[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]
Re: Class-less dinosaurs
>From: chartiem@ERE.UMontreal.CA (Chartier Michel)
> I don't think that the desire to retain the class Aves has much
> to do with what you think is their proper phylogeny (i.e. be they
> descended from dinosaurs, crocodiles or whatever) as much as it has to do
> with what kind of systematist you are, i.e. pheneticist (numerical
> taxonomist), evolutionary systematist (evolutionary taxonomist), or
> cladist (phylogenetic systematist).
I was planning on saying something like this, but you beat me to it.
This is certainly why *I* retain Aves seperate from dinosaurs.
[I am an evolutionary taxonomist, not a cladist].
> If you're a strict cladist (which I am not), you will want
> to include birds with theropod dinosaurs, just like you will want to
> include humans and pongids (chimp and gorilla) into a single taxon,
> namely hominids. But to do this is to take away what is "special"
> about the group concerned. It is clearly evident that humans are much
> more derived than either chimps or gorillas are, so why not recognize
> their unicity (is that a proper word?) and retain them in hominids, to
> the exclusion of chimps and gorillas?
And of course similarly argued for Aves.
One good book on this subject is "Principles of Systematic Zoology,
2nd Edition", by Mayr and Ashlock. [The older edition, by Mayr only,
does not address the issue as fully].
Another good reference, showing *mathematically* the weakness
of a cladistic classification, is:
Carpenter, K.E. 1993. Optimal Cladistic and Quantitiative
Evolutionary Classifications as illustrated by Fusilier
Fishes. Systematic Biology 42(2):142-154.
> You don't have to remind me that, yes, evolutionary systematics
> are very subjective in nature.
So far, but I think by pulling in the best parts of the cladistic
method along with information theoretic measures (as in Carpenter's
article), this can be much reduced.
> But to believe that cladism is totally
> objective is being naive.
In addition to what you mention, there is also the issue of the
sensitivity of cladograms to the choice of character set and to
the set of taxa being analyzed. Where this occurs, this introduces
another level of subjectivity.
Then there is the problem of multiple "best" cladograms for most
data sets, making selection or integration a highly subjective
issue. [I think strict consensus cladograms throw away too much
> why be a strict cladist, or a strict evolutionary systematist? Why not
> take what's good of both methods instead?
Actually, the way I look at it, using cladistic methods to derive
hypotheses about relationships is entirely consistant with an
evolutionary systematics approach. Evolutionary taxonomy, unlike
cladism, is not tied to one method - it is more a philosophy or
way of approaching the question. Thus I am perfectly happy to
add cladograms to the data I use in establishing taxa.
[I currently favor majority consensus and Adams trees to either
maximum parsimony or strict consensus trees].
On this subject: are there any maximum likelyhood methods for
generating cladograms from *character* data (as opposed to
sequence data)? If so, I would much appreciate references.
> One last thing. Dinosaurs as a class sounds great (and it also
> makes common sense (a la evolutionary systematics style) since the
> group as a whole is as diverse in its adaptations and morphology as are
> the mammals). But what about archosaurs as a group? If you take out
> dinosaurs, what happens to archosaurs, Phil?
I have looked intot he matter, and I have concluded that the proper
break point is to establish a high ranking taxon for the archosauri-
morphs. Thus I currently tend to favor a subclass (or class)
Archosauria that includes not only dinosaurs, pterosaurs, crocodiles,
and thecodonts, but also rhynchosaurs (or is that rhynchocephalians?).
[Note, I exclude Aves from this, as I retain it as a seperate class].
The peace of God be with you.