Dinosauria/Mammalia relationships

[Stephan Pickering <stefanpickering2002@yahoo.com>]

A few years ago -- M.J. Bishop & A.E. Friday 1988;
S.B. Hedges, K.D. Moberg, L.R. Maxson 1990; A.
Rzhetsky & M. Nei 1992 -- proposed that Dinosauria
(Aves) is the sister taxon to extant Mammalia, using
18S rRNA sequences. Bishop/Friday (in the Mike Benton
volume) demonstrating that both clades, independently,
have increased GC in some of their chromosomal
regions. To be sure, this contradicts behavioral and
ecological data...but when parsimony sites are
transformed using the LogDet methodology of Michael
Steel, dissimilarity values are clustered by
neighbor-joining techniques(bootstrap analyses, with
Jukes-Cantor or Kimura two-parameter distances), the
resulting tree congruent with the tree inferred from
other biological and cladistic data.
My question: has anyone continued the research? Do the
early Triassic dinosaurs share a sister taxon
relationship, e.g., with "mammal-like" reptiles? Or,
are we seeing, among living taxa, convergence in
chromosomal data which has no relationship to the
fossil record? Nucleotide compositions can differ
significantly, and can "fool" one into seeing
relationships that do not exist in the "real world"...
and LogDet was, ca. 5 years+ ago, demonstrating its
usefulness in creating robust cladistic trees, i.e.,
LogDet enabled one to discern historical signals in
DNA data. Take, e.g., social insects in the fossil
record. Honeybees diverged 40-50mya, and, in that
relatively short time span, AGCT interfaced (as it
were) with the amino acid compositions of some protein
sequences. Honeybee genomes have high AT in their
mitochondrial genome (Ross Crozier, to whom I give
thanks for sharing his research, has shown that there
needs to be a rigorous balance between DNA data and
the inferences drawn from the fossil record). Using
bootstrap extrapolations for the post-K/T history of
avian dinosaurs does not mean the phylogenetic tree is
correct for these dinosaurs. If the bootstrap values
are high, they probably will not change when longer
genome sequences are found (convergence), but the
bootstrap values do NOT tell one if they are
converging to the correct cladistic tree of
relationships (consistency). In dinosaur studies, the
differentiation between convergence and consistency is important.

__________________________________________________
Do You Yahoo!?
HotJobs - Search Thousands of New Jobs
http://www.hotjobs.com