|
Arghh. College leaves little time for
e-mail. Got this done over four days or so. Enjoy.
Tugulusaurus Dong 1973
T. faciles Dong 1973
Etymology- "Tugulu lizard", Tugulu being
the stratigraphic group it was found in.
Early Cretaceous
Lianmugin Formation of Tugulu Group, Xinjiang,
China
Holotype- (IVPP V 4025) dorsal rib, four mid caudal
vertebrae, metacarpal I, manual phalanx I-1, manual ungual I, manual phalanx
III-?, femora (one proximal) (210 mm), tibia (230 mm), astragalus, calcaneum,
distal metatarsal III, pedal ungual, other fragments.
Diagnosis- The coelurosaurian femur and tibia can
be distinguished from other coelurosaurs based on gross morphology, although the
poor description and figure quality preclude identification of distinct
apomorphies. The manual elements can be distinguished from Allosaurus
(which they are morphologically closest to) by features noted in the
description, but may not be properly referred to Tugulusaurus and cannot be
compared to other carnosaurs. If the elements belong to one taxon, the
combination of carnosaur and coelurosaur traits would be
apomorphic.
Description-
The description of Tugulusaurus is translated
rather poorly and the plates are difficult to make out details in.
There is a dorsal rib figured. Dorsal ribs
aren't well described or figured in the literature, but some comparisons can be
made. It is from the middle of the dorsal column based on the length and
curvature. The capitulum is broken, making it look shorter and deeper than
in other theropods.
The caudal centra are amphicoelous and longer than
wide. The absence of an "edged process" ventrally may mean there was no
ventral keel. The ventral edge is concave laterally and the neural spines
had "disappeared". Whether the latter statement implies they were
originally absent or have broken off, I don't know. There are bases of
transverse processes preserved.
A bone is figured in plate III (5, 6) that appears
to be a first metacarpal. This element is very different from coelurosaur
first metacarpals, but resembles Allosaurus quite closely. It is much more
robust than coelurosaurs and has well-defined distal condyles. The bones
only differences from Allosaurus are: narrower lateral condyle, condyles less
extensive dorsally and proximolateral process facing more laterally than
proximally.
Plate III figure 3 is a manual phalanx I-1.
It is also very similar to Allosaurus, being slightly more slender with a deeper
proximoventral process. It resembles Alxasaurus slightly as well.
The proximoventral process extends proximally past the proximodorsal process,
the shaft is straight and the distal expansion extends further proximally
ventrally.
Also among the figures (and in the text) is a
manual ungual. This ungual is well curved, has a moderate flexor tubercle,
a strongly proximally projecting proximodorsal process and no proximodorsal
lip. It resembles Allosaurus and Torvosaurus more than any
coelurosaur. Among coelurosaurs, it resembles troodontids and the Iren
Debasu therizinosaur the most.
Another manual phalanx is figured. It is from
the third digit based on robustness and size. Comparison to other taxa is
difficult as the exact identity of the phalanx is unknown. It does
resemble Allosaurus in general proportions however.
The femur is very slender and hollow with a
ridge-like fourth trochantor. The femoral head is declined and the medial
condyle projects further distally than the lateral condyle. It is
described as being bowed "to the back". It is distinctly sigmoidal in
anterior view. Besides these features, the femur is described as having a
"crista present on the small twist" (perhaps a crested lesser trochantor?) and
"muscle adherant surface rough".
The tibia is 110% of femoral length. It is
referred to as a radius in the translation, which explains the rather odd
statement that the radius exceeds femoral length in Glut's encyclopedia.
The cnemial crest is "poorly developed" and the medial condyle is more distally
projected than the lateral condyle. The shaft expands distally in
anterior/posterior view and no fibular crest or astragalar fossa can be
observed. Other dubious tibial features include "proximal end ... expands
triangularly at front and back" and "distal end ... inserting into calcaneum
(astragalus?) fovea like a sharply-cut triangle".
The astragalus and calcaneum are not illustrated,
but are stated to be tightly united and the calcaneum is crescentric.
There are said to be "twisted processes on both sides of astragalus" and "no
conspicuous tibial jacent process (ascending process?)."
The text describes two probable distal ends of
metatarsal III that are figured in plate III. I can't recognize them off
hand. Perhaps figures III-8 or 4. They are said to have shiny and
smooth articular surfaces and be concave on both sides. An important
detail is that Dong states Tugulusaurus differs from other ornithomimids in that
the proximal third metatarsal does not constrict. Thus, the
arctometatarsalian condition was probably absent.
A pedal ungual is figured (III-1). It is very
slightly curved, with a concave ventral margin, shallow dorsoventrally, has a
very small flexor tubercle and the proximodorsal corner is placed far proximally
to the proximoventral corner. It is most similar to Neovenator, Sinraptor
and Nedcolbertia. Compsognathus seems to have similar pedal unguals as
well.
Other bones are figured. Some (II-6, 7, 8,
III-7, 9, 12, 13) may be the caudal vertebrae described in the text.
III-14 is rather indeterminable. Oddly, II-9 resembles the first
metacarpal of Archaeornithomimus quite closely. However, II-7, 8 and 9,
and III-9 and 14 are labeled as Theropoda, not Tugulusaurus as the other figures
are. Perhaps more than one taxon is present.
Relationships-
Dong refers this genus to the Coelurosauria based
on hollow long bones and tibia longer than femur, and to the Ornithomimidae
based on the outline and characteristics of the tarsometatarsus and
phalanges. Molnar thought the tibiofemoral ratio was too small for an
ornithomimid. Besides this, the affinities of Tugulusaurus have been
ignored. It is generally placed as Coelurosauria incertae
sedis.
First, let's compare it to ornithomimosaurs.
The caudal vertebrae are plesiomorphic and could be ornithomimosaurian from the
description. The first metacarpal is extremely different from
ornithomimosaurs. It is much shorter (even than Harpymimus) with a more
medially divergent distal articulation. Compare the first metacarpals of
Allosaurus and any ornithomimosaur to see what I mean. Manual phalanx I-1
is not dorsally bowed, unlike advanced ornithomimids and the proximoventral
process extends much further proximally than the proximodorsal process
does. The first manual ungual has a proximodorsal process that extends
proximally much more than the proximoventral process, unlike
ornithomimosaurs. It is also much more curved and deeper than all
ornithomimosaurs except Deinocheirus. The phalanx from manual digit III is
too short to be III-3 in ornithomimosaurs and too long to be III-1 or III-2,
except in Harpymimus. The femur plesiomorphically has a fourth trochantor,
like ornithomimosaurs. The possible lack of an arctometatarsus would be
more basal than ornithomimosaurs except probably Harpymimus and
"Ginnareemimus". The pedal ungual is more curved than ornithomimosaurs,
with a more proximally projecting proximodorsal process. Thus, the only
similarities to ornithomimosaurs (amphicoelous caudal vertebrae, fourth
trochantor present, etc.) are symplesiomorphies.
As was obvious from my description, I felt the
manual elements compared very well to Allosaurus. Also, the pedal ungual
compares most closely to Neovenator and Sinraptor, also carnosaurs.
Another character that may support a non-coelurosaur affinity is the
inconspicuous "tibial jacent" process, possibly indicating a low ascending
process on the astragalus. The only features arguing against this are
the high tibiofemoral ratio and the declined femoral head (along with the small
size and gracility). Although most of these features could simply be
juvenile traits, the declined femoral head is hard to ignore.
It is also quite possible that there are actually
two taxa present, a carnosaur and an ornithomimid. This is supported by
the lack of forelimb elements in the holotype listing and by a cryptic statement
by Dong. "It is likely this claw and this digit belong to the
forelimb. At the present, because the phalanx is too large to be regarded
to be a forelimb, it is difficult to judge". There are no scale bars or
measurement guides for the forelimb material, so this is impossible for me to
judge. The forelimbs would be expected to be larger if they were
carnosaurian however, so that would make sense if there are two animals
present. Another reason more than one theropod may be represented in the
Tugulusaurus material is the fact that two figures (II-9 and III-5, 6) seem to
be first metacarpals of different morphologies. II-9 look almost
identical to Archaeornithomimus while III-5 and 6 are almost identical to
Allosaurus. Finally, carnosaur-like theropods are also present in the
Tugulu group, as is shown by the jaw remains of Kelmayisaurus and a distal pubis
labeled as "Megalosauridae indet." (fig. VII-4) that appears carnosaurian,
perhaps allosaurid.
If the elements belong to one species, Tugulusaurus
is either a very gracile or juvenile carnosaur with an apomorphically declined
femoral head, or a derived coelurosaur with a carnosaur-like manus. I find
it more likely that two taxa are represented. The femur is coelurosaurian
and the tibia probably belongs to the same animal, as might the metatarsal
fragments. The manual elements are carnosaurian. The pedal ungual
resembles members of both groups and the rib and caudal vertebrae could be from
either taxon.
The name Tugulusaurus should apply to whatever
genus the hindlimbs belong to, as they are listed in the holotype, unlike the
forelimb material. This theropod is probably coelurosaur, due to the
declined femoral head, but otherwise hard to place phylogenetically.
Declined femoral heads are known in some ornithomimosaurs (eg
Archaeornithomimus) and eumaniraptorans. The fourth trochantor would
indicate that it is not troodontid, avian or an advanced dromaeosaurid.
The caudal vertebrae eliminate alvarezsaurids and pygostylians if they belong to
the same animal. If the metatarsals go with the femora and tibia, which is
probable, arctometatarsalian taxa are also eliminated. This leaves basal
ornithomimosaurs, Protarchaeopteryx and basal dromaeosaurids as possible
candidates. The listed holotype material is too poorly
described/figured to assign Tugulusaurus to either of these taxa. If
the first metacarpal in II-9 does belong to this taxon however, it would be
almost certainly ornithomimosaurian.
As for the forelimb material, it resembles
Allosaurus more than megalosaurids, based on the less robust first metacarpal
and much longer phalanx I-1. The ungual resembles Allosaurus and
Torvosaurus more than Baryonyx or Afrovenator. Carcharodontosaurid mani
are poorly known and I lack figures of (?)Szechuanosaurus campi and S?
zigongensis. The pedal ungual resembles sinraptorids and
carcharodontosaurids more than Allosaurus, but may be coelurosaurian.
Thus, I would say the material is probably carnosaurian, but unassignable to a
family until more information is available.
Well, that was certainly a surprise. If
anyone wants the plates (two pages), contact me offlist. And if anyone has
other ideas for the identity of figure III-5/6 or II-9, please tell me.
Next is Phaedrolosaurus, described in the same paper. Is it
dromaeosaurid? I guess we'll have to wait and see.
Mickey Mortimer
|