Mongolian Microfossils Push Animal Origins Back 200 Million Years

Animals may have originated between 800 and 700 million years ago, roughly 200 million years earlier than the leading fossil-calibrated estimate, according to a study published in Science Advances on 2 October 2026. The argument rests not on a new animal fossil but on the continued absence of one in a deposit where animals ought to have turned up.

The paper, led by Orin M. Lole Durbin with Ross P. Anderson of the Oxford University Museum of Natural History as senior author, examined phosphatised microfossils from the Kheseen Formation on the western shore of Lake Khuvsgul in northern Mongolia. The team studied more than 140 samples using scanning electron microscopy, according to a University of Oxford statement issued the same day, and raised the number of acritarch species known from the Kheseen Biota to 16, eight of them new. Co-authors came from the University of California, Berkeley, ETH Zurich and Yale.

The problem with dating an absence

Molecular clocks estimate when lineages diverged by comparing genetic differences between living species and anchoring the result to dates from the rock record. A minimum age is straightforward: animal body fossils in the Ediacara Macrobiota reach back to about 574 million years ago. A maximum age is harder, because it must be inferred from a deposit good enough to have preserved animals that nonetheless contains none. The paper identifies the phosphatised Weng'an Biota of China, dated in the study to 590.4 million years ago plus or minus 2.6 million, as the calibration doing most of that work in recent analyses.

The Kheseen Biota is the test case. The authors date its fossil-bearing strata to between 534 and 526 million years ago, and no older than 550 million, which places it well inside the window when animals are documented elsewhere. Preservation is comparable to Weng'an and the two assemblages share species. Animals are still missing.

With Weng'an set aside, the authors fall back on older deposits that combine fossil richness with chemical conditions favourable to preserving soft tissue: the Svanbergfjellet Formation in Svalbard, the Bitter Springs Group in Australia and the Chuar Group in Arizona, spanning the late Tonian from about 850 to 717 million years ago. Recalibrated on those, the clock returns an origin for crown Metazoa of roughly 800 to 700 million years ago. That would put the first animals before, or possibly during, the Cryogenian glaciations that began around 720 million years ago and are known as Snowball Earth.

Reconstructing a planet's deep past from a patchy rock record is a recurring difficulty, as Parallax Nexus reported when NASA described three separate episodes of water in the margin rocks of Jezero Crater on Mars.

What the authors do not claim

Durbin was direct about the limits of the result.

The strongest check on the paper is that it replaces one argument from absence with another. The late Tonian deposits now carrying the calibration contain no animal fossils either, and the confidence that they would have preserved them rests on reasoning about burial conditions rather than a positive find. Corroboration is thinner than the headline figure implies: the paper notes that lipid biomarkers place animals in the final 10 to 20 million years of the Cryogenian, which ended 635 million years ago, still well short of 800 million. The honest reading is that this is a well-argued demolition of a calibration rather than a new birth date for animal life.

Anderson's team argues the search should now move to pre-Ediacaran deposits representing different environments and modes of fossilisation, weighing body fossils, traces of activity and chemical biomarkers together rather than any one line of evidence.