The Water That Held Its Shape

The Ediacara biota should not exist in the fossil record.

They had no bones, no shells, no mineralized parts of any kind. They were soft -- some of them apparently little more than quilted bags of tissue lying flat on the seafloor. Dickinsonia, the most famous, looked like a ribbed oval doormat up to four feet across, with a faint ridge down the center and no mouth, no gut, no evidence of internal organs. Charnia resembled a fern frond rooted to the deep ocean floor, swaying in water too dark for photosynthesis, bearing no relation to any fern or coral or anything alive today.

They lived between 575 and 539 million years ago -- the Ediacaran period, the last chapter before the Cambrian explosion rewrote the rules. More than 200 species. Found on every continent except Antarctica. Discs, tubes, fronds, mud-filled bags, quilted mattresses. Adolf Seilacher called them "strange as life on another planet, but easier to reach."

Soft things don't fossilize. That's the rule. Jellyfish wash up and vanish. Worms decompose in hours. The fossil record is an archive of hard parts -- teeth, bone, shell -- because those are what resist the indifference of deep time. Sand is the worst medium: porous, coarse, storm-churned. Anything soft buried in sand should dissolve into chemistry and leave nothing behind.

The Ediacara biota fossilized in sandstone. In extraordinary detail.

For decades, the explanation was that the organisms themselves must have been different -- tougher than they looked, chemically resistant, armored in some invisible way. The alternative was that microbial mats on the seafloor acted as a sealing layer, entombing the bodies before decay could finish its work. Both ideas were plausible. Neither was proven.

In January 2026, Lidya Tarhan's team at Yale published the answer in Geology, and it had nothing to do with the organisms at all.

It was the water.

Tarhan's team measured lithium isotopes in Ediacaran fossils from Newfoundland and northwest Canada. Lithium isotopes can distinguish between detrital clays -- particles washed off the continents -- and authigenic clays, which precipitate directly from seawater. What they found was a two-step process: detrital clay particles were already present in the burial sediment, and these served as nucleation sites for authigenic clays to crystallize out of the surrounding seawater. The Ediacaran ocean was unusually rich in dissolved silica and iron. When these elements met the detrital clay seeds in the sediment, they formed a mineral cement that locked the sand grains together around whatever happened to be buried there.

The ocean itself was the preservative. Not the organisms. Not the mats. The water chemistry of the Ediacaran period created conditions for a kind of fossilization that has no modern analog. The seawater held the shapes of things that the sand alone would have erased.

This reframes the entire Ediacaran record. If the preservation was environmental, not biological, then the organisms didn't need to be unusual in their composition -- they just needed to die in the right water. And it raises Tarhan's most unsettling question: did the Ediacara biota actually go extinct at 539 million years ago, or did the ocean chemistry change? Did they vanish, or did the world simply stop being able to remember them?

The end-Ediacaran extinction is one of the major mass extinction events, conventionally understood as a biological crisis that cleared the stage for Cambrian animals. But if the preservation window closed because the ocean's chemistry shifted -- silica and iron concentrations dropping, authigenic clay formation declining -- then the fossil record wouldn't just fail to preserve new Ediacaran organisms. It would retroactively erase the evidence that they were ever there.

An absence of fossils is not evidence of absence. That's the first thing paleontology teaches. But it's one thing to know it intellectually and another to hold the possibility that an entire kingdom of life -- 200 species on six continents, 36 million years of existence -- might have kept going, unrecorded, because the water forgot how to hold their shape.

Tarhan plans to extend the lithium isotope approach to fossil beds from other periods, testing whether similar preservation mechanisms operated elsewhere. The method could reveal other windows -- times when the ocean was an unusually good archivist, and times when it was an unusually poor one. The fossil record may be less a chronicle of what lived and more a chronicle of what conditions allowed to be remembered.

The Ediacara biota were quilted and frondose and strange. They absorbed nutrients through their skin. Some of them moved, leaving faint trails. Some of them grew to four feet across on nothing but osmosis. They existed for 36 million years in a world that hadn't invented predation yet.

And they might still be waiting, in some gap in the record, for the right chemistry to notice them.