Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!sdd.hp.com!mips!dimacs.rutgers.edu!rutgers!mbcl!kliman
From: kliman@mbcl.rutgers.edu
Newsgroups: sci.bio
Subject: Re: Coelocanth and evolution:x
Message-ID: <476.2856b24b@mbcl.rutgers.edu>
Date: 13 Jun 91 03:46:19 GMT
References: <17580003@hpfcdj.HP.COM> <18@tdatirv.UUCP> <471.284d6041@mbcl.rutgers.edu> <29@tdatirv.UUCP> <472.28511bf9@mbcl.rutgers.edu> <34@tdatirv.UUCP>
Lines: 50

In article <34@tdatirv.UUCP>, sarima@tdatirv.UUCP (Stanley Friesen) writes:
> In article <472.28511bf9@mbcl.rutgers.edu> kliman@mbcl.rutgers.edu writes:
>>In article <29@tdatirv.UUCP>, sarima@tdatirv.UUCP (Stanley Friesen) writes:
  
>>What I'm saying is purely hypothetical.  I'm inclined to agree that all extant
>>chordates probably differ from their ancestors from 400 MY.  All I originally
>>wanted to convey was disagreement with a statement that I considered too
>>rigid.  We simply don't know what the phylogenies of all living species look 
>>like, and all "rules" are bound to be broken occasionally.

> Hmm, well, I wouldn't exactly call it a 'rule', more like a generalization
> from observation.  The longest spans known for individual morpho-species
> are on the order of 10 million years, so on statistical grounds, I reject
> the likelyhood of a species so *far* out on the tail of the distribution.
> [Typical species last about 1 million years or so].
> 
> Certainly it is strictly speaking *possible* that a form remains so
> stable in all aspects of its biology as to remain viable with its
> ancestors for many milions of years, but it is *very* unlikely (P << .01).
> [In math the symbol '<<' means 'much less than'].

Okay.  I was simply trying to make a point.  But let's play with the numbers.
If we assume there are only 1 million extant species, and if we make the
assumption that the likelihood of a species being 400 MY old is .000001 
(which I hope satisfies the criterion of P<<.01), then the probability
that *no* species is that old is 38%.  If there are 5 million species, the
probability becomes .7% (i.e., 99.3% probability that some species is 400 MY
old).  If we change .000001 to .00001, then the probability that 1 million
species are all recent is .0000454.  

Admittedly, I have made up these probabilities.  I don't know what the value
should actually be.  Without the frequency distribution of known species, I
can't estimate the value.  If you know of a good source, I'd like to take
a look.  

To the point: do I believe that some species is 400 MY old?  I'm not sure.
There's just not enough data to make an informed conclusion.  With all of
those plants, animals, protists, prokaryotes.... who knows.  Perhaps some
regions of the planet remain stable for eons, e.g., deep in the oceans.  
Maybe there's no stability, but particular environments move about slowly
enough for adapted species to move along with them.  How many different
extant species are there, anyway?  (Probably a lot more than 5 million).

So what if the coelocanth isn't 400 MY old.  Or any vertebrate, or any
terrestrial animal (probably). The original question posed - can a species
remain unchanged for that long - is fun to think about. I think it's possible,
but if the author of the question wanted a definitive answer, he won't get one
from this devil's advocate.

- Rich Kliman