Mixosaurus cornalianus Bassani, 1886

Renesto, Silvio, Sasso, Cristiano Dal, Fogliazza, Fabio & Ragni, Cinzia, 2020, New findings reveal that the Middle Triassic ichthyosaur Mixosaurus cornalianus is the oldest amniote with a dorsal fin, Acta Palaeontologica Polonica 65 (3), pp. 511-522 : 513-519

publication ID	https://doi.org/ 10.4202/app.00731.2020
persistent identifier	https://treatment.plazi.org/id/03A58B5E-4D6E-EE4A-FCC4-FA865096FCEB
treatment provided by	Felipe
scientific name	Mixosaurus cornalianus Bassani, 1886
status

Treatment

Figs. 3–8 View Fig View Fig View Fig View Fig View Fig .

Material. —Two nearly complete and articulated specimens: BES SC 1000 in right lateral view ( Fig. 3A View Fig ), BES SC 1001 in dorsal view ( Fig. 3B View Fig ). Coded as 20.S288-2.3 and 20.S288-2.4, respectively, in the Inventario Patrimoniale dello Stato (State Heritage Database). From Sasso Caldo site (45°54’03.7” N 8°55’10.6” E, elev. 650 m), Besano, Varese Province, NW Lombardy, N. Italy; Middle Besano Formation (sensu Bindellini et al. 2019), Uppermost Anisian, Nevadites secedensis Zone (sensu Brack et al. 2005), Middle Triassic.

Description.— The premaxilla is caudally pointed, only forming the rostral margin of the external naris; the maxilla shows a high and wide postnarial process contacting the prefrontal, and excluding the lacrimal from the external naris; the latter is very small and slit-like; a long sagittal crest is present on the skull roof, along with a large rostral terrace of the upper temporal fenestra, both reaching the nasals; the nasals are short, not extending caudally beyond the rostral margin of the orbit; the prefrontal and postfrontal are large, each forming about one half of the dorsal margin of the orbit; the upper temporal fenestra is small and slit-like, formed by supratemporal and parietal only; the cheek region is made up of a large supratemporal and a large squamosal. The dentition is subthecodont and isodont; the posterior maxillary and dentary teeth are slender, pointed, and widely spaced; there are 17 maxillary teeth (in BES SC 1000); the maxillary tooth row extends backwards beyond the caudal margin of the postnarial process; the pubis is more than twice the size of the ischium, the neural spines are high and narrow, approximately three times higher than the centra except for the last caudals; the caudal peak occurs at the level of the caudal vertebrae 17–18; the mid caudal vertebrae are larger than the preceding ones; the length of the hindfins is less than two thirds that of the forefins; in the forefins, five metacarpals are present with constricted shafts.

The body size, proportions, and degree of ossification of the two skeletons, compared with a larger mixosaurid sample ( Brinkmann 1996, 1997, 2004), indicate they had almost reached a fully adult age.

Soft tissues: Preservation of vertebrate soft tissues has been reported for several specimens from the Grenzbitumenzone/Besano Formation and the Meride Limestone Formation, such as the tanystropheids Macrocnemus ( Peyer 1937; Renesto and Avanzini 2002) and Tanystropheus ( Renesto 2005; Renesto and Saller 2018), the sauropterygian Neusticosaurus ( Sander 1989) . and the fish Saurichthys ( Renesto and Stockar 2009; Maxwell et al. 2013; Argyriou et al. 2016). Preservation of soft parts is likely due to authigenic mineralization induced by bacterial activity in oxygen-depleted environments, often in fully reducing conditions ( Martill et al. 1992), in the presence of organic matter that may also represent the source of the mineralizing phosphates. Microbial activity creates microenvironments within the decaying carcass ( Noto 2009; Sagemann et al. 1999). Depending on the microbial activity, phosphatization of soft parts may be limited only to some areas of the carcass, according to local variations of the pH ( Briggs and Wilby 1996) which may (or may not) allow precipitation of phosphates that replace integumentary structures or even internal organs (e.g., Dal Sasso and Maganuco 2011).

In both specimens here described, patches of skin are preserved over the body, along with remains of other soft tissue (probably belonging to the stomach) lying inside the rib cage. In BES SC 1000, the upper lobe of the caudal fin is also preserved, along with the trailing edge of the dorsal fin ( Fig. 4 View Fig ). In BES SC 1001, a possible tract of intestine is exposed ( Fig. 5 View Fig ), and traces of the dorsal lobe of the caudal fin are also present. It is the first time that the presence of these structures is reported for Mixosaurus , and their preservation is good enough to allow a description of their internal microstructure.

Dorsal and caudal fins: In BES SC 1000, a well-delimited, straight, elongate, and apparently stiff structure made by tightly packed fibrous elements is positioned above the longest presacral neural spines, together with its basal portions ( Fig. 4B View Fig ). Given its specific identity, shape, position, and size (68 mm from base to its well-preserved apex), we exclude from being a loose flap of decaying tissue, as seen in rarely preserved cutaneous and subcutaneous tissues of Stenopterygius , which have a clearly wavy, not stiffened aspect (CDS personal observation on SMNS 54051, SMNS 81958). Therefore, the structure seen in BES SC 1000 most likely represents the leading edge of the dorsal fin. It lies above dorsal vertebrae 15–23 and at the level of the trailing edge of the forefins, thus it is placed more cranially than in most of the tentative depictions published so far (e.g., Kuhn-Schnyder 1974; Motani et al. 1996). At a closer look, an array of collagen fibres (sensu Lingham-Soliar and Plodowski 2007; that is, each visible fibre may represent a bundle of packed collagen microfibers) densely packed together is clearly visible under binocular microscope ( Fig. 4D View Fig ). These fibres extend parallel to each other, and are inclined at an angle of approximately 26° with respect to the body axis. All fibrous structures are in relief, giving a near three-dimensional appearance; their nature and microstructure is confirmed by SEM analysis (see below).

The dorsal lobe of the caudal fin of BES SC 1000 ( Fig. 4C View Fig ) lies at the level of the peak of the caudal vertebrae, just above caudal vertebrae 14–22, it is almost complete and shows a triangular outline. The preserved portion of the fin is 43 mm high ( Table 1), the cranial margin is 66 mm long, and the ventral margin (i.e., the base) measures 51 mm. The concave ventral margin mirrors exactly the outline of the caudal peak of the vertebral column, showing the original in vivo position of the fin lobe. In fact, the fin has been set off dorsally and slightly rotated counterclockwise, with post-mortem shifting of a few centimetres. The fossilised skin surface that covers most of the fin seems to retain few details, but in areas where the skin is missing, the underlying array of fibres becomes visible ( Fig. 4E, F View Fig ). As in the dorsal fin, these fibrous structures are in relief. At the cranioventral corner of the base of the fin, fibres can be observed diagonally oriented at an angle of approximately 32° with the body axis. Other almost straight, feebly waved arrays of fibres are visible in the central portion and cranial margin of the fin, with an inclination of 58–53°. A small portion of the dorsal lobe of the caudal fin is present also in BES SC 1001, at the level of caudal vertebrae 19–22, but only few patches of skin are preserved.

Under SEM observation of a tissue microsample from the base of the caudal fin of BES SC 1000, each array consists of parallel fibres with a diameter of 60–70 μm each, which are part of the innermost dermal layer. The lack of transverse segmentation and the size range of the fibres are consistent with the collagen microstructure (e.g., Laurent 2018), unlike somatic muscular myofibres, which are striated and have larger diameters (e.g., Muhl et al. 1976). Apparently homologous and better-preserved fibres, sampled from the trailing edge of the dorsal fin of BES SC 1000, are composed of very fine striations, likely corresponding to collagen fibrils ( Fig. 5A–C View Fig ).

SEM analysis of the microsample from the base of the caudal fin also recovered remnants of skin with an excellent degree of preservation, fossilised through replication in authigenic calcium phosphate ( Fig. 5G View Fig ). No scales or scutes are evident, and the external surface is smooth, as in Stenopterygius ( Lindgren et al. 2018) . Moreover, the skin of Mixosaurus shows a multi-layered subsurface architecture ( Fig. 5D View Fig ), corresponding to the laminated epidermis and dermis of modern tetrapods, which is also seen

Fig. 6. Mixosaurid ichthyosaur Mixosaurus cornalianus Bassani, 1886 (BES SC 1001), Sasso Caldo quarry, Besano, Italy, upper Anisian. A. General view → of the specimen. B. Back-scattered SEM element microanalysis of the possible intestine. C. Close up of the abdominal area with the protruding portion of the organ. D. Close up of the possible intestine, showing the longitudinal folds of the organ (left) and the internal granulosity (right). Arrows (in A, C) point to the organ position; asterisk indicates location of the back-scattered electron element microanalysis of the tissue.

in Stenopterygius ( Lindgren et al. 2018) . In our sample the first layer is smooth and continuous, whereas the second layer appears vacuolar or fine-grained. Theoretically there should be a layer of fibrous connective tissue between the right and left dermal layers, and internal to a thin fat layer Erin Maxwell, personal communication 2020), which our surficial non-invasive sampling did not penetrate.

In any case, our SEM images and elemental microanalyses clearly demonstrate that the fibres of BES SC 1000 are fossilised anatomical structures, well in relief within the ichthyosaur skin tissue, and not sedimentary cracks or tool marks, as it was questioned recently for similar finds in Jurassic ichthyosaurs ( Smithwick et al. 2017).

Intestine: A finely corrugated structure of apparent soft tissue nature protrudes off the ventral margin of the abdominal region on the right side of BES SC 1001, approximately at the level of the presacral vertebra 40, i.e., cranial to the pelvic girdle (Fig. 6). It shows surficial corrugations and has a U-shaped tubular structure, it measures 31 mm along its long axis, 17 mm perpendicular to it and 9–11 mm in transverse diameter. Secondary ridges, perpendicular to the longitudinal ones, are visible under grazing light and seem to define millimetric internal subdivisions, or contents.

In our opinion, this structure may represent a tract of the Mixosaurus guts, namely a lithified intestinal segment, filled by remains of ingested prey. Some of the granularities glimpsed within the organ look subrectangular in shape and would be compatible in size with the small vertebrae contained in the stomach region; the transverse diameter of the organ is also consistent with the dimensional range of the intestinal tube in a 0.9–1.0 m long reptile ( Vitt and Caldwell 2009).

The main corrugations would represent the longitudinal folds of the original external layer of muscular tissue, which might have become fossilised thanks to a prolapse outside the abdominal cavity, under direct contact with the sediment undergoing diagenesis. This event might explain why there is no apparent continuity of gut remains cranial and caudal to this section of the organ, within the abdomen of BES SC 1001, where different chemical conditions likely made the decay of soft tissues prevalent.

Fluorescence induced by UV light did not show similar soft tissue remnants nearby; a microsample of the organ, taken along its broken section and analysed under SEM, did not recover any diagnostic anatomical microstructure. Contrary to the skin and collagen fibres, this organ is extensively colonised by a number of pyrite crystals, and in cross-section it appears amorphous. Fossil preservation, in this case, does not reach the cellular level and is limited to replication of the external morphology. Nevertheless, the organic origin of this structure is confirmed by SEM element microanalysis (Fig. 6B), which gives a chemical signal very similar to skin and collagen of BES SC 1000 Fig. 6E, G), and also to the visceral tissue fossilised in the theropod dinosaur Scipionyx ( Dal Sasso and Maganuco 2011: fig. 143).

If our interpretation is correct, this find confirms, after indirect hypotheses based on coprolites (e.g., Pollard 1968) that, in ichthyosaurs, the ingested bones could pass through the intestines. Bones are much harder than scales, and more difficult to process throughout the guts. Not all reptiles are able to do that, as the pyloric valve at the end of the stomach can be too narrow to allow transit of solid material (for a complete account see Dal Sasso and Maganuco 2011). In this respect, BES SC 1001 might represent an important addition to our knowledge of digestive physiology in the Ichthyosauria .

Stomach contents: In the thoracic region of BES SC 1000, several hooklets of cephalopods are detectable ( Fig. 7A View Fig ), as is frequent in Mixosaurus specimens; in addition, a finely rough area of dark material, 100 mm long and 40 mm wide, overlies the left ribs and contains a number of tiny, ebonybrownish polygonal to cylindrical elements, whose diameter varies from 0.1 to 0.8 mm. These elements represent vertebral centra ( Fig. 7B View Fig ), often grouped in clusters and embedded in the rough material, whereas the sediment surrounding the Mixosaurus skeleton is devoid of them; thus it is feasible that this area would represent stomach remains and contents. The attribution of these centra is difficult because they differ significantly from the C-shaped or ring-shaped vertebral centra of Triassic actinopterygians, and are also too small and different from vertebral centra of marine reptiles, while there is some resemblance to some Triassic neoselachian vertebrae ( Mears et al. 2016: fig. 8). Neoselachian remains are known since the Lower Triassic ( Koot et al. 2015), but in the Besano Formation only hybodont sharks have been found so far, thus the attribution to neoselachians must be considered as tentative.

In BES SC 1001, isolated or partially articulated scales ( Fig. 7C, D View Fig ) are preserved in the stomach area. They show a pattern of ridges that suggest they belonged to some actinistian, a group that is well represented in the Besano Formation ( Rieppel 1980, 1985; Cavin et al. 2013; Ferrante et al. 2017; Renesto and Stockar 2018).

Given that all other reported gastric contents of mixosaurids from Monte San Giorgio consist exclusively of cephalopod hooklets ( Rieber 1970; Brinkmann 1997), the two new specimens testify that adult Mixosaurus also preyed on small fishes, as medium-large ichthyosaurs did (e.g., Pollard 1968; Druckenmiller et al. 2014; Dick et al. 2016).