Big birds in the Cretaceous of Central Asia: say hello to Samrukia
By Darren Naish | August 9, 2011 | blogs.scientificamerican.com/tetrapod-zoology/2011/08/09/say-hello-to-samrukia/
Two possible body shapes for the new bird - read on. Images by John Conway.It’s not uncommon in palaeontology to discover isolated, even fragmentary, specimens that seem not only to represent new species, but also to tell you a lot of interesting stuff. Today sees the publication of a new paper in Biology Letters in which I and a team of colleagues describe a remarkable new Cretaceous bird, discovered a few decades ago in the Kyzylorda District of Kazakhstan (Naish et al. 2011). Represented only by the two halves (or rami) of its large lower jaw, this fossil provides new information on Cretaceous bird evolution and diversity, and perhaps on the composition of Cretaceous faunas and ecosystems.
The adventure started back in August 2010 when Pascal Godefroit of the Institut royal des Sciences naturelles de Belgique (IRSNB) in Brussels began corresponding with Gareth Dyke and myself about this most interesting specimen. Complete, in great shape and just over 30 cm long, it possessed a rounded, ‘U’-shaped symphyseal region (the section where the two rami meet anteriorly) and was completely toothless. Seen from above, it looked superficially like the lower jaw of a caenagnathid oviraptorosaur.
While the jaw clearly belonged to a theropod dinosaur, it was also evident right from the start that it was specifically from a bird, and actually not from an oviraptorosaur at all. The mandibular cotyle – the concave region of the lower jaw that articulates with the quadrate bone – was very obviously biconcave, with a diagonally aligned ridge separating the two parts. Mandibular fenestrae (window-like openings in the outer wall of each ramus) were totally absent, the bones were all fused together, and a large, apparently pneumatic opening was present on the posterior, post-articular region of each ramus. Gareth and I needed to check it out in person, so we made the trek to Brussels to examine the specimen first hand (and a trip to the IRSNB is great, especially if you’re interested in iguanodontians).
The jaw as it originally appeared. It's huge! (for a Mesozoic bird).Sure enough, we could confirm that it was indeed a bird jaw. A big one. Those biconcave cotyles, fully fused-up bones, pneumatic foramina and so on all screamed ‘bird’. One or two of these features might be present in members of other clades, but not all of them. Gareth codes the specimen for analysis. How did we do science before laptop computers?
So ideas that the jaw might be from any of the other Mesozoic tetrapod groups known to have evolved toothless rami (like turtles or pterosaurs) could be immediately eliminated. What about the general oviraptorosaur-like demeanour of the specimen? By now it was now obvious that the specimen was only superficially oviraptorosaur-like, and all because of that ‘U’-shaped, toothless symphyseal region. And this was a painted plaster reconstruction! Yes, it seems that someone (we don’t know who) deliberately modelled the symphyseal region after that of a caenagnathid oviraptorosaur, perhaps because they assumed that this is the sort of animal the jaw rami belonged to. In all of its other details, the jaw is emphatically unlike that of an oviraptorosaur, and it also differs substantially from the lower jaws of the other theropod clades that evolved toothlessness (ceratosaurs and ornithomimosaurs).
With the fake symphyseal region removed, we were left with two elongate rami. These are toothless, but – lacking the animal’s real symphyseal region – we can’t be sure that the jaws were completely toothless since at least a few Mesozoic birds (Jeholornis prima is the classic example) lack teeth in their rami, but still have a few teeth at the jaw tips.
The partial lower jaw rami of our new giant Cretaceous bird. (a) posterior region of left ramus as seen from above. (b) posterior region of left ramus as seen from below. (c) right side of lower jaw, seen from the outside. (d) right side of lower jaw, seen from the inside.
Placing the bird in the tree
By eyeballing the specimen and comparing its characters with those of other theropods (and other animals) we were already confident that this was a big Cretaceous bird. But what sort of bird? The large pneumatic foramen at the back of the jaw suggested that it was close to Ichthyornis and neornithines. Gareth and I coded the specimen and threw it into a large cladistic analysis of Mesozoic birds (O’Connoret al. 2011). It was resolved as a member of Ornithuromorpha, the avialian clade that includes crown-birds (Neornithes), Patagopteryx, Vorona,Ambiortus, Apsaravis and the Yixianornis+ Yanornis clade.
In order to be even more confident about this phylogenetic placement, we worked with Andrea Cau (of Theropoda) to include the specimen within a far more extensive analysis of all of Theropoda. The placement was exactly the same: the specimen not only grouped within Avialae (the bird branch of Theropoda), but closer to neornithines than to confuciusornithids or enantiornithines, and somewhere round about the ‘base’ of Ornithuromorpha (Naish et al. 2011).
Archaeopteryx not a “bird”?
As an interesting aside, one peculiarity of our tree is that Archaeopteryx was recovered as a basal paravian, in a polytomy with scansoriopterygids, deinonychosaurs and avialians (Naish et al. 2011). According to this tree [shown below: click to enlarge], neither Archaeopteryx nor scansoriopterygids are members of Avialae. Life restoration of Xiaotingia zhengi, (c) Xing Lida and Liu Yi.
As you’ll know if you’ve been keeping up with the news, Xu et al. (2011) recently published the new Liaoning maniraptoran Xiaotingia and recovered a phylogeny where Archaeopteryx is a deinonychosaur, not an avialian (Xu et al. 2011). Hence all those headlines like “Archaeopteryx knocked off its perch” (groan), “Flap about Archaeopteryx” (double groan) and so on. We didn’t includeXiaotingia in our data set and our result is completely independent of Xu et al.’s: had our paper been published just a couple of weeks earlier, we’d have been first to recover Archaeopteryx as a non-avialian. Our 'whole theropod' analysis, with (again) our new big bird (Samrukia) within Ornithuromorpha. Some other taxa of special interest are marked with arrows, including Oviraptorosauria, Archaeopteryx and Avialae. From Naish et al. (2011, online supp. info).
Note that our result actually isn’t the same as Xu et al.’s seeing as, unlike them, we didn’t find Archaeopteryx to be a deinonychosaur. I’d also like to take this opportunity to note that the moving ofArchaeopteryx out of Avialae really isn’t a big deal, or a surprise – it amounts to a shift of a node or two, and quite a few theropod workers have been saying for years that it’s probably only a matter of time before phylogenies start findingArchaeopteryx to fall outside of Avialae. Given what we now know about early dromaeosaurids, troodontids and oviraptorosaurs, and about Anchiornis, scansoriopterygids and so on, it’s clear that working out the relationships among these confusing and often very similar feathered little maniraptorans is not going to be easy. Indeed, don’t go thinking that the notion of a non-avialian Archaeopteryx is necessarily here to stay!
Aves or Avialae? (assuming that the taxon shown is a member of the bird branch of Maniraptora). People are still fighting. Image by H. Raab, from wikipedia.
Fans of phylogenetic nomenclature will note that we use the term ‘Aves’ throughout our paper (Naish et al. 2011). I really dislike the use of the term Aves for the lineage that includes neornithines and all maniraptorans closer to them than to deinonychosaurs: Avialae, in my opinion, is superior since it should be understood – right from its earliest use (Gauthier 1986, p. 36) – to be maximally inclusive (Aves, in contrast, has been restricted by some authors to the avialian crown). However, some people who work on fossil birds (including one of our reviewers) really hate the term Avialae, and in this case we decided to make our lives easier and just change our favoured nomenclature (I won’t bore you with details of the long, tedious, pre-publication history of this paper. As is usual these days, it went round the houses before finally being accepted at Biology Letters).
Anyway, both our ‘birds only’ and ‘total theropod’ analyses recovered the same phylogenetic position for the Kazakh giant. Furthermore, this position matched what we had already concluded from our non-computer-assisted assessment of the specimen’s character distribution. This is far from the first time that I’ve had a parsimony analysis produce results that mostly matched the conclusions already reached by unassisted human brain power – a very similar thing happened with the neosauropod Xenoposeidon (Taylor & Naish 2007).
Even though we only have partial lower jaw rami for our new ornithuromorph, a few unique features were immediately apparent. The specimen possesses elongate sulci on the medial sides of its rami and a shelf-like lamina connects the bone next to the bottom part of this sulcus with the medial cotyle (see close-up photo below). Furthermore, the anatomy of the bone around the specimen’s mandibular cotyles is unique and distinctive.
Close-up posteromedial view of the articular region of the specimen's left ramus, showing the diagnostic medial sulcus and associated structures.
While I agree with the majority of my colleagues that naming new taxa for very scrappy remains should be avoided where possible, I’m also of the opinion that you should give names to things when you can clearly distinguish them from other things. If, in a fossil animal, you find autapomorphies – that is, unique, previously unreported anatomical features – you’re obliged to name what you have. Based on incomplete mandibular rami or not, our giant Kazakh bird jaw therefore needs a name, and hence Samrukia nessovi was born. The generic name is a nod to the Samruk, a mythological, phoenix-like Kazakh bird, while the specific name honours the contribution to Central Asian vertebrate palaeontology of the late Lev Nessov (1947-1995).
Big – but how big?
The most remarkable thing about Samrukia is its size. The longest of those incomplete rami is a whopping 27.5 cm long, and even the anterior-most tip of this ramus appears well short of the mandible’s original symphyseal region. The entire mandible was, therefore, damn near certainly over 30 cm long: the cranium was likely longer, since that’s typically the case. So, this Cretaceous bird had a skull over 30 cm long.
Two possible body shapes for the gigantic Samrukia, with a human and 'normal-sized' Mesozoic bird for scale. Image by John Conway.
How big was the whole animal? That, unfortunately, is just about unanswerable, since it’s impossible to work out anything for certain about the overall appearance of a bird when all you have is its (incomplete) lower jaw. We can’t even say whether Samrukiawas flightless or flight-capable – there aren’t any reliable inferences you can draw on this issue from jaw structure, nor from bone histology, bone wall thickness or anything like that. At the moment, it’s possible that Samrukia was a large, perhaps condor-shaped flying bird with a giant wingspan. Based on rough comparisons with other big, flying birds, it could have had a wingspan of 4 m or so and weighed at least 12 kg (Naish et al. 2011). But it’s equally plausible that it was a flightless bird shaped like a ratite, gastornithid or dromornithid; if so, it could have weighed more than 50 kg (Naish et al. 2011) and been somewhere between 2 and 3 m in standing height. These suggestions are highly speculative and, as we state in the paper, we can’t say anything with certainty about the size or appearance of this bird in the absence of better remains.
The majority of Mesozoic birds, and certainly those known from terrestrial environments, were small animals less than about 2 kg in mass and typically similar in size to modern finches, thrushes or crows. Note that some of the marine hesperornithines were very big, hence my specific reference to ‘terrestrial’ Mesozoic birds throughout this discussion. I should also note (given the fondness blog readers have for pickiness) that a few Mesozoic terrestrial birds, like Sapeornis and the enantiornithines Enantiornis, Martinavis and Avisaurus, had wingspans of between 1 and 1.5 m. Based on comparison with similar-sized modern hawks, eagles and gulls, such birds would still be less than 2 kg in mass.
We haven’t forgotten you, Gargantuavis
The Gargantuavis holotype synsacrum and partial pelvis (r) and referred femur (l). Images by Ghedoghedo, from wikipedia.Anyway, Samrukia was obviously a giant, and hence very special. But we’re careful to note in our paper that it isn’t unique. Gargantuavis philoinosfrom the Late Cretaceous of France is already well established as a terrestrial giant (Buffetaut et al. 1995, Buffetaut & Le Loeuff 1998, 2011). So, Samrukiaisn’t the first giant, terrestrial Cretaceous bird; it’s the second.
Actually, when we started our research on Samrukia a minor disagreement over the avialian status of Gargantuavis was occurring – some palaeornithologists were suggesting that it might not be a bird at all but actually an azhdarchid pterosaur (Mayr 2009). We didn’t agree with this (in my opinion, Gargantuavis has always very obviously been a bird – and I speak as someone who works on azhdarchids as well as on Mesozoic and Cenozoic birds). However, it seemed only appropriate to note that the status of Gargantuavis was at least under suspicion, or under discussion, at the time of writing; the best cause of action was to say that Samrukia confirmed the existence of giant terrestrial birds in the Cretaceous. And while our paper was in review, a new article on Gargantuavis appeared in which Eric Buffetaut and Jean Le Loeuff argued strongly against suggestions that Gargantuavis might be a pterosaur (Buffetaut & Le Loeuff 2011).
Speculative reconstruction of Gargantuavis, by Hyrotrioskjan. While shown here as an ostrich-like, small-headed form, it could have been big-skulled like a gastornithid or dromornithid.The phylogenetic position ofGargantuavis is uncertain, but various of its characters suggest that it might be a basal ornithuromorph and hence from the same approximate region of the avialian cladogram as Samrukia. Could Gargantuavis and Samrukia be close relatives, or even the same thing? For now, we lack evidence to evaluate this idea further, but I’m certainly not averse to the idea that they could be sister-taxa. As we state in the paper, however, “the restriction of Gargantuavis to a younger, western European fauna with no close biogeographic ties to the Santonian-Campanian of Central Asia renders it unlikely that Gargantuavis and Samrukia are congeneric” (Naish et al. 2011, p. 3).
Friends and neighbours
Map of the Republic of Kazakhstan, from Naish et al. (2011). Several sites in the south of the country are well known for yielding Cretaceous vertebrate fossils, notably Akkurgan (marked with a star), Shakh-Shakh and Syuk-Syuk.What I regard as one of the most interesting things about bothGargantuavis and Samrukia is that neither was living in isolation on some Cretaceous island. Rather, both were living alongside assemblages of non-avialian dinosaurs and also alongside large pterosaurs. Samrukia comes from the Bostobynskaya Formation (also known as the Bostobe Formation). Hadrosaurs and tyrannosaurids are known from the Akkurgan locality in Kazakhstan that yielded Samrukia, but ornithomimids, therizinosaurs, caenagnathids (how ironic), dromaeosaurids, ankylosaurs and rare sauropods have all been reported from the Bostobynskaya Formation as well (Dyke & Malakhov 2004, Averianov 2007a). All should be imagined as close neighbours of this giant bird. The azhdarchid pterosaurAralazhdarcho is also from the Bostobynskaya Formation (Averianov 2004, 2007b).
Virtually all of these Kazakh fossils come from floodplain habitats, but associated wood fragments indicate that forests were present nearby. Sharks, salamanders, turtles and the remains of other aquatic organisms show that pools, lakes and large, occasionally quasi-marine meandering rivers were also present across the region.
Samrukia (in both speculative volant and flightless forms) to approximate scale with some contemporaneous taxa: a therizinosaur, tyrannosaurid and azhdarchid pterosaur. Samrukia images by John Conway, azhdarchid by Mark Witton, therizinosaur and tyrannosaurid by me.
What does this mean for Samrukia, and for the composition of Cretaceous ecosystems as a whole? If Samrukiawas flightless, maybe it was able to avoid tyrannosaurids and other predatory theropods by being cursorial, but if it was flight-capable then we have to imagine it soaring over the heads of its terrestrial cousins and literally sharing the skies with similar-sized azhdarchid pterosaurs. Overall, the discovery of Samrukia provides additional evidence for a more diverse Late Cretaceous world than the one we’ve been inclined to imagine. The Late Cretaecous continental realm wasn’t a ‘non-avialian-dinosaurs-only theme park’; there was ecological ‘space’ for big birds and also for reasonably big, terrestrial crocodilians, squamates and even synapsids (and if any of this sounds familiar, it’s because I covered the same subject back in 2007 when talking about Mesozoic sebecosuchian crocodilians).
So, hello Samrukia, welcome to the ranks. What’s next? As regards Samrukia… as is so often the case, we need more material before we can go any further.