Thalattosuchia Fraas, 1901

Young, Mark T., Wilberg, Eric W., Johnson, Michela M., Herrera, Yanina, Brandalise, Marco de Andrade, Brignon, Arnaud, Sachs, Sven, Abel, Pascal, Foffa, Davide, Fernández, Marta S., Vignaud, Patrick, Cowgill, Thomas & Brusatte, Stephen L., 2024, The history, systematics, and nomenclature of Thalattosuchia (Archosauria: Crocodylomorpha), Zoological Journal of the Linnean Society 200 (2), pp. 547-617 : 584-587

Thalattosuchia Fraas, 1901 (Zoological Code)

Thalattosuchia Fraas 1901: 410 , converted clade name (PhyloCode).

RegNum registration number 1012.

Etymology

‘Sea crocodiles’. Thalatto- is from the classical Ancient Greek (θᾰ́λᾰ́ττᾰ́, thálatta) for sea. Suchus is the Neo-Latinized form of the Greek Soukhos (σοῦΧος), which appears to have been the name of an individual tamed crocodile that lived in Arsinoite nome, in Ancient Egypt ( Larcher 1844: 286 on Herodotus’ journey through Late Period Ancient Egypt). ‘Sonchis’ was used to refer to the tamed crocodiles that lived in Arsinoite nome ( Larcher 1844: 286), although Larcher also quoted Damascius in stating that the Ancient Egyptians used ‘sonchis’ to refer to a species of crocodile (as two species of crocodile lived in Egypt at that time). Larcher (1844: 286) considered Herodotus’ use of ‘champsa’ to be the generic Egyptian term for crocodile. However, Anonymous (1821: 160) noted that ‘Herodotus must have been a very imperfect master of the Egyptian language. In no instance does he write accurately an Egyptian name’. The suffix - suchus is today used to refer to crocodiles, crocodylian relatives, or crocodylian analogues. The Neo-Latin suffix - ia denotes an abstract noun of feminine grammatical gender.

Geological range

Early Jurassic (Hettangian–Sinemurian) to Early Cretaceous (earliest Aptian) ( von Huene and Maubeuge 1952, 1954, Gasparini 1985, Godefroit 1994, Gasparini et al. 2000, Chiarenza et al. 2015, Sachs et al. 2020, Hicham et al. 2023).

PhyloCode phylogenetic definition

The largest clade within Crocodylomorpha containing Macrospondylus bollensis ( Jäger 1828) and Thalattosuchus superciliosus (Blainville in Eudes-Deslongchamps, 1852), but not Protosuchus richardsoni Brown, 1933 , Notosuchus terrestris Woodward, 1896 , Peirosaurus tormini Price, 1955 , Anteophthalmosuchus hooleyi Salisbury and Naish, 2011 , Deltasuchus motherali Adams et al., 2017 , Pholidosaurus schaumburgensis von Meyer, 1841 , Dyrosaurus phosphaticus ( Thomas, 1893) , and Crocodylus niloticus ( Laurenti, 1768) .

Reference phylogeny

Fig. 3 .

PhyloCode diagnostic apomorphies

Crocodylomorphs with the following unique combination of characters (11): loss of planar skull table morphology (reversal to non-crocodyliform and non-hallopodid condition) (185.0); lack of longitudinal groove on the squamosal (shared with Iharkutosuchus , reversal from the crocodyliform condition) (250.0); temporal bars oblique and anteriorly convergent, giving the skull roof a trapezoidal outline in dorsal view (shared with Dyrosauroidea, paralligatorids, and numerous other acquisitions within Notosuchia and Neosuchia ) (255: 1); paroccipital process with a proportionally robust, thickened lateral/ventrolateral edge (shared with Goniopholididae, Tethysuchia , Paralligatoridae , Bernissartiidae , Hylaeochampsidae , and Allodaposuchidae ) (416: 1); large region of the exoccipitals exposed ventral to the paroccipital processes (shared with Protosuchidae , Notochampsidae , Notosuchidae , and Sphagesauridae ) (433: 1); proötic exposed in dorsal view within the supratemporal fossa (reversal to the non-crocodyliform condition) (443.0); lack of contact between the laterosphenoid and quadrate (reversal to non-crocodyliform and non-hallopodid condition) (470.0); quadrate anteroventral process free of bony attachment along its anteromedial surface, but contacts the pterygoid ventrally (472.2); tympanic membrane fossa restricted to the posterolateral corner of the skull (477.1); retroarticular process is triangular in shape (shared with the notosuchian Araripesuchus tsangatsangana , the goniopholidid Sunosuchus , the pholidosaurids Sarcosuchus and Terminonaris , the paralligatorid Kansajasuchus, Bernissartiidae , Susisuchidae , and Eusuchia ) (571: 0); absence of laminar presacral hypapophyses (optimization of character unclear) (688.0).

Potentially diagnostic characters

The following 46 characters are diagnostic for Plagiophthalmosu chus + Neothalattosuchia but cannot be scored for Turnersuchus . Therefore all, or a subset thereof, could be diagnostic for Thalattosuchia . However, without more complete specimens from the pre-Toarcian, these characters could also define the Pla giophthalmosuchus + Neothalattosuchia clade.

Maxilla conspicuously ornamented with pits and grooves (18.3); anteroposterior elongation (36.1) and sub-horizontal orientation of the pituitary fossa chamber (37.1); enlargement of the cerebral carotid and orbital vasculature canals (43.1); hypertrophy of the transverse dural venous sinuses (44.2); stapedial-temporoorbital vasculature canals enlarged (45.1); dorsal alveolar canals posteriorly start off being medial to the maxillary alveoli, and shift to a dorsomedial position (48.1); internal antorbital sinus diverticula do not invade the maxilla surrounding the posterior alveoli (i.e. the cavity is a simple tube) (50.1); the palatal processes of the maxillae are apneumatic (53.0); pharyngotympanic diverticula reduced and largely confluent (66.2); subtympanic portion of the infundibular diverticulum cavity reduced/confluent with the pharyngotympanic sinus (68.2); absenceofsubtympanicforamina(sharedwithearly diverging crocodylomorphs and dyrosaurids) (70.0); absence of cavities for the quadrate diverticula (72.0) along with the absence of foramina aërum (73.0); otoccipital diverticula restricted to the ventral half of the otoccipital (78.1); proötic diverticula reduced (81.1); absence of intertympanic diverticula cavities (84.0); presence of intranarial fossa (111.1); premaxilla contributes less than 25% of total rostrum length (reversal in Metriorhynchidae ) (120.0); premaxilla and nasals not in contact in dorsal view (shared with Meridiosaurus and Gavialis ) (134.1–3); nasal anterior margins terminate posterior to the third maxillary alveoli (shared with Pholidosauridae ) (136.0); anterior margins of the nasals are triangular, with the lateral margins being strongly confluent anteriorly (shared with Notosuchia) (137.0); maxillae ventral margins are straight (shared with Pholidosauridae , Dyrosauridae , and Gavialoidea) (146.0); antorbital cavity closer to the alveolar margin than to the orbit, or equidistant (177.1); frontoparietal fossa forms a ‘flat platform’ on the dorsal surface of the frontal (reversal to non-crocodyliform and non-hallopodid condition) (191.1); postorbital is longer than the squamosal (shared with Sphenosuchus and Almadasuchus ) (240.1); postorbital forms at least 50% of the supratemporal bar (shared with Sphenosuchus , Almadasuchus , most European goniopholidids, some dyrosaurids, and Crocodylidae ) (241.1); upper temporal bar ventrally displaced relative to the intertemporal bar, coincident with the horizontal plane, and rotated, with the dorsal surface exposed laterally and the ventral surface medially (257.1); absence of palpebrals (264.0); lateral surface of the postorbital bar only formed by the postorbital, with the jugal exposed on the medial surface of the bar (315.1); in the postorbital bar, the postorbital is lateral to the jugal (shared with Mahajangasuchus ) (316.2); dorsal end of the postorbital bar broadens dorsally, being continuous with the dorsal part of the postorbital (reversal in Metriorhynchidae ) (318.0); anterior border of the suborbital fenestra forms a sharp angle, forming a notch (339.1); presence of paired longitudinal palatal grooves on the maxillae and palatines (356.1); increase in size of the carotid foramina (reversal in the unnamed Chinese teleosauroid, Indosinosuchus potamosiamensis , and Machimosaurus buffetauti ) (428.1); cranioquadrate canal enclosed by the squamosal laterally, quadrate ventrally, and the exoccipital medially, posteriorly and partly ventrally (482.1); squamosal descending process at least partially separates the cranioquadrate canal and the external auditory meatus (483.1); presence of a coronoid process on the surangular (shared with Tomistoma and Iharkutosuchus , process often overlooked in teleosauroids) (546.1); retroarticular process divided into medial and lateral portions (shared with Dyrosauridae and Crocodylia) (572.1); diastema between the fourth and fifth dentary alveoli (shared with Sarcosuchus ; reversal in Dakosaurus + Plesiosuchus subclade) (622.1); transverse processes of sacral vertebra 1 are arched lateroventrally (729.1); ‘fan’-shaped coracoids (741.1– 2); forelimbs (humerus + ulna + metacarpal III) are between 20–35% of total trunk (presacral vertebra minus the atlas-axis) length (shared with Gavialis and Pietraroiasuchus ) (750.3); sigmoidal femur forming a shallow ‘S’-shape (811.1); fourth trochanter ridge absent, instead a flattened rugose area is present (819.0); no appendicular osteoderms (character poorly sampled in Crocodyliformes ) (867.0).

Composition

The early diverging taxa Turnersuchus hingleyae and Plagiophthalmosuchus gracilirostris , as well as the subclade Neothalattosuchia (which is composed of Teleosauroidea and Metriorhynchoidea ).

Comments

Authorship: The nomen Thalattosuchia was first used by Fraas (1901), and Fraas (1901) is the nominal authority of the clade under both nomenclatural codes (note, when higher clades are written in italics, we are referring to the PhyloCode variant). As Thalattosuchia is above the family-group, only Articles 1–4, 7–10, 11.1–11.3, 14, 27, 28, and 32.5.2.5 of the Zoological Code apply (as per Article 1.2.2).

Prior phylogenetic definition: Young and Andrade (2009) defined Thalattosuchia as the most inclusive clade consisting of Teleosaurus cadomensis ( Lamouroux, 1820) and Metriorhynchus geoffroyii von Meyer, 1832 , but not Pholidosaurus s chaumburgensis von Meyer, 1841, Goniopholis crassidens Owen, 1842 , or Dyrosaurus phosphaticus ( Thomas, 1893) . Note, here the internal specifiers for Thalattosuchia , and the external specifier for Goniopholididae , were changed from type species to species that are better preserved and more accessible to workers. We have also expanded the number of external specifiers to include a protosuchid, a eunotosuchian, a peirosaurid (sebecian), a paluxysuchid, and an extant representative. This ensures that our intended definition of Thalattosuchia does not ‘favour’ one phylogenetic positional hypothesis of the clade over another.

Typological errors: The nomen Thalattosuchia is frequently misspelt. The most common typological error is the ‘double l’ instead of the ‘double t’, i.e. ‘Thallatosuchia’/ ‘thallatosuchian’ (e.g. De Beer 1928: 488). This is followed by using both the ‘double l’ and the ‘double t’, i.e. ‘Thallattosuchia’/ ‘thallattosuchian’ (note these appear in recent papers where Thalattosuchia is correctly spelt, so appear to be typological errors). A less common typological error is ‘Thalassosuchia’, where the stem Thalasso - from Laconian Greek is used, e.g. van de Wiele (1905: 102, 108) and Sauvage (1916: 48).

Discussion: Thalattosuchians are remarkably diagnostic. Unfortunately, Turnersuchus hingleyae cannot be scored for 46 characters that also unite Plagiophthalmosuchus gracilirostris and other thalattosuchians (due to the incomplete preservation of its holotype). Of the 11 characters that can be scored for Turnersuchus , nine are cranial, one is from the lower jaw, and one is vertebral. Of the 46 characters that cannot be scored for Turnersuchus , 37 are from the cranium, three are from the lower jaw, and only six are from the postcranial skeleton (one vertebral character, one pelvic girdle character, one forelimb character, two hindlimb characters, and an osteoderm character).

When we look at the external cranial characters, they relate to the extensive modification to the supratemporal and postorbital bars, the temporal region, the braincase, and the loss of the palpebrals ( Fig. 17 ). There are also numerous internal cranial characters, as thalattosuchians have remarkably apneumatic braincases and rostra, and many hypertrophied vascular canals (see Figs 18 , 19 ; Fernández and Herrera 2009, Fernández et al. 2011, Herrera et al. 2013 a, 2018, Brusatte et al. 2016, Pierce et al. 2017, Schwab et al. 2021, Bowman et al. 2022, Cowgill et al. 2022, Wilberg et al. 2022, Young et al. 2023b).

Hicham et al. (2023) described a specimen from the Hettangian or Sinemurian of Morocco that is remarkably similar to many early Toarcian teleosauroids ( Johnson et al. 2020a). As we show below, the specimen shares two characters with our Teleosauroidea diagnosis, which supports Hicham et al.’s systematization (the other characters cannot be scored due to preservation). This, therefore, extends the confirmed geological range of Teleosauroidea , and Thalattosuchia , into the Hettangian– Sinemurian. As such, the teleosauroid–metriorhynchoid split is far older than previously thought, occurring either within the first 5 million years of the Jurassic or during the Late Triassic. This also makes the hypothesis that thalattosuchians are noncrocodyliforms more plausible (for more details, see: Wilberg 2015 b, Wilberg et al. 2023)—as this hypothesis requires thalattosuchians to have been present in the middle Norian (the age of the oldest known protosuchid crocodyliforms; e.g. see Martínez et al. 2019 and the references therein).

Wilberg et al. (2023) presented two Bayesian time-scaling analyses to estimate the origination time for Thalattosuchia , one based on the phylogenetic dataset from Herrera et al. (2021a) and another a modification of the Wilberg et al. (2019) dataset. The 95% highest posterior density (HPD) for the Herrera et al. analysis spanned the Norian to Pliensbachian, with the median age within the Sinemurian, while the 95% HPD for the Wilberg et al. dataset spanned the Norian to Hettangian, with the median age within the Norian. Given that the newly discovered teleosauroid from the Hettangian–Sinemurian ( Hicham et al. 2023) was not included in either dataset, origination-time estimates that include the Sinemurian and Pliensbachian can be rejected. This only strengthens a Late Triassic, and Norian in particular, origination time for Thalattosuchia .

The presence of teleosauroids prior to the Toarcian should not come as a surprise, as by the early Toarcian there had been an extensive diversification of this clade. With the split between Teleosauridae and Machimosauridae having had occurred, and within Teleosauridae the subfamily Teleosaurinae was already distinct ( Johnson et al. 2020a). Therefore, Teleosauroidea , and its major subclades must have been present in the Pliensbachian, and by extension Metriorhynchoidea . Furthermore, if our phylogenetic analyses are correct, and the Moroccan specimen from the Hettangian–Sinemurian ( Hicham et al. 2023) is indeed an early diverging machimosaurid, then the origination times of Teleosauroidea and Metriorhynchoidea could go back into the Triassic [supporting the time-scaling analyses of Wilberg et al. (2023)].

Our limited understanding of the origins of Thalattosuchia is mirrored by our poor understanding of their extinction. Teleosauroids are clearly present in the Late Jurassic , with both teleosaurids and machimosaurids known from the Tithonian of Western Europe (e.g. Johnson et al. 2020 a, Young and Sachs 2021). However , our knowledge of teleosauroids outside of Europe remains limited, particularly near the end of the Jurassic. Two specimens are used as evidence that teleosauroids continued into the Early Cretaceous : Machimosaurus rex Fanti et al., 2016 from the Hauterivian of Tunisia, and an indeterminate specimen from the late Barremian of Colombia ( Cortés et al. 2019). While there is disagreement over the exact age of Machimosaurus rex ( Cortés et al. 2019, Martin et al. 2019, Young and Sachs 2021), the specimen described by Cortés et al. (2019) does appear to be evidence of teleosauroid survival into the Early Cretaceous ( Young and Sachs 2021). With a body length estimate of 9.6 m ( Cortés et al. 2019), the Barremian teleosauroid is by far the largest known thalattosuchian and rivals the giant pholidosaurid Sarcosuchus imperator in length (for the revised body length of the latter, see: O’Brien et al. 2019). This specimen is an excellent example of how the Eurocentrism of thalattosuchian palaeobiology limits our understanding of thalattosuchian diversity and geological range.

When and why Metriorhynchidae became extinct also remains elusive. The youngest known metriorhynchid fossil is an isolated tooth crown from the earliest Aptian of Sicily referred to Plesiosuchina indet. by Chiarenza et al. (2015). The referral of this tooth crown to Plesiosuchina was disputed by Fischer et al. (2015), who used a superficial similarity-based approach to suggest the tooth crown could belong to a brachauchenine pliosaurid. However, as pointed out by Sachs et al. (2020), the list of apomorphies Chiarenza et al. (2015) used to refer the tooth crown to Metriorhynchidae , and Plesiosuchina in particular, was never addressed by Fischer et al. (2015). Moreover, Sachs et al. (2020) suggested that ‘ Fischer et al. (2015) inadvertently strengthened the referral of the Sicilian tooth to Plesiosuchina (as Cretaceous pliosaurids did not seem to evolve the apomorphies seen in metriorhynchids), not the reverse’. Therefore, this specimen remains the youngest known metriorhynchid, and thalattosuchian. However, our understanding of post-Valanginian metriorhynchids remains exceptionally poor and we cannot make any definitive inferences on their diversity or biology.

In spite of the recent discoveries of thalattosuchians outside of Europe (or particularly the Western European countries of Germany, France, and the UK) in the earliest Jurassic and in the post-Valanginian Early Cretaceous (e.g. Fanti et al. 2016, Cortés et al. 2019, Hicham et al. 2023) our understanding of Thalattosuchia remains Eurocentric. We caution workers from making grand palaeobiological or biogeographical hypotheses for Thalattosuchia , given we do not know when the clade may have originated (or where) and when the clades that survived into the Cretaceous may have become extinct (or where). The thalattosuchian fossil record outside of Western Europe, and Europe itself, is still patchy and poorly sampled—although Chile and Argentina are increasingly becoming better sampled (e.g. Gasparini 1973, 1980, 1985, Gasparini and Dellapé 1976, Gasparini and Chong 1977, Vignaud and Gasparini 1996, Gasparini et al. 2000, 2006, 2008, Fernández and Herrera 2009, 2022, Herrera et al. 2009, 2013a –c, 2015, 2021b, Pol and Gasparini 2009, Fernández et al. 2011, 2019). Given that thalattosuchians were present in northern Africa, western Europe, and South America by the Sinemurian ( von Huene and Maubeuge 1952, 1954, Gasparini 1985, Godefroit 1994, Gasparini et al. 2000, Hicham et al. 2023), and in China, India, and Madagascar by at least the Toarcian ( Owen 1852, Buffetaut et al. 1981, Johnson et al. 2020a), so much of their fossil record is simply unknown.

Figures

Figure 3. The equal weights, no constraints strict consensus topology (C0). Length = 4164, CI = 0.303, RI = 0.831, RC = 0.251, HI = 0.697, number of most parsimonious cladograms hit the hold limit.Phylogeny pruned to highlight Thalattosuchia (see Supporting Information, Appendix S2, Fig. S1 for the entire topology). A, Crocodylomorpha; B, Crocodyliformes; C, Thalattosuchia; D, Neothalattosuchia; E, Teleosauridae; F, Teleosaurinae; G, Aeolodontini; H, Machimosauridae; I, Machimosaurinae; J, Machimosaurini; K, Euthalattosuchia; L, Metriorhynchidae; M, Metriorhynchinae; N, Rhacheosaurini; O, Geosaurinae; P, Geosaurini. Green lines highlight non-neothalattosuchian thalattosuchians, orange lines highlight teleosauroid thalattosuchians, and blue lines highlight metriorhynchoid thalattosuchians.Images from PhyloPics.org, with the teleosauroid by Scott Hartman, and the metriorhynchid by Dmitry Bogdanov (vectorized by T. Michael Keesey), the only change made was to their colouration, and their license is: https://creativecommons.org/licenses/by/3.0/.

Figure 17. Comparative plate of thalattosuchian skulls in lateral view.A, non-neothalattosuchian Plagiophthalmosuchus gracilirostris, paratype NHMUK PV OR 15500; B, the teleosaurid teleosauroid Mystriosaurus laurillardi, referred specimen NHMUK PV OR 14781 (holotype of Steneosaurus brevior); C, the early diverging metriorhynchoid Pelagosaurus typus, referred specimen NHMUK PV OR 32599; D the metriorhynchid ‘Metriorhynchus’ brachyrhynchus, referred specimen NHMUK PV R 3804 (holotype of ‘Met’. cultridens). Highlighting that only teleosauroids had dorsolaterally oriented orbits (convergent with extant crocodylians), and that orbit size dramatically increased within Metriorhynchoidea.

Figure 18. Comparative plate of extant crocodylian and thalattosuchian internal braincase anatomy.The extant gavialid Gavialis gangeticus, referred specimen UF-herp-118998, in lateral view (A) and in occipital view (B). The non-neothalattosuchian Plagiophthalmosuchus cf. gracilirostris, NHMUK PV OR 33095, in lateral view (C) and in occipital view (D). The early diverging metriorhynchoid Pelagosaurus typus, referred specimen NHMUK PV OR 32599, in lateral view (E) and in occipital view (F). The rhacheosaurin metriorhynchid Cricosaurus araucanensis, holotype MLP 72-IV-7-1, in lateral view (G) and in occipital view (H). All scale bars equal 5 cm. Highlighting the reduction

Figure 19. Comparative plate of extant crocodylian and thalattosuchian internal rostral anatomy (A–G in dorsal view and H–I in lateral view).A, extant gavialid Gavialis gangeticus, referred specimen UF-herp-118998; B, non-neothalattosuchian Plagiophthalmosuchus gracilirostris, paratype NHMUK PV OR 15500; C, extant gavialid Tomistoma schlegelii, referred specimen USNM 211322; D, early diverging metriorhynchoid Pelagosaurus typus, referred specimen NHMUK PV OR 32599; E, extant crocodylid Crocodylus rhombifer, referred specimen MNB AB50.0171; F, H, rhacheosaurin metriorhynchid Cricosaurus araucanensis, holotype MLP 72-IV-7-1; G, I, geosaurin metriorhynchid Torvoneustes coryphaeus, holotype MJML K1863. All scale bars equal 5 cm. Highlighting how apneumatic thalattosuchian rostra were, that the preorbital fenestrae were decoupled from the paranasal sinus system, and how the enlarged salt glands expanded the posterior nasal cavity in metriorhynchoids. For details on the computed tomography scans and internal descriptions, see: Cowgill et al. 2023. Abbreviations:ant fen, antorbital fenestra; antorb, antorbital sinus of the paranasal sinus system; exp, expansion of the nasal cavity; max, maxillary diverticulum of the paranasal sinus system; max cec, maxillary cecal recess; nc, nasal cavity; pre fen, preorbital fenestra; post, postvestibular sinus.