Feud over family ties in evolution

A furor has broken out among biologists over ant specialist E.O. Wilson’s latest attack on a concept used to explain the origins of self-sacrifice in the dog-eat-dog world of evolution.

The debate centers around an idea called kin selection, which biologists use to understand altruistic behaviors such as honeybee workers raising the queen’s young but never having their own. These selfless workers would seem to lose out in the evolutionary struggle to pass along genes to the next generation. But according to the idea of kin selection, workers without young more than compensate by sharing in the reproductive success of relatives, with whom they share genes.

In the Aug. 26 Nature, Wilson and two Harvard colleagues argue that the concept of kin selection is “limited” and “unnecessary.” And they propose steps for the evolution of ants, honeybees and other highly social species with such altruistic behaviors by just the broad “survival of the fittest” forces of natural selection without specifically invoking the power of kinship.

In recent years, Wilson has argued that the close family ties in ant colonies and other highly social groups may be consequences, rather than causes, of the evolution of such extreme social forms. In the new paper he combines his perspective with two coauthors’ mathematical critique of the methods used to calculate kinship effects, arguing that the techniques are as unnecessarily complicated as Ptolemaic astronomy.

“Babylonian astronomers look up in the heavens and they see the planets moving in ‘epicycles,’” says paper coauthor and mathematical biologist Martin Nowak. “But if you put the sun in the center, there are no epicycles.”

Some kin selection adherents are firing back that, even with new math, the challenge itself is old-fashioned. “This is such a tired old debate,” says Ben Oldroyd of the University of Sydney, who studies social insects.

The new analysis attacks the core of kin selection by examining how biologists calculate what’s called inclusive fitness. Evolutionary biologists measure fitness not by push-ups but by progeny, and inclusive fitness counts not just an individual’s direct offspring but some share of relatives’ youngsters. The closer the relative and the more the altruist helps to raise the young, the greater the share that counts.

Looking over decades of papers calculating inclusive fitness, coauthor and Harvard mathematician Corina Tarnita says that she was surprised to find no rigorous mathematical analysis had been done to assess the conditions under which an inclusive fitness calculation can be used. “I find that very dangerous,” she says.

Nowak and Tarnita did such an analysis for a hypothetical population in which organisms use either of two strategies (which could be genes or behaviors, like cooperation). Mathematically, the researchers argue, inclusive fitness works only under very limited conditions, such as when two strategies occur at nearly the same numbers in the population, and when interactions among individuals occur straightforwardly in pairs.

Even in those special circumstances, Nowak says, mathematical approaches like those used in game theory can predict the same outcomes without using inclusive fitness.

Nowak clarifies that the team isn’t arguing that kinship is irrelevant in biology. “Relatedness does matter,” he says. What he and Tarnita are challenging is the accounting method of inclusive fitness, which he contends is overly complicated.

Michael Doebeli of the University of British Columbia in Vancouver says the analysis is “a very welcome attempt at clarifying several key issues in the theory of social evolution, yet it will surely take time before its conclusions will be embraced by more traditional evolutionary biologists.”

But others, like evolutionary biologist Andrew Bourke of the University of East Anglia in Norwich, England, disagree with the new argument. Bourke laments that the researchers take “a very narrow view of inclusive fitness theory, focusing on details of algebraic expression of the theory at the expense of its essential insights.”

Whether kin selection has indeed been useful is one of the more contentious arguments of the new paper. The authors dismiss most earlier work as finding mere correlation between kinship and social behavior without strict analysis of costs and benefits. “We can find no case that presents compelling evidence for the explanatory adequacy of kin selection and inclusive fitness theory,” they write.

As an alternative, Wilson, Nowak and Tarnita describe how extreme societies based on nonreproductive workers could evolve without invoking kin selection. Organisms might cluster at first for any reason, such as a good feeding site, and then accumulate traits that eventually prove important to a social structure, such as building defensible nests. Eventually crucial colony-building behaviors could evolve, such as adult children sticking around the home nest instead of striking out on their own.

The key argument in this scenario, says William Hughes at the University of Leeds in England, is that cooperation between individuals that share a gene vital for extreme social living would be enough to spur the evolution of altruism.

Hughes objects that previous studies have considered this scenario but found such genes vulnerable to cheating or to conflicts with other parts of the genetic instruction book.

“Previously I always had a nagging concern that maybe Wilson had thought of something that everyone else had missed,” Hughes says. “Having read this paper, I’m now quite confident that’s not the case.”

In all the furor, however, Tarnita says she’s looking forward to the comments she might hear several weeks after the publication. By then, she says, people should have had time to do the math.