Re: DNA news stories: bird and mammal evolution

["Matthew Bonnan" <mbonnan@hotmail.com>]

Toby White has done a nice job summing up some of the many problems present
in DNA/RNA analysis for phylogenetic studies.  If I may, I'd like to add a
little more information.

I am addressing specifically the DNA-DNA hybridization data for the bird
study.  The idea behind DNA-DNA hybridization is pretty straightforward --
the closer the similarity between two strands of DNA from two different
organisms (in this case, birds), the closer the two organisms are supposedly
related to one another.  Champions of this method suggest if the rate of
evolutionary divergence were constant (which we cannot know), than we have a
so-called "evolutionary clock."

I think it would be really, really cool if DNA and RNA data worked this way,
and I was sad to discover over time that DNA and RNA data have just as many
problems, if not more, than morphological data in reconstructing
phylogenies.

Some problems with DNA-DNA hybridizations methods speak to the concept of
homology.  Essentially, we are looking at distance data, or how different
each genome is from one another and then assigning kinship based on
similarity.  But this has nothing to do with homology.  A homologous
structure is not just similar form (or sequence, in this case).  The humerus
of a bird, human, whale, dog, etc., is a homologous structure not because it
is the same shape, but because it has a particular position (anterior
pectoral girdle) and similar basic development embryonically in all of these
animals.  Furthermore, we have fossils of the earliest known tetrapods that
show that the humerus is a shared derived or unique feature inherited by all
these animals from a common ancestor.

DNA-DNA hybridization distances are not the same thing.  We have no idea
whether the similarities we see are convergent (homoplasy, parallel
evolution) or homologous (suggesting the inhertied sequences were inherited
from a common ancestor).  The evaluation of morphological character data is
not simply a matter of comparing overall similarity -- we are looking for
shared dervied or unique structures that separate certain taxa from others.
Therefore, the two methods are not comparable and we must be very careful
regarding our reading of DNA data.

Although the tools of my trade are necessarily morphological, I would love
to see more DNA and RNA studies.  The more data we can collect and analyze
on living and fossil organisms, the better for our hypotheses of evolution
and function.  However, as with any new set of tools, we must recognize the
limitations as well as the strengths of those tools.  I do not doubt that in
years to come, methods more on level with the homology concept as expressed
in morphology will come about, and we will eventually have a common language
for phylogenetics and systematics.  Until that time, we must be cautiously
optimisic.

Matt Bonnan
_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com