Research Article
Print
Research Article
A new Eurhinosaurus (Ichthyosauria) species from the Lower Jurassic (Toarcian) of Mistelgau (Bavaria, Southern Germany)
expand article infoGaël E. Spicher§|, Feiko Miedema, Jelle Heijne|, Nicole Klein|
‡ JURASSICA Museum, Porrentruy, Switzerland
§ University of Fribourg, Fribourg, Switzerland
| Rheinische Friedrich-Wilhelms-Universität Bonn, Bonn, Germany
¶ Naturkundemuseum Bamberg, Bamberg, Germany
Open Access

Abstract

Eurhinosaurus is a European Lower Jurassic longirostrine ichthyosaur, characterized by its remarkable overbite. Despite the long history of the genus, the taxonomy of Eurhinosaurus is still under debate, and its morphology is poorly understood. Over the past two decades, three specimens of Eurhinosaurus have been discovered in the Mistelgau clay pit in Bavaria, southern Germany, from Upper Toarcian layers. This makes these specimens the youngest stratigraphic occurrence of the genus Eurhinosaurus. The examined specimens include two nearly complete skeletons and a partial snout, preserved three-dimensionally in a semi-articulated state, with elements exposed in multiple orientations. The Mistelgau specimens exhibit clear morphological similarities to known Eurhinosaurus, confirming their affiliation to the genus based on numerous characteristics. However, the Eurhinosaurus specimens from Mistelgau exhibit notable differences in the basioccipital and rib morphology compared to known species. While not significant at the genus level, these distinctions allow recognition as a new species: Eurhinosaurus mistelgauensis sp. nov. These well-preserved fossils from Mistelgau further provide valuable insights into Eurhinosaurus morphology and significantly contribute to our understanding of this historically important ichthyosaur.

Key Words

3D Preservation, pathology, parvipelvia, skull morphology, taphonomy, taxonomy

Introduction

Ichthyosauria is a diverse group of reptiles well-adapted to marine life that thrived from the Early Triassic to the early Late Cretaceous (e.g., Sander 2000; McGowan and Motani 2003; Motani 2005; Fischer et al. 2016). Ichthyosaurs exhibit morphological features such as limbs modified into fins, elongated snouts, deeply amphicoelous vertebrae, and the development of a tail fin that shows that they were fully adapted to aquatic life (e.g., Sander 2000; McGowan and Motani 2003; Motani 2005). Ichthyosaurs had a cosmopolitan distribution during their entire existence (Sander 2000). Southern Germany, especially, is famous for its extensive record of fossil ichthyosaurs from the Lower Jurassic, mainly the Posidonienschiefer Formation, which has yielded thousands of specimens (Urlichs et al. 1994). The Lower Jurassic ichthyosaurs from southern Germany show a great diversity, with numerous genera and species recognized. At least five established genera and one more recently described genus are currently identified: Temnodontosaurus, Suevoleviathan, Stenopterygius, Hauffiopteryx, Eurhinosaurus, and Magnipterygius (von Huene 1922; Swinton 1930; Maisch 1998a, 1998b, 2001, 2008; Maxwell 2012, 2018; Maxwell and Cortés 2020; Maisch 2020, 2022; Maisch and Matzke 2022). Among these Lower Jurassic ichthyosaurs, Eurhinosaurus is a relatively rare faunal component (Hauff 1921; von Huene 1922; Maxwell and Vincent 2016). This genus is notable for its distinctive elongated and slender snout with an extreme overbite by having the mandible drastically shorter than the upper jaw (more than 60% according to McGowan and Motani 2003; see also, e.g., Abel 1909; McGowan 1994; Maisch 2022). The biology of Eurhinosaurus is poorly understood and discussed controversially. According to McGowan (1979) and Sander (2000), Eurhinosaurus, with its snout resembling the extant swordfish Xiphias that also shows a similar extensive overbite, might have fed in a similar way to this fish by preying on schools of small fish. The teeth of Eurhinosaurus are conical, slender, and pointed, and correspond to the “Pierce 1” predatory guild, which implies predation on soft prey (Massare 1987).

Eurhinosaurus is a representative of the clade Leptonectidae (Maisch 1998b; McGowan and Motani 2003; sensu Moon 2017), a group that has elongated and slender limbs, a long and slender snout and a skull with very large orbits and a short postorbital portion (Maisch 1998b). According to recent studies (e.g., Moon 2017; Maxwell and Cortés 2020), the clade Leptonectidae includes only a few genera: Eurhinosaurus Abel, 1909, Excalibosaurus McGowan, 1986, Leptonectes McGowan, 1996, and Wahlisaurus Lomax, 2016. Recent phylogenetic studies have also retrieved Temnodontosaurus azerguensis Martin et al., 2012 as a representative of the clade Leptonectidae or closer related to species classically interpreted as leptonectid when the clade Leptonectidae was not recovered (Moon 2017; Maxwell and Cortés 2020; Laboury et al. 2022).

Despite being a historical genus known since the mid-19th century (e.g., Mantell 1851; von Theodori 1854; von Jaeger 1856), recent detailed morphological descriptions of Eurhinosaurus specimens are limited, as well as our understanding of the morphological variation (von Huene 1951; McGowan 1979; Maisch 2022). Additionally, Eurhinosaurus suffers from problematic and unstable taxonomy (e.g., McGowan 1994; Maisch 2022). Moreover, the lack of detailed bone-by-bone descriptions and studies assessing morphological variation in detail among Eurhinosaurus specimens leads to the current misunderstanding regarding the diversity within this genus.

Eurhinosaurus is so far restricted to the Toarcian (Lower Jurassic) (McGowan and Motani 2003). Its paleogeographic distribution is restricted to what is today Europe, with fossil evidence from Germany (von Theodori 1854; von Jaeger 1856; von Huene 1922, 1926, 1928, 1931a, 1951; Hauff et al. 2017; Kosma 2018; Maisch 2022), the United Kingdom (Mantell 1851; McGowan 1994), France (Lamaud 1979; Pharisat et al. 1993; Fischer et al. 2011), Luxembourg (Godefroit 1994), and Switzerland (Reisdorf et al. 2011; Klug et al. 2024).

Taxonomic history

The taxonomic history of Eurhinosaurus was already extensively reported and discussed in McGowan (1994) and Maisch (2022). Therefore, we herein only briefly review the history of the three species of Eurhinosaurus currently discussed as being valid by different authors. After a revision by McGowan (1994), Eurhinosaurus was considered monospecific with E. longirostris being the only valid species (see also McGowan and Motani 2003) despite a certain observed morphological variability (von Huene 1931a, 1951; McGowan 1994; McGowan and Motani 2003; Maisch 2022). A recent study by Maisch (2022), however, questioned the monospecific status of E. longirostris, attributing this doubt to the poor preservation of the holotype from Whitby (partial skull, vertebral column, and forefin) and its lack of a detailed description, although the latter is not a sufficient basis for questioning the name’s validity. Maisch (2022) claimed that the Eurhinosaurus specimens found in Germany cannot be categorically associated with E. longirostris just on the basis of having a strong overbite because this feature is generic and only allows affiliation with the genus Eurhinosaurus and not with the species E. longirostris. Maisch (2022), therefore, proposed that E. longirostris must be considered a nomen dubium.

The earliest valid name for a species of Eurhinosaurus exclusively from Germany is E. huenei Swinton, 1930 based on a complete specimen from Holzmaden (NHMUK PV R 5465). However, this species was later invalidated by von Huene (1931a) and then by McGowan (1994), claiming that there were no notable differences between the proposed holotype of E. huenei and other known complete specimens of E. longirostris from Germany, except for the number of notched fin elements, which was considered an individual anomaly. Maisch (2022) agrees with McGowan’s (1994) conclusions regarding the validity of the characters of Eurhinosaurus huenei brought forward by Swinton (1930) but disagrees with McGowan (1994) by still considering Eurhinosaurus huenei valid.

The problem remains that the most recent diagnosis for E. longirostris and E. huenei is based on the same specimens and, if Eurhinosaurus should not be monotypic as suggested by Maisch (2022), that the species diagnosis of E. longirostris is not separated from the diagnosis of the genus. Maisch (2022) argued that a detailed description of the Whitby holotype of E. longirostris is needed to gain more evidence regarding the taxonomic relationship between the English and German specimens. Further on, Maisch (2022) erected a new species in the same work where he recognizes Eurhinosaurus huenei as valid and declares E. longirostris invalid. This new species, E. quenstedti Maisch, 2022, is based on a specimen from the Lower Toarcian of Ohmden in southern Germany, discovered in 1999, which Maisch (2022) found to be distinct from E. huenei, primarily based on differences in forelimb proportions, the number of presacral vertebrae, and the morphology of the neural spines in the mid-trunk region (see Maisch 2022 for the detailed diagnosis).

Thus, at present, we recognize three species in the literature with their state of validity depending on the respective authors: E. longirostris Mantell, 1851, E. huenei Swinton, 1930, and E. quenstedti Maisch, 2022. Whether German Eurhinosaurus specimens, aside from the two affiliated with Eurhinosaurus quenstedti, should be assigned to Eurhinosaurus longirostris or Eurhinosaurus huenei remains an open question until a comprehensive revision of the material from Baden-Württemberg is conducted. Additionally, the validity of Eurhinosaurus huenei and E. quenstedti remains uncertain when based on the most recent diagnoses by Maisch (2022). Many of the diagnostic features proposed by Maisch (2022) can be attributed to taphonomic deformation, particularly cranial compression caused by headfirst impact with the seafloor, a process documented in several similarly preserved ichthyosaur skeletons (e.g., Hänggi and Reisdorf 2007; Maxwell et al. 2022; Erin Maxwell, pers. comm. 01/2025). Additionally, preservation and taphonomic biases may account for some of the listed characteristics, such as characters based on deformed limbs or vertebral count-based characters, which are unreliable when the vertebral column is disarticulated (e.g., E. quenstedti diagnosis in Maisch 2022). Due to this taxonomic confusion and the lack of adequate diagnoses for genus and species, detailed descriptions of Eurhinosaurus specimens are essential for clarifying the genus’ taxonomic diversity and understanding the influence of intra- and interspecific morphological as well as ontogenetic variation. Due to their exceptional 3D preservation and the fact that they were exposed in their original position with different cranial and postcranial elements exposed in various views, the herein described specimens are particularly important, considering the above-mentioned aspects.

Therefore, the objective of this study is to provide a comprehensive bone-by-bone description of the Eurhinosaurus specimens from Mistelgau, thereby enhancing our understanding of the morphology of Eurhinosaurus. Furthermore, this study aims to discuss the taxonomical affiliation of the Mistelgau specimens when considering the above-discussed taxonomical issues.

Institutional abbreviations: UMO, Urwelt-Museum Oberfranken, Bayreuth, Germany; BRSMG, Bristol City Museum and Art Gallery, United-Kingdom; GPIT, Paläontologische Sammlung der Universität Tübingen, Germany; MAMSPLP, Musée des Amis de la Mine in Saint-Pierre La Palud, Rhône department, France; NHMUK, Natural History Museum, London, United Kingdom; SMNS, Staatliches Museum für Naturkunde Stuttgart, Germany.

Geological and stratigraphical setting

The studied specimens were excavated in the Mistelgau clay pit of Jurassic age in Mistelgau, Bavaria, southern Germany (Fig. 1). The coordinates of the clay pit are 49°54'16"N, 11°27'54"E (Schulbert 2013). The stratigraphy and geology of the Mistelgau clay pit have been thoroughly documented (Schulbert 2001, 2013), while the Jurensismergel Formation, the primary geological unit of interest, has been comprehensively studied by Arp et al. (2021). The stratigraphic sequence of the Mistelgau clay pit spans from the Lower Jurassic to the Middle Jurassic, including strata from the Pliensbachian, Toarcian, and Aalenian (Schulbert 2011, 2013; Arp et al. 2021) (Fig. 2A). However, due to past exploitation and backfilling, parts of the sequence, particularly the older Pliensbachian layers, are no longer accessible (Schulbert 2013). These Pliensbachian deposits were represented by the Amaltheenton Formation, characterized by clay sediments and calcareous concretions with high pyrite content near the boundary with the Posidonienschiefer Formation (Schulbert 2001, 2013). The currently exposed sequence in the Mistelgau clay pit extends from the upper part of the Lower Toarcian to the lower part of the Aalenian (Schulbert 2001, 2013). The Lower Toarcian is represented by the Posidonienschiefer Formation (Fig. 2A, B), which consists mainly of finely laminated bituminous clay and calcareous marl layers (Simonsen 2013). The Upper Toarcian is dominated by the Jurensismergel Formation, while the Lower Aalenian comprises the Opalinuston Formation (Schulbert 2001, 2013). No sharp lithological boundary is observed between the Jurensismergel and Opalinuston formations, as both are primarily composed of marl and clay. However, they can be distinguished by sediment color: the Jurensismergel Formation exhibits bluish to grayish hues, whereas the Opalinuston Formation appears more yellowish to brownish (Schulbert 2001, 2013). The boundary between the Lower and Upper Toarcian does not coincide with the boundary between the Posidonienschiefer Formation and the Jurensismergel Formation but instead lies between the Bifrons and Variabilis ammonite zones (Fig. 2A, B) (Arp et al. 2021). Most of the Variabilis Zone, including the Illustris, and Vitiosa Subzones, belongs to the Jurensismergel Formation. However, the basal Variabilis Zone, corresponding to the Variabilis Subzone, remains part of the Posidonienschiefer Formation (Fig. 2B) (Arp et al. 2021). In Mistelgau, the Variabilis Zone forms a highly condensed layer of only a few centimeters, marking the transition between the Bifrons Zone (Posidonienschiefer Formation) and the Thouarsense Zone (Jurensismergel Formation) (Schulbert 2001, 2013; Arp et al. 2021). The Variabilis Subzone consists of a layer consisting mainly bituminous calcareous marl sediments (Fig. 2A, C), while the overlying condensed layer of calcareous marl contains a belemnite battlefield (characteristic of the Vitiosa Subzone) and phosphatic nodules (characteristic of the Illustris Subzone) (Schulbert 2001, 2003; Arp et al. 2021).

Figure 1. 

Locality of the Eurhinosaurus specimens from Mistelgau. A. Map showing the location of Bavaria in Germany; B. Close-up of Bavaria showing the location of the Mistelgau locality.

Figure 2. 

Stratigraphy of the Mistelgau locality. A. Lithographic section of the Mistelgau locality (modified after Arp et al. 2021). B. Stratigraphic chart depicting the ammonite zones of northwestern Europe and the corresponding ammonite subzones of southwestern Germany, which are applicable to the Mistelgau locality in southeastern Germany, along with the associated geological formations (modified after Arp et al. 2021). The two subzones in which the Eurhinosaurus of Mistelgau were unearthed are highlighted in yellow. C. Close-up of the lithographic section of the Mistelgau locality showing the exact stratigraphic position of the Eurhinosaurus specimens (modified after Arp et al. 2021). The red line corresponds to the maximum regression surface and the dashed blue line to the maximum transgression surface (after Arp et al. 2021).

All three Eurhinosaurus specimens described in this study were found in the lower part of the Jurensismergel Formation (Fig. 2B). UMO BT 011 221.00 and UMO BT 011 240.00 were unearthed from the marly facies of the Thouarsense ammonite Subzone (Fig. 2C). UMO BT 011 235.00 is stratigraphically slightly older and was discovered directly on the top of the belemnite battlefield of the Vitiosa Subzone from the upper Variabilis Zone (Fig. 2C). Thus, the specimens studied here are of Upper Toarcian age (Fig. 2). Moreover, the two specimens (UMO BT 011 221.00 and UMO BT 011 240.00) that are from the Grammoceras thouarsense ammonite Zone and Subzone, represent the youngest stratigraphical occurrence of the genus Eurhinosaurus.

The Mistelgau vertebrate fauna

Since 1999, in addition to the Eurhinosaurus described in this study, numerous marine vertebrate fossils have been discovered during excavations organized by the Urwelt-Museum Oberfranken, such as other ichthyosaurs like Stenopterygius and Temnodontosaurus, as well as other marine vertebrates such as a so far undescribed teleosauroid, the plesiosaur Franconiasaurus, and several fishes (Rabold and Eggmaier 2013; Pardo-Pérez et al. 2018; Sachs et al. 2024).

Material

The three specimens described here originate from the Mistelgau clay pit, southwest of Bayreuth in northern Bavaria, Germany (Fig. 1A, B). The specimens were excavated by the team of the Urwelt-Museum Oberfranken (Bayreuth) and prepared by the museum’s preparator Stefan Eggmaier. All three specimens are part of the collection of the Urwelt-Museum Oberfranken. Measurements of the Mistelgau specimens can be found in Suppl. material 1: table S1.

UMO BT 011 221.00 was excavated in several slabs in 2007 (Rabold and Eggmaier 2013). The specimen is a nearly complete, semi-articulated skeleton (Fig. 3; Suppl. material 3: fig. S1). Numerous cranial elements are preserved in three dimensions but are disarticulated and exposed in various views (Fig. 3, 6, 8A; Suppl. material 3: figs S2–S4). The skull roof is mostly missing, and the palatal bones are absent. The vertebral column is disarticulated, with numerous isolated centra and neural arches distributed across multiple slabs (Fig. 3; Suppl. material 3: figs S1, S5). Additional material found in its vicinity was assigned to UMO BT 011 221.00 (Suppl. material 3: fig. S1), such as few post-flexural centra from the tail fin which are preserved in semi-articulation on separate slabs near the main skeleton. Dorsal ribs are largely preserved near their anatomical position, while cervical and caudal ribs are widely scattered (Fig. 3; Suppl. material 3: figs S6–S8). The specimen includes pectoral girdle elements preserved in semi-articulation, as well as pelvic girdle elements preserved disarticulated (Figs 3, 1115). The forelimbs are well-preserved in semi-articulation, while the hindlimbs are less well-preserved with numerous elements missing and strong disarticulation (Figs 3, 13, 15).

Figure 3. 

Holotype specimen of Eurhinosaurus mistelgauensis sp. nov. (UMO BT 011 221.00). A. Photograph; B. Interpretative line drawing. Abbreviations: fc, first caudal centrum (number 51, first centrum with fused rib facets).

UMO BT 011 235.00 was unearthed in numerous slabs in 2014 (Stefan Eggmaier, pers. comm. 01/2024). It is a semi-articulated skeleton (Fig. 4). The skull is laterally compressed, with the right side articulated and exposed in internal view (Figs 4, 7). The nasal opening is preserved, but the anterior margin is missing. Its rounded shape is surely not representing the original shape of the narial opening, as it is oval to subtriangular (pointing anteriorly) in better preserved Eurhinosaurus skulls (e.g. von Huene 1928, 1951; Maisch 2022) and thus might represent an artefact of preservation. Broken bony material is visible within the nasal opening (Fig. 7). Skull roof and palatal elements are either missing or too fragmentary to identify. Some mandibular bones and a few bones from the left side of the skull are preserved but scattered. The vertebral column is disarticulated, with numerous centra and neural arches scattered across the slabs. However, the centra forming the distal-most portion of the tail fin are well-articulated (Fig. 4; Suppl. material 3: fig. S9). Dorsal ribs are largely preserved near their anatomical position, while cervical and caudal ribs are widely dispersed (Fig. 4; Suppl. material 3: fig. S8). Pectoral and pelvic girdle elements, as well as fore- and hindlimb elements, are preserved in various views and are completely disarticulated (Figs 4, 1115).

Figure 4. 

Specimen UMO BT 011 235.00 from Mistelgau. A. Photograph; B. Interpretative line drawing. Note that UMO BT 011 235.00 lays on a belemnite battlefield. Abbreviations: at/ax, atlas/axis complex; cl, clavicle; co, coracoid; f, femur; h, humerus; il, ilium; is, ischium; la, lacrimal; pb, pubis; sc, scapula.

UMO BT 011 240.00 was discovered in 2002 (Rabold and Eggmaier 2013). It is the least complete of the three specimens, consisting solely of both premaxillae with numerous teeth still in articulation with the jaw bones, as well as associated rib fragments (Fig. 5).

Figure 5. 

Specimen UMO BT 011 240.00 from Mistelgau. A. Photograph; B. Interpretative line drawing. The area highlighted by the dashed line visible on the posterior portion of the right premaxilla in medial view indicates the nasal facet. Abbreviations: nf, nasal facet; pmx, premaxilla.

Note that for clarity, the description of the specimens is structured bone by bone, while the taxonomic assessment is discussed in detail later.

Methods

Specimen measurements were obtained using calipers and are provided in Suppl. material 1: table S1. Statistical analyses, including the linear regression, were conducted using the R statistical environment (version 4.4.2) (R Core Team 2024) with the R package ggplot2 (Wickham 2016).

To estimate body size, Scheyer et al. (2014, fig. 3) provided a linear regression comparing humerus length and body length in Triassic ichthyosaurs. In this study, we apply a similar linear regression analysis to estimate the body sizes of specimens UMO BT 011 235.00 and UMO BT 011 221.00, using body length and humerus length measurements from a selection of complete Eurhinosaurus specimens reported in the literature (see Suppl. material 2: table S2 for specimen details and measurements). Specimen UMO BT 011 240.00 lacks preserved humeri and thus could not be included in the size estimation. The results of these estimations, along with additional details (e.g., standard error), are provided in Suppl. material 2: table S3.

FM, JH, and GES examined specimens firsthand relevant for comparison in the following collections: Urwelt-Museum Oberfranken, Bayreuth, Germany; Paläontologische Sammlung der Universität Tübingen, Germany; Natural History Museum, London, United Kingdom; Staatliches Museum für Naturkunde Stuttgart, Germany.

Systematic paleontology

ICHTHYOSAURIA de Blainville, 1835

PARVIPELVIA Motani, 1999

Eurhinosaurus Abel, 1909

Type species and holotype.

Eurhinosaurus longirostris (Mantell, 1851), NHMUK PV OR 14566.

Revised diagnosis of the genus.

Mandible considerably shorter than skull, <60 percent of skull length; snout, premaxillary, and prenarial ratios each >1.0; orbital ratio ≥0.20. Unpaired carotid foramen in the parabasisphenoid*. Pelvic girdle is tripartite, without fusion of pubis and ischium. Long-bodied, preflexural vertebrae >85, but probably not >95; presacral vertebrae probably >44. Fore- and hindfins both long and slender; number of elements in longest digit of forefin >17; forefin at least half as long as skull; hindfin well developed, approaching length of forefin. Forefin with 4 or 5 digits (5th digit might be an additional postaxial digit). Presence of sheathed-bicapitate dorsal ribs*. Presence of ossified haemal arches in the anterior caudal vertebrae*. Large adult body size (up to 7 meters in the largest individuals).

Remarks.

Characters marked with an asterisk (*) indicate newly added diagnostic features. Maisch (2022), in his description of Eurhinosaurus quenstedti and validation of Eurhinosaurus huenei, did not provide a diagnosis for the genus Eurhinosaurus. Hence, the most recent diagnosis for this genus is that from McGowan and Motani (2003), who indicated that the diagnosis for Eurhinosaurus longirostris (monospecific at the time of publication) is the same as for the genus. Accordingly, this diagnosis remains the most recent diagnosis for the species Eurhinosaurus longirostris.

This revised diagnosis for the genus is largely based on the diagnosis provided by McGowan and Motani (2003, pages 81–83). However, Maisch and Matzke (2000) originally described the unpaired carotid foramen in the parabasisphenoid and the dorsal ribs that are not clearly double-headed as diagnostic traits of Eurhinosaurus longirostris (considered monotypic at the time of publication). We here add these features considered characteristic of Eurhinosaurus (highlighted by an *) to the most recent diagnosis established by McGowan and Motani (2003). Note that the not clearly double-headed condition of the dorsal ribs mentioned by Maisch (2022) is recognized in this study to be highly similar to the condition described in Argovisaurus (Miedema et al. 2024) (see description). Miedema et al. (2024) described this specific condition as “sheathed-bicapitate” and we therefore apply this same terminology to Eurhinosaurus.

Although notching in fin elements of the leading digit is part of the diagnosis for the genus Eurhinosaurus provided by McGowan and Motani (2003) and also frequently described in other ichthyosaurs (e.g., Maxwell et al. 2014; Lomax et al. 2017; Anderson 2019) and notched elements are present in the specimens of this study (Figs 3, 13) the presence or absence of notching as well as the number of notched elements (when present) is highly variable in Eurhinosaurus and this variability has already been reported in previous studies (von Huene 1951; McGowan 1994; McGowan and Motani 2003; Maisch 2022). Furthermore, von Huene (1951) hypothesized that the variation in number of notched elements in the limbs of Eurhinosaurus could possibly be explained by sexual dimorphism. Therefore, due to the highly variable nature of this character and the uncertainty regarding its significance, we do not consider the characters that regard the notching of fin elements to be truly informative for diagnosing Eurhinosaurus and they are here excluded from the revised diagnosis. It should be noted that the basioccipital morphological traits identified by McGowan and Motani (2003) have been excluded from the revised generic diagnosis, as our study indicates that basioccipital morphology holds specific significance rather than generic.

As originally reported by von Huene (1926), Eurhinosaurus shows the characteristic presence of ossified haemal arches in the anterior caudal vertebrae. This observation is confirmed by the specimens from Mistelgau and we therefore added this characteristic condition to the generic diagnosis.

Type locality and horizon.

Lower Jurassic (Lower Toarcian), Whitby, Yorkshire, England (Mantell 1851; McGowan 1994).

Distribution.

Holzmaden, Ohmden, Bad Boll, Aalen, Dotternhausen, Schömberg, in Baden-Wuerttemberg (von Jaeger 1856; von Huene 1922, 1926, 1928, 1931a, 1951; Maisch 2022), Banz (von Theodori 1854; von Huene 1922), Mistelgau (this study), in Bavaria, Schandelah and Hondelage, in Lower Saxony (Hauff et al. 2017; Kosma 2018), Germany; Whitby (Mantell 1851; McGowan 1994), Yorkshire, in the United Kingdom; Dudelange, in Esch-sur-Alzette (Godefroit 1994), Luxembourg; Pic-Saint-Loup, in Hérault (Lamaud 1979), Noirefontaine, in Franche-Comté (Pharisat et al. 1993), Marcoux, in Alpes-de-Haute-Provence (Fischer et al. 2011), France; Staffelegg, in Aargau, Switzerland (Reisdorf et al. 2011; Klug et al. 2024).

Stratigraphic distribution.

Toarcian (Lower Jurassic).

Eurhinosaurus mistelgauensis sp. nov.

Etymology.

The specific epithet is derived from Mistelgau, the locality in Bavaria where the specimens were discovered.

Holotype.

UMO BT 011 221.00, an almost complete skeleton (Figs 3, 6).

Figure 6. 

Close-up on the skull of UMO BT 011 221.00. A. Photograph; B. Interpretative line drawing. Abbreviations: ang, angular; ar, articular; at/ax, atlas/axis complex; atna, neural arch of the atlas; axrb, rib of the axis; boc, basioccipital; cb1, ceratobranchial I; cp, cultriform process; cr, cervical rib; den, dentary; epi, epipterygoid; ex, exoccipital; hc, hyoid corpus; j, jugal; mx, maxilla; op, opisthotic; pbs, parabasisphenoid; pmx, premaxilla; pre, prearticular; prf, prefrontal; pro, prootic; q, quadrate; qj, quadratojugal; sp, splenial; sq, squamosal; sta, stapes; sur, surangular.

Paratype.

UMO BT 011 235.00, an almost complete skeleton (Figs 4, 7).

Figure 7. 

Close-up on the skull of UMO BT 011 235.00. A. Photograph; B. Interpretative line drawing of the cranial bones. The dashed lines show uncertain sutures. Note that the suture between the right premaxilla and right maxilla is not visible. Abbreviations: ang, angular; boc, basioccipital; j, jugal; la, lacrimal; mx, maxilla; na, nasal; no, narial opening; pa, parietal; pmx, premaxilla; po, postorbital; pof, postfrontal; pre, prearticular; prf, prefrontal; pt, pterygoid; scp, sclerotic plate; sp, splenial; sq, squamosal; sta, stapes; sur, surangular.

Referred specimens.

UMO BT 011 240.00, a partial snout (both premaxillae) with preserved articulated teeth and rib fragments (Fig. 5).

Locality.

Mistelgau, Bavaria, Germany (Fig. 1).

Stratigraphic horizon.

Jurensismergel Formation (Upper Toarcian), Haugia variabilis Zone, Vitiosa Subzone (UMO BT 011 235.00), and Jurensismergel Formation (Upper Toarcian), Grammoceras thouarsense Zone and Subzone (UMO BT 011 221.00 and UMO BT 011 240.00) (Fig. 2).

Diagnosis.

Eurhinosaurus species with the combination of the following autopomorphic characters: basioccipital with the ventral extracondylar area (ECA) greatly extending anteroposteriorly and absence of ventral protrusion so that the extracondylar area is not visible in posterior view (no dorsoventral orientation of the extracondylar area); ventral extracondylar area slightly concave; absence of basioccipital peg; prominent basioccipital condyle forming the entire posterior surface of the basioccipital; markedly thickened and robust ribs that are round in cross-section, lacking a longitudinal groove along the shaft, particularly in the dorsal region.

Description.

Skull

Orbit and sclerotic ring. The internal orbit is partially preserved in UMO BT 011 235.00 (Figs 4, 7). The orbit is approximately 14 centimeters in diameter (Suppl. material 1: table S1). The orbit is extremely large compared to the size of the skull (Figs 4, 7). The orbit is surrounded by the jugal ventrally, the lacrimal anteriorly, the prefrontal anterodorsally, the postfrontal dorsally, and the postorbital posteriorly (Fig. 7). Some bony material visible in the orbital region can tentatively be attributed to fragments of sclerotic plates (Fig. 7).

Premaxilla. The premaxilla is a paired bone preserved in all three specimens (Figs 3, 4, 5, 6, 7; Suppl. material 3: fig. S2). In UMO BT 011 235.00 and UMO BT 011 221.00, it is heavily damaged and fragmentary (Figs 3, 4, 6, 7). In UMO BT 011 235.00, solely the posterior portions of both premaxillae are preserved. The right premaxilla is articulated and visible in medial view. A small portion of the left premaxilla partially covering the right premaxilla is visible in lateral view and is slightly shifted upwards. In UMO BT 011 221.00, a large portion of the right premaxilla is preserved in medial view. However, the anterior and posterior ends are missing. In the same specimen, a small portion of the left premaxilla is preserved in lateral view. The premaxillae are disarticulated in UMO BT 011 221.00. In UMO BT 011 240.00, the premaxillae, exposed in both lateral (left element) and medial (right element) views, are well-preserved and articulated (Fig. 5). The most visible element, the left premaxilla, measures approximately 63 centimeters as preserved (the posterior portion is broken). (Suppl. material 1: table S1).

The premaxilla is an extremely elongated, straight, and slender bone forming most of the snout (anterior to the narial opening). The posterior portion of the premaxilla is dorsoventrally wide but narrows gradually towards its anterior tip (Figs 3, 5, 6). In UMO BT 011 221.00, the right premaxilla preserved in medial view, shows a medial flange parallel to the lateral wall (corresponding to the ventral margin of the premaxilla) along the entire length of the premaxilla (Fig. 6; Suppl. material 3: fig. S2). The medial shelf and lateral wall of the premaxilla form a shallow alveolar groove, in which the teeth insert when in articulation. No individual alveoli are observable in the premaxilla.

Laterally, the premaxilla carries a longitudinal groove extending over its entire length: the fossa premaxillaris (Fig. 5), like already observed in known Eurhinosaurus specimens (e.g., von Huene 1922, 1928; Maisch 2022). The fossa premaxillaris is discontinuous anteriorly and alternates between external exposure with small open pits and intraosseous pathways (Fig. 5).

As mentioned above, premaxillae articulated to the rest of the skull are uniquely preserved in UMO BT 011 235.00, but the posterior portion of the bone is too damaged to identify the exact contact with the maxilla and nasal (Fig. 7). However, the right premaxilla, visible in medial view in specimen UMO BT 011 240.00, preserves a clearly defined facet for the nasal, located on the posterior mediodorsal surface of the bone (Fig. 5). In addition to this posterior mediodorsal contact with the nasal, the premaxilla also articulates posteroventrally with the maxilla (Fig. 7). Both premaxillae contact each other medially along most of their length, as also observed in UMO BT 011 240.00 (Fig. 5). In this specimen, the medial suture accounts for approximately 64% of the preserved premaxillary length, while the nasal facet comprises the remaining ~36%. However, the posterior-most region of the premaxilla is absent, precluding a complete assessment of the nasal facet’s full extent. Since the anterior margin of the rounded nasal opening is not preserved and the overall nasal region is heavily damaged, the contribution of the premaxilla’s posterior end to the anterior margin of the narial opening is not visible in UMO BT 011 235.00 (Fig. 7).

Maxilla. The maxillae are disarticulated in UMO BT 011 221.00 and articulated in UMO BT 011 235.00 (Figs 3, 4, 6, 7; Suppl. material 3: fig. S2). In UMO BT 011 221.00, the left maxilla is exposed in lateral view, and in UMO BT 011 235.00, the right maxilla is visible in medial view. In UMO BT 011 235.00 specifically, despite being articulated, the maxilla is strongly damaged. In lateral view, the maxilla has an anteroposteriorly elongated outline (Fig. 6). Both posterior and anterior portions of the maxilla narrow towards the respective ends. In UMO BT 011 221.00, on the anterior projection of the maxilla, there is an elongated and deep groove laterally that faces dorsally, into the direction of the suture with the premaxilla (Suppl. material 3: fig. S2). The limits of the anterior end and dorsal suture of the maxilla are not visible, and the posterior suture is uncertain in UMO BT 011 235.00. Anteriorly and anterodorsally, the maxilla contacts the premaxilla (Fig. 7). In medial view, the posterior extension of the maxilla runs under the lacrimal so that it contacts the latter posterodorsally. Also in medial view, the posterior end of the maxilla contacts the jugal along a short contact (Fig. 7). In UMO BT 011 235.00, due to the heavily damaged region of the narial opening, it is uncertain whether the dorsal portion of the maxilla contributes to the ventral margin of the narial opening in medial view. In the maxilla of UMO BT 011 221.00, the anteroposterior extent of the facets for the premaxilla and the jugal indicates that these do not contact one another. This suggests that the dorsal surface of the maxilla at its midpoint, between the premaxilla and jugal facets, contributed to the ventral margin of the narial opening in lateral view. The alveolar groove of the maxilla is not visible in either specimen and since no teeth are found articulated in the maxilla, it remains unknown how many teeth the maxilla bears.

Jugal. The left jugal is completely preserved in medial view in UMO BT 011 235.00 but disarticulated (Figs 4, 7), while the right jugal is preserved in medial view and partially articulated to the skull (Fig. 7). The right jugal is heavily damaged in UMO BT 011 235.00 and its posterior portion is missing. In UMO BT 011 221.00, two fragments of bones are tentatively identified as anteriorly incomplete jugals (Fig. 6; Suppl. material 3: fig. S3). However, based on its general morphology, the bone fragment identified by us as left jugal may also represent the left postorbital (Fig. 6; Suppl. material 3: fig. S3). Nevertheless, the fragment cannot be assigned to either element with certainty. The fragment identified as a right jugal is also incomplete posteriorly in UMO BT 011 221.00. In the same specimen, both jugals are preserved in lateral view. The jugal is a long and curved bone that forms the ventral mid-orbital margin when in articulation, extending posterodorsally beyond the orbit (Fig. 7). The anterior ramus of the jugal is thin and tapers towards its anterior process (Fig. 7). The anterior end of the jugal contacts the posterodorsal end of the maxilla anteroventrally and the lacrimal anterodorsally (Fig. 7). Based on the jugal facet of the maxilla of UMO BT 011 221.00, the contact between the jugal and the maxilla was shorter than the one between the maxilla and the premaxilla (Suppl. material 3: fig. S2). The posterodorsal ramus ends in a broad dorsal facet that articulates with the descending process of the postorbital (Fig. 7). The postorbital facet of the jugal presents numerous thick ridges and grooves on its lateral surface (Fig. 7). In UMO BT 011 221.00, the jugals are more angular than in UMO BT 011 235.00 (Suppl. material 3: fig. S4). The postorbital facet is also smaller in lateral view, as preserved in UMO BT 011 221.00, than in medial view, as preserved in UMO BT 011 235.00 (Fig. 6; Suppl. material 3: fig. S3). The overall shape and posterodorsal ridges are very reminiscent of Argovisaurus (Miedema et al. 2024).

Lacrimal. Lacrimals are solely preserved in UMO BT 011 235.00 (Figs 4, 7; Suppl. material 3: fig. S3). The left lacrimal is disarticulated and preserved in lateral view (Fig. 4). The right lacrimal is part of the skull bones that are preserved in articulation and is exposed in medial view (Fig. 7). The lacrimal is thin and has a triangular shape (Suppl. material 3: fig. S3). The lacrimal forms the anteroventral margin of the orbit (Fig. 7). Its dorsal process contacts the nasal and prefrontal and forms the posteroventral margin of the narial opening (Fig. 7). Despite the poor preservation of this region in UMO BT 011 235.00, it appears that, in medial view, the ventral lacrimal mostly contacts the anterodorsal jugal, as in lateral view, based on the jugal facet morphology of the maxilla of UMO BT 011 221.00. (Fig. 7, Suppl. material 3: fig. S2). The posteroventral lacrimal contacts the anterior process of the jugal (Fig. 7).

Nasal. The nasal is solely preserved in UMO BT 011 235.00. The right nasal is found articulated in medial view (Figs 4, 7) and the left nasal is missing. As mentioned above, the skull has been laterally compressed, and damages obscure the exact shape of the nasal. Therefore, the suture between the nasal and the premaxilla is not clearly visible. In medial view, the nasal contacts the premaxilla anteriorly and anteroventrally. Posteriorly, the nasal has a long vertical contact with the prefrontal (Figs 4, 7). The nasal further contacts the dorsal process of the lacrimal posteroventrally (Figs 4, 7) and forms the dorsal margin of the narial opening (Figs 4, 7).

Prefrontal. The prefrontal is preserved in UMO BT 011 235.00 and UMO BT 011 221.00 (Figs 3, 4, 6, 7; Suppl. material 3: fig. S3). In UMO BT 011 221.00, the right prefrontal is disarticulated and partially covered posteriorly by the dentary and prearticular (Figs 3, 6; Suppl. material 3: fig. S3). The bone is still mostly embedded in the matrix and is partially broken so that only the posterior part of the prefrontal is visible in dorsal view. In UMO BT 011 235.00, the right prefrontal is partially in articulation with other elements and visible in medial view (Fig. 7). However, the right prefrontal is damaged; a portion of the posterior part, visible in ventral view, is broken off and located now within the orbit (Fig. 7). The posterior portion of the left prefrontal is exposed in ventral view and disarticulated from the rest of the skull, but still in contact with fragments of the left postfrontal and parietal (Figs 4, 7; Suppl. material 3: fig. S3). Nevertheless, in the left prefrontal of UMO BT 011 235.00, the contacts with the postfrontal and parietal do not appear entirely genuine, as these bones seem to have overlapped with thin, fragile sheets of bone that appear to have partially broken, likely altering the genuine shape of the suture in some areas. The prefrontal is curved in medial view and slightly crescentic. It comprises a descending process that contributes to the anterior rounded orbit, and a flattened dorsal portion that contributes to the skull roof and borders the anterodorsal orbit margin (Fig. 7). The posterior prefrontal portion contributes to the skull roof and has a broad ridge ventrally, running longitudinally, so that the prefrontal appears thick in this portion. This ridge is not visible on the dorsal surface, so the prefrontal appears flat in dorsal view (Fig. 6). In UMO BT 011 221.00, the dorsal surface of the prefrontal is slightly concave and almost entirely covered with striations, suggesting that it was largely overlapped by a bony sheet, most likely the parietal and postfrontal (Suppl. material 3: fig. S3). Dorsolaterally, the prefrontal presents a thin flange that contacts the postfrontal. In medial view, along its anteroventral surface, the prefrontal has a contact with the dorsal process of the lacrimal (Fig. 7). The prefrontal extensively contacts the nasal anteriorly along a curved suture (Fig. 7). This contact excludes the prefrontal from contributing to the narial opening in medial view. The prefrontal contacts the parietal posteromedially. The prefrontal likely contacted the frontal medially, but no frontal is preserved in the Mistelgau specimens. Consequently, the morphology of the frontal remains unknown in Eurhinosaurus, as it has never been described for other specimens of the genus. The prefrontal forms the major portion of the anterodorsal margin of the internal orbit (Fig. 7).

Postfrontal. The anterior part of the left postfrontal is preserved in ventral view in UMO BT 011 235.00 (Figs 4, 7). In the articulated part of the skull in UMO BT 011 235.00, bone fragments are potentially identified as the partial right postfrontal in medial view, due to their location in the skull between the prefrontal and the postorbital (Fig. 7). However, in this specimen, in addition to the damage, the postfrontal fragments are partially covered by the dislocated right angular, which obscures the exact shape of the postfrontal. Nevertheless, despite the damage, it is observable that the postfrontal contributes to the dorsal margin of the orbit (Fig. 7). In UMO BT 011 235.00, the left postfrontal is thin and flat anteriorly. Anteromedially, the postfrontal contacts the posterolateral part of the prefrontal (Fig. 7). Posterior to the contact with prefrontal, a short contact between the medial margin of the postfrontal and the lateral margin of the parietal is visible in ventromedial view (Fig. 7).

Squamosal. An unambiguous squamosal is only preserved with UMO BT 221.00 (Fig. 6; Suppl. material 3: fig. S3). In UMO BT 011 235.00, a small fragment of bone in contact with the dorsal process of the postorbital is visible and can be identified as the right squamosal in medial view, based on its location (Fig. 7). The bone is hidden behind the postorbital and only a small strip is visible, preventing a detailed description. In UMO BT 221.00, the squamosal is the left element, disarticulated and preserved in dorsolateral view. The squamosal is largely complete with partially broken anterodorsal and ventral edges. The element is triangular in outline with the largest portion directed anteriorly (Suppl. material 3: fig. S3). The squamosal is convex in dorsal view, forming part of the natural curve of the posterior cheek and skull roof (Suppl. material 3: fig. S3). Ventrolaterally a distinct, slightly offset round process is visible (Suppl. material 3: fig. S3). The process is concave medially, which would have held part of the dorsal lamella of the quadrate. Just dorsomedial to the quadrate process the posterior margin of the squamosal is slightly concave (Suppl. material 3: fig. S3).

Postorbital. The right postorbital is preserved in medial view in UMO BT 011 235.00 (Figs 4, 7). The postorbital is still in correct anatomical position and oriented vertically (Fig. 7), resulting in a semi-lunate shape with well-developed dorsal squamosal and ventral jugal facets. The anterior margin of the postorbital forms most of the posterior margin of the orbit (Fig. 7). The postorbital appears broader dorsally and slightly narrows gradually towards its ventral tip. The squamosal facet of the postorbital is broad and extends anterolaterally (Fig. 7). A small anterodorsal contact of the postorbital with the postfrontal is visible. The jugal facet of the postorbital is slightly roughened and oriented ventrally (Fig. 7). The jugal facet presents a shallow groove that contacts the postorbital facet of the jugal when articulated. The jugal facet is slightly longer and less broad than the squamosal facet. A contact with the quadratojugal is not preserved, and the quadratojugal facet is not clearly identifiable. A short ridge located dorsal to the jugal facet could correspond to a posteriorly oriented quadratojugal facet, but this is highly uncertain due to the preservation and may also be an artefact of taphonomy.

Quadratojugal. The right quadratojugal is preserved in UMO BT 011 221.00 and visible in lateral view (Figs 3, 6; Suppl. material 3: fig. S3). The bone has a dorsally slightly convex sub-rectangular plate-like shape (Fig. 6; Suppl. material 3: fig. S3). Almost the entire dorsal surface of the bone is covered by long and thin striations, suggesting that the bone was partially covered by other bones, most probably the squamosal and postorbital (Suppl. material 3: fig. S3). The posterior rim of the bone is slightly concave, and the anterior rim is roughly serrated (Fig. 6). The quadrate facet is directed ventrally like in Stenopterygius and Temnodontosaurus (McGowan 1973; Maisch and Hungerbühler 2001; Miedema and Maxwell 2022) and the offset process of the quadrate facet is not visible and most probably still embedded in the matrix.

Parietal. Only a small portion of the left parietal is preserved in ventral view in UMO BT 011 235.00 (Figs 4, 7). The fragment corresponds to the anterior-most parietal. The parietal contacts the posteromedial aspect of the prefrontal in medial view (Fig. 7). Posterior to this contact with the prefrontal, the parietal has a minute contact with the postfrontal laterally (Fig. 7).

Pterygoid. In UMO BT 011 235.00, an isolated elongated bone can tentatively be identified as a highly fragmented left pterygoid in dorsal view (Figs 4, 7). However, the bone is heavily damaged, preventing a detailed description. The anterior portion of the pterygoid is broken off and missing. The posterior portion of the pterygoid is round and forms a wide bulge (Fig. 7). Anterior to the bulge, the pterygoid is constricted before expanding mediolaterally on the anterior portion of the bone (Fig. 7).

Epipterygoid. The epipterygoids are well-preserved in lateral view in UMO BT 011 221.00 but completely disarticulated (Figs 3, 6, 8A, B; Suppl. material 3: fig. S3). They are well-ossified and triradiate with two diverging dorsal and posterior projections, and a short but broader ventral foot, giving the element a forked shape (Figs 6, 8A, B; Suppl. material 3: fig. S3). The dorsal projection is referred to here as the epipterygoid dorsal ramus and the posterior projection as the posterior process of the epipterygoid (Fig. 8A, B). The dorsal ramus and posterior process that probably contact the parietal dorsally and posterodorsally when in articulation, diverge from the pedicel-like ventral foot that likely contacts the pterygoid when articulated. Both epipterygoids of UMO BT 011 221.00 differ in shape: the left epipterygoid has a broad bony bridge that connects the dorsal and posterior projections, making them less apparent and forming a small, rounded epipterygoid foramen at the center of the bone (Figs 6, 8A, B). The right epipterygoid lacks such a bony bridge, and the lack of damage to the bone suggests a primary absence of the bridge (Fig. 8B), which could indicate a difference in ossification in that specific bone. The dorsal ramus is long and slender (Fig. 8A, B). The posterior process is much shorter (more than twice as short as the other divergent dorsal ramus) and wider (Fig. 8A, B). This is less apparent in the left epipterygoid due to the presence of the afore-mentioned bony bridge. The long epipterygoid ramus narrows towards its dorsal end (Figs 6, 8A, B). The short dorsal epipterygoid process slightly widens towards its dorsal end (Figs 6, 8A, B). The ventral foot is the broadest portion of the epipterygoid and expands ventrally (Figs 6, 8A, B). The epipterygoids were unknown in Eurhinosaurus and the Mistelgau specimens are the first record of such bones in this genus. Preserved epipterygoids are rare in the ichthyosaurian fossil record and have been found ossified in the Triassic Chaohusaurus (Yin et al. 2021) and Besanosaurus (Bindellini et al. 2021), as well as in the Jurassic Ichthyosaurus (McGowan 1973), Stenopterygius (Miedema and Maxwell 2022), and Argovisaurus (Miedema et al. 2024). However, the morphology of the epipterygoids in Eurhinosaurus from Mistelgau, by showing the presence of a diverging dorsal portion and a foramen (in the left element) dorsal to the foot of the epipterygoid, differs greatly from the other known epipterygoids preserved in the afore-mentioned taxa. Nevertheless, based on its location on the epipterygoid of UMO BT 011 221.00, the shorter dorsal process may be analogous to the medial bulge and potential prootic facet present in the epipterygoids of Stenopterygius (Miedema and Maxwell 2022, supplementary figure 11). However, the comparison is limited due to different preservational states (3D preserved in this study but flattened in Miedema and Maxwell 2022) and the scarce knowledge of ichthyosaurian epipterygoids.

Figure 8. 

Selections of photographs of skull elements in UMO BT 011 221.00. A. Left epipterygoid in lateral view; B. Right epipterygoid in lateral view; C. Parabasisphenoid in dorsal view; D. Parabasisphenoid in ventral view; E. Opisthotic in medial view and cervical neural arch in anterior view; F. Isolated tooth in lingual view; G. Identified hyoid corpus in ventral view. Abbreviations: bp, basipterygoid process; ds, dorsum sellae; ebb, epipterygoid bony bridge; edr, epipterygoid dorsal ramus; ef, epipterygoid foramen; epp, epipterygoid posterior process; evf, epipterygoid ventral foot; hsc, horizontal semicircular canal; icf, internal carotid foramen; na, neural arch; przy, prezygapophyses; psc, posterior semicircular canal; sa, sacculus; st, sella turcica; tc, tooth crown; tr, tooth root.

Quadrate. The right quadrate is well-preserved and exposed in anterior view in UMO BT 011 221.00 although partially covered by the quadratojugal (Figs 3, 6; Suppl. material 3: fig. S3). The left quadrate is broken, and only a small portion of the dorsal part is preserved (Fig. 6). The quadrate is one of the most massive skull elements and has a general crescentic outline (Fig. 6; Suppl. material 3: fig. S3). The dorsal and ventral portions of the quadrate in UMO BT 011 221.00 are composed of a laterally directed dorsal supratemporal ramus and a ventral pterygoid lamella. The pterygoid lamella is not accessible for description as it is entirely covered by the quadratojugal. The supratemporal ramus and articular area are very similar in outline. Both supratemporal ramus and articular area are mediolaterally wide and show a convex dorsal and ventral margin. The surface of the entire anterior quadrate is roughened (Suppl. material 3: fig. S3). The quadrate is concave laterally (Fig. 6).

Parabasisphenoid. The parabasisphenoid is preserved in UMO BT 011 221.00 in dorsal view (Figs 3, 6, 8C, D). However, it was prepared from the other side of the slab as well so that the ventral surface of the parabasisphenoid is exposed (Fig. 8C, D). The parabasisphenoid has a quadrangular shape. A long, robust rod-like bone fragment is preserved near the parabasisphenoid and is interpreted as the potential cultriform process that has broken off (Fig. 6). The partial cultriform process is broad posteriorly and narrows gradually towards its anterior end in ventral view. The internal carotid foramen passes through the parabasisphenoid and is located on the anteroposterior midpoint on the ventral side and anteriorly on the dorsal side (Figs 6, 8C, D). The internal carotid foramen is oval and unpaired (Figs 6, 8C, D), as already described in Eurhinosaurus (Maisch and Matzke 2000). Note that, despite the unpaired condition of the internal carotid foramen, the lateral indentations for the carotid arteries are visible (Fig. 8D). In UMO BT 011 221.00, in ventral view, the basipterygoid processes are slightly bulged and anterolaterally oriented. The basipterygoid processes form laterally directed pterygoid facets (Figs 6, 8C, D). The shape is similar to Ichthyosaurus although the basipterygoid processes do not protrude anteriorly as strongly (McGowan 1973). A small, rounded to subtriangular depression oriented laterally is visible on the dorsal surface of the basipterygoid process. The dorsal surface of the parabasisphenoid comprises a small dorsum sellae located posterior to the internal carotid foramen (Fig. 8C). On the dorsal surface of the parabasisphenoid, the sella turcica is extremely small and located just anteriorly to the carotid foramen (Fig. 8C, D). The basioccipital facet of the parabasisphenoid is projecting posterodorsally and forms the entire posterior side of the parabasisphenoid. The basioccipital facet has a strongly pitted texture. The basioccipital facet also shows a narrow and shallow groove that runs dorsoventrally along the entire posterior side of the parabasisphenoid.

Prootic. The paired prootics are preserved in UMO BT 011 221.00 (Figs 3, 6; Suppl. material 3: figs S2, S3). Both elements are still mostly embedded in the matrix, and only visible in anterior view. The prootic has a rounded to oval shape and its anterior surface is slightly domed (Fig. 6; Suppl. material 3: figs S2, S3). It shows an overall smooth surface except for a few rare and shallow pits that are heterogeneously distributed (Suppl. material 3: figs S2, S3).

Opisthotic. Solely the right opisthotic is preserved in medial view in UMO BT 011 221.00 (Figs 3, 6). The opisthotic is partially covered by a neural arch, obscuring most of its shape (Fig. 6), and therefore only the impression of the membranous labyrinth is accessible for description (Fig. 8E). The impression is complete and well-preserved. The texture visible within the impression is smooth. The impression of the membranous labyrinth is deep and V-shaped (Figs 6, 8E). The V-shaped impression shows the horizontal and posterior semicircular canals that meet ventrally while contacting the impression of the sacculus (Fig. 8E). The posterior semicircular canal is slightly broader than the horizontal semicircular canal. Both semicircular canals are slightly curved. The horizontal semicircular canal curves anteriorly and the posterior semicircular canal posteriorly. Neither semicircular canal contacts the medial surface of the opisthotic (Figs 6, 8E). The sacculus is broad and has a rounded shape (Figs 6, 8E).

Exoccipital. Both exoccipitals are preserved in UMO BT 011 221.00 (Fig. 6; Suppl. material 3: figs S3, S4). The left exoccipital is visible in dorsolateral view and the right exoccipital in ventral view. The left element is slightly broken laterally and is slightly compressed vertically. The exoccipitals are small and shaped like a short column with a ventral articulation facet for the basioccipital and dorsal facet for the supraoccipital (Suppl. material 3: figs S3, S4). Both the basioccipital and supraoccipital facets show a roughened texture, but the supraoccipital surface is more heavily pitted than the basioccipital surface. Both facets have a rounded outline. The supraoccipital facet is slightly convex, while the basioccipital facet is more flattened, but still slightly bulged. The shaft of the exoccipital is short and constricted so that the exoccipital is the broadest at its dorsal and ventral ends (Suppl. material 3: fig. S3). The shaft of the exoccipital slightly extends anteroventrally. In ventral view, this small extension is subtriangular in shape and is clearly distinguishable from the basioccipital facet by its slightly roughened texture (Suppl. material 3: figs S3, S4). Therefore, the shaft of the exoccipital is broader ventrally than dorsally. The medial surface of the shaft forms the wall of the foramen magnum. On the lateral surface of the exoccipital, two foramina for the hypoglossal nerve (XII) are present.

Basioccipital. The basioccipital is exposed in dorsal view in UMO BT 011 221.00 (Figs 3, 6, 9A, D; Suppl. material 3: fig. S4). In UMO BT 011 235.00, most of the ventral surface of the basioccipital is exposed (Figs 4, 7, 9B, E). In this specimen, the occipital condyle is clearly visible (Figs 7, 8B, E). The basioccipital is one of the largest and most massive bones among the posterior bones of the skull. In dorsal view, the basioccipital is flattened and has an overall rounded outline with a slightly convex anterior surface (Figs 6, 9A, D). On the dorsal surface, the basioccipital presents two large circular exoccipital facets that have a pitted texture (Figs 6, 9A, D). The exoccipital facets are separated by the floor of the foramen magnum that runs anteroposteriorly (Figs 6, 9A, D). The exposed portion of the floor of the foramen magnum is approximately as long as the exoccipital facets. The anterior half of the dorsal surface of the basioccipital (anterior to the exoccipital facets) has rugose and pitted texture and presents a small anterior notochord pit (Fig. 9A, D). The facet for the parabasisphenoid is formed by the entire anterior surface of the basioccipital. On the anterior face of the basioccipital, a basioccipital peg is absent, which contrasts with the condition described for other species of Eurhinosaurus in which a basioccipital peg is present (McGowan and Motani 2003). Anterolaterally, the basioccipital shows large rounded articulation facets for the stapes (Fig. 9B, E). The posterior face of the basioccipital is entirely formed by the occipital condyle. The occipital condyle is prominent and has a rounded and strongly convex shape. The occipital condyle in the Mistelgau Eurhinosaurus is larger and more massive than in other Eurhinosaurus specimens, especially in posterior view (Fig. 9A–F) (e.g., SMNS 18648, SMNS 56632). A posterior notochord pit is visible dorsally on the occipital condyle (Fig. 9A, D). On the ventral surface of the basioccipital, the extracondylar area is extensive (Figs 6, 9B, E). It extends greatly posteriorly and is reduced laterally. It is also slightly concave medially (Figs 6, 9B, E). However, the condition of the ventral extracondylar area in the Mistelgau specimens differs greatly from the condition described in Eurhinosaurus (Fig. 9B, C, E, F) (McGowan and Motani 2003). As mentioned, the ventral extracondylar area extends greatly anteroposteriorly but does not protrude ventrally so that it is not visible in posterior view (Figs 6, 9B, E), while in Eurhinosaurus the basioccipital is described to present an extensive extracondylar area that greatly extends ventrally so that in posterior view, the extracondylar area is well-visible ventral to the condyle (Fig. 9C, F) (von Huene 1949, 1952; McGowan 1979; McGowan and Motani 2003), like in closely related ichthyosaurs such as Excalibosaurus (McGowan 2003), Leptonectes tenuirostris (Lomax and Massare 2012), and Wahlisaurus (Lomax 2016). The peculiar condition of the basioccipital in the Mistelgau specimens is similar to the condition observed in Hauffiopteryx (Maxwell and Cortés 2020) and, to some extent, Temnodontosaurus zetlandicus (Laboury et al. 2022) and Temnodontosaurus trigonodon (Fraas 1913; von Huene 1931b).

Figure 9. 

Selection of morphological characteristics and pathologies of Eurhinosaurus mistelgauensis sp. nov. A. Photograph of the basioccipital in UMO BT 011 221.00 in dorsal view; B. Photograph of the basioccipital in UMO BT 011 235.00 in ventrolateral view; C. Photograph of the basioccipital of SMNS 18648 in posterior view, photograph by Erin Maxwell (Staatliches Museum für Naturkunde, Stuttgart); D. Interpretative line drawing of the basioccipital in UMO BT 011 221.00; E. Interpretative line drawing of the basioccipital in UMO BT 011 235.00; F. Interpretative line drawing of the basioccipital in SMNS 18648; G. Photograph of a dorsal rib of UMO BT 011 221.00 in cross-section at midshaft; H. Photograph of the anterior cervical rib with pseudarthrosis in UMO BT 011 221.00. Abbreviations: anp, anterior notochord pit; co, occipital condyle; eca, extracondylar area; exf, exoccipital facet; fmf, floor of the foramen magnum; pop, posterior notochord pit; stf, articulation for stapes.

Stapes. Stapes are preserved in UMO BT 011 221.00 and in UMO BT 011 235.00 (Figs 3, 4, 6, 7; Suppl. material 3: fig. S4). In UMO BT 011 221.00, the right stapes is visible in posterior view, and the left stapes in anterior view. In UMO BT 011 235.00, it is uncertain whether the stapes is the left or right element. Stapes are small bones presenting a lateral and medial articular facet which are separated by a slightly constricted shaft (Figs 6, 7; Suppl. material 3: fig. S4). The medial head of the stapes, contacting the basioccipital when articulated, is massive and broad. (Figs 6, 7; Suppl. material 3: fig. S4). It is convex in anterior and posterior view and shows a roughened texture. The shaft is slightly constricted at the midshaft (Figs 6, 7; Suppl. material 3: fig. S4). The lateral end of the stapes is the articulation facet for the quadrate (Suppl. material 3: fig. S4). The quadrate facet extends dorsoventrally and has a sub-rectangular shape that is slightly convex. This morphology is reminiscent of other Early Jurassic ichthyosaurs such as Stenopterygius and Hauffiopteryx (Miedema and Maxwell 2019; Marek et al. 2015).

Dentary. A single partial left dentary is preserved in lateral view in UMO BT 011 221.00 (Figs 3, 6; Suppl. material 3: fig. S2). The preserved bone is crushed, missing the anterior and posterior parts. The dentary is an elongated bone that forms most of the anterior part of the mandible. The bone narrows towards its posterior and anterior ends and is the broadest at its midpoint (Fig. 6). The lateral surface of the dentary is convex. The shallow dentary fossa is present on the entire length of the lateral side of the bone (Fig. 6; Suppl. material 3: fig. S2). The fossa is shallower posteriorly than anteriorly. Anteriorly, the fossa is discontinuous, showing an alternation of external furrows and foramina with passage inside the bone (Suppl. material 3: fig. S2), as also observed in the premaxilla. The dentary normally bears teeth, but teeth are completely disarticulated in UMO BT 011 221.00, and the alveolar grooves are obscured by matrix.

Splenial. The splenial is preserved in both UMO BT 011 221.00 and UMO BT 011 235.00 (Figs 3, 4, 6, 7). The splenials are well-preserved in medial view in UMO BT 011 221.00 but their posterior end is obscured by sediment. In UMO BT 011 235.00, both splenials are damaged and fragmented. In UMO BT 011 235.00, the right splenial is visible in medial view and the left splenial is too damaged to distinguish which view of the bone is visible. The splenial is a long bone that forms most of the medial aspect of the mandible. The splenial is deep but narrows dorsoventrally towards its anterior end. The anterior end of the bone is forked, showing the two symphyseal processes (Fig. 6). The symphyseal processes run from the dorsal and ventral edges of the splenial (Fig. 6). The ventral process is more than twice the length of the dorsal process, which is also less robust than the ventral process (Fig. 6). This is standard morphology for parvipelvian ichthyosaurs, except for Ophthalmosaurus icenicus in which the processes are closer to equal length (Moon and Kirton 2016). A short sheet of bone is present between the two major symphyseal processes (Fig. 6), which to our knowledge could be an autapomorphy of Eurhinosaurus, although the morphology is unclear in many taxa e.g., Excalibosaurus (McGowan 1989a). The processes are rounded in cross-section.

Surangular. The surangulars are preserved in medial view in UMO BT 011 235.00 (Figs 4, 7). In UMO BT 011 221.00, the left surangular is preserved in medial view, and the right surangular in lateral view (Figs 3, 6; Suppl. material 3: fig. S2). The surangular is a massive and elongated bone that forms the major part of the posterior lower jaw. The right surangular in UMO BT 011 221.00 is still articulated with the right angular. The posterior and ventral margins of the surangular contact the dorsal margin of the angular. The suture between the two bones is well-visible laterally (Fig. 6). Anteriorly, the surangular narrows towards its anterior end, becoming thin and reducing greatly in height. On the medial surface of the surangular, the Meckelian fossa and canal extend longitudinally along the entire length of the bone. The Meckelian fossa becomes shallower towards the posterior end of the bone and disappears at the posterior-most aspect of the surangular (Fig. 6). On the posterior portion of the surangular, the dorsal surface gives place to a rounded preglenoid process (sensu Bindellini et al. 2021) that extends dorsomedially (Figs 6, 7; Suppl. material 3: fig. S2). Anterior to the preglenoid process, the dorsal margin forms a broad paracoronoid process (Suppl. material 3: fig. S2). On the posterior-most portion of the surangular, the dorsal margin develops a thin and convex articular facet (Figs 6, 7; Suppl. material 3: fig. S2). The articular facet is covered on its medial surface by minute and thin ridges for articulation with the articular. The preglenoid process and articular facet are distinctly separated by a small glenoid fossa on the dorsal margin (Suppl. material 3: fig. S2). The depth of the fossa varies between the specimens, being slightly less apparent in UMO BT 011 235.00 than in UMO BT 011 221.00 (Figs 6, 7). A large foramen is located anterior to the preglenoid process on the medial side of the surangular (Figs 6, 7; Suppl. material 3: fig. S2).

Angular. The right angular is preserved in lateral view in UMO BT 011 221.00 and the left in medial view (Figs 3, 6). In UMO BT 011 235.00, the right angular is preserved in medial view (Figs 4, 7). The right angular in UMO BT 011 221.00 is complete and still preserved in articulation with the surangular. The suture between the two bones is well-identifiable on the lateral side of the mandible. The angular is an elongated bone that curves dorsally in its posterior portion, where it reaches its greatest dorsoventral height (Figs 6, 7). The angular tapers anteriorly (Figs 6, 7, and its medial surface, anterior to the curvature, is slightly concave in UMO BT 011 221.00.

Articular. The left articular is preserved in medial view in specimen UMO BT 011 235.00 (Suppl. material 3: fig. S2). It is relatively small and robust and exhibits a slightly quadrangular outline. The medial surface is strongly concave anteroposteriorly, resulting in maximal width at the anterior and posterior ends and giving the element a saddle-like appearance in medial view (Suppl. material 3: fig. S2). The ventral articulation facet for the prearticular is slightly concave and elongate anteroposteriorly (Suppl. material 3: fig. S2). The anterior surface of the articular is entirely formed by the glenoid, which is dorsally slightly convex and displays a roughened, heavily pitted texture (Suppl. material 3: fig. S2). The posterior surface is well defined, convex, and contributes to the posterior margin of the mandible. Like the glenoid, it is roughened, although it lacks pitting.

Prearticular. In UMO BT 011 221.00, the right prearticular is preserved in medial view (Figs 3, 6; Suppl. material 3: fig. S2). In the same specimen, the left prearticular is potentially preserved in lateral view (Figs 3, 6; Suppl. material 3: fig. S2). In UMO BT 011 235.00, one bone fragment can tentatively be identified as a right prearticular preserved in medial view (Figs 4, 7). However, in this specimen, the bone is too fragmentary to be able to state this with certainty. In both specimens, most of the anterior and posterior portions are missing. The prearticular is thin, but the ventral margin is thicker than the dorsal margin (Fig. 6). The prearticular narrows towards its anterior end and its posterior portion is slightly angled (Fig. 6).

Dentition. Teeth are mainly well-preserved in UMO BT 011 240.00 and UMO BT 011 221.00. In UMO BT 011 240.00, numerous teeth are preserved in articulation with the two premaxillae (Fig. 5), but teeth are completely disarticulated and dispersed in UMO BT 011 221.00. In UMO BT 011 240.00, the articulated teeth are placed vertically in the premaxilla. Very few dispersed teeth are also preserved with UMO BT 011 235.00. Due to the disarticulation of the teeth and the preservational state of the maxilla and the dentary in UMO BT 011 235.00 and UMO BT 011 221.00, the condition of the dentition and teeth alveoli in the maxilla and lower jaw (dentary) is unknown for the Mistelgau specimens. In the Mistelgau specimens, the premaxilla shows a shallow alveolar groove (see section on the premaxilla).

The exact number of teeth per ramus is unknown in the Mistelgau specimens. In UMO BT 011 240.00, the best-preserved specimens regarding the teeth, 44 teeth are preserved in the left premaxilla (none are identified in the right premaxilla) (Fig. 5). In this same specimen (UMO BT 011 240.00), the roots of most of the in-place teeth are visible (Fig. 5). The teeth are mainly straight (Fig. 8F), but some are slightly curved (Fig. 5). The exact distribution of straight and curved teeth along the jaw remains uncertain, primarily due to the high degree of disarticulation. However, some teeth of UMO BT 011 221.00, found in situ, suggest that the anterior portion of the upper jaw was primarily composed of straight teeth, with the exception of those at the tip, which appear slightly curved. The teeth show no carina and have a narrow conical shape (Fig. 8F). The surface of the crown is completely smooth, unlike the root, which is formed by plicidentine on its entire length (Fig. 8F). The shape of the teeth can typically be associated with the pierce 1 guild, which corresponds to a diet of soft prey such as cephalopods or small fish (Massare 1987). No graduation in tooth size is observable and the replacement teeth are randomly distributed along the rami. The functional tooth size ranges from less than 15 millimeters to 20 millimeters in height for the largest (root of the tooth included) (Suppl. material 1: table S1). The largest crowns reach 11 millimeters in height, and the smallest are approximately 8 millimeters in height (Suppl. material 1: table S1). The higher the tooth, the greater the diameter. The largest diameter reaches 6 millimeters for the root and 5 millimeters for the diameter at the base of the crown (Suppl. material 1: table S1).

Hyoid apparatus. Parts of the hyoid apparatus are preserved in ventral view in UMO BT 011 221.00 (Figs 3, 6; Suppl. material 3: fig. S4). The hyoids consist of two slightly curved rod-like elements identified as ceratobranchial I (Fig. 6; Suppl. material 3: fig. S4). At approximately their midpoint, the cross-section is round, becoming flatter towards both ends. (Suppl. material 3: fig. S4). The ventral surface of the ceratobranchial is smooth. Another isolated element in UMO BT 011 221.00 close to the ceratobranchials, is interpreted as the hyoid corpus (Figs 6, 8G). It is a small element that has a rounded subtriangular shape with a roughened, domed, and convex base (Fig. 8G). The other sides of the bone show coarser roughening and slightly convex surfaces. The general morphology of the hyoid corpus of the Mistelgau specimens corresponds to the rare hyoid corpus findings in other ichthyosaurs (Hauffiopteryx typicus: Motani et al. 2013; Maxwell and Cortés 2020; Stenopterygius quadriscissus: Miedema and Maxwell 2022; Ichthyosaurus sp.: Delsett et al. 2023; Mixosaurus cornalianus: Miedema et al. 2023).

It is important to note that the general morphology and size of the element we interpret as the hyoid corpus is also similar to that of an atlantal intercentrum. However, the element lacks a dorsal mediolateral concavity that would typically accommodate the rounded shape of the centra, which supports our current interpretation as a hyoid corpus. Still, we recognize that both the hyoid corpus and the atlantal intercentrum are rarely preserved and seldom described in detail, which makes distinguishing between them particularly challenging. Due to this limited comparative material, we do not rule out the possibility that this element could represent an atlantal intercentrum. Future discoveries and more comprehensive data may allow for a clearer distinction, but based on our current observations, we provisionally maintain the identification as a hyoid corpus.

Unidentified cranial bones. Several preserved bony remains are too fragmentary or damaged to be correctly identified (mainly in UMO BT 011 235.00). These bone fragments surely correspond to some of the missing, undescribed skull bones. These missing bones are the following: supraoccipital, supratemporal, frontal, palatine, and vomer.

Axial skeleton

Atlas-Axis complex. The atlas-axis complex is well-identifiable in UMO BT 011 221.00 (Figs 3, 6; Suppl. material 3: figs S4, S6). The atlas-axis complex lacks cortical bone and thus exhibits a roughened texture that differs from the other preserved centra. In UMO BT 011 221.00, the centra of the atlas and axis are still mostly embedded in sediment and so they are mostly visible from dorsolateral view. In UMO BT 011 235.00, the atlas-axis complex is heavily damaged and still mostly embedded in the matrix, as in UMO BT 011 221.00, thus exposing only the axis in posterior view. Due to extensive damage, identifying the element interpreted as the atlas-axis complex in specimen UMO BT 011 235.00 remains challenging. The identification is supported by morphological congruence with the atlas-axis complex of UMO BT 011 221.00, as well as the lack of cortical bone on the element. The atlas has a polygonal outline. In lateral view, the centra of the atlas and axis are fused; however, the two elements remain differentiated due to a shallow groove demarcating their junction (Fig. 6; Suppl. material 3: figs S4, S6). In UMO BT 011 221.00, the anterior atlantal facet is concave for the articulation with the occipital condyle of the basioccipital. In UMO BT 011 235.00, the visible posterior axial facet is slightly concave as well.

The atlas and axis are approximately the same length. The articulation facets of the corresponding cervical ribs are poorly defined and are strongly damaged. However, it is visible on both atlas and axis that only one articulation facet for the rib is present. These broad and rounded articulation facets are anterodorsally placed on both centra. Only one half of the atlantal neural arch is preserved in lateral view in UMO BT 011 221.00 (Fig. 6, Suppl. material 3: fig. S3). It is strongly damaged and has a slight triradiate outline (Fig. 6; Suppl. material 3: fig. S3). The pedicel is short (Suppl. material 3: fig. S3). The neural spine is about twice the height of the pedicel (Suppl. material 3: fig. S3). The neural spine is oriented posteriorly and slightly narrows towards its dorsal tip (Fig. 6; Suppl. material 3: fig. S3). Pre- and postzygapophyses are not visible. No neural arch of the axis is preserved in any of the specimens.

Postaxis vertebrae. The vertebral columns of UMO BT 011 221.00 and UMO BT 011 235.00 are well-preserved (Figs 3, 4). They range from disarticulated to well-articulated, depending on the region and specimen (Figs 3, 4; Suppl. material 3: figs S5, S8). The condition of the vertebral column does not permit an exact length measurement in any of the two specimens. Due to disarticulation, an exact count of vertebrae cannot be given in either of the specimens. UMO BT 011 235.00 has a total of 128 centra preserved, and UMO BT 011 221.00 has 107. The caudal region is very incomplete in UMO BT 011 221.00 but mostly preserved in UMO BT 011 235.00. In UMO BT 011 235.00, 47 preserved postflexural centra, among which 25 are articulated, form the ventrally descending tail fluke that is elongated (Fig. 4; Suppl. material 3: fig. S9). However, no wedge-shape centra from the tail flexure are observable. In UMO BT 011 235.00, 81 centra are preflexural vertebrae. In UMO BT 011 221.00, among the 107 preserved centra, 89 centra are preflexural and 18 are interpreted to be postflexural, based on their small size (compared with postflexural centra in UMO BT 011 235.00) and absence of rib articulation facets. 15 of the 18 postflexural centra are preserved on two slabs found isolated from the rest of the specimen, one is completely isolated, and two are preserved on the main slabs composing the specimen (Suppl. material 3: fig. S1).

Due to the exposure in articular view of numerous centra, the exact determination of the position of the centra is not always possible. Nevertheless, a general distinction between cervical, dorsal, and caudal vertebrae is still possible based on the general morphology, position of the rib facets, and position of the centra on the different slabs composing the specimens. The caudal region of the vertebral column might have started in UMO BT 011 221.00 around the 51st preserved centrum (Fig. 3) since it is the first vertebra showing fused diapophysis and parapophysis (not visible in UMO BT 011 235.00). Thus, all posterior vertebrae are considered caudals.

All centra are amphicoelous, but the overall shape of the centra varies throughout the vertebral column. The cervical centra are rounded-pentagonal with a distinct keel ventrally, while the dorsal and caudal centra are rounded. The posterior postflexural centra have a more oval shape, with the dorsoventral height being slightly greater than the lateral width. Dorsally, the centrum presents two facets to which the neural arch attaches. Between the two facets, there is the floor for the neural canal. The cervical and anterior dorsal centra are of moderate height compared to more posterior dorsal vertebrae (Fig. 10A). The dorsoventral height of the centra increases throughout the vertebral column until the anterior caudal centra, which are the highest (Fig. 10A). Then the height of the centra starts to reduce gradually until the posterior portion of the tail where the centra reduce rapidly in height up to the tail bend region (Fig. 10A). The anterior-most postflexural centra are small, and their size reduces slowly until the end of the tail fin, where the centra are the smallest in height (Fig. 10A). The variation of the height of the centra along the body of the Mistelgau specimens (Fig. 10A) is similar to the condition observed in Eurhinosaurus (Buchholtz 2001). The width of the centra of Mistelgau specimens follows the same variation pattern as the height of the centra described above (Fig. 10B). The length of the centra is difficult to assess because the vast majority of them are either lying on their anterior or posterior side, making measurements of the length impossible. However, it is observable that the length of the centra appears to have a similar trend as the height and width of the centra. The length increases from the cervical centra to the posterior dorsal centra. The posterior dorsal and anterior caudal centra are the longest. After the sacral region, the length of the centra decreases rapidly up to the postflexural centra which are the shortest. The di- and parapophyses, located on the lateral sides of the centrum, also follow a pattern throughout the vertebral column. Anteriorly in the cervical centra, the diapophysis is contacting the dorsally placed facets for articulation with the neural arch. Throughout the vertebral column, the diapophysis shifts gradually more ventrally and becomes separated from the facet for the neural arch. All cervical and dorsal centra have a distinctly separated diapophysis and parapophysis. Anteriorly, the diapophysis is larger than the parapophysis, but both facets tend to be approximately the same size in the dorsal centra. As the position of the facets lowers throughout the vertebral column, the distance between the diapophysis and parapophysis narrows until the two facets fuse. While both facets fuse, it appears that the parapophysis reduces in size at the same time. The diapophysis and parapophysis of the anterior centra are connected to the anterior rim of the centra by a bony bridge. This bridge disappears in the posterior dorsal centra, and the facets remain isolated from the edges of the centra. The first centrum with a synapophysis is considered to be the first caudal vertebra (as mentioned above). All the caudal centra have a synapophysis placed lower on each lateral side of the centra. They continue to lower throughout the tail and gradually reduce in size until they disappear. The posterior-most preflexural centra and the postflexural centra have no articulation facets for the ribs. The apophyses of the cervical and dorsal centra have a rounded shape. Most of the caudal centra show rounded apophyses as well. However, the shape of the synapophyses of the anterior caudal centra is irregular, where the diapophysis and parapophyses fuse. In these specific centra, the shape is initially hourglass-like and becomes rounder throughout the fusion of the two apophyses.

Figure 10. 

Charts showing the size variation of the centra along the vertebral column in the Mistelgau Eurhinosaurus. A. Variation of the centra height; B. Variation of the centra width. Note that the graphs are based on the measurements provided in Suppl. material 1: table S1. As centra are missing in both specimens (particularly in the caudal region, see description), the centra numbers do not necessarily exactly correspond to their anatomical positions, but to the preserved consecutive order (measured from anterior to posterior).

The neural arches are well-preserved in UMO BT 011 221.00, but much less in UMO BT 011 235.00, where most of them are missing or damaged. UMO BT 011 221.00 has 71 neural arches preserved, and UMO BT 011 235.00 has 6. In both UMO BT 011 221.00 and UMO BT 011 235.00, no neural arches are preserved in articulation with their respective centrum, and all are dispersed around the specimens (Suppl. material 3: fig. S5). Ventrally, the neural arches have broad and slightly convex articulation facet for the centrum. The neural arches slightly narrow toward their dorsal portion where they bear the zygapophyses and the neural spine. The neural arches in the posterior caudal region show a porous surface, unlike the other neural arches which have a smoother surface (Suppl. material 3: fig. S5). The same size variation throughout the vertebral column as in the centra is expected for the neural arches in the Mistelgau specimens. However, since numerous neural arches are missing, it is difficult to assess in detail their size variation. Nevertheless, based on the preserved neural arches in UMO BT 011 211.00, the neural arches are higher in the cervical and dorsal region than in the caudal region where they rapidly reduce in height. The prezygapophyses and postzygapophyses are located dorsally on the neural arch pedicels (Suppl. material 3: fig. S5). In the presacral vertebrae, the prezygapophyses project anterodorsally and the postzygapophyses project posterodorsally (Suppl. material 3: fig. S5). The prezygapophyses and postzygapophyses have an oval to rounded shape (Fig. 8E; Suppl. material 3: fig. S5). Both pre-and postzygapophyses are paired, one on each side of the neural arch in the presacral vertebrae. However, each pair tends to gradually converge medially along the vertebral column and be confluent in anterior caudal vertebrae up to the tip of the tail. The size variation of the pre- and postzygapophyses follows a similar pattern as the size of the neural arches. They are well-developed and large in the cervical and dorsal region, but rapidly reduce in size in the caudal region, to almost completely disappear in the posterior-most caudal neural arches, where they only appear almost as a notch in the neural arch (Suppl. material 3: fig. S5). The neural spine that is projected posteriorly throughout the vertebral column extends dorsally to the arch (Fig. 8E; Suppl. material 3: fig. S5). All neural spines have a sub-rectangular shape with a distal end that is often slightly convex (Suppl. material 3: fig. S5). The sub-rectangular shape becomes less apparent in the posterior-most caudal neural spine due to the almost complete reduction of the spines (Suppl. material 3: fig. S5). The cervical and anterior dorsal neural spines are mediolaterally broader than the more posterior spines which are thinner (Fig. 8E; Suppl. material 3: fig. S5). The neural spines increase in height rapidly in the first anterior vertebrae. The neural spines are the highest in the dorsal region and they can almost reach twice the height of the neural canal in the largest vertebrae. The neural spines drastically reduce in height in the caudal region and are the smallest in the posterior-most caudal region, where they are extremely short and reduced (Suppl. material 3: fig. S5). No neural arches of postflexural vertebrae are preserved, although neural arch facets are visible on postflexural centra of UMO BT 011 221.00 (Suppl. material 3: fig. S1).

Ribs. Numerous ribs are preserved in UMO BT 011 235.00 and UMO BT 011 221.00 (Figs 3, 4). Only one well-preserved rib is found in UMO BT 011 240.00 (Fig. 5), next to two smaller rib fragments. In UMO BT 011 235.00 and UMO BT 011 221.00, the ribs are disarticulated and mainly completely preserved (Figs 3, 4; Suppl. material 3: figs S6–S8). However, the ribs of the anterior dorsal region are less dispersed than the cervical ribs, posterior dorsal ribs, and caudal ribs (Figs 3, 4). Ribs of the atlas centrum are not preserved in either specimen. However, the two ribs of the axis centrum are preserved in UMO BT 011 221.00 (Fig. 6; Suppl. material 3: figs S4, S6). Additionally, one rib of the third cervical centrum is possibly preserved (Fig. 6; Suppl. material 3: fig. S4). The two ribs of the axis are the smallest and shortest preserved anterior ribs. They are curved and narrow greatly towards their distal end (Fig. 6; Suppl. material 3: figs S4, S6). The single head of the ribs of the axis is well-developed and wide and has a shallow groove proximally that extends on the shaft of the rib (Suppl. material 3: figs S4, S6). The preserved third cervical rib is about twice as long as the axis rib (Suppl. material 3: fig. S4). The third cervical rib is single-headed as the axis rib but contrasts with the latter by showing no groove proximally that extends towards its distal end. The third cervical rib has a round cross-section.

In UMO BT 011 221.00, an element exhibiting an unusual and presumably pathological morphology is preserved in the region of the anterior cervical ribs (Fig. 9H). Although its general morphology is consistent with that of a rib, the element is highly elongate and needle-shaped, characterized by a rounded proximal extremity and a shaft that tapers progressively distally (Fig. 9H). In its proximal portion, the element displays a notably greater diameter than adjacent anterior cervical ribs, supporting the interpretation of an abnormal condition. Accordingly, this element is interpreted as a pathologically modified anterior cervical rib, most likely exhibiting pseudarthrosis.

Posterior cervical ribs and dorsal ribs show a similar morphology. Both show a form of bicapitate head with oval-shaped tuberculum and capitulum which are confluent proximally (Suppl. material 3: fig. S6). This morphology is highly similar to the “sheathed-bicapitate” condition present in the ribs of Argovisaurus martafernandezi (Miedema et al. 2024). The tuberculum and capitulum are of equal size and form an obtuse angle (Suppl. material 3: fig. S6). This angle gradually reduces towards the posterior dorsal ribs, and finally gradually disappears in the caudal ribs. The posterior-most dorsal ribs show more proximally fused but still recognizable tuberculum and capitulum than the anterior or mid-dorsal ribs, showing a gradual transition to caudal ribs. In some dorsal ribs, a shallow groove restricted to the head of the rib is observable between the tuberculum and capitulum, making them appear more distinct, while this groove is absent in other dorsal ribs (Suppl. material 3: fig. S6). However, no trend in the distribution of this proximal groove is observable. Cervical ribs and dorsal ribs show no groove along their shaft, and the shaft is completely round or oval in cross-section (Suppl. material 3: figs S6–S8). The dorsal ribs are the longest and have the greatest diameter at midshaft (Suppl. material 1: table S1).

The Mistelgau specimens show no preserved posterior caudal ribs. The most posterior preserved ribs are the anterior caudal ribs. The caudal ribs are single-headed and the tuberculum and capitulum are completely fused to form a synapophysis which is oval and flat. The caudal ribs are round to oval in cross-section and are greatly reduced in length compared to the dorsal ribs. All ribs of the Mistelgau specimens narrow towards their distal end, except for the very last ribs (posterior-most ribs in the anterior caudal region), which show a widened distal end (Suppl. material 3: fig. S5). In these posterior-most ribs, the head of the rib is as wide as the distal end, and the midshaft is slightly constricted (Suppl. material 3: fig. S5). As the cervical and dorsal ribs, the caudal ribs show no groove along their shaft and are round in cross-section. A variation in the length of the ribs is observable in the Mistelgau specimens. The length of the ribs rapidly increases in the cervical region to reach a maximum in the dorsal region. The last dorsal ribs slightly decrease in length compared to mid-dorsal ribs. The caudal ribs drastically decrease in length compared to dorsal ribs. Within the caudal ribs, their length gradually decreases, whereas the very last ribs are the shortest of the entire body. The same variation applies to the diameter of the ribs, where the thickest ribs are in the dorsal region and the thinnest ribs are in the caudal region (Suppl. material 1: table S1). Additionally, the ribs are curved in the cervical and dorsal regions, but become straighter in the caudal region, as they reduce in length.

The ribs of the Mistelgau specimens appear to be especially thick and massive. Especially the dorsal ribs appear pachyostotic compared to the rest of the body of the Mistelgau specimens with a very thickened cortex and small medullary cavity (Fig. 9G; Suppl. material 3: figs S6–S8). At midshaft, the cortex of the dorsal ribs is moderately thick, with radii ranging from 3 to 4.7 mm in UMO BT 011 235.00 and from 3.5 to 5.4 mm in UMO BT 011 221.00. The radius of the medullary cavity varies between 1.8 and 2.3 mm in UMO BT 011 235.00 and between 1.5 and 3.1 mm in UMO BT 011 221. These measurements are based on the best-preserved rib cross-sections and are presented as indicative intervals (see Suppl. material 1: table S1); poorly preserved specimens not included in the measurements may have affected the actual range. It is noteworthy that the absence of a groove along the shaft may also enhance the perception of the thickness of the ribs. Nevertheless, cross-sections of the ribs confirm that the bone is particularly thick (Fig. 9G; Suppl. material 3: fig. S6). The morphology of the ribs in the Mistelgau Eurhinosaurus specimens, especially in the dorsal region, differs notably from that observed in the holotype of Eurhinosaurus longirostris from Whitby (NHMUK PV OR 14566), which is stratigraphically older than the material from Mistelgau. The dorsal ribs of the holotype do not appear to have undergone significant compression. The holotype exhibits thin ribs with longitudinal grooves along the shaft, resulting in an hourglass-shaped cross-section (Suppl. material 3: fig. S7). In contrast, the Mistelgau specimens show rib morphology more similar to that of specimens from Baden-Wuerttemberg which are also stratigraphically older. These southwestern specimens exhibit some variability, with ribs that may be grooved or ungrooved (typically oval in cross-section in the latter case) (e.g., von Huene 1922, 1926, 1928, 1951; Suppl. material 3: figs S7, S8), although they are generally thin (e.g., von Huene 1926, 1928, 1951). In this regard, the Mistelgau specimens more closely resemble these specimens from Baden-Wuerttemberg than the holotype material from England.

However, none of the previously known Eurhinosaurus specimens exhibit ribs as thick as those in the Mistelgau material. In the Mistelgau specimens, dorsal rib diameters range from 9 mm to 14 mm at midshaft, with the majority exceeding 10 mm. By contrast, in the holotype material from Whitby (NHMUK PV OR 14566), dorsal ribs do not exceed 10 mm at midshaft and are often thinner. Even in the southwestern German material, where ribs are occasionally preserved ungrooved, they do not appear as robust as those from Mistelgau, although they are generally thicker when compared to those of the holotype.

Haemal arches. In the proximal portion of the tail in UMO BT 011 221.00, a few small curved and rod-like elements are preserved near the vertebral centra (Fig. 3; Suppl. material 3: fig. S5). Some of these elements are distinctly smaller and shorter than the rod-like caudal ribs. These elements are interpreted as haemal arches (Suppl. material 3: fig. S5). The haemal arches show a slightly porous surface. Thus, the ossification level of these small elements tends to vary. Some elements resemble the description made by von Huene (1926), being short, straight to curved rod-like elements that, paired, would have formed the haemal arch. However, in the posterior-most portion of the tail, the ribs tend to vary greatly in shape, which makes it sometimes difficult to distinguish between haemal arches and posterior-most caudal ribs. Since the vertebral material is disarticulated, it is difficult to identify the exact true location of the caudal ribs to distinguish which element is a rib or potentially a partial haemal arch. However, very few elements distinctly differ from the rib morphology by being much more curved and presenting a distinct proximal articulation facet (Suppl. material 3: fig. S5). Due to disarticulation and different preservational states, their exact number and position remain unknown. Based on the preserved material, there are at least two vertebrae with haemal arches in the proximal portion of the tail in UMO BT 011 211.00.

Gastralia. Gastralia are thin, long, and sometimes slightly curved rod-like bones that are preserved in UMO BT 011 221.00 and UMO BT 011 235.00 (Figs 3, 4). The gastralia are completely disarticulated and dispersed in both specimens. In both specimens, UMO BT 011 221.00 and UMO BT 011 235.00, the gastralia have a rounded cross-section. The diameter is the largest at the midshaft and decreases towards both ends (Figs 3, 4). In UMO BT 011 221.00, the gastralia are all isolated, but in UMO BT 011 235.00, some gastralia are exceptionally preserved and are showing articulation; some pairs of gastralia are fused to each other medially, forming a single boomerang-shaped element that slightly expands anteroposteriorly, as described in the genus Hauffiopteryx (Maxwell and Cortés 2020).

Appendicular skeleton

Pectoral girdle. Specimen UMO BT 011 221.00 preserves nearly all elements of the pectoral girdle, which are disarticulated; however, the interclavicle is missing. In specimen UMO BT 011 235.00, the pectoral girdle elements are disarticulated and highly dispersed, with only the left clavicle and the interclavicle missing.

Clavicle. The clavicles are preserved in ventral view in UMO BT 011 221.00 (Fig. 11). In UMO BT 011 235.00, the right clavicle is preserved in dorsal view (Fig. 11). The clavicle is an elongated, slender bone with a tapering lateral end bearing striations that indicate the attachment point for the scapula (Figs 3, 4). The medial part is flared (Fig. 11) and shows striations. The clavicle curves laterally after the flatter medial part (Fig. 11). The medial and distal portions of the clavicle are somewhat equally thick. When articulated, the clavicle possibly overlapped the interclavicle medially and contacts the proximal portion of the scapula posterolaterally along a long contact. The medial and lateral flanges of the clavicle are approximately equal in thickness. This is common in early diverging parvipelvians (Temnodontosaurus, Excalibosaurus and Suevo­leviathan, but not in Stenopterygius and Hauf­fiopteryx (McGowan 1989a; Maisch 1998a; Maisch 2008; Maxwell and Cortés 2020).

Figure 11. 

Close-up on the pectoral girdle of UMO BT 011 221.00. A. Photograph; B. Interpretative line drawing. Abbreviations: cl, clavicula; co, coracoid; sc, scapula.

Coracoid. The coracoids are preserved in ventral view in UMO BT 011 221.00 (Figs 3, 11), and in dorsal view in UMO BT 011 235.00 (Fig. 4). The coracoid is plate-like and slightly rounded (Figs 3, 4, 11). It has a small semi-circular notch anterolaterally (Figs 3, 4, 11), as in most parvipelvians (McGowan and Motani 2003). However, the coracoid in Wahlisaurus greatly differs in shape from Eurhinosaurus (Lomax 2016; Lomax et al. 2019), since in addition to an anterior notch, it has a posterior notch and an ovoid foramen, which are absent in the specimens described in this study. The coracoid presents three long articulation facets (Fig. 11). The first one, the intercoracoid facet, is placed medially and the other two laterally (Fig. 11). The articulation facets are especially deep and thus make the coracoid dorsoventrally thicker at its medial and lateral portions. The intercoracoid facet slightly extends on the dorsal surface, forming a small rim on the medial margin of the coracoid (Fig. 11). The anterior-most lateral facet is the scapular facet (Fig. 11). Posterior to the scapular facet is a longer, slightly concave articulation facet, the glenoid facet. Both articulation facets show a rough texture of numerous large pits. The anterior notch is covered with cortical bone.

Scapula. The scapulae are visible in external view in UMO BT 011 221.00 (Figs 3, 11) and in internal view in UMO BT 011 235.00 (Fig. 4). The scapula is an anteroposteriorly elongated bone with a fan-like, flared proximal and slightly flared distal end (Fig. 11). At midshaft, the scapula is rounder than at its ends, showing here an oval cross-section. The proximal flared end of the scapula is much wider and convex proximally than the distal end (Fig. 11). The posterior shaft of the scapula is strongly curved, unlike the anterior shaft, which is almost straight. The wide proximal portion of the scapula takes part in the glenoid articulation. The entire proximal margin of the scapula is not strongly differentiated and thus no well-defined acromion process is visible (Fig. 11). Consequently, the acromion process of the scapula of the Mistelgau specimen is not prominent like in other leptonectid ichthyosaurs (e.g., Leptonectes moorei: McGowan and Milner 1999; Excalibosaurus: McGowan 2003, and Wahlisaurus: Lomax 2016). The texture of the articular surface of the scapula is roughened. The proximal blade of the scapula articulates with the coracoid anteromedially. When in articulation, the scapula also contacts the clavicle dorsally from its proximal end towards the mid-shaft.

Pelvic girdle. The pelvic girdle is almost completely preserved in UMO BT 011 221.00 (Figs 3, 12) including the right ilium, both pubes, and both ischia (Fig. 12). UMO BT 011 235.00 shows a left ilium, the right ischium, and both pubes (Fig. 4). The pubis and ischium show no fusion and are separate elements, as in most non-baracromians (McGowan and Motani 2003).

Figure 12. 

Close-up on the pelvic girdle of UMO BT 011 221.00. A. Photograph; B. Interpretative line drawing. Abbreviations: il, ilium; is, ischium; pb, pubis.

Ilium. The right ilium is preserved in medial view in UMO BT 011 221.00 (Figs 3, 12), and the left ilium in lateral view in UMO BT 011 235.00 (Fig. 4). The ilium is an elongated and slender rod-like bone with a rounded cross-section. It is thinnest dorsomedially but becomes wider towards its ventrolateral acetabular end that contacts the pubis and ischium when in articulation, forming the acetabulum. The bone is slightly curved, showing a slight angle in the middle of its length (Fig. 12). The ilium is slightly longer than the ischium and pubis (Fig. 12).

Ischium. Both ischia are preserved in external view in UMO BT 011 221.00 (Figs 3, 12), and the right ischium in UMO BT 011 235.00 is visible in internal view (Fig. 4). The ischium is a flattened, plate-like bone, and the largest bone of the pelvic girdle (Figs 3, 4, 12). The bone is mediolaterally wider than anteroposteriorly long (Fig. 12). The ischium is narrowest at midshaft, showing a widening towards both ends. The widening of the medial end of the ischium is more developed than the one of the lateral acetabular end. (Fig. 12). The medial margin is strongly curved, unlike the lateral margin, which shows an almost straight but oblique margin (Fig. 12).

Pubis. The pubes are preserved in lateroventral view in UMO BT 011 221.00 (Figs 3, 12), and in mediodorsal view in UMO BT 011 235.00 (Fig. 4). The pubis is an elongated flat bone with a constricted shaft and widened ends (Fig. 12). The lateral acetabular end is slightly wider than the medial end forming the pubic symphysis when in articulation. The pubis is the narrowest at the midshaft (Fig. 12). The lateral margin is slightly curved, and the medial margin is straight (Fig. 12).

Humerus. In both specimens, UMO BT 011 221.00 and UMO BT 011 235.00, the humeri are robust and elongated elements. In both specimens, the diaphysis of the humerus is constricted, giving the margins a concave outline (Fig. 13; Suppl. material 3: fig. S10). Proximally, the humerus is slightly expanded and exposes a large surface for articulation with the glenoid (Figs 13, 14A, B, C). This proximal expansion is more developed in UMO BT 011 235.00 and almost non-existent in UMO BT 011 221.00 (Figs 13, 14B). In UMO BT 011 235.00, the proximal articulation of the humerus is slightly convex (Fig. 14A). In UMO BT 011 221.00, the proximal surface is flattened and extremely pitted (Figs 13, 14B), which is considered abnormal and interpreted as being potentially pathological (see section on pathologies in the discussion). Note that the proximal end of the left humerus in UMO BT 011 235.00 shows a zone of subsidence above the dorsal process (Fig. 14C, Suppl. material 3: fig. S10), which is interpreted as an indicator of avascular necrosis (see section on pathologies in the discussion). The dorsal process is only visible in UMO BT 011 235.00 (Fig. 14C; Suppl. material 3: fig. S10). The dorsal process is more developed than the deltopectoral crest and runs further distally along the diaphysis, while the deltopectoral crest is short and does not extend along the diaphysis. Proximally, the dorsal process forms a descending sub-triangular surface that extends the glenoid facet (Suppl. material 3: fig. S10). The texture of the dorsal process is roughened proximally. The deltopectoral crest differs between both specimens. While in UMO BT 011 221.00, it appears as a rectangular structure that is confluent with the edge of the proximal humeral head (Figs 13, 14B). The deltopectoral crest, especially in UMO BT 011 235.00, is broad and less differentiated from the humeral shaft than the dorsal process (Fig. 14A). The texture of the deltopectoral crest of UMO BT 011 235.00 is slightly roughened (Fig. 14A). In UMO BT 011 221.00, the texture is heavily pitted as visible on the proximal head of the humerus (Figs 13, 14B). In both specimens (UMO BT 011 221.00 and UMO BT 011 235.00), the deltopectoral crest exhibits limited distal extension along the diaphysis and is confined to the proximal half of the humerus.

Figure 13. 

Close-up on the right forelimb of UMO BT 011 221.00 in ventral view. A. Photograph; B. Interpretative line drawing. Abbreviations: 2–4, distal carpals; II–IV, metacarpals; dc, deltoptectoral crest; H, humerus; i, intermedium; R, radius; re, radiale; U, ulna; ue, ulnare.

Figure 14. 

Morphology and pathologies of the proximal surface of the humeri in the Mistelgau Eurhinosaurus. A. Right humerus in UMO BT 011 235.00 in ventral view; B. Right humerus in UMO BT 011 221.00 in proximoventral view, C. Left humerus in UMO BT 011 235.00 in dorsal view showing a subsidence area indicative of avascular necrosis, D. Left humerus in Ichthyosaurus larkini in dorsal view with subsidence area compatible with avascular necrosis (photograph modified after Lomax and Massare 2017), E. Right humerus in Ichthyosaurus sp. in dorsal view with avascular necrosis (Copyright Pardo-Pérez et al. 2017, CC BY 4.0). The dashed lines highlight the zones of subsidence congruent with avascular necrosis. Abbreviations: dc, deltopectoral crest; dp, dorsal process.

Distally, the humerus is greatly expanded and bears two large articulation facets for articulation with the radius and ulna. These facets form an obtuse angle (Figs 3, 4, 13). The radial facet is slightly longer than the ulnar facet (Fig. 13). In distal view, both articulation facets are sub-rectangular to sub-pentagonal in shape and are concave. The texture of the articulation facets is roughened and pitted, but not to the extent observed in the proximal articulation facet in UMO BT 011 221.00. At midpoint of the humeral anteroposterior width, where the radial and ulnar facets meet, the texture is smoother.

Epipodials of the forelimb. In UMO BT 011 221.00, the radius is larger than the ulna and has a sub-rectangular shape (Fig. 13). It is thick and anteroposteriorly wider than proximodistally long (Fig. 13). The radius contacts the radiale distally, the intermedium posterodistally and the ulna posteriorly. The anterior margin of the radiale is notched (Figs 3, 12). The posterior margin of the radius is strongly convex in ventral view (Fig. 13). The ulna is polygonal and is as thick as the radius. However, both elements are thicker at their edges than at the center of their surface, giving their ventral surface a concave shape. This concavity of the surface is visible on all limb elements distal to the humerus. This is likely a compression artifact caused by trabecular collapse in the least dense part of the bone. The ulna contacts the ulnare distally and the intermedium anterodistally (Fig. 13). The ulna lacks a notch, and the posterior margin is convex (Fig. 13). The texture of the articular surfaces of the radius and the ulna is slightly pitted.

Carpals. UMO BT 011 221.00 preserves three proximal carpals (radiale, intermedium, and ulnare) and three distal carpals (Figs 3, 13). The intermedium has a deformed pentagonal shape and is slightly smaller than the ulna (Fig. 13). It is anteroposteriorly wider than proximodistally long (Fig. 13). The intermedium contacts the radius anteroproximally and the ulna posteroproximally. Anterodistally, the intermedium contacts the radiale, and posterodistally it likely contacts the ulnare when in articulation (Fig. 13). Distally, it contacts the distal carpal III (Fig. 13). The radial and ulnar facets are shorter than the facet for distal carpal III (Fig. 13). The facets for the radiale and ulnare are the shortest (Fig. 13).

The radiale is a sub-rectangular element about the same size as the intermedium and similar in shape to the radius (Fig. 13). As in the radius, the radiale is notched anteriorly (Fig. 13). The elements distal to the radiale lack notches (Fig. 13). The facet for the intermedium of the radiale is posteriorly slightly convex (Fig. 13). The anterior margin is longer than the posterior margin (Fig. 13). The texture of the articular facets of the radiale is slightly pitted. This condition is also visible in the ulnare. The radiale contacts the radius proximally, the distal carpal II distally, the intermedium posteroproximally, and the distal carpal III posterodistally (Fig. 13). The ulnare is oval and is the smallest proximal carpal (Fig. 13). It contacts the ulna proximally, the intermedium anteriorly, and distal carpal IV distally (Fig. 13). Since the ulnare is not preserved in perfect articulation, a contact with the distal carpal V is not visible but cannot be ruled out.

The distal carpals are slightly smaller than the proximal carpals (Fig. 13). The distal carpal II is the largest, and the distal carpal IV is the smallest (Fig. 13). The distal carpals, by their oval shape, are rounder than the proximal carpals. This follows the general trend in the shape of the paddle elements which become increasingly rounded distally (Fig. 13). The distal carpals are slightly thinner dorsoventrally than the more proximal elements, and the texture of their articular surfaces is slightly pitted.

Metacarpals and phalanges. Metacarpals II-IV are oval-to-round elements smaller than the proximal limb elements (Fig. 13). Their shape is similar to the shape of phalanges, making their identification in complete disarticulation of the limbs difficult. The articular surfaces of the metacarpals have a roughened texture and, in some places, are pitted. These elements are much thinner than the more proximal elements. However, all limb elements distal to the humerus are thicker at their edges than at their center, giving them a dished surface (Fig. 13). The proximal phalanges are oval, and the more distal the phalanges, the smaller, dorsoventrally thinner, and rounder they become (Fig. 13). The more distal preserved phalanges are completely discoidal in ventral view (Fig. 13). However, these phalanges are shaped more spool-like when observed in articular view. This shape is caused by a slight constriction of the articular surface compared to the ventral and dorsal margins of the phalanx.

Femur. Both femora are preserved in ventral view in UMO BT 011 221.00, and the left femur is preserved in dorsal view in UMO BT 011 235.00 (Figs 3, 4, 15). The shape of the femur resembles that of the humerus; however, the femur is smaller than the humerus (Figs 3, 4; Suppl. material 1: table S1). The femur is wider distally than proximally, with a proximally flat acetabular articulation facet. At the proximal end, it bears a ventral and a dorsal process. Both are positioned near the proximal margin, with the ventral process located closer to the anterior edge and the dorsal process extending across the entire anteroposterior width of the proximal femur. The ventral process is relatively small and proximodistally short (Fig. 15), while the dorsal process is more prominent and extends farther distally. The ventral process is primarily recognizable by the angle it forms where it contacts the acetabular facet (Fig. 15). Due to its offset from this facet, the ventral process appears triangular in lateral view (Fig. 15). Distally, the femur exhibits two articulation facets for the tibia and fibula, with the tibial facet being slightly larger than the fibular one.

Figure 15. 

Close-up of the left hindlimb of UMO BT 011 221.00 in ventral view. A. Photograph; B. Interpretative line drawing. Abbreviations: 4, fourth distal carpal; cal, calcaneum; F, femur; Fi, fibula; sle, semi-lunate element; te, tibiale; Ti, tibia; vp, ventral process.

Epipodials of the hindlimb. The tibia and fibula are well-preserved in semi-articulation only in the left hindlimb of UMO BT 011 221.00 (Fig. 15). Both tibia and fibula are preserved in ventral view. The tibia has an oval shape (Fig. 15). However, it has a straight distal margin, while the other margins are convex. The tibia is smaller than the fibula (Fig. 15). The articular surfaces of the tibia and fibula are roughened and slightly pitted. No notching is visible in the tibia.

The tibia contacts the femur proximally (Fig. 15). The tibia further contacts the tibiale distally (Fig. 15). A contact with the astragalus posterodistally is not preserved. The fibula has a rounded shape and is the largest epipodial of the hindlimb (Fig. 15). The fibula contacts the humerus proximally (Fig. 15).

Distal elements of the hindlimb. The limb elements distal to the epipodials (tarsals, metatarsals, and digits) are poorly preserved in both UMO BT 011 221.00 and UMO BT 011 235.00 (Fig. 15). Numerous elements are missing in both specimens, and the preserved elements are completely disarticulated and dispersed. Therefore, the exact number of digits and elements per digit in the hindlimbs is unknown. UMO BT 011 221.00 shows the best-preserved distal elements of the hindlimb. The tibiale of the left hindlimb of UMO BT 011 221.00 is preserved (Fig. 15). The tibiale presents four slight angles, giving it a sub-rectangular to oval shape (Fig. 15). It is anteroposteriorly wider than proximodistally long. All other elements present a rounded or oval outline (Fig. 15). This similarity in shape makes the precise identification of the disarticulated and dispersed elements difficult. Semi-lunate, notched anterior phalanges are preserved (Fig. 15). However, the tibia and tibiale are unnotched (Fig. 15). Distal phalanges are rounder and thinner than the more proximal elements. As in the forefin, the phalanges gradually decrease in size distally (Fig. 15).

Discussion

Generic affiliation of the Mistelgau specimens

As already mentioned, most of the cranial elements of the Mistelgau specimens are either preserved disarticulated in various views or in articulation. However, articulated skull elements are mostly visible from internal view, which makes the comparison of sutures and contacts difficult because most of the other described Eurhinosaurus specimens preserve cranial elements visible from external view (e.g., von Huene 1922, 1926, 1928, 1952; Maisch 2022).

The Mistelgau specimens are morphologically highly similar to Lower Toarcian Eurhinosaurus specimens. However, numerous characters diagnostic for the genus Eurhinosaurus are based on ratios (of skull or fin proportions) or cranial characters based on articulated skulls visible in external view (McGowan and Motani 2003; Maisch and Matzke 2000). Given the preservation status of the cranial material and fins in the Mistelgau specimens, these characteristics cannot be applied. Nevertheless, several other characters (see below) from McGowan and Motani (2003) and Maisch and Matzke (2000) allow us to test the specimens’ referral to the genus Eurhinosaurus. In the following paragraphs, we list these characters that permit us to identify the Mistelgau specimens and examine those requiring further discussion.

The specimens from Mistelgau, mostly based on the most recent diagnosis for Eurhinosaurus by McGowan and Motani (2003) (see systematic paleontology section), can be identified as representative of the genus Eurhinosaurus by the combination of the following characters: the mandible is considerably shorter than the skull, the presence of an unpaired internal carotid foramen in the parabasisphenoid (character taken from the diagnosis of Eurhinosaurus by Maisch and Matzke 2000), the presence of a tripartite pelvic girdle with the absence of fusion between the ischium and pubis, the presence of a long body more than 44 presacral vertebrae, the presence of sheathed-bicapitate dorsal ribs, and presence of ossified haemal arches in the anterior caudal vertebrae, the latter representing a new addition to the diagnostic characters.

The degree of disarticulation and incompleteness of the three specimens from Mistelgau obscures the characteristic overbite of Eurhinosaurus. Furthermore, due to the same reasons, no calculation of the overbite is possible, as it is defined to be more than 60% in Eurhinosaurus (McGowan and Motani 2003). However, based on the length of the premaxillae compared to the lower jaw elements in UMO BT 011 221.00, it can be deduced that a well-developed overbite was present, like in Eurhinosaurus (McGowan 1994) and the closely related Excalibosaurus (McGowan 1986, 1989a, 2003).

Although notching in fin elements of the leading digit is part of the diagnosis for the genus Eurhinosaurus (McGowan and Motani 2003) and notched elements are present in the specimens of this study (Figs 3, 13), the presence or absence of notching as well as the number of notched elements (when present) is highly variable in Eurhinosaurus and has already been reported in previous studies (McGowan 1994; McGowan and Motani 2003; Maisch 2022). We, therefore, do not consider the characters that consider the notching of fin elements to be truly diagnostic for Eurhinosaurus.

Notwithstanding, there are two diagnostic features (from the diagnosis of McGowan and Motani 2003) for the genus Eurhinosaurus that do not match the Mistelgau specimens. According to McGowan and Motani (2003), based on the specimens from Baden-Württemberg, Eurhinosaurus shows the presence of a basioccipital peg and an extensive extracondylar area lying largely ventral to the condyle, from which it is clearly set off. In the Mistelgau Eurhinosaurus, the basioccipital drastically differs in shape from the diagnosed condition for Eurhinosaurus. A basioccipital peg is absent, and the extracondylar area, despite being extensive, extends anteroposteriorly and not ventrally. Furthermore, the basioccipital condyle of the Mistelgau Eurhinosaurus further differs from known Eurhinosaurus specimens by being relatively larger and forming the entire posterior surface of the basioccipital.

Another notable difference observed in the Mistelgau specimens is the morphology and thickness of the ribs, which appear particularly thick and robust (Fig. 9G; Suppl. material 3: figs S6–S8) compared to other Eurhinosaurus specimens, especially the holotype material from Whitby. The Mistelgau ribs are morphologically more similar to specimens from Baden-Wuerttemberg, which display either grooved or ungrooved (oval in cross-section) ribs (e.g., von Huene 1922, 1926, 1928, 1951). It is possible that, in some of these southwestern German specimens, the longitudinal grooves may not be genuine anatomical features but rather taphonomic artifacts resulting from post-mortem compression. This is supported by the observation that other specimens from the same age and locality lack such grooves. Nonetheless, the presence of variation in rib morphology cannot be ruled out, and the potential taxonomic significance of such variation remains unknown.

Although the ribs of the southwestern German specimens generally appear thinner than those of the Mistelgau material, this difference might be influenced by preservation biases. The Mistelgau specimens preserve uncompressed ribs, while most other specimens from Baden-Wuerttemberg exhibit signs of compression. However, it is worth noting that compressed ribs are expected to display greater width due to collapse, which may complicate comparisons. Therefore, the apparent similarity in rib thickness between Mistelgau and some other southwestern German specimens may be misleading. Whether the differences in rib morphology between the holotype of Eurhinosaurus longirostris from England and the specimens from Baden-Wuerttemberg have taxonomic significance remains uncertain. A comprehensive revision of all available material would be required to resolve this, which falls outside the scope of the present study.

Despite the current ambiguity in interpreting rib morphology across German Eurhinosaurus specimens, the Mistelgau material can be clearly distinguished from other German specimens by its markedly different basioccipital morphology, as discussed above. Hence, although the basioccipital morphology does not align with the diagnosis of the genus Eurhinosaurus, and the ribs also differ from what is observed in at least the holotype specimen of E. longirostris, we consider that the Mistelgau specimens nevertheless belong to the genus Eurhinosaurus. This is supported by the large number of characters described above that are consistent with the genus Eurhinosaurus. However, the observed differences are considered to be significant at the species level. Consequently, the characters related to the shape of the basioccipital should no longer be included in the generic diagnosis of Eurhinosaurus but rather be incorporated into the specific diagnoses.

Comparative discussion of morphologically similar taxa

Temnodontosaurus azerguensis

Eurhinosaurus is the only ichthyosaur identified as leptonectid currently known from the Toarcian, increasing the evidence, in addition to the above-mentioned characteristics, that the specimens from Mistelgau belong to the genus Eurhinosaurus. However, Temnodontosaurus azerguensis, known from a unique specimen with an extremely slender and elongated rostrum (MAMSPLP), is also known from the Bifrons Zone (Martin et al. 2012). Furthermore, recent phylogenetic analyses have recovered Temnodontosaurus azerguensis as a close relative of Eurhinosaurus (Maxwell and Cortés 2020; Laboury et al. 2022). It is therefore necessary to compare the Mistelgau specimens with this species. According to Martin et al. (2012), Temnodontosaurus azerguensis differs from Eurhinosaurus in having a paired internal carotid foramen, numerous hexagonal elements in the fins, and a constricted humerus. Based on the figure provided by Martin et al. (2012), the humerus of Temnodontosaurus azerguensis shows a comparable constriction to the humeri of the Mistelgau specimens. Maisch (2022) pointed out that the shape of the humerus and femur in Eurhinosaurus can greatly vary. We agree with the latter, and therefore, characters based on humeri and femora should be treated with caution when applied in taxonomic and phylogenetic analyses.

As mentioned above, the Mistelgau specimens present numerous rounded elements in the fins and an unpaired internal carotid foramen, which contrasts with the description of Temnodontosaurus azerguensis by Martin et al. (2012). Nevertheless, these differences lack evidence. Based on the photographs and illustrations provided in Martin et al. (2012), the hexagonal elements of the fin do not differ from the range of fin elements’ morphology observable in Eurhinosaurus and the Mistelgau Eurhinosaurus. Furthermore, based on the illustrations from the study, the vast majority of the fin elements look oval and round with the presence of smooth angles in some elements (as visible in the proximal fin elements in the Mistelgau Eurhinosaurus, see Figs 13, 15) but no significant and clearly hexagonal elements are observable.

We did not observe a paired carotid foramen in the parabasisphenoid of Temnodontosaurus azerguensis as previously described and illustrated by Martin et al. (2012). Instead, our examination suggests a single, large, and deep carotid foramen located approximately at the center of the parabasisphenoid, consistent with the unpaired condition observed in Eurhinosaurus. We note a smaller, shallow depression adjacent to this foramen, which may have been misinterpreted as a second carotid foramen in the original study, leading to the description of a paired condition of the carotid foramen in T. azerguensis. Thus, the morphological differences initially noted by Martin et al. (2012) between T. azerguensis and Eurhinosaurus are neither fully consistent nor clearly substantiated.

Additionally, based on the illustrations provided by Martin et al. (2012), T. azerguensis appears to lack an overbite. Martin et al. (2012) state that the dentary and premaxilla are both preserved to their respective tips but both are indistinguishable. Despite this assertion, the illustration (Martin et al. 2012, Fig. 4) labels both bones distinctly, despite a questioned identification for the dentary. Given their indistinguishability, we question the identification of the fragment labeled as the dentary. This fragment, preserved at the snout’s tip adjacent to the premaxilla, may instead represent the second premaxilla. If this is the case, the two elongated bones forming the snout’s tip in T. azerguensis may both be premaxillae, suggesting the potential presence of an overbite. However, due to the fragmented and poorly preserved nature of the skull, any morphological interpretation warrants cautious consideration until a detailed revision of the holotype material is provided.

In conclusion, the morphological distinctions proposed by Martin et al. (2012) to differentiate T. azerguensis from Eurhinosaurus, such as humeral constriction, parabasisphenoid carotid foramen arrangement, and fin element outlines appear inconsistent. These inconsistencies, along with the possible presence of an overbite in T. azerguensis, cast doubt on the validity of T. azerguensis as a distinct taxon and suggest it may represent a specimen of Eurhinosaurus or at least a closely related taxon. Finally, based on figure 2 of Martin et al. (2012), T. azerguensis exhibits a basioccipital with an extensive dorsoventrally oriented extracondylar area, a feature diagnostic of Eurhinosaurus yet absent in the Mistelgau Eurhinosaurus specimens. This observation suggests that the Mistelgau Eurhinosaurus is likely not a representative of the same taxon as T. azerguensis, at least at the specific level.

Excalibosaurus costini

Despite being temporally separated, Eurhinosaurus (Toarcian) and Excalibosaurus (Sinemurian) are frequently considered closely related due to their strikingly similar morphology, particularly the prominent overbite they both share. McGowan (1986, 1989) even suggested Excalibosaurus as a possible ancestor of Eurhinosaurus, and Maisch and Matzke (2000) went further, synonymizing Excalibosaurus with Eurhinosaurus, arguing that the only distinguishing feature between them is the relative length of the overbite. Given the high morphological similarity, we compare Excalibosaurus costini (monospecific) with the Eurhinosaurus specimens from Mistelgau. According to McGowan and Motani (2003), who rejected the synonymy of the two genera claimed by Maisch and Matzke (2000), Excalibosaurus can be distinguished from Eurhinosaurus based on several characteristics: a less developed overbite, a preflexural vertebral count of 98 (as opposed to the 91–95 range in Eurhinosaurus; McGowan and Motani 2003 argued that vertebral count is relatively stable among ichthyosaurs), smaller hindfins compared to the forefins, a forefin that is less long and slender than in Eurhinosaurus (with its length being less than half that of the skull), a relatively longer body compared to Eurhinosaurus, and a significant temporal gap of 16 million years.

Despite the remarkable preservation of the Mistelgau Eurhinosaurus, the incompleteness of the specimens (e.g., lack of a complete and well-articulated skull, fins, or vertebral column) limits detailed comparisons with the above-mentioned distinguishing characters. The only confirmable differences are the presence of a well-developed basioccipital peg in Excalibosaurus costini (McGowan 2003), which is absent in the Mistelgau Eurhinosaurus, and their temporal separation, although the possibility that Excalibosaurus persisted into the Toarcian cannot be entirely excluded.

The holotype of Excalibosaurus costini (BRSMG Cc881) presents highly angular forefin elements, such as a rectangular radiale, which contrasts with the morphology observed in the Mistelgau Eurhinosaurus. However, because forefin element shape can exhibit variation within Eurhinosaurus, this feature is not considered a definitive distinguishing characteristic, although it remains a potential area for future confirmation.

The most recent diagnosis of Excalibosaurus by McGowan and Motani (2003) suggests an orbital ratio probably less than 0.20, which contrasts with the ratio in Eurhinosaurus, where it is equal to or greater than 0.20. However, due to the preservation state of the Mistelgau Eurhinosaurus, comparison based on these ratios is not possible. Additionally, the fore- and hindlimb size disparity (with the forelimbs being larger) diagnosed in Excalibosaurus mirrors the same pattern in the Eurhinosaurus from Mistelgau, though limb size proportions show variability even among Eurhinosaurus specimens (e.g., NHMUK PV R 5465, GPIT-PV-30024). Based on McGowan (2003), the pectoral girdle of Excalibosaurus is relatively well-developed compared to body size, whereas the Eurhinosaurus pectoral girdle appears smaller in relation to body size. However, we do not observe a distinct difference in pectoral girdle development between the Eurhinosaurus from Mistelgau. and Excalibosaurus.

The most evident difference between the two taxa is the degree of the overbite. In Eurhinosaurus, the mandible is less than 60% of the skull length, whereas in Excalibosaurus, it exceeds 60% (McGowan and Motani 2003). However, the exact degree of the overbite cannot be assessed in the Mistelgau Eurhinosaurus due to the incomplete skulls of the available specimens.

In conclusion, the comparison of the morphology between the Eurhinosaurus specimens from Mistelgau and Excalibosaurus is constrained, in part because of the preservation state of the Mistelgau material and the limited morphological description of Excalibosaurus. Consequently, while Excalibosaurus and Eurhinosaurus share a highly similar overall morphology despite significant temporal separation, their generic distinction remains unresolved. Nevertheless, both taxa are considered distinct at the species level, primarily based on the presence of a well-developed basioccipital peg in Excalibosaurus costini (McGowan 2003), a feature absent in the Mistelgau Eurhinosaurus.

Specific affiliation of the Mistelgau Eurhinosaurus

Although the assignment of the Mistelgau specimens to the genus Eurhinosaurus is well-substantiated, determining their specific affiliation is more complex, owing to the controversial validity of certain species included in the genus (see Introduction). As discussed earlier in the sections on generic affiliation, the Mistelgau Eurhinosaurus shares several diagnostic features with Eurhinosaurus longirostris. However, notable differences indicative of specific-level distinction have been identified, as outlined above. These include the morphology of the basioccipital, which, based on the diagnosis by McGowan and Motani (2003), deviates from that of both the genus Eurhinosaurus and the species Eurhinosaurus longirostris, as well as the unusual rib thickness and robustness.

The most recent published diagnosis for Eurhinosaurus huenei (Maisch 2022) cannot be completely applied to the Mistelgau specimens, since most characters are based on ratios of complete fins. Additionally, the remaining characters of the revised diagnosis of Eurhinosaurus huenei by Maisch (2022) are imprecise and applicable with difficulty to other specimens. This concerns the following remaining characters diagnostic for Eurhinosaurus huenei: elongated and slender forefins, and the neural spines in the middle trunk area, which are not massively shortened (Maisch 2022). Based on personal observations, all specimens of Eurhinosaurus show somewhat elongated and slender forefins. We, therefore, consider this character not to be pertinent for species differentiation within Eurhinosaurus without any further precision. The same applies to the neural spines from the middle trunk area that are not massively shortened in Eurhinosaurus huenei according to Maisch (2022). Without further precision, we do not consider this character as being applicable to differentiate Eurhinosaurus species. Therefore, the unique character from the revised diagnosis of Eurhinosaurus huenei provided by Maisch (2022) that is applicable for comparison to the Mistelgau Eurhinosaurus is the number of presacral vertebrae. According to Maisch (2022), this number is higher than 40 in Eurhinosaurus huenei and this is also true for the Mistelgau specimens. Nevertheless, based on the diagnosis provided by Maisch (2022) we cannot further assess the potential affiliation of the Mistelgau specimens to Eurhinosaurus huenei. However, as mentioned above, Eurhinosaurus longirostris was considered the sole species within Eurhinosaurus until the erection of Eurhinosaurus quenstedti by Maisch (2022). Therefore, the diagnosis of the genus and species Eurhinosaurus longirostris from McGowan and Motani (2003) is mostly based on German specimens and also applies to the specimens now considered to represent Eurhinosaurus huenei by Maisch (2022). Consequently, comparing the Mistelgau Eurhinosaurus with Eurhinosaurus huenei from Maisch (2022) is the same as comparing them to Eurhinosaurus longirostris based on McGowan and Motani (2003).

Except for the already mentioned differences, most notably the morphology of the basioccipital, the Mistelgau Eurhinosaurus are similar to the Baden-Wuerttemberg Eurhinosaurus, regardless of whether they are considered Eurhinosaurus longirostris or Eurhinosaurus huenei. However, the markedly distinct basioccipital morphology observed in the Mistelgau Eurhinosaurus enables taxonomic differentiation from the other German Eurhinosaurus specimens, regardless of their specific assignment.

As noted by Maisch (2022), the holotype specimen of Eurhinosaurus longirostris from Whitby, England (NHMUK PV OR 14566), lacks a detailed description, complicating direct comparisons with the Mistelgau specimens. Consequently, the degree of morphological similarity between the Mistelgau Eurhinosaurus and the Whitby material remains uncertain. Nonetheless, first-hand examination of the holotype, despite its poor preservation and current challenging storage conditions, has confirmed that the rib morphology differs from that of the Mistelgau specimens, as detailed above. This distinction supports the taxonomic separation of the Mistelgau Eurhinosaurus from Eurhinosaurus longirostris.

Referral of the Mistelgau Eurhinosaurus specimens to the most recently named species Eurhinosaurus quenstedti from southern Germany cannot be excluded solely based on the diagnosis provided by Maisch (2022). Note that the holotype of Eurhinosaurus quenstedti (SMNS 81842) was not a newly found specimen, but an already known specimen that was thus considered as Eurhinosaurus longirostris until the assignment to Eurhinosaurus quenstedti by Maisch (2022). Eurhinosaurus quenstedti is mainly diagnosed by features based on fin proportions that cannot be applied to the Mistelgau specimen, due to the lack of good preservation in the hindlimbs and distal portion of the forefins in the latter specimens. The other character of Eurhinosaurus quenstedti that is not based on the appendicular skeleton comprises, according to Maisch (2022), neural spines in the middle trunk area, which are very blunt and short. The neural spines of the middle trunk portion of the Mistelgau Eurhinosaurus, from our perspective, do not show a morphology or height that drastically differs from either Eurhinosaurus quenstedti, described as very short and blunt, or other specimens of Eurhinosaurus from Baden-Wuerttemberg. Without further precision or measurements, we do not consider the above-mentioned character diagnostic of Eurhinosaurus quenstedti, provided by Maisch (2022), to be truly useful for comparing and distinguishing Eurhinosaurus species. Nevertheless, Eurhinosaurus quenstedti shares the same basioccipital and rib morphology as the other southwestern German Eurhinosaurus specimens discussed above, and thus differs from the condition observed in the Mistelgau Eurhinosaurus, which excludes the affiliation of the Mistelgau specimens to Eurhinosaurus quenstedti.

In conclusion, despite the ongoing taxonomic uncertainty surrounding Eurhinosaurus material from southwestern Germany and England, comparative analysis indicates that the Mistelgau specimens differ from these stratigraphically older Eurhinosaurus specimens. We therefore assign the Mistelgau material to a new species, Eurhinosaurus mistelgauensis sp. nov.

Taxonomic consistency across the Mistelgau Eurhinosaurus specimens

Given that the Eurhinosaurus specimens from Mistelgau originate from two different stratigraphic layers that slightly differ in time (Vitiosa Subzone for UMO BT 011 235.00; Thouarsense Subzone for UMO BT 011 221.00 and UMO BT 011 240.00) (Fig. 2), it is important to determine whether all specimens belong to the same taxon or if multiple taxa are represented.

The most complete specimens, UMO BT 011 235.00 and UMO BT 011 221.00 do not exhibit significant morphological differences. The comparison of UMO BT 011 240.00 with the other two is more challenging due to its fragmentary nature, as it preserves only the premaxillae with teeth and rib fragments. However, the morphology of these elements is consistent with that of the other specimens, showing no distinguishing characteristics.

Among the three specimens, the only notable difference concerns the shape of the humeri between UMO BT 011 235.00 (the stratigraphically older specimen) and UMO BT 011 221.00 (the stratigraphically younger specimen).

The differences observed in the humeri of UMO BT 011 235.00 and UMO BT 011 221.00 are likely pathological in origin (see the section below on pathologies), rather than taxonomically significant.

Additionally, the basioccipital is difficult to compare between these two specimens, as it is not preserved in the same orientation. In UMO BT 011 235.00, it is primarily exposed in ventral view, with its lateral margins also visible, whereas in UMO BT 011 221.00, only the dorsal view is preserved. Consequently, it remains uncertain whether the condition of the extracondylar area (mainly the complete absence of ventral extension making the ECA not visible in posterior view), observed in UMO BT 011 235.00 and unique among Eurhinosaurus, is also present in UMO BT 011 221.00. However, the dorsal portion of the condyle appears to span the entire width of the basioccipital, as seen in UMO BT 011 235.00, suggesting a similar overall morphology.

Beyond these aspects, the specimens either show no differences or preserve skeletal elements absent in the other specimens, making direct comparisons for those bones impossible. We therefore conclude, based on the strong morphological similarities, that all three Eurhinosaurus specimens from Mistelgau belong to the same taxon, Eurhinosaurus mistelgauensis sp. nov., which has a stratigraphic range extending from the Vitiosa Subzone to the Thouarsense Subzone (see the geology section).

Size and ontogeny

UMO BT 011 235.00 is the largest specimen from Mistelgau with a measured preserved length of 419 cm. UMO BT 011 221.00 measures 355 cm, as preserved (Tab. 1). Both specimens are incomplete and were originally larger, since the snout is almost completely missing in UMO BT 011 235.00, as well as the posterior caudal portion in UMO BT 011 221.00 (Figs 3, 4). Since solely a portion of the snout is preserved in UMO BT 011 240.00, no body length measurement is possible. However, its incompletely preserved premaxilla is approximately 63.25 cm long which is slightly longer than the greatly broken premaxilla in UMO BT 011 221.00 (Suppl. material 1: table S1). Interestingly, UMO BT 011 235.00 shows lengths and widths of the stylopodial elements slightly shorter than in UMO BT 011 221.00. Compared to a selection of known Eurhinosaurus specimens, the Eurhinosaurus from Mistelgau appear to be among the smaller representatives of the genus, based on their body length (Tab. 1). However, note that the comparison is limited since the Mistelgau specimens are not complete. That the Mistelgau specimens are among the smaller individuals of Eurhinosaurus is also true when comparing the lengths and widths of the stylopodium, pectoral girdle, and pelvic girdle of the Mistelgau specimens with the available measurements for other Eurhinosaurus specimens in the literature (e.g., von Huene 1922, p. 35, 1926, p. 76, 1928, p. 472–473, 1951, p. 282; Maisch 2022, p. 121, tab. 1).

The body size of fairly complete Eurhinosaurus specimens documented in the literature ranges from approximately 385 to 712 cm (Fig. 16A; Suppl. material 2: table S2). Using a linear regression model (Fig. 16B), we estimate the body length of UMO BT 011 235.00 and UMO BT 011 221.00 based on their humeral lengths (see Suppl. material 1: table S1 for measurements). The resulting approximate values are 423 cm for UMO BT 011 235.00 and 465 cm for UMO BT 011 221.00 (Fig. 16B; see Suppl. material 2: table S3 for exact values, standard error and 95% confidence interval estimates). The mean estimated body length of approximately 423 cm for UMO BT 011 235.00 appears unexpectedly low. Even when accounting for the standard error (Suppl. material 2: table S3), which increases the estimated length to approximately 439 cm, the result remains surprising given that the preserved portion of the specimen already measures about 419 cm, despite missing most of the snout. This suggests that the total body length was likely somewhat greater than the mean estimate. This discrepancy probably highlights variability in the relationship between humerus length and body length, indicating that humerus length may exhibit a degree of plasticity among individuals of similar size.

Figure 16. 

The size of complete Eurhinosaurus specimens (list of selected best-preserved specimens and measurements provided in the Suppl. material 2: table S2). A. Boxplot of body length of Eurhinosaurus specimens. The boxplot displays the distribution of body length (in cm) across a sample of complete Eurhinosaurus specimens. The central box represents the interquartile range (IQR), with the lower and upper edges corresponding to the first (Q1) and third (Q3) quartiles, respectively. The horizontal line inside the box represents the median. Whiskers extend from the box to the minimum and maximum values within the data, with outliers marked in red. The black “X” symbol indicates the mean body length, and the numerical values for the minimum, Q1, median, Q3, and maximum are annotated adjacent to the plot. B. Scatter plot showing the relationship between humerus length (cm) and body length (cm) in a selection of complete Eurhinosaurus specimens and estimation of body length of UMO BT 011 235.00 and UMO BT 011 221.00 based on a linear regression model (y = 0.02958x – 1.30029; R² value of 0.9043; p-value = 0.00028).

The specimens of Eurhinosaurus mistelgauensis sp. nov. are not the smallest representatives of Eurhinosaurus but fall within the lower range of known body sizes (Fig. 16B). The comparatively smaller body size of the Mistelgau Eurhinosaurus does not necessarily indicate osteologically immature individuals. The Mistelgau specimens are all well-ossified and show no morphological features that could be assigned to osteological immaturity (e.g., sensu Miedema and Maxwell 2019 for Stenopterygius). On the contrary, based on Miedema and Maxwell (2019), the Mistelgau specimens show certain features that are characteristic of adult specimens in Stenopterygius, such as the relatively small size of the exoccipitals, a well-defined V-shaped impression of the membranous labyrinth, and a well-demarcated floor of the foramen magnum on the dorsal surface of the basioccipital. Note that the presence of an anterior notochord pit in the basioccipital in UMO BT 011 221.00 could suggest an early developmental stage. However, this feature is only visible prenatally in Stenopterygius (Miedema and Maxwell 2019), whereas it is retained into maturity in Argovisaurus (Miedema et al. 2024). In UMO BT 011 235.00, the proximal head of the humeri is convex, which is, according to Johnson (1977), a characteristic of mature individuals in Stenopterygius. The convex proximal head of the humeri in UMO BT 011 235.00 thus could further indicate that the latter specimen was a mature individual. However, it is important to note that Stenopterygius is phylogenetically distant from Eurhinosaurus and morphologically different, which might limit the comparison regarding their respective ontogeny. Furthermore, some morphological aspects can vary due to taxonomy, taphonomy, or pathology and limit comparisons.

Pathologies

As mentioned in the description, the morphology of the proximal head of the humeri in the Mistelgau Eurhinosaurus shows some peculiarities. In UMO BT 011 221.00, the proximal humeral head greatly differs from what can frequently be observed in ichthyosaurs (McGowan and Motani 2003) and looks abnormal. In this specimen, as previously mentioned, the proximal head of the humerus is completely flat in both humeri and shows a heavily pitted texture on its articular surface (Fig. 14B). The pits are extremely large (up to 1 cm in diameter) and deeply concave, resembling small craters (Fig. 14B). This morphology is atypical; however, we cannot conclusively determine whether it originates from a pathological condition. Among the Mistelgau Early Jurassic vertebrate fauna, pathologies in ichthyosaurs have already been reported in Temnodontosaurus specimens showing a high number of diverse pathologies (Pardo-Pérez et al. 2017, 2018). However, none of the reported pathologies in the Mistelgau Temnodontosaurus (Pardo-Pérez et al. 2018) appears to correspond to the morphology of the humerus in UMO BT 011 221.00. The condition visible in UMO BT 011 221.00 resembles to some extent the condition visible in humeri of Stenopterygius showing articular pathologies (Pardo-Pérez et al. 2017: fig. 5). However, the pits in UMO BT 011 221.00 appear larger and deeper. Pathologies are frequent in the forelimbs of ichthyosaurs, with a high tendency for avascular necrosis (Pardo-Pérez et al. 2017). However, without further histological analysis of the humeri of UMO BT 011 221.00, which is beyond the scope of this study, no further paleopathological inferences are possible for this specimen. In addition to the humeral abnormalities, UMO BT 011 221.00 also preserves a cervical rib with unusual and presumably pathological morphology (see description) (Fig. 9H), which is interpreted as pseudarthrosis, based on comparison with pseudarthrotic ribs documented in other ichthyosaurs that show similar morphology to the Mistelgau specimens (Pardo-Pérez et al. 2017).

In specimen UMO BT 011 235.00, the left humerus exhibits a localized zone of subsidence, characterized by a pronounced concavity on the proximal head proximal to dorsal process (Fig. 14C). In contrast, the corresponding area in the right humerus appears relatively flat, lacking any discernible concavity, despite the presence of well-developed bone surface rugosity (Fig. 14A). This distinct morphological feature of the left humerus closely resembles the condition observed in the femora of ichthyosaurs exhibiting avascular necrosis (Fig. 14D, E) (Pardo-Pérez et al. 2017: fig. 4; Rothschild et al. 2012: figs 1, 2). Therefore, the subsidence zone in the left humerus of UMO BT 011 235.00 is interpreted as indicative of avascular necrosis. Similar instances of avascular necrosis have been documented in other ichthyosaur taxa of the Early Jurassic, such as Leptonectes, Ichthyosaurus, and Temnodontosaurus (Motani et al. 1999; Rothschild et al. 2012; Pardo-Pérez et al. 2017), as well as in other geologically younger ichthyosaurs (see Motani et al. 1999; Rothschild et al. 2012; Arkhangelsky and Zverkov 2014). The presence of avascular necrosis has been suggested as proxy for inferring deep-diving behaviors in aquatic animals (Motani et al. 1999; Rothschild et al. 2012). Consequently, this morphological evidence suggests that Eurhinosaurus mistelgauensis sp. nov. may have also exhibited deep-diving behavior.

Taphonomy of the Mistelgau specimens

Specimen UMO BT 011 221.00. As mentioned in the material section, the specimen is mostly complete, with the exception of most of the distal caudals. Articulation, or semi-articulation (i.e. preservation of elements in natural position with slight displacement) is only seen in the pectoral girdle, forelimbs, the left side of the ribcage, and some caudal vertebrae. As it was prepared from the stratigraphic underside, the carcass shows a ventral landing on the seafloor. It is likely that the pectoral girdle was in direct contact with the sediment, resulting in relatively good preservation. The same pattern is seen in the ribs, especially on the left side of the body. A relatively well-preserved ribcage with well-preserved paddles and a fully disarticulated vertebral column is indicative of a ventral embedding position, according to Hofmann (1958) and illustrated by Martill (1993). Most of the missing skull elements are part of the skull roof, the portion of the skull that was the furthest away from the sediment, allowing the individual bones to disperse after disarticulation along the suture and subsequent collapse. The dispersion of the dorsal ribs of the posterior abdominal region is prominent and might be caused by a rupture of the body cavity resulting in ongoing decay and subsequent transport by current activity. The left side of the ribcage being almost fully intact hints at a leftward position on the sediment. This is further supported by the slightly better preservation of the left hindlimb and the dispersion of the ribs of the right side of the body. With the heavy pectoral region acting as an anchoring point (see Hofmann 1958) to the substrate, this area has been least affected by current activity. The rest of the carcass, especially the posterior dorsal and anterior caudal regions, did not adhere to the substrate and was susceptible to extensive displacements. The posterior-most caudals (Suppl. material 3: figs S1, S5) display semi-articulation, resulting from the ventral position of the carcass on the seafloor. The ventral, bony lobe of the semi-lunate tailfin was likely to have sunk into the substrate to some degree protecting the distal caudals from currents.

In contrast to other specimens of Eurhinosaurus, especially from the Posidonienschiefer Formation, the bones are generally three-dimensionally preserved. No abrasion caused by prolonged seafloor exposure is observed, either on the visible side nor on the single bone prepared from both sides, the parabasisphenoid (Fig. 8C, D). Compression and deformation of bones appear to be limited as overlapping bones do not bend or break according to the shape of the under- or overlying element with the exception of the very thin gastralia. Epibionts and associated fauna that clearly indicate a carcass-fall community (e.g. Dick 2015; Maxwell et al. 2022) are not identified.

Specimen UMO BT 011 235.00. In terms of completeness, the stratigraphically slightly older UMO BT 011 235.00 is highly comparable to UMO BT 011 221.00. In contrast to the latter specimen, it is prepared from the stratigraphic upper surface in order to preserve the directly underlying belemnite battlefield. Although almost entirely disarticulated, the bones are closely associated, and some retain their natural position. The only articulation is found within the skull, mainly the right side, and the posterior-most caudals (postflexural centra) (Fig. 7; Suppl. material 3: fig. S9). As the skull is seen in left lateral view, and the ribs are much better preserved than the vertebral column, a right lateral embedding combined with a slight ventral component is hypothesized. However, the dispersion of the pectoral girdle contradicts this hypothesis so that the dorsoventral component cannot be identified unambiguously. The right lateral position is confirmed by the position and direction of the preserved tail bend. It is likely that the skull acted as an anchoring point in the sediment, whereas everything posterior to it did not sink into the substrate. Whereas the right side of the skull is mostly complete and well-articulated, the left side is very incomplete and heavily disarticulated. It is likely that the soft sediment into which the carcass sank was not as thick as the width of the skull, leaving the left side exposed to current activity for a prolonged time period. Although the vertebral column is not highly flexible, it is likely that the posterior caudal region bent downward to form a secondary anchoring point on the sediment after initial landing on the substrate. This must have happened relatively early, as the articulation indicates flesh was still attached to the tailfin upon embedding. An anterior skull-first landing is possible, as the premaxillae seem to bend down into the belemnite battlefield layer. However, due to the relatively reduced weight of the skull, it could not penetrate in a similar way to Temnodontosaurus from Mistelgau (Pardo-Peréz et al. 2018, Fig. 9). Although a belemnite battlefield is often associated with a high-energy environment in which winnowing or redeposition takes place (e.g. Doyle and Macdonald 1993) and does not often preserve semi-articulated remains, the completeness is remarkable. It is interpreted here that the formation of the belemnite battlefield was in its final phase, as almost no belemnites were deposited on top of the Eurhinosaurus skeleton. Nevertheless, the bones seem to have been abraded to a slightly higher degree than in UMO BT 011 221.00. Admittedly, the preparation of the surface exposed to the water column generally does not favor the best surface-preservation of the bones. No obvious epifauna is present on the bone surface. A single tooth of Temnodontosaurus sp. is associated with the skeleton, but no direct evidence of scavenging can be observed. The compression imprinting of over-and underlying bones is slightly more present as in UMO BT 011 221.00, but still not to a degree comparable to most Posidonienschiefer Formation specimens (Martill 1993; Cooper and Maxwell 2022). Although the observed compaction of ichthyosaur skeletons in the Posidonienschiefer Fm. is variable between stratigraphic horizons, three-dimensionally preserved skeletons are outnumbered by far.

Specimen UMO BT 011 240.00. Of the three specimens of Eurhinosaurus, UMO BT 011 240.00 is by far the least complete. However, it shows the semi-articulation of both premaxillae with some of the teeth still implemented in the alveolar groove. A lateral embedding position of the skull is likely based on the position of the premaxillae. The presence of a dorsal rib on top of the premaxillae suggests an overall state of disarticulation, which conforms to the general trend seen in the Jurensismergel Formation. Whether the rest of the body was dispersed in a similar fashion as in the other two specimens cannot be determined beyond speculation. Solely the disarticulation of the premaxillary suture can be observed, which caused the right premaxilla to have shifted dorsoposteriorly in relation to its left counterpart. Whether this occurred on the sediment-water interface or post-burial cannot be determined.

Overall preservation compared to the Posidonienschiefer Formation. Sachs et al. (2024) attribute the extensive disarticulation of the holotype of Franconiasaurus brevispinus, a plesiosaur from the Mistelgau clay pit, to slow sedimentation rates and current activity. The disarticulation patterns observed in UMO BT 011 221.00, with partially articulated portions and heavy disarticulation otherwise, are very similar to those observed in the holotype of Franconiasaurus. As these specimens are from the same stratigraphic level (Thouarsense Zone) a similar taphonomic setting is highly plausible. Highly complete skeletons with a low rate of articulation are indicative of an absent- or very short floating phase, which has been reported for Ichthyosauria from various deposits (Beardmore and Furrer 2016). This is in accordance with the available data for the Posidonienschiefer Formation.

A major difference between the Jurensismergel and the Posidonienschiefer Formation is the thickness of the finely laminated sediments. Both the Variabilis and Thouarsense Zones are very thin (5–10 cm) (Schulbert 2001; Sachs et al. 2024), especially in contrast to the much thicker vertebrate-bearing (sub)zones of the Posidonienschiefer Formation (e.g. Röhl et al. 2001). Although the sedimentation rates in the Posidonienschiefer Formation have been similarly low (Röhl et al. 2001; Reisdorf et al. 2012; Ruebsam et al. 2024), the preservation of marine vertebrates tends to reach perfect completeness and articulation in some specimens, in contrast to the Jurensismergel Formation. Hofmann (1958) and Martill (1993) attribute this preservation to the presence of a soft substrate in which the carcasses could sink in, partially or completely, resulting in exceptional preservation. Depending on the thickness of this layer, Hofmann (1958) illustrates that the articulation and preservation are influenced heavily. Judging by the absence of perfectly articulated material from Mistelgau, and a likely prolonged exposure on the sediment, it can be assumed that the soft layer in the Mistelgau depositional environment must have been much thinner than in the Posidonienschiefer Formation of southern Germany and similar Lagerstätten with pristine preservation (Martill 1993). Based on the articulation of the body-areas that are assumed to have sunken into the sediment, the soft layers of the Variabilis and Thouarsense Zones would have likely had a thickness less than the width or dorsoventral height of the skull of an adult Eurhinosaurus.

Despite articulation being somewhat inferior to a similar fauna from the Posidonienschiefer Formation, the three-dimensional preservation of the elements is much better in the specimens from the Mistelgau pit. This is mainly due to the extremely high (90%) compaction rate of the Posidonienschiefer Formation (Martill 1993), which seems to have been less pronounced in the Jurensismergel Formation. Furthermore, massive improvements in preparation techniques allow for the visualization of fine structures that had been damaged by crude preparation applied to historic specimens. This allows us to describe delicate skull bones in great detail.

Conclusion

In conclusion, three ichthyosaur specimens from the Mistelgau clay pit can be identified as belonging to the genus Eurhinosaurus. These specimens were found in layers from the early Upper Toarcian (Variabilis Zone and Thouarsense Zone) and thus represent the youngest stratigraphic occurrence of the genus. The preservation of these specimens is exceptional, presenting the bones in three dimensions, which is extremely rare and underlines the importance of describing these specimens. However, the different preservation and thus views of exposed bones often hamper detailed comparison to other Eurhinosaurus specimens. Nevertheless, the Mistelgau specimens show differences from other described species (independent if currently valid or not) of Eurhinosaurus, notably a basioccipital whose morphology differs greatly from previously known material mainly by the presence of an anteroposteriorly extending ventral extracondylar area (in contrast to the dorsoventrally oriented ECA in Eurhinosaurus longirostris) and thickened and robust ribs. Therefore, despite the pending revision of the holotype Eurhinosaurus material, the distinct morphological differences observed in the Mistelgau specimens justify their attribution to a new species within the genus Eurhinosaurus. Given their excellent preservation and diagnostic features, we propose the establishment of a new taxon, Eurhinosaurus mistelgauensis sp. nov. This new species provides valuable insights into the morphology of this enigmatic ichthyosaur.

Funding

Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – Project number 461387258 (SA 469/59-1) (FM and JH). Additionally, JH has received funding from the University of Bonn.

Authors’ contributions

GES and NK designed the study. GES, JH, FM interpreted the anatomy and conducted direct, first-hand examinations of the specimens described. JH interpreted the taphonomy and wrote the respective chapter. GES prepared the figures. GES wrote the first draft of the manuscript. All authors completed, read, and approved the final manuscript.

Competing interests

The authors declare that they have no conflict of interest.

Acknowledgments

We thank Joachim Rabold, Ulrike Albert, and Stefan Eggmaier from the Urwelt-Museum Oberfranken Bayreuth for providing access to specimens under their care and for kind support during our visits. Our thanks to Stefan Eggmaier for his excellent preparation of the specimens. Furthermore, we would like to thank Valentin Fischer (University of Liège) for general discussion. GES wants to thank Paulina Jimenez-Huidobro (University of Bonn) for general assistance during this study. FM wants to thank Matt Riley (Cambridge Sedgwick Museum) and Deborah Hutchinson (Bristol Museum and Art Gallery) for their assistance in collection visits studying Leptonectidae material relevant to this study. JH and GES thank Ingmar Werneburg and Henrik Stöhr (both University of Tübingen) as well as Marc Jones (Natural History Museum London) for access to the collections including Eurhinosaurus material. We thank the two anonymous reviewers for their constructive comments, which greatly improved the manuscript.

References

  • Abel O (1909) Cetaceenstudien. 1. Mitteilung: Das Skelett von Eurhinodelphis cocheteuxi aus dem Obermiozän von Antwerpen. Sitzungberichte der kaiserlichen Akademie der Wissenschaften, Wien 118: 241–253.
  • Anderson KL, Druckenmiller PS, Erickson GM, Maxwell EE (2019) Skeletal microstructure of Stenopterygius quadriscissus (Reptilia, Ichthyosauria) from the Posidonienschiefer (Posidonia Shale, lower jurassic) of Germany. Palaeontology 62(3): 433–449. https://doi.org/10.1111/pala.12408
  • Beardmore SR, Furrer H (2016) Evidence of a preservational gradient in the skeletal taphonomy of Ichthyopterygia (Reptilia) from Europe. Palaeogeography, Palaeoclimatology, Palaeoecology 443: 131–144. https://doi.org/10.1016/j.palaeo.2015.11.049
  • Bindellini G, Wolniewicz AS, Miedema F, Scheyer TM, Dal Sasso C (2021) Cranial anatomy of Besanosaurus leptorhynchus Dal Sasso and Pinna, 1996 (Reptilia: Ichthyosauria) from the Middle Triassic Besano Formation of Monte San Giorgio, Italy/Switzerland: taxonomic and palaeobiological implications. PeerJ 9: e11179. https://doi.org/10.7717/peerj.11179
  • De Blainville HMD (1835) Description de quelques espèces de Reptiles de la Californie, précédée de l’analyse d’un système général d’erpétologie et d’amphibiologie. Nouvelles Annales de Muséum d’Histoire Naturelle 4: 236–296.
  • Campos L, Fernández MS, Herrera Y (2020) A new ichthyosaur from the Late Jurassic of north-west Patagonia (Argentina) and its significance for the evolution of the narial complex of the ophthalmosaurids. Zoological Journal of the Linnean Society 188(1): 180–201. https://doi.org/10.1093/zoolinnean/zlz095
  • Cooper SL, Maxwell EE (2022) Revision of the pachycormid fish Saurostomus esocinus Agassiz from the Early Jurassic (Toarcian) of Europe, with new insight into the origins of suspension‐feeding in Pachycormidae. Papers in Palaeontology 8(6): e1467. https://doi.org/10.1002/spp2.1467
  • Delsett LL, Pyenson N, Miedema F, Hammer Ø (2023) Is the hyoid a constraint on innovation? A study in convergence driving feeding in fish-shaped marine tetrapods. Paleobiology 49(4): 684–699. https://doi.org/10.1017/pab.2023.12
  • Fischer V, Guiomar M, Godefroit P (2011) New data on the palaeobiogeography of Early Jurassic marine reptiles: the Toarcian ichthyosaur fauna of the Vocontian Basin (SE France). Neues Jahrbuch für Geologie und Paläontologie. Abhandlungen 261(1). https://doi.org/10.1127/0077-7749/2011/0155
  • Fischer V, Bardet N, Benson RB, Arkhangelsky MS, Friedman M (2016) Extinction of fish-shaped marine reptiles associated with reduced evolutionary rates and global environmental volatility. Nature Communications 7(1): 10825. https://doi.org/10.1038/ncomms10825
  • Fraas E (1913) Ein unverdrückter Ichthyosaurus-Schädel. Jahreshefte des Vereins für vaterländische Naturkunde in Württemberg 69: 1–12.
  • Godefroit P (1994) Les reptiles marins du Toarcien (Jurassique inférieur) belgo-luxembourgeois. Mémoires pour servir à l’explication des cartes géologiques et minières de la Belgique 39: 1–98.
  • Hänggi H, Reisdorf A (2007) Der Ichthyosaurier vom Hauensteiner Nebelmeer: wie eine Kopflandung die Wissenschaft Kopf stehen lässt. Mitteilungen der Naturforschenden Gesellschaft des Kantons Solothurn 40: 7–22.
  • Hauff B (1921) Untersuchung der Fossilfundstätten von Holzmaden im Posidonienschiefer des Oberen Lias Württembergs. Palaeontographica 64: 1–42.
  • Hauff RB, Heunisch C, Hochsprung U, Ilger JM, Joger U, Klopschar M, Kosma R, Krüger FJ, Thies D, Zellmer H (2017) Jurameer: Niedersachsens versunkene Urwelt. Verlag Dr. Friedrich Pfeil, München, 96 pp.
  • Hofmann J (1958) Einbettung und Zerfall der Ichthyosaurier im Lias von Holzmaden. Meyniana 6: 10–55.
  • Huang JD, Motani R, Jiang DY, Tintori A, Rieppel O, Zhou M, Ren X-X, Zhang R (2019) The new ichthyosauriform Chaohusaurus brevifemoralis (Reptilia, Ichthyosauromorpha) from Majiashan, Chaohu, Anhui Province, China. PeerJ 7: e7561. https://doi.org/10.7717/peerj.7561
  • Johnson R (1977) Size independent criteria for estimating relative age and the relationships among growth parameters in a group of fossil reptiles (Reptilia: Ichthyosauria). Canadian Journal of Earth Sciences 14(8): 1916–1924. https://doi.org/10.1139/e77-162
  • Kirton AM (1983) A review of British Upper Jurassic ichthyosaurs. PhD Thesis. University of Newcastle, Upon-Tyne.
  • Kosma R (2018) The Paleontological Collection of the State Museum of Natural History (SNHM) in Braunschweig, Germany. In: Beck LA, Joger U (Eds) Paleontological Collections of Germany, Austria and Switzerland: The History of Life of Fossil Organisms at Museums and Universities. Springer, 81–92. https://doi.org/10.1007/978-3-319-77401-5_9
  • Klug C, Sivgin T, Miedema F, Scheffold B, Reisdorf AG, Stössel I, Maxwell EE, Scheyer TM (2024) Swiss ichthyosaurs: a review. Swiss Journal of Paleontology 143(31): 1–29. https://doi.org/10.1186/s13358-024-00327-4
  • Laboury A, Bennion RF, Thuy B, Weis R, Fischer V (2022) Anatomy and phylogenetic relationships of Temnodontosaurus zetlandicus (Reptilia: Ichthyosauria). Zoological Journal of the Linnean Society 195(1): 172–194. https://doi.org/10.1093/zoolinnean/zlab118
  • Lamaud P (1979) Les ichthyosaures et la mer toarcienne du Pic Saint–Loup. Minéraux et fossiles. Le guide du collectionneur 58: 42–49.
  • Lomax DR (2016) A new leptonectid ichthyosaur from the Lower Jurassic (Hettangian) of Nottinghamshire, England, UK, and the taxonomic usefulness of the ichthyosaurian coracoid. Journal of Systematic Palaeontology 15(5): 387–401. https://doi.org/10.1080/14772019.2016.1183149
  • Lomax DR, Massare JA (2012) The first reported Leptonectes (Reptilia: Ichthyosauria) with associated embryos, from Somerset, England. Paludicola 8(4): 263–276.
  • Lomax DR, Massare JA (2017) Two new species of Ichthyosaurus from the lowermost Jurassic (Hettangian) of Somerset, England. Papers in Palaeontology 3(1): 1–20. https://doi.org/10.1002/spp2.1065
  • Lomax DR, Massare JA, Mistry RT (2017) The taxonomic utility of fore fin morphology in Lower Jurassic ichthyosaurs: Protoichthyosaurus and Ichthyosaurus. Journal of Vertebrate Paleontology 37(5): e1361433. https://doi.org/10.1080/02724634.2017.1361433
  • Lomax DR, Evans M, Carpenter S (2019) An ichthyosaur from the UK Triassic–Jurassic boundary: A second specimen of the leptonectid ichthyosaur Wahlisaurus massarae Lomax 2016. Geological Journal 54: 83–90. https://doi.org/10.1002/gj.3155
  • Maisch MW (1998a) Kurze Übersicht der Ichthyosaurier des Posidonienschiefers mit Bemerkungen zur Taxionomie der Stenopterygiidae und Temnodontosauridae. Neues Jahrbuch für Geologie und Paläontologie-Abhandlungen: 401–431. https://doi.org/10.1127/njgpa/209/1998/401
  • Maisch MW (1998b) A new ichthyosaur genus from the Posidonia Shale (Lower Toarcian, Jurassic) of Holzmaden, SW-Germany with comments on the phylogeny of post–Triassic ichthyosaurs. Neues Jahrbuch für Geologie und Paläontologie–Abhandlungen: 47–78. https://doi.org/10.1127/njgpa/209/1998/47
  • Maisch MW (2001) Neue Exemplare der seltenen Ichthyosauriergattung Suevoleviathan Maisch 1998 aus dem Unteren Jura von Südwestdeutschland. Geologica et Palaeontologica 35: 145–160.
  • Maisch MW (2008) Revision der Gattung Stenopterygius Jaekel, 1904 emend. von Huene, 1922 (Reptilia: Ichthyosauria) aus dem unteren Jura Westeuropas. Palaeodiversity 1: 227–271.
  • Maisch MW (2020) The best-preserved skeleton of Suevoleviathan integer (Bronn, 1844) (Reptilia: Ichthyosauria) from the lower Jurassic of south-western Germany, with a discussion of the genus. Neues Jahrbuch für Geologie und Paläontologie. Abhandlungen 297: 153–172. https://doi.org/10.1127/njgpa/2020/0918
  • Maisch MW (2022) Ein neuer Eurhinosaurus (Reptilia: Ichthyosauria) aus der Posidonienschiefer–Formation (Unteres Toarcium) von Südwest–Deutschland mit Bemerkungen zur Nomenklatur und Taxonomie der Gattung. Jahreshefte der Gesellschaft für Naturkunde in Württemberg 178: 117–148.
  • Maisch MW, Hungerbühler A (2001) New evidence for a discrete supratemporal bone in the Jurassic Ichthyosaur Temnodontosaurus. Historical Biology: A Journal of Paleobiology 15(4): 335–345. https://doi.org/10.1080/08912960127806
  • Maisch MW, Matzke AT (2000) The Ichthyosauria. Stuttgarter Beiträge zur Naturkunde, Serie B 298: 1–159.
  • Maisch MW, Matzke AT (2022) Magnipterygius huenei n. gen. n. sp., a new small stenopterygiid (Reptilia: Ichthyosauria) from the Posidonienschiefer Formation of SW Germany. Neues Jahrbuch für Geologie und Palaontologie-Abhandlungen 303(2). https://doi.org/10.1127/njgpa/2022/1042
  • Mantell GA (1851) Petrifactions and Their Teachings: Or, A Hand-book to the Gallery of Organic Remains of the British Museum. H. G. Bohn, London, 496 pp. https://doi.org/10.5962/bhl.title.19821
  • Marek RD, Moon BC, Williams M, Benton MJ (2015) The skull and endocranium of a Lower Jurassic ichthyosaur based on digital reconstructions. Palaeontology 58(4): 723–742. https://doi.org/10.1111/pala.12174
  • Martill DM (1993) Soupy substrates: a medium for the exceptional preservation of ichthyosaurs of the Posidonia Shale (Lower Jurassic) of Germany. Kaupia 2: 77–97.
  • Martin JE, Fischer V, Vincent P, Suan G (2012) A longirostrine Temnodontosaurus (Ichthyosauria) with comments on Early Jurassic ichthyosaur niche partitioning and disparity. Palaeontology 55(5): 995–1005. https://doi.org/10.1111/j.1475-4983.2012.01159.x
  • Maxwell EE (2012) New metrics to differentiate species of Stenopterygius (Reptilia: Ichthyosauria) from the Lower Jurassic of southwestern Germany. Journal of Paleontology 86(1): 105–115. https://doi.org/10.1666/11-038.1
  • Maxwell EE (2018) Redescription of the ‘lost’ holotype of Suevoleviathan integer (Bronn, 1844) (Reptilia: Ichthyosauria). Journal of Vertebrate Paleontology 38(2): e1439833. https://doi.org/10.1080/02724634.2018.1439833
  • Maxwell EE, Cortés D (2020) A revision of the Early Jurassic ichthyosaur Hauffiopteryx (Reptilia: Ichthyosauria), and description of a new species from southwestern Germany. Palaeontologia Electronica 23: a31. https://doi.org/10.26879/937
  • Maxwell EE, Vincent P (2016) Effects of the early Toarcian Anoxic Event on ichthyosaur body size and faunal composition in the Southwest German Basin. Paleobiology 42(1): 117–126. https://doi.org/10.1017/pab.2015.34
  • Maxwell EE, Scheyer TM, Fowler DA (2014) An evolutionary and developmental perspective on the loss of regionalization in the limbs of derived ichthyosaurs. Geological Magazine 151(1): 29–40. https://doi.org/10.1017/S0016756812001070
  • Maxwell EE, Cortés D, Patarroyo P, Ruge MLP (2019) A new specimen of Platypterygius sachicarum (Reptilia, Ichthyosauria) from the Early Cretaceous of Colombia and its phylogenetic implications. Journal of Vertebrate Paleontology 39(1): e1577875. https://doi.org/10.1080/02724634.2019.1577875
  • Maxwell EE, Cooper SL, Mujal E, Miedema F, Serafini G, Schweigert G (2022) Evaluating the existence of vertebrate deadfall communities from the Early Jurassic Posidonienschiefer Formation. Geosciences 12(4): 158. https://doi.org/10.3390/geosciences12040158
  • McGowan C (1973) The cranial morphology of the Lower Liassic latipinnate ichthyosaurs of England. Bulletin of the British Museum (Natural History), Geology 24(1): 1–109. https://doi.org/10.5962/p.313822
  • McGowan C (1979) A revision of the Lower Jurassic ichthyosaurs of Germany with the description of two new species. Palaeontographica 166: 93–135.
  • McGowan C (1989a) Computed tomography reveals further details of Excalibosaurus, a putative ancestor for the swordfish-like ichthyosaur Eurhinosaurus. Journal of Vertebrate Paleontology 9(3): 269–281. https://doi.org/10.1080/02724634.1989.10011762
  • McGowan C (1989b) Leptopterygius tenuirostris and other long-snouted ichthyosaurs from the English Lower Lias. Palaeontology 32(2): 409–427.
  • McGowan C (1990) Computed tomography confirms that Eurhinosaurus (Reptilia: Ichthyosauria) does have a tailbend. Canadian Journal of Earth Sciences 27(11): 1541–1545. https://doi.org/10.1139/e90-164
  • McGowan C (1993) A new species of large, long-snouted ichthyosaur from the English lower Lias. Canadian Journal of Earth Sciences 30(6): 1197–1204. https://doi.org/10.1139/e93-101
  • McGowan C (1994) The taxonomic status of the Upper Liassic ichthyosaur Eurhinosaurus longirostris. Palaeontology 37(4): 747–754.
  • McGowan C (1996) The taxonomic status of Leptopterygius Huene, 1922 (Reptilia: Ichthyosauria). Canadian Journal of Earth Sciences 33(3): 439–443. https://doi.org/10.1139/e96-033
  • McGowan C (2003) A new specimen of Excalibosaurus from the English Lower Jurassic. Journal of Vertebrate Paleontology 23(4): 950–956. https://doi.org/10.1671/1860-20
  • McGowan C, Motani R (2003) Part 8 Ichthyopterygia. Sues H–D (Ed.) Handbook of Paleoherpetology. München: Verlag Dr. Friedrich Pfeil: 175.
  • Miedema F, Maxwell EE (2019) Ontogeny of the braincase in Stenopterygius (Reptilia, Ichthyosauria) from the Lower Jurassic of Germany. Journal of Vertebrate Paleontology 39(4): e1675164. https://doi.org/10.1080/02724634.2019.1675164
  • Miedema F, Maxwell EE (2022) Ontogenetic variation in the skull of Stenopterygius quadriscissus with an emphasis on prenatal development. Scientific Reports 12(1): 1707. https://doi.org/10.1038/s41598-022-05540-0
  • Miedema F, Bindellini G, Dal Sasso C, Scheyer TM, Maxwell EE (2023) Ontogenetic variation in the cranium of Mixosaurus cornalianus, with implications for the evolution of ichthyosaurian cranial development. Swiss Journal of Palaeontology 142(1): 27. https://doi.org/10.1186/s13358-023-00289-z
  • Miedema F, Bastiaans D, Scheyer TM, Klug C, Maxwell EE (2024) A large new Middle Jurassic ichthyosaur shows the importance of body size evolution in the origin of the Ophthalmosauria. BMC Ecology and Evolution 24(1): 1–25. https://doi.org/10.1186/s12862-024-02208-3
  • Motani R, Ji C, Tomita T, Kelley N, Maxwell E, Jiang DY, Sander PM (2013) Absence of suction feeding ichthyosaurs and its implications for Triassic mesopelagic paleoecology. PLOS ONE 8(12): e66075. https://doi.org/10.1371/journal.pone.0066075
  • Pardo-Pérez JM, Kear BP, Gómez M, Moroni M, Maxwell EE (2017) Ichthyosaurian palaeopathology: evidence of injury and disease in fossil ‘fish lizards’. Journal of Zoology 304(1): 21–33. https://doi.org/10.1111/jzo.12517
  • Pardo-Pérez JM, Kear BP, Mallison H, Gómez M, Moroni M, Maxwell EE (2018) Pathological survey on Temnodontosaurus from the Early Jurassic of southern Germany. PLOS ONE 13(10): e0204951. https://doi.org/10.1371/journal.pone.0204951
  • Pharisat A, Contini D, Frickert J-C (1993) L’ichthyosaure de la base des schistes–cartons du Toarcien inférieur de Noirefontaine (Doubs). Bulletin de la Société d’Histoire Naturelle du Pays de Montbéliard: 193–198.
  • R Core Team (2024) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna.
  • Rabold JM, Eggmaier S (2013) Grosse Wirbeltierfunde aus der Grube Mistelgau. Der Steinkern 15: 70–76.
  • Reisdorf AG, Maisch MW, Wetzel A (2011) First record of the leptonectid ichthyosaur Eurhinosaurus longirostris from the Early Jurassic of Switzerland and its stratigraphic framework. Swiss Journal of Geosciences 104: 211–224. https://doi.org/10.1007/s00015-011-0069-x
  • Reisdorf AG, Bux R, Wyler D, Benecke M, Klug C, Maisch MW, Fornaro P, Wetzel A (2012) Float, explode or sink: postmortem fate of lung-breathing marine vertebrates. Palaeobiodiversity and palaeoenvironments 92: 67–81. https://doi.org/10.1007/s12549-011-0067-z
  • Riess J (1986) Fortbewegungsweise, Schwimmbiophysik und Phylogenie der Ichthyosaurier. Palaeontographica. Abteilung A, Paläozoologie, Stratigraphie 192(4–6): 93–155.
  • Röhl HJ, Schmid-Röhl A, Oschmann W, Frimmel A, Schwark L (2001) The Posidonia Shale (Lower Toarcian) of SW-Germany: an oxygen-depleted ecosystem controlled by sea level and palaeoclimate. Palaeogeography, Palaeoclimatology, Palaeoecology 165(1–2): 27–52. https://doi.org/10.1016/S0031-0182(00)00152-8
  • Rothschild BM, Xiaoting Z, Martin LD (2012) Adaptations for marine habitat and the effect of Triassic and Jurassic predator pressure on development of decompression syndrome in ichthyosaurs. Naturwissenschaften 99: 443–448. https://doi.org/10.1007/s00114-012-0918-0
  • Ruebsam W, Schmid-Röhl A, Al-Husseini M (2023) Astronomical timescale for the early Toarcian (Early Jurassic) Posidonia Shale and global environmental changes. Palaeogeography, Palaeoclimatology, Palaeoecology 623: 111619. https://doi.org/10.1016/j.palaeo.2023.111619
  • Sachs S, Eggmaier S, Madzia D (2024) Exquisite skeletons of a new transitional plesiosaur fill gap in the evolutionary history of plesiosauroids. Frontiers in Earth Science 12: 1341470. https://doi.org/10.3389/feart.2024.1341470
  • Schulbert C (2001) Die Ammonitenfauna und Stratigraphie der Tongrube Mistelgau bei Bayreuth. Beihefte zu den Berichten der Naturwissenschaftlichen Gesellschaft Bayreuth 4: 1–183.
  • Schulbert C (2013) Die Tongrube Mistelgau. Der Steinkern 15: 4–17.
  • Simonsen S (2013) Fossilien aus dem Posidonienschiefer der Tongrube Mistelgau. Der Steinkern 15: 28–46.
  • Swaby EJ, Lomax DR (2021) A revision of Temnodontosaurus crassimanus (Reptilia: Ichthyosauria) from the Lower Jurassic (Toarcian) of Whitby, Yorkshire, UK. Historical Biology 33(11): 2715–2731. https://doi.org/10.1080/08912963.2020.1826469
  • Urlichs M, Wild R, Ziegler B (1994) Der Posidonien-Schiefer und seine Fossilien. Stuttgarter Beiträge zur Naturkunde C 36: 1–95.
  • Von Huene F (1916) Beiträge zur Kenntnis der Ichthyosaurier im deutschen Muschelkalk. Palaeontographica 62: 1–68.
  • Von Huene F (1922) Die Ichthyosaurier des Lias und ihre Zusammenhänge (Vol. 4). Gebrüder Borntraeger, Berlin, 114 pp.
  • Von Huene F (1926) Neue Ichthyosaurierfunde aus dem Schwäbischen Lias. Neues Jahrbuch für Mineralogie, Geologie und Paläontologie 55; 66–86.
  • Von Huene F (1928) Ein neuer Eurhinosaurus aus dem oberen Lias von Holzmaden. Neues Jahrbuch für Mineralogie, Geologie und Paläontologie: 471–484.
  • Von Huene F (1931a) Neue Studien über Ichthyosaurier aus Holzmaden. Abhandlungen der Senckenbergischen naturforschenden Gesellschaft 42: 345–382.
  • Von Huene F (1931b) Neue Ichthyosaurier aus Württemberg. Neues Jahrbuch für Mineralogie, Geologie und Paläontologie 65: 305–320.
  • Von Huene F (1949) Das Hinterhaupt der Ichthyosaurier. Neues Jahrbuch für Mineralogie, Geologie und Paläontologie, Monatshefte 1949: 215–221.
  • Von Huene F (1951) Ein neuer Fund von Eurhinosaurus longirostris. Neues Jahrbuch für Geologie und Paläontologie. Abhandlungen 93: 277–284.
  • Von Huene F (1952) Kurzer Überblick über die Ichthyosaurier des Schwäbischen Lias. Jahreshefte des Vereins für vaterländische Naturkunde in Württemberg 107: 51–59.
  • Von Jaeger GF (1856) Über eine neue Spezies von Ichthyosauren (Ichthyosaurus longirostris Owen et Jäger) nebst Bemerkungen über die übrigen in der Liasformation Württembergs aufgefundenen Reptilien. Nova acta Academiae caesareae Leopoldino–Carolinae germanicae naturae curiosorum 25: 937–967.
  • Von Theodori C (1854) Beschreibung des kolossalen Ichthyosaurus trigonodon in der Lokal–Petrefakten–Sammlung zu Banz, nebst synoptischer Darstellung der übrigen Ichthyosaurus–Arten in derselben. München (G. Franz): XVII + 81.
  • Yin YL, Ji C, Zhou M (2021) The anatomy of the palate in Early Triassic Chaohusaurus brevifemoralis (Reptilia: Ichthyosauriformes) based on digital reconstruction. PeerJ 9: e11727. https://doi.org/10.7717/peerj.11727

Supplementary materials

Supplementary material 1 

Measurements

Gaël E. Spicher, Feiko Miedema, Jelle Heijne, Nicole Klein

Data type: xlsx

Explanation note: This file provides the morphological measurements of the Mistelgau Eurhinosaurus.

This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Download file (25.41 kb)
Supplementary material 2 

Measurements and statistical data

Gaël E. Spicher, Feiko Miedema, Jelle Heijne, Nicole Klein

Data type: xlsx

Explanation note: This file provides the list of Eurhinosaurus specimens and their body length measurements (table S2) used in the linear regression analysis, as well as the results of the analysis (table S3).

This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Download file (13.25 kb)
Supplementary material 3 

Images

Gaël E. Spicher, Feiko Miedema, Jelle Heijne, Nicole Klein

Data type: pdf

Explanation note: This file provides supplementary figures highlighting some important morphological apsects of the Mistelgau Eurhinosaurus.

This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Download file (6.17 MB)
login to comment