The Case Against Bipedal Pterosaurs

Two Camps
The subject of pterosaur bipedality remains contentious. In the modern era, Stokes (1957 attributed odd quadrupedal tracks to pterosaurs. For the next 25 years no one argued against it. Then Padian and Olsen (1983) determined that Pteraichnus tracks were crocodilian in origin. This convinced several other paleontologists along the way (Unwin 1989). In 1995 two papers independently determined that Pteraichnus tracks could only be pterosaurian in origin (Mazin et al. 1995; Lockley et al. 1995), which turned attitudes around universally. No one else bought into the bipedal story except Bennett (1997, 2003) who illustrated the oddly proportioned and fingerless Nyctosaurus as a biped, but otherwise fell in with the quadrupedal folks.

The Bipedal Ancestor Irony
Virtually all the quadrupedal theorists also insist on a close relationship with the obligate biped, Scleromochlus and a close relationship to basal bipedal dinosaurs. Hmm…

The Heretic
My own phylogenetic work (Peters 2000, 2011) shows that basal fenestrasaurs (including pterosaurs) were occasionally to obligatorily bipedal and that several clades of derived pterosaurs were quadrupeds. All pterosaurs would have been capable of bipedal locomotion in the manner of lizards that can attain a bipedal configuration (Fig. 1, that’s how they spread their wings in preparation for a take-off). However many clades preferred quadrupedal locomotion during beachcombing/feeding/ordinary walking as demonstrated by their quadrupedal ichnites and the relative length of the fore and hind limbs. This form of locomotion was secondarily derived, as demonstrated by the orientation of the manual fingers. Manual digit 3 often is oriented posteriorly, the opposite of all other terrestrial tetrapods, and digits 4 and 5 were elevated off the substrate.

The quad proponents dispute virtually all suggestions of bipedality (see below). Hone and Benton (2007) went so far as to say, Cosesaurus is treated as a biped by Peters (2000) with characters coded based on this assumption.” Not sure how that could possibly affect bone traits and ratios, but that’s the attitude and paradigm out there. The authors were aware that Peters (2000) stated Cosesaurus was an occasional biped based on matching its feet to Rotodactylus ichnites, which are occasionally bipedal.

The Evidence from Ichnites
So far we have quadrupedal ichnites for pre-germanodactylids, ornithocheirids, pterodactylids, ctenochasmatids and azhdarchids. We have pedal ichnites without manus impressions for anurognathids. We have pedal ichnites with occasional manus ichnites for cosesaurids that are called Rotodactylus (Peters 2011).

A 2003 Argument Against Bipedal Pterosaurs
Darren Naish is a brilliant paleontologist with many discoveries to his credit, but on this subject he was in the “all quadruped virtually all the time” camp. In 2003, Darren discussed on the DInosaur Mailing List (DML) bipedality in lizards compared to that of pterosaurs in response to the publications of Peters (2000 and 2002) and several posts I had made to the DML. I have abridged his arguments (in yellow below, but you can read his full post here).

Darren: …here I state more clearly why I think the lizard/pterosaur analogy is flawed. then he quotes an earlier post I made to the DML , “If lizards can do it, irrespective of the math, pterosaurs could do it because they have superior equipment (increased sacral number, anteriorly hypertrophied ilium = bigger thigh muscles). As in birds or bipedal lizards, the CoG can be easily manipulated to be either head heavy or tail heavy by moving the tail, head, femur, tibia or angle of the back. Nothing out of the ordinary is required to balance a pterosaur. And the forelimbs are always within a whisker of touching the substrate to deal with momentary lapses.”

Then Darren reports, “– Why do certain lizard species run bipedally? Is it just so that they are faster? Probably not: the fastest lizards are quadrupedal runners.”

Indeed. Speed is not the reason.

Darren: “What appears most likely is that bipedal running in lizards has evolved to circumvent Carrier’s constraint: by relying on the hindlimb complex alone, bipedal running lizards are not compressing the thorax as they run, and they are therefore able to maintain breathing while sprinting (in contrast to quadrupedal running lizards).

Unfortunately not true. According to Christopher Clemente“When you see a lizard running bipedally, it’s just a consequence of its acceleration.”  Clemente found that lizards trotting on two legs ran out of steam quicker, indicating that bipedalism does not serve to conserve energy.

Bipedal lizard video marker

Figure 1. Click to play video. Just how fast can quadrupedal/bipedal lizards run? This video documents 11 meters/second in a Callisaurus at the Bruce Jayne lab.

If you want to see a video of a quadrupedal lizard screaming along at a breakneck pace, then, for reasons known only to itself, lifts up its forelimbs and continue running, click here. This is not an acceleration situation. Some lizards run more upright than this one.

Then Why Go Bipedal?
Clemente said, “If you’re not using your front legs anymore, they can develop for something else.” Pterosaur predecessors, like Cosesauruswere flapping their forelimbs according to their bird-like/ pterosaur-like, strap-like scapula and locked down stem-like coracoids (both very un-lizard-like).

Sharovipteryx, another pterosaur sister, could not have contacted the substrate with such short forelimbs and long hindlimbs. It too flapped forelimbs anchored with a stem-like coracoid.

Longisquama was similar, but with much larger, more pterosaur-like hands. Certain pterosaurs were secondarily quadrupedal, as demonstrated by phylogeny and the lack of pronation in the manus. Certain pterosaurs left only pedal impressions.

Darren: Given that pterosaurs would clearly not have needed to avoid Carrier’s constraint (because, even if they were to run quadrupedally, they would not be compressing the thorax with each stride), why run bipedally?

Bipedal lizards do not avoid Carrier’s constraint.
Following Clemente’s comments (above), bipedal lizards “run out of gas” sooner rather than later. So, they’re still holding their breath while running bipedally. In pterosaurs a short stiff torso was one trait to help avoid Carrier’s constraint. Combine that with a lack of large caudofemoral muscles. Lizards use these to propel their sprawling hind limbs with alternate lateral pulls of the femur. Instead pterosaurs and their fenestrasaur antecedents had large, dinosaur-like thighs as long as the extent of their elongated ilia. They ran using fundamentally different muscles than lizards do (thigh muscles vs. tail muscles), and basal pterosaurs had more erect hind limbs than more derived pterosaurs.

Darren: Furthermore, given that pterosaurs exhibit features associated with leaping (see, e.g., Bennett 1997) and/or scansoriality, it is probable that they wouldn’t need to sprint in order to take off.

Not sure why Darren discounts running in creatures capable of leaping. I can think of several mammals that are great leapers AND runners (rabbits, deer, roadrunners, squirrels, big cats, Michael Jordan. With birds some can take off with a simple leap. Others require a runway.

Darren: Bottom line: there is no inherent ‘need’ for good bipedal running abilities in pterosaurs, in contrast to the situation in lizards.

Unfortunately, Darren did not take into account the possibility of a secondary sexual characteristic, flapping, that pterosaur predecessors practiced (based on their stem-like coracoids). Flapping to show off is how pterosaurs developed the necessary ‘equipment’ to flap to fly. Ironically, Darren’s remarks seem to be indicating there IS a need for bipedality in lizards, after arguing there was no good reason for it in lizards either (see above) – yet they do it! (but not for sex~). Evolution does not proceed based on “need” in any case, but on random changes, some of which prove to enhance survivability.

Darren asks, “What allows certain lizards to run bipedally? Dave is fond of stating that pterosaurs could run bipedally because they exhibit an increased number of sacrals relative to their probable outgroups, and a large preacetabular process on the ilium. As has been pointed out several times in the literature, it?s relatively easy for a lizard to run bipedally IF it combines these two features WITH (1) hindlimbs that are proportionally longer than its forelimbs (and consequently the animal has proportionally short forelimbs), (2) a proportionally short thorax and (3) short neck*, and (4) a long muscular tail (see Synder 1954, 1962, Bellairs 1969, Rieppel 1989 etc). Note that most of these features are to do with reducing the mass of the foreparts and thus shifting the CoG caudally. On point (4), as shown by Russell and Bauer (1992), the most important anatomical correlate of bipedality in lizards is the presence of a large m. caudofemoralis longus that inserts relatively distally on the tail (thus explaining why the lizards that run bipedally are the same ones that don?t practise caudal autotomy). *Apparently _Chlamydosaurus_ has a longer neck than most other agamids. Its neck is still not as proportionally long as that of a pterosaur though. Pterosaurs obviously don’t have proportionally short forelimbs, but more importantly they don’t have the short neck seen in bipedal lizards, nor do they have a tail that would have supported a large m. caudofemoralis longus: even in basal long-tailed forms, transverse processes (and hence a reliance on m. caudofemoralis longus) are extremely reduced (and, incidentally, there is no indication that pterosaurs switched to the knee-based retraction system seen in birds). On the relevance of this reduction in caudofemoral musculature to bipedal locomotion, Synder (1954) writes “while a long, heavy tail does not necessarily indicate bipedal habits, a short, lighter tail precludes the possibility of this type of locomotion? (p. 9). Given then the profound differences evident here between pterosaurs and bipedal lizards, I think the analogy is seriously suspect.”

Unfortunately Darren (like so many others) completely ignores or overlooks the origin of bipedality in pterosaur ancestors that I described in 2000. Bipedality appeared with Cosesaurus, which had feet which matched occasionally bipedal and always narrow-gauge, digitigrade tracks. Cosesaurus, Longisquama and basal pterosaurs (including anurognathids) all had short necks. Longer necks evolved later. Sharovipteryx, an obvious biped, had a long neck. So that’s not an issue. Basal pterosaurs with short forelimbs and long hindlimbs, like MPUM6009 would have been awkward quadrupeds. The longer forelimbs developed AFTER pterosaur ancestors were already flapping, leaping and running about bipedally.

Darren continues, “– So what of the alleged correlates of bipedality present in pterosaurs? Dave suggests that an increased number of sacral verts and a hypertrophied preacetabular process on the ilium are indicative of “improved: bipedality in pterosaurs. The problem is that, firstly, the features discussed above are needed as well (viz, proportionally short neck, big m. caudofemoralis longus etc), and, secondly, when the sacral and iliac features are present without these others, they may not be indicative of bipedality but of quadrupedality. Look at (e.g.) ceratopsians. Relative to basal ceratopsians, ceratopsids have a longer preacetabular process and an increased number of sacrals (10-11 compared to 6), so according to your criteria ceratopsids might be better suited to bipedality than psittacosaurs. Parareptiles come to mind too: in nycteroleterids, nyctiphruretids, procolophonoids and sclerosaurs there are (usually) 3 sacrals and a short or absent preacetabular process, but in pareiasaurs – most notably in big derived forms like _Scutosaurus_ – there are 4-6 sacrals and the preacetabular process may be so hypertrophied that the pelvis looks much like that of a pterosaur (see Fig. 14E in Lee 1997). As in ceratopsians, these sacral and iliac trends are to do with improved quadrupedal abilities.”

Well, when you start comparing pterosaurs to pareiasaurs and ceratopsians, I think we can all agree, Darren is really reaching here. Certainly a high sacral number is a convergent feature here, but for different reasons. Note, he avoids mention of Sharovipteryx and Cosesaurus, both with an increased sacral count and much closer to pterosaurs than pareiasaurs and ceratopsians.

In bipeds, like pterosaurs and dinosaurs (both bipeds and descendants of bipeds), the sacrum is put under greater stress as the fulcrum balancing all the weight anterior to the sacrum and all the weight aft. Elongating the ilium, adding to the muscle mass of the thigh, along with strengthening the fulcrum is the reason for adding sacrals (often coosified) in pterosaurs. No one championing the quadrupedal configuration has ever proposed another reason for increasing the sacral count and ilium length in pterosaurs. Sure cynodonts added sacrals. The also reduced their caudofemoralis muscles and their tails while elongating their ilia, all convergent with pterosaurs – without going bipedal. Anyone can take parts and make arguments any way they want to, but they can’t take the whole suite of characters and make the same argument. And, as Darren and most paleontologists would agree, parsimony only rules when you look at the sum of a taxon’s traits, not just a few of its convergent parts.

Darren continued, “A better way of testing for bipedality in Pterosauria might be to look at intermembral indices (viz, forelimb:hindlimb ratios), at the CoG (as I mentioned, Sangster has been working on this), at unambiguous soft tissue evidence (e.g., the Crato azhdarchoid with its preserved brachiopatagium), or at trackway evidence? and right now the evidence from all of these areas shows that quadrupedality is better supported, or in other words that pterosaurs were more likely quadrupedal.

Here Darren is talking about most pterosaurs, not basal taxa or their ancestors. As above, all pterosaurs more derived than MPUM6009 were capable of placing their hands on the substrate – without bending over an iota! (a fact typically overlooked in most other quadrupedal reconstructions of pterosaurs). Even so, the most primitive pterosaurs for which we have ichnites, the anurognathids, preserve no manus impressions.

Darren asked, “– Why be bipedal anyway when the forelimbs are plenty
long enough? Dave notes that, in bipedal pterosaurs, “the forelimbs are always within a whisker of touching the substrate”. Well, if that;s so, it seems more likely to me that the animals would have employed the forelimbs in locomotion. I can’t think of a group of living animals in which the forelimbs are close to the substrate, and are not then employed in locomotion (think monkeys and apes). Again, the hard evidence we have (trackways) shows that the forelimbs were deployed in quadrupedal locomotion.

Darren was not aware of the other hard evidence of pedal impressions of pterosaurs without manus impressions.  In virtually all pterosaurs the forelimbs were plenty long enough to touch the substrate and the tracks show that many clades were quadrupedal. I never argued against that. Some pterosaurs reverted to quadrupedal locomotion, always retaining the ability to walk or run bipedally, simply by lifting the forelimbs off the substrate. Their toes were already planted beneath their centers of gravity, the shoulder glenoid, as ALL of my pterosaur reconstructions demonstrate. It’s that simple. As in Darren’s example, think lemurs, monkey and apes, all of which can go bipedal whenever they want to.

Darren concludes with, “The presence of a well-developed iliopubic ligament (=ligamentum inguinale) might indicate that pterosaurs were good at elevating the thorax, but given that everyone agrees that pterosaurs must have been bipedal when opening or closing their wings (and they thus would have needed to elevate the thorax at least occasionally), this doesn’t necessarily indicate bipedal running. Incidentally, see Hutchinson (2001, pp. 156-8) for a discussion of iliopubic ligament distribution in Reptilia. Because reptiles including ceratopsids, pareiasaurs and turtles appear to have had a hypertrophied iliopubic ligament as well, the correlation between this structure and an enhanced bipedal ability is not immediately clear.

I appreciate the half-hearted concession Darren made regarding opening the wings. Nice to hear. Yes, pterosaurs had to stand bipedally to open the wings. (Forelimb launch was not a consideration when Darren wrote this, and it has major flaws in any case.) The anterior extension of the ilium does not always signal a bipedal configuration (e.g. basal mammals). I never argued that it did. But in the case of pterosaur antecedents, a long ilium is one of a long list or suite of traits  shared with pterosaurs. In Peters (2002), I was moved to report that other than the twist of the wing finger, nearly every pterosaurian trait could be found in fenestrasaur antecedents. No has ever argued against that hypothesis and presented a more parsimonious series of antecedents from some other distant clade. And it has been ten years!

Darren further concluded, “The long forelimbs of pterosaurs, combined with the morphology of the patagia and the evidence from trackways, show that the interpretation of pterosaurs as predominantly quadrupedal is better supported and less speculative than interpretation of them as bipedal.”

Cosesaurus and Rotodactylus, a perfect match.

Figure 2. Click to enlarge. Cosesaurus and Rotodactylus, a perfect match. Elevate the proximal phalanges along with the metatarsus, bend back digit 5 and Cosesaurus (left) fits perfectly into Rotodactylus (right).

Predominate? Yes. However, Darren was not, at the time, aware of certain pterosaur pedes imprints attributed to anurognathids that were not attended by manus imprints. Those were described in Peters (2011), but “Sauria aberrante” (Casamiquela 1962) and Rotodactylus (Peabody 1948, Fig. 2) have been known for decades. Even though they are in the minority at present, they still count.

In summary, looking for reasons to go bipedal is probably not the way to go. Looking at bits and pieces by themselves is also not the way to go. Cladistic analysis and judging a taxon as a whole are the ways to go. And I like the example of quadrupedal/bipedal primates. That’s a good analogy!

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

Darren Naish, School of Earth & Environmental Sciences University of Portsmouth UK, PO1 3QL

Click to read Darren Naish’s complete comment to the DML
Bennett SC 1997. Terrestrial locomotion of pterosaurs: a reconstruction based on Pteraichnus trackways. Journal of Vertebrate Paleontology, 17: 104–113.
Bennett SC 2003. New crested specimens of the Late Cretaceous pterosaur Nyctosaurus.Paläontologische Zeitschrift 77: 61-75.
Casamiquela RM 1962.
Sobre la pisada de un presunto sauria aberrante en el Liassico del Neuquen (Patagonia). Ameghiniana, 2(10): 183–186.
Lockley MG, Logue TJ, Moratalla JJ, Hunt AP, Schultz RJ and Robinson JW 1995. The fossil trackway Pteraichnus is pterosaurian, not crocodilian; implications for the global distribution of pterosaur tracks. Ichnos, 4: 7–20.
Mazin J-M, Hantzpergue P, Lafaurie G and Vignaud P 1995. Des pistes de pterosaures dans le Tithonien de Crayssac (Quercy, France). Comptes rendus de l’Academie des Sciences de Paris, 321: 417–424.
Peabody FE 1948.  Reptile and amphibian trackways from the Lower Triassic Moenkopi formation of Arizona and Utah.  University of California Publications, Bulletin of the  Department of Geological Sciences 27: 295-468.
Peters D 2000.
 Description and Interpretation of Interphalangeal Lines in Tetrapods.  Ichnos 7(1): 11­-41.
Peters D 2002. A New Model for the Evolution of the Pterosaur Wing – with a twist. Historical Biology 15: 277-301
Peters D 2007. 
The origin and radiation of the Pterosauria. In D. Hone ed. Flugsaurier. The Wellnhofer pterosaur meeting, 2007, Munich, Germany. p. 27.
Peters D 2010. In defence of parallel interphalangeal lines. Historical Biology 22:437-442.
Peters D 2011. A Catalog of Pterosaur Pedes for Trackmaker Identification. Ichnos, 18: 2, 114 —141.
Stokes WL1957. Pterodactyl tracks from the Morrison Formation. Journal of Palaeontology, 31: 952–954.
Unwin DM 1997. Pterosaur tracks and the terrestrial ability of pterosaurs. Lethaia, 29: 373–386.

Darren’s References:
Bellairs, A d’A 1969. The Life of Reptiles, Vol. 1_ Weidenfeld & Nicolson (London), pp. 282.
Bennett SC 1997. The arboreal leaping theory of the origin of pterosaur flight. Historical Biology 12, 265-290.
Hutchinson JR 2001. The evolution of pelvic osteology and soft tissues on the line to extant birds (Neornithes). Zoological Journal of the Linnean Society 131, 123-168.
Lee MSY 1997. Pareiasaur phylogeny and the origin of turtles. Zoological Journal of the Linnean Society 120, 197-280.
Reeder TW, Cole CJ and Dessauer HC 2002. Phylogenetic relationships of whiptail lizards of the genus_Cnemidophorus_ (Squamata: Teiidae): a test of monophyly, reevaluation of karyotypic evolution, and review of hybrid origins. American Museum Novitates 3365, 1-61.
Rieppel O 1989. The hind limb of Macrocnemus bassanii (Nopcsa) (Reptilia, Diapsida): development and functional anatomy. Journal of Vertebrate Paleontology 9, 373-387.
Russell AP and Bauer AM 1992. The m. caudifemoralis longus and its relationship to caudal autotomy and  locomotion in lizards (Reptilia: Sauria). Journal of Zoology 227, 127-143.
Synder RC 1954. The anatomy and function of the pelvic girdle and hind limb in lizard locomotion. American Journal of Anatomy 95, 1-45.
Synder RC 1962. Adaptations for bipedal locomotion of lizards. American Zoologist 2, 191-203.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.