Some Thoughts on Darwinopterus

By now, the blogosphere has had it’s chance on Darwinopterus. PZ Myers had a shot over at Pharyngula summarizing it, Darren Naish at Tetrapod Zoology approached the ecological angle by exploring the potential for predation (also handled by Mark Witton’s illustration of the aerial predator it is envisioned to be), while Dave Hone at Archosaur Musing’s set up the pterosaur general story and Darwinopterus‘ impact on it. Eventually, the hubbub will die down, but for now, I wanted to say a piece or three.

Lü J.-c., Unwin, D. M., Jin X.-s., Liu Y.-q. and Ji Q. 2010. Evidence for modular evolution in a long-tailed pterosaur with a pterodactyloid skull. Proceedings of the Royal Society of London, B 277(1680): 383-389.  (DOI: 10.1098/rspb.2009.1603)

For the purposes of taxonomy, this paper is not technically published, and will not be until it is produced in print. However, the taxonomy has been fairly prevalent and the name is in such abundant use that it seems unavoidable to use the nomenclature, so it is followed here.

Firstly, to reiterate what everyone else has already done, and that is to clarify the distinction between two grades of pterosaur: The pterodactyloid and the rhamphorhynchoid. This figure, linked from Dave Hone’s Archosaur Musings, replicates the rhamphorhynchoid model or bauplan on the left, and the pterodactyloid bauplan on the right. They are generally and immediately distinguishable by the tail and the neck, but in finer detail also by the architecture of the skull openings in front of the orbits and by the form of the wrist bones. The “rhamphorhynchoids” represent a grade of pterosaur that retained an elongated tail, short skull, short carpal suite (forming the wrist), and possessed a long fifth toe, while the “pterodactyloid” grade developed a longer neck, shorted the tail, elongated and fused portions of the carpal suite, and virtually lost all traces of a fifth toe.

Our most recently described pterosaur shows development of many of the traits of the “pterodactyloids,” but did not lose many traits of the “rhamphorhynchoids.” It’s been assumed by most people who studied the evolution of pterosaurs for the last century that this is what happened, but only recently now have we found fossils that show this transition occuring. What is more interesting, however, is that (like Naish points out) when one of these groups loses one feature that seems suited for predation, but gains another, interesting things may have been happening.

In 2002, Stephen Czerkas published a book that described several fossils, including a pterosaur named Pterorhynchus (Czerkas and Ji, 2002). This pterosaur, like Darwinopterus, combines a long tail, elongated cervical vertebrae (although still a short neck), and a retracted nasal bone that may or may not have lost contact with the maxilla, forming a nasoantorbital fenestra. However, it still possessed a short carpal suite, among other things, and was seemingly much less like a “pterodactyloid” than it was a “rhamphorhynchoid,” but like Darwinopterus, it appears to have been “in transition.” So it, too, shows mosaic, moduler transformation. (It is worthwhile to note that the authors spent some time considering the nature of Pterorhynchus, and probably for space concerns and the impact of the paper, it was ommited from all by the phylogenetic analysis.)

Skeleton of Pterorhynchus, approximated after Czerkas and Ji, 2002, with attention to preparation performed after publication.

Darwinopterus is relatively unique because, unlike many recent pterosaur discoveries from China, it is described on the basis of a number of corroborating specimens, which validates many features of its description, and prevent confusion on the basis of misinterpretation of a variety of details. It is unfortunate that Pterorhynchus does not receive more attention, for no other reason that if it had been published in a more mainstream manner, it (rather than Darwinopterus) would have been the model of the transition between the “rhamphorhynchoid” and “pterodactyloid” grades, part of neither but bridging them. It shows from a basic model (for example, Dorygnathus here) that the skull has elongated the nasal and antorbital fenestrae, retracted the nasal bones and probably combined the two fenestrae into a single one, rotated the occiput under the orbit and moving the jaw joint anteriorly so that it lies below the anterior portion of the orbit; the length of the skull compared to the dorsum length and the length of the cervical column is greater; it has elongated the metacarpal, although other features of the wrist seem obscured; and the teeth are more numerous and much smaller than classic “rhamphorhynchoides” but not as numerous or small as in Eudimorphodon (representing the earliest morphology known for pterosaurs). Thus it is a shame to overlook it, but it only adds to the complex that we see in Darwinopterus.

In Lü’ et al’s paper, Pterorhynchus is closer to the base of Pterosauria than is Darwinopterus, and they are separated by several placements of splitting-off taxa. Moreover, they are distinguished from one another by 9 distinct features, although some of these (such as the confluent nasoantorbital fenestra and the position of the jaw joint under the orbit) are present in Pterorhynchus (or at least seem to be). However, it seems Pterorhynchus is nonetheless more basal and less “pterodactyloid”-like than Darwinopterus is, and this helps us consider the evolution of the group in a whole. For example, it allows us to test the model of evolution of the flight apparatus and potential the prey acquistion mode that Naish delved into, by describing the steps in which certain features were acquired.

As of Pterorhynchus, its lineage had increased the length of the cervical column and of the skull (the latter far greater than the former), and had reduced the size of the dentition but increased its number. This emphasizes the head over the body, unlike in classic “rhamphorhynchoids” which have been characteristically “barrel-chested” and “pointy-headed.” A much longer, more slender, and toothier jaw lightened by larger openings of the skull and skull/neck that was more angled than linear orients the skull downward. If, as a recent paper by Mike Taylor et al. (Taylor, M. P., Wedel., M. J. and Naish, D. 2009. Head and neck posture in sauropod dinosaurs inferred from extant animals. Acta Paleontologica Polonica 54(2):213-220) suggests, the neck was inflected up from the dorsal series, this would still orient the skull slightly downward, and given the animal a high-up view (which has also been derived from work by Witmer et al. (Witmer, L. M., Chatterjee, S., Franzosa, J. and Rowe, T. 2003. Neuroanatomy of flying reptiles and implications for flight, posture and behavior. Nature 425:950-953; available online from Witmer Labs). This cranial position differs from that of the “rhamphorhynchoid” model as shown for one taxon in Witmer et al., in which the head was held almost parallel to the ground. Many “rhamphorhynchoids” exhibit skulls in which the teeth are elongate and procumbent (i.e., they are tilted towards from the front of the jaws) while the tips of the jaws were apparently also toothless, and while it has been proposed by some but tested by few, it has been suggested the animal employed a feeding strategy akin to some skimmers (Rhynchops sp.) which let the lower jaw drop into the water while the bird flies above (also employed by some birds that do not possess the rhynchopid extreme lower jaw, such as terns and gulls). Early references to pterosaur skimming include works cited in Humphries et al. (Humphries, S., Bonser, R. H. C., Witton, M. P. and Unwin, D. 2007. Did pterosaurs feed by skimming? Physical modelling and anatomical evaluation of an unusual feeding adaptation. PLoS Biology 5(8):e204. doi:10.1371/journal.pbio.0050204) and project a model of the narrow distal jaw as a skimming “prow,” although as noted with opportune skimmers like gulls, it is not required for a thin, deep mandible to perform this behavior successfully.

With Pterorhynchus, this feeding mechanism does not seem possible: the mandible is robust, but relatively shallow, and is toothed to its tip with smaller teeth. It thus seems more ideal for capturing or holding prey in a manner more reminiscent of Pterodactylus, of which one of the most famous specimens is shown here, as well as being the example in the grade comparison linked to at the beginning of this post). Cervical elongation improves reach and potential flexibility of the long skull, implying the requirement of mobility and reach in prey acquisition, and this seems more akin to another avian predator, the stork, which are stalking birds whom seek out and grab prey with their long beaks.

The next major transformation in this gradient is the retraction of the nasals and the opening of the antoantorbital fenestra but as yet no argument has yet been published linked reasonably to the purpose this played in pterosaur evolution. One major concept involves the increasing pneumatization of the skeleton, and when it reached the skull, would have decreased weight of the skull while removing unnecessary bone from the support structure of the skull. Thus, the seprate fenestrae were unnecessary, nor the interveneing bone, and their confluence is simply an artefact of this pneumatic process. Currently, studies on the extensiveness of pneumatic tissues in pterosaurs are sparse, but the msot recent projects size increase in pterosaurs was enabled by pneumatization, increasing the size of bones while retaining or even improving the structural capabilities of the bones (see, for example, Claessens, O’Connor and Unwin 2009 here, and Habib 2009 in the [hard-to-get-a-hold-of] Wellnhofer Pterosaur Meeting volume [B 28] of Zitteliana).

While the elongated tail is present, and potential coupling of the wing to the hindlimb as proposed for “rhamphorhynchoids” was retained, we can assume that the flight apparatus was generally the same even through to Darwinopterus. But by the time we get to Darwinopterus in the lineage, addition cranial elongation has occured, the wrist is clearly longer even while the wing seems the same size, and the foot retains its fifth toe. Fundamentally, not much has changed, and the features in which Darwinopterus is distinct from Pterorhynchus are almost certainly centered in the cranium. Here, the jaw is more slender, and this trend continues for Pterodactylus, while the occiput is rotated further ventrally and the jaw joint is slightly more anterior. Thus, we consider here that an important cranial transition started of the gradient transformation and the mosaic condition started, not with Darwinopterus, but with Pterorhynchus; and that the most complex change involved was a slightly longer neck, a much longer skull, and a reduction of the tooth size and thus prey acquisition varied considerably from the likely piscivorous “rhamphorhynchoid” grade to the more terrestrially or aerially oriented “pterodactyloid” grade. This grade in transition could, in honor of the recent taxon’s namesake, even be called “Darwinian” — I would propose “darwinopterygoid” for this grade, but I might be too presumptuous to put that in print.