These RobertCharrueC

Institut de Gologie et dHydrogologie
Universit de Neuchtel
Suisse
Gologie structurale de
lAnti-Atlas oriental,
Maroc
Thse de doctorat prsente la Facult des Sciences
Universit de Neuchtel
Pour lobtention du grade de docteur s sciences
par
Charles Robert-Charrue
- Soutenue le 19 septembre 2006 -
Accepte sur proposition du jury :

Prof. Karl Fllmi, Universit de Neuchtel, Directeur de thse remplaant
Prof. Thierry Adatte, Universit de Neuchtel
Prof. Dominique Frizon de Lamotte, Universit de Cergy-Pontoise, France
Prof. Abderrhamane Soulamani, Universit Cadi-Ayyad, Marrakech, Maroc
Dr. Mahmoud Zizi, ONHYM, Maroc
Gologie structurale de lAnti-Atlas oriental
Thse de doctorat
II
Thse de doctorat
CE TRAVAIL EST DDI LA MEMOIRE DE
MARTIN BURKHARD
QUI NOUS A QUITT PRMATURMENT EN AOT 2006.
NOUS GARDERONS LE SOUVENIR

DUN HOMME DE SCIENCE AU SERVICE DE LA SOCIT
DUN PDAGOGUE TOUJOURS DISPONIBLE
DUN HOMME SOURIANT LA VIE
DUN AMBASSADEUR DE LA TERRE
Thse de doctorat
III
IV
Thse de doctorat
RESUME
Mots-cls : Anti-Atlas, Maroc ; Tectonique dinversion ; Orogne varisque ; Palozoque ;
Interfrence de plis ; Profondeur de dtachement ; Dformation interne ; Plis associs aux failles
extensives.
La chane de lAnti-Atlas du Maroc est situe sur la bordure nord du craton ouest
africain. Sa forme gnrale est un vaste anticlinal dorientation ENE-WSW, form
en son cur de boutonnires de socle et sur ses flancs dune couverture
plisse. Le socle Prcambrien recle les traces des orognes Eburnen (2 Ga)
et Pan Africain (600 Ma). Ds la fin du Noprotrozoque, un rift initie un
nouveau cycle de Wilson. A lissue de la sdimentation syn-rift, le rift est avort,
mais une subsidence rgulire permet le dpt dune paisse couverture
Palozoque dans un environnement de mer peu profonde. Ce bassin
intracontinental est invers au Carbonifre moyen, crant ainsi la chane plisse
de lAnti-Atlas. La particularit de cette chane est labsence de dcollement
majeur.
Les tudes gologiques menes dans lAnti-Atlas oriental ont permis de dcrire
sa structure et de caractriser son style structural. Les deux paramtres qui
contrlent le type de la dformation sont (1) la pr-structuration du socle lors du
rift Noprotrozoque Cambrien et (2) la rhologie particulirement
incomptente de la couverture. La compression varisque inverse les anciennes
failles normales du socle. A grande chelle, la couverture accommode la
dformation par drapage, mais elle est galement affecte par un
raccourcissement interne qui se traduit par la formation de petites structures
compressives, visibles dans les quelques formations plus comptentes qui
jalonnent la srie Palozoque. Lanalyse des structures a permis de discerner
deux orientations distinctes de dformation. Lune correspond lorientation
gnrale de lAnti-Atlas, lautre adopte lorientation de la chane de lOugarta
situe au sud-est. La superposition de ces deux orientations provoque une figure
dinterfrence en dmes et bassins.
Dix coupes dcrivent la structure gnrale de lAnti-Atlas oriental. La restauration
diffrents stades de lune dentre elle, a permis de calculer une profondeur de
dtachement mso-crustal de -18 -20 km. Une nouvelle mthode
dquilibration intgrant le raccourcissement interne a t dveloppe et a
permis de mettre en vidence un raccourcissement interne de 11 17%.
Une extension tardive attribue louverture du bassin Msozoque du Haut
Atlas (au nord de lAnti-Atlas oriental) ractive les structures profondes et cre
dans la couverture des failles normales raides et des plis associs.
Les relations avec les sdiments post-varisques montrent que linversion du Haut
Atlas ne provoque pas de dformation mais une lgre surrection dans lAntiAtlas oriental. Ce soulvement combin celui induit par une anomalie
thermique rcente explique la haute topographie de lAnti-Atlas oriental.
Thse de doctorat
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 AntiAtlas fold belt. The main feature of this chain is the lack of a major dcollement
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.
Ten cross-sections describe the structure of the eastern Anti-Atlas. The
restoration at different stages of one of these cross-sections, has allowed for the
calculation of a mid-crustal depth to detachment of 18 to 20 km. A new
balancing method, integrating bed-length shortening has been developed. This
indicated an internal shortening of 11 to 17 %.
A later extension, attributed to the opening of the Mesozoic High Atlas basin,
reactivated the deep structures, creating normal faults and extensional faultrelated 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.
VI
Thse de doctorat
SOMMAIRE
Imprimatur
Ddicace
Rsum
Abstract
Sommaire
Sommaire des annexes
1.
2.
3.
4.
5.
6.
I
III
V
VI
VII
VIII
Introduction
1.1.
Le projet
1.2.
Situation gographique et bases gologiques
1.3.
Buts de ltude
1.4.
Mthodes, outils, donnes
1.5.
Logistique et approche
1
2
3
5
6
Cadre gologique Etat des connaissances

2.1.
Tectonique globale
2.2.
Elments morpho-gologiques de lAnti-Atlas oriental
2.3.
Le socle
2.4.
La couverture palozoque
2.5.
Lorogne varisque
2.6.
Les vnements post-varisques
9
11
14
16
22
22
Structures Description des observations

3.1.
Introduction
3.2.
Stratigraphie mcanique
3.3.
Structure gnrale Coupes
3.4.
Structures compressives
3.5.
Structures extensives
3.6.
Relations avec les roches rcentes et topographie
25
27
31
35
44
49
Modle Gomtrique Interprtations des observations

4.1.
Style structural
4.2.
Restaurations et calcul de la profondeur de dtachement
4.3.
Interfrences et gradient de dformation
4.4.
Structures dextension tardives
4.5.
Relations Anti-Atlas roches post-varisques
4.6.
Interprtations des donnes sismiques
53
56
66
69
71
73
Modle Dynamique Chronologie des vnements

5.1.
Structuration Prcambrienne
5.2.
Bassin Palozoque
5.3.
Inversion et interfrence varisques
5.4.
Extension Msozoque
5.5.
Rjuvnation Cnozoque
77
79
79
81
81
Conclusions Apports de ltude

6.1.
Style structural
6.2.
Structures
6.3.
Tectonique
6.4.
Implications et perspectives
83
84
85
85
Rfrences
87
Table des Figures, Tableaux et Graphiques
95
Remerciements
97
Thse de doctorat
VII
SOMMAIRE DES ANNEXES

A. Donnes brutes de la Figure 22
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 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.
- Projet dun troisime article en premier auteur.
C. Rsums et posters de confrences et de congrs
- Late Variscan Inversion Tectonics in the Paleozoic Tafilalt basin of South-Eastern Morocco.
Abstract. MIOG 2002.
- Inversion tectonics in the hercynian Anti-Atlas of Morocco. Abstract. SGM 2003.
- The Anti-Atlas fold-belt of Morocco : Thick skin inversion tectonics in the hinterland of the
alleghenian orogeny. Abstract. RST 2004.
- Tectonique dinversion dans lAnti-Atlas oriental. Abstract. 3Ma 2005.
- The Anti-Atlas fold belt of Morocco : variscan inversion tectonics and interference pattern of an
intracratonic basin. Abstract. SGM 2005.
- The Anti-Atlas fold-belt of Morocco. Poster. RST 2004.
- 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 : dcoupage et lgendes de la mosaque
- Mosaque de lAnti-Atlas oriental
NB :
VIII
Les figures marques dun astrisque dans le texte, sont de taille rduite. Elles sont
prsentes dans lannexe D dans un format plus grand.
Thse de doctorat
1
INTRODUCTION
1.1. Le projet
Ltude structurale de lAnti-Atlas oriental sinsre dans le projet du Professeur
Martin Burkhard de lUniversit de Neuchtel. Le projet dbute en 1997 avec
Sverine Caritg et Urs Helg, deux doctorants qui concentreront leurs tudes sur
la partie sud-occidentale de la chane. De ces tudes dcoulent deux articles
principaux : Helg U., M. Burkhard, S. Caritg, C. Robert-Charrue (2004), Folding
and inversion tectonics in the Anti-Atlas of Morocco, Tectonics, Vol. 23 et Caritg
S., M. Burkhard, R. Ducommun, U. Helg, L. Kopp, C. Sue (2004), Fold
interference patterns in the Late Palaeozoic Anti-Atlas belt of Morocco, Terra
Nova, 16. En 1999, trois diplmants (Romain Ducommun, Lionel Kopp et moimme) prennent part au projet. Tandis que Romain Ducommun et Lionel Kopp
sattaquent aux plis dinterfrence de la rgion de Tata dans le sud-ouest, mes
tudes me mnent dans le Tafilalt. Ce choix est influenc par la possession de
trois lignes sismiques leves dans cette rgion par lactuel ONHYM (Office
National des Hydrocarbures et des Mines) et interprtes par A. Bally et M. Zizi.
Une partie des donnes rcoltes dans cette extrmit de la chane, servent
lcriture de Robert-Charrue & Burkhard (2006), Variscan inversion tectonics and
interference pattern in the Tafilalt, Anti-Atlas, Morocco soumis au Journal of
African Earth Sciences. A lissue de ce travail de diplme, sur proposition de
Martin Burkhard, mes recherches sont tendues lAnti-Atlas oriental dans le
cadre dun travail de doctorat. Larticle : Tectonics of Anti-Atlas of Morocco
Thse de doctorat
1. Introduction
(2006) de M. Burkard, S. Caritg, U. Helg, Ch. Robert-Charrue et A. Soulamani

trate en particulier du bassin Palozoque et de la tectonique de linversion
varisque. Il parat dans un numro des Comptes Rendus Geoscience consacr
la gologie de lAfrique du Nord avec comme rdacteurs invits D. Frizon de
Lamotte, A. Michard et O. Saddiqi.
1.2. Situation gographique et bases gologiques

Au Nord-ouest du continent africain, lAnti-Atlas constitue le domaine austral du
Royaume du Maroc. Comme le suggre limage satellite (Figure 1 gauche), ces
rgions arides et semi-dsertiques prsentent une qualit daffleurement
exceptionnelle.
Figure 1. Situation de lAnti-Atlas du Maroc. Gauche : mosaque de vue satellite. Droite : carte
gologique simplifie.
La chane de lAnti-Atlas se prsente sous la forme dun vaste bombement

anticlinal comportant en son cur des affleurements de socle appels
boutonnires, elles-mme bordes par une couverture sdimentaire palozoque
plisse (Figure 1 droite). Laxe de cet anticlinal suit une orientation ENE-WSW.
Thse de doctorat
1. Introduction
Figure 2. Carte gnrale du Craton Ouest Africain. Daprs Dallmeyer et Lcorch (1991) cit
dans Piqu (2001).
Au sud-ouest lAnti-Atlas se prolonge par la chane du Zemmour et les

Mauritanides avec une orientation NE-SW. A lest, il est relay par la chane de
lOugarta dorientation NW-SE. Ces diffrentes chanes dessinent un arc qui
constitue la bordure septentrionale du craton ouest africain (WAC, West-African
craton) (Figure 2). Au sud le bassin de Tindouf est constitu de la mme
couverture palozoque que lAnti-Atlas la seule diffrence quelle nest point
dforme. Au nord, lAnti-Atlas est limit aujourdhui par le Haut-Atlas, un bassin
msozoque invers lors de lorogne alpin. Lhistoire de la Meseta (Figure1) est
contemporaine de celle de lAnti-Atlas, mais non identique. Sa dformation a eu
lieu dans une situation beaucoup plus interne de lorogne varisque.
1.3. Buts de ltude

Les tudes publies sur la dformation varisque de lAnti-Atlas oriental sont rares
(A. Piqu dans son ouvrage sur la gologie du nord-ouest de lAfrique (2001)).
Thse de doctorat
1. Introduction
Le projet dans sa globalit vise dcrire la structure de cette chane plisse et

plus particulirement caractriser les paramtres contrlant la dformation
varisque en terme de style tectonique. Une question fondamentale est celle de la
relation entre le socle et la couverture. La dformation est-elle de type thick-skin
ou thin-skin ? En dautres termes, la dformation affecte-t-elle autant la
couverture que le socle ou la couverture se dsolidarise-t-elle du socle par un
dcollement basal comme on peut lobserver dans bien des chanes davantpays typiques ? A premire vue, labsence de front ou de chevauchement
majeur font de cette chane un objet dtude singulier.
Quelques critres de terrains fourniront quelques lments de rponse la
question pose en dbut de projet, de la cause de la topographie tonnamment
leve et du relief extrmement jeune de lAnti-Atlas. Mais les travaux
spcifiques cette problmatique de Frizon de Lamotte et de son quipe, sur la
base de donnes gophysiques fourniront une rponse plus taye (Missenard
et al., 2006 ; Saddiqi et al., 2005).
Les premires tudes faites dans le sud-ouest se sont davantage focalises sur
les plis et les relations entre les longueurs donde diffrentes observes dans les
bancs marqueurs de la couverture. Cette dformation de type flambage
(buckling), o chaque banc se plisse indpendamment selon ces propres
caractristiques, nest pas observe grande chelle dans lAnti-Atlas oriental
o la dformation semble affecter lensemble de la couverture dune manire
homogne. Un des buts est donc de caractriser cette diffrence. LAnti-Atlas
oriental se poursuit au SE par la chane de lOugarta (Figures 1 et 2), il diffre
galement par la prsence de structures extensives qui naffectent pas la partie
occidentale. Linterfrence de ces diffrents traits gologiques constitue une clef
dans la comprhension de ldifice. Les objectifs pour cette tude sont de fournir
une image de la gomtrie globale de lAnti-Atlas oriental, den dfinir le style
structural et den quantifier la dformation. Lambition ultime tant de proposer
un modle dynamique expliquant la chronologie et les paramtres contrlants la
mise en place de cette partie de la chane.
Thse de doctorat
1. Introduction
1.4. Mthodes, outils, donnes

Etant donns la grande taille de lobjet et ltat des connaissances, limit en
terme de structures, les mthodes visant dgrossir la structure de lAntiAtlas oriental sont lmentaires.
Loutil principal est la cartographie. Pratiquement, la boussole permet la mesure
de lorientation des plans de stratification, des plans de failles et des stries. Ces
mesures servent dessiner les coupes, calculer lorientation des axes de plis et
dduire lorientation des contraintes qui ont gnr ces dformations (plis et
failles). La boussole utilise est une Brunton GEO Transit. La dclinaison
magntique utilise provient des cartes du British Geological Survey [1]. Le
positionnement gographique se fait laide des cartes topographiques au
1 :100'000 du Maroc et dun GPS Garmin III. Le systme godsique choisi est
le WGS84. Les coordonnes sont sous la forme longitude/latitude en degr,
minute, millime de minute (000 00.000). Un millime de minute pour cette
rgion correspond environ 0.9 m en latitude et 0.7 m en longitude. Les
mesures dpaisseur sont calcules cartographiquement par le dessin de coupes
locales gnralement au 1 :25'000. Malgr les lacunes de donnes structurales,
les cartes gologiques au 1 :200'000 sont un outil prcieux, elles fournissent
dexcellentes descriptions stratigraphiques.
Les images satellites sont une source dinformation inestimable vu la qualit
daffleurement de ces rgions. Les images utilises sont celles de la gnration
Landsat 7 ETM+ (Enhanced Thematic Mapper). Elles sont traites partir de 3
bandes de frquence. Elles sont orthorectifes, gorfrences et ont une
rsolution de 14.25 m. Ces vues sont tlchargeables gratuitement depuis
internet [2] et peuvent tre exportes en image (.tif) laide du programme
GeoViewer de Lizardtech.
Les modles numriques de terrain (MNT) proviennent du modle GTOPO 30
(Global 30 Arc Second Elevation Data) qui a une maille dapproximativement 1
Thse de doctorat
1. Introduction
km2. Ces donnes sont gratuites et disponibles sur internet [3]. Le programme
Surfer 8.01 de Golden Software, Inc. permet la gnration de cartes et la
cration de vues tridimensionnelles.
Trois lignes sismiques fournies par lONAREP (actuellement ONHYM) ont permis
de mieux contraindre le modle gomtrique en profondeur. Une premire
interprtation de la ligne RS 8 a t ralise dans le cadre du travail de diplme.
Cette interprtation a t revue et corrige pour le premier article.
1.5. Logistique et approche

Cette tude a ncessit plusieurs voyages au Maroc :
Printemps 2002
-Terrain, Anti-Atlas oriental du 7 au 29.03.

-Anti-Atlas central avec Martin Burkhard, 30.03 - 6.04.
-Congrs Marrakech International Oil & Gas , 7 - 10.04.
Automne 2002
-Terrain, Anti-Atlas oriental du 22.09 au 12.10.
Printemps 2003
Automne 2003
Hiver 2003-2004
-Anti-Atlas occidental avec M. Burkhard, S. Caritg et Zeshan

Ismat (Assistant Professor au Franklin & Marshall College,
Lancaster PA, USA) 30.12.2003-11.01.2004.
Printemps 2004
-Excursion de lInstitut de Gologie, UniNe, 5 - 21.04.
Printemps 2005
-Congrs 3Ma (Magmatisme, Mtamorphisme et

Minralisations associs), Agadir, 5 - 7.05.
-Excursion du congrs 3Ma dans lAnti-Atlas occidental et
central, 8 - 11.05.
-Terrain, Anti-Atlas oriental du 12 au 18.05.
Les autorisations de mission de terrain ont t dlivres par la Division de la

Gologie du ministre de lEnergie et des Mines Rabat.
Thse de doctorat
1. Introduction
Le travail de terrain a t effectu seul. La carte daccs (Figure 3) montre les

endroits visits lors des diffrentes sessions de terrain dans lAnti-Atlas oriental.
A lexception de la dernire session, les dplacements pour le travail de terrain
se sont faits en Landrover Defender amnag, permettant une autonomie de
plusieurs jours, afin dviter des trajets inutiles.
Figure 3. Carte daccs des tudes menes dans lAnti-Atlas oriental.
Lapproche du terrain a t dicte par les recherches. Outre les mesures de la

gomtrie gnrale, on a, dans un premier temps, cherch plus particulirement
observer le contact socle - couverture et dfinir lpaisseur des formations
dans les diffrentes rgions de lAnti-Atlas oriental. Des profils de coupes ont t
tracs perpendiculairement laxe des boutonnires, selon les variations de la
structure gnrale et les possibilits daccs. Le terrain est trs facilement
abordable par les routes nord (Erfoud-Tinejdad-Tineghir-Boumalne) et centrale
Thse de doctorat
1. Introduction
(Rissani-Alnif-Tazzarine) (Figure 3). Les possibilits de traverses des massifs

du Saghro et de lOugnate ainsi que les explorations vers le sud sont limites
de rares pistes parfois abandonnes. Lobservation et la recherche de structures
plus petites (invisibles sur les cartes et les images satellites) a t
systmatiques.
Thse de doctorat
2
CADRE GEOLOGIQUE
Etat des connaissances
2.1. Tectonique globale
LAnti-Atlas est situ sur la bordure nord du craton ouest africain (Figure 2, p. 3).
Cette zone est particulire dans le sens que de nombreux vnements
tectoniques affectent cette rgion au cours des temps gologiques. Sur la base
des compilations palogographiques de Stampfli et Borel (2002) (Figure 4),
voici un aperu des mouvements tectoniques globaux prcdents et suivants
lpisode varisque.
A lissue de lorogense Pan-Africaine, la marge nord du Gondwana voit le
dtachement successif de plusieurs terrains et donc louverture de diffrents
ocans. Aprs locan Prototthys, les terrains Avaloniens quittent le Gondwana
et font place locan Rhic (stade 490 Ma de la Figure 4). Durant le Silurien, de
lautre ct de locan, la collision des terrains Avaloniens, du continent Baltica
et de Laurentia (craton nord amricain) cre le continent Laurussia (440 Ma).
Durant cette mme priode, le dpart des terrains Huniques de la marge du
Gondwana, font place locan Palotthys (400 Ma). Locan Rhic se referme
alors sur la marge du continent Laurussia (340 Ma), puis le Palotthys se
referme son tour (320 Ma). La collision des continents Gondwana et Laurussia
forment le super-continent Pange (280 Ma). Ces deux derniers stades
correspondent la dformation varisque tudie dans lAnti-Atlas. A la fin du
Thse de doctorat
2. Cadre gologique
Trias et durant le Jurassique (200 Ma et 180 Ma), la Pange est disloque par
louverture de locan Atlantique.
Les vnements dcrits ci-dessus ne constituent quun seul cycle de Wilson (rift
- ouverture ocanique - fermeture du bassin - collision - rift). Dautres cycles ont
prcd et suivent ce cycle varisque.
Figure 4. Palogographies de Stampfli et Borel (2002). Ltoile indique la position de lAntiAtlas.
10
Thse de doctorat
2. Cadre gologique
Si lAnti-Atlas est toujours proximit des limites actives, il faut noter quil ne se
situe jamais sur ces limites. Les sries plus marines correspondent des
niveaux marins levs. La forme du bassin Anti-Atlas peut tre reprsent par un
indentation marine ou un golfe. Toujours sur le continent Gondwana, cette zone
enregistre plus ou moins intensment les vnements tectoniques, mais ne les
subit pas de manire directe.
2.2. Elments morpho-gologiques de lAnti-Atlas oriental

La carte de la Figure 5 reprsente les grandes formes gologiques de lAntiAtlas oriental.
Figure 5. Elments gologiques de lAnti-Atlas oriental. Daprs la Carte Gologique du Maroc

au 1 : 1'000'000 (1985).
Le socle apparat sous la forme de boutonnires. Douest en est : la boutonnire

de Bou Azzer (Anti-Atlas central), celle du Saghro et celle de lOugnate. Avec
son orientation NNW-SSE, la boutonnire de Bou Azzer dvie de lorientation
gnrale (NE-SW) des boutonnires de lAnti-Atlas (Figure 1, p.2). Le plissement
de la couverture palozoque est mis en valeur par la formation du Bani, crte
Thse de doctorat
11
2. Cadre gologique
ordovicienne qui peut tre suivie dans tout lAnti-Atlas. Limage grande chelle
du Palozoque montre une srie de dmes et bassins. Douest en est : la
cuvette de Tazzarine, les bassins du Mader et du Tafilalt. Cette structure
concorde avec celle de la chane de lOugarta dorientation NW-SE. De grandes
structures dextension tel que le graben de Zagora apparaissent dans les
niveaux rigides de la couverture. Le tout est scell par des sdiments du Crtac
ou plus rcents. On distingue le plateau Crtac des Monts Kem-Kem au sudest, la Hamada du Guir Mio-Pliocne lest, le Crtac du nord de lOugnate et
le bassin Cnozoque de Ouarzazate sur la bordure NW du Saghro. Au nordouest de lAnti-Atlas, le Haut Atlas est limit au sud par la faille sud atlasique.
Ces formes gomorphologiques sont visibles sur les vues tridimensionnelles de
la Figure 6.
6 A : Vue vers le nord
12
Thse de doctorat
2. Cadre gologique
6 B : Vue vers louest
6 C : Vue vers le sud
Thse de doctorat
13
2. Cadre gologique
6 D : Vue vers lest
Figure 6. Vues tridimensionnelles de lAnti-Atlas oriental. Modle numrique de terrain :

GTOPO30 avec plaquage de limage satellite Landsat 7.
2.3. Le socle
Avec une impressionnante varit de roches, le socle est le tmoin dune activit
gologique multiphase qui explique sa complexit. Dun point de vue
varisque , le socle comprend toutes les roches antrieures aux sdiments
dposs lors du rift Prcambrien (dbut du cycle de Wilson varisque). Il nen va
pas toute fait de mme en terme de rhologie et de dformation.
Pour les chercheurs sintressant au Prcambrien , lorogense Anti-Atlasique
dsigne celle du cycle Pan-Africain. Les roches du socle de la prsente tude
sont subdivises en trois parties (Thomas et al., 2004 ; Gasquet et al., 2005) : (1)
Le socle burnen
14
Paloprotrozoque (~2 Ga) compos principalement de
Thse de doctorat
2. Cadre gologique
granitodes intrusifs dans des sries mtamorphiques complexes, (2) le supergroupe de lAnti-Atlas comprend des roches volcano-sdimentaires, ophiolitiques
et intrusives impliques dans laccrtion Pan-Africaine au Noprotrozoque, et
(3) le super-groupe de Ouarzazate est compos de roches volcaniques,
intrusives (Barbey et al., 2004) et dtritiques dans un contexte fini- postcollisionnel. Des sdiments syn-rift indiquent lextension Post-Panafricaine
(Piqu et al., 1999 ; Piqu, 2003 ; Soulamani et al., 2003).
Figure 7. Limites cratoniques de lAnti-Atlas. SAF : South Atlas Fault ( Faille sud atlasique),
AAMF : Anti-Atlas Major Fault (Accident Majeur), SGA : Saoura gravimetric anomaly (Anomalie
gravimtrique du Saoura).
Suivant le linament du Tibesti (Guiraud et al., 2000), lAccident Majeur

(Choubert et Faure-Muret, 1971) (AAMF, Figure 7) est certainement la plus
importante trace de lorogne Pan-Africain dans lAnti-Atlas. Anciennement
considre comme la suture dlimitant le nord du craton ouest africain
Thse de doctorat
15
2. Cadre gologique
(Saquaque et al., 1989), cette structure est aujourdhui dcrite comme la limite
australe dun bassin aulacogne du Noprotrozoque infrieur (Hefferan et al.,
2000 ; Ennih & Ligeois, 2001; 2003). Les ophiolities trouves sur cet accident,
sont en fait charries sur le craton (et le bassin aulacogne) lors de laccrtion
Pan-Africaine dun arc volcanique. Ainsi la limite septentrionale du craton ouestafricain est la faille sud-atlasique (SAF). La zone intermdiaire peut tre qualifie
de mtacratonique (Ligeois et al., 2005). La limite orientale de ce mtacraton
semble concider avec lanomalie gravimtrique du Saoura (SGA, Figure 7)
(Bayer et Lesquer, 1978 cits dans Ennih et Ligeois, 2001).
La nouvelle phase de rift du Noprotrozoque terminal utilise peut-tre
danciennes structures, mais cre
en tous cas dimportantes htrognits
dans
la
ce
socle
varisque
sous
forme
de
failles
normales.
Cette
compartimentalisation ou prstructuration jouera un rle essentiel lors de la

dformation varisque.
2.4. La couverture Palozoque

La srie sdimentaire Palozoque de lAnti-Atlas est impressionnante par sa
puissance. Elle atteint plus de 10 km dans le sud-ouest. Dans la partie orientale,
lpaisseur est de 8 km louest et dcrot jusqu 4 km lest (Figure 8). Ceci
est d la rduction des formations elles-mmes mais galement la disparition
de certains niveaux (Figure 9). La puissance de la couverture dcrot galement
vers le nord. La mesure des paisseurs sy complique par la prsence du Haut
Atlas et de roches sdimentaires rcentes. Sur la bordure nord-ouest de la
boutonnire du Saghro, les sdiments Palozoques sont masqus par le bassin
Cnozoque de Ouarzazate (Figure 5). Au nord de lOugnate, ce sont les roches
Crtaces qui recouvrent une partie du Palozoque et mme des pointements
de socle comme le suggrent de petits affleurements de Noprotrozoque dans
la partie nord-est du Tafilalt.
16
Thse de doctorat
2. Cadre gologique
Figure 8. Situation de lAnti-Atlas par rapport la chane des Appalaches la fin du Palozoque.
Les contours isopaques (en km) sont donnes pour les bassins sdimentaires palozoques qui
nont subi que peu ou pas de dformation. Le mme code couleur est appliqu la chane
plisse de lAnti-Atlas afin dindiquer la profondeur estime de ce bassin avant son inversion. Les
massifs de socle allghaniens sont indiqus en bleu. Les parties internes, mtamorphiques et
plus anciennes, de la chane des Appalaches-Mauritanides-Meseta marocaine sont colores en
vert et en rose.
Toute la srie Palozoque est domine par des sdiments dposs dans un
environnement de mer peu profonde. Labsence de sdiment de bassin, de
talus ou de plate-forme externe exclut une interprtation de marge passive.
Lenvironnement de dpt de lAnti-Atlas sapparente plutt un bassin de type
intracontinental (Burkhard et al., 2006), d une subsidence thermique post-rift.
Mais le rgime thermique, la structuration pr-rift et les effets dune compression
lointaine qui contrlent la formation du bassin (Cloething et al., 1995 ; Ziegler et
Cloething, 2004) sont mal contraints. Le bassin Palozoque de lAnti-Atlas est la
partie tectonise du gigantesque bassin de Tindouf (Coward et Ries, 2003).
Thse de doctorat
17
2. Cadre gologique
Figure 9. Formations et colonne stratigraphique de lAnti-Atlas. Colonne de droite est dessine

daprs Soulamani et al., 1997. Elle est typique de lAnti-Atlas occidental.
Cambrien
La srie de lAdoudounien (Buggish et Flgel, 1988 ; Algouti et al., 2001) se
dpose aprs les Conglomrats de base couronnant le super-groupe de
Ouarzazate. LAdoudounien comprend les Calcaires infrieurs, la srie Lie de vin
et les Calcaires suprieurs (Figure 9). Dans le sud-ouest, des signes du rifting
amorc au Noprotrozoque terminal persistent jusqu la fin du Cambrien
infrieur (Benssaou et Hamoumi, 2003). Cette srie est fortement rduite dans
18
Thse de doctorat
2. Cadre gologique
lAnti-Atlas oriental et mme absente son extrmit la plus orientale (Figure 5,

p.11). Ensuite, une srie de schistes argileux se termine par le niveau marqueur
des Grs roses. Le sommet des Grs roses marque la fin du Cambrien infrieur,
absent sous cette forme lest de lOugnate. Les premiers sdiments prsents
sur lensemble du domaine anti-atlasique sont les Schistes Paradoxides
(trilobites) et les Grs verts, ensemble correspondant au Cambrien moyen. Tout
lest, les grs trouvs la base du Cambrien moyen sont attribus au
Cambrien infrieur et non au super-groupe de Ouarzazate. Le hiatus du
Cambrien suprieur correspond la phase de rift qui spare les terrains
avaloniens du Gondwana par louverture de locan Rhic.
Ordovicien
LOrdovicien demeure toujours tendance dtritique. LOrdovicien infrieur
dbute par une puissante srie de schistes argileux et se termine par les grs et
quartzites du 1er Bani. LOrdovicien suprieur ressemble beaucoup linfrieur,
mais dune plus petite paisseur. Le niveau marqueur qui le couronne est le 2me
Bani. Il comprend des microconglomrats et des surfaces rodes tmoignant
dune glaciation (Choubert et Faure-Muret, 1983).
Silurien
Tectoniquement, cette priode correspond louverture de locan Proto- et
Palotthys et la drive des terrains Huniques. Aprs la dominance du
dtritisme provenant du craton africain, le Silurien marque le passage une
sdimentation carbonate. Ce changement est expliqu par une hausse du
niveau marin qui commence par dposer une paisse srie de black shales,
constituant une importante roche mre en Afrique du nord (MacGregor, 1996 ;
Boote et al., 1998). Sensuit lapparition de bancs calcaires qui vont dominer
durant le Dvonien.
Dvonien
Durant cette priode, la marge du Gondwana bnficie dun calme tectonique.
Les Richs Dvoniens tmoignent dun niveau marin toujours haut. La
composante calcaire est plus dveloppe dans la partie orientale. Les calcaires
Thse de doctorat
19
2. Cadre gologique
gris-bleuts noirs terminent la formation des Richs. Ces niveaux condenss

abritent une riche faune palontologique de goniatites et dorthocres
notamment. Ceci tmoigne dune lvation du niveau marin et dun
environnement marin plus ouvert.
Carbonifre
Les sdiments Carbonifres tmoignent dun retour marqu du dtritisme. Aprs
une paisse srie de schistes argileux, la tendance devient grseuse. Des
figures de chenaux, des marques de courant et des bioturbations dcrivent un
environnement deltaque. On trouve dans la partie nord de lAnti-Atlas oriental
des figures sdimentaires interprts comme des tempestites. Les calcaires
couronnant la srie dans le sud-ouest sont absents dans la partie orientale.
La Figure 10 est une charte chronostratigraphique. Elle montre linairement la

rpartition des dpts sdimentaires selon le temps (et non selon lpaisseur).
Elle met en vidence les priodes de dpt et de hiatus. Attention toutefois car
les trous (hiatus apparent) peuvent galement provenir de la tectonique par
lrosion de structures.
20
Thse de doctorat
2. Cadre gologique
Figure 10. Schma chronostratigraphique de lAnti-Atlas. Compilation daprs les cartes

gologiques au 1 : 200'000.
Thse de doctorat
21
2. Cadre gologique
2.5. Lorogne Varisque

La chane de lAnti-Atlas est une partie externe de lorogne varisque
(Appalaches-Ouachita-Mauritanides) (Figure 8, p.17). En terme de style
tectonique, de chronologie et de godynamique, les relations entre la chane de
lAnti-Atlas et les parties internes de cet orogne restent lucider.
La compression varisque affecte lAnti-Atlas vraisemblablement au Carbonifre
suprieur. La datation relative de cet vnement est difficile contraindre
puisque lorogne varisque marque le dbut dune longue priode de non
sdimentation. Les prochaines roches sceller cette dformation, par
recoupement de structures, sont des dolrites. Elles apparaissent principalement
sous la forme de dykes. Mais elles se dveloppent galement en sills, ce qui
peut tre droutant, car ils pousent la forme des plis. Lge de mise en place de
ces dolrites est dat la limite Trias Jurassique entre 206 et 195 Ma (Sebai et
al., 1991). Les plus anciennes roches sdimentaires post-dater le plissement
sont du Crtac. Pour linstant la date la plus plausible de cet vnement est
donne par les compilations tectoniques globales, telles que celles de Stampfli
et Borel (2002). Elles proposent un ge denviron 320 Ma pour la collision des
continents Laurussia et Gondwana au niveau de lactuelle Afrique du Nord.
Sil est vrai que la Meseta (Figure 8, p.17) fait galement partie de ldifice
varisque, des corrlations avec lAnti-Atlas sont hasardeuses.
Malgr sa
proximit actuelle, sa position dans lorogne varisque a t nettement plus

interne. Elle a subi des dformations bien avant lAnti-Atlas (Hoepffner et al.,
2005). Aujourdhui la relation gomtrique qui lie lAnti-Atlas la Meseta est
masque par le Haut Atlas.
2.6. Les vnements post-varisques

Du Permien au Jurassique, aucun sdiment ne se dpose sur le domaine de
lAnti-Atlas. En revanche dans le Haut-Atlas central (au Nord de lAnti-Atlas
22
Thse de doctorat
2. Cadre gologique
oriental), plusieurs kilomtres de sdiments syn-rift se dposent du Trias la fin

du Jurassique (Manspeizer et al., 1978 ; Beauchamp et al., 1999). On trouve
galement dans cette srie sdimentaire des roches intrusives basaltiques,
contemporaines des dolrites de lAnti-Atlas, aux alentours de la limite Trias
Jurassique (Sebai et al., 1991). Ces intrusions sont les prmices de louverture
de lAtlantique centrale au Jurassique infrieur et moyen (At Brahim et al.,
2002).
Les prochains sdiments se dposer dans lAnti-Atlas, scellant la dformation
varisque, sont dge Crtac suprieur. Ils forment la base du Hamada. Ils
dbutent par des grs grossiers et des conglomrats ; suivent des marnes
rouges, puis des calcaires. Aucun sdiment Palocne nest prsent sur lAntiAtlas et la Hamada. Aprs ce hiatus, des sdiments continentaux et lacustres se
dposent durant le Miocne, puis au Pliocne.
Sur la bordure nord de lAnti-Atlas oriental (sud du Haut Atlas) la srie
Cnozoque est plus complte. Elle enregistre linversion du Haut Atlas. Les
conglomrats, calcaires marins et marnes rouges du Crtac suprieur
correspondent la phase post-rift atlasique (Beauchamp et al., 1999). Ils sont
suivis par des sdiments Palocnes et Eocnes denvironnement lagunocontinental (calcaires lacustres, marnes grseuses, oolithes). La premire des
deux phases dinversion du Haut Atlas a lieu ds lEocne suprieur et durant
lOligocne. Les sdiments Oligo-Mio-Pliocnes continentaux (molasses du
bassin de Ouarzazate) se dposent ensuite (Grler et al., 1988). La deuxime et
la plus importante phase dinversion du Haut Atlas intervient durant le
Plistocne et le Quaternaire infrieur (Frizon de Lamotte et al., 2000 ; Ellouz et
al., 2003). Le raccourcissement de linversion du Haut Atlas est plus important
sur les bords du bassin atlasique en particulier sur sa limite sud (Teixell et al.,
2003). Mais cette dformation naffecte pas directement lAnti-Atlas oriental.
Le relief particulirement dvelopp du Haut Atlas central est partiellement d
linversion et un paississement crustal (Arboleya et al., 2004). Une origine
thermique est responsable de la topographie actuelle de lAnti-Atlas et pour une
Thse de doctorat
23
2. Cadre gologique
partie du Haut Atlas (Saddiqi et al., 2005 ; Ayarza et al., 2005). Les modles
gophysiques (gravimtriques) de Missenard et al. (2006) dmontrent un
amincissement lithosphrique (<70 km) orient NE-SW. Cette anomalie est
oblique par rapport au systme atlasique et correspond gographiquement au
volcanisme alcalin Nogne et Quaternaire, ainsi qu une bande dactivit
sismique. Initie par remonte mantellique durant le Miocne (Missenard et
al.,2006), lanomalie est actuellement davantage entretenue par la convergence
tectonique des plaques Ibro-Africaine et du craton saharien (Teixell et al.,2005).
24
Thse de doctorat
3
STRUCTURES
Description des observations
3.1. Introduction
LAnti-Atlas oriental est analys selon diffrentes structures.
-
La structure gnrale sobserve sur la carte gologique gnrale (Figure 5,

p.11 et Figure 11), sur les vues Landsat (Figure 12). Elle est dcrite par les
coupes de la Figure 17, p.33).
Par structures compressives et extensives, on entend celles que lon observe

sur le terrain. Elles sont typiquement mtriques hectomtriques
(reprsentes par les photographies). Vu leurs petites tailles, elles ne figurent
pas sur les coupes ; en revanche leur mesure permet de caractriser
lorientation de la dformation. La Figure 22 (p.40) reprsente une partie des
mesures effectues.
Sans parler de structures, une analyse spcifique de la bordure de lAntiAtlas oriental permet dtudier ses relations avec les roches sdimentaires
crtaces et cnozoques.
Thse de doctorat
25
3. Structures
Figure 11*. Carte gologique de lAnti-Atlas central et oriental. Cette carte nillustre que les
roches incorpores la dformation varisque. Ainsi toutes les roches post varisques
apparaissent en noir. Le rouge a t attribu 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.
26
Thse de doctorat
3. Structures
3.2. Stratigraphie mcanique

Le comportement mcanique des roches est mise en valeur par lanalyse des
structures. Elle sera nanmoins discute au pralable, puisquelle est un des
principaux paramtres qui contrlent le style de dformation.
Socle
Malgr une composition des plus htrognes et une forte anisotropie, les blocs
de socle ragissent de manire homogne, comme des units rigides avec de
trs faibles dformations internes. Ceci est visible sur les images Landsat
(Figures 3 et 12). Mise part la boutonnire de Bou Azzer, aucune relation
gomtrique directe ne lie la surrection des boutonnires et les diffrents
lments qui les composent, autant en terme dge (Eburnen, super-groupe de
lAnti-Atlas ou de Ouarzazate) quen terme ptrologique (mtamorphique,
plutonique, volcanique ou sdimentaire). Par contre, les failles normales
dveloppes lors du rift prcambrien prdfinissent la structure du rejeu de la
dformation varisque. Elles naffleurent malheureusement pas dans lAnti-Atlas
oriental, mais certaines structures dans la couverture fournissent des indices sur
leurs positions. Pour la partie orientale, le rift prcambrien est davantage dduit
de lanalyse de la nature des sdiments que de lobservation directe de
structures. Il nest pas non plus exclu que ces structures extensives soient dj
hrites dvnements Pan-Africains ou Eburnens.
Contact socle - couverture
Dans les premiers temps de ltude, un accent a t mis sur lobservation de
cette limite et la srie incomptente sus-jacente pouvant potentiellement contenir
un dcollement basal. Mais, outre certaines discordances angulaires, ce contact
ne montre ni chevauchement, ni dcollement basal. La Figure 13 montre ce
contact sur diffrents sites. Cette absence de dcollement basal est
certainement la caractristique la plus intrigante de la chane plisse de lAntiAtlas.
Thse de doctorat
27
3. Structures
Figure 13. Contact socle-couverture. A : Tazelaft, sud-est de la boutonnire du Saghro. Le

socle est compos dandsite et la couverture dbute par les schistes argileux du Cambrien
infrieur, surmont des Grs roses ; B : sur la bordure nord du Saghro, le bton rouge et blanc
mesure 1 m, sa base repose sur la discordance, le socle est andsitique, suivent 2.50 m de Grs
roses (conglomratiques) puis les schistes Paradoxides du Cambrien moyen ; C : lextrmit
est de la boutonnire de lOugnate. Rhyollite surmonte de grs conglomratiques.
Couverture
La couverture palozoque est trs nettement domine par de puissantes sries
de faible comptence. Elle est ponctue de niveaux marqueurs comptents.
Ceci explique la morphologie typique de la couverture sous la forme de
succession de cuestas (Figure 14).
28
Thse de doctorat
3. Structures
Figure 14. Cuesta au sud de la boutonnire de lOugnate. Ici sous le 2me Bani (Ordovicien
suprieur).
Units lithologiques structurales

La rhologie des roches impliques dans la dformation varisque est
reprsente dans la Figure 15. Cette colonne est une compilation des sries les
plus compltes de lAnti-Atlas oriental. Le trac soft-stiff dcrit la comptence
relative des roches. Elle a t dtermine daprs les observations de terrain, les
coupes et des critres de morphologie des affleurements (topographie).
Chaque unit lithologique structurale comprend un membre dominant dont la
comptence et lpaisseur relative au matriel environnant dterminent le
comportement de lunit. Les limites des units sont comprises dans les
intervalles incomptents. Les membres ou units de conformation sont ceux dont
la forme est impose par un membre dominant. Selon ces termes dfinis par
Currie et al. (1962), les deux membres dominants majeurs sont le socle et le 1er
Bani. Les autres membres comptents sont davantage des membres de
conformation, puisque leur forme gnrale pouse celle des membres
dominants. Toutefois plus petite chelle, chaque membre dveloppe sa propre
forme de structure de raccourcissement interne (layer-parallel shortening).
Autrement dit, les structures de dformation interne sont caractristiques de
chaque membre dominant et indpendantes des membres sous- et sus-jacents.
Thse de doctorat
29
3. Structures
Figure 15. Stratigraphie mcanique et units lithologiques structurales du socle et de la

couverture varisque de lAnti-Atlas oriental. FM : formations ; Soft-Stiff : Incomptent-comptant ;
MD : membre dominant ; ULS : unit lithique structurale. EB ; socle burnen ; OZ : supergroupe de Ouarzazate et lAnti-Atlas ; CB : Conglomrats de base ; GR : Grs roses ; GV : Grs
verts ; B1 : 1er Bani ; B2 : 2me Bani ; RC : Richs ; OK : Ouarkziz.
En ce sens on peut parler de dformation disharmonique et dcouper la colonne

(Figure 15) en cinq units lithiques structurales, avec une limite plus marque
dans les argiles de lOrdovicien infrieur. A grande chelle, nanmoins, toute la
30
Thse de doctorat
3. Structures
couverture palozoque se conforme au socle. Lensemble de ces roches ne

constitue ainsi quune seule grande unit lithologique structurale. La couverture
est drape sur les mouvements relatifs des blocs de socle. Ainsi dans le cas dun
pli forc (Figure 16), cette mcanique prvoit un modle avec des blocs de
socle rigides forant une srie incomptente et crant dans les squences de la
couverture une dformation disharmonique.
Figure 16. Effet de la mcanique sur un pli forc ou de drapage (daprs Groshong, 2002). En
rouge : unit rigide, en jaune : srie incomptente, en bleu : de comptence intermdiaire.
Cette mcanique particulire explique en partie labsence de chevauchement

majeur. Les puissantes sries incomptentes permettent la couverture de se
dformer de manire plastique. Ce phnomne damortissement sobserve
galement dans les structures en extension.
3.3. Structure gnrale - Coupes

La Figure 12 permet de se rendre compte de la structure gnrale de lAnti-Atlas
oriental. Laffleurement des boutonnires donne la structure globale savoir une
forme anticlinale dorientation ENE-WSW louest puis E-W lest (structure
antiforme majeure). Au cur de cet anticlinal se trouve le socle, puis de part et
dautre la couverture palozoque. Les affleurements sur le flanc nord sont
Thse de doctorat
31
3. Structures
rduits, car la couverture est elle-mme rduite, mais aussi parce quils ont t
masqus par linversion du Haut Atlas.
Lanalyse du trac du sommet de lOrdovicien suprieur (2me Bani) et celui de
lOrdovicien infrieur (1er Bani) montre une structure dun ordre infrieur
(structure majeure secondaire). Une succession de synclinaux et danticlinaux
dont les axes sont orients globalement N-S NW-SE et plongent vers le sudest. Mise en valeur par les membres dominants de lOrdovicien, cette structure
est galement souligne par les affleurements de socle (resserrement du
Saghro, ensellement
Saghro-Ougnate et dpression entre lOugnate et
laffleurement de lextrmit est) et par les diffrents bassins Dvoniens et

Carbonifres (Cuvette de Tazzarine, bassin de Mader et bassin du
Tafilalt). Cette structure globale est galement visible grce la carte gologique
remanie (Figure 11, p.26).
Construction des coupes

La structure de lAnti-Atlas oriental est reprsente par dix coupes (Figure 17)
perpendiculaires laxe de lanticlinal donn par lalignement des boutonnires.
Originellement, elles ont t construites par dessin une chelle de 1 : 50'000.
Le profil topographique provient des cartes topographiques du Maroc au
1 : 100'000. Les donnes gologiques proviennent dabord des donnes de
terrain, puis des cartes gologiques au 1 : 200'000 couples aux vues satellites
(limites lithologiques, failles) et aux cartes topographiques (calcul de pente sur
les pentes structurales). Les interprtations en profondeur et au-dessus de la
topographie ont t faites par simple projection. Sans preuves de variation, les
paisseurs ont t maintenues constantes. De mme, lorientation des failles
normales a t conserve en profondeur. En revanche lors dindications
latrales videntes, certaines failles ont t amorties en remontant dans la srie
ou inversement des failles ont t dessines sous des structures non failles.
Aucune distinction stratigraphique na t faite dans le socle.
32
Thse de doctorat
3. Structures
Figure 17*. Coupes gologiques de lAnti-Atlas oriental. Image Landsat avec la position des
coupes. Epaisseurs des sries dans les diffrentes coupes.
Thse de doctorat
33
3. Structures
Description des coupes

Alors que les cartes gologiques et les images satellites fournissent une image
dun plissement modr, les coupes font apparatre une structure gnrale
particulirement hache par des failles normales. Sur le terrain, ces failles
normales sont bien ou peu apparentes selon la srie. Elles sont trs marques
depuis le socle jusquau sommet de lOrdovicien (1er Bani). Beaucoup dentre
elles
sont
amorties
par
le
Silurien
et
le
Dvonien
infrieur
(sries
particulirement incomptentes). Seuls des rejets importants permettent aux

failles de traverser les sries Dvoniennes et Carbonifres.
Les failles normales tendent minimiser la structure gnrale de cette partie de
la chane. La structure anticlinale est bien visible sur les coupes 5 10, bien que
son amplitude soit diminue par leffet des failles normales. Les coupes 1 4
paraissent plus monotones et il faut restaurer les failles normales pour voir
apparatre cette structure principale. La coupe 1 montre toutefois une remonte
de socle un peu plus importante qui taye lide de la prsence dune
boutonnire plus lest, aujourdhui sous la Hamada.
Les coupes occidentales (10 5) montrent des failles normales trs localises
crant notamment le graben de Zagora. A lest (coupes 1 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 interprtation.
Ils servent contre-balancer la dformation particulirement intense des
sdiments Carbonifres dans cette zone. Le pli sur rampe chevauchement de
la coupe 8 est lui observ en surface ; sa gomtrie en profondeur reste
toutefois interprtative.
Les parties du Haut Atlas sont dessines partir des coupes de Frizon de
Lamotte et al. (2000).
34
Thse de doctorat
3. Structures
3.4. Structures compressives

Les observations lchelle de laffleurement des diffrents bancs marqueurs
ont permis de mettre en vidence plusieurs types de structures compressives.
On distingue ici deux types : les plis et les failles chevauchantes. Les unes ou
les autres apparaissent selon la puissance et la comptence des roches en
prsence.
Figure 18. Raccourcissement interne de la couverture. Chevauchement des bancs comptents

et plissement des bancs incomptents. Interpntration en queue de poisson (Fishtail
structure). Dvonien moyen. Vue vers le SE. Bordure ouest du Tafilalt. N 3116.800 / W
00420.400. Marteau dans le synclinal du banc incomptent.
La Figure 18 montre les deux cas. Les bancs de calcaire sont plus cassants, la
dformation provoque un chevauchement. Au-dessus de cette structure, la
lithologie est plus incomptente, le banc nest que pliss. La dformation est
diffuse, elle se rpartit dans tout le volume de la couche. Lorsque les conditions
daffleurement permettent lobservation des sries incomptentes, on distingue
une multitude de petits chevauchements presque parallles au litage (Figure 19).
Thse de doctorat
35
3. Structures
Figure 19. Plans stris de chevauchements mineurs dans des schistes argileux (Ordovicien
infrieur), illustrant le style caractristique de dformation interne dans les sries incomptentes.
Sur la bordure nord du Saghro : N 3124.361 / W 00530.192. La boussole est oriente au
Nord.
La dformation (pli ou chevauchement) est extrmement localise, latralement

le long du mme banc et lintrieur dune succession de bancs. Ainsi un banc
peut tre trs rgulier et seule une portion limite peut tre affecte par un pli ou
un chevauchement (Figure 20). Dautre part dans une srie de bancs spars
par des intervalles incomptents, seul un banc peut tre dform localement, la
dformation du prochain banc se situant ailleurs.
36
Thse de doctorat
3. Structures
Figure 20. Schma illustrant lindpendance des petites structures et leur diffrentes formes en
relation avec leur comptence. A : Structure en queue de poisson (Fishtail), illustre par la
Figure 17. B : Plissement harmonique de type buckling . Prsence de cisaillement banc sur
banc. C : Plissement dun banc dominant, prsentant des poly- ou des disharmonies. D :
Multitude de microchevauchements sub-parallles la stratification (Figure 19). E : Faille inverse
dans un banc massif.
Les petits plis

Au sud des boutonnires, les plis sont isols. Ils ne montrent pas de vergence
particulire. Au nord des boutonnires par contre, ils apparaissent beaucoup
plus rgulirement, voire en continu. Leur vergence est clairement oriente vers
le sud. La Figure 21 montre quelques plis typiques. Les mesures des plis sont
reprsentes par les strogrammes jaunes en hmisphre infrieur (Figure 22).
Dans une zone donne, les mesures de plusieurs plis sont reportes sur un seul
strogramme. Les points sont des ples de plans de stratification mesurs dans
ces plis. Laxe du pli est le ple (perpendiculaire) du plan dessin par la somme
des ples de la stratification. Les axes ne sont pas dessins sur la Figure 22.
Les donnes sont brutes, donc non retro-bascules par rapport au pendage
gnral des bancs dans lesquels les plis ont t mesurs. Ceci explique le
caractre bimodal du strogramme au sud de lOugnate ( lest de la longitude
W00500). Dans cette zone, les mesures des plis ont t prises dans des bancs
avec un pendage gnral plongeant tantt vers le sud, tantt vers le nord.
Thse de doctorat
37
3. Structures
Figure 21. Photographies de petits plis (Planche 1/2).
38
Thse de doctorat
3. Structures
Figure 21. (Planche 2/2).

Lgendes de la Figure 20 :
A : Plis dans le Dvonien moyen. Vue vers le SE. Bordure ouest du Tafilalt.
N 3116.800 / W 00420.400. Marteau en bas de limage.
B : Pli dans le Dvonien moyen. Vue vers le NW. Bordure ouest du Tafilalt.
N 3115.750 / W 00423.400. Echelle 1 m.
C : Pli dans le Dvonien moyen. Vue vers le NW. Bordure ouest du Tafilalt.
N 3115.550 / W 00423.540. 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 lOugnate. Depuis N 3117.194 / W 00455.591. Le synclinal au
premier tiers infrieur mesure environ 10 m de large.
E : Plis disharmoniques dans le Dvonien. Vue vers le Nord. Sud de lOugnate.
N 3115.020 / W 00448.360. Marteau au centre.
F + G : (la vue G est la base de la vue F) Plis dans le Dvonien. Vue vers le NW et NNW. Au
sud de lOugnate. N 3113.220 / W 00453.760. Echelle 1 m.
Thse de doctorat
39
3. Structures
H : Plis dans le Carbonifre. Vue vers le NE. Jebel Tisdafine (Carbonifre au nord de
lensellement Saghro Ougnate). N 3127.150 / W 00518.950. Echelle 1 m.
J : Pli-faille dans le 1er Bani. Vue vers le WSW. Bordure NNW du Saghro. Depuis N 3124.120 /
W 00537.640. Lanticlinal fait 100 m de large.
K+L : (vue gnrale et dtail) Pli dans le Carbonifre. Vue vers le NE. Jebel Asdaf, lambeaux
Carbonifre isol dans lOued Todra louest de l Ougnate. N 3125.455 / W 00512.368. K :
250 m de large, L : chelle 1 m.
M : Pli dvers vers le SE dans le Dvonien. Vue vers le NE. Au nord du Saghro. N 3127.240 /
W 00530.950. Echelle 1 m.
N : Pli couch dans le Dvonien. Vue vers lest. Au nord du Saghro.
N 3128.200 / W 00531.400. Marteau en bas gauche.
Figure 22*. Strogrammes des structures mesures. Jaune :

chevauchements. Rouge : structures extensives. Vert : structure gnrale.
Projection de Wulff, hmisphre infrieur.
petits
plis.
Bleu :
Lanalyse de lorientation de ces petits plis met en vidence deux orientations

prfrentielles. Lune axe E-W (les ples des plans de stratification sont
aligns N-S) lautre axe NW-SE (ples des plans de stratification aligns NESW). Les petits plis axe NW-SE dominent la moiti est, au sud des
boutonnires ( lest de la longitude W00500 et au sud de la latitude N3130).
Les deux orientations coexistent dans la rgion de Tazzarine (entre W005 et
40
Thse de doctorat
3. Structures
W006 de longitude et N3030 et N3100 de latitude), alors qu louest de

W006 lorientation E-W domine. Cette dernire domine galement au nord des
boutonnires avec une tendance ENE-WSW, lexception dun cas isol au nord
de lOugnate qui prsente un axe N-S.
Failles chevauchantes - inverses

Les chevauchements sont plus rares que les plis. Les structures en queue de
poisson (Fishtail) sont un encastrement de bancs plus massifs. Dans les sries
massives des failles inverses trs raides se dveloppent. Alors que dans les
srie de faible comptence, les chevauchements ont tendance se parallliser
la stratigraphie. Il nest pas rare de voir le plan de faille se matrialiser par une
minralisation, gnralement siliceuse. Ceci donne lieu des plans stris
rsistants lrosion. Les cristaux de roche (Quartz) y abondent. Au nord des
boutonnires, les petits chevauchements sont beaucoup plus intenses, en terme
de nombre et de rejet. La Figure 23 montre quelques exemples de
chevauchements. Leurs strogrammes apparaissent en bleu sur la Figure 22.
Les grands cercles reprsentent les plans de failles et les points, les stries. Les
chevauchements se dveloppent dans les sries plus comptentes et massives.
De part et dautre de la lithologie rigide, le chevauchement se perd dans les
sries incomptentes. On observe trois directions de contrainte principale. Une
contrainte dorientation NE-SW domine la partie est, elle correspond aux plis
daxe NW-SE. Une deuxime orientation est perpendiculaire la premire avec
une contrainte oriente NW-SE. Elle est trs marque au nord des boutonnires
et concide avec les orientations des petits plis et de la structure gnrale. Une
troisime orientation dans louest traduit des contraintes N-S.
En terme de structures compressives, notons encore la prsence dune
schistosit (Figure 24) au nord des boutonnires. Elle est absente dans le reste
de lAnti-Atlas oriental. Cette schistosit est trs raide et plonge vers le Nord.
Son orientation suit celle des petits plis de cette zone.
Thse de doctorat
41
3. Structures
Figure 23. Photographies de chevauchements (Planche 1/2).
42
Thse de doctorat
3. Structures
Figure 23. (Planche 2/2).

A : Chevauchement dans un banc massif du Dvonien moyen. Vue vers le SE. Centre du
Tafilalt. N 3119.900 / W 00410.050. Marteau au centre.
B+C+D : Chevauchements dans le 2me Bani. Sud de lOugnate. N 31 11.250 / W 00455.250.
B : Plan de chevauchement avec un angle aigu par rapport la stratification, le plan de faille est
matrialis par une minralisation siliceuse, vue au SSE, le second plan est 100 m. C + D :
Chevauchement et crochon dans le toit, vue au SE, bton 1 m.
E+F+G+H : Chevauchement dans le 1er Bani louest de Tazzarine. N 3044.250 / W
00604.450. E+F : Chevauchement et crochon, vue vers le SE, le chevauchement mesure 30 m.
G : Plan stri dans le crochon, vue vers le Sud, boussole oriente. H : Crochon suprieur,
diffusion du chevauchement, vue vers lest, longueur du chevauchement 30 m.
J : Structure Fishtail dans le Dvonien. Sud de lOugnate. N 3115.000 / W 00448.375. Echelle
1 m.
K : Faille inverse et chevauchement dans le Cambrien. Vue vers le NE. Au SE du Saghro (nord
de Tazzarine). Depuis N 3056.610 / W 00531.210. Palmier dans loued.
L : Faille inverse minralise et crochon suprieur dans lOrdovicien infrieur. Vue vers le WSW.
Bordure NNW du Saghro. N 3124.115 / W 00537.525. Echelle 1 m.
M : Chevauchement et pli sur rampe dans le Cambrien moyen. NNW du Saghro. N
3123.320 / W 00537.930 (sommet). La falaise fait 200 m de long.
Thse de doctorat
43
3. Structures
Figure 24. Schistosit dans les plis dverss Dvoniens. Vue vers lest. Au nord de la
boutonnire du Saghro. N 3128.200 / W 00531.400. Marteau au centre.
3.5. Structures extensives

Les observations directes des failles normales sont rares. La Figure 25 en
reprsente quelques exemples. Les failles normales ont davantage t mises en
vidence par des zones de contact de deux formations contrastes espaces
dans la colonne lithostratigraphique. Leur rejet est estim grce la
connaissance des puissances des formations. La plupart des mesures de
lorientation des failles normales ont t prises non pas sur la faille principale,
mais sur des failles secondaires dans la zone de faille avec des rejets plus petits.
44
Thse de doctorat
3. Structures
Figure 25. Photographies des structures extensives (Planche 1/2).
Thse de doctorat
45
3. Structures
Figure 25. Photographies des structures extensives (Planche 2/2).

A : Faille normale dans le sud de lOugnate. Vue vers louest. Depuis N 3116.590 / W
00456.025. Le sommet de gauche est 4 km. Sans considrer le premier plan, les deux
cuestas (au sud et au nord) sont du Cambrien moyen. Cette rptition est due une faille
normale qui plonge au nord. On trouve du socle au pied de la falaise de gauche.
B : Faille normale dans lOugnate. Vue vers lest. Depuis N 3120.945 / W 00500.800.
Landrover en bas gauche. La faille normale plonge au nord et met en contact le socle et la
cambrien infrieur.
C+D : Faille normale au nord de lensellement Saghro-Ougnate. N 3120.300 / W 00516.850.
C : vue vers louest, le bloc central ( 300 m) est le 1er Bani de lOrdovicien, il est en contact (au
sud) avec du Cambrien moyen, la faille normale plonge vers le Nord. D : vue vers lest, la faille
normale est dans lOrdovicien infrieur, la particularit est que les stries de la faille normale sont
sur-imprimes de stries indiquant un mouvement dextre. Sac dos pour chelle.
E : Faille normale dans lOrdovicien infrieur. Vue vers le NE. Bordure nord du Saghro. Depuis N
3123.995 / W 00537.360. La faille plonge au SE.
F+G+H : Graben lest de la cuvette de Tazzarine. Zone N 3045 / W 00513. F : Limite nord
du graben. Vue au Nord. Au fond les falaises du 1er Bani, plus en avant la faille normale
46
Thse de doctorat
3. Structures
plongeant au SSE, la plaine avec quelques affleurement de Dvonien et Silurien et nouveau le

sommet du 1er Bani sous la khasbah. G : Faille normale et pli dentranement dans le Dvonien
aux abords de la limite sud du graben. Echelle 1 m. H : Limite sud du graben. Vue vers le sud, la
falaise est le 1er Bani qui plonge au sud, devant la falaise la faille normale plonge au nord, au
premier plan les affleurements Dvoniens plongent au nord.
J+K : Vues sur limportante faille normale du sud de lOugnate. Depuis N 3115.480 / W
00445.990. J : Vue vers lest. La faille spare le Dvonien moyen (au nord) du Cambrien moyen
(au sud). Le Dvonien plonge localement au nord, le Cambrien plonge au sud, la srie est
rgulire et continue, au loin droite les falaises du 1er Bani. K : Vue vers le WSW. Au sud la
cuesta de Cambrien suprieur plonge au sud. Les diffrents affleurements sur la droite sont du
Dvonien qui plonge globalement au nord. A droite au loin sort une cuesta du 2me Bani
(Ordovicien suprieur), plongeant au sud. En plein centre au loin un synclinal perch de
Dvonien tmoigne de la faille.
Figure 26. Vue Landsat et en coupe dun synclinal li une faille normale au sud de la
boutonnire de lOugnate.
Thse de doctorat
47
3. Structures
Une toute autre manire de signaler leur prsence sont les synclinaux que les
failles normales provoquent dans les formations du toit, lorsque ces dernires,
de faible comptence, glissent sur des roches du mur trs rigides. Ces plis
dentranement ont souvent t interprts comme plis varisques. Ces structures
lies lextension sont dautant plus droutantes que comme tous les plis, elles
prsentent 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 caractrisent par un flanc limit par la faille
(gnralement le plus raide) et un flanc en concordance avec le reste de la srie
(Figure 26). Ces structures ont t observes dans le systme du rift de la Mer
Rouge (Khalil et McClay, 2002) et sont connues dans dautres systmes en
extension (Schlische, 1995). Elles sont galement dcrites dans des modles
exprimentaux (Withjack et al., 1990 ; Hardy et McClay, 1999 ; Finch et al.,
2004 ). Les modles avec une couverture ultra-incomptente semblent le mieux
correspondre la configuration observe dans lAnti-Atlas oriental. Selon Xiao et
Suppe (1992), le dveloppement de ces structures est induit principalement par
une gomtrie de faille de type listrique.
Toutes les failles normales observes sont trs raides, leur plongement est
typiquement compris entre 70 et 80. Lorientation de leur trace est constante
dans tout lAnti-Atlas oriental : E-W avec de lgres variations de WSW-ENE
WNW-ESE. Aucune autre orientation na t observe. Outre les grabens
symtriques, une majorit des failles normales plongent vers le nord,
lexception de la partie nord des boutonnires o les plongements sud dominent.
Les donnes de structures extensives sont reprsentes en rouge sur la Figure
22 (p.40). Un seul strogramme reprsente directement les failles avec leur
grand cercle et leur strie. Les autres strogrammes montrent les ples de
stratification mesurs dans les synclinaux provoqus par ces failles. Laxe de pli
que lon peut dduire, donne la direction de la trace de la faille. Le trs fort
pendage, que peut obtenir le flanc en contact tectonique, tmoigne du fort
plongement de la faille normale.
48
Thse de doctorat
3. Structures
3.6. Relations avec les roches rcentes et topographie

Les premires roches rellement sceller la dformation de lAnti-Atlas sont des
conglomrats crtacs. Cette discordance tectonique est bien visible grande
chelle sur les cartes gologiques (Figures 5 et 11). Elle est particulirement
bien mise en vidence lest du Tafilalt o les roches de la base de la Hamada
ctoient successivement toutes les formations de la srie palozoque et mme
des parties de socle.
Cette discordance est visible au nord du Saghro (photographies A et B de la
Figure 27). Une discordance relativement plate spare le Carbonifre
intensment dform, des sdiments crtacs. Ces couches conglomratiques
ont aujourdhui un faible pendage vers le nord, mais elles reposent sur la surface
rosive de manire conforme, sans discordance angulaire. Cette configuration
nest toutefois pas applicable lensemble du domaine. En effet, si lon projette
linairement le plan de stratification crtac, il butte contre les affleurements de
socle. Malgr des recherches, ce contact socle Crtac na pas t
directement observ, lrosion provoquant le retrait des cuestas de Crtac
(Figure 27 D).
Au-dessus des sdiments crtacs, reposent en concordance, des couches du
Palogne. Elles sont suivies des sdiments nognes, avec une importante
discordance angulaire sur une limite rosive, permettant par endroit leur dpt
directement sur le Crtac. Aux abords nord de la boutonnire du Saghro, les
sdiments crtacs et palognes ont typiquement des pendages de 10 15
vers le nord, alors que les sdiments nognes plongent de 5 (Figure 27 C et
D). Notons encore la prsence de galets de rhyolite provenant du socle varisque
dans les conglomrats nognes (galement observs par Schmidt 1988 et
1992).
Thse de doctorat
49
3. Structures
Figure 27. Photo des sdiments post-varisques.

A+B : Conglomrats Crtacs recouvrant en discordance des plis Carbonifre. Au nord du
Saghro. Vue vers le ENE. N 3128.450 / W 00533.100. Marteau pour chelle.
C : Butte prsentant une discordance angulaire entre les sdiments Eocnes (10) et Moipliocnes (< 5). Au loin gauche la boutonnire de Saghro. Vue vers le WSW. N 3122.310 / W
548.410. La butte est 400 m.
D : Contact rod Anti-Atlas sdiments post-varisque. Vue vers le sud-ouest. A gauche des
Rhyollites du Super-groupe Ouarzazate (plongement 18). Au centre et droite les couches
rouges conglomratiques du Crtac (12) surmontes de conglomrats Moi-Pliocnes (6) en
discordance angulaire. Depuis N 3121.103 / W 00548.860. La butte du centre est 450 m.
50
Thse de doctorat
3. Structures
Figure 28*. Modle numrique de terrain de lAnti-Atlas oriental. Donnes de GTOPO 30.
La Figure 28 permet de se rendre compte de la topographie dveloppe de

lAnti-Atlas oriental et plus particulirement de la boutonnire du Saghro et de
son voisin septentrional, le Haut Atlas central. Lanalyse de lorientation des
couches du pourtour nord, est et sud de lAnti-Atlas oriental, montre un
bombement gnral, visible galement par larrangement radiaire des rseaux
hydrographiques (modle numrique de terrain Figure 28 et vues Landsat
Figures 3 et 6). Les couches des Monts Kem-Kem au sud et du Hamada lest
plongent globalement en direction du sud-est. Au nord de lOugnate, les couches
crtaces plongent au nord.
Thse de doctorat
51
3. Structures
52
Thse de doctorat
4
MODELE GEOMETRIQUE
Interprtation des observations
4.1. Style structural

Laffleurement des boutonnires et des roches profondes de la couverture,
prouve limplication du socle lors de la formation de la chane de lAnti-Atlas.
Ainsi cette inversion massive est de type thick skin. Linversion des blocs de
socle provoque la dformation dans la couverture qui ragit de manire passive,
exception faite de la partie nord o lon observe de vritables chevauchements
lintrieur de la couverture.
Les modles numriques trishear (Figure 29) de Erslev (1991), Allmendinger
(1998), Johnson et Johnson (2002, A) ou Finch et al. (2003) semblent trs bien
correspondre la gomtrie globale des structures compressives. Le fait que
lon observe pas directement la faille inverse de la base rigide dcoule de trois
faits : (1) linversion des failles normales induit des raccourcis dans le mur de la
faille, vraisemblablement dans les sdiments syn-rift, de manire que la faille
shorizontalise avant lintersection du contact socle-couverture (Figure 30, stade
C). Ceci amne considrer une gomtrie de faille de type listrique. Linversion
induit le raccourci lendroit o la faille normale devient trop abrupte (langle
Thse de doctorat
53
4. Modle gomtrique
entre la prstructuration et la contrainte est trop obtus) (Bump, 2003 ; Yamada &
MacClay, 2003). (2) Une part importante du raccourcissement horizontal nest
pas localise en un seul horizon, mais est accommode de manire diffuse dans
limportant volume de roche incomptente. (3) Le niveau drosion actuel ne
permet pas encore dobserver ces failles inverses. Elles se trouvent au large
des boutonnires, sous la couverture. La prsence et la localisation de ces
failles inverses sont toutefois trahies par les failles dextension tardive qui
affectent la couverture au moins jusquau 1er Bani.
Figure 29. Principe de trishear (daprs Erslev, 1991). A : Solution classique. Elle ncessite un
dcollement basal. Il en rsulte un excs de volume de la couverture. B : Le modle trishear
palie ces problmes. C : Rpartition des vecteurs de dplacement dans la zone triangulaire.
54
Thse de doctorat
4. Modle gomtrique
Figure 30. Modle volutif du style structural de lAnti-Atlas oriental (stades A C repris de
Bump, 2003). A : Rift et sdiments syn-rift. B : Dpt des sdiments post-rift. C : Inversion avec
raccourcis dans le socle et drapage de la couverture (trishear). D : Extension. Les failles
normales recoupent la couverture.
Diffrence majeure avec lAnti-Atlas du sud-ouest

Malgr une stratigraphie globalement similaire, le style structural de la partie
orientale et celui du sud-ouest sont diffrents. Dans lAnti-Atlas occidental, des
dtachements se produisent dans la majeure partie des sries incomptentes.
La srie y comprend quatre membres dominants : le socle, les calcaires du
Cambrien infrieur (Adoudounien), le Bani Ordovicien et les Richs Dvoniens.
Chacun de ces membres dveloppe sa propre frquence et amplitude de
Thse de doctorat
55
4. Modle gomtrique
plissement, ce qui aboutit un plissement polyharmonique multi-couches (Helg

et al., 2004). Dans la gomtrie globale de lAnti-Atlas oriental, la couverture se
comporte comme une couche unique drapant les mouvements verticaux du
socle (Soulamani et al., 1997).
4.2. Restaurations et calcul de la profondeur de dtachement

Choix de la coupe
Le calcul de la profondeur de dtachement a t appliqu sur la coupe 8. Cette
coupe a t choisie, car elle offre le niveau daffleurement le plus bas dans la
srie. Ainsi le minimum derreur dinterprtation sur la morphologie du contact
socle-couverture est garanti. La coupe 8 montre un bon dveloppement de la
structure anticlinale majeure. Ceci implique quelle coupe un bloc de socle en
son milieu. La coupe 7 passant dans lensellement entre les boutonnires du
Saghro et de lOugnate, naurait pas t significative, puisque ce secteur peut
tre interprt comme une zone de transition entre les deux blocs.
Restauration
La Figure 31 illustre les diffrents stades dans la restauration de la coupe. Afin
dtablir la situation la fin du paroxysme varisque, il faut tout dabord restaurer
les failles normales tardives (passage du stade B au stade C). Ce profil est
ensuite remis plat (D). A cette tape, les couches contiennent encore une
composante importante de raccourcissement interne. Il faut par consquent les
tirer pour obtenir la configuration la fin de la sdimentation Palozoque. Le
profil A montre leffet de la surrection Nogne. Ces donnes proviennent du
modle dvelopp par Missenard et al. (2006) partir de donnes
gravimtriques. Cette composante rcente est prsente dans le relief actuel et
doit tre dduite du profil pour obtenir la configuration relle du paroxysme
varisque.
56
Thse de doctorat
4. Modle gomtrique
Figure 31. Restauration de la coupe 8. A : Surface diffrentielle due la surrection Nogne Su,
dduite de Missenard et al. (2006, Figure 6.C, profil oriental). B : Coupe actuelle ; Sx est la
surface en dfaut due lextension tardive. C : Coupe de la fin de lvnement varisque. Ce profil
contient toutefois encore leffet de la surrection rcente. D : Coupe restaure montrant la
situation avant linversion mais comprenant encore le raccourcissement interne. E : Coupe
restaure, sans raccourcissement interne.
Principe du calcul de la profondeur de dtachement

Pour estimer la profondeur du dtachement, on utilise la mthode des surfaces
dplaces (Epard et Groshong, 1993 ; Groshong, 1994, 2002). Cette mthode
dquilibrage analyse, sur une coupe en deux dimensions, la surface en excs
provoque par une compression ou une surface en dfaut, dans le cas dune
extension par rapport au niveau rgional virtuel reprsentant le niveau non
dform. Le principe est que cette surface en excs ou en dfaut est gale la
surface dplace, qui est elle-mme dfinie par la profondeur du dtachement
multiplie par le dplacement horizontal. Cette mthode ne donne toutefois
aucune information sur la gomtrie ni sur la dynamique de la mise en place du
dtachement. Ainsi dcrite, la mthode ne prend en compte aucune dformation
interne.
Le principe de notre dmarche est de calculer tout dabord la profondeur de
dtachement grce aux structures extensives. La premire hypothse est que
Thse de doctorat
57
4. Modle gomtrique
lextension ne provoque aucune dformation interne. On peut donc simplement

appliquer le calcul comme dcrit ci-dessus. Pour traiter le cas de la compression
nous modifions la mthode pour quelle prenne en compte la dformation
interne, nous ajoutons ainsi une inconnue. La seconde hypothse est que les
structures profondes et donc la profondeur de dtachement, sont les mmes
dans le cas de lextension tardive que dans le cas de linversion varisque. Nous
pouvons rduire dune inconnue, notre nouvelle formule, en injectant la
profondeur de dtachement trouve et ainsi dduire le raccourcissement interne
provoqu par linversion varisque.
Calcul de la profondeur du dtachement par les structures extensives

La surface en dfaut provoque par lextension tardive est appele Sx. Elle est
dfinie comme la diffrence daire entre le profil actuel (B dans la Figure 31) et le
profil restaur sans les failles normales (C). A ce stade de la restauration, la
surface provoque par le soulvement rcent (Su) na pas besoin dtre prise en
compte puisquelle est implicitement contenue dans les deux profils.
Figure 32. Dtachement provoqu par extension. Dfinition des lments gomtriques utiliss
dans le calcul de la profondeur de dtachement.
58
Thse de doctorat
4. Modle gomtrique
La surface en dfaut Sx, situe sous le niveau rgional Hr (Figure 32) rsulte
dune extension. Elle est compense par la surface dplace dfinie par la
longueur du dplacement horizontal Dx multiplie par hx la profondeur du
dtachement sous le niveau rgional Hr :
(1)
Sx = Dx hx ainsi
hx =
Sx
Dx
Le dplacement horizontal Dx est la diffrence entre la longueur originale Lx et

Wx la largeur de la structure :
(2)
Dx = Lx Wx
La profondeur du dtachement en absolu, par rapport laltitude 0 m, est

donne par :
(3)
Hx = Hr hx = Hr
Sx
Lx Wx
Notez que dans le cas de lextension, Sx et Dx sont tous deux ngatifs.

Cette profondeur de dtachement a t calcule pour diffrentes structures de la
coupe 8 : le graben de Zagora, la boutonnire du Saghro et la coupe entire.
Pour chaque structure le calcul a t appliqu aux diffrents horizons. Le
Tableau 1 prsente les mesures et les rsultats.
La profondeur de dtachement pour le graben de Zagora varie de -12 -18 km.
Les valeurs les plus profondes sont obtenues pour le Silurien et lOrdovicien
suprieur qui sont les mieux contraints puisquils sont affleurants. Le Graphique
1 reprsente la relation surface-profondeur. On reporte laltitude de chaque
niveau rgional (Hr) en fonction de sa surface en dfaut (Sx). Ce graphique
permet dextrapoler la profondeur du dtachement basal. Cette profondeur est
dfinie comme celle laquelle la surface en dfaut est nulle (Sx = 0 km2). Ainsi
Thse de doctorat
59
4. Modle gomtrique
la droite pointille dsigne une profondeur de -16 km. La pente de cette droite
donne linverse du dplacement (1/Dx). On obtient une valeur pour Dx de 0.204
km ce qui concorde avec les valeurs mesures.
Niveau
Hr (km)
Sx (km2)
Dx (km)
hx (km)
Hx (km)
Graben de Zagora
Silurien
1.920
3.743
-0.185
-20.232
-18.312
Ordovicien suprieur
1.330
3.533
-0.185
-19.097
-17.767
Ordovicien infrieur
0.725
3.290
-0.185
-17.784
-17.059
Cambrien moyen
-1.555
3.047
-0.260
-11.719
-13.274
Cambrien infrieur
-2.125
2.840
-0.245
-11.592
-13.717
Socle
-2.735
2.500
-0.250
-10.000
-12.375
Silurien
5.000
28.279
-1.245
-22.714
-17.714
Ordovicien suprieur
4.460
30.285
-1.435
-21.105
-16.645
Ordovicien infrieur
3.815
30.317
-1.490
-20.347
-16.532
Cambrien moyen
1.500
30.850
-1.490
-20.705
-19.205
Socle
0.355
25.623
-1.355
-18.910
-18.555
Silurien
2.100
68.000
-2.155
-31.555
-29.455
Ordovicien suprieur
1.445
92.000
-2.365
-38.901
-37.456
Ordovicien infrieur
0.800
135.000
-2.480
-54.435
-53.635
Cambrien moyen
-1.425
135.000
-2.545
-53.045
-54.470
Socle
-2.560
144.000
-2.635
-54.649
-57.209
Saghro
Coupe entire
Tableau 1. Profondeur de dtachement sur les structures extensives.
La profondeur de dtachement obtenue par la boutonnire du Saghro varie de

-17 -19 km. Contrairement au graben de Zagora, ce sont les horizons infrieurs
qui sont les mieux contraints.
Les rsultats obtenus sur la totalit de la coupe sont irrationnels. La profondeur
de dtachement varie de -29 -57 km. En terme de relation surface-profondeur,
le niveau le plus lev (ici le Silurien) devrait obtenir une surface en dfaut
suprieure au niveau le plus bas (socle). Cette dviation rsulte certainement
des raisons suivantes. (1) Lorientation des failles normales. Avec notre
interprtation des failles normales (avec des plongements de 70 80 maintenus
constants en profondeur), les dplacements horizontaux Dx sont faibles. Avec
60
Thse de doctorat
4. Modle gomtrique
des dplacements absolus plus grands, la profondeur de dtachement serait

moindre. Pratiquement, le dplacement change selon la gomtrie de la faille
normale. Laplatissement des failles en profondeur (type listrique) gnre deux
effets : des dplacements horizontaux plus grands et des surfaces en dfaut
plus petites par rapport des failles droites. (2) Aucune dformation interne na
t considre. En ralit, la longueur mesure Lx (Figure 32) est plus longue
que la longueur initiale avant lextension (tirement). Si lon applique la formule
(7) dveloppe pour les structures compressives (voir paragraphe suivant), une
faible dformation interne suffit pour changer significativement le rsultat de la
profondeur de dtachement. Avec une dformation interne de 3 %, la profondeur
de dtachement calcule pour le sommet du Cambrien infrieur passe de -54 km
seulement -19 km. Linfluence de lamincissement des bancs sur la surface en
dfaut est ngligeable puisque lamincissement vertical est compens par
lallongement horizontal (volume constant).
Graphique 1. Relation surface profondeur sur le graben de Zagora.
Thse de doctorat
61
4. Modle gomtrique
Profondeur de dtachement et dformation interne sur les structures

compressives
Les calculs sont effectus sur un profil restaur, montrant la gomtrie suppose
de la fin de linversion varisque. Pratiquement, les failles normales et leurs effets
ont t restaurs pour obtenir une gomtrie rgulire. Le niveau rgional Hr
(Figure 33) a t dfini lextrmit sud de la coupe. La surface diffrentielle
due la surrection tardive (Su) a t soustraite de la surface en excs mesure.
La valeur de Su pour la partie Saghro est de 12.9 km2 et de 61.2 km2 pour la
totalit de la coupe.
Figure 33. Dtachement par compression. Cas de figure sans et avec raccourcissement interne
(Dce).
Le calcul de la profondeur de dtachement sur les structures compressives sans

dformation interne utilise des quations semblables celles dmontres pour
lextension.
Sc = Dc hc ainsi
(1)
hc =
Sc
Dc
Le dplacement Dc est dfini comme tant la diffrence entre la longueur

originale Lc et la largeur de la structure Wc.
Dc = Lc Wc
(2)
62
Thse de doctorat
4. Modle gomtrique
Dans le cas de la compression, Dc et Sc sont positifs. La profondeur de

dtachement par rapport au niveau de la mer est donne par lquation
suivante :
Hc = Hr hc = Hr
(3)
Sc
Lc Wc
Applique telle quelle, cette formule fournit des valeurs extrmement profondes
de -111 -126 km pour le partie Saghro et de -172 -212 km pour la coupe
entire (Tableau 2). Etant donn les excellentes conditions daffleurement,
labsence totale de chevauchement ou dautres grandes structures est un fait
avr. Lquilibrage de ces coupes ncessite la considration dune composante
plus discrte. Le style structural, dict par une rhologie particulire, ne permet
pas de changer les valeurs des dplacements horizontaux autrement que par
raccourcissement interne. Cette dformation interne est par ailleurs trs bien
documente par les structures de plus petite taille ne figurant pas sur les
coupes.
Niveau
Hr (km)
Sc (km2)
Dc (km)
hc (km)
Hc (km)
Saghro
Silurien
1.550
206.090
1.730
119.595
-118.045
Ordovicien suprieur
1.000
212.090
1.725
122.951
-121.951
Ordovicien infrieur
0.340
222.090
1.925
115.371
-115.031
Cambrien moyen
-1.890
249.090
2.005
124.234
-126.124
Socle
-3.030
264.090
2.425
108.903
-111.933
Silurien
0.650
454.785
2.590
175.593
-174.943
Ordovicien suprieur
0.075
449.785
2.445
183.961
-183.886
Ordovicien infrieur
-0.585
411.785
2.400
171.577
-172.162
Cambrien moyen
-2.805
431.785
2.445
176.599
-179.404
Socle
-4.000
458.785
2.205
208.066
-212.066
Coupe entire
Tableau 2. Profondeur de dtachement sur les structures compressives.
Thse de doctorat
63
4. Modle gomtrique
Pour prendre en compte le changement de longueur dans le calcul de la

profondeur de dtachement, nous incluons le raccourcissement e :
e=
(4)
Dce
Lo
ainsi
Dce = Lo e
O la longueur Dce est le raccourcissement absolu et Lo la longueur originale

(voir Figure 32).
La longueur Lc est mesurable, elle est dfinie comme la longueur originale
moins le raccourcissement :
Lc = Lo Dce
(5)
En introduisant lquation (4) dans la (5), on obtient :
Lc = Lo Lo e = Lo(1 e) or
(6)
Lo =
Lc
1 e
On introduit ensuite cette composante de raccourcissement interne, lquation

(1) de la profondeur de dtachement devient :
hce =
(7)
Sc
Dc + Dce
Dc est remplac par lquation (2), Dce par la (4) et Lo par la (6). Cette quation
se simplifie ensuite comme suit :
hce =
(7)
Sc
Sc
Sc
=
=
e
Lc Wc + Lo e Lc Wc + Lc e
Lc1 +
Wc
1 e
1 e
Finalement pour calibrer la profondeur du dtachement laltitude :

Hce = Hr hce
(8)
Le Tableau 3 donne les valeurs de la profondeur de dtachement (Hce) pour

chaque niveau avec une dformation allant de 0 0.25 (gamme compatible avec
les observations de terrains). Ces rsultats sont reports sur le Graphique 2.
64
Thse de doctorat
4. Modle gomtrique
Considrant une profondeur de dtachement raisonnable de -18 -20 km

(dtermine sur les structures en extension), on obtient des valeurs de
dformation interne de 0.11 0.17.
Silurien
Ordov.sup
Ordov.inf
Camb.moy
Socle
0.00
-174.943
-183.886
-172.162
-179.404
-212.066
0.01
-109.169
-112.435
-104.84
-110.83
-126.226
0.02
-78.801
-80.501
-75.032
-80.175
-90.015
0.03
-61.313
-62.402
-58.216
-62.798
-70.045
0.04
-49.945
-50.748
-47.416
-51.609
-57.391
0.05
-41.962
-42.617
-39.895
-43.802
-48.655
0.06
-36.049
-36.621
-34.356
-38.045
-42.262
0.07
-31.492
-32.017
-30.107
-33.624
-37.380
0.08
-27.874
-28.371
-26.744
-30.122
-33.530
0.09
-24.931
-25.411
-24.016
-27.280
-30.416
0.10
-22.490
-22.961
-21.759
-24.928
-27.846
0.11
-20.434
-20.900
-19.860
-22.948
-25.688
0.12
-18.677
-19.141
-18.241
-21.259
-23.851
0.13
-17.159
-17.623
-16.844
-19.801
-22.269
0.14
-15.834
-16.299
-15.626
-18.530
-20.890
0.15
-14.667
-15.135
-14.555
-17.412
-19.680
0.16
-13.633
-14.103
-13.606
-16.421
-18.608
0.17
-12.709
-13.181
-12.759
-15.536
-17.652
0.18
-11.878
-12.354
-11.998
-14.741
-16.794
0.19
-11.128
-11.607
-11.311
-14.024
-16.021
0.20
-10.447
-10.929
-10.688
-13.373
-15.319
0.21
-9.826
-10.311
-10.120
-12.779
-14.680
0.22
-9.257
-9.745
-9.601
-12.236
-14.095
0.23
-8.734
-9.226
-9.123
-11.737
-13.558
0.24
-8.252
-8.746
-8.683
-11.277
-13.064
0.25
-7.806
-8.303
-8.276
-10.851
-12.606
Tableau 3. Profondeur de dtachement (km) en fonction de la dformation compressive (e) pour

les diffrents niveaux de la coupe 8.
Thse de doctorat
65
4. Modle gomtrique
Graphique 2. Profondeur de dtachement des diffrents niveaux sur la coupe entire en fonction
de la dformation compressive.
4.3. Interfrence et gradient de dformation

Lanalyse de lorientation des petits plis et des chevauchements reprsents
dans la Figure 22 (p.40) permet les interprtations suivantes :
- Les plis axe NW-SE, qui dominent la moiti orientale au sud des
boutonnires, peuvent tre associs la dformation de la chane de lOugarta.
Ils correspondent galement la structure secondaire majeure.
- Les plis axe E-W (tendance ENE-WSW) au nord des boutonnires sont
parallles la structure antiforme majeure de lAnti-Atlas.
- Les plis dorientation E-W (tendance ESE-WNW) sont le rsultat de la somme
de la formation de la structure antiforme majeure (Anti-Atlas) et de la structure
majeure secondaire (Ougarta). En effet les faisceaux ougartiens divergent. Ainsi
66
Thse de doctorat
4. Modle gomtrique
ils sont perpendiculaires la structure antiforme majeure dans la partie orientale

et pratiquement parallles cette dernire dans la partie occidentale.
- A chelle gale, les deux orientations de plis E-W et NW-SE coexistent dans la
rgion de Tazzarine (N 3100 / W 00530). Lorientation E-W est attribue la
structure
antiforme
majeure,
lorientation
NW-SE
lOugarta.
Les
chevauchements au nord de Tazzarine, avec des contraintes orientes NW-SE

sont galement attribus la structure anticlinale majeure.
- Lexception au nord de lOugnate avec une orientation N-S est attribue la
dformation dorientation ougartienne. Elle pourrait galement provenir dun
poinonnement dun coin de socle provoquant des plis de drapage.
Limage globale de la compilation de ces interprtations est reprsente par la
Figure 33. Cest linterfrence des deux orientations (Anti-Atlas et Ougarta) qui
donne cette structure en dmes et bassins. Cette particularit de lAnti-Atlas
oriental est due au fait de la proximit de lOugarta et la perpendicularit des
deux orientations dans cette zone. Ceci conforte galement le concept de
mtacraton pour cette partie du craton ouest-africain.
Considrant la direction anti-atlasique (ENE-WSW), le nord des boutonnires et
le Jebel Tisdafine en particulier (Carbonifre au nord de lensellement SaghroOugnate), sont caractriss par un plissement beaucoup plus intense quailleurs
dans lAnti-Atlas oriental. Les couches y sont dformes sur toute leur longueur
et non pas uniquement en un segment trs localis. Le dversement des plis
vers le nord, les chevauchements et le dveloppement dune schistosit sont
autant de particularits propres cette rgion nord de lAnti-Atlas oriental. Ces
observations amnent dcrire un fort gradient de dformation qui dcrot du
nord au sud.
La direction ougartienne prsente galement un gradient de dformation. Ce
gradient est moins marqu que celui de lAnti-Atlas. Il napparat pas par
lanalyse des observations de terrains, mais par linterprtation gologique
grande chelle (Figures 11 ou 12, p.26). Si lon considre la structure majeure
secondaire dorientation ougartienne, on saperoit que les plis de la partie
Thse de doctorat
67
4. Modle gomtrique
Figure 34* : Faisceaux dans lAnti-Atlas oriental. Avec et sans le fond Landsat.
68
Thse de doctorat
4. Modle gomtrique
orientale sont caractriss par une longueur donde plus courte et une amplitude
suprieure par rapport la zone plus louest. En effet, lanticlinal sparant le
Tafilalt du Mader est nettement plus troit que celui sparant le Mader et la
cuvette de Tazzarine (longueur donde). Les variations damplitude sont
montres par laffleurement des roches plus ou moins profondes. Dans la partie
orientale, des roches profondes (Cambriennes au moins) se trouvent proximit
gographique de roches Carbonifres, alors que dans la partie occidentale, les
affleurements varient du 1er Bani (Ordovicien) au Dvonien (voire mme jusquau
Silurien uniquement). Cette observation peut tre explique par les variations
dpaisseur qui affectent la couverture. Il est vident que si la couverture est plus
fine, la probabilit de voir affleurer des roches plus ancienne est plus grande.
Mais cette constatation est rfute par lobservation des pendages. Les
pendages sont en effet beaucoup plus forts lest et plus plats louest. Ces
observations dmontrent galement limplication du socle dans la dformation
dorientation ougartienne. Linterfrence obtenue dans cette partie de la chane
dcoule donc du croisement de deux dformations toutes deux de type thick
skin.
4.4. Structures dextension tardive

Les failles plongeant au nord sont majoritaires, on les interprte comme les
principales, alors que les failles qui plongent au sud sont leurs conjugues. Cette
interprtation est influence, mais galement supporte, par la connaissance de
la configuration du craton.
Malgr le fait que toutes ne recoupent pas toujours lentire srie Palozoque,
les failles normales sont attribues une extension post-varisque. La rhologie
particulire de la couverture explique le fait quelles naffleurent pas
systmatiquement. Le modle trishear en extension de la Figure 35 illustre ce
phnomne. Les niveaux marqueurs suffisamment spars de la faille par des
sries incomptentes, forment des plis de drapage ou plis de revtement. Ceci
explique labsence de faille normale dans les sdiments suprieurs de la
Thse de doctorat
69
4. Modle gomtrique
couverture. Certains larges plis sont interprts comme tant le seul rsultat des
failles normales, comme par exemple les plis Dvoniens et Carbonifres 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 lest qu louest, sexplique
par le simple fait que la couverture y est moins paisse.
Figure 35. Modle trishear en extension.
Avec lhypothse que ces failles rutilisent les anciennes discontinuits du socle,
leur emplacement permet la localisation de ces structures profondes. La forte
inclinaison des failles normales dans la couverture permet une extrapolation
aise. Toutefois, mise part la projection, ces failles ne dfinissent pas
directement
le front des blocs inverss lors de lvnement varisque. Tout
comme les failles inverses prennent un raccourcis dans le socle, les failles
normales empruntent un chemin lgrement diffrent. Elles se dveloppent de
manire verticale et recoupent la couverture (Figure 30.D, p.55), sans utiliser le
chemin cr par linversion. La faille inverse tant trop plate en surface
(Faccenna et al., 1995).
Toutes les failles normales suivent lorientation anti-atlasique, aucune faille
nindique une orientation ougartienne. Le Tafilalt montre un nombre de failles
normales plus important quailleurs. Ceci explique les rptitions de cuestas
dune mme formation. Les rejets sont plus faibles, mais la frquence des failles
70
Thse de doctorat
4. Modle gomtrique
est plus grande. On peut en dduire que la taille des blocs de socle est infrieure
dans cette zone et donc que le socle y est plus fractur quailleurs. Il est possible
que cette intensit de fracture traduise un interfrence des structures profondes
des deux orientations (Anti-Atlas et Ougarta).
Cette phase extensive naffecte que lAnti-Atlas oriental. Elle est interprte
comme le tmoignage de louverture du bassin adjacent du Haut-Atlas. Ceci
explique le fait que cette phase napparaisse pas dans le reste de lAnti-Atlas.
Les premiers sdiments se dposer dans ce bassin datent du Trias. Toutefois
comme aucun sdiment Triassique ou Jurassique ne se dpose sur lAnti-Atlas,
il est possible que cette extension soit uniquement un effet de bord et que cette
extension naffecte pas forcment lAnti-Atlas dans les premiers stades de
louverture de ce bassin.
4.5. Relations Anti-Atlas roches Post-varisques

La base des sdiments Crtacs dcrit une surface de pnplaine. Exception
faite des abords des boutonnires, les cuestas des formations comptentes
montrent des crtes dune grande rgularit topographique, tmoignant dun
palo niveau drosion. Lobservation de la discordance tectonique, sparant
les roches Carbonifres plisses et les conglomrats Crtacs, permet daffirmer
que lintense dformation, observe dans les sdiments Palozoques de la
partie nord, est due exclusivement la phase varisque et non linversion du
Haut Atlas. Les diffrents pendages mesurs dans les sdiments post-varisques
tmoignent toutefois dune certaine surrection. Lvolution chronologique de ce
soulvement est mis en valeur par la discordance angulaire sparant les
sdiments Crtacs-Palognes et Nognes. Les calculs de Schmidt (1988 ;
1992) sur le recul des cuestas indiquent le dbut de la surrection lEocne
suprieur.
Thse de doctorat
71
4. Modle gomtrique
Considrant la projection des couches Crtaces dans la partie nord, deux

configurations du contact socle Crtac sont possibles. (1) Soit les sdiments
Crtacs buttent sur le socle, dcrivant ainsi une discordance angulaire ; le socle
apparaissant alors comme une le dans cette surface Crtace. (2) Soit les
couches Crtaces pousent la topographie du socle, montrant ainsi une
surrection tardive (partielle) des boutonnires. Lanalyse du pendage des
sdiments Crtacs et Cnozoques parle en faveur de cette configuration. La
prsence de roches du socle varisque dans et par-dessus les sdiments
Palognes prouve lexistence dune le de socle (configuration 1). La
situation relle est donc une configuration intermdiaire. Cette thorie se limite
aux boutonnires du Saghro et de lOugnate. Dans lextrmit orientale ( lest
du Tafilalt), le Crtac repose sur le socle montrant exclusivement la surrection
tardive (configuration 2).
Figure 36*. Topographie et gologie de surface selon les coupes de lAnti-Atlas oriental. La
coupe 10 comprend galement une courbe montrant linfluence de la surrection Nogne
(dessine daprs Missenard et al. (2006, Figure 6.C, profil oriental).
72
Thse de doctorat
4. Modle gomtrique
Le bombement, observable par lanalyse de lorientation des couches postvarisques sur les vues satellite (Figure 3, p.7) et le modle numrique de terrain
(Figure 27, p.51), provient du soulvement rcent d une anomalie thermique
(Missenard et al., 2006), mais il nest pas lunique responsable de la forte
topographie. La courbe du soulvement modlis par Missenard et al. (2006) a
t reporte sur le profil topographique de la coupe 10 (Figure 36), cale au sud,
la base des sdiments Crtacs. Il apparat que la topographie du Bani et de la
boutonnire du Saghro dpasse de cette limite rosive thorique. Ces hauts
reliefs ont donc d constituer des les ds la fin de linversion varisque.
4.6. Interprtations des donnes sismiques

Le modle gomtrique est tay par lanalyse de trois lignes sismiques
ralises par lONAREP (actuellement ONHYM) dans la rgion du Tafilalt. RS 8
et RS 10 sont orientes N-S (Figure 37). La ligne RS 3 na pas pu tre
clairement positionne. Elle a une orientation E-W, passant vraisemblablement
au sud de Rissani. La ligne RS 8 permet une excellente corrlation avec les
donnes de surface, cest pourquoi elle a t choisie pour linterprtation. RS 10
laisse une trop grande part linterprtation tant donn lpaisse squence de
Carbonifre qui en compose la surface.
RS 8 est perpendiculaire lextension et oblique par rapport aux deux structures
majeures : WSW-ENE dans la partie nord et NW-SE dans la partie sud. Le dix
premier kilomtres (Figure 38) depuis le sud montrent de larges plis. Ils sont
interprts comme dtachs du socle la base de la pile palozoque, le
raccourcissement de la couverture tant compens par linversion des anciennes
failles normales dans le socle. Ces plis correspondent lorientation
ougartienne.
Le reste du profile est domin par des failles normales tardives. Une srie de
pseudo-bassins (p. ex. km 25) ont t anciennement interprts comme des
Thse de doctorat
73
4. Modle gomtrique
Figure 37. Image satellite du Tafilalt avec la position des lignes sismiques RS 8 et RS 10.
74
Thse de doctorat
4. Modle gomtrique
Figure 38*. Ligne sismique RS 8. En haut : vierge. Au centre : interprtation en profondeur

daprs les donnes de surface (terrain et carte gologique). En bas : profile aprs restauration
des failles normales.
bassins syn-sdimentaires de chevauchements. Notre interprtation prvoit un

modle en extension, dans lequel la gomtrie de ces pseudo-bassins est
explique
comme
incomptente
la
consquence
additionne
dun
dune
couverture
particulirement
artefact
sismique.
Premirement,
lamincissement des couches du toit de la faille et lpaississement de celles du

mur est un fait observ spcialement dans les modles couverture
incomptente sur socle rigide (Finch et al., 2004). Deuximement, laxe z des
lignes sismiques est donn en temps (Two Ways Time) et non en profondeur. La
vitesse des ondes dans les marnes et argiles du Carbonifre doit tre
extrmement faible (probablement infrieure 2000 m/s), alors quelle peut
atteindre 3000 m/s dans les couches comptentes telles que dans celles des
calcaires dvoniens et des quartzites ordoviciennes. Cette diffrence de vitesse
provoque un artefact (velocity pull down) qui tend exagrer la profondeur de
ces pseudo-bassins .
Thse de doctorat
75
4. Modle gomtrique
La troisime partie de la Figure 38 est le profile restaur qui illustre la

configuration prsume avant la mise en place des failles normales. Le
raccourcissement mesur avoisine les 10 %, mais il ne prend pas en compte le
raccourcissement interne observ plus petite chelle.
76
Thse de doctorat
5
MODELE DYNAMIQUE
Chronologie des vnements
5.1. Structuration Prcambrienne

Lorogne Pan-Africain (650 580 Ma) termine lassemblage du Gondwana
(Boote et al., 1998). Si cet vnement na pour ainsi dire aucune influence sur la
pr-structuration de lvnement varisque dans lAnti-Atlas occidental, il nen va
pas de mme pour ce qui est de la partie orientale. En effet, le socle de lAntiAtlas oriental peut tre schmatis en une zone triangulaire, limite au NNW par
la faille sud atlasique, lest par lanomalie du Saoura et au SSW par lAccident
Majeur. Cette dernire structure dcoule directement de ldifice Pan-Africain.
Son orientation concide avec le linament du Tibesti qui traverse lAfrique en
direction ESE. Si cette structure est dcrite comme une suture, le socle de lAntiAtlas oriental nest pas pour autant allochtone par rapport au craton ouestafricain. Cette zone dcoulant dun bassin aulacogne du Noprotrozoque
infrieur, correspond plutt une bordure moins rigide du craton, do cette
qualification de mtacratonique.
Le rift dbutant au Noprotrozoque terminal et qui dure jusquau Cambrien
moyen, marque le dbut du super-cycle du Gondwana (Boote et al., 1998). Cette
phase dextension affecte tout lAnti-Atlas. La prstructuration en blocs par des
failles normales est un des paramtres dterminant la gomtrie de linversion
Thse de doctorat
77
5. Modle dynamique
varisque. La prsence de zones de faiblesse crustale permettra linversion en

type thick skin. La trace des failles normales suit la structure gnrale de lAntiAtlas. Les failles plongent le plus gnralement vers lextrieur du craton soit
vers le nord (Figure 39, stade A). Les plongements symtriques sont interprts
comme leurs conjugus. Les sdiments associs au rift (Figure 30, stade A,
p.55), trs puissants dans la partie occidentale, fournissent dimportants horizons
de dcollement potentiel. Cet aspect ne concerne toutefois pas lAnti-Atlas
oriental, o ces sries sont fortement rduites voire absentes.
Figure 39. Modle du dtachement mdio-crustal pour la coupe 8. Sans exagration verticale.
La couverture comprend le Cambrien, lOrdovicien et le Silurien. A : Profil avant la compression
avec les failles normales hrites du rift Prcambrien. B : Profil au paroxysme de linversion
varisque. Les failles normales sont rutilises mais recoupent le socle et saplatissent avant
datteindre la couverture. C : Extension. Les failles normales atteignent la couverture. D :
Soulvement diffrentiel rcent d une anomalie thermique (donne de Missenard et al. 2006,
Figure 6.C).
78
Thse de doctorat
5. Modle dynamique
5.2. Bassin Palozoque

Durant pratiquement la totalit du Palozoque, une puissante srie de
sdiments se dpose dune manire rgulire dans cette partie septentrionale du
bassin de Tindouf. Quelques hiatus refltent probablement une activit
tectonique distante (Burkhard et al., 2006). La varit de sdiments est faible.
Lenvironnement de dpts passe dun milieu domin par le dtritisme jusqu
lOrdovicien suprieur un milieu carbonat avant un retour du dtritisme la fin
du Dvonien (Figure 10, p.21). Aucun sdiment ne traduit des conditions de mer
profonde, ce qui conforte lide dun bassin de type intracontinental au mme
titre que dautres bassins nord africains comme celui de lOugarta, de Reggane
ou de Taoudenni par exemple (Figure 3, p.7).
La puissance de cette couverture atteint 8 km dans lAnti-Atlas oriental et mis
part quelques bancs marqueurs, la srie est trs incomptente (Figure 15, p.30).
Ce sont ces deux caractristiques particulires qui vont dicter le style de la
dformation dans la couverture.
5.3. Inversion et interfrence varisques

Lvnement varisque provoque une inversion massive de type thick skin
(Figures 39, stade B et 30 C, p.55). Les anciennes failles normales rejouent en
failles inverses sans affecter directement la couverture. La datation de cet
vnement est difficile contraindre, mais on peut le dater raisonnablement la
fin du Carbonifre infrieur, selon les extrapolations grande chelle de Stampfli
et Borel, 2002 (Figure 4, p.10) ou de Veevers, 2004.
La chronologie relative des deux dformations dorientations diffrentes (AntiAtlas et Ougarta) na pas t tablie de faon certaine. Des indices daxes de
plis replisss dans la rgion de Tata dans lAnti-Atlas occidental (Caritg et al.,
2004), indiquent que lorientation anti-atlasique (axe NE-SW) prcde la
dformation dorientation ougartienne (E-W). Mais pour autant que ces directions
Thse de doctorat
79
5. Modle dynamique
soient correctement interprtes (Figure 40), cette chronologie relative nest pas
forcment identique dans lAnti-Atlas oriental, vu galement la proximit qui lie
lAnti-Atlas oriental la chane de lOugarta. Dailleurs, dans sa partie la plus
orientale, la prfrence est donne pour une orientation ougartienne prcdant
la dformation anti-atlasique. En effet, les axes des petits plis dorientation
ougartienne sont basculs par la structure majeure de la dformation antiatlasique (voir le strogramme le plus oriental de la Figure 22, p.40). Dans la
rgion de Tazzarine, lanalyse des deux strogrammes de petits plis semble
indiquer le contraire. Les petits plis dorientation ougartienne ont un axe de pli
horizontal, alors que laxe des petits plis dorientation anti-atlasique est bascul
vers louest. Ce basculement tant provoqu par la structure majeure secondaire
attribue la dformation ougartienne. Mais ceci peut tre expliqu par le fait
que les dformations internes (petits plis) prcdent la dformation de plus
grande envergure (structure majeure). Au sud de la boutonnire de lOugnate
ces relations sont brouilles par les synclinaux provoqus par lextension venir.
Figure 40*. Interprtation grande chelle de la carte gologique en terme de plis ou de

faisceaux.
80
Thse de doctorat
5. Modle dynamique
5.4. Extension Msozoque

La formation de failles normales affectant la couverture est post-varisque, car
elles recoupent les structures majeures anti-atlasiques et ougartienne (Figure 39,
stade C, p.78 et Figure 30, stade D, p.55). De plus les synclinaux crs dans le
toit des failles normales replissent les axes des petits plis varisques. Ces failles
normales ne recoupent pas toujours la totalit de la couverture Palozoque,
mais elles laffectent nanmoins. En effet, le rejet est parfois amorti dans les
sries sommitales, crant ainsi des plis deffondrement drapant les
mouvements verticaux de la mme manire que lors de lvnement varisque.
Cette extension naffecte pas les sdiments Crtacs. Le fait que lon trouve ces
failles uniquement dans la partie bordant le Haut Atlas permet raisonnablement
dattribuer cette phase dextension louverture du bassin du Haut Atlas au Trias
suprieur. Quelques cas montrent une ractivation de ces failles dans un
systme dextre.
5.5. Rjuvnation Cnozoque

Linversion du Haut Atlas ne provoque pas de dformation telle que plissement
ou chevauchement dans lAnti-Atlas oriental. Mais il induit un soulvement
progressif identifiable par les diffrentes discordances angulaires observes
entre les sries Crtaces, Eocnes et Mio- Pliocnes sur la bordure nord de
lAnti-Atlas oriental. Ces tapes dans le soulvement correspondent celles de
linversion du Haut Atlas. Ce bombement lisible par lanalyse de lorientation des
sdiments Crtacs et Cnozoques (Figure 41) provient de surcrot dune
surrection actuelle due une anomalie thermique. Cette conjonction explique la
topographie dveloppe de lAnti-Atlas oriental et plus particulirement celle du
Saghro. Cela nannule toutefois pas les effets de ldifice varisque puisque
lanalyse des constituants des sdiments post-varisques dmontre que la
boutonnire du Saghro du moins, a toujours t mergeante depuis lvnement
varisque, expliquant ainsi labsence de sdiments Triasiques et Jurassiques sur
Thse de doctorat
81
5. Modle dynamique
le domaine de lAnti-Atlas alors en extension. Ceci est galement dmontr par

la comparaison de luplift modlis (Missenard et al., 2006) avec le relief actuel
(Figure 36, p.72).
Figure 41*. Modle tridimensionnel du modle numrique de terrain. Exagration verticale 10 :1.
Donnes GTOPO 30.
82
Thse de doctorat
6
CONCLUSIONS
Apports de ltude
6.1. Style structural

Le style structural est contrl par deux paramtres principaux :
Pr-structuration du socle. Les failles normales du rift Noprotrozoque
Cambrien constituent des zones de faiblesse qui seront ractives lors de
linversion varisque et lors de lextension Msozoque. Le fait que le socle de
lAnti-Atlas oriental soit mtacratonique, le diffrencie du reste de la chane et
favorise linterfrence Anti-Atlas Ougarta.
Rhologie de la couverture. Lincomptence de la srie Palozoque permet
lamortissement des mouvements de socle, lors de linversion varisque selon le
modle trishear. Ce drapage est galement observ, dans la partie suprieure
de la couverture, lors de lextension tardive. Cette incomptence rend possible
un important raccourcissement interne discret grande chelle : 3 km de
raccourcissement observ (2%) contre 11 17 % de dformation interne (16
25 km). Elle explique labsence de dcollement majeur.
Thse de doctorat
83
6. Conclusions
La diffrence de style structural avec lAnti-Atlas occidental provient du fait que la

srie y est encore plus paisse et que les sries incomptentes comprennent
des dtachements qui permettent le buckling polyharmonique.
6.2. Structures
Linversion varisque est de type thick skin. On ne trouve aucun dcollement
majeur au contact socle couverture, ni dans dautres horizons de dcollement
potentiel. Le dtachement senracine en milieu de crote (-18 -20 km). Les
failles inverses se prolongent en chevauchements horizontaux, probablement
localiss dans les sdiments syn-rift du sommet du socle (Figure 39, stade B,
p.78). Ceci peut toutefois crer, au front de la structure majeure, une dformation
de type thin skin. La gomtrie globale de la couverture est donc un drapage
induit par les mouvements relatifs des blocs de socle.
Une multitude des petites structures compressives plus ou moins isoles
tmoigne dune part importante de dformation interne dans la couverture.
Lanalyse de lorientation des structures majeures et des petites structures
montre une image dinterfrence due au croisement de deux dformations
dorientations distinctes : anti-atlasique et ougartienne. Cette partie de la chane
est donc davantage un carrefour plutt quun arc relayant lAnti-Atlas lOugarta
(Figure 40, p.80).
Lintensit de la dformation indique clairement un gradient dcroissant vers le
sud pour ce qui est de lorientation anti-atlasique. Un gradient ougartien
dcroissant vers louest est moins marqu.
Une phase dextension tardive cre des failles normales dans la couverture et
provoque localement des synclinaux (anciennement interprts comme plis
varisques). Ces structures extensives trs raides permettent dextrapoler les
limites des blocs de socle.
84
Thse de doctorat
6. Conclusions
6.3. Tectonique
Le caractre mtacratonique du socle de lAnti-Atlas, la suture Pan-Africaine et
les failles normales du rift Noprotrozoque Cambrien sont autant de
structures dterminantes dans lagencement de linversion varisque.
LAnti-Atlas
est
la
partie
dforme
du
bassin
de
Tindouf.
Lanalyse
sdimentologique et la corrlation avec dautres bassins nord-africains,

permettent de conclure un bassin intracontinental. Il bnficie dune
subsidence rgulire durant le Palozoque, certaines variations tant attribues
des fluctuations du niveau marin, des vnements climatiques ou des
phases orogniques lointaines. Il se distingue des autres bassins par une
inversion plus massive et une dformation plus intense. Il est nanmoins
connect aux chanes inverses des bordures du craton ouest africain, comme
lOugarta ou lAhnet-Mouydir (Haddoum et al., 2001).
LAnti-Atlas oriental enregistre louverture du bassin du Haut Atlas par la
formation de structures extensives. Mais linversion alpine ne provoque aucune
dformation directe dans lAnti-Atlas oriental, except un soulvement gnral
combin leffet dune anomalie thermique rcente.
6.4. Implications et Perspectives

Daprs son style structural, lAnti-Atlas a des analogues tels que les bassins
inverss de lorognse Laramide de Uinta ou de Wind River, dcrits par Mitra et
Mounts (1998) ou Stone (1999). Ils ont en commun une dformation de type
trishear, induite par des mouvements verticaux de blocs de socle rigides, forant
une couverture de faible comptence. LAnti-Atlas dvie pourtant de ces
exemples, par laplatissement en surface des failles inverses. Ceci est rendu
possible grce aux sries post-Panafricaines et syn-rift qui caractrisent la partie
Thse de doctorat
85
6. Conclusions
suprieure du socle varisque. Cette dviation de toute autre chane dcrite

jusqu prsent, fait de lAnti-Atlas un objet gologique particulier. Cela peut tre
intressant dans le domaine appliqu de la recherche de ressources
nergtiques et pourrait amener (r)interprter la structure dautres bassins.
La nouvelle mthode de calcul de la profondeur de dtachement, prenant en
compte le raccourcissement interne, permettra la comparaison avec dautres
bassins inverss. Elle pourrait tre teste sur des chanes mieux connues et de
styles structuraux diffrents.
Lentreprise dune comparaison avec les autres bassins nord-africains en terme
de style structural et de raccourcissement interne pourrait aider la
comprhension de lassemblage de la Pange.
Ltude de la dnudation tectonique du Saghro par des mthodes danalyse de
traces de fission ou de rayonnements cosmiques permettrait de dfinir plus
prcisment lhistoire de sa topographie.
Enfin, des essais en modles analogiques, jouant les multiples ractivations des
blocs de socle, en variant la comptence et la puissance de la couverture,
seraient une exprience fort intressante.
86
Thse de doctorat
Rfrences
REFERENCES
AIT BRAHIM L., P. CHOTIN, S. HINAJ, A. ABDELOUAFI, A. EL ADRAOUI, C. NAKCHA, D.

DHONT, M. CHARROUD, F. SOSSEY ALAOUI, M. AMRHAR, A. BOUAZA, H. TABYAOUI,
A. CHAOUNI (2002), Paleostress evolution in the Moroccan African margin from
Triassic to present, Tectonophysics, 357, pp. 187-205.
ALGOUTI AB., AH. ALGOUTI, B. CHBANI, M. ZAIM (2001), Sdimentation et
volcanisme synsdimentaire de la srie de lAdoudounien infra-cambrien
travers deux exemples de lAnti-Atlas du Maroc, Journal of African Earth
Sciences, Vol. 32, N 4, pp. 541-556.
ALLMENDINGER R.W. (1998), Inverse and forward numerical modeling of trishear
fault-propagation folds, Tectonics, Vol. 17, N. 4, pp. 640-656.
ARBOLEYA M.L., A. TEIXELL, M. CHARROUD, M. JULIVERT (2004), A structural
transect through the High and Middle Atlas of Morocco, Journal of African Earth
Sciences, 39, pp. 319-327.
AYARZA P., F. ALVAREZ-LOBATO, A. TEIXELL, M.L. ARBOLEYA, E. TESN, M.
JULIVERT, M. CHARROUD (2005), Crustal structure under the central High Atlas
Mountains (Morocco) from geological and gravity data, Tectonophysics, 400, pp.
67-84.
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.
BAYER R., A. LESQUER (1978), Les anomalies gravimtriques de la bordure
orientale du craton ouest-africain: gomtrie dune suture pan-africaine, Bull.
Soc. Gol. Fr. 20, pp. 863-876.
BEAUCHAMP W., R.W. ALLMENDINGER, M. BARANZAGI, A. DEMNATI, M. EL ALJI, M.
DAHMANI (1999), Inversion tectonics and evolution of the High Atlas Moutains,
Morocco, based on a geological-geophysical transect, Tectonics, Vol. 18, N 2,
pp. 163-184.
BENSSAOU M., N. HAMOUMI (2003), The lower-Cambrian western Anti-Atlasic
graben: tectonic control of palaeogeography and sequential organisation, C. R.
Geoscience, 335, pp. 297-305.
Thse de doctorat
87
Rfrences
BOOTE D.R., D. CLARK-LOWES, M. TRAUT (1998), Palaeozoic petroleum systems

of North Africa. In: Macgregor, D.S., Moody, R.T.J. & Clark-Lowes, D.D. (eds).
Petroleum Geology of North Africa. Geological Society, London, Special
Publication No. 132, pp. 7-68.
Buggish, W., E. Flgel (1988), The Precambrian / Cambrian boundary in the
Anti-Atlas (Morocco) discussion and new results, in The Atlas System of
Morocco, edited by V.H. Jacobshagen, Lecture Notes Earth Sci., vol. 15, pp.
361-404, Springer-Verlag, New-York.
BUMP A.P. (2003), Reactivation, trishear modeling, and folded basement in
Laramide uplifts: implications for the origins of intracontinental faults, GSA
Today, pp. 4-10.
BURKHARD M. S. CARITG, U. HELG, C. ROBERT-CHARRUE, A. SOULAMANI (2006),
Tectonics of the Anti-Atlas of Morocco, Comptes Rendus Geoscience, 338, pp.
11-24.
CARITG S., M. BURKHARD, R. DUCOMMUN, U. HELG, L. KOPP, C. SUE (2004), Fold
interference patterns in the Late Palaeozoic Anti-Atlas belt of Morocco, Terra
Nova, 16, pp.27-37.
CHOUBERT G., A. FAURE-MURET (1971), Epoque hercynienne, In: Tectonique de
lAfrique, Sciences de la Terre, Paris, 6, UNESCO, Paris, France, pp. 353-371.
CHOUBERT G., A. FAURE-MURET (1983), Anti-Atlas, In: Lexique Stratigraphique
International, Nouvelle srie n1, Afrique de lOuest, J. Fabre (Editor),
Pergamon, pp. 80-95.
CLOETINGH S., J.D. VAN WEES, P.A. VAN DER BEEK, G. SPADINI (1995), Role of prerift 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.
COWARD M., A. RIES (2003) Tectonic development of North African basins in :
Arthur, T. J., MacGregor, D. S. & Cameron, N. R. (eds) Petroleum Geology of
Africa : New Themes and Developing Technologies. Geological Society, London,
Special Publications, 207, 61-83.
CURRIE J.B., H.W. PATNODE, R.P. TRUMP (1962), Development of folds in
sedimentary strata, Geological Society of America Bulletin, v. 73, pp. 655-673.
ELLOUZ N., M. PATRIAT, J.-M. GAULIER, R. BOUATMANI, S. SABOUNJI (2003), From
rifting to Alpine inversion: Mesozoic and Cenozoic subsidence history of some
Moroccan basins, Sedimentary Geology, 156, pp. 185-212.
ENNIH N., J.-P. LIGEOIS (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.
88
Thse de doctorat
Rfrences
ENNIH N., J.-P. LIGEOIS (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.
ERSLEV E.A. (1991), Trishear fault-propagation folding, Geology, v. 19pp. 617620.
FACCENNA C., T. NALPAS, J.-P. BRUN, P. DAVY (1995) The influence of preexisting thrust faults on normal fault geometry in nature and in experiments,
Journal of Structural Geology, Vol. 17, N 8, pp. 1139-1149.
FINCH E., S. HARDY, R. GAWTHORPE (2003), Discrete element modelling of
contrational fault-propagation folding above rigid basement fault blocks, Journal
of Structural Geology, 25, pp. 515-528.
FINCH E., S. HARDY, R. GAWTHORPE (2004), Discrete-element modelling of
extensional fault-propagation folding above rigid basement fault blocks, Basin
research, 16, pp. 489-506.
FRIZON DE LAMOTTE D., B. SAINT BEZAR, R. BRACNE, E. MERCIER (2000), The two
main steps of the Atlas building and geodynamics of western Mediterranean,
Tectonics, Vol. 19, N 4, pp. 740-761.
FRIZON DE LAMOTTE D.,CRESPO-BLANC A., SAINT-BEZAR B.,COMAS M., FERNANDEZ
M., ZEYEN H., AYARZA P., ROBERT-CHARRUE C., CHALOUAN A., ZIZI M., TEIXELL
A., ARBOLEYA M.L., ALVAREZ-LOBATO F., JULIVERT M., MICHARD A. (2004)
TRANSMED Transect I. In: Cavazza W, Roure F, Spakman W, Stampfli GM, and
Ziegler PA (eds) The TRANSMED Atlas: the Mediterranean Region from Crust to
Mantle, Springer Verlag, 2004, XXIV, 141 pp. ISBN : 3-540-22181-6.
GASQUET D.,G. LEVRESSE, A. CHEILLETZ, M.R. AZIZI-SAMIR, A. MOUTTAQI (2005),
Contribution to geodynamic reconstruction of the Anti-Atlas (Morocco) during
Pan-African times with the emphasis on inversion tectonics and metallogenic
activity at the PrecambrianCambrian transition, Precambrian Research, 140,
pp. 157-182.
GRLER K., F.-F. HELMDACH, P. GAEMERS, K. HEISSIG, W. HINSCH, K. MDLER, W.
SCHWARZHANS, M. ZUCHT (1988), The uplift of the central High Atlas as deduced
from Neogene continental sediments of the Ouarzazate province, Morocco, in
The Atlas System of Morocco, edited by V.H. Jacobshagen, Lecture Notes Earth
Sci., vol. 15, pp. 361-404, Springer-Verlag, New-York.
GROSHONG R.H. (1994), Area balance, depth to detachment, and strain in
extension, Tectonics, v. 13, No. 6, pp. 1488-1497.
Thse de doctorat
89
Rfrences
GROSHONG R.H., 2002, 3-D structural geology: 2nd printing, Springer-Verlag, 324 p.
GUIRAUD R., J.-C. DOUMNANG MBAIGANE, S. CARRETIER, S. DOMINGUEZ (2000),
Evidence for a 6000 km length NW-SE-striking lineament in northern Africa: the
Tibesti Lineament, Journal of the Geological Society, London, Vol. 157, pp. 897900.
HADDOUM H., R. GUIRAUD, A. MOUSSINE-POUCHKINE (2001), Hercynian
compressional deformations of the Ahnet-Mouydir Basin, Algerian Saharan
Platform: far-field stress effects of the Late Palaeozoic orogeny, Terra Nova, 13,
pp. 220-226.
HARDY S., K. MCCLAY (1999), Kinematic modelling of extensional faultpropagation folding, Journal of Structural Geology, 21, pp. 695-702.
HEFFERAN K.P., H. ADMOU, J.A. KARSON, A. SAQUAQUE (2000), Anti-Atlas
(Morocco) role in Neoproterozoic Western Gondwana reconstruction,
Precambrian Research, 103, pp. 89-96.
HELG U., M. BURKHARD, S. CARITG, C. ROBERT-CHARRUE (2004), Folding and
inversion tectonics in the Anti-Atlas of Morocco, Tectonics, Vol. 23, TC4006,
doi:10.1029/2003TC001576.
HOEPFFNER C., A. SOULAIMANI, A. PIQU (2005), The Moroccan Hercynides,
Journal of African Earth Sciences, 43, pp. 144-165.
JOHNSON K.M., A.M. JOHNSON (2002, A), Mechanical models of trishear-like folds,
Journal of Structural Geology, 24, pp. 277-287.
KHALIL S.M., K.R. MCCLAY (2002), Extensional fault-related folding, northwestern
Red sea, Egypt, Journal of Structural Geology, 24, pp. 743-762.
LIGEOIS J.-P., A. FEKKAK, O. BRUGUIER, E. ERRAMI, N. ENNIH (2005) Lower
ediacaran age (630-610 Ma) for the Sarhro group. Implications for the
metacratonic evolution of the Anti-Atlas, correlations with the Tuareg shield and
the evolution of the pan-african orogeny in the NW Africa (abs.), 4me Colloque
International 3Ma, Agadir 2005, Abstract Volume, pp. 2.
MACGREGOR D.S. (1996), The hydrcarbon systems of North Africa, Marine and
Petroleum Geology, Vol. 13, N 3, pp. 329-340.
MANSPEIZER W., J.H. PUFFER, H.L. COUSMINER (1978), Separation of Morocco
and eastern North America: A Triassic-Liassic stratigraphic record, Geological
Society of America Bulletin, v. 89, pp. 901-920.
MISSENARD Y., H. ZEYEN, D. FRIZON DE LAMOTTE, P. LETURMY, C. PETIT, M.
SRBIER, O. SADDIQI (2006), Crustal versus asthenospheric origin of relief of the
Atlas Mountains of Morocco, Journal of Geophysical Research, Vol. 111,
B03410, 13 p.
90
Thse de doctorat
Rfrences
MITRA S., V.S. MOUNT (1998), Foreland Basement-Involved Structures, AAPG

Bulletin, V. 82, N. 1, pp. 70-109.
PIQUE A., M. BOUABDELLI, A. SOULAMANI, N. YOUBI, M. ILIANI (1999), Les
conglomerates du P III (Noprotrozoque suprieur) de lAnti-Atlas (Sud du
Maroc): molasses panafricaines, ou marqueurs dun rifting fini-protrozoque ?,
C. R. Acad. Sci. Paris, Sciences de la Terre et des plantes, 328, pp. 409-414.
PIQUE A. (2001) Geology of Northwest Africa, Beitrge zur regionalen Geologie
der Erde ; Bd. 29, Gebrder Borntraeger, 310 p.
PIQU A. (2003) Perspective, Evidence for an important extensional event during
the latest Proterozoic and Earliest Paleozoic in Morocco, C. R. Geoscience, 335,
pp. 865-868.
ROBERT-CHARRUE C., M. BURKHARD (soumis), Variscan inversion tectonics and
interference pattern in the Tafilalt, Anti-Atlas, Morocco, Journal of African Earth
Sciences.
SADDIQI O., N. ZIAD, Y. MISSENARD, D. FRIZON DE LAMOTTE, B. GOFFE, P.
LETURMY, R. EL MAMOUN, S. SEBTI, G. POUPEAU (2005), Age Miocne moyen
suprieur du soulvement thermique du Haut-Atlas Central : apport des traces
de fission sur apatite, 4me Colloque International 3Ma, Agadir 2005, Abstract
Volume, pp. 45.
SAQUAQUE A., H. ADMOU, J. KARSON, K. HEFFERAN, I. REUBER (1989),
Precambrian accretionary tectonics in the Bou Azzer-El Graara region, AntiAtlas, Morocco, Geology, v. 17, pp. 1107-1110.
SCHLISCHE R.W. (1995), Geometry and Origin of Fault-Related Folds in
Extensional Settings, AAPG Bulletin, V. 79, N. 11, pp. 1661-1678.
SCHMIDT K.-H. (1988), Rates of scarp retreat: a means of dating neotectonic
activity, in The Atlas System of Morocco, edited by V.H. Jacobshagen, Lecture
Notes Earth Sci., vol. 15, pp. 445-462, Springer-Verlag, New-York.
SCHMIDT K.-H. (1992), The tectonic history of the Pre-Saharan depression
(Morocco) a geomorphological interpretation, Geologische Rundschau, 81/1,
pp. 211-219.
SEBAI A., G. FERAUD, H. BERTRAND, J. HANES (1991), 40Ar/39Ar dating and
geochemistry of tholeiitic magmatism related to the early opening of the central
Atlantic rift, Earth and Planetary Science Letters, 104, pp. 455-472.
SOULAMANI A., C. LE CORRE, R. FARAZDAQ (1997), Dformation hercynienne et
relation socle/couverture dans le domaine du Bas-Dra (Anti-Atlas occidental,
Maroc), Journal of African Earth Sciences, Vol. 24, N 3, pp. 271-284.
Thse de doctorat
91
Rfrences
SOULAMANI A., M. BOUABDELLI A. PIQUE (2003), Lextension continentale au NoProtrozoque suprieur Cambrien infrieur dans lAnti-Atlas (Maroc), Bull.
Soc. Gol. France, 174 (1), pp. 83-92.
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.
STONE D.S. (1999), Foreland Basement-Involved Structures: Discussion,
Discussion & Reply, AAPG Bulletin, V. 83, N. 12, pp. 2006-2016.
TEIXELL A., M.-L. ARBOLEYA, M. JULIVERT, M. CHARROUD (2003), Tectonic
shortening and topography in the central High Atlas (Morocco), Tectonics, Vol.
22, N 5.
TEIXELL A., P. AYARZA, H. ZEYEN, M. FERNANDEZ, M.-L. ARBOLEYA (2005), Effects
of mantle upwelling in a compressional setting:the Atlas Mountains of Morocco,
Terra Nova, 17, pp. 456-461.
THOMAS R.J., A. FEKKAK, N. ENNIH, E. ERRAMI, S.C. LOUGHLIN, P.G. GRESSE, L.P.
CHEVALLIER, J.-P. LIGEOIS (2004), A new lithostratigraphic framework for the
Anti-Atlas Orogen, Morocco, Journal of African Earth Sciences, 39, pp. 217-226.
VEEVERS J.J. (2004), Gondwanaland from 650-500 Ma assembly through 320 Ma
merger in Pangea to 185-100 Ma breakup: supercontinental tectonics via
stratigraphy and radiometric dating, Earth-Science Reviews, 68, pp. 1-132.
WITHJACK M.O., J. OLSON, E. PETERSON (1990), Experimental Models of
Extensional Forced Folds, AAPG Bulletin, V. 74, N. 7, pp. 1038-1054.
XIAO H., J. SUPPE (1992), Origin of Rollover, AAPG Bulletin, V. 76, N. 4, pp.
509-529.
YAMADA Y., K. MCCLAY (2003), Application of geometric models to inverted listric
fault systems in sandbox experiments. Paper 2: insights for possible along strike
migration of material during 3D hanging wall deformation, Journal of Structural
Geology, 25, pp. 1331-1336.
ZIEGLER P.A., S. CLOETINGH (2004), Dynamic processes controlling evolution of
rifted basins, Earth-Science Reviews, 64, pp. 1-50.
92
Thse de doctorat
Rfrences
Cartes gologiques
S.E. M. Fetah, M. Bensad, M. Dahmani (1986) carte gologique du Maroc
1:200'000, Tafilalt Taouz (N 244), Ministre de lEnergie et des Mines,
Royaume du Maroc. J. Destombes, H. Hollard.
S.E. M. Fetah, Bensad, M. Dahmani (1988) carte gologique du Maroc
1:200'000, Todrha Mader (N 243), Ministre de lEnergie et des Mines,
Royaume du Maroc. R. du Dresnay, G. Dubar, J. Hindermeyer, G. Choubert, A.
Emberger, J. Caa, J. Destombes, H. Hollard.
S.E. M. Fetah, Bensad, M. Dahmani (1989) carte gologique du Maroc
1:200'000, Zagora Coude du Dra Hamada du Dra (p.p.), Ministre de
lEnergie et des Mines, Royaume du Maroc. G. Choubert, A. Faure-Muret, J.
Destombes, H. Hollard.
S.E. M. Saadi, E.A. Hilali, M. Bensad (1975) carte gologique du Maroc
1:200'000, Jebel Saghro Dads (N 161), Ministre de lEnergie et des Mines,
Royaume du Maroc. J. Hindermeyer, H. Gauthier, J. Destombes, G. Choubert, A.
Faure-Muret, E. Roch, G. Dubar, P. Fallot, E. Laville, J.L. Lesage, R. du
Dresnay.
S.E. M. Saadi, E.A. Hilali, M. Bensad (1982) carte gologique du Maroc
1:200'000, Bou Haara Zegdou (N 259), Ministre de lEnergie et des Mines,
Royaume du Maroc. J. Destombes, H. Hollard.
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
Thse de doctorat
93
Rfrences
94
Thse de doctorat
TABLES DES FIGURES

TABLEAUX ET GRAPHIQUES
Figures
Page
1.
2.
3.
4.
5.
6.
7.
8.
2
3
7
10
11
12
15
9.
10.
11*.
12.
13.
14.
15.
16.
17*.
18.
19.
20.
21.
22*.
23.
24.
25.
26.
27.
28*.
29.
30.
31.
Situation de lAnti-Atlas du Maroc

Carte gnrale du Craton Ouest Africain
Carte daccs des tudes dans lAnti-Atlas oriental
Palogographies de Stampfli et Borel
Elments gologiques de LAnti-Atlas oriental
Vues tridimensionnelles de lAnti-Atlas
Limites cratoniques de lAnti-Atlas
Situation de lAnti-Atlas par rapport la chane des
Appalaches la fin du Palozoque
Formations et colonne stratigraphique de lAnti-Atlas
Schma chronostratigraphique de lAnti-Atlas
Carte gologique de lAnti-Atlas central et oriental
Image satellite Landsat avec certains contacts
Contact socle-couverture
Cuesta au sud de la boutonnire de lOugnate
Rhologie et units lithiques
Effet de la rhologie sur un pli forc ou de drapage
Coupes gologiques de lAnti-Atlas oriental, avec position
des coupes et paisseurs des sries
Raccourcissement interne de la couverture
Plans stris de chevauchements mineurs
Schma illustrant lindpendance des petites structures
Photographies de petits plis
Strogrammes des structures mesures
Photographies de chevauchements
Schistosit dans les plis dverss Dvoniens
Photographies des structures extensives
Vue Landsat et en coupe dun synclinal
Photographies de sdiments post-varisques
Modle numrique de terrain de lAnti-Atlas oriental
Principe de trishear (daprs Erslev, 1991)
Modle volutif du style structural de lAnti-Atlas
Restauration de la coupe 8
17
18
21
26
26
28
29
30
31
33
35
36
37
38
40
42
44
45
47
50
51
54
55
57
____________________
* Figures galement reprsentes dans lannexe D en format plus grand.
Thse de doctorat
95
32.
33.
34*.
35.
36*.
37.
38.
39.
40*.
41*.
Dtachement provoqu par extension

Dtachement par compression
Faisceaux dans lAnti-Atlas oriental
Modle trishear en extension
Topographie et gologie de surface selon les coupes
Image satellite du Tafilalt avec la position des lignes
sismiques RS 8 et RS 10
Ligne sismique RS 8
Modle du dtachement mdio-crustal pour la coupe 8
Interprtation de la carte gologique (plis et faisceaux)
Modle tridimensionnel du modle numrique de terrain
Tableaux
1.
2.
3.
58
62
68
70
72
74
75
78
80
82
Page
Profondeur de dtachement sur des structures extensives 60

Profondeur de dtachement sur des structures compressives 63
Profondeur de dtachement en fonction de la dformation
compressive
65
Graphiques
Page
1.
2.
61
66
96
Relation surface - profondeur sur le graben de Zagora

Profondeur de dtachement de diffrents niveaux
Thse de doctorat
REMERCIEMENTS
Mes remerciements posthumes Martin Burkhard qui ma toujours fait confiance

et encourag dans toutes mes entreprises. A celui qui ma inculqu la gologie
structurale, la cartographie, la gologie des Alpes. A celui qui ma rendu attentif
limportance de la gologie dans notre socit, aux problmes
environnementaux et politiques, la vulgarisation. A celui qui ma permis de
dcouvrir un bout du monde.
Au Fonds National de la Recherche Scientifique. Pour le financement des projets
N 21-52516.97 et 20-63790.00.
Au Fonds Matthey-Dupraz pour le soutien financier lors de mon travail de
diplme.
Au ministre de lIndustrie, de lEnergie et des Mines de Rabat, Maroc.
Notamment MM Dahmani et Haddane chefs de la Division de la Gologie, ainsi
que leur secrtaire et M. Scalante.
A lONHYM (anciennement ONAREP) pour nous avoir fournit les donnes
sismiques. A MM B. Bally et M. Zizi pour leurs interprtations.
A Dominique Frizon de Lamotte pour mavoir fait participer au le projet
Transmed, pour son avis critique et son soutient.
A MM les jurs A. Soulamani et M. Zizi pour avoir accept de faire partie de
mon jury de thse, pour leur critiques constructives.
Au Pr Jean-Paul Schaer pour ses conseils clairs et nos discussions
encourageantes.
Au Pr Jean-Paul Ligeois, pour ses changes intressants lors du congrs 3Ma
2005 Agadir et lors de sa visite Neuchtel.
Au Pr Karl Fllmi davoir spontanment pris le relais de directeur de thse.
Au Pr Thierry Adatte pour ses corrections pertinentes et constructives.
Thse de doctorat
97
A lInstitut de Gologie pour les belles annes passes. Une thse tant
galement le fruit de tout un environnement. :
A Mme la directrice le Pr A. Kalt, aux secrtaires Sabine et Franca, Andr,
Eli, Manu.
A mes collgues et amis assistants, enfants de linstitut (anciens et nouveaux) :
Baba, Nat, Kaspar, Rachel, Sverine, Urs, Jack, Julien, Feddi, Bas, Alex
(Corinne), Stphane (Justine), Beni (sixpack), Stafele (Gaminou), Mary-Alix (arcen-ciel de linstitut, avec qui jai aim partager mon bureau), Laurette (bonne
humeur de linstitut, photophile et super co-pilote en Tunisie), Flurin (10 nights in
Tunisia), Laure, Laurent, Haydon, Ccile, Pascal, Raoul, Nico, Sverine,
Christina ainsi quaux palo-matres-assistants Virginie et Erwan.
A mes collgues de voles : Lionel, Pauline, Christophe, Philippe (2x), Stork,
Gregory, Meryem, Virginie, Yves, Galaad, Julien, Mtro, ...
Aux tudiants avec qui jai partag TP et excursions : Aline, Pauline, Lucien,
Laurent, Rza, Patrick, Samuel, Tomaso, Roger, Robin, Esther, Laureline, JeanLuc, Johann, Jonathan (2x), Vincent (3x), Baptiste, Claude, Gal, Naomi, Marina,
Mickael, Amstutz, Cdric, Graldine (2x), Martine, Ivan, Pascal, Pascale, Affoltre,
Christophe, Anne, Nicolas, Ekim, Tiziano, Yannick, Brnice, Sophie, Laurent,
Rodrigo, Paul, Samuel, Raphal, Guillaume, Clara-Marine, Clmentine,
Cristobal, Julien, Philipp ... et jen oublie.
A mes amis dhydrogologie : Franz, Romain, Michiel, Ludovic, Daniel, Vincent,
Stphane, Franois-David, Franois Z, ...
A toute lquipe du Garage Blaser pour mon initiation la mcanique.

A tous les Marocains, amis de longue date ou dun instant, pour leur chaleur et
leur hospitalit. Hassane, Abdulla, Redoine, Tarik, Smal, Mohammed, Ahmed,
Mansour, Mehdi, Somia, Kaoutar, Isham, Amine, Adnane, Abdulwahed, Aziz, ...
A mes parents
A Rachel
98
Thse de doctorat
A
DONNEES BRUTES
DE LA FIGURE 22
PETITS PLIS
Strogrammes jaunes
Plan de stratification :
Donnes : Azimut (000-360) et Plongement (00-90) de la ligne de plus grande
pente.
Reprsentation sur les strogrammes : Ple du plan, hmisphre infrieur.
Kreb Lethel
356
306
024
311
002
294
358
032
002
338
287
331
018
005
38
42
15
55
39
34
42
60
64
28
32
20
10
61
69
47
17
30
21
26
23
069
053
046
225
226
228
198
173
207
253
005
358
033
060
309
341
045
048
082
198
248
Aoufital
058
054
043
044
247
059
079
056
225
075
064
081
154
243
253
097
062
228
042
045
178
N3120 / W00350
54
54
68
43
53
41
38
79
71
46
52
59
56
64
004
030
240
031
306
344
324
66
60
64
77
47
53
49
330
339
012
017
002
46
52
48
45
43
10
45
19
08
23
44
46
24
25
18
26
15
18
51
56
14
23
48
54
25
21
208
039
202
055
225
112
075
254
079
077
215
228
234
087
074
057
230
211
043
221
224
39
35
27
22
33
21
40
12
34
24
19
18
52
36
39
36
51
22
55
40
39
N3055 / W00405
Thse de doctorat
30
19
28
20
28
35
30
16
72
38
18
28
22
30
18
19
22
51
24
22
27
035
062
069
149
052
073
094
109
166
168
222
049
075
039
230
187
218
201
049
095
153
Annexe
083
057
044
050
047
054
217
11
39
40
31
78
60
16
Tamjrani
042
047
229
228
066
078
083
190
234
063
063
071
074
124
168
216
12
41
20
40
19
37
58
207
061
052
044
209
212
055
56
49
44
38
48
53
26
217
057
225
064
063
21
44
55
37
20
68
56
33
20
39
65
71
74
53
56
60
28
20
29
34
63
053
087
121
152
166
172
228
192
067
204
103
089
075
219
229
34
22
23
32
46
44
59
22
28
33
47
40
59
55
73
23
34
30
43
66
42
60
15
57
42
15
34
14
13
20
30
49
50
19
22
45
44
45
40
215
029
042
045
043
233
221
215
213
039
214
060
041
043
022
111
074
239
258
172
121
40
86
84
59
66
88
76
47
48
49
10
06
53
49
37
30
59
24
35
63
30
N3100 / W00410
60
61
83
32
48
37
29
22
37
34
55
53
26
17
15
43
Bou Tchrafine
173
161
060
057
073
125
190
192
195
043
189
039
207
197
155
212
044
038
222
209
107
058
228
111
231
252
053
236
053
231
049
26
38
51
75
32
25
40
42
39
57
24
21
60
38
10
85
46
32
42
28
29
87
63
14
227
225
057
226
049
048
067
142
197
222
019
062
228
193
054
227
79
69
64
84
47
78
30
14
21
38
24
74
69
25
53
84
217
062
207
090
241
051
228
054
237
091
226
191
136
075
051
048
N3125 / W00410
149
186
214
346
093
224
167
142
213
149
122
127
229
077
062
053
065
063
215
212
214
219
179
073
25
32
48
28
63
53
27
22
36
34
31
37
60
33
63
22
16
19
42
48
54
37
20
32
078
059
052
045
029
033
025
038
019
359
220
189
152
190
099
206
236
230
158
097
227
212
232
235
Thse de doctorat
Jebel Irhs
171
205
199
206
059
103
181
197
193
196
048
064
207
235
231
225
223
111
086
103
156
207
225
226
095
224
217
040
031
189
161
045
047
201
204
025
040
N3120 / W00420
35
53
58
77
53
24
44
18
20
12
71
25
71
29
47
73
57
29
67
78
29
32
41
30
09
47
30
43
29
64
13
26
46
46
83
84
72
Jebel Rheris
253
316
265
222
195
258
261
58
72
61
50
78
68
81
Msissi est
042
019
Annexe
046
041
037
146
110
220
227
152
218
226
041
089
205
203
172
124
084
059
054
055
052
056
057
220
205
203
215
175
193
139
029
112
077
074
117
116
242
21
54
45
18
24
44
64
13
59
82
38
22
42
47
11
12
10
19
20
26
28
23
21
37
36
35
87
18
46
36
67
19
33
36
13
05
20
238
249
186
242
090
276
093
272
090
097
077
065
208
242
311
175
102
235
113
057
057
188
141
080
207
044
125
051
233
051
218
209
088
048
066
202
091
32
41
34
42
08
48
60
58
42
30
33
29
06
20
04
30
43
55
22
39
58
28
27
38
23
35
12
41
70
79
53
47
55
56
45
36
10
064
069
060
110
105
208
042
240
045
062
035
019
239
003
280
229
260
023
281
255
244
254
074
221
075
243
252
059
063
063
061
228
228
210
19
31
61
20
31
83
43
40
43
78
27
20
82
40
47
61
43
49
21
48
45
53
80
18
58
24
40
51
62
59
49
22
50
53
57
65
77
82
46
45
54
010
019
351
327
287
269
013
62
60
49
45
65
63
65
55
84
299
005
60
67
N3120 / W00440
256
255
258
279
257
288
295
52
56
52
50
64
49
60
270
234
312
177
270
276
338
N3120 / W00445
79
62
Thse de doctorat
049
265
70
70
352
051
Annexe
258
318
018
31
26
58
Msissi ouest
247
331
357
279
342
227
355
108
180
013
297
257
11
65
197
045
07
25
21
Tazzarine est
201
225
212
21
15
18
Tazzarine nord
243
326
202
184
181
197
214
17
09
09
25
12
19
16
Nekob nord
42
64
72
234
088
116
035
74
23
19
44
51
19
30
200
192
175
050
056
042
74
65
29
33
45
44
223
020
211
016
342
219
77
43
51
52
20
59
45
25
041
204
64
44
211
42
46
50
17
232
220
213
29
21
12
226
28
28
19
42
24
17
17
68
324
333
320
321
277
52
25
52
49
27
182
63
54
40
24
N3045 / W00520
163
166
234
12
08
55
055
197
215
N3050 / W00525
168
173
351
307
175
168
198
27
25
59
09
23
19
25
187
205
339
334
278
203
158
N3105 /W00545
71
76
88
22
22
17
24
011
185
148
159
78
59
38
228
194
206
10
13
14
N3045 / W00605
Tamnougalt
012
345
31
37
60
N3120 / W00455
10
40
61
Achmas
141
164
153
156
270
283
265
N3110 / W00450
Toufassemane
143
156
148
30
61
69
N3115 / W00450
41
42
46
17
29
70
Fezzou
156
210
005
015
002
69
23
68
70
50
27
172
022
157
011
26
68
21
25
N3010 / W00625
002
183
44
67
010
003
38
40
Thse de doctorat
Agdz
354
358
001
011
349
349
338
N3045 / W00630
82
71
63
37
42
28
22
Agdz nord
172
155
17
19
Touroug
308
298
280
25
20
52
11
60
72
62
59
28
10
42
25
16
Tisdafine sud
003
058
022
146
084
60
18
38
70
33
Tisdafine ouest
006
001
171
358
001
178
59
52
50
55
50
78
Tineghir sud-est
009
060
13
43
42
49
37
19
174
156
338
020
004
010
53
03
11
23
29
33
017
006
007
179
186
188
50
59
89
68
80
65
35
14
57
072
282
273
41
11
54
27
39
27
29
156
162
002
013
45
59
49
41
18
14
27
74
52
127
138
335
158
35
64
17
38
36
48
57
39
27
130
149
092
009
31
47
22
40
39
49
27
82
59
172
156
082
027
006
58
40
35
50
87
34
09
105
112
72
83
N3050 / W00630
042
193
16
36
031
33
062
348
078
23
24
81
045
351
267
N3125 / W00510
Tizin Boujou
337
15
135
347
335
257
196
152
174
154
152
N3140 / W00440
Asdaf
039
153
152
344
352
Annexe
27
19
Thse de doctorat
353
055
044
154
52
23
19
59
015
013
025
034
N3125 / W00515
174
142
320
343
342
17
41
33
17
25
348
1
122
135
120
N3130 / W00520
014
015
014
164
158
46
68
37
43
89
107
021
025
036
106
N3110 / W00520
163
022
040
014
352
65
45
21
53
67
061
016
054
354
174
N3130 / W00525
021
051
21
09
182
236
Annexe
083
110
119
015
27
51
68
40
Tineghir sud
103
130
020
53
60
40
Tineghir SSW
240
302
342
351
30
20
40
83
174
072
173
178
85
12
80
54
033
334
028
094
26
50
19
67
356
009
044
129
55
33
24
45
65
39
15
015
046
015
63
50
50
33
58
40
20
162
065
027
57
37
56
37
27
55
23
32
328
335
294
200
24
25
18
64
N3135 / W00530
033
067
342
42
48
72
357
105
087
N3130 / W00530
201
195
204
195
62
57
35
48
295
012
003
132
Tineghir sud-ouest N3125 / W05540

003
341
003
354
341
22
29
29
26
38
346
165
353
349
228
88
77
12
28
21
272
304
215
299
340
CHEVAUCHEMENTS
Strogrammes bleus
Plans de chevauchements :
Donnes : Azimut et plongement de la ligne de plus grande pente.
Reprsentation sur strogrammes : Grand cercle, hmisphre infrieur.
Stries :
Donnes : Azimut et plongement de la direction de la strie.
Reprsentation sur strogramme : Direction de la strie, hmisphre infrieur.
Jebel Irhs
N3120 / W00420
Plans de chevauchement
100
182
078
32
26
61
227
056
159
64
85
30
064
138
126
35
45
15
095
231
30
40
28
23
60
230
074
332
41
02
64
082
068
077
29
34
13
267
36
Stries
078
201
063
Thse de doctorat
Msissi
Annexe
N3110 / W00450
074
068
152
36
33
35
233
047
226
40
71
54
278
051
273
68
54
52
050
59
27
29
39
207
044
070
24
69
51
278
230
295
68
47
20
032
57
Stries
035
032
215
Tazzarine nord
N3050 / W00525
133
113
27
29
114
123
17
12
101
132
60
33
261
332
18
33
27
26
137
140
15
12
116
144
59
33
321
317
08
32
Stries
135
152
Achmas
N3045 / W00605
222
176
25
34
001
182
52
44
018
183
70
50
192
011
53
25
35
39
022
200
48
40
038
218
68
42
198
002
49
23
Stries
206
182
Tineghir sud
N3135 / W00530
119
046
343
002
170
65
20
45
31
31
005
332
338
341
45
72
59
21
321
105
160
339
46
23
10
11
331
352
325
302
40
79
50
24
61
03
36
28
30
355
279
016
328
44
65
55
20
002
159
162
337
38
14
10
11
021
020
296
291
28
77
48
23
Stries
165
326
299
331
164
Thse de doctorat
Annexe
Tineghir sud-ouest
N3125 / W05540
353
341
348
010
19
29
29
49
092
013
321
068
24
27
13
30
339
348
295
349
30
21
23
47
004
017
327
256
35
23
19
20
17
28
27
45
157
333
334
165
09
21
12
01
350
346
334
329
29
21
21
43
330
326
321
324
28
15
18
09
Stries
342
344
334
338
PENDAGES GENERAUX
Strogrammes verts
Plans de stratification :
Reprsentation sur strogrammes : Ple du plan, hmisphre infrieur.
Aoufital
019
036
030
039
047
041
054
057
073
070
064
041
055
079
N3055 / W00405
25
24
27
29
36
32
17
22
17
19
18
23
26
12
050
043
028
069
068
074
082
031
057
037
039
042
051
063
20
22
26
17
27
40
29
19
19
18
15
18
18
17
29
21
25
24
30
37
30
123
122
145
116
053
253
230
14
50
20
17
35
17
18
Msissi
126
129
130
130
146
131
123
065
073
052
065
076
091
072
037
064
049
073
038
058
071
23
23
20
22
20
22
30
41
20
33
08
29
29
24
033
032
050
067
054
087
054
033
099
094
057
138
083
058
21
18
09
18
18
20
21
08
17
19
16
39
20
32
137
199
141
108
134
146
120
20
30
14
14
34
11
11
N3115 / W00450
204
200
198
205
197
200
211
28
26
42
38
38
49
55
Thse de doctorat
122
112
120
100
082
106
089
101
096
090
067
121
13
20
11
16
37
18
19
17
22
16
39
12
229
208
208
208
209
213
219
231
191
209
211
28
26
31
24
23
18
19
15
18
28
29
30
14
09
17
17
15
12
141
125
164
150
256
194
12
10
17
12
09
15
07
38
11
116
354
336
07
61
69
Tazzarine nord
139
136
144
136
152
154
155
189
203
156
164
187
208
150
166
163
161
139
32
51
34
39
37
65
22
19
12
16
13
160
092
155
176
176
169
144
120
111
156
148
22
13
39
38
20
24
21
13
15
25
21
N3050 / W00525
Tazzarine
163
005
142
Annexe
303
136
306
138
139
181
03
08
04
18
11
09
260
137
146
165
149
155
16
35
26
11
10
27
N3045 / W00535
116
134
142
18
25
21
149
37
STRUCTURES EXTENSIVES
Strogrammes rouges
Failles :
Reprsentation sur strogrammes : Grand cercle, hmisphre infrieur.
Stries :
Donnes : Azimut et plongement de la direction de la strie.
Reprsentation sur strogramme : Direction de la strie, hmisphre infrieur.
Tizin Boujou
N3120 / W00515
Failles
339
40
333
39
328
45
322
41
328
47
40
054
09
334
44
338
40
049
03
Stries
334
Thse de doctorat
Annexe
Tineghir sud-ouest
N3125 / W05540
Failles
190
53
161
38
137
89
138
55
164
64
71
141
35
225
37
196
51
220
37
Stries
176
Plans de stratifications des synclinaux associs une faille normale :

Reprsentation sur strogramme : Ple du plan, hmisphre infrieur.
Ifarherioun
153
180
032
139
146
188
175
181
023
210
039
031
007
024
027
029
013
049
049
018
29
80
51
28
39
50
52
35
80
29
54
46
60
51
50
47
34
53
54
73
N3110 / W00415
032
202
219
331
343
330
333
023
012
002
184
193
221
171
187
189
008
181
173
180
33
37
44
29
33
43
47
66
76
29
72
38
54
61
48
56
53
50
62
33
Msissi
356
359
10
31
54
64
59
26
67
50
07
44
71
12
33
88
51
36
51
73
67
72
50
012
016
026
016
012
351
277
157
176
103
263
250
219
180
178
345
207
205
040
050
37
33
59
51
27
24
34
18
19
22
26
17
47
59
64
31
43
42
33
33
034
061
006
004
028
007
043
051
038
033
040
031
022
030
349
037
173
354
45
44
79
51
51
70
64
45
50
53
46
40
42
52
58
36
22
30
30
44
348
356
40
30
57
33
44
40
39
37
63
58
000
023
003
166
330
337
63
42
53
37
42
59
N3110 / W00450
59
31
339
002
17
19
176
354
142
348
176
178
140
339
17
39
51
50
57
41
09
21
Jebel Morf
186
305
168
167
163
187
138
156
335
170
194
185
059
202
032
001
182
019
352
191
192
247
182
169
008
004
354
010
22
41
32
27
19
18
35
20
009
155
13
52
148
358
N3045 / W00510
338
356
315
003
004
017
353
339
32
41
27
31
34
33
50
78
013
009
006
008
022
325
007
354
Thse de doctorat
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.
- Inversion tectonics in the eastern Anti-Atlas of Morocco. Soumis le 05.09.2006

au AAPG Bulletin.
- 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.
- Projet dun troisime article.
Thse de doctorat
Variscan inversion tectonics and interference pattern

in the Tafilalt, Anti-Atlas, Morocco
Charles Robert-Charrue, Martin Burkhard*
1
Institut de Gologie et dHydrogologie, Universit de Neuchtel,
11, rue Emile-Argand, CH-2009 Neuchtel, Switzerland

*Corresponding author. Tel. : +41 32 718 26 52 ; fax : +41 32 7182601.
E-mail address : [email protected]
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 (Piqu et al., 1999, Piqu, 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. Introduction
The Moroccan Anti-Atlas belt could be regarded
as a fold-and-thrust belt in the hinterland of
the Appalachian orogeny as the thin-skinned
Alleghenian chain in the western foreland. The
lack of major basal dcollement, deformation
front or thrust fault makes the Anti-Atlas an
unusual type of fold belt, which does not easily
fit with classic schemes of ramp flat folding style,
nor does it have any obvious analog in other well
known orogens. The relationship between the
basement involvement in the large inliers and
the gentle folds of the surrounding cover has not
found a satisfactory tectonic explanation sofar.
The aim of this study is to elucidate the tectonic
evolution of this uncommon chain in terms
of basement / cover relationships. Is the AntiAtlas primarily thick skinned or is the basement
involvement a late stage of an ordinary thin
skinned belt ?
Detailed geological mapping has been conducted
in the eastern Anti-Atlas of the Tafilalt area in
order to characterize the type and orientation
of folds, faults and successive generations of
paleo-stress axes from fault/striae populations.
The interpretation of a regional grid of reflexion
seismic lines from ONAREP (National Office
for Research and Oil Exploitation) provided
additional constraints for the establishment
of cross-sections and to establish a regional
kinematic model.
2. Geological framework
2.1. Location and setting
The Anti-Atlas of Morocco is part of the larger
Appalachian Orogen system, time equivalent with
the Alleghenian front to the West. The Anti-Atlas
appears as a huge anticlinorium oriented NESW (Choubert and Faure-Muret, 1971). Locally,
the basement is cropping out as inliers, but the
main part is made of an important Paleozoic
cover, which is gently folded and of low grade
metamorphism. The Anti-Atlas belt (Fig. 1) is
located along the northern border of the WestAfrican Craton (WAC). It is limited to the South
by the Tindouf basin, unaffected by late Paleozoic
folding, and to the North by the High-Atlas where
Meso-Cenozoic sediments are involved in strong
Cenozoic-lower Quaternary deformations. The
2
Fig. 1. Geographic situation and geological context of the

Anti-Atlas. AAMF: Anti-Atlas Major Fault ; SAF: South
Atlas Fault ; SGA: Saoura gravimetric anomaly.
Anti-Atlas s.s. is laterally limited by the Atlantic

Ocean to the South-West and disappears eastward
below the Hamada in the Tafilalt region. On a
larger scale, the Anti-Atlas can also be seen as an
arch (or rather a recess) continued to the SouthWest into the Zemour and Mauritanides thrust
belts and to the South-East into the Ougarta fold
belt (Haddoum et al., 2001). The studied area, the
Tafilalt, is the eastern end of the Anti-Atlas s.s..
Structurally, this area differs markedly from the
western part of the Anti-Atlas chain (Soulaimani
et al., 1997, Helg et al., 2004, Cartig et al., 2004).
In these semi-arid regions, a Landsat 7 view
can be used as a geologic map (see the succinct
description, Fig. 2).
2.2. Stratigraphic units and lithology
Basement
In the Tafilalt area proper, basement outcrops
are very sparse, limited to a few tiny outcrops of
rhyolites and andesites of late P III, located along
the north-east edge of the Tafilat half-window, 30
km east of Erfoud. An larger inlier is certainly
hidden below the Cretaceous and Neogene
sediments of the Hamada, in prolongation of an
ENE alignment of inliers further west.
The crystalline basement has a complex history.
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.
It was last consolidated during the Panafrican

orogeny. Exposed further west in the Bou Azzer
inlier, the Major Anti-Atlas Suture or AntiAtlas Major Fault (AAMF) (named Accident
Majeur) was interpreted as the old border of the
african shield (Choubert and Faure-Muret, 1971).
Recent workers, however, considered it as part
of the metacratonic evolution and the northern
boundary of the WAC is now defined further
north, along the South Atlas Fault (Hefferan et al.,
2000, Ennih and Ligeois, 2001, 2003, Ligeois
et al., 2005), coincident with a major Neogene
tectonic feature, the southern deformation front of
the High-Atlas chain. At the end of Proterozoic,
a major unconformity is interpreted as the onset
of a new rifting cycle (of the Paleozoic era). The
successive openings of Rheic and Paleotethys
oceans (Stampfli and Borel, 2002) along the NW
border of the WAC must have left their imprint in
the area of the future Anti-Atlas belt. Rifting in
this area was aborted early, but it is structurally
very important since lead to the preconfiguration
of blocks which were inverted during the Variscan
orogeny. Syn-rift sediments of the P II-III unit
are mainly volcaniclastics, but the thickness of
this unit is very thin in the Tafilalt compared
to the western Anti-Atlas. Traditionally, from a
structural point of view, the basement englobes
the crystalline basement and the P II-III unit.
Paleozoic cover
Throughout the Paleozoic, the Anti-Atlas basin
is dominated by shallow marine conditions. In the
western part the series reaches 12 km thickness,
while in the Tafilalt it is only about 6 km thick.
The major part of the Paleozoic sediments are
made of fine grained detrial shales, siltstones
and mudstones (Fig. 3). Competent marker beds
are subordinate in abundance and thickness, and
distributed among the series. These relatively
thin competent beds appear morphologically as
cuestas and some of these crests can be followed
throughout the Anti-Atlas range. The local name
of each crest has given the name to the formations
(Ordovician Jebel Bani, Devonian Jebel Rich,
Carboniferous Jebel Ouarkziz). These formation
names are not formally accepted lithostratigraphic
terms but they are very useful, widely known and
applied by all the workers in the region.
In the studied area, the whole Adoudounian series
(local formation name for Cambrian from Basal
Conglomerates to Upper Limestones) is lacking
4
from the stratigraphy of the Tafilalt. Above some

thin basal conglomerates, the first competent layers
to appear are the Cambrian Green Sandstones.
The Ordovician 1st Bani is very thin compared
to the Zagora region and the major Ordovician
cuestas, creating the topographic heights in the
Tafilalt, can be linked to the 2nd Bani, mostly made
of quartzites and sandstones. The Silurian shales
are very weak and outcrops are exceedingly rare.
They are responsible for large valleys inbetween
the two cuestas of the 2nd Bani and the Devonian
Richs. However, a single limestone marker bed in
the middle of the Silurian shales is well known
for its echinoderms and some graptolithes at its
base. This bed can be seen as the first member of
a carbonate series that follows in the Devonian,
in contrast to the first half of the Paleozoic
series which is mainly detrial. The Devonian
series is dominated by fossiliferous blueish dark
carbonates. To the delight of the Filallis (Tafilalt
inhabitants), slabs and handicrafts from this area
are exported worldwide. The Upper Devonian is
marked by a return to detrital sedimentation that
dominates throughout the Carboniferous where
marker beds are fluvial and/or deltaic sandstones
in contrast to the carbonates found further west.
The structural style of the cover is dictated by two
key-parameters, namely: the total thickness and
the incompetent / competent ratio.
2.3. Variscan orogeny
The closure of Paleotethys leads to the collision
of Gondwana and Laurussia to form the
supercontinent Pangea. The former normal
faults are reactivated and inverted. Because of
its particular incompetent rheology, the Paleozoic
cover folds. For this reason, the deformation is
characterised by a gradient rather than by a major
thrust-front. The timing of the inversion in the
Anti-Atlas is hard to constrain since the oldest
formation sealing the deformed Carboniferous
sediments is of Upper Cretaceous age. Dolerites
dykes and sills postdate the Variscan folding and
occur between 206 and 195 Ma (Sebai et al.,
1991) . However deformation can be estimated
to late Carboniferous between 320 and 300 Ma
according to the new paleoreconstructions of
Stampfli and Borel (2002).
2.4. Post Variscan
The next Wilson cycle begins with tholeitic
magmatism at the end of Triassic, which is the
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

Atlantic ocean. The new Mesozoic High-Atlas
basin opens just north of eastern Anti-Atlas.
This extension affects the Anti-Atlas by the
intrusions of dolerite dykes and sills and by
normal faulting. This normal faulting corresponds
to a third re-activation of faults inherited from
the late Proterozoic (Piqu and Laville, 1996)
(post-tectonic intrusion of dykes between 206
and 195 Ma (Sebai et al., 1991)). The AntiAtlas Major Fault is essentially Precambrian,
but must be reactivated in the Mesozoic since
it produces the graben from Zagora to Taouz in
the Paleozoic sedimentary cover. No Permian to
Jurassic sediments are found in the Anti-Atlas.

The structural basement of the High-Atlas is of
Paleozoic age. It is mostly covered by a thick
deformed Mesozoic series but must include
Variscan deformations. So the northern part of
the eastern Anti-Atlas is hidden under the HighAtlas.
The actual high topography of the Anti-Atlas (up to
2500 m) does not result from the Variscan orogeny
alone. The Cretaceous and Neogene hamada east
and south of Tafilalt is clearly tilted to the South
and in the north of Tafilalt the Cretaceous hamada
is tilted to the North-East (see Fig. 2, on the DEM
5
and the river patterns on the Landsat view). The

scarp retreat analysis of Schmidt (1992) indicates
a beginning of uplift in late Eocene times with an
increase in late Miocene / early Pliocene times.
These geomorphological interpretations fit nicely
with structural observations in the High-Atlas and
with an analysis of the deformation of Neogene
sediments (Grler et al., 1988). The deformations
of the High-Atlas during Eocene Oligocene and
mainly Pliocene Pleistocene epochs (Frizon de
Lamotte et al., 2000, Ellouz et al., 2003) just a
few kilometres north, affects also the adjacent
strongly compartmentalized Anti-Atlas basement.
This rejuvenated relief could come partly from
a Neogene thermal uplift (Anguita and Hernan,
2000, Saddiqi et al., 2005, Teixell et al., 2005).
3. Field data
3.1. Small scale structures
The dominant structures in the study area are
apparently regular cuestas. On closer inspection,
however, internal imbrications and small scale
folds such as those shown in Fig. 4 are quite
common. These folds and thrusts are highly
localised. The spacing is on the order of 500
to 1000 m, with virtually undeformed regular
intervals inbetween. Structural perturbations are
not only laterally localised but seem to be confined
to competent horizons within the stratigraphy.

Below and above a folded or thrust competent
layer, the next immediate competent layers may
well be unaffected. In terms of structural style,
folding and thrusting occurs within the competent
bed which may be folded in one place and thrust
in an other locality. Folds in competent layers are
parallel class 1B folds and class 3 for the weak
interbeds, resulting in a characteristic chevron
fold pattern (Ramsay and Huber, 1987). Bedding
parallel slickenside surfaces can be found in
folded competent layers indicating flexural slip.
Thrust faults are very localised both laterally
and vertically within the stratigraphy. The
identification of contractional faults is limited to
thick competent layers; but wherever observed,
thrusts seem to be rapidly attenuated within
adjacent incompetent layers. In the right picture
of Fig. 4, the competent layers are seen to be
duplicate by a thrust while the weaker beds
above (hammer) are just folded. This example
demonstrates that in absence of competent
marker beds, the layer parallel shortening is hard
to constrain.
In contrast to the apparent ENE-WSW global
trend of the Anti-Atlas (Fig. 1), small scale
structures show a clear NW-SE trend as shown
on the stereogramms of Fig. 4. Note that the
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.
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 smallscale structures has been observed on southverging cuestas (hanging wall syncline). Tilting
those limbs back to horizontal also restores the
fold axes to a sub-horizontal attitude. This is a
strong argument for these small-scale structures
to be older then the hanging wall synclines. This
applies to the entire Tafilalt area, which has been
affected by a NE-SW compression throughout.
This NE-SW compression direction contrasts
with the overall Anti-Atlas trend, but it fits nicely
with the structural trend of the Ougarta chain to
the South-East.
3.2. Hanging wall synclines
The synclines shown in Fig.5 have often been
mistaken as Variscan folds. Strike and dip
measurements of these meso-scale structures do
not fit with the results obtained from the smallscale folds and thrusts discussed above. Synclines
are typically some 5 km long, 2 km wide (half

wave-length) and are gently plunging to either
eastward or westward. Their repetition on a northsouth axis is quite intriguing (especially west of
Rissani, Fig. 2). These synclinal structures are
well developed only within the northern half of
Tafilalt, mainly because their surface expression
is related to the underlying Devonian carbonate
beds of the Rich (see the three northern red
squares of Fig. 2). In other words, the synclines
owe their existence to a particular mechanical
stratigraphy rather then a geographically distinct
part of the Tafilalt basin. The same extension is
expressed in a different way. Ordovician and
Cambrian sandstones and quartzites are too thick
and too competent for such gentle fold structures
to develop. Silurian and Carboniferous shales, in
the other hand, are too incompetent. Hanging wall
synclines appear discreetly within Carboniferous
sandstones (red square south-east of the sand
7
dunes on Fig. 2). In fact they are extensional

fault-related folds.
A detailed study of the Jebel Amelane region
revealed a bimodal distribution of the bedding
plane orientations (see stereogramm of the
bedding poles, Fig. 5). This is interpreted to be
a result of kinked or trisheared layers on a
normal fault as described in the northwestern
Red Sea (Khalil and McClay, 2002). The fact
that the synclinal axes are plunging to the east,
is explained by the super-structure. The Jebel
Amelane is located on the eastern flank of a huge
anticlinorium. The most important evidence for
extensional fault-related folding is stratigraphic,
however. Minor normal faults do appear but
the major fault zones are mostly obscured by
a thick soft interval of Silurian shales. Major
fault zones can be directly observed only where
the slip places more competent layers side by
side as for instance at the Jebel Bou Ifarherioun
(red square, 15 km south of Rissani on Fig. 2).
The dip of the faults seems to vary according
to the competence, flat in incompetent, steep in
competent layers. The localisation and repetition
of these faults is interpreted as due to pre-existing
faults in the basement. Hanging wall synclines
show flexural-slip movement along bedding
plane surfaces, which can be misleading since
the relative shear sense of such layer parallel slip
is always indicting thrusting independent of
the underlying mechanism of folding in either
compression and/or extension. These extensional
fault-related folds postdate the small scale folds
by folding their axis.
3.3. Superstructures
The eastern Anti-Atlas shows interference of
two superimposed superstructures. First, the
ENE-WSW anticlinorium trend of the entire
Anti-Atlas belt is materialized by the large scale
shape of basement inliers (Fig. 1). The Hamada
half-window apparently sealed to by Upper
Cretaceous sediments to the east, inherited this
ENE-WSW trend (Fig. 1 and 2). In fact the large
scale Anti-Atlas superstructure appears more like
a huge monocline. From NNW to SSE rocks get
increasingly younger, from basement inliers,
across the folded Paleozoic cover, gradually
passing into the flat lying basin fill of the Tindouf
basin, partly covered by younger sediments of the
Hamada. The northern limit of this superstructure
is defined by the thrust front of the inverted
8
Mesozoic sediments of the High-Atlas. The

contribution of High-Atlas inversion and the
thermal uplift in the exaggeration of this general
large scale Anti-Atlas trend is difficult to quantify
(Helg et al., 2004). Secondly, the Tafilalt region is
clearly affected by folding along Ougartian NWSE trend. This trend is materialized by the global
antiform axis linking the Ougnat inlier to Taouz,
separating the synforms of Tafilalt and Mader,
and disappearing under the Hamada (Fig. 2). This
large scale orientation is identical with the one
measured in small scale structures found in the
entire Tafilalt (3.1.).
While these two orientations are clearly
distinguished in terms of their geometry and
interference patterns, the relative chronology
remains open to debate. In the Tafilalt, both
structural trends are of similar, weak overall
shortening intensity (estimated to ca. 10%) and
intuitively either structure could be first or second.
Further west, however clear evidence is found for
a relative chronology between folds of the AntiAtlas trend preceding the Ougartian trend (Caritg
et al., 2004). Our own, unpublished observations
in the Tata area have since provided additional
strong evidence for such a relative chronology.
On the other hand rather than separate these two
deformations in time, an alternative explanation
could be found in the shape of the WAC
boundary. The eastern Anti-Atlas is located along
the northern edge of the WAC metacraton. It is
limited by the South Atlas Fault (ENE-WSW) to
the North and by the Saoura gravimetric anomaly
(NW-SE) (Ennih and Ligeois, 2001) to the
East. On the very large scale, the arc defined by
the Ougarta Anti-Atlas system thus seems to
be an inherited feature, mimicking the shape of
the WAC. The late Paleozoic fold belts would
have been formed along the boundaries of this
craton, specifically along its metacratonic borders
which were once again re-mobilized in inversion
tectonics during the assembly of Pangea.
4. Seismic data
The ONAREP RS8 seismic reflexion line (Fig. 6)
crosses the Tafilalt with a north-south orientation
(see location on fig. 2). Accordingly, this line is
perpendicular to extensional folding and oblique
with respect to both superstructures : WSW-ENE
in the northern part and NW-SE in the southern
part. The interpretation of this profile has been
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.
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 dcollement at
the base of the Paleozoic sedimentary pile. Cover
shortening is compensated below the dcollement
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
of the hanging wall layers is observed in many
models specially for ultraweak cover on rigid
basement (Finch, 2004). The image obtained by
discrete-element modelling is very similar to this
seismic line. Secondly, the seismic line shows the
Two Ways Time and not the depth. The velocity
of Carboniferous shales and mudstones must
be particularly slow (probably less than under
2000 m/s), while competent layers (Devonian
limestones, quartzites) may attain 3000 m/s or
more. This relationship leads to a pronounced
velocity pull down, which tends to exaggerate the
depth of the pseudo basins.
The third part of Figure 6 is the presumed profile,
restored to illustrate the situation at the paroxysm
of the Variscan deformation, i.e. without the later
normal faults. The overall shortening is minimum
of 10 %, but this value is measured at regional
scale and does not include the layer-parallel
shortening observed at smaller scale. Helg et
al. (2004) argue that such distributed small
scale deformation add a substantial amount of
shortening of similar magnitude.
10
Fig. 7. Schematic illustration of the proposed kinematical

model (inspired from Bump, 2003). South-North
orientation perpendicular to inliers alignment. But parallel
to the Ougartian trend. a: Late Proterozoic rifting with synrift sediments; b: Deposition of Paleozoic cover; c: Variscan
inversion. The thrust fault takes a shortcut in the basement,
the cover reacts in a trishear mode. The Ougartian trend do
not appears on this sketch; d: New extension phase (Atlantic
opening). Normal faults short cut the Paleozoic cover, its
dips vary according to the competence of rocks and produce
hanging wall synclines.
5. Conclusions
The eastern Anti-Atlas structure is a complex
interference between two trends of Variscan
folding which lead to an egg box pattern of
basins and domes. This fold interference pattern
is subsenquently cut by a series of normal faults
most likely in response to the opening of the
Atlantic in Liassic times. The ENE-WSW AntiAtlas global trend is the best developed in terms
of deformation and inversion. The Anti-Atlas is
clearly thick skin. The Paleozoic cover reacts
in a trishear way (Fig. 7.c) to accommodate

basement shortening which is inverting former
late Proterozoic normal faults. The deformation
gradient decreases strongly from NNW to SSE,
with increasing distance from basement inliers.
The NW-SE Ougarta trend is materialized
throughout the Tafilalt by small scale folds and
thrust faults. This deformation has a important
thin skin component, detachment along weak
layers; although the underlying cause is equally
thick skin basement inversion. The interference
between these two trends is thought to be due
to the shape of the northern corner of the WAC,
and particularly a renewed re-mobilisation of
its weak meta-cratonic borders in the late
Paleozoic times. Former normal and inverted
faults are again reactivated during the Atlantic
opening event. This leads to the formation of
new steep normal faults within the Paleozoic
cover series. The overall weakness of the
thick Paleozoic series favours the formation of
extensional drape folding (Fig 7.d). The HighAtlas inversion and Neogene thermal uplift do
not significatively deform the eastern Anti-Atlas.
Only a light doming is underlined by the radial tilt
and retreat of cretaceous cuestas.
Acknowledgments
Ongoing support by the Ministre de lIndustrie
et des Mines, Direction de la gologie by M.
Dahmani and M. Haddane and the moroccan
national petroleum agency ONAREP is
gratefully acknowledged. Financial support by
Swiss National Science Foundation grants N
21-52516.97 / 20-63790.00 and by the Fonds
Matthey-Dupraz local fund for geology
students.
References
Anguita, F., Hernn, F., 2000. The Canary Island origin:
a unifying model. Journal of Volcanology and
Geothermal Research 103, 1-26.
Bump, A.P., 2003. Reactivation, trishear modeling, and
folded basement in Laramide uplifts: implications
for the origins of intracontinental faults. GSA Today,
4-10.
Caritg, S., Burkhard, M., Ducommun, R., Helg, U., Kopp,
L., Sue, C., 2004. Fold interference patterns in the Late
Palaeozoic Anti-Atlas belt of Morocco. Terra Nova 16,
27-37.
Choubert, G., Faure-Muret, A., 1971. Epoque hercynienne.
In: Tectonique de lAfrique, Sciences de la Terre,
Paris, 6, UNESCO, Paris, France, 353-371.
Ellouz, N., Patriat, M., Gaulier, J.-M., Bouatmani, R.,

Sabounji, S., 2003. From rifting to Alpine inversion:
Mesozoic and Cenozoic subsidence history of some
Moroccan basins. Sedimentary Geology 156, 185212.
Ennih, N., Ligeois, 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, 289302.
Ennih, N., Ligeois, J.-P., 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, 185-189.
Erslev, E.A., 1991. Trishear fault-propagation folding.
Geology 19, 617-620.
Fetah, M., Bensad, M., Dahmani, M., 1986. Carte
gologique du Maroc 1:200000 Tafilalt-Taouz, Notes
et Mmoires du Service Gologique du Maroc N244.
Finch, E., Hardy, S., Gawthorpe, R., 2004. Discrete-element
modelling of extensional fault-propagation folding
above rigid basement fault blocks. Basin research 16,
489-506.
Frizon de Lamotte, D., Saint Bezar, B., Bracne, R.,
Mercier, E., 2000. The two main steps of the Atlas
building and geodynamics of western Mediterranean.
Tectonics 19 (4), 740-761.
Grler, K., Helmdach, F.-F., Gaemers, P., Heissig, K.,
Hinsch, W., Mdler, K., Schwarzhans, W., Zucht, M.,
1988. The uplift of the central High Atlas as deduced
from Neogene continental sediments of the Ouarzazate
province, Morocco. In: V.H. Jacobshagen, SpringerVerlag (Ed.), The Atlas System of Morocco. Lecture
Notes Earth Sci. 15, 361-404.
Haddoum H., Guiraud, R., Moussine-Pouchkine, A., 2001.
Hercynian compressional deformations of the AhnetMouydir Basin, Algerian Saharan Platform: far-field
stress effects of the Late Palaeozoic orogeny. Terra
Nova 13, 220-226.
Hefferan, K.P., Admou, H., Karson, J.A., Saquaque, A.,
2000. Anti-Atlas (Morocco) role in Neoproterozoic
Western Gondwana reconstruction. Precambrian
Research 103, 89-96.
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.
Khalil, S.M., McClay, K.R., 2002. Extensional fault-related
folding, northwestern Red sea, Egypt. Journal of
Structural Geology 24, 743-762.
Ligeois, J.-P., Fekkak, A., Bruguier, O., Errami, E., Ennih,
N., 2005. Lower ediacaran age (630-610 Ma) for
the Sarhro group. Implications for the metacratonic
evolution of the Anti-Atlas, correlations with the
Tuareg shield and the evolution of the pan-african
orogeny in the NW Africa. 4me
Colloque
International 3Ma, Agadir 2005, Abstract Volume, 2.
NASA and Earth Satellite Corporation, Landsat 7 images
download website (https://zulu.ssc.nasa.gov/mrsid/ ),
last visited 14.03.2006.
ONAREP (Office National de Recherches et dExploitations
Ptrolires) Morocco. Seismic reflexion line RS 8.
11
Piqu, A., Laville, E., 1996. The central atlantic rifting :

reactivation of Palaeozoic structures ?. J. Geodynamics
21 (3), 235-255.
Piqu, A., Bouabdelli, M., Soulamani, A., Youbi,
N., Iliani, M., 1999. Les conglomerates du P III
(Noprotrozoque suprieur) de lAnti-Atlas (Sud du
Maroc): molasses panafricaines, ou marqueurs dun
rifting fini-protrozoque ?. C. R. Acad. Sci. Paris,
Sciences de la Terre et des plantes 328, 409-414.
Piqu, A., 2003. Perspective, Evidence for an important
extensional event during the latest Proterozoic and
Earliest Paleozoic in Morocco. C. R. Geoscience 335,
865-868.
Ramsay, J.G., Huber, M.I., 1987. The techniques of modern
structural geology, vol. 2: Folds and Fractures,
Academic Press Ltd. London, pp. 462.
Saddiqi, O., Ziad, N., Missenard, Y., Frizon De Lamotte,
D., Goffe, B., Leturmy, P., El Mamoun, R., Sebti, S.,
Poupeau, G., 2005. Age Miocne moyen suprieur
du soulvement thermique du Haut-Atlas Central :
apport des traces de fission sur apatite. 4me Colloque
International 3Ma, Agadir 2005, Abstract Volume,
45.
Schmidt, K.-H., 1992. The tectonic history of the PreSaharan depression (Morocco) a geomorphological
interpretation. Geologische Rundschau 81 ( 1), 211219.
12
Sebai, A., Feraud, G., Bertrand, H., Hanes, J., 1991. 40Ar/
39
Ar dating and geochemistry of tholeiitic magmatism
related to the early opening of the central Atlantic rift.
Earth and Planetary Science Letters 104, 455-472.
Service Topographique du Maroc, Aerial Photography, ST
1/40000, Mission 4-9-73, N 16_10 / 16_11.
Soulamani, A., Le Corre, C., Farazdaq, R., 1997.
Dformation hercynienne et relation socle/couverture
dans le domaine du Bas-Dra (Anti-Atlas occidental,
Maroc). Journal of African Earth Sciences 24 (3), 271284.
Soulamani, A., Piqu, A., Bouabdelli, M., 2001. La srie
du PII-III de lAnti-Atlas occidental (Sud marocain)
: un olistostrome la base de la couverture postpanafricaine (PIII) du Protrozoque suprieur. C.R.
Acad. Sci. Paris, Sciences de la Terre et des plantes
332, 121-127.
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, 1733.
Teixell, A., Ayarza, P., Zeyen, H., Fernandez, M.,
Arboleya, M.-L., 2005). Effects of mantle upwelling
in a compressional setting: the Atlas Mountains of
Morocco. Terra Nova 17, 456-461.
Inversion tectonics in the eastern

Anti-Atlas of Morocco
Charles Robert-Charrue, Martin Burkhard*
Institut de Gologie et dHydrogologie, Universit de Neuchtel,
11, rue Emile-Argand, CH-2009 Neuchtel, Switzerland
*Corresponding author. Tel. : +41 32 718 26 52 ; fax : +41 32 7182601.
E-mail address : [email protected]
ACKNOWLEDGMENTS
Ongoing support by the Ministre de lIndustrie et des Mines, Direction de la gologie by M.
Dahmani and M. Haddane are gratefully acknowledged. Financial support by Swiss National
Foundation grants N 21-52516.97 / 20-63790.00.
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-andthrust belt.
INTRODUCTION
The aim of this study is to elucidate the
structure of the Eastern Anti-Atlas belt and to
understand the structural style and its primary
controls which govern the deformation. A
specific look at the depth to detachment
is used to propose a large scale model of
the tectonic evolution of this chain through
time. Detailed field mapping and structural
analyses lead us to new a interpretation for
what was for a long time considered as a
fold and thrust belt. Actually, the absence of
dcollement or thrust at the basement-cover
contact is one of the particularities of the AntiAtlas fold belt.
The Paleozoic Anti-Atlas fold belt of southern
Morocco is located at the northern rim of the
West African Craton. Northward, the AntiAtlas is juxtaposed against the Cenozoic
High Atlas mountains along the so called
South Atlas Fault (Figure 1). Southward, the
Anti-Atlas forms the northern border of the
Tindouf basin, filled with the same Paleozoic
series, but not involved in Variscan orogeny.
The Anti-Atlas appears as a huge antiform
punctuated by a series of basement inliers
along a general a ENE-WSW axis (Figure
1). Southeastward, the Anti-Atlas belt is
connected by the NW-SE oriented Algerian
aulacogenic Ougarta chain. In the study
area the two chains intersect, leading to
some intricate large scale interference
pattern. The Anti-Atlas WSW-ENE trend is
dominant, however, and in this paper, the
subordinate Ougartan NW-SE trend will not
be considered.
Figure 2 shows a Landsat scene of the
Eastern Anti-Atlas. Modern false color satellite
images in these arid regions compete with
geologic maps in detail and accuracy. The
tracks of cross sections have been chosen at
a high angle to the structural grain defined by
the large basement inliers cropping out along
the northern border of the chain.
ANIT-ATLAS THROUGH TIME
Basement of the Anti-Atlas chain belongs to
the West African Craton (WAC). Core parts
of this craton consist in granites and high
grade metamorphic rocks of the so-called
2
Figure 1. Situation of The Anti-Atlas.
Eburnean, 2 Ga orogeny. The final assembly

and consolidation took place during the
Panafrican event from 650 to 580 Ma. The
Anti-Atlas Major Fault (AAMF), running in a
WNW-ESE direction across the Anti-Atlas
belt, is the site of a panafrican suture zone
with the accretion of an oceanic island arc
and associated granite intrusions (Ennih
and Ligeois, 2001; Ligeois et al., 2005).
The northern limit of the West African
Craton, however, is located further north
still and coincides with the South Atlas fault
SAF (Figure 1), a long lived feature, last reactivated in Late Neogene times (Frizon de
Lamotte et al., 2000).
Late Proterozoic volcaniclastic series of
the so-called P II and P III document the
end of the Panafrican orogeny (Benssaou
and Hamoumi, 2003) and the onset of
a new Wilson cycle. Evidence for active
rifting is found up to Middle Cambrian. P III
series of generally less than 1 km (0.6 mi)
thickness are widespread in halfgrabens and
patches of ill defined orientation all along
the northwestern border of the West African
Craton. In contrast, Cambrian series display
a well defined elliptical basin of WSW-ESE
orientation with up to 4 km (2.5 mi) thickness
in the Western Anti-Atlas south of Agadir
(Bouda and Choubert, 1972; Benssaou and
Hamoumi, 2003). Post-rift series of Middle
Cambrian to Middle Carboniferous build a
fairly regular layercake of shallow marine
Figure 2. Landsat view of the Eastern Anti-Atlas and cross sections localization. False coulor Landsat 7
downloaded from :< https://zulu.ssc.nasa.gov/mrsid/ >.
sediments with a total thickness of 12 km

(7.5 mi) in the western Anti-Atlas decreasing
to less than 8 km (5 mi) in the study area, i.e.
the Mader and Tafilalt. The Anti-Atlas is best
described as an intracratonic basin (Burkhard
et al., 2006); it can be regarded as the
northern, tectonized part of the much larger
Tindouf basin (Coward and Ries, 2003).
The basin fill is dominated by an abundance
of shales and silts deposited in a shallow
marine environment (see the stratigraphic log
on Figure 3). Competent and morphologically
outstanding marker beds are sandstones and
quartzites of the Cambrian and Ordovician.
Most of the Silurian is represented by soft
black shales, very rarely seen in outcrop.
Carbonate plateforms dominated the AntiAtlas sea in Lower Cambrian and again
in Devonian times. The return of detrital
conditions in Upper Devonian times could
be interpreted as the onset of convergence
in relation to Variscan tectonics. The
Carboniferous is mainly made of shales and
rare stiff beds are pericontinental deltaic
sandstones with an abundance of flute casts
figures and channel fills.
Variscan orogeny inversion
Most if not all the Paleozoic basins of Northern
Africa are involved to some degree in Variscan
tectonics in Carboniferous to Permian times
(Coward and Ries, 2003). The Anti-Atlas fold
belt has generally be considered as a foldand-thrust belt developed in the hinterland
of the larger Appalachian Variscan orogen.
It is indeed time equivalent to and bears
some structural resemblance with the
Alleghenian Valley and Ridge province (Helg
et al., 2004). The direct link between internal,
metamorphic zones of the Variscan chain
cropping out north of the South Atlas Fault
and the Anti-Atlas belt remains obscure,
however. No basal detachment has been
observed anywhere within the Paleozoic
cover series of the entire Anti-Atlas. The
thick skinned involvement of the basement at
very short distances behind the deformation
front have been used as arguments in favour
of a basin inversion- rather than a frontal
fold-and-thrust belt- deformation style
of the Anti-Atlas (Burkhard et al., 2006).
Anti-Atlas basin inversion is younger than
4
middle Carboniferous as documented by

the youngest sediments involved in folding
and tilting along the northern border of the
Tindouf basin. Anti-Atlas folds are cross-cut
by the widespread Late Triassic (200 Ma)
intrusions of dykes and sills (Sebai et al.,
1991), in relation with Atlantic opening. The
peneplained Anti-Atlas folds are sealed only
by a generalized marine transgression in
Upper Cretaceous times, however. Not far
to the north of the Anti-Atlas, rifting in Late
Triassic Early Jurassic times lead to the
opening of the High Atlas trough, filled with
thick series of Triassic to Jurassic sediments
(Manspeizer et al., 1978). This rifting event
lead to some extensional reactivation of
faults within the eastern Anti-Atlas. The major
event of Cenozoic inversion responsible for
uplift and folding of the High Atlas chain did
not lead to any noticeable folding and/or
faulting within the Anti-Atlas domain. This is
documented by a smooth configuration of
the basal Upper Cretaceous unconformity.
The Anti-Atlas chain is standing high,
however, and it has long been speculated
that this renewed uplift and relief could be in
response to alpine shortening deformation at
a crustal scale. Alternatively, however, high
present day topography in both the Anti- and
High Atlas have recently been interpreted as
being due to a large scale thermal anomaly
located within the upper mantle (Missenard
et al., 2006).
RHEOLOGY
The basement is a complex assembly
of igneous, metamorphic, volcanic and
sedimentary rocks. Despite these differences
in composition and despite the existence of
internal layering and strong anisotropy, in
terms of rheology, basement blocks behave
as overall rigid units, and display very little
internal deformation. This is quite obvious in
satellite images such as the Landsat scene
shown in Figure 2. Normal faults of the Late
Proterozoic - Middle Cambrian rifting event
and possibly older structures, predefine the
structural grain of the variscan inversion.
The Paleozoic cover series are quite weak
and essentially behave as one soft layer
draped over the basement blocks. Details
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 bedinternal layer parallel shortening (LPS).
Figure 4. Effects of layer parallel shortening in

competent beds. (A) Fishtail structure. (B) Buckling.
(C) Disharmonic oblique fold(s). (D) Multitude of small
thrusts. (E) Reverse fault.
Figure 3. Synthetic stratigraphic log for the Eastern

Anti-Atlas. FM: Formations, OZ: Ouarkziz, RC:
Richs, 2B: Second Bani, 1B: First Bani, GS: green
Sandstones, PS: Pink Sandstones, BC: Basal
Conglomerates, P III: Precambrian III, PB: PanAfrican basement. DM: dominant members, in black:
dominant members of structural lithic unit (Currie et
al., 1962), in grey: dominant members of conforming
structural lithic unit. SLU: structural-lithic units, in fine
arrows: conforming structural lithic units with the more
marked boundary zone in Lower Ordovician, in large
arrow: the dominant structural lithic unit which is the
whole represented log.
While the rheology log of the cover can

be subdivided in five conforming units, the
primary structural-lithic unit is in fact the entire
column of Paleozoic sediments including
basement without any clear boundary zone,
let alone any tectonic dcollements. This
can also be observed in the Landsat view
(Figure 2) and cross sections. Despite a
very similar overall stratigraphy, there is
a marked difference between the eastern
and the western Anti-Atlas. In the western
Anti-Atlas, most of the weak units are used
as detachments (boundary zones) and four
distinct dominant members, namely the
basement-P III, the Cambrian limestones,
the Ordovician Bani, and the Devonian
Rich developed their own fold geometries
in what is best described as a multilayer of
polyharmonic buckle folds (Helg et al., 2004).
5
In contrast, the overall geometry observed in

the Eastern Anti-Atlas is single layer drape
folding, dominated by the vertical component
of movement of the rigid forcing member.
STRUCTURAL STYLE
Thick skinned fold belt
The lack of any basal dcollement is probably
the most intriguing feature of the entire
Anti-Atlas fold belt. The basement-cover
contact (Figure 5) is a major unconformity
which is easily identified in satellite images,
aerial photography and in the landscape.
Compressional structures are increasingly
abundant higher up in the stratigraphy, but
even in the western Anti-Atlas, no major
thrust nor any duplexing has been identified.
The geometry of compressional structures

observed within the Paleozoic cover series
is best explained by the trishear models
of Erslev (1991) or Johnson and Johnson
(2002), albeit with the difficulty to clearly
identify the underlying reverse faults. This
leads to conceive that most of the inverted
basement faults must have induced footwall
shortcuts (Bump, 2003), most likely within
in Late Proterozoic syn-rift sediments,
in order to horizontalize before cutting
the basement-cover contact. In addition,
an important component of horizontal
shortening deformation is not localised but
accommodated in a more diffuse manner
within the volume of abundant thick soft
layers of the conforming structural lithic units
(Figure 4).
Figure 5. Basement cover contact at the southern limit of Saghro inlier (cross section 08), view to North-East
(houses for scale).
Compressional structures
On the outcrop scale, the dominant members
show isolated meso-scale structures such as
folds and minor wedging fishtail structures
(Figure 6). The controlling factors are the
bed thickness, the abundance of weak layers
and the relative thickness of the two. The
6
pictures A and B (Figure 6), taken from the

same layers of Devonian carbonate illustrate
two characteristic layer parallel shortening
features, namely buckle folding and wedging
(fishtail). In the fold case A, the thickness
involves more layers then in case B where
the thrust is limited to one single competent
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).
Figure 7. Small scale thrust planes

with striation (in dark) within Early
Ordovician shales, illustrating the
characteristic style of internal layer
parallel shortening deformation;
location N 3124.361 / W 005
30.192 between cross sections 08
and 09 (compass for scale).
bed. The bigger thrust on picture C cuts off

the whole lower part of the 1st Bani dominant
member. Made of sandstones and quartzites
this series has very few weak interlayers, and
reacts as a single layer. The soft layers permit
the development of a drag fold at the base of
the cliff whereas the thick stiffer beds at the
top dont. All these layer parallel shortening
deformations are too small to appear in the
cross sections, their role in balancing will be

considered however.
Bed internal deformations within the thick
weak zones are more subtle in appearance
(Figure 7). When looking closer at shales in
good outcrop conditions, they are invariably
affected by a multitude of small thrusts,
generally at a sharp angle with respect to
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.
Overall,
compressional
layer
parallel
shortening structures are best developed in
the north and their abundance and intensity
decreases progressively southward. This is
interpreted as a regional gradient inbetween
the supposed internal zones of the Anti-Atlas
orogen to the north and the stable continental
interior to the south. Internal parts of the AntiAtlas orogen are obscured by, and may have
been substantially shifted in position during,
Mesozoic and Cenozoic events of High Atlas
formation, however. The spatial relationships
between the Variscan Anti-Atlas belt and
the basement blocks found within and to
the north of the High Atlas are a matter of
ongoing debate (Hoepffner et al., 2005).
strike such as the graben in the middle of

cross sections 07 and 08. Devonian folds
of cross section 07 are interpreted as forced
drape folds resulting from extension in the
underlying graben, but attenuated within the
soft Silurian shales accommodating the fault
throw in a trishear style.
Extensional structures
For calculations of the depth to detachment,

the cross section 08 has been chosen
because it crops out at the deepest level.
Therefore, interpretation errors are minimal
for the geometry of the basement-cover
contact. For the reconstruction at different
stages, we first restore the latest normal
faults in order to establish the situation at the
end of the variscan paroxysm. In a second
step, the variscan folding is restored to a flat
lying configuration. At this stage the layers still
contain a component of early Variscan layerparallel shortening which is compensated by
stretching back to a restored configuration at
the end of Paleozoic sedimentation.
The eastern Anti-Atlas of the Tafilalt and

Mader basins is affected by an important
late event of extension. This is in stark
contrast to the western Anti-Atlas where
no such extensional structures have been
identified. Early authors (Choubert et al.,
1952) depicted the general structure of the
Tafilalt area as a set of tilted blocks inbetween
north dipping normal faults. Normal faults
are quite steep and cut mostly straight
through the entire Paleozoic cover (pictures
in Figure 8). Normal faults are responsible
for a partial collapse of the basement inlier
antiforms. Locally, extensional fault-related
folds (Withjack et al., 1990; Schlische, 1995;
Finch et al., 2004) and roll over folds (Xiao
and Suppe, 1992) developed within weak
Paleozoic cover rocks of the hanging-wall.
Late normal faults are systematically rooted
within the old Late Proterozoic normal
faults, they were reactivated across the
compressional footwall shortcut thrust of the
variscan event.
CROSS SECTIONS
Cross sections have been constructed at
the 1:50000 scale, not allowing to take
mesoscale deformation structures into
account (Figure 9). Interpretations below and
above topography have been made by simple
projection (no change in fault orientation and
constant layer thickness). Some features
have also been laterally projected along
The fault-bend fold in the northern part of

cross sections 06 and 07 is an interpretation.
It has been drawn such as to balance
the extremely deformed Carboniferous
sediments. However, the thrust in cross
section 08 has been observed, but its
interpretation remains ill constrained at
depth. The High-Atlas structures are redrawn
according to Frizon de Lamotte et al.(2000).
DEPTH TO DETACHMENT
Depth to detachment is estimated using
the displaced-area method (Epard and
Groshong, 1993; Groshong, 1994, 1999).
This 2-D cross-section balancing method
(Figures 10 and 11) is based on the uplifted
area above regional caused by compression
(excess area Sc) or a downdropped area
below regional caused by an extension (lost
area Sx). Excess or lost area must be equal
to the displaced area, which is defined by
the depth to detachment times horizontal
displacement. Excess or lost area is
measured above or below a regional horizon
(Hr in Figures 10 and 11). Such simple area
balancing methods do neither require any
detailed information about the shape of the
detachment nor about the kinematical model.
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.
Figure 11. Compressional deformation. Geometric

elements used in the depth to detachment calculation,
including layer parallel shortening Dce.
to be zero, but in the following we will develop

a modified displacement area method in
order to evaluate the role of layer parallel
shortening. In this study, the excess and lost
areas are measured from restored sections
constructed for different stages of the AntiAtlas evolution through time (Figure 12). In
the following discussion, we will concentrate
on cross-section number 08, one of the best
constrained and representative sections
across the eastern Anti-Atlas chain. The
present day cross-section has last been
affected by Neogene tilting and differential
uplift, without any associated faulting. The
excess area resulting from this uplift (Su in
Figure 12.A) has been removed from section
08 according to the lithospheric model of
Missenard et al. (2006), which is based on
gravity data.
The youngest faulting event in Triassic times
caused the formation of the so-called Zagora
graben as well as an extensional reactivation
of crustal scale faults throughout the eastern
Anti-Atlas anticlinorium.
The calculation of the depth to detachment
is first applied to these latest extension
12
features (stages C to B in Figure 12). The

differential uplift (Figure 12.A) does not have
any influence on the lost area in this stage
because this uplift is identical in both profiles.
However, for the next step backward, the
differential uplift excess area (Su) must be
subtracted from the excess area produced
by compressional deformation (stage C)
because the pre-compression profile D does
not include the recent uplift.
Depth to detachment on extensional
structures
The depth to detachment calculation for
extension structures is made under the
classical assumption of constant bed length.
The area Sx below the regional Hr (Figure
10) resulting from extensional deformation
is compensated by an equal area defined
by the horizontal displacement Dx times hx
the depth to detachment below the regional
horizon Hr:
(1)
thus
Where the horizontal displacement Dx is the
Figure 12. Restoration of cross section 08. (A) Differential uplift in Neogene to recent times; cross sectional
area according to Missenard et al. (2006; figure 6.C, eastern profile). (B) Simplified present day cross section;
Sx = area lost by the latest extension in Early Mesozoic times. (C) Cross section at the end of the Variscan
orogeny. (D) Restored section at the onset of Variscan inversion tectonics including early Variscan layer-parallel
shortening. (E) Restored section (Late Devonian, Early Carboniferous).
difference between the original bed length Lx

and the width of the structure Wx:
(2)
Accordingly, the depth of detachment Hx with
respect to the zero altitude is given by:
(3)
Note that in the case of extension, Sx and Dx
are both negative.
Such a depth to detachment calculation
has been made for three different features
: the Zagora graben, the Saghro inlier and
the entire cross-section (Table 1; Figures
2 and 9 for localisation). For the Zagora
graben detachment depth lies in the range
12 to 18 km (-7.5 to 11 mi). The higher
values are those obtained from Silurian and
Late Ordovician which are best constrained.
The results of this depth to detachment
calculation are shown in Graphic 1 by
plotting the altitude of each regional level
(Hr) against the lost area (Sx). Extrapolating
the dotted line back to Sx = 0 km2 yields a
best fit of -16 km (10 mi) for detachment
depth. The slope of this line is the inverse
of the displacement (formula (1)), and yields
a modest horizontal extension of D = 0.204
km (670 ft) in agreement with the measured
values of Dx for the different levels.
Table 1. Depth to detachment on extensional features

of cross-section 08.
Graphic 1. Area-Depth relationship of Zagora

graben.
13
For the Saghro inlier the calculated depth

to detachment varies from 17 to 19 km.
In contrast to the Zagora graben, the lower
horizons (Cambrian and basement) are best
constrained by outcrop observations. For
the entire cross section each layer has been
reconstructed individually. Simple depth
to detachment calculations for this crosssection yield unreasonably high values in
the range -29 to -57 km (-18 to -35 mi). From
an area-depth relationship, the absolute Sx
area for the top of the Silurian level should
be greater then Sx of basement. These
discrepancies may result from either of the
following reasons. First, the overall cross
section is not oriented perpendicular to all the
normal faults; some of them are cut at a low
angle, leading to a minimal apparent offsets
and minimal Dx values. With higher absolute
displacements, the detachment depth
would be shifted upward. Displacement Dx
is dependent on the assumed geometry of
the normal faults at depth. In the interpreted
cross sections, normal faults have been
projected to depth with a constant angle of
70 to 80 as observed in outcrop. Assuming
a listric geometry at depth would lead to
greater displacements and smaller lost area
with respect to the straight fault interpretation.
Secondly, internal deformations (bed length
change) have not been considered. In reality,
the length Lx (Figure 10) is greater then
the initial length before any extension and
stretching. If equation (7) described in the
later section on compressional structures
were applied, a very small amount of strain
would be sufficient to yield a significant
change in the resulting depth to detachment.
With a strain of only -0.03 (3 % bed internal
stretching ), the detachment depth calculated
for the top of Early Cambrian would change
from an initial -54 km (34 mi) upward to only
-19 km (-12 mi). The bed thinning influence
(of only three percent) on the lost area Sx is
negligible.
Depth to detachment and Layer parallel
shortening on compressional structures
The following calculations are made on a
restored profile of the assumed geometry at
the end of the variscan shortening. Practically,
normal faults and their effects on the top of the
14
sedimentary cover have been restored to a

smooth unfaulted configuration. The regional
horizon Hr (Figure 11) has been defined at
the southern end of the cross section. The
late (Neogene) differential uplift area Su has
been subtracted from the measured excess
area. The values of Su is 12.9 km2 (8 mi2)
for the local Saghro section and 61.2 km2 (38
mi2) for the entire cross-section.
The calculation of the depth to detachment on
compressional structures with constant bed
length based on the very same equations as
those demonstrated for extension.
(1)
thus
Displacement Dc is defined as the difference

between the originlal length Lc minus present
day widths of the structure Wc.
(2)
In the case of compression, Dc and Sc
are both positive. Depth to detachment
with respect to sea level, is defined by the
following equation :
(3)
When applied to our cross-sections (Table
2) such considerations yield exceedingly
high values for depth to detachment in the
range -111 to -126 km (-69 to -78 mi) for the
Saghro inlier and -172 to -212 km (-107 to 132 mi) for the entire cross section. Given the
excellent of outcrop conditions, the complete
absence of thrusting, duplexing and other
large scale compressional structures is
a well established fact. Balancing these
cross-sections clearly requires more subtle,
hidden components of strain to be taken
into account. The structural style dictated by
the particular rheology of the cover does not
permit to change the displacement values
other than by the assumption of an important
component of layer-parallel shortening. This
layer-parallel shortening is well documented
in the field, but the small meso-scale of
these deformation features does not allow
them to be shown in cross-sections.
In order to take such bed length changes
into account in the depth to detachment
Table 2. Depth to detachment

on compressional structures
calculation, we include the stretch e :

(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)
The combination of equations (4) and (5) yields :
(6)
or
Including a component of layer-parallel shortening, equation (1) for depth to detachment is
modified to :
(7)
Further, Dc is replaced by equation (2), Dce by equation (4) and Lo by equation (6). This
yields :
(7)
and depth to detachment :
(8)
We examined the influence of layer-parallel
shortening on the depth to detachment by
varying e from 0 to 0.25, in a range compatible
with observations made in outcrop and on the
thin section scale. Resulting values of depth
to detachment Hce are given in Table 3 and

illustrated in Graphic 2.
This exercise shows that depth to detachment
calculations applied to weakly folded section
such as the Anti-Atlas cross-sections
15
are strongly dependent on layer-parallel

shortening. With an assumed moderate
amount of layer-parallel shortening on the
order of 0.11 to 0.17, the calculated depth
to detachment is shifted upward from more
than -100 km to a reasonable depth of -18
to -20 km (-11 to -12 mi). This depth range

is the same as the one initially determined
for late extensional structures. Taking into
account the observed deformation gradient,
the layer-parallel shortening must be higher
in the north and decreases southward.
Table 3. Detachment depth (Hce)

vs stretch (e) for compressional
deformation on the entire crosssection.
Graphic 2. Depth to detachment vs stretch

for compressional deformation on the entire
cross section.
16
CONCLUSIONS
The structural style of this inverted
intracratonic basin is controlled by two
properties:
1. The weak rheology of the cover. This
allowed for a distributed shortening
deformation within shales and siltstones
and the formation of isolated mesoscale
structures within competent layers. The
weakness of the cover series prevented the
development of any strong localization major
dcollements. This soft pile forms a single
structural-lithic unit which adopts a drape fold
allure with respect to the basement structures.
2. Pre-existing inherited structures within the
basement. Precambrian rift normal faults
constitute weak zones (Figure 13, stage
A) which are repeatedly reactivated in later
events be it compressional or extensional.
The variscan shortening lead to the inversion
of Precambrian normal faults. Thrusts faults
systematically shortcut the footwalls to flatten
out within Lower Cambrian shales and never
step up any higher into the Paleozoic cover
(Figure 13, stage B). Shortening in the cover
series is accommodated by an overall drape
folding and widely distributed layer-parallel

shortening of up to 17 %. Extensional reactivation in Early Mesozoic times (stage C)
lead to the formation of new set normal faults
within cover rocks, crosscutting the variscan
compressional cover structures at a high
angle.
The depth to detachment calculation and its
error analysis provide significant informations.
On the extension structures, calculations
yield a detachment depth in the range of -18
to -20 km (-11 to -12 mi). Positive deviation
from this value indicates that the dip of the
normal faults must decrease with depth.
Alternatively, a minimal amount of 3% bed
internal stretching could equally well explain
the discrepancies in depth to detachment
obtained from the overall cross section.
Balancing Variscan shortening deformations
yields unreasonably high values for depth to
detachment in excess of -100 km . Taking a
component of 11 to 17 % distributed layerparallel shortening into account allows to
reconcile surface observations and balancing
with a reasonable depth to detachment of -18
to -20 km.
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 AntiAtlas. 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
Benssaou, M., N. Hamoumi, 2003, The lowerCambrian western Anti-Atlasic graben: tectonic control
of palaeogeography and sequential organisation:
Comptes Rendus Geoscience, 335, p. 297-305.
Boudda, A., G. Choubert, 1972, Sur la limite infrieure
du Cambrien au Maroc: C. R. de lAcadmie des
Sciences, Paris, Serie II 275 (1), p. 5-8.
Bump, A.P., 2003, Reactivation, trishear modeling,
and folded basement in Laramide uplifts: implications
for the origins of intracontinental faults: Geological
Society of America Today, p. 4-10.
Burkhard, M., S. Caritg, U. Helg, C. Robert-Charrue,
A. Soulamani, 2006, Tectonics of the Anti-Atlas of
Morocco: Comptes Rendus Geoscience, 338, p. 1124.
Choubert G., L. Clariond, J. Hindermeyer, 1952,
Livret-guide de lexcursion C36, Anti-Atlas central
et oriental: Congrs gologique international, XIXe
session-Alger-1952, Srie : Maroc N 11, 89 p.
Coward, M.P., A.C. Ries, 2003, Tectonic development
of North African basins : Petroleum geology of Africa;
new themes and developing technologies. T. Arthur,
S. MacGregor Duncan and N.R. Cameron, Geological
Society of London. London, United Kingdom, 207, p.
61-83.
Currie, J. B., H. W. Patnode, R. P. Trump, 1962,
Development of folds in sedimentary strata: Geological
Society of America Bulletin, v. 73, p. 655-673.
Ennih, N., J.-P. Ligeois, 2001, The Moroccan AntiAtlas: the West African craton passive margin with
limited Pan-African activity. Implications for the
northern limit of the craton: Precambrian Research,
112, p. 289-302.
Epard, J.-L., R. H. Groshong, 1993, Excess area and
depth to detachment: The AAPG bulletin, v. 77, no. 8,
p. 1291-1302.
Erslev, E. A., 1991, Trishear fault-propagation folding:
Geology, v. 19, p. 617-620.
Finch, E., S. Hardy, R. Gawthorpe, 2004, Discreteelement modelling of extensional fault-propagation
folding above rigid basement fault blocks: Basin
research, 16, p. 489-506.
Frizon de Lamotte, D., B. Saint Bezar, R. Bracne,
E. Mercier, 2000, The two main steps of the Atlas
building and geodynamics of western Mediterranean:
Tectonics, v. 19, no 4, p. 740-761.
18
Groshong, R. H., 1994, Area balance, depth to

detachment, and strain in extension: Tectonics, v. 13,
no. 6, pp. 1488-1497.
Groshong, R. H., 1999, 3-D structural geology:
Springer-Verlag, 324 p.
Helg, U., M. Burkhard, S. Caritg, C. Robert-Charrue,
2004, Folding and inversion tectonics in the Anti-Atlas
of Morocco: Tectonics, v. 23, TC4006, doi:10.1029/
2003TC001576.
Hoepffner, C., A. Soulaimani, A. Piqu, 2005, The
Moroccan Hercynides: Journal of African Earth
Sciences, 43, pp. 144-165.
Johnson, K. M., A. M. Johnson, 2002, Mechanical
models of trishear-like folds: Journal of Structural
Geology, 24, p. 277-287.
Ligeois, J.-P., A. Fekkak, O. Bruguier, E. Errami, N.
Ennih, 2005, Lower ediacaran age (630-610 Ma) for
the Sarhro group. Implications for the metacratonic
evolution of the Anti-Atlas, correlations with the Tuareg
shield and the evolution of the pan-african orogeny in
the NW Africa (abs.): 4me Colloque International
3Ma, Agadir 2005, Abstract Volume, p. 2.
Manspeizer, W., J.H. Puffer, H.L. Cousminer, 1978,
Separation of Morocco and eastern North America:
A Triassic-Liassic stratigraphic record: Geological
Society of America Bulletin, v. 89, pp. 901-920.
Missenard, Y., H. Zeyen, D. Frizon de Lamotte, P.
Leturmy, C. Petit, M. Srbier, O. Saddiqi, 2006,
Crustal versus asthenospheric origin of relief of the
Atlas Mountains of Morocco: Journal of Geophysical
Research, v. 111, B03410, 13 p.
NASA and Earth Satellite Corporation, < https://
zulu.ssc.nasa.gov/mrsid/ >
Accessed June 30, 2006.
Sebai A., G. Feraud, H. Bertrand, J. Hanes, 1991, 40Ar/
39
Ar dating and geochemistry of tholeiitic magmatism
related to the early opening of the central Atlantic rift:
Earth and Planetary Science Letters, 104, p. 455472.
Schlische, R. W., 1995, Geometry and Origin of FaultRelated Folds in Extensional Settings: The AAPG
Bulletin, v. 79, no. 11, p. 1661-1678.
Withjack, M. O., J. Olson, E. Peterson, 1990,
Experimental Models of Extensional Forced Folds:
The AAPG Bulletin, v. 74, no. 7, p. 1038-1054.
Xiao, H., J. Suppe, 1992, Origin of Rollover, The
AAPG Bulletin, v. 76, no. 4, p. 509-529.
C. R. Geoscience 338 (2006) 1124

http://france.elsevier.com/direct/CRAS2A/
Tectonics
Tectonics of the Anti-Atlas of Morocco

Martin Burkhard a, , Sverine Caritg a , Urs Helg b , Charles Robert-Charrue a ,
Abderrahmane Soulaimani c
a Institut de gologie, universit de Neuchtel, rue mile-Argand 11, CP 2, 2007 Neuchtel, Switzerland
b BUWAL, Bundesamt fr Umwelt, Wald und Landschaf, 3003 Bern, Suisse
c Facult des sciences Semlalia, universit Cadi-Ayyad, av. Moulay-Abdellah, BP S20, Marrakech, Maroc
Received 4 October 2005; accepted after revision 17 November 2005

Available online 4 January 2006
Written on invitation of the Editorial Board
Abstract
The Anti-Atlas is reviewed and examined in the light of its geodynamic significance as a Palaeozoic basin and fold belt. Shortening 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 NWSE to northsouth and maybe NESW 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 Acadmie des sciences. Published by Elsevier SAS. All rights reserved.
Rsum
Tectonique de lAnti-Atlas marocain. LAnti-Atlas est revu et examin sous langle de sa signification godynamique comme
bassin palozoque et comme chane plisse palozoque. Le raccourcissement est accommod par le plissement polyharmonique
de la couverture, avec une nette implication du socle. Aucun systme significatif de chevauchement ni duplex ne sest dvelopp.
LAnti-Atlas est un bassin intracratonique fortement invers plutt quune partie de la marge passive de la Palotthys. Linversion
doit dater du Carbonifre tardif/Permien prcoce. La direction du raccourcissement a chang au cours du temps depuis une direction
NWSE vers une direction nordsud et peut-tre mme NESW, ce qui conduit la formation de figures dinterfrences de plis en
dmes et bassins aux chelles allant de 100 m 10 km. Pour citer cet article : M. Burkhard et al., C. R. Geoscience 338 (2006).
2005 Acadmie des sciences. Published by Elsevier SAS. All rights reserved.
Keywords: Folds and folding; Continental contractional orogenic belts; Africa; Morocco; VariscanHercynian
Mots-cls : Plis et plissement ; Chanes orogniques continentales de contraction ; Afrique ; Maroc ; VarisqueHercynien
1. The Anti-Atlas of Morocco Introduction
* Corresponding author.
E-mail address: [email protected] (M. Burkhard).
The Anti-Atlas fold belt of the south-western Moroccan desert (Figs. 1 and 2) offers vast expanses of
beautifully exposed bare outcrops. This is due to a recent phase of uplift and concomitant erosion that led
1631-0713/$ see front matter 2005 Acadmie des sciences. Published by Elsevier SAS. All rights reserved.
doi:10.1016/j.crte.2005.11.012
12
M. Burkhard et al. / C. R. Geoscience 338 (2006) 1124
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
ApplachianMauritanidesMoroccan Meseta are coloured in green and pink tones [44,73].
Fig. 1. Situation de lAnti-Atlas par rapport la chane des Appalaches la fin du Palozoque [80]. Les contours isopaques en kilomtres sont donns pour les bassins sdimentaires palozoques qui nont subi que peu ou pas de dformation [102]. Le mme code couleur est superpos la chane
plisse de lAnti-Atlas afin dindiquer la profondeur estime de ce bassin avant linversion. Les massifs de socle allghaniens sont indiqus en bleu
(inspir du Blue Ridge appalachien). Les parties internes, mtamorphiques et plus anciennes, de la chane des AppalachesMauritanidesMsta
marocaine sont colores en vert et en rose [44,73].
to a rejuvenation of relief with summits of 2500 m

and higher [35,47]. Geomorphologists have coined the
French term relief appalachien to characterize the pattern of deeply eroded fold trains, and there is indeed a
striking similarity between the geomorphic expression
of the Appalachian Valley and Ridge and the Anti-Atlas.
Both chains are external parts of the larger Variscan
AppalachianOuachitaMauritanides orogen (Fig. 1).
The relationship between the Anti-Atlas fold belt and
the internal parts of this orogen remains to be elucidated
in terms of tectonic style, timing and geodynamics.
At first sight, the Anti-Atlas shares many common
features with its American counterpart of the Valley and
Ridge in general and with the Alleghany Basin in particular. Both are located on the craton side of the orogen involving a thick and fairly regular layer cake of
mostly shallow marine Palaeozoic sediments. Both fold
provinces also have their non-folded time-equivalent intracratonic basins further inland: Michigan and Illinois
on the American side, Taoudenni, Tindouf and others on
the African side. On closer inspection, however, and in
stark contrast to the frontal Appalachian chain, the AntiAtlas fold belt does not easily conform with the standard anatomy of foreland fold-and-thrust belts worldwide [77]. The most striking difference is the existence
of major basement domes at a very short distance behind
the deformation front [76]. Similarities exist with Wind
River-style basement uplifts of the frontal Rocky Mountains, but in the Anti-Atlas, the basement uplifts occur
amidst a tightly folded thick Palaeozoic cover series.
The style of cover folding is quite unique too, with a
dominance of upright detachment folds and a complete
absence of any thrusting and duplex structures [47],
with the exception of the westernmost parts of the Anti-
13
Fig. 2. Geologic overview map of the Anti-Atlas, compiled from the geologic map series (1:200 000) of the Service gologique du Maroc
[105114]. NWSE 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 gologique de lAnti-Atlas, synthtise partir de la srie des cartes 1:200 000 du Service gologique du Maroc [105114]. Les
transects NWSE, utiliss pour la compilation de la Fig. 4, sont indiqus le long de la bordure sud. Une coupe travers lanticlinorium de lAdrar
Zouggar est donne sur la Fig. 3.
Atlas along the Atlantic coast [8,65]. There is no thin

skinned basal dcollement level and the western AntiAtlas does not conform with a foreland fold and thrust
belt system in the sense of Boyer and Elliott [15].
An exhaustive review of the geology of the AntiAtlas has been presented by Michard [64]. While many
authors have interpreted the Anti-Atlas fold belt in
terms of a predominance of strike-slip movements [43,
61,74,99], recent structural analyses depict the western Anti-Atlas folds as highly cylindrical frontal folds
related with a head-on collision in a NNWSSE direction [20,47,91].
In this paper the Anti-Atlas system is revisited in
light of recent data regarding the Pre-Cambrian basement, the Palaeozoic cover, structural observations
within the Anti-Atlas belt as well as plate tectonic reconstructions on a global scale. Questions of particular interest concern the Palaeozoic sedimentary basin
history and subsidence mechanisms as well as the evolution of this Anti-Atlas basin near the border of the
West-African Craton through time.
1.1. The basement

The Anti-Atlas basement is a complex assemblage of
crystalline, metamorphic and sedimentary rocks. Note
that we use the term basement here in the sense of
the petroleum geologists [58], including all rocks older
than the Gondwana Megacycle. The oldest rocks of
the West-African Craton (WAC) are granitoids, gneisses
and a complex series of metamorphic rocks, attributed
to the Eburnean orogeny at around 2000 Ma [2,96,
98]. The final assembly of most of the African continental crust takes place during the Panafrican orogeny,
lasting roughly from 700 to 600 Ma [46]. Remnants
of a Panafrican suture zone are present as a dismembered ophiolite series in the Bou Azzer inlier of the
central Anti-Atlas [45,55,81,96] a structure recently
reinterpreted as an aulacogen within the WAC [32].
Elsewhere in the Anti-Atlas, the Panafrican event left
a more subtle imprint in the form of strike-slip shear
zones and thrusts [43]. Post-Panafrican continental extension is well documented for the entire Anti-Atlas re-
14
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 dcollement at about 25 km depth.
Fig. 3. Coupe travers lanticlinorium de lAdrar Zouggar [27] ; pour la localisation, voir Fig. 2. La gologie de surface est construite partir
de la carte gologique 1:200 000 [110], augmente par nos propres cartographies et mesures de plis et dautres structures [20]. comparer
avec Michard [64, (Fig. 20)]. Cette coupe est approximativement quilibre en ce qui concerne les aires. Une dtermination du raccourcissement
horizontal minimal lchelle de la carte est de lordre de 13 15% par le seul plissement msoscopique. Quelque 10% de raccourcissement
additionnel par dformation intra-couche sont fort probables. Un raccourcissement total de 25 km ncessite une profondeur de dcollement
25 km de profondeur dans le socle.
gion [72,89] where indications for synsedimentary tectonics are found in the clastic series of the Saghro-group
(PII3 ) [92], the Ouarzazate group (PIII) and, progressively fading upward within lowermost Cambrian series. The interpretation of events during the Late Neoproterozoic (600540 Ma) is still a matter of debate,
due in part, at least, to the scarcity of reliable absolute
ages. The significance of the basement/cover relationships and the geodynamic context are also still open
to discussion: interpretations range from syn-orogenic,
Late Panafrican molasse series shed in a collision context, to post-orogenic extension and collapse with the
formation of tilted blocks and halfgrabens [55,87,91].
In a most recent interpretation [90], based on the type
and volumes of volcanic rocks associated with the PIII
Ouarzazate series, the Late Proterozoic extension event
is compared with a basin and range extensional setting,
with an important production of lower crustal melts.
Soulaimani and Piqu [88] even go so far as to propose the present-day Anti-Atlas boutonnires as being
re-activated former metamorphic core complexes.
1.2. The cover
Prior to folding, the Palaeozoic cover series of the
Anti-Atlas must have represented a fairly regular layer
cake configuration. Overall thickness reaches 10 km
and more in the westernmost Anti-Atlas near Tiznit
and decreases to about 6 km and less in the easternmost Anti-Atlas of the Tafilalt. In terms of rheo-
logical basement/cover relationships, the sedimentary

cover series is best defined as starting with the first
carbonate-bearing units of the lowermost Adoudounian,
following in most places concordantly upon the coarsegrained PIII conglomerates of the Ouarzazate series.
This limit makes a major colour contrast in the field
and it is easily visible in satellite images. The PIII
behaves as a competent basement-like unit, while the
well-layered carbonates above it are increasingly detached and folded at different wavelengths from meters up to kilometres. Despite some prevailing uncertainties, the PIII/Adoudounian transition just above coincides with the Latest Proterozoic/Cambrian boundary [18,38]. There is good evidence for some extensional tectonics continuing upward into the Earliest
Cambrian (Lower carbonate series of the Adoudounian)
at least in the western Anti-Atlas region [3,9,10]. Lateral correlations are greatly facilitated by the presence of fossils and characteristic facies assemblages
and the Anti-Atlas cover figures among the best studied Palaeozoic series worldwide [11,17,29,37,50,64,73,
101]. A schematic compilation in the form of a time
chart along a strike parallel line is shown in Fig. 4.
Overall, the Anti-Atlas cover series are predominantly deposited in a shallow marine environment. Important platform carbonate build-ups are observed during the Lower Cambrian of the western Anti-Atlas [13,
38]. From Middle Cambrian to Late Silurian times,
sedimentation is dominated by detrital input from the
African craton, i.e. from the east and/or the south [17,
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. Schma chronostratigraphique pour lAnti-Atlas. Laxe vertical est le temps [40], laxe horizontal est une coupe ouestest le long de la
chane (voir la Fig. 2 pour la localisation des transects). Pour chaque colonne verticale, les donnes stratigraphiques ont t compiles et projetes
depuis le nord-ouest (vieux) et le sud-est (jeune). Le code couleur indique le caractre dominant de la sdimentation (voir lgende). Les lacunes de
non dpt ou drosion sont laisses en blanc. Les appellations informelles des formations ainsi que certains vnements majeurs sont indiqus.
16
28]. Carbonate sedimentation is resumed at the end of

the Silurian [50] and lasts for most of the Devonian, being combined with clastic inputs, throughout the AntiAtlas region and beyond [100]. The Lower Carboniferous [59] is marked by an renewed increase in detrital
input from the east, south and in places from the north
[66,73]. Regional-scale facies and thickness changes
point toward an open ocean to the west and northwest
throughout the Palaeozoic. It is therefore tempting to interpret the Anti-Atlas basin as the landward tier of the
former passive margin of Gondwana, facing west and
northwestward. In the preserved stratigraphic record,
the shelf break, continental slope and rise of this ancient
passive margin wedge are missing, however. In comparison with contemporaneous North-African Palaeozoic basins such as Tindouf, Reggane, Bchar, Ahnet,
Ghadames, Illizi, Hamra, Murzuq [12], the Anti-Atlas
cover series appears as just another intracratonic basin,
which happened to be close enough to the continental
edge to be more massively involved in Late Palaeozoic
collision tectonics than its neighbours.
1.3. Cover series as a mirror of distant tectonic
events?
Palaeozoic sedimentation of the Anti-Atlas region is
fairly continuous throughout, and minor disconformities
are most easily explained in terms of sequence stratigraphy, i.e. sea-level and/or climatic changes. Probably
the most important climatic events are the Late Ordovician glaciations, recorded within the 2nd Bani microconglomerates [39,56,95] and an associated erosional
discordance. The following sea-level rise is hold responsible for the deposition of the thick interval of Silurian black shales, a major source rock of the NorthAfrican realm [12,68] and a potential dcollement horizon. The famous Devonian Kellwasserevent at the
Frasnian/Famennian boundary [16,54] is probably a
meteorite impact that left its imprint in the faunal assemblages.
As illustrated in Fig. 4, there is only one noteworthy
interruption of the Palaeozoic sedimentary cycle in the
Late Cambrian. Classically, it has been correlated with
a sardic phase of orogeny [64] or with an epeirogenetic
uplift [29]. The widespread (although not absolute) nature of this hiatus, common to all North-African basins
from Morocco through Tunisia, Algeria to Libya [25]
speaks for a general, eustatic, rather than a local, tectonic origin. Crossley and McDougall [26, (p. 160)]
quote evidence for block faulting and marked angular
unconformities in Algeria; similar evidence is missing
in the Anti-Atlas of Morocco, where some lateral thick-
ness changes are subtle at best, and strongly overprinted

by later folding. Higher up in the stratigraphic column, subtle lateral facies changes, sedimentary wedges
and minor erosional disconformities have long been
interpreted as echoing distant tectonic phases such as
Taconic, Acadian, Bretonne and Sudte [64], known
from either the Appalachian or the European Variscides.
New palaeogeographic reconstructions (Fig. 6) of the
evolution of Gondwana and the reassembly of Pangea
provide a reference frame against which such statements
can be tested [83,86,93].
The Late Cambrian hiatus could be interpreted as
related to a new phase of rifting further north and concomitant erosion on a southern rift shoulder. At least
two important Palaeozoic rifting phases have indeed
been postulated to occur along the northern border of
Gondwana [93]. Avalonian (and Armorican, etc.) terranes or continental fragments are supposed to originate
from the northern margin of Gondwana in Cambrian
times, leading to the opening of the Rheic Ocean. A second rifting event of Hunic terranes during the Lower
Silurian would have led to the formation of the PalaeoTethys ocean [93, (Fig. 3)]. In this proposal, future
Avalonian and Hunic terranes represent former active
subduction margins off the northwestern African craton. Both terranes would have drifted away after two
successive stages of back-arc rifting giving way to two
(Rheic, Paeothethys) Palaeozoic oceans. This proposal
opens new perspectives in the interpretation of events
known in the various Palaeozoic terrains of Morocco.
The Anti-AtlasTindouf basin would not have been anywhere close to the open ocean prior to Late Silurian
times. We see little evidence for rifting phases from
Middle Cambrian times onward, however, and the easiest interpretation is to assume that none of those rifting
events did take place anywhere close to the Anti-Atlas
basin, but rather in a more external position or in a different place along the northern margin of Gondwana.
The former passive margin of Gondwana would be entirely obscured by the future Atlantic rifting and hidden
below the Mesozoic onshore and offshore basins. Alternatively, the Anti-Atlas basin could be interpreted
as the southeastern half of an oblique back-arc basin
that evolved into a passive margin in SilurianDevonian
times, after a (right lateral?) strike-slip departure of Hunic terranes.
On a local scale, geologists have also long tried
to correlate events between the Meseta, north of the
present-day High Atlas chain, and the Anti-Atlas [70,
73,74]. There is no direct evidence for the existence of
any kind of tectonic deformation within the Anti-Atlas
at least up to Middle Carboniferous times. It is thus
quite obvious that the Moroccan Meseta underwent a

very different tectonic history as compared to that of
the Anti-Atlas, from the Ordovician onward [69,71].
Many parts of the complex assemblage of units in the
Meseta have suffered several intense phases of deformation, metamorphism including the intrusion of granites
in Silurian times and from the Lower Carboniferous onward [49]. In this respect, the Meseta is comparable
to the internal Appalachian chain, with its Taconic and
Acadian belts. The most important question still open
in this context regards the relative position of the Moroccan Meseta block with respect to the stable African
craton. In the light of the most widely accepted plate
tectonic models [86], the Meseta might be regarded as
a small continental fragment (or terrane) rifted off the
northern margin of Gondwana during an ill-defined period in the Lower Palaeozoic, comparable to Avalonia
and the more recently postulated Hunic terranes [93].
Such fragments would have been accreted again to
the African craton only during the latest stages of
continentcontinent collision in Late Carboniferous [14,
33]. Such mobilistic proposals have been made as early
as 1971 by Schenk [82], but the lack of an identified suture between the Anti-Atlas domain and the Meseta and
the similarities in the Lower Palaeozoic stratigraphic
record are quoted as evidence against such a scenario [49,51,52,71,73,75]; a recent confirmation of the
Meseta block being of African affinity is provided by
crustal xenoliths found in Triassic lamprophyres [31].
2. Anti-Atlas: what kind of sedimentary basin?
The Anti-Atlas Basin has been involved in Late
Palaeozoic folding, uplift and erosion and it is therefore
difficult to establish its true initial shape and extent to
the west and to the north, where the Alpine High Atlas
is a further obstacle to restorations. East and southeastward, the transition towards the neighbouring petroleum
bearing Palaeozoic basins of Bechar and Reggane are
somewhat obscured by a thin blanket of Upper Cretaceous sediments (Hamada), but compilations from reflection seismics reveal the gross trends in the form
of isopach contour maps shown in Fig. 1 [6,102]. The
Anti-Atlas actually appears to be the northern half of
the Tindouf Basin. It is not clear if the two are separated from each other by some sort of swell or basement
high, nor is there any known reason for the localization of the remarkably smooth trend of the Ouarkziz
chain (Fig. 1), which represents the deformation front
of the Anti-Atlas belt to the SSE (Figs. 2 and 3)
and the northern border of the Tindouf basin. A strikeslip boundary (of Alpine age) has been proposed to run
17
along Dra lowland north of the Jbel Ouarkziz [99],

but field evidence seemingly rules out any significant
wrenching of any age to occur in this area [20,47,91].
To the east, the Anti-Atlas belt or basin turns gradually into the Ougarta chain [30,48]. Just as in the case of
the Anti-Atlas belt, folding, uplift and erosion prohibit
the establishment of original isopach contours for the
former Ougarta Basin, but it is quite clear that this intratratonic chain is localized along a former trough of NW
SE orientation. The same structural trend is also present
in neighbouring basins of Reggane and Bechar and it is
generally admitted that this direction is inherited from
the Panafrican orogeny [24]. Our compilation in Fig. 1
further illustrates the location of the different Palaeozoic
basins with respect to the AppalachianVariscan chain
and the subsequent rifting axis of the Atlantic [80]. The
question arises of what was the geodynamic setting of
these different North-African basins. Is the Anti-Atlas
Basin a remnant part of a passive margin of Gondwana (Fig. 6)? The total thickness of the sedimentary
pile alone might be used as an argument in favour of a
passive margin setting for the Anti-Atlas with the freeboard to the WNW. Other Palaeozoic basins, however,
at hundreds of kilometres inboard, accumulated similar
total sediment thicknesses in clearly intracratonic settings (Taoudenni > 8 km, Murzuq > 6 km). The same is
true for large Palaeozoic basins on the American craton
such as the Michigan (ca. 5 km) and to a lesser degree
Williston, Illinois, and Hudson Bay [57]. A major argument against a passive margin interpretation is the lack
of any sedimentary record for deposits from the outer
shelf, talus and continental rise. Despite some general
tendency to more open marine conditions westward,
there are no deep water sediments preserved anywhere
in the Anti-Atlas (with the exception of some debatable
flysch deposits of Carboniferous age at the northern border near the High-Atlas front [66]. The determination
of palaeo-water depths within the Carboniferous clastic
series of the Anti-Atlas remains a matter of controversy,
however. Sedimentary structures, such as graded bedding and intense convolute bedding as well as the presence of olistrostromes has been used as an indication for
turbiditic flysch type sediments laid down in deep water in a compressional context [66]; we could not find
any convincing flysch sequences, however, and a shallow water origin in a coastal and deltaic environment
seems an equally appropriate interpretation of the sedimentary structures [41,59].
A neutral reading of the stratigraphic record of the
Anti-Atlas (Figs. 4 and 5) reveals an eventless subsidence history very comparable to contemporaneous
intracratonic basins of North America [7,53]. The to-
18
late. Alternatively, and more probably, the Anti-Atlas

was the locus of a major broad thermal dome with limited syn-rift deposits in widely distributed small graben
and half-graben systems, rather than a well defined narrow rift trough with associated shoulders. Subsequently,
this broad swell (or basin and range-type collapse?)
would not have led to any successful rifting in the AntiAtlas region but to a long phase of thermal subsidence
throughout the Palaeozoic.
Fig. 5. Sediment accumulation curve for southwestern Anti-Atlas (between the Goulimine and Foum el Hassan transects). Horizontal axis
is time [40], vertical axis is sediment thickness as observed today,
compiled from a vast body of stratigraphic literature regarding the
western Anti-Atlas. No decompaction, palaeo-water-depths, eustatic
corrections nor back-stripping has been applied. Principal informal
formation names are given along the top of the chart, while major
events of the Gondwana margin are indicated at the bottom.
Fig. 5. Courbe daccumulation sdimentaire pour lAnti-Atlas
sud-ouest entre les transects de Goulimine et de Foum el Hassan (voir
Fig. 4). Laxe horizontal donne le temps, linaire en millions dannes,
laxe vertical donne lpaisseur des sdiments en kilomtres. Les donnes ont t compiles partir dun grand nombre de publications
concernant la stratigraphie de lAnti-Atlas occidental. Aucune correction de dcompaction, de palo-profondeur deau, deustatisme ou de
dlestage na t applique. Les principaux noms de formations, informels, sont donns en haut du graphique, tandis que les vnements
tectoniques majeurs sur la bordure de Gondwana sont indiqus en bas.
tal accumulation curve shown in Fig. 5 illustrates the

general difficulty in the interpretation of such linear
on and off subsidence trends. In the case of the AntiAtlas, more than 10 km of shallow marine sediments
were accumulated over a time span of almost 200 Ma.
Even though we have not applied any decompaction
nor backstripping procedure, it is quite clear that this
curve lacks tell-tale features of rapid tectonic rifting followed by thermal subsidence as expected in the case of
a riftdrift passive margin evolution [4]. Just as in the
case of the Michigan [53] and of many other intracratonic basins, the underlying mechanism of subsidence
remains enigmatic [57,85]. Explanations put forward include thermal decay, uncompensated masses, imposed
load as well as the far field, flexural response to tectonic
compressional forces and lithospheric buckling [85,103,
104]. The only rifting phase that left its imprint in the
rock record is ill constrained in its age; the terminal Proterozoic PIIILower Cambrian basal conglomerate and
lower limestone series are all deposited in a rifting context as confirmed by a magmatic suite of calco-alkaline
to tholeiitealkaline volcanics [90]. Even this rifting
event does not clearly show up in Fig. 5 either, with less
than 2 km preserved thickness of PIII series, unless we
assume the Late Proterozoic rifting was short and very
3. The Anti-Atlas chain what kind of fold belt?

The western Anti-Atlas chain is dominated by highly
cylindrical, upright fold trains. In contrast to the Appalachians and other foreland fold-and-thrust belts
worldwide [77,78], there is no evidence for any thinskinned thrusting with the exception of the westernmost
parts of the Anti-Atlas along the Atlantic coast [8,65]
and some blind thrusting below the Ouarkziz ridge. In
our interpretation [19,47], the Ouarkziz ridge, a slightly
curved monocline of more than 400 km length, is the
surface expression of a triangle structure [97]. We postulate the existence of a blind thrust to end somewhere
below the Tindouf Basin and a major NNW-vergent
backthrust to re-emerge within a thick series of shales
above the Devonian Rich, tightly folded and pushed
under the Carboniferous Ouarkziz monocline (Fig. 3).
This mountain front of the southwestern Anti-Atlas is
progressively loosing its significance eastward, however [36]. East of Tata, Anti-Atlas fold trains are decreasing in number and individual folds are decreasing
in amplitude, tightness and cylindricity. There is also
a change in general orientation from SWNE to east
west and the Anti-Atlas fold belt seems to merge with
the Ougarta chain of SENW orientation [24,30,48].
Some layer-parallel dcollements and multiple detachments are obviously required between different
structural levels, each folded with its own characteristic
wavelength and amplitude, but no such dcollement has
ever been mapped to step up in a ramp-flat geometry
across competent beds. Our recent structural analyses
[20,47] confirm previous studies [23,64] in the sense
that individual marker beds are just folded, but never
thrust upon each other in a duplex style. We conclude
that the Anti-Atlas folds are a natural case of truly polyharmonic buckle folding. The abundance of incompetent shales allowed the subordinate competent beds such
as the Ordovician Bani quartzites and the Devonian
Richs carbonates to develop their own characteristic
folds. Overall map scale shortening on the order of 10 to
20% was accommodated without (cover-)thrusting [47]
and the Anti-Atlas defies any straightforward interpre-
tation in terms of a classic thin-skinned thrust system

organization [15].
On the crustal scale, however, the cover shortening determined across the western Anti-Atlas fold belt,
poses a serious balancing problem: where and how is
this shortening accommodated within the basement?
Large structural domes, the so-called boutonnires [22]
of Proterozoic basement rocks, crop out at a very short
distance behind the south-eastern front of the southwestern Anti-Atlas and these basement inliers punctuate the tightly folded Palaeozoic cover. The lack of
any mapped thrusts has led earlier authors to consider
even these basement uplifts as just another, deeper level
of crustal scale folds [22,23] dubbed plis de fonds by
Argand (Fig. 5AC) [5]. Balancing and rheology considerations make such a gentle folding interpretation
improbable [67], however, and we infer that the AntiAtlas basement uplifts must be associated with a series
of major crustal scale reverse faults, supposedly in a
Windriver-style well known in the foreland of the Rocky
Mountains [67,76,94]. This uncommon proposal is illustrated in a section across the western tip of the Adrar
Zouggar anticlinorium (Fig. 3) [27,110]. The backbone
of this structure is made of Ordovician quartzites [27]
and drilling in the 1960s has documented the existence
of a basement core at 3.4-km depth, at least 2.5 km
above the regional basement top (5 km) as constructed
between the synclinorium to the north and the Tindouf
basin to the south. Gently plunging fold axis allows the
projection of the Jbel Rich (Devonian carbonates) folds
above the top of this anticlinorium. Southward, deformation ends below the Jbel Ouarkziz triangle structure,
while northward, another major basement uplift of the
Tagragra dAkka boutonnire has led to a complete removal of all cover series. The tightly folded Bani (Ordovician) clearly projects above this basement uplift,
however. The Adrar Zouggar and Tagragra dAkka anticlinoria nicely illustrate the balancing problem associated with such basement folds. A conservative unfolding of the well-exposed Devonian and Ordovician
marker beds reveals a minimum of 12 to 15% shortening ratio this amounts to some 17 km of minimum
line-length shortening (Fig. 3). In order to compensate
this shortening at the basement level, we postulate the
existence of a set of rather steep (40 to 60 ) faults, possibly Late Proterozoic normal faults that would have
been strongly inverted. This structural style cannot be
mapped in the case of the Adrar Zouggar, but it is compatible with structural observations made along the borders of more internal, deeply eroded basement inliers
such as the Tagragra de Tata [21,47]. Simple area balancing considerations allow us to estimate the depths
19
of detachment. With an excess area of ca. 380 km2 of

basement above regional, a minimum shortening estimate of 17 km requires a basal detachment in the lower
crust, at ca. 32 km depth (Fig. 3). Note that presentday Moho depth is thought to be at ca. 35 km [36] at
least. A more likely estimated horizontal shortening of
some 25 km, taking into account layer-parallel shortening features and other bed internal strains [47], would
lead to a lesser depth of detachment at mid-crustal levels of ca. 25 km (Fig. 3).
The Appalachians also used to have their basement
problem in the form of the Blue Ridge (Fig. 1), a linear belt pre-Cambrian basement rocks of Grenvillian,
i.e. American craton affinity [44,63,79] cropping out
at some distance behind the thin-skinned frontal Valley
and Ridge fold-thrust belt. While these basement uplifts
have long been considered as more or less autochthonous too, crustal scale seismic reflection profiling has
provided clear evidence of a truly allochthonous nature of these basement slivers, detached from the former
edge of the passive margin and thrust craton-ward by
up to 200 km! [44, plate 1, section C], a tectonic style
anticipated by Argand [5, (Fig. 5, case E)], [63]. In comparison, our inversion interpretation of the Anti-Atlas
basement domes remains very modest and autochthonous indeed and the question arises if a Blue Ridge
type interpretation would not be more appropriate in the
case of the Anti-Atlas? We have not found any positive
arguments in favour of such an analogy, and we conclude that the similarities between the Appalachians and
Anti-Atlas fold belts are not extending far below the geomorphic expression of a common Appalachian relief.
We further propose the structural style of the Anti-Atlas
belt being a rather unique combination of basement uplifts in a Windriver style that happened to occur below a
more than 10-km-thick series of mostly soft, shale dominated sediments. Tectonic compression of this Palaeozoic basin leads to a massive inversion of the underlying
basement structures and a simultaneous polyharmonic
buckle folding of the basin fill. Despite its proximity to
the AppalachianVariscan chain, the Anti-Atlas is not
easily classified as a foreland fold-thrust belt to this orogen. In many respects, it is rather to be considered as a
severely inverted intracratonic basin related to the orogen on a crustal to lithospheric scale, supposedly with a
floor thrust at mid to lower crustal level. The Anti-Atlas
belt has a direct connection to the OugartaAhnet fold
belts, but other North-African basins suffered a similar inversion event at the end of the Palaeozoic [24,42]
in more isolated, clearly intra-cratonic settings. Some
striking analogies in structural style and complexity also
exist with the Iberian chain, an intraplate Cretaceous
20
inversion structure in the southern foreland of the Pyrenees [84].

Interestingly, there are no intracratonic structures of
Palaeozoic age known cratonward of the Alleghenian
mountain front of the Appalachians. However, the very
same craton reacted quite sensitively to Rocky Mountain deformations and subduction along its western
border, with the formation of basement uplifts and
inversion of many old graben structures hundreds of
kilometres eastward, well within the North American
craton [60].
4. Conclusion: plate tectonic setting through time
(Cambrian to Permian)
The tectonic evolution of the Anti-Atlas basin and
fold belt in relation to plate tectonics on a global scale
is illustrated in Fig. 6, and summarized below:
during the Panafrican orogeny, a series of terranes

are accreted to the West African craton on its northern and probably western side. While the northeastern suture (Bou Azzer) and terranes to the northeast will remain in place, northwestern and western
borders are subsequently reactivated and a series
of terranes or continental fragments will be ripped
off again during the Palaeozoic. The southwestern
Anti-Atlas, however, at the margin of the Saharan metacraton [1] remains attached to Gondwana
throughout its Palaeozoic history;
in Late ProterozoicEarly Cambrian, the Anti-Atlas
area is in extension with the formation of many
widely distributed graben and halfgraben structures, filled in with coarse clastics (PIII) mostly
of igneous origin. The youngest volcanism is tholeiiticalkaline 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 (Appalachians) are according to Fichter [34].
Fig. 6. volution schmatique de lAnti-Atlas (toile) en comparaison avec la chane des Appalaches travers le Palozoque, une chelle globale.
Les cartes palo-tectoniques ont t redessines et simplifies partir de Stampfli et Borel [93]. Les coupes schmatiques gauche (Appalaches)
sont selon Fichter [34].
setting. The geodynamic significance of this extensional event is not entirely clear, however. It could
be related to a southeast dipping, major and longlived subduction zone on the northwestern margin
of Gondwana, causing a wide area of extension cratonward in a basin and range style. Alternatively,
extension could be due to a (series of) hot spot(s)
within Gondwana;
from Middle Cambrian through Middle Carboniferous, the western Anti-Atlas basin is characterized by a strong and essentially linear subsidence
trend, leading to the accumulation of more than
10 km of mostly fine-grained clastic sediments,
shed into an epicontinental sea from the African
craton. There is little evidence in this stratigraphic
record for tectonic events postulated to have taken
place along the active northwestern plate margin of Gondwana [93]. The departure of Avalon,
Armorica and Hunic terranes from this margin
in successive events of back-arc spreading must
have brought the Anti-Atlas Sea increasingly closer
to the open ocean(s) (Rheic and Palaeo-Tethys).
From Silurian times onward, the Anti-Atlas Basin
could thus represent the passive margin of the
Palaeo-Tethys ocean, but very little if any sediments of the more distal parts of this passive margin are preserved anywhere (with the possible exception of terrains west of Guelmin [8] and near
Tineghir [66]);
in Late CarboniferousPermian (?) compression
leads to an event of strong inversion and folding. Basement is uplifted and folded into huge antiformal culminations (boutonnires) which punctuate the southwestern Anti-Atlas fold belt. The
structural relief of the basement culminations is
in excess of 10 km; minimum estimates of total
shortening are 15 to 25 km. The Anti-Atlas belt
does not represent a classical frontal, thin-skinned
foreland fold-and-thrust-belt of the Appalachian
Variscan orogen, however, but rather an intracratonic, thick-skinned basement inversion belt. Similar time-equivalent belts occur further east into
the African craton (Ougarta, Ahnet), but no such
structures are known on the American side of the
Appalachian chain.
In conclusion, the Anti-Atlas was a fence-rider
throughout, watching the action from a distance, not
paying any tribute, nor suffering too much from the
events going on all along the very active margins of
Gondwana [49,62].
21
References
[1] M.G. Abdelsalam, J.-P. Ligeois, R.J. Stern, The Saharan
Metacraton, J. Afr. Earth Sci. 34 (34) (2002) 119136.
[2] H. Ait Malek, D. Gasquet, J.-M. Bertrand, J. Leterrier,
Gochronologie UPb sur zircon de granitodes burnens et
panafricains dans les boutonnires dIgherm, du Kerdous et
du Bas Dra (Anti-Atlas occidental, Maroc), C. R. Acad. Sci.
Paris, Ser. IIa 327 (1998) 819826.
[3] A. Algouti, A. Algouti, B. Chbani, M. Zaim, Sdimentation et volcanisme synsdimentaire de la srie de base de
lAdoudounien infra-cambrien travers deux exemples de
lAnti-Atlas du Maroc, J. Afr. Earth Sci. 32 (4) (2001) 541
556.
[4] P.A. Allen, J.R. Allen, Basin Analysis, Blackwell Scientific
Publications, Oxford, 1990.
[5] E. Argand, La tectonique de lAsie, in : 13th International Geological Congress, Brussels, 1924.
[6] T. Arthur, D.S. MacGregor, N.R. Cameron, Petroleum geology
of Africa, new themes and developing technologies, Geological
Society, London, Spec. Publ. 207 (2003).
[7] A.W. Bally, Phanerozoic basins of North America, in: A.-W.
Bally, A.-R. Palmer (Eds.), The Geology of North America: An
Overview, Geol. Soc. Am., 1989, pp. 397446.
[8] M.A. Belfoul, F. Faik, B. Hassenforder, Evidence of a tangential tectonic event prior to the major folding in the Variscan belt
of the western Anti-Atlas, Morocco, J. Afr. Earth Sci. 32 (4)
(2002) 723739.
[9] M. Benssaou, N. Hamoumi, The Lower-Cambrian western
Anti-Atlasic graben: tectonic control of paleogeography and sequential organisation, C. R. Geoscience 335 (2003) 297305.
[10] M. Benssaou, N. Hamoumi, Les microbialites de lAnti-Atlas
occidental (Maroc) : marqueurs stratigraphiques et tmoins des
changements environnementaux au Cambrien infrieur, C. R.
Geoscience 336 (2004) 109116.
[11] J. Bertrand-Sarfati, A. Moussine-Pouchkine, P. Affaton,
R. Trompette, Y. Bellion, Cover sequences of the West African
Craton, in: R.D. Dallmeyer, J.-P. Lcorch (Eds.), The West
African Orogens and Circum Atlantic Correlatives, SpringerVerlag, Berlin, 1991, pp. 6582.
[12] D.R.D. Boote, L.D.D. Clark, M.W. Traut, Palaeozoic petroleum
system of North Africa, in: D.S. MacGregor, R.T.J. Moody,
D.D. Clark Lowes (Eds.), Petroleum Geology of North Africa,
Geological Society, London, 1998, pp. 768.
[13] A. Boudda, G. Choubert, A. Faure-Muret, Essai de stratigraphie
de la couverture sdimentaire de lAnti-Atlas : Adoudounien
Cambrien infrieur, Notes Mm. Serv. gol. Maroc 271 (1979)
96.
[14] J. Boulin, M. Bouabdelli, M. El-Houicha, volution palogographique et godynamique de la chane Palozoque du moyen
Maroc : un essai de modlisation, C. R. Acad. Sci. Paris, Ser.
II 306 (1988) 15011506.
[15] S.E. Boyer, D. Elliott, Thrust systems, AAPG Bull. 66 (9)
(1982) 11961230.
[16] W. Buggisch, The global FrasnianFamennian Kellwasser
Event, Geol. Rundsch. 80 (1) (1991) 4972.
[17] W. Buggisch, R. Siegert, Paleogeography and facies of the grs
terminaux; uppermost Lower Cambrian, Anti-Atlas, Morocco,
in: V.H. Jacobshagen (Ed.), The Atlas System of Morocco;
Studies on its Geodynamic Evolution, Springer-Verlag, Berlin
HeidelbergNew York, 1988, pp. 107121.
22
[18] W. Buggisch, E. Fluegel, The Precambrian/Cambrian boundary in the Anti-Atlas (Morocco); discussion and new results,
in: V.H. Jacobshagen (Ed.), The Atlas System of Morocco;
Studies on its Geodynamic Evolution, Springer-Verlag, Berlin
HeidelbergNew York, 1988, pp. 8190.
[19] M. Burkhard, S. Caritg, U. Helg, C. Robert-Charrue, Forced,
disharmonic multilayer buckle folding in the Late Variscan
Anti-Atlas of Morocco, in: AAPG Annual Meeting, 2001, p. 30.
[20] S. Caritg, Gologie structurale dans lAnti-Atlas occidental du
Maroc. Implications tectoniques sur les relations entre dmes
de socle et couverture plisse en front de chane de montagne,
PhD thesis, University of Neuchtel, 2003.
[21] S. Caritg, M. Burkhard, R. Ducommmun, U. Helg, L. Kopp,
C. Sue, Fold interference patterns in the late Paleozoic AntiAtlas of Morocco, Terra Nova 16 (1) (2003) 2737.
[22] G. Choubert, Histoire gologique du Prcambrien de lAntiAtlas, Notes Mm. Serv. gol. Maroc 162 (1963) 352.
[23] G. Choubert, A. Faure-Muret, Anti-Atlas (Maroc), in: Tectonic
de lAfrique, in: UNESCO Earth Sci. Series, 1971, pp. 163
175.
[24] M.P. Coward, A.C. Ries, Tectonic development of North
African basins, in: T. Arthur, S. MacGregor Duncan, N.R. Cameron (Eds.), Petroleum Geology of Africa; New Themes and
Developing Technologies, Geological Society, London, 2003,
pp. 6183.
[25] R. Crossley, N. Mcdougall, Lower Palaeozoic reservoirs of
North Africa, in: D.S. MacGregor, R.T.J. Moody, D.D. Clark
Lowes (Eds.), Petroleum Geology of North Africa, Geol. Soc.
London, 1998, pp. 157166.
[26] R. Crossley, N. McDougall, D.S.E. MacGregor, R.T.J.E.
Moody, D.D.E. Clark-Lowes, Lower Palaeozoic reservoirs of
North Africa Petroleum geology of North Africa, Geol. Soc.
London, Spec. Publ. 132 (1998) 157166.
[27] F. Desthieux, Etude tectonique et mtallognique du Jbel Addana, Ordovicien des plaines du Dra, Maroc prsaharien, Notes
Serv. gol. Maroc 268 (38) (1977) 209236.
[28] J. Destombes, The Ordovician of the Moroccan Anti-Atlas, in:
M.G. Bassett (Ed.), The Ordovician System, Univ. Wales Press
and Natl. Mus. Wales, 1976, pp. 411413.
[29] J. Destombes, H. Hollard, S. Willefert, Lower Paleozoic rocks
of Morocco, in: C.H. Holland (Ed.), Lower Palaeozoic of
North-Western and West-Central Africa, Trinity Coll., Dep.
Geol., Dublin, Ireland, 1985, pp. 91336.
[30] M. Donzeau, J. Fabre, Tectonique des monts dOugarta, in:
Lexique stratigraphique international, Afrique de lOuest : introduction gologique et termes stratigraphiques, N.S.N., 1983,
pp. 118120.
[31] J. Dostal, J.D. Keppie, J. Hamilton, E.M. Aarab, J.P. Lefort,
J.B. Murphy, Crustal Xenoliths in Triassic lamprophyre dykes
in western Morocco: tectonic implications for the Rheic Ocean
suture, Geol. Mag. 142 (2) (2005) 159172.
[32] N. Ennih, J.-P. Ligeois, The Moroccan Anti-Atlas; the West
African craton passive margin with limited Pan-African activity; implications for the northern limit of the craton, Precambr.
Res. 112 (34) (2001) 289302.
[33] H. Feinberg, T. Aifa, J.-P. Pozzi, D. Khattach, J. Boulin,
Courbes de drive apparente des ples magntiques de
lAfrique et de la Meseta marocaine pendant le Palozoique,
C. R. Acad. Sci. Paris, Ser. II 310 (1990) 913918.
[34] L. Fichter, Geologic History of Virginia, http://Csmres.jmu.
edu/geollab/vageol/vahist/introduc.html, http://csmres.jmu.edu
/geollab/vageol/vahist/K-LatPal.html, accessed January 2005.
[35] D. Frizon de Lamotte, B. Saint-Bezar, R. Bracene, E. Mercier,

The two main steps of the Atlas building and geodynamics of
the western Mediterranean, Tectonics 19 (4) (2000) 740761.
[36] D. Frizon de Lamotte, A. Crespo-Blanc, B. Saint-Bzar, M. Comas, M. Fernndez, H. Zeyen, P. Ayarza, C. Robert-Charrue,
A. Chalouan, M. Zizi, A. Teixell, M.-L. Arboleya, F. Alvarez-Lobato, M. Julivert, A. Michard, TRANSMED Transect
I: Iberian MesetaGuadalquivir BasinBetic CordilleraAlboran SeaRifMoroccan MesetaHigh AtlasSahara Platform,
in: W. Cavazza, et al. (Eds.), The TRANSMED Atlas: The
Mediterranean Region from Crust to Mantle, Springer, Berlin,
2004, p. 141 & CD-ROM.
[37] G. Geyer, E. Landing (Eds.), MOROCCO95 The Lower
Middle Cambrian standard of western Gondwana, Beringeria,
Special Issue 2, University of Wrzburg, Germany, 1995.
[38] G. Geyer, E. Landing, W. Heldmaier, Faunas and depositional
environments of the Cambrian of the Moroccan Atlas region,
in: G. Geyer, E. Landing (Eds.), MOROCCO95 The Lower
Middle Cambrian Standard of Western Gondwana, Beringeria, Special Issue 2, University of Wrzburg, Germany, 1995,
pp. 47119.
[39] J.-F. Ghienne, Late Ordovician sedimentary environments,
glacial cycles, and post-glacial transgression in the Taoudenni Basin, West Africa, Palaeogeogr. Palaeoclimatol. Palaeoecol. 189 (34) (2003) 117145.
[40] F.M. Gradstein, et al., International Stratigraphic Chart, http://
www.stratigraphy.org/chus.pdf, Stratigraphy I.C.O. Editor,
2004.
[41] J.R. Graham, Wave-dominated shallow-marine sediments in the
Lower Carboniferous of Morocco, J. Sediment. Petrol. 52 (4)
(1982) 12711276.
[42] H. Haddoum, R. Guiraud, A. Moussine-Pouchkine, Hercynian
compressional deformations of the AhnetMouydir Basin, Algerian Saharan Platform: far-field stress effects of the Late
Palaeozoic orogeny, Terra Nova 13 (3) (2001) 220226.
[43] B. Hassenforder, La tectonique panafricaine et varisque de
lAnti-Atlas dans le massif du Kerdous (Maroc), PhD thesis,
University of Strasbourg, France, 1987.
[44] R.D. Hatcher Jr., W.A. Thomas, P.A. Geiser, A.W. Snoke,
S. Mosher, D.V. Wiltschko, Alleghanian Orogen, in: R.D.
Hatcher Jr., W.A. Thomas, G.W. Viele (Eds.), The Geology
of North America, The AppalachianOuachita Orogen in the
United States, Geol. Soc. Am., 1989, pp. 233318.
[45] K.P. Hefferan, H. Admou, R. Hilal, J.A. Karson, A. Saquaque,
T. Juteau, M.M. Bohn, S.D. Samson, J.M. Kornprobst, Proterozoic blueschist-bearing melange in the Anti-Atlas Mountains,
Morocco, Precambr. Res. 118 (34) (2002) 179194.
[46] K.P. Hefferan, H. Admou, J.A. Karson, A. Saquaque, AntiAtlas (Morocco) role in Neoproterozoic western Gondwana reconstruction, Precambr. Res. 103 (12) (2000) 8996.
[47] U. Helg, M. Burkhard, S. Caritg, C. Robert-Charrue, Folding
and inversion tectonics in the Anti-Atlas of Morocco, Tectonics 23 (2004) 17, TC 4006.
[48] Y. Hervout, G. Due, Analyse morphostructurale par imagerie satellitaire et coupes structurales modlises des monnts
dOugarta (Sahara occidental, Algrie) : une chane hercynienne chevauchante plis passifs, Mm. Serv. gol. Algrie 8
(1996) 127173.
[49] C. Hoepffner, M.R. Houari, M. Bouabdelli, Tectonics of the
north African Variscides (Morocco, Western Algeria), an outline, C. R. Geoscience 338 (2006).
[50] H. Hollard, Tableaux de corrlations du Silurien et du Dvonien

de lAnti-Atlas, Notes Serv. gol. Maroc (1981) 42.
[51] H. Hollard, J.P. Schaer, Southeastern Atlantic Canada, Northwestern Africa, and Continental Drift; discussion, Can. J. Earth
Sci. 10 (4) (1973) 584586.
[52] M.R. Houari, C. Hoepffner, Structures des terrains palozoiques la limite sud de la chane hercynienne du Maroc (Haut
Atlas oriental), Afr. Geosci. Rev. 7 (1) (2000) 3953.
[53] P.D. Howell, B.A. Van der Pluijm, Structural sequences and
styles of subsidence in the Michigan basin, Geol. Soc. Am.
Bull. 111 (7) (1999) 974991.
[54] M.M. Joachimski, R.D. Pancost, K.H. Freeman, H.C. Ostertag,
W. Buggisch, Carbon isotope geochemistry of the Frasnian
Famennian transition, in: G. Racki, M.R. House (Eds.), Late
Devonian Biotic Crisis; Ecological, Depositional and Geochemical Records, Elsevier, Amsterdam, 2002.
[55] M. Leblanc, J.R. Lancelot, Interprtation godynamique du
domaine pan-africain (Prcambrien terminal) de lAnti-Atlas
(Maroc) partir de donnes gologiques et gochronologiques,
Can. J. Earth Sci. 17 (1) (1980) 142155.
[56] P. Legrand, Palogographie du Sahara algrien lOrdovicien
terminal et au Silurien inferieur, Bull. Soc. gol. France 174 (1)
(2003) 1932.
[57] M.W. Leighton, D.R. Kolata, Selected interior cratonic basins
and their place in the scheme of Global Tectonics A Synthesis, in: M.W. Leighton, et al. (Eds.), Interior Cratonic Basins,
AAPG, Tulsa, 1990, pp. 729797.
[58] D.S. MacGregor, R.T.J. Moody, D.D. Clark-Lowes, Petroleum
Geology of North Africa, Geological Society London, Spec.
Publ. 132 (1998).
[59] B. Mamet, G. Choubert, L. Hottinger, Notes sur le Carbonifre
du Jebel Ouarkziz. tude du passage du Visen au Namurien
daprs les Foraminifres, Notes Mm. Serv. gol. Maroc 27
(1966) 421.
[60] S. Marshak, K. Karlstrom, J.M. Timmons, Inversion of Proterozoic extensional faults: An explanation for the pattern of
Laramide and Ancestral Rockies intracratonic deformation,
United States, Geology 28 (8) (2000) 735738.
[61] M. Mattauer, F. Proust, P. Tapponnier, Major strike-slip fault of
Late Hercynian age in Morocco, Nature 237 (1972) 160162.
[62] P. Matte, Variscides between the Appalachians and the Urals;
similarities and differences between Paleozoic subduction and
collision belts, in: J.R. Martinez Catalan, et al. (Eds.), Variscan
Appalachian Dynamics; the Building of the Late Paleozoic
Basement, Geol. Soc. Am., 2002.
[63] J.H. McBride, J.H. Knapp, Review of seismic reflector signatures of crustal deformation in the Appalachian-Caledonide
Orogen with reference to the Spanish Variscides and the
Uralides, in: J.R. Martinez Catalan, et al. (Eds.), Variscan
Appalachian Dynamics; the Building of the Late Paleozoic
Basement, Geol. Soc. Am., 2002, pp. 281300.
[64] A. Michard, lments de gologie marocaine, Notes Mm.
Serv. gol. Maroc 252 (1976) 408.
[65] A. Michard, J. Sougy, Lorogense hercynienne la lisire
nord-ouest de lAfrique (Structure des chanes primaires du
Maroc au Sngal), in : Colloque international CNRS : La
Chane varisque dEurope moyenne et occidentale, Rennes,
France, 1977, pp. 605640.
[66] A. Michard, A. Yazidi, F. Benziane, H. Hollard, S. Willefert,
Foreland thrusts and olistostromes on the pre-Sahara margin of
the Variscan Orogen, Morocco, Geology 10 (5) (1982) 253
256.
23
[67] G. Mitra, V.S. Mount, Foreland basement involved structures,

AAPG Bull. 82 (1) (1998) 70109.
[68] A.M. Morabet, R. Bouchta, H. Jabour, An overview of the
petroleum systems of Morocco, in: D.S. MacGregor, R.T.J.
Moody, D.D. Clark Lowes (Eds.), Petroleum Geology of North
Africa, Geological Society, London, 1998, pp. 283296.
[69] A. Piqu, Variscan terranes in Morocco, Spec. Pap. Geol. Soc.
Am. 230 (1989) 115129.
[70] A. Piqu, Gologie du Maroc, Les domaines rgionaux et leur
volution structurale, PUMAG, Marrakech, 1998.
[71] A. Piqu, Geology of Northwest Africa, Beitrge zur regionalen
Geologie der Erde, vol. 29, Gebrder Borntrger Verlag, Berlin,
Stuttgart, 2001.
[72] A. Piqu, M. Bouabdelli, A. Soulaimani, N. Youbi, M. Iliani,
Upper Neoproterozoic PIII conglomerates in the Anti-Atlas,
southern Morocco; Pan-African molasses, or indicators of Upper Proterozoic rifting?, C. R. Acad. Sci. Paris, Ser. IIa 328 (6)
(1999) 409414.
[73] A. Piqu, A. Michard, Moroccan hercynides: A synopsis. The
Palaeozoic sedimentary and tectonic evolution at the northern
margin of West Africa, Am. J. Sci. 289 (1989) 286330.
[74] A. Piqu, J.-J. Corne, J. Mller, J. Roussel, The Moroccan Hercynides, in: R.D. Dallmeyer, J.O.P. Lcorch (Eds.),
The West African Orogens and Circum Atlantic Correlatives,
Springer-Verlag, Berlin, 1991, pp. 229263.
[75] M. Roddaz, S. Brusset, J.-C. Soula, D. Beziat, M. Ben-Abou,
P. Debat, Y. Driouch, F. Christophoul, A. Ntarmouchant, J. Deramound, Foreland basin magmatism in the western Moroccan
Meseta and geodynamic interferences, Tectonics 21 (5) (2002)
23.
[76] J. Rodgers, Lines of basement uplifts within cratons marginal
to orogenic belts, Am. J. Sci. 287 (1987) 661692.
[77] J. Rodgers, Fold-and-thrust belts in sedimentary rocks. Part 1:
Typical examples, Am. J. Sci. 290 (1990) 321359.
[78] J. Rodgers, Fold-and-thrust belts in sedimentary rocks. Part 2:
Other examples, especially variants, Am. J. Sci. 291 (1991)
825886.
[79] J. Rodgers, Lines of basement uplifts within the external parts
of orogenic belts, Am. J. Sci. 295 (1995) 455487.
[80] M. Sahabi, D. Aslanian, J.-L. Olivet, Un nouveau dpart pour
lhistoire de lAtlantique central, C. R. Geoscience 336 (2004)
10411052.
[81] A. Saquaque, H. Admou, J.A. Karson, K. Hefferan, I. Reuber,
Precambrian accretionary tectonics in the Bou Azzer-El Grara
region, Anti-Atlas, Morocco, Geology 17 (1989) 11071110.
[82] P.E. Schenk, Southeastern Atlantic Canada, Northwestern
Africa, and Continental Drift, Can. J. Earth Sci. 8 (1971) 1218
1251.
[83] C.R. Scotese, Paleomap project, http://www.scotese.com/
Default.htm, 2005.
[84] J.L. Simon, Superposed buckle folding in the eastern Iberian
Chain, Spain, J. Struct. Geol. 26 (2004) 14471464.
[85] L.L. Sloss, Epilog (to Interior Cratonic Basins), in: M.W.
Leighton, et al. (Eds.), Interior Cratonic Basins, AAPG, Tulsa,
1990, pp. 799805.
[86] A.G. Smith, Gondwana: its shape, size and position from Cambrian to Triassic times, J. Afr. Earth Sci. 28 (1) (1999) 7197.
[87] A. Soulaimani, Dynamique et interactions Socle/Couverture
dans lAnti-Atlas occidental (Maroc) : Rifting fini-protrozoque et orogense hercynienne, 1998.
[88] A. Soulaimani, A. Piqu, The Tasrirt structure (Kerdous inlier,
Western Anti-Atlas, Morocco): a Late Pan-African transtensive
dome, J. Afr. Earth Sci. 39 (2004) 247255.
24
[89] A. Soulaimani, M. Bouabdelli, A. Piqu, Lextension continentale au No-Protrozoque suprieurCambrien infrieur dans
lAnti-Atlas (Maroc), Bull. Soc. gol. France 174 (1) (2003)
8392.
[90] A. Soulaimani, A. Essaifi, N. Youbi, A. Hafid, Les marqueurs
structuraux et magmatiques de lextension crustale au Protrozoque terminalCambrien basal autour du massif de Kerdous
(Anti-Atlas occidental, Maroc), C. R. Geoscience 336 (16)
(2004) 14331441.
[91] A. Soulaimani, C. Le Corre, R. Farazdaq, Deformation hercynienne et relation socle/couverture dans le domaine du Bas-Draa
(Anti-Atlas occidental, Maroc), J. Afr. Earth Sci. 24 (3) (1997)
271284.
[92] A. Soulaimani, A. Piqu, M. Bouabdelli, La srie du PIIPIII
de lAnti-Atlas occidental (Sud marocain) : un olistostrome la
base de la couverture post-panafricaine (PIII) du Protrozoque
suprieur, Earth Planet. Sci. Lett. 332 (2001) 121127.
[93] G.M. Stampfli, G.D. Borel, A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic plate boundaries
and restored synthetic oceanic isochrons, Earth Planet. Sci.
Lett. 196 (12) (2002) 1733.
[94] D. Stone, Morphology of the Casper Mountain uplift and related subsidiary structures, central Wyoming: Implications for
Laramide kinematics, dynamics, and crustal inheritance, AAPG
Bull. 86 (8) (2002) 14171440.
[95] O.E. Sutcliffe, J.A. Dowdeswell, R.J. Whittington, J.N. Theron,
J. Craig, Calibrating the Late Ordovician glaciation and mass
extinction by the eccentricity cycles of Earths orbit, Geology 28 (11) (2000) 967970.
[96] R.J. Thomas, L.P. Chevallier, P.G. Gresse, R.E. Harmer, B.M.
Eglington, R.A. Armstrong, C.H. De Beer, J.E.J. Martini, G.S.
De Kock, P.H. Macey, B.A. Ingram, Precambrian evolution of
the Sirwa Window, Anti-Atlas Orogen, Morocco, Precambr.
Res. 118 (2002) 157.
[97] I.R. Vann, R.H. Graham, A.B. Hayward, The structure of mountain fronts, J. Struct. Geol. 8 (34) (1986) 215227.
[98] G.J. Walsh, J.N. Aleinikoff, F. Benziane, A. Yazidi, T.R. Armstrong, UPb zircon geochronology of the Paleoproterozoic
Tagragra de Tata inlier and its Neoproterozoic cover, western
Anti-Atlas, Morocco, Precambr. Res. 117 (2002) 120.
[99] R. Weijermars, Estimation of paleostress orientation within deformation zones between two mobile plates, with Suppl. Data
9329, Geol. Soc. Am. Bull. 105 (11) (1993) 14911510.
[100] J. Wendt, Disintegration of the continental margin of northwestern Gondwana; Late Devonian of the eastern Anti-Atlas
(Morocco), Geology 13 (11) (1985) 815818.
[101] J. Wendt, Z. Belka, E. Fluegel, Age and depositional environment of Upper Devonian (Early Frasnian to Early Famennian)
black shales and limestones (Kellwasser facies) in the eastern
Anti-Atlas, Morocco, Facies 25 (1991) 5189.
[102] Tectonic Map of the World, AAPG Foundation, R.W. Wiener,
F.L. Wehr, G.M. Skerlec, I.O. Norton, 1985.
[103] P.A. Ziegler, S. Cloetingh, Dynamic processes controlling evolution of rifted basins, Earth Sci. Rev. 64 (12) (2004) 150.
[104] P.A. Ziegler, S. Cloetingh, R. Guiraud, G.M. Stampfli, PeriTethyan platforms; constraints on dynamics of rifting and
basin inversion, in: P.A. Ziegler, et al. (Eds.), Peri-Tethyan
Rift/Wrench Basins and Passive Margins, ditions du Museum
national dhistoire naturelle, Paris, 2001.
Geological Maps
[105] D. Alaoui-Mdaghri, M. Bensaid, M. Dahmani, Carte gologique du Maroc 1:100 000, Bou Izakarn, Notes Serv. gol.
Maroc (1992).
[106] G. Choubert, Y. Ennadifi, Carte gologique du Maroc
1:200 000, Flanc sud de lAnti-Atlas occidental et des plaines
du Dra, Akka-Tafragount-Tata, Notes Serv. gol. Maroc (1970).
[107] G. Choubert, A. Faure-Muret, Carte gologique du Maroc
1:100 000, Igherm, feuille NH-29-XVI-2, Notes Serv. gol.
Maroc (1983).
1:100 000, Tafraout ; feuille NH-29-X-3, Notes Serv. gol.
Maroc (1983).
1:100 000, Taroudannt, Feuille NH-29-XVI-I, Notes Serv. gol.
Maroc (1983).
[110] M.A. El-Alaoui, Y.C. Chefchaouni, M. Diouri, Carte
gologique du Maroc 1:200 000, Flanc sud de lAnti-Atlas et
des plaines du Dra, Foum el Hassane-Assa, Notes Serv. gol.
Maroc (1963).
[111] S.E.M. Fetah, M. Bensaid, M. Dahmani, Carte gologique du
Maroc 1:200 000, TafilaltTaouz, ministre de lnergie et des
Mines, royaume du Maroc (1986).
Maroc au 1:200 000, TodrhaMader, ministre de lnergie et
des Mines, royaume du Maroc (1986).
Maroc au 1:200 000, ZagoraCoude du DraHamada du Dra,
ministre de lnergie et des Mines, royaume du Maroc (1986).
[114] A. Guerraoui, M. Bensaid, M. Dahmani, Carte gologique du
Maroc 1:100 000, Fask, Notes Serv. gol. du Maroc (1997).
TECTONICS, VOL. 23, TC4006, doi:10.1029/2003TC001576, 2004
Folding and inversion tectonics in the Anti-Atlas

of Morocco
Urs Helg, Martin Burkhard, Severine Caritg, and Charles Robert-Charrue
Institut de Geologie, Universite de Neuchatel, Neuchatel, Switzerland
Received 25 August 2003; revised 30 April 2004; accepted 14 May 2004; published 22 July 2004.
[1] The late Variscan Anti-Atlas of Morocco shows

some conspicuous deviations from the standard
anatomy of foreland fold-and-thrust belts. Large
basement inliers crop out at a very short distance of
less than 50 km behind the southeastern front of the
fold belt, reminiscent of Windriver-style basement
uplifts. In contrast to the Rocky Mountain foreland,
however, the Anti-Atlas basement uplifts punctuate
tightly folded Paleozoic cover series similar in tectonic
style to the Appalachian Valley and Ridge province.
Cover shortening is exclusively accommodated by
buckle folding, and the Anti-Atlas fold belt lacks any
evidence for duplexing or thrust faults other than the
occasional steep reverse fault found near basement
inliers. Basement domes have classically been
considered as the result of vertical tectonics in a
horst and graben fashion, or, alternatively, as large
plis de fond [Argand, 1924], basement folds.
Unfolding of a large portion of an Ordovician
quartzite marker bed reveals a minimum shortening
of 17% (30 km). Balancing this section at the crustal
scale indicates a lower crustal detachment level at 18
to 25 km depth. Basement shortening is inferred to be
accommodated through massive inversion of former
extensional faults, inherited from a Late ProterozoicINDEX TERMS: 8005
Lower Cambrian rifting phase.
Structural Geology: Folds and folding; 8015 Structural Geology:
Local crustal structure; 8102 Tectonophysics: Continental
contractional orogenic belts; 8159 Tectonophysics: Rheology
crust and lithosphere; 9305 Information Related to Geographic
Region: Africa; KEYWORDS: Anti-Atlas, tectonics, balanced cross
section, folding, foreland fold and thrust belt. Citation: 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.
1. Introduction
[2] The Anti-Atlas foreland fold belt of southwestern
Morocco is part of the Variscan Appalachian-OuachitaMauretanides chain. It shows some important deviations
from typical foreland fold and thrust belts [Rodgers, 1990].
Large structural domes, the so-called boutonnieres
Copyright 2004 by the American Geophysical Union.
0278-7407/04/2003TC001576$12.00
[Choubert, 1963] of Proterozoic basement rocks, crop

out at a very short distance behind the southeastern front
of the orogen. Detailed mapping of the Anti-Atlas did not
reveal any evidence for major thrusts within this fold belt.
Layer-parallel decollements or detachments are required
between different structural levels, but no such decollement has ever been mapped to step up in a ramp-flat
geometry. The complete lack of thrusts lead earlier authors
to consider the entire basement uplifts as crustal-scale
folds [Choubert, 1963; Choubert and Faure Muret,
1971], inspired by the plis de fonds of Argand [1924,
Figures 5a, 5b, and 5c].
[3] The sedimentary cover of the Anti-Atlas includes the
uppermost Proterozoic and an up to 10 km thick Paleozoic
series. These mildly deformed and nearly unmetamorphosed
strata of the Anti-Atlas have received much attention from
stratigraphers and paleontologists [Destombes et al., 1985;
Bertrand-Sarfati et al., 1991; Villeneuve and Cornee, 1994;
Pique, 2001]. Special attention has been given to the
Proterozoic-Cambrian boundary [Boudda et al., 1979;
Bertrand-Sarfati, 1981; Buggisch, 1988b; Latham, 1988;
Geyer and Landing, 1995; Benssaou and Hamoumi, 2003].
In contrast to this wealth of stratigraphical and paleontological literature, structural publications dealing with the
tectonics of the western Anti-Atlas remain scarce [Leblanc,
1972; Michard, 1976; Soulaimani, 1998; Guiton et al.,
2003].
[4] On the basis of paleomagnetic observations near the
South Atlas and the TizinTest faults, a strike-slip origin for
the entire Anti-Atlas chain has been proposed [Mattauer et
al., 1972]. This view is widely shared by French authors
[Leblanc, 1972; Donzeau, 1974; Jeannette, 1981], and
folding of the sedimentary cover has mostly been explained
as draping over vertically uplifted basement blocks and/or
as wrenching above inherited, subvertical zones of weakness [Faik et al., 2002]. Horizontal shortening in the
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
depict the basement inliers as horst and graben structures,
indistinct broad basement domes or delimited by vertical
strike-slip faults [Michard, 1976; Michard and Sougy, 1977;
Pique et al., 1991]. Clearly, these cross sections imply no
crustal-scale shortening within the basement, and fold trains
within the synclinoria remain unexplained in terms of
material balance, a question which has never been
addressed in the Anti-Atlas. The striking difference between
the Devonian Jbel Rich folds and the virtually undeformed 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, Draa 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

strike-slip fault [Weijermars, 1993], of a Neogene age. The
apparent en echelon arrangement of the Rich folds was
further used to deduce a right lateral shear sense and amount
of shear, assuming a subvertical plunge of fold axes
[Weijermars, 1993]. In reality, however, fold axes plunges
are very gentle, less than 30, and the asymmetric fold
pattern of the lower Draa valley implies very weak, if any,
wrenching deformations [Soulaimani et al., 1997]; paleostress indicators are systematically orientated nearly perpendicular to the fold axial planes, rather than at an inferred
angle of some 50 [Weijermars, 1993]. New observations in
the southwestern Anti-Atlas have recently led to a general
revision of the classic model of vertical and strike-slip
tectonics [Soulaimani et al., 1997; Soulaimani, 1998].
Convincing evidence exists for southwestward thrusting of
the Lower Draa inlier onto the sedimentary cover, and the
dextral strike-slip boundary postulated by Weijermars
[1993] is not substantiated by any field evidence.
[5] In this study, we present a detailed structural analysis
of a large portion of the SW Anti-Atlas. A continuous fold
train in a competent Ordovician quartzite bed is mapped in
detail and rendered in a three-dimensional (3-D) structural
model. Unfolding of this marker bed provides minimum

estimates of horizontal shortening as well as a measure of
regional variations in folding intensity. The consequences of
this tectonic shortening are discussed at the larger, crustal
scale of this thick skinned foreland fold belt.
2. Geological Framework of the Western

Anti-Atlas
2.1. Basement Preconfiguration
[6] The Anti-Atlas fold belt is located at the northwestern
border of the West African craton and continues southward
into the Mauretanides. Large-scale basement cored domes
with irregular shapes form a continuous area of positive
structural relief (Figures 1 and 2). The long axis of the AntiAtlas belt axis is oriented SW-NE and measures some
700 km.
[7] The pre-Cambrian basement has a complex geologic
history. It was consolidated during the Panafrican orogeny
between 620 and 580 Ma [Choubert, 1963; Choubert
and Faure Muret, 1971; Leblanc and Lancelot, 1980;
Hassenforder, 1987]. In detail it shows considerable
2 of 17
TC4006
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

West African craton and a Proterozoic mobile belt to the
north. Remnants of this suture zone are exposed in several
basement domes, for example in the Bou Azzer and in the
El Graara inliers. In the central Anti-Atlas this suture zone
is known as accident majeur de lAnti-Atlas. Southwest
of this suture zone, Eburnean (approximately 1900 Ma)
augen gneisses, metadolerites, and metamorphic rocks, the
so-called Zenaga and Kerdous Series, crop out in several of
the structural domes. The Proterozoic rocks are further
characterized by numerous synorogenic granite and granodiorite intrusions.
[8] The Panafrican orogeny is followed by an important
phase of continental sedimentation, which results in locally
thick clastic and volcaniclastic deposits, the so-called PII-III
unit, separated from the crystalline basement by a major

unconformity. The geodynamic context of the PII-III is still
a matter of debate [Pique et al., 1999; Pique, 2001]. In the
case of the overlying PIII unit of clastics and volcaniclastics, separated by a minor unconformity from the PII, there
is increasing evidence in favor of an extensional tectonic
regime, in relation with a new rifting cycle starting in
the Late Proterozoic [Pique et al., 1995; Pique, 2001;
Soulaimani et al., 2003]. Extensional structures have been
described within the Proterozoic basement in the westernmost Anti-Atlas [Soulaimani et al., 1997] and large-scale
half-grabens are documented in the central Anti-Atlas
[Azizi-Samir et al., 1990]. The post-Panafrican geological
history of the Anti-Atlas is now considered as an aborted rift
[Pique et al., 1995; Soulaimani et al., 2003].
3 of 17
TC4006
TC4006
contact an easy target for mapping both in the field and on

satellite images (Figure 2). The same unconformity is also
used to define the outline of the basement inliers.
[10] A new sedimentary cycle leading to the formation
of the Anti-Atlas basin starts in the uppermost Proterozoic
and ends in the upper Carboniferous. In the western AntiAtlas, a total of up to 12 km of varied clastic and
shallow marine sediments were deposited, while eastward,
in the central Anti-Atlas, the thickness is decreasing
to about 8 km [Michard, 1976; Pique and Michard,
1989; Kogbe, 1998]. The Anti-Atlas basin fill is characterized by a high amount of fine-grained detrital, clay
rich sediments: muddy siltstones and shales (Figure 3).
Competent quartzites, sandstones, limestones, and conglomerates form thin but continuous beds at different
stratigraphic levels. Coarse-grained beds are increasing
in thickness and abundance toward the easternmost
Anti-Atlas.
[11] The thin competent marker beds form very continuous spectacular crests easily visible in aerial and satellite
images (Figure 2). Major stratigraphic intervals such as the
Ordovician and the Devonian series have been named
according to the mountain belts they form: Jbel Bani (for
Ordovician), Jbel Rich (Devonian), and Jbel Ouarkziz
(Carboniferous). We continue to use this terminology in
order to describe the main competent units, despite the fact
that they are not formally accepted lithostratigraphic
formation names.
2.3. Variscan Orogeny
Figure 3. Synthetic stratigraphic column of the western

Anti-Atlas. Incompetent units are shown in black. The
asterisks marks the slight unconformity which is classically
used to define the outlines of the boutonnieres (basement
inliers). The commonly used, informal lithostratigraphic/
mechanical units are labeled at the right-hand side;
chronostratigraphy is shown to the left.
2.2. Sedimentary Cover

[9] Above the PIII conglomerates, a slight unconformity
marks the onset of epicontinental shallow marine sedimentation in the Anti-Atlas, a limit often referred to as the
basement cover contact. A marked contrast in color,
sediment type and rheology makes the PIII-Adoudounian
[12] The age of deformation within the western AntiAtlas is generally considered as middle to late Carboniferous on the basis of the youngest age of (slightly) deformed
and tilted sediments on the one hand, and superregional
considerations and comparisons on the other hand
[Michard, 1976; Pique and Michard, 1989; Pique, 1994].
The oldest sedimentary formations sealing eroded AntiAtlas folds are of Upper Cretaceous age and do not provide
any useful timing constraints. Dolerite dikes such as the
famous dike of Foum-Zguid and sills clearly postdate AntiAtlas folding [Sebai et al., 1991], but their lower Jurassic
age does not put any tight time constraint on the deformation age of Anti-Atlas folding either, since folding might
still be anything between middle Carboniferous and Late
Triassic. Some attempts at a direct age determination of
the tectonic cleavage have been made using radiometric
methods, applied to fine grained illite-muscovite assemblages of the westernmost Anti-Atlas [Bonhomme and
Hassenforder, 1985]. Isotopic ages of 370 Ma and 290 Ma
were attributed to a pre-Variscan phase of deformation and
the peak of regional metamorphism, respectively. Similar
results have been obtained in isotopic studies of the Variscan
Meseta of northern and central Morocco [Huon, 1985].
Since metamorphism in many parts of the Variscan chain
of Morocco is barely of lower greenschist facies, and
deformation only locally penetrative, any K-Ar and Rb-Sr
ages should be regarded very cautiously [Villa, 1998].
Correlations between the northern Moroccan Meseta and
the Anti-Atlas are not straightforward either. The two
4 of 17
TC4006
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 Draa plain corresponding to a lowland eroded into Silurian shales; view is to the north.
provinces are separated by the South Atlas shear zone, a

major terrane boundary within the Variscan chains of
Morocco [Mattauer et al., 1972]. An overall geodynamic
interpretation of a complete Canadian-Moroccan transect
through the entire Appalachian-Meseta-Anti-Atlas system
has been presented as early as 1971 [Schenk, 1971].
2.4. Anti-Atlas Foreland Fold Belt
[13] In terms of folding and deformation style, four
structural units with different wavelengths and amplitudes
can be distinguished in the Anti-Atlas. They correspond to
four distinct stratigraphic levels separated by the thick
incompetent units of the Middle Cambrian, Silurian, and
Upper Devonian respectively. None of the fold levels shows
a clear and consistent vergence.
[14] The lowermost unit is represented by the basement
domes and their autochthonous cover consisting of Late
Proterozoic PIII conglomerates and Lower Cambrian limestones and dolomites of the so-called Adoudounian.
These basement domes have accessible, mapped amplitudes
of up to 2 km, and the corresponding wavelength is on the
order of several kilometers to tens of kilometers. The true
amplitude of basement folds is in excess of 10 km,
however, larger than the total thickness of Paleozoic cover
series. Locally, thin layers of incompetent shales and
evaporites within the autochthonous Adoudounian series
act as minor detachment horizons, resulting in second-order
folding at a much smaller scale (10 to 100 m). Toward the
central Anti-Atlas the Lie de vin formation acts as a
distinct local decollement horizon (Figure 2). Fold axis
trends within the Adoudounian are strongly influenced by
the preconfiguration of the underlying basement domes, and
they often deviate considerably from the large-scale SW-NE
Anti-Atlas orientation [Soulaimani et al., 1997; Soulaimani,
1998]. Higher up in the stratigraphic column, two structural
units are characterized by detachment folds of the Ordovi-
cian Jbel Bani quartzites (Figure 4) and the Lower Devonian

Jbel Rich limestones, respectively. In the western Anti-Atlas
they are separated from each other by about 1000 to 1500 m
of Silurian shales and mudstones. The folding style of these
two units is very similar, albeit with different wavelengths
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
Rich folds display smaller wavelengths and amplitudes at
the 500 m scale. The uppermost structural unit, a thick
series of Carboniferous sandstones and limestones of the
Jbel Ouarkziz, is separated from the Jbel Rich folds by a
thick series of Upper Devonian shales. Morphologically, the
Jbel Ouarkziz defines a very continuous monocline, where
the Carboniferous series dip gently south-southeastward
into the adjacent Tindouf basin (Figure 2). Some authors
(e.g., M. Zizi, personal communication, 2000; see also
http://www.onarep.com) consider the lack of folding within
the tilted Jbel Ouarkziz as evidence for a Late Devonian age
of folding in the Anti-Atlas, an interpretation inspired by
comparisons with northern and central Morocco and with
the Canadian Appalachians [Pique and Michard, 1989].
Direct field evidence in the form of an unconformity has
never been identified in the sedimentary series of the AntiAtlas, where a tectonic interpretation (decollement) is more
generally accepted [Soulaimani, 1998]. While the deformation front of the Anti-Atlas system is classically seen at the
base (NNW) of the Jbel Ouarkziz, we propose a new
interpretation as a triangle structure with a forethrust ending
somewhere blindly SSE of the Jbel Ouarkziz, at depth
within the Tindouf basin. From this most external tip point
of the Anti-Atlas thrust system, a NNW vergent backthrust
climbs north-northwestward to emerge into the shale series
separating the Jbel Ouarkziz monocline from the Jbel Rich
fold train within the Lower Draa valley. This interpretation
is corroborated by structural observations, notably the tilting
of the Jbel Ouarkziz monocline and bed internal layer
5 of 17
TC4006
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 decollement is inferred within the incompetent Lie de vin
formation of Lower/Middle Cambrian age. The structural style below this decollement 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

youngest outcropping Carboniferous limestones [Burkhard
et al., 2001].
3. Anti-Atlas Cross-Section Balancing

3.1. Theoretical Considerations
[15] In areas of tectonic activity, deformation typically
occurs at all scales, ranging from structures in crystal
lattices to structures forming mountain ranges [Ramsay
and Huber, 1983; Mitra, 1992; Mitra, 1994]. Three scales
of observation are commonly used to describe structures:
microscale (<0.01 m), outcrop scale (0.01 100 m), and map
scale (>100 m). Primary sedimentary structures are used as
proxies for the qualitative description and quantitative
determination of strain at all scales.
[16] Cross-section balancing is dealing with map-scale
structures such as folds, faults, and thrusts. Smaller-scale
structures accommodate additional deformation, which
should be considered in overall estimations of tectonic
shortening. Over the years, various investigators have
shown that small-scale structures do make a significant
contribution to total tectonic deformation. Mitra [1994]
measured strains in sections parallel to the transport direction across the Sevier fold-and-thrust belt. He could demonstrate that failure to include microscale deformation in the
restoration of regional cross sections results in overestimation of the wedge taper for the thrust belt. Another study
examined the contribution of both microscale and outcropscale data to regional deformation using the foreland thrust
belt of the central Appalachians where small-scale structures account for more than 75% of the shortening in the
roof sequence of this thrust belt [Smart et al., 1997]. Hogan
and Dunne [2001] conducted a study in the Upper Devonian Chemung-Brallier Formation boundary in the western
Valley and Ridge province, where map-scale structures
account for about 50% of the total shortening, whereas
outcrop-scale and microscale shortening yield each close to

25% of the overall shortening.
[17] Qualitatively, layer parallel shortening features are
widely recognized as predating folding and thrusting and
this is very nicely demonstrated in the Jura mountains
[Plessmann, 1972; Homberg, 1996].
[18] For the estimation of the total shortening in the AntiAtlas we concentrated on map-scale and outcrop-scale
structures. Since there are other important unknowns of
the shortening at the scale of the orogen we were content
with a qualitative assessment of microscale shortening. In
order to minimize the effects of LPS, we choose a very
competent quartzite horizon, devoid of any LPS features
visible at the outcrop to hand specimen scale. We believe
that this marker bed deformed essentially by buckling at the
outcrop to map scale.
3.2. Study Area and the Ordovician Quartzite
Marker Bed
[19] Our estimation of the shortening is mainly based
on measurements and observations made in the Bani
mountain range, an area of folded Ordovician rocks between
the towns of Bou Izakarn, Aouinat Torkoz, and Icht
(Figure 1b) with excellent outcrop conditions (Figure 2).
The structural style is quite homogeneous, and fold trends
are easily recognized.
[20] We concentrated on a very competent horizon within
the so-called 2nd Bani formation of Ashgillian age. This
uppermost competent unit of the Jbel Bani is labeled or6c
according to Guerraoui et al. [1997] (Figure 3); on older
maps, however, this thin marker bed is not always distinguished from the immediately underlying or6b or or6s. It
consists of massive microconglomeratic quartz arenites,
interpreted as being of glacial origin. This particular horizon
is ideal for the reconstruction of the map-scale structures:
(1) It is ubiquitous in the study area and has a relatively
small thickness of 50 to 90 m [Destombes et al., 1985],
6 of 17
TC4006
which is rather thin compared to the average wavelength of

the folds. (2) The massive or6c quartzites are much more
resistant to weathering and erosion than the Silurian shales
immediately above. The top of the or6c horizon is often
very well preserved and forms the present-day surface, often
in the form of dip slopes. In consequence, where the or6cquartzites crop out, the topographic contours can directly be
used as structure contours (Figure 4). Owing to very
favorable outcrop conditions, large areas could be checked
for outcrop-scale structures through the examination of
aerial photographs and verification in the field. (3) By
coincidence, the present-day valley floors in the western
Anti-Atlas lie close to the inflection point of the folded Jbel
Bani series. The dip at the inflection point is one of the most
important pieces of information for the reconstruction of a
fold, especially when considering variations of fold vergence. Outcrop conditions allow for a straightforward
construction of fold closures below and above topography.
3.3. Shortening Estimates Based on Folding
[21] We constructed first a structure contour map of the
top or6c surface at the 1:50,000 scale, using the most recent
geologic and topographic maps available, with an equidistance of 25 m. Additional information includes our field
observations, bedding dip measurements, and interpreted
aerial photographs.
[22] Because of the abundance and the continuity of the
outcrops we were able to reconstruct most of the fold limbs
and many fold closures quickly and with high accuracy. In a
second step, the structure contour lines for the eroded and
subsurface parts of the folds were added using auxiliary
cross sections and classical techniques of down-plunge
projection along the measured fold axis direction [Ramsay
and Huber, 1987, p. 356ff ]. Since the stratigraphic thicknesses of all the Ordovician formations are very well
known, and not internally deformed to any measurable
degree, we were able to reconstruct the hidden parts of
the folds with nearly the same level of accuracy as the
outcropping limbs.
[23] The structure contour map was then transformed to a
digital 3-D surface with standard digitizing procedures. For
most of our further analyses, we used this 3-D surface in a
TIN format, in which a network of small flat, but irregular
triangles approximates the real surface. In a TIN, the size of
the triangles is variable and depends, in our case, on the
spacing of the structure contour lines. A TIN is ideal for
representing surfaces with features at variable scales since
even smaller structures are adequately represented. The
error stemming from replacing a curve by straight line
segments (or in 3-D: a curved surface replaced by
interconnected triangles) is negligible. In our case, even
smaller folds consist of numerous triangles.
[24] Using a common geographic information systems
software package (ArcInfo, ArcView) allowed calculation
of statistics for many of the relevant fold parameters such as
amplitude, wavelength, symmetry, vergence, and variations
in the direction of the fold axis. Similarly, the extraction of a
section at a random point in the model and the calculation of
the curves length is a more or less automatic procedure. In
TC4006
this way, we calculated the map-scale shortening for a

virtually unlimited number of cross sections.
[25] In a second phase, sections were extracted from the
model and enlarged for those locations where outcrop-scale
second-order folds had been observed in the field. These
structures were then added by hand in order to estimate
their relative contribution to overall shortening.
[26] The assessment of outcrop-scale and microscale
shortening is mainly based on field observations. The
fieldwork had essentially two goals: (1) an assessment of
the outcrop-scale structures and the estimation of their
frequency and distribution and (2) the search for macroscopic indicators of microscale deformation, their description, and the estimation of their frequency and relative
importance.
[27] Outcrop-scale deformation features include meterscale minor folds and, more frequently, minor faults with
displacements of the order of centimeters to decimeters. We
could hardly find any indicators for microscale deformation
features within the quartzites. Axial planes of folds in the
Jbel Bani quartzites are materialized through the development of systematic sets of joints [Guiton et al., 2003] rather
than a cleavage, let alone a schistosity. Even shaly interlayers still mostly display primary bedding and the occurrence of LPS feature in the form of a weakly developed
pencil cleavage or other tectonic cleavage is rather exceptional. Generally, deformation intensity increases downward
in the stratigraphic pile, and the development of cleavage is
more frequently observed near the basement inliers. Higher
up in the stratigraphy, within limestones of the Devonian
Jbel Rich and the Carboniferous Jbel Ouarkziz, we identified layer parallel stylolite peaks as clear indicators of early,
layer parallel shortening in a SSE-NNW direction, at a high
angle to the strike of this monocline.
4. Results
4.1. Map-Scale Shortening
[28] The top of the Ordovician or6c horizon has been
reconstructed for the entire western part of the Jbel Bani
(Figures 5 and 6). Our structure model covers an area of
approximately 1600 km2. It nicely illustrates the low
complexity of this folded structure. Because of favorable
outcrop conditions even some of the larger outcrop-scale
structures are resolved, especially in the southern quadrangle (see Figure 7 for the use of quadrangles and parts).
Clearly, folding is the predominant deformation style at the
map scale. A few minor tear faults appear in the northern
quadrangle, but they are local phenomena, normally limited
to one fold limb.
[29] Throughout the area the folds show a high lateral
continuity and cylindricity. Individual fold hinges are near
horizontal over several kilometers. Plunging fold hinges are
generally found in pairs of en echelon lateral transitions.
The overall orientation of fold axes is very constant at
approximately N038.
[30] Fold limbs near the inflection point have dips
between 35 and 75 (Figure 8). Most folds, irrespective
of wavelength and amplitude are symmetrical and upright.
7 of 17
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
TC4006
Figure 7. The location of the different areas mentioned in

the text: N, northern quadrangle; E, eastern quadrangle; S,
southern quadrangle; W, western quadrangle.
The differences of dip of corresponding fold limbs (at the

inflexion point) is normally less than 5. At the map scale
there are no overturned limbs and there is no clear fold
vergence.
[ 31 ] The wavelengths and amplitudes of the folds
decrease significantly from north to south. In the northern
quadrangle amplitudes of up to 1100 m are attained (northernmost anticline; Adrar Tailoukst in Figure 6), but
decrease toward the eastern quadrangle to typical values
of 450 to 550 m. The wavelengths for the northern quadrangle varies between 3.5 km and 4 km. Especially in the
northern high-amplitude quadrangle the limb dips at the
inflexion point are up to 60. Fold limbs are rather straight
and the deformation is concentrated to relatively narrow,
angular hinge zones. The folds tops are flat and large. In
this quadrangle the folds are very near to the box fold endmember of detachment folds [Namson, 1981].
TC4006
[32] The eastern and southern quadrangles are characterized by a higher variability in wavelength and amplitude.
The larger and more continuous folds have amplitudes of
200 to 300 m and wavelengths of 1.5 to 2 km. The
amplitudes of the smaller folds vary strongly; their wavelengths are on the order of 500 m or less. Some of these
folds are at the transition to outcrop-scale structures. In the
southern quadrangle the fold hinges become more rounded.
The limb dips vary between 35 and 45 but they rarely
exceed 40. Only the limb dips of folds at the transition to
outcrop scale attain high values up to 60.
[33] A striking feature of the 3-D model of the Ordovician
marker bed is the complete absence of thrust faults and/or
duplex structures. This absence is real and not just an
interpretation. Outcrop quality over large distances is
excellent and allows ruling out any hidden, unresolved
structures at this level. The low complexity of the structures
at the map scale, essentially a continuous surface without
thrusts or faults, allows for a straightforward calculation of
the shortening. Sections at the borders of the model that
contain gaps (because of holes of missing data) were
ignored. The shortening estimates obtained from unfolding
the structure contour map in different parts of the model are
shown in Figure 9.
[34] The calculated shortenings correspond to the mapscale portion of the total shortening. In the southern
quadrangle some structures at the transition to outcrop
scale are integrated. The shortening shows considerable
variability. It varies between less than 5% and more than
30%. However, if the model is subdivided into a northern
high-amplitude domain (corresponding to the northern
quadrangle) and a low-amplitude domain (eastern
and southern quadrangles), the shortening within each of
these domains is rather homogeneous. The northern quadrangle displays shortening between 12 and 20% with an
average of 16.3% (Figure 9a). The eastern and the southern
quadrangle together show shortening between 4% and 14%
with an average of 8.7% (Figure 9b). Accordingly, the
(long) sections through the whole northeastern part show
shortening values between the two mentioned above. They
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
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
TC4006
vary between 10% and 20% with an average of 13.2%

(Figure 9c).
4.2. Outcrop-Scale Shortening
[35] Field observations and the examination of aerial
photographs allowed checking different structural positions
for outcrop-scale structures. Also, very continuous outcrops, such as the Jbel Oumloulles Anticline west of Assa
(Figure 6, southern quadrangle) allowed checking for
outcrop-scale structures at several structural positions of
the same fold.
[36] The most common outcrop-scale structures are wide
flexures of the or6c horizon. Typically, they are characterized by very low amplitude to wavelength ratio. Shortening
associated with these structures is minimal, on the order of a
few meters in the case of individual, outcrop-scale structures. Their spacing is on the order of hundreds of meters to
kilometers. More complex outcrop-scale structures such as
tight minor folds and reverse faults are very rare. Such
structures could add another few tens of meters of shortening per kilometer of cross section at the map scale. Together,
outcrop-scale shortening is estimated to account for less
than 5% total shortening at the map scale.
TC4006
those observed in the Appalachian Tuscarora Sandstone.

In one site near Targoumait (Figure 5), we measured a
grain shape preferred orientation of the detrital quartz grains
of Rs = 1.15 within the bedding plane. This shortening is
oriented NW-SE at a high angle to the local fold axis and it
could well be interpreted as due to an early, prefolding
shortening of up to 13%, if it was accommodated by pure
volume loss. In any case, LPS shortening in quartz
arenites seems to be much more restricted than the highly
variable values reported from shales and other clay-rich
lithologies [Kisch, 1991].
[38] In summary, an absolute minimum horizontal shortening for the studied area, based on the map-scale balancing
of folds (13%) with the addition of a conservative estimate
of outcrop-scale shortening (2%), is 15%. It is more difficult
to give an estimate for the most likely total tectonic
shortening. It could be as large 30%, twice the minimum
shortening, if microscale deformation was of the same
magnitude as reported from comparable Appalachian
Tuscaroro Sandstone. A more conservative estimate, based
on our preliminary thin section observations, is in the range
15 to 25%.
4.3. Microscale Shortening
5. Extent of the Folding
[37] The or6c horizon exhibits no metamorphic overprint.

Illite cristallinity measurement in the Ktaoua clays below
and the Silurian shales above indicate that the horizon is
located near the transition from diagenetic to anchizone
conditions (U. Helg et al., Illite cristallinity patterns on the
southeastern flank of the Variscan Anti-Atlas, Morocco,
submitted to Terra Nova, 2004; hereinafter referred to as
Helg et al., submitted manuscript, 2004) estimated at 200C
[Frey, 1987]. The or6c quartz arenites lack any cleavage.
Rare stylolites are clearly of diagenetic origin, with peaks at
high angle to bedding and no tectonic stylolites have been
found within the Jbel Bani quartzites despite an intense
search for such structures. Synfolding deformation is mostly
by brittle fracturing and veining, most prominent in fold
hinges, where this deformation is leading to some layerparallel stretching, rather than shortening [Guiton et al.,
2003]. Prefolding LPS fabrics have not been identified. If
anything, there might be a hidden component of layer
parallel shortening in the rearrangement of the quartz grains
on the truly microscopic scale [Engelder and Marshak,
1985]. Studies in the Appalachian Tuscarora Sandstone, a
framework-supported quartz arenite, have shown that significant amounts of up to 22% horizontal shortening may be
accommodated by such subtle microscale deformations
[Onasch, 1993]. A similar study was conducted in the
Tuscaroro Sandstone of the western margin of the Appalachian foreland near Keyser, West Virginia [Harrison and
Onasch, 2000], where pressure solution together with
micro fracturing and crystal plastic strain results in 10%
layer parallel shortening, 10% less than the LPS values
previously reported by Onasch [1993] from a more internal
position within the Appalachian chain. Our own preliminary
thin section observations made within the or6c quartzite
reveal a very similar style of microscale deformations as
[39] In northwestern parts of the Anti-Atlas belt, Ordovician quartzites have been removed by erosion leaving
only small isolated remnants, while southward, they disappear below younger Devonian and Carboniferous strata.
The question therefore arises to what extent shortening
estimates based on the tightly folded Jbel Bani can be
extrapolated to the rest of the Anti-Atlas mountain belt.
[40] In contrast to many previous authors, we assume that
similarly folded Ordovician sediments were present above
at least the western basement inliers. In our interpretation,
the present-day Jbel Bani is but a small erosional remnant of
a much larger Ordovician sedimentary basin, preserved only
in structural depressions in between basement uplifts. The
present-day outline of the folded belt is by no means
coincident with the original extent of the folded belt. There
is no change in folding style or shortening from the northern
quadrangle toward the eroded eastern quadrangle. Similarly,
there is no significant decrease in shortening from the
eastern toward the southern quadrangle, toward the outline
of the Jbel Bani. This is also corroborated by our field
observations outside of the study area: most of the mapped
fold terminations along the outline of the Jbel Bani are
erosional and markedly different, in terms of cylindricity,
geometry, fold axis plunge, from real, conical en echelon
fold terminations occasionally observed within the fold belt.
Remnants of folded Ordovician series are also found outside
of the Jbel Bani: The Middle Cambrian Tabanit sandstones,
which are structurally coupled to the overlying Ordovician,
form an open syncline at the Jbel Tayyert mountain which is
the lowermost, not yet eroded part of a large, map-scale
fold.
[41] Another indication for a much larger extent of the
folding is the Jbel Ouarzemine mountain (JO on Figure 1b),
which consists of a map-scale fold in Ordovician rocks. It is
11 of 17
TC4006
a nearly isoclinal east vergent overturned syncline, located

approximately 60 km north of the Jbel Bani. Its axis trends
N021, which is in perfect continuity to the northernmost
Bani folds. On its western side it is bounded by a large
thrust. Obviously, this structure accommodates more shortening than the open upright folds of the northernmost Jbel
Bani, and we conclude that the observed trend of increasing
shortening from southeast to northwest continues northwestward outside of our study area [Belfoul et al., 2002].
[42] Folding does not stop at the southeastern border
of the Jbel Bani either. A general south-southwestward
decrease in deformation intensity is expected, however,
and we take this into account in our conservative cross
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
Draa plain comes from the Adrar Zouggar and Addana
mountains [Desthieux, 1977] (AZ in Figure 1b). They
represent the most external Ordovician outcrops of the
Anti-Atlas chain. Both of these Anticlinoria are basement
cored domes as documented by drilling in the case of Adrar
Zouggar [Michard and Sougy, 1977]. Unfortunately, this
area is inaccessible for field studies at present. Detailed
cross sections of the Addana and Adrar Zouggar mountains
[Desthieux, 1977] document folding of the same style,
wavelength, and intensity as our own observations at
the southern border of the Jbel Bani.
[43] We conclude that the folded Ordovician strata initially occupied a much larger area than seen today. Folding
of the Jbel Bani is not restricted to structural lows where it is
preserved from erosion, but covered also large parts above
the present day basement inliers. However, as seen in our
model and on the basis of additional information from
isolated further outcrops, both amplitude and wavelength
generally decrease from northwest to southeast. This
decrease is most probably related to regional-scale facies
changes. The thickness of the Middle Cambrian shales, the
decollement horizon, also decreases toward the southeast
[Boudda et al., 1979; Benssaou and Hamoumi, 2003],
thereby progressively losing its ability as an efficient
detachment horizon. Toward the southeastern front, the
spacing between individual folds becomes more pronounced. This is seen as a consequence of the combined
effect of a decrease in layer thickness and an increase in the
ratio of competent versus incompetent beds.
[ 44 ] Another important unknown in terms of total
tectonic shortening at the orogen scale is the southwestern
border of the Anti-Atlas, the so-called Jbel Ouarkziz (OM
in Figure 1b). There is strong geomorphic evidence for this
monocline being the surface expression of a major triangle
structure. The Ouarkziz mountain chain forms a straight
ridge with altitudes of up to 500 m (200 300 m above the
valley floors) over more than 300 km length, a landmark
feature often photographed from space (images available at
http://earthobservatory.nasa.gov/Newsroom/NewImages/
images.php3?img_id=7285). This ridge is also used over a
large distance as the natural boundary between Morocco,
Algeria, and the western Sahara, respectively. There is only
one road cut (Assa-Zag) at present, where the Ouarkziz chain
TC4006
is accessible for in situ study. In structural terms, it is

essentially a large continuous monocline, dipping at 15
to 30 south-southeastward, forelandward into the Tindouf
basin. This monocline stands in stark contrast to the tightly
folded Devonian Jbel Rich underneath, from which it is
separated by the lowland of the Draa plain, running along a
thick series of Upper Devonian-lower Carboniferous shales
and siltstones. The Ouarkziz monocline bears all the key
features of a typical triangle structure [Banks and Warburton,
1986]. Unfortunately, in the absence of seismic lines, we have
no means to determine the additional shortening accommodated by underthrusting below this structure other than
minimum estimates based on total structural relief.
6. Tectonic Implications and Discussion

6.1. Crustal-Scale Balancing
[45] Having gained some new information about the
shortening in the sedimentary cover of the western AntiAtlas, some further implications for the basement are now
considered. For the illustration of our concepts a section
has been chosen where the interaction between basement
and sedimentary cover can be best demonstrated. This
section runs across the central part the Jbel Bani (see
Figure 1b), where our 3-D model has been constructed.
We will concentrate first on the tectonic implications,
which come from the minimal shortening recorded in the
sedimentary cover. A conservative section has been constructed. For the folded Ordovician only map-scale and
outcrop-scale shortening were considered. Northwest of
the outcropping Jbel Bani folds, now eroded from above
the basement high, the structural style is assumed to be the
same geometry as within the 3-D model, i.e., open, upright
folds with steep limbs. In order to account for an observed
increase in thickness of the Ordovician toward the northwest [Destombes et al., 1985], folds were drawn with
a slightly increasing amplitude and wavelength in this
direction.
[46] Southeast of the model, the Bani folds are depicted
as dying out rapidly below the Draa plain. The Jbel
Ouarkziz triangle structure is interpreted as stemming from
one single thrust stepping up south-southeastward from the
Middle Cambrian shales to Upper Devonian shales before
ending blindly somewhere below the Tindouf basin in a tip
point, from which a postulated backthrust/decollement horizon must lead back up to the Earths surface, most likely
into the lower Draa valley. We hold this backthrust responsible for the marked disharmony between the folded Jbel
Rich below and the nonfolded Jbel Ouarkziz monocline
above.
[47] The top of the lower Cambrian has been reconstructed using down-plunge projections and depth to
detachment calculations [Epard and Groshong, 1993;
Bulnes and Poblet, 1999]. In the western Anti-Atlas the
Adoudounian (Lower Cambrian) consists mainly of massive
platform carbonate series [Destombes et al., 1985; Benssaou
and Hamoumi, 2003]. Incompetent horizons are scarce and
sediments up to the Middle Cambrian are strongly coupled
12 of 17
TC4006
to the basement (compare Figures 2 and 3). A decollement

within these sediments, at least in the western Anti-Atlas, is
very unlikely and not corroborated by any structural observations. Therefore the whole pile from the basement up to
the top of the Lower Cambrian can be considered as one
structural unit, the basement-Cambrian unit in our rheologic terminology.
[48] As seen on most classical cross sections of the AntiAtlas and confirmed by our own field observations, the
folded Ordovician is clearly decoupled from this basementCambrian unit. Folding is not restricted to structural lows
between the basement inliers, however, and entire fold
trains can be projected sideways above the larger basement
domes. This also means that Bani folding at any given
location cannot be linked to one specific (assumed) thrust
fault in the basement. Shortening in the Ordovician needs to
be compared to the overall shortening on a much larger
scale, so as to include the entire basement-cover contact. In
order to balance the section at the crustal scale, an equal
amount of shortening is necessary in the sedimentary cover
and the basement. Shortening within the basement is
assumed to take place along widely spaced, mostly hidden
large-scale reverse faults. Geometric constraints will be
discussed below.
6.2. Thin Skin Interpretation
[49] With the Ordovician strongly decoupled from the
basement-Cambrian unit and a shortening increasing toward
the northwest, the question arises to which degree the
observed folding could be the result of a thin skin thrustingfolding phase, predating the formation of the basement
uplifts. The westernmost part of the Anti-Atlas, west of
the Bas Draa inlier, is known as a thin skin domain indeed
[Belfoul, 1991; Soulaimani, 1998; Pique, 2001; Belfoul et
al., 2002]. It is characterized by map-scale thrusts and
reverse faults within Lower Cambrian units. Accordingly,
shortening observed in the Ordovician Jbel Bani might be
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
Bani as very unlikely, however. Such a thin skin interpretation still requires two new postulates, which are not
substantiated by any observations. First, a regional-scale
basal decollement within Middle Cambrian shales would
have to be increasingly important westward, prior to cutting
down into the basement-Cambrian unit, in locations where
neither direct nor indirect evidence for any major basement
ramp is present. Secondly, the basement uplifts of the SW
Anti-Atlas would have to be explained by a second deformation phase, in an essentially vertical tectonics fashion,
without any sizeable amount of horizontal shortening. This
latter view of the Anti-Atlas basement uplifts does not
provide any explanation for the observed crustal-scale
thickening.
[50] The fold axis directions within the Adoudounian of
the western Anti-Atlas show considerable heterogeneity
[Soulaimani, 1998]. In many places, they parallel in a
conspicuous manner the outlines of the basement domes.
This results in places in fold interference patterns with
TC4006
nearly perpendicular fold axes [Caritg et al., 2003]. Most

authors agree that the intimate relationship of fold axes
directions and basement outlines is due to forced folding
of the sedimentary cover, accommodating differential
movements of basement blocks. It is difficult to envisage
these complicated fold pattern as the result of an early, thin
skin folding phase. Thus, at least in the Adoudounian of the
western Anti-Atlas, there exists no evidence for folding
independent of the main deformation phase, which is clearly
thick skinned, involving basement. The tectonic style of
these folds at the lowermost stratigraphic levels is very
similar to the Rocky Mountains of Wyoming [Stone, 1993,
2002].
[51] Higher up in the stratigraphy, however, within the
Ordovician Jbel Bani, fold axes directions are very regular
and increasingly independent of the underlying basement
structures. Some minor variations in fold axis directions can
still be explained as deflections around the basement inliers,
albeit strongly attenuated above the Lie de vin detachment
level. No polyphase deformation history is required to
explain these variations in fold axis trends in the western
Anti-Atlas. Note that further east, near Tata, there is
evidence for a two-phase history in the form of conspicuous
fold interference patterns [Caritg et al., 2003], but even
there, both folding phases are best explained in a thick
skinned fashion, because of reactivation of former basement
faults, in close analogy with Laramide uplifts of the Rockies
[Stone, 1993; Marshak et al., 2000; Bump, 2003].
[52] Paleostress measurements in the western Anti-Atlas
fold belt reveal that the local axes of major compressive
stress are subperpendicular to the fold axes throughout the
Jbel Bani. The paleostress field also shows some heterogeneity and it is characterized by a predominant local component. It seems therefore that the paleostress field is also
conditioned by the basement blocks rather than being the
result of a distant push. In conclusion, an early thin skin
phase is regarded as a very improbable mechanism to
explain the folding of the Ordovician Jbel Bani.
6.3. Thick Skin Interpretation
[53] A rapid glance at the tectonic map of the SW AntiAtlas reveals the existence of thick skin tectonics, necessary to explain the large domes of Proterozoic-Cambrian
[Rodgers, 1995]. The central question is not the existence,
but the nature and origin of these basement domes, as well as
their relationship with the buckle folding observed in the
cover series around and above them. Direct evidence for
large-scale thrusts within the basement-Cambrian unit is
elusive in the western Anti-Atlas. This lack of direct evidence should not be used as an argument against the
existence of thrusts or reverse faults at depth, however.
There is a series of explanations why such thrusts have not
been mapped, nor postulated. The most important reason is
the present-day erosion level, revealing only a very small,
topmost 2 km of more than 10 km of structural relief of these
basement uplifts. This estimate is based on the known depth
to basement in the distant foreland of the Tindouf basin,
where it is located at approximately 8 km [Michard, 1976;
Bertrand-Sarfati et al., 1991]. Similarly, the preshortening
13 of 17
TC4006
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
must have been at some 10 km below sea level in order to
accommodate the total thickness of Paleozoic sediments,
which comprise shallow marine series up to the middle
Carboniferous (Visean).
[54] Another reason why no large-scale thrusts were
found in the SW Anti-Atlas is the presence of widespread
thick Middle Cambrian shales. As depicted in Figure 10, we
assume a tectonic style where thrust ramps, breaking
through the Lower Cambrian Adoudounian, turn to horizontal flats within incompetent Middle Cambrian shales.
Minor thrusts may level out even deeper in the stratigraphy,
within thin evaporitic shales of the Adoudounian unit.
[55 ] An important reverse fault or thrust has been
described in the westernmost Anti-Atlas, at the southern
border of the Lower Draa inlier, where Proterozoic basement is thrust toward the SE onto the sedimentary cover
[Soulaimani, 1998]. Here, intense deformation under lower

greenschist facies lead to the development of a foliation
with steeply plunging stretching lineations and a marked
internal deformation of rhyolite pebbles of the PIII; all these
deformation features are compatible with a south-southeastward thrust.
[56] At the scale of the entire Anti-Atlas, crustal-scale
balancing considerations are used to constrain permissible
basement geometries at depth. One of the key features of the
Anti-Atlas is its high structural relief. A conservative
estimate of the total thickness of the Paleozoic cover series
of the SW Anti-Atlas yields 10 km thickness. Most of this
thickness is preserved today in first-order synclinoria, in
between the basement domes. However, there is good
evidence for a similar amount of erosion from atop the
westernmost basement inliers (Bas Draa, Ifni, Kerdous),
which were invariably at lower greenschist facies metamor-
14 of 17
TC4006
phic conditions during late Variscan deformation. This

metamorphism is obvious from the deformation style at
the brittle/ductile transition and from our own illite crystallinity measurements [Buggisch, 1988a; Burkhard et al.,
2001; Helg et al., submitted manuscript, 2004]. In contrast
to previous authors, who mostly considered these basement
domes as all-time horsts, with a reduced thickness of
total Paleozoic cover, we assume an essentially continuous
thickness of sedimentary series across these basement
domes. We therefore assume a similar structural relief to
exist at the basement cover contact, where it is documented,
and at the Devonian/Carboniferous boundary, where it is
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
reactivation in relation with High-Atlas inversion tectonics
[Beauchamp et al., 1999; Frizon-de-Lamotte et al., 2000].
The Anti-Atlas is also the location of Pliocene volcanic
activity, e.g., in the Jbel Siroua, and broad thermal uplift at a
regional scale has to be taken into consideration. Accordingly, the removal of more than 10 km of Paleozoic sediments from above the basement inliers may not be solely the
result of late Variscan tectonics but the combined effect of
late Variscan tectonics, Neogene Atlas inversion, and thermal hot spot related uplifts and concomitant erosion. The
total amount of non-Variscan uplift/erosion in the Anti-Atlas
is ill constrained. Upper Cretaceous sediments of the
Hamada clearly seal Variscan folds; their present-day
elevation is about 400 m above sea level in the eastern AntiAtlas but close to sea level near the Atlantic coast. In
between, this same marker horizon is completely eroded
from above the SW Anti-Atlas, leaving some room for
speculation. Thermal uplift, related to the volcanic activity
of the Jbel Sirhoua, can only be estimated from general
considerations. Surface uplift related to plates moving over
hot spots may reach 1 to 3 km, affecting large areas of up to
2000 km across [Coward, 1994; Eisbacher, 1996].
[58] Nevertheless, even with the assumption of up to 3 km
of Neogene surface uplift and High Atlas reactivation, the
leveling of more than 7 km of structural relief remains to be
explained by late Variscan tectonics with concomitant uplift
and erosion. This same amount of structural relief is not
only attained at the scale of the orogen but also at the scale
of individual basement domes. Between the Jbel Bani and
the surrounding basement inliers Kerdous, Lower Draa, and
Ifni (Figure 2), the structural relief is in excess of 7 km.
Given the modest amount of cover shortening observed in
the Anti-Atlas fold belt, this structural relief is difficult to
explain by an antiformal stack of thin basement slabs above a
relatively shallow crustal detachment [Boyer and Elliott,
1982]. As an example, if we assume 5 km thick basement
slabs overlain by a 10 km thick sedimentary series, a
horizontal shortening of roughly 95 km would be necessary
to create the observed cross sectional excess area of 475 km2
(Figure 10b, shallow detachment diagram labeled b2).
This exceeds by far our estimations of the horizontal
shortening observed within the sedimentary cover. The
discrepancy is too large to be explained by ill evaluated,
underestimated outcrop-scale or microscale shortening.
TC4006
Even if we assume a total submap-scale shortening of the

same magnitude as map-scale shortening, the calculated
depth to detachment remains at about 18 km (Figure 10b,
maximum shortening diagram labeled b1). Considering
only the minimal shortening (the total observed shortening),
the detachment depth comes to lie at about 26 km depth.
[59] As an alternative to the antiformal stack of thin
basement slabs we postulate basement inversion of thick
block-like basement parts with the dimension of the present
day boutonnieres, uplifted along steep reverse faults. The
depth of the detachment horizon again determines the
thickness of such basement blocks.
6.4. Inversion Tectonics
[60] The term inversion is used here in the sense of
Coward [1994] to describe regions which have experienced
a reversal in uplift or subsidence. The Anti-Atlas nicely fits
this definition; after an extended period of subsidence
during the Paleozoic, the Anti-Atlas basin is the locus of
tectonic shortening, uplift, and erosion during the later
stages of the Variscan orogeny, possibly with some reactivation in Neogene times [Frizon-de-Lamotte et al., 2000].
[61] Most authors agree that Variscan basements tectonics
of the Anti-Atlas is dominated by the reactivation of former
weakness zones. In earlier publications these weakness
zones are invariably considered to be subvertical [Michard,
1976]. Sedimentological evidence for an important postPanafrican extensional phase has accumulated during the last
two decades [Soulaimani et al., 2001; Benssaou and
Hamoumi, 2003]. Decametric fault-bounded graben structures are documented within the Proterozoic basement in the
westernmost Anti-Atlas [Soulaimani, 1998]. Regional
changes in thickness and facies of Cambrian series
have been used to map the location and general orientation
of graben structures [Benssaou and Hamoumi, 2003]
(Figure 11).
[62] In the light of this post Panafrican rifting phase a
Variscan inversion of former PIII normal faults seems the
most likely explanation for the mostly still hidden reverse
faults bounding the present day basement inliers. Rather
than all-time horsts [Michard, 1976; Pique, 1994; Benssaou
and Hamoumi, 2003], basement inliers in our interpretation
represent the former basin floors. The fault reactivation
nicely explains the strong basement involvement and the
high basement topography in an otherwise only mildly
deformed tectonic setting.
[63] Tectonic inversion is, at least partly, corroborated by
the thickness distribution of the rift/early postrift sediments
[Boudda et al., 1979]. The thickness of the lowermost Adoudounian unit, also called Lower Limestones
(Figure 11), is highest in the area of the basement domes of
the western Anti-Atlas, decreasing both east and southward,
forelandward. In other words, the location of maximum
subsidence by rifting corresponds roughly with the area of
maximum uplift by late Variscan tectonic inversion.
[64] The widespread occurrence of PIII rift sediments
found along the borders of almost all boutonnieres is an
additional indication for these to be former grabens, rather
15 of 17
TC4006
Figure 11. Thickness distribution of the lowermost

sedimentary unit of the Paleozoic Anti-Atlas basin according to Boudda et al. [1979]. The Bas Draa-Bou IzakarneFouanou graben structure of Middle Cambrian age is shown
according to a recent detailed stratigraphic analysis of
Benssaou and Hamoumi [2003, Figure 2].
than horsts. We have to admit, however, that there is no
observational control, nor is there any subsurface data
available about the regional-scale variations of PIII series
for the large synclinal areas in between the present day
basement domes. According to our inversion hypothesis,
there should be only a very limited amount of PIII series in
these areas of the former, Late Proterozoic horsts.
7. Conclusions
[65] The Anti-Atlas of Morocco represents a special type
of foreland fold belt with a striking absence of observable
thrusts. Cover shortening is accommodated by tight detachment folding of competent quartzite and carbonate layers,
but no ramping or duplexing has ever been observed. This
tectonic style is dictated by the mechanical stratigraphy of
the Paleozoic cover series, dominated by thick intervals of
TC4006
shale and siltstones. Four competent units are distinguished

(Figure 3), from bottom to top: (1) basement-Lower Cambrian, (2) Jbel Bani (Ordovician), (3) Jbel Rich (Devonian),
and (4) Jbel Ouarkziz (lower Carboniferous). The basementLower Cambrian unit (1) is involved in massive inversion of
Late Proterozoic normal faults, leading to Laramide style
basement domes of several tens of kilometers wavelength,
with amplitudes or structural relief on the order of 7 to
10 km: the first-order basement folds or so-called plis
de fond [Argand, 1924]. A more than 1 km thick series of
Middle Cambrian shales allows for a complete decoupling
and disharmony between this basement unit (1) and the
overlying Paleozoic cover series (2 and 3), which react
essentially by polyharmonic multilayer folding [Ramsay
and Huber, 1987]. Larger wavelengths (1 km) and amplitudes (500 m) are observed within folds of the thicker
Ordovician sandstones of the Jebel Bani than within the
Devonian Jbel Rich folds (500 m, 200 m). Carboniferous
series (4) are only preserved at the very mountain front,
where they are involved in a triangle structure. The total
width of the SW Anti-Atlas foreland belt is about 150 km
and basement domes occur at less than 30 km behind the
mountain front. A minimum total amount of 30 km (17%)
of cover shortening is obtained from restoration of the Jbel
Bani fold train. Balancing considerations require a middle to
lower crustal detachment at 18 to 25 km depth for the
basement inversion.
[66] Acknowledgments. We wish to thank A. Soulaimani for introducing us to the geology of the Anti-Atlas, his hospitality, and for many
inspiring discussions. Ongoing support by the Moroccan Ministe`re de
lEnergie et des Mines, Direction de la Geologie and the state petroleum
agency ONAREP is gratefully acknowledged. We wish to thank personally M. Dahmani, A. M. Morabet, and M. Zizi for their help. We are
grateful to Bert Bally, Andre Michard, Alain Pique, Jean-Paul Schaer, and
Dieter Schroeder for encouragement and for sharing their ideas and data
about the Anti-Atlas of Morocco. Financial support by the Swiss National
Science Foundation, grants 21-52516.7 and 20-63790.00 is gratefully
acknowledged.
References
Argand, E. (1924), La tectonique de lAsie, Proc. Int.
Geol. Congr., 13th, 171 372.
Azizi-Samir, M. R., J. Ferrandini, and J. L. Tane
(1990), Late Pan-African (580 560 M.A.) tectonic
activity and volcanism in the Anti-Atlas Mountains
of Morocco: Geodynamic interpretation on the
scale of north-west Africa, J. Afr. Earth Sci., 10,
549 563.
Banks, C. J., and J. Warburton (1986), Passive-roof
duplex geometry in the frontal structures of the
Kirthar and Sulaiman mountain belts, Pakistan,
J. Struct. Geol., 8, 229 237.
Beauchamp, W., R. W. Allmendinger, M. Barazangi,
A. Demnati, M. Alji, and M. Dahmani (1999),
Inversion tectonics and the evolution of the High
Atlas Mountains, Morocco, based on a geologicalgeophysical transect, Tectonics, 18, 163 184.
Belfoul, M. (1991), La structuration hercynienne de la
re gion de Labiar-Oued Boussafene (Guelmim,
Anti-Atlas sud-occidental), the`se 3e`me cycle thesis,
162 pp., Univ. de Rabat, Rabat, Morocco.
Belfoul, M. A., F. Faik, and B. Hassenforder (2002),
Evidence of a tangential tectonic event prior to the
major folding in the Variscan belt of the western
Anti-Atlas, Morocco, J. Afr. Earth Sci., 32, 723
739.
Benssaou, M., and N. Hamoumi (2003), The LowerCambrian western Anti-Atlasic graben: Tectonic
control of paleogeography and sequential organisation, C. R. Geosci., 335, 297 305.
Bertrand-Sarfati, J. (1981), Proble`me de la limite
Pre cambrien-Cambrien: La section de Tiout
(Maroc): Les stromatolithes et leur biostratigraphie,
Newsl. Stratigr., 10, 20 26.
Bertrand-Sarfati, J., A. Moussine-Pouchkine,
P. Affaton, R. Trompette, and Y. Bellion (1991),
Cover sequences of the West African Craton, in
The West African Orogens and Circum Atlantic
Correlatives, edited by R. D. Dallmeyer and J. P.
Lecorche, pp. 65 82, Springer-Verlag, New
York.
Bonhomme, M., and B. Hassenforder (1985), Le
me tamorphisme hercynien dans les formations
tardi-et post-panafricaines de lAnti-Atlas occidental (Maroc): Donne es isotopiques Rb/Sr et
K/Ar des fractions fines, Sci. Geol. Bull., 38,
175 183.
Boudda, A., G. Choubert, and A. Faure-Muret
(1979), Essai de stratigraphie de la couverture
se dimentaire de lAnti-Atlas: Adoudounien Cambrien Inferieur, Notes Mem. Serv. Geol. Maroc,
271, 96.
16 of 17
Boyer, S. E., and D. Elliott (1982), Thrust systems,

AAPG Bull, 66, 1196 1230.
Buggisch, W. (1988a), Diagenesis and very low-grade
metamorphism of the Lower Cambrian rocks in the
Anti-Atlas, Morocco, in The Atlas System of
Morocco: Studies on Its Geodynamic Evolution,
Lecture Notes in Earth Sci., vol. 15, edited by
Volker H. Jacobshagen, pp. 123 128, SpringerVerlag, New York.
Buggisch, W. (1988b), The Precambrian-Cambrian
boundary in the Anti-Atlas (Morocco), in The Atlas
System of Morocco: Studies on Its Geodynamic
Evolution, Lecture Notes in Earth Sci., vol. 15, edited by Volker H. Jacobshagen, pp. 81 90, Springer
Verlag, New York.
Bulnes, M., and J. Poblet (1999), Estimating the detachment depth in cross sections involving detachment
folds, Geol. Mag., 136, 395 412.
Bump, A. P. (2003), Reactivation, trishear modeling,
and folded basement in Laramide uplifts: Implications for the origins of intra-continental faults, GSA
Today, 13(3), 4 10.
Burkhard, M., S. Caritg, U. Helg, and C. RobertCharrue (2001), Forced, disharmonic multilayer
buckle folding in the late Variscan Antiatlas of
Morocco, paper presented at American Association
TC4006
of Petroleum Geologists 2001 Annual Meeting,

Tulsa, Okla.
Caritg, S., M. Burkhard, R. Ducommmun, U. Helg,
L. Kopp, and C. Sue (2003), Fold interference patterns in the late Paleozoic Anti-Atlas of Morocco,
Terra Nova, 16, 27 37.
Choubert, G. (1963), Histoire geologique du precambrien de lAnti-Atlas, Notes Mem. Serv. Geol.
Maroc, 162, 352.
Choubert, G., and A. Faure Muret (1971), Anti-Atlas
(Maroc), in Tectonique de lAfrique (Tectonics of
Africa), Earth Sci. Ser., vol. 6, pp. 163 175, UNESCO, Paris.
Coward, M. (1994), Inversion Tectonics, in Continental
Deformation, edited by P. L. Hancock, pp. 289
304, Pergamon, New York.
Desthieux, F. (1977), Etude tectonique et metallogenique du Jbel Addana, Ordovicien des plaines du
Dra. Maroc presaharien, Notes Serv. Geol. Maroc,
268, 209 236.
Destombes, J., H. Hollard, and S. Willefert (1985),
Lower Paleozoic rocks of Morocco, in Lower
Palaeozoic of North-Western and West-Central
Africa, vol. 4, edited by C. H. Holland,
pp. 91 336, Trinity Coll., Dep. of Geol., Dublin,
Ireland.
Donzeau, M. (1974), LArc Anti-Atlas-Ougarta (Sahara
nord-occidental, Algerie-Maroc), C. R. Acad. Sci.,
Ser. II, 278, 417 420.
Eisbacher, G. H. (1996), Einfuhrung in die Tektonik,
374 pp., F. Enke, Stuttgart, Germany.
Engelder, T., and S. Marshak (1985), Disjunctive cleavage formed at shallow depths in sedimentary
rocks, J. Struct. Geol., 7, 327 343.
Epard, J. L., and R. H. J. Groshong (1993), Excess area
and depth to detachment, AAPG Bull., 77, 1291
1302.
Faik, F., M. A. Belfoul, M. Bouabdelli, and
B. Hassenforder (2002), The structures of the Late
Neoproterozoic and Early Palaeozoic cover of the
Tata area, western Anti-Atlas, Morocco: Polyphased deformation or basement/cover interactions
during the Variscan orogeny?, J. Afr. Earth Sci., 32,
765 776.
Frey, M. (1987), Low Temperature Metamorphism,
Chapman and Hall, New York.
Frizon-de-Lamotte, D., B. Saint-Bezar, R. Bracene, and
E. Mercier (2000), The two main steps of the Atlas
building and geodynamics of the western Mediterranean, Tectonics, 19, 740 761.
Geyer, G., and E. Landing (1995), MOROCCO 95
The Lower-Middle Cambrian Standard of Western
Gondwana, 269 pp., Inst. fur Palaontol., Univ.
Wurzburg, Wurzburg, Germany.
Guerraoui, A., M. Bensaid, and M. Dahmani (1997),
Carte geologique du Maroc 1:100000, Fask, Notes
Serv. Geol. Maroc.
Guiton, M. L. E., W. Sassi, Y. M. Leroy, and B. D. M.
Gauthier (2003), Mechanical constraints on the
chronology of fracture activation in folded Devonian sandstone of the western Moroccan Anti-Atlas,
J. Struct. Geol., 25, 1317 1330.
Harrison, M. J., and C. M. Onasch (2000), Quantitative
assessment of low-temperature deformation
mechanisms in a folded quartz arenite, Valley and
Ridge Province, West Virginia, Tectonophysics,
317, 73 91.
Hassenforder, B. (1987), La tectonique panafricaine et
varisque de lAnti Atlas dans le Massif du Kerdous
(Maroc), The`se Sci. thesis, 249 pp., Univ. L. Pasteur Strasbourg, Strasbourg, France.
Hogan, J. P., and W. M. Dunne (2001), Calculation of
shortening due to outcrop-scale deformation and its
relation to regional deformation patterns, J. Struct.
Geol., 23, 1507 1529.
Homberg, C. (1996), Analyse des deformations cassantes dans le Jura et modelisation numerique des
perturbations des contraintes tectoniques autour
daccidents majeurs, Ph.D. thesis, 282 pp., Univ.
Pierre et Marie Curie, Paris.
Huon, S. (1985), Clivage ardoisier et rehomogenisation

isotopique K-Ar dans des schistes paleozoiques du
Maroc: Etude microstructurale et isotopique: Consequences regionales, Ph.D. thesis, 124 pp., Univ.
Louis Pasteur, Strasbourg.
Jeannette, D. (1981), Influence de structures panafricaines sur des de formations hercyniennes
(Bordure septentrionale de la boutonnie`re dIgherm,
Anti-Atlas occidental), Notes Serv. Geol. Maroc,
42, 25 32.
Kisch, H. J. (1991), Development of slaty cleavage and
degree of very-low-grade metamorphism: A review,
J. Metamorph. Geol., 9, 735 750.
Kogbe, C. A. (1998), Sedimentary Basins of Morocco,
Afr. Geosci. Rev., 5, 232.
Latham, A. (1988), The Precambrian-Cambrian transition in Morocco: Problems and perspectives,
in Trace Fossils, Small Shelly Fossils, and the
Precambrian-Cambrian Boundary, edited by
E. Landon, G. M. Narbonne, and P. M. Myrow,
Bull. N. Y. State Mus., 1076, 14 pp.
Leblanc, M. (1972), Sur le style disharmonique des plis
hercyniens, a` la base de la couverture, dans lAntiAtlas central (Maroc), C. R. Acad. Sci., Ser. D, 275,
803 806.
Leblanc, M., and J. R. Lancelot (1980), Interpretation
geodynamique du domaine pan-africain (Precambrien terminal) de lAnti-Atlas (Maroc) a partir de
donnees geologiques et geochronologiques, Can. J.
Earth Sci., 17, 142 155.
Marshak, S., K. Karlstrom, and J. M. Timmons (2000),
Inversion of Proterozoic extensional faults: An explanation for the pattern of Laramide and Ancestral
Rockies intracratonic deformation, United States,
Geology, 28, 735 738.
Mattauer, M., F. Proust, and P. Tapponnier (1972), Major strike-slip fault of late Hercynian age in Morocco, Nature, 237, 160 162.
Michard, A. (1976), Elements de geologie Marocaine,
Notes Mem. Serv. Geol. Maroc, 252, 408.
Michard, A., and J. Sougy (1977), Lorogene`se hercynienne a` la lisie`re nord-ouest de lAfrique (Structure
des chanes primaires du Maroc au Senegal), in La
Chane varisque dEurope moyenne et occidentale,
Colloq. Int. Cent. Natl. Rech. Sci., vol. 243,
pp. 605 640, Cent. Natl. de la Rech. Sci., Paris.
Mitra, G. (1994), Strain variation in thrust sheets and
across the Sevier fold-and-thrust belt (Idaho-UtahWyoming): Implications for section restoration and
wedge taper evolution, J. Struct. Geol., 16, 585 602.
Mitra, S. (1992), Balanced structural interpretations in
fold and thrust belts, in Structural Geology of Fold
and Thrust Belts, edited by S. Mitra and G. W.
Fisher, pp. 53 77, Johns Hopkins Univ. Press, Baltimore, Md.
Namson, J. (1981), Structure of the western Foothills
Belt, Miaoli-Hsinchu area Taiwan: (1) Southern
Part, Pet. Geol. Taiwan, 18, 31 51.
Onasch, C. M. (1993), Determination of pressure solution shortening in sandstones, Tectonophysics, 227,
145 159.
Pique, A. (1994), Geologie du Maroc: Les Domaines
Structuraux et Leur Evolution Structurale, 283 pp.,
Ed. PUMAG, Marrakech, Morocco.
Pique , A. (2001), Geology of Northwest Africa,
310 pp., Gebruder Borntrager, Stuttgart, Germany.
Pique, A., and A. Michard (1989), Moroccan hercynides: A synopsis: The paleozoic sedimentary and
tectonic evolution at the northern margin of West
Africa, Am. J. Sci., 289, 286 330.
Pique, A., J. J. Cornee, J. Muller, and J. Roussel (1991),
The Moroccan Hercynides, in The West African
Orogens and Circum Atlantic Correlatives, edited
by R. D. Dallmeyer and J. P. Lecorche, pp. 229
263, Springer-Verlag, New York.
Pique, A., M. Bouabdelli, and J.-R. Darboux (1995), Le
rift cambrien du Maroc occidental, C. R. Acad. Sci.,
Ser. II, 320, 1017 1024.
Pique, A., M. Bouabdelli, A. Soulaimani, N. Youbi, and
M. Iliani (1999), Upper Neoproterozoic PIII con-
17 of 17
TC4006
glomerates in the Anti-Atlas, southern Morocco:

Pan-African molasses, or indicators of upper Proterozoic rifting?, C. R. Acad. Sci., Ser. II, 328, 409
414.
Plessmann, W. (1972), Horizontal-Stylolithen im franzosisch-schweizerischen Tafel-und faltenjura und
ihre Einpassung in den regionalen Rahmen, Geol.
Rundsch., 61, 332 347.
Ramsay, J. G., and I. M. Huber (1983), The Techniques
of Modern Structural Geology, 307 pp., Academic,
San Diego, Calif.
Ramsay, J. G., and I. M. Huber (1987), The Techniques
of Modern Structural Geology, 697 pp., Academic,
San Diego, Calif.
Rodgers, J. (1990), Fold-and-thrust belts in sedimentary
rocks, part 1: Typical examples, Am. J. Sci., 290,
321 359.
Rodgers, J. (1995), Lines of basement uplifts within the
external parts of orogenic belts, Am. J. Sci., 295,
455 487.
Schenk, P. E. (1971), Southeastern Atlantic Canada,
northwestern Africa, and continental drift, Can. J.
Earth Sci., 8, 1218 1251.
Sebai, A., G. Feraud, H. Bertrand, and J. Hanes (1991),
AR-40/AR-39 dating and geochemistry of tholeiitic
magmatism related to the early opening of the central Atlantic rift, Earth Planet. Sci. Lett., 104, 455
472.
Smart, K. J., W. M. Dunne, and R. D. Krieg (1997),
Roof sequence response to emplacement of the
Wills Mountain duplex: The roles of forethrusting
and scales of deformation, J. Struct. Geol., 19,
1443 1459.
Soulaimani, A. (1998), Dynamique et interactions
Socle/Couverture dans lAnti-Atlas occidental
(Maroc): Rifting fini-proterozoque et orogene`se
hercynienne, The`se detat, Univ. Cadi Ayyad,
Marrakech, Morocco.
Soulaimani, A., C. Le Corre, and R. Farazdaq (1997),
Deformation hercynienne et relation socle/couverture dans le domaine du Bas-Draa (Anti-Atlas occidental, Maroc), J. Afr. Earth Sci., 24, 271 284.
Soulaimani, A., A. Pique, and M. Bouabdelli (2001),
La serie du PII PIII de lAnti-Atlas occidental
(Sud marocain): Un olistostrome a` la base de la
couverture post-panafricaine (PIII) du Proterozoic
superieur, Earth Planet. Sci. Lett., 332, 121 127.
Soulaimani, A., M. Bouabdelli, and A. Pique (2003),
Lextension continentale au Ne o-Prote rozoque
supe rieur-Cambrien inferieur dans lAnti-Atlas
(Maroc), Bull. Soc. Geol. Fr., 174, 83 92.
Stone, D. (1993), Basement-involved thrust-generated
folds as seismically imaged in the subsurface of the
central Rocky Mountain foreland, in Laramide
Basement Deformation in the Rocky Mountain
Foreland of the Western United States, edited by
R. Chase, E. Erlsev, and C. Schmidt, Spec. Pap.
Geol. Soc. Am., 280, 271 318.
Stone, D. (2002), Morphology of the Casper Mountain
uplift and related subsidiary structures, central
Wyoming: Implications for Laramide kinematics,
dynamics, and crustal inheritance, AAPG Bull.,
86, 1417 1440.
Villa, I. (1998), Isotopic closure, Terra Nova, 10, 42
47.
Villeneuve, M., and J. J. Cornee (1994), Structure, evolution and palaeogeography of the West African
Craton and bordering belts during the Neoproterozoic, Precambrian Res., 69, 307 326.
Weijermars, R. (1993), Estimation of paleostress
orientation within deformation zones between
two mobile plates, Geol. Soc. Am. Bull., 105,
1491 1510.
M.
Burkhard,
S. Caritg, U. Helg, and C. RobertCharrue, Institut de Geologie, Universite de Neuchatel,
Case Postale, Rue E. Argand 11, CH-2007, Neuchatel,
Switzerland. ([email protected])
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
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
TC4006
PROJET DARTICLE
Un troisime article est prvu abordant les thmes suivants :
-
Interfrence Anti-Atlas Ougarta (Figures 34 et 40), avec llaboration

dun modle tridimensionnel reprsentant lipsographie du sommet de
lOrdovicien suprieur au paroxysme varisque.
Mise en vidence de la dformation dorientation ougartienne par la

construction dune coupe ENE-WSW au sud des boutonnires.
Analyse des relations gomtriques entre lAnti-Atlas et les sdiments

post-varisques (Figure 36).
Thse de doctorat
C
RESUMES ET POSTERS
- Late Variscan Inversion Tectonics in the Paleozoic Tafilalt basin of SouthEastern Morocco. Abstract. MIOG 2002.
- Inversion tectonics in the hercynian Anti-Atlas of Morocco. Abstract. SGM 2003.
- The Anti-Atlas fold-belt of Morocco : Thick skin inversion tectonics in the
hinterland of the alleghenian orogeny. Abstract. RST 2004.
- Tectonique dinversion dans lAnti-Atlas oriental. Abstract. 3Ma 2005.
- The Anti-Atlas fold belt of Morocco : variscan inversion tectonics and
interference pattern of an intracratonic basin. Abstract. SGM 2005.
- The Anti-Atlas fold-belt of Morocco. Poster. RST 2004.
- Structural study of the eastern Anti-Atlas, the variscan belt of Morocco. Poster.
SGM 2005.
Thse de doctorat
ABSTRACTS
Abstract prsent au Marrakech International Oil & Gas 2002, session poster.
Late Variscan Inversion Tectonics in the Paleozoic Tafilalt basin

of South-Eastern Morocco
The Paleozoic Tafilalt basin in south eastern Morocco is located in the
"hinterland" of the late Variscan Appalachian-Antiatlas chain. A post-panafrican
sedimentary cycle starts with rifting in Late Proterozoic (PIII), followed by thermal
subsidence during most of the Paleozoic. An increase in subsidence rates in
Late Devonian, with an associated change from marl/carbonates to
shale/siliciclastics is interpreted in terms of a change to a foreland basin setting
in response to closure of the Japetos ocean and tectonic activity within an early
Appalachian/ Variscan chain forming in the far distance to the NW. We could not
find any evidence for syn-sedimentary tectonics within this Late Paleozoic clastic
wedge. NE-SW oriented compressional deformation is entirely post- Late
Carboniferous and lead to spectactular inversion structures seen on geologic
maps and satellite images. This compression is due to a late phase of Variscan
collision tectonics expressed here as mostly thick skin intracratonic shortening
in Wind river style. Seemingly thin skin folding takes place above thick
incompetent shale horizons particularly in the southern Tafilalt. Structural
trends of the large scale broad folds and associated faults are inherited from the
Panafrican and or Late Proterozoic (PIII). This compression is followed by
marked event of extension in a NNW-SSE direction of uncertain age (Late
Variscan Permian? or early Alpine - Liassic?). The interference between the
forced folding (inversion) and renewed extension in an oblique direction lead to a
complex map pattern of folds and faults, further complicated by a third
reactivation in a dextral strike slip regime.
Thse de doctorat
Abstracts - Posters
Abstract prsent au Swiss Geoscience Meeting 2003 Ble, confrence dans

la session du Swiss Tectonic Studies Group.
Inversion tectonics in the hercynian Anti-Atlas of Morocco
Charles Robert-Charrue, Martin Burkhard
Geological Institute, Neuchtel University, CP 2 , CH-2007 Neuchtel, [email protected]
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 " Boutonnires", but the main
part is made of an important Paleozoc 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 basementcover contact.
ERSLEV, E. A.,1991, Geology, v. 19, p. 617-620.
Thse de doctorat
Abstracts - Posters
Abstract prsent la Runion des Sciences de la Terre 2004 Strasbourg,

session poster.
THE ANTI-ATLAS FOLD-BELT OF MOROCCO : THICK SKIN INVERSION

TECTONICS IN THE HINTERLAND OF THE ALLEGHENIAN OROGENY
Ch. Robert-Charrue (1), M. Burkhard (2)
Geological Institute, Neuchtel University, C.P., CH-2007 Neuchtel, Switzerland, (1)
[email protected], (2) [email protected]
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 Paleozoc 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 Proterozoc 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 Mesozoc 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.
Thse de doctorat
Abstracts - Posters
Abstract prsent au Congrs 3Ma 2005 Agadir, Confrence.
TECTONIQUE DINVERSION DANS LANTI-ATLAS ORIENTAL

Robert-Charrue Ch., Burkhard M
Institut de Gologie, Universit de Neuchtel, CP 2, CH-2007 Neuchtel, Suisse
[email protected], [email protected]
L Anti-Atlas du Maroc sinscrit dans lorogne Appalachien. Limit au sud par le craton
ouest africain (WAC) et au nord par le Haut-Atlas alpin, lAnti-Atlas apparat comme un
vaste anticlinal dorientation NE-SW. Localement, le socle affleure en boutonnires,
mais la majeure partie de son difice est faite dune importante couverture palozoque
lgrement plisse et de mtamorphisme faible.
Cette chane montre un cycle de Wilson entier. La sdimentation dbute avec une phase
de rifting au Protrozoque tardif qui mne louverture de locan Rhic plus louest,
mais dont les nombreuses structures de rifting locales sont avortes. Les sdiments postrift, du Cambrien au Carbonifre, sont domins par lapport dtritique du craton. Il en
rsulte une paisse srie dalternances de quartzites, argilites, schistes et de quelques
carbonates. Lenvironnement des dpts palozoques ainsi que la courbe de subsidence
semblent davantage correspondre un bassin intracratonique qu une marge passive,
dont aucun lment profond nest prsent dans lAnti-Atlas. Durant les derniers stades
de la fermeture de locan Palotthys, la collision entre les continents Laurussia et
Gondwana entrane la dformation de ce bassin. Cet vnement varisque provoque une
sur-inversion des failles normales, une remonte du socle prcambrien et un plissement
polyharmonique de la couverture.
Une nouvelle phase dextension affecte principalement lAnti-Atlas central et oriental,
proximit du Haut-Atlas. Elle est interprte comme un nouveau cycle : louverture de
lAtlantique au msozoque. Au sein des sries palozoques, le plongement de ces
nouvelles failles normales varie selon les lithologies ce qui provoque des structures en
Rollover souvent confondues avec des plis varisques.
La topographie actuelle a certainement t rajeunie par la compression miocne (collision
alpine et inversion du Haut-Atlas) et peut-tre par un uplift thermique nogne (hot
spot). Lrosion syn- et post-tectonique et la rquilibration isostatique ont engendr par
endroit une ablation de plus de 10 km de roche, permettant laccs des roches de facis
schistes verts, ce dans une chane que moyennement dforme. LAnti-Atlas est une
chane dinversion de type thick skinned et non une chane davant-pays plisse.
Thse de doctorat
Abstracts - Posters
Abstract prsent au Swiss Geoscience Meeting 2005 Zrich, session poster.
The Anti-Atlas fold belt of Morocco : variscan inversion

tectonics and interference pattern of an intracratonic basin.
Robert-Charrue, Ch. & Burkhard, M.
Institut de Gologie et dHydrogologie, University of Neuchtel,
CP 2, 2007 Neuchtel, Switzerland, [email protected]
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 dcollement, 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 HighAtlas (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 & Ligeois 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.
Thse de doctorat
Abstracts - Posters
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 & Ligeois 2001). AAMF: Anti-Atlas Major Fault (older WAC limit); SAF: South Atlas
Fault ; SGA: Saoura gravimetric anomaly.
REFERENCES
Ennih, N. & Ligeois, J.-P. (2001): The Moroccan Anti-Atlas: the West African craton passive margin with limited PanAfrican 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., Bracne 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.
Thse de doctorat
Abstracts - Posters
POSTERS
Poster prsent la RST 2004 Strasbourg
- Poster prsent au SGM 2005 Zrich
Thse de doctorat
The Anti Atlas fold-belt of Morocco
Charles ROBERT-CHARRUE
&
Martin BURKHARD
Swiss National Science Foundation, grant n 22-52516.97 / 20-63790.00
University of Neuchtel
Switzerland
Thick skin inversion tectonics in the hinterland of the Alleghanian Orogeny
s
la
At
gh
Hi
l a
Ou
ga
rt
Pre-Cambrian
Reguibat Shield
Adoudounian
Tindouf Basin
The rheology of the paleozoc cover

dictates the geometry of the final fold
chain. This structural style depends on
two key parameters namely the total
thickness of the cover (decreasing from
12 km thick in the west to 6 km eastward)
and the ratio of competent and
incompetent layers
Hammada
Ouarkziz
Rich (1-4)
Bani 2
Bani 1
Grs verts
Cambrian
Paleozoc sediments 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. Triassic dolerites testify the
extension phase of the opening of
Atlantic ocean. The Hammada seals the
deformed folds of Anti-Atlas, but is itself
tilted a few degress to the south.
Ordovician
Atl
ant
ic o
cea
n
le
dd
Mi
Miocene compression in relation with the

Alpine collision and High Atlas inversion,
and possibly by some Neogene hot spot
related thermal uplift.
Devonian Silurian
as
l
At
The Anti-Atlas of Morocco is formed in

the hinterland of the larger Appalachian
Orogen, time equivalent with the
Alleghanian front to the West. It is limited
to the South by the Tindouf Basin and the
north-west african craton (Reguibat
Shield 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 inliners, but
the main part is made of an important
Paleozoc cover, wich is gently folded
and of low grade metamorphism. The
Anti-Atlas system clearly includes the
Algerian Ougarta chain. Syn- and posttectonic erosion and isostatic reequilibration lead to the removal up to 10
km of overburden, bringing lower greenschist facies rocks back to the Earths
surface - allowing for deep insight into
this otherwise mildly deformed fold belt.
The high present day topography of AntiAtlas has been rejuvenated by late
Carboniferous
Rif
Geological setting
Mediterrannean Sea
Cretaceous
Kingdom of Morocco
Schistes Paradoxydes
Grs roses
Calcaires sup
Lie de Vin
Calcaires inf
Conglomrats de Base
P III
P II
Basement
Landsat view
10 km
This image shows the excellent

outcrop condition. Note (from NW to
SE) : The Alpine High Atlas, The
Ouarzazate Basin, The Saghro AntiAtlas Inliner (dark-green), the
foldedPaleozoc cover appearing as
cuestas, here and there chopped by
normal faults or graben and finally the
sealing light-green Hammada.
Landsat 4/5 1990. Bands : 7 / 4 / 2.
Original pixel is at 28.5 meters.
Source of Landsat images :
https://zulu.ssc.nasa.gov/mrsid/
Cross section
Cretaceous-Neogene Hammada
Carboniferous Ouarkziz
Devonian Rich
Upper Ordovician Bani 2
Lower Ordovician Bani 1
Upper Cambrian
Middle Cambrian
Adoudounian
P III / Basement
Conclusions
Cross section
NW
(M)
SE
5000
-5000
-5
65 km
Cross section shown on the

landsat view. Stratigraphy from
Cambrian to Devonian. The superstructure is a huge antiform. It
appears to be chopped by normal
faults. The dip of these normal faults
varies according to the lithology and
produce Rollover structures. No
major detachement is observed at the
basement-cover contact, wich
indicates a Thick Skin structural style.
Reconstructed cross section

of the Cambrian and
Ordovician series at different stages :
(1) after deposition ,(2) after
compression
and (3) actual
configuration (after normal faulting).
Stage 2 is the configuration at the end
of the variscan orogeny. The
horizontal shortening between stage
1 and 2 reaches 7.5 km (~12 %).
An entire Wilson cycle of mountain

building left its imprint in this belt.
1. Sedimentation started with a Late
Proterozoc rifting phase (opening of
Rheic and Paleotethys oceans to the
west).
2. Post-rift sediments of lower
Cambrian to middle Carboniferous
age are dominated by detrial input,
leading to the accumulation of a thick
pile of sediments.
3. 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
concomittant folding of the overlying
cover series.
4. A post-folding phase of extension is
interpreted as a new cycle of HighAtlas / Atlantic rifting.
(km)
By measuring the differential

surface between stages 1 and
2 (225 km2) and the shortening (7.5
km), the depth of detachment must be
of 33 km. But if the LPS (Layer Parallel
Shortening) observed in the field,
reaches 50 %, the detachment will be
localised at 15 km depth.
N
S
225 km^2
-3
15 km
-18
-33 km
7.5
km
65 km
Trishear model : Erslev, 1991
The Anti-Atlas is a thick skinned

inversion belt, not a thin skinned
foreland fold-thrust-belt.
Charles ROBERT-CHARRUE / Institut de gologie / Universit de Neuchtel / CP 2 / CH-2007 Neuchtel / ++41 32 718 27 39 / [email protected]
STRUCTURAL
STUDY
OF
THE
EASTERN
Anti-Atlas
variscan
belt
of
IGH
Morocco
PALEOGENE NEOG.
High-Atlas Inversion
The present day high
topography of the Anti-Atlas (up
to 2500 m) does not result from
the variscan orogeny alone.
Cretaceous to Neogene
sediments around the eastern
Anti-Atlas show some
unconformities and tilts which
correspond to the inversion of
the High-Atlas in EoceneOligocene and PliocenePleistocene times (Frizon de
Lamotte et al., 2000). A
Neogene thermal uplift may
also rejuvenate the topography.
But except doming, no
significant deformation reaffects the variscan belt.
L
RA S
A
NT
CE -ATL
H
G
HI
N 32
N
ER S
ST LA
EA I-AT
T
AN
N 31
CRETACEOUS
W 005
W 004
N 30
W 006
Digital Elevation Model from eastern Anti-Atlas and central Anti-Atlas region.
(Data from GTopo30 @ http://edcdaac.usgs.gov/gtopo30/gtopo30.asp).
A New Cycle Starts

N
4000
3000
2000
1000
JURASSIC
0
-1000
-2000
-3000
-4000
Hanging-wall syncline
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 especially
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.
Detachment Depth
5
-5
-5
-10
-10
-15
-15
PERMIAN
Egg-Box Interference Patterns

from Variscan Inversion
Blind thrusts in the shales.
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. 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
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
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.
Unsuccessful Rift
Intracratonic Basin
The entire 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 ocean. 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
in eastern parts. Note that the total thickness and the
incompetent/competent ratio are the two key parameters
which will dictate the structural style in the cover.
150505PP/12
0
031103PP/21
0
041103PP/19
0
051003PP/17
081002PP/12
0
360
290602PP/28
0
031002PP/18
0
080302PP/9
0
240902PP/24
0
161003PP/9
0
0
0
111003PP/5
0
180302PP/13
0
160303PP/10
0
0
Small scale fold. Stick is one meter.
400
121003PP/25
0
210302PP/6
0
Except north of the inliers, the

small scale folds axis have the
orientation of the Ougartian
trend (NW-SE in the eastern
area, E-W in central Anti-Atlas).
LITHOLOGICAL SYMBOLS
Conglomerate
440
Sandstone / Quartzite
View of the eastern Anti-Atlas with stereograms of the bedding plane poles
of small scale folds.
Landsat 7 downloaded @ https://zulu.ssc.nasa.gov/mrsid/
Sandstone and shale
Post Variscan
Carboniferous
Devonian
Silurian
Ordovician
Cambrian
Adoudounian
PIII
Basement
Carbonate and sandstone
WAC Limits
Carbonate
Eastern Anti-Atlas
l
tA
A
Tindouf
Basin
MF
Ou
ga
Algeria
Continent
Orogens
except Anti-Atlas sytem
Palaeozoc
Cover of Anti-Atlas system
Basement inliners
of Anti-Atlas system
rta
30 N
29 N
006 W
007 W
Oran
Meseta
Exposed further west in the

Bou Azzer inlier, the Major AntiAtlas Suture or Anti-Atlas Major
Fault (AAMF) was interpreted
as the old front of the african
shield but is newly seen as
integrated in a metacratonic
evolution. The northern
boundary of the WAC is now
defined by the South Atlas Fault
(SAF) bording the High-Atlas to
the south and by the Saoura
gravimetric anomaly (SGA) to
the east.
008 W
n
tic
an
Atl
Reguibat Shield
SW Anti-Atlas
le
d s
id la
M At
s
l a
A t
F
g h SA
H i
n
N 28
W 002
W 006
W 008
W 010
Rabat
Moroccan
Meseta
N 32
Volcanics
Mediterranean
Sea
Rif
100 km
Oc
Volcaniclastics
490
N 34
N 36
ea
Black shale
Shale
Syncline axis
31 N
Kingdom
of
Morocco
A
SG
CAMBRIAN
Shale and carbonate
Anticline axis
003 W
ORDOVIC. SILUR.
140505PP/15
0
004 W
CARBONIF.
300
EDIACARA
Reconstructed profile at the end of

Variscan deformation (without normal
faults). Detachment depth at -33 km
with mesured 7.5 km shortening and at
-18 km with 15 km shortening including
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
assumed 50 % of LPS is not over estimated and yields a detachment depth at -18 km which
corresponds to a more reasonable midcrustal detachment.
The particular rheology of the cover

dictates the deformation style. There is no
major dcollement. The deformation is
absorbed in pureshear style by an
anabundance of incompetent layers.
Competent layers are gently buckle-folded
in a trishear mode (Erslev 1991). The
Layer Parallel Shortening must be very
significant but it remains difficult to
determine.
N 30
NEOPROTEROZOIC
Cross section of Ougnate inlier. Red is basement (incl. rift sediments), khaki :
Cambrian, orange : Ordovician, purple : Silurian, green Devonian. Note the
hanging-wall syncline in the Devonian, produced by normal faulting.
200
Ultraweak Cover
DEVONIAN
PALEOZOIC
TRIASSIC
65 km
005 W
MESOZOIC
CENOZOIC
the
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.
Neoproterozoic Rifting
References
The basement is an assemblage of crystalline, metamorphic

and sedimentary rocks with a complex history (Ennih & Ligeois
2001). The Anti-Atlas corresponds to the northern limit of the
Eburnian West African Craton (c. 2 Ga), reactivated in the
Neoproterozoic continental breakup and last consolidated
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
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.
Ennih N., J.-P. Ligeois (2001), The Moroccan Anti-Atlas: the West African craton passive margin with limited PanAfrican activity. Implications for the northern limit of the craton, Precambrian Research, 112, pp. 289-302.
Erslev E.A. (1991), Trishear fault-propagation folding, Geology, v. 19, pp. 617-620.
Frizon de Lamotte D., B. Saint Bezar, R. Bracne (2000), The two main steps of the Atlas building and geodynamics of
western Mediterranean, Tectonics, Vol. 19, N 4, pp. 740-761.
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.
Charles ROBERT-CHARRUE & Martin BURKHARD

Universit de Neuchtel - 2007 Neuchtel
[email protected]
D
FIGURES
11*.
Carte gologique de lAnti-Atlas central et oriental
17*.
Coupes gologiques de lAnti-Atlas oriental, avec position

des coupes et paisseurs des sries
22*.
Strogrammes des structures mesures
28*.
Modle numrique de terrain de lAnti-Atlas oriental
34*.
Faisceaux dans lAnti-Atlas oriental
36*.
Topographie et gologie de surface selon les coupes
40*.
Interprtation de la carte gologique (plis et faisceaux)
41*.
Modle tridimensionnel du modle numrique de terrain
Thse de doctorat
Post-varisque
Carbonifre
Dvonien
Silurien
Ordovicien
Cambrien
Adoudounien
31 N
Ouarzazate / Anti-Atlas
Socle
30 N
003 W
004 W
005 W
006 W
007 W
008 W
29 N
km
4
km
01 S
4
2
-2
-2
CROSS SECTIONS
OF
EASTERN ANTI-ATLAS
OF MOROCCO
1: 200000
km
4
km
02 S
no vertical exageration
4
2
-2
-2
km
4
km
03 S
3000
Neogene
Paleocene
Cretaceous
Jurassic
Triassic
Late Carboniferous
Early Carboniferous
Devonian
Silurian
Late Ordovician
4000
Early Ordovician
-2
-2
0m
1000
2000
km
4
2
km
04 S
4
2
-2
-2
5000
Middle Cambrian
Early Cambrian
6000
10 09 08 07 06 05 04 03 02 01
km
6
4
05 S
km
6
4
-2
-2
-4
-4
km
6
4
06 S
km
6
4
2
-2
-2
-4
-4
-6
-6
km
6
4
2
07 S
km
6
4
2
-2
-2
-4
-4
-6
-6
km
8
6
08 S
km
8
6
4
-2
-2
-4
-4
km
8
6
09 S
km
8
6
-2
-2
-4
-4
-6
-6
km
8
6
10 S
km
8
6
-2
-2
-4
-4
-6
-6
0
0
0
N 3130
0
0
N 3100
0
0
0
0
W 400
W 430
W 500
W 530
W 600
W 630
N 3030
N 3000
N 32
N 31
4000
3500
3000
2500
2000
1000
500
m
N 30
W 003
W 004
W 005
W 006
W 007
1500
006 W
004 W
005 W
50 km
31 N
30 N
1500
TOPOGRAPHIE ET GEOLOGIE
1000
500
AL
T
01
TA
F
1500
1000
02
500
Trias
Carbonifre sup
Carbonifre inf
Dvonien
Silurien
AT
E
ER
AD
U
O
E
IN
0 km
Faille normale
SA
E
IN
R
ZA
TA
Z
20
40
60
80
100
Topographie due
lamincissement
crustal selon
Missenard et al. 2006
08
GRABEN DE ZAGORA
R
H
0m
Prcambrien
06
AF
500
Cambrien inf
SD
BASSIN DE
OUARZAZATE
1000
Cambrien sup / ND
TI
2500
1500
05
07
0m
2000
Ordovicien inf
Faille chevauchante
500
HAUT-ATLAS
1000
1000
1500
Jurassique
Ordovicien sup / ND
1500
2000
Crtac
04
2000
0m
Palogne
03
0m
500
Nogne
Echelle verticale 1 : 55000

Echelle horizontale 1 : 550000
Exagration verticale 10 x
IL
0m
DE SURFACE ET SUBSURFACE
DE LANTI-ATLAS ORIENTAL
2em BANI
09
1er BANI
120
140
10
160
180
Post Varisque
Carbonifre
Devonien
Silurien
Ordovicien
Cambrien
Adoudounien
AA-Ouarzazate
Socle
Axe anticlinal
Axe synclinal
Anti-Atlas
Ougarta
31 N
30 N
003 W
004 W
005 W
006 W
007 W
008 W
29 N
E
ATLAS LANDSAT 7
NW Africa
Anti-Atlas
AA oriental : dcoupage et lgendes de la mosaque
Mosaque de lAnti-Atlas oriental
Thse de doctorat
NORTHWESTERN AFRICA
LANDSAT 7 MOSAC VIEW
GEOLOGICAL
GENERAL MAP
NEAN
RRA
DITE
ME
UX
EA
AS
AT
L
TL
HP
NA
G
I
A
H
R
HA
SA
RIF
EA
IC
T
AN
OC
MESETA
LA
H AT
HIG
L
AT
SEA
TELL
GRAND ERG
OCCIDENTAL
TUNISIAN
ATLAS
GRAND ERG
ORIENTAL
INT
LA
AT
TINDOUF BASIN
CANARY
ISLANDS
OU
GA
RT
A
TADEMAIT
TASSILI
REGUIBAT SHIELD
TARGUI SHIELD
HOGGAR
Recent sedimentary rocks
Mesozoic folded or not
Recent volcanic rocks
Palaeozoic + late Precambrian
Cenozoic
Precambrian (except latest)
Views downloaded from :

NASA
Applied Sciences Directorate
John C. Stennis Space Center
https://zulu.ssc.nasa.gov/mrsid/
Landsat ETM+ bands : 7-4-2
Original pixel size 14.25 m
ANTI-ATLAS
OF
MOROCCO
100 km
004 W
005 W
006 W
Rpartition des vues
31 N
50 km
30 N
Lgendes
Palmeraie / Dune
Carbonifre
Ouarkziz
Cambrien moyen
Cnozoque
Dvonien
Rich
Cambrien infrieur
Crtac
Ordovicien sup
m
e
2 Bani
Adoudounien
Jurassique
Ordovicien inf
e r
1 Bani
Socle

These RobertCharrueC

Transféré par

Droits d'auteur :

Formats disponibles

These RobertCharrueC

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

These RobertCharrueC

Transféré par

Droits d'auteur :

Formats disponibles

Institut de Gologie et dHydrogologie

- Soutenue le 19 septembre 2006 -

Accepte sur proposition du jury :

Gologie structurale de lAnti-Atlas oriental

Gologie structurale de lAnti-Atlas oriental

Gologie structurale de lAnti-Atlas oriental

CE TRAVAIL EST DDI LA MEMOIRE DE

NOUS GARDERONS LE SOUVENIR

Gologie structurale de lAnti-Atlas oriental

Gologie structurale de lAnti-Atlas oriental

Gologie structurale de lAnti-Atlas oriental

Gologie structurale de lAnti-Atlas oriental

Cadre gologique Etat des connaissances

Structures Description des observations

Modle Gomtrique Interprtations des observations

Modle Dynamique Chronologie des vnements

Conclusions Apports de ltude

Table des Figures, Tableaux et Graphiques

Gologie structurale de lAnti-Atlas oriental

SOMMAIRE DES ANNEXES

Gologie structurale de lAnti-Atlas oriental

Gologie structurale de lAnti-Atlas oriental

(2006) de M. Burkard, S. Caritg, U. Helg, Ch. Robert-Charrue et A. Soulamani

1.2. Situation gographique et bases gologiques

La chane de lAnti-Atlas se prsente sous la forme dun vaste bombement

Gologie structurale de lAnti-Atlas oriental

Au sud-ouest lAnti-Atlas se prolonge par la chane du Zemmour et les

1.3. Buts de ltude

Gologie structurale de lAnti-Atlas oriental

Le projet dans sa globalit vise dcrire la structure de cette chane plisse et

Gologie structurale de lAnti-Atlas oriental

1.4. Mthodes, outils, donnes

Gologie structurale de lAnti-Atlas oriental

1.5. Logistique et approche

-Terrain, Anti-Atlas oriental du 7 au 29.03.

-Terrain, Anti-Atlas oriental du 22.09 au 12.10.

-Terrain, Anti-Atlas oriental du 10.03 au 2.04.

-Terrain, Anti-Atlas oriental du 5.10 au 4.11.

-Anti-Atlas occidental avec M. Burkhard, S. Caritg et Zeshan

-Excursion de lInstitut de Gologie, UniNe, 5 - 21.04.

-Congrs 3Ma (Magmatisme, Mtamorphisme et

Les autorisations de mission de terrain ont t dlivres par la Division de la

Gologie structurale de lAnti-Atlas oriental

Le travail de terrain a t effectu seul. La carte daccs (Figure 3) montre les

Figure 3. Carte daccs des tudes menes dans lAnti-Atlas oriental.

Lapproche du terrain a t dicte par les recherches. Outre les mesures de la

Gologie structurale de lAnti-Atlas oriental

(Rissani-Alnif-Tazzarine) (Figure 3). Les possibilits de traverses des massifs

Gologie structurale de lAnti-Atlas oriental

Gologie structurale de lAnti-Atlas oriental

Figure 4. Palogographies de Stampfli et Borel (2002). Ltoile indique la position de lAntiAtlas.

Gologie structurale de lAnti-Atlas oriental

2.2. Elments morpho-gologiques de lAnti-Atlas oriental

Figure 5. Elments gologiques de lAnti-Atlas oriental. Daprs la Carte Gologique du Maroc

Le socle apparat sous la forme de boutonnires. Douest en est : la boutonnire

Gologie structurale de lAnti-Atlas oriental

6 A : Vue vers le nord

Gologie structurale de lAnti-Atlas oriental

6 B : Vue vers louest

6 C : Vue vers le sud