How a tardigrade senses its world

An animal that moves has to know something about where it is. Here are seven ways a small animal could do that, what each one would mean, and what is actually known about it in a tardigrade. Four of the seven are blank. They are on this page anyway, because a page that lists only the answered ones would tell you the field is further along than it is.

Counts below are claims in the project graph, each carrying a source and a page number. They describe what we have read, not what exists — the difference matters and is spelled out at the foot of every empty section.

Light

Photoreception — seeing, or at least telling light from dark and which way it comes from. 63claims
16 open

best mapped This is the only sense in a tardigrade with real anatomy behind it. Each animal has two pigment-cup eyes of about 90 µm³, buried inside the outer lobes of the brain. Four cells make each one, and none of them carries an axon — five separate neurons alongside carry the signal out.

Dutta et al. argue that the consistent antero-dorso-lateral orientation of each pigment cup is a structural adaptation for directional light detection — a pair of single-pixel eyes that between them could tell which side light is on. Dutta et al. 2026, bioRxiv p. 10, Discussion · H. exemplaris

The genetics are richer than the anatomy. The animal carries at least 107 opsin genes — the light-catching proteins — and the authors say that is a lower bound. One of them, neuropsin, turns up in the midgut and storage cells, tissues with no known connection to vision.

Watch a signal go in

Light enters the cup, the rhabdomeric cell responds, and the five axon-bearing neurons carry it forward. Then the animation stops, because that is where the evidence stops.

dim bright

The signal stops where the evidence stops. Four cells detect, five neurons carry, and the destination in the brain has never been followed. Everything before the dashed line is from a serial reconstruction; everything after it is unknown. Brightness changes how hard the rhabdomeric cell responds — which is itself an assumption, since nobody has recorded from one.

Where it stops. This animal has eyes wired into its brain and over a hundred opsin genes, and nobody has shown it moving towards or away from light. There is no phototaxis assay for H. exemplaris in the literature we hold. The nearest behavioural evidence is from a velvet worm — Euperipatoides rowelli turns away from light at 373–465 nm but not at 591 or 631 nm (Beckmann et al. 2015). A different phylum.

The full story of how the eye is wired →

Touch

Mechanoreception — feeling contact, texture, vibration, or the pull of its own body. 17claims
2 open

presumed Tardigrades carry structures everybody calls sense organs: cirri and clavae on the heads of heterotardigrades, sensory fields and papillae on eutardigrades. They are innervated — individual nerves have been traced to individual appendages — and that innervation is the whole basis for calling them sensory.

Only 3 of the 17 claims are about the laboratory animal. The rest are Echiniscus testudo, Actinarctus doryphorus and Milnesium, and which structure corresponds to which across those groups is actively disputed in our graph.

The only behavioural test we hold is from 1983, and it failed.

Cephalic appendages of heterotardigrades, and the leg IV papilla and rostral papillae of Milnesium, are presumed sense organs, yet touching them with a fine needle produced no visible reaction. Ramazzotti & Maucci 1983, p. 62 · Tardigrada, Milnesium tardigradum

That is a negative result from a needle and a microscope, four decades old. It is not evidence that the organs do nothing. It is evidence that nobody has looked properly since.

Chemistry

Chemoreception — detecting substances in the water around it. In us this splits into smell and taste. 12claims
1 open

borrowed A tardigrade finds food, and most tardigrades find a mate, so something must be detecting chemistry. One of our twelve claims is about a tardigrade at all; the rest are velvet worms and other animals, kept for comparison.

The clavae — club-shaped head appendages of heterotardigrades — are the usual suspects, on the grounds of their shape and their nerve supply. That is an argument from anatomy, not an observation of the animal responding to anything.

Nobody has watched a tardigrade follow a chemical. No assay, no gradient, no receptor gene described as chemosensory in a tardigrade genome. The strongest honest sentence is that it almost certainly does this and nobody has shown how.

Water

Hygroreception — sensing that the water is going away. This is the one that should exist. 0claims
0 open

nothing A tardigrade's famous trick is surviving being dried out. To do it, it has to start in time: pull in its legs, fold into a tun, and make the protective proteins before the water has gone. That is a response, and a response needs a detector.

We hold 71 claims about desiccation and anhydrobiosis. Exactly three of them use the words sense, detect, trigger or cue, and none of those describes a receptor, a gene or a cell that measures water. The most famous ability of this animal, and the thing that tells it to start is unrecorded.

We cannot tell you whether this is the field's gap or ours. We have read 326 sources. Somebody may have described this in a paper we have not opened, and if you know of one we would like to hear about it. What we can say precisely is that nothing in this graph answers it, and that is a different sentence from "nobody knows".

Temperature

Thermoreception — telling hot from cold. 0claims
0 open

nothing Tardigrades survive from near absolute zero to above boiling in the dried state, and their development rate changes measurably with temperature in the wet one. Whether the animal detects temperature, as opposed to merely being affected by it, is not addressed by anything we hold.

Nothing in the graph, and no open question recorded either. The absence of even a posed question is itself informative: it suggests nobody we have read thought to raise it.

Its own body

Proprioception — knowing where its legs are without looking. 0claims
2 open

nothing measured A tardigrade walks with eight legs in a coordinated gait, and switches gait with speed. Coordination of that kind usually needs feedback from the limbs. We hold no claims about stretch receptors or any proprioceptive organ in a tardigrade — but we do hold two open questions, both asked about other animals, which is a small clue about where to look.

The gait is measured; the feedback that produces it is not. What is known about walking →

Magnetism

Magnetoreception — sensing the Earth's field, as some migrating animals do. 0claims
0 open

nothing Included because it is the kind of thing people ask about tardigrades, and because leaving it off would imply it had been ruled out. It has not been ruled out and it has not been looked for. We hold nothing.

Why the blanks are on the page

Three of these seven senses have real material behind them and four have none. A page that showed only the three would be the more impressive page, and it would give a false picture of a field that is small, young, and mostly working on one animal.

There is also a practical reason. Every blank above is something a person could do. The phototaxis assay is a lamp, a dish and an afternoon. The needle test wants repeating with a modern microscope. Hygroreception wants somebody to film a tardigrade drying and watch what moves first. None of these need a grant.

How the tardigrade eye is wired Four cells, none with an axon, and a question posed in 2013 that closed in 2026 The brain, measured, in the glass tardigrade The outer lobes that hold the eyes, from the X-ray scan, with every published statement Every published statement about the nervous system 104 parts, with sources, locators and the species each was actually seen in What it would need to live somewhere else Not what it survives — what it needs to be awake. Water, CO₂ and light, and why that is the whole list The open questions Everything our sources say is unresolved, with what would settle each one