Fwd: Different codes (and eudicots)

[Philip Cantino]

Ken Kinman wrote:

>       Correct me if this non-botanist is wrong, but isn't it relatively easy
>to distinguish a monocot from a dicot using a combination of simple
>characters (just one or two characters will usually do, except for some
>paleoherbs).  Isn't there a clear diagnostic reason that two classes have
>been recognized for so many generations?  This isn't just tradition for the
>sake of tradition.  Traditions often have good reasons behind them, not just
>historical precedence.

Yes, it is relatively easy to distinguish monocot from dicots (dicots
being a paraphyletic group comprising all angiosperms that are not
members of the monocot clade), but it is equally easy to distinguish
a composite from a non-composite, a grass from a non-grass, etc.  My
point is that there are several ways that one can subdivide the
angiosperm clade into one highly distinctive subclade and a
paraphyletic set of left-over species.  Which paraphyletic group one
chooses to recognize is a matter of tradition alone, since such
groups have no existence outside the human mind.  If someone can
demonstrate that the monocots-dicot distinction has more phenetic
support than any other way one could divide the angiosperms into two
groups, then I will concede that this this distinction is not merely
a matter of tradition.




>        On the other hand, how easy is it to tell a "eudicot" from other
>dicots?  Other than genetic sequences (and perhaps a certain type of
>pollen?) are there diagnostic characters that distinguish the eudicots?


The only morphological distinction I am aware of is that the pollen
of eudicots is tricolpate or derived from a tricolpate type.
Nonetheless, it is a well supported clade that is frequently
mentioned in the scientific literature; on this basis, I would argue
that it should have a formal scientific name.




>       All the rest of this discussion about PEST and subsidiary ranks is not
>the main issue.  That is one of the reasons I dispense with subsidiary
>(intermediate) ranks and taxa.  You could certainly use additional ranks in
>the Kinman System if you want to, but I find the coding so useful that I
>find all these extra ranks a nuisance that create more confusion than
>they're worth.  Thank you for helping to prove my point.

These issues may not be relevant in distinguishing phylogenetic
nomenclature from the Kinman system, but they are relevant in
distinguishing phylogenetic nomenclature from rank-based nomenclature
and classification as it is practiced by most of the world.



>       Finally, since the devil is always in the details, if eudicots are to
>be given a formal name with a cladistic definition, I would be interested to
>know what kind of phylogenetic definition one would give.  Presumably
>node-based (rather than stem-based), but which two genera (ooooppps, I mean
>clades) would be used to anchor this eudicot clade.  I assume one of them
>would be Ranunculus or a close relative.  Has a cladistic definition of "the
>eudicot clade" been proposed yet?

Let me start by clarifying a few points of confusion in this
paragraph.  First, there are three standard kinds of phylogenetic
definitions, not just two; the one that Ken left out is the
apomorphy-based definition.  Second, it is perfectly acceptable, and
often desirable, to use more than two specifiers (the items cited in
a definition to link a name to a particular clade).  Third, as the
draft PhyloCode currently stands (though this might change),
specifiers are neither genera nor clades; they are species,
specimens, and synapomorphies.

As far as I know, no one has published a formal name or phylogenetic
definition for the eudicot clade.  If I were to do it, I would use
either a node-based or apomorphy-based definition.  A stem-based
definition would be more problematical because the external
relationships of the eudicot clade are poorly resolved.

My understanding is that tricolpate pollen is a synapomorphy of this
clade.  I would look into the literature to confirm this before
publishing an apomorphy-based definition, but for the sake of
argument, let's suppose that this is correct.  If so, one might
define Eudicotyledonae (if that were the name chosen) as the clade
stemming from the first species to possess tricolpate pollen
synapomorphic with that in species X.  Species X could be any one of
thousands of eudicot species that have tricolpate pollen; any of them
would do, so the choice might be quite arbitrary.  On the other hand,
if a clade name based on a genus name were selected, one would want
species X to be the type species of the genus name that forms the
stem of the clade name.  For example, if the eudicot clade were named
Ranunculina, species X would have to be Ranunculus auricomus (the
type species of Ranunculus), assuming it has tricolpate pollen.  This
requirement (PhyloCode Article 11.8) prevents the kind of problem
that Ken pointed out--e.g., the clade Ornithosuchia excluding the
genus Ornithusuchus.  If Ranunculus auricomus does not have
tricolpate pollen, it could not be used as a specifier in this
definition, and the name Ranunculina could therefore not be selected
for this clade.

One could also use a node-based definition.  For example, if the
eudicots were named Ranunculina, the definition might be the least
inclusive clade that contains Ranunculus auricomus, species W,
species X, species Y, and species Z.  Species W, X, Y, and Z here are
representatives of subgroups of the eudicot clade that could
plausibly be sister to the rest of the eudicots.  If the basal split
within a clade that is being named is not well supported, it is
important to hedge one's bets and include enough specifiers to be
sure that the basal group, whatever it turns out to be, is
represented in the definition.  If this is not done, the name may
subsequently end up applying to a less inclusive clade than the
author intended.

As you can see from these examples, phylogenetic nomenclature is not
cut and dried.  The PhyloCode must be used with care, just as the
other codes must be.  Users will have to read it carefully and
understand its subtleties.  This in itself will tend to decrease the
breadth of the "PhyloZone", which should be reassuring to those of
you who are concerned about it driving the rank-based system to
extinction.

Phil



Philip D. Cantino
Professor and Chair
Department of Environmental and Plant Biology
Ohio University
Athens, OH 45701-2979
U.S.A.

Phone: (740) 593-1128; 593-1126
Fax: (740) 593-1130
e-mail: cantino at ohio.edu