The Evolution of Biological Innovation
The theory evolution by natural selection, boiled down to its bare bones, is pretty simple. All it requires is that a few conditions be met among a population of animals or plants (or any other organism for that matter): competition for resources; variation in survival and reproductive success; and a system of heredity that ensures that some of this variation is passed on from generation to generation. We now know what Darwin didn’t, that genes underlie the transmission of much of the variation seen among organisms that affects how well they thrive, and whether they pass on their genes or not.
And so natural selection can be cast as an essentially algorithmic process: when there is genetic variation in a population of organisms, and some of this variation affects how well they get on in life, the population will evolve. As new genetic variants with beneficial effects arise, their bearers will do better, pass more copies of these genes on, and after a while most or all members of the population will carry the new genetic variant and its associated benefits.
So when thinking about if and when a population will evolve, genetic variation is a crucial issue. If there is none, then there is now raw material for natural selection to work on. Although there may be variation in the outward form or behaviour of the individuals in the population, and some of this may affect whether they stay alive and fecund, it won’t be passed on to future generations — thus short-circuiting the cumulative power of natural selection.
So the extent of genetic variation in a species or population is a crucial determinant of whether it will evolve, and how it will respond to new selective pressures. To capture this in a word, we might say that genetic variation drives the ‘evolvability’ of a species of population.
‘Evolvability’ was coined, perhaps surprisingly, as recently as 1987, by Richard Dawkins, the arch-phrasemaker who also brought us the ‘selfish gene’, ‘extended phenotypes’ and ‘memes’. And while it has sometimes been used to reflect the capacity for evolutionary change under the pressures of natural selection described above, it is nowadays more commonly used to mean something more subtle, perhaps more fundamental.
Evolvability, in its modern sense, generally refers to the capacity for genetic changes to produce adaptive changes in how organisms are built and behave — their phenotypes, in the biologists’ lexicon. The issue here is not the extent of genetic variation per se, but how this genetic variation maps onto phenotypic variation — that is, whether genetic variants produce phenotypic variants that are beneficial and can be passed on to offspring. This is the key to evolutionary innovation, and the emergence of new organismal designs. So rather than focusing on how much genetic variation is knocking around, researchers interested in understanding evolvability are increasingly looking to the factors that determine the ‘genotype–phenotype map’: for it is changes in the mapping functions that determine the relevance of whatever genetic variation is present.
This is all pretty abstract and theoretical, but I put some flesh on these ideas in a piece for New Scientist this week. It’s currently available to read in all its glory here. Check it out.