Ornithothoraces deconstruction


Remember the outgroups I mentioned?

A) is the typical topology for Paraves, give or take. This particular result is derived from the Hartman 2019 version with characters for Enantiornithes and Jinguofortithidae added.

B) is what happens when you disregard the supposed homologies of Ornithothoraces. Its now proven that the enantiornithean sternum and the ornithuromorph sternum evolved independently (Zheng 2012), so revisions are in order.

C) is taking O’Connor 2019’s advice and regarding the anatomy of the digestive system as phylogenetic signal. Enants, confuciusornithids, dromaeosaurs and archaeopterygids had no crop nor gizzard while troodontids, scansoriopterygids, anchiornithids, sapeornithids and others had evidence of either pellets or gastroliths. Plus, archaeopterygids, dromaeosaurids and enants had slow growth rates while taxa close to modern birds had faster ones.

Enantiornithes: tentative phylogenetic tree



Using the dataset in Hartman 2019 albeit with the missing enantiornithean characters and additional taxa (same outgroups; the results are… interesting, and for another day). Only genera are listed.

As predicted by Mickey Mortimer, “Martinavis” is polyphyletic. North American and European remains rank with Halimornis while Lecho remains are almost identical to Elbretornis.

New Enantiornithes Cladogram (oh, and Ornithuromorpha as well)

From the Hesperornithoides paper’s supplementary we get this:

enantiorniEnantiornithes cladogram. Credit goes to Mickey Mortimer.

ornithuromorpOrnithuromorpha cladogram. Credit goes to Mickey Mortimer.

Per the study, the phylogenetic analyses presented are not meant to be objective fact, the authors repeatedly pointing out the required steps to change positions, often into more “organic” alignments. But if taken at face value, this is pretty shocking.

Some results do feel organic enough, like Lectavis, Hollanda and Qiliania forming a clade of long legged enantiornitheans while Hesperornithes are now part of Ichthyornithes, uniting the two major lineages of Cretaceous seabirds into a single monophyletic group.

Hopefully this gets further investigated.

Velociraptors on air remix

Deinonychus-ontogenetic-niche-shift-600-px-tiny-June-2015-Darren-Naish-Tetrapod-Zoology.jpgDeinonychus by Darren Naish, showcasing both flying juveniles and clueless adult.

A brief one. Everyone got acquainted with my hyper-controversial flying dromaeosaur post, right? Well, a more recent study on dromaeosaur flight has been posted. I disagree with some of its conclusions, but it did provided an immeasurably useful tool: a proper value for flight strokes in paravians.

According to said study, humerus-femur proportion rates have to be above 70% in order for a proper flight stroke to be provided and for wing-loading to be carried appropriately. To these ends there is a proper table with examples:


Some are to be expected, but other are VERY surprising. Deinonychus and Unenlagia, for example, rank 76-80 and 72 respectively. These are some big ass theropods, so to see them within this range certainly makes the hypothesis that they could fly as at least juveniles much more credible.

Then there’s Unenlagiinae being a basically cosmopolitan clade, including both european taxa as well as DAKOTARAPTOR and Rahonavis to tie it all together. So basically they were like ratites, with flying ancestors and multiple flight losses and gigantism. Nice.

1104754913_preview_Dakota (4)Flying up the tree baby!

Now the study as a whole prefers a polyphyletic origin of dinosaur flight, with basal paravians being flightless and then several clades becoming volant. Fair enough, but test on Pelecanimus first and then we’ll talk.

Can’t wait for the aneurisms this will cause on haters.

Velociraptors On Air Revisited

dromaeosauridae_sp__and_children_by_8bitaviation-d8y0pmrDromaeosauridae indet. by ‘8bitAviation’ on DA. In spite of the label, author claims that the animal probably doesn’t match Deinonychus antirrhopus per se, though it certainly helps to illustrate the putative flight capacities of young dromaeosaurs.

It’s been a long while since I wrote about theropod flight. I for the longest time lost interest in non-avian dinosaurs, but my interest has since been revived by a series of studies that have confirmed my hypotheses. Even as I began to doubt, my ego has been validated, and for that I am very happy.


Megapodes Are The Way

Birks megapodes004_4_2Newborn megapode bird; Burke Museum.

Probably one of the most controversial ideas I have spoken about was the possibility that, like modern megapode birds, juvenile dromaeosaurs and troodontids were capable of flight soon after birth, and that this extended to even forms in which adults are flightless.

Troodontids are long suspected of superprecocial habits (Varrichio et al 2002) and Enantiornithes have outright evidence of chicks with adult flight plumage, but three particular recent studies have outright confirmed that baby raptors could fly:

  • Fernández et al 2013 demonstrates that megapode-like burying behaviours were present in both Enantiornithes, some basal Euornithes and troodontid theropods. They suggest that this reproductive mechanism was basal at least in bird-like maniraptor dinosaurs, with oviraptorids and modern birds having developed brooding behaviours independently.
  • Parsons et al 2015 outright demonstrates that juvenile Deinonychus could fly(!). Admitely the study isn’t flawless, but this conclusion in particular is vindicated in light of the evidence demonstrated in both this section as well as latter on.
  • Mayr et al 2018 further reinforce the idea that basal birds as a whole, including iconic non-Enantiornithes such as Archaeopteryx and Confuciusornis were megapode-like in habits because they couldn’t incubate their eggs, and in fact brooding and altricial behaviours developed very late among birds.

Of curious relevance is the fact that both Fernández et al 2013 and Mayr et al 2018 cite megapodes as possible evidence that altricial behaviours weren’t even basal to the earliest modern birds, though Leaché et al 2014 suggests that megapodes did revert to this archaic reproductive strategy from brooding ancestors. Combined with oviraptorids brooding their eggs, it does suggest that brooding did develop independently a few times, though they are obviously the minority as the vast overwhelming majority obviously still could not incubate their eggs and had superprecocial young.

Also interesting is the possibility that dromaeosaurs did have flying juveniles but also brooded their eggs. Makovicky et al 2006 found Deinonychus eggs to be more similar to those of Citipati, an oviraptorid, than to those of Troodon, implying that its reproductive strategies were more similar to those of the brooding oviraptorids than those of the closely related-to-dromaeosaurids-but-megapode-like troodontids. However, if Parsons et al 2015 is right then baby Deinonychus were flying around anyways.

It should be noted that a few modern birds that do have brooding behaviours can, like megapodes, have good flying juveniles but downright flightless adults. These include snowcocks (McGowan 2002), flying steamer ducks, horned coots and some exceptionally large mute swans. In all of these species, the chicks are precocial, but are still flightless and dependent on their parents; they still develop the ability to fly relatively early on, and eventually become too heavy for it. If dromaeosaurids had brooding behaviours, either the young were already able to fly from birth or would have undergone a brief period of parental care before developing flight… and become flightless again. Perhaps in a much shorter and more dramatic way than their modern relatives.

As noted before, Velociraptor (and now Dakotaraptor) have quill knobs, a characteristic seen in some flying birds but almost never in flightless species and even some flying forms like flamingos; they’re also not seen in most other dromaeosaurids, such as the similarly sized Zhenyuanlong which does have massive forelimb feathers (Lü et al 2015). This suggests that, much as in modern birds, they correlate to particularly strong flight feathers, which certainly seems to indicate powered flight. While I think we’re just assuming Velociraptor‘s flightlessness (see below), I do have to concede that an adult Dakotaraptor was far too heavy to fly by any means known to animals short of aerial machinery, so this certainly lends credence to the hypothesis that the juveniles of these forms could fly and later lost the ability to do so.


Alien Equipment


Bats don’t need no keels, and neither did early birds (author: cmglee)

Also controversial, but considerably less so since this has been circulating for a while long before I even wrote anything, is the idea that early birds did more with less. Its no secret that non-Ornithothorace theropods lack the keel sternum and other skeletal features associated with flight musculature and a full flapping motion in modern birds; this is by far the main reason why the evolution of flight in birds is so controversial, since there are a variety of animals that by all means should be able to fly but seemingly lack the means to so much as move their arms up and down.

Looks can be deceptive, however. We know for a fact that some extinct birds, like Confuciusornis, relied on other muscle complexes than those seen in modern birds. Like bats and pterosaurs, there is a reliance on the deltoideus and other complexes, than anchor to the back; this is opposed to the coracoideus and other muscle complexes employed by modern birds, which attach solely to the sternum. Therefore, there was less of a pressure to develop large keeled sternums, and instead these animals focused more on having broad shoulders. A cursory glance at bat and pterosaur and volaticothere skeletons will gladly demonstrate this trend, though pterosaurs eventually also developed deep keels, just not as deep as those of Ornithothorace birds.

More complicated is the apparent incapacity of early birds to raise their forelimbs above their torsos, a problem bats and pterosaurs and volaticotheres did fix. Many studies across my lifetime have posited different conclusions: Zhou et al 2001, for example, posited that half a stroke is better than nothing, while Agnolín et al 2013 claims that the should position is in fact elevated as in modern birds.

By far the most convincing argument comes from Voeten et al 2018, which posits a rather radical solution: early birds had a different flight position and movements than modern ones, their torsos more upright and the arm moving in a more forward angle. In essence, somewhat like what people do when imitating birds.

02-archaeopteryx-flight“My hemorrhoids fuel me” by Jana Růžičková.

This has some precedent in that bats also fly at a rather different angle than modern birds, albeit in the exact opposite fashion; whereas pterosaurs and volaticotheres also had different flight positions remains to be seen, and I reckon this could offer massive insights in everything from aerodynamic efficiency to how flight developed among the various groups. It also addresses the function of the hindwing seen in so many early birds: as the picture above shows you, it comes in handy in this new flight position.

The study also offers unambiguous evidence of flight in Archaeopteryx, its bone structure and vascularisation being unambiguously more similar to that of modern birds and pterosaurs than that of flightless dinosaurs. In particular, it is in line with the flight style employed by fowl, tinamous, woodpeckers and Dimorphodon, characterised by short bursts. It may seem like poor flight capacities, but it is noted to be a rather specialised flight style that requires several unique adaptations; for instance, earlier pterosaurs are more efficient flyers than Dimorphodon (Witton 2013). Rather than be an earlier starter, Archaeopteryx is already a fairly specialised animal, implying that early birds were already diverse in terms of flight style. Certainly, looking at the swift-winged Microraptor, the soaring Sapeornis, the forest-bird like Confuciusornis and the bat winged Yi indicates that non-Ornithothoraces were flying in many different ways.

This begs the question as to why there was a switch, especially since other flying vertebrates never followed suit. Since birds are bipedal it could be a legitimate functional improvement, since they can’t launch quadrupedally and what applies to bats and pterosaurs and volaticotheres likely only makes things worse for them. Most likely, however, this was simply a matter of chance, as typical for evolution.


More Than Once?

800px-Pelecanimimus_restorationPelecanimimus by Diego Ortega. As noted by Mickey Mortimer, its sternum may indicate that it evolved from then recent flying ancestors.

A question I’ve tackled occasionally was when flight evolved in dinosaurs and how often. Unlike pterosaurs and bats and volaticotheres, which show up in the fossil record already specialised for flight and highly modified from any potential “missing link” (yes, even volaticotheres; we’ll talk about it later), theropod dinosaurs have a variety of forms that could be called “intermediary” in case of flight capacity, lots of them even. Its possible that flight evolved multiple times among theropods; compared to other vertebrate groups, they have an advantage in having free, proportionally large forelimbs with aerodynamic surfaces. But it’s just as likely that dinosaur flight only evolved once, and that the high amount of “semi-flyers” are either secondarily that way or never became more efficient flyers.

Dececchi et al 2016 makes the claim that flight evolved twice among dinosaurs: in the lineage that lead to modern birds and in microraptors. This naturally requires the stance that all maniraptors outside of Ornithothoraces and Microraptoria are flightless, which as Voeten et al 2018 and Parsons et al 2015 demonstrates is just plain not true. Therefore, flight was likely basal at least in regards to the last common ancestor between Archaeopteryx, Deinonychosauria and modern birds. That most deinonychosaurs are (supposedly) flightless is not any more indicative that microraptorines developed flight independently as tinamous did in spite of being surrounded by basically wingless ratites.

The earlier in the phylogenetic tree we go, the murkier things become. Maniraptora also includes forms like anchiornithids, which have bodily proportions similar to those of early birds and flying deinonychosaurs but have shaggy feathers that would make them basically useless to fly in, and the scansoriopterygids, strange creatures who have vastly different wing structures from those of other flying maniraptors. This should imply that the last common ancestor among all these critters was flightless, flight having thus evolved independently in scansoriopterygids* and in birds+deinonychosaurs. In the past I did try to argue for flight capacities in anchiornithids, but while I do still think that Anchiornis proper has bodily proportions more typical of flying dinosaurs I prefer to think of them as “proto-flyers”, being essentially the “missing links” between flightless theropods and flying ones.

It’s been suggested that basal oviraptorids could parachute (Currie 2004), and I do think that the bird-like wrist of more derived species is rather eyebrow raising. As their phylogenetic position is rather fluid, however, its difficult to say if they aren’t actually nested among the post-anchiornithid taxa, and at any rate I imagine they would have engaged in “proto-flight” behaviours rather than actual flight.

Other maniraptors for the moment don’t appear to be volant, and until further notice I consider them indeed ancestrally flightless or perhaps engaging only in the least sophisticated of “proto-flight” behaviours. But immediately outside of Maniraptora, we see potential adaptations for powered flight: Pelecanimimus, an early unusual basal ornithomimosaur, is noted as possessing large paired sterna, ossified ribs and uncinates, an unusual equipment compared to that of modern birds but comparable to that of other flying theropods like Archaeopteryx. This has first lead Gregory S. Paul to suggest a flying ancestry for ornithomimosaurs in his infamous book Dinosaurs of the Air, but this particular idea of him seems to have resonated as it was more recently reinforced by Mickey Mortimer.

Pelecanimimus itself seems to have been flightless as a adult, though naturally it is possible that it could fly as a juvenile. A flying ancestry for ornithomimosaurs certainly goes a long way of explaining their wings; though it is very likely that the preserved feathers of Dromiceiomimus are actually more simple than previously thought, in practise it would just mean that, much like modern ratites, they underwent a secondary simplification of their feathers, and that like modern ostriches they still retained large fluffy display devices.

Assuming this that all these taxa do not in fact have a much deeper common flying ancestor, flight evolved at least three times in dinosaurs: in ornithomimosaurs, in scansoriopterygids and in the bird+deinonychosaur lineage. Should we include oviraptorids, this increases to four; should we go with an independent origin for deinonychosaur and avian flight, this goes to five; and the number can actually go up if we consider hypotheses like an independent flight origin for Enantiornithes and Euornithes, in which case independent acquisitions of flight are almost as numerous as the taxa themselves. While this idea is amusing, I think the bare minimum is more plausible.

As before, I hold a non-gliding origin for avian flight. Birds have long been implicated in pretty good alternative models like Wing-Assisted-Incline-Running, the more classical “ground up” and a model based on the flapping motions of dromaeosaur arms. Only the most volant true birds show adaptations towards arboreality; even scansoriopterygids, typically considered a model of dinosaur tree-living, have been found to be more terrestrial and likely only periodically visited the arboreal realm (Xu et al 2015). That said, as many of these animals did live in well-forested regions and do show curved claws, I think its fair to suggest that they at least perched on trees to rest, much as modern fowl and other mostly-terrestrial birds.

A model I’ve been toying with for the origin of flight in non-bat tetrapods is the sifaka lifestyle, jumping from the ground to the trees and back again while fluttering the forelimbs, perhaps graduating into launching from the ground by flapping. The key here is the lack of actual climbing: pressure to move up and down with long jumps is the motivator to develop flight.

*An elephant in the room you might have noticed is how I’m unambiguously treating scansoriopterygids as true flyers when Xu et al 2015 thinks a gliding lifestyle is more likely. As always, the main reasoning for conclusion is the supposed lack of musculature, which I have already addressed above. Another posited argument is how the styliform bone might be cumbersome when flapping the wings; this has so far not been tested, and I wager the fact that it is far more robust than any other type of styliform bone seen in vertebrates if anything argues for flapping flight. The study also suggests that fluttering might have been employed much as in kakapos, so there is an inconsistency here. Ultimately, however, scansoriopterygids are so far only known from a specific fossil formation in mid-Jurassic China; such a restricted spatio-temporal range is more consistent with gliding animals than flying ones, though this could easily be due to the poor fossil record in this time period. Much work needs to be done in this regard.


Size Limit

Aiolornis reconstruction (artist unknown). At a wingspan of five meters, it’s nothing compared to the largest pterosaurs, but can other flying theropods match it?

An issue that is certainly relevant to flight in theropods is the size limit of flying species. The typical answer presented is that non-Ornithothorace flying theropods are generally small, something somewhat ironic considering the non-Ornithothorace Sapeornis is considerably larger than most Mesozoic flying Ornithothoraces, and now the microraptor Changyuraptor is in the same boat. The logic is nonetheless pretty straightforward: if Ornithothoraces, which have clearly superior adaptations for powered flight, struggle to become large, surely more basal theropods would be even worse off, right?

As stated before, its difficult to measure the sheer functionality of the flight musculature and skeletons of early flying theropods against anatomically modern birds. It is very likely that on the whole the Ornithothoraces system is more efficient, hence their quick dominance over Cretaceous avifaunas; however, non-Ornithothorace species still dominate the larger flying niches, so even if they were less efficient on the whole whatever they were doing allowed them to grow larger. The classical view is that these animals were inclined towards gliding, ergo they quickly exploited soarer niches; however, as discussed previously there’s no particular reason to assume that non-Ornithothorace theropods were any less specialised towards flapping flight than more derived birds, especially given their broad range of wing shapes and likely flight styles.

In many ways, this reminds me of the discussions in regards to giant pterosaurs: previously regarded as living kites, they are now seen as the most sophisticated flying animals of all time. Perhaps we can extend this same nuance to early flying theropods.

Speaking of pterosaurs, I have long argued that pterosaurs dominated larger aerial ecological niches, while Mesozoic birds remained comparatively small. This seems to indeed have been vindicated by Longrich et al 2018, which finds that Late Cretaceous birds rarely have wingspans above 1 meter, while pterosaurs dominate all size ranges above that – and conversely have few Late Cretaceous small forms. This has been compared to the similar ecological divide between non-avian dinosaurs and Mesozoic mammals, the latter classically living “in the shadow” of the former, while also preventing them from getting smaller.

However, I now find this view simplistic; many non-avian dinosaurs have been found to be small sized, while similarly we know several Mesozoic mammals reached at least wolverine size, and likely larger. Likewise, we know some Late Cretaceous azhdarchids were small (Prodvai et al 2014, Witton et al 2016), as was the mysterious Piksi. It only makes sense that some Mesozoic flying dinosaurs were large too, though on the bird side of the equation results have been inconclusive: Samrukia, a putative teratorn-sized Cretaceous flying bird, turned out to be a pterosaur, while there are footprints of a possibly crane-sized bird in Australian Cretaceous deposits (Martin et al 2013).

On the other hand, the non-Ornithothorace side seems to have been more productive, as shown by the aforementioned Sapeornis and Changyuraptor. The latter is particularly interesting because it is a dromaeosaur, retaining traits typically considered counterproductive to large flying animals like a long bony tail. Yet, at an estimated four kg it is large even by modern flying bird standards. This is considerably larger than Hesperonychus and Sinornithosaurus, microraptorines previously considered to be flightless because they are somewhat heavier than the raven-sized Microraptor. Microraptorines as a whole are fairly large flying dinosaurs when compared to contemporary birds, and probably occupied ecological niches nowadays taken by large omnivorous or carnivorous birds like corvids and raptors. Sapeornis, meanwhile, is compared to modern screamers, as a primarily herbivorous open-space soarer.

With this in mind, it’s not hard to think that many deinonychosaurs previously rendered flightless for being “too heavy” might have actually been capable of flight all along. The one non-microraptorine dromaeosaur with well preserved wing feathers, Zhenyualong, shows rather massive, asymmetrical ones. This animal is rendered nonetheless rendered flightless as the wing bones themselves don’t quite match the length of the wing feathers; as Mark Witton has pointed out recently, there is a certain balance in regards to the proportions of the bone and wing feathers. Still, the same can’t be applied to several other mid-sized dromaeosaurs, which emphatically have larger and more powerful arms.

Thus, I think judging dromaeosaur flight capacity is better judged by comparing limb proportions and the presence of quill knobs than some definite “maximum weight”. Only the largest species can safely be deemed too heavy to fly, and that may not encompass juveniles or even subadults. By contrast, troodontids probably can be dismissed as flightless more easy, since most have proportionally short arms at least as adults and generally lack quill knobs; perhaps unsurprisingly, troodontids and short armed dromaeosaurids tend to be more cursorial, as typical in the dichotomy between terrestrial and aerial birds of prey.

That dromaeosaurs are often compared to modern hawks and eagles might not be the most accurate description, but it does have several points of interest. Eagles and other large accipitrids usually hunt from a vantage point, and the same can likely be said of dromaeosaurs, which are among the least cursorial theropods. This hunting strategy is likely the main reason as to why eagles and hawks have retained the ability to fly even in ecosystems dominated by flightless birds like New Zealand, and it would certainly explain why several dromaeosaurids have retained the ability to fly at least as juveniles. Larger individuals, more likely to being able to scare off other predators from their kills, would have the pressure the remain aerial.


Velociraptor on Air

800px-Velociraptor_model_at_Museum_für_Naturkunde_in_Münster_in_GermanyVelociraptor model at Museum für Naturkunde in Münster by Phillip2001. Notice the giant ass wings that come from turkeys.

Finally, no topic with this name can be finished without Velociraptor itself. As by far the most iconic of the non-avian winged theropods, is a centerpiece in how feathered dinosaur discourse is presented, and certainly should have a thing or two to say about flight in dromaeosaurs.

The adult Velociraptor is usually held to be a flightless animal, as typical for most dromaeosaurids. For the most part this is “just because”, but works that feel the need to express in more detailed terms usually cite the arms not being long enough in proportion to the body size or it being too heavy. Traditional weight estimates render the latter explanation somewhat excuse-y, at around 15 kg – heavy, but still acceptable by flying bird standards -, but potential estimates of up to 19 kg – mute swans around this weight are thought to be flightless, but there were comparably sized flying birds, with the absolute largest like Pelagornis being twice heavier – could add some weight to these claims (Campione et al 2014).

By contrast, less can be said about the arms. As previously indicated, they bear quill knobs, which already ring bells. The arm length of the adult animal does come short compared to those of unambiguously volant theropods like Microraptor and Archaeopteryx, but it is still proportionally long and the hand is proportionally longer, which could have further compensated as a site of flight feather attachment. The overall picture are large but low-aspect ratio wings, typical for flying animals in inland settings.

My strongest bet overall is a type of inland soaring flight, which would have been facilitated by the dune desert environment it occurred on. With relatively little cover or perching sites, Velociraptor might have been more adapted to search long distances in search of food much like modern vultures, which explains why its jaws are similarly more gracile as opposed to the robust jaws of other large dromaeosaurids; this is also seen in other scavenging flyers like vultures and istiodactylids, an indication that they do not usually have to deal with struggling prey.

Large adults, being less capable of launching and more robust, would have hunted larger prey on the ground; however, if this shift even happened it must have been fairly late in life, as unlike other dromaeosaurids the jaws of Velociraptor don’t become deeper with age. The same happens with most other velociraptorines, implying that this clade as a whole occupied a guild of somewhat uncommon scavengers across the northern hemisphere, much like mid-Cenozoic gypaetine vultures.

So yes, all the lammaergeier Velociraptor pictures are warranted.



Flying raptors
I will not feel sorry for using BBC as it is evil and should not have a disclaimer.

I can never be wrong.



Fernández, Mariela S.; García, Rodolfo A.; Fiorelli, Lucas; Scolaro, Alejandro; Salvador, Rodrigo B.; Cotaro, Carlos N.; Kaiser, Gary W.; Dyke, Gareth J. (2013). “A Large Accumulation of Avian Eggs from the Late Cretaceous of Patagonia (Argentina) Reveals a Novel Nesting Strategy in Mesozoic Birds”. PLoS ONE. 8 (4): e61030. doi:10.1371/journal.pone.0061030. PMC 3629076 Freely accessible. PMID 23613776.

Parsons, William L.; Parsons, Kristen M. (2015). “Morphological Variations within the Ontogeny of Deinonychus antirrhopus (Theropoda, Dromaeosauridae)”. PLoS ONE. 10 (4): e0121476. Bibcode:2015PLoSO..1021476P. doi:10.1371/journal.pone.0121476. PMC 4398413 Freely accessible. PMID 25875499. e0121476.

D. C. Deeming; G. Mayr (2018). “Pelvis morphology suggests that early Mesozoic birds were too heavy to contact incubate their eggs”. Journal of Evolutionary Biology. in press. doi:10.1111/jeb.13256.

Harris, R. B., S. M. Birks, and A. D. Leaché. 2014. Incubator birds: biogeographical origins and evolution of underground nesting in megapodes (Galliformes: Megapodiidae). Journal of Biogeography 41:2045-2056.

Grellet-Tinner, G.; Makovicky, P. (2006). “A possible egg of the dromaeosaur Deinonychus antirrhopus: phylogenetic and biological implications”. Canadian Journal of Earth Sciences43(6): 705–719. Bibcode:2006CaJES..43..705Gdoi:10.1139/E06-033.

Madge, Steve; McGowan, J. K.; Kirwan, Guy M. (2002). Pheasants, Partridges and Grouse: A Guide to the Pheasants, Partridges, Quails, Grouse, Guineafowl, Buttonquails and Sandgrouse of the World. A. C. Black. pp. 174–180. ISBN 9780713639667.

Lü, Junchang; Brusatte, Stephen L. (2015). “A large, short-armed, winged dromaeosaurid (Dinosauria: Theropoda) from the Early Cretaceous of China and its implications for feather evolution”. Scientific Reports5: Article number 11775. doi:10.1038/srep11775PMC 4504142Freely accessiblePMID 26181289.

Zhou Z. and Farlow, J.O. (2001) “Flight capability and habits of Confuciusornis“. In: Gauthier and Gall (eds). New perspectives on the origin and early evolution of birds: proceedings of the international symposium in honor of John H. Ostrom. Peabody Museum of Natural History. Yale University, New Haven. pp. 237–254

Federico L. Agnolín & Fernando E. Novas (2013). “Avian ancestors. A review of the phylogenetic relationships of the theropods Unenlagiidae, Microraptoria, Anchiornis and Scansoriopterygidae”. SpringerBriefs in Earth System Sciences: 1–96. doi:10.1007/978-94-007-5637-3.

Voeten, Dennis F.A.E.; et al. (13 March 2018). “Wing bone geometry reveals active flight in Archaeopteryx”. Nature Communications9 (923). Bibcode:2018NatCo…9..923Vdoi:10.1038/s41467-018-03296-8. Retrieved 13 March 2018.

Mark P. Witton (2013), Pterosaurs: Natural History, Evolution, Anatomy, Princeton University Press, ISBN 978-0-691-15061-1

Dececchi, T.A.; Larsson, H.C.E.; Habib, M.B. (2016). “The wings before the bird: an evaluation of flapping-based locomotory hypotheses in bird antecedents”. PeerJ4: e2159. doi:10.7717/peerj.2159PMC 4941780Freely accessiblePMID 27441115.

Currie P.J. (2004), Feathered dragons: studies on the transition from dinosaurs to birdsIndiana University Press, p. 184, plate 16.

Xu, X.; Zheng, X.; Sullivan, C.; Wang, X.; Xing, L.; Wang, Y.; Zhang, X.; o’Connor, J. K.; Zhang, F.; Pan, Y. (2015). “A bizarre Jurassic maniraptoran theropod with preserved evidence of membranous wings”. Nature521 (7550): 70–3. doi:10.1038/nature14423PMID 25924069.

Nicholas R. Longrich; David M. Martill; Brian Andres (2018). “Late Maastrichtian pterosaurs from North Africa and mass extinction of Pterosauria at the Cretaceous-Paleogene boundary”. PLOS Biology. 16 (3): e2001663.

Prondvai, E.; Bodor, E. R.; Ösi, A. (2014). “Does morphology reflect osteohistology-based ontogeny? A case study of Late Cretaceous pterosaur jaw symphyses from Hungary reveals hidden taxonomic diversity”. Paleobiology40 (2): 288–321. doi:10.1666/13030.

Martin-Silverstone, Elizabeth; Witton, Mark P.; Arbour, Victoria M.; Currie, Philip J. (2016). “A small azhdarchoid pterosaur from the latest Cretaceous, the age of flying giants”. Royal Society Open Science3 (8): 160333. doi:10.1098/rsos.160333.

Martin, Anthony J.; Vickers-Rich, Patricia; Rich, Thomas H.; Hall, Michael (2013). “Oldest known avian footprints from Australia: Eumeralla Formation (Albian), Dinosaur Cove, Victoria”. Palaeontology57: 7–19. doi:10.1111/pala.12082.

Campione, Nicolas E.; Evans, David C.; Brown, Caleb M.; Carrano, Matthew T. (2014). “Body mass estimation in non-avian bipeds using a theoretical conversion to quadruped stylopodial proportions”. Methods in Ecology and Evolution5 (9): 913–923. doi:10.1111/2041-210X.12226.

Aquaflying vs Foot Propelled Swimming

The infamous "wing propelled cormorant" picture. Credit to "Marek Sk".
The infamous “wing propelled cormorant” picture. Credit to “Marek Sk”.

With inflexible torsos and short tails, birds are forced to engage in two main forms of underwater locomotion: aquaflying or foot propulsion. There are variations, mostly in the latter (compare the “frog-like” motions of grebes and loons to vertical strokes of cormorants, for instance), but ultimately specialised diving birds end up in two molds, the flipper winged “penguin like” forms, or the hindlimbs-at-the-end-of-the-body “loon like” forms. There isn’t much margin of overlap between the two, with nearly all aquatic birds known either opting for one method or the other. Either you develop strong oar-like wings, or you don’t and let them degenerate in favour of stronger or otherwise modified hindlimbs.

That is conventional wisdom. As it turns out, many birds previously thought to be exclusively foot-propellers actually do engage frequently in aquaflying, using it almost as common and effectively as their “conventional” swimming. Most notable among these are diving ducks and cormorants, traditionally thought to be outright incapable of wing propulsion (Upstroke Thrust, Drag Effects, and Stroke-glide Cycles in Wing-propelled Swimming by Birds), but clearly very much capable of doing such.

Indeed, one notable lineage of aquafying birds, Plotopteridae, may have evolved from foot propelled divers, being closely related to modern cormorants and darters, and possibly retained it quite late in their evolutionary history.

The differences in effect between aquaflying and foot propulsion are rather clear: aquaflying has been proven to ensure faster speeds, while foot propulsion results in higher manouverability (Schreiber, Elizabeth A. & Burger, Joanne (2001) Biology of Marine Birds). Tellingly, aquaflying is most prevalent in oceanic pursuit predators like penguins, auks, petrels and seagulls, while foot propulsion is common among freshwater and coastoal birds (grebes, loons, ducks, cormorants, darters, finfoots). There are however obvious exceptions to the rule, like the freshwater dippers and the (largely) marine Hesperornithes.

In the aforementioned “foot propeller turned into aquaflyer” cases, it appears to occur primarily in marine settings, in which foot propulsion becomes less beneficial than wing propulsion. In ducks, the dive is initiated by aquaflying, and only engaging in foot propulsion when it reaches it’s feeding grounds in the ocean bottom. Most likely, aquaflying speeds up the process of descent, indeed confirming that it is a faster method of locomotion than foot propulsion on average. In the case of the cormorants, it seems to have started in the same way, using aquaflying to reach the ocean bottom faster, but since these birds are fish predators a continuously fast speed is necessary, so the animal only seems to resume to foot propulsion if the prey is stactic. From this point on, foot propulsion is best done with altogether, the animal achieving more success by continuing to rely on it’s wings.

Most freshwater birds do not have this pressure, simply needing to round up fish against trapping environments, so aquaflying seems less favoured than sheer manouverability. However, dippers and similar birds do need the sheer wing muscle power in order to compensate for their small size and higher bouyancy, so they became aquaflyers in the first place. Generally speaking, foot propulsion seems to correlate with lack of speciation, with only loons, grebes and Hesperornithes seemingly having become more specialised in it than switching to aquaflying altogether.

For now, it is too early to establish a consistent pattern to the evolution of aquatic locomotion in birds, but it is clear that there is less of a dichotomy than previously thought

Speculative Evolution: Masticating Waterfowl

Goose lamellae up-close
Goose pecten up-close

Birds can’t chew; that is why they have gizzards. After all, you can’t do much with a beak, which doesn’t have teeth and lets the bolus fall off. Older non-avian dinosaurs like hadrosaurs, ceratopsians and even some therizinosaurs could process food with their teeth, but birds cannot, and this has been used as the main reasoning as to why modern ecosystems have a distinct lack of giant herbivorous birds, while herbivorous mammals are abundant (neverminding that ostriches, rheas, elephant birds, eogruiids and dromornithids have co-existed throught the Cenozoic with large herbivorous mammals, and that tortoises and other non-masticating critters have been able to effectively compete with mammals as well).

Except that’s not really the case. Some birds, like hoatzins and cuckoos, seemingly can engage in some rudimentary form of mastication, using their bills to triturate food before consuming it (Korzun LP et al. (2003) Biomechanical features of the bill and and jaw apparatus of cuckoos, turacos and the hoatzin in relation to food acquisition and processing. Ostrich 74: 48-57.). The main counter argument against mastication in birds, that the bolus would fall off without lips of any sort, has been counter argued before, and many birds developed tooth-like structures in their bills. So, birds with the capacity to masticate, to chew effectively, is definitely a possibility, provided the right set of circumstances occurs.

Which brings me to the most likely candidate for such an evolutionary path, anatids. Anseriformes are notable for their bizarre tooth-like serrations, the lamellae/pecten. Pecten are present in all waterfowl, and seemingly evolved for filter feeding, but they have since been refined for a variety of purposes, such as the fish-grabbing, truly tooth-like merganser pecten:

pick 007 b

In herbivorous waterfowl like geese and swans, the pecten serve mostly to help the animal rip out plant material, much like the grooves in the beaks of other herbivores like tortoises, or the teeth of non-masticating herbivorous animals like iguanas. Jumping from simply ripping to trituration would be easy; it’d just start as additional bites before consumption, then progress into full blown mastication. Some adjustments would be necessary, mainly in changing the shape or orientation of the pecten in order for them to be useful in trituration, but as seen before they are plastic enough for this to be possible.

More interestingly, waterfowl mastication could be facilitated due to another characteristic: cranial kinesis. Like many other birds, Anseriformes can flex their upper jaw, largely thanks to an articulation that seperates it from the rest of the skull. In a controlled form, this could help to progress the development of mastication a lot, as all the bird would have to do would be to move the upper jaw – or sections of it – as the mouth closes, making the pecten slide against each other, instantly pulverising any plant matter in the beak. This method of mastication would be very similar to that practised by hadrosaurian dinosaurs, which engaged in a more controlled form of maxillary kinesis:

Doing this would mean a very efficient form of chewing with minimal changes to the pecten’s orientation or shape.

Furthermore, remember the aforementioned lack of necessity for lips? Waterfowl “beaks” are largely just composed of keratin-reinforced skin, with the rhamptheothecae limited to the “nails” at the tip of the jaws. As you can see in the above pictures, “lips” of a sort already exist, so if there was an actual need for lips, it’d be solved rather quickly.

Five reasons as to why Proapteryx is interesting

Peter Schouten's Proapteryx picture accompanied by more familiar kiwi (geddit) characters like a tuatara-like sphenodontian and mysticine bats.
Peter Schouten’s Proapteryx picture accompanied by more familiar kiwi (geddit) characters like a tuatara-like sphenodontian and mystacine bats.

So 2013 was a big year for our understanding of extinct birds. We know that phorusrhacids terrorised Europe, that early birds didn’t have scaly feet, a new understanding of the enantiornithe cladogram, et cetera. One of these discoveries is that small, unassuming New Zealand bird known as Proapteryx micromeros, a bird the size of a rail (and actually compared to the Banded Rail, Gallirallus philippensis, in terms of leg proportions) that stalked Aotearoa’s forests in the early Miocene, where is now Otago. It lived in a New Zealand already populated by a strange and alien fauna, composed of animals like terrestrial mekosuchine crocodiles, turtles, pythons and even a mysterious, poorly understood non-chiropteran mammal. Birds were already dominant, however, and the site where it was found, the Saint Bathans Formation, already had a very recognisable avifauna, with ducks, seagulls, passerines, herons, owlet-nightjars, hawks and even recognisable forms of Moa and Adzebill running around. While there were a few oddbirds here and there, like a palaelodid flamingo, the avifauna of Saint Bathans was distinctively modern, conforming to the Neogene bird faunas, with the odd, ancient birds already rendered into flightless giants while more derived and familiar bird groups dominated the skies.

There was, however, one component of the familiar avifauna in Saint Bathans that was absent, and that was the kiwi, the bizarre, nocturnal, terrestrially probing, hairy ratite that we all know and love. As the name implies, Proapteryx was an early form of kiwi, but it was a rather different animal. Unlike the moas and adzebills, it wasn’t already a large, flightless giant, but a still very small animal that, based on limb proportions, was probably still capable of flight or at least had evolved from then very recently flying ancestors. This kiwi wasn’t therefore an early relic from a time when birds could afford to not be neognaths, but a recent coloniser of the island continent like the pigeons and herons it co-existed with, and by all accounts a typical member of the Australasian avifauna of the period, making the absence of volant palaeognaths from modern Oz all the more conspicuous. It’s discovery bears therefore a lot of meaning in the study of not only kiwis, but also palaeognaths as a whole, and of the changes in avifauna in a landmass once presumed to have been exceptionally conservative.

1. It confirms that the known Palaeognath clades are [mainly] not the result of vicariance

The most obvious conclusion advertised in articles about this bird is that the ages old hypothesis that ratites all evolved from a common ancestor that diverged as Gondwanna split apart is effectively disproven. Previously, it had already been under fire across the 20th century, with genetic data connecting kiwis to australian ratites instead of moas, and Hackett et al. 2008 that demonstrated that palaeognath diversity as a whole is independent from the progression of Gondwanna’s split up (i.e. tinamous are more closely related to Australasian ratites than to rheas, elephant birds are more closely related to either than the indo-african ostriches, et cetera.). Proapteryx not only offers evidence that palaeognaths cross the Tasman Sea in the Miocene, well after New Zealand had already become an island, but also that kiwis evolved independently from moas, as Saint Bathans already preserves fossils of large, flightless dinornithids while it’s sole kiwi is a dimininutive flying or then recently flightless bird.

2. It implies a higher diversity of mid-Cenozoic palaeognaths than previously thought

While palaeognath fossils are rare, period, it’s clear that these birds were quite diverse in the Palaeogene. Fossil sites in Eurasia and North America dating bear fossils of birds known as “lithornithids”, a diverse menagerie of flying palaeognaths tentatively classified as a monophyletic clade, but actually more likely to represent a paraphyletic assemblage of birds leading to modern ratites, something especially likely when one examines basal ostriches like Palaeotis. “Lithornithids” disappear from the fossil site in the Oligocene, alongside many other laurasian bird clades, and with their extinction it was presumed that the only flying palaeognaths that remained were the already poorly-flying tinamous, surrendering the skies to neognath birds for the rest of the Cenozoic.

With the discovery of a flying kiwi dating to the Miocene, this obviously wasn’t the case. Proapteryx is a small bird, actually smaller than the largest known “lithornithids” such as Paracathartes, and it clearly evolved from animals that were very competent flyers, if it wasn’t a competent flyer itself, indicating the presence of rather generalistic flying palaeognaths in the Miocene of Australia and New Zealand. Not only that, since kiwis are most closely related to autralian cassowaries and emus, the presence of the already large, flightless Emuarius in the Oligocene of Australia suggests that aussie ratites were undergoing an adaptative radiation in the Oligocene and early Miocene, branching in at least two radically different ecological niches, instead of being reduced to flightless forms as previously thought. More volant palaeognaths are expected to turn up in Oligocene and Miocene fossil sites in mainland Oz, if they haven’t already and simply have been misidentified (as sadly often the case when dealing with fragmentary bird remains).

3. It offers a possible behaviorial connection between kiwis and “lithornithid grade” palaeognaths, and perhaps an insight to the ancestral ratite condition

While Proapteryx is explicitly contrasted against Lithornis in Worthy et al. 2013 (namely, in the morphology of the quadrate bone and associated elements), there are possible similarities between the paraphyletic “lithornithids” and kiwis raised by it’s discovery. “Lithornithids”, while most likely subjected to a wide variety of lifestyles, generally possess long, slender jaws, which have been implicated as having been used in probing, often compared to those of shorebirds; given their hypothesised arboreal tendencies, they might have been ecologically analogous to hoopoes and woodcocks, introducing their jaws on the soil or in cavities in the trees to search for insects and other small prey.

This obvious raises a strong parallel to modern kiwis, which are birds specialised to probe on the New Zealand forest substrates for foraging, having sensor pits, the nostrils at the tip of the jaws, vibrissae and other adaptations that evolved to detect prey in this manner. Previously, these weird features were thought to be the product of island life, but the discovery of a volant kiwi might very well establish these as having predated flightlessness, presumably having evolved in “lithornithids” and having simple been retained, if slightly refined, in extant kiwis.

Indeed, it might very well suggest that kiwis are actually the most ecologically conservative of all extant palaeognaths, having remained in a probing, woodcock like lifestyle while rheas, emus and ostriches became plains dwelling herbivores, cassowaries became frugivores and tinamous became fowl-like omnivores and granivores.

4. It offers insight about the evolution of kiwi flightlessness

In the New Zealand that Proapteryx lived in, mekosuchine crocodiles offered predation to the native avifauna, while small mammals of uncertain affinities competed with them. It is easy to understand why moas, waterfowl and adzebills became massive and wingless, as there simply weren’t any other animals occupying the roles of large herbivores and of opportunistic carnivores respectively, but the ecology of Saint Bathans did not showcase the sheer variety of flightless birds that New Zealand is famous for, implicating that the competitive and predatorial pressure from the crocodiles, mammals and perhaps even the local lepidosaurs and turtles was a restricting factor in that biota. Certainly not the environment where one would expect a flightless insectivore, and yet the kiwi outlived the competing mammals in it’s island continent.

It is possible that one of the more bizarre aspects of kiwi biology, the proportionally massive eggs, evolved in response to flightless in such an environment, allowing larger and more independent chicks, both the eggs and juveniles being less vulnerable to mammal or mekosuchine predation. It is now clear that this isn’t the result of allometry, as kiwis evolved from small flying birds and aren’t dwarfs as previously thought. Thus, kiwis may have responded to a rather competitive environment, instead of being the result of 20-16 million years of lack of negative pressures.

5. It offers insight about the progression of palaeognath avifaunas in the Neogene

As showcased before, Proapteryx is a game changer in our understanding of palaeognaths, implicating diversity previously unheard off. It is also an indicator of a rather conspicuous faunal turn over, as there aren’t any more flying palaeognaths in Oz, in spite of the continent’s overall rather conservative bird faunas. Given the rather obvious diversity of many derived bird groups in the Oligocene and Miocene of Australia, and the definite presence of many recognisable birds in the Saint Bathans Formation, such as passerines, rails, shorebirds, owlet-nightjars, parrots and pigeons, competition and eventual ecological displacement by neognaths seem to be unlikely, at least on on a systematic, large scale, especially when “lithornithid grade” palaeognaths don’t seem to have an equivalent in modern australasian avifaunas. Rather, it is possible that Australia’s late Miocene/Pliocene climatic changes spelt doom for Proapteryx like birds, as they have for several other volant terrestrial insectivores like mystacine bats. Flying palaeognaths seem to generally have fared poorly in conditions of mass drying, as evidenced by the disappearence of “lithornithids” from Laurasia and possibly of non-aepyornithid palaeognaths from Madagascar, the only clade seemingly adapting well being nothurine tinamous, which diversified in South America’s Miocene grasslands. By contrast, flightless palaeognaths have basically prospered, albeit in rather different ecological niches, the patterns of which worth investigating in the near future.


Miocene fossils show that kiwi (Apteryx, Apterygidae)
are probably not phyletic dwarves (Worthy et al. 2013)

Paleognathous Birds from the Early Tertiary of the Northern Hemisphere (Peter W. 1988)

Jones et al. 2009

Miocene mammal reveals a Mesozoic ghost lineage on insular New Zealand, southwest Pacific (Tennyson et al. 2006)

Hackett et al. 2008

Turvey et. al 2005

Phillips et al. 2010

Notes on The Speculative Dinosaur Project’s Enantiornithes

In keeping with scientific accuracy:

– No retrices. Only tweeties and some Allospiziformes have well developed tail-feathers, and they’re more ribbon-like feathers like those found in longipterygids and kin. I’m considering expanding this to some avisaurids, but generally enants shouldn’t have tails beyond their long pygostyles.

– With the exception of tweeties and Allospiziformes, most have well developed wing claws. Note that the fingers are fused as in euornithes, so there should only be two visible claws.

– With the exception of a few oddballs, most are highly precocial, with at most a week or so of parental care. This doesn’t change from old!Spec.

– There should be some notice to their aerial locomotion. Enantiornithes are pretty much really weird when it comes to sternal and shoulder anatomy, and it’s been made very clear that these things were flying in a radically different way from euornithes. Tweeties might converge with euornithes in flight style, but this depends on whereas interpretations on longipterygid flight resembling that of euornithe birds more than that of other enants is a reasonable assumption or not.

– Most should have hindwings or feathery hindlimbs; no scutes whatsoever. This should be particularly interesting on aquatic forms like ebergs.

Established Enantiornithe clades:

– Gondwanaviformes: weird arboreal sloth-hoatzin things that include the Falsa Panha, global tropical folivores.

– Avisauriformes: the classical “raptors” + the frigatebird-like jarilos.

– Enabaptiformes: grebe-like ebergs. More concepts can be made from these, as it is a pretty underdeveloped clade.

– Twitiaviformes: tweeties and kin. Basically I think they can be invisioned as longipterigids with beaks, though of course some are highly divergent, such as the swift-like mistriders. Speaking of swifts, Spec Apodiformes may be remade into mistriders.

– Allospiziformes: parrot-like forms that are far more speciose than HE parrots, such as finch-like otherworld finches.

A few odd relics can be present here and there, though ideally all modern enants are within these clades, all originating as far back as the Cretaceous.


– Twitiaviformes diverged first; within it Twitiavidae is the most basal (if a valid clade and not a paraphyletic grade), with the remaining four “families” being group as Australoenantiornithes, composed of an Arborogemnidae + Pseudartamidae clade and a Therizinorhynchidae + Balaclavidae clade. Other clades are welcome.

– All other Enantiornithes belong to Euenantiornithes. Enabaptiformes diverged first, followed by Gondwanaviformes; this leaves a clade composed of Avisauriformes + Allospiziformes, something akin to the falcon + parrot + passerine clade within Neornithes.

– Within Avisauriformes, Avisauridae diverged first (again, if a valid clade and not a paraphyletic grade), leaving an Allostrigidae + Eupelagornithidae clade.

– Within Allospiziformes, otherworld finches are well established to be a paraphyletic grade. There may or may not be a crown clade of Civilispizidae + Apatopsittaciformes.

On The Origins Of Flight: No Gliders?

All things considered, vertebrate flight is the ultimate triumph of animal bioengineering, producing active locomotion on the thinnest of availiable mediums, releasing the individual from the ground and opening it to an entirely new realm to navigate upon. Flight has led to massive adaptative radiations and ecological niches completly impossible for terrestrial or aquatic animals to explore, and it leads to exaptations for other credible feats, like extremely effective lungs and large brains. Flight is freedom, and there are ten thousand reasons as to why we pitiful land primates have envied birds, bats and insects, and most surely pterosaurs if they were still alive.

Yet, when one looks at the groups of flying animals that have existed, one thing becomes apparent: it’s very, very rare for an organism to develop true, powered flight. Birds, bats, insects and pterosaurs are and were incredibly speciose and morphologically diverse, but immense adaptative radiations were all resulted from a mere four successful attempts at being an aeronaut. All the +10,000 living bird species, and their possibly millions of extinct relatives? Descendents from a single dinosaurian aeronaut back in the Jurassic, whose aerial prowess quickly kickstarted an adaptative radiation more or less at the same time pterosaurs were during their golden age. Pterosaurs themselves, from the miniscule anurognathids to the ginormous azhdarchids and ornithocheiroids, all descended from a currently unidentified form of sauropsid that took to the skies in the Triassic, a single species possibly akin to Scleromochlus. Going hundreds of million years further in time, we have the single common ancestor of over 60% of all animal life on the planet, the winged arthropods, and forward in time to the earliest Cenozoic we have a small, shrew like mammal who has made ghastly wings out of it’s hands.

Four species. That’s not even 0.01% of all animal life, no matter how mind crushingly diverse their descendents would be.

The sheer minority of successful aeronauts becomes even more puzzling when one witnesses the incredible diversity of gliding and parachuting animals, both from the past and presence. Among synapsids, extensive gliding has evolved at least 11 times, with the earliest known gliding mammal known from the Jurassic/Cretaceous boundary, the bizarre Volaticotherium, while squamates have produced gliders among all major lineages, to say nothing of the sheer amount of Triassic gliding reptiles. Some frogs developed long toes and webs that seem like smaller mockeries of bat wings, while countless fish species have turned their pectoral fins with airfoils and taken to the air; in fact, one lineage of said fish, Thoracopterigids, are quite possibly the oldest vertebrate group to have taken into the air. To say nothing of all the spiders and other arhtropods that glide and parachute, with webs or sheer chitinous extentions, as well as of the pelagic flying squid.

Yet, once again, all this incredibly diverse menagerie has achieved nothing akin to the powered flight of birds, insects, bats and pterosaurs, with the possible exception of a few fish, which propell themselves into the air with their pectoral fins. Indeed, many of these groups have become extinct, with no known close relatives and certainly no radiation of aerial critters derived from them.

It has been suggested that the presence of aerial vertebrates would have limited the presence of availiable niches for experimental flyers, something that falls extremely falt when A) birds evolved at the peak of pterosaurian diversity and bats when birds were already very well established (and posssibly even further back, when pterosaurs were still around), B) the idea that any of these groups have ever outcompeted or competitively excluded each other is at best severel uncircumstancial, and C) there’s plenty of ecological niches each individual group has occupied that the others haven’t (i.e. birds haven’t ever produced robust, boar-like terrestrial molluscivores like dsungaripterids, bats have never produced freshwater filter feeders, et cetera). Gliding animals also don’t seem to competitively exclude each other, so there’s little reason why animals in between gliding and full volancy wouldn’t have enough ecological space to exist.

An alternative option is what is reflected by an examination of gliding animal groups: gliding simply does not translate into flight, simply never leads to powered volancy, and that the flying vertebrates we know and love, as well as insects, have simply evolved under extremely anomalous sets of behaviour completly unrelated to gliding.

This is what I’m going to explore today.

What Living Gliders Tell Us

Colugos are about as close to bat-grade as any known non-chiropteran mammal has ever got...and that's as far as they will ever go.
Colugos are about as close to bat-grade as any known non-chiropteran mammal has ever got…and that’s as far as they will ever go.

Living animals and their extreme speciation compared to their long gone cousins can be exceptionally deceiving when it comes to studying the past, but they nonetheless offer a few clues about how flight may have aroused within the known flying vertebrates. As mentioned previously, there is a conspicuous absence of gliding vertebrates in a state of tentative flight, while the few fish that jump using their pectoral fins are not related to any gliding species. This alone can be considered pretty damning for the hypothesis that flight evolved from gliding, though it’s obviously not enough to establish a certain correlation.

Among gliding vertebrates, we witness a very wide range of adaptations for aerial locomotion, from the simple membranes of Holaspis to the complex airfoils of gliding squirrels. Some of these gliders have gone into extra lengths to acquire aerodynamical prowess, developing elaborate tail rudders (Ptychozoon), unique structural support for their wing membranes (such as the cartilage spurs of several flying squirrels and the ribbed wings of several squamates), proportionally massive wings (colugos, Draco lizards, et cetera), and even airfoil adaptations convergent with the aerodynamic design of flying animals (Exocoetidae). Some of these critters can be considered true aerial animals of their own right, travelling long distances on air and spending a good portion of their time gliding, though of course they are far more limited than true flyers.

Within the massive variety of living gliders, it’s evident that some will never develop powered flight: the rib wings of Draco and similar squamates, for example, cannot develop the necessary musculature and articulation to form anything other than a simple parachuting airfoil, and indeed extinct sauropsids with similar adaptations didn’t seem to have gone anywhere beyond a few “one shots”. Likewise, as suggested by Mark Witton 2013, flight is presumed to be unlikely to develop in ectothermic animals, as a warm blooded metabolism is likely a necessary exaptation in order to provide for the energetic demands of true volancy. This leaves pretty much every single living non-mammalian glider alive today as invalid candidates for powered flight, and therefore ill-suited models for how flight developed in bats, birds and pterosaurs. Thankfully, the menagerie of living gliding mammals is large enough to supply enough samples as to determine whereas flight can develop from gliding or not. And within these, two groups in particular can be held in particular regard.

The first, erroneously named “flying lemurs”, are often described as the most aerially adapted of all gliding mammals and just a step beneath the full volancy of chiropterans, having a well developed protopatagium and uropatagium, very long limbs that produce a massive brachiopatagium, and even a significant cheiropatagium, the fingers being quite long and displaying well developed webbing; in total, a colugo’s patagium is as massive as geometrically possible (Kathy MacKinnon 1984), and all the animal needs to develop functional wings is to elongate the hand more. They are capable of gliding for as much as 70 meters without losing altitude significantly, and their wing membranes indeed often an effective aerodynamic profile, the propatagium angled in the same as as that of birds and bats (and, presumably, as in pterosaur), the fingers serve as bastard wings and the uropatagium as an effective rudder (see image above). Such is their aerial competence that they were for the longest time assumed to be stem-bats, a relictual lineage wrestled out of the skies by more derived chiropterans, but genetic and morphological studies firmly place them as sister-taxa to primates, meaning that they became aeronauts independently (fringe “megabats are primates” pseudo-science nonwithstanding).

Thus, instead of being “living fossils”, colugos represent a lineage of gliding mammals that came very close to volancy… and stopped right there. The fact that the oldest members of this clade may date as far back as the Paleocene, when the first recognisable bats flapped their wings, means that they presumably had enough time to go go an extra mile and become true flyers, though in the interest of fairness we do not know if any extinct dermopterans were capable of gliding and if the aerial adaptations seen in modern forms are extremely recent. Colugos are specialised folivores, so it has been argued that powered flight is simply out of reach for these mammals, as folivory provides relatively little energy input, and gliding is a supposedly cheap way to travel around, while flight is obviously very energy demanding.

However, recent studies showcase that gliding is actually very energy taxing for colugos (Byrnes, G., Libby, T., Lim, N. T.-L. and Spence, A. J. (2011)), wasting as much as 1.5 times more energy by travelling through the air than by clambering amidst the branches. Part of the reason this is so energy demanding is due to the need to climb in order to achieve enough altitude for the next glide, often needing to climb as much as 8 meters for a non-problematic 30-50 meter “flight”, trouble flying animals do not have to deal with. Indeed, based on the studies’ conclusions, flight would have been a relatively cheaper mode of locomotion, and likewise the one flying animal with the closest lifestyle, the hoatzin, manages to be a specialised folivore and waste relatively little energy by engaging in short bursts of flight, less impressive than the colugo’s majestic “soaring” but certainly without the need to climb 8 meters in order to move around.

Thus, flight in colugos would have been relatively advantageous, and yet they have not made any further leaps into volancy throught the Cenozoic.

The other group, the iconic flying squirrels, come right behind colugos in terms of aerial speciation. While not having such extensive patagia, flying squirrels have well defined propatagia, brachiopatagia and uropatagia, and most importantly they are the only gliding mammals that have elongated the forelimb – the first step into the formation of a true wing -, developing unique cartilaginous wrist spurs that support the dystalmost part of the brachiopatagium. True to form, this development makes them even better gliders than colugos, capable of “flights” of 90 meters and more, and are seemingly better at steering in the air (Flaherty, E.A.; M. Ben-David, W.P. Smith (2010)).

No similar studies have been performed on whereas they spend more energy gliding, but if the studies on colugos are of any indication then flying squirrels should logically be also subjected to the same energetic pressures. Being omnivores, they’re also in a better position to develop powered flight, and in fact they presumably would benefit more from flight, as it would reduce energy espenses and aid them massively when foraging. Flying squirrels are thought to be a monophyletic lineage dating to the late Oligocene, which would have granted these rodents enough time to develop powered flight. Yet, as obvious, they remain exclusive, if particularly specialised, gliders. Once again, competition appears to not be an issue: no bats, and indeed no birds or even other sciurids, forage in the same manner as flying squirrels.

A third, ancient lineage of living gliding mammals is represented by the anomalures. Relatively basal rodents, these animals have been around since the Eocene, and not only haven’t become true aeronauts, they are also relatively cumbersome gliders, having a small propatagium and a bizarre brachiopatagium supported by an elbow spur, that helps stabilise the membrane but reduces any significant expansion of it. Flight, in this case, might had been prevented by the “flaw in design” rather than lack of opportunity.

Gliding marsupials are recent arrivals, having evolved around the latest Miocene; with relatively simple wing membranes, gliding is still on it’s infant steps for most of these animals, let alone powered flight. Outside of these lineages, there are a vast variety of mammals that have been reported to parachute, but of these only the sifakas and their possible rudimentary patagia show any possible adaptations for aerial locomotion in living mammals.

With at least two long lived lineages of specialised gliding mammals with no significant ecological competition in the form of bats or birds, and that are subjected to ecological pressures that favour powered flight, as well as almost certainly several long gone and unreccorded gliding mammal clades, it appears that gliding in mammals is not conductive to powered flight. It is possible that the necessary step for true volancy, the wing stroke, simply does not occur to animals “hardwired” to keep the wing stable during gliding, making the very speciation for gliding ironically the very reason why powered aerial locomotion is out of reach.

Of course, to strike out living gliders is not enough to determine whereas gliding is inconsistent with flight. To this end, examining true flyers is of extreme importance.

What Bats Tell Us

Baby bats appearently flutter when airborne, which has been at times interpreted as a possible indicator that the earliest bats weren't gliding.
Baby bats appearently flutter when airborne, which has been at times interpreted as a possible indicator that the earliest bats weren’t gliding.

The idea that bats are not descendents of gliders goes as far back in time as Glenn L. Jepson’s Bats: Origins and Evolution (1970), but it was more recently explored in detail in Kevin Padian et al. 2011. Here, the observance of fluttering in juvenile bats, as well as the very early evolution of echolocation in these animals, has provoked a hypothesis that, rather than descending from gliders, bats evolved from cave-dwelling mammals that specialised in stalking aerial insects, using echolocation to guise themselves in the darkness and fluttering in order to control their descent. While without doubt a rather fantastical idea, it did correlate to Weatherbee SD et al. 2006, in which the development of the bat patagium has been noted as rather atypical, developing first the specialised cheiropatagium and latter the rest of the patagia, inconsistent with the development in other gliding mammals but certainly on par with fluttering as means to develop the wing.

Later, Padian et al. 2013 dismissed the idea of cave-dwellers, but the emphasis on the hypothesis that early bats fluttered instead of gliding remained. In this scenario, creatures like Onychonycteris see themselves in a more plausible scenario: as arboreal mammals that weren’t that different from modern gliding mammals, but that flapped their arms to control their fall rather than simply letting themselves fall with style. Combined with the impetus to capture aerial prey, the drive to develop large, webbed hands occured, and subsequently it would only be a matter of time until true flight emerged.

This hypothesis, while so far only barely supported, certainly explains why bats achieved what no other mammal managed in a such a short amount of time. If it is a correct accessment, it defenitely parts mammalian flight from mammalian gliding, relying on the presence of arm movement since it’s very earliest stages, something that by itself is very rare among parachuting mammals. This also explains why bats managed to achieve profecient aerial hawking early on when birds and pterosaurs seemingly only developed it much later, having evolved from day on to snare aerial prey.

What Birds Tell Us

RPR illustrated (author unknown; I desire clarification so I can give the proper credit)
RPR illustrated (author unknown; I desire clarification so I can give the proper credit)

How avian flight developed has been one of the most debated topics in paleontology across it’s existence as a discipline. You’ve all heard about gliding, about running and beating the wings like a manic, WAIR, et cetera.

At this moment in time, the debate is still long from settled, but it seems that a few options can be safely struck down as unlikely. The running hypothesis, for example, has never witnessed much support, and nowadays it is rarely defended outside of conjunction with WAIR; the fact that most known early paravians had hindwings also renders this possibility unlikely.

With Dyke, G., de Kat, R., Palmer, C., van der Kindere, J., Naish, D. & Ganapathisubramani, B. 2013, gliding should also be out of the picture: the most famous supposed glider, Microraptor, was appearently a really lousy aeronaut, ill suited for long distance gliding. Therefore, as with bats, avian flight was not derived from gliding, and the theropods that rose into the skies did so in methods equally as active as their chiropteran successors, by flapping their wings.

Thankfully, there is a menagerie of ideas about how birds flapped their way into true volancy:

Wing Assisted Incline Running (W.A.I.R.), in which dinosaurs runned up trees by flapping their arms, like modern terrestrial birds;

The Leaping Grapple, in which dinosaurs leaped at their prey, using the forelimbs to generate power to elevate themselves.

– RPR, or “ripper”, in which dinosaurs flapped their forelimbs to retain balance when eating their prey, much like modern raptors.

Of these, WAIR has gone under the most scrutiny due to the assumption that non-ornithothorace paravians could not provide enough of a vertical stroke in order to produce WAIR. Ignoring several examples in which this it has been argued that early paravians could provide vertical strokes, Holtz argues that a vertical stroke isn’t necessary to generate WAIR (https://www.facebook.com/thomas.holtz/posts/173481572740497), so WAIR is within the realm of possibilities. However, the fact that WAIR is only observed in Neognathae, with no evidence among living palaeognaths, may cast suspicion into it’s veracity.

The other two suggestions are reliant on the assumption that early paravians were predatory, an assumption progressively considered less plausible given the extensive evidence of omnivory in paravians like troodontids, oviraptorans and microraptorids. Likewise, forms suspected of using these, like eudromaeosaurs, were presumably too large and specialised to be ancestral to birds. However, it is very likely that similar processes, unrelated to predation, were what developed dinosaur arms into avian wings.

Regardless of the method, it appears that exaptations for flight occured very early in theropod evolution, as the evidence of scutes in Concavenator implicates the presence of hindwings in non-coelurosaur theropods.

What Pterosaurs Tell Us

A series of Hypothetical Pterosaur Ancestors (HYPTAS) by Maija Karala (Eurwentala), in turn inspired by Mark Witton's own HYPTAS.
A series of Hypothetical Pterosaur Ancestors (HYPTAS) by Maija Karala (Eurwentala), in turn inspired by Mark Witton’s own HYPTAS.

Currently, pterosaurian origins are sorrounded in mystery. The most accepted hypothesis is that they’re sister taxa with Dinosauromorpha, and are especially closely related to the small archosaur known as Scleromochlus, though even this tells us very little about the hypothetical pterosaur ancestors, given the morphological diversity of early avemetatarsalians. If we go by animals like Scleromochlus, we find ourselves in an interesting position: while pterosaurs are quadrupedal animals with powerful forelimbs and plantigrade hindlimbs, Scleromochlus is a bipedal animal with long, digitigrade hindlimbs and small forelimbs. In the 90’s, this wouldn’t have been something hard to invision, but now that we know pterosaurs quadrupedally launched and were generally happy running on all fours, it seems very bizarre indeed.

Scleromochlus does not seem like a gliding animal, to say the least, and although basal pterosaurs suggest an arboreal ancestry, gliding is not necessarily on the equation. A suggestion posited by David Peters (I know, but this doesn’t involve squamates, so it’s okay) is that the key of the pterosaurian wing, the massive fourth finger, evolved as a display device, which would make sense in these original bipedal runners, with the forelimbs free to occupy other functions. With time, these display devices would have become larger and more elaborate, and eventually be useful in WAIR or fluttering. Eventually, once the wing was large enough, flight would have become possible, and subsequently the animals would have shifted to quadrupedality as the wing membranes connected the limbs. This hypothesis would also explain the elongated fifth toe, which would have also evolved as a display device, before becoming an effective support for the cruropatagium.

As of this writting, the fog sorrounding pterosaurian origins is still thick, and therefore one can only imagine how these animals first took to the sky.


Based on birds and bats, as well as modern gliders, gliding appears to be effectively independent from flight, gliding animals seemingly unable to produce powered flight and animals that have evolved powered flight having developed it in unusual circumstances. Exceptions might exist, including possibly pterosaurs, but as it stands vertebrate flight has only been able to evolve from behaviours that prommote forelimb movement, with aerodynamic speciation not leading at all to volancy.