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

Re: Birds and mosasaurs [Rather long and theoretical]

Tom Holtz wrote:

>Actually, I'd change that first word to "Fortunately".  The problem with a
>lot of scenario building is that the schemes are NOT disprovable...

Out of curiosity, do you know of a case which *does* fit or at least better
fits the model I described, i.e., mosaic evolution of Z-like traits in group
A where Z and A's LCA was an older model without strong Z-like characters?

>>The sampling error works both ways.  Your logic is that there may have been
>>Dromeosaurids by the megaton in Early Jurassic days, or even earlier.
>Not by the megaton.  All that needs to be there is a single lineage.  And,
>of course, I do not advocate that dromaeosaurids are ANCESTRAL to birds,
>only the sister taxon to them.

Right you are.  However, that still means that the main lines of the
Maniraptoriform radiation had to be essentially complete by the mid-Jurassic
at the latest and perhaps as far back as the late Permian.  If this is the
case, we might expect to see all of the main Maniraptoriform lineages with
essentially world-wide distribution since there were no sea barriers during
most of this time.  I don't know, so I'll have to ask, is this the case?  If
not, I admit its no disproof, but its a point against your thesis.

>>on the fact that much later Dromeosaurs are relatively birdy, you'd suggest
>>(not unreasonably) that this improves the chances of some early Dromeosaur
>>being the ancestor of birds.
>No.  No.  Uh uh.  Nope.

Sorry.  I spoke loosely.

>To clarify the issue: most phylogenetic analyses indicate that
>dromaeosaurids and birds are each others closest relatives (aka sister
>taxa).  Neither is ancestral to the other, but both came from a common
>The Feduccia crowd have misled many people into thinking that Gauthier and
>company suggest an ancestral position for dromaeosaurids.  They (read we) do

Its probably not the Feduccia crowd so much as the fact that some of us (who
are already beginning to fossilize) completed our formal educations before
cladistics.  We tend to speak in terms of descent when there is no real
reason to do so.

>Well, larger than typical birds, but not particularly large sized.  My pet
>dog as a kid was bigger than any Velociraptor or Saurornitholestes, and she
>was only a golden lab-collie cross.

Jeez!  Picky, picky...  

>But there isn't that much variation there, within each coelurosaur lineage.
>Most dromaeosaurids look extremely similar to each other.  Most troodontids
>look extremely similar to each other.  Most tyrannosaurids look extrememly
>similar to each other.  And so on.

I'll take your word for it, although the male and female dromeosaurs would
probably have disagreed with you.  There is still a lot of variation between
the Maniraptoriform groups, a lot more than we see in Carnosaurs for
example.  In any case, doesn't this make matters worse for you?  You must
posit that essentially nothing happened in the way of evolutionary novelty
in this group for 100-150 My.  That's a huge amount of stability for a group
that had an average species lifetime of (if I remember the estimate) about 3My.

>Long branch attraction may result in molecular convergence, but would be
>extraordinarily unlikely to result in a consistent pattern of distribution
>of "birdlike" features between each of the coelurosaur lineages, while
>simultaneously NOT producing variation of the birdlike features within each

Can you spell this out a bit more?  Molecular-type long branch effects are
statistical anomalies resulting from random events.  Phenotypes, on the
other hand, are not random because only some are viable.  Thus, we'd not
expect to *literally* see long branch attraction in paleontological
cladograms.  Instead, we'd see independent, mosaic recurrence of permissible
(i.e. viable) novelties.  This is the phenotypic analog (and direct result)
of genotypic LBA.

>>The point being that
>>regulatory mechanisms are no respecters of cladistics.
>(Tell that to the evo-devo crowd!  They make a big case, and rightly so, of
>demonstrating how patterns of regulating mechanism among the various groups
>of metazoans can be used to resolve interrelationships among animal
>lineages, as well as to predict various common ancestors of the different

Actually, I'm trying to apply some of the evo-devo stuff to this problem.
The point is that highly conserved genetic regulatory structures (e.g. hox
boxes) and highly conserved structural genes can result in radically
different phenotypes essentially overnight by changing which regulatory
domain governs which structural gene.  This is exactly why it isn't
necessary for the LCA of  (to use neutral terms) Z and Z-like A's to have
much in the way of Z-like features.

  --Toby White