Gallery and print store

Saturday, 10 October 2015

We just can't quit you, Pterodactylus

A small flock of Pterodactylus antiquus, represented by small juveniles (left) up to big adults (right) scope out foraging options in a Jurassic marsh. The animal on the right is luring prey to the surface through paddling forefeet, a behaviour common to (at least) several modern gull species.
Pterosaur researchers are infamous for their frequent disagreements over flying reptile evolution, lifestyles and even basic anatomical interpretation. I can certainly attest that there is some truth to this: writing about pterosaurs can be frustrating because of continual need for clarification and digressions to ensure all points of view are represented. But if there's one pterosaur we must have reached a consensus over, one species we must all agree about, surely it's the Late Jurassic, Solnhofen species Pterodactylus antiquus. The holotype of Pterodactylus - below - has been known to us longer than any other pterosaur fossil, this being the specimen which kick-started flying reptile studies in 1780. Since then, it's almost a rite of passage for researchers to see this specimen along with some of the other several dozen Pterodactylus fossils in museums around the world. Pterodactylus has been looked at so much that a general agreement over what it is, and the species it is related to, has more or less been established. Characterised by a long, low skull, simple teeth, a long neck and largish feet, we consider Pterodactylus a relatively early form of pterodactyloid, and most likely a member of the ctenochasmatoid/ archaeopterodactyloid* branch of pterosaur evolution. This puts it in the same lineage as several other familiar species, such as comb-toothed Ctenochasma and twisty-jawed weirdo Cycnorhamphus. We're all basically happy with the idea that our dataset for Pterodactylus is fairly good: 30 specimens (a very conservative estimate) provide numerous complete skeletons and a growth series from small juveniles up to very large adults.

*As with many parts of the pterosaur tree, nomenclature for 'Pterodactylus-line' pterodactyloids is confused by the use of several, conflicting names and definitions. I wasn't kidding about those caveats and digressions.

All that said, some areas of Pterodactylus research remain contentious, and new insights into its anatomy and disparity are still being published more than 200 years after it was discovered. Surprisingly, unlike the way we often gain novel appreciation for familiar taxa - new specimens shedding new light on old problems - much of our recent understanding of Pterodactylus relies on the same, well-worn specimens we've been analysing for centuries. It's actually quite sobering to see specimens which have been interrogated so much still providing talking points, and it makes me wonder what we're missing from those briefly described, rarely analysed specimens comprising so much of our vertebrate fossil dataset. Here, I want to cover some of the new insights provided on Pterodactylus in just the last two years.

The specimen which started it all: the Pterodactylus antiquus holotype. The wingspan of this specimen, preserved in a 'falling forward' posture rather atypical for a Solnhofen pterodactyloid, is about 45 cm.

Anatomy and palaeobiology

Until recently, most of us have been used to seeing Pterodactylus depicted as a crestless species. A privately owned specimen described in 2003 showed that, like many other pterosaurs, this animal bore a set of soft-tissue structures associated with the top and back of the skull (Frey et al. 2003). Specifically, it seems Pterodactylus bore a soft-tissue crest along the posterior region of the head and a pointed, posteriorly-projecting 'occipital lappet' at the back of the skull. The latter, for now at least, seems unique to Pterodactylus. This information is well known to 21st century scholars, but it's less appreciated that these soft tissues were first mentioned almost 100 years ago. Pterodactylus crests were first reported in the 1920s, and the lappets in 1970 (see Bennett 2013). I find it bizarre that we didn't start restoring Pterodactylus with these interesting structures until the 2000s: this seems to be an example of artists and scientists not working together as well as they might.

Unlike other pterosaurs, the soft-tissue crest of Pterodactylus did not seem to anchor on a low, striated bony ridge. The absence of this feature, even when preservation was sublime enough to record soft-tissues and detection methods were of late 20th century quality, was likely a key factor in our general consideration of Pterodactylus as a crestless species. I always found the occurrence of soft-tissue crests without corresponding bony structures an alarming prospect, one implication being that we could be ignorant of soft-tissue crests in a huge number of pterosaur species.

It was somewhat relieving, therefore, to see Chris Bennett reporting a crest-anchoring structure for Pterodactylus in 2013. It's small, and often smooth rather than striated, but Pterodactylus definitely does have a midline ridge for crest anchorage - even the holotype has one when we look close enough. Exactly how extensive these structures were remains unknown thanks to many historic specimens being accidentally damaged during preparation. It's easy to see how this occurred: the crests are low, extremely fragile, and only 0.2 millimetre thick. They'd be hard to detect and avoid damaging even if you were looking for them. Hopefully, preparators working on unprepared specimens can recover intact crests now we know they exist.

The most extensive example of cranial soft tissues known thus far from Pterodactylus. Unfortunately, we're still some way from knowing what shape they took in life, although this specimen indicates that almost half of the skull was covered by the crest and that the lappet was also quite large. Parts of the diagram labelled 'fa' record sediments which fluoresce under UV light - they're likely matrix contaminated by organic seepage from the decaying pterosaur head. They are unlikely to tell us much about the appearance of the animal in life. From Bennett (2013).
Pterodactylus cranial soft tissues are now known to occur in a number of specimens, but it remains unclear how large or what shape the crests were. The lappets seem to be of a consistent size and position, and many curve upwards, but whether they are joined to the rest of the crest (as suggested by Frey et al. 2003) remains to be confirmed (Bennett 2013). At least some aspects of crest and lappet development matches what we see in other pterosaurs, in that we only start picking up evidence of these structures in larger Pterodactylus specimens. There also seems to be a rough correlation between crest proportions and body size. Pterodactylus thus seems to be another pterosaur species where cranial ornament signifies entry into adulthood, suggesting a function of sexual communication (Bennett 2013).

Speaking of adulthood, it was also only recently that we obtained a true sense of how large Pterodactylus may have grown. We typically imagine this animal as small bodied - maybe with a 50 cm wingspan - but a newly described skull and lower jaw (below) makes the first unambiguous case for Pterodactylus reaching at least 1 m across the wings (Bennett 2013). To put this in a modern context, large Pterodactylus would be of comparable size to smaller heron species, and large individuals would have been conspicuous components of the Solnhofen pterosaur fauna. A trend where skull, neck, and limb proportions increase with body size, first intimated by Peter Wellnhofer (1970), seems to hold up in modern interpretations of Pterodactylus specimens. Realising how variable this pterosaur's proportions might have been throughout life has been very informative to recent considerations of Pterodactylus taxonomy.

The mother of all Pterodactylus skulls. A preserved skull length of 142 mm indicates a skull of around 200 mm long in life, and an animal reaching a 1 m wingspan. From Bennett (2013).

One species, two species, or three genera?

This brings us to some of the more contentious recent developments in Pterodactylus studies: just how many species are represented in the Pterodactylus dataset? Many readers will be aware that the name Pterodactylus was once applied to almost any new pterosaur fossil, and around 80 'Pterodactylus' species have existed in the last 200 years (Ford 2013). The taxonomic history of Solnhofen pterodactyloids has been especially mixed up with the name Pterodactylus and, by the end of the 1800s, their taxonomy was in a real tangle. Work in the mid-20th century, particularly by Peter Wellnhofer (1970), streamlined systematic interpretations of Pterodactylus so that, by the 2000s, only two species were considered valid: P. antiquus and P. kochi. A couple of 'hangers on' were still knocking about ('P'. longicollum and 'P'. micronyx), but researchers universally agreed that these animals were not true members of Pterodactylus, and were simply awaiting new generic names (they now have them: Ardeadactylus and Aurorazhdarcho, respectively).

Distinguishing features between kochi and antiquus were subtle, being primarily aspects of tooth shape, tooth number, and proportions of the skull, neck and torso. This is not a new observation, and suggestions that they may represent the same taxon date back to the 1800s. Eventually, studies of Pterodactylus teeth was used to suggest outright synonymy of these two species (Jouve 2004). Many pterosaurs, as with most reptiles, increase their tooth counts with age and size. Jouve realised that the allegedly distinctive tooth count of P. kochi aligned perfectly with tooth numbers predicted for antiquus of comparable body size. At least in this respect, these two species could not be distinguished. More recently, Bennett (2013) bolstered this synonymy with an assessment of kochi proportions, noting that perceived distinctions in skull and body length were reliant on erroneously recorded measurements. Once corrected, kochi proportions were very similar to comparably sized antiquus individuals (there's a lesson there about the importance, and repetition, of basic data recording in this) and, along with Jouve's work, this study eroded the foundations of the kochi/antiquus split considerably. Remaining distinguishing features between these species are rather poorly defined, and certainly not divorceable from effects of growth, preservation and preparation. Finally, after 200 years, it was looking like Pterodactylus taxonomy had finally been tidied up: we have one Pterodactylus species, not two, or 80.

Historically considered to represent Pterodactylus antiquus, recent work argues this specimen (along with some referred material) is a wholly distinct species, and distantly related to P. antiquus. It was recently christened Aerodactylus scolopaciceps. Image from Vidovic and Martill 2014 (this particular version from Steve Vidovic's Mesozoic Monsters blog).

Except... the story doesn't end there. Last year, my University of Portsmouth colleagues Steven Vidovic and David Martill suggested that not only were 'cryptic taxa' present in the Solnhofen Pterodactylus dataset, but that the traditional phylogenetic placement of some Pterodactylus-like animals might be erroneous. Using a variety of methods, Steve and Dave proposed that Pterodactylus contained at least three taxa: antiquus (which they considered the only true member of the genus), kochi (a separate genus in their interpretation, and more closely related to other pterodactyloids than antiquus), and a resurrected Pterodactylus species from the 1800s, scolopaciceps (Vidovic and Martill 2014, see image, above). Steve and Dave created the generic name Aerodactylus for this animal, providing a diagnostic combination of over 10 character states relating to skull shape and proportions, orbit shape, tooth count, neck length, humeral curvature and limb bone robustness. Attempting to establish the relationships of these three 'Pterodactylus' taxa saw Ctenochasmatoidea/Archaeopterodactyloidea dissolve into a paraphyletic spread across the base of Pterodactyloidea. In this topology, antiquus and kochi anchor the base of Pterodactyloidea (without forming an exclusive clade themselves) and scolopaciceps is at the other end of the 'ctenochasmatoid' range, in a sister clade to the rest of Pterodactyloidea.

That's quite a shake up, contradicting virtually all other recently published opinions on the taxonomy and evolution of these animals. But although different, at least some of these ideas are not be untenable. For instance, the idea that Ctenochasmatoidea/Archaeopterodactyloidea might be paraphyletic is suggested by the 'Painten pro-pterodactyloid', an unusual pterosaur specimen revealed two years ago (below, Tischlinger and Frey 2013). This taxon, which shows a Pterodactylus-like skull combined with postcranial features somewhat like those of non-pterodactyloid pterosaurs, suggests aspects of 'ctenochasmatoid' anatomy developed outside of Pterodactyloidea proper. It therefore will not be that surprising if this taxon pulls some 'basal' ctenochasmatoids of traditional lore to the root of Pterodactyloidea once it's included in phylogenetic studies. (Those interested in the influence of the 'Painten pro-pterodactyloid' animal on our understanding of pterosaur evolution might find this previous post of interest).

The 'Painten Pro-pterodactyloid' specimen, messing up our nice, neat interpretation of pterodactyloid evolution since 2013. Notice the Pterodactylus-like posterior skull morphology alongside traditional non-pterodactyloid features - a long(ish) tail and big fifth toes. From Tischlinger and Frey (2013).
But do our few dozen Pterodactylus specimens really comprise three, distantly-related species? On this, I'm less certain. We see a lot of variation across Pterodactylus specimens reflecting those aforementioned factors of ontogeny, preservation and preparation - not to mention individual variation. Having played with Pterodactylus data a little myself, and seen a fair share of specimens relevant to these discussions (though I stress not all), I find the arguments for synonymy more compelling than those for splitting Pterodactylus apart. This said, I have no horse in this race and could be persuaded otherwise. What we really need - and a number of folks in pterosaur research have been saying this for a while now - is someone to travel the world exhaustively documenting and illustrating Pterodactylus specimens, ultimately producing a modern synthesis on its anatomy. Such a study would not only be a valuable research aid (the last attempt at this was 50 years ago, which is an age ago in terms of research and publication techniques), but would pack a lot of weight in resolving ongoing, long running disputes in this animal's taxonomy.

Talking about the future of research into Pterodactylus seems like a sensible place to leave off, and I'll summarise in saying that - as with much else in pterosaur research - we're a little while off a complete consensus on Pterodactylus for now. Clearly, although the concept of Pterodactylus is over two centuries old, there's still things learn about it. Who knows what we'll be saying about this most familiar of pterosaurs in years to come?

These Pterodactylus were produced with Patreon

The artwork and words you see here are funded by a collective of extremely cool individuals who support me through my Patreon page. From $1 a month you can also be a patron, sponsoring output on this site, getting access to cool, exclusive content and rewards, and helping me produce higher quality art and writings. Thanks to all those who've signed on already!


  • Bennett, S. C. (2013). New information on body size and cranial display structures of Pterodactylus antiquus, with a revision of the genus. Paläontologische Zeitschrift, 87(2), 269-289.
  • Ford, T. L. (2013). Is Pterodactylus monophyletic or paraphyletic? Short Communications - International Symposium on Pterosaurs, Rio Ptero 2013. 68-70.
  • Frey, E., Tischlinger, H., Buchy, M. C., & Martill, D. M. (2003). New specimens of Pterosauria (Reptilia) with soft parts with implications for pterosaurian anatomy and locomotion. Geological Society, London, Special Publications, 217(1), 233-266.
  • Jouve, S. (2004). Description of the skull of a Ctenochasma (Pterosauria) from the latest Jurassic of eastern France, with a taxonomic revision of European Tithonian Pterodactyloidea. Journal of Vertebrate Paleontology, 24(3), 542-554.
  • Tischlinger H, Frey E. 2014. Ein neuer Pterosaurier mit Mosaikmerkmalen basaler und pterodactyoider Pterosaurier aus dem Ober-Kimmeridgium von Painen (Oberpfalz, Deutschland) [A new pterosaur with moasic characters of basal and pterodactyloid Pterosauria from the Upper Kimmeridgian of Painten (Upper Palatinate, Germany)]. Archaeopteryx 31, 1-13.
  • Vidovic, S. U., & Martill, D. M. (2014). Pterodactylus scolopaciceps Meyer, 1860 (Pterosauria, Pterodactyloidea) from the Upper Jurassic of Bavaria, Germany: the problem of cryptic pterosaur taxa in early ontogeny.
  • Wellnhofer, P. (1970). Die Pterodactyloidea (Pterosauria) der Oberjura-Plattenkalke SĂĽddeutschlands. Bayerische Akademie der Wissenschaften, Mathematisch- Wissenschaftlichen Klasse, Abhandlugen, 141, 1-133.


  1. Mark, I suggest you run a phylogenetic analysis on all specimens referred to Pterodactylus as well as all those which are not. I did so (results at and found that no two included putative Pterodactylus taxa scored identically, even though several were similar in size. The resulting tree was completely resolved with several putative Pterodactylus taxa nesting far from the holotype of Pterodactylus. The tiny putative juvenile Pterodactylus generally nested at the bases of distinct clades of larger taxa and were derived from distinct taxa often with longer tails, demonstrating phylogenetic miniaturization and the multiple origin of ‘pterodactyloid’-grade pterosaurs. Falsifying your claims of allometry during ontogeny, several juveniles of established taxa are known that are virtually identical to their parents (Pterodaustro, Zhejiangopterus, Pteranodon, the largest species of Rhamphorhynchus, Tapejara). Like many of their lepidosaur ancestors, pterosaurs matured isometrically. So if two pterosaurs are different, they cannot be a juvenile/adult pairing. It is okay to lump specimens together, but you must not try to do this by eyeballing them. Clades must be established by a comprehsive phylogenetic analysis (see above for an example of this). Accurate skeletal reconstructions set side-by-side are essential to demonstrate subtle variations you may be missing. The pedal bone proportions are typically unique, like fingerprints (Peters 2011) Bennett has been loathe to run an analysis. So have others duplicating the taxa listed in Peters 2007 and added to ever since.

    1. "Must not do this by eyeballing them."

      Yup, it's similar to how eyeballing the fossils yields this:
      Wait... A... Second...