Inclusive Fitness, Population Genetics and the origin of altruism

Sparks fly over origin of altruism

When British geneticist J. B. S. Haldane was asked if he would risk his life to save another, he is alleged to have replied that he would only do so to save at least two brothers or eight cousins.

His reasoning was that this would preserve enough copies of his genes to justify his own death. This idea - that animals are more likely to show altruistic behaviour towards individuals they are related to, is called kin selection.

Haldane's colleague William Hamilton later drafted a mathematical description of the phenomenon, known as inclusive fitness, which assigns numerical values to the costs and benefits of an animal's actions. In theory, inclusive fitness makes it possible to calculate the extent of the spread of a given altruistic behaviour throughout a population e.g. staying with your parents to raise your siblings. 

Hamilton's mathematical formula has been used for decades by biologists studying cooperation in animals and was a major inspiration for Richard Dawkins's The Selfish Gene.

The problem, say Nowak and Tarnita, is that the calculations just don't work in the real world because they rely on a limiting set of conditions that nature does not stick to. For example, they are only valid for interactions between pairs of animals, which is fine for solitary species whose individuals rarely meet, but no use in studying thousands of ants sharing a colony. What's more, they do not work for populations that are under strong pressure to evolve.

These and other limitations, Tarnita says, mean that the maths of inclusive fitness is not relevant to the real world. Instead, she says biologists should use the models of population genetics, which focus on interactions between different gene variants. 

These models avoid the messiness of predicting the consequences of behaviour and don't require any dubious assumptions.

Tarnita has shown that by using standard population genetics equations, it is possible to produce an all-encompassing model. She explained that when she plugged Hamilton's conditions into her model, its equations simplified to those of inclusive fitness. 

Hamilton's maths, she concludes, describes a special case of a broader model of how all behaviours evolve: it is not wrong, but limited.