molecular phylogeny

SKÁLA Zdeněk skala at INCOMA.CZ
Thu Oct 9 09:25:21 CDT 2003


Dear Dr. Grehan,
I am not a typical supporter of molecular phylogenies but perhaps can comment a bit you questions:

-----Original Message-----
Q.1. If non-coding regions of DNA are supposed to be subject to random
mutational changes why should the genetic patterns between taxa match the
phylogenetic sequence of taxageny (or speciation)?
----------

- here the reasoning is quite similar as in the case of morphologic characters: the character-state pattern should follow the sequence of splitting of lineages (in the plesiomorphy->apomorphy manner), no matter if the characters are morphological or other. (1) The real problem, in fact, is quite opposite: what if the molecular characters are (or, worse, were) subjected to selection or other systematic forces? Then, we can expect parallel evolution in remote lineages that would distort the "neutral-theory" character evolution.
(2)Another problem can follow from the estimated mutation frequency: in a case the frequency was very high (recently or in the past), the character pattern will be noise only with no explanation value as a phylogenetic signal.

----------
Q.2. How would one deal with the possibility that there has been no molecular
evolution at all between species for a group so that all that is being
compared are the ancestral states with respect to the species being
compared and thus get a consistent pattern of molecular relationships but
one that has nothing to do with speciation sequences?
----------
- if I understand you well, you mean the case when the species are homogeneous in molecular markers. Such cases can exist and, indeed you cannot distinguish the terminal taxa as separate units in the analysis. You simply need to use another markers (be it molecular or otherwise).
However, more problematic situation can exist: the case when a "mother lineage" is polymorphic in some markers and the "daughter lineages" receive this polymorphism and loose some character states randomly. Again, resulting pattern will have nothing to do with phylogenetic relationships. Of course, all the described problematic cases apply to morphological markers as well.

Best regards!
Zdenek Skala
skala at incoma.cz




More information about the Taxacom mailing list