RE: Ceratops (was RE: Pterosaur diversity (was: Re: Waimanu))

[Tim Williams <twilliams_alpha@hotmail.com>]

Mickey Mortimer wrote:

> Well, not so fast.  Horn core size isn't necessarily the only character 
> which can be evaluated for Ceratops.

_Ceratops montanus_ is known from a pair of horn cores and an occipital 
condyle.  That's not much to go by.  If anyone can glean some more 
characters other than "big horn cores" from the type material then I haven't 
heard.

> Perhaps other skull roof or occipital characters can distinguish 
> ceratopsids from more basal taxa, or centrosaurines from chasmosaurines.

I think you may have misinterpreted what I meant.  What I was trying to say 
is that large brow horns are not exclusive to chasmosaurines, so using this 
character *alone* to refer a specimen to this group is a bad idea.

> In fact, Ryan's thesis finds that his unpublished taxon is extremely 
> similar with Ceratops and from equivalent beds.  I wouldn't be surprised if 
> they were synonymous, though Ceratops' holotype is not as obviously 
> diagnostic.

Is this the new ceratopsian from Milk River?

If this new taxon is not a chasmosaurine, but is potentially the same as 
_Ceratops_, then we cannot assume that _Ceratops_ is a chasmosaurine.  This 
undermines Sereno's argument that Chasmosaurinae should be renamed 
Ceratopsinae.

> True, but again we have other characters to consider (e.g. neural arch 
> placement).  I don't think anyone doubts Titanosaurus is a titanosaurid (in 
> the sense of being more closely related to Saltasaurus than Andesaurus; not 
> a phylogenetic definition, but you get the clade I mean).  I'd agree with 
> Salgado.  There's nothing in the ICZN saying family names shouldn't be 
> based on nomina dubia.

Yeah, _Titanosaurus_ is probably a titanosaurid under those terms.  However, 
there is no guarantee that this will always remain so.  For example, new 
sauropod material may convince us that procoelous mid-caudals and 
forward-sitting neural arches evolved more than once, and so have a wider 
distribution than we previously thought.

Using just one or two characters to refer a taxon to a group (e.g., 
_Titanosaurus_ to Titanosauridae, _Ceratops_ to Ceratopsidae) reminds me of 
the old "key characters" rationale for classification.  This is the same 
rationale that phylogenetic taxonomy tries to avoid, by using as many 
characters as possible to ascertain relationships.

Salgado subdivides Titanosauridae into two subfamilies: Epachthosaurinae and 
Saltasaurinae.  So, he doesn't even use Titanosaurinae.  That shows you what 
little faith he has in the genus _Titanosaurus_.  By contrast, the advantage 
of using a certain genus as BOTH the name-giver and the internal specifier 
for a clade is that the taxon's inclusion in the clade is guaranteed.  This 
is the advantage that Saltasauridae has over Titanosauridae.

> To play devil's advocate, both Ceratops and Titanosaurus were very 
> distinctive when they were first found.  It's only with hindsight that we 
> know strongly procoelous caudals and large brow horns have wide 
> distributions.

I agree.  It is not the fault of either the _Ceratops_ or _Titanosaurus_ 
type material that their significance was eclipsed by later discoveries.  
But if we want stability in phylogenetic usage, then we should be anchoring 
our clades in the genera that we name them after.  (This only applies to 
family-level clades.)  And if a certain taxon is too poorly-known to be used 
as a specifier, then we should name the clade after a taxon that can be used 
as a specifier.

Cheers

Tim