The “Random Genetic Drift” Fallacy — a review

by Neil Rickert

I often see suggestions that random genetic drift drives speciation.  I’m a bit skeptical of that idea.  In any case, a discussion developed elsewhere, and Will Provine’s recent book on the topic was suggested as reading.  I found the book interesting, though perhaps not decisive on the issues.  So I am still at the same point of skepticism as where I started.  But I at least have a better grasp of some of the issues.

I should add that I am not a biologist, though I have taken an interest in evolutionary biology.

What is random drift?

Let’s start by looking at the main issue.  The proponents of random drift seem to be suggesting that

  • there are near neutral mutations (neither benefiicial nor deleterious)
  • by random chance, some of those neutral mutations take over the population (the new genes become fixed).

That is the sort of drift that Provine is discussing.

Note that when there is a mutant gene, there will be several (two or more) possible genes or alleles for the particular locus on the genome.  Fixation, is when the gene takes over the entire population.  That is, it becomes the only allele remaining in that population.  So fixation amounts to the loss of all but the one variant at that particular locus.

If a gene is beneficial, then it is more likely to become fixed.  That’s natural selection at work.  The organisms with the beneficial gene are more likely to survive, so the proportion with that gene increases until it becomes the entire population.

But how can a gene become fixed if it is not beneficial?  It turns out that it can, by random luck in the sampling.  But it might take a long time if the population is large.  So this is mostly an effect for small populations.  A neutral mutant can become fixed in a large population, given enough time and the right sampling luck.  But, in practice, there are selection events and mutation events going on at the same time, so seeing fixation in a large population would be rare enough to not be important.  Random drift is mainly important is small population.


Another thing that happens in small populations, is inbreeding.  Since the population is small, organisms are likely to mate with a relative.  Inbreeding results in the same chromosome being shared among close relatives (the idea of pure breeding comes from this).  Provine describes this as the loss of one or more chromosomes to the population.

It is known that inbreeding and selection can result in particular traits.  Plant and animal breeders have used this technique for hundreds of years.

Provine defines inbreeding as the loss of chromosomes, and he defines genetic drift as the loss of alleles at specific loci.


Much of the book is spent discussing the history of these ideas, starting with some of the history of inbreeding.  Provine then goes on to discuss the history of the random drift idea, mainly starting  with Ronald Fisher and Sewall Wright, and continuing with other researchers up through Kimura and Ohta.  He also spends some time (in chapter 4) on the experimental evidence that has been used to support the idea of genetic drift.

Provine’s view is that genetic drift is being confused with inbreeding.  That is, the evidence for loss of alleles is really evidence for the loss of chromosome variety to the population, which is the expected consequence of inbreeding.

Population genetics

The final chapter of the book deals with population genetics, which is a mathematical theory of how the genetics of populations can change.  Provine is quite critical here.  He sees the researchers as using a mathematically nice model, but one that is unrealistic.  They treat reproduction as random sampling from a pool of genes.  Provine believes that they should instead be seen as sampling from a pool of chromosomes (packages of genes), and they should take into account the effects of meiosis, where there are crossover events between two (perhaps non-identical) copies of a chromosome resulting in genetic recombinations being formed and mixing things up.

My summary

I found the book very readable, even though I am not a biologist.  It did  help that I was already familiar with meiosis.  In my estimation, the book is accessible to anyone who knows enough about biology to have developed an interest in the topic of random drift.  The review of the history of the topic was quite useful.  I even learned something about breeding.

What I did not get out of the book is a final conclusion on the question of random drift.  Provine made a pretty good case that what is being called drift is really inbreeding.  But some of the proponents of the drift idea might actually accept that inbreeding is involved.  So I doubt that many drift proponents will change their minds after reading this book.

%d bloggers like this: