The ichnofossil-bearing horizon corresponds to a fill mesosequence comprising 20 m of marls and marly limestones overlain by 5 m of thickening-upwards limestone deposits (Fig. 4B and C). The overlying mesosequence correlates to the topographic point 1622 (Fig. 2) slightly above the ichnofossil-bearing horizon that forms a cornice about ten meters below. North of hill 1622, the horizon is fairly well exposed and consists of borings and a ferruginous encrustation defining a hard ground sequence at the top (Fig. 4G and H). South of hill 1622, next to a N20° trending fault, the top surface of the same horizon is widely exposed (approximately 20 m²) and displays several trace fossils (GPS N 33′27.528′, W 4′39.170′; Fig. 5, Fig. 6 and Fig. 7).

The studied horizon displays a relatively homogeneous microfacies of calcareous grainstone, with a microsparite cement and well-graded calcareous pellets and ooids ranging from 50 to 150 μm in size. Rare bipyramidal quartz grains are the nuclei for some ooids. A pellet-rich microfacies (pelmicrosparite) dominates in the lower part of the sequence, and becomes subordinate in the upper part with respect to the intraclast and bioclast-rich facies (intrabiomicrosparite). Ooids are common in the upper part of the beds and they most often have a micritic cortex (bahamite) encased by a single layer. The ichnofossil-bearing horizon contains different types of fragmented shells (bivalves, brachiopods and echinoderms), benthic foraminifers (miliolids, uniserial arenaces) and rare phosphatic fragments of bone or teeth (Fig. 4F). The associated lithophase is dominated by intraclasts and bahamites lacking spherical oolith laminations (alpha-type ooliths).

Ichnogenus Selenichnites Romano and Whyte, 1987

Ichnospecies Selenichnites tesiltus Gibb, Chatterton and Pemberton, 2011 (Fig. 5)

Type ichnospecies: S. hundalensis (Romano and Whyte, 1987) from the Jurassic Scarborough Formation of Yorkshire, United Kingdom.

Referred material: FSBM 3, 4 and 7, isolated crescent-shaped trace fossils from FSBM locality 1. All specimens are preserved in a concave epirelief form.

Description: Irregularly ridged crescent-shaped concave epirelief trace fossils. Morphological dimensions up to 20 cm long, 14–26 cm wide (Table 1). The length to width ratio is labile and is not be a useful ichnotaxonomic parameter in this case (Table 1). Shallow to deep concavities (depth from 2 to 8 cm). Specimen FSBM 7 (Fig. 5), maximum width (≈ 26 cm) with shallow lunate impression. Depth of specimens varies from shallow to moderately deep, deepest (FSBM 4) measures 21 cm long, 16 cm wide and approximately 6 cm deep, typically slightly widen than long.

Discussion : The trace fossils described herein display the morphological features of the S. tesiltus Gibb et al., 2011, from the Middle Cambrian Azlag Formation of Zagora (Anti-Atlas, Morocco). The significant character strate is the horseshoe simple shape. The type ichnospecies, S. hundalensis, (Romano and Whyte, 1987) displays a relatively flat and subtriangular morphology between two crescent-shaped paired lobes. Hardy (1970) described S. rossendalensisas lunate impressions associated with appendages, or posteriorly projecting morphological imprint and defined lateral edges to anterior and lateral edge of trace fossil. The heart-shaped doublure and telson imprints of S. cordiformis (Fischer, 1978) are not observed in these specimens. S. bradfordensis also displays more complex morphological character states, such as transverse or chevron markings and a defined median furrow. S. scagliai (Poiré and Del Valle, 1996) was discussed and synonymised by Gibb et al. (2011) to the ichnogenus Selenichnites. SG, in Late 2011, had the opportunity to visit the collection in Argentina and now the synonomization is called into question. This assignment will be addressed in a future publication anon. Further ichnospecies identified within the ichnogenus Selenichnites are discussed in Gibb et al. (2011).

Ichnogenus Selenichnites (Romano and Whyte, 1987)

Selenichnites isp. (Fig. 6 and Fig. 7)

Referred material: FSBM 1–2 and 5–6, isolated crescent-shaped invertebrate traces from FSBM locality; all specimens are preserved in concave epirelief.

Description : Small to medium size elongate crescent-shaped (more than half circle) with defined anterior and lateral edges, preserved in concave epirelief measuring 50–70 cm long and 8–38 cm width (Table 1). Ratio of length to width is less than 2 (2 > l/w > 1.4) for the specimens FSBM 1–3 (Table 1). Specimen FSBM 3 represents only the telson trace with very shallow crescentic form. Generally, all the specimens studied here correspond as well to horseshoe-like segments with locally elevated margins and arc of circle-like feature exceeding sometimes the half circle (specimens FSBM1-2; Fig. 6). The depth of the traces varies from 2 to 10 cm, with the specimen FSBM 2 as the deepest and larger trace. The last one shows the maximum length size (≈ 70 cm) with deep lunate imprints and possible telson trace (Fig. 6B). Specimens, FSBM 1-2, could be subdivided in two parts, i.e., the lunate part in the front and the telson imprint behind. Furthermore, the same morphological characters are present in the specimens FSBM 5–6 (Fig. 7), which moderately incise the bedding surface with the smallest length and width, showing sometimes three shallow lobes (FSBM 5) and measuring 8 cm in width. Generally, the invertebrate traces described herein are slightly twice as long as large, except the specimens FSBM 5–6 that we could not estimate their true length. Telson trace mark is seemingly also present in specimen FSBM 5.

Discussion : Specimens FSBM 1–3, 5–6 show the features given in the diagnosis of the ichnogenus Selenichnites (Draganits et al., 2001, Fischer, 1978, Gibb et al., 2011, Hardy, 1970, Romano and Whyte, 1987, Romano and Whyte, 2013 and Wang, 1993). Romano and Whyte (1987) evoked in their diagnosis the presence or absence of prosomal imprint and scratch/telson traces. The lack of several morphological features similarly prevented several other authors from suggesting any definite ichnospecies, i.e., the Selenichnites isp. of Wang (1993), Thomson and Weber (1999), Draganits et al. (2001), Morrissey and Braddy (2004), Lucas and Lerner (2005), Romano and Whyte (2013), and Riahi et al. (2014). Assignment to a distinct ichnospecies is not given here, taking into account that the type species S. bradfordensis (Chisholm, 1985) is basically described as Aulichnites?bradfordensis with short furrow marks. We are aware that the Skoura material might even comprise different ichnospecies. However, considering the fact that the ichnotaxonomy of Selenichnites (Romano and Whyte, 1987) is still problematic, we refrain from determining distinct ichnospecies or naming a new ichnotaxon. Therefore, we herein prefer an open nomenclature referring it to Selenichnites isp.

According to the available research and publications, the possible tracemaker of Selenichnites is commonly attributed to Xiphosurids (Chisholm, 1985, Draganits et al., 2001, Fischer, 1978, Hardy, 1970, Lucas and Lerner, 2005, Riahi et al., 2014, Romano and Whyte, 1987, Romano and Whyte, 1990, Romano and Whyte, 2013 and Wang, 1993) though Trewin and Mc Namara (1995) suggested that S. langridgei could be the behaviour of euthycarcinoids and Weber and Braddy (2004) attributed S. antarcticus to crustaceans. No body fossils are known from the Skoura syncline. Based on morphological character states of Selenichnites and the observed morphology, functional morphology and behaviours of modern limulids, the rounded prosoma and musculature to drive the prosoma into sandy substrate, we suggest limulids are the possible tracemakers for Selenichnites.

The ichnofossil-bearing layer belongs to the Middle Jurassic “Gulf of Skoura” of the South Tethyan margin (Frizon de Lamotte et al., 2008). The combination of sedimentological data and microfacies analyses provides evidence that the deposition environment of the Ich Timellaline/Bou Akrabene Formation was a regressive coastal environment in a relatively warm and low-energy setting that was periodically open sea. The horizon is conformably overlain by marls that contain spines of sea urchins and fragments of tabular i.e. isolated corals, providing further confirmation of a warm fully marine environment. Despite the variable lithology including pellets, intraclasts, bioclasts and ooids, the entire deposit is very well sorted at 50–150 μm. This type of limestone is almost exclusively derived from fine-grained, well sorted carbonate sands, which indicates regular and prolonged wave action apparently undisturbed by storms. The presence of bahamites also provides evidence that the Ich Timellaline/Bou Akrabene Formation formed in a protected environment.

The presence of borings suggests shallow water depth during depositional/early diagenetic processes, as these borings are supported by ferruginous concretions indicating the formation of a hard ground during a period of low sedimentation (Fig. 4G). We note the absence of current ripples on the ichnofossil-bearing surface, while the presence of sinuous ripples atop the two underlying beds indicates subtidal conditions (Fig. 4I). As concluded from sedimentological, palaeontological and palaeoichnological data, a warm humid climate in a shallow marine environment is inferred as the depositional environment of the ichnofossil-bearing layer.

Xiphosurans are marine group of chelicerate arthropods, based on their supposed stem lineage. They are known from the Ordovician to the present day (Martin et al., 2015, Moore et al., 2007, Rudkin et al., 2008, Van Roy et al., 2010 and Van Roy et al., 2015). Van Roy et al., 2010 and Van Roy et al., 2015 and Garassino et al. (2008) reported undetermined xiphosurid specimens from the Early Ordovician and the Late Cretaceous respectively in Morocco. The record of Xiphosurid-like trace fossils from Africa is sparse in comparison to numerous occurrences on other continents with only rare limulid-like trackways from the Late Palaeozoic Great Karoo Basin (South Africa) assigned to the ichnogenus Kouphichnium (Anderson, 1975). Indeed, Selenichnites has only recently been recorded in North Africa from the Middle Cambrian strata of the Anti-Atlas (southern Morocco; Gibb et al., 2011) and from Oligocene–Miocene deposits in northern Tunisia (Riahi et al., 2014). New decapod assemblages were studied from the Late Cretaceous (Cenomanian–Turonian) of Gara Sbaa (Kem Kem region, Anti-Atlas, southeastern Morocco), including the first report of xiphosuran fossils in North Africa (Garassino et al., 2008). The trace fossils discussed in this paper broadens the geographical distribution of possible xiphosurans on the South Tethyan margin.

Selenichnites from the Anti-Atlas Middle Cambrian deposits (Morocco) are associated with the transgression of the Rheic Ocean on the West African Craton margin (Michard et al., 2008 and Nance et al., 2012). Likewise, the Turonian xiphosurids of the Kem Kem region (Morocco) are associated with a global transgression that occurred on the northern African continent. Conversely, the Middle Atlas Selenichnites of Late Bajocian–Early Bathonian age correspond to the end of the Jurassic carbonate platform and the onset of the Middle Jurassic regression (El Mers Fm) prior to the complete emersion of the region (Charrière, 1992). In every case, Selenichnites is found in what is a shallow marine, relatively warm marine environment. Xiphosurid-like ichnofossils are well known in the northern part of the Tethys especially from Ordovician to the present day. However Selenichnites has been interpreted as a burrowing activity of horseshoe crab-like organisms searching for food such polychaete worms and/or soft-shelled molluscs that live within muddy sediments (Gibb et al., 2011, Romano and Whyte, 1987 and Wang, 1993). Four limulid burrowing behaviours were proposed by Eldredge (1970) and Wang (1993) and the Selenichnites would consist of the beginning of the burrowing process.