MALE NARRATOR: In the arid coastal region of South America known as the Atacama Desert between the Andes Mountains and the Pacific Ocean there lie the remains of ancient behemoths of the sea.

At a site known as Cerro Ballena, or Whale Hill, there are more than 40 skeletons of marine mammals, a graveyard of ocean life dating back 6 and 1/2 million to 9 million years ago in the late Miocene Epoch.

Most of these marine mammals were big baleen whales of all ages.

But there are also skeletons of two different species of seals, an extinct sperm whale, an aquatic sloth, and a weird toothed whale that looked like a walrus.

And the skeletons were deposited there in four distinct layers over a period of about 10,000 to 16,000 years, so whatever killed these animals happened more than once.

Dense bone beds of marine mammals like this are pretty rare in the fossil record, which makes this site an incredible find.

But what happened?

How did these sea creatures end up buried in the desert?

Well, back in the Miocene, Cerro Ballena was a beach, a tidal flat where dead and dying animals washed ashore over time.

As for what killed all those creatures, researchers have considered a couple of different hypotheses, from a natural disaster like a tsunami to a fast-acting disease that somehow struck four different times.

But the identity of the killer that they finally settled on might surprise you.

The hypothesis that best fits the strange evidence found at Cerro Ballena is that these marine mammals might have ultimately been killed by a quirk of South America's geography.

[MUSIC PLAYING] Today, we'd probably call what happened at Cerro Ballena a mass stranding.

These happen when whales beach themselves on land and can't get back into the water, and the whales often die as a result.

In modern times, mass strandings usually involve toothed whales like pilot whales and sperm whales, not baleen whales like most of the ones found at Cerro Ballena.

But the few mass strandings of baleen whales that have happened in the last century or so can give us important clues about what happened millions of years ago.

For example, over a five week period from 1987 to 1988, 14 humpback whales washed up on the coast of Massachusetts.

Their bodies showed no signs of attack or trauma.

They included both males and females and one calf, and they died over a short period of time in a small geographic area.

But the killer at Cerro Ballena was even less discriminating, because it affected at least five other species of marine mammals along with the baleen whales.

And, like those unfortunate humpbacks, the marine mammals of the Miocene also probably died and floated ashore pretty quickly, maybe in as little as hours or days, washed there by ocean currents and lifted onto the tidal flat by storm surges.

The paleontologists that work on the site think that's what happened, because the whale skeletons were still pretty much intact.

That means marine predators didn't have time to feast on the dead or dying whales at sea, and land predators didn't scavenge the remains either.

Most of the animals were also found with their ventral side facing up-- literally belly up-- which suggests that they were dead when they washed ashore.

But knowing how the skeletons ended up there doesn't explain what killed all the marine mammals in the first place.

Even today, it could be hard to know what causes a mass stranding.

Many theories have been suggested, like changes in climate patterns, or the difficulty that whale sonar can have in picking up the shape of the sloping coastline.

The human use of sonar might be a factor too.

So the researchers working on Cerro Ballena explored four hypotheses to explain the fossil evidence they found there.

One idea was: could it have been something about the behavior of the animals themselves?

In general, it does look like there might be a connection between whale behavior and mass strandings.

Species that are more social and live in deeper water seem to strand more often than solitary species, or ones that live in coastal waters.

But even today, researchers aren't sure why that is.

And some populations of orcas and bottlenose dolphins will beach themselves on purpose to capture prey that's tried to escape them by leaving the water.

But this isn't something that baleen whales do.

And while most of the skeletons preserved at Cerro Ballena are whales, there are still other species, like seals, that should have been able to get back off the tidal flat and into the ocean.

This means the Miocene graveyard couldn't have been caused by the behavior of a single species.

So next hypothesis-- could the animals have been killed by-- and I'm sorry to use this word now-- a pandemic?

Well, again, there are several different species at the site, so the disease would have had to have been able to infect and kill them all.

It also would have had to do it various times over a period of 10,000 to 16,000 years, which the researchers think is pretty unlikely.

Then could it have been another kind of natural disaster?

Like what about a tsunami?

Well, it turns out the sediment that the animals were buried in looks like it was deposited continuously at a regular slow pace, not all at once.

So the researchers were left with one more hypothesis.

What if they were poisoned?

This actually happens all the time in events known as harmful algal blooms.

These blooms happen when colonies of certain kinds of algae suddenly explode in size, and these algae naturally produce toxic compounds, like neurotoxins, that infiltrate marine food webs.

Marine mammals might consume these toxins directly, or they might eat prey that have consumed them, and the toxins can also be inhaled.

And we know that harmful algal blooms can be deadly.

In fact, they were the culprit behind the mass strandings of humpback whales in Massachusetts in the late 1980s.

They've also been the cause of strandings and deaths for many other species, like sea lions, bottlenose dolphins, birds, and sea turtles, so there effects aren't specific to a particular species the way a disease might be.

So at Cerro Ballena, the scientists think that the animals found there probably swam through an algal bloom and breathed in the toxins or ate prey that had been poisoned by them.

Then they likely died pretty quickly and were washed onto that tidal flat where they were preserved.

This would explain why so many different kinds of mammals were found at the site.

The killer didn't have to affect a single species.

All of the victims were just unlucky enough to have been in the wrong place at the wrong time.

But how could this have happened repeatedly within a period of 10,000 to 16,000 years?

That is where geography comes in.

Along the coastline of the Atacama region, there's what's called an upwelling system in the Pacific Ocean.

Upwelling happens when wind pushes the surface water of the ocean away from the edge of a continent, and cold, nutrient-rich water from the deep ocean rises up to take its place.

This can make the area really biologically productive, with plankton thriving on the nutrients and forming the basis of marine food webs.

However, the upwelling water doesn't always have all the nutrients that these organisms need, and one that's usually in pretty short supply is iron.

But when a bunch of iron is dumped into the ocean system, it can wreak havoc on marine life by creating a harmful algal bloom.

This is what researchers think happened in the Miocene, and it all began with runoff flowing into the ocean from the Andes mountains.

A study from 2004 was able to piece together at least one of the mechanisms behind this iron-rich runoff by studying cores of marine sediment spanning the last 100,000 years.

It confirmed that the cores had alternating layers of iron-rich and iron-poor sediment, and this pattern roughly coincided with the astronomical cycle known as procession, the change in the tilt of our planet's axis over timescales of about 20,000 years.

This cycle can cause lots of changes to climate all over the world.

In this case, it probably shifted the location of climate zones in Chile and influenced the amount of precipitation that fell on parts of the Andes mountains.

Today, most of the precipitation there happens in the winter as snow, and in the summer, that snow melts into rivers that run into the ocean.

But in the past, when Earth was tilted at a slightly different angle, this region probably got less direct sunlight in the winter and received even more snow.

If this cold winter was followed by an extra-hot summer, it would lead to even more runoff, and more runoff meant more sediment being eroded off the iron-rich rocks of the Andes and carried into the ocean.

And those times when there was a lot of iron washing into the ocean provided a big boost of nutrients for plankton-- both the good ones, like phytoplankton, but also the bad ones that produce toxins, like harmful algae.

While that study of sediment cores only looked at this cycle over the past 100,000 years, it was probably happening back in the Miocene Epoch too.

And we know from studying climate in the present that it's a complicated system.

So other factors, like that pattern of warming ocean currents known as El Niño, probably also played a role in why the strandings happened four times.

So that's the process that paleontologists think created the strange evidence we find at Cerro Ballena.

The effects of procession and other variables changed the local climate, which created spikes of iron-rich runoff, which in turn promoted the growth of harmful algal blooms at least four times off the South American coast.

And groups of unlucky marine mammals swam through the toxic blooms, died, and then floated toward the tidal flat, where storm surges lifted them onto the sand.

They eventually were buried and fossilized, preserving four tragic moments in deep time.

Now we'll probably never know for sure exactly what happened at Cerro Ballena, but the hypothesis that the researchers constructed does a good job of explaining the observed evidence by using the present as a tool for understanding the past.

And it stands today as a stark reminder of how tightly our planet's systems are connected, from big baleen whales to tiny toxic algae to the oceans they all live in, and even the far-off mountains that feed the seas.

[MUSIC PLAYING]