FEMALE NARRATOR: A bat fluttered through the sky above what is now Wyoming some 52 million years ago.

But it wasn't like the bats you and I know.

It was small like most modern bats, but it had claws on the tips of all five digits that supported each wing.

And its wings were a bit shorter while its hind legs were a little longer.

With those claws and long hind limbs, it was a better climber than most modern bats, and its teeth suggests that it ate insects.

But it probably didn't use echolocation to find and catch them.

This little flying mammal was onychonycteris, and it was definitely a bat, the most primitive bat that we have good fossil evidence of, and also one of the oldest.

And on the mammal family tree, it's set on the next closest branch to all other bats, living and extinct.

But when you trace that branch back, well, it's a mystery.

Bats pretty much appear in the fossil record as recognizable, full-on flying bats, and they show up on all of the continents except Antarctica around the same time in the early Eocene Epoch.

However, the very earliest fossils consist only of teeth and limb bones, which don't give us many clues about what the ancestors of today's bats looked like.

So where did bats come from and which of the many weird features that bats have showed up first?

Like, did these mammals learn to fly first, or echolocate first, or both?

And how do they fit into the mammal family tree?

We used to think that, based on their skeletal structure, bats were closely related to primates, like us, which is kind of strange when you think about it.

But once we were able to study their genetic history, their DNA revealed something much weirder.

Instead of being closely related to the mammals that they kind of look like, bats turned out to be much more closely related to the ones that they don't.

The fossil record of bats is both great and terrible, which is why we waited so long to do this episode.

It's great in that several of the earliest bat fossils we have are exquisitely preserved.

They're basically complete because they've been buried in an ancient lake deposit.

The best bat fossils come from these types of lagerstatten, or fossil sites with exceptional preservation.

On the other hand, the fossil record of bats is also terrible, because it's either that or nothing.

Most bats are small and have thin, fragile bones.

This keeps them light, which makes flight easier, but it also means they don't preserve well.

But we are lucky enough to have a few specimens of very old bats.

Flying around at the same time as onychonycteris and likely sharing the skies with it was icaronycteris index.

This species also comes from Wyoming.

And for a long time, it held the title of earliest known definitive bat, since it was first described in 1966.

Like onychonycteris, which was first published in 2008, icaronycteris dates to around 52 and 1/2 million years ago.

And because these two bats were already distinct from each other, we know that bats must have evolved sometime before this.

Both bats were clearly capable of powered flight, and also looked a lot like most modern insect-eating bats.

But modern bats only have claws on one or two of their digits, and onychonycteris had claws on all five digits, while icaronycteris had a well-developed claw on its second digit and bony tips on three others.

And this ancient bat was probably capable of echolocation, because it had specialized features on one of the inner ear bones called the malleus, and in the base of its skull that are linked to echolocation in living bats.

Also, early bats weren't restricted to North America.

There are at least four more spectacularly preserved taxa of bats from just a bit later, around 47 million years ago, from the Messel Pit in Germany, a site we've talked about before.

These are also complete or nearly complete skeletons, and in some of them you can even see outlines of their soft tissues.

And, like icarynycteris, they've got all the hallmarks of modern bats, and several of them seem to have been able to echolocate too.

But all of these fossil bats had already mastered powered flight, so how did their ancestors make it into the air in the first place?

The answer to that question is tied up with the evolution of another one of the defining features of most bats, echolocation.

And among the experts, there are three competing hypotheses for how bats evolved.

Either echolocation came first, or flight came first, or the two developed together.

All three hypotheses start with the same basic assumptions based on the traits that are most common in bats today.

These are that the bat ancestor was arboreal, or lived in the trees; it was insectivorous, or ate insects; and it was probably nocturnal.

In the echolocation first hypothesis, the bat ancestor would have already had ultrasonic capabilities to start off with.

This seems like it could be a possibility, because there are modern insect incestivores, like some species of shrew, that use ultrasound for communication or navigation.

So the thinking here goes that this ancestral bat might have reached out from tree branches to snatch passing insects.

Over time, the super high-pitched calls would have evolved into sonar that it could use to track its prey.

And being arboreal, its digits would have been selected to get longer, with a membrane stretched between them to more effectively capture food.

Those long, webbed hands would have then been co-opted for gliding when the animal started leaping to get at insects that flew further and further from its perch.

And eventually, it acquired adaptations for powered flight.

But the main problem with this hypothesis is that this kind of hunting behavior, catching insects that just happen to fly where you can reach them, hasn't been observed in the wild.

Plus, weirdly enough, it turns out that it takes a lot of energy to echolocate, especially when you're stationary.

So this whole feeding strategy seems pretty inefficient.

So what if they flew first instead?

In this model, powered flight evolved from a gliding ancestor, which had originally started by leaping between trees or branches.

This arboreal creature would have developed longer digits and a membrane between them as part of its gliding phase, eventually transitioning into powered flight.

Once this proto bat was flying around, it likely encountered insects, possibly scooping them up with its wings or catching them out of the air.

And from there, an energy- efficient form of echolocation developed, with bats exhaling and squeaking in time with their wing beats.

Some researchers have argued against this idea, saying that a leaping or gliding nocturnal animal without specialized senses, either vision or echolocation, wouldn't have been able to see where it was trying to land.

But here, at least, the fossil record can tell us something.

The skeleton of onychonycteris shows us that it was definitely capable of powered flight, but it didn't have the cranial features that are linked to echolocation.

Because the skull was partially crushed, we can't tell if it had larger eyes like many nocturnal gliding and leaping creatures do, such as flying squirrels, but it's still good evidence that flight likely came first.

But there's still the third option to consider, that maybe echolocation and flight evolved in tandem.

In this hypothesis, the bat ancestor originally used ultrasound to communicate, and was able to start using it like basic sonar to help it plot its nighttime leaps between branches.

As its ability to echolocate evolved in power, so too did its ability to make longer jumps, which eventually turned into gliding and powered flight.

And those two adaptations, stronger echolocation and powered flight, made bats the stealthy aerial predators of insects that many still are today.

The problem with this model is that onychonycteris didn't echolocate, but it did fly.

OK, so it seems that the fossil evidence we have is in favor of the flight-first hypothesis.

So that's one piece of the puzzle of bat origins that we can kind of snap into place.

But it still doesn't tell us where bats came from.

To figure out where bats truly fit in the mammal family tree, paleontologists have teamed up with geneticists to study the DNA of living bats.

For a long time, bats were thought to be part of the super order archonta, the group that includes tree shrews, colugos, and primates, because those are the mammals they look the most like.

Now, super orders are, by their very nature, incredibly diverse.

But members of archonta do share a lot of the same skeletal features, from the presence of a tiny bone in the inner ear to the particular way their ankle bones fit together.

And some studies even suggested that bats and colugos were more closely related to each other than to the rest of the group, based on some features of their hands, elbows, and feet.

Colugos are nocturnal, arboreal, and they glide using a membrane of skin stretched between their limbs, just like the transitional pre-bat is hypothesized to have done.

So it's easy to see why scientists thought that they were closely related.

And in the 1980s and '90s, an Australian neuroscientist even suggested that fruit bats evolved from primates, based on similarities in the patterns of connections between the retina and the brain.

But in more than two dozen molecular studies carried out since the early 1990s, bats have never grouped with archonta.

Instead, all of these studies put bats in a totally different super order, one known as Laurasiatheria.

This includes a number of the placental mammals that are thought to have originated on the supercontinent Laurasia during the late Cretaceous period.

And this group is also very diverse, including the orders containing moles, camels, horses, whales, pangolins, and bears, most of which look nothing like bats.

So that's right.

It turns out that bats are more closely related to whales than they are to us.

Within this group, analyses usually place bats with a clade that contains pangolins, carnivores, and ungulates, or as the sister group of shrews, moles, and maybe hedgehogs.

But it's still really unclear who, within Laurasiatheria, bats are most closely related to and how.

They seem to have come from some very primitive mammal near the base of Laurasiatheria that also gave rise to one or more of the other groups in the super order.

Another benefit of all these genetic data is that it can give us a sense of when bats became a thing.

According to studies based on that model known as the molecular clock, bats seem to have originated around 65 million years ago, just after the extinction of the non-avian dinosaurs.

So where does that leave us in the understanding of bat origins?

Well, we seem to be getting closer to figuring out the order in which bats evolved their most distinctive traits, and it looks like flight came first, followed closely by echolocation.

And we're still digging up wonderfully preserved early bat fossils.

As genomics continues to grow as a field, we will hopefully be able to zero in on exactly what group bats are closest to.

And this might be able to tell us what kinds of traits to look for in a bat ancestor.

It may end up looking totally different from what we've expected.

After all, it's happened before.

But for the moment, the lack of enough evidence, both in the ground and in their DNA, is keeping the true origins of bats in the dark.