Fw: cladistic and the scientific value of paleontology

[arisi <arisigiacomo@libero.it>]

 >----- Original Message -----
 >From: "David Marjanovic" <david.marjanovic@gmx.at>
 >To: "DML" <dinosaur@usc.edu>
 >Sent: Sunday, June 20, 2004 1:44 PM
 >Subject: Re: cladistic and the scientific value of paleontology


 > > It is curious  the strong difference in results
 > > of phylogenetic analysis conducted on the same subject
 > > (for example turtles).
 >
 > This is largely the result of using different data.
 >
 >> Some people not consider paleontology a real science, but rather, a
 > >philosophical discussion where the cladistic is used as an instrument,
by
 > >the researcher, to legitimate, as objective, results obtained selecting
 > >characters, taxa, and morphotypes suitable to the prefixed goals.
 >
 > It is rare that people select their characters and taxa in order to meet
a
 > prefixed goal.
 >
 > >1) Is in the cladistic practice the convergence factor underestimated?

> Let's put it the pessimistic way... if it were, how could we find out? :-)
>

 In my opinion there is the possibility that the convergence increases in
relation to the time distance  that separed the compared taxa. At least
because if the time distance is too
big is big also the probability of extinction of taxa and the consequent
reoccupation of the same niche by
other taxa that will probably evolve convergent characters for the
adaptation at the same niche

>> 2) Should the concept of parsimony be considered applicable in analysis
>> where the taxa are separated by 45 millions of years?
>
> Why not?
>
>> Perhaps parsimony should be considered a valid criterion only in
> >phylogenetic analysis involving the
> >approximately contemporaneous taxa of a clade early radiation compared
> >with an older outgroup
>
> Why?
> Hennig wrote paleontologists were moving towards comparing only
> contemporaneous taxa. This has never happened outside his head. It would
> mean to throw away important data.
>
 It is true that analyzing only contemporaneous taxa there is the risk to
lost important data, but is also true, in my opinion, that otherwise the
overwhelming
convergence could lead  astray. For example varanopid are a group of
synapsid that during the time
evolved, in some lines as mycterosaurinae, a convergent eureptilian
appearance, that is so
striking that their inclusion  in the classical amniote phylogeny of the
past years would probably cause
the collapse of the consistency index and perhaps of great part of the
trees.
>
>
>
> > and without any use of constructed primitive morphotypes,
>
> In cladistics, the construction of a primitive morphotype is _part of_
> making the cladogram. It is _not_ done a priori. In earlier times it was
> common to start with a scenario, imagine a transition sequence composed of
> successive speculative "primitive morphotypes", and to then arrange the
> taxa along this speculative transition sequence; this is what cladistics
has
>stopped. First we try to find out the phylogeny by scenario-independent
> methods, then (if at all) we try to build an evolutive scenario _based
> upon_the phylogeny; in earlier times people commonly did it the other way
> around.
>
I have some doubts about the facts that the primitive morphotype is not done
a priori. in all the phylogenetic analysis the synapsid primitive morphotype
is constructed
using eothyris skull, dimetrodon occiput, and ophiacodon skeleton, probably
because this mix fit
perfectly in a comparison with diadectomorpha. Mycterosaurinae or also
archaeovenator are
not included in the preparation of the mix, despite the fact that should be
more basal
than ophiacodon or dimetrodon, this choice show, in my opinion, the will of
maintaining
synapsida as a monophyletic assemblage, with the consequent stability of the
other clades of amniote
tree.
>
 >
> >It seems to me that the analysis of Rieppel et al. on the turtles origin
> >goes, partially, in this direction comparing the triassic turtles also
> >with  triassic taxa, but I would like to know your opinion.
>
> Does their analysis include pareiasaurs? If it doesn't, it does not test
> the hypothesis that turtles are paedomorphic dwarf pareiasaurs (Lee every
two
> years, Kordikova 2002 http://www.cmnh.org/dinoarch/2003Oct/msg00245.html)
>
 Rieppel et al. include pareiasaurs in their analysis, but if I remember
well
their consistency index result much lower than the one of Lee.

 Alberto Arisi