Skip to main content

The Romanian Dinosaur Balaur Seems to Be a Flightless Bird

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American


A Balaur bondoc pair in their forested Romanian home, 67 or so million years ago. Artwork by Emily Willoughby. Read on for more about this image.

Paravian theropods are all the rage right now, and not just because of the phylotarded, retrofitted ones that appear in a certain blockbuster movie. This week sees the publication of a lengthy paravian-themed paper by long-time collaborator Andrea Cau, former student Tom Brougham and myself on the enigmatic and wonderful Romanian paravian Balaur bondoc, originally described in 2010 by Zoltán Csiki-Sava and colleagues and discovered by my friend and colleague Mátyás Vremir in 2009 (Csiki et al. 2010).

Balaur got a rightful amount of press coverage following its discovery, with much of the attention being focused on the fact that it was regarded as an especially close relative of the east Asian, desert-dwelling dromaeosaurid Velociraptor (Csiki et al. 2010). As discussed below, Balaur is anatomically weird when interpreted as a dromaeosaurid, and the hypothesis favoured by Csiki et al. (2010) and then by Brusatte et al. (2013) in their monograph on the animal was that this weirdness is down to its isolated evolution on an island.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


Diagram accompanying the original work on Balaur by Csiki et al. (2010), focusing on many of its more unusual skeletal features.

This dromaeosaurid interpretation of Balaur looks reasonable on the basis of the animal’s size and foot anatomy, and Balaur is certainly dromaeosaurid-like in a general sense. But it’s also similar to members of another, closely related paravian lineage: Avialae, the bird lineage. And that’s the primary contention of our new study. Based on phylogenetic analyses and critique of the various unusual features of this theropod, we argue that Balaur is likely not a dromaeosaurid, but a secondarily flightless bird. If you’re at all aware of the discussion that’s surrounded the possible evolution of flightlessness in non-bird paravians (Paul 1988, 2002), the significance of this won’t be lost on you. Our paper (Cau et al. 2015) is published in the open access journal PeerJ, so is available for free to everyone.

The Balaur bondoc holotype at Transylvanian Museum Society, Department of Natural Sciences, Cluj-Napoca, Romania. The articulated right hindlimb is close to the middle of the image; the long, curved elements on the right are the pubic bones. Photo by Darren Naish.

Interpreted as a dromaeosaurid, Balaur is weird.Balaur is known from a single decent but headless partial skeleton, discovered in the latest Cretaceous red riverbed sediment of Sebeş-Glod in Alba County, Romania. A second, far more fragmentary specimen is known from a different locality in Hunedoara County (Csiki et al. 2010). Balaur was about similar in size to a goose and seems to have been similar in overall form and proportions to mid-sized dromaeosaurids, like Velociraptor. I've spent time with the specimen on several occasions and have also visited the discovery location in an effort to find more of the animal. Incidentally, this is the same place that yielded the small azhdarchid Eurazhdarcho, described by myself and colleagues in 2013 (Vremir et al. 2013).

The Balaur discovery site at Sebeş-Glod, as it looked in June 2011. Fossils are found both on the submerged bed of the river and on the adjacent banks. Photo by Darren Naish.

Balaur has always seemed super-weird when compared to other dromaeosaurids. It has a reduced third finger (lacking a claw), extensive fusion between the hand and wrist bones, a strangely broad, extensively fused pelvis with pubic bones that bow outwards for most of their length and are strongly swept back, fusion between the tibia and fibula, fusion between the tibia and ankle bones, an especially stocky, heavily built, partially fused-up foot, and a long hallux (first toe) with an especially big claw. As mentioned above, Steve Brusatte, Zoltán Csiki-Sava and colleagues ascribed these features to island endemism (Csiki et al. 2010, Brusatte et al. 2013). That is, they proposed that these were weird features that evolved as the dromaeosaurid lineage concerned evolved in isolation in an island environment. As is well known, islands have a tendency to produce weird members of otherwise not-especially-weird lineages.

The pelvis of Balaur (shown in A and B) is one of several parts of its skeleton that are quite bird-like. (A) shows the Balaur pelvis in left lateral view. (B) shows the pelvis in anterior view: compare those bowed-out pubic bones to the condition in (C) the bird Sapeornis and (D) the dromaeosaurid Velociraptor. Diagram from Cau et al. (2015).

Interpreted as a bird, Balaur is not quite as weird. However, the weird features of Balaur actually aren't weird if we look around in the paravian family tree. All of the features I just mentioned - yes, all of them - are present in Avialae, the lineage that includes modern birds and all of their especially close relatives (as theropod aficionados will know, Archaeopteryx and a few other very bird-like paravians are found to be members of Avialae in some studies, but to belong elsewhere within the clade in others). Looking at Balaur with Mesozoic bird specialist Gareth Dyke, we often noted how certain parts of Balaur’s skeleton seemed so reminiscent of early birds. And this was also noted (independently) by Andrea Cau, who’s been saying ever since 2010 that Balaur might actually be an archaic flightless bird, not a dromaeosaurid at all. Aided and abetted by Gareth, a team formed of myself, Tom Brougham and Andrea decided to collaborate in a test of Balaur’s phylogenetic position.

Substantially simplified cladogram depicting approximate relationships of Mesozoic birds. Most of the lineages shown here came and went during the Cretaceous, only crown-birds making it through the K-Pg event. Diagram from Naish (2012).

It should be said that the concept of an avialan Balaur had already entered the literature prior to our new study. Godefroit et al. (2013), in their description of the Jurassic avialan Aurornis, published a phylogeny where Balaur is an avialan, closer to Pygostylia (the short-tailed bird clade) than is Archaeopteryx. And Foth et al. (2014), in their study of a new Archaeopteryx specimen, also found Balaur to be a member of Avialae, again closer to crown-birds than Archaeopteryx.

Part of one of our cladograms (from Cau et al. 2015).

We found Balaur to belong to the same approximate region of the cladogram as Jeholornis, Sapeornis and other birds close to – but outside of - Pygostylia (Cau et al. 2015). This position means that Balaur has to be interpreted as secondarily flightless, its stocky feet, forelimb proportions and body size being derived from smaller, flight-capable ancestors. Several secondarily flightless birds are already known from the Mesozoic (for review see Naish 2012), but Balaur is the one furthest from crown-birds.

It’s a bird, but that’s not actually a big deal. We incorporated Balaur into an updated version of Andrea’s giant, evolving theropod matrix – the one we previously used in our Lee et al. (2014) study of body size over time – and also into the latest version of the TWiG matrix (TWiG = Theropod Working Group, a team of researchers mostly affiliated with the AMNH). As is explained at length in the paper, Balaur is better supported as a avialan in all of these analyses, and a placement within Dromaeosauridae was statistically weaker and thus less likely to be correct (Cau et al. 2015).

Here’s an important thing which needs saying. When you reinterpret a given dinosaur species as a member of the bird lineage or, conversely, when you remove a given dinosaur species from the bird lineage, people think of it as a big deal. “OH”, they say, “you’re not saying it’s a dinosaur... you’re saying it’s a bird now?”, they say. The implication is that movement of the given animal from the (say) dromaeosaurid branch to the bird branch is a big deal, that it’s a fundamental ‘reinterpretation’ or ‘correction’. Here’s what you should remember: it’s not. We’re talking about moving an animal something like two, three or four nodes on the cladogram – not a big deal at all. Other animals get moved much further than this on cladograms all the time and nobody cares.

So, Balaur has been moved around a bit in the cladogram. What I want you to note from this substantially simplified depiction of the competing phylogenetic positions is that we're talking about a very minor change. Other animals are routinely moved much greater distances across cladograms. Images by Darren Naish except for Balaur, which is by Emily Willoughby.

A point we also make in the paper is that an avialan identification of Balaur doesn’t make much difference as goes how we imagine the look of this animal. After all, abundant fossil evidence has demonstrated that dromaeosaurids and the members of other paravian (and maniraptoran) lineages were extremely similar in life appearance to archaic birds. As I like saying, were you to go back in time and look at live, early members of Dromaeosauridae, Troodontidae, Avialae and other maniraptoran lineages you almost certainly wouldn’t be able to ‘spot the bird’; in other words, to recognise members of the one lineage that was destined to survive the K-Pg event and give rise to a dynasty of thousands and thousands of species.

Speculative life reconstruction of Balaur by Emily Willoughby, taking account of data known about the life appearance of the maniraptorans that surround it in the phylogeny.

A specimen of the stem-bird Jeholornis prima from the Jiufotang Formation of Liaoning Province contains over 50 ovules in its body cavity. Diagram from Naish (2014).

An omnivorous, Sapeornis-like flightless bird? Interpreted as a bird from the region of the cladogram described above, we have to conclude that Balaur was not the arch-predator previously imagined, but an omnivore or generalist. We know from stomach contents (Naish 2014), and we infer from skull and tooth shape, that birds like Sapeornis and Jeholornis were omnivores or herbivores that ate seeds and fruits. One Jeholornis specimen has over 50 fruits preserved in its stomach contents – a strong indication that the feeding event which this represents was very deliberate. We should now assume that Balaur was similar in terms of dietary preferences and behaviour.

An omnivorous or herbivorous lifestyle is also in line with the widened, backswept pubic bones of Balaur: these imply larger guts and a broader, heavier body not especially suited for speedy, predatory behaviour. The proportionally short, stocky foot is also line with a non-predatory lifestyle. But – waitaminute – doesn't Balaur have two sickle-claws, showing that it was an even more badass predator than dromaeosaurid predators like Velociraptor and Deinonychus? Well, not so fast. As we explain in the paper (Cau et al. 2015), and as Brusatte et al. (2013) noted previously, those curved claws on the first and second toes are actually not especially curved: the second toe claw of Balaur is less curved than is typical for dromaeosaurids, the claw looks less suited for use in predation, and second toe claws of similar shape and curvature are present in a whole list of Mesozoic birds (including Bohaiornis, Fortunguavis, Jixiangornis, Parabohaiornis, Patagopteryx, Qiliania, Sulcavis and Zhouornis).

Those 'double sickle claws' might not have been 'sickle claws' at all, but structures used in climbing and perching, as is typical for birds. Based on Sapeornis and other birds from the same region of the cladogram, we should expect an extensively feathered foot like that shown here. Illustration by Emily Willoughby, based on a concept diagram by Andrea Cau.

Furthermore, the hallux in Balaur looks less like a ‘weaponised’ digit and more like a hallux of the sort typical for avialans (Cau et al. 2015). The articular joints between the bones show that Balaur’s hallux was highly mobile (as is typical for birds, but not for dromaeosaurids), and the length of the hallux relative to the other toes is also typical for birds but very unlike the condition present in dromaeosaurids (Cau et al. 2015). These enlarged foot claws and mobile hallux suggest that Balaur might have have been good at climbing and perching. You can see these ideas about omnivory and a possible climbing ability were incorporated into the fantastic life reconstructions produced by Emily Willoughby, who worked closely with us in completing the pieces shown here. A rich diversity of plant fossils are known from the sediments of latest Cretaceous Romania, various of which were incorporated by Emily into her reconstruction.

Bonus image of fossil flora discovered in the same sedimentary unit as Balaur (and the azhdarchid Eurazhdarcho): a partial trunk of a large tree, my boot for scale. This particular fossil trunk was found about 10 m from the spot where Balaur was discovered. Photo by Darren Naish.

Incidentally, the manual unguals of Balaur are not as strongly curved as those of most dromaeosaurids either, and the flexor tubercles (the bony lumps on the undersides of the unguals that anchor ligaments used in ungual flexion) are comparatively weakly developed. Combine this with the reduced, claw-less third digit and fused nature of the fingers and wrist, and we have clear indications that Balaur was in possession of a non-raptorial hand (Cau et al. 2015).

Speculative skeletal reconstruction for Balaur bondoc, showing known elements in white and unknown elements in grey. Note that the integument would presumably have substantially altered the outline of the animal in life. Produced by Jaime Headden, used with permission.

Phylogenetic hypotheses are... hypotheses – they posit an evolutionary position for an organism on the basis of an interpretation of evidence, and they’re liable to be modified or overturned as new data comes in. We find Balaur to be better supported as a member of Avialae than as a member of Dromaeosauridae, and our ideas about its appearance, ecology and lifestyle are based on this hypothesis. But, as ever, this mustn’t be regarded as a ‘last word’ on the affinities of this marvellous and remarkable dinosaur. One thing that would go a long way towards helping out here would be the discovery of a Balaur skull – hey, we’re looking!

There’s more to say about our reinterpretation of Balaur. If a dromaeosaurid-like maniraptoran has turned out to be a flightless bird, does this have implications for any of the other flightless paravians of the Jurassic and Cretaceous? And why does Balaur share that weird, wrinkled bone texture with a selection of other Romanian maniraptorans, including Bradycneme and Elopteryx (Naish & Dyke 2004, Csiki et al. 2010)? There is still much to find out.

For previous Tet Zoo articles on Mesozoic paravians and other maniraptorans, see...

Refs - -

Brusatte, S. L., Vremir, M., Csiki-Sava, Z., Turner, A. H., Watanabe, A, Erickson, G. M. & Norell, M. A. 2013. The osteology of Balaur bondoc, an island-dwelling dromaeosaurid (Dinosauria: Theropoda) from the Late Cretaceous of Romania. Bulletin of the American Museum of Natural History 374, 1-100.

Cau, A., Brougham, T. & Naish, D. 2015. The phylogenetic affinities of the bizarre Late Cretaceous Romanian theropod Balaur bondoc (Dinosauria, Maniraptora): dromaeosaurid or flightless bird? PeerJ 3:e1032.

Csiki, Z., Vremir, M., Brusatte, S. L., Norell, M. A. 2010. An aberrant island-dwelling theropod dinosaur from the Late Cretaceous of Romania. Proceedings of the National Academy of Sciences of the United States of America 107, 15357-15361.

Foth, C., Tischlinger, H. & Rauhut, O. W. M. 2014. New specimen of Archaeopteryx provides insights into the evolution of pennaceous feathers. Nature 511, 79-82.

Godefroit, P., Cau, A., Dong-Yu, H., Escuillié. F., Wenhao. W. & Dyke, G. 2013. A Jurassic avialan dinosaur from China resolves the early phylogenetic history of birds. Nature 498, 359-362.

Lee, M. S. Y., Cau, A., Naish, D. & Dyke, G. J. 2014. Sustained miniaturization and anatomical innovation in the dinosaurian ancestors of birds. Science 345, 562-565.

Naish, D. 2012. Birds. In Brett-Surman, M. K., Holtz, T. R. & Farlow, J. O. (eds) The Complete Dinosaur (Second Edition). Indiana University Press (Bloomington & Indianapolis), pp. 379-423.

- . 2014. The fossil record of bird behaviour. Journal of Zoology 292, 268-280.

- . & Dyke, G. J. 2004. Heptasteornis was no ornithomimid, troodontid, dromaeosaurid or owl: the first alvarezsaurid (Dinosauria: Theropoda) from Europe. Neues Jahrbuch für Geologie und Paläontologie, Monatshefte 2004, 385-401.

Paul, G. S. 1988. Predatory Dinosaurs of the World. Simon & Schuster, New York.

- . 2002. Dinosaurs of the Air: the Evolution and Loss of Flight in Dinosaurs and Birds. Johns Hopkins University Press, Baltimore.

Vremir, M., Kellner, A. W. A., Naish. D. & Dyke, G. J. 2013. A new azhdarchid pterosaur from the Late Cretaceous of the Transylvanian Basin, Romania: implications for azhdarchid diversity and distribution. PLoS ONE 8(1): e54268.

Darren Naish is a science writer, technical editor and palaeozoologist (affiliated with the University of Southampton, UK). He mostly works on Cretaceous dinosaurs and pterosaurs but has an avid interest in all things tetrapod. His publications can be downloaded at darrennaish.wordpress.com. He has been blogging at Tetrapod Zoology since 2006. Check out the Tet Zoo podcast at tetzoo.com!

More by Darren Naish