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

Re: The validity of cladograms (was Re: giant birds)

Martin Barnett wrote:
> Chris Brochu wrote:
> > Also - on what basis do you regard DNA tests as "more reliable?"
> Well, there are more features to compare with DNA tests than there are with
> bone tests - on the order of millions of base pairs in the former, as
> opposed to the (comparatively) mere thousands of features you get in the
> latter.

This isn't always a "plus" for molecular data.  In fact, for most DNA
sequences, a relatively small percentage of sites actually vary within
the ingroup - in other words, there may be "millions" of base pairs, but
only a few tens or hundreds are useful.  (Actually, most genes are much
smaller than "millions" of base pairs.)  Furthermore, several
simulations have shown that if a data set is positively misleading you,
then adding more of it will only make the situation worse.

Most systematists today would argue that morphology and molecules, on
average, are equally likely to yield a significant phylogenetic signal. 
You'll still run across the occasional "molecules are always better"
type, but not like you might have back in the late 1980's.

  That's just with modern bone tests.  With fossils you also have the
> problem of random bones being missing. 

Agreed, this can be a problem - but missing data can be a problem with
the living, too.  We can have unscorability because of transformation
(see Gauthier et al. 1988 for more on that), and as more combined data
sets are assembled, there will be gaps were some genes are not sequenced
for some taxa (Wiens and Reeder had a great paper about this one not
long ago in Systematic Biology).  And sequences also can have gaps,
which many treat as a kind of "missing data," at least as far as the
algorithm is concerned.

 Dinogeorge mentioned doing an
> only-skull analyses.  I wonder how accurate the sauropod branch of that
> would be given there are few sauropod heads.

A good question.  

  Not only bones missing, but
> entire species - In essence, the material included in your test was picked
> for you with it's own preservation bias agenda.  That's gotta throw out the
> results by some.

Another good point - but again, from my experience, it doesn't "throw
out" the results as much as render some nodes weak.  And it makes a
great case for more field work.

  My knowledge on ancient DNA is probably more fragmentary
> than the material in question, so I won't comment on that.  So in summary I
> will say that I regard DNA tests more reliable because of the DNA, not
> because of the test.
> Out of curiosity, are there any organisations like Genbank
> http://helix.nih.gov/science/genbank.html  set up for researchers doing
> cladistic trees using bone material?  If not, why not?

First, a clarification - most morphology-based systematists don't rely
exclusively on bone.  Soft tissue can be as informative as hard.  We
paleontologists rely on bones, of course, but that's our problem.

I don't know of any general data bases exclusively for morphological
matrices, but there are several places where you can get them -
Systematic Biology usually puts up a .pdf or MacClade version of each
matrix it publishes on its web site, and a consortium at Harvard was
busily assembling as many matrices as it could get.


Christopher A. Brochu
Department of Geology
Field Museum of Natural History
Roosevelt Road at Lake Shore Drive
Chicago, IL 60605

voice: 312-665-7633  (NEW)
fax: 312-665-7641 (NEW)
electronic:  cbrochu@fmppr.fmnh.org