Why does so much of the ocean glow?

A 2017 study went through seventeen years of remotely operated vehicle video from the Californian coast, identifying animals to the lowest possible taxonomic level and scoring whether each is known to produce light.

Across more than 350,000 observations from the surface to 3,900 metres, 76 per cent of the animals that could be assigned were probably capable of light; counting the rest one way or the other gives between 69 and 78 per cent of all animals seen. The proportion is remarkably uniform with depth: the shallow water column is as luminous as the deep, by this measure. Within particular groups it does change, for comb jellies, true jellyfish, arrow worms and crustaceans.

Before this, the figures quoted came from studies of a few groups, from general descriptions of which lineages can glow, and from estimates that were not quantitative.

The same group later surveyed the seafloor beneath those waters and found the share far lower there: 30 to 41 per cent of the animals seen. It is the open water that glows most.

Light usually comes from a light-emitting molecule, a luciferin, oxidised under the control of an enzyme. The enzymes are diverse and particular to their lineages. The luciferins are not: four of them produce most of the light in the ocean, and one, coelenterazine, is the light emitter for nine different phyla.

Counting distinct chemical mechanisms across lineages, the review’s authors estimate that light has evolved at least 40 times, and likely more than 50. A 2021 recount across all of life, not only the sea, puts it at 94 or more. The review reads the number two ways: the trait matters to the animals, and it is relatively easy to evolve, partly because a predator whose prey already carries a luciferin needs only the enzyme.

A 2010 review of bioluminescence in the sea sets out its functions, “ranging from defense to reproduction”, and sorts them into defence, offence and communication within a species.

The review is a good measure of the balance in this field. Some functions have been shown in laboratory experiments on particular species. Many more are thought, suggested or hypothesized: warning displays have not been “convincingly demonstrated”, and for stunning prey there is “little to no experimental evidence”. Why so many animals glow, its authors write, “still does not have a satisfactory answer”.

Counterillumination is well studied: matching the faint light from above so the animal casts no silhouette. Shrimp, squid and the midshipman fish have been shown under controlled conditions to match the light coming down. Burglar alarms, where a flash draws a larger predator onto the attacker, have been shown in laboratory tanks, and the courtship signals of Caribbean ostracods are the best-studied mating system in the sea.

For most of the rest the function is inferred rather than watched. The review’s own caution is that light provoked in a laboratory may not be how an animal uses it in nature, and watching one behave normally in the wild is hard too: a vehicle’s lights and noise can send some fishes, crustaceans and squid away.

Being capable of producing light is not the same as producing it. Some light has been recorded with no vehicle in the water: deep neutrino telescopes, arrays of light sensors built to detect something else, register bioluminescence for between about 1 and 30 per cent of their active time, depending on the site. Those signals rise with the current, as more animals are carried into the detector, which makes them an indirect measure. One record of two and a half years at a single telescope found seasonal blooms up to a hundred times brighter than the background, which its authors put down most likely to luminous bacteria.

The survey measured capability, which is what video can score. Watching animals glow in place, its authors wrote two years later, “remains a technical challenge”. How often each kind flashes of its own accord is the next question, and it needs instruments that watch without illuminating.