The Hill-Robertson effect (or Hill-Robertson interference) is a population genetics phenomenon first identified by Bill Hill and Alan Robertson in 1966.[1] It describes an evolutionary advantage to genetic recombination.
In a population of finite size which is subject to natural selection, random linkage disequilbria will occur. These can be caused by genetic drift or by mutation, and they will tend to slow down the process of evolution.[2] This is most easily seen by considering the case of disequilibria caused by mutation:
Consider a population of individuals whose genome has only two genes, a and b. If an advantageous mutant (A) of gene a arises in a given individual, that individual's genes will through natural selection become more frequent in the population over time. However, if a separate advantageous mutant (B) of gene b arises before A has gone to fixation, and happens to arise in an individual who does not carry A, then individuals carrying B and individuals carrying A will be in competition. If recombination is present, then individuals carrying both A and B (of genotype AB) will eventually arise. Provided there are no negative epistatic effects of carrying both, individuals of genotype AB will have a greater selective advantage than aB or Ab individuals, and AB will hence go to fixation. However, if there is no recombination, AB individuals can only occur if the latter mutation (B) happens to occur in an Ab individual. The chance of this happening depends on the frequency of new mutations, and on the size of the population, but is in general unlikely unless A is already fixed, or nearly fixed. Hence one should expect the time between the A mutation arising and the population becoming fixed for AB to be much longer in the absence of recombination. Hence recombination allows evolution to progress faster.[2]
Joe Felsenstein (1974)[3] showed this effect to be mathematically identical to the Fisher-Muller model proposed by R.A. Fisher (1930)[4] and H.J. Muller (1932),[5] although the verbal arguments were substantially different.