More about social insects and their decision-making and evolution

There are four main kinds of eusocial insects: ants, termites, wasps, and bees. The word "eusocial" refers to an organization in which only one individual, the "queen," performs the vast majority of all reproduction, while the others perform non-reproductive duties. The entire hive is thus a super-organism.

Asterisk: bees depart slightly from the ideal of being a superorganism because not all worker bees are forever sterile, and a very small percentage of the reproduction is done by workers. If a fertile worker is caught laying eggs, she is killed and her eggs are eaten! This kind of "policing" behavior could only have evolved if worker reproduction was once much more common than it is now – so earlier bees were "less superorganismic."

Nest-relocation decisions are not made by a large community in the ants, termites, and wasps: instead the queen acting alone founds the new colony (albeit in the case of termites, in collaboration with a single "king"). So in those cases there is no need for "voting." Only bees vote.

NEWS FLASH! Sorry, that conclusion was premature. It turns out certain ants also will vote on a hive relocation decision if their original nest site is damaged. (Might the same perhaps be true for other social insects?) And amazingly, these ants also appear to use range voting.

There are thousands of species of bees. Most are not social. The bumblebees are social but only in a comparatively small-scale way (maximum hive population 50-400 usually) and their new hives are started by a single queen acting alone – just like ants and wasps – and last only 1 year.

There are two main types of sophisticated eusocial bees, which live in large hives with tens of thousands of inhabitants and multiyear tenancy: the honeybees and the stingless bees. The stingless bees appear older (a trigona worker fossilized in amber was found that is 74-94 million years old) and are more diverse (about 200 species known). The honeybees (apis) seem younger (oldest known fossil 22-25 million years old) and less speciated (about 10 species known). The stingless bees were the only kind native in Australia and South America until humans introduced honeybees from Europe and Africa, and they now appear to be faring badly in competition with honeybees.

Note that the fact that those fossilized bees were sterile worker bees strongly indicates they had already become eusocialized. Flowers appeared on the scene 120-140 million years ago. Hence the (co)evolution of modern eusocialized bees happened in 50 million years or less.

For economic reasons, the honeybees are much more studied than the stingless bees. Honeybees produce much more honey, build quite different kinds of hives (involving wax honeycombs), are anatomically rather different (have stingers), and exhibit considerably different behavior. Honeybees communicate by a "dancing" sign language. The stingless bees are also known to communicate (for at least some species), but do so by olfactory or auditory means which have not yet been decoded by humans.

Honeybee hives "fission" into two, with the new "swarm" picking its new nest site a considerable distance (typically several miles) away by means of a "scouting and voting" collective decision-making process that is nowadays well understood, and is in fact almost equivalent to the CRV's recommended form of range voting.

Once the honeybee swarm has left the old hive, it is a total divorce; they do not return. The old queen moves to the new nest-site (and construction of the new nest by the swarm's workers commences) while new virgin queens fight to the death for control of the old nest.

The stingless bees, in contrast, build new hives fairly nearby, with the new hives being occupied by a new virgin queen after its construction by workers. There is a great deal of interaction between the new and old hives, including exchanged honey and materials, for a period of several months. The process of determining where to locate the new hive also appears to be a collective one involving scouts eventually settling on one choice, but the manner in which that consensus is reached is not presently understood.

An open question: It is generally believed that the honeybees did not evolve from the stingless bees but rather independently from some other (asocial) bee type (and this speculation should be provable from DNA evidence); and that any social features the two now share are the result of convergent rather than divergent evolution. If so, then the stingless eusocial bees are pretty much an independent Darwinian experiment and so it would be very interesting to understand what voting method the stingless bees independently invented. Is it also range voting? Is it something else? Is it a single-winner method at all (or will stingless bees settle into two new homes with two new queens generated for the occasion if both look good enough)?


Henry R. Hermann (ed.): Social Insects (4 vols.) Academic Press, New York 1979-.

C.D.Michener: The social behavior of the bees; a comparative study, Belknap Press of Harvard University Press, Cambridge MA 1974.

J.Philip Spradbery: Wasps; an account of the biology and natural history of solitary and social wasps, University of Washington Press, Seattle 1973.

E.O.Wilson: The insect societies, Belknap Press of Harvard University Press, Cambridge MA 1971.

Bees and range voting

Return to main page