In 1948, in the Gobi Desert of Mongolia, a Soviet fossil-hunting expedition uncovered three flat pieces of bone about a meter long.

The paleontologist who described them thought that they must have been the claws of some big species of turtle or turtle-like reptile that used them to forage for seaweed.

So he gave the fossils a name that translates to “turtle-shaped scythe lizard,” a.k.a.

Therizinosaurus cheloniformis.

But then, over twenty years later, the identity of these claws would change…a lot.

Because they were actually the enormous hand-claws of some kind of dinosaur.

But with so little other material, exactly what type of dinosaur they belonged to was still a mystery.

Over the next few decades, other bizarre fossils would be discovered in the same area.

Paleontologists found hips, neck bones, and a skull that looked like they all came from plant-eaters, along with more big claws, and odd four-toed feet.

It seemed like there was no way these fossils  could’ve all come from a  single group of dinosaurs.

And if they did, where did they belong in the dino family tree?

Figuring out the identity of these dinosaur mash-ups would take the discovery of more complete fossils in the 1980s and 1990s.

But the mystery didn’t end there… Because how the therizinosaurs lived and evolved  ended up being just as weird  as their mixed-up anatomy.

For decades, speculation reigned about where all these seemingly-different fossil animals fit in the family tree of dinosaurs.

Some researchers suggested that the strange remains from the Gobi came from theropods, the group that includes carnivorous dinosaurs  like Tyrannosaurus rex (maybe  you've heard of it), and Allosaurus.

Others went in a completely different direction – thinking that they might be late-surviving prosauropods, the ancestors of the long-necked giant plant-eaters known as sauropods.

Or maybe they were sauropods!

And it even seemed possible that they were from another, previously unknown branch of the dinosaur family tree.

The mystery continued until 1993, when a new dinosaur dated to between 125 and 113 million years ago was described from China.

It was named Alxasaurus and it had all of the strange features of those earlier fragmented finds combined in one animal, like slender jaws and a long neck.

If the previously discovered fossils from the Gobi Desert were like the mixed-up pieces of a jigsaw puzzle, then  Alxasaurus gave paleontologists  a look at the picture on the box.

This new skeleton showed how the fossils found before, including Therizinosaurus, could actually be from different species of the same family of just extremely weird-looking dinosaurs.

And this group was unlike anything anyone could have imagined.

These dinosaurs had the hips of a plant-eater, long arms that ended in long claws like a giant ground sloth, and four-toed feet like prosauropods.

If those features together weren’t strange enough, they also had long necks and small heads that contained tiny, leaf-shaped teeth like a sauropod, and beaks.

And, along with the discovery of Alxasaurus, came another study that would change the shape of the dinosaur family tree by figuring out where these weirdos fit into it.

It found that Therizinosaurus and its relatives were actually theropods after all.

And later work would place them in the subgroup  that also includes  dromaeosaurs, like Velociraptor, Deinonychus, and birds.

But these theropods looked nothing like  T.rex or any of their other  known carnivorous relatives.

Dromaeosaurs had more narrow hips and were  more lightly built than  Therizinosaurus, for example.

They also walked on three-toed feet, had shorter necks, and had larger heads relative to their body size with serrated, meat-cutting teeth.

And while dromaeosaurs also had long arms, they didn’t have the big claws on their hands that got Therizinosaurus mistaken for a clawed turtle.

So how and why did the therizinosaurs end up looking so different from their cousins?

Well, the answers to these two questions are actually linked together: these theropods look the way they do because they evolved into herbivores.

And in 2005, an early relative of Therizinosaurus  that was uncovered in Utah helped  us understand how this happened.

It’s called Falcarius and it dates back to at least 125 million years ago, in the Early Cretaceous Period.

It shows an in-between  phase between therizinosaurs  and their more carnivorous ancestors.

Falcarius was much smaller  than other therizinosaurs,  weighing around 100 kg.

It had more narrow hips, some pointed teeth that could’ve helped it eat prey, and the standard curved hand claws of a theropod.

But it also had some leaf-shaped teeth that point to it having a more omnivorous diet than a typical theropod, because this tooth shape is what we see in some plant-eating dinos.

And Falcarius had a much longer neck than most theropods of its size, too.

This let it reach about 1.5 meters off of the ground, which may have allowed it to   browse on leaves.

By the Late Cretaceous, therizinosaurs took these adaptations one step further.

Their hips evolved to become wider and more robust, which made room for a bigger gut to digest plant material, a trait we see in many herbivores.

Their hands began to have a greater range of motion, and their claws gradually became better-adapted as grasping hooks for foraging,  similar to the claws of  extinct giant ground sloths.

And their necks trended longer to browse on even higher foliage, while they also increased in overall body size.

But how could a theropod evolve to eat plants in the first place?

We’ve seen how their bodies changed over time, but what drove that change?

Well, in 2009, a study proposed two possible scenarios for how this could’ve happened.

One scenario pointed out that we already know of several groups of theropods that shifted from a more carnivorous diet to a more omnivorous one.

These include the ornithomimosaurs and the oviraptorosaurs.

So it’s possible that this change in diet happened multiple times independently in different groups of theropods.

The second scenario pointed out that these two groups and the therizinosaurs belong to a specific subgroup of  theropods called Coelurosauria,  along with dedicated carnivores like T. rex.

This subgroup is thought to share a common ancestor, which might’ve had a more omnivorous diet to start with.

This would mean that the therizinosaurs went from  omnivory to herbivory, and  were never carnivorous at all!

And the thing that drove this group to herbivory?

Well, Therizinosaurus and its ancestors probably adapted to eat plants to avoid competition with their carnivorous relatives.

Falcarius shared an environment with a host  of dromaeosaurs, troodontids,  and early tyrannosaur  relatives, so competitive  pressure among carnivores was likely pretty high.

And the specific adaptations of the therizinosaurs – like long necks and claws – probably also helped them avoid competition with  other herbivorous dinosaurs,  like hadrosaurs and ceratopsians.

Therizinosaurs that lived in the same place at the same time even became more specialized to avoid competition with each other!

So, amid all of this competition, the  therizinosaurs were pushed into their weird combination of features to carve out their own ecological niche.

And in solving this fossil mystery, the big picture that the adaptations of Therizinosaurus and its relatives paint for us is ultimately a much more diverse picture of theropods – they’re not all just variations of T. rex!

Not only that, but they’re also pretty weird for herbivorous dinosaurs, combining many of the traits we see in  other groups of plant-eaters  into one unique family.

From their initial mistaken  identity as giant-clawed  turtles, to the debates about how  they evolved, the therizinosaurs   have been surprising paleontologists for decades.

They’ve both shaken up our view of the dinosaur  family tree and shown us how much of the  fossil record we still have to explore.

Thanks to this month’s claw-some Eontologists!

I'm just getting used to it now.

Like there's just gonna be a pun every week, fine.

Gale Brown, Juan M., Jacksy  Weiss, Melanie Lam Carnevale,   Annie & Eric Higgins, John Davison Ng,  Colton, Chase Archambault, and Jake Hart.

Become an Eonite at patreon.com/eons and you  can get fun perks like  submitting a joke for me to read.

Like this one from The Burgins.

"How many dinosaurs does it  take to screw in a lightbulb?"

I'm not even gonna look.

I'm gonna say...11.

"Ha, trick question!

No dinosaur has paid  its electric bill in over 65 million years."

Then, there's the question of like, is the  electric bill - did they use fossil fuels???

Riddle me this.

Now the joker has become the jokee.

Or is it...?

Now the jokee has become the joker.

And as always thanks for joining me in the Adam Lowe studio.

Subscribe at youtube.com/eons  for more evolutionary escapades.

These dinosaurs had the hips of a...

I don't know why that cracks me up... has the hips of a plant eater.

[sighs] I'm unhinged now and that's  the kind of stuff I'm getting in my... [Paige, the script editor of this  episode:] "Hips Don't Lie..." Exactly!

The hips don't lie!

Now I have to get back on track.

I don't get paid by the hour.