For the Love of Bees and Beekeeping

July 2015

Things Are Getting Messy: What’s the Unit of Selection?

by Keith Delaplane


For the last few months I’ve been unpacking a natural history of the western honey bee Apis mellifera with the aim of pointing out clues for modern beekeeping along the way. The practical clues have been few to this point because we’ve been dealing with the most fundamental groundwork of the species – its morphology and social structure. But in the months ahead, as we follow our bee out of southeast Asia into temperate Europe, things will start looking more familiar. Something that looks like the modern honey bee will begin coming into focus. Hopefully by the end of the exercise we’ll have a better empathy for the bee’s priorities and its peculiar struggles. More to the point, we’ll be better managers of those honey bees we have taken into our own care.

In Figure 1 I’ve summarized our journey to this point. In January I gave a primer of taxonomic classification of the modern honey bee, its natural range, and the evolutionary processes of natural selection and speciation – the birth of new species from parent species. In February I explained the criteria for eusociality or colonial life and talked about the emergence of the first bees from their wasp ancestors in synchrony with the first flowering plants in the Cretaceous epoch. In March we faced the evolutionary puzzle of the self-sacrificing worker and showed mathematically how worker altruism makes sense if it permits the worker to pass on more of her genes to the next generation. In April I took a break from genetics to talk about nesting – the ability to make a nest and the ability to navigate and return to it – and underscored that this ability is a prerequisite to social life. In May I returned to the genetics of social evolution and noted that the most highly-related colonies result when a mother is mated to one male and her daughters stay at the nest to help her rear more siblings; these are, in fact, the necessary conditions for the emergence of eusociality. In June I pointed out that even with the high relatedness possible with singly-mated mothers that some of the daughters can still be counted on to act selfishly. “Selfish” in this sense means a worker choosing to produce her own sons and abandon work for the colony. What keeps this from happening on a large scale is coercion – the queen or other workers eating the eggs of workers. In this way the “outlaws” who would exploit the system are checked – but checked in a generous kind of way because if they play by the rules they still get to pass along 50% of their genes via helping their mother rear siblings. Once coercion settled the problem of altruism, we left off last month with the way cleared for the evolution of specialized castes – the egg-laying mother (now properly called a queen) and her task-oriented daughters who gradually lost the ability to mate (now properly called workers).

This is where we pick up this month. And this is where we must introduce the idea of the evolutionary unit of selection. In its broadest sense a unit of selection is an entity at any level of biological hierarchy that can be acted upon by natural selection. To quote Richard Lewontin.1

“The generality of the principles of natural selection means that any entities in nature that have variation, reproduction, and heritability may evolve.”

Lewontin goes on to say:

“This . . . makes clear that the principles can be applied equally to genes, organisms, populations, [or] species . . .”

To restate things, to qualify as a unit of selection the entity must possess (1) variation – ie., be one unit among many possessing a range of “fit” or “not fit” characteristics, (2) the ability to reproduce itself, coupled with (3) the ability to pass along its heritable fit characteristics to the next generation.

Now traditionally, the unit of selection has been thought to be the organism, the likes of you and me. Organisms are unique packet of genes contained in contiguous organic systems of one or more cells. Organisms are self-regulating, capable of reproduction, interact with and respond to their environments, and usually undergo individual growth and development. Species are made up of populations of organisms. It was organisms that Charles Darwin was thinking about when he put together his theory of natural selection. To Darwin, it was the organism that was fit or not fit: the faster rabbit, the stronger tree, the taller giraffe, the peacock with the brightest feathers. But as I mentioned in March, Darwin knew things were a little different when it came to the social insects and their “neuters or sterile females2” and he came presciently near getting it right when he said:

“. . . selection may be applied to the family, as well as to the individual, and may thus gain the desired end.”

Thus, by more than a century Darwin anticipated Lewontin’s exposition of the different levels at which natural selection may occur.

I diverged to talk about units of selection because at this point in our evolutionary history of the honey bee we can begin asking What exactly is the unit of selection? Is it the workers? Well maybe, but even though they are quick to lay eggs when the forces of coercion relax, the vector of their evolution has been toward less, not more, fecundity. Is it the queen? This too has problems because the queen is utterly reliant on the workers. The truth is, neither the worker nor the queen meets Lewontin’s criteria as a unit of selection on her own. Neither is reproductively autonomous. Each is mutually dependent. This insight made ...