The Repository Deep Time Archives

Oman's Sponge Molecule Survived. Its Date Did Not.

October 8, 2026 · Runtime 10:18 · Watch on YouTube

Two and a half kilometres under the desert of Oman, a mudstone from before the end of the last global ice age held a molecule announced in 2009 as the oldest trace of animal life. Three attacks later, the molecule is still standing. The date is not — and the scientists who made the claim have moved it themselves.

Transcript

A mudstone under Oman 0:00

Two thousand four hundred and twenty metres under the desert of Oman, an oil well called GM-1 cut through a grey lime mudstone. It was laid down before the last global ice age had ended. In 2009, a molecule inside it was announced as the oldest trace of animal life on Earth. Sixteen years later, even the scientists who announced it no longer put that date on their timeline.

Three attacks, one molecule 0:26

The molecule is called twenty-four-isopropylcholestane. Its parent compound is made in abundance by one group of living animals: sponges. By the end, you will know which part of the 2009 claim has survived three separate attacks, and which part its own defenders have now set aside.

Cores drilled for oil 0:46

The rock comes from the South Oman Salt Basin, one of the best-preserved sequences of its age anywhere. It lies one to five kilometres down, under salt, and the only reason anyone has touched it is oil. Petroleum Development Oman drilled the cores. In 2009, a team led by Gordon Love and Roger Summons analysed sixty-four of its source rocks.

Two freezes, then the sea 1:11

Those rocks are the Huqf Supergroup: about a hundred and seventy million years of sediment, from roughly seven hundred and ten to five hundred and forty million years ago. Near the bottom are the deposits of two global freezes, dated in Oman to about seven hundred and thirteen million years, and to younger than six hundred and forty-five. After the second one, the sea came back.

Well GM-1, 2,420 metres 1:34

The oldest sample in the study sits below that second freeze. In well GM-1, the mudstone lies about a hundred and fifty metres beneath the glacial deposit, and beneath the cap carbonate that marks the end of the ice. So the team read it as younger than seven hundred and thirteen million years, and older than about six hundred and thirty-five. One zircon grain inside it sets a hard limit: no older than seven hundred and fifty-one.

Forked or straight 2:02

Here is what they were measuring. Two steranes with thirty carbon atoms, identical except for a three-carbon branch on the side chain. In one, the branch forks: isopropyl. In the other, it runs straight: n-propyl. Sponges make the forked kind as a major sterol. A small class of algae, the pelagophytes, make the straight kind.

99 percent of a membrane 2:27

Living demosponges still produce the forked version, across three of the main branches of their family tree. In one genus, Pseudaxinyssa, it and related sterols make up ninety-nine percent of everything in the membrane. The pelagophyte algae make the forked kind only as a trace: under half a percent of their sterols, when it shows up at all.

The 2009 ratio 2:49

So the test is a ratio: forked over straight. In ordinary marine rocks and oils of any age, it sits around two or three tenths. In living pelagophytes, it is one hundredth or less. In the Huqf rocks, across eighty-one analyses, it ran from about half to sixteen, and averaged one and a half. In the GM-1 extract itself, it was three point two six.

Two tests against contamination 3:16

The obvious worry is contamination: drilling fluid, migrated oil, younger rock. The team had two answers. The molecule was also locked inside kerogen, the insoluble organic matter that forms in the sediment soon after burial, and released only by hydrogen pyrolysis. And where drilling lubricant had leaked into a sample, the ratio went down, not up.

Sixty million years early 3:41

That gave the claim its weight. In February 2009, in Nature, the team concluded that demosponges were living before the end of the Marinoan glaciation, more than six hundred and thirty-five million years ago. The oldest animal-like body fossils are about five hundred and seventy-one million years old. The molecule put animals more than sixty million years earlier.

A second sponge marker 4:05

In 2018, the same circle of scientists added a second molecule: twenty-six-methylstigmastane, with an extra carbon at the very tip of the side chain, from a sterol then known only in living demosponges. It turned up alongside the first, in the same Neoproterozoic rocks and oils. Two sponge markers, one sequence. The case looked closed.

Rhizaria make it too 4:30

It stayed closed for less than five months. In March 2019, Benjamin Nettersheim, Jochen Brocks, Christian Hallmann and colleagues reported that the precursors of both molecules are made by single-celled organisms: early-branching Rhizaria, predatory protists common in the modern ocean. Neither marker, they argued, could be read as a sponge anymore.

Could the rock make it? 4:54

Then a second line of attack, led by Ilya Bobrovskiy, the chemist whose method, three years earlier, had made Dickinsonia an animal. The question was different: not who else makes the molecule, but whether the rock itself could make it. Green algae dominated the oceans of that time, and their main sterols have twenty-nine carbons. Add one carbon in the ground, and you have thirty.

Heating sea lettuce 5:19

So they heated algal sterols. In one run, the sterol fraction of sea lettuce, a common green alga, was pyrolysed with no mineral catalyst, and thirty-carbon steranes appeared. Of the two propyl forms, sixty-six percent came out forked: about two forked for every straight one, above the Huqf average, made from an alga. The amount was tiny, under a third of a percent of all steranes.

The heat did it 5:47

The control matters. When the same algal sterols were only hydrogenated, without the heating step, the thirty-carbon steranes were below detection, or orders of magnitude rarer. The heat did it. The authors are careful about what that proves: a lab run of seconds to weeks cannot reproduce half a billion years of burial. It shows a plausible route, not a measured one.

The raw material was in the rock 6:12

And the raw material was already in the rock. The 2009 supplement itself notes that the GM-1 mudstone is the earliest marine sediment found with twenty-nine-carbon steranes making up more than half of all steranes. In its extract, seventy-two percent. Green algae dominated that sea, and the ingredient for the rival explanation sat in the same sample as the claim.

Not diagnostic, 2021 6:37

In 2021, Bobrovskiy's team concluded that twenty-four-isopropylcholestane is not diagnostic for sponges, and that it probably formed in the sediment from the sterols of green algae. Their title said it plainly: the finding negates the oldest evidence for animals in the rock record. It sent that evidence back to the latest Ediacaran, the age of Dickinsonia.

Thirty-one carbons, 2025 7:01

The other side did not concede. In autumn 2025, in PNAS, Lubna Shawar, Gordon Love, Roger Summons and colleagues reported a third kind of molecule: thirty-one carbons, in two forms, n-butyl and sec-butyl. Living demosponges make the sterols behind both. José-Luis Giner synthesised eight candidate versions, so each fossil peak could be matched against a known standard.

Pattern, not presence 7:29

Their case against geological methylation is about pattern, not presence. Random chemistry in a sediment should make a messy crowd of side-chain variants. The rocks show two dominant thirty-one-carbon forms, paired with the forked thirty-carbon one, the way living demosponges make them. And thirty-carbon steranes do not always appear where the algal twenty-nine-carbon ones do.

Sample OMR 027 7:54

The clearest match came from sample OMR 027, in the Masirah Bay Formation of Oman: the best-preserved rock in the set, with a forked-to-straight ratio as high as twenty-one. Masirah Bay is early Ediacaran. It sits above the cap carbonate, after the ice, not below it. The pattern also turned up in rocks and oils from eastern Siberia and western India.

The date is set aside 8:19

And here is the part the headline never carried. In that same paper, two of whose authors made the 2009 claim, the GM-1 sample is described as previously interpreted as Cryogenian, and as requiring verification. A gap in the Oman sequence may span tens of millions of years. To be conservative, they write, their timeline begins in the early Ediacaran.

Measured, argued, moved 8:45

So what can the rock actually say? The molecules are measured: the Huqf numbers are not in dispute, and the contamination tests held. The source is argued: a sponge pattern against algal chemistry, and the field is still split. The age has moved: from before the last global freeze to after it, on both sides of the argument.

No body this time 9:08

Last time, a molecule made Dickinsonia an animal, and it held because it sat inside a body, tested against a named rival. Here there is no body. The molecule has to carry the whole identification alone, and the same chemist stands on opposite sides of the two cases. The method is not the verdict. What it is tested against is.

The molecule survived 9:31

The molecule survived. The date that made it famous has not, at least for now. Sponges may well have lived in early Ediacaran seas; the 2025 chemistry argues they did. But if they did, they left almost nothing else behind. The oldest fossils that are diagnostically sponge, their mineral spicules, are only about five hundred and thirty-five million years old.

Next: a glass sponge with no glass 9:57

Next time, the fossil that may explain that silence: a sponge from the Yangtze Gorges, about five hundred and fifty million years old, described in 2024 as a glass sponge with no glass. If its skeleton really was organic, the missing spicules are not missing evidence. They were never made.

Description and sources

24-isopropylcholestane, a 30-carbon sterane whose parent sterol living demosponges make in abundance. Love et al. (2009, Nature) measured it in 64 source rocks of the Huqf Supergroup, South Oman Salt Basin: across 81 analyses the forked-to-straight ratio (24-ipc/24-npc) ran 0.52-16.1, average 1.51, against 0.2-0.3 in marine rocks and oils of any age and 0.00-0.01 in pelagophyte algae. The oldest sample, a lime mudstone from well GM-1 at 2,420 m, about 150 m below the Marinoan glacial deposit, was read as 713-635 Ma. Kerogen-bound molecules and a contaminated sample whose ratio fell, not rose, argued against contamination. Zumberge et al. (2018) added 26-methylstigmastane. Nettersheim et al. (2019) found the precursors of both in unicellular Rhizaria. Bobrovskiy et al. (2021) generated 30-carbon steranes by heating green-algal sterols (sea lettuce: 0.28% C30, 66% of propyl forms forked) and concluded the marker is not diagnostic — and the 2009 supplement itself records GM-1 as the earliest marine rock dominated by algal 29-carbon steranes (72%). Shawar et al. (2025, PNAS) answered with two 31-carbon steranes paired with 24-ipc as in living demosponges — and wrote that GM-1's Cryogenian age 'requires verification', starting their timeline in the early Ediacaran.

PRINT-READY, FROM THIS CHANNEL

The Deep Time Record, Vol. 01 — the cores, the curves and the clocks

https://therepository.gumroad.com/l/deep-time-record

PRIMARY SOURCES

Love, Grosjean, Stalvies, Fike, Grotzinger, Bradley, Kelly, Bhatia, Meredith, Snape, Bowring, Condon & Summons 2009. Nature 457:718. https://doi.org/10.1038/nature07673
Zumberge, Love, Cárdenas, Sperling, Gunasekera, Rohrssen, Grosjean, Grotzinger & Summons 2018. Nat. Ecol. Evol. 2:1709. https://doi.org/10.1038/s41559-018-0676-2
Nettersheim, Brocks, Schwelm et al. 2019. Nat. Ecol. Evol. 3:577. https://doi.org/10.1038/s41559-019-0806-5
Bobrovskiy, Hope, Nettersheim, Volkman, Hallmann & Brocks 2021. Nat. Ecol. Evol. 5:165. https://doi.org/10.1038/s41559-020-01334-7
Shawar, Love, Uveges, Zumberge, Cárdenas, Giner & Summons 2025. PNAS 122:e2503009122. https://doi.org/10.1073/pnas.2503009122
Wang, Liu, Chen, Wu, Liu, Wan, Pang, Zhou, Yuan & Xiao 2024. Nature 630:905. https://doi.org/10.1038/s41586-024-07520-y
Wagner & Kelley 2016. Marine Biodiversity. https://doi.org/10.1007/s12526-016-0508-z

ARCHIVE IMAGES — author, date, archive and licence.

Bengt Littorin ( Havspappan ) — 1992 | Wikimedia Commons | CC BY 2.0
JPL — 1999-04-15 | NASA / JPL | Public Domain (NASA)
K. Rasmussen — 1984-08 | Wikimedia Commons | Public domain
Ernst Haeckel — 1904 | Wikimedia Commons | Public domain

... and 4 more, each credited on screen in the video and listed in this project's manifest.

The 2 generated plates were made for this channel with FLUX.1-schnell. No person is depicted in any of them.

Deep Time Archives rebuilds the pre-human Earth from the physical record — outcrops, drill cores, golden spikes, catalogued specimens. Every episode starts from an object you can go and check, and the sources are listed above.

About the images. No image of the Oman cores is openly licensed: the basin appears from orbit (NASA/JPL radar of central Oman; NASA/JSC photograph of dunes on the Saudi Arabia–Oman border). Every organism shown is modern: a living demosponge (Halichondria panicea), demosponges with coralline algae, sea lettuce (Ulva lactuca, the alga used in the 2021 experiment), radiolarians drawn by Ernst Haeckel in 1904 (one branch of Rhizaria, not the species tested in 2019), and, in the closing seconds, a modern glass sponge (NOAA Ocean Exploration) as a preview of the next episode. The ice front and the returning sea are AI-generated reconstructions of places; no organism is generated. This video is labelled as containing altered or synthetic content. The diagrams are drawn in code from the published numbers.

#DeepTime #EarthHistory #Ediacaran #Paleontology #Evolution