The area of interest is well covered by 2D seismic survey acquired by the Maxus Ltd in 1983 and 1994. These seismic profiles are of moderate quality and some of them are not migrated. Thus we present thereafter careful interpretations.

We use the seismic profile L1 (with additional information from profile L2, Fig. 2) trending north–south. This profile crosses the Utique fold and extends for about 15 km from northern Utique to Djebel Ammar to the south. The profile has been depth-converted thanks to stratigraphic data coming from wells Utique1 and Bm1, closely located (2.2 and 1.8 km, respectively; Fig. 2). The Ut1 well is 2500 m deep.

The seismic profile L1 exhibits an erosional unconformity, which separates two stratigraphic packages (Fig. 5). This unconformity is marked by the truncation of north-dipping reflectors corresponding to a lower stratigraphic package and by the onlap of the upper stratigraphic package. Ut1 well data indicates the lower package is made of Mesozoic sediments whose strata in the upper part could be seen, its base being probably made of Triassic rocks. The overall Mesozoic sequence slightly dips to the north. A thick evaporitic section of the Mellaha formation (Serravalian) seals the north-dipping Triassic strata. Such an unconformity was previously reported by Brusset (Brusset, 1999) and Khomsi et al. (Khomsi et al., 2009). The Mellaha fm. is overlain by strata from Upper Miocene upward, which are merely isopach and constitute the upper package. These are only deformed by a thrust branching at depth into the Mellaha formation acting as a decollement level (see thereafter and Fig. 5). The structural style can be tracked on the seismic line. It consists obviously of a triangle zone defined by a deep-seated fault-bend fold and a shallow fault-propagation anticline (Utique anticline). Both folds are associated with south-verging thrusts, which sole-out in Triassic and Mellaha evaporates, respectively (Fig. 5). One must note that if the Mesozoic and the Lower Miocene formations are missing in the Utique structure, they are found at outcrop just 5 km to the south in the Djebel Ammar (Devolvé et al., 1980; Pini et al., 1971).

The seismic section has been analyzed using typical concepts of thrust tectonics (Dahlstrom, 1969; Dahlstrom, 1990; De Sitter, 1964; Jamison, 1987) in order to propose a balanced cross-section of the Utique fold, which gave important insights into the fold history. The Utique fold developed reactivating a pre-Miocene structure. Slip transfer is owing to an intermediate duplex developed in the lower stratigraphic package.

The cross-section was further analyzed by stepwise unfolding. This exercise is represented in Fig. 5. After a quiet period marked by Upper Miocene deposition over previously folded Mesozoic strata, the area has been shortened again. It is not possible to know when this shortening stage began, nor it is possible to be sure this shortening stage has continued up to today, even if such a hypothesis is probable regarding to the background local seismicity. During this last deformation stage, deformation is accommodated at depth by the Mellaha formation decollement (Figs. 4 and 5) and meanwhile by fault propagation folding in the uppermost, from Miocene upward, formations. Interestingly, this scenario implies a connection to deeper structures further south. It allows calculating a total shortening since the Mio-Pliocene boundary of about 690 m (0.14 mm/yr on average) in the basal part and of about 210 m in the uppermost formations (0.04 mm/yr). Note that individual formation offset decreases upward, as predicted by the fault propagation geometry (Ramsay and Huber, 1987; Rouvier, 1985; Suppe and Medwedeff, 1990). This implies that at surface, the maximum offset observable is much less than these 690 m. Actually, it must be less than the uppermost offset observed on the cross-section, which corresponds to the 170 m Mio-Pliocene boundary offset.

Geomorphic investigations from Utique anticline were conducted through satellite imagery analysis, and by analysis of a digital elevation model (DEM) of the Utique fold (Fig. 3). The study of this DEM shows that the Utique anticline is trending east–west in the western part, and its trend changes to ENE–WSW in the eastern part. It is about 3 km wide and 200 m high. It is characterized by a steeply dipping forelimb (reaching 45°) and a gently dipping backlimb (not exceeding 30°). Both forelimb and backlimb exhibit a break in slope dividing them into two slope segments separated by a short “flat” with gentle slope (usually used for farming). In both limb the lower slope is shorter than the higher one, both slopes having close dip (Fig. 3). Gullies developed in fold limbs in Quaternary and Upper Pliocene (Porto Farina fm) rocks mantling the fold. In these gullies Quaternary rocks are discordant over the Pliocene ones.

On the south limb of the fold, at least two calcretes horizons are found, attributed to Pleistocene and Holocene, respectively (Oueslati, 1993). These horizons contain calcareous nodules generated around plant roots. Originally, these nodules trend vertically; this is not necessarily the case for the horizon itself, which could have mantled an inclined slope. A careful measurement of calcareous concretion orientation has been carried in two locations on the southeastern limb of the fold (Fig. 6). They indicate a possible slight tilt (less than 5 degrees) with an axis perpendicular to the fold limb. If true, it would imply the tilt has occurred since Pleistocene there.

A fault that can be associated with the fold has been observed along the Tunis-Bizerte motorway at the time of its building, but unfortunately the outcrop has never been described and is no more visible (Ben Ayed, 1986; Dlala, 1995; Dlala and Rebai, 1994). In turn, after careful exploration of the fold, another outcrop was found in a quarry now turned into a dump (37°01.80’N; 10°00.00’E). A fault outcrops in a gully ∼5 m wide and ∼5m deep (Fig. 7). The gully seems to have experienced a complicated story with a first incision stage followed by recent filling and renewed incision that continues currently. This explains some strong differences observed on both sides of the gully; the southeastern side is the most representative and is schematized in Fig. 3 in which were added some observations made in the northwestern side. A fault plane was found, displaying important variations in trend and dip, the average values being a N150°E trend and an 80°N dip, with uncertainties evaluated to be ∼20°. It is accompanied by small faults that display an average N150°E-azimut and 60°N-dip. The fault offsets clearly some conglomerates made of Pliocene marine formation reworking and gray clay levels, observed elsewhere lying on top of calcrete. It is mantled by the recent gully infilling made of angular debris, which could have been affected by the fault (Fig. 7). The major fault and its associated minor faults are affecting, with sometimes offsets observed, the following formations (Fig. 7): some sandstones and conglomerates coming from Pliocene strata erosion (2 on Fig. 7); grey clays (3 on Fig. 7); red clays with calcrete remnants, typical form the Quaternary (Oueslati, 1993) (5 on Fig. 7). A disturbed formation, that can be considered as a fault gouge, made of clay and calcrete-derived sands, contains a piece of pottery, less than 2000 years-old after archeologist expertise (Slim Khosrof, National Institute of Patrimony, Tunis, personal communication).

To sum up, the fault observed appears to have been active very recently, during the Holocene; it has a clear reverse component.

Observations around this gully indicate that the calcrete underwent a total ∼12 m vertical offset (Fig. 7). The calcrete often mantles the fold topography, but it is found faulted in numerous places in the lower fold slope described in the previous section. One may hypothesize that this lower slope corresponds to the total vertical calcrete offset, evaluated to be ∼40 m, corresponding to ∼45 m of fault motion if fault dip is 60°.