Thaumastosaurus DE STEFANO, 1903

Rage, Jean-Claude, 2016, Frogs (Amphibia, Anura) From The Eocene And Oligocene Of The Phosphorites Du Quercy (France). An Overview, Fossil Imprint 72 (1 - 2), pp. 53-66 : 57-59

Thaumastosaurus DE STEFANO, 1903

Among amphibians from the Phosphorites, the genus

Thaumastosaurus has been the most studied taxon and its history is complicated. Thaumastosaurus was erected by de Stefano (1903a) for his species T. bottii , which is the type species of the genus. De Stefano regarded T. bottii as a lizard and he described it under the heading ‘genus incertae sedis’. The taxon was based on a single specimen from the Phosphorites, i.e. a posterior braincase plus incomplete frontoparietals and otic capsules. The specimen was subsequently lost and de Stefano was most likely the only paleontologist who studied it. Fortunately, the specimen was illustrated ( de Stefano 1903a: pl. IX, figs 11 and 15) and its characteristics clearly appear in the figures. The specimen came from the old collections, therefore neither its precise provenance nor its age are known.

Nopcsa (1908) regarded the name Thaumastosaurus DE STEFANO, 1903 as a homonym of Thaumattosaurus VON MEYER, 1841 (in fact, Nopcsa misspelled the name of Thaumatosaurus VON MEYER, 1841 , a plesiosaur today synonymized with Rhomaleosaurus SEELEY, 1874 ). Nopcsa (1908) coined the replacement name Enigmatosaurus NOPCSA, 1908 , but he still regarded it as a lizard. Piveteau (1927) described a relatively complete anuran skull coming also from the old collections. He associated this skull (now MNHN.F.QU17736) with the dentary of a gekkonid lizard from the old collections and identified the two specimens as Amphignathodon sp. (i.e., an extant genus). The latter genus today is referred to the synonymy of Gastrotheca FITZINGER, 1843 , a South American hemiphractid frog genus that includes the only species of anuran bearing teeth on its lower jaw. On the basis of the original figures, Hoffstetter (1945) showed that the fossil described by de Stefano was an anuran, not a lizard, and he referred to it as Enigmatosaurus bottii (DE STEFANO, 1903) . In addition, he assigned to this species the skull MNHN.F.QU17736 described by Piveteau (1927). Based on the ornamentation of the dermal bones, Hoffstetter suggested that this anuran may be a Pelobatidae . Subsequently, the skull MNHN.F.QU17736 was temporarily lost.

In the 1970s, new excavations in the Phosphorites yielded several specimens belonging to Thaumastosaurus . Surprisingly, as with de Stefano’s lost specimen, they were all represented by the posterior part of a braincase plus otic capsules and fused posterior parts of the frontoparietals. Fossilization of this part of the skull of Thaumastosaurus , which is marked by hyperossification, clearly is favored. Based on this new material, Crochet et al. (1981: tab. 2 -1) reported the taxon (as Enigmatosaurus bottii ) from six (plus perhaps another two) localities of late Eocene age in the Phosphorites. The species then was regarded as a possible Leptodactylidae , an assignment that was not discussed in the article.

The tentative referral to the ‘leptodactylids’, an assemblage now regarded paraphyletic, was briefly discussed by Rage (1981). The assignment was based on the similarity of the skull with that of ‘ceratophryines’ (then included in the ‘leptodactylids’) and on the presence, in the same localities as the skull bones, of fragments of bony plates bearing a pustular ornamentation similar to that of the dorsal shield of ‘ceratophryines’. Thaumastosaurus was therefore regarded as an anuran with South American affinities (‘ceratophryines’ being restricted to South America). Thereafter, such biogeographical affinities were supported by Roček and Lamaud (1995), Rage and Roček (2007), Evans et al. (2008, 2014) and Agnolin (2012). Roček and Lamaud (1995) provided the first detailed description of Thaumastosaurus bottii , based on various skull bones from La Bouffie, a late Eocene (MP 17) locality in the Phosphorites. They also showed that the one-letter difference between Thaumastosaurus DE STEFANO, 1903 and Thaumatosaurus VON MEYER, 1841 (or the misspelled Thaumattosaurus) prevents homonymy, and that the name Enigmatosaurus NOPCSA, 1908 is a junior synonym of Thaumastosaurus DE STEFANO, 1903 . By the early 2000s, the skull MNHN. F.QU17736, formerly described by Piveteau (1927) and temporarily lost, was found in the collections. Rage and Roček (2007) described it, showed that it belonged to Thaumastosaurus and demonstrated that it represents a second species, T. gezei . At that time, the latter species was known only by two specimens from the old collections: the skull (i.e. the holotype) and a squamosal.

Laloy et al. (2013), using tomography, studied the so-called ‘mummy’ of Rana plicata FILHOL, 1876 (i.e., Rana cadurcorum MARTÍN , ALONSO- ZARAZAGA et SANCHIZ, 2012). They also scanned the ‘mummy’ of a forelimb, which might have broken off the main ‘mummy’ according to Filhol (1877). It was not possible to confirm whether the forelimb belongs to the ‘mummy’. The tomographic study revealed that the skull of the ‘mummy’ is identical to that of Thaumastosaurus gezei and, on that basis, the ‘mummy’ was referred to the latter species. Thanks to a large part of the post-cranial skeleton being preserved in the ‘mummy’, much of that region is now reliably known for T. gezei (e.g., Text-fig. 4 : 1, 2). A recent phylogenetic analysis including post-cranial characters demonstrated that Thaumastosaurus does not belong to a South American clade, but that it belongs instead to the ranoid assemblage; more specifically, it appears to be related to pyxicephalids, an endemic African group ( Laloy et al. 2013). This is a good example of how an over-reliance on cranial features related to hyperossification, which can be convergent among unrelated groups of anurans, may adversely affect phylogenetic analyses ( Báez and Gómez 2014, Evans et al. 2014).

New knowledge about part of the post-cranial skeleton now can be used to identify isolated post-cranial bones in the material collected during recent excavations. Unfortunately, post-cranial bones are known only in T. gezei , thus differences with post-cranial elements of T. bottii remain unknown. Therefore, in this paper, all disarticulated post-cranial bones from the Phosphorites similar to those of T. gezei are referred to as Thaumastosaurus sp. Such is the case, for instance, of scapulae (characterized by a marked dorsoventral elongation; Text-fig. 4 : 3) and one humerus (characterized by an articular ball that rather weakly protrudes and is slightly shifted laterally, and by a moderately developed lateral epicondyle; Text-fig. 4 : 4). This is also the case for various cranial bones that cannot be allocated at species level ( Rage and Roček 2007).

It is worth noting that, although a part of the ilium is preserved in the ‘mummy’ of T. gezei , it is not possible to rank this bone among the post-cranial elements of Thaumastosaurus whose morphology is known. Ilia are among the most frequently recovered anuran bones and among the most useful ones for purposes of identification ( Roček et al. 2013, Gómez and Turazzini 2016). Unfortunately, in T. gezei only the anterior extremity of one ilium is known, and it shows only that the shaft bears a medially inclined dorsal crest ( Laloy et al. 2013). Unfortunately, this feature is not sufficient for identification within ranoids; consequently, no isolated ilium from the Phosphorites or elsewhere may be confidently assigned to Thaumastosaurus .

For the time being, in the Phosphorites we must distinguish between Thaumastosaurus bottii (the type species), T. gezei and Thaumastosaurus sp. Outside of the Phosphorites, two species were assigned to Thaumastosaurus : T. wardi and T. sulcatus , described by Holman and Harrison (2002) and Holman and Harrison (2003), respectively. Both non-Phosphorites species were recovered from the late Eocene (MP 17) of southern England.

Figures

Text-fig. 4. Ranoids from the Phosphorites du Quercy. 1: Thaumastosaurus gezei, urostyle in left lateral (a), dorsal (b), and anterior (c) views. 2: T. gezei, left coracoid in medial view (the urostyle and coracoid are parts of the ‘mummy’ MNHN.F.QU17279). 3: Thaumastosaurus sp., left scapula UM-ECA 2537, from the late Eocene of Escamps (MP 19), in lateral (a) and medial (b) views. 4: Thaumastosaurus sp., left humerus UM-MAL 619, from the late Eocene of Malpérié (MP 17), in ventral (a) and medial (b) views. 5: Ranoid indeterminate, small-sized form, right humerus UM-LAV 1280, from the late Eocene of Lavergne (MP 16), in ventral view. 6: Ranoid indeterminate, mid-sized form, left humerus UM-ECA 2558, from the late Eocene of Escamps (MP 19), in ventral view. 7: Ranoid indeterminate, large sized form, left humerus UM-ECA 2557, from the late Eocene of Escamps (MP 19), in ventral (a) and medial (b) views. 8: Ranoid indeterminate, mid-sized form, left scapula UM-ECA 2560, from the late Eocene of Escamps (MP 19), in lateral (a) and medial (b) views. 9: Ranoid indeterminate, large-sized form, left coracoid UM-ECA 2559, from the late Eocene of Escamps (MP 19), in medial view. (1 and 2, redrawn from images segmented by F. Laloy in 2012). Each scale bar = 3 mm.