Evolution experts in heated row over recent theory


“To borrow from Dorothy Parker, this is not a book to be tossed lightly aside. It should be thrown with great force,” wrote Richard Dawkins in a scathing review for Prospect magazine of Edward O. Wilson’s 2012 book The Social Conquest of Earth. Two of the most influential behavioural biologists of our time – Richard Dawkins, well-known for his popular books on the mechanisms of evolution and opinionated views on religion, and Edward O. Wilson, world expert on ants and commonly referred to as the ‘father of sociobiology’ – held a vehement argument on the evolution of social behaviour a couple of years ago. Recent comments made by Wilson during an interview on BBC2’s Newsnight have rekindled the altercation. It is forgivable to interpret the exchange as a spat between two veterans of the field, each set in their ways. However, Wilson and Dawkins are not alone, and both present compelling arguments for a long-standing debate.

The topic at hand is the evolution of eusociality, a special case of animal cooperation that involves division of labour – picture a colony of ants: there is a queen along with workers, soldiers etc. Crucially, the lower ‘castes’ in the colony are usually sterile, so this poses a Darwinian puzzle: how is it possible that non-reproductive individuals endure through evolutionary time if natural selection can only favour those who pass on their genes? The answers to this question have generated conflicting opinions among evolutionary biologists since the early twentieth century.

Dawkins, like many others, defends the concept of ‘kin selection’. Evolutionary theory tells us that for a trait, like cooperation, to become more common it must increase the probability that the genes responsible for it are passed on to the next generation. We all think of ‘passing on genes’ as having children, but the truth is that this is not the only way: if you help someone else who shares your genes to have children, you are indirectly enabling (some of) your genes to live on in the next generation. J. B. S. Haldane, an important kin selection theoretician, famously joked that he would willingly die for two siblings or eight cousins.

Kin selection was tentatively laid out by Darwin himself in his 1859 book The Origin of Species, but the mathematical models to support it didn’t come until the following century. This has been the classical approach to answer the puzzle of eusociality because kin selection is consistent with Darwinian views of evolution that place the gene at the centre of selection, as popularised by Dawkins’ concept of the ‘selfish gene’. However, the theory is also difficult to back with empirical evidence due to the assumptions of the mathematical models it relies on.

Traditionally, the problem has been simplified by demonstrating a positive correlation between cooperation and relatedness of individuals, which is well documented. Wilson argues that this approach provides no proof of kin selection at work, as he believes that relatedness is a consequence and not a cause of eusocial behaviour.

His own theory, explained in a 2010 paper in Nature with two co-workers, is based on a ‘multilevel selection’ mechanism, in which natural selection acts on groups of organisms, not just on individuals. Group selection is not a new idea, but Wilson et al. have described a full mathematical model to account for advanced social behaviour using multilevel selection theory for the first time, replacing the widely accepted model based on kin selection, which Wilson argues is mathematically unsound and only applicable in a limited number of cases. This has not gone down well with everyone, including Dawkins, who believes the new theory is the product of a misunderstanding of kin selection.

In Wilson’s model, eusocial behaviour arises in a series of ordered evolutionary stages. Key to his theory is the idea that before becoming eusocial, a species must acquire a set of traits that will allow it, given the right conditions, to become eusocial. For example, a species of solitary bee may possess the necessary adaptations for division of labour but will not express them unless forced to live in partnership with other bees. Wilson et al. say that these individuals are ‘spring-loaded’, that is, they possess the potential to start living eusocially. Then all that is necessary is for a mutation to arise in the population of ‘spring-loaded’ individuals that will force them to live in permanent groups. This will ‘activate’ the adaptations for eusociality that they already possess. They suggest, for example, a mutation responsible for loss of wings in ants. Wingless ants cannot disperse from the nest and will remain to help their mother; if she no longer has to venture out of the nest to forage because her daughters do it for her, she will be able to rear more young and bypass the risk of predation, hence becoming the queen of a eusocial colony.

An important point of Wilson’s proposal is that the different roles of the reproductive parent and the non-reproductive offspring are not genetically determined. Rather, mothers and daughters share the genes for division of labour, but the resulting role depends on environmental factors. Thus, the whole colony can be viewed as a ‘superorganism’, and selection is considered to be acting on the group as a whole – in reality, it is acting upon the queen, but the workers are, at least partially, a projection of her genes.

Interestingly, Wilson et al.’s model predicts that eusociality is unlikely to arise in a solitary species, but will probably persist once it has appeared. Kin selection has been less successful at explaining why there are so few eusocial groups in the animal kingdom, despite the obvious advantages of eusociality. This model also accounts for the correlation between cooperation and relatedness, as individuals become related because they live eusocially, and not the other way around.

Much work has yet to be done in the field of social behaviour before either theory can be rejected, but the future of sociobiology promises to be an interesting one. For the time being, there is insufficient empirical evidence to conclusively support any model of advanced sociality, and biologists continue to argue from either side of the divide.

Photograph: USDA / Lance Cheung

2 thoughts on “Evolution experts in heated row over recent theory

  • “Wilson et al. say that these individuals are ‘spring-loaded’, that is, they possess the potential to start living eusocially.”
    Isn’t this begging the question, the question being why this potential to live eusocially should have evolved, especially because it is only a potential and so is not being tested by the environment for selection or otherwise?

  • Mike, the potential to live eusocially appears in these organisms as all new traits appear in nature: randomly by mutations and sexual recombination. In Wilson et al.’s paper the acquisition of “spring-loading” traits is written off as a case of “adaptive radiation” (http://www.nature.com/nature/journal/v466/n7310/pdf/nature09205.pdf).

    But you are right in questioning the use of the word “evolved”, as these traits have not yet been selected for at this stage – they have merely entered the pool of traits available for selection (natural selection acts on existing variability but does not create it). It is only when the organisms are forced to live together that these traits are favoured by selective pressures and true eusociallity appears.


Leave a Reply

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.