Thursday, January 14, 2021

Hopeful Pterosaur(omorph)s

Lagerpetid by Rodolfo Nogueira

As we’ve seen, the Triassic was a period filled with all manner of charismatic animals, including reptilian platypuses and whales, praying mantis-like drepanosaurs, semi-aquatic dragons, pseudosuchian dinosaur mimics, and everything in between. We have not, however, in our Triassic adventures, yet directly touched on the typical touchstones of Mesozoic diversity: pterosaurs and dinosaurs. In part, that has been intentional: why focus on dinosaurs when there are so many other wonderful animals to learn about during the Triassic? But another, perhaps more interesting, reason is that dinosaurs and pterosaurs simply weren’t a particularly large or diverse component of Triassic ecosystems until the very end. I would argue that the dinosaur's reign didn’t really begin until the Triassic-Jurassic extinction event removed their pseudosuchian competitors. Pterosaurs, too, diversified during the Jurassic, although I'm not sure if anything was keeping them down before. Anyway, until the Jurassic, these would-be prehistoric paragons were living in the shadows of predatory rauisuchians, herbivorous aetosaurs, giant dicynodonts, and semi-aquatic phytosaurs. No revolution, however, comes from nothing. Dinosaurs and pterosaurs had ancestors too, and in this essay, we’ll focus on one their most distant relatives, the lagerpetids.

Remember that the Archosauria, which today consists of crocodilians and birds, got its start during the Middle Triassic, when those two branches split. The bird line, called the Avemetatarsalia, got its start with Teleocrater and its aphanosaur cousins, who are the earliest known members of that large group. Non-aphanosaur avemetatarsalians are called ornithodirans, which means “bird necks,” named (by Gauthier in 1986) for the common ancestor of pterosaurs and dinosaurs (and all its descendants), which tend to have S-curved necks.

Within Ornithodira, there are two clades: Pterosauromorpha, which unambiguously includes all pterosaurs, and the Dinosauromorpha, which contains a hodgepodge of dinosaur-like animals, including the silesaurids, who we’ve met before, and who may in fact be basal ornithischian dinosaurs, and an animal called Lagosuchus (or Marasuchus) (or Lagosuchus after all). Other players in the ornithodiran story are an animal in a perpetual state of reassessment called Scleromochlus and the topic of today’s essay: lagerpetids.

Lagerpetids are generally small, bipedal animals that would have superficially resembled very small theropods at first glance. While we once thought that all lagerpetids were itty-bitty, that view is changing thanks to recent discoveries. Their geographic range is impressive, having been found in Brazil, Argentina, Texas, Arizona, and Madagascar (so far). Perhaps the most frustrating aspect of this group is they are often known primarily through leg bones, because the leg (and specifically the femur) is the most-often preserved part of the skeleton for some reason. One lagerpetid genus, which includes three species, is known from leg bones alone!

The first lagerpetid ever discovered was Lagerpeton chanarensis from Argentina, named and preliminarily described by prominent paleontologist Alfred Romer in 1971; he also named Lagosuchus talampayensis, who we’ll meet in a future essay, in the same paper. The Lagerpeton fossil in question is a complete, articulated hindlimb (right). Romer referred an additional specimen to Lagerpeton and Lagosuchus the following year based on material preserved on two slabs that also contained a cynodont (Romer, 1972). Sereno & Arcucci (1993) would later clarify that, of the many bones preserved in these slabs, only the fragmentary left and right femora can be unambiguously referred to Lagerpeton--and so the pattern begins. Additional specimens were described in the following years by Bonaparte (1984) and Arcucci (1986). All told, Lagerpeton is known from its hindlimb, pelvis, sacrum, some posterior dorsal vertebrae, and anterior caudal vertebrae. As we’ll see, that’s pretty good for a lagerpetid.

Lagerpeton has a unique femur that will later become diagnostic for the whole family: the head of the femur is notably hook-shaped. Additional unique features include a blade-shaped flange (later called a “pubo-ischiadic plate”) connecting the pubis and ischium, and ankle bones (astragalus & calcaneum) that are partially co-ossified. The foot of Lagerpeton is unique among not just ornithodirans, but reptiles generally: the first toe (dewclaw) is very short, as would be expected, but the second toe is also short and may not have reached the ground while Lagerpeton was walking around (assuming it walked--read on). Instead, the third and fourth toes are quite long--the fourth is a bit longer--and gives Lagerpeton a distinctly didactyl foot.

A few other oddities are worth mentioning: the neural spines of the preserved dorsal vertebrae are angled forward, which is unique among archosaurs, and the pelvis is surprisingly small (especially front-to-back) relative to the length of the hindlimb.

In light of these unique features, Sereno & Arcucci (1993) hypothesize that Lagerpeton was a hopper, or saltator. “Several characters in L. chanarensis,” they write, “in particular, are consistent with saltatory habits, although none are conclusive.” In bipedal saltatorial mammals, anteriorly-angled neural spines on the posterior lumbar vertebrae may provide better leverage for the muscles that support the trunk. Regarding the small pelvis, “there is some suggestion in saltatory rodents that shortening of the distance from pelvic girdle to femur enhances force production during hip extension.” The functionally didactyl foot, too, “may increase the strength of the distal segment of the hind-limb without increasing its weight.” Their reconstruction of the Lagerpeton material is at the left. The saltator hypothesis is a good one for smaller members of the family, but we might question its applicability towards larger lagerpetids. While Romer (1971) considered Lagerpeton (and Lagosuchus) to be “pseudosuchians,” he did recognize that they were both “in some fashion...leading toward the dinosaurs and, particularly, towards the coelurosaurian group of the Saurischia.” It should be noted that Romer’s concept of “pseudosuchians'' was more of generic mish-mash of unrelated archosaur grpi[s instead of the diagnostic, monophyletic group it is today. Furthermore, his conception of dinosaurs was similarly broad. At the time, it was widely thought that different groups of dinosaurs originated from different archosaur “stock.” For example, coelurosaurs had different ancestors than did the “carnosaurs,” or larger theropods. Ornithischians, sauropods, and even basal sauropodormophs like Plateosaurus (“prosauropods”) all had different ancestors, making the “Dinosauria” something of a useless concept.

Today, we utilize more restrictive definitions for these groups, but Romer was basically correct in his reasoning: Lagerpeton really was “leading towards the dinosaurs” (as opposed to, say, crocodiles).

Lagerpeton remained a singular curiosity for more than thirty years years before another lagerpetid was described in 2007 by Randall Irmis and colleagues. That taxon, Dromomeron romeri, is from the Chinle Formation in New Mexico (above). As will become something of a trope for the group, Dromomeron romeri was described from only from leg bones. What’s arguably more important than the discovery of a second lagerpetid (although that’s still significant) is the authors’ finding that it existed alongside a silesaurid (Eucoelophysis) and a true dinosaur (Chindesaurus). “Our investigations,” they write, “have found the same co-occurrences in several other Chinle Formation and Dockum Group localities in Arizona, New Mexico, and Texas” (Irmis et al., 2007, p.361). Additionally, the age ranges of these sites indicate that lagerpetids, silesaurids, and true dinosaurs cohabitated for 15-20 million years. Indeed, Dromomeron is from younger rocks than Lagerpeton, suggesting this was a long-lived group. 

A second species of Dromomeron, D. gregorii, was named two years later for material from Texas’ Dockum Group (Nesbitt et al. 2009) (above). It is also known only from femora, but the authors documented a number of other leg bones, including toe bones, from both it and D. romeri, as well as assigning a number of specimens from other localities as belonging to Dromomeron or Lagerpetidae generally. Intriguingly, D. gregorii is known from a series of six femora which form a growth series. Finally, the authors formally defined the Lagerpetidae for the first time. I will note that Arcucci (1986) named the group “Lagerpetonidae” but he did not provide a phylogenetic definition.

Ricardo Martinez and colleagues (2013) reported a lagerpetid femur fragment from Argentina’s famous Ischigualastro Formation, the “Valley of the Moon,” where one of the earliest dinosaurs, Eoraptor, was discovered. Unfortunately limited to the distal end of the femur, the specimen provides little taxonomic information apart from “here be lagerpetids.”

2016 was a banner year for lagerpetids. Martinez et al. named a third species of Dromomeron, D. gigas, from Ischigualastro (left). The holotype is a single femur preserved in two pieces--a small section of the mid-shaft is missing. However, it is extremely similar to Dromomeron specimens from the United States. D. gigas was quite a bit larger than any other lagerpetid known at the time, however: With a femur length of 7.5 inches, D. gigas was more than twice the size of Lagerpeton’s 3 inch femur. D. gigas, at least, grew to be as large as some silesaurids.

The second lagerpetid published that year was Ixalerpeton polesinensis, from the Santa Maria Formation of Brazil (Cabriera et al. 2016) (right). It was found in the same quarry as the basalmost sauropodomorph dinosaur, Buriolestes schultzi, and briefly described in the same paper. Ixalerpeton is arguably the most important lagerpetid found so far, as its beautifully-preserved remains include many bones never before described for a lagerpetid, including a complete series of dorsal and sacral vertebrae, the front third of the tail, the back half of the cervical series, parts of the skull roof and braincase, and a scapula and humerus. Unfortunately, these remains have not been fully described. It appears to be quite small, though, with a femur length comparable to Lagerpeton. Once it is fully described, Ixalerpeton will be able to fill in a lot of blanks regarding the anatomy of lagerpetids.

There was one more noteworthy lagerpetid paper published in 2016--Sarıgül reported additional femora from D. gregorii and D. romeri from the lower member of the Dockum Group. Especially in the case of D. romeri, this demonstrates that Dromomeron “existed during most of the Late Triassic” (Sarıgül, 2015). Furthermore, the paper documents the first co-occurrence of the two species of Dromomeron. Previously, the two had been reported from successive rock units and did not overlap in time, but these new discoveries show that the hypothesized replacement of D. gregorii by D. romeri later in the Triassic must be incorrect if the latter had been around the whole time. In fact, D. gregorri appears to have gone extinct long before D. romeri did.

Known lagerpetids from Muller et al. (2018)

Muller et al.
attempted the first in-group phylogeny of lagerpetids in 2018, using data from all six known species and Specimen PVSJ 883 from Martinez et al. (2012). They found that
Lagerpeton was the basalmost member of the group (and also the oldest) while Dromomeron was the most derived, longest-lived, and youngest taxon.

The next year, Garcia et al. (2019) published an extensive description of the fauna of Brazil’s Santa Maria Formation where a variety of dinosauromorphs lived together, including a new (but unnamed) lagerpetid. Although concurrent with the rocks that produced Ixalerpeton, the new lagerpetid--known from a distal femur fragment--does appear to be distinct. The authors also report that, through successive faunal assemblages, lagerpetids, silesaurids, and basal dinosaurs lived together for at least 21 million years, similar to the situation in North America at this time.

As reported by Kammerer et al. (2020), lagerpetids also lived on Madagascar. This team described the smallest lagerpetid yet--little Kongonaphon kely is, like so many of its relatives, known primarily from a femur, but other fragments of the skeleton were also found, including a maxilla bone with several teeth (above). The extremely small size of this animal is evident from the femur, which, at 38 mm long, is almost half the size of Lagerpeton (77.8 mm). Perhaps this tiny individual was a baby or juvenile? To test that idea, the authors put a slice of the femur under the microscope to estimate its age, and found that its growth had slowed and was most likely an adult. They do note, however, that Kongonaphon may have been able to reach larger sizes, but given that they only have one example of it, it’s hard to say.

The teeth of Kongonaphon are the first teeth recorded for a lagerpetid and show wear patterns consistent with an insectivorous diet. Given that lagerpetids tend to be smaller animals, this isn’t terribly surprising, but good to have confirmed.

More evidence for dinosauromorph cohabitation was reported by Beyl et al. (2020). This team looked at a new site in the Santa Maria Formation in New Mexico and documented the animals living there. Among the various reptiles are several surprisingly large individuals of
Dromomeron (species unknown) represented by fragmentary hindlimb elements--as per tradition--and a silesaurid. When scaled against other Dromomeron femora, one individual (NMMNH P-18091) would have been as large or larger than the early theropods Chindesaurus and Coelophysis (above)! “Unless new discoveries provide evidence for a unique lagerpetid ecological niche,” they write, “their large-bodied forms may have competed with small-bodied carnivorous dinosaurs for prey.”

The point, of course, is that dinosauromorphs did not go through periods of replacement, where more derived species simply outcompeted their predecessors. It’s clear that, in many areas, these animals all lived alongside each other, and must have occupied unique ecological roles to avoid direct competition. Intriguingly, lagerpetids occupy a large spectrum of body sizes, from extremely small (Kongonaphon) to the size of early theropods (NMMNH P-18091). One would assume that, as lagerpetids increased in size, the hopping gait advocated by Sereno & Arcucci (1993) would eventually become less favorable unless they switched from a kangaroo rat method to something more akin to wallabies. I will note, however, that most lagerpetids, and certainly the largest of them, are known only from fragmentary hindlimb remains, so any more skeletal adaptations in favor of a saltatory lifestyle may simply be unknown at this time.

For most of their paleontological history, lagerpetids were thought to be closer relatives of dinosaurs than pterosaurs, forming the basalmost group of a clade called Dinosauromorpha. This position has been advocated by Sereno (1991), Sereno & Arcucci (1993), Brusatte et al. (2010), Nesbitt (2011), Ezcurra (2016), Cabreira et al. (2016), and Muller et al. (2018), and I'm sure I'm forgetting some. Nesbitt (2011), however, did note that the ankle structure of lagerpetids more closely resembled the basal pterosaur Dimorphodon than dinosaurs. However, he considered that convergence and did not outweigh all the other features which tied lagerpetids to other dinosauromorphs.

In their description of Kongonaphon, however, Kammerer et al. (2020) recovered lagerpetids in an intriguing, though weakly-supported, position:

“In analyses where Scleromochlus was not included, Lagerpetidae was recovered in its traditional the earliest-diverging clade of dinosauromorphs. When Scleromochlus was included, this genus and Lagerpetidae were both recovered as early-diverging pterosauromorphs (with Scleromochlus as the sister-taxon of Pterosauria).” The authors suggest caution, and that more complete lagerpetid specimens will be required to more fully vet this idea. “At present,” they write, “we...prefer to depict the base of Ornithodira as an unresolved polytomy between Lagerpetidae, Sclermochlus, Pterosauria, and Dinosauriformes.” (Kammerer et al., page 3).

Scleromochlus as a doswellid, by Matt Celeskey
A rundown of the many issues surrounding Scleromochlus is outside the scope of this essay, but for the most part, Dr. Mark Witton has you covered. All I can add is that, since Mark's post, Chris Bennett (2020) has suggested that it is a saltatorial doswellid or something “outside of the clade containing the most recent common ancestor of Erythrosuchidae and Archosauria and all its descendants.” (Bennett, 2020, page 1).

Awaiting a more complete--or at least non-cast specimen--of Scleromochlus, however, may be unnecessary (though still welcome), for Kammerer et al.’s wish for better lagerpetid data was not far off. At the blessed end of 2020, Ezcurra et al. brought a treasure trove of lagerpetid material to light (below), including endocasts from Ixalerpeton and, surprisingly, Dromomeron that robustly supports a basal pterosauromorph position for Lagerpetidae.

Some of the more impressive similarities include:

  • Anterior tip of the dentary in Ixalerpeton and (!) Lagerpeton is toothless and comes to a point, as in early pterosaurs Seazzadactylus, Carniadactylus, and Raetiodactylus (and, as we’ve seen, silesaurids). The dentary also curves downward towards its tip, as in Austriadactylus and Peteinosaurus.

  • Contrary to most archosauriformes, lagerpetids and pterosaurs both have multicusped teeth and lack interdental plates.

  • The endocasts for Ixalerpeton and Dromomeron are similar to, but not as developed as, early pterosaurs. The same can be said for the shape of the inner ears.

  • That pubo-ischiadic plate I mentioned occurs in Lagerpeton, Ixalerpeton, and several early pterosaurs like Austriodraco and Dimorphodon.

  • Lagerpetids and early pterosaurs have hook-shaped femoral heads, and Nesbitt (2011) was right about their ankles, although in isolation, the ankle features themselves provide an ambiguous signal, as similar adaptations are found in silesaurids and some dinosaur groups.

In all, Ezcurra et al. document features across the skeleton that connect lagerpetids to pterosaurs (some shown below). Interestingly, when the authors included Scleromochlus in the analysis, it showed up as a sister taxon to the (Lagerpetidae + Pterosauria) clade.

In each case, lagerpetid on the left pterosaur on the right.

The authors note that the structural changes in the inner ear between lagerpetids and early pterosaurs mirrors the same transformation between non-avian theropods and birds. Do lagerpetids show any arboreal tendencies? Apparently Dromomeron romeri has strongly-recurved manual digits, suggesting to the authors that it may have used its forelimbs to capture prey, climb in trees, or both. “Our observations suggest that functional forelimb versatility became widespread in ornithodirans, allowing the evolution of disparate behaviours such as manual processing of food resources in dinosaurs and active flight in pterosaurs.” (Ezcurra et al., page 5).

This does not mean, of course, that lagerpetids ARE the ancestors of pterosaurs, but as an outgroup, they can give us insights into the transformations that must have taken place between the two groups. And in fact, recognition of lagerpetids as basal pterosauromorphs still allows 18 million years for lagerpetid-like ancestors to develop into their final forms--I can only hope that representatives from that critical anatomical transition are one day discovered (or maybe they already have been, and are sitting in a museum someday).

I apologize for the formatting weirdness in this post--I'm getting sick of wrestling with the HTML.


  1. Zach, good history and review of lagerpetids. Did you miss this abstract? Novas FE and Agnolin FL 2016 Lagerpeton chanarensis Romer (Archosauriformes): A derived proterochampsian from the middle Triassic of NW Argentina. Simposio. From Eventos del Mesozoico temprano en la evolución de los dinosaurios”. Programa VCLAPV. Conferencia plenaria: Hidrodinámica y modo de vida de los primeros vertebrados. Héctor Botella (Universitat de València, España) 2016.

    This confirms when more taxa are added Lagerpeton nests with the proterochampsian Tropidosuchus.

    Speaking of taxon exclusion, Ezcurra et al. cited Peters 2000, then omitted the pterosaur precursors recovered by there by adding Cosesaurus and kin to four previously published analyses. Likewise Hone and Benton 2007, 2008 claimed they were going to test the results of Bennett 1996 vs. Peters 2000 in the first of their two-part study, but omitted all reference to Peters 2000 in the second part of their study when Cosesaurus and kin nested closer to pterosaurs in their supertree. Benton 1999 published that Scleromochlus was the sister to pterosaurs and he didn't want to be wrong. This is akin to the Oculudentavis scandal. Taxon omission (cherry-picking taxa) is a problem in paleontology. More details here:

    1. "This confirms when more taxa are added Lagerpeton nests with the proterochampsian Tropidosuchus."

      Nope. Ezcurra et al. (2020) has eight proterochampsid OTUs, so clearly the new lagerpetid material has made the 2016 results moot. Have you added the new cranial and forelimb material to your Lagerpeton and Ixalerpeton OTUs?