What tardigrades actually survive
We went back to the sources behind the famous numbers and checked each one: which species, which life stage, how long, and how many survived. Dry tardigrades (tuns) survive extraordinary doses of heat, pressure, radiation and vacuum, for minutes to days. Active ones are ordinary small animals, and a hot afternoon can kill them.
What would a tardigrade need to live somewhere else? — not what it survives, what it needs to be awake and eating. The list collapses from three requirements to three completely different ones, and the last step is a loop.
The limits, stressor by stressor
Each dot is one measurement we checked in its source. Colour gives the animal's state during the exposure, and shape its life stage. Switch between the highest exposures animals survived and the exposures that killed half or all of them. Pick a dot for species, conditions and source.
The grey band on the radiation row is the human LD50 at 30 days, 2.5–4.5 Gy (Jönsson et al. 2018, Table 12.1, after Bolus 2001). Survival checks ran from 1 day to a few weeks after exposure; each dot gives the check time where the source does.
Show the measurements as a table
| Stressor | Value | Species, state | Exposure and result | Source |
|---|
How this is built
- Data
- One measurement per dot, each read in its source: open-access papers, the three survival chapters of Water Bears (Schill 2018), and abstracts where that is all we could read. The human radiation band is from Jönsson et al. 2018, Table 12.1.
- Built with
- We recorded the values as claims in our knowledge graph, each with its page, and as rows in the page's own table, which the script draws as SVG on log or linear scales. Each dot carries its species, state, life stage, exposure and source.
- Checked
- Each value by hand against its source. Values we found only through a review are drawn as "reported, not verified".
- Limits
- Exposure times and the delay before survival was scored differ between studies, so dots in one row are not strictly comparable.
- More
- How the whole project is built · the claims in the graph
Heat: time and state decide
We redrew the heat experiments on Ramazzottius varieornatus from the per-replicate counts Neves and colleagues deposited: about 100 animals per temperature, scored 48 hours after the heat (after rehydration, for tuns). Our logistic fits to their counts give the same 50% mortality temperatures as the papers (37.0, 81.8, 63.0 and 56.0 °C, against 37.1, 82.7, 63.1 and 56 °C).
The same species that shrugs off 80 °C for an hour as a tun loses half its active animals at 37 °C, and a week of dry storage at 56 °C does what an hour at 82 °C does. Heat tolerance in the dry state has a clock: Hengherr et al. 2009 link the ceiling to the glass the dry cytoplasm forms, which melts at about 80 °C in Paramacrobiotus and Macrobiotus.
How this is built
- Data
- Per-replicate activity counts from the GitHub repository deposited with Neves et al. 2020 (no licence stated), about 100 animals per temperature, scored 48 hours after the heat.
- Built with
- We averaged the replicates at each temperature and drew the curves as SVG; lines join the observed means. We fitted logistic curves to the counts to recover the 50% mortality temperatures.
- Checked
- Our fits give 37.0, 81.8, 63.0 and 56.0 °C against the papers' 37.1, 82.7, 63.1 and 56 °C, all within 1 °C. Checked by hand, not yet a test in the graph.
- Limits
- One species from one lab. "Active" means moving 48 hours later, not a normal life afterwards.
- More
- How the whole project is built · the claims in the graph
Popular claims, checked
They survive near absolute zero.
Dry, briefly, and mostly unverifiedThe figure traces to Becquerel (1950), whose note is titled "life suspended below 1/20 K", about 0.05 K, in high vacuum, on dried organisms. Reviews quote it as −272.8 °C (0.35 K), which does not match the title, and give no exposure time; we have not read the note itself. Rahm in the 1920s immersed animals in liquid air, nitrogen and helium. Measured on hydrated animals: Richtersius coronifer survives cooling to −196 °C, and active H. exemplaris ~98% at −10 °C but only 45% after 24 h at −20 °C.
Sadowska-Bartosz & Bartosz 2024 (ref. 51, Becquerel's title); Erdmann & Kaczmarek 2017; Hengherr & Schill 2018 pp. 296–298; Lyons et al. 2024 (preprint, abstract)
They survive 150 °C.
A 1921 report, never repeatedRahm reported dried Milnesium tardigradum and Ramazzottius oberhaeuseri reviving after 110–151 °C. Reviews disagree on how long: up to 15, 30 or 35 minutes. The 1983 monograph that summarised Rahm's work describes something else again: tuns in dried moss cooled to −190 °C for 5 hours, then raised "suddenly, within 15 minutes, to 151 °C" (Ramazzotti & Maucci 1983, p. 86). That is a 15-minute ramp across 340 degrees, not a 15-minute hold at 151 °C, and it may be where the "15 minutes" in later reviews comes from. Under controlled 1-hour tests of nine species, no macrobiotid or echiniscid survived 100 °C, and the hardiest, M. tardigradum, went from 92% survival at 100 °C to about 1% at 110 °C.
Neves et al. 2020 (introduction); Sadowska-Bartosz & Bartosz 2024; Hengherr et al. 2009 (abstract); Schill & Hengherr 2018 pp. 283–284
Radiation can't hurt them.
It sterilises them long before it kills themAdults do tolerate enormous doses: gamma LD50s of 4,400 Gy (dry) and 5,000 Gy (hydrated) in M. tardigradum and about 4,200 Gy in active H. exemplaris, a thousand times the human dose. Hydrated animals are about as tolerant as dry ones, so the tun is not the trick. But above 1,000 Gy M. tardigradum became sterile, H. exemplaris fertility collapses above 100 Gy, and hydrated eggs of R. varieornatus have an LD50 of 509 Gy.
Horikawa et al. 2006 (abstract); Beltrán-Pardo et al. 2015; Jönsson et al. 2018 pp. 314–320
They survive in open space.
For 10 days, dry, out of the sunlightIn 2007, dry R. coronifer and M. tardigradum spent 10 days outside a FOTON-M3 capsule in low orbit. Vacuum (10⁻⁶ Pa) and cosmic radiation (100 mGy) had no significant effect. Full solar UV killed every animal of both species; with UVA and UVB only, 12% of M. tardigradum survived. Eggs exposed to everything at once did not hatch.
Jönsson 2019 §2.3–2.4; Jönsson et al. 2018 p. 322 (the 2008 paper itself we could not download)
They withstand 6,000 atmospheres.
As tuns, in oil-like liquidThe 600 MPa figure (about 6,000 atmospheres) is from Seki & Toyoshima 1998, read at first hand on 24 September 2026, and the species are not the ones usually quoted: dry tuns of Macrobiotus occidentalis (95% of 20 survived) and Echiniscus japonicus (80%), in perfluorocarbon. Active animals, in water or in perfluorocarbon, were all dead above 200 MPa. Checked since: dry M. tardigradum tuns in a perfluorocarbon survived 7.5 GPa for up to 13 hours, and 1.2 GPa for 20 minutes, which killed hydrated animals. 20 GPa for 30 minutes killed tuns.
Ono et al. 2012; Horikawa et al. 2009 (abstract); Ono et al. 2016 (abstract)
Tardigrades are living on the Moon.
Unknown, and they could not wake up thereThe Israeli lander Beresheet crashed on the Moon in April 2019 carrying dehydrated tardigrades; nobody knows their fate. In gun tests, frozen Hypsibius all survived hitting sand at 0.73 km/s (0.86 GPa shock) and none at 0.90 km/s (1.14 GPa), with only 2–3 animals per shot. Beresheet was still doing 3,000 km/h (0.83 km/s) at 150 m altitude. Without liquid water, any survivor stays a tun.
Traspas & Burchell 2021; Erdmann et al. 2021; Palka 2024; Grady 2019
They revive after 120 years in dry moss.
Doubted; the records are 20 and 30 yearsIn 1948 Franceschi saw a leg move in a tardigrade from 120-year-old museum moss, before it died. Later authors think the movement was passive swelling as it took up water. The longest reliable dry survival is 20 years (Echiniscus testudo). Frozen at −20 °C, Acutuncus antarcticus revived after 30.5 years: two animals and one egg, and they went on to reproduce.
Schill & Hengherr 2018 p. 275; Jönsson & Bertolani 2001 (abstract); Tsujimoto et al. 2016 (abstract)
They're indestructible, or even immortal.
Only while dry, and not for everActive R. varieornatus lose half their number at 37 °C in a day. Hydrated H. exemplaris die at a UVC dose (2.5 kJ/m²) that a hardier species survives, and almost none recover from 24 h without oxygen. A kept M. tardigradum lives about 83 days. Drying pauses that clock, but DNA damage builds up the longer they stay dry. And not every species can dry out: freshwater species have little or no cryobiotic ability, and H. exemplaris needs slow drying to survive at all.
Neves et al. 2020; Jönsson 2019 §2.3; Hagelbäck & Jönsson 2023; Schill & Hengherr 2018 pp. 276, 285; Guidetti et al. 2011 (abstract); Boothby et al. 2017
How we checked
We read each number in the source's own text: the full text of open-access papers and the three extreme-tolerance chapters of Water Bears (Schill 2018). Where we could read only an abstract, the source line says so; where a number reached us only through a review, the dot is marked as reported, not verified. The rest, including how cryptobiosis works, is on the cryptobiosis page.
Limits: most numbers come from one lab and one population, often with a few dozen animals. "Survived" usually means moving again 1–2 days later, not living a normal life. Species differ a lot, and the lab species H. exemplaris is among the less tolerant.
How this is built
- Data
- The popular claims as reviews and news pieces state them, traced back through the monograph chapters, Ramazzotti & Maucci 1983 and the original papers we could open.
- Built with
- For each claim we followed the citations to the earliest source we could read, and kept disagreeing accounts side by side as separate claims in the graph, linked as contradictions. Verdicts are written from what those sources say.
- Checked
- By hand against each source. Our reading of the "15 minutes at 151 °C" as a heating ramp is a finding we are still checking against Rahm's own papers.
- Limits
- Some sources we know only through later citations or abstracts: Becquerel's note, Seki & Toyoshima 1998 and the 2008 space paper among them. Each source line says so.
- More
- How the whole project is built · the claims in the graph