Honeybees Vote on Where to Live, and the Losing Side Gets Head-Butted Into Silence

A honeybee swarm hanging off a branch looks like the aftermath of something going wrong. It is the opposite. It is a colony in the middle of the most consequential decision it will ever make, and it is running a genuine collective process to make it.

Thomas Seeley spent roughly forty years working out how, much of it on an island off the coast of Maine chosen because it had almost no natural nest cavities, which let him supply the options and control the experiment.

What a Swarm Actually Is

When a colony outgrows its home, it reproduces by splitting. The old queen leaves with something like half the workers. What stays behind is the remaining bees and a developing new queen.

The departing group flies a short distance and forms a cluster, often on a branch, sometimes somewhere far less convenient. That cluster is a temporary staging point. It has no comb, no stores, and only the honey the bees carried in their own stomachs on the way out. They have a few days at most.

In that window they have to find and agree on a permanent cavity. Pick badly and the colony will not survive the winter.

The Scouts Campaign

A few hundred of the oldest and most experienced foragers act as scouts. They leave the cluster and search, sometimes several kilometres out, for candidate cavities: hollow trees, wall voids, anything enclosed.

Bees are specific about what counts as good. The preferences Seeley measured are consistent. Roughly forty litres of volume, because too small means no room for winter stores and too large is impossible to keep warm. A small entrance of a few square centimetres, easier to defend. The entrance low on the cavity rather than at the top, and preferably facing south. Several metres off the ground, away from most predators.

A scout that finds something inspects it thoroughly. It walks the interior, flies around the outside, and can spend the better part of an hour assessing a single site. Then it returns to the cluster and dances.

The dance is the same waggle dance foragers use for flowers, with direction and distance encoded in the angle and duration of the waggle run. What matters here is the vigour. A scout that found a good site dances longer and repeats more circuits. A scout that found a mediocre site does a shorter, less enthusiastic version.

Other uncommitted scouts read a dance, fly out to inspect that site for themselves, and if convinced, come back and dance for it too. This is the crucial part: they do not simply copy the dance. They go and look, then advertise based on their own assessment. Good sites accumulate advocates faster than poor ones, and the imbalance compounds.

The Head-Butt

For a long time it was unclear how the process converged rather than deadlocking between two decent options. The answer turned out to involve active suppression.

Seeley and colleagues documented what is called the stop signal. A scout committed to one site will approach a bee dancing for a rival site, deliver a brief head-butt, and produce a short vibrational pulse of around 350 Hz. A bee that receives enough of these stops dancing.

Crucially, scouts direct stop signals at dancers for competing sites rather than at random. It is targeted. The effect is to break up the enthusiasm behind losing options, so a swarm that might otherwise split its support converges on one.

It functions as cross-inhibition, the same broad pattern that turns up in neural decision-making, where competing populations suppress each other until one wins. A swarm of bees and a brain arrive at similar architecture for the same reason: both need to commit to one option and stop deliberating.

The Quorum, Not the Majority

The final step is not a vote count at the cluster. The bees never assemble the totals.

Instead, scouts at a candidate site monitor how many other scouts are there with them. When that number crosses a threshold, in Seeley's measurements around fifteen to twenty bees present at once, the scouts at that site treat the decision as made.

They return to the cluster and switch behaviour entirely. They produce a high pitched signal called piping, and they perform buzz-running, shoving through the cluster while vibrating. Both have the same function: warming the flight muscles of thousands of bees that have been hanging still and are too cold to fly.

Once the cluster is warm enough, it lifts off, ten to twenty thousand bees at once, and flies to a location the overwhelming majority of them have never visited and cannot navigate to. Only the small minority of scouts know the way, and they appear to steer by streaking rapidly through the swarm in the correct direction while the rest follow the general drift.

How Often Do They Get It Right

This is the part that makes it more than a curiosity. Seeley ran the test by offering swarms a set of nest boxes of known quality, including one clearly best option.

Swarms chose the best available site around eighty percent of the time. Failures generally happened when two sites were close in quality, which is exactly when a poor decision costs least.

Ten thousand insects, no leader, no one individual with an overview of the choices, and a success rate at picking the best option that most committees would be pleased with. The queen, for the record, does not participate. She is carried along with the result.

Sources

• Seeley, T. D. (2010). Honeybee Democracy. Princeton University Press.
• Seeley, T. D. and Buhrman, S. C. (1999). Group decision making in swarms of honey bees. Behavioral Ecology and Sociobiology.
• Seeley, T. D., Visscher, P. K., Schlegel, T. et al. (2012). Stop signals provide cross inhibition in collective decision-making by honeybee swarms. Science.
• Seeley, T. D. and Visscher, P. K. (2004). Quorum sensing during nest-site selection by honeybee swarms. Behavioral Ecology and Sociobiology.