Comparative endocranial anatomy in the crocodylians Leidyosuchus canadensis and Stangerochampsa mccabei from the upper Cretaceous of Alberta, Canada

Abstract

Crocodylians evolved a high variety of rostral morphologies during their evolutionary history, highlighting the strong links between morphological plasticity and environmental and ecological parameters. Two Late Cretaceous alligatoroids, the mesorostrine Leidyosuchus canadensis Lambe, 1907, and the brevirostrine Stangerochampsa mccabei Wu et al., 1996, from Alberta, Canada, preserve a large groove‐shaped recess on the posterior part of the maxilla that has not been documented in other alligatoroids. Despite the potential phylogenetic and paleoecological significance of this neurovascular feature, internal and endocranial structures remain under‐explored among stem alligatoroids. The endocranial morphology, including the paratympanic sinus system of Leidyosuchus canadensis and Stangerochampsa mccabei, was compared to those of extant crocodylians and of the extinct alligatoroid Diplocynodon ratelii based on computed tomography data. The Cretaceous alligatoroids share endocranial features, such as a posteroventral neurovascular projection of the labiolateral canal that connects to the groove‐like recess at the posterior edge of the maxilla and a paratympanic sinus system most similar to those of small‐bodied and young extant crocodylians, suggesting that these pedomorphic features may reflect the ancestral crocodylian condition. Future phylogenetic studies should consider internal and endocranial characters alike to improve our understanding on the relationships among crocodylians.

Leidyosuchus canadensis and Stangerochampsa mccabei share endocranial features such as posterior projection of a neurovascular canal in the maxilla and a paratympanic sinus system most similar to those of small‐bodied and young extant crocodylians, suggesting that these pedomorphic features may reflect the ancestral crocodylian condition. Future phylogenetic studies should consider internal and endocranial characters alike to improve our understanding on the relationships among crocodylians.

1. INTRODUCTION

High levels of disparity in rostral shape exist among extant and extinct Crocodylia (Brochu, 2001; Drumheller & Wilberg, 2020; Godoy, 2020; Sadleir & Makovicky, 2008) and can be described through Ecomorph Shape Categories (ESCs). Three to seven different ESCs have been described in the literature (Brochu, 2001; Busbey, 1995; Drumheller & Wilberg, 2020; McHenry et al., 2006; Morris et al., 2019; Pierce et al., 2008), with different terminology and metrics being used between studies but sharing similarities (Brochu, 2001; Busbey, 1995; Drumheller & Wilberg, 2020; Morris et al., 2019; Pierce et al., 2008). The three main ESCs can be described as brevirostrine (blunt or short snout), longirostrine (slender and elongated snout), and mesorostrine (generalized snout, neither blunt nor slender). Each rostral morphology evolved multiple times in the evolutionary history of crocodylians, and phylogenetic topologies reveal artificial clustering of longirostrine species apart from brevirostrine species (Brochu, 2001; Drumheller & Wilberg, 2020; Sadleir & Makovicky, 2008). Moreover, the three main ESCs have been interpreted as reflecting differences in diet and interaction with the environment (Bontemps et al., 2016; Cott, 1961; Wallace & Leslie, 2008). While rostrum shape and dietary specialization appears to be partially uncorrelated in crocodylians (Balaguera‐Reina et al., 2018; Bontemps et al., 2016; Cott, 1961; Grigg & Kirshner, 2015; Platt et al., 2013; Shirley et al., 2017; Taylor, 1979; Villegas & Schmitter‐Soto, 2008; Waitkuwait, 1986; Wallace & Leslie, 2008; Webb, 1982; Whitaker & Basu, 1982), the rostrum is one of the main body part interacting with the environment through numerous sensory organs (Grigg & Kirshner, 2015). These organs can have an impact on the morphological structure of the bone elements that house them. For instance, integumentary sensory organs (ISO) (Leitch & Catania, 2012; Soares, 2007) play an important role in hunting, parental care, and reproduction, and are hosted by neurovascular canals in the rostrum (Grap et al., 2015; Grap et al., 2020; Grigg & Kirshner, 2015; Leitch & Catania, 2012). Thus, cladistic clustering of species with similar ESCs can be the result of close phylogenetic relationships or convergence due to similar environmental conditions. Testing this assumption can be difficult, especially in fossil taxa for which the ecology and ethology cannot be studied. This is particularly true for species displaying similar cranial characters while their rostrum is classified in different ESCs, one notable example being that of the mesorostrine Leidyosuchus canadensis Lambe (1907) and the brevirostrine Stangerochampsa mccabei Wu et al. (1996) (Figure 1).

FIGURE 1.

Comparison of Leidyosuchus canadensis (TMP1986.221.1) and Stangerochampsa mccabei (TMP1986.61.1). (a) 3D models of Leidyosuchus canadensis (TMP1986.221.1) in dorsal (top) and right lateral (bottom) views, (b) 3D models of Stangerochampsa mccabei (TMP1986.61.1) in dorsal (top) and right lateral (middle) views produced from CT scans. Fo: Closeup on the enlarge foramina described by Wu et al. (1996, 2001) as a “recess for blood vessels and nerves”. (c) Phylogenetic framework. Gray corresponds to bones, dark yellow to jugal canal hosting neurovasculature, light yellow to maxillary canals hosting neurovasculature and light blue to alveoli. The black line on the skull represents the anterior suture of the jugal with the maxillary laterally and the lacrymal medially. Synthetic phylogenetic framework after Brochu (1999), Bona et al. (2018) and Cossette and Brochu (2020). Scalebars represent 5 cm for the skulls and 1 cm for the closeup.

Both taxa are from the Upper Cretaceous formations of Alberta, Canada, and are considered alligatoroids (Brinkman, 1990; Brochu, 1999; Erickson, 1972; Erickson, 1973; Lambe, 1907; Norell et al., 1994; Wu et al., 1996). Leidyosuchus canadensis is from the upper Campanian Dinosaur Park Formation (76.5 to 74.5 Ma; Eberth et al., 2023; Ramezani et al., 2022) and Stangerochampsa mccabei is from the uppermost Campanian to Maastricthian Horseshoe Canyon Formation (73.1 to 68 Ma; Eberth & Kamo, 2020). Both taxa were (re)described and diagnosed by Wu et al. (1996, 2001). Long comprised of six species, the genus Leidyosuchus has been revised by Brochu (1997) and now consists of a monospecific genus, which is usually recovered as the basalmost member of Alligatoroidea close to Diplocynodon (Brochu, 1997; Brochu, 1999; Brochu, 2011; Brochu, 2012; Burns et al., 2013; Cossette & Brochu, 2020; Farke et al., 2014; Hastings et al., 2016; Wu et al., 1996) although a recent study recovered it as a basal crocodylian (Walter et al., 2025). S. mccabei, also a monospecific genus, is usually considered as more derived than L. canadensis, either as a basal Globidonta (Brochu, 1997; Brochu, 1999; Brochu, 2011; Brochu, 2012; Cossette & Brochu, 2018; Hastings et al., 2016), an Alligatorinae (Wu et al., 1996), a sister taxon of Caimaninae (Bona et al., 2018; Farke et al., 2014), or as a Caimaninae (Cossette & Brochu, 2020; Rio & Mannion, 2021). Regardless of its exact phylogenetic position, it is always recovered in Globidonta as a sister taxon to other brevirostrine taxa Albertochampsa and Brachychampsa (Figure 1) (Bona et al., 2018; Brochu, 1997; Brochu, 1999; Brochu, 2011; Brochu, 2012; Burns et al., 2013; Farke et al., 2014; Hastings et al., 2016; Wu et al., 1996). Despite their phylogenetic distance, L. canadensis and S. mccabei share a unique feature: a perforation on the lateral surface of the maxilla, described as “a series of linearly arranged nutrient and nerve foramina that are enlarged into a groove‐like recess that extends posteriorly to a point just beyond the suture with the jugal” (Wu et al., 2001: 2). This feature has not positively been identified in the sister taxon of S. mccabei, Brachychampsa montana, and in Borealosuchus sternbergii, a taxon close to L. canadensis (Gilmore, 1910; Norell et al., 1994), and is neither described nor illustrated in other closely related taxa, like Albertochampsa langstoni, Bo. formidabilis, Bo. wilsoni or Bo. acutidentatus (Erickson, 1972; Erickson, 1973; Mook, 1959; Sternberg, 1932). Despite its apparent taxonomic utility, this anatomical character has never been included in phylogenetic studies of alligatoroids. Previous cladistic analyses were conducted primarily based on observable external characters and ignored internal or endocranial anatomy, which is more difficult to observe. As cranial morphology is often the result of competing selective forces driven by ecology, environment, and phylogeny (Godoy, 2020; Hermanson et al., 2022; Herrel et al., 2013; Holmes et al., 2016; Kraatz et al., 2015; Salas‐Gismondi et al., 2016; Werner & Seifan, 2006; Wilberg et al., 2019), such practice masks the importance of internal characters and their links with external characters, known to carry taxonomic and/or phylogenetic information (Kuzmin, 2022; Kuzmin et al., 2021; Kuzmin et al., 2024; Perrichon, 2024; Perrichon et al., 2023), which could bias our understanding of the phylogenetic relationships of early alligatoroids.

Here, we describe the osteological correlates of the trigeminal nerve system (i.e., the bony structures hosting the trigeminal system that correlate with its morphology) associated with the groove‐shaped recess in S. mccabei and L. canadensis using CT‐scan imagery and 3D model reconstructions. We subsequently compare this recess to that of extant crocodylian species and of the extinct crocodylian Diplocynodon ratelii, from the Neogene of Europe, to assess the anatomical similarities and phylogenetic proximity of S. mccabei and L. canadensis. Finally, we used CT‐scan imagery to describe the paratympanic sinuses and neuroanatomy of the latter two taxa. Assessment of the phylogenetic relationships of basal alligatoroids is beyond the scope of this study and will be the subject of a future work.

2. MATERIALS AND METHODS

2.1. Taxa and specimens studied

We studied one well‐preserved specimen of Leidyosuchus canadensis, TMP1986.221.1, and the well‐preserved holotype of Stangerochampsa mccabei TMP1986.61.1 (Table 1). Both were CT‐scanned at Canada Diagnostics, Calgary, Alberta, Canada, on a Light Speed Ultra medical scanner, following the protocol and parameters recommended by Ridgely and Witmer (2006). For comparative purposes, the endocranial anatomy of TMP1986.221.1 and TMP1986.61.1 was compared to that of adult specimens of the basal alligatoroid Diplocynodon ratelii (MNHL‐LA86) and the extant crocodylians Alligator mississippiensis (OUVC_9761_M39878‐71998), Alligator sinensis (NHMW‐Zoo‐HS‐37966), and Crocodylus niloticus (MHNL_50001399) (Table 1). These taxa were selected based either on their phylogenetic proximity (Diplocynodon), as extant representatives of alligatoroid (A. mississippiensis, A. sinensis), and as a representative of the outgroup (Crocodylus niloticus), and the specimens were chosen based on the availability and completeness of the tomography scans. They were CT‐scanned, either with a medical scanner or a microCT scanner, at different venues (Table 1). For clarity, the specimens studied in this work are not mentioned by their curation numbers but referred to by the name of the corresponding species.

TABLE 1.

Information about the specimen used.

Species	Identification number	Type of CT scan	CT scan location	Voxel size
Leidyosuchus canadensis	TMP1986.221.1	Medical scanner	Canada Diagnostics, Calgary	538 × 497 × 657 μm3
Stangerochampsa mccabei	TMP1986.61.1	Medical scanner	Canada Diagnostics, Calgary	428 × 499 × 673 μm3
Diplocynodon ratelii	MHNL‐LA86	Microtomograph	Phoenix, Lyon	86 μm3
Alligator mississippiensis	OUVC_9761_M39878‐71998	Medical scanner	Ohio Health O'Bleness Hospital	500 × 500 × 1000 μm3
Alligator sinensis	NHMW‐Zoo‐HS‐37966	Microtomograph	Naturhistorisches Museum Wien	84 μm3
Crocodilus niloticus	MHNL_50001399	Microtomograph	DTHE, INSA Lyon	102 μm3

All the CT scans were converted into raw files using ImageJ (RRID:SCR_003070), without any compression or size reduction (Figure 2). The files were imported subsequently into Amira 3D (RRID:SCR_007353) to virtually reconstruct the sinuses, cranial nerves, and neurovascular rostral canals, and then exported in PLY file format (Figures 1, 2, 3, 4, 5, 6, 7, 8, 9, 10). For the neurovascular canals, the profusion of small canals in A. sinensis have led us to stop segmentation before reconstructing the whole system in order to preserve the readability of the model. All reconstructed structures in Leidyosuchus canadensis and Stangerochampsa maccabei are showed in Figure 3.

FIGURE 2.

CT‐scan slices of Leidyosuchus canadensis (TMP1986.221.1) (a, b) and Stangerochampsa mccabei (TMP1986.61.1) (c, d) in transversal view. The slices are located at the posterior edge of the maxillary and in a medial position in the rostrum. Scalebars represent 5 cm.

FIGURE 3.

Neurovascular structures of Leidyosuchus canadensis (TMP1986.221.1) and Stangerochampsa mccabei (TMP1986.61.1). Dorsal view of L. canadensis (a) and S. mccabei (b). Dorsolateral view of L. canadensis (c) and S. mccabei (d). C, cerebral carotid artery; cAs, superior alveolar canal; cCU, canal for the passage of the cutaneous branch of the trigeminal nerve; cJU, canal for the passage of the jugal branch of the maxillary division of the trigeminal nerve; cL, labial canal; cLac, lacrymal canal; cPm, canal for the passage of the premaxillary branch of the maxillary division of the trigeminal nerve; cPNV, posterior neurovascular canal; cVL, ventrolingual canal; En, endocast; gT, trigeminal ganglion; intS, intertympanic sinuses; ptS, pharyngotympanic sinuses. Light blue represent tooth alveoli. Numbers on the laterodorsal views represent the position of maxillary teeth. Scalebars represent 5 cm.

FIGURE 4.

Comparison of the braincases of studied crocodylians. Leidyosuchus canadensis (TMP1986.221.1) in left‐lateral (a), dorsal (c) and ventral (e) views. Stangerochampsa mccabei (TMP1986.61.1) and in left‐lateral (b), dorsal (d) and ventral (f) views. Alligator mississippiensis (OUVC_9761_M39878‐71998) in left‐lateral (g) and ventral (i) views (c). Alligator sinensis (NHMW‐Zoo‐HS‐37966) in left‐lateral (h) and ventral (j) views (d). Crocodylus niloticus (MHNL_50001399) in left‐lateral (k) and ventral (l) views (e). C, cerebral carotids artery; cbl, cerebellum region; cbr, cerebrum; CN IX‐XI, cranial nerves IX to XI (glossopharyngeal and accessory nerves); CN V, cranial nerve V (trigeminal nerve); CN VII, cranial nerve VII (facial nerve); CN XIIa, cranial nerve XIIa (anterior hypoglossal nerve); CN XIIb, cranial nerve XIIb (posterior hypoglossal nerve); mid, midbrain; mo, medulla oblongata; np, nerve projection of the trigeminal ganglion; ob, endocast of the olfactory bulb; ocs, otoccipital venous sinous; ot, olfactory tracts; pf, pituitary fossa; pons, pons region; vls, ventral longitudinal venous sinus. The asterisk represent the concave shape linked to the otic capsule. Scalebars represent 2 cm.

FIGURE 5.

Virtual braincase reconstruction of Diplocynodon ratelii (MHNL‐LA86). cbl: cerebellum region, cbr: cerebrum, CN II: cranial nerve II (optic nerve), mid: midbrain, mo: medulla oblongata, ocs: otoccipital venous sinous, ot: olfactory tracts, pf: pituitary fossa, pons: pons region, vls: ventral longitudinal venous sinus. The asterisk represents the concave shape linked to the otic capsule. Scalebars represent 1 cm.

FIGURE 6.

Comparison of the neurovascular canals of the jugal branch of the maxillary division of the trigeminal system of L. canadensis (TMP1986.221.1) in dorsal (a), and lateral (c) views, S. mccabei (TMP1986.61.1) in dorsal (b), and lateral (d) views, Alligator mississippiensis (OUVC_9761_M39878‐71998) in dorsal (e), dorsal (c) and posterior (e) views, Alligator sinensis (NHMW‐Zoo‐HS‐37966) in lateral (b), dorsal (d) and posterior (f) views, Crocodylus niloticus (MHNL_50001399) in lateral (g), dorsal (i) and posterior (k) views and Diplocynodon ratelii (MHNL‐LA86) in lateral (h), dorsal (j) and posterior (l) views. AntcMJ, anterior branch of the main jugal canal; cCU, canals for the passage of cutaneous branch of the maxillary division of the trigeminal nerve; cJANV, jugal anterior neurovascular canal; cJUf, foramen linked to the canal for the passage of the jugal nerves; PostcMJ, posterior branch of the main jugal canal. Skulls are illustrated at the same scale. Scalebars represent 2 cm.

FIGURE 7.

Neurovascular canals 3D reconstruction of Leidyosuchus canadensis (TMP1986.221.1) (a) and Stangerochampsa mccabei (TMP1986.61.1) (b). cAs, superior alveolar canal; cCU, canals for the passage of cutaneous branch of the maxillary division of the trigeminal nerve; cJANV, jugal anterior neurovascular canal; cJU, canal for the passage of the jugal branch of the maxillary division of the trigeminal nerve; cL, labial canal; cPm, canal for the passage of the premaxillary branch of the maxillary division of the trigeminal nerve; cPNV, posterior neurovascular canal; cVL, ventrolingual canal; infLacR, infra lacrimal recess; lVL, ventrolingual projection of the cAs, light blue represent tooth alveoli. Skulls are illustrated at the same scale. Scalebars for the global neurovascular canals represent 40 cm. Scalebars for the closeup represent 2 cm.

FIGURE 8.

Comparison of paratympanic sinus system of Leidyosuchus canadensis (TMP1986.221.1) in lateral (a), posterolateral (c), dorsal (e) and posterior views (g) and Stangerochampsa mccabei (TMP1986.61.1) in lateral (b), posterolateral (d), dorsal (f) and posterior views (h) in. Skulls are illustrated at the same scale. AntMC, anterior branch of the median pharyngeal canal; BoR, basioccipital recess; BsR, basisphenoid recess; C, cerebral carotid artey; cbr, cerebrum; IntR, intertympanic recess; LsR, laterosphenoid recess; mo, medulla oblongata; MPhC, median pharyngeal canal; MPhf, median pharyngeal foramen; ob, olfactory bulb; ot, olfactory tracts; OtoR, otoccipital recess; pf, pituitary fossa; phar int., pharyngeal intersection; PhR, pharyngotympanic recess (middle ear cavity); PhT, pharyngeal tubes; PostMC, posterior branch of the median pharyngeal canal; ppc, prootic‐parietal canal; PR, parietal recess; ProR, prootic recess; QR, quadrate recess; REt, recessus epitubaricus; RhR, rhomboidal recess; ScR, subcaritod recess; sf, subtympanic foramen; Siph, siphonium; vls, ventral longitudinal venous sinus. Black arrowhead points the canals connecting the rhomboidal recess to the middle ear cavity. Dotted black square represent the small subspherical cavity dorsal to the basioccipital recess. Scalebars represent 2 cm.

FIGURE 9.

Paratympanic sinus system 3D reconstruction of Alligator mississippiensis (OUVC_9761_M39878‐71998) in lateral (a), dorsal (c) and posterior (e) views, Alligator sinensis (NHMW‐Zoo‐HS‐37966) in lateral (b), dorsal (d) and posterior (f) views, Crocodylus niloticus (MHNL_50001399) in lateral (g), dorsal (i) and posterior (k) views and Diplocynodon ratelii (MHNL‐LA86) in lateral (h), dorsal (j) and posterior (l) views. Skulls are illustrated at the same scale. BoR, basioccipital recess; BsR, basisphenoid recess; IntR, intertympanic recess; LsR, laterosphenoid recess; MPhC, median pharyngeal canal; MPhf, median pharyngeal foramen; OtoR, otoccipital recess; pf, pituitary fossa; PhR, pharyngotympanic recess; PhT, pharyngeal tubes; PostMC, posterior branch of the median pharyngeal canal; ppc, prootic‐parietal canal; PR, parietal recess; ProR, prootic recess; PtPR, pterygoid recess; QR, quadrate recess; REt, recessus epitubaricus; RhR, rhomboidal recess; ScR, subcaritod recess; sf, subtympanic foramen; Siph, siphonium. Scalebars represent 2 cm.

FIGURE 10.

Comparison of maxillary neurovascular system in the crocodylians studied within a simplified phylogenetic framework. cAs, superior alveolar canal; cCU, canal for the passage of the cutaneous branch of the maxillary division of the trigeminal nerve; cIS, canal of the integumentary system; cL, labial canal; cPNV, posterior neurovascular canal; lp, lingual projection. The black arrowhead points to the location of the posterior neurovascular canal. The dotted line in the closeup marks the posteriormost extent of the neurovascular canals. Dotted circles shows the cL. Scalebars for skulls represent 5 cm, scalebars for neurovascular closeup represents 2 cm.

2.2. Preservation of the specimens

The skull of Leidyosuchus canadensis is slightly dorsoventrally compressed at the level of the orbit and exhibits numerous cracks and breaks along the skull, with sediment present in the braincase and in the canals of the jugal bone. In the rostrum, cracks are blurring the outline of internal cavities in the jugal and in the maxilla between the 3rd and 5th maxillary alveoli. The ventral part of the braincase is crushed by the upward thrust of the basisphenoid, but the anterior part of the pituitary fossa is intact (Figure 4a,e). Ventral cracks have created spaces infilled by sediment in the basisphenoid and basioccipital, which may blur or fill internal cavities or sutures in some areas.

The braincase of Stangerochampsa mccabei is dorsoventrally crushed due to taphonomic deformation, which pushed the pterygoid–basisphenoid–basioccipital complex inside the brain endocast approximately 1 cm dorsally. Furthermore, numerous cracks and breaks are present throughout the skull, resulting in sediment infilling the braincase and the jugal, and bone fragments being present in the maxillary and paratympanic sinuses. In the rostrum, fractures blur the identification of internal cavities in the jugal. The right maxilla is damaged internally between the 3rd and 5th teeth, and so is the anterior part on the left maxilla. In the braincase, the ventral edge of the brain cavity is crushed at the level of the cerebrum and pituitary fossa, creating a deep depression in the cerebrum and flattening the midbrain. The right side of the braincase is filled with opaque, high‐density sediment that prevents an accurate reconstruction of the paratympanic sinuses and nerves in this area. Moreover, both prootic are crushed, preventing an accurate reconstruction of the dorsal part of the recessus epitubaricus, the prootic recess, and the inner ear. The left side of the braincase is better preserved except for a large crack on the ventral part of the middle ear.

2.3. Measurements

Measurements of the endocast and neurovascular canals were realized with the Meshlab measurement tool, following the method described by Ristevski (2022) (Table 2). These measurements do not represents the original dimensions of the neural structures, but of the chamber that housed them (Ali et al., 2008; Jirak & Janacek, 2017).

TABLE 2.

Anatomical abbreviations.

A	Alveoli
AntcMJ	Anterior branch of the main jugal canal
BoPR	Basioccipital pneumatic recess
BsPR	Basisphenoid pneumatic recess
C	Cerebral carotid artery
cA	Canal for the passage of alveolar nerves
cAs	Superior alveolar canal (canal hosting the superior alveolar nerve)
cbl	Cerebellum
cbr	Cerebrum
cCU	Canal for the passage of cutaneous branches of trigeminal nerve
cIS	Canal of the integumentary system
cJANV	Jugal anterior neurovascular canal
cJU	Canal for the passage of the jugal branch of the maxillary division of the trigeminal nerve
cL	Labial canal
cMJ	Jugal main canal
CN II	Cranial nerve II, optic nerve
CN V	Cranial nerve V, trigeminal nerve
CN V1	Ophtalmic division of the trigeminal nerve
CN V2	Maxillary division of the trigeminal nerve
CN V3	Mandibular division of the trigeminal nerve
CN VII	Cranial nerve VII, facial nerve
CN V IX‐XI	Cranial nerves IX and XI, common canal for glossopharyngeal, vagus and spinal accessory nerves
CN XIIa	Cranial nerve xiia, anterior hypoglossal nerve
CN XIIb	Cranial nerve xiib, posterior hypoglossal nerve
cPNV	Posterior neurovascular canal
cVL	Ventrolingual canal
En	Endocast
IntPR	Intertympanic pneumatic recess
intS	Intertympanic sinuses
L	Labyrinth
Lc	Lacrymal canal
LsRP	Laterosphenoid pneumatic recess
lp	Lingual projection
mid	Midbrain
mo	Medulla oblongata
MPhC	Median pharyngeal canal
MPhS	Median pharyngeal sinus
ob	Olfactory bulb
ocs	Otoccipital venous sinous
ot	Olfactory tracts
OtoPR	Otoccipital pneumatic recess
pf	Pitituary fossa
PhT	Pharyngeal tubes
Pmn	Premaxillary nerves
pons	Pons region
‍PostcMJ	Posterior branch of the main jugal canal
PostMC	Posterior median canal
ppc	Prootic‐parietal canal
PPR	Parietal pneumatic recess
PREt	Pneumatic Recessus epitubaricus
ProPR	Prootic pneumatic recess
ptS	Pharyngotympanic sinuses
QPR	Quadrate pneumatic recess
RhPR	Rhomboidal pneumatic recess
ScPR	Subcaritod pneumatic recess
sf	Subtympanic foramen
Siph	Siphonium
nT	Trigeminal nerve
vls	Ventral longitudinal venous sinus

2.4. Institutional abbreviations

CCEV: Centre de Conservation et d'Étude des Confluences, Lyon, France.

MHNL: Musée d'Histoire Naturelle de Lyon, Lyon, France.

NHMW: Natural History Museum of Wien, Wien, Austria.

OUVC: Ohio University Vertebrate Collection, Athens, Ohio, United States.

ROM: Royal Ontario Museum, Toronto, Ontario, Canada.

TMP: Royal Tyrrell Museum of Palaeontology, Drumheller, Alberta, Canada.

2.5. Anatomical abbreviations and terminology

All the abbreviations used in the text are summarized in Table 2. The abbreviations for the brain endocast and the paratympanic sinuses used in this study follow Kuzmin et al. (2021) with clarifications from Perrichon et al. (2024). For rostral osteological correlates of the nervous system, we follow the terminology and abbreviations of Lessner and Holliday (2020) and supplement it with new terminology when necessary. Contrary to Lessner and Holliday (2020) who described the nervous system itself, we consider its osteological correlates, which are the neurovascular canals (Porter et al., 2016; Sedlmayr, 2002). Thus, some of their abbreviations were adapted according to the osteological nature of the structures we describe (Table 2).

3. RESULTS

3.1. Endocranial structures of Leidyosuchus canadensis

3.1.1. Brain endocast

The brain endocast of L. canadensis is sigmoidal in lateral view. Its dorsal surface is flat from the olfactory bulb to the posterior end of the cerebrum. The ridge of the dorsal longitudinal venous sinus is not visible on the midbrain due to low resolution of the tomography. The posterior part of the brain endocast is slightly bent ventrally (Figure 4a). The dorsal surface of the L. canadensis brain endocast is similar to that of A. mississippiensis and differs from the flat brain endocast of C. niloticus (Figure 4k,l) and D. ratelii (Figure 5) and from the stocky brain endocast of A. sinensis (Figure 4h,j).

The olfactory apparatus (tract and bulb) is delimited by the prefrontal pillars anteriorly, the frontal dorsally and the posterior limit of the olfactory tract is considered to extend to the most anterior sector of the laterosphenoid. The olfactory apparatus represents ~50% of the total length of the brain endocast (Table 2), as in A. mississippiensis and C. niloticus. Because the olfactory tract is not delimited lateroventrally by bone, the reconstruction of the lateral and ventral limits are arbitrary. The endocast of the olfactory bulb is rather large (~67% of the cerebral hemispheres width) and shows a slight longitudinal depression in its dorsal part, which corresponds to the olfactory sulcus. The endocast of the olfactory tract is 4.8 cm long, which represents 35.5% of the total length of the brain endocast. The dorsal part of the olfactory tract has a slight dorsal tilt (7°) and a constant width until it reaches the cerebral hemispheres (Figure 4a). This tilt is less pronounced than in the two Alligator species studied (~17°) and D. ratelii (~20°) but not null as in C. niloticus.

The cerebral hemispheres consist of wide bulges, widening from the olfactory tract to a maximum just anterior to an abrupt narrowing, giving the cerebrum a kite shape in dorsal view. The dorsal tilt observed in the olfactory apparatus extends along the cerebrum (Figure 4a). The pituitary fossa is located in the posteroventral part of the cerebrum, supported by the basisphenoid and oriented ventrally (Figure 4a,e). Its connection to the cerebral carotid arteries is severed due to the partial breakage of the basisphenoid and infislling of sediments. The palatine branch of the facial nerve CN VII can be seen ventrally to the pituitary fossa and the cerebral carotid arteries (Figure 4a,c,e).

The midbrain (or mesencephalon) of L. canadensis is the shortest part of the brain endocast, as in other crocodylians. In L. canadensis, the midbrain represents around 5% of the total length of the brain endocast, which is shorter than in other species studied herein. However, its width is ~67% of the cerebrum width, similar to the width of the olfactory bulb endocast. The junction between the forebrain and the midbrain is not clearly delimited in L. canadensis, as in A. sinensis (Figure 4a,c,e,h,j). The midbrain hosts the optic lobes and otoccipital venous sinus, but these structures do not protrude on the surface of the brain endocast, such as in A. sinensis (Figure 4a,c,h). However, contrary to A. sinensis, the dorsal longitudinal venous sinus does not protrude from the midbrain in Leidyosuchus.

The hindbrain is divided into the metencephalon anteriorly and the myelencephalon posteriorly. The metencephalon hosts the cerebellum dorsally and the pons ventrally (Figure 4a,c,e). The bulge of the cerebellum protrudes only slightly laterally in L. canadensis and has a straight dorsal profile, descending toward the medulla oblongata as in other crocodylians. The otoccipital venous sinus (~0.5 cm wide) is marked by the constriction linked to the otic capsule (Figure 4a,c,e). The trigeminal ganglion (ganglion of the cranial nerve V, CN V) is located in a fossa lateral to the brain at the level of the transition between the cerebellum and the cerebrum (Figure 4a,c,e). It is housed outside the primary bony lateral wall of the brain endocast, formed by the postorbital and the laterosphenoid. The trigeminal ganglion is of constant width and shows two nerve projections: one oriented laterodorsally at the posterior end of the ganglion, most probably the tympanic branch of the trigeminal nerve, and the other oriented anteriorly, visible only on the right side of the specimen, likely the ophthalmic branch of the trigeminal nerve (Figure 4a,c,e).

On the ventral side of the hindbrain, a slight swelling runs from the pons, posterior to the trigeminal ganglion, and all the way to the medulla oblongata; it corresponds to the ventral longitudinal venous sinus (vls) (Figure 4a,e). This bulge is artificially exaggerated by an internal breakage on the right side of the brain endocast, creating a larger cavity. In lateral view, the metotic foramen can be distinguished, as well as the posterior part of the sympathetic nerve.

The myelencephalon contains the medulla oblongata, in which both the occipital venous sinus and the ventral longitudinal venous sinus end. It represents ~9% of the total length of the brain endocast in L. canadensis and is ellipsoidal in cross section, with a dorsal flattening similar to the condition observed in other crocodylians studied herein. The ventral part of the medulla oblongata becomes inclined ventrally, with an anterior height of 12 cm and a posterior height of 15 cm. Branches of the hypoglossal nerve XIIa and XIIb exit the medulla oblongata in its anterior part (Figure 4a,c,e).

Lessner et al. (2024) segmented the brain endocast of another specimen of Leidyosuchus canadensis (ROM 1903). The brain endocast of ROM 1903 appears to have the same general shape as TMP1986.221.1, with some small differences probably due to specimen preservation, such as the presence of the optic nerve (CN II) ventral to the cerebral hemisphere, a higher and more ventroposteriorly located pituitary fossa, and a more pronounced protrusion of the cerebellum.

3.1.2. Rostrum neurovasculature

Neurovascular canals for the passage of the branches of the maxillary division of thetrigeminal nerve (CN V₂) are present in the jugal (canal for the passage of the jugal nerves (cJU), Figures 3a and 6). The precise structure of the cJU is subject to interpretation due to the numerous cracks on both sides of Leidyosuchus canadensis. The description here is mostly based on the left side of L. canadensis, which is the less damaged. The CN V₂ enters the jugal posterior to the postorbital bar (Figure 6a,c). The cJU is composed of a main canal, the jugal main canal (cMJ) extending posteriorly and anteriorly and of numerous smaller canals ramifying from the cMJ (Figure 6a,c). The anterior part of the cMJ is 3.5 cm long, 0.5 cm high, and up to 0.3 cm large, and the posterior part is 2 cm long, 0.3 cm high, and 0.2 cm large. The anterior part of the cMJ is straight over 1.6 cm and curved anteroventrally ending in a jugal foramen (Figure 6a,c). Numerous neurovascular canals for the passage of cutaneous branch of the CN V₂ (cCU) ramified from the anterior part of the cMJ (Figure 6a,c). The anteriormost projections of cCU reach the level of the posteriormost maxillary teeth. Unfortunately, it is impossible to determine whether there is continuity between the cJU and the canal for the passage of the maxillary branch of the CN V₂ due to insufficient CT scan resolution. The posterior part of the cMJ ramify in small canals that do not run posteriorly to the posteriormost part of the cMJ (Figure 6a,c).

The canals for the passage of the maxillary branch of the CN V₂ can be divided into three parts: the superior alveolar canal (cAs), the labial canal (cL), and the ventrolingual canal (cVL) (Figures 3a,c and 7a). We do not focus on the link between the paranasal cavities and the canals for the passage of the maxillary branch of the CN V₂ because the preservation of theses sinuses is incomplete, and therefore, no link with the bony canal for the passage of nerves and blood vessels is visible in the specimen studied. The cAs extends dorsally from the 14th maxillary tooth to the maxilla/premaxilla suture and continues into the premaxilla. The medial part of the cAs is not visible due to fractures in the bone. The anterior part of the cAs in the maxilla is enlarged between the 3rd and 1st maxillary tooth. The cAs is straight posteriorly and curves around the 6th maxillary tooth. The cAs forms a bridge without branches in the notch between the first maxillary tooth and the last premaxillary tooth (Figures 3a,c and 7a). This bridge is slightly enlarged at the level of the maxillary/premaxillary suture and strongly enlarged against the 4th premaxillary tooth. The anterior part of the bridge projects a thin canal dorsally, parallel to the notch bridge, and ventrally in the alveolar region (Figures 3a,c and 7a). The cAs continues anteriorly along the dental alveoli to the 2nd premaxillary tooth, and then project canals cutaneously and in the anteriormost alveoli (Figures 3a,c and 7a).

The labial canal (cL) is large and flat, with a bean‐shaped cross section (Figures 3a,c and 7a). It extends connections through dental alveoli and cCU. It runs from the 17th to the 2nd maxillary tooth where it meets the cAs dorsally. The cL projects posteriorly a large, 2.6‐cm‐long branch, called the posterior neurovascular canal (cPNV), to the penultimate maxillary tooth, through the large foramen previously described as a “recess for blood‐vessels and nerves” (Wu et al., 2001: 2), which is part of the diagnosis of Leidyosuchus canadensis (Figure 1a,c). The cPNV consists of the most ventroposterior neurovascular labial projection of the cL. It runs posterior to both the dorsalmost part of the cL and the posteriormost part of the cAs, lacks cCU ramifications and ends in a large and long foramen at the posterior edge of the maxilla (Figures 3 and 7). The cAs and the cL are connected at the level of the 8th or 7th maxillary teeth posteriorly and anteriorly at the level of the 2nd maxillary tooth.

The ventrolingual canal (cVL) originates dorsally from the cAs at the level of the 11th maxillary tooth and extends anteroventrally to the 8th maxillary tooth, and ramifies at the level of the 9th maxillary tooth (Figures 3a,c and 7a). This lingual part of the maxillary neurovascular canals is potentially better developed, but a tomographic scan with higher resolution would be needed to assess the presence of thinner projections.

Ventral to the lacrimal canal, a small recess is present that could house a segment of the CN V₂, as seen in Alligator spp. (see Lessner & Holliday, 2020) (Figures 3c and 7a). No canal enters the maxilla anterior to this lacrimal recess. Although this absence could be due to poor CT‐scans resolution, we can suppose these putative nervous projections have no connections to the maxillary neurovascular canals. This recess recalls a structure interpreted as an olfactory expansion of the internal surface of the prefrontals and lacrimals in the works of Cowgill et al. (2023) and of Burke et al. (2025) (Figure 2; Figure S1). No similar well‐defined recess is present in D. ratelii and C. niloticus (Figure S1).

3.1.3. Paratympanic sinus system

Paratympanic sinuses are epithelial extensions filled with air (diverticulum) generating a cavity on the bone (recess) as described by (Kuzmin et al. (2021). The paratympanic sinus system, as described by Kuzmin et al. (2021) and refined by Perrichon et al., 2024, 2023), is composed of two ontogenetically different systems: the median pharyngeal system and the pharyngotympanic system. The median pharyngeal system is composed of the median pharyngeal canal and foramen, the basiphenoid recess, the precarotid recess, the postcarotid recess, the subcarotid recess, the rostral recess, and the basioccipital recess. The pharyngotympanic system is composed of the pharyngotympanic recess and the intertympanic system. This subdivision into pharyngotympanic and intertympanic system are anatomical and do not reflect differences in the sinuses that created these recesses. The intertympanic system is composed of the intertympanic recess, the posteromedial pre‐parietal recesses and the anterolateral pre‐parietal recesses.

All parts of the paratympanic sinus system of the Leidyosuchus canadensis specimen studied are well‐preserved, except for both connections of the recessus epitubaricus (equivalent to the lateral anterior branch of the pharyngotympanic sinus system) with the middle ear cavity due to taphonomic damage to the anteroventral part of the prootic.

L. canadensis is characterized by a dorsoventrally short paratympanic sinus system, thick ventral canals, a dorsoventrally large middle ear cavity, a well‐developed intertympanic system, and a lack of pneumatization of the laterosphenoid (Figure 8a,c,e,g).

The median pharyngeal system of L. canadensis, composed of the median pharyngeal canal and the basisphenoid recesses, shows thick canals. The median pharyngeal foramen (situated at the suture between the basisphenoid and the basioccipital) is large, occupying one‐third of the basicranial width. Dorsally, this aperture narrows to become the median pharyngeal canal, which is ellipsoidal in cross section and retains constant dimensions throughout its length (Figure 8a,c,g). The median pharyngeal canal represents ~35% of the total length of the paratympanic sinus system, which is shorter than in D. ratelii and most extant crocodylian species studied but longer than in extant Caimaninae (Perrichon, 2024). The pharyngeal intersection (the angle formed by the anterior median pharyngeal canal and the posterior median pharyngeal canal) is located in the dorsoventral center of the basisphenoid, which differs from its ventral position in C. niloticus and its dorsal position in Alligator spp. and D. ratelii. The pharyngeal intersection occupies half of the median pharyngeal canal length (Figures 8a and 9a,b,h).

The anterior branch of the median pharyngeal canal is straight, oriented dorsally, and of the same width as the median pharyngeal canal. Dorsally, the left and right branches of the basisphenoid recess are short, flat, straight, and oriented laterally. The basisphenoid recess is positioned slightly dorsal to the basioccipital recess. Laterally, the basisphenoid recess connects to the recessus epitubaricus (lateral canals directed lateroposteriorly on each side of the braincase) (Figure 8a,e). There are no additional recesses (precarotid, subcarotid, postcarotid, or rostral) connected to the basisphenoid recess.

The recessus epitubaricus is dorsoventrally elongate and anteroposteriorly narrow (i.e., leaf‐shaped), and does not enter the pterygoid ventrally (Figure 8a,e). The overall shape of the recessus epitubaricus in L. canadensis is similar to that of Alligator spp. although thinner and less elongate, and similar to that of D. ratelii but with a less pronounced leaf shape (Figures 8a,e and 9a–j). It differs from that of C. niloticus, which is much thinner, more rectangular, and vertically oriented (Figure 9g,i,k).

The pharyngotympanic sinus system comprises the basioccipital recess, the rhomboidal recess, the lateral pharyngotympanic tubes, the middle ear cavity, and the quadrate recesses. L. canadensis does not possess a prootic facial recess, contrary to Alligator spp. and D. ratelii but similar to C. niloticus (Figure 9).

The posterior branch of the median pharyngeal canal (ventral part of the basioccipital recess) is particularly large in L. canadensis, being confluent with the median pharyngeal canal for half of its length and forming a high pharyngeal intersection (Figure 8a,c). The posterior median pharyngeal canal is oriented posteriorly and forms a large dorsoventral ellipsoidal bulge that pneumatizes almost all of the main body of the basioccipital. This bulge is connected dorsally to two small subspherical cavities, which are the paired basioccipital recesses. The latter show a flat roof and are directed laterally, connecting to the rhomboidal recess on each side of the braincase (Figure 8a,c,g).

The lateral pharyngotympanic tubes are mediolaterally wide and anteroposteriorly elongate. They exit the basicranium through the paired lateral pharyngeal foramina, which are dorsoventrally wide foramina positioned dorsal to the median pharyngeal foramen. These tubes are oriented dorsally and curve posteriorly in the dorsalmost part of the basioccipital upon entering the rhomboidal recess (Figure 8c,g).

The rhomboidal recess is large and round. It is connected to the middle ear dorsally via two canals of different sizes. The larger canal runs along the ventral margin of the quadrate before entering the middle ear lateroventral to the inner ear. The second, slightly smaller canal goes around the cerebral carotid arteries to join the middle ear cavity in its posteroventral part, which is severed on the right side of the braincase due to taphonomic damage (Figure 8c,g).

The middle ear cavity of L. canadensis is anteroposteriorly narrow but dorsoventrally and laterally wide. A peculiar feature that distinguishes it from those of extant taxa is its subdivision into two large, confluent ellipsoidal cavities oriented mediolaterally (Figure 8c,g), which contrasts from the dorsoventral orientation observed in D. ratelii and in the extant crocodylians studied herein (Figure 9). The ventral bulge of the middle ear cavity creates a wide cavity in the quadrate along its suture with the otoccipital; it is connected to the rhomboidal recess ventrally and to the otoccipital recess posterodorsally through a wide ostium. The dorsal bulge of the middle ear cavity holds the columella and directly connects the tympanum to the inner ear. This bulge is connected laterodorsally to the intertympanic recess (Figure 8c,g).

Laterally, the quadrate recess is visible on the right side of the skull and forms a bulbous cavity connected to the siphonium (Figure 8a). The siphonium is thin and exits the skulls through the foramen aereum, which is situated on the dorsal surface of the quadrate, between the two hemicondyles as in Alligator spp.

The infundibular recess and the subtympanic foramen are not visible on this specimen.

The intertympanic system occupies the dorsal part of the braincase. In L. canadensis, it is anteroposteriorly short and the intertympanic system is not as stocky as in A. mississippiensis nor as elongated as in C. niloticus (Figures 8a,c,e,g and 9a–i,k). The intertympanic recess, located in the supraoccipital, links the two middle ear cavities together dorsal to the braincase. In L. canadensis, the recess has an ellipsoidal shape in longitudinal view and a more or less constant width with no anteroposterior or dorsoventral compression visible in its medial part (Figure 8a,e).

The lateral parts of the intertympanic recess are largely confluent with the paired otoccipital recesses that inflate the otoccipital (Figure 8e). The otoccipital recess is, however, still largely inflated ventrally and posteriorly, forming a triangular shape in transverse view, with a laterodorsal protrusion directed toward the temporal canal (Figure 8c,e,g). The prominent posterior development of the otoccipital recess is also seen in Alligator spp. and D. ratelii. The laterodorsal protrusion is absent in extant crocodylian species and D. ratelii (Figure 9).

Finally, the parietal bone of L. canadensis is pneumatized. However, the paired posteromedial pre‐parietal recesses are very small, with only the right recess reaching the parietal‐supraoccipital suture, contrary to Alligator spp., D. ratelii and C. niloticus in which they are generally well‐developed (Figures 8a and 9). The parietal recess is dorsoventrally high, ellispoidal in longitudinal section and connected to the intertympanic recess through paired anterolateral circular ostia (Figure 8a). The parietal recess is less developed anteriorly than in D. ratelii (Figure 9d). In addition, L. canadensis possesses a prootic‐parietal canal, which is an additional connection between the anterior part of the parietal recess and the prootic part of the intertympanic recess (Figure 8a). This prootic‐parietal connection is present in all extant Alligatoridae but absent in Longirostres.

3.2. Endocranial structures of Stangerochampsa mccabei

3.2.1. Brain endocast

The brain endocast of S. mccabei is sigmoidal in lateral view, more similar to that of A. mississippiensis than of other crocodylians studied herein. Its anterior part is inclined posteordorsally with a slight ventral curvature until the middle of the cerebrum. The posterior part of the brain endocast is ventrally curved (Figure 4b). The dorsal longitudinal venous sinus is not visible on the midbrain, potentially due to low resolution of the tomography.

In the forebrain, the endocast of the olfactory tract is 3 cm long, which represent only 30.9% of the total length of the brain endocast. This is smaller than in Leidyosuchus canadensis and extant crocodylians, but similar to D. ratelii (Figure 5). The olfactory apparatus is bounded by the same bones as in L. canadensis and extant crocodylians and, similarly, the lateral and ventral parts of olfactory tract have been reconstructed with and arbitrary limit due to the lack of bony walls lateroventrally. The endocast of the olfactory bulbs is large, representing around 71% of the width of the cerebral hemispheres. The dorsal part of the olfactory bulbs shows a slight dorsal tilt toward the olfactory tract (~8°), as in L. canadensis but unlike extant crocodylians (see above). The olfactory sulcus is not visible. The dorsal tilt of the olfactory tract is deformed by dorsoventral taphonomic compression, with a slight ventral curvature in its middle. The anterior portion of the olfactory tract is 1 cm wide over a length of 1.1 cm and narrows posteriorly to a width of 0.6 cm before widening again at the anterior part of the cerebral hemispheres (Figure 4b,d,f). The abrupt widening anteriorly is probably artifactual, but the presence of an anterior widening of the olfactory tract is real and is a condition that is not seen in any other crocodylian species studied herein.

The cerebral hemispheres of S. mccabei have a smoother and rounder shape than in L. canadensis, giving the cerebrum a heart shape in dorsal view, with the long sides oriented anteriorly, similar to A. sinensis. The dorsal slope observed in the olfactory apparatus continues along the cerebrum (Figure 4b), until it flattens at the top. The ventral parts of the cerebrum are completely crushed due to taphonomic deformation, but the pituitary fossa is partially preserved (Figure 4b,f), although its connection to the cerebral carotid arteries is broken. The palatine branch of the cranial nerve VII (CN VII) is visible ventral to the pituitary fossa and the cerebral carotid arteries (Figure 4b,f).

The junction between the forebrain and the midbrain is not clearly defined in S. mccabei. As in L. canadensis, the optic lobes and otoccipital venous sinus do not protrude on the surface of the brain endocast (Figure 4b,d). The midbrain is the shortest part of the brain, representing ~6% of the total length of the brain endocast. Its width is ~71% of the cerebrum width, similar to the endocast of the olfactory lobes. The small ventrolateral protrusion is probably artifactual, due to cracks in the braincase. Ventrally, the midbrain is damaged by ventral taphonomic compression.

The dorsal part of the metencephalon is convex above the cerebellum. In lateral view, the dorsal outline of the cerebrum is rounded until the medulla oblongata, most similar to the condition observed in A. sinensis. Unfortunately, the height of the brain endocast cannot be compared due to dorsoventral taphonomic compression in S. mccabei. The otoccipital venous sinus (0.5 cm wide) is marked by the constriction linked to the otic capsule (Figure 4b,d,f).

The trigeminal ganglion is located in a fossa lateral to the brain at the level of the cerebellum/cerebrum transition (Figure 4b,d,f). It is housed outside the primary bony lateral wall of the brain endocast, formed by the postorbital and the laterosphenoid. The width of the trigeminal ganglion is not homogenous, the medial part has a width of 0.35 cm and the lateral part a width of 0.8 cm, Only one neurovascular canal, projecting laterodorsally from the posterior end of the ganglion, is observed, which most likely represents the tympanic branch of the trigeminal nerve. On the ventral side of the hindbrain, a small groove extends from the trigeminal ganglion to the pons, before being replaced by the swelling of the ventral longitudinal venous sinus (Figure 4b,f), which is partially hidden by the taphonomic compression. The trigeminal foramen in the braincase is ~0.35 cm in diameter. On the left side, the posterior part of the sympathetic nerve exits the brain endocast.

The myelencephalon contains the medulla oblongata, in which both the otoccipital venous sinus and the ventral longitudinal venous sinus end. It represents ~13% of the total length of the brain endocast and is of the same shape as in other crocodylians studied. The branch of the hypoglossal nerve CN XIIb exits the medulla oblongata in its anterior part (Figure 4d,f).

3.2.2. Rostrum neurovasculature

Neurovascular canals for the passage of the branches of the maxillary division of the CN V₂ are present in the jugal (canal for the passage of the jugal nerves (cJU), Figures 3b and 6b,d). The precise structure of the cJU is subject to interpretation due to the numerous cracks on both sides of Stangerochampsa mccabei. The measurements have been done on the left side, which is the least damaged.

The CN V₂ enters the jugal posterior to the postorbital bar. The cJU are composed of a main canal, called the jugal main canal (cMJ), of a jugal anterior neurovascular canal (cJANV), and of numerous smaller canals branching off the cMJ. The cMJ extends posteriorly and anteriorly from its entry point and ramifies into smaller canals. The anterior part of the cMJ is ~3 cm long, 1.4 cm high and up to 0.7 cm large, and the posterior part is ~0.7 cm long, ~0.2 cm high and 0.1 cm large. It forms a wide cavity on the right jugal and it is filled with bone in its center on the left jugal, giving it a “donut” shape. This cavity has a “step shape” anteriorly, the ventral part going more anteriorly than the dorsal part of the cMJ (Figure 6d). Numerous cCU ramify from the anterior part of the cMJ, exiting the bone laterally through foramina (Figures 3d and 6b,d). The anteriormost projections of the cMJ is the cJANV. It is 1 cm long, 0.3 cm wide and reaches a large foramen in the anteriormost part of the jugal. This jugal foramen contacts the foramen of the cPNV for the passage of the maxillary branch of the CN V₂ in the maxilla (Figures 3d and 6b,d). Unfortunately, it is impossible to determine if there is continuity between the cJU and the canal for the passage of the maxillary branch of the CN V₂ at the jugal‐maxilla suture, due to insufficient CT scan resolution. The posterior part of the cMJ ramifies into one small canal running laterally (Figure 6b,d).

The maxillary part of the CN V₂ is divided in two main canals, the canal of the superior alveolar nerve (cAs) and the labial canal (cL) (Figures 3 and 7). As for Leidyosuchus canadensis, the preservation of the paranasal cavities is incomplete, and no link with the bony canal for the passage of nerves and blood vessels is visible in S. mccabei. The cAs runs from the 13th maxillary tooth to the maxilla/premaxilla suture and extends into the premaxilla. At the level of the12th tooth, the cAs projects a branch ventrolingually which reconnects with the cAs at the level of the 9th maxillary tooth, forming a lenticular loop. On the posterior part, the cAs has a wide connection with the cL at the level of the 11th maxillary tooth. The medial part of the cAs is not visible due to taphonomic fractures in the skull. Anteriorly, the cAs connects with the cL and nearby tooth alveoli. In the premaxillary‐maxillary notch, the cAs is enlarged labially (Figures 3b,d and 7b,d). Numerous cCU and alveolar canals (cA) project anteriorly from there to reach the 1st premaxillary tooth alveolus, extending laterally and dorsally to the alveoli (Figures 3b,d and 7b).

The cL runs from the 13‐14th maxillary tooth to the anteriormost part of the maxilla and projects a large branch posteriorly, called the posterior neurovascular canal (cPNV), that reaches the last maxillary tooth (Figures 3b,d and 7b). The cPNV is 0.54 mm wide and 3 cm long and ends in the large foramen at the surface of the maxilla described by Wu et al., (1996: 2) as a “recess for blood vessels and nerves” (Figure 3). The cL is particularly wide and flat, with dorsal projections to tooth alveoli and lateral projections to maxillary foramina (cCU). Some of these projections extend through tooth alveoli and reach palatal foramina (Figures 3b,d and 7b).

Contrary to what is seen in L. canadensis, no recesses are observed ventral to the lacrimal canal, but this could be due to taphonomic fractures and deformation.

3.2.3. Paratympanic sinus system

(See Section 3.1 for a detailed description of the paratympanic sinus system).

Stangerochampsa mccabei possesses a dorsoventrally shallow paratympanic sinus system with thick ventral canals, a dorsoventrally deep middle ear cavity, and a well‐developed intertympanic system. This condition is similar to that of Leidyosuchus (Figure 8) but shows a few unique features. The anatomy of S. mccabei will not be compared to the conditions observed in extant crocodylians and D. ratelii for the points that are discussed relative to Leidyosuchus canadensis (see above).

The median pharyngeal sinus is characterized by thick canals. The median pharyngeal foramen (at the basisphenoid‐basioccipital suture) is large, occupying one third of the basicranial width. The median pharyngeal canal, with an elliptical cross section of constant width (Figure 8b,d,h), occupies ~30% of the total length of the sinus system. The pharyngeal intersection is located in the dorsoventral middle of the basisphenoid and occupies half of the median pharyngeal canal height. This condition is very similar to that of L. canadensis.

The anterior branch of the median pharyngeal canal is straight, oriented dorsally, and of the same width as the median pharyngeal canal. Dorsally, the anterior branch of the median pharyngeal canal widens to form a bulbous basisphenoid recess, which contrasts to the thin basisphenoid recess of L. canadensis. On the right side of the specimen, a small sub‐carotid recess is located ventral to the cerebral carotid artery (Figure 8b,d,h). The basisphenoid recess is positioned dorsal to the basioccipital recess. Anterolaterally, the paired recessus epitubaricus form branches that are wider anteroposteriorly than dorsoventrally, resulting in a tubular shape rather than the leaf shape seen in L. canadensis (Figure 8). On both sides, the connection of the recessus epitubaricus to the middle ear cavity is severed due to taphonomic compression of the basicranium that broke the anterior part of the prootic.

There is a small laterosphenoid recess in the ventral part of the laterosphenoid in S. mccabei that connects ventrally to the median pharyngeal canal (Figure 8b,d,f).

Similar to L. canadensis, the posterior branch of the median pharyngeal canal (ventral part of the basioccipital recess) is particularly large in S. mccabei, being confluent with the median pharyngeal canal for half of its length. The canal is directed posteriorly and slightly dorsally, resulting in an ellipsoidal cross section, with a greater height than width (Figure 8b,d). The posterior branch of the median pharyngeal canal is connected dorsally to the paired, subspherical basioccipital recesses and to the rhomboidal recess via a thick canal that is almost of the same diameter as the basioccipital cavity. Peculiarly, the basioccipital hosts another small subspherical cavity dorsal to the basioccipital recess in S. mccabei (Figure 8d,h). This small cavity is connected to the roof of the basioccipital recess and to the rhomboidal recess laterally through to small left and right canals (Figure 8d,h).

The lateral pharyngotympanic tubes are mediolaterally wide and greatly stretched anteroposteriorly in cross section. They bend posteriorly in the dorsalmost part of the basioccipital upon entering the rhomboidal recess, exactly as seen in L. canadensis (Figure 8d,h).

The rhomboidal recess is large and spherical. It connects dorsally to the middle ear cavity via two canals. The first canal is thin and connected anteriorly to the ventral side of the middle ear cavity (Figure 8d). The second canal, oriented posterodorsally, is large and forms a sub‐spherical cavity in the ventral part of the otoccipital, which corresponds to the otoccipital part of the pharyngotympanic recess. This cavity connects dorsally to the intertympanic sinus and anterolaterally to the middle ear cavity (Figure 8d,h).

The middle ear cavity of S. mccabei is similar to the condition seen in L. canadensis: it is oriented mediolaterally and displays a large ventral bulge. (Figure 8d,h). The middle ear cavity is large, both dorsoventrally and mediolaterally. The ventral part of the middle ear cavity forms a wide cavity in the quadrate along its suture with the otoccipital. The middle ear cavity directly connects the tympanum to the inner ear, but the columella is not preserved (Figure 8d,h).

The prootic facial recess is subspherical and connects to both the middle ear cavity posteriorly and to the intertympanic system dorsally (Figure 8b).

The quadrate recess is a large capsule in the lateral part of the quadrate, which is anteroposteriorly elongate and connects to a thick siphonium. In S. mccabei, the quadrate recess is connected to the siphonium posteriorly and to the middle ear cavity via two short canals, one oriented anteroventrally and the second oriented dorsomedially (Figure 8d,f). This is similar to the condition seen in A. sinensis, although the latter has a thinner anteroventral canal, but different from that of A. mississippiensis, in which the anterior canal is thicker, less ventrally located, and the dorsomedial canal is absent. Just like S. mccabei, D. ratelii also possesses an anteroventrally oriented canal, but it connects to the infundibular recess and not directly to the middle ear cavity (Figures 8 and 9). The siphonium of S. mccabei is wide and short (Figure 8d,f,h), oriented posteroventrally, and the corresponding foramen aereum is located on the dorsal surface of the quadrate as in L. canadensis and extant alligatoroid.

The infundibular recess, located anterior to the quadrate recess, is a thick canal with a circular cross section that tapers rapidly anteriorly and connects to the anterior wall of the middle ear cavity. This condition differs form that observed in Alligator spp., C. niloticus, and D. ratelii in which the canal is of constant diameter (Figures 8d,f and 9). No infundibular recess can be observed in L. canadensis. The subtympanic foramen is circular and measures 2 mm in diameter (Figure 8f).

The intertympanic recess of S. mccabei display no anteroposterior compression in its medial part. The lateral parts of the intertympanic recess are largely confluent with the paired otoccipital recesses, which inflate the otoccipital as in L. canadensis (Figure 8f).

The otoccipital recess expands ventrally, but flattens posteriorly, adopting a triangular shape in transverse view, with a laterodorsal protrusion directed toward the temporal canal. This protrusion cannot be seen in the other crocodylian species studied herein (Figure 8d,h).

The parietal recess of S. mccabei is dorsally flattened, and anteroposteriorly and mediolaterally expanded, similar to the condition seen in Caiman species (Perrichon, 2024). The paired posteromedial pre‐parietal recesses linked to the roof of the intertympanic recess have a circular cross section and extend dorsally for a short distance prior to turning anteriorly toward the parietal (Figure 8b,d,f,h). The orientation of these recesses is similar to that observed in D. ratelii and Alligator spp., but differs from C. niloticus and L. canadensis in which the recesses are more horizontal (Figures 8 and 9). The paired anterolateral pre‐parietal recesses are wider and directed anterodorsally. All these recesses enter the parietal and merge to form the parietal recess (Figure 8d,f,h). Finally, S. mccabei possesses a wide prootic‐parietal canal, similar to L. canadensis and extant Alligatoridae.

4. DISCUSSION

4.1. The rostral neurovascular system of crocodylians

The rostrum of crocodylians contain cavities and canals. The cavities hosts the paranasal sinuses and the nasal canal (Cowgill et al., 2022; Witmer & Ridgely, 2008), while the canals host both blood vessels and nerves (Porter et al., 2016; Sedlmayr, 2002); thus, they are here referred to as the neurovascular system. Unfortunately, it is currently impossible to determine more precisely the function of these canals because the proportion of nerves and blood vessels vary between species; thus, changes in the osteological canals can represent changes in innervation, in the vascular system, or both. The topology of these structures, can, however be compared between different taxa. No link between the paranasal cavities and the neurovascular canals is visible in either specimen studied (TMP1986.221.1 and TMP1986.61.1). Moreover, the diameter, position, and continuity of the reconstructed bony canals is consistent with such structure previously observed in crocodyilans (Bowman et al., 2022; Cowgill et al., 2022; Lessner & Holliday, 2020).

4.1.1. Jugal canals of the neurovascular system

As the different specimens studied have not been CT scanned with the same resolution and contrast, it is not always possible to reconstruct the smaller canals, but the general shape of the canals is always visible. The foramen allowing the entry of the jugal branch of the CN V₂ into the jugal is located on the posterior face of the postorbital bar in all species studied except A. sinensis, in which there are two foramina, exclusively on the anterior face of the postorbital bar (Figure 6) although it is not clear if this is a specific feature or linked to this particular specimen. Stangerochampsa mccabei also possesses a foramen in the anterior face of the postorbital bar, as well as C. niloticus and Alligator sp. but it is absent in Leidyosuchus canadensis and Diplocynodon ratelii (Figure 6). We note that, in C. niloticus, the two foramina on the anterior part of the postorbital bar are much larger than the foramen on the posterior surface (Figure 6i,k). In L. canadensis, the overall shape of the cMJ is similar to that of Alligator spp.: the posterior part has thin circular branches whereas the anterior part is curved ventrally and has numerous branches of different sizes (Figure 6). However this condition differs from that of C. niloticus and D. ratelii. In C. niloticus, the overall shape of the cMJ is straight, without curvature on its anterior part. All the cCU branches are thin and circular in cross section. Moreover the anterior part of the cMJ, shows branches in the anteriormost part of the jugal and is connected with canals in the lacrimal (Figure 6j,k). The overall shape of the cMJ in D. ratelii resemble more the one of S. mccabei: it is thick, with a “step shape” anteriorly, and it has a large anterior canal with an enlarged anterior projection (Figure 6b,d,j,l). The cMJ of D. ratelii differs from that of S. mccabei in its anterior projection, which does not end in an enlarged foramen as the cJANV of S. mccabei, and in its posterior part, which is longer and exhibits more branching than in S. mccabei (Figure 6b,d,j,l).

In summary, the jugal neurovasculature of L. canadensis resembles that of modern Alligator spp., while that of S. mccabei is closer to D. ratelii. Finally, the jugal neurovasculature of C. niloticus differs from all studied alligatoroids (Figure 6).

4.1.2. Maxillary canals of the neurovascular system

The comparison of the general features of the maxillary neurovascular system between the Upper Cretaceous alligatoroids and the other species studied herein is summarized in Table 4.

TABLE 4.

Comparison and general features of the endocast, paratympanic sinuses and rostral neurovasculature of the species studied herein.

	Leidyosuchus canadensis	Stangerochampsa mccabei	Alligator mississippiensis	Alligator sinensis	Diplocynodon ratelii	Crocodylus niloticus
Endocast
Overall shape	Sigmoidal	Sigmoidal, with rounder shapes	Sigmoidal	Sigmoidal, with rounder shapes	Sigmoidal	Sigmoidal
Olfactory tracts	No special feature	Wider anteriorly	No special feature	No special feature	No special feature	No special feature
Cerebrum	Kite shape in dorsal view	Heart shape in dorsal view	Heart shape in dorsal view	Heart shape in dorsal view	Heart shape in dorsal view	Round shape in dorsal view
Paratympanic sinus system
Overall shape	Dorsoventrally short	Dorsoventrally short, round and wide	Dorsoventrally high	Dorsoventrally high	Dorsoventrally high	Thin, dorsoventrally high
Pharyngotympanic system
Laterosphenoid pneumatization	Absent	Present, with ventral connection	Present, with posterior connection	Present, with ventral and posterior connection	Present, with posterior connection	Absent
Pharyngeal canals	Thick	Very thick	Intermediate	Intermediate	Intermediate	Thin
Pharyngeal intersection	Large	Large	Short	Large	Medium	Short
Posterior median canal	Large ellipsoid bulge	Large, posteriorly oriented, ellipsoidal in cross section	Posteriorly oriented, ellipsoidal in cross section	Dorsally oriented, ellipsoidal in cross section	Posteriorly oriented, ellipsoidal in cross section	Dorsally oriented, circular in cross section
Basioccipital recess	No special feature	Topped by a small cavity connected to the basioccipital recess and rhomboidal recess	No special feature	No special feature	No special feature	Thin
Siphonium	Intermediate	Thick	Intermediate	Thin	Intermediate	Thin
Middle ear	Mediolaterally oriented, with a large ventral bulge	Mediolaterally oriented, with a large ventral bulge	Medioventrally to laterodorsally oriented	Medioventrally to laterodorsally oriented	Medioventrally to laterodorsally oriented	Medioventrally to laterodorsally oriented
Intertympanic system
Overall shape	No special feature	No medial compression	No medial compression	No special feature	No special feature	No special feature
Volume	Low degree	High degree	High degree	High degree	High degree	Very low degree
Prootic‐parietal canal	Present	Present	Present	Present	Present	Absent
Otoccipital recess	Developed ventrally with a triangular shape in transverse view, laterodorsal protrusion toward the temporal canal	Developed ventrally with a triangular shape in transverse view, laterodorsal protrusion toward the temporal canal	Developed ventrally	Developed ventrally	Developed ventrally	Dorsoventrally short
Parietal recess	Very small posteromedial preparietal processes oriented horizontally, and anterolateral pre‐parietal processes merging to form a dorsoventrally hight parietal recess	Dorsoventrally flattened and mediolaterally expanded. Posteromedial pre‐parietal processes extending dorsally then anteriorly toward the parietal. Anterolateral pre‐parietal processes directed anterodorsally	Posteromedial pre‐parietal processes extending dorsally then anteriorly toward the parietal	Posteromedial pre‐parietal processes extending dorsally then anteriorly toward the parietal	Posteromedial pre‐parietal processes extending dorsally then anteriorly toward the parietal	Posteriomedial pre‐parietal processes oriented horizontally.
Rostral neuravoscular sytem
Posterior neurovascular canal	Present	Present	Absent	Absent	Absent	Absent
Labial canal	Present, large	Present, large	Present, discontinuous	Present, discontinuous	Present, very fragmentary	Absent
Ventrolingual canal	Present	Absent	Absent	Absent	Absent	Absent
Superior alveolar canal	Present	Present	Present	Present	Present	Present

Previous studies about phylogenetic topologies reveal artificial clustering of longirostrine species apart from brevirostrine species, questioning the possible biases caused by snout morphology (Brochu, 2001; Drumheller & Wilberg, 2020; Sadleir & Makovicky, 2008). A similar clustering is also obserbed in alligatoroids, with Stangerochampsa mccabei clustering with other brevirostrine species (e.g., Brachychampsa montana, Albertochampsa langstoni) (Sadleir & Makovicky, 2008). However, even though Leidyosuchus canadensis and Stangerochampsa mccabei are mesorostrine and brevirostrine species respectively, they both possess a large foramen at the posterior edge of the maxilla, and are the only species to possess such a feature (Wu et al., 1996, 2001) (Figure 1). In both species, this foramen is connected to the neurovascular canals in the maxilla via the cPNV (Figures 3 and 7). The presence of the cPNV in L. canadensis and S. maccabei raises questions about their phylogenetic proximity. The similarities in shape and position of this canal and its link to the cL led us to hypothesize a secondary homology between this structure in both species. As such, it would be interesting to test the impact of this character on phylogeny in future studies.

Overall, the general orientation and volume of the neurovascular canals in L. canadensis and S. mccabei follow the shape of the rostrum, explaining numerous differences between the two species, notably at the premaxilla‐maxilla suture and in the premaxilla (Figures 3a,b and 7). Differences are also observed in the jugal, such as the proximity of the foramen of the cJANV with the foramen of the cPNV in S. mccabei that is not visible in L. canadensis, although taphonomic fractures in L. canadensis prevent a precise comparison.

Both L. canadensis and S. mccabei possess a cAs and a cL (Figures 3, 7 and 10). The cAs is also present in the extant crocodylians species studied herein and seems to be a shared feature in all crocodylians, despite terminological differences (Bowman et al., 2022; Lessner et al., 2022) (Figures 7 and 10). On the other hand, the cL is only visible in alligatoroids. The cL is absent in extant Longirostres (Crocodylus rhombifer, Gavialis gangeticus and Tomistoma schlegelii), as they possess instead a circular canal, named canal of the integumentary system (cIS), in the labioventral part of the maxilla (Bowman et al., 2022); our observations of perfectly circular canals in Crocodylus niloticus (Figure 10) are consistent with this observation. The labial position of both cL and cIS raises doubts about their possible homology but ontogenetic work will be needed in order to assess the validity of this hypothesis. The morphology of the cL differs between alligatoroid species, especially in its continuity along the maxilla. It is continuous in L. canadensis and S. mccabei, almost running along the entire length of the maxilla (Figures 3 and 7). In Diplocynodon ratelii, the cL is mostly present posteriorly and it is not generally as flattened as in L. canadensis and S. mccabei, except at three locations, two on the posterior edge of the maxilla, and at the junction with the cAs (Figure 10). Serrano‐Martínez et al. (2025) reconstructed part of the neurovascular system of D. tormis, but they only focused on the cAs (which they called the dorsal alveolar canal after Bowman et al. (2022)), which precludes comparison with the cL or the cPNV. The cL is mostly visible posteriorly in Alligator sinensis, running from the 12th maxillary tooth to the 4th maxillary tooth, and adopts a large and flatten bean shape in transverse view; anteriorly, the shape is less typical and a bit more fragmentary. In Alligator mississippiensis, the cL runs over a longer distance than in D. ratelii and A. sinensis but is discontinuous (Figure 10).

Lastly, both L. canadensis and S. mccabei possess a posteroventral, labial projection of the cL, for which we introduce the new term: cPNV, that extends posteriorly to the cAs up to the penultimate maxillary tooth and ends in a wide ellipsoid foramen on the posterolateral edge of the maxilla (Figures 7 and 10). We consider the cPNV as a modified cCU, due to its link with a maxillary foramen, allowing nerves to innervate the skin as for cCU. The cPNV is remarkable and differs from cCU by its important size, its posterior most position and its lack of ramifications. cCU are also present in the posterior region of the maxilla in the other crocodylian species studied herein; however, these cCU do not exhibit the same characteristics as the cPNV. The posterior cCU in D. ratelii is very short and thin, and connects to a medium‐sized foramen (relative to the other foramina in the maxilla) on the posterior edge of the maxilla, and does not extend posterior to the dorsal part of the cL (Figures 7 and 10); the posterior cCU in A. mississippiensis extends farther posteriorly than in any other crocodylian species studied herein, but it is very thin and connects to a medium‐sized foramen relative to the other foramina in the maxilla (Figure 10); in A. sinensis, the posterior cCU is very large but several other cCU branches separate from it, and it does not extend posterior to the cAs et the dorsal branch of the cL (Figure 10); Crocodylus niloticus possesses a short and thin posterior cCU with several branches extending labially and a long, thin cCU projecting lingually (Figure 10). Thus, the cPNV is a unique feature shared only by L. canadensis and S. mccabei.

In summary, the morphology of the rostral neurovascular system is similar among alligatoroids, which all share a cL, and differs from that of Longirostres, which lack a cL. Furthermore, L. canadensis and S. mccabei share a unique feature, the cPNV, which appears to be a developed version of the labial cCU projection seen in more derived alligatoroids. As this projection is linked to the neurovascular system, we suppose it played an ecological role in connection with neurosensorial organs, like the integumentary sensory organs (Leitch & Catania, 2012; Soares, 2007), or with thermoregulation. Lessner et al. (2023) showed that mandibular neurovascular canals host more nervous tissue than vascular ones in tactile taxa, thus the hypothesis of a specialized sensitive organ seems plausible. However, the precise function of that characteristic cannot be currently determined.

4.2. Endocranial comparison of Leidyosuchus canadensis and Stangerochampsa mccabei

4.2.1. Brain endocast

The general features of the brain endocasts of the Late Cretaceous alligatoroids and other species crocodylians herein are summarized in Table 4. Several endocranial features differ between Leidyosuchus canadensis and Stangerochampsa mccabei. The cerebrum of L. canadensis exhibits a distinctive kite‐shaped outline in dorsal view (Figure 4c), contrasting with the rounder shape observed in all other crocodylians studied herein (Figures 4 and 5). The cerebral hemispheres and the curvature of the midbrain and hindbrain in S. mccabei are rounder and smoother than in L. canadensis (Figure 4a–f). In addition, the endocast of the olfactory bulbs, the cerebrum and the midbrain are proportionally larger in S. mccabei than in L. canadensis (Table 3). Lastly, the endocast of the olfactory tract of S. mccabei shows a wider anterior part narrowing abruptly at the anterior third of the olfactory tract, which is not seen in L. canadensis nor in the other crocodylians studied herein (Figure 4d,f).

TABLE 3.

Comparative measurement of the brain endocast of the species studied herein.

	Leidyosuchus canadensis	Stangerochampsa mccabei	A. mississippiensis	A. sinensis	C. niloticus	Diplocynodon ratelii
Brain endocast (lenght)	135	97	113	75	123	92
Olfactory apparatus (tract and bulb) (lenght)	70	37	52	31	64	37
Olfactory apparatus, % of total length	51.8	38.1	46	41	52	40
Olfactory bulbs (width)	18	17	13	11	15	8.2
Cerebral hemispheres (width)	27	24	20	20	31	21
Cerebral hemispheres (height)	19	?	19	15	26.6	15
Midbrain region (width)	18	17	12.7	11.7	16.5	12
Midbrain region (length)	~7	~6	8.1	7.5	7.6	6
Midbrain region (height)	17	?	17	19	23	16
Hindbrain region (width)	18	10	12	12	18	14
Hindbrain region (height)	23	16	22.5	23	27	21
Medulla oblongata (length)	12	13	12	?	12	13
Pituitary fossa (length)	8.3	5	20	14	?	13
Pituitary fossa (width)	8.2	4	4	5.8	?	6
Pituitary fossa (height)	4.8	?	7.6	9	?	6

Note: Values in mm.

Surprisingly, the overall shape of the brain endocast of L. canadensis and S. mccabei differs from that of D. ratelii (Figure 5) and D. tormis (Serrano‐Martínez et al., 2019; Serrano‐Martínez et al., 2025), in which it is much straighter posteriorly, with no tilt between cerebrum, midbrain and cerebellum. The endocast of the olfactory bulbs of D. ratelii is narrower than in L. canadensis and S. mccabei, but the length of the endocast of the olfactory apparatus is similar to that of S. mccabei (Figures 4 and 5).

Both Late Cretaceous alligatoroids share similarities with the extant genus Alligator. L. canadensis and S. mccabei have an overall sigmoidal‐shaped brain endocast, as in other crocodylians, and a dorsal tilt of the olfactory apparatus toward the cerebrum (Figure 4). The brain endocast of L. canadensis is more similar to that of A. mississippiensis than to A. sinensis in terms of overall length and shape, being more elongate and slenderer. Conversely, the brain endocast of S. mccabei is most similar to that of A. sinensis, characterized by a rounder curvature of the cerebrum and midbrain, and heart‐shaped cerebral hemispheres in dorsal view. These similarities are more likely the result of skull construction constraints than reflecting shared traits as explained below.

Ontogenetic studies of crocodylians have shown that the brain and its corresponding brain endocast are smoother, wider, stockier, and have a prominent sigmoidal profile in early developmental stages, while they become straighter, flatter, more elongate and largely similar in terms of overall shape in adult forms (Jirak & Janacek, 2017; Lessner et al., 2022). Furthermore, comparison between avian and crocodylian brain endocast have shown that avian brain endocasts are pedomorphic compared to crocodylian brain endocasts (A. mississippiensis and C. niloticus), and that the shape changes observed between juvenile and adult crocodylian brains are on the same exaggerated trend as those observed between avian and crocodylian brains (Beyrand et al., 2019; Hu et al., 2021). Indeed, differences in brain endocasts shape among crocodylian species appear to represent allometric responses related to specimen size, possibly similar to patterns observed in endosseous labyrinth morphology (Pochat‐Cottilloux et al., 2024). Consequently, the overall similarities in brain shape observed between L. canadensis and A. mississippiensis, two medium‐snouted species (Elsey & Woodward, 2010; Wu et al., 2001), and between S. mccabei and A. sinensis, two short‐snouted species (Jiang et al., 2010; Wu et al., 1996), could be most parsimoniously explained as allometric responses related to their similar skull size rather than reflecting ecological, behavioral, neurological or phylogenetic signals. As such, the morphological characteristics of the brain endocasts may not be relevant to assess phylogenetic relationships among crocodylian species.

4.2.2. Paratympanic sinuses

Leidyosuchus canadensis and Stangerochampsa mccabei share similarities unique to both: a middle ear oriented mediolaterally and with a large ventral bulge; an otoccipital recess developed ventrally with a laterodorsal protrusion directed toward the temporal canal. Moreover, L. canadensis and S. mccabei share several characteristics with the distantly related brevirostrine alligatoroid Caiman spp., and the brevirotrine crocodylid Osteolaemus spp. (Perrichon, 2024). These features are a dorsoventrally short paratympanic sinus system with thick ventral canals, a dorsoventrally large middle ear, large pharyngotympanic tubes, and thick pharyngeal canals that meet at a large pharyngeal intersection located halfway in the basisphenoid, and whose transversal section occupies half of the median canal, all of which are pedomorphic features (Figure 8) (Perrichon, 2024). Additionally, S. mccabei shares other pedomorphic characters with Caiman spp. and Osteolaemus spp., namely a ventral connection of the laterosphenoid recess and the rounder and wider paratympanic sinuses cavities (Perrichon, 2024). Given that cranial shape is generally correlated with body size in crocodylians (Godoy, 2020), we can suppose that the features shared by L. canadensis, S. mccabei, Caiman spp., and Osteolaemus spp. may be pedomorphic convergences linked to brevirostry. If true, the presence of such “pedomorphic features” in the mesorostrine L. canadensis, may be interpreted in terms of phylogenetic trait rather than a growth constraint, supporting the hypothesis of a brevirostrine ancestor of alligatoroids (Brochu, 2001). Lastly, L. canadensis and S. mccabei both share alligatoroid characters, such as the presence of a prootic‐parietal canal in the intertympanic sinus system (Figure 8) (Perrichon, 2024). They also show a position of the foramen aereum on the dorsal surface of the quadrate (Figure 8), although this character is now debated as a synapomorphy of alligatoroids (Rio & Mannion, 2021).

The paratympanic sinuses of the Leidyosuchus canadensis and Stangerochampsa mccabei can be differentiated from each other based on: (1) the absence of pneumatization in the laterosphenoid in L. canadensis (similar to the condition seen in D. ratelii; Figure 9) and D. tormis (Serrano‐Martínez et al., 2019), (2) the low degree of pneumatization of the intertympanic sinuses of L. canadensis (unlike D. ratelii, D. tormis (Serrano‐Martínez et al., 2019, Serrano‐Martínez et al., 2025) and extant alligatoroids), and (3) a thicker siphonium in S. mccabei (Figure 8, Table 4). L. canadensis and S. mccabei also possess autapomorphic characters of the paratympanic sinuses not seen in any other crocodylians. L. canadensis possesses a posterior median canal that forms a large ellipsoidal bulge. On the other hand, S. mccabei exhibits a small subspherical cavity dorsal to the main basioccipital recess and connected to the rhomboidal recess via left and right canals, as well as a subspherical bulge located in the canal connecting the rhomboidal recess to the middle ear, in the ventral part of the otoccipital. The comparison of the general features of the sinus system between L. canadensis and S. mccabei and the other species studied herein are shown in Figures 3, 8 and 9, Figure S2 and resumed in Table 4.

4.3. Implications for phylogenetic studies

In most studies of crocodylian relationships, especially those involving fossil taxa, anatomical description and diagnostic characters are primarily based on the external features of the skull (Brochu et al., 2022; Groh et al., 2020; Perrichon et al., 2024; Rio & Mannion, 2021; Ristevski, 2022; Salas‐Gismondi et al., 2016). Alligatoroids are diagnosed by four unambiguous cranial synapomorphies according to Rio and Mannion (2021), which are also shared by both Leidyosuchus canadensis and Stangerochampsa mccabei. The position of L. canadensis as a basal alligatoroid has been recently questioned by Walter et al. (2025), raising doubts about the validity of the current synapomorphies of alligatoroids. The position of S. mccabei within the larger clade Globidonta appears quite strong, as it possesses all unambiguous synapomorphies of the clade (all cranial characters, mostly linked to the rear part of the skull) sensu Brochu (1999). A recent study recovered the unnamed clade formed by S. mccabei and other Late Cretaceous North American brevirostrines alligatoroids (i.e., Brachychampsa, Ceratosuchus, Wannagonosuchus) inside the Caimaninae (Rio & Mannion, 2021). However, this phylogenetic placement is supported by only one exclusive synapomorphy (presence of foramina on the medial parietal wall), which has been scored for Brachychampsa montana only and not for S. mccabei, Ceratosuchus nor Wannagosuchus (Rio & Mannion, 2021). Thus, the exact position of S. mccabei within Globidonta is far from resolved.

We found that the alleged phylogenetically distant L. canadensis and S. mccabei share several unique internal cranial features, such as the posterior neurovascular canal exteriorly connected to the recess for blood vessels and nerves in the maxilla described by Wu et al. (1996, 2001), the orientation of the middle ear cavity, and an extension of the otoccipital recess in the sinuses. Taking into account the internal skull features of L. canadensis and S. mccabei could allow to determine more precisely their phylogenetic position. Given the anatomical similarities between those taxa, it is possible that S. mccabei would occupy a more basal position, possibly as one of the basalmost Globidonta, or alternatively as a stem crocodylian if the position of L. canadensis recovered by Walter et al. (2025) is confirmed in future studies. These phylogenetic questions will have to be evaluated in a future study that considers the internal anatomy of a larger sample of fossil crocodylians.

The endocranial characteristics documented herein could be phylogenetically informative and would require to be tested in future work. Moreover, they could shed light on the paleoecology of fossil species, especially for the rostral cavities. The posterior neurovascular canal shared by L. canadensis and S. mccabei exits in the enlarged foramen situated on the posterior edge of the maxilla described by Wu et al. (1996, 2001). Interestingly, the neosuchians Goniopholididae and Pholidosauridae, and potentially the Jurassic mesoeucrocodylian Hsisosuchus all possess a similar structure in the same area, called the maxillary depression (Martin & Buffetaut, 2012). This deep, horizontally ovoid depression contains three to five subdivisions with marked borders and a vertically ovoid shape. Although the maxillary foramen connected to the posterior neurovascular canal in L. canadensis and S. mccabei and the maxillary depression of Goniopholididae and Pholidosauridae have clearly different shapes, both are suspected to have had a sensory function (Andrade, 2009; Martin & Buffetaut, 2012). We can hypothesize that an ecological convergence occurred between the clade containing Goniopholididae and Pholidosauridae and the one containing L. canadensis and S. mccabei.

5. CONCLUSIONS AND PERSPECTIVES

Endocranial structures, namely the brain endocast, the paratympanic sinus system, and the rostral neurovascular canals of Leidyosuchus canadensis and Stangerochampsa mccabei, are described for the first time and compared with those of other crocodylians, the extinct Diplocynodon ratelii, and the extant Alligator sinensis, Alligator mississippiensis, and Crocodylus niloticus. L. canadensis and S. mccabei possess characters observed in Alligator spp. and D. ratelii, but also share unique characters of the paratympanic sinus system and the posterior neurovascular canal. The brain endocast of L. canadensis is similar to that of A. mississippiensis whereas that of S. mccabei is most similar to A. sinensis, likely due to convergence in snout morphology (mesorostrine and brevisrostrine, respectively). Furthermore, the paratympanic sinus systems of L. canadensis and S. mccabei display pedomorphic characteristics, namely large and short paratympanic sinus system with particularly large pharyngotympanic tubes. These shared pedomorphic features could bring S. mccabei into a more basal position within Globidonta, or possibly in a stem‐crocodylian position with L. canadensis. Finally, L. canadensis and S. mccabei share a unique morphology of the posterior maxillary neurovascular canals for the CN V₂, unseen in our dataset in other crocodylians.

Future studies should aim to include endocranial characters in phylogenetic matrices, especially aspects of the paratympanic sinus system and of the posterior neurovascular canal. Moreover, endocranial reconstruction of Brachychampsa montana and Albertochampsa langstoni, the sister taxa of S. mccabei, would allow a more in‐depth comparison of their internal cranial structures.

AUTHOR CONTRIBUTIONS

JM: Conceptualization. GD, GP, PV, FT, JM: Investigation and Analysis. FT, JM: Data Curation. GD, GP, PV, FT, JM: Methodology. FT, JM: Resources. GD, GP: Visualization. GD, GP, PV, FT, JM: Validation. JM and PV: Supervision. GD: Writing original draft. GD, GP, PV, FT, JM: Writing review and editing.

Supporting information

Figure S1. CT‐scan slices of Leidyosuchus canadensis (TMP1986.221.1) (a) Stangerochampsa mccabei (TMP1986.61.1) (b), Alligator mississippiensis (OUVC_9761_M39878‐71998) (c), Alligator sinensis (NHMW‐Zoo‐HS‐37966) (d), Crocodylus niloticus (MHNL_50001399) (e) and Diplocynodon ratelii (MHNL‐LA86) (f) in transversal view. clac: lacrymal canal, infLacR: infra lacrimal recess. Scalebars represent 5 cm.

Figure S2. Paratympanic sinus system 3D reconstruction in posterolateral view of Alligator mississippiensis (OUVC_9761_M39878‐71998) (a), Alligator sinensis (NHMW‐Zoo‐HS‐37966) (b), Crocodylus niloticus (MHNL_50001399) (c) and Diplocynodon ratelii (MHNL‐LA86) (d). BoR, basioccipital recess; BsR, basisphenoid recess; IntR, intertympanic recess; MPhC, median pharyngeal canal; MPhf, median pharyngeal foramen; OtoR, otoccipital recess; PhR, Pharyngotympanic recess (middle ear cavity); PhT, pharyngeal tubes; PostMC, posterior branch of the median pharyngeal canal; PR, parietal recess; QR, quadrate recess; REt, recessus epitubaricus; RhR, rhomboidal recess; sf, subtympanic foramen; Siph, siphonium. Skulls are illustrated at the same scale. Scalebars represent 2 cm.

DATA AVAILABILITY STATEMENT

3D models of the described specimens and additional figures are available online at Morphomuseum (MorphoMuseuM e284. doi: https://doi.org/10.18563/journal.m3.284). The original CT scan of NHMW‐Zoo‐HS‐37966 is available on the NHMW data repository (https://doi.org/10.57756/fw24d5). The other original CT scan data are available from the corresponding authors at a direct request.