The forgotten clade: the REAL proximal ancestors to Dinosauria

Ignored by Baron et al. 2017, and everybody else
the Junggarsuchus clade (including Pseudhesperosuchus, Carnufex and Trialestes in order of increasing quadrupedality, Figs. 1–4) nests as the proximal ancestors to Herrerasaurus (Fig. 1) and the rest of the Dinosauria (Fig. 5) in the large reptile tree (LRT). That cladogram tests a wider gamut of taxa in greater detail than any other reptile cladogram ever published, attempting to not overlook anything. The Junggarsuchia is a sister clade to the Crocodylomorpha with both arising from a taxon near Lewisuchus (Fig. 1). Traditional paleontology (see Wikipedia) nests this largely ignored clade with the sphenosuchian crocodylomorphs (Fig. 4)… and for two good reasons!

Figure 1. Members of the Junggarsuchus clade were derived from a sister to the basal crocodylomorph, Lewisuchus and produced one line that includes Pseudhesperosuchus and Trialestes. The other line produced dinosaurs. These taxa are shown to scale. Note the evolution from a bipedal configuration to a quadrupedal stance.

Figure 1. Members of the Junggarsuchus clade were derived from a sister to the basal crocodylomorph, Lewisuchus and produced one line that includes Pseudhesperosuchus and Trialestes. The other line produced dinosaurs. These taxa are shown to scale. Note the evolution from a bipedal configuration to a quadrupedal stance.

One: Paleontologists never seem to include Dinosauria
in their smaller gamut croc analyses because they’re looking at crocs!~. So once again, taxon exclusion is holding some workers back from seeing ‘the big picture’. ReptileEvolution.com and the blog you are currently reading is all about examining ‘the big picture.’

Figure 2. Skulls of the Junggarsuchus clade not to scale. Herrerasaurus is the basalmost dinosaur.

Figure 2. Skulls of the Junggarsuchus clade not to scale. Herrerasaurus is the basalmost dinosaur, closely related to Junggarsuchus.

Two: Junggarsuchians ALSO have elongate proximal wrist bones
Elongate proximal carpals are found in both sphenosuchian crocs and derived members of the Junggarsuchus clade. Paleontolgists wrongly assumed such odd wrist bones were homologous. It’s an easy mistake to make. However, the LRT makes clear that intervening taxa, including Junggarsuchus, do not have elongate wrist bones.

Among taxa that preserve the manus,
(Fig. 3) it is Junggarsuchus that nests closest to Herrerasaurus and the Dinosauria.

Figure 3. Hands of Lewisuchus, Herrerasaurus, Junggarsuchus, Pseudhesperosuchus and Trialestes. The proximal carpals (radiale and ulnare) were elongate by convergence with a line of crocodylomorphs. This has confused paleontologists and mentally removed them from possible ancestry to the Dinosauria. Note the very short proximal carpals in Junggarsuchus.

Figure 3. Hands of Lewisuchus, Herrerasaurus, Junggarsuchus, Pseudhesperosuchus and Trialestes. The proximal carpals (radiale and ulnare) were elongate by convergence with a line of crocodylomorphs. This has confused paleontologists and mentally removed them from possible ancestry to the Dinosauria. Note the very short proximal carpals in Junggarsuchus.

Like the basal members of the Crocodylomorpha
the Junggarsuchus clade (the Prodinosauria here) transition from bipedal basal members to quadrupedal derived members, with the longest forelimbs belonging to the most derived member, Trialestes (Fig. 3). Distinct from the others and contra the original interpretation, I think Trialestes may have had a larger ulnare than radiale, to match its larger ulna.

Figure 4. Crocodylomorph manus and carpus samples including Terrestrisuchus, Erpetosuchus, Hesperosuchus and Dibothrosuchus along with Scleromochlus documenting the elongate radiale and ulnare on derived taxa. Ticinosuchus is the closest example of an ancestral/plesiomorphic manus in the LRT.

Figure 4. Crocodylomorph manus and carpus samples including Terrestrisuchus, Erpetosuchus, Hesperosuchus and Dibothrosuchus along with Scleromochlus documenting the elongate radiale and ulnare on derived taxa. Ticinosuchus is the closest example of an ancestral/plesiomorphic manus in the LRT.

Let’s not forget
PVL 4597 (Fig. 6) which was mistakenly considered a specimen of Gracilisuchus by (Lecuona and Desojo 2011), but under phylogenetic analysis in the LRT, still nests as the proximal outgroup to Herrerasaurus. It is tiny specimen, supporting the hypothesis of phylogenetic miniaturization at clade origin. And it retains a small proximally oriented calcaneal tuber, as found in other Junggarsuchians.

Figure 1. Subset of the LRT focusing on the Archosauria (Crocodylomorpha + Dinosauria and kin). Gray areas document specimens with elongate proximal carpals (radiale and ulnare).

Figure 5. Subset of the LRT focusing on the Archosauria (Crocodylomorpha + Dinosauria and kin). Gray areas document specimens with elongate proximal carpals (radiale and ulnare).

We looked at
phylogenetic miniaturization at the origin of several pterosaur clades. Well, it happens here too, at the base of the Dinosauria (Fig. 1) with PVL 4597 (Fig. 6), easily overlooked, easily mistaken for something else.

One should not ‘choose’ outgroup taxa
based on paradigm, tradition, guessing, convenience or opinion. Rather outgroup taxa should ‘choose themselves’ based on rigorous testing of a large gamut of outgroup candidates in phylogenetic analysis. To minimize selection bias, the LRT provides 858 outgroup taxa the opportunity to nest close to dinosaurs.

Figure 6. The closest known taxa to the Dinosauria, PVL 4597, is a tiny taxon (phylogenetic miniaturization) with erect hind limbs, a large and deep pelvis and a tiny calcaneal tuber.

Figure 6. The closest known taxa to the Dinosauria, PVL 4597, is a tiny taxon (phylogenetic miniaturization) with erect hind limbs, a large and deep pelvis and a tiny calcaneal tuber.

 

References
Baron MG, Norman DB, Barrett PM 2017. A new hypothesis of dinosaur relationships and early dinosaur evolution. Nature 543:501–506.
Bonaparte JF 1969. 
Dos nuevos “faunas” de reptiles triásicos de Argentina. Gondwana Stratigraphy. Paris: UNESCO. pp. 283–306.
Butler RJ. et al. 2014. New clade of enigmatic early archosaurs yields insights into early pseudosuchian phylogeny and the biogeography of the archosaur radiation. BMC Evol. Biol. 14, 128.
Clark JM et al. 2000. A new specimen of Hesperosuchus agilis from the Upper Triassic of New Mexico and the interrelationships of basal crocodylomorph archosaurs. Journal of Vertebrate Paleontology 20 (4): 683–704.
doi:10.1671/0272-4634(2000)020[0683:ANSOHA]2.0.CO;2.
Clark JM, Xu X, Forster CA and Wang Y 2004. A Middle Jurassic ‘sphenosuchian’ from China and the origin of the crocodilian skull. Nature 430:1021-1024.
Lecuona A and Desojo, JB 2011. Hind limb osteology of Gracilisuchus stipanicicorum(Archosauria: Pseudosuchia). Earth and Environmental Science Transactions of the Royal Society of Edinburgh 102 (2): 105–128.
Nesbitt SJ 2011. The early evolution of archosaurs: relationship and the origin ofmajor clades. Bull. Amer. Mus. Nat. Hist. 352, 1–292.
Novas FE 1994. New information on the systematics and postcranial skeleton of Herrerasaurus ischigualastensis (Theropoda: Herrerasauridae) from the Ischigualasto
Reig OA 1963. La presencia de dinosaurios saurisquios en los “Estratos de Ischigualasto” (Mesotriásico Superior) de las provincias de San Juan y La Rioja (República Argentina). Ameghiniana 3: 3-20.
Sereno PC and Novas FE 1993. The skull and neck of the basal theropod Herrerasaurusischigualastensis. Journal of Vertebrate Paleontology 13: 451-476. doi:10.1080/02724634.1994.10011525.
Zanno LE, Drymala S, Nesbitt SJ and Schneider VP 2015. Early Crocodylomorph increases top tier predator diversity during rise of dinosaurs. Scientific Reports 5:9276 DOI: 10.1038/srep09276.

wiki/Pseudhesperosuchus
wiki/Junggarsuchus
wiki/Carnufex
wiki/Herrerasaurus
wiki/Sanjuansaurus

 

The cervical/dorsal transition in Herrerasaurus

Figure 1. Lewisuchus cervical/dorsal transition at top photo and the same for Herrerasaurus drawings, including a foreword shift of the pectoral girdle in a 2-frame GIF movie. The cervical ribs are imagined.

Figure 1. Lewisuchus cervical/dorsal transition at top photo and the same for Herrerasaurus drawings, including a foreword shift of the pectoral girdle in a 2-frame GIF movie. The cervical ribs are imagined in the drawing. See text for details. Drawings from Novas and Sereno 1994.

In their report on the basal dinosaur Herrerasaurus,
Sereno and Novas 1993 reported, “The cervical column (Fig. 1) was preserved in articulation with the skull. The anterior cervical vertebrae are better preserved than the posterior cervical vertebrae, and nearly all the ribs are lacking.”

“Because the cervical-dorsal transition in vertebrae or ribs is not preserved, we regard the first ten presacral vertebrae as cervical vertebrae, based on the condition in other basal dinosaurians.”

Despite their assessment, 
Novas and Sereno appear to have reconstructed Herrerasaurus with no more than six or possibly seven cervicals (Fig. 1, original).  With phylogenetic scoring at issue, a deeper look was warranted.

Novas and Sereno 1993 considered Herrerasaurus
a member of the ‘Ornithodira’ thus related to pterosaurs and Lagerpeton. That hypothesis is not supported by the present study.

By contrast,
in the large reptile tree (LRT) Herrerasaurus arises from another list of taxa, including Junggarsuchus, Pseudhesperosuchus, LewisuchusTurfanosuchus and further distantly Decuriasuchus. Bittencourt et al. 2014 identify seven cervicals in Lewisuchus (Fig. 1) with the eighth having ribs descending into the torso. Seven is a number common to crocodylomorpha* and Lewisuchus nests at its base. Turfanosuchus and Decuriasuchus each have eight cervicals, a plesiomorphic number going back at least to basal archosauriformes, like Proterosuchus and to basal diapsids, like Petrolacosaurus.

Based on available data,
Herrerasaurus had but seven cervicals as a basal dinosaur. Based on data from Tawa, Marasuchus, Eodromaeus and Eoraptor, all slightly more derived basal dinosaurs had 9 or 10.

* Among Crocodylomorpha, Scleromochlus has six cervicals, Gracilisuchus, Litargoschus and Terrerstrisuchus have eight, as do modern crocs and their relatives by convergence. Intervening taxa often have seven.

References
Bittencourt JS, Arcucci AB, Maricano CA and Langer MC 2014. Osteology of the Middle Triassic archosaur Lewisuchus admixtus Romer (Chañares Formation, Argentina) its inclusivity, and relationships amongst early dinosauromorphs. Journal of Systematic Palaeontology. Published online: 31 Mar 201. DOI:10.1080/14772019.2013.878758
Nesbitt SJ. et al. 2010. Ecologically distinct dinosaurian sister group shows early diversification of Ornithodira. Nature 464(7285):95-8
Novas FE 1994. New information on the systematics and postcranial skeleton of Herrerasaurus ischigualastensis (Theropoda: Herrerasauridae) from the Ischigualasto Romer AS 1972. The Chañares (Argentina) Triassic reptile fauna; XIV, Lewisuchusadmixtus, gen. et sp. nov., a further thecodont from the Chañares beds. Breviora 390:1-13
Sereno PC and Novas FE 1993. The skull and neck of the basal theropod Herrerasaurusischigualastensis. Journal of Vertebrate Paleontology 13: 451-476. doi:10.1080/02724634.1994.10011525.

wiki/Lewisuchus
wiki/Herrerasaurus

 

Return of the short-face Gracilisuchus MCZ 4116

Earlier we looked at the MCZ 4116 specimen attributed to Gracilisuchus (Fig. 1).

Figure 1. MCZ 4116 a short-faced Gracilisuchus compared to the holotype with a longer face.

Figure 1. MCZ 4116 a short-faced Gracilisuchus compared to the holotype with a longer face. These two nest as sister taxa at the base of the Archosauria.

 

Gracilisuchus (Romer 1972) 
nests at the base of the Archosauria in the large reptile tree. Scleromochlus and Saltopus are sister taxa. So are these short-faced specimens (Fig. 1), MCZ 4116 and 4117 (Brinkman 1981). That short rostrum looks juvenile, but note these specimens are not smaller than the holotype (Fig.1). Romer and Parrish (year?) restored the snout tip with a Gracilisuchus-like big round nasal and a very short, transverse premaxilla. As an option, I just followed existing contours and added a premaxilla similar in length to the holotype.

Could this be a juvenile of a much larger adult?

Gracilisuchus

Figure 2. A basal archosaur, Gracilisuchus.

References
Brinkman D 1981. The origin of the crocodiloid tarsi and the interrelationships of thecodontian archosaurs. Breviora 464: 1–23.
Romer AS 1972. 
The Chañares (Argentina) Triassic reptile fauna. An early ornithosuchid pseudosuchian, Gracilisuchus stipanicicorum, gen. et sp. nov. Breviora 389:1-24.

New clade of enigmatic early archosaurs? No.

Updated one day after publication. The taxa come from the Supp. Data, most not shown in the greatly simplified chronological cladogram.

Recently, Butler et al. (2014)
recovered a “new clade of enigmatic early archosaurs” comprised of Yonghesuchus, Gracilisuchus and Turfanosuchus.

Figure 1. Do these taxa nest in a single clade? No. Turfanosuchus, Gracilisuchus and Yonghesuchus. Each nests more closely with other taxa.

Figure 1. Do these taxa nest in a single clade? No. Turfanosuchus, Gracilisuchus and Yonghesuchus. Each nests more closely with other taxa. Yonghesuchus nests with the crocodylomorph Dromicosuchus. Gracilisuchus nests with Saltopus as a much more basal crocodylomorph. Turfanosuchus nests at the base of the Poposauridae.

Unfortunately they added the unrelated Mesosuchus (lepidosaur), Vancleavea (thalattosaur) and two pterosaurs (lepidosaurs).

And they did not add the true sisters of Gracilisuchus (Pseudhesperosuchus, Decuriasuchus, Lewisuchus, Saltopus, the SMNS 12591 specimen and Scleromochlus).

Red Flags
In the Butler et al. (2014) tree the following purported sister taxa are all “odd bedfellows” that do not look like one another.

  1. Prolacerta is derived from Mesosuchus (and presumably the rhynchosaurs)
  2. Euparkeria is derived from Tropidosuchus and Chanaresuchus.
  3. Tropidosuchus and Chanaresuchus are derived from Vancleavea.
  4. Vancleavea is derived from Erythrosuchus.
  5. Pterosaurs are derived from parasuchians!!!!!!
  6. Lagerpeton is derived from pterosaurs.
  7. Ornithosuchia is derived from Lewisuchus.
  8. Theropoda is derived from Ornithischia.
  9. Ornithosuchia is a sister to Pterosauria, also derived from Parasuchia.
  10. Revueltosaurus is a sister to the Aetosauria and derived from Ornithosuchia
  11. The new Gracilisuchus clade is derived from Revueltosaurus.
  12. Poposaurus and the poposaurs are derived from Qianosuchus, Xilosuchus and Arizonasaurus
  13. Prestosuchus and the Rauisuchidae is derived from Ticinosuchus.
  14. Hesperosuchus and the Crocodylomorpha are derived from Rauisuchidae.

Say it ain’t so!
As you can see, many of these relationships don’t make sense. Sister taxa share very few traits with one another (pterosauria and parasuchia, is the worst such example). Many relationships are upside down with basal taxa, like theropods, derived from derived taxa, like ornithischia. (M. Mortimer also had this problem a few years ago).

What is needed is a large reptile tree in which basal taxa are basal to derived taxa and all sisters look alike (share most traits). In the large reptile tree, sister taxa look quite a bit like one another. The authors should have cast a critical eye on these results, which are very similar to those of Nesbitt (2011), who also recovered many strange bedfellows.

If I had proposed that pterosaurs arose from parasuchians,
the ridicule would be endless and justified, as it is here. Taxon exclusion seems to be the culprit again, along with the tradition of using previously published matrices, even those riddled with Red Flags and strange bedfellows.

In the large reptile tree, Gracilisuchus nests with the SMNS 12591 specimen, Saltopus and Scleromochlus at the base of the Archosauria. Turfanosuchus nests at the base of the Poposauridae, between Decuriasuchus and the base of the Archosauria, not far from GracilisuchusYonghesuchus, nests between Dromicosuchus and Protosuchus.

And, because this is Science, you can repeat these experiments to see for yourself which taxa share more traits — that make sense.

References
Butler et al. 2014. New clade of enigmatic early archosaurs yields insights into early  pseudosuchian phylogeny and the biogeography of the archosaur radiation. BMC Evolutionary Biology  14:128. doi:10.1186/1471-2148-14-128
Nesbitt SJ 2011. The early evolution of archosaurs: relationships and the origin of major clades. Bulletin of the American Museum of Natural History 352: 292 pp.

Is Nyasasaurus the oldest dinosaur?

Nesbitt et al. (2014) report, “Here, we describe Nyasasaurus parringtoni gen. et sp. nov., which is identified as either the earliest known member of, or the sister–taxon to, Dinosauria. Nyasasaurus possesses a unique combination of dinosaur character states and an elevated growth rate similar to that of definitive early dinosaurs.”

Nesbitt et al. (2012) did not provide a reconstruction. The specimen is only known from a few Middle Triassic (Anisian) vertebrae and a proximal humerus. Here (Fig. 1) is a reconstruction that seems to fit pretty well based on the holotype and referred specimens of Nyasasaurus placed onto the bauplan of a large and slightly larger (for the cervicals) specimen of Turfanosuchus, a basal archosaur at the base of the Poposauridae (now nesting basal to archosaurs in the large reptile tree). This was, what you might call, an early ‘false start’ mimicking the actual rise of the Dinosauria.

Figure 1. Click to enlarge. Nyasasaurus bones placed on an enlargement of Turfanosuchus, a middle Triassic basal archosaur, not a dinosaur. Dinos and crocs all started out as tiny bipeds.

Figure 1. Click to enlarge. Nyasasaurus bones placed on an enlargement of Turfanosuchus, a big. middle Triassic basal archosaur, not a dinosaur. Dinos and crocs all started out as tiny bipeds, also derived from Turfanosuchus, but smaller.

Nesbitt et al. ran their analysis and found Nyasasaurus to nest between 1) Lewisuchus and dinosaurs; 2) basal to Ornithischia; or 3) as the sister to Dilophosaurus, a Jurassic dinosaur. I didn’t run an analysis as my characters would not resolve relationships based on so few parts.

The large reptile tree finds two small specimens, Trialestes and an unnamed specimen incorrectly referred by Lecuona and Desojo (2011) to Gracilisuchus (PVL 1259) at the base of the Dinosauria. The latter is as old as Nyasasaurus, fulfilling chronological predictions.

Nesbitt et al. note a ventrally elongate deltopectoral crest on the humerus, but that assumes a short humerus. The crest is not so elongate if the humerus is a little longer.

Nesbitt et al. note three sacral vertebrae, but basal dinos don’t have three sacrals, only two. Turfanosuchus also has only two, but look at the size difference! Poposaurus, a sister taxon, has five sacrals. So Nyasasaurus is something else. Nesbitt et al. note hyposphene–hypantrum intervertebral articulations in the pre sacral vertebrae. Sorry, not much about that in Turfanosuchus data. In Turfanosuchus, as in Nyasaurus, the cervical vertebrae are laterally concave.

References
Lecuona A and Desojo, JB 2011. Hind limb osteology of Gracilisuchus stipanicicorum (Archosauria: Pseudosuchia). Earth and Environmental Science Transactions of the Royal Society of Edinburgh 102 (2): 105–128.
Nesbitt SJ, Barrett PM, Werning S, Sidor CA and Charig AJ (posthumously) 2012. The oldest dinosaur? A Middle Triassic dinosauriform from Tanzania. Biology Letters 9: 20120949.

 

 

Poposaurs – Topology Shift

Yes, I was wrong. And it’s time to man-up.
Earlier, based on available data, the large reptile tree nested poposaurids with phytodinosaurs and attributed the appearance and growth of the calcaneal tuber in certain poposaurs to convergence with the Crocodylomorpha. So the earlier data recovered poposaurs as dinosaurs with an odd ankle, not crocs with a “massive convergence” with dinosaurs, which is still the widespread hypothesis (see Nesbitt 2011 and others).

That seemed to make sense — except some poposaurs, like Lotosaurus and Asilisaurus (Fig. 1), appeared a little too early in the Triassic. They seemed to be anachronistic, and that can be a red flag.

So going back to the phylogenetic analysis,
I reexamined certain specimens, discovered a few items not originally presented (I trusted original tracings instead of making my own from in situ photos) and I found several bad scores. The newly recovered tree finds poposaurids derived from Turfanosuchus, a taxon that earlier stood alone at the base of the Archosauria (basal to crocs and dinos). Now things seem to make more sense, phylogenetically, chronologically (Fig.1) and morphologically. See if you agree…

Figure 1. Poposauridae revised for 2014. Here they are derived from Turfanosuchus at the base of the Archosauria, just before crocs split from dinos.

Figure 1. Poposauridae revised for 2014. Here they are derived from Turfanosuchus at the base of the Archosauria, just before crocs split from dinos. Among these, only Silesaurus and Asilisaurus lost the calcaneal tuber.

Figure 2. The Euarchosauriformes featuring a new nesting for the Poposauridae.

Figure 2. The Euarchosauriformes featuring a new nesting for the Poposauridae.

Now the odd thing is: 
Poposaurs appear to provide a sort of preview to what would eventually evolve in the Dinosauria itself, likely filling similar niches in earlier strata.

Turfanosuchus and Poposaurus (Fig. 1) were convergent with theropods. The larger Asilisaurus was convergent with sauropodomorphs. The remainder were convergent with various ornithischians, even down to the toothless predentary they shared by convergence. Lotosaurus was a stegosaur mimic. Shuvosaurus was a Dryosaurus mimic. Silesaurus was a Camptosaurus mimic, down to losing the calcaneal tuber. Sacisaurus was a little Agilisaurus mimic. Effigia was still the oddball with those vestigial hands and back-sloped braincase.

So poposaurids are not dinosaurs. They are also not basal to rauisuchidae (contra Nesbitt 2011), but were derived from basal rauisuchia like Decuriasuchus and Vjushkovia. They are the most basal archosaurs. Basal poposaurs were the last common ancestors of crocs and birds. From their basalmost taxon, a sister to little Turfanosuchus, both tiny basal bipedal crocs and tiny bipedal basal dinos evolved.

Size
Poposaurs, in the form of Nyasasaurus, Asilisaurus and Lotosaurus (Fig. 1), were the first archosaurs to evolve substantial size in the Middle Triassic. Crocs and dinos remained small until the late Triassic (mid-Triassic for the basal Herrerasaurus) when they had their great radiation and poposaurs began to fade. This is an unrecognized faunal turnover.

Discovering and correcting errors is what scientists do. 
And I was happy that these new insights appeared.

What took so long?
Inattention to red flags. We should all look more closely at problems. They lead to new insights.

M.M. I hope this helps the cause. And yes, I have made and will make changes to earlier posts on this subject.

References
Brusatte SL, Benton MJ, Desojo JB and Langer MC 2010. The higher-level phylogeny of Archosauria (Tetrapoda: Diapsida), Journal of Systematic Palaeontology, 8:1, 3-47.
Irmis RB, Nesbitt SJ, Padian K, Smith ND, Turner AH, Woody D and Downs A 2007. A Late Triassic dinosauromorph assemblage from New Mexico and the rise of dinosaurs. Science 317 (5836): 358–361. doi:10.1126/science.1143325. PMID 17641198.
Nesbitt SJ 2011. The early evolution of archosaurs: relationships and the origin of major clades. Bulletin of the American Museum of Natural History 352: 292 pp.
Nesbitt SJ, Irmis RB, Parker WG, Smith ND, Turner AH and Rowe T 2009. Hindlimb osteology and distribution of basal dinosauromorphs from the Late Triassic of North America. Journal of Vertebrate Paleontology 29 (2): 498–516. doi:10.1671/039.029.0218

Lewisuchus – new reconstruction

Be wary of the scale bars, but the new data on Lewisuchus by Bittencourt et al. (2014) brings a fresh look to old Romer (1972) data. Here (Fig. 1) the scale of the posterior skull and tibia are pure guesswork until I hear back from the authors on my query.

Figure 1. Lewisuchus from Romer 1972. Tibia and skull do no match published scale bars. Pelvis and femur are hypothetical. Skull above left per Romer. Below left per Bittencourt et al.

Figure 1. Lewisuchus from Romer 1972. Tibia and skull do no match published scale bars. Pelvis and femur are hypothetical. Skull above left per Romer. Below left per Bittencourt et al.

Unfortunately the authors did not include basal archosaurs (Decuriasuchus and Pseudhesperosuchus) in their analysis, but restricted their taxon list to dinos and pre-dinos. Not surprisingly, Lewisuchus nested with Marasuchus. The large reptile tree nests Lewisuchus with Pseudhesperosuchus.

Lewisuchus is the reason why basal crocs and dinos were bipeds. It nests at the base of all archosaurs. The palate is a close match to that of the basal bipedal croc Scleromochlus.

References
Bittencourt JS, Arcucci AB, Maricano CA and Langer MC 2014. Osteology of the Middle Triassic archosaur Lewisuchus admixtus Romer (Chañares Formation, Argentina) its inclusivity, and relationships amongst early dinosauromorphs. Journal of Systematic Palaeontology. Published online: 31 Mar 201. DOI:10.1080/14772019.2013.878758
Romer AS 1972. The Chañares (Argentina) Triassic reptile fauna; XIV, Lewisuchusadmixtus, gen. et sp. nov., a further thecodont from the Chañares beds. Breviora 390:1-13

wiki/Lewisuchus