The Case for Maniraptorid Tyrannosaurs

From: "King, Norm" [nking.ucs@smtp.usi.edu]

Can someone convince me that tyrannosaurs are maniraptorans? I know they might be, but I'm still skeptical.

A sticking point for me is that tyrannosaurs don't have (am I wrong about this?) a semi-lunate carpal--at least their carpals are not shaped that way--and it is only POSSIBLE that one was derived from the semi-lunate condition, and apparently clear that it was not derived from an allosaur-like carpal.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

There is only one really well-preserved tyrannosaurid carpus that I know of, and that is from Albertosaurus(?). It shows five carpals, one of which is evidently a degenerate semi-lunate.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

Does "evidently" mean that one really looks like a degenerate semi-lunate carpal, or is it evidently so because it must be to fit the classification?

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

As I recall, it's in Holtz's 1994 paper. Semi-lunate carpals appear to be a primitive tetanuran or neotetanuran feature. Allosaurs have them, and apparently Afrovenator does, too.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

There is a big flat degenerate distal carpal, which covers the base of mcI & mcII. It could either be a degenerate semi-lunate carpal block, or a degenerate fusion of the distal carpals in the homologous position in other tetanurines (Allosaurus, for example).

From: "King, Norm" [nking.ucs@smtp.usi.edu]

But isn't it also possible that tyrannosaur carpals are derived directly from Compsognathus carpals, which we unfortunately know almost nothing about?

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Why are you suggesting this? Does it have anything to do with the didactyl manus in that genus? (which, incidentally, has a reconstructed phalangial formula of 2,3,3,0,0, more "derived" than any tyrannosaur). Didactyly is pretty easy to develop in parallel.

On other grounds, Compsognathus is *way* more primitive than tyrannosaurs.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

No. Actually, I have seen the claim that the manus of Compsognathus is too poorly preserved (I don't have a reference handy for that one) to say it is unquestionably didactyl, although I realize that Ostrom has looked at it recently and pronounced it so.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Having seen the casts of the specimens, it looks like someone took a shotgun to its hands and blew them up! Gauthier, who has seen the real things, is equally unconvinced of the didactyly. I think it is a distinct possibility, but unresolved.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

I was concerned about reversals of characters reported by Holtz. The Maniraptora developed flexed cervical zygapophyses, but this was reversed at Arctometatarsalia (so, that character considered alone [ok, I know!], we go back to the next step before Maniraptora, which "picks up" Compsognathus). Then, at the next node beyond Arctometatarsalia, but before we get to the tyrannosaurs, three more characters are reversed, including another unambiguous synapomorphy of maniraptorans--posterodorsal margin of ilium curved ventrally in lateral view.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

I'm not all that thrilled with this particular character. Some tyrannos and ornithomimes have dorsally curved ilia, While others show a double curve: ventrally in back; dorsally in front.

Troodonts are almost universally considered quite advanced, yet Sinornithoides, the earliest known troodont, has a dorsally curved ilium!

Archaeopteryx, Sinornis, and Confuciusornis, at least, have dorsally curved ilia, too.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

One of the other reversals at the node beyond Arctometatarsalia is for a synapomorphy at the next node beyond Maniraptora. I didn't "grow up" with cladistics, but rather with numerical taxonomy, back in the late 60's, so maybe I still don't have command of cladistic principles. But I thought we should be suspicious of reversals. Am I wrong on that? If I'm right, why aren't we suspicious of these?

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Suspicious about reversals? I've got to get that "reversals happen" bumper sticker :-)

Incidentally, with elmisaurs and oviraptorids over on the Maniraptora vera side, many of the previous "reversals" disappear.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

IMHO, we ought not shut out the possibility of reversals entirely. Troodonts look manifestly more birdlike than dromaeosaurs to me, yet they have a propubic pelvis. I think the most logical explanation is that they are birds more derived than dromaeosaurs that reevolved a propubic configuration for some reason or another.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

It seems that to make tyrannosaurs and ornithomimosaurs more derived than maniraptors, we have to undo a lot.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

They weren't more derived than maniraptors; they WERE maniraptors.

It all depends on what you look at. In the pelvis and wrist, tyrannos and ornithomimes look fairly primitive; but if one looks at the detailed anatomy of the foot and braincase, they are clearly rather derived birds.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

This MIGHT be a more parsimonious hypothesis, since we don't have to assert that tyrannosaur carpals passed through a semi-lunate state, for which we have no evidence (I say MIGHT, because we know almost nothing about Compsognathus carpals, and maybe an assumption is involved here that makes this hypothesis just as complicated). The same comments would apply to ornithomimosaurs, if they also lack a semi-lunate carpal (that's right, isn't it?).

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Ornithomimosaurs also show a degenerate semi-lunate. I advanced this as a potential synapomorphy (one of several) linking tyrannosaurs and ornithomimosaurs within the Arctometatarsalia, but my idea was not widely accepted.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

The carpal situation in ornithomimosaurs and tyrannosaurids is VERY uncertain. They are simple nubbins of bone. The most found in any specimen is five per hand (in one ornithomimid and in one tyrannosaurid). Given that they are highly reduced, it is equally parsimonious that they could be derived from either carpus form. This can only be resolved by the parsimonious distribution of other characters.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

According to Holtz, tyrannosaurs are grouped with troodonts and ornithomimosaurs due to their arctometatarsalian feet (due to this largely or entirely[?], once it is assumed they are maniraptors).

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Nope. Tyrannosaurs also show a good many more maniraptoran autapomorphies, particularly in the skull.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Within the 'maniraptorans' (soon to be maniraptoriforms, but that's another issue), yes: cursorial features such as limb proportions, arctometatarsi, iliac blades that meet on the dorsal surface, etc. are among the more common features uniting the two. However, there are some cranial features (d-shaped premaxillary teeth, and others listed in the 1994 paper) which suggest an affinity of this group.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

If we undo all of those characters--like assume they never developed in the first place--the arctometatarsalian feet certainly become much more important (ok, I've left out one group of synapomorphies--at the node just before Arctometatarsalia); "entirely" was too strong a word.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

If you run the dataset without those cranial characters, tyrannosaurs fall back to basal coelurosaurs. Same with the arctomet feet. "Entirely" is too strong a word.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

In fact, tyrannos show innovations in foot structure and neural pathways strongly indicating that they are closer to birds than dromaeosaurs are, and maybe even closer than Archaeopteryx!

From: "King, Norm" [nking.ucs@smtp.usi.edu]

You mean Bakker was right???

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Sure looks that way. The feet appear to be a little more birdlike, but it's hard to tell about Archie's feet.

I sure wish someone would CAT-scan an Archie and look at the neural pathways, assuming they would be preserved at all.

In almost all theropods (and almost all tetrapods, as far as I can tell), nerve V1 exits the braincase out the side, along with a bunch of other nerves. In birds, it exits out the front of the braincase, though its own hole. Dromaeosaurs, allosaurs, and most other theropods show the primitive pattern; tyrannosaurs, troodonts, and ornithomimosaurs show the bird pattern. I have no idea whether this feature has ever been checked out in Compsognathus or Archaeopteryx, but if not, it's certainly high time.

This unusual modification, apparently of little functional significance, is what convinced me firmly of the avian status of the arctomets. The idea that the position of the V1 would persist through almost the entire history of the Tetrapoda and then simultaneously and spontaneously shift in several separate Cretaceous theropod lineages seemed pretty untenable.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

I remain unconvinced of such a position [as Bakker's]. I am among the last of the people on this net, it seems, who believes that dromaeosaurids are the sister group to the Archie + later bird clade...

From: "King, Norm" [nking.ucs@smtp.usi.edu]

This can be inferred to indicate they are highly derived maniraptorans, and maybe that makes it easier to believe that they once had a semi-lunate carpal, then lost it (since they're highly derived). Primitive ornithomimosaurs (e.g., Harpymimus) lack the arctometatarsalian condition (isn't that right?).

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Nope. Harpymimus' foot is smeared out of shape: it is uncertain at best. Garudimimus' foot IS arctometatarsalian, despite illustrations otherwise. It was correctly illustrated in Currie & Russell's 1988 Chirostenotes paper, under the name 'Oviraptor'! Thanks to Halska Osmolska (for photos & comments) and Mark Norell (pers. commun. a few years ago) for the above.

Pelecanimimus, sadly, lacks feet (okay, the only specimen lacks feet).

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Garudimimus has been shown to be fully arctometatarsalian, and Harpymimus is apparently very poorly preserved. It is also more primitive in the hand than the much earlier Pelecanimimus, and this, along with some other features, including the enlarged preacetabular blade of the ilium, suggests to me that this genus may not be an ornithomimosaur at all, but possibly a relative of the oviraptorosaurs.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

Wait a minute! Why is this counter example invalid? If it's more primitive in hand, why can't it also be more primitive of foot, and legitimately so? Sounds like Johnny Cochran--if it doesn't fit, it's not legit! ;-)

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

There are questions as to the ornithomimosaur nature of the really scrappy Harpymimus type. I still think it is an ornithomimosaur, myself.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

I didn't mean to imply that; the ornithomimosaur status of Harpymimus just sounds fishy to me. Pelecanimimus had *far* more teeth (more primitive) than Harpymimus, but its thumb metacarpal was nearly as long as the others (apparently more advanced). Harpymimus just doesn't fit easily into ornithomimosaur evolution the way it used to. It still may be an ornithomimosaur, but it would be very much off the evolutionary mainline.

If Harpymimus is, in fact, an ornithomimosaur (quite possible), it probably lost its teeth independently of the rest, and it most likely secondarily rebroadened its feet. I haven't seen most of the bones, and as I said they are apparently poorly preserved, so I can't say anything definite on the subject.

To put it simply, Pelecanimimus seems to consist of better material than Harpymimus, and its ornithomimosaurian affinities seem more secure than those of the latter genus. Therefore I will look to Pelecanimimus for primitive ornithomimosaur characters before I look to Harpymimus.

I just wish we had a Pelecanimimus foot!

Troodonts and ornithomimosaurs seem to be pretty obviously closely related and very birdlike. I'm not quite as sure about a special relationship between these and tyrannosaurs, but tyrannosaurs are certainly at least as birdlike as the other two groups.

All three groups most likely arose from one or two advanced Jurassic or earliest Cretaceous birds. And I'm not talking birds in the vague Olshevskian sense. I mean birds more advanced than Archaeopteryx. The propubic pelvis and arctometatarsalian foot may indicate that they are one group, but the tyrannosaurs would evidently have split off early.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

Now, if troodonts are truly maniraptorans and have arctometatarsalian feet, and advanced ornithomimosaurs have arctometatarsalian feet, but the primitive ornithomimosaurs did not, then why is it not possible that troodonts and ornithomimosaurs developed arctometatarsalian feet convergently (and, perhaps tyrannosaurs, as well)?

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Because

1. primitive ornithomimosaurs *did* have arctometatarsalian feet;
2. ornithomimosaurs, troodonts, and tyrannosaurs also share many other derived characters; and
3. the fine details of the foot in all three groups match up perfectly, bone for bone, curve for curve.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

Since ornithomimosaurs also lack a semi-lunate carpal, and developed arctometatarsalian feet independently from derived (advanced) maniraptorans, perhaps they are not maniraptorans at all.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Certainly a possibility. See Gauthier 1986 for the most cogent argument towards this end.

However, in the most parsimonious distribution of derived characters in analyses by myself, by the American Museum of Nat. Hist. paleontologists, by Hans Sues, and Perez-Moreno et al. 1993, a tyrannosaur-troodont-ornithomimosaur clade was found (including therizinosauroids for the AMNH, troodonts not included in Perez-Moreno 1993).

From: "King, Norm" [nking.ucs@smtp.usi.edu]

According to Holtz (J. Paleo., v.68, no. 5, p. 1107), an unambiguous synapomorphy of maniraptors (sensu Gautier only?) is a long and slender metacarpal III, yet tyrannosaurs have a very short metacarpal III.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Clearly an autapomorphy, wherever they belong!

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

It was as long as metacarpal II and quite slender.

From: Dinogeorge@aol.com

The metacarpal III of Daspletosaurus was much shorter (and much, much smaller) than metacarpal II: see the photo in D. A. Russell 1970 (by the way--if you ever get to see that specimen, tell me [classroom exercise] what's wrong with the [incomplete] manus as depicted).

From: "King, Norm" [nking.ucs@smtp.usi.edu]

Also, in maniraptors metacarpal I is one-third or less the length of metacarpal II, but that is not what is shown in illustrations of tyrannosaur manus ("manuses") that I've seen.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Another potential synapomorphy linking tyrannosaurs and ornithomimosaurs...

From: "King, Norm" [nking.ucs@smtp.usi.edu]

Since tyrannosaurs lack these unambiguous synapomorphies

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Unambiguous with the exclusion of tyrannosaurs, ambiguous with the most parsimonious distribution of characters in the above data sets.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

and also lack the semi-lunate carpal of the Maniraptora sensu Holtz, and we know(?) that some coelurosaurs developed an arctometatarsalian condition independently of others,

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Untrue.

From: "King, Norm" [nking.ucs@smtp.usi.edu]

I am not convinced that tyrannosaurs are maniraptorans. Coelurosaurs, yes--maniraptorans, maybe.

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Fair enough. I don't think they are maniraptorans anymore, anyway (my clade "Maniraptora" is NOT Maniraptora).

From: "King, Norm" [nking.ucs@smtp.usi.edu]

What do I have wrong, or don't know, that is keeping me from accepting the maniraptoran status of T. rex?

From: Thomas_R_HOLTZ@umail.umd.edu (th81)

Hell if I know. Also STOP looking at T. rex, the most derived member of the clade. Alectrosaurus, Stygivenator, Shanshanosaurus, Gorgosaurus: these are the primitive tyrannosaurs to look at.

From: "Nicholas J. Pharris" [pharrinj@PLU.edu]

Actually, T. rex cannot be a maniraptoran as the term was originally defined. It was defined as "all theropods closer to birds than to Ornithomimus." Well, ornithomimosaurs show many attributes which put them closer to birds than dromaeosaurs, which would limit membership in the Maniraptora to the birds alone. Arctometatarsalians may well, in fact, belong to about the same evolutionary grade as the enantiornithians, reducing the membership of the Maniraptora down to the Ornithurae alone.