17 million years ago in  the woodlands of Argentina,   a predator crept ever closer  to its unsuspecting prey.

Its eyes were locked on a small herbivore  taking delicate bites out of the bushes.

But this was not a familiar predator –  neither big cats nor wolves had evolved yet.

Instead, it was the otter-shaped  Arctodictis, a 2 meter-long marsupial.

And the animal it stared  hungrily at was Thoatherium,   a half-meter-tall plant-eater  with no living descendants.

Now, if this were a nature documentary, you know  that this could only go one of two ways, right?

Either Thoatherium would be  toast, or it would escape.

And these little guys and their relatives are  sometimes called ‘False Horses,’ for their hoofed,   horse-like legs, so a speedy escape  was a definite possibility – in theory.

But, I don't know if you know this, running is  a terrible adaptation for living in the woods.

It works well in open terrain where animals  can gain speed across vast distances.

But finding a clear path in thick,  brushy woodlands is much harder.

Today, most animals with similar adaptations  to Thoatherium live in open grasslands.

But these landscapes didn’t dominate until  the Pliocene, some 5 million years ago,   when grasslands came to cover close to  30% of North America – around the same   time as grasslands exploded in Argentina, too.

And Thoatherium didn't live 5 million  years ago.

It lived 17 million years ago.

So how did this false horse and its  relatives survive when they seemed to   be adapted for an ecosystem that didn't exist yet?

The story of Thoatherium starts nearly 66  million years ago, during a family breakup.

Thoatherium was a member of the  order Litopterna which were weird,   hooved mammals from South America, most  of whom had long limbs and few toes.

Their features were so strange  that, in the early 1900s,   scientists called their legs and toes  “inadaptive modification” – basically   saying that they were a bizarre mutation and  weren't really shaped by evolution at all.

Which is kind of silly when you think  about it, considering that they share   these traits with perissodactyls, the group  that includes horses, rhinos, and tapirs.

Both groups are herbivores, both have hooves, and  both tend to have an odd number of toes - 5, 3,   or 1.

So it's no surprise that they are related.

But that relationship ended in the smoking  crater of a dinosaur-killing asteroid around   66 million years ago, and the two groups  didn't spend much time together after that.

While perissodactyls spread  across Eurasia and North America,   litopterns stayed in the Southern Hemisphere.

The earliest litoptern fossils we've found so far  are nearly 63 million years old, from Patagonia.

Over their 60 million year lifespan,   this group split into nine families  and spread out across South America.

One of these families was the Prototheriidae,  or 'first beasts'.

This was the family that   contained Thoatherium and its  other small-bodied cousins.

And despite Thoatherium first  appearing more than 40 million   years after Litopterna broke up with  the ancestors of horses and rhinos,   they still had some…stunning similarities  – like, they both hated toes.

The trend of toe-loss is strongest in the  Prototheriidae, a.k.a.

the 'false horses'.

All of the members of this family  have some sort of monodactyly,   which means that they bear their weight  on one single toe, like modern horses do.

In some cases, this meant that they only  had one toe per limb like Thoatherium,   but in others they were functionally monodactyl.

This means that they had three toes,  but carried their weight exclusively   on the middle one – the side toes  were smaller and weren’t used much.

Horses lost their toes as grasslands  expanded in North America, having one   main toe by just over 10 million years ago,  while prototheriids lost toes much earlier,   leading to full monodactyly  by about 20 million years ago.

Ok but, why lose toes at all?

Well, for animals with hooves, there  are a lot of good reasons to get   rid of extra toes - and weirdly,  most of them are linked to grass… Specifically, to how grass and  grasslands change an ecosystem.

The most obvious change with grasslands  is that animals lose cover and places   to hide.

But grasses change a lot more  than that, right down to the dirt itself.

Instead of the soft, muddy soil that you see  in forests, grasslands are typically hard,   solid surfaces held together  by thousands of small roots.

And the stability hypothesis suggests hard  soil and toe loss are connected.

Essentially,   the idea is that having lots of toes  is only advantageous on softer ground.

Think of it like snowshoes, but for mud.

Big, wide feet help keep you from sinking.

But when the ground is hard and flat, that  wide foot isn't helpful anymore.

Instead,   skinnier feet and fewer toes  become more advantageous.

A second hypothesis deals with being  efficient.

See, when grasslands spread,   they tend to leave resources and hiding  spots distributed across vast distances.

Horses and litopterns also  have relatively long legs,   and they walk on the tips of their toes  like hooved ballerinas, if that helps you.

This makes them faster and  less prone to stubbing toes.

Isn't that... something... there's a  Disney movie in there, "Hooved Ballerinas".

And having longer limbs that weigh less  makes their movement more efficient.

The idea that this causes toe loss is  called the locomotor economy hypothesis.

Now, is it just me?

[stammers] I say this like every other episode:   paleontology is chock-full  of great unused band names.

"Locomotor Economy Hypothesis" I think I saw them in Seattle,  in like '92, you know?

But while these two hypotheses  seem to work for horse evolution,   they don't match the pattern or  timing of toe loss in litopterns.

Litopterns are preserved in fossil forests,   ecosystems that have neither hard ground  nor great distances to travel for resources.

So if toe loss isn’t linked to  grasslands, then what could it be?

Well, a final hypothesis for losing  toes has to do with body size.

Bigger bodies need better support,   especially against bending forces  which could cause their bones to snap.

Strong single digits resist that force better  than multiple skinny digits, because of physics.

Which may partially explain why horses evolved  larger sizes at the same time as they lost toes.

But many litopterns weren't  very big.

And Thoatherium,   the most horse-like of the false  horses was a measly 30 kg at most.

And another piece of the mystery  is that, unlike their toes,   Thoatherium's teeth are far better  adapted to eat bushes than grass.

Why be adapted for life in the  grass when you don’t even eat   it… especially when your bones  seem to be found in the woods?

The key may lie in an unexpected difference  between North America and South America.

Thoatherium and many other litoptern  fossils come from Patagonia, a region   of Argentina with a rich fossil record  that shows incredible change over time.

As mountains grew and shifted  and the planet slowly cooled,   Patagonia's ecosystem followed suit.

In the Eocene epoch, around 56 to 34 million  years ago, fossils show an incredible diversity   of plant life, including massive rainforests.

And, weirdly, also quite a lot of grass.

That in and of itself isn't too unexpected.

Grass has existed since the Early Cretaceous.

But what hasn't existed until  a lot more recently is grass   thick enough to change the environment entirely.

But in Patagonia, we see patches of mollisols,   a type of soil that only forms  when you have pretty thick grass.

Those don't show up in North  America until the Oligocene,   some 11 to 20 million years  after they appear in Patagonia.

But we see in the Eocene of Patagonia  both grass and tropical rainforests,   in the exact same formations at the  exact same time…a weird, contrasting   pattern that also continues throughout  the Oligocene, 34 to 23 million years ago.

And as the Oligocene gave way to the Miocene  epoch, from 23 to 5 million years ago,   forests were still present but shrinking…  while the grasslands were growing.

In North America, most formations just have one   or the other.

So how can you  have a region that has both?

It's called a mosaic landscape.

This is still common in some regions of  South America today, where patches of   forest are interspersed with patches of grassland.

This means that instead of evolving in  a place with vast open plains or dense,   impenetrable forest, litopterns  evolved in a place that had… both!

One type of mosaic landscape is called  a gallery forest, where forests grow   along the rivers and waterways, but are  separated by broad patches of open land.

Also in "Gallery Forest" the  first Friday of every month,   they have like the boxed wine and the little  cheese and crackers, it's open to the public.

But then after that, you have to... Thoatherium may have lived in this kind of  forest, which could explain its adaptations.

Its home was a region that was slowly drying  up, but still retained patches of lush forest,   full of tasty leaves that were a lot easier  to eat than the silica and grit-filled grass.

But those forests weren't continuous - so being  able to quickly cross stretches of open grasslands   to find safety and food in a neighboring chunk  of forest would still have been advantageous.

And while we don't have an estimate  for how fast Thoatherium was,   there's some research that suggests its other,   bigger relatives could reach sprinting speeds  of around 40 km/hr over short distances.

That's not quite as fast as a  horse, but still plenty fast.

So, litopterns like Thoatherium  may have essentially adapted to   cross an environment they didn't  live or eat in – grasslands – to   get to the environment that they  did live and eat in – forests.

It’s almost like if you imagine  whales as land dwellers who just   swam in between islands so much that  swimming became really important.

Their seemingly inexplicable mosaic of features   may have been a direct result  of the mosaic of the landscape.

Some adaptations were for life in the forests,   while others were focused on navigating  the grasslands to get to those forests.

And when we look at evolution,  especially adaptations to habitats,   we have to keep in mind that not  all places follow the same trends.

Much of the initial research on litopterns  and other South American fossils came from   the United States - where people  were used to the vast expanses   of grasslands and open prairies  that characterize North America.

So when they looked at South America,  they expected to see the same trends - so   they were confused when the fossil animals  didn't seem to follow those same patterns.

The lesson of Thoatherium is that  animals may have evolved similar   traits for very different reasons or the  same ones under different circumstances.

And sometimes, understanding why  requires adjusting your own expectations… Because it just might be  that, for some adaptations,