RE: Nomina Dubia Part II: Rapator

[Michael Mortimer <mickey_mortimer111@msn.com>]

Tim Williams wrote-

> Yes, and here we come to the crux of why we have different opinions. For the 
> vast majority of characters, I would tend to treat them as potentially 
> non-diagnostic. These include shape or length of horn cores, length and depth 
> of sulci, size and orientation of metacarpal processes, degree of fusion of 
> metatarsals, and so on.
>
> So for our old friend _Ceratops_, I think we can safely assume (and I concede 
> that it is *just* an assumption) that because of the huge variation in the 
> length and orientation of postorbital horn cores known within *other* 
> ceratopsian species, that _Ceratops_ cannot be diagnosed based on the length 
> and orientation of its postorbital horn cores. I think we can say this even 
> in the absence of a specimen-level statistical analysis of variation in 
> ceratopsian horn cores.
>
> Yes, I agree. It's all very warm-and-fuzzy, and so often in the eye of the 
> beholder. But I think at a qualitative level that we should diagnose 
> characters that we suspect may be prone to intraspecific (including 
> ontogenetic) variation. I know there is no standard operating procedure for 
> this; your _Australovenator_ is a perfect example of this.
>
> That's an excellent question. If push comes to shove: you can't. Not for most 
> characters anyway. I suspect that a taxon based on a partial skeleton gets 
> treated far more leniently than a taxon based on just one or a few elements. 
> But given we have multiple elements for _Australovenator_, the sum total of 
> 'diagnostic' characters at least offers the chance of a unique combination of 
> characters.
>
> It's perhaps unfair that _Rapator_ is considered a nomen dubium when we 
> haven't actually tested the validity of its metacarpal characters. Instead 
> we've just *assumed* that its characters are not reliable for diagnosing a 
> genus.
>
> But let's say we did erect a new name for *every* fragmentary specimen that 
> has one or two characters that are possibly diagnostic. The result would be a 
> proliferation of new genera and species. Such a policy would lead to an 
> explosion of highly suspect taxa, and spawn a thousand arguments over whether 
> this-or-that genus had priority. Your question regarding whether 
> _Australovenator_ should be referred to _Rapator_ (based on shared characters 
> mentioned by Agnolin &c) would be repeated dozens of times over. This sounds 
> like a
> without us throwing more dodgy names into the mix.
>
> Your approach of giving any and all characters the benefit of the doubt would 
> justify naming taxa such as _Serendipaceratops_, based on a single ulna that 
> was said to have unique proportions. I would say that _Serendipaceratops_ 
> should probably not have been named, because the fact that the ulna is shaped 
> a little differently to all other known ornithischian ulnae does not stack up 
> as a 'good' diagnostic character.
>
> I think the current approach, whereby names like _Rapator_ and _Ceratops_ and 
> _Serendipaceratops_ are regarded as nomina dubia even though nobody has 
> actually *tested* whether or not their characters are subject to 
> intraspecific variation, is an attempt at damage control. While it might seem 
> unfair, from a taxonomic perspective these names tend to cause more trouble 
> than what their worth.

To me, your approach seems more like following the status quo than actually 
doing science.  You give characters the benefit of the doubt too, but only when 
they're listed as being diagnostic by their authors.  If you've ever gone 
through the diagnosis of any taxon, you'd understand why I don't value 
characters listed in diagnoses more than others.
 
As an example I've studied more than Australovenator, take Struthiomimus.  
You'd think it was obviously diagnostic.  Known from complete specimens and 
such.  However...
 
Osborn (1917) erected Struthiomimus for Ornithomimus altus because it possessed 
metatarsal V, which he incorrectly thought was absent from Ornithomimus velox. 
Later authors often realized Osborn's error and synonymized the genera, though 
Parks (1926, 1928, 1933) did name four additional species in Struthiomimus 
because they possessed the metatarsal (S. ingens, S. currelli, S. brevetertius 
and S. samueli). Russell (1972) revised ornithomimid taxonomy and was the first 
author to use real morphological differences to validate the separation of 
Struthiomimus from Ornithomimus edmontonicus (and his new genus 
Dromiceiomimus), though several characters he cites are now seen as invalid due 
to the recent synonymization of Dromiceiomimus with Ornithomimus edmontonicus. 
Notably, Dromiceiomimus samueli also has a humerus shorter than its scapula, 
antebrachium length overlaps Dromiceiomimus', and preacetabular, tibial, 
metatarsal and pedal digit length are no longer distinct. Furthermore, all the 
characters proposed by Russell to be apomorphies of Struthiomimus are 
symplesiomorphies when viewed in a cladistic context. The robust forelimb is 
plesiomorphic, being seen in all ornithomimosaurs except Dromiceiomimus, 
Gallimimus bullatus and Sinornithomimus. The curved manual unguals are also 
plesiomorphic, present in all ornithomimosaurs except Dromiceiomimus, 
Anserimimus and "Gallimimus" "mongoliensis". Dromiceiomimus is the only 
ornithomimosaur known with a presacral column not longer than its hindlimb, as 
opposed to Struthiomimus, Anserimimus, Gallimimus and Sinornithomimus. The 
proximal caudal centra are also posteriorly wide (over half their length) in 
Garudimimus, "Grusimimus", Gallimimus, Ornithomimus sp. nov. and O? sedens. The 
transition point is also posterior to the fourteenth caudal in Harpymimus, 
Anserimimus and Gallimimus. Metacarpal I is shorter than metacarpal II in all 
ornithomimosaurs except Anserimimus, Dromiceiomimus and Ornithomimus. The 
elongate manual ungual I (longer than ungual II) may be primitiv!
e for orn
homimids, also being present in "Grusimimus", Sinornithomimus and Anserimimus. 
The elongate manual ungual III (longer than phalanx III-3) is here found to be 
synapomorphic of a larger clade also containing Gallimimus bullatus, "G." 
"mongoliensis" and Ornithomimus? sedens. Makovicky et al. (2004) proposed an 
elongate manus (>7% longer than humerus) as an additional apomorphy of 
Struthiomimus, and this seems to be true as it is otherwise present only in the 
somewhat distantly related Ornithomimus? sedens, Harpymimus and Pelecanimimus. 
Kobayashi et al. (2006) and Longrich (2008) both proposed a small skull (<50% 
of femoral length) is unique to Struthiomimus, but this is also found here to 
be characteristic of the larger clade also containing Gallimimus bullatus, "G." 
"mongoliensis" and Ornithomimus? sedens. Longrich also proposed many characters 
to distinguish Struthiomimus from Dromiceiomimus (Ornithomimus in his use) and 
his new large Dinosaur Park ornithomimid, but most were not examined in a 
broader context. The convex ventral maxillary edge is also seen in Gallimimus 
bullatus, Sinornithomimus and probably "Gallimimus" "mongoliensis", for 
instance. Gallimimus bullatus shares the short postorbital portion of the 
frontal and dorsally flat distal caudal centra (in posterior view). The latter 
may be plesiomorphic however, as it is also found in Patagonykus and 
Aniksosaurus. Broad pedal unguals also seem plesiomorphic, being present in 
Anserimimus, Gallimimus bullatus and Garudimimus. Finally, the elongate 
proximodorsal process on its pedal unguals is primitive, found in 
Sinornithomimus, "Grusimimus", Garudimimus and Harpymimus as well.
 
So of the characters listed by Osborn (1917) and Russell (1972), none withstand 
scrutiny.  Nor did over half of Longrich's (2008).  Which ones are we left with 
as valid?
 
Diagnosis- (after Longrich, 2008) frontal with an orbital rim that is 
completely convex in dorsal view; frontals abruptly expanding posteriorly, with 
the anterolateral edge angled 40 degrees to the midline; strongly flattened 
dorsal edge of distal caudal vertebrae; ventrolateral edges of pedal unguals 
rounded (unknown in most taxa).
(after Makovicky et al., 2004) manus 8% longer than humerus (also found in 
Ornithomimus? sedens).
 
Gee, those seem like exactly the kinds of characters you tend to dismiss.  
Minor proportions and angles, convexities of surfaces that aren't known or 
examined in most taxa.  So unless I want to blindly follow professional 
authority and say "Struthiomimus is valid because everybody has thought so, 
it's complete and has had long diagnoses; Rapator is a nomen dubium because the 
most recent reviews say so, it's known from one bone and has no explicit 
diagnosis published", I have to value these as much as any other character 
until they've been shown to be invalid.
 
You admit that you have no objective reason to value Australovenator's supposed 
diagnostic characters over Rapator's, you just 'feel' some characters are 
better than others and figure Australovenator has a better chance at validity 
due to its relative completeness.  Then you defend your approach with a 
pragmatic reason rather than a scientific one.  I agree throwing out 
fragmentary specimens would make my job easier, but I don't study these issues 
because they're easy, I study them because I actually care about the identity 
and relationships of the animals.  If it takes hard work to find out which name 
to call them, then so be it.  It makes it more interesting and fun.
 
Mickey Mortimer                                           
_________________________________________________________________
Hotmail has tools for the New Busy. Search, chat and e-mail from your inbox.
http://www.windowslive.com/campaign/thenewbusy?ocid=PID28326::T:WLMTAGL:ON:WL:en-US:WM_HMP:042010_1