Dinosaurs Quadrupedal

Well, the time has come to talk about dinosaurs. In a blog about evolution, it was inevitable. Dinosaurs are, so to speak, the elephant in the room.

It is widely believed that dinosaurs are big and go “Rawr!’. While it is certainly true that dinosaurs go “Rawr!” (of course, there’s a Santa Claus, boys and girls!), it happens that many were not very large.

Some were, of course, brain-meltingly huge. But others were the size of a chicken. In fact dinosaurs came in such a diversity of sizes and shapes, that talking about dinosaurs is a bit like talking about mammals: there are just too many of them to be able to generalize much. Mammals can be giant like an elephant – or a whale – or they can be tiny like a shrew. And they’ve changed dramatically over evolutionary time. Same thing with dinos: the first dinosaurs were quite different from the ones that got smacked down by that asteroid at the end of their reign.

But there’s one thing that seems to have remained the same over the entire age of the dinosaurs: the carnivorous ones always had a bipedal stance. From Allosaurus to Velociraptor, they all seem to have run on two beautifully engineered legs. Herbivores came in a surprising array of shapes both two-legged and four; but carnivores? Just the two-legged variety.

Let’s think about why this would be, just for the fun of it. Like the blind men of mixed metaphor, let us grope the elephant in the middle of the room. It is an evolutionary question, of course. Why did one form evolve, but not another? The object here is not to solve the problem, but to look at how scientists go about solving problems of this general kind.

When asking a question like this, first we have to be sure that the pattern we are trying to explain is real. So we need to answer two questions:

1) Is it true that all carnivorous dinosaurs were bipedal (while herbivores walked on two or four feet)? Were there any exceptions?

2) Are we sure we have our categories straight? What is a dinosaur, anyway? Because if crocodiles count as dinosaurs, then we’ve just found an exception! Similarly, we need to ask what a carnivore is.

Let’s start with the second question. What is a dinosaur? What is a carnivorous dinosaur?

A group of species, such as Dinosaurs or Mammals, is defined biologically in terms of common ancestors and their descendants. Any grouping is considered valid so long as it includes the latest common ancestor of the group and all of that ancestor’s descendants. these valid groups are called “clades.” If you have a group of species and you exclude their common ancestor, then you don’t really have one valid group. You really have some other number of groups that are in a sense unrelated. Say for example that you are studying primates. The primates include (very roughly) apes, lemur-like things and a couple of kinds of monkeys. Oh, and us humans. To say that all primates are “related” is to say that the common ancestor of all primates was itself a primate. If we exclude the latest common ancestor from the group, then we are saying that “primateness” evolved in some other critter, possibly more than once. So, if we exclude the common ancestor of all primates, then somehow monkeys became primates in a way very different from how apes became primates. If we accepted that kind of group, we wouldn’t be able to draw any conclusions about the evolution of the group as a whole.

The group also has to include all the ancestor’s descendants. Thus, because we humans evolved from the common ancestor of all primates, we are primates. And because we evolved from the common ancestor of all apes, we are apes. Apes and primates are both clades, so long as they include humans. It doesn’t matter that our later evolution took us in quite a different direction, and it doesn’t matter that we think humans are special among all the species of the world. Fair’s fair!

So how do we apply this to dinosaurs? …Well, it’s messy.

For a long time, a dinosaur was anything that a paleontologist said was a dinosaur. Oh look! Harry dug up a new giant reptilian dead thing! Must be a dinosaur!

Eventually, paleontologists got serious about naming things correctly and systematically. They had a huge pile of species that they called dinosaurs, but which were quite different from each other. And it was clear that some were more closely related than others. What they did was to choose a set of extinct critters that most paleontologists could agree was a dinosaur, and then define a group based on those species alone, using our fair criteria about common ancestors. They defined dinosaurs as the set of all descendants of the common ancestor of a select group of agreed upon species. So, if you can show that your new fossil is descended from a dinosaur, then it too is a dinosaur.

If we do a bit of evolutionary calculation, we find that the common ancestor of all dinosaurs had a long S-shaped neck, and special features in its pelvis and ankle. Based on these criteria, we can establish this cladogram:


A cladogram – which is sort of mostly like an evolutionary tree – showing the relationships of the true dinosars to several other kinds of reptiles, including lizards, alligators (and crocodiles) and the pterosaurs, icthyosaurs and other “weird” reptiles. Each node (intersection of lines) represents a common ancestor. the Dinosaur node is marked with a labeled dot.

So, traditional dinosaurs such as Tyranosaurus, Apatosaurus, Triceratops and Iguanodon are still dinosaurs, but some of the other weird critters are something else. Excluded from Dinosauria, we have the group of flying reptiles called Pterosaurs, and the fully marine reptiles such as Mesosaurus, Mososaurus, Ichthyosaurus and Plesiosaurus. There are also a bunch of curious-looking terrestrial crocodile-like beasties that don’t make the cut.

In the process, though, we learned something interesting and new: the first dinosaur, the ancestor of all the others in the valid group, seems to have been bipedal! More on this later.


So we’ve more or less defined the dinosaurs. What about carnivores? The most important criterion for figuring out what a fossil species ate is its teeth. Teeth for herbivory are usually designed for either nipping or grinding: chisels or mortars-and-pestles. Carnivorous teeth are usually pointy, triangular things, preferably with a serrated edge. One variant of this is the fish-eating teeth, which are also pointy, but they are longer and conical, generally speaking – like having a bunch of fishhooks in your mouth.

Now, how do we know whether a dinosaur is bipedal? Basically, we look at the relative lengths of the front and hind legs, as well as trying to reconstruct the center of balance.

Fig 1

Note, front legs are shorter than hind legs.  Images copyright The Natural History Museum. See full credits at end of piece.

With all these criteria in mind, we can review the list of all carnivorous dinosaurs, and indeed we find… an exception.

Have you heard of Spinosaurus? It was hailed recently as one of the biggest carnivorous dinos ever. However, it seems that Spinosaurus may have had features that make it not quite bipedal. Its front legs are a bit too long and robust compared to the hind legs, for instance. But it also has many features that suggest that it was partly aquatic: nostrils on the top of its head and dense bones. For now, let us assume that being aquatic puts it outside our discussion of bipedal locomotion. In other words, let’s ignore it temporarily.

So provisionally, with one questionable exception, it looks like our pattern holds. Carnivorous dinosaurs were all bipedal, and all the quadrupedal dinos, such as Triceratops, Apatosaurus and Stegosaurus, were herbivores. There were bipedal plant-eaters too, but that’s ok; our question is about the absence of quadrupedal dinosaurs. If the pattern is real, we can finally start to ask questions about what shapes it!

Now, there are two distinct ways for a species to acquire a character: it can evolve it via natural selection, or it can just receive it from an ancestor. The latter is known as phylogenetic inertia: the tendency for species to keep an old, inherited characteristic, regardless of its value. We have to remember that changes don’t happen because a species looks ahead and plans something creatively. Each species has to work with what it has, modified only by more or less random changes. The random changes that work (or at least don’t suck too much) get kept. The ones that don’t work will fade away.

So in this light, it’s very important to note that the ancestor of all dinosaurs was bipedal. The quadrupedal herbivores seem to have settled down on four legs after the fact. In fact, we can still see the remnants of their bipedal ancestors in their skeletons. For example, the front legs tend to be a bit shorter than their back legs.

But why didn’t the carnivores do the same? Why didn’t some of them drop down on all fours? We don’t know. As far as I can tell, it hasn’t been studied extensively or conclusively (if somebody knows of a relevant study please tell me!).

So, I get to guess. Only, because this is science, I have to call it hypothesizing, and I have to try to back it up with experimental data. If there’s no data, it’s just speculaating.

There are four most obvious hypotheses.

1. There is no good reason. They inherited their bipedality from the Ur-dinosaur and were unable to lose it, even if they “wanted” to. This is called phylogenetic inertia.

2. Two legs are better than four when it comes to running fast to catch prey.

3. Running on two hind legs frees up the front legs for grabbing.

4. Some combination of hypotheses #2 and #3.

We need to keep the first hypothesis in mind, but it is very hard to test. In effect, the first hypothesis requires us to exhaust all other “adaptive” hypotheses, or find some other constraint. The very fact that loads of non-carnivorous dinos did become quadrupedal suggests that it can’t have been that hard to achieve. So, I’m going to wait until the end of this analysis to discuss it.

Choices 2 and 3 make sense, but how do they fit the pattern? Let’s start with an observation: plants don’t run very fast. So, some herbivores could have become quadrupedal because they didn’t need hands, which is consistent with hypothesis number 3. However, it should be noted that herbivores do have to occasionally run away from carnivores. This seems to constrain hypothesis 2. Can we find any other data to distinguish between hypotheses 2 and 3?

Sure. Let’s first ask whether bipedality is inherently faster or slower than quadrupedality. We have plenty of data on the running speeds of modern species. Scientists take into account things like the size of the animal, leg length, and even how springy the vertebrae are (the cheetah uses its back like a huge leaf-spring). Suffice it to say that when we consider the top speeds of the fastest land animals, it doesn’t seem that bipedality has an effect. However, it looks like bipedality might confer a slight advantage in acceleration. The cheetah achieves a higher top speed than the ostrich, but it is possible that the ostrich reaches its top speed more quickly (I don’t think there have been any direct comparisons of cheetah and ostrich acceleration to top speed). Also, bipedal animals might be able to turn more quickly.

Ok. what about the claws that catch? Here, bipedality has a distinct advantage. If you aren’t using your front legs for propulsion, they are free to grab your fleeing prey. And yet, the modern quadrupeds do a pretty decent job of catching a meal. There seem to be three different strategies: either you get a good hold with your teeth (as the crocodillians do), or you make your bite extra deadly, using venom or bacteria (viz snakes and the komodo dragon), or you adapt your front legs to do double duty, as cats do (and dogs to a lesser extent). In this last case, the evolution of cats serves to illustrate how useful it is to have front claws for grabbing.

So, where does that leave our dinosaurs? At a guess, I’d have to say that there are several factors that might have contributed to the bipedal pattern. The ancestor was bipedal, so there may have been a bit of phylogenetic inertia. But running bipedally might have been a very useful strategy. Once gained, there might have been no reason to lose it. Sometimes, it’s better to stick with a good strategy than to try to find a better one.

Finally, we have to consider our exception: Spinosaurus. Let us assume that Spinosaurus was fully quadrupedal. If these guys (Ibrahim et. al. 2014) are right about Spinosaurus being aquatic, then we probably don’t have a problem. Clearly, locomotion in the water is very different from walking or running on land, as is the prospect of catching prey while swimming. Pending some new study that says bipedal running is a good strategy for getting around under water, I think we can ignore Spinosaurus. We might even be able to use it to support our hypothesis (this was the original sense of the phrase “exception that proves the rule”. To “prove” in this sense is to “test”).

All of the above still leaves us with one niggling question: what about the herbivores? These were species that began with a bipedal stance and then did evolve a reason to lose it: to eat plants low on the ground, or to support very large bodies. Which of these was the most important driver of herbivore evolution? And why did some of the herbivores keep their bipedal stance? That’s the other side of the elephant.

Literature Cited

Ibrahim1, Nizar, Paul C. Sereno1, Cristiano Dal Sasso2, Simone Maganuco2, Matteo Fabbri3, David M. Martill4, Samir Zouhri5, Nathan Myhrvold6, Dawid A. Iurino7 2014. Science 26 Sept. Vol 345(6204): 1613-1616.

Image Credits

Stegosaurus. Via Wikimedia Commons,

Maidment SCR, Brassey C, Barrett PM (2015) The Postcranial Skeleton of an Exceptionally Complete Individual of the Plated Dinosaur Stegosaurus stenops (Dinosauria: Thyreophora) from the Upper Jurassic Morrison Formation of Wyoming, U.S.A. PLoS ONE 10(10): e0138352. doi:10.1371/journal.pone.013835

Cladogram drawn with the help of:

  1. Mesquite – Maddison, W. P. and D.R. Maddison. 2015. Mesquite: a modular system for evolutionary analysis.  Version 3.04  http://mesquiteproject.org
  2. GIMP (GNU Image Manipulation Program) – http://www.gimp.org


Special Thanks to James Farlow for providing advice about carnivorous dinosaurs. If I said anything inaccurate or just plain stupid in this piece, it wasn’t his fault!

Edit: This article first appeared with an inaccurate cladogram. I’ve replaced it with something that will make paleontologists less unhappy. The original cladogram was definitely not James Farlow’s fault, and he had nothing to do with the new one either.


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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