Skip to content
NatureHQ

Insectsadaptation

Eyespots

The spots do work — birds really do leave the butterfly alone. They just do not seem to work by looking like an owl.

Eyespots do deter birds — that part is measured and real. But apparently not by looking like eyes: when researchers varied eye-likeness and conspicuousness separately, equally conspicuous squares and bars worked just as well as circles.

The owl story is one of the most repeated explanations in natural history. A butterfly opens its wings, two large ringed spots appear, and a small bird flinches because it has apparently just been confronted by something with a face. It is vivid, it is intuitive, and it accumulated a century of supportive-looking observation without anybody testing it against the obvious alternative — that a sudden large high-contrast pattern is startling whether or not it resembles anything. Both halves of the modern answer are worth stating together. The deterrent is real: peacock butterflies presented to blue tits survived far more often with their eyespots intact than with them painted over, so the display genuinely stops birds attacking rather than merely delaying them. And the explanation is not eyes. When artificial prey were given marks that varied independently in conspicuousness and in how eye-like they were — circles, bars, squares, at matched and mismatched salience — survival tracked conspicuousness, and eye resemblance added nothing detectable. Two conclusions follow, and they are different in kind. The mechanism is salience: a large, sudden, high-contrast pattern triggers a startle or hesitation response, which for a small bird deciding whether to attack an unfamiliar object is enough. And the case is a good demonstration of how an explanation can persist. The owl hypothesis was never absurd; the flaw was that the evidence people cited for it — birds do flinch, the spots do look like eyes — is equally consistent with the alternative, and nobody separated them until 2008.

Early coverage · 43% complete · reviewed 2026-09-03

What this page covers

Eyespots occur in butterflies and moths, some fish, some frogs and a few birds. Nearly all the experimental work is on lepidopteran patterns tested against bird predators.

Often confused with: Mimicry of a specific predator’s face, which the evidence does not support; Warning coloration, which signals a real defence rather than startling; Deflection marks, which are a separate proposed function of the same structures

Quick facts

The effect is real
Butterflies with eyespots painted out were attacked far more
The mechanism is salience
Conspicuous bars and squares protected as well as eye-like circles
Why the owl story lasted
Its evidence was equally consistent with the alternative nobody tested
A separate proposal
Deflection — drawing attacks to a wing edge — is a different function, not addressed by these tests

The display works

Establish that first, because the correction is about the explanation and not the effect.

How we know

Painting out the spots

Peacock butterflies flash large eyespots when disturbed. Does the display actually stop a bird attacking, or merely delay it?

Peacock butterflies were presented to blue tits with their eyespots either intact or painted over, and with the sound-producing wing hiss either intact or prevented. Whether the bird attacked, and whether the butterfly survived the encounter, was recorded.

What happened

Butterflies with intact eyespots survived far more often than those whose eyespots were painted out. The wing hiss added to the effect, but the visual display accounted for most of it.

What it shows

That the display genuinely deters attack rather than buying a moment. This matters because the later correction is about *why* eyespots work, and it is easy to misread that correction as saying they do not.

What it does not show

It cannot say why they work. Painting over an eyespot removes its conspicuousness as well as its resemblance to an eye, so this design alone cannot separate the two explanations — which is what the next experiment was for.

The controls — what makes this evidence rather than a story
  • Paint applied in both conditions so the manipulation is the pattern rather than the presence of paint.
  • The acoustic component manipulated separately, so the visual and sound contributions can be told apart.
  • Wild-caught birds making ordinary decisions about an unfamiliar insect.

From Prey survival by predator intimidation: an experimental study of peacock butterfly defence against blue tits

It is worth being clear about this before the correction, because "eyespots do not work like an owl" is easy to misread as "eyespots do not work". They do. The difference in survival between intact and painted-over butterflies is large, and the birds involved are ordinary wild-caught predators making ordinary decisions about whether an insect is worth attacking.

But not because they look like eyes

The two explanations were tangled for a century, then separated in one experiment.

Butterfly eyespots do deter birds — but apparently not by looking like eyes. When conspicuousness and eye-likeness were varied separately, equally conspicuous bars and squares protected prey just as well as circles.

Well supported

Good evidence backs this, though some details remain open.

Field experiments varying eye resemblance and conspicuousness independently found survival predicted by conspicuousness, with highly conspicuous non-circular markings conferring protection equivalent to eye-like circles. The intimidation effect is real and attributable to salience rather than to mimicry of vertebrate eyes.

Who this applies to
Demonstrated with artificial prey against wild birds; the intimidation effect itself is separately demonstrated in a real butterfly.
Studied in
Lepidoptera, Aves
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

The independent manipulation is exactly the right test and gives a clear answer. It rests on artificial prey with abstract marks, and does not exclude eye mimicry operating in particular species, which is why this is well-supported rather than established.

How far it can be extended

The conspicuousness result comes from artificial prey and applies to how avian predators respond to salient markings generally.

Caveats

  • The deterrent effect itself is well established; what is not supported is the owl-face explanation for it.
  • Deflection — drawing attacks to a wing edge rather than the body — is a separate proposed function of eyespots and is not addressed by these experiments.

Still unanswered

  • Whether eye mimicry contributes in specific cases with unusually detailed eyespots, which the general result does not exclude.

Last reviewed 2026-09-03

The evidence (3 studies)

How we know

Circles, bars and squares

Do eyespots deter birds because they look like eyes, or because they are simply very hard to ignore?

Artificial prey were presented to wild birds carrying marks that varied independently along two dimensions: how eye-like they were — circles versus bars versus squares — and how conspicuous they were. Survival was scored across the combinations.

What happened

Survival was predicted by conspicuousness. Highly conspicuous bars and squares protected prey as well as eye-like circles of equivalent conspicuousness; eye resemblance added nothing detectable.

What it shows

That the deterrent effect of eyespots is attributable to salience rather than to mimicry of a vertebrate eye. It is also a clean demonstration of why accumulating supportive observations cannot settle a question when a rival hypothesis predicts the same observations.

What it does not show

Artificial prey with abstract marks rather than real butterflies. It does not exclude eye mimicry contributing in particular species with unusually detailed eyespots, and it does not address deflection — the separate proposal that eyespots draw attacks to a survivable part of the wing.

The controls — what makes this evidence rather than a story
  • Eye-likeness and conspicuousness varied independently, which is the entire point: a century of observation had left them confounded.
  • Non-circular shapes matched to circles for conspicuousness, so shape can be assessed at equal salience.
  • Wild bird predation as the measurement rather than human or captive-bird responses.

From Conspicuousness, not eye mimicry, makes eyespots effective antipredator signals

Two explanations, and what separates them
ExplanationPredictsWhat was found
Eye mimicryCircles protect better than equally conspicuous non-circlesNo advantage for circles
ConspicuousnessProtection tracks salience regardless of shapeSurvival tracked conspicuousness

The reason the owl explanation survived so long is instructive rather than embarrassing. Every piece of evidence commonly offered for it — that birds startle, that the spots are strikingly eye-like to us, that the effect is strongest in the largest spots — is exactly what the conspicuousness account also predicts. Two hypotheses that make the same predictions cannot be separated by more observation. They need an experiment that pulls them apart, and that took until 2008.

Diagram

Two explanations that predicted the same observations

Which is why more observation could never have settled it.

Two explanations that predicted the same observationsEye mimicryPredicts: circles beat equallyconspicuous non-circlesNot supportedConspicuousnessPredicts: protection trackssalience, whatever the shapeSupportedThe display does work: spots painted out, and attacks went up.More observations could never have separated these. It needed an experiment.
The same explanation in words

Two panels setting the hypotheses against their predictions. Eye mimicry predicts that circular eye-like marks should outperform equally conspicuous non-circular ones; shown with a ringed circle, and marked not supported. Conspicuousness predicts that protection tracks salience whatever the shape; shown with a filled circle, a square and a bar of comparable prominence, and marked supported. When the two were varied independently against wild birds, survival tracked conspicuousness and eye resemblance added nothing detectable. The display itself does work — butterflies with eyespots painted out were attacked far more.

Related

Where this connects

  • Warning coloration

    Where conspicuousness signals something real

  • Mimicry

    Resembling something specific, which this turns out not to be

  • Camouflage

    The receiver principle, running the other way

Claims about this, checked

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

The research behind this page

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

2008Behavioral Ecology

Conspicuousness, not eye mimicry, makes eyespots effective antipredator signals

Survival was predicted by conspicuousness rather than by resemblance to eyes.

2005Biological Reviews

The role of eyespots as anti-predator mechanisms, principally demonstrated in the Lepidoptera

Both functions have some support, but the evidence for eye mimicry as the mechanism of intimidation is weaker than commonly assumed, and much of the literature had not distinguished eye resemblance from simple conspicuousness.

2005Proceedings of the Royal Society B

Prey survival by predator intimidation: an experimental study of peacock butterfly defence against blue tits

Butterflies with intact eyespots survived far more often than those whose eyespots were painted out.

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 43% completeness against what we would call a finished subject, and was last reviewed on 2026-09-03. It carries 1 claims and answers 3 mapped search questions.

  • no research from the last few years is attached — check for newer work
  • Deflection — the proposal that eyespots draw attacks to a survivable part of the wing — is a separate function and is not resolved here.
  • Eyespots in fish, frogs and birds are mentioned only; the evidence is lepidopteran.
  • Whether unusually detailed eyespots work differently from simple ones has not been tested.