In a world cloaked in darkness, these fish may have a unique way of seeing colors


Scientists have long presumed that the creatures in the deep ocean experienced a dark, colorless world. But some of the fish who live there may be able to see colors thanks to a newly discovered visual system that’s never been seen in vertebrates before.

The find, reported in the journal Science, challenges long-held assumptions about how these animals perceive colors.

“We didn’t expect this at all,” said study co-leader Fabio Cortesi, a marine biologist at the University of Queensland in Brisbane, Australia. “We were like, ‘Whoa, what’s happening?’”


Christopher Kenaley, a biologist at Boston College who wasn’t involved in the study, said the report should prompt scientists to reconsider the conventional wisdom that deep-sea fish have very limited vision.

“There’s some important questions in the deep sea about how animals communicate down there,” he said. “This gives us insight about how they might be able to detect one another.”

Vertebrate vision is made possible by photoreceptor cells in the back of the eye. These cells — called rods and cones — include pigment proteins that detect different types of light and relay that information to the brain.

A typical vertebrate eye has multiple types of cones that work in bright conditions — each capable of sensing a certain range of colors — and one type of rod that senses light when the environment is dim. The rods can’t distinguish between colors because they all have the same pigment protein, which is why humans and most other animals are said to be colorblind at night.

Cortesi and his colleagues wondered if they could find some exceptions among fish who lived in perpetually dark environments. Their question was prompted by a 2015 study of mostly shallow-water fish that turned up several species with more genes for cone pigment proteins than scientists had expected.


“We just thought if other fish are more variable in their visual system than previously thought, we should look at the deep-sea fishes,” said Walter Salzburger, an evolutionary biologist at the University of Basel in Switzerland who oversaw both the 2015 study and the new one. After all, if any fish stood to benefit by having more ways to see in dark conditions, it would be fish that live in water so deep that light barely reaches them.

Very little is known about fish that reside more than 1,000 meters below sea level. Some developed large pupils and very long rods help them catch whatever light is around. (At those depths, most of the light is produced by fish themselves through bioluminescence.)

For the new study, the researchers started by counting the number of genes for both rod and cone pigment proteins in the genomes of 101 species of fish living in a diverse array of habitats. Although they found a dozen species with up to seven cone pigment genes, what really struck them was the discovery of 13 species that had more than one rod pigment gene.

Four of those species stood out with five or more of the genes: the tube-eye (Stylephorus chordatus), the glacier lanternfish (Benthosema glaciale), the longwing spinyfin (Diretmoides pauciradiatus) and the silver spinyfin (Diretmus argenteus).

All four of these fish live 1,000 meters to 2,000 meters below sea level. Their most recent common ancestor dates to more than 100 million years ago, so the researchers think the additional genes evolved independently in each lineage.

“Is it to see prey species? Or to find mates in a completely dark, or almost dark, environment? Or to avoid predators?” Salzburger said. “These are the three main evolutionary advantages we can think of.”


But were these fish actually using their extra pigment proteins? To answer that question, the team examined specimens representing 36 different fish species. Some tissue samples were already preserved in laboratories, and others were acquired on fishing expeditions.

Cortesi and other researchers dragged a net through the ocean from Perth to Sri Lanka. They trawled at night so the fish wouldn’t encounter sunlight that might damage their eyes. It could take six hours to fill just one little bucket of the thumb-size fish, Cortesi said.

Most of the 36 species had only one active gene for producing rod pigment proteins. The species with at least five rod pigment genes had at least three that were active.

The star was the silver spinyfin. It had 38 genes for rod pigment proteins, and 14 of those proteins were actually at work inside the eye. (For the sake of comparison, most humans use only three types of cone pigment proteins to see the world in color.)

It’s not clear how the silver spinyfin uses all of these rod pigments, but the scientists suspect they may increase their sensitivity to light, Salzburger said.

To get an idea of what colors the silver spinyfin might see, the researchers enlisted bacteria to reproduce some of its rod pigment proteins in a petri dish. Then they shined a light on each of them to see what portion of the spectrum the pigment proteins were able to absorb. They found that they could detect light across the entire spectrum of bioluminescence — from different shades of blue and green to yellow.


Finally, they used those results to predict the colors other deep-water fish with multiple rod pigment proteins could see. The shapes of those proteins were the key, since different shapes are sensitive to different wavelengths of light.

Their work suggested that the lantern fish, tube-eye and longwing spinyfin probably could detect blue light, as well as shades of green and yellow-green. But they would not have as wide a range as the silver spinyfish.

Without behavioral experiments, the scientists can’t know for sure whether these fish really do use their rods to see color. The experiments would be difficult to pull off because the fish aren’t just hard to get, they don’t live long once they are brought to the surface, Salzburger said. (The water pressure at sea level is much lower than what they’re used to in the deep ocean.)

Nevertheless, scientists who were not involved in the study agreed that identifying fish with multiple rod pigment proteins was a novelty in itself.

Biologist David Hunt, a professor emeritus at the University of Western Australia who has specialized in the evolution of vertebrate vision, called its findings “quite astounding.”

“That is something that is unknown and really totally unexpected,” he said. “I’m still trying to get my head around really what it means.”