When I say one phylum can't give rise to another, I'm talking about pattern
not process. Again, it is a problem of ranks. Lets take the Tridacnidae
(giant clams) and the Cardiidae (cockles). When you see these names, you
picture them as two separate groups (C=Cardiidae; T=Tridacnidae):


T1  T2  T3  C1  C2  C3  C4
 \  /   /    \   \   \  /
  \/   /      \   \   \/
   \  /        \   \  /
    \/          \   \/
     \           \  /
      \           \/
       \          /
        \        /
         \      /
          \    /
           \  /
            \/
            /
           /


But it turns out that giant clams are modified cockles, based on anatomical
and molecular evidence. The diagram below (simplified from work of Jay
Schneider) is the currently accepted pattern:

C1  C2  C3  C4  T1  T2  T3
 \   \   \   \   \   /  /
  \   \   \   \   \ /  /
   \   \   \   \   \  /
    \   \   \   \   \/
     \   \   \   \  /
      \   \   \   \/
       \   \   \  /
        \   \   \/
         \   \  /
          \   \/
           \  /
            \/
            /
           /


This means that the giant clams arose from within the cockles. The rules of
phylogenetic classification do not allow me to call C1 through C4 the
Cardiidae while excluding T1 to T3, because then I have excluded some of
the descendants from the taxon. Instead I must include them all in
Cardiidae. The rules of classification also don't allow a taxon of a
particular rank to contain other taxa of the same rank. I can call T1-T3
Tridacninae (a subfamily), but Tridacnidae is a synonym of Cardiidae.

I could call the whole group Cardiiacea, which allow me to use Tridacnidae
for the giant clams, but that's just playing with ranks. Also, there are so
many levels in the tree of life that there aren't enough ranks to go
around. And whose to say that a class of vertebrates is the same kind of
thing as a class of mollusks. Better to abandon ranks altogether and just
deal with taxa.

Now, returning to macroevolution. Why is that for so many years the giant
clams were considered so distinct that they had their own superfamily,
Tridacnacea? Because they seem so different in morphology. What allowed
that morphology to evolve? A symbiosis with zooxanthellae than provides the
giants clams with extra fuel. The symbiosis is the hypothesized
macroevolutionary event.

>There weren't any Chordates in the Cambrian.

Cathaymyrus is the oldest known chordate at 535 million years old and
Pikaia from the Cambrian Burgess Shale is slightly younger
<http://www-geology.ucdavis.edu/~GEL3/Cathaymyrus.html>; Yunnanozoon from
the lower Cambrian of China
<http://www.lu.org/symposium/cambrian_chordate.html> is also thought to be
an early chordate.

Gary

---------------------------------------------------------------------
Gary Rosenberg, Ph.D.                     [log in to unmask]
Malacology & Invertebrate Paleontology    gopher://erato.acnatsci.org
Academy of Natural Sciences               http://www.acnatsci.org
1900 Benjamin Franklin Parkway            Phone 215-299-1033
Philadelphia, PA 19103-1195 USA           Fax   215-299-1170