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

Re: New Jeholornis specimen



> Well this is nice. With its long tail Jeholornis kills yet another 
> Archaeopteryx character used to defend the latter's position closer to 
> modern birds than longer tailed dromaeosaurs and so forth. 
 
Well... if we have, irrespective of the position of dromies and everything 
else, this here 
 
+--+--Archie 
|  `--+--*Rahonavis* 
|     `--*Shenzhouraptor/Jeholornis* 
`--Avebrevicauda 
 
, then it is not clear if the +--+ part had a long tail (in which case 
Archie's shorter one is a convergent feature) or a short one (in which 
case that of *S/J* is a reversal); both are equally parsimonious, and 
unfortunately even a complete tail of *Rahonavis* wouldn't change that. 
 
If we have 
 
+--+--dromies 
|  `--+--Archie 
|     `--+--*Rahonavis* 
|        `--*S/J* 
`--Avebrevicauda 
 
, however, and if we qualify the tails of *Microraptor* and 
*Sinornithosaurus* as "long", then, yes, it is most parsimonious to assume 
that Archie's short tail is convergent. Although the alternative would 
only require one more step. 
        This doesn't change if we reroot the tree to 
 
+--+--Archie 
|  `--+--*R.* 
|     `--*S/J* 
`--+--dromies 
   `--Avebrevicauda 
 
-- 2 independent tail reductions are 2 steps, 1 reduction + 2 reversals 
are 3. 
 
> Likewise Confuciusornis  
> has a much less derived postorbital bar than less derived Archaeopteryx. 
 
Really? I mean, important parts of Archie's are unknown? 
In phylogenetic bracketing, we need 2 things to argue for a reversal in 
*C.*: a lack of slightly more derived akinetic birds, and at least 2 
prokinetic less derived birds. 
        If alvarezsaurs and *Avimimus* are closer to Avebrevicauda than 
(to) Archie, then the latter case is fulfilled. For *Sapeornis*, if known 
(there is a skull), the temporal region is embargoed. I assume it isn't 
preserved in *Omnivoropteryx*? 
 
> Yet another example that cladistics is at best hard pressed to handle 
> the degree of reversal, parallalism, and convergence going on in the 
> mosaic development of basal birds 
 
Yet another example that cladistics can't make data, only interpret them. 
And yet another example that shows that there is so far no cladistic 
analysis that includes the many characters that are listed in DA. :-| 
About time one is done. Might become my diploma thesis or dissertation, 
although the former is due in 2 years, so I wouldn't regard it as scooping 
if anyone else would do it. :-) 
 
> Cladistics is one important tool for restoring 
> paleophylogentics, but not the only one, and in this case 
> is inferior to assessing the relative derived degree of 
> flight adaptations that are present. 
 
Where is the problem? Enter those flight adaptations into the matrix. You 
can polarize characters ( = forbid reversals for them), outside of PHYLIP 
you can make ordered multistate characters*, you can write step matrices 
and thereby introduce even more complex weighting of transitions within 
one character. You can weight a posteriori to reduce convergence (there's 
a very interesting case in Mesozoic Birds, chapter 5 by Novas & Pol where 
they got Metornithes back after a posteriori weighting). And of course you 
can use good old a priori weighting, as long as you can defend why you 
weight which character how much. 
 
* Extreme case: Sacral vertebrae in my analysis: 5 (0), 6 (1), 7 (2), 8 
(3), 9 (4), 10 (5), more (6). The Nexus Data Editor 
http://taxonomy.zoology.gla.ac.uk/rod/NDE/manual.html supports up to 3 
more states. 
 
While I am at it... "primordial bird", singular: Urvogel, plural: Urvögel. 
No ü involved. 

-- 
+++ GMX - Mail, Messaging & more  http://www.gmx.net +++
Bitte lächeln! Fotogalerie online mit GMX ohne eigene Homepage!