Re: Pan-clades, good or bad?

[Christopher Taylor <ck.taylor@auckland.ac.nz>]

On 16/6/04 8:30 am, "David Marjanovic" <david.marjanovic@gmx.at> wrote:

> Well. Whether Panarthropoda _is_ an example depends on whether we consider
> Onychophora and Tardigrada well-known names. The idea seems to be to define
> all well-known names that contain living representatives as crown-groups and
> to then make Pan-clade names based on them.
> 
    <Sputter, sputter> I'd say that both Onychophora and Tardigrada are
*very* well-known names. Onychophora, in particular, is even more so these
days as a result of work on the Burgess Shale (Hallucigenia and its ilk).
And both are well-estabished on what might be called the  'standard'
textbook list of phyla. Not to mention that tardigrades include perhaps the
cutest of all invertebrates.... :)


>       I really don't like implied names. The Pan- business could become
> like today's confusing rule of the ICZN that says if one names a family
> (-idae), one automatically names a superfamily (-oidea), subfamily (-inae)
> and tribe (-ini), and perhaps (I don't know) a subtribe (-ina), and when 100
> years later someone uses any of those implied names for the first time, they
> have to be ascribed to the author of the family.
> 
Not so confusing at all. It ensures that if any family is divided into
subfamilies, then at least one subfamily has the same type genus as the
family, and it can be inferred simply from that. Also, the type subfamily
ranks the same in terms of priority as the family. So we don't get a
situation where the type subfamily is based on a different genus name from
the type family (if A-idae (publ. 1890) with no subfamilies was synonymised
with B-idae (1891) containing subfamilies B-inae and C-inae, and the type
genus A was regarded as the same subfamily as C, we wouldn't get a family
A-idae containing subfamilies B-inae and C-inae, with no indication from the
names that C-inae was the type subfamily).

> While, in fact, it's just an argument for _naming_ crown groups at all. :-)
> 
>> (it avoids the inherent possibility for error in common statements
>> such as 'all mammals produce milk').
> 
> While it doesn't avoid the inherent possibility for error in other common
> statements such as "mammals give birth to live young" or "marsupials are
> characterised by their marsupial bones". Or -- a real example -- "poison
> spurs are a synapomorphy of Monotremata". (They are at least one of the
> crown group of Mammalia, with at least one secondary loss; for monotremes
> poison spurs are plesiomorphic.)
> 
    These are merely examples of being just plain wrong, then. The
crown-clade restriction argument is derived from characters which _cannot_
be determined one way or another in the stem group - most physiological,
genetic characters, etc.


> Of course, "Aves is the sister-group to Crocodylia" becomes even wronger, if
> I may write that, when those terms are restricted to their crown-groups.
> 
The statement 'Group A is sister to Group B' is, unless dealing with two
sister stem clades, always a matter of context, though - which taxa are
considered in the tree. Aves _is_ sister to Crocodylia in a tree using only
living taxa, for instance. By now, I think this is so much the case that to
insist on a strictly correct use of 'sister taxa' would merely cause big
problems in communication, Pragmatism or idealism?

> BTW... about half of all occurrences of *Crocodylia* in the IPNM abstract
> booklet http://www.ohiou.edu/phylocode/IPNM.pdf are spelled *Crocodilia*.
> Perhaps this spelling will win? I'd like this, because it's etymologically
> correct, and people seemingly never get used to the y version anyway.
> 
But the type genus is spelt _Crocodylus_. I realise that typification and
the like don't apply to higher names, but I still think it complicates
matters unnecessarily if taxa disagree with the genus names they're based
on. Similarly, I dislike the use of the name 'Ornithosuchia' for a clade
which doesn't include _Ornithosuchus_.

>> The point is, do you want taxonomy to
>> agree with the statements of 99+ percent of natural scientists,
> 
> (Are there really so many zoologists and botanists out there? What about
> "80+ %"?)
> 
Not to mention ecologists, molecular biologists, microbiologists...

> 
> The latter can be stopped; note how people stopped calling dinosaurs
> reptiles before (within the last 5 years or so) people started digging up
> the definition of Reptilia from Jacques Gauthier's dissertation. The
> former... perhaps we should adopt the definition of *Mammalia* by Luo,
> Cifelli & Kielan-Jaworowska 2002 -- {*Sinoconodon* + crown-group}. This
> would contain *Morganucodon*.

    Which, funnily enough, they didn't use in that very paper -
_Adelobasileus_ was included in Mammalia, despite falling outside the
phylogenetic definition.

>> BTW: I DO feel Lissamphibia should be abandoned. For one thing, it was
>> originally coined to *exclude* frogs (!).
> 
> 1. Nobody knows this anymore. All usages I've seen include frogs.

    A good point. After all and for instance, Insecta, when first used by
Linnaeus, included crustaceans as well as insects. I can't imagine anyone
arguing for a redefinition of Insecta based on this.


>> So, perhaps it's a good idea to link corresponding crowns and panstems in
>> some way, but I don't see how this could be it. Perhaps we could allow
>> hyphens in this one instance (_Pan-mammalia_)? Or perhaps there's
>> another solution.
> 
> Hyphens... why actually not...
> 
In this age of computer searches, the need for avoiding punctuation,
dialectics, etc. in names is even more important then ever, I think. A
computer can't recognise that Panmammalia and Pan-mammalia are the same
thing. And punctuation is the first thing to get dropped by sloppy editing,
or by misspelling.

    Cheers,

        Christopher Taylor