The Togolese coastal basin basement corresponds to part of the western domain of the so-called “internal zone” of the Pan-African Dahomeyides belt [7]. It consists of gneisso-migmatitic units associated with rare late Pan-African granitic intrusions [8] [9] [10]. These lithostuctural units containing ortho- or para-derived petrographic rocks are considered to be of Eburnean age (2000 ± 200 Ma; [6] ), remobilized during the Pan-African orogeny (600 ± 50 Ma; [11] [12] ). They represent the frontal part of the Benino-Nigerian Shield that collided with the eastern margin of the WAC during the ultimate stage of the Dahomeyides edification [7]. The fracture network in the gneisso-migmatitic units is attributed to the two last Pan-African deformation phases [13] [14] [15]. These major fractures have probably influenced the coastal basin setting up.

The Meso-Cenozoic cover of the Togolese coastal sedimentary basin is defined as a monoclinal mega-sequence with very low dip to the SE where it approaches 600 m thickness (Figure 2). Both basement and sedimentary cover are compartmentalized by normal faults. Formations outcrop in NE-SW trending parallel bands corresponding to different stages of marine evolution [4]. These outcrops become blurred toward the north of the basin by an erosional limit beyond which occur outliers of the Tohoun-Tado Sandstones (Figure 1).

According to [4] and [18], the mega-sequence of the Togolese coastal basin consists of three series, namely:

1) The lower series or “Tabligbo Group”, consisting, from base to top, of Campanian-Maastrichtian detrital deposits (sands and sandstones), Paleocene-Eocene limestones and laminated mudstones;

2) The middle series of lower to middle Eocene age, called the “Hahotoe-Kpogame phosphatic Complex” [4], bringing together, from bottom to top, phosphatic marls, phospharenites (phosphate deposits in exploitation) and phosphatic mudstones;

3) The upper series (upper Eocene to Quaternary) representing the “Continental terminal s.l.” and consisting of sandy clay detrital deposits with pebbly horizons. The latter, is separated from the preceding series by an erosion surface (Figure 2).

The genesis and evolution of the Togolese coastal basin, like its West African equivalents, are associated with the process of the opening of the South Atlantic equatorial segment. This opening occurred in the global dynamic of the dislocation of Gondwana that separated the African and South American continents since the Lower Cretaceous [19] - [29]. For [30], these basins have several tectono-sedimentary similarities. They are made of continental (fluvio-lacustrine) pre- to syn-rift valley filling deposits overlain by marine sediments generally transgressive on anterior deposits. Basin filling generally ends with a detrital continental series.

Two faulted sites were studied. The first site (N07˚03'13.7''; E01˚36'39.6'') includes sandstone outcrops attributed to the lower members of the sequence. The planes surveyed bear striations whose geometry indicates normal fault movements (Figure 5(a)) associating a more or less important strike-slip fault displacement (oblique slip fault, Figure 5(b)), or exclusively strike-slip fault (Figure 5(c)). Outcrops at the second site (N07˚02'12.5''; E01˚37'36.5'') belong to the upper member of the sandstone sequence and only show normal fault type striated planes (Figure 5(d)).

Several plane directions make up the fracture networks at the two sites. Based on the intersection criteria and superposition of incompatible striae [32] [33] [34], the planes are attributed to many fracturing episodes. A total of 334 striated planes were recorded. From this fault population, only 163 planes are retained considering conformity with movement criteria [35]. The measurements were first subjected to manual stereographic processing to complete striae characterization.

Data on the striated planes were processed using the numerical calculation software Tectonics FP 1.75 [36]. This software requires as inputs, the geometrical parameters of planes (Dip direction/dip) and striae on planes, as well as fault type and the precision for each plane measurement. Three methods are proposed for the calculation of tensors namely: 1) Direct Inversion [37] [38] [39] [40]; 2) Numerical Dynamical Analysis (NDA; [41], review in [42] ); and the 3) Dihedra calculation [43] [44] [45]. The Direct Inversion methods are used for inherited faults systems [44] [46], while the NDA allows determining the parameters of deformation axes rather than stressing characteristics. Thus we have chosen the third method (Dihedra calculation) which is recommended for conjugated fault systems [43] [47] [48]. After the treatments, the states of the paleostresses obtained for each system or fracturing episode are synthesized in Table 1.

At site 2, the intersection criteria allowed to establish a relative chronology of at least three episodes of development of striated planes (Figure 6). The first episode is represented by the E-W to WNW-ESE (N80 to N100) planes. The second generation expresses itself in the NW-SE to NNW-SSE (N120 to N160) directions and finally, the last generation corresponds to the NE-SW (N40 to N70) planes. The same generations of normal faults were recorded at site 1, but one notes the presence of families of NE-SW dextral and NW-SE sinistral strike slip faults. Each generation of normal faults is considered as a system allowing determination of the responsible theoretical stress tensor. The system of strike slip faults at the site 1 is considered to be a set of fault reactivation planes linked to the last episode. These planes were analyzed separately.

Site 1

The striated planes recorded at the site 1 have medium to high dips (40˚ to 80˚) in many directions. The pitch of striae and associated tectoglyphes allow identifying normal, normal strike-slip and strike-slip faults. The network is made up of three major normal fault directions which correspond to three conjugated fault systems associated with different fracturing episodes. The NW-SE planes are more represented than the NE-SW and E-W normal faults. The stress tensors determined for each system are the following:

­ The system of E-W conjugated normal faults express a subvertical maximum principal stress axis σ1, oriented N036-85 and a minor stress σ3 oriented N177-04 (Figure 7(a));

­ The second generation system of NW-SE conjugated normal faults corresponds to a paleostress tensor with principal axes σ1 N171-86 and σ3 N044-02 (Figure 7(b));

­ All the NE-SW normal faults assume a tensor with paleostresses σ1 N253-64 and σ3 N344-00 (Figure 7(c)).

The strike-slip faults are organized in a system of conjugated families with NE-SW (N40 to N70) dextral and NW-SE (N115 to N130) sinistral planes. This system of strike-slip faults seems to correspond to a tensor of paleostresses, with subhorizontal σ1 and σ3 oriented N087-06 and N178-07 respectively (Figure 7(d)).

Site 2

Sandstones belonging to the upper member of the sequence (“Tohoun Sandstones”) bear decimetric to decametric normal fault planes in E-W, NW-SE and NE-SW directions. Thus, in the image of site 1, site 2 shows imprints of the three fracturing episodes. Stress tensor is also determined for each episode.

The primary episode is expressed by the E-W planes generally filled in by a greyish brown to blackish fine grained ferruginous sandstone. Analysis of the system of normal faults indicates a tensor of paleostresses with subvertical σ1 N319-85 and σ3 N188-04 (Figure 7(e)).

The second episode corresponding to several NE-SW striking planes filled with whitish silica, results from a tensor with σ1 N079-76 and σ3 N209-09 (Figure 7(f)).

The last concretization of fault tectonics is defined by a population of planes with NE-SW directions. These planes suggest the principal stress axes σ1 N277-85 and σ3 N153-03 (Figure 7(g)).