Skip to content

Velociraptors on air: analysis of Deinonychosaur clades

October 12, 2011

Tianyuraptor by Nobu Tamura. Given it's short arms, to suggest that it was capable of gliding would still be rather unnecessarily generous, even despiste its small size.

Previously we saw how flight was defenitely present in deinonychosaurs, that juveniles of assumed flightless species were probably capable of flight and that there is proof to suggest the latter in at least two iconic dromaeosaurs. Now, we analyse how aerial locomotion was present across Deinonychosauria.


Xiaotingia, by Nobu Tamura. Much like Anchiornis, it is presumed to have been a mere glider, but it's wing feather arrangement might had been useful for flapping flight.

“Archaeopterygidae” is, in my opinion, nothing more than a wastebasket clade. While a sister-taxa relationship can be inferred for Archaeopteryx/Wellnhoferia and Anchiornis/Xiaotingia (perhaps it is apropriate to classify Wellnhoferia and Xiaotingia as species within Archaeopteryx and Anchiornis respectively?), the connection between the two groups into a four-genera clade seems rather contrived. They exhibit just about as much similarity in skeletical traits between themselves as they do with basal troodontids; in fact, historically the primitive troodontid Jinfengopteryx has been considered an “archaeopterygid”, and the “archaeopterygid” Anchiornis has been considered a basal troodontid. If anything, it is more likely that these animals are naught but stem-troodontids, with either Anchiornis/Xiaotingia or Archaeopteryx/Wellnhoferia being closer to Troodontidae, although they are sufficiently basal to suggest that they might predate the dromaeosaur/troodontid split, with either genus/sister-genera being more basal than the other.

Previously, we’ve established that Archaeopteryx was clearly volant; asymmetrical feathers aside, it’s insular distribution and known specimens from deep water areas suggest that it was capable of powered flight; the fact that it has quill knobs also suggests that it could fly, as flightless birds lack quill knobs. We did not debate Wellnhoferia‘s flight capacities because AFAIK we lack well preserved wing feathers, but given it’s similarity (and possible synonymity) to Archaeopteryx, it is very likely that it was perfectly volant as well. In fact, given it’s shorter tail, it was probably a better flyer. It is also very likely that not only it belonged to the genus Archaeopteryx, but it was probably synonymous with the type species, A. lithographica, it’s specimen representing the full grown adult of Archaeopteryx  rather than a different species (as previously discussed, deinonychosaurs were superprecocious and grew slowly, and considerable anatomical changes did occur in forms like Deinonychus).

Size comparations between different Archaeopteryx specimens (Wellnhoferia is the largest shown, being coloured orange); the smallest specimen was previously considered an idependent species, Jurapteryx, before it was classified as a juvenile of Archaeopteryx. Thus, it is perfectly reasonable to suggest that Wellnhoferia was also part of the same species.

While the urvogel is clearly volant, the same is less clear in Anchiornis/Xiaotingia. Unlike in Archaeopteryx, the wing feathers are symmetrical, a feature of flightless birds, not of volant ones (this is in part why I think Anchiornis/Xiaotingia was more basal than the urvogel). However, as we previously discussed, the unique feather arrangement (well evident in the Xiaotingia picture above) might suggest that they were capable of limited powered flight or WAIR. So far, only the evidence of quill knobs will make it clear whereas the wings were used for flapping or not.

Due to their generally poor adaptations for climbing (the sickle claw certainly indicates that they would have climbed trees, but probably just as a means to escape terrestrial predators), they presumably spent most of their time foraging in the ground. Since it is known that troodontids were omnivorous, it is possible that these animals behaved like modern small terrestrial birds like partridges, tinamous or buttonquails, relying on flight or WAIR to escape predators. Archaeopteryx‘s natural habitat (an arid archipelago) suggests that it might had been analogous to modern insular rails in habits (and, given how many specimens were found in deep water areas, it probably was capable of long distance flights, hence it colonised the islands by air); given that it lived on the shores of lagoons with high levels of salinity, it might had predated on brine flies in the same way as modern Laysan ducks.

The first accurate picture of Archaeopteryx to date. The "ground fowl" vibe is quite evident.

 Archaeopteryx is unique in that it’s hindwings were not as well developed as in other unambiguously volant deinonychosaurs (sans perhaps for Rahonavis, although it’s integrument is currently not preserved); considering that it is considered a less efficient flyer than volant dromaeosaurs, this might be an indicator that it was evolving towards terrestriality, maybe even actual flightlessness, which would fit it’s insular habitat. It’s slow growth rate was clearly a usefull exadaptation to it’s environment, as it came alongside a metabolism slower than that of modern birds, something usefull in an area with little nourishment. The fact that Archaeopteryx remained volant even despiste adquired terrestriality is both an indicator that flight was usefull in times when resources were minimal and that complete flightlessness in deinonychosaurs evolved slowly, befitting of animals that developed flight since birth like megapodes and pterosaurs.

The forest dwelling Anchiornis and Xiaotingia from Asia, on the other hand, exhibited well developed hindwings. Whereas they were capable of powered flight or not, they clearly had an aerodynamic function in these animals; they made running quite difficult, and due to the previously mentioned lack of adaptations for climbing an arboreal life style was extremely unlikely. Most likely they foraged in the ground, and when threatened either took off or climbed to the highest availiable spot and glided away (although unlikely to have lived on trees, quick climb would obviously still had been within their range of capacities, mainly by using the sickle claw; if using WAIR like their modern Galliforme equivalents, this would have been easier). Regardless of the method to take off, gliding would have been the main form of locomotion while on air, with the extra wing surface provided by the hindwings allowing the animal to stay on air for longer periods of time with minimal efford; with wing flaps, they could be airborne very long long periods of time.

“Archaeopterygids” co-existed with several pterosaur taxa; Archaeopteryx lived side by side with the nocturnal aerial insectivore Anurognathus, the nocturnal filter feeding Ctenochasma, the diurnal opportunist carnivore Pterodactylus, the nocturnal piscivore Rhamphorhynchus, the diurnal carnivore Scaphognathus and the possibly cathemeral piscivore Cynorhamphus and the possibly omnivorous/opportunist carnivore Germanodactylus, while Anchiornis and Xiaotingia co-existed with the carnivorous Archaeoistiodactylus, Fenghuangopterus, Jianchangopterus and the fabled wukongopterids. Predictably, a large number of these pterosaurs probably predated on the deinonychosaurs; competiton betwen pterosaur juveniles and the “archaeopterygids” was unlikely as they occupied different ecological niches. While pterosaurs had a similar number of diurnal and nocturnal forms, “archaeopterygids” were exclusively diurnal, further decreasing competition.

Other basal, but poorly understood forms

Pedopenna. Basal avialan or deinonychosaur, it is only known from the hindlimbs.

Several basal deinonychosaurs have been described from the Jurassic and Cretaceous periods, some even dating as recently as the Maastrichtian. Because of the fragmentary or incomplete nature of these fossils, it is hard to infer how they looked like, let alone how they lived. Due to their small size, it is possible to infer that they were capable of gliding or flying, although without integrument and well preserved forelimbs it is not honest to claim that this was proven.

One of these animals, however, is worth discussing: Pedopenna. It’s exact placement in the paravian evolutionary tree is unknown, and it has been suggested to have been either a basal deinonychosaur, a basal avialan or even a dinosaur that diverged before Deinonychosauria and Avialae. Regardless of it’s affinities, Pedopenna had very well preserved hindlimbs (the only parts of it’s body that became fossialised, in fact), which show clear hindwings. However, unlike other dinosaurs with hindwings, the wing feathers of Pedopenna were not just asymmetrical and short, but considerably poorly developed and without barbicels, like the wings of ostriches and rheas. Thus, much like in Archaeopteryx, they had very little in the way of aerodynamic function. This can either be interpreted in two ways; either Pedopenna represented a truly basal form with primitive flight feathers, or represented a more derived species which began loosing it’s hindwings, either by specialised in using exclusively the forelimbs, or by becoming flightless.


Byronosaurus by "FunkMonk". While nest parasitism is unlikely, this animal clearly displayed precociality, if not outright superprecociality.

As previously mentioned, at least basal troodontids are very similar to “archaeopterygids”, so it is perhaps not surprising that Jinfengopteryx has been originally mistaken for a close relative of the urvogel, for instance, or that Anchiornis had been considered a troodontid; it is more likely that “archaeopterygids” were naught but stem troodontids. Being generally omnivorous, troodontids can therefore be considered “Mesozoic ratites”, since they were the flightless, cursorial descendents of partridge like dinosaurs. It is easier to make an argument that troodontids were flightless more easier than in any other deinonychosaur clade; their arms were generally shorter than those of other deinonychosaurs, and no quill knobs have so far been reported – however, this is probably because few well preserved troodontid forelimbs have been found. Regardless, it is possible troodontids might not had been completly flightless.

Byronosaurus and Troodon both offer several examples of troodontid embryos or newborns. In all known cases, the young show evidence of superprecociality, thus the possiblity that young troodontids might had been able to glide or even fly has some logical evidence. Troodontid forelimbs seem poorly adapted to grasp, so an exclusive use as wings was likely, either for aerial locomotion or display. Troodontids in general were relatively small animals, thus the possibility that they glided and/or used WAIR as juveniles or even as adults (in the case of the smallest species) can be inferred to some extent. Such a case would be the already mentioned Jinfengopteryx, which had strong sickle claws, apropriate for climbing, but had forelimbs poorly equiped for that, so it most likely used WAIR. Jinfengopteryx is also unique in that it lacks hindwings, indicating that troodontids were more likely to have been limited to WAIR and parachuting in the smallest species and the chicks of the larger ones. However, since it is a single specimen, it can probably be inferred that it is equally as likely that it lost the hindwings post-mortem. If troodontids preserved their hindwings, then most of the availiable wing surface would have come from them, considering the length of the legs, a situation similar to that of microraptorines (see below).

At least derived troodontids are inferred to have been generally nocturnal or crepuscular; they generally exhibit proportionally large eyes, and some species have ear asymmetry (do not, however, that, unlike in owls, said asymmetry is superficial, so whereas they amplied hearing is not known). This is a strong contrast to the exclusively diurnal “archaeopterygids”, and unlike in other cathemeral/nocturnal deinonychosaurs (see below), this was probably due to competition with other omnivorous theropods, including birds, which were generally more diurnal, instead of deliberate specialisation. Otherwise, these animals seem to have behaved like modern terrestrial omnivorous birds, spending most time on the ground but using WAIR if running didn’t work. Some specialisation to stalk prey like owls might had occured, specially in Byronosaurus; in said species, there could be some higher degree of aerial locomotion, like being able to actively glide and fly, than in their omnivorous relatives. Such a lifestyle would had been similar to that of microraptorines (see below).


Rahonavis and Microraptor are generally considered to have been better flyers than Archaeopteryx. Rahonavis in particular seems to have been a dedicated soarer, given its unusually long arms, even by volant dromaeosaur standards (in this illustration by Nobu Tamura, said arms are folded, but you can see hints of their length; look at the metacarpals, for instance, which are almost pterosaur-like).

Dromaeosaurs are well known for having had volant species; both Microraptor and Rahonavis are not only known to have been capable of flying, but being better flyers than Archaeopteryx, indicating that these animals became more specialised for flight. Furthermore, aerial locomotion seems to have been more widespread across Dromaeosauridae than previously thought, having possibly occured even in eudromaeosaurines. Amusingly enough, the earlier forms are the ones most likely to have been flightless.

Flightless basal taxa

Although co-existing with the much derived Velociraptor, Mahakala was less likely than it to have flown, despiste being more basal. Illustration by Nobu Tamura.

The dromaeosaurs most likely to have been flightless are several basal taxa, rather ironic considering that dromaeosaurs are generally considered to have been the flightless descendents of basal flying taxa. Of these ancient flightless animals, Mahakala and Tianyuraptor stand out the most. Both animals have extremely short arms, making gliding very unlikely at best. WAIR might had been a possibility, specially given their powerful sickle claws, but even that would perhaps be too generous. The possibility that juveniles could fly or glide is still availiable.

These animals, as well as microraptorines are often noted as troodontid like in several aspects. Presumably, both Tianyuraptor and Mahakala led similar lifestyles to those of the more terrestrial troodontids.


Sinornithosaurus was closely related to Microraptor, and might had been similar in many aspects.

Microraptorinae is iconic for it’s type genus, which is one of the two unambiguously volant dromaeosaurs. These animals are notable for their hindwings, the most well developed among known species that had them. In Microraptor itself, they might even be broader than the forewings, which were triangular in shape like those of modern fast flying birds. Though clearly capable of powered flight, these animals were likely well adapted to gliding and soaring, making use of their wings to stay airborne as long as possible without loosing energing by flapped flight. Indeed, in Microraptor the shape of the wings was so efficient that it appearently capable of gaining altitude by correctly adjusting the angle of the hindlimbs, making U shaped loops.

Larger species like Sinornithosaurus have been claimed as being flightless due to their larger size. However, size is actually a factor in the microraptorine style of flight, since gaining altitude in U shaped loops would have been easier with more weight. Thus, size alone not only would not render Sinornithosaurus as flightless, but would had made taking advantage of the hindwings easier. The equally big Hesperonychus also shows an important adaptation for this style of flight; it’s hip sockets are directed partially upwards, an unique adaptation among dinosaurs. While this could had been vestigial, it would have made terrestrial locomotion harder, to the point of jeopardising a terrestrial dinosaur’s survival. Thus, it is likely that Hesperonychus too was capable of at least gliding. Furthermore, quill knobs are present.

Based on scleral ring studies, Sinornithosaurus was cathemeral, while Microraptor was nocturnal, indicating that, like troodontids, microraptorines became specialised to forage at night (indeed, microraptorines are among the dromaeosaurs most often compared to troodontids, and the smaller and more carnivorous troodontids like Byronosaurus probably led similar lifestyles). This, and the findings of mammal remains in the digestive system of Microraptor, seems to suggest that these animals lived like modern owls and frogmouths, predating on small tetrapods at hours of darkness. Their specialisation to gliding would have been useful to stalk their prey silently, without requiring to develop specialisations like the specialised feathers in owl wings and the specialised pycnofibrils in anurognathid wings. Said adaptations most likely did evolve in later species like Hesperonychus, however.


Buitreraptor is often infered to have been the closest known relative to the flying Rahonavis. It seems less suited for flight given it's proportionally thinner arms, but has interesting adaptations that cannot be ignored.

Unenlagiinae is the other dromaeosaur clade to have a known volant genus, Rahonavis. Historically, it has been suggested that it’s arm actually belonged to a different animal, the contemporary bird Vorona; however, this seems unlikely because of the nature of it’s position in relation to the animal’s body, and also because the arm shows several features that Vorona lacks. Most notably, the difference difference in thickness in the radius and ulna is less extreme than in Vorona, and the arm is also proportionally longer. Thus, Rahonavis was a fully volant deinonychosaur, and the length of the arms as well as the presence of quill knobs indicates that it was probably adapted to soaring flight, a logical conclusion to the gliding capacities so prevalent in Deinonychosauria.

Unenlagiines in general show adaptations for powered flight. The shoulder girdle was even more apropriate for flapping than in other dromaeosaurs, and with the exception of Austroraptor these animals were generally very lightly built. However, most species have forelimbs that are too short for powered flight, meaning that, like troodonts, these animals were generally more likely to have engaged in WAIR and parachuting than in gliding/flying. But again, juvenile animals were probably Rahonavis like animals that engaged in powered flight, as it is likely that the volant unenlagiine was just the juvenile of a larger species.

The closest relative of Rahonavis was Buitreraptor, an animal that is considered to have been less volant. However, it has a few adaptations that are of interest in it’s forelimbs. There are well developed crests in the base of the humerus, where the deltoideus complex would had been attached in life, about as well developed as those present in Microraptor, suggesting that the wing muscles would have been relatively well developed. The manus is also unusual, in that the second and third fingers were roughly of the same length; it is possible that they were probably beginning to be united in order to provide a more solid base for wing feathers, just like in birds. Whereas these are remains of volant ancestors or juveniles (probably unlikely considering that they don’t seem to be vestigial; in dromaeosaurs like Deinonychus the forelimb morphology changes considerably with age to suit the animal’s change in lifestyle) or signs of behaviours the adult animals engaged on is unknown.

Buitreraptor by Nobu Tamura, with wing feathers probably long enough for it to parachute. Although the pink colour is the result of a campaign with a pink dinosaur motiff, modern birds like scarlet ibises can obtain red pigments from their diet, and Buitreraptor was likely also a predator of aquatic invertebrates.

Unenlagiines occupied a similar ecological niche to that of azhdarchid pterosaurs. It is possible that direct competiton wasn’t present, however, as azhdarchids favoured drier habitats while unenlagiines have known piscivorous species like Austroraptor. If directly competing, it is unlikely unenlagiines would have had the success they had, as stork-like niches favour species more capable of flying (as they can escape predators easily and find new food sources in times of hardship), as evidenced by neornithe evolution; regardless of whereas unenlagiines were volant or not, azhdarchids were still unambiguously the better flyers, as pterosaurs had more efficient wings, more developed musculature and the quadrupedal launch method that would have preserved far more energy that the bipedal dinosaurian take-off. It is also possible that unenlagiines were nocturnal, further avoiding competition with the clearly diurnal azhdarchids.


Linheraptor, a member of Velociraptorinae, the clade within Eudromaeosauria that provides the most evidence that at least their juveniles could fly

 We’ve already discussed the evidence for flight in the juveniles of two iconic eudromaeosaurines, so it is already made clear that the possibility of flight in juvenile eudromaeosaurs has some basis on reality.

Eudromaeosaurids are the most derived dromaeosaurs, and the fact the juveniles were capable of flight or at least WAIR indicates that their adaptations to kill large prey are exadaptations from their juvenile aerial phase. The strong forelimbs, presumably originated as requirements for flapping strokes, became usefull to grasp prey, while the sickle claw used to aid in climbing became refined as stabbing weapons. Microraptorines showed some evidence of doing the same, but they never went to the extreme eudromaeosaurs shown, presumably because microraptorines still retained flight as adults.

Most evidence for aerial tendencies on juvenile eudromaeosaurs comes from velociraptorines (notably, unlike other eudromaeosaurs, they managed to colonise the european islands and produce the bizarre Balaur, which would probably had been truly a “dodo raptor”, both in terms of behaviour as well as the possibility of being the descendent of animals that reached the islands while still flying juveniles), but saurornitholestines probably also engaged in aerial locomotion. Bambiraptor is perhaps the most obvious example, as the well described juvenile is on par in terms of size with the average microraptorine, and well full grown it would probably still had been the apropriate size for gliding and possibly also flapped flight; the same applies to Atrociraptor. The bigger Saurornitholestes presumably would have reached the limit once near adult size, but would probably still have been capable of at least parachuting at subaldult size. The only clade of eudromaeosaurs that was presumably fully terrestrial would had been Dromaeosaurinae, which included the largest known eudromaeosaurs, and even then it is perhaps possible to infer that juveniles where arboreal and thus possibly able to parachute or glide, as it happened with the velociraptorine Deinonychus.

Eudromaeosaurs, like microraptorines, seem to have been nocturnal or cathemeral animals, according to scleral ring studies. The superficial similarity between both clades can maybe be use to infer than both groups occupied similar ecological niches, but while microraptorines became more specialised for flight, eudromaeosaurs only exhibited flight in the smaller species and the juveniles of the larger ones, with the general tendency of adult animals sacrificing aerial locomotion for strength to kill prey.


Unlike most neornithes, deinonychosaurs were clearly capable of compromising between flight and flightlessness by exploiting different niches at different stages of their life, and even then several taxa were still capable of aerial locomotion across their lifetime. A large number of species within Troodontidae, Microraptorinae and Eudromaeosauria show clear nocturnal tendencies, suggesting that these animals might have been specialised to hunt forage like modern nocturnal birds. This might also had evolved as means to avoid competiton with true birds, which are generally diurnal. Indeed, it can perhaps be infered that birds expanded more easily into these niches once deinonychosaurs became extinct. Their presence may also explain why predatory birds were not very common in the Mesozoic; microraptorines and juvenile eudromaeosaurs might also had exploited the disappearence of predatory non-pterodactyloid pterosaurs like scaphognathines, which occupied similar niches. However, said pterosaurs were generally diurnal, not nocturnal.

4 Comments leave one →
  1. BananaBoyL337 permalink
    December 21, 2011 5:20 pm

    Hey JhonFaa (If you are not JhonFaa I apologize for my mistake) is it possible that flight in Dromeasaurs, Troodonts and birds evolved independently from a flightless ancestor?

    • December 21, 2011 5:25 pm

      Absolutely, and possibly already a proven fact, given that microraptorines are clearly well adapted to aerial locomotion while less derived deinonychosaurs like Anchiornis are possibly mere gliders (if my hypothesis that they were facultatively volant is wrong).

  2. ajukar permalink
    April 29, 2013 6:45 am

    Hi gwawinapterus,

    I was wondering if you could provide citations for your claim that troodontids and dromaeosaurs were capable of powered flight. As far as my research goes, no one has really explored the possibility of a deltoid driven upstroke. The arguments against troodontid and dromaeosaur powered flight are that the orientation of the gelno-humeral joint, which in pretty much every group other than pygostylia is ventrally or medially oriented. We only start seeing more dorsal orientations of the joint in confuciusornithidae, ornithothoraces and ornithurae. The only real documentation of the flight capabilities of microraptor and rahonavis are in the book by Agnolín and Novas (2013). Also, what is your opinion of the eosinopteryx paper that came out earlier this year. Also, just out of curiosity, are you a professional paleontologist/paleo artist or an amateur/student?

    • April 29, 2013 11:58 am

      Deltoideus based wing strokes are what pterosaurs and bats do, for starters; not only has Microraptor always been noted for it’s well preserved deltoideus complexes, but Chatterjee even compared them to a bat’s. As for the gelno-humeral joint:

      – As Agnólin and Novas pointed out, a more dorsal condition is present in “deinonychosaurs” in general.

      – Forelimb studies on alligators show that archosaur forelimb flexibility cannot be judged by joints alone

      Eosinopteryx may or may not suffer from preservation issues, but at any rate it seems very “Galliforme-ish” to me. Finally, I’m a person with both hobby-oriented and moral duties to be interested in palaeobiology, but I am not interested in a professional career in the area.

Leave a Reply

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

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: