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

Fw: The Papers That Ate Cincinnati & (Par)Avian anatomy questions

Forwarded on request, my commments interspersed.

----- Original Message ----- From: "evelyn sobielski" <koreke77@yahoo.de>
To: <david.marjanovic@gmx.at>
Sent: Sunday, May 06, 2007 5:38 AM

> This is a crude representation of the internal
> pattern of the _Triticum_-_Aegilops_ clade.
> What would PhyloCode do with it?

Apply Articles 2 and/or 16; in this case, 2.

As per I.2.: "Taxon names should be unambiguous in their designation of particular taxa. Nomenclatural clarity is achieved through explicit definitions, which describe the concept of the taxon designated by the defined name."

That is Principle 2...

III.I.2.1 does describe but not resolve the problem of
ambiguity. Phylogenetically, either clade (ABX vs CDX)
is equally justified.

I can't find out what article you mean.

Article 2, specifically Note 2.1.3 with its figure, says that not only are both justified, both are correct: X and Y are part of both clades. The clades overlap.


It works well in most or possibly all animals; I am doubtful that it would be that simple in plants. Common wheat would use 2 x-es concatenating 3 binomials. Genetically, 4 of the 6 sets of chromosomes are very similar and the third one is more distinct. Overall, the third genome confers little of significance as regards natural selection (its phenotyspic contribution is decidedly less than 1/3), but (>)1/3 of the significance as regards evolution: In time, the 4 Triticum s.str. genomes would tend to equalize themselves through recombination, but the 2 Triticum s.l. genomes do not recombine as flawlessly IIRC. If a dense fossil record would *correctly* be mapped to a cladogram of phenotypic characters, lineages would have to emerge out of the blue, because that's what can happen in plants (at least in theory; the fossil record is not dense enough but it can be demonstrated in crown orchids IIRC).

Keep in mind that the PhyloCode will not regulate species names, at least not in the first edition. These will be left to the existing codes. It will only regulate clade names.

If lichens are monophyletic, they are monophyletic.
If they are diphyletic
and produced by convergence, then they are
diphyletic. Where is the problem?

What phylogenetic analysis would be able to find such a pattern? It is easier to describe and explain trees like in II.I.2.1. than to actually calculate them with the present algorithms. But gut feelings are not part of total evidence.

What do you mean? I thought it was about whether the symbiosis between fungus and cyanobacterium/green alga evolved once or several times, which is easy to find out by a sufficiently large phylogenetic analysis of fungi. Lichens are symbioses, not hybrids.

(In principle, clades of mitochondria or chloroplasts could be named. It's just not going to happen because it would be redundant.)

> Then, anything that is unicellular and reproduces
> clonally. In these cases, ontogenetic genetic
> changes (eg by HGT) *are* possible phylogenetic
> changes. How to define clades in these?

I don't see where the problem is. Please explain.

Redundancy; a lot of technically correct but "forgotten" taxa.

(A little game I'll do on some raind Sunday afternoon:
most taxa established for the Archaeopteryxes up to
and including Sauriurae are based on "contains some
particular Archie specimen" + (supposed)
autapomorphies. These can be transcribed into
apomorphy-based clades conforming to the PC. What
would be their content today? Would they be valid?
Would they be considered useful?)

Ah. This is only possible with apomorphy-based definitions; most definitions will not be apomorphy-based. Most of the Archie taxa will never be converted because they are already known to be redundant...

But, you see, we don't even try to use them as a
classification framework.
We don't "translate the tree into a classification".
We have stopped
classifying in the first place. Instead, we tie
labels to defined places on
the tree.

"Framework" was not a good word. "Clade-delimitation cluesheet" might express better what I meant. If the tree's not right, the labels are bound to be wrong.

The labels cannot be wrong, by definition. They can only move. "Rather than arguing over whether a group should include a particular species, we should argue whether a particular species belongs to a group." (Brochu & Sumrall 2001)

Names are defined, not clades :-)

All that can happen is that a label may end up with no place to put it. That's called a potentially self-destructive definition. If, for instance, *Reptilia* will be defined as "everything closer to (a tuatara), (a lizard), (a croc) and (a turtle) than to (a bird) and (a mammal)", there will be no reptiles because nothing fits that definition; in mathematical terms, *Reptilia* will be a null set.

It is less the ability to construct a theroretically
working PhyloCode than to implement it successfully in
a competitive environment that I am not sure about*.
No reason not to try it, but don't be disappointed if
it won't work.

The situation has been explained to me as follows:
Mayr wrote in a readable
style but was not a deep thinker; Hennig is a pain
to read, but he was right.

At present ;-) no, actually, he was as much on the spot as one could be back then. We had an old biology textbook from around the time when Hennig had studied. To quote, roughly, "the information of heredity is apparently contained in the protein bodies that are held together by filaments composed of the nucleic acids."

> He should have stuck to the BoPs.

(Birds of paradise, right?)


For something completely different, is it correct that
the pygostyle in Ornithurae is distinct in form
(plough- vs rod-shaped), (inferred) function (lack of
evidence for crown-Aves-type tail in latter), anatomy
(markedly fewer vertebrae) and possibly ontogeny from
that in Nomingia, Sapeornis, Confuciusornis,
Enantiornithes, ...?

Not in ontogeny AFAIK. Function is not clear. The rest is different, yes.

And as regards lower vertebral column: is there
anything on when the uropygial gland evolved?

Apparently not.

* The current trend seems to be to build a cladistic
delimitiation onto Linnean taxa. I.e., keep the
onomatophore, delimit the borders (which hasn't been
done in most cases yet). See for example Sangster
(2005)'s definitions of Mirandornithes.

Yes, why is that competitive?