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

Re: Phyl tax misunderstandings (was Re: tetrapoda) (LONG)

Thomas R. Holtz, Jr. wrote:
> At 06:37 AM 11/14/97 -0800, Johnathon Woolf wrote:
> >OK, time for another of those stupid questions for which I am probably
> >becoming infamous: what is the point of the arguments over different
> >cladogram-based definitions of taxa?
> Actually, what us systemicists are trying to do is to establish explicity
> definitions for taxon names so that there WON'T be any arguments over them.
> However, these are the formative years, which (we all no) is the time when
> most of the tantrums take place.

More power to you, then. <g>  I'm all for clarity instead of confusion.

> >As I understand it, "crown-based"
> >definitions say "all animals closer to modern X than to modern Y."
> Actually, no.  "Crown-based" taxon definitions are the a special type of
> node-based definitions, where the 'anchor' taxa are currently living: as in
> Mammalia as all descendants of the most recent common ancestor of
> Monotremata and Theria.
> (Note that some people have misapplied the term "crown-based" and refered
> to, for example, Hadrosauridae as a "crown group" within Iguanodontia...).
> >"Node-based" definitions say "the last common ancestor of groups X and
> >Y, and all descendants of that ancestor."
> Yes.
> >"Stem-based" definitions say
> >"X and all organisms closer to X than to Y," where X and Y may be living
> >or extinct species.
> Yes.

Two out of three?  Not bad.  Maybe I understand this stuff better than I

> >All depend on cladograms, which means the
> >definition is only as good as the cladogram is.
> Nope.  Time to do the mind-shift.  Accepting, for the moment, the reality of
> evolution, there will ALWAYS be some clade which can be described by these
> different definitions.  

Always?  I suppose so, but you have to guard against two
always-significant possibilities.  One is that the anchor species are
more closely related than you think they are, and the definition turns
out to be much too limited.  The other is that the two anchor species
are more distantly related than you think, and the definition turns out
to be grossly paraphyletic or even polyphyletic.  For example, suppose
you define Mammalia as the last common ancestor of the platypus and
{insert placental-of-choice here} and all descendants of that ancestor. 
Well, current evidence rather weakly points to the mid-Cretaceous as the
time that Monotremata branched off from the main mammalian line.  So, by
that definition nothing older than mid-Cretaceous can be called a
mammal.  No matter the fact that many Jurassic and early Cretaceous
animals are more mammal-like than the platypus is.  

OTOH, suppose you define Mammalia that way and then find that the
monotremes actually descend form a line that branched off from cynodonts
in the Late Triassic, several million years before even _Morganucodon_
lived.  Do you then include cynodonts in the Mammalia, despite the fact
that they had few or none of the apomorphies that are used to
distinguish either modern or fossil mammals?

> >All also depend totally on having a good fossil record to work with.
> Definitely not true!!  Most invertebrate workers can produce useful
> phylogenetic taxonomies for groups with little or no fossil record.
> >For example, we can't build a proper crown- or node-based definition of
> >mammals because nobody knows when monotremes branched off from the basal
> >mammal line, so we don't know the last common ancestor of modern
> >mammals.
> Sure we can, and it has been done!  

No, it hasn't, because we can't.  The most recent work on mammal
classification that I've read, Hopson's 1994 paper in MAJOR FEATURES OF
VERTEBRATE EVOLUTION, gives _three_ possible definitions of Mammalia. 
One is a crown-based definition.  One is a node-based definition.  One
is based on morphology.  Guess which one he uses?  
> >I don't understand why people got away from the plain old-fashioned idea
> >of basing definitions on physiology:  if X and Y have features 1, 2, and
> >3 in common, then X and Y are in the same taxon.  And taxa are defined
> >as "the first organism to have features 1, 2, and 3, and all of its
> >descendants."
> Okay, actually there ARE such definitions (the "apomorphy-based" taxon
> definitions), but they have a failing.  If you use multiple characters to
> define a taxon, what happens when you get to the origins of that clade and
> some (but not all) of the features are present.
> Real World Example: Dinosauria was once defined by upright limbs,
> asymmetrical hands, three or more sacral vertebrae, and an open acetabulum.
> This worked fine when Pisanosaurus, Thecodontosaurus, and Coelophysis were
> among the most primitive dinosaurs known.  Now we have Lagerpeton,
> Lagosuchus, etc. (which have upright limbs), Eoraptor (with upright limbs,
> asymmetrical hand, three sacrals, but uncertain acetabulum), and
> Herrerasaurus (upright limbs, asymmetrical hands, only two sacrals, and an
> open acetabulum).  Which ones are dinosaurs?

Simple (IMO): if it doesn't have _all_ the features that have been used
to define Dinosauria, it ain't a dinosaur.  Two or three out of four
isn't enough.  

> >Tetrapoda then would be the first four-legged partially
> >land-dwelling animal and all of its descendants.
> Ichthyostega & company's legs are a lot more fin-like than previously
> believed.  Panderichthys fins are a lot more leg-like than previously
> believed.  Ichthyostega & company were a LOT more aquatic than once thought.
> How "leggy" does the limb have to be for such a definition?  How partial
> does the "land-dwelling" have to be?

Ichthyostega wouldn't be a tetrapod under this scheme.  "Leggy" means
"capable of supporting the animal's weight out of water" -- which
_Ichthyostega's_ weren't.  "Partial" means "it could spend more than a
few minutes at a time there" -- which _Ichthyostega_ apparently couldn't
do, since it had no lungs.

> >Amniota would be the
> >first animal that laid hard-shelled eggs, and all of its descendants.
> Assigning eggs to fossil organisms is EXTRAORDINARILY difficult.  In fact,
> dinosaurs are one of the few groups for which eggs are well studied (others
> are some lizards and turtles).  There are Permian eggs known, but nobody
> knows who laid them.  How could such a definition be used to determine if
> Dimetrodon, Diadectes, or Limnoscelis is or isn't an amniote?

My understanding is that whether we have actual fossil eggs or not,
there are definite differences in the skeleton between an animal that
lays soft eggs like an amphibian does and an animal that lays
hard-shelled eggs. 

> >Aves would be the first animal to have flight
> >feathers, and all of its descendants.
> But feathers are only preserved in a VERY limited environmental setting.
> (We are damned lucky Archaeopteryx lived near lagoons!!).

OTOH, archy is identified as the first bird ONLY because it has those
feathers.  There is not one other feature that definitively places archy
within the birds.

> >Ceratopsia would be the first
> >animal to have the parrot beak and neck frill, and all of its
> >descendants.
> "Neoceratopsia", actually,  although a simple apomorphy-based definition for
> Ceratopsia would just be "the first taxon to possess the rostral bone and
> all of its descendants".

> I hope this helps.

It does, thanks.

-- JSW