Science1 distinct publisher3 min readPublished
A team at Nanjing University read sulfur and molybdenum ratios in Conotubus, a 550-million-year-old tube from southern China, and inferred bacterial partners feeding on hydrogen sulfide: a metabolic signature, not a preserved body.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
The interpretive weight here rests on modern analogues. Microbes generate energy for corals and deep-sea tube worms today, and break down cellulose and lignin inside animals as different as cows and termites [11]. Conotubus is a tube [2]. What the Nanjing University team measured, though, is not bacteria: it is sulfur and molybdenum isotopes in remarkably well preserved fossils, read against the isotopic signatures that organisms and environments produce now [4]. Isotope ratios record which elements are present and how biological and other processes have shifted their proportions [15], and the ratios in Conotubus are consistent with bacteria that used hydrogen sulfide as an energy source [5].
That is a reasonable chain, and it is worth naming which link carries the load. Even exceptionally preserved fossils almost never retain the microbes that lived inside an animal [9], so a standard that demands preserved cells will never date symbiosis at all. Geochemistry gets around that preservation bias. What it does not tell you is where the bacteria sat. Cells within tissue, a lining in a body cavity, and a film on the outside of the tube all run the same metabolism, and on my reading the data support an ecological association more firmly than anatomical intimacy.
The dating claim is arithmetic worth doing out loud. The fossil is 550 million years old [1], and the previous earliest documented evidence of symbiosis came from rocks 100 million years younger [3], which puts the old record near 450 million years [16]. The Ediacaran closes at roughly 541 million years ago [6], so Conotubus sits about 9 million years inside it [17], in the interval whose sediments carry the first large animal fossils visible without a lens [8]. Oxygen was much lower than today, and in some marine settings hydrogen sulfide accumulated to concentrations highly toxic to animals [7]. The authors' reading is that a partnership with sulfur-oxidizing bacteria turned that hostility into an opening, supplying nutrition while helping the animal tolerate the water [12].
Two cautions on how far to carry it. The account I am working from was written by one of the study's co-authors [14], and it gives no sample count and no uncertainty on the isotope measurements; the paper in the Proceedings of the National Academy of Sciences is where those belong [1]. And Ediacaran fossils of this kind are rare to begin with [10], so the sample is unlikely to be large. One taxon from one region can move the earliest known date without establishing that host-microbe coupling was general among the first animals, however tidy the line looks from Conotubus to the observation that about half the cells in a human body are microbial [13].
Ranked by verification strength, evidence, and original report placement.
The results suggest Conotubus lived in partnership with bacteria that could use hydrogen sulfide as an energy source.
Research led by Zhenfei Wang and Yongbo Peng of Nanjing University found evidence of close animal-microbe relationships in a 550-million-year-old fossil from southern China; the work is published in the Proceedings of the National Academy of Sciences.
The study focused on Conotubus, a tubular fossil that lived during the Ediacaran period.
Previously, the earliest evidence of symbiosis was documented in fossils 100 million years younger than this one.
The team analyzed sulfur and molybdenum isotopes in remarkably well-preserved fossils from southern China, and interpreted them by comparing the ancient signatures with those produced by organisms and environments today.
The Ediacaran period extends from approximately 635 million to 541 million years ago.
Distinct publishers with included, body-backed reporting in this cluster.
phys.org
1 article · September 2, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
science
A PNAS study says X's feed learns from your arguments, not your likes2 distinct publishers
science
Whale-watch footage catches 1,417 humpbacks feeding along a corridor theory calls a fast1 distinct publisher
science
Sex has at least five axes. Eligibility rules and reference ranges have one field.2 distinct publishers
science
Police AI reached the evidence file before anyone measured how often it fabricates1 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 peer-reviewed paper, told by its own author
Underneath the coverage sits a real PNAS paper with a DOI, and the geochemical reasoning is described clearly enough to argue with. But the retelling comes from a co-author, and the numbers a skeptic would want are absent: how many tubes were measured, what the isotope values were, how far they sit from a plausible sediment-only signal. The finding is an inference from chemistry, which the authors say themselves, and the peer review is the only external check anywhere in view.
No independent uptake on record
Nothing in this reporting shows the result being used, cited, challenged or replicated by anyone outside the research team. A paper published this year has had no time to accumulate that, and inventing a figure from its absence would be worse than leaving it blank.
Headline 'reveals', body says 'suggest'
The text is more careful than the packaging around it. Inside, results 'suggest' a partnership and the ecological reading is offered as something that 'may' have happened; at the top, a fossil 'reveals' animal-microbe partnerships. The overreach is modest and mostly lives in the framing, but the 100-million-year record shift is presented as established when it is a single, uncited assertion by the people making it.
The author is on the paper, and says so
This is a scientist writing up his own result, which is what The Conversation exists to publish and which phys.org passes along intact. The disclosure is voluntary and unambiguous, and there is no commercial or funding interest visible. Still, the person deciding which caveats survive the translation into plain English is the same person whose paper is being described, and priority claims are exactly where that pressure bites.
Single account, one voice, real paper
Confidence here is limited less by sloppiness than by narrowness. The descriptive facts, dates, fossil, isotope systems, are as reliable as their source, and that source is the study team. What we cannot yet judge is whether other geochemists accept that these sulfur and molybdenum ratios demand a symbiotic host, and until someone outside Nanjing says so on the record, our reading stays provisional.