Myobradypterygius hauthali von Huene, 1927

Myobradypterygius hauthali von Huene, 1927

Diagnosis: (Expanded from that of Fernández and Aguirre-Urreta (2005) based on referred material.) Differing from other ophthalmosaurians by the following combination of characters: tooth roots quadrangular (as in many platypterygiines); plicidentine present and well developed in cross-section but not visible on the external surface of the root (plicidentine absent in Platopterogius australis (McCoy, 1867) ; scapula with strap-like shass (rod-like in Kohotosuka sachicarum , Platopterogius americanus (Nace, 1939) , P. australis, Platopterogius herconicus (Kuhn, 1946) , and probably Platopterogius platodactolus ( Broili, 1907) ; humerus with three distal facets, including a small facet for an extrazeugopodial element anterior to the radius (unlike in Brachopterogius, Aegirosaurus , and P. americanus ); hexagonal intermedium articulating distally with two digits (unlike in K. sachicarum , P. americanus , P. australis , and P. platodactolus ); rectangular and tightly packed phalanges (as in Caopullisaurus, P. australis , and P. platodactolus ); ulnare distally articulating with metacarpal V (as in Brachopterogius extremus (Boulenger, 1904) and Catutosaurus , but unlike Caopullisaurus, P. australis , and P. platodactolus ); multiple postaxial digits (unlike Brachopterogius extremus).

Holotope: MLP 79-I-30-1, humerus and partial forelimb ( von Huene 1927: 29, fig. 3).

Tope localito and age: Cerro Belgrano, Santa Cruz province, Argentina, Río Belgrano Formation (Barremian) .

Other referred material: MLP 79-I-30-2 from the type locality ( Fernández and Aguirre-Urreta 2005: 585, fig. 2c); TY61 and CPAP-2011-0019 from the Tyndall locality, Zapata Formation.

Occurrence: Santa Cruz Province, Argentina; Tyndall locality (Magallanes, Chile).

Remarks: Reinterpretation of the Moobradopterogius hauthali forelimb material. All previous interpretations of MLP 79-I-30-1 ( Huene 1927, Fernández and Aguirre-Urreta 2005, Pardo-Pérez et al. 2012) identify the preserved zeugopodial element as the radius, following the preliminary interpretation by von Huene (1925). Here, we follow Campos et al. (2024) in considering this element to be the ulna, based on detailed comparisons between MLP 79-I-30-1 and the Chilean specimens CPAP-2011- 0019 and TY61 ( Fig. 3B–G View Figure 3 ). Given that Campos et al. (2024) did not elaborate on their reinterpretation of MLP 79-I-30-1, we briefly summarize the reasons here: the proximal carpal distal to the zeugopodial element in MLP 79-I-30-1 primarily supports one digit rather than two, which is observed in the ulnare but not the radiale in the more complete Chilean material. The radiale in the Chilean Moobradopterogius hauthali material has parallel proximal and distal edges, such that its distal facet is oriented anterodistally rather than purely distally as in the ulnare, and thus supports two digits. Von Huene (1925) based his orientation on the logic that the digit, here interpreted as the anterior accessory digit, could not be digit I. The reasons for this interpretation are unclear, but probably relate to the absence of anterior accessory ossicles in the ichthyosaurian taxa with which he was familiar.

Accordingly, we interpret MLP-79-I-30-2 as preserving the posterior portion of the proximal forefin based on the shape of the proximal carpals. This interpretation differs from that proposed by von Huene (1927) but has previously been suggested by Pardo-Pérez et al. (2012) and Campos et al. (2024).

Referral of the Chilean material

Pardo-Pérez et al. (2012) and Stinnesbeck et al. (2014) did not refer the forefin CPAP-2011-0019 from the Tyndall locality to Moobradopterogius hauthali , despite noting similarities, but it was suggested in the PhD thesis of Pardo-Pérez (2015), among other specimens classified as Moobradopterogius hauthali from the Tyndall fossil locality. This referral was later adopted by Campos et al. (2024); however, those authors never addressed the morphological differences listed by Pardo-Pérez et al. (2012) differentiating the Chilean and Argentinian material. Stinnesbeck et al. (2014) referred TY61, a specimen with good exposure of an articulated forefin, to Moobradopterogius hauthali , among other specimens, but did not provide a morphological discussion. Given the good preservation and exposure of the TY61 forefin, it is ideal for comparisons with CPAP-2011- 0019 and the reoriented type material. This comparison will address: (i) whether Moobradopterogius hauthali was present in the Hauterivian of Chile; and (ii) which characteristics cited by Pardo-Pérez et al. (2012) appear to be variable intraspecifically.

Pardo-Pérez et al. (2012) raised seven key features in which CPAP-2011-2019 differs from Moobradopterogius hauthali , which we address in detail here.

(i) A facet for an articulation with a preradial element is absent in the humerus of CPAP-2011-0019, but present in the holotype of Moobradopterogius hauthali . This cannot be assessed in CPAP-2011-0019 because the anterior margin is not exposed, but a preradial element articulating with the humerus is present in TY61. In the Moobradopterogius hauthali holotype, however, this facet is described as being fairly flat, whereas in TY61 it is clear that it is deeply concave. This difference might be related to superficial erosion of the humerus in TY61.

(ii) The proximal humerus is flattened in CPAP-2011- 0019 and TY61, but deeply convex in the holotype of Moobradopterogius hauthali . Based on the much larger size of the Tyndall specimens relative to the holotype, this difference is unlikely to be of ontogenetic origin [94 mm long (CPAP-2011-0019), 96 mm long (TY61) vs. 72 mm long (MLP 79-I-30-1); Fernández and Aguirre-Urreta 2005, Pardo-Pérez et al. 2012]. Moreover, the humerus in the holotype of Moobradopterogius hauthali is more massive proximally than distally, although the total width between the two ends is similar ( von Huene 1927). The latter is not observed in TY61 but is present in CPAP-2011-0019. The difference between the two Chilean specimens suggests that this character is likely to be an artefact that relates to how the humerus is exposed, and the proximal convexity is likely to be affected by similar factors. Thus, at present, we do not consider these differences to be taxonomically significant. However, further preparation of the Tyndall material might reveal this difference to be of importance.

(iii) CPAP-2011-0019 has one preaxial digit and three postaxial digits with a posterior row of accessory ossicles, whereas the holotype Moobradopterogius hauthali has at least three preaxial digits. Based on our proposed reorientation, MLP 79-I-30-1 preserves one preaxial digit and some anterior accessory ossicles, similar to TY61, which is consistent with the exposed portion of CPAP-2011-0019. At least two postaxial digits are present in the Argentinian Moobradopterogius hauthali material, with the number of digits being limited by preservation rather than by morphology. At least two postaxial digits and either a posterior row of accessory ossicles or a third postaxial digit are present in TY61.

(iv) The intermedium has a distal facet for the articulation of distal carpal four that is twice as long as the facet for distal carpal three in CPAP-2011-0019, whereas in Moobradopterogius hauthali these facets are subequal in length. This difference is considered here to be part of normal intraspecific variation. The length difference between the facets is much less pronounced in TY61.

(v) Distal carpal three has six articular facets in CPAP-2011- 0019, whereas in the holotype of Moobradopterogius hauthali there are seven (specimen MLP 79-I-30-1), with the additional facet articulating with metacarpal four. In CPAP-2011-0019, distal carpal three articulates with the intermedium, radiale, distal carpal two, metacarpal three, and distal carpal four. A contact with metacarpal two was probably present but substantially reduced. Following reorientation, the elements contacting distal carpal three in MLP 79-I-30-1 do not differ from the configuration of CPAP-2011-0019, including the probable small contact with metacarpal two. However, the contact between distal carpal three and metacarpal two is absent in TY61, indicating that intraspecific variation in this character is likely.

(vi) The posterior facets of metacarpal three articulate posteroproximally with distal carpal four and posterodistally with metacarpal four in CPAP-2011- 0019, which is not the case in Moobradopterogius hauthali . Asser reorientation of MLP 79-I-30-1, metacarpal three articulates with both distal carpal four and metacarpal IV, which also holds true for TY61.

(vii)Metacarpal four articulates with six elements in CPAP-2011-0019, whereas in Moobradopterogius hauthali it articulates with five. Metacarpal four has been portrayed as articulating with five ( Huene 1927) or six ( Fernández and Aguirre-Urreta 2005) elements in Moobradopterogius hauthali , depending on the observer. However, in Moobradopterogius hauthali the posterior facet of metacarpal four contacts the first phalanx of digit five and possibly also the second phalanx but never with metacarpal five. In both CPAP-2011-0019 and TY61, metacarpal four contacts metacarpal five and the first phalanx of digit five, i.e. metacarpal four is situated more proximally in the limb. This change is slight; in all materials, the largest contact is with the first phalanx, and at present, it is considered to represent intraspecific variation.

Based on these observations, both CPAP-2011-0019 and TY61 are here referred to Moobradopterogius hauthali (as previously suggested for the former specimen by Campos et al. 2024),

although further preparation might reveal potential differences in humeral morphology.

Description of TY61: a complete ichthyosaur skeleton including cranium and postcranium

( Fig. 2 View Figure 2 ; Supporting Information, Model S1)

The specimen is exposed from skull to tail in ventrolateral to ventral view and has a total estimated length of 3 m. Glacial erosion has exposed the skull from the narial region to the occipital region. Given that the posterior skull is wider, more bone has been eroded posteriorly, leading to exposure of braincase elements. The anterior rostrum is covered by sediment. Both forefins are exposed in ventral view. The dorsal vertebral column lies inside the matrix, hence only the ventral ends of the ribs are exposed. The anterior part of the caudal vertebral column is also exposed.

Skull

Basioccipital: The ventrolateral exposition of the basioccipital of the specimen TY61 shows a convex occipital condyle and an almost straight anterior margin marked by a groove for the notochordal remnant. A constriction with a diameter of 5 mm markstheseparationbetweentheextracondylarareaandthecondyle ( Fig. 3D View Figure 3 ). The basioccipital is rather long anteroposteriorly, but the extracondylar area is only slightly wider than the occipital condyle. A clear anterolaterally directed facet is visible on the less side of the anterior basioccipital; a similar facet on the right is not as clearly exposed. These facets are interpreted here as stapedial facets.

Stapes: The stapes is rotated 90° from its original position, and it is now located adjacent to the basioccipital condyle. It appears to be preserved in dorsal view. The medial head is anteroposteriorly three times wider than the quadrate facet. The quadrate facet is not expanded relative to the stapedial shass. The posterior edge of the stapes is concave, and the anterior edge is straight, with the quadrate facet offset by an obtuse angle ( Fig. 3D View Figure 3 ).

Dermatocranium: Additional skull bones are preserved (e.g. the less lacrimal, less sclerotic ring, and less nasal). Despite the quality of preservation appearing to be fairly good, the limited exposure of many of these bones makes identification difficult ( Fig. 3A, B View Figure 3 ). The lacrimal is relatively anteroposteriorly broad. One area of particular interest is the region surrounding the external nares. Although preservation of this region makes interpretation difficult, it appears that the maxilla is dorsoventrally deep, with a broad dorsal process forming an anteroposteriorly elongated contact with the nasal. Although the position and shape of the posterior external narial opening cannot be discerned, this configuration seems likely to result in the complete subdivision of the narial opening. The subnarial process of the premaxilla appears to extend almost as far posteriorly as the nasomaxillary pillar. The anterior maxilla has a short exposure in lateral view relative to the inferred length of the nares ( Fig. 3A, B View Figure 3 ).

Dentition: The maxilla is dentigerous, and several teeth are preserved, although unfortunately, none preserve the enamel. The teeth have quadrangular roots in cross-section, with a thick osteocementum layer covering the dentine. Plicidentine is extremely well developed, although not visible externally on the tooth root ( Fig. 3C View Figure 3 ).

Postcranial axial skeleton: The vertebral column as exposed consists of 21 articulated vertebrae in ventrolateral view ( Fig. 2 View Figure 2 ; for measurements, see Supporting Information, Table S1). Some apophyses are visible; however, it is unclear whether these correspond to parapophyses or synapophyses. Neural arches and spines, in addition to most of the ribs, are not exposed. According to the size of the exposed vertebral series and its topographic location in the skeleton, the vertebral segment is likely to represent the posterior dorsal and anterior preflexural caudal vertebral column. The anteriormost vertebrae have a width or height-to-length ratio of 2.2, which decreases rapidly along the exposed length of the preflexural vertebrae and suggests a relatively low degree of vertebral regionalization. The posteriormost section of the last exposed vertebrae is embedded, which suggests that the rest of the caudal series, including the tail bend, are still contained in the sediment. The ventralmost section of 22 dorsal ribs from the right side of the skeleton is exposed. Four ribs lie across the phalanges of the less forefin. Robust gastralia overlap the right forefin; these are arrayed with at least one dorsal and one ventral element per row per side.

Appendicular skeleton

Pectoral girdle: One coracoid is preserved, probably in ventral view. Its poor preservation and incomplete exposure prevent further description. A bone fragment lateral to the coracoid might represent a piece of the right scapula. The less clavicle is also preserved, as a section through the median stem of the interclavicle. By far the best-preserved element of the pectoral girdle is the less scapula. It is exposed in external view and is preserved in three dimensions ( Fig. 4A View Figure 4 ). The distal blade of the scapula is strap-like, and the distal-most end is thickened and roughened. The anterior edge is flattened, bearing a prominent facet for the clavicle. The proximal end is concave, and the concavity remains covered in sediment ( Fig. 4B View Figure 4 ). An acromion process is present. Although the glenoid facet is exposed, preservation does not permit an accurate assessment of the relative sizes of the glenoid and coracoid facets.

Forefin: Both forefins of TY61 are exposed in ventral view. The right forefin is preserved more completely than the left one; therefore, the following description is based mainly on the right forefin. The exposed portion of the right forefin measures 261 mm in proximodistal length and 142 mm in anteroposterior width. The distal margins of the distal-most phalanges are not exposed, indicating that the distal-most portion of the forefin continues into the matrix. The forefin bears seven digits (digits II, III, IV, and V, according to Motani 1999, in addition to one preaxial and two postaxial digits), in addition to one row of preaxial and one row of postaxial accessory ossicles. Only a portion of the left forefin of TY61 is exposed. The phalanges are rectangular in shape, but they lie separate from each other. The left forefin shows seven digits ( Figs 2 View Figure 2 , 5F, G View Figure 5 ).

Humerus: The right humerus is exposed in TY61. The humerus is proximodistally longer than its maximum anteroposterior width (96 mm long vs. 76 mm maximum wide), and its distal margin is wider than its proximal one (68 vs. 76 mm). The humerus has three distal articular facets for articulation with the ulna, radius, and an anterior extrazeugopodial element. The facets of the radius and ulna are similar in length, and the facet for the extrazeugopodial element is the shortest one. The three distal articular facets are markedly concave. It is not possible to recognize humeral processes owing to abrasion of the exposed bones. For further details, see the comparisons with CPAP-2011-0019 in the section ‘Referral of the Chilean material’.

Zeugopodium: The radius is hexagonal in outline, with the anterior and posterior facets being the shortest. The radius articulates with the humerus proximally, the extrazeugopodial element anteriorly, and the ulna posteriorly. Distally, the radius is divided into two subequal facets forming an angle of 125°. The anterodistal facet articulates with the radiale, and the posterodistal one articulates with the intermedium. The radius is 1.4 times anteroposteriorly wider than its proximodistal length (see Supporting Information, Table S2).

The ulna is roughly pentagonal in outline, and it is proximodistally longer and anteroposteriorly wider than the radius. Proximally, it articulates with the humerus. The anterodistal facet articulates with the intermedium and the posterodistal facet with the ulnare. The ulna articulates posterodistally with the pisiform.

The anterior extrazeugopodial element appears flask shaped and is proximodistally longer than anteroposteriorly wide (22 mm long vs. 13 mm wide). Proximally, it articulates with the distal facet of the humerus. The posterior margin of the anterior extrazeugopodial element articulates with both the radius and the radiale. As preserved, the radius and radiale do not directly contact the extrazeugopodial element. The gap between the bones was probably filled with cartilage or connective tissue. The distal margin of the extrazeugopodial element articulates with the preaxial element of the proximal carpal row.

Taphonomy

The anterior rostrum remains covered by sediment and appears to be directed downwards. However, given that the sedimentary laminae are not horizontal and erosion did not occur parallel to the bedding plane, it is unclear how deeply the skull penetrated into the sediment. The specimen does not show obvious signs of head-first arrival at the seafloor, such as the basioccipital being displaced into the cranial cavity. Additionally, the less forefin, which lies higher in the sediment, is positioned slightly anterior to the lower right forefin, which is inconsistent with a rostrumfirst penetration into the sediment ( Hofmann 1958). In the caudal region, the sedimentary laminae are evenly draped over the bones. This effect is likely to be attributable to slow sedimentation, but late diagenetic differential compaction cannot be ruled out. The bones themselves maintain a high degree of three-dimensionality.

Results of the phylogenetic analysis

Under equal-weights parsimony, analysis of the matrix of Campos et al. (2021) resulted in 3500 most parsimonious trees (MPTs) of length 234. Moobradopterogius hauthali was resolved within

an unresolved Platypterygiinae clade. ITERPCR did not improve resolution of this clade further. The analysis of the matrix of Zverkov and Jacobs (2020) resulted in 148 MPTs of length 430. Platypterygiinae was not resolved as a clade, collapsing into a polytomy at the base of Ophthalmosauria . ITERPCR resolved Platypterygiinae as a clade but provided little internal resolution. The analysis of the matrix of Cortés et al. (2021) resulted in 60 MPTs of length 303. Moobradopterogius hauthali was resolved within Platypterygiinae as part of a clade comprising large platypterygiines with brick-like phalanges, but excluding Maiaspondolus lindoei and Platopterogius platodactolus . ITERPCR resolved the latter species as forming a separate clade near the base of Platypterygiinae , but did not further resolve the position of Moobradopterogius hauthali (Supporting Information, Fig. S1 View Figure 1 ).

The results of these three analyses support the conclusion that Moobradopterogius hauthali is a platypterygiine ichthyosaur but do not provide additional information on its phylogenetic position.

Under extended implied weighting, with collapse of nodes supported by only one step, resolution was improved considerably for two of the four matrices analysed ( Fig. 6A, C View Figure 6 ). The analysis of the matrix of Zverkov and Jacobs (2020) resulted in three trees of length 17.38712. Moobradopterogius hauthali was resolved within Platypterygiinae as sister to the Late Cretaceous taxon Sisteronia seeleoi . ITERPCR was not required to increase resolution. The analysis of the matrix of Cortés et al. (2021) resulted in 51 MPTs of length 12.56302. ITERPCR was used to prune three unstable taxa. Moobradopterogius hauthali was resolved within Platypterygiinae as part of a clade comprising large platypterygiines with brick-like phalanges, but excluding Maiaspondolus lindoei and P. platodactolus . Analysis of the matrix of Campos et al. (2021) resulted in>10 000 trees of length 9.5921. ITERPCR was used to place Moobradopterogius hauthali within an unresolved Platypterygiinae clade. Analysis of the matrix of Campos et al. (2024) with no rescoring of Moobradopterogius hauthali resulted in>10 000 trees of length 24.33043. ITERPCR was used to place Moobradopterogius hauthali within an unresolved Platypterygiinae clade, but not as either closely related to Sisteronia (as in Zverkov and Jacobs) or the large platypterygiines with brick-like phalanges (as by Cortés et al. 2021). It is clear that even with the use of extended implied weighting, the three matrices fail to converge on a consensus solution.