a few, late night queries for Thomas Holtz

[Stephan Pickering <stefanpickering2002@yahoo.com>]

     During the past three years, or so, several
non-avian theropod phylogenies have, albeit slowly,
sorted out some relationships of taxa where
skulls/skeletons are known, although each effort may
use character states not found in another (i.e., there
appears to be some inconsistency, even if the overall
picture of Theropoda is agreed upon). Mickey
Mortimer's unfolding analyses, e.g., using PAUP and
347+ characters, for Coelurosauria. checks each taxon
in the literature to verify if a recorded character is
actually present. This appears to correlate with your
own pioneering efforts (in the Gauthier/GAIA volumes),
and those of Gregory Paul.
    N. Saitou & M. Nei, in 1987, proposed the
neighbor-joining method (NJ) for molecular data sets
of large phylogenies, including shallow (more recent)
and deep (inferred evolutionary) branches (alas, NJ
trees have not roots). Since we have, as yet, no blood
samples for DNA sequencing from pre-K/T taxa, there is
an ideational gap between Mesozoic and Recent dinosaur
phylogenies. The latter is, I think, becoming more
cohesive, so that DNA studies are eradicating most of
the Victorian classifications (often based only on
morphology and physical appearances; e.g., the condor
is a highly derived crane, not a vulture). And,
Richard Prum's recent AUK paper is breaking through
the ideological walls of ignorance.
    Could not, however, the Saito/Nei NJ concepts
parallel fossil dinosaur phylogenies? In 1996 (NATURE
381:226-229), S. Blair Hedges et al. posited (with
molecular estimates) that only two avian theropod
clades (Ciconiiformes and Anseriiformes) predate the
K/T extinction events. This means that, during the
Cretaceous, these two clades co-existed with other
avian theropods who did not survive the extinction
processes, leaving open the question of which dinosaur
clades actually survived the bolide impact. This
means, too, that during the Cretaceous, the avian
theropods were proliferating as Pangaea broke up, sea
levels rose (reducing available land areas by @ 25%),
environmental stresses (and disease) perhaps killing
entire (as yet unknown) dinosaur breeding populations.
At any rate, the NJ method has, as a principle: an
inferred tree's topology is better analyzed when
compared with a realized tree vs. an expected tree.
    For example, the Sereno/Pisani efforts to present
a classification of non-avian Dinosauria need to be
re-examined rigorously. As mentioned above, there is
inconsistency in characters used in several analyses
of the same clades.
    Why is there not ONE cladistic data base of ALL
KNOWN and ACCEPTED characters of all known
skeletons/skulls, from the tip of the snout to the tip
of the tail? (Setting aside permanently the
paleoastrology of avian/non-avian bone scraps of
metatarsals, vertebrae, teeth, e.g., which have been
given names.)The realized tree would contain
unresolved (i.e., incertae sedis)multifurcations, but
they would be recognized as such, until better
specimens clarify relationships (speciation is
bifurcation, and we are dealing with, in the main,
bone without soft tissue evidence).
    Using computer simulations -- NJ stacked trees to
composite trees, viz., recognizing a taxon's position
relaive to other outgroups -- could not questions be
resolved? We would be discussing monophyletic groups
for statistical comparisons. And, we just might be
able to make speciation guesses using extant taxa.
Among living avians, e.g., a "low" evolutionary rate
is an rbcL interior branch of about 0.00625
substitution/size, ranging up to about 10-fold rate
differences. Within a good framework, we can guess the
time it takes for a new species to appear, isolate
itself enough to become a breeding population of a new
genus. Homoplasy may hinder complete knowledge of
"higher-level" relationships (after all, extant
dinosaurs have been evolving for 65my since the bolide
impact), but pairwise distances can be formulated
independent of trying to reconstruct, e.g., the 65my
history, full of gaps and bone fragments, of ducks.
    A. Rzhetsky has done interesting speculation re:
the minimum-evolution principle: one may have
inaccuracies within analyses, say, of one monophyletic
family, but these may not necessarily change
higher-level relationships if the group's monophyly is
provable.




__________________________________________________
Do You Yahoo!?
Yahoo! Health - Feel better, live better
http://health.yahoo.com