Social insects are often considered prototypes of group selection, in which the evolutionary interests of individual organisms are forced to defer to the needs of their social group. Now, the authors of a new study of honeybees argue that colony-level traits can override the genetic predispositions of individual bees [$-a].
Do the needs of the many outweigh the needs of the one? Photo by Max_xx.
The study’s authors, Linksvayer et al.,
made use of artificially-selected colonies of bees that were first developed for a 1995 study [$-a]. The original selection experiment crossed queen bees with drones to create lines of honeybee colonies that collected and stored more pollen (“high pollen” lines) or less pollen (“low pollen” lines) than un-selected colonies do. The total amount of pollen a colony stores is supposed to be a “group” trait — an emergent property of the individual foraging decisions of every worker bee in the hive. But the genetics underlying that trait is encapsulated within the individual workers.
In the new experiment, Linksvayer et al. placed larvae from “high pollen” lines in “low pollen” colonies, and vice-versa. The larvae developed under the care of workers from the adoptive colony; when transplanted larvae reached adulthood, the team dissected them and measured the size of their ovaries — apparently big-ovaried workers collect lots of pollen. They found that “high pollen” larvae reared by “low pollen” workers had smaller ovaries than than those raised by workers of their own type. “Low pollen” larvae reared by “high pollen” workers didn’t end up with larger ovaries, though; and the “high pollen” larvae had substantially larger ovaries than the “low pollen” larvae regardless of who raised them.
There was a statistically significant effect of rearing environment, even if it was (apparently) entirely driven by the change seen in “high pollen” larvae. The authors conclude that this points to a mechanism whereby a bee colony keeps its workers in line with the colony-wide policy:
Thus, our results show that the network of social interactions that shapes development and expressed phenotypes has changed as a result of the colony-level selection program on pollen hoarding. Just as selection shapes physiological networks within organisms, our study shows that selection also shapes regulatory networks of superorganisms.
So the metaphor, then, is that the authors have observed in the hive something like what happens to a transplanted organ — the new host system incorporating the transplant for its own needs. I’m not sure the observed effect is strong enough to justify the meaning they assign to it; but it is an interesting observation.
As a postscript, I’m not sure social insects are a good model of group selection, because we know that they’re probably also experiencing kin selection, in which each worker’s fitness comes from helping the closely-related queen produce more sisters who share the same genes. Rarely, “anarchic” workers are born fertile and mate with drones [$-a] (there’s an open-access paper on the genetics underlying this trait); but in hives without anarchists, “group fitness” is hard to separate from the fitness of individual workers. A paper published in Nature this June showed that in another classic group selection system (parasites within a single host) kin selection is really the more important process.
Linksvayer, T., Fondrk, M., & Page Jr., R. (2009). Honeybee social regulatory networks are shaped by colony-level selection. Am. Nat., 173 (3) DOI: 10.1086/596527
Oldroyd, B., Smolenski, A., Cornuet, J., & Crozler, R. (1994). Anarchy in the beehive. Nature, 371 (6500) DOI: 10.1038/371749a0
Oxley, P., Thompson, G., & Oldroyd, B. (2008). Four quantitative trait loci that influence worker sterility in the honeybee (Apis mellifera). Genetics, 179 (3), 1337-1343 DOI: 10.1534/genetics.108.087270
Page, R., & Fondrk, M. (1995). The effects of colony-level selection on the social organization of honey bee (Apis mellifera L.) colonies: colony-level components of pollen hoarding Behavioral Ecol. & Sociobiol., 36 (2), 135-44 DOI: 10.1007/BF00170718