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

Re: was: Wukongopterus and Darwinopterus now: cladistics.



More on this thread later this week.

 Hardly - it simply means that purely dichotomous systems are often
 inadequate.  Other types of codings include things like multi-state
 characters, ordered characters, and continuous characters.  See:
 Pimentcl and Riggins, 2008; Jenner, 2002; Rouse & Fauchald, 1997;
 Rouse, 2001; Strong & Lipscomb, 1999; Forey & Kitching, 2000; Hawkins
 et al., 1997; Hawkins, 2000

The matrix I'm currently working on has lots and lots of characters that follow the pattern "interesting state absent (0), present (1)" -- this loses, in several cases, a lot of phylogenetic signal, because the authors didn't bother to find out how many states the character actually has. In many cases the single state 0 covers a lot of radically different morphologies, the distributions of which contain phylogenetic signal.

> Turtles ancestors had temporal fenestra? Which is the ancestor with
> fenestra that your are thinking of?

 Many recent phylogenies place turtles somewhere within Diapsida
 (deBraga & Rieppel, 1996, 1997; Rieppel & Reisz, 1999; Hedges &
 Poling, 1999; Mannen & Li, 1999).  In such trees, the lack of
 fenestra in turtles is recovered as a reversal.

It is rather useless to cite any molecular phylogeny that dates from before 2001, because they all consist mostly of long-branch attraction... though all more recent molecular analyses, except for Frost et al. (2006) which has other problems, have so far found the turtles next to or inside Archosauria. I'll look at this, and at the morphological analyses that find turtles and lepidosaurs close together, during... it's going to be a postdoc, most likely. I'm supposed to finish my thesis in June -- theses are limited to 3 years in France. :-| So far I think that turtles and procolophonoids look like close relatives (at least more so than turtles and dwarf pareiasaurs), and note that even *Odontochelys* shows no sign of having ever had temporal fenestrae.

But anyway, fenestrae can definitely disappear. The dorsal temporal fenestra closes during the late ontogeny of extant *Crocodylus*, the ontogeny of *Pachycephalosaurus*, and so on; it's also closed in several squamates (don't know about the ontogeny of those). In thalattosaurs and *Vancleavea*, the dorsal temporal bar approaches the skull roof to make the dorsal temporal fenestra slit-shaped in some thalattosaurs and absent in the other thalattosaurs and *Vancleavea*. The lateral temporal fenestra is closed, and most likely secondarily so based on the shapes of the bones in that region, in *Araeoscelis*, and an analogous (and partly or entirely primarily homologous) fenestra closes during the ontogeny of *Milleretta*. The antorbital fenestra shows a trend to closure in ornithischians and crocodiles. The orbit is closed in scolecomorphid caecilians... the nares are closed in some waterbirds such as pelicans... the pineal foramen (a fenestra in all but size!) has closed again and again, once in our ancestors... the posttemporal fenestra is closed in anthracosaurs yadda yadda yadda... you name it, it can close.

I don't understand the point behind this, though. That point was something about ordering. For logical reasons, a character with only two states cannot be ordered in a meaningful sense (all the programs let you do it, but it doesn't change anything), and you were proposing such binary characters (given fenestra absent/present), right?

Surely you didn't believe that "ordered" means "irreversible"?