In 2010 three authors—Martin Nowak, Corina Tarnita, and E. O. Wilson—published a paper in Nature (reference and link below) purporting to explain the evolution of eusociality in insects: the phenomenon whereby a colony contains different “castes” that perform different tasks, and at least one caste is sterile. In bees, for example, there is usually a single fertile queen, who produces all the offspring, a bunch of sterile working females (“workers”) who defend the nest and tend the brood, and fertile male “drones” who do basically nothing but compete to mate with the next generation of queens. Nonreproductive castes are general (though not ubiquitious) in the Hymenoptera (ants and wasps), as well as in some other animals like termites (Isoptera) and naked mole rats.
It’s hard to understand how it could be advantageous for some individuals to evolve sterility, which, of course, seems patently maladaptive. Darwin was the first to notice this problem. At any rate, one solution involves the notion of kin selection: the idea that a gene can promote sacrificing your own personal reproduction if it more than compensates for that loss by increasing the number of relatives you have—relatives who also carry copies of the “sacrifice” gene. It turns out that under a simple calculus that involves weighing the reprodutive benefit to relatives (discounted by the degree of relatedness) versus the cost of sacrificing your own reproduction, you can indeed evolve genes that cause you to lose reproductive ability—so long as they increase it in your relatives.
Such kin selection was an important explanation for the evolution of eusociality. Some think it’s because of the peculiar “haplodiploid” nature of inheritance in Hymenoptera, whereby the male who fertilizes the queen is haploid (has only a single set of chromosomes), and the fertile queen is diploid, with the normal two sets. In such a case, the female workers are more related to their sisters than to their own offspring, which may help them evolve the tendency to stop having their own offspring and produce more sisters; i.e., become sterile and help the queen raise her brood. Others question the importance of haplodiploidy in eusociality.
But the evidence for kin selection and relatedness is still clear. For example, eusociality in Hymenoptera has evolved several times, but always occurred in an ancestral lineage in which queens mated singly rather than multiply: a statistically significant finding (Hughes 2008; reference and free link below). That’s important because in such cases offspring are more related to each other than offspring produced by different fathers. Further, Bob Trivers showed that other patterns in bees, ants, and wasps—especially the observed ratios of males to reproductive females in colonies—also followed the dictates of what kin selection predicted. There are still other behavioral recognition experiments of kin versus non-kin supporting the importance of relatedness.
Nowak et al.’s paper, however, attacked this body of knowledge, claiming that kin selection and relatedness were not only unimportant in the evolution of eusociality, but were unimportant in general. (Ed Wilson has spent the last five years, for instance, arguing in his books and talks that kin selection is a misguided notion in evolutionary biology, and that group selection is far more important.) Nowak et al. used a rather complicated model of colony selection in which mother and all offspring were genetically identical and argued that relatedness was a consequence of the evolution of eusociality and not a driver of eusociality.
As I noted at the time, their dismissal of relatedness and kin selection from their model seemed bizarre, since they didn’t vary relatedness in their model. If you don’t do that, how can you say it’s unimportant in evolving eusociality? And people in the field found other problems with both Nowak et al.’s model and their conclusions about the uselessness of kin selection (go here for all my many posts on this issue). Over 120 experts in the field, for example, wrote a letter to Nature criticizing the conclusions of Nowak et al. But Nowak, Tarnita, and Wilson have remained obdurate.
Now, a new paper in PLOS Biology by Xiaoyun Liao, Stephen Rong, and David Queller (reference and free link below) shows not only that the model of Nowak et al. was bizarre, with little obvious relationship to the evolution of eusociality, but was also flatly wrong its three major claims about the evolution of eusociality.
What Liao et al. did was simply vary the relatedness and other assumptions of Nowak et al.’s model. After doing that, they found that the original authors’ claims about the generality of their model were incorrect. Keep in mind that Liao et al.’s conclusions were based simply on manipulating the very model that Nowak et al. used to claim the irrelevance of kin selection, or on deriving new equations using Nowak et al.’s exact modeling strategy.
Here is what Liao et al. found, contra Nowak et al.:
1. Relatedness does help the evolution of eusociality, so kin selection is not irrelevant. Unlike Nowak et al., Liao et al. varied total relatedness by allowing a certain fraction of offspring in the nest to be unrelated to the “queen” rather than simply her clones. (This could occur by immigration of insects from other nests, or by queens laying eggs in other queens’ nests.) What they found is that relatedness indeed makes a big difference: under conditions in which worker behavior is affected by their own genes rather than just the queen’s, eusociality evolves much more easily when relatedness between queen and “worker” is higher. In other words, higher relatedness (kin selection) is causal in this circumstance, not just a consequence of the evolution of eusociality. Nowak et al. were wrong, and all the statements of this group about the uselessness of kin selection based on this model are also wrong.
2. Workers’ evolutionary interests can differ from those of the queen. Nowak et al. saw workers as “evolutionary robots” who could not have an evolutionary strategy differing from that of the queen. This was a bit weird, since we’ve known from experimental and natural history data that there is a conflict between queen and workers (predicted by kin selection theory), and the results of that conflict are seen in the sex ratios produced in nests. Further evidence for that conflict occurs when the queen produces more males than the workers “want” (i.e., more than is in the workers’ genetic interests), and in that case the workers kill those males. This wouldn’t happen if the interests of queens and workers were coincident.
Liao et al. showed that varying how the genes for worker behavior and sterility were expressed—whether in queens alone or in offspring alone—had a substantial effect on the probability of evolving eusociality. As kin-selection theory predicts, eusociality evolves much more easily when the queen has absolute control over the behavior (and fertility) of her workers. When workers get a say—that is, when their own genes rather than just the queen’s genes control their own behavior—it’s not so easy to evolve eusociality, for workers sometimes have an evolutionary impetus to produce their own offspring rather than just raise the queen’s. This shows, as kin selection predicts and observation shows, that there is indeed a conflict between the interests of the queen and the workers.
3. Eusociality isn’t as hard to evolve as Nowak et al. assert. In their paper Nowak et al. claimed that eusociality is “hard to evolve.” It’s difficult to evaluate this claim because you have to ask, “Hard relative to what?” But Liao et al. showed that some of the difficulty in the Nowak et al. model is because of two wonky assumptions: 1). Below a certainly colony size no worker can add anything to the offspring production of the colony, while 2). above that threshold there is a fixed output of offspring that does not change with the addition of more workers.
This seems completely unrealistic, for why wouldn’t two workers add more offspring than one, and why, over the threshold, wouldn’t more workers help produce more offspring by defending the nest better and tending more brood? Further, why would workers in large colonies remain in those colonies, since their presence adds nothing according to the “threshold” model? They should, instead, join smaller colonies, pushing them over the threshold. And indeed, when Liao et al. added more realistic assumptions to Nowak et al.’s model—a “stepwise” feature whereby, up to some limit, each worker adds an increment to the offspring production of the colony—eusociality evolved more readily.
So what is the upshot? First, that kin selection, i.e., the relatedness between queen and her offspring, plays an important causal role in the evolution of eusociality. Nowak et al. were dead wrong in denying this. And since the subsequent statements of both Nowak and Wilson on the evolutionary worthlessness of kin selection were based on a model that could not show what they claimed to show (because relatedness wasn’t allowed to vary), we should not take their dismissal of kin selection seriously. Kin selection remains a viable and valuable view in evolutionary biology—indeed, one of the most important advances since the 1950s—and, as I’ve shown in my earlier posts, has helped us understand a wide variety of biological phenomena.
The other points are less important, but still show that Nowak et al.’s model was too narrow to support their generalizations about no conflict between queens and workers, or about the “difficulty” of evolving eusociality. Yes, it may indeed be hard to evolve such a bizarre system, and it may require uncommon ecological and/or genetic circumstances, but it’s not as hard, at least in theory, as Nowak et al. maintained.
The final lesson is that one’s biological conclusions from a model are only as good as the biological assumptions built into it. Because Nowak et al.’s assumptions were flawed, and because they failed to examine the robustness of the model to varying its assumptions, they arrived at faulty conclusions. But because Nowak and Wilson were already famous evolutionary biologists (particularly Wilson, who is an iconic figure in the field), and because the paper was published in Nature, their conclusions were taken far too seriously. The paper should have been reviewed by more critical reviewers in the field. Even I, who do not work on the evolution of eusociality, could see that you can’t dismiss the value of genetic relatedness from a model in which relatedness isn’t allowed to vary!
Hughes, W. O. H., B. P. Oldroyd, M. Beekman, and F. L. W. Ratnieks. 2008. Ancestral monogamy shows kin selection is key to the evolution of eusociality. Science 320:1213-1216.
Liao, X., Rong, S., and D. Queller, 2015. Relatedness, conflict, and the evolution of eusociality. PLOS Biology | DOI:10.1371/
Nowak, M. A., C. E. Tarnita and E. O. Wilson. 2010. The evolution of eusociality. Nature 466: 1057-1062.