[[e-mail]: Re: Relaxed Selection, a b-level posting]

[Russell Standish]

It seems I have to forward this message rather than bouncing it. Oh well...

----- Forwarded message from Russell Standish <[hidden email]> -----

Date: Sun, 12 Oct 2008 21:17:17 +1000
From: Russell Standish <[hidden email]>
To: Nicholas Thompson <[hidden email]>
Subject: Re: [FRIAM] Relaxed Selection, a b-level posting
In-Reply-To: <[hidden email]>
User-Agent: Mutt/1.4.2.3i

On Fri, Oct 10, 2008 at 01:28:38AM -0600, Nicholas Thompson wrote:
> Russell,
>
> Allow me to use CAPS in your text to distinguish my text from yours in the
> dialogue below.
>
> Some Santa Fe locals have cautioned me severely that it is impossible to
> use CAPS in an email message without shouting; but I am hoping that that
> convention does not extend to the southern hemisphere, but if it does, I
> apologize in advance. .  

Well, yes CAPS is a form of shouting Internet wide. The usual
convention for responding to someone is as I have done here, with the
quoted text having a '>' in front of it.

>
> Nick
>
> Nicholas S. Thompson
> Emeritus Professor of Psychology and Ethology,
> Clark University ([hidden email])
>
>
>
>
> > [Original Message]
> > From: Russell Standish <[hidden email]>
> > To: <[hidden email]>; The Friday Morning Applied Complexity
> Coffee Group <[hidden email]>
> > Date: 10/10/2008 7:16:11 PM
> > Subject: Re: [FRIAM] Relaxed Selection, a b-level posting
> >
> > One should not confuse economics with biological selection. It would
> > seem plausible that good economic times might lead to rapid evolution
> > of economies, such as during the recent Internet bubble for instance,
> > but not that it would have any influence on us at the genetic level.
>
> WELL, EVEN IN THE ECONOMIC DOMAIN, I AM TROUBLED BY THE SUGGESTION THAT THE
> REMOVAL OF K-SELECTION (YES, FOLKS, i ADMIT THIS IS A B-LEVEL POST) MEANS
> THAT ALL SELECTION HAS BEEN RELAXED.  OBVIOUSLY, THERE IS STILL R-SELECTION
> TO CONTEND WITH. SO EVEN IN A POST EXTINCTION (OR POST RECESSION
> ENVIRONMENT), THERE COULD BE INTENSE COMPETITION IN THE SPEED WITH WHICH
> FIRMS EXPAND OR ORGANISMS REPLICATE.    

You were the first to bring up K and r selection, the names of which come
from the logistic equation (K being carrying capacity, and r being net
reproduction rate). The equation in question looks something like:

  dx/dt = rx(1 - x/K)

The distinction between r and K selection is whether x << K or x ~ K
when selection occurs.

For an n-species ecosystem, one needs to replace the logistic equation
with something like a generalised Lotka Volterra equation:

  dx/dt = r*x*(1-Ax)

where x and r are now an n-dimensional vectors, A a matrix and * is
elementwise multiplication. One could consider K to be the inverse of
the diagonal element A_{ii}, and then r & K selection would be as above.

> >
> > The sort of idea that David Green was proposing was that ecosystems
> > (aka foodwebs) would cycle between a chaotic and a stable phase. My
> > take on this is that immediately after a mass extinction, just about
> > any foodweb is stable, because there are not enough connections to
> > make it chaotic.
>
> I AM EVEN MORE IGNORANT THAT USUAL IN THIS, HERE, DOMAIN, BUT ISNT IT
> POSSIBLE THAT A WEB COULD BE LESS STABLE BECAUSE IT HAD FEWER CONNECTIONS?
> DOESNT IT DEPEND ON THE NATURE OF THE WEB ... LIKE WHETHER ITS EDGES (IS
> THAT THE RIGHT TERM) ARE MORE SERIAL OR PARALLEL, FOR INSTANCE?  
>

There are many different notions of stability. The two main ones are
the mathematical notion of stability (aka linear stability of the
fixed point x_0 = A^{-1}1) and permanence (no extinctions occur).

Both of these properties become less likely the larger the foodweb is,
due to the increased likelihood of a positive feedback cycle which
causes the dynamics to fluctuate wildly eventually causing the
extinction of one or more species.

However, even a two species foodweb need not be stable (which L-V
showed back in the 1930s) nor even persistent. With 3 or more species,
the dynamics can also be chaotic. Nevertheless smaller foodwebs are
likely to be stable, or exhibit persistent limit cycles than large
ones. The effect is so dramatic, that any random assemblage of species
will suffer an extreme extinction event. Such things have happened
in the past when continents collided, or now when human being have built great
vessels ferrying plants and animals around the world.

Large, complex ecosystems get built by a different, no so well
understood, evolutionary process. I think specialisation has a lot to
do with building complex food webs.

> Selection under such circumstances would be fairly
> > relaxed.
>
> MY BASIC INTUITION HERE IS THAT IT DEPENDS ON WHICH  SORT OF CATASTROPHE
> [NEARLY] TAKES OUT THE SPECIES.  SELECTION IS JUST DIFFERENTIAL REPLICATION
> WITH RESPECT TO SOME HERITABLE PROPERTY OF THE INDIVIDUALS IN THE SPECIES
> [OR IN SPECIAL CASES, GROUPS OF THE SPECIES].  FOR INSTANCE, IT HAS BEEN
> ARGUED THAT AT THE EXTREME BOTTLENECK WHERE HOMO SAPIENS EMERGED, THERE WAS
> TREMENDOUS K SELECTION ON GROUPS.  IN  A FAMINE SITUATION, THOSE GROUPS
> THAT COULD HOLD TERRITORY OR TAKE IT FROM COMPETING GROUPS WERE THE FEW
> THAT SURVIVED.  HENSE THE EXTREME GROUPISHNESS OF THE HUMAN SPECIES.   SO
> THAT WOULD BE A POSITIVE INSTANCE FOR YOUR ARGUMENT.  BUT IF THE SPECIES
> HAD BEEN TAKEN OUT BY SOME OTHER EVENT .... ROCKS FALLING OUT OF THE SKY
> .... THEN THERE WOULD HAVE BEEN NO SELECTION.  DOES THIS MAKE SENSE, OR AM
> I MISUNDERSTANDING THE ARGUMENT?  
>
> IS THERE ANY REASON TO BELIEVE THAT FOOD WEB COLLAPSE WOULD BE ANY MORE
> SELECTIVE THAN THE CHALLENGES OF RAPID REPRODUCTION????
>

Any extinction event is a collapse of the food web. And selection only
proceeds by means of extinctions. So I'm not really quite sure what
you're trying to nuance here.

> As evolution proceeds, the foodweb becomes more
> > complex until such a time as chaotic behaviour sets in. Extinction
> > becomes increasingly likely, and corresponding selection becomes
> > "fierce".
>
> > Cycles of mass extinction followed by species radiation _may_ be a
> > driving cause of ecosystem complexity.
>
> MY ANSWER WOULD BE, "SOMETIMES".  PERHAPS SOMEBODY COULD HELP ME OUT, HERE.
> >
> > I'm trying a slightly different tack with Tierra, of artificially
> > inducing mass extinctions every now and then. I have also tried
> > reducing parsimony pressure from time to time (I'm not sure what would
> > be the biological world equivalent of this - possibly variation in
> > background radioactivity or cosmic rays). But currently my simulation
> > code is broken, so I haven't got too far with this to date :(
>
> I OF COURSE KNOW NOTHING OF THE INTRICACIES OF SIMULATION.  BUT THIS
> INTERESTS ME AND I AM GLAD YOU ARE WRITING ABOUT IT.
>
> AND I PROMISE I AM NOT SHOUTING.
>
>
> NICK

--

----------------------------------------------------------------------------
A/Prof Russell Standish                  Phone [phone] (mobile)
Mathematics                        
UNSW SYDNEY 2052                 [hidden email]
Australia                                http://www.hpcoders.com.au
----------------------------------------------------------------------------

----- End forwarded message -----

--

----------------------------------------------------------------------------
A/Prof Russell Standish                  Phone [phone] (mobile)
Mathematics                        
UNSW SYDNEY 2052                 [hidden email]
Australia                                http://www.hpcoders.com.au
----------------------------------------------------------------------------

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org