TYRANNOSAURIAN IMPLOSION [long; part 1 of 2]

[Dinogeorge@aol.com]

Sorry for lack of italics in this mini-treatise. If you read this, please 
mentally italicize the generic and species names, and also such Latin terms 
as nomen dubium, etc.
--------------------------------------------------------------

Over the past few weeks Tracy Ford and I have taken a cold, hard look at the 
tyrannosaurians in the light of recent opinions (e.g., Carr, 1999) that the 
smaller "species" are based on juvenile and subadult specimens of the larger 
"species." After rereading some of the important papers and studying some of 
the specimens available to us (for example, Tracy photographed the 
Stygivenator molnari type specimen in Los Angeles from several viewpoints), 
we've concluded that this is indeed very likely the case. Some interesting 
taxonomic implications follow, the least of which is the sinking of my three 
1995 tyrannosaurian genera. The following is a summary of our conclusions. 
I'll continue to run the descriptions of those three genera in the back of 
future printings of Mesozoic Meanderings #3 for a while, just for the record, 
but this note (in edited form) will become part of a revised Appendix 
starting with the second printing; and the tyrannosaurid genera and species 
in the North American and Asiatic dinosaurs lists will be reorganized to 
conform with this note.

We're now pretty sure that Stygivenator molnari is based on the front end of 
a juvenile Tyrannosaurus rex snout, somewhat distorted in preparation; that 
Nanotyrannus lancensis is based on a slightly more juvenile Tyrannosaurus rex 
skull; and that Dinotyrannus megagracilis is based on the partial skeleton of 
an 80%-grown subadult Tyrannosaurus rex. There is considerable individual 
variation among all the more than two dozen fully adult specimens presently 
referred to Tyrannosaurus rex (as shown, for example, by Carpenter, 1990), 
and if all the minor distinctions among these specimens were taken into 
account, one might well parition them into a dozen or more "species." But the 
existence of a dozen very similar sympatric species of large Lance/Hell Creek 
tyrannosaurids during the last couple of million years of the Maastrichtian 
seems much less likely than the existence of a single species, perhaps 
exhibiting microevolutionary or regional variability. (A similar situation 
may occur with the coeval genus Triceratops, and a similar range of 
individual variation has been noted in the theropod genus Coelophysis 
[Colbert, 1989].)

Characters that seem stable for all Tyrannosaurus rex specimens include (1) 
"jugal spread": the strong, pronounced, and sometimes abrupt increase in the 
width of the skull across the jugals, evident even in Nanotyrannus lancensis 
and, to a lesser extent, in Dinotyrannus megagracilis; (2) "hidden condyle": 
the occipital condyle is blocked from view by the overhanging supraoccipital 
in dorsal aspect when the skull is held with nasals horizontal, which compels 
the skull to incline forward and downward in relaxed position; (3) 
"hornless": lacrimal horn (or "cornual process of the lacrimal") absent, also 
evident despite plaster in Nanotyrannus and in Dinotyrannus, although in the 
latter the relevant skull material is fragmentary; (4) 13â14 (almost always 
13) dentary teeth, most very large rostrally, becoming smaller caudally, with 
some crowns about as tall as the dentary itself. Note that in Tyrannosaurus 
rex the dentary is much larger than in the earlier North American species, 
yet has fewer teeth, which leads to very much larger dentary teeth in this 
species (and maxillary teeth to match). (The dentary teeth cannot be seen in 
the Nanotyrannus lancensis type skull, but preparation may eventually open 
the jaws and establish a dentary count for this specimen.) These characters 
appear to be unique to this genus and species. (5) Adult Tyrannosaurus rex 
skulls show a distinct, pendant suborbital process that intrudes into the 
orbit below the eyeball, giving the orbit a distorted keyhole shape ("keyhole 
orbit"). This character also occurs in the Asiatic genus Tarbosaurus, 
although the process is not pendant. (And it also occurs convergently in the 
South American abelisaurid Carnotaurus, so its taxonomic value may be 
limited.)

Characters 1â3 form a suite that seems related to the acquisition of 
overlapping visual fields for the eyes for stereoscopy, since the eyes point 
forward (and above the muzzle) more in Tyrannosaurus rex than in any other 
tyrannosaurid species. Further examination of known Tyrannosaurus rex 
specimens will doubtless reveal more characters. But many other characters, 
particularly those concerning sizes and shapes of cranial foramina and 
fenestrae, cranial pneumaticity, and locations of cranial sinuses, etc., are 
either not yet well studied and mapped or seem too variable (ontogenetically, 
individually, or even pathologically) for reliable taxonomy. It does not 
appear that the so-called "robust" and "gracile" morphs are anything other 
than age-related: the "robust" individuals are simply older individuals than 
the "gracile" individuals, a hypothesis that could be tested by looking for 
age-related features (growth rings, e.g.) in bones. So far, we find only one 
valid species in the genus Tyrannosaurus, namely, Tyrannosaurus rex. All 
Asiatic species previously referred to Tyrannosaurus lack the aforementioned 
features and are thus referable to other tyrannosaurid genera.

With Stygivenator sunk as a Tyrannosaurus juvenile, it becomes almost 
inescapable that Aublysodon-like teeth, that is, premaxillary teeth with 
D-shaped cross-section, lingual ridge, and unserrated carinae, are simply the 
teeth of juvenile to subadult individuals of whatever species of 
tyrannosaurid happens to occur in the same formation or at the same 
stratigraphic level. (The Stygivenator snout includes an associated 
Aublysodon-like unserrated premaxillary tooth.) If so, then the presence or 
absence of a lingual ridge or denticles on such teeth would have little 
taxonomic significance within Tyrannosauridae. It is peculiar that although 
Aublysodon-like teeth are fairly common (many specimens are known: see Molnar 
& Carpenter, 1989), and are found in practically all horizons where 
tyrannosaurid body fossils are found, not one theropod skull or skeleton is 
known from North America with such a tooth actually in place in a premaxilla. 
If, however, such teeth belonged predominantly to juvenile tyrannosaurids and 
showed up only rarely in large individuals, then the general scarcity of 
juvenile tyrannosaurid material would account for the particular scarcity of 
material with Aublysodon-like teeth in place.

It is easier to tell where in the jaw a tyrannosaurid tooth comes from than 
it is to tell what species it belongs to. Their relative thickness readily 
distinguishes tyrannosaurid teeth from those of other large theropods, but 
isolated teeth cannot yet be reliably identified below family level. Thus, 
Albertosaurus-size tyrannosaurid teeth in a formation from which only 
Tyrannosaurus is known should conservatively be considered teeth of subadult 
Tyrannosaurus and not evidence of an Albertosaurus-size tyrannosaurid in the 
formation. Tooth replacement was continuous in tyrannosaurids, and as an 
animal grew and its skull elements became increasingly robust, its 
replacement teeth kept pace, becoming larger and relatively thicker with age. 
Juvenile tyrannosaurids, obviously, had smaller and narrower teeth than 
adults of the same species, although tyrannosaurid maxillary and dentary 
teeth in general are "incrassate" (relatively thick) and appear "oversized" 
relative to their jawbones. So a single animal could generate a whole range 
of sizes and shapes of shed teeth over its lifetime. Maxillary and dentary 
tooth counts may decrease by one or two with increasing age in tyrannosaurids 
(Carr, 1999), which is unusual in dinosaurs (the trend in herbivores is 
opposite). Denticle size and shape may eventually prove taxonomically 
significant (Currie, Rigby & Sloan, 1990), but the relevant studies are not 
yet completed.

Since Aublysodon-like teeth likely belong to juvenile and/or subadult 
tyrannosaurids, this must demolish any concept of a distinct family 
Aublysodontidae or subfamily Aublysodontinae for tyrannosaurians with 
unserrated premaxillary teeth. Aublysodontidae (Nopcsa, 1928) and 
Aublysodontinae become junior subjective synonyms of Tyrannosauridae and 
Tyrannosaurinae, respectively. Tyrannosaurian genera referred to these 
family-level taxa because they are known with unserrated premaxillary teeth 
(e.g., Alectrosaurus, Shanshanosaurus, Oklahoma Museum "Aublysodon") are now 
best referred to Tyrannosauridae or, perhaps, to some more basal family-level 
taxon within Tyrannosauria. The genus Aublysodon itself, being based on a 
single premaxillary tooth from the Campanian Judith River Formation of 
Montana, is almost certainly a juvenile Albertosaurus, species indeterminate. 
Thus, ironically (given how the genus Aublysodon has fared since it was 
coined), Leidy (1865, 1868) was probably correct to have originally referred 
the Aublysodon type tooth to the same tyrannosaurid genus and species as the 
other teeth he described from the Judith River, namely, Deinodon horridus.

And so, with only one genus and species of tyrannosaurid securely known from 
the late Maastrichtian of North America, what of the earlier forms? Dale 
Russell (1970) listed a number of features that purportedly distinguish 
Daspletosaurus from Albertosaurus, but I frankly don't see them as having 
generic significance. At best they define Daspletosaurus as a robust, rather 
long-forelimbed species of Albertosaurus distinct from Albertosaurus 
sarcophagus. One apomorphy shared by all known pre-Lance/Hell Creek North 
American tyrannosaurids is the presence of a lacrimal horn. In 
Daspletosaurus, the horn was positioned above the vertical ramus of the 
lacrimal, whereas in Albertosaurus the horn lay forward of the vertical 
ramus. Is this really a generic distinction? Were there no individuals in 
which the lacrimal horn occupied an intermediate position? Russell listed a 
few other minor features to distinguish Daspletosaurus from Albertosaurus, 
but they seem no more important than the features that distinguish 
Albertosaurus sarcophagus from Gorgosaurus libratus; and I think what we have 
here, as with Tyrannosaurus, is just one genus, though including three 
species that may correlate stratigraphically. We all know that generic 
distinction is entirely subjective, and that one person's genus may be 
another person's tribe or subtribe. This said, I think that generic 
distinctions among skeletal fossils should be apparent at a glance, 
immediately, not via detailed analysis of suites of minor features; the 
latter kinds of distinctions are more appropriate to distinguishing species 
within a genus.

Several features are shared by Daspletosaurus torosus, Albertosaurus 
sarcophagus, and Gorgosaurus libratus that contrast with the Tyrannosaurus 
rex features noted above. Some are surely primitive to the family 
Tyrannosauridae: (1) orbit without suborbital bar ("round orbit") even in 
adult animals; (2) skull not abruptly widened across the jugals ("no jugal 
spread"); (3) occipital condyle visible in dorsal view of skull ("visible 
condyle"), because the skull was carried more horizontally than in 
Tyrannosaurus; (4) lacrimal horn present ("horned"); (5) dentary tooth count 
14â17. There is a clear generic distinction between this group of species and 
Tyrannosaurus, but no really clear generic distinctions among the three 
species themselvesâjust a bunch of what are essentially minor proportional 
and topological differences (see Carr, 1999 for a lengthy list), the kinds of 
distinctions one might observe between species rather than genera. I would 
sink both Daspletosaurus and Gorgosaurus into Albertosaurus, but retain their 
three species as Albertosaurus sarcophagus, Albertosaurus libratus, and 
Albertosaurus torosus. Considering that Albertosaurus torosus and 
Albertosaurus libratus occur in the same formation, it is possible that one 
is a sexual dimorph of the other or that their type specimens represent 
extremes within a continuum of individual variation of a single species. But 
in view of Russell's list of distinctions between the two forms, it is at 
least equally likely that there were indeed two slightly different sympatric 
species that preyed upon two different kinds of prey (ceratopians and 
hadrosaurus, for example).

The as-yet-unnamed Two Medicine tyrannosaurid, which has been described 
loosely as a transitional form between the genera Albertosaurus (or 
Daspletosaurus) and Tyrannosaurus, may indeed represent a distinct 
tyrannosaurid genus. But since it seems to retain a low lacrimal horn, 
perhaps it should simply be described as a fourth species of Albertosaurus. I 
haven't seen the specimens, only a sketch of the skull, and I cannot comment 
further. Other North American tyrannosaurid species whose descriptions are 
presently in preparation may prove to belong to the genus Albertosaurus, too.

The oldest available name for the genus Albertosaurus is Leidy's Deinodon, 
which, since I would recognize only one genus of Campanian tyrannosaurid from 
western North America, could be resurrected as its senior subjective synonym. 
But in the absence of good tyrannosaurid skeletal topotypes from the Judith 
River Formation of Montana, and in the interest of nomenclatural stability, 
Deinodon and its type species Deinodon horridus (which, as implied above, can 
once again be regarded as a senior subjective synonym of Aublysodon mirandus) 
should remain isolated as nomina oblita (forgotten names). We cannot tell 
from the type teeth whether Deinodon horridus is distinct from or the same 
taxon as either of the coeval species Albertosaurus torosus or Albertosaurus 
libratus, which makes Deinodon horridus a nomen dubium. The Deinodon 
situation parallels the Iguanodon situation in Great Britain, except that the 
name Iguanodon is now no longer based on a few isolated teeth but on one of 
the excellent Bernissart skeletons (ICZN Opinion #1947). The genus Deinodon 
provides a convenient wastebasket for disposing of the several Cope Laelaps 
tyrannosaurid tooth species, too.

The oldest available generic name for Tyrannosaurus is Cope's (1892) 
Manospondylus, whose type species Manospondylus gigas is based on a worn 
Tyrannosaurus rex cervical centrum (originally two but one is lost). Since 
only one large theropod species is known from Lance-age western North 
America, Manospondylus gigas is no longer a nomen dubium, and Cope should be 
credited with discovering and describing as a distinct taxon the first 
Tyrannosaurus rex specimen. But again in the interest of nomenclatural 
stability, the names Manospondylus and Manospondylus gigas should remain 
nomina oblita. (Imagine what would happen if kids everywhere woke up one day 
to find that their favorite dinosaur, Tyrannosaurus rex, had been renamed 
Manospondylus gigas!)