atlas [Soulaimani, 1998]. Regional
of the Jbel Sirhoua, can only be estimated from general changes in thickness and facies of Cambrian series
considerations. Surface uplift related to plates moving over have been used to map the location and general orientation
hot spots may reach 1 to 3 km, affecting large areas of up to of graben structures [Benssaou and Hamoumi, 2003]
2000 km across [Coward, 1994; Eisbacher, 1996]. (Figure 11).
[58] Nevertheless, even with the assumption of up to 3 km [62] In the light of this post Panafrican rifting phase a
of Neogene surface uplift and High Atlas reactivation, the Variscan inversion of former PIII normal faults seems the
leveling of more than 7 km of structural relief remains to be most likely explanation for the mostly still hidden reverse
explained by late Variscan tectonics with concomitant uplift faults bounding the present day basement inliers. Rather
and erosion. This same amount of structural relief is not than all-time horsts [Michard, 1976; Piqué, 1994; Benssaou
only attained at the scale of the orogen but also at the scale and Hamoumi, 2003], basement inliers in our interpretation
of individual basement domes. Between the Jbel Bani and represent the former basin floors. The fault reactivation
the surrounding basement inliers Kerdous, Lower Drâa, and nicely explains the strong basement involvement and the
Ifni (Figure 2), the structural relief is in excess of 7 km. high basement topography in an otherwise only mildly
Given the modest amount of cover shortening observed in deformed tectonic setting.
the Anti-Atlas fold belt, this structural relief is difficult to [63] Tectonic inversion is, at least partly, corroborated by
explain by an antiformal stack of thin basement slabs above a the thickness distribution of the rift/early postrift sediments
relatively shallow crustal detachment [Boyer and Elliott, [Boudda et al., 1979]. The thickness of the lower-
1982]. As an example, if we assume 5 km thick basement most Adoudounian unit, also called ‘‘Lower Limestones’’
slabs overlain by a 10 km thick sedimentary series, a (Figure 11), is highest in the area of the basement domes of
horizontal shortening of roughly 95 km would be necessary the western Anti-Atlas, decreasing both east and southward,
to create the observed cross sectional excess area of 475 km2 forelandward. In other words, the location of maximum
(Figure 10b, ‘‘shallow detachment’’ diagram labeled b2). subsidence by rifting corresponds roughly with the area of
This exceeds by far our estimations of the horizontal maximum uplift by late Variscan tectonic inversion.
shortening observed within the sedimentary cover. The [64] The widespread occurrence of PIII rift sediments
discrepancy is too large to be explained by ill evaluated, found along the borders of almost all boutonnieres is an
underestimated outcrop-scale or microscale shortening. additional indication for these to be former grabens, rather
15 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
References
Argand, E. (1924), La tectonique de l’Asie, Proc. Int. Benssaou, M., and N. Hamoumi (2003), The Lower- Boyer, S. E., and D. Elliott (1982), Thrust systems,
Geol. Congr., 13th, 171 – 372. Cambrian western Anti-Atlasic graben: Tectonic AAPG Bull, 66, 1196 – 1230.
Azizi-Samir, M. R., J. Ferrandini, and J. L. Tane control of paleogeography and sequential organisa- Buggisch, W. (1988a), Diagenesis and very low-grade
(1990), Late Pan-African (580 – 560 M.A.) tectonic tion, C. R. Geosci., 335, 297 – 305. metamorphism of the Lower Cambrian rocks in the
activity and volcanism in the Anti-Atlas Mountains Bertrand-Sarfati, J. (1981), Problème de la limite Anti-Atlas, Morocco, in The Atlas System of
of Morocco: Geodynamic interpretation on the Précambrien-Cambrien: La section de Tiout Morocco: Studies on Its Geodynamic Evolution,
scale of north-west Africa, J. Afr. Earth Sci., 10, (Maroc): Les stromatolithes et leur biostratigraphie, Lecture Notes in Earth Sci., vol. 15, edited by
549 – 563. Newsl. Stratigr., 10, 20 – 26. Volker H. Jacobshagen, pp. 123 – 128, Springer-
Banks, C. J., and J. Warburton (1986), ‘‘Passive-roof’’ Bertrand-Sarfati, J., A. Moussine-Pouchkine, Verlag, New York.
duplex geometry in the frontal structures of the P. Affaton, R. Trompette, and Y. Bellion (1991), Buggisch, W. (1988b), The Precambrian-Cambrian
Kirthar and Sulaiman mountain belts, Pakistan, Cover sequences of the West African Craton, in boundary in the Anti-Atlas (Morocco), in The Atlas
J. Struct. Geol., 8, 229 – 237. The West African Orogens and Circum Atlantic System of Morocco: Studies on Its Geodynamic
Beauchamp, W., R. W. Allmendinger, M. Barazangi, Correlatives, edited by R. D. Dallmeyer and J. P. Evolution, Lecture Notes in Earth Sci., vol. 15, edi-
A. Demnati, M. Alji, and M. Dahmani (1999), Lecorche, pp. 65 – 82, Springer-Verlag, New ted by Volker H. Jacobshagen, pp. 81 – 90, Springer
Inversion tectonics and the evolution of the High York. Verlag, New York.
Atlas Mountains, Morocco, based on a geological- Bonhomme, M., and B. Hassenforder (1985), Le Bulnes, M., and J. Poblet (1999), Estimating the detach-
geophysical transect, Tectonics, 18, 163 – 184. métamorphisme hercynien dans les formations ment depth in cross sections involving detachment
Belfoul, M. (1991), La structuration hercynienne de la tardi-et post-panafricaines de l’Anti-Atlas occi- folds, Geol. Mag., 136, 395 – 412.
région de Labiar-Oued Boussafene (Guelmim, dental (Maroc): Données isotopiques Rb/Sr et Bump, A. P. (2003), Reactivation, trishear modeling,
Anti-Atlas sud-occidental), thèse 3ème cycle thesis, K/Ar des fractions fines, Sci. Geol. Bull., 38, and folded basement in Laramide uplifts: Implica-
162 pp., Univ. de Rabat, Rabat, Morocco. 175 – 183. tions for the origins of intra-continental faults, GSA
Belfoul, M. A., F. Faik, and B. Hassenforder (2002), Boudda, A., G. Choubert, and A. Faure-Muret Today, 13(3), 4 – 10.
Evidence of a tangential tectonic event prior to the (1979), Essai de stratigraphie de la couverture Burkhard, M., S. Caritg, U. Helg, and C. Robert-
major folding in the Variscan belt of the western sédimentaire de l’Anti-Atlas: Adoudounien - Charrue (2001), Forced, disharmonic multilayer
Anti-Atlas, Morocco, J. Afr. Earth Sci., 32, 723 – Cambrien Inférieur, Notes Mem. Serv. Geol. Maroc, buckle folding in the late Variscan Antiatlas of
739. 271, 96. Morocco, paper presented at American Association
16 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
of Petroleum Geologists 2001 Annual Meeting, Huon, S. (1985), Clivage ardoisier et réhomogénisation glomerates in the Anti-Atlas, southern Morocco:
Tulsa, Okla. isotopique K-Ar dans des schistes paléozoiques du Pan-African molasses, or indicators of upper Pro-
Caritg, S., M. Burkhard, R. Ducommmun, U. Helg, Maroc: Etude microstructurale et isotopique: Con- terozoic rifting?, C. R. Acad. Sci., Ser. II, 328, 409 –
L. Kopp, and C. Sue (2003), Fold interference pat- séquences régionales, Ph.D. thesis, 124 pp., Univ. 414.
terns in the late Paleozoic Anti-Atlas of Morocco, Louis Pasteur, Strasbourg. Plessmann, W. (1972), Horizontal-Stylolithen im fran-
Terra Nova, 16, 27 – 37. Jeannette, D. (1981), Influence de structures pan- zösisch-schweizerischen Tafel-und faltenjura und
Choubert, G. (1963), Histoire géologique du précam- africaines sur des déformations hercyniennes ihre Einpassung in den regionalen Rahmen, Geol.
brien de l’Anti-Atlas, Notes Mem. Serv. Geol. (Bordure septentrionale de la boutonnière d’Igherm, Rundsch., 61, 332 – 347.
Maroc, 162, 352. Anti-Atlas occidental), Notes Serv. Geol. Maroc, Ramsay, J. G., and I. M. Huber (1983), The Techniques
Choubert, G., and A. Faure Muret (1971), Anti-Atlas 42, 25 – 32. of Modern Structural Geology, 307 pp., Academic,
(Maroc), in Tectonique de l’Afrique (Tectonics of Kisch, H. J. (1991), Development of slaty cleavage and San Diego, Calif.
Africa), Earth Sci. Ser., vol. 6, pp. 163 – 175, UN- degree of very-low-grade metamorphism: A review, Ramsay, J. G., and I. M. Huber (1987), The Techniques
ESCO, Paris. J. Metamorph. Geol., 9, 735 – 750. of Modern Structural Geology, 697 pp., Academic,
Coward, M. (1994), Inversion Tectonics, in Continental Kogbe, C. A. (1998), Sedimentary Basins of Morocco, San Diego, Calif.
Deformation, edited by P. L. Hancock, pp. 289 – Afr. Geosci. Rev., 5, 232. Rodgers, J. (1990), Fold-and-thrust belts in sedimentary
304, Pergamon, New York. Latham, A. (1988), The Precambrian-Cambrian transi- rocks, part 1: Typical examples, Am. J. Sci., 290,
Desthieux, F. (1977), Etude tectonique et métallogén- tion in Morocco: Problems and perspectives, 321 – 359.
ique du Jbel Addana, Ordovicien des plaines du in Trace Fossils, Small Shelly Fossils, and the Rodgers, J. (1995), Lines of basement uplifts within the
Dra. Maroc présaharien, Notes Serv. Geol. Maroc, Precambrian-Cambrian Boundary, edited by external parts of orogenic belts, Am. J. Sci., 295,
268, 209 – 236. E. Landon, G. M. Narbonne, and P. M. Myrow, 455 – 487.
Destombes, J., H. Hollard, and S. Willefert (1985), Bull. N. Y. State Mus., 1076, 14 pp. Schenk, P. E. (1971), Southeastern Atlantic Canada,
Lower Paleozoic rocks of Morocco, in Lower Leblanc, M. (1972), Sur le style disharmonique des plis northwestern Africa, and continental drift, Can. J.
Palaeozoic of North-Western and West-Central hercyniens, à la base de la couverture, dans l’Anti- Earth Sci., 8, 1218 – 1251.
Africa, vol. 4, edited by C. H. Holland, Atlas central (Maroc), C. R. Acad. Sci., Ser. D, 275, Sebai, A., G. Feraud, H. Bertrand, and J. Hanes (1991),
pp. 91 – 336, Trinity Coll., Dep. of Geol., Dublin, 803 – 806. AR-40/AR-39 dating and geochemistry of tholeiitic
Ireland. Leblanc, M., and J. R. Lancelot (1980), Interpretation magmatism related to the early opening of the cen-
Donzeau, M. (1974), L’Arc Anti-Atlas-Ougarta (Sahara geodynamique du domaine pan-africain (Precam- tral Atlantic rift, Earth Planet. Sci. Lett., 104, 455 –
nord-occidental, Algérie-Maroc), C. R. Acad. Sci., brien terminal) de l’Anti-Atlas (Maroc) a partir de 472.
Ser. II, 278, 417 – 420. donnees geologiques et geochronologiques, Can. J. Smart, K. J., W. M. Dunne, and R. D. Krieg (1997),
Eisbacher, G. H. (1996), Einführung in die Tektonik, Earth Sci., 17, 142 – 155. Roof sequence response to emplacement of the
374 pp., F. Enke, Stuttgart, Germany. Marshak, S., K. Karlstrom, and J. M. Timmons (2000), Wills Mountain duplex: The roles of forethrusting
Engelder, T., and S. Marshak (1985), Disjunctive clea- Inversion of Proterozoic extensional faults: An ex- and scales of deformation, J. Struct. Geol., 19,
vage formed at shallow depths in sedimentary planation for the pattern of Laramide and Ancestral 1443 – 1459.
rocks, J. Struct. Geol., 7, 327 – 343. Rockies intracratonic deformation, United States, Soulaimani, A. (1998), Dynamique et interactions
Epard, J. L., and R. H. J. Groshong (1993), Excess area Geology, 28, 735 – 738. Socle/Couverture dans l’Anti-Atlas occidental
and depth to detachment, AAPG Bull., 77, 1291 – Mattauer, M., F. Proust, and P. Tapponnier (1972), Ma- (Maroc): Rifting fini-protérozoı̈que et orogenèse
1302. jor strike-slip fault of late Hercynian age in Mor- hercynienne, Thèse d’état, Univ. Cadi Ayyad,
Faik, F., M. A. Belfoul, M. Bouabdelli, and occo, Nature, 237, 160 – 162. Marrakech, Morocco.
B. Hassenforder (2002), The structures of the Late Michard, A. (1976), Eléments de géologie Marocaine, Soulaimani, A., C. Le Corre, and R. Farazdaq (1997),
Neoproterozoic and Early Palaeozoic cover of the Notes Mem. Serv. Geol. Maroc, 252, 408. Deformation hercynienne et relation socle/couver-
Tata area, western Anti-Atlas, Morocco: Poly- Michard, A., and J. Sougy (1977), L’orogenèse hercy- ture dans le domaine du Bas-Draa (Anti-Atlas occi-
phased deformation or basement/cover interactions nienne à la lisière nord-ouest de l’Afrique (Structure dental, Maroc), J. Afr. Earth Sci., 24, 271 – 284.
during the Variscan orogeny?, J. Afr. Earth Sci., 32, des chaı̂nes primaires du Maroc au Sénégal), in La Soulaimani, A., A. Piqué, and M. Bouabdelli (2001),
765 – 776. Chaı̂ne varisque d’Europe moyenne et occidentale, La série du PII – PIII de l’Anti-Atlas occidental
Frey, M. (1987), Low Temperature Metamorphism, Colloq. Int. Cent. Natl. Rech. Sci., vol. 243, (Sud marocain): Un olistostrome à la base de la
Chapman and Hall, New York. pp. 605 – 640, Cent. Natl. de la Rech. Sci., Paris. couverture post-panafricaine (PIII) du Protérozoic
Frizon-de-Lamotte, D., B. Saint-Bezar, R. Bracene, and Mitra, G. (1994), Strain variation in thrust sheets and supérieur, Earth Planet. Sci. Lett., 332, 121 – 127.
E. Mercier (2000), The two main steps of the Atlas across the Sevier fold-and-thrust belt (Idaho-Utah- Soulaimani, A., M. Bouabdelli, and A. Piqué (2003),
building and geodynamics of the western Mediter- Wyoming): Implications for section restoration and L’extension continentale au Néo-Protérozoı̈que
ranean, Tectonics, 19, 740 – 761. wedge taper evolution, J. Struct. Geol., 16, 585 – 602. supérieur-Cambrien inférieur dans l’Anti-Atlas
Geyer, G., and E. Landing (1995), MOROCCO ’95— Mitra, S. (1992), Balanced structural interpretations in (Maroc), Bull. Soc. Geol. Fr., 174, 83 – 92.
The Lower-Middle Cambrian Standard of Western fold and thrust belts, in Structural Geology of Fold Stone, D. (1993), Basement-involved thrust-generated
Gondwana, 269 pp., Inst. für Paläontol., Univ. and Thrust Belts, edited by S. Mitra and G. W. folds as seismically imaged in the subsurface of the
Würzburg, Würzburg, Germany. Fisher, pp. 53 – 77, Johns Hopkins Univ. Press, Bal- central Rocky Mountain foreland, in Laramide
Guerraoui, A., M. Bensaid, and M. Dahmani (1997), timore, Md. Basement Deformation in the Rocky Mountain
Carte géologique du Maroc 1:100’000, Fask, Notes Namson, J. (1981), Structure of the western Foothills Foreland of the Western United States, edited by
Serv. Geol. Maroc. Belt, Miaoli-Hsinchu area Taiwan: (1) Southern R. Chase, E. Erlsev, and C. Schmidt, Spec. Pap.
Guiton, M. L. E., W. Sassi, Y. M. Leroy, and B. D. M. Part, Pet. Geol. Taiwan, 18, 31 – 51. Geol. Soc. Am., 280, 271 – 318.
Gauthier (2003), Mechanical constraints on the Onasch, C. M. (1993), Determination of pressure solu- Stone, D. (2002), Morphology of the Casper Mountain
chronology of fracture activation in folded Devo- tion shortening in sandstones, Tectonophysics, 227, uplift and related subsidiary structures, central
nian sandstone of the western Moroccan Anti-Atlas, 145 – 159. Wyoming: Implications for Laramide kinematics,
J. Struct. Geol., 25, 1317 – 1330. Piqué, A. (1994), Géologie du Maroc: Les Domaines dynamics, and crustal inheritance, AAPG Bull.,
Harrison, M. J., and C. M. Onasch (2000), Quantitative Structuraux et Leur Évolution Structurale, 283 pp., 86, 1417 – 1440.
assessment of low-temperature deformation Ed. PUMAG, Marrakech, Morocco. Villa, I. (1998), Isotopic closure, Terra Nova, 10, 42 –
mechanisms in a folded quartz arenite, Valley and Piqué, A. (2001), Geology of Northwest Africa, 47.
Ridge Province, West Virginia, Tectonophysics, 310 pp., Gebrüder Bornträger, Stuttgart, Germany. Villeneuve, M., and J. J. Cornee (1994), Structure, evo-
317, 73 – 91. Piqué, A., and A. Michard (1989), Moroccan hercy- lution and palaeogeography of the West African
Hassenforder, B. (1987), La tectonique panafricaine et nides: A synopsis: The paleozoic sedimentary and Craton and bordering belts during the Neoprotero-
varisque de l’Anti Atlas dans le Massif du Kerdous tectonic evolution at the northern margin of West zoic, Precambrian Res., 69, 307 – 326.
(Maroc), Thèse Sci. thesis, 249 pp., Univ. L. Pas- Africa, Am. J. Sci., 289, 286 – 330. Weijermars, R. (1993), Estimation of paleostress
teur Strasbourg, Strasbourg, France. Piqué, A., J. J. Cornée, J. Müller, and J. Roussel (1991), orientation within deformation zones between
Hogan, J. P., and W. M. Dunne (2001), Calculation of The Moroccan Hercynides, in The West African two mobile plates, Geol. Soc. Am. Bull., 105,
shortening due to outcrop-scale deformation and its Orogens and Circum Atlantic Correlatives, edited 1491 – 1510.
relation to regional deformation patterns, J. Struct. by R. D. Dallmeyer and J. P. Lecorche, pp. 229 –
Geol., 23, 1507 – 1529. 263, Springer-Verlag, New York.
Homberg, C. (1996), Analyse des déformations cas-
santes dans le Jura et modélisation numérique des
Piqué, A., M. Bouabdelli, and J.-R. Darboux (1995), Le
rift cambrien du Maroc occidental, C. R. Acad. Sci.,
M.
Burkhard,
S. Caritg, U. Helg, and C. Robert-
perturbations des contraintes tectoniques autour Ser. II, 320, 1017 – 1024. Charrue, Institut de Géologie, Université de Neuchâtel,
d’accidents majeurs, Ph.D. thesis, 282 pp., Univ. Piqué, A., M. Bouabdelli, A. Soulaimani, N. Youbi, and Case Postale, Rue E. Argand 11, CH-2007, Neuchâtel,
Pierre et Marie Curie, Paris. M. Iliani (1999), Upper Neoproterozoic PIII con- Switzerland. ([email protected])
17 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 2. Landsat image of the westernmost Anti-Atlas chain. Compare with Figure 1 for major
landmark features such as the basement inliers (dark green) with their ‘‘autochthonous’’ Cambrian cover
in light green and tan colors. Quartz-rich lithologies of the Jbel Bani, the Jbel Rich, and Jbel Ouarkziz
appear in dark purple. Image courtesy of NASA and Earth Satellite Corporation (available at http://
zulu.ssc.nasa.gov/mrsid/mrsid.pl).
3 of 17
TC4006 HELG ET AL.: FOLDING AND INVERSION TECTONICS IN THE ANTI-ATLAS OF MOROCCO TC4006
Figure 6. Three-dimensional rendering of a structure contour map constructed for the top of the or6c
marker horizon in the western part of the Jbel Bani. The small white circle shows the location of the
photograph of Figure 8. Note the decreasing amplitude and wavelength toward the south.
8 of 17
Géologie structurale de l’Anti-Atlas oriental
PROJET D’ARTICLE
C
RESUMES ET POSTERS
- Structural study of the eastern Anti-Atlas, the variscan belt of Morocco. Poster.
SGM 2005.
ABSTRACTS
Abstract présenté au Marrakech International Oil & Gas 2002, session poster.
The Anti-Atlas of Morocco is a foreland belt, limited to the south by the west african craton and to the
north by the alpine High Atlas. This chain is included in the Appalachian orogeny. The Anti-Atlas appears
as a huge NE-SW anticlinorium. Locally, the basement is cropping out as " Boutonnières", but the main
part is made of an important Paleozoïc cover, wich is gently folded and of low grade metamorphism.
An entire Wilson cycle of mountain building left its imprint in these rocks. Sedimentation started with a
Late Proterozoic rifting phase, leading to the opening of the Japetos ocean to the west. Post-rift sediments
of Lower Cambrian to Middle Carboniferous age are dominated by detrial input from the cratonic
hinterland of Africa, leading to the accumulation of a thick pile of alterning quartzites, siltstones, shales
and occasional carbonates. During the last stages of closure of the Japetos ocean, this passive margin is
squeezed in the collision between Euramerica and North Africa. This collision leads to a massive inversion
of former normal faults, enormous uplifts of the Precambrian basement and a simultaneous folding of the
overlying cover series. Syn- to post-tectonic erosion and isostatic re-equilibration lead to the removal of
beetween 2 and 15 km of material, bringing lower green-schist facies rocks back to the earth's surface.
Minor phases of reactivation include far field effects of a new cycle of High Atlas / Atlantic rifting during
Mesozoic, Late Miocene compression in relation with the Alpine collision and High Atlas inversion, and
possibly some recent hot spot related thermal uplift responsible for the high present day topography and
relief.
In the studied central and eastern part of Anti-Atlas, the paleozoic serie is thinning gradually to about 6
to 8 km, while more then 12 km are found in the south-west. This thinning affects the behaviour of the
cover. Important rheologic contrasts determine the strutural style. The total shortening of the cover also
decreases gradually, from some 30 km in the SW to a maximum of 10 km in the eastern part. Excess areas
in sections pass from 500 to 250 km2. These two changes (thinning and decreasing shortening) go along
with a gradual change in structural style. The whole of the Anti-Atlas is affected by inversion tectonics in a
thick skin mode. Paleozoic cover series of the southwestern Anti-Atlas are tightly folded into upright
detachement folds in a disharmonic multilayer buckling fashion. Further east, cover folding is more closely
related to the underlying basement inversion in a typical Windriver - Rocky Mountain style [ERSLEV]. An
additional post-folding phase of extension, interpreted as related to the opening of the High Atlas Rift
basin, is more developped in the central and eastern part of the belt, where the High Atlas is immediately
adjacent to the North. The dip of these normal faults varies according to the lithology and produces
Rollover structures, which have often been mistaken as "folds".
Structural sections across the central and eastern parts of the Anti-Atlas are constructed from field
observations, existing geological maps and litterature (regarding the stratigraphy). Field observations
include detailed dip measurements in order to constrain the fold geometry, systematic measurement of
faults and their striation, meso- and micro- structural observations and a detailed mapping of the basement-
cover contact.
The Anti-Atlas of Morocco formed in the hinterland of the larger Appalachian Orogen,
time equivalent with the Alleghenian front to the West. It is limited to the south by the
north-west african craton and to the north by the Alpine High Atlas. The Anti-Atlas
appears as a huge NE-SW anticlinorium. Locally, the basement is cropping out as inliers,
but the main part is made of an important Paleozoïc cover, which is gently folded and of
low grade metamorphism.
An entire Wilson cycle of mountain building left its imprint in this belt. Sedimentation
started with a Late Proterozoïc rifting phase, leading to the opening of first the Rheic and
second the Paleotethys oceans to the west. Post-rift sediments of lower Cambrian to
middle Carboniferous age are dominated by detrial input from the cratonic hinterland of
Africa, leading to the accumulation of a thick pile of alternating quartzites, siltstones,
shales and occasional carbonates. During the last stages of closure of the Paleotethys
ocean, this passive margin is squeezed in the collision between Laurussia and Gondwana.
This variscan collision lead to a massive inversion of former normal faults, substantial
uplift of the Precambrian basement and a concommittant folding of the overlying cover
series. A post-folding phase of extension, interpreted as a new cycle of High Atlas /
Atlantic rifting during Mesozoïc times, is more developed in the central and eastern part
of the belt, where the High Atlas is immediately adjacent to the North. The dip of these
normal faults varies according to the lithology and produces Rollover structures, which
have often been mistaken as variscan folds. Late Miocene compression in relation with
the Alpine collision and High Atlas inversion, and possibly some Neogene hot spot
related thermal uplift are responsible for the high present day rejuvenated topography.
Syn- and post-tectonic erosion and isostatic re-equilibration lead to the removal up to 10
km of overburden, bringing lower green-schist facies rocks back to the earth's surface -
allowing for deep insight into this otherwise mildly deformed fold belt. The Anti-Atlas is
a thick skinned inversion belt, not a thin skinned foreland fold-thrust-belt.
L’ Anti-Atlas du Maroc s’inscrit dans l’orogène Appalachien. Limité au sud par le craton
ouest africain (WAC) et au nord par le Haut-Atlas alpin, l’Anti-Atlas apparaît comme un
vaste anticlinal d’orientation NE-SW. Localement, le socle affleure en boutonnières,
mais la majeure partie de son édifice est faite d’une importante couverture paléozoïque
légèrement plissée et de métamorphisme faible.
Cette chaîne montre un cycle de Wilson entier. La sédimentation débute avec une phase
de rifting au Protérozoïque tardif qui mène à l’ouverture de l’océan Rhéic plus à l’ouest,
mais dont les nombreuses structures de rifting locales sont avortées. Les sédiments post-
rift, du Cambrien au Carbonifère, sont dominés par l’apport détritique du craton. Il en
résulte une épaisse série d’alternances de quartzites, argilites, schistes et de quelques
carbonates. L’environnement des dépôts paléozoïques ainsi que la courbe de subsidence
semblent d’avantage correspondre à un bassin intracratonique qu’à une marge passive,
dont aucun élément profond n’est présent dans l’Anti-Atlas. Durant les derniers stades
de la fermeture de l’océan Paléotéthys, la collision entre les continents Laurussia et
Gondwana entraîne la déformation de ce bassin. Cet événement varisque provoque une
sur-inversion des failles normales, une remontée du socle précambrien et un plissement
polyharmonique de la couverture.
The Anti-Atlas fold belt is formed in the larger Appalachian Orogen system, time equivalent with
the Alleghenian front to the West. Anti-Atlas appears as a huge anticlinorium oriented NE-SW.
Locally, the basement is cropping out as inliers. The important palaeozoic cover is gently folded
and of low grade metamorphism. The lack of major “décollement”, deformation front or thrust fault
makes the Anti-Atlas an unusual type of belt, which does not fit with classic schemes.
The Anti-Atlas is located (Fig.1) on the northern border of the West African Craton (WAC),
limited to the South by the Tindouf basin, tectonically not affected, and to the North by the High-
Atlas (Cenozoic to lower Quaternary deformation of Meso-Cenozoic sediments). The studied area
is the eastern part of the Anti-Atlas which structurally differs strongly from the rest of the chain
further west (Helg et al. 2004).
For this study, the basement is defined in terms of rheological “basement/cover” relationships.
The basement is an assemblage of crystalline, metamorphic and sedimentary rocks with complex
history (Ennih & Liégeois 2001). Note that the Anti-Atlas correspond to the northern limit of the
Eburnian WAC (c. 2 Ga), which replayed in the Neoproterozoic continental breakup and was last
consolidated during the Pan-African orogeny. Late Proterozoic extension produces normal faults
and halfgraben structures filled with coarse clastics which are linked to the basement from a
rheological (variscan) point of view. These normal faults and the WAC limits are important
structures for the comprehension of the variscan phase deformation, since they preconfigure the
later inverted blocks.
The whole palaeozoic cover is dominated by shallow marine mostly fine-grained clastic
sediments. The departure of terranes from the WAC margin, brought the Anti-Atlas basin closer to
the open sea. Detrital input from the WAC dominates sedimentation from Cambrian to Silurian
times. Carbonate sedimentation takes place from the end of Silurian to the end of Devonian. The
late Devonian is marked by a return to detrital sediments that will dominate during all the
Carboniferous. This strong subsidence leads to the accumulation of 12 km in the SW Anti-Atlas, 6
km at the most eastern part. Note that the total thickness and the incompetent/competent ratio are
the two key parameters which will dictate the structural style in the cover.
In Late Carboniferous, the last stages of closure of the Paleotethys ocean (Stampfli & Borel
2002),cause the collision between Laurussia and Gondwana. This compression leads to a massive
inversion of the former normal faults, substantial uplift of the Precambrian basement and a
concommittant folding of the overlying cover series, by the intervention of thick detachment
horizons, in a trishear mode (Erslev 1991). The major variscan trend of the Anti-Atlas is given by
the outcropping inliers as a huge anticlinorium with NE-SW axis. Well developed in the eastern
Anti-Atlas, a regional gradient in deformation style, intensity and orientation is observed from north
(thick skinned inversion style with ENE-WSW fold axis) to south (thin skinned detachment folding
with NW-SE fold axis). This egg-box interference pattern is to be linked to the shape of the WAC.
Some folded folds show that the NE-SW Anti-Atlas global trend has developed before the NW-SE
Ougartian trend. However, these two trends are interpreted as the product of the underlying shape
of the WAC and not as two distinct deformation phases.
The next Wilson cycle begins at the end of Triassic. The new “Mesozoic High-Atlas basin” opens
just north of the Anti-Atlas. It affects the whole belt by dolerites dykes and sills intrusions. Normal
faulting is specially developed in the eastern Anti-Atlas to produce extensional fault-related folding
(hanging-wall synclines) in the palaeozoic cover. The next sedimentary layers which really seal the
variscan deformation are cretaceous conglomerates, no permian to jurassic sediments are found in
the Anti-Atlas.
The actual high topography of the Anti-Atlas (up to 2500 m) does not result only from the
variscan orogeny. Cretaceous to neogene sediments around the eastern Anti-Atlas show some
unconformities and tilts which correspond with the inversion of High-Atlas at Eocene-Oligocene
and Pliocene-Pleistocene epochs (Frizon de Lamotte et al., 2000). A neogene thermal uplift may
also rejuvenate the topography. Except a light doming, no significant deformation affects the
variscan belt.
Figure 1. Geographic situation and geological context of the Anti-Atlas. The northern limit of the WAC is newly defined
by the SAF and the SGA (Ennih & Liégeois 2001). AAMF: Anti-Atlas Major Fault (older WAC limit); SAF: South Atlas
Fault ; SGA: Saoura gravimetric anomaly.
REFERENCES
Ennih, N. & Liégeois, J.-P. (2001): The Moroccan Anti-Atlas: the West African craton passive margin with limited Pan-
African activity. Implications for the northern limit of the craton. Precambrian Research 112: 289-302.
Erslev E.A. (1991): Trishear fault-propagation folding. Geology 19: 617-620.
Frizon de Lamotte D., Saint Bezar B., Bracène R. (2000): The two main steps of the Atlas building and geodynamics of
western Mediterranean. Tectonics 19 (4): 740-761.
Helg U., Burkhard M., Caritg S., Robert-Charrue C. (2004): Folding and inversion tectonics in the Anti-Atlas of
Morocco. Tectonics, 23, TC4006, doi:10.1029/2003TC001576.
Stampfli, G.M. & Borel, G.D. (2002): A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic
plate boundaries and restored synthetic oceanic isochrones. Earth Planet. Sci. Lett. 196: 17-33.
POSTERS
Cretaceous
Mediterrannean Sea
Hammada
Carboniferous
The Anti-Atlas of Morocco is formed in Miocene compression in relation with the Ouarkziz
Rif the hinterland of the larger Appalachian Alpine collision and High Atlas inversion,
Orogen, time equivalent with the and possibly by some Neogene hot spot
Devonian Silurian
Alleghanian front to the West. It is limited related thermal uplift.
las Rich (1-4)
n
At
to the South by the Tindouf Basin and the
cea
Ordovician
Atl
At
huge NE-SW anticlinorium. Locally, the
siltstones, shales and occasional
gh basement is cropping out as inliners, but Bani 1
Hi the main part is made of an important
carbonates. Triassic dolerites testify the
s extension phase of the opening of
l a
Paleozoïc cover, wich is gently folded Grès verts
t Atlantic ocean. The Hammada seals the
A and of low grade metamorphism. The Schistes à Paradoxydes
Ou deformed folds of Anti-Atlas, but is itself
Cambrian
i ga Anti-Atlas system clearly includes the
t rt tilted a few degress to the south. Grès roses
n a Algerian Ougarta chain. Syn- and post-
A
tectonic erosion and isostatic re- The rheology of the paleozoïc cover
equilibration lead to the removal up to 10 dictates the geometry of the final fold Calcaires sup
Lie de Vin
Adoudounian
km of overburden, bringing lower green- chain. This structural style depends on
schist facies rocks back to the Earth’s two key parameters namely the total Calcaires inf
Tindouf Basin
surface - allowing for deep insight into thickness of the cover (decreasing from Conglomérats de Base
this otherwise mildly deformed fold belt. 12 km thick in the west to 6 km eastward) P III
Pre-Cambrian
The high present day topography of Anti- and the ratio of competent and P II
Atlas has been rejuvenated by late incompetent layers
Reguibat Shield Basement
Landsat view
Cross section
Cretaceous-Neogene Hammada
Carboniferous Ouarkziz
Devonian Rich
Upper Cambrian
Middle Cambrian
Adoudounian
P III / Basement
Anti-Atlas
the variscan belt of Morocco IGH
High-Atlas Inversion
PALEOGENE NEOG.
CENOZOIC
W 005°
W 004°
also rejuvenate the topography.
m But except doming, no
Digital Elevation Model from eastern Anti-Atlas and central Anti-Atlas region. significant deformation re-
(Data from GTopo30 @ http://edcdaac.usgs.gov/gtopo30/gtopo30.asp).
affects the variscan belt.
-1000
-2000
intrusions. Normal faulting is especially
-3000
-4000 developed in the eastern Anti-Atlas to produce 5
N S 5
Hanging-wall syncline
extensional fault-related folding (hanging-wall
synclines) in the Palaeozoic cover. The next 0 0
Cambrian, orange : Ordovician, purple : Silurian, green Devonian. Note the -18 km with 15 km shortening including
200
Ultraweak Cover
hanging-wall syncline in the Devonian, produced by normal faulting. a 50 % of layer parallel shortening.
The first calculation of detachment depth gives -33 km : Lower crustal scale detachment. In
fact, this calculation does not take the Layer Parallel Shortening (LPS) into account. An
The particular rheology of the cover
assumed 50 % of LPS is not over estimated and yields a detachment depth at -18 km which
dictates the deformation style. There is no
corresponds to a more reasonable midcrustal detachment.
major “décollement”. The deformation is
absorbed in pureshear style by an
anabundance of incompetent layers.
PERMIAN
PALEOZOIC
and a concommittant folding of the overlying cover series, by the intervention of thick
detachment horizons. The major variscan trend of the Anti-Atlas is given by the outcropping
inliers as a huge anticlinorium with a NE-SW axis. Well developed in the eastern Anti-Atlas, a
Unsuccessful Rift
regional gradient in deformation style, intensity and orientation is observed from north (thick
skinned inversion style with ENE-WSW fold axis) to south (thin skinned detachment folding with
NW-SE fold axis). Some folded folds show that the NE-SW Anti-Atlas global trend has
Intracratonic Basin developed before the NW-SE Ougartian trend. However, these two trends are interpreted as the
product of the underlying shape of the WAC and not as two distinct deformation phases.
The entire Palaeozoic cover is dominated by shallow marine 140505PP/15
0
150505PP/12 031103PP/21
0
041103PP/19
0
051003PP/17
0
081002PP/12
0
closer to the open ocean. Detrital input from the WAC 031002PP/18
290602PP/28
0
0 0
080302PP/9
0
180302PP/13 160303PP/10
0 0
WAC Limits
Silurian
Ordovician
Carbonate Cambrian
Adoudounian
PIII
Basement
W 002°
Shale N 34°
Rabat
le
Oran
Fault (AAMF) was interpreted
W 010°
490 d s
id la
M At
Meseta
as the old front of the african
Volcaniclastics Moroccan
n
s
ea
Meseta
A t
l a shield but is newly seen as
Oc
30° N
N 32° F
g h SA integrated in a metacratonic
tic
H i
an
s
SG
a
A
l
A
tA
A Ou boundary of the WAC is now
N 30° MF ga
i
rta defined by the South Atlas Fault
t
n Algeria (SAF) bording the High-Atlas to
NEOPROTEROZOIC
A
N 28°
Continent the south and by the Saoura 29° N
Tindouf
Basin
Orogens
except Anti-Atlas sytem gravimetric anomaly (SGA) to
Palaeozoïc
008° W
007° W
006° W
005° W
004° W
003° W
Geological map of the eastern Anti-Atlas showing an Egg-Box interference pattern. Colors cool
down up the stratigraphy, from red Precambrian to blue Carboniferous. Post variscan in black.
Anticlines (red) and synclines (blue) indicate fold axis of the superstructure folds.
Eastern Anti-Atlas
The basement is an assemblage of crystalline, metamorphic Ennih N., J.-P. Liégeois (2001), The Moroccan Anti-Atlas: the West African craton passive margin with limited Pan-
African activity. Implications for the northern limit of the craton, Precambrian Research, 112, pp. 289-302.
and sedimentary rocks with a complex history (Ennih & Liégeois Erslev E.A. (1991), Trishear fault-propagation folding, Geology, v. 19, pp. 617-620.
2001). The Anti-Atlas corresponds to the northern limit of the Frizon de Lamotte D., B. Saint Bezar, R. Bracène (2000), The two main steps of the Atlas building and geodynamics of
Eburnian West African Craton (c. 2 Ga), reactivated in the western Mediterranean, Tectonics, Vol. 19, N° 4, pp. 740-761.
Neoproterozoic continental breakup and last consolidated Stampfli , G.M., G.D. Borel (2002), A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic plate
boundaries and restored synthetic oceanic isochrones, Earth Planet. Sci. Lett. , 196, pp 17-33.
during the Pan-African orogeny. Late Proterozoic extension
produces normal faults and halfgraben structures filled with
coarse volcaniclastics which are linked to the basement from a Charles ROBERT-CHARRUE & Martin BURKHARD
rheological (variscan) point of view. These normal faults and the
WAC limits are important structures for the comprehension of Institut de Géologie et d’Hydrogéologie
the variscan phase deformation, since they preconfigure the Université de Neuchâtel - 2007 Neuchâtel
later inverted blocks. [email protected]
Géologie structurale de l’Anti-Atlas oriental
D
FIGURES
30° N
29° N
008° W
007° W
006° W
005° W
004° W
003° W
km km CROSS SECTIONS
4 N 01 S 4 OF
2 2 EASTERN ANTI-ATLAS
0 0 OF MOROCCO
-2 -2
1: 200’000
no vertical exageration
km km
4 N 02 S 4
2 2
0 0
-2 -2
km km
4 N 03 S 4 Neogene
Paleocene
2 2
Cretaceous
0 0 Jurassic
0m Triassic
-2 -2 Late Carboniferous
1000 Early Carboniferous
Devonian
2000
Silurian
km km
3000 Late Ordovician
4 N 04 S 4
4000 Early Ordovician
2 2
5000
0 0 Middle Cambrian
6000 Early Cambrian
-2 -2 10 09 08 07 06 05 04 03 02 01
km km
6
4
N 05 S 6
4
2 2
0 0
-2 -2
-4 -4
km km
06 S
6 6
4 N 4
2 2
0 0
-2 -2
-4 -4
-6 -6
km km
07 S
6 6
4 N 4
2 2
0 0
-2 -2
-4 -4
-6 -6
km km
08 S
8 8
6 N 6
4 4
2 2
0 0
-2 -2
-4 -4
km km
8
6
N 09 S 8
6
4 4
2 2
0 0
-2 -2
-4 -4
-6 -6
km km
8
6
N 10 S 8
6
4 4
2 2
0 0
-2 -2
-4 -4
-6 -6
0 0 0 0 0
0
0
N 31°30’
0 0
0
0 0 0
0
0 0
0 0
N 31°00’
0
0
0
N 30°30’
W 4°30’
W 4°00’
W 5°00’
W 6°00’
W 5°30’
W 6°30’
N 30°00’
N 32°
N 31°
4000
3500
3000
2500
2000
1500 N 30°
1000
W 005°
W 003°
W 006°
W 004°
W 007°
500
m
006° W
50 km
005° W
004° W
31° N
30° N
1500
N S TOPOGRAPHIE ET GEOLOGIE
1000
DE SURFACE ET SUBSURFACE
500 DE L’ANTI-ATLAS ORIENTAL
01
T
0m
AL
Néogène
Echelle verticale 1 : 55’000
IL
1500 Paléogène
Echelle horizontale 1 : 550’000
F
02
TA
Crétacé
1000
Exagération verticale 10 x
Jurassique
500 Trias
0m 03 Carbonifère sup
Carbonifère inf
Dévonien
2000 04 Silurien
Ordovicien sup / ND
E
AT
1500 Ordovicien inf
05
ER
1000 Cambrien sup / ND
G
U Cambrien inf
AD
E
500
O
HAUT-ATLAS
IN
Précambrien
06
AF
M
0m
SD
Faille normale
TI
Faille chevauchante
2000
07
Topographie due à
l’amincissement
1500 crustal selon
Missenard et al. 2006
1000
500
08
O
GRABEN DE ZAGORA
0m
R
H
G
2500
SA
OUARZAZATE
2000
2em BANI
E
09
IN
BASSIN DE
1500
R
1000 ZA
Z
TA
500
1er BANI
0m 10
0 km 20 40 60 80 100 120 140 160 180
Post Varisque Axe anticlinal
Carbonifère Axe synclinal
Devonien
Anti-Atlas
Silurien
Ougarta
Ordovicien
Cambrien
Adoudounien
AA-Ouarzazate
Socle
31° N
30° N
29° N
008° W
007° W
006° W
005° W
004° W
003° W
Géologie structurale de l’Anti-Atlas oriental
E
ATLAS LANDSAT 7
NW Africa
Anti-Atlas
HOGGAR
100 km
Répartition des vues
006° W
004° W
005° W
31° N
50 km
30° N
Légendes
Carbonifère Cambrien moyen
Palmeraie / Dune
Ouarkziz
Cénozoïque Dévonien
Cambrien inférieur
Rich
Ordovicien sup
Crétacé è m
e Adoudounien
2 Bani