Original Research
Frontiers in Mammal Science ·
Volume 5 - 2026 | https://doi.org/10.3389/fmamm.2026.1869886
Cranial anatomy of the baurioid therocephalian Choerosaurus dejageri with comments on secondary palate convergence in therocephalians and cynodonts
- Alessandro Gigliotti1,2*corresponding author
- Luisa C. Pusch2,3
- Christian F. Kammerer4,5
- Julien Benoit5
- Jörg Fröbisch1,2
- 1Department of Biology, Humboldt-Universität zu Berlin, Berlin, Germany
- 2Museum für Naturkunde, Leibniz-Institut für Evolutions- und Biodiversitätsforschung, Berlin, Germany
- 3Department of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ, United States
- 4North Carolina Museum of Natural Sciences, Raleigh, NC, United States
- 5Evolutionary Studies Institute (ESI), University of the Witwatersrand, Johannesburg, South Africa
Abstract
Permian members of the therocephalian clade Baurioidea remain much more poorly known than the extremely mammal-like Triassic representatives. Here, we describe the cranium of the single known specimen of the lycideopid baurioid Choerosaurus dejageri, based on high-resolution CT-imaging and provide comparisons with closely related taxa. Besides examining external anatomical characters like its characteristic cranial and dentary bosses, we provide important details on the previously obscure palatal morphology. The reconstruction of the holotype shows that this species represents an important transitional stage in the evolution of the therocephalian secondary palate. Complete secondary palates evolved independently three times in therapsids (incl. baurioid therocephalians and cynodonts). In Choerosaurus, the vomer plays the role of an important bracing element in the palate through a long suture with the maxillary shelves, known as the maxillo-vomerine bridge. This differs from the early cynodont condition and highlights different evolutionary origins of therocephalian and cynodont secondary palates. Moreover, similarities in the palatal morphology with Lycideops support the hypothesis of a close relationship between these two taxa.
Keywords
1. Introduction
Baurioidea represents one of the major clades of therocephalian therapsids (Hopson and Barghusen, 1986) and is known from the late Permian through Middle Triassic. Middle Triassic baurioids were extremely mammal-like and have been the topic of recent study (e.g., Abdala et al., 2014), but the Permian members of the group are much more poorly known, and many basic aspects of their anatomy and taxonomy remain obscure. Among Permian baurioids, the family Lycideopidae exhibits the most extreme morphologies, including highly longirostrine taxa that show a mosaic of cynodont-like (e.g., presence of a naso-frontal contact; Sigurdsen et al., 2012) and primitive therocephalian traits (e.g., simple postcanine morphology; Sigurdsen et al., 2012).
Baurioid therocephalians show trends toward closure of the secondary palate and the development of postcanine occlusion convergent on cynodonts (Kemp, 1982; Hopson and Barghusen, 1986; Huttenlocker, 2009; Sigurdsen et al., 2012). Within Synapsida, a closed secondary palate evolved multiple times independently: once within Anomodontia, multiple times within Therocephalia, and again early in cynodont history (Angielczyk, 2009). This structure is typically interpreted as facilitating the separation of the nasopharyngeal portion of the snout from the oral cavity, thereby allowing simultaneous mastication and respiration (Werneburg and Maier, 2019). Baurioids further exhibit many characteristics generally associated with juvenile morphology, such as enlarged orbits positioned dorsally on the skull roof, weakly developed caniniforms, the retention of multiple precanines, and a general trend toward smaller body sizes and more gracile skulls. Moreover, a long and slender mandible is often retained into later ontogenetic stages (Huttenlocker, 2009; Sigurdsen et al., 2012). However, many baurioid taxa may have been diagnosed only by juvenile or sub-adult specimens. This potential sampling bias toward younger individuals raises the question of whether these characteristics are truly paedomorphic. Determining whether baurioid evolution was shaped by neotenic processes will require further ontogenetic studies and increased sampling within this clade (Huttenlocker, 2009; Huttenlocker et al., 2022). The evolutionary parallelism with cynodonts and the potentially paedomorphic skull development, including the feeding apparatus, represent characteristic tendencies in baurioid evolution around the Permo-Triassic boundary. Baurioids account for the vast majority of therocephalian diversity in the Triassic, both in species richness and morphological disparity (including herbivorous and potentially semi-aquatic forms). Non-bauriamorph baurioids neither regained their Permian diversity thresholds nor underwent renewed radiation after the end-Permian extinction. However, bauriamorphs, notably the herbivorous subclade Bauriidae, became diverse, exhibited a trans-Pangaean range, and survived into the Middle Triassic (Grunert et al., 2019; Botha and Smith, 2020; Hancox et al., 2020).
The monophyly of Baurioidea is supported by most analyses focusing on therocephalian phylogeny (e.g., Hopson and Barghusen, 1986; Huttenlocker, 2009; Huttenlocker et al., 2011; Sigurdsen et al., 2012; Huttenlocker and Sidor, 2016; Huttenlocker and Smith, 2017; Liu and Abdala, 2019, 2020, 2022). However, the taxonomy of many baurioids and the interrelationships within this clade are far more disputed and still remain uncertain (e.g., Huttenlocker and Sidor, 2016; Huttenlocker and Smith, 2017; Liu and Abdala, 2019, 2020, 2022). Due to their small size and other diagnostic features that may instead represent juvenile attributes, stemward baurioids were traditionally grouped under the name ‘scaloposaurs’ (Sigurdsen et al., 2012). This concept has since been abandoned, and these taxa have been reassigned to several families now recognized within Baurioidea in the most recent analyses of the group. These include Ictidosuchidae, Regisauridae, Karenitidae, Lycideopidae, Ericiolacertidae, and Bauriidae (Huttenlocker, 2009, 2014). Hopson and Barghusen (1986) reintroduced the family Lycideopidae Boonstra, 1934 into baurioid taxonomy, a placement later supported by Sigurdsen et al. (2012) based on several cranial features, including a long and narrow snout with the orbits positioned more posterior on the skull, a relatively shallow and straight zygomatic arch, a nasal-lacrimal contact, and a maxillary-vomerine contact. Lycideopidae currently includes the four taxa Lycideops longiceps, Choerosaurus dejageri, and the sister taxa Tetracynodon darti and Tetracynodon tenuis (Huttenlocker, 2009, 2014; Sigurdsen et al., 2012). This family is of particular importance because it represents a case example for the evolutionary trajectories of Permo-Triassic non-cynodont therapsids, indicating a trend toward smaller body sizes and more delicate skulls in the Triassic – an instance of the so-called ‘Lilliput Effect’ following the Permo-Triassic mass extinction (PTME). Tetracynodon darti is so far only known from Early Triassic deposits (Sigurdsen et al, 2012; Botha and Smith, 2020), whereas all other lycideopid taxa are of Permian age (Day and Smith, 2020; Smith, 2020; Viglietti, 2020).
The most unusual lycideopid is the enigmatic Choerosaurus dejageri, known from a single specimen (SAM-PK-8797) consisting of a largely complete cranium with occluded partial lower jaw and associated partial postcranium. The specimen was first briefly described by Haughton (1929), who assigned it to the family Scaloposauridae, but it was later revised and reassigned to Lycideopidae by several authors due to its delicate skull structure and derived dentition (Sigurdsen et al., 2012; Sigurdsen et al., 2012; Huttenlocker et al., 2015; Huttenlocker and Sidor, 2016). Uniquely among therocephalians, this taxon exhibits exaggerated bosses on its cranium and dentary, similar to those of some dinocephalian and burnetiamorph therapsids and modern mammals (e.g., suids). A recent CT-assisted study by Benoit et al. (2016) provided morphofunctional information on these bosses, suggesting their use in combat or, more likely, display.
Despite these functional insights, the broader craniomandibular anatomy of this taxon has not been described in detail since Haughton’s (1929) original account. Here, we present a comprehensive redescription of the craniomandibular anatomy of Choerosaurus dejageri, based on a computed tomographic (CT) reconstruction of SAM-PK-8797. This provides new insights into the previously obscured palatal regions and novel information on some previously undescribed features of the endocranial and mandibular anatomy of this taxon. Integrating these new anatomical data will refine our understanding of lycideopid morphology and contribute to broader discussions of therocephalian evolution and therapsid survival across the Permo–Triassic extinction event.
2. Materials and methods
Specimen
SAM-PK-8797 is housed at the Iziko South African Museum (Cape Town, South Africa). It was recovered on Kuilspoort Farm, north of Beaufort West (Central Karoo District, Western Cape Province) (Haughton, 1929). According to Kitching (1977) and the museum database, this locality exposes rocks that encompass almost the whole of the late Permian, as it has delivered faunal elements ranging from the Endothiodon to the Daptocephalus Assemblage Zones (AZs). Haughton (1929), however, states that the specimen came from a “shale bank just above the house”, which belongs to the Hoedemaker member of the Teekloof Formation (Adelaide Subgroup) of the Beaufort Group (SAGS 1:250–000 geological map, Beaufort West 3222). This corresponds to the Wuchiapingian Tropidostoma–Gorgonops Subzone of the Endothiodon AZ (~259 Ma) (Day and Smith, 2020) and is consistent with the age provided by Benoit et al. (2016). The skull of SAM-PK-8797 is slightly sinistrally sheared along its anteroposterior axis and mediolaterally compressed. The right side of the snout is generally well preserved, although several parts are missing, whereas the left side is poorly preserved by comparison. The occipital area is badly damaged and compressed, and the left part of the external braincase is almost completely missing. The right mandible ramus is largely complete compared to the left ramus, which only preserves the anterior portion of the dentary.
SAM-PK-8797 exhibits several paedomorphic features frequently observed in baurioids, including an incomplete postorbital bar, large interpterygoid vacuity, and multiple precanine maxillary teeth, raising the possibility that it represents a juvenile with an unknown adult morphology (as was until recently the case for fellow baurioid Scaloposaurus constrictus; Huttenlocker et al., 2022). However, the presence of prominent inferred display features would be at odds with a juvenile identification (Benoit et al., 2016), as is its relatively large size in comparison to most ‘scaloposaur’ crania (including those of Tetracynodon). Poor preservation of the braincase in SAM-PK-8797 makes it impossible to tell to what extent it was fused in this individual, a frequent indicator of maturity in early synapsids (Araújo et al., 2017), so although we consider this specimen a likely adult, we recognize the need for additional research on this point. A variety of appendicular elements (including a partial femur broken at mid-shaft) are preserved with the postcranium of this specimen, making it a good candidate for osteohistological study.
Computed tomography data
SAM-PK-8797 was CT-scanned at the Stellenbosch CT Scanner Facility of the Stellenbosch University using a Phoenix VTomeX L240 CT system with an isotropic voxel size of 0.050 mm. The visualization of the slices, virtual 3D rendering, and segmentation of selected structures were performed using VGStudio Max 3.4 (Volume Graphics GmbH, Heidelberg, Germany) in the 3D Visualization Laboratory at the Museum für Naturkunde Berlin.
3. Results
General preservation
SAM-PK-8797 is missing large parts of the snout tip, skull roof, subtemporal region, dorsal occiput, and most of the left mandibular ramus. Each premaxilla is anteriorly damaged and lacks the prenasal (intranarial) process, which would have extended dorsally to contact the nasals as is present in other synapsids. The left premaxilla is generally better preserved than the right, which is missing most of its anterolateral portion housing upper incisors 1-5. Except for a small anterior fragment on the left side, both septomaxillae are not preserved, the maxillae are moderately well preserved, although their full dorsal and posterior extents—including contacts with the lacrimal, prefrontal, nasal, and jugal—cannot be reconstructed due to missing elements on both sides. The nasals are badly damaged, with only parts of the posterior region and fragments of the lateral portion on both sides preserved. The left zygomatic arch, the anterodorsal margins of the lacrimals and anteroventral margins of the right prefrontal, and most of the left prefrontal are not preserved. Also, the posterior portion of the frontals and nearly the entire parietal bones are missing, leaving little information on their mutual suture. The postorbitals are preserved on both sides but are disarticulated from the frontal. The palatal region of SAM-PK-8797 is in comparably better condition. The left side of the maxillo-vomerine bridge is slightly damaged, and this anterior portion of the vomer is obliquely cracked. The palatines are well preserved except for the posterolateral portion on the left side. Only a fragment of the left ectopterygoid is retained, and the quadrate process of the pterygoid is only visible on the right side. The occipital and temporal regions are severely damaged and compressed. Because the density contrast between bone and matrix is minimal, much of this region could not be visualized. The prootics can be visualized superficially, along with the exoccipitals, the orbitosphenoid, and the anterior portion of the fused parabasisphenoid, including the cultriform process of the parasphenoid. No identifiable traces of the left squamosal, left quadrate, quadratojugals, basioccipital, opisthotics, tabulars, supraoccipital, and postparietal are present. In the lower jaw, the right dentary is generally well-preserved, whereas the left dentary is only preserved up to the level of the tenth postcanine. Both dentaries are broken at the level of the symphysis, with the left dentary having been rotated and shifted approximately 45° relative to the right. Among the postdentary bones, only the right splenial, angular, surangular, prearticular, and both coronoids are preserved, with the left coronoid having been displaced dorsally into the region of the nasal cavity. Thus, descriptions of the postdentary bones are primarily based on the right side.
Snout
Premaxilla
The premaxilla forms the anterior limit of the snout and the anteromedial margin of the external naris (Figures 1, 2). A section of the premaxilla-septomaxilla contact is visible on the left side of the skull, where a small fraction of the septomaxilla is preserved (Figures 2, 3). Internally, each premaxilla has a palatally extending vomerine process, which accommodates the anterior process of the vomer medially, almost up to the anterior limits of the paracanine fossa (Figures 4, 5). The vomerine process of the premaxilla is anteroposteriorly elongated and forms much of the medial wall of the anterior portion of the internal naris. Its posterior half, which overlies the anterior process of the vomer on the palatal surface, is slightly damaged, exposing portions of the vomer that would otherwise have been covered by the vomerine process in intact condition of the latter. In dorsal view, the vomerine process of the premaxilla features a distinct anteroposteriorly oriented ridge running medial to the lower canine fossa (Figure 5). Comparable to some other baurioids such as Regisaurus jacobi (Mendrez, 1972) and the lycideopid Tetracynodon darti (Sigurdsen et al., 2012), the vomerine process of the premaxilla in Choerosaurus extends between the anteromedial portion of a stout medial process of the maxilla located medial to the upper canine alveolus, forming a maxillo-vomerine bridge (Figure 5). This differs from the condition observed in the bauriamorph Notictoides absens (Sidor et al., 2022) and whaitsiid eutherocephalians such as Theriognathus microps (Huttenlocker and Abdala, 2015) and Moschowhaitsia vjuschkovi (Tatarinov, 1965; Liu and Abdala, 2023), in which the vomerine process of the premaxilla is anteroposteriorly shorter and does not reach the maxillo-vomerine bridge. The facial portion of the premaxilla is posteriorly overlain by the maxillae.
Septomaxilla
The septomaxilla is an elongated, thin element along the anterolateral edge of the snout. On the left side, only a small anterior fragment of the septomaxilla is preserved, whereas the right one is completely missing. The preserved fragment, which may represent an intranarial process, is located on the anteriormost tip of the septomaxilla and faces medially (Figures 1–3). In this region, the septomaxilla contacts the premaxilla and maxilla medially.
Maxilla
The maxilla is a large bone that forms most of the lateral margin of the snout and side wall of the nasal cavity. It bears enigmatic bony outgrows consisting of conspicuous hemispheric bosses, one on either side positioned rostrally (Figures 1, 2). These structures have recently been described in detail by Benoit et al. (2016), who investigated their potential function and concluded that both display behavior and low energy combat are plausible, highlighting the possible role of such structures in complex behaviors such as sexual selection and sexual dimorphism in premammalian evolution. The bosses extend over approximately one third of the length and nearly the full height of the maxilla. Each boss extends from the level of the last upper incisor to the level of the first upper postcanine tooth. According to measurements by Benoit et al. (2016), the right boss measures 14 mm in length and 11 mm in height, whereas the left measures 12 mm in length and 9 mm in height. The bosses consist of two types of spongy bone, both distinct from the compact bone of the surrounding maxilla. The core is more cancellous and hollowed, whereas the denser outer layer is traversed by radially oriented canals, giving the surface of the boss a rough texture. The right boss is completely preserved, whereas the left boss is missing parts of its posterodorsal margins and has been reattached to the maxilla. The surfaces of both bosses are perforated by numerous larger foramina, likely associated with branches of the infraorbital nerve. Benoit et al. (2016) argued that the high perforation by the maxillary canal, the gracile and lightly built skull of Choerosaurus, and the fact that the canal opens directly onto the surface of the bosses make their use in high energy combat unlikely. By contrast, the denser and less vascularized anatomy of the mandibular boss suggests that lateral flank butting may have been possible, as this mode of combat reduces the risk of head injuries. The high innervation in the bosses also supports a role in display behavior involving tactile recognition. Moreover, these structures may have been covered by a cornified sheath and used for intimidation or ritualized sexual behavior (Benoit et al., 2016; Bolton et al., 2025).
The anterolateral margin of the maxilla would have been overlain by the septomaxilla, but this contact is not preserved. The ventral margin of the maxilla extends in a straight line and forms a long posterior process, which is poorly preserved on both sides (Figures 1, 2), and would have contacted the anterior process of the jugal in intact condition. Posterior to the maxillary boss, the facial surface of the maxilla rises dorsally to meet the nasal (Figures 1, 2), but due to poor preservation it cannot be determined whether the maxillae constricted the lateral extent of the nasals in external view, nor can the shape of the dorsal maxillary extension be reconstructed. The posterodorsal margin of the maxilla contacts the anterior extension of the lacrimal, but no contact with the prefrontal is preserved on either side of the skull.
Internally, the maxilla projects medially and shares a long suture anteriorly with the vomer, together forming a maxillo-vomerine bridge, which extends anteroposteriorly from the level of the upper canine to the level of the third upper postcanine (Figures 4, 5). Anterior to its contact with the vomer, the maxilla also meets the palatal process of the premaxilla, together forming the paracanine fossa (Figure 5). Posterior to the large canine boss and the maxillo-vomerine bridge lies a depression, the posterior maxillary fossa (van den Heever, 1994), also referred to as the anterior maxillary sinus (after Sigurdsen, 2006, Figures 4A, C). No anterior maxillary fossa is present posterior to the premaxilla, as in Tetracynodon (Sigurdsen et al., 2012). Immediately posterior to this depression, the maxilla contacts the anterolateral process of the palatine ventrally, which overlaps the medial surface of the maxilla to approximately the midpoint of the choana (Figure 5). Dorsally, the internal surface of the maxilla is slightly overlapped by the nasal and, on the left side, also by the lacrimal. However, in the intact condition of the maxilla, this contact would presumably be more extensive. In dorsal view, the maxilla shows no marked curvature and extends posteriorly in a relatively straight course.
Nasal
The paired nasals are elongated elements that form the roof of the snout and the nasal cavity. Owing to the poor state of preservation, little can be said about their anterior proportions or the degree of lateral constriction. Posteriorly, the nasals contact the frontals in a V-shaped suture (Figure 3). A posterolateral fragment, isolated from the main preserved portion of the nasal, contacts the anteriormost tip of the lacrimal, as well as the dorsal margin of the maxilla on either side of the skull (Figure 3). The better-preserved left nasal shows evidence of a nasal-lacrimal suture – a characteristic feature of Lycideopidae representing convergence with cynodonts (Sigurdsen et al., 2012; Huttenlocker, 2013). No contact with the prefrontal is preserved on either side of the skull due to bone loss around the anterior orbit. Internally, the midline suture of the nasals is slightly convex and is bordered posteriorly by two anterior extensions of the frontal. Parallel to the midline suture, ridges occur on the posterolateral margins of the ventral surface of the nasals, which continue posteriorly onto the frontals and flare laterally at the level of the anterior orbital margin (Figure 6). Such ridges have frequently been interpreted in therapsids as attachment sites for cartilaginous nasoturbinals associated with olfactory function (e.g., Kemp, 1979; Hillenius, 1991, 1994; Ruben, 1996; Hillenius and Ruben, 2004; Sigurdsen, 2006; Ruben et al., 2012; Crompton et al., 2015, 2017; Bendel et al., 2018; Pusch et al., 2019, 2020, 2021; Huttenlocker and Sidor, 2020; Gigliotti et al., 2023).
Lacrimal
The lacrimal forms the anteroventral margin of the orbit. Posteroventrally, it contacts the anterior process of the jugal bone, although this contact is only preserved on the right side in SAM-PK-8797. Its posterodorsal portion shares a straight suture with the prefrontal, likewise, only preserved on the right side (Figures 1, 2). Anterior to this point, the remaining dorsal margin of the right lacrimal is missing, whereas on the left side the entire dorsal margin is absent. Anteriorly, the lacrimals contact the dorsal margin of the maxilla (laterally) and part of the posterolateral margin of the nasal (medially). Internally, an opening is present at the base of the anterior portion of the lacrimal, which extends posteriorly along the entire length of the bone and exits into the orbit (Figures 4, 7). This opening corresponds to the naso-lacrimal duct, which is accompanied medially by an anteroposterior extending ridge on the internal surface of the lacrimal. Internally, the lacrimal forms a short anterior process that contacts the posteroventral margin of the nasal, but due to damage, it cannot be determined whether this internal nasal–lacrimal contact was restricted to the anterior process alone or whether a broader contact was present, as observed in some other eutherocephalians such as the akidnognathid Olivierosuchus (Gigliotti et al., 2023) and the whaitsiid Theriognathus (Pusch et al., 2024), and in cynodonts for which cranial CT-reconstructions are available (Pusch et al., 2024).
Prefrontal
The prefrontal contributes to the anterodorsal margin of the orbit. It borders the anterolateral part of the frontal, although some medial parts of that margin are missing on both sides. On the left side, only the posterodorsal part of the bone is preserved, and is broken into two separate anterior and posterior pieces (Figures 2, 3, 6). The anterodorsal extension of the bone that would have curved slightly inwards to contact the posterodorsal part of the nasal is also missing. On the right side, the prefrontal contacts the lacrimal posteroventrally in a straight suture to build the anterior margin of the orbit (Figures 1, 4). In intact condition, the ventral contact with the lacrimal would proceed anteriorly to overlap the bone until a possible contact with maxilla and nasal. The internal expansion of the prefrontals cannot be observed due to the severe crushing, but it is possibly of a lesser extent and without a descending flange, as is also observed in other therocephalians for which cranial CT reconstructions are available, including Alopecognathus, Olivierosuchus, and Theriognathus (Pusch et al., 2024). This condition contrasts with that of many cynodonts, in which the prefrontal has a distinct flange descending along the posteromedial margin of the lacrimal (Pusch et al., 2024).
Jugal
The jugal is preserved only on the right side of the skull. It is a slender bone forming the posterolateral floor of the orbit and the ventral portion of the zygomatic arch. Its anterior process, which in intact condition would have been overlain by the posterolateral process of the maxilla, overlaps the posteroventral part of the lacrimal (Figure 1). In internal view, this process meets the lateral process of the ectopterygoid, which also approaches the posteromedial margin of the lacrimal (Figures 4, 5, 7). Posteriorly, the jugal extends into the temporal region as a posterior process, or subtemporal ramus, which flares laterally to form an expanded lamina that defines the ventral margin of the subtemporal fenestra. The posteriormost tip of this ramus contacts the anterior extension of the squamosal. A dorsal process appears to be absent, and thus there is no indication of an ossified postorbital bar, although the postorbital bones themselves are not preserved in articulation. An unossified postorbital bar is also present in the baurioids Ericiolacerta, Bauria, Tetracynodon, Ictidosuchops, and the juvenile regisaurid described by Kemp (1986), but contrasts with the well-ossified bar observed in Silphoictidoides and Microgomphodon (Kemp, 1986; Huttenlocker, 2009; Sigurdsen et al., 2012; Maisch, 2017).
Skull roof
Frontal
The anteroposteriorly elongated frontals form the anterodorsal portion of the skull roof. Anteriorly, they are overlapped by the V-shaped posterior margin of the nasals (Figure 3). The posterior half of each frontal is missing from the point at which it would have contacted the postorbitals laterally. Anterolaterally, the frontal contacts the medial margin of the prefrontal. Although the postorbitals are not in articulation, it is clear that they did not meet the prefrontals to exclude the frontal from the orbit; instead, the frontal contributes to the dorsomedial margin of the orbit. The ventral surface of the frontals bears two prominent lateral crests that accommodate the posterior margins of the prefrontals and flare laterally. Between these crests is a prominent depression that housed the olfactory bulbs of the brain, with the crests defining their lateral and (partially) anterior limits (Figure 6). A comparable morphology has also been described for other therapsids such as the gorgonopsian Cynariops (Bendel et al., 2018) and the cynodonts Procynosuchus (Kemp, 1979), Galesaurus (Pusch et al., 2019), Bolotridon (Pusch et al., 2021), and Chiniquodon (Kemp, 2009). Anteriorly, additional lateral ridges extend forward and appear to continue onto the nasal, potentially representing attachment sites for cartilaginous nasoturbinals as described above.
Postorbital
The postorbital contributes to the dorsal margin of the orbit and the postorbital bar, where it typically articulates with the posterolateral portion of the frontals and the anterolateral portion of the parietals. In SAM-PK-8797, the postorbitals are displaced from their in-situ position, where they seem to “float” within the braincase. Because the parietals are almost completely missing and the posterior portion of the frontals are not preserved, the precise position of the postorbital relative to the frontal–parietal suture cannot be determined with confidence. The postorbital has a narrowing descending process that terminates in a pointed tip and would likely have contributed to an open postorbital bar. This process originates from an anteroposteriorly oriented crest on the dorsal surface of the bone, which probably continued onto the parietal to accommodate the pineal opening.
Palate
Vomer
The vomer is an elongated, medially positioned element that extends from the premaxilla anteriorly to the pterygoid posteriorly and divides the nasopharyngeal passage along the midline. Its anterior process tapers into a sharply pointed tip that fits between the palatal plates of the premaxilla. Anterolaterally, the vomer bears paired transverse processes that contact the maxilla (damaged on the left side), forming the maxillo-vomerine bridge associated with a consolidation of the anterior portion of the soft palate (Figure 5). This structure has also been described in whaitsiids (e.g., Tatarinov, 1965; Huttenlocker and Sidor, 2016; Liu and Abdala, 2023) and other baurioid eutherocephalians (Mendrez, 1972; Ivakhnenko, 2011; Sigurdsen et al., 2012; Huttenlocker and Sidor, 2016; Sidor et al., 2022). The maxillo-vomerine bridge divides the internal naris into anterior (lower canine fossa) and posterior (choanal) compartments and is continued posteriorly by a sharp crista choanalis that extends onto the palatine and terminates just anterior to the suborbital vacuity. Unlike whaitsiids, in which the maxillo-vomerine bridge forms only a short, finger-like process at the level of the upper canine (Tatarinov, 1965; Huttenlocker and Sidor, 2016; Liu and Abdala, 2023), the maxillo-vomerine bridge in SAM-PK-8797 is an anteroposteriorly more expansive structure, spanning from the level of the upper canine to around the level of PC3. This condition resembles that of other baurioids including the regisaurid Regisaurus jacobi (Mendrez, 1972), the lycideopid Tetracynodon darti (Sigurdsen et al., 2012), and the bauriamorph Notictoides absens (Sidor et al., 2022). The anterior portion of the vomer has the typical isosceles-triangle shape seen in many basal therocephalians (Mendrez, 1974; Ivakhnenko, 2011; Liu and Abdala, 2017, 2019, Gigliotti et al., 2023) and is obliquely broken at the level of the maxilla-vomerine bridge, where the bone also reaches its greatest thickness. Posterior to the maxillo-vomerine bridge, the interchoanal process of the vomer narrows, reaching its minimum width at the posterior margin of the choanae, before flaring laterally again where it contacts the palatines. The posterior portion of the vomer extends between the palatines and tapers towards its contact with the pterygoid. A distinct median crest extends from the interchoanal process onto the posterior portion of the vomer, as is also present in other baurioids (e.g., Mendrez, 1972; Sigurdsen et al., 2012; Huttenlocker and Sidor, 2016) and other non-baurioid eutherocephalians (e.g., Mendrez, 1974; Durand, 1991; Huttenlocker and Sidor, 2016; Liu and Abdala, 2017; Gigliotti et al., 2023). On the dorsal surface, the vomer bears a median ridge that runs anteroposteriorly along its entire length. This ridge has a shallow median groove that may have supported the ventral margin of a cartilaginous nasal septum, as has also been described for other therocephalians and cynodonts (Fourie, 1974; van den Heever, 1994; Crompton et al., 2015, 2017; Pusch et al., 2019, 2020, 2024; Gigliotti et al., 2023). In dorsal view, at the level where the vomer contributes to the posterior margin of the choana, the bone typically expands laterally to overlap the medial margins of the palatine. This expansion is moderate in other therocephalians such as the akidnognathid Olivierosuchus (Gigliotti et al., 2023), the lycideopid Tetracynodon (Sigurdsen et al., 2012) and the whaitsiid Theriognathus (Huttenlocker and Abdala, 2015). In contrast, the anterior process of the vomer in the akidnognathid Cradognathus albanensis appears much wider and its expansion at the suture with the palatines is less prominent, as the width remains relatively constant anteroposteriorly (Lloyd and Durand, 2025). In Choerosaurus, the lateral expansion is even more subdued, widening only at the posterior margin of the choana and constricting again posteriorly to nearly the width of the interchoanal process (Figure 5). In contrast, many cynodonts and the basal therocephalian Lycosuchus show an extreme lateral expansion of the vomer in this region, forming lateral “wings” that create a transverse lamina together with the transverse processes of the palatines (Crompton et al., 2015, 2017; Pusch et al., 2019, 2020). Ventrally, the vomer is keeled and has a ventromedian crest, a typical eutherocephalian feature (Gigliotti et al., 2023). This crest appears to originate at the level of the maxillo-vomerine bridge and continues posteriorly along the entire length of the bone via the interchoanal process.
Palatine
The paired palatines make up the majority of the palatal plate. Medially, they share a long anteroposterior extending suture with the vomer, and their lateral expansion is constrained by the maxilla. The anterior margin of the palatine forms the posterior border of the choana. In dorsal view, the palatine bears three lateral processes (Figure 5). The first is the anterolateral process, which extends dorsally to meet the internal surface of the maxilla. The second is the medial process, which rises almost vertically to contact the lacrimo-palatine ridge of the lacrimal and typically also contacts the maxilla, although this latter contact is not preserved in the present specimen. This process develops from a nearly vertically oriented ridge that curves posteriorly to run parallel to the palatine-vomer suture. Medially, this ridge forms a distinct lamina, which is more clearly visible on the left side. A comparable lamina has also been described in the akidnognathid Olivierosuchus parringtoni and the whaitsiid Theriognathus microps, although it is less pronounced than in the former and more closely resembles the condition in the latter (Gigliotti et al., 2023; Pusch et al., 2024). As in these taxa, the lamina shows some resemblance to the pronounced transverse laminae of cynodonts, which are formed by the medially directed plates of the palatines and the lateral “wings” of the vomer (Crompton et al., 2015, 2017; Pusch et al., 2019, 2021). However, unlike cynodonts, none of the therocephalian taxa mentioned above exhibit corresponding lateral expansions of the posterior vomer. The third process is the posterolateral process of the palatine, which contacts the anterior process of the ectopterygoid posteriorly and contributes to the anterolateral border of the suborbital vacuity (Figure 4C). Posteromedially, the palatines overlap the anterior process of the pterygoid laterally in dorsal view, and this posterior extension forms the anteromedial margins of the suborbital vacuity. In ventral view, the palatines exhibit a prominent medial depression representing the posterior portion of the crista choanalis. This depression is laterally bound by a sharp anteroposterior crest that accommodates the internal surface of the maxilla on both sides.
Ectopterygoid
The ectopterygoid is a tripartite element that forms the lateral margin of the suborbital vacuity and consists of anterior, posterior, and dorsolateral processes (Figure 4). The anterior and posterior processes are preserved only on the right side, whereas of the left ectopterygoid only the dorsal process is preserved. Its elongate shape and vertical orientation correspond to the condition described in other therocephalians such as Lycosuchus, Alopecognathus, Olivierosuchus, and Theriognathus, but differs from the diminutive state seen in early cynodonts (Huttenlocker and Abdala, 2015; Gigliotti et al., 2023; Pusch et al., 2024). The anterior process of the ectopterygoid contacts the posterolateral process of the palatine, and the posterior process extends to meet the anterolateral portion of the transverse flange of the pterygoid. The distinct dorsolateral process of the ectopterygoid articulates with the anterior process of the jugal and the posteroventral margin of the lacrimal, as in the many other eutherocephalians including Ictidosuchoides, Olivierosuchus, Microwhaitsia, Theriognathus, Tetracynodon, Notictoides, Mupashi, and Bauria (Brink, 1963; Sigurdsen et al., 2012; Huttenlocker and Abdala, 2015; Huttenlocker and Sidor, 2016; Huttenlocker and Smith, 2017; Sidor et al., 2022; Gigliotti et al., 2023; Pusch et al., 2024). In this region, the dorsolateral process becomes coalescent with the lacrimo-ectopterygoidal ridge of the lacrimal.
Pterygoid
The pterygoid is a large plate-like element that forms the posterior margin of the palate. It consists of a tongue-shaped anterior process, the lateral transverse flanges, a posterior portion, and a long quadrate process that is only preserved on the right side. The tongue-shaped anterior process of the pterygoid is relatively long and narrow (Figure 5), more closely resembling the condition present in Tetracynodon than in Lycideops, which possesses a shorter and stouter anterior process of the pterygoid (see Sigurdsen et al., 2012). A comparable condition of a long and narrow anterior process of the pterygoid as is present in Choerosaurus and Tetracynodon has also been described for other non-lycideopid baurioids, including the regisaurids Regisaurus and Urumchia, the bauriamorph Ordosiodon (Sun, 1991), and the karenitid Mupashi (Huttenlocker and Sidor, 2016). By contrast, in the bauriamorph Hazhenia, the process is unusually long and extremely broad (Sun, 1991), whereas a very short and broad anterior process of the pterygoid is observed in the basal baurioid Ictidosuchops (Huttenlocker, 2013) and the bauriids Bauria (Brink, 1963) and Microgomphodon (Abdala et al., 2014). In ventral view, the tongue-shaped anterior process of the pterygoid is laterally overlapped by the palatines and tapers anteriorly where it would have been bifurcated by the posterior process of the vomer, although the exact nature of the contacts among these three bones is difficult to determine due to poor preservation. The ventromedian crest of the vomer does not continue onto the anterior process of the pterygoid. Instead, this region of the pterygoid exhibits an oval depression that develops a subtle crest along its posteromedian margin, just anterior to the oval interpterygoid opening. Unlike the typical condition described for therocephalians (Huttenlocker, 2009), this crest does not extend into a ventromedian process that would give the interpterygoid vacuity a heart-shaped outline. Posteriorly, the interpterygoid vacuity is bisected by a short cultriform process of the parasphenoid, which extends anteriorly to approximately the midpoint of the vacuity (Figure 5). This region resembles that of the basal baurioid Silphoictidoides in its tongue-shaped outline and oval depression. Anterior to the vacuity, the pterygoid in therocephalians commonly features a distinct median tubercle or eminence on its ventral surface (e.g., Huttenlocker, 2009; Maisch, 2017; Gigliotti et al., 2023). However, this tubercle is only faintly evident in Choerosaurus (Figure 5B). In dorsal view, the anterior process of the pterygoid does not show a continuation of the dorsal vomerine crest, unlike the condition in Olivierosuchus (Gigliotti et al., 2023). Only the right transverse flange of the pterygoid is preserved. It forms the posteromedian margin of the suborbital vacuity together with the anterior margin of the pterygoid and meets the posterior process of the ectopterygoid. The quadrate process supports the epipterygoid, which rests on it with an anteroposteriorly elongate ventral footplate (Figure 7). This process broadens dorsoventrally toward the posterior end, giving it a bladelike appearance. It would have contacted the anterior margins of the quadrate, but due to postmortem displacement it now lies beneath the ventral surface of the trochlea. A lacrimal-pterygoid contact is absent in Choerosaurus; a contact between these bones is commonly observed in most cynodonts for which cranial CT reconstructions are available (Pusch et al., 2024). As in the lycideopid Tertracynodon and various other non-lycideopid eutherocephalians, pterygoid teeth are absent in Choerosaurus, a condition that differs from Lycideops in which a reduced number of pterygoid boss teeth are present (Huttenlocker and Sidor, 2016).
Braincase and occiput
Orbitosphenoid
The sphenethmoidal complex is preserved but not in its in-situ position; instead, it appears to “float” within the braincase between the pterygoid ventrally and the frontal dorsally. This complex consists of the paired, wing-shaped orbitosphenoids, which fuse ventrally to form the ventromedian interorbital septum (Figure 8), as has also been described for many non-cynodont therapsids, including gorgonopsians (Benoit et al., 2017) and the eutherocephalians Olivierosuchus (Gigliotti et al., 2023), Tetracynodon (Sigurdsen et al., 2012), Microgomphodon (Abdala et al., 2014), and Mupashi (Huttenlocker and Sidor, 2016). In its in-situ position, the complex would have contacted the frontals loosely, forming part of the lateral walls of the anterior braincase and accommodating the olfactory bulbs in non-mammalian therapsids (e.g., Benoit et al., 2017; Bendel et al., 2018; Crompton et al., 2018). In SAM-PK-8797, the entire complex is shifted to the right side of the braincase, rotated posteriorly by approximately 180° and tilted dorsoventrally by nearly 90°. Each orbitosphenoid consists of a short plate that is fused ventrally into a keel. Unlike the long, hemicylindrical orbitosphenoids of some gorgonopsians, those of Choerosaurus are shorter, thinner, and more U-shaped in cross-section, resembling the condition present in the akidnognathid Olivierosuchus (Gigliotti et al., 2023), the baurioids Tetracynodon (Sigurdsen et al., 2012), Microgomphodon (Abdala et al., 2014), and Mupashi (Huttenlocker and Sidor, 2016), and cynodonts (e.g., Benoit et al., 2017; Pusch et al., 2024). The orbitosphenoids are indistinguishably fused ventrally to the interorbital septum and rostrally to the mesethmoid.
Epipterygoid
The epipterygoid is a vertically oriented, flat, and elongate element situated dorsal to the quadrate process of the pterygoid, forming part of the lateral wall of the braincase (Figures 2, 7). Its general morphology closely resembles that described for other therocephalians, such as the akidnognathids Olivierosuchus (Gigliotti et al., 2023), Cradognathus (Lloyd and Durand, 2025) and Promoschorhynchus (Mendrez, 1974; Huttenlocker et al., 2011), the karenitid Mupashi (Huttenlocker and Sidor, 2016), and the baurioids Regisaurus (Mendrez, 1972) and Silphoictidoides (Maisch, 2017). In these taxa, the epipterygoid exhibits a considerably narrower exposure than is observed in the whaitsiid Theriognathus (Huttenlocker and Abdala, 2015) and cynodonts (e.g., Fourie, 1974; Kemp, 1979; Abdala et al., 2019; Pusch et al., 2019, 2023, 2024; Huttenlocker and Sidor, 2020), in which the epipterygoid is more extensively expanded anteroposteriorly. However, its anteroposterior extent is noticeably wider than in the lycideopid Tetracynodon, in which the epipterygoid more closely resembles a thin rod (Sigurdsen et al., 2012). The ascending process of the epipterygoid extends dorsally and expands to presumably contact the parietal, although the parietal is preserved only as a small fragment, preventing a confident reconstruction of the exact dimensions of this contact. The anterior margin of the epipterygoid is dorsoventrally concave and contributes to the posterior rim of the orbital vacuity (Figure 7), which transmitted the ophthalmic branch (V1) of the trigeminal nerve—a common condition in therocephalians and cynodonts (Mendrez, 1972; Kemp, 1979; Ivakhnenko, 2011; Sigurdsen et al., 2012; Huttenlocker and Sidor, 2016; Maisch, 2017; Huttenlocker and Smith, 2017; Pusch et al., 2023, 2024). The posterior margin of the epipterygoid forms the anterior border of the trigeminal foramen associated with the trigeminal nerve (V). Ventrally, the footplate of the epipterygoid rests on an anteroposteriorly oriented ridge along the dorsal surface of the quadrate process of the pterygoid, which is only preserved on the right side. No contact between the footplate and the prootic or squamosal is visible and appears rather unlikely in intact condition. Furthermore, there is no suture between the epipterygoid and the anterodorsal process of the prootic. A posterior apophysis is absent, and there is no evidence for an epipterygoid foramen.
Prootic
The prootic lies posterior to the epipterygoid and forms part of the lateral braincase wall. It is positioned anterior to the otic region and contributes to the anterior margin of the bony labyrinth. In SAM-PK-8797, both prootics are considerably displaced toward the right side of the braincase, with the left prootic being better preserved than the right (Figures 2, 7). Only the left prootic preserves the pila antotica, an anterodorsally directed process, which lies just medial to the epipterygoid in its in-situ position. The base of this process is broad and thickens posteriorly. Under normal circumstances, it would overlap the posterodorsal margin of the basisphenoid to accommodate the sella turcica, but this region is not preserved in SAM-PK-8797. The prootic further has an anterodorsal process that extends medially. Together with the pila antotica, it forms the incisura prootica, which constitutes the posterior margin of the trigeminal nerve (V) opening. Posteroventral to the incisura, a small foramen pierces the lateral lamina of the left prootic and probably transmitted the facial nerve (VII). In intact condition, the prootic likely would have contacted the supraoccipital, tabular, and squamosal posterodorsally, the basioccipital ventrally, and the opisthotic posteriorly as in other therapsids.
Parabasisphenoid
The fused parabasisphenoid complex forms the anterior floor of the braincase. In SAM-PK-8797, only the anterior portion of this complex is preserved including the short cultriform process of the parasphenoid, which projects anteriorly between the pterygoids and bisects the interpterygoid vacuity (Figures 5, 7). Poor preservation of the bone prevents further description.
Exoccipital
The exoccipital is a paired bone that forms much of the lateral margins of the foramen magnum and the occipital condyle. The bone is extremely dislocated on both sides, and the right one is better preserved than the left (Figure 7). Its ventral process is a thin, sheet-like structure that is pierced by a foramen at its base. The dorsal process has a clubbed medial extension and a dorsal extension, which probably would have contacted neighboring bones in the occiput, like the opisthotic and the supraoccipital respectively.
Lower jaw
Dentary
Only the right dentary provides sufficient information about its morphology to describe. It is a horizontally orientated, elongated and slender bone, lying lateral to the postdentary bones. Its posterior horizontal ramus is relatively vertically oriented and gently curved (Figures 9A, B). It lacks a pronounced, tall coronoid process that tapers dorsally, unlike other therocephalians such as Glanosuchus, Microgomphodon, Moschorhinus, Mirotenthes, Theriognathus, Ictidosuchoides, Bauria, Perplexisaurus, Microwhaitsia, Ophidostoma, Silphoictidoides, Lycosuchus, and Olivierosuchus (Abdala et al., 2008, 2014; Huttenlocker, 2009; Ivakhnenko, 2011; Huttenlocker and Abdala, 2015; Huttenlocker and Smith, 2017; Maisch, 2017; Pusch et al., 2020; Gigliotti et al., 2023). Instead, the coronoid process of SAM-PK-8797 is low and rounded, rising only slightly above the level of the surangular, similar to the condition present in other lycideopids such as Tetracynodon darti (Sigurdsen et al., 2012) and other baurioids such as the undetermined baurioid AMNH FARB 9542, Mupashi, and Notictoides (Huttenlocker and Sidor, 2012, 2016; Sidor et al., 2022). Huttenlocker (2013) considered the coronoid extending just above the surangular to be a diagnostic feature for lycideopids. Approximately at the posterior third of the bone, the dentary bears a rounded bony outgrowth on the posteroventral angle of the right dentary (Figures 9A, C), comparable to the bosses on the maxillae. A corresponding structure was likely also present on the left dentary, but it cannot be described because of the poor preservation of the left mandibular ramus. However, the mandibular boss is more oval in outline than the hemispheric maxillary bosses and is elongated posterodorsally to anteroventrally. It consists of two different layers of bone tissue–a dense, compact inner layer and a more vascularized outer layer that produces a pitted external surface—similar to the maxillary bosses, though less pronounced (Benoit et al., 2016). This two-layered structure is continuous throughout the entire dentary. Anterior to the boss, the ventral margin of the dentary thickens before gradually diminishing toward the tooth bearing ramus. Although both dentaries are broken just anterior to the symphysis, there is no indication of a steep dorsal tapering of the anterior tooth-bearing portion to form a distinct chin, unlike the condition in Promoschorhynchus and Olivierosuchus (Mendrez, 1974; Gigliotti et al., 2023). The lateral surface of the dentary is smooth and lacks anterior foramina, in contrast to many other therocephalians. Medially, the dentary bears an anteroposteriorly extending Meckelian groove that accommodates the splenial (Figure 9B).
Splenial
The splenial is an anteroposteriorly elongated, flat element that rests medial to the horizontal ramus of the dentary within the Meckelian groove, an anteroposterior-extending groove on the ventromedial margin of the dentary (Figure 9B). This groove would also have housed the Meckelian cartilage in life. The preserved left splenial lacks its anterior portion, but it likely extended far enough anteriorly to contribute to the mandibular symphysis, whereas posteriorly it reaches beyond the level of the dentary boss. The bone thickens dorsoventrally toward its midpoint, attaining its greatest height just posterior to the last postcanine. Its posterior process tapers ventrally to a pointy tip and meets the anterior processes of the prearticular and angular dorsomedially.
Angular
The angular is a dorsoventrally tall bone that tapers anteriorly, where it rests medial to the posteroventral margin of the dentary. The anterior process of the angular meets the anterior process of the surangular dorsally and the posterior processes of the splenial ventrally. It also exhibits a medial shelf along its entire length which accommodates the anterior prearticular (Figure 9B). The posterior portion of the angular bears a well-developed, posterodorsally oriented ridge—the angular crest—which is positioned at the anterior edge of the dorsal notch, as is typical for therocephalians (Olroyd and Sidor, 2022). The typical pattern of radiating ridges and four fossae described for other therocephalians on the reflected lamina (Olroyd and Sidor, 2022) cannot be fully discerned in SAM-PK-8797 due to damage to the ventral portion of the angular. However, the dorsalmost contact between the reflected lamina and the body of the angular is clearly visible, where it curves slightly dorsally to form the dorsal notch (Figure 9A). The reflected lamina is platelike and broad, but its dorsoventral extent is reduced relative to that of other non-epicynodont eutherocephalians, resembling the condition in other lycideopid baurioids (Huttenlocker, 2013). As in other eutherocephalians, the mandibular fenestra is present between the angular and the dentary, but it is comparatively small and is almost entirely obscured in lateral view by the coronoid process of the dentary.
Surangular
The surangular is a flat, plate-like element that arches posterior and medial to the coronoid process, where it rests dorsal to the prearticular and medial to the angular, contributing to the dorsal border of the mandibular fenestra in medial view (Figure 9B). Anterior to the fenestra, it curves gently downward to underlie the coronoid. Posteriorly, the surangular exhibits a medially projecting, hook-like extension at mid-height, which supports an anterior extension of the quadrate and a posterodorsal extension of the prearticular. In lateral view, the surangular is almost completely obscured by the dentary and the main body of the angular, except for a narrow posterior strip (Figure 9A). The posteriormost fringe of the surangular has a narrow dorsoventral thickening.
Coronoid
Only the right coronoid is preserved in place, with the left coronoid having been displaced dorsally into the nasal cavity. It is a thin, arrow-shaped element. The dorsal portion of the right coronoid lies medial to the dentary, just anterior to the coronoid process, and its ventral portion rests lateral to the anterior processes of the surangular and prearticular (Figure 9B). Anteriorly, the coronoid narrows dorsally and tapers to a pointed tip, forming an anterior process approximately at the level of the posteriormost margin of the splenial. Posteriorly, it broadens and divides into a wider posteroventral and a narrower posterodorsal extension, separated by a curved indentation.
Prearticular
The prearticular is an elongated, rod-like element that rests medially on an elongated medial shelf of the angular and contributes to the ventral margin of the mandibular fenestra (Figures 9A, B). Its long anterior process meets the posteriormost tip of the splenial. The posterior margin of the prearticular has a short medial extension, and on that medial surface it forms a sharp, perpendicular crest. The area posterior to that crest would have accommodated the articular when intact.
Dentition
Upper dentition
Only the left upper tooth row is almost completely preserved. Because most of the tooth-bearing ramus of the premaxilla is missing on the right side, only the two posteriormost right incisors are preserved (Figure 10A). By contrast, the left premaxilla preserves seven conical incisors (Figure 10B), a condition also observed in the closely related Tetracynodon (Sigurdsen et al., 2012), but unlike most baurioids, which possess six upper incisors, and derived bauriamorphs such as Bauria and Microgomphodon, which have only four (Maisch, 2017). Among these, only I5–I7 are in intact condition, whereas I1–I4 have damaged crowns. The two preserved posteriormost right incisors and the left I5–I7 appear similar in size and lack serrations, as in the lower dentition. Each maxilla bears three precanine teeth, of which only the right precanines are intact, with the second tooth being considerably longer than the first and the third tooth (Figure 10A). Both maxillae preserve a single, large canine positioned at the level of the posterior margin of the maxillary boss. The crown of this canine is broken on both sides (Figures 10A, B), preventing assessment of its full size or the presence of cutting edges. Twelve conical postcanines are present in each maxilla. On the right side, PC4–PC12 are completely preserved, although PC12 has slipped out of its socket. In the left maxilla, only PC1, PC2 and PC7–PC10 appear to be in intact condition. The postcanines increase slightly in size and diameter toward the posterior end of the tooth row.
Lower dentition
Because a major portion of the left dentary is broken away, only the right lower tooth row is nearly completely preserved. Like in the upper dentition, the lower incisors and postcanines are simple, conical teeth, without distal or mesial serrations. Based on the left dentary, SAM-PK-8797 possessed four lower incisors (Figure 11B), consistent with the typical eutherocephalian condition (Huttenlocker, 2009). Because the anterior portions of the dentaries and the symphysis are missing, only the left i2–i5 remain in place, whereas the left i1 is represented only by its root. By contrast, in the right dentary only the roots of the possible i2 and i3 are preserved (Figure 11A). The lower incisors are similar in size and shape, and slightly recurved posteriorly, although the cutting margin of i4 appears sharper and its tip is more pointed. They have a smooth surface, resembling the condition present in the closely related Tetracynodon, but differing from other more stemward baurioids, in which the surface is striated, and from akidnognathids, in which the surface is faceted (Sigurdsen et al., 2012). SAM-PK-8797 has a single, large lower canine preserved on both sides, which is pointed and more than twice the height of the other lower teeth (Figures 11A, B). The right dentary preserves a total of 16 postcanines, of which pc2, pc5, and pc7 have fallen out. Most right lower postcanines have damaged or missing crowns, with only pc12–pc15 being completely preserved. By contrast, the left dentary retains pc1–pc7, of which pc1–pc6 are well preserved, whereas pc7 is represented only by its root.
Ear region
Quadrate
The quadrate forms the posterolateral margin of the occiput. It typically consists of two distinct portions: a dorsal plate and a ventral trochlea. In SAM-PK-8797, only the trochlea of the right quadrate is preserved (Figure 9). As in other therocephalians, the trochlea is broad (Botha-Brink and Modesto, 2011), and its dorsal surface is concave. Anteriorly, it bears a thin lamina that is inclined ventrally in this specimen. In life, this lamina would have contacted the quadrate process of the pterygoid, but due to postmortem displacement, the pterygoid process now lies beneath the entire quadrate and contacts the ventral surface of the trochlea. The trochlea itself would normally have formed a broad articulation with the dorsal surface of the articular to form the jaw joint. The dorsal plate of the quadrate, although not preserved, would have been positioned between the ventrolateral margin of the squamosal posteriorly and the lateralmost margin of the mastoid process of the opisthotic medially. The quadratojugals are not preserved.
Stapes
Both stapes are preserved, with the right stapes lying roughly in their original position, whereas the left stapes has been displaced far from its in-situ orientation (Figure 9C). The stapes of Choerosaurus is dumbbell-shaped, with the lateral and medial ends being similar in size. It likely had a transverse orientation in life, with the lateral footplate contacting the quadrate and the medial plate articulating with the fenestra ovalis. The horizontal shaft is broader anteroposteriorly and narrower dorsoventrally. As in other therocephalians with preserved stapes, the stapes in Choerosaurus lacks a stapedial foramen (Huttenlocker and Abdala, 2015; Huttenlocker and Sidor, 2016; Abdala et al., 2014; Kammerer and Masyutin, 2018).
4. Discussion
Evolution of the secondary palate in baurioids
Several bones contribute to the secondary palate in SAM-PK-8797: the premaxillae, maxillae, vomer, and palatines. The baurioid anterior margin of the nasal cavity or secondary palate shows a pattern that is distinct from other non-mammalian synapsids. Whereas basal ictidosuchid baurioids like Ictidosuchoides (Hillenius and Ruben, 2004) show almost no internal extension of the maxilla to contribute to the secondary palate (which is more similar to the condition in more basally positioned therocephalians such as the eutherocephalian Ichibengops munyamadziensis, Figure 12), more derived representatives like lycideopids have a long sutural connection between the vomer and palatal process of the maxilla to build an incipient hard palate (Huttenlocker, 2013). In Choerosaurus, that maxillo-vomerine bridge shows a long extension from the level of the large canine to the fourth postcanine (Figures 5, 12), very similar to what is observed in its close relatives Tetracynodon darti (Sigurdsen et al., 2012) and Lycideops longiceps (Mendrez, 1975), with a possibly longer suture in Lycideops (Huttenlocker, 2013). This condition also resembles what was described for Mupashi migrator, where the maxillary crista choanalis contacts the widened anteroventral margin of the vomer (Huttenlocker and Sidor, 2016). In regisaurid baurioids, the extension of this bridge is not as well developed as in Lycideopidae and Karenitidae. For example, in Regisaurus jacobi that structure prolongs from the level of the canine to the level of the first postcanine only (Mendrez, 1972). The scaloposaurid Scaloposaurus constrictus, which is typically recovered just basal to the Bauriamorpha (Sidor et al., 2022), exhibits a broadened maxillary shelf that contacts the vomer in a deep suture from the canine to the first postcanine. However, the anterior palate differs from most other post-ictidosuchid baurioids in that the vomerine process of the premaxilla does not contact the maxilla medially to the large canine fossa, allowing the vomer to contribute to that fossa posteromedially (Huttenlocker et al., 2022). A similar condition is observed in the Antarctic and Chinese bauriamorphs Notictoides absens and Hazhenia concava, respectively (Sidor et al., 2022). In contrast, in the lycideopids Choerosaurus, Tetracynodon, and Lycideops, as well as in the karenitid Mupashi and the regisaurid Regisaurus, the vomer is excluded from the canine fossa. Only in baurioids more derived than Scaloposaurus do the maxillary shelves meet at the midline to create a completely closed hard palate and thus exclude the anterior vomer in ventral view. This is the earliest appearance of a true hard palate in therocephalians resembling what is seen in mammals. In Ericiolacertidae, the earliest-diverging bauriamorphs, the suture between both shelves ends more anteriorly compared to other representatives (Figure 12), allowing the vomer to form a wedge in-between those structures (based on Ericiolacerta parva; Mendrez, 1975). However, at the anterior margin of the palate, the premaxillary and maxillary shelves close completely, obscuring the vomer and forming the anterior and posterior halves of the canine fossa respectively. In the more derived bauriamorph Notictoides absens, the maxillary shelves meet on the midline only up to the level of the canine in anterior direction, so that the anterior process of the vomer still contacts the palatal process of the premaxilla. Here, the vomer shares a contact with the dorsal maxillary surface, but without contribution to the secondary palate. Posteriorly, the maxillary shelves reach the level of about the fourth postcanine (Sidor et al., 2022). Bauriidae represent the only herbivorous clade of therocephalians and exhibit transversely expanded postcanine teeth convergent with the molars of mammals and the postcanines of gomphodont cynodonts (Abdala et al., 2014). The closed osseous secondary palate reaches posteriorly to the level of the sixth postcanine in Bauria cynops (Figure 12; Brink, 1963) and has a proportionally similar length in Microgomphodon oligocynus (Abdala et al., 2014).
The development of a secondary palate represents one of the most significant morphological innovations among therapsids and evolved convergently at least four times within the group: in anomodonts (within Dicynodontia), in therocephalians (within Whaitsiidae and Baurioidea), and during early cynodont evolution Brink, 1960; Mendrez, 1975; Hopson and Barghusen, 1986; Maier, 1999; Angielczyk, 2009; Huttenlocker, 2009). In dicynodonts, the secondary palate is formed by a posterior extension of the fused premaxilla (Angielczyk and Walsh, 2008). All cynodonts exhibit a secondary palate formed by the medial projections of the maxillae and the associated medial projections of the palatines. However, most of the earliest (Permo–Triassic) cynodonts lack a complete secondary palate, as the palatal processes of the maxillae and palatines do not meet along the ventral midline of the palate (e.g., Kemp, 1979; Sidor and Smith, 2004; Ivakhnenko, 2013; Abdala et al., 2019; Pusch et al., 2019, 2023, 2024; Huttenlocker and Sidor, 2020). The medial extent of these palatal processes vary along cynodont phylogeny: epicynodonts, such as the galesaurids Progalesaurus (Sidor and Smith, 2004) and Galesaurus (Pusch et al., 2019) as well as the possible galesaurid Nythosaurus (Pusch et al., 2023), exhibit greater transverse expansion of these medial palatal processes than non-epicynodont cynodonts such as charassognathids, Dvinia, and Procynosuchus (Pusch et al., 2024). The apparent contact of the palatal processes of the maxillae observed in the type and only known specimen of the Permian charassognathid Charassognathus gracilis (Pusch et al., 2024: Figure 2E) is clearly an artefact of deformation caused by strong lateral compression of the snout in SAM-PK-K10369. Similar deformation-induced contacts between the palatal processes of the maxillae have also been reported in some specimens of the Permian Procynosuchus (Abdala and Allinson, 2005) and the Early Triassic Galesaurus (Pusch et al., 2019). Undistorted skulls of Procynosuchus and Galesaurus show that these taxa did not have a complete secondary palate (Jasinoski and Abdala, 2017). Crompton et al. (2015) proposed that the midline gap between the medially projecting palatal shelves of the maxillae and palatines in these taxa was probably occupied by soft tissue in life.
The Early Triassic Thrinaxodon represents the most stemward occurrence of a definite complete secondary palate within Cynodontia (e.g., Fourie, 1974; Hopson, 1991; Crompton et al., 2015, 2017). In this taxon, the medial projections (transverse laminae) of the maxillae and palatines meet along the midline. Anteriorly, the maxillae and premaxillae together enclose the paired incisive foramina, which partially separate the premaxillae from the rest of the palate. The maxillae are the main contributor and form the bulk of the palatal plate. From that point of the phylogeny onward, the basic architecture of the secondary palate is maintained in all crownward cynodonts (including mammals), only altered by a posterior extension of the palatine plate contribution, a further premaxillary expansion, and a decrease in incisive foramen size resulting in a broader occlusion of the vomer in probainognathian eucynodonts. Those trends reach the final state observed in the modern mammalian condition beginning with chiniquodontids in the late Middle/early Late Triassic (Angielczyk and Kammerer, 2018).
Similar to cynodonts, where the secondary palate evolved gradually towards a closed state in the late Permian and only reached that condition in the Triassic, late Permian baurioids also show a trend towards closure of the palate. However, the intermediate state within therocephalians is characterized by a sutural contact between maxillary shelves and the anterior vomer forming the maxillo-vomerine bridge, as seen in lycideopids, regisaurids, and scaloposaurids. Direct contact of the maxillary shelves at the midline is seen in Early to Middle Triassic bauriamorphs like ericiolacertids and Chinese ‘ordosiids’ such as Hazhenia concava and Ordosiodon youngi (Figure 12; Sun, 1991; Huttenlocker, 2013). In those forms the palatal processes of the maxillae cover the vomer ventrally, excluding it from the external surface of the palate. Finally, bauriids like Bauria exhibit a condition quite similar to what is observed in basal eucynodonts, with a complete palatal plate that extends around halfway through the posterior tooth row (Abdala et al., 2014). Despite the morphological similarity of that region between early cynodonts and advanced baurioid therocephalians, a key difference represents the participation of the palatine in the secondary palate, which is unique to Cynodontia among synapsids (palatine pads posterior to the premaxillary palatal plate are present in Dicynodontia, but do not contact each other on the midline; Angielczyk and Kammerer, 2018; Pusch et al., 2024).
The functional aspects of the independently derived secondary palates in different synapsid clades are likely to be somewhat similar – strengthening the snout, providing an enhanced surface for food processing, and maintaining respiratory function during food intake (Thomason and Russel, 1986; Angielczyk, 2009) – and in all three groups an argument can be made that this structure represents a ‘key innovation’ related to their diversification (cynodonts as the only surviving synapsids, dicynodonts as the most abundant and species-rich Permo-Triassic synapsids, and bauriamorphs as the only successful Triassic radiation of therocephalians). Trends toward strengthening the snout region may have been tied to different selective pressures in different groups, such as increasing bite force in predatory forms or permitting prolonged mastication in herbivorous forms. The evolution of specialized mastication techniques that generate increased torsional loads on the rostrum, such as lateral jaw movements, might require a stronger snout as well (Preuschoft et al., 2024). In hypercarnivorous forms, those adaptations may alternatively be related to prey handling techniques like lateral head shaking and holding of resisting prey (McCurry et al., 2015). In dicynodonts, the palate was rugose and likely at least partially cornified, producing a broad triturating surface for processing plant matter (Crompton and Hotton, 1967; Angielczyk, 2004; Angielczyk and Kammerer, 2018). In addition to permitting attachment of a hypothetical keratinous covering, the thickening and sculpturing of the palatal bones in dicynodonts may also have strengthened the palate as a whole. For instance, a study of phenotypical diversity in rabbits demonstrated that individuals fed harder food items developed thicker palatal bones (Menegaz et al., 2009). As the only herbivorous subclade of therocephalians, bauriids may also have benefited from having a more resistant palatal surface while processing high fiber forage. In archosaur evolution, a fully closed secondary palate comparable to that of cynodonts among synapsids is present in crocodylomorphs. Hypercarnivorous diets in this group have been proposed as a driver of palatal closure, as a closed secondary palate may reinforce the snout and thereby facilitate the generation of greater bite forces (Dollman and Choiniere, 2022). The repeated independent evolution of an ossified secondary palate among synapsids may therefore reflect increased masticatory loading along several trophic trajectories. In Cynodontia, the secondary palate is also considered a prerequisite for other evolutionary innovations in modern mammals that increased their survival chances throughout time, such as higher metabolic rates und ultimately endothermy, and simultaneously playing an important role in the development of suckling behavior due to generation of negative pressure in the mouth (Maier et al., 1996; Maier, 1999). This level of integration of the secondary palate into a suite of broader physiological aspects of the lineage did not occur in therocephalians or anomodonts, however, which went extinct during the Triassic.
Further research is needed on the underlying causes of the development of closed secondary palates in synapsid groups and their potential importance for these groups’ long-term survival (Angielczyk and Walsh, 2008; Maier, 1999). Vital to such work is a better understanding of the anatomy of the palate in basal members of the involved groups (early anomodonts, baurioid therocephalians, and cynodonts), which is generally more poorly known than that of their highly derived subclades (dicynodonts, bauriids, eucynodonts). The description here of the palate of Choerosaurus is one such contribution to understanding the initial evolutionary history of synapsid secondary palate development, building on recent comprehensive sampling in the earliest cynodonts (Pusch et al., 2024). Future CT-assisted descriptive research targeting other early baurioids and non-dicynodont anomodonts would be helpful in terms of understanding the morphological changes involved in palatal evolution in these groups. More functional and paleoecological studies of the involved clades are also clearly needed to help understand the selective pressures underlying these changes.
5. Conclusions
The craniomandibular anatomy of lycideopid therocephalians has historically been understudied. Here we have presented the first detailed description of the cranial anatomy of the aberrant Permian therocephalian Choerosaurus dejageri, providing novel insights into the anatomy of this enigmatic animal using CT-modelling. Like other lycideopids, Choerosaurus presents a mixture of cynodont-like and primitive therocephalian features, supporting the idea of widespread convergence between these groups. The most striking such convergent character is the medial expansion of the maxillary shelves that make sutural contact with the anterolateral vomers, forming a maxillo-vomerine bridge. The presence of an expansive maxillo-vomerine bridge in lycideopids such as Choerosaurus represents an important morphological novelty within baurioid therocephalians, presaging the fully closed secondary palate in advanced bauriamorphs such as Bauria. The baurioid secondary palate differs in the architecture of the component bones from that of dicynodonts and cynodonts, but it is likely that similar selective pressures underlie the origin of this feature in the different eutheriodont lineages.
Statements
- Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors JB, JF declared that they were an editorial board member of Frontiers at the time of submission. This had no impact on the peer review process and the final decision.
- Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by a grant from the Deutsche Forschungsgemeinschaft (FR 2457/8-1) to Jörg Föbisch and by one from Humboldt-Universität zu Berlin, Open Access Publication Fund to Alessandro Gigliotti.
- Data availability
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. For requests concerning access to CT scan data on SAM-PK-8797, please contact Zaituna Skozan (Collection Manager, Iziko Museum Cape Town: zskosan@iziko.org.za).
- Ethics approval
Ethical approval was not required for the study involving animals in accordance with the local legislation and institutional requirements because the studied specimen is a fossil housed at the Iziko Museum in Cape Town and was acquired legally.
- Author contributions
AG: Software, Methodology, Data curation, Formal analysis, Visualization, Conceptualization, Investigation, Writing – review & editing, Writing – original draft. LP: Writing – review & editing, Supervision, Methodology, Investigation, Formal analysis. CK: Methodology, Supervision, Writing – review & editing, Investigation, Formal analysis. JB: Data curation, Writing – review & editing. JF: Resources, Writing – review & editing, Funding acquisition, Project administration, Supervision.
- Acknowledgements
The authors wish to thank Z. Skosan (SAM) for facilitating access to the specimen SAM-PK-8797, and the financial support of the DSTI-NRF GENUS Centre of Excellence in Palaeosciences and African Origins Platform (AOP240418214774). Furthermore, A. Gigliotti wishes to thank the entire J. Fröbisch AG and colleagues at the MfN for the continuous support throughout the years.
References
- Abdala F., Allinson M. (2005). The taxonomic status of Parathrinaxodon proops (Therapsida: Cynodontia), with comments on the morphology of the palate in basal cynodonts. Palaeontologia Africana 41, 45 – 52.
- Abdala F., Gaetano L., Smith R. M. H., Rubidge B. S. (2019). A new large cynodont from the Late Permian (Lopingian) of the South African Karoo Basin and its phylogenetic significance. Zool. J. Linn. Soc 186, 983 – 1005. doi: doi:10.1093/zoolinnean/zlz004
- Abdala F., Jashashvili T., Rubidge B. S., van den Heever J. (2014). “ New material of Microgomphodon oligocynus (Eutherapsida, Therocephalia) and the taxonomy of Southern African Bauriidae,” in Early Evolutionary History of the Synapsida. Vertebrate Paleobiology and Paleoanthropology. Eds. Kammerer C., Angielczyk K., Fröbisch J. (Springer, Dordrecht), 209 – 231.
- Abdala F., Rubidge B. S., van den Heever J. (2008). The oldest therocephalians (Therapsida, Eutheriodontia) and the earliest diversification of Therapsida. Palaeontology 51, 1011 – 1024. doi: doi:10.1111/j.1475-4983.2008.00784.x
- Angielczyk K. A. (2004). Phylogenetic evidence for and implications of a dual origin of propaliny in anomodont therapsids (Synapsida). Paleobiology 30, 268 – 296. doi: doi:10.1666/0094-8373(2004)030<0268:PEFAIO>2.0.CO;2
- Angielczyk K. D. (2009). Dimetrodon is not a dinosaur: using tree thinking to understand the ancient relatives of mammals and their evolution. Evolution: Educ. Outreach 2, 257 – 271. doi: doi:10.1007/s12052-009-0117-4
- Angielczyk K. D., Kammerer C. F. (2018). “ 5. Non-mammalian synapsids: the deep roots of the mammalian family tree,” in Mammalian Evolution, Diversity and Systematics. Eds. Zachos F. E., Asher R. J. (de Gruyter, Berlin), 117 – 198.
- Angielczyk K. D., Walsh M. L. (2008). Patterns in the evolution of naris size and secondary palate length in anomodont therapsids (Synapsida): implications for hypoxia as a cause of end-Permian tetrapod extinctions. J. Paleontol. 82, 528 – 542. doi: doi:10.1666/07-051.1
- Araújo R., Fernandez V., Polcyn M. J., Fröbisch J., Martins R. M. (2017). Aspects of gorgonopsian paleobiology and evolution: insights from the basicranium, occiput, osseous labyrinth, vasculature, and neuroanatomy. PeerJ 5, e3119. doi: doi:10.7717/peerj.3119
- Bendel E. M., Kammerer C. F., Kardjolov N., Fernandez V., Fröbisch J. (2018). Cranial anatomy of the gorgonopsian Cynariops robustus based on CT-reconstruction. PloS One 13, e0207367. doi: doi:10.1371/journal.pone.0207367
- Benoit J., Jasinoski S. C., Fernandez V., Abdala F. (2017). The mystery of a missing bone: revealing the orbitosphenoid in basal Epicynodontia (Cynodontia, Therapsida) through computed tomograophy. Sci. Nat. 104, 1 – 10. doi: doi:10.1007/s00114-017-1487-z
- Benoit J., Manger P. R., Fernandez V., Rubidge B. S. (2016). Cranial bosses of Choerosaurus dejageri (Therapsida, Therocephalia): earliest evidence of cranial display structures in eutheriodonts. PloS One 11, e0161457. doi: doi:10.1371/journal.pone.0161457
- Bolton A. D., Mangera T., Benoit J. (2025). 150 years of synapsid paleoneurology: the origins of the mammalian brain, behavior, sense organs and physiology. J. Paleontol., 1 – 29. doi: doi:10.1017/jpa.2025.10121
- Boonstra L. D. (1934). A contribution to the morphology of the mammal-like reptiles of the Suborder Therocephalia. Ann. S. Afr. Mus. 31, 215 – 267.
- Botha-Brink J., Modesto S. P. (2011). A new skeleton of the therocephalian synapsid Olivierosuchus parringtoni from the Lower Triassic South African Karoo Basin. Palaeontology 54, 591 – 606. doi: doi:10.1111/j.1475-4983.2011.01048.x
- Botha J., Smith R. M. H. (2020). Biostratigraphy of the Lystrosaurus declivis Assemblage Zone (Beaufort group, Karoo supergroup), South Africa. S. Afr. J. Geol. 123, 207 – 216. doi: doi:10.25131/sajg.123.0015
- Brink A. S. (1960). On some small therocephalians. Palaeontologia Africana 7, 155 – 182.
- Brink A. S. (1963). The taxonomic position of Synapsida. S. Afr. J. Sci. 59, 153 – 159.
- Crompton A. W., Hotton N. (1967). Functional morphology of the masticatory apparatus of two dicynodonts (Reptilia, Therapsida). Postilla 109, 1 – 51.
- Crompton A. W., Musinsky C., Owerkowizc T. (2015). “ Evolution of the mammalian nose,” in Great Transformations in Vertebrate Evolution. Eds. Dial K. P., Shubin N., Brainerd E. (University of Chicago Press, Chicago, IL), 189 – 203.
- Crompton A. W., Oowerkowicz T., Musinsky C. (2017). Structure of the nasal region of non-mammalian cynodonts and mammaliaforms: speculations on the evolution of mammalian endothermy. J. Vertebr. Paleontol. 37, e1269116-3. doi: doi:10.1080/02724634.2017.1269116
- Crompton A. W., Musinsky C., Rougier G. W., Bhullar B. A. S., Miyamae J. A. (2018). Origin of the lateral wall of the mammalian skull: fossils, monotremes and therians revisited. Journal of Mammalian Evolution, 25, 301 – 313.
- Day M. O., Smith R. M. H. (2020). Biostratigraphy of the Endothiodon Assemblage Zone (Beaufort group, Karoo supergroup), South Africa. S. Afr. J. Geol. 123, 165 – 180. doi: doi:10.25131/sajg.123.0011
- Dollman K. N., Choiniere J. N. (2022). Palate evolution in early-branching crocodylomorphs: implications for homology, systematics, and ecomorphology. Anat. Rec. 305, 2766 – 2790. doi: doi:10.1002/ar.24993
- Durand J. F. (1991). A revised description of the skull of Moschorhinus (Therapsida, Therocephalia). Ann. South. Afr. Museum 99, 381 – 413.
- Fourie S. (1974). The cranial morphology of Thrinaxodon liorhinus. Ann. South. Afr. Museum 65, 337 – 400.
- Gigliotti A., Pusch L. C., Kammerer C. F., Benoit J., Fröbisch J. (2023). Craniomandibular anatomy of the akidnognathid therocephalian Olivierosuchus parringtoni from the Early Triassic of South Africa. Palaeontologia Africana 56, 142 – 170. Available online at: https://hdl.handle.net/10539/37185.
- Grunert H. R., Brocklehurst N., Fröbisch J. (2019). Diversity and disparity of Therocephalia: macroevolutionary patterns through two mass extinctions. Sci. Rep. 9, 5063. doi: doi:10.1038/s41598-019-41628-w
- Hancox P. J., Neveling J., Rubidge B. S. (2020). Biostratigraphy of the Cynognathus Assemblage Zone (Beaufort group, Karoo supergroup), South Africa. S. Afr. J. Geol. 123, 217 – 238. doi: doi:10.25131/sajg.123.0016
- Haughton S. H. (1929). On some new therapsid genera. Ann. South. Afr. Museum 28, 55 – 78.
- Hillenius W. J. (1991). The evolution of nasal turbinates and mammalian endothermy. Paleobiology 18, 17 – 29. doi: doi:10.1017/s0094837300012197
- Hillenius W. J. (1994). Turbinates in therapsids: evidence for Late Permian origins of mammalian endothermy. Evolution 48, 207 – 229. doi: doi:10.2307/2410089
- Hillenius W. J., Ruben J. A. (2004). The evolution of endothermy in terrestrial vertebrates: Who? When? Why ? Physiol. Biochem. Zoology: Ecol. Evolutionary Approaches 77, 1019 – 1042. doi: doi:10.1086/425185
- Hopson J. A. (1991). “ Systematics on the nonmammalian synapsida and implications for patterns of evolution in synapsids,” in Origins of the Higher Groups of Tetrapods: Controversy and Consensus. Eds. Schultze H. P., Trueb L. (Comstock Publishing Associates, Ithaca, NY), 635 – 693.
- Hopson J. A., Barghusen H. (1986). “ An analysis of therapsid relationships,” in The Ecology and Biology of Mammal-Like Reptiles. Eds. Hotton N., MacLean P. D., Roth J. J., Roth E. C. (Smithsonian Institution Press, Washington, DC), 83 – 106.
- Huttenlocker A. (2009). An investigation into the cladistic relationships and monophyly of therocephalian therapsids (Amniota: Synapsida). Zool. J. Linn. Soc 157, 865 – 891. doi: doi:10.1111/j.1096-3642.2009.00538.x
- Huttenlocker A. K. (2013). The Paleobiology of South African Therocephalian Therapsids (Amniota, Synapsida) and the Effects of the End-Permian Extinction on Size, Growth, and Bone Microstructure (University of Washington, Seattle, WA).
- Huttenlocker A. K. (2014). Body size reductions in nonmammalian eutheriodont therapsids (Synapsida) during the End-Permian mass extinction. PloS One 9, e87553. doi: doi:10.1371/journal.pone.0087553
- Huttenlocker A. K., Abdala F. (2015). Revision of the first therocephalian, Theriognathus Owen (Therapsida: Whaitsiidae), and implications for cranial ontogeny and allometry in nonmammaliaform eutheriodonts. J. Paleontol. 89, 645 – 664. doi: doi:10.1017/jpa.2015.32
- Huttenlocker A. K., Botha J., Browning C., Kulik Z., Tshibalanganda M., du Plessis A. (2022). A Gulliver Scaloposaurus (Therapsida, Therocephalia) from the Katberg Formation of South Africa and its implication for Lilliput assemblages during the Early Triassic recovery. J. Afr. Earth. Sci. 196, 104720. doi: doi:10.1016/j.jafrearsci.2022.104720
- Huttenlocker A. K., Sidor C. A. (2012). Taxonomic revision of therocephalians (Therapsida, Theriodontia) from the Lower Triassic of Antarctica. Am. Mus. Novit. 3738, 1 – 19. doi: doi:10.1206/3738.2
- Huttenlocker A. K., Sidor C. A. (2016). The first karenitid (Therapsida, Therocephalia) from the upper Permian of Gondwana and the biogeography of Permo-Triassic therocephalians. J. Vertebr. Paleontol. 36, e1111897. doi: doi:10.1080/02724634.2016.1111897
- Huttenlocker A. K., Sidor C. A. (2020). A basal nonmammaliaform cynodont from the Permian of Zambia and the origins of mammalian endocranial and postcranial anatomy. J. Vertebr. Paleontol. 40, e1827413. doi: doi:10.1080/02724634.2020.1827413
- Huttenlocker A. K., Sidor C., Angielczyk K. D. (2015). A new eutherocephalian (Therapsida, Therocephalia) from the upper Permian Madumabisa Mudstone Formation (Luangwa Basin) of Zambia. J. Vertebr. Paleontol. 35, e969400. doi: doi:10.1080/02724634.2015.969400
- Huttenlocker A. K., Sidor C. A., Smith R. M. H. (2011). A new specimen of Promoschorhynchus (Therapsida: Therocephalia: Akidnognathidae) from the Lower Triassic of South Africa and its implications for theriodont survivorship across the Permo-Triassic boundary. J. Vertebr. Paleontol. 31, 405 – 421. doi: doi:10.1080/02724634.2011.546720
- Huttenlocker A. K., Smith R. M. H. (2017). New whaitsioids (Therapsida: Therocephalia) from the Teekloof Formation of South Africa and therocephalian diversity during the end-Guadalupian extinction. PeerJ 5, e3868. doi: doi:10.7717/peerj.3868
- Ivakhnenko M. F. (2011). Permian and Triassic therocephals (Eutherapsida) of Eastern Europe. Paleontol. J. 45, 981 – 1144. doi: doi:10.1134/S0031030111090012
- Ivakhnenko M. F. (2013). Cranial morphology of Dvinia prima Amalitzky (Cynodontia, Thermomorpha). Paleontological Journal 47, 210 – 222. doi: doi:10.1134/S0031030113010048
- Jasinoski S. C., Abdala F. (2017). Cranial ontogeny of the Early Triassic basal cynodont Galesaurus planiceps. Anat. Rec. 300, 353 – 381. doi: doi:10.1002/ar.23116
- Kammerer C. F., Masyutin V. (2018). A new therocephalian (Gorynychus masyutinae gen. et sp. nov.) from the Permian Kotelnich locality, Kirov Region, Russia. PeerJ 6, e4933. doi: doi:10.7717/peerj.4933
- Kemp T. S. (1979). The primitive cynodont Procynosuchus: functional anatomy of the skull and relationships. Philos. Trans. R. Soc. 285, 73 – 122. doi: doi:10.1098/rstb.1979.0001
- Kemp T. S. (1982). Mammal-like reptiles and the origin of mammals. Syst. Biol. 31, 526 – 527. doi: doi:10.1093/sysbio/31.4.526
- Kemp T. S. (1986). The skeleton of a baurioid therocephalian therapsid from the Lower Triassic (Lystrosaurus Zone) of South Africa. J. Vertebr. Paleontol. 6, 215 – 232. doi: doi:10.1080/02724634.1986.10011617
- Kemp T. S. (2009). The endocranial cavity of a nonmammalian eucynodont, Chiniquodon theotenicus, and its implications for the origin of the mammalian brain. J. Vertebr. Paleontol. 29, 1188 – 1198. doi: doi:10.1671/039.029.0430
- Kitching J. W. (1977). The distribution of the Karroo vertebrate fauna (Bernard Price Institute for Palaeontological Research, University of the Witwatersrand, Johannesburg).
- Liu J., Abdala F. (2017). The tetrapod fauna of the upper Permian Naobaogou Formation of China: 1. Shiguaignathus wangi gen. et sp. nov., the first akidnognathid therocephalian from China. PeerJ 5, e4150. doi: doi:10.7717/peerj.4150
- Liu J., Abdala F. (2019). The tetrapod fauna of the upper Permian Naobaogou Formation of China: 3. Jiufengia jiai gen. et sp. nov. a large akidnognathid therocephalian. PeerJ 7, e6463. doi: doi:10.7717/peerj.6463
- Liu J., Abdala F. (2020). The tetrapod fauna of the upper Permian Naobaogou Formation of China: 5. Caodeyao liuyufengi gen. et sp. nov. a new peculiar therocephalian. PeerJ 8, e9160. doi: doi:10.7717/peerj.9160
- Liu J., Abdala F. (2022). The emblematic South African therocephalian Euchambersia in China: a new link in the dispersal of late Permian vertebrates across Pangea. Biol. Lett. 18, 20220222. doi: doi:10.1098/rsbl.2022.0222
- Liu J., Abdala F. (2023). Late Permian terrestrial faunal connections invigorated: the first whaitsioid therocephalian from China. Palaeontologia Africana 56, 111 – 117.
- Lloyd J. K., Durand F. (2025). A new akidnognathid synapsid specimen from the Permian of Cradock, South Africa and the revision of Hewittia albanensis. Acta Palaeontol. Pol. 70, 339 – 355. doi: doi:10.4202/app.01187.2024
- Maier W. (1999). On the evolutionary biology of early mammals – with methodological remarks on the interaction between ontogenetic adaptation and phylogenetic transformation. Zool. Anz. 238, 55 – 74.
- Maier W., van den Heever J., Durand F. (1996). New therapsid specimens and the origin of the secondary hard and soft palate of mammals. J. Zool. Syst. Evol. Res. 34, 9 – 19. doi: doi:10.1111/j.1439-0469.1996.tb00805.x
- Maisch M. W. (2017). Re-assessment of Silphoictidoides ruhuhuensis von Huene 1950 (Therapsida, Therocephalia) from the Late Permian of Tanzania: one of the most basal baurioids known. Palaeodiversity 10, 25 – 39. doi: doi:10.18476/pale.v10.a3
- McCurry M. R., Mahony M., Clausen P. D., Quayle M. R., Walmsley C. W., Jessop T. S.. (2015). The relationship between cranial structure, biomechanical performance and ecological diversity in varanoid lizards. PloS One 10, e0130625. doi: doi:10.1371/journal.pone.0130625
- Mendrez C. H. (1972). “ On the skull of Regisaurus jacobi, a new genus and species of Bauriamorpha Watson and Romer 1956 (= Scaloposauria Boonstra 1953), from the Lystrosaurus-zone of South Africa,” in Studies in Vertebrate Evolution. Eds. Joysey K. A., Kemp T. S. (Oliver and Boyd, Edinburgh), 191 – 212.
- Mendrez C. H. (1974). A new specimen of Promoschorhynchus platyrhinus Brink 1954 (Moschorhinidae) from the Daptocephalus -Zone (Upper Permian) of South Africa. Palaeontologia Africana 17, 69 – 85.
- Mendrez C. H. (1975). Principales variations du palais chez les therocephales sud-africains (Pristerosauria et Scaloposauria) au cours du Permien superieur et du Trias inferieur. Colloque Int. CNRS 218, 379 – 408.
- Menegaz R. A., Sublett S. V., Figueroa S. D., Hoffman T. J., Ravosa M. J. (2009). Phenotypic plasticity and function of the hard palate in growing rabbits. Anat. Rec. 292, 277 – 284. doi: doi:10.1002/ar.20840
- Olroyd S. L., Sidor C. A. (2022). Nomenclature, comparative anatomy, and evolution of the reflected lamina of the angular in non-mammalian synapsids. J. Vertebr. Paleontol. 42, e2101923. doi: doi:10.1080/02724634.2022.2101923
- Preuschoft H., Martin T., Werneburg I. (2024). Remarks on the functional morphology of the earliest mammals. Mamm. Biol. 104, 739 – 749. doi: doi:10.1007/s42991-024-00437-z
- Pusch L. C., Kammerer C. F., Fernandez V., Fröbisch J. (2023). Cranial anatomy of Nythosaurus larvatus Owen 1876, an Early Triassic cynodont preserving a natural endocast. J. Vertebr. Paleontol. 42, e217441. doi: doi:10.1080/02724634.2023.2174441
- Pusch L. C., Kammerer C. F., Fröbisch J. (2019). Cranial anatomy of the early cynodont Galesaurus planiceps and the origin of mammalian endocranial characters. J. Anat. 234, 592 – 621. doi: doi:10.1111/joa.12958
- Pusch L. C., Kammerer C. F., Fröbisch J. (2021). Cranial anatomy of Bolotridon frerensis, an enigmatic cynodont from the Middle Triassic of South Africa, and its phylogenetic significance. PeerJ 9, e11542. doi: doi:10.7717/peerj.11542
- Pusch L. C., Kammerer C. F., Fröbisch J. (2024). The origin and evolution of Cynodontia (Synapsida, Therapsida): reassessment of the phylogeny and systematics of the earliest members of this clade using 3D-imaging technologies. Anat. Rec. 307, 1634 – 1730. doi: doi:10.1002/ar.25394
- Pusch L. C., Ponstein J., Kammerer C. F., Fröbisch J. (2020). Novel endocranial data on the early therocephalian Lycosuchus vanderrieti underpin high character variability in early theriodont evolution. Front. Ecol. Evol. 7, 464. doi: doi:10.3389/fevo.2019.00464
- Ruben J. (1996). “ Evolution of endothermy in mammals, birds and their ancestors,” in Animals and Temperature: Phenotypic and Evolutionary Adaptation. Eds. Johnston I. A., Bennett A. F. (Cambridge University Press, Cambridge), 347 – 376.
- Ruben J. A., Hillenius W. J., Kemp T. S., Quick D. E. (2012). “ The evolution of mammalian endothermy,” in Forerunners of Mammals: Radiation, Histology, Biology. Ed. Chinsamy-Turan A. (Indiana University Press, Bloomington, IN), 272 – 286.
- Sidor C., Kulik Z. T., Huttenlocker A. K. (2022). A new bauriamorph therocephalian adds a novel component to the Lower Triassic tetrapod assemblage of the Fremouw Formation (Transantarctic Basin) of Antarctica. J. Vertebr. Paleontol. 41, e2081510. doi: doi:10.1080/02724634.2021.2081510
- Sidor C. A., Smith R. M. H. (2004). A new galesaurid (Therapsida: Cynodontia) from the Lower Triassic of South Africa. Palaeontology 47, 535 – 556. doi: doi:10.1111/j.0031-0239.2004.00378.x
- Sigurdsen T. (2006). New features of the snout and orbit of a therocephalian therapsid from South Africa. Acta Palaeontol. Pol. 51, 63 – 75. Available online at: http://app.pan.pl/acta51/app51−063.pdf.
- Sigurdsen T., Huttenlocker A. K., Modesto S. P., Rowe T. B., Damiani R. (2012). Reassessment of the morphology and paleobiology of the therocephalian Tetracynodon darti (Therapsida), and the phylogenetic relationships of Baurioidea. J. Vertebr. Paleontol. 32, 1113 – 1134. doi: doi:10.1080/02724634.2012.688693
- Smith R. M. H. (2020). Biostratigraphy of the Cistecephalus Assemblage Zone (Beaufort Group, Karoo Supergroup), South Africa. S. Afr. J. Geol. 123, 181 – 190. doi: doi:10.25131/sajg.123.0013
- Sun A. (1991). A review of Chinese therocephalian reptiles. Vertebrata PalAsiatica 29, 85.
- Tatarinov L. P. (1965). New late Permian therocephalian. Int. Geol. Rev. 7, 1094 – 1109. doi: doi:10.1080/00206816509474759
- Thomason J. J., Russell A. P. (1986). Mechanical factors in the evolution of the mammalian secondary palate: a theoretical analysis. J. Morphol. 189, 199 – 213. doi: doi:10.1002/jmor.1051890210
- van den Heever J. A. (1994). The cranial anatomy of the early Therocephalia (Amniota: Therapsida). Ann. Univ. Stellenbosch 1994, 1 – 59.
- Viglietti P. A. (2020). Biostratigraphy of the Daptocephalus Assemblage Zone (Beaufort group, Karoo supergroup), South Africa. S. Afr. J. Geol. 123, 191 – 206. doi: doi:10.25131/sajg.123.0014
- Werneburg I., Maier W. (2019). Diverging development of akinetic skulls in cryptodire and pleurodire turtles: an ontogenetic and phylogenetic study. Vertebr. Zool. 69, 113 – 143. doi: doi:10.26049/VZ69-2-2019-01
Citation
Gigliotti, A., et al. (2026). Cranial anatomy of the baurioid therocephalian Choerosaurus dejageri with comments on secondary palate convergence in therocephalians and cynodonts. Frontiers in Mammal Science, Volume 5. https://doi.org/
- Received
- Accepted
- Published
- Volume
- 5 - 2026
Copyright
Copyright © 2026 Gigliotti, Pusch, Kammerer, Benoit and Fröbisch.
License: https://creativecommons.org/licenses/by/4.0/
*Correspondence: Alessandro Gigliotti
Article metrics
- Views
- 0
- Downloads
- 0
- Citations
- 0
- Shares
- 0
