[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Extinction

From: John Bois <jbois@umd5.umd.edu>

 > On Sat, 30 Nov 1996, Steve Jackson wrote:
 > > Interesting discussions about the "stochastic" mode of extinction,
 > > either species or mass 
 > This idea doesn't apply to dinosaurs at the K/T--at least no one here
 > has yet explained how it applies.

Why *doesn't* it apply?  It is the normal mode of extinction, and
a claim that it *doesn't* apply requires evidence, as that is the
more extraordinary claim.

Also, I have presented my analysis of the extinction data which suggest
an essentially random component among the susceptible groups. (This
pattern is even more clear if it turns out that some of the crocodiles
that died were terrestrial).

 >  Merely stating the idea doesn't make it fit.  For example, if I
 > have a hundred green marbles representing mammal species at the K/T
 > and a hundred red marbles representing dino species (arbitrary
 > numbers), the probability of picking ALL red marbles at random is
 > close to zero!!

Try a different model: in the Lancian deoposits there were about 10-12
species of dinosaurs and more than 20 species of mammals, perhaps as
many as 30.  Thus 300 green marbles and 100 red is more like reality.
Now add 50 pink marbles, and 150 blue marbles, and 100 yellow ones,
and some more of several additional colors.

Now, select 75% of the marbles.  What is the probability that all of
*at* *least* *one* color are chosen (not necessarily red)? [From a
statistical point of view, this si the correct question, not the
probability of choosing all red]. There is actually quite a high
probability that at least one group will be completely wiped out. But
the mammals, being the most diverse group, are the least likely to
buy the farm.

In fact, after leaving out the groups showing no extinction, the
distribution of extinction among the various groups is quite close
to what is expected from the above model.

In order to show the stochastic model is inviable you have to show
that the *distribution* of extinction between the various groups
differs significantly from that predicted by the stochastic model.
This is what is called a statistical test.

 >  And yet this is what happened at the K/T.  Some non-random
 > causative force was at work.

This is debatable, at best.  This is what you have to demonstrate:
thet there *was* indeed a non-random causative force.  To do so you
must actually perform a statistical test of the observed extinction
pattern against the predicted extinction pattern.

And to do so you cannot just arbitrarily choose to look at just the
dinosaurs and the mammals, you have to include all other groups with
non-zero extinction.

swf@elsegundoca.ncr.com         sarima@ix.netcom.com

The peace of God be with you.