Since January in this column I have been telling an evolutionary history of the honey bee we all know and love. In June I explained that daughter bees even in highly-related families can still be predicted to act selfishly, that is, lay their own eggs and abandon work for the colony. But this kind of selfish behavior is controlled and group cohesion encouraged by enforced coercion – the mother eating the eggs of “outlaw” daughters. Enforced coercion is not a bad deal for daughters because each can still, on average, pass on 50% of her genes by helping their common mother produce more siblings. It was enforced coercion that freed our primitive honey bee colony from internecine conflicts and allowed the appearance of stable altruism seen and marveled at by beekeepers to this day. More to the point, this pax apis freed daughters and mothers to follow divergent evolutionary paths, one focused on nest tasks and the other on laying eggs. Eventually, the one could no longer perform the duties of the other; the two had become two separate castes. It is only after this point that we can rightly distinguish the two as workers vs. queens.
Next, I spent July and September talking about the evolutionary unit of selection which I defined as an entity at any level of biological hierarchy that can be acted upon by natural selection. A unit of selection can be as simple as a gene or as complex as a superorganism – and maybe beyond. As long as it has characteristics that land somewhere along a continuum of “fit” or “unfit,” if those characteristics are heritable, and if the entity can reproduce, it’s in the club. I also talked about the tendency for lower levels of biological organization to incorporate into higher levels: genes get packaged into chromosomes; chromosomes get bundled into cells, cells into organisms, organisms into superorganisms. The stability of each higher level depends on eliminating or suppressing the competing interests of levels below it. When the genomes of the lower and higher levels become sufficiently harmonized that the emerging entity is stable, physically contiguous, and internally coordinated, we witness the emergence of a new level of individuality* in a narrow biological sense of the word. When this stability falters we see conflicts within the genome, whether it’s cancer cells that rebel against other cells or sister queens that fight to the death in a honey bee colony.
Now in the case of social insects, the concepts of caste evolution and units of selection come together in a dramatic way when we consider the fact that the queen cannot reproduce without the workers, and workers cannot reproduce without the queen. Neither is reproductively autonomous; neither is a proper unit of selection. It was ruminating on this problem that led Harvard entomologist William Morton Wheeler in 1928 to coin the term “Superorganism” to explain the idea of a colony as the unit of selection, not the members that make it up.2 Among Wheeler’s writings on the subject, it is clear he understood several characteristics of a social insect colony to qualify it as an organism:
- It behaves as an integrated unit.
- It is similar enough to other such units to qualify it as the same
- It is different enough from other units to qualify as an individual (in the general sense of the word) of that species.
- It undergoes a period of growth and reproduction.
- It possesses “cells” specialized for reproduction (queens and drones) and “cells” differentiated into the various requirements of the body (workers).
The superorganism concept has had an up and down history. For a few decades in the mid-20th century it enjoyed a period of prominence, but by 1971 E.O. Wilson3 could dismiss it as a “mirage” rendered irrelevant by a younger generation of scientists bent on reductionist approaches in preference to overarching systems and starry-eyed with the new and unfolding powers of kin selection theory (see March and May 2015 columns). At that time it seemed that all biological adaptation could be explained at the level of the gene. But the pendulum began swinging back in 1989 when David Wilson and Elliott Sober4 pointed out that there’s no logical demand that genes-in-individuals be treated more prominently than individuals-in-groups and that superorganisms are not just theoretical constructs but really do exist in nature. In that same year Tom Seeley added that “It seems correct to classify a group of organisms as a superorganism when the organisms form a cooperative unit to propagate their genes, just as we classify a group of cells as an organism when the cells form a cooperative unit to propagate their genes5,” and he linked the emergence of superorganism to the suppression of genomic conflicts among its constituent parts: “. . . it appears that ….