Reviving group selection

Group selection

"Group selection" refers to differential reproductive success of groups.

It is the idea that groups of organisms vary naturally - and exhibit differential reproductive success (in terms of spawning new groups) on the basis of the traits of the parent group.

Group selection - problems

Habitat-specific selection is of some interest since it can help make models of group selection more viable.

Differential reproductive success of groups consisting of sexually-reproducing individuals depends on the groups maintaining their own identity in the face of sexual recombination constantly stirring groups together.

Partial group isolation and population viscosity have been invoked to explain how groups could maintain their individuality in the past.

However, theoretical considerations have indicated that the rate of between-group migration would have to be very low - or - equivalently - population viscocity would have to be very high in order to make the effect work - and observations of actual groups of individuals suggest such partial-isolation conditions are rarely met in practice.

Group selection - solutions

A simple observation puts quite a spanner in this theoretical argument. The above argument is essentially asserting that sexual groups don't vary from place to place - because the stirring effect of sexual recombination is too great for this to happen.

However human races present a fairly clear counter-example to this - where there exists very significant geographical variation in the human population.

Where did the analysis go wrong?

Several effects make group selection more viable than the analasis above might suggest.

Among them are:

  • Habitat-specific selection;

  • Xenophobia;

  • Divergent selection;

These effects act together synergetically:
  • Habitat-specific selection acts to generate between-group variability.

  • Xenophobic tendencies act to amplify existing variability between groups, and to reinforce group boundaries.

  • Divergent selection arises when hybrid offspring between members of the two groups have reduced viability.

To consider the effects in turn:

  • Habitat-specific selection

    Habitat-specific selection arises when a single population is exposed to a range of different environments - and is subjected to different selection pressures in the different locations.

    Selection on the basis of environmental properties is likely to result in adaptations to local environments.

    This effect can be seen clearly in the case of the manner in which human skin colour is associated with lattitude.

  • Xenophobia

    It has been previously suggested (by Pagel and Mace) that xenophobia is an adaptation to facilitate the existence of cooperative groups. Distrust of outsiders has the effect of biasing favourable interactions towards previous givers of altruistic events - and results in a healthy distrust of outsiders.

    There may also have been other reasons to distrust outsiders historically. For one thing an outsider is normally less likely to be a cousin than someone in an agent's existing social group.

  • Divergent selection

    Divergent selection arises when hybrid offspring between members of the two groups have reduced viability. That could happen if the offspring contained a mixture of traits - and winds up not being well adapted to the habitat of either of its parents.

    Such selection pressure typically causes females of one group to reject the advances of males not of their own group.

    Divergent selection is normally a form of sexual selection. It can happen due to females rejecting the advances males from other groups - in order to avoid having to invest in "half-breed" offspring - that do not fit into either group - or who may have reduced fertility for other reasons.

    Divergent selection reduces the effect of sex in stirring the population by reducing gene flow across group boundaries.

    It invalidates the assumption that group membership does not influence breeding prospects.

Causes of habitat-specific selection

Different habitats may contain:

  • Different sources of food;

  • Different predators;

  • Different competitors;

  • Different pathogens;

  • Different temperature, humidity, illumation - and other physical factors;

Such factors can create significantly selection pressures in different environments.

For example, an organism that eats slow moving sedentary creatures on one continent - but is forced to chase fast- moving creatures on another one may develop different adaptations in the two environments do deal with the different problems - despite the two populations remaning genetically linked by gene flow between them - with the differences being continually reinforced by selection.

Note that groups adapted to particular prey can be obliterated when their prey goes extinct.

Similarly, groups which have lacked recent exposure to a particular predator may be obliterated if the predator invades their environment.

Environmental inheritance

Habitat-specific selection is made possible because of the effect of environmental inheritance.

As well as inheriting genes from their parents, organisms also inherit their environment.

The environment contains a lot of information - far more than is in any organism's genome.

However the environment is not usually static. Mobile organisms can uften easily move to new locations - and when that can happen the significance of environmental inheritance is reduced.

Environmental inheritance can expected to be powerful when movement to new environments is more difficult.

For example, if the organisms are land-dwellers living on an island, the chances that they will migrate to a different environment from the one their parents inhabited decrease.

There's a similar effect when environments are large, compared to the distance organisms can travel. Latitude is often strongly inherited for this reason, along with its associated weather patterns.

Genetic model

Consider a model where habitat-specific selection acts on one gene. Different alleles can be expected to be favoured in their respective environments - and will be found preferentially in them.

However other loci are likely to be less strongly affected. The effect on them will be due to linkage with the affected genes. Genes on other chromosomes are likely to escape any effect - while genes nearby on the same chromosome are likely to be most affected.

In practice, traits are often complex - and are associated with more than one locus. Environmental differences may create a whole suite of associated adaptations - with a large number of genes being involved. The genes involved may well be numerous - and spread across many chromosomes.

In that way, selection for a particular suite of traits by the environment may create selection pressure that affects practically every locus - via linkage.


I don't expect these considerations such as these to lead to a radical revival of group selection. However, group selection does seem to me to have more going for it than many suspect, and does seem to have suffered from some unfair public relations in the past.


  1. Steve Connor - Racism and xenophobia linked to biological fear of outsiders in Stone Age

Tim Tyler | Contact |