Re: Metaves and Coronaves

[Mickey Rowe <rowe@psych.ucsb.edu>]

Peter Houde has seen our discussion in the Gmane archive of the list.
He tried to respond, but listproc bounced his e-mail to me since he's
not currently subscribed.  Anyhoo, let this be a lesson to you any
time you're discussing someone's work... your discussion might get
back to them!

--
Mickey Rowe     (rowe@psych.ucsb.edu)

------- Start of forwarded message -------
Date: Sun, 12 Dec 2004 07:19:43 +0000 (UTC)
From: Peter Houde <phoude@nmsu.edu>
Subject: Re: Metaves and Coronaves
To: dinosaur@usc.edu

Thanks, everyone, for your interest in Matt Fain's and my recent paper.  You 
might be surprised to learn how difficult it is to get such unorthodox views 
published.  Consequently, we had to adhere strictly to one message and skip 
over a lot of issues that only fuel reviewers' criticisms.  A number of 
questions some of you have raised are issues that we are attempting to deal 
with in ongoing research and manuscripts in preparation.  I can fill you in on 
some of this.

For starters, I wouldn't expect anyone to accept our hypothesis without 
corroborating evidence.  I am also certain that for a variety of reasons some 
people wouldn't accept our result no matter what evidence eventually was 
marhalled in its support.  Fortunately, I think we are well on our way to 
achieving this goal with an analysis of a multilocus supermatrix (i.e., like a 
supertree, unfortunately assembled from only partly overlapping data sets).  
And of course other well-funded groups are working on the same.  

Our focus is and has always been the so-called "gruiformes".  For those and 
all their now-apparent potential sister-taxa, we have assembled a large 
complete multilocus data set in collaboration with Carey Krajewski of Southern 
Illinois University.  "Gruiformes" are unequivocally polyphyletic.  Expect to 
see very high bootstrap values to support this assertion.  This will have 
substantial implications for the interpretation of the many fossils of 
superficially crane/trumpeter/seriema-like terrestrial birds that have been 
attributed to "gruiformes".  

There is at least one novel clade within Coronaves for which the gene 
phylogeny makes some very clear sense out of a previously inexplicable mosaic 
of characters in a fairly well-known group of fossil birds.  I'm sorry to be 
cryptic, but the fossils are too readily available for me to be more specific 
before we get something submitted on them.  I have high hopes that greater 
objectivity and circumspection will lead to significant advances in avian 
paleontology and systematics, much as has been done recently for mammals.

Its not that we aren't interested in the biogeographic origins of Metaves or 
that we've given up on trying to sort it out.  The reality, however, is that 
previous attempts to address avian biogeographic origins have been flawed by 
questionable or patently incorrect hypotheses of phylogeny, and complete 
reliance on the modern distributions of extant taxa.  It would be 
irresponsibly premature for us to speculate too much about the distributions 
of fossils whose relationships to modern birds are even more poorly known than 
the relationships among extant "gruiformes".  Our current attempts to further 
explore biogeography involve the calibration of "relaxed" semi-parametric 
molecular clocks to our data, based on the calibration of other groups by 
other researchers for which we have signicantly overlapping data sets and some 
evidence for correlation of lineage-specific rates of evolution.  I am quick 
to point out, though, that the vast majority of existing calibrations are 
based on unverified hypotheses of biogeography and phylogeny, and thus there 
is quite a bit of circularity of reasoning.  Our main objective in addressing 
biogeography in our paper was to temper others from jumping to conclusions 
about what might seem on the surface (and what truly may end up proving to be) 
a Gondwana origin of Metaves.

I refer you to the work of Paton et al, among others, for corroborating 
resolution of Charadriiformes, including the inclusion of Turnicidae.  We 
sequenced several genes of this group fairly extensively owing to our 
incorrect supposition that "gruiformes" are monophyletic and Charadriiformes 
are their sisters.  The molecular data are 100% in agreement on charadriiform 
phylogeny.  Again, expect to see very high bootstrap values in our upcoming 
publications.  

The "courser" that is not is Pluvianus.  Several people have speculated in the 
past based on morphology and other traditional characters that Pluvianus was 
misplaced in the Glareolidae.  We simply corroborated that (now from multiple 
loci).  It is interesting in retrospect that part of the difficulty people had 
in the past in determining the sister-group of glareolids was because they 
were dealing with a polyphyletic group.  Pluvianus had always been the problem 
taxon 'linking' glareolids to burhinids and charadriids.  Although we did not 
study any glareolids other than this and Rhinoptilus, others have.  To the 
extent that they have been studied, coursers (excluding Pluvianus) and 
pratincoles are indeed monophyletic.

Pteroclidae was originally described as Pteroclididae but recently shown to be 
grammatically incorrect.  Refer to Sibley and Monroe and the taxonomy data 
base at NCBI Entrez.

We tried to sequence a specimen of Leptosomus that was nearly 200 years old, 
but without luck.  Shannon Hackett at the Field Museum in Chicago has a fresh 
sample of Leptosomus and she will undoubtedly publish something of interest on 
it soon.  I was always suspect that Leptosomus would be metavian because of 
its endemism to Madagascar and the fact that its morphologically as wierd and 
taxonomically as problematic as all the other Metaves.  A the very least, I am 
confident that it has nothing to do with Coraciiformes.  We were able to 
include it in the supermatrix analysis relying on loci other than beta-
fibrinogen.  We obtained conflicting results depending on which loci and taxa 
were compared, though, and neither result was well-supported.  In one case it 
came out with Metaves, just as Mayr had placed it near the poly- or 
paraphyletic "caprimulgiformes" (all of which group in Metaves).  However, in 
another analysis it came out as sister to owls - which incidentally is an idea 
that has been advocated previously and that is eminently believable from 
osteological similarities.  

I'll leave you with one closing thought.  There is no doubt that birds have 
experienced the sort of ecological shifts we propose within Metaves.  They 
must have because birds are monophyletic.  The question is - how many times 
has this happened?  If extant birds were truly representative of all the 
extinct intermediate forms between these ecological shifts, then parsimony 
would be a direct arbiter of phylogeny.  Hence, Cracraft united loons, grebes, 
and Hesperornithiformes.  But clearly modern birds are not representative of 
all extinct intermediates.  I for one balked when I read several years ago 
that grebes are the sister of flamingos.  I'm not balking anymore; the 
evidence is overwhelming.  Mayr even argued this recently from cladistic 
analysis of morphological characters.  So, the nearest relative of foot-
propelled divers are not necessarily other foot-propelled divers.  Has anyone 
out there compared the skeleton of a tropicbird with that of a grebe lately?  
I have.  Its less of a leap of faith than grebes-flamingos.  See if you can 
even tell the difference between the humerus of Columba and Caprimulgus (both 
Metaves).  The biggest ecological shift within Metaves would probably be 
between arboreal and aquatic birds; yet, that transition may be indicated 
between sunbittern and mesites.  It will probably be a long time before any 
consensus is reached about the validity of Metaves and Coronaves, but these 
sure are exciting times.

------- End of forwarded message -------