Playing with Wilson and Sereno

[Adam Yates <A.Yates@zoo.latrobe.edu.au>]

I'm fascinated with sauropods, so it was with glee that I obtained my copy
of the recent Wilson and Sereno monograph. Let me say at the outset that
despite the fact that I disagree with some details of the phylogeny, this is
a magnificent piece of work, what a cladistic analysis should be like. It is
precisely because it so clearly set out, the characters are well described
and the codings for each of the taxa are provided in a datamatrix, that
specific areas of disagreement can be located and tested. 
I guess that the biggest problem with the analysis is the absence of certain
taxa (notably Mamenchisaurus and Patagosaurus). These may not be quite as
completely known as some of the other taxa but are complete enough so that
they don't come out just anywhere on the cladogram. The other problem is
'hiding' what I consider to be significant homoplasy by subsuming many lower
level taxa into single large terminal taxa (Diplodocoidea, Titanosauria and
to a lesser extent Brachiosauridae). This effect is most notable with the
inclusion of Rayososaurus/Rebbachisaurus/Limaysaurus tessonei - ;) within
the Diplodocoidea. This strongly affects the postion of Haplocanthosaurus.
Rayososaurus exhibits two of the three apomorphies that are used to diagnose
the Macronaria, and one that diagnoses the Camarasauromorpha (which is the
valid name for the Camarasaurus + Titanosauriformes clade, even if it wasn't
used by Wilson and Sereno).
These are proximally open anterior chevrons, coplanar ischial blades and a
deep quadrate fossa. Indeed if analysis is rerun when the Diplodocoidea is
boken into Rayososaurus, Dicraeosauridae and Diplodocidae (and three new
characters supporting the monophyly of the Diplodocoidea are included) then
we find that the characters used to support the inclusion of
Haplocanthosaurus within the Macronaria have a more ambiguous distribution
and Hap now forms part of an unresolved trichotomy including Hap.,
Diplodocoidea and Camarasauromorpha. I would bet that if you then went on to
include the admittedly weak characters of Calvo and Salgado, you find that
Haplocanthosaurus is indeed a basal diplodocoid.
One interesting feature of Patagosaurus is that it exhibits both of the so
called autapomorphies of Barapasaurus (slit shaped neural canal in the
posterior dorsal vertebrae and fenestrae in the neural arches opening into
the neural canal below the transverse processes). If these characters are
added and Patagosaurus is included, the two genera do indeed form a clade
(the Barapasauridae?)which is interesting biogeographically (an early
exclusively gondwanan radiation of eusauropods?) Perhaps Rhoetosaurus should
be examined in the light of this.
There is more I would like to say but this is already an overblown post. I
would just like to reiterate that it is only because Wilson and Sereno set
out so clearly what they had done that specific claims (eg. the
relationships of Haplocanthosaurus) can be tested in other words it is
cladistics as it should be.

cheers

Adam Yates