Ocean Digest

How deep-sea animals make their own light

Ocean Digest has a glow-in-the-dark set alongside its goldfish and bettas: deep-sea jellyfish, anglerfish and hatchetfish. All three are real animals, and all three produce light in genuinely different ways. The anglerfish, in the species that have been studied, does not make its light itself. The hatchetfish shines downwards, at nothing, for one of the neatest reasons in biology. And the jellyfish that gave science its most famous glowing protein is, in that particular respect, not glowing at all.

This page is natural history, not fishkeeping. It explains the chemistry in plain language, separates bioluminescence from fluorescence, and gives a number only where a source we read actually gives one. Where the published answer is that nobody has observed it yet, it says that instead.

Bioluminescence is a chemical reaction. A molecule called a luciferin is oxidised, and an enzyme — a luciferase, or a photoprotein — controls the rate. It has arisen separately many times, so there is no single glowing chemistry. In deep-sea anglerfishes the light is not even the fish's own: it comes from bacteria living in the lure.

Light from a reaction, not from a glowing material

Nothing in the deep sea is coated in glow-in-the-dark paint. Light is released as energy during a chemical reaction, and the standard description is short. In the words of a review of marine bioluminescence in the Annual Review of Marine Science, "In most cases, this reaction is the oxidation of a light-emitting molecule, a luciferin," and "The reaction rate for the luciferin is controlled by an enzyme, either a luciferase or a photoprotein."

That is the whole scheme. A fuel molecule, an enzyme and oxygen. The consequence worth keeping hold of is that light costs the animal something each time, which is why almost everything that glows can also stop glowing.

  • Luciferin — the molecule that is oxidised and gives off the photon. The same review notes that luciferins are "highly conserved across phyla", so very distant animals often share one.
  • Luciferase — the enzyme that controls the reaction. Unlike the luciferins, luciferases and photoproteins are "unique and derived from many evolutionary lineages".
  • Photoprotein — described as "a luciferase variant in which factors required for light emission (including the luciferin and oxygen) are bound together as one unit". It fires when it binds something else, such as a calcium or magnesium ion, which gives very precise control.
  • Photophore — the light organ itself. These are not simple bulbs: the review notes that animals "can turn their photophores on and off, but they can also modulate the intensity, color, and even angular distribution of light".
  • Not a glow stick. The reaction needs a supply of fuel. The Nobel Foundation makes the same point from the other direction when it says the fluorescent protein GFP "does not need any additives to glow, in contrast to aequorin and other bioluminescent proteins, which require a continous supply of energy rich molecules".

There is not one glowing chemistry, there are several

This is the part that most short explanations flatten. Bioluminescence is not a single invention that spread; it is a set of chemically distinct systems in animals that are not closely related. The review puts a figure on the fuels: "Four luciferins are responsible for most light production in the ocean," while adding that "there are undoubtedly many other light-emitting reactions yet to be discovered."

One of those four, coelenterazine, turns up everywhere, including in the jellyfish Ocean Digest draws on. But sharing a luciferin does not mean sharing a mechanism, and the cnidarians are the clearest demonstration of that.

  • All cnidarians use coelenterazine as the luciferin — jellyfish, corals, sea pens, siphonophores.
  • But the enzymes differ inside that one group. Hydrozoans use photoproteins; scyphozoans, the group that includes the true jellyfish, use "primarily luciferases"; and anthozoans use "unrelated luciferases, sometimes in conjunction with a luciferin-binding protein".
  • Some animals do not make their own luciferin. The review discusses the role of coelenterazine moving through the food chain, and in one experiment midshipman fish raised without dietary luciferin began counter-illuminating immediately once they were fed luminous ostracods.
  • Bacteria are a fifth route entirely. Several fish groups do not run the chemistry themselves at all and keep luminous bacteria instead — the anglerfishes, flashlight fish such as Photoblepharon, and shallow-water ponyfishes.
  • How many separate origins? Often you will see "at least 40 times". We are not repeating it, because the review we read declines to give a number: "It is difficult to calculate the number of times that bioluminescence has evolved independently, and there is a potential for both over- and underestimation."

Bioluminescence and fluorescence are different things

These two get used interchangeably and they are not the same. Bioluminescence makes new light out of chemical energy. Fluorescence makes no light of its own: it absorbs a photon and re-emits it at a longer wavelength. The review's glossary is one line — fluorescence is the "emission of light derived from the energy of an absorbed photon". Switch off the light falling on a fluorescent animal and it goes dark. A bioluminescent one does not care.

The jellyfish Aequorea victoria is where the confusion starts, because it does both at once. Its bioluminescent reaction, coelenterazine with the photoprotein aequorin, "would normally emit blue light at around 470 nm". The animal also carries green fluorescent protein, GFP, packed in alongside it, and the Nobel Foundation describes the result plainly: "In jellyfish, the GFP's chromophore simply transforms the blue light from aequorin into green light." The chemistry makes blue. The fluorescent protein re-emits it as green.

That protein, not the glow, is what won the prize. The Royal Swedish Academy of Sciences awarded the 2008 Nobel Prize in Chemistry jointly to Osamu Shimomura, Martin Chalfie and Roger Y. Tsien, in its own words "for the discovery and development of the green fluorescent protein, GFP". GFP was first observed in Aequorea victoria in 1962, in the same work in which Shimomura and Johnson isolated aequorin. Shimomura first saw the blue flash by accident, after pouring jellyfish extract into a sink that happened to contain seawater, whose calcium set the reaction off.

The practical warning is worth keeping. Many photophores can be found by their fluorescence under blue or ultraviolet light, which is how the ventral light organs of hatchetfishes are often photographed. But the review is explicit that "fluorescence does not automatically indicate the presence of luminescence", because chitin and calcium phosphate are fluorescent too.

The anglerfish lure: the light is not the fish's own

The deep-sea anglerfishes are the suborder Ceratioidei — 167 species in 11 families, and the most species-rich fish suborder of the bathypelagic zone. They do not have a luciferin and a luciferase in the lure. They have tenants.

The sentence to quote is from a 2019 study in eLife: "Most female ceratioid anglerfishes host extracellular luminous symbiotic bacteria in a lure-like projection (esca) above the animal's head." Extracellular means the bacteria sit in the lure rather than inside the fish's cells. The lure glows because they do.

  • Which bacteria. Two have been named from sequenced escas: 'Candidatus Enterovibrio luxaltus' in Cryptopsaras couesii, and 'Candidatus Enterovibrio escacola' in Melanocetus johnsonii — host lineages that split roughly 100 million years ago.
  • Where they come from. Not from the mother. The 2019 study found host and symbiont family trees "lacked congruence, and there was no statistical support for codivergence broadly", recovered symbiont-specific gene sequences from seawater collected near the fish, and concluded that "diverse anglerfishes share symbionts that are acquired from the environment".
  • The odd part. Those same symbiont genomes are "reduced ~50% relative to closely related free-living bacteria", a pattern normally seen in bacteria trapped inside a host for good. The authors say outright that it "is unclear why anglerfish symbionts, even from distinct hosts and geographic locations, are so genetically similar".
  • Females, in the species documented. The esca is described as a female structure in this literature, and juveniles have not been found with symbiotic bacteria in their developing lures. We have not found a source we can read that states no male ceratioid of any species carries one, so we are not saying that.
  • What the light is for is still not settled. The study's wording: bioluminescent symbiosis "is thought to be essential to the survival of adult anglerfishes, although the exact function has not been observed". Attracting prey, confusing predators and signalling to mates have all been proposed. The bacteria themselves have never been successfully cultured.

Hatchetfish, and the best idea in the deep sea

A hatchetfish lives in the twilight. There is a little daylight filtering down from above, not enough to see by properly, but enough that anything solid between a predator below and the surface above shows up as a silhouette. The counter to that is to glow downwards at exactly the brightness of the sky.

It is called counter-illumination, and the review's description is compact: "This form of camouflage involves using ventral (lower) photophores to match the dim light coming from the surface, thus making a potential shadow disappear." It adds that this can work either "through a uniform match to the light field or by sufficiently disrupting the silhouette". There is a real audience for the trick: many midwater predators have upwardly directed eyes, "sometimes tinted yellow, to search for silhouettes".

A 2017 study of hatchetfish skin in the Journal of the Royal Society Interface worked on Argyropelecus sp., collected between 300 and 1,000 m, where the authors note ambient light is "5–10 orders of magnitude dimmer than at the surface" and still "1–2 orders of magnitude brighter travelling in the downward direction than horizontally". Their finding adds a second mechanism on top of the glow: the skin's layered structure scatters light that strikes the fish's flank downwards and out through the ventral photophores, so that "about 20% of the light impinging horizontally on the flank of the fish ultimately radiates downwards". Some of the fish's camouflage is recycled light, including light from a hunting predator's own searchbeam.

One honest limit. Matching the sky to order has been measured directly in sergestiid shrimp, in squid and in the midshipman fish Porichthys notatus, which were shown to match the intensity, angular distribution or colour of downwelling light. The hatchetfish work we read is about the optics of the skin, not a behavioural demonstration that the fish tracks the sky's brightness. The photophores point the right way and the physics works; we are not claiming the experiment that has been done on other species has been done on this one.

Why nearly all of it is blue

Seawater is a colour filter, and it is a steep one. NOAA Ocean Exploration states the order bluntly: "Blue light penetrates best, green light is second, yellow light is third, followed by orange light and red light," and "Red light is quickly filtered from water as depth increases and red light effectively never reaches the deep ocean." By 100 m, NOAA says, red light does not penetrate at all, which is why a red fish at that depth looks black.

Above that filter sits a thin lit layer. NOAA puts the euphotic, or sunlit, zone in the upper 200 m, the dysphotic twilight zone from 200 m to 1,000 m, and the aphotic zone below 1,000 m, where "Sunlight does not penetrate". The hatchetfish operates in the twilight band. An anglerfish lure is working in water where the only light is the light animals make.

So the light animals evolved to make, and to see, sits in the band the water lets through. The review's summary is that despite the enormous taxonomic diversity of bioluminescence, "its spectral properties are largely constrained to blue-green wavelengths, centered around 470 nm", tightly coupled to the colour sensitivity of marine eyes.

Blue is the rule rather than a law. The same review records planktonic Tomopteris worms producing golden-yellow light, the amphipod genus Scina emitting unusually short-wavelength light around 440 nm, and some individuals of the amphipod Cyphocaris emitting orange at 595 nm by an unknown mechanism.

Three animals, three different answers

The three creatures in the app's glow set are a useful spread, because no two of them solve the problem the same way. The third column is deliberately cautious: function is much harder to establish than chemistry, and for the anglerfish it has not been observed at all.

AnimalWhere the light comes fromWhat the light appears to be for
Deep-sea anglerfishes (Ceratioidei), e.g. Melanocetus johnsoniiSymbiotic bacteria, living extracellularly in the esca. Not the fish's own chemistryThought to be essential to adult survival, but the exact function "has not been observed". Lure, predator confusion and mate signalling are all proposed
Hatchetfishes (Argyropelecus sp.)Its own photophores, on the underside, plus skin that scatters ambient light down through themCounter-illumination: matching the dim light from above so no silhouette shows to a predator below
Crystal jelly (Aequorea victoria)Its own chemistry: coelenterazine with the photoprotein aequorin, blue at around 470 nm, re-emitted as green by GFPNot established. Most hydrozoans are thought to use light defensively or as a warning
Deep-sea coronate jellyfish (Atolla spp., Periphylla periphylla)Its own chemistry: coelenterazine, with scyphozoans using luciferases rather than photoproteinsProduce "cascading waves of light" and can exude luminous particles; defensive roles are the usual interpretation

How common is it, really

There is one published figure worth quoting, and it comes with its own caveats. A 2017 paper in Scientific Reports annotated 350,536 water-column observations from 240 remotely operated vehicle dives off the California coast between 1999 and 2016, from the surface down to 3,900 m, and reported that "76% of the observed individuals have bioluminescence capability". More than 97% of the cnidarians were bioluminescent, and the authors note that "The percentage of bioluminescent animals is remarkably uniform over depth" — the deep is not glowier than the shallows, it is just darker.

Read the word "capability" carefully. Those animals were not filmed glowing. Each observation was classified from what the literature says about that taxon, which is why the paper's own figures include a large Undefined category, and why the authors note that the upper water column has a lower share of undefined organisms than the less well known deeper water. The 76% is also one region over 17 years, not a global census.

That is the honest state of the number: a real study, a real figure, and a method that depends on prior knowledge of species that are often known from a handful of specimens. Anyone quoting a precise global percentage of luminous deep-sea animals is going beyond what has been measured.

What the app is, and what it is not

Ocean Digest builds an aquarium you keep on your Home Screen, and its glow set is a drawing of these three animals, not a simulation of them. Everything on this page happens in water between a few hundred and a few thousand metres down, at pressures and temperatures no tank reproduces, in animals that in several cases have never been kept alive — the anglerfish's own bacteria have never even been cultured. None of this is advice about keeping anything: there is no aquarium version of a ceratioid anglerfish, and this page is natural history rather than husbandry.

The app's other creatures are freshwater and marine species you will recognise by name — goldfish, guppies, clownfish, seahorses, koi, cichlids and bettas — and naming them is as far as we will go here.

If you came for the widget rather than the biology, the useful pages are elsewhere: our guide on putting a swimming aquarium on your Home Screen covers the sizes and what a widget can show, and the one on how iPhone widgets refresh explains, with Apple's own numbers, why an animated widget sometimes looks frozen. There is also a free fish tank you can build in a browser, with no app required.

Questions

How do deep-sea fish glow?

Most do it chemically. A molecule called a luciferin is oxidised, and an enzyme — a luciferase, or a photoprotein with the components bound together — controls the reaction, releasing light in a light organ called a photophore. A review in the Annual Review of Marine Science describes four luciferins as responsible for most light production in the ocean, so there is no single glowing chemistry shared by everything that glows. A minority of fishes skip the chemistry entirely and keep luminous bacteria instead.

Do anglerfish make their own light?

In the deep-sea anglerfishes that have been studied, no. A 2019 study in eLife states that most female ceratioid anglerfishes host extracellular luminous symbiotic bacteria in the esca, the lure above the head. Two symbionts have been named, 'Candidatus Enterovibrio luxaltus' in Cryptopsaras couesii and 'Candidatus Enterovibrio escacola' in Melanocetus johnsonii, and the same study concluded they are acquired from the environment rather than inherited from the mother. The bacteria have never been successfully cultured.

Is glowing the same as fluorescence?

No. Bioluminescence makes new light from a chemical reaction. Fluorescence absorbs a photon and re-emits it at a longer wavelength, so it needs light shining on it and goes dark without one. Green fluorescent protein is fluorescent, which is what the name says. In the jellyfish Aequorea victoria the two sit together: the chemical reaction emits blue light at around 470 nm, and GFP converts it to green.

What did the glowing jellyfish win the Nobel Prize for?

The Royal Swedish Academy of Sciences awarded the 2008 Nobel Prize in Chemistry jointly to Osamu Shimomura, Martin Chalfie and Roger Y. Tsien "for the discovery and development of the green fluorescent protein, GFP". GFP was first observed in Aequorea victoria in 1962. The prize was for the protein's use as a tag that lets researchers watch individual proteins and cells, not for the jellyfish's glow itself.

What is counter-illumination?

Glowing downwards to cancel your own shadow. In the twilight zone there is still faint light from above, so a solid body shows as a silhouette to anything looking up. Animals with photophores on the underside match that dim downwelling light, which, in the words of one review, makes "a potential shadow disappear". It has been measured in shrimp, squid and the midshipman fish, and hatchetfishes carry exactly the right arrangement of ventral photophores for it.

Why is deep-sea light nearly always blue?

Because that is what seawater transmits. NOAA Ocean Exploration states that blue light penetrates best, followed by green, yellow, orange and red, and that red light effectively never reaches the deep ocean. Light production has converged on the same window: one review reports that marine bioluminescence is largely constrained to blue-green wavelengths centred around 470 nm, matched to the colour sensitivity of marine eyes. There are documented exceptions, including golden-yellow and orange emitters.

What proportion of deep-sea animals glow?

A 2017 paper in Scientific Reports annotated 350,536 observations from 240 ROV dives off California, surface to 3,900 m, and found that 76% of observed individuals had bioluminescence capability, with the proportion remarkably uniform over depth. Treat it as the best available figure rather than a settled one: capability was assigned from the literature rather than from seeing each animal glow, a large share of deeper animals were classed as undefined, and it covers one region.

Can you keep a glowing deep-sea fish in an aquarium?

This page does not give fishkeeping advice, and nothing here should be read as suggesting these animals can be kept. They live at depths where sunlight does not reach, several are known from very few specimens, and in the anglerfishes even the bacteria responsible for the light have never been cultured. The glow set in Ocean Digest is an illustration of them on a screen.

Sources, and which claim came from where. All four research sources below are peer-reviewed. Haddock SHD, Moline MA, Case JF, "Bioluminescence in the Sea", Annual Review of Marine Science 2:443–493, 2010 (https://doi.org/10.1146/annurev-marine-120308-081028, read as the full PDF) for the luciferin/luciferase description, the definition of a photoprotein and its calcium or magnesium trigger, the statement that four luciferins are responsible for most light production in the ocean, luciferins being conserved while luciferases and photoproteins are not, all cnidarians using coelenterazine while hydrozoans use photoproteins and scyphozoans primarily luciferases, animals modulating intensity colour and angular distribution, the refusal to give a number of independent origins, bacterial symbiosis in anglerfishes and flashlight fish and ponyfishes, the midshipman fish raised without dietary luciferin, the definition of fluorescence and the warning that fluorescence does not indicate luminescence, Aequorea victoria emitting at around 470 nm and the aequorin-to-GFP energy transfer, aequorin's original isolation from Aequorea victoria, the counterillumination description and the predators with upward-directed yellow-tinted eyes, the counterillumination experiments on sergestiid shrimp, squid and Porichthys notatus, emission being largely constrained to blue-green centred around 470 nm, the Atolla and Periphylla cascading waves and luminous particles, hydrozoans likely using light defensively, hatchetfish photophores being visible by fluorescence, and the Tomopteris, Scina 440 nm and Cyphocaris 595 nm exceptions. Baker LJ et al., "Diverse deep-sea anglerfishes share a genetically reduced luminous symbiont that is acquired from the environment", eLife 8:e47606, 2019 (https://doi.org/10.7554/eLife.47606, read via https://pmc.ncbi.nlm.nih.gov/articles/PMC6773444/) for the 167 species in 11 families and Ceratioidei being the most speciose bathypelagic fish suborder, the sentence that most female ceratioid anglerfishes host extracellular luminous symbiotic bacteria in the esca, the named symbionts Candidatus Enterovibrio luxaltus and Candidatus Enterovibrio escacola and their hosts Cryptopsaras couesii and Melanocetus johnsonii, the roughly 100 million year host divergence, the lack of phylogenetic congruence and absence of statistical support for codivergence, the recovery of symbiont gene sequences from seawater near hosts, the conclusion of environmental acquisition, the ~50% genome reduction, the symbionts never having been cultured, juveniles not having been found with symbiotic bacteria in developing lures, the function not having been observed, and the authors' statement that it is unclear why the symbionts are so genetically similar. Rosenthal EI, Holt AL, Sweeney AM, "Three-dimensional midwater camouflage from a novel two-component photonic structure in hatchetfish skin", Journal of the Royal Society Interface 14:20161034, 2017 (https://doi.org/10.1098/rsif.2016.1034, read as the full PDF) for the study species Argyropelecus sp., the 300 to 1,000 m collection depth, light being 5 to 10 orders of magnitude dimmer than the surface and downwelling light 1 to 2 orders brighter than horizontal, and about 20% of light striking the flank being redirected downwards through the ventral photophores. Martini S, Haddock SHD, "Quantification of bioluminescence from the surface to the deep sea demonstrates its predominance as an ecological trait", Scientific Reports 7:45750, 2017 (https://doi.org/10.1038/srep45750, read via https://pmc.ncbi.nlm.nih.gov/articles/PMC5379559/) for 76% of observed individuals having bioluminescence capability, the 350,536 water-column observations from 240 ROV dives between 1999 and 2016, the surface to 3,900 m range and 17 years of video, more than 97% of cnidarians being bioluminescent, the percentage being remarkably uniform over depth, and the lower share of undefined organisms in the upper water column. NobelPrize.org, press release for the Nobel Prize in Chemistry 2008 (https://www.nobelprize.org/prizes/chemistry/2008/press-release/) for the laureates Osamu Shimomura, Martin Chalfie and Roger Y. Tsien, the citation "for the discovery and development of the green fluorescent protein, GFP", and GFP first being observed in Aequorea victoria in 1962; and NobelPrize.org, popular information for the same prize (https://www.nobelprize.org/prizes/chemistry/2008/popular-information/) for the sink accident and the calcium-triggered blue flash, the statement that GFP's chromophore transforms the blue light from aequorin into green light, and GFP needing no additives unlike aequorin and other bioluminescent proteins. NOAA Ocean Exploration, "Why are so many deep-sea animals red in color?" (https://oceanexplorer.noaa.gov/ocean-fact/red-color/) for blue light penetrating best followed by green, yellow, orange and red, red light being quickly filtered out and effectively never reaching the deep ocean, and red light not penetrating at 100 metres; and NOAA Ocean Exploration, "How far does light travel in the ocean?" (https://oceanexplorer.noaa.gov/ocean-fact/light-distributed/) for the euphotic zone in the upper 200 m, the dysphotic zone from 200 to 1,000 m, and sunlight not penetrating below 1,000 m. Cut as unconfirmed: the widely repeated figure that bioluminescence evolved independently at least 40 times, which the Annual Review paper we read explicitly declines to quantify; any general statement that only females of all ceratioid anglerfish species carry a lure, since the source we read says "most female ceratioid anglerfishes" and we found no readable source for the stronger claim; any behavioural demonstration that hatchetfishes match the brightness of downwelling light, which has been shown in other species but not in the hatchetfish work we read; any global percentage of luminous deep-sea animals, since the 76% figure is one region over 17 years and is based on capability assigned from the literature rather than observed emission; and any species-level chemistry for hatchetfish photophores, which our sources do not specify. Species names follow the sources: claims about anglerfish symbionts are about Cryptopsaras couesii and Melanocetus johnsonii, the hatchetfish optics are about Argyropelecus sp., and the aequorin and GFP claims are about Aequorea victoria.

Read next: put a swimming aquarium on your Home Screen

Get Ocean Digest on the App Store

More about Ocean Digest · Contact us