"Tim Tyler"  wrote in message
news:ce4oh8$1mg9$1{at}darwin.ediacara.org...
>
> I can see how assortive group formation might generate variation between
> groups - e.g. if all the green skinned folks seek each other out as
> bedfellows - and all the blue-skinned folks do the same.
>
> I'm not sure "outbreeding" is a good name for that, though -
you soon get
> groups consisting primarily of each type - and then
"outbreeding" rapidly
> becomes a very misleading term for what's going on.
>
> Genetic drift is bound to produce some between-group variation.
>
> Regularly ripping groups up and reassembling them at randomly, seems
> likely to generate more variation.  Unfortunately, while the number of
> groups selection looks at goes up - the time selection has to
> evaluate each one goes down.
> -- 

Michael Wade, the group selector of Tribolium, has a good article -
"Critical Review of the Models of Group Selection" in the Quarterly
Review of Biology, Vol 53, No. 2 (June 1978) pp 101-114.

Worth a look, even if it is 25 years old.  He discusses both the old
and the new models of group selection (he calls them "traditional"
and "interdemic") and gives an illuminating analysis of why they both
work so poorly.  He points out some assumptions common to both kinds
of theoretical models that were violated in his empirical studies, and
suggests that theoretical models without those assumptions are needed.
(He also mentions the Triboleum cannibalism that I referred to
elsewhere.)

One point that caught my attention is that while group selection does
tend to "run out of gas" due to an inevitable decrease in inter-group
variance, this only applies to the "additive component" of epistatic
variation.  The non-additive component is subject to selection between
groups, but it is unaffected by selection within groups.  Therefore,
he suggests, group selection for traits that are "emergent" at the group
level can be quite effective, at least in the "traditional" models.
That seems to support the position I was taking in the discussion of
group and species selection involving you, me, Hoelzer, and Wilkins.
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 8/1/04 5:56:35 AM