Science1 distinct publisher2 min readPublished
Radiodonts sat near the top of the Cambrian food web, and one of them appears to have been bitten and survived. Whether that mark is a bite or a tear from moulting rests on one fossil flap of a kind that hardly ever survives.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
The useful work here is subtractive. Two tests stand between a notch in a soft fossil and a bite. The first asks whether the gap is original: X-ray fluorescence mapping found the chemical composition inside the notch consistent with the rest of the flap, which is not what happens when a fossil flakes apart after burial and exposes a fresh surface [10]. The second asks when the damage happened. A black band runs around the edges of the shape, and modern arthropods lay down exactly that darkening around tissue they are regrowing, which is why Pates places the injury in the animal's lifetime [11].
The alternative that survives both tests is moulting. Arthropods injure themselves routinely while shedding an exoskeleton, and Moysiuk's point is that a flap damaged during a moult would also present as a wound [18]. Pates answers with geometry rather than chemistry: a moult tear is usually a single rip and can land anywhere on the body, whereas this wound has two points and sits on the part of the flap that an attacker behind a fleeing animal would reach [19].
Then the denominator. Pates counts only about ten isolated radiodont flaps ever recovered [12], which makes the injured one roughly a tenth of the world's supply of this material [23]. Set that against trilobites, which carry hundreds of probable predation wounds [14]. Trilobites had hard exoskeletons and their record continues to 250 million years ago, about 230 million years past the last radiodonts [15][22], and they were small animals nearer the bottom of the food chain [16]. The distance between hundreds and one measures preservation and group longevity much more than it measures how dangerous Cambrian water was for large predators.
The thing this doesn't tell you is who did the biting. Pates allows another radiodont species, or a member of Amplectobelua symbrachiata itself [13]. The victim's size is not settled either. If the fossil is that animal's largest flap, the individual was at least 23 centimetres long, and because the flaps shrink toward the tail, a smaller flap implies a longer animal [7]. Twenty-three centimetres is under a quarter of the near-metre maximum known for the species [24], which leaves room for the bitten radiodont to have been a juvenile rather than a full-grown adult [4].
Pates says he would never have expected to see something like this preserved in the fossil record [21], and that surprise is the right register for the claim on offer: one large Cambrian animal was wounded by something with two points, and lived on afterwards [1].
Ranked by verification strength, evidence, and original report placement.
Radiodont swimming flaps were softer than their grasping appendages, resulting in very few fossils of them; a team recently found a single flap at the Yu'anshan Formation in China, a site where soft tissues are exceptionally preserved.
Pates says only around 10 isolated radiodont flaps like this have ever been found, and that with more complete specimens you would not necessarily be able to see such wounds even if they were there.
Pates says: "I would never have expected to see something like this preserved in the fossil record."
The researchers in China contacted Stephen Pates at University College London because he has studied similar injuries in trilobites; Pates says you want to rule out that the feature was broken during the fossilisation process.
X-ray fluorescence mapping shows the chemical composition of the W-shaped part is consistent with the rest of the fossil, meaning it is not an artefact caused by flaking; Pates says this tells us it is likely a flap preserved as it is rather than altered later.
Joe Moysiuk at the Manitoba Museum in Canada says there have been other reports of healed injuries in Cambrian animals, though never in a radiodont, and that with only one specimen he does not think other potential causes can be ruled out.
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 28, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
science
Physicists attach a target precision to each of 132 ways matter could break Lorentz symmetry1 distinct publisher
science
Sleep debt gets a wiring diagram: mouse neurons that add and erase recovery sleep2 distinct publishers
product
After Arup, a face on a video call is not a credential1 distinct publisher
science
August 2027's eclipse delivers 84 per cent of the longest totality the orbits allow1 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One flap, one outlet, no paper to read
The physical case is a single isolated appendage from the Yu'anshan Formation, and the analytical support, X-ray fluorescence showing the notch is chemically continuous with the slab, reaches us only as Pates describing it to New Scientist. No publication, figure, or measurement is named anywhere in our coverage. What keeps this well above thin is the specificity of the observations and the fact that the report carries its own dissent rather than smoothing it away.
No uptake to observe yet
Nothing in this reporting tells us whether the interpretation has been formally published, cited, or accepted by other workers on Cambrian arthropods. The only outside voice quoted is a sceptical one. We would rather record nothing than invent a trajectory for a claim that is a few days old in public.
Headline decides what the body leaves open
New Scientist's headline announces a healed fang wound as found; nine paragraphs later Pates is saying you cannot rule out moulting and that this is a balance of probabilities. The gap is a framing gap, not a fabrication, since the caveats are all present and Moysiuk gets to state them. Still, a reader who stops at the headline has a firmer fact than the reporting supports.
The vouching expert is on the team; the sceptic is not
Worth noticing who is speaking. Pates was brought in by the Chinese researchers because of his trilobite injury work, so the specialist most quoted in defence of the bite reading is party to the interpretation, and a first-ever healed wound in a radiodont is a striking thing to have your name on. The genuinely detached voice, Moysiuk at the Manitoba Museum, is the one raising moulting. New Scientist's own pull towards a vivid Cambrian predator headline sits on top of that, but the piece does not hide the disagreement.
Firm on the fossil, soft on the fangs
We are reasonably sure of what exists: a radiodont flap with a W-shaped notch, chemically continuous, dark-rimmed. We are much less sure of what caused it, and with one publisher and one specimen there is no second line of sight on either. Confidence lands mid-range and would move quickly on a named paper or a second wounded flap.