Re: Evolution appears fulfilled-molecular clocks

David Campbell (bivalve@mailserv0.isis.unc.edu)
Tue, 15 Dec 1998 19:18:07 -0400

>At 04:34 PM 12/15/98 -0400, David wrote:

>>Also in this context, it should be noted that most people who try to do
>>molecular clock estimates of the time of metazoan origins do not seem to
>>understand the fossil record or statistics.
>
>You are gunning down some big names in evolutionary biology with that
>comment. But, oh, well, thats nothing new to me!

As far as I can tell, it's mostly big or little names in molecular biology
that are the culprits I am referring to. I wonder about papers that cite
only one date (or perhaps a few) from the fossil record, without citing a
range for error, and use that to say that some other part of the fossil
record is wrong.

[snipping a summary of the discussion at GSA in 1996]

Depending on the gene, the assumptions, and the mistakes made (in one case
cited by Levinton at the 1998 GSA meeting, at least), you can get molecular
clock dates for the origin of metazoans ranging from about 300 to 1500
million years. I think that using better statistics, careful consideration
of the choice of genes (e.g., not something that evolves too fast or too
slow to be useful), and involving people who know something about the
fossil record will greatly improve the credibility of molecular clock
estimates. Right now, there are too many examples of "garbage in, garbage
out."

David C.