Skip to content
NatureHQ

Senses and abilitiesability

Ultraviolet vision

Two birds identical to a human eye can be obviously different to each other. Put a filter over the window and the choosing stops.

Many animals have a receptor class sensitive to wavelengths shorter than human vision reaches, so they can make distinctions we cannot — between two flowers, or two birds, that look identical to us. What that adds up to as an experience is not something any experiment has established.

Human colour vision is built from three cone classes, and everything you have ever seen is a set of three numbers compared against each other. A bird has four, including one sensitive into the ultraviolet, which does not mean a bird sees your colours plus one more: adding a fourth channel changes every comparison, so the whole space is a different shape rather than a larger version of ours. The consequence is concrete and testable. Two birds that look the same to a human can differ sharply in ultraviolet reflectance, and if that difference matters to the animals, removing it should change their behaviour — which is exactly what a filter over an aviary window does to female blue tits choosing between males. The reason humans are not part of this is partly the retina and mostly the lens, which yellows with age and absorbs ultraviolet before it arrives; people whose lens has been removed report seeing shorter wavelengths, usually as a whitish violet. That last observation is worth holding onto, because it is the one place in this subject where a first-person report exists at all, and what it describes is not a new colour. It is the familiar end of the spectrum, extended.

Developed coverage · 67% complete · reviewed 2026-09-02

What this page covers

Ultraviolet sensitivity is widespread: most birds, many fish and reptiles, and a great many insects. Among mammals it is the exception rather than the rule, and adult humans are among the animals that block it hardest.

Often confused with: Colour vision in general, which is about comparing channels rather than about which wavelengths reach the eye; Seeing in the dark, which is a different problem entirely and mostly about rods

Quick facts

Bird cone classes
Four, one of them ultraviolet-sensitive — a differently shaped space, not a bigger one
Bees
Three channels, shifted: ultraviolet, blue, green — and no red
Mantis shrimp
Twelve receptor classes, and coarser colour discrimination than you have
Why not us
The lens absorbs it before the retina gets a chance

A fourth channel is not a fourth colour

What adding a receptor class actually does to a visual system, which is not what the phrase suggests.

Diagram

Where the receptors sit, and why the space is a different shape

Curve positions are indicative. Where a receptor peaks varies between species and even between individuals.

The same light, divided up three different waysUVbluegreenyellowredshorter wavelengthslongerHumanthree conesBeethree, shifted — no redBirdfour, plus oil-droplet filtersCurve positions are indicative. A bird’s fourth channel is not a bonus colour — it makes afour-dimensional space that no three-primary screen can show you.
The same explanation in words

A horizontal wavelength axis running from 300 to 700 nanometres, with the human visible range marked from about 400 to 700 and the region below 400 marked as ultraviolet. Above the axis, three overlapping sensitivity curves are drawn for human cones — short, medium and long wavelength — all sitting inside the visible range. Below, four curves are drawn for a typical bird: the same broad arrangement plus a fourth peaking in the ultraviolet, outside anything a human eye receives. A note explains that the extra curve does not add a colour to the end of the human set; because colour is produced by comparing channels against each other, a fourth channel changes every comparison, making the space a different shape rather than a longer version of ours.

Birds have four kinds of colour cone to a human’s three, and the fourth reaches into the ultraviolet. Take the ultraviolet away and females choose different males.

Established

Specialists would state this without hedging. Multiple independent lines of evidence agree.

Birds are tetrachromatic, with a fourth single-cone class maximally sensitive either in the violet or the ultraviolet depending on lineage, each cone filtered by a coloured oil droplet. Removing ultraviolet wavelengths alters mate-choice outcomes in zebra finches.

Who this applies to
Tetrachromacy is general across birds; the behavioural demonstration is from one species.
Studied in
Aves, Taeniopygia guttata
Why we rate it this way, and what the caveats are
EstablishedHigh confidence

The receptor physiology is thoroughly characterised across many species, and the behavioural consequence is demonstrated by a within-individual manipulation.

How far it can be extended

The four-cone arrangement is documented across the avian radiation; whether ultraviolet affects choice has been tested in a small number of species.

Caveats

  • Not all birds have an ultraviolet-sensitive fourth cone; many have a violet-sensitive one instead, which shifts the whole system.
  • Oil droplets filter each channel before the light arrives, trading sensitivity for discrimination in a way no human eye does.
  • Removing a waveband also alters brightness slightly, which is hard to control perfectly.

Still unanswered

  • What females are actually assessing in the ultraviolet, as opposed to that they assess something.

Last reviewed 2026-09-02

The evidence (2 studies)

Two details of bird eyes usually get left out and both matter. The first is that every cone sits behind a coloured oil droplet that filters the light before it arrives, narrowing what each receptor responds to and throwing away a good deal of light to do it — a trade of sensitivity for sharper separation between channels. The second is that "the ultraviolet cone" is really two arrangements: some birds have a receptor peaking well into the ultraviolet, others one that sits at the violet edge, and the two produce measurably different colour spaces. Which arrangement a species has turns out to track what it needs to tell apart.

Words used here
Cone class
A type of photoreceptor with its own sensitivity curve. Colour comes from comparing the outputs of different classes, which is why the number of classes sets the shape of the colour space.
Tetrachromacy
Colour vision built on four receptor classes rather than three. Standard in birds, and present in some fish, reptiles and insects.
Oil droplet
A coloured droplet sitting in front of a bird’s cone, filtering the light that reaches it. It sharpens the separation between channels at the cost of losing light.

Does it change anything the animal does?

Detection is a measurement of a retina. Whether it matters is a question about behaviour, and it needs a different experiment.

It is possible for a receptor to respond to something an animal never uses. Showing that ultraviolet matters means changing only the ultraviolet and watching what happens to a decision the animal was already making — which is a harder experiment to design than it sounds, because almost any filter you put over a light source also changes brightness.

How we know

Taking the ultraviolet out of a bird and asking a female to choose again

Birds have a fourth cone class that responds to ultraviolet. Does what it detects change anything they do?

Female zebra finches were given a choice between males, seen through a filter that transmitted the whole spectrum including ultraviolet. Each female was then offered the same males again through a filter that removed the ultraviolet while altering the rest of the light as little as possible. If ultraviolet carries no information, the same female should rank the same males the same way twice.

What happened

The preferences changed. Females did not rank the same males the same way once ultraviolet was removed.

What it shows

That the fourth cone is not decorative. Something in the ultraviolet is part of what a female is assessing, and removing it changes a real decision. The wider consequence is uncomfortable for a century of ornithology: every plumage colour ever described by eye, and every photograph ever taken of a bird, was recorded through an instrument blind to part of the signal.

What it does not show

It does not identify what the females are reading — brightness, a specific patch, or something correlated with condition — and it does not show what ultraviolet looks like to a bird, which is not a question a choice test can reach. Removing a waveband also changes overall brightness slightly, which is difficult to control away entirely. One captive species, long domesticated.

The controls — what makes this evidence rather than a story
  • Each female tested under both conditions, so preferences are compared within an animal rather than between animals.
  • Filters matched as closely as possible outside the ultraviolet, so the manipulation is the waveband rather than the brightness.
  • The males themselves unchanged — nothing was done to the birds being assessed.

From Ultraviolet vision and mate choice in zebra finches

The same logic runs through the rest of the subject. Ultraviolet reflectance has been shown to matter in kestrels hunting voles, in fish choosing mates, and in flowers whose patterns are largely invisible to us and strongly patterned to a bee. In each case the demonstration is not that the animal can detect ultraviolet; it is that removing the ultraviolet changes the choice.

The other property of light we cannot read

The bee: shifted, not richer

A bee has three colour channels like us, and they sit in different places: ultraviolet, blue and green, with nothing sensitive to red. Its colour world is shifted, not expanded.

Established

Specialists would state this without hedging. Multiple independent lines of evidence agree.

Honey bee colour vision is trichromatic with photoreceptor peaks in the ultraviolet, blue and green, and no long-wavelength receptor. Behavioural discrimination against a full brightness series establishes true colour vision rather than intensity discrimination, with characteristic confusions where human and bee colour spaces diverge.

Who this applies to
Honey bees; the ultraviolet–blue–green arrangement is widespread among insects with some variation.
Studied in
Apis mellifera
Why we rate it this way, and what the caveats are
EstablishedHigh confidence

Behavioural demonstration with the grey-series control, later confirmed by direct receptor measurement — two independent lines agreeing.

How far it can be extended

The same three-receptor arrangement occurs across many hymenopterans and other insect groups.

Caveats

  • Three channels, like ours — the difference is where they sit, not how many there are.
  • Bees are effectively red-blind, which is why red flowers are typically bird- rather than bee-pollinated.
  • Von Frisch could not have identified the ultraviolet channel; it was established decades later.

Still unanswered

  • How much of floral colour diversity is explained by bee colour space as opposed to by other pollinators.

Last reviewed 2026-09-02

The evidence (2 studies)

How we know

One coloured card among a hundred greys

Does a bee returning to a blue card see blue, or is it only telling light from dark?

Bees were trained to feed from a card of a single colour. Then the reward was removed and the coloured card placed in an array of grey cards spanning every shade from near-white to near-black — including, necessarily, a grey of exactly the same lightness as the colour. An animal that discriminates only brightness has no way of picking the coloured card out of that array; an animal with colour vision does.

What happened

Trained bees went to the coloured card regardless of the greys around it — and made a specific, repeatable confusion, treating blue-green as indistinguishable from grey.

What it shows

True colour vision, and a colour system that is not ours. The grey array is what makes it evidence: without it, a bee flying to a yellow card has demonstrated nothing except that it can see a difference. The confusion is the more interesting half, because it says the bee’s colour categories are drawn in different places from a human’s.

What it does not show

It does not identify the receptors, and von Frisch could not have found the most important one: the bee’s ultraviolet channel was established decades later, which means the colour space this experiment mapped was shifted further from ours than it was possible to know at the time. Nor does it show what a bee experiences.

The controls — what makes this evidence rather than a story
  • A full brightness series of greys, so a matching lightness is always present.
  • The reward removed during testing, so the bee is choosing on appearance rather than on food.
  • Card positions changed between trials, removing position learning.

From Der Farbensinn und Formensinn der Biene

The bee is the case that should stop anyone from ranking colour vision on a single axis. A bee has three channels, as we do. Its channels sit in different places, so it reads ultraviolet patterns on petals that we cannot see at all, and it is effectively blind to red. Neither system contains the other, and neither is a degraded version of the other. They are two different three-number summaries of the same light.

What the bee does with it

Twelve receptors, and worse colour vision than you

The mantis shrimp is the single most instructive result in sensory biology, and almost every popular account of it is upside down.

The short answer

Does a mantis shrimp see more colours than a human?

No. It has twelve photoreceptor classes to our three, and when it was actually asked to tell two colours apart it performed considerably worse than a human — it needs wavelengths far further apart before it treats them as different.

The reason is that colour discrimination does not come from receptors; it comes from comparing them, and comparison is expensive. A human visual system takes three channels and does a great deal of neural work on the differences, which is why we can separate wavelengths a few nanometres apart. The proposal for the mantis shrimp is that it does almost none of that comparing, and instead reads which of its twelve narrow channels fired — a fast, coarse scan of the spectrum rather than a fine measurement of it. For an animal that strikes in milliseconds, quick and rough may be the better trade. What the result establishes for certain is the negative: the number of receptor classes tells you almost nothing about what an animal can discriminate, and every headline that ranks eyes by receptor count is ranking on the wrong quantity.

Check it for yourself

Diagram

Counting receptors tells you almost nothing

Discrimination thresholds are approximate and depend on where in the spectrum the test sits.

Twelve receptors, and worse colour vision than yoursColour receptor classesHumanMantis shrimpSmallest colour difference each can actually tell apartHumana few nanometresMantis shrimpabout 15 to 25 nanometresColour discrimination comes from comparing channels, not from counting them.This animal appears to skip the comparison — fewer distinctions, made faster.Bars are proportional to the measured thresholds.
The same explanation in words

Two rows compared. The first shows a human eye: three receptor classes, drawn as three marks, and a discrimination threshold of a few nanometres — two wavelengths that close can be told apart. The second shows a mantis shrimp: twelve receptor classes, drawn as twelve marks, and a threshold of tens of nanometres, several times coarser. Between them a note explains why the intuition fails: colour discrimination comes from comparing channels rather than from having them, and the proposal for the mantis shrimp is that it reads which of its twelve narrow channels fired instead of comparing them — a fast, coarse scan rather than a fine measurement.

Mantis shrimp have twelve classes of colour receptor and can tell colours apart less finely than a human with three. Counting receptors does not measure colour vision.

Well supported

Good evidence backs this, though some details remain open.

Wavelength discrimination thresholds in Haptosquilla trispinosa are approximately 15–25 nm, substantially coarser than human thresholds of a few nanometres, despite twelve photoreceptor classes. The evidence is consistent with wavelength recognition by receptor identity rather than by opponent comparison between channels.

Who this applies to
Measured in one stomatopod; the general point about receptor count applies wherever colour vision is discussed.Do not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Haptosquilla trispinosa
Why we rate it this way, and what the caveats are
Well supportedHigh confidence

A direct behavioural threshold measurement using the standard method, giving a number comparable with those from other species.

How far it can be extended

Stomatopods are diverse and the discrimination threshold has been measured in very few of them.

Caveats

  • This does not make mantis shrimp vision poor: the same eyes detect linear and circular polarisation, which no vertebrate eye does.
  • The proposed mechanism is an interpretation of the threshold, not a recording from the brain.
  • One species, few individuals, in a trained task.

Still unanswered

  • What the twelve channels are actually for, if not fine colour discrimination.

Last reviewed 2026-09-02

The evidence (2 studies)

How we know

Asking the animal with twelve colour receptors to tell two colours apart

Mantis shrimp have twelve classes of colour receptor to a human’s three. Does that mean they see finer colour differences?

Mantis shrimp were trained to strike a fibre-optic target glowing at one particular wavelength, rewarded when they hit it. They were then offered pairs of targets whose wavelengths differed by progressively smaller amounts, and the smallest difference each animal could still get right was measured. This is the standard test of colour discrimination and it has been run on humans, bees, birds and fish, so the numbers are directly comparable.

What happened

The shrimp could distinguish wavelengths about 15 to 25 nanometres apart and failed below that. Humans, with three receptor types, manage differences of a few nanometres across much of the spectrum.

What it shows

That counting receptors does not tell you how well an animal discriminates colour, and that the popular version of this animal has it backwards. Colour vision comes from *comparing* channels, and comparing twelve of them is expensive; the evidence here suggests the shrimp does not compare at all, but reads a wavelength off directly from which receptor it excites. Fewer distinctions, made faster — which suits an animal whose entire hunting strategy is one strike measured in milliseconds.

What it does not show

It does not establish the mechanism, which remains an interpretation of the threshold rather than a recording from the brain. It does not mean mantis shrimp vision is poor — the same eyes do things no vertebrate eye can, including reading circular polarisation. And it is one species, in a trained task, with few individuals.

The controls — what makes this evidence rather than a story
  • A trained rewarded task, so the animal is reporting a discrimination rather than a preference.
  • Wavelength pairs narrowed step by step to find the threshold rather than assuming one.
  • Target brightness controlled, so the animals cannot succeed on intensity.

From A different form of color vision in mantis shrimp

The pictures, and what they are actually showing you

Pictures captioned "what a bee sees" are translations into human colour of something a human eye cannot receive. They can be accurate about the information and cannot be accurate about the experience.

Established

Specialists would state this without hedging. Multiple independent lines of evidence agree.

Rendering a non-human visual scene requires mapping receptor excitations from one colour space into another with a different dimensionality and different primaries. The mapping is not unique, and no rendering can represent a percept whose receptor basis the viewer lacks.

Who this applies to
A methodological point about visual simulations, applying wherever a non-human sensory world is depicted.
Studied in
Animalia
Why we rate it this way, and what the caveats are
EstablishedHigh confidence

A four-dimensional colour space cannot be displayed on a three-primary screen without loss, and the choice of what to lose is made by the person producing the image. That is arithmetic, not opinion.

How far it can be extended

The constraint is mathematical rather than biological: it holds for any mapping between colour spaces of different dimensionality.

Caveats

  • False-colour images are genuinely useful: showing where ultraviolet reflectance sits on a flower conveys real information, and the objection is to the caption rather than the picture.
  • The same limit applies to every modality, not only vision — there is no honest rendering of what a magnetic compass feels like either.

Still unanswered

  • Whether receptor-based models predict animal colour discrimination well enough to be used as ground truth, which they do only for a handful of well-tested species.

Last reviewed 2026-09-02

The evidence (3 studies)

NatureHQ does not publish simulated animal vision, and the reason is narrow rather than purist. An ultraviolet photograph of a flower is a measurement and we would publish one; a false-colour rendering of it captioned as a bee’s view is a measurement plus an invented mapping plus a claim about experience, and only the first part survives scrutiny. Where the ultraviolet pattern on a flower matters to a page here, it is described and diagrammed rather than rendered.

What is still unsettled

  • Why do some birds have a true ultraviolet receptor and their close relatives a violet one?

    Why it matters: The switch has happened repeatedly and in both directions, which suggests it is under selection rather than accidental — but what is selecting it is not established.

    What would settle it: Comparative work relating receptor type to the discriminations a species actually has to make, in enough lineages to separate the pattern from the phylogeny.

  • How much of the colour signalling we describe is visible to the animals it is supposedly aimed at?

    Why it matters: Plumage and petal descriptions in the literature are overwhelmingly human descriptions, and the ultraviolet component is routinely absent from them.

    What would settle it: Spectrophotometry as standard practice rather than as a specialist add-on, which is happening slowly.

  • Is the coarse-scan account of mantis shrimp vision correct?

    Why it matters: It is the best available explanation of a genuinely strange result, and it remains an interpretation of behavioural data rather than something observed in the animal’s nervous system.

    What would settle it: Recording from the processing stages behind those twelve channels, which has barely been attempted.

Claims about this, checked

Things people have heard, and what the evidence actually supports.

The research behind this page

5 studies, newest first. Each one has a page explaining what it found and what it could not show.

2014Science

A different form of color vision in mantis shrimp

The animals discriminated wavelengths differing by roughly 15–25 nanometres and failed below that — far coarser than humans, who discriminate differences of a few nanometres across much of the spectrum, despite the shrimp having twelve photoreceptor classes to a human’s three.

2013Oxford University Press

Sensory Ecology, Behaviour, and Evolution

Sensory systems are shaped by the physics of the environment and by the specific tasks an animal performs, and no sensory system is general-purpose: sensitivity in one dimension is routinely traded against resolution, speed or energy in another.

2001Progress in Retinal and Eye Research

The visual ecology of avian photoreceptors

Birds are tetrachromatic, with a fourth cone class sensitive either to violet or to ultraviolet depending on lineage, and each cone sits behind a coloured oil droplet that narrows its effective sensitivity.

1996Nature

Ultraviolet vision and mate choice in zebra finches

Female preferences changed when ultraviolet was removed, and the ranking of males under the two conditions did not agree.

1914Zoologische Jahrbücher

Der Farbensinn und Formensinn der Biene

Trained bees selected the coloured card regardless of the brightness of the surrounding greys, and confused blue-green with grey — the pattern expected of an animal with colour vision and a different set of receptors from ours.

This page is a stop on a longer route

A guided journey reads several subjects in a deliberate order, with an argument for why one follows another. You can join in the middle.

Where to go from here

Each of these follows from something on this page — a relationship in the evidence, a claim people ask about, or the next mechanism along.

How complete this page is, and what it is still missing

NatureHQ publishes its own gaps. This page is at 67% completeness against what we would call a finished subject, and was last reviewed on 2026-09-02. It carries 4 claims and answers 19 mapped search questions.

  • no research from the last few years is attached — check for newer work
  • Ultraviolet vision in fish and reptiles is named rather than treated, and the fish literature is large.
  • Nothing here covers ultraviolet damage or the cost of admitting short wavelengths to a retina, which is part of why some animals block it.
  • Human aphakic vision is described from published reports and deserves its own careful treatment.