Postosuchus kirkpatricki, Chatterjee, 1985

Gower, David J., 2002, Braincase evolution in suchian archosaurs (Reptilia: Diapsida): evidence from the rauisuchian Batrachotomus kupferzellensis, Zoological Journal of the Linnean Society 136 (1), pp. 49-76 : 64-66

POSTOSUCHUS KIRKPATRICKI

Postosuchus kirkpatricki Chatterjee is known from the Upper Triassic of the south-western United States of America ( Chatterjee, 1985; Long & Murry, 1995). It has a confused taxonomic history ( Long & Murry, 1995; Gower, 2000), but it seems that the documented braincase material from the type locality can be relatively confidently associated with the species name P. kirkpatricki . Chatterjee (1985: fig. 7) figured three views of a composite restoration of the braincase of P. kirkpatricki and described and interpreted aspects of the osteology. The braincase of the holotype (TTUP 9000) is currently largely obscured by paint and plaster applied during the restoration of the skull. My observations of this specimen were aided by access to photographs (provided by S. Chatterjee) taken of the specimen prior to restoration. I also examined the other TTUP P. kirkpatricki braincase, no. 9002.

My observations agreed with the view presented by Chatterjee (1985) in terms of the general proportions and in many of the sutural relations and detailed features. I concur that there is a single opening for cranial nerves IX-XI (metotic foramen), a laterally positioned entrance of the cerebral branch of the internal carotid artery, and an unossified gap between the basal tubera of the basioccipital and parabasisphenoid that was probably filled with cartilage in life. Some other details merit further discussion

There is a strong subvertical ridge on the lateral surface of the exoccipital and upper part of the basioccipital, but this lies posterior, not anterior, to the single external foramen for the hypoglossal nerve (seen on the left side of TTUP 9002), so that this foramen is not (contra Chatterjee, 1985: fig. 7) visible in posterior view. This ridge runs down onto the dorsolateral edge of the basal tubera of the basioccipital. It is stronger than the corresponding ridge in B. kupferzellensis and, as in Stagonolepis robertsoni ( Gower & Walker, 2002) and crocodylomorphs ( Walker, 1990), it extends further laterally than the ventral ramus of the opisthotic. Unlike the situation in B. kupferzellensis , the basal tubera of the basioccipital are barely bilobed, and there is no clear recess on the posterior surface of the basioccipital between the tubera. The sutural relations between opisthotic and prootic on the paroccipital process are not clear. The foramina for the cerebral branches of the internal carotid arteries are within depressions that lie close to the midline, where the parabasisphenoid is very narrow. Chatterjee identified a separate presphenoid ossification articulating with the laterosphenoid and parabasisphenoid. Chatterjee’s (1985: fig. 7) reconstruction of this region presents a highly unusual arrangement for a Triassic archosaur. The known part of the laterosphenoid is shown as sandwiched between the prootic and the base of the cultriform process of the parabasisphenoid, and as forming the anteroventral half of the border to the trigeminal foramen and even part of the crista that overhangs that part of the parabasisphenoid pierced by the foramen for the cerebral branch of the internal carotid foramen. This region is incompletely preserved in TTUP 9002, and it does not provide support for this reconstruction. The surface of TTUP 9000 (photographs and direct observation) is poorly preserved and difficult to interpret. At least that part of the laterosphenoid-prootic suture depicted as lying immediately posteroventral to the trigeminal foramen might be instead interpreted as an artefactual crack. I suggest that determination of the exact pattern of elements here, including the possible presence of a presphenoid, requires additional material. There are several features for which it is not possible, in the known specimens, to ascertain the condition with any confidence including the presence/ absence of a discrete epiotic, the exact path of the abducens and facial nerves through the braincase wall, and the entire shape of the cultriform process of the parabasisphenoid.

The endocranial cavity is exposed in TTUP 9002. The exoccipitals meet along the midline and, anteriorly, they appear to be separated here from the prootics by exposure of the parabasisphenoid and basioccipital. As in B. kupferzellensis , there is a abrupt drop in the level of the floor of the endocranial cavity from the hind part formed by the exoccipitals to the mid part in front of this. On each side, close to the lateral edge of the floor of the endocranial cavity, there are two clearly defined cup-like depressions. The more posterior of these (visible in Parrish, 1993: fig. 7A) lies just in front of the base of the exoccipital pillar and is here interpreted as the floor of the metotic foramen. The more anterior depression lies a short distance in front of this and is here interpreted as the lagenar/ cochlear recess. This is not as deep as in crocodylomorphs, but it is better defined than in non-crown-group archosaurs and phytosaurs ( Walker, 1990; Gower & Weber, 1998). Its upper, anterior limit is not marked by any sign of a crista demarcating the border between the lower end of the vestibule and the upper end of the lagenar recess. On the left side of TTUP 9002, the lagenar/cochlear recess lies along the basioccipitalparabasisphenoid suture. Anteriorly, in the same specimen, the opposite prootics seems to make midline contact. TTUP 9002 also presents evidence that P. kirkpatricki closely resembled B. kupferzellensis in possessing a separate foramen possibly for the passage of the posterior cerebral/cephalic vein. Seen best on the left side, the wall of the endocranial cavity immediately above the metotic foramen bears a crescent-edged depression that is interpreted as the position of a venous sinus. A foramen within this depression passes through the thin anterior edge of the exoccipital pillar to emerge immediately above the external opening of the metotic foramen. This is also visible on the right side of TTUP 9002, but the medial surface is partly broken.

In addition to TTUP 9000 and TTUP 9002, Chatterjee (1985: 407) referred a third braincase to P. kirkpatricki, UMMP 7473. This had previously been figured by Case (1922: pl. 13), who referred this isolated braincase to the dinosaur Coelophysis . Chatterjee considered it to differ from the braincase of Known Coelophysis and instead found that UMMP 7473 “corresponds so well” with the TTUP braincases referred to P. kirkpatricki that he proposed that it represented the same taxon. I have examined UMMP 7473 and believe that Chatterjee’s taxonomic identification is not correct. The specimen is incomplete (see figures in Case, 1922), consisting of the occipital condyle, the endocranial cavity, and the proximal parts of the parabasisphenoid and paroccipital processes. The region of the fenestra ovalis, ventral ramus of the opisthotic and perilymphatic foramen has been obliterated during preservation/preparation on both sides and is marked now by only a single large hollow. The suture between exoccipitals and the condylar portion of the basioccipital cannot be detected. The structures labelled ‘bpt’ by Case are not the basipterygoid process as such, but rather the incomplete parts of the parabasisphenoid that would have supported these processes. The part of the parabasisphenoid that has been preserved does resemble that of P. kirkpatricki in that it consists of a pair of thin subvertical plates of bone that border a transversely narrow but deep median pharyngeal recess. Additionally, a small depression is positioned just above this recess and between the ventral edge of the basal tubera of the basioccipital, resembling a similar feature in TTUP 9000 and 9002. The endocranial cavity of UMMP 7473 is also exposed. Features observable here include well ossified otic bullae, well marked fossae indicating the position of venous sinuses and the passage of the dorsal head veins, shallow auricular fossae, and an abrupt step down on the floor of the endocranial cavity anterior to the front end of the exoccipitals. These features are consistent with this specimen being P. kirkpatricki , but they are also found in a wider range of Triassic archosaurs.

Foramina that are interpreted as being for the entrance of the cerebral branches of the internal carotid arteries are positioned on the lateral surfaces of the parabasisphenoid – one of the few indications ( Gower & Weber, 1998; Parrish, 1993) that the braincase of UMMP 7473 probably belongs to a crowngroup archosaur. However, the one important difference between this specimen and the known braincases of P. kirkpatricki lies in the position of the external foramen for the hypoglossal nerve. This is posterior to a very strong subvertical ridge on the lateral surface of the exoccipital. Among crurotarsans, this arrangement is known to certainly occur only in aetosaurians and crocodylomorphs (see below; Gower & Walker, 2002). It is not possible for me to completely rule out that UMMP 7473 might be from an ornithodiran, but this may partly reflect my unfamiliarity with braincase osteology in taxa from that group.

Apart from Chatterjee’s (1985) original figures of the braincase of P. kirkpatricki , two other interpretations have been published ( Chatterjee, 1991: fig. 28b), Parrish, 1993: fig. 7). Chatterjee (1991: fig. 28b) presented a modified version of his previous (1985) reconstruction of the braincase of P. kirkpatricki in lateral view. Sutural relations remained largely the same (except for a smaller prootic contribution to the paroccipital process), but a few changes in detail were made. A small foramen was now indicated positioned above the metotic foramen, this was identified as the posterior tympanic recess ( Chatterjee, 1991: 335, fig. 28b), a pneumatic diverticulum of the tympanic cavity. This is here interpreted as possibly representing the passage of the posterior cerebral/cephalic vein through the braincase (see above). In addition, two further possible pneumatic depressions were accentuated relative to the earlier reconstruction, posterodorsal (not visible in TTUP 9002, pers. obs.) to the trigeminal opening, and in the region of the foramen for the cerebral branch of the internal carotid artery.

Parrish (1993: 300) claimed that P. kirkpatricki possesses a ‘wedgelike, dorsoventrally expanded parasphenoid rostrum’ and that this is a derived feature shared with rauisuchian and crocodylomorph suchians including Lotosaurus and Dibothrosuchus . The exact meaning of ‘wedgelike’ in this context is unclear. This description has also been used to describe the condition seen in crocodilians where the rostrum does not taper proximally, but rather expands dorsoventrally before ending at an abrupt blunt edge. This contrasts with the condition in which the rostrum is longer than it is high and it tapers to a point proximally. This latter condition is probably plesiomorphic for suchians (and archosaurs as a whole) because it is present in erythrosuchids and proterosuchids ( Gower & Sennikov, 1996a; Gower, 1997), Euparkeria capensis ( Welman, 1995; Gower & Weber, 1998) and phytosaurs ( Camp, 1930; Chatterjee, 1978; pers. obs.). The parabasisphenoid rostrum is incomplete in the known specimens of P. kirkpatricki (including that figured by Parrish, 1993: fig. 7) as well as of Dibothrosuchus elaphros ( Wu & Chatterjee, 1993: 68) . The rostrum of the crocodylomorph Sphenosuchus acutus tapers anteriorly in lateral view and is not wedge-shaped ( Walker, 1990).

Parrish (1993: 302) also identified a synapomorphy for his clade Paracrocodylomorpha (comprising Crocodylomorpha + Poposauridae , the latter represented by P. kirkpatricki ): “Foramen at basioccipital/basisphenoid juncture junction... In crocodylomorphs and Postosuchus , this is the exit for the eustachian tubes”. A similar opening/pit (present in many, even noncrown-group archosaurians) is more probably an unossified gap that was plugged with cartilage in life (see review by Gower & Weber, 1998).

Long & Murry (1995) proposed a revision of material referred to P. kirkpatricki . This included referral of additional braincase material to this species. However, these specimens (e.g. UCMP A269/27479; Long & Murry, 1995: fig. 125) are fragmentary and I suggest that their identification must remain in question for the time being.