Skip to content
NatureHQ

Claim check

Can bees find their way on a cloudy day?

Not supportedNo good evidence supports this, or the evidence points the other way.

A patch of blue sky is enough. Scattered sunlight is polarised in a pattern fixed by where the sun is, insects read that pattern, and it still gives the sun’s position when the sun itself is hidden. Rotate a polarising filter over an ant by forty degrees and it walks off forty degrees wrong.

The claim as it circulates

“Bees navigate by the sun, so they are lost when the sky clouds over.”

Where you may have met it: Beekeeping guidance about foraging weather; School material on the waggle dance; Popular accounts of insect navigation

Taking it apart

What was claimed
That insect navigation depends on a direct view of the sun, and that overcast conditions therefore leave a bee or an ant without a compass.
What was actually observed
Sunlight scattering off air molecules acquires a direction of vibration arranged in bands around the sun. A strip at the top of an insect eye — the dorsal rim — carries receptors at different angles to one another and is strongly sensitive to that direction. Homing ants shown only a patch of sky through a polarising filter changed their heading by the angle the filter was rotated. Bees shown a patch of blue sky continued to dance in the correct direction and changed direction when the filter above them was turned.
What the evidence supports
That the sky itself carries the sun’s position, not merely the sun, and that insects read it. The compass therefore survives cloud so long as some clear sky remains visible — and it survives the sun being below the horizon, which is when the pattern is strongest and when several species do their navigating.
What it does not support
It does not mean weather is irrelevant: complete overcast leaves no pattern to read, and bees fly less in poor conditions for reasons that have nothing to do with navigation. Nor does the sky compass amount to a navigation system on its own. It supplies direction; distance is tracked separately, and the animal needs an internal clock to compensate for the sun having moved. The compass is one input among several.

How we know

Rotating the sky above an ant and watching it turn

Desert ants run home in a straight line across ground with no landmarks at all. What are they steering by?

Sunlight scattered by the atmosphere is polarised in a pattern that circles the sun, and the pattern is invisible to us. A foraging desert ant, on its way home, was covered by a screen that hid the sun and left only a patch of sky visible through a filter. Rotating that filter rotates the plane of polarisation the ant can see, without changing the brightness, the colour, or anything else. The ant’s homeward heading was recorded against the rotation.

What happened

The ants turned by the angle the filter had been rotated. With polarisation removed and the sun hidden, they were lost.

What it shows

An animal steering by a property of light that humans cannot perceive at all. Turning the filter turns the ant, by the matching amount, which leaves nothing else it could be reading. It is also a good demonstration of why the sense exists: this ant forages on salt pan where there is nothing to navigate by, and a compass that works from any patch of sky is the difference between getting home and dying of heat.

What it does not show

A compass is not a map — this tells the ant which way it is pointing, and the distance it has run comes from a separate step-counting system that this experiment says nothing about. It is one genus in an extreme habitat, which is precisely why the compass is so dominant in it, and the experiment covers a patch of sky rather than all of it.

The controls — what makes this evidence rather than a story
  • The sun itself screened, so the ant cannot fall back on it.
  • The filter rotated by known angles, giving a quantitative prediction rather than a direction of effect.
  • Trials with polarisation removed entirely while other cues stayed blocked, testing what happens when the compass is taken away.

From Polarization vision

The rest of the answer

The claims underneath

Each one carries its own evidence, scope and caveats. Expand any of them to reach the studies.

Scattered sunlight is polarised in a pattern that circles the sun. Insects with the right region of eye read it as a compass, and it works through cloud and after the sun has set.

Established

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

Rayleigh scattering polarises skylight in a pattern determined by solar position. The dorsal rim area of many insect compound eyes contains photoreceptors with aligned microvilli and untwisted rhabdoms, providing polarisation analysis used for compass orientation. Rotating the plane of polarisation over a homing desert ant rotates its heading by the same angle.

Who this applies to
Demonstrated most decisively in desert ants; dorsal rim specialisation occurs widely across insects.
Studied in
Cataglyphis, Insecta
Why we rate it this way, and what the caveats are
EstablishedHigh confidence

A quantitative manipulation with a quantitative prediction: rotate the filter by an angle and the animal turns by that angle. Little room is left for an alternative reading.

How far it can be extended

The dorsal rim area and polarisation-guided orientation are documented in ants, bees, locusts, crickets and dung beetles.

Caveats

  • A compass is not a map: the ant’s distance information comes from a separate step-counting mechanism.
  • Polarisation sensitivity arises almost incidentally from invertebrate photoreceptor architecture; a compass requires the specialised eye region as well.
  • Desert ants are an extreme case, which is why the compass dominates so completely in them.

Still unanswered

  • How polarisation, sun position and landmark information are weighted against one another when they disagree.

Last reviewed 2026-09-02

The evidence (2 studies)

  • Supports · primary

    Polarization vision

    Wehner and Labhart, 2006 · Invertebrate Vision (Cambridge University Press)

    The rotation experiment on homing ants.

  • Supports · primary

    Polarization vision — a uniform sensory capacity?

    Wehner, 2001 · Journal of Experimental Biology

    The comparative account, and the warning that shared sensitivity does not imply shared function.

Go deeper

Polarised light vision

Rotate a polarising filter over an ant by forty degrees and the ant walks home forty degrees wrong. There is nothing else it could be responding to.

Related things people ask

Last reviewed 2026-09-02