Re: Heritability of fitness
- From: joe@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx (Joe Felsenstein)
- Date: Wed, 4 Jan 2006 11:41:19 -0500 (EST)
In article <dp57d5$1818$1@xxxxxxxxxxxxxxxxxxx>,
Perplexed in Peoria <jimmenegay@xxxxxxxxxxxxx> wrote:
>In the course of preparing my next posting on my "Breadbox" thread,
>I came to the conclusion that there should be a relationship between
>the rate at which information 'about the environment' is accumulated
>in the genome and the heritability of total fitness. Since the
>heritability of fitness strikes me as one of the cornerstones of
>Darwin's theory, I assumed that there must be a host of measurements
>of this important parameter in a variety of wild and laboratory
>populations. So I went to the web to see if I could find them.
>
>What I found instead is that it is generally known and/or believed
>that the heritability of fitness in the wild is too small to measure.
>Some empirical studies report that total fitness heritability was
>undetectable "as expected". Apparently I don't understand evolutionary
>theory as well as I thought I did. The following abstract seems to
>summarize the conventional thinking (and also points out that heritability
>is probably not EXACTLY zero):
....
>So my current understanding of this is that the heritability of fitness
>in the wild is near zero at most times and places for most species.
>But occasionally, due to environmental change or the freak appearance
>of a rare beneficial mutation, the heritability of fitness in a population
>rises to a level high enough to actually DO something.
>
>Which leads me to wonder - why was there ANY opposition in the pop gen
>community to Gould's and Lewontin's ideas regarding punctuated
>equilibrium?
>
>One other comment - I would have thought that there must be enough natural
>environmental change out in the wild, and enough variation capable
>of exploiting that change, that fitness heritability would be large
>enough to measure - at least over timescales in which the new environment
>is constant. But apparently the 'signal-to-noise' ratio is just too high.
I don't see why there is a problem. Natural selection would be expected to
exhaust additive genetic variance in fitness. And this is seen -- in
livestock, heritabilities of fitness components are low (litter size in pigs
may have a heritability of 5% while backfat thickness is about 50%). The
former is a fitness component. The higher heritabilities of other characters
(which is expected) means that if the environment changes and one or more
of them become(s) selected it can change.
A slightly more sophisticated view has almost all characters having optimum
selection with fitness declining away from their optima. Each has some
amount of genetic variation maintained by mutation, with selection
acting to hold the variation down. Nevertheless there is genetic variation.
Expecting all heritabilities to be high is unreasonable -- natural selection
can be very effective even when a trait has 5% heritability, especially
on a long time scale.
This view does not force us to envisage most change as punctuational change
caused by species selection (it can have a punctuational pattern, but the
process would still be within-species selection). The difficulty with
species selection is that the turnover rate of species would have to be
high to have the effects of species selection be noticeable compared to
within-population natural selection.
I think that to establish that there is some problem here, you should
attempt a quantitative argument about rates of change.
--
Joe Felsenstein joe@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Department of Genome Sciences and Department of Biology,
University of Washington, Box 357730, Seattle, WA 98195-7730 USA
.
- Follow-Ups:
- Re: Heritability of fitness
- From: Perplexed in Peoria
- Re: Heritability of fitness
- Prev by Date: Re: Which Came First
- Next by Date: Re: Announcing A New Unifying Principle of the Evolutionary
- Previous by thread: Re: Heritability of fitness
- Next by thread: Re: Heritability of fitness
- Index(es):
Relevant Pages
|
