The West African Orogens and Circum-Atlantic Correlatives

The geology of Africa is dominated by the Kalahari, Congo (Zaire) and West African cratons (Fig. 1). These are composed of late Archean and early-middle Proterozoic crystalline basement rocks which are nonconformably overlain by middle to late Proterozoic-earliest Paleozoic shelf sedimentary sequences and successions within younger Paleozoic basins (Fig. 1). Kennedy (1964) and Clifford (1968) pointed out that most regions surrounding the African cratons record affects of ca. 750–500 Ma tectonothermal activity (Pan-African). The history of these orogenic belts and the role of plate tectonic processes in their evolution has been uncertain and controversial (e.g. Kroner 1977, 1979). Results of recent, collaborative field, geochemical, geophysical and geochronological investigations in the West African orogenic terranes have significantly enhanced our understanding of the tectonothermal evolution of the enigmatic Pan-African orogens. These data are summarized in this volume.

West Africa, facing the Laurentian basement of North America and the Brazilian craton of South America in a Bullard-type predrift reconstruction, is largely identified to the West African craton (a Precambrian granitized basement) and its surrounding Upper Proterozoic (Pan-African) mobile belts. In contrast with relatively complete geological studies (Fabre 1983), West Africa did not benefit from a large number of extensive geophysical studies. Only gravity measurements have resulted in a nearly complete regional mapping and thereby are useful for investigation of the crustal structure and clarification of geological interpretation. In places, seismological studies such as the detection of teleseismic travel time anomalies or surface waves velocity dispersion measurements, magnetotelluric soundings, and heat flow measurements have been made. Considering the total absence of deep seismic profiles, all these data furnish a precious adjunct to the knowledge of the crustal and upper mantle structures of West Africa.

The West African craton is a very extensive portion of Precambrian crust (~4500000 km²), stable since 1700 Ma ago, bounded on all sides by more recent mobile belts mainly of Pan-African age, such as the Mauritanide fold belt (Fig. 1) on the western edge.

The West African craton, as defined by Kennedy (1964), represents a very large area stabilized after the Eburnean orogeny, 2150±150 Ma ago (Black and Fabre 1983, Bessoles 1983). These old rocks outcrop today in two separate shields: the Reguibat Shield and the Man-Leo Shield. The cover sequences, beginning after a long period of erosion, appear in roughly concentric belts, in the Taoudenni basin, between the two shields. Younger deposits are found in the eastern part of the craton toward the Pan-African suture zone.

The Pan-African belt of central northwest Africa is part of the Transaharan belt (Cahen et al. 1984) and has been interpreted by Black et al. (1979) and Caby et al. (1981) as a collision belt formed during a Wilson cycle. The Tilemsi suture zone separates the stable West African Craton (passive margin) from the mobile belt (active margin). Pan-African deformation and metamorphism affected the whole Hoggar shield from the Tilemsi region at least to the 8°30′E lineament and is recognized on a 1000-km-wide area (Fig. 1). Actually, the Transaharan belt appears as a complex assemblage of NS trending geological domains which have undergone different evolutions. Large areas are still poorly known and several interpretations have been proposed for some others. In this review paper, I will present the geological evolution of each domain and discuss these different interpretations. Correlations of these domains with the Togo-Benin belt and with the Nigerian shield have been proposed by several authors (Bessoles and Trompette 1980; Caby et al. 1981; Ajibade et al. 1987; Caby, in press).

The Dahomeyide orogen is located along the southeastern margin of the West African craton, and is exposed throughout eastern Ghana, Togo, Benin, Nigeria and Cameroon (Fig. 1). The West African craton has remained tectnically stable since ca. 2000 Ma (Black 1985 a, b; Camil 1984). Western sectors of the craton are characterized by an Archean domain represented by grey gneisses and migmatitic complexes (amphibolite and granulite metamorphic facies), acid to basic granulites and charnockitic complexes, together with subordinate greenstone rocks, supracrustal metasedimentary and metavolcanoclastic rocks and granites. These rocks were affected by orogenic events at ca. 3000 Ma (Leonian) and ca. 2800 Ma (Liberian). Southeasternmost segments of the Archean domain were partially remobilized at ca. 1800 Ma during Eburnean orogenesis. Western sectors of this domain were variably affected by late Proterozoic (Pan-African) orogenic events.

The term “Rokelide” (Allen et al. 1967) was proposed for the orogenic belt in Sierra Leone that was deformed during the Pan-African tectonothermal event, about 550 Ma (Kennedy 1964). Upper Archean basement gneisses and upper Proterozoic to lower Cambrian clastics were deformed and metamorphosed during this orogenic event.

The Bassarides of eastern Senegal and northern Guinea constitute a segment of a continuous orogenic terrane which borders the western edge of the West African craton. The name, given by Villeneuve (1984), refers to the NNE-SSW Bassaris mountain ridge that runs across eastern Senegal and northern Guinea. We distinguish two tectonic branches. The eastern one, called the Bassaris branch, has a NNE-SSW trend, while the eastern one, called the Koulountou branch, has a NE-SW trend. These two branches were separated by the triangular Youkounkoun Basin (Fig. 1). The Bassaris branch is separated from the West African craton, represented here by the Kedougou inlier and the Madina-Kouta Basin, and by the Faleme and the Komba Basins. All of these geological elements are overlain by the Bové Basin which represents the Paleozoic cover of the Bassaride orogen.

The West African fold belt (Sougy 1962 b), has generally been presented as including the Mauritanide, Bassaride and Rokelide orogens (Fig. 1). In this paper we shall consider the Mauritanide belt sensu stricto, that is to say the “Akjoujt-Bakel bow” and its northern, poorly known, extensions in the Western Sahara and Zemmour. The orogen is bordered on the west by unconformably overlying Mesozoic and Cenozoic sedimentary sequences of the Senegal-Mauritanian and El Ayoun coastal basins. Large segments of the orogen are covered by Quaternary sand dune complexes.

Geological exploration of Morocco began at the end of the last century and developed rapidly. The first reliable synthesis of much of the country was made by Gentil and illustrated by a map at a scale of 1/500000 (Gentil 1920). Several regional descriptions were carried out by the Geological Survey, progressing roughly from north to south. Choubert summarized the geology with a synoptic map at 1/500000 (Choubert 1952), followed in 1985 by the current 1/1000000 map, which includes the southern provinces. More recent work has often involved collaboration between the Geological Survey and other organizations such as the rapidly growing Moroccan universities. Present research is devoted to thematic studies that are largely structural, and detailed mapping at scales of 1/100000 and 1/50000.

The Iberian Peninsula consists of two very different parts: (1) the eastern half which constitutes the westernmost extension of the Alpine orogen; and (2) the western half (Iberian Massif) representing the southwesternmost and largest piece of the so-called European Hercynian Foldbelt. Although Precambrian and Paleozoic inliers are scattered within the former, they were generally so strongly reactivated during the Alpine orogeny that any attempt of correlation with the Iberian Massif has as yet failed. The latter, on the other hand, has been extensively surveyed in recent decades, and a summary of the present state of the art constitutes the main objective of this contribution. A tectonostra-tigraphic terrane analysis perspective is preferred for this purpose. This will allow a more accurate unravelling of each individual terrane evolutionary history, which will in turn greatly facilitate correlation.

The pre-Mesozoic basement of central-western (CW) Europe is exposed in several scattered massifs (Fig. 1). This is because of covering by post-hercynian sedimentary basins, superimposition of the Alpine fold belt, and disruption by the opening of the Bay of Biscay and the Mediterranean system.

The nature and age of the Alpine basement has been a controversial topic of the Alpine geology (e. g. von Raumer 1988). The fossil record begins in the Late Ordovician (for a review, see Schönlaub 1979) except for scarce sporomorphs and acritarchs in metamorphic areas (Giorgi et al. 1979; Reitz and Höll 1988; Sassi et al. 1984) which are not related to continuous Late Ordovician to Carboniferous sections. These data suggest the existence of a distinct late Proterozoic to Early Ordovician sedimentary cycle separated from the later one. Geochronological data from crystalline complexes structurally removed from low-grade sediments favour the presence of a basement metamorphosed and cooled during Early Paleozoic times (for a recent review, see Ebner et al. 1987; Sassi et al. 1987). The relationship between weakly metamorphosed fossiliferous Late Ordovician to Carboniferous strata and crystalline complexes is reviewed in this chapter.

Possible correlatives of the West African Orogens in the northern Appalachians occur along the eastern seaboard of North America, where they comprise the Avalon Composite Terrane and the Meguma Terrane. The Avalon Composite Terrane is defined by the presence of a lithostratigraphically correlative, subaerial-shallow marine, Cambro-Ordovician overstep sequence containing an Acado-Baltic fauna (Keppie 1985). In general, these Cambro-Ordovician rocks are only preserved in outliers, so overstep relationships are inferred from their close lithostratigraphic similarity. Where the Cambro-Ordovician rocks have been removed by erosion, it is possible to use the presence of the lithostratigraphically correlative, Silurian-Gedinnian overstep sequence containing the distinctive Rhenish-Bohemian fauna (Boucot 1975). By this definition, the Avalon Zone includes New England southeast of the Hope Valley-Bloody Bluff fault zone; Maine and New Brunswick south of the Turtle Head and Honeydale faults; most, if not all of northern Nova Scotia, and Newfoundland east of the Dover-Hermitage fault (Fig. 1). Terranes with Late Precambrian rocks northwest of these faults, such as the Hope Valley and Nashoba Terranes (O’Hara and Gromet 1985), have not been included in the Avalon Composite Terrane because the distinctive, Avalonian, Early Paleozoic rocks are absent. Should future work show that the Early Paleozoic overstep sequence was present before erosion, these terranes would need to be added to the Avalon Composite Terrane. In the Avalon Composite Terrane, the Early Paleozoic rocks rest unconformably — disconformably — conformably upon a variety of Late Precambrian igneous rocks and turbidites overlying Middle-Late Proterozoic or older metasediments and gneisses.

Central and southern segments of the Appalachian orogen have long been subdivided into several northeast-trending lithotectonic “belts” (Fig. 1) on the basis of contrasts in lithologic character, metamorphic grade, and/or structural style (e.g. King 1955; Hatcher 1972,1978b, 1987; Rankin 1975; Glover et al. 1983). The nomenclature and regional configuration of these belts as well as the character of intervening boundaries has changed markedly over the last 30 years (e.g. King 1955; Overstreet and Bell 1965a, b; Hatcher 1972, 1978b, 1987; Butler 1980). Such subdivision of the orogen has become increasingly more difficult to reconcile with results of recent detailed geologic mapping where, for example, protolith sequences within a low-grade belt may be traced directly into an adjacent high-grade belt (e. g. Overstreet and Bell 1965b; Griffin 1972; Secor et al. 1986a). In addition, contrasting times of regional metamorphism have been documented within different areas of the same belt (e. g. Dallmeyer et al. 1986; Noel et al. 1988). These and other similarly contradictory relationships have led to the suggestion by some workers (e. g. Zen 1981; Williams and Hatcher 1982, 1983; Secor et al. 1986b; Higgins et al. 1986, 1988; Noel et al. 1988; Dallmeyer 1986, 1989a; Horton et al. 1989, 1991) that the central-southern Appalachian orogen is comprised of several markedly different, fault-bounded structural units (Fig. 2) with tectonic significance similar to that of tectonostratigraphic terranes defined within Mesozoic-Cenozoic orogens of the Circum-Pacific realm (e.g. Coney et al. 1980; Jones et al. 1983a, b; Howell and Jones 1984).

Northeast Brazil comprises two structural provinces defined by Almeida et al. (1981). To the north the Parnaiba province includes the São Luis craton as a probable remnant of the West Africa craton in Brazil (Hurley 1967) and the Gurupi belt (Hasui et al. 1984). Most of northeast Brazil belongs to the Borborema province (Fig. 1) which extends for about 400000 km². The major tectonic feature of this province is a system of sinuous and branched, anastomosing shear zones, which divide the province into elongate domains that often differ in lithology, metamorphic grade and structure, suggesting large-scale relative displacements (Schobbenhaus et al. 1981).