re: E and P of pterosaurs - more comments

[david peters <davidrpeters@earthlink.net>]

David Unwin wrote:
Here are some short responses to Peters comments on my earlier mailing. 

>First of all, let me mention one compliment I forgot to pass on earlier
>regarding David's observations. He was exactly right in anticipating
>that Preondactylus would have a higher premaxilla ascending process than
>previous reconstructions showed -- more like Dimorphodon. A
>reconstruction justifies David's foresight.

DU: I would be gratified if any actual new evidence was cited to support my 
reinterpretation. It isn't, so I'm not. 

DP: I mentioned a reconstruction, David. All you have to do is ask to see it. 
It's unpublished, but that's okay. And it was based on a viewing of the actual 
specimen in Italy.

>I have to wonder, though, about David's comment above. What will strike
>the reader upon even a casual glance at his 2003 figures is that they
>all appear to be copies of extant figures, rather than new drawings --
>starting from scratch. The same holds true of many of David's papers,
>excluding his 1988 report on Dimorphodon, which had no precedent. I have
>no doubt that David has actually seen the fossils. A carefull tracing of
>some or many of the fossils will reveal flaws in their precedents. A
>tracing of the figures themselves will not.

DU: So, the plan is that I travel around the world with my sketch pad and 
pencil, redrawing each of the 50 or so line drawings that appear in my E 
and P of Pterosaurs paper on the basis of the original specimens. Would 
someone like to step forward with the cash to do this, oh and at the same 
time you can try convincing my boss to give me a few months of work for 
the travelling. And don't mention that at the scale they are drawn, my 
sketches will look almost exactly like the published versions.

DP: As I understand it, you've already been around the world examining the 
various specimens. 

DU: Fortunately, the validity and reliability of cladograms (including my 
own) do not depend in any way whatsoever on my drawing skills. 


>David's matrix has a higher proportion of filled boxes than any
>preceding it, and this is one reason given for using suprageneric OTUs.
>If that's the goal, then great. I still think it breeds chimaerae.

DU: So prove it. 

DP: Would a few examples be enough to illustrate the chimaera problem? 

1. Let's code the Ornithocheiridae for a "notarium [fused dorsal vertebrae]". 
It appears in Istiodactylus and "Coloborhynchus" spielbergi, but not in 
"Anhanguera" piscator or Arthurdactylus. Do we code a "0"  or a "1"? We can 
suppose that in those specimens without a notarium that they are juvenile, but 
if you look at the big picture, you'll see that the notarium is restricted to 
one clade within the Ornithocheirdae.

2. Let's code Campylognathus for "skull smaller than glenoid-acetebulum 
distance." It appears in C. liasicus, but not in C. zitteli. 

3. Let's code Dorygnathus for "manual digit III longer than metacarpal IV". It 
appears in SMNS 50164, but not in the Donau specimen.

There's more, of course... but I think you see the problem.



>I think the pterosaur interrelations problem is more difficult than
>anyone imagined two years ago when David's paper was probably due to the
>publisher, but thankfully and fortunately, it appears that there now is
>a critical mass of taxa to tell the tale -- if they are all employed. I
>applaud David's efforts to divide the spectrum of pterosaur
>interrelationships further, but I am disheartened to hear that the
>cladogram will be the same as before. 

DU: This doesn't make any sense at all. Peters argued that analyses should be 
done at finer taxonomic scales in an earlier post, but is unhappy with 
the result when someone does just that. 

DP: I was disheartened because you didn't go deep enough, evidently. Because 
when you start including phalangeal proportions and palate morphology -- and 
alot more specimens --  your cladogram will start to change. First it will 
become incredibly messy with 1000s of possible trees, then as the data becomes 
more refined, you'll see new patterns sifting out. If you have more than one 
tree in the end, as in Unwin 2003, there's something wrong, you didn't dig deep 
enough, because in nature there is only one tree. Of course, you can have one 
tree with a bad matrix, but that's another story. A good test for your tree is 
to input a new taxon, even if incomplete, and see if it falls into place or 
messes things up. A robust tree will accept new taxa without a problem. It will 
also alert you to observational problems by exposing "bogeys" in the matrix. 
And that's the beauty of PAUP. 


>Mine was similar, about a year
>ago, when it had a single clade "Pterodactyloidea" as all published
>cladograms do now. Here's what happened: the morphological gap between
>the highest rhamphs and the lowest pterodacs bothered me enough to dig
>deeper, and I hope David and others do the same. Evolution works in tiny
>steps and if you look hard enough you'll see the tiny steps, as I have.

DU: Cladograms as I expect readers of this list are now well aware, are not 
intended to, and indeed do not, show evolution. Peters statement above 
exposes some extremely muddled thinking.   

DP: I do believe that cladograms reveal evolutionary patterns better and more 
precisely than any other currently available method. As I understand it, the 
sister taxon approach to systematics using an unprejudiced computer program 
actually works better than previously efforts done by hand. The program reveals 
which taxa are primitive and which are derived. Most importantly, it exposes 
convergence, which is much more rampant in the Pterosauria than you may think. 
The warped deltopectoral crest appears twice, for instance. The notarium, at 
least three times. If cladograms do not reveal family trees [relationships] and 
evolutionary patterns, what do they show? Please clarify my muddled thinking! 

And I remind you, David, that as a professor, you have taken on the mantle of a 
teacher. In that role you are entrusted to explain things to your students, no 
matter their age. In my experience, the best teachers always follow accusations 
of muddled thinking with clarifying statements.

>I would urge future geneologists to consider a number of Dorygnathus and
>Scaphognathus specimens as terminal taxa and not to consider Dory or
>Scapho a single taxon each. Here is where the former 'Pterodactyloidea'
>splits into four clades. Pay special attention to the manus and pedes of
>each. Some reassembly will be required. In all four cases, a size
>crunch follows that ultimately leads to the larger taxa we know and
>love. So, here's where the wee ones make their entrance before giving
>rise to the big ones. The Teyler's Museum in the Netherlands cares for
>some of these key taxa.

DU: So this is based on intensive anatomical studies, accompanied by detailed 
drawings of each of the actual fossil specimens concerned right?   

DP: Often, yes. And you can see them whenever you wish!!  

DU: I'll finish with another contentious proposal, which one might call 
Hennig's Law: 

'The quality, reliability and robustness of a cladogram is directly 
related to the amount of time and effort spent studying the specimens on 
which it is based.' 


DP: I've already spent a few years on the specimens, and _still_ they are 
revealing hidden data. So I _fully_ agree.

       I think JFK Jr. said it better [from heaven]: "Trust your instruments."

David Peters
St. Louis