Nocturnal Bottleneck

nocturnal bottleneck


1. Rhodopsin

The 4 opsins that vertebrates had at the time of the cambrian explosion was just one step short of what we consider the full modern complement of opsins, which includes the 4 classes of cone opsins, plus RH1, used in Rhodopsin. Towards the end of the Cambrian, about 500 million years ago, the Rh opsin would split into the Rh2 opsin, present in the medium wavelength cone in most vertebrates, and the Rh1 opsin, present in the rods of almost all vertebrates.

2. Lose Which Opsins?

There doesn’t seem to be too much rhyme or reason to which opsin is lost. When the sharks became monochromats, sometimes they lost the LWS gene and sometimes their Rh2 gene. When mammals each lost one of their short-wave genes, the monotremes lost the SWS1 gene and we lost the SWS2 gene. It doesn’t matter too much, no matter the gene that is lost, the remaining opsins will in time become tuned towards an ideal overlap between the remaining opsins. If opsins are too close or too far apart, there is not a lot of color contrast. When mammals lost the SWS2 gene and became dichromats, our SWS1 gene shifted up from a 370nm “ultraviolet sensitive” gene to a 420nm VS gene to achieve a more ideal separation with our 560nm LWS gene, to optimize color contrast.

3. Silver Spinyfish

For all known animals, color is determined through interactions of different classes of cones. Through losing their cones that cannot detect light in the deep sea, these fish have lost their color vision. Almost all animals only have a single rod opsin, Rh1, and color vision without cones is not possible. However, some deep sea vertebrates, such as the silver spinyfish, have evolved multiple rod opsin genes. This provides the possibility that this fish could have rod-mediated color vision, but it has been impossible so far to behaviorally test deep sea fish, so we’ve still found no confirmed case of an animal that can use its Rods to see color.

4. Mesopic Bottleneck

Davies (2012) offers a competing theory to the nocturnal bottleneck. Looking at opsin loss in other vertebrates, mammalian color vision actually best resembles the color vision of SNAKES, which are also dichromatic and mostly possess the same two cone opsins as mammals. Snakes – especially those that most resemble the common snake ancestor like boas and pythons – are most active in the twilight, also known as mesopic – or intermediate – light conditions.
Because of the similarities in the visual systems, Davies argues that the mammalian ancestor could have experienced rather a MESOPIC BOTTLENECK – not a nocturnal one – which formed the dichromatic color vision. This helps to explain why mammals did not become monochromatic during the bottleneck.

5. Rod Monochromats

By extension, one would expect this trend of opsin loss to continue with mammals that remained nocturnal… that they would continue losing their opsins until they are left with rod-only eyes.
However, through all of the independent gene deletions in mammals over the eons, no mammal has ever lost the LWS opsin (Davies 2012).
Even the eyes of the blind mole rat, which don’t even have the ability to react to visual stimuli, still retain functional LWS opsins. LWS was the first visual opsin, and it is theorized that perhaps there is some unknown non-visual function for the L-opsin that causes it to always be the last opsin standing.
So would we expect all nocturnal animals to at least become monochromatic? Retaining just the single LWS cone opsin? Afterall, the three families of marine mammals – toothed whales, balleen whales and seals – all independently became lost the SWS1 opsin to become monochromatic.
Indeed, Owl Monkeys, kinkajou & raccoons – all nocturnal – have lost the SWS1 opsin, leaving only LWS cones. However, Lorises, Galagos and the nightmaric Aye Aye are all nocturnal and yet still dichromatic.
Things get even more counterintuitive when you look at bats. Of bats that eat at twilight, some are monochromats, but bats that are fully nocturnal are always dichromats. Why haven’t they lost their color vision, while the twilight bats have? Just goes to show, its complicated.


INTRO (00:00)

I want you all to say hello, to the pouched lamprey. He’s got a cool bag under his eye to hold rocks, an insatiable lust for fish blood and a ring of teeth like a miniature sarlacc… but probably the most magnificent aspect of this terrifying worm fish is its color vision, which- at least on the surface – is better than yours. That’s because this creature – a real living fossil largely unchanged from the ancestor that gave rise to all reptiles, birds, mammals, you and me – has tetrachromatic super-color vision.

Compared to this guy, ALL humans are colorblind. Then why did mammals ditch their super color vision for MEDIOCRE color vision? Today on chromaphobe, we’ll be looking at part 2 of this series on the evolution of color vision: the nocturnal bottleneck. In part 1, we traced back the first appearance of color vision in our ancestors by using phylogenetics and the fossil record. I recommend watching part 1, but it’s not strictly necessary for understanding this video, as long as you generally know what opsins ARE and what they do.


So let’s reset the stage to about 500 million years ago. The common ancestor of all vertebrates, something very similar to this modern lamprey has just evolved, complex, image-forming eyes. Even more impressive, it has photoreceptors featuring the 4 opsins that would give them – and the vast majority of their descendants – tetrachromatic color vision.

For the next half billion years, this organism would then go on to evolve into fish, amphibians, reptiles, birds and mammals… but this arrangement of 4 opsins – the so-called cone complement – stayed amazingly stable.

Almost every descendent would inherit these same 4 cone opsins. Sure, they’d made tweaks along the way to tune the opsin sensitivities to their habitat, but otherwise, they’d leave well enough alone and preserve that gift of color vision. Hell, even the move from water to land didn’t affect our ancestor’s cone complement. Only in incredibly rare cases would an animal ever think that 4 was not enough and evolve a fifth opsin class.

But losing 1 of the 4 opsin classes altogether… is surprisingly common, occurring in snakes, sharks and important for this video… mammals. Of the 4 cone opsins that were available to our lamprey-like ancestors, our later mammalian ancestors would first lose this opsin gene (Rh2) and then this opsin gene (SWS2). As a result, mammals are typically DICHROMATS, a name which references our two remaining cone opsins. This is in addition to our Rh1 rod opsin that doesn’t contribute to color vision.

But why would mammals lose what many consider to be objectively better vision, so much so that so many different mammals would later re-evolve trichromacy?


For most other vertebrates, including our ancestors, modern birds, etc… there was – or is – a selective evolutionary pressure where the environment essentially rewards tetrachromatic color vision with higher reproduction rates. This is the basis of natural selection. For example, most birds rely on color to find mates, a task that would generally be more difficult for trichromatic birds.

This pressure is not only required for the original evolution of tetrachromatic color vision, but also for retaining it. In the simplest sense, Genes follow a “use it or lose it“ principle. If there is no selective pressure for a gene, it will eventually accumulate enough mutations that it falls into disrepair, and then disappears. There needs to be enough selective pressure that weeds these mutated genes out of the genepool, otherwise you will eventually lose the functional genes.

The most common way to relax the evolutionary pressure for color vision is for a species to transition to an environment where light is not bright enough to excite their cones. Because if you’re not using your cones, you’re losing them. This applies to deep sea, nocturnal, cave-dwelling or burrowing animals… any habitat where there is little light…


Just look at the burrowing blind mole rats… whose mammalian ancestors were also tetrachromatic, but have not only lost the opsins necessary for color vision, but have lost their vision entirely as their eyelids have fused together and eyes atrophied to almost nothing.


Many examples of cave dwelling salamanders and fish have also completely lost their eyes so quickly in fact, that they can still otherwise breed with the surface variants of their species, because essentially everything needed for reproduction between the two – which excludes the eyes – is still compatible.

DEEP SEA (4:30)

In the deep sea, below 1000m depth, there is virtually no surface light at all. Many of these fish in the so-called midnight zone DO lose their eyes, such as the cruelly-named Flabby Whalefish. However, many of them retain their vision if they need it to recognize bioluminescence, like this hungry little fella, just chasing after some glow-y food… oh… well I guess he would have also been better off WITHOUT eyes.

If we ascend from the midnight zone into the twilight zone of the ocean, around 200m, where very small amounts of sunlight penetrate, most animals have also found it useful to keep their eyes.

However, in both of the cases of these deep sea fish who have kept their eyes, they have typically lost their cones, which just aren’t sensitive enough to be useful in those dim conditions. Instead, they rely exclusively on their RODS to enable their more sensitive scotopic vision… This means that the fish that CAN see, most likely cannot see color.

DIM SEA (5:33)

A little bit higher in the ocean as we move into the sunlight zone, maybe around 50m, there is enough light to excite cones, but that light is still quite monochromatic. Ocean water is simply much better at letting blue light pass through than other wavelengths. If we look at this graph, which shows how deep certain colors of light can travel underwater: red light can only penetrate to about 10m, while blue light can penetrate to almost 200m. Take a look at this color palette as it descends from the surface to a depth of 50m. After a while, the red target turns black because there is simply no more red sunlight for it to reflect at the camera. This means that the deeper you go in the ocean, the narrower the band of visible light and the less colorful the habitat can be.

So even when cone-driven, photopic vision is possible at 50-100m since the light is still bright enough for cones to function, fewer cones are required to experience the “less complex” colors at this depth. As a result, these fish tend to have only 1 or 2 cones, instead of the 4 that their shallow water counterparts have. Actually, in essentially all families of fish, the deeper a given species lives in the ocean, the fewer cones they tend to have… meaning… less light = fewer opsins… So let’s get back to mammals.


What happened to our mammalian ancestors that made them lose half of their cone opsins? Did they spend a few millennia chilling in the deep sea while they waited for their opsins to mutate away?
That would make this video way more interesting, but it was actually the NOCTURNAL BOTTLENECK.

315MYA, our mammalian ancestors diverged from the dinosaurs’ ancestors. The proto-mammalians couldn’t directly compete with the proto-dinosaurs, so evolution took a different tack. By 250MYA, they evolved to be warm-blooded. This step was probably primarily to avoid parasitic fungi… But not requiring the constant warmth of the sun had a useful side effect: it allowed them to nope outta that dangerous daytime role as readily available appetizers for dinosaurs and quickly fill a niche as small, nocturnal insectivores.

It’s kinda like when I became nocturnal as a 16 year old to avoid seeing my parents. But MY phase only lasted 2 years. Mammals spent over 100 MILLION years as nocturnal, during which they adapted many traits that benefited this nocturnal lifestyle, such as rod-dominant retinas and reflective tapetums – both of which increase their eyes’ sensitivity to the dim, scotopic nighttime environment. Since they rarely – if ever – needed to use their eyes in daylight, the cones became useless and eventually they lost TWO of the cone opsin genes, as described earlier.

And it wasn’t just the two cone opsins… mammals lost all sorts of features that help birds, fish and reptiles detect a richer color spectrum:

  • We lost the elaborate system of colored oil droplets that pre-filter and tweak light color before it hits the opsins;
  • We lost about ⅔ of the non-visual opsins that have various light detecting tasks in the body;
  • We lost double-cones, where pairs of cones directly communicate with each other in a way that we still don’t even understand.

It’s no wonder then that mammals are routinely described as having “impoverished” color vision. “Please sir I want some more COLOR…”


One commenter on the first evolution video claimed that they wanted to “slap that ancestor animal that got rid of its tetrachromatic vision”.

This is certainly a relatable feeling for most colorblind people… but it’s maybe disingenuous to think of them as some deadbeat animal who lost their opsins because they couldn’t “keep up with the payments” or something. Much more likely, they were better off without them, meaning that there was not only a neutral evolutionary pressure, but a negative evolutionary pressure, pushing them to ditch the opsins.

In life, nothing is free, and this includes your body parts. Think back to the blind mole rat. They didn’t lose their eyes simply because they didn’t NEED them – after all, this guy is still on the surface – but because keeping them was a lot of wasted energy in their development and maintenance. Plus, eyes are also a big infection risk when you are literally using your teeth to burrow through dirt.

Likewise, color vision is not free. There are so many sacrifices that a visual system makes to be able to see color:

  • First, more cone types means lower visual acuity, since the ‘pixels’ in your eye essentially become larger to hold four cones instead of one.
  • Second, to calculate color, your eye needs some intermediate cells between the photoreceptors and the optic nerve to compare the signals from multiple photoreceptors. This slows down the propagation of visual information from the eye to the brain.
  • Third, those intermediate cells can also add noise at each step, decreasing the ‘dynamic range’ of the vision.
  • Fourth, more types of cones generally means more cones. More cones means less space for rods. And fewer rods means worse scotopic ‘dim-light’ vision. Dark-adapted animals like the Tarsier or the Giant Squid tend to evolve giant eyes to collect as many precious photons as possible. Clearly, there is a limit to eye size, so they also have to allocate as much of the available retinal real estate to rods. To them, any photon caught by a cone is a waste, so they minimize both the cones and cone types to see in dimmer and dimmer light.

So all else equal, when compared to a tetrachromat, a monochromat could essentially have

  • higher visual resolution
  • better contrast sensitivity
  • faster reaction to vision and
  • stronger night vision.

With all those advantages: higher… better… faster… stronger… Maybe the more appropriate question is: “Why WOULD you keep your color vision?”

OUTRO (11:29)

After the K-T extinction 66MYA, many mammals proliferated to re-fill the daytime niche left vacant by the now extinct dinosaurs. So did their new habitat cause them to re-evolve the opsins required for tetrachromatic color vision? In the words of Bon Jovi, “woah… we’re halfway there!” In the next video of the series, we will look at how primates re-evolved TRIchromacy, and why colorblindness was a necessary side effect.

This is Chromaphobe.

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Post