Bawitius, Meunier & Eustache & Dutheil & Cavin, 2016

Meunier, François J., Eustache, René-Paul, Dutheil, Didier & Cavin, Lionel, 2016, Histology of ganoid scales from the early Late Cretaceous of the Kem Kem beds, SE Morocco: systematic and evolutionary implications, Cybium 40 (2), pp. 121-132 : 128-130

Bawitius

Weiler (1935) attributed to Stromerichthys isolated maxillae from Bahariya that Cavin et al. (2015) later referred to a polypterid, possibly Bawitius Grandstaff et al., 2012 . A fragment of a maxilla similar in shape was also described from the Kem Kem beds in Morocco ( Cavin et al., 2015). Schaal (1984) attributed an isolated ectopterygoid from Bahariya to Polypterus? bartheli and Grandstaff et al. (2012) later coined the genus Bawitius for this species. In Bahariya isolated scales similar to those described here were referred to a polypterid ( Stromer, 1925, 1936; Schaal, 1984; Grandstaff et al., 2012). In the Kem Kem beds, Dutheil (1999a) recorded the occurrence of polypterids represented by isolated pinnulae and he described a small species of polypterid, Serenoichthys kemkemensis Dutheil, 1999b from a clay lens called OT1 interpreted as a quiet lake or pool. Cavin et al. (2015) and Benyoucef et al. (2015) referred scales similar to those described here from the Kem Kem beds in Morocco and from coeaval beds in the Bechar basin in Algeria, respectively, to a polypterid, possibly Bawitius . Therefore, the scales described here are referred to Bawitius Grandstaff et al., 2012 .

Contrary to the opinion of Smith et al. (2006), we do not regard as secondary and resulting from a resorption of a primary continuous ganoine layer, the interrupted surface of ganoine in the scales referred here to Bawitius . This interrupted arrangement of ganoine is instead the standard and original state in this genus. This interpretation rests on the absence of resorption traces, which should have been present either in the lateral sides of the ganoine patches or in any stratum at the surface of a pile. On the contrary, the scales of Callipurbeckia tendaguruensis , for example, show typical resorption lacunae (= Howship’s lacunae) on their ganoine surface ( Arratia and Schultze, 1999, fig. 12). Moreover the dentine that fringes the ganoine is also deprived of resorption traces that normally appear as discontinuities in the structure (compare our Fig. 3E to the scales of Lepidotyle in Meunier and Gayet, 1992, figs 9-12). Moreover, no scales showing a regular continuous surface of ganoine associated with the characteristic histological organisation of a “palaeonisciform” fish have been found in the Kem Kem beds, an observation that reinforces our hypothesis of an initial discontinuity rather than a resorption of the ganoine.

In the two palaeoniscoid-types of scales, we observed the presence of a plywood-like organisation of collagen fibres. The regular arrangement of these strata, located between the dentine layer and the bony basal plate, has been recorded in the scales of extant Polypteriformes (see Sire 1989, 1990), as well as in fossil ones ( Gayet and Meunier, 1991; Meunier and Gayet, 1996; Daget et al., 2001). This observation is strengthened by the presence of polarized osteocytes that surround the ascending vascular canals as described in a number of Polypteriformes . This peculiar organisation has been first noticed in several fossil Palaeonisciformes ( Schultze, 1968; Ørvig, 1978), in extant polypterids ( Ørvig, 1957; Sire, 1989, 1990) and then in fossil Polypteriformes ( Meunier and Gayet, 1996) . Our isolated scales from the Tabaste’s collection and from new collection from the Kem Kem beds resemble various other scale material described in the literature, such as the Cenomanian of Bahariya, Egypt ( Stromer, 1925, 1936; Smith et al., 2006), and the lower Cretaceous of Ubangui formation ( Congo) ( Casier, 1961: Pl IV, figs 10, 11).

The studied thick rhomboid “patches” and “ridged” scales from upper Cretaceous Moroccan Sahara, with the three superimposed layers – ganoine, dentine and bony basal plate – are ‘true’ ganoid scales ( Goodrich, 1907; Kerr, 1952; Schultze, 1977; Meunier, 1980; Sire et al., 2009). But the important reduction of the superficial shiny ganoine, form- ing either parallel ridges, with or without ramifications, or thick irregular patches, gives them an original status among the ganoid scales. The bony basal plate of the two series of scales is totally devoid of canaliculi of Williamson. Consequently, this fossil material cannot be allocated to a holostean fish, which are characterized by the occurrence of such canalicles ( Nickerson, 1893; Goodrich, 1913; Ørvig, 1951; Moss, 1964; Schultze, 1966; Francillon-Vieillot et al., 1990; Sire and Meunier, 1994; Meunier, 2011). Accordingly, the attribution of these scales by Tabaste (1963) and Weiler (1935) to semionotids is wrong. In return, the three described layers in types III and IV fit well with a paleoniscoid scale type ( Goodrich, 1907; Schultze, 1968, 1977, 1996; Meunier, 1980; Meunier and Brito, 2004; Sire et al., 2009). These scales, moreover, share characters with the polypteriforms: i) in the basal plate, there are ascending vascular canals surrounded by osteocytes closed to the osteocytic canalicles, which are polarized. Such an organisation has already been described in the scales of the paleoniscid Scanilepis ( Ørvig, 1978) and in the polypteriform Latinopollia ( Meunier and Gayet, 1996, 1998); it is also known in extant Polypteridae ( Meunier and Gayet, 1996; Daget et al., 2001); ii) there is an elasmodine layer ( Schultze, 1996; Sire et al., 2009), a plywood-like organisation of collagenous fibres between the dentine layer and the oldest part of the bony basal plate ( Fig. 5C, G). Such a structure has been first described in the scales of the extant polypterid Polypterus senegalus ( Sire, 1989, 1990). Later this typical elasmodine layer has been also observed in various fossil Polypteridae from the Cretaceous of Africa ( Gayet et al., 1988; Daget et al., 2001; Smith et al., 2006), in Dagetella sudamericana ( Gayet and Meunier, 1992) and Latinopollia suarezi ( Meunier and Gayet, 1996) from the Cretaceous of Bolivia, and in Bartschichthys sp. from Maranhão state of Brazil ( Dutra and Malabarba, 2001). The elasmodine is regarded as an autapomorphy of the Polypteriformes . Consequently, the presence of a plywood-like organisation between the dentine layer and the bony basal plate provides further evidence to include Bawitius into Polypteriformes ( Daget et al., 2001) .

The absence of morphological and histological differences between the structure of large and small scales of types III and IV indicates that both kinds likely belong to the same taxon, possibly Bawitius . Presence of an interrupted ganoine layer in the small scales, i.e. from young individuals, is further evidence that the ganoine is not resorbed during ontogeny.

Acknowledgements. – We are grateful to the Laboratoire de Paléontologie (Département Sciences de la Terre, Muséum national d’Histoire naturelle, Paris) that allowed the first author to study the material collected by N. Tabaste. We thanks Claude Ferrara for the macrophotographs, Vincent Romevaux (MNHN, Paris), Pierre- Alain Proz (MHN, Genève) for their technical assistance for the ground sections, and Sylvain Pont (service SEM-EDS de la Direction des Collections du Muséum national d’Histoire naturelle, Paris) for SEM pictures. This research was partly funded by grants from the Swiss National Science Foundation (SNSF 129923 and 146868). We are also very grateful to the anonymous referees who greatly helped us to improve our manuscript.

Figures

Figure 3. - Scales of Adrianaichthys pankowski (MNHN-Histos 1959). A: Cross section of a scale (transmitted natural light) showing two superimposed layers: pluristratified ganoine (black arrow), and basal plate (bp) constituted of an avascular bony tissue. Dentine is lacking; B-D: Cross sections of scales (transmitted natural light); B: Detail of pluristratified ganoine (g). In the underlying bony basal plate (bp) one can see canaliculi of Williamson (cw) and a growth mark (gm); C: Detail of the bony basal plate showing numerous Sharpey’s fibres and two canaliculi of Williamson (arrow) and a growth mark (gm); D: Detail of the bony basal plate showing a dichotomy on a canaliculus of Williamson (arrowhead). Scale bars:A = 1 mm; B = 100 µm; C = 50 µm; D = 20 µm).