Subject: Re: orioles/genetics
Date: May 3 22:35:38 1995
From: Burton Guttman - guttmanb at

Teresa caught me saying something silly (How do you make an embarrassed
smiley?) about our having X and Y chromosomes. (Very embarrassing for
someone who regularly teaches genetics. Don't know what I was
[non]thinking.) Leaving the blunder aside, there are some important
points to be made.

First, it doesn't matter which chromosomes have the genes that produce a
character, such as a plumage. What's important for sexual dimorphism is
that the Z and W chromosomes have _regulatory_ genes. The factors that
turn genes on and off are proteins that diffuse freely through the
nucleus and bind to regulatory sites (they're called promoters and
enhancers) on any of the chromosomes. Teresa's models seem quite
sensible, but what will be on the Z and W chromsomes won't
be genes "for" blue or green plumage but rather regulatory elements that
will turn on, or off, a bunch of other genes probably scattered among the
other chromosomes that actually encode proteins, such as enzymes, that
produce the plumage.

That's why I'm troubled when David Wright says, "Teresa Michelson
correctly interpeted the assumptions of the Fisher-Lande model for
evolution of sexually dimorphic characters via sexual selection. If the
genes responsible for the new, sexually selected character are *not* on
one of the sex chromosomes (i.e., they are on an autosome, thus
"autosomal," as in my previous description), then the new, sexually
selected character will appear in both sexes (by default, in a sense)."
Y'see, I don't believe it. (Fisher was writing long before we knew
anything about gene regulation; I don't know about Lande.) The genes could
be on any chromosome; it's pretty unlikely, in fact, for genes that create
new features to pop up suddenly on the sex chromosomes, especially in
birds, because, as I recall, birds tend to have many (rather small)
chromosomes. The sex chromosomes have been making significant differences
between males and females all along, before the appearance of any new
characteristics, and they should be able to continue making those
differences if it's advantageous.

And with plumage it often _is_ advantageous, isn't it? Think about a
scenario. A complicated characteristic like plumage will surely evolve
rather slowly, through the action of many genes. Now suppose a new gene,
or new combination of genes, starts to produce just a little bit of bright
plumage that is expressed equally in males and females, and suppose
females who show it have reduced fitness. Then right from the beginning
there's going to be strong selection (natural selection) to bring the
relevant genes under control of the female regulatory systems and repress
those features in females.

So, David, when you say, "In models such as that of Lande 1980
(Evolution 34:292-20) "changing expression" (i.e., getting rid of it in
females) of such a character in only one sex can take millions of generations,
given high genetic correlations between the sexes for homologous characters,"
my reaction is to say that there's something terribly wrong with the model.

I'm still trying to understand everything David is saying (and I
apologize for not having time to give it all the attention it deserves
right now), but maybe we could try to get some agreement on this point.

Burt Guttman guttmanb at
The Evergreen State College Voice: 360-866-6000, x. 6755
Olympia, WA 98505 FAX: 360-866-6794