Animals may have evolved up to 200 million years before they appear clearly in the fossil record, according to a study led by the University of Oxford, which places their origin between 800 and 700 million years ago.
The findings, published on 2 October in Science Advances, come from an international team that also included scientists from the University of California, Berkeley, ETH Zurich and Yale University. They would move the emergence of animal life deep into the Neoproterozoic Era, possibly before some of the most severe ice ages in Earth's history.
The puzzle the work addresses is a long-standing one. Animals seem to appear relatively abruptly in the fossil record shortly before the Cambrian Period, which ran from 539 to 487 million years ago, even though other lines of evidence have suggested their evolutionary story began much earlier.
At the centre of the study is an assumption about how old fossil deposits are used to set a maximum age. One important site is the Weng'an Biota in China, a roughly 590-million-year-old deposit of Ediacaran age that contains microscopic organisms preserved in exceptional detail, none of which can be definitively identified as an animal. Because of that absence, scientists had argued that animals likely evolved after Weng'an formed: if animals had already existed, so unusually well-preserved a deposit should have captured evidence of them.
The team tested the assumption against a second site. The Kheseen Biota in Mongolia is more than 40 million years younger than Weng'an and shares some species with it. Critically, animals are already known from other locations around the world by the time Kheseen formed, including sites in Namibia and South China.
Researchers examined more than 140 samples, including material collected from previously undocumented locations, using scanning electron microscopy. They found exquisitely preserved new microfossil species, among them acritarchs, tiny spherical organisms with spines and branching projections, along with embryo-like fossils containing internal cells. Despite the remarkable preservation, none of the fossils can confidently be identified as an animal.
"The Kheseen Biota breaks the argument that the exceptional microfossils of Weng'an mean we would have seen animal fossils in the assemblage had they existed at the time," said the senior author, Associate Professor Ross Anderson of the Museum of Natural History at Oxford University. "The Kheseen microfossils are just as well-preserved, yet animals continue to be absent - despite the fact we know at that point they existed."
The likeliest explanations are that animals occupied environments different from those represented in the Kheseen deposits, or that the chemical conditions responsible for fossilisation there were not suitable for preserving animal remains.
With the Weng'an constraint removed, the team turned to much older deposits dating from about 850 to 730 million years ago: the Svanbergfjellet Formation in Norway, the Bitter Springs Group in Australia and the Chuar Group in Arizona. All are fossil-rich and considered capable of preserving animals, although no confirmed animal fossils have yet been found in them. Using those older geological constraints in a molecular clock analysis moved the estimated origin of animals backward by about 200 million years.
Other evidence points the same way. Chemical fossils, or biomarkers, preserved in ancient rocks contain evidence consistent with sponges living at least 650 million years ago.
The reasoning here is sound, and it is also, stated plainly, remarkable. A deposit containing no animals was used to prove that animals did not yet exist. A second deposit containing no animals, from a period when animals demonstrably did exist, has now cancelled the proof.
The quantity doing the work in both cases is the same quantity: nothing. What changed is its standing. The nothing at Kheseen now outweighs the nothing at Weng'an, because the nothing at Kheseen is known to be wrong.
That gives the method a direction. Every superbly preserved deposit that fails to contain an animal makes animals older, because failure to preserve is no longer evidence of absence and the constraint moves to the next bed down. The study duly moves it to beds of 850 to 730 million years, which also contain no confirmed animals.
Put in the discipline's own terms, the maximum age of animal life is now set by the oldest rock in which animals have most convincingly failed to turn up. On that principle the figure can travel in one direction only, and each new empty deposit of excellent quality will carry it further.
The Department of Unnecessary Research has extended the method to its limit and reports the result without comment: the oldest animal on Earth is the one that has never been found anywhere.