Aerosteon riocoloradensis: A Very Cool Dinosaur from Argentina

Fossils of a newly discovered species of dinosaur -- a 10-meter-long, elephant-weight predator -- were discovered in 1996 along the banks of Argentina's Rio Colorado, and are now being reported after a long period of careful study. This dinosaur dates to about 85 million years (which falls within the Cretaceous period).

Perhaps the most interesting feature of Aerosteon riocoloradensis is that it demonstrates the evolution of a bird-like respiratory system in an animal that is definitely not bird-like in most other ways. Indeed, the authors of this paper imply that this dinosaur's respiratory system represents an early phase in the evolution of the bird's respiratory system. This is a case of an adaptation arising in one context and later being used in an entirely different context.

This is how evolution often works: An adaptation arises for one reason, and then that adaptation is employed in a similar way but for a very different purpose (it is "exapted" as Stephen Gould would say) in a later organism. In this case, we see a lung adapted for a certain kind of efficiency in a terrestrial organism. Later, if the authors are correct, this same efficiency-enhancing adaptation is used in birds as an adaptation related to flight, an entirely different reason.

This diagram shows a reconstruction of the dinosaur Aerosteon riocoloradensis in outline, with the features identified as related to the pneumatic (breathing-related) adaptations indicated. You can access the original paper on line if you wish to study these features in more detail. Suffice it to say that there are a lot of pneumatic features of interest!

In order to understand the significance of this find, it may be helpful to first discuss some interesting things about tetrapod lungs, and bird lungs in particular.

Let's start with a useful digression about breathing. Obviously, the rate of energy demand by an air breathing organism (which might relate to the organism's movement pattern) determines the required breathing rate. It is interesting to note, then, that the breathing rate of many organisms is determined to some degree ... sometimes completely ... by the movement pattern of the organism. The classic example of this is the medium to large size cursorial (ground-walking) mammalian quadruped, such as a horse or a cheetah (perhaps the two most studied such animals in relation to running). When a horse or cheetah is running, the animal automatically exhales when it hits the ground, and thus inhales when it is in the air-borne part of its stride. Thus, the rate of breathing is determined by the rate of running. That may sound useful because one would want to increase breathing rate when running faster, but in truth, this is NOT even close to the best engineered system. The fact is that as a cheetah or a horse runs father, the rate of breathing should increase much more than it is allowed to by the movement pattern. This is the limiting factor on distance and/or speed of these mammals.

This limiting factor (breathing rate) affects two different systems, and which one conks out first will depend on the situation: 1) This limits the rate of oxygen intake; and 2) this limits the cooling effects of breathing, which for many quadrupedal mammals would be the main effect. Cheetahs do not stop running after a quick burst because they are tired out. They stop because they are overheated.

Humans do not have this limitation. Being bipedal, humans can increase their breathing rate as they run. For this reason a fit human with proper experience can run for hours without stopping, while a typical quadruped (like a deer, antelope, horse, or cheetah) must stop much more frequently. Some human groups today and in recent times are known to have hunted this way: Running the animal down. Don't try this with a cheetah, they can bite and pant at the same time!

If this was a constraint on birds ... say, if every flap of the wing required a breathing cycle ... then you would see birds standing around resting all of the time, and there would be no long-range migratory species of birds. Also, many birds fly (typically while migrating) at very high altitudes, where there is less oxygen per breath. And, of course, generally speaking flight is very demanding in terms of energy (this affects a wide range of adaptations. See this for example.)

Birds manage this problem by having a very different respiratory system than other tetrapods. Bird lungs are fairly rigid. They do not expand and contract, letting air in and pushing air out in a tidal fashion, as mammalian lungs do. Rather, they are set up to have air pass through them in one direction (from front to back) continuously. The movement of air in mammals is facilitated by the diaphragm contracting and expanding, and thus inflating and deflating the lung. In birds, the movement of air is facilitated by a system of air sacs. Many of the air sacs act as bellows moving the air at a continuous and controlled rate through the lungs. This rate can be increased or decreased as needed irrespective of the movement of the animal's body.

From Figure 1 of the paper: Cranial sinus and postcranial air sac systems in birds. All pneumatic spaces are paired except the clavicular air sac, and the lungs are shaded. Abbreviations: aas, abdominal air sac; atas, anterior thoracic air sac; cas, cervical air sac; clas, clavicular air sac; hd, humeral diverticulum of the clavicular air sac; lu, lung; pns, paranasal sinus; ptas, posterior thoracic air sac; pts, paratympanic sinus; t, trachea.

One of the interesting features of bird air sacs is that some part of some of the air sacs grow into some of the bird's bones.

This is the best scenario for a paleontologist. Soft tissue adaptations are nice, but typically do not preserve as fossils. Therefore, we don't know what early bird lungs look like. But since the air sacs involve the bones, which do preserve as fossils, there is hope to understand the evolution of this trait in birds.

It turns out that Aerosteon riocoloradensis has numerous pneumatic adaptations in its skeleton (see the diagram at the top of this post). The PLoS paper clearly documents these; There is no question whatsoever that these are present, and there is no question as to what they are for. It is almost certainly the case that Aerosteon riocoloradensis had a somewhat bird-like respiratory system.

But, if you look at Aerosteon riocoloradensis as reconstructed, it clearly did not fly. Also, as a biped, it may not have been as constrained in its breathing cycle as a mammalian horse or cheetah would be (I'm oversimplifying that argument a bit). But there are at least two reasons that this dinosaur would benefit from this adaptation: First, long distance high energy cursorial hunting would be facilitated by an extremely efficient respiratory system. In other words, this large allosaurus like dinosaur would have been able to prance along after its prey for longer than the less aerobically efficient, possibly quadrupedal tidal-lung equipped prey would be able to prance away. What a great hunting adaptation: To not have to stop and rest as often as your prey. Second, this adaptation would have been very effective for cooling down.

The authors of this paper have added a third benefit of the rigid lung facilitated by air sacs: Reduction of upper body mass to allow the bipedal animal to be better balanced. This idea relates to one of the odd features of Aerosteon riocoloradensis's air sacs: Some of the air sacs are located in an unusual place, involving the "belly ribs" on the lower part of the body. The authors see this as evidence that the front of the animal had a system of air tubes in it's skin, which they view as an adaptation for cooling.

The authors have combined the results of this analysis with other information to provide a model for the evolution of avian air sacs and one way bellows-driven lung ventilation. Here (paraphrased from the original paper for easier reading) are the four phases they outline, which you can link to the diagram provided below.

Phase I
Elaboration of paraxial cervical air sacs in basal theropods (bipedal dinosaurs) no later than the earliest Late Triassic.
Phase II
Differentiation of so-called avian ventilatory air sacs, including both cranial (clavicular air sac) and caudal (abdominal air sac) divisions, in basal "tetanurans" (theropod dinosaurs) during the Jurassic. A heterogeneous respiratory tract with flexible air sacs that could be used as bellows, in turn, suggests the presence of rigid, dorsally attached lungs with flow-through ventilation, as found today in birds.
Phase III
Evolution of a primitive air pump involving the sternum and ribs in maniraptoriform theropods, the group that includes today's birds, before the close of the Jurassic.
Phase IV
Evolution of an advanced air pump involving the sternum and rib in maniraptoran theropods before the close of the Jurassic.

In addition, the authors conclude:

  • The advent of avian unidirectional lung ventilation is not possible to pinpoint, as osteological correlates have yet to be identified for uni- or bidirectional lung ventilation.
  • The origin and evolution of avian air sacs may have been driven by one or more of the following three factors: flow-through lung ventilation, locomotory balance, and/or thermal regulation.


In addition to the importance of this find in relation to the evolution of respiratory systems in dinosaurs and birds, Aerosteon riocoloradensis is also important for another reason. From the paper:

The new theropod is particularly interesting for another reason; it represents a previously unrecorded lineage of large-bodied predator in the early Late Cretaceous (Santonian, ca. 84 Ma) of South America. Most large-bodied Cretaceous theropods on southern continents (South America, Africa, Madagascar, India) pertain to one of three distinctive and contemporary clades: abelisaurids..., spinosaurids..., or carcharodontosaurids .... This predatory triumvirate persisted for millions of years on both South America and Africa.... Although initially thought to be a late-surviving carcharodontosaurid..., Aerosteon preserves cranial bones that bear none of the distinguishing features of carcharodontosaurids.... Rather, Aerosteon represents a distinctive basal tetanuran lineage that has survived into the Late Cretaceous on South America and is possibly linked to the allosauroid radiation of the Jurassic.

See also this post at A Blog Around the Clock

Sereno PC, Martinez RN, Wilson JA, Varricchio DJ, Alcober OA, et al. (2008) Evidence for Avian Intrathoracic Air Sacs in a New Predatory Dinosaur from Argentina. PLoS ONE 3(9): e3303. doi:10.1371/journal.pone.0003303.

You can get the paper here.

More like this

I saw a brief video clip about this on the Daily Planet, Discovery Canada's daily science news show. A new model of the skeleton was "unveiled" at one of their museums. Thanks very much for providing the science background to go with the video of curious children gazing at the skeleton.

Interesting, but the conclusions might be somewhat overhyped. . .many of the observations on pneumaticity were made by other paleontologists in previous papers!!! See work by Pat O'Connor, Leon Claessens, Matt Wedel, etc. . .(of course, these guys are cited - I think much of the problem has arisen in the translation to mass media. . .)

Not that Sereno and company don't add some interesting details/speculation/lines for future research, of course. . .and kudos on another cool theropod (published in PLoS One, too!)!

Are there any known four-legged predatory dinosaurs? All of the ones I can think of ran around on two legs.....

By oscar zoalaster (not verified) on 30 Sep 2008 #permalink

In this case, we see a respiratory system adapted for a certain kind of efficiency in a quadruped.

No, a biped, as you get right later.

Are there any known four-legged predatory dinosaurs?


By David Marjanović (not verified) on 30 Sep 2008 #permalink

This was an amazing post. I learned a lot about the lungs of birds, which is very interesting. Also I would have thought that there would have been four legged predators, but having two more legs could slow them down. While you are chasing your prey you want to be fast and agile but having four legs can really slow you down.

Absolutely fascinating post! Thanks so much for putting all this stuff in language even I can understand. Most informative and keep up the good work.

By Nebularry (not verified) on 01 Oct 2008 #permalink

having four legs can really slow you down.

What? How?

There were lots and lots of quadrupedal predators in the Triassic. They just weren't dinosaurs; they were more closely related to the crocodiles. And what about today's mammalian predators...?

By David Marjanović (not verified) on 01 Oct 2008 #permalink

Cool post! Well written and very informative, thank you!

By galapagos (not verified) on 01 Oct 2008 #permalink

Wow this is very informative. I never thought about birds lungs, and how the birds lungs and the dinasaurs lungs were related. I think that it is really cool how dinasaurs ran faster on two legs rather then four legs, because it is quite the opposite now. Now the fastest land animal is the cheetah, and that has 4 legs, and we have two legs, but we run much slower than the cheetah. This is very confusing to me,could you please explain why this is?

Awesome! Evolution is absolutely mind boggling. This find is an interesting link in the lineage leading to the modern day bird.
If being bipedal is more desirable than being a quadruped why then do we not see an abundance of bipedal predators today? Concerning balance, isn't that what the tail of the dinosaur is for. Also it is unfair to compare running of the cheetah to human running. The cheetah is built for speed, using bursts to catch it's prey. Though the human can run for hours on end, similarly to the cheetah we cannot full out sprint for extended periods of time or we to will become overheated. If the fastest animals back then were bipedal why are the fastest today quadrupedal?
Thank you for the fascinating post.

I'm not sure that dinosaurs with two legs ran faster than dinosaurs with four legs. Maybe they just ran for longer, or maybe they had some other thing going on that was not speed or even running per se.

In comparing humans and quadrupeds, remember that we are comparing an animal evolved in a lineage of four legged runners that probably goes back almost to the beginning of mammals, vs. an animal derived from climbers (primates).

The better comparison might be between bipedal and quadrupedal rodents, or between kangaroos and deer. Off hand I don't know what that comparison would yield in terms of relative speed or endurance.

Bipeds are traditionally thought to be faster than quadrupeds for a given size and approximate physiology, though this is of course incredibly hard to make comparisons. One (admittedly exceptional) example is the kangaroo which can produce both high speeds and excellent endurance thanks to it's wonderfully locomotor system (that actually gets *more* efficient at high speed). The reason most quadrupeds are faster than bipeds in the modern world is simply that bipeds are rare and tend to be quite unusual or not modified for fast running (like apes) whereas there are tons and tons of quadrupeds to be able to pick a few fast ones from (like cheetah, pronghorn or springbok).

Comparing like with like is all but impossible unless you have a biped and quadruped of near identical species (which of course we don't rather by definition). Simply becuase you can reduce the overall mass an power requiremnts when you make a biped vs a quadruped you should (in theroy) be able to make them faster.

By Dave Hone (not verified) on 06 Oct 2008 #permalink

Dave .... yeah, but ... I can't think of a single cursorial quadruped that is not faster than humans, so while humans may be odd (and are) in terms of locomotion, they manage to be slower than ALL comparable quadrupeds (we are excluding the non-cursorial forms, so the sloth is out).

More to the point, this range of quadrupeds includes a very full range of body forms.

OK, maybe the giant ground pangolain is slower, but I'm not sure.

The idea that back in the day bipedal dinosaurs are faster than quadrupedal dinosaurs may be an example of one of those fundamental shifts in nature, such as the shift from when animals were mostly sessile (attached) and plants not to the opposite.

Hey Greg, Super post. Thanks for sharing information on the evolution of bird's lungs that I'm sure I never would have known without it.

Ha! Just you wait... As more and more similar discoveries are made it will be proven that dragons did exist.