Jellyfish

A jellyfish is two thin layers of cells with a jelly-like layer between them, and that arrangement removes the need for most of the organs an animal normally has. Oxygen diffuses straight in because nothing is more than a short distance from the surface, so there is no respiratory system and nothing to pump. Digestion happens in a single cavity with one opening, which serves as both mouth and anus. Around 95% of the animal is water.

The nervous system is a net rather than a hierarchy. There is no brain and no central processor; nerve cells are distributed through the bell and coordinate swimming contractions between them. Box jellyfish complicate this usefully: they have four rhopalia, each carrying six eyes, and each processes what it sees more or less independently. Twenty-four eyes, four processing centres, and nothing joining them into a single view.

The name is doing too much work, which matters because findings do not transfer. True jellyfish, box jellyfish and hydrozoan medusae are different classes. Comb jellies are a different phylum entirely and sting nothing. And the Portuguese man o’ war is not an animal in the ordinary sense at all — it is a colony of specialised zooids, one of which is the float, none of which could live alone.

Box jellyfish hunt among mangrove roots and steer around them by sight, judging distance from how much contrast a root shows against the water. That is a solvable problem for a visual system with lenses — and it is a problem whose right answer changes, because water clarity changes. An animal that cannot recalibrate would spend the murky days colliding with roots.

The experiment lowered the contrast so that roots appeared further away than they were, and the animals began to collide. Within about five minutes they had increased their distance from the wall and their rate of successful avoidance turns. The change required the visual cue and the collision to be paired, which is what distinguishes learning an association from a reflex simply wearing out.

What this does not mean is worth as much as what it does. Nothing here shows memory lasting more than minutes, and nothing locates where the learning is held — which in an animal with four unconnected processing centres is the obvious question. It is one species, in small numbers, in one laboratory, and box jellyfish have far better eyes than jellyfish generally. The honest headline is not "jellyfish are smarter than we thought" but "a capacity we assumed required a brain turns out not to".

Most jellyfish have a two-stage life: a polyp fixed to the seabed, which buds off medusae, which swim away and reproduce. The sequence was assumed to run one way, as development generally does. Turritopsis does not respect that. Starved, injured or shocked by temperature, a medusa contracts, settles onto a surface and turns back into a polyp — and the polyp then buds new medusae.

The mechanism is the part that startled developmental biologists. The cells do not return to a stem-like state and start again; they change directly from one differentiated type into another, which is transdifferentiation. An adult cell becoming a different kind of adult cell was not supposed to be available to an animal.

It is called biological immortality, and the phrase means something narrower than it sounds. It says the animal does not have to die of getting old. It says nothing about being eaten, which happens constantly, or about disease, or about being crushed. There is no known limit on how many times the reversal can happen and no measurement of how often it happens at sea. An immortal animal that is also a common food item is a strange thing to describe, and both halves are true.

Most jellyfish stings are painful and harmless. A small number of species are genuinely dangerous, and they are geographically concentrated: the Australian box jellyfish and the much smaller Irukandji are the serious cases, and both occur in tropical Indo-Pacific waters. Elsewhere, the realistic outcome of a sting is discomfort.

The first-aid folklore is unusually bad, and the reason is that different species need different treatment while the advice circulates as though jellyfish were one animal. Urine is not a treatment and can make discharge worse. Fresh water does the same, by changing the osmotic balance around undischarged nematocysts. Rubbing the area fires more of them.

Blooms are the other way jellyfish reach the news, and the picture is less settled than coverage suggests. Jellyfish populations swing enormously between years, and the sampling record in most places is too short to distinguish a long-term rise from a large natural oscillation. Warming, overfishing and nutrient run-off are all plausible drivers with real supporting evidence in specific places, and "jellyfish are taking over the oceans" outruns what the data support.

• Where is learning held in an animal with no centre?

  Why it matters: Four rhopalia process vision independently. If a box jellyfish learns, something has to store it, and there is no obvious structure to store it in.

  What would settle it: Recording from rhopalia during and after learning, and testing whether learning in one transfers to another.

• How long does that learning last?

  Why it matters: Minutes is what has been measured. Whether it persists across hours is the difference between recalibration and memory.

• Are jellyfish actually increasing?

  Why it matters: Blooms are dramatic and natural variability is enormous. Most time series are too short to separate a trend from an oscillation, and policy arguments are being made on both readings.

  What would settle it: Consistent long-term monitoring, which barely exists.

• How often does Turritopsis actually revert in the wild?

  Why it matters: The reversal is a laboratory observation under induced stress. Its ecological significance depends on a frequency nobody has measured.