meiotic drive: more hierarchical thinking

arlin at arlin at
Fri Dec 6 11:33:21 EST 1991

In a prior posting about concerted evolution, I gave the following

>Consider a population of 100 individuals, 25 BB, 50 Bb, and 25 bb. We
>can consider a case of no deterministic differences in reproduction:
>each type has the same fitness, and thus the same expected number of
>offspring.  And we can further stipulate no stochastic differences:
>each type not only has the same *expected* number of offspring, they
>each have exactly the same *realized* number of offspring. Thus, each
>generation, each of 100 individuals has a number of offspring such
>that exactly 1 survives to reproduce in the next generation, which
>will also number 100 individuals.

>Can the frequencies of alleles B and b ever change?  Indeed they can,
>though they cannot change due to differences in organismal
>reproduction, since we have excluded this possibility. If something
>like gene conversion is allowed to happen, at some time a Bb can
>become a BB: thus a small change in allele frequencies.

Risto Vainola (Univ. of Helsinki) has pointed out to me that there is
another way for allele frequencies to change: by segregation.  In the
scenario above, no sexual mating scheme was specified.  However, if
one imagines that the organisms are reproducing sexually, by means of
fusion of haploid gametes, then only one chromosome from each
individual will survive in a reproducing individual of the next
generation.  Any Bb individual will pass on either B or b-- and it is
a matter of chance which one is passed on.  Thus, if we specify sexual
reproduction for the scheme above, segregation will result in varying
allele frequencies, even in the absence of gene conversion.  (If we
have asexually-reproducing diploids, then the scheme above does not
allow any effect of segregation.)

Thus, another part of the reproductive hierarchy has been brought into

>concerted evolution... is the combined result of selection/drift at
>two or more different levels of a reproductive hierarchy (the
>hierarchy consists of the fact that individuals reproduce only if
>individuals' chromosomes reproduce, only if chromosomes' sequences
>reproduce: but the three processes do not occur in lock-step).

In the previous posting, the subject was the reproduction of
*sequences* out of step with chromosomes and individuals.  The kind of
effect suggested by Risto Vainola deals with the reproduction of
*chromosomes* as distinct from that of individuals.  He reminds us
that one chromosome, but not its sister chromosome, is segregated into
a gamete when one individual reproduces sexually.  For the case of
random segregation, I know of no prior name for the effect on allele
frequencies; *biased segregation*, however, has been called "meiotic
drive," and there are documented examples of its occurrence in nature.

IMHO even if one chooses to use "meiotic drive" to describe the
evolutionary implications of biased segregation, one should not think
of "meiotic drive" as an additional force in evolution, but as
positive selection (due to biased segregation) acting at the level of
chromosome reproduction as distinct from organismal reproduction.
Similarly, IMHO if one chooses to employ the term "molecular drive",
it should be reserved for cases of positive selection (due to biased
conversion, slippage, etc), and it should not be thought of as a third

Arlin Stoltzfus

Arlin at

More information about the Mol-evol mailing list