Species Senescence

Species selection

The existence of species selection is well established.

It is sometimes known as "species sorting".

Species - like genes and individual organisms - reproduce - by dividing into separate species. They undergo variation, and selection (extinction).

There are some questions about how strong a force species selection is:

Species reproduce more slowly than individuals.

Consequently fewer generations of species will pass, and such sophisticated adaptations cannot be expected.

Then there's the issue of what happens if individual selection and species selection conflict with one another. It would seem that if individuals benefit at the expense of the species they will come to spread through the population. No doubt this happens. However, note that species where this does not happen will have less chance of going extinct. Therefore we can expect species selection to attempt to arrange things so that either individual selection is rarely in conflict with species selection, or that it does not have the required variation to work on - since safeguards are so well built in.

One candidate for a species-selected trait is small size. Many lineages tend to grow with time. However, asteroid impacts tend to repeatedly systematically obliterate them. As a result it seems possible that the species that best resist the temptation to increase in size are the ones we can expect to see around.


Peter Medawar described a theory of senescence of organisms using an analogy involving a population of self-reproducing agents [1].

He showed that - if there was some finite probability death - such a population will mainly consist of young agents - i.e. at any time the proportion of very old agents present in the population will be small.

Therefore, genes that express themselves in young organisms will be more deleterious than later-acting ones.

Consequently mutations that delay the expression of deleterious genes will be selected.

In particular, genes that favour diverting resources into reproduction - and away from maintenance can be favoured - under some circumstances.

This leads to organisms that senesce - and age and die.

Species senescence

Species senescence is sometimes known by the technical name of phylogeronty. It is the theory that species - like organisms - tend to age and die.

In the light of how senescence evolves in organisms, we can see that species senescence is a logical consequence of species selection, if there is some finite probability that species will go extinct at random.

While we don't have much experimental evidence which bears on this question, it's hard to imagine randomness not playing some role.

For example the number of individuals in a species undergoes fluctuations with a random component. If the number becomes zero the species goes extinct.

More important factors that may result in species extinctions occurring with a strong random component include disasters - fires, meteorite strikes, floods, etc.

Also, predators, parasites attack species - and cause variations in population size that may sometimes be terminal. The older a species gets the more chance other organisms get to exploit it.


  • Aren't old species are less likely to go extinct than younger ones?

    It seems likely that an old species will have learned the ropes - and have a low chance of extinction - whereas a young one is likely to have few members with little chance of finding mates - and be likely to suffer from inbreeding. What gives?

    This situation is much the same with organisms. Children are more likely to die than adults. However this does not stop senescence from arising.

  • Species reproduce by dividing - thus aren't all species extremely old?

    If that were true, it would not significantly affect the theory. As it is, in practice, it is often possible to identify a "mother" species and a "daughter" species.

  • Isn't species selection is just another name for group selection?

    No. The term "group selection" is generally applied to groups of individuals within a species. Within sexual species individuals tend not to form sufficiently distinct groups for group selection to be very effective. Even tiny quantities of gene flow within groups practically destroys any effect. As a consequence of this, group selection is widely regarded as being a weak force.

    Species selection does not suffer from this problem.

  • Isn't species selection always going to be swamped by individual selection?

    Not necessarily. This will happen sometimes, but species selection has the final word, and will favour lineages where individual selection does not get the required variation to work on - or is not opposed to species selection in the first place.

    If that were true, it would not significantly affect the theory. As it is, in practice, it is often possible to identify a "mother" species and a "daughter" species.

  • Species selection works very slowly. After battles with individual selection are considered, has species selection ever managed to produce any significant adaptations?

    It's not easy to identify anything with great certainty - but it seems quite possible. Species reproduction is not that uncommon - and we have observed many instances of it in modern times.

    It is unfortunate that the fact that species selection tends to work in such a way that individual selection comes to be aligned with it, acts to cover its tracks.

  • Species X has apparently remained unchanged for hundreds of millions of years - doesn't that disprove species senescence?

    We don't claim that all species senesce.

    Just as there are organisms that show no signs of senescence with passing age - no doubt some species can live for long periods as well.

    Also not that it's unlikely that we know that "Species X" is actually ancient. It may well be a relatively recent species that has invaded the ancestral species' habitat completely.


Species senescence suggests that species will be likely to prioritise attempts at speciation over their own longevity - to some extent.

This means, for example:

  • Habitats that increase the chances of reproductive isolation will be preferred. In practice this often means coastal regions.

  • Events that change ploidity will have an unexpectedly-high chance of arising.

  • Organisms are likely to have complex genitalia - increasing the probabilities of incompatibilities between them arising.

  • Mating rituals will be complex - increasing the possibility of behavioral isolation.


  1. Medawar, Peter B. "An Unsolved Problem of Biology." H. K. Lewis, 1952.

Tim Tyler | tim@tt1.org | http://alife.co.uk/