In 2003, a fossil was discovered in an  ancient lakebed in what’s now China.

It was an almost complete skeleton about the size  of a platypus, with teeth that resembled a seal’s.

And in the rock surrounding  it, there was a shadow.

That shadow was the preserved  impression of skin and other   soft tissue – including a furry pelt,  webbed feet, and a paddle-shaped tail.

It looked like a typical aquatic mammal, the  kind you might see near a river or lake today.

But this creature lived in the Jurassic period,   164 million years ago, and it  wasn’t technically a mammal.

It was a mammaliaform, part of the group that  includes mammals and our ancient relatives.

And for a long time, scientists had thought  of the mammaliaformes that lived during the   Mesozoic Era as small, shrew-like insectivore  generalists scurrying in the shadows of dinosaurs.

But this discovery complicated  the history of mammaliaforms...

It painted a picture of their explosive  diversification, their mysterious disappearance,   and how our own ancestors might have survived  thanks to a leg up from some leafy allies… Now, we know that placental mammals like us  humans belong to a group called Eutheria,   while marsupials like kangaroos and koalas  belong to a group called Metatheria.

Eutherians and metatherians share  a more recent common ancestor than   other mammal groups, so together  we’re referred to as ‘therians.’ And monotremes, like the platypus and echidna,  are our next closest living relatives.

Together, therians, monotremes and  all the animals back to our most   recent common ancestor form what’s  called the ‘crown-mammal’ group.

The mammaliaformes are basically  one more step removed from there.

It’s the group that includes crown-mammals  and their most recent common ancestor – plus   all the mammal-like animals that are more closely  related to mammals than they are to anything else.

The common ancestor of the mammaliaformes would  have looked something like Morganucodon– a small,   insect-eating generalist that scurried through the   undergrowth of the Late Triassic,  around 200 million years ago.

And the descendants of mammaliaformes  like Morganucodon would eventually   give rise to our mammal ancestors,  which were also small insectivores.

For a long time, it was thought that  all Mesozoic mammaliaformes were built   like this because of the imposing  ecological presence of the dinosaurs.

Many thought it wasn’t until the  Cretaceous-Paleogene extinction,   that space was freed up for the  mammals to adapt to new niches.

But… mammaliaformes were  rare in the fossil record.

And paleontologists were lucky to  find more than a few teeth or parts   of a jaw.

So this hypothesis wasn’t  exactly based on a ton of material.

Occasionally, more substantial fossils were  uncovered, but even they were mostly tiny,   like the genus from China  known as Zhangheotherium.

And while it’s true that most  mammaliaformes did fit this description,   a site in Northern China has  recently flipped the script.

There, layers of Jurassic lake sediment and   volcanic ash deposits have preserved the  fossils of a huge variety of organisms.

And that’s where our mysterious,  paddle-tailed mammaliaform was discovered.

It belonged to an ancient group called the  docodonts, and was named Castorocauda – from   castor meaning beaver and cauda meaning  tail, after its, well, beaver-shaped tail.

Measuring at least 42.5 centimeters  long and weighing up to 800 grams,   it may not have been huge by our standards, but  it’s the biggest Jurassic mammaliaform ever found.

Castorocauda also sported the oldest evidence of  a furry pelt, which indicates it was warm-blooded.

This suggests that some of the  characteristics we associate   with crown-mammals are actually ancestral  traits that existed long before we emerged.

But its size and its fur are only part  of what made this fossil so special.

Castorocauda was a far cry from  a tiny, scurrying generalist.

With its webbed feet, paddle tail  and teeth adapted for catching fish,   the scientists who described it recognized  it as a semi-aquatic specialist.

And this pushed the emergence of the first aquatic  mammaliaformes back by over 100 million years.

Which meant that mammaliaformes were developing  specialized traits much earlier than we thought,   despite the supposedly restrictive  presence of the dinosaurs.

And it soon became clear that Castorocauda wasn’t  the only fuzzy innovator in this fossil deposit.

Over the next few years,  scientists began uncovering   more evidence that they’d been kinda  wrong about these ancient critters.

Mammaliaformes had actually experienced  a period of major ecological radiation.

And not only that, this radiation revealed   fast rates of evolutionary change  that peaked in the mid-Jurassic.

But why?

What caused it?

Well, when separate lineages  start to radiate at the same time,   it suggests that something is going on in the  environment that benefits multiple groups.

During this time, Pangea was in the process  of breaking up, which transformed ecosystems.

And scientists soon realized that this  breakup gave docodonts and other groups   the chance to radiate into new niches  and develop brand new adaptations... Like the tunneling Docofossor, with its sprawling  limbs and paws like a mole’s for shoveling earth.

Or the squirrel-like, sap-eating Agilodocodon,   with its curved claws and flexible elbow,  wrist, and ankle joints for climbing.

Members of another mammaliaform group,  Haramiyida, were also getting in on the action.

The fossilized imprint of skin membranes  stretching between their limbs indicates   that these animals could glide from tree to tree.

In addition to the environmental change,   this mammaliaform radiation might have  also been a result of good timing.

Because, ever since the Triassic, mammaliaform  groups had been gaining adaptations like fur,   warm-bloodedness, lactation, and modifications to  the jaw and inner ear – sometimes convergently.

Developing this “critical mass”  of advantageous traits could   have made it easier for them to explore new  environments and adapt to specialized roles.

Now, the ancestors of crown-mammals were  also a part of this mid-Jurassic radiation.

But while they saw a major increase in  species diversity, they stayed shrew-like… For example, the late Jurassic Juramaia,   another mammal from that lakebed  formation and the earliest known therian.

If Juramaia wasn’t our direct ancestor,  it was at least a close relative,   and would have looked a lot  like our Jurassic ancestors.

So if other mammaliaformes weren’t nearly  as constrained by dinosaurs as we thought,   why did our ancestors stay so small?

Well, a recent study found that other  mammaliaformes actually had a greater   role in suppressing the diversification  of the therians than the dinosaurs did,   because of their good timing.

Those mammaliaformes had shown up on the  scene earlier than the crown-mammals,   and occupied the available specialist niches  before therians had a chance to compete.

Now, that doesn’t mean that  the dinosaurs had no influence.

As far as we know, they did  prevent ancient mammaliaformes   from ever getting bigger than a hefty housecat.

But it goes both ways – the presence of  mammaliaformes in these smaller-bodied   roles would have also limited the ability  of dinosaurs to exploit those same niches.

And our ancestors were perfectly suited  to the insectivore niche they occupied,   which may have been their saving grace  for the changes that were to come.

After all, there’s a reason that we’re here now.

So why did our ancestors survive  while other mammaliaformes didn’t?

Trying to figure out how, when and why most  of the others disappeared is challenging.

Because, we can’t identify a line  in the fossil record with diverse   mammaliaform remains on one side, and  only therians and monotremes on the other.

Even with all the new discoveries,  mammaliaform fossils are still relatively rare.

And when they do turn up, it’s usually only teeth.

But those teeth allow us to identify  which group these critters belonged   to and draw conclusions about their lifestyle,  distribution, and diversity at different times.

For example, from the mid to late Jurassic,  about 175 to 145 million years ago,   there are teeth belonging to all kinds  of mammaliaformes in the fossil record.

But around this time, many mammaliaformes  were also starting to experience a decline.

This decline was not marked  by a destructive cataclysm,   but instead by a gradual, two-phase process.

A study analyzing mammaliaform  fossil distribution throughout   the Mesozoic found that they first  experienced a restriction in range.

And then during the Cretaceous,  around 125 million years ago,   the record reveals a decrease in species richness.

During this time, in most areas,   two types of teeth replaced all  the others in the fossil record.

The first type were cimolodont molars,  belonging to the multituberculates,   a rat-like group of crown-mammals  whose teeth mark them as plant-eaters.

The second type were tribosphenic molars.

These are the kind of molars you have,   and they’re the kind of molars  ancestral placentals and marsupials had.

Studies indicate that this kind  of tooth has a greater capacity   to evolve into varied forms than  some of the other tooth types.

And in the late Cretaceous, they took  on a form that could cut, pierce,   and chew exoskeletons -- these were the  teeth of a small, insectivorous generalist.

This change coincided with the  Cretaceous Terrestrial Revolution,   a time of ecological upheaval that  lasted from 125 to 80 million years ago.

It was an event that would alter ecosystems  forever: the spread of flowering plants.

Flowering plants, or angiosperms,   made up less than 5% of terrestrial  plant species at the start of the period.

But by the end of the Cretaceous,   they dominated the land, comprising  over 80% of terrestrial plant species.

As flowering plants spread, small insecty  mammals were in the perfect position to   take advantage of the insect pollinators that  diversified in response to this expanding niche.

The ability of angiosperms to  outcompete other plants in many   regions even helped make the climate  more humid, expanding the world’s wet,   tropical habitats, which began to spread  into what’s now North America and Eurasia.

And mammals’ ability to climb trees  and travel across uneven terrain   likely helped them to navigate  the spreading tropical forests.

At the same time, the growing abundance  of fleshy fruit and seeds provided new   nutritional opportunities for the  herbivorous multituberculates.

And their high energy needs may have driven  them to compete more fiercely for resources.

Such a dramatic change in the environment would  have been a stressor for highly specialized   animals that relied on specific resources,  leaving many mammaliaformes struggling to adapt.

But small, generalist therians that were  less reliant on stable conditions persisted.

In this way, evidence suggests that the spread  of angiosperms helped our ancestors succeed,   by tipping the scales against their  ancient mammaliaform relatives.

Some recently-published fossil  finds suggest that during this time,   monotremes had radiated into  diverse forms in Australia.

But therians weren’t free to take over  quite yet – other crown-mammal groups   were still around and had started to specialize.

The most successful of these was the  multituberculates, a group that had taken   advantage of the new abundance of vegetation  brought on by the spread of angiosperms.

They and several other mammal groups would  go on to survive the K-Pg mass extinction,   diversifying and competing with  placentals, marsupials, and monotremes.

But we now know that this era of evolutionary  rivalry was nothing new for the mammals.

Instead, it was the next chapter of a  story that had played out over millions   of years of mammaliaform history, from the  diverse docodonts of the Jurassic to the   angiosperms that helped our ancestors  get a foot in the evolutionary door.

Eventually, competition and a changing  climate caused the last of those other   mammal groups to disappear, leaving the  monotremes and therians to diversify.

The diversification resulted in the return  of countless forms, from the shovel paws,   to the gliding membranes, and the paddle  tails that our ancient relatives pioneered.

And that shift left our own group, the placentals,   to become the most successful,  diverse mammaliaformes to ever live.