Why Does Half a Hive Leave? A Colony Reproducing Itself

A honeybee colony is the reproductive unit rather than an individual bee, and swarming is how that unit reproduces. When a colony becomes crowded and conditions are right, workers raise new queens and the existing queen leaves with roughly half the workforce, settling temporarily as a cluster on a branch or a wall while scouts search for a permanent home. The colony left behind retains the comb, the brood and the stores and produces a new queen from the cells prepared beforehand. The swarm therefore takes the workforce and the experience while the original colony takes the infrastructure, which is a division that gives both halves a reasonable chance. A hanging swarm is generally docile, since the bees have no nest to defend and are full of honey.

The decision is made collectively and the process has been studied in detail:

• •A few hundred scouts search independently over an area of many square kilometres
• •A scout finding a candidate returns and advertises it by dancing on the swarm surface
• •The vigour and duration of the dance reflect how good the site is
• •Other scouts visit advertised sites and assess them independently rather than taking the dancer's word
• •Support accumulates for better sites and dancers for poorer ones stop advertising
• •When enough scouts agree on one site, the swarm takes off and flies to it

Scouts measure specific properties of a cavity and the criteria have been established experimentally by offering boxes that differ in one variable at a time. Volume matters, with around forty litres preferred, since smaller cavities cannot hold enough stores to survive winter and larger ones are hard to keep warm. The entrance should be small, positioned low on the cavity and facing south in northern latitudes. Height above ground is preferred, which reduces predation. The cavity should be dry and should ideally have held bees before, since residual comb and odour save work. A scout assesses a cavity by walking over its interior surfaces and flying within it for many minutes, which is how it measures volume, and the accuracy of those assessments has been demonstrated directly.

Losing half a colony costs a beekeeper the season's honey, so preventing the departure is a central part of the craft and the methods follow from the biology. Inspecting for queen cells every week or so during the season catches the preparation before the queen leaves, since the cells are built well in advance and are conspicuous. Giving the colony more room reduces the crowding that triggers it. Removing the queen with a few frames into a separate box, which is an artificial swarm, satisfies the impulse by imitating what the bees were about to do. Collecting a cluster that has already left is straightforward, since a hanging swarm can usually be shaken into a box and will stay if the queen is inside.

The mechanism has attracted attention well beyond beekeeping because it solves a hard problem reliably. No bee compares the options, since each scout knows only the site it visited, and no bee has authority. Quality is converted into advertising effort, and independent verification by other scouts prevents an enthusiastic individual from carrying the decision. Support for poor options decays because their advocates stop dancing. The threshold for action is reached only when enough independent assessors agree. Research on the process, principally by Thomas Seeley over several decades, has identified it as a case of a decentralised group reliably choosing the best available option, and the parallels drawn with human group decision-making are frequently overstated and genuinely suggestive.

Swarming is a colony reproducing, with the old queen and half the workforce leaving while the remainder keeps the comb and raises a new queen. Scouts assess cavities independently and advertise by dancing with vigour reflecting quality, and other scouts verify rather than accepting the recommendation. Around forty litres with a small low entrance is what they prefer.